HomeMy WebLinkAboutAPA2836"'"-=--.'::'"""=------------"----~-------------------
--
....
....
........
ALASKA DEPARTMENT OF FISH AND GAME
SUSITNA HYDRO AQUATIC STUDIES REPORT SERIES
,-
I
I
r-
I
r
I
ALASKA DEPARTMENT OF FISH AND GAME
SUSITNA RIVER AQUATIC STUDIES PROGRAM
REPORT NO.7
RESIDENT AND JUVENILE ANADROMOUS FISH
INVESTIGATIONS (MAY -OCTOBER 1984)
PARTS 1 AND 2
Editors:Dana C.Schmidt,Stephen S.Hale,
and Drew L.Crawford
Prepared for:Alaska Power Authority
334 W.Fifth Avenue,Second Floor
Anchorage,Alaska 99501
July 1985 ARLIS
Alaska Resources
Library &InformatIOn ServIces
Anchorage,Alaska
-.
I~
-..-
0)
o:::t
No:::t,c~o:::t
0
0
0
l!)
l!)
I"
('I)
('I)
r--
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 feasib"llity of the
proposed Susitna Hydroelectric Project.The ADF&G Susitna River Aquatic
Studies Program was initiated in November 1980.
The report covers studi es conducted from May through October 1984 of
juvenile salmon and resident fish species of the Susitna River.In
addition,some information is included on overwintering of resident fish
radio-tagged in 1983.The majority of the effort during the 1984
open-water season was on the lower river (from the mouth to the Chulitna
River confluence).No studies were conducted this year in the area
above Devil Canyon.This volume consists of three parts.
Part 1 (RSA Tasks 16A and 16B)covers the migration and growth of
juvenile salmon.Coded wire tagging of chum and sockeye fry in the
middle river (Chulitna River confluence to Devil Canyon)and collecting
of a 11 species of outmi grati ng fry at Tal keetna Station were simil ar to
1983 studies.In addition,a mark-and-recapture cold branding study was
conducted in tributaries,sloughs,and side channels of the middle river
to obtain estimates of chinook and coho juvenile salmon population size
and residence time in these rearing areas.Also,outmigrant traps were
operated near the mouth of the ri ver at Fl athorn Station (River Mi 1e
22.4)to obtain a timing index of outmigration from the lower river.
A statistical time series analysis of 1983 and 1984 discharge,
turbidity,and juvenile salmon outmigration data from the middle river
is included as an appendix.
Studies of the distribution and relative abundance of juvenile salmon
and modelling of rearing habitat in the lower river are discussed in
Part 2 (RSA Tasks 14 and 36).These studies were similar to those
conducted in the middle river in 1983.Habitat suitability criteria
developed for the middle river were used for the lower river unless
evidence of different conditions in the lower river necessitated modifi-
cations.Results from habitat modelling at 14 RJHAB model sites and 6
IFIM model sites are presented.The RJHAB and IFIM models were compared
by using both at two sites.IFIM model calibration is contained in
Appendix D.
Part 3 (RSA Task 14)presents the results of resident fish studies in
both the mi dd1 e and lower river.Moni tori ng of fi sh movement through
use of radio tags was continued.Index sites in the middle river were
sampled as part of the long term monitoring effort.Population esti-
mates for rainbow trout,Arctic grayling,round whitefish,and longnose
suckers in the middle river were made from multiple year mark-recapture
data.
Questions concerning this report should be directed to:
Alaska Power Authority
334 West 5th Avenue
'Anchorage,Alaska 99501
Telephone:(907)276-0001 ARLIS
Alaska Resources
Library &InformatIOn SenrtCf'
Anchorage,Alaska".
TITLES IN THIS SERIES
Report Publication
Number Title Date
~1 Adult Anadromous Fish Investigations:April 1984
May -October 1983
~2 Resident and Juvenile Anadromous Fish July 1984
Investigations:May -October 1983
r-3 Aquatic Habitat and Instream Flow September 1984
Investigations:May -October 1983
4 Access and Transmission Corridor Aquatic September 1984
Investigations:May -October 1983
5 Winter Aquatic Investigations:March 1985
,rQ...September 1983 to May 1984
6 Adult Anadromous Fish Investigations:June 1985
,.....May -October 1984
7 Resident and Juvenile Anadromous Fish July 1985
,-,Investigations:May -October 1984
8 Availability of Invertebrate Food Sources 1985
for Rearing Juvenile Chinook Salmon
in Turbid Susitna River Habitats
9 Summary of Salmon Fishery Data for June 1985
N Selected Middle Susitna River Sites
;8
-
--
.-
Part 1.
Pa rt 2.
Pa rt 3.
CONTENTS OF REPORT NO.7
The Migration and Growth of Juvenile Salmon in the
Sus itna Ri ver.
The Relative Abundance,Distribution,and Instream Flow
Relationships of Juvenile Salmon in the Lower Susitna
Ri ver.
Resident Fish Distribution and Life History in the
Susitna River below Devil Canyon.
......
-
-
-
.-
-
PART 1
The Migration and Growth of Juvenile Salmon
in the Susitna River
-
,..,.
-.
-
THE MIGRATION AND GROWTH OF JUVENILE SALMON
IN THE SUSITNA RIVER
Report No.7,Part 1
by Kent J.Roth and Michael E.Stratton
Alaska Department of Fish and Game
Susitna River Aquatic Studies Progra~
620 East 10th Avenue,Suite 302
Anchorage,Alaska 99501
ABSTRACT
Studies of salmon spawning,embryo incubation,and juvenile rearing are
all critical in understanding the current life history and habitat
dynamics of salmon in the Susitna River.However,the final measure of
the value of a reach of river to the freshwater life stages of salmon is
the number and condition of the fry which outmigrate from the reach to
the ocean.Baseline data on salmon outmigration have been collected at
Talkeetna Station (river mile 103.0)for the past three years.The data
from 1982 and 1983 have shown that a substantial number of chinook,
coho,and sockeye fry outmigrate from the middle river during their
first sUlllJler.Because the majority of returning adults have spent at
least one winter rearing in freshwater,an important question was
whether these age 0+fish overwintered in the lower river or had a low
survival rate.To help answer this question,outmigrant traps were also
operated near the mouth of the Susitna River (RM 22.4)during 1984.
Mark and recapture studies gave population estimates for chum and
sockeye fry (marked by coded wire tags)in the Susitna River above
Talkeetna Station (middle river)and for chinook fry (marked by cold
branding)in Indian River and other rearing sites.The cold branding
study also monitored outmigration timing from Indian River and obtained
estimates of juvenile chinook residence time in mainstem rearing areas.
The Talkeetna River and Deshka River were intermittently sampled to help
explain the mainstem outmigrant trap data.A portion of the age 0+
chinook fry apparently outmigrate from the middle river upon reaching a
critical size but a large number remain to overwinter and then out-
migrate during their second summer.Coho fry outmigrate at a wider
range of lengths than chinook fry so the cumulative biomass of coho fry
lags behind the cumulative numbers of individuals by one or two weeks.
Age 0+chinook and coho fry grow about 30 mm in 1ength duri ng the
i
-open-water season.Juvenile sockeye salmon appear to seek out lake-like
rearing areas at a size of about 50 mm.The limited amount of this
habitat type in the middle river is the major influence on their redis-
tribution to the lower river.The estimated 1984 middle river
population size was about 300 t OOO for age 0+sockeye and 2 t 040,OOO for
chum fry.Chum fry rearing in the middle river was demonstrated by
their growth and by analysis of stomach contents.~
-
-
.....
-
,.....
ii
I."
ABSTRACT ••.••••••••••
TABLE OF CONTENTS
LIST OF FIGURES •••
.............................................
.'.
i
vi
LIST OF TABLES ••••••••;•••••.•••
-
...........'.
LIST OF APPENDIX TABLES ••••••••••.•••••••••..••••••
xii
xiii
1.0
2.0
I NTROD.UCT ION •••••••••••••••••••••••••••••••••••••••••••'••'••••
METHODS •••••••.•••••
1
3
Study Locations ••••••••2.1 ....................................3-2.1.1 Flathorn Station ••••••...................'..
2.1.2 Deshka River ......................
2.1.3 Talkeetna Ri ver .•••.••................
2.1.4 Talkeetna Station ••••••••.............
2.1.5 Coded wire tagging ••••..............
2.1.6 Cold branding ........................................
3
3
3
3
3
10
Recording •••2.2 Field Data Collection and .....................11
2.2.1
2.2.2
2.2.3
2.2.4
2.2.5
2.2.6
Flathorn Station outrrtigrant traps •••••••
Deshka River outmigrant weir ••••••.••••.
Talkeetna River beach seining •••••••••.•••••••
Talkeetna Station outmigrant traps ••.
Coded wire tagging •••.••••••••
Cold,branding ••••••••••••••••~••••••••••••••••
11
11
12
12
12
13
2.3 Data Analysis •••••••.••••••14
3.1 Chinook Salmon .
3.0 RESULTS •••.••••••
2.3.1
2.3.2
2.3.3
Juvenile salmon catch per unit effort.
Population and survival estimates.
Time series analysis •••••••••••••••.•••.•
............................................
14
16
16
17
17
3.1.1 Catch per unit effort...................................17
3.1.L1
3.1.1.2
Age 0+••.•.•••••..•••••••
Age 1+•....•••••••.••••..
17
19
3.L2 Growth •.••••27
3.1.2.1
3.1.2.2
Age 0+•.•
Age 1+••.
iii
27
27
TABLE OF CONTENTS (Continued)
3.1.3 Cold branding .
3.1.4 Population estimates .
Page
27
31
'"""
3.2 Coho Salmon •••••••..•••••..••...•••••••c •••••••••e ••••••••o 35
3.2.1 Catch per unit effort..................................35
3.2 .1.1 Age 0+c ••••••1!I •iii • • • • • • • • •~5
3.2.1.2 Age 1+and older o.........................39
3•2•2 Growt h•••••••••••••••••••••••••••• ••••••••••••••••••••• • 39
3•2•2•1 Ag e .0+D •
3 ..2.2.2 Age 1+and older e •••••••••D.
3.2.3 Cold branding ...•••..........•.•.•.•.•.....••.o ••eoeosc
3.2.4 Population estimates oa •••••••••o.c ••••••••
39
44
44
44
3.3 Sockeye Sa 1mon '00 ••0 •••CI a ••••••••••••0 •eo.••44
3.3.1 Catch per unit effort..................................44
3•3 •1.1 Ag e 0+0 ••••••••••••••••e •••c ••D iii)e e ••e _.
3.3.1.2 Age 1+0 ••••••••c •••c •eo •e III ••••••••••
3 e 3•2 Growt h••••••••••••••••••••••••• •e ••&••II •II • • • • ••D •C •••••
3.3.3 Coded wire tagging and recovery .•••••••••••••..••..•.••
3.3.4 Population estimates and survival rates of outmigrants.
47
47
54
54
60 -
3.,4 Chum Salmon _o ••••••••••__._•••••o •••_.60
304el Catch per unit effort ..o•••••_••••ea.oe ••••eo.m ••••••••
3•4•2 Growt h••••CI ••••CI ••• •e ••••••••G CI • • ••••••Ii!•••e eo ••••e _C'•••
3.4.3 Coded wire tagging and recovery ••.•.•••••••••..••••.•.•
3.4.4 Population estimates and survival rates of outmigrants.
60
64
64
64
3.5 Pink Salmon ..__ __..67
3.6 Descriptive Statistics for Catch and Environmental Variables.67
4.0 DISCUSSION...................................................74
4.1 Chinook Salmon __74
_.
4.1.1 Ou tmi gra t ion _C'•eo •••••••••••••e ••••a _••••••0 CI 0 ••••
4.1.2 Freshwater life history •....•••......•..•.•..•.•.•••...
4.1.3 Estimates of population size and residence time .••....•
4.1.4 Growt h__ _ _ _.__
iv
74
74
78
79
-
TABLE OF CONTENTS (Continued)
4.2 Coho Salmon................................................79
4.2.1 Outmigrati-on..79
4.2.2 Freshwater life history .••.•.......••••...•..••..~.....83
4 .2.3 Growth.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83
-
,~
4.3 Sockeye Salmon -.
4.3.1 Outmigration · ·.
4.3.2 Freshwater life history .•••...•••.•......•......•......
4.3.3 Estimates of population size and survival.....•.•....•.
4.3.4 Growth e •••••••III
88
88
88
91
92
4.4 Chum Salmon-••••.•••.••••.•••••••..•••••••••••••-••••••D.....96
4.4.1 Outmigration 10.....96
4.4.2 Freshwater life history................................96
4.4.3 Estimates of population size and survival..............96
4.4.4 Growth.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96
4.5 Pi-nk Salmon................................................99
4.5.1 Outmigration --99
4.5.2 Freshwater life history................................99
5.0 CONTRIBUTORS ••••.••_••••••••.••_................................101
6.0 ACKNOWLEDGEMENTS ••••••••.e • • • • • • • • • • • • • • • • • • • • • • • •••• • • • • • • • •103
7.0 LITERATURE CITED •••••••-.......................................104
8.0 APPENDICES
Appendix A Juvenile Salmon Catch and Length
Data,1984
Appendix B The Schaefer Estimate of Population
Size
Appendix C Time Series Analysis of Discharge,Turbidity,
and Juvenile Salmon Outmigration in the
Susitna River~Alaska
v
L1ST OF FIGURES
Figure Title
1 Map of juvenile salmon outmigration study
field stations in the Susitna River basin,
1984 "ID •••••ell.If!•.ilI a 4
River,1984 ...............................•0..........9
6 Ma p of coded wi re taggi ng and cold brandi ng
sites in the middle reach of the Susitna
2
3
4
5
7
8
9
10
Map of the stationary outmigrant trap and the
mobile outmigrant trap sampling points on the
Susitna River at FlathornoStatjon,1984 ••.•.•.••••...•
Bottom profile of the Susitna River at the
stationary and mobile outmigrant trap
sampling points at Flathorn Station ••.•.•••....•••.•..
Map showing the location of the fyke net weir
on the Deshka River,1984 ••.••...•••.•.•••••.••••••..•
Map showing the reach where juveni 1e salmon
mark-recapture sites are located (RM 122.2 to
144.8 and Indian River)and the locations of
the Talkeetna stationary outmigrant traps (RM
103.0)and the Talkeetna River sampling site
(TRM 1.0 ),1984 ....-.. . . . . . •. . . •. •. . . . . . . . . . . ...e ••••S G
Branding locations and sample brands used for
cold branding chinook and coho salmon
j uven i 1es,1984 e •"••••••••IIlI ••••••••
Chinook salmon (age 0+)average catch per
minnow trap by sampl i ng peri od and survey
section in Indian River,1984 •••.•••••.•••.•••••••••••
Chinook salmon (age 0+)smoothed daily catch
per unit effort and adjusted cumulative catch
recorded at the Talkeetna stationary
outmigrant traps,May 14 through October 6,
1984.. .ill S •••••"DOD ••••••e •
Chinook salmon (age 0+)smoothed daily catch
per unit effort and adjusted cumulative catch
recorded at the Flathorn stationary
outmi grant trap,May 20 through October 1,
1984 0 8 Q •'II •••••••••'"••••10 e •••••
vi
5
6
7
8
15
18
20
21
-
-
-
_.
-
LIST OF FIGURES (Continued)
-
-
-
-
Figure
11
12
13
14
15
16
17
18
19
20
Title
Chinook salmon (age 0+)daily catch per unit
effort recorded at the Fl athorn mobil e
outmi grant trap,July 12 through August 30,
1984 .
Chinook salmon (age 0+)percent of total
catch by sampling point recorded at the
Flathorn mobile outmigrant trap,1984 ••••.•.••••.••.••
Chinook salmon (age 0+)catch per unit effort
by sampling period recorded at JAHS sites in
the lower reach of the Susitna River,1984 •.••••••••.•
Chinook salmon (age 1+)smoothed daily catch
per unit effort and adjusted cumulative catch
recorded at the Taol keetna stationary
outmigrant traps,May 14 through October 6,
1984 ..''..
Chinook salmon (age 1+)smoothed daily catch
per unit effort and adjusted cumulative catch
recorded at the Fl athorn stationary
outmi grant trap,May 20 through October 1,
1984 '' .
Chinook salmon (age 0+)mean length and range
of lengths by sampling period for fish
collected in the lower and middle reach of
the Susitn-a River,1984 ·.".
Wei9ht/length relationship for juvenile
chi nook sa 1mon co 11 ected at the Ta 1keetna
stationary outmigrant traps,1984 .•••••••...•••••••...
Catch,estimated population size,and main-
stem discharge level at Moose Slough,August
8 -August 12,1984 .
Catch,estimated population size,and main-
stem discharge level at Lower Side Channel
llA,July 29 -August 2,1984 .•.•••••••......••.•...••
Coho salmon (age 0+)average catch per minnow
trap by sampling period and survey section in
Indian River,1984 •••..........•.••.•.•....•..•.•.•...
vii
22
23
24
25
26
28
29
33
34
36
LIST OF FIGURES (Continued)'-
Figure
21
22
23
24
25
Title
Coho salmon (age 0+)smoothed daily catch per
unit effort and adjusted cumulative catch
recorded at the Talkeetna stationary
outmi grant traps,May 14 through October 6,
1984 eo C III 0 ..............• '.
Coho salmo~(age 0+)smoothed daily catch per
unit effort and adjusted cumulative catch
recorded at the Fl athorn stationary
outmi grant trap,May 20 through October 1,
1984 0 ••••••••••••••••••••••
Coho salmon juvenile catch per unit effort by
sampling period recorded at JAHS sites in the
lower reach of the Susitna River,1984 •.•••••••.•..•.•
Coho salmon (age 1+and older)smoothed daily
catch per unit effort and adjusted cumulative
catch recorded at the Talkeetna stationary
outmi grant traps,May 14 through October 6,
1984 II
Coho salmon (age 1+and older)smoothed daily
catch per unit effdrt and adjusted cumulative
catch recorded at the Flathorn stationary
outmi grant trap,May 20 through October 1,
1984 ~II-C ..
37
38
40
41
42
-
-
26 Coho salmon (age 0+)mean length and range of
lengths by sampling period for fish collected
in the lower and middle reach of the Susitna
R;ve r,1984 ..e •0 0 Gl •G 1&0 e 0 Q Cl g IP "..CD.II •.... ..43
27
28
29
Coho salmon (age 1+)mean length by month for
fish collected in the lower and middle reach
of the Susitna River,1984 •••••••.•••.••••.•••••.•.•••
Weight/length relationship for juvenile coho
salmon collected at the Talkeetna stationary
outmigrant traps,1984 ............•.................1Il.
Sockeye salmon (age 0+)smoothed daily catch
per unit effort and adjusted cumulative catch
recorded at the Talkeetna stationary
outmigrant traps,May 14 through October 6,
1984 ..
viii
45
46
48
~,
-.
LIST OF FIGURES (Continued)
Figure
30
31
Title Page
Sockeye salmon (age 0+)smoothed daily catch
per unit effort and adjusted cumulative catch
recorded at the Flathorn stationary
outmigrant trap,May 20 through October 1,
1984 -0 49
Sockeye salmon (age 0+)daily catch per unit
effort recorded at the Flathorn mobile
outmi grant trap,July 12 through August 31,
1984 ..,50
32
33
34
Sockeye salmon (age 0+)percent of the total
catch by sampling point recorded at the
Flathorn mobile outmigrant trap,1984 .•.•••.•....•.••.
Sockeye salmon juvenile catch per unit effort
by sampling period recorded at JAHS sites in
the lower reach of the Susitna River,1984 .•...•..•...
Sockeye salmon (age 1+)smoothed daily catch
per unit effort and adjusted cumulative catch
recorded at the Fl athorn and Talkeetna
stationary outmi grant traps,May 14 through
October 6,1984.e_••••••••••'.
51
52
53
35 Sockeye salmon (age 0+)mean length and range
of lengths by sampling period for fish
collected in the lower and middle reach of
the Susitna River~1984 -..55
....,
36
37
38
39
Weight/length relationship for juvenile
sockeye sa lmon co 11 ected at the Ta 1keetna
stationary outmigrant traps,1984 •••••.••.•.•••••.••..
Length of time between mark and recapture of
coded wire tagged sockeye salmon juveniles in
the middle reach of the Susitna River,1984 .••••••.•••
Chum salmon fry smoothed daily catch per unit
effort and adjusted cumulative catch recorded
at the Talkeetna stationary outmigrant traps,
May 14 through October 6,1984 •••.•.•••..•.•••••..••••
Chum salmon fry smoothed daily catch per unit
effort and adjusted cumulative catch recorded
at the Flathorn stationary outmigrant trap,
May 20 through October 1,1984 .•••.••.••••••••.•••••.•
ix
56
58
61
62
LIST OF FIGURES (Continued)
66
Figure
40
41
Title Page
Chum salmon fry catch per unit effort by
sampling period recorded at JAHS sites in the
lower reach of the.Susitna River,1984................63
Length of time between the mark and recapture
of coded wire tagged chum salmon juveniles in
the middle reach of the Susitna River,1984 •..•.•.•.••
42
43
44
45
46
47
48
49
50
Pink salmon fry smoothed daily catch per unit
effort and adjusted cumulative catch recorded
at the Talkeetna stationary outmigrant traps,
May 14 through October 6,1984 ••••••••••••••••••••••.•
Pink salmon fry smoothed daily catch per unit
effort and adjusted cumulative catch recorded
at the Flathorn stationary outmigrant traps,
May 20 through October 1,1984 ••••••••.•••••••••••••••
Mainstem discharge,water temperature,and
turbidity in the middle reach of the Susitna
River,1984 CG.
Mainstem discharge in the lower reach of the
Susitna River measured at the USGS gaging
station at Susitna Station,1984 •••..••••.•••••••••••.
Chinook salmon (age 0+)adjusted cumulative
catch recorded at the Talkeetna stati onary
outmigrant traps,1983 and 1984 ••••••••••.•••.•••.•••.
Chinook salmon (age 1+)adjusted cumulative
catch recorded at the Talkeetna stati onary
outmigrant traps,1983 and 1984 ••.•••.•.•••••••..•••..
Chinook salmon (age 0+)mean length and range
of mean 1engths by sampl ing period recorded
at the Talkeetna stationary outmigrant traps
during 1982,1983,and 1984 ••••••••.•.••••••••...•.•••
Chinook salmon adjusted cumulative catch and
biomass by age class recorded at Talkeetna
and Flathorn stations,1984 ••••.••••.••.•..•...•••.•.•
Coho salmon (age 0+)adjusted cumulative
catch recorded at the Tal keetna stationary
outmigrant traps,1983 and 1984 ••........•.•••••.•.•••
x
68
69
72
73
75
76
80
81
82
-
-
LIST OF FIGURES (Continued)
Figure Title
-
-
-
51
52
53
54
55
56
57
58
59
60
61
Coho salmon (age 1+)adjusted cumulative
catch recorded at the Talkeetna stati onary
outmigrant traps,1983 and 1984 ••....•••.•...•••.••••.
Coho salmon (age 0+)mean length and range of
mean 1engths by sampl i ng period recorded at
the Talkeetna stationary outmigrant traps
during 1982,1983,and 1984 ••••••••.••••...••••••.••••
Coho salmon (age 1+)mean length and rang·e of
mean lengths by sampling period recorded at
the Talkeetna stationary outmigrant traps
during 1982, 1983,and 1984 .•..••••.•••••••••.••••••••
Coho salmon adjusted cumul ative catch and
bi amass by age class recorded at Talkeetna
and Fl athorn Stati ons,1984 ••••••••••••••••••.•.•••.•.
Sockeye salmon (age 0+)adjusted cumulative
catch recorded at the Talkeetna stationary
outmigrant traps,1983 and 1984 .•••••••••••••.•••••.••
Mean length of coded wire tagged sockeye
salmon fry at recovery sites in,the middle
reach of the Susitna River by week,1984 .••••.•••••.••
Sockeye salmon (age 0+)mean length and range
of mean lengths by sampling period recorded
at the Talkeetna stationary outmigrant traps
during 1982,1983,and 1984 •••••••••••••.•.•••.••••.••
Sockeye salmon adjusted cumulative catch and
biomass by age cl ass recorded at Talkeetna
and Flathorn Stations,1984 •..•••••••••.••••••.•.•.•••
Chum salmon fry adjusted cumulative catch
recorded at the Talkeetna stationary
outmigrant traps,1983 and 1984 ••••.••..•.••••..••..•.
Mean length of coded wire tagged chum salmon
fry at recovery sites in the middle reach of
the Susitna River by 5 day period,1984 •.••••••.•.••.•
Pink salmon fry adjusted cumulative catch
recorded at the Talkeetna stationary
outmigrant traps,1983 and 1984 ••.••...•....•.••••....
xi
84
85
86
87
89
93
94
95
97
98
100
LIST OF TABLES -
Table
1
Title Page
The number of chinook salmon fry marked and
recovered in Indian River by sampling period,
1984 ~e ••••••••IZI •e •00 3a -
2
3
4
5
6
7
8
Chinook salmon fry population estimates by
site for middle Susitna River sloughs and
side channels and for Indian River,1984 •..•.•••••••.•
Coded wire tag release data for sockeye
salmon fry on the Susitna River by tagging
site and release date,1984 .•••••••••.•.••••••••••.••.
Recoveries of coded wire tagged sockeye
salmon fry at mainstem river sites between
Talkeetna and Devil Canyon,1984 ••••••••••••••••••••••
Coded wi re tag release data for chum salmon
fry on the Susitna River by tagging site and
re 1ea se da te·".1984 •••••••••••••••••Go •Co •••••••••••••••III
Summary statistics for juvenile salmon catch
per hour by species and age class recorded at
the Talkeetna Station outmigrant traps,May
14 through October 6,1984 •••••••••••••••••.••••••••.•
Summary statistics for habitat variables
recorded on the Susitna River between the
Chul i tna Ri ver confl uence and Devi 1 Canyon,
May 14 through October 6,1984 ••••••••.•••••••••••••••
Summary statistics for juvenile salmon catch
per hour by species and age class recorded at
the Flathorn Station outmigrant traps,May 20
through October 1,1984 .
xi i
32
57
59
65
70
70
71
-
~I
LIST OF APPENDIX TABLES
Appendix Table Title
A-1 Wei r catches of juveni 1e chi nook and coho
salmon on the Deshka River,May 10 through
September 19,1984....................................A-1
A-2 Results of incidental minnow trapping in the
Deshka River,1984....................................A-2
A-3 Mean 1ength and range of lengths for age 0+
chinook salmon by sampling period in the
lower reach of the Susitna River,1984................A-3
A-4 Mean length and range of lengths for age 0+
chinook salmon by sampling period in the
Tal keetna River and the middle reach of the
Susitna River,1984 ...~...............................A-4
Mean length and range of lengths for age 0+
coho salmon by sampling period in the lower
reach of the Susitna River,1984......................A-6
Mean length and range of lengths for age 1+
chinook salmon in the Susitna River,1984.............A-5
Mean length and range of lengths,for age 0+
coho salmon by sampling period in the middle
reach of the Susitna River,1984......................A-7
A-5
.-
A-6
r-
A-7
A-8 Mean 1ength and range of 1engths for age 1+
coho salmon by sampling period in the lower
reach of the Susitna River,1984......................A-8
-
....
......
A-9
A-I0
Mean length and range of lengths for age 1+
coho salmon by sampling period in the middle
reach of the Susitna River,1984......................A-9
Mean length and range of lengths for age 2+
coho salmon by sampling period in the Susitna
Ri ver between Cook In 1et and Dev i 1 Ca nyon ,
1984 •.•.•...•••...••...•••••.•.••••...•.••..•..•••.••.A-I0
A-ll Daily catches of outmigrant chum and sockeye
salmon fry in a fyke net located at the mouth
of Slough 21,May 23 to June 12,1984.................A-ll
A-12 Mean 1ength and range of 1engths for age 0+
sockeye salmon by sampling period in the
Susitna River between Cook Inl et and Devi 1
Canyon,1984...........................................A-12
xiii
LIST OF APPENDIX TABLES (Continued)
Appendix Table Title
A-13
A-14
Mean 1ength and range of 1engths for age 1+
sockeye salmon by sampling period in the
Susitna River between Cook Inl et and Devi 1
Canyon,1984 -o ••••••••o8 0Iloo ••e.-o.A-13
Mean 1ength and range of 1engths for chum
salmon fry by sampling period in the Susitna
River between Cook Inlet and Devil Canyon,
1984.........•••••• ••.•.•••••••.•..•.•••.•.•••••••.•••A-14
.....
B-1 Data collected on the coded wire tag,mark-
recapture experiment for sockeye salmon fry
to prOVide a Schaefer population estimate.............8-2
B-2 Computati on of the sockeye sa 1man for
outmigrant population from the data presented
in Appendix Table B-l ••.•••••e •••••••~•••••e.e •••ee.8e B-3
B-3 Data call ected on the coded wi re tag,mark-
recapture experiment for chum salmon fry to
provide a Schaefer population estimate................8-4
8-4 Computation of the chum salmon for outmigrant
population from the data presented in
Appendix Table B-3 ••••••••~•.•••e.~~••••e$••o~o ••••o..B-5
xiv
-
~I
-
"""'
1.0 INTRODUCTION
Studies of the migration and growth of juvenile salmon in the mainstem
Susitna River are a part of the ongoing investigations being conducted
by the Resident and Juvenile Anadromous Fish Project (RJ)of the Susitna
River Aquatic Studies Program.The scope of these studies has been to
describe the periods of freshwater residence,growth,and timing of
outmigration for juvenile salmon in the Susitna River and to provide
population estimates for the reach of river between the Chulitna Riv~r
confluence and Devil Canyon.This report presents the results of
juvenile salmon outmigration stl,ldies conducted on the Susitna River
between Cook Inlet and Devil Canyon during the 1984 open-water season.
Five Pacific salmon species are addressed in this report:chinook
(Oncorhynchus tshawytscha),coho (Q.kisutch),sockeye (0.nerka),chum
(Q.keta),and pink (0.gorbuscha).
Investigations of the distribution,abundance,and migration of juvenile
salmon during 1982 and 1983 were focused primarily on the Susitna River
reach above the Chulitna River confluence (ADF&G 1983a;Schmidt et ale
1984).These studies included the operation of stationary outmigrant
traps at Talkeetna Station,river mile (RM)103.0,during 1982 and o1983
and a mark-recapture program for post-emergent chum and sockeye sal mon
fry using half-length coded wire tags in 1983 (Roth et ale 1984).These
techniques have provided valuable information on the success of previous
spawning runs,the effect of di scharge on redi stri bution of young-of-
the-year salmon juveniles,and the population size and egg-to-outmigrant
fry survival rates for chum and sockeye salmon fry.
During the 1984 open-water season,additional tasks were added to
further describe juvenile salmon growth,migration timing,and response
to changing habitat conditions.The study area was expanded to include
the entire river between Cook Inlet and Devil Canyon.New tasks begun
in 1984 were the addition of stationary and mobile outmigrant traps at
Flathorn Station (RM 22.4),intermittent trapping of migrating chinook
sa 1mon juven i1es in the Deshka and Talkeetna ri vers,and mark-recapture
by cold branding of juvenile chinook and coho salmon in the Curry
Station to Devil Canyon reach.
Investigations of the migration and growth of juvenile salmon in the
Susitna River above the oChul itna River confl uence duri ng 1982 and 1983
indicated extensive migration of pre-smolt juveniles of all species to
areas below this reach.This migration of pre-smolt chinook salmon was
also observed in the Deshka River in 1980 (Delaney et ale 1981).If this
movement is common in the major tributaries entering the Susitna River,
extensive rearing and growth of juveni1e salmon,parti cul arly chinook,
may occur in habitats associated with the mainstem river.Small habitat
changes in the reach of river below Talkeetna could impact large numbers
of rearing salmon.
The combined studies of juvenile salmon growth and migration conducted
during the 1984 open-water season were developed to provide data to meet
the following objectives:
1
o Estimate the timing,relative abundance,and size of out-
migrating juvenile salmon in the Susitna River above the
Chulitna River confluence.
o Estimate the population size of outmigrating chum and sockeye
salmon fry and egg-to-outmigrant fry survival in this reach of
.river.
o Estimate the timing and size of outmigrating chum salmon from
the Talkeetna River.
o Estimate the timing and rate of movement of juvenile chinook
and coho salmon out of Indian River and their residence time
at selected macrohabitats associated with the mainstem Susitna
River.
-
o
o
o
Estimate the timing and rate of outmigration of chinook salmon
juven'iles from the Deshka River into/the mainstem Susitna.
!
Estimate the timing and rate·of \outmigration of juvenile
salmon from the Susitna River into Coo~Inlet.
Estimate the rate of growth of juvenile chum and chinook
salmon from the time they enter the lower river (below the
Chulitna River confluence)until they enter the marine
environment.
o Estimate the relationship of mainstem Susitna discharge and
other environmental variables to juvenile salmon outmigration.
Sampling of chum salmon fry in the Talkeetna River was hindered by
equipment failure;insufficient data were collected for this species,
although some growth and relative abundance data for chinook salmon were
collected.
Although initially designed as a survey of Portage Creek using a sta-
tionary outmigrant trap,the cold branding study was relocated to Indian
River with minnow traps serving as the primary collection technique.
The design of the original collection equipment did not lend itself well
to the continually fl uctuating hydraul ic conditions present at Portage
Creek.The low numbers of juvenile salmon observed in Portage Creek
after June 15,combined with the comparative logtstical inaccessibility
of this stream,made Indian River a better choice.
The data presented in this report provide information that can be used
to determine the size of the present fishery resource,potential changes
caused by the proposed hydroelectric development,and mitigation
requirements necessary to compensate for any reductions of the juvenile
salmon populations in the Susitna River.
2
.-
2.0 METHODS
2.1 Study Locations
Studies on the migration and growth of juvenile salmon in the mainstem
Susitna River were conducted at survey sites from Flathorn Station (RM
22.4)upstream to Slough 22 (RM 144.3)during the 1984 open-water season
(Fig.1).
2.1.1 Flathorn Station
A stationary outmi grant trap was operated on the west bank of the
Susitna River at Flathorn Station (RM 22.4)and a mobile outmigrant trap
was used to sample a total of ten points along transects spanning three
channels of the mainstem river at this station (Fig.2).Five sampling
points were located in the west channel (RM 22.8),one in the middle
channel (RM 22.8),and four in the east channel (RM 23.9).A bottom
profile of the Susitna River at these sampling points is provided in
Fig.3.
2.1.2 Deshka River
An outmigrant fyke net weir was operated in the Deshka River (RM 40.6)
between tributary river mile (TRM)2.5 and TRM 5.0 to estimate the
timing and rate of outmigration for juvenile chinook salmon (Fig.4).
2.1.3 Talkeetna River
A beach seine sampl ing site for outmigrants was located in the north
channel of the Talkeetna River (RM 97.5)approximately one mile upstream
from the river's mouth (Fig.5).
2.1.4 Talkeetna Station
Two stationary outmigrant traps were deployed on the mainstem Susitna
River above the Chulitna River confluence at Talkeetna Station (RM
103.0)at the same locations used in 1983.One trap was set off the
east bank (Trap 1)and the other off the west bank (Trap 2)of the river
(Fig.5).
2.1.5 Coded wire tagging
Coded wire tagging sites were selected from those locations above the
Chul itna River confluence where hi gh density spawning by adults was
recorded (Barrett et ale 1984),and from surveys of the availability of
sufficient numbers of post-emergent chum and sockeye salmon fry for
collection and tagging (Fig.5).Specific coded wire tagging sites (Fig.
6)were:
3
TALKEETNA~~~la
STATION
~
0 10 20 30
I I r I
MILES
(Appro •.Scale)
/
~;
-
-
Figure 1.Map of juvenile salmon outmigration study field stations in
the Susitna River basin,1984.
4 -
SAMPLING POINTS
•Stationary Outmigrant Trap
•Mobile Outmigrant Trap
4
,'.;.
~...
,f I">•
"
~~
'".?:V,
~
CI)
......
"f
~
~".~
~."
".'
'=...!~
~
~
J
"
.'
~.
FlATHORN
STATION
.-
-
Figure 2.Map of the stationary outmigrant trap and the mObile outmigrant
trap sampling points on the Susitna River at Flathorn Station t
1984.
5
EAST CHANNEL IHM23.91
~IICW
0 ••••0
e ••-iE
Go '0l!f l'
1.l'I 11.110 '"~100 .1'710 .71 'GOO 111.
HORIZONTAL DISTANCE ern
,~
-
IIDDLE CHANNEL IBI 2%.81
o
•
SAM PLINK POINTS
x STATIONARY OUTMlGAANT TRAP
•MOBLE OUTMIGRANT TRAP
10 10
o 111 110 17.
HORIZONTAl.DISTANCE am
WEST CHANNEL IBM 22.41·
-
o
•
Raw••• • •
f l'-%
l-
Go
~,.
I.+---r-......,r---~-....,..---r--~-....,..---r--.,..---r----r--+-tl
o 11.110 .7.too II'710 11.'000 "1'',10 '11'1100
HOAIZONTAL DI8TANeE (FT)
Figure 3.Bottom profile of the Susitna River at the stationary and mobile
outmigrant trap sampling points at Flathorn Station.Measured
on August 23,1984 at a mainstem discharge of 114,000 cfs at the·
USGS gaging station at Susitna Station.
6
MILES
(ApprOll.Scale)fJ eRM45
Fyke Net------~n
Weir Site
(TRM 2.5)
r-
!
-
Figure 4.Map showing the location of the fyke net weir on the Deshka
River l 1984.
7
-
~
0 10
I I I
MILES
~Coded Wire Tagging a
Cold Branding Sites
"'0
.....0'"co ,.••,f
"-.-Talkeetna River Sampling Site
Talkeetna Station
Out migrant
Traps
Figure 5.,Map showing the reach where juvenile salmon mark-recapture sites are
located (RM 122.2 to 144.8 and Indian River)and the locations of the
Talkeetna stationary outmigrant traps (RM 103.0),and the Talkeetna _
River sampling site (TRM 1.0),1984.
8
•SAMPLING SITE
LOUGH 21
SIDE CHANNEL SLOUGH 21
SLOUGH 20
SLOUGH 19
SLOUGH II
UPPER SIDE CHANNEL II
SI DE CHANNEL 10_
SLOUGH 9
SLOUGH 8A
MOOSE SLOUGH
SLOUGH 88
R.
SLOUGH 22
INDIAN RIVER
SLOUGH 17
SLOUGH 16
SLOUGH 15
4t11 of
JulyCr.
__'"'==I-TALKEETNA STATION OUTMIGRANT TRAPS
.....
Figure 6.Map of coded wire tagging and cold branding sites in the
midole reach of the Susitna Rivers 1984.
-
9
CODED WIRE TAGGING
SITES RIVER MILE
Slough 8B 122.4
Slough 8A 125.3
Slough 9 129.2
Slough 11 135.3
Slough 15 137.3
Indian River 138.6
Slough 20 140.1
Slough 21 142.0
Slough 22 144.3
2.1.6 Cold branding
A col d brand mark-recapture study was conducted at the mouth and at
numerous side channels and side sloughs of Indian River (RM 138.6)which
were found to contain large concentrations of juvenile chinook and coho
salmon.Indian River was divided into three sections for this study.
Section I included the mouth upstream to TRM 0.5,Section II was the
portion of Indian River from TRM 0.5.to 7.5 and Section III was from TRM
7.5 upstream to TRM 12.3 (Fig.5).
Cold branding was also used to estimate the populations and study the
movements of juvenile salmon at the following study sites (Fig.6):
COLD BRANDING
SITES
Moose Slough
Side Channel 10
Lower Side Channel llA
Slough 16
Slough 17
Slough 19
Slough 20
Side Channel 21
Slough 22
10
RIVER MILE
123.2
133.8
135.9
137.7
138.9
139.7
140.1
141.1
144.3
-
--
2.2 Field Data Collection and Recording
2.2.1 Flathorn Station outmigrant traps
The stationary outmigrant trap on the west bank of the Susitna River at
Flathorn Station (RM 22.4)was operated from May 20 through October 1,
1984.A description of this outmigrant trap is provided in ADF&G
(1985).The trap was checked at least twice each day to remove the
captured fish and to clean the trap.
The mobile outmigrant trap.at Flathorn Station was.operated for 43
days during the period July 12 through September 13,1984.A description
of the trap and its operation is presented in ADF&G (1985).The trap
was fished for 20-minute periods at ten different transect points during
a fishing day.
Habitat and biological data recorded for each check of the stationary
outmigrant trap included fishing effort (hours),trap depth (feet),
distance from shore (feet),and catch by species and age class.Main-
stem stage was recorded once each day.The fi rst 25 fi sh of each
species and age class collected daily were measured for total length
(tip of snout to tip of tail)in millimeters (mm).
Biological and habitat data for the stationary trap were entered
directly into an Epson HX-20 microcomputer in the field.·Operational
procedures f.o..r the microcomputer and the associated data form program
are presented in ADF&G (1985).Computer entri es were made for each trap
check throughout the fi~ld~eason.Printouts and cassettes were
periodically transferred to Data Processing to be entered into a main-
frame computer for later data retrieval and analysis.
Transect number,fishing effort,total water column depth,set velocity,
and drift velocity (if the trap was not held stationary during the set)
were recorded for each individual transect point at which the mobile
outmigrant trap was fished.Total catch by species and age class was
also recorded,and total length measurements were taken for all captured
fish.Data were recorded on a field data form for later analysis.
2.2.2 Deshka River outmigrant weir
A weir was established on the Deshka River (RM 40.6)using a fyke net
(3/16 inch square mesh)to block a portion of the river.The fyke net is
described in ADF&G (1985).The weir was operated at varying tributary
miles (TRM 2.0 -5.0)periodically from May 10 through June 22.The
weir was moved to TRM 2.5 on July 11 and was fished periodically through
September 18.Minnow traps were fished intermittently from 1ate June
through mid October to supplement the weir data.
Fishitlg effort and total catch by species and age class were recorded
for the outmigrant weir and the minnow traps.A sample of each species
and age class captured were measured for total length and scale samples
were collected for age determination.
11
2.2.3 Talkeetna River beach seining
Beach seining (1/8 inch square mesh)was conducted one to two times each
week from June 5 through September 15.Sampling was conducted to obtain
a sufficient sample for comparative length and outmigration timing data.
An attempt was made to use a Fyke net weir in late May and June.This
did not work,so we changed to a beach seine.
Total catch by species and age class was recorded.All.captured fish
were measured for total length and released.
2.2.4 Talkeetna Station outmigrant traps
Two inclined plane outmigrant traps were operated continuously in the
mainstem Susitna River at Talkeetna Station (RM 103.0)from May 14
through October 6,1984 using the methods outl ined by Roth et al.
(1984).
Measurements of the following habitat parameters were recorded daily at
the outmigrant traps:air and surface water temperature (OC),turbidity
(NTU),water velocity (ft/sec),and mainstem stage data.The equipment
and methods used to collect the habitat data are given in ADF&G (1985).
Trap fishing depths and distances from shore were adjusted to maximize
catches while maintaining trap efficiency.All juvenile fish captured
were anesthetized using MS-222 (Tricaine methanesulfonate).Field
specimens were identified using the guidel ines set forth by McConnell
and Snyder (1972),Trautman (1973),and Morrow (1980).Juvenile chinook
and coho salmon collected at the traps were checked for a cold brand
mark and all recovered marks were recorded.Chum and sockeye salmon
juveniles with a clipped adipose fin were passed through a detector to
veri fy the presence of a coded wi re tag.A11 coded wi re tagged fi sh
recovered at the traps were preserved and tags were 1ater removed and
decoded using a reading jig and a binocular microscope.All other fish
recovered at the traps were held until anesthetic recovery was complete
and then released downstream of the traps.
Scales were collected from a sub-sample of fish captured for comparison
to length frequency data for final age class determination.Biological
and habitat data were entered directly into an Epson HX-20
mi crocomputer.
Length and weight relationship data were also collected from samples of
juvenile chinook,coho,and sockeye salmon collected in the outmigrant
traps at Talkeetna Station.Total length was recorded to the nearest
millimeter and live weights were determined to the nearest 0.1 gram.
2.2.5 Coded wire tagging
The coded wire tagging was conducted at Slough 11 (RM 135.3)from May 16
through June 20,1984.The fish were transported from the collection
areas to Slough 11 in an aerated tub,tagged,held for at least 24
hours,and then returned to the collection areas •.The fish were also
held overnight at the collection areas prior to release.
12
....,
-
-
-
-c
~
I
-
Beach seines were used to weir off the downstream end of the collection
area and were checked at least once each day to collect fish and remove
debris.Beach seining and dip netting supplemented the weir catches at
sites where wei ring alone did not provide enough fish for the tagging
operation.
The coded wire tagging equipment and implantation procedures are similar
to those outlined by Roth et a1.(1984)using the guidelines provided by
Koerner (1977)and Moberly et al.(1977).One,;"ha1f length binary coded
wire tags measuring 0.02 inches (0.533 mm)in length and 0.01 inches
(0.254 mm)in diameter were used in the study.Separate head molds were
required for each species and length class of fish.Fifty fish of each
group were measured to determine mean length and the proper head molds
for the tagging procedure.The adipose fin was clipped from each fish
prior to tagging to provide a visual indicator of the presence of a
coded wire tag.At the end of each tagging day,a subsamp1e of 100
tagged fish were anesthetized and passed through the quality control
device to determine the tag retention rate.Mortalities were recorded
the following day and again just prior to release.A single tag code
was used for each species tagged and for each collection site.Six
distinct tag codes were used for juvenile sockeye salmon and fourteen
distinct tag codes were used for juvenile chum salmon.
Coded wire tagging data recorded at each site included date tagged,tag
code,speci es,number of fi sh tagged,percent tag retent i on,mortal i ty,
and date and time of release.Total numbers of fish tagged by species,
collection site,and release date as well as final tag retention and
mortality were tabulated for each tag code.
2.2.6 Cold branding
Mark-recapture studies of chinook and coho salmon populations were
conducted from July through mid October.Cold branding was used as a
marking technique because it is less expensive than coded wire tagging.
Cold branding was not used on chum and sockeye because it has not been
proven effective on these fish at the post-emergent stage.Sites in
Indian River were sampled twice a month and fish were captured,branded,
and released continually throughout the field season.Sampling in the
sloughs and side channel s of the Susitna River was conducted for five
consecutive days and captured fish were either branded and released the
same day or held until the end of the five day period before release.
Minnow traps,beach seines,and dip nets were used to capture fish which
were then transported from the areas of collection to the Gold Creek
field camp for cold brand marking.Cold branded fish from all sites
except Indian River were held for 24 hours to determine marking
mortality before being released at the area of collection.Fish col-
lected in Indian River were marked,held for 24 hours,and then released
at a side slough at TRM 7.2.
The brands consisted of single brass letters or symbols measuring
approximately three millimeters in height which were soldered onto
threaded brass caps.Liquid nitrogen was used as the cooling agent and
13
branding procedures were similar to those outlined by Raleigh et al.
(1973).The cold branding equipment is described in ADF&G (1985).
Juvenile chinook and coho salmon were marked with a distinctive brand to
signify the collection site and date of their capture.Fish were marked
on one side of th~body at one of three target branding areas (Fig.7),
and a branding time of two seconds was used.-
Date,collection site,gear type,fishing effort,species,number of
fish captured,and brand symbol were recorded for each site.The number .~
of recaptures by species and the symbols for previously marked fish were
also recorded.Total length was measured for 50 fish of each species
during each sampling trip.
2.3 Data Analysis
2.3.1 Juvenile salmon catch per unit effort
The catch per unit effort (CPUE)data collected for juvenile salmon at
the stationary outmigrant traps are presented as the average catch per
hour for each calendar day of sampling effort.The catch was expanded
to 24-hour intervals by dividing the number of hours fished on a given
day into 24 and then multiplying this ratio by the catch for each
species and age class.
The catch rates plotted for each species and age class of juvenile
salmon collected at the stationary traps were smoothed using the von
Hann linear filter (Dixon et al.1981).The equation is:
Z(t)=iY(t_1)+iY(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 were adjusted for days not fished by tabu-
lating the mean of the total catches recorded for the three days
preceding and the three days following an unsampled day.
Length frequency distribution and scale analysis data were used to
determine the age class composition of chinook,coho,and sockeye salmon
juveniles.
A regression was done on the natural logarithm of weight versus the
natural logarithm of length for chinook,coho,and sockeye salmon.The
regression equations were used to provide estimates of the total biomass
passing the Talkeetna and Flathorn station outmigrant traps by sampling
period through the season.
14
Six·8 randing Locations
Left Side Right Side
a)anterior to dorsal fin
b)beneath dorsal fin
c)posterior to dorsal fin
,..,.,
I
Sample Cold-Brands
U ::::>n c
,...,
3EI.LI ITt
T I-.L ~
A L 1....J r
I -L--1 S
-Figure 7.,_Branding locations and sample brands used for cold branding chinook
and coho salmon juveniles,1984.
15
2.3.2 Population and survival estimates
Potential egg deposition for chum and sockeye salmon was calculated by
multiplying the average fecundity for each species by the estimated
number of female spawners that passed Curry Station in 1983 (Barrett et
al.1984).The chum,sockeye,and chinook salmon adult population
estimates were reduced by 40%,39%,and 7%respectively,to account for
milling fish which eventually spawned below the Chulitna River
confluence (Barrett 1984;Barrett et al.1984).The following formula
was used to determine egg deposition:
Total potential egg deposition =(E)x (l-M)x (p)x (F)
where:
E =Adult population estimate at Curry Station
M=Proportion milling
P =Proportion females
F =Average fecundity
Population estimates for chum and sockeye outmigrants were calculated by
the Schaefer (1951)method (Appendix B).Estimates of survival for both
species were determined by dividing the population estimates by the
calculated potential egg deposition for each species.Only valid tagged
fish were used in the calculations.The total number of valid tagged
fish was determined by subtracting the mortalities for each day of
tagging from the total number of fish tagged and then multiplying this
by the tag retention rate.Total tag recoveries at the Talkeetna
Station outmigrant traps include only those fish with a coded wire tag.
Fish having a clipped adipose fin but no tag were not considered in the
population estimates.
Population estimates for chinook salmon were calculated from the data
collected during the cold branding study by using the Petersen,
Schaefer,or Jolly-Seber methods (Ricker 1975).The Schaefer and
Jolly-Seber methods were used at sites where conditions allowed five
consecutive days of sampling.The Peterson method was used when there
was one marking period and one recapture period.Confidence limits for
the Jolly-Seber estimate of population size were developed using the
method of Manly (1984).The Jolly-Seber model was run on a commercial
spreadsheet program for microcomputers.The potential egg deposition
for chinook salmon in Indian River was determined-using the technique
1i sted above except that the estimate was reduced to represent the
percentage of chinook (determined from peak spawning counts)which
spawned in Indian River.Fecundities used were those measured by Healy
and Heard (1984)for Kenai River chinook salmon.
2.3.3 Time series analysis
The 1983 and 1984 discharge,turbidity,and age 0+chinook and sockeye
salmon outmigration time series are analyzed in Appendix C.
16
-
.....
-.
-
,....
""",
-
.-
.....
3.0 RESULTS
The results of the juvenile salmon outmigration studies are presented by
species.The catch per unit effort (CPUE)data are presented as a
percentage of the highest CPUE (after smoothing)recorded at the sta-
tionary traps during 1984.The cumulative catch data are presented as a
percentage of the total adjusted ·cumu1ative catch after app1 ication of
the smoothing functions.Juvenile salmon length data collected at
Flathorn Station are from both the stationary and mobile traps and the
length information presented for Talkeetna Station is from both
stationary traps located at this site.
3.1 Chinook Salmon
3.1.1 Catch per unit effort
3.1. 1.1 Age 0+
Chinook salmon fry collected incidentally during the coded wire tagging
study in May and June were observed to be most abundant at Slough 22 and
Indian River.
The cold branding study captured 26,823 chinook salmon fry in Indian
River from July 1 through October 15.Fifty-eight percent of this catch
was recorded near the mouth of the river (section I),30%in the lower
portion (section II)and 12%in the upper portion (section III).Beach
seining of sections II and III during July captured 3,280 chinook salmon
fry;66%in section III and 34%in section II.Minnow trapping begun in
Indian River in late July collected a total of 23,543 chinook fry during
947 minnow trap days (defined as one trap day for each overnight minnow
trap set)for a season average of 24.9 fish per trap day.
Catch rates in Indian River (Fig.8)were generally highest in section
II except during late August when high and turbid water conditions
reduced trapping effectiveness.The CPUE for chinook fry in Indian
River for all sections combined was highest during late July (average of
36 fish per trap day)and steadily declined through the season to a low
of 15 fish per trap day in early October .
A total of 11,875 chinook salmon fry were captured in sloughs and side
channels in the middle reach of the Susitna River during the cold
branding ~tudy from July 1 through October 15.Sloughs accounted for
84%of the catch while the remaining 16%were collected in side
channels.Beach seining during July and August collected 39%of the
total catch at these sites while minnow trapping begun in early
September captured 61%of the chinook fry.
The 7,291 chinook salmon fry captured by minnow trapping at slough and
side channel sites in the middle river were collected during 378 minnow
trap days for an average of 19 fish per trap day.Mean CPUE by study
site ranged from a hi gh of 48 fi sh per trap day at Slough 22 during
early October to a low of 3 fish per trap day at Side Channel 21 in late
September.
17
INDIA~l RIVER 1 984
-
-
E OCTLSEPESEP
SAlVIPLlNG PERIOD
CHI~JOOK 0+
45
40
35
~a
a.30
«a::
f-
a::25
wa.
:I:20(j
~
(j
'15
10
5
L JUL E AUG L AUG
Figure 8.Chinook salmon (age 0+)average catch per minnow trap by
sampling period and survey section in Indian River,1984.
18
-
-
A total of 14,110 chinook salmon fry were collected at the Talkeetna
Station outmigrant traps.Peak catches were recorded from late June
through early August and the highest catch rate of 17.3 chinook fry per
hour was recorded on July 26 (Fig.9).Fifty percent of the catch was
recorded by July 20.Catches decreased after early August and the last
capture of chinook fry at this site was recorded on September 29.
A total of 2,118 ch"jnook salmon fry were captured in the stationary
outmigrant trap at Flathorn Station.Catch rates were greatest between
late June and late August (Fig.10).The chinook fry catch rate at this
site peaked at 7.8 fish per hour on July 23,50%of the captures were
recorded by July 13,and the last capture was recorded on September 30.
The highest catch rate of the Flathorn Station mobile trap was 16.2 fish
per hour,recorded on July 23 (Fig.11).Of the 189 chinook fry
coll ected in the mobile trap duri ng 1984,60%were captured at bank
transect sampling points and the remaining captures occurred at center
channel sampling sites (Fig.12).
The Deshka River weir captured 1,808 chinook salmon during 1984 (Appen-
dix Table A-I).Eighty-eight percent of the captures were recorded
during July and the peak catch rate of 21.2 "fi sh per hour was recorded
on July 25.Minnow trap catches at this site were highes"t during late
June at 8.7 fish per trap (Appendix Table A-2).
A total of 1,356 chinook salmon fry were collected in the lower reach of
the Susitna River by the Juvenile Aquatic Habitat Studies (JAHS)surveys
from June through early October (see Part 2 of this report).Catch
rates for all sites combined peaked in August and then decreased through
early October (Fig.13).
3.1.1.2 Age 1+
Age 1+chinook salmon were captured incidentally during the coded wire
tagging study in May and June and were most abundant at Indian River and
Slough 11.No age 1+chinook were captured during the cold branding
study begun in July,as most of these fish had outmigrated by that time.
Peak catch rates of the 1,321 age 1+chinook captured at the Talkeetna
Station outmi grant traps we·re recorded during the deployment of the
traps in mid May and again in mid and late June (Fig.14).Fifty
percent of the season catches occurred by June 23.The highest catch
rate for this age class was 3.6 fish per hour recorded on May 15 and the
last age 1+chinook was captured in the traps on August 7.
Catch rates for the 346 age 1+chi nook salmon captured at Fl athorn
Station were highest during early June (Fig.15).The highest CPUE of
6.4 fish per hour was recorded on June 14 (50%of the season total by
this date)and the last age 1+chinook was collected at this site on
August 23.
Nine age 1+chinook salmon were collected in the Deshka River during
weir and minnow trap sampling,with the last capture recorded on October
10.
19
TALKEETNA BOTH TRAPS
~
Z
l.U
()a::w
Q.
100 ~
90
80 170
60
50
40
30
20
10
o~IMAY
\
I
%CUMULATIVE
%OF HIGHEST CPUE
~
AUG r SEP
~14.8
[
It ~
t@
L ~
I..J..O.O
Figure 9.Chinook salmon (age 0+)smoothed daily catch per unit effort
and adjusted cumulative catch recorded at the Talkeetna
stationary outmigrant traps,May 14 through October 6,1984.
20
FLATH OR !'J STATION CHINOOK 0+
a::
;:)
o
:::I:
.......
:::I:
U
r-«
u
-%CUMU LATIVE
-%OF HIGHEST CPUE
...,------·--------r-------~~----__r_5.1
A-+-----f--"':::::-...,.---+------+--------:::~--::=...-----""=t_o.o
JUL SEP'
100
gO
80
70
50I-
Z
W
(J 50ccwa.
40
30
20
10
0 MAY
Figure 10.Chinook salmon (age 0+)smoothed daily catch per unit effort
and adjusted cumulative catch recorded at the Flathorn
stationary outmigrant trap,May 20 through October 1,1984.
21
-
AGE 0+CH I hj()(]~<CF'lJE -,
100 -a'-Notsa~~~:--l90 ~b -No Fish Captured
w 80
::}
11.
()70
I-
tl)
W
I 60
I~.,1\;,.
1:
lJ..50
0
W
G 4-0 ,-~
Z
lJJ 30u
I:r
\.LJ .-ll.20
10 -i
b~~.~~1~~"
0 aa aa
12 15 20 JULY 30 5 10 AUGUST 20 --'C.":;·0L,l
DATE
Figure 11.Chinook salmon (age 0+)daily catch per unit effort recorded
at the Flathorn mobile outmigrant trap,July 12 through
August 30,1984.
22
AG"E 0+CHI!'JOOK SEASON CATCH-21
20
19
"18
17
:I:16
()15I-«14(.)
...J 13
~12
0 11I-
1.1...10
0 9I-
Z 8w
0 7
0::6wc..5
4
3
2
1
0
"'1 2 3 4 "'5 "'6 "'7 8 9 "'10
Figure 12.
TRANSECT POINT NUMBER
*BANK TRANSECTS
Chinook salmon (age 0+)percent of total catch by sampling
point recorded at the Flathorn mobile outmigrant trap,1984.
23
CHINOOK'CPLIE 1"984
100 -,.....-----------...,..,.......,......,....,--------------,
90
W
:::::l
0-o
t;j
W
:I:
C)
:I:
u..a
~
80
70
60
50
40 r-,,......,......,.,
30
20
10
E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAMPLI NG PERIOD -
-
Figure 13.Chinook salmon (age 0+)catch per unit effort by sampling period
recorded at JAHS sites in the lower reach of the Susitna River,
1984.
24
-
-
TALKEETNA 80TH TRAPS CHII\-IOOK ~l +
%CUMU LAT1\1E
%OF HIGHEST CPUE
v\_---+_------11.o
JUL -4 AUG SEP IJUN
20
60
50
30
70
40
80
90
1:X _
,MAY
l-
Z
LlJo
0::
W
0..
Figure 14.Chinook salmon (age 1+)smoothed daily catch per unit effort
and adjusted cumulative catch recorded at the Talkeetna
stationary outmigrant traps,May 14 through October 6,1984.
25
FLATHORN STATION CHINOOK 1 +
T----,-----==:::::::;:=====-----------r3 .3 -
-
a::
~
o
::t:.....
::t:
U...
0<
u
SEPAUG
%CUMULAnVE
%OF HIGHEST CPUE
+........L..1f-------,-___t'-..:.....:=--=~....:=...-o<:::>t--==---=--___t'-------+-O.O
100
90
80
70
60
I-
Z
W
0 50~wa.
40
30
20
10
0
Figure 15.Chinook salmon (age 1+)smoothed daily catch per unit effort and
adjusted cumulative catch recorded at the Flathorn stationary
outmigrant trap,May 20 through October 1,1984.
-
26
....,
-
-
,-
--
3.1.2 Growth
3.1.2.1 Age 0+
Chinook fry collected between the Chulitna River confluence and Devil
Canyon (middle river)averaged 43 mm during late May and showed a steady
growth through the season to a mean 1ength of 64 mm by early October
(Fig.16).Age 0+chinook collected between Cook Inlet and The
Chulitna River confluence (lower river)during the same period averaged
consistently larger than fry collected in the middle river.Chinook fry
in the lower river increased from a mean length of 41 mm in late May to
75mm in early October.The number of fish measured,mean length,and
range of lengths by sampling period for chinook salmon fry are presented
for each data collection area in Appendix Table A-3 and A-4.
3.1.2.2 Age 1+
Age 1+chinook salmon for all sites sampled averaged 78 mm during May
and the mean length increased to 90 mm during early June (Appendix Table
A-5).Average 1engths for thi sage cl ass stayed the same through 1ate
July by which time most of the age 1+chinook had migrated out of the
Susitna River.
The length/weight relationship of juvenile chinook (both age classes)at
Talkeetna Station is shown in Fig.17.
3.1.3 Cold branding
A total of 23,406 chinook salmon fry were cold branded in Indian River
between July 1 and October 15,1984 (Table 1).One hundred forty-seven
of these marked fish were later recaptured in Indian River,five were
captured in the Talkeetna Station outmigrant traps,and five were
captured below Indian River in side channels and sloughs associated with
the mainstem Susitna River.The time between release of marked chinook
fry in Indian River at TRM 7.2 and their subsequent recapture at the
mouth of this tributary ranged from nine to 70 days with a mean of 30
days.The five chinook fry branded in Indian River which were collected
in the outmigrant traps at Tal keetna Station averaged 17 days between
release and recapture with a range from 8 to 26 days.
A total of 9,802 chinook salmon fry were cold branded in sloughs and
side channels in the middle river between July 1 and October 15.Of
these fish,643 (6.6%)were later recaptured;637 in the same slough
where they were originally marked and released,·seven fish in sloughs
and side channels downstream from their release sites,four fish in the
Talkeetna Station traps and two fish at sites upstream from their points
of release.Of the 637 fry recaptured in the same slough where they
were marked,136 were caught 5 to 30 days later,and 113 were caught
30-60 days later.The branded chinook fry collected in the Talkeetna
outmigrant traps averaged 12 days between release and recapture with a
range from 8 to 17 days.
27
CHINOOK 0+
80
75
70
,,-...
::E 65~......,
:I:
I-
0 60z
W
....J
Z 5S<{w
~
50
45
40
L M;\Y E JUN L JUN
1984
E JUL L JUL E AUG L AUG E SEA..SEP-E OCT
SAMPLI NG PERIOD
'"""
-
-
-
......
Figure 16.Chinook salmon (age 0+)mean length and range of lengths by
sampling period for fish collected in the lower and middle
reach of the Susitna River,1984.
28
-
Figure 17.Weight/length relationship for juvenile chinook salmon
collected at the Talkeetna stationary outmigrant traps,
1984.
29
Table 1.The number of chinook salmon fry marked and recovered in Indian River by
sampling period,1984.
-.
Recapture Period
Number
Marki n9 of Fi sh July August August Sept.Sept.Oct
Period Marked 16-31 1-15 16-31 1-15 16-30 1-15 Total
July 1-15 2,093 26 10 5 2 3 3 49
July 16-31 1,924 5 4 5 5 2 21 -,
August 1-15 6,735 8 17 8 8 41
August 16-31 3,806 4 5 2 11
September 1-15 5,492 17 7 24
September 16-30 3,356
TOTALS 23,406 26 15 17 28 38 23 147
-
-
-
30
-.
_.
-
-
.-
-
3.1.4 Population estimates
Using the mark-recapture data of Table 1 with the Schaefer method
(Appendix B),there were an estimated 3,211 ,000 age 0+chinook fry in
Indian River after mid July.Females comprised 41.7%of the estimated
population of 8,482 (9120-7%milling)adult chinook salmon (greater than
350 mm)which passed Curry Station in 1983 [95%confidence interval
(C.l.)on estimate of 9120 of 6,148 to 14,212 fish;Barrett et ale
1984J.Indian River chinook comprised 27%of the peak spawning survey
counts (Barrett et ale 1984).Using a fecundity estimate of 10,622 eggs
p.er female (Healy and Heard 1984),an estimated 10,143,000 eggs were
deposited in Indian River during 1983.It is not possible to calculate
the egg to outmigrant survival rate because of unknowns in both the
adult and the fry population estimates.
Population estimates were made at three sloughs and two side channels in
the middle river during the cold branding study (Table 2).Populations
were estimated at a high of 47,000 chinook fry in Slough 22 to a low of
3,400 in Lower Side Channel 11A.No Jolly-Seber estimate of population
size was made for August 11 at Moose Slough because the head of site
closed the night of August 11 and almost all of the fish left.Only one
chinook fry was captured on August 12;there were no recaptures.
The effect of fluctuating discharge levels on the density (beach seine
catch with constant effort)and total number (population estimate).of
chinook fry in sloughs and side channels can be seen in Figs.18 and 19.
Estimates of population size were made using the Jolly-Seber method
which allows for inmigration,recruitment,outmigration,and mortal ity.
Recruitment does not occur,so all gains to the population were a result
of migration into the site.Similarly,assuming that mortality during a
five day period is negligible,all losses to the population were a
result of migration from the site.
The total number of fry in Moose Slough during these five days
paralleled the density of fry and the discharge level (Fig.18).This
pattern suggests that habitat qual ity was best at the highest observed
flow and declined with a drop in discharge level.As the surface area
of the site and the habitat quality decreased,so did the total number
of fish at the site.Evidently,the site is of little rearing value to
chinook salmon when the head of the site is not breached.A partial
explanation is that the water clears when the head is closed;there is
little cover other than turbid water at this site.The marked/unmarked
ratio for each day was diluted by the entry of new fish into the site
through the slough head,until the head closed.By that time,most of
fish that had been at the site the previous four days had left.
Residence time in this slough was low.This site probably acts mainly
as an outmigration corridor and temporary rearing area.
At Lower Side Channel 11A,the density of fry stayed relatively constant
over the five days even though the discharge level steadily decreased
(Fig.19).Meanwhile,the total number of fry at the site declined with
the lowering in discharge level.The table of recaptures (Fig.19)
indicates a longer residence time than at Moose Slough.This fact,and
the fairly constant density,suggests that the habitat quality at this
31
Table 2.Chinook salmon fry population estimates by site for middle Susitna River sloughs and side channels and,for Indian River,
1984.
Sampling Branding Recapture Estimate Population 95%Confidence
Site Dates Dates Method Estimate Interval
Lower Side Channel 11A 7/29 -8/1 7/30 -8/2 Schaefer 3,420
7/30 Jolly-Seber 4,962 2,466 -14,441
7/31 Jolly-Seber 1,370 1',038 -2,106
8/1 Jolly-Seber 1,245 958 -1,874
Side Channel 10 7/16 -7/19 7/17 -7/20 Schaefer 7,630
Moose Slough 8/8 -8/11 8/9 -8/12 Schaefer 4,990
8/9 Jolly-Seber 5,884 3,888 -11,141
8/10 Jolly-Seber 1.455 1.159 -2,071
Slough 22 9/8 -9/13 10/8 Petersen 47,050 39,000 -56.750
Schaefer 43.761
Slough 19 8/29 9/26 Petersen 4,550 3,200 -6.700
(.oJ
N Indian River 7/1 -9/30 7/15 -10/15 Schaefer 3.211,000
•'.~)J J J ~)t J )I 1 j D ,!.~J
33
-
RECAPTURE
29
DATE NO.7/30 7/31 8/1 8/2 TOTALMARK
(!)22 13 6 46 ....
Z 7/29 130 5
~7/30 209 20 27 17 64
28 a:«7/31 24 26 50:lE 179 -,
8/1 173 30 30
-27 0 -5
0
0 0
0 -0 0
)(·0
.,26 :-4 ~...
)(u .-.-POPULATION 0 -••·w ESTIMATE ·w.·l-.C)25 =-3 ct r:3a::0.•:E I -.ct -:I:l-I
(,)0 -en t-o ·en p--- -~~- ---
LIJ I 0-tBREACHING /:-2-Z ~2~-0 24 .......
""'0
FLOW //·0 I )(·-I -/·t-/.....ct ...:I:r/·Io·...11 •••••e ·...J I (,)
:-':;)r 1 I-23 a..ct·0 I Ua..I -,-t-·I
I
22 0 '0 1"""'1
2 :3 4 5
DAY -Figure 19.Catch,estimated population size,and main-
stem discharge 1evel at Lower Side Channel
llA,July 29 -August 2,1984.
-
34
.....
site is relatively.unaffected by changes in level of discharge.How-
ever,the total number of fry at the site necessarily declines with a
lowering discharge level because the amount of habitat (surface area)
available decreases.The constant density of fry even after the head of
the site closed is perhaps attributable to a greater amount of object
cover at this site than at Moose Slough.
3.2 Coho Salmon
3.2.1 Catch per unit effort
3.2.1.1 Age 0+
Juvenile coho salmon were observed during the coded wire tagging study
to be most abundant at Indian River.Catch rates were not recorded.
The cold branding study collected 1,548 coho salmon fry in Indian River
from July 1 through October 15.Of thi s catch,31%of the coho were
captured in Section I,44%in section II and 26%in section III.Beach
seining of sections II and III during July captured 444 juvenile coho
salmon;76%in section II and 24%in section III.Minnow trapping begun
in late July captured 1,129 juvenile coho salmon during 947 minnow trap
days for a season average of 1.2 coho per trap day.Of these catches,
43%were recorded in the lower section,31%in the middle section,and
26%in the upper section.
The catch per unit effort for all Jndi an Ri ver secti ons combi ned was
steady through the season rangi ng from 1.1 to 1.5 fi sh per trap day
(Fig.20).Coho fry catches were highest in section III with an average
of 5.0 coho per trap day over the season.Season average CPUE in
section II was 1.4 coho per trap day and Section I averaged 0.8 coho per
trap day..
A total of 90 coho salmon fry were captured during the cold branding
study in sloughs and side channels in the middle Susitna River.Ninety-
five percent of the coho catch was recorded in slough habitats in this
reach.Beach seining during July and August captured 40%of the
season's total catch while minnow trapping during September and early
October collected the remaining 60%(average of 0.2 coho per trap day).
Daily minnow trap CPUE ranged from a low of 0.01 at Slough 22 and Side
Channel 21 in September to a high of 7.6 coho per trap day at Slough 14
on September 10.
Peak catches for the 1,830 age 0+coho salmon collected at the Talkeetna
Station outmigrant traps were recorded during late July and August,and
the highest catch rate of 2.9 coho fry per hour was recorded on July 30,
by which time 50%of the season.total had been recorded (Fig.21).The
last coho fry was captured in the traps on October 4.
A total of 441 age 0+coho salmon were captured at the Flathorn
stationary outmigrant trap during 1984.Catch rates were highest during
late August and late September and the peak catch rate of 1.5 fish per
hour was recorded in the trap on September 30 (Fig.22).Fifty percent
of the catch at this site occurred by August 26.Only 16 age 0+coho
were captured in the mobile trap at Flathorn Station.
35
-
.....
-COHO JUVENILES I!\J DIAf'J RI\/ER ~I 984
8..,.---------------------------,'"'"
7
~Section UI
-
-E OCT
Combined...-~I
L SEP
..,..,j--~--___io.I_...!ectlon I
L AUG E SEP
SAMPLlNG PERIOD
E AUG
6
~a
~5
«a:::
l-
a:::4
w
0..
::r:3(j
I-
<:(
(j
2
1
0
L JUL
Figure 20.Coho salmon (age 0+)average catch per minnow trap by sampling
period and survey section in Indian River,1984.
~,
36
TALKEETNA BOTH TRAPS COHO 0+
100 T------------l--------:::::::=;;;=;:-i2 .3
90
SEPAUGJUL
%CUMULAnVE
%OF HIGHEST CPUE
I
l~
I~I ~
!\I
I \I
/'~J '\~!o -t-~~~~=:---+_----+---:::....-+_-~~~~.)...O.O
10
40
60
20
30
50
70
80
I-
Z
W
()a::w
0.
-
--
-Figure 21.Coho salmon (age 0+)smoothed daily catch per unitieffort and
adjusted cumulative catch recorded at the Talkeetna stationary
outrnigrant traps,May 14 through October 6,1984.
37
FLATHORN STATION AGE 0+COHO
""'"
""",
-.---------------------------.._1.1
~
l:t::
~
0
:z:
......
%
U
l-
e(-(,,)
~
Figure 22.Coho salmon (age 0+)smoothed daily catch per unit effort and
adjusted cumulative catch recorded at the Flathorn stationary
outmigrant trap,May 20 through October 1,1984.
38
-
..."
