HomeMy WebLinkAboutAPA1784r
-
,--------------------------------------
REPORT NO. 2
RESIDENT AND JUVENILE ANADROMOuS FISH
INVESTIGATIONS (MAY -OCTOBER 1983)
ALASKA DEPARTMENT OF FISH AND GAME
SUS ITNA HYDRO AQUATIC STUDIES REPORT SERIES
-
-
Document No. 1784
Susitna File No. 4.3.1.6
ALASKA DEPARTMENT OF FISH AND GAME
SUSITNA HYDRO AQUATIC STUDIES
REPORT NO. 2
RESIDENT AND JUVENILE ANADROMOuS FISH
INVESTIGATIONS (MAY -OCTOBER 1983)
Edited by:
Dana C. Schmidt
Stephen S. Hale
Drew L. Crawford
Paul M. Suchanek
Prepared for:
Alaska Power Authority
334 W. Fifth Avenue, Second Floor
Anchorage, Alaska 99501
July 1984 ARLIS
Tf\
l'-f2t;
.$9
,+,g
no.l!{f3lf
Alaska Resources
Library & Information Services
lUlchorage,AJaska
NmiCB
ANY QUESTIONS OR COMMENTS CONCERNING
THIS RBPOR~ SHOULD BE DIRECTED '1'0
THE ALASKA POWER AUTHORITY
SUSITRA PROJECT OPPICB
--------------~----------------------------------------~-------~---------------------
-
-
PREFACE
This report is one of a series of reports prepared for the Alaska Power
Authority (APA) by the Alaska Department of Fish and Game (ADF&G) to
provide information to be used in evaluating the feasibility of the
proposed Susitna Hydroelectric Project. The ADF&G Susitna Hydro Aquatic
Studies program was initiated in November 1980. Beginning with the
reports for the 1983 open water season, all reports will be sequentially
numbered as part of the Alaska Department of Fish and Game Susitna Hydro
Aquatic Studies Report Series.
Report
Number
1
2
3
4
TITLES IN THIS SERIES
Title
Adult Anadromous Fish Investigations:
May -October 1983
Resident and Juvenile Anadromous Fish
Investigations: May -October 1983
Aquatic Habitat and Instream Flow
Investigations: May -October 1983
Access and Transmission Corridor Aquatic
Investigations: ~1ay -October 1983
Publication
Date
April 1984
July 1984
1984
1984
Questions concerning this report should be directed to:
Alaska Power Authority
334 W. 5th Avenue
Anchorage, Alaska 99501
Telephone (907) 276-0001
ARLIS
Alaska Resources
Library & Information Services
J\nchorage,AJaska
-
-
-
-
CONTENTS OF REPORT NO. 2
INTRODUCTION
Part 1. The Outmigration of Juvenile Salmon from the Susitna River
above the Chulitna River Confluence.
Part 2. The Distribution and Relative Abundance of Juvenile Salmon in
the Susitna River Drainage Above the Chulitna River
Confluence.
Part 3. Juvenile Salmon Rearing Suitability Criteria.
Part 4. Juvenile Salmon Rearing Habitat Models.
Part 5. Resident Fish Distribution and Population Dynamics in the
· Susitna River Below Devil Canyon.
Part 6. Resident Fish Habitat Studies.
Part 7. Modelling of Juvenile Salmon and Resident Fish Habitat.
-
.....
INTRODUCTION TO REPORT NO. 2
This volume of the series includes juvenile salmon and resident species
studies conducted during the period May to October, 1983. The majority
of these studies took place in the Sus i tna River reach between the
Chulitna River confluence and Devil Canyon, but a small amount of
sampling (primarily for resident species) was conducted below the
Chulitna River confluence. -
We have used a format for presenting the 1983 data which is different
from that of previous years. The studies are organized into individual
papers (Parts 1 to 7) which are essentially complete reports by them-
selves. The papers contain summary tables and figures; no long
appendices of raw field data are included. Printouts of the raw data or
access to computer files of raw data are available upon request.
There are four general categories of studies included in this volume.
The first category covers basic distribution and relative abundance
information on each species, similar to the studies from previous years.
This information is contained in Part 2 for juvenile salmon species and
in Part 5 for resident species. However, the emphasis this year is on
distribution by macrohabitat type. This frequency of use data may be
coupled with the total surface areas of these macrohabitat types at
different levels of discharge (which is being compiled by Trihey and
Associates) to provide an estimate of the· habitat potential of the
reach. Another difference is that the apparent causes of the observed
distributions are analyzed in greater detail than in reports from
previous years.
The second category of studies includes movement and migration data.
Information on the outmigration of juvenile salmon is contained in Part
1 and data on movement and migration of resident species can be found in
Part 5. With an eye toward new technology, we used a battery-powered
portable microcomputer to store data on outmigrating salmon. This
eliminated several steps in the process of transferring field data to
the final computer data base and also reduced the number of data pro-
cessing errors. Radio-tagging of selected resident species made it
possible to determine the amount of time these fish spend in each
macrohabitat type; this information can be used in determining the
relative value of each macrohabitat type for the species.
Radiotelemetry also made it possible to track resident species to their
spawning areas and then obtain data on spawning habitat.
The third category of studies included in this volume covers population
dynamics, including population estimates. A new technique which yielded
interesting results was used this year to obtain population estimates
and percent survival information for chum and sockeye salmon juveniles.
We captured newly-emergent chum and sockeye salmon at their natal areas
and tagged them with coded wire tags. A sample of the fish were
subsequently recaptured in two downstream migrant traps. This work is
described in Part 1. Population estimates for several species of
resident fishes were attempted using a capture-recapture technique.
----------------~------------------
These data were analyzed by the CAPTURE computer program which calculat-
ed capture probabilities and maximum likelihood estimates of population
size (Part 5). A version of this model was implemented on a portable
microcomputer so that biologists would have on-site verification that
the juvenile salmon sampling techniques were providing appropriate
capture probabilities (Part 2).
The fourth and most emphasized category of studies includes the habitat
relationships of each species. The primary factors examined in these
studies are discharge and the relation of species/life stages to
discharge-influenced variables such as depth and velocity. However,
other variables, especially cover, are also examined. The influences of
habitat parameters on juvenile salmon outmigration is examined in Part 1
and the effect of habitat variables on the distribution and relative
abundance of juvenile salmon is covered in Part 2. Habitat data for
spawning resident species are presented in Part 6. Suitability criteria
curves for several variables are developed for juvenile salmon in Part 3
and for resident species in Part 6.
These suitability criteria are used in habitat models described in Part
4 and Part 7. Results of the Instream Flow Group (IFG) hydraulic models
in simulating habitat (weighted useable area) are presented in Part 7.
In Part 4, we develop a new kind of habitat model which requires
significantly less field data collection than the IFG models and which
runs on a microcomputer rather than the mainframe. These two kinds of
models are evaluated and compared in Part 7. Finally, Part 7 discusses
the implications of the models and all the other data in determining the
instream flow requirements of juvenile salmon and resident species.
-
-
-
-
-
-
-
-
PART 1
The Outmigration of Juvenile Salmon from the
Susitna River Above the Chulitna River Confluence
-
-
-
-~
ABSTRACT
THE OUTMIGRATION OF JUVENILE SALI\101~ FROM THE
SUSITNA RIVER ABOVE THE CHULITNA RIVER CONFLUENCE
1984 Report No. 2, Part 1
by Kent J. Roth, Daniel C. Gray, and Dana C. Schmidt
Alaska Department of Fish and Game
Susitna Hydro Aquatic Studies
2207 Spenard Road
Anchorage, Alaska 99503
Population estimates of juvenile salmon were obtained by mark-recapture
using a unique application of the coded wire tagging technique during
1983. One-half length coded wire tags were used to mark 24,287
post-emergent chum and 17,963 post-emergent sockeye salmon fry at four
sloughs and one tributary of the Susitna River between the Chulitna
River confluence and Devil Canyon. Tag retention rates averaged 96% and
total mortalities caused by the capture and tagging procedure were 1%.
Sixty-two coded wire tagged chum salmon fry and 394 tagged sockeye
salmon fry were recovered in two downstream migrant traps located in the
Susitna River five miles above the Chulitna River confluence. The
mark-recapture estimates indicated that 3,322,000 chum salmon fry and
560,000 sockeye salmon fry migrated downstream past the outmigrant traps
during 1983. Estimated survival rates between potential egg deposition
and outmi grati on for chum and sockeye salmon fry were 14% and 41%,
respectively. The downstream migrant traps collected all five species
of Pacific salmon during the open water period. Pink salmon trap
catches were highest in early June, and peak outmigration of chum salmon
occurred in mid June. Chinook, coho, and sockeye salmon juveniles were
collected at the traps throughout the sampling season, with peaks
occurring during high mainstem discharge levels in early June, early
July, and mid August. The rate of outmigration of chum salmon showed a
higher correlation with discharge than that of other species.
i
-
TABLE OF CONTENTS
ABSTRACT. . . • • • . . . . • • . . . . • . . . . . • . . . • • • . . . . . . . • • . • . . • • . • . . . . . . . . • . . i
LIST OF FIGURES •...•.•....•.••.•.••...••..•••••••..••...•.. ~..... iv
LIST 0 F TABLES • • • • . . . • • • • . . • • • . . • • . • . • • . . . • • • • . • . . • • . • . • . . . . • . . • . v
LIST OF PLATES................................................... vi
LIST OF APPENDIX TABLES.......................................... vii
1.0 INTROOUCTI01'4................................................ 1
2.0 ~IETHODS ••••••••••••••••••••••••••••••••••••••••••••••••••••• 3
3.0
2.1 Study Locations........................................ 3·
2.2 Field Data Collection.................................. 3
2.3
2.2.1 Coded wire tagging •....•........•...............
2.2.2 Downstream migrant traps ..••••••..•••..•....••..
2 • 2. 3 Oy e mark i ng .•••••...•.••••••..•••••.••.....•....
Data Recording ........................................ .
3
6
8
9
2.3.1 Coded wire tagging.............................. 9
2.3.2 Downstream migrant traps........................ 9
2. 3. 3 Oy e ma-rking. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 Data Analysis.......................................... 9
2.4.1 Population and survival estimates............... 9
2.4.2 Juvenile salmon catch per unit effort........... 11
2.4.3 Relation of outmigration to habitat variables... 12
RESULTS •••••••••.••••••••••••••••••••••••••••••••••••••••••• 13
3.1 Coded Wire Tagging and Recovery .•..•.....•.•...•.••.... 13
3.2 Population Estimates and Survival Rates of Outmigrants. 13
3.3 Outmigration Rates From Selected Sloughs............... 17
3.4 Juvenile Salmon Catch Per Unit Effort.................. 17
3.5 Relation of Outmigration to Habitat Variables .•.•••.•. 22
3.5.1 Interrelationship of mainstem discharge,
temperature and turbidity....................... 22
3.5.2 Effects of mainstem discharge on outmigration... 28
ii
4.0 DISCUSSIOI~ ••••••••••• o ...................................... .
4.1 Coded Wire Tagging and Recovery ....•..•.•.••.....•••...
4.2 Dye Marking and Outmigration Rates ••.•..••••••.•.......
4.3 Survival of Outmigrants •....•..•...•.....•.••.••...••.•
4.4 Comparison of Trap Catch Rates .••....••.••••.•.•••..••.
4.5 Relation of Outmigration to Habitat Variables ........ ..
5.0 CONTRIBUTORS •••.•••..•••.•••••.••.••••.•..•.••..•.•.•..••.••
6. 0 ACKJ~OWLEDGEMENTS •..............................................
7.0 LITERATURE cr-rED ................... ee·•······················
8.0 APPENDICES
Appendix A Sampling Selectivity of the Outmigrant Traps •..••
Appendix B The Schaefer Estimator of Population Size ...••••.
Appendix C Comparison of Daily Catch Per Hour Between
Outmigrant Trap 1 and Trap 2 •.••.•.••.•.•.•...•..
iii
~
Page
33
33
34
35 -
36
40
42
43
44
49
52 -
57 -
-
""" i
-
LIST OF FIGURES
Figure Title
1 Map of the Susitna River from Talkeetna upstream
to Devil Canyon showing the coded wire tag deploy-
ment and recovery sites..................................... 4
2 Chum salmon fry daily catch per hour recorded at
the downstream migrant traps, May 18 through
August 20, 1983........ .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3
4
5
6
7
8
Sockeye salmon fry daily catch per hour recorded
at the downstream migrant traps, May 18 through
August 30, 1983 ............................................. .
Chinook salmon age 0+ and 1+ daily catch per hour
recorded at the downstream migrant traps, May 18
through August 30, 1983 .................................... .
Coho salmon age 0+ and age 1+ or older daily catch
per hour recorded at the downstream migrant traps,
May 18 through August 30, 1983 •.•••.•••..•••.•...••..•...•••
Pink salmon fry daily catch per hour recorded at
the downstream migrant traps, May 18 through July
8' 1983 .................................................... .
Cumulative catch for age 0+ chi nook, coho, sock-
eye, chum, and pink salmon recorded at the down-
stream migrant traps, May 18 through August 30,
1983 ....................................................... .
Mainstem discharge, water temperature, and turbid-
ity recorded on the Susitna River, May 18 through
September 25, 1983 ......................................... -· •
9 Relationship of mean daily discharge with mean
daily chum salmon fry catch per hour at the down-
stream migrant traps, May 18 through July 15, 1983 •...••••..
10 East bank outmigrant trap (Talkeetna Station, RM
103.0) cumulative catch recorded for juvenile coho,
chinook, chum, and sockeye salmon during 1982 and
1983 ......................................................... .
iv
20
21
23
24
25
26
32
39
DRAFT /PAGE 6 7/13/84
4/15/84, 4/16/84, 6/14/
SER3J/Part 1 -Contents
LIST OF TABLES
Table
1
2
3
4
5
6
7
Title
Coded wire tag release data for chum salmon fry on
the Susitna Riverby site and date, 1983 ••....•...•...•.....
Coded wire tag release data for sockeye salmon fry
on the Susitna River by site and date, 1983 ••..•••••.••••.•.
Comparision of release and recovery data for coded
wire tagged chum salmon fry on the Susitna River
by site and date, 1983 ..................................... .
Comparison of release and recovery data for coded
wire tagged sockeye salmon fry on the Susitna
River by site and date, 1983 •.•••.•.••••.••••••..••..••.....
Population size, rate of emigration , and rate of
emergence of chum salmon fry at Slough 11 as esti-
mated by Bailey•s Deterministic Model using mark-
recapture data collected June 5, 6, and 7, 1983 ...••.•.•....
Outmigration rates of chum and sockeye salmon fry
from Slough 11 and Slough 21 determined by 24 hour
weir catches, 1983 ......................................... .
Listing of age separation values by length for
juveni 1 e chinook, sockeye, and coho sa 1 man
captured by two week periods on the Susitna River
between the Chulitna River confluence and Devi 1
14
14
15
15
18
19
Canyon, 1983. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
8 Summary statistics for habitat variables recorded
on the Susitna River, between the Chulitna River
confluence and Devi 1 Canyon, May 18 through
September 25, 1983.......................................... 27
9 Summary statistics for juvenile salmon catch per
hour by species and age class recorded at the
downstream migrant traps, May 18 through September
25, 1983..................... .. . . . . . . . . .. . . . . . . . . . . . . . . . . . . . 27
10 Correlation coefficients between discharge and
temperature, and discharge and turbidity, for the
Susitna River between the Chulitna River con-
fluence and Devil Canyon, 1983.............................. 29
11 Correlation coefficients ·between discharges and
juvenile salmon daily catch per hour by species
and age class, for the Susitna River between the
Chulitna River confluence and Devil Canyon, May 18
through August 30, 1983..................................... 30
v
""'1
' i
-
-
-
-
-
-
-I
-
-
-
-
-
DRAFT/PAGE 7 7/13/84
4/15/84, 4/16/84, 6/14/
SER3J/Part 1 -Contents
LIST OF PLATES
Plate
1
2
3
Title
A weir set near the mouth of Slough SA (Rm 125.3)
to co 11 ect outmi grating chum and sockeye sa 1 man
fry for coded wire tagging, 1983 .•..•.•.••.•••..••.•..•.....
Separation of salmon fry by species and length
prior to the implantation of coded wire tags,
1983 ....................................................... .
The east bank downstream migrant trap at its
fishing location on the mainstem Susitna River at
River Mile 103.0 ........................................... .
4 A dorsal view of a one-half length coded wire tag
in the snout of a sockeye salmon fry recovered in
5
5
7
the downstream migrant traps, 1983.......................... 10
5 A side view of a one-half length coded wire tag in
the dissected snout of a sockeye salmon fry
recovered in the downstream migrant traps, 1983............. 10
vi
DRAFT/PAGE 8 7/13/84
4/15/84, 4/16/84, 6/14/
SER3J/Part 1 -Contents
LIST OF APPENDIX TABLES
Appendix Table Title
A-1 Correlation coefficients (r) for juvenile salmon
catch per hour compared to trap ve 1 oci ty at each
of the downstream migrant traps, by species and
age class, 1983 ••••••••••••••• ~............................. 51
B-1 Data collected on the coded wire tag, mark-
recapture experiment for sockeye sa 1 man fry to
provide a population estimate using the methods
outlined by Schaefer (1951) .••••••••••••••••••••••••••••••••
B-2 Computation of the sockeye salmon fry outmigrant
population from the data presented in Appendix
Tab 1 e B -1 ........................ : .................... e •••••
B-3 Data collected on the coded wire tag, mark-
recapture experiment for chum salmon fry to
provide a population estimate using the methods
outlined by Schaefer (1951) ••••••••.•••••••••••••••.••••••..
B-4 Computation of the chum salmon fry outmigrant
population from the data presented in Appendix
Table B-3 •••••••••••••••••••••••••••••••••••••••••••••••••••
C-1 Comparison of daily catch per hour of juvenile
salmon in trap 1 versus trap 2 by species and age
class ...................................................... .
vii
55
55
56
56
58
~
-
~
-
_:_")
!I'll!!
.I
-
-
-
-
-
1.0 INTRODUCTION
Since November 1980, studies of the distribution, relative abundance and
timing of outmigration of juvenile salmon in the Susitna River have been
part of the Susitna Hydro Aquatic Studies program. A portion of these
studies have been directed towards determining the interactions of
outmigrating juvenile salmon with their habitat to provide the data
necessary to predict their response to environmental changes associated
with hydroelectric development. This report presents the results of the
juvenile salmon outmigration studies conducted on the Susitna River
between the Chulitna River confluence and Devil Canyon during the open
water period of 1983. Five Pacific salmon species are addressed in this
report: sockeye (Oncorhynchus nerka), chum (0. keta), chinook (0.
tshawytscha), coho (Q. kisutch), and pink (.Q.. gorbusefi"a}. -
Previous distribution and abundance studies of juvenile salmon in the
Susitna River were conducted by Barrett (1974), Friese (1975), and Riis
and Friese (1978) as part of preliminary environmental assessments of
the proposed hydroelectric development. Juvenile salmon life histories
including outmigration timing have also been studied on the Susitna
River (ADF&G 1981, 1983b, 1983c) and its major tributary streams
including the Deshka River (Delaney et al. 1981), Willow Creek (Engel
and Watsjold 1978) and Montana and Rabideux creeks (Kubik and Wadman
1978).
The effects of discharge fluctuations on juvenile salmon during the
periods of incubation, emergence and outmigration have been reported by
White (1939), Neave (1953), Gangmark and Broad (1956), Wickett (1958),
Andrew and Geen (1960), and McNeil (1966). Other factors affecting
survival and timing of outmigration include the size of smelts (Foerster
1937 and Barnaby 1944), predation (Neave 1953; Roos 1958; Hunter 1959;
and Thompson 1964), and water temperature (Foerster 1968 and McCart et
al. 1980). Changes in photoperiod have also been reported to influence
the timing of juvenile salmon outmigration (Hunter 1959; McDonald 1960;
Burgner 1962; Heard 1964; and Hartman et al. 1967).
To provide a clearer understanding of the relationship between present
production and natural changes in habitat conditions of the Susitna
River, a portion of the 1983 aquatic studies were directed toward
quantifying the rates of survival and the rates and timing of
outmigration of juvenile salmon in the Susitna River between the
Chulitna River confluence and Devil Canyon.
Specific objectives of this portion of the 1983 program were as follows:
A. Estimate the current numbers of chum and sockeye salmon
juveniles outmigrating from the study reach.
B. Estimate the egg-to-outmigrant survival for chum and sockeye
sal man for the peri ad spent in the study area under present
environmental conditions.
C. Determine the periods of freshwater residence and the timing
of outmigration for all species of juvenile salmon in the
- 1 -
·study area and the relationship of outmigration and habitat
parameters.
D. Continue the collection of biological data including species,
age class and length frequency distribution to determine the
condition and stage of development for each species during
outmigration.
E. Provide descriptions of the variability of biological
development and outmigration behavior among the different
species and within a given species.
Data were collected at downstream migrant traps in 1983 to determine the
outmigration timing windows and periods of freshwater residence for
juvenile salmon (objectives C, D and E). Information was also collected
on the migration and redistribution of juvenile resident fish species
within the study reach (See Part 5 of this Report).
A coded wire tag, mark-recovery program was initiated during 1983 to
estimate the population size and survival rate of juvenile sockeye and
chum sal man during the peri ad they spend above the outmi grant traps
(Objectives A and B). These population estimates may be compared with
estimates of egg production in order to calculate survival rates for
sockeye and chum salmon during the period of freshwater residence in the
study area. By correlating survival rates with habitat conditions
at the individual study sites, it is possible to evaluate the
contribution that these sites make to the overall production of chum and
sockeye salmon outmigrants from this reach.
The coded wire tagging program will also assist in determining the
viability and importance of sockeye salmon stocks between the Chulitna
River confluence and Devil Canyon. Although not an integral part of
this study, the future recove-ry of tagged adult salmon will provide
definitive evidence concerning the contribution of sockeye salmon
spawning in this reach of river to the number of returning adults.
Through the continued monitoring of the survival and distribution of
existing stocks as a function of natural environmental changes, more
accurate predictions can be made on the subsequent effects of habitat
changes on juvenile salmon production in this reach of river. Continued
monitoring will also provide weighted values for the different species
during certain critical periods of their freshwater residence. This
data coupled with data collected by other portions of the Susitna Hydro
Aquatic Studies program-will assist in developing mitigation require-
ments necessary to maintain existing salmon stocks.
- 2 -
-
-
-
.....
!""'
i
I
-
2.0 METHODS
2.1 Study Locations
The coded wire tag deployment sites and tag recovery sites are shown in
Figure 1. Coded wire tagging sites were selected from locations which
had previous high density spawning history {ADF&G 1983a), and from
surveys of the availability of sufficient numbers of post-emergent chum
and sockeye salmon for collection and tagging. The tagging sites were
Sloughs 8A (RM 125.3), 9 (RM 129.2), 11 {RM 135.3), and 21 (RM 142.0),
and one tributary site at the mouth of Indian River (RM 138.6). Tag
recovery efforts were conducted at two downstream migrant traps located
on opposite banks of the mainstem Susitna River at RM 103.0. Dye
marking and data collection on outmigrant rates were conducted at Slough
11 and Slough 21.
2.2 Field Data Collection
2.2.1 Coded wire tagging
The sample sizes required to provide valid population estimates for each
species were calculated prior to the tagging program using the estimator
provided by Robson and Regier {1964). The actual numbers of fish tagged
for each species was ultimately determined by the availability of fish
at the collection sites and the time constraints of the field program.
The coded wire tagging program was conducted by five fisheries personnel
based at the Gold Creek camp (RM 136.8) from May 16 through June 19,
1983. Tagging operations were conducted mainly at the individual
collection sites, and the primary tagging equipment and personnel were
staged in a six-man portable wall tent. However, if logistical or
equipment problems occurred, the fish to be tagged were transported from
the collection area to the base camp and then returned to the collection
site for release following tagging.
The primary fisheries collection techniques were beach seines, dipnets,
and backpack electrofishing units. Beach seines were used to weir off
the downstream end of the study site and were checked periodically to
collect fish and remove debris (Plate 1). Beach seining, dipnetting,
and backpack electrofishing supplemented the weir catches at sites where
weiring did not provide enough fish for the tagging operation or at
those sites where the weirs were not deployable.
The coded wire tagging equipment was leased from Northwest Marine
Technology, Inc. of Shaw Is 1 and, Washington, and operated in accordance
with the manufacturer• s instructions and operation manuals. The leased
equipment was the NMT MK2A tagging unit and included the following:
o Coded wire tag injector with 1/2 length tag capability
o Quality Control Device {QCD)
o Water pump
o Portable power supply
The equipment was field portable and included a more compact prototype
of the standard quality control device.
- 3 -
II
e SAMPLING SITE
Figure 1. Map of the Susitna River from Talkeetna upstream to
Devil Canyon showing the coded wire tag deployment
and recovery sites.
-4-
-
""'l!
I
....
-
-
--
-
-
Plate 1. A weir set near the mouth of S1ough 8A (RM 125.3) to
col1ect outmigrating chum and sockeye salmon fry for
coded wire tagging~ 1983.
Plate 2. Separation of salmon fry by species and 1 ength prior
to the implantation of coded wire tags, 1983.
One-half length binary coded wire tags measuring 0.02 inches (0.533 mm
long and 0.01 inches (0.254 mm} in diameter were obtained from Northwes
f.1arine Technologies, Inc. The one-half length tags were used due to the
small size of the fish to be tagged. The total length of post emergent
chum salmon averaged 40 mm (1,500 fry/lb) and the total length of
sockeye fry averaged 32 mm {3,000 fry/lb). Tag injector head molds were
constructed by the manufacturer from samples ·of fish of the species and
size ranges to be tagged.
The coded wire tag implantation procedures were similar to those
outlined by Moberly et al. (1977) and Koerner (1977). The captured fish
were separated by species and length prior to tagging (Plate 2), as
physical differences between fish required the use of separate head
molds for each species and length class. A sample of 50 fish of each
group was measured for total length to determine the proper headmolds
for the tagging procedure. The adipose fin \-Jas clipped from each fish
prior to tagging to provide a visual indicator to the presence of a
coded wire tag during recovery efforts. At the end of eaCh tagging day,
a subsample of 100 tagged fish were anesthetized and passed through the
quality control device to determine the tag retention rate. r~ortalities
were recorded the following day. All tagged· fish were released at the
sites of collection. The number of valid tagged fish was determined
daily by subtracting the number of mortalities from the number of tota1
tagged fish and then multiplying this by the tag reten~ion rate.
Only one tag code was used for a given site during a single tagging
period, which ranged from one to six days. The same tag code was used
for both sockeye and chum salmon fry at a site during each tagging
period, but physical differences between fish required the use of
separate head molds for each species and length class. Up to three
different code groups were used at a single collection site during the
entire program with a minimum of ten days separating the re 1 eases of
different tag codes at the same site.
2.2.2 Downstream migrant traps
A two to three person crew recovered coded wire tagged fish using bJo
downstream migrant traps (Plate 3) operated at Talkeetna Station on the
mainstem Susitna River (RM 103.0) ~ 23 miles downstream from the nearest
coded wire tagging site (Figure 1). The traps were opet~ated off the
east bank (Trap 1) and the west bank (Trap 2) of the river on a
continuous 24 hour schedule from May 18 through August 30, with short
periods of down time due to high water and debris, manpower limitations,
and trap repair. The traps were checked from two to nine times daily,
depending on the capture rate and the debris load. The traps were
operated on an abbreviated schedule during September. A description of
the inclined plane traps is presented in the FY84 procedures manual
(ADF&G 1984).
Trap fishing depths and distances from shore were adjusted to maximize
catches and minimize mortalities. All juvenile fish captured were
anesthetized using MS-222 (Tricaine methanesulfonate). Field specimens
were identified using the guidelines set forth by Trautman (1973),
- 6 -
-
-
-
-
-
-
-
-
1 1 . l 1
Plate 3. 1he east ban~ downstream migrant trap at its fishing location on the mainstem susitna River
at River Mi1e 103.0t 1982.
1 1 1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
' 1
1
1
1
1
McConnell and Snyder (1972) and Morrow (1980). Chum and sockeye salmon
juveniles having an adipose fin clip were passed through a Northwest
Marine Technologies FSD-1 field sampling detector to verify the presence
of a coded wire tag. The detector sensed the magnetic field emitted by
the tag and provided an auditory cue when a tagged fish was passed
through. All coded wire tagged fish recovered at the traps were pre-
served in 10% formalin for later tag removal and decoding. All other
fish were retained until anesthetic recovery was complete and then
released downstream of the traps.
Daily habitat data measured at the downstream migrant traps were air and
surface water temperatures (°C), turbidity (NTU), pH, dissolved oxygen
(ppm), specific conductance (umho/cm), water velocity (ft/sec), and
mainstem stage data. The equipment and methods used to collect and
measure the habitat data are contained in the FY84 procedures manual
(ADF&G 1984).
Scales were collected from a sample of juvenile fish captured in the
traps for comparison with length frequency data for final age determina-
tions. Scales were placed between two microscope slides, and age
determination from the call ected seale samples was conducted at the end
of the field season with a Micron 780 portable microfiche reader using
the guidelines provided by fl.1osher (1969) and Lux (1971).
2.2.3 Dye marking
Bismark Brown dye was used to mark a portion of the juvenile salmon
collected at the coded wire tagging sites to determine the dye retention
rates and the ability to observe the dye mark on recovered fish. The
fish were soaked for 30 minutes in a solution of one gram of dye fot
each 30 1 Hers of water.
The dye was also used in conjunction with coded ~tire tagging on chum
salmon fry in a pilot study to determine the feasibility of providing
population estimates of outmigrating fry from individual sites. The
mark and recovery experiment was conducted over a three day period using
the guidelines set forth by Ricker (1975).
Fish were collected in a beach seine set across the 1ower portion of
Slough 11. On the first day, captured chum fry were coded wire tagged
and then dyed and released. Marked fish ~-Jere randomly distributed in
the study site above the collection net. All chum collected on the
second day were checked for marks. Unmarked fish were dyed and then
released with the previously marked fish. On the third day, captured
chum fry were separated into the following groups and totaled: coded
wire tagged and dyed fish, dyed fish with no coded wire tag, and
unmarked fish. All fish were released at the end of the experiment.
Outmigration rates were also monitored during six 24-hour periods at
sloughs 11 and ·21 using beach seines set across the lower portions of
each site.
- 8 -
-
-
-
-
-
-
2.3 Data Recording
2.3.1 Coded wire tagging
Coded \vire tagging data recorded at each site included species, mean
tota 1 1 ength, numbers of fish tagged, percent tag retention, and mor-
tality. Date, tag code, and time of release were also recorded. Total
numbers of fish tagged by species, collection site, and release date as
well as final tag retention and mortality were tabulated for each code
group. Total valid tagged fish were determined by subtracting the
mortalities for each days tagging from the total number of fish tagged
and then multiplying this by the tag retention rate.
2.3.2 Downstream migrant traps
Biological data collected at the downstream migrant traps included catch
by species, age clas-s, total length, presence of a coded wire tag, fate,
and scale sampling. Up to 50 fish of each species and age class were
measured for total length {tip of snout to tip of tail) in millimeters
(mm) daily and all remaining fish were tallied for total catch. Trap
depth and distance from shore were recorded for each trap at every
check. All other habitat parameters {Section 2.2.2) were measured once
daily. Refer to Appendix A for a discussion of the sampling selectivity
of the traps.
Biological and habitat data were entered in the field directly into an
Epson HX-20 microcomputer which provided a magnetic tape and paper
printout of the data. Operational procedures for the microcomputer and
the associated data form program are presented in the FY84 procedures
manua 1 {ADF&G 1984). Computer entries were made for each trap check
throughout the field season. Printouts and cassettes were periodically
transferred to Data Processing. These data were then transferred to a
mainframe computer for later data retrieval and analysis.
Coded wire tags were dissected from preserved fish at the end of ·the
field season and were decoded using a reading jig and an American
Optical binocular microscope {Plates 4 and 5).
2.3.3 Dye t~arking
Total numbers of dyed fish, date of release, date of recapture. and
periods of dye retention were recorded.
2.4 Data Analysis
2.4.1 Population and survival estimates
Potential egg deposition refers to the total number of eggs carried
upstream by a given spawning run and is determined by multiplying the
average fecundity by the number of female spawners. The estimated
number of young fish emigrating from the study reach is expressed as a
percentage of the potential egg deposition and represents the percentage
survival between these points in the life cycle.
- 9 -
-~----·---·------·
Plate 4. A dorsal view of a one-half length coded wire tag
(arrow) in the snout of a sockeye salmon fry recovered
in the downstream migrant traps, 1983.
Plate 5. A side view of a one-half length coded wire tag (arrow)
in the dissected snout of a sockeye salmon fry re-
covered in the downstream migrant traps, 1983.
-10-
-
-
-
r
-
Potentia 1 egg deposition for chum and sockeye sa 1 mon in the Sus i tna
River between the Chulitna River confluence and Devil Canyon was gen-
erated from the 1982 adult population data collected at Curry Station
(RM 120). One hundred percent of the sockeye and over 99% of the chum
salmon spawning in the study reach used the habitats located above this
survey site during 1982 (ADF&G 1983a). The chum salmon population
estimates of adults at Curry Station were reduced by 40% to account for
milling fish which eventually spawned below the Chulitna River conflu-
ence; no milling factor was suggested for sockeye spawning in 1982
(Bruce Barrett, personal communication). The number of female spawners
was determined from sex ratios recorded at Curry Station during 1982
(ADF&G 1983a). Fecundities of Susitna River chum and sockeye salmon
were determined from egg counts conducted in 1983 (Barrett et al. 1984).
Population estimates for chum and sockeye salmon outmigrants were
calculated using the adjusted Petersen estimate outlined by Chapman
(1951) and the marking experiments provided by Schaefer (1951). Final
survival estimates for both species were determined by taking the
population estimates and dividing by the calculated potential egg
deposition for each species. Only the numbers of valid tagged fish (as
described in Section 2.2.1) were used in the calculations. Total tag
recoveries at the traps include only those fish which had a coded wire
tag. Clipped fish with no tag were not considered in the estimates.
Population and recruitment estimates for the dye marking experiment were
calculated using the multiple mark-recapture technique outlined by
Bailey {1951), as discussed by Ricker (1975). Mortalities were low
during the experiment and were not factored in the estimates.
2.4.2 Juvenile salmon catch per unit effort
The catch per unit effort (CPUE) data collected on juvenile salmon at
the downstream migrant traps are presented as the combined trap catch
per hour for each calendar date of sampling effort. The number of fish
of a given species and age class which were caught on a particular day
was divided by the number of hours the trap fished that day.
The catch per hour rates plotted for each species and age class of
juvenile salmon collected at the traps during 1983 were smoothed using
the von Hann linear filter (Dixon et al. 1981). The equation is:
Z(t) = iY(t-1 ) + tY(t) + iY(t+1)
where: Z(t) : smoothed catch per hour for day (t) and
Y(t) -observed catch per hour for day (t)
This is similar to a three day moving average except that the current
day is weighted twice as heavily as the preceding and subsequent days.
The cumulative catch totals for each species are for both traps combined
and were adjusted to 24 hour intervals for the sampliAg conducted from
May 18 through August 30. The totals were adjusted for the periods not
sampled (six days in all) by tabulating the mean of the total catches
-11 -
recorded for the three days preceding and the three days following each
unsampled period.
2.4.3 Relation of outmigration to habitat variables
Correlation analysis of the relationships between outmigration timing of
juvenile salmon and environmental variables recorded for the Susitna
River at the downstream migrant traps was conducted using the 1983 data.
Turbidity and water temperature were recorded daily at the traps through
the sampling period. Discharge levels are provisional data collected by
the U. S. Geological Survey at the Gold Creek gaging station (RM 136.6).
Temperature values for days the traps were not fished were provided by a
thermograph located at Talkeetna Station (RM 103.0).
Correlation analysis for chinook, coho, and sockeye salmon included the
106 days of trap fishing effort which occurred between May 18 and
September 25. Correlation analysis on chum and pink salmon catch data
was performed only for the period from May 18 through July 15 as 98.4%
of the chum and 100% of the pink salmon were captured during this
period. Discharge and catch per hour data were smoothed by the linear
filter described above. The significance test for all correlations was
to determine whether the correlation coefficient was significantly
greater or less than zero.
Because some of the variables appeared to lag behind discharge, dis-
charge correlations were included with one day (discharget 1 ) and two
day (discharget 7 ) lags. The season was separated into tnree periods
early (May 18 ttl June 15), middle (June 16 to August 31), and late
(September 1 to 25) because of different climatological and hydrological
processes occurring during these periods. The early period follows
break-up and is a time of melting ice and snow and increasing solar
insulation. Glacial melting occurs mainly during the middle period.
Also, there often are large amounts of rainfall during this period.
September is a time of rapidly declining water temperature and
tu rb i d i ty • ·
Autocorrelation coefficients were calculated for each variable on both
raw and transformed (log (X+l)) data for the period May 18 through
August 30. Catch per hour for the six days with no sampling data during
this period were interpolated to provide a continuous time series.
September data were not included in this portion of the analysis because
of the limited sampling conducted during this period.
-12 -
-
-
-·
-
-
-
-
-
-
-
3.0 RESULTS
3.1 Coded Wire Tagging and Recovery
A total of eight distinct tag code groups were implanted in chum salmon
fry at five study sites during 1983. Table 1 presents the total chum
salmon fry tagged by site and tag code and includes tag retention and
mortality rates. A total of 24,287 valid tagged chum fry averaging 40
mm total length were released between May 24 and June 19. Tag retention
rates ranged from 91.7 to 100% and averaged 97.7%. Mortality rates
between tagging and release averaged 1.1% and ranged from 0.1 to 2.4%.
A total of 17,963 valid tagged sockeye salmon fry averaging 32 mm total
1 ength were released between May 24 and June 20. Six tag codes were
distributed at three study sites (Table 2). Tag retention rates for
sockeye fry averaged 96.3% and ranged from 92.6 to 100%. Tagging
mortality averaged 1.2% for sockeye salmon fry and ranged from 0.3 to
6.3%.
Of the 8,616 chum salmon fry captured and examined for tags at the
downstream migrant traps during 1983, 62 tagged chum salmon fry (0.3% of
the total tagged chum released) were recovered (Table 3). Trap recov-
eries of tagged chum fry were made from 0 to 28 days fo 11 owing their
release at the tagging sites. In addition, two chum fry with clipped
adipose fins but no coded wire tags were recovered in the traps. When
compared to the total tagged chum salmon fry recovered, this provides a
tag retention rate at the traps of 96.9%.
A tot a 1 of 394 tagged sockeye sa 1 man fry ( 2. 2% of the tota 1 tagged
sockeye released) were recovered from the 12,312 age 0+ sockeye captured
and examined for tags at the outmigrant traps (Table 4). Tag recoveries
occurred within zero to 113 days following the release of sockeye at the
tagging sites. Nineteen sockeye salmon fry with clipped adipose fins
but no coded wire tags were also captured, providing a tag retention
rate of 95.4% for sockeye fry at the traps.
A test of ad·ipose fin clip efficiency conducted at the traps during a
48-hour period of recovery efforts showed no captures of tagged fish
that did not also have an adipose fin clip. No partial fin clips or
regeneration of the adipose fin were observed during the recovery
efforts. Also, no sockeye or chum salmon fry were observed to have
naturally missing adipose fins during the fin clipping operation.
A t-test comparison of daily recoveries of coded wire tagged chum and
sockeye sa 1 mon to the tota 1 daily captures of each species showed no
significant difference (p < 0.05) in recovery rates between the two
downstream migrant traps.
3.2 Population Estimates and Survival Rates of Outmigrants
The total potential egg deposition for chum and sockeye salmon in the
study area during 1982 was calculated using the following formula:
-13 -
Table 1. Coded wire tag release data for chum salmon fry on the Susitna River
by site and date, 1983,
Percent
Tagging Site Dates of Number of Dates of Tag Percent
(River Mile\ Ta~~in~ Fish Ta!!ged Release Retention Mortality
Slough 21 5/25-29 8,555 5/27-30 99.5 0. l
(IDI 142. 0) 6/15-16 2,149 6/19 99.5 1.2
Indian River 6/4-5 1,131 6/5 91.7 ? 4!!/
(R.\1 138.6) 6/18 2,541 6/19 93.0 2:.~/
Slough 11 5/21-22 2,579 5/24 93.9 2.2!1
(RM 135.3) 6/4-9 2,409 6/5-10 99.8 0.3
Slough 9 5/30 l3 6/5 100.0 0.0
(RM 128.3)
Slough SA 6/10-14 4,910 6/13-15 99. I I. 7!/
(RM 125.3)
TOTAL -ALL SITES 5/21-6/18 24,287 5/24-6/19 97.7 I .1
a/ -Mortalities were due to oxygen loss, thermal stress, or anesthetic.
Table 2. Coded wire tag release data for sockeye salmon fry on the Susitna
River hy site and date, 1983.
Percent
Tagging Site Dates of Number of Dates of Tag Percent
(River Mile' Tagging Fish Ta!!!!ied Release Retention Mortalit:t:
Slough 21 5/27-29 288 5/29-30 100.0 0.3
(RM 142.0' 6/15-16 884 6/19 100.0 1.0
Slough 11 5/23-24 4,264 5/24-25 92.9 0.3
(RM 135.3) 6/5-9 8,491 6/6-10 96.7 0.5
6/19 1,928 6/20 99.0 0.9
Slough 8A 6/10-14 2.108 6/13-15 98.0 6.32..1
(RM 125.3)
TOTAL -ALL SITES 5/23-6/19 17,963 5/24-6/20 96.3 1.2
'!!../Mortalities were due primarily to oxygen loss during transfer.
-14-
-
-
~I
"""
-
~
-
-
-
-
......
!"""
r-
--
-
Table 3. Comparison of release and recovery data for coded wire tagged chum salmon fry on the Susitna Riv
by site and elate, 1983 .
Number Percent Days Bet
Tagging Site Dates of of Fish Dates ofa/ Number of Tags Release
(River Mile) Release Tagged Recovery-Recovered Recovered Recove
Slough 21 5/27-30 8,555 5/30-6/ ~4 12 o. 1 0 to
(R~! 142.0) 6/19 2,149 6/20-7/8 12 o.o 0 to
Indian River . 6/5 1, 131 6/20-21 2 0 ., 15 to
(!U~ 138.6) 6/19 2,451 6/20-26 12 0.5 tu
Slough 11 5/24 2,579 5/25-6/18 9 0.3 l to
(Rl-1 135.3 6/5-10 2,409 6/10-15 3 0. I 0 tc
Slough 9 6/5 13 0 0.0
(RM 128.3)
Slaugh SA 6/13-15 4_, 910 6/15-7/2 12 0.2 0 to
(RH 125.3)
TOTAL -ALL SITES 5/24-6/19 24,287 5/25-7/8 62 0.3 0 to
~/Recoveries were made at the two downstream migrant traps (RM 103.0).
Table 4. Comparison of release and recovery data for coded wire tagged sockeye salmon fry on the Susit:
River by site and date, 1983.
Number Percent Days Bet"'
Tagging Site Dates of of Fish Dates of a/ Number of Tags Releuse
(River Mile) Release Tagged Recovery-_ Recovered Recovered Recover
Slough 21 5/29-30 288 5/31-7/29 4 1.4 to h
(RM 142. 0) 6/l9 884 6/21-8/12 7 O.ll 7 to 'i -
Slough 11 5/24-25 4,264 5/25-9/14 93 2.2 0 to 11
(&.'! 135. 3) 6/6-10 8,491 6/6-8/25 181 2. I 0 to 8
6/20 I ,928 6/22-8/30 22 1.! :> to 7
Slough 8/1 o/13-15 2,108 6/!6-8/21 87 4.1 to 6
(~) 125.3)
TOTAL -ALL SITES 5/24-n/20 17,963 5/25-9/14 394 2 .. 2 0 to 11
~/ ~ecoveries were made at the two downstream migrant traps (RM 103.0).
-15-
Total potential egg deposition = (E) X {M) X {P) X {F)
100
where:
E =Adult population estimate at Curry Station
P = Percent females
F = Average fecundity
M = Percent milling
Adult population estimates at Curry Station during 1982 were 17,648 chum
salmon (adjusted for 40% milling) and 1,261 sockeye salmon (ADF&G
1983a). Females comprised 46.7% of the chum salmon and 32.4% of the
sockeye salmon at the survey site. Fecundities of Susitna River fish
were determined during-1983 to be 2,850 for chum salmon and 3,350 for
sockeye salmon (Barrett et al. 1984). Total potential egg deposition
was calculated to be 23,490,000-eggs for chum salmon and 1,370,000 eggs
for sockeye salmon.
Adjusted Petersen population estimates were generated for outmigrant
chum and sockeye salmon fry from the mark-recapture data using the
formula by Chapman (1951):
where:
N = Estimate of population
M = Number of fish marked
C = Number of fish captured and examined for marks
R = Number of marked fish recaptured
For chum salmon, this fonnula provided an outmigrant population estimate
of 3,322,000 fish with a 95% confidence interval {Ricker 1975) of
2,633,000 to 4,327,000 fish. The age 0+ sockeye sa 1 man outmigrant
population was estimated to be 559,976 fish with a 95% confidence
interval of 508,632 to 619,641 fish.
Si nee tag re 1 eases and trap recoveries were extended over a peri ad of
time, the method outlined by Schaefer (1951) was also used to estimate
the outmigrant populations. The calculations to determine the Schaefer
estimate are provided in Appendix B. This method provided population
estimates of 3,037,000 chum salmon and 575,000 sockeye salmon outmi-
grants.
Using the above data, calculations of survival were made for both
species. An egg-to-outmigrant survival rate of 14.1% was calculated for
chum salmon with the adjusted Petersen estimate and a rate of 12.9% was
determined using the Schaefer estimate. Sockeye salmon survival rates
-16 -
-
-
-
.l!ll!lffi\
-
were calculated to be 40.9% with the Petersen estimate and 42.0% with
the Schaefer estimate.
3.3 Outmigrant Rates From Selected Sloughs
A mark-recapture experiment based on Bailey's Deterministic Model
(Ricker 1975) was conducted at Slough 11 to estimate the population and
the rates of emergence and emigration of chum salmon fry at the study
site. The results of the pilot study are presented in Table 5. A
population of 2,068 chum fry was determined for Day 2 and the daily
emigration rate was estimated to be 32.7% of the population. The daily
recruitment or emergence rate of chum salmon fry during the survey was
estimated at 1.84.
Outmigrant rates for chum and sockeye salmon fry at Sloughs 11 and 21
determined by fyke net catches are presented in Table 6.
3.4 Juvenile Salmon Catch Per Unit Effort
Length frequency distribution and scale analysis data were used to
determine the age class composition for chinook, coho and sockeye salmon
juveniles. The points of length separation of age classes for each
species by two week periods are presented in Table 7. The graphs
presented in this section represent smoothed data, but the catch rates
given in the text of this section are the raw data. A comparison of
unsmoothed daily catch per hour of juvenile salmon for Trap 1 versus
Trap 2 by species and age class is presented in Appendix C.
The catch per unit effort (CPUE) for chum salmon fry collected by the
two downstream migrant traps during 1983 is presented in Figure 2. Peak
catches of chum fry were recorded during late May and early June, and a
second peak was observed in early July. The highest daily catch rate of
16.1 chum per hour was observed on July 6. The major outmigration of
chum salmon fry had occurred by July 15 and the last chum was captured
in the traps on August 20. The total catch for the season was 8,611
juvenile chum salmon.
Sockeye salmon CPU£ at the traps was highest during late June and early
July (Figure 3). Sixty-two percent of the total catch of sockeye salmon
juveniles occurred during this period. The highest catch rate of 16.8
sockeye per hour was recorded on July 1. Age 0+ sockeye salmon ( 1982
brood year) comprised 99.3% of the total trap captures (12,312 fish)
while age 1+ {1981 brood year) comprised the remaining 0.7% (83 fish).
The outmigration of age 1+ sockeye from the study reach was completed by
the end of June.
Chinook salmon juveniles were collected in the traps throughout the open
water period. Small peaks in CPUE were recorded during early June, late
June, and early July, and a large peak was observed during early August
(Figure 4). The highest catch rate of 21.0 chinook per hour was record-
ed on August 11. Age 1+ chinook salmon comprised 7.0% (434 fish) of the
total juvenile chinook salmon catch (6,202 fish) during 1983, and the
outmigration of this age class from the study reach was essential1y
complete by the middle of July. ·
-17 -
Table 5. Population size, rate of emigration, and rate of emergence of chum
salmon fry at Slough 11 as estimated by Bailey•s Deterministic Model
using mark-recapture data collected June 5, 6, and 7, 1983.
Day 1 Marked and released 648 chum fry
Day 2 -Examined 1,081 chum fry for marks
Recaptured 227 chum fry marked on Day 1
Marked and released 854 chum fry
Day 3 -Examined 1,513 chum fry for marks
Recaptured 172 chum fry marked on Day 1
Recaptured 336 chum fry marked on Day 2
Captured 1005 unmarked chum fry
Chum fry population present at Day 2
Emigration rate of chum fry = _M_z_R_13 _____ = 0.67~
M1 ( R23 + 1)
Emergence rate of chum fry R12 (C3 + 1 ) I = -"-------= 1.8~
~ Proportion of the population on a daily basis.
-18 -
-
-
-
....
.....
-
-
-
F"
-
-
f"""
-
Table 6. Outmigration rates of chum and sockeye salmon fry from Slough 11 and
Slough 21 determined by 24 hour weir catches, 1983.
SLOUGH 11 SLOUGH 21
Date ~ Socke;t:e nate Chum Socke;t:e
May 24 1, 111 2,500 May 2l 1,996 45
~ay 25 716 2. 175 May 25 963 8
June 4 649 4. 118 May 26 I ,590 47
June 5 542 1, 623 May 27 798 44
June 6 1,083 2,466 May 28 !, 785 93
June 1,005 4,043 ~lay 29 1,851 63
MEAN 851 2,821 1,497 50
Table 7. Age separation values by length for juvenile chinook, sockeye, and
coho salmon captured over two week intervals on the Susitna River
between the Chulitna River confluence and Devil Canyon, 1983.
Total Len~::th (mm)
Survey Chinook Sockeye Coho 1~/ Period Ase 0+ Age 1+ Age 0+ Age I+ Age 0+ Age
May 1-15 ~55 :::-56 ! 55 ~56 ~ 45 ~ 46
May 16-31 ~ 65 ~ 66 ~ 60 ~ 61 ~50 ~ 51
June 1-15 ~ 70 ~ 71 ~ 65 == 66 !: 60 !! 61
June 16-30 ~ 75 2: 76 ~ 70 ~71 ~ 65 == 66
July 1-15 ~ eo ?! 81 All None ~ 70 '= 71
July 16-31 <!: 85 ~ 86 All None : 75 ~ 76
August 1-15 All None All None ~ 80 i'!: 81
August 16-31 All None All None ~ 85 ~ 86
September 1-15 All None All None ~ 90 >-91
September 16-30 All None All None ~ 95 ~ 96
!_/ Includes all coho age 1+ or older.
-19-
10
9
a: 7
:::::>
0 :r: 6
a: lt 5
:z:::
t.)
1-.q
t.)
4
3
18 2:3 2S 2 7 12 17 22 Z7 2 7 12 17 22. 27 6 11 16 21 2.6
f.--MAY JUNE JUL'l'-----"1-----AUGUST-----j
DATE
Figure 2. Chum salmon fry daily catch per hour recorded at the
downstream migrant traps, May 18 through August 20, 1983,
smoothed by Z(t}=!Y(t-l)+tY(tl+*Y(t+l)·
16.5
15.0
13.5
12.0
~ 10.15
0 :z:
a:
1.LI
ll.
::t: u
1-.q u
9.0
7.5
60
4!i
30
I .IS
0
18 23 28 2 7 12 17 22 27 2 7 12 17 22 27 6 16 21 26 31
~MAY ·-..;....---JUNE:-----1+----JULY----....._---AUGUST
Figure 3.
DATE
Sockeye salmon fry daily catch per hour recorded at the
downstream migrant traps, May 18 through August 30, 1983,
smoothed by Z(t)=!Y(t-l)+tY(t)+!Y(t+l)·
-20-
-
-
-
N .....
I
--~---~ l l
0::
:>
0
:I:
0:: w n..
:I:
()
1-<
()
15.0 -CHINO-OK, AGE tO
-·-CHINOOK 1 AGE+ I
12.5
10.0
7.5
5.0
2.5
18 26 3 II 19 27 5 13 21 29 6 14 22 30
f..-MA Y--t-1+-----JUNE -----+-1--o---------JUL Y---------o--if+----AUGU ST ---~
DATE
Figure 4. Chinook salmon age a+ and age 1+ daily catch per hour recorded
at the downstream migrant traps, May 18 through August 30,
1983, smoothed by Z(t)=!Y(t-l)+!Y(t)+!Y(t+1).
]
Catch rates for coho salmon juveniles were generally low throughout the
survey period with peaks observed during late May and early June, early
July and mid-August (Figure 5). The highest CPUE for this species was
9.6 coho per hour recorded August 11. Age 0+ fish comprised 91.6%
(5,170 fish) of the total trap captures of coho salmon juveniles while
age 1+ and older fish made up the remainder (476 fish) of the catches.
Small numbers of pink salmon fry (245 fish) were collected during May
and June in the outmigrant traps (Figure 6). The highest catch rate of
1.3 pink per hour was recorded on June 3 and the last trap capture of
pink salmon fry was recorded on July 8.
The adjusted cumulative catch rates for age 0+ salmon by species at the
outmigrant traps from ~1ay 18 through August 30, 1983 are presented in
Figure 7. This figure graphically represents the freshwater residence
·times and patterns of redistribution and outmigration for each of the
species.
3.5 Relation of Outmigration to Habitat Variables
The time series of mainstem discharge, water temperature, and turbidity
data collected during 1983 are depicted in Figure 8 and summarized in
Table 8. A summary of the juvenile salmon catch per hour statistics by
species and age class is presented in Table 9.
Adjacent daily values of discharge, water temperature, and turbidity
were closely related as shown by the high autocorrelation coefficients
in Table 8. The coefficient for discharge was slightly less than that
for the other two variables, indicating that discharge showed more day
to day variation than did temperature or turbidity.
In contrast with the habitat variables, the daily catch per hour time
series for all species and age classes showed more abrupt fluctuations.
The autocorrelation coefficients for all species by age class, with two
exceptions, ranged from 0.60 to 0.66 {Table 9). The first exception was
age 1+ sockeye salmon, which had a low coefficient of 0.43, but the
sample size was small (only 83 age 1+ sockeye salmon were captured).
The low coefficient could indicate that these fish outmigrate in sharper
pulses than do other species and age classes, perhaps because of school-
ing tendencies. The other exception was age 0+ coho salmon, which had a
higher coefficient than the other species and age classes, indicating a
more constant outmigration.
A logarithmic transformation {1og(X+1)) considerably
autocorrelation coefficients of the catch per hour time
little to improve that of the habitat variables, again
sharp fluctuations of the catch rates.
improved the
series but did
indicating the
3.5.1 Interrelationship of mainstem discharge, temperature
and turbidity
The climatic conditions (air temperature, solar insolation, and rain-
fall) which influence mainstem discharge also influence mainstem water
temperature and turbidity. Hence, these three mainstem variables were
correlated with one another.
-22 -
-
-
-
-
-
-
l
a:
:::>
0
r
a:
I w
N a_
w
I r u
1-
<! u
j 1
10 -COHO, AGE tO
9
-·-COHO, AGE+ I
6
7
6
5
4
3
2
18 23 28 2 7 12 17 2 2 27 2 7 12 17 2 2 2 7 6 It 16 2 I 26 31
!-'--MAY JUNE JULY-------o.-ii-+----AUGUST---a-1
DATE
Figure 5. Coho salmon age a+ and age 1+ or older daily catch per hour
recorded at the downstream migrant traps, May 18 through
August 30, 1983, smoothed by Z(t)=aY(t-l)+tY(t)+iY(t+l)'
N
+:>
I
] J .J
a::
::I
0
::z:::
a::
UJ a.
:I:
u
1-
<(
u
0.8
0.6
0.4
0.2
19 24 29 3 e 13 18 23 28 3 8
]'*""--MAY __ _,...--------JUN E--------1---· JULY ---j
DATE
Figure 6. Pink salmon fry daily catch per hour recorded at the
downstream migrant traps, May 18 through July 8, 1983,
smoothed by Z(t)=!Y(t-1)+!Y(t)+!Y(t+l)'
l I I .) .I I J
-
-
-
1-z w u a: w a..
w
IOO,-------------------~--~=r----~-~-~-7-~-~-~-~-~-~-~-~-=~~~~~~~
/ ~ / ,
, ' . I .... Pink'\.. / ,' "Chum / .. , ..... 'I -/ /1
I
I
/ ,.,.,
I
I
I
I
.
I
I
I
I
, ,
r
' I . ,
'
/
I
--/
/ . .......__
I Sockeye ; I
I
;I
.'I ;,
···-···-···-.. ..,...~
Discharge
>· J , , .... .,.~··-., ---1-
<t
...J
:::::>
:!!
:::::> u
I
18 23 28 Z 1 12 17 22 Z1 Z 1 IZ 17 ZZ 27 I 6 II 16 Zl 26 31
~MAY·-.~----JUNE:----~-----JULY'----~~---AUGUST----4
DATE
Figure 7. Cumulative catch for age 0+ chinook, coho, sockeye,
chum and pink salmon recorded at the downstream
migrant traps, May 18 through August 30, 1983,
adjusted to 24 hour periods.
-25-
N
"' i
liS
u
0
I.JJ Ul
a::
::l ~ a:: w a..
.::!E w
.....
"
-4aaraa
1"'\
UJ 3/alllala ....
u
v
w
(.!) Zlillillala
a::
<( :r:
(.)
(/)
1-1 1 lalllalil c
•• n
MAY
Fiqure 8 .
. I
• ·~ "" 11 '• zs: 15: l!ll:
JUNE JULY AUGUST
1983
IE 1• zo
SEPT.
61illl
3111!
>-
t-
2111!8
1-1
lO
ll!lil §5
I-
,_
-
-
I"""'
-
-
Table 8. Summary statistics for habitat variables recorded on the Susitna
River between the Chulitna River confluence and Devil Canyon,
May 18 to September 25, 1983.
Auto-
carrel-
Min Max Hean Std.Oev. n ation _n_
Discharge(ft 3 tsecl~/ 10,500 36,000 21,964 4965.5 106 0.87
Water temperature (co~/ 0.0 14.5 10.2 2.8 106 0.92
Turbidity (NTU)£/ 13 560 167 119.6 105 0.93
~ USGS provisional data at Gold Creek, 1983, 15292000.
Ef AOF&G data at Talkeetna Station downstream migrant traps, 1983.
Table 9. Summary statistics for juvenile salmon catch per hour by species
and age class recorded at the downstream migrant traps, May 18
through September 25, 1983.
Auto-
Catch per hour, carrel-
both traps Min Max Mean Std.Dev. n at ion
Chinook 0+ 0.0 21.0 1. 4 2.6 106 0.66
Chinook 1+ o.o 1.8 0.1 0.3 106 0.64
Coho 0+ 0.0 9.4 1.3 1.8 106 0.73
Coho 1+ !!_/ o.o 1.3 0. 1 0.2 106 0.60
Sockeye 0+ o.o 9.4 2.4 2.1 106 0.65
Sockeye 1+ 0.0 0.3 0.2 0.5 106 0.43
Chum o.o 16.1 2.2 J.J 106 0.65
Pink o.o l.J 0. l 0.2 105
~/ Includes all juvenile coho age 1+ or older.
-27-
104
104
104
n
104
104
104
104
104
104
87
During the four weeks following ice-out (May 18 to June 15), there was
no relationship between mainstem discharge and water temperature (Table
10). Discharge was negatively correlated with temperature during the
middle part of the season (June 16 to August 31), but positively cor-
related in September. A similar pattern was observed in 1982 when
discharge and temperature were a mirror image during the middle part of
the season (ADF&G 1983d). This pattern results from differences among
the various thermal inputs -melting ice and snow, rainwater, solar
insolation, and air temperature. Correlations were best when there was
no time lag {lag=O) between the two variables.
Correlations between mainstem discharge and turbidity were highest when
turbidity was lagged one day behind discharge {Table 10). The relation-
ship was strong during the early and late periods but the two variables
were not statistically related during the June 16 to August 30 period.
During this middle period, turbidity levels increased in late June and
decreased in late August (Figure 8), coinciding with the level of solar
insolation and the melting of glaciers. However, discharge remained at
a more constant level during the same time period as a result of ice and
snow melt in the spring and rainfall in late August. A good correlation
between discharge and turbidity resulted when the two transition times
were eliminated by shortening the time window to the period from June 25
to August 10.
3.5.2 Effects of mainstem discharge on outmigration
Correlation analysis showed that discharge is an important factor in
influencing the rate of outmigration (Table 11). This was especially
true for chum salmon, which outmigrated primarily during the two dis-
charge peaks which occurred in early June and in early July (Figure 2
and Figure 8). During the period May 18 to July 15 (by which date 98.4%
of the total season catch of chums had outmigrated) chum salmon catch
rates were strongly correlated with discharge (r = 0.89), as shown by
Figure 9.
The correlation coefficients for the other species and age classes,
except for sockeye salmon, ranged from 0.41 to 0.55. These values
suggest that discharge has an important effect on timing of salmon
outmigration. The relationships with discharge for both age classes of
chinook, coho, and sockeye salmon were strongest when the catch per hour
was compared with the discharge of the previous day. Chum and pink
salmon correlations were best when there was no lag between discharge
and catch per hour. Smoothing the daily catch per hour with the linear
filter (see Section 2.4.2) improved the correlation coefficient for all
species and age classes except for sockeye juveniles.
The correlation between trap mouth water velocity and mainstem
discharge, as recorded at the Gold Creek gaging station, was 0.37 at
Trap 1 and 0.30 at Trap 2. Comparing trap velocity with the previous
day•s discharge did not improve the correlations (the discharge lag
between the Gold Creek gaging station and the outmigrant trap is less
than one day). The correlations of discharge with trap velocity would
have been higher if the traps were fixed in place. However, the traps
-28 -
-
-
-
-
-
-
-
~'
-
·-
Table 10. Correlation coefficients between discharge and temperature, and
discharge and turbidity, for the Susitna River between the Chulitna
River confluence and Devil Canyon, 1983. The data were not smoothed.
Correlation Significance Sample
Variables Period Coefficient{r} Level Size
Discharge/temperature May 18-Jun 15 0.07 NsY 29
Jun 16-Aug 31 -0.40 0.01 77
Sep 01-Sep 25 0.53 0.01 25
May 18-Sep 25 0.39 0.01 131
Discharge(t-1 )/turbidity May 18-Jun 15 0.95 0.01 27
Jun 16-Aug 31 0.04 NS 76
Sep 01-Sep 25 0.86 0.01 12
May 18-Sep 25 0.38 0.01 115
a/ NS = Not significant
-29 -
Table 11. Correlation coefficients between discharge and juvenile salmon
catch per hour by species and age class for the Susitna River
between the Chulitna River confluence and Devil Canyon, May 18
through August 30, 1983. Both discharge and catch per hour were
smoothed by the linear filter: Z(t) = !Y(t-1 ) + iY(t) + !Y(t+1).
Discharge(t-1)/ Significance
catch per hour, Correlation Level Sample
both traes Coefficient ( r) {p) Size
Chinook age 0+ 0.50 0.01 102
Chinook age 1+ 0.44 0.01 102
Coho age 0+ OA1 0.01 102
Coho age 1+ 0.47 0.01 102
Sockeye age 0+ 0.34 0.01 102
Sockeye age 1+ 0.24 0.01 102
Discharge/
catch per hour
both traps
ChumV 0.89 0.01 57
Pink~/ 0.55 0.01 54
~Sampling dates -May 18 through July 15, 1983.
-30 -
~'
-
-I
-
!II'W
-
~
-
,400;j
'I
""""!
-
-
-
-
-
-!
-
were moved closer to shore as mainstem discharge increased in order to
maintain that range of velocities through the traps which minimized
mortality. Although a rise in mainstem discharge did increase the trap
mouth water velocity, correlations between trap velocity and the catch
per hour of age 0+ salmon for most species/trap combinations were 1 ow
and not statistically significant. This indicates that the relationship
shown in Figure 9 is not simply a function of fishing a greater volume
of water at the higher discharge levels. In contrast, the catch per
hour of age 1+ chinook, coho, and sockeye salmon juveniles was
positively correlated with trap mouth water velocity. This may relate
to trap avoidance by the larger fish and is discussed further in
Appendix A.
The discharge/catch per hour correlations for chinook, coho, and sockeye
were calculated for the entire season and those far chum and pink were
calculated from mid-May to mid-July. The relationship during shorter
time periods than these was stronger, as is graphically demonstrated in
Figure 7. Inflections in the cumulative discharge curve correspond to
inflections in the cumulative catch· curves. During the early August
discharge peak (Figure 8), there were few chum or pink juveniles left in
the reach; the three remaining species all responded to the discharge
increase. Only age 0+ chinook fry responded to the late August dis-
charge peak.
-31 -
0::
::::>
0
J:
0::: w
0..
J:
(.)
1-
<(
u
z
0
:E
_J
<(
I C/)
w
N ::E I
::::>
:r;
u
.I I -J
12.Q CPH • 0.606 ( Dischar<;Je x 10-3 )-9.841
r = o. 8 9
~ <0.0 I
10.0 • •
•
8.0
••
• •
6.0 •• • •
4.0 • •• • • • • 2.0
••
0.0~--~--~----~--~--~----~----~--~----~..--.--~----~----~
'14 16 18 20 22 24 26 28 30 3 2 34 3 6 3 8
SUSITNA RIVER DISCHARGE (ft 3/ sec x 10-3 )
Figure 9. Relationship of mean daily discharge with mean daily chum
salmon fry catch per hour at the downstream migrant traps, May
18 through July 15, 1983.
J 1 .. J .. J .J .. 1 __ .. I J -.I J J .]
-
.....
-
-
r
4.0 DISCUSSION
4.1 Coded Wire Tagging and Recovery
Coded wire tagging has been used primarily as a tool to mark salmon
smelts prior to their entrance into the marine environment by programs
emphasizing the return of adults. The objectives of these programs have
been to determine the contribution and timing of specific stocks such as
hatchery releases to the overall return of adults to a commercial
fishery, or to determine the success of various timings of hatchery
smo lt re 1 eases.
The program conducted on the Susitna River during 1983 was a unique
use of coded wire tag methodology. This was the first study to use
coded wire tags to mark post-emergent salmon fry in the field rather
than under cantrall ed hatchery conditions, and was a 1 so the first to use
the tags on the small size of fish observed during this study. The
sockeye salmon fry were a minimum length of 27 mm total length and
averaged up to 3,000 fish per pound. ·
The objectives of the 1983 program were to quantify the populations and
survival rates of outmigrating chum and sockeye salmon fry rather than
determining their contributions to the total number of returning adults.
Although not an integral part of this study, adult recovery by
fishwheels and spawning ground surveys would be useful in determining
rates of marine survival and is still very much a possibility but is
dependent on future program funding.
Coded wire tagging provided a mark-recovery method which could be
successfully incorporated with the current fisheries investigations on
the Susitna River. However, for the methods to be useful in providing
valid estimates of outmigrant populations and egg-to-outmigrant survival
rates, certain assumptions had to be met.
First, neither mortality rates nor catchability should vary between
marked and unmarked fish. Previous studies such as Hagar and Jewel
(1968), Jefferts et al. (1963) and Opdycke and Zajac (1981) and have
shown that marking juvenile salmon with coded wire tags does not affect
mortality or catchability.
Secondly, tag retention rates must not vary significantly between
tagging and recovery. This assumption was met during 1983 as tag
retention rates averaged 97.7% for chum salmon fry at release and were
96.9% during recovery efforts. Sockeye salmon ta·g retention rates were
96.3% at release and 95.4% during trap recovery.
A third assumption was that the marked fish were randomly distributed
within the total outmigrant population at the point of recovery. A
camp a rison of the numbers of rna rked-to-unma rked fish captured at the
traps showed that this assumption was valid. Although the traps were
fished on opposite banks of the river, the ratios of recovery of tagged
versus ~ntagged fish at each trap were essentially the same.
-33 -
The fourth assumption was that all marks were recognized and reported
during recovery. The efficiency of the field sampling detector to
detect the tags and the test of fin clip efficiency showed that all
tagged fish were recogni zab 1 e during the recovery efforts.
The combined mortality rate of 1.2% recorded for chum and sockeye salmon
fry during the coded wire tagging procedures was not entirely due to the
implantation procedures. Two-thirds of the mortalities were a direct
result of handling stress or decreased oxygen levels during capture, or
over-exposure to the anesthetic solution. The mortalities related
directly to the coded wire tag implementation procedures averaged 0.4%
over all the sampling sites.
Although the tagging of small fish worked well for this study, applica-
tion of tl}ese methods to other programs, especially when emphasizing
adult returns, should be done cautiously. Our program covers only one
season of data and does not provide information concerning changes in
tag retention and mortality rates which may occur during the period of
marine residence.
4.2 Dye Marking and Outmigration Rates
The dye marking experiments showed the period of dye retention ranged
from 12 hours to five days after marking. Most of the dye had faded
within 24 hours but was visible on the fins and lower jaw for longer
periods. The fish were under stress during the period of dye immersion
as shown by the continued gulping of air, flashing, and darting of the
fish, but mortality rates were less than one percent. Marking with
Bismark Brown dye is effective for short-term marking experiments in
which detection is necessary for only a few days, but would not provide
an adequate mark for studies extending over longer periods.
The mark-recapture experiment conducted on chum salmon fry at Slough 11
(Section 3.3} demonstrated the possibility of estimating outmigrant
rates and populations at specific sites on the Susitna River. This
study was time consuming due to the problem of distinguishing dyed fish
from coded wire tagged fish which had also been dyed. The use of more
distinct marks to delineate groups of fish would minimize this problem.
It would be beneficial to conduct these outmigrant estimates during the
1984 sampling program at numerous study sites over the entire period of
outmigration. These data would provide a comparison of outmigration
rates by study site and, when compared to the habitat variables recorded
at each site, the factors influencing outmigration could be more clearly
determined.
Survival rates could also then be generated for each site using the
adult spawner counts recorded during the previous season. By comparing
these survival rates to the habitat parameters recorded at each site
during the period of incubation and emergence, the environmental factors
affecting the egg-to-outmigrant survival could also be more clearly
defined.
-34 -
-
-
-
-
-
-
-
,r"""
-
-
-
The above data when used in conjunction with trap population estimates
and survival rates could ultimately be used to determine the contribu-
tion which an individual site or macrohabitat type makes to the total
production of juvenile salmon from the reach of river between the
Chulitna River confluence and Devil Canyon. This would provide weighted
values for each habitat type for use in project flow mitigation.
4.3 Survival of Outmigrants
The survival rates of 12.9 to 14.1 percent estimated for Susitna River
chum salmon from potential egg deposition to outmigration are similar to
the rates reported for chum salmon survival in other systems. Neave
(1948) reported chum salmon freshwater surviva.l rates as 1 ow as 0.4
percent while Beacham and Starr (1982) observed chum survival to be as
high as 35.4 percent. Hunter (1959} recorded survival rates from 1.0 to
19.4% over a ten year period for chum salmon in a small coastal stream
in British Columbia.
Sockeye salmon egg-to-outmigrant survival rates are more difficult to·
determine due to the more complicated freshwater life history for this
species. While chum salmon are strictly age 0+ outmigrants, most
sockeye juveniles spend one to two winters in freshwater before outmi-
grating. Thus, the survival calculations for the period of freshwater
residence for sockeye must be made for two or more age classes of
outmigrants.
Most previous studies have reported survival rates for sockeye salmon
associated with lake systems. In such systems, spawning occurs along
the lake shore and in the inlet and outlet streams. Following emer-
gence, the sockeye fry enter the lake, first feeding along the shoreline
and later entering the pelagic areas to rear and overwinter (McCart
1967}. Outmigrating sockeye smelts are then enumerated as they move
through the outlet stream to the ocean. Survival rates reported for
these sockeye salmon stocks during the period from egg deposition to
outmigration as age 1+ and age 2+ smelts have ranged from 0.6 percent
(Russell 1972) to 8.5 percent (Meehan 1966}.
In large river systems such as the reach of the Susitna between the
Chulitna River confluence and Devil Canyon, the sockeye salmon spawn in
sloughs and side channels and, following emergence, the fry rear in
these areas and the mainstem river. A major portion of the sockeye
salmon juveniles in this reach migrate as young-of-the-year fish to
areas located below the Chulitna River confluence. It was for the
period from egg deposition through this emigration of age 0+ fish out of
the study reach that survival rates of 40.9 to 42.0% were determined for
Susitna River sockeye. Thus, the high survival rates determined for
Susitna River sockeye cover a shorter period of the life cycle and are
not comparable to other studies which have determined survival rates
through the entire period of freshwater residence.
The survival rates recorded for the Susitna River do, however, provide
an indication of the relative productivity of various salmon ·spawning
habitats used in the study reach. The accuracy of the survival rate
estimates is dependent upon the accuracy of the adult escapement counts,
-35 -
by the lower survival rates observed for chum salmon compared to sockeye
salmon for the same period of their life cycles are probably a result of
the habitat conditions present at the spawning and incubation sites for
each species. The sockeye salmon in the study reach spawn almost
exclusively in sloughs associated with the mainstem river and the high
observed survival rates for this species are primarily a result of the
productivity of these sloughs. Chum salmon spawning occurs in the
tributaries and sloughs, and the survival to outmigrating fry is
determined by the habitat conditions present at a broader range of
sites.
Previous studies have shown that natural survival of salmon between the
periods of egg deposition and the time of smolt emigration to the ocean
is highly variable. and is dependent on numerous conditions present in
the freshwater environment (Wickett 1958; Hunter 1959). Most mortal-
ities of salmon occur during this critical period of their life cycle
and often have the most profound effect on the numbers of returning
adults (Henry 1953). ·
The discrepancy between survival in tributaries and in the side sloughs,
as suggested by the differences in egg to outmigrant survival of sockeye
and chum salmon, suggests an approach to understand the importance of
environmental factors in influencing survival. An examination of the
critical habitat components during spawning and incubation at the major
tributaries, compared with the sloughs, should suggest the habitat
variables that are responsible for these differences. Those factors
most apparently different, and that are the subject of other investiga-
tions by ADF&G, include:
0 Access of adults to sloughs as a function of mainstem flows.
o Winter ground water flows and the prevention of freezing.
o Adverse effects of temperature on development and survival
caused by ice processes which lead to overtopping of sloughs.
0 Density-dependent mortality because of redd superimposition at
both sloughs and tributaries (affected by access or brood year
survival).
o Inter-specific competition for redds (chinook, pink, and coho
spawn in streams near chum spawning areas).
0 Spawning occurs during high flow periods and redds are
deposited at areas that are subsequently dewatered and frozen.
All of the factors listed, except for species composition, are affected
by mainstem discharge and consequently may be affected, either
beneficially or negatively, by flow regulation of the Susitna River.
4.4 Comparison of Trap Catch Rates
A comparison of catch rates of juvenile chum and sockeye salmon collect-
ed in the two downstream migrant traps during 1983 showed that catches
-36 -
-~
-
-J
-
-
-
-
-
-I
-
....
were not proportional to population size for the two species. Chum
salmon comprised only 41 percent of the total captures of both species
at the traps~ while population estimates from the coded wire tagging
program indicated that almost six times as many chum salmon fry migrated
past the traps during 1983. This trap selectivity observed for sockeye
and chum fry is probably due to the difference in migration patterns
between the two species. Chum salmon fry migrate primarily near the
water surface and in the center of the channel where water velocity is
greatest (Hunter 1959). McCart (1967) observed that downstream migrat-
ing sockeye fry were associated with the river banks during the
migration.
As the east bank trap (Talkeetna Station~ RM 103) was fished during both
1982 and 1983~ we compared the catch rates at this trap between the two
years for juvenile salmon collected during the same calendar dates.
Chinook, coho, and chum salmon catch rates indicate relative abundances
were related to the estimated populations of parent spawners at Curry
Station. Chum salmon fry catch rates at the east bank trap for the
period from June 18 through August 15 averaged 0.7 fish per hour during
1982 and 1.6 fish per hour (2.3 times as high) during 1983. The parent
spawners estimated for the 1983 outmigrant population were 2.3 times the
number of estimated parent spawners for the 1982 outmi grants (ADF&G
1983a). A comparison of east bank trap catch rates for juvenile chinook
and coho salmon captured between June 18 and August 30 to the estimated
number of parent spawners showed simi 1 ar results. Adult coho salmon
were estimated to be 2.1 times as abundant in 1982 as 1981 and the trap
catch rates were 2.8 times as high in 1983 than in 1982. Although no
population estimates were provided for adult chinook salmon during 1981,
it appears that the spawning escapement was much smaller than that
observed during 1982 (Bruce Barrett, personal communication). Trap
catch rates of juvenile chinook salmon were over four times as great in
1983 than for the same calendar period in 1982. These data indicate
that the traps provide a comparative index of annual differences in the
relative abundance of outmigrants.
East bank trap catch rates for sockeye salmon juveniles during 1983 were
1.4 times higher than the rates recorded during the same calendar period
in 1982. Conversely~ the estimates of sockeye parent spawners at Curry
Station during 1982 were less than half the estimated number past this
site in 1981. As the sockeye salmon in the study reach spawn only in
the sloughs, the discrepancy between catch rates for this species is
probably caused by the en vi ronmenta 1 factors previously 1 i sted, with the
most like causes being: (l)The large number of adult sockeye observed
during 1981 may have resulted in the superimposition of redds and a
density-dependent mortality of eggs. (2)The 1981 spawn·ing occurred
during a period of high flows, and as winter progressed, many of the
redds may have dewatered and frozen during this low flow period
resulting in high mortalities of the incubating eggs.
The survival rates of 1982 brood year sockeye salmon (1,261 adults) from
egg deposition to fry outmigration determined during 1983 were very high
(over 40%). During years of high adult escapement such as 1981 (2,804
adults), the number of eggs deposited may exceed the productive capacity
-37 -
of·the spawning sloughs and result in lower survival rates. Conversely
chum, coho., and chinook salmon spawn primarily or entirely in the
tributaries which are capable of sustaining much larger spawning escape-
ments because of the larger amount of available habitat.
These data and the comparisons of sockeye and chum salmon fry catch
rates at the traps show that although the outmigrant traps can provide
an index of relative abundance, they are selective and cannot be used to
accurately determine outmigrant population estimates without the inclu-
sion of a mark-recovery program. Trap selectivity also influenced the
catch rates of age 1+ salmon juveniles (Appendix A). Transect sub-
sampling as a mechanism to apportion catches would assist in quantifying
the extent of trap selectivity.
A comparison of the cumulative catch rates adjusted to 24 hour periods
for the east bank trap for the same calendar periods during 1982 and
1983 (June 18 through August 30) showed similar patterns of chum and
sockeye outmigration for the two open water periods. Over 90 percent of
the chums were captured by July 15 during both years and their out-
migration from the study reach was completed by the middle of August
(Figure 10). Sockeye salmon juveniles showed an initial pulse of
downstream movement during 1 ate June and early July, but the emi gra-
tional redistribution of this species continued throughout the open
water period during both 1982 and 1983 (Figure 10).
Cumulative catch rates for chinook and coho salmon juveniles at the east
bank trap were not as similar during the two sampling seasons. Both
species showed more even patterns of outmigration during 1982 than in
1983 (Figure 10). Trap catch rates for juvenile chinook and coho salmon
were low during July and early August of 1983 and then dramatically
increased beginning on August 10. This corresponds to an increase in
mainstem discharge from less than 23,000 cfs during July to a peak of
32,000 cfs on August 10. July was also a period of low flows in the
primary chinook and coho salmon spawning tributaries (Indian River and
Portage Creek), but during early August, significant increases in water
levels were recorded for both streams (Report Series 3, Part 1).
The observed high catch rates of juvenile chinook and coho salmon
recorded at the outmigrant traps after early August are a result of two
factors: (1) Rearing juveniles in Indian River and Portage Creek may
have been trapped in side channels and pools and were unable to emigrate
to the mainstem river until the high flow periods in early August. This
situation was recorded on August 3, when hundreds of juvenile chinook
and coho salmon trapped in small pools were observed in Indian River,
and (2) The abrupt increase in tributary and mainstem discharge during
this period and the subsequent extensive breaching of mainstem rearing
areas caused a flushing and downstream displacement of rearing chinook
and coho salmon.
As shown in Figure 10, less than 50 percent of the adjusted cumulative
catches of chinook and coho salmon juveniles was recorded between June
18 and August 9, and the remaining captures occurred between August 10
and August 30. These data indicate that chinook and coho salmon were
-38 -
-
-
-
-
!"""
-f-z
L1J
(.)
a::
L1J a..
L1J
>
f-<t
_J
:::>
::E :::>
(.)
.-
-
~
~
~
-
' ' , , ,
7 I
,-
' I
, ' ~ ' 50 I
~-2 , ,
' ,
40 , 4
I
I
I
30 I
I
I
20 COHO CHUM
10
0
18 28 8 18 2.8 7 17 2.7 18 28 8 18 2.8 7 17 27
100 10
,-, ,
90 , , 90 . .
' eo ' eo ' . -, ,
70 , ,
, . . .
I
60 ' ,
!
' 50 ' ' .
40
30
20 CHINOOK SOCKEYE
10
0
18 28 8 18 28 7 17 2.7 18 2.8 8 18 28 7 17 27
~JUNE~JULY AUGUST--+i t-.JUN~.JULY AUGUST ----1
DATE
Figure 10. East bank outmigrant trap (Talkeetna Station, RM 103.0)
cumulative catch recorded for juvenile coho, chinook, chum,
sockeye salmon during 1982 and 1983, adjusted to 24 hour per
for the calendar period from June 18 through August 30.
-39-
still predominantly in the natal tributaries or in mainstem habitats
above the traps until the high flow period in August. Studies of
juvenile salmon outmigration at the major spawning tributaries would be
valuable in determining the residence time and growth of juvenile salmon
at habitats associated with the mainstem Susitna.
4.5 Relation of Outmigration to Habitat Variable
Discharge was an important facto.r influencing the timing and rate of
outmigration of juvenile salmon during 1983. Chum salmon outmigration
showed the highest correlation to discharge (Section 3.5.2). Calcu-
lations were made for the entire sampling season but higher correlations
exist between discharge and outmigration when analyzed during short
periods of time. High catch rates for chinook, coho and sockeye juve-
niles recorded during the middle of August, for example, coincided with
a period of high discharge in the mainstem river and major tributaries
(Figures 8 and 10). Similarly,-catch per unit effort peaks for chinook
and chum fry in the Skagit River coincided with peaks in river discharge
(Congleton et al. 1981).
Raymond {1968) showed that lower migration rates occurred during periods
of low discharge than at moderate discharge levels. Adequate river
stage is necessary at the sloughs to allow the outmigrating juveniles
access to the Susitna River mainstem. An increase in migration time
required for juveniles to reach their marine rearing areas may result in
increased predation and a decreased ability of the migrants to make the
transition to salt water (Andrew and Geen 1960; Foerster 1968).
Water temperatures at the emergence and rearing areas are also an
important factor in triggering outmigration. (Foerster 1937, 1968)
found that outmigration of sockeye in lakes begins as temperatures rise
above a minimum level during the spring {4.4 to 5.0°C) and may cease
during the summer if temperatures become unacceptably high {13.0°C)
Mihara (1958, cited by Bakkala 1970) found that in streams in Hokkaido,
Japan, chum fry changed from a positive rheotaxis to a negative
rheotaxis and moved quickly downstream when the water temperature
reached 15°C. This was interpreted as an adaptive response to avoid the
high summer stream temperatures. Similar results have been demonstrated
by Keenleyside and Hoar (1955). Unseasonably high winter and spring
water temperatures resulting from dam operation could trigger juvenile
salmon outmigration before optimum downstream and marine habitat con-
ditions are present (McCart et al. 1980).
Turbidity is an important factor in providing cover to outmigrating
salmon in large rivers such as the Susitna. Andrew and Geen {1960)
suggested that reduced sediment loads (turbidity) might expose migrating
juveniles to abnormally high predation levels. It can be speculated
that an increase in turbidity occurring when the heads of natal sloughs
are overtopped by a rising mainstem discharge could induce juveniles to
leave the object cover available in the slough and move to the mainstem.
The correlations of mainstem temperature and turbidity with the daily
catch per hour of juvenile salmon were generally low during 1983. This
does not mean that these two variables are not important factors in
-40 -
-
-
-
-
-
-
~
I
-
-
-
F"~
, I
influencing outmigration but, rather, reflects the fact that the temper-
ature and turbidity data were taken at the same location as the
outmigrant traps. It is likely that the major effect of the variables
as outmigrant stimuli would occur at the rearing areas.
In summary, the time between egg deposition and outmigration is the most
critical period in the life history of salmon populations (Henry 1953),
and ultimately it has the greatest effect on the numbers of adult fish
returning to the commercial and sport fisheries, and the spawning
grounds. The development of population estimates for chum and sockeye
salmon has allowed estimates of the survival of these species from egg
to outmigration. These differences suggest that slough spawners, if
they have an opportunity to deposit eggs, have a high probability of
producing viable fry and may contribute proportionately more offspring
than their counterparts spawning in the tributaries. This is probably
because slough discharge during the winter is more stable because of the
large groundwater influences. The strong correlation of outmigration
with short term discharge peaks suggests discharge changes can be
expected to affect the rearing in mainstem habitats and the successful
outmigration of smolts. High flows at the proper period (late May and
early June) could stimulate outmigration of smolts to ensure minimal
freshwater mortality. Similar events in later summer could possibly be
detrimental as rearing 0+ fish might be displaced from habitat upstream
(Hartman et al. 1982). If optimum habitat were maintained by flows
after the fish were displaced, the benefits would be reduced because of
the previous downstream displacement of the population.