~I
,-
A tota 1 of 380 age 0+coho sa 1mon were captu red in the lower Sus i tna
River during the JAHS study (see Part 2 of this report).Catch rates
were highest during the late summer sampling and the peak catch rates
were recorded in early October (Fig.23).
The Deshka River weir captured 95 coho salmon fry during 1984;the peak
catch rate of 1.3 fish per hour was recorded on July 25 (Appendix Table
A-I).Minnow trap catches at this site were highest during late August
at 2.6 coho per trap (Appendix Table A-2).
3.2.1.2 Age 1+and older
Age 1+coho salmon were collected sporadically during the coded wire
tagginq study in May and June with the highest concentrations observed
in Slough 11 and Indian River.The cold branding study from July
through early October captured 25 age 1+coho at Indian River and 18 at
middle river slough and side channel sites during the season.
Peak catches for the 1,425 age 1+coho salmon juveniles captured at the
Talkeetna Station outmigrant traps were observed in mid June and were
again high in late July and late August (Fig.24).Fifty percent of the
catch was recorded by June 25.The highest catch rate for these age
classes was 1.6 fish per hour recorded on June 18 and the last capture
was on October 2.
Catch rates'for the 291 age 1+coho salmon juveniles captured at the
Flathorn stationary outmigrant trap were highest during late August and
September (Fig.25)and the highest CPUE of 0.8 coho per hour was
recorded on September 3.Fifty percent of the tota 1 ca tch was recorded
by August 30 and the 1ast capture of these age cl asses was October 1.
The mobile outmigrant trap captured 10 age 1+coho salmon during the
season.
The JAHS study in the lower ri ver coll ected 62 age 1+coho salmon
juveniles with most of the captures being recorded at tributary sites in
this reach.
The Deshka River weir collected 26 age 1+coho while minnow trapping at
this site captured 119 fish.Catches were observed throughout the
season with a peak rate of 6.2 coho per trap recorded in late August.
A total of 44 age 2+coho salmon juveniles were collected during the
1984 studies.Talkeetna Station,Flathorn Station,and the Deshka River
accounted for 95%of the captures of this age class.
3•2•2 Growth
3.2.2.1 Age 0+
Coho fry collected in the lower river were consistently larger than the
fry collected in the middle river throughout the season (Fig.26).Coho
fry collected in the middle river averaged 40 mm total length during
late May and showed a steady growth to a mean of 58 mm by late August.
Coho fry in the lower river averaged 42 mm in early June and had grown
39
COHO.CPUE 1984
-
....
-
-
100
90
80
w 70
::::J
Q.
()60
l-
UIw 50:I:
C)
:I:
U.40
0
~30
20
10
0
E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAMPLING PERIOD
.....
-
Figure 23.Coho salmon juvenile catch per unit effort by sampling period
recorded at JAHS sites in the lower reach of the Susitna
River,1984.
40
-
.,.,..
""""
......
--
l~
TALKEETNA BOTH TRAPS COHO 1+ex:2-+
1-
Z
tJ.J
()
0:
W
CL
100 :---------r-------------=:=::;==------r 16
90 1 I'
80 l /.%ClJtviU LATIV'E I
70 1 %OF HIGHEST CPUE I
:J ~!
40 ~\I :::I:
1\~
o -hAY JUN JUL ~u;hE/~O'O
-
Figure 24.Coho salmon (age 1+and older)smoothed daily catch per unit
effort and adjusted cumulative catch recorded at the Talkeetna
stationary outmigrant traps,May 14 through October 6,1984.
41
100
90
80
70
60
f-
Z
Wu 50
0::
Wa.
40
30
20
10
0
FLATHORN STATION COHO 1 +&2+
.....---------------------,,----'----r---r-O.6
%CUMUlATIVE
%OF HIGHEST CPUE
-l-__----lJ.:q:::..-+-....IL..-t-_:-:-::--_-+_-:-:::-=-__H-0.0
MAY
-
""'"
-
Figure 25.Coho salmon (age 1+and older)smoothed daily catch per unit
effort and adjusted cumulative catch recorded at the Flathorn
stationary outmigrant trap,May 20 through October 1,1984.
42 -
.-
COHO 0+1984
72 -,',..-----------------------------,
Middle Sualtna River
L AUG E SEP L SEP
E OCT
E JUL L JUL E AUG
SAMPLING PERIOD
L J·UNEJUN
56
54
52
50
48
46
44
42
40
38 -t-----r----r-----,------......---.,----.----,.-----!
l MAY
70
68
66
64
62........
:::E 60~
......,58
I
I-o
Z
W
-l
Z
~
::E
Figure 26.Coho salmon (age 0+)mean length and range of lengths by
sampling period for fish collected in the lower and middle
reach of the Susitna River,1984.
"....
43
to a mean length of 71 mm by late September.The number of fish
measured,mean length,and range of lengths by sampling period for coho
fry are presented for each data collection area in Appendix Table A-6
and A-7.
3.2.2.2 Age 1+and older
The average length of age 1+coho salmon juveniles collected in the
lower river during the open water season was greater than that of fish
of the same age class collected in the middle river (Fig.27).Age 1+
coho averaged 70 mm total length in both reaches during May and
increased to 104 mm in the middle river and 111 mm in the lower river by·
early October.Length data by collection area and sampl ing period are
provided in Appendix Table A-8 and A-9.
Age 2+coho salmon juveniles collected during the 1984 studies averaged
137.1 mm and ranged from 114 to 176 mm (Appendix Table A-10).
A sample of juvenile coho salmon were measured at Talkeetna Station to
provide a relationship between length and weight for fish passing this
site (Fig.28).
3.2.3 Cold branding
A total of 1,480 juvenile coho salmon were cold branded in Indian River
from July 1 through October 15.Of these fish,five were recaptured in
Indian River and two were recovered at the Talkeetna Station outmigrant
traps.The marked coho recaptured in Indian River were branded and
rel eased at TRM 11.5 on July 17 and recaptured at TRM 2.2 between
September 9 and 11,for an average of 55 days between rel ease and
recovery.The two branded coho recovered at Talkeetna Station were
released in Indian River on August 12 and were recovered in the outmi-
grant traps on August 31 and September 22;19 days and 41 days,
respectively,between release and recovery.
A total of 106 juvenile coho salmon were cold branded at slough and side
channel sites,and the only recapture was recorded at Talkeetna Station.
The recaptured fish was marked and released at Slough 14 on September 10
and was recovered in the traps on September 16.
3.2.4 Population estimates
Since only 100 to 200 of the estimated 750 adult coho passing Curry
Station in 1983 entered Indian River,and since juvenile coho of the
same brood year outmigrate as age 0+,1+,and 2+fish,few juvenile coho
salmon were captured for marking during the 1984 cold branding studies.
Too few branded coho salmon were recaptured to provide population
estimates for any of the sites surveyed.
3.3 Sockeye Salmon
3.3.1 Catch per unit effort
44
~,
-
-
-
COHO 1+1984
115
110 ,,//
105 //
/,-100 ff........
:i
:::iE 95.......//::I:
I-
0 90 ///zw /...l
z 85
«:(w
:::iE 80
75
,,-.70
65
MAY JUNE JULY AUG SEP-OCT
SAMPLING PERIOD
Figure 27.Coho salmon (age 1+)mean length by month for fish collected
.-.in the lower and middle reach of the Susitna River,1984.
45
COHO SALMON
~,
~,
10g e y =-12.21 +3.12 loge x
r 2 =O.98
22
20
18
16
14
E
Cl'
12
4-'.s::.
())10'4)
3:..,8
~
6
4
2
0
40 60 80
Total length (mm)
100
o
120 -
Figure 28.Weight/length relationship for juvenile coho salmon
collected at the Talkeetna stationary outmigrant traps,
1984.
46
-
!~
,...,.,
I
3.3.1.1 Age 0+
Sockeye salmon fry were collected during the coded wire tagging study in
May and June at sloughs 8A,9,11,and 21 but catch rates were recorded
only for Slough 21.These data were determi ned from 24 hour fyke net
catches and are presented in Appendix Table A-11.
A total of 248 sockeye salmon 'fry were captured at slough and side
channel sites in the middle river and in Indian River during beach seine
sampling conducted in July and August.Of these fish,94%were col-
lected in sloughs and the remaining 6%were collected in Indian River
and at mainstem side channels.
Peak catch rates for the 7,484 age 0+sockeye salmon fry collected at
the Talkeetna Station outmigrant traps were recorded in mid June and
early July with the highest da"ily catch rate of 13.0 sockeye fry per
hour occurring on June 18 (Fig.29).The major downstream redistri-
bution of sockeye fry in this reach had occurred by mid July (50%by
July 4).The last sockeye fry at Talkeetna Station was observed on
October 4.
Juvenile sockeye catches at the Flathorn stationary outmigrant trap were
greatest during May and June but the downstream movement of sockeye fry
continued through the open water season (Fig.30).A total of 2,315
sockeye fry were collected in the trap during 1984,and the peak catch
rate of 4.6 fish per hour was recorded on June 8.Fifty percent of the
catches had occurred by June 29 and the last capture was October 1.
Mobile trap catches of sockeye fry at Fl athorn Stati on were hi ghest
during June and the peak catch rate of 5.4 fish per hour was recorded on
July 12 (Fig.31).Of the 114 sockeye collected in the mobile trap
during 1984,59%were captured at bank transect points (Fig.32).
A total of 412 sockeye salmon fry were collected in the lower river
during JAHS surveys from June through.October (see Part 2 of this
report).Catch rates at JAHS sites peaked in late June and then were
low throughout the remainder of the season (Fi g.33).An increase in
catch rates was recorded at some sites including Rolly Creek'(RM 39.0)
and Beaver Dam Slough (RM 86.3)in late August and September,indicating
the movement of sockeye int,o these sites during late summer.
47
-
""'Ii
T)A,LK~EThJA BOTH TRAPS SOCKEYE {~"+\_.1
100 1 12•7 -I r-
gO 1 I
f
l-
80 !
(I
70 I l %CUMULATIVE ~~
!%OF HIGHEST CPUE I a:60 r~I-\0z ~w ~%(j SO !......0::
W /i::z:::Q.\(.)40 1-1-~,
\
I «
30 ~(.)
~-~-20 Il-I
10 Ir ~
~0
,
0.0
JUN JUL AUG SEP
Figure 29.Sockeye salmon (age 0+)smoothed daily catch per unit effort
.and adjusted cumulative catch recorded at the Talkeetna
stationary outmigrant traps,May 14 through October 6,1984.
~,
48
FLATHORN ST.ATION SOCKEYE 0+
100
90
80
70
60
~
Z
I.&J
<J 50
0::
I.&J
Q.
40
30
20
10
0
T--r;--------------=::;:::::::::::===:::;;::;:~13.5
%CUMU LATI\iE
%OF HIGHEST CPI)E
A/"--J-F-----il-------t------+------+-~~-~-+-0.0
"'UN JUL AUG SEP
Figure 30.Sockeye salmon (age 0+)smoothed daily catch per unit effort
and adjusted cumulative catch recorded at the Flathorn
stationary outmigrant trap,May 20 through October 1,1984.
49
-
-
-
-
/
-
-
-
--
,
-",If IF",v 1 r
IflRRaabaaaa••V ~b a a -
100
90
l.i.l 80
.~.
CL
e..;.70
I-
[J)
W
:r:60
D
I:
l.L 50
0
LLI
13 4-0<I-
Zw ,30
1:-'1
0:::
W
Q..:20
10
0
1215 20 JULY .30 1 5
DATE
-----------1
a -Not Sampled I
b -No FI8h Captured
I
a~ba ~bb baababb~-rT-rr-J-
10 AUGUST 20 2530
Figure 31.Sockeye salmon (age 0+)daily catch per unit effort recorded
at the Flathorn mobile outmigrant trap,July 12 through
August 31,1984.
50
~I'
AGE 0+SOCKEYE SEASON CATCH
24
22
20
:J:18
0.....«16<J
..J 14~
0....12
l.L.
0....10z
W
<J 8a::wa.6
4
2
0
"'1 2 3 4 "'5 *6 "'7 8 9 "'10
TRANSECT POINT NUMBER
'"BANK TRANSECTS
Figure 32.Sockeye salmon (age 0+)percent of the total catch by
sampl ;ng po;ntr :recorded at the Fl athorn mob;1e outmigrant
trap,1984.
51
90
1 00 -""---"""T?""~""""------------------------'
w
~
Q..o
tiiw:r:(..=l
:J:
1.La
~
80
70
60
.so
40
30
20
10
SOCKEYE CPUE 1984
~-
-
"'""
E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAMPLI NG PERIOD
Figure 33.Sockeye salmon juveniles catch per unit effort by sampling
period recorded at JAHS sites in the lower reach of the
Susitna River,1984.
52
TALKEETNA &FLATHORN SOCKEYE 1 + .
100 Tr-----::::=::::==::;;=::;::::;:r----------i
SEPAUG
%CUMULATI\IE
%OF HIGHEST CPUE
70
10
40
20
60
30
80
O.....f+'----+------'~--J...I.---L.\.-....u...f__----_t_-----_H
90
....zw.()50a::wa.
Figure 34.Sockeye salmon (age 1+)smoothed daily catch per unit effort
and adjusted cumulative catch recorded at the Flathorn and
Talkeetna stationary outmigrant traps,May 14 through
October 6,1984..
-
-
53
~,'3.3.2 Growth
The mean length and range of lengths for age 0+sockeye salmon by reach
of river and sampling period is presented in Fig.35.During May and
June,sockeye fry collected in the middle river reach had a smaller mean
length than the same age class sockeye collected in the lower river.By
early July,sockeye fry averaged the same length (49 mm)in both -
reaches,and by late August,middle river sockeye fry were averaging
larger than fish collected in the lower river.This trend continued
through the remainder of the season.The number of fi sh measured,the
mean length and range of lengths by sampling period for sockeye saJmon .
fry are presented for each of the data collection areas in Appendix
Table A-12.
'""'"
The 90 age 1+sockeye salmon collected during 1984 ranged from 56 to
102 mm total length (Appendix Table A-13).A coded wire tagged sockeye
fry released in 1983 and recaptured in 1984 had increased from 32 mm to ~
81 mm.
A sampl e of juveni 1e sockeye were measured at Talkeetna Station to
provide a relationship between length and weight for fish passing this
site (Fig.36).
3.3.3 Coded wire tagging and recovery
A total of 14,532 tagged sockeye salmon fry averaging 33 mm total length
were released between May 22 and June 22,1984 (Table 3).Tag retention ~
rates for sockeye fry averaged 97.1%and ranged from 92.3 to 99.0%.
Tagging mortality ranged from 0.6 to 2.6%and averaged 1.3%.
.....
A total of 366 tagged sockeye salmon fry (2.5%of the total tagged
sockeye released)were recovered from the 7,484 age 0+sockeye captured
and examined for tags at the Talkeetna Station outmigrant traps during
1984.In addition,15 sockeye fry with clipped adipose fins but no
coded wire tags were recovered in the traps.When compared to the total
tagged sockeye salmon fry recovered,this provides a tag retention rate
at the traps of 96.1%.
Trap recoveries of coded wire tagged sockeye fry were made from a to 109
days (mean =35 days)following their release at the tagging sites (Fig.
37).In addition,one tagged sockeye fry which was released from Slough
21 on May 28 was recaptured at Flathorn Station on July 7.Seven coded
wire tagged sockeye fry were recovered during the cold branding study in
early August (Tabl e 4).Six of these fi sh were recovered at Moose
Slough (RM 123.2)and one tagged sockeye fry was recovered at a side
channel below Slough 11 (RM 135.2).
A single coded wire tagged sockeye salmon marked and released during
1983 was recovered during the 1984 sampling season.This fish was
released June 8,1983 at Slough 11 and was recovered at Talkeetna
Station on July 21,1984.
_.
54
,-
SOCKEYE 0+"1984
L AUG E S EP L S EP
E OCT
L JUL E AUG
SAMPLING PERIOD
E JULLJUNEJUN
52
50
48"
46
44
42
40
38
36
34
.32
30 -+---.......---,---,.-----.---.......----,,.-----.-----1
L I'v1AY
62 .,--------------------------...,
60
58
56
54
Figure 35.Sockeye salmon (age 0+)mean length and range of lengths
by sampl ing period for fi sh call ected in the lower and
middle reach of the Susitna River,1984.
55
SOCKEYE SALMON
6-r------------------------------.
10Qe y=-12.33+3.IIIOQ.K
r 2 =O.97
4
1
5
~
E
0'
3 1If%~
4-'
.J:en
'tj
~
tJ 2>:J
907050
O-l----.,.----,...-----,~--___,.---__r---_r_----1
30
Total length (mm)
Figure 36.Weight/length relationship for juvenile sockeye salmon
collected at the Talkeetna stationary outmigrant traps,
1984.
56
Table 3.Coded wire tag release data for sockeye salmon fry on the Susitna River by
tagging site and release date,1984.
Tagging Site Number of Date of Percent Tag Percent
(River Mil e)Fish Tagged Release Retention Mortality
Slough 21 3,736 5/28 97.9 2.6a
(RM 142.0)
Slough 11 2,327 5/22 92.3 1 .1
(RM 135.3)2,732 5/24 97.7 0.7
1,537 6/22 96.6 1.1-Slough 9 2,052 6/9 99.0 1.0
(RM 128.3)
Slough 8A 2,148 6/19 99.0 0.6
(RM 125.3)
TOTAL -ALL SITES 14,532 5/22-6/22 97.1 1.3
a Mortality due to handling,thermal,and anesthetic stresses.
57
CODED WIRE TAGGED SOCKEYE
80
70
I-
::I:
C)60~
(J
::I:
(J)50
ii:
Clw 40C)
0<t:
l-
lL.300
0::
W
CD 20~
::::l
Z
10
0
5 20 35 50 65 80 95 110
NUMBER OF DAYS AFfER RELEASE
(Grouped by 5 Day Period)
Figure 37.Length of time between the mark and recapture of coded
wire tagged sockeye salmon juveniles in the middle reach
of the Susftna River,1984.
58
,~
-
-
..!,
Table 4•Recoveries of coded wire tagged sockeye salmon fry at
mainstem river sites between Tal keetna and Devil Canyon,
1984 •
....Collection.Collection Release Release
Site Date Site Date
Moose Slough 1 8/8 Slough 21 5/28
r-Moose Slough 8/8 Slough 21 5/28
Moose Slough 8/8 Slough 11 6/22
Moose Slough 8/8 Slough 9 6/9
Moose Slough 8/8 Slough 8A 6/19
Moose Slough 8/8 Slough 8A 6/19
Slough 11 Side Channe1 2 8/3 Slough 21 5/28
1 River Mile 123.2~
2 River Mile 134.9
59
The ratio of coded wire tagged sockeye fry to total sockeye fry was the
same (0.05:1.00)in both traps at Talkeetna Station.This indicates
that the coded wire tagged fish were uniformly mixed in the total
population by the time they migrated past the traps.
3.3.4 Population estimates and survival rates of outmigrants
Females comprised 38.5%of the population of 1,900 adult sockeye salmon
estimated past Curry Station in 1983 (95%C.!.-1,600 to 2,300 adu1 ts)
and the fecundity of Susitna River sockeye averaged 3,350 eggs per
.female,with a 95%C.I.of 3131 to 3569 (Barrett et al.1984).Milling
activity was estimated at 30%(Barrett 1984).These data provided a
calculation of total potential egg deposition for sockeye salmon of
1,715,000 eggs during 1983.
Using the method outlined by Schaefer (1951),the number of age 0+
sockeye salmon fry above Ta"lkeetna Station during 1984 was estimated to
be 299,000 (Appendix Table B-1 and B-2).A comparison of this estimate
to the calculated potential egg deposition (dividing the estimated
number of fry by the number of eggs)gave an egg-to-outmigrant fry
survival rate of 17%.The reliability of this estimate is not currently
known because there is no way to estimate the variance of the adult
mi 11 i ng estimate and because we do not currently have a method of
estimating the variance on the Schaefer estimate of the fry population
size.
3.4 Chum Salmon
3.4.1 Catch per unit effort
Chum salmon were collected during the coded wire tagging study in May
and June and during beach seine sampling of Indian River in July.Catch
rates were not generally recorded during these studies except for 24
hour fyke net sets at Slough 21 (Appendix Table A-10).
Peak catches of chum fry collected at the Talkeetna Station outmigrant
traps were recorded during late May and mid June,with the highest daily
catch rate of 8.0 fish per hour occurring on June 14 (Fig.38).Ninety-
fi ve percent of the 3,590 chum fry captured at Tal keetna Stati on were
recorded by July 15.The major outmigration had occurred by the end of
June (50%by June 13),although the migration continued until September
11.
Chum salmon fry catches at Flathorn Station were greatest during June
with a peak catch rate of 10.9 fish per hour recorded on June 14 by
which time 50%of the season catch had occurred (Fig.39).By July 1,
97%of the chum fry collected at this site had been captured;the last
chum fry was captured on July 22.
Beach seining and electrofishing at side channel,slough,and tributary
sites in the lower river reach collected chum salmon fry during June and
July (see Part 2 of this report).Chum fry were abundant in this reach
during early June but catches steadily decreased through July (Fig.40).
60
-
-~
-
-
~
T,A.Lk~EFTI',JA BOTH TR.A.PS CHUM FR"'(
100 -[
t··
2~.....
/9°1
\so -I
70 ~\%CUMU LATI\/E
.I
~~I ~\I
%OF HIGHEST CPUE
80 _I (~I~-zw
0 50 -10::-0
W
t:i...~
40-et
U
Ii r I
-
30 -./~~I"'""
2')-1'1/1'I V\j~-l:V ~)
I JUL ~I 0.0
MAY I JUN AUG SEP
~
....,
Figure 38.Chum salmon fry smoothed daily catch per unit effort and
adjusted cumulative catch recorded at the Talkeetna
stationary outmigrant traps,May 14 through October 6,1984.
-
61
~,
~
FLATHORN STATION CHUM FRY
100 5.8 -
90
l -~80
70
60 %CUMULAnYE l~l-%OF HIGHEST CPUEzw ~
()50
0:I :r:w I <.iIi.
40 I-
<l:-u
30
20
10
0 0.0JUNJULAUGSEP
~,
Figure 39.Chum salmon fry smoothed daily catch per unit effort and
adjusted cumulative catch recorded at the Flathorn
stationary outmigrant trap,May 20 through October 1,1984.
-
62
r-.
CHUM CPUE ·1984
100 -rr~;o-r------------------------...,
90
.-
-
.-
I
W
:::Ja.u
t-
(.I!
W:::c
Cl
:::c
lJ...o
te
80
70
60
50
40
30
20
10
E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAMPLING PERIOD
-
Figure 40.Chum salmon fry catch per unit effort by sampling period
recorded at JAHS sites in the lower reach of the Susitna
River,1984.
63
3.4.2 Growth
At both Talkeetna and Flathorn Stations,chum length ranged from emer-
gent lengths (less than 35 mm)to lengths greater than 60 mm for May,
June,and July (Appendix Table A-14).Chum salmon spawn in both tribu-
taries and sloughs and there is a wide range in emergence timing.The
fish caught at 30-40 mm are probably recent emergents.The 50-60+mm
fish have gained over 20 mm in length.
During June,Indian River chum fry averaged 40 mm and had increased to a
mean length of 48 mm by early July.Limited sampling of the Talkeetna
River during June and July indicated a mean length of 43 mm for chum fry
outmigrating from this tributary.
3.4.3 Coded wire tagging and recovery
A total of 31,396 tagged chum fry averaging 43 mm total length were
released between May 22 and June 22,1984 (Table 5).Tag retention
rates ranged from 93.0 to 100%and averaged 96.4%.Mortality rates
between tagging and release averaged 0.9%and ranged from 0.0 to 2.7%.
Fifty-one tagged chum salmon fry (0.2%of the total tagged chum
released)were recovered from the 3,590 chum salmon fry captured and
examined for tags at the Talkeetna Station outmigrant traps during 1984.
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 96.2%.
Trap recoveries of tagged chum fry were made from 0 to 29 days (mean =8
days)following their release at the tagging sites (Fig.41).
The ratio of coded wire tagged chum fry to the total number of fish
caught at each trap at Talkeetna Sta ti on was 0.016:1 at Trap 1 and
0.013:1 at Trap 2,indicating that the tagged chum fry were randomly
distributed with the untagged population by the time they migrated past
the traps.
3.4.4 Population estimates and survival rates of outmigrants
Adult population estimates at Curry Station during 1983 were 21,100 chum
salmon with 95%confidence limits of 19,200 to 23,500 adults.Females
comprised 34.5%of these fish and chum salmon milling was estimated at
40%(Barrett et ale 1984).Fecundity of Susitna River chum salmon was
determined during 1983 to be 2,850 eggs per female (95%confidence
limits of 2,666 to 3,034).These data provided an estimated total
potential egg deposition of 12,448,000 eggs.
The population estimated using the Schaefer (1951)method was 2,039,000
chum salmon fry outmigrating past Talkeetna Station during 1984 (Appen-
dix Table B-3 and 8-4).Using the above data,an egg-to-outmigrant fry
survival rate of 16%was calculated for chum salmon.As with sockeye
salmon,there is no way of knowing the reliability of the estimate
64
-
-
-
-
"""':
Table 5.Coded wire tag release data for chum salmon fry on the Susitna River by
tagging site and release date,1984.
Tagging Site Number of Date of Percent Tag Percent
(River Mile)Fish Tagged Release Retention Mortality
Slough 22 2,383 6/1 98.0 0.5
(RM 144.3)
r-
Slough 21 2~201 6/3 96.6 1.4
(RM 142.0)
Slough 20 1,255 6/11 96.9 0.6
(RM 140.1)
Slough 15 351 6/14 100.0 0.0
(RM 137.3)
Indian River 4~612 6/1 94.5 0.7
(RM 138.6)341 6/1 93.0 O.Oa
4~592 6/21 93.8 2.7
-"2,511 6/22 95.0 0.4
Slough 11 2~031 5/22 97.7 0.1
(RM 135.3)2~203 5/24 93.9 0.3
572 5/24 99.0 0.2
1~916 6/16 98.0 0.4
Slough 9 5~122 6/6 99.4 0.7
(RM 128.3)
Slough 86 1~306 6/13 98.0 0.8
(RM 122.4)
,...
TOTAL -All SITES 31~396 5/22-6/22 96.4 0.9
a High mortality due to injury from improper headmold.
-
-
.-
65
CODED WIRE TAGGED CHUM'SALMON
-27246 9 12 15 18 21
NUMBER OF DAYS AFTER RELEASE
3o
15 ""'T"""---------------------------,
14
13
12
11
10
9
8
7
6
5
4
3
2
1
O-"r-l..I!r.l-"r-l.l~~f.J_l',,..J_I',..u:,..L.r__.___.__.,..._,....._I!,...l..lIT.uyuyJiT_U~T.L..r__I'r...l.....r...y.....,......__._..,_JlT-UY
.......:r.:
<=J
~
(J
.:r.:
CJI
Li:
o
I.LI
<=Jo
;!
La..o
0::
I.LI
CD
~
:::J
Z
Figure 41.Length of time between the mark and recapture of coded
wire tagged chum salmon juveniles in the middle reach of
the Susitna River,1984.
66
-
,I"'-
.
1"-
,.-
because the variance of the adult milling estimate and the variance of
the fry population estimate are not known.
3.5 Pink Salmon
Sixty-eight pink salmon fry were captured between May 15 and July 18 at
the Talkeetna Station outmigrant traps during 1984,with the peak catch
rate of 0.8 fish per hour being recorded on June 18 (Fig.42).Pink fry
migrating past Talkeetna Station averaged 36 mm total length with a
range from 29 to 53 mm.
A total of 405 pink salmon fry were collected in the stationary outmi-
grant trap at Fl athorn Station.Catches occurred from May 21 through
July 6 and the peak catch rate of 4.0 fish per hour was recorded on June
5 (Fig.43).Fifty percent of the catches at this site were recorded by
June 11.Pink fry collected at Flathorn Station averaged 34 mm and
ranged in length from 25 to 46 mm.
No pink salmon fry were collected during the cold branding studies in
the middle river,during sampling of the Deshka River,or at JAHS sites
in the lower river during 1984 •
3.6 Descriptive Statistics for Catch and Environmental Variables
Summary statistics for Talkeetna Station catch are given in Table 6 and
for environmental variables in Table 7.Flathorn data are summarized in
Table 8.The influence of discharge peaks on the level of outmigration
can be seen by comparing the seasonal discharge level (Fig.44;Fig.45)
with the outmigration plots presented earlier.Results of a statistical
time series analysis of 1983 and 1984 discharge,turbidity,and age 0+
chinook and sockeye salmon outmigration are presented in Appendix C.
67
TALKEETNA BOTH 'TRAPS PI i'.n<
-
100 0.2
!I
90 I
!80 I
I
70
1%CUMULATIVE a::
60 I
I Jroo%OF HIGHEST CPUE 0z
LU 1 ::c
()50 t-,0::
W I :t:n.j l)40 I-
et
l)
30
10
o ..pC-.i.......L---L..lf----U.+-...l-I.-L---"--_-++-+-_+_.0.0
AUG SEP
Figure 42.Pink salmon fry smoothed daily catch per unit effort and
adjusted cumulative catch recorded at the Talkeetna
stationary outmigrant traps,May 14 through October 6,1984.
68
'"""
-
-
-
..-
-
Fi gure 43.Pink sal men fry smoothed daily catch per unit effort and
adjusted cumulative catch recorded at the Flathorn
stationary outmigrant traps,May 20 through October 1,1984.
69
Table 6.Summary statistics for juvenile salmon catch per hour by
species and age cl ass recorded at the Tal keetna Station
outmigrant traps,May 14 through October 6,1984.
Catch Per Hour,Both Trapsa
-
Chinook 0+
Chinook 1+
Coho 0+
Coho l+b
Sockeye 0+
Sockeye 1+
Chum
Min
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Max
17.2
3.5
2.9
1.7
13.0
0.3
8.0
Mean
2.2
0.3
0.3
0.3
1.2
0.0
0.7
Std.Dev.
3.2
0.6
0.4
0.3
1.8
0.0
1.2
a n =146
b includes all juvenile coho age 1+or older.
....,
Table 7.Summary stati.stics for habitat variables recorded on the
Susitna River between the Chul itna River confl uence and
Devil Canyon,May 14 through October 6,1984.
"""i
Min -Max Mean Std.Dev.n
Discharge (ft3 jsec)a 6,780 52,000 19,405 8160.0 146
Water Temperature (oC)b 2.0 13.5 8.8 3.0 145
Turbidity (NTU)b 13 400 115 92.0 145
a USGS provisional data at Gold Creek,1984.
b ADF&G data at Talkeetna Station outmigrant traps,1984.
70
-
-
-
Table 8.Summary statistics for juvenile salmon catch per hour by
species and age class recorded at the Flathorn Station
outmigrant traps,May 20 through October 1,1984 .
..-
'Catch Per Hour a Min Max Mean Std.Dev.
Chinook 0+0.0 7.8 0.7 1.1
Chinook 1+0.0 6.5 0.1 0.6
Coho 0+0.0 1.5 0.1 0.3
Coho l+b 0.0 0.8 0.1 0.1
Sockeye 0+0.0 4.6 0.8 0.8
Sockeye 1+0.0 0.4 0.0 0.1
~Chum 0.0 10.9 0.3 1.1
Pink 0.0 4.0 0.2 0.5
Discharge {ft 3 /sec)c 40,800 166,000,93,122 28,887.5
.-a n =134.
b Includes all juvenile coho age 1+or older.
/"""
c USGS provisional data at Susitna Station,1984.
-
-71
-
,-
IS'"U
0-~-
W 1121~
::)
~.".,«
51313~
W S -0-
4121121 :;:)-L I-W Z~-121 31210
>-
I-
21313 H6eeeeCl
H--seeee 1130 enTurbidity~en :::>-0 4eeee a I--
W
(!)3eeee
Q::
<C::c 2eeeeu
(J)
H leee00
i0 172431 7
MAY JUN
14 21 28 5 12 19 26 2 9
JUL AUG
1984
16 23 30 6 I 3 213 27 4
SEP OCT
-
Figure 44.Mainstem discharge,water temperature,and turbidity in the middle
reach of the Susitna River,1984.Discharge was measured at the
USGS gaging station at Gold Creek.Water temperature and turbidity ~
were measured at Talkeetna Station.
-
72
--
-
,.....-OJ 16eeee-o-W .281/188
CJ
ct<~
(J
f1J-C
-
-
II 17 14 111 7 '4 II ••.1 I'•11 I .1 • •I .1 •117 4
MA Y JUNE JULY AUGUST SEPTEMBER OCTOBER
1984
Figure 45.Mainstem discharge in the lower reach of the Susitna River
measured at the USGS gaging station at Susitna Station,1984.
73
4.0 DISCUSSION
4.1 Chinook Salmon
4.1.1 Outmigration
Fifty percent of the outmigration of age 0+chinook salmon past
Talkeetna Station during both 1983 and 1984 had occurred by mid July,
but the rates and timing were different between the two years (Fig.46).
Duri ng 1983,two pul ses of ch"j nook fry movement were recorded,one in
late June.and the second in mid August.Conversely,the 1984 out-
migration did not start until mid June and was then relatively steady
through late August.
Low tributary flows during July of 1983 trapped chinook fry in pools and
side channels in Indian River until high tributary flows from heavy
rainfall in mid August allowed access or flushed fry to the Susitna
Ri ver (Roth et a 1.1984).In 1984,mi nnow trap catches of marked and
unmarked chinook in Indian River during the cold branding study showed
the movement of chinook fry out of this tributary continued from July
through early October.
In 1984,age 0+chinook salmon in the middle river that had outmigrated
from the tributaries were found predominately in shallow,turbid,rocky
bottom areas in breached sloughs and side channels during July and
Augus t.Not unt i 1 mi d August,when rna i nstem flows had decreased and
many of these sloughs and si de channel s were no longer breached,did
catches of juvenile chinook increase at clear water sloughs and side
channels.In early September,juvenile chinook were concentrated at the
mouths of clearwater sloughs and side channels,but as water tempera-
tures and stage continued dropping through September and early October,
these fish slowly dispersed within these sites with the major concen-
trations being found in areas with non-imbedded substrate and a
groundwater source.
The rates of outmigration of age 1+chinook salmon past Talkeetna
Station were similar in 1983 and 1984 (Fig.47),but the date by which
half of the total seasonal outmigration occurred was ten days earlier in
1983 than in 1984,primarily because of the late start of outmigration
in 1984.
The chinook fry appear to associate with the banks of the river during
their downstream movement.Although juvenile chinook were captured
across the entire river at Flathorn Station,60%of the total mobile
trap captures were recorded at bank transect sites.
4.1.2 Freshwater life history
Chinook salmon juveniles in the middle river appear to group into three
separate categories.The first group are those juveniles which rear and
overwinter in their natal tributaries and outmigrate to the ocean as age
1+fi sh duri n9 the spri ng of thei r second yea r.The second group of
chinook juveniles spend a portion of their first summer in their natal
tributaries and then,probably because of density dependent interaction,
74
-
-
-
~
I
j.
--,
-
-
-
.....
,~
--
1983 &
100
90
80
l1J 70>3 60
:::J
:E
:::J 50Q
t-
Z
l1J 40Qa::
l1Ja..30
20
10
0
MAY
1984 TALI'<EETNA CHINOOK 0+
-
.-
Figure 46.Chinook salmon (age 0+)adjusted cumulative catch recorded
at the Talkeetna stationary outmigrant traps,1983 and 1984.
75
1984 TALKEETNA CHII'JOOK-1983 &
100
90
80
w 70>i=
:5 60:J
~
::::J 50()
I-
Zw 400
0::
Wa.30
20
10
0
JUL AUG
l+
~-
-
-
Figure 47.Chinook salmon (age 1+)adjusted cumulative catch recorded
at the Talkeetna stationary outmigrant traps,1983 and 1984.
76
-
,....
behavioral changes related to downstream re-distribution~or flushing by
high flows~enter the mainstem river.These fish actively search out
suitable habitats as they move downstream.Many of the fish enter
sloughs and side channels in the middle river to overwinter while others
continue downstream to the lower river.Since 80%of the Talkeetna
Station trap catch had occurred by August 1~and high catches were still
being recorded at Indian River and selected sloughs.above Talkeetna
Station in August~September,and October~it appeared that a signi-
ficant percentage of 1983 brood year chinook salmon belonged to one of
these two groups.We do not know what this percentage was.
A third group of chinook salmon juveniles may be present in the Susitna
River.Data collected at the Flathorn Station outmigrant trap showed
that a porti on of the age 0+chinook were movi ng downstream past thi s
site.Many o-f these were probably fry from the Deshka River.Although
it is possible that these fish overwintered in freshwater habitats below
Flathorn Station~it appeared that many of these fish entered the ocean
as age 0+fish because few rearing chinook fry were found at sites below
the Deshka River during 1984 (see Part 2 of this report).
Intermittent operation of an outmigrant weir on the Deshka River during
1984 showed that a large number of age 0+chinook fry were outmigrating
from this tributary during July and August.Similar data were collected
in 1980 by Delaney et ale (1981),who postulated that the observed
outmigration was a size related response as the fish reached approxi-
mately 80 mm.It is not known whether these fish remain in habitats
associated with the mainstem river or if they continue to the ocean as
age 0+fish.
Scale samples collected from returning adults at Sunshine Station and
above indicated that the age 0+class of outmigrants represented less
than 3%of the middle river returning chinook during 1983 (Barrett et
ale 1984)and less than 1%in 1984 (Barrett et ale 1985).However,no
adult chinook scale samples were taken in 1984 at Flathorn Station,
which did not begin operation until early July.It may be that a
significant proportion of the adults bound for lower river tributaries
such as the Deshka did outmigrate during their first summer.
Otherwise~if it is assumed that a significant percentage of Susitna
River chinook salmon migrate to the ocean as age·0+fish~then either
the marine survival of this age class is very low or the freshwater life
histories on adult scales were not interpreted correctly.Ri chards
(1979)reported that a major portion (72%)of the adult scales analyzed
from the Deshka River during 1978 indicated that the fish had migrated
to the ocean during their first summer as age 0+fish.Scale analysis
from creel census samples collected in the Deshka River have classed
these fish as predominantly age 1+outmigrants (Kubik 1967;Kubik and
Wadman 1978;Kubik and Delaney 1980)..
There are many unanswered questions about chinook fry life history in
the Susitna River.Aging of adult chinook at Flathorn Station during
1985 will help answer the question of whether there is a significant
proportion of returning adults which outmigrated during their first
77
summer.However,we still do not know the proportion of returning
adul ts whi ch,as fry,followed one or the other of the three 1i fe
history strategies discussed above.The answer to this question is of
major importance in assessing dam-related effects on the population.
4.1.3 Estimates of population size and residence time
The Schaefer population estimate of 3.2 million chinook salmon juveniles
in Indian River in 1984 must be qualified.A successful method of
sampling large numbers of juvenile chinook and a location containing
large numbers were not found unti 1 mj d July,at whi ch time over 50%of
the Tal keetna Station trap catch of age 0+chinook fry had occurred.
Therefore,this estimate is only for those fish in Indian River for the
period July 15 to Oct.15.
The efficiency of minnow traps decreases when flows are high.Because
the marked fish were not randomly re-introduced into the system,we have
to assume that the.recapture was random.However,there is some reason
to believe that the unmarked fish were more l'ikely to redistribute
downstream during high flows than were the marked fish,which were
re-introduced into side sloughs.
Having two separate groups of juvenile chinook within Indian River,
those fish which overwinter in Indian river and the middle Susitna River
and those fish which migrate out of this reach,further complicates the
population estimate.Most marked fish were marked near the mouth of
Indian;it is likely that fish captured near the mouth were going to
migrate out of Indian River during the first summer.Also,it has to be
assumed that these fish,when transported back upriver,randomly mixed
with the other fry.The estimate of 3.2 million fry for Indian River
should be used as a rough approximation,obtained by an experimental
project.Information gathered during the 1984 season will enable a more
refined estimate for the 1985 season.
The chinook fry population estimates made for sloughs and side channels
give a general idea of how many fry these sites can support.The
day-to-day variation in total number of fish at these sites,which
results from variation in discharge level,is striking.Another impor-
tant result of this study is the residence time of rearing chinook fry
at these sites because of the implications this has on the results of
the IFIM and RJHAB models of rearing habitat {presented in Part 2 of
this report}.Habitat value from the models is measured by weighted
usable area {WUA},which depends only on water depth,water velocity,
cover,and substrate.The model will predict discharge levels at which
habitat value of a site is high.However,there may not be many fish at
a site,even when WUA is high,because of previous flushing of the site·
by a high discharge or because of a seasonal effect in level of out-
migration.More importantly,if the fish are using a site only as an
outmigration corridor,as appeared to be the case at Moose Slough in mid
August,then it really doesn't matter if the WUA is high or low,because
WUA measures only rearing habitat quality.On the other hand,if the
fish have a longer residence time at a site,such as at Lower Side
Channel llA in late July,then the amount of WUA is important.
78
-
.....
-
t~
-
......
Of the 643 chinook fry which were captured in a slough or side channel,
cold-branded,and later recaptured at the same site,113 were still
present 30-60 days later.This indicates that a substantial amount of
chinook fry rearing occurs at these sites.
4.1.4 Growth
The increase in mean length of age 0+chinook by sampling period for the
combined data collected at the Talkeetna Station outmigrant traps during
1982,1983 and 1984 is presented in Fig.48.Chinook fry,which emerge
from the gravel at an average length of approximately 37 mm,had
increased to an average of 44 mm by early June.By the end of the
open-water season,their mean length was 63 mm.Chinook fry collected
in the lower river in 1984 averaged from two to ten rnm larger than their
counterparts in the middle river through the season (Fig.16).
Chinook fry which overwinter in Indian River show little growth between
late October (when they are a little less than 70 rnm long)and late
March (ADF&G,unpublished data).Outmigrating age 1+fish at Talkeetna
station averaged 90 mm during the peak of outmigration,so they had
grown about 20 mm during April,May,and June.
Examination of the downstream redistribution of juvenile chinook salmon
in'the Susitna River by age class during 1984 shows that chinook fry in
the middle river averaged approximately the same length (50 to 55 mm)
throughout the period of peak outmigration (late June through early
August).This results in very little separation between cumulative
movements recorded for catch and biomass at Talkeetna Station (Fig.49).
The outmigration of chinook fry in the middle river appears to be
triggered,in part,by the fish reaching a critical size.As they reach
this critical size (estimated at 55 mm),chinook fry redistribute down-
stream to other rearing areas.
In the lower river,total biomass movements were delayed in comparison
to the total number of chinook fry moving past Flathorn Station (Fig.
49).This was due to the growth occurring in the lower river and
because of the mixed stocks present in this reach.
4.2 Coho Salmon
4.2.1 Outmigration
The downstream movement of coho salmon fry past Ta 1 keetna Stati on is
compared for 1983 and 1984 in Fig.50.Although the outmi grati on from
May through early July was slower during 1984,50%of the total season
outmigration was recorded ten days earlier in 1984 than in 1983.The
delay in downstream movement observed during July of 1983 was due in
part to low tributary water levels during this period,and the high
rates of downstream movement recorded in mi d August corresponded to a
period of heavy rainfall and high tributary discharges •
79
CHINOOK 0+MEAN LENGTH
80
75
.-70
:::::E
:::::E.......,
::I:65I-
Cl
Z
W
....J 60
....J
<{
I-a 55I-
z
~
~50
45
<)Maxlmum
Combined
+Minimum
E SEP L SEP
OCT
L AUG
+4O-j-----'T----r------r-----r---..-----,------r-----l
L MAY E JUNE L JUNE E JULY L JULY E AUG
SAMPU NG PERIOD
Figure 48.Chinook salmon (age 0+)mean length and range of mean lengths
by sampling period recorded at the Talkeetna stationary
outmigrant traps during 1982,1983,and 1984.
~,
-
80
1 )J J J )))))J I 1 J J 1
~
::l
::i
::lo
~wo
'"~
'1984 TALKEETNA CHINOOK 0+
100 •:::;;JF='"fB II I
90
80
70
60
50
40
30
20
10
o •{:Ii'I I I i I
L MAY E JUNE L JUNE E JULY L JULY E Aue L AUG E SEP l SEP E OCl
SAMPLING PERIOD
1984 FLATHORN CHINOOK 0+
1984 TALKEETNA CHINOOK 1 +
100
90
80
~70
~
::l 60.::Ii
::l
0
!i;;50
w
0
'"40w
Q.
30
20
10 T ,.-.-I I I 1 ,,I
L M!W E JUNE L JUNE E JULY L JULY E AUG L AUG E SEP l SEP E OCT
SAMPLING PERIOD
1984 FLATHORN CHINOOK 1 +
co
I-'
w
?;
~
::l
::i
::lo...z
Wo
'"W
Q.
100 I :;)i T II I
90
80
70
60
50
40
30
20
10
01 If --r-I I,I I I I
L MAY E JUNE L JUNE E JULY L JULY E AUG L AUG E SEP l SEP E OCT
SAMPLING PERIOD
~
::l
::i
::lo...z
~
W
Q.
.100
1
~II I
90
80
70
60
50
40
30
20
10
o T'i I I Iii I i I
L MAY E JUNE L JUNE E JULY L JULY E Aue LAue E SEP L SEP E OCT
SAMPLING PERIOD
Figure 49.Chinook salmon adjusted cumulative catch and biomass by age class recorded at Talkeetna and
Flathorn stations,1984.
1983 &1984 TALKEETNA COHO 0+
100
90
80
w 70>;:=
:5 60
:=l
2
:J 500
I-
Zw 400
0::
Wa..30
20
10
0 MAY
Figure 50.Coho salmon (age O+)adjusted cumulative catch recorded at
the Talkeetna stationary outmigrant traps,1983 and 1984.
82
-
~,
-
--
-
-
-
-
The downstream movement of age 1+coho salmon past Talkeetna Station was
approximately two weeks later in 1984 than in 1983 while the rates of
movement were fairly stabl~throughout both seasons (Fig.51).
4.2.2 Freshwater life history
Most coho salmon juveniles spend one or more years in the Susitna River
before migrating to the ocean.Analysis of scales from returning adults
indicate that most juvenile coho outmigrate as either age 1+or age 2+
but the proportion of each age c1ass has varied between years (ADF&G
1982;ADF&G 1983;Barrett et al.1984;Barrett et al.1985).
Coho salmon in the middle Susitna River spawn almost exclusively in the
tributaries.The fry,after emergence,re.ar in their natal tributaries
or enter the mainstem river in search of suitable habitats.Outmigrant
trap data collected at Talkeetna Station have shown a downstream redis-
tribution of juvenile coho occurring throughout the open-water season.
During the fall,coho fry move into tributaries,sloughs,beaver ponds,
or other habitats to overwinter.Similar redistributions of juvenile
coho were observed by Delaney and Wadman (l979)and by Tschaplinski and
Hartman (l983).
Trap catches recorded at Talkeetna Station during 1982 and 1984 showed
that hi gh catches of age 0+and 1+juveni 1e coho occurred duri ng
September or early October.It was presumed that these fish were
redistributing to habitats in the lower river to overwinter,but the
data collected at Flathorn Station in 1984 indicate that a portion of
these fish may migrate to the ocean during the fall (Fig.22).
4.2.3 Growth
The change in mean 1ength for age 0+coho by samp 1i ng peri od for the
combined data collected at the Talkeetna Station outmigrant traps during
1982,1983,and 1984 is presented in Fig.52.Coho salmon in the middle
river emerge from the gravel at approximately 35 mm and grow to 45 mm by
early July.By the end of the open-water season,coho fry have obtained
a mean length of approximately 68 rnm.Throughout the season,age 0+
coho in the lower river averaged at least five millimeters larger than
fish collected in the middl~river (Fig.26).
Age 1+coho sa 1mon in the mi ddl e ri ver also showed a steady growth
through the season (Fig.53)increasing approximately 45 mm between late
May and early October.Similar to age 0+coho,age 1+coho collected in
the lower river averaged larger than fish captured in the middle river
reach (Fig.27).
The downstream redistribution (as shown by the cumulative biomass)of
juvenile coho salmon in the Susitna River by age class during 1984
averaged one to two weeks later than the redistribution of the total
number of individuals recorded at both the Talkeetna and Flathorn
stations outmigrant traps (Fig.54).The difference between the cumu-
lative biomass movement and the movement of total numbers of fish
results from the growth of juvenile coho occurring during the open-water
83
1983 &1984 TALKEETNA COHO 1+
-
100
90
80
IJ.J 70>i=
:5 60
::l
::i
::l 50<.:l
t-
Zw 40Q
0::
Wa.30
20
10
0
-
.....
Figure 51.Coho salmon (age 1+)adjusted cumulative catch recorded at
the Talkeetna stationary outmigrant traps,1983 and 1984.
84
-
-
.,~'
COHO 0+MEAI'J LENGTH
70
68
66
·64
62 (>Maxlmum...-..
:::E 60
:::.i:......,58
:::c 56I-Combined
C)z 54 <>+w Minimum +...J 52
...Jg 50 +
l-48 +
z 46
L5 44 (>:::.i:
42
40
38
36
34
L MAY E JUNE L JUNE E JULY L JULY E AUG L AUG E SEP L SEP
SAMPLING PERIOD OCT
!~'
.-.
Figure 52.Coho salmon (age 0+)mean length and range of mean lengths
by sampling period recorded at the Talkeetna stationary
outmigrant traps during 1982,1983,and 1984.
"
85
60 +-----,r-------'--,-----.,----,------r---,-----r----1
L MAY E JUNE L JUNE E JULY L JULY E ,~UG L AUG E SEP L SEP
140 I
·130 J
::t:110
l-
e)zw
..J 100
..J
«:(
I-o
I-90
z
l5
2 80
70
COHO 1+M~AN LENGTH
SAMPLING PERIOD
j
oel ~I
Figure 53.Coho salmon (age 1+)mean length and range of mean lengths
by sampling period recorded at the Talkeetna stationary
outmigrant traps during 1982,1983,and 1984.
86
-
}1 )1 1 l )1 J )1 )1 j ]
1984 TALKEETNA COHO 1+
100
90 -,
"k80-l
~70 i
j 60::::I
:::i
:J 500.-
%
W 400
0:
Wn.30
20
10
COHO 0+1984 TALKEETNA
~o
10
50
40
90
20
70
80
30
100 I ~,
~
3
:J::;
:Jo.-zwo
0:wn.
O~I ,ii,I •I
l MAY E JUNE l JUNE E JULY l JULY E AUG l AUG E SEP l SEP E OCT
SAMPLING PERIOD
o I i I I Iii I I I
l MAY E JUNE l JUNE E JULY l JULY E AUG llWG E SEP l SEP E OCT
SAMPLING PERIOD
ex:>
-.....J
1984 FLATHORN COHO 0+
100
90
80
w 70
2
j 60
:J::;
:J
0 50
.-zw 400
0:wn.30
20
10 -l -~-----/
~
~
:J
:::i
:Jo.-z
tJ
0:
Wn.
1984 FLATHORN COHO 1 +
100 T-------,-------.:.--------......,
90
80
70
60
50
40
30
20
10
o I ii'I Iii I I
L MA.Y E JUNE L JI)NE E JULY L JULY E AUG L AUG E SEP L SEP E OCT
SAMPLlI'JG PERIOD
Figure 54.Coho salmon adjusted cumulative catch and biomass by age class recorded at Talkeetna and
Flathorn stations,1984.
season.The cumulative biomass curve is probably a better indicator of
the value of coho rearing habitat in the reach than is the cumulative
numbers curve.That is,the greater the amount of time the fry spend
rearing in a particular reach of river,the greater the benefit they
have gained from that particular reach.Not only are they larger,
having consumed more food in this reach,they also have a higher proba-
bility of survival than smaller fry and therefore are of more value.
Any management determination for these fish should consider the timing
of movement of total biomass in the river rather than formulating
actions only from the catch data.
4.3 Sockeye Salmon
4.3.1 Outmigration
The migration of sockeye salmon fry past Talkeetna Station during 1984
was similar to the timing recorded during 1983 (Fig.55).Fifty percent
of the total outmigration was recorded by the end of June during both
seasons.Sockeye fry were steadily redistributing to areas below the
sampling site from break-up through late August.Sampling of sloughs
and side channels in the middle river during the cold branding study
showed that sockeye fry were not actively outmigrating but were entering
habitats along the margins of the river as they moved downstream.The
fry probably remain at these sites until (1)they are displaced by flows
or density interactions,(2)adequate food supplies are no longer
available,(3)the habitats become otherwise unsuitable,or (4)the
critical size is reached.
The tendency of sockeye fry to ori ent along the banks of the ri ver
during their downstream migration was observed at Flathorn Station where
59%of the total sockeye fry collected in the mobile trap were captured
at bank transect points.
The rates of downstream movement for coded wi re tagged sockeye fry
during 1984 showed that fry in the middle river,after tagging,spent an
average of 35 days (range from 0 to 109 days)in the middle river before
migrating past Talkeetna Station.
4.3.2 Freshwater life history
Outmigrant trap data collected at Talkeetna Station during the past
three seasons (1982-1984)show that a 1arge number of sockeye fry
migrate out of this reach as age 0+fish,but scale analysis of adult
sockeye coll ected at Curry Stati on showed that thi sage cl ass repre-
sented only 6.4%of the returning adults during 1984 (Barrett et al.
1985)•The 1a rgest percentage of returni ng adul ts were compri sed of
fish which had spent one winter in freshwater before going to the ocean.
There fore,the majority of age 0+fry from the middle river either rear
in the lower river or have a low survival rate.
-
"'"
Bernard et ale (1983)analyzed scale patterns from samples of adult
sockeye sa 1mon collected from fOIJ r di fferent sites in the Sus i tna Ri ver
watershed in an attempt to delineate the differences in scale patterns ..."
'88
"'"'
1983 &
100
90
80
w 70
~
~
:3 60
=:J
2
=:J 50(,)
~zw 40<Ja::w
~30
20
10
0
1 984 TALKEETNA.SOCKEYE 0+
--
.....
Figure 55.Sockeye salmon (age 0+)adjusted cumulative catch recorded
at the Talkeetna stationary outmigrant traps,1983 and 1984.
89
for the period of freshwater growth for each of the sites.Samples were
collected from escapements of sockeye salmon at Curry and Talkeetna
stations on the Susitna River,from the outlet of Larson Lake on the
Talkeetna River,and from the Tokositna River which is a tributary to
the Chulitna River.One of the results of this study was that sockeye
salmon scale samples collected from the Susitna River sites could not be
-dtstinguished from those of Tokositna or Larson Lake fish.
Six hypotheses were suggested by Bernard et ale (1983)for the lack of
unique differences in the scale patterns between Susitna River fish and
those collected from the other sites.In general,these hypotheses can
be placed into two groups:1)The Susitna River fish are a unique stock
but the fry rear in environments similar to those found in Larson Lake
or the Tokositna River,or 2)the sockeye salmon spawning in the Susitna
River are strays from either the Talkeetna or Chulitna watersheds and
their fry move into these watersheds to rear or are displaced downstream
and enter the ocean as age 0+fish.If these fish enter the ocean as
age 0+fish,scale analysis of returning adults indicates that survival
of these fish is very low.
However,the study conducted by Bernard et ale was based on the
assumption that sockeye fry did not rear in the middle Susitna River.
Data collected at the Talkeetna Station outmigrant traps during the past
three years have shown that a si gnifi cant amount of sockeye rea ri ng
occurs in this reach.The Susitna River samples collected by Bernard et
ale were taken at the fishwheel sites rather than at the spawning
grounds.Barrett (1984)has pointed out that a high percentage of these
fish (30%estimated in 1983)are milling fish which eventually spawned
in areas other than the middle Susitna River.Comparisons of the scales
of fish collected at the spawning grounds in these rivers may provide
more accurate differentiation of Susitna River fish from those observed
in the Talkeetna and Chulitna rivers.Also,Bernard et ale analyzed
scales from only 1.3 age fish (European formula);Barrett et ale (1984)
have shown that multiple age classes are present in the middle Susitna
River escapements.Juvenile sockeye salmon outmigrating from Larson
Lake predominantly spend two winters in freshwater before outmigrating
from the lake as smolts (Mar~uson 1985).
Although it is possible that sockeye salmon which spawn in the middle
reach of the Susitna River are strays from the stocks originating from
the Ta"lkeetna and Chulitna rivers,it is more likely that the Susitna
sockeye are a separate and viable stock.However,the amount of rearing
habitat in this reach is limited.The age 0+fish which outmigrate from
the middle reach of the Susitna probably imprint to their natal areas in
the early stages after hatching and then later distribute to suitable
habitats throughout the expanse of the lower river to overwinter.These
fish then enter the ocean during their second year of life and finally
return to their natal areas as adults to spawn.Also,a limited amount
of overwintering by sockeye fry in the middle reach does occur,as shown
by the capture of age 1+fry at Talkeetna Station.
More definitive information on the viability of middle Susitna River
sockeye may be obtained through the continued monitoring of returning
90
-
-
-
,-
r
!
I
adul ts at the fi shwheel sites and duri ng spawni ng ground surveys to
collect returning fish which were marked with coded wire tags as fry.
Juvenile sockeye salmon life histories in the middle Susitna River can
be grouped into three categories.The first group includes those fish
which spend their entire freshwater period rearing in the middle river,
overwintering in this reach and then migrating to the ocean during the
spring of their second year (age 1+).The second group includes those
fish which rear for a portion (one to four months)of their first summer
in,the middle river and then migrate to areas below the Chulitna River
confluence to overwinter and then enter the ocean during the spring of
their second year.The third group of juvenile sockeye spend a portion
of their first surrmer rearing in the middle river and then begin a
downstream mi gration,eventua 11y enteri ng the marine envi ronment duri ng
their first summer or fall as age 0+fish.
Currently,it is not known what contribution each group provides to the
total outmigration of juvenile sockeye from the middle Susitna River.
Outmigrant trap data collected at Flathorn Station during 1984 collected
a large number of age 0+sockeye;most of these fish were probably
destined for the ocean as 0+fish.
Although trap catches of age 1+sockeye at Talkeetna Station have been
low (only 19 fish during 1984),it is possible that this age class
(group 1)migrates out of the middle river prior to the initiation of
spring sampl ing or that they differ from their age 0+counterparts in
that they migrate further from shore and are not intercepted by the bank
traps in proportion to their relative abundance.Also,the bank traps
are less effective at capturing these larger fish (Roth et ale 1984)~
4.3.3 Estimate of population and survival
The estimated 1983-1984 egg-to-emergent fry survival rate of 17%,based
on an estimated 299,000 sockeye fry produced dur-ing 1984 from the
approximately 1,900 adults which migrated past Curry Station in 1983,
was lower than the 1982-1983 estimate of 42%,based on the 1,300 adult
sockeye past Curry Station during 1983 which produced an estimated
575,000 fry.The substantial differences between the estimates of
survival in 1983 and 1984 are due in part to the data used in the
calculations.During both years,survival rates were calculated by
divi ding the number of fry produced by the estimated number of eggs
carried by adults past Curry Station during the previous season.
Ba rrett et a 1.(1984)poi nted out that the estimates provi ded at Curry
Station represent only the fish which passed this site but do not
necessarily reflect the number of fish which actually spawned in the
middle river reach.As sockeye salmon in this reach are almost strictly
slough spawners,~ore reasonable estimates were calculated by Barrett et
ale (1984)by comparing slough escapement counts to observation life
data to estimate the total slough escapement in the middle river.
During 1983,this comparison provided an estimate that 1,060 adult
sockeye had spawned in sloughs in the middle river.The stream 1 ife
data were then used to provi de comparable estimates for 1982 showi ng
approximately 1,500 sockeye had spawned in the sloughs that year.These
91
data were then used to recalculate the sockeye egg-to-outmigrant sur-
vival rates.A survival rate of 22%was estimated for 1983-1984 and a
rate of 35%was calculated for 1982-1983.
4.3.4 Growth
The weekly growth rate for sockeye fry which were coded wire tagged in
1983 and 1984 (Fig.56)most accurately represent the growth rates for
sockeye salmon fry in the middle river because the dates of release and
recovery and the mean lengths for each period were known.
These fry grew approximately three mi 11 imeters each week unti 1 they
reached a critical size and then the growth rates slowed (Fig.56).
Schmi dt (1984)postul ated that the cessati on of sockeye growth after
reaching a certain size was associated with evolved behavioral patterns
and morphological changes.Schmidt suggested that the sockeye fry were
able to rear in the middle river habitats for part of the summer but
began a downstream migration in search of plankton rich environments
after reaching a critical size.The small number of habitats which
provide this type of environment in areas associated with the Susitna
River is a major factor in controlling the production of sockeye in the
middle river.
A comparison of the length data collected at Talkeetna Station during
1982,1983,and 1984 and during the previous winter studies above
Tal keetna in 1981 and 1982 show that Susitna River sockeye average
approximately 32 mm total length at emergence,35 mm by early June,and
have increased to approximately 50 mm by late July (Fig.57).From late
July through August,no significant growth was observed for sockeye fry
collected at Talkeetna Station,indicating that the critical size
postulated by Schmidt (1984)may be 50 to 55 mm in the middle river.
The apparent growth of sockeye fry after late August (Fig.57)is
attributed to the collection of fish which had continued rearing in the
small number of sites in the middle river which provide the necessary
food and habitat requi rements.These fi sh were probably forced to
migrate out of these areas as water levels and available habitat
decreased.The number of sockeye collected after late August represent
less than 2%of the total outmigration of age 0+fish from this reach.
A comparison of the downstream redistribution of sockeye salmon in the
Susitna River by age class during 1984 as the percent cumulative of the
total catches recorded at Tal keetna and Fl athorn stations compared to
the calculated percent cumulative biomass moving past these sites,
indicated that the redistribution by weight of sockeye in the Susitna
River was up to two weeks later than the redistribution observed when
comparing only total numbers of fish (Fig.58).
Age 1+sockeye salmon collected during 1984 averaged apprOXimately
75 mm.Thi sis apprOXimately 10 mm longer than the average 1ength of
sockeye fry collected at the end of the open-water season indicating
that the fry are growing through the winter and early spring prior to
outmigrating as smolts.The average length of age 1+sockeye migrating
out of the Susitna River was approximately 10 mm smaller than the same
92
....1
I
~!
CWT SOCKEYE MEAN LENGTH
10
56
987
7~•
85432
58
58
54-52:::E
::E 50-:J:
I-48
CJ
Z 48W
-I
-I 44
<C
I-42O·
I-40Z
<C 38W
:::E
38
34
32
1
-.
.....
t"'"
WEEKS BETWEEN RELEASE AND RECOVERY
Figure 56.Mean length of coded wire tagged sockeye salmon fry at
recovery sites in the middle reach of the Susitna River by
week,1984.Number of fish shown by data points.
"""!
93
SOCKEYE 0+MEAN LEf'.JGTH
70,----------------------------,
65
.,..,
5 55
z
~
....J 50g
~45
CIz
L5~40
35
Maximum
Q
r.Io-_--1°~--____:combined+
++
Minimum
+
30 +-----,----,...----r----.,.---.....------r------.----'--l
L MAY E JUNE L JUNE E JULY L JULY E AUG L AUG E SEP L SEP
SAMPlI NG PERIOD OCT
Figure 57.Sockeye salmon (age 0+)mean length and range of mean lengths
by sampling period recorded at the Talkeetna stationary
outmigrant traps during 1982,1983,and 1984.
94
h,mll
])1 1 1 }1 t 1 J )
1984 TALKEETNA SOCKEYE 0+1984 TALKEETNA SOCKEYE 1 +
o r •I I •Ii',I
l MAY E JUNE l JUNE E JULY l JULY E AUG l AUG E SEP L SEP E OCT
SAMPLING PERIOD
100
90
80
~70
~60::I
]i
::I 500...Z
hi 400
lr
hi
0-30
20
10
~.~~
100
95
90
~~65
::I
::IIi
::I 600...Z
hi
0 75Ii'
hi
Q.
70
65
60 I I •i I I Iii I
L MAY E JUNE L JUNE E JULY L JULY E AUG l AUG E SEP L SEP E OCT
SAMPLING PERIOD
1.0
U"l
~
::I
::IIi
::Io...
Z
hio
Ii'
hi
Q.
1984 FLATHORN SOCKEYE 0+
100-/~.T
90
80
70
60
50
40
30
20
10
~f i i J f I ,I j
L MAY E JUNE L JUNE E JULY L ,JULY E AUG l AUG E SEP L SEP E OCT
SAMPLING PERIOD
~
~
~
::Io
!Z
~
Ii'
hi
Q.
1984 FLATHORN SOCKEYE 1 +
100 /~I
90
60
70
60
50
40
30
20
10 ,I Iii J I i I I
L M".Y E ,jUNE l ,JUNE (JULY L ,JULY E AUG L AUG E SEP L SEP E OCT
SAMPLlI~G PERIOD
Figure 58.Sockeye salmon adjusted cumulative catch and biomass by age class recorded at Talkeetna and
Flathorn stations,1984.
age fish outmigrating during 1984 from Larson Lake,a major spawning
site in the Talkeetna River (Marcuson 1985).
4.4 Chum Salmon
4.4.1 Outmigration
The migration of chum salmon fry past Talkeetna Station during 1984 was
similar to the timing recorded during 1983 (Fig.59).Fifty percent of
the total outmigration past this site had occurred by mid June and over
95%of the chum fry had migrated out of the middle river by mid July.
At Flathorn Station,the peak chum fry outmigration also occurred in mid
June during 1984.
4.4.2 Freshwater life history
Chum salmon fry spend from one to eight weeks in the middle Susitna
River before outmigrating from the reach.A portion of the population
of chum fry probably begins outmigration shortly after emergence whereas
other fry stay in the river to rear for a few weeks before outmigrating.
It is not possible to determine the percentage which each group provides
because of the difficulty in sampling outmigrant fishes prior to and
during breakup,a time when many newly emerged chum fry may outmigrate.
4.4.3 Estimates of population and survival
The estimated 1982-1984 egg-to-outmigrant fry survival rate of 16%,
based on an estimated 2,039,000 chum sa 1mon fry produced duri ng 1984
from the approximately 21,100 adults past Curry Station in 1983,was
similar to the estimated 1982-1983 rate of 14%,based on the 17,600
adult chum which passed Curry Station during 1982 which produced an
estimated 3,322,000 fry.
The calculation of survival rates is based upon the estimated number of
parent spawners which is difficult to obtain because of the extent of
tributary spawning by chum salmon.Also a substantial percentage of
chum salmon passing Curry Station are milling fish which eventually
spawn below this site,and although estimates have been provided for
1982 and 1983 (Barrett 1984),these percentages are,at best,only
indicators of the amount of chum salmon milling occurring.As these
estimates have a large influence on the calculated rates of survival,
the rates presented for 1983 and 1984 should be used to compare differ-
ences between years rather than absol ute val ues of mi ddl e river chum
salmon survival.
4.4.4 Growth
Many chum fry from the middle reach move downstream at lengths not much
longer than their emergence length (less than 35 mm),but there are also
many that spend several weeks in freshwater and attain lengths of over
60 mm,an increase of more than 20 mm.The mean 1ength by one-week
periods of recovery after release for coded wire tagged chum fry which
were tagged.and recaptured during 1983 and 1984 (Fig.60)most
96
-
-
,~
100 i------------------:::::::::=;;;;;=--!!!!!!!!!!::===------,
,-
~,
-1 983 &1 984 TALKEETNA CHUM FRY
.-
.....
I
90
80
w 70>~~60
::::J
:2
~50
I-
Zw 40
<J
Cl::wa.30
20
10
0-1=---1--------+----------+--------,
AUG
-,,
-
Figure 59.Chum salmon fry adjusted cumul ati ve catch recorded at the,
Talkeetna stationary outmigrant traps,1983 and 1984.
97
CWT CHUM MEAN LENGTH -
5 to 10 11 to 15 1 5 to 20 2 1 to 25
DAYS BETWEEN RELEASE AND RECOVERY
(Grouped by 5 Day Period)
5251~
50 ~-:E 49:E-:::t 48
I-
0 47Z
W
..J 46
..J<45I-
0
I-44
Z<43W
:E
42
41
40
o to 5
19 15
2
25 to 29
Figure 60.Mean length of coded wire tagged chum salmon fry at recovery
sites in the middle reach of the Susitna River by 5 day
period,1984.Number of fish shown by data points.
98
....
!
-
accurately represent the growth rates of chum fry in the middle river
because the dates of release and recovery and the lengths for the fish
for each period were known.The 15%increase in length by fish captured
more than 20 days after release (mean length significantly different
from release length at 95%confidence level)would correspond to an even
larger percentage increase in weight.The chum fry greater than 50 mm
in length collected.during the three years of this program had a
noticeably greater girth than shorter fry.Similarly,chum fry in the
Tokachi River of Japan grew 1.0 to 1.3 times in length and 1.0 to 3.1
times in weight during April and May (Kaeriyama et al.1978).
These data indicate that the chum fry in the middle river are actively
rearing after emergence.Chum fry rearing was al so shown from the
analysis of stomach samples from tagged fish recovered at Talkeetna
Station during 1983.These fish had been eating various life stages of
mayflies,stoneflies,blackflies,midges,and other dipterans.
4.5 Pink Salmon
4.5.1 Outmigration
The rates of downstream migration of pink salmon fry past Talkeetna
Station for 1983 and 1984 were very similar between the two years but
the timing was approximately two weeks later in 1984 than in 1983 (Fig.
61).Differences in spawning times,winter temperatures,and spring
breakup account for the differences in timing between the two years.
The low catches of juvenile pink salmon recorded at Talkeetna Station
during the past three seasons is due to the pattern and timing of
outmigration.Pink salmon fry outmigrate shortly after emergence and
most of the fry probably have migrated past the traps prior to the
initiation of sampling.Those fish which are still in the middle river
after breakUp appear to outmigrate in association with center channels
and high velocities.I
4.5.2 Freshwater life history
Pink salmon fry in the Susitna River outmigrate to the ocean shortly
after emergence during a relatively short (in comparison to the other
species)timing window whose boundaries are determined by the timing of
spawning the previous season,incubation temperature,and the level of
discharge.The pink fry collected during 1984 averaged approximately 35
111m which is similar to their mean length at emergence.A few pink fry
which ranged in length from 40 to 50 mm were collected,indiCating that
a small percentage of fry may be feeding for a short period of time in
freshwater before outmigrating to the ocean.
99
1983 &1984 TALKEETNA PINK FRY ~,
100 _.
90
80 -
w 70>i="""1
j
I
60
::J
::i
::J 50 ~,()
~zw 40()
c:r::1"""1wa.30
20
10
""'"'0 JUL AUG
Figure 61.Pink salmon fry adjusted cumulative catch recorded at the Talkeetna
stationary outmigrant traps,1983 and 1984.
-
100
"...
-
-
-
5.0 CONTRIBUTORS
Resident and Juvenile Anadromous Fish
Project Leader
(Acting Project Leader,
Jan.to Jun.1985)
Task Leader
Talkeetna Station
Flathorn Station
Coded Wire Tagging
Cold Branding
Deshka River Weir
Data Base Management
Jolly-Seber Model
Analysis of Stomach Samples
101
Dana Schmidt
Stephen Hale
Kent Roth
Chuck Blaney
(Crew leader)
Albert Badgley
Patricia Harris
Tom Crowe
Li nda Soquet
Larry Dugan
(Task leader)
Roger Harding
(Crew leader
Jeff Bigler
Jim Anderson
Diane Roche
James Gruber
Dan Sharp
Linda Soquet
Doug Patrick
Aimee Weseman
Mi ke Stratton
(Task leader)
Dan Gray
Dave Sterritt
(Crew leader)
John McDonell
Allen Bingham
-(Project leader)
Kathrin Zosel
Alice Freeman
Chuck Mi 11 er
Donna Buchholz
Stephen Hale
Dana Schmidt
Stephen Hale
Tim Hansen
(Crew leader)
Craig Richards
Drafting
Typing
Text
Report Coordinators and
Editors
iidlll,
102
Carol R.Hepler
Skeers Word Processing
Kent Roth
Mike Stratton
Stephen Hale
Drew Crawford
Paul Suchanek
-
~:
"""
-
.....
,....
-
-
-
6.0 ACKNOWLEDGEMENTS
Funding for this study was provided by the Alaska Power Authority.
We are grateful to the various consulting agencies working on the
Susitna Hydroelectric Project for helpful comments on a draft of this
report.
103
Adult
Alaska
Report
.-
-
-
-
-
7.0 LITERATURE CITED
Alaska Department of Fish and Game (ADF&G).1981a.Adult anadromous
fisheries project (June -September 1981).Phase 1 final draft
report.Subtask 7.10.Alaska Department of Fish and Game Susitna
Hydro Aquatic Studies.Anchorage,Alaska.
•1981b.Juvenile anadromous fish study on the Lower Susitna----~River (November 1980 -October 1981).Phase 1 final draft report.
Subtask 7.10.Alaska Department of Fish and Game Susitna Hydro
Aquatic Studies.Anchorage,Alaska.
•1983a.Adult anadromous fish studies,1982.Susitna Hydro
-----aquatic studies phase II final report.Volume 2.Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies.
Anchorage,Alaska.
•1983b.Resident and juvenile anadromous fish studies on the
--";:;"Susitna River below Devil Canyon,1982.Susitna Hydro aquatic
studies phase II basic data report.Volume 3.Alaska Department
of Fish and Game Susitna Hydro Aquatic Studies,Anchorage,Alaska.
•1985.(unpublished draft).Resident and juvenile anadromous
--studies.Procedures manual draft (May 1984 -April 1985).Susitna
Hydro Aquatic Studies Program,Alaska Department of Fish and Game.
Anchorage,Alaska.
Barrett,B.M.1984.Summary of abundance and distribution of adult
salmon in Susitna river sub-bas,ins.Presented at Aquatic Habitat
Workshop No.1,Susitna Hydroel ectri c Project,February 15,1984.
Anchorage,Alaska,
Barrett,B.M.,F.M.Thompson,and S.N.Wick (eds.).1984.
anadromous fish investigations:May -October 1983.
Department of Fish and Game Susitna Hydro Aquatic Studies.
No.1.Anchorage,Alaska.
Bernard,D.R.,G.01 iver,W.Goshert,and B.Cross.1983.Comparison
of scale patterns from sockeye salmon sampled from different rivers
within the Susitna River watershed in 1982.Alaska Department of
Fish and Game,Division of Commercial Fisheries,Statewide Stock
Biology Group.Anchorage,Alaska.
Chapman,D.G.1951.Some properties of the hypergeomtric distribution
with appl ications to zoological sample censuses.University of
California Publication Statistics 1:131-160.
Delaney,K.,K.Hepler,and K.Roth.1981.Deshka River chinook and
coho salmon st~dy.Alaska Department of Fish and Game,Division of
Sport Fish.Federal Aid in Fish Restoration,Project AFS-49,Vol.
22.
104
~,
Delaney,K.J.,and R.Wadman.1979.Little Susitna River juvenile
chinook and coho salmon study.Alaska Department of Fish and Game,
Division of Sport Fish.Anchorage,Alaska.~
Dixon,W.J.,M.B.Brown,L.Engelman,J.W.Frane,M.A.Hill,R.1.
.Jennrich,and J.D.Toporek (eds.).1981.BMDP Statistical
Software 1981.University of California.Berkley,California.
Healy,M.e.,and W.R.Heard.1984.Inter-and intra-population
variation in the fecundity of chinook salmon (Oncorh*nchus
tshawytscha)and its relevance to the life history t eory.
Canadian Journal of Fisheries and Aquatic Sciences 41:476-483.
Kaeriyama,M.,S.Sato,and A.Kobayashi.1978.Studies on the growth
and feeding habit of the chum salmon fry during seaward migration
in the Tokach i Ri ver system.1.I nfl uence of thaw on the growth
and feeding habit of the fry.Sci.Rep.Hokkaido Salmon Hatchery
32:27-41.(in Japanese,English summary)..
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.
Kubik,S.1967.Population studies of anadromous species with emphasis
on upper Cook Inlet drainage.Alaska Department of Fish and Game,
Division of Sport Fish.Federal Aid in Fish Restoration,1966-
1967,Project Report 8:117-128.
Kubik,S.,and K.Delaney.1980.Inventory and cataloging of sport
fish waters of the lower Sus·itna River and central Cook Inlet
drainages.Alaska Department of Fish and Game.Federal Aid in
Fi sh Restorat ion,Annual Report of Progress,1979-1980,Proj ect
F-9-12,21(G-I-H).
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).
Manly,B.F.J.1984.Obtaining confidence limits on parameters of the
Jolly-Seber model for capture-recapture data.Biometrics
40:749-758.
Marcuson,P.1985.Larson Lake project progress report.Cook Inlet
Aquaculture Association,Anchorage,Alaska.
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
Marine Fisheries Service CIRC-366.
105
-
~,
-
-
-
-
-
.-I
Moberly,S.A.,R.Miller,K.Crandall,and S.Bates.1977.Mark-tag
manual for salmon.Alaska Department of Fish and Game.Fisheries
Rehabilitation and Enhancement Division.
Morrow,J.E.1980.The freshwater fishes of Alaska.Alaska Northwest
Publishing Company,Anchorage,Alaska.
Raleigh,R.F.,J.B.McLaren,and D.R.Graff.1973.Effects of topical
location,branding techniques and changes in hue on recognition of
cold brands in Centrarchid and Salmonid fish.Transactions of the
American Fisheries.Society 102:637-641.
Richards,K.1979.Aspects of the juvenile life history of spring
chinook salmon (Oncorhynchus tshawytscha)in Deshka River,Alaska
determined from adult scale analysis and migrant trapping.M.S.
Thesis.Oregon State University,Corvallis,Oregon.
Ricker,W.E.1975.Computation and interpretation of biological
statistics of fish populations.Bulletin of the Fisheries Research
Board of Canada.191.
Roth,K.J.,D.C.Gray,and D.C.Schmidt.1984.The outmigration of
juvenile salmon from the Susitna River above the Chulitna River
confluence.Part loin D.C.Schmidt,S.S.Hale,D.L.Crawford,and
P.M.Suchanek (eds--:).1984.Resident and juvenile anadromous
fish investigations (May -October 1983).Susitna Hydro Aquatic
Studies.Report No.2.Alaska Department of Fish and Game.
Anchorage,Alaska.
Schaefer,M.B.1951.Estimation of the size of animal populations by
marking exper.iments.United States Fish and Wildlife Service,
Fisheries Bulletin 52:189-203.
Schmidt,D.C.1984.Riverine rearing of slough spawned sockeye salmon
in the Susitna River.Paper presented at annual meeting of the
American Fisheries Society,Alaska Chapter.November,1984.
Juneau,Alaska.
Schmidt,D.C.,S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.).
1984.Resident and juvenile anadromous fish investigations (May -
October 1983).Alaska Department of Fish and Game Susitna Hydro
Aquatic Studies.Report No.2.Anchorage,Alaska.
Trautman,M.B.1973.A guide to the collection and identification of
pre-smolt Pacific salmon in Alaska with an illustrated key.
National Oceanographic and Atmospheric Administration Technical
Memorandum.NMFS ABFL-2.
Tschaplinski,P.J.,and G.F.Hartman.1983.Winter distribution of
juvenile coho salmon (Oncorhynchus kisutch)before and after
logging in Carnation Creek,British Columbia,and some implications
for overwinter survival.Canadian Journal of Fisheries and Aquatic
Sciences 40:452-461.
106
Vining,L.J.,J.So Blakely,and GoM.Freeman.198.50 An evaluation of
the incubation 1 ife-phase of chum salmon in the middle Susitna
River,Alaska.Winter Aquatic Investigations:September,1983 -
May 1984.Susitna Hydro Aquatic Studies.Report No.50 Alaska
Department of Fish and Game,Anchorage,Alaska.
107
-
-
-
-
-
APPENDIX A
JUVENILE SALMON CATCH AND LENGTH DATA,1984
Appendix Table A-1.Weir catches of juvenile chinook and coho salmon on the Deshka River,
May 10 through September 19,1984.-
Chinook Coho
Tributary Hours Daily Catch Dai ly Catch
Date River Mile Fished Catch Per Hour Catch Per Hour
,-May 10 2.0 21.5 2 0.1 0 0.0
12 2.0 15.0 9 0.6 1 0.1
13 2.0 21.0 3 0.1 0 0.0
27 5.0 12.0 50 4.2 1 0.1
28 5.0 12.5 7 0.6 0 0.0
r-29 4.5 12.5 3 0.2 0 0.0
31 5.0 12.0 4 0.3 0 0.0
June 1 5.0 12.5 21 1.7 0 0.0
~.21 5.0 11.5 1 0.1 0 0.0
22 5.0 21.5 3 0.1 0 0.0
July 11 2.5 14.5 209 14.4 5 0.3
12 2.5 24.0 144 6.0 2 0.1
13 2.5 24.0 268 11.2 3 0.1
14 2.5 23.5 186 7.9 4 0.2
15 2.5 24.0 27 1 .1 0 0.0
16 2.5 24.0 130 5.4 1 0.0
~25 2.5 15.0 318 21.2 21 1.4
26 2.5 24.0 149 6.2 8 0.3
31 2.5 20.0 168 8.4 4 0.2
August 13 2.5 14.0 45 3.2 15 1 .1
14 2.5 23.0 4 0.2 2 0.1
15 2.5 23.0 5 0.2 5 0.2
16 2.5 23.0 27 1.2 12 0.5
31 2.0 21.5 5 0.2 22 1.0
September 11 1.5 13.5 1 0.1 0 0.0
12 1.5 23.0 6 0.3 0 0.0
13 1.5 23.0 8 0.3 1 0.0
14 1.5 23.0 2 0.1 0 0.0
15 2.5 18.0 1 0.1 2 0.1
16 2.5 24.0 0 0.0 6 0.3
17 2.5 24.0 1 0.0 0 0.0
18 2.5 23.0 1 0.0 2 0.1-
Season Totals 621.0 1,808 2.9 117 0.2
,-
A-I
Appendix Table A-2.Results of incidental minnow trapping in the Deshka River.1984.
-
Chinook Coho
Tributary Number Catch Catch
River Hours of Daily Per Daily Per
Date Mile Fished Traps Catch Trap Catch Trap
June 21 5.5 16 6 56 9.3 14 2.3 .....
August 28 2.5 9 o 6 15 2.5 48 8.0
29 2.7 7 7 23 3.3 50 7.1
September 17 5.5 24 4 20 5.0 4 1.0 -,
October 10 2.2 24 2 1 0.5 2 1.0
10 6.0 24 4 30 7.5 4 1.0
11 5.0 27 7 23 3.3 21 3.0
13 2.0 to 6.0 54 5 2 0.4 10 2.0
14 2.0 to 6.0 28 5 1 0.2 4 0.8
15 4.0 24 5 41 8.2 9 1.8
~I
Season Totals 51 212 4.2 166 3.3
-,
A-2
))}1 1 --,1 1 ~-~1 1 J 1
Appendix Table A-3.Mean length and range of lengths for age 0+chinook salmon by sampling period in the lower reach of the Susitna
River,1984.
Lower Susitna
Flathorn Station Deshka River JAHS Sitesa
Sampling
Period Mean Range of Mean Range of Mean Range of
n Length Lengths n Length Lengths n Length Lengths
May 0 --77 42.7 36-49 b
June 1-15 24 56.6 40-67 21 42.4 40-46 74 48.5 34-63
June 16-30 374 58.5 39-74 56 55.7 46-69 63 52.0 36-70
July 1-15 357 62.0 40-84 236 66.8 52-83 84 54.5 39-74
July 16-31 436 64.3 43-88 201 69.7 52-93 171 58.1 39-80
August 1-15 189 66.6 47-89 53 74.4 60-91 330 58.9 40-82
»August 16-31 193 72.7 46-94 65 71.7 55-89 238 61.5 42-94Iw
September 1-15 8 77.3 68-84 15 77.9 69-88 52 66.8 52-95
September 16 -October 15 10 78.7 68-95 102 76.0 68-85 53 73.2 51-92
-a Includes all mainstem,slough and side channel sites sampled during the JAHS study in the Susitna River between Cook Inlet and the
Chulitna River confluence.
b Not sampled.
Appendix Table A-4.Mean length and range of lengths for age 0+chinook salmon by sampling period in the Talkeetna River and the
middle reach of the Susitna River,1984.
Talkeetna River Talkeetna Station
Middle Susitng
Indian RiverMarkingSites
Sampling
Period Mean Range of Mean Range of Mean Range of Mean Range of
n Length Lengths n Length Lenghts n Length Lengths Ii Length Lengths
May b --2 55.5 53-58 60 40.8 35-45 b
June 1-15 0 --54 48.6 36-66 b - -
b
June 16-30 26 52.2 43-64 475 53.0 37-70 b --b
July 1-15 159"56.0 44-70 538 56.2 38-75 100 47.8 38-67 50 48.9 42-64
Jul Y 16-31 155 56.1 40-74 1131 55.5 37-80 50 52.2 42-69 50 54.9 47-67
August 1-15 257 60.7 44-84 748 57.9 40-90 50 52.4 40-77 100 58.8 47-90
>August 16-31 114 65.2 51-84 612 59.5 39-95 100 56.1 43-72 100 61.1 49-80I.,t::.
September 1-15 0 --119 62.7 45-91 100 57.6 47-88 100 63.8 47-90
September 16 -October 15 b - -
13 60.8 51-90 200 61.0 45-90 300 65.5 50-89
-
a Includes all mainstem,slough,and side channel sites sampled during the coded wire tagging and cold branding studies in the
Susitna River between the Chulitna River confluence and Devil Canyon.
b Not sampled.
,J )J )1 J »)J J 1 ),)_J 1 ,)J
))1 1 1 "1 J -)..-1 -J l ).,1 1
Appendix Table A-5.Mean length and range of lengths for age 1+chinook salmon by sampling period in the Susitna River,1984.
Flathorn &Talkeetna
Flathorn Station Talkeetna Stations Stations Combined
Sampl i ng Mean Range of Mean Range of Mean Range of
Period n Length Lengths n Length Lengths n Length Lengths
May 11 79.7 67-105 209 77.9 61-101 220 78.0 61-105
Early June 104 89.1 70-122 126 89.6 71-112 230 89.7 70-122
Late June 101 85.2 75-122 335 88.4 71-107 436 87.7 71-122
Early July 17 94.1 86-113 218 85.7 76-117 235 86.3 76-117
Late July 4 97.5 95-102 96 87.7 81-115 100 88.1 81-115
Early August 8 98.6 90-113 1 91.0 91 9 97.8 90-113
Late August 2 96.0 95-97 0 - -
2 96.0 95-97
~
I
U1
Appendix Table A-6.Mean length and range of lengths for age 0+coho salmon by sampling period in the lower reach of the Susitna
River,1984.
Flathorn Station
Sampling
Period Mean Range of
n Length Lengths n
May 0 --0
June 1-15 10 42.7 32-60 0
June 16-30 19 48.7 32-64 0
July 1-15 11 49.3 36-65 0
July 16-31 38 58.6 44-73 21
August 1-15 30 62.1 49-79 19
August 16-31 181 66.8 40-89 59
:i-
&September 1-15 84 75.0 55-94 2
September 16 -October 15 67 75.1 57-94 29
Deshka River
Mean Range Of
Length Lengths
57.3 47-65
63.6 53-72
71.2 51-89
68.0 67-69
77.0 60-95
Lower Susitna
JAHS Sitesa
Mean Range of
n Length Lengths
b
18 40.9 33-50
9 46.2 34-61
26 50.7 35-65
33 50.2 37-65
45 49.6 41-68
71 59.1 40-85
59 62.2 49-86
105 66.7 49-95
a Includes all mainstem,slough and side channel sites sampled during the JAHS study in the Susitna River between Cook Inlet and the
Chulitna River confluence.
b Not sampled.
I J •,J .J J J .J J )t I "J ~J t ,
1 1 "1 ])]J -1 --1 )1 1 ]
Appendix Table A-7.Mean lengths,and range of lengths for age 0+coho salmon by sampling period in the middle reach of the Susitna
River,1984.
Talkeetna Station
Sampling
Period Mean Range of
n Length Lengths
May 35 39.7 35-46
June 1-15 40 39.6 30-51
June 16-30 156 43.9 31-58
July 1-15 242 47.8 32-63
July 16-31 439 51.8 33-69
August 1-15 221 54.1 41-74
)::-August 16-31 198 61.5 42-80
I
'-l September 1-15 212 60.5 42-85
September 16 -October 15 39 69.1 51-90
Middle Susitn9
Indian RiverMarkingSites
Mean Range of Mean Range of
n Length Lengths n Length Lengths
b --b
b --b
b --b
0 - -
62 38.0 34-51
0 - -
10 44.1 42-49
0 --80 48.0 39-58
38 50.8 39-62 46 49.0 42-61
41 56.8 40-70 90 50.9 44-64
5 59.4 48-76 166 55.1 44-73
a Includes all mainstem,slough,and side channel sites sampled during the coded wire tagging and cold branding studies in the Susitna
River between the Chulitna River confluence and Devil Canyon.
b Not sampled.
Appendix Table A-8.Mean length and range of lengths for age 1+coho salmon by sampling period in the lower reach of the Susitna
River,1984.
Lower Susitna
Flathorn Station Deshka River JAHS Sitesa
Sampling
Period Mean Range of Mean Range of Mean Range of
n Length Lengths n Length Lengths n Length Lengths
May 0 --5 69.8 58-89 b
June 1-15 7 87.4 62-110 0 --1 70 70
June 16-30 15 78.1 65-96 14 78.6 58-108 11 97.4 62-111
July 1-15 12 84.9 70-111 13 79.0 62-95 6 81.3 72-101
July 16-31 39 89.8 75-120 6 101.7 65-118 4 85.3 73-92
August 1-15 16 92.8 80-112 2 97.5 83-112 4 102.0 98-109
August 16-31 68 103.4 91-122 68 98.2 90-123 11 105.2 90-123
):0
I September 1-15 68 109.4 95-129 1 118.0 118 3 105.3 104-108co
September 16 -
October 15 53 112.9 95-133 31 111.8 92-134 4 112.0 99-110
a Includes all mainstem,slough and side channel sites sampled during the JAHS study in the Susitna River between Cook Inlet and the
Chulitna River confluence.
b Not sampled.
~-,I J J ),,~J I J J ~~.t J )~.)
J 1 J 1 J 1 )J 1 1 )1 I
Appendix Table A-9.Mean lengths,and range of lengths for age 1+coho salmon by sampling period in the middle reach of the Susitna
River,1984.
Middle Susitng
Marking Sites
Mean Range of
length Lengths
b
b
b
2 67.0 64-70
7 85.7 79-90
17 86.1 74-99
0
0
0
Talkeetna Station
Sampling
Period Mean Range of
n length lengths
May 139 69.4 51-105
June 1-15 332 71.8 52-102
June 16-30 340 76.1 59-115
July 1-15 192 77.8 64-118
July 16-31 252 82.2 70-125
August 1-15 28 93.5 79-120
):=0 August 16-31 96 101.9 81-131
I
1.0 September 1-15 14 99.6 86-127
September 16 -October 15 21 114.4 93-135
n
18
b
b
o
o
o
2
10
4
63.0
103.5
93.2
93.5
52-85
102-105
83-101
90-99
n
Indian River
Mean Range of
Length lengths
a Includes all mainstem,slough,and side channel sites sampled during the coded wire tagging and cold branding studies in the Susitna
River between the Chulitna River confluence and Devil Canyon.
b Not sampled.
Appendix Table A-10.Mean length and range of lengths for age 2+coho
salmon by sampling period on the Susitna River
between Cook Inlet and Devil Canyon,1984.~
~~
Sampling n Mean Range of
Period Length Lengths -.
May 5 133.2 120 -160
~
E.June 7 135.6 114 -157
L.June 1 136.0 136 ~
E.July 2 130.0 130
L.July 0
E.August 1 126.0 126
""'"L.August 13 138.0 125 -176
E.September 2 134.0 134 ""'"
L.September -
E.October 13 141.0 135 -150
~
All Season 44 137.1 114 -176
~
A-IO ~,
Appendix Table A-11.Daily catches of outmigrant chum and sockeye salmon fry in a fyke
net located at the mouth of Slough 21,May 23 to June 12,1984.
Check Date Sockeye Chum Check Date Sockeye Chum
May 23 1,005 74 June 3 155 8
24 694 83 4 140 8,-
25 810 60 5 164 10
26 2,150 355 6 419 12
27 1,479 399 7 1,024 82
28 400 83 8 570 85
29 1,777 198 9 761 59
30 253 89 10 31 34
~.June 156 44 11 23 8
2 344 33 12 29 8
13a 2 1
a Slough breached allowing fish passage around net.Net pulled.
r
....
A-ll
Appendix Table A-12.Mean length and range of lengths for age 0+sockeye salmon by sampling period on the Susitna River between Cook
Inlet and Devil Canyon,1984.
Lower Susitna a Middle Susitna b
Flathorn Station JAHS Sites Talkeetna Station Marking Sites
Sampling --Mean MeanPeriodMeanRangeofMeanRangeofRangeof Range of
n Length Lengths n Length Lengths n Length Lengths n Length Lengths
May 134 32.8 27-45 c --213 32.0 26-41 100 30.5 25-37
June 1-15 284 40.4 29-60 15 36.0 26-52 305 36.5 28-60 100 35.2 29-49
June 16-30 343 42.7 25-70 80 40.1 26-66 509 41.9 25-71 50 34.2 28-44
JUly 1-15 313 49.2 25-8p 20 43.6 30-65 570 48.8 30-75 0
July 16-31 337 52.2 30-85 54 43.5 28-76 748 53.4 35-87 8 53.1 47-68
August 1-15 239 53.0 29-85 38 47.9 30-76 547 51.8 33-88 49 51.4 43-62
):-August 16-31 185 52.8 30-93 106 53.0 28-86 90 58.6 42-79 50 56.2 36-69
I......
N September 1-15 41 55.6 42-75 20 61.2 45-71 95 59.8 40-91 0
September 16 -October 15 37 57.2 38-81 62 60.3 35-79 15 60.4 48-90 0
a Includes all mainstem,slough,and side channel sites sampled during the JAHS study in the Susitna River between Cook Inlet and the
Chulitna River confluence.
b Includes all mainstem,slough,and side channel sites sampled during the coded wire tagging and cold branding studies in the Susitna
River between the Chulitna River confluence and Devil Canyon.
c Not sampled.
)I _J __J _c_J ~J J
'.
)J •J -)J J J l
/'~
I~
,
Appendix Table A-13.Mean length and range of lengths for age 1+
sockeye salmon by sampl ing period on the Susitna
River between Cook Inlet and Devil Canyon,1984.
A-13
Appendix Table A-14.Mean length and range of lengths for chum salmon fry by sampling period on the Susitna River between Cook Inlet
and Devil Canyon,1984.
Lower Susitnaa Middle Susitna b
F1athorn Station JAHS Sites Talkeetna Station Marking Sites
Sampling
Range of Mean Range of Range ofPeriodMeanRangeofMeanMean
n Length Lengths n Length Lengths n Length Lengths n Length Lengths
May 35 42.7 36-62 c --367 40.1 32-52 150 39.9 33-47
June 1-15 198 41.9 30-55 298 43.2 31-58 357 45.6 35-68 300 44.5 36-60
June 16-30 209 42.7 32-63 109 39.4 31-50 427 42.9 36-62 50 40.2 36-48
July 1-15 17 42.5 30-59 37 42.3 33-57 337 44.0 35-65 50 48.2 39-54
July 16-31 3 43.3 31-52 21 40.4 36-47 172 44.6 36-59 10 46.5 40-51
Includes all mainstem,slough,and side channel sites sampled during the JAHS studies in the Susitna Riyer between Cook Inlet and
the Chulitna River confluence.
a
b Includes all mainstem,slough,and side channel sites sampled during the coded wire tagging and cold branding studies in the Susitna
River between the Chulitna River confluence and Devil Canyon.
c Not sampled.
»
I.....
~
~J J J 1 ,t I !J .-j J ~jJ ~j ~)j
-
APPENDIX B
THE SCHAEFER ESTIMATE OF POPULATION SIZE
The Schaefer method of estimating population size is useful with
migrating fish which can be sampled and marked at one point and
recovered later at a different point on the migratory route (Ricker
1975).The Schaefer estimate of population size (N)is given by Ricker
as:
number of fish marked during a single tagging period.
total recaptures of fish tagged in the ith period
number of fish captured and examined for marks during
a recovery period.
N =2'N..=~0".Mi .Cj )lJ L..lJ r If:"
1 J
number of fish which were marked during a tagging
peri od (i)and subsequently recaptured duri ng a recovery
period (j).
where:R..=lJ
F"'.
Mi =
R.=
1
C.=
J
R.=number of marked fish which were recaptured during a
J recovery period.
Nij =estimate of the number of fish available for marking
during a period (i)and the number available for recovery
in a period (j).
..-
-
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 are tabulated by the Schaefer
method in Appendix Table B-1.The computation of the population esti-
mate is presented in Appendix Table B-2 .
Because only age 0+sockeye fry were tagged and because some of these
remained in the middle river to overwinter (therefore,there was no
chance of recapturing them as age 0+fry at Tal keetna Station),we had
to assume that the marked/unmarked ratio was the same for the fry that
outmigrated as it was for the fry that remained to overwinter.The
purpose of sampl ing at Talkeetna Station was to estimate this ratio.
Data collected so far indicate that the number of overwintering sockeye
fry in this reach is low in comparison to the number that outmigrate,so
the consequences of violating this assumption are not severe.
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.
-
B-1
Appendix Table B~1.Data collected on the coded wire tag,mark~recapture experiment for sockeye salmon fry to provide a Schaefer
population estimate.Tagging and recovery periods are by eight day intervals,May 22 through September 18,
1984.
Period of Period of Tagging (i)Tagged Fish Total Fish
Recovery Recovered Recovered
(j)1 2 3 4 (Rj)(Cj)Cj/Rj
1 27 ~--27 339 12.6
2 4 --~4 71 17.8
3 7 -~-7 414 59.1
4 26 -6 5 37 1,293 34.9
5 21 -5 24 50 931 18.6
6 70 -16 15 101 1,627 16.1
7 32 -9 7 48 976 20.3
8 16 .1 3 20 428 21.4
9 29 -5 10 44 693 15.8
OJ 10 6 2 4 12 360 30.0I-
N 11 6 -~-7 173 24.7
12 - -
1 -1 20 20.0
13 1 ---1 46 46.0
14 2 ~--2 60 30.0
15 1 ---1 31 31.0
Total Tagged
Fish Recovered
(Ri)248 0 45 69 362 7,462
Total Fish
Tagged
(Mi)8,795 0 2,052 3,685 14,532
Mi /Ri 35.5 -45.6 53.4
~_~J J J 1 I J 1 o_J !),j )J J 1 J
..-
Appendix Table B-2.Computation of the sockeye salmon for outmigrant population from the
data presented in Appendix Table B-1 •
B-3
Appendix Table B-3.Data collected on the coded wire tag,mark-recapture experiment for chum salmon fry to provide a Schaefer
population estimate.Tagging and recovery periods are by eight day intervals,May 22 through July 24,1984.
Period of Period of Tagging (i)Tagged Fish Total Fish
Recovery Recovered Recovered
(j)1 2 3 4 (Rj)(Cj)Cj/Rj
1 11 ---11 932 84.7
2 -1 --1 104 104.0
3 3 4 2 -9 860 95.6
4 -3 3 6 12 526 43.8
5 1 3 -8 12 361 30.1
6 ---1 1 334 334.9
7 ---4 4 154 38.5
8 - - -
1 1 132 132.0
OJ Total Tagged
I Fish Recovered
~(Ri)15 11 5 20 51
Total Fish
Tagged
(Mi)4,806 12,276 5,295 9,019 31,396
Mi/Ri 320.4 1 )116.0 1,059.0 451.0
)I .~)J )J J J ~,.1 ,J •
Appendix Table B-4.Computation of the chum salmon for outmigrant population from the
data presented in Appendix Table B-3.
Period of Tagging (i)
-..
"....
....
-
Period of
Recovery
(j)
1
2
3
4
5
6
7
8
TOTAL
298,517
91,891
9,644
400,052
2
116,Q64
426,758
146,642
100,775
790,239
B-5
3
202,481
139,153
341,634
4
118,523
108,601
150,634
69,454
59,532
506,744
Total
298,517
116,064
721,130
404,318
219,020
150,634
69,454
59,532
2,038,669
~,
.-
APPENDIX C
Time Series Analysis of Discharge,
Turbidity,and Juvenile Salmon Outmigration
in the Susitna River,Alaska
-
-
-
TIME SERIES ANALYSIS OF DISCHARGE,TURBIDITY,AND JUVENILE
SALMON OUTMIGRATION IN THE SOSITNA RIVER,ALASKA
by:Stephen S.Hale
Alaska Department of Fish and Game
Susitna River Aquatic Studies Program
620 East lOth Avenue,Suite 302
Anchorage,Alaska 99501
ABSTRACT
During the three years of study of juvenile salmon outmigration from the
middle reach of the Susitna River,a correspondence has been noted
between the peaks of river discharge and the peaks of outmigration.
Further investigation of the relationship of outmigration to discharge
was required because two large hydroelectric dams have been proposed for
a region above the salmon rearing areas.These dams will markedly
change the downstream discharge and turbidity regimes,factors which
influence not only salmon outmigration, but almost all fish species and
life stages including juvenile salmon rearing.Box-Jenkins models were
developed for the 1983 and 1984 time series of river discharge,tur-
bidity,and chinook and sockeye salmon fry outmigration rates in order
to better understand the forces that shape the seri es and to stati s-
tically describe the natural conditions as a baseline against which
future changes can ,be measured.The time series examined were described
by relatively simple models,using mostly first-order autoregressive
terms.About 85%of the variance in turbidity for one day was explained
by the value for turbidity of the previous day.This figure was 44%for
chinook salmon outmigration and 43%for sockeye salmon outmigration,the
lower numbers indicating the effect of behavioral decisions on bio-
logical time series.Although the form of the time series plots of
discharge and chinook salmon outmigration was different between the two
years,the underlying stochastic processes which generated these series
were the same.Bivariate transfer function model s were constructed for
turbidity and salmon outmigration rates which explain present values of
these variables in terms of their own past values as well as past values
of discharge.
-
-
.-,
-~
TABLE OF CONTENTS
ABSTRACT ••.•e·'.• • • • • • • • • •i
LIST OF FIGURES ••••••••••••••••••••••••••••••••••••••••••••••.•••••i1-i
1.0 INTRODUCTION •••••••••••••••••••••••••••••••••••_...............1
1.1 Time Series Analysis •••.•••.••••••.•••••••••~................4
1.2 Applications of Time Series Analysis........................5
1.3 O-bjectives,•••'•.,••••••••••-•••,•••••••••••••••••••••••••••••••.,5
2.0 METHODS •••••••••••••••••••••••••••••__•••• ••••••••••••••••••••• _7
2.1 The Data ••••••••••••••••••••••••••••••••••••••••••.••••••'...7
2.2 Identification and Estimation of Time Series Models.........7
2.3 Transfer Function Models....................................10
3.0 RESUL 1S •••••••'•••••••••••••••••••••••••••••••••• •• ••••••••••••11
3.1 Univariate Model for Mean Daily Discharge...••••••••••••••••11
3.2 Univariate Model for Turbidity..............................17
3.3 Univariate Model for Age 0+Chinook Salmon Outmigration.....22
3.4 Univariate Model for Age 0+Sockeye Salmon Outmigration.....31
3.5 Discharge -Turbidity Transfer Function ModeL..............31
3.6 Discharge -Chinook Transfer Function Model ••••~............35
3.7 Discharge -Sockeye Transfer Function ModeL................37
4.0 DISCUSSION.....................................................40
5.0 ACKNOWlEDGEME-NTS •••••••••••••••••••••••••••••••••••••••8.• • • • •44
6~O LITE.RATURE CITED..............................................45
7.0 BOX-JENKINS ARIMA AND TRANSFER FUNCTION MODELS................50
ii
- ----~-~---------_._--
-
LIST OF FIGURES
Figure Title
1
2
3
Map of the Susitna basin study region ..•.•..•••••.•....••
Discharge,turbidity,and chinook and sockeye
salmon outmigration rate,1983 ••••••••••.•••••••••••.••..
Discharge,turbidity,and chinook and sockeye
salmon outmigration rate,1984 .••••••••.•.•••.••••••.••••
2
8
9
-
Susitna River discharge time series at the Gold
Creek gaging station,1983 and 1984......................12
5
4
Plots of autocorrelations and partial auto-
correlations for 1983 discharge time series..............13
6 Log-transformed discharge time series,1983 and
1984 'COl •••••••••••••••••••0 •e _II Ql 14
7 Plots of autocorrelations and partial auto-
correlations for 1983 log-transformed discharge
time series .•.•.•...•.•........•.•.•......•.•......eoGo.o.15
8 Spectrum of 1983 discharge time series...................16
9
10
Plots of autocorrelations and partial auto-
correlations for 1984 discharge time series..............18
Plots of autocorrelations and partial auto-
c~rrelat!ons for 1984 log-transformed discharge
tlme serles B •••••••••••••••••a..................19 ~,
11 Turbidity time series at Talkeetna Station,
1983 and 1984 o..os.................20
12 Plots of autocorrelations and partial auto-
correlations for 1983 turbidity time series..............21
13 Differenced turbidity time series,1983..................23
14 Plots of autocorrelations and partial auto-
c~rrelat!ons for differenced 1983 turbidity
tlme ser1es..............................................24
-
15 Age 0+chinook salmon outmigration rate time
series,1983 and 1984 ..e.................................25
16 Plots of autocorrelations and partial auto-
correlations for 1983 chinook salmon outmi-
gration time serles......................................26
.....
iii
--.
LIST OF FIGURES (Continued)
Figure
17
18
19
20
Title Page
Log-transformed age 0+chi nook salmon outmi -.
gration rate,1983 and 1984..............................28
Plots of autocorrelations and partial auto-
correlations for log-transformed 1983 chinook
salmon outmigration time serles..........................29
Plots of autocorrelations and partial auto-
correlations for log-transformed 1984 chinook
salmon outmigration time series..........................30
Age 0+sockeye salmon outmigration rate time
series,1983 and 1984....................................32
22
21 Plots of autocorrelations and partial auto-
correlations for 1984 sockeye salmon outmi-
gratian time series......................................33
Plot of cross correlations between the resi-
duals of the ARMA (1,1)discharge model and the
prewhitened turbidity time series,1983 data.............34
23 Plot of cross correlations between the residu-
als of the ARMA (1,1)discharge model and the
prewhitened chinook salmon outmigration time
series,1983 data .......•.........-.......................36
-
.-
24 Plot of cross correlations between the residu-
a 1s of the ARMA (1,1)di scha rge model and the
prewhitened sockeye salmon outmigration time
series,1984 data ••••••••••••,•••••••••••••~..............38
iv
-
-
-
1.0 INTRODUCTION
While examining the plots of daily catch rate of outmigrating juvenile
salmon at the Talkeetna Station outnligrant traps,an apparent correspon-
dence was noted between the peaks of the time series of mean daily
discharge and the time series of salmon outmigration (Hale 1983;Roth et
al.1984).Correlation analysis showed that there was a relatively
strong relationship between discharge and the outmigration rates of
various species/age classes of salmon during certain periods of time.
The term outmi grati on rate is used here to mean the number of outmi-
grating fry captured at the traps per hour,not the distance travelled
per hour.This relationship is not simply a matter of a greater volume
of water being fished at higher discharges.The correlations of catch
rate of age 0+salmon with water velocity at the mouths of the traps
were not significantly different from zero (Roth et al.1984,Appendix
A).There was in fact a greater number of fry per unit volume of water
at high levels of discharge than at low levels.
A correspondence between discharge rate and salmonid outmigration has
also been reported by other investigators (Cederholm and Scarlett 1982 -
coho salmon;Congleton et al.1982 -chum and chinook salmon;Godin
1982;Grau 1982;Solomon 1982b).The selective advantages of this
behavi.or,according to Solomon (l982b),include easier passage over long
distances or shallow areas and protection from predators provided by
increased turbidity and by the large numbers resulting from a coor-
dinated mass migration in response to an environmental cue.
There are probably two mechanisms which account for this relationsh"jp in
the Susitna River.One is that the fish,which have gradually become
physiol ogica lly ready for outmi grati on by growth and in response to
photoperiod and temperature,are stimulated by a rise in mainstem
discharge to begin that outmigration (Grau 1982).The second mechanism
is that high flows physically displace the fish downstream.This latter
mechani sm may frequently occur for fry rea ri ng ins i de slough s,pa rt icu-
larly for chum salmon (Oncorhynchus ketal and sockeye salmon (0.nerka).
The natal sloughs for many chum anCfSOckeye sa.lmon have berms at the
heads which prevent water from the mainstem from entering the site at
low levels of discharge.When high flows occur,the slough heads are
overtopped and the fry which had been rearing in low velocity water are
subjected to a strong current.
Because two large hydroelectric dams have been proposed for the Susitna
River in an area upstream of the rearing areas of the juvenile salmon
(Fig.1),and because these dams would markedly alter the natural dis-
charge and turbidity regimes,it is necessary to quantify the relation-
ship between the di scharge and turbidity regimes and the outmi gration
patterns of the juvenile salmon.After the dams begin operation,the
annual patterns of river discharge and turbidity level would be smoothed
-both would be lower than normal in the SUmnler and higher than normal
in the winter.Also,the high frequency (daily)oscillations of these
two time series would be dampened;there would be less day to day
variation.
1
N
INLET
f7
•10 RIIIe,mll,Incn""nis
•P,opoilid Dam
Figure 1.Map of the Susitna basin study region.
Data Center).
(Source:Arctic Environmental Information
J I J I ".,.J 5 I _I j )~J j )J )J
-I
-
-
--
There are many factors other than di scharge and turbi di ty whi ch affect
the outmigration timing of juvenile salmon including time of year,size
of fish,photoperiod,light intensity,and temperature (Brannon and Sa10
1982);however,discharge and turbidity bear further investigation
because of the changes in these two va ri ab 1es whi chwou1 d be caused by
the proposed dams.Changes in river flow can affect the survival rate
of young salmon (Stevens and Miller 1983).Potential negative effects
of an altered flow regime include accelerated or delayed timing of
outmigrations.Changes in outmigration timing may place the fish in
their rearing areas at an unfavorable time from the standpoint of food
supply,which could cause reduced survival (Hartman et a1.1967).Lower
discharge levels can result in a shorter distance covered per day
(Raymond 1968).Decreasing mainstem flows can lead to stranding of fish
in pools which have been isolated from the mainstem (Solomon 1982a).
Lower flows and clearer water than normal may also result in increased
predation (Stevens and Miller 1983).
Turbidity level in the Susitna River probably does not have much direct
effect on the daily number of fry which outmigrate or on the initiation
of outmigration.In clear water streams,however,an increase in
turbidity level can directly increase the number of outmigrating salmon
by providing cover from predators (Solomon 1982b).Turbidity level in
the Susitna River does change outmigration timing because fry in turbid
water outmigrate during the day as well as during the night (Godin 1982;
Roth et a1.1984).Clearing of the water.cou1d force the fry to shift
to a nocturnal outmigration to avoid predators.However,this would be
of marginal benefit for fry during the continuous daylight in June and
July at 63 0 N latitude.
To avoid or alleviate the above problems,it is necessary to understand
the mechanisms producing the present discharge,turbidity,and outmi-
gration regimes.Knowledge of the discharge-outmigration relationships
will be useful in trying to establish a post-project flow regime which
will not interfere with the natural outmigration timing.
Also,because discharge and turbidity level are important variables
affecting salmon life stages other than the outmigration phase as well
as other species,it is necessary to statistically describe the natural
discharge and turbidity regimes as a baseline against which .future
changes in these variables can be measured.Turbidity provides cover
for salmon fry (Suchanek et a1.1984;Part 2 of this report)but also
decreases primary producti on and affects the feeding,movement,and
distribution of many of the fish species present in the river.Turbi-
dity level after the dams begin operation will not only be influenced by
a changed discharge regime,but will also be directly changed by the
dams because settling of suspended sediment in the reservoir will create
a turbidity regime substantially different from the present regime.
Turbidity was included as a variable of interest in this paper more
because of its effect on other life stages and species than because of
its effect on salmon outmigration.
Further,discharge is the major variable in the extensive instream flow
habitat modeling effort which has been conducted in the Susitna River;
turbidity is also an important factor (Hale et a1.1984;Suchanek et ale
3
1984;Part 2 of this report).The current discharge and turbidity
regimes that are driving these habitat models must be accurately
described so that the models can be put into a proper perspective.
1.1 Time Series Analysis
The statistical methods collectively known as time series analysis.are a
logical choice for analyzing the natural discharge,turbidity,and
outmigration regimes.A time series is a collection of observations
ordered in time such as daily water temperature measurements.Time
series analysis includes frequency domain (spectral analysis)and time
domain problems.Spectral analysis is concerned with transforming a
time series with a Fourier transform to a sum of sines and cosines (see
Priestley 1981)and is appropriate with periodic·series such as the
classical example of the Canada lynx/snowshoe hare ten year cycle
(Bulmer 1978).Methods for time domain problems (or Box-Jenkins models)
a re referred to as ARIMA (autoregressive,integrated,moving average)
models (Box and Jenkins 1976).ARIMA models have been used extensively
in economic forecasting (Nelson 1973;Granger and Newbold 1977).
Time series are shaped by both deterministic and stochastic (random)
events.The series has a II memo ry II of the random events (or Il s hocks ll
)
operating on the series,that is,the effect of these disturbances may
be apparent for several time units after the event occurred.One aspect
of time series analysis consists of removing deterministic trends from a
time series so that the values fluctuate around a mean level.A trans-
formation may be necessary to ensure a constant variance.The random
processes that generated the observed series can then be mathematically
defined.The residuals left over after this model is fitted should be
lI white noisell (completely random)if the model is adequate.
Time series can be passed through a mathematical fi 1ter which changes
the form of the input series.A Ill ow pass filter ll dampens high frequency
perturbations and allows low frequency perturbations to pass unchanged.
This is useful in smoothing noisy time series so that the basic pattern
may be more readily observed.High pass filters are used when it is
desirable to remove obvious (low frequency)trends in order to focus on
the high frequency events.
Box-Jenkins models can be constructed using only the information con-
tained in the time series itself.For example,although the discharge
time series results from several independent variables including rain-
fall,air temperature,and solar insolation on the glaciers,it is not
necessary to quantify these inputs in order to model the output (dis-
charge).Information on the effects of all the inputs is already
contained in the past history of the discharge record.However,infor-
mation on the input series can be used in a transfer function model to
obtain an equation with more predictive power.This is a model where an
output seri es is a functi on of one or more independent input seri es as
well as its own past history.
An observed series is one realization of all possible time series which
could have been generated from a random process.Time series analysis
examines the nature of the probablistic process that generated the
4
-
-
~.
I
~,
-
.....
-
-
-
-
,....
-
..-
observed series.The model should have similar properties to the
generat"ing mechanisms of the stochastic process (Granger and Newbold
1977).Then!one can form summary statistics about the series and make
inferences about the nature of the stochastic process.After a model
has been developed!it can be used to test some hypothesis about the
generating mechani sm of the time series!to forecast future values of
the series!or to make decisions on how to control future values of the
series (Granger and Newbold 1977).
1.2 Applications of Time Series Analysis
Time series analysis has been extensively used in examining physical
data!particularly in oceanography.Salas and Smith (1981)demonstrated
that ARIMA models can be used to model the time series of annual flows
in streams.Srikanthan et al.(1983)analyzed the time series of annual
flows in 156 streams in Australia.Time series models have also been
used to examine the effect of the Aswan dam on the discharge of the Nile
River and the effect ofa hydroelectric dam on the discharge regime of
the Saskatchewan River (Hipel et al.1978).
Time series analysis methods have been also been used in examining time
series of abundance and catch in marine fisheries (Van Winkle et al.
1979;Botsford et al.1982;Peterman and Wong 1984;and Taylor and
Prochaska 1984).These methods have been used by Saila et al.1980,
Mendelssohn 1981!Stocker and Hilborn (1981),Kirkley et al.(1982),and
Jensen (1985)for forecasting future abundance or catch of marine fi sh
stocks.Mendelssohn (1981)used transfer function models in addition to
univariate Box-Jenkins models to forecast fish catch.Botsford et al.
(1982)focused on searching for causal mechanisms of observed cycles in
salmon fisheries in California rather than on defining models for the
fisheries.
Applications to freshwater fish ecology problems are much more limited.
Saila et al.(1972)used time series methods to cross correlate upstream
migration activity of the alewife to solar radiation and water tempera-
ture.O'Heeron and Ellis (1975)considered a time series model for
judging the'effects of reservoir management on fish.Applications of
spectral analysis to ecological problems have been reviewed by Platt and
Denman (1975)and time series analysis in ecology was the subject of a
.symposium proceedings edited by Shugart (1978).
1.3 Objectives
The objective of this paper was to develop mathemati ca 1 model s for the
times series of mean daily Susitna River discharge at the Gold Creek
gaging station (river mile 136.7),daily turbidity level!and daily
outmigration rates of chinook salmon (Oncorhynchus tshawytscha)and
sockeye salmon (0.nerka)at the Talkeetna Station outmigrant traps
(river mile 103.0)during the open water seasons of 1983 and 1984.
Because time series analysis can provide an efficient summarization of a
data set by a few parameters (Hipel et al.1978),these models will be
used to statistically describe the present conditions as a baseline
against which future changes can be measured.The discharge and tur-
bidity information will be useful for examining their relationship with
5
salmon fry outmigration as well as with other species and life history
stages.In addition,discharge was used as an input in transfer func-
tion models of discharge-turbidity,discharge-chinook outmigration and
discharge-sockeye outmigration in order to describe the relationship
between these variable and to be used as a possible technique to fore-
cast futu re values or to exami ne the probable effects of the proposed
dams.
Turbidity was chosen as a variable of interest because of its rela-
tionship with discharge and because of its importance in determining the
distribution of rearing juvenile salmon (Suchanek et al.1984;Part 2 of
this report)and other species.It was selected more for this reason
than for its effect on salmon outmigration,so it was not used as an
input in a transfer function model with salmon outmigration.Chinook
salmon were chosen because this species rears in sloughs and side
channels affected by mainstem discharge and because chinook salmon have
been selected as the evaluation species of the impact assessment study
(EWT&A 1985).The sockeye salmon time series was chosen because mainstem
discharge affects sloughs which are both natal and rearing areas for
this species.While chinook salmon spawn mainly in tributaries in this
system,sockeye salmon spawn mostly in mainstem sloughs.
6
-
~,
-
~I
.-
,-,
,
-
/"'''
2.0 METHODS
2.1.The Data
Mean daily discharge values for 1983 and 1984 (Fig.2,Fig.3)were
obtained from the U.S.Geological Survey gaging station on the Susitna
River at Gold Creek,river mile 136.7 (Still et a1.1984;U•.S.Geolog-
ical Survey provisional data,1984).The time frame examined was May 18
to August 30 (105 observations).Discharge levels begin to decline in
September when glacier melting decreases;hence,a .10nger series would
not be stationary.Throughout this paper,the unit for discharge is one
thousand cubic feet per second.
Daily water samples for turbidity (Fig.2,Fig.3)were taken at the
outmigrant trap station and measured with an HF Instruments Model No.
ORT-15S field turbidometer (Roth et al.1984).Units are in nephelo-
metric turbidity units (NTU).Only the 1984 turbidity series was
examined..
Outmigration rate (Fig.2,Fig.3)was measured by two outmigrant traps,
one on each bank,located at river mile 103.0 (Roth et a1.1984).The
rate·is reported as number of fish per trap hour with catch from the two
traps combined.Only age 0+fry were used in the analysis because the
traps were not efficient at capturing age 1+fry and,consequently,the
numbers were low.Further,age 1+chinook and sockeye salmon have
essentially completed their outmigration from this reach of river by the
end of July so the time series are shorter.
The chinook salmon time series for.1983 runs from May 18 (shortly after
ice-out)to August 30 (when outmigration is winding down),a total of
105 observations.The 1983 sockeye salmon data were not examined.
There were six days during the 105 day series when the outmigrant traps
were not fished - a one day,a two day,and a three day period.Although
values for gaps in time series can be estimated by a spl ine method,the
gaps in the outmigration series are short enough so that a s"imple
interpolation of values is sufficient (Sturges 1983).
In 1984,the traps were continuously operated from May 14 to October 6.
However,the series were cut off at the end of August in order to be
comparable to 1983 and to achieve a stationary series.About 98%of the
cumulativeoutmigration of age 0+chinook and sockeye fry in 1984 had
occurred by the end of August.
2.2.Identification and Estimation of Time Series Models
Univariate models were developed for the four time series:discharge,
turbidity,and chinook and sockeye salmon outmigration.Methods for
developing Box-Jenkins ARIMA and transfer function models are described
in section 7.0.Basically,there are three steps in developing an ARIMA
model:model identification,parameter estimation,and diagnostic
checking (Box and Jenkins 1976).The autocorrelation (AC)and partial
autocorrelation (PAC)plots for each series were examined to help
identify possible autoregressive (AR)and moving average (MA)com-
ponents.A tentative model was developed and the parameters estimated.
7
-
~-
.....
-
150
SUSITNA RIVER DISCHARGE
!lQ
!000
J.;;.
8.;301ia:e.
~20~u
'0
0 ......,Juft 1!5 .luI ,oIu.1:5.Au<:!,_l!5
1983
~
SUSITNA RIVER TURBIDITY
400
~300
~
.ili!200~
'00
0
Jun ,......'5 .suI ,Jul lS Au<:!1 _'5
1983
...
AGE 0+CHINOOK"••
'5
14
13
'2~":I:'0,gQ.
'"II~7u
IS
5
4
3
2
•
0
.Iu~•Juft 1!5 .Iu',""",r 1:5
1983
...
.7 AGE 0+SOCKEYE'.'IS
.4
'"'2~"'"'019
~IS
;;,
u ....
4
"2,
0
'9113
Figure 2.Discharge,turbidity,and chinook
and sockeye salmon outmigration
rate,1983.
8
so,.---=::"='==~-=~=:---:-:~....,
SU,SITNA RIVER DISCHARGE
50
'0
,,.,..,
-
o -In.'''''''rmm......'''''''''''''''''"'''""'''''''mmmm'''''''"""'mmmmmm.......l
,Ain 1 Jun 1~Jul 1 ,Jul 15 AU9'Au9'~
loge.
""".-------------~~--'----.....SUSITNA RIVER T..,RBIDITY
-
100
o-lm.'""'mmrmm......mmrmm"""'rmm......mm"""'rmm,;,.",mm......~
.Jun 1 Ju"'5 Jut 1 Jul'5 .Auq 1 Au9 15
'984
oJun 1 Jun 1.$.Jut 1 Jul'e Au9'AUtJ'5
198'4
O-irmrrrmrrrmmmnrnTmmmmmmrmmrmm""""rmmrmm""""",""""",rmm,,,..l
.I!!.----:-'="=~-..".,....,--:----'-----...,
u AGE 0+,CHINOOKI.
1ll...
1.3
12
10
10
1I
a
7
"II..
.3
2
-
II!!,.------__--------....,
17 .~GE 0+SOCKEYE
III
15,..
1.3
12
tl
'0
1I
I!!
7•5
4
.3
2,
O-M1lmmnitft;;(;","""'rmmmmmmmm"'"""""'rmmmm.....;.~~l'n11fI
...km'Jutt HI ..luI 1 .A.I(1!5 .Auq'~'S
'lIS4
"...
Figure 3.Discharge,turbidity,and chinook
and sockeye salmon outmigration
rate,1984.
9
Insignificant components were removed from the model.The residuals were
checked to see if there was significant departure from the assumption
that they were white noise.If the residuals were white noise,the model
was considered to be adequate.If not,a new model was identified and
the process repeated until the residuals were reduced to a white noise
process.
All of the time series work was done using the BMDP statistical package
(Dixon et al.1981).The BMDP Box-Jenkins program estimates parameters
by both the conditional least squares method and the backcasting method.
The estimates chosen for this paper were from whichever method gave the
lowest residual mean square.
The time series of mean daily discharge from May 18 to August 30 ap-
peared to be stationary so no differencing was done.A plot of the range
of sub-groups of the series against the mean of the sub-groups (as
suggested by Hoff (1983)indicated that a logarithmic transformation of
the data would be helpful in stabilizing the magnitude of the fluctua-
tions throughout the series;therefore,a model was also developed for
the natural log of the raw data.As the turbidity time series was
questionably stationary,models were developed for both the original
series and for a differenced series.
Models were developed for the chinook and sockeye salmon olJtmigration
rate time series on both the raw data and on data transformed by ln
(x +1).This transformation was used to avoid taking logarithms of
zero;there was zero catch on some days.
2.3 Transfer Function Models
Transfer function models (see section 7.0)were developed for discharge/
turbidity,discharge/chinook outmigration,and discharge/sockeye out-
migration.Only one input (discharge)was used.Multiple input transfer
function models (Liu and Hanssens 1980)or multivariate time series
models (Mendelssohn 1982)can be developed,but are substantially more
complex.
10
-
-
-
-
-
-
I~
3.0 RESULTS
3.1.Univariate Model for Mean Daily Discharge
The time series of mean daily discharge during the summer of 1983 is
shown in Fig.4;the log-transformed data are in Fig.6.Autocorre-
lation function (ACF)and partial autocorrelation function (PACF)plots
for the raw data are given i nFi g.5 and for the log-transformed data
in Fig.7.In all the ACF and PACF plots 5 the "+"symbol on either side
of the vertical axis indicates the 95%confidence interval.The first
order autoregressive ·component was strong in both the raw and the trans-
formed series.The ACF and PACF plots for the raw data indicated that a
moving average component was required.Models containing various combi-
nations of first and second order AR and MA terms were examined.Of the
acceptable models identified 5 the model with the lowest standard errors
on the parameter estimates and the least significant residuals was an
ARMA(2 52).However 5 the ARMA(I51)was nearly as good as the ARMA (2 52)
S05 in keeping with Box and Jenkins'(1976)advice that a parsimonious
model (i.e.5 the one with the fewest possible parameters)is desirable,
the ARMA(1 51)is considered the "best"model for the non-transformed
data.Parameter estimates were:
=.992 with std.error of .0135
-
h
E9,=-.580 with std.error of .0807
The model is:
where:Yt is the discharge level at time t and
a~is a white noise process at time t
Neither the mean nor any of the autocorrelations or partial autocorre-
lations of the residuals was significant;therefore 5 the model is
considered to be adequate.This equation can be interpretted as:The
discharge level for any given day is a function of (the mean 1eve1 5 22.7
cfs,of discharge during the period)plus (most of the previous day's
discharge level minus the mean level)minus (about half of the previous
day's noise component)plus (the given day's noise component).
The plots of both the ACF and PACF on the res i dua 1s from th is model
showed a slightly significant spike at a lag of 15 or 16 days.This
could indicate that the discharge time series has a periodicity of about
15 days,or sl ightly more than two weeks.Thi s possibi 1i ty was further
examined by spectral analysis.The spectrum of discharge (Fig.8)did in
11
Susitna River Discharge.1983 ~
60
50
"'0 ~C
0 40u
~
~"'O
!.S 30III
-~1110~~......u:a 20=s
0
10
~J
a
Jun 1 Jun 15 ..lui 1 ..luI 15 Aug 1 Aug 15
1983
Susitna River Discharge.1984
60
50
~
"'0
C 400
~-.
..."'0!.~30-~"0.f!§.
u:a 20~
0
.-10
a ~
Jun 1 Jun 15 ..lui 1 ..luI 15 Aug 1 Aug 15
1984
Figure 4.Susitna River discharge time series at the Gold Creek gaging
station,1983 and 1984.
12
13
12
11.5
11
'10.5......
0>..c.u..10is...,
3
9.5
9
a.5
8
Log-transformed discharge.1983
-
Jun 1 Jun 15 Jul 1 Jul 15 Aug 1 Aug 15
1983 ~
LOG-TRANSFORMED DISCHARGE.1 984
12..,.....-----------------------,
11.5
11
......10.5t&J
Cl0:
!10<.J
!!!
0...,z 9.5....I
9
8.5
8
Jun 1 Jun 15 Jul 1 Jul 15
1984
Aug 1 Aug 15
Figure 6.Log-transformed discharge time series,1983 and 1984.
-
14
AUTOCORRELAT ION S
•1.0 -.8 ••6 •.4 -.2 .0 .2 .4 •S ••1.0
LAG
t +
t to
".,..++-
+·t
++
++..t
,~++..t
+t
+......+-...~......
++
++
+.........
+++...
++
+...
++
+.........
~..
++
-++
~...
50 +...
+.........
r-
!
PARTIAL AUTOCORRELATIONS
.1.0 -.8 -..-.4 -.2 .0 .2 .4 .S o •1.0
LA.
.....,7iI'I'Rr\
-+.....
-++..........
.........+
+...
10 ......
++
++...+
++
III ...+++"'-......
t-+++
20 ++
+++.........
+...
25 ++
+...
+
+
30 +
+
,.1lJlII.'!!Il +
Figure 7.Plots of autocorrelations and partial
~autocorrelations for 1983 log-trans-
formed discharge time series.
15
.....
O"l
2.4
1.8..,
:E
;:)
a::1.2....
()
W
0-0.6
t/)
CD 0.00
..J
-0.6
+
...
++
++++
+++
++
+
++++++t
++t
+++
+++++
++++-t
++
++++++++++++,+
I , ,8 ,
.00 .05 .10
I • ,,, I ,
.15 .20 .25 .30
FREQUENCY
,
.35
, I I
.40 .45
I
.50
Figure 8.Spectrum of 1983 discharge time series.
],J J ~J )J t ~J I J J ))J ~j
fact indicate apeak at a frequency of .065 (a period of 15 days).It is
not known at this time if this periodicity is urea"'.It may be related
to weather patterns in the basin which control solar insolation (cloud
cover)and rainfall.A much longer time series of discharge would have
to be examined to answer this question.A periodic term could be added
to the ARMA(l,l)model (Box and Jenkins 1976)but,given the low signi-
ficance level of the periodicity,it does not seem appropriate at this
stage of model development.
Carrying the idea of parsimony a step further,it can be seen that an
ARMA(I,O)model using the log-transformed data is adequate and has the
lowest number of parameters.The parameter estimates for this model
were:
A
¢l =.994 with std.error of <.00005
giving
a •qq (~I"tIt _\-10.0)+4.+Nw\1t =10.0 +d \;
A A
The parameter ¢,was very close to unity.If ¢.were equal to 1.000,the
model would be reduced to a random walk model (Chatfield 1984).That is,
the log of the discharge for today is the same as the log of the dis-
charge for yesterday pl us a random error term.When q;1 approaches 1.000
in a model with only one AR term,the series could be non-stationary
(Hoff 1983).To test this,the series was differenced.The residuals
from an ARIMA(l,1 ,0)model showed significant spikes,so the differenc-
ing did not help;the ARIMA(l,O,O)model is better.
The AC's on the residuals of the ARMA(l,O)model were a little better
than those of the ARMA(l,l)on the non-transformed data.However,the
mean of the residuals was slightly significant,so the ARMA(l,l)model
on the raw data is probably superior to this one.
The 1984 discharge time series is shown in Fig.4 and Fig.6.The ACF
and PACF plots (Fig.9)were similar to those of 1983.An ARMA(l,l)
model on the 1984 raw data ~?s adequate,as it was in 1983.Para~ter
estimates were:y =23.2;¢>,=.808 (std.error =.0638);and 6,=
-.692 (std.error =.0750).An AR(l)model on the log-transformed data
was also adequate but,again,had a slightly significant mean residual.
The ACF and PACF plots,using log-transformed data (Fig.10),were
similar to those of 1983,but perhaPas showed less indication of a moving
average process.The estimate for ~,was .994 (exactly the same as the
1983 data),with a standard error of 0.0001,and the estimate for y was
10.0.
3.2.Univariate Model for Turbidity
The time series for turbidity in 1983 (Fig.11)was more complex than
that of discharge.The ACF and PACF plots (Fig.12)indicated a strong
AR(l)component.However,AR{l),AR(2),and ARMA(l,l)models were not
adequate to explain the series.
17
.'i'_4 AWi
""'"
AUTOCOR RE LATIONS
-1.0 -..-.S -."-.2 .0 .2 ...•S ••1.0
LAG
t ++t
++--
t +-~,;
++
t +
+t
+t
+t
10 t +
+t
of'....+..+
15 +...
++
++
+...
++
20 1-+
+t.....
+...
t +-25 +.....+
++..i'....
+...~50 ..;...-+
"'"'
PARTIAL AUTOCORRELATIONS
-1.0 -.'a ••a .4 •.2 .0 .2 •of ••••1.0
..+
--+..
1-..+
1-..
++
++
++
++
t +..+•+
++....
++
t +
+•............+.....
++
+t++..1-..1-
t +
+l'...1-.....
+..
++
Figure 9.Plots of autocorrelations and partial
autocorrelations for 1984 discharge
time series.
18
AUTOCORRELATIONS
-1.0 -.8 -..-.4 -.2 .0 .2 ••••••1.0
LAG
++
++
++--
+..-...t
++
++
++++.......
++++++~+...
.++
++......
++
++......
++
++
+...
++
+to
++
++
++-++
+......+
++
PARTIAL AUTOCORRELATIONS
-1.0 -.S -..-.4 -.2 .0 .2 .4 ••••1.0
LAG,-
+•-++..+
++
~.............+..+...+..-+......+..+
++
to ...
t .....to....+..+...+++~++I
I +...I ..++++1"~+.....+.....
+...
30 +.....+.....
.....Figure 10 .Plots of autocorrelations and partial
autocorrelations for 1984 log-trans-
formed discharge time series .
.-
19
Susitna Riv~r Turbidity,1983
soo ,...-------------------------,
400 ~
"~300
....,
;;.
~
..Q...200~
100
Figure 11.Turbidity time series at Tal keetna Station,1983 and 1984.
20
AUTOCORRELATIONS
.--1.0 -..-.s -.4 -.2 .0 .2 .4 .S .S 1.0
LA.
..+
++.-+......
+.-++-
++--+..
t .......
+t....
"""..+..+...+....•..
~..+.....
+•+.....+
+..
.~..............+
++
++...+......
+.......
The series appears to border on being non-stationary because it in-
creases in the spring as glacier melt increases and then declines in the
fall.(This series would certainly be non-stationary over a longer time
frame because the turbidity level is very low in the winter).The slow
decay of the autocorrelations in the ACF (Fig.12)also indicated
non-stationarity.
Further investigation using the raw data showed that the series had a
significant second order MA term,while the first order MA term was not
significant.Both first and second order AR terms were significant.This
gives the model:.
~t::r'1(,.1 t I qCof (l"t-r -It~.r)+.0<'0 (1t-.l -It"./)
+ ..2 3 2tt-~+2Lt.
A
with std.errors:on ~,=.0122
J\
on tJ2.=.0234
1\on e =.0988
~
Note that even though the same notation is used,the white noise process
(~~)here is different from that in section 3.1.
While this ARMA model is adequate for the time frame examined,in
general,an integrated model (i .e.,one with a differencing operation)
is probably more appropriate because of the suspected non-stationarity
of the raw data.The differenced series (Fig.13),which represents
consecutive changes in the original series values,is clearly stationary
with a mean close to zero.The ACF and PACF plots for the differenced
series (Fig.14)showed that the differenced series could be adequately
mode 1ed with just the second order MA term;the fi rst order autore-
gression term was not significant in the differenced series.The
equation is:
.....
~,
•.23 ~t-~t
w 'hu"e ~Z t::'#Iv ~-""'t-If ..d-I
I\.with std.error on el\=.0972 and the mean of the residuals insignifi-
cant.<.-
3.3.Univariate Model for Age 0+Chinook Salmon Outmigration
The time frame chosen for Age 0+chinook salmon was the same as that of
discharge (Fig.15).The plots of the ACF and the PACF for 1983 (Fig.
16)showed a strong first order autoregresssive component.In fact,an
ARMA(l,O)model,abbrevi ated as AR(l),adequately represents the data.
Although the plot of the range of sub-groups against the mean of the -
22
1 ]1 1 )l l -I t l J ,t,~"J
ORIGINAL DATA
+...
++++++...+
++...
++++++..t++
+
+
++
++...+.+
+++
++..........+
1-
'+
...
++++
+ ++++
+...'"...'"++++++
++++.+...+
++
++
...
+++++++
+...++++..+..++
++++++++
...0
>-....uo
o
CD
0:110
:::>....
:::>
l-z a.o-
-
o I I I I
18
MAY JUN JUL AUG
N
W S leo
t-
Z
FIRST DIFFERENCES
...
+
......+
+ +++.........
+
+ •
"t
..
...++
..
+...++++++++++
+++...
++
+
...
+
+
...
+
+
..
+
+
+
..
+
.....
++
+
+.
...
++ +
.....
+
+
++..+
...+..+t +t ++..+
+
+
++
...
+
...+...++++..++t-+
.0>-t-
O
-0.0CD
0:
~
t--80
-
+
I.
MAY JUN JUL AUG
Figure 13.Differenced turbidity time series,1983.
AUTOCORRELAT10NS """',-1.0 -.8 -.IS -.4 -.2 .0 .2 .4 ••••1.0
LAG
~+-+~
~+++++++..+"""++
++
++
+..
+......++-++..+........
++""'"+..
20 ....
++
"t +++....
++
25 ..+..+....
++
+T++
++
PARTIAL AUTOCORRELAT10NS '~
-1.0 -.8 -..-.4 -.2 .0 •2 .4 •• ••
1.0
LAG
+..-,-++....
++
!5 ++++++..+++
10 ++..+
4-....+
-+..
+....
+..
++~,+....
25 +
+++
+
50 ..
+
Figure 14.Plots of autocorrelations and partial
autocorrelations for differenced 1983
turbidity time series.-
24
-
Age 0+Chinook Salmon.1983
18 -.--------------------r------,
17
16
15
14-
1.3
12
·11
10
9
8
7
6
5
4
.3
2
1
o-Hmrr1'm'rfb~rtm:rr:¥rmm'T11TrmmTTTlTl~;:;;m:iTimmm~TTTlTI1lTITI~rr_rl
-
Jun 1 Jun 15 Jul 1 Jul 15
1983
Aug 1 Aug 1,5
Age 0+Chinook Salmon.1984
Aug 1 Aug 15Jul1Jul15
1984
Jun 1 Jun 15
18.,....-----:---'---.
17
16
15
14
1.3
12
1 1
10
9
B
7
6
5
4
.3
2
1
O-lTrrrrrmmmTTTTTTTl1l1mmT1'lTrnTTTITrmmrnmmTTTTTrrrmrnmT1'lTrmmmmTrmmTTrrrrI
...
~
:I:
ba.
.cu....ou
"""
,-
Figure 15.Age 0+chinook salmon outmigration rate time series,1983 and 1984.
25
AUTOCORRELATIONS
-1.0 -.'-.8 -.4 -.2 .0 .2 .4 .•.1 1.0
LA.r--I---lf----l--+--+--+--+--t---t--;---r-
&0
..
++
?
+
+
++++....
+..
+
++
++
+.....
++......
+++...
+
++----
t---+-
...
+...
+
+
++...
+...
i-+...
+
i-..
++++...
+++...
++++
",.,..
-
-I
PARTIAL AUTOCORR ELATIONS
-1.0 -.'-..-.._.2 .0 .2 .4 .,••1.0LAG....
+i-
+...
++.........+
++......
++..+
++++
++..+++
++..+.....
++++
++......
++++...+
1-++.........++..
++++
Figure 16.Plots of autocorrelations and partial
autocorrelations for 1983 chinook salmon
outmigration time series.
26
-
sub-groups indicated the need for a logarithmic transformation,the
residual AC's of an AR(1)model on the log-transformed data (Fig.17)
were slightly larger (but still insignificant)than those of the AR(l)
model on the raw data.The standard error on (DI 'however,was lower
with the log-transformed data.ACF and PACF plots for the log-
transformed data are shown in Fig.18.The AR(I)model for the raw data
is:
·~,("1 t-f -J •5 <)t
-
......
""'"i
-
"with standard error on ¢,=.0743.
TheAR(l)model for the log-transformed data is:
"with standard error on ¢,=.0363.
The mean of the residuals was not significant.
The time series plot for age 0+chinook salmon outmigration in 1984
(Fig.15)shows a different pattern frnm that of 1983.The fry did not
begin to migrate in 1984 until about June 12.The low level of out-
migration early in the season causes a time series which is non-
stationary.To avoid this problem,the time frame selected for 1984 ran
from June 12 to August 31 (79 cases).Analysis of this shorter series is
not as strong as that of the longer seri es in 1983 but the seri es is
long enough from a statistical point of view;Hoff (1983)suggests that
about 40 or 50 observations is the minimum necessary for attempting an
ARIMA model.Although logarithmic transformation did not appear to be
strictly necessary for the 1983 data,it was requi red (to produce an
AR(l)model)with the 1984 data,perhaps because of the shorter time
series in 1984.
The ACF plot for 1984 on the log-transformed data (Fig.19)was similar
to that of 1983,although it did decay a little more quickly.The 1984
PACFplot (Fig.19)was very similar to that of 1983 in indicating a
strong AR(I)component.The estimated value of¢,in 1984 was 0.973 (very
close to that of 1983),with a standard error of 0.0265.The 1984 model
is:
27
,""",
Log-transformed chinook,1983
4
3.5
::5
.......2.5..-
+
J:2a-
8z
..l 1.5
0.5
04n\-rffrrlirrrmmmTn1Tl1TmmmTnTlTTTmmmTn1Tl1TmmmTnTlTTTmmmTn1Tl1TT1TT111TTT'11
-
..Iun 1 ..Iun 15 ..luI 1 ..luI 15
1983
Aug 1 Aug 15
LOG-TRANSFORMED CHINOOK,1984 -4-,---------.
-
Aug 1 Aug 15..luI 1 ..lui 15
1984
Jun 1 ..Iun 15
O-fmrnmmmmmrrrmrrfmrrrmmmrnmmmrrrmrrrmmmrnmrnmmmmmmmmmmm-mrrl
0.5
3.5
::5
--2.5..-
+
J:2a-u
'-'3 1.5
Figure 17.Log-transformed age 0+chinook salmon outmigration
rate,1983 and 1984.
28
,-
AUTOCORRELATfONS
I"--1.0 -..-.'-.4 -•2 ~.0 .2 .4 ••••1.0
LA.
t t
+..
to +--
++-
++
++
++++
t +
++
++
++++++
++
++
++
++-++
++
++
++
-++-t -+
15 +....
++
++
+....
++
++
++
++
29
15
10
115
20
AUTOCORRELATIONS
·1.0 -..-..-.4 -.1 .0 .1 .4 ••••1.0
+
++
++-..+
++...+...t
+++....+..+
++
++...+
++
++
+...
+......+..+
++..++++..
+...
++..........
++...+
++
++
-
"""
-
PARTIAL AUTOCORRELATIONS
-i.O -.....-.4 -.2 .0 .2 .4 ••.,1.0
l.AG
+...
++
++..+
5 ++...+
++~...+++
10 ++........."~.....
++
15 ++...+...+~++
++
20 ...+........i...+++
25 +...
+..
++..+~
++
50 ...+
+t+...
Figure 19.Plots of autocorrelations and partial
autocorrelations for log-transformed
1984 chinook salmon outmigration time
series.-
-
30
"""
.-
The mean of the residuals was insignificant.This model does not differ
from that of 1983,except that the mean 1eve1 was higher.This was a
result of a higher escapement of adult chinook salmon in 1983 than in
1982.
All three of the ACF plots for chinook fry outmigration (Figs.16, 18,
and 19)had AC's after lag 18 which did not appear to decay further.
This may indicate the presence of a weak non-stationary or periodic
element which should be explored with subsequent data sets.
3.4.Univar;"ate t10del for Age 0+Sockeye Salmon Outmigration
Age 0+sockeye salmon outmigration was examined from May 23 through
August 31,1984 (Fig.20).This time series showed a strong AR(l)compo-
nent (Fig.21),similar to that of the chinook salmon time series.
However,neither an AR(l)model on the raw data or on the log-
transformed data was adequate.A MA(l)component was also significant in
the raw data,1eadi ng to the model:
....
i
1t'::'.1(,t .18 (-d"t ..,
.-
i
i
-
-
-
,"'"
A ~
The standard error on 'i/J,(.775)was .0681 and on €II (-.567)was .0883.
Although the mean of the residuals was slightly signlficant,none of the
autocorrelations or partial autocorrelations were,so the model is
reasonable .
3.5.Discharge-Turbidity Transfer Function Model
The cross correlations for the residuals from the 1983 discharge series
and the 1983 turbidity series,both filtered by the ARMA(I,I)model for
discharge,had a significant spike at lag =1 day (Fig.22).This
suggested a candidate model (Box and Jenkins 1976;McCleary and Hay
1980):
~o 8
1-6.B
where:Yt is the output series (turbidity)
w 0 and £.are transfer functi on parameters
B is the backward shift operator
x t is the input series (discharge)
Nt is the noise component,an ARIMA model
31
-
Age 0+Sockeye Salmon,1983
18
17
16
15
14
13 ~\
12
~:::r 110
J:10U9Q.
.J:.8u....7Cl
0
6 ~
5
4
.3
2 ~
1
.0
Jun 1 Jun 15 Jul 1 Jul 15 Aug 1 Aug 15 -1983
-Age 0+Sockeye Salmon.1984
18
17
16
15
14
13
12
'-:::r 11Q
J:10
b 9Q.
.c Su..70
-0 ~16
5
4
3 -2
1
0
Jun 1 Jun 15 Jut 1 ..luI 15 Aug 1 Aug 15 ~
1984
~
Figure 20.Age 0+sockeye salmon outmigration rate time series,1983 and 1984.