-41 -
5.0 CONTRIBUTORS
We would like to extend special thanks to the following individuals for
their assistance with this study:
Fisheries Technicians Albert Badgley and Patricia Harris, and
Fishery Biologists James Mauney and James Quinn for their assis-
tance with the operation of the downstream migrant traps.
James Patrick, Fishery Technician; Dan Sharp, Fishery Biologist;
and Linda Soquet and Aimee Weseman, Fish Culturists, for their
assistance with the coded wire tagging efforts.
Stephen Hale for his assistance in the statistical analyses and
preparation of the information on the influence of habitat
variables.
Bruce Barrett, Adult Anadromous Studies Project Leader, for his
cooperation in obtaining personnel and camp equipment during the
coded wire tagging program and for reviewing the draft and
providing helpful suggestions.
Allen Bingham and staff, especially Kathrin Zosel and Gail
Heineman, for data processing services and providing review
comments on the draft; Skeers Word Processing Services for typing
this report; and Sally Donovan and Carol Kerkvliet for drafting the
figures.
-42 -
~:
-
-
-
-
-
-
-
-
-
6.0 ACKNOWLEDGEMENTS
Funding for this study was provided by the State of Alaska, Alaska Power
Authority.
We would like to express our gratitude to all the people and organiza-
tions that provided information or assistance to the outmigrant studies
during the past season.
We are especially grateful to Timothy Joyce, David Gaither, and Rick
Krolick, Fish Culturists, and the personnel at Kitoi Bay Hatchery for
their patient instruction on coded wire tagging procedures, techniques,
and trouble shooting.
We are also grateful to Dr. Keith Jefferts and the personnel at North-
west Marine Technologies, of Shaw Island, Washington, for their timely
assistance in obtaining and operating the coded wire tagging equipment.
We wish to thank the various staff at E. Woody Trihey and Associates
(EWT&A), the Arctic Environmental Information Data Center, Harza-Ebasco
Susitna Joint Venture, Woodward-Clyde Consultants, the Alaska Power
Authority, and especially Cleve Steward (EWT&A) for reviewing the draft
of this paper and providing helpful suggestions.
We would like to express our appreciation for the logistical services
provided by Air Logistics.
We waul d a 1 so 1 ike to extend our thanks to the 1 oca 1 residents and
property owners of Talkeetna~ and especially to Band K Trading Post,
Gene and Rose Jenne of Three Rivers Union, and Bill Blakely.
-43 -
7.0 . LITERATURE CITED
Alaska Department of Fish and Game (ADF&G). 1981. Resident fish
investigations on the lower Susitna River. Phase I. Final Draft.
Prepared for Acres American, Inc., by Alaska Department of Fish and
Game/Susitna Hydro Aquatic Studies Program. Anchorage, Alaska.
1983a. Adult anadromous fish studies, 1982. Volume 2 of Phase
II Final Report. Alaska Department of Fish and Game/Susitna Hydro
Aquatic Studies Program. Anchorage, Alaska.
• 1983b. Resident and juvenile anadromous fish studies on the
--s·usitna River below Devil Canyon, 1982. Volume 3 of Phase II Basic
Data Report. Alaska Department of Fish and Game/Susitna Hydro
Aquatic Studies Program. Anchorage, Alaska.
. 1983c. Winter aquatic studies {October, 1982 -May, 1983).
--Phase II Data Report. A 1 aska Department of Fish and Game/Susi tna
Hydro Aquatic Studies Program. Anchorage, Alaska.
• 1983d. Synopsis of the 1982 aquatic studies and analysis of
--fish and habitat relationships, Appendices. Phase II Report.
Alaska Department of Fish and Game/Susitna Hydro Aquatic Studies
Program. Anchorage, Alaska.
• 1984. Aquatic studies procedures manual. Phase III. Final
--·Draft. Alaska Department of Fish and Game/Susitna Hydro Aquatic
Studies Program. Anchorage, Alaska.
Andrew, F.J., and G.H. Geen. 1960. Sockeye and pink salmon production
in relation to proposed dams in the Fraser River system. Interna-
tional Pacific Salmon Fisheries Commission Bulletin 11: 259 pp.
Bailey, N.J.J. 1951. On estimating the size of mobile populations from
recapture data. Biometrika 38: 293-306.
Bakkala, R.G. 1970. Synopsis of biological data on the chum salmon
Oncorhynchus keta (Walbaum) 1792. FAO Species Synopsis No. 41.
United States--,;rsh and Wildlife Service, Bureau of Commercial
Fisheries Circular 315. Washington, D.C. 89 pp.
Barnaby, J. T. 1944.
Oncorhynchus nerka
States Fish and
237-295.
Fluctuations in abundance of red
(Walbaum), of the Karluk River, Alaska.
Wildlife Service, Fishery Bulletin,
salmon,
United
50(39):
Barrett, B.M. 1974. An assessment study of the anadromous fish
populations in the upper Susitna River watershed between Devil
Canyon and the Chulitna River. Cook Inlet Data Report No. 74-2.
Alaska Department of Fish and Game. Division of Commercial
Fisheries 56 pp.
-44 -
-
-
-
-
-
-
-
-
-I
Barrett, B.M., F.M. Thompson, and S.N. Wick. 1984. Adult anadromous
fish studies: May-October 1983. Alaska Department of Fish and
Game Susitna Hydro Aquatic Studies Report No. 1. Prepared for
Alaska Power Authority. Anchorage, Alaska.
Beacham, T.D., and P.Starr. 1982. Population biology of chum salmon,
Oncorhynchus keta, from the Fraser River, British Columbia.
Fishery Bulletin 80(4): 813-825.
Burgner, R.L. 1962. Studies of red salmon smolts from the Wood River
Lakes, Alaska. University of Washington Publication Fisheries {New
York) 1: 247-314.
Chapman, D.G. 1951. Some properties of the hypergeometric distribution
with applications to zoological sample censuses. University of
California Publication Statistics 1: 131-160.
Chapman, D.W. 1962. Effects of logging upon fish resources of the west
coast. Journa 1 of Forestry 60(8): 533-537.
Congleton, J.L., S.K. Daviss and S.R. Foley. 1981. Distribution,
abundance and outmigration timing of chum and chinook salmon fry in
the Skagit salt marsh. Pages 153 -163 in E.L. Brannon and E.O.
Salo, editors. Preceedings of the Salmon and Trout Migratory
Behavior Symposium. University of Washington, Seattle, Washington.
Delaney, K.J., K. Hepler, and K. Roth. 1981. Deshka River chinook and
coho salmon study. Alaska Department of Fish and Game, Division of
Sport Fish. Federal Aid in Fish Restoration, Project AFS-49, Vol.
22.
Dixon, W.J., M.B. Brown, L. Engelman, J.W. Frane, M.A. Hill, R.I.
Jennrich, and J.D. Toporek (edits.). 1981. BMDP Statistical
Software 1981. University of California. Berkley, California.
725 pp.
Engel, L.J. and D.A. Watsjold. 1978. Fisheries studies prepared for
the Alaska Capital Site Planning Commission. Alaska Department of
Fish and Game, Division of Sport Fish 122 pp.
Foersters R.E. 1937. The relation of temperature to the seaward
migration of young sockeye salmon (Oncorhynchus nerka). Journal of
the Biological Board of Canada 3{5): 421-438 •
• 1968. The sockeye salmon, Oncorhynchus nerka. Bulletin of the
--,Fisheries Research Board of Canada 162.
Friese, N.V. 1975. Preauthorization assessment of anadromous fish
populations of the upper Susitna River watershed in the vicinity of
the proposed Devil Canyon hydroelectric project. Cook Inlet Data
Report No. 75-2. Alaska Department of Fish and Game. Division of
Commercial Fisheries 121 pp.
-45 -
Gangmark, H.A., and R.D. Broad. 1956. Further observations on stream
survival of king salmon spawn. California Fish and Game 42:
37-49.
Hager, R.C., and E.D. Jewell. 1968. Field evaluation of coded wire tag
detection and recovery techniques. Washington Department of
Fisheries, Fisheries Division, Marine Fisheries Research Manage-
ment, Olympia 20 pp.
Hartman, G.F., B.C. Anderson, and J.C. Serivener. 1982.
movement of coho salmon (Oncorhynchus kisutch) fry in
Creek, an unstable coastal stream in British Columbia.
Journal of Fisheries and Aquatic Sciences. 39: 588-597.
Seaward
Carnation
Canadian
Hartman, W.L., W.R. Heard, and B. Drucker. 1967. Migratory behavior of
sockeye salmon fry and smolts. Journal of the Fisheries Research
Board of Canada 24(10): 2069-2099.
Heard, W.H. 1964. Phototactic behavior of emerging sockeye salmon fry.
Animal Behavior 12(2): 382-388.
Henry, K.A. 1953. Analysis of factors affecting the production of chum
salmon (Oncorhynchus keta) in Tillamook Bay. Fisheries Commission
of Oregon. Contribution No. 18. 37 pp.
Hunter, J.G. 1959. Survival and production of pink and chum salmon in a
coastal stream. Journal of the Fisheries Research Board of Canada.
16(6): 835-886.
Jefferts, K.B., P.K. Bergman, and H. Fiscus. · 1963. A coded wire
identification system for macro-organisms. Nature {London) 198:
460-462.
Keenleyside, M., H.A. and W.S. Hoar. 1955. Effects of temperature on
the responses of young salmon to water currents. Behavior 7{2/3):
77-87.
Koerner, J.F. 1977. The use of the coded wire tag injector under
remote field conditions. Alaska Department of Fish and Game,
Informational Leaflet No. 172. 24 pp.
Kubik, S., and R.D. Wadman. 1978. Inventory and cataloging of sport
fish waters of the lower Susitna River and central Cook Inlet
drainages. Alaska Department of Fish and Game. Federal Aid in
Fish Restoration, Annual Report of Progress, 1978-1979, Project
F-9-11, 20(G-I-H). 21 pp.
Lux, F.E. 1971. Age determination of fishes (revised). National
Marine Fisheries Service, Seattle, Washington. Fishery Leaflet No.
637. 7 pp.
-46 -
-
-
-
-
-
-
~
'
I
i
·-'
I
i
""""'
!
; !
McCart, P. 1967. Behavior and ecology of sockeye salmon fry in the
Sabine River. Journal of the Fisheries Research Board of Canada
24(~): 375-428.
McCart, P.J., D.W. Mayhood, M.L. Jones, and G.J. Glova. 1980.
Stikine-Iskut Fisheries Studies, 1979. P. McCart Biological
Consultants Ltd. Nanaimo, British Columbia. A report to British
Columbia Hydro and Power Authority. April, 1980.
McConnell, R.J., and G.R. Snyder. 1972. Key to field identification of
anadromous juvenile salmonids in the Pacific Northwest. National
Oceanic and Atmospheric Administration Technical Report, National Ma
Fisheries Service CIRC-366. 6 pp.
McDonald, J. 1960. The behavior of Pacific salmon fry during their
downstream migration to freshwater and saltwater nursery areas.
Journal of the Fisheries Research Board of Canada 17(5): 655-676.
McNeil, W.J. 1966. Effect of the spawning bed environment on
reproduction of pink and chum salmon. United States Fish and
Wildlife Service, Fisheries Bulletin 65: 495-523.
Meehan, W.R. 1966. Growth and survival of sockeye salmon introduced
into Ruth Lake after removal of resident fish populations. United
States Fish and Wildlife Service, Special Scientific Report
Fisheries 532, 18 pp.
Moberly, S.Aq R. Miller, K. Crandall, and S. Bates. 1977. Mark-tag
manual for salmon. Alaska Department of Fish and Game. Fisheries
Rehabilitation and Enhancement Division. 56 pp.
Morrow, J.E. 1980. The freshwater fishes of Alaska. Alaska Northwest
Publishing Company, Anchorage, Alaska. 248 pp.
Mosher, K.H. 1969. Identification of -Pacific salmon and steelhead
trout by scale characteristics. United States Fish and Wildlife
Service, Bureau of Commercial Fisheries Circular 317. 16 pp.
Neave, F. 1948. Fecundity and mortality in Pacific salmon. Royal
Society of Canada Transaction Section 5, Serial 3 42: 97-105.
--• 1953. Principles affecting the size of pink and chum salmon.
populations in British Columbia. Journal of the Fisheries Research
Board of Canada 9(9): 450-491.
Opdycke, J.D., and D.P. Zajac. 1981. Evaluation of half-length binary-
coded wire tag application in juveni 1 e chum salmon. Progressive
Fish-Culturist 43(1): 48.
Raymond, H.L.
relation
Rivers.
356-359.
1968. Migration rates of yearling chinook salmon in
to flows and impoundments in the Columbia and. Snake
Transactions of the American Fisheries Society 97(4):
-47 -
1-
Ricker, W.E. 1975. Computation and interpretation of biological
statistics of fish populations. Fisheries Research Board of
Canada, Ottawa Bulletin 191. 382 pp.
Riis, J.C., and N.V. Friese. 1978. Preliminary environmental
assessment of hydroelectric development on the Susitna River.
Alaska Department of Fish and Game. Division of Sport Fish and
Commercial Fisheries 116 pp.
Robson, D.S., and H.A. Regier. 1964. Sample size in Petersen mark-
recapture experiments. Transactions of the ·American Fisheries
Society 93(3): 215-226.
Roos, J.F. MS, 1958. Red salmon smolt studies at Chignik in 1958.
Fisheries Research Institute, University of Washington 17 pp.
Russell, P.A. 1972. Frazer Lake sockeye investigations, 1970. Alaska
Department of Fish and Game. Division of Commercial Fisheries
Informational Leaflet No. 159. 83 pp.
Schaefer, M.B. 1951. Estimation of the size of animal populations by
marking experiments. United States Fish and Wildlife Service,
Fisheries Bulletin 52: 189-203.
Thompson, S.H. 1964. The red salmon (Oncorhynchus nerka) of Copper
River, Alaska. United States Fish and Wildlife Service, Auke Bay
Manuscript Report, 64-12.
Trautman, M.B. 1973. A guide to the collection and identificQtion of
presmolt Pacific salmon in Alaska with an illustrated key.
National Oceanographic and Atmospheric Administration Technical
Memorandum. NMFS ABFL-2. 20 pp.
White, H.C.
smolts.
323-326.
1939. Factors influencing descent of Atlantic salmon
Journal of the Fisheries Research Board of Canada 4(5):
Wickett, W.P. 1958. Review of certain environmental factors affecting
the production of pink and chum salmon. Journal of the Fisheries
Research Board of Canada 15{5): 1103-1126.
-48 -
-
APPENDIX A
Sampling Selectivity of the Outmigrant Traps ....
F"''
!
-49 -
-
~
I
The downstream migrant traps were designed to capture juvenile resident
and anadromous fish as they outmigrated from the Susitna River between.
the Chulitna River confluence and Devil Canyon. The first trap was
deployed at Talkeetna station (RM 103.0) during the 1982 open water
season and the second trap was added during 1983. The traps have
provided the most effective technique for capturing migrating juveniles
in the mainstem, and have been important in collecting information on
the biology and timing of emigration of juvenile fishes of the Susitna
River.
Beginning in 1983, velocity measurements were collected daily at the
mouth of each trap. Velocities for the east bank trap (Trap 1) ranged
from 1.4 to 3.1 feet per second ( fps) and, over the season, averaged 2.1
fps. The west bank trap (Trap 2) had a higher mean velocity of 2.3 fps,
with a range from 1.2 to 4.0 fps .
. Large numbers of age 0+ salmon fry have been collected in the traps
during the past two seasons, but fewer age 1+ and older fish were
captured in the traps. This is a direct result of relative abundance of
the two age classes but may also be affected by trap selectivity. In
other words, the traps may be more effective at catching the younger,
smaller fish than at collecting the larger fish. Thus, the relative
abundance of older fish determined from trap catch rates may be less
than the actual abundance of these fish passing the traps.
A test of the correlation by species and age class between the raw daily
catch per hour and daily water velocity was conducted on the 1983 data
to determine if a relationship exists between trap velocity and the
resulting collection of different age classes of juvenile fish. The
results of these tests are presented in Appendix Table A-1.
The correlations of catch per hour for age 0+ chinook and coho (both
traps), and sockeye (one trap) with trap velocity were not significant
at the 95% confidence level. Conversely, the correlations of catch per
hour for age 1+ chinook, coho, and sockeye salmon to trap velocity were
significant (0.31 to 0.56). These relationships were ~ost apparent in
Trap 2.
The higher correlations for age 1+ salmon to trap velocity could be a
result of the following factors:
1) The high trap velocities and resulting higher catches of age 1+
fish occurred during periods of high mainstem discharge. The
larger age 1+ fish may migrate predominantly during these high
discharge periods.
2) The higher velocities result in more water passing through the
traps per unit time resulting in an increase in catch per hour
of the older fish.
3) The· traps are more effective_ at catching the larger fish when
the trap velocities are higher, because the migrating fish are
less able to avoid capture.
-50 -
The outmigrant traps do not appear to be selective in the collection of
age 0+ salmon, but the relative abundance of age 1+ and older fish may
be biased due to trap avoidance by the larger fish. The traps do,
however, provide a measure of the seasonal timing of outmigration and
comparative changes in relative abundance for the older fish.
Appendix Table A-1. Correlation coefficients (r) for juvenile salmon
catch per hour and trap velocity at each of the
~ownstream migrant traps, by species and age class,
1983. The data were not smoothed.
Trae 1 Trae 2
Corr. Corr.
seecies Age Class Coeff(r) ---L_ n Coeff(r) ---L_
Chinook 0+ 0.09 0.20 95 -0.02 0.44
Chinook l+ 0.39 o.oo 95 0.56 0.00
Coho 0+ 0.15 0.07 95 -0.07 0.26
Coho 1+ 0.40 o.oo 95 0.53 0.00
Sockeye 0+ 0.22 0.01 95 -0.11 0.15
Sockeye 1+ 0.31 o.oo 95 0.44 0.00
Chum 0+ 0.29 0.02 54 -0.03 0.41
Pink 0+ 0.38 o.oo 54 0.44 o.oo
-51 -
n
91
91
91
91
91
91
52
51
-
-
-
-
-
APPENDIX B
-The Schaefer Estimate of Population Size
-
""'1'
I
I
""f
. I
I
-52 -
-
-
il
!
!
-r·
I
I
' :
' !
-
One of the assumptions of a mark-recapture program which must be met to
provide a valid population estimate is that, during tagging and recov-
ery, the marked individuals are randomly distributed within the unmarked
population. A biased Petersen estimate would result if the marking and
recapture efforts were selective. Schaefer (1951) pointed out that when
generating a population estimate for migrating fishes, the fact that
some fish do not always migrate as a single population should be
considered, so that the mixing of marked and unmarked fish between the
time of tagging and recovery may be incomplete.
Schaefer (1951) provided a method for estimating the population, when
using numbered tags, by estimating the relation between time of tagging
and recovery when migration extends over a considerable period of time.
By using numbered tags, both the date of tagging and date of recovery is
known for each fish recovered and the population can be divided into a
series of distinct units.
Specific to the coded wire tag, mark-recapture program conducted on the
Susitna River during 1983, there may be a tendency for fish which emerge
earliest to outmigrate earliest, resulting in a positive correlation
between time of tagging at the emergence sites and the time of migration
past the recovery site. When such a correlation exists, the recovery
during any single period would not be a random sample of the whole
population.
The method proposed by Schaefer uses the summation of populations for
individual periods of tagging and recovery to estimate the total popu-
1 ati on. A tab 1 e is first generated which shows the number of fish
tagged and recovered during each time interval. Using these data, a
second table can be formed which estimates the population for each
period; the sum of these being the total population estimate.
The population estimate (N) was determined from the formula from
Ricker•s (1975) modification of Schaefer•s (1951) equation:
where: R .. lJ
Mi =
Ri =
R. = J
=
N = N - R Mi C; ij -ij • "if: . -R.-
1 J
the number of .fish which were marked during a tagging
period (i) and subsequently recaptured during a recovery
period {j).
the number of fish marked during a single tagging period.
the total marked fish recaptured from a single tagging
period.
the number of fish captured and examined for marks during
a recovery peri ad.
the number of marked fish which were recaptured during a
recovery peri ad.
N .. = the estimate of the available for marking during a period
lJ (i) and available for recovery in a period (j).
-53 -
Tagging and recovery periods for the Susitna River study were grouped by
eight day intervals. The data collected for the estimate of the popu-
lation of sockeye salmon outmigrants is tabulated by the Schaefer method
in Appendix Table B-1. The computation of these data and the resulting
population estimate are pre~ented in Appendix Table B-2. This estimate
is very close to the population determined from the Petersen estimate
(Section 3.2}, indicating a random distribution of marked and unmarked
sockeye salmon fry between the time ·of tagging and time of recovery
during 1983.
The mark-recovery data for chum salmon are presented in Appendix Table
B-3, and the computations and final population estimate are provided in
Appendix Table B-4. This estimate is lower than the population
determined for chum salmon fry by the Petersen estimate (Section 3.2).
The difference is probably a result of incomplete mixing of marked and
unmarked chum fry between tagging and recovery, due to the comparatively
shorter time interval of chum outmigration compared to that of sockeye
salmon fry.
With the use of distinct marks, successive groups of tagged fish main-
tain a separate identity and can be treated as separate populations.
Using the methods provided by Schaefer (1951), it is possible to gener-
ate population estimates for each time interval both at tagging and
recovery. This allows the comparison of population estimates not only
between years, but between given time periods of the outmigration during
a single year.
-54 -
-
-
-
-
-
-
-
-
Appendix Table B-1.
Per ind -of Ped od
Recovery
_ill_
24 0
2 8 0
3 9 0
4 I 0
5 28 0
6 14 0 -7 8 0
8 2 0
9 0
10 I 0
11 0 0
12 0 0
13 0 0
14 0 0
15 0
Total Tagged -Fish Recovered
(R.1 97 0
l
Total Fish
Tagged
(Mil 4,553 0
M/Ri 46.938 0
Appendix Table 8-2.
"""' Period
of
Recovery
__ill_ -26,051
21,854
5,635
4 2,871 -5 41,799
6 21,898
7 15,430
8 4,635
9 1,945
10 3,732 -11
12
13
14
15 376
Total 146,226
Data collected on ·the cod€d. wire tag,
mark-recapture experiment for sockeye salmon fry to
provide a population estimate using the methods
outlined by Schaefer (1951). Tagging and recovery
periods are by eight day intervals, May 23 through
September 27, 1983.
Tagged Total cJ
o[ Tagging (i) Fish Fish I
Recovered Recovered Rj
_l 4 (R ) (C )
0 0 24 555 23.125
2 0 10 582 58.200
88 0 97 1,294 13.340
15 2 18 1' 101 61.167
72 7 107 3,403 31.804
45 3 62 2,066 33.323
20 5 33 1,356 41.091
fi 0 8 395 49.375
3 3 7 290 41.429
3 2 6 477 79.500
8 4 12 445 37.083
6 2 8 278 34.750
0 1 1 16 16.000
0 0 0 0 0
0 0 1 8 8.000
268 29 394 12,666
10,599 2,881 17.963
39.549 96.931
Computation of the sockeye salmon fry outmigrant
population from the data presented in Appendix
Table B-1.
Period of Tagging (il
2 _l 4 Total
26,051
4,604 26,458
46,427 52,062
36,286 11,858 51,015
90,563 21,580 153,942
59,305 9,690 90,893
32,502 19,915 67,847
11,716 16,351
4, 915 12,047 18,907
9,432 15,412 28,576
11,733 14,378 26,111
8,246 6,737 14,983
1,551 1,551
376
Jl5,i29 113,168 575,123
-55-
Appe.,dix Table B-3. Data collected on the coded wire tag,
mark-recapture experiment for chum salmon fry to
provide a population estimate using the methods
outlined by Schaefer (1951}. Tagging and reco~ery
periods are by eight day intervals, May 19 through
July 13, 1983.
Period Tagged Total c.
of Fish Fish J/
Recovery Period of Ta~gin~ ( i) Recovered Recovered Rj
__ij_2._ 1 2 _1 4 (R.) (C )
l 328 328.000
2 5 5 725 145.000
] 6 2 l 9 1,301 144.556
4 2 2 2 1 7 640 91.429
5 3 2 25 30 I, 751 58.367
6 9' 9 2 ,ll4 234.889
7 1,396 1,396.000
Total Tagged
Fish Recoverd
(Ri) 9 12 36 62 !1,255
Total Fish
Tagged
Cl\) 2. 579 8,555 3,553 9,600 24,287
~1./R. 286.556 712.917 710.600 266.667
1 1
Appendix Table B-4. Computation of the cnum salmon outmigrant
population from the data of Appendix Table B-3.
Period
of
Recovery Period of Taggin!! (i)
__ill_ ...1 ..1 4 ~
93,990 93,990
2 516,152 516,!52
3 248,540 206,113 102,721 557,374
4 52,399 !30,363 !29,939 24,381 337,082
5 124,8:!2 82,951 389 ,1!4 596,897
6 563,734 563.734
7 372,267 372;267
Total 394,929 977,460 315 ,6ll 1,349,496 3,037,496
-56-
-
-
-
-
~
-i
-
-
~
-
-
-
-I
-
-
APPENDIX C
Comparison of Daily Catch Per Hour Between Outmigrant Trap 1 and Trap 2 -
-
-
--57 -
....
The raw daily mean catch per hour of Trap 1 was compared with that of
Trap 2 for all species by paired t-tests. The means between traps for
half of the species by age class groups were significantly different
(Appendix Table C-1). Smoothing the data with a three day moving
average to reduce the possibility of daily peaks causing a difference
did not change the results. Trap 2 had a higher catch per hour for the
majority of fishing days for all species by age class except age 0+
coho; however, the Trap 1 to Trap 2 proportion varied throughout the
season.
We can conclude from these results that juvenile salmon do not outmi-
grate in a uniform manner across the breadth of the ma·instem river.
Rather, individual groups appear to follow one shore or another or
perhaps the mid-channel; their location can change depending on the
level of discharge, the origin of the fish, and several other factors.
This pattern of outmigration should be considered when interpreting the
results from the data collected at the outmigrant traps.
Appendix Table C-1. Comparison of unsmoothed daily catch per hour of
juvenile salmon in Trap 1 versus Trap 2, by species
and age class.
Carr.
t-test of means£/ Species by coeH
Age Class .£!2-_ .!! t value df Signif •
Chinook, 0+ 0.84 97 -3.48 96 p<0.01
Chinook, I+ o. 90 97 0.4 7 96 NSE_/
Coho, 0+ 0.47 97 o. 72 96 NS
Coho, ~ 1+ 0.67 97 2.65 94 p< 0.01
Sockeye, 0+ 0.64 97 -4.89 96 p<0.01
Sockeye, 1+ 0.43 97 -1.45 96 NS
Chum 0.69 97 -2.59 93 p < 0.01
Pink 0.74 96 -0.98 92 NS
~~ May 18 -Sep 25, 1983; all significant at 95% confidence level
-May 22 -Aug 30, 1983
E.l NS = Not significant at 95% confidence level.
-58 -
Percent
of Days when
Trap 1 catch/hr >
Tra12 2 catch/hr
32.6
45.8
80.0
63.5
20.7
21.4
41.4
19.7
PART 2
The Distribution and Relative Abundance
of Juvenile Salmon in the Susitna River
Drainage above the Chulitna River Confluence
---------~---
·,~
·-
....
THE DISTRIBUTION AND RELATIVE ABUNDANCE
OF JUVENILE SALMON
IN THE SUSITNA RIVER DRAINAGE
ABOVE THE CHULITNA RIVER CONFLUENCE
1984 Report No. 2, Part 2
by Lawrence J •. Dugan, David A. Sterritt, and Michael E. Stratton
ABSTRACT
Alaska Department of Fish and Game
Susitna Hydro Aquatic Studies
2207 Spenard Road
Anchorage, Alaska 99503
The Juveni 1 e Anadromous Habitat Study was undertaken to determine the
seasonal distribution and abundance of juvenile salmon by macrohabitat
type in the Susitna River drainage between the Chulitna River confluence
and Devil Canyon. Thirty-five sites representing four macrohabitat
types were sampled from May through September, 1983; limited sampling
was conducted in October and November. Side channels and tributaries
were found to be important rearing areas for juvenile chinook salmon
with tributaries important early in the summer and side channels of the
mainstem Susitna increasing in importance as the summer progressed.
Coho salmon were most abundant in tributaries and upland sloughs. Natal
side sloughs and backwater areas provided rearing areas for chum and
sockeye salmon fry. Upland sloughs, the most lake-like environment, had
concentrations of sockeye and coho salmon juveniles. Macrohabitat type
and time of year were found to be significantly (p < 0.10) related to
the distribution of all species •
i
TABLE OF CONTENTS
ABSTRACT. . . . • • • . • . • . • • . . . • . . • . . . . • . . . . . . • . . • • . . . . • . • • . . . . . . • . • . . . . . i
LIST OF FIGURES.................................................... iv
LIST OF TABLES..................................................... vi
LIST OF APPENDIX TABLES •••••••• vi
1.0 INTRODUCTION •••••••••••••••••••••••••••••••••••••••••••••••••• 1
2.0 METHODS ••••••••••••••••••••••••••••••••••••••••••••••••••••••• 3
2.1 Field Sampling Design ••••••.•.••••••••••.••••••.•••• 3
2.1.1 Study site locations and selection criteria .. 3
2.1. 2 Field data collection ••••••••••.••••••••.••.•••.•• 6
2 .1. 3 Schedule of activities and frequency of sampling •• 6
2.2 Data Recording and Analysis ••••••••••••••••.•.•.••••••. 8
2.2.1 Macrohabitat use ••••.•.••••••••••.•••••••••••.•••. 8
2.2.2 Analysis of variance •••••••.••••••••• 8
3.0 RESULTS ••••••••••••••••••••••••••••••••••••••••• . . . . . . . . . . . . . . 10
3.1 Distribution of Juvenile Chi nook Salmon •••• .............. 10 -3.2 Distribution of Juvenile Coho Salmon ••••••• .............. 10
3.3 Distribution of Juvenile Chum Salmon ................ 14
3.4 Distribution of Juvenile Sockeye Salmon .................. 20
3.5 Analysis of Variance ..•••...•........ ~·············· 20
4.0 DISCUSSION •••••••••••••••••••••••••••.•••••••••••••••••••••••• 27
4.1 Limitations of the Data .....•.............•............ 27
4.1.1 Sampling limitations .• 27
4.1.2 Gear efficiency. 27
4.2 Chinook Salmon •• 28
4.3 Coho Sa 1 rna n ••••••••••••••••• 31
4.4 Chum Salmon •...•... 34
ii
-
TABLE OF CONTENTS (Continued)
4. 5 Sockeye Salmon................................ . . . . . . . . . . . . . 35
5 .0 CONTRIBLITORS ••••••••••••••••••••••••••••• e • • • • • • • • • • • • • • • • • • • • 38 -
6.0 ACKNOWLEDGEMENTS................................................ 39 -7.0 LITERATURE CITED.............................................. 40
8.0 APPENDICES
Appendix A Summary statisitics for transformed catch/cell
data ............................................... . 43
Appendix B Gear efficiency experiments....................... 48
-
-
-
-
-iii
,_
'
LIST OF FIGURES
Figure Title Page
1 Juvenile Anadromous Habitat Study (JAHS)
sites sampled more than three times by
macrohabitat type, 1983.................................... 5
2 Arrangement of transects, grids, and cells at
a Juvenile Anadromous Habitat Study (JAHS)
s i te . . . . . • • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3 Seasonal distribution and relative abundance
of juvenile chinook salmon on the Susitna
River between the Chulitna River confluence
and Devil Canyon, May through November 1983................ 11
4 Density distribution of juven·ile chinook
salmon by macrohabitat type on the Susitna
River between the Chulitna River confluence
and Devil Canyon, May through November 1983................ 12
5 Juvenile chinook salmon mean catch per cell
at tributaries and upland sloughs by sampling
period, May through November 1983.......................... 13
6 Juvenile chinook salmon mean catch per cell
at side slough and side channels by sampling
period, May through November 1983....... •• •• . • • • • • • • . • . • . • • 13
7 Seasonal distribution and relative abundance
of juvenile coho salmon on the Susitna River
between the Chulitna River confluence and
Devil Canyon, May through November 1983.................... 15
8 Density distribution of juvenile coho salmon
by macrohabitat type on the Susitna River
between the Chulitna River confluence and
9
Devil Canyon, May through November 1983.................... 16
Juvenile coho salmon mean catch per cell at
tributaries and upland sloughs by sampling
period, May through November 1983.......................... 17
10 Juvenile coho salmon mean catch per cell at
side sloughs and side channels by sampling
period, May through November 1983.......................... 17
11 Seasonal distribution and relative abundance
of juvenile chum salmon on the Susitna River
between the Chulitna River confluence and
Devil Canyon, May through November 1983.................... 18
iv
Figure Title
12 Percentages of the total juvenile chum salmon
catch by sampling period, May thro~gh October
1983 ......................................... 5 •••••• ,., ••••• 0 • 19
13 Juvenile chum salmon mean catch per cell at
the four macrohabitats by sampling period,
14
15
16
17
18
May through October 1983................................... 19
Density distribution of juvenile chum salmon
by macrohabitat type on the Susitna River
between the Chulitna River confluence and
Devil Canyon, May through October 1983..................... 21
Seasonal distribution and relative abundance
of juvenile sockeye salmon on the Susitna
River between the Chulitna River confluence
and Devil Canyon, May through November 1983.... •• • • • . • • • • • • 22
Percentages of the total juvenile sockeye
salmon catch by sampling period, May through
October 1983............................................... 23
Juvenile sockeye salmon mean catch per cell
at three macrohabitats by sampling period,
May through October 1983................................... 23
Density distribution of juvenile sockeye
salmon by macrohabitat type on the Susitna
River between the Chulitna River confluence
and Devil Canyon, May through October 1983................. 24
19 Seasonal deviation of catch per unit effort
of juvenile chinook salmon on the Susitna
River between the Chulitna River confluence
20
and Devil Canyon, May through September 1983............... 30
Seasonal deviation of catch per unit effort
of juvenile coho salmon on the Susitna River
between the Chulitna River confluence and
Devil Canyon, May through September 1983 •••••••.••••...•••.
v
-
-
-
, ....
-
-
-
-
-
-
-
LIST OF TABLES
Table Title
1 Juvenile Anadromous Habitat Study (JAHS) sites
sampled on the Susitna River between the
Chulitna River confluence and Devil Canyon,
May through November 1983 .•....••••••••..•.• ~·······~······ 4
2 Results of analysis of variance of juvenile
salmon distribution by selected habitat
variables.................................................. 26
LIST OF APPENDIX TABLES
Appendix Table Title
A-1 Summary statistics for transformed catch/cell
data of each species, by groups for each habitat
parameter. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . 44
B-1 Capture probabilities for chum and sockeye
salmon at Slough 11 as a function of percent
cover.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
B-2 Comparison of beach seining and backpack
electrofishing juvenile chinook catches at
five cells fished at two different turbidity
levels ..................................................... 50
vi
....
-
-
1.0 INTRODUCTION
The Resident and Juvenile Anadromous Fish Studies (RJ) have been direct-
ed toward accomplishing the general objectives described in 1979 by the
Alaska Department of Fish and Game for the Susitna Hydroelectric Project
(ADF&G 1979). These objectives are stated below:
A. Define seasonal distribution and relative abundance of resi-
dent and juvenile anadromous fish in the Susitna River between
Cook Inlet and Devil Canyon.
B. Characterize the seasonal habitat requirements of selected
anadromous and resident species within the study area.
Five species of Pacific salmon spawn in the reach of the Susitna River
above the Chulitna River confluence. With the exception of pink salmon,
substantial freshwater rearing and growth occur in this reach of river.
The Resident and Juvenile Anadromous Fisheries Studies began in November
1980 with general surveys of the Susitna River mainstem and associated
habitats between Cook Inlet and Devil Canyon conducted during the open
water season of 1981~ Beginning in the winter of 1981 and the spring
and summer of 1982, the studies concentrated on those areas of the
mainstem and associated habitats that may be most affected by the
development of the Susitna Hydroelectric Project.
The data collected during 1981 and 1982 outlined the general dis-
tribution patterns of these species and their habitat utilization (ADF&G
1981b, 1981c, 1983c). The 1982 studies also investigated the response
of selected macrohabitat areas to mainstem discharge changes and demon-
strated species differences in the use of "hydraulic zones" (ADF&G
1983d). These zones were subsections of the slough and tributary mouth
areas that were affected by backwater of the mainstem Susitna River,
mixing areas of the mainstem with slough or tributary flow, and free-
flowing tributary or slough water above ·the back water. The relative
use of the hydraulic zones by each species of juvenile salmon was
analyzed to provide an incremental index of habitat availability for
each species. This analysis provided evidence that the relative use by
juvenile salmon of these macrohabitat areas was affected by changes in
mainstem flow. During the course of the 1982 study, observations of the
distribution of juvenile salmon indicated certain microhabitat parame-
ters within the zone may respond to discharge changes at a higher rate
than does zone surface area. These microhabitat factors include cover
and turbidity, with depth and velocity having a somewhat lesser impor-
tance.
The objectives of the 1983 Juvenile Anadromous Habitat Study (JAHS)
program were to correlate juvenile salmon habitat use to microhabitat
parameters and further document the seasonal distribution and relative
abundance of juvenile salmon (except pinks) in macrohabitat types
(tributaries, upland sloughs, side sloughs and side channels) associated
with the Susitna River above the Chulitna River confluence. Pink salmon
are not discussed because of the short time they spend in this reach of
the river between emergence and outmigration. The purpose of this paper
- 1 -
is to present the data on spatial and seasonal distribution and relative
abundance for each species and to discuss the causative factors behind
the observed distributions.
Juvenile salmon distribution and abundance data will be used to deter-
mine the proportion of use of the macrohabitats associated with the
mainstem river. In addition, the data can be used in the assignment of
dam flows throughout the summer to minimize the effects on life stages
of different juvenile anadromous species. Furthermore, the data will be
integrated into macrohabitat indices compiled by E.W. Trihey and
Associates which project the percentages of suitable rearing habitat for
each juvenile salmon species over a range of mainstem flows between
9~000 cfs and 23~000 cfs. Distribution and abundance data were also
used in conjunction with microhabitat studies including the juvenile
salmon habitat suitability functions (Part 3 of this report), the
juvenile salmon habitat modelling (Part 4), and the IFG-4 modelling
{Part 7).
- 2 -
-
-
_,1
-
.....
-
.....
-
-
-' I
-
r
I
2.0 METHODS
2.1 Field Sampling Design
Two Juvenile Anadromous Habitat Study (JAHS) field crews collected
distribution and abundance data at rearing habitats used by juvenile
salmon. Selected side sloughs, upland sloughs, tributaries and mainstem
side channels of the Susitna River between the Chulitna River confluence
(RM 98.5) and Portage Creek (RM 148.8) were sampled during the open
water season. Crews operated out of tent camps and used river boats for
transportation with helicopter support when necessary.
2.1.1 Study site locations and selection criteria
Thirty-five study locations on the Susitna River and its major tribu-
taries between the Chulitna River confluence and De vi 1 Canyon were
sampled (Table 1). Rearing habitat at thirteen of the sites was subse-
quently modelled using either RJHAB (Part 4) or an IFG model (Part 7).
Sites sampled more than three times are shown in Figure 1.
Sites selected for study included: (1) sites where relatively large
numbers of spawning adult salmon were recorded in 1982 (ADF&G 1983b),
(2) sites where concentrations of rearing juvenile salmon were observed
or collected in 1981 and 1982, and (3) sites representing macrohabitat
types associated with the Susitna River that are affected by changes in
mainstem flow.
In 1982, sampling sites were classified on the basis of morphological
features into one of four macrohabitat types: tributary, upland slough,
side slough, or side channel. Upland sloughs are areas which have heads
vegetated with trees and brush that are rarely overtopped. Side sloughs
are sites with unvegetated heads that are sometimes overtopped by
mainstem flows during the open water season of a normal year. Side
channels convey mainstem flows overtopped, during most of the open water
season of a normal year. ·
Side sloughs are morphologically and hydraulically distinct from side
channels for several reasons. A mainstem backwater area is frequently
present at the mouths of side slo~ghs. Fewer backwater areas occur at
the mouth of side channels because the gradient of the side channels is
typically greater than that of sloughs. The infrequency of strong flows
in the s 1 oughs over the course of severa 1 years has a 11 owed silt,
debris, and deadfall to accumulate. Debris and silt is often flushed
out of the side channels and sometimes the streambed may become armored.
The water in sloughs is often clear and moving slowly and is therefore
more conducive to the growth of aquatic and emergent vegetation.
In 1983, side sloughs and side channels were distinguished using a
discharge-based classification scheme which depends on the status of the
head of the. site. Under this criterion, sites are classified as side
sl0ughs only when the head is not overtopped by mainstem discharge.
When the head is overtopped by the mainstem, these sites are classified
as side channels. Classification of upland sloughs did not change.
- 3 -
Table 1. Juvenile Anadromous Habitat Study (JAHS) sites sampled on the
Susitna River between the Chulitna River confluence and Devil
Canyon, May through November 1983.
Site
Whiskers Creek
Slough
*Whiskers Creek
*Slaugh 38
*Mainstem at head of
Whiskers Creek Slough
Chase Creek
Slough 5
Oxbow I
Slough 6A
*Mainstem above
Slough 6A
*Lane Creek
Slough 8
Mainstem II
*Lower McKenzie Creek
*Upper McKenzie Creek
*Side Channel below
Curry
*Oxbow II
Slough 8A
Side Channel lOA
Slough 9
Side Channel 10
*Lower Side Channel 11
Slough 11
*Upper Side Channel 11
Indian River -Mouth
Indian River-TRM 10.1
*Slough 19
*Slough 20
Side Channel 21
Slough 21
Slough 22
*Jack Long Creek
Portage Creek Mouth
Portage Creek TRM 4.2
Portage Creek TRM 8.0
~ T -Tributary
US -Upland Slough
SS -Side Slough
SC -Side Channel
River
Mile
101.2
101.2
101.4
101.4
106.9
107.6
110.0
112.3
112.4
113.6
113.6
114.4
116.2
116.7
117.8
119.3
125.3
127.1
129.2
133.8
134.6
135.3
136.2
138.6
138.6
140.0
140.1
140.6
142.0
144.3
144.5
148.8
148.8
148.8
Macro-
habits;
~
SS/SC
T ss
sc
T us
SC/SS us
sc
T ss
SC/SS
T
T
sc
SC/SS ss sc
SS/SC
SC/SS sc ss sc
T
T us
SS/SC sc
SS/SC
SS/SC
T
Total
T
T
T
Fish RJHAB IFG
Distri-Model-Model-
bution ing ing
Site Site Site
X X
X
X
X
X
X X
X
X X
X
X
X X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X X
X
X
X
X
35 6
X
X
X
X
X
X
X
7
*These sites sampled three times or less.
- 4 -
-
-
-
-
-
/'""
"""
;"""
-
'~
-
-
Macrohabitat Type
j;jf,(// :;,0. ~ •
... ,1 Q:;,o. ~ ,l
b ~ ..... ·(J~
.._o"' be .~ ...
q .... 0.~
.:::. Go] ~"
Porto~:~ a Creek 4
Slou~:~h 22 4 ...
SIOuQh 21 4 ...
Side Channel 21 ...
Indian River 4
SlouQh II •
Side Channel 10 ... 4
Side Channel lOA •
Slough 9 • ...
SlouQh SA • ... •
4
• ... • ... ...
Whiskers Cr. S I. • ...
Figure 1. Juvenile Anadromous Habitat Study (·JAHS) sites sampled more
than three times by macrohabitat type, 1983.
-5-
This is the classification method which was used by E.W. Trihey and
Associates to measure the total surface area of each macrohabitat type
in this reach of river; this method is used in all parts of this report.
The discharge-based method is useful when considering fish distribution
because of the major habitat changes which occur when the head of a
slough is overtopped. The geomorphological-based method is useful
because the frequency of overtopping has an important influence on the
distribution of substrate and object cover which are important to
juvenile and spawning salmon. A classification based on the discharge
acknowledges the instantaneous effect of mainstem discharge, while one
based on geomorphological differences emphasizes long-term consequences.
Both effects are important.
2.1.2 Field data collection
Each of the study sites was divided into one or more grids. Grids were
located to keep water quality (temperature, turbidity) within the site
as uniform as possible and to encompass a variety of depth, velocity,
cover, and substrate types. Each grid consisted of a series of
transects which intersected the channels of the study sites at right
angles (Figure 2). There were one to three cells (6 ft. in width by 30
ft. in length = 300 sq. ft.) at every transect within the grid. An
attempt was made to confine uniform habitat within each cell. Further
descriptions of the grid system used are detailed in the 1983-84 Proce-
dures Manual (ADF&G 1984). Habitat data collection methods are further
described in Parts 3 and 4 of this report.
Backpack electrofishing units (Coffelt, Model BP1C and Smith-Root, Model
XVBPG) and beach seines were used to collect fish. Procedures used for
sampling with these techniques are described in the 1982-83 Procedures
Manual (ADF&G 1983a). Juvenile salmon collected were identified to
species, measured for total length in millimeters and released in the
cell from which they were captured. A few specimens were preserved in
10% formalin for later identification.
Fish were usually sampled from a minimum of seven cells within each grid
at each site. The cells were selected to represent the complete range
of habitat types available within the grid. Fish density was estimated
by electrofishing or beach seining the entire cell, attempting to
capture all fish. Catch per unit effort (CPUE) was defined as the catch
(number of fish) per cell. With this definition, electrofishing and
beach seining effort could be compared; also, the extra time required to
capture fish in difficult locations would not bias the results as it
would have had if we defined CPUE as catch per unit time.
2.1.3 Schedule of activities and frequency of sampling
The sampling schedule was dependent on the target species. Sites that
predominantly had juvenile chum, pink, and sockeye salmon were sampled
in May and June. In late June and early July, sampling efforts were
redirected toward sites previously identified ·in 1981 and 1982 as
rearing areas for chinook and coho salmon. The chinook and coho salmon
sites were sampled until freezeup in early November. Because the
- 6 -
.,
~.
-
-
-
-
"""' I
LEFT
BANK
TRANSECT 6
TRANSECT 5
TRANSECT 4
TRANSECT 3
TRANSECT I
D 6x 50 Foot
Ce II U r, it -
Area Sornpled
RIGHT
BANK
0
a::
e>
-.. .. u..
0
0
0
~
0
u
0..
"" 1-
Figure 2. Arrangement. of transects, grids, and cells at a Juvenile
Anadromous Habitat Study (JAHS) site.
-7-
primary objective of the JAHS study was microhabitat suitability and
habitat modelling, there was not equal sampling effort at all sites,
which would be more desirable, from the standpoint of a distribution and
relative abundance study. This problem was partially solved by using
catch per unit effort data.
2.2 Data Recording and Analysis
All field data were recorded on data forms and transmitted to the
office, where they were entered into a rna in frame computer data base.
Data sorts and summary retrievals were extracted from this data base as
needed.
2.2.1 Macrohabitat use
Percentage distribution of each salmon species among macrohabitat types
was calculated by dividing the catch/cell for each type by the sum of
the catch/cell for all types. The equations are:
Percentage; = n
L:
i=1
(Total Fish);/(Tota1 Cells);
(Total Fish)i/(Total Cells);
where: i = each macrohabitat type
n = number of macrohabitat types = 4
2.2.2 Analysis of variance
X 100
An analysis of variance (ANOVA) was conducted to examine the effect of
several habitat variables on the distribution of each species. The two
major variables considered were macrohabitat type and time of year.
Site habitat characteristics {which contribute to differences among
macrohabitat types) considered were: mean water depth, mean water
velocity, mean percent cover, water temperature, and turbidity. All of
these can be influenced by discharge level. Temperature and turbidity
are influenced by time of year; the other variables are indirectly
influenced by time of year in that discharge levels have a seasonal
pattern.
All sites were grouped into the four macrohabitat types -tributary,
upland slough, side slough, or side channel. Periods were taken as the
nine half-month periods from late May (May 16-r~ay 30) to late September
(Sept. 16-Sept. 30). Study site depth, velocity, and percent cover were
calculated as the mean values of all 300 sq ft cells sampled in a
particular interval of each parameter, such as 0.1 to 0.6 ft. There
were usually at least seven cells sampled at each sampling site on each
occasion. Because the cells were not randomly distributed at the site,
the ANOVA is weakened for the three variables (depth, velocity, cover)
which were taken as means of the cells sampled. However, it appeared
that the means of these three would generally characterize each site.
- 8 -
-
-
-
-
All variables were transformed by natural log (x+l) prior to running the
ANOVA. The intervals and frequencies for all the variables are given in
Appendix Table A-1. The intervals were selected to be physically or
biologically meaningful while still allowing for an adequate sample size
in each interval. For example, the first interval for turbidity is 0 to
10 NTU, which covers the non-flood tributary conditions.
Fish density data were taken as the total number of fish captured in a
particular interval, divided by the number of 300 sq. ft. cells sampled
in that interval. Mean catch per cell for each species was transformed
by natural log (x+l).
The analysis of variance was run on BMDP Statistical Software, using the
regression approach. One run was conducted for macrohabitat type and
period, with fish catch/cell as the dependent variable and a second run
was conducted for mean depth, mean velocity, mean percent cover, water
temperature, and turbidity, with fish catch/cell as the dependent
variable. Because of empty cells in the analysis of variance table,
interactions among variables were not calculated.
- 9 -
3.0 RESULTS
3.1 Distribution of Juvenile Chinook Salmon
A tot a 1 of 4,443 juveni 1 e chi nook salmon were captured at JAHS sites
located between the Chulitna River (RM 98.6) confluence and Portage
Creek (RM 148.8) from May 1 to November 15, 1983. Approximately 99% of
these fish were Age 0+ and the rest were Age 1+. Chinook juveniles were
captured at all of the study sites surveyed at least four times (Figure
3). Chinook juvenile salmon were widely distributed from early July
through September. Portage Creek and Indian River produced the highest
densities of chinook salmon through the ice free field season. In-
creases in densities were apparent as the season progressed at several
sites.
Chinook juveni1e salmon were unequally distributed among macrohabi tats
Side channels contributed 22.6 percent of the catch per unit effort
(CPUE), the highest percentage of the three macrohabitats influenced by
mainstem flows (Figure 4). The CPUE of chinook juveniles captured from
side channels was twice that of side sloughs, and twelve times that of
upland sloughs. (See also Appendix Table 1, which gives the means used
in the analysis of variance). Four side channels (Slough 22, Side
Channel lOA, Oxbow I and Slough 9) accounted for 80.8 percent of the
juvenile chinook captured at 13 side channels sampled during the 1983
field season. Side channel lOA (RM 127.1) contributed 31.1 percent of
the chinook juvenile captured at this macrohabitat type.
Chinook juvenile salmon CPUE by macrohabitat type ranged from less than
one fish per cell in May at upland slough and side slough study sites to
26.4 fish per cell at tributary macrohabitats in early July (Figure 5).
Consistently higher densities of chinook salmon were recorded for
tributary sites than for upland slough, side slough, or side channel
sites 'from May through early August. Peak densities of 26.4 fpc and
19.5 fpc were recorded at tributary sites in early July and August,
respectively. Chinook juvenile densities were higher in tributaries in
July and August than in side sloughs or side channels. Chinook juvenile
densities increased at mainstem associated macrohabitats in late July.
Chinook juveniles were redistributing into mainstem side channels, side
sloughs and to a lesser extent upland sloughs during this time following
outmigration from tributaries. Comparison of chinook juvenile salmon
densities between side slough and mainstem side channel macrohabitats is
illustrated in Figure 6. In general, side channel CPUE•s were higher
than those in side sloughs. Chinook juvenile densities in both areas
gradually increased until late August or early September. Side channel
densities of juvenile chinook salmon gradually decreased after August.
Densities at side sloughs were higher in September and October than
earlier in the season. Densities were five times greater at side
sloughs in surveys conducted during September through November than
before September.
3.2 Distribution of Juvenile Coho Salmon
A total of 2,023 juvenile coho salmon were captured at sites located
between the Chulitna River (RI~l 98.6) and Portage Creek (RM 148.8).
-10 -
ilf!lillm'i
-
-
-
-
.~
-
F""'
....
i
"""'
l.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Portage Cre.ek (a 11
Slough 22
sites)
Slough 21
Side Channel 21
Indian River (all sites)
Slough 11
Side Channel 10
Side Channel lOA
Slough 9
Slough SA
Mainstem 2
Slough 8
Slough 6A
Oxbow One
Slough 5
Chase Creek
Whiskers Creek Slough
CHI NOOK
MEAN CATCH PER CELL
RELATIVE ABUNDANCE KEY
oo.o
QO.I-0.49
~0.5-4.99
• s.o-49.99
-No Sample
JULY AUG. SEPT. OCT.
r II I II I 1I I
••••••• ~ 'iii
·~·~· . 000-~-~-~-
-~-Cl-Q.,--Q
~~ -....... ii
000-~---'iii
-~ ~ ~ ~.
----~.-...
~0 CJ -'iii·~ 'iii iii
0-Cl-CJO-ii
-CJ-~-iiiCJ'ii'iiiii
-Cl-0-~Cliii'ii-
~Cl-iii-ii~CliiiCJ
-----iiiiiiliiiiiil iii
-0-0 -~Cliii iiiQ
-Q-0-1:1-CJ
-0-'ii-'iiiil···
Figure 3. Seasonal distribution and relative abundance of juvenile/
chinook salmon on the Susitna River between the Chulitna
River confluence and Devil Canyon. May through November 1983.
-11-
I .......
N
I
j
Mainstem II
9. 3% Oxbow One
Eight Sites 8.2%
Combined 4.0%
Whiskers Creek
Slough
Side Channel
SIDE SLOUGHS
Five Tributaries
Combined 10.4%
Slough 22
~
Oxbow One
10.7%
_--,........_ Side Channel
SIDE
CHANNELS
10 .,j 17.9%
Twelve Sites
Combined
Slouoh 9
S.'~------..._UPLAN 0 S L 0 UGHS
COM 8 IN ED MACROHABITAT
TYPES
Figure 4. Density distribution and juvenile chinook salmon by
macrohabita t type on the Sus itna River between the Chulitna
River confluence and Devil Canyon, May through November 1983.
Percentages are based on mean catch per cell.
J
,....
I
!
-
..J
..J
LLI
0
a:
LLI
0..
::c
0
1-
<(
0
z
<(
LLI
::l:
30.
II TRIBUTARIES
~ UPLAND SLOUGHS * NO £fFOIU
n = CHINOOK CATCH
Figure 5. Juvenile chinook salmon mean catch per cell at tributaries and upland sloughs
by sampling period, May through November 1983.
,--6-.01
Ill ~~B~fAL.f~~~OITIONS
rldJ ~~B~T~l~~~51 TIONS
••629
fl!!!l!m
..J
..J
LLI
0
a:
LLI
0..
::c
~
0
z
<(
LLI
::::s
--.5.0
4.0
1.0
---* NO EFFORT
n •CHINOOK CATCH
Figure 6. Juvenile chinook salmon mean catch per cell at side sloughs and side channels
by sampling period, May through November 1983.
-13-
Three age classes of juvenile coho salmon from the 1980, 1981 and 1982
brood years (age 2+, 1+, and 0+ respectively) were captured. Ninety-
seven percent of the coho juvenile salmon captured at JAHS sites in 1983
were from the 1982 brood year (age 0+), three percent were age 1+, and
less than one percent were age 2+ fish.
In general, coho juvenile salmon were widely distributed in low den-
sities at many sites in the Chulitna River to Devil Canyon reach of the
Susitna River, although high tributary densities were observed in early
July and August (Figure 7). Juvenile coho CPUE 1 s were frequently
highest at sites located in the lower segment of the Chulitna River to
Devil Canyon reach.
The comparative distribution of coho juvenile salmon by macrohabitat
types is depicted in Figure 8. Coho juveniles were captured mainly in
tributaries and upland sloughs, with Whiskers Creek and Chase Creek
being the primary tributary capture sites and Slough 5 and Slough 6A
being the primary upland slough capture sites. Coho juvenile salmon
were rarely encountered in side channels. Twelve side channel sites
were sampled during 1983 and less than one percent of the juvenile coho
salmon were captured at this macrohabitat type. Side channels appear to
function as a pathway for redistribution of fish from tributaries
macrohabitat into upland sloughs and side sloughs such as Whiskers Creek
Slough and Slough 8. Side sloughs contributed 10% of the coho juvenile
salmon total CPUE. Whiskers Creek Slough and Slough 8 contributed 99
percent of the juvenile coho captured at side sloughs.
Coho juvenile salmon catches ranged from 20 fish per cell at tribu-
taries, to less than one fish per cell at side channels and side sloughs
(Figure 9). Densities were higher in upland and side sloughs during
late summer than in early summer or in autumn.
The highest densities of coho juvenile salmon were captured at tribu-
taries in late June. Upland slough catch rates were higher from late
July through late September than the catch rates for the other macrohab-
itat types. The highest densities of coho juvenile salmon at upland
sloughs occurred in late July and then catch rates gradually declined
through late September.
Seasonal trends in juvenile coho salmon in densities in side slough and
side channel macrohabitats were not observed (Figure 10). Side slough
densities of coho juvenile salmon were consistently higher than
densities in side channels except during late June.
3.3 Distribution of Juvenile Chum Salmon
A total of 1,174 juvenile chum salmon were captured by electrofishing
and beach seining at the JAHS sites from early May through July. During
this same time period, the downstream migrant trap captured 8,555
juvenile chum salmon. The outmigration of chum salmon from this reach
of river by early August is apparent from Figure 11.
The percent of total juvenile chum catch by two week period is presented
in Figure 12. Catches at JAHS sites peaked in late May, by which time
-14 -
-
-
-
-
-
-
-
!""""
-
-
1. Portage Ct·eek (;;ll s i tes)
2. Slough 22
3. Slough 21
4. Side Channel 21
5. Indian River (oll sites)
6. Slough 1!
7. Side Channel 10
R. Side Channel lOA
9. Slough 9
10. Slough 8A
11. Mainstem 2
12. Slough 8
13. Slough 6A
14. Oxbow One
15. Slough 5
16. Chase Creek
17. Whiskers Creek Slough
COHO
MEAN CATCH PER CELL
RELATIVE ABUNDANCE KEY
oo.o
c:J 0. I -0. 4 9
iiiiil 0.5-4.99
• s.o-49.99
-No Sample
JULY AUG. SEPT. OCT.
r rr r. rr r II I
I. -Q 0 0 0 0 • 0 ii 0 ji
----00000 0
DOD -0-c.:J-0
-0-0-0 0
c.:JO -iiiiil•iiiiiliiiiiliir::l ii
000-0-c.:J-::J
-00000
-00-000
'---9. 0 0 0 -0 0 0 0 0
0-0-~0-ii
-0-0-00000
-o-o-•iiiiiliiiiiiil
00-iiiiil-··liliiiiil.
-----0000 ~
-0-0-iiiiill::lilliliiiiil
-~-·-·-~ ---·-·iiiiiliiiiiliiiiiliiiiil
Figure 7. Seasonal distribution and relative abundance of juvenile coho
salmon on the Susitna River between the Chulitna River
confluence and Devil Canyon, May through November 1983.
-15-
I
I--' , en
I
-J
Whiskers Creak
Slouoh ---\-
./Eleven Slouoh•
-----/ Combined
1.2%
Whisker•
Craek --t--
Siouoh
Fourteen Mainstem Side
Channel• Combined
O.S%
Indian River
6.9% Slouohl9
4%
Whiskers
Cruk
Chase
Creak~
TRIBUTARIES~ ~
COMBINED MACROHABITAT
TYPES
U PLANO SLOUGHS
Figure 8. Density distribution of juvenile coho salmon by macrohabitat
type on the Susitna between the Chulitna River confluence and
Devil Canyon, May through November 1983. Percentages are
based on mean catch per cell.
f""'
F"""
r-
-
-
..J
..J ...
(,.)
a:: ...
0..
::c
(,.)
..... <
(,.)
z
< ...
:::.
..J
..J
1&1
(,.)
a::
11.1
0..
:t:
(,.) ..... < (,.)
2 <
11.1
:E
n•SI4
·M~ I • TRIIUTARIIS
IS.O ~ UPLAND SLOUGHS
* NO EFFORT ----___,___
1 • COHO CATCH 10.0
9.0
8.0
T.O
6.0
s.o
4.0
3.0 __..,___
Figure 9. Juvenile coho salmon mean catch per cell at tributaries and upland sloughs by
5ampling period, May through November 1983.
.5.0
• SIDE SLOUGH
••tl7 HABITAT CONDITIONS
fl SIDE CHANNEL
HABITAT CONDITIONS
4.0 * NO EFFORT
1 • COHO CATCH
1.0
Figure 10. Juvenile coho salmon mean catch per cell at side sloughs and 5ide channels by
5ampling period, May through November 1983.
-17-
!. Portage Creek (all s i tes)
~ Slough 22 L •
3. Slough 21
4. Side Channel 21
5. Indian River (all sites)
6. S l oug~. 11
7. Side Channel 10
!L Side Channel lOA
9. Slough 9
10. Slough 8A
11. Mainstem 2
12. Slough 8
13. Slough 6A
14. Oxbow One
15. Slough 5
16. Chase Creek
17. Whiskers Creek Slough
JULY AUG. SEPT. OCT.
I II I. II I II I
1. -Oc:JOO'iiiiOOOO 0
2. -- - -0 0 c:J 0 0 0
~-•• 'iii -0-0-0
-iii-D-O 0
'iii• -•O'iiiiOOO 0
'iiii·'iiii-c:J-0-0
-00000
-00-000
'---9. 'iii • 'iii -0 0 0 0 0
'iiii-•-'iiiiD-0
-'iiii-'iiii-ODOOO
-'iiii-'iiii-0000-
1 3. 0 'iii -0 -'iii 0 0 0 0
I 4. -0 0 0 0 -. 0
CHUM Is. -0-0 -0 0 0 0 0
MEAN CATCH PER CEL 1 6 · -If -0 -0 -0
RELATIVE ABUNDANCE KEY 11. ---If -[] 0 0 0 0
0.00 0
0.01·0.251:J
0.26·5.00 iii
> 5.00.
-No Sample
Figure 11. Seasonal distribution and relative abundance of juvenile
chum salmon on the Susitna River between the Chulitna River
confluence and Devil Canyon, May through November 1983.
-18-
-
-
-
-
'~ ..
-
-
-
-
:I: u
1-
<l u
..J
<l
1-
0
1-
~
0
1-z
UJ u
0::
UJ
Q..
MAY JUNE JULY AUGUST SEPT. OCT.
Figure 12. Percentages of the total juvenile chum sa1mon catch by
sampling period, May through October 1983.
..J
..J
UJ
<.J
0::
UJ
Q..
:I:
<.J
1-
<l
<.J
z
<l
UJ
:!:
8.58
4.0
f!j-Side Slouohs
~-Upland Sloughs
• -Side Channels
tz1-Tri butariu
• -No Effort
0 -No Cotch
Figure 13. Juveni1e chum salmon mean catch per cell at the four macro-
habitats by sampling period~ May through October 1983.
-19-
--~~--~--------.•.. ····· .,. . . -
over 60% of the total catch had occurred. The downstream migrant trap
recorded two peaks, one in early June and one in early July.
Juvenile chum salmon were abundant during May and June at sites having
previous year spawning and were absent from the study sites by the end
of July. Catch rates were highest in side slough and tributary macro-
habitats and low in upland slough and side channel macrohabitats (Figure
13). Only 5% of the total catch was captured in these latter macrohabi-
tats.
The comparative distribution of juvenile chum salmon densities is
presented in Figure 14. Juvenile chum salmon were most dense at tribu-
taries and side sloughs. As catches at side sloughs decreased, catches
at upland sloughs used for rearing increased.
3.4 Distribution of Juvenile Sockeye Salmon
A total of 1,010 juvenile sockeye salmon were captured by electrofishing
and beach seining at the JAHS sites from early May through September.
All juvenile sockeye salmon actually captured at JAHS sites were age 0+.
A few Age 1+ fish were visually observed at Slough 11.
The downstream migrant trap, located at RM 103.0 captured 12,395 juve-
nile sockeye between May 18 and September 25. Juvenile sockeye salmon
were captured at 12 (71%) of the 17 JAHS sites sampled at least four
times (Figure 15). They were absent from the study sites above Slough
SA after mid August; catches were still being made at sites below this
until the end of September. The percent of total juvenile sockeye catch
by two-week period is presented in Figure 16. Two peaks occurred in the
catches, one in late May-early June and one in early August. The major
peak at the downstream migrant trap occurred in mid-July.
Catch rates were highest in side sloughs and upland sloughs and lowest
in side channels and tributaries (Figure 17). A single catch of four
juvenile sockeye occurred in early June in Portage Creek, the sole
tributary found to contain juveni'le sockeye salmon.
The relative distribution of juvenile sockeye salmon among macrohabitat
types is given in Figure 18. Juvenile sockeye salmon were predominantly
found at side sloughs and upland sloughs. Almost all of the sockeye
were caught at either upland sloughs or near their natal areas (side
sloughs). The higher densities observed at Slough 11 are attributable
to the amount of spawning occurring there in 1982 (ADF&G 1983b).
3.5 Analysis of Variance
The mean values of the transformed catch per cell which were compared
among the i nterva 1 s of each parameter are shown for each species in
Appendix Table 1. If any one of the means within a parameter is signif-
icantly different from any of the other means, then the parameter is
considered to influence the varying levels of catch associated with the
distribution of that species. The confidence level for this analysis
was taken to be 90%.
-20 -
.....
-
-
-
-
....
-
I
N
1-'
I
Seven Slouohs
Combined 2.5%
s louoh 8 6.5"/o
Slouoh 9 7.8"/o
Slouoh SA
Sl DE SLOUGHS
l
Slouoh II
1
Whiskers Creek s louoh
Slouoh 22
Eleven
Mainstem
Sites Combined
9.5%
MAINSTEM SIDE CHANNEL
Chase Creek
TRIBUTARIES
COMBINED
M ACROHABITAT
TYPES
UPLAN.D SLOUGHS
Figure 14. Density distribution of juvenile chum salmon by macrohabitat
type on the Susitna River between the Chulitna River
confluence and Devil Canyon, May through October 1983.
Percentages are based on mean catch per cell.
l. Portage Creek (,,11 sites)
;!. Slough 22
3. Slough 21
4. Side Channel 21
5. Indian River (all sites)
6. Slough 11
7. Side Channel 10
B. Side Channel lOA ·
9. Slough 9
10. Slough BA
11. Mains tem 2
12. Slough 8
13. Slough 6A
14. Oxbow One
15: Slough 5
16. Chase Creek
17. Whiskers Creek Slough
SOCKEYE
MEAN CATCH PER CELL
RELATIVE ABUNDANCE KEY
o.ooD
0.01-0.25 c:J
o. 26-5.00 ii
> 5.00.
-No Sample
JULY AUG. SEPT. OCT.
I II I. II I II I
r. -D D iii D D D D D D 0
""~ "-r 5.
~16.
I 7,
--DDDDD-D
DCJiii-iii-Cl-D-
-D-0-0.,... 0
DO -DDDDDO D
ii··---iii-D
-DDODD
-CJD-OOD
OOiii-DiiiiiiDO
Q-iii-D0-0
-0-ii-L:Jii 'ii DO
-ii---iiiiiCiiii-
Oiii-iii-···iiiD
-----iiiDiiiD-
-0 -iii -iii ill 0 iii 0
-o-D-D-o
---iii-iiiiiiCJDO
0
Figure 15. Seasonal distribution and relative abundance of juvenile
sockeye salmon on the Susitna River between the Chulitna
River confluence and De vi 1 Canyon, May through November
1983.
-22-
....
-
-
-
.....
-
I"""
~~
-
25
:c: u
~
<( 20 u
.....1
<(
~
0
~
I.L..
0
~ z w u a: w a.. 5
Figure 16.
6.0
5.0
.....1
.....1 w u 4.0
a:
w a..
:c: 3.0 u
~
<(
u
z 2 .0 <(
w
~
I 0
0
n= 1,010
MAY JUNE JULY AUGUST SEPT. OCT.
Percentages of the total juvenile sockeye salmon catch b
sampling period, May through October 1983.
ill
I -·~
'!$ ~\
~ ' "
'~ I
JUNE
v
¥· 1i~
lt~
JULY
!!] -Side Slougns
0 -Upland Slougns
• -Side Cnannels
• -No Effort
0 -No Catcn
T r.-Trace
n: 1,010
AUGUST OCT.
Figure 17. Juvenile sockeye salmon mean catch per cell at three macro-
habitats by sampling period, May through October 1983.
-23-
I
N _p,
I
J .I
Nina SlouQh&
Combined
SIDE SLOUGHS
Seven Mainstem
Sites Combined
MAIN STEM Sl DE
CHANNELS
s louoh ~
Mainatem II
Slough 19
8.4%
TRIBUTARIES
COMB I NED MACROHABITAT
TYPES
UPLAND SLOUGHS
Figure 18. Density distribution of juvenile sockeye salmon by
macrohabitat type on the Susitna River between the Chulitna
River confluence and Devil Canyon, May through October 1983.
Percentages are based on mean catch per cell.
J
-
Both macrohabitat type and sampling period were significantly linked to
the distribution of all four species (Table 2). These results lend
credence to the pie charts presented earlier in this section in which
the catch per cell for each species is compared among different macro-
habitat types and sampling periods. All species show preferences for
certain macrohabi tat types over others. They a 1 so exhibit seasona 1
differences in their distribution.
The analysis suggests that mean catches/cell for chinook and coho were
significantly different for different levels of turbidity. The power of
the analysis to detect significant differences ·in depth, velocity, and
percent cover was weakened because of the non-randomness of the cells
from which the means of these three variables were calculated. The
effect of percent cover is compounded by the fact that fish use turbid-
ity as cover. Because of many empty cells in the analysis of various
table, interactions among variables were not calculated. Consequently,
conclusions about the parameters other than macrohabitat type, sampling
period, and turbidity are provisional.
-25 -
------------------------------------------------------------------·
Table 2. Results of analysis of variance of juvenile salmon catch/cell
by selected habitat variables. A parameter is considered to
be significant if the probability is less than 0.10. The
first two parameters were run together and then the next five
parameters were run together. Catch/cell was the response
variable in both runs.
Probabi 1i ties for each Seecies
Parameter Chinook Coho Chum Sockeye
Macrohabitat type 0.00 0.00 0.09 0.01
Sampling period 0.00 0.00 0.00 0.01
Mean depth 0.42 0.01 0.53 0.47
Mean velocity 0.01 0.87 0.87 0.05
Mean percent cover 0.24 0.40 0.43 0.51
Water temperature 0.35 0.21 0.37 0.32
Turbidity 0.03 0.02 0.60 0.98
-26 -
~
-
~
~
-
-
-
4.0 DISCUSSION
4.1 Limitations of the Data
4.1.1 Sampling limitations
The macrohabitat types depicted in the pie charts do not include the
mainstem macrohabitat, a type which constitutes a large portion of the
wetted surface area in this reach of river. The mainstem was not
included because of the difficulty in effectively sampling deep, fast,
turbid water for juvenile salmon and because these high velocity waters
have little potential for rearing salmon. The side channels which were
sampled were relatively small, near shore side channels, with riparian
vegetation and often with some kind of clear water input such as a small
tributary, an upwelling area, or hillside runoff. Large portions of the
surface area of the river which can be classified as side .channel are
larger or mid-channel side channels which are devoid of cover other than
substrate. Also, the heads of side channels where the best data were
collected as a rule tend to overtop at a higher level of discharge than
many mid-river side channels. Therefore, the fish collection side
channels were actually side sloughs a higher proportion of the time than
are many of the mid-river side channels.
The overall distribution of juvenile salmon in this reach of river can
be classified as a contagious (clumped) distribution. There are areas
of fish concentrations in areas such as natal sloughs or tributary
mouths and there are other areas where fish density is much lower.
Sampling sites have not been selected randomly throughout the reach.
The Susitna River has clear water sloughs and tributary mouths and
vegetated side channels interspersed amongst large areas of fast, turbid
mainstem water. These main channel areas are important as pipelines
between rearing areas and as an outmigration corridor. Their overall
value as rearing areas is unknown but the amount of rearing habitat in
these areas is limited by velocity.
4.1.2 Gear efficiency
Minnow traps, beach seines and electrofishing equipment have been used
extensively as sampling methods for conducting fisheries surveys
(Bennett 1970; Delaney et al. 1981; ADF&G 1981b, 1983c). However,
minnow traps are se 1 ecti ve for j uveni 1 e chi nook and coho sa 1 man and
beach seining and electrofishing appear to be selective for smaller
sized juvenile salmon (ADF&G 1983c). Burger et al. (1982) and Dauble
and Gray (1980) have concluded that beach seining and electrofishing,
when used in conjunction, provide a reliable index of species diversity,
distribution, and relative abundance for juveniles of all salmon species
except pink salmon. Minnow traps were not used in the Juvenile
Anadromous Habitat Study (JAHS) in 1983. However, as with any sampling
technique, the data collected were affected by gear bias and
limitations. Electrofishing and beach seining methods were sometimes
difficult to use in sampling the entire range of the available habitat
utilized by juvenile salmon.
-27 -
Results from two preliminary gear efficiency experiments presented in
Appendix B indicate that (1) the capture efficiency of electrofishing
decreases as percent cover increases and {2) that beach seining was more
effective in water with high turbidity and electrofishing was more
effective in water with low turbidity. However, these experimetns are
not considered to be definative tests. Until these experiments can be
repeated with a larger number of cells for all salmon species, we
consider the above findings preliminary.
Differences in gear efficiency undoubtedly exist, however these
differences are thought to be small in comparison to the seasonal
variation in numbers of fish at a given site and the variations in
numbers of fish among sites.
4.2 Chinook Salmon
The low numbers of age 1+ chinook salmon captured can be attributed to
sampling gear bias and to the outmigration of this age class from the
study area before July 15. Outmigrant trap data collected during the
same time period indicated that a higher number of age 1+ chinook were
present in the study area above the Chulitna River and subsequently
rearing in the four macrohabitat types than the data from the dis-
tribution study indicated. Seven percent of the seasonal catch at the
outmigrant trap consisted of age 1+ chinook. Of course, since age 1+
chinook would be most 1 ikely to outmigrate, one would expect a higher
proportion of age 1+ chinook at an outmigrant sampling location.
Early in the summer, densities {fish per cell) of the two age classes of
chinook salmon were considerably higher at tributaries as compared to
upland sloughs, side sloughs, and side channels. Tributaries provided
the highest concentrations of chinook early in the summer with side
channel concentrations increasing in July.
Heavier cover in tributaries and the turbidity in side channels probably
reduced gear effectiveness. The data presented reflect min·imum den-
sities at those sites. The effects of gear efficiency were probably not
as important at side sloughs. In general, sites which represented this
macrohabitat type such as Slough 22 and Whiskers Creek Slough, consisted
of shallow, relatively clear water habitats with low to moderate cover
which permitted effective use of electrofishing gear.
Densities of age 0+ chinook salmon were higher at side sloughs from July
through November than before July. Lower densities at side sloughs
before June were due to the tributary outmigrations which had not yet
occurred.
One percent of the seasonal catch was collected in upland sloughs.
Preference for habitat conditions that optimize rearing and proximity of
study sites to natal tributaries were the two major factors which
affected distribution. Previous studies conducted by Delaney and Wadman
(1979), ADF&G (1983c), and Burger et al. {1983) concluded that the
preferred habitat included moderate water velocities and water depths.
Low densities of chinook salmon at upland sloughs may have resulted from
the avoidance of this habitat type because of their preference for areas
-28 -
....
-
-
-
-
-
-
-
-
.....
-
,. ..
with moderate flow. The analysis of variance confirmed this preference.
(See also Part 3 of this report which presents suitability criteria
curves for each species).
Habitat conditions at side channels were more favorable for chinook
salmon juveniles and, consequently, significantly more fish were found
rearing in this habitat type. Fish collected from side channels were
actively feeding at these sites although they were never directly
observed in this activity. Examination of stomach contents indicated
that some feeding was occurring at these sites in spite of the relative-
ly high water turbidity. Turbidity was found by the analysis of vari-
ance to be a significant factor affecting distribution. We have ob-
served that chinooks in side slough/side channels such as Slough 22 are
widely distributed at the site when the head is overtopped and the water
is therefore turbid. When the head is no 1 anger overtopped and the
water clears, the fish either move to the available cover such as cobble
or leave the site.
Chinook salmon juveniles occurred in large numbers at tributary sites,
because these fish originated in these tributaries and were rearing to
attain sufficient size prior to dispersing into side channel or side
slough macrohabitat •
The high densities of chinook juvenile salmon observed at side sloughs.
in September was a response to changes in side channel conditions.
Decreasing side channel water temperatures may have stimulated chinook
juveniles to migrate into side sloughs where conditions were more
favorable for over-wintering. Also, as mainstem discharges decreased,
some side channels, which harbored large numbers of juveniles, became
side sloughs and fish moved into any available cover or outmigrated. It
can be speculated that they may have stayed in higher densities than
waul d norma 11y occur when temperatures were higher and there was more
competition for ava i 1 ab 1 e food. Although water temperature was not
found by the analysis of variance to be a significant factor in affect-
ing chinook distribution during the open water season, our observations
suggest that temperature is a factor during the fall re-distribution.
A comparison of outmigration from the tributaries or out of the lower
river may provide some insight as to how catch rates are related to
migration. Two peaks in catch rates for chinook juvenile salmon oc-
curred at the four macrohabitat types and the outmigrant trap located at
RM 103.0 (Figure 19). The first peak in catch rates was recorded at
tributary study sites in early July. Large numbers of age 0+ fish left
the natal tributaries to redistribute into the other major macrohabitats
(upland sloughs, side sloughs, and side channel). Some of these fish
outmigrated from the study area above the Chulitna River. A second peak
in catch rates occurred at tributaries and the outmigrant trap in mid
August. A substantial number of the juvenile chinook salmon in August
apparently moved into mainstem associated areas as catches at these
locations peaked in late August. Although overall catch rates declined
in September for juvenile chinook in the study area, relatively high
densities were recorded at side sloughs at this time. Apparently, fish
were immigrating into side sloughs to overwinter prior to freeze up
possibly because of the warmer temperatures associated with upwelling
groundwater in the side sloughs.
-29 -
1.1..
0
w
(!)
<t
f-w
z::l
Wa_
uu
0:: w
Q...J
(f)<t
<tz
wo
I ::l(f)
w a..<t
'C) uw I (f)
>-
.J
~ w w
3: -CD
] l
300
250
200
150
100
I
50 I
I
I
I
I
I
I
I-15 16-30
MAY
I
I
I
I
/
/
I
/
/
/
~
I \
I \
/ \
I \
I I I
I
I
I
I
/
/
/
\
\
\
\
\
\
1-15 16-30 1-15 16-31
JUNE JULY
/'
OUT MIGRANT
TRAP
----TRIBUTARIES
--MAINSTEM SIDE
' ' '
CHANNELS 1 SIDE
SLOUGHS AND
UPLAND SLOUGHS
\
\
\
\
1-15 16-31 16-31
AUGUST
Figure 19 Seasonal deviation of catch per unit effort of juvenile
chi nook sa 1 mon on the Sus itna River be tween the Chu 1 itna
River confluence and Devil Canyon, May through September
1983.
I
-
-I
A decline in catch rates was reported by Riis and Friese (1978) at
tributaries and side sloughs. Furthermore, Riis and Fries concluded
that juvenile chinook overwinter in side channels as opposed to tribu-
taries or side sloughs. However, the conclusions were based on a small
sample size. Surveys conducted in October and November 1983 by the
present study encountered substantial numbers of chinook juvenile salmon
utilizing tributaries, side sloughs and, to a lesser extent, side
channels. ·
Although exact comparisons of the relative abundance of chinook salmon
fry among the three open water seasons sampled to date cannot be made
because of different gear and effort it is apparent that 1982 was a year
of low abundance of chinook juveniles in this reach, relative to 1981
and 1983.
4.3 Coho salmon
Juvenile coho salmon were distributed primarily in tributaries, upland
sloughs, and side sloughs associated with the Susitna River above the
Chulitna River confluence. The highest densities of juvenile coho were
found in natal tributaries such as Chase Creek and Indian River which
were documented as spawning areas for adult coho salmon by ADF&G
(1983b). Tributaries are only affected by changes in Susitna River
mainstem flows at areas located near the mouths of the tr·ibutaries
{ADF&G 1983c). Consequently, macrohabitat types which are critical
rearing areas for juvenile coho salmon and were affected by mainstem
flows consisted of upland sloughs and side sloughs. Changes in flows·
can affect access to and usability of these sloughs and consequently the
distribution and abundance of juvenile coho.
Upland sloughs, such as Slough 6A {RM 112.3) and Slough 5 (RM 107 .6),
and side sloughs are generally warmer than mainstem side channels or
tributaries. Delaney and Wadman (1979) and Northcote (1969) concluded
that warmer water attracted juvenile salmonids. Furthermore, Balchen
(1976) argued that fish migration and redistribution was a behavioral
response to seek optimal temperatures to maximize 11 comfort 11
•
Upland sloughs probably enhance the survival of coho juvenile salmon by
providing shelter from high discharges common for the Susitna River
durin~ the summer months. Skeesick (1970) and Cederholm and Scarlett
(1981) concluded that juvenile coho immigration into lateral tributaries
and riverine ponds was a behavioral response to high mainstem flows, to
assure the viability of individuals under adverse flow conditions, and
to escape high flow levels and turbid water.
Side sloughs and upland sloughs are generally clear to slightly turbid
water environments, in contrast to mainstem or side channel water.
Water clarity in the sloughs is not affected by turbidity levels in the
mainstream Susitna River, except at backwater zones near the mouths of
these macrohabitat types. Juvenile coho apparently immigrate into these
macrohabitat types for rearing, since mainstem turbidity levels within
the 70-100 NTU range may impair feeding (A 1 abaster 1972; Bisson and
Bilby 1982). Sigler et al. (1984) found, in a laboratory study, that
turbidity as low as 25-50 NTU caused a reduction in juvenile coho salmon
growth; also, more coho juveniles emigrated from channels with this
-31 -
----------------~----------·----------------------·--
level of turbidity than from channels with clear water. The analysis of
variance confirmed the preference of Susitna River juvenile coho for
waters with a lower turbidity level.
Studies conducted by Delaney and Wadman (1979) in the Little Susitna
River found high densities of post emergent fry near the spawning areas
of adult coho salmon from April through June. After that, the fry
disperse from the redds.
Substantial increases in coho fry density at upland sloughs and, to a
lesser degree, at side channels were detected during the same sampling
periods when high densities were recorded for tributaries. Increases in
the number of coho juveniles occurred in late July at Slough 8, Slough
6A, and Whiskers Creek Slough. Although Delaney and Wadman (1979)
concluded that 60mm was the average length for coho juveniles before
indications of outmigration from tributaries and redistribution into
suitable habitat, data collected in 1983 indicated that mobility size
was considerably less (37mm -45mm). The smaller size age 0+ coho
salmon captured at upland sloughs and side sloughs were fish probably
displaced from natal tributaries because of high flow events, intraspe-
cific competition with other juvenile coho and or interspecific competi-
tion with juvenile chinook salmon. Small coho juveniles were also
captured at the Talkeetna outmigrant trap from late June through July.
The deviations in catch rates of coho juvenile salmon are compared
between tributaries, mainstem influenced macrohabitats, and the
Talkeetna outmigrant trap (RM 103.0) in Figure 20. Although direct
comparisons of catch rates are impossible, because of the different
units used to calculate catch per unit effort (catch/hour, trap;
catch/cell, macrohabitat types), an examination of variability in the of
catch rates gives some indication allows comparisons of seasonal abun-
dance.
The distribution and outmigrant patterns do not provide clear trends.
Catch rates at the sites sampled in both tributaries and adjacent to the
mainstem had similar catch rate variations but were not duplicated at
the outmigrant traps.
Outmigrant trap catch rates declined sharply after mid August as
compared to catch rates at side and upland sloughs during the same time
period. This decline at the outmigrant trap may be attributed to
redistribution of coho juvenile salmon into suitable rearing macrohabi-
tat at sites above the location of the trap or a decline in the number
of age 0+ coho outmigrating from the upper reaches of the Susitna River.
The higher rates of catch recorded at habitats adjacent to the mainstem
suggest use of these areas for wintering.
Catch rates of coho juveniles generally declined at all macrohabitats
sampled from summer to winter. Similar decreases in catch rates were
also reported by Riis and Friese (1978) at tributaries and side sloughs.
Furthermore, Riis and Friese concluded that coho juveniles probably over
winter in mainstem sidechannel s, as opposed to tributaries or side
sloughs because of reductions in rearing habitat resulting from lower
flows. However, data co11ected during the 1981 through 1983 studies
-32 -
-
-
-
-
....
-
-
,....
f.!"llllilll.