.-
32
AUTOCORRELATIONS
r-.-1.0 -..-..-.4 -.2 .0 .t .4 ••••1.0
LAB
+++.-++
++
++
++
++
++
++
++
++
"....++
+-+++
++
++
++
++
++
++
++
++
25 ++
++
++++
++
r""'"SO ++
++++
..-
i-'
i
PARTIAL AUTOCORRELAT ION S
-1.0 -..-..-.4 -.2 .0 .t .4 ••••1.0
LAG
+-to
-i +++-
++..+
++++
++++++
++
++++++++++++++++
++
++
+++++++
++++
++
+
++++++
Figure 21.Plots of autocorrelations and partial
autocorrelations for 1984 sockeye salmon
outmigration time series.
33
-
CORR ELATIONS ~CROSS
.0 .2 .4 .f
,.
.1 1.0.1.0 ..I -.s -.4 •.2
LAG
~
•20 +•++
++
+•+..-.-Ie ++
++•..
++
++
++
+++'t
++
++
-15 ..+~,..+++
++++..+
++-
+++
++++-+..
++
+..
++
++-.++++
++..+
++-..'t
++++
-+..
20 ....-
Figure 22.Plot of cross correlations between the
residuals of the ARMA (1,1)discharge
model and the prewhitened turbidity
time series,1983 data.
34
-
-
The assumption that the ARIMA component of the model was white noise led
to significant AC's in the residuals series and was therefore rejected.
The ACF and PACF plots on the residuals from this model suggested an
AR(l)model for the Ntcomponent~leading to the full model:
OJ o (j
J -J"BI
+
1-¢,B
Parameter estimates were:
A
00 0 =8.349 with std.error of 1.7044
-'\
S,=-0.559 with std.error of 0.1718
95,=0.993 with std.error of 0.0009
The t statistic for each of these estimates was significant~leading to
the conclusion that discharge and turbidity are related by the equation:
t.35 B7t=I+-.5"B /"I t +-.qq B
--
The ACF and PACF plots on the residuals from this model showed no
significant spikes;therefore~the model is adequate.
3.6.Discharge-Chinook Transfer Function Model
After both the input series (discharge)and the output series (chinook
salmon outmigration rate)from 1983 were filtered by the ARMA(l~I}model
for the discharge series and the residuals from both series were cross
co.rrelated~there was a significant correlation at lag =1 day (Fig.
23).This suggested the transfer function model ~as given by McCleary
and Hay (1980):
or~using the backward shift notation of Box and Jenkins (1976):
35
CROSS CORRELATIONS
•1.0 •.8 ~..~.4 ~.2 .0 .2 .4 .S .8 1.0
LAS
-20 +t
++
+i-
++++
-15 +..
++
++
++
++
-10 +-+..+.......+++
+ +
++++
++
++..+..1'-
+..
+......
+........+
+-+
++....
+of
+..
+......
++
++..+
+....+.......
Figure 23.Plot of cross correlations between the
residuals of the ARMA (1,1)discharge
model and the prewhitened chinook salmon
outmigration time series,1983 data.
36
-
-
-
~,
-
This model implies that the current day's discharge rate has an effect
on the next day·s outmigration rate.The estimate of W o was 0.02.The
residual ACF using this model suggested that the assumption of white
noise for the Ntcomponent was not valid;it appeared that an ARMA(I,O)
mode 1 wou 1d be appropri ate.The fu 11 model is:
+
1-¢,B
The parameters for this model were estimated as:
.1\
W o =.025 with std.error of .0249
~l =.667 with std.error of .0751
The t statistic on the estimate for tOo was not significant.However,
because the practice of prewhitening the output series with the model
for the input series tends to underestimate the si gnifi cance of the
results (Botsford et al.1982)and because there was a significant cross
correlation between discharge and outmigration rate at a lag of one day,
it seemed best to leave this term in the model.This would have to be
verified with more years of data.The model is:
-
-
The ACF and PACF for the residuals from this model showed no significant
spikes so we may conclude that the model is adequate.
This model does not imply that the discharge series is a strong predic-
tor for the outmigration series.But adding discharge does result in an
expression which has more predictive value than would be obtained by
looking at the outmigration series by itself.
3.7.Discharge-Sockeye Transfer Function Model
As with the di scharge-chinook rel ationshi p,the cross-correl at;ons of
the 1984 discharge and sockeye series,filtered by an ARMA(l,I)model
for discharge,showed a significant spike when the sockeye series was
lagged one day behind the discharge series (Fig.24).This spike was
stronger for sockeye than it was for chinook.A candidate model (Box and
Jenkins 1976;McCleary and Hay 1980)was:
I ~8 t3,!'It +-
37
CROSS CORRELATIONS
•I.0 -.8 -.6 •.4 -.2 .0 .2 .4 .e .8 1.0LAG
-20 t +...+...+++
++
-15 ......+++of-
+i-
+of-
-10 ++
T +...to
++++
-5 .....
+...+...
+...-++
0 ++++-
+......+-t +
++++......
++++
10 ........+
++++
+ot-++
+...
++
++....t-
20 -++
Figure 24.Plot of cross correlations between the
residuals of the ARMA (1,1)discharge
model and the prewhitened sockeye salmon
outmigration time series,1984 data.
38
-
-
....,
The ACF and PACF plots on the residuals from this model suggested an
ARMA(l,l)model for the Nt component,leading to the full model:
lUI)B
1-ef,B +((-$1 8)
(I-fl.B)
-
""'",
Parameter estimates were:
"-Wo =.206 with std.error <.00005
A
8,=-.190 with std.error .1848
1\
(/>,=.952 with std.error .0483
"e,=-.318 with std.error .1078
A
The t statistic for each of these estimates except J was significant,
giving:'
-3whereIYt=discharge X 10
The ACF and PACF plots on the residual series from this model showed no
significant spikes and the mean of the residuals was barely significant;
therefore,the model is deemed adequate.
-,
-
,,-t -(l+.3.2B)
( \ -.15 G)ttt
39
4.0 DISCUSSION
I
Time series analysis is a useful method for dealing with time ordered
data sets,including ones that do not appear to make much sense at first
glance because they are too noisy or because they drift as a result of
.random events.The·modeling effort helps us to understand why the plots
look as they do and what factors shape them.It also is useful in trying
to understand what effect a change in the controll ing factors might
produce.
The influence of discharge level on turbidity and chinook and sockeye
salmon outmigration is clearly seen upon inspection of Fig.2 and Fig.
3.Of course,these latter three series are shaped by several factors
other than discharge,so the correlation coefficient between them and
discharge is not normally expected to be high,unless a relatively short
section is examined.For example,the discharge peak in early June of
1983 is reflected in the other three series (Fig.2).The bimodal
discharge peak in August of 1983 is reflected in the turbidity and the
chinook outmigration series,but only the first August peak is shown by
the sockeye outmigration series.Thi s was because most age 0+sockeye
salmon in the reach above the traps had left by the middle of August.
Similarly,the late August discharge spike in 1984 had no effect on the
sockeye series (Fig.3).However,the high discharge peak in mid June
of 1984 is strongly reflected in the sockeye series because this was a
time when many age 0+sockeye salmon were present in the reach.
Another example of a change in the relative effect of a discharge spike
is shown by the 1984 chinook salmon series.The high discharge peak is
mid-June had less effect on chinook outmigration than did the lower
discharge peak in late July,a time when more age 0+chinook fry were
ready,because of physiological and behavioral reasons,to outmigrate.
The segments of the time series examined (discharge,turbidity,chinook
and sockeye salmon outmigration)were described by relatively simple
Box-Jenki ns models,usi ng mostly fi rst-order terms.The useful ness of
Box-Jenkins models is shown by the relative simplicity of the models
developed for the salmon outmigration series;a visual inspection of the
plots of the raw data for these series (Figs.15 and 20)gives the
impression of an erratic series of events.None of the series appeared
to require differencing (although turbidity was on the borderline)to
achieve stationarity nor did they appear to have a periodic component
(discharge being a possible exception)which would require seasonal
differencing.However,this should be re-examined when subsequent
seasons of data are available.All of the series showed a strong first
order autoregressive term,indicating that the value for anyone day is
greatly influenced by the value for the previous day.Similar results
for the flow regimes of several streams in Austral ia was reported by
Srikanthan et al.(1983),who found that most of the discharge series
which were not white noise had a first order autoregressive term.
Examination of the autocorrelation coefficients of the four time series
at lag =1 day (adjacent values)gives an idea of the smoothness of the
time series.Typically,the coefficient for physical/chemical variables
is higher than that of biological variables and the time series for
40
-
-
-
-
discharge (Fig.4)and turbidity (Fig.11)are less jagged than those
for chinook salmon outmigration rate (Fig.15)and sockeye salmon
outmigration rate (Fig.20).Saila et al.(1972)reported similar
results for the autocorrelations of alewife upstream migration activity
in relation to incident solar radiation and water temperature.
The square of the autocorrelation coefficient at lag =1 gives a measure
of the percentage of the variance of the value for today which is
explained bY2 what was measured yesterday (Murray and Farber 1982).In
1983,(.86)=74%of the variance of discharge for one day was
explained by the value for diicharge on the previous day.The percen-
tage for turbidity was (.92){85%while,for chinook salmon outmi 2grationrate,it was only (.66)=44%,and,for sockeye salmon,(.65)
=42%.
So,although fish tend to move in pulses more so than water or suspended
sediments,fish outmigration is far from being a random event.That is,
when an outmigration pul se occurs,the impetus has affected many fish
and the phenomenon extends over a three or four day peri ode When we
look at an outmigration time series,we are seeing the integrated
results of several factors operating on sub-groups of the population in
different locales.The fry in one slough may have emerged two weeks
earl ier than those of another slough because of a higher intragravel
temperature.Or the head of one slough may have overtopped at a lower
discharge level than the head of another slough,thus providing an
environmental cue to the two groups at different points in time.But
there is also a behavioral effect in that fry are stimulated to migrate
when they see other fry migrating.This is particularly true for those
species that form schools during outmigration.
The turbidity time series was the only one examined which included a
second order term.The second order moving average term is 1ikely
re'lated to the random "s hock"caused by a rising discharge (which is in
turn caused by rainfall)which resuspends sediment.It takes a few days
after the rainfall is over for this perturbation in turbidity level to
drop to the pre-rainfall 1evel.
The discharge-turbidity transfer function model does not necessarily
imply that discharge level is a strong causal factor for turbidity.
These two variables are correlated largely because when glacial melting
is high,both discharge and turbidity are high.This phenomenon pro-
vides the seasonal trend of the two series;the discharge of clear water
tributaries such as Portage Creek and Indian River (which increases
discharge but not turbidity)is a noise component.However,discharge
does in fact have some direct causal effect on turbidity by resuspending
sediments and other particles during a rapid rise in discharge level.
Certainly turbidity is not a cause of discharge,so it makes sense to
take discharge and noise as the input and turbidity as the output of a
transfer function model.The value of the model is that it allows
levels of turbidity for a few days ahead to be predicted from past
values of both turbidity and discharge.
41
Turbidity level after the dams begin operation will not only be influ-
enced by a changed discharge regime,but will also be directly changed
by the dams because of settling of suspended sediments in the reservoir.
By building Box-Jenkins models for these four time series,a better
understanding of the processes·which control these variables was
developed in that the structure of the random processes which generate
an observed series has now been specified.Also,we have statistically
described the natural time series as a baseline against which future
changes can be assessed.Thi s descri pt i on of the di scharge and tur-
bidity regimes is important not only because of their effects on salmon
outmigration,but also because of their effects on other life stages and
species.It is important to explore the effect on salmon outmigration
of a construction project which will change the basic rules,that is,
change the underlying physical processes.Whereas the present discharge
regime can be described as a mixed first order autoregressive and moving
average process,the di scharge regime under a post-project scenario
could include entirely different terms.
An important point is that the underlying processes (the autoregressive
and moving average components)were essentially the same in 1983 and in
1984 even though the actual time series,or llrealizations,ll looked very
different between the two years.This was true for both discharge and
for chinook salmon outmigration;only a single year of turbidity and
sockeye salmon outmigration was examined.Even though the discharge
peaks do not match between the two years and the mean levels between
years may have been different,the process which generated these peaks
in both years was the same and can be described by an ARMA(l,l)model
with similar parameter estimates for both years.
In a sense,the proposed dams would operate like a gigantic low pass
filter on the discharge regime,dampening out the high frequency pertur-
bations and letting the low frequency (annual cycle)events pass,but at
a reduced amplitude.In other words,there are two effects of intro-
ducing a reservoir into this system:1)the day-to-day changes in
discharge would be smoothed and 2)the general discharge level would be
higher than normal in winter and lower than normal in summer.However,
this is an oversimplification because a new element would be present if
the dams are built -namely,power demand.Power demand is not in phase
with the natural discharge fluctuations,so dam operation to accommodate
power demand wi 11 change the mechani sms whi ch generate the current
discharge regime.
The important question is,how would the salmon outmigration rates be
affected if these di scha rge spi kes were not present,as wi th a dam-
regulated discharge regime?Further,what effects would these changes
have on the population survival rate?Relatively high levels of dis-
charge,and possibly four or five day peaks,in the late spring and
early summer may be necessary to faci 1itate normal outmi grati on timi ng
of juvenile salmon.On the other hand,very high discharge levels at
this time of year,which occur naturally,may be harmful to juvenile
chi nook salmon if these floods di spl ace the fry downstream from what
would otherwise be their rearing areas.
42
-
""'"
-
-
-
-
-
-
-
-
-
.-
.-
Time series analysis is a statistical tool which has many potential
applications to the Susitna River Aquatic Studies Program.It would be
useful to build Box-Jenkins models for the 36 year record of discharge
at Gold Creek gaging station.Because this information is continuous,it
can be digitized as monthly,weekly,daily,or even hourly means.
Turbidity,temperature,and dissolved gas time series could also be
modeled in:this manner.Develop"ing time series models for.the proposed
post-project di scharge regime to see whether the post-project di scharge
regime is also an ARMA(1,l)process would be informative in assessing
d~m-relatedeffects.Intervention analysis,which is an extension of
Box-Jenkins models concerned with a natural or human caused change to a
system,woul d be an appropriate method to use (Box and Ti ao 1975;Hi pel
et ale 1978;Thompson et ale 1982).One could determine if the inter-
vention (construction of the dams)would have a significant effect on
the time series processes.This method has been used to model the
effects of the Aswan dam on the Nile River and of the Gardiner dam on
the South Saskatchewan River in Canada (Hi pel et ale 1978).Before and
after mean levels can not be compared using normal analysis of variance
because the observations are serially correlated..
Developing forecast models for the annual return of adult salmon or the
annual total number of outmigrants would be an excellent use of time
series analysis.The adult salmon return of a particular year is
strongly related to the return of the previous year (that is,when catch
is high one year,it tends to be high for several years)and there is
probably a periodic component based on strong year classes.With such a
model,one could predict the size of next year1s adult salmon return,a
piece of information whi ch woul d be very useful to both fishery and
hydroelectric dam managers.However,the time series of adult salmon
return to the Susitna River is not long enough (only seven or eight
years of data)to develop Box-Jenkins models.A minimum of about 40 or
50 observations is necessary (McCleary and Hay 1980;Huff 1983),
although the method has been applied by Jensen (1985)to fish catch data
with as few as 32 observations.The annual abundance of adult chinook
and coho salmon in the California marine fishery has been successfully
examined with time series analysis by Botsford et aT.(1982)and
Peterman and Wong (1984)have looked at sockeye salmon cycles in British
Columbia and Bristol Bay.For the present,analysis of salmon time
series in the Susitna River will have to be restr·icted to daily rates of
a single year .
43
5.0 ACKNOWLEDGEMENTS
I thank Kent Roth,who has run the outmigrant operation since its
beginning in 1982,and Dana Schmidt, former Project Leader of the
Resident and Juvenile Anadromous Fish project,for their valuable
discussions on some of the ideas in this report.Allen Bingham,Paul
Suchanek,and Dave Bernard also made helpful comments on a draft copy of
the report.
Much of thi s work was done as a projec.t for a course on time seri es
analysis taught by J.Horowitz of the Department of Mathematics and
Stati sti cs,University of Massachusetts.Hi s assistance with the time
series analysis and review of this paper are appreciated.
I am grateful to Mary Ferber of the Al aska Resources Library for con-
ducting a computerized literature search on ecological and fisheries
appl ications of time series analysis.Drew Crawford and Andy Hoffmann
helped compile this report,the figures were drafted by Carol Hepler,
and Skeers Word Processing Services did the typing.
44
-
.~
-
-
-
-
-
• I
6.0 LITERATURE CITED
Botsford,L.W.,R.D.Methot,Jr.,and J.E.Wilen.1982.Cyclic co-
variation in the Cal ifornia king salmon,Oncorhynchus tshawytscha,
silver salmon,O.kisutch,and dungeness crab,Cancer magister,
fisheries.Fishery Bulletin 80:791-801.
Box,G.E.P.,and G.M.Jenkins.1976.Time series analysis.Fore-
casting and control.Holden-Day,San Francisco •
Box,G.E.P.,and GrC.Tiao.1975.Intervention analysis with applica-
tions to economic and environmental problems.Journal of the
American Statistical Association 70:70-79.
Brannon,E.l.,and E.O.Salo.(eds.).1982.Proceedings of the salmon
and trout migratory behavior symosium.June 3-5,1981.University
of Washington,Seattle,Washington.
Bulmer,M.G.1978.The statistical analysis of the ten year cycle.
Pages 141 ..153 in H.H.Shugart,Jr.(ed.).Time Series and Eco-
logical Processes.SIAM-SIMS Conf.Sere 5.Society for Industrial
and Applied Mathematics,Philadelphia,Pennsylvania.
Cederholm,C.J.,and W.J.Scarlett.1982.Seasonal immigrations of
juvenile salmonids into four small tributaries of the Clearwater
River,Washington,1977 -1981.Pages 98 -110 in LL.Brannonand
LO.Salo (eds.).Proceedings of the salmon and trout migratory
behavior symposium.June 3-5,1981.University of Washington,
Seattle,Washington.
Chatfield,C.1984.The analysis of time series:an introduction.
Chapman and Hall.london.
Congleton,J.L.,S.K.Davis,and S.R.Foley.1982.Distribution,
abundance and outmigration timing of chum and chinook salmon fry in
the Skagit salt marsh.Pages 153 -163 in E.L.Bra.nnon and E.O.
Salo (eds.).Proceedings of the salmon and trout migratory behav-
ior symposium.June 3-5,1981.University of Washington,Seattle,
Washington.
Dixon,W.J.,M.B.Brown,L Engelman,J.W.Frane,M.A.Hill,R.I.
Jennrich,and J.D.Toporek.(eds.).1981.BMDP statistical
software.1981.University of California Press,Berkely,
California.
EWT&A.1985.Instream Flow Relationships Report.Volume No.1.
Prepared for Harza-Ebasco Susitna Joint Venture by E.Woody Trihey
and Associates and Woodward-Clyde Consultants,Anchorage,Alaska.
Godin,J-G.Y.1982.Migrations of salmonid fishes during early life
history phases:daily and annual timing.Pages 22-50 .i!!.E.l.
Brannon and E.O.Salo (eds.).Proceedings of the salmon and trout
migratory behavior symposium.June 3-5,1981.University of
Washington,Seattle,Washington.
45
Granger,C.W.J.,and P.Newbold.1977.Forecasting economic time
series.Academic Press,New York.
Grau,E.G.1982.Is the lunar cycle a factor timing the onset of salmon
migration?Pages 184-189 in E.L.Brannon and E.O.Salo (eds.).
Proceedings of the salmon and trout migratory behavior symposium.
June 3-5,1981.University of Washington,Seattle,Washington.
Hale,S.S.1983.Habitat relationships of juvenile salmon outmigra-
tion.Appendix H "in Synopsis of the 1982 aquatic studies and
analysis of fish ana-habitat relationships.Susitna Hydro Aquatic
Studies.Alaska Department of Fish and Game,Anchorage,Alaska.
Hale,S.S.,P.M.Suchanek,and D.C.Schmidt.1984.Modelling of
juvenile salmon and resident fish habitat.Part 7 in D.C.Schmidt,
S.S.Hale,D.L.Crawford,and P.M.Suchanek.leds.).1984.
Resident and juvenile anadromous fish investigations (May -October
1983).Susitna Hydro Aquatic Studies.Report No.2.Alaska
Department of Fish and Game,Anchorage,Alaska.
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 24:2069-2099.
Hipel,K.W.,D.P.Lettenmaier,and A.!.McLeod.1978.Assessment of
environmental impacts.Part one:intervention analysis.Environ-
mental Management 2:529-535.
Hoff,C.1983.A practical guide to Box-Jenkins forecasting.
Wadsworth,London.
Jensen,A.L.1985.Time series analysis and the forecasting of
menhoden catch and CPUE.North American Journal of Fisheries
Management 5:78-85.
Kirkley,J.E.,M.Pennington,and B.E.Brown.1982.A short-tenn
forecasting approach for analyzing the effects of harvesting
quotas:application to the Georges Bank yellowtail flounder
(Limanda ferruginea)fishery.J.Cons.into Explor.Mer.
40:173-175.
Liu,L-M.,and D.M.Hanssens.1980.Identifi cation of multi pl e-i nput
transfer function models.BMDP statistical software.Technical
Report No.68.Los Angeles.
McCleary,R.,and R.A.Hay,Jr.1980.Applied time series analysis for
the social sciences.Sage Publications,Beverly Hills,California.
Mendelssohn,R.1981.Using Box-Jenkins models to forecast fishery
dynamics:identification,est"imation,and checking.Fishery
Bulletin 78:887-896.
46
-
-
-
Mendelssohn,R.1982.Environmental influences on fish population
dynamics:a multivariate time series approach.Paper presented at
a meeting of the American Statistical Association.August,1982.
Cincinnati,Ohio.
Murray,L.C.,and R.J.Farber.1982.Time series analysis of an
historical visibility data base.Atmospheric Environment
16:2299-2308.
Nel son,R.
ing.
1973.Applied time series analysis for managerial forecast-
Holden-Day,San Francisco,California.
,fIi,fOA
~
I
Q'Heeron,M.K.,Jr.,and D.B.Ellis.1975.A comprehensive time series
model for studying the effects of reservoir management on fish
populations.Transactions of the American Fisheries Society
104:591-595.
Peterman,R.M.,and F.Y.C.Wong.1984.Cross correlations between
reconstructed ocean abundances of Bri stol Bay and British Col umbi a
sockeye salmon (Oncorhynchus nerka).Canadian Journal of Fisheries
and Aquatic Sciences 41:1814-1824.
Platt,T.,and K.L.Denman.1975.Spectral analysis in ecology.
Annual Review of Ecology and Systematics 6:189-210.
Priestley,M.B.1981.Spectral analysis and time series.Vol 1:
univariate series,Vol 2:multivariate series,prediction and
control.Academic Press,London.
Raymond,H.L.1968.Migration rates of yearling chinook salmon in
relation to flows and impoundments in the Columbia and Snake
Rivers.Transactions of the American Fisheries Society 97:356-359.
Roth,K.J.,D.C.Gray,and D.C.Schmidt.1984.The outmigration of
juvenile salmon from the Susitna River above the Chul itna River
confluence.Part 1 in D.C.Schmidt,S.S.Hale,D.L.Crawford and
P.M.Suchanek (eds.).-1984.Resident and juvenile anadromous fish
investigations (May -Qctober 1983).Susitna Hydro Aquatic
Studies.Report No.2.Alaska Department of Fish and Game.
Anchorage,Alaska.
Saila,S.B.,1.T.Polgar,D.J.Sheehy,and J.M.Flowers.1972.Corre-
lations between alewife activity and environmental variables at a
fishway.Transactions of the American Fisheries Society
101:583-594.
Saila,S.B.,M.Wigbout,and R.J.lermit.1980.Comparison of some
t"ime series models for the analysis of fisheries data.J.Cons.
into Explor.Mer.39:44-52.
47
Salas,J.D.,and R.A.Smith.1981.Physical basis of stochastic models
of annual flows.Water Resources Research.17:428-430.
Shugart,H.H.,Jr.(ed.).1978.Time series and ecological processes.
Proceedings of SIAM-SIMS Conference.Society for Industrial and
Applied Mathematics,Phil?delphia.
Solomon,D.J.1982a.Migrat'ion and dispersion of juvenile brown and
sea trout.Pages 136-145 in LL.Brannon and LO.Salo (eds.).
Proceedings of the salmon and trout migratory behavior symposium.
June 3-5,1981.University of Washington,Seattle,Washington.
Solomon,D.J.1982b.Smolt migration in Atlantic salmon (Salmo salar
L.)and sea trout (Salmo trutta L.).Pages 196-203 in £oL.
Brannon and E.O.Salo (eds.).Proceed"ings of the salmon arld trout
migratory behavior symposium.June 3-5,1981.University of
Washington,Seattle,Washington.
Srikanthan,R.,LA.McMahon,and J.L.Irish.1983.Time series
analyses of annual flows of Australian streams.Journal of Hydrol-
ogy.66:213-226.
Stevens,D.E.,and L.W.Miller.1983.Effects of river flow on abun-
dance of young chinook salmon~American shad,longfin smelt,and
delta smelt in the Sacramento-San Joaquin River system.North
American Journal of Fisheries Management 3:425-437.
Still,P.J.,R.D.Lamke~J.E.Vaill,B.B.Bigelow~and J.L.VanMaanen.
1984.Water resources data.Alaska.Water year 1983.U.S.G.S.
Water-Data Report AK-83-1.U.S.Geological Survey~Anchorage,
Alaska.
Stocker,M.,and R.Hilborn.1981.Short-term forecasting in marine
fish stocks.Canadian Journal of Fisheries and Aquatic Science.
38:1247-1254.-
..
Suchanek,P.M.,R.P.Marshall,S.S.Hale~and D.C.Schmidt.1984.
Juvenile salmon rearing suitability criteria.Part 3 in D.C.
Schmidt,S.S.Hale,D.L.Crawford~and P.M.Suchanek (eds.)-.1984.
Resident and juvenile anadromous fish investigations (May -October
1983).Susitna Hydro Aquatic Studies.Report No.2.Alaska
DepQrtment of Fish and Game.Anchorage,Alaska.
Taylor,T.G.,and F.J.Prochaska.1984.Incorporating unobserved
cyclical stock movements in fishery catch equations:an applica-
tion to the Florida blue crab fishery.North American Journal of
Fisheries Management 4:67-74.
Sturges,W.1983.On interpolating gappy records for time
analysis.Journal of Geophysical Research.88:9736-9740.
series
-
-
48
-
Thompson,K.W.,M.L.Deaton,R.V.Foutz,J.Cairins,Jr.,and A.C.
Hendricks.1982.Application of time series intervention analysis
to fish ventilatory response data.Canadian Journal of Fisheries
and Aquatic Sciences 39:518-521.
Van Winkle,W.,B.L.Kirk,and B.W.Rust.1979.Periodicities in
Atlantic Coast striped bass (Marone saxatilis)commercial fisheries
data.Journal of the Fisheries Research Board of Canada 36:54-62.
49
7.0 BOX-JENKINS ARIMA AND TRANSFER FUNCTION MODELS
Box-Jenkins models can be summarized as follows (Box and Jenkins 1976;
McCleary and Hay 1980;Chatfield 1984).Suppose there is a time series
Y:t ' t =l..N.Then Y1;is a moving average process of order q (or an
MA(q)process)if
where 9 4 are constants and eo =l.The term at.is a whi te no;se pro-
cess.White noise consists of a series of ranoom shocks,each dis-
tributed normally and independently about a zero mean with a constant
variance.The series Yt is an autoregressive Rrocess of order p (or an
AR(p)process)if
¢.'i t-I +-¢:J.l't-~t-."
t
where 9;are constants.This is similar to a multiple regression model
except that Yt.is regressed not on independent variables but on past
values of itself.A first order autoregressive process,AR(1),has the ~
form:
t tlt
Box and Jenkins (1976)define a backward shift operator Bas:
For m =1,
or,the previous value.
-50
,~
Using B,the AR(I)equation can be written:
Time series resulting from a mixture of ARand MA processes are called
ARMA(p,q)models and have the form:-t .. .
+t +
,..,..
-
6.2L t _1 t-.'.+e t 2t t -'t
Using the backward shift operator B,an ARMA (1,1)may be written as:
ARMA (p,q)models are appropriate only when the time series is station-
ary.Stationary in an ARMA model means that there is no systematic
change in the mean or the variance over time and that there are no
strictly periodic variations (Chatfield 1984);in other words,the mean,
variance,and autocovaria.nce are not dependent on time.Time series
which are not stationary can sometimes be handled by IIdifferencing"the
series.Taking the difference of adjacent values gives a differencing
order,d,of one:
)
Such models are said to be "integrated ll and are denoted by ARIMA(p,d,q)
where pis the order of the autoregressive component,d is the order of
differencing,and q is the order of the moving average component.
Time series with seasonal variations,such as would occur in a multiple
year series of daily water temperature measurements,can be made sta-
tionary by seasonal differencing.For example,the value for April 15 of
one year is subtracted from the value for Apri 1 15 of the fo 11 owi ng
year,and so on for all days of the year.
It has been assumed above that the time series had a mean value of zero.
With stationary time seri es whi ch have a non-zero mean,the mean has to
51
be subtracted from every y i term.For exampl e,the form of an AR{I)
model would be:
1't::',+4>I (~t -I -,..)+~t
The autocorrelation function plays a major role in identifying and
building time series models.A regular correlation coefficient measures
the correlation between N pairs of observations on two variables.The
autocorrelation coefficient is somewhat similar except that it measures
the correlation between all observations of the same variable at a given
distance apart in time (that is,between Yt and 'ft.-jc.for all values of
t!t where k =time lag).Also,the covariance is estimated only over N-k
pairs of observations (McCleary and Hay 1980).Autocorrelation coeffi-
cients at different lags indicate the extent to which one value of the
series is related to previous values and can be used to evaluate the
duration and the degree of the "memoryll of the process.The autocorre-
lation function (ACF)is the set of autocorrelation (AC)coefficients at
different lags associated with a time series;a plot of the ACF is
called a correlogram (Chatfield 1984).
The ACF is defined as:
(OV'fU-i 4'1({.(Yt )YttAl
V Hi'1lo~(e.C.Yt.')
and is estimated by:
-
-
-
ACr.k.-
N
•N-Jt
A partial autocorrelation (PAC)coefficient measures the excess corre-
lation at lag k which is not accounted for by an autoregressive model of
order k-l.The set of PAC I S at different 1ags associated with a time
series is called the partial autocorrelation function (PACF).~
There are three steps in developing an ARIMA model:model identifica-
tion,parameter estimation,and diagnostic checking (Box and Jenkins
1976).ARIMA model building is an iterative process.The first thing to
do is to look at a plot of the time series.Time series that are not
stationary must be made so by trend removal which can be accompl ished by
52
r-
i
such methods as differencing the series or by polynomial (or other)
regression.Examination of the autocorrelation function (ACF)and the
partial autocorrelation function (PACF)of a stationary series helps to
identify a possible ARIMA model.The next step is to estimate the
parameters of the model and again examine the ACF and PACF plots,this
time on the residuals from the model.This process is repeated until the
residuals show no significant AC's or PAC's at any lag,which indicates
that the residuals consist of only a white noise process.
When there is an independent variable which is also a time series,a
transfer function model can be developed.This model consists of the
transfer function component from the independent variable as well as the
ARIMA component (or noise component)from the dependent variable
(McCleary and Hay 1980)and can be represented as:
where:Yt is the output time series
Xt is the input time series
f(X t -b )is the transfer function component
Ntis the noi se or ARIMA component
Transfer function models can be bivariate (when there is one independent
variable)or multivariate (more than one independent variable).
The steps to take in developing a transfer function model (Box and
Jenkins 1976;McCleary and Hay 1980;Dixon et al.1981)are:(1)develop
an ARIMA model for the input series,obtaining the pre-whitened "input
(residuals),(2)filter the output series by the model for the input
series,(3)cross-correlate the residuals from the first two steps,(4)
identify the form of the transfer function component from the cross
correlation function,(5)assuming the errors are white noise,estimate
the values for the parameters,(6)identify an ARIMA model for the
residuals,(7)if the ARIMA component is not white noise,combine the
ARIMA component with the transfer function comRonent to form a new
model,(8)estimate the parameter values,and (9)examine the ACF and
PACF plots on the residuals from the new model to see if the model is
adequate.
53
PART 2
The Relative Abundance,Distribution,and Instream
Flow Relationships of Juvenile Salmon
in the Lower Susitna River.
r
I
-
THE RELATIVE ABUNDANCE,DISTRIBUTION,AND INSTREAM
FLOW RELATIONSHIPS OF JUVENILE SALMON
IN THE LOWER SUSITNA RIVER
Report No.7,Part 2
by Paul M.Suchanek,Karl J.Kuntz,and John P.McDonell
Alaska Department of Fish and Game
Susitna River Aquatic Studies Program
620 East 10th Avenue,Suite 302
Anchorage,Alaska 99501
ABSTRACT
Juvenile salmon abundance and distribution were studied in the lower
Susitna River (below the Chulitna River confluence)and juvenile salmon
rearing habitat was modelled at 20 sites within the reach.Chinook,
chum,and sockeye salmon juveniles made use of side channels;however,
high turbidity limited use of side channels located in the Chulitna
River plume.Coho salmon juveniles were found primarily in tributary
mouths;sockeye~chinook,and chum salmon also were present in these
areas.Sloughs,which were limited in occurrence,were not used heavily
by any of the salmon species.
Both tributary mouths and side channel/slough sites were modelled using
one of two habitat models.At tributary mouths,an.increase in weighted
usable area with a rise in mainstem discharge resulted from the forma-
tion of backwater areas which led to lower velocities and an expansion
of the area and amount of cover inundated.At si de channel s ~chi nook
weighted usable area increased after overtopping due to a gain in cover
suitability (turbidity)~velocity,and area.The weighted usable area
response to a ri se in rna i nstern di scha rge for sockeye and chum salmon
juveniles at side channels was also usually positive.Habitat indices
at side channels for chinook~chum,and sockeye juveniles at mainstem
discharges and side channel flows above the overtopping discharge
declined as velocities became unsuitably high.Weighted usable area for
these species did not always decline at high discharges,however,
because of the compensating effect of a larger surface area.
i
TABLE OF CONTENTS
ABSTRACT ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••".i
LIST OF FIGURES....................................................iv
LIST OF TABLES.....................................................vii;
LIST OF APPENDIX FIGURES...........................................ix
LIST OF APPENDIX TABLES............................................xv
1.0 INTRODUCTION.. . •. . . •... . . . . •. •. •. •. ••••. •. •••. ...•. •. . . . . . . •. . . 1
2.0 METHODS •••••IiI.................................................3
2.1 Field Sampling Design ,.......................3
2.1.1 Study locations and selection criteria .•..•.•..••~....•.
2.1.2 Field data collection •.........................•........
2.1.2.1 Resident Juvenile Habitat (RaHAB)model sites •......
2.1.2.2 Instream Flow Incremental ~1ethodology (IFIM)sites •.
2 ..1.2.3 Opportunistic sites ••••••••••••••••.•••~•••••.•••••.
3
5
5
9
9
-
2.1.3 Schedule of activities and frequency of sampling..•••.•.10
2.2 Data Analysis -....10
2.2.1 Physical data -........................10
2.2.2 Abundance and distribution..............................11
2.2.3 HabitQt modelling of rearing salmon........•.••.•••••.••11
2.2.3.1 Suitability criteria development....................11
2.2.3.2 Instream Flow Incremental Methodology (IFIM)models.11
2.2.3.3 Resident Juvenile Habitat (RJHAB)models............12
2.2.3.4 Mode-l verification..................................14
3.0 RESULTS •••••~•••••••••. •••. ••. . . . •. ••••••. ••. •••••••. •• . •. . ••.15
3.1 Seasonal,Spatial,and Discharge Related Variations
i n Ha bita t.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
3.1.1 Macrohabitat type classifications of study sites........15
3.1.2 Chulitna and Talkeetna River plume influences
on turbidity of side channels...........................19
3.1.3 Physical responses of sampling sites to
mainstem discharge variations...........................19
3.1.3.1 Area 19
3.1.3.2 Cover...............................................21
ii
TABLE OF CONTENTS (Continued)
3.2 Distribution and Abundance of Juvenile Salmon...............25
3.2.1 Chinook salmon •.•.•.•.•........•.•.•.........o..........25
3•2•2 Co h0 sal ma n••••.•••••••••••••II Q ••t.'I • ••• •25
3.2 CI 3 Chum sa 1man •••.••••••••••.•a 0 ••Ii'III •••••••II Q ••••0 CI ••••II • • •28 ~
3.204 Sockeye salmon •••...••Go ••••••••••••0 0 •••••••••••••••GI...28
3.3 Habitat Modelling of Rearing Juvenile Salmon................33
I
3.3.1 Chinook salmon ee e o a.........39
3.3.2 Coho salnlon a 0 ••••••••••••••••e ••••••"...42
3.3.3 Chum salmon ••••.••••••..••.•••.••ecoe ••"'••••••••••••••o...48
3.3.4 Sockeye sa 1man ••••••••.•••co II e 1&III GI a II a 0 ••e ....eo.II •GI •II •••co ••54
4.0 DISCUSSION ••••••••••••••••..•••••••••s~••••o •••oog.e..••••••••64
4.1 Chi nook Sa 1man •••III ••••••.,•••••e e G co ..1:1 e •D •••G ea •••••1:1 Cit .....Q • ••64
4.2 Coho Salmon co ••.,0 a ••Q ••a.GI ••.,•••e 'II e •••••co Cl •e g ••••"•Q a ••II!Io .,• • ••65
4.4 Sockeye Salmon ••••••••••••••e ...e.".10 0 co 8 •••.,••ill ••e 8 ill).,p ••0 CI 0..68
5.0 CONTRIBUTORS •••••••••••••••••••o ••••••e ••••e.ee.e.coo •••080.e.,.70
6.0 ACKNOWLEDGEMENTS ••••••••e e ••••••••••••D •••Q •Cl •Cl •e ••0 0 ••••"•-II • •72
7.0 LITERATURE CITED .•••••••••••••.•••••.•••.,D-.o ••ClCl-.co •••oo ...o ••"o.73
8.0 APPENDICES
Appendix A Lower Susitna River Juvenile Salmon
Rearing Suitability Criteria
Appendix B Modelled Site Turbidities,
Juvenile Salmon Catches,Areas,
Weighted Usable Areas,and
Habitat Indices
Appendix C Comparison of the IFIM and RJHAB
Modelling Techniques at Two Selected
Sites
Appendix 0 Hydraulic Models for Use in Assessing
the Rearing Habitat of Juvenile
Salmon in Six Side Channels of the
Lower Susitna River
iii
-
-
"'..
LIST OF FIGURES
Figure Title Page
-1 Location of study sites on the lower Susitna
River at which juvenile salmon habitat was
modelled,June through October 1984....................4
2 Arrangement of transects and sampl ing cell s
within a grid at a hypothetical RJHAB model-
ling site 6
.....
3
4
Turbidities at modelled side channels and
tributary mouths on the lower Susitna River,
June through October 1984..............................18
Compari son of turbi di ti es in the lower
Susitna River below the Chulitna and Tal-
keetna River confluences on July 19 and
August-16.,.1984 .•............•...•.•...........•....•.•20
5 Area within modelled tributary mouths as a
function of mainstern discharge at the USGS
Sunshine gaging station,1984..........................22
-
-
,.....
6
7
8
9
10
11
Area within modelled sloughs and side chan-
nels as a function of mainstem discharge at
the USGS Sunshine gaging station,1984.................23
Instream cover response at Beaver Dam Slough,
Rolly Creek,and Caswell Creek mouths as a
fun~ti on of rna i nstem di scharge at the USGS
Sunshine gaging station,1984 •..••.•.•....•...•....•..•24
Seasonal distribution and relative abundance
of juvenile chinook salmon on the lower
Susitna River,June through mid-October 1984 26
Juvenile chinook salmon mean catch per cell
at si de channel s and tributary mouths on the
lower Susitna River by sampling period,June
through mid-October 1984 .•....•...............•.....••.27
Juvenile chinook salmon mean catch per cell
at modelled side channels on the lower
Susitna River by turbidity increment,June
through mid-October 1984 ...•......•.•...•.......•..•...27
Seasonal distribution and relative abundance
of juvenile coho salmon on the lower Susitna
River,June through mid-October 1984 29
iv
LIST OF FIGURES (Continued)
Fi gure Title
-
--
12 Juvenile coho salmon mean catch per cell at
four tributary mouths on the lower Susitna
River by sampling period,June through
mid-October 1984 D •••••••••••o ••••••30
.....
13 Seasonal distribution and relative abundance
.of juvenile chum salmon on the lower Susitna
River,June through mid-October 1984 ••••••••.••••••••••31
14
15
Juvenile chum salmon mean catch per cell at
modelled side channels and tributary mouths
on the lower Susitna River by sampling
period,June through mid-October 1984 ••.•.••••••••...••32
Juvenile chum salmon mean catch per cell at
modelled side channels on the lower Susitna
River by turbidity increment,June through
mid-July 1984 e •••_32
""""
-
-
20
18
19
16 Seasona 1 di stri buti on and re1 ative abundance
of juvenile sockeye salmon on the lower
Susitna River,June through mid-October 1984 •••.•...•••34
17 Juveni 1e sockeye salmon mean catch per cell
at side channels and tributary mouths on the
lower Susitna River by sampling period,June
through mid-October 1984 •••••••••••••••••••••••••••••..35
Juveni 1e sockeye salmon mean catch per cell
at modelled side channels on the lower
Susitna River by turbidity increment (with
and wi thout Beaver Dam Si de Channel),June
through mid-October 1984 •••••••••.••••..••••••••.•••.••.35
Juvenile sockeye salmon mean catch per cell
at Beaver Dam Slough,Beaver Dam Si de Chan-
nel,and Rolly Creek Mouth by samp1 ing
period,June through mid-October 1984 •••••••••••.••••••36
Weighted usable area for juvenile chinook
salmon at the Rolly Creek Mouth,Kroto Slough
Head,and Sucker Side Channel study sHes as
a function of mainstem discharge,1984 ••••••••••••....•40
21 Wei ghted usab1 e area and habitat i ndi ces for
juvenile chinook salmon at tributary mouth
study sites as a function of mainstem dis-
charge,1984...........................................41
v
-
-
.....
LIST OF FIGURES (Continued)
Figure
22
23
24
Title Page
Weighted usable area and habitat indices for
juvenile chinook salmon at side channell
slough study sites as a functfon of mainstem
dischar.9'e,1984 -............43
Turbidity adjusted weighted usable area and
habitat indices for juvenile chinook salmon
at side channel/slough study sites as a
function of mainstem discharge,1984 .•....•....•....•..45
Juvenile chinook salmon mean catch per cell
versus seasonal mean habitat indices at side
channel and tributary mouth modelling sites
on the lower Susitna River,1984.......................46
,""",
26
I~
27
25 Weighted usable area for juvenile coho salmon
at the Caswell Creek,Rolly Creek,and Beaver
Dam Slough tributary study sites as a func-
tion of mainstem discharge,1984 ...•.•.....•...•.......47
Weighted usable area and habitat indices for
juvenile coho salmon at tributary mouth study
sites (excluding Birch Creek)as a function
of mainstem discharge,1984........•..•....•.•..•••....49
Juvenile coho salmon mean catch per cell
versus seasonal mean habitat indices at
tributary mouth modelling sites on the lower
Susitna River~1984,......•.•.•.•....•,......•.•..•.•....50
28 Weighted usable area for juvenile chum salmon
at the Rustic Wilderness and Last Chance Side
Channel study sites as a function of mainstem
discharge,1984 ......................•.................51
29 Weighted usable area for juvenile chum salmon
at the Trapper Creek and Sunset Si de Channel
study sites as a function of mainstem dis-
ch'.arge,1984...........................................52
-
-
30
31
Weighted usable area and habitat indices for
juvenile chum salmon at side channel/slough
study sites as a function of mainstem dis-
charge,1984 53
Turbidity adjusted weighted usable area and
habitat indices for juvenile chum salmon at
side channel/slough study sites as a function
of mainstem discharge,1984............................56
vi
33
32
Figure
37
LIST OF FIGURES (Continued)
Title Page
Juvenile chum salmon mean catch per cell
versus seasonal mean habitat indices at side
channel and slough modelling sites on the
lower Susitna River,1984 ••...•.•'.~~•.•...•.......•..•.57
Weighted usable area for juvenile sockeye
salmon at the Rolly Creek Mouth and Sucker
Side Channel study sites as a function of
mainstem discharge,1984 .•......•.•.....•.•..••..•.....58
34 Weighted usable area for juvenile sockeye
salmon at the Beaver Dam and Sunset Side
Channel study sites as a function of mainstem
discharge,1984 10................59
35 Weighted usable area and habitat indices for
juvenile sockeye salmon at tributary mouth
study sites on the lower Susitna River as a
function of mainstem discharge,1984 .••.•••••••.••••••.60
36 Weighted usable area and habitat indices for
juvenile sockeye salmon at side channel and
slough study sites on the lower Susitna River
as a function of mainstem discharge,1984 .•.•••••••..••61
Juvenile sockeye salmon mean catch per cell
versus seasonal mean habitat indices at side
channel and tributary mouth modelling sites
on the lower Susitna River,1984 •.....••••••.•.....••••63
vii
~I
-
~,
LIST OF TABLES
Table Title
1 Percent cover and cover type categories................8
2 Partitioning of wetted channel width into
s tream ce 1.1s... . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . .. .13
-
3
4
Classifications and habitat characteristics
of study sites on the lower Susitna River at
which juvenile salmon habitat was modelled,
June through October 1984..............................16
Percentages of lower river habitat modelling
sites associated with nine cover-type cat-
eg,orles................................................17
7
8
-
.....
-
.-
5 Evaluation of R.JHAB model quality for extrap-
olating WUAs over the range of 12,000 to
75,000 cfs as measured at Sunshine gaging
s ta t ion ,.1984 .......•.•.•......'..'.....•......~. . . . . . . . . .3~
6 Discharge ranges of IFIM models at lower
Susitna River sites for which hydraulics are
rated acceptable,1984 ••••••••••.•••.•••••••.•.•.•••...38
Preliminary juvenile chinook salmon turbidity
criteria derived from lower Susitna River
side channel distribution data for turbidi-
ties greater than 100 NTU •.••••••••••.•••••.•.•••••.•..44
Weighting factors for turbidity by side
channel site for analysis of juvenile chinook
salmon habitat use,1984...............................44
9 Weighting factors for turbidity by site for
analysis of juvenile chum salmon habitat use,
1984 •.•.••.••••.••••••••••••••••.•.•'....................5'5
viii
LIST OF APPENDIX FIGURES
Appendix
Figure
A-I
A-2
A-3
A-4
A-5
Title Page
Mean catch of juvenile chinook salmon per
cell by percent cover category (bars)in
clear water of the lower Susitna River,1984
and comparison of fitted suitability indices
(lines)calculated in 1984 and for the middle
Susitna R1-ver,1983 A-7
Mean catch of juvenile chinook salmon per
cell by velocity intervals (bars)in clear
water of the lower Susitna River,1984 and
fitted suitability index (line)developed for
the middle Susitna River,1983 ••••••••••0...............A-8
Mean catch of juvenile chinook salmon per
cell by velocity intervals (bars)in clear
water of the lower Susitna River,1984 and
fitted suitability index (line)developed for
turbid water in the middle Susitna River,
1983 •••••••••.•••••••••••••••••••••••••••••••••••••••••A-8
Comparison of cover type suitability indices
for juvenile chinook salmon in clear water
calculated from 1984 lower Susitna River
distribution data and 1983 middle Susitna
River distribution data ••••••••••••••••••••••••••••••••A-9
Cover type suitability indices for juvenile
chinook salmon in clear water calculated from
1984 lower Susitna River distribution data •••••••••••••A-I0
.-
-
-
A-6 Mean catch of juveni 1e chi nook salmon per
cell by depth intervals (bars)in clear water
of the lower Susitna River,1984 and fitted
suitability index (line)developed for the
middle Susitna River,1983 •••••••••••••••••••••••••••••A-II 0
A-7 Mean catch of juveni 1e chinook sa.l man per
cell by depth intervals (bars)in clear water
of the lower Susitna River,1984 •••••••••••••••••••••••A-12
A-8 Mean catch of juvenile chinook salmon per
ce 11 by velocity i nterva 1s (bars)in turbi d
water of the lower Susitna Ri ver,1984 and
fitted suitability index (line)developed for
the middle Susitna River,1983 •••••••••••••••••••••••••A-15
ix
~,
.-
LIST OF APPENDIX FIGURES (Continued)
Appendix
Fi gure
A-9
Title
Me.an catch of juvenile chinook salmon per
cell by percent cover category (bars)in
turbid water of the lower Susitna River,1984
and fitted suitability index (line)calcu-
lated for the middle Susitna River,1983 ••••••••••.••••A-IS
A-I0 Cover type suitability indices for juvenile
chinook salmon in turbid water developed from
1984 lower Susitna River chinook clear water
distribution data -.....•.....•.•....•...........A-I?
-
-
-
A-ll
A-12
A-13
A-14
Mean catch of juvenile chinook salmon per
cell by depth intervals (bars)in turbid
water of the lower Susitna River,1984 and
fitted suitability index (line)developed for
the middle Susitna River,1983 •••••••••••••••••••••••••A-I?
Mean catch of juvenile chinook salmon per
cell by depth intervals (bars)in turbid
water of the lower Susitna River,1984 •••••••••••••••••A-18
Mean catch of juvenile coho salmon per cell
by velocity intervals (bars)in the lower
Susitna River,1984 and fitted suitability
index (line)developed for the middle Susitna
Riv-er,19·83 _ _ ' '...•.................A-21
Mean catch of juvenile coho salmon per cell
by percent cover category (bars)in the lower
Susitna River,1984 and comparison of fitted
suitabi 1 ity i ndi ces (l i nes)calculated in
1984 and for the middle Susitna River,1983 ••••••••••••A-21
A-IS Comparison of cover type suitability indices
for juvenile coho salmon calculated from 1984
lower Susitna River distribution data and
1983 middle Susitna River distribution data ••••••••••••A-22
A-16 Cover type sui tabil ity indices for juveni 1e
coho salmon calculated from 1984 lower
Susitna River distribution data ••••••••••••••••••••••••A-23
-
A-I?Mean ca tch of j uven i 1e coho sa 1mon per cell
by depth intervals (bars)in clear water of
the lower Susitna River,1984 and fitted
su i tabi 1 i ty index (1 i ne)developed for the
middle Susitna River,1983 ••••••••••••••.•••••••••••••••A-24
x
--------------------------------------
LIST OF APPENDIX FIGURES (Continued)
Appendix
Figure Title
A-18 Proportion of cells with juvenile sockeye
salmon present by velocity intervals (bars)
in the lower Susitna River,1984 and fitted
suitability index (line)developed for the
middle Susitna River,1983 and revised in
1984 for the lower river using professional
judgement.•. •. •••••. ••••••••. . ••••••••••••••••••. ••••••A-26
A-19 Proportion of cells with juvenile sockeye
salmon by percent cover category (bars)in
the lower Susitna River,1984 and comparison
of fitted suitability indices (lines)cal-
culated in 1984 and for the middle Susitna
River,1983 ••••••••••••.•••.••••••••••••••••••.•••••••.A-27
-
A-20 Comparison of cover type suitability indices
for juvenile sockeye salmon calculated from
1984 lower Susitna River distribution data
and 1983 middle Susitna River distribution
data ••••.•...••••••••••••••••••••••••••.•••••••••••••..A-27
A-21 Proportion of cells with juvenile sockeye
salmon present by depth intervals (bars)in
the lower Susitna River,1984 and fitted
suitability index (line)developed for the
middle Susitna River,1983 •••••••••••••.•••••••••••••••A-28
A-22
A-23
Proportion of cells with juvenile chum salmon
present by velocity intervals (bars)in the
lower Susitna River,1984 and fitted suita-
bility index (line)developed for the middle
Susitna River,1983 .••••••••••••.••••••••••••..••••••••A-30
Proportion of cells with juvenile chum salmon
present by percent cover category (bars)in
the lower Susitna River,1984 and fitted
suitability index (line)calculated for the
middle Susitna River,1983 ••••••••••..••.•.•.••••••••••A-31 .....
A-24 Proportion of cells with juvenile chum salmon
present by cover types (ba rs)in the lower
Susitna River,1984 A-32
A-25 Proportion of cells with juvenile chum salmon
present by depth intervals (bars)in the
lower Sus itna River,1984 and fi tted suita-
bi 1 ity index (l i ne)developed for the mi ddl e
Susitna River,1983 A-33
xi
Title
-
LIST OF APPENDIX FIGURES (Continued)
Appendix
Figure
B-1 Weighted usable area for juvenile chinook
sa lmon at the Caswell Creek and Beaver Dam
tributary study sites·as a function of
mainstem discharge ••........•..•..•....................B-12
B-2 Weighted usable.area for juvenile chinook
salmon at the Hooligan and Bearbait Side
Channel study sites as a function of mainstem
discharge ......•.•.•••..........•......................B-13
.-
-
B-3
B-4
Weighted usable area for juvenile chinook
salmon at the Last Chance and Rustic Wilder-
ness Side Channel study sites as a function
of rna instem di scharge ...•..............................B-14
Weighted usable area for juvenile chinook
salmon at the Island Side Channel and
Mainstem West Bank study sites as a function
of rnainstem d,i-scha-rge .•.•........•...•.........•'8-15
.....
B-5 Weighted usable area for juvenile chinook
salmon at the Goose 2 and Circular Side
Channel study sites as a function of mainstem
discharge B-16
-
.-.
-
B-6
B-7
B-8
B-9
Weighted usable area for juvenile chinook
salmon at the Sauna and Beaver Dam Side
Channel study sites as a function of mainstem
di scha,rge B-17
Weighted usable area for juvenile chinook
salmon at the Sunset and Sunrise Side Channel
study sites as function of mainstem dis-
charge •.•.......••••........•.•.......•.........•...•..B-18
Weighted usable area for juvenile chinook
salmon at the Trapper Creek Side Channel
study site as a function of mainstem dis-
charge ......................•............•.•...........B-19
Weighted usable area for juvenile chum salmon
at the Hooligan Side Channel and Kroto Slough
Head study sites ad a function of mainstem
di scha rge ,_. . . . . . .... . . . . ...8-23
-
B-I0 Weighted usable area for juvenile chum salmon
at the Bearbait and Island Side Channel study
sites as a function of mainstem discharge 8-24
xii
LIST OF APPENDIX FIGURES (Continued)--
Appendix
Figure Title
B-ll Weighted usable area for juvenile chum salmon
at the Mainstem West Bank and Goose 2 Si de
Channel study sites as a function of mainstem
di scharge ..••••...•.••••••......••...•..~••••.•.....•.•B-25
B-12 Weighted usable area for juvenile chum salmon
at the Circular and Sauna Side Channel study
sites as a function of mainstem discharge •.•.••.•.•....B-26
B-13 Weighted usable area for juvenile chum salmon
at the Sucker and Beaver Dam Si de Channel
study sites as a function of mainstem dis-
charge ...•...••••.•..••.•.•.•••....•••.•.•.•.•.•.•••..•B-27
-
8-14
8-15
Weighted usable area for juvenile chum salmon
at the Sunrise Side Channel study site as a
function of mainstem discharg~....•••.•.••.•••••.....••B-28
Weighted usable area for juvenile sockeye
salmon at the Caswell Creek and Beaver Dam
tributary study sites as a function of
mainstem discharge .••••.•••.•••...•••.•.•..•...•••••.••B-30 ""'"
8-16 Weighted usable area for juvenile sockeye
salmon at Sunrise Side Channel as a function
of mainstem discharge •.•••••...•••••.•.••.•.•••••..•••.B-31
C-l
C-2
C-3
C-4
Comparison of site areas calculated with the
RJHAB and IFIM model 1i ng techniques for the
Trapper Creek and Island side channel study
sites .••...••••••.••.•••••••••••.•.•.•.••••••.•....••••C-2
Comparison of weighted usable areas calcu-
lated with the RJHAB and IFIM modelling
-techniques for juvenile chinook and chum
salmon at Trapper Creek Side Channel,1984 •...•.•..••.•C-3
Comparison of habitat indices calculated with
the RJHAB and IFIM modell ing techniques for
juvenile chinook and chum salmon at Trapper
Creek Side Channel,1984 .•.••••••.••••••••••.•.•.•.••••C-4
Comparison of weighted usable areas calculat-
ed with the RJHAB and IFIM modell ing tech-
niques for juvenile chinook and chum salmon
at Island Side Channel,1984 ••••••••••.••••....••...•.•C-5
xiii
-
-
-
LIST OF APPENDIX FIGURES (Continued)
Appendix
Figure Title
-
,.....
.....
-
C-5 Comparison of habitat indices calculated with
the RJHABand IFIM modelling techniques for
juvenile chinook and chum salmon at Island,
Side Channel,1984 ••.•.•••••.•.••.~......•.••.•.....••.C-6
xiv
LIST OF APPENDIX TABLES
Appendix
Table Title
A-I Percent cover and cover type categories ...•.•....••....A-2
A-2 Kendall correlation coefficients between
habitat variables and chinook catch by cell
(N=744)for all gear types,in turbid water .•....•.....A-5
A-3 Kendall correlation coefficients between
habitat variables and chinook catch by cell
(N=396)for all gear types,in clear water •............A-14
A-4 Calculations of turbidity factors for 1984
lower river data .....•.••.•.......•.....••.••....•.....A-16
A-5 Kendall correlation coefficients between
habitat variables and coho catch by cell
(N=345)in clear water ...•••......•...•••.••.......•.••A-2D
A-6 Kendall correlation coefficients between
habitat variables and sockeye catch by cell
(N=922)-A-26
A-7 Proportional presence of sockeye salmon
associated with the composite weighting
factor calculated by multiplying velocity and
cover suitabilities together .•..•........•••....•••••..A-29
A-8
A-9
A-I0
Kendall correlation coefficients between
habitat variables and chum catch by cell
(N=249)for all gear types,turbidity below
200 NTU................................................A-3D
Proportional presence of chum salmon fry
associated with several composite weighting
factors.•.•••. .•.•••.. .. •••••••••..••••.•.. .•.•••••.••.A-33
Summary of revisions of 1983 middle river
juvenile salmon criteria for use in the lower
Susitna River,1984 ••....••.••.•.•••••••.••..•••..•.•••A-35 """"
A-ll Suitability indices for juvenile salmon for
velocity,depth,and cover in the lower
Susitna River,1984 ••...•...•.........•..•••.•.........A-36
B-1 Turbidities within modelled side channels of
the lower Susitna River,June through August
1984 .•...................III ••• ••• ••••••••••••••• • • • • • • • •B-3
xv
-LIST OF APPENDIX TABLES (Continued)
-
Appendix
Table
B-2
B-3
Title Page
Catch and catch per ce 11 ,(CPUE)of juvenil e
salmon within lower Susitna River sampling
si te s,1984............................................B-4
Lengths of RJHAB model sites in the lower
Susitna River,1984 .••••••••~•••.•••••••.••••...•••••.•B-5
B-4
-
,.,..
-
-
Side channel flows at the 15 modelled side
channels as a function of mainstem discharge B-6
B-5 Weighted usable areas and habitat indices for
juvenile chinook salmon in lower Susitna
Rive r mode 1 site s,1984................................B-9
B-6 Weighted usable areas and habitat indices for
juvenile coho salmon in lower Susitna River
model sites,1984 ...•..~.......•.•••.•.•.•....•........B-20
B-7 Weighted usable area and habitat indices for
juvenile chum salmon in the lower Susitna
River model sites,1984 ............•••..•....•.........B-20
B~8 Weighted usable areas and habitat indices for
juvenile sockeye salmon in lower Susitna
River model sites,1984 •...•.•...••..••••......•.•.....B-29
xvi
-
-
-
,-
~,
-
1.0 INTRODUCTION
The Susitna Ri ver Aquatic Studies Program juveni 1e anadromous di stri-
bution and abundance studies initiated during 1981 and 1982 outlined the
general distribution patterns of juvenile salmon and their habitat
utilization within the Susitna River (ADF&G 1981a,1981b,1983a,1983b).
The 1982 studies also investigated the response of selected areas to
mainstem discharge changes and demonstrated species differences in the
use of "hydraulic zones"(ADF&G 1983c).These zones were subsections of
slough and tributary mouth areas.Some zones were affected by mainstem
backwater,other zones were above the backwater,and other zones
included mixing areas of the mainstem with slough or tributary flow.
The relative use of the hydraulic zones by each species of juvenile
salmon was analyzed to provide an incremental index of habitat availa-
bility at each site for each species.This analysis provided evidence
that the relative use by juvenile salmon of these sites was affected by
changes in mainstem discharge.Also,the distribution of juvenile
salmon suggested certain microhabitat factors within the zone such as
turbidity and the amount of instream cover responded to discharge
changes at a higher rate than did zone surface area.
Studies conducted during the 1983 open-water season concentrated on the
instream flow relationships of juvenile salmon in the middle reach of
the Susitna River between the Chulitna River confluence and Devil Canyon
(Schmidt et al.1984).Suitability criteria for juvenile salmon were
developed and these were used in two types of habitat models to model
the site-specific response of juvenile salmon habitat to variations in
mainstem discharge.Additional information was gathered on juvenile
salmon abundance and distribution in the middle reach.
The 1983 studies suggested that juvenile chinook salmon made heavy use
of mainstem side channels and used the turbid water in these areas as
cover.Juvenile coho,chum,and sockeye salmon tended to occupy areas
that were less influenced by mainstem flow.
In the Susitna River below the Chulitna River confluence (lower river),
the braided nature of the river and lower gradient provides large
amounts of potential side channel habitat for juvenile salmon.A study
plan was formulated,therefore,to examine juvenile salmon distribution
and the usability of different morphological components of the lower
Susitna River for juvenile salmon during the 1984 open-water season.
The results of these studies,which include the responses of rearing
juvenile salmon and their habitat within these morphological components
to variations in mainstem discharge,are detailed in this paper.These
results will be integrated with responses of side channel and slough
complex wetted surface areas to variations in mainstem discharge in
order to estimate the response of juvenile salmon habitat in the lower
river to flow regulation.
Large scale aerial mapping of lower Susitna River side channel and
slough complex changes in area with variations in mainstem discharge has
been done by Ashton and Klinger-Kingsley (1985).Habitat types identi-
fied in the mapping included tributaries,tributary mouths,side
1
sloughs~primary side channels~secondary side channels~clearwater
areas~and turbid backwaters.Tributaries,tributary mouths t and side
sloughs were defi ned as in the mi ddl e ri ver by Kl i nger and Tri hey
(1984).Primary side channels have characteristics similar to the
mainstem in the middle river and therefore offer little potential
habitat for juvenile salmon and are not discussed in this report.
Turbi d backwaters are unbreached channel s whi ch contai n turbid water
from being overtopped at higher mainstem discharges and therefore are a
transitional habitat type between secondary side channels and side
sloughs or clearwater areas.Turbid backwaters are not addressed in
this report but their habitat values are probably similar to barely
breached side channels.Clearwater areas were also not sampled but are
thought to have habitat value similar to that of side sloughs.
The major emphasis of this report is the evaluation of juvenile salmon
use of secondary side channels and their related habitat values.Some
of the larger secondary side channels are considered primary side
channels at higher mainstem discharges.Juvenile salmon use of tribu-
tary mouths and side sloughs was also evaluated.The macrohabitat
evaluation data presented here will be integrated with the aerial
mapping data contained in Ashton and Klinger-Kingsley (1985)in later
reports to formulate the reach-wide response of juvenile salmon habitat
to discharge variations.
2
"""
-
-
-
-
--
-
2.0 METHODS
2.1 Field Sampling Design
Three Juvenile Anadromous Habitat Study (JAHS)field crews,composed of
two biologists,examined rearing habitats used by juvenile salmon at
selected side channels,tributary mouths,sloughs,and mainstem sites of
the Susitna River between the Yentna River confluence (RM 28.5)and
Chulitna River confluence (RM 98.5).JAHS sampling was conducted from
river boats during the open-water season,with helicopter support
enl i sted as needed.The crews operated out of camps located on the
Susitna River at the Oeshka River (RM 40.6),Sunshine Station (RM 79.0),
and Ta 1keetna (RM 97.5).
The JAHS field crews sampled three categories of sampling sites.Most
of the sampling occurred at Resident Juvenile Habitat (RJHAB)model
sites where the response of the site to changes in mainstem discharge
was evaluated along with juvenile salmon use of the site.Crews also
sampled Instream Flow Incremental Methodology (IFIM)model sites for
fish distribution and abundance at which hydrau1 ic habitat model s were
developed.The third category of sites,at which further data on fish
distribution and habitat were gathered,were known ~s 1I0pportunistic"
sites.Further details on specific sampling techniques and methods used
in the JAHS studies are given in earlier reports (AOF&G 1984a,1984b).
2.1.1 Study locations and selection criteria
The sampling sites modelled were chosen from side channels,tributary
mouths,and side sloughs,which met the following basic criteria:
The effects of mainstem discharge (stage and flow)on the
sites are measurable.
The sites are documented or thought to contain potential
habitat for rearing juvenile salmon.Sites with extremely
high (>3 feet/sec)velocities were assumed to have little
value and were not evaluated.
C.The sites are accessible by boat at normal mainstem discharges
during the open-water season.
The 20 sites modelled with RJHAB and IFIM models were distributed
between the Yentna River confluence and Talkeetna (Figure 1).Fourteen
of the sites were modelled only with the RJHAB model,four with only
IFIM models,and two with both RJHAB and IFIM models.Eight of the
sites are located within slough or side channel complexes which were
picked by R&M Consultants and E.W.Trihey and Associates as representa-
tive of lower Susitna River slough or side channel complexes for extra-
polation purposes.For purposes of extrapolation,the side channel
complex area data need to be integrated with the habitat modelling data
by comparing breaching flows and channel size and type between modelled
sites and individual channels within the representative complexes.
3
RIVER MODEL
SITE MILE RJHA IFIM
Trapper Creek S.C.91.6 X X
Birch SloughC!88.4 X
Sunrise S.C.a 87.0 X
Sunset 5.C.°86.9 X.
Beaver Dam SloughO 86.3 X
Beaver Dam S.C.a 86.3 X
Sucker S.C.a 84.5 X
Sauna S.C.79.8 X
Circular S.C.75.3 X
·Goose 2 S.C.74.8 X
Mainst8m West Bank 74.4 X
Island S.C.63.2 X X
Caswell Creek Mouth °63.0 X
Rustic Wilderness S.C.59.5 X
Lost Chance 5.C.44.4 X
Bear Bait S.C.42.9 X
Rolly Creek Mouth 39.0 X
Krato Slough Head 36.3 X
Eagles Nest S.C.a 36.2 X
Hooligan S.C.a 35.2 X
a LOCATED WITHIN REPRESENTATIVE
SIDE CHANNEL OR SLOUGH
COMPLEXES MAPPED BY ASHTON
a KLINGER -KINGSLEY (1985).
Coolt Inltlt
Figure 1.Location of study sites on the lower Susitna River at
which juvenile salmon habitats were model1ed,June
through October 1984.
4
-
-
....
.-
-
-
-
-
-
,..,..
,....,
,..,..
-
-
,..,..
-
-
Proportionately more sampling effort was expended within smaller side
channels in this study because that is where potential habitat is
greatest.Only a portion of the habitat modelling sites were selected
to occur wi thi n the representative compl exes because further data on
distribution of juvenile salmon at locations throughout the lower river
were desired.
Four of the sites were normally clear-water sloughs or tributary mouths
while the other sites were turbid secondary side channels at normal
summer flows.Secondary side channels selected for sampling ranged
greatly in size,shape,and overtopping discharge.The majority of the
habitat model sites selected were secondary side channels because most
of the potential habitat for juvenile fish in areas of the lower Susitna
River affected by the mainstem is composed of secondary side channels.
Primary s'ide channel and mainstem velocities were so high that they were
not considered viable habitat.
Opportunistic sampling sites were selected by sampling crews as poten-
tial habitat which upon sampling might provide for a better analysis of
fish abundance and distribution.Sites sampled were more diverse than
the RJHAB and IFIM model sites and included areas within alluvial island
complexes.
2.1.2 Field data collection
2.1.2.1 Resident Juvenile Habitat (RJHAB)model sites
Two types of data were collected at RJHAB model sites.Habitat data
were collected for the purpose of modelling the response of the site to
changes in mainstem discharge.Fish distribution data were collected
for use in verifying the habitat model data,documenting abundance and
distribution,and modifying suitability criteria,if necessary.A
discussion of the techniques used in the collection of habitat modelling
data will be followed by a discussion of methodology used in the col-
lection of fish sampling data.
Each o~the RJHAB sites was sampled within a grid consisting of a series
of transects with associated sampling cells which intersect the channel
of the study site at right angles (Figure 2).Grids were located so
that water quality within them was uniform and so that they encompassed
a variety of habitat types.Survey stakes and orange flagging were used
to mark each transect within a grid.Initial measurements within each
grid included distances and angles between transect bench marks.
Transects were spaced from 50 to 300 feet apart in order to encompass a
variety of habitat types within each grid.Aerial photos of all the
RJHAB sites showing placement of all transects within each site are
presented in Quane et al.(1985).
Up to four 6-by-50 foot rectangular sampl ing cells extending upstream
from every transect within each grid were characterized by habitat
measurements (Figure 2).If the top width of the wetted channel was
greater than 42 feet,two of the four cells paralleled both edges of the
channel and the third and fourth cells were located parallel to the
shoreline cells so as to split the channel into thirds.If the channel
5
~
J.-,.
{;J
~
iTRANSEC\
TRANSECT 2
TRANSECT
I Cell Unit
Area Sampled
-
-
-
Figure 2.Arrangement of transects and sampling cells within
a grid at a hypothetical RJHAB modelling site.
6
.....
.....
.-
--
,-
-
measured 30 to 41 feet in width ·at the transect,there was a cellon
each shoreline of the channel and one cell located approximately mid
channel.If the wetted edge was 18 to 29 feet in width ,there was one
cellon each side of the channel parallel with the bank.If the channel
was 1ess than 18 feet in width,there was only one cell .
Transects were numbered consecutively beginning with the transect
furthest downstream 'withi n the site.Cells were a 1so numbered consecu-
tively from right to left looking upriver.If there were less than four
cells within a transect,cells were numbered as if the missing cells
were present.
One or more staff gages were installed by Aquatic Habitat and Instream
Flow Project (AH)personnel at each site to document changes in the
stage at each site with changes in mainstem discharge.These gages
provided an index to the changes in habitat and hydraulic conditions at
the site between sampling occasions.AH staff also developed mainstem
stage and site flow relationships and mapped the thalweg at selected
sites.
Habitat modelling data were collected over a broad range of mainstem
discharges.Emphasis was placed on data collection at rnainstem dis-
charges of 30,000 to 60,000 cfs as measured at the Sunshine USGS gaging
station.When staff gage readings and observations indicated that the
stage at the site had changed little from a previous sampling occasion,
no habitat data were taken •
Habitat data taken at each grid on a modelling occasion included the
following.At each transect,the distance between the left and right
edge of water and the left bank transect marker was measured.If the
water quality within the grid or grids was uniform,one measurement of
water pH,temperature,conductivity,and dissolved oxygen was taken.A
turbidity sample was collected in a 250 ml plastic bottle and stored in
a cool dark location for up to two days prior to analysis.Turbidity
was measured in nephelometric turbidity units (NTU)with an HF Instru-
ments Model No.DRT-15B field turbidometer.If the water quality within
the grid appeared to vary because of mixed water sources,additional
water quality and turbidity measurements were taken as necessary to
describe these within grid variations.
In addition to the above measurements,each sampling cell within the
grid was characterized by several habitat measurements.A representa-
tive depth and velocity were measured by taking one or more point
measurements along the midline of each cell.The entire cell was walked
so measurements taken were representative . A vel oci ty measurement was
taken at 0.6 of the distance from the top of'the water column at one
representative location for the entire cell.
Additionally,cover type and amount were estimated in each cell and
coded into categories (Table 1).Aquatic vegetation was defined as
aquatic plants which are normally completely submerged and do not stand
upright.Emergent vegetation consisted of plants such as Equisetem sp.
which normally are only partially submerged and stand upright.Over-
hanging riparian vegetation consisted of vegetation whose roots are
7
submerged only at flood stage and which typically grow in moist or dry
soil.Initially,the total amount of cover of all types was estimated
for the enti re cell.Next,the primary and secondary cover type was
recorded along wi th a percentage of the total for each.Cover was
defined as hiding or escape locations for fish less than or equal to 100
mm in total length.
Table 1.Percent cover and cover type categories.
....
Group #
1
2
3
4
5
6
%Cover
0-5%
6-25%
26-50%
51-75%
76-96%
96-100%
Group #
1
2
3
4
5
6
7
8
9
Cover Type
No object cover
Emergent vegetation
Aquatic vegetation
Debris or deadfall
Overhanging riparian vegetation
Undercut banks
Gravel (l"to 3"diameter)
Rubble (3"to 5"diameter)
Cobble (larger than 511 diameter)
....
In September,when the water levels in the Susitna River were low,the
cover on all the transects within each site was systematically recorded.
One person did the systematic cover coding for all the sites so that
between site observer bias was minimized.The cover was recorded by
distance from the left bank transect marker along the transect line.
Fish distribution data were normally collected from a minimum of seven
cells with·in each RJHAB site during each sampling occasion.Cells to be
sampled were selected randomly by using a random numbers table (ADF&G
1985).If a cell was missing or could not be sampled due to high·
velocities or large depths,an additinna1 cell was randomly chosen for
samp1 ing.Consequently,the samp1 ing was not totally random.Each cell
selected was then sampled for fish with one pass through the entire cell
with a backpack e1ectroshocker or beach seine.The gear type used was
considered the most efficient for sampling the cell.Typically,beach
seines are more efficient in turbid water while e1ectrofishing gear is
most efficient in clear water (Dugan et a1.1984).The area of the ce1~
was recorded so that catches in cells with areas different than 300 ft
could be adjusted to this standard cell size.Sampling efficiency of
e1ectrofishing and beach seining was assumed to be equal.
Additional selected cells were occasionally fished at the site if
samp1 ing of the random cell s failed to capture many fi sh because the
cells had high water velocities.In this case,the sampling crew
fished areas which had more suitable water velocities.Areas fished
were not limited to cells on the transects.These data were pooled with
the randomly selected cell data for analysis.
8
~,
-
After each cell was sampled,juvenile salmon captured were identified to
speci es and then released.The total 1ength of each of the fi rst 50
fish of each species in each size class was measured in millimeters.
If staff gage readings indicated the stage at the site had not changed
from a previous sampling period only limited habitat measurements were
taken.These included water chemistry data and a turbidity sample.
Fish distribution data were taken during each visit to the site,how-
ever.Each cell sampled for fish was al so characterized by a represen-
tative velocity,depth,and estimate of cover type and abundance.
2.1.2.2 Instream Flow Incremental Methodology (IFIM)sites
In addition to the RJHAB model sites,there were also six sites modelled
for juvenile fish using the Hinstream flow incremental methodologyn
(IFIM)(Bovee 1982).A Summary of this methodology and specific data
collection and modelling techniques are presented in Appendix D of this
report.All habitat data used in the IFIM models were collected and
analyzed by Aquatic Habitat (AH)personnel.Two of the IFIM sites were
also modell edwith RJHAB models using the same transects in order to
compa re output from the two mode 11 i ng methods.At these two sites,RJ
personnel collected the RJHAB and fish distribution data and AH person-
nel collected the IFIM data,so the two models were independent.
Fish abundance and distribution data were also collected at the other
four IFIM model sites.Sampling effort at these sites was secondary in
importance to the sampling of the RJHAB sites.Cells were sampled for
fish using the transects placed for the IFIM models.Cells were ran-
domly sel ected and then sampl ed wi th the same procedures used at RJHAB
sites.Cell numbering was the same as that used in the RJHAB studies.
The distance from the transect end markers to the cell edge was mea-
sured,however,so that the location of the cellon the transect was
spec ifi ed.Other data collected at each ce 11 fi shed included amount and
type of cover,water depth,and water velocity.Water chemistry mea-
surements and a turbidity sample were also taken at a selected location
within the site..
2.1.2.3 Opportunistic sites
In addition to the RJHAB and IFIM sites,other sites were sampled for
fish as time permitted to gather juvenile abundance and distribution
information at a wider variety of sites and to obtain further data for
juvenile suitability criteria.Selected 6-by-50 foot cells were sampled
for juvenile salmon at opportunistic sites but no permanent grids or
transects were marked.Water chemistry was measured at mid-site.If
time permitted,each cell sampled for fish was characterized to amount
and type of cover,water depth,and water velocity as were cells sampled
at RJHAB and IFIM sites.
Early in the sampling season,large differences in turbidity.were noted
between sites located on the east and west banks of the Susi tna River
mainstem below the Chulitna River confluence.In order to better
understand the reason for these differences,turbi diti es were taken
9
within the Talkeetna and Chulitna rivers just above their respective
confluences with the Susitna and also in the middle Susitna River above
its confluence with the Chulitna River.The turbidity measurements were
then repeated in the lower Susitna River below the Chulitna River on the
left (west)bank channel,center channel,and right (east)bank channel
at several locations from RM 92.7 downstream to RM 60.6.Blueline maps
detailing the precise sampling locations are available at the Susitna
Aquatic Studies office.Two sets of measurements were taken,on July 19
and on August 16.The measurements were recorded within a four hour
period on each date.Turbidity samples were taken at least 30 feet off
shore near the middle of the channel.
2.1.3 Schedule of activities and frequency of sampling
Field sampling trips,lasting approximately 7-10 days,were conducted
bimonthly from June through mid-October.Each RJHAB site was sampled
for fish on each sampling occasion if fish habitat was present.Habitat
data were collected on at least three occasions when staff gage readings
or observations suggested a change in the habitat within a site.The
collection of habitat data was therefore dependent upon mainstem dis-
charge.
The IFIM sites were sampled at least once a month during the open-water
season.Opportunistic sites were sampled as time permitted and some
were only sampled once.Opportunistic sites were sampled mainly in
September and early October when many of the RJHAB and IFIM sites were
dewatered.
2.2 Data Analysis
All fi el d data were recorded on the appropri ate data forms and trans-
mitted to the office where the fish distribution data and much of the
habitat data were entered into a mainframe computer data base.Data
sorts,summary retrievals,and selected computer files were extracted
from this data base as needed.Other habitat data were entered directly
into basic programs or commercial software on a personal computer.
2.2.1 Physical data
Overtopping flows at the study sites were observed or estimated from
staff gage measurements and flow observations.Data were grouped into
nine half-month sampling periods from early June (June 1 -June 15)to
early October (October 1 -October 15).Due to logistical constraints,
the actual sampling periods did not always run from the 1st to the 15th
and 16th through the end of the month.
An index to the amount and type of cover within the RJHAB and IFIM model
sites was calculated by totalling the linear feet of all the cover types
along the transects at a mainstem discharge within the range of 49,000
to 57,000 cfs.In addition,at Rolly Creek mouth,Caswell Creek mouth,
and Beaver Dam Slough,the response of phys i ca 1 cover to changes in
mainstem discharge was plotted by totalling the cover along the tran-
sects at all measured discharges.
10
"""
...."
-
~
,
,....
"'""
,.....
-
-
The response of RJHAB site wetted areas to ma i nstem di scharge was
plotted using a BASIC language geometry program to calculate wetted area
at each transect within a site on each modelling occasion.After
fitting these points by hand using professional judgement,site areas at
3000 cfs increments were measured on the graphs with a digitizer.The
IFG HABTAT program calculated wetted .areas at the six IFIM sites as a
function of side channel flow,and these were also plotted us"ing a
mainstem discharge-side channel flow ~e1ationship.
2.2.2 Abundance and distribution
The same classification of macrohabitats was used to examine differences
in fish distribution among the sites as that discussed in Dugan et a1.
(1984).The sites were classified as tributary mouths,side sloughs,
and side channels.Tributary mouths are sites which are influenced by
tributary flows and backwater effects from the mainstem.Side channels
are channels whose upstream berms (heads)are breached by the mainstem
while side sloughs are channels whose heads are not breached and whose
water sources are upwelling,local runoff,or small tributaries.Side
sloughs transform to side channels when their heads are breached by the
mainstem.Birch Creek Slough was classified as a tributary mouth in
1984 because road building activities in the upper part of the slough
closed the head off from the mainstem.Beaver Dam Slough was also
classified asa tributary mouth because it only overtops at discharges
greater than 80,000 cfs and normally runs clear.Beaver Dam Slough is
much more similar to Rolly Creek mouth than to any of the other side
sloughs in the lower reach.
Catches within cells with areas other than the standard 300 ft 2 were
adjusted to correspond to this standard cell area.The analysis was
then based on the adjusted mean catch per cell.
2.2.3 Habitat modelling of rearing salmon
2.2.3.1 Suitability criteria development
Suitability criteria have been developed to model the response of
juvenile salmon habitat to variations in mainstem discharge at sites
located in the middle reach of the Susitna River (Suchanek et a1.1984).
As habitat data collection techniques used in the lower river in 1984
were similar to those used during 1983,the middle river suitability
criteria were compared to the lower river distribution data and mod-
ified,if necessary,in Appendix A.The suitability criteria developed
in Appendix A are used in all subsequent habitat modelling for the lower
river.
2.2.3.2 Instream Flow Incremental Methodology (IFIM)models
The IFIM PHABSIM system of computer programs was developed by the U.S.
Fish and Wildlife Service as a means of describing the mosaic of phys-
ical features of a stream which includes hydraulic variables such as
depth and velocity and other features such as substrate or cover (Bovee
1982).A hydraulic model is first calibrated which describes the
response of hydraulic variables such as depth and velocity to stream
11
-----------------------
flow (Milhous et al.1981).The HABTAT program is then used to incorpo-
rate output from the hydraulic model and substrate data with the suita-
bil ity cri teri a to produce estimates of the habi tat potenti a1 (weighted
usable area)for a given life stage of a species.Weighted usable area
(WUA)is calculated as follows (Bovee 1982):
-
WUA =Ci,s X Ai
where:C., ,s
A.,
=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)
=the surface area of the cell
Each cell is a small section of the study channel which is bounded by
other cells or the shoreline and extends midway between transects.The
WUA for the study site at a given discharge was calculated by,totalling
a11 the i ndi vi dua 1 cell WUA IS.The compos ite wei ghti ng factor was
calculated by multiplying the suitability indices for cover,velocity,
and depth of the cell together.WUA'S at each study site were calculat-
ed at flows whi ch corresponded to 3,000 cfs increments of rna instem
discharge as measured at Sunshine gaging station.
Much more detailed descriptions of the IFIM data analysis methods and
hydraulic simulation results are presented in Appendix D.Only selected
WUA results as a function of mainstem discharge are presented here.All
species and site combinations were run and are available on request but
space limitations prevent presentation here.Site/species combinations
presented were selected on the basis of fish catches at the site.
2.2.3.3 Resident Juvenile Habitat (RJHAB)models
The original RJHAB model was designed to calculate weighted usable areas
for the habitat within a site Without using hydraulic models (Marshall
et al.1984).The model divided the site into shorelfne and mid-channel
sections,and calculated weighting factors for cover and velocity for
each section which were then multiplied together with area to produce a
weighted usable area estimate at each of the discharges measured.
The original RJHAe model was greatly modified for the 1984 analyses.
These changes were made so that the RJHAB model cal cul ates wei ghted
usable areas similarly to the HABTAT program described by Milhous et al.
(1981)that is used in IFIM analysis.Also the cover coding has been
standardized so that observer variations in rating cover at different
discharges do not lead to variations in cover estimates unrelated to
changes in wetted area.
The current RJHAB model is a spreadsheet developed on commercial soft-
ware.Though no hydraulic model is developed,the current RJHAB model
12
"""I
~,
closely resembles the HABTAT model in its procedures for calculating
weighted usable areas within a site.Instead of calculating weighting
factors for cover and velocity in shoreline and mid-channel sections on
a given sampling occasion as did the original RJHAB model,each site is
partitioned into "s tream cellsll each with a unique area,cover type,
cover percentage,velocity,and depth.The site weighted usable area
(WUA)is ttien the sum of the "s tream cell"WUA's which are calculated by
multiplying the area,cover,velocity,and depth suitabilities together.
The velocity and depth measurements of the 6'x 50'sampling cells are
assumed to represent a much larger stream cell.The wetted surface area
between transects was partiti oned into one to four stream cell s depen-
dent upon wetted transect width (Table 2).
Table 2.Partitioning of wetted channel width into stream cells.
-
Wetted Channel Width
>42 ft
30-41 ft
18-29 ft
<18 ft
No.of
Stream Cell s
4
3
2
1
How Area Partitioned
Cellon each shoreline 6 ft in
width,two center cells split
the difference.
Cellon each shoreline 6 ft in
width,middle cell is the rest.
Each cell with half the width.
Entire width.
--
Occasionally,islands prevented a simple partitioning of the site but in
each case,areas were partitioned so that sampling cells best repre-
sented a given stream cell.Once the wetted width of stream cells was
partitioned,a computer program written in BASIC was used to calculate
the surface area of each stream cellon each sampling occasion.The
areas of islands were estimated from width measurements,observations,
and sketch maps and then subtracted from the area of each stream cell.
Cover suitabilities for each stream cell were calculated with a BASIC
program which integrated the standard cover data taken on each transect
with the partitioned wetted width of each stream cell.The cover
su i tabi 1i ty of each cover type on the stream cell wetted wi dth was
averaged with the other cover suitabilities present (proportional to
their occurrence)to give an average cover suitability.For example,if
the stream cell was 15 feet in width and ten feet of the width was a
cover type with a suitability of 0.5 and the other five feet was a cover
type with a suitability of 1.0,the average cover suitability for the
cell would be :[(10 x 0.5)+(5 x 1.0)]/15 =0.67.
The RJHAB spreadsheet then took the stream cell areas and cover suit-
abilities,and multiplied these with the depth and velocity suitabil-
ities which .it assigned to the sampling cell depth and velocity measure-
ments.The products of these calculations (stream cell WUA's)are then
totalled to calculate site WUA's for each sampling occasion.Weighted
13
usable areas for chinook salmon in turbid and clear water and chum,
coho,and sockeye salmon were all calculated concurrently.
Weighted usable areas were plotted over the range of mainstem discharges
sampled.Since initial overtopping flows were estimated for each side
channel,WUA response was extrapolated in the range around breaching
using this information.Habitat indices were calculated by dividing the
WUA of the 'site at a given discharge by the site area at the same
discharge and these were also plotted.Only selected site and species
combinations are presented here,all other WUA calculations are avail-
able upon request.Individual sampling cell measurements are also
available upon request.
In order to compare output from the RJHAB model with that of the IFIM
methodology,two sites (Island and Trapper Creek side channels)were
modelled with both techniques.Output from both techniques were graphed
as a function of mainstem discharge and then correlated with each other
at the measured RJHAB discharges.
2.2.3.4 Model verification
Fish abundance data were collected at all of the IFIM and RJHAB sites.
High mean catches per cell (CPUE's)should reflect high densities of
fish within the site.Since WUA on a per site basis reflects the size
of a site,WUAisite is not an index to habitat quality of a site.The
habitat index calculated by dividing WUA by site area (at any given
discharge),however,does reflect site habitat quality,independently of
site area.
Variations in mainstem discharge cause fluctuations in the habitat value
of a given site.Fish populations within a site may not respond immedi-
ately to such variations in habitat value but should adjust after a
period of time.Over a season,average densities of fish (as expressed
by CPUE)should be positively correlated to the average seasonal habitat
index if there is a relationship between the two.A test of the signi-
ficance of the correlation between mean seasonal habitat indices and
mean catch per cell by species was used to verify the habitat modelling
efforts.
Mean seasonal habitat indices for each site were calculated for each
species with the following procedure.Mean daily discharges for each
day between May 15 and October 15 were rounded to the nearest 3,000 cfs
increment in the range from 12,000 to 75,000 cfs.The season for chum
salmon ran from May 15 to July 15.If the discharge was greater than
75,000 cfs,the discharge was assumed to be 75,000 cfs because WUA's
were calculated only up to 75,000 cfs.Corresponding WUA's and site
areas corresponding to these discharges were then totalled to find the
total WUA and site area for the season.The mean seasonal habitat index
was then calculated by dividing the total WUA by the total site area.
For chinook and chum salmon,WUA's were adjusted by a turbidity factor
before the habitat index was calculated.The turbidity factor was
calculated by fitting a suitability index from a to 1.0 on the dis-
tribution of mean chum and chinook juvenile salmon catch by 50 NTU
turbidity increments.Site mean CPUE's were regressed against site
habitat indices at each site.
14
~i
-
-
..-
3.0 RESULTS
3.1 Seasonal,Spatial,and Discharge Related Variations in Habitat
3.1.1 Macrohabitat type cl assifi cations of study sites
All the study sites were classified into one of three macrohabitat
types:tributary mouths,side channels,or side sloughs.Classifica-
tion and habitat characteristics of the twenty modelled study sites are
given in Table 3.Initial breaching discharges for the side channels
ranged from approximately 14,000 to 46,000 cfs with flows controlled by
the mainstem at least 50%of the time.Channels with input into the
tributary mouth sites were never breached at flows less than 54,100 cfs
and site flows were controlled by the mainstem less than 5%of the time.
Backwater effects were the only effects attributable to mainstem dis-
charge at the tributary mouths on all sampling occasions except at
Beaver Dam Slo~gh where discharges greater than 75,000 cfs caused the
head to overtop and flow to increase through the site.Even at dis-
charges greater than 75,000 cfs however,the major effect of rna i nstem
discharge on Beaver Dam Slough was a backwater response.
The side slough macrohabitat type was not represented by any of the
sites when mainstem discharges were highest during the period from late,
June through early August.Side slough habitat increased with decreases
in mainstem discharges.
Major object cover differences among the model 1 ing sites were differ-
entiated by macrohabitat type.An index of cover for each site at a
discharge of approximately 52,000 cfs (range 45,500 to 58,800 cfs)was
calculated for between-site comparisons of cover (Table 4).The per-
centage of the site with the primary cover type,submerged aquatic
vegetation,varied from 8.5%to 68.5%for the tributary mouths,while
none of the si de channel/sloughs had any submerged aquatic vegetation.
Substrate in the form of large gravel (1-3 11 diameter)and rubble (3-5 11
diameter)was the primary cover type and averaged 62%of the side
channel area whil e these two cover types only covered an average of 14%
of the area of tributary mouth sites.The density of cover at tributary
mouths was almost three times that of side channels also.Side sloughs,
which by definition are unbreached side channels,typically have less
object Cover than side channels.
Cover,in the form of turbidity was much more frequent within side
channels than at tributary mouths and side sloughs.Turbidities were
consistently higher in the side channels than in the tributary mouths
during the open-water season (Figure 3).A few turbidities of 100 to
150 NTU were recorded at Rolly Creek mouth and Beaver Dam Slough due to
rapid increases in mainstem stage which caused turbid water to intrude
into the sites,or in the case of Beaver Dam Slough,by a slight over-
topping of the channel head by mainstem water.Turbidities within the
side sloughs ranged from 1 to 19 NTU with a mean of 5.2 NTU.
15
Table 3.Classifications and habitat characteristics of study sites on the lower Susitna River at which juvenile salmon habitat was
modelled,June through October 1984.
......
01
Site
Side Channels (head open)/
Sloughs (head closed)
Hooligan Side Channel
Eagles Nest Side Channel
Kroto Slough Head
Bear Bait Side Channel
Last Chance Side Channel
Rustic Wilderness Side Channel
Island Side Channel
Mainstem West Bank
Goose 2 Side Channel
Circular Side Channel
Sauna Side Channel
Sucker Side Channel
Beaver Dam Side Channel
Sunset Side Channel
Sunrise Side Channel
Trapper Creek Side Channel
Tributary Mouths
Rolly Creek Mouth
Caswell Creek Mouth
Beaver Dam Slough 2
Birch Creek Slough
River Mile
35.2
36.2
36.3
42.9
44.4
59.5
63.2
74.4
74.8
75.3
79.8
84.5
86.3
86.9
87.0
91.6
39.0
63.0
86.3
88.4
Initial
Breaching
Discharge (cfs)
23,100
14,000
36,000
35,000
(Est.)
22,700
19,000
34,000
19,000
30,000
36,000
37,000
27,500
46,000
31,000
34,300
43,000
75,000+
54,100
Percent of
Time Flow
Controlled by 1
Mainstem in 1984
80
94
62
64
(Est.)
79
86
64
86
68
64
62
71
50
68
64
57
o
o
<5
<5
Non-mainstem
Water Sources
Pools only
Unknown
Minor upwelling
Pools only
Pools only
Pools only
Major upwell fng
Major upwell i ng
Minor upwell ing
Major upwelling
Minor upwelling
Minor upwelling
Unnamed tributary
Major upwelling
None
Cache Creek
Rolly Creek
Caswell Creek
Unnamed tributary
Birch Creek
These percentages based on controlling breaching discharges presented in Quane et al.(1985)for the period from May 15 to October
15,1984.
2 A culvert at the head of this slough is frequently blocked and therefore little mainstem water flows into the slough,even if the
slough head is breached.The effect of mainstem discharge on this site is minimal for this reason •
.,J J t .~I ])!J J J J J J J J ,
1 i }1 I 1 1 J J j
Table 4.Percentages of lower river habitat modelling sites associated with nine cover-type categories.Percentages are based on the width of transect with each cover type.
Cover Index calculated by dividing total cover by total area of site.
Percentage of Site With Primary Cover Type
River
OVerhang.Ccver 1
Discharge No Emergent Aquatic large RI parlan U.C.Density
Side Channel s/Sloughs Mile Date (ch)Cover Veg.Veg.Crave I Rubble Cobble Debris Veg.Banks Total (t)
Hooligan Side Channel 35.2 7/14 52400 18.9 0.0 0.0 72.0 0.0 0.0 8.5 0.6 0.0 100.0 13.7
Kroto Slough Head 36.3 7/17 49600 56.4 0.0 0.0 8.6 0.0 0.0 33.5 1.6 0.0 100.1 1.8
Bear Bait Side Channel 42.9 7/13 52400 0.0 0.0 0.0 66.8 0.0 0.0 28.1 3.7 1.4 100.0 11.5
last Chance Si de Channe 1 44.4 7/12 54100 23.5 0.0 0.0 63.5 0.0 0.0 12.3 0.8 0.0 100.1 5.9
Rustic Wilderness Side Channel 59.5 8/12 52900 0.0 0.0 0.0 60.9 30.0 0.0 7.8 0.8 0.5 100.0 13.7
Island Side Channel 63.2 7/19 51600 13.4 0.0 0.0 62.0 21.6 0.0 0.0 1.4 1.6 100.0 10.5
Mainstem West Bank 74.4 Extrapolated 54100 1.0 0.4 0.0 43.4 49.3 0.0 2.2 3.4 0.4 100.1 22.7
Coose 2 Side Channel 74.8 7/20 52600 2.0 0.9 0.0 24.3 51.8 13.7 3.5 3.5 0.2 99.9 22.5
Circular Side Channel 75.3 7/24 56600 20.4 0.0 0.0 48.4 21.3 0.0 5.3 4.6 0.1 100.1 9.3
Sauna Side Channel 79.8 7/23 56600 93.4 0.0 0.0 0.0 0.0 0.0 4.3 2.4 0.0 100.1 0.5
Sucker Side Channel 84.5 7/09 55400 80.2 8.4 0.0 6.6 0.0 0.0 3.9 1.0 0.0 100.1 1.1
Beaver Dam Side Channel 86.3 7/08 57100 55.9 0.9 0.0 18.6 5.9 0.0 18.6 0.0 0.0 99.9 1.9
Sunset Side Channel 86.9 7/22 57800 15.0 0.0 0.0 66.8 9.7 0.0 7.7 0.5 0.3 100.0 4.8
Sunrise Side Channel 87.0 7/07 58800 4.0 0.0 0.0 51.4 44.6 0.0 0.0 0.0 0.0 100.0 10.0
Trapper Creek 51 de Channel 91.6 8/19 57200 2.2 0.0 0.0 39.1 58.8 0.0 0.0'0.0 0.0 100.1 12.3
MEAN 25.8 0.7 0.0 42.2 19.5 0.9 9.0 1.6 0.3 100.0 9;5
Tributary Mouths
Rolly Creek Mouth 39.0 7111 55100 6.9 25.2 46.2 0.0 0.0 0.0 21.5 0.1 0.0 99.9 24;2
Caswe 11 Creek Mouth 63.0 8/18 45400 2.9 5.3 4B.2 17.6 0.0 0.0 18.4 1.6 6.1 100.1 19;0
Beaver Dam Slough 86.3 7/08 57100 6.8 9.9 68.5 0.0 0.0 0.0 11.1 3.1 0.6 100.0 57;8
Birch Creek Slough 88.4 7/20 52600 36.8 0.5 8.5 29.2 9.0 0.0 13.6 2.2 0.3 100.1 6.3
MEAN 13.4 10.2 42.9 11.7 2.3 0.0 16.2 1.8 1.8 100.0 26.8
......1 Ccver density Is the average density of object cOver within the site on a percentage basis •'i
SIDE CHANNEL TURBIDITIES
(MODELLED SITES ONLY)
1000----------=----------------------,
900
800
•MEANIRANGE
::;)700
I-
Z
600
500
>-
l-
e
CD
Q:400::>
I-
300
200
--•O..L----:----r---~-__r--_r_-~--__.__-___,r___-+----J
100
E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAMPLING PERIOD
•MEAN
IRANGE
TRIBUTARY MOUTH TURBIDITIES
(MODELLED SITES ONLY)
150 --------------------------_
140 -
130
120 -
110 -
~100-
z 90-
>-80-
l-
e 70-
~60-
::>
I-50-
40 -
30 -
20 -
10 -l ~~I t
o ....I....------rT---,Tr------.----r---...--1 --i-T ---i-----T1---"r___---l
E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAMPLING PERIOD
Figure 3.Turbidities of modelled side channels and tributary mouths on
the lower Susitna River,June through October 1984.
18
.-
3.1.2 Chulitna and Talkeetna River plume influences on turbidity
of side channels
Turbidity measurements of the lower Susitna River taken in west bank,
mid-channel,and east bank portions of the mainstem indicate that plume
influences of the Chulitna and Talkeetna Rivers extend at least 20 to 30
miles downriver (Figure 4).On September 2,turbidities at RM 83.8
ranged from 60 NTU on the east bank,to 77 NTU in 'mid-channel,and 88
NTU on the west bank.West bank turbidities are much higher than on the
east bank,because the Chulitna River is three or more times as turbid
as the Talkeetna River and middle reach of the Susitna River.
A comparison of turbidities at the modelled side channels located above
RM 70 also suggests that the plumes have major effects on turbidities
downstream.Mean turbidity at lateral side channels located on the west
bank (Mainstem West Bank,Sauna S.C.,and Trapper Creek S.C.)during
June through late August was 377 NTU.During the same time period,
lateral side channels located on the east bank (Goose 2 S.C.,Sunset
S.C.,and Beaver Dam,S.C.)had a much lower mean turbidity of 158 NTU.
Mean turbidities for all the side channels modelled with the exception
of Eagle's Nest Side Channel have been calculated in Appendix Table 8-1.
Many more turbi diti es woul d have to be taken to better del i neate the
Chulitna River and Talkeetna River plumes.The large east bank clear
water tributaries such as Montana Creek and Goose Creek make the differ-
ences in turbidity between the east and west banks of the lower river
even larger,and confound analysis of the extent of plumes from the
Chulitna and Talkeetna rivers.
3.1.3 Physical responses of sampling sites to mainstem discharge
variations
Variations in mainstem discharge cause the heads of side channels to
alternately be overtopped or dewatered,thereby altering macrohabitat
classifications due to changes in water quality,flows,wetted areas,
and the amount of cover.The relationships between side channel flows
and mainstem discharge at the sampling sites are presented in Quane et
a 1.(1985).
Changes in wetted area of sites due to variations in mainstem discharge
are important because these changes may directly increase or decrease
fish habitat.Areas measured from aerial photos have been compiled for
selected side channel and slough complexes by Ashton and
Klinger-Kingsley (1985)for a variety of discharges.Mainstem backwater
effects at tributary mouths are al so important because object cover
inundated by backwater is an'important component of these sites for
juvenile salmon.Discharge related responses of site area for all sites
pooled and cover for selected tributary mouths will be presented in the
next two sections.
3.1.3.1 Area
The areas of the RJHAB study sites were calculated geometrically at
modelled discharges,and then plotted against mainstem discharge by eye.
Measurements of area were then read from these graphs in the range
19
CHULlTNA,TALKEETNA PLUME EFFECTS
1.1
JULY 19.1984 LOWER 8U8ITNA
Cl WEST BANK
+MID-CHANNEL
o EAST BANK
~,
0.9
~0.8
.....-g 07
~i .
Q~0.6
m~~.0.5
0.4
0.3
0.2
0.1
a
CHULITNA
MIDDLE
SUSITNA
TALKEETNA
-
5 15 2S 3S
-CHULITNA
o AUGUST 18.1984
1.2
1.1
1.3 -r---,...------------:~=--------___,
LOWER 8U8lTNA
D WEST BANK
+MIO-CHANNEL
o EAST BANK
0.9
~0.8
'-oI~7~.i o.
Q~0.6
rtJ~~0.5
0.4-
0.3
0.2
TAU<EETNA
MIDDLE SUSITNA
0.1
O+---t----r--.,...--~---r--_r--,_-___,r_-___j
5 15 2S 35
MILES DOWNSTREAM FROt.f CONFLUENCE
Figure 4.Comparison of turbidities in the lower Susitna River below
the Chulitna and Talkeetna River confluences on July 19 and
Auoust 16,1984.
20
-
-
-
-
between 12,000 to 75,000 cfs at 3,000 cfs increments.Since Eagles Nest
Side Channel was modelled only at discharges less than 20,000 cfs,we
did not try to extrapolate values over this range for this site.
Similarly,area·response at the six IFIM sites were calculated by the
IFG program at side channel flows which corresponded to increments of
3000 cfs within the 12,000 to 75,000 cfs mainstem discharge range.
Individual area responses for all the modelling sites have been tabu-
lated in Appendix Table B-4 at 3,000 cfs discharge increments.Also,
side channel flows associated with these increments have been tabulated.
By summing areas of the sites by macrohabitat type,the response of the
pooled sites can be illustrated.The combined area of three tributary
mouths increased greatly at di scha rges greater than 27,000 cfs (Fi gure
5).Since sloughs transform to side channels at greater discharges,
slough habitat decreased with discharge while side channel habitat
steadily increased (Figure 6).Slough habitat was broken into two
categories:total and access ib1e.The total category i ncl udes ponded
water with no access from the mainstem while the accessible sloughs are
those with potential access from the mainstem.
3.1.3.2 Cover
Since instream cover is an important component of fish habitat,the
response of available cover tomainstem discharge at individual sites is
of interest.Increases in instream cover (debris,riparian vegetation)
at side channels were often accompanied by large increases in flows and
related water column velocities.Therefore,increases in suitable cover
at side channels were often offset by increases in velocities which made
the site unsuitable.Turbid water in side channels also provides cover
for juvenile chinook salmon and therefore,instream object cover may be
less necessary for chinook salmon under turbid conditions (Suchanek et
a 1.1984).
At tributary mouths,on the other hand,tributary flows ·are independent
of mainstem discharge,the water is often clear,and the primary effect
of mainstem discharge is the formation of a backwater zone.Increases
in mainstem stage typically decrease velocities and "inundate cover at
tributary mouths.
Cover responses to mainstem discharge at the four tributary mouths
varied.At Birch Creek Slough,there were no changes in cover as a
result of changes in mainstem stage during 1984 sampl ing because the
sampling site was located high enough (0.7 miles)up the channel that it
was not influenced by mainstem stage.At Beaver Dam Slough,increases
in total cover caused by rises in mainstem discharge were limited
because most of the cover was submerged aquatic vegetation (Figure 7).
At Rolly Creek and Caswell Creek mouths,however,the amount of cover
increased rapidly at discharges larger than 45,000 cfs.Increases in
total cover at Rolly Creek mouth were caused primarily by inundation of
emergent vegetation while both emergent vegetation and overhanging
riparian vegetation cover became more abundant at Caswell Creek mouth at
high mainstem discharges.
21
703050
(Thousands)
MAINSTEM DISCHARGE AT SUNSHINE (efs)
TRIBUTARY MOUTHS
(BIRCH SLOUGH EXCLUDED)
300 I _.'"'I
290
280
270
260
250
240
230
220
210
200
190
180
170
160
150
140
130
120
110 ~G a a B ~IIIIII8
10
?::-....It)
"tJ
'Cero14It)
-::J
~o~
N
N
Figure 5.Area within modelled tributary mouths as a function of mainstem discharqe at the
USr,S Sunshine gaging station!1984.Boundaries of the site were fixed.
1 J ,1 J t I J J },.~,s J t ~
30 50 70
(Thousands)
MAlNSTEM DISCHARGE AT SUNSHINE (cfs)
-
.....
-
-
.....
I
SIDE CHANNELS
(EAGLES NEST S.C.EXCLUDED)2..,....---------------------=::::>&----,
1.9
1.8
1.7
1.6
1.5
1.4
1.3.......
..1.2-.......&t!1 .1
~~1
~0.9
~0.8
0.7
0.6
0.5
0.4-
0.3
0.2
0.1
O-H:I-l:l-lEF-,.---.,..---.....,.---,....----.--"""""1r----l
10
Fiqure 6.Area within modelled sloughs and side channels as a function
of mainstem discharge at the US~S Sunshine qaqinq station~
1984.
23
-
~
BEAVER DAM SLOUGH
COvER RESPONSE ~,
0 .•~
0.8
0.7 ..<n:JT.....
~0.6 .AQUAT'1,t£C
"-III
'"0.5 D [wE,R VEe.0:
0:~0.4
<J
0.3
0.2
0.'
0
20 30 40 50 60 70
r.MINSTEM OISC~~~·~UN5HIHE(cts)
~
ROLLY CREEK MOUTH
COvER RESPONSE
0."•Tm"....
0.8
0.7
!~0.6
...0.50:
0:~0.4
0.3
0.2
,0.1 -40 60 !IO
MAlNSTEw ~E::r~NE (cfa)
CASWELL CREEK MOUTH
CO'w"ER RES-PeNSe:
0.9
0.8
0.7
III
Z 0.6
0"-III 0.5'"0:
0:
0.'~<J
0.3
0.2
0'
0
10
•N:JUAT
a EMDi
•0\<0<_
30 50 70
MAlNSTE....OISC~~£~·~UNSHII'fE Cd_)
Figure 7.Instream cover response at Beaver Dam
Slough t Rolly Creek t and Caswell Creek
mouths as a function of mainstem
discharge at the LJSr,S Sunshine qaging
station t 1984.
24
""'I f
-1
""'"
3.2 Distribution and Abundance of Juvenile Salmon
Chinook,coho,chum,and sockeye salmon juveniles were captured at the
twenty habitat model sites,But only one pink salmon fry was captured.
Pink salmon outmigrate early and our methods are not effective at
capturing them.A summary of the juvenile chinook,coho,chum,and
sockeye salmon catch and catch per cell (CPUE)data by site is given in
Appendix Table B-2.
3.2.1 Chinook salmon
Fourteen hundred fifty-ei ght juveni 1e chi nook salmon were coll ected in
the lower reach of the Susitna River from June through early October.
Approximately 83%of these fish were captured at the 20 habitat model
sites.Age 0+fry accounted for 93%of the chi nook salmon j uveni 1es
captured.The percentage of 0+fry increased from 66%in late June to
99%in early August.All chinook fry captured after early August were
0+fish,indicating that 1+chinooks had outmigrated from the study
reach prior to August 15.
Chinook fry were widely distributed at the modelling sites from early
June through late August (Figure 8).Last Chance Side Channel was the
only site where no chinook juveniles were captured.Chinook juveniles
were captured at 80%or more of the sites sampled in early June and late
August.In September and early October,the proportion of sites where
chinook salmon were captured decreased.
Mean juvenile chi nook CPUE was hi ghest at tri butary mouths,where 1.5
fish per cell (fpc)were captured.At side channels,the mean CPUE for
juvenile chinook was 0.8 fpc.Slough catch rates were consistently low
(0.1 fpc).Mean catch rates at side channels were relatively constant
throughout the season,whi 1e tributary mouth CPUE's peaked in August
(Figure 9).The peak CPUE for tributary mouths occurred in late August
at Caswell Creek mouth (20.2 fpc).The peak CPUE at a modelled side
channel (4.4 fpc)occurred at Sunset Side Channel.CPUE I s within the
side channels peaked at turbidities of 100 to 150 NTU (Figure 10).The
correlation (r)between mean turbidity of the modelled side channels and
mean catch per cell of chinook salmon was -0.63 (p <0.05).
Catches at Trapper Creek Side Channel appeared to reflect the effect of
turbidity upon chinook fry use.This west bank site,located below the
Chulitna River,had a high CPUE in early June (2.7 fpc)when turbidity
was low but then no chinook were captured in late June and early July
when turbidities were above 550 NTU.Chinook fry catches increased
slightly on subsequent trips when turbidities began to decrease.
3.2.2 Coho Salmon
Four hundred forty-two juveni 1e coho salmon were captured wi thi n the
lower Susitna River study areas of which only five were not captured
within the habitat model sites.Three age classes of juvenile coho
salmon were captured.Eighty-six percent of the juvenile coho captures
were age 0+and 14%were age 1+.Only one age 2+juvenile was captured.
25
I
t")
~~~•i.""/CHINOOK SALMON
~6\~-.....~\;10\SAMPLING PERIOD...~~.\.'-.:I JUN JUL AUG SEP OCT
.SOt t.II I ]I I ]I I :II:I1~~raepper Creek S.C.•.0 D l2][2]{;;ii [;ji [2]0
r1l ~Birch Slough f.ZI ~D 0 I2l ~f.ZI 0 0
,~.~Sunrise S.C.0 -Ciii CiiiiiI •Ciii - - -1~~'\Sunset s.c.---.•[;;iiI [;iii IZl -
~Beaver Oam SlouGh 0 ~DOD [;iii D 0 -
~Beaver Dam S.C.f2]•CiiiiiI •[,;iiI [,;iiI [,;iiI 10 -
~~Sucker S.C.Gil ~[ZI [;;iiI 0 IZJ I2l - - :
Sauna S.C.f7l -D 0 I2l [ZI 0 - -
"=t Circular S.C.[;;iiI 0 IZI IZI ~[;iiiI 0 - -~~Goose 2 S.C.[;iii ~[;;iiI lZJ Gi ~0 - -
\
~;;;Mainstem West Bank - - -0 0 IZl [;iii 0 -
~<~r ;I .~I--J Island S.C.0 [;;iiI I:;ii ~~I2l 0 0 -
()'J ,r~Caswell Creek Mouth . • -0 • • •IZI fZI 0
1.'f),~:.tl Rustic Wilderness S.C.[;jjJ [;;iiI •~[;jjJ IZI [;jjJ 0 -
";~~~'~~Last Chance S.C.- -0 10 0 0 - 0 -
Q~.~Bear Bait S.C.-0 f2j [][21 D·0 - -
,~...j.j'RolI'l Creek Mouth IZI !ZI 0 IZI [;jjJ [;;iiI IA Gii [ZJ
"'iY'Ii Kroto SlouGh Head 0 0 121 0 0 IZI 0 - -
.,,"Eagles Nest S.C.- - - - - -0 0 ~
j~~;;;;~:~~u:'6~NC~~EY~0 0 ~~0 0 -0
~0 0.00 [;iii 0.25-2.50
~~~[2]0.QI-O.25 •>2.50
V(]-No sample
o _3"----
Fiaure 8.Seasonal distribution and relative abundance of juvenile chinook
salmon on the lower Susitna River,June through mid-October 1984.
26
-
-
-
,-
".5
4 ,-
:l.::lS:SI TRIBUTARY MOt1l'HS
0 3.50z
J:
0 3I
..J..J 2.5I.&J
0
IS 2Q..
J:g 1.5~
~
::E
0.5
o
E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAMPUNG PERIOD
Figure-9.Juvenile chinook salmon mean catch per cell at
side channels and tributary mouths on the lower
Susitna River by sampling period,June through
mid-October 1984.
75 125 175 225 275 325 375 >400
TURBIDITY (NTU)
1.9
1.8
1.7
1.6
1.5
1.4
::i 1.3w
0 1.2
f5 1.1
Q.1J:
()0.930.8
~0.7
0.6::E 0.5
0.4
0.3
0.2
0.1
0
25
-
Figure 10.Juvenile chinook salmon mean catch per cell at
modelled side channels on the lower Susitna by
turbidity increment,June through mid-October
1984.
27
The percentage of age 1+fry captured decreased from approximately 50%
in early June to 2%in early October.
Juvenile coho salmon were unevenly distributed in the study area,being
captured at only 50%of the 20 modelled sites (Figure 11).Only one coho
was captured at four of these sites.In'most instances,juvenile coho
CPUE's tended to be higher in late summer.
Juvenile coho salmon catches varied greatly among the three macrohabitat
types.Tributary mouths had a mean juvenile coho CPUE of 1.2 fpc while
sloughs and side channels had CPUE's of 0.02 and 0.01 fpc,respectively •.
Juvenile coho were captured at all four tributary mouths,five of the 16
side channels (31%)and two of the 14 sloughs (14%)sampled.Over half
of the juvenile coho were captured at Caswell Creek mouth,with the
majority in mid to late August.The juvenile coho catch rate at tribu-
tary mouths ranged from near ten juveniles per cell at Caswell Creek in
late August to zero fish per cell at several sites during various
sampling periods throughout the open-water season (Figure 12).With the
exception of Birch Creek Slough,coho CPUE's were higher during late
summer and fall than during early summer sampl ing periods.
3.2.3 Chum salmon
Six hundred eight juvenile chum salmon were collected in the lower
Susitna River of which only ten were captured at opportunistic sites.
In early June,chum fry were captured at 13 of 15 (87%)modelling sites
sampled (Figure 13).By late July,chum were only captured at six of 19
(32%)sites sampled.Over 99%of the total catch was made prior to
August and no chum salmon fry were captured after August 15.The
majority of sites with high CPUE's were located in the reach from Island
Side Channel (RM 63.2)to Sucker Side Channel (RM 84.5).
Chum fry CPUE's declined steadily from early June to mid-August (Figure
14),reflecting outmigration of juvenile chum salmon from the Susitna
system.In a pre-study trip in~late May,chum fry were collected at a
number of lower river sites and appeared widely distributed in the
river.
Juvenile chum CPUE's were highest in side channels (0.6 fpc)and tribu-
tary mouths (0.1 fpc).,Slough CPUE's of juvenile chum were low (0.01
fpc),however,sampling effort at sloughs was limited from early June
through early July.Tributary mouth densities were unequally distri-
buted by a single site catch of 39 fry at Birch Creek Slough in late
June.Juvenile chum catches at side channels were affected by turbi-
dity.Peak chum catches were made in side channels with a turbidity of
less than 50 NTU (Figure 15).
3.2.4 Sockeye salmon
Four hundred twelve juvenile sockeye salmon were captured in the lower
Susitna River study reach.Ninety percent (369)of these fish were
captured at the habitat modelling sites.Age 0+sockeye comprised 99%
of the catch.Age 1+sockeye were found in early June at Hooligan Side
Channel,a site which produced no further sockeye juveniles all season,
28
-
-
-
-
-
COHO SALMON
.....
I
C')
::to /~~e
t
;;-~\ol
Q ~
':Q e,r-_....1
-;.".to'SAMPLING PERIOD"',-:/;\"JUN JUL AUG SEP.OCT
~Site I II I 1I I II:1 II:I~K Trapper Creek S.C.0 0 DOD 0 0 0 [ZJ
~Birch Slough •{2]j:;;j 0 [;jjJ 0 [;jjJ [;jjJ r:zJ
,~Sunrise S.C.0 -0 IZl 0 -tj - - -
\Sunset S.C.- - -0 0 ODD -
CC Beaver Dam Slough 0 ~1Zl'j:;;j I2lGiiil:LZI • -
~Beaver Dom S.C.0 C!IZI 121 0'IZJ eilD -
~,\,;;s:.=.uc.:.;.k:..:.er~S.:,,:::C':""---f!0~.!::D::!.~D~~O~O~.~O~D=~-+-~
Sauna s.c.0 -rl DOD I(J - -
Circular S.C.0 0 0 0 0 0 0 - -
Goose 2 S.C.0 0 0 12]0,0 0
IZJDDODODO-
---00000-
--0000-0-
o O'0 0 I;i 1ZI I;i fA [;;iiI
D 0 ODD fZI []ID -
------000
ODoocroo-o
-000000--
0000000--
.[;jjJ -~• • • •[;jjJ •
Island S.C.
Last Chance S.C.
Caswell Creek Mouth
Bear Belit S.C.
Moinstem West Bonk
Rustic Wilderness S.C.
Hooligan S.C.
Rolly Creek Mouth
Kroto Slough Head
Eagles Nest S.C.
MEAN CATCH PER CELL
RELATIVE ABUNDANCE KEY
00.00 Gi 0.25-2.50
[2]O.QI -0.25 •>2.50
-No sample
(ill
-
~,
If J)
-~''----
Cook InltJf
Figure 11.Seasonal distribution and relative abundance of juvenile coho
,-salmon on the lower Susitna River~June through mid-October 1984.
29
CASWELL CREEK MOUTH BEAVER DAM SLOUGH
'0 4
g,I J "'.ill .".5
0 8 0
5 g "0 ..,
I I
:l :l 2.5
6'"Q '"Q00
a:5 W a:2 W
'"Ie ....IQ.....I
J:4 a.li D.
8 ::i 3 1.5 ::i
"<<
!0 !1 ~0
::Ii 2 ~::Ii ~b00.5
Z l77'";l Z
0 0
E JUN L JUN E JUl L JUL E AUG L AUG E SEP L SE"E OCT E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAMPUNG PERIOD SAMPUNG PERIOD
BIRCH CREEK SLOUGH ROLLY CREEK MOUTH
4
W I ".5 "5a
0 0
J:"5 "0
0 0
I I
:l 2.5 ....I 2.5....I
'"'"0 0
a:2 a:2IeIe
J:J:
il 1.5 5 1.5
is
~1 !
::Ii ::Ii
0.5 O.S
0 0
E JUN L JUN E JUL L JUL E AUG I.AUG E SEP L SEP E OCT E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAIoIPUNG PERIOD SAMPUNG PERIOD •
Figure 12.Juvenile coho salmon mean catch per cell at four tributary mouths on the lower Susitna River
by sampling period,June through mid-October 1984.
~3 J J J "J J J J .1 ~I J I f J )J
-
CHUM SALMON
/
\~~:::;::e s~~~·(;iI =~fj 8 80 0 =~.Beaver Dam SlouOh 00 0 0 0 10 0 0 -
..,Beaver Dam S.C.f.2][;;iiI 121 0 tzI 0 n In -
~.~\.:S:=.uc::k~e,~s.=c.--:,;.~.~.!!!!!.f:~1Zl~D=:··:+.0;:;;.*0;+:0:=+=--+--=-1
Sauna S.C.• -rei 0 00 0 -_.
·.i~\
I)
":Q e\r--.,.'~l:.~O\SAMPLING PERIOD...,I JUN JUL AUG SEP OCT
--II"'.Site I n I n I n I 1I I~~.:..;Tr...:...;o:.....pp-e'-c-r-ee-k-s-.c-.---r::::~=-r:::D=·""D::=-,r:::[;jjjjJ:=-r:D~-=D~D:::=-r;D:;:;;"=D:::-'
Birch Slough [;jjjI .0 0 0 0 ClOD
..-
-
-
-
-
~Circular S.C.•[;j 12I IZIIOO r--_.
.~~Goose 2 S.C.I:i c;ji 0 0 0 0 r--_.
~~p,Ma.instem West Bank - - -10 0 0 0 0 _.
~i ,f1J~'-1_sl_.an_d_S_.C_.-+::::.=+~t:;i fZI (21 0 0 0 -
j lIP]Caswell Creek Mouth .0 0 0 0 0 0 0 10~~,)IarJL 'Rustic Wilderness S.C.fA 0 IZl 00 0 0 0 -
~~o:~~~lI~.~::\:;C:.c~·C..=0 ~~8BB-0 0 =-
~.Rolly Creek Mouth IZl 0 DODO 10 0 0";/Ji.Kroto SlouOh Head [21 0 0 0 °°°--
Eaoles Nest S.C.- - - - - -Old 0
Hooli~an s.c.•[;j 0 lZI ClOD -0
MEAN CATCH PER CELL
RELATIVE ABUNDANCE KEYo0.00 I:i 0.25-2.50
IZ]om ~0.25 •>2.50
-No sample
Cook Inlet
Figure 13.Seasonal distribution and relative abundance of juvenile chum
.-salmon on the lower Susitna River.June through mid-October 1984.
-31
6..,--------------------------.
5
4
3
2
_SIDE CHANNELS
lS::sJ TRIBUTARY MOUTHS
o -
E JUN L JUN E .JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAMPLING PERIOD
Figure 14.Juvenile chum salmon catch per cell at
modelled side ~hannels and tributary mouths
on thel.ower Susitna River by sampl ing
period,June through mid-October,1984.
~,
6..,-------------------------.,
125 175 225 275 325 375 400
ruRB10ITY eNTU)
7525
:iw
U ~
ct:
W
0-
J:
U'<-u
3
~
~
Figure 15.Juvenile chum salmon mean catch oer cell at
modelled side channels on the lo~er Susitna
River by turbidity increment,June through
mid-July 1984.
32
-
-
.'-
-
-
-
-
--
and in late June at Beaver Dam Slough.Sockeye juveniles were most
widely distributed within modelled sites upstream of Goose 2 Side
Channel (Figure 16).
Tributary mouths had the greatest densities of juvenile sockeye salmon
with a mean catch of 0.7 fpc.The highest CPUE for juvenile sockeye at
a tributary mouth was 1.2 fpc at Beaver Dam Slough.Side channels had a
mean sockeye CPUE of 0.1 fpc.Beaver Dam Side Channel had the highest
CPUE for a side channel of 0.7 fpc.Side slough CPUEs of sockeye
juveniles were minimal (0.03 fpc).Side channel CPUEls remained at low
levels through August in comparison to tributary mouth CPUEls which
varied greatly (Figure 17).No sockeye juveniles were captured in side
channels after August,however,sampling was limited.
sock.eye fry CPUEs were hi~hest in side channels where turbidities ranged
between 100 and 150 NTU (Fi gure 18).The numbers of sockeye juveni 1es
captured in.Beaver Dam Side Channel,immediately below and contiguous
with Beaver Dam Slough,may have been enhanced by site to site movement.
With Beaver Dam Si de Channel captures exc1 uded,the peak CPUE for
juvenile sockeye in side channels occurred at turbidities between 50 and
100 NTU.
Catches at Beaver Dam Slough and Beaver Dam Si de Channel show the
effects of turbidity as cover on the distribution of sockeye juveniles
(Figure 19).From late June through August,Beaver Dam Side Channel was
breached by the mainstem,the water was turbid,and sockeye CPUEls were
high.In early June and September,however,the head of the channel was
not breached,the water was clear,and few sockeye juveniles were caught
in this environment with little cover.In contrast,Beaver Dam Slough,
which had abundant aquatic vegetation cover,had high CPUEls of sockeye
juveniles in late August and September.Catches at Rolly Creek also
increased in 1ate August and remained fai r1y hi gh through early October
(Figure 19).
3.3 Habitat Modelling of Rearing Juvenile Salmon
The response of juvenile salmon habitat to variations in mainstem dis-
charge was modelled using two techniques:(l)the RJHAB model developed
in Marshall et a1.(1984)and (2)the IFIM hydraulic models discussed by
Bovee (l982).Suitability criteria for important microhabitat variables
are necessary as inputs to both models and criteria specific to the
lower reach of the Susitna River for juvenile chinook,coho,chum,and
sockeye salmon have been developed in Appendix A.
In the following discussion,results are presented by species.Each
presentati on inc1 udes modell ing resu1 ts from se1 ected si tes usi ng the
RJHAB or IFIM models,pooled results from all the sites modelled,and a
test of model verification.
No results from the Birch Creek Slough and Eagles Nest Side Channel
modelling sites are presented here.At Birch Creek Slough,there was no
measurable effect of mainstem discharge upon the site as mainstem
backwater at discharges less than 75,000 cfs did not extend to the site
and a blocked culvert at the head of the slough stopped mainstem water
33
-
-
-
-
/t:-t SOCKEYE SALMON~...J1I-:::l ~\~~o
':l:l>a'-...'<.~~o\SAMPLING PERIOD.....AUG SEP OCT
"I
.JUN JUL
1~K Site I II I n:I Jr I :II:I
Trapper Creek S.C.[2]0 [2]0 0 0 lZ1 0 CI
'i Birch Slough 0 ~0 0 IZ1 I2l 0 0 0
Sunrise S.C.1ZI -IZI IZI IZl [2]-- -
\Sunset S.C.---[;iii I2l [2]0 0 -
Bedver Dam Slough 0 •[;iii 0 0
(ijjjI Gil [;;jjI -~I~re-Dem s.c.0 fA [;jjJ IA ~~0 0 -
~~.Sucker S.c.[ZJ CjjjI 0 ~10 I2J 0 --~~Y'Sauna S.C.[ZI -0 0 0 [;;jjI 0 --
~Circu lor S.C.c;j 0 0 0 0 IZJ 0 --
10.,;[2]0 0 0 0 0 0 --.....Goose 2 S.C.
~Mainstem West Bank -- -0 I[]I2l 0 0 -)~1[J I"sland S.C.IZl 0 0 0 0 0 0 0 -
j Caswell Greek Mouth 0 -0 [;iii)[;;iiI [;;iiI 1121 10 0
Rustic Wilderness S.C.0 0 0 0 0 0 0 0 -
-='-Last Chance S.C.--0 0 0 0 -0 --='~~Bear Bait S.C.-0 0 0 0 0 0
--~.JZI IZI 10 (:;jjjI 0 [;iii)~[;iI (;iii....U~j Roll y Creek Mouth•-.....Kroto SIoU9h Head 121 0 0 0 0 0 "12]--~.,..Eagles Nest S.C.------0 0 0
~l HooliQdn S.C.Gil 0 10 0 0 0 0 -0
MEAN CATCH PER CELL
RELATIVE ABUNDANCE KEY
00.00 [;iii 0.25-2.50
[2]0.01-0.25 •>2.50
-No sample
\
.4•
Coolt Inlsf
-
Figure 16.Seasonal distribution and relative abundance of juvenile sockeye
salmon on the lower Susitna River,June through mid-October 1984..~
34
,_SIDE CHANNELS
IS::sJ TRIBUTARY t.40UTl-lS
.3-,------"-.;.--'-''''-''''-''''-'~~--''-'''-'--__.
2.8
2.6
2.4
2.2
2
1.8
1.6
1.4
1.2
0.8
0.6
a.'"
0.2
a
E JUN L JUN E JUL L JUL E AUG L AUG E SEP L SEP E OCT
SAMPUNG PERIOD
-Figure 17.Juvenile sockeye salmon mean catch per cell at
side channels and tributary mouths on the lower
Susitna River by samplinq period,June through
mid-October 1984.
-LESS BEAVER 0N0t s.C
cs::sJ WITH BEAVER DIW S.C
75 125 175 225 275 325 375 >400
TURBIDITY (NTU)
Figure 18.Juvenile sockeye salmon mean catch per cell at
modelled side channels on the lower Susitna
River by turbidity increment (with and without
Beaver Dam Side Channel),June through niid-
October 1984.-
35
-
-
-
..
BEAVER DAM SLOUGH
7
6
0
I
0u 5
I
.............40.....
G..
I J
~Q~w
!2 ...
G..
~::IE <
CO..
0
Z
0
E .IUN L .IUN E .lUL L JUL E AUG L AUG E SEP L SEP E OCT
SANPI.JNC PERIOD
BEA-";:R OAM SIDE CHANNEL
7
6
0
I
0
0 5
I
-'-'...40..
""a..
I J
5 Qw
~2'...
G..
::E::IE <CO..
0
Z
0
E .IUN L ./UN E .lUL L .IUL E AUG L AUG E SEP L SEP E OCT
SAMPlING PERIOD
ROLLY CREEl<MOUTH
7
6
0
I00 5
I...........40..
""II.
I 3
3
~2
::IE
0
e:.IUN L .IUN e:.lUL L .lUL E AUG L AUG E SEP L SEP E OCT
SAMPUNG PERIOD
FiQure 19.Juvenile sockeye salmon mean catch per
cell at Beaver Dam Slouqh,Beaver Dam
Side Channel,and Rolly"Creek Mouth by
sampling period,June through mid-
October 1984.
36
-
-
.i"""
-
--
.....
-
,....
from flowing through the site.The Eagles Nest Side Channel site was
modelled only twice at mainstem flows of 14,900 and 20,400 cfs and
therefore could not be readily extrapolated to discharges of 75,000 cfs.
All of the other sites were modelled at three or more discharges and
results were extrapolated to discharges ranging from 12,000 to 75,000
cfs.The WUAs and site areas at the RJHAB sites were not adjusted to a
reach length of 1,000 ft as were the IFIM WUAs.Lengths of all the
RJHAB sites are listed in Appendix Table B-3,so that the WUAs could be
adjusted if desired.
The instream flow results have been generated only to discharges of.
75,000 cfs because it is very difficult to collect data at discharges
greater than 75,000 cfs.At 75,000 cfs,most of the side channel sites
have very large flows and are poor habitat for juvenile fish.At higher
discharges,the entire flood plain becomes full and the flows are barely
constrained within the side channels.Refuge for the juvenile fish at
these times presumable include large backwater areas and small side
channels which are infrequently flooded.
At Island and Trapper Creek side channels,both RJHAB and IFIM models
were run on the same transects.Comparative resul ts for these two
models are given in Appendix C.The summary figures presented here
incorporate data from the RJHAB model at these two side channels.
The abil ity of the RJHAB model s to extrapol ate WUA between di scharges of
12,000 and 75,000 cfs was rated unacceptable ~o good (Table 5).Some
model s were rated fair because there were no habitat measurements taken
at discharges just above overtopping of the side channel.Eagle's Nest
Side Channel was rated unacceptable because measurements were taken on
only two occasions at discharges less than 21,000 cfs.
The IFIM models were evaluated according to hydraulic criteria on the
basis of excellent to acceptable (Appendix D).Acceptable ranges of the
models usually extend to over 60,000 cfs (Table 6).The models were run
and WUAs generated at side channel flows which corresponded to dis-
charges ranging to 75,000 cfs,so reliability at these flows is unknown.
At discharges below overtopping,the WUAs of IFIM sites at flows of 5 or
6 cfs were used,except at Trapper Creek Side Channel where a site flow-
mainstem discharge rating curve for unbreached conditions developed by
Quane et al.(1985)was used to estimate unbreached flows.
Since suitability criteria for chinook salmon juveniles have been
developed for both turbid (>30 NTU)and clear «30 NTU)conditions,
several assumptions were made.Tributary mouth sites were assumed to be
clear (>30 NTU)at all discharges less than 75,000 cfs.This is not
always the case,as occasionally turbid mainstem water may back up into
tributary mouths with a rapid increase in mainstem stage.Also spring
runoff or large storms may increase turbidities at tributary mouths to
over 30 NTU.Available data,however,have indicated turbidities at
tributary mouths are normally less than 30 NTU (Figure 3).At side
channel/slough sites,turbidities were assumed to be greater than 30 NTU
when the site was breached and less than 30 NTU when the site was not
breached.In early June,September,and early October,turbidities in
side channels were sometimes less than 30 NTU (Figure 3).Many of the
37
Table 5.Evaluation of RJHAB model quality for extrapolating WUAs over the range of
12~OOO to 75,000 cfs as measured at Sunshine gaging station,1984.
Site
Hooligan Side Channel
Eagle's Nest Side Channel
Kroto Slough Head
Rolly Creek Mouth
Bear Bait Side Channel
Last Chance Si de Channel
Rustic Wilderness Side Channel
Caswell Creek Mouth
Island Side Channel
Coose 2 Side Channel
Sucker Side Channel
Beaver Dam Slough
Beaver Dam Side Channel
Sunrise Side Channel
Birch Creek Slough
Trapper Creek Side Channel
Number of
Habitat Measurements
5
2
4
4
4
5
5
3
5
4
4
4
3
4
3
4
Model Quality'
Cood
Unacceptable
Fai r
Cood
Fai r
Fair
Cood
Fair
Cood
Fai r
Cood
Cood
Cood
Fair
Good
Good
....
~,
Model quality definitions:
1.Good -Side Channels:Measurements spaced so as to cover the range of mainstem
discharges above breaching to 75,000 cfs.Models include information about
unbreached,barely breached,and a minimum of two other breached flows,one near
75,000 cfs.
Tributary Mouths:Models include information when no backwater,moderate
backwater,and high backwater present.
2.Fair -Side Channels:Model missing information concerning habitat when channel is
barely breached~or other flows given above.
Tributary Mouths:Not enough measurements to accurately describe amount of
backwater effect.
3.Unnacceptable -Less than three data points -cannot describe a curve.
-
-
-
-
Table 6.Discharge ranges of IFIM models
hydraulics are rated acceptable,1984.
at lowe r Sus itna Ri ve r
Data taken from Appendix D.
sites for
Site
Island Side Channel
Mainstem West Bank
Circular Side Channel
Sauna Side Channel
Sunset Side Channel
Trapper Creek Side Channel
38
Acceptable Range
35,000 to 70,000 cfs
18,000 to 48,000 cfs
36,000 to 63,000 cfs
44,000 to 63,000 cfs
32,000 to 67,000 cfs
20,000 to 66~000 cfs
-
-
-
.-,
-
model sites were not breached during these periods of low mainstem
discharge ..Turbidities in side sloughs were usually less than 10 NTU.
3.3.1 Chinook Salmon
Chinook salmon juveniles were captured at all of the study sites with
the exception of Last Chance Side Channel (Figure 8).Since chinook
juveniles were widely distributed,results from all sites modelled with
RJHAB and IFIM techniques will be presented.
Graphs of the weighted usable'area responses to mainstem discharges for
all sites not presented here are included in Appendix B.Appendix B
also contains the tabulated values of weighted usable areas at 3,000 cfs
increments as digitized from these graphs (including site graphs pre-
sented here).Also tabulated are habitat indices which were calculated
by dividing the weighted usable area at a given discharge by the site
area at the same discharge.
At the Rolly Creek,Caswell Creek,and Beaver Dam Slough tributary mouth
sites,the responses of weighted usable area to mainstem discharge were
very similar.The Rolly Creek mouth weighted usable area response to
discharge is presented here as an example (Figure 20).The great
increase in weighted usable area with discharge above approximately
45,000 cfs is due to the effect of mainstem backwater causing large
increases in area,depth,and amount of cover.
At side channel/slough sites,the responses of weighted usable areas to
mainstem discharge was varied.Normally,the weighted usable area
increased greatly after overtoppi ng and then decreased with further
increases in mainstem discharge as at Kroto Slough Head (Figure 20).
The increase in weighted usable area after overtopping is due to in-
creases in area and also increases in cover suitability as turbidity
improves cover.As discharge increases with site flow,velocities
initially become more suitable,but then as flows continue to rise,
velocities become unsuitable and WUA decreases.
At Sucker Side Channel,backwater effects buffer the velocities from
becoming too high and so weighted usable area increases after overtop-
ping and then remains nearly the same to a discharge of 45,000 cfs after
which it rapidly increases (Figure 20).At approximately 60,000 cfs,
WUA's begin to decline at this site,however,as velocities and depths
become unsuitable.At other sites,WUA held quite constant after
overtopping or slowly increased (see Appendix B).
When WUA's from three tributary mouths are pooled there is no large
change in WUA until approximately 45,000 cfs when the WUA increases
greatly with discharge (Figure 21).By dividing the WUA at 3,000 cfs
increments by pooled area for the three sites and plotting the habitat
index,it becomes apparent that the change in WUA is not simply due to
increases in site area.Increases in habitat indices are due to in-
creases in the amount of instream cover,more suitable velocities,and
deeper water which may also provide cover.
39
ROLLY CREEl<~OUTH40
35 """
::30,;.
~
~25
~-ac
~~20~~£.15Cl
I!!~
I:
8 10
~
5 -0
10 30 50 70~USands§.......NSTE""DISC GE AT UNSHINE (cfs)
~
KROTO SLOUGH HEAD9
8 -.
::7
a
~6
~
~-a 5 -c
~~
~a 4:£
a 3~
(>
~2
"-0
10 30 50 70 90~USOnds§MAlNSTEM OISC GE AT UNSH/NE (cfs)
~
SUCKER SlOE CHANNEL
~4:<
!
~:3H ,,
c ",-~ll ,
~a:£2
Cl
I!!
I:-Cl~
0 .....
10 30 50 70
MAlNSTEM OISC~:;~a~s~UNSHINE(cfs)
Figure 20.Weighted usable area for juvenile
chinook salmon at Rolly Creek ~1outh,
Kroto Slough Head,and Sucker Side
Channel study sites as a function of
mainstem discharge,1984.
40
.....
0.16
0.15
0.14
0.1.3
X 0.12
w
0
~0.11
!<0.1~0.09
0.08
0.07
0.06
0.05
10 .30 50 70
(Thousands)
MAINSTEM DISCHARGE AT SUNSHINE (efa)
Fi gure 21.Weighted usable area and habitat indices for juvenile
chinook salmon at tributary mouth sites as a function of
mainstem discharge,1984.
41
When WUA's from the modelled side channels/sloughs are pooled,WUA's
increase greatly to approximately 40,000 cfs and then very gradually
decline (Figure 22).Habitat indices for the pooled side channels show
a similar rise to a peak at 40,000 cfs but then a rapid decrease to
approximately 60,000 cfs when the habitat index levels off.The rela-
tively more rapid decrease in the habitat index is due primarily to
velocities and depths becoming very unsuitable at the higher dis~harges.
Turbidity has been shown to be an important determinant of juvenile
chinook distribution (Figure 10).Turbidity varies in the Susitna River
from the east bank to the west bank downstream from the Chul itna and
Talkeetna river confluences (Figure 4).In formulating the pooled side
channel/slough response of juvenile salmon habitat,it was desirable to
weight turbidity as it varies from site to site.
Although turbidity data for the model sites are limited,an average
turbidity for the side channels modelled during the period from June
through August was calculated in Appendix Table B-1.A preliminary
suitability index for high turbidity was then fit to the data in Figure
10 (Table 7).This index is specific only to the turbidity regimes of
lower river side channels and is undefined for application to tur-
bidities of less than approximately 100 NTU.When the turbidity indices
and mean turbi diti es were combi ned,WUA estimates for the sites were
weighted differently (Table 8).
When the WUA estimates for each site are adjusted by these factors and
the WUA's are again totalled,the WUA and habitat index response ad-
justed for turbidity for the side channels combined can again be ex-
amined (Figure 23).There is very little change from the previous
unadjusted graph in the shape of the WUA response curve,but the magni-
tude was reduced by almost 40%.Similarly,the shape of the habitat
index responses curve has also been changed very little by these
adjustments.The lack of change in shape of these curves suggests that
the responses of the side channel WUAs and habitat indices are similar
for most of the sites.
The mean seasonal chinook salmon habitat index for the 15 side channels
and four tributary mouths were cal cul ated and compared wi th mean chi nook
catch (Figure 24).The positive relationship was statistically signifi-
cant (p <0.001)but not very strong.Most of the correlation was due to
the large catch (5.16 fpc)and habitat index (0.19)at Caswell Creek
mouth.Another outlier is Beaver Dam Slough with a habitat index of
0.17 and a mean catch of 0.17 chinook per cell.
3.3.2 Coho Salmon
Si nce coho salmon were captured in number (more than 20)only at the
tributary mouth sites,only results from these sites will be presented
here.In Appendix B,values of WUA's and habitat indices at 3,000 cfs
increments for these areas are presented.
The response of WUA to mainstem discharge at the three tributary mouths
varied (Figure 25).At Caswell Creek mouth,WUA rose with discharge due
to increases in area and the amount of preferred cover.At Rolly Creek
42
-
-
-
-
-
....,
SIDE CHANNELS /SLOUGHS
CHINOOK SALMON70
60
..........-,-IT 50~
LS.......
~.tJ 40
l:
~51
~5 30::Jt.
0~20::t::
Q
lJJ:t
10
0
r-10 30 50 70~hQUSCndS§
MAINSTEM DISC ARGE AT UNSHINE (cfs)
-0.06
--0.05
0.04
Xw
0
?;
~0.03
r-t:::
(D
~
0.02
0.01
0
10 30 50 70~USOndS~MAINSTEM DISC GE AT UNSHINE (cfs),-
-Figure 22.~/eighted usable area and habitat indices for juvenile
chinook salmon at side channel/slough study sites as a
function of mainstem discharge,1984.
-43
Table 7.Preliminary juvenile chinook salmon turbidity criteria
derived from lower Susitna River side channel distribution
data for turbidities greater than 100 NTU.These criteria
are only applicable to lower Susitna River side channels.
-
.-
Mean
Turbidity (NTU)
101 -200*
201 -250
251 -300
301 -350
350
Suitabil ity
1.00
0.65
0.55
0.40
0.15
-
*Suitability index for turbidities of less than 101 NTU is undefined
and may be greater than 1.0.
Table 8.Weighting factors for turbidity by side channel site for
analysis of juvenile chinook salmon habitat use,1984.
-
Mean
Site Turbidity (NTU)
Hooligan Side Channel 377
Kroto Slough Head 388
Bear Bait Side Channel 254
Last Chance Side Channel 365
Rustic Wilderness Side Channel 118
Island Side Channel 215
Mainstem West Bank 279
Goose 2 Side Channel 194
Circular Side Channel 241
Sauna Side Channel 266
Sucker Side Channel 140
Beaver Dam Side Channel 139
Sunset Side Channel 152
Sunrise Side Channel 121
Trapper Creek Side Channel 499
44
Turbi di ty
Weighting
Factor
0.15
0.15
0.55
0.15
1.00
0.65
0.55
1.00
0.65
0.55
1.00
1.00
1.00
1.00
0.15
....,
-
-
40
35
.........-30
cT
II'-'
~25
~-8I::
~g 20~S~15c
~CI 10~
5
0
10
0.04
0.035
0.03
x 0.025
I.&Jc
~
!<0.02
!::
III~0.015
0.01
0.005
0
10
SIDE CHANNELS /SLOUGHS
ADJUSTED CHINOOK WUA
30 50 70
(Thousands)
MAINSTEM DISCHARGE AT SUNSHINE (cfs)
30 50 70
(Thousands)
MAINSTEM DISCHARGE AT SUNSHINE (cfs)
-
Figure 23.Turbidity adjusted weighted usable area and habitat indices
for juvenile chinook salmon at side channel/slough study
sites as a function of mainstem discharge,1984.
45
CHINOOK MODEL VERIFICATION
(SIDE CHANNELS AND TRIBUTARY MOUTHS)
6 i I
54 Y =0.15 +15.04x
p <0.001
::J 4 I 2
w r =0.39
0
a::wa..
J:.35
~2
~
-I=':>
O'l
I
1
o 18 ....I T'iii iii Iii iii i I I I
o 0.02 0.04 0'.06 0.08 0.1 0.12.0.14 0.16 0.18
SEASONAL MEAN HABITAT INDEX
Fiqure 24.Juvenile chinook salmon mean catch per cell versus seasonal mean habitat
indices at side channel and tributary mouth modelling sites on the lower
Susitna River,1984.
.J ,J J ,J }J J J ,J ,J .)J ~)
Figure 25.Weighted usable area for juvenile
coho salmon at the Caswell Creek,Rolly
Creek,and Beaver Dam Slough tributary
study sites as a function of mainstem
discharge,1984.
47
mouth,the WUA first decreased with discharge due to the formation of
zero velocity backwater from a free flowing state without major changes
in cover or area.At higher discharges,the WUA increases due to a rise
in area and usable cover.At Beaver Dam Slough,these effects of
backwater formation and increases in cover inundated offset one another
so that there was little change in WUA with discharge.
When the WUA's from all three sites are summed (Figure 26),there is
little change in WUA until approximately 50,000 cfs when the WUA begins
to increase greatly with discharge.When the effect of change in area
is taken out by calculating a habitat index,site quality decreases
initially as the backwater is formed and then begins to increase as
cover is inundated by backwater.
The mean habitat index for the season (May 15 to October 15)was cal-
culated for the four tributary mouths.Since Birch Creek Slough was a
natal.area,only catches from mid-July through mid-October were used in
calculating the mean site catch.The mean catch per cell of coho
juveniles increased with the mean habitat index but a linear regression
was not statistically significant at the 0.05 level (Figure 27).None
of the side channels had mean seasonal habitat indices greater than 0.05
and most were 0.03 or less,primarily due to the lack of suitable cover
types.
3.3.3 Chum Salmon
Chum salmon were widely distributed at all of the side channel sites
sampled from early June through July 15 (Figure 13).Therefore,graphs
of the WUA response as a functi on of ma i nstem di scha rge for a11 the side
channel/slough sites not presented here are included in Appendix B.
Also tabulated in Appendix B are values of WUA's and habitat indices at
3,000 cfs increments as digitized from the graphs.
Responses of WUA's at the sites to increases in mainstem discharge were
variable.At Rustic Wilderness Side Channel,WUA greatly increased
after overtopping and then declined with further increases in discharge
as velocities and,depths became unsuitable (Figure 28).At other sites,
for example Last Chance Side Channel,the increase in WUA after overtop-
ping was considerably less while at Trapper Creek Side Channel (Figure
29),.WUA's decreased after overtopping.At Sunset Side Channel,WUA
increased after overtopping until about 53,000 cfs when WUA·quickly
declined.The other sites a.lso showed variations of these response
curves (see Appendix B figures).