-
-
-
-
450
OUTMIGRANT
TRAP
----TRIBUTARIES
400 ~ --MAINSTEM SIDE
I I CHANNELS, SIDE
I I SLOUGHS AND
I UPLAND SLOUGHS I I
1 I
I I
LIJ I I
::J r I ~ I I (..)
' I
...J I I
<{ I I A z I I 0 I I I \ en
<{ I I
LIJ I I I \ en I I
I I I 1.1... 25 I I \ 0
I I
LIJ I I I \ (!) I I <{ A 1-I I I z I I \ LIJ I I \ I (..) I I 0:: I w I I \ I
~ I I
en I I \ I I <{ I I
LIJ : I \ I I I ::J I I ~ l I \ (..) I . I
~ f I \ I I I /
I / :X: \ \ I
LIJ :I I --/ w I :: :, \ I /,
m I
~ \ \ .,
\V , .......
' i ,. ' , ' \ ,. , ' -~'.J \ ' v '
0
16--31 1-15 16-30 16-31 1-1 5 16-31 1-15
MAY JUNE JULY AUGUST SEPT.
Figure 20. Seasonal deviation of catch per unit effort of juvenile coho
salmon on the Susitna River between the Chulitna River
confluence and Devil Canyon, May through September 1983.
-33-
(ADF&G 1981b; 1983c) indicate that substantial winter rearing occurs in
side sloughs and upland sloughs.
Studies conducted by Peterson {1980) indicate that upland slough coho
juveniles incur a much lower winter mortality than the typical stream
resident. In the winter, juvenile salmon are inactive and hide in the
gravel or deep pools, ensuring that they are not carried out of the
system (Thorpe 1981).
4.4 Chum
An accurate record of the true distribution of juvenile chum and sockeye
salmon may not be shown by 1983 data due to biases associated with the
sampling techniques. During this and previous studies, beach seining
and electrofishing have been the two most effective methods of
collecting juvenile chum and sockeye salmon (ADF&G 1981b, 1983c). Beach
seining and electrofishing efficiencies are directly correlated to
mainstem discharge and turbidity levels at many macrohabitat locations.
Burger et al. (1982) found that as the discharge and turbidity of the
Kenai River increased, electrofishing efficiency decreased while beach
seining efficiency increased. Comparisons of this year's data with
previous year's studies on the Susitna River are also biased. During
the 1981 Juvenile Anadromous studies, CPUE's were based mainly on minnow
trapping, with only a minimal amount of beach seining effort. Minnow
trapping is not an effective method of capturing juvenile chum and
sockeye salmon.
A total of 1,174 juvenile chum salmon were captured in 1983 above the
Chulitna River, while 1,104 were captured in the same reach in 1982.
All of the sites where chum salmon were collected during 1982 studies
and which were sampled in 1983 again produced juvenile chums (ADF&G
1983c).
Tributaries and side sloughs accounted for 92% of the total juvenile
chum catch in 1983, of which 92% were captured in natal sloughs and
tributaries. In 1982, a large school of fish captured at upland slough
6A accounted for 81% of the total catch for all macrohabitat types.
This uneven distribution creates biases in results when catch per unit
effort data are used.
Although upland sloughs accounted for only 1% of the total catch, visual
observations both within and outside the designated study areas and 1982
catches (ADF&G 1983c) confirmed that juvenile chum use upland sloughs
for rearing, as do sockeye juveniles.
High velocity side channel and mainstem environments are not considered
prime rearing areas for juvenile chum salmon. Juvenile chums are
captured in the mainstem, but usually in lower velocity backwater zones.
Basically, juvenile chum salmon were found in high densities in natal
side sloughs and tributaries early in the season (May-early June) and in
upland sloughs and side channels in late June and July. After July,
catches and observations of juvenile chums within any of the macro-
habitats were extremely rare. Chum sa 1 mon catches at the downstream
-34 -
-
-~
-
-
-
-
-
migrant traps also plummeted after mid-July, indicating· that the bulk of
the outmigration had taken place {see Part 1 of this report).
Figure 13 illustrates the possibility of two distinct outmigrating
juvenile chum populations; one from the natal sloughs in late ~1ay and
one from the tributaries in early July. These peaks correspond with
peak catches at the downstream migrant traps (See Part 1 of this re-
port). Although the tributary chums generally spawn earlier than the
slough populations {ADF&G 1983b), the colder intragravel temperatures
found in the tributaries in the winter (Estes and Vincent-Lang 1984)
could account for a delayed emergence and outmigration.
Juvenile chums have been found to prefer the shallower, flowing waters
of side sloughs and upland sloughs, as opposed to the low flow, deeper
pools preferred by juvenile sockeye. Juvenile chum salmon were more
widely distributed than sockeye juveniles during 1983, the reason being
that chum salmon spawn in more sloughs than sockeyes. This was also
true in 1982 (ADF&G 1983b).
Although tributaries are not affected by mainstem flow, except at the
confluence, higher mainstem flows usually occurred at times of higher
tributary flows. Higher tributary flows acted as a flushing device,
with fewer fish being present in natal areas and more fish being present
at rearing and outmigrating areas after the high flows.
The first major peak of mainstem discharge in May coincided with the
highest juvenile chum catch rates. By the time the peak mainstem
discharge occurred in early June, the majority {62%) of the total
juvenile chum catch had already occurred. Juvenile chum salmon from
natal sloughs tend to take advantage of the first major rise in mainstem
discharge and start outmigrating. This was also true in 1982 when the
last juvenile chum was observed by mid July (ADF&G 1983c). The exact
stimulus for outmigration is not known, but is probably a combination of
innate behavior, increased cover (turbidity), increased water
temperatures and the higher flows. Few juvenile chum were captured at
tributary sites until early July, after the peak spring discharge in the
mainstem. Similarly, few chum juvenile were captured (using the same
methods) until late June in 1982, well before the peak mainstem
discharge.
4.5 Sockeye Salmon
Gear bias also affected the catch data for sockeye salmon. Beach
seining on the Kenai River, in areas where no sockeye juveniles were
captured in minnow traps, proved that sockeye were present (Burger et
al. 1982). The 1983 catches by location in the Susitna River can be
loosely compared with 1982 data, as beach seining was the main sampling
method used in 1982. Juvenile sockeye salmon have been found to school
in the clear waters of some of the side sloughs. Often, schools were
observed just prior to sampling, but unavoidable disturbances caused the
fish to move out of the sampling grid and few, if any, would be
captured. Sockeye juveniles· were also observed to use the deeper pools
and interstitial spaces in the larger substrate. Due to their depth,
-35 -
many of the deeper pools were inaccessible to effective sampling. Fish
using substrate as cover might remain within the substrate during
electrofishing and beach seining passes and, once again, the data would
not reflect this presence.
A total of 1010 juvenile sockeye salmon were captured in 1983 above the
Chulitna River. Distribution within this reach was similar in both 1982
and 1983, with 57% and 66% of the total catch occurring above RM 125.0
during 1983 and 1982, respectively. All of the sites where sockeyes
were collected during 1982 sampling, were found to contain sockeye in
1983.
Side sloughs accounted for 71% of the total juvenile sockeye catch in
1983, of which 65% were captured in natal sloughs. Side sloughs ac-
counted for 31% of the total catch during 1982. The major reason for
this lower number during 1982 is the large number of fish captured at
the upland slough, Slough 6A, (62% of the total catch for all habitat
types). These differences are probably a result of collection method-
ology rather than any major difference in distribution between years.
Upland sloughs accounted for 20% of the total catch in 1983, with the
highest catch rates occurring late in the summer (July-August). A
distinct redistribution of sockeye juveniles from side slough natal
areas to upland slough rearing areas at this time can be seen in Figure
18. Slough 6A, the major upland slough used by outmigrating and/or
rearing sockeye juveniles, accounted for 86% of the total upland slough
catch. Juveniles sockeye generally rear in lakes although slough
populations are not uncommon (Foerster 1968, McCart et al. 1980). With
the exception of the unique habitat at Slough 6A, including low veloci-
ty, clear water, depth and abundant cover and aquatic vegetation, major
concentrations of juvenile sockeye salmon were found in natal side
sloughs. Slough 5, an upland slough with shallow depths and low
gradient banks, did not have large numbers of sockeye. This slough was
broadly covered with emergent vegetation.
With the exception of backwater areas, side channel and mainstem en-
vironments are not used extensively as rearing areas by juvenile
sockeye. Mainstem 2 and Oxbow I are both side channels that were
breached during much of the 1983 season and both had these backwater
zones. Sockeye juveniles were captured at both of these sites. The
preference of sockeye juveniles for low velocity water was clearly
demonstrated by the analysis of variance.
Tributary spawning by sockeye salmon is rare in the Chulitna confluence
to Devil Canyon reach. During the past three years, six adult sockeyes
have been observed in the tributaries, four of them in Portage Creek
during 1982 (ADF&G 1981a, 1983b; Barrett et al. 1984). Few juveniles
have been captured in tributaries during the past three years because of
this lack of tributary spawning (ADF&G 1983c). Basically, juvenile
sockeye salmon in the study reach primarily use side and upland sloughs
for rearing.
Two of the major natal areas of sockeye salmon (Sloughs 9 and 21) were
directly affected by mainstem discharges overtopping the head of the
sloughs in 1983. Slough 11, the major sockeye spawning area in the
-36 -
-
~'
-
-
-'
-
upper Susitna River is only breached by very high flows, the last time
in 1981 (ADF&G 1981c). Small changes occur at the mouths of side
sloughs which are not breached, with increases in depth, turbidity, pool
sizes and cover occurring at higher flows. Sockeyes have been found to
utilize lower velocities and greater depths than the other juvenile
salmon species. (See Part 3 of this report).
As mainstem discharges increase in May and June, catch rates also
increased (Figure 16). The peak catch rate in the primary natal sloughs
occurred in early June when the discharge was at its seasonal peak of
34,000 cfs. Sockeye juveniles may respond to increases in water depth,
velocity, and turbidity in the breached slough (now a side channel) by
outmigrating. Whatever ~he stimulus, lower catch rates in natal sloughs
after head breaching reflects outmigration.
Intraspecific competition for available rearing habitat could also
initiate outmigration. The highest catch/hour of sockeye juveniles at
the downstream migrant trap occurred in early July, corresponding to the
highest catches at natal sloughs before July and at rearing sites during
and after July.
Observations at rearing sites and downstream migrant catch data indicate
that some overwintering in this reach by juvenile sockeye salmon does
occur. Age 1+ sockeye were captured and observed in Slough 11 during
1981, 1982 and 1983. The downstream migrant trap juvenile sockeye
catches included 1.1 and 0.7 percent catches of Age 1+ fish in 1982 and
1983, respectively. During the past three years of study, Age 1+
sockeyes have been observed at Slough 9, Slough 11 and Slough 6A (ADF&G
1981b, 1983c).
The capture at non-natal sites of juvenile sockeye during August and
September that were coded wire tagged in early June suggests that
overwintering in sloughs 6A and 11 and presumably other sites may occur.
Sockeye 0+ fry have been observed to remain in the shallower waters near
shore both in rearing areas and while outmigrating early in the summer.
As they grow, they start using the deeper waters. Age 1+ fish, if they
follow the same pattern, may be using the deepest waters of the macro-
habitats for both rearing and outmigrating and therefore would not be
susceptible to our sampling technique.
-37 -
5.0 CONTRIBUTORS
Field work for the project was conducted by Larry Dugan, Paul Suchanek,
Bob Marshall, and Dave Sterritt.
Dana Schmidt and Steve Hale assisted with the study design and analysis.
The data base was keypunched by Donna Buchholz and managed by Allen
Bingham, Gail Heineman, and Alice Freeman.
The analysis of variance section was prepared by Allen Bingham and Steve
Hale. Steve Hale and Paul Suchanek wrote the section on analysis of
gear efficiency.
Bruce Barrett and Larry Bartlett reviewed the draft copy of this paper
and provided helpful comments.
Sally Donovan and Carol Kerkvliet drafted the figures and the typing was
done by Skeers Word Processing.
-38 -
.....
-
-
-
-
-
-
-
-
-
6.0 ACKNOWLEDGEMENTS
Funding for this study was provided by the Alaska Power Authority.
We wish to thank the various staff of E. Woody Trihey and Associates
(EWT&A), the Arctic Environmental Information Data Center, Harza-Ebasco
Susitna Joint Venture, Woodward-Clyde Consultants, the Alaska Power
Authority, and especially Cleve Steward (EWT&A) for reviewing the draft
of this paper and providing helpful suggestions.
-39 -
7.0 LITERATURE CITED
Alabaster, J.S. 1972. Suspended solids and fisheries. Proceedings of
the Royal Society of London. B. 180: 395-406.
Alaska Department of Fish and Game (ADF&G). 1979. Preliminary final
plan of study. Fish and wildlife studies proposed by the Alaska
Department of Fish and Game. Alaska Department of Fish and Game.
Anchorage, Alaska.
1981a. Phase I final draft report. Subtask 7.10. Adult
anadromous fish study on the lower Susitna River. Alaska Depart-
ment of Fish and Game Susitna Hydro Aquatic Studies. Anchorage,
Alaska.
__ • 1981b. Phase I final draft report. Subtask 7 .10. Juvenile
anadromous fish study on the lower Susitna River. Alaska Depart-
ment of Fish and Game Sus itna Hydro Aquatic Studies. Anchorage,
Alaska.
• 1981c. Phase I final draft report. Volume 1. Subtask 7 .10.
--Aquatic habitat and instream flow project. Alaska Department of
Fish and Game Susitna Hydro Aquatic Studies. Anchorage, Alaska.
--,· 1983a. Aquatic studies procedures manual. Phase II (1982-83)
Subtask 7 .10. Alaska Department of Fish and Game Susitna Hydro
Aquatic Studies. Anchorage, Alaska.
__ . 1983b. Susitna Hydro aquatic studies phase II final report.
Volume 2 (2 parts). Adult anadromous fish studies, 1982. Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies.
Anchorage, Alaska.
1983c. Susitna Hydro aquatic studies phase II basic data
report. Volume 3 (2 parts). Resident and juvenile anadromous fish
studies on the Susitna River below Devil Canyon, 1982. Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies.
Anchorage, Alaska.
__ • 1983d. Susitna Hydro aquatic studies phase II report. Synopsis
of the 1982 aquatic studies and analysis of fish and habitat
relationships (2 parts}. Alaska Department of Fish and Game
Susitna Hydro Aquatic Studies. Anchorage, Alaska.
--• 1984. Susitna Hydro Aquatic Studies May 1983 -June 1984
procedures manual. Alaska Department of Fish and Game Susitna
Hydro Aquatic Studies. Anchorage, Alaska.
Balchen, J.G. 1976. Principles of migration in fishes. Teknisk notat
nr. 81:33p. SINTEF, Trondheim, Norway.
Barrett, B.M., F.M. Thompson, and S.N. Wick. 1984. Adult Anadromous
Fish studies: May-October 1983. Alaska Department of Fish and
-40 -
-
-
-
Game. Susitna Hydro Aquatic Studies Report No. 1. Prepared for
Alaska Power Authority. Anchorage, Alaska.
Bennett, G.W. 1970. Management of lakes and ponds. Van Nostrand
Reinhold Company. Second Edition. New York, New York.
Bisson, P.A., and R.E. Bilby. 1982. Avoidance of suspended sediment by
juvenile coho salmon. North American Journal of Fisheries Manage-
ment. 4:371-374.
Burger, C.V., D.B. Wangaard, R.L. Wilmot, and A.M. Palmisano. 1982.
Salmon investigations in the Kenai River, Alaska 1979-1981. U.S.
Fish and Wildlife Service, Anchorage, Alaska.
Cederholm, C.J., and W.J. Scarlett. 1981. Seasonal immigrations of
juvenile salmonids into four small tributaries of the Clearwater
River, Washington. Pages 98-110 In: E.l. Brannon and E.O. Sale,
editors. Salmon and Trout Migratory Behavior Symposium. Universi-
ty of Washington, Seattle, Washington.
Dauble, D.O., and R.H. Gray. 1980. Comparison of a small seine and a
backpack electroshocker to evaluate nearshore fish populations in
rivers. The Progressive Fish Culturist 42{2): 93-95.
Delaney, K.J., K. Hepler, and K. Roth. 1981. Deshka River chinook and
coho salmon study. Alaska Department of Fish and Game, Division of
Sport Fish. Federal Aid in Fish Restoration, Project AFS-49, Vol.
22.
Delaney, K.J., and R. Wadman. 1979. little Susitna River juvenile
chinook and coho study. Alaska Department of Fish and Game.
Division of Sport Fish. 41 pp.
Estes, C.C. and D.S. Vincent-lang editors. 1984. Aquatic habitat and
· instream flow investigations, May-October 1983. Alaska Department
of Fish and Game Susitna Hydro Aquatic Studies. Report No. 3.
Prepared for Alaska Power Authority. Anchorage, Alaska.
Foerster, R.E. 1968. The sockeye salmon Oncorhynchus nerka. Bulletin
of the Fisheries Research Board of Canada. 162.
McCart, P.J., D.W. Mayhood, M.l. Jones, and G.J. Glora. 1980.
-Iskut Fisheries Studies, 1979. Report prepared for
Columbia Hydro and Power Authority. P. McCart Biological
tants Ltd. Nanaimo, British Columbia.
Stikine
British
Consul-
Morrow, J.E. 1980. The freshwater fishes of Alaska. Alaska Northwest
Publishing Company, Anchorage, Alaska.
Northcote, T.G. 1969. Patterns and mechanisms in the 1akeward
migratory behavior of juvenile trout. Pages 183-203. In: Sympo-
sium on Salmon and Trout in Stream, T.G. Northcote, (ed.): Univer-
sity of British Columbia, Vancouver, British Columbia.
-41 -
Peterson, N. P. 1980. The ro 1 e of spring ponds in the winter eco 1 ogy
and natural production of coho salmon (Oncorhynchus kisutch) on the
Olympic Peninsula, Washington. M.S. Thesis, University of
Washington, Seattle.
Platts, W.S., W.F. Megahan, and G.W. Minshall. 1983. Methods for
evaluating stream, riparian, and biotic conditions. Gen. Tech.
Rep. INT-138. Ogden, UT: U.S. Department of Agriculture, Forest
Service, Intermountain Forest and Range Experiment Station.
Riis, J.C., and N.V. Friese. 1978. Preliminary environmental assess-
ment of hydroelectric development on the Susitna River. Alaska
Department of Fish and Game. Div. of Sport Fish and Comm. Fish.,
Anchorage, Alaska.
Sigler, J.W., T.C. Bjornn, and F.H. Everest. 1984. Effects of chronic
turbidity on density and growth of steelheads and coho salmon.
Transactions of the American Fisheries Society. 113: 142-150.
Skeesick, D.G. 1970. The fall immigration of juvenile coho salmon into
a small tributary. Research Report Fisheries Commission of Oregon
2: 90-95.
Thorpe, J.E. 1981. Migration in salmonids, with special reference to
juvenile movements in freshwater. Pages 86-97 In: E.L. Brannon
and E.O. Salo, editors. Salmon and trout migratory behavior
symposium. University of Washington, Seattle, Washington.
-42 -
....
-
-
.....
-
-
-
-
-
APPENDIX A
Summary Statistics for Transformed Catch/Cell Data
-·
. 1"""
-43 -
I
]
Appendix Table A-1. Summary statistics for transformed catch/cell data of each species, by groups for each
habitat parameter.
PA G[ 13 E!MDP10 STATISTICS OF GROUPED JAHS DATA CRJ83011 -BY HAIHTAT VARIAALES
V.ARIARLE GROUPING TOTAL STANDARD ST.ERR COEFF. or s H A L L E s T L A R G E S T
~·o. •• A~E VARIABLE LEVEL FREQUEtJCY ,..EAN DEVIAT!u~ (if MEAN VARIATI~N VALUE Z-SCORE VALUE l-SCORE
1 5 LCHIN LD 1.112 .905 .0785 .81361 o.ooo -1.23 3.9&5 3.15
MACNUM UP SLOUGH 24 ob21t o5Alt • 1.192 .93668 o.ooo -1.07 2.079 2. 49
SISLOUGH 42 • 744 • 703 • 1084 .94480 o.ooo -1.06 2ollt0 1.99
SICHANNE 39 1.233 • 63'1 • .1016 .51431 o.ooo -1.94 2.6'!5 2o54
TR I!lUTAR 28 1.914 1.133 .2141 .59183 o.ooo -1.69 3.<J65 1. 81
PERlt1D · L"AY 15 .33'+ .4':16 .1280 1.48690 o.ooo -.67 1o609 2.57
EJUN 6 .516 .868 .• 3542 1o681 H o.ooo -.59. 2·230 1· 96
LJUN 10 .618 .610 .1929 .98733 0. 000 -1.01 1.504 1o45
EJUL 16 1.629 1. 347 .3367 .82651 o.ooo -1 • :n 3.965 1o73
LJUL 19 1o2'16 • 852 .1955 .68397 o.ooo -1.'16 2.868 1.90
EAUG 18 1.128 • 907 • 21.31 .80364 o.ooo -1.24 3.186 2.27
LAUG 20 1o274 o/:129 .1853 .65014 o.ooo -1.54 2o945 1o90
ESE:P 20 1o 3'-3 .570 ol27't •. H910 .531 -1.'13 2o230 ___ 1.56
LSEP 9 1o 21f 8 o707 • 2356 o5f>622 o262 -1.39 2.542 lo83
I ··rAtJOEP 0.1-0.6 52 1.214 le01B .1412 .83878 o.ooo -1.19 3.965 2. 70 .r:::. .;::.. 0.7-0.9 46 1.188 .883 .1302 .74350 o.ooo -1.34 3.&40 2.78
I 1o0-1o2 17' .779 • u.~ • 18 50 .97957 o.ooo -1.02 2.845 2. 71
1.3-1.5 q .&87 .8'+8 • 2<l2B .95620 o.ooo -1.05 2o701 2o1't
1. 6+ Cl o993 .472 • 15 72 oliHA9 a. no -2.11 1.649 1· 39
MEA"JCOV o-sx 71 1.100 • 796 • 0944 .72306 o.ooo •1.38 3.186 2.62
f>-25X 53 1o255 1.042 .1431 .82996 o.ooo -1.21 3.965 2.60
26-100X 9 .364 • 389 .1298 1o07142 o.ooo -.93 1o099 1o 89
M[ANVEL o .. o-o.5 103 .995 o860 .OB48 .86494 o.ooo -1.16 3.965 3.45
o.b• 30 lo 515 .952 .1738 .62821 c.ooo -1.59 3.487 2. 07
SWA TTEMP o.o-s.o 13 1o21!3 .• .751. •. 2 0 ll2 .58429 OoO(IO -1.71 2o542 1. 69
5.1-10.0 63 1o247 1.061 .1336 oR5061 o.ooo •1 o18 3.965 2.5&
10.1+ 5& .925 • 714 o095~ o77173 c.ooo -1.30 3. 640 3·80
TUPA 0-10 85 .987 .9:'18 • 1017 .94969 o.ooo -1.05 3.640 2. 83
>10-50 16 lo 2 07 .744 • 18 59 .•• 61589 o.ooo -1.62 2.701 2. 01
)50-100 f, 1.208 .537 ·2190 .44430 • "70 -1.37 1o841 1 .tB
)100-200 11 1. 66/j • 629 .1896 .37785 .993 -1.07 2o8/t5 1. 88
200+ 10 oR 57 o361 o1142 .42149 .262 -1.65 1.308 lo25
1
RANGE
3.965
2.079
2.14 0
2.!!45
3.96!:1
1o609
2. 2 3 0
1.504
3.965
2.86!1
3.186
2 ·9 4!:1
1ef>9'}
2.279
3. 9 6!:1
3.6'+0
2.94!:1
2.701
1of>'t9
3e1R6
3.96!:1
1.09'J
3.96!:1
3.4 A.,
2o5it2
3o965
3.640
3.~40
2.701
1· 371
1.852
1.0'+6
Appendix Table A-1 (cont.).
I~' IJMOPlD STATISTICS :)f GR CIUPED JAHS
; ~ bLE GROUPING TOTAL
•; A11E VARJ ABLE LEVlL FREQUENCY
l COHO 133
MACNUM UP SLOUGH 24
SJSLOUGH '12
SICHANNE 39
TRIBUTAR 28
PERIOD L"'AY 15
E,JU~I f,
LJUN 10
EJUL 16
LJUL 19
EAUG 18
LAUG 20
ESEP 20
I LSEP 9
-:::0. MEANDEP 0.1-0.6 52
L..1 0.7-0.9 46 I 1.0-1·2 17
1.3-1.5 9
1.6+ 9
l~rANCOV o-sx 71
6-25X 53
2f.-1 oox 9
'IEANVEL o.o-o.s 103
0.6+ 30
SwA TTEMP o.o-s.o 13
5.1-10.0 63
10 ·1 + 56
TURA 0-10 85
)10-50 16
)50-1 00 6
)100•200 11
200+ 1 0
.J
Summary statistics for transfonned catch/cell data of each species, by groups
for each habitat parameter.
OA TA l RJ83011 -BY HABITAT VARIA£1LES
STANDARD ST.ERR CO[ff • Of s M A L L E s T L A R G E s T
MEAN DEVJATIGtJ (:f MEAN VARIATION VALUE Z-SCORE VALUE Z -SCORE RANGE
.587 .899 .0780 1o53114 o.ooo -.&5 3o'l21 3.15 .3 .• 4 21
lo 161 .9'1'1 .1926 .81247 o.ooo -1.23 3.258 2o22 3.256
.361 .715 .1103 1.9BU.3 o.ooo -.so 2oR45 3.'18 2.e.,,
·199 .5&6 .0906 2.81!859 o.ooo -.35 2.380 3.85 2.380
.976 1o105 • 2 oe e 1.1.3132 o.ooo -.as 3.'121 2. 21 3." 21
• 2'1'1 .591 .1526 2.419.38 o.ooo -.'1 1 1o758 2o56 }.758
o.ooo o.ooo o.oooo o.ooooo o.ooo o.oo o.ooo n.oo o.cou
1.256 1. 29'1 .'1092 1.03025 0 • 0 DO -.97 3.'121 1.67 3o'l 21
.127 .368 • 09 21 2.90231 o.ooo -.3'1 1o't82 3. 68 lo'IB2
1.037 1.310 .3 0 05 1·26296 o.ooo -.79 3.258 1.70 3.258
.7'56 .965 • 22 76 1.27631 o.ooo -.78 2.398 1o 70 2 • .39!!
.56'1 • 675 .1509 1.19666 o.ooo -.84 1.988 2.11 1.98tl
o't&9 .707 .• 15 81 1o.50.582 o.ooo -.6& 2ol75 2.41 2o175
.652 • 661 • 22 02 1.01305 o.ooo -.99 1.792 1o 72 1.792
.380 • 712 .0988 1.87315 o.ooo -.53 2.fl't5 3o46 2.8'15
.535 .92'1 .1.36.3 1.72801 0. 0 QO -.58 3.266 2.96 3.2&&
.891 1o120 .2716 1.25738 o.ooo -.8o 3.'121 2.26 3.421
.633 • 710 .2365 1o12067 o.ooo -.89 1o75B 1o58 1.758
1.~33 .998 • .3:325 o69625 o.ooo -1.'1'1 2·6&7 1o2'1 2.~&/
.'I 06 • 7Blf • 09 31 1.93026 o.ooo -.52 3.256 3.6'1 3.25!!
.777 1.037 ol't 24 1.33379 o.ooo -.75 3." 21 2.55 3o421
.897 .581 .1938 .64827 .182 -1.23 1.988 1.88 lo90b
.6'19 .961 .0947 1o'I8171J o.ooo -.67 3.'121 2.88 3. 4 21
.376 .609 • 1112 1.61840 o.ooo -.62 1o 792 2..32 1.792
.558. ..658 ol8 2'1 1.17.850 Oo'JOO -.85 1.792 1.88 1o792
.53'1 .858 .1081 1o605't2 o.ooo -. 62 3.258 3.18 3o25B
.662 1.002 .1339 1.51200 o.ooo -.66 3.'121 2o75 3.421
.76'1 .979 .1062 lo28176 o.ooo -.78 3o't21 2o7I 3 ... 21
.450 .809 .2o2q 1.797'11 o.ooo -.56 2.313 2.30 2.31 j
.2'14 .314 o1281 1.28808 o.ooo -.78 .788 1. 7'1 .788
.2BB .798 .2'107 2.772~9 o.ooo -.36 2o667 2.9!! 2of.67
o.ooo o.ooo J.LOOU o.ooooo o.ooo 0.0(1 r..ooo o. 00 o.ooo
·.· .. I
I
-I=»
"' I
Appendix Table A-1 (cont.).
PAGE 16 BHDPlD STATISTICS Of GkOUPED
VAq JlllllE GROUPING TOTAL
-l
Summary statistics for transformed catch/cell data of each species, by groups
for each habitat parameter.
JAHS OAT A I R.J& 30 ll -BY bABJTAT VARIABLES
STANDARD ST,ERR CnEFF, OF s ~ A L L E s T L A R G !: s T
Nr'. "A~[ VARIABLE LEVEL FREQUENCY MEAN DEVIATION CF MEAN VARIATION VALUE Z-SCJRE VALUE Z-SC.)RE
11> l CI-'UM 133 .21t6 .5Ab .0510 2,391t8J o.ooo -.lt2 2.856 ~.qq
HACNUM UPSL(IUGH 21t .035 ol 01 .0207 2.A6181 o.coo -.35 .'105 3.65
SISLOUGH lf2 olt67 .806 • 121tlt 1.72529 o.ooo -.58 :>,A56 2.9&
SICHANNE 30 • 1 02 o287 • 04 60 2o827A7 o.ooo -.35 lolt35 It. 61t
TRIBUTAR 21! .294 .65P .l21t3 2.23501 o.ooo -.'15 2.715 3.68
PER I 00 LMAY 15 1.0;:>9 l.Cl4-.261P. .98556 o.ooo -1 .o 1 ;>.856 loBO
EJUN f, t.lJO .757 o3089 • 669 33 .095 -1.37 2.001 1. 15
L.JUN 10 .41t8 olt91t • 1563 1.10252 o.ooo -.91 t.ltJ5 2. 0 0
tJUL 1& .248 .673 .16£:2 2.70800 0.001} -.37 2.715 3. 66
LJUL 19 ,(187 • 2 01 • 0 4 f.:? 2.31837 o.ooo -.43 .788 3.1t9
£AUG lll .020 .065 .0152 3.:?47911 ·o.ooo -.31 .262 3.76
LAUG 20 o.ooo o.ooo 0.000(1 o.ooooo 0. 000 o.oo o.ooo o.oo
ESEP 20 o.ooo o.ooo o.oooo o.ooooo o.ooo o.oo o.ooo o.oo
LSEP 9 o.coo o.ooo o.oooc o.ooooo o.o~o o.oo o.ooo 0. 00
11EANDEP o.t-o.6 52 o399 • 774 .1 on 1.93835 o.ooo -.52 2.fl56 3 .t 7
0.7-0.9 '16 ·125 .400 .0590 3.20910 o.ooo -• .31 2. 0 01 4.69
lo0-1o2 17 ·194 .~10 .12n 2.635'17 o.ooo -,38 2.001 3.5'1
1.3-1.5 'I .27';' olf2C .13'11' lo51t322 o.ooo -.65 1o030 1. 81
t.&• 9 .0'19 .too • 0334 2.02522 O,GOO -.49 .2 62 2.13
HEANCOV o -5X 71 .217 ·520 .0617 2.'10068 o.ooo -.lt2 2.60.3 Ito 59
6-25X 53 .327 • 705 • 096£; 2.1589'1 o.ooo -.46 2.856 ~.59
26-1 00" 9 o.ooc o.ooo o.oooo o.ooooo o.ooo o.oo !J,(IOO o.oo
MEANVEL o.o-o.s 10 3 .254 .588 • (15 79 2.:no5e o.ooo -.43 :>.856 llo<t3
0.6+ 30 .216 .600 ol09f. 2. 77718 O.I}OO -.36 2.715 4.16
SIIA TTEHP o.o-s.o 13 .l51t • 555 .1540 3.&0555 o.ooo -.28 2.001 3.33
s.t-to.o 63 • 373 .755 .0951 2.02046 o.ooo -.49 2.856 lo29
10 ol + 56 .}2~ .29'1 • 0392 2.29794 o.ooo -.44 lo435 4.45
TURR 0-l(l 85 .33P • 6% .075!' 2. 06021t o.oo~ -.49 2.P.56 3o62
>10•50 16 .143 .365 • 0913 2.55629 o.ooo -.39 1.435 3.54
)50-100 6 .159 • 39 0 .1593 2olt4949 o.ooo -.41 .956 ;>. 04
>100-200 11 .049 .092 .0277 1.117422 o.ooo -.53 .262 2·32
200+ 10 .010 .030 .0095 3.16228 o.ooo -.32 .095 2.85
RANG[
2.65b
,q():;.
;>.856
1o4J!J
2. 71::.
2ofi5f.
lo9Cf.
lo'IJ~
2.71~
.7P.P.
.26;>
o.OOl1
o.oor
o.oo~
2 ,S Sf
2. 0 01
2. 0 01
lo03l
,Zf;>
2. 6 0!
2 • B !'.~
o.ooo
2.8 'Sf
2.71 ~
2.001
2.856
1 0 4 3::J
2ol\5f
1 • 11]:;,
·"~(
• 2 f.~
• 0 '11.
I .r-.
\ ........
i
Appendix Table A-1 (cont.). summary statistics for transformed catch/cell data of each species, by groups
for each habitat parameter.
PAG[ l'! BMDPJD STATISTICS Of GROUP ED .JAHS DATA IR.J8J011 -BY HABITAT VARIARLES
VARIABLE GROUPING TO TAL STANDARD SToERR COEFF. CF s M A L L E s T L A " G E s T
NO• NAME VARIABLE LEVEL FR EOUENCY ~[AN DEVJATIJN Of MEAN VARIATION VALUE Z-SCORE VALUE Z-SCORE
16 L SGCJ( 133 .300 .621 • 053!' 2.06598 0.'>00 •.411 3. 24(, 4.75
MACNU" UP SLOUGI-l 24 .456 .69/i • H 17 lo52396 o.ooc -.66 2.557 3.03
SJSLuUGH 42 .452 .819 ·1263 1.81076 o.ooo -.55 3.246 3.4)
SICHANNE 39 o245 • "1,3 • 074 2 lo"'8967 o.ooo -.53 ;>.197 4. 21
TR IBUTIIR 2!' .011 • 089 • 0 16P 5.29150 o.ooo -.19 .470 5.10
PERIOD L~1 AY 15 .297 .683 .1763 2.30000 o.ooo -.4 ;:o; 2.6.32 3.'12
EJUN 6 • 8 75 lo201 • 4901 1o37235 o.ooo -.73 3.246 1o98
LJUN 10 o661 .773 o24'11\ 1o1b9'17 o.ooo -.86 2.282 2.10
EJUL 16 .234 .592 • JltBO 2.53521 o.ooo -.39 2.3 61 3o59
LJUL 19 .397 • 653 .1497 1o61\390 o.ooo -.61 1.960 2.'10
£AUG 18 .476 • 783 .1844 1.64385 o.ooo -.bl 2.557 2o6ft
LAUG 20 • 0 7& .139 .0312 1·82'163 o.ooo -.55 .336 1.87
ESEP 20 .t 09 .27& .0617 2o54!'12 o.ooo -.39 lo163 3o82
LSEP 9 • 011 .032 • 0106 3o00000 a.ooo -.33 .095 ;?.67
'IEANDEP 0.1-0.6 52 .279 • 685 • 0950 2.'15872 o.ooo -.41 3.246 ~. 33
0.7~0.9 1+6 ol75 ·380 .0561 2ol7911 o.ooo -. 46 2.197 5.:32
1.0-1.2 17 .35f .553 ·1342 1.55408 o.ooo -.611 1.629 2.30
1.3-1.5 9 • 6 39 of1()2 .2~75 1.25624 o.ooo -.eo 2·282 2o05
1.6+ 9 .627 .973 .32411 1 .• 55257 o.ooo •.64 2.557 1.98
ME ANCOV 0·5X 71 • 2 'I 0 .52'1 .0622 2o18115 o.ooo -·'If> 2.632 11.57
6•25X 53 .:nJ .738 ·1 013 1o97905 o.ooo -.51 3.21+6 3.90
26-100::1: 9 .350 o591 •1970 lob8730 o.ooo -·59 lo6()9 2. 13
MEAtJVEL o.o-o.s J[l3 .?IJ6 • 681\ • 06 711 1·82012 o.coo -.55 3. 24 6 4.20
o.&+ 30 .042 ol36 .02117 3.25£,65 o.ooo -. 31 .588 4.03
SIIA TTEMP o.o-s.o 13 • 0 07 .o 26 ~0073 3.60555 o.ooo -·28 o095 3o33
5.1-10.0 63 .359 • 748 .0943 2.08359 0. 0 0 0 -.48 3o246 3.86
10.1+ 56 .3 08 .517 .0691 lo67991 o.ooo -.60 2.197 3ob5
TURB 0•10 B5 ·303 o661\ • 0720 2.189'12 o.ooo -.46 3.246 ~.413
>10-50 16 .353 • 6 78 o1695 1o92212 a.ooo -.'i2 2.'557 3o25
)50•100 6 .419 • 363 .1481 .86600 o.ollo -1. 15 !.099 lo87
>100-200 11 .lf31 .709 • 2138 1ob4459 o.ooo -.61 1.960 2o1ft
200+ 10 .o 86 .141 .04'1'1 1o63690 o.ooo -. !.1 o'l 05 2o27
.I I 1 I j
RANGE
3o2H>
2o55f
3.24(,
2. 19 -,
e47G ·
2.632
3.2%
2.282
2o3£.1
1.9&0
2.55.,
.~Jb
1.163
.09~
3.246
2. 19'
lo62 ':!
2.282
2.551
2.632
3.246
1o60':1
3o24f:
.581:1
o095
3.21+6
2ol97
3o246
2.557
1.09~
l • 96 0
.4 0::>
-
APPENDIX B
Gear Efficiency Experiments
-
-
-
.....
-
-
-48 --
-
-
-
INTRODUCTION
Realizing that beach se1n1ng and electrofishing have different capture
efficiencies and that these efficiencies vary with the turbidity level,
amount of cover, and other factors, we conducted two small experiments
in an attempt to be better able to interpret the catch data.
METHODS
The first experiment was designed to determine if backpack electro-
fishing was equally efficient in cells with different amounts of cover.
Previous experience had suggested that capture efficiencies might be low
in cells with little cover because the fish are easily disturbed and
leave the area. Capture efficiencies might also be low in cells with a
large amount of cover because all the fish could not be extracted from
the substrate or dense vegetation.
We approached this problem by calculating the capture probabilities of
fish in cells which ranged from low percent cover cells to high percent
cover cells. Capture probabilities would remain relatively constant
over this range if percent cover had no effect on capture efficiency.
Capture probabilities were calculated by a computer program designed to
estimate population size from multiple removal data (Platts et al.
1983). This program was implemented on a portable battery-powered
microcomputer (Epson HX-20) so that th~ biologists would have on-site
verification that they were using appropriate sampling techniques.
This experiment was conducted at Slough 11 on June 8th and at Slough 8
on August 2nd. Seven cells with a typical range of cover available to
juvenile salmon were sampled at each site with a backpack electrofishing
unit on three successive trials. At the completion of each trial, the
fish were identified and counted and held until the end of the third
trial. Successive trials were separated by about one hour. Turbidity
was low at both sites and did not provide cover.
In the second experiment, five cells at Side Channel lOA were first
sampled with beach seines and then with backpack electrofishing gear.
This was done on two different dates, once when the turbidity level was
high (150 NTU) and once when the turbidity level was low (24 NTU). The
objective was to study the effect of turbidity on the sampling efficien-
cy of the two gear types.
RESULTS
Effects of Cover Density on Electrofishing Efficiency
Only chum and sockeye salmon at Slough 11 were captured in sufficient
numbers to compare capture probabilities among cells with different
percentages of cover. The low numbers of other species captured at this
site and at Slough 8 led to high standard errors on the capture proba-
bility. All species/cells combinations where the standard error was
greater than 2.0 were rejected from this analysis. The capture pro-
bability for chum salmon was high in cells where the percent cover was
low and then steadily declined as the percent cover increased (Appendix
-49 -
Table B-1). The capture probability for sockeye salmon also decreased
as percent cover increased. These results should be regarded as
preliminary because most percent cover categories are represented by
only one cell.
Appendix Table B-1. Capture probabilities for chum and sockeye salmon at
Slough 11 as a function of percent cover.
Species
Chum
Sockeye
Percent cover
0-5
6-25
26-50
51-75
6-25
26-50
Capture Standard
Probability Error
0.9 0.06
0.8 0.12
0.8 0.13
0.7 0.10
0.9 0.03
0.3 0.12
0.9 0.09
0.7 0.14
Comparison of Beach Seining with Backpack Electrofishing
On two occasions when turbidity levels were very different, five cells
at Side Channel lOA were first sampled with beach seines and then with
backpack electrofishing gear {Appendix Table B-2}. A comparison of the
mean catches of chinook salmon fry suggests that beach seining was more
effective in water of high turbidity {150 IHU}, while electrofishing was
more effective in clearer waters (24 NTU}. The Wilcoxon Rank Sum test
failed to reject the null hypothesis that the means are equal; however,
the sample size was only five. Electrofishing at 150 NTU was difficult
even though the cells where the comparisons were made only ranged to 0.4
ft. in mean depth.
Appendix Table B-2. Comparison of beach seining and backpack electro-
fishing juvenile chinook catches at five cells
fished at two different turbidity levels.
Beach
Electrofishing Seining Wilcoxon
Catch/Cell Catch/Cell Rank
Chinook Chinook Sum Test
Turbidity Salmon Salmon Significance
Date (NTU) (Mean ± S.E.} {Mean± S.E.~ Level)
9/07 24 1.6 ± 0.8 0.2 ± 0.2 0.27
7/22 150 1.2 ± 0.6 2.4 ± 0.4 0.19
-50 -
-
-
~
~
-
-
-
-
-
-
-
-'
DISCUSSION
Results from the preliminary experiment on the effect of percent cover
on electrofishing efficiency indicate that capture efficiency decreases
as percent cover increases. This is probably attributable to the
difficulty of seeing fish when cover is abundant and also to the in-
creased likelihood of stunned fish not rising to the surface in dense
cover.
Although the standard errors of the capture probabilities were high,
capture probabilities also appeared to be lower in the 0-5% cover
category for both sockeye at Slough 11 and coho at Slough 8. When cover
is not abundant, the fish are perhaps more likely to flee the cell being
sampled.
The lowest capture probabilities for all three species occurred in the
51-75% cover category (the highest percent cover category sampled in .
this experiment). However, cells with high percent cover were infre-
quently encountered during the 1983 juvenile salmon sampling. Only 13%
of cells sampled at all sites throughout the season had greater than 50%
cover. Therefore, the unequal sampling efficiency over cells with
different amounts of cover was probably not much of a problem, although
it is likely that catch/cell was probably underestimated for cells with
a high percentage of cover. This experiment should be repeated with a
larger number of cells for all species of salmon.
The test conducted of beach seining and electrofishing efficiency at
different levels of turbidity indicated that beach seining was more
effective in water with a high turbidity and electrofishing was more
effective in water with a low turbidity. ·Beach seining is not as
effective in clear water because the fish are often hiding in deadfall,
cobb 1 e, or other cover where the beach seine can not reach them.
Electrofishing is not as effective in water with a high turbidity level
because the samplers can not see the shocked fish.
In conclusion, it may be assumed that estimates of fish density, as
determined by beach seining or electrofishing catches, are often
underestimated. This contrasts with our minnow trap data (for chinook
and coho) of previous years in that minnow traps attract fish to an
area.
-51 -
PART 3
Juvenile Salmon Rearing Suitability Criteria
...
-
....
JUVENILE SALMON REARING SUITABILITY CRITERIA
1984 Report No. 2, Part 3
by Paul M. Suchanek, Robert P. Marshall, StephenS. Hale,
and Dana C. Schmidt
ABSTRACT
Alaska Department of Fish and Game
Susitna Hydro Aquatic Studies
2207 Spenard Road
Anchorage, Alaska 99503
Changes in flow regimes in the Susitna River may affect the habitat used
by rearing juvenile salmon. In order to model changes in habitat
usability, data were collected for development of suitability criteria
for the habitat attributes of cover, velocity, and depth used by juve-
nile chinook, coho, sockeye, and chum salmon. Representative sites
between the Chulitna River confluence and Devil Canyon were sampled for
juvenile salmon and habitat attributes were measured. Analysis was
primarily univariate and data were pooled over site and season. Turbid-
ity was apparently used by chinook salmon as cover prompting development
of suitability criteria for clear (<30 NTU) and turbid (>30 NTU) con-
ditions. Catches were insufficient for analysis of the other species by
turbidity level. Suitability criteria for percent cover, cover type,
velocity, and depth were developed for all four species of salmon.
Composite weighting factors were formulated and correlated or compared
with observed fish catch. Limitations of the suitability criteria and
possible uses in habitat analysis are discussed.
-
-
-
-·
TABLE OF CONTENTS
ABSTRACT •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••
LIST OF FIGURES ••..
LIST OF TABLES •.•...••.....•..•..
LIST OF APPENDIX TABLES •....•
1.0
2.0
3.0
INTRODUCTION •..
METHODS ....................................................... .
2.1
2.2
2.3
Study Locations ...•.•..
Field Data Collection •. ..................................
2.2.1
2.2.2
Biological ..••..
Phys i ca 1 ••.
Data Analysis .....
2.3.1
2.3.2
2.3.3
Cover analysis •..•.•.••.••...•.•.•.•••..••••.•...•
Velocity and depth analysis .•.••••...•.•••••...•.•
Tests of data fit ..••.•.••..••...••.•••..•••..••••
RESULTS •••••••••••••••••••••••••••••••••••••••••••••••••••••••
3.1
3.2
3.3
3.4
3.5
Sampling Effort and Catch .•••....•.•.
Analysis of Chinook and Coho Salmon
Distribution in Clear Water Using Electrofishing Data .•
3.2.1
3.2.2
Cover .•••••.
Depth and velocity ............................... .
Analysis of Chinook Salmon Distribution in Turbid Water
Using Beach Seine Data ...••....•...•........•..........
3.3.1
3.3.2
Cover ............................................ .
Depth and velocity ••.••••.••.•••••••••.••
Analysis of Sockeye and Chum Salmon Distribution
Using Pooled Electrofishing and Beach Seine Data •.•.
3.4.1
3.4.2
Cover ............................................ .
Depth and velocity ..........•....•..........
Tests of Fitted Habitat Criteria to Observed Fish
Catches •••••.••....•.•••••..••..••••.•••.
3.5.1
3.5.2
Chinook and coho salmon .•.•.••...••.•••.••.
Sockeye and chum salmon ..•.•..••.....•••••.
i i
iv
vi
vii
1
2
2
2
2
4
4
7
7
8
10
10
14
14
16
19
19
20
20
20
22
22
22
26
TABLE OF CONTENTS (Continued)
4.0 DISCUSSIQN ..•..•.•....• ., .. e••••"•••••e•••e•e-e••••••••••••••••e 28
4.1 Limitations of the Suitability Criteria.................. 28
4.2 Chinook and Coho Salmon.................................. 29
4.2.1 Chinook salmon.................................... 29
4.2.2 Coho salmon....................................... 30
4.3 Sockeye and Chum Salmon.................................. 31
4.3.1 Sockeye salmon.................................... 31
4.3.2 Chum salmon....................................... 32
4.4 Recommended Applications for the Suitability Criteria.... 32
5.0 CONTRIBUTORS.................................................. 35
6.0 ACKNOWLEDGEMENTS.............................................. 36
7.0 LITERATURE CITED ...••..••••.••.. ~~····························· 37
8.0 APPENDICES
Appendix A
Appendix B
Appendix C
Calculations of suitability of cover type for
chinook and coho salmon in clear water •..•..•••
Calculations of effect of cover type on
distributions of sockeye and chum salmon .....•.
Suitability indices for juvenile salmon for
cover, velocity, and depth •..•••....•.....•.•••
iii
40
44
47
....
-
.....
l
~
~
-
-
-
-
-
LIST OF FIGURES
Figure
1
2
3
4
5
6
7
Title Page
Location of the study sites sampled more than
three times for juvenile salmon suitability
criteria development~ May through October 1983......... 3
Outline of data analysis pathways for determina-
tion of juvenile salmon suitability criteria........... 5
Mean catch of juvenile chinook and coho salmon per
cell by percent cover category (bars) and fitted
suitability index (lines) in low turbidity waters~
Chulitna River confluence to Devil Canyon reach of
the Susitna River...................................... 15
Comparison of cover type suitability indices for
juvenile chinook and coho salmon in low turbidity
waters, Chulitna River confluence to Devil Canyon
reach of the Susitna River............................. 15
Mean catch of juvenile chinook salmon per cell by
velocity and depth intervals (bars) in low
turbidity waters~ Chulitna River confluence to
Devil Canyon reach of the Susitna River ..•..•..........
Mean catch of juvenile coho salmon per cell by
velocity and depth intervals (bars) in low
turbidity waters, Chulitna River confliJence to
17
Devil Canyon reach of the Susitna River................ 18
Mean catch of juvenile chinook salmon per cell by
percent cover categories (bars) and fitted
suitability index (line) in high turbidity waters~
Chulitna River confluence to Devil Canyon reach of
the Susitna River...................................... 19
8 Mean catch of juvenile chinook salmon per cell' by
velocity and depth intervals {bars) in high
turbidity waters, Chulitna River confluence to
Devil Canyon reach of the Susitna River................ 21
9 Proportion of cells with juvenile sockeye and chum
salmon present by percent cover category (bars)
and fitted suitability indices (lines) Chulitna
River confluence to Devil Canyon reach of the
Susitna River.......................................... 23
10 Comparison of cover type suitability indices for
juvenile sockeye and chum salmon, Chulitna River
confluence to De vi 1 Canyon reach of the Sus i tna
River....................................................... 23
iv
LIST OF FIGURES (Continued)
Figure
11
Title
Proportion of cells with juvenile sockeye salmon
present by velocity and depth intervals
(bars), Chulitna River confluence to Devil Canyon
reach of the Susitna River............................. 24
12 Proportion of cells with juvenile chum salmon
present by velocity and depth intervals (bars),
Chulitna River confluence to Devil Canyon reach of
the Susitna River...................................... 25
v
-
-
-
-
-
-
,...
-
-
LIST OF TABLES
Table Title
1 Percent cover and cover type categories ....•...•...•..... 4
2 Velocity and depth groupings for suitability criteria
deve 1 opment .•.•.•..•••••.•••.•.••• .:.......... . . • • • • . • • • • • 8
3 Sampling effort (number of cells fished) and catch
by gear type. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
4 Comparison of mean catch/cell for cells without
-object cover above and below 30 NTU turbidity............ 11
5 Sampling effort and catch by gear type and turbidity
level .................................................... 12
6 Data sets used for suitability index development......... 12
7 Kendall correlation coefficients between habitat
variables and chinook and coho catch by cell (N=813)
in_ clear water for electrofishing data................... 13
8 Kendall correlation coefficients between habitat
variables and chinook catch in turbid water by cell
9
(N=320) for beach seine data............................. 13
Analysis of variance between cover type, percent
cover, and chinook or coho catch transformed by ln
( x+ 1 ) ............ ·. . . . . . . . . . . . . . . . . . . . . . . . . . • . . . . .. . . . . . . . . 14
10 Analysis of variance between depth, velocity, and
11
12
13
14
chinook or coho catch transformed by ln (x+l)............ 16
Analysis of variance between depth, velocity, and
chinook catch transformed by ln (x+l) in high
turbidity water.......................................... 20
Chi-square tests for differences in proportions of
sockeye or chum presence between habitat attribute
groupings of percent cover, cover type, velocity and
depth.................................................... 22
Correlations between composite weighting factors
generated using various combinations of suitability
indices and transformed (ln (x+1)) chinook and coho
catch..................................................... 26
Proportional presence of sockeye and chum salmon fry
associated with several composite weighting factors...... 27
vi
LIST OF APPENDIX TABLES -Appendix Table Title
C-1 Suitability indices for juvenile salmon for cover,
velocity, and depth ........... o., •••• ., ••••••••••••••••• ~ •• 48 -
-
-
-
-
-
vii -
-
-
-
1.0 INTRODUCTION
Studies to date (ADF&G 1983a) of the rearing salmon species which occur
between the Chulitna River confluence and Devil Canyon, indicate that
successful rearing is dependent on a variety of physical parameters.
The instream flow incremental methodology has been developed for use in
evaluating fish habitat (Bovee 1982) and can be used in the Susitna
River basin to evaluate effects of mainstem discharge on sites used by
rearing juvenile salmon. In order to implement this methodology,
habitat suitability criteria need to be developed which express the
optimum, marginal, and unusable ranges of habitat variables on a one
(optimum) to zero (unusable) basis. These criteria are then coupled
with hydraulic models by using a system of computer programs called the
Physical Habitat Simulation (PHABSIM) system (Bovee 1982). Output from
PHABSIM includes calculations of the amount of equivalent optimum
habitat called weighted usable area.
The present work develops suitability criteria for four species of
juveni 1 e salmon in the Chulitna River to De vi 1 Canyon reach of the
Susitna River for application in incrementa 1 simulations of rearing
habitat as a function of mainstem flows. Criteria developed for these
species are univariate suitability functions for cover type and percent
cover, depth, and velocity. Functions for each of these environmental
attributes were developed for juvenile chinook, coho, sockeye and chum
salmon rearing. Different criteria for low and high turbidity water were
developed as data permitted. Pink salmon were not considered because
they do not rear in the study reach.
Suitability criteria have been formulated in a variety of ways (Bovee
1982) although most methods have been oriented towards describing the
requirements for readily observable individuals in a relatively uniform
or predictable macrohabitat. Since rearing juvenile salmon are neither
easily observed nor sampled in the Susitna River 1 s diverse glacial
environment and related salmon rearing habitats, alternate criteria
development techniques were used in this study. The criteria developed
are specific to the Susitna ·River reach between the Chulitna River
confluence and Devil Canyon.
The criteria developed in this report have been used with hydraulic
models for seven sites on the Susitna River to provide weighted usable
area projections at a wide variety of discharges (see Part 7 of this
volume). They also have been used to study changes in the usability of
habitat at six habitat model sites as natural mainstem discharge changes
(see Part 4 of this volume). These results will be used in combination
with other information to develop estimates of total usable rearing area
for the Chulitna confluence to Devil Canyon reach of river at incre-
mental levels of mainstem discharges.
- 1 -
2.0 METHODS
2.1 Study Locations
Locations selected as fish preference sites had substantial numbers of
rearing juvenile salmon in 1981 and 1982 or were thought to be typical
sites having the potential for juvenile rearing. The sites are located
on the Susitna River reach between Whiskers Creek (RM 101.2) and Portage
Creek (RM 148.8). Seven tributary sites, two upland sloughs, and 12
other sites which naturally oscillate between being side sloughs or side
channels were sampled at least four times (Figure 1). There were also
nine sites sampled only once and five sites sampled two or three times
(see Part 2 of this report for a listing). These sites were thought to
represent a wide cross section of habitat conditions experienced by
rearing juvenile salmon in this reach of the Susitna River since
tributaries, upland sloughs, side sloughs, and side channels were all
intensively sampled. A 1 imited amount of sampling ·was done in the
mainstem channel and large side channels because of the difficulty in
sampling these areas and because we believed high velocities limit
juvenile rearing habitat.
2.2 Field Data Collection
2.2.1 Biological
Detailed descriptions of the site layout and data collection techniques
are available in other reports (ADF&G 1984, and Part 2 of this report).
Eight to 10 day field samplings were made twice monthly between May and
October 1983. Twenty-three sites were sampled from three to seven times
while the other 12 sites were only incidentally sampled once or twice.
About eight staked transects from 75 to 200 feet apart were established
across the study site. Upstream f2om each transect, sampling cells 50
feet long by six feet wide (300 ft ) were delineated along each shore-
line. Another mid-channel cell was located between the shoreline cells.
The grid of transects and cells was normally located in areas of rela-
tively uniform water temperature, pH, dissolved oxygen, conductivity,
and turbidity. Transects were placed to maximize within site variabil-
ity of habitat types sampled while also attempting to maintain uniform
physical habitat within individual sampling cells. Cells were selected
to represent a wide range of habitat types and approximately 20 cells
were sampled per day.
During the field season, we directed sampling effort towards sites where
rearing fish were numerous based on knowledge of seasonal movements.
Sampling frequency was reduced if efforts to catch 30 or more juveniles
of a species in a grid of transects were unsuccessful. Backpack elec-
trofishing units and 1/8" mesh beach seines were used to sample the
entire cell for fish. Typically, beach seining was limited to turbid
water samplings and electrofishing to clear water conditions. Electro-
fishing was the preferred sampling method, but was found to be ineffec-
tive in turbid water. Each captured fish was identified to species and
measured in total length to the nearest millimeter. Those cells sampled
for fisheries data were subsequently individually characterized by a set
of habitat measurements even if no fish were captured.
- 2 -
-
-
-
....
-
-
-
-
Portaoe Creek (Mouth ond TRM 4.2)
Slouoh 22
Slouoh 21
Sidechannel 21
Indian River (Mouth and TRM 10.1)
Slouoh II
S idechannel 10
Sidechannel lOA
s louob 9
Slouoh SA
---------16. Chase Creek
Whiskers Cr. Sl.
Figure 1. Location of the study sites sampled more than three times for
juvenile salmon suitability criteria development, May through
October, 1983.
-3-
2.2.2 Physical
We determined an average depth and ve 1 oci ty, and a 1 so est i rna ted the
total amount of available cover (expressed in percent areal coverage),
and the dominant type of cover available for juvenile salmon in each
cell. Codes for nine cover types and six categories of percent cover
were developed (Table 1). Prior to the sampling season, a field trip
was made to promote consistent ratings among the raters. Estimates of
cover were made on the basis of cover specifically available to juvenile
salmon for concealment or protection. Cells without objective cover
(cover type group #1) wi 11 be referred to as "no cover" or "zero cover 11
ce 11 s.
Table 1. Percent cover and cover type categories.
Group If :r; Cover Group If Cover Type
1 0-5% 1 No object cover
2 6-25% 2 Emergent vegetation
3 26-5m; 3 Aquatic vegetation
4 51-75% 4 Debris or deadfall
5 76-96~ 5 Overhanging riparian
6 96-100% 6
vegetation
Undercut banks
7 Gravel (1" to 3" diameter)
8 Rubble (3" to 5" diameter)
9 Cobble (larger than 5"
diameter)
Water temperature, dissolved oxygen, pH, conductivity, and turbidity
were measured at one point in the grid. If an obvious water quality
gradient existed across the grid, another measurement of these
parameters was taken. Detai 1 ed descriptions of the water chemistry
measurement procedures are available in ADF&G (1984).
2.3 Data Analysis
Data were separated by gear type because both beach se1n1ng and electro-
fishing effectiveness are influenced by water quality and hydraulic
attributes and because each gear was used selectively, dependent upon
the sampling conditions. Since no resources were available for a major
study of gear effectiveness, we did not attempt to quantify gear effi-
ciency under various sampling conditions. Beach seines were used
because backpack electrofishing is ineffective in highly turbid water.
The bias inherent in both gear types influenced our pathway of analysis
and affected our interpretation of results and subsequent conclusions.
Figure 2 details the data analysis pathways and final products of
criteria development as presented in the results section.
- 4 -
-
~·
-
-
-
I
01
I
Figure 2.
1
COVER TYPE weighted mean& within
~ percent cover categorf es
SECTION 3.2 ~PERCENT COVER -re_g_r_e
5
-
6
-:-1-
0
n---------
Low Turbidity -....__ Analysi& of Variance
/ El ectrofi shi ng ---.._ _
/ Data .'-.... VELOCITY
CHINOOK . ~ ~ hand·flt curves
AND COHO ~DEPTH hand ... fit curves
SALMON
(Suitability
= mean catch
per cell)~
SOCKEYE ~
AtiD CHUM
SALMON
(Sui tabi 1 i ty
. --------PERCENT COVER regress ton
SECTION 3.3 --------.
High Turbidity Analy•ls of Variance--VELOCITY ----,------,.-----
Beach Seining Data~ hand-fit curves
(Chinook Only)
SECT I ON 3.4
Low Turbl dl ty
Electrofi shing
Data
Pooled
DEPTH
COVER TYPE
hand· I It curves
divide by proportional presence
for zero cover eel h
PERCENT COVER regression
=proportional
presence) ~High Turbidity
VELOCITY
Beach Seining
Data DEPTH hand ... fit curves
l
Figure 4.
Figure 3.
Figure& 5. and 6.
Figures 5. and 6.
Fl gure 7.
Figure 8.
Figure 8.
Figure 10.
Figure 9.
Figure• 11. and IZ.
Fi gure• II. and IZ.
Outline of data analysis pathways for determination of juvenile salmon suitability criteria.
We used different types of ana lyses for chi nook and coho sa 1 mon in
comparison to sockeye and chum salmon. Chinook and coho salmon are
territorial or at least exhibit some forms of agonistic behavior (Stein
et al. 1972) and normally disperse themselves as individuals while
sockeye and chum salmon are usually distributed in schools which move
about as a cohesive social unit.
Suitability was derived for chinook and coho salmon by taking total fish
catch for each value of attribute (utilization) and dividing by the
number of cells fished having the same attribute value (effort). For
example, if 50 chinook salmon fry were captured in 25 cells of 0.0
velocity sampled, mean catch per cell {suitability) was 50/25 = 2.0 for
0.0 velocity cells. Fish density was assumed to be a function of mean
catch per cell. Differences in mean catch per cell by habitat attribute
value were analyzed with analysis of variance and least squares re-
gression.
Sockeye and chum salmon suitability was derived by taking the total
number of cells with fish present by value of habitat attribute (uti-
lization) and dividing by the number of cells fished (effort). For
example, if chum salmon fry were captured in 10 of 50 cells of 0.0
velocity fished, then proportional presence (suitability) was 10/50 =
0.2 for 0.0 velocity cells. Suitability was derived differently for
sockeye and chum salmon because these fish school normally and capture
of a large school within a cell might disproportionately affect mean.
catch per cell as the habitat might be only as good as another cell
nearby without any fish but the cell with fish would be ranked much
higher than if rated on a proportional presence basis. Differences in
proportional presence by habitat attribute value were analyzed with
chi-square tests of association.
Data from all sites over the entire season were pooled by species for
analysis. Data from tributary sites where no major runs of sockeye
salmon are present were excluded from the sockeye suitability criteria
development, as were data collected between May 1 and 15, when only a
small percentage of sockeye had emerged. Since the vast majority of
chum salmon outmigrate from the upper Susitna River prior to July 15
(ADF&G 1983b), only data collected before July 15 were used to develop
suitability relationships for this species.
Statistical analyses used included analysis of variance, linear re-
gression and chi-square tests of association. Most statistical analyses
were conducted using the Statistical Package for the Social Sciences
(SPSS) {Nie et al. 1975). Transformations by natural log (X+1) were
used to help equalize variances and normalize catch per cell of chinook
and coho salmon for analysis of variance (Dixon and Massey 1969).
Chi-square tests of association were used to examine proportional
presence data for differences in use of categories of habitat attributes
by sockeye and chum salmon. Expected values in these tests were cal-
culated with standard contingency table techniques. Kendall rank-order
correlations were carried out between the habitat variables to check for
intercorrel ations. The particular procedure utilized in each analysis
is presented within the appropriate results section.
- 6 -
-
-
-
-
t4ost of the analysis was geared toward a univariate analysis and devel-
opment of suitability criteria but some multivariate comparisons were
made. t4ultiway analyses of variance were conducted to find if inter-
action effects were significant. All velocity and depth criteria were
fit to the data by hand using professional judgement to give the best
fit. The rationale and judgements used for criteria development are
discussed according to the individual relationship.
2.3.1 Cover analysis
Cover is an important factor influencing the distribution of juvenile
salmon (Reiser and Bjornn 1979). Rocks, debris, and vegetation are
types of object cover; turbidity is another form of cover. vie examined
the effects of both the type and amount of object cover on the dis-
tribution of juvenile salmon. Turbidity effects were inferred from
differences in catch in· cells without object cover over the range of
turbidities sampled. We pooled percent object cover categories 76-95%
and 96-100% for the analysis because of small sample sizes and then
regressed percent cover categories against catch per cell for chinook
and coho salmon. The proportion of cells with fish present were re-
gressed against the percent cover categories for sockeye and chum
salmon.
The relative importance of object cover type for chinook and coho salmon
in clear water was addressed by examining mean catch per cell by cover
type within each percent cover category. Each mean catch/cell for a
cover type within a percent cover category was divided by the mean catch
for that percent cover category for all cover types combined. These
ratios were then pooled over all percent cover categories for a cover
type by taking a weighted mean adjusted by the number of cells of that
cover type within each percent cover category to give an average effect
of cover type. The weighted mean was then used to rank cover types by
suitability on a scale from 0 to 1. The equations used and an example
are given in Appendix A. Cover type suitability differences were not
addressed with the beach seine data since we believed seine effective-
ness was strongly affected by cover type.
Because of the smaller sample sizes and use of proportional presence
data, cover type suitability differences were calculated in a different
way for chum and sockeye salmon. Sockeye and chum cover type
suitability differences were addressed by pooling the incidence of catch
by cover type over all percent cover categories and then dividing
through by the proportional presence for cells without object cover.
Sometimes, the proportional presence for some cover types was less than
the proportional presence for zero cover cells. In these instances,
cover type was assumed to have no effect on distribution and was ranked
with the zero cover type in the suitability ratings. The equation used
and an example are given in Appendix B.
2.3.2 Velocity and depth analysis
Velocity and depth were measured in intervals of 0.1 ft/sec and 0.1 ft,
respectively. Since sample sizes were small and variances were high,
- 7 -
these values were pooled into groups (Table 2). Baldridge and Amos
(1983) listed a number of criteria of use in grouping data for criteria
development but since we analyzed four species of salmon, one standard
grouping interval was used for all criteria development.
Table 2. Velocity and depth groupings for suitability criterid develop-
ment.
Velocity Depth
(ft/sec) (ft)
Group # Grouping Group # Grouping
0 0.1 -0.5
2 0.1 -0.3 0.6 -1.0
3 0.4 -0.6 3 1.1 -1.5
4 0.7 -0.9 4 1.6 -2.0
5 1.0 -1.2 5 2.1 +
6 1.3 -1.5
1.6+
Mean catch/cell was again used as the measure of suitability for chinook
and coho criteria development. Sockeye and chum suitability was
measured using proportional presence.
2.3.3 Tests of data fit
In the PHABSIM system, univariate suitability indices are combined to
provide a composite weighting factor which reflects the habitat poten-
tial of a cell at a given discharge (Bovee 1982). Suitability criteria
are normally combined by multiplying suitability indices together to
formulate these weighting factors but other combinations are possible
(Milhous et al. 1981). Regardless of· the composite weighting factor
formulation used, one of the assumptions of the instream flow incre-
mental methodology is that there is a positive linear relationship
between weighted usable area and habitat use (Orth and Maughan 1982).
We attempted to eva 1 uate various combinations of univariate suitability
indices by comparison with observed fish catches.
For chinook and coho salmon, we compared observed catches by cell with
composite weighting factors calculated using suitabi1 ity indices from
various combinations of habitat attributes. Pearson correlation coeffi-
cients were calculated between various composite weighting factor
indices and coho and chinook catch per cell. We again transformed catch
per cell with natural log (X+l) to normalize the data. Since propor-
tional presence was used as a measure of suitability for chum and
- 8 -
-
-
.....
-
-
-
sockeye salmon, correlation coefficients could not be used to test for
data fit. Instead, we ca 1 cul a ted severa 1 composite weighting factors
using only a few combinations of univariate suitability indices and then
divided the data into four groups of approximately equal size by value
of composite weighting factor. Chi-square tests were then run to see if
proportional presence was associated with the composite weighting factor
value intervals .
- 9 -
3.0 RESULTS
3.1 Sampling Effort and Catch
Fish suitability criteria data were collected at a total of 1,260 cells
over the entire season, with about 70 percent of the sampling done with
backpack electrofishing gear and 30 percent with beach seines (Table 3).
Some of the cells fished were subsequently eliminated from the sockeye
and chum suitability criteria development because of seasonal and site
factors discussed in the methods section.
Table 3. Sampling effort (number of cells fished} and catch by gear
type.
Electrofishing Beach Seinin9 Total
Effort Catch Effort Catch
(cells all age (cells a 11 age
fished) classes fished classes Effort Catch
Chinook 871 3066 389 1329 1260 4395
Coho 871 1907 389 113 1260 2020
Sockeye 658 814 355 192 1013 1006
Chum 408 1152 106 5 514 1157
Field observations and examination of the catch data indicated that
chinook salmon distribution was very different in turbid water than in
clear water. Scatter plots of juvenile salmon catch by species in cells
without object cover versus turbidity were examined. An inflection
point at approximately 30 NTU was noted for juvenile chinook salmon.
The catch rate at turbidities greater than 30 NTU was much higher than
the catch rate below 30 NTU, indicating that turbidity is used for cover
in lieu of object cover. Sample sizes for the other species were too
small to indicate whether other inflection points were evident. Subse-
quently, mean catch/cell was examined for cells without object cover for
each of the four species both above and below 30 NTU (Table 4). Catches
of chinook were significantly higher in high turbidity cells without
object cover than in similar cells with turbidities of less than 30 NTU.
Chum salmon were caught in significantly higher numbers in clear water.
-10 -
-
-
-
-
-
-
-
-I
-
~
-
-
Table 4. Comparison of mean catch per cell for cells without object
cover above and below 30 NTU turbidity.
Total
catch Total
in zero
zero cover ~1ean Mean
cover cells catch catch
cells fished :30 NTU >30 NTU ! Significance
Chinook 312 155 0,19(N=42) 2.69(N=ll3) 14.99 <0.001
Coho 5 155 O.OO(N=42) 0.04(N=ll3) 1. 35 0.25
Sockeye 64 144 0.23(N=35) 0.51(N=109) 0.76 0.39
Chum 52 57 1.8l(N=21) 0,39(N=36) 5.15 0.03
Since the distribution of chinook is different in waters with turbidit-
ies greater than 30 NTU, when compared to. clearer water, we grouped the
data by both turbidity level and gear type (Table 5). The only data set
deemed sufficient in size for suitability criteria development in high
turbidity conditions was the chinook beach seine data. Although chum
salmon may have a different distribution in turbid water, sample sizes
were insufficient for suitability criteria development. Coho catches
were very small in turbid water and no turbidity dependent suitability
criteria could be generated from the data. The electrofishing data in
clear water cells was ample for criteria development, and therefore the
small amount of beach seine data were not pooled with the electrofishing
data. Similarly, chinook electrofishing data from clear water were used
exclusively for low turbidity criteria development.
Small sample sizes made it necessary for gear types and turbidity levels
to be pooled for development of chum and sockeye suitability criteria
development for two reasons. The amount of electrofishing data for
sockeye and chum salmon was smaller than for chinook and coho salmon
because some cells fished were eliminated due to season or spawning
distribution as previously discussed in the methods. Also since propor-
tional presence was used as the measure of suitability, sample sizes
need to be large for good estimates of proportions. We therefore
assumed that seining and electrofishing were equally effective at
catching at least one fish in a cell if fish were present. Table 6
summarizes the data sets used for criteria development.
-11 -
Table 5. San1pling effort and catch by gear type and turbidity level.
Clear (Turbidity ~ 30 NTU)
Electrofishin~ Beach Seine
Effort atch Effort Catch
Chinook 813 2574 41 39
Coho 813 1699 41 62
Sockeye 611 757 24 84
Chum 366 1107 16
Turbid (Turbidity > 30 NTU)
El ectrofi shin~ Beach Seine
~ atch ~ Catch
Chinook 44 61 320 1241
Coho 44 206 320 23
Sockeye 44 57 303 101
Chum 29 44 90
Note -Cells where turbidity was not recorded (14 electrofished cells
and 28 beach seined cells) were excluded from this data set.
Table 6. Data sets used for suitability index development.
Number
Turbidity Suitability of cells
Species Level* Gear T,l~ Measure Fished
Chinook Clear Electrofishing Catch/cell 813
Turbid Beach Seine Catch/cell 320
Coho Clear Electrofishing Catch/cell 813
Sockeye Both Pooled Proportion of 1013
cells with catch
Chum Both Pooled Proportion of 514
cells with catch
* Clear -Turbidity = 30 NTU
Turbid -Turbidity > 30 NTU
-12 -
-
-
"""'!
-
-
-
-
...,
'""'l
.....
-
Correlations among the values of habitat attributes and catch were
examined for the data sets used in criteria development. The resulting
Kendall rank-order correlation coefficients are listed in Table 7 for
the low turbidity electrofishing data. There are a number of
statistically significant correlations among the habitat attributes but
none are greater in absolute value than 0.18. Correlations between the
habitat attributes and fish catch are also small, none being over 0.22
in absolute value. Large correlations among the habitat variables would
necessitate a multivariate approach or elimination of selected habitat
attributes from consideration.
Table 7. Kendall correlation coefficients between habitat variables and
chinook and coho catch by cell (N=813) in clear water for
electrofishing data.
Percent Cover
~ ~ Ve loci tl: Depth Chinook
Percent cover 1.00
Cover Type 0.11'** 1.00
Velocity 0.13*'* 0.18** 1.00
Depth 0.03 -0.11** -0.17** 1.00
Chinook 0.21** 0.18** 0.20** -0.04 1.00
Coho 0.22** -0.18** 0.02 0.21** 0.20**
*Significantly different from 0 at P< 0.05
**Significantly different from 0 at p< 0.01
Kendall rank-order correlations among the high turbidity beach seine
data were very similar to the electrofishing data {Table 8). The
correlation between percent cover and cover type was fairly high {0.40)
but small sample sizes and beach seine inefficiency in high object cover
conditions caused the analysis of cover type in turbid water to be only
qualitative.
Table 8. Kendall correlation coefficients between habitat variables and
chinook catch in turbid water by cell (N=320) for beach seine
data.
Percent Cover
cover ~
Percent cover 1.00
Cover Type 0.40** 1.00
Velocity 0. 12** 0.20**
Depth 0.01 -0.05
Ch1nook 0.12** -0.02
*Sfgn1f1cantly different from 0 at p < 0.05
**Significantly different from 0 at p < 0.01
-13 -
Velocity Depth
1.00
0.08* 1.00
-0.19** 0.12**
3.2 Analysis of Chinook and Coho Distribution in Low Turbidity
Waters
3.2.1 Cover
Two-way analyses of variance (using the regression approach) were run on
the catch/ ce 11 data to examine the effects of cover type and percent
cover on the transformed chinook and coho catch/cell (Table 9). The
effects of both cover type and percent cover were significant but the
amount of explained variation was small.
Table 9. Analysis of variance in clear water between cover type,
percent cover, and chinook or coho catch transformed by ln
(x+l). Due to empty cells or a singular matrix, interactions
could not be calculated.
Chinook Sum of Mean Significance
Source of Variation Squares df Square f. of F
Main Effects 113.852 12 9.488 10.805 ~ 0.001
Cover type 45.871 8 5.734 6.530 < 0.001
Percent cover 54.897 4 13.724 15.630 "0.001
Explained 113.852 12 9.488 10.805 "'-0.001
Residual 702.482 800 0.878
Total 816.334 812 1.005
Coho Sum of Mean Significance
Source of Variation Squares df Square £. of F
Main Effects 90.738 12 7.561 11.402 < 0.001
Cover type 56.793 8 7.099 10.705 .(; 0.001
Percent cover 35.058 4 8.765 13.216 . < 0.001
Explained 90.738 .12 7.561 11.402 ..:. 0.001
Residual 530.550 800 0.663
Total 621.288 812 0.765
Least squares regressions were then run between chinook and coho catch
per cell and the percent cover categories to quantify the relationship
to cover categories where there is only a small amount of data. The fit
of the regression to the actual mean catches and derived suitability
indices by cover category is shown in Figure 3. The effects of cover
type by species were then quantified by taking a weighted mean of the
effect of cover type over all percent cover categories to derive a suit-
ability index for cover type (Figure 4).
-14 -
.....,
-
~
.""'1
~
""'l
-
-
-
1
I ......
Ul
I
Figure 3.
100
...J
....1 9.0 w
u 8.0
a:
lr 7.0
~ 6.0
0
~ 5.0
u
:1: 4.0
~ 3.0
u
z 2.0
<I w 1.0
:::E
...J
..J
1.0
"' 6.0 0
"' "' 5.0 Q.
0 6 4.0
0
~ 3.0 ... ..
0 2.0
:0:
<(
"' :::;; 1.0
l
[! ·Standatd Error
1.00
0.90 X
0.80 ~
0.10 z
0.60 >-
1-
0.50 ....1
0.40 Cll
0.30 <I
1-
0.20 ~
0.10
L..J_..L....L....-'-...L.-L..J __ .__.__--I.._..L 0
f!l•l50 N•29l' N•:l:S.2 t.~-=i 5 H • ~9
(o-5'lbl (6-25,) (28-50%1 ,,, .. 7~%1176-100%!
PERCENT COVER CATEGORIES
[! -Standard Error
1.00
0.90 >(
0.80 ~
070 3
0.60
050
0.40
0.30
020
0.10
'-"o._;;;-'---'--L..J'--...L...L......JL.l_...J...._.L 0.00
N=l50 frrll=l!H N=Z32 Jrrll:9~ loll • )9
(Q-5,.-,J (6-~,'1.) (26~~ (51-7~'"/.J (7'6-100'1.)
PERCENT COVER CATEGORIES
Mean catch of juvenile chinook and coho salmon per cell by percent
cover category (bars) and fitted suitability index (lines) in low
turbidity waters, Chulitna River to Devil Canyon reach of the
Sus itna River.
>-
1-
...J
Ql
<{
I-
:>
V>
N=GB N•32
:>
U"" ~c:
"'"' ..,on
c:
::;)
u g
~:; "-"'"' <I"' .. >
COVER TYPE
fJ CHINOOK
C2J COHO
.. o;
0 ..
~ a .. ~
...J<!l
... c: c: ·o. .~ " ... a o -e -~
: 0:
0
c
0 c ·-.,-oo ~-., ..
e ~ LoJ>
.,
"" 0 u
0 z
L------------------------·-·-----'
Figure 4. Comparison of cover type suitability indices for juvenile chinook
and coho salmon in low turbidity waters, Chulitna River to Devil
Canyon reach of the Susitna River.
3.2.2 Depth and velocity
Since depth and velocity were not expected to be linearly related to
fish habitat suitability, depth and velocity effects were analyzed in a
two-way analysis of variance for chinook and coho catch per cell
(Table 10). Depth and velocity were singly not significant for chinook
at the 0.05 significance level after adjusting for the effects of the
other, but taken together, they were significant for chinook as was the
interaction between depth and velocity. Depth, velocity, and the
interaction between these two attributes were all significant for coho.
The total amount of explained variation was again relatively small for
both species.
Table 10. Analysis of variance in clear water between depth, velocity,
and chinook or coho catch transformed by ln (x+1).
Chinook Sum of Mean Significance
Source of Variatlon Squares df Square £. of F
Main Effects 27.¢26 10 2.743 2.990 L. 0.001 Depth 8.099 4 2.025 2.207 0.067 Velocity 7.549 6 1.258 1.372 0.223
Interaction Effects 25.216 16 1.576 1. 718 0.039
Exp Ja i ned 95.271 26 3.664 3.994 < 0.001
Residua 1 721.062 786 0.917
Total 816.334 812 1.005
Coho Sum of Mean Significance Source of Variation Squares df Square £. of F
Main Effects 35.505 10 3.551-5.242 < 0.001 Depth 8.318 4 2.079 3.070 0.016
Velocity 19.343 6 3.22¢ 4.760 < 0.001
Interaction Effects 40.079 16 2.505 3.699 < 0.001
Explained 88.957 26 3.¢21 5.052 < 0,001
Residual 532.331 786 0.677
Total 621.288 812 0.765
Since the data base was not large enough, given the amount of varia-
bility in the data, to fit a multivariate function with any confidence,
we examined depth and velocity only on a univariate basis. Professional
judgement was used to fit a curve to the data by hand and suitability
indices were normalized to the fitted data (Figures 5 and 6). The
functions were fit so that they followed the means most closely over the
intervals where sample sizes were greatest. On the depth curves, we
-16 -
-
-
-
-
-
"""!
-
,.,.,.
8.0
7.0
..J
..J -ILl
0 6.0
' J:
~ 5.0 <
0
~ 4.0 0
0 z
J: 3.0
0
z
< 2.0 LLJ
::IE
1.0
,..-
0
F"'"
6.0
...J
...J
LLJ .-(.) 5.0 a:
ILl
a.
:I: 4.0
(.)
1-< 3.0 (.)
::.:::
I""' 0
0 2.0 z -:I:
(.) 1.0 -z
<
LLJ
::::IE 0
0
Figure 5.
I± Standard Error 1.00
0.90
0.80
)(
0.70 LLJ c z
0.60
~
0.50 1-
:::!
0.40 CD < I-
0.30 5 I
(J)
0.20
W· 0.10
'W' 0.00
0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 3.0
VELOCITY (ft. /sec.)
I± -Standard Error
1.00 X
ILl
0.80 c
z
0.60 ~
1-
0.40 ...J
CD w-<
0.20 1-
N•2.97 N•47 'W' :::>
N•17 (/)
'W' 0.00
0.2 0.4 0.6 0.8 LO 1.2 1.4 1.6 1.8 2D 2.2 2.4 3.5
DEPTH (ft.)
Mean catch of juvenile chinook salmon per cell by velocity and
depth intervals (bars) in low turbidity waters, Chulitna River to
Devil Canyon reach of the Susitna River. Suitability indices
(lines) fitted by hand.
-17-
...J
...J w u
a:: w a..
0
:I:
0 u
:I: u
1-
<t: u
z
<t:
ILl
:::E
:I: u
1-<t:· u
z
<t: w
:::E
6.0
Figure 6.
I! -Standard Error
1.00 (/)
c
-i
0.80 )>.
(D
0.60 r
-i
0.40 -<
z
0.20 0
CTI
X
0
0 0.3 0.6 0.9 1.2 1.5 1.8 2.0 3.0
VELOCITY (ft/sec}
r~ -Standard Error 1.00 X
w
0.80 0 z
0.60 >-
1-
0.40 ..J
CD
<(
0.20 1-
::>
C/)
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 3.5
DEPTH (ft.)
Mean catch of juvenile coho salmon per call by velocity and depth
intervals (bars) in low turbidity waters, Chulitna River to Devil
Canyon reach of the Susitna River. Suitability indices (lines)
fitted by hand.
-18-
"""
-
~I
-
r-
'
"""" I
believed that gear inefficiency was becoming a factor at the greatest
depths sampled and therefore the curves were drawn in at a higher
suitability than a close fitting of the data would warrant. The depth
curves did not drop. off to zero at the high ranges because we thought
depths did not limit juvenile distribution and we had no data for large
depths.
3.3 Analysis of Chinook Salmon Distribution in High Turbidity
Waters Using Beach Seine Data
3.3.1. Cover
Cover analysis of beach seine catch data is complicated by the fact that
gear effectiveness is reduced by the amount and type of object cover. A
least squares regression line was taken as a reasonable estimate of the
relationship between suitability and percent cover, however, and a
suitability index was normalized to the regression line (Figure 7). We
did not try to analyze the effect of object cover type on suitability
for chinook as it was obvious that the chinooks were using turbidity for
cover and thus the type of object cover present was probably not as
important.
10.0
_j 9.0
_j
w u 8.0
0::
w 7.0 a..
:::s:::
0
0 z
I u
6.0
5.0
I 4.0
u !;;t 3.0
u
z 2.0
<l: w ~ 1.0
I~ -Standard Error
1.00
0.80
X
W· a z
0.60 >-
1-
_j
0.40 CD
<l:
1-
0.20 ::J
(J)
0 ~-7.-'-:'::-::'---'--:-:-:~~--==-1..--l--:-:--::-1---L.~....L----1. 0.00
PERCENT COVER CATEGORIES
Figure 7. Mean ca~ch of juvenile chinook salmon per cell by percent cover
cate~o~1es {bars) and _fitted suitability index (line) in high
turb1d1ty waters, Chulltna River to Devil Canyon reach of the
Susitna River.
-19 -
3.3.2 Depth and velocity
Depth and velocity have much less effect on beach seine effectiveness
than does the amount and type of cover within the· range samp 1 ed and so
analysis of depth and velocity was identical to that used for the
electrofishing data. A two-way analysis of variance between depth,
velocity and catch per cell showed velocity to be significant (Table
11). Depth was not significant by itself as an effect and interactions
could not be assessed due to empty cells (in the analysis of variance
table classification).
Table 11. Analysis of variance between depth, velocity, and chinook
catch transfonned by ln (x+1) in high turbidity water. Due to
empty cells or a singular matrix, interactions could not be
calculated.
Chinook Sum of Mean Significance
Source of Variation Squares .ti. Square .E of F
Main Effects 43.617 10 4.362 5.160 <.. 0.001
Depth 5.965 4 1.491 1.764 0.136
Velocity 35.617 6 5.936 7.022 "'0.001
Explained 43.617 10 4.362 5.160 <. 0.001
Residual 261.212 309 0.845
Total 304.828 319 0.956
Even though depth was not statistically significant by itself, a curve
was fit by hand to the data for depth using professional judgement
because a trend was evident {Figure 8). A curve was a 1 so fit to the
velocity data by hand using professional judgement and a suitability
index derived (Figure 8). The data indicate that in turbid water,
chinook use shallower and slower moving water than they do in clear
water.