When WUA's from all modelled side channel/slough sites are pooled,the
peak in WUA's for the sites occurs at a discharge of 40,000 to 52,000
cfs (Figure 30).Above this discharge range,WUA's decrease rapidly due
to unsuitable velocities and depths.Habitat indices for the same
pool ed sites are constant through about 24,000 cfs and then decrease
steadily.
Chum salmon use of side channels was affected by turbidity (Figure 15),
and since turbidity varied from site to site,WUA's for each site were
adjusted for turbidity.Since chum salmon outmigration is mostly
48
"""
-
~
TRIBUTARY MOUTHS
20 (BIRCH SLOUGH EXCLUDED)
~
19
"18.J--0-17.,.......
h 16
fr~c
~g 15~~~14
a~13:r
()
!i 12
11
10
~10 30 50 70~USOncls~tAAlNSTEM DISC GE AT UNSHINE (cts)
0.1
0.09
~
0.08
0.07
P"'''xw.0.06a
?;
!<0.05
!:::::
lD 0.04~
0.03
0.02
0.01
0
10 30 50 70~usand.~MAtNSTEM DISC GE AT .NSHINE (eta)
Figure 26.Weighted usable area and habitat indices for juvenile
coho salmon at tributary mouth sites (excluding Birch
Creek Slough)as a function of mainstem discharge,1984.
49
COHO MODEL VERIFICATION
(TRIBUTARY MOUTHS ONLY)
5 I I
4
:i I CASWELLwcU
It:3 -IJ.J
Q..
I
U
3 2-g
:::!.
<..n
C)
1 4 BEAVER DAM
a
BIRCH ROLLY
a a
o +I !!i !!iii iii iii
o 0.02 0.04·0.06 0.08 0.1 0.12 0.14
HABITAT INDEX
Figure 27.Juvenile coho salmon mean catch per cell versus seasonal mean habitat
indices at tributary mouth modelling sites on the lower Susitna River,
1984.
J J J ~J I .J !J j )J J 1 )J J )
r'"
CHUM WUA
RUSTIC WILDERNESS SIDE CHANNEL34
32
30
......28::::26cT..·24......
i5.-..22
~.g 20c
~~18~~16
~IIN.chad
0 14 t~12
CI 10iii~8
6
4
2--10 .30 50 70
»>~hOUSOndS~
MAINSn:t.4 DISC ARGE AT UNSHINE (ets)
-
LAST CHANCE SIDE CHANNEL22
20
......18
::::
cT 16 Bra.ched,,-411......ti5.-..14-
~-8 12c
~g
~5 10
:£
0 8
~6CI
iii~4
2 ~--
0
10 .30 50 70 90~usandS~MAINSn:t.4 DISC GE AT UNSHINE (ets)
-
Figure 28 .Weighted usable area for Juvenile chum>salmon at Rustic
Wilderness and Last Chance Side Channel study sites as
a function of mainstem discharge,1984.
51
CHUM WUA
TRAPPER CREEK SIDE CHANNEL
50
---045
,...t...40-a-
01....,35
~~-ll 30c
~Hl
~~25
::3t:.
0 20~
J:
0
iij 15:t
10
5
10 .30'59 70
SUNSET SIDE CHANNEL
54
52
50 -,...
48...-a-46
01....,
~44
42c
~g 40ID:I~~38 ~
::3t:.
0 36
~34J:
0
iij 32:t
30
28
26
10 .30 50 70
~hOUSOndS~
MAINSTEt.l DISC ARGE AT UNSHINE (cfs)
Figure 29.Weighted usable area for juvenile chum sallmn at the
Trapper Creek and Sunset Side Channel study sites as
a function of mainstem discharge,1984.
52
-
420
410
400
......390:::
r:T 380
n
'-'370
i:S,...360~-8c:350~~
~~340
:£330
CI 320~r 310C)
W
3:300
290
280
270
10
0.7
0.6
0.5
Xw
CI
~
~0.4-
!:::::
lD
:f
0.3
0.2
0.1
10
SIDE CHANNELS /SLOUGHS
CHUM SALMON
30 50 70
(Thousonds)
MAINSlEM OISCHARGE AT SUNSHINE (c:fs)
30 50 70
(Thousonds)
MAINSlEM DISCHARGE AT SUNSHINE (cts)
Figure 30.Weighted usable area and habitat indices for juvenile
chum salmon at side channel/slough study sites as a
function of mainstem discharqe,1984.
53
completed by July 15,turbidity data contained in Appendix Table B-1
were examined through July 15.Since turbidities greater than 200 NTU
appear to affect use greatly (Figure 15),site WUA's were adjusted for
periods when the turbidity exceeded 200 NTU.Adjustment factors for the
sites ranged from 0.50 to 1.0 (Table 9).
When the chum salmon WUA 's were adjusted for turbidity and again to-
talled,very few changes were noted in the shape of the WUA of habitat
index response curves although both WUA'S and habitat indices decreased
(Figure 31).Since there was little change in these curves,it appears
that the shapes of the chum WUA responses at all the side channels are
very similar and therefore weighting the sites differently by turbidity
only changes the magnitude of the response.
Mean chum salmon adjusted habitat indices were calculated for the period
from May 15 through july 15 and compared with mean chum catch during the
same time period (Figure 32).There was no sampling effort at two of
the side channels,Mainstem West Bank and Sunset Side Channel,during
this time so they are not included in this graph.The correlation
(0.54)between the seasonal habitat index and chum catch was significant
at the 10%probability level but not at the 5%probability level.
3.3.4 Sockeye Salmon
Sockeye salmon were most numerous at the tributary mouth sites with most
side channels having some use (Figure 16).Presented here or in Appen-
di x B are graphs of the WUA responses to di scharge of the three tribu-
tary mouths and the four side channels (Beaver Dam,Sucker,Sunrise and
Sunset)whi ch were found to have sockeye salmon present more than half
the times sampled.
The typical response of WUA at the tributary mouths to increases in
discharge was a steady increase as shown here by the modelling results
from Rolly Creek (Figure 33).The WUA increased as the backwater zone
increased because sockeye find zero velocity water most suitable and
because site area and cover also increased greatly with discharge.The
WUA response at Sucker Side Channel was similar to that of the tributary
mouths as WUA generally increased wi th di scharge after overtopping.
Th iss i te is i nfl uenced grea t 1y by bac kwa ter effects from the side
channel at its mouth.At Beaver Dam Side-Channel,WUA increased after
overtopp"ing and then declined somewhat (Figure 34).At Sunset Side
Channel,WUA fluctuated irregularly with discharge as the small amount
of usable habitat along the margins of the site moved back and forth
with flow changes.
At the combined tributary mouth sites,both WUA and habitat indices
increased above discharges of approximately 30,000 cfs (Figure 35).At
the pooled side channel/sloughs,on the other hand,WUA's also increased
after approximately 30,000 cfs while habitat indices generally declined
from the peak at 12,000 to 24,000 cfs (Figure 36).The decrease in the
habitat index is due to the steadily increasing velocities in the side
channels with increases in flow.No adjustments in turbidity are
necessary for the four side channel/slough sites as these have very
54
~l
Table 9.Weighting factors for turbidity by site for analysis of
juvenile chum salmon habitat use,1984.
Site
Hooligan Side Channel
Kroto Slough Head
Bear Bait Side Channel
Last Chance Side Channel
Rustic Wilderness Side Channel
Island Side Channel
Mainstem West Bank
Goose 2 Side Channel
Circular Side Channel
Sauna Side Channel
Sucker Side Channel
Beaver Dam Side Channel
Sunset Side Channel
Sunrise Side Channel
Trapper Creek Side Channel
55
Sampling Period
When Turbidity
Exceeds 200 NTU
June 16-30
June 16-30
June 16-30
June 16-30
July 16-30
July 1-15
June 16-30
July 1-15
July 1-15
June 16-30
July 1-15
July 1-15
July 1-15
July 1-15
June 16-30
Turbidity
Weighting
Factor
0.50
0.50
0.50
0.50
1.00
0.75
0.50
0.75
0.75
0.50
0.75
0.75
0.75
0.75
0.50
Figure 31.Turbidity adjusted weighted usable area and habitat
indices for juvenile chum salmon at side channel/slough
study sites as a function of mainstem discharge~1984.
56
}----}l 1 J 'I -t
[]
o
o
o
I
0.34
a
o
..
0.3•n
I I ,J
0.14 0.18 0.22 0.26
SEASONAL MEAN HABITAT INDEX
J
o
CHUM MODEL VERIFICATION
(SIDE CHANNELS/SLOUGHS ONLY)
-r
0.1
4.5
4 -
3.5 -
::fw 3-()
ffia..2.5 -
J:
I a
()
3 2 -
01
~1.5 -
"I
~
1 -
0.5 -,a
o I o 0
1
0.06
Figure 32.Juvenile chum salmon mean catch per cell versus seasonal mean habitat
indices at side channel and slouqh modellinQ sites on the lower Susitna
River,1984..
SOCKEYE WUA
ROLLY CREEK MOUTH
110
100
"90
..;-0-ao
II.....
~70
~.g 60c:
~g
~~50
:€
0 40
~.30~
ILl~20
10 --------'l
0
10 .30 50 70
~usancls~MAtNSTEM OISC GE AT UNSHINE (c:fs)
SUCKER SIDE CHANNEL
6
5.........;-0-
Il.....4
~
~.gc:
~5l :3~~:£Sre.cud
0 2 t~
"G:i~
----
~,
.30 50 70
(Thousands)
IAAINSTEf.4 DISCHARGE AT SUNSHINE (cts)
O+----,.--""'-.,....----r---r----r---.,r-----!
10
Figure 33.Weighted usable area for juvenile sockeye salmon at
Rolly Creek '·1outh and Sucker Side Channel study sites
as a function of mainstem discharge,1984.
58
~I
SOCKEYE WUA
BEAVER DAM SIDE CHANNEL
,....
4.5
4
,........3.5-0-
Il.....3 ---------------
ii-.~-8 2.5l:
~g
~g 2:liS
0 1.5~
J:
Cl
i
0.5
Breached
.:5()50 70
(Thousands)
MAlNSTEM DISCHARGE AT SUNSHINE (c15)
O-t-----,.---,..----r--....,;,.,..r---.....----,,----!
10
SUNSET SlOe:CHANNEL10.-------------------------,
,....9....-0-
,.!!,.
ii-.8~~l:
~g
~5=£.7
0~J:
!""""Qw~6
.30'50 70
(Thousands)
MAINSTEM DISCHARGE AT SUNSHINE (cfs)
5+-----,.---T-:----,---,-r----r-----,----j
10
Figure 34.Weighted usable area for juvenile sockeye salmon at the
Beaver Dam and Sunset Side Channel study sites as a
function of mainstem discharge,1984.
59
~
TRIBUTARY MOUTHS
130 (BIRCH SLOUGH EXCLUDED)-120
110.........-100
cT
~90
~~~80 -c
~g 70~5 60:£
c 50~
(.)40~~
30
20
10
10 30 50 70~USandS~MAtNSTEM DISC GE AT UNSHINE (cfs)
~
0.42
0.4-
~0.38
0.36
0.34-
x 0.32....c
~0.3
~0.28t::
~0.26
0.24-
0.22
0.2
0.18
~
0.16
10 30 50 70
~usands~MAlNSTEM DISC GE AT UNSHINE (cfs)
~
Figure 35.Weighted usable area and habitat indices for juvenile
sockeye salmon at tributary mouth study sites on the lower
Susitna River as a function of mainstem discharge.1984.
60 ~,
SIDE CHANNELS /SLOUGHS
SOCKEYE SALMON
30 50 70
(Thousands)
MAINSTEM DISCHARGE AT SUNSHINE (efs)
20
19
""'18~
0-17II.....,
~16~-8c
ljg 15~g:£14
a
~13:r
8
3:12
11
10
10
7030'50
(Thousands)
MAINSTEM DISCHARGE AT SUNSHINE (efs)
0.2 -r-----,-----------------------,
0.19
0.18
0.17
0.16
0.15
0.14
0.13
0.12
0.11
0.1
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
O-l-----,----..----,----r---~---._--__T--__;
10-
Figure 36.Weighted usable area and habitat indices for juvenile
sockeye salmon at side channel and slough study sites on
the lower Susitna River as a function of mainstem discharge,
1984.
61
similar turbidity regimes,being located on the same general location on
the river.Use of many of the other side channels is probably limited
by turbi di ty.
The mean seasonal habitat index for sockeye salmon at the four tributary
mouths and four side channel sites was calculated for the period from
May 15 to October 15,1984.The mean catch of sockeye salmon juveniles
was positively related to the mean habitat index (Figure 37).High
turbidities and velocities within the other side channels presumably
limited use by sockeye salmon juveniles.
62
-
.....
-
,~
~"
-
-.i·
$1 1 1 I ))~J j ~l J
SOCKEYE MODEL VERIFICATION
(SIDE CHANNELS AND TRIB MOUTHS)
y =0.10+2.21x
P <0.001
2r ;:0.78
o
o
o
o
0.2
SEASONAL MEAN HABITAT INDEX
0.4 '
Figure 37.Juvenile sockeye salmon mean catch per cell versus seasonal mean habitat
indices at side channel and tributary mouth modelling sites on the lower'
Susitna River,1984.
4.0 DISCUSSION
4.1 Chinook Salmon
Chinook salmon were widely distributed throughout tributary mouths and
side channels of the lower Susitna River.Densities of juvenile chinook
were highest within tributary mouths.This distribution of chinook fry
substantiates earlier observations (ADF&G 1981a;Dugan et al.1983)'that
densities of chinook are generally highest at tributary mouths.Caswell
Creek mouth had the highest CPUE of juvenile chinook salmon in the lower
river and'appears to be a major rearing or holding area.
Chinook salmon juveniles used side channels for rearing in both the
middle and lower Susitna River after moving from tributary natal areas.
Redistribution of chinook fry from natal areas to lower density rearing
areas has also been observed in the Deshka River (Delaney et ale 1981)
and Montana Creek (Ri is and Fre i se 1978).Th is phenomenon refl ects a
downstream movement or dispersal of the 0+age fish (ADF&G 1981c).Most
of the 1+chinook juveniles have outmigrated by August l.
Use of tributary mouths is limited by the amount of instream cover and
suitable velocities.Also,depth may be important to chinook juveniles
in tributaries because it probably provides cover in slightly turbid
water (10 to 20 NTH)(Appendix A).At Caswell Creek mouth,catches of
juvenile chinook were low in September when the mainstem water stage
dropped and depths decreased,velocities increased,and amount of cover
was reduced.
Use of Susitna River side channels by chinook juveniles for rearing is
widespread although it is limited by turbidity in portions of the lower
reach.Side channels located in the Talkeetna River plume had much
higher use than those located in the more turbid Chulitna River plume or
those located further downstream where the water of these two tri-
butaries are mixed.Si de channel catch rates of juvenil e chi nook (i n
similar habitat)in the middle Susitna River in 1983 were approximately
four times higher than those in the lower river in 1984 (Dugan et al.
1984).
Since lower Susitna River side channels are used less by ,chinook juve-
niles than middle river side channels,it is not surprising that sloughs
are also used less in the lower reach than in the middle reach.As
water levels decreased in the fall and side channel heads dewatered,
there were very few chinook fry at slough sites in the lower river to
take advantage of the lowered turbidity.Also the side sloughs in the
lower river contain little cover.
Instream flow effects upon juvenile chinook salmon are related to
backwater effects at the tributary mouths and side channel/slough sites
and to breaching and side channel flows.When a side slough is not
overtopped by the mainstem,access is usually poor and cover is limited.
64
-'
-
-
......
,~
.-
At tributary mouths,backwater effects increase chi nook use si gni fi-
cantly because of increases in instream cover and depth and decreases in
water velocity.Also,turbid backwater from the mainstem sometimes
intrudes into the sites with rapid rises in mainstem stage.Pooled data
from three tributary mouths showed major increases inWUA at mainstem
discharges greater than 45,000 cfs.
If the study sites had been chosen further upstream in the tributary
mouths,WUAs would have begun to increase at a higher discharge,so the
45,000 cfs figure is not absolute.At Birch Creek Slough,for example,
there were no measurable effects of backwater to mainstem discharges of
72,000 cfs.In general,increases in mainstem discharge increase the
amount of juvenile chinook salmon habitat at tributary mouths.Also,
these backwaters may increase access into tributaries where rearing
could occur by decreasing water velocities at the mouth.
Within side channel/slough sites,mainstem discharge is very important.
When sloughs are breached,the water becomes turbid and cover for the
chinook juveniles is improved.High turbidities,however,may also
limit use of side channels (Figure 10).High turbidities generally
occur from mid-June through September (especially duri ng hi gh di s-
charges),while turbidities are much lower during the rest of the year.
Turbidity also varies spatially within the river.Chulitna and
Talkeetna river plume effects extend at least 20 miles downriver (Figure
4).Sites located within the Talkeetna River plume have much lower
turbidity and higher juveni 1e chi nook salmon use.
Mainstem discharge initially increases chinook WUA within a side chan-
nellslough after it overtops but with further increases in flow,WUA
usually remains constant or .declines while the proportion of usable
chinook habitat declines.The RJHAB model shows a decline in WUA with
increasing discharge which is greater than that shown by the IFIM model
(Appendix C).
The results obtained by pooling WUA from all modelled sites should not
be directly extrapolated to represent the entire lower reach.If the
modelling sites would have been chosen randomly,many more large,high
velocity side channels with extremely little usable habitat would have
been modelled.This study was designed to sample proportionately more
side channels with.usable habitat which would represent a diversity of
channel types in the lower river.The modelled side channels represent
a wide range of sizes and shapes of channels with diverse breaching
flows,and so these resul ts need to be coupl ed wi th a stratification of
lower river side channels by breaching discharge and channel size and
type.The most important side channel complexes in the lower Susitna
River for juvenile chinook salmon rearing are located within the low
turbidity plume of the Talkeetna River.Other side channels or side
channel complexes should be weighted according to their mean turbidity
level.
4.2 Coho Salmon
Juvenile coho salmon in the lower river were found mostly within tribu-
tary mouths.Tributaries and tributary mouths were al so the most
65
important rearing areas for juvenile coho salmon in the middle Susitna
River (Dugan et a1.1984).Upland sloughs were also used by coho salmon
for rearing in the middle river t but upland slough habitat is limited in
the lower river and was not sampled during this study.
The heavy use of tributary mouths by juvenile coho is due in part to
coho in tributary mouths rearing near their natal areas.Their limited
use of side channels maybe due to their documented tendencies to favor
waters with relatively low turbidities.Sigler et a1.(1984)t for
example,found that a larger number of juvenile coho salmon emigrated
from exper'imenta1 laboratory channels with turbidities of 25-50 NTU than
from clear water channels.In another laboratory study,Bisson and
Bilby (1982)established that coho salmon avoided turbidities exceeding
70 NTU.Turbidities in lower Susitna River side channels during June
through August often greatly exceeded 100 NTU.
Use of tributary mouths by juvenile coho varied greatly seasonally and
from site to site.Rolly Creek and Beaver Dam Slough CPUEls of coho
salmon generally increased from early summer to late fall (Figure 12).
This occurrence may be due to both the immigration of coho juveniles and
a decrease in site area.The area of Rolly Creek was reduced by approx-
imately 63%from late June and July to September and early October t
while the area of Beaver Dam Slough was reduced by approximately 33%
between these two time periods.In Birch Creek Slough t on the other
hand,a relatively high CPUE occurred in early summer with much smaller
values throughout the summer and fall.The relatively high CPUEls in
early summer at Birch Creek Slough are probably due to a natal effect.
Barrett et a1.(l985)reported that Birch Creek has a spawning run of
coho salmon.
A comparison of juvenile coho catch rates between tributary mouths and
the Talkeetna outmigrant trap (RM 103.0)suggests that a redistribution
of juvenile coho into suitable rearing habitat peaks from late July to
early August.The catch per hour of age 0+coho at the Tal keetna
outmigrant trap increased during this time period while CPUEts at
tributary mouths also changed greatly.Birch Creek Slough t which
habitat modell ing indicates to be rel ative1y poor coho tributary mouth
rearing habitat (Figure 27),shows a reduction in CPUE in late July,
following natal emigration t while Caswell Creek,a site evaluated as
having relatively good rearing habitat,had increasing CPUEls beginning
in late July.A study conducted by Delaney and Wadman (1979)in the
Little Susitna River found emigration of emergent fry from natal areas
after the end of June.
Instream flow effects of the lower Susitna River upon juvenile coho
sa 1mon are 1imited to the backwater zone effects at tri buta ry mouths
because coho juveniles make little use of the side channel/slough sites.
Initi ally,backwater may decrease the amount of habitat sl i ght1y as
tri butary mouths change from free flowing to a backwater zone but then
WUA generally increases with mainstem stage as cover is inundated.
Also,the backwater can provide access into small tributaries and beaver
ponds where rearing and overwintering can occur.
66
....,
-
-
f'""'
I
Studies of coho salmon distribution in 1982 by hydraul ic zone showed
that coho generally preferred free-flowing tributaries over backwater
zones (ADF&G 1983).Cover in the free-flowing tributaries is often
better than in the backwater areas.For example,Birch Creek Slough
generally has poor cover while Bi rch Creek itsel f has abundant emergent
and aquatic vegetation in which coho were abundant.
4.3 Chum Salmon
The use of minnow trapping during 1981 and 1982 juvenile anadromous
studies makes comparisons of lower river catch and CPUE data with 1984
studies difficult because chum salmon are rarely captured in minnow
traps.The necessity for very early sampling,almost concurrent with
ice-out,becomes important when studying chum salmon juveniles.Their
early season movement and short time in the Susitna River system makes
detailed conclusions difficult.
The large catches of chum salmon fry in side channels in the lower river
contrast with the 1983 distribution data from the middle reach.Dugan
et al.(l984)indicated that chum fry CPUE's were greatest at tribu-
tari es and side sloughs.The 1983 catch rates,.however,refl ect the
prevalence of natal sloughs in the middle reach,while the lower reach
contains few natal side sloughs.Also,side channels in the middle
reach were not extensively sampled until July in 1983.
In 1984,chum salmon spawning was observed in several side channell
slough sites where none had been observed previously (Barrett et ale
1985)indicating that under certain conditions,lower river side chan-
nels do provide some suitable spawning habitat.Chum salmon fry
observed in some of the side channels may be rearing near their natal
areas.
The exact stimulus for the outmigration of chum salmon from the Susitna
River is not known,but probably reflects a combination of factors (Roth
et al.1984).Mainstem discharge was highly positively correlated with
chum salmonCPUE at the Talkeetna outmigrant traps in 1983.The sharp
decline in CPUE at the lower r5ver sites from early June (3+fpc)to
late June (1+fpc)in 1984 followed the peak June discharge on June 17
at Sunshine Station,and the mid-J~ne peak of chum outmigration past the
Talkeetna traps.
Since juvenile chum salmon outmigration is mostly completed by mid-July,
flow effects are limited to spring and early summer for this species.
Juvenile chum salmon used side channels heavily during this time while
use of the tributary mouths was limited.Apparently,chum salmon do not
move into the tributary mouths as they gradually move downstream and out
of the system.Most of the us.e of side channels for rearing occurs
before high turbidities occur.
Use of side channels by juvenile chum salmon is limited by depth and
velocity.The presence or lack of instream cover in side channels is
67
not important to juvenil e chum (Appendix A).Chum fry were captured
primarily in shallow sampling cells (S 1.0 ft)which had a relatively
low velocity and low to moderate cover.After breaching,side channel
WUA's may increase or decrease but the proportion of the area that is
suitable generally decreases as velocities and depths become unsuitably
large.Turbidities show sharp seasonal increases and some side channels
become turbid earlier in the season than others depending upon the
turbidity regimes in the Chulitna,Talkeetna,and Susitna rivers.
Since chum salmon side channel WUA's respond very similarly to those of
chinook salmon at individual sites,it appears that an analysis of
response to changes in mainstem discharge for chinook would also hold
for chum salmon.An analysis of flow regimes,would only need to take
place through mid-July for chum salmon,however,while chinook salmon
fry occur throughout the season in side channels.
4.4 Sockeye Sa.lmon
Tributary mouths were the primary capture sites for sockeye salmon in
the lower river.In the middle river,sockeye salmon were captured
primari ly at side sloughs (Dugan et a 1.1984)•.Si de sloughs were the
primary spawning areas for sockeye salmon in the middle river,and
tributary/l ake systems were the major sockeye spawni ng areas in the
lower reach (Barrett et al.1985).Relatively large catches of juvenile
sockeye in the middle river side sloughs were due to fish rearing in
their natal areas.
Few sockeye juveniles were captured in early June at modelled JAHS
sites.This low incidence was probably due to lack of natal habitat in
mainstem influenced areas of the lower river.Outmigrant trap catches
at Talkeetna (RM 103.0)and Flathorn (RM 22.4)indicate that sockeye fry
were redistributing in the system by the middle of June (Part 1 of this
report).The greatest catch per cell of juvenile sockeye occurred at
the modelled sites during late June.
The consistently low CPUE's in lower river side channels suggest these
areas are of l'imited value for juvenile sockeye rearing.Possibly these
juvenil e sockeye catches represent trans ient popul ations.Exceptions
include Beaver Dam Side Channel and other side channels located in the
Talkeetna River plume where lower turbidities allow juveniles to rear.
Since turbid glacial lakes are much less productive for sockeye salmon
than are clearwater lakes (Lloyd 1985),the productivity of these side
channels for sockeye is probably low in comparison to similar clearwater
streams.
The larger catches (21 to 101)of sockeye at tributary mouths indicate
that these sites are of some value for juvenile sockeye rearing.Beaver
Dam Slough had moderate numbers of sockeye present throughout much of
the season.Beaver Dam Slough resembled a lake system as it had low
velocities,large amounts of cover,and relatively warm temperatures
during the open-water season.CPUE1s of sockeye fry at Rolly Creek
mouth was low until early August.Emergent and aquatic vegetation were
profuse at this site during mid-season,making sampling difficult.
After late August,CPUE's of sockeye juveniles increased.Although high
68
"""'.
-
-
-
.-m,
numbers of these salmon fry were caught late in the season~we do not
know if they overwinter.
Instream flow effects upon sockeye salmon rearing occur at both tribu-
tary mouths and side channels.Occurrence of sockeye juveniles in side
channels appears to be limited by factors such as turbidity and velo-
city.Juvenile sockeye were captured more than half the times sampled
only in four side channel sites in the Talkeetna River plume.Even at
these four sites,the number of sockeye fry captured was less than 20 at
each,except at Beaver Dam Side Channel where 71 were captured.Typi-
cally,WUAs for sockeye increase after overtopping of the side channels
but then gradually decrease with further increases in discharge as side
channel velocities became unsuitable.Sometimes backwater areas may
form at the mouths of side channels (for example, Sucker Side Channel)
and modify this relationship somewhat so that WUA may rise with
increases in discharge for much longer periods.Generally,the
proportion of area that is usable within side channels decreases with
flow as velocities become less suitable.
At tributary mouths,the formation of backwater zones has a major effect
in increasing WUA for sockeye salmon juveniles.The response of the
increase in WUA for sockeye is similar to that of chinook salmon.
Access into suitable rearing and overwintering areas may also occur with
the increase in backwater or the amount of overtopping.For example,
access into potential rearing areas such as Whitsol Lake may be
inhibited if Kroto Slough is not overtopped.Also several other small
tributaries along the Kroto Slough side channel may be inaccessible if
flows are below those required for overtopping.
69
-
-
..-
5.0 CONTRIBUTORS
Resident and Juvenile Anadromous Fish
Project Leader
(Acting Project Leader,
Jan.to Jun~,1985)
Task Leader
Data Collection
IFIM Data,Models,and Text
Stage Data
IFIM Site Selection and Transect
Placement
Data Base Management
70
Dana Schmidt
Stephen Hale
Paul Suchanek
Rich Sundet
(Task Leader)
Stuart Pechek
Dave Sterritt
(Crew Leader)
John McDonell
Karl Kuntz
(Crew Leader)
Bob Ma rsha 11
Dale Corzine
Mike Domeier
Andy Hoffmann
(Task Leader)
Jim Anderson
Jeff Bigler
Fred Metzler
Tim Quane
(Task Leader)
Pat Morrow
(Crew Leader)
Isaac Quera1
Tommy Wi th row
Glenn Freeman
Doug Sonnerholm
Sharie Methvin
Chris Kent
Di ane Hi 11 i ard
E.Woody Tri hey
Allen S-i ngham
(Project Leader)
Chuck Miller
Alice Freeman
Kathrin Zosel
Ga il Hei nemann
Donna Buchholz
Drafting
Typing
Text
Report Coordinators and Editors
71
Carol R.Hepler
Roxanne Peterson
Skeers Word Processing
Paul Suchanek
Karl Kuntz
John McDonell
Stephen HaJe
Drew'Crawford -
-
-,
-
"""'
-
-
f""""
I
i
-
......
6.0 ACKNOWLEDGEMENTS
Funding for this study was provided by the State of Alaska,Alaska Power
Authority.
We thank the various consulting agencies working on the Susitna
Hydroelectric Project for helpful comments on a draft of this report.
72
-
-
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 Fi sh and Game.1981a.Juvenil e anadromous fi sh
study'on the Lower Susitna River (November 1980-0ctober 1981).
Phase 1 final draft report.Subtask 7.10.Alaska Department of
Fish and Game Susitna Hydro Aquatic Studies.Anchorage,Alaska.
·1981b.Aquatic habitat and instream flow project.Phase 1
---;:-final draft report.Volume 1·(December 1980-0ctober 1981).
Subtask 7.10.Alaska Department of Fish and Game Susitna Hydro
Aquatic Studies.Anchorage~Alaska.
•1983a.Resi dent and juvenile anadromous fi sh studi es on the
--=Susitna River below Devil Canyon,1982.Susitna Hydro aquatic
studies phase II basic data report.Volume 3 (1 of 2).Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies.
Anchorage,Alaska.
•1983b.Resident and juvenile anadromous fish studies on the
--""'Susitna River below Devil Canyon,1982.Susitna Hydro aquatic
studies phase II basic data report.Volume 3 (2 of 2:Appendices
A-H).Al aska Department of Fi sh and Game Susitna Hydro Aquati c
Studies.Anchorage,Alaska.
•1983c.Synopsis of the 1982 aquatic studies and analysis of
--'"lrfish and habitat relationships (2 of 2:Appendices A-K).Susitna
Hydro aquatic studies phase II report.Alaska Deparbnent of Fish
and Game Susitna Hydro Aquatic Studies.Anchorage,Alaska.
__........1984a.Susitna Hydro aquatic studies procedures manual (May
1983 -June 1984)(l of 2).Alaska Department of Fish and Game
Susitna Hydro Aquatic Studies.Anchorage,Alaska.
__."...1984b.Susitna Hydro aquatic studies procedures manual (May
1983 -June 1984)(2 of 2:Appendices).Alaska Department of Fish
and Game Susitna Hydro Aquatic Studies.Anchorage,Alaska.
•1985.(In Preparation).Susitna aquatic studies procedures
--=manual (May 1984 -June 1985).Alaska Department of Fish and Game,
Susitna Aquatic Studies.Anchorage,Alaska.
Ashton,W.S.,and S.A.Klinger-Kingsley.1985.Response of aquatic
habitat surface areas to mainstem discharge in the Yentna River
confluence to Talkeetna reach of the Susitna River.Draft report
prepared for Alaska Power Authority.Anchorage,Alaska.
Barrett,B.M.,F.M.Thompson,and S.N.Wick.1985.1984 salmon escape-
ment studies in the Susitna River drainage {Draft}Report No.4.
Alaska Department of Fish and Game Susitna Hydro Aquatic Studies.
Anchorage,Alaska.
73
1979.Little Susitna River juvenile
Alaska Department of Fish and Game.
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.
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.
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.
chinook and coho study.
Division of Sport Fish.
Dugan,L.J.,D.A.Sterritt,and M.E.Stratton.1984.The distribution
and relative abundance of juvenile salmon in the Susitna River
drainage above the Chulitna River confluence.Part 2 in D.C.
Schmidt,S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.)--.1984.
Resident and juvenile anadromous fish investigations (May -October
1983).Alaska Department of Fish &Game Susitna Hydro Aquatic
Studies.Report No.2.Anchorage,Alaska.
Klinger,S.,and E.W.Trihey.1984.Response of aquatic habitat
surface areas to mainstem discharge in the Talkeetna to Devil
Canyon reach of the Susitna River,Alaska.E.Woody Trihey &
Associates.Anchorage,Alaska.
Lloyd,D.S.1985.Turbidity in freshwater habitats of Alaska.A review
of pub 1 i shed and unpub 1i shed 1 i terature relevant to the use of
turbidity as a water quality standard.Alaska Department of Fish
and Game Habitat Division Report No.85-1.Juneau,Alaska.
Marshall,R.P.,P.M.Suchanek,and D.C.Schmidt.1984.Juvenile salmon
rearing habitat models.Part 4 in D.C.Schmidt,S.S.Hale,D.L.
Crawford,and P.M.Suchanek (eds:T.1984.Resident and juvenile
anadromous fish investigations (May -October 1983).Alaska
Department of Fish and Game Susitna·Hydro Aquatic Studies..Report
No.2.Anchorage,Alaska.
Milhous,R.T.,D.L.Wegner,and T.Waddle.1981.User's guide to the
physical habitat simulation system.United States Fish and Wild-
life Service.Biological Services Program FWS/OBS-81/43.
Mundie,J.H.1969.Ecological implications of the diet of juvenile
coho in streams.p.135-152.In T.G.Northcote (ed.),symposium
on salmon and trout in streams.H.R.MacMillan Lectures in
Fisheries,Univ.B.C.,Vancouver.
74
-
~.
-
-
"...
-
Noggle,C.C.1978.Behavioral,physiological and lethal effects of
suspended sediment on juvenile salmonids.Master's thesis,Univer-
sity of Washington,Seattle,Washington,USA.
Quane,T.,P.Morrow,1.Queral,T.Keklak,and I.Withrow.1985.
Technical memorandum in support of Task 14 (Lower River Resident
and Juvenile Anadromous Fish Studies).Alaska Department of Fish
and Game Susitna Aquatic Studies.Anchorage,Alaska.
Riis,J.C.,and N.V.Friese.1978.Preliminary environmental
assessment of hydroelectric development on the Susitna River.
Alaska Department of Fish and Game.Div.of Sport Fish and Comm.
Fish.,Anchorage,Alaska.
Roth,K.J.,D.C.Gray,and D.C.Schmidt.1984.The outmigration of
juvenile salmon from the Susitna River above the Chul itna River
confluence.Part 1 in D.C.Schmidt,S.S.Hale,D.L.Crawford and
P.M.Suchanek (eds.).-1984.Resident and juvenile anadromous fish
investigations (May -October 1983).Alaska Department of Fish and
Game Susitna Hydro Aquatic Studies.Report No.2.Anchorage,
Alaska..
Schmidt,D.C q S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.).
1984.Resident and juvenile anadromous fish investigations (May -
October 1983).Alaska Department of Fish and Game Susitna Hydro
Aquatic Studies.Report No.2.Anchorage,Alaska.
Sigler,J.W.,I.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.
Suchanek,P.M.,R.P.Marshall,S.S.Hale,and D.C.Schmidt.1984.
Juvenile salmon rearing suitability criteria.Part 3 in D.C.
Schmidt,S.S.Hale,D.L.Crawford and P.M.Suchanek (eds.)-.1984.
Resident and juvenile anadromous fish investigations (May -October
1983).Alaska Department of Fish and Game Susitna Hydro Aquatic
Studies.Report No.2.Prepared for the Alaska Power Authority.
Anchorage,Alaska.
75
-
-
APPENDIX A
LOWER SUSITNA RIVER
JUVENILE SALMON REARING SUITABILITY
CRITERIA
""""
-
-
INTRODUCTION
Habitat suitability criteria are necessary for evaluating fish habitat
using the instream flow incremental methodology (Bovee 1982).The
criteria express the value of a habitat variable such as velocity on a
zero (unusable)to one (optimum)basis for a given fish species and life
stage.The suitability criteria are coupled with the habitat present
within a study site to produce estimates of equivalent optimal habitat
called weighted usable area (WUA).
Juvenile salmon rearing suitability criteria have been used to model the
response of juvenile salmon habitat to variations in mainstem discharge
of the middle reach (Chulitna River confluence to Devil Canyon)of the
Susitna River (Hale et al.1984,Marshall et al.1984).The suitability
criteria used in these studies were developed specifically for the
middle Susitna River by Suchanek et al.(1984).EWT&A (1985)modified a
few of the same suitability criteria for use in impact analysis of
chinook salmon rearing in the middle Susitna River.
In 1984,some of the juven"ile salmon habitat modeling effort was direct-
ed toward evaluating responses of juvenile salmon habitat in the lower
Susitna River (below the Chulitna River confluence)to discharge varia-
tions.Since habitat data collection techniques used in 1984 were
similar to those used during the 1983 studies,suitabil ity criteria
specific to the lower reach can be developed.The purpose of this
appendix is to verify the applicability of the suitability criteria
developed in 1983 by Suchanek et al.(1984)for use in the lower river
habitat studies.The general philosophy was to use the 1983 middle
river criteria curves for the lower river unless the 1984 studies in the
lower river provided evidence for modifications.
METHODS
The field sampling methods used are detailed in Section 2.1 of this
report.These methods are very similar to those used during the 1983
studies (Suchanek et al.1984)and will only be summarized briefly here.
Sampling sites included:(1)20 habitat model sites which were normally
sampled twice a month and (2)31 opportunistic sites which were usually
sampled only once.
At each site,6 ft x 50 ft rectangular cells were sampled for fish and
then habitat variables were measured in each cell.Cells were selected
randomly at model sites,although sometimes additional selected cells
with "goo dll habitat were also sampled.At opportunistic sites,cells
were selected to encompass a variety of habitat conditions within
potentially usable habitat.Habitat measurements taken at each cell
sampled included a representative depth,mean column velocity,and
estimates of primary cover type and percent cover (Appendix Table A-I).
The data collected were examined for suitability criteria development by
using the procedures described in Suchanek et ale (1984),with a few
modifications.
A-I
Suitability was represented by mean catch per cell for chinook and coho
salmon and proportional presence (proportion of cells sampled in which
~I
8fi@I
Appendix Table A-I.Percent cover and cover type categories.
Group #%Cover Group #Cover Type
1 0-5%1 No object cover
2 6-25%2 Emergent vegetation
3 26-50%3 Aquatic vegetation
4 51-75%4 Debris or deadfall
5 76-96%5 Overhanging riparian vegetation
6 96-100%6 Undercut banks
7 Gravel (III to 311 di ameter)
8 Rubble (3 11 to 511 diameter)
9 Cobble (larger than 51l diameter)-
fish were captured)was used as the suitability measure for chum and
sockeye salmon.Data were pooled by species for analysis.Some data
were excluded from analysis by using results from the distribution and
abundance analysis (Section 3.2)which indicated factors other than the
microhabitat variables of velocity,depth,and cover were greatly
affecting distribution.Macrohabitat type and turbidity were two
factors which greatly affected distribution and were used as a basis for
excluding cells fished.Cells which were excluded from the analysis
varied by species and are detailed in the results section.The beach
seine and electrofishing data were pooled for analysis because these
sampling methods were both thought to be equally as effective given the
sampling conditions.Although sampling efficiency varies by gear type
and conditions fished,we assumed equal efficiency under all conditions
as analysis of sampling efficiency was beyond the scope of this study.
Groupings of habitat variables were identical to those used in 1983.
Percent object cover categori es 76-95%and 96-100%were pool ed because
of small sample sizes.Velocity and depth were pooled in groups identi-
cal to those used in 1983 with the exception that cells with depths of
0.1 feet were examined separately.In 1983,only two cells with a depth
of 0.1 feet were sampled,and therefore insufficient data were available
for examination of suitability of this depth.
Comparisons of the 1983 data with the 1984 data were made by plotting
the suitabi 1ity criteri a derived in 1983 on the same graph with com-
parable 1984 data.On the depth and velocity graphs this was done by
normalizing the suitability to 1.0 for the 1984 depth or velocity
increment with the highest suitability and then plotting the 1983
suitability criteria normalized to the same scale.The 1984 percent
cover data were first regressed against catch per cell or proportional
presence,and,if significant,the regression line was plotted and the
suitability normalized to 1.0 for the highest cover category.The 1984
percent cover suitability line was then plotted on the same graph,by
using the normalized 1.0 as the starting point.The suitability of
A-2
-
~l
-
,...,
-
"....
cover type for each species was calculated with the 1984 data using the
methods described in Suchanek et al.(1984).The suitabilities cal-
culated were then graphed against the cover type suitabilities calcu-
lated in 1983.
Variations in histogram distributions are to be expected on.a ,univariate
basis given that percent cover,cover type,velocity,and depth together
affect suitabilities of a cell.Therefore,composite weighting factors
were calculated for each cell using the 1983 suitability criteria and
.revi~ed 1984 criteria and then these weighting factors were compared
with catch.Composite weighting factors were calculated by multiplying
suitability indices for cover type,percent cover,and velocity togeth-
er.For chinook and coho salmon,Pearson correlation coefficients were
calculated between composite weighting factors and catch per cell
(transformed by natural log (X +1)].Chi-square association tests were
run between chum and sockeye proportional presence and composite weight-
ing factor value intervals calculated using the 1984 criteria data.
Intervals of composite weighting factors were specified by dividing the
data into four groups of approximately equal sizes by value of the
composite weighting factor.Pearson correlation coeffi ci ents and
results of the chi-square analysis were then compared with the same
analyses done in 1983.Most of the statistical tests and data manipu-
lations were done with the Statistical Package for the Social Sciences
(SPSS)(Nie et al.1975).
If the fit of the 1984 data to the 1983 suitability criteria did not
seem close upon visual inspection,the 1983 criteria were modified.One
of the procedures for modifi cation was as foll ows.If,for exampl e,the
1984 velocity distribution data appeared to match closely the 1983
velocity criteria,the 1983 velocity criteria were input as suitabil-
ities and averaged over each increment ofa variable such as depth for
which a modification of suitability was desired.These averages were
then multiplied by the mean catch of fish per cell divided by the mean
suitability.The actual mean catches per cell by depth increment were
then divided by the adjusted mean velocity suitability.If this ratio
was less than 1.0,this would indicate less use of a depth increment
than expected,given the average suitability for velocity.If the ratio
was greater than 1.0,the use would be more than expected by adjusting
for the effect of velocity.Sometimes this procedure..would be effective
in taking out variation caused by the other variable.If necessary,
this procedure was used to adjust for effects of two or more variables.
If the above procedure was not effective in discounting the extraneous
variation,then the criteria were modified using professional judgement.
Correlations or chi-square association tests were then calculated
between mean catch and calculated composite weighting factors using the
modified criteria.
RESULTS
Abundance and distribution data (Section 3.2)have shown that the number
juvenile chinook,coho,chum and sockeye salmon was very small at side
sloughs in the lower reach.Even sampling cells at sloughs with good
A-3
habitat failed to have any significant number of fish present in compar-
ison with similar cells at the other macrohabitat types (tributary
mouths and side channels).Fish were therefore responding to factors
other than the availability of suitable microhabitat in their use of
sloughs.For this reason,data collected at sloughs were eliminated
from suitability criteria analyses to avoid comparing similar cells with
large differences in mean catch.
Chinook Salmon
Chinook salmon suitability criteria were developed for both clear «30
NTU)and turbid (>30 NTU)water in 1983 because the catch in cells
without object cover was much greater in turbid water than in clear
water (Suchanek et ale 1984).Data collected in the lower river in 1984
have shown that turbidity may limit the distribution of chinook salmon
by being too high (Figure 10).Since cells with good habitat were
sampled when high turbidity was limiting use by chinook salmon fry,we
decided to eliminate sampled cells with turbidities greater than
350 NTU.
After-eliminating cells in side sloughs and cells with turbidities
greater than 350 NTU,1155 cells were available for analysis of chinook
distribution.Of the 1155 cells,400 were sampled in water with a
turbidity of 30 NTU or less.Mean adjusted catch (catch adjusted to a
standard cell size of 300 ft 2 )per cell of chinook fry in the 400 clear
water cells was 1.3,while mean adjusted catch per cell in the 755
turbid cells was 1.1.
A scatter plot of chinook salmon catch in cells without object cover
versus turbidities ranging to 100 NTU was examined.No notable inflec-
tions in catches of chinook salmon fry were noted over this range,
although gradual increases in catches occu rred across the range.It
seemed reasonable,therefore,to keep the same 30 NTU breakpoint between
high and low turbidity data for this year1s analysis.
Clear Water
Correlations among the values of habitat attributes and clear water
«30 NTU)chinook catch range to 0.32 in absolute value and a number of
the correlations are statistically significant (Appendix Table A-2).In
addition to these data,partial habitat data were recorded for four
additional clear water cells and these additional data are used in
subsequent analyses.
Composite weighting factors for all cells sampled were calculated by
using the 1983 suitabil ity criteria and al so with modification of the
velocity criteria as proposed by EWT&A (1985)and then correlated with
chinook catch transformed by natural log (x +1).In clear water,the
correlation in 1983 was 0.43 but the correlation with the 1984 data was
only 0.31 for the original criteria data and 0.26 with the change in
velocity criteria proposed by EWT&A (1985).It was therefore deemed
desirable to modify t~e criteria where large differences in individual
criteria were found.
A-4
-
--
~,
-
-
~
Appendix Table A-2.Kendall correlation coefficients between habitat
variables and chinook catch by cell (N=396)for all gear
,~types,in clear water.
f""""
Percent Cover
Cover Type Velocity Depth Chinook
Percent Cover 1.00
Cover Type 0.08*1.00
F'"'Velocity -0.32**0.04 1.00
.Depth 0.03 -0.08*-0.04 1.00
Chinook 0.07 0.09*-0.09*0.21**1.00
....*Significantly different from 0 at p <0.05 .
**Significantly different from 0 at p <0.01.
--
-A-5
Least squares regressions were run between chinook catch per cell and
the percent cover categori es in cl ear water.There was a si gnifi cant
positive regression which is very similar to the suitability line
developed in 1983 when the Y axis is normalized to a suitability of one
(Appendix Figure A-I).The 1983 suitability criteria was therefore
retained as a good estimate of this relationship.
The distribution of mean catch per cell of chinook fry by velocity
interval in clear water in 1984 shows that peak catches were made in
sampling cells with a velocity ranging from 0.1 to 0.3 fps (Appendix
Figure A-2).After normalizing this peak in catch to a sui·tability of-
1.0 and then plotting the 1983 suitability criteria on the same graph,
it appears that chi nook used lower velocity water in the lower reach
than in the middle reach under clear conditions.It was noted that the
1984 clear water distribution of catch by velocity interval was more
similar to the 1983 turbid water velocity suitability criteria and
therefore the 1983 turbid velocity criteria were plotted against the
1984 data (Appendix Figure A-3).Since the two distributions were
similar,the 1983 turbid water velocity criteria were taken as a good
estimate of the lower river velocity suitability for chinooks in clear
water.
Cover type suitabilities derived in 1984 for juvenile chinook in clear
water contrasted sharply with those derived in the middle reach in 1983
(Appendix Figure A-4).Debris was used less by chinook in the lower
reach for cover and emergent vegetation was used more.The sample size
of the cobble/boulder cover category was only one and therefore this
cover type could not be evaluated.Catches in the cells without object
cover were also relatively higher in 1984 than in 1983.
Therefore,it appeared that 1983 suitabi 1 ity for cover types woul d not
apply in the lower reach.By adjusting for the effects of velocity and
percent cover,better estimates of cover type suitability for the lower
river were formulated from the 1984 data (Appendix Figure A-5).Since
cobble and boulder sample sizes were low,suitabilities for these cover
types were kept proportional in suitability to large gravel as was the
case in 1984.Since the "no cover"catches were relatively large
because fish were using relatively deep cells without object cover (see
next paragraph),we lowered the suitability for no cover cells to 0.10,
the suitability found in 1983.
A heavy use of deep,clear water by chinooks was found in 1984 while in
1983 the data suggested a peak in use of cells 1.0 to 1.5 feet deep
(Appendix Figure A-6).In 1983,an evaluation of depth found it had
little effect on increasing the correlation of fish catch with composite
weighting factors using it.Depth was used in the 1983 modelling
efforts as having no value if less than 0.14 ft and having a suitability
of 1.0 if greater than 0.15 ft.In order to evaluate depth,sUitability
criteria were fit to the data using professional judgement after first
adjusting for mean velocity and percent cover suitabil ity (Appendix
Figure A-7).
After the modifications to the cover suitability and depth criteria were
made,we then correlated transformed chinook catch with the composite
A-6
-
-
-
--
-
18.9
10.0
I .:t STANDARD ERROR
N =NUMBER OF CELLS SAMPLED
-=1983
--=1984
~
0
0
Z 8.0-~
0
..f
..J
I.LI 6.00.....
:r:
0
~
c(
0
Z 4.0
c(
I.LI:e
2.0
1.0
X
I.LI0.8 0z
0.6 )-
~-..J
0.4 CD
c(
~
0.2 :;
(I)
0.0 --'-...L-.__....L-....L-__.l.--.l.--_---I"---I_~-..L-..L-~--'--"-0.0
N-168 N=125 N-61 N-34 N-12
(0-5%)(6-25%)(26-50%)(51-75".)(76-100%)
PERCENT COVER CATEGORIES
Appendix Figure A-I.Mean catch of juvenile chinook salmon per cell by
percent cover category (bars)in clear water of the
lower Susitna River,1984 and comparison of fitted
suitability indices (lines)calculated in 1984 and
for the middle Susitna River,1983.
A-7
-
e.2 Ha8S
:0::I!.STA~DARD ERROR0S.o0
Z N -NUMf!ER OF CELLS SAMPLED
J:---1983 1.0
u 4.0.X
..J 0.8 ~~..J Z1&1 3.0
U 0.6 >-....
J:I-
U 2.0
0.4 ~I-et CD
U et
Z 1.0 I-
<0.25
1&1 (/)
~H:a;'4
0.0 0.0
0.0 0.3 0.6 0.9 1.2 1.!5 1.8 2.1 3.4
VELOCITY (U/••cl
Appendix Figure A-2.Mean catch of juvenile chinook salmon per cell by
velocity intervals (bars)in clear water of the
lower Susitna River,1984 and fitted suitability
index (line)developed for the middle Susitna River,
1983.
7.0 I~STANDARD ERROR
:0::N-88
0 Ii.O N-NUMBER OF CELLS SAMPLED0
Z -aI9B3(TURBIO)
J:!l.0U
1.0
..J X..J 4.0
1&1 1&1
u 0.8 0....z
J:3.0
U 0.1i >-I-<I-
U 2.0 ..J
Z 0.4 ai
et <
lI.I 1.0 I-
~0.2 ::;
(/)
N s t4
0.0 0.0
0.0 0.3 0.6 0.9 1.2 1.!5 1.8 2.1 3.4
VELOCITY (ft/sec I
Appendix Figure A-3.Mean catch of juvenile chinook salmon per cell by
velocity intervals (bars)in clear water of the
lower Susitna River,1984 and fitted suitability
index (line)developed for turbid water in the
middle Susitna River,1983.
A-8
-
,.-
o
_CHINOOK,1983
VZ2I CHINOOK,I984
N =NUMBER OF CELLS
SAMPLED
<D
II
Z
x
~cz
0.25
1.00
0.75
>-
~-0.50
..J
OJ
<t
~
::>
(/)
,-0.00
tJ)~¥'0:1.&1 OZ I.&I..J (!)Z ~Z a:
a:jZ 1.&11.&1 ..J -0 (!)w zet ZO 1.&1
oet ..Jo m ~-o:~tLI->CD a:m m.J m et~(!)a:(!)!;i 01.&1 1.&1 03::>::>::>~eta:Zet a:~00.J(!)etQ.Q 00 0:01.&1 %-1.&11.&1 0Zomet(!)0:0:2!C>z::>1.&1 1.&1 ILl ILl>>>
0
COVER TYPE
Appendix Figure A-4.Comparison of cover type suitability indices for
juvenile chinook salmon in clear water calculated
from 1984 lower Susitna River distribution data and
1983 distribution data.
A-9
-....--N,.:..!.!...O ..-
-
-
-
-
-N=87
r--
-N=14~
N=I....--N=18-..-
N=71
r---
-N=62~
1.00
0.80
X
LIJ
Q 0.60z
,....-..J-OJ 0.40
oct...-~
U)
0.20
0.00
N=6
N=NUMBER OF.CELLS SAMPLED
N=31
-
-
en 1-:lI:::......0:11.1 UZ 1&.I..J ~z I-Z 0:
0::::)z lIJlI.I ..J -0 ~1&.1 z«zo 1&.1
In 0«..Jc m !i-0:>~o::1&.1->
1&.1 O:lD ID..J In :::)!i et«z«(!)I-0
C 1&.1 m:J :::)01-..JO:0:«0c000:c!>«a.11.11-z «11.1 x-:::lEu.0:::)om ~0:0:1&.1~Z11.1 11.1>>>
0
COVER TYPE
Appendix Figure A-5.Cover type suitability indices for juvenile
chinook salmon in clear water calculated
from 1984 lower Susitna River distribution
data after adjusting for velocity and percent
cover.
A-IO
--)j )-))1 --]<I 1
1.0
0.6 >-
I-
..J
0.4 iii
~
0.2 :;
C/)
X
LLI
0.8 az
4.5
N=36
N=55
I±STANDARD ERROR
N·NUM8ER OF CELLS SAMPLED
-=1983
O 0 i I -,I !!I\,I I 00. I I I I I I I I I I II .
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
DEPTH (ft)
1.0
12.7
9.0
~o 8.0oz:E 7.0
o
•..J 6.0
..J
LLIo 5.0
......
:ro 4.0
~o 3.0
Z<tLLI 2.0
~
;p
I..............
Appendix Figure A-G.Mean catch of juvenile chinook salmon per cell by depth intervals
(bars)in clear water of the lower Susitna River,1984 and
fitted suitability index (line)developed for the middle Susitna
River,1983.
N=36
0.6 >-
t-
...J
0.4 CD
<tt-
0.2 :5
CJ)
1.0
X
iJJ
0.8 0z
- - -....l\.r
N=!55
/'
/'
I
/
I
I
I
Ir--------f--F-J !0=,30
I N:127 N:144 I
I±STANDARD ERROR
N:NUMBER OF CELLS SAMPLED
--:1984
12.7
9.0
~
0o 8.0
z-::I:7.0u
I
...J 6.0
...J
W
U 5.0
.......
::I:
U 4.0
)::a I ~
I
1-'"U 3.0
N Z
<t
W 2.0
~
1.0
4.5
00"""-I I I I ~OO. I i I I I I I I «I I .
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
DEPTH (ft)
Appendix Figure A-7.Mean catch of juvenile chinook salmon per cell by depth intervals
(bars)in clear water of the lower Susitna River,1984.
Suitability index (line)fitted by hand usin9 professional
judgement.
1 J j i J )J ~J ]I j J I J 3 I
weighting factors calculated with the 1983 percent cover criteria and
turbid water velocity criteria along with the 1984 lower river cover
type and depth suitability criteria.The correlation was 0.61,
substantially higher than the original 1983 criteria.If depth was
eliminated from the calculations,the correlation dropped to 0.26 and if
primary cover type was dropped the correlation dropped to 0.52.There-
fore,it seemed reasonable to keep the newly modified cover type and
depth criteria as inputs.
Turbi d Water
Correlations between the values of habitat attributes and chinook catch
in turbid water range to 0.39 in absolute value and a number are statis-
tically significant (Appendix Table A-3).Partial habitat data were
recorded for 11 additional turbid cells and these additional data were
used in subsequent univariate histograms.
Correlations between composite weighting factors calculated with the
1983 turbid water criteria and 1984 chinook catch was 0.31,wh"ile
composite weighting factors calculated by incorporating the cover
modifications proposed by EWT&A (1985)were correlated with an r-va1ue
of 0.26.Comparable correlation with the 1983 data was 0.38.These
data again suggested that some modifications could be made,especially
given the changes already made °in the clear-water cover type suitabil-
ities.
A comparison of 1984 velocity distribution data and the 1983 velocity
suitability criteria for chinook salmon showed few differences (Appendix
Figure A-8),and therefore the 1983 velocity criteria were accepted as
the 1984 criteria curve.
Least squares regressions were run between chinook catch per cell and
the percent cover categories in turbid water.There was no significant
relationship between catch per cell and percent cover category and mean
catch per cell decreased with increases in cover (Appendix Figure A-9).
By adjusti ng for vel oci ty,a s1i ght trend upwa rd was noted over the
first three categories.The percent cover criteria developed in 1983
was therefore accepted as reasonable,as increases in the amount of
object cover would seem more desirable for fish and sample sizes were
very small in the 51-75%and 76-100%cover categories.
In 1983,cover type for chinook in turbid water was not evaluated.
EWT&A (1985)modified the turbid water criteria,however,so that they
more closely reflected the clear water criteria developed in 1983.In
1984,mean catches of chinooks in turbid water were highest in the
emergent vegetation,rubble,and debris-deadfall categories,but catc~es
were only slightly higher than in the cover category "no cover".
Cover type was evaluated in 1984 by using the method of EWT&A (1985)for
calculating turbidity factors from the fitted regressions of percent
cover in clear and turbid water and their associated chinook mean
catches.Turbidity factors were calculated (Appendix Table A-4)and
then applied to the revised lower river cover suitability data.These
A-13
""'",
~
Appendix Table A-3.Kendall correlation coefficients between habitat
variables and chinook catch by cell (N=744)for all gear
types~in turbi d wa ter.~
-
Percent Cover
Cover Type Velocity Depth Chinook ~
Percent Cover 1.00
Cover Type 0.39**1.00
Velocity 0.05*0.16**1.00 I!'!A':
Depth 0.06*0.26**0.21**1.00
..".,.
Chinook -0.02 0.00 -0.17**-0.15**1.00
*Significantly different from 0 at p <0.05.
**Significantly different from 0 at p <0.01.
""'"I,
-
-
A-14
f"""2.5
N-20T I:!:STANDARD ERROR
N·NUMBER OF CELLS SAMPLED
~,---1983 1.0
¥2.0
0
0
Z
~
Q
O.B
1.5 )(.J
.J IU
IU 0
Q Z,...,....0.6
:r 11-39 >-
Q 1.0 N'1I2 ~~
c(.J
Q 1II
Z NoS2 0.4 c(
r~c(~
IU 5
:IE 1II
0.5
0.2
11-97
0.0 0.0
0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 4.8
VELOCITY (ft/sec)
.-
Appendix Figure A-8.Mean catch of juvenile chinook salmon per cell by
velocity intervals (bars)in turbid waters of the
lower Susitna River,1984 and fitted suitability
index (line)developed for the middle Susitna
River,1983.
2.0
lIl:
0
0
Z U5::c
0,-..,
..J
..J
W 1.00
"-::c
0...
~
0 0.5
Z
~
lIJ
::I
0.0
;F
I~STANOARD ERROR
N-NUMBER OF CELLS SAMPLED
---1983
N-525 N'161 N-46
(0-5%)(6-25-1.){26-50%1 (51-7S%1 (76-10<W.)
PERCENT COVER CATEGORIES
1.0
Xw0.8 0z
0.6 ~
!::
0.4 ...J
ai
~
0.2 !::
::I
U)
0.0
Appendix Figure A-9.Mean catch of juvenile chinook salmon per cell by
percent cover category (bars)in turbid water of the
lower Susitna River,1984 and fitted suitability
index (line)calculated for the middle Susitna
River,1983.
A-IS
revised suitabilities were much too low for many categories given
observed catches and therefore a suitability of 0.15 was assigned as a
minimum for cover type suitability in turbid water based on observed
mean catches.Using this method,none of the suitabil ities for cover
type in conjunction with percent cover in turbid water are greater than
0.40 (Appendix Figure A-10).
Appendix Table A-4.Calculations of turbidity factors for 1984 lower
river data.
Number of Fish Per Cell (Fitted to a Line
Percent Turbidity
Cover Clear Turbid Factor
0-5%
6-25%
25-5m~
51-75%
76-100%
0.5
1.5
2.5
3.5
4.5
1.1
1.3
1.5
1.7
1.9
2.2
0.9
0.6
0.5
0.4
~I
In turbid water,peaks in chinook use were found in water less than 0.5
ft deep in both 1983 and 1984 (Appendix Figure A-ll).In 1983,since
fitting the depth suitability line to the data did not increase the
composite weighting factor much,the depth criteria used for clear water
(0 if less than 0.14 ft,1.0 if greater than 0.15 ft)was used for
modelling.
In 1983 there was only one turbid cell sampled with a depth of 0.1 feet
and therefore the value of cells with this depth could not be evaluated.
For purposes of IFIM modelling,this depth was assigned a suitability of
0,while in the RJHAB model data this depth did not occur.In turbid
water,21 cells of 0.1 feet depth were fished in 1984 and the mean catch
was 0.5 chinook juveniles per cell.These data suggest that under
turbid conditions the value of 0.1 feet cells is greater than o.A
suitability criteria line was fit to the 1984 turbid water depth data by
first adjusting for the effects of velocity (Appendix Figure A-12).The
optimum depth"ranged from 0.3 to 1.5 feet.
Once all the criteria were modified,correlations were calculated
between catch transformed by natural log (x +1)and the composite
weighting factor calculated by multiplying the suitabilities for ve-
locity,cover,and depth together.The correlation was 0.33,and if
depth were removed the correl ation dropped to 0.28.If cover was
removed from calculations of the composite weighting factor,the corre-
lation increased to 0.36.Since instream cover has value as a velocity
break in turbid water,it seemed reasonable to keep velocity,cover,and
depth in the modelling.
A-16
~\
-
....
--)(
llJ 0.4-0
Z
>--
~
..J
lD 0.2-
~
~
::I -en
1""".
PERCENT COVER
.1 0 -~
.2 6 -25
.3 26 -~O
.4 ~I -7~
.~76-100
PERCENT COVER BY COVER TYPE
Appendix Figure A-lO.Cover type suitability indices for juvenile
chinook salmon in turbid water developed from
1984 lower Susitna River chinook turbid water
distribution data.
1.75
¥o 1.50o
!:z:u 1.25
..J
~1.00
U....
:z:0.75
u
t-
<{
u 0.50
z
<{
~0.25
T I!STANOARD ERROR
N-NUMBER OF CELLS SAMPLED
---1983
1.0
)(
0.8 ~
Z
0.6 >-
l-
..J
0.4 iii
<{
t-
0.2 :;
en
4.3~:
0.00 +...J--r--r-~-"",-""".....,r--r-""'-,.-.....,rJ...-r-..-J\r-+..J..0.0
0.0 0.2 0.4 0,6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
DEPTH (ftl
Appendix Figure A-II.Mean catch of juvenile chinook salmon per cell by
depth intervals (bars)in turbid water of the
lower Susitna River,1984 and fitted suitability
index (line)developed for the middle Susitna
River,1983.
A-17
\.0
Xw0.8 0
z-
0.6 >-
~
....I
0.4 m
oct
~
0.2 :5
en
0.0
LED
-N=1I6 It STANDARD ERROR
N=252 -
-N=220 N-NUMBER OF CELLS SAMP
--"1984
,-
,-r--1-----_.1----~
\r-I
1\I-I l.-
e-I \-I \,-I \r·I \N=61 ,...
I \...I-
II -----Jy-....r-
j
N=76 r-- I
T t\I
J,.t..'If 1 ....
V
\IIII.0.00
I I I I I I I
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 4.3
DEPTH (ft)
1.75
....I....I 1.00
UJu
.....075X.
<.)
~ctU 0.50
z
ct
LLJ 0.25
~
~1.50
o
Z
:i:1.25
u
:too
I........
ex>
Appendix Figure A-12.Mean catch of juvenile chinook salmon per cell by depth intervals
(bars)in turbid water of the lower Susitna River,1984.
Suitability index (line)fitted by hand using professional
judgement.
»._I J J ,
••J I .t J 1 il J ,I -~
-
-
-
-
-
-
Coho Salmon
Juvenile coho salmon suitability criteria were developed only for clear
water in 1983.Very few coho were captured in macrohabitat types other
than tributary mouths in the lower reach and therefore only tributary
mouth data were used in suitab-llity criteria comparisons.Most of the
turbidities in the'tributary mouths were less than 30 NTU although on
two occasions,turbidities were over 100 NTU.
A total of 345 cells with complete habitat data were sampled in tribu-
tary mouths and another 2 cells with partial habitat data were sampled.
Mean adjusted catch in the cells sampled was 1.2 fpc.Kendall corre-
lations among the values of habitat attributes and coho catch ranged to
0.43 in absolute value (Appendix Table A":5).Cover type was most highly
correlated with coho catch.
The distribution of mean coho catch per cell by velocity interval in
1984 matched quite closely with the suitability criteria derived in 1983
for the middle river (Appendix Figure A-13).The 1983 velocity criteria
were therefore chosen as representative for the lower river.
A regression of coho catch to percent cover category was significant
(Appendix Figure A-14).When the 1983 and 1984 data were normalized to
1.0 on the V-axis for the 76-100%category,the 1983 SUitability line
had a much greater slope,and suitability for 0-5 percent cover in 1983
was 0.12,while in 1984 it was 0.33.After adjusting for the effect of
velocity,the distribution of catches by percent cover interval appeared
to be more similar to the 1983 distribution and since the sample size in
1983 was larger,the 1983 percent cover suitability relationship was
chosen for use in the lower river.
Initial calculations of the suitability of cover type for coho salmon
indicated that suitabilities in the lower river were similar to those
found in 1983 (Appendix Figure A-15).After adjusting for the effects
of velocity and percent cover,these estimates of cover suitability for
the cover types were revi sed for use in the lower ri ver in 1984
(Appendix Figure A-16).Since sample sizes for the three substrate
cover types were small,the suitability of 0.10 calculated in 1983 for
rubble and boulders was used for these three categories.
The distribution of CPUE 's for depth was very different from that found
in 1983 (Appendix Figure A-I?).By adjusting for the effects of ve-
locity,percent cover,and cover type there still was no trend in depth
suitabilities and therefore depth suitability was not changed from that
used in 1983.
The correlation between transformed coho catch and the composite weight-
ing factor calculated by multiplying the velocity,cover,and depth
suitabilities together was 0.32.
Sockeye Salmon
Juvenile sockeye salmon suitability criteria were developed by pooling
data over gear type and turbidity level in 1983.Since abundance and
distribution data have indicated that sockeye salmon use of lower river
A-19
Appendix Table A-5.Kendall·correlation coefficients between habitat
variables and coho catch by cell (N=345)in clear water.
A-20
-
-
-
-
N~35 T~STANDARD3.0 ERROR
J..
0 N-NUMBER OF CELLS SAMPLED::z::2.5 N=&7 -.-)9B30
0 1.0
...J 2.0 X...J
kJ O.B kJ
0 Q
"-1.5 N-9 :!!:-%"'0147 0,60 ~
~!:c:(1.0 ...J00.4 ai
Z c:(
~0.5 0.2 !:
::Iii:;:)
en
0.0 0.0
0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.2..-VELOCITY (tt/,.c)
Appendix Figure A-13.Mean catch of juvenile coho salmon per
ce 11 by vel oei ty i nterva 1s (bars)in the
lower Susitna River,1984 and fitted
suitability index (line)developed for
the middle Susitna River,1983.
3.5
o 3.0
Xo
o 2.5
I!STANDARD ERROR
No NUMBER OF CELLS SAMPLEO
--01983
--01984
1.0
)(
0.8 kJ
Q
Z
0.6
~
~
0.4 ...J
iii
0.2 ~
;:)
en
0.0
1
...J
~2.0
o
"-
%1.5o
~
C.:J 1.0
Z
c:(
~0.5
0.0 -l..~N~_';';13;-:;7""'''''''7.N~0'::'8;-1L..J-N::"6::':6;:-l-~N::-0--4:-:0::-L...l..:N::-o-:2~2...L....l-
(O·S'IIo1 (6-25'110)126-50'1101151-75'1101 (76-100'll01
PERCENT COVER CATEGORIES
-
Appendix Figure A-14.Mean catch of juvenile coho salmon per
ce 11 by percent cover category (bars)
in the lower Susitna River,1984 and
comparison of fitted suitability indices
(lines)calculated in 1984 and for the
middle Susitna River,1983.
A-21
~,
-,
-<D
II
1.00 z
iWll COHO,1983
~COHO,1984
N =NUMBER OF CELLS SAMPLED
~,
0.75 -
X
LLI
C
Z
>-....0.50 ~
-I ID-alcs:....-::>en
0.25
-
0
ID
"Z ~
0.00
en ~~.....0:::ILl oz ILI-J C)Z ~z a:
a:::::>z ILl ILl -J -0 C)ILI z«zO ILl0«-J c m ~-a:::>--ILl->m m-J «~C)a:::C)!;tl.LI a:::m m::::>m ::::>c:(««zc:(0 ~ILl ::::>at--Ja:::a:t-0c0000:::«ILl C)«4-WILlomx-0zC)a:::a::::lEC)::::>ILl ILl ILl ILl Z
>>>
0 ~;
COVER TYPE
Appendix Figure A-IS.Comparison of cover type suitability indices for
juvenile coho salmon calculated from 1984 lower
Susitna River distribution data.
-
-
A-22
,-
1.00
0.80
x
~LLI 0.60c
Z
~>-...
oJ
CD 0.40
1""'"t!
;:)
en-
0.20
0.00
lD
II
Z--
-
-
co
II-Z.---
It)-II-Z-
0-10
V
01 II
(\/Z
11 r----Z
r---
-
0 co =II II II-Z Z Z.----.---
0-In
II
Zn
!!!~~.....a::LU OZ LU~<.!)Z ~z a::
a::~z WlIJ ~-0 C)lIJ z<zo LU0<~o ID ~-0::>LU->ID O::ID CD~ID <I-«(!)a::ClI-oW~<z<a::<0 lIJ CD;:)~a~~a::0
0 00 a::<a-wI-z Om <lIJ (!)z-:::l!!E
LU 0
~(!)a::a::LU~zLULU>>>
0
COVER TYPE
Appendix Figure A-16.Cover type suitability indices for juvenile
coho salmon developed for the lower Susitna
River in 1984.
A-23
~205j I±STANDARD ERROR
N"NUMBER OF CELLS SAMPLED
o 2.0 N=131 -=1983
I
..J N=26..J X
~1.5 1.0 ~
......0.8 z:I:o 1.0 >-I-0.6 I-«-0 ..J0.4 -)::-I z 0.5 CDI«N «+::0 lLJ 0.2 I-
~0.0 :::>0.0 U)
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 4.5
DEPTH (ft)
Appendix Figure A-17.Mean catch of juvenile coho salmon per cell by depth
intervals (bars)in clear water of the lower Susitna River,
1984 and fitted suitability index (line)developed for the
middle Susitna River,1983.
J ,~J J J J l t J I ..~i J J ]J
side channel s is 1 imited by hi gh turbidities (Figure 18),cell s with
turbidities greater than 250 NTU were e1 iminated from suitabi1 ity
criteria development.
After cells with turbidities greater than 250 NTU were eliminated,922
cells with complete habitat data were available for analysis.Sockeye
were captured in 117 (12.7%)of these cells.Correlations among the
habitat variables ranged to 0.65 in absolute value and velocity was most
highly correlated with sockeye catch (Appendix Table A-6).In addition
to these cells,partial habitat data were collected at six additional
cells and these data are used in subsequent univariate histograms.
The distribution of proportional presence by velocity interval was very
similar to that found in 1983 (Appendix Figure A-18).There was no use
of velocities greater than 1.2 fps,however,and in 1983 there also was
no use of velocities greater than 1.2 fps although sample sizes were
smaller.Since these high velocities are not used,the lower river
velocity suitabi 1 ity criteria were modifi ed so that velocities greater
than 1.2 fps have 0 suitability (Appendix Figure A-18).
Distribution of proportional presence by percent cover categories was
similar to that found in 1983 (Appendix Figure A-19).The.1983 suita-
bility relationship was therefore selected for use in 1984.
The distribution of proportional presence by cover type categories was
somewhat different than that found in 1983 (Appendix Figure A-20).
Suitabi1 ities for the cover types used in the lower river in 1984 w"ill
be those developed in 1984 wi th the fo 11 owi ng two exceptions.Si nce
sample sizes were small (less than 25)for the cover type categories,
undercut banks and overhanging riparian vegetation,the suitabilities
calculated in 1983 were averaged with the 1984 suitabilities to give a
value intermediate between the two.
No trend was noted in the 1984 depth distribution data and therefore no
suitability criteria were fit to these data (AppendiX Figure A-21).Of
the 20 cells sampled with 0.1 ft depth,fish were sampled in 2 suggest-
i ng that thi s depth does have value.Therefore any depth wi 11 be
assumed to have a suitability of 1.
Composite weighting factor intervals calculated by multiplying cover and
velocity suitabilities together were associated with proportional
presence of sockeye salmon (AppendiX Table A-7).
Chum Salmon
Juvenile chum salmon suitability criteria were deve1qped by pooling data
over gear type and turbidity in 1983.Abundance and distribution data
indicate that chum salmon use of lower river side channels is limited by
high turbidities (Figure 15).Cells with turbidities greater than 200
NTU were eliminated from sUitability criteria development.Also,since
most chum salmon outmigrate before July 16,only data collected before
this date were retained for sUitability criteria analysis.