3.4 Analysis of Sockeye and Chum Salmon Proportional Presence
Using Pooled Electrofishing and Beach Seining Data
3.4.1. Cover
Since proportional presence was used as a measure of suitability instead
of catch per cell, standard analysis of variance techniques were not
used. Instead, chi-square tests of association were used to test for
differences in proportional presence among categories of percent cover
and cover type {Table 12). All these tests were significant and suita-
bility criteria were fit to the data. The five points of proportional
-20 -
-
-
-
..J
..J
UJ
u
a::
UJ a..
~
8 z
:X: u
z
<(
UJ
:!:
7.0
..J 6.0
..J
UJ
u 5.0
a::
UJ a..
~ 4.0
0
0
z ,3.0
i· u
:1: 2.0 u
1-
<t
u 1.0
z
<t
UJ
0
I ~-Standard Error
0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4
VELOCITY (ft. I sec.)
I~ -Standard Error
:!: 04-~T-~~~---.--~--~~--~----~~~~--~
0 0.2 0.4 0.6 0.8 lO 1.2 1.4 1.6 1.8 2.0 2.2
DEPTH (ft.)
l.OO
0.90
0.80 X
UJ
0.70 0 z
0.60
>-
0.50 1-
0.40 ::!
IIl
<t
0.30 1-
~
0.20 (f)
0.10
000
1.00
X
UJ
0.80 0
z
0.60 >-
1-
..J
0.40 m
<t
1-
0.20 ~
(f)
0.00
Figure 8. Mean catch of juvenile chinook salmon per cell by velocity and
depth intervals (bars) in high turbidity waters, Chulitna River to
Devil Canyon reach of the Susitna River. Suitability indices
(lines) fitted by hand.
-21-
presence were regressed to the percent cover categories and the re-
gression line was normalized to a suitability index {Figure 9). Cover
type suitability criteria were formed by dividing through by the percent
presence for zero cover ce1ls and then normalizing {Figure 10). Some
cover types were not any more suitable than the zero cover cells.
Table 12. Chi-square tests for differences in proportions of sockeye or
chum presence between habitat attribute groupings of percent
cover, cover type, velocity and depth.
Habitat
Species Attribute df Chi-square
Sockeye
Cover type 8 41.11**
Percent cover 4 19.05**
Velocity 6 28.68**
Depth 4 15.73*
Chum
Cover type 8 21.18*
Percent cover 4 23.65**
Velocity 5 11.06*
Depth 3 20.09**
*S1gnificant at p < 0.05
**Significant at p < 0.01
3.4.2 Depth and velocity
Chi-square tests indicated that the depth and velocity group intervals
were associated with both sockeye and chum proportiona1 presence (Table
12). Curves were fit to the data by hand using professional judgement
{Figures 11 and 12) and suitability indices normalized to the lines.
Velocity criteria were similar for both species but the depth criteria
indicated that sockeye salmon found deeper water more suitable while
chum used shallower water.
3.5 Tests of Fitted Habitat Values to Observed Fish Catches
3.5.1 Chinook and coho salmon
Once suitability indices were fitted to the data, various formulations
of composite weighting factors were corre 1 a ted with actua 1 fish catches
to evaluate their fit. Catches were transformed by ln (X+1) and Pearson
correlations were then run between the transformed catch and various
composite weighting factor combinations of habitat variables (Table 13).
-22 -
-
""'' '
""'!'!
~
"""
-1
-
"'""
-
I
N w
I
1
0.50 J!-Standard Error
0.40
0.30
020
0.10
PERCENT COVER CATEGORIES
070
~ I!· Standard Error
::: 0.60 ..
f
2 050
" r
0
~ 0.40
ji
3 0.30 ..
0
I'; 0.20
15
;::010 g
1.00
" 080 ~ z
0.60 ,.. ... :::;
!
~
0.40
ozo
1.00
0.80 )( ...
0
0.60 ~
>-...
0.40
:::;
;;; ... ...
0.20
:; ..
a: ... OJL~.~.~,.~7~N""•"•n~~.~., .. ,.~~•"-•75o~~.~.~z"•L--LO
{0-5"'1.1 (6~2.S%1 (26-50"1oJ (51-75'"1~1 (76-~/ .. 1
PERCENT COVER CATEGORIES
L_ ______________________________ __
Figure 9. Proportion of cells with juvenile sockeye and chum salmon present
by percent cover category (bars) and fitted suitability indices
(lines), Chulitna River to Devil Canyon reach of the Susitna
River.
1.00
>-
1-
::::! co
<t
1-
5
(f)
Figure 10.
1
------------.~-~-----
10 U> <r (\J ~ SOCKEYE II
2 2
r<l Ill CHUM .
<r 0
U>
2
0
<r
U>
10 I'-
<r 10
.. :; ..... ., .!:! 6 ~"i.i -~"' -c: -·.:: .,., :0 <=o Q) u>< -'0 g2 at 000 .. ·-..
Ll -c .c_ Ll <::·-.,.-0 <> GO LI::O " ...J-o-_o u ss """' Cl ~0 0>1ij Q ..,m "' c <f<JO _a. Eo 0 !t .,.-::> ,.a:: UJ~ z
0
COVER TYPE
Con1parison of cover type suitability indices for juvenile sockeye
and chum salmon, Chulitna River to Devil Canyon reach of the
Susitna River.
l
0.30
(/)I-
..Jz
..JUJ
UJ<J)
UUJ
0.20 a:
u..~
0 UJ
>-zUJ
Olil::
-0 1-0 a: <I)
~::1:
o!::
0.10
if3:
0
0.40
1-
z
UJ
fl')
UJ a: a.
UJ 0.30 >-UJ
~
0
0
(/)
::1:
1-
i 0.20
fl')
..J
..J
UJ
0 ....
0
z 0.10 0
i= a:
& a: a.
0
0
Figure 11.
I!-Standard Error
1.00
X 0.80 UJ
0
~
0.60
>-
1-
0.40 ..J
Ill < " 1-z 5 0.20 (/)
0
0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4
VELOCITY (ft /sec)
I'!: Standard Error
1.00
0.80 X
IJJ c
z
0.60 >-
1-
...J
0.40 CD
<t
1-
20 :::>
(/)
0.00
0.2 0.4 0.6 0.8 LO 1.2 1.4 1.6 1.8 2.0 3.6
·(Ma~}
DEPTH (ft,)
Proportion of cells with juvenile sockeye salmon present by
velocjty and depth intervals (bars), Chulitna River to Devil
Canyon reach of the Susitna River. Suitability indices (lines)
fitted by hand.
-24-
-
-
-
-
-
-
-
en
...JI-
...Jz ww
UU'l
lJ.J U.a::
Oa..
z~
~:::>
I-::I:
a::u
~:I:
o!::
a::3 a..
~1-...JZ ww ucn w u.a:: oa..
z~
0:::>
-::~: ~(.)
Oz a..l-0-C::3 a..
0.50
0.40
0.30
0.20
0.10
0
0.50
0.40
0.30
0.20
0.10
0
I : -Stand a rd Error
I
rt')
U') ..
z
0 0.3 0.6 0.9 1.2 1.5 1.8
VELOCITY (ft./sec.)
I-:: Standard Error
0 0.2 OA-0.6 0.8 1.0 L2 1.4 1.6 1.8
DEPTH (ft.)
1.00
)(
0.80 w
0 z
0.60 >-
1-
0.40 ...J
al
<t
1-0.20 5
(/')
0
2.1 2.4
1.00
)(
w
0.80 0
z
0.60 >-
1-
0.40 ...J
Cl
0.20 <(
1-
:::>
0.00 en
3.5
Figure 12. Proportion of cells with juvenile chum salmon present by velocity
and depth intervals (bars), Chulitna River to Devil Canyon reach
of the Susitna River. Suitability indices (lines) fitted by
hand.
-25-
The correlations range from 0.16 to 0.42, and all were statistically
greater than zero.
Table 13. Correlations between composite weighting factors generated
using various combinations of suitability indices and trans-
formed rln {X+l)] chinook and coho catch.
Composite Weighting Factor Calculation
(Percent cover)x(cover type)x(velocity)x(depth)
{Percent cover)x(cover type)x(velocity)
(Percent cover)x(cover type)
(Velocity)x(depth)
Limiting factor (minimum of
(percent cover x cover type), (velocity),
or (depth) taken as weighting factor)
Pearson correlations (r)*
Chinook Coho Chinook
(clear) (clear) (turbid)
0.42 0.36 0.31
0.41 0.38 0.30
0.35 0.37 0.16
0.28 0.30 0.28
0.43 0.39 0.32
N=813 N=813 N=813
*All correlations significantly greater than zero at the 0.01 signifi-
cance level.
Combinations of habitat variables with the highest carrel at ions are the
most likely candidates for applications in habitat modelling studies.
The low correlations are due to the fact that actual fish numbers are
influenced greatly by other factors such as season and site.
3.5.2 Sockeye and chum salmon
Sockeye and chum salmon proportional presence increased significantly
with increased magnitude of several composite weighting factor intervals
(Table 14). The largest composite weighting factor interval had an
associated proportional presence which was three to seven times the
proportional presence associated with the lowest composite weighting
factor interva 1.
-26 -
-
-
-
-
-
....
-
-
I"""'
-
-
Table 14. Proportional presence of sockeye and chum salmon fry
associated with several composite weighting factors.
Composite Composite Proportion
Weighting Weighting Total with
Factor Factor No. of Fish
Species Calculation Interva 1 cells Present Chi-Square
Sockeye Minimum factor of 0.0-0.12 269 0.12 62.9*
{percent cover x cover 0.12-0.2.0 321 0.08 df=3
type), (velocity) 0.20-0.33 312 0.22
or (depth) 0.33+ 111 0.38
Sockeye {Percent cover) x {cover 0.0-0.04 312 0.09 49.6*
type) x (velocity) 0.04-0.08 260 0.13 df=3
x (depth) 0.08-0.17 330 0.20
0.17 + 111 0.36
Sockeye (Percent cover) x (cover 0.0-0.08 341 0.09 50.8*
type) x (velocity) 0.08-0.14 253 0.12 df=3
0.14-0.30 308 0.22
0.31 + 111 0.35
Chum Minimum factor of 0.0-0.33 79 0.18 32.6*
(percent cover x cover 0.33-0.50 177 0.25 df=3
type} (velocity), 0.50-0.67 178 0.37
or (depth} 0.67+ 80 0.55
Chum (Percent cover) x (cover 0.0-0.17 77 0.09 49.6*
type) x (velocity) 0.17-0.31 171 0.26 df=3
x (depth) 0.31-0.53 177 0.37
0.53 + 89 0.56
Chum (Percent cover) x (cover 0.0-0.26 71 0.14 32.7*
type) x (velocity) 0.26-0.44 183 0.27 df=3
0.44-0.64 175 0.36
0.64 + 85 0.54
*All s1gnificant at p < 0.001
-27 -
4.0 DISCUSSION
Suitability criteria for juvenile salmon in the Susitna River have been
deve 1 oped by integrating stati sti ca 1 methods with profession a 1 judge-
ment. Somewhat novel design and analysis methods were used to overcome
problems that prevented the use of traditional applications in the
Susitna River system. Bovee (1982) reviewed the popular methods of
describing preference curve construction. The methods range from the
binary criteria used by Collings et al. (1972) to multivariate suitabil-
ity techniques explored by Voos {1981) and Prewitt (1982). Perhaps the
most widely used methods have been the probabi 1 i ty-of-use curves con-
struction techniques described by Bovee and Gochnauer {1977).
Baldrige and Amos (1983) have expanded Bovee and Cochnauer 1 s approach to
produce univariate suitability descriptions which minimize environmental
and sampling bias. Our techniques merge these authors 1 concepts of
environmental suitability, availability, and usability with an infre-
quently applied approach. Usability descriptions (defined as suita-
bility times avail abi 1 i ty) are commonly derived from collecting point
specific habitat measurements at locations where fish are observed.
These data are the probability of observing a value for an environmental
attribute {E), given fish (F), which is P[E/F] (Bovee 1982). This
practice cannot be easily ·implemented for juvenile salmon in large
turbid glacial systems. Instead, we have compiled the description
P[N/E], the probability of one or more fish (N), given a set of environ-
mental attribute values. This method, has the benefit of collecting
fish and physical habitat data in a manner that can be used to sub-
sequently verify model outputs. This was accomplished by establishing
the grid and cell sampling scheme over important rearing areas in the
reach. Bovee notes that two assumptions are made when P[N/E] distri-
butions are calculated directly: systematic random sampling is employed
and that the entire population is sampled. We view our experimental
design as stratified random sampling of selected areas· of the most
important macrohabitats available in the reach above the Chulitna
confluence. While we did not observe the whole population we believe
that representative data have been collected.
4.1 Limitations of the Suitability Criteria
Not all the factors which could have a major effect on the distribution
of juvenile fish were addressed in this study. We evaluated cover,
depth, and velocity but such factors as water quality and food produc-
tion also influence juvenile salmonid distribution (Reiser and Bjornn
1979). We may have addressed food production indirectly as Reiser and
Bjornn reported that velocity, depth, and substrates are correlated with
food supply. The water quality suitability differences within and
between sites are probably minimal with the exception of turbidity as
measured water quality attributes of dissolved oxygen and temperature
normally do not vary greatly from optimum ranges presented by Reiser and
Bjornn (1979).
These criteria are also specific to the Susitna River reach studied and
if used outside that reach they might not be valid. The suitability
criteria developed are also limited to the open-water time period from
-28 -
-
'!"'!!,
I
I
-
-
-
-
-
-
May to mid-October. Winter rearing habitat preferences are probably
different as fe.eding and activity of the fish are reduced. Bjornn
(1971) reported that juvenile salmon enter large rubble substrate when
stream temperatures drop below 4-6°C and will leave the area if this
cover type is not present.
The criteria are also limited by the values of the habitat attributes
which could be effectively sampled by the methods used. Velocities over
three feet per second and depths over two to three feet could not be
effectively sampled, for example. A preliminary experiment described in
Part 2 of this report suggested that sampling efficiency also decreased
slightly in cells with large amounts of cover.
Single habitat measurements used to describe a cell with diverse values
of habitat attributes like depth and velocity are often inadequate
descriptions. Since the curves are univariate, they also do not account
for interactions between variables such as depth and velocity.
Criteria also were not developed specifically by age class; however,
over 99% of the fish captured were 0+ fish and 1+ fish were pooled with
these to increase sample sizes. Suitability criteria might also shift
as a function of within year life history: larger fish of a given
species may prefer different habitat conditions as food sources and
behaviors change. (Chapman and Bjornn 1969; Everest and Chapman 1972).
4.2 Chinook and Coho Salmon
Chinook and coho salmon low turbidity suitability indices were developed
from the same data set. Electrofishing is perhaps the best method for
collecting juvenile fish in clear water as seining efficiency is affect-
ed strongly by cover. Because the backpack electroshocker is most
effective in shallow water, the depth curves were drawn so that the
suitability in deep water was actually higher than indicated by the
data. Wiley and Tsai (1983) concluded that the electroshocker (and also
beach seine) was more effective and consistent than seines for est·imat-
ing fish populations. Dauble and Gray (1980) concluded that electro-
fishing was better than beach seining for sampling irregular substrates
and higher velocities.
4.2.1 Chinook salmon
Chinook salmon were the. only species for which enough data were collect-
ed to generate suitability indices for both clear and turbid conditions.
Some shifts in preferences for habitat conditions are apparent. Lower
velocity waters are preferred under turbid conditions than under clear
conditions, as are shallower depths (Figures 5 and 8). Juvenile chinook
salmon possibly prefer lower velocities in turbid water because when
using the turbid water as cover, they have no velocity breaks to hide or
·rest behind. Cover might still be useful, however, as a break from
velocity. A shift in depth preference may be due to the fish reacting
to high suspended solid concentrations by staying near the surface
(Wallen 1951 as cited in Beauchamp et al. 1983).
-29 -
The preference for object cover appears 5tronger in clear water than in
turbid water for chinook salmon because of the higher suitability for
low cover cells and lesser slope of the cover regression line in turbid
than in clear water. This limited preference for object cover in turbid
water is partly due to gear bias as beach seining is quite ineffective
where large amounts of object cover are present. However, the distribu-
tion of chinook salmon is clearly different in clear than in turbid
water. In turbid waters, such factors as depth and velocity most limit
and influence distribution while in clear water, object cover seems more
important. MacCrimmon · (1954) noted Atlantic salmon fry use of turbid
water for cover.
The velocity probability-of-use curves for juvenile chinook salmon
presented in Bovee (1978) and Burger et al. (1982) are almost identical
with the curve developed for chinooks in clear water of the Susitna
River with the peaks at approximately 0.2 to 0.6 ft /sec. Minnow trap
chinook catch data from the Little Susitna River also suggest the
optimum velocity for chinook salmon to be approximately 0.3 to 0.6
ft /sec with little use of velocities greater than 1.8 ft /sec
(Delaney and Wadman 1979).
Depth criteria developed in other systems for juvenile chinook salmon
vary significantly from those presented here, where optimum depths were
1.0 to 1.5 ft in clear water and less than 0.5 ft in turbid water. A
depth probability-of-use curve presented in Bovee (1978) for chinook
salmon shows an optimum range from 1.2 ft up to at least 3.0 ft in
depth, while data presented in Delaney and Wadman's (1979) data suggest
·an optimum of 2.5 to 3.2 ft Burger et al. (1982) observed chinook fry
in pools to ten feet in depth and thought depths of less than 0.2 ft
were avoided. Correlations of depth with other important distributional
factors which may vary from river to river probably cause much of this
variation in the form of the depth suitability functions.
4.2.2 Coho salmon
In contrast to chinook salmon, coho salmon do not appear to use turbid
water as cover. Bisson and Bilby (1982} reported that coho salmon
avoided turbidities of 70 to 100 NTU under experimental conditions and
Sigler et al. (1984) found, in a laboratory study, that more juvenile
coho salmon emigrated from channels with a turbidity level of 25-50 NTU
than from clear water channels. These turbidity levels are frequently
exceeded during the ice free months in side channels of the Susitna
River. Catches of coho salmon were very low in turbid side channels
(see Part 2 of this volume). Cover types preferred by coho, i.e. debris
and undercut banks, are also very scarce at these sites, however, and
almost impossible to sample effectively with beach seines. It may be
that coho usually leave a site when turbidities exceed a certain level.
The distribution of coho salmon fry may be 1 imited greatly within a
clear water area by the lack of suitable cover type, as very strong
preferences for a few cover types were noted (Figure 4). In contrast to
chinook salmon, substrate was little used as cover while preferred
velocities and depths were also somewhat different. Bustard and Narver
(1975) also noted that coho preferred bank cover in the form of undercut
-30 -
-
-
-
-
banks rather than instream cover. Social interactions between the two
species could cause these differences (Stein et al. 1972) but
intraspecific interactions and microhabitat preferences might be most
important (Allee 1981).
Bovee (1978) presented a velocity suitability curve for coho fry very
similar to that presented in this report with a slightly higher optimum
of 0.5 ft /sec. and a minimum at 2.3 ft /sec. Burger et al. (1982)
presented utilization curves with optimums at 0.0 ft /sec , but which
then quickly dropped to very low suitabilities at velocities greater
than 0.2 ft /sec. Habitat suitability criteria from the Terror and
Kizhuyak Rivers for coho salmon juveniles also presented optimum veloc-
ities at 0.0 to 0.4 ft /sec (Baldridge 1981) as do those suggested by
Delaney and Wadmans• (1979) data. Optimum velocities for coho derived
in this report are therefore very similar to velocity criteria developed
for coho in other streams.
Depth criteria, on the other hand, vary greatly from stream to stream.
On the Terror and Ki zhuyak rivers, optimum depths for coho fry ranged
from near 0.0 ft to 1.0 ft and then declined rapidly to zero at 2.5 ft
(Baldrige 1981). Data presented in Bovee (1978), however, indicate very
1 ittle use until 1.0 ft in depth with an optimum at 2.0 ft and a
gradual decline to zero use at 5.0 ft. In the Susitna River, the
optimum suitability appeared to occur at approximately 1. 6 to 2.0 ft
with limited data above this depth. These conflicting data show that
depth suitability may vary greatly from river to river for unknown
reasons, a 1 though carrel ati ens of depth with other important factors
influencing distribution are probable.
4.3 Sockeye and Chum Salmon
The sockeye and chum suitability indices are less reliable than for
chinook and coho as the numbers, distribution, and seasonal use of
habitat is smaller for these species. The seasonally reduced sampling
and need for large sample sizes also made it necessary to pool gear
types to adequately address the range of habitat conditions encountered
during the study. The schooling behavior of these species also caused
us to put catch on a presence-absence basis for purposes of analysis.
4.3.1 Sockeye salmon
Sockeye salmon were apparently much less dependent on cover than were
chinook or coho salmon because they occur in schools and use the school-
ing as a means of predator avoidance. Schools of sockeye were observed
ranging throughout areas which varied from heavy cover to no cover at
all. Depth and velocity, therefore, could have a much larger effect on
their distribution. However, from the analysis, the distribution of
junvenile sockeye salmon did appear to be related with cover. The
suitability curves for depth and velocity both indicate a fish that
rears in a lacustrine environment. The effect of turbidity on sockeye
salmon distribution is unknown. A limited review of the literature
indicated that suitability criteria for stream rearing sockeye
populations have not been developed. Burger et al. (1982) presented a
velocity probability-of-use curve for sockeye in the Kenai River with an
-31 -
optimum at 0.0 ft /sec and very little use at velocities greater than
0.6 ft /sec.
Sockeye salmon have a 1 imited distribution in the upper Susitna River
basin. Most of the rearing appears to be limited to sites along the
mainstem Susitna which offer lacustrine environments. However, we had
no means of effectively sampling these types of habitat areas in this
study.
4.3.2 Chum salmon
Of the four species of salmon which rear in the middle Susitna River,
chum salmon rear for the shortest period of time (ADF&G 1983b). Little
is known about the rearing requirements of chum salmon but they have
been reported to use substrate as cover initially (Neave 1955) and then
after schooling, use the protection of the schools (Hoar 1956). Both
these behaviors of chum salmon fry were observed in the Susitna River
and the suitability indices reflect a larger relative use of large
substrate for cover by chum salmon than for sockeye salmon. As the
amount of cover increased greatly, however, the change in use by juve-
nile chum salmon was very similar to sockeye salmon. Shallow depths and
low velocity water were found most suitable for chum salmon fry in this
study. Mean catches of juvenile chum salmon were less in cells without
object cover in turbid water which suggests avoidance of turbid con-
ditions. On the other hand, this may also have been an artifact of the
influences of natal areas on distribution with clear water near emer-
gence areas affecting the results.
4.4 Recommended Applications for the Suitability Criteria
The suitability criteria for juvenile salmon in the Susitna River reach
between the Chulitna River confluence and Devil Canyon which are
recommended for use in calculating weighted usable area are listed in
Appendix Table C-1.
Suitability criteria, in conjunction with hydraulic models, are one
means of calculating changes in habitat with changes in flow. Typical-
ly, weighted usable areas (WUA's) are calculated for a series of dis-
charges and these are taken as representing changes in the desirability
of habitat. There are several standard methods for calculating WUA's by
multiplying area with composite weighting factors which are combinations
of suitability indices of factors believed to have major effects on dis-
tribution. Suitability indices can be multiplied together, the geo-
metric mean can be taken, or the lowest suitability index for attributes
of importance can be used as the composite weighting factor (Mil hous
e t a 1. 1981 ) .
We have calculated composite weighting factors for various combinations
of habitat attributes and compared the composite weighting factor to
observed fish catch (Tables 13 & 14). The geometric mean was not used
for integrating suitability indices as this implies a compensatory
effect that does not seem biologically reasonable for juvenile
salmonids. The correlations are very similar for various combinations
and are consistently low. Other formulations of composite weighting
-32 -
-
-
-
-
-'
factors are possible and these could produce better correlations, but
time constraints prevented further testing.
Effects of depth on the distribution of juvenile salmon are probably
limited as depth typically by itself would not limit the distribution of
fish. Correlations with other factors like site, season, or velocity
may make depth seem more important than it is. When depth was eliminat-
ed from calculations of the composite weighting factor, little reduction
in the correlations of catch with weighting factors was noted. By
including depth in the calculations, however, equal weight is given to
depth with cover and velocity and this weighting can drive changes in
WUA with discharge as was noted in trial runs with models discussed in
Part 4 of this report. Since depth is not as limiting in a behavioral
or physical sense as cover and velocity are, its applicability to
habitat modelling as equally weighted with velocity or cover is dubious.
Analyses of variance, however, suggested that depth and velocity inter-
actions were sometimes significant and that fish were not selecting
habitat on the basis of velocity independent of depth (Table 10).
Interactions of depth and velocity have been shown in at least one other
study (Orth and Maughan 1982) to affect WUA 1 s when depth and velocity
were multiplied together to generate composite weighting factors.
Because the inclusion of depth in the composite weighting factors did
not improve the correlation with fish density, we decided to discount
the effect of depth at depths greater than 0.15 ft in the composite
weighting factors which were used in projecting weighted usable area in
Part 4 and Part 7 of this report. This was done by setting the
suitability index to 1.0 for all depths greater than or equal to 0.15
ft. and represents a departure from the depth suitability indices
presented in the results section. The 0.15 point is somewhat arbitrary,
but there is little data to go on. Burger et. al (1982) as previously
suggested that chinook salmon avoided depths of less than 0.2 ft.
Obviously, a depth of 0.0 ft. has a suitability index of 0.0.
If turbidity is used as cover, then depth suitability is not independent
of turbidity. At shallower depths, water of a given turbidity may not
provide cover, while deeper waters may provide excellent cover. Secchi
disc transparencies measured in Eklutna Lake decreased from 3.0 to 1.4
ft. over a turbidity range of 18 to 36 NTU (R & M Consultants, 1982).
Cover for fish would be provided at shallower depths than indicated by
Secchi disc readings due to their cryptic coloration. The relationship
of turbidity to light penetration, water depth, and related cover value
has not been quantified in the Susitna River.
The minimum factor approach which implies that the habitat is no better
than the most limiting attribute is biologically reasonable. The
calculated fit with the observed data was as good as the other approach-
es used. When the minimum factor was used as the composite weighting
factor, cover was often the minimum factor for chinook and coho salmon
in clear water, velocity was secondarily important, and depth was only
occasionally the minimum factor. Reiser and Bjornn (1979) reviewed the
importance of cover in the literature and found that salmonid abundance
-33 -
declined and increased as cover was removed or added to streams in a
number of instances. Burger et al. (1982) reported that velocity was
perhaps the most limiting factor for juvenile chinook in the Kenai River
but that the fry also moved from areas where suitable cover types in the
form of steep vegetated banks no longer existed. Depth was not men-
tioned in these studies as having much of an influence on distribution,
and therefore probably should not be weighted the same as cover or
velocity. If cover and velocity are weighted with equal importance and
depth suitability is held constant, determinations of WUA 1 s for juvenile
salmon will perhaps be most valid.
The suitability criteria which have been developed in this paper
represent a compendium of the data from the 1983 field study and three
years of experience in observing and sampling these populations.
Although there are limitations to the suitability criteria technique, we
are confident that the curves presented are reasonably accurate for this
reach of river and will lead to weighted usable area projections which
are of value in predicting effects of changes in flow on juvenile salmon
habitat.
-34 -
-
-
-
-
-
-
-
-3
1
-
-
5.0 CONTRIBUTORS
Field sampling was conducted by Paul Suehanek, Larry Dugan, Robert
Marshall, and David Sterritt. Carol Kerkvl iet drafted the figures.
Allen Bingham provided assistance with the data analysis. Donna
Buchholz keypunched the data and Gail Heineman organized the database.
Larry Bartlett, Allen Bingham, and Kathrin Zosel reviewed the
manuscript.
-35 -
6.0 ACKNOWLEDGEMENTS
We thank Ken Voos, Arctic Environmental Information and Data Center
(AEIDC), and Deborah Amos (ADF&G), for taking the time to discuss
analysis procedures. Funding for this study was provided by the State
of Alaska, Alaska Power Authority.
We wish to thank the various staff of E. Woody Trihey and Associates
(EWT&A)-, the Arctic Environmental Information Data Center, Harza-Ebasco
Susitna Joint Venture, Woodward-Clyde Consultants, the Alaska Power
Authority, and especially Cleve Steward (EWT&A) for reviewing a draft of
this paper and providing helpful suggestions.
-36 -
-
-
-
-
-
-
-
....
7.0 LITERATURE CITED
ADF&G. 1983a. Susitna Hydro aquatic studies phase II report. Synopsis
of the 1982 aquatic studies and analysis of fish and habitat
relationships. Alaska Department of Fish and Game Sus.itna Hydro
Aquatic Studies. Anchorage, Alaska.
--,· 1983b. Susitna Hydro aquatic studies phase II basic data report
Volume 3. Resident and juvenile anadromous fish studies below
Devil Canyon, 1982. Alaska Department of Fish and Game Susitna
Hydro Aquatic Studies. Anchorage, Alaska.
• 1984. Susitna Hydro aquatic studies May 1983 -June 1984 --procedures manual. Alaska Department of Fish and Game Susitna
Hydro Aquatic Studies. Anchorage, Alaska.
Allee, B.A. 1981. The role of interspecific competition in the dis-
tribution of salmonids in streams. Pages 111-122 in E.L. Brannon
and E.O. Salo, editors. Salmon and Trout Migratory Behavior
Symposium.
Baldrige, J. E. 1981. Development of habitat suitability criteria.
Appendix 3. in Wilson et al. 1981. An assessment of environ-
mental effects-of construction and operation of the proposed Terror
Lake hydroelectric facility, Kodiak, Alaska. Arctic Environmental
Information and Data Center, University of Alaska, Anchorage,
Alaska.
Baldrige, J. E., and D. Amos. 1983. A technique for determining fish
habitat suitability criteria: A comparison between habitat uti-
lization and availability. Paper presented at the Symposium on
Acquisition and Utilization of Aquatic Habitat Inventory Informa-
tion. (American Fisheries Society, Portland, Oregon. October
28-30, 1981).
Beauchamp, D. A., M. F. Sheperd, and G. B. Pauley. 1983. Species
profiles: Life histories and environmental requirements (Pacific
Northwest) chinook salmon. U. S. Department of the Interior, Fish
and Wildlife Service. FWS/OBS-83/1.
Bisson, P. A., and R. t. Bilby. 1982. Avoidance of suspended sediment
by juvenile coho salmon. North American Journal of Fisheries
Management 4:371-374 •
Bjornn, T. C. 1971. Trout and salmon movements in two Idaho streams as
related to temperature, food, stream flow, cover, and population
density. Transactions of the American Fisheries Society 100:
423-438.
Bovee, K. D. 1978. Probabi 1 ity-of-use criteria for the family
Salmonidae. Instream Flow Information Paper. No. 4. U.S. Fish and
Wildlife Service. FWS/OBS-78/07.
Bovee, K. D. 1982. A guide to ·stream habitat analysis using the in-
stream flow incremental methodology. Instream Flow Information
Paper. No. 12. U.S. Fish and Wildlife Service. FWS/035-82/26 .
. -37 -
Bovee, K. D., and T. Gochnauer. 1977. Development and evaluation of
weighted criteria, probability-of-use curves for instream flow
assessments: fisheries. Instream Flow Information Paper 3. USDI.
U.S. Fish and Wildlife Service. FWS/OBS-77/63.
Burger, C. V., D. B. Wangaard, R. L. Wilmot, and A. N. Palmisano. 1982.
Salmon investigations in the Kenai River, Alaska 1979-1981. U. S.
Fish and Wildlife Service, Anchorage, Alaska.
Bustard, D. R., and D. W. Narver. 1975. Aspects of the winter ecology
of juveni 1 e coho salmon (Oncorhynchus ki sutch) and stee·l head trout
(Salmo gairdneri). Journal of the Fisheries Research Board of
Canada 31:667-680.
Chapman, D. W., and T. C. Bjornn. 1969. Distribution of salmonids in
streams with special reference to food and feeding. Pages 153-176
in T. G. Northcote, editor. Symposium on Salmon and Trout in
Streams. H. R. MacMillan Lectures in Fisheries, University of
British Columbia, Vancouver, Canada.
Collings, M. R., R. W. Smith, and G. T. Higgins. 1972. The hydrology
of four streams in western Washington as related to several Pacific
salmon species. U. S. Geological Survey Water Supply Paper 1968.
Dauble, D. D., and R. H. Gray. 1980. Comparison of a small seine and a
backpack electroshocker to evaluate nearshore fish populations in
rivers. The Progressive Fish-Culturist 42(2): 93-95.
Delaney, K., and R. Wadman. 1979. Little Susitna River juvenile
chinook and coho salmon study. Alaska Department of Fish and Game,
Sport Fish Division. Anchorage, Alaska.
Dixon, W. J., and F. J. Massey, Jr. 1969. Introduction to statistical
analysis. McGraw-Hill Book Company, New York, USA.
Everest, F., and D. Chapman. 1972. Habitat selection and spatial inter-
action by juvenile chinook salmon and steelhead trout in two Idaho
streams. Journal of the Fisheries Research Board of Canada
29:91-100.
Hoar, W. S. 1956. The behavior of migrating pink and chum salmon fry.
Journal of the Fisheries Research Board of Canada 13(3):309-325.
MacCrimmon, H. R. 1954. Stream studies on planted Atlantic salmon.
Journal of the Fisheries Research Board of Canada 11:362-403.
Milhous, R. T., D. L. Wegner, and T. Waddle. 1981. User 1 S guide to the
physical habitat simulation system. United States Fish and ltJild-
life Service. Biological Services Program FWS/OBS-81/43.
Neave, F.
fry.
374.
1955. Notes on the seaward migration of pink and chum salmon
Journal of the Fisheries Research Board of Canada 12(3):369-
-38 -
._,1
....
.....
.....
-
-
-
Nickelson, T. 1976. Development of methodologies for evaluation of
instream flow needs for salmonid rearing. Pages 588-596 in J.F.
Orsborn and C.H. Allman, editors. Instream flow needs, VolUme 2.
Western Division, American Fisheries Society, Bethesda, ~1aryland,
USA.
Nie, N. H., C. H. Hull, J. G. Jenkins, K. Steinbrenner, and D. H. Bent.
1975. Statistical package for the social sciences. 2nd ed.
McGraw-Hill Book Co., New York,USA.
Orth, D. J., and 0. E. Maughan. 1982. Evaluation of the incremental
methodology for recommending instream flows for fishes. Trans-
actions of the American Fisheries Society 111:413-445.
Prewitt, C. G. 1982. The effect of depth-velocity correlations on
aquatic physical habitat usability estimates. Doctoral dis.serta-
tion, Colorado State University, Fort Collins, Colorado, USA.
R & M Consultants, Inc. 1982. Task 3-Hydrology Glacial Lakes Studies
Interim Report. Alaska Power Authority, Susitna Hydroelectric
Project, Anchorage, Alaska.
Reiser, D. W., and T. C. Bjornn. 1979. Influence of forest and
rangeland management on anadromous fish habitat in ~!estern North
America -1. Habitat requirements of anadromous salmonids. Gen.
Tech. Rep. PNW-96 Pacific Northwest Forest and Range Experiment
Station, Forest Service, U.S.D.A. Portland, Oregon.
Sigler, J.W., T.C. Bjornn, and F.H. Everest. 1984. Effects of chronic
turbidity on density and growth of steelheads and coho salmon.
Transactions of the American Fisheries Society 113:142-150.
Stein, R. A., P. E. Reimers, and J. D. Hall. 1972. Social interaction
between juvenile coho (Oncorhynchus kisutch) and fall chinook
salmon (~ tshawhscha) in Sixes River, Oregon. Journal of the
Fisheries Researc Board of Canada 29:1737-1748.
Voos, K. A. 1981. Simulated use of the exponential polynomial/maximum
likelihood technique in developing suitability of use functions for
fish habitat. Doctoral dissertation. Utah State University,
Logan, Utah, USA.
Wallen, I. E. 1951. The direct effect of turbidity on fishes. Doctoral
dissertation. University of Michigan, Ann Arbor, Michigan, USA.
Wiley, M. L., and C. Tsai. 1983. The relative efficiencies of
electrofishing vs. seines in piedmont streams of Maryland. North
American Journal of Fisheries Management 3:243-253.
-39 -
-
APPENDIX A
Calculations of Suitability of Cover Type for
Chinook and Coho Salmon in Clear Water
-40 -
.~
.....
Calculations of suitability of cover type for chinook and coho salmon in
clear water.
X c .. ) Weighted mean 2 ., J
effect of cover i = I Ki =
type j l(
2. c ..
i = I
a, J
N .. Number of fish captured 1 ,J =
in percent cover category i and cover type category j
c. . = Number of cells sampled
1 ,J
in percent cover category i and cover type category j
i = Percent cover category
j = Cover type category
X = Number of percent cover categories = 5
y y
K. = 2_ N; ~~ C; . = Mean catch for all cover types pooled in percent 1 J. = I ' J. = I ,J cover category i
y = Number of cover types = 9
-41-
Hypothetical example:
1. Samp 1 e data
Percent Cover
Category
1) 0-5%
2) 6-25%
Primary
Cover Type
1) Emergent vegetation
2) Undercut banks
3) Boulders
3 I 3
1< -~ Nl .. 2 l -,J
j=l j = I
1) Emergent vegetation
2) Undercut banks
3) Boulders
3 3
K2=L N2 . ~~ , J j=l
Chinook
Captured
( N. . )
1 ,J
1
5
4
c, . = ,J
5
10
15
c2 .
' J =
Cells
Sampled
(c. . )
1 ,J
5
10
5
I 0 I 20 = 0.5
10
10
10·
3 0 I 30 :: 1.0
2. Calculations of average effect of cover types on chinook distribution
2
( N~,i 1 )
_I_ + 5
Weighted mean ~ = 0.5 TO
= effect of = = = 0.47
emergent vegetation 2 5 + 10
~ C. I I ,
i =
Weighted mean 2
( N~ ,i2 )
5 10
effect of ~ --= 0.5 + 1.0 i = I
~o~ndercut. banks = = I. 00
2 10 + 10 ~ c. 2 I t
i = I
-42-
~I
I
-
-
-
-
2
( :i~3) Weighted mean ~ 4 15 +
effect of i = I 0.5 1.0 = = = 1.53 boulders 2 5 + 10 > C· 3
I ' 1
i = I
3. Normalize to 1.0 by dividing each effect by the largest effect
Emergent Vegetation
Undercut banks
Boulders
Weighted Mean
Effect
0.47
1. 00
1. 53
-43-
Suitability
0.47/1.53 = 0.31
1.00/1.53 = 0.65
1.53/1.53 = 1.00
-
r
-
APPENDIX B
Calculations of Effect of Cover Type on Distributions
of Sockeye and Chum Salmon
-44 -
r-
-I
r-
-
-
-
Calculations of effect of cover type on distributions of sockeye and
chum salmon.
Effect of
cover type j = E . = J
R
If less than
1. 0 then E j = I. 0
= effect of no cover
pj
cj
=
=
Number of cells of cover type j sampled with fish present
Number of cells of cover type j sampled
R = N1;c 1 = Proportional presence of fish in cells without object cover
Hypothetical example:
1. Sample data
Primary Cells Number of Cells
Cover Type Sampled { Cj) Sampled with Sockeye Present
1) No object cover 15 5
2) Emergent vegetation 20 5
3) Undercut banks 20 8
4) Boulders 50 25
-45-
( Nj)
2. Calculations of average effect of cover type on sockeye distribution.
Effect of
emergent =
vegetation
Effect of
undercut =
banks
Effect of =
boulders
P 2 I. C 2 5 I 20
=
R 0. 33
Since less than I. 0 change
p 3 I c3 8 I 20
:: ::
R o. 33
p4 I c4 25 I 50
:: =
R o. 33
3. Normalize to 1.0 by dividing each effect
Effect
No cover 1.00
Emergent vegetation 1.00
Undercut banks 1.21
Boulders 1.52
-46-
= 0. 76
to equal 1. 0.
I. 21
1.52
by the largest effect
Suitabi 1 ity
1.00/1.52 = 0.66
1.00/1.52 = 0.66
1.21/1.52 = 0.80
1.52/1.52 = 1.00
-
-
-
-
"""1
-
....
-
APPENDIX C
Suitability indices for juvenile salmon for cover,
velocity, and depth
-47 -
.... l 1
Appendix Table C-1. Suitability indices for juvenile salmon for cover, velocity, and depth.
Chinook
Cover Suitabilit~
Chinook
% Cover 1 PHABSIM (high ( 1 ow
Cover t~~e Code turbidit~} turbi dit~} Coho Socke~e Chum
No cover 0-5% 1.1 0.45 0.01 0.00 0.11 0.29
Emergent vegetation 0-5% 2.1 0.57 0.01 0.03 0.18 0.29
76-100% 2.5 1.00 0.12 0.29 0.47 0.53
Aquatic vegetation 0-5% 3.1 0.57 0.07 0.07 0.39 0.29
76-100% 3.5 1.00 0.68 0.65 1.00 0.53
Debris/ dead fa 11 0-5% 4.1 0.57 0.11 0.10 0.19 0.47
""" 76-100% 4.5 1.00 1.00 0.90 0.49 0.87
(X)
Overhanging 0-5% 5.1 0.57 0.06 0.04 0.30 0.40
riparian vegetation 76-100% 5.5 1.00 0.61 0.38 0.78 0.74
Undercut banks 0-5% 6.1 0.57 0.10 0.12 0.11 0.40
76-100% 6.5 1.00 0.97 1.00 0.29 0.74
Large gravel {1-311) 0-5% 7.1 0.57 0.07 0.03 0.17 0.37
76-100% 7.5 1.00 0.63 0.24 0.44 0.68
Rubb 1 e ( 3-5'') 0-5% 8.1 0.57 0.09 0.02 0.12 0.54
76-100% 8.5 1.00 0.81 0.18 0.30 1.00
Cobble or boulder 0-5% 9.1 0.57 0.09 0.02 0.11 0.46
( ) 511) 76-100% 9.5 1.00 0.89 0.18 0.29 0.86
1 With the exception of the "no cover" cover type, there are three other precent cover categories
for each cover type between the 0-5% and 76-100% categories. Suitability values for these
rnvPr tvnP~ ~rP linP~rlv intPrnnlAtPrl frnm thP two Pndnnints oiven. PHABSIM codes for the
Appendix Table C-1 (continued)
VELOCITY
Chinook (turbid) Chinook (clear} Coho Socke~e Chum
Velocity Suita-Velocity Suita-Velocity Suita-Velocity Suita-Velocity Suita-
{ft/sec} bil it~ {ft/sec} bi lit~ {ft/sec) bi lit~ (ft/sec) bi lit~ (ft/sec) bil it~
0.00 0.42 0.00 0.18 0.00 0.29 0.00 1.00 0.00 0.86
0.05 1.00 0.20 0.57 0.05 1.00 0.05 1.00 0.05 1.00
0.35 1.00 0.35 1.00 0.35 1.00 0.20 0. 71 0.35 1.00
0.50 0.80 0.65 1.00 0.50 0.88 0.50 0.48 0.50 0.87
+=> 0.80 0.38 0.80 0.68 0.80 0.55 0.80 0.36 0.80 0.70
1.0 1.10 0.25 1.10 0.44 1.10 0.32 1.10 0.27 1.10 0.56
1.40 0.15 1.40 0.25 1~40 0.12 1.40 0.17 1.40 0.37
1.70 0.07 1. 70 0.18 1. 70 0.04 1. 70 0.09 1.70 0.15
2.00 0.02 2.00 0.12 2.00 0.01 2.00 0.02 2.00 0.03
2.30 0.01 2.30 0.06 2.10 0.00 2.10 0.00 2.10 0.00
2.60 0.00 2.60 0.00
DEPTH (A 11 Species)
De~th (ft} Suitabilit~
0.00 0.00
0.14 0.00
0.15 1.00
10.00 1.00
.I 1 .. J J I
PART 4
Juvenile Salmon Rearing Habitat Models
r
JUVENILE SALMON REARING HABITAT MODELS
1984 Report No. 2, Part 4
by Robert P. Marshall, Paul M. Suchanek, and Dana C. Schmidt
ABSTRACT
Alaska Department of Fish and Game
Susitna Hydro Aquatic Studies
2207 Spenard Road
Anchorage, Alaska 99503
The effects of mainstem discharge on rearing habitat of juvenile salmon
in the Susitna River reach between the Chulitna River confluence and
Devil Canyon were quantified by use of habitat models. Six slough and
side channel sites were sampled at four to seven different levels of
mainstem discharge during the 1983 open water season. Data were
collected on hydraulic characteristics, cover, water quality, water
surface area, and fish density. Suitability criteria were integrated
with the habitat data to calculate weighting factors for cover and
velocity for selected species at each site. These weighting factors,
which were calculated for both shoreline and mid-channel areas, were
then combined with area to produce weighted usable areas for the site.
A habitat index was then calculated for site comparisons. Peaks in
habitat indices for chinook salmon occurred when slough or side channel
heads were overtopped. Upland slough habitat indices steadily increased
with mainstem discharge. Lack of cover may limit juvenile salmon use of
many of the sites.
i
-
JUVENILE SALMON REARING HABITAT MODELS
TABLE OF CONTENTS
ABSTRACT ••••••••••••••••••••••••••••••••••••••••••••••••••••
LIST OF FIGURES •••••.••••..•••••••••••••••••••••.•••••••.•••••
LIST OF TA.BLES ••••••.••.•••••••.••••.•••••••••.•••••••••.••.••
LIST OF PLATES •••••••••••••••••••.••••..••••••••.•••••.••.••••
LIST OF APPENDIX TABLES •••••••••••••••••.•••••.••
1.0
2.0
3.0
INTRODUCTION ••••.••
METHODS ••••••••••••
2.1
2.2
Field Sampling Design •..••••••..••••••.••.••.••••..•
2 .1.1
2 .1.2
2 .1.3
Study site location and selection criteria •••
Sampling grid design ••••••••.•••••••••.•
Cell measurements .•••••...•••••••••.••..
Data Analysis ..•...•....•.......
2.2.1
2.2.2
2.2.3
Surface areas ............................... .
Mode 1 ••••••••••••••••••••••••••••••••••••••••
Model verification •••••••••••••••.••••••••.••
RESULTS ••••••••••••••••••••••••••••••••••••••••••.•••••••
3.1 Surface Areas .••••.•••.•.
3.2 Side Channel lOA .••••....•••.••..•••••.••..•.•••.•.•
3.3 Slough 22 ••••••••••••••••.•.•••••••••••••...•.••••••
3.4 Whiskers Creek Slough ••••••••••.••••.•••
3.5 s 1 ou g h 8 ••••••.••••••••.•.•••.••..•.•.•.•.••••..••..
3.6 Slough 5 .....•..•...................•...............
3.7 Slough 6A .......................................... .
3.8 Model Verification .••••••••••.•••••••.••.•••••.•.•.•
i i
i
iv
vi
vii
viii
1
3
3
3
3
12
13
13
16
18
20
20
20
20
29
29
29
35
35
4.0 DISCUSSION ••.•. o ••••••••••••••••••••••••••••• ~............ 39
4.1 Chinook Salmon ............ 6 •••••••••••••••••••••••• o.. 39
4.2 Coho Salmon ••..•...••.••.••..•.•....•...•• "'.......... 39
4.3 Sockeye and Chum Salmon •.•••....•••.•••...•..•••••..
4.4 Limitations of the Models Regarding
Meth~do 1 ogy ....................................... .
42
42
4.5 Model Verification.................................. 42
5. 0 CONTRIBUTORS ••••••••••••••••••••••••••••••••••••••• 8 ••••• 45
46
47
6.0
7.0
. ACKNOWLEDGEMENTS ...••...••..•.••...•••..•....•••...•.....
LITERATURE CITED. G •••••••••••••••••••••••••••••••••••••••
8.0 APPENDICES
Appendix A Weighted Usable Area and Habitat Indices
Tabulated by Site and Species .......•••..
iii
48
-
-
....
-
-
-
......
-
-
LIST OF FIGURES
Figure
1
2
3
4
5
6
7 -
8
9
10
11
12
-
Title
River mile and relative location of the
juvenile salmon rearing habitat model
study sites......................................... 4
Illustration of the grid and cell sampling
scheme employed at habitat modelling study
sites .............................................. .
Data analysis flow chart for juvenile
11
salmon rearing habitat models....................... 14
Total wetted surface areas measured and
extrapolated in the Side Channel lOA
habitat model study site............................ 21
Tota 1 wetted surface areas measured and
extrapolated in the Slough 22 habitat
model study sites................................... 22
Tota 1 wetted surface areas measured and
extrapolated in the Whiskers Creek Slough
habitat model study site............................ 22
Tota 1 wetted surface areas measured and
extrapolated in the Slough 8 grid 1
habitat model study site............................ 23
Tot a 1 wetted surface areas measured and
extrapolated in the Slough 8 grid 2
habitat model study site............................ 23
Total wetted surface areas measured and
extrapolated in the Slough 6A habitat
mode 1 study s i te. • . . • . • • • • • . • • • . . . • . . • • • . . . • • • . . . • • • 2 4
Tota 1 wetted surface areas measured and
extrapolated in the Slough 5 habitat model
study s i te . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 4
Weighted usable area projections for
juvenile chinook salmon at the Side
Channe 1 lOA mode 11 i ng site. . • . • . • • . • • • • • • • • . . . • . • • • • 25
Mid-channel area weighting factors for
juvenile chinook salmon at the Side
Channel lOA modelling site.......................... 26
iv
LIST OF FIGURES (Continued)
Figure
13
14
15
16
17
18
19
20
21
22
23
24
25
Title Page
Shoreline area weighting factors for
juvenile chinook salmon at the Side
Channel lOA modelling site.......................... 27
Weighted usable area projections for
juvenile chinook salmon at the Slough 22
mode 11 i ng site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Weighted usable area projections for
juvenile chinook salmon at the Whiskers
Creek Slough modelling site......................... 30
Weighted usable area projections for
juvenile coho salmon at the Whiskers Creek
Slough modelling site............................... 31
Weighted usable area projections for
juvenile coho salmon at the Slough 8
modelling site...................................... 32
Weighted usable area projections for
juvenile so.ckeye salmon at the Slough 8
modelling Slte •••••• ~............................... 33
Weighted usable area projections for
juvenile chum salmon at the Slough 8
modelling site...................................... 34
Weighted usable area projections for
juvenile coho and sockeye salmon at the
Slough 5 modelling site............................. 36
Weighted usable area projections for
juvenile coho and sockeye salmon at the
Slough 6A modelling site............................ 37
Habitat indices for juvenile chinook
salmon.............................................. 40
Habitat indices for juvenile coho salmon............ 41
Habitat indices for juveni 1 e sockeye
salmon.............................................. 43
Habitat indices for juveni~e chum salmon............ 43
v
-
-
-
-
-
-
-
-
-
.....
-
-
LIST OF TABLES
Table
1
2
3
4
Percent cover and cover type categories............. 12
Catch, catch per cell, and delineation of
site and species combinations modelled.............. 15
Correlations between composite weighting
factors and catch transformed by natural
log (X+1) for juvenile coho and chinook
salmon by site and by all sites pooled.............. 38
Chi -square contingency tests of juvenile
sockeye and chum salmon proportional
presence by composite weighting factor
intervals........................................... 38
vi
I""'"
I
I
,....
I
I
LIST OF PLATES
Plate
1
2
3
4
5
6
Title
Aerial photograph of Side Channel lOA
{ RM 132.1), September 1983 •••.•••••••••....••..••.• 5
Aerial photograph of Slough 22 (RM 144.3),
May 1982 ..••....•••.••...••..•....•••••.•.......... 6
Aerial photograph of Whiskers Creek Slough
(RM 101.2), September 1983.......................... 7
Aerial photograph of Slough 8 (RM 113.6),
August 1982 ............................... '0........ 8
Aerial photograph of Slough 6A (RM 112.3),
May 1982 . . • • . . . • • . . • • • . . • • • • • • • • • • • . • • • • • • • . . . • . . • • 9
Aerial photograph of Slough 5 (RM 107.6),
September 1983 ...............•..............•...... 10
vii
LIST OF APPENDIX TABLES
Appendix Table Title Page
A-1 Weighted usable area and habitat indices for
Side Channel lOA.................................. 49
A-2 Weighted usable area and habitat indices for
Slough 22 ••••••••••••••••••••••••••••••••••••••••• 50
A-3 Weighted usable area and habitat indices for
Whiskers Creek Slough .....•.••...••......•........ 50
A-4 Weighted usable area and habitat indices for
Slough 8 ••.•.••...•••••.•..••••..••..••••••...•.•• 51 -A-5 Weighted usable area and habitat indices for
Slough 5 ........................................... . 51
A-6 Weighted usable area and habitat indices for
Slough 6A .......................................... ... 51
viii
-
1.0 INTRODUCTION
Five species of Pacific salmon spawn in the Susitna River between the
Chulitna River confluence and Devil Canyon. This reach of river
provides rearing habitat for chinook, coho, sockeye, and chum salmon
during the juvenile portion of their life cycle. Pink salmon outmigrate
immediately after emergence. The proposed hydroelectric project on the
Susitna River will create turbidity, temperature, discharge, and other
physical-chemical conditions which are substantially different from
preproject conditions {Acres, 1982). This is one of three interrelated
studies attempting to determine the effects of lowered flows on the
capability of this reach of the Susitna River to support juvenile salmon
rearing during the ice-free season.
Studies during 1981 and 1982 (ADF&G 1981; 1983a) demonstrated large
scale distribution and habitat utilization patterns of these species.
Other studies (ADF&G 1983b, appendices E, F and G) investigated the
response of selected macrohabitat areas to mainstem discharge using
11 hydraul ic zones 11 to characterize sections of the slough and tributary
mouth areas. The surface area of these zones, as a function of mainstem
discharge, were compared to the relative use of the zones by each
species. The result of the analysis was an index of habitat availabil-
ity for each species as a function of mainstem discharge. During the
course of that study we noticed that microhabitat parameters within the
zones were responding to discharge changes at rates higher than the zone
surface areas being evaluated. These microhabitat factors included
cover and turbidity.
The present study incorporates these microhabitat parameters into
simulations of mainstem Susitna River discharge effects on juvenile
salmon rearing habitat. Our experimental design emphasizes the measure-
ment of cover at sites that are characteristic of the macrohabitats
utilized by juvenile salmon. Otherwise, the methodology is similar to,
but less data intensive than Instream Flow Group (IFG) hydraulic methods
{Bovee 1982) of calculating the amount of optimum habitat called weight-
ed usable area. Each site/discharge description is developed from
parameters measured in shoreline and mid-channel area cells specified by
a fixed sampling grid. Our experimental design evolved because it
enabled us to develop models at several sites encompassing the ful 1
range of macrohabitat types. The intensive effort required to develop
IFG models would have limited the number of sampling sites.
Concurrent with the collection of habitat modelling data, fisheries data
were collected at less rigidly specified grids at 29 additional sites.
The two data bases were used to develop estimates of: 1) abundance of
cover type and percent cover, turbidity, velocity and depth versus
mainstem discharge at the six sites, and 2) univariate suitability
functions for velocity, depth, cover type, and percent cover for sampl-
ing cells at all sites. The suitability function study is reported
separately {Part 3 of this volume). In this report, the environmental
descriptions are combined with the suitability functions to yield
weighted usable rearing areas for the species as a function of mainstem
discharge at the six sites. The weighted usable areas for each species,
site, and mainstem discharge were then divided by the surface area of
- 1 -
the site at a typical midsummer mainstem discharge of 23,000 cfs to
produce habitat indices. The index values are plotted as a function of
mainstem discharge by species so that the weighted usable areas can be
c.ompared independently of each site • s surface area at a fixed mains tern
flow.
The results of these calculations have application to two concurrent
projects. The results from juvenile habitat simulation studies using
IFG hydraulic models {Part 7 of this volume) will be integrated with
those presented here to produce best estimates of habitat indices for
the juvenile salmon species at the macrohabitat types identified in the
Susi tna River reach between the Chulitna River confluence and Devi 1
Canyon.
Secondly, incremental estimates of total usable rearing area in the
Chulitna River confluence to Devil Canyon reach impacted by mainstem
flows will be made from the product of the integrated indices and
macrohabitat abundance as a function of mainstem Susitna River dis-
charge. To accomplish this, the area of each macrohabitat type is being
mapped from aerial photographs taken at different mainstem flows. The
total area of each macrohabitat type in the reach as a function of
mainstem discharge will be provided by E. Woody Trihey and Associates.
- 2 -
-
-
-
-
....
-
.....
-
-
-
....
'
2.0 METHODS
2.1 Field Sampling Design
2.1.1 Study site location and selection criteria
Much of the juvenile salmon studies program has been directed towards
collection of CPUE data over widely ranging spatial and temporal habi-
tats of the species (ADF&G, 1982; 1983c). A product of these studies
has been the identification of critical juvenile rearing 11 macrohabitat 11
types affected to varying degrees by variation in mainstem flow. These
areas of the riverine environment, depending on the mainstem stage, are
characterized as side channels, side sloughs or upland sloughs. For
this study, six study sites representative of these three macrohabi tat
types were chosen to complement the IFG hydraulic modelling sites. All
these macrohabitats are affected by mainstem stage and flow and contain
significant numbers of rearing juvenile salmon. Side Channel lOA was
chosen because it possessed potential habitat for rearing juvenile
chinook salmon and represented side channel macrohabitats strongly
affected by mainstem discharge. Two upland slough sites, Slough 5 and
Slough 6A, were chosen because juvenile sockeye salmon rear in these
areas and because they are representative of sites that do not have
mainstem discharge passing through; the predominant influence of the
mainstem on these sites is the backwater created by mainstem stage at
the mouth of the site. Three sites, Slough 8, Slough 22, and Whiskers
Slough, which progressed from side sloughs to side channels at high
mainstem flows, were also modelled (Figure 1). A side slough is
considered a side channel when turbid mainstem water flows through
(overtops) the head of the site. These six sites represented a cross
section of three morphological habitat types present in this reach which
are known to support significant rearing of juvenile salmon.
2.1.2 Sampling grid design
Habitat data at the modelling sites was collected at a grid of fixed
transect markers. The 1 ocati ons of the transects at each site are
illustrated on aerial photographs in Plates 1 through 6. The grids at
each site were placed to maintain a relatively uniform water chemistry
condition and to maximize the diversity of cover, depth, and velocity
parameters to be sampled in the area.
The ei!]ht or nine pairs of the transect markers spanning the selected
reach {typically 1,000 ft) of the site were installed during the first
visit to the sampling site. The location of up to three cells {6 ft by
50 ft) per transect were specified for each subsequent sampling.
(Figure 2). Two shoreline and one mid-channel area cells were always
specified if the wetted area at the transect crossing was 18 or more
ft in width. When the site was between 12 and 18 ft in width, two
shoreline cells were specified; for widths under 12 ft, one shoreline
cell was specified .
- 3 -
Figure 1.
e SAM P Ll N G S I T E
SITES RM
A Slough 22 144.3
B Side Channel lOA 132.1
c Slough 8 113.6
D Slough 6A I I 2.3
E SloughS 107.6
F Whiskers Slough I 0 1.2
River mile and relative location of the juvenile salmon
rearing habitat model study sites.
-4-
-
"""'
I
(J1
I
l l 1 ... 1 1 ... 1 1
Plate 1. Aeria·l photograph of Side Channt~·l lOA (R.M 132.1)" September 1983. The pool betw(~en transE!cts 1-5
and t .. hP ·isli'lnd l,,,,.r,::il''; l'"·}iCli.ldF~d fv·'nl'i'l' t·h,p. (::·t·11•rl·'·' ;,,,,...,,,.,
-
N .....
0::: a;
,......;
»
It :;:::
---M
.......
;j. -,.....;
::E:
0:::
N
N
..s::
C"J
=::;
0 ....-
</)
4-
0
..s:: ~ c.. n::s
~
C"J
0
-+-'
0 ..s::
0..
rr:l .....
~ Pi!FI Q) ...:
N
<J)
-+-'
n::l ,....
0.. -
-6--
I
t
I
I
I
. --....j
I
.. 1 I 1 1
Plate 3. Aerial photograph of Whiskers Creek Slough (RM 101.2), September 1983.
-8-
cc
..c:
CT>
~ ..-
(./}
'+-
0
..c:
0...
~
c::n
0
-!-'
0 ..c: c..
"""''
-
-
-
-
-
-
-
-
-
-
-
N co en ->,
<'0
::E
........
('"'}
C""....l ,_, -::.E: c::
c:r:
\.0
.c
t:n
;:,
0 .--
U')
I;-
0
.c
0.
<'0 s-
O'l
0
+-'
0 r-
.c
0.
<'0 .,_
s....
Q) c:r:
-LO
Q)
+> m
0.... --
-9-
-----------------,~~~---------------------------
.~
~.
(Y'j
cc o-, .......
~
(1,) ~
..0
E
<lJ .....,
0..
Q.l
(/)
1,0
i"'--~ 0 .......
:;;: c:::
.~
L{)
.J::
O"l :;:,
0 ...... """!
V')
'+-
0
.J::
0. ""''' I'd
~
C'l
0
+-'
0
.J:: ~. c.
I'd .,...
s-
Q) -· <::<:
!.0
a; ~I
+-' co
r-c..
-10--
-
"""
-
,.-
LEFT
BANK
TRANSECT 7
TRANSECT 6
TRANSECT 5
TRANSECT 4
TRANSECT 3
TRANSECT I
~6K50 Foot
Cell Unit-
Area Sampled
RIGHT
BANK
c -
0::
<!)
-.. ..
"-
0
0
~
"" c;
-~
Q. ,... ...
~Figure 2. Illustration of the grid and cell sampling scheme employed
at habitat modelling study sites.
--11-
Characterization of the physical parameters of each site by the cell
measurements was made over as wide a ranqe of mainstem discharqe as was
practically possible. Relative water surface elevations at each study
site were recorded from staff gages at each sampling. Mainstem Susitna
River discharges for each sampling \vere taken from USGS provisional
records of flows measured at the Gold Creek gaging station, 15292000.
2.1.3 Cell measurements
Eight or nine mid-channel cells and 16 to 18 shoreline cells were
created by the grid of transects established at each site. During each
sampling, average depth, and mean water column velocity was measured in
each ce 11 and tot a 1 percent cover, and the dominant cover type was
estimated. In nearly all cases, cells in a grid were assigned a common
water chemistry measurement of temperature, turbidity, pH, dissolved
oxygen, and conductivity. If obvious water quality differences existed
across the grid, two or more groupings of the cells .,.,ere made by water
quality parameters. In one case (Slough 8), two grids of transects were
used to sample regions having similar water quality but very different
morphological characteristics.
The mean depth of a cell \'Jas estimated from several measurements taken
with a graduated wading rod midway along the length of the cell. Cell
velocity was determined using a Price Model AA velocity meter at one to
three characteristic mid-cell locations. The total percent of object
cover available to juvenile fish was visually estimated, as was the
primary object cover type. Nine cover types and six categor·ies of
percent cover (Table 1} were developed. Prior to the sampling season, a
field trip was made to promote consistent ratings among the four raters.
Percent cover in this study is defined as the ratio of horizontal or
obliquely viewed conc~aling, hiding or protecting area potentially
available to a {30-100 mm) juvenile fish, relative to the surface area
of the cell. To reduce variances introduced by raters, rating cat-
egories were kept broad and the training introduced common concepts of
how to rate percent cover. The percent cover rating is thus an estimate
of the square feet of cover per cell (300 ft 2 ).
Table 1. Percent cover and cover type categories
Percent Cover
0-5%
6-25%
26-50%
51-75%
76-95%
96-100%
Cover Type
No object cover
Emergent vegetation
Aquatic vegetation
Debris/deadfall
Overhanging riparian vegetation
Undercut banks
Gravel 1" to 3" (in diameter)
Rubble 3" to 5"
Cobble or boulders> 5"
-12 -
-
-
-
-
-
Water temperature, pH, dissolved oxygen, and conductivity were measured
at mid-site with a Hydrolab model 4001 multiparameter meter. Poly-
propylene bottles stored grab samples for turbidity measurements using
an HF model DRT-15 turbidometer calibrated over a 0 to 200 NTU range.
The procedures and techniques used to collect the fisheries data have
been described in detail in ADF&G (1984) and also are summarized in Part
2 of this report.
2.2 Data Analysis
An overview of the data analysis performed in this study is shown in
Figure 3. Field procedures and recording forms specified in ADF&G
(1984) were used throughout. The field data were initially input to a
mainframe computer data base management system and reformatted for ex-
amination.
Following completion of the field season, the catch per unit effort data
for the juvenile salmon species at the six model sites were examined to
determine which sites should be integrated with the species suitability
data for weighted usable area (WUA) projections (Table 2). All sites
with species catches greater than mean catch per cell for all six sites
combined were selected for modelling. Mean catch at these six sites was
very similar to mean catch at all sites sampled during 1983 even though
very high mean catches were recorded at tributary sites. Slough 5 was
modelled for coho and sockeye as these two species were most abundant at
this site. Whiskers Creek Slough, Slough 8, and Slough 22 were modelled
during both their side slough and side channel states (clear and turbid
conditions).
2.2.1 Surface areas
Surface areas were calculated from the distance between each transect
bench marker and the wetted edge of the water measured dut'ing each field
sampling (during one visit to each habitat site the distances and
compass bearings between transect bench markers were measured). These
data were input to a computer program which calculated the wetted
surface area of the study site on each occasion. The .,mid-channel" area
present bebJeen six feet wide 11 Shoreline area" strips was also calculat-
ed and by subtracting this area from the total surface area for each
sampling, the wetted shoreline area was computed.
Tot a 1 surface areas of each of the study sites for rna in stem flows
outside the range of conditions observed during the 1983 open water
season were estimated using a variety of techniques. The methods used
at each study site are noted on figures presented in the results sec-
tion. Since a wide range of mainstem discharges was desired for the
incremental analysis (6)000 to 45,000 cfs), an extrapolation of the
measured surface area curve shapes based on a knowledge of general study
site morphology was required in some cases. Surface area projections at
high mainstem flows were not made for the Slough 8 site.
-13 -
' I-' .p.
'
SUITABILITY CRITERIA CALCULATE COMPOSITE
FOR COVER a WEIGHTING FACTORS
CORRELATION TEST OF MODEL
VELOCITY BY SPECIES ~ t ANALYSIS
BY CELL
( FROM PART 3) ~
JUVENILE SALMON
CATCH DATA
~
SITE WEIGHTED USABLE HABITAT DATA WEIGHTING FACTOR RJ HAB
(COVER a VELOCITY RJHAB AREA CALCULATIONS ~
MEASUREMENTS) MODEL CALCULATIONS BY MODEL BY DISCHARGE LEVEL
i DISCHARGE LEVEL
HABITAT INDICES
GOLD CRE~K -BY SITE 6 SPECIES
MAIN STEM -AS FUNCTION OF
DISCHARGE DATA MAINSTEM DISCHARGE
TRANSECT ANGLE CALCULATIONS OF
6 WETTED EDGE COMPUTER SITE AREA BY
MEASUREMENT DATA PROGRAM DISCHARGE LEVEL
Figure 3. Data analysis flow chart for juvenile salmon rearing habitat models.
I J J .I
-
-
-
-
Table 2. Catch, catch per cell, and delineation of site and species combina-
tions modelled.
No of cells Catch (catch ~er cell)
fished
Site (effort) Chinook Coho Socke~e Chum
Whiskers Creek Slough 67 260(3.9)Y 291(4.3) 24{0.4) 5(0.1)
Slough 5 50 20(0.4) 88{1.8) 27(0.5) 0(0.0)
Slough 6A 77 108(1. 4) 286{3.ij 169{2.2) 11(0.1)
Slough 8 72 65(0.9) 198{2.8} 131(1.8} 73{ 1.0)
Side Channel lOA 64 406(6.3) 0(0.0) 1(0.0) 0(0.0)
Slough 22 52 260{S.Oi 5(0.1) 0(0.0) 1(0.0)
Total (model sites) 382 1119{2.9) 868(2.3) 352(0.9) 90(0.2
Total of all cells .
sampled during 198#/
1260 4395{3.5) 2020( 1. 6) 1006(0.8) 1157{0.9
Yif underlined, the species response to mainstem discharge was modelled
at the site.
b/Taken from Part 3 of this report.
-15 -
2.2.2 Resident Juvenile Habitat (RJHAB) Model
The Resident Juvenile Habitat (RJHAB) model presented here is a sim-
plified method for calculating weighted usable area (WUA) without using
hydraulic models. Our method divides the modelling site into shoreline
and mid-channel areas and then calculates a WUA for both of these
portions of the site. The site WUA is the sum of the shore 1 i ne and
mid-channel WUA. The WUA for a shoreline or mid-channel portion of the
site (i) having area {A) at Susitna River discharge (q) for rearing
species (s) is calculated as follows:
(1) WUAi,s,q = WF(c) x WF(v) x WF{d) x A
The weighting factors WF(c), WF(v) and WF{d} are shoreline or mid-
channel overall suitability values for cover (both amount and type
integrated), velocity, and depth for any given i, s, and q. The depth
weighting factor was set to 1.0 because data from part 3 of this report
indicated it had 1 ittle effect on fish distribution in comparison to
velocity and cover. Examples of the calculations required to obtain the
weighting factors for cover and velocity are described in text and
equations 2 and 3 below. The factors i, s, and q are held constant in
the following equations.
The weighting factor for cover (WF(c)) can be calculated in the form:
(2) WF(c) =
Where:
S{a)j
S(t)j
c.
J
n
n
L: (CJ. x S(a)J. x S(t)J.)
j=l
n
L CJ.
j=l
=
=
=
=
Value of the habitat suitability function
for value of percent cover in cell #j.
Value of the habitat suitability function
for measured value of cover type t in cell
#j
surface area of cell #j.
number of cells sampled in either shore-
line or mid-channel portions
Since there were nine cover types (t) and five present cover categories
{a), a total of 45 percent cover by cover type combinations were possi-
ble.
The weighting factor for velocity was calculated by expressing the
velocity data as proportiona1 frequencies of occurrence after measured
values were grouped into 0.3 ft/sec categories (intervals) with 0.0
-16 -
-
-
-
-
-1
-
,....-
-
-
-
rema1n1ng a unique data point. The weighting factor for velocity
(WF{v)) is_ calculated as follows:
{4) WF(v) =
Where:
m
k
pk =
r =
n =
m 2: (Pk x S(v)k)
k=l
= number of velocity categories
= velocity category code
value of the habitat suitability function
for velocity in interval k
r n
'[ cl l l: cj = proportion of cells within
1=1 j=l velocity interval k
number of ce 11 s in velocity interval k
number of cells in either shoreline or mid-channel
portions
These computations were carried out on a microcomputer using commercial
spreadsheet software. The calculated weighting factors ~JF(c) and HF(v)
were output as graphs for each site and species for both shoreline and
mid-channel areas of the site as a function of mainstem discharge. FOi~
chinook salmon juveniles, the weighting factors were also plotted fm~
both low and high turbidity mainstem conditions. These plots were
interpreted with respect to the changing environmental conditions and
data scatter and a line was fit to the data by hand. This interpreta-
tion required that the frequency distribution of each attribute's values
(in the shoreline and mid-channel areas of the site), at each discharge
be viewed with respect to the suitability curve for the attribute. The
analysis of the weighting factor plots enabled some conclusions to be
drawn from the data which were not obvious from the plots. Following
slough breaching for example, chinook salmon mid-channel area velocity
weighting factors at two similar discharges may have been about the same
value. The two velocity frequency distributions, however, occasionally
had median points falling on opposite sides of the peak in the velocity
suitability function plot; hence, the implication of peak suitability
between the two points and falling suitability (with increasing veloc-
ities) after. Similarly, the slight displacement of maximum suitabil-
ities for high and low turbidity chinook salmon velocity values
occasionally inferred refinements between the plots. For example, a
downwards trend of the weighting factors (with increasing discharge) in
a low turbidity plot could be used to project the slope of a downwards
-17 -
trend in the high turbidity plot at higher velocities where no data were
available. Based on the shape of the composite weighting factor (f(c) x
f(v)) plot, WUA curves were drawn to fit the data.
Weighting factors for flows well beyond those observed were estimated
from the trends occurring in the cover and ve 1 oci ty data and from the
shape of the suitability criteria curves. Accumulated field experience
at ·the site'" and comparisons to other sites where similar conditions
existed were additional criteria used to make the projections. The
velocity weighting factors extrapolated for side channels at high
mainstem discharges are the most uncertain of these projections.
The last step in the data analysis was to calculate "habitat indices~~
for the species. Habitat indices were calculated as the WUA divided by
the surface area present in the study site sampling grid at a mainstem
discharge of 23,000 cfs. The 23,000 cfs figure was chosen because it is
a representative summer streamflow and it also may be integrated 1tlith
macrohabitat abundance information provided by E. Woody Trihey and
Associates from aerial photographs of the upper Susitna reach at this
discharge.
The individual cell measurements and weighting factor plots are not
presented in this report. Bound volumes of the data can be obtained for
inspection at the Susitna Hydro Aquatic Studies office.
During the analytical process the data base was screened for errors and
inconsistencies. Some data collected at closely related mainstem flows
were averaged to eliminate scatter not related to mainstem discharges.
The largest single change made to the raw data was to substitute a
representative mean cell cover value for the individual (instantaneous)
mid-channel cell readings. The desirability for this change arose
because of the considerable difficulty with consistently determining
substrate cover values in deep'" rapid or turbid water mid-channel cell
areas. Roughly 750 habitat cells were characterized for the analysis.
Several field observations were changed because we believed they were
recorded erroneously.
2.2.3 Model verification
Data on fish abundance and distdbution were collected at the sites to
validate WUA projections. However, time constraints prevented an
intensive sampling effort. A composite weighting factor was calculated
for each cell sampled for fish and this factor was correlated with fish
catch in the cell. If cells with high composite weighting factors are
associated with higher densities of fish as expressed in the catch, then
it can be assumed that if changes in mainstem discharge raise or lower
an entire site•s composite weighting factor, the associated potential
for fish use will also be raised or lowered.
In order to test for a relationship between cell composite weighting
factors and fish catch, the following procedures were carried out. The
composite weighting factor in each cell was calculated by multiplying
suitability values for cover and velocity together. Coho and chinook
catches were transformed by natural log (X+l) in an attempt to normalize
-18 -
-
....
-
-
-
variances. Pearson correlation coefficients were then calculated
between the composite weighting factor and coho and chinook catch by
cell. For chum and sockeye, chi-square contingency tables \'Jere run
between proportional presence and composite weighting factor value
intervals (to test for associations between these two factors). Sampl-
ing occasions when less than three fish were captured in all the cells
within a site (in a day of sampling) were deleted from analysis. This
was done because low densities of fish are often due to seasonal
movements rather than to within site habitat conditions. If fish
sampling data from sites without fish were used in a correlation
analysis, the correlations might become statistically insignificant even
if the correlations between composite weighting factor and fish catch
were large.
-19 -
3.0 RESULTS
3.1 Surface Areas
The total wetted surface areas at each site are plotted as a function of
mainstem Susitna River discharge on Figures 4 through 10. These figures
also contain schematic notes concerning important changes which occurred
over the range of flows which were observed. The range of surface areas
calculated from observational data are highlighted with solid·lines.
Extrapolated data are noted with dotted lines.
The total weighted usable areas {WUA's) calculated for the species at
sites where fisheries data support projecting habitat use are presented
in Section 3.2 through 3. 7. The total weighted usable areas projected
for each site and species at mainstem discharge inct"'ements of 3,000 cfs
are also tabulated in Appendix A.
3.2 Side Channel lOA
Chinook salmon were the only juvenile species captured in abundance at
this site. Because suitability functions for cover, velocity, and depth
at turbidities above and below 30 NTU were different for this species of
juvenile salmon, \~UA projections for high and low turbidity mainstem
flows are calculated (Figure 11). All WUA units are in square feet.
The solid line labelled 11 calibrated range 11 in the WUA plots is the
estimated WUA at observed flows. The dotted line labelled 11 extrapolated
rangeu is the extrapolated l~UA at flows which were not observed during
the open water season of 1983. The total weighted usable area in each
plot is the sum of the WUA's calculated for the shoreline and mid-
channel areas of the study site. At any mainstem discharge, the WUA for
the shoreline or mid-channel area is the product of the weighting
factors WF(c) and WF(v) and the surface area for the shoreline or
mid-channel area at that mainstem discharge. The weighting factor plots
calculated for this species and site under high and low turbidity
mainstem flow conditions are included here (Figures 12 and 13) as an
example. Weighting factor plots for the other sites are available at
the Su Hydro Aquatic Studies office.
The difference between the WUA • s projected for high and low turbidity
conditions reflects the differences in suitability for the cover and
velocity values measured at the study site over the range of observed
and extrapolated mainstem flows. Especially noticeable are the effects
of suitability for cover: under the low turbidity condition the weight-
ing factors (and thus the WUA's) are greatly reduced. Similarly, the
difference in the shape of the velocity weighting factor curves for the
two turbidity conditions explains much of the differences between the
shapes of the two plots.
3.3 Slough 22
Chinook salmon were the only juvenile species captured in abundance at
this site. Weighted usable area projections for juvenile chinook salmon
were calculated for both high and low mainstem turbidities (Figure 14).
At mainstem flows above 20,200 cfs, the head of this slough is
-20 -
-
-
.....
-
-
-
-'I
I
....
0
~
I(
t:;
1&1
IL
1-IIJ
"" 0::
"" <
!'::) 1&1 a 0:: (I) < ....
IIJ <( u Q ~ a: ..J
!'::) 1&1 I
N (I) z ,_. 2:
I < ::z: u
1&1
0 a;
I
l
· SIDE CHANNEL lOA
SURFACE AREA IN STUDY GRID -1!1
140--_, -----ItO· ~ /' 100-
/' so-
• OBSERVED
Gl ESTIMATED
60-• I Surface area at high flows as-40-• timated from aerial photograph
I
I
20-I
I
I
01--~~-4~1-6r-1~81,-ll--~l-~l-r-l~lr-~T--~T--~~---Tr--Tr--IT--~1 _---1T-_-r-1 -r~--~1 _--T_1 -
0 2. I 0 12. 14 16 18 2.0 2.2. 24 2.6 2.8 30 32. 34 36 38 40 42. 44
SUSITNA R. DISCHARGE (CFS x IOOOIAT GOLD CREEK
USGS PROVISIONAL DATA 1983 1&292000
Figure 4. Total wetted surface areas measured and
extrapolated in the Side Channel lOA habitat model
study site.
0 a .. ...... <Cbl ..... ......
"'"' «<a: ... ~ uo
~Ill ... _ .... ::. ...
"':z:
0
:l
0 .J
"'
0
0
2 .. ... ... ...
1-....
~ .....
0:
"""" w::>
o:O .. ~
"':z: u.,
• ::> ... 0
"' .J ::>,. .,
"' a: ..
"" ~ :z:
it<
110
U)O
90
eo
70
&0
50
0
70
60
50
40
30
zo
1.0~
0
0 2 ~
SLOUGH 22
SURFACE AREA IN STUDY GRI 0
~ GROUND1UTER
Af low mainJ.Iem. flowa...tiG!II-QA
WSE :b.-com•• -s.tattt• beca.use
of berm-Gt do..-gh rnoutl'l.
• DB SERVED
ElESTUIATED
SutfQc e oreos o1 r.iqft Uows es-
timoted from ouial ph-o1o: on-
umin q s.cou.nd ana re pra·u nh
hiqb ·flow 'IIIIa fer tint:.
a 10 IZ 1• 16 18 20 Z2 U t6 28 30 31 34 36 38 40 ~1 44
SUSITNA R. DISCHARGE ICFS" 1000) AT GOLD CREEK
USGS PROVISIOKAL DATA 1983 IOZ8ZOOO
Figure 5. Total wetted surface areas measured and
extrapolated in the Slough 22 habitat model study
sites.
WHISKERS CREEK SLOUGH
SURFACE AREA STUDY GRID
C~EAR II~~ ::~~:ATER
I~~ JHC:R£ASING-
0ROUNDWATER _-MAINST£11
/ BR .ED ........... CHARACTER
WSE AT LOWER TRANSECTS -_.e.-
UP FROII RAIN SWOL~EN -• --
!il
/
WHISkERS CREEK ;•
_ - _ .:-:' ________ • -WSE I$ R£L.O.TIV£LY ST.O.BLE BELOW
22.000 CFS 8£CAUS£ OF lEA >I HEAR
TRAHS£CT 2
Svrte~ce areas. at high flow-. •••imaltd
from oerio-1 ;hoto-ond tram -u:fro_poto-
1iofl of mid ltou-qh. wSE with a"plica•
fiofl1o mid -.Jo-vgJ\ c:to.ss-teet ton profile
• OSSEI<VED
Iii £$TIMAT£D
0 .. & 10 IZ I~ 16 18 20 ZZ 2~ 16 U lO J2 l~ 36 l8 40 42 ~~
SUSITNA R. DISCH ... RGE (CFS aiOOOl AT GOLD CREEK
USGS PROVISIONAL DATA 1983 15292000
Figure 6. Total wetted surface areas measured and
extrapolated in the Whiskers Creek Slough habitat
model study site.
-22-
I
I
-I
-
-1
-
~
" t-... ... .. ... -.. ... ... "' "' c .. .. ... "' .., 0 .. .. ...
"' ~
"' :a:: .,
0
"' 0
..J ..
0
0
!2
M ... ... ... .. ... ... ...
"' ... "' 0: .. <( -... "' 0 <.> ~ <( ... "' a:
"' :z: ., 0 :::>
0
..J .,
-
10
60
50
40
30
20
IQ-
0
SLOUGH 8 GRID I
SURFACE; AREA IN STUDY GRID
/
INCREASING
MAJNST£M
CHARACTER
/
BREACH EO
CLEAR~ TURBID I .. :~ .. :-<>
\"f"" \l
'1 '£,... • e~c,...," /"
~ ~ •OBSERVED
E..'li--\,.\~£. ••
cP.f. . ------
-__ •--•~ Water b61JOn overtoppinq banlris
01 32:00 ~h.
0 ~ 8 10 ~~ 14 16 18 20 2Z 24 26 Z8 30 32 34 36 38 40 42 44
16
, ..
1:!
.a
II
10
9
6
5
..
3
SUSlTNA R. DISCHARGE ( CFS •1000) AT GOLD CREEK
USGS PROVISIONAL DATA 1983 1529:ZDOO
Figure 7. Total wetted surface area~ measured and
extrapolated in the Slough 8 grid 1 habitat model
study site.
0
SLOUGH 8 GRID 2
SURFACE AREA IN STUDY GRID
_ Tlte rapid ris-e in
1ur-ta.ce ar•a gt flows above 32,000
c:f1,_ indicates walu a11ertoppino banks.
/
INCREASING
MAIMST[M
CHARACTER
CLEAR Ill TURBID
RUNOFF AND ~~~):: ACKWATER
GROUNDWATER ~ •
----riJ. • LIKE !!REACHED
____ ... _., __ ....;;.. _____ ..
8 10 12 14 16 18 20 22 24 28 Z8 30 u 34 36
SUSITNA R.DISCHARGE (CFSaiOOOI AT GOLD CREEK
USGS PROVISIONAL OAT A 19113 152.9 2000
• OBSERVED
9 ESTIWATEO
Figure B. Total wetted surface areas measured and
extrapolated in the Slough 8 grid 2 habitat model
study site.
-23-
110
100
§ 90 .. ..... oe..,
o(W ....... 10 a::..,
oCa: ..,oe ..,::. oea 70 ... !!
"'-c ~ ..
%
"' 60 ::> 0
...l ..
50
~
140
IZO
a
0
!2 100 ... ,_ ...
<I; ... ...
"' ... ao
<( ... "" "' <(
<( ::> ... a
u ., •o : ... tr :1: ::>
"' "' ::1 40 s
"'
20
0 z
SLOUGH 6A
SURFACE AREA IN STUDY GRID
-------~·--·--
------
------·-·
Su<loet oroo• at high ond low llo>ws
•atrapofoted from mtaaured tr•nd.
• 088£RVEO
C ~ 10 IZ 14 IS II ZO 22 24 26 28 ~0 ~z 34 ~~ 38 40 4Z 44
SUSITHA R. DISCHARGE (CFS 11000) AT GOLD CREEK
USGS PROVISIONAL DATA 1983 15292000
Figure 9. Total wetted surface areas measured and
extrapolated in the Slough 6A habitat model study
site.
0
SLOUGH 5
SURFACE AREA IN STUDY GRIO ------B
• OBSERVEQ
l!IESTIMATEO
l
/
I
./
. .
-,..-
Surf.aca <1na at hiQh f1ows •IIi"'
timohd from a11ial pnoto co.n·
sidorin; I hOI slough bonk• Md
nearly con lei nod wollr (width)
at ZB,OOOds
R41N SWOLLEN/REEK /"
IN UrPER TRANSECTS
__ ____.
• 6 10 12 14 16 18 20 ZZ 24 26 28 30 32 H 36 38 40 42 4<
SUSITNA R. DISCHARGE (CFS x 1000) AT GOLD CREEl<
USGS PROVISIONAL DATA 1983 15292000
Figure 10. Total wetted surface areas measured and
extrapolated in the Slough 5 habitat model study
site.
-24-
---~ ·----------------------
-
-
-
-
,.-
30
28
26
-24
0 22
20] < w a: 18 I
< 16 .J uJ I -'
14 i "' < .. 12i :J
Q tO, w .... j
:r B-j
!::! 6~ w -~ i
4
2
0
5 -
12
11
10 -
0 9 -8
< w a: 7 <
w 6 ....
0::
<
"' 5 ::3
0 4 w ....
J:
S1 3
w ;: 2
0
5
-
CHINOOK WEIGHTED USABLE AREA
SIDE CHANNEl. lOA
TURBID MAINSTEM
I
I
I
I WElGWTEO U.SA8-LIE AA(A
It~ 0 .... , .. u ...
I I +
+ NI••CII-•"""1
0 Tohl
I I -Colfttnt-H Raaqt
I 1 + --:Est.-a"•ot•CI ROI'Q•
:; '\ I I
l I ,,
/ I ' ........
II ............ _
I I --I}"' --tl ,---------::-::::::-
' 1: .... ...... _ --'• --..._... ____
15 25 35
SUSITNA R. DISCHARGE (CFS • 1000) AT GO\.D CREEK
USGS PROVISIONAL DATA 1983 15292000
CHINOOK WEIGHTED USABLE AREA
SIDE CHANNEl. lOA
I
I /+ /,+-/. ~
:: ., I I
I :
I I
I 1
I I
I l
I I ,
W£10.HTEO liSA8LE AR(.A
a lifll.oren••
+ Nld•C.tr.••"•'
0 Tot•l
-CoHitroted Ro•t;•
--Ealrat~~-alattd-RoftQe
I: +, ', ., '~',,
45
.. .... ...................... __ _
,,_ --~-... -----
tl ------~!' .....
ll
~ --------------
15 25 35
SUSITNA R. DISCHARGE (CFS • 1000) AT GOLD CREEK
USGS PROVISIONAl. DATA 1983 15292000
45
Figure 11. Weighted usable area projections for juvenile
chinook salmon at the Side Channel lOA modelling
site.
-25-
CHINOOK
MID-CHANNEL AREA WEIGHTING FACTORS
SIDE CHANNEL lOA 1.0,---------------------~~------------------------------~ TURBID MAINSTEM
0.9
~ :t----:.:-,;~ '
g o.s~ '-....,, \-_..__----a--~ ----oo ---------
0 0.5 ' ::
~ +
J: Q 4 '-IIIII'[,QiotlUtO IACTOA ~ . "+ C Cower ~ '-+ V•ht4ttr
O.J., ~ .0. C.OIIIIPG-•••·
r + -ColiDfOSitd RQftill
j ~ --EsltQJoolah.C Ftan91
::I . ~ "·~~~;o--~
15 25 35
SUSITNA R. DISCHARGE (CFS •1000) AT GOLD CREEK
USGS PROVISIONAL DATA 1963 15292000
CHINOOK
MID-CHANNEL AREA WEIGHTING FACTORS
45
t.O,-------------------~S~IO~E~C~H~A~N_N~E~L~tO~A~--------------------,
t.OW TURBIDITY MAINSTEM
0.9
0.8
Wf.tGHTINQ I'A.CTOA
0 CD"*'
-+ Yet.ocur
.0. C:IIIIIIIP-••H•
-C4liD-r-otw4 Aa•t•
--£•t-rcpotoled A.Gn;e
,----------...... -------------o.o+------.------,------.~-----.----~r------.---~--~--=r~-~~~
5 15 25 35
SUSUNA R. DISCHARGE (CFS • 1000) AT GOlD CREEK
USGS PROVISIONAL DATA 1963 15292000
45
Figure 12. Mid-channel area weighting factors for juvenile
chinook salmon at the Side Channel lOA modelling
s~tP..
-26-
....
-
-
-CHINOOK
SHORELINE AREA WEIGHTING FACTORS
I.O-r----------____;S:::Ic=Dc=£c..C;:.H:..:..;_:A;.:N;.:N;:.E;:.L_t;.:O:..:.A;__ _________ --,
TURBID MAtNSTEM
0.9
0.8
<r 0. 7
0 ---.... ----+--------~ ~---.::----c---------.... u
"' ...
0
!!: ....
0.5 -----
0.61
§ 0.4 ______ ___,._--------
w
~ 0.3l
0.2,
0. I -Colib•tllftd Roa-g-1
--(nrapolau4 .Fto"~'
0.0~5.-----~-----r----~r----,-----~-----.-----~----~
IS 25 35 45
SUSITNA A. DISCHARGE (CFS •tOOO) AT GOLD CREEK
USGS PROVISIONAl DATA 1983 15292000
CHINOOK
SHORELINE AREA WEIGHTING FACTORS 1 .0,------------S;:.I;.:O;.:E~C;.:H;.:A;.:N;.:N~E;:.L_1;.:0:..:.A;__ __________ -,
LOW TURBIDITY MAINSTEM
0.9
0.8
!
a: 0.7, -----·---
0 t .----+-.... ~ 0.6, / + ----~ . ..... ......... , ~ ::: ___ ./ :"':::::G HCTOO .._'
S! ........... . ~ 0.3lj ----4 Co~paail•
-C~litlf'0114 Rongr
j --iE111Q='olottd R:on;r
0.2 _______ ...,____
0.1 ----------------------o.o,_--~r----r---r---~---~--~----,---~ 5 15 25 ---35 45
SUSITNA R. DISCHARGE (CFS • 1000} AT GOLD CREEK
USGS PROVISIONAl DATA 1983 15292000
Figure 13. Shoreline area weighting factors for juvenile
chinook salmon at the Side Channel lOA modelling
site.
-27-
-
0
<
lU c:
<
Ul _,
Ol
<
"' =>
0 w ....
l:
S! w
;:
45
40
35]
::i
20~
I : lSi
I 10!
_J
"I
'
0 !
5
CHINOOK WEIGHTED USABLE AREA
SLOUGH 22
TURBID MAINSTEM
BREACHED T
Wf:IGJ'IlEO fJSA8L.E .lAEA
Q $har•U"'•
+ Mh:I-CII.,nn•l
0 fatal
C.ctl.ib4'0ftd Ranq•
--[111tG-pcrlohd Reno•
It ',,
~~~::~~~~~::k~~~~~~.=~~-1
15 25 35 45
SUSITNA R. DISCHARGE {CFS • 1000) AT GOLD CREEK
USGS PROVISIONAL DATA 1983 15292000
CHINOOK WEIGHTED USABLE AREA
SLOUGH 22
11,------------------------------------------------------------~ LOW TURBIDITY MAINSTEM
9
0 8 WEIG~T£-D USA8-lE ARt.._
0 Shoa-·nlin•
< w a: <
w
.J
T
8
+ wt.a-~Cft:•Aa .. i
-CatjtltQ-U•G RO-"'Qt
--[str~polo.le4 .R-1ln-Q-e ri ~ :!]==============-~==:.-~
2 I +
1 i -~---------------
0 -------------~ -u-
s 15 25 35
__ _§_UStTNA_F!, DtSCHARqE (CFS • __ 1_000) AT GOLD CREEK
USGS PROVISIONAL DATA f983 15292000-
Figure 14. Weighted usable area projections for juvenile
chinook salmon at the Slough 22 modelling site.
-28-
45
....
-
-
-
-
-
-
-
-
--
....
.....
-
overtopped, and in both the low and high turbidity models WUA is affect-
ed by the changing velocity conditions. A greatly increased suitability
for cover at the higher turbidity is again manifest in the projected
WUA's.
3.4 Whiskers Creek Slough
The shapes of the weighted usable area plots projected for chinook
salmon juveniles at this site (Figure 15) are very similar to those for'
the Slough 22 site. The Whiskers Slough site has more cover and hydrau-
lically approaches mainstem conditions at a faster rate following the
breaching event than does the site at Slough 22 •
Weighted usable areas were also projeCted for coho salmon at this site
(Figure 16). Preferences fat~ different turbidity conditions for juve-
nile coho salmon were not demonstrated because of the lack of occurrence
of juvenile coho at turbid sites. The WUA plots for this species do not
reflect use of turbid conditions. Compared to chinook WUA's for the
site, cohos WUA's are roughly 25% smaller under low turbidity slough
conditions, and 50 to 80% smaller during either low or high turbidity
side channel conditions.
3.5 Slough 8
Juvenile coho, sockeye, and chum salmon were captured in abundance at
this site. Seventy-five percent of the chums were captured during the
one sampling in May, so the seasonal mean catch/cell data presented in
Table 2 fer chum salmon are somewhat misleading. Modelling at mainstem
discharges above the calibrated range was dropped for lack of supporting
fisheries data and because projections for surface areas at high
mainstem discharges \'/ere so uncertain that robust predictions for t~UA' s
were impossible.
Weighted usable areas for coho, sockeye, and chum salmon in both study
grids were calculated up to a mainstem discharge of 31,900 cfs (Figures
17 through 19). The shapes of these plots largely reflect velocity
changes as backwater moved into and nearly covered the site before the
head breached. The cover weighting factors however, are responsible for
the very large differences in the WUA's calculated for each species.
Weighted usable areas around 4,400 ft 2 for chum salmon are associated
with mean cover weighting factors of 0.44 and 0.34 for the shoreline and
mid-channel areas of the site,. respectively. Weighted usable areas
around 1,400 ft 2 for sockeye salmon are associated with mean cover
factors of 0.27 and 0.12 for the shoreline and mid-channel areas of the
site. The site is least suitable to coho. WUA's for that species are
around 380 ft 2 with mean cover factors of 0.14 and 0.02 for the shore-
line and mid-channel areas.
3.6 Slough 5
Slough 5 is an upland slough which is not normally connected with the
mainstem Susitna River except at its mouth. Juvenile coho and sockeye
salmon were captured in moderate abundance at this site. At mainstem
-29 -
-
~
0 .
< w
0: <
w
_J
"' <
"' ::l
0
11.1 ,...
J:
!:! w
3:
22
20
18 ~
16
"~ 12
10-j
I I
8-i
J
4
2
0
5
CHINOOK WEIGHTED USABLE AREA
WHISKERS CllEEK SLOUGH
TURBID MAINSTEM
0 sruu•U••
+ M~of-C haBtl-•1
0 To-I•J
-CGiiDI4t•d RC.I'Iq..-
--(:afr(J;:IOh:J1•d ACift'Q't
--...... , ............... -----
15 25 35
SUSITNA R. DISCHARGE (CFS > 1000) AT GOLD CREEK
USGS PROVISIONAL DATA 1983 15292000
CHINOOK WEIGHTED USABLE AREA
WHISKERS CREEK SLOUGH
LOW TURBIDITY MAINSTEM ..,.--..,
..... .....
" BRE ... CHEDf
" ' \
\.
\.
'\
wEt0Hif£D U.SA&l..e AA£11
'\
--
' '\
' \.
5
SUSITNA R. DISCHARGE (CFS • 1000) AT GOlD CREEK
USGS PROVISIONAL-DATA 1983 15292000 .
Figure 15. Weighted usable area projections for juvenile
chinook salmon at the Whiskers Creek Slough
modelling site.
-30-
45
-
-
-
-
-
.....
-
-
-
-
-
r-
!
3.8
..... 3.6
3.4
......
N --.., 3.0 0
X 2.8
<: 2.6
w a: 2.4
<:
w 2.2
..J 2.0 Cl
<:
r/) 1.8
;::) -1.6 0
UJ
1-1.4
:c
""""
(!). 1.2
UJ ;: 1.0
0.8
0.6
0.4 ,-
-
-
COHO WEIGHTED USABLE AREA
WHISKERS CREEK SLOUGH
-......
\
BREACHED y \
\
\ WEIGHTED USABLE AREA \
0 Shoreline \ + Mld-ChanRel
0 Total \
\ -Calibrated RCinQa
--Extrapolated Ronqe \
\ .,..-+...., I+ ,\ \
0 \
\
\ + \
) \ ' ...-----,---....... ' --...' ' -..::::._ ' ' c ' ' ' ' ......
5 15 25 35
SUSITNA R. DISCHARGE (CFS x 1000) AT GOLD CREEK
USGS PROVISIONAL OAT A 1983 15292000
Figure 16. Weighted usable area projections for juvenile coho
sa-lmon at the Whiskers Creek Slough modelling-
site.
-31-
----·-·---·--------
45
0
<(
w a::
<
w
..J
"' <
"' ;::,
0
w ,...
I
" w ;::
-
0
< w cr <
\IJ
-'
"' ...
"' ::;:,
0
w ,...
S1 w ;::
COHO WEIGHTED USABLE AREA
S~OUGH 8 GRID 1 . 700-,------------------------------------------------~----~
BREACHED T 0
0
W£1GHT.£0 US..t.8._E ..t.RE..t.
0 Sftoullhl•
+ IIUd-CI'oenn•r
0 TdoUt
-Calibroh6 Rol'l-;t-
--E•tt'Oj:~OIOI•4 R4-"Q'"I
04------~-------.------~------r------.-------r------~----~
5
sao
I I
I
400!
i I
I
300-l
I 1---
'"1 I
I
100-l
0
5
15 25 35
SUSITNA R. DISCHARGE (CFS • 1000) AT GOLO CREEK
USGS PROVISIONAL DATA 1983 15292000
COHO WEIGHTED USABLE AREA
SLOUGH 8 GRID 2
BREACHED f
a
•••vw 2t.oaa (::1• "o
.IOor•.cl•bt• C••c ,.,,.
,.~ ....... '" '"'• lltU.cfJ'
WEJGN"'€0 U$AI!n.E AAEA
0 .SIIOfWHIII•
+ Mld-CIIo,.nl'lal
Q-_ Ttald
-C~~ttibrohd Ronqt1
--E.111rapololttd Aono•
SUSITNA R. DISCHARGE (CFS >1000) AT GOLD CREEK
USGS PROVIStONA~ DATA 1983 15292000
45
45
Figure 17. Weighted usable area projections for juvenile coho
salmon at the Slough 3 modelling site.
-32-
-
-
-
-
SOCKEYE WEIGHTED USABLE AREA
SLOUGH 8 GRID 1
5.0,---------~--------------------------------------~-----,
-6REACttEO T
4.5
-4.0~
0 I
-3.5 i
" I ~ 3.01
+ Llll.O:tCII•~t"•l
0 Tot4t
-CaHbrat•d Ro~:~nQ!Ii
--E..ltrapc-lafld Ran9.-
()
0
: 2.5J /" ~ I +
< I o
fJ1 2.0 -r ~ !
0 ~ t.si_! __ ~/~
"' 1.0-s: r--·
0.5 -l _.---· j--+-+
o.o41-------.------,-------r------,~-----.------~------~----~
s 1s 25 3s
1,01
0.9,
: 0.8~
o I
• 0.7 i
< I
"' c::
<
ut
..../
"' <
"' ~
0
uJ ...
5
w s:
0.6i
O.Sl
0.4
I
0.3 -j
J
0.2~
i
a 1 i
SUSITNA R. DISCHARGE (CFS x 1 000) AT GOLD CREEK
USGS PROVISIONAL OATA 1983 15292000
SOCKEYE WEIGHTED USABLE AREA
WE:t.QMT€0 0$118lE ARE:.A
0 $t\o••Un4
+ llti<:I-CPI•-~tft•t
0 ,., •.
-Cd.litlfCU•d R'lnq1
--£.-rrCJ:Id.IOt•d" RonQ•
---a
SLOUGH 8 GRJO 2
9AE,._CHEoY
:>.'lila• l2.COC cl& nc
<~<DDt•cl•bl• 4eeg .. , ...
liiiM&In•d In til• ttwdy
ar•• ..
45
~ ..
0.04-----~r------r------.-~~=T~----~-----,------,-----~ ts -25-----45
SUSITNA R. DISCHARGE (CFS • 1000) AT GOLD CREEK
USGS PROVISIONAL DATA 1983 15292000
Figure 18. Weiqhted usable area projections for juvenile
soc~eye salmon at the Slough 8 modelling site.
-33-
---------·-----------------------------------~-------------F-------------------~
I.
j
121
11 l
! : ':l
~ a I
; 71
~ 6-1
(/)
:0
0
w
'"" J:
<;!
"' :;::
5
CHUM WEIGHTED USABLE AREA
WEtGtt,E.O USABLE ARIEA
0 S>iltr•Hn•
+ wld-Ch•""'•'
0 To•••
-i:o:~l~croud ~0110•
--E1tt~Ocolol•d A-oi'I'Q•
SLOUGH a GRID 1
8REACHE"O l
15 25 35
SUSITNA R. DISCHARGE (CFS • 1DOO) AT GO~O CREEK
USGS PROVISIONAl DATA 198J 15292000
CHUM WEIGHTED USABLE AREA
45
SLOUGH B GRID 2 2.0~------------------~~~~~~~------------------~
BREACH EDT
8
+
!5 25 35
SUSITNA R. DISCHARGE (CFS •1000) AT GOLD CREEK
USGS PROVISIONAl DATA 1983 15292000
45
Figure 19,. '..leighted usable area projections for juvenile chum
salmon at the Slough 8 modelling site.
-34-
....
-
-
"""
~i
-
....
discharges under about 15,000 cfs, the majority of Slough S's v1etted
surface is divided between a steep-sided channel at the mouth and a
shallorJ meandering stream, often only a few feet in width. At higher
discharges, rising back\vater progressively floods large areas of the
study site. The increase in WUA for both species, with increasing
mainstem stage, was projected to be lower than the physical measurements
indicated (Figure 20). The downwards adjustment of WUA • s was made to
reflect less than optimal conditions which existed following the initial
flooding event when submerged vegetation was so dense that it restricted
juvenile movements and caused the water to stagnate. Because increasing
water depths improved habitat conditions in the flooded areas, the
weighted usable areas indicated by the physical data at mainstem
discharges around 28,000 cfs were used for the species at 45,000 cfs.
This adjustment is reflected in the projected cover indices at
discharges greater than 25,000 cfs. Relatively lower velocity and cover
weighting factors are responsible for the lower WUA's calculated for
cohos than those calculated for sockeye at this site.
3.7 Slough 6A
Slough 6A is an upland slough with steep banks which prevent large
changes in surface areas from occurring over the range of mainstem
discharges observed. All species of juvenile salmon except pink salmon
were captured at the site, although only coho and sockeye juveniles were
captured in abundance relative to catches at other sites.
Smaller WUA's for both species (Figure 21} at mainstem discharges below
25,000 cfs reflect loss of cover in the shoreline areas of the site.
Differences in the magnitude of the cover and velocity weighting factors
in all areas of the site ue responsible for the much lower overall
suitabilities calculated for coho juveniles.
3.8 Model Verification
Strong positive (i.e., significantly greater than 0.0) correlations
between coho and chinook catch and combined weighting factors by ce11
were found for most sites modelled (Table 3). Correlations between
chinook catch and combined weighting factors in low turbidity waters
ranged from 0.61 to 0.81. In high turbidity water, the correlations
were much lower in absolute value and sometimes not significant by site
at the 0.05 level although the correlation coefficient for the sites
pooled was highly significant. Coho salmon catches were significantly
correlated with combined weighting factors at all sites, and ranged from
0.48 to 0.63.
Sockeye proportional presence was strongly associated with large values
of the combined weighting factor (Table 4). Chum salmon were not
significantly associated with the combined weighting factors but the
sampling effort was very small.
-35 -
COHO WEIGHTED USABLE AREA
s,--------------------------S~l~O~U~G~H-5~------------------------,
...
w a: ...
w ...
"' <
.:J
0
'l
4i
I 31
! ~---o
zJ
,J
I
0
+
......... ._ --
WUA-4bQVIt ;t:S.,OOQ o::l• I•
adl•n·t•d betv• '"dlcaced
••t"~•• C••• t••f for
d'I•.CYIIo•tort)
--0~-------.-------.------,-------,-------.--------.-~-:=~=T~~==~
5 15 25 35
SUSITNA A. DISCHARGE (CFS >< 1DOOl AT GOLD CREEK
USGS PROVISIONAl DATA 1983 15-292000
SOCKEYE WEIGHTED USABLE AREA
SLOUGH 5
45
40,---------------------------------------~------------------;
I H~ D ~
~ I + /.f1/
: 30 1-:-<': Wt::ICifofT£0 U.S AB-U .I.REA O ,/!f'
! : :~:~;~::... ~,.
< 25jl 0 'ht•l ~
w .._ Colle'~"• Ro•o• + 7 '/ ~ 1 --Est,.il"pc-la-tcd R:aR-Q'I f ~ 20 .,.,..."/?
"' --/ : --/ ::> 15-l ,/'
c I o /
W 1' +"'/ wv• ••••• u.oon ch .... // .......... ····-~ ·: ---0 ~D _: __ rSJ::~•,•:~~·.::::::,!~;• ,'
-=..:;.:: .................. _________ _
Or -~~ 1 1 i i
5 15 25 35 45
SUSITNA R. DISCHARGE (CFS • 1000) AT GOLD CREEK
USGS PROVISIONAL DATA 1983 1592000
Figure 20. Weighted usable area projections for juveni1e coho
and sockeye salmon at the Slough 5 modelling site.
-36-
-
-
-
r
COHO WEIGHTED USABLE AREA I
SLOUGH 6A
-
0
<(
w
c:
<(
w _,
"' <(
"' :1
0 w ,_
C!
"' ;::
3.2,,--------------------------------------------------~~--~---.l J.O~ ~---------------
2.8~ /
2.sJ "" ~ ' #'..,'*" ----------
2.4-l ... +-+----
2.2-f -+~ 2.o-L .... -" I
1.8~ WEIGHT!.!J U:SAlHE AREA I
1 0 St>oreliin 1
1.6...4 + Mtd-C"•ron•t I
1.4 J_.• o rout j,l'! -Ci:!libtQI•d Ro.n<;•
--(JIIrqpol.ol•cS ROIIQe
1.21 I
1.0, J
:::~1: ~---------------i
+----~-~~-.----T----.----~---.----.----41 0.4 1 I
5 15 25 35 45
SVSITNA R. DISCHARGE (CFS • 1000) AT GOLD CREEK
USGS PROVISIONAL DATA 1983 15292000
SOCKEYE WEIGHTED USABLE AREA
SLOUGH 6A
28 I ------I 26i ~-----~--!
;;-__ 2 4 ~ _..r'
I ---------~
0 22+------------,i
I +----t .: 201-----+--+---
<( i ~ 181
: 161 WEIGHTED IJS.ABLE ~;qE.-,
~ J 0 $11or•Hn" ~ 1124 ~~~ ! ~:~:~~•nul
:1 ~ -Cal;o<oloo Ra•o•
~ 10 ~ --E '"" ooto u d Ronqo I j
~ :i I I 4 l-----~------------------,~ I
2 I I I I I
5 --15-25 35 45
SUSITNA R. DISCHARGE {CFS •1000) AT GOLD CREEK
USGS PROVISIONAL DATA 1983 15292000
Figure 21. Weighted usable area projections for juvenile coho
and sockeye salmon at the Slough 6A modelling
site.
-37-
Table 3. Correlations between composite weighting factors and catch
transformed by natural log (X+l) for juvenile coho and chinook
salmon by site and by all site~ pooled.
Chinook
Low Turbidit~ ftigh Turbidity
(do NtU) ~ ( >-30 NTU)
!1. .!: !1. .!: ~
Whiskers Creek Slough 30 0.61 0.001 37 0.40 0.066
Slough 22 35 0.81 0.001 17 0.73 0.001
Side Channel lOA 14 0.77 0.001 so 0.19 0.065
Pooled 79 0.72 0.001 104 0.29 0.009
Coho
!1. !. ~
Whiskers Ct·eek Slough 67 0.48 0.001
Slough 6A 62 0.50 0.001
Slough 8 51 0.63 0.001
Slough 5 39 0.58 0.001
Pooled 219 0.45 0.001
21 Significance level for rejection of hypothesis that there is a
positive correlation between composite weighting factors and catch.
Table 4. Chi-square contingency tests of juvenile sockeye and chum salmon
proportional presence by composite weightiny factor intervals.
Sockeye (Data from Sloughs B. 6A, and 5 pooled)
Combined weighting
factor interva 1
0.03-0.12
0.13-0.22
0.23-1.00
x• = 16.7 df = 2
Significant at P < 0.001
Chum (Data from Slough 8)
Combined weighting
factor interva 1
0.24-0.34
0.41-0.66
X2 = 0.0 df = 1
6
12
24
No. of cells
30
25
15
No. of cells
4 4
5 5
Not significant at 0.05 level
-38 -
36
37
39
8
10
Proportion
Present
0.17
0.32
0.62
Proportion
Present
G. SO
0.50
~
-
-
·~"""
-
i.O DISCUSSION
The weighted usable area models for juvenile salmon at critical upland
slough, side slough~ and side channel habitat locations indicate that
both species-specific and site-specific trends exist. The trends
reflect fish suitability for hydraulic conditions, including changes in
surface area. Significantly~ most of the weighted usable area estimates
are affected strongly by the availability of suitable cover. In the
environments modelled~ suitable cover for juvenile chinook salmon
includes turbidity. In all three side channel habitats, peaking of the
weighted usable area function occurs in a narrow range of flows which
occur following the overtopping event. In side and upland slough
habitats, the changes in WUA values for all juvenile salmon species are
related to mainstem backwater effects.
Habitat indices were calculated from the smoothed ~JUA projections
(Appendix A). In this calculation, the weighted usable areas
interpolated at 3,000 cfs increments of mainstem discharge are expressed
as the fraction of the total area available at the site when mainstem
discharge was 23,000 cfs. Plotting these normalized values as a
function of mainstem discharge results in habitat indices by
macrohabitat type for each juvenile salmon species. Habitat index
values are compared with the IFG modelling results in Part 7 of this
report.
4.1 Chinook Salmon
Juvenile chinook habitat was w~delled at three study sites for turbidity
levels above and below 30 NTU (Figure 22). The difference in habitat
index values for the two turbidity conditions largely reflects the
differences in suitability for cover at the sites. Slough 22 appears
roughly as usable as Whiskers Creek Slough under turbid conditions but
is much less usable with low turbidity flows. This reflects the rela-
tively cover-poor environment at Slough 22. The shape of all three side
channel plots shows that the available habitat becomes less suitable fm~
juvenile chinooks as velocity increases at large mainstem discharges.
Since each side channel habitat is breached by mainstem water at slight-
ly different mainstem discharges, a larger sampling of side channels
which are breached by mainstem water at different discharges is required
to formulate average index values for a particular mainstem discharge.
4.2 Coho Salmon
Habitat indices for coho salmon at four sites are plotted in Figure 23.
The habitat indices are much lower than those for chinook and reflect
generally poor rearing habitat for coho in mainstem influenced environ-
ments of the Susitna River. The index for Slough 5 increases primarily
because of a large increase in surface area of the site. These low
indices in generally are primarily the result of a lack of suitable
cover for coho.
The Whiskers Creek Slough site was unusual among the sites sampled
because coho were captured there when turbid water was present. This
was related to the proximity of the slough to a natal area~ Whiskers
Creek.
-39 -
X w
0
!
<
Q
<
J:
>< w
0
! ...
< ....
Cl <
J:
0.6
0.5
0.4
0.3
0.2
0.1
0.0 ' 5
0.6
0.5
0.4
0.3
0.2
0.1
0.0
·5
I
I
CHINOOK HABITAT INDICES
HIGH TURBIDITY MAINSTEM
M.ll91,.11f i-NO.CES
0 Slouol'l 2'2
+ Whle-h•~· Ct•ell $IOYQI!I.
0 5141• CII•Aft•l 1 DA
-C:•IIDUie-CI fhnv•
--£anaaolat•CI Aanv•
,
I
7
I
I
'""---// "'"---,._____
---
+--t--""---t---+--~--·
o---o __
-o---o---c,...--a---o---o---
1
15 25 35
SUSlTNA R. DISCHARGE (CFS • 1000) AT GOLD CREEK
USGS,PRO\IISIONAL DATA 1983 15292000
CHINOOK HABITAT INDICES
LOW TURBIOlTY MAINSTEM
0 :uvva111 ll
+ 'fUthketa cr • .-IL SJv~vn
0 Sl4• C.fl.•""'•' Jo-111
-c;aue,at•'lt R.-noe-
--[ao'"*CIOf•t•ci 'llf•no•
15 25. 35
SUSITNA R. DISCHARGE (CFS • 1 000) AT GOLD CREEK
USGS PROVISIONAL DATA 1983 15292000
45
45
Figure 22. Habitat indices for juvenile chinook sa1mon.
-40-
-
...,
-
-
-
0.13
)~ 0.12
0.11
0.10
0.09
X 0.08 w -0 z 0.07
1-
< 0.06 !-
..::::
<( 0.05 J:
0.04
0.03
0.02
0.01
0.00
.6---
¢---
5
COHO HABITAT INDICES
HABITAT I>IDICES
0 Whlakel"& Creek Slouoh
+ Slough 6A
0 Slough 6
1:> Sl~ugh 5
-Callbraled Range
--E.x.lfapota•ed Range
o"
/ ---o " ........
"'"' / \
/ \
I' \
" a \
<:>" /
,/
\
\
\ a
'
o-,..
,/
\
' 'o
<:>-----
' -+---+---+---+---+-~~0~--1
15 25 35
SUSITNA R. DISCHARGE CCFS x 1000) AT GOLD CREEK
USGS PROVISIONAL DATA 1983 15292000
--
45
Figure 23. Habitat indices for juvenile coho salmon.
-41-
4.3 Sockeye and Chum Salmon
Sockeye salmon habitat indices increased with discharge over the
modelled range (Figure 24}. Upland sloughs (Slough 6A and 5} become
increasingly important habitats for juvenile sockeye salmon as mainstem
discharge increases due to the backwater effects. These two sloughs
represent the extremes in changes of conditions possible for this type
of habitat; Slough 6A has a steep banked, well defined channel and
Slough 5 has very low gradient banks which are quickly overtopped by
backwater. Only Slough 8 was modelled for chum salmon and the habitat
index increased with mainstem discharge (Figure 25). With further
increases in mainstem discharge, however, the indices for both chum and
sockeye at Slough 8 would decline due to velocity becoming important in
limiting distribution.
4.4 Limitations of the t4odels ·Regarding Methodology
The methods employed in this study were intended to provide a rapid and
quantitative estimation of the overall effects of mainstem Susitna River
discharge on the suitability of selected rearing habitats for juvenile
salmon. Simultaneously, IFG-2 and IFG-4 models were developed at
companion side slough and side channel sites (Part 7)~ Because habitat
parameters were measured at only three cells along each transect in this
study, we do not expect that these predictions 'IIi 11 provide the same
degree of resolution that will result from using well calibrated
multi-cell hydraulic models.
The WUA calculations projected for mainstem flows not observed are
generally subject to review. In the case of projections for lov1
mainstem flows at side sloughs, however, conditions were nearly static
so that extrapolations to 6,000 cfs (mainstem discharge) are reasonably
solid. In contrast, forecasts for high flow conditions at mainstem side
channels should be used as preliminary estimates.
However, we believe that in large glacial systems, like this reach of
the Susitna River, catastrophic hydraulic events and the availability of
cover are major factors related to the distribution and relative
abundance of juvenile saTmonids. Our model is designed to provide the
resolution necessary to observe overall changes related to these phenom-
ena, and we believe that it does.
4.5 Model Verification
Chinook salmon distribution in low turbidity waters was strongly
correlated to the composite weighting factor index but the correlations
for chinook salmon in turbid water were much lower. The lower corre-
lations in turbid \'later may reflect gear efficiency problems because
beach seines were used in turbid water and their efficiency varies
-42 -
-
,....
-
-
>C ...
Q
3!:
1-<
!::
CD < :r
-
.. ...
Q :;
..... < 1-
iii < :r
SOCKEYE HABITAT INDICES
0.~
0.8
0.7 + t••u•h IS
0.6 -c ..... , .. t.o~~ ...... ,.
--l•tfe"'•t•tl "••••
0.5
0.~
0.3
0.2
0.1j •:=:
0.0~-----.-----..------r-----•• -----.-----~-----.----~
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0.35
0.30
0.25
5 15 25 35
SUSlTNA R. DISCHARGE (CF.S >1000) AT GOLD CREEK
USGS PROVISIONAL OA TA 1983 15292000
Figure 24. Habitat indices for juvenile sockeye salmon.
a---
5
CHUM HABITAT INDICES
15 25 35
SUSITNA R. DISCHARGE (CFS • 1000) AT GOLD CREEK
USGS PROVISIONAL OAT A 1983 15292000
Figure 25. Habitat indices for juvenile chum salmon.
-43-
45
•5
widely with cover type and other habitat conditions (Part 2 of this
report}. Electrofishing gear, used as a sampling method in clearer
waters, was believed to be more reliable when sampling diverse habitat.
Coho and sockeye salmon also were correlated to or associated with the
calculated composite \'oieighting factors. Chum salmon catches were so
limited at the six model sites that the relationship of composite
weighting factors to fish use remains unproven. Factors such as season~
of course, are strongly related to fish abundance and obscure the
relationships. The analysis is also specific to the ice free months and
no analyses of winter processes have been made. Since there is a
positive relationship bet\oJeen the composite weighting factors and fish
catch at the cell level and by inference between WUA and fish use at the
site level, the models are verified on at least a general basis although
many refinements in the model are possible.
-44 -
-
~·
-
-
"""
-
-
-
-
-
....
5.0 CONTRIBUTORS
l4oody Trihey and Steve Hale provided helpful discussions. Larry Dugan
and Dave Sterritt helped collect the field data. Tommy Withrmv, Jodi
Miller, Pat r~orrow, and Chris Kent installed our staff gages. Allen
Bingham and staff managed the mainframe computer data base. Sally
Donovan and Carol Riedner did the art work for the final copy. Woody
Trihey, Allen Bingham, and Kathrin Zosel reviewed a draft of this paper
and provided helpful comments.
-45 -
6.0 ACKNOWLEDGEMENTS
Funding for this study was provided by the State of Alaska, Alaska Power
Authority.
We wish to thank the various staff at E. Woody Trihey and Associates
(EWT&A), the Arctic Environmental Information Data Center, Harza-Ebasco
Susitna Joint Venture, Woodward-Clyde Consultants, The Alaska Power
Authority, and especially Cleve Steward (EWT&A) for reviewing the draft
of this paper and providing helpful suggestions.
-46 -
-
-
-
....
....
-
-
....
7.0 LITERATURE CITED
Acres Americans Inc. (Acres). 1982. Susitna hydroelectric project draft
FERC license application, Exhibit E, Chapter 3, Anchorage, Alaska.
Alaska Department of Fish and Game (ADF&G) 198L Juvenile anadromous
fish investigations on the lower Susitna River. Phase I. Final
Draft. Prepared for Acres American, Inc., by Alaska Department of
Fish and Game/Su Hydro, Anchorage, Alaska.
• 1982. Aquatic studies program. Phase I. Final Draft.
--...... S,.-u'btask 7.10. Prepared for Acres Americans Incq by the Alaska
Department of Fish and Game/Su Hydro. Anchorage, Alaska.
---....,.,....· 1983a. Resident and juvenile anadromous fish studies on the
Susitna River below Devil Canyon, 1982. Volume 3 of Phase II Basic
Data Report. ADF&G Su Hydro Aquatic Studies Program. Anchorage;
Alaska.
----~· 1983b. Synopsis of the 1982 aquatic studies and analysis of
f1sh and habitat relationships. Appendices. Phase II Report.
ADF&G Su Hydro Aquatic Studies Program. Anchorage, Alaska.
• 1983c. Synopsis of the 1982 Aquatic Studies and Analysis of
----=-r Fish and Habitat Relationships. Phase II Report. ADF&G Su Hydro
Aquatic Studies Program. Anchorage, Alaska.
--~· 1984 Aquatic Habitat Studies Procedures ~1anual. Phase III.
ADF&G Su Hydro Aquatic Studies Program. Anchorage, Alaska.
Bovee, K.O. 1982. A guide to stream habitat analysis using the
instream flow incrementa 1 methodology. Instream Flow Information
Paper. No. 12. U.S. Fish and Wildlife Service FWS/035-82/26.
-47 -
......
, ....
-
-
-
APPENDIX A
Weighted Usable Area and Habitat Indices
Tablulated by Site and Species
-48 -
,....
"""'
!"""
~"""'
,....
!"""
-
Weighted usable area (WUA} and habitat index (HI) projections for
species captured in abundance at the juvenile salmon rearing habitat
model study sites during the summer of 1983 (Appendix Tables Al through
A6). The habitat index is calculated as the weighted usable area
divided by the sites surface area at a mainstem Susitna River discharge
of 23,000 cfs.
Appendix Tabte_A-1. Weighted usable area and habitat indices for Side Channel
lOA.
Chinook Salmon ·
Mainstem
Discharge Turbi dit,!!: 30 NTU Turbidit;t 30 NTU
(cfs) WUA !!!. WUA HI I
5,000* 0 0.000 0 0.000
I 9,000* 18,580 0.171 8,400 0.078
12,000 27,700 0.256 11,000 0.102
15,000 25,500 0.236 11,000 0.102
18,000 24,400 0.226 11,500 0.106 J 21,000 23,300 0.216 10,800 0.100
24,000 21,100 0.195 9,500 0.088 l 27,000 16,800 0.156 7,600 0.070
30,000 n ,Joo 0.105 4,600 0.043 l
33,000* 9,000 0.083 3,500 0.032 I 36,000* 7,500 0.069 3,000 0.028 I I 39,000* 6,400 0.059 2,700 0.025 I 42,000* 5,700 0.053 2,400 0.022 I 45 ,000* 5,100 0.047 2,300 0.021
I
The surface area at 23,000 cfs was 108,000 ftZ
* Data at this discharge extrapolated.
-49 -
Appendix Table A-2. Weighted usable area and habitat indices for Slough ~
224
-Chin'OOk Salmon
Hainstem
Ofscliarge
Turbidit;i> 30 NTU Turbidity§ 30 NTU
~ WUA !!l WUA !!!.
6000* 2500 0.035 2500 0.035
_,
9000* 2500 0.035 2500 0.035
12000* 2500 0.035 zsoo 0.035
15000* 2500 0.035 2500 0.035 -18000* 2500 0.035 2500 0.035
21000~/ 2500 0.035 2500 0.035
24ooo!Y 27100 0.382 6000 0.085
27000 40500 0,570 10100 0.142 ~
30000 18200 0.256 5800 0.082
33000* 13300 0.187 4800 0.068
36000* 11500 0.162 4100 0.058 ~
39000* 10000 0.141 3600 0.051
42000* 8800 0.124 3400 0.048
45000* 7600 0.107 3100 0.044
The surface area at 23,000 cfs was 71,000 ft.
y : Side slough condition ~
£1 : Side channel condition
*Data at this discharge extrapolated -
Appendix Table A-3. Weighted usable area and habitat indice for
Whiskers Creek Slough. -
Chinook. Salmon Coho Salmon
Mainstem Turbidit,r:) 30 NTU Turbidit;i ~ 30 NTU All Turbidity
Discharge
..lilll... WUA !!!. WUA HI ~ Hf
6000* 2300 0.059 2300 0.059 1600 0.041 -9000* 2300 0,059 2300 0.059 1600 0.041
12000 2300 0.059 2300 0.059 1600 0.041
15000 2300 0.059 2300 0.059 1600 0.041
18000 2300 0.059 2300 0.059 1600 0.041 _,
2100oY 2400 0.062 2400 0.062 1600 0.041
2400o!?i 18200 0.467 5600 0.144 2700 0.069
27000 20100 0.515 8900 0.228 3600 0.092
30000* 18900 0.485 9600 0.246 3600 0.092 -
33000* 15500 0.397 9300 0.238 2900 0.074
36000* 11200 0.287 8400 0.215 2200 0.056
39000'" 8500 0.218 7300 0.187 1600 0.041
42000'" 6900 0.177 5700 0.146 1200 0.031
45000* 5900 0.151 4100 0.105 1100 0.028
~
The surface area at 23,000 cfs was 39,000 ft.
!Y : Side slough condition
'!Y -= Side channel condition
*Oata at this dischar9e extrapolated
-50
-
Appendix Table A-4. Weighted usable area and nabitat indices for Slough
8.
Mains tern Chum Salmon Coho Salmon Sockexe Salmon
Discharge
...illll.. WUA HI WUA HI WUA HI
6000* 5300 0.273 670 0.035 1750 0,090
9000 5400 0.278 6go 0.036 1780 0.092
12000 5800 0.299 960 0.049 1910 0.098
15000 6900 0.356 1010 0.052 2160 0.111
18000 7300 0.376 890 0.046 2550 0.131
21000 7400 0.381 580 0.030 3200 0.165
24000 7800 0.402 540 0.028 3860 0.199
27000 9350 0.482 790 0.041 4600 0.237
30000 11800 0.608 1020 0.053 5320 0.274
33000 13200 0.680 1140 0.059 5780 0.298
ihe surface area at 23,000 cfs was 19,400
*Oata at this discharge extrapolated
ft2•
Appendix Table A-5. Weighted usable area and habitat indices for Slough
5.
l•lainstem Coho Sa]l!l(lrl
Discharge
__!fill_ WUA HI
6000* 2500 0.058
9000 2400 0.056
12000 1400 0.033
15000 1200 0.028
18000 1600 0.037
21000 2100 0.049
24000 2600 0,060
27000 3200 0.074
30000* 3700 0.086
33000* 4200 0,098
36000* 4600 0.107
39000"" 5000 0.116
42000* 5200 0.121
45000* 5300 0.123
The surface area at 23,000 cfs was 43,000 ft.
*Data at this discharge extrapolated
Sockei:e Salmon
WUA !!!.
4200 0.098
4700 0.109
5000 0.116
6700 0.156
9400 0.219
13000 0.302
15900 0.370
17400 0.405
18800 0.437
21200 0.493
26000 0.605
29200 0.679
32800 0.763
36900 0.858
Appendix Table A-6. ~Jeighted usable area and habitat indices for Slough 6A.
Mainstem
Oi scharge Coho Salnnn
.tml WUA ~
6,000* 2,350 0.024
9,000* 2,510 0.026
12,000 2,670 0.028
15,000 2,870 0.030
18,000 2,970 0.031
21 ,000 3,000 0.031
24,000'* 3,020 0.031
27,000'* 3,040 0.031
30,000'* 3,060 0.032
33 ,000'* 3,080 0.032
36,000* 3,110 0.032
39,000* 3,140 0.032
42,000* 3,170 0.033
45,000'* 3,200 0.033
The surface area at 23,000 cfs was 96,800 ft'
* Data at this discharge extrapolated.
-51 -
Sockexe Salmon wuA !!!.
22,000 0.227
22,600 0.233
23,200 0.240
24,100 0.249
25,400 0.262
26,200 0.271
26,400 0.273
26,600 0.275
26,900 0.278
27,000 0.279
27,200 0.281
27,400 0.283
27,500 0.284
27,600 0.285
...
...
PART 5
Resident Fish Distribution and Population Dynamics
in the Susitna River below Devil Canyon
-
F"'
I
ABSTRACT
RESIDENT FISH DISTRIBUTION AND POPULATION DYNAMICS
IN THE SUSITNA RIVER BELOW DEVIL CANYON
1984 Report No. 2, Part 5
by Richard L. Sundet and Mark N. Wenger
Alaska Department of Fish and Game
Susitna Hydro Aquatic Studies
2207 Spenard Road
Anchorage, Alaska 99503
Studies of resident fish during 1983 were concentrated on the reach of
the Susitna River between the Chulitna River confluence and Devil
Canyon. With the use of radio telemetry and mark and recapture methods,
the seasonal distribution of rainbow trout and estimates of local
abundance were obtained. Examination of recapture data over the past
several years suggests that the rainbow trout population in this reach
is probably less than 4,000 fish. Most of the concentrations are in the
smaller tributaries, particularly Fourth of July Creek, which also has
the only significant amount of successful spawning documented so far in
this portion of the Susitrra basin. The large tributaries, Portage Creek
and Indian River, had comparatively small numbers of rearing rainbow
trout. This species spends much of its annual life cycle in the main-
stem Susitna near tributary mouth areas or mixing zone confluences of
sloughs. Much of the migratory movements during the summer appear to be
in response to the influx of adult salmon spawners, whose eggs apparent-
ly provide a major source of food. Radio tagged rainbow trout movement
data suggests that the mainstem is important for overwintering. Limited
data from tagged rainbow trout below the Chulitna River confluence
suggests the reach of river between RM 78.0 and Talkeetna may also be an
important overwintering area for Talkeetna River stocks as well. Spawn-
ing of round whitefish in October and probably burbot in January is
directly influenced by mainstem flows. Young age class Arctic grayling
and round whitefish appear to reside in the mainstem Susitna, usually
near tributary or slough mouths. Nearly all of the spawning and most of
the rearing of older age class Arctic grayling occurs in tributaries.
Arctic grayling overwinter in the mainstem Susitna. Dolly Varden are
rare in this reach of the Susitna. Selected sites have been established
that can be used to monitor catch per unit effort of the resident
species, and consequently their response to flow regulation of the
proposed hydroelectric project.
i
TABLE OF CONTENTS
ABSTRACT •••••••••••
LIST OF FIGURES ...••.•..•••••
LIST OF TABLES ..................................................... .
LIST OF PLATES •••.•.•••••••••.••••••.•••••••••••.•.•.••••.•.•.• Ill •••
LIST OF APPENDIX FIGURES ••••••••••••••••••••
LIST ·OF APPENDIX TABLES •••••••••.••••••
1.0
2.0
3.0
INTRODUCTION •••• ..............................................
METHODS •••••••••••••.•••
2.1
2.2
2.3
Study Locations ••••
2.1.1
2.1.2
2.1.3
Relative abundance measurements.
Population estimates ••••••••••••••••
Radio telemetry ............................. .
Data Collection •.•••••••••• ...............................
2.2.1
2.2.2
2.2.3
Relative abundance ••
Population estimates •• ; •••.•••..•••••••••••
Radio telemetry ........................... .
2.2.3.1
2.2.3.2
2.2.3.3
Equipment •....•....•..•.•..••..•.•..••...
Transmitter implantation •••••••••••••••••
Tracking ................................. .
Data Recording and Analysis ••••••.••••••.•••••.••••••••••
RESULTS •••••••••••••••••••••••••••••••••••••••••••••••••••••••
3.1
3.2
Rainbow Trout ......•..............•.............
3.1.1
3.1.2
3.1.3
Distribution and relative abundance ••••••••••••••.
Movement and migration .....•..............•.......
Population estimates .....•..••.............•......
Burbot .........••..........•....•..........•.............
3.2.1
3.2.2
3.2.3
Distribution and relative abundance ••••••••••••.••
Movement and migration ..........•.................
Population estimates ••••••••••••••.•••••••••••••••
i i
i
v
vi
vii
viii
xi
1
3
3
3
3
3
6
6
8
9
9
9
10
13
14
14
14
14
18
18
18
20
20
3.3
3.4
3.5
3.6
3.7
Arctic Grayling .............................. Cl •••••••••••
3.3.1
3.3.2
Distribution and relative abundance •.••..•.••••.••
Movement and migration .••••...••.••.•..•.•..••.•••
Round Whitefish •••••••••••..•.••.•.••••.••.••.•••..••.•.•
3.4.1
3.4.2
Distribution and relative abundance .••.•••••••...•
Movement and migration ..•••••••.•••.•..•••••••••..
Humpback Whitefish •••••••••••.•••••••••••••••••••.•.•••
3.5.1
3.5.2
Distribution and relative abundance ..••••.••.•.•••
Movement and migration ••••••••••••••••.•••...••.••
Longnose Sucker ..
3.6.1
3.6.2
Distribution and relative abundance •••..•••••.•.•.
Movement and migration ........................... .
Other Species ........................................... .
3.7.1
3.7.2
3.7.3
Dolly Varden •..•....••.••...••.....••..•..•..•..••
Threespine stickleback •••••••••.•••••••••••••••.••
A ret i c 1 amprey ..............•.....................
20
20
20
23
23
23
25
25
25
25
25
27
27
27
27
27
4. 0 DISCUSS I ON • . • • . . . . • . • . . • . . . . . . . . • . . . . • • . . . . . . . . . • . • . . . • . . . . • . • . 2 8
5.0
6.0
7.0
4.1
4.2
4.3
4.4
4.5
4.6
4.7
Rainbow Trout ••••••..••..••..••••••••.
Burbot .................................................. ,.,
Arctic Grayling •...•••••••••••.
Round Whitefish ••••••••
Humpback Whitefish •.•.•
Longnose Sucker •....•.....•....•.....•..
Other Species ........................................... .
4.7.1
4.7.2
4.7.3
Dolly Varden •••.••..•••••••..•.••••.••.•••..••..••
Threespine stickleback •••••••......••.....••.•.••.
Arctic lamprey ................................... .
CONTRIBUTORS •••••••••
ACKNOWLEDGEMENTS ••.•.••••.. ...................................
LITERATURE CITED •..•....•.•
iii
28
32
34
36
38
40
41
41
41
42
43
44
45
-
-
-
-
-
-
.....
-
-
-
-
8.0 APPENDICES
Appendix A Gear efficiency and selectivity and tag retention
efficiency........................................ 49
Appendix B Radio tagged fish movement data.................. 62
Appendix C Population and biological characteristics......... 65
Appendix D Population estimates.............................. 94
iv
--
-
....
-
LIST OF FIGURES
Figure Title
1 Resident fish study sites on the Susitna River
between the Chulitna River confluence and Devil
2
Canyon, 1983........ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Movement of five radio tagged rainbow trout in the
Susitna River below Devil Canyon, May to December
1983 . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3 Movement of four radio tagged rainbow trout in the
Susitna River below Devil Canyon, June to December
1983........................................................ 16
4 Movement of six radio tagged rainbow trout in the
Susitna River below Devil Canyon, June to December
1983 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5 Movement of eleven radio tagged rainbow trout in the
Susitna River below Devil Canyon, September to
December 1983 .•••..•..•.••...••.••••.•..••.•••• ·. • • • . • . . • . . • • 17
6 Movement of four radio tagged burbot in the Susitna
River below Devil Canyon, July to December 1983............. 21
v
LIST OF, TABLES
Table
1
2
3
4
5
Title
Resident fish population estimate sites on the
Susitna River between the Chulitna River
confluence and Devil Canyon, 1983........................... 5
Rainbow trout catch on the Susitna River between the
Chulitna River confluence and Devil Canyon, May to
October 1983................................................ 15
Burbot catch on the Susitna River between the
Chulitna River confluence and Devil Canyon, May to
October 1983................................................ 19
Arctic grayling catch on the Susitna River between
the Chulitna River confluence and Devil Canyon, May
to October 1983............................................. 22
Round whitefish catch on the Susitna River between
the Chulitna River confluence and Devil Canyon, May
to October 1983............................................. 24
6 Humpback whitefish catch on the Sus.itna River
between the Chulitna River confluence and Devil
7
Canyon, May to October 1983................................. 26
Longnose sucker catch on the Susitna River between
the Chulitna River confluence and Devil Canyon, May
to October 1983............................................. 26
vi
-'
-
-
-
r
-
LIST OF PLATES
Plate Title Page
1 Electrofishing with a boat mounted electroshocking
unit at mainstem Susitna gravel bar opposite
Montana Creek (RM 78.0)..................................... 7
2 Implanting a radio tag into the abdomen of a rainbow
trout....................................................... 11
3 Implant·ing a radio tag under the skin of a rainbow trout •••. 12
vii
-----------------------------------•-w-•~-· ---------------------------------------------
-
LIST OF APPENDIX FIGURES
Appendix Figure Title Page
A-1 Gear selectivity for rainbow trout in
the Susitna River, May through October
1983. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
A-2 Gear selectivity for burbot in the
Susitna River, May through October
1983.............................................. 54
A-3 Gear selectivity for Arctic grayling in
the Susitna River, May through October
A-4
1983 .............................................. 55
Gear selectivity for round whitefish in
the Susitna River, May through October
1983 .............•................................ 56
A-5 Gear selectivity for humpback whitefish
in the Susitna River, May through
A-6
C-1
C-2
C-3
October 1983...................................... 57
Gear selectivity for longnose suckers in
the Susitna River, May through October
1983............. .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Age and length relationship for spawning
rainbow trout captured in the Susitna
River between the Chulitna River con-
fluence and Devil Canyon, May 11 to July
18, 1983........................... .. . . . . . . . . . . . . . 67
Length frequency composition of rainbow
trout captured in the Susitna River
between the Chulitna River confluence
and Devil Canyon by all gear type, May
to October 1983. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
Age and length relationship for rainbow
trout captured in the Susitna River
between the Chulitna River confluence
and Devil Canyon, May to October 1983............. 70
C-4 Survival rate curve for rainbow trout-
captured in the Susitna River between
the Chulitna River confluence and Devil
C-5.
Canyon, 1983...................................... 71
Length frequency composition of burbot
captured in the Susitna River between
the Chulitna River confluence and Devil
Canyon by all gear types, May to October
1983.............................................. 72
viii
LIST OF APPENDIX FIGURES (Continued)
Appendix Figure
C-6.
Title
Survival rate curve for pooled burbot
catch data from the Susitna River
between Cook Inlet and Devil Canyon,
1981 to 1983...................................... 73
C-7. Age and length relationships for spawn-
ing Arctic grayling captured in the
Susitna River between the Chulitna River
confluence and Devil Canyon, May 20 to
June 22, 1983..................................... 75
C-8 Length frequency composition of Arctic
grayling captured in the Susitna River
between the Chulitna River confluence
and Devil Canyon by all gear types, May
to October 1983................................... 76
C-9
C-10
Age and length relationship for Arctic
grayling captured in the Susitna River
between the Chulitna River confluence
and Devil Canyon, May to October 1983 .•.••••.••••.
Survival rate curve for Arctic grayling
captured in the Susitna River between
the Chulitna River confluence and Devil
77
Canyon, 1983...................................... 79
C-11 Age and length relationship for spawning
round whitefish in the Susitna River
between the Chulitna River confluence
and Devil Canyon, October 4 to October
7 ' 1983. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
C-12 Length frequency composition of round
whitefish captured in the Susitna River
between the Chulitna River confluence
and Devil Canyon by all gear types, May
C-13
to October 1983................................... 81
Age and length relationships for round
whitefish captured in the Susitna River
between the Chulitna River confluence
and Devil Canyon, May to October 1983............. 84
C-14 · Survival rate curve for round whitefish
captured in the Susitna ·River between
the Chulitna River confluence and Devil
Canyon, 1983...................................... 85
ix
-
-
-
LIST OF APPENDIX FIGURES (Continued)
Appendix Figure
C-15
Title
Length frequency composition of humpback
whitefish captured in the Susitna River
between Cook Inlet and Devil Canyon by
all gear types~ May to October 1983............... 86
C-16 Age and length relationship for humpback
whitefish captured in the Susitna River
between Cook Inlet and Devil Canyon, May
C-17
C-18
to October 1983 ••• ~............................... 88
Age and length relationships for spawn-
ing longnose suckers captured in the
Susitna River between the Chulitna River
confluence and Devil Canyon, May to
October 1983....................................... 89
Length frequency composition of longnose
suckers captured in the Susitna Rivet
between the Chulitna River confluence
and Devil Canyon by all gear types, May
to October 1983................................... 90
C-19 Age and length relationship for longnose
suckers captured in the Susitna River
between the Chulitna River confluence
and Devil Canyon, May to October 1983............. 93
X
-
-
LIST OF APPENDIX TABLES
Appendix Table
B-1
B-2
C-1
Title
Summary of tagging data for radio tagged
rainbow trout on the Susitna River
between the Cook Inlet and Devil Canyon,
May to December 1983. . • . . • . • • . . . . . . . . . . . . • . . . • . . . . 63
Summary of tagging and tracking data for
radio tagged burbot on the Susitna River
between Cook Inlet and Devil Canyon,
July to December 1983............................. 64
Rainbow trout age-length relationships
on the Susitna River between the
Chulitna River confluence and Devil
Canyon, 1983. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 9
C-2 Arctic grayling age-length relationships
on the Susitna River between the
Chulitna River confluence and Devil
Canyon, May to October 1983....................... 78
C-3 Round whitefish age-length relationships
on the Susitna River between the
Chulitna River confluence and Devil
Canyon, May to October 1983....................... 83
C-4 Humpback whitefish age-length relation-
ships on the Susitna River between Cook
Inlet and Devil Canyon, May to October
1983.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
C-5 Longnose sucker age-length relationships
on the Susitna River between Chulitna
Inlet confluence and Devil Canyon, May
D-1
to October 1983................................... 91
Biases, corrections, and assumptions
which affect the resident fish
population estimate below Devil Canyon,
1 983 . . • . • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
xi
,----a----------------,-·-----""""""1-----------
-
"""" I
-
1.0 INTRODUCTION
Study of resident fi s~/ species began in the fa 11 of 1980 to co 11 ect
baseline data to meet the following objectives:
A. Define seasonal distribution and relative abundance of resi-
dent fish species in the Susitna River between Cook Inlet and
Devil Canyon.
B. Characterize the seasonal habitat requirements of selected
resident fish species within the study area.
During the 1983-84 season, the Resident Fish Studies were refined to
also address the following sub-objective:
C. Quantify the important habitat parameters associated with
spawning and rearing (growth) of selected resident fish
species and measure fish density in spawning and rearing
habitats to provide an estimate of habitat quality.
The rationale behind these objectives is that often there can be
changes in fish populations after the construction of a hydroelectric
dam. These postproject effects result from changes in water
temperature, flow, turbidity, and other water quality parameters.
Preproject baseline fisheries data and their correlation to habitat
conditions, therefore, are necessary to evaluate the potential impact to
these fisheries.
Studies on how resident fisheries are affected by hydro-projects similar
in magnitude to the Susitna proposals are limited. One of the better
pre-and post-project studies was conducted by the Montana Department of
Fish, Wildlife, and Parks on the Kootenai River below the Libby Dam site
(MDFW&P 1983). The overa 11 effects of_ the dam were conducive to
increased production of rainbow trout and mountain whitefish but
adversely affected sturgeon. A quality sport fishery has arisen in the
regulated waters below the project after an initial five year problem
with supersaturation of dissolved gas. In recent years, however, the
average size of the rainbow trout have decreased, which may be related
to sport fishing and perhaps to changes in invertebrate comunity
structure caused by power peak·i ng fluctuations. The system remains one
of the more productive rivers in this portion of the state of Montana.
Provision for proper downstream flow is considered by these researchers
to be the primary reason the fisheries have deve 1 oped favorably after
project operation.
Sport fishing for rainbow trout and Arctic grayling in the Susitna River
drainage occurs throughout the open water season, primarily around the
mouths of clearwater tributaries. Burbot fishing occurs mostly in the
mainstem Susitna River or at the mouths of clear water tributaries
during both summer and winter. In the Chulitna River confluence to
For the purposes of this report "resident fish" will be defined as
any fish species which spend their entire 1 ife cycle within the
Susitna River drainage.
- 1 -
Devil Canyon reach of the Susitna River, the reach that will probably be
most affected by the proposed hydroelectric project, sport fishing
occurs at Whiskers Creek [river mile (RM) 101.4], Lane Creek (RM 113.6),
Fourth of July Creek (RM 131.1), Indian River (RM 138.6), and Portage
Creek (RM 148.8). Current information on the extent of the harvest of
these resident fish species is limited to data available from Mills
(1982) for the entire Susitna River basin. These catches have been
stab 1 e for the past five years, with the average harvest of rainbow
trout and burbot at 20,000 and 700 fish respectively. The level of
fishing effort will probably increase in the Susitna River drainage
during the next decade.
- 2 -
-
-
-
-
-
....
-
-
-
-
2.0 METHODS
This report addresses resident fish studies conducted during the open
water period of 1983, spawning surveys done in early May, and radio
telemetry results through December 1, 1983. Telemetry results are
presented through December 1 to show the movement patterns during the
transition period from open water to winter conditions. Although most
of the sampling occurred in the mainstem Susitna River between the
Chulitna River confluence to Devil Canyon, a few other areas were also
studied.
2.1 Study Locations
2.1.1 Relative abundance measurements
Thirteen index sites were sampled twice per month by boat electrofishing
to monitor seasonal trends in relative abundance of resident fish
(Figure 1). In addition, other mainstem, side channel, slough, and
tributary sites on the Susitna River between the Chulitna River conflu-
ence and Devil Canyon were also sampled intermittently.
The upper reaches of Fourth of July Creek (RM 131.1), Indian River (RM
138.6), and Portage Creek (RM 148.8) were sampled to determine the
extent of resident fish spawning and rearing. These tributaries were
selected because of their size, their proximity to Devil Canyon, and
their relatively high abundance of resident fish species. Fourth of
July Creek was sampled in May, June and July between tributary river
mile (TRM) 0.0 and TRM 2.3. Indian River was sampled in June and August
between TRM 1.5 and TRM 14.0, while Portage Creek was sampled in June at
TRM 6.0 and TRM 10.0.
Resident fish catches recorded at four fishwheel sites, two downstream
migrant traps (RM 103.0), and 35 juvenile salmon rearing study sites
were also examined to evaluate trends in relative abundance and seasonal
movements.
2.1.2 Population estimates
Resident fish population estimates were attempted at five sites on the
Sus i tna River between the Chulitna River confluence and De vi 1 Canyon
(Table 1). These sites included a slough, a side channel, a tributary,
a tributary mouth, and a one-mile reach of the mainstem Susitna River.
2.1.3 Radio telemetry
Selection of radio tagging sites in the mainstem Susitna between the
Chulitna River confluence and Devil Canyon were based on resident fish
distr·ibution data collected during the 1981 and 1982 open water field
seasons (ADF&G 1981c; 1983b). Primary efforts to capture rainbow trout
(Salmo gairdneri Richardson) in the mainstem were focused at the mouths
of Whiskers Creek (RM 101.4), Lane Creek (RM 113.6), Fourth of July
Creek (RM 131.1) and Indian River (RM 138.6). Backwater areas in the
mainstenrwere sampled for burbot (Lata lata Linnaeus). The upper
- 3 -
Slough 6A
Talk~~tna R.
usitna Moinstem-West Bank
Site
~hiskers Creek Slough -Mouth
Slough 6A
lane Creek -Mouth
Skull Creek-Mouth
Slough SA
Fourth of July Creek
Susitna Mainstem -~est Bank
:ncian River -Mouth
Slough 20 -Mouth
Jack Long Creek -Mouth
Susitna Mainstem
Portage Cr~ek -Mouth
Sus itna r~a ins tem -Eddy
River Mile
101.2
112.3
113.6
124.7
125. 3
131. 1
137.3-138.3
!38.6
!40.1
144.5
!47.0-148.0
148.8
150.1
Figure 1. Resident fish study sites on the Susitna River between the
Chulitna River confluence and Devil Canyon, 1983.
-4-
-
-
.....
-
....
-
-
....
-
Table 1. Resident fish population estimate sites on the Susitna River
between the Chulitna River confluence and Devil Canyon, 1983.
Location RM TRM Dates Occasion Methods
Slough 8A 125.3 7/15-7/17 6 boat electro-
fishing
4th of July 131.1 0.0-0.8 7/19-7/21 3 hook & line
Creek
Mainstem 131.0-7/15-7/16 4 gi 11 net and
131.8 hoop net
Mainstem 138.9-7/1-7/4 4 trotline, burbot
140.1 sets, and hoop
nets
Jack Long
Creek 144.5 0.0 8/10 3 boat electro-
fishing
Note -Population estimates were also begun at seven other locations
in 1983 but were not completed due to insufficient captures of
fish.
- 5 -
reaches of Fourth of July Creek, Indian River, and Portage Creek were
also sampled for spawning or rearing rainbow trout.
2.2 Data Collection
2.2.1 Relative abundance
Resident fish were collected at mainstem and tributary sites primarily
with a boat mounted electrofishing unit (Plate 1). A Coffelt Model
VVP-3E boat electrofishing unit powered by a 2,500 watt Onan generator
was used for boat electrofishing and techniques used are described in
· the 1982-83 procedures manua 1 (ADF&G 1983a). Secondary gear types used
included downstream migrant traps at RM 103.0, backpack electrofishing
units, gill nets, hook and line, hoop nets, trotlines, and catfish
traps. Baited hoop nets, trotlines and catfish traps were used mainly
to capture burbot. Catfish traps were introduced as a new sampling
technique in 1983. They were set and fished using techniques similar to
those described for hoop nets (ADF&G 1983a).
All resident fish were identified to species. Biological data (age,
length, sex, and sexual maturity) were collected as outlined in the
1982-83 procedures manual. Scales for age determination were taken from
a representative sample of rainbow trout, Arctic grayling (Thymallus
arcticus Pallas), round whitefish (Prosopium cylindraceum Pallas),
humpback whitefish (Coregonus pidschian Gmelin), and longnose suckers
(Catostomus catostomus Forster).
Survival rates for selected resident fish species were calculated using
catch and age data following the methods of Everhart et al. (1975). The
log of the number of fish for each age class was plotted. Then, a
regression 1 ine was fit to the descending leg of the graph. Points
(numbers in an age class) in the descended leg were used after the peak
and to the oldest age class consisting of greater than three points.
The equations are:
loge S = Z
S = e-z = eb
where: S = survival
Z = instantaneous mortality rate
b = slope of regression between the log of the number of
fish and year classes
Resident fish spawning data were collected whenever gravid female fish
were captured. A gravid female ·fish was defined in this study as one
which expelled eggs when . its abdomen was palpated. Because of
turbidity, direct observations of redds was not possible.
A tag-and-recapture program was continued in 1983 to monitor the season-
al movements of adult resident fish. Flay anchor tags were used to tag
seven species of adult resident fish: humpback whitefish, round
- 6 -
....
-
-
-
-
....
-
-
1 I .. -1 J
Plate 1. Electrofishing with a boat mounted electroshocking unit at Mainstem Susitna-gravel bar
opposite Montana Creek (RM 78.0).
whitefish, burbot, longnose suckers, rainbow trout, Arctic grayling, and
Dolly Varden {Salvelinus malma Walbaum). All resident fish that ap-
peared heal thy after capture and were large enough to be tagged were
tagged. Burbot with a total length of 225 millimeters (mm) or greater
were tagged. All other resident fish with fork lengths greater than 200
mm were tagged. Tag recoveries were made by the resident fish study
group, the adult salmon fishwheel crews, and the angling public.
2.2.2 Population estimates
Population estimates for rainbow trout, Arctic grayling, burbot, round
whitefish, and longnose suckers were attempted at five representative
sites (Table 1). The study design followed that outlined by Otis et al.
(1978) and White et al. (1982) which uses a computer program called
CAPTURE to calculate the population estimates and associated statistics.
Fourth of July Creek was sampled with hook and line gear to capture
rainbow trout and Arctic grayling. Trotlines and hoop nets were used at
Mainstem (RM 138.9 -140.1) to collect burbot. Boat electrofishing and
gill nets were used at the remaining three sites to capture resident
fish species. Each site was sampled on three to six occasions over a
peri ad of one to four days. Resident fish over 200 ITD1l in 1 ength were
Flay anchor tagged while smaller fish were marked by clipping the upper
tip of the caudal fin. Catch and recapture information from 1982
indicated that resident fish movement is at a minimum during late July
and early August {ADF&G 1983b). This is important because the CAPTURE
model is only valid for closed populations. Population estimates for
some species were not obtained at all study sites because of insuffi-
cient capture of fish.
The CAPTURE program indicates whether the data set meets the assumption
if a closed population (i.e., no in-or out-migration during the sampl-
ing period). The program selects one model which best fits the data set
out of several possible models. The different models allow for various
effects on capture probability such as behavioral effects (for example,
fish that are hook-shy or will not take a lure after having done so
once). The program also calculates capture probabilities and provides
confidence limits on the population estimates.
Population estimates for all species except burbot were made by a
capture-recapture model from the CAPTURE computer program. Population
estimates for burbot were made using a multiple removal model instead of
the capture-recapture model because of the lack of burbot recaptures.
Although population estimates were attempted at five sites, population
estimates were only able to be calculated for rainbow trout at Fourth of
July Creek and burbot at mainstem Susitna (RM 138.9 -140.1). Popu-
lation estimates of resident fish at Jack Long Creek and at the mainstem
site between RM 131.0 -131.8 were not generated due to insufficient
numbers of fish captured. Population estimates of resident fish at
Slough 8A were also not generated due to low numbers of fish captured
for three species, whi 1 e for two species ( 1 ongnose suckers and round
whitefish) population estimates were inaccurate due to the wrong CAPTURE
models u.sed.
- 8 -
-
....
-
....
-
-~
-
-
-
....
-I
'In addition to the five sites sampled three or more times, population
estimate-s were stopped at seven other sites in 1983 due to insufficient
fish captures during the first sampling occasion. Two of these sites
were sampled for burbot in the mainstem at RM 128.3 -129.3 and at RM
147.0 -147.3. The remaining five sites were in Indian River between
TRM 1.5 -14.0.
2.2.3 Radio telemetry
2.2.3.1 Equipment
Radio telemetry receiving equipment used in this study was developed by
Smith-Root Incorporated in Vancouver, Washington. Receiving equipment
consisted of a low frequency (40 MHz) radio tracking receiver (Model
RF-40) and scanner (Model SR-40), and a loop antenna (Model LA-40).
Radio transmitters manufactured by Smith-Root Incorporated and Advanced
Telemetry Systems (Bethel, Minnesota) were used in the 1983 study.
Advanced Telemetry System radio tags with a nine month life expectancy
were used in rainbow trout. Smith-Root radio tags with a six month life
expectancy were implanted in burbot and a few large rainbow trout.
Advanced Telemetry System transmitters (model BEI 10-35) were cylin-
drically shaped, encapsulated in epoxy, and had flexible 30 em external
antennas. The copper wire antennas were cut down to 15-20 em to make
implanting easier yet still provide a suitable receiving range. The
Advanced Telemetry System transmitters measured 5.6 em in length, 1.2 em
in diameter and had a dry weight of approximately 13.3 gm. The power
source for the transmitters were 3.4 volt lithium batteries providing
life expectancies of 200-270 days, depending on the pulse rate. Trans-
mitter frequencies ranged between 40.600 and 40.770 MHz and had pulse
rates between 1.0 and 2.0 per second. Radio frequencies from 40.680 -
40.700 MHz were not used to avoid interference with transmitting Alascom
radio signals on frequency 40.690.
Smith-Root transmitters were identical to those used in previous resi-
dent fish telemetry studies with exception of the pulse rates (ADF&G
1981d; 1983a;1983b). Smith-Root transmitters used in the 1983 studies
had pulse rates of 3.0 pulses per second and a life expectancy of 180
days.
All radio tags were immersed in cold water (1-5°C) for 48 hours to
ensure they were transmitting properly before they were implanted in
fish.
2.2.3.2 Transmitter implantation
Rainbow tro1Jt used for radio telemetry studies were captured by drift
gill net, boat electrofishing, or hook and line. All burbot used in
radio telemetry studies were captured by boat electrofishing. Based on
personal communications with Carl Burger (USFWS) and experience gathered
from the previous two years of radio telemetry studies, minimum lengths
of rainbow trout and burbot radio tagged were set at 380 mm fork length
and 525 mm total length, respectively. No injured or lethargic fish
- 9 -
were radio tagged. Each fish radio tagged was placed in a 14 gallon
cooler filled with a solution of river water and an anesthetic MS-222
(tricaine methane-sulfonate). After the fish were anesthetized, their
lengths were measured to the nearest millimeter {fork length for rainbow
trout and total length for burbot). Scales were taken from rainbow
trout for aging. All radio tagged fish were marked with Flay anchor
tags to identify them during subsequent recaptures.
With the exception of two rainbow trout, transmitters were surgically
implanted in the coelom using a procedure described in Ziebell (1973).
An incision was made on the midline of the ventral surface midway
between the pectoral and pelvic ffns, and a half capsule of ampicillin
{an antibiotic used to prevent infection) was sprinkled into the body
cavity. The length of the incision for the Advanced Telemetry System
tag was 2.0-2.5 centimeters (em) and a 3.0-3.5 em incision was made for
the Smith-Root tag. The radio tags were inserted anteriorly with the
antenna extended fully toward the posterior of the fish. Incisions were
closed with four to seven individual sutures of commercial silk (Plate
2).
Two rainbow trout received subcutaneous implants of Advanced Telemetry
System radio transmitters using techniques which had been tested on
rainbow trout in the Elmendorf Hatchery. The procedure involved making
a 2.0-2.5 em perpendicular incision through the skin below the posterior
of the dorsal fin. A 1.0 em diameter sharpening steel was used to
tunnel anteriorly beneath the skin and separate the skin from the
muscle. The radio tag was then inserted through the incision under the
skin to the anterior end of the tunneled area. This positioned the
anterior end of the radio tag approximately 3-5 em behind the base of
the fish's head with the antenna trailing out the incision. The inci-
sion was closed with 3-4 silk sutures {Plate 3).
After the surgical implantation of the radio tag, the fish was placed
into a live box and held upright until it regained its equilibrium. The
fish was then held overnight for observation. The following day the
sutures were checked and the transmitter's signal was tested before
releasing the radio tagged fish near the point of capture.
2.2.3.3 Tracking
Biologists radio tracked fish by boa:t, by aircraft and by ground sur-
veys. Radio tracking by boat and ground surveys was conducted in the
mainstem Susitna from Talkeetna (RM 97.0) to Devil Canyon (RM 150.5)
once every 10-14 days from mid-May until mid-October 1983. Ground
tracking was conducted primarily at tributary mouths and in the lower
reaches of tributaries.
Aerial tracking, using methods described in Adult Anadromous Investiga-
tions {ADF&G 1981b), was conducted twice per month from mid-May through
October 1983. In November and December 1983, aerial tracking was
conducted once per month.
-10 -
-
-
-
-
-
-
-
-
-
-
-c
Plate 2. Implanting a radio tag into the abdomen of a rainbow trout.
11
-
-
+l
:::5
0 s.... -+l
3:
0 .a >= -ro s....
ro
4-
0 ~
>=
~
Vl
Q) -~
+l
s....
Q)
"'0 c
:::5
O'l ro
+>
0 -"'0 ro s....
ro -C"l >=
+> c
ro ~ .....-c..
E .......
("'")
Q)
+> ro .....-
0.. -
-
-
12
-
-
-
2.3 Data Recording and Analysis
Biological data and catch data were recorded at relative abundance
study sites as specified in the 1983-84 procedures manual (ADF&G 1984).
Habitat data were also collected at resident fish spawning sites and are
presented in Part 6 of this report. These data included, but were not
limited to, species, length, sex, water velocity, substrate, location,
time sampled, and gear type used. Biological and catch data were also
recorded at sites where population estimates were obtained and where
fish were collected for the radio telemetry study.
Data collected for resident fish relative abundance, population esti-
mates, and radio telemetry were checked for accuracy and completeness
following each sampling trip. Relative abundance data were submitted to
the data processing unit for key punching. Radio telemetry data was
filed for hand compilation at a later date. Printouts of the initial
relative abundance data were returned to the individuals who collected
the data to be rechecked for errors before befng incorporated into the
computer data base for analysis.
Analysis of relative abundance, length frequency and catch per unit
effort data were provided by the data processing group. Population
estimates for resident fish species were computed using the computer
program CAPTURE, described by Otis et al. (1978) and White et al. (1982).
An analysis of variance of juvenile salmon catch rate at the juvenile
salmon study sites was also run on juvenile round whitefish which were
relatively abundant at those sites. Details of the analysis are given
in Part 2 of this report.
-13 -
3.0 RESULTS
3.1 Rainbo~ Trout
3.1.1 Distribution and relative abundance
Four hundred twenty-eight rainbow trout were captured by Susitna Hydro
study groups using various methods between Cook Inlet and Devil Canyon
from May to October 1983 (Table 2). Most of these fish were captured on
the Susitna River above the Chulitna River confluence by hook and line
(43.2%) or boat electrofishing (35.3%).
One hundred sixty-three rainbow trout were caught by a resident fish
study crew at 12 selected sites between the Chulitna River confluence
and Devil Canyon. Most (80.4%) of these fish were captured by boat
electrofishing. The highest catches of rainbow trout at these sites by
all gear types were at Fourth of July Creek (RM 131.1) and Indian River
(RM 138.6) where 46 and 45 fish were caught respectively. Other sites
where relatively high rainbow trout catches were made included Whiskers
Creek Slough (RM 101.2), Lane Creek (RM 113.6) and Portage Creek (RH
148.8).
Two hundred twenty-eight rainbow trout were captured by the resident
fish crew at sites other than the twelve selected sites. Most (78%) of
these fish were captured in Fourth of July Creek between TRM 0.1 and TRM
1.5. In addition to the 391 rainbow trout captured by the resident fish
crew, other Su Hydro study groups captured 37 rainbow trout.
The maximum seasonal catch of 168 rainbow trout (all gear types) was
recorded in late July. Relatively high catches were also recorded in
early (43) and late (41) September (Table 2).
3.1.2 Movement and migration
Twenty-nine rainbow trout were radio tagged at ten different sites on
the Susitna River between the Chulitna River confluence and Devil Canyon
from May 12 to October 5, 1983. Eighty-three percent of these radio
tagged rainbow trout were captured and released at the mouths of tribu-
tary streams. Appendix Table B-1 presents a surrrnary of capture and
biological data for the individual radio tagged fish. Individual
movements of radio tagged rainbow trout during 1983 are presented in
Figures 2-5. During the tracking period, ten radio tagged rainbow trout
moved downriver over 0. 5 mi 1 e, four moved upriver aver 0. 5 mi 1 e and
seven had both downstream and upstream movements over 0.5 mile. The
remaining five radio tagged rainbow trout moved less than 0.5 mile
throughout the tracking period. Eighteen rainbow trout moved downstream
from 0.1 to 26.7 miles (average of 6.9 miles), with most of the down-
stream movement occurring after September 1. Eleven rainbow trout moved
upstream from 0.4 -12.0 miles, with an average upstream move of 2.4
miles.
During 1983, one radio tagged rainbow trout was reported caught by a
sport fisherman. This rainbow trout {648-1.6) was radio tagged on June
7th in Whiskers Creek {TRM 0.1) and recaptured by a sport fisherman on
-14 -
-
.....
-
-
......
(.11
J 1 1
Table 2. Rainbow trout catch on the Susitna River between the Chulitna River confluence and
Devil Canyon, May to October 1983.
Study Group May May June June July July Aug Aug Sept Sept
1-15 16-31 1-15 16-30 1-15 16-31 1-15 16-31 1-15 16-30
Resident Fish Study 17~/ Boat Electrofishing 14 11 5 15 4 5 26 30
Other Gear 6 1 22 21 0 145 2 17 15 9
Juvenile Anadromous
Habitat Studies(JAHS) 0 0 1 0 1 4 1 1 1 2
Downstream
Migrant Trap 0 0 0 2 3 4 3 0 0
Fishwheel sites 1 1 5 1 2 3 1
Total 6 18 38 33 13 168 13 29 43 41
- = No effort
~I One rainbow was captured below the Chulitna River confluence.
l
Oct Total
1-15
24 151
2 240
0 11
12
14 E_/
26 428
b/ Seven rainbows were captured in fishwheels below the Chulitna River confluence. Yentna Station (RM 27.5,
TRM 4.0) capturing three in early July. The remaining four were captured during early June, early
August, late August, and in September at Sunshine Station (RM 79.0).
]
140
Indian Cr. 630"1.0
.r!J
---------718. '· 5
1&.1 4th of Jul Cr. -------------738·1.4 .J
;::! a:
1&.1
>
a:
< z
1-
Cl)
:::1
Cl)
1&.1
.J
:1
a:
w
>
a:
< z
1-
Cl)
:::1
Cl)
130
120
X • TAG AND RELEASE SITE
e FISH LOCATED IN MAINSTEM
SUSITNA RIVER
0 FISH LOCATED IN TRIBUTARY
~ NO LOCATIONS BETWEEN THESE POINTS
1101---~~~--r--,--~--~---r--,---~--~--~-,--~--~--~
MAY 15
Figure 2.
140
130
120
JUNE 1!1 JULY 15 AUG. 15 SEPT. I !I OCT. 1!1 NOV. 1!1 DEC 1!1
1983
~!ovement of five radio tagged rainbow trout in the Susitna
River below Devil Canyon, May to December 1983.
'-.,..../
Moou
Slou~h
63t •1.4
Lane Cruk
__ .. __ .......,,
~--e----0---..-~~-,,~ 1a1-'·'
110
• ____ _.,... . ....-......_ _______ _.Ha-2.0
Whiskers ~···0·0····<:>·······(!)······ ·0················•····-........... 870•1.4 Creek 100
90
0• FISH LOCATED IN
TRIBUTARY OR SLOUGH
X • TAG AHO RELEASE SITE
•• FISH LOCATED IN
MAIHSTEN SUSITH4 R.
80~--~--~--~--~---~--T---~--T---r---r---r----~-,
JUNE 1!1 JULY 1!1 AU~~ SEPtl5 OCT. 1!1 NOV.I5 DEC. 1!1
1'983
Figure 3. Movement of four radio tagged rainbow trout in the Susitna
River below Devil Canyon, June to December 1983.
-16-
-
-
-
-
-
-
-
-
-
-
r
-
w
..J
2
a: w
>
a:
< z
~
Cl)
;)
Cl)
140
Indian R. 819-1.4)( •• ·· ..
4th of July Cr. 619.10 \ .
130-...... •• ._ _______ ... Ftoh trapped behind ice dam, e frozen in ice olonv bonk
··· ...
120
Bear Kill, Ia~ ·._
748 • 1.6 on bonk at Sl. 9 , ·. ·.
La no Cr.
110-
~ -----... .:.. ......... -..... --~-~ • ···-709"1.5 .................. ... '\ ·.
\ 'a..
Whi Ikon Cr. 648• 1.6
0-·-·-·-0-· ... Sport CauQhl at
mouth of Whiskers Cr.
\
roo
90
0• FISH LOCATI!D IN
TRIBUTARY OR SLOUGH
X • TAG AND RI!LEASE SITE
• • FISH LOCATED IN
MAINSTEM SUSITNA R.
"
80~---r---r--~---~.--~--~--~---r-,--T-,--T---~,--~.--~.-
JUNE 15 JULY 15 AUG. 15 SEPT. 15 OCT. 15 NOV.I5 DEC.I5
1983
Figure 4. Movement of six radio. tagged rainbow trout in the Susitna
River below Devil Canyon, June to December
1983.
Figure 5.
150
PORTAGE CR.
14Q-
INDIAN R.
4th of Julv Cr.
j( ........... ~ · ...
••••• • .. 660-3.0 .. . .. ·······•····
X: TAG AND RELEASE SITE
e• FISH LOCATI!D IN
MAIHSTEM SUSITNA R.
,1(~749•1.0
~. \'"· ...... _______ • 607"1.!
'· \\
.._ -·"': .. ·--····-608 "I.! \ ....... ______ 600• 1.<
\,
\,
'-----62.11-u
13~~--r---~,--T---~--~---r---,
SEPT. 15 OCT, 15 NOV.I5
1983
Movement of eleven radio tagged 1·ainbow trout in the Susitna
River below Devil Canyon, September to December 1983.
-17-
DEC. 15
--·-----------------·-··----~----~-----------------
August 8th at the mouth of Whiskers Creek (TRM 0.0). The angler report-
ed that the rainbow trout was in excellent condition and that the
sutured incision had healed nicely. Radio tracking data showed that
this rainbow trout did move short distances above and below the tagging
site before being recaptured, but it largely stayed in the same general
area for summer rearing.
Three of the 29 radio tagged rainbow trout provided little or no move-
ment and migration data. One rainbow trout (668-1) radio tagged by the
under-the-skin method either dropped its transmitter or died in Moose
Slough (RM 123.5). When the slough 1 s water became clear during Septem-
ber, neither the rainbow trout or transmitter could be found. Thereaf-
ter, rainbow trout radio tags were surgically ·implanted. Only one
rainbow trout (628-2) was presumed to have been injured from the tagging
or capture process during 1983. Immediately following its release, this
rainbow trout moved rapidly downriver and was extremely lethargic when
recaptured by boat electrofishing 20 days later. A third radio tagged
rainbow trout (659-1.8) was ·injured when it was accidentally recaptured
by boat electrofishing and it also moved rapidly downstream. With the
exception of these three rainbow trout, it appeared that the remaining
radio tagged rainbow trout exhibited normal behavior after being radio
tagged.
Flay anchor tagged rainbow trout also provided information on rainbow
trout movements. During 1983, 275 rainbow trout were Flay anchor tagged
and 35 recoveries were made. Five rainbow trout were recovered at the
same site where they were tagged. Sixteen rainbow trout were recovered
within 5.0 miles of their tagging site. The remaining 14 rainbow trout
were recaptured an average of 18.7 miles from where they were tagged.
Ninety-four percent of the recaptured rainbows were recovered in or at
mouths of tributaries such as Fourth of July Creek (12, RM 131.1) and
Clear Creek (4), a tributary 6.0 miles up the Talkeetna River (RM 97.0).
The most rapid movement recorded for a rainbow trout in 1983 was an
upstream movement of 37.4 miles in 40 days during the spring. The
maximum movement documented for a 11 rainbow trout tagged to date was
53.0 miles by a rainbow trout tagged on July 19, 1982 at Jack Long Creek
(RM 144.5) and recaptured at Clear Creek (TRM 0.0) on June 30, 1983.
3.1.3 Population estimates
The population estimate of rainbow trout in Fourth of July Creek between
TRM 0.0-0.8 using the behavioral model from the CAPTURE computer program
was determined to be 107 rainbow trout. The standard error of this
estimate was 15.10 and the 95% confidence interval was from 82-137. The
catch during the three day sampling period was 42, 22 and 18 respec-
tively; in addition, eight fish were recaptured.