A-25
Appendix Table A-6.Kendall correlation coefficients between
variables and sockeye catch by cell {N=922}.
habitat
Percent Cover
Cover Type Velocity Depth
Percent Cover 1.00
Cover Type 0.30**1.00
Velocity -0.18**0.65**1.00
Depth 0.05*-0.01 0.07**1.00
Sockeye 0.04 -0.06*-0.21**0.02
*Significantly different from 0 at p<0.05.
**Significantly different from 0 at p<O.01.-
0.30 .,I:.STANDARD ERROR(ij
M -,...z
~~0.25 N •NUMBER OF CELLS SAMPLED
...I<n -"1983 \,0lULU
u~0.20 --·1984 (REVISED)X
LLl -lL.0.8 0OLU~>-~LUO.15 I'l >-_lll::2 0.6 ......U
0::0 ...I
~<n 0.10 en
0:1:0.4 <l
0::)-)-
Q.i 0.015 ~0.2 (I)""'"'
0.00 0.0
0.0 0.3 0.6 0.9 1.2 \,5 1.8 2.1 4.8
VELOCITY (ftlsee)
-Appendix Figure A-lB.Proportion of cells with juvenile sockeye salmon present by
velocity intervals (bars)in the lower Susitna River,1984
and fitted suitability index (line)developed for the middle
Susitna River,1983 and revised in 1984 for the lower Susitna
River using professional judgement.
A-26
]"J ]j ))))'J 1 1 1 )
1,00
...z
C1 SOCKE'I'E,1983
lZ2'SOCKEYE,1984
N •NUMBER OF CELLS
SAMPLED
....z
'''1 I :,"-::'
0.31:1.,I:t STANDARD ERROR
'.~~\;
)("t Oi...~:~.-
0 ~.z
0.30 N-NUMBER OF CELLS SAMPLED ~....z
(l)Z -'198~::0.150...JIIJ ---1984iii(I)0.25 ::::i
01lJ iii0::1.0 ~J!l",,11.o 0.20 XIIJ (I').!
Z>-0.8 ~~!
I 01lJ 0.215 ;~)::-j::lIC 0.11:1 z
I -
N 0::0 0.6 >-
-....J 0 0
11.(1)...o 0.10 :::io::Z 0.4 iiiQ.!:<l:..lli~0.01:1 ...0.00 .~
0.2 :5 II)1-'""'0:....OZ ........1 t!>Z 1-2 ca::
iii:=>Z ............I ~2 <lI ....zo(ZO '"00(....10 III o::~;i2 ",->(I)III <lit(...."'Ill 111....1 III =>le 0(0::Z4 0
0 UJ Ill=>=>01-....It!>4Q.0::1-0
0.00 I I I I I I I ,I I I ,0.0 0 00 0::UJUJ4",:z:iii:~N'1:I31 N-233 N-IOI N'41 N·22 z 0111 <!I 0::Et!>
(0-1:1%)(6-25%)(26-1:10%)(51-75%)(76-100%)
=>w III wUJ
>>>
0
PERCENT COVER CATEGORIES COVER TYPE
Appendix Figure A-19.Proportion of cells with juvenile sockeye salmon
present by percent cover category (bars)in the
lower Susitna River,1984 and comparison of fitted
suitability indices (lines)calculated in 1984 and
for the middle Susitna River,1983.
Appendix Figure A-20.Comparison of cover type suitability
indices for juvenile sockeye salmon
calculated from 1984 lower Susitna
River distribution data and 1983 middle
Susitna River distribution data.
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 4.5
DEPTH (ft)
I±STANDARD ERROR
N '"NUMBER OF CELLS SAMPLED
N=131 -=1983
):a
I
N
00
r-t/)Z 0.20
...I~
...It/)
~~lOXoIfO.I 5 N=307 N=302 N:90 r+'.~
~ZOLLJ0.8_
~~O.IO 0.6~
r-~_
0:g 0.4 I
0(/)0.05 m
~:J:0.2 ~Q:r-5
a..~0.00 0.0 t/)
0.0
Appendix Figure A-21.Proportion of cells with juvenile sockeye salmon present by
depth intervals (bars)in the lower Susitna River,1984 and
suitability index (line)developed for the middle Susitna
River,1983.
I )J j t j J )1 I )..~,)]})J
-Appendix Table A-7.Proportional presence of sockeye salmon associated
with the composite weighting factor calculated by
multiplying velocity and cover suitabilities to-
gether.
Composite Weighting Total Number Proporti on With
Factor Interval of Cell s Fish Present Chi-Square
o -0.06 244 0.02 X2 =55.3.....0.07 -0.11 213 0.08
0.12 -0.19 228 0.17 p<O.OOl
0.20 -1.00 241 0.23
:1f1Iii'i'II!A,
The number of cells available for analysis of juvenile chum distribution
totaled 249 after elimination of the cells outlined above.Chum salmon
were captured in 98 (39.4%)of these cells.Correlations among the
habitat variables and chum fry catch ranged to 0.32 in absolute value
(Appendix Table A-8).Partial habitat data were collected at two
additional cells.
.-
.....
-
-
The chum salmon distribution by velocity interval in 1984 was similar to
1983 (Appendix Figure A-22).Therefore,the suitability criteria for
chum salmon developed in 1983 was selected for use in 1984.
In 1983,the relationship of chum salmon use to percent cover and cover
type was the weakest of any of the four species.In 1984,the 0-5%
cover category and the "no cover"type had the highest proportional
presence within their respective distributions (Appendix Figures A-23
and A-24).These data indicate that chum salmon fry do not orient to
cover during rearing.Even when velocity suitability was adjusted for,
no real trends in percent cover and cover type utilization were noted,
although large gravel and rubble were used sl ightly more than was the
"no cover"type.Since there were no trends,cover type and percent
cover will not be used in the 1984 analysis of chum habitat use.
The distribution of chum proportional presence by depth intervals in
1984 was similar to that found in the 1983 studies (Appendix Figure
A-25).Since the distributions were similar,the criteria fit in 1983
was used to test for the value of depth in increasing the associations
with chum catch.Therefore velocity was first used alone and then with
depth to form categories which were associated with chum proportional
presence.
Although composite weighting factors calculated by velocity alone and
velocity and depth together were both significantly associated with chum
proportional presence,the composite weighting factor calculated by
depth and velocity together seemed to fit the observed distribution data
better (Appendix Table A-9).Therefore both velocity and depth suita-
bility criteria will be used to model chum salmon habitat.
A-29
Appendix Table A-B.Kendall correlation coefficients between habi tat
variables and chum catch by cell (NzZ49)for all gear
types,turbidity below 200 NTU.-
.....
Percent Cover
Cover Type Velocity Depth Chum
Percent Cover·1.00
Cover Type 0.13·...1.00
Velocity -0.25.....0.15**1.00
Depth -0.05 -0.03 0.07 1.00 ....,
Chum -0.20**-0.07 -0.04
-0.32*·1.00
..Si gnificantly different fl"Olll 0 at p<0.05•-Significantly different from 0 at p<O.OL
--
~,
-
I~STANDARD ERROR
N '"NUMBER OF CELLS SAMPLED
-·1983
N-or
IIz
1.0
1.0 X
ILl
0.8 ~
0.6 >-
l-
0.4 ;J
a::l
0.2 ~
:5
0.0 -I.+,~-r--,.....-.....,...r--"""T..L-""'="""I'"'L-_....,..~_~.T\.r-+-....L.0.0 C/)
0.0 0.3 0.6 0.9 1.2 I.e 1.8 2.1 4.8
VELOCITY (ftlsec)
Appendix Figure A-Z2.Proportien of cells with juvenile chum salmon present by
velocity intervals (bars)in the lower Susitna River,1984
and fitted suitability index (line)developed for the middle
Susitna River,1983.-
A-30 -
0.0 -I±.STANDARD ERROR-
(I)....0.5 -N =NUMBER OF CELLS SAMPLED r-\.0:lz -LlJLlJ I-
,.-0(1)0.4 -XLlJI-o.a LlJu..o::-~0OCL.-I-Z
z:t 0.3 -I-0.62::::>>-....%...
I-....
0::0 -
0%0.2 --0.4 ..J-~....CD-I-et0::-....CL.~0.1 -I-0.2 ::::>
f"""-en
I-
0.0 0.0
N-119 N"'74 N-37 N-12 N - 9
j.t~(0 -5%)(0-25%)(26-50%)(51-75%)(76-100%)
PERCENT COVER CATEGORIES
A~pendix Figure A-23.Proportion of cells with juvenile chum salmon
present by percent cover category (bars)in the
lower Susitna River,1984 and fitted suitability
index (line)calculated for the middle Susitna
River,1983.
-A-31
0.8
0.7
N=4 I~STANDARD ERROR
N =NUMBER OF CELLS SAMPLED
-
0.3
0.5
...zau
(/)aua:0.6
a..
::E
::;)
%:o
(/)
-I 0.4
-Iauo
lLo
zo...a:o
0..0.2oa::
0..
0.1
0.0
(J)
0::m
LLIc
N=29
N=53
COVER TYPE
N=44
0::'
LLI>8
oz
Appendix Figure A-24.Proportion of cells with juvenile chum
salmon present by cover type (bars)in
the lower Susitna River,1984.
A-32
~,
~,
-
..-
I±STANDARD ERROR
N'NUMBER OF CELLS SAMPLED
-.1983
1.0
1.0 x
w
0.8 ~
0.6 ~
0.4 :::!
!D
~
0.2 t
0.0 +Lr--,---.l.,..---r-.,J-r--.,.---.Jl.,---,--,Jc:::;:::±~::::::j....LO.0 ~
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 4.3
DEPTH (ft)
Appendix Figure A-25.Proportion of cells with juvenile chum salmon present by depth
intervals (bars)in the lower Susitna River,1984 and fitted
suitability index (line)developed for the middle Susitna River,
1983.
Appendix Table A-g.Proportional presence of chum salmon fry associated
with several composite weighting factors.
A-33
Summary
A summary table of reV1S10ns of the middle river suitability criteria
for use in the lower river reveals that about half the criteria were not
changed or changed only slightly (Appendix Table A-I0).The velocity
and percent cover relationships were often not changed while the depth
and cover type criteria have often been modified •.Point specific values
for all the sultability criteria developed for use in the lower river
are presented in Appendix Table A-II.
DISCUSSION
Chinook Salmon
The turbid water velocity criteria developed in 1983 were used for both
clear and turbid chinook distributions in the lower river in 1984.The
reason that there was no shift in velocity optima from cl ear to turbi d
water may be due to several factors.In the middle river,substrate is
·much larger and therefore,juvenile chinooks may find higher velocities
because suitable as there is always some substrate cover to hide under
or behind.In the lower river,however,very little substrate cover is
present and therefore chinook use lower velocity water much more.
In the lower river,cover suitabilities were often somewhat different
than in the middle river.Part of this difference may be due to the
actual cover in cover type categories being of a different type.For
instance,the aquatic vegetation in Caswell Creek,which harbored large
numbers of chinook fry,was not present in any of the sampled streams in
the middle river.Also the debris cover type in the lower river was
often much more silted in than in the middle river and therefore less
suitable.The primary cover type is associated with a variety of
secondary cover types and it is likely that,on the average,secondary
cover types associated with a primary cover type in the lower river are
different than the secondary cover types most common in the middle
river.If these secondary cover types are more suitable for fish,then
they might raise the suitability of the primary cover type.
Most notable in the analysis of chinook suitability criteria was the
effect of depth upon the distribution of chinook salmon.In the lower
river,chinook salmon found deep,water m~ch more suitable than in the
middle river (Appendix Figure A-7).This 1S probably due to the tribu-
taries in the lower river having a turbidity of approximately 10 to 20
NTU and therefore depth mi ght have a cover value in deeper waters.In
the middle river,much of the data were collected in Portage Creek,
Indian River,and other areas where the turbidity was usually less than
5 NTU and depth would not provide cover at depths which can be sampled.
Sometimes juvenile salmon thought to be chinook fry could be seen
feeding on the surface in tributary mouths such as Rolly Creek where
depths were greater than 5.0 ft.
In turbid water,on the other hand,depths greater than 1.5 ft were less
suitable than shallower cells (Appendix Figure A-Il).This trend was
A-34
-
-
-
-
-
-
-
-
"""
-
-
Appendix Table A-10.Summary of reV1Slons of 1983 middle river juvenile
salmon criteria for use in the lower Susitna River,
1984.
A-35
Appendix Table A-II.Suitability indices for juvenile salmon for velocity,depth,and cover in the lower
Susitna River,1984.
VELOCITY
Chinook Coho Sockeye Chum
Velocity Suita-Velocity Suita-Velocity Suita-Velocity Suita-
(ft/sec)bi 1ity (ft/sec)bi 1ity (ft/sec)bil ity (ft/sec)bil ity
0.00 0.42 0.00 0.29 0.00 1.00 0.00 0.86
0.05 1.00 0.05 1.00 0.05 1.00 0.05 1.00
0.35 1.00 0.35 1.00 0.20 0.71 0.35 1.00
0.50 0.80 0.50 0.88 0.50 0.48 0.50 0.87
0.80 0.38 0.80 0.55 0.80 0.35 0.80 0.70
1.10 0.25 1.10 0.32 1.10 0.14 1.10 0.56
):0 1.40 0.15 1.40 0.12 1.30 0.00 1.40 0.37,
w 1.70 0.07 1.70 0.04 1.70 0.150\
2.00 0.02 2.00 0.01 2.00 0.03
2.30 0.01 2.10 0.00 2.10 0.00
2.60 0.00
DEPTH
Chinook (turbid)Chinook (clear)Coho Sockeye Chum
Depth Suita-Depth Suita-Depth Suita-Depth Suita-Depth Suita-
(ft)bil ity (ft)bil ity (ft)bi 1ity (ft)bil ity (ft)bil ity
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.10 0.29 0.15 0.00 0.14 0.00 0.10 1.00 0.10 1.00
0.30 1.00 0.20 0.25 0.15 1.00 10.00 1.00 0.50 1.00
1.50 1.00 1.50 0.25 10.00 1.00 0.80 0.68
1.80 0.33 1.80 0.80 1.30 0.50
10.00 0.33 2.10 1.00 1.80 0.38
10.00 1.00 10.00 0.38
I !J J J J ,)J J )J J J )J I
J 1 I 1 ))].J »)))J ...1 ."j
Appendix Table A-II (Continued)
Percent Chinook Chinook
Cover Type Cover (turbid)(clear)Coho Sockeye Chum
No cover 0-5%0.15 0.01 0.00 0.18 1.00
Emergent Vegetation 0-5%0.23 0.11 0.05 0.39 1.00
6-25%0.30 0.33 0.14 0.54 1.00
26-50%0.33 0.55 0.24 0.70 1.00
51-75%0.39 0.78 0.33 0.85 1.00
76-100%0.40 1.00 0.42 ~.OO 1.00
Aquatic Vegetation 0-5%0.23 0.10 0.04 0.23 1.00
6-25%0.30 0.32 0.13 0.32 1.00
~26-50%0.33 0.53 0.21 0.41 1.00Iw51-75%0.39 0.76 0.30 0.50 1.00-...I
76-100%0.40 0.97 0.38 0.59 1.00
Debris or Deadfall 0-5%0.15 0.05 0.08 0.21 1.00
6-25%0.20 0.17 0.24 0.29 1.00
26-50%0.20 0.28 0.39 0.37 1.00
51-75%0.20 0.39 0.55 0.45 1.00
76-100%0.20 0.50 0.70 0.53 1.00
Overhanging Riparian 0-5%0.15 0.04 0.07 0.25 1.00
Vegetation 6-25%0.20 0.13 0.20 0.34 1.00
26-50%0.20 0.21 0.33 0.44 1.00
51-75%0.20 0.30 0.46 0.54 1.00
76-100%0.20 0.38 0.59 0.63 1.00
Undercut Banks 0-5%0.23 0.11 0.12 0.25 1.00
6-25%0.30 0.33 0.34 0.34 1.00
26-50%0.33 0.55 0.56 0.44 1.00
51-75%0.39 0.78 0.78 0.54 1.00
76-100%0.40 1.00 1.00 0.63 1.00
Appendix Table A-II (Continued)
Percent Chinook Chinook
Cover Type Cover (turbid)(clear)Coho Sockeye Chum
Large Gravel (1-3 11
)0-5%0.15 0.02 0.02 0.18 1.00
6-25%0.20 0.08 0.06 0.24 1.00
26-50%0.20 0.13 0.10 0.32 1.00
51-75%0.20 0.18 0.14 0.38 1.00
76-100%0.20 0.23 0.18 0.45 1.00
Rubble (3-5 11
)0-5%0.15 0.03 0.02 0.18 1.00
6-25%0.20 0.10 0.06 0.24 1.00
26-50%0.20 0.17 0.10 0.32 1.00
51-75%0.20 0.23 0.14 0.38 1.00>76-100%0.20 0.30 0.18 0.45 1.00,
wco
Cobble or Boulder 0-5%0.15 0.03 0.02 0.18 1.00
(>511 )6-25%0.20 0.11 0.06 0.24 1.00
26-50%0.20 0.18 0.10 0.32 1.00
51-75%0.20 0.25 0.14 0.38 1.00
76-100%0.20 0.32 0.18 0.45 1.00
-,~i )),J •J J J J ]]J )J I J
--
-
..-
also found in 1983 although discounted at the time.This difference may
be due to fish reacting to high suspended solid concentrations by
staying near the surface (Wallen 1951 as cited in Beauchamp et al.
1983).It also could be due to fish not being able to feed at depths
where there is very little light,whereas in shallower water a greater
amount of light may enable fish to feed.
Coho Salmon
The suitability criteria developed for coho salmon juveniles in the
middle river were modified only slightly in cover suitability for use in
the lower reach.The fit of the data to the composite weighting factor
was not very high (r=0.32)however,which suggests that coho respond to
other factors than those studied.These factors include food supply or
seasonal movements.
Sockeye Salmon
Since sockeye normally rear in lakes (Morrow 1980),it is not surprlslng
that velocity is one of the most important variables affecting their
distribution.In both the lower and middle Susitna river,no sockeye
were captured in cells with velocities greater than 1.2 ft/sec.The
highest catches of sockeye in the lower river were made at Beaver Dam
Slough,which is a backwater site with minimal velocity.
Instream cover also has an effect on juvenile sockeye salmon distri-
bution and it appears they use turbidity as cover (Section 3.2.4).In
lakes which are turbid due to glacial input,however,production of
sockeye smolts on an area basis is much smaller than that of clear lakes
(Lloyd 1985).Deep water in the clear lakes would provide cover while
in the Susitna,depths of 10 feet or more are infrequently found,and
therefore turbidity would be used as cover.Cover type suitabilities
were somewhat different in the lower reach than in the middle reach,
perhaps due to differences in the primary or secondary cover type within
the categories between the two reaches.
Chum Salmon
Chum salmon,in contrast to the other species,did not show any positive
response to the presence of cover.The response shown,which is a
negative one,is probably partly a function·of gear efficiency.They
did respond to velocity and depth,however.The lack of relationship
with cover may partly be a function of schooling behavior which reduces
the need for cover.It is also possible that since chum fry rear in
fresh water for only a short period,they usually are searching for food
instead of hiding in cover.
The reason for the heavi er use of sha 11 ower depths by chum juveni 1es
found in both years not known.It could be due to a use of shallow
depths and low velocities in side channels where some of the suspended
solids may settle out.Perhaps these areas also are somewhat warmer
than adjacent areas because the sunl ight strikes the substrate and is
absorbed heating the water above.
A-39
-
LITERATURE CITED
Beauchamp,D.A.,M.F.Sheperd,and G.B.Pauley.1983.Species pro-
files:Life histories and environmental requirements (Pacific
Northwest)chinook salmon.U.S.Department of the Interior,Fish
and Wildlife Service.FWS/OBS-83/1.
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.
E.Woody Trihey and Associates (EWT&A)and Woodward-Clyde Consultants.
1985.Instream flow relationships report.Vol.I.Working Draft.
Al aska Power Authority Susi tna Hydroel ectri c Project.Report for
Harza-Ebasco Susitna Joint Venture,Anchorage,AK.1 vol.
Hale,S.S.,P.M.Suchanek,and D.C.Schmidt.1984.Modelling of
juvenile salmon and resident fish habitat.Part 7 in D.C.Schmidt,
S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.Y:-1984.Resi-
dent and juvenil e anadromous fi sh investigations (May -October
1983).Alaska Department of Fish and Game Susitna Hydro Aquatic
Studies.Report No.2.Anchorage,Alaska.
Lloyd,D.S.1985.Turbidity in freshwater habitats of Alaska.A
review of published and unpublished literature relevant to the use
of turbidity as a water quality standard.Alaska Department of
Fish and Game,Habitat Division.Report No.85-1.l.luneau,Alaska.
Marshall,R.P.,P.M.Suchanek,and D.C.Schmidt.1984.Juvenile salmon
rearing habitat models.Part 4 in D.C.Schmidt,S.S.Hale,D.L.
Crawford,and P.M.Suchanek,(eds:T.1984.Resident and juvenile
anadromous fish investigations (May -October 1983).Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies.Report
No.Z.Anchorage,Alaska.
Morrow,J.E.1980.The freshwater fishes of Alaska.Alaska Northwest
Publishing Company,Anchorage,Alaska.
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.
Suchanek,P.M.,R.P.Marshall,S.S.Hale,and D.C.Schmidt.1984.
Juvenile salmon rearing suitability criteria.Part 3 in D.C.
Schmidt,S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.)-.1984.
Resident and juvenile anadromous fish investigations (May -October
1983).Al aska Department of Fi sh and Game Susi tna Hydro Aquati c
Studies.Report No.2.Anchorage,Alaska.
Wallen,I.E.1951.The direct effect of turbidity on fishes.Doctoral
dissertation.University of Michigan,Ann Arbor,Michigan,USA.
A-40
-
-
-
APPENDIX B
MODELLED SITE TURBIDITIES~JUVENILE
SALMON CATCHES~AREAS~SIDE CHANNEL FLOWS~
WEIGHTED USABLE AREAS~AND HABITAT INDICES
.-
,
-
This appendix is a compilation of data arranged into a number of graphs
and tables.The first three tables (Appendix Tables B-1, B-2,and B-3)
present:modelled side channel turbidities;modelled site catches and
CPUE1s of juvenile salmon;and lengths of RJHAB model sites;respec-
tively.Appendix Table B-4 presents modelled side channel flows as a
function of mainstem discharge at 3,000 cfs increments.
Next weighted usable areas and habitat indices are presented by species
in the following order:
Chinook Salmon
Tabulation of weighted usable areas and habitat indices for 18
sites (Appendix Table B-5).
Graphs of weighted usable area versus mainstem discharge for sites
not presented in Section 3.3:
Caswell Creek Mouth (Appendix Figure B-1)
Beaver Dam Slough (Appendix Figure B-1)
Hooligan Side Channel (Appendix Figure B-2)
Bearbait Side Channel (Appendix Figure B-2)
Last Chance Side Channel (Appendix Figure B-3)
Rustic Wilderness Side Channel (Appendix Figure 8-3)
Island Side Channel (Appendix Fi gure 8-4)
Mainstem West Bank (Appendix Fi gure B-4)
Goose 2 Side Channel (Appendix Figure 8-5)
Circular Side Channel (Appendix Figure 8-5)
Sauna Side Channel (Appendix Figure 8-6)
Bearbait Side Channel (Appendix Figure 8-6)
Sunset Side Channel (Appendix Figure B-7)
Sunrise Side Channel (Appendix Figure B-7)
Trapper Creek Si de Channel (Appendix Fi gure B-8)
Coho Salmon
Tabulation of weighted usable areas and habitat indices for three
sites (Appendix Table 8-6).
B-1
Chum Salmon
Tabulation of weighted usable areas and habitat indices for 15
sites (Appendix Table 8-7).
Graphs of weighted usable area versus mainstem discharge for sites
not presented in Section 3.3:.
Hooligan Side Channel
Kroto Slough Head
Bearbait Side Channel
Island Side Channel
Mainstem West Bank
Goose 2 Side Channel
Circular Side Channel
Sauna Side Channel
Sucker Side Channel
Beaver Dam Side Channel
Sunrise Side Channel
(Appendix Figure 8-9)
(Appendix Figure 8-9)
(Appendix Figure 8-10)
(Appendix Figure 8-10)
(Appendix Figure 8-11)
(Appendix Figure 8-11)
(Appendix Figure 8-12)
(Appendix Figure 8-12)
(Appendix Figure 8-13)
(Appendix Figure 8-13)
(Appendix Figure 8-14)
Sockeye Salmon
Tabulation of weighted usable areas and habitat indices for seven
sites (Appendix Table B-8).
Graphs of weighted usable area versus mainstem discharge for sites
not presented in Section 3.3:-Caswell Creek Mouth
Beaver Dam Slough
Sunrise Side Channel
8-2
(Appendix Figure B-15)
(Appendix Figure 8-15)
(Appendix Figure B-16)
J })1 -)-----)1 1 )J )j }
Appendix Table B-1.Turbidities within modelled side channels of the lower Susitna River,June through August,1984.Values within
parentheses were calculated by inputting the overall mean for all the side channels during a given two week period.
Site June 1-15 June 16-30 July 1-15 July 16-30 Aug 1-15 Aug 16-301 Mean
West Bank Lateral Side Channels
Kroto Side Channel
Bear Bait Side Channel
Mainstem West Bank
Sauna Side Channel
Trapper Side Channel
Middle Side Channels
Hooli9an Side Channel
Last Chance Side Channel
Island Side Channel
Circular Side Channel
Sucker Side Channel
Sunrise Side Channel
(64)
(64)
(64)
120
96
(64)
(64)
55
89
26
18
394
392
(227)
(227)
576
365
(227)
126
122
64
112
(369)
284
(369)
496
940
288
296
334
592
276
180
272,704
312
368
364
470
296
672
336
288
118
88
784
328
324
244
306
704
352
228
216
292
280
126
142
324
156,256
608
544
576
(209)
78,304
44,163
44,124
388
254
279
266
499
377
365
215
241
140
121
CD
I
W East Bank Lateral Side Channels
Rustic Wilderness
Side Channel
Coose Side Channel
Sunset Side Channel
Beaver Dam Side Channel
OVERALL MEAN
(64)
41
(64)
(64)
64
120
140
(227)
90
227
130
384
(369)
224
369
160
300
114
134
312
196
188
100
170
314
38
64,244
41,146
150
209
118
194
152
139
1 Two turbidities are given in this column for six sites because there were two sampling trips during this two week period in the
Sunshine area.Turbidities were dropping rapidly in late August and so turbidities taken on the first late August trip were much
higher than those taken during the second trip in late August.
Appendix Table B-2.Catch and catch per cell (CPUE)of juvenile salmon within lower Susitna River
sampling sites,1984.Cells have been standardized to an area of 300 ft 2.
No.f.1t-
cells Chinook Coho Chum boc:keye Chinook Coho Chum Soc:I,:eyl:
S!te sampled catch C:<i\tch catch catch CPUE CPUE CPUE Cf"UE
---------------------------------~--_.__...-.-n.__......____._
-----.....-q .....
.......~_....._..~_....,---_.-..-._---,~......,.--~.q •••----------_.
Hooligan Side Channel '77 21 0 lEI 3 0.27 0.00 1.01 0.04
Eagles Nest Side Channel 30 5 (I 0 0 0.17 0.00 0.00 0.00
I<roto Slaugh Head 56.5 4 0 1 2 0.07 0.00 0 ..02 0.04
Rolly Creek Mouth 91 53 ,39 2 87 0.58 0.4:5 0.02 0.96
Bearbait Side Channel 49.4 4 0 3 0 0.08 0.00 0.06 0.00
Last Chance Side Channel 50 0 0 1 0 0.00 0.00 0.02 0.00
Rustic Wilderness Side Channel 65 55 1 11 0 0.85 0.02 0.17 O.OQ
Caswell Creel,:MCluth '74 419 245 (l 21 5.66 s.:'~l 0.00 0.28
OJ I~land Side Channel 82 39 1 74 ....0.4[1 0.01 0.90 0.02I"-
+:00 Mainstem West Bank 45 7 (I I)1 0.16 I).(10 0.00 0.02
Goose 2 Side Channel 82 l4 1 30 "0.90 0.01 0.37 0.02"-
Circular Side Channel 88 2B (I 114 6 0.32 0.00 1.30 0.07
Sauna Side Channel 44 3 (I 41 5 0.07 0.00 0.9,3 O.11
Sucker Side Channel 77.1 23 0 112 15 0.30 0.00 1.45 0.19
Beaver Dam Slough 83 14 67 0 101 0.17 0.81 0.00 1.22
Beaver Dam Side Channel 102 153 9 23 71 1.50 0.09 0 ..23 0.70
Sunset Side Channel 73.5 121 0 0 1 ','1.65 0.00 0.00 0.16
Sunrise Side Channel 73 120 1 43 8 1.64 0.01 0.59 0.11
Birch Creek Slaugh 'J6 2:::;'71 45 29 0.24 O.'74 0.47 0.30
Trapper Creek Side Channel 96 43 2 20 4 0.45 0.02 0.21 0.04
SUBTOTAL 1434.5 1209 437 598 369 0.84 0.30 0,42 0.26
Opportunistic sites 163.7 249 5 10 43 1 "')0.03 0.06 0.26•i.I..:.
TOTAL 1598.2 1458 442 608 412 0.91 0.28 0.38 0.26
],~t J ~..J J J .~t )I I "t I,)cJ J
.-
Appendix Table B-3.Lengths of RJHAB model sites in the lower Susitna
River,1984 •
.-
.....
-
-
Site
Hooligan Side Channel
Eagle's Nest Side Channel
Kroto Slough Head
Rolly Creek Mouth
Bearbait Side Channel
Last Chance Side Channel
Rustic Wilderness Side Channel
Caswell Creek Mouth
Island Side Channel
Goose 2 Side Channel
Sucker Side Channel
Beaver Dam Slough
Beaver Dam Side Channel
Sunrise Side Channel
Birch Creek Slough
Trapper Creek Side Channel
B-5
Length (feet)
1377
490
748
1437
496
961
1169
712
769
1030
658
436
608
1003
841
968
Appendix Table 8-4.Side channel flows at the 15 modelled side channels in the lower Susitna River as a
function of mainstem discharge,1984.Flows calculated from rating curves presented
in Ouane et al.(1985).
HOOLIGAN S.C.KRDTD SLOUGH HEAD BEARBAIT SIDE CHANNEL LAST CHANCE S.C.RUSTIC WILDERNESS S,C.
-~~~--~~--~~-~----------------------------------------~----------
lUi I tiSTEN SITE SITE SITE SITE SITE
DISCHARGE AREA FLOW AREA FLOW AREA FLOW AREA FLOW hREA FLOW
12000 63400 (I 48200 0 3100 0 17500 0 48(1\J (I
15000 63400 0 4B200 0 3100 (I 175(/(1 0 480i)0
18000 65400 0 48200 0 3100 0 17500 0 4800 0
21000 63400 0 48200 C,3100 (;17500 0 31900 54
24000 79800 50 48200 0 3100 0 20000 1 49500 It!
27000 86900 72 48200 0 3100 (I 22000 t 60700 10~,~
30000 908(1)100 48200 0 3100 0 2700iJ 5 6S1700 1)4
330(iO 9&500 135 48200 (I 2,100 0 34000 B 711800 171
36000 104BOO 178 50000 {S a 5700 .33 46500 13 83s00 2n
39000 113700 22'1'61900 74 10800 48 70000 21 89900 261
42000 122900 2BB 77500 98 14600 67 81000 .31 9700(1 7i~:.hJ
45000 1:31300 358 86800 128 I79(h)93 91000 46 104000 375
I 48000 141200 439 95100 163 21100 ('1'"94000 67 1(6000 442OJLJ
I 51000 152000 531 102200 206 23BOG 166 9630u 95 1140(10 5160'1
5401;0 163000 636 106700 25~t 26400 217 98500 131 j I 74ll:.1 59c
57000 174100 '"'C'-,110200 ~14 29000 2i9 100200 li8 t1920()684(oJ;:'
60000 186800 B85 113500 381 31500 ~.54 101800 2.38 12li i'OO 'J7'i'
6301)0 200800 1032 116600 459 339(1(;44:1 103200 m 121700 b
660(;(i 213.300 lb'4 1190lj(l 547 .)6300 552 104400 408 12~'20(:h
69(;00 22bOOO 1373 120100 648 38})(1 b 105560 ~;:d:!1127(JO b
72(ii)t)239000 1570 121000 761 40iJOO b 106300 609 12~;OOO t
7~;£j0(;25090(:1785 12WH)889 4150('t;1070liO 844 1235(il~j b
a =Flow estimated
b =Ratino curve not available
c =IFIM model rated unacceptable at this site flow
d =Modelled at flow of 6 cfs for IFIM
e =Modelled at flow of 5 cfs for IFIM
f =These flows are approximate because they are heavily influenced by Cache Creek flow
»}J i .1 ...1
••I )t J J J J ._J I •j
l 1 -))J i -})-)))))J
Appendix Table B-4.Continued.
iSLAND SIDE CHANNEL MAINSTEH WEST BANK GOOSE 2 SIDE CHANNEL CIRCULAR SlDE CHANNEL SAUNA SIDE CHM~i~EL
------------------------------------~-----------------------------------,--------,--
MA1NSTEH SITE SITE SITE SiTE SITE
DISCHARGE AREA FLOW AREA FLOW AREA FLOW AREA FLOil AREA FLD~j
12000 3150(l <1 d 6160.3 {1 d (I i)59464 <.1 d 4209.5 <l e
15000 315(10 <i d 61603 ·0 d 0 (:59464 <1 d qLi)9:3 <l e
18000 31500 <1 d b16!)}{1 d I)0 59464 ::1 d 42093 <1 e
21000 3150(i d d 73426 19 13 (l 59464 <1 d 42093 q e
24000 .31500 <1 d B0904 1:,(I 0 594t:,4 <1 d 42093 <I e,J,_I
27000 31500 {1 d 93353 134 0 (I 59464 -<1 d 420S'3 <1 e
.30000 31500 <1 d 108613 .307 9600 'I:59464 <1 d 42093 <i E\...!a
TSi)()(1 31500 <'1 d 1147SB 470 21500 24 59464 <1 d 42(i'{3 <.1 e
36(100 392(1)6'1 117696 559 34300 {'~71590 27 -4209.3 (j e'-'"
59(iOO 45300 94 .120505 657 47800 41 7653'1 38 49127 21 c
42000 51000 126 1;J~-'"'~762 61400 52 80557 54 497:,8 25 [LC)J1i
4500 11 58500 166 129211 &74 720(j(f 65 85140 73 50289 2~
48000 65500 215 1~~T 995 81400 81 92944 DC'50889 34OJI·:i~'u '1 '"I 51000 720(11)2n i36885 1123 c 87800 98 102530 129 51451 39-....a
54000 79400 342 140761 1260 [93200 liB 113323 j67 52011 H
5700(1 86700 424 144269 1404 [97100 141 125753 rjf1'S2b]1;50":'.11,}
D(,O(H)93100 520 147B99 1555 c 9S'900 166 1:34218 Lo8 532q4 56
63000 9980(1 /;31 151842 1]1:;t 102000 195 143575 334-5427:,'10...·
66(iOO 106200 758 154205 1882 [i03200 226 150869 412 [55184 ;0 c
c,90UfJ 1i 1900 904 156425 2{i57 [11)420(1 261 154657 502.c 56053 '7:i c
721)01:'118200 1070 c 158522 2241 [j(l4800 300 i57074 bl0 [57142 85 -[
?51)O(J 123300 125&c 160818 24:)1 [10,51 i)(1 )42 15'1'211 733 c 61018 O'l r
/,_11,..
a =Flow estimated
b =Ratino curve not available
c =IFIM model rated unacceptable at this site flow
d =Modelled at flow of 6 cfs for lFIH
e =Modelled at flow of 5 cfs for IFIM
f =These flows are approximate because they are heavily influenced by Cache Creek flow
Appendix Table 8-4.Continued.
SUCKER 5IDE CHANNEL BEAVER DAN SIDE CHANNEL SUNSET SIDE CHANNEL SUNRISE SIDE CHANNEL fRAPPER CREEK j,L.
-------------------------------------------------------~-------_.~----------------
l1ii WS lEN 8m SITE SITE SITE SITE
DISCHARGE AREH FLOW AREA FLOW AREA FLOW AREA FLOW AREil FLDfi
12000 (I t)18900 d 49562 j e (I 0 73,300 C',-
!T
151)00 I)0 1890(1 U 49562 1 e ,.0 72;30(:E f'.'
180i,/O I)0 189ilO q 49562 1 e (1 I)73300 14 f
21\;0\}(i (i 18900 <l 49562 1 e (i (,733()(i 16 f
2400i)(;0 18900 <1 49562 1 e (I (;73300 J8 f
270(i(,(J 0 18900 {1 49562 1 E'i}(I 73300 20 t
30000 8500 13 16900 <1 49562 1 e 0 0 733iJO 12 f
53000 14900 18 18900 (j 78488 47 (,0 73300 24 f
3",000 [6900 24 18'100 (1 89472 68 19000 [9 73300 26 f
3S'i)(i1)1940(;31 1890(1 (j 117943 96 53900 29 73300 28 f
42(il}O 23600 39 18900 <1 106320 132 78500 41 73300 30 f
4:;(i(;(1 29600 48 1890(J \1 122:3,)8 17B 9710(1 58 7760(,39
OJ I 480(Ji)37100 59 22400 7 j 35476 235 115400 79 91200 72
I SlUO!)46600 71 280{JCl 14924B 305 106co\,131100 1081iY,129d
54(1)(1 57900 86 32600 lB 165990 3'10 14tl900 139 12:5)1)0 221
S/UDO 6690t)101 .357{Ki 2~;173483 49:16('600 18i 137100 370
60000 ·71300 119 38i)00 45 188419 614 175600 233 15120(i 564
63000 7390(,139 39600 68 194419 7;::j 192(1(11)29~i 158000 683_oJ}
66(1uO 75900 161 40800 101 20:\000 925 207300 370 163100 81 \'
b9(n)t)7730t)185 41500 148 206972 111'!c 2214(i!j 457\1669,)0 ~17~i c
/~Ol):)7GlOO <:11 41900 213 210m 1345 c 2'29000 5;:,4 1hHOO 1151 c
75000 18300 24i)4210(,302 21586l i 6(i3 r.233300 688 173500 i 351 c
a =Flow estimated
b =Ratina curve not available
c =IFIM model rated unacceptable at this site flow
d =Modelled at flow of 6 cfs for IFIM
e =Modelled at flow of 5 cfs for IFIH
f =These flows are approximate because they are heavily influenced by Cache Creek flow
~I c..1 J )1 D I J J 1 ]}J '.,1 J J
::c
-
-
Appendix Table 8-5.Weighted usable areas and habitat indices for juvenile
chinook salmon in lower Susitna River model sites,1984.
RQl.LY CREEK IHlIlTH .CASlIEll CREEK IIOIITII BEAVER DA~SLOUGH-----------_..-----_._---------------------_._-----------,..--------.------------------------------...------
IlAI1iSTElI mE CHINOOK CIlIIlOOK "AINS1£"SITE CHINOOK CIII.NIIDK "AINSTElI SHE CHINlJOK CI!!lI00K
DISCHARGE AIlEA MilA Il.I.DISCHARGE ARU IlIIA II.I.DISCHARGE AREA wu~H.I.
12000 ..84'100 3900 0.05 12000 16200 800 0.05 12000 llbOO 1300 v.ll
15000 '~900 3900 0.05 15000 16200 800 0.'05 15000 llbOO 1300 0»11
18000 84900 3900 0.05 18000 10200 800 o.D5 IBOoo llWO 1300 O.ll
21000 B~'I00 3900 0.05 21000 110200 800 0.05 21000 moo 1300 O.ll
2~OOO 85300 3900 0.05 2~OOO 16200 800 0.05 HOOO 11900 1300 v.lI
21000 8lI300 3900 O.O~27000 16300 aoo 001)5 27000 12200 13M 0.11
30000 93200 3900 0.04 30000 110700 l100 0,07 30000 125\)0 1300 0.10
33000 99800 4100 0.04 33000 11300 11000 0.09 33000 13000 1300 0.10
36000 108900 4200 0.04 36000 18000 .2200 0.12 310000 13400 .1300 0.10
39000 121000 4300 0.04 39000 18900 ..2100 0.14 39000 13900 1400 0.10
42000 135000 4400 G003 42000 1'1800 3200 0.16 42000 14400 1500 0.10
45000 152/000 4500 0.03 45000 21000 1700 0.19 45000 15000 1BOD 0.12
48000 178500 7300 0.04 48000-2UIGO 4200 0.19-48000 15700 2100 0.13
51000 1'18110O WOO 0.07 51000 -mot 4700 0.21 51000 111JOO 2/000 0.16
54000 213000 20100 0.09 54000'moo.sm·0.22 54000 10800 3000 0.18·
57000 223200 23400 ll.10 57000 2_5700 0.23 57000 11600 1700 0.21
60000 229800 2S900 0.11 DOOOO 25500 6200 0.24 60000 19500 ·4200 0.23
63000 23500&28000 0.12:63000 2~6700 0.25 03000 19700 4600 0.23
66000 238700 30000 -0.13 D6000 27200 7200 0.26 610000 ~.4BOO 0.23
69000 241bOO~31500 0.13 69000 27900 7600 0.27 69000 21600 5ODO 0.23
.12000 243200 32900 0.13 ..12000 .28900 9000 0.28 72000 22100 5100 0.23
75000 243600 33500 0.14 75000 29700 8400 0.28 75000 22600 5200 0.23
HDOUGAII SIDE CHt1NIIEl ---.KRDTD S1.1lU6H HEAl IlEARBAIT SIDE CIlAIIlEi.-_..._----_._------------------------------------......-------
IIAINSTElI sm CHIlIIIOK CHIIIlIOt:lIAINSTEJr SITE CHIIlIllIl CHIIIIOK ItAI~STElI SITE CHI!IIOK CHINIlOl
DISCHARGE AIlEA MIlA H.I.DIS£llAli6E .AREA 1IIl~K.I.DISCHARGE AREA IIUA H.I.
12000 63400 500 0.01 12000 49200 100 .00 12000 3100 20 0.01
15000 63400 500 0.01 15000 4820t 100 .00 15000 3100 20 0.01
18000 bJ400 500 0.01 18000 48200 100 .00 18000 3100 20 0.01
21000 .bJ400 500 0.~1 21000 48200 100 .00 21000 3100 20 0.01
24000 79800 7600 -0.10 24000 48200 100 .00 2+000 3100 20 GoOl
21000 86900 7200 0.09 27000 48200 100 ,00 27000 3100 20 0.01
30000·90800 6700 0.07 30000 48200 100 .00 30000 3100 20 0.01
33000 moo 6100 0.06 33000 48200 100 .00 33000 3100 20 0.01
36000 104800 5500 0.05 3hOOO 50000 2000 o.e4 30000 5700 200 0.04
39000 113700 4900 0.04 39000 67900 4800 0.07 39000 10800 350 0.03
42000 122900 4200 0.03 42000 77500 6200 0.09 42000 14000 530 0.04
45000 131300 3600 0.03 45000 86800 7300 0.08 45000 17900 6S0 0.04
48000 141200 2900 0.02 48000 95100 BUIO 0.09 48000 21100 720 0.03
51000 152000 2200 0.01 51000 102200 7900 0.08 51000 23800 790 0.03
54000 1&3000 2000 0.01 54000 106700 6900 O.Ob 54000 26+00 aoo 0.03
51000 174100 2000 0.01 moo 110200 6000 0.05 57000 29000 750 0.03
60000 186800 1900 0.01 60000 113500 5100 0.04 60000 31500 700 0.1)2
6300(1 200800 IBOO 0.01 63000 116600 4300 a.04 63000 moo b50 0.02
610000 2lJ300 1800 0.01 06000 moDO 3400 0.03 66000 36301)610 0.02
69000 220000 1800 0.01 69000 120100 2900 0.02 69000 38300 590 0.02
72000 139000 1800 0.01 72000 121000 2500 0.02 72000 .liOOO 570 0.01
75000 250900 1800 0.01 15000 121400 22<tO 0.02 75000 41500 5bO 0.01
8-9
Appendix Table 8-5.Continued.
LAST CIiA~CE S.C.~USTIC ~ILDERNESS S.C.15L~HD SIDE CH"NNEL-----------------------------------------------------------------------------------------------------------
IlAIIISTEII SITE CHlNilIlk CHINOOK ~AIMSTEM sm CflINO[l(CHINiJOK ~AIIiSTEII sm CHI/JOOK CHINOOK
DISCHliR&E "RE"~iIli Ii.I.DISCHARGE ARE"~u"n.1.DISCHAR6E "REA liUH H.i.
12600 moo 110 O.vl 12(1)0 4600 30 •O.iil 120W 315M 400 0.01
15000 ·mol)110 0.u1 15000 4601)30 ii.Ol 15(100 31500 400 0.01
18000 '17500 110 0.01 180M 4800 30 ii.Ol 16000 315M 400 0.01
21000 1751)0 110 0.01 21000 31901}4800 ii.15 21000 31500 400 0.01
24000 200M 1200 0.011 24000 49500 5100 O.lf)241)00 31500 400 0.01
2700(1 22000 mo O.Ob 27000 b0700 4300 0.07 moo 31500 4VO 0.01
30000 27000 1370 0.05 30000 moo 3700 0.1l5 30000 31500 400 0.01
33000 34000 1400 0.04 33000 76800 3000 D.04 33000 31500 400 0.01
3bOOO 46500 1420 0.03 36000 83300 2400 0.03 3bOOO 39M 3500 0.09
39000 70000 1440 0.02 39000 89900 1900 0.02 39000 45300 4800 0.11
42000 81&00 1470 0.02 42000 97000 1500 0.02 42000 51000 4100 0.08
45000 91000 1500 0.02 45000 104000 1260 0.01 45000 58500 3400 0.011
48000 94000 1610 11.02 48000 10'1000 900 o.vl 48000 &5500 2900 0.04
51000 96300 2050 0.02 51000 114000 100 0.01 51000 72000 2400 0.03
54000 98500 2560 0.03 54000 117400 SOO .00 54000 79400 2100 0.03
57000 1*200 2620 0.03 51000 119200 500 .00 57000 811700 18GO 0.02
60000 101800 2540 0.02 60000 120700 61)0 .00 60000 93100 .1700 0~02
63000 103200 mo 0.02 63000 121700 61)0 .00 63000 99800 1800 0~02
66000 104400 2350 0.02 116000 122200 &00 .00 66000 106200 2100 0.02
69000 105500 2240 0.02 119000 122700 700 0.01 69000 111900 2400 0.02
72000 1l>l.300 2100 0.02 12000 123000 700 0.01 72000 110200 2600 0.02
75000 107000 1900 0.02 75000 123500 BOO 0.01 75000 123300 2700 0.02
"AIIISTE"\lEST BAIIX SOOSE 2 SIDE CHANNa CIRCULAR SIDE CHAIIIl£L------------------------------------------------------------------
"AINsm SHE CIiIIfOOK CIIlllooK "AI/lSTElI SHE CHINOOK CIHNlIIIk "Al/lSTElI SHE CHllIOIIK CHtHOllK
DISCHARGE AREA IIIIA H.I.DISCHARIiE AREA 1iU"H.I.DtSCHARSE AREA IIIIA H•.I.
12000 61603 1082 0.02 12000 0 0 0.00 12000 59464 i47 O~OI
15000 61603 1082 0.02 15000 I)0 0.00 15000 5941>4 747 -0.01'
lBOoo 61603 1082 0.112 18000 0 0 0.00 18000 59464 747 0~01
21000 734211 10041 0.14 21000 0 0
0.00 21000 594&4 747 0.01
24000 80904 8325 0.10 24000 0 0 0.00 24000 59%4 747 0.01
27000 93353 5224 0.0iI 21000 0 I)0.00 27000 59464 747 0.01
30000 108613 4045 0.04-30000 9600 1500 0.16 30000 594&4 747 0.01
33000 11mB 3959 0.03 33000 21500 2900 0.13 33000 594&4 747 0.01
311000 117696·3861 0.03 36000 34300 4000 0.12 36000 71590 8117 0.12
39000 120505 3175 0.01 39000 47000 5100 0.11 39000 7&534 B404 0.11
42000 123397 3855 0.03 42000 &1400 &100 0.10 42000 80557 0013 O.HI
45000 129211 4113 0.03 45000 72000 6900 0.10 45000 85140 7472 0.09
48000 133649 4630 0.03 48000 81400 7000 0.09 48000 n944 7077 0.08
51000 136885 5080 0.04 51000 8711oo 6700 0.li8 51000 102530 b9911 0.07
54000 140761 5554 G.a4 54000 moo 6000 0.06 54000 1U323 119'19 0.06
57000 1442&9 6211 0.04 57000 97100 41100 0.05 57000 12575,6634 0.05
&001lQ 147899 &728 0.05 110000 99900 3100 0.03 bOOOO 134218 651&0.05
63000 1511142 7i}92 0.05 &3006 102000 2700 0.03 63000 143575 6906 0.05
6/,000 1542il5 7598 0.05 "MOO 103200 2400 0.02 06000 150869 7926 (1.05
69000 1511425 7913 O.{IS 69000 104200 21M '),02 69000 154&57 8561 0.06
nooo 158522 8078 0.05 72000 104800 1800 0.02 72000 157074 8840 0.0&
75000 1110818 6438 v.G5 75000 105100 11100 0.02 75000 15921l 81154 0.06
8-10
-
-
,..,..
Appendix Table 8-5.Continued.
,.,...
SAlJIU\SIDE CHANIl£L SUCKER SInE CHAHIl£l BEAYER vA"SIDE CHllMIIEl------------------------_...._..-----------------....-----.....----------------------_.._--...-----------
MIM5TEN SITE CHINOOK CHIllooK IbUNSTEN SITE CHINOOK CHINOOK IIAINSTEN SITE CHINOOK CHIIIOOK
r-o.DISCHARGE AREA IIlJ/l H.I.DISCHAII6E AIlO IIUA H.I.DISCHARGE MEA IlIJA H.1.
12000 42093 Ib5 .00 12000 0 0 0.00 12000 18900 50 .00
15000 >~;~2093 165 .00 15000 0 0 0:00 1~00 18900 50 .00
I BOllO
.
42093 Ib5 .00 18000 0 0 0.00 18000 18900 50 .00
21000 42093 165 .00 2101»0 0 0.00 21000 18900 50 .00
24000 42093 Ib5 .00 24000 0 0 0.00 24000 IB9flO 50 .00
:nvoo 42093 165 .00 moo 0 0 ERR 27000 IB900 50 .00
30000 41093 1&5 .00 ZOOOO 8500 lObO 0.12 3ססoo 18900 50 .00
33000-42093 165 .00 33<lOO 14900 1600 0.11 33000 18900 50 .00,....,
3&000 42093 1&5 .00 3l:IOOO 1&900 1570 0.09 3&000 189flO 50 .00
3'1000 49127 5759 0.12 3~19400 1510 0.08 39000 18900 50 .00
42000 49758 5740 0.12 42000 23600 1450 0.06 42000 18900 50 .00
45000 50289 5503 0.11 ~.29600 1550 US"45000 IB'IOO 50 .00
4Bm S08119 4980 0.10 -4IIIlIf.37100 20700 .0.011:
'.
22400 1120p-·48000 0.04
51000 51451 4<'10 0.09 ""1M ·46600.2940·0.0ii·.'StOOO:-2B600 2310 0.08
54000 52011 4046 0.08 .~S790f .4230 0.07 54000 moo·3560 0.1t
57000 521171l 3MS 0.07 S7OIO 1lIl900 4680·0.07 .'57000 35700 3840 0.11-60000 53294 3365 .0.06 600tll1 moo 4-490 0.011:....&0000 18000 3570 0.0'/
63000 54275 .3116 ..0••0&1130OO 73900 4230 0.06<63000..39&00 3060 0.08
&6000 5518.-2947 o.~6&000 159flO 3940 0.05 &&000 40800 2510 '0.0&
&9000 .5&053 ·.275T O.lr.i 69000 moo 1610 0.05 69000.''41SOt 2260 0;05
72000 ,5m2 .~-2'7B 0.05 72000 78100 3270 O.O~72000 419flO 2100 0.05
'75000 !1018 2714 O.O~..75000 78300 3010 0;04 1sooa'j2100 '.200t 0.05,.
...".-'
SIIIISET SID£CYAilm SUIIlISE SlllECllAllllEl TIbVftircRm s.c;
,-----------_..__....-.---'-.-~......._------------
flAINSTU SITE C1UHIllllC .CH1lIOOK --IIAIIISIEI SITE CHINOOK CHINOOK IIAIIISTEIf'..SHE'CIlINlllJl('CHUIODt:
DISCIIARGE AREA 11IM 8;1.DISCIlAIl6E AREA IlUA II.I~DIDR6£AREA 11IM H.I.
12000 495h2 5bi 0.01 l2GOO 0 0 0;00 12000 moo HOO 0.02
15000 495&2 568 0.01 lseot 0 0 0.00 15000'73300 1100 0.02.-18000 495&2 568 0.01 1BOOO ..0 0 0.00 ,IBOOO moo 1100 0.02
21000 49562 5bi 0.01 2100e 0 0 0.00 21000 73300 1100 0.02
24000 495&2 568 0.01 2400t 11 0 0.00 24000 n300 1100 0.02
27000 495&2 ..Shit MI 27000 0 0 0.00 27000 73300 1100 0.02
~30000 495112 568·0))1 30000 II 0 0.00 30000'73300 1100 0.02
33000 78488 4378 0.0&31000 11 0 0.00 33000 moo 1100 0.02
3&000 89472 4420 0.05 36000 19000 ..610 0.03 3&000 moo 1100 0.02
39000 97943 4&30 0.05 39000 53900 3250 0.011 39000 moo 1100 0.02....,42000 10&320 4984 0.05 4200G 78S00 56110 0.07 42000 73300 HOO 0.02
45000 122338 5U&0.04 45000 .97100 .6090 0.0&45000 moo 9300 0.12
48000 13547&584b 0.04 41lOOO 115400 4270 0.04 48000 moo 9000 0.10
51000 149248 581>8 0.04 51000 131100 3820 0.03 51000 108100 7500 0.07
54000 165990 5768 0.03 54000 1-46900 3540 0.02 54000 i23;100 5&00 0.05
"*"57000 173483 5487 0.03 57000 160&00 3250 0.02 57000 137700 2900 0.02
60000 188419 5931 0.03 bOOOt l75bOO 3180 0.02 60000 151200 1300 0.01
63000 194419 6000 0.03 &3000 192000 3460 0.02 &3000 15800.0 1330 0.01
b6000 203000 6231 0.03 6&000 207300 3700 0.02 66000 lb3100 1360 0.01
~/,9000 20&972 6263 0.03 69000 221400 4080 0.02 69000 166900 1390 0.01
12000 210128 6157 0.03 72000 229000 4190 0.02 72000 170700 1400 0.01
moo 2158&1 584S 0.03 1'5000 233300 4210 0.02 75000 173500 1400 0.01
-
.-
8-11
-
CHINOOK WUA
CASWELL CREEK MOUTH
10
~
9 ,'"
,,;p-'"
......8::",..,
a-7n.....,
l5.-.6~-8c:
~~5M4-
0~.3
<:)
~2
~
0
10 59 70
-.
BEAVER DAM SLOUGH
6
~
............-----:;5-a-
".....,-,ii 4
c:
~UI
~&~
~.3
0~
:I:
8 ~
~2
------~'
10 30 50 70
~housands~MAINSTD4 DISC ARGE AT UNSHINE-(cfs)
Appendix Figure B-1.Weighted usable area for juvenile chinook salmon
at Caswell Creek and Beaver Dam tributary study
sites as a function of mainstem discharge.
B-12
-
-
-
"...
0.9
.......0.8..;-d'0.7It.....,
~0.6~~c~5:0.5~~~0.4
a~0.3"5~0.2
0.1
0 -_._---
10
BEARBAlT SIDE CHANNEL
3Q SO 70
(Thousands)_
MAlNSTEM DISCl-t~GE AT SUNSHINE (cfs)
Appendix Figure 8-2.Weighted usable area for juvenile chinook salmon
at the Hooligan and Bearbait Side Channel study
sites as a f~nction of mainstem discharge.
~8-13
2.B
2.6
2.4-
,...
2.2i
0-2
II)
'-'
~1.8
~-S 1.6c:
~:Ul 1".4-~~1.2::£.
0~O.BG
(;j 0.6~
0.4-
0.2
0
10
CHINOOK WUA
LAST CHANCE SIDE CHANNEL
Br••clt.d
t
.30 50 70
(Thousands)
MAINSTEM DISCH~GE:!<T SUNSHINE:(cts)
RUSTIC WILDERNESS SIDE CHANNEL
90
-
....
6
5,...
..;-cT
I)
'-'4-
~~-Sc:~g 3 Bre8ched~~t::£.
0 2~
()
~
-
o -----
10 30 50 70
.(Thousands).
MAlNSTEM DISCli!'RGE;!'oT SUNSHINE:(cts)
""'"
Appendix Figure B-3.Weighted usable area for juvenile chinook salmon
at last Chance and Rustic Wilderness Side Channel
study sites as a function of mainstem discharge.
B-14
-
-
Appendix Figure B-4.Weighted usable area for juvenile chinook salmon
at the Island Channel and Mainstem West Bank study
sites as a function of mainstem discharge.
8-15
'""'"
~
CHINOOK WUA
GOOSE 2 SIDE CHANNELe ""'"
7
::;--6e-
li.....
~5
c
~i 4
~~~.:5 .-0~:r
~2
~----
~
0
10 30 59 70 ".,.,
CIRCULAR SLOE CHANNEL
10
IlI'ucfMtd
9
......e...-U 7II.....-~6~-8c
~i 5~.~~
:£.4-
0
e!.:5:r
Cl -iii~2
1
0
10 30 50 70
~uSandS~MAINSTEM DISC GE AT UNSHINE (cfs)-
Appendix Figure 8-5.Weighted usable area for juvenile chinook salmon
at the Goose 2 and Circular Side Channel study
sites as a function of mainstem discharge.
8-16
70
........
.......................
BEAVER DAM SIDE CHANNEL4-.------..--:=.:..::..::.:.:....:....:..--------------,
2
3
1
o ----------------
10 30 50
(Thousands)
MAlNSTEW DISC~GE:AT SUNSHINE (cfs)
2.5
1.5
3.5
0.5
,...
I
--
-
Apoendix Figure 6-6.Weighted usable area for juvenile chinook salmon
at the Sauna and Beaver Dam Side Channel study
sites as a function of mainstem discharge.
~B-17
30 50 70
(Thousands).
MA1NSlEM DISCHARGE AT SUNSHINE (cfs)
CHINOOK WUA
SUNSET SIDE CHANNEL
7
"""
6
......
~..,.
1:1 5II......
~~-a 4 Breaclledct~i!~3~
a ..-
~2J:~~
~
0
10 30 59 70 1"''''
~
SUNRISE SIDE CHANNEL
7
6
......Breeched~t1:1 5
~iII.....
~~-a 4c
~i ~~~3:£
a~2
C)
iii~
o-l----_---.....-..-.I!..---r---r-----r-----,r------i
10
Appendix Fiqure 8-7.Weighted usable area for juvenile chinook salmon
at the Sunset and Sunrise Side Channel study
sites as a function of mainstem discharge.
8-18
)1 -1 ))1 1 1 .j -1 l }
10
9
.-;.8....-a 7I/)
'-"
~6
c
~i 5~~:>t:,4
co I 0I~......31.0 J:<:>w
~2
1
0
10
CHINOOK WUA
TRAPPER CREEK SIDE CHANNEL
Projected WUA ..:
(Head barely overtopped)
~----
30 50 70
(Thousands)
MAINSTEM DISCHARGE AT SUNSHINE (efs)
Appendix Figure 6-8.Weighted usable area for juvenile chinook salmon at the Trapper
Creek Side Channel study site as a function of mainstem discharge.
Appendix Table B-6.Weighted usable areas and habitat indices for juvenile
coho salmon in lower Susitna River model sites,1984.
-
-
ROLLY CREEK IIOOTH CASliELL CREEk rtOUTH BEAVER DA~SL IlUSH
..-.-------------------...-_..._----_..----_______a ____________________________-_..--------------------------_..-----
~AIIiSTE"SITE COHO COHO ""IIISTE"SHE COHO COHO "AIIISTE"sm COHO ~OHD
OlSCHAA6E AREA lIUA H.I.~ISCHARSE ARE ..lIUA H.L DISCHARGE AREA lIUA H.I.
12000 84900 7900 0.09 12090 lb200 1350 i).os 12000 11600 1700 1).15
15000 84M:1'100 0.09 1500t}lb20{)1350 (i.03 15(IC'{1 !tbOO 170')0.15
18000 .84900 mo 0.09 18(/1)0 Ib200 1350 0.08 180(;('llbOO 1700 0.15
21000 84900 7900 O.M 21000 16200 1350 o.oe moo woo 1700 0.15
24000 85300 7900 0.j)9 24000 Ib200 135(i 0.08 24000 moo 170C'1).14
27000 88300 /700 0.09 27000 16300 1500 u.O'1 27000 12200 mil 0.14
30000 moo 7500 .O.OB 30000 16700 1700 0.10 10000 moo 1700 0.14
33000 998(1)mo 0.07 33000 moo 2000 (I 1"1 33000 13000 1700 0.13...
36000 109900 6900 0.00 30000 18000 2300 0.13 36000 13400 17n0 0.13
39000 121000 .1I~00 0.05 390liO 18900 2500 0.13 39000 moo 1700 0.12
42000 135000 .5900 0.04 42000 1'1800 2800 0.14 ~2000 1~400 1610 0.12
45000 152600 5500 Q.04 45000 21000 3000 0.14 4liOOO 15000 16liO 0.1l
48000 178500 51100 0.03 48000 21800 3200 0.15 48000 15100 1610 0.10
51000 198800 1300 0.04 51000 22700 340G 0.15 510GO 16300 1s.40 0.09
54000 213000 9200 0.04 5-4000 moo 3600 0.15 54000 16BOO 1480 0.09
57000 223200 10100 0.05 51000 246GO JBOG 0.15 510GO moo 'IUO 0.08
IIGOOO 229900 1070G G.OS 60000 25500 ~ooo 0.16 bOOOO IBliOO Hao O.OB
63000 23'.'i001)11200 0.05 moo 20300 ~300 0.16 63000 1910G 1540 0.G8
66000 238100 11700 G.05 60000 moo 4400 0.10 60000 20BGO 1630 0.08.
69000 24t1100 120W 0.05 09000 279GO ~700 0.17 69GOO 21600 1740 0.08
72000 m200 12300 G.05 72000 28900 4'100 0.J7 72000 22100 1780 0.08
75000 243000 12500 G.OS 75000 29100 5100 0.17 75000 22600 1B1'>0.li8
Apoendix Table B-7.Weighted usable areas and habitat indices for juvenile
chum salmon in lower Susitna River model sites,1984.
HOOLlIlAll SIll£CIIIlNIIEL KRIITO SlIllf6ll HEAD BEARBAIT SIIIE CHANNEL---------_._----------_...---_..--------------------------------------------------
MIHSTEII SITE ..CIlUIl &HUll IlIlINSTEII SITE CHUII CIIUII ~AINSTEIl sm (HUll COOl!
IIISCHARGE A~"lIUA H.I.IIISCHARGE AREA IIUA H.r.IIISCI!/lRBE AREA WUA H.1.
12000 63400 28500 0.45 12000 48200 39600 0.92 12000 31\)0 13GO 0.42
15000 63400 28500 0.45 15000 48200 39600 0.82 15000 3100 1300 0.42
18000 63400 28500 0.45 18000 48200 39600 0.82 18000 3100 1300 0.42
21000 63400 28500 0.45 21000 48200 39~0 0.82 21000 3100 13GO 0.42
24000 79900 47'100 0.60 24000 48200 39600 o.B2 24000 3100 1300 0.42
27000 i6900 46700 0.54 27000 48200 39600 0.82 27000 3100 1300 0.42
3ססoo 901100 44000 0.48 30000 ~82OO 39000 0.82 3ססoo 3100 1300 0.42
33000 96500 moo Q.43 l3OllO 48200 39600 o.a2 33000 3100 1300 0.42
36000 104l1oo 38400 0.31 -_.-3bOOO.5000&39600 0.79 36000 5700 I~OO 0.25
l'iOoo Imoo 34700 0.31 ~19000 61900 42000 0.62 39000 108GO 1900 0.19
42000 122'100 30300 0.25 -,;:,'42000 11500 44500 0.51 42000 ·14600 2600 0.18
45000 111300'16100 o.~·-·:r:-.B6IIOO "100 .0.53:45000 17900 330Ct 0.18
48000 141m 21900-<t.lo·..-4801»95100 4160t 0.50 48000 moo 4100-0.19
~1000 152000 ill900 0.12
.-
51000 102200 4lJ5Oec c-Q.45 51000 23800 5300 .0.22
54000 163000 19100'.0.11 54000 106100·-.42300 0.40 54000 2&400 5700 0.22
57000 174100 .17600.:0.10 57000 110200 38lOO 0.35 57000 29000 5500 0.19
6ססoo 18llBOO:.11.200.0.09 60000 113500 ·34400 0.10 60000:31500.5100 0.1&
03000 200800 10900 0.08 63000 116600 29100 0.21 .6lOOO 33900 4700 0.14
66000 113300 16700 0.09 66000 H'lOOO 24100 0.20 66000 l6300 «00 0.12
69000 226000 IMoo 0.07 6'1000 120100 19800 0.16 69000 38300 4200 0.11
72000 239000 16100 0.07 .12000 121000 17800 0.15 72000 4000t ,4100 0.10
75000 250900 ·l511oo 0.011 15000 121400 15200 0.13 1liOOO'41500 4000 tl.l0
8-20
-
--
Appendix Tab1e 8-7.Continued:
LAST CIIAIlCE S.t.-.RUSTlC 1I1LDERNESS S.C.·ISlAilD SIDE CIIAlfIIEl.-----_.---~----------.---------------______________...__..:.._-.a.____
IVIIIISTU SITE OWl!.CHIlli IIAIIISTEJI SITE CHUII ClIIII IIAINSTEJI SITE CHill .CIIJII
DISCIlARSE AREA lRIA H.I.DISClfAR6E AREA IIIlA H.I.IlISCHAR6E AREA MIlA H.I.
12000 17500 11500 o.oll 12000 4800 3600 0.75 12000 .moo 19300 0./11
15000 17500 11500 0.66 15000 4800 3600 0.75 15000 31500 19300 0.111
IBOOO .17500 11500 0.66 laooo 4BOO 3600 0.75 18000 31500 1'1300 0.61
21000 11500 1/500 0.66 21000 31900 30800.0.97 21000 31500 19300 0.61
24000 20000 11500 0.58 ·24000 49500 32500 D.66 24000 31500 moo 0.61
27000 22000 moo 0.52 27000 60700 27600 0.45 27000 31500.19300 0.61
30000 27000 11500 0.43 30000 69700 22700 0.33 30000 31500 19300 G.61
33000 34000 moo 0.34 33000 70800 18100 0.24 33000 31500 19300 0.61
36000 46500 11500 0.25 36000 83300 13700 0.16 36000 39200 28100 0.72
39000 70000 1/500 0.16 39000 89900 10600 0.12 3~00 45300 moo 0.lI4
42000 81000 11500 0.14 42000 97000 8000 0.09 42000 51000 25800'0.51
45000 91000 11500 0.13 45000 104000 7400 0.07 45000 sa500 22700 0.39
48000 94000 moo 0.12 48000 109000 saoo 0.05 49000 65500 /9700 0.30
51000 96300 15100 O.lb 51000 114000 4200 0.04 51000 72000 17400 0.24
54000 'tlI500 20200 0~21 54000 117400 3300·0.03 54000 moo 15100 0.19
57000 100200 moo 0.19 57000 119200 3000 0.03 57000 86700 l3200 0.15
60000 101800 18000 0.18 60000 120700 3000 0.02 60000 moo 12400 0.1l
moo 103200 16200 0.16 63000 121700 3000 0.02 63000 99800 12700 0.13
66000 104400 13600.0.13 66000 122200 3000 0.02 6bOOO 106200 13000 0.12
69000 105500 10500 0.10 69000 122700 30UO 0-02 69000 111900 moo 0.12
72000 106300 8800 0.08 72000 123000 3000 0.02 moo JJB20D moo 0.12
75000 107000 7600 0.07 75000 123500 3000 0.02 75000 123300 13600 0.11
"AINSTEII iiEST BAlik 600SE 2 SIDE CHAIIIlEL CIRCULAR SiDE CHANNEL
-------_..._-------------------------------------------------------_._--......--------------_._-------------
~AINSTEII SITE (Ill!"CHIlli IIAINSTa SITE l:HUII CHIJIl ~AINSTElI SITE GJiU"CHUII
DISCHARBE AREA lIUA H.1.DISCHARGE AREA ~UA H.L DISCHARGE AREA MUA H.I.
12000 6160~47090 ,).76 J2000 0 0 Uo 12000 59464 411109 0.78
150M bl~"4mo 0.76 15000 0 0 0.00 15000 59404 46109 0.78
18000 II 160l 470%iJ.76 18000 0 0 0.00 181)00 59464 46109 0,78
2100\)73426 53955 0.73 21000 0 0 0.00 21000 59464 46109 0.78
24000 80904 43289·0.54 24000 0 0 0.00 24000 59464 46109 0.79
27000 93353 316011 0.34 27000 0 0
(;.00 27000 59464 46109 0.78
30000 108613 27151 0.25 30000 9600 4900 0.51 30000 59464 411109 0.78
33000 114738 23420 0.20 33000 21500 11000 0.51 33000 59464 46109 0.78
36000 117696 m82 0.19 36000 34300 17400 0.51 36000 71590 44495 0.62
39000 .120505 21096 0.18 39000 47800 25500 0.53 39000 76534 4%06 0.58
42000 J23397 2/218 0.17 42000 moo 31800 0.52 42000 .·80557 42269 0.52
45000 129211 22389 0.17 45000 72000 37900 0.53 45000 85140 42176 0.50
48000 133649 26710 0.20 4BOO0 81400 moo 0.51 48000 92944 43074 0.46
51000 136885 27661 0.20 51000 87800 42600 0.49 51000 102530 45026 0.44
54000 140761 30382 ·0.22 54000 93200 40700 0.44 54000 113323 50073 0.44
57000 .144269 :U815 0.22 57000 97100 33400 O.M 57000 115m 50248 0.40
60000 147B9'1 33950 0.23 60000 99900 24000 0.24 60000 134218 46305 0.34
63000 151842 35'153 0.24 63000 102000 18600 0.18 63000 143575 49339 0.34
66000 154205 364.89 0.24 66000 103200 13800 0.13 66000 150869 49565 0.33
09000 1511425 30211 0.23 119000 104200 10400 0.10 119000 154657 50346 0.33
72000 158522 37029 0.23 72000 104800 Il3OO 0.08 72000 157074 48491 0.31
75000 160818 3680'/0.23 75000 105100 7400 0.07 75000 159211 .46797 0.29
\
8-21
Appendix Table B-7.Continued.
SAUIlA SIDE CIWllIEl SUO:ER SIDE 0IAIIIlEI.B£AYER llM 51*CJlAtlIIEL
-----------------------------------------------------------------------
MIIISTEII SITE CIlIlIl CIilII IlAINSm SITE CHUIl C!lIJII lIAINSTEII SITE CHUIl CHIllI
DISCHAR6E AREA ~H.I.DlSCHAfl6E AAEjl IItIA H.1.DISCHAR6E AREA IiIIA H.I.
12000 420'13 31754 0.75 12000 0 0 0.00 12000 18900 11900 0.113
15000 42093 31754 0.75 15000 0 0 0.00 15000 1890ct .11900 0.113
18000 42093 31154 0.75 18000 0 0 0.00 18000 18900 11900 0.113
21000 420'13 31754 0.75 21000 0 0 0.00 21000 18900 11900 0.113
24000 420'13 31154 0.15 24000 0 0 0.00 24000 18900 11900 0.113
27000 42093 31754 0.15 27000 0 0 ERR 27000 18900 11900 0.113
30000 42093 3175'0.75 30000 8500 7300 0;86 30000 18900 11900 0.03
33000 .42093 31754 0.15 33000 14900 11800 0.79 33000 18900 11900 0.113
30000 42093 31754 0.75 311000 111900 12700 0.15 311000 18900 11900 0.113
39000 49127 27301 0.511 39000 19400 13200 0.08 39000 18900 11900 0.03
42000 49758 20413 0.53 42000 231100 13400 0.57 42000 18900 11900 0.113
45000 50289 25204 0.50 45000 29boo 14300 0.48 45000 18900 11900 0.03
48000 5OBll9 231170 0.47 48000 moo 19900 0.54 48000 22400 13200 0.59
51000 51451 22565 0.44 51000 4&000 27700 0.59 51000 28000 15700 0.5b
54000 52011 218311 0.42 54000 57900 33100 0,58 54000 moo 17500 0.54
57000 52078 21381 0.41 57000 6&900 34#0 0.51 51000 35700 18800 0.53
00000 53294 20990 0.39 110000 71301)32900 0.411 110000 38000 18200 0.48
113000 54275 206&9 0.38 03000 73900 W800 0.42 63000 39000 Ib400 0.41
66000 55184 20938 0.38 116000 75900 28200 0.37 cbOOO 40800 14000 0.34
&9000 511053 21017 0.37 &9000 moo 25000 0.32 09000 41500 12100 0.29
72000 57142 21153 0.37 72000 78100 21800 0.28 72000 41900 11300 0.27
75000 c10t8 :3075 Q.38 moo 78300 19200 0.25 75000 42100 10700 0.25
SU/iSEl SIDE eHA~~El SUNRISE SIDE CHANNEL TRAPPER CREEK S.C.