3.2 Burbot
3.1.2 Distribution and relative abundance
A total of 163 burbot were captured in the Susitna River between the
Chulitna River confluence and Devil Canyon during 1983 (Table 3). Most
(78 of 118) of the burbot captured by resident fish biologists were
-18 -
-
-
-
......
lO
-.. I
Table 3. Burbot catch on the Susitna River between the Chulitna River confluence and
Devil Canyon, May to October 1983.
Study Group May May June June July July Aug Aug Sept
1-15 16-31 1-15 16-30 1-15 16-31 1-15 16-31 1-15
Resident Fish Study
Boat Electrofishing 7 5 3 4 13 10 0 10
Other Gear 0 16 0 6 13 0 5 0 0
Juvenile Anadromous
Habitat Studies(JAHS) 0 2 0 5 2 2 2 4 1
Downstream
Migrant Trap 1 8 3 1 1 4 4 0
Fishwheel sites 0 0 0 4~/ 0 0 1£/
Total 0 26 13 17 20 20 21 8 12
- = No effort
a/ One burbot was captured in a fishwheel at Yentna River Station (RM 27.5, TRM 4,0).
~/One burbot was captured in a fishwheel at Sunshine Station (R}I 79.0).
Sept Oct Total
16-30 1-15
8 2 62
16 0 56
0 0 18
0 22
5
24 2 163
caught in the mainstem Susitna River or side channel sites. Burbot were
most abundant at mainstem RM 139.6 (18 burbot), mainstem RM 102.5 (16
burbot), and mainstem RM 147.0-148.0.
3.2.2 Movement and migration
From August 18 to September 3, 1983, four burbot were radio tagged on
the Susitna River between RM 113.6 and RM 147 .5. A summary of 1983 data
for radio tagged burbot is presented in Appendix Table B-2.
Radio tagged burbot movements were variable (Figure 6). One radio
tagged burbot (610-3) remained within 3.6 miles of its capture site for
three months. Two other radio tagged burbot (639-3 and 720-3) moved
slowly downstream after their release 11.9 and 13.6 miles, respectively,
and remained at these locations. Between its release on September 1 and
October 21, radio tagged burbot (670-3) moved 36.5 miles downstream.
Three radio tagged burbot also made small movement upstream. Burbot
(610-3) moved upstream 2.5 miles, burbot (720-3) moved upstream 0.6
miles, and burbot (670-3) moved upstream 0.4 miles.
One hundred eight burbot were Floy anchor tagged and three burbot were
recaptured in 1983. Movements exhibited by these burbot were minimal.
A 11 three Fl oy anchor tagged burbot were recaptured with 0.1 mi 1 es of
their tagging location.
3.2.3 Population est·imates
The burbot population estimate for the mainstem Susitna River between RM
138.9-140.1 was 15 burbot with a standard error of 4.18 and a 95%
confidence interval of 13-24 burbot. The catch was 6, 1, 4 and 2
respectively for the four days sampled; no burbot were recaptured.
3.3 Arctic Grayling
3.3.1 Distribution and relative abundance
A total of 1,165 Arctic grayling were captured on the Susitna River
between the Chulitna River confluence and Devil Canyon in 1983 (Table
4). Arctic grayling were most abundant at a mainstem site (RM
137.3-138.3) where 195 Arctic grayling were captured. Other sites where
more than 60 Arctic grayling were captured are Lane Creek (RM 113.6),
Indian River (138.6) and Portage Creek (RM 148.8). Catches of Arctic
grayling were high in the spring at Whiskers Creek Slough (RM 101.2) and
at RM 150.1 in the mainstem. During the summer, most Arctic grayling
were captured in late May -early June and in September. The maximum
Arctic grayling catch by all gear types (307 fish) was recorded in late
September.
3.3.2 Movement and migration
Seven hundred sixty-five Arctic grayling were Flay anchor tagged and
forty-one Arctic grayling were recaptured in 1983. Sixty-one percent of
the recovered fish were from fish tagged in 1981 or 1982. Recaptured
Arctic grayling movements ranged from 0.0 to 29.4 miles with an average
-20 -
-
-
-
-
-
-
-
I""'
-
.....
-
150
720-3.0
125 670-3.0 )( •••• ·•.
)( ·· .. 1&.1 -·--....... •---·-·-------=:..---610·3.0
..J ·.
~ . . a:: ·.
1&.1 .
> ................. ...•
a::
c( 75
z .....
C(J
::;:)
Cl) 50
X c TAG AND RELEASE SITE
e:: Fl SH LOCATED IN
25 MAINSTEM SUSITNA R.
JULYI5 AUG. 15 SEPT. 15 OCT. 15 NOV. 15 DEC.I5
1983
Figure 6. Movement of four radio tagged burbot in the Susitna
River below Devil Canyon, July to December 1983.
-21-
N
N
~--'1
Table 4. Arctic grayling catch on the Susitna River between the Chulitna River confluence and
Devil Canyon, May to October 1983.
Study Group May May June June July July Aug Aug Sept Sept
1-15 16-31 1-15 16-30 1-15 16-31 1-15 16-31 1-15 16-30
Resident Fish Study 13~/ Boat Electrofishing 126 72 19 89 57 12 169 299
Other Gear 0 29 7 17 6 5 4 7 2 8
Juvenile Anadromous
Habitat Studies(JAHS) 0 0 0 0 9 3 6 3 0 0
Downstream
Migrant Trap 1 5 13 8 4 5 1 0 0
Fishwheel sites 1 2 0 1 l 2 5
Total 0 166 139 104 42 102 73 25 176 307
-"" No effort.
Oct Total
1-15
35 1, 014
1 86
0 21
37
12 E_/
36 1,170
!I Two Arctic grayling were captured below Chulitna River confluence.
b/ Three Arctic grayling were captured in fishwheels at Sunshine Station (RM 79.0). One was caught in late
August and two were caught in September.
I J " I I J I
-
-
movement of 5.4 miles. About ha1f (19) of the 43 recaptured Arctic
grayling were recaptured at their tagging sites. -Another six Arctic
grayling were recovered within 5.0 miles of their tagging sites. The
remaining 18 Arctic grayling recaptures moved an average of 12.5 mi1es
from their tagging locations. Thirty of the 43 recoveries were made in
tributaries or at tributary mouths. Eight Arctic grayling were
recaptured at Fourth of July Creek (RM 131.1) and seven at Lane Creek
(RM 113.6).
3.4 Round Whitefish
3.4.1 Distribution and relative abundance
A total of 4,917 round whitefish were captured in 1983 on the Susitna
River between the Chulitna River confluence and Devil Canyon (Table 5).
Many of the round whitefish were juveniles (<200 mm) captured by two
downstream migrant traps at RM 103.0.
The analysis of variance on the round whitefish catch at juvenile salmon
rearing sites (JAHS sites) , which was almost all juvenile fish, showed
that time of year had a significant effect on the catch rate (Part 2 of
this Report). Juveniles were captured mainly in July and August at the
JAHS sites; however, sampling efforts in their preferred habitat (turbid
side sloughs and side channels) was minimal in June. The fish were in
the river and moving earlier than July as evidenced by the catch at the
downstream migrant traps (also almost all juveniles) in June.
Adult round whitefish (: 200 mm) were most abundant at a mainstem site
between RM 147.0-RM 148.0. Other sites where over 100 adult round
whitefish were captured were Slough 8A (RM 125.3), a mainstem site
between RM 137.3-138.3, Indian River (RM 138.6), Jack Long Creek (RM
144.5), and Portage Creek (RM 148.8). Boat electrofishing catches of
round whitefish were the highest in early September. Relatively high
catches were also made in early June, late July, late September, and
October.
3.4.2 Movement and migration
During 1983, 1,081 round whitefish were Floy anchor tagged and 73 round
whitefish were recovered. Most of the 36 recoveries were from round
whitefish tagged in 1982. The maximum downstream movement for round
whitefish was 69.5 miles and the maximum upstream movement was 17.0
miles.
Thirty round whitefish were recaptured at sites where they were orig-
inally tagged. Twenty-seven were recaptured within 5.0 miles of their
tagging locations. The remaining 16 tagged round whitefish moved an
average of 18.5 miles downstream before being recaptured.
Thirty-three of round whitefish tag recaptures were made at tributary
mouths and two were made 3.0-5.0 miles upstream of tributary mouths.
Another 29 were r~covered in the mainstem and the remaining nine were
recovered in sloughs.
-23 -
N
~
Table 5, Round whitefish catch on the Susitna River between the Chulitna River confluence and
Devil Canyon, May to October 1983.
Study Group May May June June July July Aug Aug Sept Sept
1-15 16-31 1-15 16-30 1-15 16-31 1-15 16-31 1-15 16-30
Resident Fish Study sa~/ Boat Electrofishing 138 60 106 244 100 8 270 174
Other Gear 0 6 21 0 4 3 0 0 1 6
Juvenile Anadromous
Habitat Studies(JAHS) 0 0 0 0 307 99 172 41 9 1
Downstream
Migrant Trap 5 56 871 1,539 295 66 59 9 1
Fishwheel sites 2 4 0 3 0 23 16
Total 0 69 217 935 1,956 644 338 131 305 182
- = No effort.
~/ Three round whitefish were captured below the Chulitna River confluence.
Oct Total
1-15
161 1,319
2 43
0 629
2,901
48 .!?_/
163 4,940
£! Twenty round whitefish were captured below the Chulitna River confluence. Fishwheels at Yentna Station
(RM 27.5, TRM 4.0) captured two in August. Fishwheels at Sunshine Station(RM 79.0) captured one in early
June; one in late June, six in August, and 10 in September.
-J .J
,j
-
3.5 Humpback Whitefish
3.5.1 Distribution and relative abundance
Eight hundred twenty humpback whitefish {Coregonus idschian) were
captured in the Susitna River during 1983 with most 83.5% being
captured above the Chulitna River confluence {Table 6). Downstream
migrant traps (RM 103.0) and fishwheels captured the majority (92.6%) of
the humpback whitefish.
A total of 466 juvenile humpback whitefish (< 200 mm) were captured by
two downstream migrant traps. The maximum catch of humpback whitefish
at the downstream migrant traps occurred during late July. Relatively
high catches were also recorded during early July and early August.
Fishwheels captured 293 adult humpback whitefish. Fishwheels at Yentna
River station (RM 28.5, TRM 4.0) captured 60.8% of the humpback white-
fish caught by fishwheels. The maximum seasonal humpback whitefish
catch (137 fish) by fishwheel was recorded in late A~gust.
Boat electrofishing catches of humpback whitefish (36) were most numer-
ous at the mouth Slough 8A ( RM 125.3}. Gi 11 net and hoop net humpback
whitefish catches (14) were greatest in Slough 6A (RM 112.3). JAHS
crews captured nine juvenile humpback whitefish in Slough 22 (RM 144.3)
with beach seines.
3.5.2 Movement and migration
In 1983, 329 humpback whitefish were tagged with Fl oy anchor tags.
Three tagged humpback whitefish were recaptured in 1983. One recaptured
humpback whitefish moved upriver 17.0 miles in two days. A second
tagged humpback whitefish moved downriver 11.0 miles in 43 days. The
third humpback whitefish, tagged in 1982, moved downriver 8.7 miles in
one year.
3.6 Longnose Suckers
3.6.1 Distribution and relative abundance
A total of 713 longnose suckers were captured in the Susitna River in
1983 {Table 7). All but 20 of these were captured in the Susitna River
between the Chulitna River confluence and Devil Canyon.
Boat electrofishing longnose sucker catches were most abundant at Slough
8A {RM 125.3), Lane Creek (RM 113.6), Fourth of July Creek {RM 131.1), a
mainstem site between RM 147.0-RM 148.0, and Portage Creek {RM 148.8)
during late July and early August.
Juvenile longnose suckers (< 200 mm) were captured incidentally by beach
seines and backpack electroshocker at mainstem and slough sites by JAHS
crews. Longnose sucker juveniles captured at JAHS sites were most
abundant at Mainstem II (RM 114.4). The downstream migrant traps at RM
103.0 also captured 111 juvenile longnose suckers.
-25 -
Table 6. Humpback whitefish catch on the Susitna River between the Chulitna River confluence
and Devil Canyon, May to October 1983.
Study <.roup May May June June July July Aug Aug Sept Sept
1-15 16-31 l-15 16-30 1-15 ~ 1-15 16-31 1-15 16-30
Resident Fish Study
Boat Electrofishing 0 0 18 2 0 3 4
Other Gear 0 0 14 0 0 0 0 0 c 0
Juvenile Anadromous
Habitat 5 tudies !,IAHSl 0 0 0 0 9 0 0 0
Downstream
Migrant Trap 0 0 11 93 228 92 40 2 0
Fishwheel sites 6 33 81 15 137 18
Total c 0 l7 19 142 328 llO 177 23 4
-• No effort.
Oct Total
.!..:.!2.
0 36
0 14
0 ll
466
29#1
0 820
!/ A total of 235 humpback whitefish were captured below the Chulitna River confluence, Yentna Station
fishwheels (RM 27.5, TRK 4.0) captured 178 and Sunshine Station fishwheels (RM 79,0) captured 57. Yentna
Station humpback whitefish catch by two week periods from early July to early September was 28, 59, ll,
76, and 4, respectively, Catch at Sunshine Station by two week periods from early June to early
September was 3, 1, 0, 1, 2, 45, and 5, respectively.
-.., ~o effort
Three fish were captureC below the Chulitna River confluence with one being captured in late May and two
in early June.
Seventeen fish were captured below the confluence with Yentna station (RM 27.5, TRM 4.0) capturing two in
early July, six in late July and one in early Septembere The remaining nine fish were captured at
Sunshine. station (RM 79.0) with one being captured in early June, two in early July, one in late July,
three in late August, and one in early Septembber.
-26-
-
-
-
-
~
-
i ~:
3.6.2 Movement and migration
During 1983, 467 longnose suckers were tagged with Flay anchor tags and
24 longnose suckers were recaptured. Six longnose suckers were
recaptured at their tagging sites and another seven were recaptured less
than 5.0 miles from their tagging sites. Six tagged longnose suckers
moved downriver (5.0 to 47.6 miles) and five moving upriver (5.0 to 36.9
miles). The average movement of the 11 fish which moved over 5.0 miles
was 18.5 miles.
The most rapid movement recorded for a tagged longnose sucker was 25.5
miles over a period of 15 days. This longnose sucker was tagged on June
6 at Slough 6A (RM 112.3) and recaptured on June 21 at mainstem RM
137.8.
3.7 Other Species
3.7.1 Dolly Varden
A total of 47 Dolly Varden were captured in the Susitna River in 1983.
Most (89%} of these were captured in the Susitna River between the
Chulitna River confluence and Devil Canyon. The largest Dolly Varden
catches in this reach of river were made at the mouth of Portage Creek
(30%) and at the mouth of Indian River (19%).
During 1983, 12 Dolly Varden were tagged and two were recaptured. One
fish was recaptured at Kashwitna River (RM 61.0) and the other re-
captured at Clear Creek, a tributary of the Talkeetna River (RM 97 .0,
TRM 6.0). Both fish had moved upriver {2.5 miles and 10.0 miles,
respectively) from their tagging site.
3.7.2 Threespine stickleback
A total of 1,834 threespine stickleback (Gasterosterus aculeatus
Linnaeus) were captured in 1983. Downstream migrant traps at RM 103.0
captured 1,601 and the remaining fish were captured incidentally by JAHS
crews with beach seines or backpack electroshockers. Among the JAHS
sampling sites threespine stickleback were most abundant at Slough 5 (Rt·1
107.6}. Most threespine stickleback young of the year were captured in
early August.
3.7.3 Arctic lamprey
A total of 69 Arctic lamprey (Lampetra japonica Martens) were captured
in the Susitna River in 1983. Forty-four were captured by the down-
stream migrant trap at RM 103.0. Arctic lamprey catches at the down-
stream migrant traps were highest in late May and late June. The
remaining Arctic lamprey were captured with a backpack electroshocker at
Chase Creek {RM 106.9} in late August.
-27 -
4.0 DISCUSSION
4.1 Rainbow Trout
The 1983 studies provided considerable information about the distribu-
tion of rainbow trout in the Susitna River between the Chulitna River
confluence and Devil Canyon. The deployment of radio tags yielded over
6 months of data on the distribution of rainbow trout and gave new
insights into their movement which previously had been hypothesized from
catch per unit effort data. In Part 6 of this report, the distribution
of this species by macro and microhabitat is described. Although our
data is somewhat limited in the early spring, the seasonal distribution
of rainbow trout within the Susitna River system is reasonably well
documented. The following discussion includes descriptions of what we
have learned about the life history of this species and its vulnerabil-
ity to altered conditions in the mainstem Susitna River. We have also
established index areas (Table 1) and have estimated the population size
of rainbow trout in one of the tributaries {Fourth of July Creek)
important to this species.
Rainbow trout catch rates in 1981 and 1982 in the mainstem Susitna
rapidly dropped off after June suggesting movement out of mainstem areas
and probably into tributaries. This movement was verified by random
sampling of the upper reaches of tributaries during 1983 and reinforced
by studies of radio tagged fish during the summer. The highest catches
of rainbow trout were recorded in Fourth of July Creek where significant
spawning activity was documented. Minnow trap catches of juveniles
rainbow trout during 1983 was the highest recorded since the onset of
these studies in 1981. Spawning occurred in late May-early June as
suggested by the capture of pre-and post-spawned adults and movements
into Fourth of July Creek by two radio tagged fish. Movements of radio
tagged fish out of this tributary after spawning suggests that at least
some of the fish will emigrate from their spawning tributaries to other
forage areas.
Random sampling for rainbow trout was conducted during 1983 in most
tributaries of the Susitna River between the Chulitna River confluence
and Devil Canyon. Fourth of July Creek had the highest concentration of
rainbow trout as reflected by the CPUE. These data suggest that adult
rainbow trout move into tributaries during the spring to spawn and some
of these fish remain in the tributaries throughout the summer.
Examination of the 1 imiting factors during the 1 ife cycle of rainbow
trout will help evaluate the vulnerability or the enhancement potential
of this species under postproject conditions. The comparatively small
numbers of juvenile rainbow trout collected, during the three years of
this study suggests reproduction could be limiting or survival of
juvenile is very low. Our survival data suggests this species shows a
relatively high turnover rate compared with other species but not
necessarily a younger age of maturity.
Catch rates of juvenile (<200 mm, Age 3) rainbow trout in Indian River
and Portage Creek have been very low suggesting poor rearing or low
spawning success in these major tributaries {ADF&G 1981c; 1983b). In
-28 -
-
...
-
-
-
-
-
....
contrast, the catch of juvenile rainbow trout in Fourth of July Creek in
1983 was the highest recorded since resident fish studies began in 1981.
Because so few juvenile rainbow trout have been captured in the mainstem
it appears that the juveniles primarily rear in the upper reaches of
tributaries and move little.
Radio tagged rainbow trout using the mainstem Susitna for summer rearing
were often located near tributary mouths, especially from August through
mid-September. The association of rainbow trout with tributaries during
this period coincides with the timing of spawning chum and pink salmon
{Barrett et al. 1984). The concentration of rainbow trout at tributary
mouths and their periodic ascents into tributaries is believed to be due
to the abundance of food {salmon eggs) in these areas. Rainbow trout,
presumably feeding on sa 1 man eggs, were observed being chased from
spawning redds by male chum salmon {Part 6 of this report). The abnor-
mally expanded ventral body cavities of other rainbow trout captured in
August and September in both 1982 and 1983 a 1 so pro vi de evidence of
rainbows foraging on salmon eggs.
In addition to the concentration of rainbow trout at tributaries during
summer periods, radio tagged rainbor trout were observed holding in
several sloughs [Moose {RM 123.5), A {RM 124.6), 8A (RM 125.3), and 9
{RM 128.3)]. The use of these sloughs by radio tagged rainbow trout in
August and September coincided with the presence of spawning chum salmon
in these same sloughs (Barrett et al. 1984). Although high turbidities
prevented actual observation in most of these instances, it is suspected
that these fish were in the sloughs to feed on salmon eggs. This
hypothesis is substantia:f-ed in one case; one radio tagged rainbow trout
was observed in Slough A milling around spawning chum salmon in an area
of clear water {Barry Stratton pers. comm.)
Areas of the mainstem Susitna River not influenced by tributaries or
sloughs were also used during summer months by radio tagged rainbow
trout. The mainstem, however, appears to be more of a migration path
between tributaries and sloughs rather than a holding area during the
open water season.
By mid-September, all radio tagged rainbow trout in tributaries had
descended to the mouths. This movement supports the hypothesis that
most adult rainbow trout outmigrate from tributaries during fall to
overwinter in the mainstem {ADF&G 1983b). The hypothesis is further
supported by the increased catch rate of rainbow trout at tributary
mouths in September. Rainbow trout in the middle Susitna River are
vulnerable to sport fishing during these fall outmigrations. Local
anglers take advantage of the outmigration at the mouth of Indian River
(RM 138.6) each fall. As the Susitna River basin continues to develop,
the rainbow trout population may decline from the increased fishing
pressure.
Beginning in October, radio tagged rainbow trout began to move away from
tributary mouths into the mainstem Susitna River. By early December
only six of 20 radio tagged rainbow trout were within the influence of a
tributary. Because of the difficulty of characterizing winter habitat,
-29 -
------·------------------
we are uncertain why radio tagged rainbow trout seek mainstem areas in
the winter.
The recaptures of six Floy anchor tagged rainbow trout at Clear Creek in
the Talkeetna River drainage suggests that this tributary may be an
important summer rearing area for adult rainbow trout. Tag deployment
data indicated that these rainbow trout also overwinter in the mainstem
Susitna River between RM 77.0 and RM 87.0.
The final activity pursued during the 1983 studies was the establishment
of index areas to monitor annual changes in the populations of rainbow
trout and other species. Population (density) estimates were planned
for five sites but were found to be unfeasible because of low capture
rates. Only the lower reach of Fourth of July Creek had sufficient
numbers of rainbow trout recaptures to generate a population estimate
(107 fish greater that 150 mm FL.). A discussion of the methodological
·problems of estimating population sizes for resident fish in this system
and other areas are included in Appendix D. Catch per unit effort data
will probably have to suffice as an estimator of site specific densities
of resident species. An examination of the annual recovery of tagged
fish as a percentage of tags deployed provides a more robust perspective
of the population of rainbow trout in this reach. A true 11 popu1ation 11
estimate cannot be made from this data because of lack of randomness of
the sample over the entire reach, mortality between years, emigration,
etc. Neverthelesst our tagging efforts have been broadly distributed in
habitats associated with the mainstem Susitna in this reach. The
movements of radio tagged fish also suggests that our samples include
fish from throughout the basin rather than representing only the specif-
ic locale where they were collected. Of 92 rainbow trout tags deployed
in this reach in 1981, only seven out of 221 rainbow trout captured in
1982 were tagged recaptures from fish tagged in 1981. If no mortality
or recruitment were considered, this would provide an estimate of about
2,581 rainbow trout. Using 1982 2tnd 1983 data the population estimate
for rainbow trout (5,057) is low.-However, our mortality estimate for
rainbow trout suggests high mortality of the post-spawning fish t which
when coupled with recruitment would substantially reduce this estimatet
probably by over half. This must be tempered with the non-randomness of
the sampling effort, which probably eliminated significant portions of
the population from sampling effort and decreased the estimate.
Y In 1983, 10 out of 365 rainbow trout ( > 200 mm) recaptures were
tagged in 1982. A total of 151 rainbow trout were tagged on the Susitna
River in 1982 between the Chulitna River confluence and Devil Canyon.
The population estimate equation used was:
where N = Population estimate
X = Number of fish tagged in preceding year
Y = Number of fish tagged in current year
Z = Number of recaptures made in current year
from fish tagged in preceding year
-30 -
-
....
-
-
-
....
....
....
-
This order of magnitude estimate provides an approximation of the extent
of the resource at stake in this basin and can be used as a starting
point to assess potential management concerns if increased sport fishing
pressure follows development of the hydroelectric project.
Current data indicates that rainbow trout in the Susitna River between
the Chulitna River confluence and Devil Canyon use three primary tribu-
taries for spawning [Whiskers Creek {RM 101.4), Lane Creek {RM 113.6)
and Fourth of July Creek (RM 131.1)]. It is not known why only a few
rainbow trout spawn in the 1 a rger Indian River ( RM 138.6) and Portage
Creek (RM 148.8) except that these rivers are close to the northernmost
range of the species. With a better knowledge of rainbow trout spawning
or rearing limitations in these two systems, possible enhancement of
habitat within these tributaries could be made to increase rainbow trout
populations.
While few rainbow trout have been captured during the springs of 1981 to
1983, data shows that spawni 119 primarily occurs between 1 ate May to
mid-June and that both sexes spawn after Age 5+.
The occurrence of so few juvenile rainbow trout ( < 100 fish captured or
observed) in the mainstem or at tributary mouths suggests that spawning
probably occurs in the upper reaches of tributaries. The low numbers of
juveniles found in mainstem areas further implies that primary rearing
of juvenile rainbow trout occurs in the upper reaches of tributaries.
Catch data from the upper reaches of three tributaries [Fourth of July
Creek (RM 131.1, TRM 0.0-2.3), Indian River (RM 138.6, TRM 0.0-14.0) and
Portage Creek (RM 148.8, TRM 0.0-10.0)] indicates a higher incidence of
rainbow trout spawning in Fourth of July Creek than in the other two
tributaries.
A further indication of the importance of Fourth of July Creek to
rainbow trout spawning was made by examining the movements of two radio
tagged rainbow trout captured and tagged in mid-May 1983 at the mouth of
Fourth of July Creek. After their release, both fish migrated to the
upper reaches of the tributary between TRM 1.0 and TRM 1.5. The radio
tagged rainbow trout were prevented from movin~ upstream beyond TRM 1.8
by an apparent fish barrier; two waterfalls (2.1 and 3.9 meters high
respectively) that are located back-to-back in the main channel with no
plunge pool between them. No juvenile or adult resident fish or salmon
were observed or captured above this barrier. Presumably both of these
rainbow trout spawned between TRW s 1.0 and 1.5 in early June. After
spawning, one of these fish dropped out of Fourth of July Creek and
moved upriver into Indian River between late June and mid-July for
summer rearing.
With habitat enhancement, Fourth of July could potentially become a
greater producer of rainbow trout. While there are numerous pools for
juvenile rearing in Fourth of July Creek from TRM•s 0.6-1.8, there are
few areas that appear to have suitable spawning gravel. Suitable
spawning habitat does exist, however, above the barrier. Therefore a
potential mitigation measures to enhance rainbow trout in the Susitna
River between the Chulitna River confluence and Devil Canyon would be to
-31 -
remove the fish barrier at TRM 1.8 and allow rainbow trout to migrate
further upstream and utilize the abundance of spawning gravel which
exists there.
Rainbow trout growth and length data also suggest that reproduction is
the major limiting factor to rainbow trout populations in the Susitna
River. Age-length data taken during 1981-83 show rainbow trout are fast
growers over all age classes (AOF&G 1981c; 1983b) Growth of Susitna
stocks have been found to be similar to other nothern populations (ADF&G
1983f). A 1 though Susitna rainbow trout are relative fast growers, they
appear to have a short life span. Since 1981, the largest and oldest
rainbow trout captured was 612 mm in fork length and nine years old.
Using data from fish captured by hook and line and boat electrofishing,
the survival rate for rainbow trout in the Susitna River was found to be
only 33.3%. Reasons for the low survival rate are not known, however,
hatchery personnel at Elmendorf report that mortalities of post-spawning
male and female rainbow trout are exceedingly high, as do Scott and
Grossman (1973). This may also be due to low egg and juvenile survival.
In addition, another possible reason for the low survival rate of
rainbow trout may be high overwintering mortalities. High winter
mortalities of rainbow trout are most likely to result from physical
catastrophes such as dewatering, collapsed snow banks, and anchor ice
formation (Needham and Jones 1959; Needham and Slater 1945). Reimer
(1957) found that physical catastrophes caused more mortalities than the
lack of food availability.
4.2 Burbot
Burbot occupy the turbid waters of the mainstem Susitna and apparently
rear and spawn in reaches directly influenced by mainstem flow. In the
Susitna River, this species appears to avoid clear water areas although
it is found over a broad range of conditions in other areas. Because of
winter effects of regulated flow on water temperature and the potential
for clearing of the mainstem Susitna, this species has a relatively high
potential to be adversely affected by habitat alterations although
increases in prey species may be a net benefit. Because alternative
modes of operation of the project will probably influence turbidity
levels appreciably, and the behavioral response to turbidity changes is
the most likely effect on this species, we have focused our studies on
monitoring this species to determine the extent of the resource at risk.
The presence of juveniles in this reach suggests spawning occurs in this
area but our efforts at data collection during the spawning season in
January have not been sufficient to locate specific spawning sites. The
spawning does not appear to be as important or concentrated as in major
spawning areas in the lower river, such as the mouth of the Deshka
River.
Burbot catches between 1981 and 1983 indicate that burbot seem to prefer
mainstem sites or slough mouths rather than tributary mouths or tribu-
taries in the Chulitna River confluence to Devil Canyon reach. In this
reach, burbot are found more often in backwater areas, however they have
also been captured in fast, shallow water.
Burbot movements in the Susitna River occur primarily before and after
their spawning period in late January. Data collected during three
-32 -
-
-
-
-
-
-
-
years (1981-83) of monitoring 20 radio tagged fish show that instream
migrations begin in September and last until March (ADF&G 1983b; 1983e).
While most of the radio tagged burbot moved little during the spawning
period, some have moved over ten miles with one moving 113.6 miles in
1982-83. This movement has been discussed previously in the 1982-83
winter report and fish tagged in 1983 show similar behavior (ADF&G
1983e). Although most movement information for burbot to date has been
from fish radio tagged during the fall, one fish was monitored through-
out the surmner in 1983. This burbot (610-3.0) moved only 3.6 miles from
its tagging site between July 19 to October 21 {Figure 6).
It appears that there is an adequate food supply for burbot in the
mainstem Susitna during the summer. During 1982 and 1983, electro-
fishing crews captured few burbot near spawning salmon compared to other
resident fish species. Although necropsied burbot have been found with
salmon eggs in their stomachs, Morrow {1980) states that burbot are an
omnivorous carnivore with a strong preference for fish.
A burbot population estimate study conducted in a one-mile reach of the
mainstem estimated a population of 15 burbot. Because no recaptures
were made, the confidence in this value is very limited. Although the
removal method used in the estimate is quite robust, the low probability
of capture makes the methodology somewhat suspect. A very high trap
avoidance appears to be a characteristic of this species. This aspect
of burbot behavior also limits the value of interpreting our annual tag
recoveries with respect to population estimates of the entire reach.
The very small percentage of tags deployed that were recovered suggest
either high avoidance to recapture, high mortality of tagging, or very
large populations. Monitoring changes in population by catch per unit
effort appear to be the most reliable method for long term study of this
species.
Catch data from 1981-83 shows few adult burbot captured in the Susitna
River above the Chulitna River confluence compared to below the conflu-
ence (ADF&G 1981c, 1983b). In addition, relatively few juvenile burbot
have been captured in the reach above the Chulitna River confluence.
This leads us to believe that few burbot spawn in the Susitna River
between the Chulitna River confluence and Devil Canyon. During inten-
sive sampling by Juvenile Anadromous Habitat Studies (JAHS) in 1983 at
35 sites above the confluence, only 18 juvenile burbot were captured by
beach seining or by backpack electroshocking. Catch data from the
downstream migrant traps at RM 103.0 in 1982 (70 juvenile burbot) and
1983 {22 juvenile burbot) also supports the hypothesis that little
spawning occurs above the confluence.
The exact spawning locations and numbers of burbot spawners in the reach
above the Chulitna River confluence is not known. It is speculated that
burbot spawning in this reach occurs primarily at the mouths of sloughs
and in deep backwater areas influenced by ground water. Support for
this theory are the juveniles found at Slough 9 in 1982, and the high
numbers of adult fish found in deep backwater areas compared to· other
types of habitat. In addition, prior winter studies on the Susitna
below the confluence suggest that spawning and rearing burbot seek areas
of upwelling. This behavior could apply to areas above the confluence
as well (ADF&G 1983e).
-33 -
Age-length data for burbot captured between 1981 and 1983 show that
Susitna River burbot grow rapidly up through Age 4 and then their growth
rate slows to approximately 40 mm a year (ADF&G 1983e). To date, the
oldest resident fish captured in the Susitna River was an Age 15 burbot.
Pooled age-length data from burbot captured between 1981 and 1983 showed
that the survival rate is relatively high (70.5%). To pool the data in
determining the instantaneous survival rate, we assumed that the
survival rate was constant between years sampled. Since burbot live
long and the mainstem where they reside is relatively stable between
years, we believe the assumption was met.
Morrow (1980) states that burbot have a high reproductive capacity and
their survival rate is quite high. Therefore the limiting factor for
the burbot population in the Susitna River between the Chulitna River
confluence and Devil Canyon may be the amount of acceptable habitat for
spawning or rearing, or lack of food. Burbot production in this reach
may be 1 imited by one or several of these factors. Burbot are less
numerous and appear to be slightly smaller for a given age class in this
reach of river in comparison to the reach of river downstream of the
Chulitna confluence (ADF&G 1981c, 1983b, 1983e). Susitna River burbot
appear to grow faster than burbot studied in interior Alaska by Chen
(1969). The mean total length of Age 5 burbot in the Susitna River was
453 mm and Chen reported a mean total length of 355 mm for the same age
class in interior.Alaska.
4.3 Arctic Grayling
Arctic grayling provide local sport fisheries at tributary mouths in
this reach of the Susitna. Our data suggest that overwintering in
mainstem areas may be of major importance for this species. Summer
rearing of Arctic grayling in the mainstem Susitna appears to be limited
to younger age class fish, apparently unable to maintain territories in
the more favorable habitat of the clear water tributaries. The data we
have obtained provides a basis to evaluate the population trends over
time and changes in the populations in response to mainstem habitat
changes and overwintering conditions.
Six sites which were sampled consistently by boat electrofishing in 1982
and 1983 and produced relatively high numbers of Arctic grayling were
Whiskers Creek Slough (RM 101.2), Lane Creek (RM 113.6), Fourth of July
Creek (RM 131.1), Indian River (RM 138.6), Jack Long Creek (RM 144.5),
and Portage Creek (RM 148.8).
Tag and recapture data support the theory that most Arctic grayling
spawn in tributaries. Recoveries of tagged fish in May and early June
show movement into tributaries.
Boat electrofishing catch data in 1982 suggests that most of the large
Arctic grayling move into tributaries immediately after ice out (ADF&G
1983b). In 1981, adult Arctic grayling were gillnetted in early May at
open water tributaries when the mainstem was still partially covered
with ice (ADF&G 1981c), indicating that Arctic grayling begin moving
prior to the open water sampling. Boat electrofishing data from 1983
-34 -
-
-
-
-~
-
support 1981 findings. We did not monitor tributary temperatures which
probably influence Arctic grayling movements more than ice cover on the
mainstem and may also account for the differences in timing between
years. Arctic grayling elsewhere in Alaska begin to migrate as the
water temperature increases to about 1°C (Armstrong 1982).
Data from 12 spawning Arctic grayling captured at RM 150.1 in late May
1983 suggest that either mainstem spawning occurs there or that spawning
occurs nearby.· Since no Arctic grayling recaptures have been made above
Devil Canyon (RM 150.1-161.0) from fish tagged below Devil Canyon and no
tagged fish have been observed in the tributaries in the canyon
[Cheechako Creek (RM 152.5), Chinook Creek (RM 156.8), and Devils Creek
(RM 161.0)], it appears unlikely that lower or middle river Arctic
grayling spawn above RM 150.1.
Higher CPUE's for Arctic grayling were recorded in late July during 1983
than in past years at the mouths of several tributary sites such as
Indian River (RM ·138.6) and Jack Long Creek (RM 144.5). We are not
certain why this occurred, however, the drought which decreased the
water levels in these tributaries during 1983 may have caused some
Arctic grayling to move out of the tributaries earlier than in 1982.
Recaptures of Floy anchor tagged Arctic grayling show that a strong
spring migration of Arctic grayling occurs in the Susitna River. In the
summer, most Arctic grayling have been recaptured at or near their
tagging locations. This suggests that Arctic grayling do not move far
from their summer rearing areas. The outmigration of adult Arctic
grayling from tributaries to the mainstem occur in September. Boat
electrofishing CPUE' s in 1982 and 1983 increased steadily from late
August through late September and then decreased in early October. This
suggests that most of the Arctic grayling have moved into the mainstem
by the end of September.
Little is known about Arctic grayling distribution during the winter in
the Susitna River. It is believed that many Arctic grayling overwinter
in the mainstem Susitna, however, specific overwintering areas in the
mainstem have not been identified. It is also believed that significant
numbers of Arctic grayling overwinter in Portage Creek. This tributary
is characterized by many deep (20 feet) pools which may provide adequate
overwintering conditions for Arctic grayling. The proportion of the
population that uses this habitat is not known.
The survival rate of Arctic grayling between the Chulitna River
confluence and Devil Canyon is 56%, which is similar to the population
above Devil Canyon. Although few individuals grow past 400 mm fork
length or Age 8, there appears to be a high recruitment from the younger
age classes, notably Ages 3 and 4.
Since reproduction is relatively high for Arctic grayling, the avail-
ability of rearing habitat may be a critical factor for this species
{Scott and Crossman 1973). Studies in 1982 indicate that younger fish,
Age classes 2 to 4, use the mainstem Susitna to a limited extents,
probably due to their displacement from tributaries by the territorial
behavior of the larger fish {ADF&G 1983b). Future changes in the
-35 -
availability of rearing habitat may be expected to directly affect the
population size of Arctic grayling in the Susitna River.
The congregation of older Arctic grayling (>300mm) at the mouths of only
a few selected streams between the Chulitna River confluence and Devil
Canyon makes them vulnerable to overfishing. Local residents have
stated that fishing for Arctic grayling has deteriorated since 1970
because of increased fishing pressure (Harold Larsen pers. comm.).
4.4 Round Whitefish
The distribution and abundance of round whitefish in the Susitna River
between the Chulitna River confluence and Devil Canyon in 1983 was
similar to findings in 1981 and 1982.
The catch of round whitefish has increased substantially each year since
1981 because of increased electrofishing efforts and the addition of
downstream migrant traps. The deployment of a second downstream migrant
trap off the west bank of the Susitna River (RM 103.0) contributed
significantly to the increased round whitefish catch in 1983.
Pooled CPUE rates based on boat electrofishing data from 1982 and 1983
showed that CPUE's at tributary or slough sites were much higher than at
mainstem sites above the Chulitna River confluence (ADF&G 1983b).
During both years sampling efficiency appeared to be the same for
mainstem and tributary or slough sites. Although boat electrofishing
CPUE' s of round whitefish are generally lower at mains tern sites compared
to tributary sites, high CPUE's were recorded in the mainstem during
June in both 1982 and 1983. Relatively high catch rates in the mainstem
were also recorded in September of both years. Pooled boat electro-
fishing data from 1982 showed higher catch rates of round whitefish at
all sites above the Chulitna River confluence than below. We speculated
this was due to more preferable habitat in this reach of river. In
1983, mainstem boat electrofishing data pooled into three subreaches (RM
98.5 -115.5, RM 115.6 -138.5, and RM 132.6 -150.1) showed that round
whitefish are most abundant in the area between RM 132.6 -RM 150.1 in
the Susitna River above the Chulitna River confluence.
Extensive sampling by JAHS crews above the Chulitna River confluence in
1983 showed that juvenile round whitefish are found more frequently at
mainstem and slough sites than at tributary sites. Although it is
unknown where they hatched, it is probab 1 e that round whitefish prefer
areas with slow velocities and turbid water for rearing.
Seasonal boat electrofishing CPUE's at tributary sites above the
Chulitna River confluence during 1982 were the highest in late June,
late August and late September (ADF&G 1983b). It was speculated in 1982
that the high catches during June and September were due to migration of
fish into and out of tributaries. A similar trend in movement was
observed in the 1983 boat electrofishing CPUE data.
Most of the recaptured round whitefish from 1981-83 showed 1 ittl e
movement. During this time, only 26 of llO recaptured round whitefish
moved over 5.0 miles (ADF&G 1981c, 1983b). Round whitefish recaptured
-36 -
-
-
-
-
-
-
-
-
-
in 1981 and 1982 exhibited a pronounced fall movement. In 1983 round
whitefish exhibited a general downstream movement throughout the summer.
The longest move documented for a tagged round whitefish was 69.5 miles
downriver from its tagging site. This fish was recaptured in 1983 by a
sport fisherman at Willow Creek (RM 49.1).
While round whitefish spawning has not been observed in the mainstem,
the distribution of sexually ripe males and females captured suggests
that spawning probably occurs within mainstem areas. Sexually ripe male
and female round whitefish have been found in the mainstem Susitna River
during early October in 1981, 1982 and 1983.
Although few sexually ripe round whitefish were captured in 1981 and
1982, over 50 were captured in 1983. This was due to differences in
sampling efficiencies rather than variability in timing of spawning. In
1983 extensive boat electrofishing was done in early October, while in
1981 and 1982 mechanical breakdowns of electrofishing equipment limited
sampling during this time. -
Since 1981, nine locations have been determined to be spawning sites for
round whitefish in the mainstem Susitna according to the criterion used
to determine a spawning site (female fish able to discharge eggs upon
palpation). In 1981 and 1982 spawning was observed at RM 100.8 and RM
102.6, respectively. In 1983 seven sites were found including four
mainstem sites (RM 102.0, RM 114.0, RM 142.0 and 147.0) and three
tributary mouth sites [Lane Creek (RM 113.6), Indian River (RM 138.6)
and Portage Creek {148.8)].
Catch data suggests that round whitefish spawning may occur throughout
the mainstem. Sexually mature fish { > 300mm) have been captured during
October in locations characterized by slow to moderate water velocities
with silt to rubble substrate. Most sexually ripe fish have been
captured in pairs or small groups. Mass spawning behavior of round
whitefish has been reported elsewhere (Normandeau 1969; Bryan and Kato
1975).
Large schools of adult round whitefish have also been captured at the
mouth of Portage Creek and Indian River in late September. This may
indicate that some round whitefish use these tributaries to spawn.
While catch data suggests that spawning areas of round whitefish are
widespread in the mainstem, the selection of specific spawning sites may
not be random. Anchor ice, water fluctuations and ice cover can all
limit egg survival. Due to these reasons, round whitefish in the
Susitna River may seek out areas which have an adequate influx of ground
water. Habitat data taken at one mainstem site (RM 147.0 in 1983), where
eight sexually ripe males and females were captured, supports this
hypothesis. Specific conductance was relatively high {160 umhos/cm) in
this area indicating an area of upwelling. This hypothesis is also
believed to be true for another mainstem spawning species in the Susitna
River, chum salmon (ADF&G 1983c).
-37 -
There is probably an upstream spawning run of round whitefish in the
fall. Spawning takes place at temperatures slightly above 0°C (Morrow
1980). Many of the juveniles subsequently migrate to the lower river
for rearing during their first year as evidenced by the catch rate of
juveniles in the downstream migrant traps.
Comparisons of 1981-1983 age-length data for round whitefish shows
considerable differences in each age class. Although results are
similar between 1981 and 1982, we believe the findings in 1983 are more
accurate. Fish were probably underaged in 1981 and 1982. Although
positive aging cannot be verified for fish of all three years, compari-
sons of the annuli of" scales from fish initially tagged in 1982, and
recaptured in 1983 provided better information on when round whitefish
in the Susitna River form their annuli.
Age-length data in 1983 show that round whitefish are one of the older
living resident fish species in the lower Susitna River with fish older
than Age 8 occurring rather often. The oldest round whitefish found in
the Susitna River by our crews was Age 12. Subsamples of aged fish also
show that the population appears stable with fish captured frequently
over all spawning age classes Age 5 and older.
Most round whitefish in the Susitna River have rather slow growth rates.
This slow growth begins at Age 3, decreases steadily thereafter, and
becomes almost non-existent after Age 10. Few round whitefish in the
Susitna River attain fork lengths greater than 390 mm. However, scale
analysis showed four fish experienced periods of extremely rapid growth.
For example, one fish aged at four years old was 265 mm fork length
while the mean fork length of 33 aged fish was 187 mm and the 95 percent
confidence intervals ranged from 141-233 mm. This fish showed extremely
rapid growth during the first and second years of its life. Based on
recapture data and reports of round whitefish being found in brackish
water (McAllister 1964; Morin et al. 1982} we believe that this fish may
have migrated from the estuary. Tag-and-recapture data from 1981 to
1983 show that some round whitefish migrate long distances in the
Susitna River.
4.5 Humpback Whitefish
Humpback whitefish have been found in the Susitna River from RM 10.1 to
RM 150.1, however, they are captured ·infrequently except during certain
time periods (ADF&G 1981c; 1983b). Sampling in 1981 and 1982 in the
reach of river below and above the Chulitna River conflu~nce {RM 98.5)
further showed that humpback whitefish were more numerous in the reach
of river below the Chulitna River confluence than above.
Although boat electrofishing in 1983 was limited to sampling above the
confluence, the data show a simil~r humpback whitefish distribution and
abundance in this reach of river as in prior years. Pooled boat
electrofishing /CPUE data in 1982 and 1983 reveal generally higher
humpback whitefish densities at tr·ibutary or slough sites than at
mainstem sites {ADF&G 1983b).
-38 -
-
~.
.....
-
-
-
-
-
-
,..,.
I
Fishwheel catches in 1982 and 1983 indicate similar yearly distributions
and abundance of adult humpback whitefish. Peak catches at fishwheels
during both years were in late August with 148 and 137 fish captured in
1982 and 1983 respectively.
Few juvenile humpback whitefish have been captured from 1981 to 1983
except by the downstream migrant traps (RM 103.0). It is currently
unknown where most young juvenile humpback whitefish rear.
Morrow (1980) reports that adult humpback whitefish move little except
during the spawning run beginning in June and lasting throughout Septem-
ber. In the Susitna River, fishwheel catches in 1982 and 1983 also
reveal a spawning run occurs during this time period. Catches during
both years peaked at Yentna (RM 28.5, TRM 4.0) and Sunshine (RM 79.0) in
late August (AOF&G 1983b). High catches were also recorded at Talkeetna
(RM 103.0) and at Curry (RM 120.0) in late August or early September.
Fi shwheel catch data recorded at Sunshine in 1981 reflect a simi 1 ar a
mid-September peak in catch (ADF&G 1981c). Susitan River humpback
whitefish spawning is presumed to occur in October in tributaries.
Tag-recapture data on humpback whitefish is limited but seems to indi-
cate a spawning or overwintering movement. Three fish tagged in Septem-
ber 1981 were recaptured in May or early July 1982, presumably before
they migrated again in fall 1982. Since these fish were recaptured long
distances (16-38 miles) downriver, it is thought that these fish were
originally tagged during their upstream migration in September. After
spawning, they returned downriver to overwinter where they were re-
captured in 1982. In addition, two fish tagged and recovered in 1983
also show an upstream movement. One fish moved 11.0 miles from late
June to mid-August, while another moved 17.0 miles in two days in
mid-July, possibly an early spawning movement.
While little is known of juvenile humpback whitefish distribution and
movement, downstream migrant trap catches in 1983 suggest that there is
a downstream movement of juvenile humpback whitefish during late· July.
Nearly all of these fish were young of the year.
Comparisons of mean lengths of humpback whitefish by age class between
1981, 1982, and 1983 shows little differences. However, comparisons of
humpback whitefish age-length data by reach indicate that fish below the
Chulitna River confluence appear to be 1 a rger than fish between the
Chulitna confluence and Devil Canyon (ADF&G 1981c; 1983b).
Scale analyses indicated that some humpback whitefish undergo a period
of very rapid growth during their first two years of 1 if e. The data
suggest that some humpback whitefish may spend part of their life
history rearing in an estuarian environment. Elsewhere in Alaska, AOF&G
(unpublished~, Alt (1979) and Berg (1948), report that l· pidschian does
not venture 1nto estuary zones as often as other species of the humpback
whitefish complex.
-39 -
4.6 Longnose Sucker
Longnose suckers occur throughout the Susitna River below Devil Canyon,
however, they appear to be more abundant in the reach of river below the
Chulitna River confluence (RM 98.5) (ADF&G 198lc; 1983b). Boat electro-
fishing catches in 1982 and 1983 were higher at tributary and slough
sites than at mainstem sites. Boat electrofishing data in both years
showed higher CPUE's at tributary and slough sites above the confluence
in August and September than in June or July. Longnose suckers may move
into tributary and slough sites in August and September to feed on
salmon eggs.
Recapture data indicate that adult longnose suckers are relatively
sedentary. Thirty-two of 45 longnose suckers recaptured from 1981 to
1983 did not move over 5.0 miles from their tagging locations {AOF&G
1981c; 1983b).
t-1ovements of the remaining 13 recaptured longnose suckers suggest an
upstream migration occurs in the spring and a downstream movement occurs
in the fall to overwintering areas.
Catch per unit effort data also support the hypothesis that there is a
spring and fall movement. Boat electrofishing catch rates at sites
sampled above the Chulitna River confluence progressively increased in
the spring and the fall in 1982 and 1983 (ADF&G 1983b).
Inferences of population dynamics for longnose suckers aged between 1981
and 1983 are difficult due to problems with aging this species accurate-
ly by scale analysis. While longnose sucker age data from 1983 is
similar to 1981 data up to Age 7, results from 1982 are similar to 1983
data only up to Age 3 and to 1981 data only after Age 6. Bond {1972)
found that he could accurately determine the ages of a closely related
species of sucker (white sucker, C. commersoni) by scale analysis up to
Age 9. However, since the mean fengths of several longnose sucker age
classes from our data vary considerably from year to year, we believe
that scale analysis is not an accurate technique for aging longnose
suckers on the Susitna River.
Another indication of the problem relating to age determination of
longnose suckers was provided by examining scales from two recaptured
fish one year later. One of the recaptured longnose suckers was accu-
rately aged for both years and the other was misaged both years. By
comparing scales from the two years, no new annulus was formed on the
1983 scale. Other studies of longnose suckers show similar results in
regard to the failure of tagged fish to form an annulus (Geen et al.
1966). Bucholz and Carlander (1963) suggest that when there is little
or no growth, fish do not forming a scale annulus. Evidently, this is
prevalent among longnose suckers in the Susitna River.
Several authors suggest alternate methods to age suckers. Beamish and
Harvey (1969) found that by using cross sections of pectoral fin rays
they were able to age older fish. Quinn and Rose {1982) found that ·
aging by pectoral fin rays for slower growing populations of suckers
this method was reliable only up to Age 7 suckers. They further imply
that otoliths are the best method to age older suckers.
-40 -
.~
.....
-
-
-
....
-
-
-
-
~.
I
......
-
While it is difficult to characterize the oldest age classes of Susitna
River longnose suckers, it appears that above the Chulitna River conflu-
ence annual growth increments decline steadily after Age 5 (ADF&G 1981c,
1983b). Age-length data from longnose suckers captured in the Susitna
River below the Chulitna River confluence in 1981 and 1982 indicate that
fish continue to grow steadily after Age 5. Catch data from these two
years also show a higher frequency of larger fish being caught below the
Chulitna confluence. This is probably due to more favorable habitat
conditions in this reach which allows for more growth.
4.7 Other Species
4.7.1 Dolly Varden
Dolly Varden occur throughout the Susitna River drainage, however,
extremely low catches have been made from 1981 to 1983. The most
productive areas are the Kashwitna River (RM 61.0), Lane Creek (RM
113.6), Indian River (RM 138.6), and Portage Creek {RM 148.8).
Catch data from 1982 show that Dolly Varden move out of the mainstem and
into tributaries by late June (ADF&G 1983b). After June, catch rates at
all sites influenced by the mainstem river stayed low all summer in 1982
and 1983. It is thought that Dolly Varden rear in the upper reaches of
tributaries until fall and then migrate back into the mainstem to
overwinter. Although it is not known when the exact timing of the fall
outmigration occurs, anglers at the mouth of the Talkeetna River and
Kashwitna River report high catches after mid-September (S. Kreuger and
R. Bloomfield pers. comm.).
Tag-recapture data from a sma 11 number of Dolly Varden recovered in 1982
and 1983 show an upstream spring movement as well as a summer movement
(ADF&G 1983b). In 1982 it was speculated this may be due to a spawning
movement.
Two out of nine Dolly Varden recaptured between 1981 and 1983 were
recovered in Clear Creek, suggesting that this tributary creek may be an
important producer of Dolly Varden in the lower Susitna River.
4.7.2 Threespine stickleback
Distribution and abundance of threespine stickleback appears to be
variable in the Susitna River. In 1981 sticklebacks were found upstream
as far as RM 146.9, in 1982 they were found upriver only to RM 101.2,
and in 1983 upriver to RM 112.3 (ADF&G 1981c; 1983b). A comparison of
catches at several sites sampled all three years suggest that catches
peaked in 1981 and increased again in 1983. While over 2,000 threespine
sticklebacks were captured at Slough 6A in 1981, none were captured in
1982 and 77 were caught in 1983.
Capture data in 1981 and 1982 suggest an upstream migration begins to
occur during late May (ADF&G 1981c; 1983b). This movement is presumed
to originate from the estuary as a spawning migration.
-41 -
Downstream migrant trap data suggest that threespine stickleback outmi-
grate in the summer following emergence. Thirty-two age 0+ {under 40
mm) threespine stickleback were captured in 1982 by a downstream migrant
trap, while approximately 1,406 of 1,601 threespine stickleback captured
by these traps in 1983 were age 0+.
The catch in 1982 was lower
spawning population in 1982.
hatching, young of the year
downstream to brackish water.
4.7.3 Arctic lamprey
than in 1983 probably due to a smaller
Morrow (1980) also reports that after
threespine stickleback immediately move
Arctic lamprey are believed to be abundant in the Susitna River below R~,i
50.5 and decrease in abundance above this river mile (ADF&G 1983b).
Most Arctic 1 ampreys have been captured at the mouths of sma 11 tribu-
taries such as Chase Creek (ADF&G 1981c; 1983b). Arctic lamprey dis-
tribution and abundance data from 1983 was similar to 1981 and 1982 for
the reach of river above the Chulitna River confluence (RM 98.5). Less
than 100 Arctic lamprey have been captured each year.
-42 -
~I
-
-
·-
-
-
5.0 CONTRIBUTORS
Data was collected by Richard Sundet and Mark Wenger with help from
Kathrin Zosel.
Dana Schmidt provided the study design. Steve Hale assisted with
running the CAPTURE program.
Data processing was done by Allen Bingham~ Gail Heineman, Donna
Buchholz, Carol Kerkvliet, Kathrin Zosel, and Alice Freeman.
Bruce Barrett reviewed the draft of this report and provided helpful
comments.
Drafting was done by Sally Donovan and Carol Kerkvliet.
Typing was done by Skeers Word Processing.
-43 -
6.0 ACKNOWLEDGEMENTS
Funding for this study was provided by the State of Alaska, Alaska Power
Authority.
We would like to express our gratitude to all the people and organiza-
tions that provided information or assistance to the Resident Fish Study
during the past years.
We are especially grateful to Carl Burger (USFWS) for his technical
expertise and advise on radio telemetry investigations. We would also
like to thank Carl for providing photographs showing the implantation of
radio tags into rainbow trout.
We are grateful to the employees at the Palmer ADF&G, Sport Fish office
and state veterinarian Burt Gore, D.V.M. who showed us several tech-
niques to radio tag fish. We are also grateful to ADF&G, FRED Division
personnel at Elmendorf hatchery who provided the rainbow trout for this
radio tagging clinic.
We wish to thank the various staff of E. Woody Trihey and Associates
(EWT&A), the Arctic Environmental Information Data Center, Harza-Ebasco
Susitna Joint Venture, Woodward-Clyde Consultants, the Alaska Power
Authority, and especially Cleve Steward (EWT&A) for reviewing the draft
of this paper and providing helpful suggestions.
-44 -
-
-
-
-
-
-
-
-
-
~
I
-
7. 0 LITERATURE CITED
Alaska Department of Fish and Game (ADF&G). 1981a. Aquatic studies
procedures manual. Phase 1. Subtask 7 .10. Alaska Department of
.Fish and Game Susitna Hydro Aquatic Studies. Anchorage, Alaska.
___ . 1981b. Phase 1 final draft report. Subtask 7.10. Adult
anadromous fisheries project. Alaska Department of Fish and Game
Susitna Hydro Aquatic Studies. Anchorage, Alaska.
--..,.· 1981c. Phase 1 final draft report. Subtask 7.10. Resident
fish investigtion on the Lower Susitna River. Alaska Department of
Fish and Game Susitna Hydro Aquatic Studies. Anchorage, Alaska.
__ ..,.. 1981d. Phase 1 final draft report. Subtask 7 .10. Resident
fish investigation on the Upper Susitna River. Alaska Department
of Fish and Game Susitna Hydro Aquatic Studies. Anchorage, Alaska.
--~· 1983a. Aquatic studies procedures manual. Phase II (1982-83).
Subtask 7 .10. Alaska Department of Fish and Game Susitna Hydro
Aquatic Studies. Anchorage, Alaska.
1983b. Susitna Hydro aquatic studies Phase II basic data
report. Volume 3 (2 parts). Resident and juvenile anadromous fish
studies on the Susitna River below Devil Canyon, 1982. Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies.
Anchorage, Alaska.
1983c. Susitna Hydro aquatic studies Phase II basic· data
report. Volume 4 (3 parts). Aquatic habitat and instream flow
studies, 1982. Alaska Department of Fish and Game Susitna Hydro
Aquatic Studies. Anchorage, Alaska.
1983d. Susitna Hydro aquatic studies Phase II basic data
report. Volume 5. Upper Susitna River impoundment studies, 1982.
Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.
Anchorage, Alaska.
• 1983e. Susitna Hydro aquatic studies Phase II data report.
--"Winter aquatic studies (October 1982 -May 1983). Alaska Depart-
ment of Fish and Game Susitna Hydro Aquatic ~Studies. Anchorage,
Alaska.
• 1983f. Age-length relationships for Arctic grayling and
---rainbow trout. Appendix J in Susitna Hydro aquatic studies Phase
II report. Synopsis of the 1982 aquatic studies and analysis of
fish and habitat relationships (2 parts). Alaska Department of
Fish and Game Susitna Hydro Aquatic Studies. Anchorage, Alaska.
• 1984. Susitna Hydro aquatic studies May 1983 -June 1984
--procedures manual. Alaska Department of Fish and Game Susitna
Hydro Aquatic Studies. Anchorage, Alaska.
-45 -
Alaska Department of Fish and Game. (Unpublished manuscript).
Whitefish investigations of the Yukon-Kuskokwim Delta. Commercial
Fisheries Division, Bethel.
Alt, K. T. 1979. Contributions to the life history of the humpback
whitefish in Alaska. Transactions of the American Fisheries
Society. 108(2): 156-160.
Armstrong, R. H. 1982. Arctic grayling studies in Alaska. Alaska
Cooperative Fishery Research Unit, University of Alaska, Fairbanks.
Barrett, B. M., F. M. Thompson, and S. N. Wick. 1984. Adult anadromous
fish investigations: May -October 1983. Alaska Department of
Fish and Game Susitna Hydro Aquatic studies Report No. 1. Prepared
for Alaska Power Authority. Anchorage, Alaska.
Beamish, R.J., and H.H. Harvey. 1969. Age determination in the white
sucker. Journal of the Fisheries Research Board of Canada
26:633-638.
Berg, L.S. 1948. Freshwater fishes of the U.S.S.R. and adjacent
countries. Guide to the fauna of the U.S.S.R. No. 27. Vol. 1, 4th
ed. Akad. Nauk SSSR Zool. Inst., Moscow. Translated from Russian
by Israel Program for Scientific Translations, 1962. Office of
Technical Services, U.S. Dept. Commerce, Washington, D.C.
Bond, W.A. 1972. Spawning migration, age, ~rowth and food habits of
the white sucker, Catostomus commersoni (Lacepede), in the Bigoray
River, Alberta. Master 1 s thesis, Un1versity of Alberta, Edmonton,
Canada.
Bryan, J.E., arrd D.A. Kato. 1975. Spawning of lake whitefish,
Coregonus clupeaformis, and round whitefish, Prosopium
cylindraceum, in Aishihik Lake and East Aishihik River, Yukon
Territory. Journal of the Fisheries Research Board of Canada
32(2):283-288.
Bucholtz, M.M., and K.D. Carlander. 1963. Failure of yellow bass,
Roccus mississippienisis, to form annuli. Transactions of the
American Fisheries Society 92:384-390.
Chen, L. C. 1969. The biology and taxonomy of the burbot, Lata lata
leptura in interior Alaska. Biological Paper. Univers1ty--crf
Al aska 11:1-51.
Estes, C. C. and D. S. Vincent -Lang editors. 1984. Aquatic habitat
and instream flow investigations, May -October 1983. Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies. Report
No. 3. Prepared for Alaska Power Authority. Anchorage, Alaska.
Geen, G.H., T.G. Northcote, G.F. Hartman, and C.C. Lindsey 1966. Life
histories of two species of catostomid fishes in Sixteenmile Lake,
British Columbia, with particular interest to inlet stream spawn-
ing. Journal of the Fisheries Research Board of Canada
23:1761-1788.
-46 -
-
-
-
-
-
-'
-
-
-
-
Jacobs, K.E., and W.O. Swink. 1982. Estimations of fish population
size and sampling efficiency of electrofishing and rotenone in two
Kentucky tailwaters. North American Journal of Fisheries Manage-
ment. 2:239-248.
McPhail, J.D., and C.C. Lindsey. 1970. Freshwater fishes of north-
western Canada and Alaska. Bulletin of the Fisheries Research
Board of Canada 173: 1-381.
Mills, M.J. 1982. Alaska statewide harvest study-1981 data. Alaska
Department of Fish and Game. Federal Aid in Fish Restoration and
Anadromous Fish Studies, Annual Performance Report. Study No.
SW-1, Job No. SW-1-A, Vol. 23:1-115.
Montana Department of Fish, Wildlife and Parks (MDFW&P). 1983.
Kootenai River fisheries investigations final report 1972-1982.
Mont. Dept. Fish, Wildlife & Parks. Kalispell, Mont.
Morin, R., J.J. Dodson, and G. Power. 1982. Life history variations of
anadromous cisco (Coregonus artedii), lake whitefish (C.
clueeaformis), and round whitefish (Prosopium cylindraceum) popu-
latlons of Eastern James-Hudson Bay. Canadian Journal of Fisheries
and Aquatic Sciences. 39:958-967.
Morrow, J.E. 1980. The freshwater fishes of Alaska. Alaska Northwest
Publishing Company, Anchorage, Alaska.
Needham, P.R., and D.W. Slater. 1945. Seasonal changes in growth,
mortality and condition of rainbow trout following planting.
Transactions of the American Fisheries Society. 73:117-124.
Needham, P.R., and A.C. Jones. 1959. Flow, temperature, solar ra-
diation, and ice in relation to activities of fishes in Sagehen
Creek, California. Ecology 40(3):465-474.
Normandeau, D.A. 1969. Life history and ecology of the round whitefish,
Prosohium cylindraceum (Pallas), of Newfound Lake, Bristol, New
Hamps ire. Transactions of the American Fisheries Society.
98(1):7-13.
Otis,D.L., K.P. Burnham, G.C. White; and D.R. Anderson. 1978. Wild-
life Monographs 62:1-135.
Quinn, S.P., and M.R. Ross. 1982. Annulus formation by white suckers
and the reliability of pectoral fin rays for aging them. North
American Journal of Fisheries Management 2:204-208.
Rawstron, R.R. 1973. Comparisons of disk dangler, ·and internal anchor
ta~s on three species of salmonids. California Fish & Game,
59(4):266-280.
-47 -
Rawstron, R.R., and R.J. Pelzman 1978. Comparison of. Flay internal
anchor and disc dangler tags on largemouth bass (Micropterus
salmoides) at Merle Collins Reservoir. California Fish & Game,
64(2):121-123.
Reimers, N. 1957. Some aspects of the relation between stream foods and
trout survival. California Fish & Game, 43(1):5-42.
Scott, W.B., and E.J. Crossman. 1973. Freshwater fishes of Canada.
Bulletin of the Fisheries Research Board of Canada 184. Ottawa,
Canada.
White, G. C., D.R. Anderson, K.P. Burnham and D.L. Otis. 1982. Cap-
ture-recapture and removal methods for sampling closed populations.
Los Alamos National Laboratory. Los Alamos, New Mexico, U.S.A.
Zeibell, C.D. 1973. Ultrasonic transmitters for tracking channel
catfish. The Progressive Fish Culturist. 35(1):28-31.
-48 -
.~
-
-·
-
-
-
-
APPENDIX A
Gear Efficiency and Selectivity and Tag Retention
--49 -
r
INTRODUCTION
Between August 9 and October 7, 1983, the responses of 13 radio tagged
fish (12 rainbow trout and one burbot) to boat motors, electrofishing,
and the generator in the electrofishing boat were observed.
METHODS
Gear efficiency
Boat electrofishing efficiency was analyzed by reviewing field notes
concerning observed effects of electroshocking on radio tagged fish.
Several radio tagged fish were also tested to observe their responses to
other noises associated with boat electrofishing such as boat motors and
the electric generator which powers the electrofishing unit.
Gear selectivity
Gear selectivity of the different gear types was evaluated by examining
length frequency distributions by gear type.
Tag retention efficiency
The external Flay anchor tag (model FD-67) has been used to tag resident
fish since January 1981 to determine seasonal and yearly movements. The
dimensions of the tag and tagging procedure are explained in the 1981
procedures manual (ADF&G 1981c). Disc dangler tags were used to tag
burbot for several months during 1981 and spring 1982.
The efficiency of the Flay anchor tag was evaluated for Arctic grayling
and round whitefish by comparing the number of fish with tag scars to
the total number of fish with tag scars and Flay anchor tags of that
species recaptured in 1983. By subtracting this ratio from 1.00, Flay
·anchor tag retention efficiencies were determined. Tag retention
efficiencies for rainbow trout and longnose suckers were not determined
because the smaller scales on these species regenerate rapidly and make
it difficult to detect tag scars.
RESULTS
Gear efficiency
electrofishing
Response of radio tagged fish to boat
During these 13 observations, all radio tracking was conducted by the
electrofishing boat.
Two of the rainbow trout and one burbot were recaptured and the others
fled from the sound of the boat or generator, or the electric field and
avoided capture.
Rainbow trout (659-2.0) and burbot (639-3) were accidentally recaptured
during routine sampling. Rainbow trout (628-2.0) had moved 10.9 miles
downriver in 20 days and it appeared healthy when it was recaptured, but
-50 -
it was late presumed to have died due to tagging injuries. The remain-
ing ten radio tagged fish moved away from the electrofishing boat during
the experiment. The 1 oca t ion of each fish was pinpointed before and
after each experiment to observe their behavior.
Six fish moved away from the sampling area when electrofishing occurred
in their vicinities. Three of these fish (rainbow trout 718-1.5,
738-1.4 and 748-1.6) were located at the mouth of Fourth of July Creek
(RM 131.1) on August 14. After 20.0 minutes of electrofishing at the
mouth of the creek the tagged fish all moved out of the area. Rainbow
trout (718-1.5) was relocated 0.6 miles downriver on the opposite bank
of the Susitna River. Rainbow trout (738-1.4) moved 200 yards into a
side channel. Rainbow trout (748-1.6) moved 150 feet downriver and into
the main channel of the Susitna. All three returned to the mouth later
that day. Rainbow trout (639-1.4) was located at Moose Slough (RM
123.5) on August 14. After electrofishing the area for 19.0 minutes,
the fish was relocated 20 feet from its original location in a deeper
section (10 feet) of the slough. Another rainbow trout (670-1.4) was
located at the mouth of Whiskers Creek Slough (RM 101.2) on October 7.
This area was shocked for 12 minutes and the tagged fish was not
captured. After shocking, the fish was found to have moved 20 feet into
the main channel. The remaining rainbow trout (660-3) was located at
the mouth of Portage Creek (RM 148.8) on September 19. This area was
shocked for 26.5 minutes. This fish was seen moving in 3.5 feet of
water away from the electric probe. After electrofishing, this fish was
found approximately 20 feet from its previous location in deeper water.
At all sites where these six radio tagged fish were located, other non-
radio tagged fish were captured during electrofishing.
On September 17 three fish were tested for responses to the sound of the
boat's electrofishing generator. These fish (rainbow trout 597-1.3,
709-1.5 and 768-1.5) were located next to the bank of the mainstem river
within 100 yards of each other at RM 114.3.
After locating the fish, the boat was positioned approximately 10 feet
away from each fish and the generator was started. All three fish moved
100-200 feet downriver after the generator was started. This was done
twice for each fish and the response was the same each time.
Ten fish were tested to observe their responses to the boat • s motor.
The ten fish included the six which fled during electrofishing (rainbow
trout 718-1.5, 738-1.4, 748-1.6, 639-1.4, 670-1.4, and 660-3), the three
that fled during the operation of the generator (rainbow trout 597-1.3,
709-1.5 and 768-1.5), and one other fish (rainbow 649-1.2). All but one
fish (649-1.2) remained in the same area when the boat was near them.
The estimated distance between the boat and each fish was from 10-30
feet.
Rainbow trout (649-1.2) was located at the mouth of Indian River (RM
138.6) on September 19. While moving towards the fish and monitoring at
the same time, the fish moved across the river (200 yards). After
locating and moving towards the fish on the opposite side, the fish
returned to the mouth. The closest distance the boat came to the fish
was estimated at 100 feet.
-51 -
-
-
-
-
-
......
Gear Selectivity
Rainbow trout
Rainbow trout were captured by nine of the 11 sampling techniques used
during the 1983 resident fish studies. The length frequencies of the
rainbow trout captured by the four methods accounting for 95% of the
total catch are shown in Appendix Figure A-1. Hook and 1 ine and boat
electrofishing techniques sampled a wide range of lengths (89-612 mm),
while minnow and migrant traps captured only juvenile fish (30 -191 mm).
Burbot
Burbot were captured by seven of the 11 sampling techniques used during
the 1983 resident fish studies. Ninety-three percent of all the burbot
caught were captured by the four techniques shown in Appendix Figure
A-2. Boat electrofishing sampled the widest range of lengths (107 -751
mm), while the migrant trap call ected only juvenile fish (26 -134 mm).
Arctic grayling
Arctic were captured by five of the 11 sampling techniques used during
the 1983 resident fish studies. Boat electrofishing accounted for 90%
of the total Arctic grayling catch. The five techniques which captured
Arctic grayling are shown in Appendix Figure A-3. Boat electrofishing
sampled the widest range of lengths (97 -444 mm) and the smelt trap,
with the exception of a few incidental adult catches, only sampled the
juveniles (36 -175 mm). The other methods only sampled the fish
between 200 and 400 mm.
Round whitefish
Round whitefish were captured by five of the 11 sampling techniques used
during the 1983 resident fish studies. The length frequencies of the
round whitefish captured by the four major methods (hook and 1 i ne
captured only one fish) are shown in Appendix Figure A-4. Boat
electrofishing and the migrant traps accounted for 98% of the total
catch. Boat electrofishing sampled a wide range of lengths (94 -403mm)
while the migrant trap captured mainly juveniles (23 -208mm).
Humpback whitefish
Humpback whitefish were captured by four of the 11 sampling techniques
used during the 1983 resident fish studies. The length frequencies of
the humpback whitefish captured by these four methods are shown in
Appendix Figure A-5. The migrant traps accounted for 77% of the total
catch, most being juvenile (30 -145mm). The other methods were selec-
tive for fish between 140 and 480mm.
Longnose sucker
Longnose sucker were captured by five of the 11 sampling techniques used
during the 1983 resident fish studies. The length frequencies of the
-52 -
U1 w
I
>-u z w
:::>
0 w
0:: u..
1-
z w
u
0::
10
w 2
0..
RAINBOW TROUT
HOOKond UNE
n=l48
mean• 285mm
BOAT ELECTROFISHING
n= 143
mean= 330 mm
30
20
10
80
FORK LENGTH (mm)
MINNOW TRAP
n 11 33
meon=l30mm
DOWNSTREAM MIGRANT TRAP
n" 12
mean= 63 mm
Appendix Figure A-1. Gear selectivity for rainbow trout in the Susitna River, May through October 1983 .
I ·~.· ~J ~-J .J .... J
>-
0 z w
:::>
0 w
0:::
lL.
t-z
' w <.n u .;:..
I 0:::
w
0...
1
30
20
10
0
"' r ~
0 .,
30
20
10
o n~~~~~~
"' "' "' "' ' 0 .,
0"13
J
BUR BOT
BOAT ELECTROFISHING 30 HOOP NET
n=62 n~22
mean= 411 mm meon=389mm
20
10
0 ll ... "' "' ..
!. 2
DOWNSTREAM MIGRANT TRAP
n=22
mean=74 mm
TROT.LINE
n=l9
mean=505mm
TOTAL LENGTH (mm)
Appendix Figure A-2. Gear selectivity for burbot in the Susitna River, May through October 1983.
50
•o
30
20.
IQ
0
BOAT Et..ECTROFISHING
fl•9-61
mean •247 mm
. : ~ .
~
* ~ ~
:e 2 ~ i
..•
~ . ~ ill ; .
0 ~ ~ ~ iO ~
30
20
10
0 . i ~ . ~ f ~ ~ ~
FORK LENGTH (mml
~ 5 ~ ~ 2 :
ARCTIC GRAYLING
. !l ~
~ ~
DOWNSTREAM MIGRANT TRAP
n•35
m1an • 3.5 mm
Appendix Figure A-3. Gear selectivity for Arctic grayling in the
Susitna River, May through October 1983.
~55-
-
-
-
-
-
1
ROUND WHITEFISH
50
40
30
20
J
"'
.. r
0 ..
FORK LENGTH (mm)
DOWNSTREAM MIGRANT TRAP
n•1,204
rneon=55 mm
Appendix Figure A-4. Gear selectivity for round whitefish in the Susitna River, May through October 1983.
l
30
20
10
>-0 z w
:::> a w
0:::
lL..
(n . ......., 1--
I z w
0
0::: w
0.
zo
10
0 ...
HUMPBACK WHITEFISH
BOAT. ELECTROFISHING
n = 36
mean" 276 mm
FISHWHEELS
n=90
mean= 367 mm
40
30
FORK LENGTH (mm)
SET GILL NET
1'1" 12
mean= 299mm
DOWNSTREAM MIGRANT TRAP
n= 466
mean=57mm
Appendix Figure A-5. Gear selectivity for humpback whitefish in the Susitna River, May through October
... J
-
-
-
longnose suckers captured by these four major methods are shown in
Appendix Figure A-6 (hook and line captured only 2 fish). Boat
electrofishing accounted for 66% of the total catch and captured the
widest range of lengths (133-407mm). The migrant trap once again
captured mainly juvenile longnose suckers (21-175mm) while the net
methods were selective for the median lengths (200-380mm).
Tag retention efficiency
The Fl oy anchor tag efficiency determined for round whitefish in the
Susitna River during 1983 was 77.5 percent with 20 of 89 recaptured
round whitefish showing a tag scar. The tag efficiency~ meanwhile, for
Arctic grayling during 1983 was 69.4 percent with 15 of 49 recaptured
Arctic grayling showing a tag scar.
-58 -
>-
(.)
z
30
20
10
LONGNOSE SUCKER
BOAT ELECTROFISHING
n=377
meon=308
40
30
20
10
HOOP NET
n=26
meon; 303mm
w 0 -L,---,__,......,..J~~
:;:,
0 w
0::
LL
~ I-
I Z w
(.)
ffi 30
Q.
20
10
SET GILL NET
n= 55
meon=.309
801
70~
~
FORK LENGTH (mm)
DOWNSTREAM MIGRANT TRAP
n = Ill
meon= 51 mm
Appendix Figure A-6. Gear selectivity for longnose suckers in the Susitna River, May through October
1983.
,)
-
-
DISCUSSION
Gear efficiency
electrofishing
Response of radio tagged fish to boat
Observed responses of 13 radio tagged fish to boat electrofishing
equipment suggest that fish learn to avoid recapture. A similar hypoth-
esis has been reported elsewhere {Jacobs and Swink 1982).
Only three of the 13 radio tagged fish were recaptured and the others
avoided the electrofishing boat. Twelve of these fish were originally
captured by electrofishing and one by hook and line (670-1.4).
Since only one of ten fish moved away from the sound of the boat motor,
it appeared that they disassociate the effects of the electric field and
capture to the sound of boat motors. This was probably due to the
constant "traffic" on the river between the time of capture and when the
experiment occurred. This enabled the fish to become acclimated to the
sound of boat motors.
While most of the fish did not respond to the sound of boat motors, they
did respond to generator noise. All of the fish tested for a response
to generator noise moved away from the source of the noise. Prior to
these observations we believed that the radio tagged fish would not
associate the generator sound with the electric field because of the
extended periods of time between successive samplings.
It appears that while boat electrofishing provides a good method to
capture fish for collection of biological data, it is a poor method by
itself for a tag-and-recapture program since fish learn to avoid the
field.
Gear selectivity
For each of the six species that the gear selectivity study was conduct-.
ed on, there was always at least one sampling technique which sampled a
wide range of lengths, one .that sampled only the juveniles and others
that sampled a small segment of the population between the smallest and
largest. Boat electrofishing was generally the best technique in
sampling a wide range of lengths, while the downstream migrant traps was
often the most effective means of capturing juveniles.
Tag retention efficiency
Studies in 1983 show that the Floy anchor tag, model FD-67, is lost from
25 percent of recaptured round whitefish and Arctic grayling. Other
studies have also reported tag losses using the model FD-67 anchor tag.
Wilbur and Duchow (1973) reported tag losses on largemouth bass up to 78
percent using the model FD-67 tag. Arctic grayling tagging studies in
the Chena River and the upper Susitna River basin reported 10 percent
tag losses (R. Holmes and M. Stratton, pers. comm., respectively).
Rawstroms (1973) reported that the primary reason for tag shedding is
improper securement. He found that tag retention rates increase
-60 -
if the tag is inserted behind the interneurals rather than into the
dorsal musculature. Rawstroms also stated that secondary causes of tag
loss occur due to breakage of the T-section of the tag or to separation
of the vinyl tube from the monofilament anchor.
Our studies also suggest that the primary cause of tag loss is improper
placement of the tag. Very few (under five) tagged fish in our study
have been found without the vinyl tube. Observations of recaptured
round whitefish and Arctic grayling show that an ulcer forms around the
area where the tag has been inserted. Since both these species have
large scales, regeneration may be impeded due to the constant movement
of the external part of the Flay tag.· The constant movement impedes
regeneration, and the wound ultimately enlarges. With the greater hole
from the wound, the tag falls out enabling the scales to regenerate or
to form a scar. Other resident fish species such as rainbow trout and
longnose suckers probably have higher tag retention rates than Arctic
grayling and round whitefish. This may be due to their smaller scales
which adhere to the tag better.
Although some Flay anchor tags are lost due to shedding it is still the
best tag to use for our studies because it can be deployed rapidly and
because it is more economical to use than other types of tags.
Tag losses during our 1983 studies appeared to decrease due to better
placement of tags. In 1982 most of the tags were injected into the
dorsal musculature. In 1983, tags were anchored at the base of the
dorsal fin.
-61 -
....
-
-
....
-
-
-
-
-
-
-
-
-
....
I
APPENDIX B
Radio Tagged Fish Movement Data
-62 -
-
Appendix Table B-1. Summary of tagging data for radio tagged rainbow -trout on the Susitna River Between Cook Inlet and
Devil Canyon, May to December 1983.
-Radio Fre-
quency/Fork Age/ Method Location River Date Date
Length (mm) Sex captured Captured Mile Capt 1 d Rels 1 d
597-1.3/424 6, F EF Lane Creek 113.6 7/18 7/19
600-1.0/508 -' F HL Indian River 138.6 9/2 9/2 -607-1.5/385 7, M HL Indian River 132.6 9/18 9/19
608-1.2/444 8' -EF Indian River 138.6 10/4 10/5
610-1.0/548 -, M ON 4th of July Cr 131.1 5/11 5/12 -619-1.0/440 -, M HL 4th of July Cr 138.6 9/1 9/2
619-1.4/387 5' -EF Indian River 138.6 9/2 9/3
628-1.2/423 6' -EF Indian River 113.6 10/4 10/5
630-1.0/558 -, M ON 4th of July Cr 131.1 5/11 5/12 .... 639-1.0/382 6' -EF Indian River 138.6 9/2 9/3 I
639-1.4/460 ' EF Slough 8A 125.3 7/16 7/17
648-1.6/405 6, F HL Whiskers Cr TRM 0.2 6/5 6/6
649-1.2/427 7' -EF Indian River 138.6 9/2 9/3
660-3.0/508 8, F EF Protage Cr 148.8 9/2 9/3
670-1.4/391 7' -HL Whiskers Cr TRM 0.2 6/6 6/7
709-1.5/418 ' EF Lane Creek 113.6 7/18 7/19 -718-1.5/376 5' -EF Indian River 138.6 6/8 6/9
719-1.0/455 5' -HL Indian River TRM 5.0 8/11 8/11
729-1.0/455 -' F HL 4th of July Cr 131.1 9/1 9/2 -729-1.3/446 6, M HL Indian River 138.6 9/2 9/3
738-1.4/455 EF Indian River 138.6 6/8 6/9
748-1.6/442 -' F EF Skull Creek 124.5 7/18 7/19
F"" 749-1.0/438 7' -HL Indian River 138.6 9/2 9/3
758-20/416 7, EF Lane Creek 113.6 7/18 7/19
767-1.5/435 6' -EF Lane Creek 113.6 7/18 7/19
768-1.0/432 6, F EF Indian River 138.6 10/4 10/5 -- = Not sexed or not aged, EF = Electrofishing, HL = Hook & Line,
ON = Drift Net -
-63 -
Appendix Table B-2. Summary of tagging and tracking data for radio tagged burbot on the Susitna River between Cook Inlet and Devil
Canyont July to December i983.
Radio Method Date River Date
Frequency/ Captured Capt'd Mile Rels'd J2ly Auf!st Seetember October Nov Dec
Total length BB
8 1 29 5 15 19 3 6 21 IO 1
(mm) pa B p B p B p B p p _!'_
610-3,0/550 Electroshock 7/18 113.6 7/19 112.3 110.0 112.5 112.0 112.0 111.3 112.0 112.0 112.0 112.0 NSC NS
639-3.0/728 Electroshock 9/18 142.0 9/19 140.0 140.0 134.3 134,3 131.8
670-3.0/677 Electroshock 9/1 123.5 9n 123.5 120.5 118.6 110.2 110.2 88.0 87.3 87.7
720-3.0/750 Electroshock 9/3 147.5 9/3 146.9 146.7 147.3 147.0 144.0 NS NS 134.8
~ Tracked by plane
c ~~a~~=~a~y boat
.I J --]
APPENDIX C
Population and Biological Characteristics
-
-
-
-
-
-65 -
-
Rainbow Trout
The sexual maturity of 28 rainbow trout from the Susitna River were
. examined between May 11 and July 18, 1983. Sexually ripe pre-spawners
were captured from May 11 to June 7. Spawned out rainbow trout were
captured from June 5 to July 18.
Fork lengths of 16 male rainbow trout examined for sexual maturity
ranged from 260-558 rnm with a mean of 403 mm. The fork lengths of
twelve sexually mature female rainbow trout ranged from 325-454 mm with
a mean of 399 mm.
Ages of twenty-one rainbow trout ranged five to eight (Appendix Figure
C-1).
A total of 424 rainbow trout were captured between the Chulitna River
confluence and Devil Canyon during 1983. The length frequency composi-
tion for rainbow trout is presented in Appendix Figure C-2. Fork
lengths ranged from 30-612 mm with a mean of 284 mm.
Scale analysis was used to determine the ages of 265 rainbow trout
captured on the Susitna River between the Chulitna River confluence and
Dev"il Canyon. A~es ranged from one to nine. Ages 3 (18.1%), 4 (18.1%),
5 (25.3%) and 6 {17.7%) rainbow trout were the most abundant age classes
(Appendix Table C-1). A graphical presentation of age-length data in
Appendix Figure C-3 shows a steady growth rate for rainbow trout.
Two hundred forty-four of the 265 rainbow trout aged were captured by
boat electrofishing or hook and line. Data from fish captured by these
two methods, were used to calculate an instantaneous survival rate of
33.3 percent by using age versus catch (Appendix Figure C-4).
Burbot
One hundred sixty one burbot were captured in the Susitna River between
the Chulitna River confluence and Devil Canyon during 1983. Total
lengths measured on 135 burbot ranged from 26-815 mm with a mean of 366
mm (Appendix Figure C-5). Most of the burbot measured ranged from 330
mm to 510 mm in total length.
Few juvenile burbot ·(total length< 200 mm) were captured in 1983. The
majority (22 of 24) of the juvenile burbot measured were caught by the
downstream migrant traps at Rr4 103.0.
The instantaneous survival rate for burbot was calculated using pooled
data from fish aged from otoliths from January 1981 to March 1983. The
instantaneous survival rate for burbot aged in this time period was
calculated to be 70.5 % (Appendix Figure C-6).
Arctic Grayling
The sexual maturities of 51 Arctic grayling from the Susitna River were
examined between May 20 and June 22, 1983. Sexually ripe pre-spawners
-66 -
Appendix Figure C-1. Age and length relationship for spawning rainbow
trout captured in the Susitna River between the
Chulitna River confluence and Devil Canyon, May 11
through July 18, 1983.
-67-
-
-
>-(.) z
IJ.J ::::>
0 w a:
(L,
1-z
IJ.J
(.) a:
IJ.J a.
1
2 • 0
:!
..,
I
0
"' N
FORK LENGTH (mm)
RAINBOW TR
-All gear tn
n= 353
mean=284
Appendix Figure C-2. Length frequency composition of rainbow trout
captured in the Susitna River between the Chulitnc
River confluence and Devil Canyon by all gear
types, May to October 1983.
-68-
0'\
1.0
. I
Appendix Table C-1. Rainbow trout age-length relationships on the Susitna River between the Chulitna River
confluence and Devil Canyon, May to October 1983.
Total No. Standard
Length (mm)
95%
Age of Fish Mean Deviation Confidence Range
(,~ears} Sampled Intervals
Fish Captured b~ Boat Electrofishing and Hook and Line
1 5 97 9.43 85 -109 93 -106
2 12 155 15.51 145 -165 124 -180
3 46 210 31.54 201 -219 159 -260
4 45 274 33.55 264 -284 205 -329
5 65 331 36.62 322 -340 260 -455
6 45 377 38.84 365 -389 301 -446
7 21 423 31.45 409 -437 366 -471
8 5 452 43.67 398 -506 390 -508
9 1 612
Total 244 306 193 -612
Fish Captured b~ All Methods
1 9 92 7.95 86 -98 84 -106
2 18 150 14.96 143 -157 124 -180
3 48 210 31.15 201 -219 159 -260
4 48 275 33.50 265 -285 205 -329
5 67 330 36.00 321 -339 260 -455
6 47 378 38.41 367 -389 301 -446
7 21 423 31.45 409 -437 366 -471
8 6 462 46.86 413 -511 390 -515
9 1 612
Total 265 298 84 -612
.I .. J J .l
.....
-
650
600
550
500
-450
E
E
-400
I
I-
(!)
z
w
_...J
~
0::
0 u..
co v
II
~r c
I
2
I
3
co v
II
c
...... c.o
II
c
C\1
II
c:
II
II •
c
!=Range of Fish, n= 265
e =Mean
I I
4 5 6
AGE
I
8
I
9
Appendix Figure C-3. Age and length relationships for rainbow trout
captured in the Susitna River between the Chulitna
River confluence and Devil Canyon, May to October
1983.
-70-
----------------------~---------------------------------------
100
• r 2 = 0. 97
50 a = 36698.59 • • b = -1.10
Survival = 3 3.3%
0 n .= 244 w a = Points used for (!) regression ana I ysis <t:
I
Cf)
lL • lL 10 0
Cf) a:: w
CD 5 ~
~ • z
2 3 4 5 6 7 8 9
AGE (years)
Appendix Figure C-4. Survival rate curve for rainbow trout captured in
the Susitna River between the Chulitna River
confluence and Devil Canyon, 1983.
-71-
-
-
~
....
.....
-
""'"'
i
I ......
N
I
1
>-u z w => 0 w
0:
lL.
1--z w u
0: w a..
l
I I I
0 0 0 m CIJ 10
-J
0
(\J
q-
TOTAL LENGTH (mm)
BUR BOT
-All gear types.
n = 135
mean= 366
m
1'-0
1'-CJ)
I I ~ &3
1'-1'-CJ)
Appendix Figure C-5. Length frequency composition of burbot captured in the Susitna River between the
Chulitna River confluence and Devil Canyon by all gear types, May to October 1983.
100
50
a
w
(!)
<(
:t:
C/)
LL.
lL.
0 10
C/)
0:::
w
Ill
;:i! 5
::::> z
•
•
•
2 = 0.96 r
a = 391.20
b = -0.35
Survival = 70.5%
n = 262
0 = Points used for
regress ion analysis
3 4 5 6 7 8 9 10 II
AGE (years)
12 13 14 15
Appendix Figure C-6. Survival rate curve for pooled burbot catch data
from the Susitna River between Cook Inlet and
Devil Canyon, 1981 to 1983.
-73-
....
~
-
~
-'
-'
-
~
....
-
-
-
....
were captured from May 20 to May 24. Post spawners were captured from
May 21 to June 22.
Fork lengths for 30 male Arctic grayling which spawned in 1983 ranged
from 308-444 mm with a mean length of 367 mm. Twenty-one female Arctic
grayling spawners had fork lengths ranging from 320-386 mm with a mean
of 349 mm.
Ages of 29 of the 30 male Arctic grayling examined for spawning condi-
tion ranged from Age 5 to Age 10. Ages of 19 female Arctic grayling
spawners ranged from Age 5 to Age 8 (Appendix Figure C-7).
A total of 1,168 Arctic grayling were captured on the Susitna River
between the Chulitna River confluence and Devil Canyon during 1983.
Fork lengths of 1,071 of those fish were measured to the nearest milli-
meter. Arctic grayling fork lengths ranged from 30 mm to 444 mm with a
mean of 246 mm (Appendix Figure C-8). Juveni 1 e Arctic grayling (fork
length under 200 mm) made up 26.4% of the catch.
Age analysis from scales of 523 Arctic grayling captured on the Susitna
River between the Chulitna River confluence and Devil Canyon yielded
ages which ranged from age 0+ to Age 10 (Appendix Figure C-9). Ages 3
{27.0%) and 4 {31.4%) were sampled most often {Appendix Table C-2).
Five hundred sixteen of the 523 Arctic grayling aged were captured by
boat electrofishing, hook and line, and hoop net. The instantaneous
survival rate for Arctic grayling captured by these three methods was
calculated at 56.0% {Appendix Figure C-10) •
Round Whitefish
Sexual maturity was determined for a subsample of round whitefish
captured on the Susitna River between the Chulitna River confluence and
Devil Canyon from October 3 to October 7, 1983. Forty males and 12
female round whitefish were sampled, all were pre-spawners. Fork
lengths of the males ranged from 266 mm to 380 mm with a mean of 319 mm.
Fork lengths for the females ranged from 319 mm to 403 mm with a mean of
355 mm. Ages of seventeen of the spawning males ranged from Age 5 to
Age 8 {Appendix Figure C-11). One female was Age 7.
In October 1983 three spawning sites for round whitefish were found.
Two sites were at the mouth of tributaries, Lane Creek (RM 113.6) and
Portage Creek (RM 148.8), and the other site was in the mainstem Susitna
at RM 147.0 off an island. At each site several extremely ripe females
and rna 1 es were captured. Fema 1 e round whitefish expe 11 ed eggs when
their abdomens were palpated. No spent fish were captured at these
sites.
Fork lengths of 2,497 round whitefish ranged from 23-403 mm with a mean
of 167 mm. Appendix Figure C-12 illustrates the length frequency
composition of all fish measured.
Four hundred fifty-six round whitefish were aged using scale analysis.
Ages ranged from Age 1 to Age 12 and Ages 4 (12.3%), 5 (16.2%), 6
-74 -
450
440-(l)e
430-
420-
410-. NI II c: -~I E 400-
E -390-..
I ·r-1-
(!) 380-r" co <0 z II II
w c: c: ~I _J 370-
~ 360-. a:: 1'-
0 II
350-c: 1.1.. <0 ~
II II 0)
340-c: c: II
c:
330-. (I >• I.
320-
n=29 n = 19
3 I 0-..
300-'5 6 7 8 9 10 5 6 7 8
1AGE I
I I . AGE
Spawning Males Spawn1ng Females
Appendix Figure C-7. Age and length relationships for spawning Arctic
grayling captured in the Susitna River between the
Chulitna River confluence and Devil Canyon, May 20
to June 22, 1983.
-75-
-
-
~
""'!
"""'
'"""'
....
-
-
-
-
>-0 z w
:::>
0 w
0::
I..L
1--z w
0
ffi a..
I
0 en
I
2
I
f6 0
ct
FORK LENGTH (mm)
ARCTIC GRAYLING
-All gear types.
n;; I, 071
mean=246 mm
-I
Appendix Figure C-8. Leng!~ !~~quency composition of Arctic grayling captured in the Susitna River between
-E
E -
:I:
t-
(!)
z:
w
_J
~
ll:
0
ll...
300
250
ue c:
I
0
m
C\1
II
c:
~I
I
2
v
II
c:
I
3
~
co
II
c:
~
<0
II
c:
I I
4 5
AGE
co v
II
c:
I
6
"' If)
II
c:
C\1
N
II c:
10 ... c:
C\1 ··t c:
I = Rang e Q f Fish , = 52 ,
• =Mean
I
7
I
8
I
9
I
10
Appendix Figure C-9. Age and length relationship for Arctic grayling
captured in the Susitna River between the Chulitr
River confluence and Devil Canyon, May to October
1983.
-77-
-
-
-
.....
-
-
~
-
-
......
O'J
Appendix Table C-2. Arctic grayling age-length relationships on the Susitna River between the Chulitna
River confluence and Devil Canyon, May to October 1983. Fish aged were captured by
boat electrofishing, hook and line and hoop net.
Total No.
Length (mm}
Age of Fish Standard 95% Confidence
(.~ears) Sampled Mean Deviation Intervals Range
0 1 108
1 5 113 9.63 101 -125 97 -122
*1 12 105 12.83 97 -113 80 -122
2 29 160 19.92 152 -168 126 -212
3 141 207 25.38 203 -211 142 -265
4 164 254 24.76 250 -258 198 -315
5 64 301 28.72 294 -308 245 -365
6 46 341 19.45 335 -347 290 -380
7 37 364 23.52 356 -372 315 -409
8 22 390 19.87 381 -399 362 -444
9 5 396 6.2.8 388 -404 390 -405
10 2 411 7.78 341 -481 405 -416
*Total 523 261 80 -444
*Aged fish caught by all sampling methods.
Cl w
(9
<!
J:
(J)
LL.
LL.
0
(J)
a:: w
CD
~
::> z
500
400
300
200
100
50
•
•
• • r 2 = 0.85
a = 1548.94
b = -o.58
Survival= 56.0%
n = 516
o : Points used for
regression analysis
0
•
I -l--~~----...-----r------r---r--T----.--.-"-T---'-..;;....;..;;;;··
0 2 3 4 5 6 7 8 9 10
AGE {years)
Appendix Figure C-10. Survival rate curve for Arctic grayling captured
in the Susitna River between the Chulitna River
confluence and Devil Canyon, 1983.
-79-
-
-
....
....
-
·-
-
410-----------------r--------------~
390
370
350
-e 33o
e -
:I: _..,_
(!) z
llJ
_J
~
0::
0
310
290
lL 270
250
190
180
n= 40
I 6 Males were aged
from scales.
6 7 a
Spawning Males Spawning Females
Appendix Figure C-11. Age and length relationship for spawning round
whitefish in the Susitna River between the
Chulitna River confluence and Devil Canyon,
October 4 to November 7, 1983.
-80-
34 n=843
32
30
28
26
>-u z w :::>
0 w
0::
lJ..
1-z w u 12
0:: w
Q_ 10
' I -0
111
N
FORK LENGTH
' I 0 0 .,. ,....
N N
(mm)
I ROUND WHITEFISH I
-All gear types.
n=2,497
mean= 167mm
I I I
0 0 0
0 <D m
rt) rt) pt)
Appendix Figure C-12. Length frequency composition of round whitefish
captured in the Susitna River between the
Chulitna River confluence and Devil Canyon by all
gear types, May to October 1983.
-81-
-
-
-
_,
-
-
-
-
Four hundred nineteen round whitefish were captured
electrofishing and aged. The instantaneous survival rate
whitefish captured by boat electrofishing was determined to
(Appendix Figure C-14}.
Humpback Whitefish
by boat
for round
be 58.3 %
E"ight hundred twenty humpback whitefish were captured in the Susitna
River between Cook Inlet and Devil Canyon during 1983. Fork lengths of
604 humpback whitefish were measured to the nearest millimeter. Fork
1 engths ranged from 30-480 mm with a mean of 125 mm. The 1 ength fre-
quency composition of the humpback whitefish catch is presented in
Appendix Figure C-15.
Ages of 78 humpback whitefish captured in the Yentna River (TRM 4.0} and
41 humpback whitefish captured in the Susitna between the Chulitna River
confluence and Devil Canyon were determined by scale analysis. Ages
from fish captured on the Yentna River ranged from Age 5 to Age 12 with
Ages 6 (25.6%), 7 (18.0%) and 8 (20.5%) predominating (Appendix Table
C-4}. Humpback whitefish were captured between the Chulitna River
confluence and De vi 1 Canyon ranged from Age 1 to Age 8 with Ages 4
(26.8%) and 5 (22.0%) predominating. The age-length relationship of
humpback whitefish presented in Appendix Figure C-16 shows that humpback
whitefish are slow growing with a wide range of fork lengths occurring
at several age classes.
Longnose Suckers
Sexual maturity was determined for 55 longnose suckers captured on the
Susitna River from May 22 to September 20, 1983. Sexually ripe male
1 ongnose suckers were captured throughout the summer. Sexually ripe
female longnose suckers were captured during June and September.
Spawned out males and females were captured from June 6 to July 18.
Fork lengths for the spawning male longnose suckers ranged from 282-392
mm with a mean of 332 mm. Spawning female longnose suckers ranged from
300-408 mm with a mean of 348 mm.
Thirteen of the male longnose suckers were aged by scale analysis with
ages ranging from six to nine (Appendix Figure C-17). Eight female
longnose suckers aged ranging from seven to ten years old.
Fork 1 engths of 571 1 ongnose suckers were measured. Fork 1 engths of
1 ongnose suckers ranged from 21-411 11111 with a mean of 258 mm. The
length frequency composition of longnose suckers captured in 1983 is
presented in Appendix Figure C-18.
One hundred thirty-six longnose suckers were aged by scale analysis.
Ages ranged from Age 1 to Age 11 and Ages 7 (23.5%) and 8 {25.0%) were
the most abundant age classes encountered (Appendix Table C-5).
-82 -
00 w
...... · .J J
Appendix Table C-3. Round whitefish age-length relationships on the Susitna River between the Chulitna
River confluence and Devil Canyon, May to October 1983. Fish aged were captured by
boat electrofishing.
Length (mm)
Total No.
Age of Fish Standard 95% Confidence
(.~ears} Sampled Mean Deviation Intervals Range
1 4 102 4.57 95 -109 95 -105
*1 41 89 11.90 85 -93 67 -110
2 11 152 15.94 141 -163 135 -187
3 33 187 22.34 179 -195 154 -265
4 . 56 222 20.13 217 -227 174 -271
5 74 262 20.74 257 -267 184 -302
6 52 290 42.67 278 -302 248 -332
7 61 311 21.65 305 -317 260 -366
8 53 332 19.15 327 -337 276 -386
9 42 342 19.44 336 -348 282 -390
10 16 362 19.70 352 -372 327 -384
11 13 376 19.45 364 -388 388 -403
12 4 382 23.96 344 -422 346 -397
*Total 456 267 67 -403
*Aged fish caught by all sampling methods •
I J .J I .. 1 ! .J ]
.-
-
r
.....
-
-
400 !I ~I 1'1')
II
c: 350 N
1'1') v
It) II
II c: -(0 c:
E300 N II
It) E c:
II ........ c:
I v
r2so ......
(!) U) II·
c: z It)
1'1') II w 10 c
_j 200 II
c:
~
0::
0
lL 150 ~I
I= Range of Fish , n = 4 56
d •= Mean
50
0
Q_ 2. 3 4 _5 6 7 8 9 10 II. 12.
AGE
Appendix Figure C-13. Age and length relationships for round whitefish
captured in the Susitna River between the
Chulitna River confluence and Devil Canyon, May
to October 1983.
-84-
0 w
(.!)
<t
:r:
(/)
LL.
LL.
0
en
0:: w
CD
~
:::J z
100
• 50
•
10 •
5
•
2 3 4
•
•
0
r2 = 0.92
a = 3620.58
b = -0.54
Survival ::: 58.3%
n = 419
0 = Points used for
regression analysis
5 6 7 8 9
AGE (years)
10 II 12
Appendix Figure C-14. Survival rate curve for round whitefish captured
in the Susitna River between the Chulitna River
confluence and Devil Canyon, 1983.
-85-
-
-
-
-
-
-
.... ,
-
->-u z w
::::>
0 w a=
l.L
1-z w u
ll: w a.
FORK LENGTH (mm)
HUMPBACK WHITEFISH
-All gear types.
n=604
mean= 125 mm
Appendix Figure C-15. Length frequency composition of humpback
whitefish captured in the Susitna River between
Cook Inlet and Devil Canyon by all gear types,
May to October 1983.
-86-
co
"-1
I
I
Appendix Table C-4. llumpback whitefish age-length ~elat1onsh1ps on the Susitna River between Cook inlet and Devil Canyon, Hay to Octobe~ 1983, Fish aged we~e captured by all
sampling methods.
Yentna River (TRH 4.0) Chulitna Confluence to Devil Canyon Yentna River to Devil Can:l:on
Total No. Len&th (mm) Total No. Lenath (10M) Total No.
Age of fish Standnd 95% Confidence of fish Standard 95% Confidence of fish
Len11th (mm)
Standard 95% Confidence
(years) Sampled Mean Deviation Intervals Range Sampled ~ Deviation Intervals Range Sampled He an Deviation Intervals Range
1 3 121 60.72 0 • 272 77 -190 3 121 60.72 0 -272 77 • 190
159 10.07 69 -249 153 -165 159 10.07 69 -249 153 -165
3 4 251 18.96 221 -281 228 -268 4 251 18.96 221 -281 228 -268
4 11 270 22.04 255 • 285 236 -311 11 270 22.04 255 -285 236 -311
11 334 25.08 317 -351 286 -363 9 303 13,82 292 -314 281 -322 20 320 25.54 308 -332 281 -363
6 20 . 348 22.74 337 • 359 316 • 390 6 330 18.23 311 • 349 303 -358 26 343 22.80 334 -352 303 -390
14 367 25,51 352 -382 318 -404 4 322 29.18 276 -368 288 -356 18 350 31,82 334 -366 288 -404
8 16 367 22.25 355 -379 329 • 400 402 49.50 0 -847 367 -437 18 371 26.63 358 -384 329 -437
9 397 22.22 376 -418 369 -410 397 22.22 376 -418 369 -410
10 6 416 31.06 383 -449 377 -458 6 416 31.06 383 -449 377 -458
11 3 430 20.03 380 -480 409 -449 3 430 20.03 380 -480 409 -449
12 1 419 1 419
Total 78 367 286 -458 41 279 17 • 437 119 337 11 -458
.' J .I J J
-
450
400
350
-E 300 E
::t: .... 250 (!)
z w
....l
~ 200
a::
0
IJ....
150
100
50
0
0
T
I
I
I
I
I
I • I
I
.L
NT ...
"'J.
T
~· c:l
.l
•• ..
I =Fish Below Chulitna River
Confluence {RM 98.5), n =78
: =Fish Above Chulitna River
1 Confluence,n=41
I =Range
•=Mean
2 3 4 5 6 7 8 9 10 II 12
AGE
Appendix Figure C-16. Age and length relationship for humpback
whitefish captured in the Susitna River between
Cook Inlet and Devil Canyon, May to October
1983.
-88 -
E
E
:I:
1-
(.!)
420-
400-
380-
z 360-
UJ
....J
::::.:::: a::
0
LL.
340-
320-
300-
(I) •
Ole
6 7
Spawning
n=l3
v
II
c
8 9
I I
Males
(I )e
v
II
n=8
c "'
(I )e
" c
7 8 9 10
I AGE I . I
Spawning Females
Appendix Figure C-17. Age and length relationships for spawning
longnose suckers captured in the Susitna River
between the Chulitna River confluence and Devil
Canyon, May to October 1983.
-89-
-
-
-
"""' I
-
-
!"""'
-
.....
1'1=20 20 LONGNOSE SUCKER
-All gear types.
18 n= 571
mean= 258 mm
16
>-14
(.)
z w
:::::l 12
0 w a:: u.. 10
1-z
w 8
(.) a:: w a.. 6
4
2
0
I I ' N
I 0 0 I I
'"" (1) 0 0 0 0 IC) ,.._
'"" <n N N .., '""
FORK LENGTH (mm)
Appendix Figure C-18. Length frequency composition of longnose suckers
captured in the Susitna River between the
Chulitna River confluence and Devil Canyon by all
gear types, May to October 1983.
-90-
1.0 .......
Appendix Table C-5. Longnose sucker age-length relationships on the Susitna River between the Chulitna
River confluence and Devil Canyon, May to October 1983. Fish aged were captured by all
methods.
Lensth (mm)
Total No.
Age of Fish Standard 95% Confidence
(~ears) Sampled Mean Deviation Intervals Range
1 3 81 11.37 53 -109 68 -90
2 2 127 10.28 35 -219 120 -133
3 7 196 18,51 179 -213 168 -219
4 2 244 3.54 212 -276 241 -246
5 10 245 23.97 228 -262 208 -282
6 16 291 21.74 279 -303 256 -321
7 32 320 25.90 311 -329 276 -370
8 34 347 27.60 337 -357 307 -408
9 17 364 24.36 351 -377 330 -407
10 10 363 20.72 348 -378 336 -403
11 3 372 16.26 332 -412 360 -383
Total 136 312 68 -408
.J J .J
-
Appendix Figure C-19 shows that the growth rate of longnose suckers in
the Susitna River between the Chulitna River confluence and Devil Canyon
is relatively slow.
Dolly Varden
Seventeen Dolly Varden were captured on the Susitna River in 1983.
Eight fish were captured by boat electrofishing and seven by the down-
stream migrant traps at RM 103.0. The downstream migrant traps Dolly
Varden catches were all juveniles {< 200 mm). Fork lengths of boat
electrofishing Dolly Varden catches ranged from 146-320 mm.
Threespine Stickleback
Five hundred and seventy-four threes pine stickleback were captured by
the downstream migrant traps at RM 103.0 in 1983. Total lengths of
these threes pine stickleback ranged from 11-93 mm with a mean of 31 mm. ·
-92 -
450
400 il ~I ... ~~ .,
350 " c:
N ,.,
~ " c:
300
E ~ E
;350 ~I 1-c:
(!) ~I z
UJ 200
..J
~ a::
0 150 I = Range of Fish, n= 136 lL
(1,1 \!:I
• =Mean
100
~~
50
0
2 3 4 5 6 7 8 9 10 fl
AGE
Appendix Figure C-19. Age and length relationship for longnose suckers
captured in the Susitna River between the
Chulitna River confluence and Devil Canyon, May
to October 1983.
-93-
-
....
~
"""1
~
-I
-
-
--
APPENDIX D
Population Estimates
-94 -
-
-
·"""
Dur-ing the course of the 1983 Resident Fish Studies, biases and as-
sumptions relating to the population estimates of resident fish were
identified. These biases fall into two general categories, those caused
by behavior or other attributes of the biology of the fish and those
caused by the sampling technique (Appendix D-1). The biases for each of
the population estimates made were shown to be different depending on
the species, area, and gear type used for sampling, or by a combination
of these three factors.
The major bias associated with the rainbow trout population estimate in
Fourth of July Creek (RM 131.1) was behavioral, the avoidance of recap-
ture. After a fish was captured and marked, the capture probability of
that fish decreased substantially since it learned to avoid the lure.
This was observed during the second and third occasion of sampling.
Although the lure was put before the marked fish, it did not strike. To
correct for this bias, a behavioral model (a type of removal model)
which allowed for decreases in capture probabilities was used in cal-
culating the population estimate.
A secondary bias of the population estimate for rainbow trout at Fourth
of July Creek was the size selectivity of sampling gear, resulting in
variations in individual capture probabilities. Smaller fish have been
reported to have a smaller capture probability than larger ones in other
population estimates (ADF&G 1983d). This was also true for rainbows in
Fourth of July Creek; angling was ineffective in capturing fish under
151 mm in fork length.
The population estimate of 107 rainbows in Fourth of July Creek there-
fore pertains only to rainbow trout over 150 mm.
Similar biases were shown at a mainstem site between RM 138.9-140.1
where a burbot population estimate was made. Since no burbot were
recaptured at this site during the four day sampling period, a removal
model was used to generate a population estimate. Other tag and recap-
ture data from 1981-83 have also ·shown that burbot evidently learn to
avoid recapture since less than ten have been recaptured during three
years of sampling.
A secondary bias of size selectivity as found for rainbow trout in
Fourth of July Creek, for the population estimate of burbot was evident
since no burbot under 300 ~n total length were captured. The population
estimate of burbot in this reach of the mainstem river should therefore
be applicable only to burbot over 300 mm in length.
To minimize the effects of in-or outmigration, sampling for rainbow
trout was done in July. Electrofishing dur·ing July and August 1982
captured few rainbow trout in the mainstem indicating that rainbow trout
are residing in the tributaries during this time period.
To minimize the possibility of in-or outmigration for burbot, sampling
was done in July because catch results from 1981-82 and radio tagged
burbot data from 1982 show that burbot move only from September to
March.
-95 -
Appendix Table D-1 Riases, corrections, and assumptions which affect
the resident fish population estimates below Devil
Canyon, 1983.
Bias:
roiTection:
Assumption:
Bias:
Correction:
Assumption:
Bias:
"'C"'"rre c t i on :
Assumption:
Bias:
Correction:
Assumption:
Lack of randomness of mark or recapture effort.
Stratification of habitat location by habitat type.
Random mark and recapture effort.
Unequal recapture probability due to time between census-
ing.
Use of multiple census estimator during a short time
period.
Time does not affect recapture probability.
Population is open geographically.
Use of July and August data only; period of minimal
movement.
Population is closed geographically.
Heterogeneity; variance in the probability of capture and
recapture between age classes.
Stratification of age class for entire population,
develop correction factor for populations.
Population estimates limited to Age IV and older fish due
only to insufficient sample sizes of smaller fish.
-96 -
"""' I
-
-
-
Although population estimates were generated for burbot in the mainstem
Susitna, problems were incountered with calculating population estimates
for other resident species in the mainstem during 19S3. For instance,
catch information shows the major biases associated with the population
estimates made at Slough SA (RM 125.3) were probably that the fish
migrated in and out of the site during the sampling {not a closed
population) and that there was an avoidance of fish to electrofishing
which was the method of capture used in Slough SA. Sampling was done at
this site during only a 72 hour period (twice a day for three days) to
correct for the geographical bias, however, failed. The resultant
population estimate, for example, of round whitefish at this site was
believed inaccurate since the estimate was S96 but had a standard error
of 294.43 using the population model selected by the computer as best
fitting the data. The low catch of round whitefish at Slough 8A on two
occasions compared to the other four occasions (25, 3, 38, 2S, 2S, and
S) showed that fish were moving in and out of the slough during at least
these two time periods.
The ·movements of round whitefish as well as other species during these
two time periods, meanwhile, were probably due to the changing turbidity
in Slough SA during the sampling period. The mainstem river was approx-
imately 0.5 feet lower on those two occasions compared to the other four
occasions. As the mainstem water decreased, the slough became clearer.
The decreased round whitefish catches on these two occasions suggests
that the fish moved into the mainstem when the water in the slough was
no longer turbid enough to provide adequate cover.
Resident fish also appeared to avoid electrofishing and this avoidance
was not anticipated prior to conducting the estimates. Of 130 round
whitefish captured in Slough SA during six occasions only nine (6.9%)
were recaptured. Similar recapture percentages and speculation on fish
avoidance to boat electrofishing were reported by Jacobs and Swink
(19S2). They found, however, that differences in turbidities did not
affect capture efficiencies, although this.may have been due to their
study area not having as large changes in turbidities as our study did.
They further point out that use of electrofishing alone for mark and
recapture estimates in large rivers are generally unsuccessful because
not enough fish are recaptured.
In order to make accurate population estimates for resident fish other
than burbot in the mainstem Susitna River, methods have to be changed
from those used in 1983. Jacobs and Swink (1982) suggested using boat
electrofishing coupled with rotenone but this is not applicable to the
Susitna River. Electrofishing coupled with baited trapnets may prove
more successful, or large seining nets could be used to block the ends
of channels and sloughs • Another more difficult method would be the
use of population estimate models that allow for in-and outmigration
{open population models).
Population estimates for resident fish in tributaries to the Susitna
River can be made if enough fish of a given species are captured.
Population estimates of rainbow trout in Fourth of July Creek succeeded
because relatively large numbers of rainbow trout were captured and
recaptured and because there was little or no in-or outmigration during
-97 -
·------·------------------------
the sampling period. The time period of sampling was very important at
Fourth of July Creek. Sampling was conducted during mid-July because
the flows were extremely low and no adult salmon were in the tributary
(Estes and Vincent-Lang 1984). Biologists, therefore, had easy access
along the stream and the fish were easily caught because less food in
the form of salmon eggs was present in the system.
-98 -
-
-
....
-
-
-
-
PART 6
Resident Fish Habitat Studies
-
ABSTRACT
RESIDENT FISH HABITAT STUDIES
1984 Report No. 2, Part 6
by Paul M. Suchanek, Richard L. Sundet and t~lark N. Wenger
Alaska Department of Fish and Game
Susitna Hydro Aquatic Studies
2207 Spenard Road
Anchorage, Alaska 99503
The macrohabitat distribution and microhabitat suitability for rainbow
trout, Arctic grayling, round whitefish, and longnose suckers in the
Susitna River drainage between the Chulitna River confluence and Devil
Canyon were evaluated using electrofishing, beach seine, and hook and
line catch data and habitat data collected at radio telemetry relocation
sites (rainbow trout and burbot) and spawning sites (round whitefish).
Turbidity had important effects on distribution of both adult and
juvenile resident fish. Longnose suckers and juvenile round whitefish
wer.e found in highest numbers in turbid water. Adult rainbow trout,
Arctic grayling, and round whitefish found clear water more suitable,
but used turbidity for cover. Suitability criteria for velocity, depth,
and object cover were fit to the distribution of resident fish. The lo-
cation of radio tagged rainbow trout among macrohabitat types varied
greatly by season.
i
----------------·~
TABLE OF CONTENTS
Aas·rRACT •••••••••••••••••••••••••••••••••••••.••••••••••••••••••• i
LIST OF FIGURES................................................. iv
LIST OF TABLES ••••••.••.••.•••••.••••••••.••••.••••.
LIST OF APPENDIX TABLES ••••••••••••••••••••
1.0
2.0
3.0
INTRODUCTION ••••••••••••••••••••••.•••.••••••••••••••••••••
METHODS •••••••••••••••••••••••••••••••••••••••••••••••• o •••
2.1
2.2
2.3
Study Locations •••••••••••••••••••.•••••••••••••••••••
2 .1.1
2.1. 2
Macrohabitat studies •••••••••••••••••••••••••••
Microhabitat studies •••••••••••••••.••••••.••••
Field Data Collection •••••••••
2.2.1
2.2.2
Biological .................................... .
Habitat ........................................ .
Data Analysis ........................................ .
2.3.1
2.3.2
Macrohabitat studies ••••••••••••••••••••••.••••
Microhabitat studies •••••••••••••••••••••••••••
2.3.2.1
2.3.2.2
Adult resident fish •••.•••••••••••••••••
Juvenile resident fish ••••••••••••••••••
RESULTS ••••••••••••••••••••••••••••••••••••••••••••••••••••
3.1
3.2
Macrohabitat Distribution •••••••••••••••••••••••••••••
3.1.1
3.1.2
Adult resident fish ••••.•......•••••.•..•.•••..
Juvenile resident fish •••••••••••••••••••••••••
Mi crohab ita t Su itab i 1 i-ty •••••••••.••••••
3.2.1
3.2.2
Adult resident fish ••••••••••••••••••••••••
3.2.1.1
3.2.1.2
3.2.1.3
3.2.1.4
3.2.1.5
Rainbow trout .•.•.•.•..•...•.•..••.
Arctic grayling ••••••••••••.•.•••••
Round whitefish ••••••••••••••••••••
Longnose suckers •••••••••••••••••••
Burbot ••••••••••••.••••••••••••••••
Juvenile resident fish •.
ii
v
v
1
2
2
2
2
5
5
5
8
8
8
8
10
12
12
12
12
17
17
17
23
23
23
23
27
4.0 DISCUSSION................................................. 29
4.1 Adult Resident Fish................................... 29
4.2 Juvenile Resident Fish................................ 30
5.0 CONTRIBUTORS •••••••••••••••••••••••••••••••••••••• o........ 32
6.0 ACKNOWLEDGEMENTS........................................... 33
7.0 LITERATURE CITED .•••..•••• ,., ••.•.••••••••.•.••••••.••..•••.•
8.0 APPENDICES
Appendix A
Appendix B
Suitability indices for resident
species for cover, velocity,
and depth •...••...••... II ••• "' • "' ••••••••••••••
Round whitefish spawning micro-
habitat data ................................ .
iii
34
36
38
-
-
-
-
.....
.....
-
-
LIST OF FIGURES
Figure Title
1 Map of the Susitna River from the Chulitna River confluence
to Devil Canyon showing major tributaries and sloughs, 1983... 3
2 Arrangement of grids and cells at a hypothetical adult
resident fish microhabitat study site......................... 6
3 Frequency distribution of radio tagged rainbow trout lo-
cations in tributaries, at tributary mouths and in the
mainstem Susitna River during 1983 •••••••••••••••••••••••••••• 15
4 Rainbow trout, round whitefish, Arctic grayling, and
longnose sucker boat electrofishing mean catch per 1000 ft 2
in cells without object cover and all cells sampled by
turbidity level ................... ~··························· 20
5 Rainbow 2trout boat electrofishing mean catch (bars) per
1000 ft by habitat attribute values of depth, velocity,
percent cover, and cover type •.•••••••••••••••••••..••..•••..• 21
6 Rainbow trout hook and line mean catch per 1000 ft 2 by
habitat attribute values of depth, velocity, percent cover,
and cover type................................................ 22
7 Arctic ~rayling boat electrofishing mean catch (bars) per
1000 ft by habitat attribute values of depth, velocity,
percent cover, and cover type ••••. ,......... . • • • • • . . . • • • • • • • • • • 24
8 Round w2itefish, boat electrofishing mean catch (bars) per
1000 ft by habitat attribute values of depth, velocity,
percent cover, and cover type ••....••••••••.••••••••••••••..•• 25
9 Longnos~ sucker boat electrofishing mean catch (bars) per
1000 ft by habitat attribute values of depth, velocity,
percent cover, and cover type •...•.••.•••••.•••••••..•.•••.•.• 26
10 Juvenile round whitefish beach seining mean catch (bars) by
habitat attribute values of velocity, depth, and percent
cover ........................................................ .
iv
28
~~l
F""
~'ll!lll.lilj
-
LIST OF TABLES
Table
1
2
3
4
5
6
7
8
Title
Adult resident fish microhabitat sites on the Susitna River
between the Chulitna River confluence and Devil Canyon, 1983.. 4
Substrate, cover, and percent cover classifications used for
resident fish microhabitat studies •••••.••••••.••••••• ~....... 7
Habitat attribute groupings for analysis of boat electro-
fishing and hook and line data................................ 9
Boat electrofishing catch and catch per unit effort (CPUE) of
five resident fish species by three types of macrohabitats ••.• 13
Incidental catch of juvenile resident fish in cells by macro-
habitat sites on a mainstem discharge basis during Juvenile
Anadromous Habitat Study sampling ••••••••••.•••••••••••••.••• ~ 16
Percent catch per unit effort (CPUE) by macrohabitat type on a
mainstem discharge basis, for juvenile resident fish species
for which at at least 20 specimens were captured ••••...•..•••• 16
Catches and effort for boat electrofishing and hook and line
sampling of adult resident fish ••.••••••.••.•••••.•.•••••••••• 18
Kendall correlation coefficients {tau) between grouped habitat
variables and resident fish catches •••••••••.••••••.••••..•••• 19
LIST OF APPENDIX TABLES
Appendix Table Title
A-1 Suitability indices for resident species for cover,
velocity, and depth ........ -.................................. 37
B-1 Physical and chemical habitat characteristics of
spawning round whitefish in the Susitna River
during October 1983 .............................. ,. ....... ~~··· 40
v
1.0 INTRODUCTION
The Resident Fish Study was initiated in the fall of 1980 to gather
preliminary data concerning the following general objectives described
in 1979 by the Alaska Department of Fish and Game for the Susitna
Hydroelectric project:
A. Define seasonal distribution and relative abundance of resi-
dent fish in the Susitna River between Cook Inlet and Devil
Canyon.
B. Characterize the seasonal habitat requirements of selected
resident fish species within the study area.
During 1981, the primary emphasis was placed upon gathering seasonal
distribution and relative abundance data. In 1982, more effort was
placed upon characterizing the seasonal habitat requirements. During
the 1983 field season, the resident fish studies were refined. We
attempted to quantify the important habitat parameters associated with
spawning and rearing (growth) of selected resident fish species and
measure fish density in spawning and rearing habitats to provide an
estimate of habitat quality.
There can be positive or negative effects upon fisheries after the
construction of a hydroelectric dam (MOFW&P 1983). Postproject effects
may include changes in water temperature, flow, and turbidity.
Preproject baseline fisheries data and their correlation to habitat
conditions, therefore, are necessary to evaluate the overall potential
impact to these fisheries. One of these impacts can be the effect on
rearing fish.
Successful rearing of resident fish in the Susitna River is dependent
upon a variety of habitat conditions that may be substantially altered
under postproject flow regimes (AOF&G 1983c; 1983d). Four major macro-
habitats influenced by the mainstem were identified as possible rearing
areas in the Susitna River for resident fish (ADF&G 1983e). These four
major habitat types are tributary mouths, side sloughs, upland sloughs,
and mainstem channels or side channels. Macrohabitat information
reported in this report supplements ADF&G (1983e) as much less boat
electrofishing was done in 1983.
Microhabitat suitability criteria are one means of quantifying the
relationship of a 1 ife stage of a fish species to its habitat. The
present work develops preliminary suitability criteria by species and
river reach for application in incremental simulations of rearing
habitat as a function of mainstem flows (see Part 7 of this report).
Preliminary data presented for rainbow trout, Arctic grayling, round
whitefish, and longnose suckers are univariate functions for cover type,
percent cover, depth, and velocity. Frequency distributions by habitat
attribute were not generated for other resident fish species such as
burbot due to small catches. Differences between distributions in low
and high turbidity water were detailed as data permitted.
- 1 -
----··----·---·-----=--------------------........ -----------
2.0 METHODS
A two man crew conducted sampling on the Susitna River between the
Chulitna River confluence and Devil Canyon from May to October 1983 to
capture resident fish for micro-and macrohabitat studies (Figure 1).
Sampling was performed largely from a river boat, with occasional use of
helicopters. The primary sampling methods were boat electrofishing and
hook and line. Habitat data collected included water depth and veloc-
ity; cover, substrate, and water chemistry parameters.
2.1 Study Locations
2.1.1 Macrohabitat studies
Relative abundances of selected resident fish species were determined by
boat electrofishing at various macrohabitats in the Susitna River from
May to October. These macrohabitats included mainstem channels and side
channels, upland sloughs, side sloughs, and tributary mouths in the
reach of river between the Chulitna River confluence and Devil Canyon.
Also, 26 radio tagged rainbow trout were located in four major macrohab-
itats in 1983. These macrohabitats included tributaries, upland and
side sloughs, tributary mouths, and the mainstem. Radio tagged fish
were located at these sites in the Susitna River between RM 100.7 and
RM 148.8 from May 19 to October 21, 1983.
2.1.2 Microhabitat studies
Thirteen adult resident microhabitat study sites were sampled from July
to October to develop habitat suitability curves. These sites were
located between the Chulitna River confluence and Devil Canyon and
included six tributary mouths, three tributaries, three side s 1 oughs,
and one upland slough (Table 1).
Nine sites at sloughs and tributary mouths were selected for sampling by
boat electrofishing because relatively high numbers of adult resident
fish exist in these areas (ADF&G 1983b). The nine sites were sampled
with boat electrofishing gear twice a month from mid-July to October.
The upper reaches of four tributaries were irregularly sampled by hook
and line in conjunction with rainbow trout population estimates or
studies of radio tagged rainbow trout. (Presented in Part 5 of this
report).
Juvenile and a few adult resident fish were captured incidentally at 35
sites sampled during the juvenile anadromous studies reported in parts 2
and 3 of this report.
Microhabitat was also measured at relocation sites of 24 radio tagged
rainbow trout and burbot. These data were recorded at tributary mouths,
sloughs and sites in the mainstem Susitna River between RM 100.8 and RM
148.7 and at three tributaries.
- 2 -
-
-
-
-
-
-
~'
-
Slough 6A
Figure 1. Map of the Susitna River from the Chulitna River confluence
to Devil Canyon showing major tributaries and sloughs, 1983.
-3-
a/TRM = tributary river mile
-
-
-
- 4 -
-
-
....
-
2.2 Field Data Collection
2.2.1 Biological
Adult and a few juvenile (under 200mm) resident fish were captured at
accessible locations in the Susitna River with a boat mounted electro-
fishing unit. Electrofishing equipment consisted of a Coffelt, model
VVP-3E, boat electrofishing unit powered by a 2500 watt Onan portable
generator. Boat electrofishing procedures are described in AOF&G
(1983a). Adult resident fish were also captured by hook and line in
tributaries. Juvenile resident fish at upland slough, side slough,
mainstem and tributary sites were collected with beach seines and
backpack electroshockers.
All resident fish were identified to species. Biological data collected
included length, sex, and sexual maturity. Ages were determined by
readi-ng scale samples. All healthy adult resident fish were tagged with
a Floy anchor tag and released in continuance of a resident fish
migrational study described in part 5 of this report. Spawning sites of
resident fish species were determined when captured female fish expelled
eggs upon slight palpation of the abdomen.
Juvenile resident fish were captured incidentally during juvenile
anadromous sampling of cells and grids located at a greater diversity of
sites. Techniques differed somewhat as beach seining and backpack
electrofishing were used (see Part 2 of this report for details on
collection methods).
Microhabitat data were collected from relocations of four burbot and 20
rainbow trout radio tagged in 1983. Tagging techniques are presented in
ADF&G (1981, 1983a) and part 5 of this report. Radio tagged fish were
tracked from airplanes and boats. A summary of capture and tracking
locations of the tagged fish are presented in Part 5 of this report •
Habitat measurements were taken after a radio tagged fish was relocated
by boat to an area of no greater than 30 feet by 30 feet. In some
cases, radio tagged fish were observed.
2.2.2 Habitat
Each microhabitat study location was divided into one to three grids.
Grids were located so that the water quality within them was as uniform
as possible and so that the grids would encompass a variety of habitat
types. At tributary mouths, one grid was located in the mainstem
Susitna River above the confluence of the tributary, another grid was
set up within or below the confluence where the tributary was the
primary water source, and a third grid was situated where the mainstem
and tributary waters mixed (Figure 2). Sites located in sloughs and
tributaries had one to three grids depending on the water quality within
the slough. Since grid location was dependent upon specific hydraulic
characteristics, grid locations were redetermined during each sampling
trip based on differences in turbidity and water chemistry readings.
Grids were subdivided into cells. Cells were rectangular and the 1ength
and width of each ce 11 varied. The 1 ength boundaries of ce 11 s within
- 5 -
lli§
0
II
MA.INSTEM SUSITNA
R IV E R W ATE R ( G R I 0 I )
TRIBUTARY WATER
(GRID 2)
MIXING ZONE WATER
(GRID 3)
Figure 2. Arrangement of grids and cells at a hypothetical adult
resident fish macrohabitat study site.
-6-
-
-
-
-
....
-
-
-
-
each grid were marked with orange flagging prior to sampling. The width
of cells in tributaries, which were sampled by hook and line, was the
width of the stream. Cell widths at sloughs and tributary mouths, which
were sampled by boat electrofishing, were determined to be five feet or
a multiple of five feet. Five feet was chosen as a standard cell width
because it is the average effective capture width of the boat
electrofishing equipment used.
This method of sampling was designed to approximate the method that the
"instream flow incrementa 1 methode 1 ogy" uses to generate estimates of
usable habitat (Bovee 1982, also see Part 7 of this report). The
correlation of fish occurrence in cells with a particular set of phys-
ical parameters can be compared with the calculated usability of the
habitat.
Habitat parameters measured within cells and at radio tagged fish
relocations included dissolved oxygen, specific conductance, pH,
turbidity, water temperature, water velocity, and water depth.
Substrate type, cover type, and percent cover were estimated (Table 2).
Intragravel temperatures were also recorded at all spawning sites.
Table 2: Substrate, cover, and percent cover classifications used for
resident fish microhabitat studies.
Substrate
Silt
Sand
Small Gravel (l/8 11
-1")
Large Gravel (1" -3")
Rubble (3" -5")
Cobble (5" -10")
Boulder ( > 10")
Cover Type
No Cover
Emergent Vegetation
Aquatic Vegetation
Debris/Deadfall
Overhanging Riparian
Undercut Banks
Large Gravel 1" -3"
Rubble 3 11
-5"
Cobble or Boulder> 5"
% Cover
0 -5%
6 -25%
26 -50%
51 -75%
76 -95%
96 -100%
The mean depth of cells and radio tagged fish relocation sites was
measured to the nearest tenth of a foot with a topsetting wading rod.
The mean velocity was measured with a Price Model AA velocity meter.
Turbidity measurements were made with an HF Instrument Model DRT -15
turbidometer in Nephelometric Turbidity Units (NTU•s). Water quality
measurements were taken with a Hydrolab model 4001 multi parameter
meter.
Habitat parameters were recorded for each cell at resident. fish micro-
habitat study sites. However if the water quality within a grid were
relatively constant, only one measurement was taken to represent all
cells within that grid. Specific data collection methodology is sum-
marized in ADF&G (1984).
- 7 -
2.3 Data Analysis
2.3.1 Macrohabitat studies
Biological and catch per unit effort (CPUE) data were compiled by
macrohabitat type from boat electrofishing sampling data recorded in
conjunction with distribution and re 1 ati ve abundance studies presented
in Part 5 of this report. Macrohabitat CPUE data were also compiled by
pooling the catch from all the cells at microhabitat study sites sampled
by boat electrofishing. The macrohabitat type of radio tagged fish
relocation sites was also recorded.
Catch data recorded by Juvenile Anadromous Habitat Study (JAHS) crews
were also compiled by macrohabitat type for incidentally captured
juvenile resident fish. Mean CPUE•s were calculated by macrohabitat
type, summed, and then each CPUE by type was expressed as a percentage
of the total to equalize sampling effort. These percentages were then
used to analyze distribution by macrohabitat type. Macrohabitat types
were defined with the discharge based classification scheme discussed in
Part 2 of this report.
An analysis of variance (ANOVA) was run to determine whether macrohabi-
tat type had a significant effect on the relative abundance of juvenile
round whitefish (see Part 2 of this report for further details).
2.3.2 Microhabitat studies
2.3.2.1 Adult resident fish
Biological, habitat and catch data were recorded at microhabitat study
sites according to ADF&G (1984). Adult fish microhabitat studies used
two gear types, boat electrofishing and hook and line. Hook and line
was used in tributaries, while boat electrofishing was used elsewhere.
Hook and line data were analyzed separately from boat electrofishing
data since the area each gear type sampled was very different in water
quality and habitat characteristics.
Values of habitat attributes measured had to be pooled for analysis
because of small sample sizes. Groupings for the boat electrofishing
and hook and line data are detailed in Table 3. Groupings for the
rainbow trout hook and line catch data were somewhat different than the
boat electrofishing data because of small sample sizes and different
cover types sampled.
Turbidity values were also grouped into three categories to determine
the effects of low, moderate and high turbidities on resident fish
distribution. The three turbidity groupings used were: 1 to 9 NTU, 10
to 30 NTU and greater than 30 NTU. Turbidity inflection points at 9 NTU
and at 30 NTU were used because light penetration changes considerably
at these points in other glacial systems in Alaska (Jeffery Koenings,
pers. comm.) and because chinook salmon fry used turbidities of greater
than 30 NTU for cover (see Part 3 of this report).
- 8 -
,..,.
-
.....
-
-
-
-
-
-
....
!"""'
r
-
....
r-
-
Table 3. Habitat attribute groupings for analysis of boat electro-
fishing and hook and line data.
Boat Electrofishing Habitat Attribute Groueings
Velocity Depth
Grouping Grouping Percent
No. {ft/sec) 1f.tl Cover Cover txEe Substrate
1 0 0.7-2.0 0-5% No cover Silt -1"
2 0.2-1.0 2.1-2.9 6-25% Emergent or 1-3"
aquatic vegetation
3 1.1-2.0 3.0-4.4 26-50% Debris or overhanging 3-5"
riparian vegetation
4 2.1-3.0 4.5 + 51+% Large gravel (1-3") sn+
5 3.1 + Rubble (3-5 11 )
6 Cobble or boulder (5"+)
Hook and Line Habitat Attribute GrouEings
Velocity Depth
Grouping Grouping
No. ( ft/sec} 1!12. Cover txee
1 0-0.5 0. 5-l. 0 No cover
2 0.6-1.0 1.1-2.0 Debris, under cut banks
or overhanging riparian
vegetation
3 1.1-1.5 2.1 + Cobble or boulder (5"+)
4 1.6+
Percent cover and substrate groupings same as for boat electrofishing
data.
- 9 -
After habitat attribute values were grouped, Kendall rank-order corre-
lation coefficients were calculated between the habitat attributes and
catch for the resident species for both the boat electrofishing and hook
and line data. Since cells varied s~nificantly in size, catch was put
on an area basis as catch per 1000 ft of surface are2. Density of fish
was assumed to be a function of catch per 1000 ft . Suitabi 1 ity of
habitat was reflected by this number as fish density can be assumed to
reflect fish habitat suitability.
The distributions of mean catches by species were examined for the
habitat attributes of velocity, depth, cover type, and percent cover.
Velocity was thought to be an important determinant of distribution and
therefore suitability criteria were fit by hand using professional
judgement to the distributions of catch by grouped velocity interval for
all four species. Since we had no data for velocities greater than 4.3
ft/sec, we assumed that suitability for all species was 0 for velocities
greater than 4.5 ft/sec.
Depth was not thought to be as important a determinant of distribution
and therefore we did not fit su itabi 1 i ty criteria to any of the depth
distributions. Depth, however, may be important in limiting dis-
tribution on the shallow end. Wesche (1976), for example, reported that
adults of three trout species preferred depths greater than 0.5 ft.
Raleigh et al. (1984) reported that rainbow trout found depths of less
than 1.5 ft less suitable than greater depths. We conservatively set
depth suitability to 1.0 for all depths greater than 0.6 ft and
suitability to 0 for depths less than 0.5 ft.
Percent cover and cover type both were believed to have potentia 1
importance in determining adult fish distribution, however, sample sizes
limited us to consider only cover type. We believed the cover type data
were most reliable and also these data showed clear differences in
usability of the different cover types. Since the turbidity data
indicated that as turbidity increased, suitability of no cover cells
increased, we integrated these data into suitability indices for cover
type by turbidity level. Cover type suitability indices for both clear
( ~ 10 NTU) and turbid ( > 30 NTU) conditions were developed. The
suitability of 11 no cover" cells {cells without object cover) at these
two levels was different. The suitability of the uno coveru cells was
set as a minimum, therefore if other cover types had mean catches less
than those of the no cover cells then suitability for these types were
changed to the suitability value for the no cover cells. Since there
were no boat electrofishing data for the cover type, undercut banks, we
assumed that undercut banks had a suitability equal to that for over-
hanging riparian vegetation and debris which provide a somewhat similar
type of cover.
2.3.2.2 Juvenile resident fish
Only round whitefish juveniles were captured in sufficient numbers at
the juvenile salmon study sites to warrant development of microhabitat
suitability indices. The habitat attributes of velocity, depth, percent
cover and cover type were examined for criteria development. Beach
seining data from water over 30 NTU in turbidity were used in the
-10 -
-
-
-
-
-
·-
-
analysis as catches were highest for this gear type at this turbidity
level.
Due to small sample sizes, groupings of velocity values were by 0.3
ft/sec increments and depths by 0.5 ft increments. Cover type analysis
was only qualitative due to small sample sizes and the inefficiency of
beach seines in different cover types. Round whitefish suitability was
measured as mean catch per cell, as this numbe~was assumed to reflect
density because cell size was constant at 300ft. In general, analysis
was the same as that used to develop criteria for juvenile chinook
salmon in turbid water (see Part 3 of this report).
An analysis of variance (ANOVA) was run to determine the effect of the
site parameters: mean depth, mean velocity, mean percent cover, water
temperature, and turbidity on the relative abundance of juvenile round
whitefish (see Part 2 of this report for further details on the methods
used).
-11 -
3.0 RESULTS
3.1 Macrohabitat Distribution
3.1.1 Adult resident fish
Boat electrofishing catch and catch per unit effort (CPUE) for five
resident fish species in three types of macrohabitats was detennined in
1983 (Table 4). Since sampling was not as intensive in 1983 as in 1982,
the category "sloughs 11 includes both upland sloughs and side sloughs.
Sampling effort in 1983 (45.9 boat electrofishing hours) was small in
comparison to 1982 efforts (177.6 total boat electrofishing hours, with
63.9 hours above the Chulitna River confluence).
Radio telemetry was used to study movements of rainbow trout among
macrohabitat types. Movements of adult rainbow trout in the Susitna
River can be placed into three major categories based on their annual
life history, those associated with spawning (April-June), those associ-
ated with summer rearing (July-September) and those associated with
overwintering (October-March). Distribution of radio tagged rainbow
trout in or at the mouths of tributary streams and at mainstem areas
changed with season (Figure 3). Radio tagged rainbow trout were located
in tributaries and at tributary mouths more often during spawning and
summer rearing periods than during the winter. Between April and June,
67% of the radio tagged rainbow trout locations were associated with
tributaries, the majority being in tributaries (52%). During July
through September, 61% of the radio tagged rainbow trout were associated
with tributaries, the minority being located in tributaries. By October
1, all radio tagged rainbow trout had outmigrated from tributaries and
sloughs into mainstem influenced areas. About 33% of the radio tagged
rainbow trout remained at tributary mouths from October to December.
Besides the high incidence of rainbows using tributaries from April to
September, about 10% used Slough 9 (RM 128.3), Slough 8A (RM 125.3),
Slough A (RM 124.7), and Moose Slough (RM 123.5) during July through
September.
Often radio tagged rainbow trout moved from one tributary or slough to
another tributary or slough (refer to Part 5 of this report for indi-
vidual trout movements). For example, five radio tagged rainbow trout
migrated 7.5 miles downriver from the mouth of Indian River (RM 138.6),
to the mouth of Fourth of July Creek (RM 131.1). In addition, a rainbow
trout moved 6.5 miles upriver from the mouth of Skull Creek (RM 124.7)
to the mouth of Fourth of July Creek, and then 2.6 miles downriver to
Slough 9. Another rainbow trout spent over one week in two different
sloughs (8A and A) before holding in Moose Slough for over three weeks.
Finally, a rainbow trout outmigrated from Fourth of July Creek (TRM 1.5)
and moved 7.5 miles upriver to Indian River where it was last located at
TRM 4.5.
3.1.2 Juvenile resident fish
Incidental catches of juvenile and a few adult resident fish were made
during juvenile anadromous habitat study (JAHS) sampling (Table 5).
Large differences in the distribution of juvenile fish by macrohabitat
-12 -
-
-
-I
I
j
-
-
I
1-' w
I
l .I
Tabla 4. Boat electrofishing catch and catch per unit effort (CPUE) of five resident fish species by three
types of macrohabits. Resident fish species sampled are rainbow trout, burbot, Arctic grayling, round
whitefish, and longnose suckers. CPUE is in parentheses, and the units are catch per minute.
-------------------------------· -----------------
HACROIIABITAT TYPE
RAINBOW TROUT
HAIKSTEH
SLOUCII
TRIBUTARY HOUTH
TOTAL
BUR BOT
tiAINSTEM
SLOUGH
TRIBUTARY MOUTII
TOTAL
MAY
lli-31
6( .0)
4( .I)
7( .0)
JUN
1-15
5( .o>
2( .0)
7( .1)
JUN
16-30
I( .0)
I( .0)
9( .I)
17( .0) 14( .0) 11( .1)
6( .0)
I ( .0)
o< o.o>
7( .o)
3( .0)
O( ().0)
2( .0)
5( .0)
O( 0.0)
o< o.o)
3( .0)
3( .o)
JUL
1-lS
0( 0.0)
I ( .0)
JUL
16-31
I( .0)
3( .0)
4( .1) II( .2)
5( .0) 15( .1)
0( 0.0)
4( .I)
0( 0 .o)
4( .o)
6( .o>
3( ,1)
AUG
1-15
AUG
16-31
0( 0.0) ---(----)
I( .1)
SEP
1-15
SEP
16-30
7( .0) 13( .0)
o< o.o> 1( .1)
OCT
1-15 TOTAL
8( .o> 41( .o>
2( .1) 16( .0) I( .0)
3( .0)
4( .0)
4( .3) 19( ,2) 16( .2) 14( ,2) 94( .I)
5( .2) 26( .1) 30( .1) 24( ,1) ISH .1)
I( ,0) -(--)
I ( .0)
8( .I)
0( 0,0)
0( 0,0)
9( .o>
I( .0)
0( 0.0)
1( ,0) 31( .0)
0( 0.0) 14( ,0)
1( .0) 17( ,0)
4( .0) 13( .0) 10( .0) O( 0.0) 10( .0)
1( .o)
1( .1)
O( 0.0)
8( .0) 2( ,0) 62( .0)
ARCTIC GRAYLING
MAINSTEH
SLOUGH
TRIBUTARY MOUTH
TOTAL
-= No effort .
. 0 = Trace.
63( .2) 78( .4) 40( 1.1) 0( 0.0) 28( .3) 32( .6) ---(--) 99( .4) 195( .7) 19( .1) 554( .4)
23( .3) 22( .4) I( .0) I( .0) 5( .0) 1( .0) 5( .3) 4( .1) 17( 1.3) 2( .1) 81( ,2)
50( .3) 26( .2) 31( .3) 18( .3) 56( .9) 24( .2) 7( ,5) 66( .6) 87( 1.1) 14( .2) 379( .4)
136( .3) 126( .4) 72( ,4) 19( .1) 89( .3) 57( .2) 12( .4) 169( .4) 299( .8) 35( .1) 1014( .4)
Table 4 continued.
---------------KACROHABITAT TYPE HAY JUN JUN JUL JUL AUG AUG SEP SEP OCT
16-31 1-15 16-30 i-1!) . 16-31 1-15 16-31 1-l!i 16-30 1-lS TOTAL -----------------.-----------------------------------------------------------
·~<~h
ROUND WHITEFISH
MAINSTEM 2S( .1) 82( .4) 21( .6)
\
O( o.o) 31( .3) 20( .4) --(--) 147( .6) 101( ·.4) 78( .4) SOS( ,4)
SLOUGH 7( .1) 11( .2) 3( .1) 4S( ,6) 142( 1.0) 8( .2) 3( .2) 1 S( .4) 7( .S) 8( .4) 249( .S)
TRIBUTARY HOUTII 26( ,2) 4S( .4) 36( .4) 61( 1 .2) 71( 1.2) 72( .S) S( ,3) 108( 1.0) 66( • 8) n< 1.0) S6S( .6)
I
~ TOTAL 58( .1) 138( .4) 60( .4) 106( .7) 244( • 8) 100( ,4) 8( .3) 270( .7) 174( .s> 161( .6) 1319( .5) """ I
LotiGNOSE SUCKER
MAINSTEH 1( .o) 3( .o> 5( .I) 0( 0.0) 29( .3) 13( .2) -<--> 6S( .3) 16( .1) 3( .o) 135( .1)
SLOUGH 2( .o) 13( .2) 9( .3) 33( .4) 51 ( .4) 16( .4) O( o.o) 7( .2) 4( .3) 0( 0.0) 135( .3)
TRIBUTARY MOUnt 0( o.o> 4( .0) IS( .1) 4( .I) 10( .2) 56( ,4) 0( o.o> 18( .2) 23( .3) 2( .o) 132( .1)
TOTAL 3( .0) 20( .1) 29( .2) 37( .3) 90( .3) 85( .4) 0( 0,0) 90( .2) 43( .1) S( .o) 402( .1)
,J j ] ) I I I I J
-
r-
-
>-Spawning n = 21
(.) 80 (April-June) z
IJ.I
:::::> 60 0
IJ.I
If 40 ... z
t&J 20 (.) a:
LLJ a.
>-Summer Rearing n= 112
(.) 80 (July-September) z
LLJ
:::::> 60
0 LLJ a:
LL. 40 ... z
IJ.I 20 (.) a: LLJ a.
>-Over-Wintering
n=98 (.) (October-December) z
LLJ
:::::>
0 LLJ
0::
LL. 4
t-z
UJ 2 (.)
0::
LLJ a.
LOCATIONS IN LOCATIONS IN LOCATIONS IN
TRIBUTARIES OR MAINSTEM AT MAINSTEM
SLOUGHS TRIBUTARY MOUTHS
Figure 3. Frequency distribution of radio tagged rainbow trout
locations in tributaries, at tributary mouths, and in the
mainstem Susitna River during 1983.
-15-
Table 5. Incidental catch of juvenile resident fish in cells by
macrohabitat sites on a mainstem discharge basis during
Juvenile Anadromous Habitat Study sampling.
Mains tern
Upland Side Side-
SEecies Tributaries Sloughs Sloughs channels Total
Rainbow trout 6 3 1 1 11
Arctic grayling 1 20 21
Round whitefish 1 20 7 601 629
Longnose sucker 20 33 66 119
Dolly Varden 21 21
Burbot 9 3 6 18
Humpback whitefish 11 11
Effort (cells fished) 236 131 455 463
Table 6. Percent catch per unit effort (CPUE) by macrohabitat type on a
mainstem discharge basis for juvenile resident fish species
for which at least 20 specimens were captured.
Mains tern
Upland Side Side-
Tributaries Sloughs Sloughs channels
Arctic grayling (n=21) 8.9% 0.0% 0.0% 91.1%
Round whitefish (n=629) 0.3% 10.4% 1.0% 88.3%
Longnose sucker (n=119) 0.0% 41.5% 19.7% 38.8%
Dolly Varden (n=21) 100.0% 0.0% 0.0% 0.0%
-16 -
-
-
-
-
-
~
~
-
-
-·
-
,...
-
type are evident in this tab 1 e. The analysis of variance of round
whitefish distribution showed that macrohabitat type does have a signif-
icant (p < 0.01) effect on distribution. In order to adjust for differ-
ences in sampling effort among the macrohabitat types, CPUE on a per-
centage basis was calculated for the four species for which more than 20
individuals were captured (Table 6). Arctic grayling and round
whitefish juveniles were most numerous at mainstem side channels while
Dolly Varden were captured only in tributaries. Longnose suckers were
distributed primarily in upland sloughs and mainstem side channels
although they were also caught in side sloughs.
3.2 Microhabitat Suitability
3.2.1 Adult resident fish
Boat electrofishing catches of rainbow trout, Arctic grayling, round
whitefish, and longnose suckers were sufficient to be analyzed for
microhabitat suitability criteria development. Hook and line catches of
rainbow trout were also sufficient. Total catches by species and number
of cells fished are listed in Table 7. Additional measurements of
microhabitat were taken at telemetry locations of 20 rainbow trout and
four burbot and these are available at the ADF&G Susitna Hydro Aquatic
Studies office. These telemetry data cannot be used for criteria
development but they supplement our knowledge of microhabitat use.
Kendall rank-order correlation coefficients between grouped habitat
attributes and fish ~atches are listed in Table 8. Since substrate is
partially a subset of cover type and also was highly correlated
(tau=0.61) with velocity, it was dropped from consideration for further
analysis.
Turbidity was the habitat attribute most highly correlated with longnose
sucker mean catch. Graphs of turbidity level versus mean catch indicat-
ed turbidity has an influence on distribution of rainbow trout, round
whitefish, Arctic grayling, and longnose suckers (Figure 4). Plots of
catch in the 11 no cover 11 cells by turbidity value also suggest that these
four species use turbidity for cover. Mean tzainbow trout, Arctic
grayling, and round whitefish catches per 1000 ft were lower in turbid
waters, however.
3.2.1.1 Rainbow trout
Rainbow trout were typically captured by boat electrofishing in cells
with water velocities less than 1.5 ft/sec (Figure 5). Favored cover
types included rocks with diameters over 311
, and secondarily, debris and
overhanging riparian vegetation. Rainbow trout used cells with 6 to 25%
and greater than 50% object cover in the highest densities.
Hook and line sampling data suggested that rainbow trout preferred pools
with velocities less than 0.5 ft/sec and depths greater than 2.0 ft
(Figure 6). Rainbow trout captured by hook and line sampling used
debris, undercut banks, and riparian vegetation more than they did
cobble or boulders. An abundance of cover also appeared to be tied to
rainbow distribution.
-17 -
Table 7. Catches and effort for boat electrofishing and hook and line
sampling of adult resident fish.
Boat electrofishing sampling
No. of cells sampled = 176
Species Catch
Rainbow trout 44
Arctic grayling 138
Round whitefish 384
Longnose sucker 157
Burbot 18
Humpback whitefish 15
Dolly Varden 2
-18 -
Hook and line sampling
No. of cells sampled = 79
Species
Rainbow trout
Arctic grayling
Catch
99
2
-
-
-
-
1
........
1..0
J ] 1
Table 8. Kendall correlation coefficients (tau) between grouped habitat variables and resident fish
catches.
Boat Electrofishing Data (n = 176}
Percent Cover Rainbow Arctic
Turbidity Cover ~ Velocity Depth Substrate Trout Grayling
Percent cover -0.07 1.00
Cover type -0.22** 0.45** 1.00
Velocity -0.08 0.10* 0.45** 1.00
Depth -0.27** 0.16** 0.43** 0.34** 1.00
Substrate -0.16** 0.33** 0.61** 0.54** 0.32** 1.00
Rainbow Trout -0.14* 0.21** 0.22** 0.11 0.11 0.20** 1.00
Arctic grayling -0.13 0.18** 0.36** 0.33** 0.27** 0.29** 0.20** 1.00
Longnose sucker 0.34** 0.19** -0.15* -0.25** -0.22** -0.25** -0.04 -0.07*
Round whitefish 0.05 0.19** 0.20** 0.10 0.11 0.10 0.15* 0.34**
Hook and Line Data (n == 79}
Percent Cover
Cover ~ Velocit~ Depth Substrate
Cover type -0.10
Velocity -0.30** 0.38**
Depth 0.59** -0.09 -0.42**
Substrate -0.04 0.53** 0.28** -0.02
Rainbow Trout 0.42** 0.04* -0.29** 0.35** 0.08
* Signif1cantly different from 0 at p ~ 0.05
** ~innifirnntlv rliffPrPnt frnm 0 at n < 0,01
Long nose
Sucker
1.00
0.18**
N -.....
0
0
Q
a:: w
tL.
%
(,)
I !;( N a (,)
I z
ct w
::E
~ Cot~:h in cella without object cover fl Catch in all celle sampled
I :!: Standard Error
RAINBOW TROUT
ARCTIC GRAYLING
I0-30
ROUND WHITEFISH
3.0
N=46
LONGNOSE SUCKER
10-30
TURBIDITY (NTU)
N=31
N=75
Figure 4. Rainbow trout, round whitefish, Arctic grayling, and longnose
sucker boat electrofishing mean catch per 1000 ft2 in cells
without object cover and all cells sampled by turbidity
I )
01 0.25 --0
0
0 0
a::
ILl
Q.
:::.:
0
I ti N ,_. 0
I z
c(
ILl
2 0.75
0.50
0.25
0
l
N=IO
1.0 2.0 3.0 4.0 5.0 6.0 0.5
DEPTH (ft)
N=IO
N=46
0.25
0
o-sa;., 6-25% 26-50"/o 51t% NO
COVER
PERCE NT COVER
1.0 1.5 2.0
RAINBOW TROUT
BOAT ELECTROFISHING I :!:Standard Error
2.5 3.0 3.5 4.0 4.5
VELOCITY (ft/aec)
N=54
EMER. DEBRIS GR., fYs·• ~c5o,.
OR OR 1•3
AQUATIC RIPARIAN
VEG. VEG.
COVER TYPE
Figure 5. Rainbow trout boat electrofishing mean catch (bars) per 1000
ft 2 by habitat attribute values of depth, velocity, percent
cover, and cover type. Suitability index (line) for velocity
0.50
0.25
(/)
c:
-i
l>
!:!:!
!:
-i -<
z
0 ...,
X
0.25
6.0
N --0
0 2.0 Q
a: w
Q..
1.0
:1:
~ < 0
I z N < N w I :E
o-5%
Figure 6.
. I .J
N=20
2.0 3.0 5.0 0.5 1.0 1.5 2.0 2.5
DEPTH (ft) VELOCITY Ut/aec)
N=l4
N=l8
4.0
2.0
0
s-25% 26-50% 51•% NO DEBRIS COBBLE
COVER OR >5"
RIPARIAN
VEG.
PERCENT COVER COVER TYPE
Rainbow trout hook and line mean catch per 1000 ft 2 by
habitat attribute values of depth, velocity, percent cover,
and cover type .
) .I J · ... 1
....
Since electrofishing data were collected at more cells in a wider
variety of habitat types, velocity and cover type suitability indices
were fit to the boat electrofishing data (Figure 4). Since the hook and
line data suggested that cover types of debris, overhanging riparian
vegetation, and undercut banks were more suitable than cobble or boul-
ders (Figure 5), suitabilities for these cover types were changed to the
suitability of cobble and boulders which was 1.0. A listing of
suitability criteria point values for rainbow trout (along with all
' other suitability criteria developed in this report) is contained in
Appendix Table A-1.
3.2.1.2 Arctic grayling
Adult Arctic grayling often used rocks for cover and water with high
velocities and deep depths (Figure 7). Arctic grayling may avoid high
turbidity waters and make little use of turbidity for cover {Figure 4).
Suitability criteria were fit to the velocity and cover type dis-
tributions of catch (Figure 7 and Appendix Table A-1).
3.2.1.3 Round whitefish
Distribution of round whitefish was influenced by turbidity as they used
it for cover (Figure 4). Round whitefish also used object cover in the
form of cobble or boulders, debris, and overhanging riparian vegetation
most highly {Figure 8). The hydraulic attribute of velocity was not
strongly tied to distribution, although optimum velocities ranged from
two to three ft/sec. Suitability criteria were fit to the velocity and
cover type distributions of catch (Figure 8 and Appendix Table A-1).
Seven spawning sites for round whitefish were found in October 1983.
Three of the sites were at tributary mouths while the other four sites
were in the mainstem. Microhabitat data collected at these sites are
presented in Appendix B, along with a brief discussion of round
whitefish spawning in the Susitna River.
3.2.1.4 Longnose suckers
Longnose suckers often used turbid water for cover (Figure 4), but they
also used emergent or aquatic vegetation, debris and overhanging
riparian vegetation cover {Figure 9). Shallow depths and waters of low
velocity were most suitable for longnose suckers. Suitability criteria
were fit to the velocity and cover type distributions of catch (Figure 9
and Appendix Table A-1).
3.2.1.5 Burbot
Burbot prefer areas of moderate to high turbidities since catch data
show they are always in the mainstem during the summer {ADF&G 1983e).
Telemetry data also showed they were always found in the mainstem.
While in these mainstem areas, radio tagged burbot appeared to prefer
low velocities {under 1.5 fps) and shallow depths (approximately 2.5
feet). They also appeared to prefer areas with rubble or cobble
substrate, however, nearly all of the mainstem river between the
Chulitna River confluence and Devil Canyon, where the radio tagged fish
were found, has a predominately rubble or cobble substrate.
-23 -
3.5
3.0
2.5
"' ... ..
g
0
0:
LLI a.
I J:
N 0
+:> l;i I
0
z
~
LLI
2
.]
N=24
3.0
2.0
1.0 2.0 3.0 4.0 5.0 6.0
DEPTH (ft)
2.!1
N=IO
2.0
N=46
1.5
1.0
0.!1
0
PERCENT COVER
ARCTIC GRAYLING
BOAT ELECTROFISHING I:!: Slandard Errar
0.5 1.0 1.5 2.0
N=40
2.5 3.0 3.5
VELOCITY (ft /ate)
NO EMER. DEBRIS GR. Ru ••
COVER OR OR 1-3 3-5
AQUATIC RIPARIAN
VEG. VEG.
COVER TYPE
4.0
Figure 7. Arctic grayling boat electrofishing mean catch (bars) per
1000 ft 2 by habitat attribute values of depth, velocity,
0-. .-., ,,.. ,..., ,.... '"' ..t-'"",...._, , _ , ,., ! ("' • t ' • , • o I f , • \ ,-.
.. I
en c
4.5 ;!
CD
r-::;
-<
z
N=54 0
I'll
)(
1.00
0.75
0.50
0.2!1
0
.,c~
I
N
U1
I
0
~
0:
11..1 a..
J:
0
!;(
0
z
c(
11..1
::E
4.0
3.0
2.0
t.O
0
4.0
3.0
4.0
3.0
2.0
0
1.0 2.0 3.0 4.0 5.0 6.0
DEPTH (ft)
4.0
3.0
N=46 N=32
2.0
1.0
PERCENT COVER
0.5 1.0
ROUND WHITEFISH
BOAT ELECTROSHOCKING J :! Standard Error
1.5 2.0 2.5 3.0 3.5 4.0 4.5 VELOCITY ( ft /lee)
N=54
NO EMER. DEBRIS I~~-..
COVER OR OR ~
co.
>5"
AQUATIC RIPARIAN
VEG. VEG.
COVER TYPE
Figure 8. Round whitefish boat electrofishing mean catch (bars} per
1000 ft 2 by habitat attribute values of depth, velocity,
percent cover, and cover tvoe. Suitabilitv index (line) for
1.00
0.75
0.50
0.25
z
0
fTI
)(
I
N
0'1
I
]
C\1 --0
0
Q
a::
"' Q.
:I:
0
!;(
0
z
oc( w
2
1.0
2.0
LONGNOSE SUCKER
BOAT ELECTROFISHING It Stondord Error
1.0 20 3.0 40 5.0 6.0
DEPTH (ft)
0.5 1.0 1.5 2.0 2.5 3.0 35 4.0 4.5
Figure 9.
VELOCITY (ftlaec)
N=IO
NO EMER. DEBRIS GR,.
COVER OR OR I -3
AQUATIC RIPARIAN
VEG. VEG.
PERCENT COVER COVER TYPE
Longnose sucker boat electrofishing mean catch {bars) per
1000 ft2 by habitat attribute values of depth, velocity,
I
~g ..
(/) c
i.!
CD
r
=i -<
2:
0
fT1
X
""""
'-+'
.....
3.2.2 Juvenile resident fish
The analysis of variance showed that turbidity had a significant
(p< 0.01) effect on the relative abundance of juvenile round whitefish.
Catch rates in water with a turbidity less than 30 NTU were extremely
1 ow.
The total catch of round whitefish by beach seines in turbid (greater
than 30 NTU) water was 569, and most of these were 0+ juveniles. Mean
catches by velocity, depth and percent cover interva 1 suggest that
velocity had the largest effect on distribution in the 320 cells fished
(Figure 10). Juvenile round whitefish greatly preferred water without a
significant velocity. Catches in cells with little object cover were
higher than in cells with large amounts of cover. This suggests that
object cover is not very significant in influencing round whitefish
habitat use. Beach seining efficiency is greatly reduced, however, by
the amount and type of cover present, and therefore catch distribution
by cover type has not been presented. The data suggest that round
whitefish fry also find shallow depths most suitable.
A suitability index was fit to both the depth and velocity catch dis-
tributions by hand using professional judgement. Pearson correlation
coefficients between the fitted suitability criteria for depth, veloci-
ty, and (depth x velocity) and juvenile round whitefish catch by cell
were calculated. The correlations between juvenile round whitefis.h
catch and depth, velocity, and (depth x velocity) were 0.23, 0.42, and
0.50 (n=320, p< 0.001 for all three), respectively. Since depth was
correlated with catch, we decided to use depth as fitted in subsequent
habitat modelling. Suitability for turbid water for all cover types was
set to 1.0 and suitability for all cover types in clear water was set to
0 (Appendix Figure A-1).
Catches were insufficient for any other species of juvenile resident
fish to be analyzed for criteria development.
-27 -
JUVENILE ROUND WHITEFISH
BEACH SEINING I !Standard Error
(/)
1.0 c:
::::j
)>
0.8 m
F
.....
0.6 -<
z
0.4 0
1'11
X
0.2
W" 0.0
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 3.0
VELOCITY (ft /sec)
.J
.J
1&.1 u
a: (/)
1&.1 c:
Q. .....
%
)>
u m
.... r c :::j u -<
z c z
1&.1 0
2 1'11
X
1.0 1.!5 2.0 2.!5 3.!5
DEPTH (ft)
0-5% 6-25% 25-50% 51-75% 75-IOO"'a
PERCENT COVER
Figure 10. Juvenile round whitefish beach se1n1ng mean catch (bars) by
habitat attribute va 1 ues of velocity, depth, and percent
cover. Suitability indices (lines) for depth and velocity
were fit by hand using professional judgement.
-28-
-
""""
-
""'!\
"""·
-
-
~
-
-
.....
-
-
4.0 DISCUSSION
4.1 Adult Resident Fish
Boat electrofishing and hook and line sampling have provided a limited
'set of data by habitat attr·ibute which were used to generate suitability
criteria for adult resident fish. These suitability criteria are
preliminary as sampling effort was limited. Since most sampling was
done by boat electrofishing a bias toward the capture of large fish was
probable. There may have also been some bias in the capture rates of
fish in clear versus turbid water because of differences in boat
electrofishing efficiency between these two habitat types but it did not
appear to be large. The boat electrofishing microhabitat suitability
data were collected near tributary and slough mouths during July to
October and therefore are applicable only during the open water season.
Additional information about rainbow trout and burbot microhabitat
distribution was also collected during radio· telemetry locations of
tagged fish and these data were used to supplement the other data
because they were free of sample efficiency bias.
Use of macrohabitats at tributaries and slough mouths could be due to
food input in the form of salmon eggs, fry or invertebrates drifting out
of the sloughs or tributaries. Species interactions could also play a
role in distribution as each species competes best within a niche. All
the species showed different responses to the habitat variables and this
may be due to these interactions rather than an actual preference.
lntercorrelations among habitat variables might also cause apparent
preferences as fish might actually be selecting for something else.
Turbidity was an important habitat attribute which had large effects on
adult resident fish distribution. Rainbow trout, Arctic grayling, and
round whitefish apparently avoided turbid water. Longnose suckers
avoided clear water. Turbidity also provided cover for all species and
therefore was desirable from this aspect.
Analysis of radio tagged rainbow trout distribution among the macrohabi-
tats of the Susitna River provided insights not obtainable by other
sampling methods. These data are not subject to the collection gear
bias inherent ·in other collection methods. Rainbow trout apparently
ascend tributary streams from mid-May through early June to spawn. Some
rainbow trout remain in the tributaries but others outmigrate to
mai nstem influenced macrohabitats. Tributary mouths are used heavily
for summer rearing especially during periods of salmon spawning.
Rainbow trout may also ascend tributaries and move into sloughs while
following spawning salmon. Rainbow trout were observed being chased
from spawning redds by male chum salmon while presumtbly feeding on
salmon eggs. One radio tagged rainbow trout in Slough A and another in
Lane Creek were observed milling around spawning pink and chum salmon.
The mainstem, per se, is probably used mainly as a migration path
between tributaries and sloughs at this time. By mid-September, howev-
er, all radio tagged trout which had been in tributaries had descended
to the mouths. The occurrence of this outmigration during a short time
period makes rainbow trout in the upper Susitna River extremely vulnera~
ble to sport fishing. Local anglers take advantage of the outmigration
-29 -
at the mouth of Indian River (RM 138.6) each fall. As the Susitna River
basin continues to develop, the rainbow trout population may suffer from
the increased fishing pressure. Most adult rainbow trout apparently
overwinter in the mainstem.
Rainbow trout distribution within microhabitat was correlated with
velocity and cover (Figures 5 and 6). Lewis (1969) found that rainbow
trout populations in pools were most highly correlated with higher
velocities, rather than the amount of cover. Shirvel1 and Dungey (1983)
found velocity to be the most important factor determining brown trout
position choice but that positions were chosen with optimum combinations
of depth and velocity. Observations of radio tagged fish, however,
revealed that rainbow trout distribution within microhabitat may be
dependent upon food source. In areas where rainbow trout were feeding
on salmon eggs, rainbow trout were closely associated with the spawning
salmon and therefore used shallow water riffles with cobble substrate
for cover. In other areas where there were no adult salmon, rainbow
trout were presumably feeding primarily on aquatic insects. In these
areas they were found in plunge· pools or deep pools using turbulent
water and depth, along with rubble/cobble substrate and debris as cover.
Turbulent water in plunge pools was observed to be excellent cover.
4.1.2 Juvenile Resident Fish
Juvenile resident fish use of macrohabitat present on the Susitna River
during the ice free months was found to vary greatly by species (Tables
5 and 6). Juvenile Dolly Varden, for example, were found only in
tributaries while round whitefish juveniles were found most abundantly
in mainstem side channels. The tributary sites are not influenced by
mainstem discharge so Dolly Varden rearing would be little affected by
changes in discharge. Round whitefish, on the other hand, might be
affected by changes in discharge. Juveniles of this species apparently
find turbid, mainstem conditions most suitable as they infrequently
occur in sloughs when the heads are not overtopped. Large numbers of
rearing juvenile Arctic grayling and round whitefish have been found
during previous Susitna studies to prefer mainstem mixing zones of
either sloughs or tributaries and secondarily mainstem waters (AOF&G
1983d). Longnose suckers were found in mainstem waters primarily but
data collected during 1983 indicate that juvenile longnose suckers also
find upland and side sloughs suitable for rearing.
Turbidity is the one factor which most distinguishes side slough habi-
tats from mainstem side channel habitats and turbid water increases the
suitability of mainstem side channels for such species as juvenile
Arctic grayling and round whitefish. Turbidity provides suitable cover
in environments which lack large amounts of object or overhead cover.
If lack of suitable cover limits rearing of juvenile fish, major de-
creases in the amount of turbid rearing areas may adversely affect
habitat used by juvenile Arctic grayling, round whitefish, and possibly
longnose suckers.
Round whitefish fry find ·turbid, mainstem side channels as the preferred
macrohabitat. Within these side channels, they use shallow, slow moving
microhabitats. Apparently the turbid water provides all the cover
-30 -
-
-
-
-
-
-
necessary. Little, if any, literature is available concerning juvenile
round whitefish rearing microhabitat needs.
Very little data are available concerning the microhabitat preferences
of other resident species which make use of mainstem influenced environ-
ments for rearing. Juvenile Arctic grayling under 200mm perhaps have
microhabitat preferences similar to that of chinook salmon fry or other
salmonids. Juvenile longnose suckers probably use microhabitat very
similar to that used by juvenile round whitefish as adult longnose
suckers also prefer shallow, slow moving, turbid habitats.
-31 -
5.0 CONTRIBUTORS
Field data were collected by Rich Sundet and Mark Wenger. Larry Dugan,
Paul Suchanek, Dave Sterritt, and Bob r4arsha11 collected the juvenile
round whitefish data. Dana Schmidt provided the study design.
Data processing was done by Allen Bingham, Gail Heinemann, Donna
Buchholz, Kathrin Zosel and Alice Freeman. Figures were drafted by
Carol Kerkvliet, Carol Riedner, and Sa11y Donovan. The typing was done
by Skeers Word Processing Services.
-32 -
-
-
-
-
-
-
-
-
-
. -
-
6.0 ACKNOWLEDGEMENTS
Funding for this study was provided by the State of Alaska, Alaska Power
Authority. We would like to express our gratitude to all the people and
organizations that provided information or assistance to the Resident
Fish Study during the past year.
We are especially grateful to Carl Burger (USFWS) for his technical
expertise and advice on radio telemetry investigations.
We are grateful to the employees at the Palmer ADF&G, Sport Fish office
and state veterinarian Burt Gore D.V.M. who showed us several techniques
to radio tag fish. We also thank ADF&G, FRED Division personnel at
Elmendorf hatchery who provided rainbow trout to us for a radio tag
implantation clinic •
-33 -
7.0 LITERATURE CITED
Alaska Department of Fish and Game (ADF&G). 1981. Phase 1 final draft
report. Subtask 7.10. Adult anadromous fisheries project. Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies.
Anchorage, Alaska.
--.,...· 1983a. Aquatic studies procedures manual. Phase II (1982-83).
Subtask 7.10. Alaska Department of Fish and Game Susitna Hydro
Aquatic Studies. Anchorage, Alaska.
1983b. Susitna Hydro aquatic studies phase II basic data
report. Volume 3 (2 parts). Resident and juvenile anadromous fish
studies on the Susitna River below Devil Canyon, 1982. Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies.
Anchorage, Alaska.
1983c. Susitna Hydro aquatic studies phase II basic data
report. Volume 4 (3 parts). Aquatic habitat and i nstream flow
studies, 1982. Alaska Department of Fish and Game Susitna Hydro
Aquatic Studies. Anchorage, Alaska.
___ • 1983d. Influence of habitat parameters on distribution and
relative abundance of juvenile salmon and resident species.
Appendix F in Susitna Hydro aquati.c studies phase II report.
Synopsis of the 1982 aquatic studies and analysis of fish and
habitat relationships (2 parts). Alaska Department of Fish and
Game Susitna Hydro Aquatic Studies. Anchorage, Alaska.
. 1983e. Use of major habitat types by juvenile salmon and
---resident species. Appendix G in Susitna Hydro aquatic studies
phase II report, synopsis of the 1982 aquatic studies and analysis
of fish and habitat relationships (2 parts). Alaska Department of
Fish and Game Susitna Hydro Aquatic Studies. Anchorage, Alaska.
• 1983f. Susitna Hydro aquatic studies phase II data report.
--..,.Winter aquatic studies (October 1982 -May 1983). Alaska Depart-
ment of Fish and Game Susitna Hydro Aquatic Studies. Anchorage,
Alaska.
___ • 1984. Susi tna Hydro aquatic studies May 1983 -June 1984
procedures manual. Alaska Department of Fish and Game Susitna
Hydro Aquatic Studies. Anchorage, Alaska.
Bovee, K.D. 1982. A guide to stream habitat analysis using the Instream
Flow Incremental Methodology. Instream Flow Information Paper 12.
U.S. Fish and Wildlife Service. FWS/OBS-82/26.
Everhart, W.H., A.W. Eipper, and W.O. Youngs.
fishery science. Cornell University Press.
1975. Principles of
Ithaca, New York.
Lewis, S. L. 1969. Physical factors influencing fish populations in
pools of a trout stream. Transactions of the American Fisheries
Society 98(1):14-19.
-34 -
-
-
-
-
-
.....
-
-
-
.....
I
,_
Montana Department of Fish, Wildlife and Parks (1'.1DFW&P). 1983.
Kootenai River fisheries investigations final report 1972-1982.
Mont. Dept. Fish, Wildlife & Parks. Kalispell, Mont. 224pp.
Raleigh, R.F., T. Hickman, R.C. Solomon, and P.C. Nelson. 1984.
Habitat suitability information: Rainbow trout. U.S. Fish &
Wildlife Service FWS/OBS-82/10.60.
Shirvell, C. S. and R. G. Dungey. 1983. Microhabitats chosen by brown
trout for feeding and spawning in rivers. Transactions of the
American Fisheries Society 112:355-367.
-35 -
APPENDIX A
Suitability Indices for Resident Fish Species
for Cover, Velocity, and Depth
-36 -
-
-
-
-
-
-
-
~.
Appendix Table A-1.
Cover Suitabilitx
PHABSIM
Cover. txpe Code
:-
No cover l.
Emergent
vegetation 2. -Aquatic
vegetation 3.
Debris/
dead fa 11 4.
Overhanging
riparian
vegetation 5.
Undercut
banks 6.
large gravel
(1-3") 7. -Rubble (3-5") 8.
Cobble or
boulder (>5") 9.
VELOCITY
-Rainbow
Velocity trout
( ft/sec) suitabi 1 itx -0.00 0.1B
0.05 1.00
1.05 1.00
1.55 0.50 .-2.55 0.33
3.55 0.20
4.50 o.oo
Suitability indices for resident fish species for cover, velocity, and depth.
Adult Juvenile
Rainbow trout Arctic grayling Round whitefish Longnose suckers Round whitefish
clear turbid clear turbid clear turbid clear turbid clear turbid
0 0.29 0 0.07 0 0.26 0 0.47 0 1.00
0 0.29 0 0.07 0.47 0.47 1.00 1.00 0 1.00
0 0.29 0 0.07 0.47 0.47 1.00 1.00 0 1.00
1.00 1.00 0.14 0.14 0.65 0.65 0.46 0.47 0 1.00
1.00 1.00 0.14 0.14 0.65 0.65 0.46 0.47 0 1.00
1.00 1.00 0.14 0.14 0.65 0.65 0.46 0.47 0 1.00
0 0.29 0 0.07 0.33 0.33 0 0.47 0 1.00
0.77 0.77 0.69 0.69 0.41 0.41 0 0.47 0 1.00
1.00 1.00 1.00 1.00 1.00 1.00 0 0.47 0 1.00
Adult Juvenile
Arctic Round longnose Round
Velocity grayling Velocity whitefish Velocity sucker Velocity whitefish
( ft/sec)
0
0.55
1.55
2.55
3.05
4.30
4.50
Depth
.l!.U_
0
0.5
0.6
10.0
suitabilitx (ft/sec)
0.04 0
0.25 0.55
0.46 1.55
O.B6 2.05
1.00 3.05
1.00 3.55
0.00 4.50
Adult
resident fish
suitabil itx
0
0
1.00
1.00
-37-
suitabi 1 itx (ft/sec) suitabil itJ: (ft/sec) suitabilitJ:
0.45 0 1.00 0 1.00
0.46 0.05 1.00 0.05 1.00
0.51 0.55 0.47 0.20 0.52
1.00 1.55 0,31 0,50 0.16
1.00 2.55 0.20 0.80 0.07
0.70 3.55 0.10 1.10 0.04 o.oo 4.3 0.00 1.40 0.00
Juvenile
Depth
1f.tl
0
0.15
0.50
0.75
1.25
1. 75
2.50
10.0
Round
whitefish
sui tabi 1 ity
0.00
1.00
1.00
0.42
0.35
0.33
0.31
0.31
APPENDIX B
Round Whitefish Spawning Microhabitat Data
-
-
-38 -
,,.,.
Since 1981, nine locations have been determined to be spawning sites for
round whitefish. In 1981 and 1982 one site was found each year at RM
100.8 and RM 102.6, respectively. In 1983 seven sites were found
including four mainstem sites: RM 102.0, RM 114.0, RM 142.0 and Rt~
147.0; and three tributary mouth sites: Lane Creek (RM 113.6), Indian
River (RM 138.6) and Portage Creek (RM 148.8) (Appendix Table 8-1).
While catch data and the incidence of sexually ripe fish suggest that
spawning of round whitefish might occur nearly anywhere in the mainstem,
selection of spawning sites may not be random. Anchor ice, water
fluctuations and ice cover can all limit egg survival. Due to these
reasons, round whitefish in the Susitna River may seek out areas which
have adequate ground water. Habitat data taken at one mainstem site (RM
147.0 in 1983) where eight sexually ripe males and females were captured
support this hypothesis. Specific conductance was relatively high, 160
umhos/cm, in this area, indicating an area of upwelling. Chum salmon,
another mainstem spawning species in the Susitna River, also seek areas
of upwelling for spawning (ADF&G 1983c).
-39 -
Appendix Table B-1. Physical and chemical habitat characteristics of spawning round whitefish in the Susitna River durfng October 1983,
Water gualitx
Water intraw sur-specific
Ve1ocit;t Substrate Turbid-gravel face conduct-
Area 1 River Mile Date .§.!.!! depth 0.2 0.8 x/0.6 Primary Secondar:t ....ilL temp temp e!:! .QQ !!!£.!.
Lane Creek (RM 113.6) Oct 7 1 3.2 1.8 1.6 cobble(511 -1011 ), rubble(311 -511 ) 12 .o 0.4
2 2.2 1.5 rubble(3 11 -5 11 ), gravel (1 11 -311 ) 12.0 0,4
Portage Creek (RM 148,8) Oct 5 1 4.2 1.4 1.2 rubble(3 11 -511 ), cobble(511 -1011 ) 4.2 1 .2 7.5 15.1 133
2 2.2 0.4 rubble (3 11 -5 11 ), silt 2.0 1,7 7.4 13~7 104
Mainstem (RM 147.0) Oct 5 1 2.1 0.7 s f1 t, cobb 1 e (511 -1 011 ) 14.0 0,6 0,0 7,5 15.1 159.0
2 1.9 0,7 silt. cobble (5 11 -1011 ) 14.0
I 3 2.3 0.7 s f1 t, cobb 1 e ( 511 -1 011 ) 14.0 0.6 o.o 7.5 15.0 160.0
~ 4 2,2 0,7 silt, cobble (511 -1011 ) 14.0
0 5 1.8 1 .2 cobble(511 -1011 ), boulder(over 1011 ) 14.0 . 0.6 o.o 7.5 15.iJ 161 .() I 6 1.7 1.2 cobble(5"-1011 ), boulder( over 1011 ) 14.0
l
PART 7
Modelling Of Juvenile Salmon
And Resident Fish Habitat
!"""
I
""" I
ABSTRACT
MODELLING OF JUVENILE SALMON AND
RESIDENT FISH HABITAT
Report Series No. 2, Part 7
by Stephen S. Hale, Paul M. Suchanek, and Dana C. Schmidt
Alaska Department of Fish and Game
Susitna Hydro Aquatic Studies
2207 Spenard Road
Anchorage, Alaska 99503
Output from the Instream Flow Group hydraulic models of rearing habitat
for juvenile salmon and resident species at seven sites in the Chulitna
River confluence to Devil Canyon reach of the Sus itna River 1 eads to
similar conclusions as those drawn from a habitat model developed by the
Susitna Hydro Aquatic Studies group for six additional sites. Overtop-
ping of side slough heads by mainstem discharge causes abrupt changes in
rearing habitat which are of positive benefit for some species/1 ife
stages and negative for others. Rearing habitat for chinook salmon at
the study sites is greatest when the head of the site is slightly
overtopped, thus providing turbid water for cover and moderate water
velocities. The portions of this reach which are directly influenced by
the mainstem provide only 1 imited rearing habitat for coho and sockeye
salmon during the open water season, but are likely to be of major
importance for all overwintering species. Resident species are associ-
ated with levels of turbidity, velocity, and food supply and in general
are not abundant in side sloughs when the head is closed unless a
tributary is present.
i
-
-
-
TABLE OF CONTENTS
ABSTRACT •••.••••••••••••••••••••••••••••••••••••••••••••••••••
LIST OF FIGURES ••••••.•.•
LIST OF TABLES •••.•••••
1.0
2.0
3.0
4.0
5.0
6.0
7.0
INTRODUCTION .••.••
METHODS ••••••••••••••••••••••••••••••••••••••••••••••••••••
2.1
2.2
2.3
2.4
2.5
Study Locations ...................................... .
Physical Habitat Modelling •.•.•••••••..•••••••••••••.•
2.2.1
2.2.2
Instream Flow Group (IFG) PHABSIM Models •••••.•
RJ Habitat Model {RJHAB} •.•••.•.•••••.•••.•••••
Sui tab i 1 i ty Criteria ...... -........................... .
Weighted Usable Area Projections ••••••••••••••.••.••••
Model Verification .••••.•..••••.•.••••.••••••••.•.
RESULTS •••••••••••••••.••••••••••••••••••••••••••••.•••••••
3.1
3.2
3.3
IFG Weighted Usable Area •••••
3 .1.1
3.1.2
3.1.3
' Chinook salmon •.••••.••••••.•.••••••••.••..••••
Chum and sockeye salmon •••••••••.....•••••.••.•
Resident species .••••••..••••.•••...••••••..•••
IFG Model Verification ••.•..•....•..........•.•.•....•
Habitat Indices ...................................... .
3.3.1
3.3.2
Juvenile salmon .•••••.•••••.•••••.••••••••••..•
Resident species ••.•••••••.•••••••••.•.••••...•
DISCUSSION ••••.•••..•••••.•••.•.•••••.••••••••••.•••••.•.••
4.1
4.2
4.3
Limitations of the Data ••.••.•.••••••••.••••.•••.••••.
Comparison of IFG Models with RJHAB •••.••.
4.2.1
4.2.2
Model characteristics •.•••.••.••.•.•••••.•.••.•
Model output .................................. .
Summary of Seasonal Habitat Projections for
Rearing Salmon and Resident Fish ••••.••••..••.••••••
CONTRIBUTORS .•..•.••.••••••••.•.•
ACKNOWLEDGEMENTS ••••..
LITERATURE CITED ..•.•..••. .................................
ii
iii
v
1
2
2
5
5
5
6
6
9
10
10
10·
17
17
22
22
31
31
35
35
36
36
39
41
45
46
47
-
....
-
LIST OF FIGURES
Figure Title
1 Location of IFG and RJHAB modelling sites •..•••••.•• 3
2 Percent of time that the heads of study sites
3
were overtopped by mainstem discharge ••••.•.•••••••• 4
Weighted usable area for chinook salmon at
the Slough 9 study site by level of mainstem
discharge at Gold Creek and by date, 1983 .•••••.•••• 11
4 Weighted usable area for chinook salmon at
the Side Channel 10 study site by level of
mainstem discharge at Gold Creek and by date,
1983 ................................................ 12
5 Weighted usable area for chinook salmon at
the Lower Side Channel 11 study site by level
of mainstem discharge at Gold Creek and by date,
1983 .............•.................................. 13
6 Weighted usable area for chinook salmon at
the Upper Side Channel 11 study site by level
7
8
of mainstem discharge at Gold Creek and by date,
1983 ................................................ 14
Weighted usable area for chinook salmon at
the Side Channel 21 study site by level of
mainstem discharge at Gold Creek and by date,
1983................................................ 15
Weighted usable area for chinook salmon at
the Slough 21 study site by level of mainstem
discharge at Gold Creek and by date, 1983 •••.••••••. 16
9 Weighted usable area for chum salmon at
the Slough 9 study site by level of mainstem
discharge at Gold Creek and by date, 1983 ••••.•..••. 18
10 Weighted usable area for chum salmon at
the Slough 21 study site by level of mainstem
11
discharge at Gold Creek and by date, 1983 .••••.••••• 19
Weighted usable area for sockeye salmon at
the Slough 9 study site by level of mainstem
discharge at Gold Creek and by date, 1983 •••.•.••..• 20
12 Weighted usable area for sockeye salmon at
the Slough 21 study site by level of mainstem
discharge at Gold Creek and by date, 1983 ••••.•••••• 21
iii
13 Weighted usable area for adult Arctic grayling
at the.Slough 9 and Side Channel 21 study
sites ..................... e•••••eo•••················· 23
14 Weighted usable area for adult Arctic grayling
15
and rainbow trout at the Slough 21 study site .•••.•• 24
Weighted usable area for adult round white-
fish and longnose suckers at the Slough 21
study site ........................................... 25
16 Weighted usable area for juvenile round
whitefish at the Slough 9 and Side Channel 10
study sites .....................................•... 26
17 Weighted usable area for juvenile round
whitefish at the Lower Side Channel 11 and
18
Upper Side Channel 11 study sites •••.••••••••••••••• 27
Weighted usable area for juvenile round
whitefish at the Side Channel 21 and Slough
21 study sites ................................. -...... 28
19 Habitat indices for juvenile salmon at IFG
20
21
22
23
modelling sites ..................................... 32
Habitat indices for juvenile round whitefish
and adult Arctic grayling at IFG modelling
sites ................................................ 33
Habitat indices for adult rainbow trout, round
whitefish, and longnose suckers at the Slough
21 modelling site ••••••••••••••..••••••••••••••.•..• 34
Time duration curves and mean monthly
discharges for June, July, August, and
September based on the 30 year record of
Susitna River discharge at Gold Creek •••••••.••••.•. 37
Comparison of RJHAB and IFG habitat indices
for juvenile chinook salmon .•••••••.•••••••••••••••• 40
iv
-
-
-
-
.....
-
~I
-
-
-
-I
-
-
LIST OF TABLES
Table
1
Title
Tota 1 catch and .catch per unit effort of
juvenile salmon at the IFG sites, open water
season, 1983........................................ 8
2 Correlations between composite weighting
factors and catch transformed by natural log
(x+l) for juvenile chinook salmon by selected
sites and by all sites pooled....................... 29
3 Chi-square contingency tests of chum and
sockeye salmon proportional presence by
composite weighting factor intervals................ 30
4 Comparison of model characteristics of IFG
mode 1 s and RJHAB •••.•••••••••••••••••••..•.•• lflo • • • • • • 38
v
.....
-
-
1.0 INTRODUCTION
The effects of flow regulation on downstream fisheries have long been
the subject of investigations whose goal was to predict the status of
fisheries after development of hydro power or other types of instream
flow regulation. The Instream Flow Incremental Methodology developed by
the U.S. Fish and Wildlife Service (Bovee 1982) has gained wide accep-
tance and is the method most often applied to these types of studies.
This method comprises the IFG {Instream Flow Group) PHABSIM (Physical
Habitat Simulation System) and has been used in Alaska by Estes et al.
(1980), Wilson et al. (1981), and ADF&G (1983a). The Susitna Hydro
Aquatic Studies group has used this method for ·two seasons to simulate
changes in available spawning habitat of chum and sockeye salmon as a
function of mainstem discharge.
Beginning in the open water season of 1983, we used these IFG hydraulic
models and another habitat model developed by ourselves (RJHAB) to
investigate the effects of mainstem discharge variations on rearing
habitat for juveniles of four species of salmon and juveniles and adults
of several resident fish species in the Susitna River.
This paper presents the results of the IFG model habitat simulations for
juvenile salmon and resident fishes, compares the IFG models with the
RJHAB model (presented in Part 4 of this report), and discusses in
general the usefulness and implications of these habitat models in
understanding and predicting the effects of discharge changes on rearing
habitat •
- 1 -
2.0 METHODS
2.1 Study Locations
Seven IFG model sites and six RJHAB sites located on the Susitna River
reach extending from the Chulitna River confluence to Devil Canyon were
modelled (Figure 1). Criteria used in IFG model site selection are
detailed in Estes and Vincent-Lang (1984). Sloughs 8A, 9, and 21 were
selected in 1982 to quantify the response of adult chum and sockeye
salmon spawning habitat in sloughs to variations in mainstem discharge.
These sloughs are representative of side sloughs in general and also
contain critical spawning habitat. In 1983, four IFG side channel study
sites were selected as representative sites for the study of responses
of mainstem salmon spawning and rearing habitat to variations in
mainstem discharge. The RJHAB sites were selected as representative or
important juvenile salmon rearing sites (see Part 4 of this report).
Figure 2 shows the sites ordered by the mainstem discharge required to
overtop the head of the sites. The two upland slough sites (Slough 5
and Slough 6A) are not included on this figure. It can be seen that,
generally, sites which have heads overtopped more than 60% of the time
have been named side channels; sites with less frequent overtopping have
been called sloughs. All sites to the left of the vertical line were
overtopped on more than half the days between June 1 and September 30.
The mainstem discharge required to overtop the head of each site is as
follows:
Site Model Overtopping Discharge~
Lower Side Channel 11 IFG-2 5,000
Side Channel lOA RJHAB 9,000
Side Channel 21 IFG-4 9,000
Upper Side Channel 11 IFG-4 13,000
Slough 9 IFG-4 16,000 b/
Slough 21 IFG-4 18,000 -
Side Channel 10 IFG-4 19,000
Slough 22 RJHAB 20,000
Whiskers Slough RJHAB 22,000
Slough 8 RJHAB 25,000
Slough 8A IFG-4 33,000
Slough 5 RJHAB upland slough
Slough 6A RJHAB upland slough
~ Cubic feet per second (cfs). Source: Estes and Vincent-Lang
{1984).
This is the discharge level at which a side channel entering the
Slough 21 study site begins to convey mainstem water. The head of
Slough 21 proper is not overtopped until a discharge level of
23,000 cfs.
- 2 -
-
-
-
-
....
.... IFG HABITAT SITES '
SITES RM
I Slough 21 141.8
2 Side Channel 21 140.6
3 Upper Side Channel II 136.0
4 Lower Side Channel II 134.6
5 Side Channel 10 133.8
6 Slough 9 128.3
7 Slough SA 125.3
t,f.
'" RJ HABITAT SITES -~
~'t SITES RM "\
~~ C:J' c
\) '/ A Slough 22 144.3 ·c., B Side Channel lOA 132.1 ..... c Slough 8 II 3.6
0 Slough 6A II 2.3.
E Slough 5 107.6
F Whiskers S l.ough I 0 1.2
-
Figure 1. Location of I FG and RJHAB mode 11 i ng sites.
- 3 -
\
o .....
........ ......... .. ...... ........
......
'o
0 5 10 IS 20 25 30 35
SUSITNA RIVER DISCHARGE AT GOLD CREEK ( x 10 3 cfs )
Figure 2. Percent of time that the heads of study sides were overtopped by mainstem discharge.
Sources: 30 year record -Bredthauer and Drage (1982); 1983 discharge -USGS provisional data.
I .I J J
-r
2.2 Physical Habitat Modelling
The models used have been described in other reports (see below) and
will only be summarized here. Basically, transects are established at a
site and then measurements of depth, mean water column velocity, and
cover are made across the transects. Also, the top width of the wetted
surface at each transect is measured so that wetted area may be cal-
culated. This is done on three or four different occasions over a range
of flows and the information is then input to the models. Output from
the models provides either simulated physical parameters and habitat
values (IFG) or interpolated habitat values (RJHAB) for any level of
discharge over a wide range of discharge.
2.2.1 Instream Flow Group (IFG) PHABSIM Models
Two hydraulic simulation models were used by the Aquatic Habitat section
and E. Woody Trihey and Associates during the 1983 open water season
(Estes and Vincent-Lang 1984). The IFG-4 model simulates depth and mean
water column velocity across horizontal transects at a site over a
discharge range from 40% of the lowest calibration flow to 250% of the
highest calibration flow (Bovee and Milhous 1978). The IFG-2 model is a
water surface profile model that provides the same information as the
IFG-4 model but which requires less field data. The IFG-4 model was
used for all of the sites except for Lower Side Channel 11, where the
IFG-2 model was used.
The models also allow the input of substrate data. However, cover data
rather than substrate information were input because it was determined
that cover was more important than substrate in influencing the dis-
tribution of juvenile salmon (see Part 3 of this report). Substrate was
frequently the primary cover type in the cover coding. Consistently
good cover data were not obtained at the IFG model sites because most of
the sites were primarily intended to be used for simulating habitat for
adult spawners. Consequently, cover for some of the transects had to be
estimated and may therefore lead to some error in the weighted usable
area (WUA) predictions. The cover values on these transects wi 11 be
obtained during the open water season of 1984 and the output modified
accordingly.
2.2.2 RJ Habitat Model (RJHAB)
The RJ Habitat Model, which modelled juvenile salmon habitat at six
sites, was described in Part 4 of this report. Transects were estab-
lished at these sites but, rather than using detailed depth and mean
column water velocity measurements across each transect, as do the IFG
models, these models use the average depth and average mean water column
velocity of 300 sq ft (6 ft wide by 50 ft long) cells which were estab-
lished along each transect. Usually, there were three cells per
transect, but sometimes only two when the channel became too narrow
(less than 18 ft in width). This model does not simulate hydraulic
characteristics of the site as do the IFG models; instead, it estimates
weighted usable area for shoreline and mid-channel portions of the site
for those discharge levels at which physical habitat attributes were
measured. Estimates of WUA for other discharges are then interpolated.
- 5 -
2.3 Suitability Criteria
The suitability criteria for juvenile salmon input into the models were
developed in Part 3 of this report. Suitability indices for cover,
velocity, and depth input into the PHABSIM models are presented in
Appendix Table A-1 of Part 3. The PHABSIM models linearly interpolate
between the point values for depth and velocity input. The cover
suitability indices were put into the model in place of substrate; these
indices reflect both amount and type of cover. Depth was not thought to
be as important as cover and velocity in affecting distribution; there-
fore, suitability for depth for all species was fixed at 1.00 (i.e., it
had no effect on the results) except when depth was less than 0.14 ft
and then suitability was fixed at 0.00.
Velocity suitability criteria input into the RJHAB models differed
slightly from those input to the IFG models. Suitability indices were
constant over an interval of 0.3 ft/sec for velocity. This grouping was
made because the limited number of velocity measurements was only an
index to hydraulic conditions present and finer resolution was deemed
unnecessary. Depth suitability for the RJHAB model was set to 1.0
because depths less than 0.2 ft did not occur.
Suitability criteria for resident fish input into the IFG models were
developed and presented in Part 6 of this report. Habitat of juvenile
round whitefish and adult rainbow trout, Arctic grayling, round
whitefish, and longnose suckers was modelled. The RJHAB models were not
run for any resident species. Because of limited data collection, the
suitability functions for resident fish are only preliminary.
2.4 Weighted Usable Area Projections
The PHABSIM system can be used to describe the mosaic of physical
features of a stream which includes substrate or cover and hydraulic
parameters such as depth and velocity. The HABTAT program of PHABSIM
incorporates the physical model and the suitability criteria to produce
weighted usable area, the habitat potential for a given life stage of a
species. Weighted usable area (WUA) is calculated (Bovee 1982) by:
WUA = Ci,s X A;
where: c. 1 ,s = the composite weighting factor (sometimes
called the joint preference factor) for cover,
velocity, and depth of the cell (i) for the
species and life stage (s)
A. = the surface area of the cell 1
The WUA for the study site at a given discharge was calculated by
totalling all the individual cell WUA's. The composite weighting factor
was calculated by multiplying the suitability indices for cover,
velocity, and depth of the cell together. WUA's at each study site were
calculated at 10 to 40 incremental flows over the recorrnnended extrapo-
lation range of the hydraulic model..
- 6 -
-
-
-:
-
-
-
-
-l
-
-
-
At RJHAB sitesy WUA•s were calculated for shoreline and mid-channel
portions of the site each time the site was measured. Data were pooled
to yield a discharge-specific site WUA instead of calculating individual
cell WUA 1 s as in the IFG PHABSIM models. WUA 1 s calculated for the RJHAB
sites are generated from habitat measurements which provide an index to
conditions at the site. The IFG WUA is standardized to a 1000 ft reach
while the RJHAB WUA is dependent on the size of the site.
The output from the IFG models consists of weighted usable area and
total surface area predictions for incremental levels of site flow which
was in turn related to mainstem discharge by rating curves provided by
Estes and Vincent-Lang (1984). RJHAB provides the same information at
measured discharges and then plots WUA as a function of discharge. All
of the output from RJHAB was presented in Part 4 of this report.
We entered the output of the IFG models into a microcomputer worksheet
program to perform additional manipulations of the data. Firsty plots
were constructed of WUA as a function of mainstem discharge. Then we
matched WUA predictions with each of the mean daily discharge levels
observed from June 1 to September 30y 1983 to obtain a time series of
WUA at each of the sites during the open water season. This time series
was compared with the catch data at these sites and the outmigration
timing data from the downstream migrant traps to better understand the
relation between WUA and fish behavior.
All of the possible site/species combinations were run through the IFG
modelsy but only certain ones are presented in this paper because of
space limitations; all raw model output is available on request. With a
few exceptionsy the basic criterion used to select species/site combina-
tions for presentation was that mean catch per cell for the species for
the entire season at the site had to be greater than the mean catch per
cell at all sites (Table 1). Hencey we are not including weighted
usable area predictions for a species at those sites where very few
individuals of the species were captured. There are some exceptions to
this practice for resident species because the sampling methods used at
the modelling sites were not intended for capture of adult resident
fish. The species/life stages for which weighted usable area predic-
tions are presented include juveniles of four salmon species (chinook,
cohoy chumy and sockeye)y juvenile and adult round whitefish, and adult
rainbow trouty Arctic graylingy and longnose suckers.
To make comparisons among sites which would be independent of the size
of the site, we divided the site weighted usable areas at each level of
discharge by the total surface area of the site when the mainstem
discharge was 23,000 cfs (the area was interpolated from the PHABSH1
output of total area as a function of flow). The 23YOOO cfs figure was
chosen because it is a typical mid-summer discharge (Bredthauer and
Drage 1982; Klinger and Trihey 1984) and because it may be integrated
with macrohabitat abundance information which was digitized from aerial
photographs by E. Woody Trihey and Associates. The resulting habitat
index is comparable to the habitat index calculated for the RJHAB sites
in Part 4 of this report.
- 7 -
Table 1. Total catch and catch per unit effort of juvenile salmon at
the IFG sites, open water season, 1983.
Catch (catch/cell)
No. of Chinook Coho Chum Sockeye
IFG Site Cells 0+ 0+ 0+
Slough 21 86 91(1.1)* 1{0.0) 417(4.8)* 23(0.3)*
Side Channel 21 23 38{1.6)* 0(0.0) 0(0.0) 0(0.0)
Upper Side
Channel 11 21 101(4.8)* 0(0.0) 0(0.0) 0{0.0)
Lower Side
Channel 11 21 39(1.9)* 0(0.0) 0(0.0) O(O.O)
Side Channel 10 62 279{4.5)* 0{0.0) 2(0.0) 0(0.0)
·Slough 9 123 227(1.8)* 0{0.0) 74(0.6)* 30(0.2)*
Slough 8A 66 6(0.1) 26(0.4) 129(2.0) 24(0.4)
Sum of IFG sites 402 781 27 205 77
Mean of IFG sites 112(1.9) 4{0.1) 29(0.5) 11(0.2)
Mean of all sites sampled
Backpack electrofishing (3.4) (2.3) ( 1.3) (0.9)
Beach seining (3.4) (0.3) (0.0) (0.5)
* = Site/species combination selected for presentation.
- 8 -
-
-
"""'
"'"'l
-
~
iOIVli
-
-
-)
-
-
2.5 Model Verification
Data on fisheries abundance and distribution were collected at the
sites; however, program constraints prevented ·intensive sampling ef-
forts. Composite weighting factors were calculated for each 6 ft X 50
ft cell sampled for fish and this index was then correlated with fish
catch in the cell. If cells with large composite weighting factors are
associated with higher densities of fish, then it can be assumed that
WUA does reflect habitat potential. Correlations or associations
between catch and composite weighting factors at the RJHAB sites have
been presented in Part 4 of this report. Data were available at the IFG
sites for verification of composite weighting factors for juvenile
salmon and round whitefish, but not for adult resident species.
The specific hypothesis tested was whether the correlation between a
composite weighting factor and catch of chinook and coho salmon/cell
[transformed by natural log (x+l)] was greater than zero (in other
words, whether there was a significant positive relationship). For
sockeye and chum sa 1 man, the null hypothesis was that there was no
association between the composite weighting factor and fish presence.
Sampling occasions when less than three fish were captured in all cells
within a site sampled during a day were deleted from the analysis. This
was done because seasonal variations in outmigration from natal areas
can 1 ead to 1 ow fish density, even in areas that provide good rearing
habitat, and inclusion of data from these times could lead to spurious
correlations.
-9 -
3.0 RESULTS
3.1 IFG Model Weighted Usable Area
Juvenile salmon catches and catch per unit effort (CPUE) varied greatly
at the seven IFG modelling sites (Table 1). Since discharge levels of
more than 33,000 cfs (the 1 eve 1 required to overtop the head of the
Slough SA study site) occurred infrequently during the 19S3 open water
season, this site was not modelled for any species. Juvenile salmon at
this slough were primarily caught below the modelling site. The Slough
SA IFG modelling site harbored few juvenile fish because access was
restricted from below by several beaver dams and access was restricted
from above because the head was only infrequently overtopped.
Juvenile coho catches and CPUE were very low at all the modelling sites
and, therefore, no results for coho WUA's are presented. In general,
WUA's calculated for coho salmon at the sites were less than 2% of the
total surface area of the site. The primary reason for low coho density
was the preference of cohos for non-turbid water and cover types infre-
quently found in the sites modelled (see Parts 2 and 3 of this report).
All of the IFG modelling sites, with the exception of Slough SA, har-
bored significant numbers of chinook salmon and results from these six
sites are presented. Sockeye and chum WUA's are presented for sloughs
21 and 9 as these were the only two sites where these species were
relatively numerous. Unfortunately, the four mai nstem side channel
sites were not sampled for fish density until July; most chum and large
/numbers of sockeye had moved down river by this time (see Part 1 of this
report).
In the time series plots that follow, if a mean daily discharge exceeded
the extrapolated range of the model, no WUA value was plotted. No
weighted usable areas of zero occurred. If the discharge was less than
the extrapolated range, then the WUA was set equal to the WUA value for
the 1 owest discharge in the extrapolated range. WUA at four of the
sites was extrapolated to some point below the overtopping flow. WUA
did not change very much at flows less than the overtopping flow because
the surface areas of the sites remained relatively constant, being
affected mainly by site morphology and local hydrology. The lower end
of the extrapolated range at Slough 9, Slough 21, and Lower Side Channel
11 was above the overtopping flow.
3.1.1 Chinook salmon
Weighted usable areas for six IFG modelling sites as a function of
mainstem discharge and as projected over the June 1 to September 30 time
period are presented in Figures 3 through 8. There were two different
sets of suitability criteria for chinook salmon; one for a low turbidity
level and one for a high turbidity level (Part 3 of this report). We
used the low turbidity criteria when the head of a site was closed and
the high turbidity criteria when the head was overtopped by mainstem
flow. The point of overtopping was taken as the point when mainstem
water just began to flow through the head, raising the turbidity level
of the site. Chinook juveniles preferred the high turbidity condition
when other cover types were not abundant. Therefore, the weighted
-10 -
~
'
.....
-
M'l!
,......
CIJ -._
34 ~ 32
30
28
26
24
22
20
181 16
14
12
SLOUGH 9
CHINOOK SALMON
OVERTOPPED
10 ,_-----,------.-0----,------.-----~------~----~----~
5 15 25 35
(Thou so nds)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs)
45
-
34
32
30
36
34
2a :c ,......
CIJ 26 ~ -... z
'-" 24 IJJ
~ < 22 rT'I .... :c
a::..-. <w 20 Q
"0 l.oJc: IJJ
....lc 1a n cc..., I
<::J 16 >
lllo ::0
::)..c: (;)
14 rT'I ct.. .... 12 ,.-..
1-::c
S2 ....
3:
X
10 -0
a 0
0
.... 9 n -Ill 2
4 '-"
2
0 a
JUNE JULY AUG SEPT
DATE
Figure 3. Weighted usable area for chinook salmon at the Slough 9 study
site by level of mainstem discharge at Gold Creek and by
date, 1983. In the lower graph, daily WUA's are plotted as
bars. No WUA value is plotted if the mean daily discharge
exceeded the extrapolated range of the model.
-11 -
:zo
19
18
17
16 ......
C'i. 15 ....
'-' t4
< 13 I,J
0::.-., <Ill 12
I,J"C 1 1 ....Jc:
CJC 10 <Ill
Vl:::l 9 ::Ji
oC 8
I,J 7 ~
I 6 Q
I,J 5 3: 4
3
2
1
0
1 8
1 7
1 6
1 5 -14 N -.... 13 -< 12 w
0::,-.. <Ill 1 1
w"C 1 0 ....Jc: ale 9 <(II
(ll:::l
8 ::Ji
oC 7
I,J ,... 6 I
(.!) 5 L;j
~ 4
3
2
1
0
SIDE CHANNEL 1 0
CHINOOK SALMON
OVERTOPPED
5 15 25 35 45
MAINSTEM
(Thousonds)
DISCHARGE AT GOLD CREEK (ch)
------------------------------------36
JULY AUG SEPT
DATE
...
0
0
0
n ....
Ill ,_.
Figure 4. Weighted usable area for chinook salmon at the Side Channel
10 study site by level of mainstem discharge at Gold Creek
and by date, 1983. In the lower graph, daily WUA 1
S are
plotted as bars. No WUA value is plotted if the mean daily
discharge exceeded the extrapolated range of the model.
-12 -
-I
-
-
-
-
-
-
-
LOWER SIDE CHANNEL 1 1
CHINOOK SALMON
27 ,-.---------------------------------------------------~
26
25
22
21
20
19
18
17
1 6
Lo.J 15 3::
,....
C\.1 -... ........
14
13
12
11
5
OVERTOPPED
45 15 25 35
(Thousands)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs)
28,..-----------------------------------------------------~36
26 34
24 32 3::
22
20
30
28
26
24
22
~
20 ,....
6
4
2
0 ~mnrrrrmm~rnmmmnrrmmmmrr*mm~rnmmmnnmmmnnrrmmmnrrrn~
JUNE JULY AUG
DATE
SEPT
18
16
14
12
10
a
Figure 5. Weighted usable area for chinook salmon at the Lower Side
Channel 11 study site by level of mainstem discharge at Gold
Creek and by date, 1983. In the lower graph, daily wuA•s are
plotted as bars. No WUA value is plotted if the mean daily
discharge exceeded the extrapolated range of the model.
-13 -
X -0
0
0
n ...
til .._,
UPPER SIDE CHA~H~EL 11
CHINOOK SALMON
32
c---' \ f\ I (~~
I
I
I A "''"<lPeED
30
28
....... 2:6 -N -........ 2:4 <(
w a:: 2:2-<(-;;;"
w"t'l 20 -__.c: m" <((/) Vlg i 8 -
::>.c
at:-i 6 -w
~
I 14-Q
w
~ 12-
i 0 -
8 -
6
5 15 25 35 45
{Thousands)
MAJNSTrM DISCHARGE (cfs)
.32 .36
.30
28
3:: 26 ~ ,....... z N 24 [J) -.... -4 -22 r"'1
< 3::
l.o.l 20 S1 0::,-... <Ill [J)
l.o.I1J 18 0
I ..Jc: > a:Jc 16 <Ill ::0
lfl:::J c;) ::J_g 14 r"'1
ot:.. ,--.
12 X l.o.l
1--J: 10 0
Q 0
l.o.l 8 0
~ n .... 6 "' -4
2
0
JUNE JULY AUG SEPT
DATE
Figure 6. Weighted usable area for chinook salmon at the Upper Side
Channel 11 study site by level of mainstem discharge at Gold
Creek and by date, 1983. In the lower graph, daily wuA•s are
plotted as bars. No WUA value is plotted if the mean daily
discharge exceeded the extrapolated range of the model.
-14 -
-
.....
-
-
-
SIDE CHANNEL 21
CHINOOK SALMON
45
40
..-..
"i. 35 -
"' w
0::.,......_ 30
<("'
11..11:1 ...Jc wo 25 "'"' V):l
=-i:
at:. 20 11.1 ,_
I
Q
11.1 15 3:
10
OVERTOPPED
5
5 15 25 35
(Thousands)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs}
45
.36
.34
32 r
!:::
30 z
Ul
-1
28 r"' r
26 Q
Ul
24 0
I >
22 ::0
Q
r"'
20 r-
X
18
0
16 0
0
Cl 14 -C/1 ......
12
10
8
JUNE JULY AUG SE:PT
DATE
Figure 7. Weighted usable area for chinook salmon at the Side Channel
21 study site by level of mainstem discharge at Gold Creek
and by date, 1983. In the lower graph, daily WUA's are
plotted as bars. No WUA value is plotted if the mean daily
discharge exceeded the extrapolated range of the model.
-15 -
SLOUGH 21
CHI NOOK SALMON
30
29
28
27
26 ,.... 25 "i. 24 ... .._,
<( 23
w 22 a::.-..
<C~n 21
t..I'O 20 ..Jc
r:IJC 19 <(In
C/15 18 ::J..c: 17 c.!:. w 16 1-
I 15
~ 14 w
3: 13
12
11
10
9 OVERTOPPED
8
5 1 5 25 35 45 ~Thousands)
MAIN STEM DISC ARGE: AT GOLD CREEK (cfs)
30 36
28 34
26
32 ~
~ 24
30 z .......
Ul C\1
-4 -22 ...
28 1""1 .._,
~ <( 20
!:! w 26 0::..-.. 18 Ul <(Ill n t..l'tl 24 I ..Jc 1 6 ~ r:IJC ::0 <(Ill 22 (;) VI ::I 14 1""1 ::J_g 20 ,..... c.!:. 12 X w
18 -1-10 0 I
0 g 16 0 w 8
~ (l
14 -6 Ill
'-"
4 12
2 10
0 8
JUNE: JULY AUG SEPT
DATE:
Figure 8. Weighted usable area for chinook salmon at the Slough 21
study site by level of mainstem discharge at Gold Creek and
by date, 1983. In the lower graph, daily WUA•s are plotted
as bars. No WUA value is plotted if the mean daily discharge
exceeded the extrapolated range of the model.
-16 -
-
-
.....
-
-
....
....
-
usable area for chinooks drops sharply when discharge levels become low
enough so that the head of the site is no longer overtopped by turbid
mainstem water. At mainstem discharges less than those required to
overtop the head of the site, there is no strong relationship between
slough flow and mainstem discharge unless groundwater flow is signifi-
cantly related to discharge. Calibration ranges of the model at many of
the sites limited the calculated responses of WUA to a small range of
mainstem discharges. The three peak discharges which occurred in early
June and in early and late August exceeded the calibration range of all
the sites except for Slough 21.
Typically, peaks in weighted usable area were found at mainstem dis-
charges slightly (within a few thousand cfs) greater than the
overtopping discharges. The Slough 21 study site appears (Figure 8) to
be an exception to this trend but in fact is not. A small side channel
which entered the Slough 21 study site conveyed mainstem water at
discharge levels greater than 18,000 cfs·, but the amount of mainstem
water entering the site did not become substantial until the head of
Slough 21 proper became overtopped at 23~000 cfs. ·
The time when the WUA peaks occurred and, hence~ the period when the
site was theoretically able to support the maximum number of fish, can
be seen from the time series plots. With a few exceptions, sites at
which the overtopping flow occurred at a middle level of discharge
provided more habitat during the open water season of 1983 than sites
which had either a relatively low overtopping flow or a relatively high
overtopping flow. With the exception of the two side channels which had
low overtopping discharges (Lower Side Channel 11 and Side Channel 21),
weighted usable area was low to all sites in September because low
mainstem discharge (down to 9,000 cfs) led to reduced velocity, depth,
and surface area at these study sites.
3.1.2 Chum and sockeye salmon
Plots of weighted usable area for chum and sockeye salmon as a function
of mainstem discharge showed very similar trends (Figures 9 through 12).
Chum and sockeye WUA plots were almost identical at both Slough 9 and
Slough 21. At both sites, WUA • s for chum and sockeye peaked rapidly
with small increases in discharge, held constant over a range of approx-
imately 5,000 cfs in mainstem discharge, and then decreased rapidly with
further increases in mainstem discharge. At a given site, sockeye WUA's
peaked slightly before chum WUA's because slightly lower velocities were
more suitable to the sockeye salmon juveniles. Chum and sockeye salmon
WUA at these two sites remained relatively high in September as compared
to chinook WUA~ because chum and sockeye salmon have a preference for
lower velocities. However, the chum WUA in September is never used
because this species has basically outmigrated from this reach by the
end of July.
3.1.3 Resident Fish Weighted Usable Area
Only limited sampling for resident fish was conducted at the IFG model-
ling sites and, therefore~ no site-specific data on adult resident use
of the sites are available. Many of the sites are inaccessible to
-17 -
SLOUGH 9
CHUM SALMON
27
26
25
24 ,-..
"i. 23 ....
'-'
< 22 Ll.l c::,..., <Ill 21
LI.I"C -lc: 20 ceo <Ill 19 Ul:::l ::::l_g
at:-18
Ll.l
1-17 :::c
Q 16 Ll.l
~ 15
14
13
OVERTOPPED
12
5 15 25 35 45
(Thousands)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs)
28 36
26 34
24 32
22 ,....... 30 ~ .... 20 ....... 28
< Ll.l 18 26 a::,.... <Ill
LI.I"C 16 24 ...~c:
CCC 14 <Ill 22 Ul:::l ::l_g 12
at:-20
Ll.l 10 18 1-
I
Q 8 16
Ll.l
~ 6 14
4 12
2 10
0 8
JUNE JULY AUG SEPT
DATE
Figure 9. Weighted usable area for chum salmon at the Slough 9 study
site by level of mainstem discharge at Gold Creek and by
date, 1983. In the lower graph, daily WUA 1 s are plotted as
bars. No WUA value is plotted if the mean daily discharge
exceeded the extrapolated range of the model.
-18 -
~ -~ z
(/)
~
rr'l
~
Q
(/) n
I
>
::0
(i) -rr'l
,.-..
X ...
0
0
0
n -(II
'-"
-
.....
-
SLOUGH 21
CHUM SALMON
25
24
23
22
21 ,....,
~ 20 ....
'-" 19
<( 18 1.&.1 c::,.....
<Ill 17
1.&.1-,:J 1 6 ...Jc
IIJO 15 <(Ill
t/]::::1 14 ~ ::J_g
aC 13 l
1.&.1 1 2 -j 1-
::t
1 1 j 5:2
1.&.1 '! ~ ~
7 I
sJ OVERTOPPED
5
5 15 25 35 45
(Thousands)
MAINSTEM DISCHARGE: AT GOLD CREEK (cfs)
-~--·-----·-··-
.24-36 -----22 34
20 32 3::
!:: ,....,
N 1 8 -....
'-"
<( 1 6 1.&.1 a::,....
<Ill 1 4-1.&.1-,:J ...Jc
IIJO 1 2 <(Ill
t/]::::1 :::J_g 10 aC
30 z
Ul
~
28 ,.,
3::
26 2
Ul
C"l 24 I
)lo
:tl 22 " ,.,
20 ,.....
X
1.&.1
1-8 ::t
5:2
1.&.1 6 ~
18 -0
0
16 0
n
14 ....
" 4-'-"
12
2 10
0 8
JUNE JULY AUG SEPT
DATE
Figure 10. Weighted usable area for chum salmon at the Slough 21 study
site by level of mainstem discharge at Gold Creek and by
date, 1983. In the lower graph, daily WUA•s are plotted as
bars. No WUA value is plotted if the mean daily discharge
exceeded the extrapolated range of the model.
-19 -
SLOUGH 9
SOCKEYE SALMON
6.5
6
,-... 5.5
(\J--.......
<( 5
UJ
0:::,-...
<(Ill
UJ"C 4.5 ...Jc: IDe <Ill
tll::l 4 ::J_g
oC
l.&.i
1-3.5 I
Q
UJ
3: 3
2.5
OVERTOPPED
2
5 15 25 35 45
(Thousands)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs)
7 36
34
6 32
,-...
(\J 30 -... 5 ._, 28
<(
UJ
C:::,-... <Ill
26
UJ"C 4
...Jc: IDe <Ill
tll::l ::J_g 3
oC
24
22
20
UJ
1-18
I
Q 2
UJ 16
3: 14
12
10
0 a
JUNE JULY AUG SEPT
DATE
Figure 11. Weighted usable area for sockeye salmon at the Slough 9
study site by level of mainstem discharge at Gold Creek and
by date, 1983. In the lower graph, daily WUA 1 s are plotted
as bars. No WUA value is plotted if the mean daily
discharge exceeded the extrapolated range of the model.
-20 -
-
.....
3::
~ z
til
-1
1"1
3::
a
til n
I >
::0
(;)
1"1
,-..
X
0
0
0
0 -" .......
.-
....
-
-
-
SLOUGH 21
SOCKEYE SALMON
5.5
5
_.....
~ 4.5 -.......
< w 4 a:,...... <Ill
w"C ....Jc:
lllc 3.5 '< (/J (/]:I ::::~1
at:. 3 U..l
~
I
Q
U..l 2.5 ~
2
OVERTOPPED
1.5
5 15 25 35 45
MAINSTEM
(Thousands)
DISCHARGE AT GOLD CREEK (cfs)
~~--
6 36
34
5 32
_..... 30 C\1 --28 .......
< 4-w 26 a:,...... <Ill
U..I"C 24 ....Jc:
lllc 3 <Ill 22
(/]:I ::::~1 20 at:. w 18 ~ 2 I
Q 16 w
~ 14
1 12
10
0 8
JUNE JULY AUG SEPT
DATE
Figure 12. Weighted usable area for sockeye salmon at the Slough 21
study site by level of mainstem discharge at Gold Creek and
by date, 1983. In the lower graph, daily WUA's are plotted
as bars. No WUA value is plotted if the mean daily
discharge exceeded the extrapolated range of the model.
-21 -
r
~·
z
VI
-t
r'1 r
Q
VI
(")
I
l>
::0
C)
r'1
r"
X -0
0
0
n ...
(/J .......
electrofishing boats except.during high mainstem discharges. Slough 21
was selected as a representative site to present responses of adult
resident fish habitat to changes in mainstem discharge. The relation-
ships between WUA and mainstem discharge for adult rainbow trout, Arctic
grayling, round whitefish, and longnose suckers are shown in Figures 14
and 15. Since Arctic grayling are frequently found in side channels
during the ice-free months, responses of WUA to mainstem discharge for
Arctic grayling at Slough 9 and Side Channel 21 are also presented
(Figure 13). Within the extrapolated flow ranges of the site or sites,
WUA•s for adult rainbow trout, Arctic grayling, and round whitefish
increased with flow. WUA for longnose suckers, which prefer low veloc-
ities ahd turbid water, peaked with the overtopping of the site by
mainstem discharge and then rapidly decreased with further increases in
discharge.
At least 16 juvenile round whitefish were captured at every site with
the exception of Slough SA where none were captured. Results from WUA
calculations for juvenile round whitefish are presented for six sites in
Figures 16 to 18.
3.2 Model Verification
Slough 9 and Side Channel 10 were the only two IFG sites where both a
relatively large amount of sampling and catch of juvenile chinook
occurred. Correlations between chinook catch and composite weighting
factor at Slough 9 and for all seven sites pooled for both clear and
turbid conditions were significantly greater than 0.0 (Table 2). At
Side Channel 10, however, there was no significant correlation between
chinook catch in turbid water and the composite weighting factor.
Data from Sloughs SA, 9 and 21 were pooled for chi-square contingency
tests of chum and sockeye proportional presence by composite weighting
factor interval (Table 3). Chum salmon presence was associated with
larger composite weighting factors; however, sockeye salmon presence was
not.
Correlations between round whitefish catch in turbid ( '/' 30 NTU) water
and composite weighting factors were all significantly greater than 0.0
at the 0.01 level. The correlations were 0.35 (n =54) at Side Channel
10, 0.46 (n = 63) at Slough 9, and 0.52 (n = 188) for all seven IFG
sites pooled.
3.3 Habitat Indices
In order to compare modelling sites with one another and to compare IFG
model results with RJHAB model results independently of site surface
area, habitat indices were calculated by dividing WUA by the total
surface area of th~ site at a mainstem discharge of 23,000 cfs. This
discharge level was chosen because it represents typical mid-summer
discharge conditions in this reach (Klinger and Trihey 1984).
-22 -
-
-
....
-
-
-'
~
I
I
SLOUGH 9
ADULT ARCTIC GRAYLING
21
20
19
18
17 --16 C\J -... 15 .._,
< 14
~-~ a::,... 13 <UJ 12 -c ~c: -lo 11 £IlUI <::::~ 10 Vlo
9 ::l.c:
ot:. 8 ~
1-7 I
Q 6
"'-1 5 ~
4
3
2
1 OVERTOPPED
0
5 15 25 35 45
MAINSTE:M
~Thousands)
DISC ARGE: AT GOLD CRE:E:K (cfs)
SIDE CHANNEL 21
ADULT ARCTIC GRAYLING
150
140 1 /
130 ~ -120 --C\J -11 0 ... .._,
< "'-1 100
a::--. <UJ 90 -c U.lc: 80 -lc
£Dill <::::~
70 l rJJo
::l.c:
ot:. 60
~
1-50 ,.,.. I
' Q
"'-1 40
~
30 ~ I
20
10
0
5 15 25 35 45
(Thousands)
(cfs) MAINSTEM DISCHARGE AT GOLD CREEK
Figure 13. Weighted usable area for adult Arctic grayling at the
Slough 9 and Side Channel 21 study sites.
--23 -
SLOUGH 21
ADULT ARCTIC GRAYl-ING
18
17 I 16
15
,...... 14
"' --13 '-'
I <( 12
1.1..1 a::,..... 1 1 <(Ill
w"tl 10 ..~c:
CDC 9 <(Ill
(/]:::1
:::Jl 8 I oG 7
LI.J
1-6 :r: I CJ 5
LI.J
3: 4
3
21 ,/ 1 OVERTOPPED
0
5 15 25 35 45
MAINSTEM
(Thousands)
DISCHARGE AT GOLD CREEK (cfs)
SLOUGH 21
ADULT RAINBOW TROUT
33
32 ~
31 __J I
30 -i
29 j ~
t'i. 28 -._, 27
<( 26 w a::,.... 25 J I <(Ill
"0 We: 24 I -lc
CD Ill 23 <C::J
!llo 22 :::J..c:
oG 21
w 20 1-:r: 19 ~
LI.J 18 3: 17
1 6
15
1 4 -j OVERTOPPED A
13 I I
5 15 25 35 45
(Thousands)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs)
Figure 14. Weighted usable area for adu1t Arctic grayling and rainbow
trout at the Slough 21 study site.
-24 -
-
.....
"""'
~
~
-
~
-
"""'
-
-
SLOUGH 21
ADULT ROUND WHITEFISH
23
22
21
20 --19
"' -18 ... .......
<( 17
Uol 1 6 a::_._
<~n 15 Uol"U ..Jc 14 IDe <111 13 ({1::::1 ::J_g 1 2 ot:.
Uol 1 1
1-
I 10
52 9 Uol
~ 8
7
: ~ OVERTOPPED
4
5 15 25 35 45
MAINSTEM
(Thousands)
DISCHARGE AT GOLD CREEK (cfs)
SLOUGH 21
ADULT LONGNOSE SUCKER
19
18
--"' 17 -...
'-'
<(
Uol a::_._ 16
<~n
Uol"U ..Jc:
IDe 15 <In
({1::::1 ::J_g
ot:.
14 ~ \
l.iJ
1-
I
52
Uol 13 ~
12
OVERTOPPED
1 1
5 15 .25 35 45
MAIN STEM
(Thousands)
DISCHARGE AT GOLD CREEK (cfs)
Figure 15. Weighted usable area for adult round whitefish and longnose
suckers at the Slough 21 study site.
-25 -
30
28
26 j
""" 24 N --'-' 22 < L.J ~: ~ a::""" <Ill
L.J'"O ....~c
a:Jo <Ill
(/]:l 161 ::l~
ot.. 14 L.J
~ , 2l I
Q
L.J 1: ~ ~
6
4
5
20
19 -
18
17
16 ........
N 15 ~ .._,. 14
< 13 -1.&1
11:.,--.. <rn 12 -
1.&1"1:1 1 1 -_.~C mo 10 -<"' V):J 9 -,o -.c
at:-8 -
1.&1 7 -1-
I 6 -0
1.&1 5 -~ 4 -
3 -
2 -
1 -
0
5
SLOUGH 9
JUVENILE ROUND WHITEFISH
I
OVERTOPPED
15 25 35
(Thousands)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs)
SIDE CHANNEL 10
JUVENILE ROUND WHITEFISH
(
I
' /
I
OVERTOPPED
15 25 35
(Thou sends)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs)
45
45
Figure 16. Weighted usable area for juvenile round whitefish at the
Slough 9 and Side Channel 10 study sites.
-26 -
-
.....
....
-
-
-
-·
-
~·
LOWER SIDE CHAN"'IEL 11
JUVENILE ROUND WHITEFISH
6 ,----.-----------------------------------------------, ~~~ 7.5 -
;;--7---'-'
~ 6.5-
a::_..., <Ul
1.1.1 "tl 6-..Jc
1Il 0 I <U] I (/]::J
:J1 s.sl
o.!:.
1.1.1
~ 51
~ 4.5 ~
l
I
4-
I
I v
OVERTOPPED
3.5 ~----~------,-----~------~-----.------~----~----~
5 15 25 35 45
(Thousands)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs)
UPPER SIDE CHANNEL 11
JUVENILE ROUND WHITEFISH
24.---------~f\=-------------------------------------~ I \ 22-
I \
\../~ ___
20-
18-
16-
14-
12
10
8 -
6 -
4-
2 -.4 OVERTOPPED
0 ,_--------~~~------------.------.------r-----~----~
5
Figure 17.
, 5 25 35
(Thousands)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs)
45
Weighted usable area for juvenile round whitefish at the
Lower Side Channel 11 and Upper Side Channel 11 study sites.
-27 -
........
"' --.._,
<
U.i
0::.-.., <111
U.i, ...~c
Cllc <111
(/]:J ::J_g
ot:.
U.i
~
I
Cl
U.i
~
SIDE CHANNEL 21
JUVENILE ROUND WHITEFISH
24 .-------------------------------------------------------~
2:2
20
18
1 6 -
14
12
10
8
6
4
2
0
<NERTOeem ,-----
5 1 5 25 35 45
(Thousands)
MAINSTEM DISCHARGE AT GOLD CREEK (ds)
SLOUGH 21
JUVENILE ROUND WHITEfiSH
30 ,---------------------------------------------------------.
28
26
24
22
20
18
1 6
14
12
10
a
6
\
OVERTOPPED
44------,------~--~~------~-----.------.------.----~
5 15 25 35 45
(Thousands)
MAINSTEM DISCHARGE AT GOLD CREEK (cfs)
Figure 18. Weighted usable area for juvenile round whitefish at the
Side Channel 21 and Slough 21 study sites.
-28 -
....
-
-
-
""'l
I
r-
-
r
I
Table 2. Correlations between composite weighting factors and catch
transformed by natural log {X+1) for juvenile chinook salmon
by selected sites and by all sites pooled.
Chinook
Low turbidity High turbidity
(-' 30 NTU)
Sig a/
( / 30 NTU)
Site n r n r Sig
Slough 9 48 0.35 0.008 63 0.48 . < 0.001
Side Channel 10 (Insufficient data) 54 -0.08 0.28
All 7 sites
pooled 99 0.40 < 0.001 192 0.25 < 0.001
21 Significance level for rejection of hypothesis that there is no
positive correlation between composite weighting factors and catch.
-29 -
Table 3. Chi-square contingency tests of chum and sockeye salmon
proportional presence by composite weighting factor intervals.
Data from Sloughs 9, 21, and 8A pooled.
Chum
Composite
weighting No. of Cells Proportion
factor i nterva 1 Present Absent Tota1 Present
0.00-0.28 13 28 41 0.32
0.29-0.44 15 21 36 0.42
0.45-0.55 14 21 35 0.40
0.56-1.00 33 10 43 0.77
x2 = 20.05 df = 3
p < 0. 001
Sockeye
Composite
weighting No. of Ce 11 s Proportion
factor interval Present Absent Total Present
0.00-0.07 9 25 34 0.26
0.08-0.14 7 28 35 0.20
0.15-0.38 11 26 37 0.30
x2 = 0.92 df = 2
p < 0.37
-30 -
-'
.-.m
3.3.1 Juvenile salmon
The response of chinook salmon habitat indices to mainstem discharge
varied by site (Figure 19). Habitat indices for juvenile chinook salmon
in Sloughs 9 and 21 showed prominent peaks. Side Channel 10 and Upper
Side Channel 11 chinook salmon habitat indices increased sharply after
the heads were overtopped and then remained fairly constant because
velocities did not become limiting at high discharge levels. Chum
salmon habitat indices at Slough 9 and Slough 21 were very similar and
showed distinct peaks. Sockeye salmon habitat indices at these two
sloughs were very low and decreased slowly with discharge.
3.3.2 Resident species
The response of resident fish habitat indices to changes in discharge
varied greatly by species. Juvenile round whitefish habitat indices
changed in a similar way to chinook salmon habitat indices while Arctic
grayling habitat indices steadily increased with discharge {Figure 20).
Rainbow trout habitat indices at Slough 21 increased with mainstem
discharge while adult longnose sucker habitat indices began to decrease
at the higher mainstem discharge levels (Figure 21).
-31 -
X
la.l
0 z
....
..:(
!:::: m
..:(
I
>< w
0 ::::
0.6 ~-------------------------------------------------------,
---SLOUGH 9
-SIDE CHANNEL 10
0.5 ••••• • UPPER SIDE CHANNEL ll
-·-· SLOUGH 21
-SIDE CHANNEL 21
0.4
0.3
0.2
0.1
/".
/ ' i \
i '\
I i \
\
\
CHINOOK SALMON
\
\
\
\
\.
\
\
0 4------,------.-----~------.------.------.------.----~
5 15 25 35 45
(Thousands}
hiAINSTEM DISCHARGE AT GOLD CREEK (cfs)
0.6~---------------------------------------------------,
--CHUM SALMON
----SOCKEYE SALMON
0.5
0.4
0.3 .
0.2
0.1 SLOUGHS\/ ~
------::---',-,-,,, /SLOUGH 21 ---.... -.......... ____ '--...
04------,------.-----~------.------.------.------.----~
45 5 15 25 35
(Thcusonds)
II\IAINSTEk\1 DISCHARGE AT GOLD CREEK (efs)
Figure 19. Habitat indices for juvenile salmon at IFG modelling sites.
-32 -
-
-
-
-
-
-
-
F""'
.....
r
X
LIJ
0 z
1-< 1-
iii < :t:
X
LIJ
0 z
1-< ~
IIl < :t:
0.6
---SLOUGH9
-SIDE CHANNEL 10
0.5 e. Cl OCI •• UPPER SIDE CHANNEL II
-·--SLOUGH 21
--SIDE CHANNEL 21
0.4
0.3
:·~'·· 0.2 : ·. ·. ... . ....
0.1
\
\
\
\
\
\
\.
'
JUVENILE
ROUND WHITEFISH
\
\
\ ·· ...... .\, ..... .
\
\
\
' ' ...
\
\
\
\
\
\
\ ":-.. ."G.,...
0,_--~-.~---.----L,----~------------~----T---~
O.B
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
5
15 25 35
(Thousands)
MAINSTEM DISCHARGE: AT GbLD CREEK (cfs)
ADULT
ARCTIC GRAYUNG
' /
fo--SLOUGH 21
·./
~'i
SLOUGH 9 ----1 /
I /
I /
1/
I • r/
I.
.1/
,.1• .,-;-"' ................. _____ ............
15 25 35
(Thousands)
MAlNSTE:M DISCHARGE: AT GbLD CREEK (cfs)
45
45
Figure 20. Habitat indices for juvenile round whitefish and adult
Arctic grayling at IFG modelling sites.
-33 -
0.6 ~------------------------------------------------------~
0.5
0.4
X
I.J
0 z ADULT
1-0.3 ROUND WHITEFISH
< !:: ----III < ' ,~ADULT :r: ...... i LONGNOSE SUCKER 0.2
0.1
0,_------.-----~------~----~------~------r-----~----~
5 15 25 35
(Thousands)
MAINSTEM DISCHARGE AT GbLO CREEK (cfs)
45
Figure 21. Habitat indices for adult rainbow trout, round whitefish,
and longnose suckers at the S1ough 21 modelling site.
-34 -
-
-
-
-I
I
-
.....
I
I
-
4.0 DISCUSSION
4.1 Limitations of the Data
The assumptions of the incremental method of habitat analysis by cal-
culating weighted usable areas have been outlined by Orth and Maughan
(1982). As applied here, these assumptions are (1) cover, velocity, and
depth are the most important variables affecting fish abundance when
flow regime changes are considered; (2) the stream channel is not
altered by changes in flow; {3) cover, velocity, and depth are indepen-
dent in their influence on habitat selection by juvenile salmon; (4) the
reach can be modelled by reference to a few study areas; and (5) there
is a positive relationship between weighted usable area and habitat use.
The initial assumption is a difficult one to evaluate as changes in flow
regime may have important effects on such factors as the food supply by
affecting water quality. Turbidity is a factor which may have major
direct and indirect effects on fish distribution but which was addressed
only for chinook salmon indirectly by its use as cover. Analysis is
also specific to, the ice-free months and no analysis for effects of
winter processes has been made. The importance of shoreline area cover
to the suitability of offshore areas for rearing juvenile coho is
similarly unknown.
Channel morphometry of the sites studies appeared to be stable during
the period of study. At Slough 9, however, an IFG-4 modelling site,
1 arge amounts of si 1t were deposited during a-flood event in September
1982 (Estes and Vincent-Lang 1984). Long term changes in channel
morphometry are therefore possible.
Cover, velocity, and depth are probably not independent in the·ir ·influ-
ence on habitat selection by young salmonids. Analysis of variance
indicated that there is a significant interaction between depth and
velocity for juvenile chinook and coho salmon catch (Part 3 of this
report). Since depth was set to 1.0 over most of the range, this
interaction became of little importance. Interactions between cover and
velocity are also likely but should not have large effects on WUA
projections.
The fourth assumption of the representativeness of the sites studied was
probably not met because of several reasons. The study sites showed
large variations in response to discharge which makes the concept of a
representative site difficult to formulate. The two upland sloughs, in
particular, showed large differences in response to changes in mainstem
discharge (Part 4 of this report). The Susitna River reach under
consideration is a vast mosaic of side channels, side sloughs, and
upland sloughs which overtop at many different discharges. The thirteen
sites modelled are representative of a large part of the habitat in this
reach but do not include the mainstem or the mid-river side channels.
The correlations and proportional presence by composite weighting factor
interval for the four species suggest that there is a positive relation-
ship between the weighted usable area and habitat use at the cell level
and, by inference, at the site level. Such factors as season and site
-35 -
are also importants however (see Part 2)s and much of the variation in
catches of fish is not explained by the composite weighting factors.
In summarys some of the assumptions of incremental analysis of habitat
may be violated but the effects of these violations on the analysis are
difficult to evaluate. The correlation and contingency table analysis,
however, suggest that the simulations are related to actual fish use of
the sites.
When interpreting the results of the habitat models presented in this
paper, it is helpful to consider how close the discharge regime of the
open water season of 1983 was to a typical year. Figure 22 shows that
June, July, and September discharges were a little lower than the 30
year mean and that the August discharge was higher.
4.2 Comparison of IFG Models with RJHAB
4.2.1 Model characteristics
A comparison of the characteristics of the IFG models and RJHAB as used
in this study is summarized in Table 4. The IFG models are based on an
underlying theory of hydraulics which enables a simulation of physical
conditions that were not actually measured. RJHAB can not simulate
physical conditions because cell measurements were not ta~en in exactly
the same physical location each time, and therefore can not be used to
project velocities or depths at a study site. It doess however, model
habitat which is based on physical measurements and this habitat can be
interpolated between actual measurements.
The enormous capacity of the IFG models to predict detailed information
on depths and velocities is perhaps overkill when the question to be
answered is the availability of rearing habitat. Juvenile salmon and
resident fish do not necessarily respond to increments of velocity and
depth on the order of 0.1 ft/sec or 0.1 ft. Fish will select an area
that has a general range of velocities· or depths. Further, factors
other than the variables simulated by the IFG modelss such as· food
availability, probably override small differences in depth or velocity
in influencing fish density. Restricted access into Slough 8As for
examp 1 e, caused by beaver dams and 1 a ck of overtopping flows 1 i mited
juvenile chinook use of the site. · The IFG models are probably more
useful in modelling salmon spawning habitat, where the variables which
the IFG model is good at simulating (depths velocitys substrate) are
also of primary importance to the fish. The IFG models in 1983 were
mainly used to model salmon spawning habitat; hence, the quality of
cover data obtained was lower than would have been desirable from the
standpoint of rearing habitat. RJHAB was specifically designed to
consider the effect of discharge on cover.
Another benefit of RJHAB is that the field data collection effort
required is considerably less than of the IFG models. This enabled us
to sample a larger range of habitat types in the reach. Also, RJHAB can
be used in more complex sites or sites such as upland sloughs which are
primarily backwater areas.
-36 -
-
-
-
-
"""
l
IO'
d ...
"' "' "' .. :z: u
"' 10
0
IO'
.: ..
"' "' a:
<( :z:
u
~ 10'
0
]
JUNE Mean Monthly DischorQo
(USGS Gating $lotion "2920001
• •1981
4 oi9U
• •1983.
•o 10 ~o •o so eo JO eo 80 100
% Of TIME DISCHARGE EQUALLED OR UCHDEO
AUGUST Mean Monthly Diachoroe
CUSGS Gaging Slalion 1~2SZOOO)
• '1981
.... 1982
• •1983 1
~ R ro ~ ~ W N W H 100
% OF TIME DISCHARGE EQUALLED OR EXCEEDED
.. l ]
10 Moon Monthly Di1chorgo
(USGS Gaging Station 15212000)
•• 1981
... •1082
• •1913 •
-;
~ ..
"' "' "' "" :z: u
!! •••
0
0 10 l:o ;,o .co $0 eo 10 eo 'o 100
% Of TIME DISCHARGE EQUALLED OR EXCEEDED
IO' SEPTEMBER Moon Monthly Diachorgo
I USGS Gaging Slotlon IDZiZOOOl
•• till
4 ai98Z
• •1983 •
-;
;;
"' "' a:
"" :z: u
"' 10'
0
10 20 .30 "'0 M 60 70 10 90 100
% Of TIME DISCHARGE EQUALLED OR EXCEEDED
Figure 22. Time duration curves and mean monthly discharges for June, July, and August, and September based
on the 30 year record of Susitna River discharge at Gold Creek. Sources: time duration curves -
Bredthauer and Drage (1982); mean monthly discharges -USGS (1982), Lamke et al. (1983), and USGS
provisional data.
Table 4. Comparison of model characteristics of IFG models and RJHAB.
Parameter
Transects
Measurements
Data collection
No. of calibration
measurements
Extrapolated
range ·
Total surface area
Physical simulation
Resolution
Computer
Cost
Upland sloughs
WUA
IFG Model
4 to 11
point specific
intensive
1 to 4
40-250% of
calibration range
yes
yes
fine
mainframe
more
no
standardized to
1,000 ft reach
-38 -
RJHAB
8 to 9
300 sq ft cells
less intensive
4 to 6
5,000 to 45,000 cfs
yes
no
coarse
micro
1 ess
yes
depends on size of
site but could be
standardized to a
1,000 ft. reach
-
-
-
-
-
-I
-
4.2.2 Model output ·
The output from the IFG models and RJHAB can be directly compared in at
least two different ways: 1) compare percent change in weighted usable
area over similar increments of mainstem discharge, and 2) compare the
habitat index plots. The actual values of WUA are not comparable
without modification because the IFG WUA•s are standardized to a linear
reach of 1,000 ft while RJHAB was calculated based on the size of the
site.
Generally, the shape of the habitat index curves for chinook salmon
juveniles are similar for side sloughs and side channels modelled by the
IFG models and RJHAB (Figure 23). The RJHAB curves have been smoothed
and extrapolated to the discharge range 5,000 to 45,000 cfs. The
habitat index for chinook juveniles is the highest at a discharge level
which is slightly (within a few thousand cfs) higher than that required
to overtop the head of the site. This is because chinooks prefer
moderate flows and moderately turbid water. As the discharge levels
increase further, the velocity at the sites becomes too great and the
habitat index decreases.
The habitat indices calculated for coho salmon from RJHAB are generally
low. The same would be true from the IFG models, had we calculated
them. The highest habitat indices are from the two upland slough sites,
Slough 5 and Slough 6A. This is in agreement with the observed dis-
tribution of coho salmon; the density of this species in turbid waters
is low (see Part 2 of this report).
Chum habitat indices were similar to those for chinook in that a dis-
charge slightly over the overtopping point produced the maximum habitat
index.
Sockeye habitat indices were generally low. The highest indices were
for upland sloughs, which are the most lake-like of all the macrohabitat
types. Generally, this reach of river is not prime sockeye rearing
habitat (see also discussion in Part 1 and Part 2 of this report).
There are not very many upland sloughs available. Neither the IFG model
or RJHAB successfully predicted the heavy use of side sloughs by sockeye
juveniles. This use is more a result of side sloughs being the dominant
sockeye spawning grounds in this reach of river than it is a result of
the quality of the rearing habitat available in side sloughs.
Sockeye habitat indices increased in side sloughs with increasing
discharge as surface area increased. After the heads of the sites were
overtopped by mainstem water, the habitat index started to decline
sooner than did the habitat indices for chinooks and chums. This
reflects the preference of sockeye juveniles for lower velocity water
than the other two spe.ci es.
Habitat indices for all species in upland sloughs increase steadily as
mainstem discharge increases. This is mainly a function of increased
surface area attributable to the backwater effect of mainstem stage at
the mouth of these sites. Similar results were obtained by the 1982
-39 -
X ....
0 z
.... -< !::: m -< :r
X w
Q
3:
1-< t:
Ill < :z:
0.8,--------------------------------------------------------,
0.5
0.4
0.3
0.2
RJHAB SITES
0 II•• .. II
+ WllleiLen CrulL II••••
0 lillie Cllee•el 1 OA
-C•llllretelll llue•
--••t••••••t•lll ......
/
,
" /
I
I
I
0.1 I
I ++--·---+---+---+---
0#---o---o---o---o---a
CHINOOI< SALMON
+'
\
' \ + \
\
\
\
\.
' ' ' \
' ' o,
....... 0 ......
.......... o--...... o __
"-,
' '+ .........
..... a .........
..... a ............
.... o ___ o __ _
I 0.04-~---,------~----~------~----~------.------.----~
5 15 25 35
SUSITNA R. DISCHARGE (CFS x 1000) AT GOLD CREEK
0.6 ~---------------------------------------------------,
0.5
0.-t
0.3
0.2
0.1
lFG SITES
---SLOUGH 9
--SID£ CHANNEL 10
••• • • • UPPER SIDE CHANNEL II
-·-·SLOUGH 21
--SIDE CHANNEL 21
/".
/ " i
.I
\
\
CHINOOK SALMON
\
\
\
\
\
\
\
\
\.
' \
0~-----r----~------.-----T-----.------.-----.----~
5 15 25 35
(Thousands)
... AINSTE ... DISCI-fARGE AT GOLD CREEK (cfs)
45
Figure 23. Comparison of RJHAB and IFG habitat indices for
juvenile chinook salmon.
-40-
-
-
-
-
-
-
-
-
study that specifically examined the effect of the backwater phenomenon
on rearing habitat (ADF&G 1983c). At very low mainstem discharges,
cover may also be lost around the shoreline of sites such as Slough 6A
where undercut banks and overhanging riparian vegetation are present.
4.3 Summary of Seasonal Habitat Projections for Rearing Salmon and
Resident Fish
An examination of the figures in which chinook weighted usable area is
plotted versus mainstem discharge and versus time of season shows that
some sites provide the most weighted usable area when discharge is low
{e.g., Lower Side Channel 11), some when discharge is at an intermediate
level {e.g., Slough 9), and some when discharge is high. The control-
ling factor is the discharge at which the head of the site is
overtopped. The maximum weighted usable area for chinook at most sites
occurred at a discharge slightly greater than the overtopping discharge.
Therefore, chinook weighted usable area in this reach of ·river would
theoretically be the highest at the discharge level which just overtops
the maximum number of sites (the size of each site must also be c·on-
sidered).
There is undoubtedly a correlation between a decline in weighted usable
area at the rearing sites and re-distribution of juvenile salmon. If a
rearing area is essentially saturated by fish and then weighted usable
area decreases, some fish are forced to leave. We have observed this at
sites such as Slough 22 where chinook juveniles were abundant when the
head was overtopped and 1 ess abundant when the water c 1 eared after
ma·instem water no longer entered the slough. Also, we have demonstrated
a positive correlation between composite weighting factors and juvenile
salmon density.
The fish that are forced out of a certain site must either seek a new
rearing site or, under more extreme conditions, migrate out of that
reach of river. In the latter situation, there should be an increase in
the capture rate at the downstream migrant traps. It is difficult to
discern such a relationship with the 1983 data. The outmigration rate
of chinook juveniles was relatively low when the weighted usable area at
Slough 9 was high and the outmigration rate was high when WUA at Slough
9 was lowest (disregarding the month of September, when discharge was
low). However, this relationship was reversed at other sites. Ideally,
only the best rearing sites should be considered in this approach. This
relationship may also be obscured by major outmigrations from the
tributaries which have little to do with changes in mainstem conditions.
There is also the larger question of whether in fact rearing habitat is
limiting to salmon. If the number of fry emerging from the gravel is
not enough to saturate the available rearing habitat, then there would
be more flexibility with regard to varying discharges. In our
experience on the Susitna River, both saturation and under-uti 1 ization
of rearing habitat occurs. A partial explanation is that there is no
substantial amount of spawning above the upper end of this reach.
Therefore, when waves of juvenile chinook and coho migrate out of
Portage Creek, they probably saturate a certain portion of the available
rearing habitat in the Susitna River downstream of the Portage
-41 -
Creek confluence until they have had sufficient time to re-distribute
further downstream. During other periods of time, when few fish are
migrating out of Portage Creek, these same rearing areas may not be
saturated, especially if an intervening per:>iod of poor habitat (dis-
charge too low or too high) has caused the previous occupants to leave
the area. We have observed this at such sites as Slough 22 and Slough
21 on occasions when habitat conditions appeared to be relatively good
(and weighted usable area was high); yet, fish density was low relative
to other times of apparently equal habitat quality.
It seems almost certain that rearing habitat is 1 imiting for sockeye
juveniles in this reach of river. The deeper, low velocity, relatively
clear water that they prefer does not occur in the reach in large
quantities (Klinger and Trihey 1984). A high proportion of the
young-of-the-year fish 1 eave this reach (based on downstream migrant
trap catch rates, see Part 2). The Age 0+ fish must either rear in the
lower river or die, because only a miniscule number of adult sockeyes
migrating upstream past the Talkeetna Station outmigrated to the ocean
as Age 0+ fish. The majority of adults are 42 •s (Barrett et al. 1984).
It has been conclusively shown (Part 1) that chum salmon rear in this
reach of river because they show substantia 1 growth between emergence
and outmigration. The correlation of fhum catch per hour at the outmi-
grant traps and discharge was high (r = 0.79, see Part 2), suggesting
that high water events displace or trigger outmigration by chums rather
than contribute to suitable habitat. If rearing habitat became re-
stricted because of low discharge, the fish would probably leave this
reach later rather than sooner because of the lack of a high water pulse
that might trigger outmigration.
Although few data on winter distribution are available, there are strong
indications of substantial changes in macrohabitat use during the
winter. Discharge levels are much reduced and the mainstem water
becomes clear. Many chinook and coho juveniles move out of tributaries
to overwinter in the mainstem. There appears to be a trend in the fall
that has been noticed for three consecutive years in which chinook and
coho move into the deeper slough areas. There may be a thermal at-
traction produced by upwelling water in the sloughs.
Resident fish use of both microhabitat and macrohabitat is closely
linked to turbidity and apparently to food supply. Juvenile round
whitefish are found in the small side channels which have a low flow, so
distribution is tied to discharges at which the heads of these side
channels are slightly overtopped.
The use of side sloughs by most species of adult resident fish is
probably limited by the very small amount of flow through these sites.
As heads are overtopped and flows increase, the sites rapidly become
more favorable for adult resident fish. These fish also use portions of
the mainstem for rearing. The rearing habitat may be limiting but this
is not likely due to lack of suitable open water season cover, depths,
or velocities. It is more likely to be attributable to other factors
such as overwintering mortality or food supply, as densities of resi-
dents are low almost everywhere in mainstem-influenced sites with the
-42 -
-
-
-
-
-
-
-
-
exception of selected tributary or slough mouths where fish may gather
to feed on salmon eggs, outmigrating juvenile salmon, or invertebrates.
In conclusion, the results presented in this part and the data and
analysis from parts one through six of this report suggest the following
trends:
(1) Of the salmon juveniles rearing in the Susitna River, chinook and
chum appear to make the best use of habitats associated with the
mainstem and also have the most abundant adult returns {even year
pink salmon excluded) in this reach of the river. Juvenile coho
salmon apparently rear primarily in tributaries, but wi 11 take
advantage of the upland slough habitat that is available.
{2) Sockeye salmon appear to be most heavily limited by rearing habitat
with highly successful incubation, but limited rearing, occurring
in this reach of river. Either rearing survival is low or rearing
takes place in the lower river. Successful rearing does occur
within limited portions of some of the upland and clear water
sloughs but is probably minor when compared to the total population
of emergent fry. Apparently, sockeye rearing does not occur in
tributaries to any great extent.
(3) Of the habitats affected by mainstem discharge, microhabitat within
side channels/side sloughs is most affected, primarily by dewater-
ing, lowered turbidity, and lower water velocity after the head is
no longer overtopped by mainstem flows. This habitat is heavily
used by chinook juveniles, who appear to be limited by cover when
the sites are not turbid (generally associated with the heads not
being overtopped). Maximum habitat value for chinook salmon is
obtained at a discharge level slightly greater than the overtopping
discharge level.
{4) Wintering habitat for all rearing species is heavily dependent on
mainstem habitats as indicated by spring and fall migratory move-
ments. The models presented have not been designed to evaluate
habitat conditions during the winter.
(5) Resident species using mainstem habitat areas are most predictively
associated with levels of turbidity and appear limited by food
supply. They often associate with the mouths of clear water
tributaries or with spawning salmon. · The response of primary
productivity of the system may be more indicative of the response
of resident species than the values generated by habitat simulation
based on hydraulic models.
The results and discussion presented in this report do not conclude the
analytical effort required to use this information in a decision making
process. It remains to integrate these results with the studies con-
ducted on adult anadromous spawning and to further extrapolate our study
sites to the entire reach of river which they were chosen to represent
using the surface area information provided by Klinger and Trihey
(1984). Further, these results must be weighted with respect to the
importance of the harvestable adults of each species. Finally, these
-43 -
results must be portrayed in such a manner as to depict the effects of
alternative flow regimes on different species so that the flow require-
ments of different management goals can be ascertained. Future reports
prepared by other investigators wi 11 use this report to ultimately
provide the above information.
-44 -
-
-
-
-
-
[
I
r
-
5.0 CONTRIBUTORS
IFG hydraulic model data collection was done by the Aquatic Habitat
Group of the Su Hydro Aquatic Studies. Kim Sylvester of the AH group
and Diane Hilliard of E. Woody Trihey and Associates calibrated the
hydraulic models. Diane Hilliard input cover data, the suitability
criteria, and ran the PHABSIM habitat models which generated weighted
usable areas. Bob Marshall made some helpful suggestions.
Drafting was done by Sally Donovan and Carol Riedner. The typing was
done by Skeers Word Processing Services.
-45 -
''"'-~-------------"--------..;..,..,---------------------
6.0 ACKNOWLEDGEMENTS
Funding for this study was provided by the Alaska Power Authority.
We wish to thank the various staff at E. Woody Trihey and Associates
(EWT&A)~ the Arctic Environmental Information Data Center, Harza-Ebasco
Susitna Joint Venture, Woodward-Clyde Consultants~ the Alaska Power
Authority~ and especially Cleve Steward (EWT&A) for reviewing a draft of
this paper and providing helpful suggestions.
-46 -
-
-
-
-
-
-
-
7.0 LITERATURE CITED
Alaska Department of Fish and Game. 1983a. Susitna Hydro aquatic
studies phase II basic data report. Volume 4 (3 parts). Aquatic
habitat and instream flow studies, 1982. Alaska Department of Fish
and Game Susitna Hydro Aquatic Studies. Anchorage, Alaska.
__ 1983b. Use of major habitat types by juvenile salmon and resident
species. Appendix G to Synopsis of the 1982 aquatic studies and
analysis of fish and habitat relationships (2 parts). Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies.
Anchorage, Alaska.
1983c. Influence of habitat parameters on distribution and
relative abundance of juvenile salmon and resident species.
Appendix F to Synopsis of the 1982 aquatic studies and analysis of
fish and habitat relationships (2 parts). Alaska Department of
Fish and Game Susitna Hydro Aquatic Studies. Anchorage, Alaska.
Barrett, B.M., F.M. Thompson, and S.N. Wick. 1984. Adult anadromous
fish investigations: May -October 1983. Alaska Department of
Fish and Game Susitna Hydro Aquatic Studies. Report No. 1.
Prepared for Alaska Power Authority. Alaska Department of Fish and
Game, Anchorage, Alaska.
Bovee, K.D. 1982. A guide to stream habitat analysis using the In-
stream Flow Incremental Methodology. Instream Flow Information
Paper 12. U.S.D.I. Fish and Wildlife Service, Office of Biological
Services. FWS/OBS -82/26.
Bovee, K.D., and R.T. Milhous. 1978. Hydraulic simulation in instream
flow studies: theory and techniques. Instream Flow Information
Paper 5. U.S.D.I. Fish and Wildlife Service, Office of Biological
Services. FWS/OBS -78/33.
Bredthauer, S., and B. Drage. 1982. Susitna Hydroelectric Project.
Appendix B.9. River Morphology. Report prepared for Acres Ameri-
can, Inc. by R&M Consultants, Anchorage, Alaska.
Estes, C., K. Hepler, and A. Hoffman. 1981.. Willow and Deception
Creeks instream flow demonstration study. Volume 1. Alaska
Department of Fish and Game, Habitat Protection and Sport Fish
Divisions. Prepared for the U.S. Department of Agriculture, Soil
Conservation Service, Interagency Coop. Susitna River Basin Study.
Estes, C.C., and D.S. Vincent-Lang (eds.). 1984. Aquatic habitat and
instream flow investigations, May -October, 1983. Alaska Depart-
ment of Fish and Game Susitna Hydro Aquatic Studies. Report No. 4.
Prepared for Alaska Power Authority. Alaska Department of Fish and
Game, Anchorage, Alaska.
-47 -
Klinger, S., and W. Trihey. 1984. Response of aquatic habitat surface
areas to mainstem discharge in the Talkeetna to Devil Canyon reach
of the Susitna River, Alaska. Report prepared for Alaska Power
Authority by E. Woody Trihey and Associates, Anchorage, Alaska.
Lamke, R.D., P.J. Still, 8.8. Bigelow, H.R. Seitz, and J.E. Vail. 1983.
Water resources data. Alaska. Water Year 1982. U.S. Geological
Survey Water -Data Report AK-82-1. USGS, Anchorage, Alaska.
Orth, D.J., and O.E. Maughan. 1982. Evaluation of the ·incremental
methodology for recommending instream flows for fishes. Trans-
actions of the American Fisheries Society. 111: 413-445.
U.S. Geological Survey. 1982. Water resources data. Alaska. Water
Year 1981. U.S. Geological Survey Water -Data Report AK-81-1.
USGS, Anchorage, Alaska.
Wilson, W.J., E.W. Trihey, J.E. Baldrige, C.D. Evans, J.G. Thiele, and
D.E. Trudgen. 1981. An assessment of environmental effects of
construction and operation of the proposed Terror Lake hydroe 1 ec-
tric facility, Kodiak, Alaska. Instream Flow Studies final report.
Arctic Environmental Infonnation and Data Center. University of
Alaska, Anchorage.
-48 -
-
-
-