------------...------------------------------------,..----------------------------------_..------
"AINSlEIl SiTE CHUII CHIlli IlAINSTEIl SITE CHUIl CHUIl IlAIIISTEIl SITE CHUIl eHU"
DISCHARGE AREA WA H.I.DISCHAR6E AREjl lIUA H.1.DISCHARGE AREA WUA H~L
12000 49562 27135 0.55 12000 I)0 0.00 J2000 moo 45400 0.62
15000 ·mo2 27135 0.55 15000 0 0 O.Oll 1S{)00 73300 45400 0.62
lBOOO 49562 27135 0.55 18000 II 0 0.00 18000 moo 45400 O.b2
21000 49562 27135 0.55 21000 0 0 0.00 21000 ;3300 45400 0.62
24000 4951>2 27135 0.55 24000 0 0 0.00 24000 73300 45400 O.b2
27000 49502 27135 0.55 27000 0 0 0.00 27000 moo 45400 0.112
30000 495112 27135 0.55 30000 0 0 0.00 300Qtl 73300 45400 0.112
33000 78488 34059 0.43 33000 0 0 0.00 33000 73300 45400 0.62
36000 89472 34808 11.39 36000 19000 11200 0.33 36000 73300 45400 0.il2
39000 97943 37649 0.38 39000 53900 32400 0.00 3'1000 73300 45400 0.62
42000 106320 39888 0.38 42000 ·7~464110 0.5'1 42000 13300 45400 0.62
4SOOO 122338 46376 0.38 45000 97100 moo 0.51 45000 71000 44800 0.58
48000 1354711 51185 0.38 48000 115400 44500 0.39 4BOOO 91200 41200 0.45
51000 149248 52D7t o.~51000 131100 37500 o.:n 51000 108100 3%00 0.32
54000 11.59'!0 53786 0.32 54000 146900 31100 0.21 54000 123300 27500 0.22
57000 1734113 48410 0.28 57000 16~·20600 0.17 57000 .·137700 1'l5OO 0.14
00000 18841'1 50093 o,'n 60000 1151>00 25200 0.14 60000 151200 10700 0.07
03000 19441'1 .43299 0.22 1>3000 192000 25300 0.13 113000 158000 10200 0.06
66000 203000 41715 0.21 66000 207300 211200 0.13 116000 163100 10000 0.06
69000 206972 31100 0.18 69000 221400 21700 0.13 11'1000 166900 9800 o.ob
72000 210728 33481 0.16 72000 229000 28500 0.12 72000 170700 91100 0.011
75000 215861 32949 0.15 75000 233300 29000 0.12 75000 173500 9500 0.05
I
B-22
~I
....
-
-
-
-
-CHUM WUA
HOOUGAN SIDE CHANNEL
50
45
.....
i 40
0-
".....
~35
~-ac
~lil .30.~a:£250~
Cl 20i>i~
15
10
.-10 30 50 70 90 110
.~USQndS~
MAINSTEM DISC _GE AT .UNSHINE (cfs)
KROTO SLOUGH HEAD
50
~,45
........40 -------
~0-
".....35
~~-a .30c
~g
~a 25:£
0 20~~:I:
Q
~15
10
~~,
5
10 30 50 70 90
.-~hOUSQndS~.
MAINSTEU DISC ARGE AT UNSHINE (cfs)
Appendix Figure B-9.Weighted usable area for juvenile chum salmon at
the Hooligan Side Channel and Kroto Slouqh Head
study sites as a function of mainstem discharge ..
B-23
CHUM WUA
BEARBAIT SIDE CHANNEL
7
6
..........-0-5II).....,
~~~4c
~a~5 3::£
~2:I:
Q
l.lJ~-
0 -10 S9 70
~
ISLAND SIDE CHANNEL
36
34
~
32........;30-0-281ft.....,
ii 26
c 24~g
~5 22 ~
::£20
~18~16
14
12
10
10 30 50 70
~OUSandS~MArNSTEM DISC ARGE AT UNSHINE (cfs)~
Appendix Figure 8-10.Weighted usable area for juvenile chum salmon at
Bearbait and Island Side Channel study sites as
a function of mainstem discharge.
8-24
-CHUM WUA
MAINSTEM WEST BANK
54-52
50
""48-46-0-44II
'-'
~42
40
c:38WO
...Jill
~2 36
::It 34
,-0 32~30z
Cl 28i£l
3:26
24
22
20
10 30'59 70
-
Appendix Figure 8-11.Weighted usable area for juvenile chum salmon at
the Mainstem West Bank and Goose 2 Side Channel
study sites as a function of mainstem discharge.
B-25
-
""'"
CHUM WUA
CIRCULAR SIDE CHANNEL.55 ~
54
53
,....52 -~
a-51
II 50....."
~49 Breached~~c 48 t~51
~5 47
:>t 46 """"0 45~
J:44C)
iii~43
42
41
40
10 30 59 70
-SAUNA SIDE CHANNEL
35
34 ar••cmeci
33 t,...32~
cT 3~
II 30.....,
~29~-8c 28~~
~5 27
::£.26
0 25~J:240
iii~23
22
21
20
10 30 50 70
ghousandS~
MAINSTEIvI DISC ARCE AT UNSHINE Cds)
Appendix Figure B-12.Weighted usable area for juvenile chum salmon at
the Circular and Sauna Side Channel study sites
as a function of mainstem discharge.-
B-26
.--
CHUM WUA
SUCKER SIDE CHANNEL
40
35-"....,
~30
0-......,
~25
~~c'jg 20 ,,~5 Breached ""~1S ta
~
~
0 10i:jj
~
S
0~
10 30 50 70
SEAVER DAM 510£CHANNEL
20
19
16
17
........16~15
0-14.......,
i1 13
12
c 11
~i 10~~9~6
0 7~6~
""'"Cl 5i:jj
~4
3
2
1
a
10
Breached
t
....---------------"....
"..................
30 50 70
(Thousands)
MAINSTEM DISCHARG~AT SUNSHINE (cts)
Appendix Figure B-13.~Jeighted usable area for juvenile chum salmon at
the Sucker and Beaver Dam Side Channel study sites
as a function of mainstem discharge.
B-27
CHUM WUA
SUNRISE SIDE CHANNEL
60 I I
50,-...
i
cT
(I)
'-"40
li-..
~-8c
'j~30 -,I I "---------~~:>t,
0 20coI~I
N ::r:(Xl e"
W~
10
30 50 70
.(Thousands)
MAINSTEM DISCHARGE AT SUNSHINE (cfs)
a I-T "T.1 iii I I
10
Appendix Figure B-14 ..Weighted usable area for juvenile chum salmon at the Sunrise Side
Channel study site as a function of mainstem discharge.
)_I J J j j _J J -J t I J ))I J
1 1 J I J 1 1 ))J j ))~
J
o::J
I
N
l.O
~ppend1~T~~le ~-~.~~19~~~~y~~~le a.r~~~~n~h~~1~~t 1n91~~$,fer j~Y~n11~SockeY~s~'~9n 10 19~~r ~Y$ftn~B1Y~F ~9gel s1t~S!1984.
""..,",,'.''''-,,-,,,,''''''''''''''''''''''''''''''-''''''",','"''"","""'",'",""."""
ROLLY CREEK "OUTH CASWELL CRW IIOUT~IUVER Mil SLOIJ6H--...-----_...........----------..__..._..._......-...--_.._--_...~.....------_...............................----...~_..----...-------_......_...,..-........_...........--_..
nAI1l5IEIl SUE ,SOCKEIE SGUHE "A INsTEn SUE SOCKEYE SOCKEIE itA INSIEft SITE SOCKEYE SUCKE,E
~ISCHARGE AREA NUA H.I.DISC~AR6£AREA'lIUA H.I.DISCHAR6E AREA NUA H.J.
12000 84900 10600 0.12 12000 16200 1350 0.08 12000 11600 6200 0.53
15000 84900 10600 0.12 15000 16200 1350 0.08 15000 11600 620\)O.S~
18000 94900 10600 0.12 18000 10200 1350 0.08 18000 11600 6200 O.SJ
21000 84900 10600 0.12 21006 WOO 1350 0.08 21000 11700 6200 0.5:
24000 85300 10600 0.12 24000 16200 1600 0.10 24000 11900 mo (0.52
27000 88300 11000 0.12 27000 16300 1700 0.10 27(100 12200 64(0)O.S:
30000 moo 13400 0.14 30000 16700 1900 0.11 30000 12500 660·)&553
33000 99800 17600 0.18 33000 17300 2300 0.13 moo 13000 6700 v.52
36000 108900 22900 0.21 36000 18000 2600 0.14 36000 13400 7000 0.5:
39000 121000 28900 0.24 39000 18900 3100 0.16 39000 moo 1100 0.51
42000 135000 35500 0.26 42000 19800 3100 0.19 42000 14400 1300 1).51
45000 152600 43400 0.28 45000 21000 4300 0.20 45000 15000 1500 0.50
48000 178500 52100 0.29 48000 21800 5000 0.23 48000 15700 7700 0.49
51000 198800 64400 0;32 51000 22700 5700 0.25 51000 16300 BOOO 0.49
54000 213000 75300 0.35 54000 23700 6400 0.27 moo 16800 8200 O.H
57000 223200 B2800 0.37 57000 24600 7200 0.29 57000 17600 B600 0.49
60000 '229800 88200 0.18 60000 25500 7900 0.31 60000 18500 8900 0.48
63000 235000 93000 0.40 63000 26300 8600 0.33 63000 19700 9400 0.4S
66000 238700 moo 0.41 66000 27200 9200 0.34 66000 20800 10200 0.4'1
09000 241600 99900 0.41 69000 27900 10000 0.36 69000 21600 10800 0.5~
72000 243200 100700 0.41 72000 28900 10600 0.37 72000 22100 11000 0.5e
75000 243600 101500 0.42 75000 29700 11400 0.38 •75000 22600 11000 0.4~
SUCHR SIDE CHIlH~EL BEAVER DAft SIDE CHANhEL SU~SET SID£CHANNEL SUNRISE SIDE CHANNEL
----------_........_---...-------_.._..._---_....----------_..._....._-_........._----_..-------.._.._-.....----------_....._-..._--------...-----_..._---------------------------...-
ItAINSIEft SITE SOCKEYE SOCKEYE ItAINSlEH SITE SOCKEYE socmE "AINSTEH SITE SOCKEYE SOCKEYE "AI~STE"SITE SOCKEYE SOCKEYE
DISCHARGE AREA WUA H.I.D[SCHAR6E AREA NuA H.I.DISCHARGE AREA llUA H.I.DISCHARGE AREA ~UA H.I.
12000 0 0 D.Qo moo 18900 300('0.16 12000 49562 7182 0.14 12000 0 0 0.00
15000 0 0 0.00 15000 18900 1000 0.16 15000 I 49562 7182 0.14 15000 o .0 0.00
18000 0 0 0.00 18000 18900 3000 0.16 18000 49562 7182 0.14 18000 0 0 0.00
21001i '0 0 0.00 21000 18900 3000 0.16 21000 49562 7182 0.14 21000 0 0 0.00
24000 0 0 0.00 24000 18900 1000 0.16 24000 49562 7182 0.14 24000 0 0 0.00
27000 0 0 ERR 27000 IBM 3000 0.16 27000 49562 7182 0.14 27000 0 0
0.00
30000 8500 1200 0.14 30000 18900 3000 0.10 30000 49562 71112 ··0.14 30000 0 0 0.00
33000 14900 1800 0.12 33000 18900 3000 0.16 33000 78488 6738 0.09 33000 .0 0 0.00
16000 16900 1100 0.10 36000 18900 3000 0.16 36000 89472 6493 0.07 36000 19000 400 0.02
19000 19400 1500 0.08 39000 18900 3000 0.16 39000 97943 6639 0.01 39000 53900 4700 0.09
42000 moo 1200 0.05 42000 18900 3000 0.16 42000 106320 6828 0.06 42000 78500 5800 0.01
45000 29600 1200 0.04 45000 18900 3000 0.16 45000 122338 7412 0.06 45000 '91100 51100 0.06
48000 37100 2600 0.07 48000 22400 3200 0.14 48000 135476 7529 0.06 48000 115400 3400 G003
51000 46600 4000 0.09 51000 28000 3700 0.13 51000 149248 7108 0.05 51000 131100 3200 0.02
54000 57900 5000 0.09 54000 32600 4100 o.n 54000 165990 6643 0.04 54000 146900 3100 0.02
57000 66900 5300 0.08 57000 35700 4300 0.12 57000 173483 600b 0.03 57000 160600 3000 0.02
60000 moo 5400 0.08 60000 18000 4100 0.11 60000 188419 6662 0.04 60000 175600 3000 0.02
63000 moo 5500 0.07 63000 39600 3900 0.10 61000 194419 6275 0.03 61000 192000 3100 0,02
66000 15900 5600 0.07 66000 40800 3600 0.09 66000 203000 6740 (1.03 66000 207301)31(10 0.01
69000 moo 5600 0.07 69000 41500 3200 0.08 69000 206972 6650 0.03 69000 211400 3200 0.01
72000 78100 5600 0.07 72000 moo 3000 0.07 nooo 210728 7m 0.03 72000 229000 3200 0,01
75000 78300 5600 0.07 75000 4-2100 2800 0.07 75(100 215861 7661 0.04 75000 233300 :;200 0.01
SOCKEYE WUA
CASWELL CREEK MOUTH ~,14
13 ,
'"
,12
'".-.......
11i
&10
-!.
~9
~
~-S 8c:
~g 7
~5 6 """:>E..
0 5
/!:!4J:
Qw 3~
2
0 ~
10 30 59 70
-BEAVER DAM SLOUGH
13
~
12
.-.
i
...-&11 .....-
~,~
~
~-S 10c:
~g
~~5 9:>E..
0
/!:!8 ~J:
Cl
iii
~,
------
6
10 30 50 70
ghousandS~.
MAINSTEM DISC ~G_E .~T_UNSHINE (cts)
Appendix Figure B-15.Weighted usable area for juvenile sockeye salmon
at Caswell Creek and Beaver Dam tributary study
sites as a function of mainstem discharge.
8-30
)))1 1 1 j .•~J 1 )]j I ]»1 J
SOCKEYE WUA
SUNRISE SIDE CHANNEL
7 ,,
6.-
"'"':'..........
r:r 5 -
lD
'-'"
~
~-8 4~
c
~~3~Sreached,0...-----"-------~~t:It
0
OJ I ~2 -
,:x:w
Q.....
~
1 -
o-+-i i J i I I I I
10 30 50 70
.(Thousands)
MAINSTEM DISCHARGE AT SUNSHINE (cis)
Appendix Figure B-16.Weighted usable area for juvenile sockeye salmon at Sunrise Side
Channel study site as a function of mainstem discharge.
-
APPENDIX C
COMPARISON OF THE IFIM AND RJHAB
MODELLING TECHNIQUES AT TWO
SEL ECTED SITES
-
.....
INTRODUCTION
In 1983,two techniques were used to model the effects of mainstem
discharge on juvenile salmon habitat within the middle Susitna River.
The Instream Flow Incremental Methodology (I FIM)(Bovee 1982)was used
at seven sites (Hale et al.1984)and the RJHAB habitat model developed
in Marshall et al.(1984)was used to model six other sites.Since
studies of the effects of mainstem discharge on juvenile salmon habitat
within the lower Susitna River were begun in 1984,it was desirable to
compare these two modelling methods.Both methods were used,therefore,
at the same transects withi n two sites to compare resul ts from the two
techniques.
METHODS
Trapper Creek Side Channel (RM 91.6)and Island Side Channel (RM 63.2)
were selected as sampl ing sites for this comparative study because they
represent two different channel types of the lower Susitna River.
Trapper Creek Side Channel is a simple straight channel.Island Side
Ch~nnel is a more complex,winding channel.Further descriptions and
photos of these two sites are contained in Quane et al.(1985).
Descriptions of the two modelling techniques will not be presented here.
Detailed descriptions of the IFIM are presented in Appendix D of this
report and Bovee (1982),and summarized in Section 2.0 of this report.
The original RJHAB model was first developed and described in Marshall
et al.(1984)and modifications were described in Section 2.0 of this
report.
Both techniques entail taking depth,velocity,and cover or substrate
measurements spaced at intervals across transects runni ng at ri ght
angles to the channel.Hydraulic models which have been developed for
use in the IFIM include the IFG-2 model which is based on open channel
flow theory and one set of field data and the IFG-4 model which is based
more strongly on fi el d data as three sets of fiel d measurements are
recommended (Milhous et al.1981).Fewer measurements are taken for
each RJHAB field data set than for the IFIM models but up to seven data
sets are taken.No hydraulic model is developed by the RJHAB and the
model runs on a spreadsheet with a microcomputer.The IFIM models can
generate estimates of equivalent optimum habitat called weighted usable
areas (WUA's)with any flow within their calibration range,while the
RJHABmodel only cal culates WUA I S at di scharges for whi ch measurements
are taken.Therefore,it is necessary to interpolate between point
measurements generated by the RJHAB model.The RJHAB model does have
the advantage of being able to run in areas heavily influenced by
mainstem backwater or sloughs with flows less than 5 cfs.The measure-
ments and data analysis for the RJHAB model were taken by different
investigators than those who took the IFIM measurements and analyzed
them.
The RJHAB model uses measurements at an additional upper transect within
each of the sites.This upper area was very similar to lower sections
of the site,and therefore would not change comparabil i ty of the two
C-1
-
methods.The IFIM presents resul ts of the ana lysi s on the basi s of a
1000 foot reach,while the RJHAB model presents WUA's for the site.
Therefore,the length of each site as used in the RJHAB model was
calculated and WUA's were adjusted to the basis of a 1000 foot reach.
At Island Side Channel,two additional partial transects were put in for
IFIM analysis of the site (see Appendix D),and no RJHAB measurements
were taken at these transects;A trial run which minimized the effect
of these two additional transects showed only very minor changes in WUA.
RESULTS
An IFG-2 IFIM model was run at Island Side Channel and hydraulic data
were collected at a side channel flow of 338 cfs (Appendix D).At
Trapper Creek Side Channel,hydraulic data for an IFG-4 IFIM model were
collected at flows of 16,32,and 389 cfs.Habitat data for the RJHAB
model were collected four times at Trapper Creek Side Channel and five -
times at Island Side Channel and the RJHAB models at both sites were
evaluated as "good"(Table 6)..
The modelled response of area at the Trapper Creek and Island side
channel sites to changes in discharge was almost identical for both the
IFIM and RJHAB modelling techniques (Appendix Figure C-1).Differences
in areas below the overtopping flow at Island Side Channel are probably
due to the IFIM not being able to model flows below 5 cfs while the
RJHAB WUA was measured at a flow of less than one cfs.Other differ-
ences are readily attributable to sampling error.Since juvenile
chinook and chum salmon are the two salmon species which make the
heaviest use of side channels for rearing,only WUA results from these
two species will be presented here.
At Trapper Creek Si de Channel,the shape of the WUA curves for both
speci es were basi cally the same for both mode"i ng methods (Appendix
Figure C-2).The RJHAB model appears to consistently underestimate the
amount of WUA in comparison to the IFIM model.The underestimation of
WUA by the RJHAB model leads to smaller habitat indices although the
shapes of the habitat index curves are similar for both techniques
(Appendix Figure C-3).
At Island Side Channel,on the other hand,WUAs from the two modelling
methods do not compare closely (Appendix Figure C-4).The chinook and
chum WUA response curves look more similar to each other than do the
modelling techniques.Peaks in WUA for the RJHAB model occur at approx-
imately 40,000 cfs while the IFIM model predicts a peak WUA at approxi-
mately 60,000 cfs.The IFIM model does predict a chinook salmon WUA of
6,230 ft 2 to 6,600 ftz at side channel flows of 6 to 11 cfs which
corresponds to the peak in the RJHAB model where a measurement was taken
at a side channel flow of approximately 10 cfs.
When habitat indices are calculated for both methods at Island Side
Channel,differences between the two techniques appear smaller (Appendix
Figure C-5).The RJHAB model shows a peak habitat index for chinook
salmon at approximately 39,000 cfs which the IFIM model would also show
at side channel flows of 6 to 11 cfs.Chum habitat indi ces for both
C-2
....,
-
Appendix Figure C-l.Comparison of site areas calculated with the RJHAB
and IFIM modelling techni~ues for the Trapper Creek
and Island Side Channel study sites.
C-3
TRAPPER CREEK SIDE CHANNEL
CHINOOK WEIGHTED USABl:..E AREA
-
11
10
9
8
x
loJ'fi'7
0."Zc;;:a 6~~m~5~....
4-
3
2
10
c 1Fl...
+RJHAB
30"50 70
....
CHUM WEIGHTED USABLE AREA
60
50
40
X
Wfi"0."~a
!<!l 30
!:::olDt:~--
20
10
0
10 3d SO 70~OUSandS~"
IIllAINSTE...DISC ARGE AT UNSHINE (cfs)
Appendix Figure C-2.Comparison of weighted usable areas calculated with
the RJHAB and IFIM modelling techniques for juvenile
chinook and chum salmon at Trapper Creek Side Channel.
1984.
C-4
-
TRAPPER CREEK SIDE CHANNEL
0.12
CHINOOK HA81TAT INDICES
..-""
0.11
0.1 a 1Fl..
0.09 +RJHAB
0.08
xw 0.07Cl
~
!<0.06
I-'
iii 0.05~
0.04
0.03
0.02
0.01
0
10 50 70
-CHUM HA81TAT INDICES
0.7
<~0.6
a 1Ft..
+RJHAS
0.5-xw
Cl 0.4~
!<~0.3
0.2
r
0.1
0
10 30 50 70
~usondS~MAINSTEM DISC GE AT UNSHINE (cfs)
-Appendix Figure C-3.Comparison of habitat indices calculated with the
RJHAB and IFIM modelling techniques for juvenile
chinook salmon at Trapper Creek Side Channel,1984.
C-5
ISLAND SIDE CHANNEL
CHINOOK WEIGHTED USABLE AREA
7
6
5
XW'ii'4o-g
~a
1-01~5 3aJf=~'"'
2
a IFlM
+RJHAB -
7030
0;----,----..----,-----,---.,----,-----1
10
CHUM WEIGHTED USABLE AREA
42
40
38
36
34-
l:k-32
0"z"O 30-6
~~28!::~~c 26
24-
22
20
18
16
10
a IFlt.l
+RJHAS
30 50 70
(Thousands)
t.lAlNSTEt.I DISCHARGE AT SUNSHINE (cfs)
Appendix Figure C-4.Comparison of weighted usable areas calculated with
the RJHAB and IFIM modelling techniques for juvenile
chinook and chum salmon at Island Side Channel,1984.
C-6
ISLAND SIDE CHANNEL
CHINOOK HASffAT INDICES
0.11
0.1
a FlM0.09
~+RJHA8
o.oa
)(w 0.070..-!;;
!<0.06
~0.05--0.04
0.03
.-
0.02
0.01
10 30 50 70
CHUM HA8lTATINDICES
a lFlM
+RJHA8
.-
30 50 70
(Thousonds)
MAlNSTEM DISCHARGE AT SUNSHINE (cfs)
Appendix Figure C-5.Comparison of habitat indices calculated with the
RJHAB and IFIM modelling techniques for Juvenile
chinook and·chum salmon at Island Side Channel,1984.
C-7
techniques decrease after overtopping although the RJHAB habitat indices
drop off more steeply.
DISCUSSION
The two modell"j ng methods compared very favorably at cal cul ati ng areas
within the two sites.The shape of the chum and chinook WUA and habitat
index responses at Trapper Creek Side Channel were very similar.The
RJHAB model consistently underestimated WUA in comparison to the IFIM
model.This is probably due to the RJHAB model not taking into account
the area between-the shoreline cell and the cell located one-third of
the way across the channel.This area was often marginal habitat with
barely suitable velocities.
At Island Side Channel,large differences in WUA can also be attributed,
in part,to the RJHAB model not taking into account peripheral marginal
habitat more than six feet from shore.This difference is also reflect-
ed in the habitat indices where the proportion of usable area drops off
more quickly for the RJHAB model.The differences in WUA below the
overtopping flow can be attributed to the fact that the IFIM model does
not run at flows less than five cfs while actual flows at discharges
below the overtopping one are less than one cfs (Quane et al.1985).
The effects of sampling errors in data collection on WUA estimates from
both the RJHAB and IFIM techniques are unknown.Since many more meas-
urements are taken for the IFIM,it should be less susceptible to
sampling errors.Because only one IFIM measurement was taken at Island
Side Channel at a flow of 338 cfs,however,the reliability of modelling
flows as small as 5 cfs is unknown.It seems reasonable to assume that
an IFG-4 model at Island Side Channel would have given somewhat differ-
ent results than did the IFG-2 model.The RJHAB model works well in
situations where the primary effect of discharge is due to backwater and
the IFIM model cannot be used or works poorly.
In summary,the RJHAB model generally gives lower WUA estimates than
does the IFIM methodology.Also peaks in WUA are often narrower for the
RJHAB model.Both models show the same general trends in the habi tat
indices for chum and chinook salmon although the RJHAB model is more
sensitive to increases in velocity and depth which decrease the habitat
indices more quickly.Since the habitat indices for both sites cal-
culated using both techniques are not appreciably different,analysis of
trends and optimal flows by use of habitat indices would lead to similar
cone 1us ions us i ng both methods.Compa ri sons of the I FIM wi th other
instream flow methodologies have also shown differences in output,and
no one method has yet been proven best (Annear and Conder 1984).
C-8
-
-
-
,...
LITERATURE CITED
Annear,T.C.,and A.L.Conder.1984.Relative bias of several fish-
eri es i nstream flow methods.North Ameri can Journal of Fi sheri es
Management 4:531-539.
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/OBS-82/26.
Hale,S.S.,P.M.Suchanek,and D.C.Schmidt.1984.Modelling of
juvenile salmon and resident fish habitat.Part 7 in D.C.Schmidt.
S.S.Hale,D.L.Crawford,and P.M.Suchanek (eds.;:-1984.Resi-
dent and juveni le anadromous fish investigations (May -October
1983).Alaska Department of Fish and Game Susitna Hydro Aquatic
Studies.Report No.2.Prepared for the Alaska Power Authority.
Anchorage,Alaska.
Marshall,R.P.,P.M.Suchanek,and D.C.Schmidt.1984.Juvenile salmon
rearing habitat models.Part 4 in D.C.Schmidt,S.S.Hale,D.L.
Crawford,and P.M.Suchanek (eds:T:1984.Resident and juvenile
anadromous fish investigations (May -October 1983).Alaska
Department of Fish and Game Susitna Hydro Aquatic Studies.Report
No.2.Prepared for the Alaska Power Authority.Anchorage,
Alaska.
Milhous,R.T.,D.L.Wegner,and T.Waddle.1981.User l s guide to the
physical habitat simul ation system (PHABSIM).Instream Flow
Information Paper 11.U.S.Fish and Wildlife Service
FWS/OBS-81/43.
Quane,T.,P.Morrow,1.Queral,T.Keklak,and T.Withrow.1985.
Technical memorandum in support of Task 14 (Lower River Resident
and Juvenile Anadromous Fish Studies)Alaska Department of Fish and
Game Susitna Aquatic Studies.Anchorage,Alaska.
C-9
-
-
I-
,,
APPENDIX D
HYDRAULIC MODELS FOR USE IN ASSESSING THE REARING
HABITAT OF JUVENILE SALMON IN SIX SIDE
CHANNELS OF THE LOWER SUSITNA RIVER
..-
.-
--
,....,
APPENDIX D
HYDRAULIC MODELS FOR USE IN ASSESSING THE REARING
HABITAT OF JUVENILE SALMON IN SIX SIDE
CHANNELS OF THE LOWER SUSITNA RIVER
By:
James Anderson,
Andrew Hoffmann,and
Jeffrey Bigler
of
Alaska Department of Fish and Game
Susitna River Aquati c Stud;es Program
Third Floor,Michael Building
620 East Tenth Avenue
Anchorage,Alaska 99501
ABSTRACT
Six side channels (Island,Mainstem West Bank,Circular,Sauna,Sunset,
and Trapper Creek)in the lower reach of the Susitna River were evalu-
ated using an Instream Flow Incremental Methodology (IFIM)physical
habitat simul ati on (PHABSIM)mode,l1 i ng approach to describe the effects
that site flow and mainstem discharge have on rearing juvenile salmon
habitat.These sites were thought to contain potential habitat for
rearing juvenile salmon and were chosen to range greatly in size,shape,
and overtopping discharge.
Si x hydraul ic simul ati on model s (either IFG-2 or IFG-4)were calibrated
to simulate depths and velocities associated with a range of site-
specific flows at the six modelling study sites.Comparisons b~tween
correspoDding sites of simulated and measured depths and velocities
indicated that the models provide reliable estimates of depths and
velocities within their recommended calibration ranges.
The recommended of ranges of mainstem Susitna River discharge over which
these models can hydraulically simulate the habitat of rearing juvenile
salmon are:Island Side Channel from 35,000 to 70,000 cfs mainstem
discharge;Mainstem West Bank Side Channel from 18,000 to 48,000 cfs;
Circular Side Channel from 36,000 to 63,000 cfs;Sauna Side Channel from
44,000 to 63,000 cfs;Sunset Side Channel from 32,000 to 67,000 cfs;and
Trapper Creek Side Channel from 20,000 to 66,000 cfs.
D-i
.....
TABLE OF CONTENTS
ABSTRACT ••••••.•.•
TABLE OF CONTENTS ..••.•••.•
LIST OF APPENDIX FIGURES ••.
LIST OF APPENDIX TABLES •.
D-i
D-ii
D-iv
.D-viii
INTRODUCTION •.....................................................
METHODS •••.••••
0-1
D-1
Approach •••.•-Analytical ..........................,..D-1
Study Site Selection.
......................--General Techniques for Data Collection.
D-3
D-4
Genera 1 Techniques for Calibration •.D-7
Island Side Channel •••
-
General Techniques for Verification •••••••••••..•.•••••.•..••
R-ESULT'S •••••••••••••••••••••••••••••••••••••••••••••••••••••••••••
.......................................
Site Description .......•...................
Calibration ••••••
Verification •••..
Application ...•.•••.•.....o •••••••
D-10
0-10
D-13
D-13
D-19
D-19
0-25
Mainstem West Bank Side Channel ..............'.D-26
Site Description ••••
Calibration ..•.
Verification .••
Application ••••••.•
Circular Side Channel •.
Site Description •.
Calibration •••••••
Verification .••...
Application •••••••
......................
..................
D--;i
0-26
D-32
D-35
D--35
D-39
0-39
0-39
D-46
D-49
TABLE OF CONTENTS (Continued)
Sauna Side Channel •••••••••
Site Description ...
Calibration ••••••••
Verification •••.•.•.••..•
Application ••••
0-51
0-51
0-51
0-60
0-62
Sunset Side Channel .••••••••••••••••••II •••••••••e •••••••••••tJo 0-64
Site Description ..••••••..•.
Calibration .•••••••••
Verification ••.•.
Application •.•...•••••
Trapper Creek Side Channel
Site Description .••.•••••
Calibration •••••.••••••••
Verification ••••••
Application •.•.•..
0-64
0-70
0-74
0-74
0-77
0-77
0-77
0-83
0-87
SUMMARy •.•••.•••.•
ACKNOWLEDGEMENTS.
•••••••••••O ••••••c."'•••••••••O.CiI ••••0-87
0-88
LITE RATURE CITED •.•••••s ••~• '••III ••Co ••••••••••••••••e e ••CI •ill _.'•••••••
D-iii
0-89
....,
LIST OF APPENDIX FIGURES
Appendix Figure Title
-
D-1 Location of the six IFG hydraulic modelling
sites in the lower Susitna River......................0-5
D-2 Overview of Island Side Channel (RM 63.2)••••.•••••••.0-14
0-3 Location of Island Side Channel study site
(RM 63.2}•.•...•....•.•.....,................•......•...0-15
0-4 Comparison of rating curves for Island Side
Channel transect 6 (Q site)(from Quane et
a 1.1985),-.. . . . . . . . . . . . . . . . . . . . . . . . . . .D-16
D-5 Cross section of transects 1,lA,2,and 3 at
Island Side Channel (adapted from Quane eta.'.1985)'..................0-17
"'"'
0-6 Cross section of transects 4,4A,5,and 6 at
Isl and Si de Channel (adapted from Quane et
al.1985)-- '0-18
--
0-7
0-8
Compari son of observed and predi cted water
surface profiles from cal ibrated model and
surveyed thalweg profile at Island Side
Channel (adapted from Quane et al.1985)••••••••••••••D-20
Application range of the calibrated hydraulic
model at Island Side Channel..........................0-22
(RM 74.4)•..••....•••.•.••.••••••••.•••~.•...••.•.'•. •.D-27
Location of Mainstem West Bank Side Channel
stu dy site (RM 74.4)••••••••••••••••••••••••••••. •. ••• D-28
Comparison of rating curves for Mainstem West
Bank Side Channel transect 1 (Q site)(from
Quane et ale 1985).......•............................D-29
D-9 Comparison of observed and predicted.veloc-
ities from the IFG-2 hydraulic model at
Island Side Channel,using two flows at the
transect 1 discharge site ••••••••••••••••••••••••.••.•D-23
Comparison of observed and predicted veloc-
ities from the IFG-2 hydraulic model at
Island Side Channel,using two flows at the
transect 6 discharge site •••••••~•••••••••••••••••••••D-24
Overview of Mainstem West Bank Side Channel
-
0-10
P"'"
D-11
0-12
~
0-13
I"""
D-iv
LIST OF APPENDIX FIGURES
Appendix Figure Title
0-14 Cross section of transects 1,2,and 3 at
Mainstem West Bank Side Channel (adapted from
Quane et ale 1985).................................•..0-30
-
Cross section of transects 3A and 4 at
Mainstem West Bank Side Channel (adapted from
Quane et ale (1985)••••••.••••~••.••.••••ee.e~e •••e •••0-31
Comparison of observed and predicted water
surface profiles from cal ibrated model and
surveyed thalweg at Mainstem West Bank Side
Channel (adapted from Quane et al.1985).•.••..••.••.•0-33
0-17 Application range of the calibrated hydraulic
model at Mainstem West Bank Side Channel ......•••.••.•0-36
0-16
0-15
0-18 Scatterplots of observed and predicted depths
and velocities from the calibrated IFG-4
hydraul ic model at Mainstem West Bank Side
Channel •.••.•......•.••..••••••••••.•..•.•••.••.•••...0-37
"""
0-19 Overview of Circular Side Channel (RM 75.3)•.......••.0-40
0-20 Comparison of ratin9 curves for Circular Side
Channel transect 4 {from Quane et al.1985)..•.••...•.0-41
0-21 Location of Circular Side Channel study site
(RM 75.3)..e _.........................................0-42
0-22 Cross section of transects 1,2,and 2A at
Circular Side Channel (adapted from Quane et
al.1985)"..e1c •••••,.,••e •••••t1 ••••••••••D-43
·0-23 Cross section of transects 3,4,and 5 at
Circular Side Channel ..•••.••••..••..•.••...•.••••.•.•·0-44
0-25
0-24 Comparison of observed and predicted water
surface profi 1es from cali brated model and
surveyed thalweg profile at Circular Side
Channel (adapted from Quane et al.1985}..••••.•..•.••0-45
Application range of the calibrated hydraulic
model at Circular Side Channel .•......•..••...•..•....D-48 -
0-26 Scatterplots of observed and predicted depths
and velocities from the calibrated IFG-4
hydraulic model at Circular Side Channel •.•••.•.•••..•0-50
0-27 Overview of Sauna Side Channel {RM 79.8).•.•.•...••...0-52
D-v
-
LIST OF APPENDIX FIGURES
Appendix Figure Title
-
0-28 Comparison of rating curves from Sauna Side
Channel transect 2 (from Quane et ale 1985).......•..•0-53
0-29 Locations of Sauna Side Channel study site
(RM 79.8)-..•..•.......•...••.....•........0-54
0-30 Cross.section of transects 1~2~3~and 4 at
Sauna Side Channel (adapted from Quane et al.
1985).... ................... . . . . . .................... ..0-55
0-31 Comparison of observed and predicted water
surface profi 1es from calibrated model and
surveyed thalweg profile at Sauna Side
Channel (adapted from Quane et al.1985)••..••••••••.•0-56
0-32 Application range of the calibrated hydraulic
mode 1 at Sauna Si de Channe 1.. . . . •. . . . . . •. . . . . . •••. . . . .0-61
0-33 Comparison of observed and predicted veloc-
ities from the IFG-2 hydraulic model at Sauna
Side Channel using two flows at the discharge
transect..••••.••. . ..••••••.•••.•. ..•.•.•••. •••••. . . . •0-63
D-34 Overview of Sunset Side Channel (RM 86.4)••••••••..•••D-65
0-35 Comparison of rating curves from Sunset Side
Channel at transect 1 (from Quane et al.
1985)~'0-66
0-36 Location of Sunset Side Channel study site
(·RM 86.9)•.•••••••••••••••.•._••••.••••••~.•••••.•••••_.0-67
D-37 Cross section of transects 0,1~2,and 3 at
Sunset Si de Channel (adapted from Quane et
a 1.1985).............................................D-68
0-38 Cross section of transects 4~5~and 6 at
Sunset Si de Channel (adapted from Quane et
al.1985)0-69
D-39 Compari son of observed and predicted water
surface profiles from calibrated model and
surveyed thalweg profile at Sunset Side
Channel (adapted from Quane et al.1985)••.•••.•••••••0-71
0-40 Application range of calibrated hydraulic
model at Sunset Side Channel ••••••••.••.•••••••.••••••0-75
D-vi
LIST OF APPENDIX FIGURES (Continued)
Appendix Figure Title
0-41
0-42
0-43
0-44
0-45
0-46
0-47
0-48
Scatterplots of observed and predicted depths
and velocities from the calibrated IFG-4
hydraulic model at Sunset Side Channel .....••••••.•••.0-76
Overvi ew of Trapper Creek Si de Channel (RM
91.6)0-78
Comparison of rating curves from Trapper
Creek Side Channel transect 4 (from Quane et
al.1985).,...,0-79
Location of Trapper Creek Side Channel study
site (RM 91.6).....G ••e •••••••e •••••••••••••••o ••••e.o 0-80
Cross section of transects 1~2~3,and 4 at
Trapper Creek Side Channel (adapted from
Quane et al.1985)........•.•.......•..........•.•....0-81
Comparison of observed and predicted water
surface profiles from calibrated model and
surveyed thalweg profi 1e for Trapper Creek
Side Channel (adapted from Quane et ale
1985).Cl •.•••••••••••.•••••••..•..••I1 0-82
Application range of the calibrated hydraulic
model at Trapper Creek Side Channel ..........•....•.•.0-85
Scatterplots of observed and predicted depths
and velocities from the calibrated IFG-4
model at Trapper Creek Side Channel ...••....•.......•.0-86
D-vii
.~,
-
~
)
-
-
-
LIST OF APPENDIX TABLES
Appendix Table Title
0-1 The six lower river IFG modelling sites with
corresponding river mile location.....................0-6
D-2 Percent cover and cover type categories...............D-8
0-3
0-4
Substrate classifications ......••.••••....o.0 •••••••0.0-9
The six lower river side channel IFG model-
ling sites with type of hydraulic model used,
dates calibrations flows measured,and
corresponding site specific flows and
mainstem discharges for the open-water period
in 1984.0 0 0 •••0 •~0 •••0 • 0 ••••••0 • 0 • 0 • 0 ••o ••00 0 0 •••0.• • •0-12
-
0-5 Comparison of field measured and model
predicted water surface elevations at the
calibration flow of 338 cfs for Island Side
Channel ..o.••0 •••0 • 0 0 00 0 0 • 0 0 •••0 • 0 • 0 • 0 • 0 0 0 0 00 • 0 ••0 ••0 •0-21
0-6 Comparison between observed and predicted
water surface elevations,discharges,and
velocities for 1984 Mainstem West Bank Side
Channel hydraulic model.0 0 •••0 ••0 •••••••0 0 ••o'0 0 o.0 o.0 0-34
0-7 The statistical results used to evaluate the
predictive ability of the four lower river
IFG-4 hydraul ic model so 0 0 00 .....0 ....0 0 ...0"0 0"0....0-38
0-8 Comparison between observed and predicted
water surface elevations,discharges,and
velocities for 1984 Circular Side Channel
hydrauli c model 0 •••0.0 • 0 0 • 0 • 0 • 0 0 ••0 0 •••0 0 •••0 ••• • • • • • •0-47
0-9 Comparison of fiel d measured and model
predicted water surface el evations at the
calibration flow of 52 cfs for Sauna Side
Channel .••••••..•.0"0 ••••••••••••••••••••••••o.••.•••0-58
0-10 The effects of the backwater at Sauna Side
Channel,information obtained from transect2.0 0...............0-59
0-11 Comparison between observed and predicted
water surface elevations,discharges and
velocities for 1984 Sunset Side Channel
hydraul ic model ••0 ••••••••••o..•....•...••..•.... •..•.0-72
0-12 Differences between stages of zero flow input
into the model and Quane et ale (1985)
thalweg survey at Sunset Side Channel •..•...•........•0-73
D-viii
LIST OF APPENDIX TABLES (Continued)
Appendix Table Title
0-13
0-14
Comparison between observed and predicted
water surface elevations,discharges,and
velocities for 1984 Trapper Creek Side
Channel hydraulic model •.........•••....••.......•.••.0-84
Summarization of the range of mainstem
discharges that the hydraulic models can
simulate for the rearing habitats of salmon
at the six lower river IFG modelling sites •....••.....0-87
D-ix
,~
--
-
!~
I"""'
i
INTRODUCTION
About 40%of the annual discharge of the lower Susitna River at the
Parks Highway bridge originates from the mainstem Susitna River above
the confluence of the Talkeetna and Chulitna Rivers (Acres 1982).Thus,
operation of the proposed hydroelectric project will alter the natural
flow regime of this lower river reach beyond the normal variations in
flow which occur naturally during the open-water season.
One of the predominant aquatic habitat types in this lower river reach
which may be affected by such flow alterations are side channels.Side
channel areas in this river reach currently provide habitat for rearing
juvenile salmon.The quantity and quality of juvenile salmonid rearing
habitat in side channels in this river reach is dependent on a multitude
of interrelated habitat variables,including water depth and velocity,
which are intimately related to mainstem discharge.
This appendix presents results of the physical habitat modelling simu-
lation efforts that Alaska Department of Fish and Game (ADF&G)Su Hydro
personnel conducted during the open-water season of 1984.The objective
of the study was to provide calibrated hydraulic simulation models for
selected lower river juvenile salmon habitat modelling study sites.The
approach of the study was to apply a methodology which used water depth
and velocity as the dominant hydraulic variables to quantify the
responses of rearing habitat to changes in site flow and mainstem
discharge.The methodology used was the system developed by the U.S.
Fish and Wildlife Service (USF&WS)Instream Flow Group (lFG)called the
Instream Flow Incremental Methodology (IFIM)Physical Habitat Simulation
(PHABSIM)modelling system (IFG 1980,Bovee 1982).The calibrated
hydraulic simulation models will be utilized to assess how site flows
and mainstem discharge affect juvenile salmon rearing habitat in side
channels of the lower Susitna River.
METHODS
Analytical Approach
A common methodology used for assessing habitat responses to flow
variations is the IFIM,·PHABSIM modell ing system.The IFIM,PHABSIM
modelling system is a collection of computer programs used to simulate
both the available hydraulic conditions and usable habitat at a study
site for a particular species/life phase as a function of flow.It is
based on the theory that changes in riverine habitat conditions can be
estimated from a sufficient hydraulic and biological field data base.
It is intended for use in those situations where flow regime and channel
structure are the major factors influencing"river habitat conditions.
The modelling system is based on a three step approach.The first step
uses field data to calibrate hydraulic simulation models to forecast
anti ci pated changes in physi ca1 habi tat va ri ab 1es important for the
species/life phase under study as a function of flow.The second step
involves the collection and analysis of biological data to determine the
behavioral responses of a particular species/life phase to important
physical habitat variables.This information is used to develop
0-1
weighted behavioral response criteria curves (e.g.,utilization curves,
preference curves,or suitability curves).The third step combines
information gained in the first two steps to calculate weighted usable
area (WUA)indices of habitat usability as a function of flow for the
species/life phase under study.
Hydraulic modelling is of central importance to the PHABSIM system.The
primary purpose of incorporating hydraulic modelling into the analytical
approach is to make the most efficient use of limited field observations
to forecast hydraulic attributes of riverine habitat (depths and veloc-
ities)under a broad range of unobserved streamflow conditions.
The IFG developed two hydraulic models (IFG-2 and IFG-4)during the late
1970's to assist fisheries biologists in making quantitative evaluations
of effects of streamflow alterations on fish habitat.The IFG-2 hy-
draulic model is a water surface profile program that is based on open
channel flow theory and formulae.The IFG-2 model can be used to
predict the horizontal distribution of depths and mean column velocities
at 100 points along a cross section for a range of streamflows with only
one set of field data.The IFG-4 model provides the same type of
hydraulic predictions as the IFG-2 model,but it is more strongly based
on field observations and empiricism than hydraulic theory and formulae.
Although a minimum of two data sets are required for calibrating the
IFG-4 model,three are recommended.Either model can be used to fore-
cast depths and vel oeiti es occurri ng ina stream channel over a broad
range of streamflow conditions.
The IFG-4 model,which is based upon a greater number of observed sets
of field data (i.e.flow levels),generally can be used to model a
greater range of flow conditions than the IFG-2 model.Additionally,
since the IFG-4 model is more dependent upon observed depths and veloc-
ities than the IFG-2 model,predicted depths and velocities can be
directly compared with the observed values.This comparison is a useful
tool for verifying the models.
Both models are most applicable to streams of moderate size and are
based on the assu~ption that steady flow conditions exist within a rigid
stream channel.A stream channel is rigid if it meets the following two
criteria:(1)it must not change shape during the period of time over
which the cal ibration data are collected,and (2)it must not change
shape while conveying streamflows with-in the range of those that are to
be simulated.Thus a channel may be "r igid"by the above definition,
even though it periodically (perhaps seasonally)changes course.
Streamflow is defined as "s teady ll if the depth of flow at a given
location in the channel remains constant during the time interval under
consideration (Trihey 1980).
In this analysis,all streamflow rates were referenced to the average
daily discharge of the Susitna River at the U.S.Geological Survey
(USGS)stream gage at Sunshine,Alaska (station number 15292780).This
location was selected as the index station primarily because it is the
gage located near the center of the river segment that is of greatest
interest in this particular analysis.The target mainstem discharge
range for data collection was from 12,000 to 75,000 cfs.
0-2
-
-
-
...."
.-,
i
"""
-
Site specific streamflow data collected during 1984 provided the basis
for correlating flow through the various study sites to the average
daily streamflow of the Susitna River at the Sunshine gage.Detailed
site specific channel geometry and hydraulic measurements provided the
necessary data base to calibrate hydraulic models for each study site.
Informati on for two .other physical habitat vari ab 1es,substrate and
cover,were also coll ected.Substrate was not incorporated into the
models at this time,but cover,an important variable in assessing the
habitat quality for most rearing salmon juveniles,was •
These data and hydraul ic model s make up the physical habitat component
of the PHABSIM analysis.For a given discharge of the Susitna River at
Sunshi ne,the flow through each study site can be determi ned and site
specific hydraulic conditions (velocity and depth)can be predicted.
The results based on velocity,depth,and cover may be used to forecast
the effects of mainstem discharge on the weighted usable area for
juvenile rearing salmonids of these modelled side channel habitats.
Study Site Selection
Two basic approaches are comnonly used for selecting study sites to be
evaluated using the IFIM PHABSIM modelling system:the critical and
representative concepts (Bovee and Milhous 1978;Trihey 1979;Bovee
1982).Application of the critical concept requires knowledge of a
stream's hydrology,water chemistry,and channel geometry in addition to
rather extensive knowledge of fish distribution,relative abundance,and
species-specific life history requirements.Criteria for application of
the representative concept are less restrictive,enabl ing this concept
to be used when only limited biological information is available or when
critical habitat conditions cannot be identified with any degree of
certainty.
In the critical concept,a study area is selected because one or more of
the physical or chemical attributes of the habitat are known to be of
critical importance to the fish resource.That is,recognizable phys-
ical or chemical characteristics of the watershed hydrology,instream
hydraulics,or water quality are known to control species distribution
or relative abundance within the study area.Because of this,an eval-
uation of critical areas will provide a meaningful index of species
response in the overall critical study area.
The representative reach concept acknowledges the importance of physical
habitat variables throughout the entire study stream for sustaining fish
populations.Thus,under this approach,study areas are selected for
the purpose of quantifying relationships between streamflow and physical
habitat conditions important for the species/life phase under study at
selected locations (representative reaches)that collectively exemplify
the general habitat characteristics of the entire river segment.
For thi s study,an adaptation of the representative concept was the
approach used to assess how mainstem discharges affect the rearing
habitat of juvenile salmon in side channel complexes.The six lower
river IFG study sites are most representative,morphologically,of
D-3
intermediate side channels and of the habitat type designation,sec-
ondary side channel as described by Ashton and Klinger-Kingsley (1985).
The results from these six IFIM-PHABSIM models are probably most appli-
cable to these types of areas in segments I and II of the lower Susitna
River.This segmentation of the lower river is also described in Ashton
and Klinger-Kingsley (1985).The six study sites were chosen by ADF&G
.Su Hydro Resident and Juvenile Anadromous (RJ)project personnel in
conjunction with ADF&G Su Hydro Aquatic Habitat and Instream Flow Study
(AH)project and E.Woody Trihey and Associates (EWT&A)personnel from
lower river side channels which met the following basic criteria:
1.The sites were chosen to range greatly in size,shape,and
overtopping discharge;
2.The sites were thought to contain potential habitat conditions
for rearing juvenile salmon;
3.The sites were judged by AH project and EWT&A personnel to be
readily modelled using the IFIM methodology;
4.The sites were accessible by boat at normal mainstem dis-
charges during the open-water season;and,
5.The sites were above Kashwitna landing and therefore much
easier to sample for logistical purposes.
The six sites chosen for modelling complemented other sites modelled
using another habitat model (see main text).All of the six sites were
side channels,the majority of potential habitat in the lower river is
composed of this habitat.Much of the other habitat ;s difficult to
model with the IFIM methodology because it is affected primarily by
mainstem backwater.Appendix Figure 0-1 shows the location of each of
the six sites selected for study,the corresponding river mile location
is presented in Appendix Table D-l.
General Techniques for Data Collection
A study reach was selected for detailed evaluation in each of the six
side channel sites.The length of the reach was determined by placing
enough transects within the area to adequately represent the major
macrohabitat types of the particular side channel area.
Transects were located within each study reach following field methods
described in Bovee and Milhous (1978)and Trihey and Wegner (1981),and
were located to facilitate collection of hydraulic and channel geometry
measurements of importance in evaluating flow effects on salmon rearing
habitat.Field data were obtained to describe a representative spectrum
of water depth and velocity patterns,cover,and substrate composition
at each side channel reach.
The number of transects estab 1i shed at the study reaches vari ed from
four to eight.The end points of each transect were marked with 3D-inch
steel rods (headpins)driven approximately 28 inches into the ground.
The elevation of each headpin was determined by differential
0-4
.....
i i
10
I
PARKS HIGHWAY BRIDGE:~~::::::::L::'USGS STATION (15292780)
rRAPP£R CREEK
SIO£CHANNEL
COOK INLET
-
-
.-
i
Appendix Figure 0-1.Location of the six IFG hydraulic modelling
sites in the lower Susitna River.
0-5
Appendix Table 0-1.The six lower river IFG modelling sites with
corresponding river mile location.
Side Channel Site River Mil e ~
Island Side Channel 63.2 ......,
Mainstem West Bank Side Channel 74.4
Circular Side Channel 75.3
Sauna Side Channel 79.8
~
Sunset Side Channel 86.9
Trapper Creek Side Channel 91.6 -
....
~I
-
D-6
I~
r
leveling using temporary benchmarks set at assumed elevations of 100.00
feet.
Cross section profiles at each transect were measured with a 1evel ,
survey rod,and fi bergl ass tape.Hori zonta 1 di stances were recorded to
the nearest 1.0 foot and streambed elevations to the nearest 0.1 foot.
Water surface elevations at each cross section in the study site were
determined to the nearest 0.01 foot by differential leveling or by
reading staff gages located on the cross section.
Streambed elevations used in the hydraulic models were determined by
making a comparison between the surveyed cross section profile and the
cross section profiles derived by subtracting the flow depth measure-
ments at each cross section from the surveyed water surface elevation at
each calibration flow (Trihey 1980).
A longitudinal streambed profile (thalweg profile)was surveyed and
plotted to scale for each modelling site (Quane et al.1985).
The water surface elevation at which no flow occurs (stage of zero flow)
at each cross section in the study site was determined from the stream-
bed profile.If the cross section was not located on a hydraulic
control,then the stage of zero flow was assumed equal to that of the
control immediately downstream of the cross section.
Discharge measurements were made using a Marsh-McBirney or Price AA
velocity meter,topsetting wading rod,and fiberglass tape.Discharge
measurements were made using standard field techniques (Buchanan and
Somers 1969;Bovee and Milhous 1978;Trihey and Wegner 1981).Depth and
velocity measurements at each calibration flow were recorded for the
same respective points along the cross sections by referencing all
horizontal measurements to the left bank headpin.
Cover and substrate values were also determined for each cell along
model 1ing transects..Methods described in Suchanek et al.(1985)were
used to code cover (Appendix Table 0-2).Substrate categories were clas-
sified by visual observation employing the substrate classifications
presented in Appendix Table 0-3.The distribution of various substrate
types was indicated on field maps.Substrates were classified using a
single or dual code.In those instances that a dual code was used,the
first code references the most predominant (i.e.,70%rubble/30%cobble
=9/11).
General Techniques for Calibration
The calibration procedure for each of the hydraulic models was preceded
by field data collection,data reduction,and refining the input data.
The field data collection entailed establishing cross sections along
which hydraulic data (water surface elevations,depths,and velocities)
were obtained at each of the different calibration flows.The data
reduction entailed determining the streambed and water surface ele-
vations,velocity distribution,the stage of zero flow for each cross
section,and determining a mean discharge for all the cross sections in
the study site.A model was considered calibrated when:1)the
D-7
Appendix Table D-2.Percent cover and cover type categories.
..,
I
Cover Type Code %Cover Code
silt,sand (no cover)1 0-5 .1
emergent vegetation 2 6-25 .2
aquatic vegetation 3 26-50 .3 """'i
1-3"gravel 4 51-75 .4 -3-5"rubble 5 76-100 .5
5 11 cobble,boulder 6 -
debris 7
overhanging riparian vegetation 8
undercut bank 9
-
0-8 -
Appendix Table D-3.Substrate classifications.
Substrate
Type
Part;cle
Size Classification
Sil t Silt 1
2
Sand Sand 3
4
Small Gravel 1/8-1"5
6
Large Gravel 1_31/7
8
Rubble 3-5"9
10
~,
Cobble 5-10"11
12
Boulder 10",13
~
-
0-9
majority of predicted water surface profiles were within ±0.05 ft of the
observed elevations and 2)the majority of predicted velocities were
within ±O.10 ft/sec of the measured velocities.A calibrated IFG-4
model gives velocity adjustment factors in the range of 0.9 to 1.1,and
relatively few velocity prediction errors.The velocity adjustment
factor is the ratio of the computed (observed)discharge to the predict-
ed discharge.
An IFG-2 model does not have velocity adjustment factors and is reviewed
with the observed data before it is considered calibrated.
For a more detailed explanation of the general techniques used for
calibrating the IFG-2 and IFG-4 models in the lower river see Hilliard
et a1.(1985).
General Techniques for Verification
The verification of how well each of these six hydraulic models simulat-
ed their respective site flows was performed by the hydraulic engineers
at EWT&A.The approach used to assess the qual ity of each model was
based on two levels of criteria.The first was a qualitative evaluation
of four separate sub-criteria.These sub-criteria were:
1.How well does the model conform to the IFG (Main 1978 and
Milhous et ale 1984)and EWT&A (Hilliard 1985)guidelines?
2.How well does the extrapolation range of the model conform to
the desired range?
3.Are the model s appropriate for the species and 1He stage
being .considered?
4.How well do the ranges of depth and velocities of the fore-
casted data conform to the ranges of depth and velocity of the
suitability criteria curves being considered based on a
~visual~evaluation?
After the first level of qualitative evaluation was performed,an
overall rating was given to the various segments of each model.The
ratings given were excellent,good,acceptable,and unacceptable.
Figures depicting these rating are presented for each site in the
results section.The second level of the verification process required
a statistical analysis to evaluate the models calibration.It was only
performed when the forecast capabilities of either the IFG-2 and IFG-4
model were not given an excellent rating in the level one evaluation.
For a detailed explanation of the verification analysis see Hilliard
(1985).
RESULTS
The results of the physical habitat simulation modelling studies are
presented below by study site.The six lower river side channel IFG
modelling sites with type of hydraulic model used,dates calibration
flows were measured,and corresponding site specific flows and mainstem
D-I0
......
-
~,
-
-
,....
-
.-
discharges for the open-water period in 1984 are presented in Appendix
Table 0-4.The following items are presented for each study site:(1)
a general site description,(2)a summary of data collected,(3)a
description of procedures used to calibrate the model,(4)the verifi-
cation of the model,and (5)the recommended application of the model
for each study site.
0-11
Appendix Table 0-4.The six lower river side channel IFC modelling sites with type of
hydraulic model used,dates calibration flows measured,and corre-
sponding site specific flows and mainstem discharges for the open
water period in 1984.
Mainstem
Date Site Discharge
Type of Calibration Specific at aSideChannelHydraulicFlowFlowSunshine
Site (RM)Model Measured (cfs)(cfs)""'"
Island Side Channel (63.2)IFC-2 July 25 338 56,100
Mainstem West Bank (74.4)IFG-4 September 2 450 32,000
September 20 310 30,500
September 25 6 19,600
Circular Side Channel (75.3)IFC-4 July 24 204 55,200
August 17 50 42,500
Sauna Side Channel (79.8)IFC-2 July 23 52 52,000 -Sunset Side Channel (86.9)IFG-4 July 22 496 57,800
August 17 127 42,500
Trapper Creek Side Channel (91.6)IFC-4 September 18 16 20,900 p:},
August 16 32 44,000
July 21 389 57,700
a Mainstem discharge determined from provisional USCS streamflow data from the stream gage ~
at Sunshine,Alaska (station number 15292780).
-
0-12
-
-
-
Island Side Channel (RM 63.2)
Site Description
Island Side Channel is located on the east bank of the main channel of
the Susitna River at river mile (RM)63.2 (Appendix Figure D-2).This
side channel is located downstream of a braided,vegetated floodplain
and is not directly connected to the main channel Susitna River.It is
approximately 0.7 miles in length with both the mouth and head portions
adjoining side channel networks.Breaching flows in this side channel
result from overtopping of the head by an adjoining larger side channel.
Prior to breaching,flow in the side channel is small with a series of
pools remaining (Quane et ale 1985).
The IFG modelling site at Island Side Channel was 735 feet long and
located in the lower portion of the side channel (Appendix Figure D-3).
The site generally consists of a pool-riffle-pool sequence.Based on
assessments by Quane et ale (1985),an area of backwater extends through
the study site to a point at least 1,100 feet upstream from the mouth of
the side channel at a nOQ.-breaching mainstem discharge of 35,000 cfs.
During mainstem discharges of 38,000 to 66,700 cfs,the area of back-
water extends throughout the study site.
The right bank of the study site is about five feet high,and the bank
is steep due to the effects of erosion.The primary riparian vegetation
along this.bank ;s alder.There are two side pocket areas along this
bank,which become slack water areas during higher site flows (400
cfs).In contrast,the left bank of the study site is a gently sloping
depositional bank.The riparian vegetation on this bank is sparse
consi sting primarily of shrub wi 11 ow.
Substrate at the study site consists primarily of gravels and rubbles,
with substrate changing to sand and silt in slackwater areas.The
thalweg gradient of the side channel is 15.6 ft/mile (Quane et ale
1985).From an evaluation of field observations,aerial photography,
and the stage/discharge relationship developed for this side channel,an
initial breaching has been estimated to occur at a discharge of 34,000
cfs (Quane et al.1985).
Based on a .review of available rating curves (Appendix Figure D-4)it
was determined that the hydraulics within this side channel are directly
controlled by mainstem discharges exceeding 35,000 cfs (Quane et ale
1985).A side channel streamflow of 43.5 cfs has been estimated to
occur at a mainstem discharge of 35,000 cfs (Quane et al.1985).
Eight cross sections were surveyed within this site during 1984 to
define channel geometry (Appendix Figures D-5 &D-6).The upper two
transects (5 and 6)were primarily located in pool habitat.Transects
4A and 4 primarily represent riffle habitat in the main portion of the
channel.Transect 4A was placed as a partial transect originating from
the right bank.It represents the 1arger of the two sl ack water areas
in this reach.The four downstream most transects are primarily in pool
type habitat.Transect 1A was also a partial transect,representing the
smaller slack water area along the right bank.
D-13
Appendix Figure 0-2.
e River Mile
01-(__-===2....1°PO t'f·
Overview of Island Side Channel (RM 63.2).
0-14
-
.....
.....
-
-
.~-1 -1 ._--,1 -1 1 }1 ,1 -,1
CI
I
f-I
U1
Appendix Figure 0-3.Location of Island Side Channel study site (RM 63.2).
-
!
-
-
-
-
t
1
ICIl...
Ic;i
l.:os
I~
I~
WSEL Et89 feet)
Q=43.5 ct.
CiK •10-0·1431 IIlSlL •II)C.U3t
pI iii .ft."
ISLAND SIDE CHANN,EL TR6
GAGE 63,256
i'4------~-..L-...._-.__-............__
I
3=o
..J
1.L.
0'
Wa::::>
(J)
<C
W
~
3S.aoo ~II !".JOG crl
IlSIL •MoO.'",1.DIlI •II
r'l._.'!II
o.c •10·n .tI"0..I.UJt
v'.too'"
IIOt CGOfIlOUU
17'-S,S 15._~ft.
WKL .Io-a·om ".lIot •II
,.2.0."
..J
W
(J)
3=
!
-...a _
MAINSTEM DISCHARGE.SUNSHINE (x 1000 cfs),
t•~I
81q,1on
':?I
';°1
~
en I
!WSEl=91.79 I
3=o
..J
Ll..Q=68.8owa::::>
(J)
<C
W
:::E
,..,.
0'1
'0....
'e __
MAINSTEM DISCHARGE,SUNSHINE (x 1000 cfs)
Appendix Figure 0-4.Comparison of rating curves for Island Side
Channel transect 6 (Q site)(from Quane et.al.
1985).
0-16
1.-)J ,1 1 )1 1 1 }.1 1
2802402001110'208040
CROSS SECTION IA
STATION 0+44
lOll
104
103..102
CD 101CD-100
z:1111
0 lie
i=117
<t 116
>.lIl1l1J
-I il4
lJJ 93
l1J 112
>In...110
<t 1111-I
IIJ 1111
I!:117
IIIl
1I11
02802402001110UO110
............--=I "6 eft
40
CROSS SECTION I
STATION Oi-OO
lOll
104
'03..102II>10'......
'00
z:9D
0 DB
~97
DIl
l1J 911
-I 114
l1J 113
W 92>III~110
-I IlII
l1J IlllI!:117
IlS
Illl
0
DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet)
DISTANCE FROM LEFT BANK HEADPIN (feet)
'---..........<J 338 eft
CROSS SECTION 3
STATION 2+55
2802402001110UO80·40
Cl
I lOll lOllt--'104 CROSS SECTION 2 104......STATION 1+12103 ~'03;:102 ..102..II>
CD 101 III 10'---100 -100
z:89 z:1111
Q 1111 0 1111...117 j:97
~116 ~91l
I&J IIl1 IIJ 1I11
-I 114 -I il4IIJ---558 e I&J113 113
lJJ 112 IIJ 92
>9'>III
j:110 i=80<t<t 1111 -I BII-I I&JlJJ1111I!:1111
I!:87 87
liB Illl
811 115
0 40 110 120 1110 200 240 2BO 0
DISTANCE FROM LEFT BANK HEADPIN «feet)
Appendix Figure 0-5.Cross section of transects 1,lA,2,and 3 at Island Side Channel (adapted from
Quane et al.1985).
2110240200111012011040
'.....1 ..........l 338 ell
CROSS SECTION 4A
STATION 4+31
105
104
103
Z 102
tlJ 101tlJ-100-
Z 1111
2 1111
~117«911>115W
..J 114
W 113
W 112
>111
~110«1111..Jw 1111
0::117
1111
115
0280240200111012011040
'---00::I sse.to
CROSS SECTION 4
STATION 3"62
10:1
104
103
Qi 102
tlJ 101--100
Z 1111
0 1111
5 117
>1I11
W 115
..J 114W
113
W 112>IIIi=«110
..J 1111
W 11110::117
1I11
115
0
DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet I
DISTANCE FROM LEFT BANK HEADPIN (fellt I
2110240200111012080
.....",sse .11
40
CROSS SECTION 6
STATION 7+35
0
I 105 10:1......104 CROSS SECTION 5 10400
103 STATION 5+65 103--..102 -;102•101 101••-100 .-100~
Z 1111 1111
ZQ11110 118
~87 i=117
«1I11 «1111>115 >115WW..J 114 sse .11 ..J 114
W 113 W 113
W 112 W 112
>111 >111
i=80 j:110«119 «119..J ..JW1111W 1111
0::117 0::117
1I11 1111
115 115
0 40 80 120 1110 200 240 2110 0
DISTANCE FROM LEFT BANK HEADPIN (fllllt)
Appendix Figure 0-6.Cross section of transects 4,4A,5,and 6 at Island Side Channel (adapted from
Quane et ale 1985)..
)J ~I J J J ~J J ~I )J J J J l I
Calibration
Calibration data available at the close of 1984 field season were
limited to that obtained for a side channel flow of 338 cfs (56,100 cfs
mainstem discharge)(Appendix Table D-4).As a result,an IFG-2 model
was used to forecast instream hydraulics based on this single·cali-
bration flow.The streambed profile,stages of zero flow,and observed
and predicted water surface elevations for this study reach are plotted
to scale in Appendix Figure 0-7.
The original field water surface elevations (WSEL's)were compared to
the model predicted WSEL's for the calibration flow of 338 cfs (Appendix
Table D-5).At transect lA,the original field WSEL was surveyed at
93.46 feet.In examining the WSEL1s of transects 1 and 2 (93.33 and
93.41 feet in elevation respectively),it was felt that an error in
surveying occurred at transect 1A.As a result,the WSEL for this
transect was lowered by 0.1 feet to 93.36 feet.For all other tran-
sects,the di fference between the fi e 1d WSEL I S and the model pred i cted
WSEL's for the calibration flow were 0.05 ft.or less.
The two partial transects (lA and 4A)which represent slackwater habitat
were extended out to the principal velocity corridor.This corridor is
where most of the flow in the channel occurs.In order to complete the
data sets for these two partial transects for use in the model,the
associated data from transects 1 and 4 were used.At partial transect
lA,the velocities were all negative.In order to use this information
in the model,these velocities were treated as positive,as it was felt
that the direction of the current would not influence the utilization of
this area by juvenile salmon.Only 6.5 cfs or about 2%of the water
flowed through this section.
Verification
Based on the first level of verification conducted by EWT&A,the model
does an excellent job of simulating hydraulics between 35,000 and 56,000
cfs mainstem discharge (69 and 416 cfs site flow)(Appendix Figure D-8).
Above 56,000 cfs,however,the simulated depth and velocity distri-
butions begin to deteriorate in quality.As a result,the model simu-
lations were rated good between 56,000 and 64,000 cfs (416 and 692 cfs
site flow),acceptable between 64,000 and 70,000 cfs (692 and 984 cfs
site flow),and unacceptable above 70,000 cfs mainstem.Below 35,000
cfs mainstem,the site flow was less than 5 cfs,and the model does not
simulate accurately below 5 cfs.
The velocity profiles produced by the IFG-2 hydraulic model for the two
flows,338 and 520 cfs,are compared to their associated observed
velocities at two transects (Appendix Figures 0-9 &0-10).The observed
and predicted velocities are in good agreement for both flows at tran-
sect 1.At transect 6 there is also good agreement between the observed
and predicted velocities at the 338 cfs flow.But at the 520 cfs flow,
from 85 to 140 feet,there is notable differences between the observed.
and predicted values.
D-19
ISLAND SIDE CHANNEL
Thol"'IV Proflll with Obllrvld and
Predlctld Watlr Surfaci ProW ..
~
....t
(Afprol.k".~
•
-
Thalw ••G.adl.nl'11I.e f••l/mll.
Oll ••n.d waler Surfac.EI••a!l"n
SI",ulal.d Wal.r Surfoc.El ••allon
E~lrapolaled W"ler Surface EI ••ollon
EI ••allan ai Zero Fl"w.
Thalweg Profile
,.....
,"••••Of e_.
&0+00.....
_._....._._._._._._._._._._._._._._.-.-._._._.-._._._._._._._._._._._._._,_,_,_,_,_'e."']
•...I.'••SS8 c'.(....pOlo'io..r'''W•."
_._._._._._._._._._._._._._._._._...:._._._._._._._._._._._.._._._._._._._._._._._..e.of.••••••11....."
._---:;:%-------------------------------------~
%~
"'0100 i..iii iii iii iii I II • , ,i , ,i , ,iii ii'iii iii Iii iii i ,j u iii i , •iii iii
,.
II
Cl
I
N j
,.
0
'"II
0;::c>II..........>,.
~....
IE ..
STllEloMBEO STATION I'.."
Appendix Figure 0-7.Compa~ison of observed and predicted water surface profiles from calibrated model
and surveyed thalweg profile at Island Side Channel (adapted from Quane et al.
1985).
J 1 J I t J I I J ,J }.1 I J J J
.....
,-
Appendix Table 0-5.Comparison of field measured and model predicted
water surface elevations at the cal ibration flow
of 338 cfs for Island Side Channel .
Transect
Water Surface Elevation (ft)
Field Model Predicted Difference
1
1A
2
3
4
4A
5
6
93.33 a93.46
93.41
93.44
93.48
93.52
93.56
93.55
93.33
93.36
93.36
93.40
93.46
93.50
93.53
93.56
0.00
0.05
0.04
0.02
0.02
0.03
0.01
-
-
a Water surface elevation reduced by 0.1 feet to 93.36 feet.
0-21
Application Range of the
at .Island
Cali brated Hydraulic Mode I
Side Channel
RM (63.2)
Site Specific Flow,cfs
o 8 38 115 211 545 984 1283
I I I 1
n ~
1 I I Io102030405060 10 15
o
I
N
N Mainstem Discharge at Sunshine Station,cfs x 1000
_Excellent
_Acceptable
II Good
D Unacceptable
Appendix Figure D-8.Application range of the calibrated hydraulic model at Island Side Channel.
J ~i J J )},J J )_I )t ,1 I.il
)))J })1 )1 )1 1 -l
ISLAND SIDE CHANNEL J Transec t 1 '
3
..Observed velocU tes (515 cIs)
•Observed veloct Hes (338 cIs)
.6.Predtcted veloct t tes (520 cIs)
D Predtcted veloct t tes (338 cIs)
-u
OJ
V'l................2Cot--
>-f--
.......-t
(J
CJ 0
I -IN
W W:>
1 ..
o-~f·90 1110 1~0 .1~OH_170 'I-I
DISTANCE FROM LEFT BANK HEADPIN (ftl
Appendix Figure 0-9.Comparison of observed and predicted velocities from the IFG-2
hydraulic model at Island Side Channel,using two flows at the
transect 1 discharge site.
ISLAND S I DE CHANNEL}.Transec t 6
5 i i
....Observed velocl Hes (543 cfs)
•Observed velod t tes (338 cfs)
.6.Predlcted ve lOel Hes (520 cfs)
[J Predkted veloct t tes (338 cfs)
4
0'ee,6,,,,I ,,,1
1
-U
OJ
lfI........3--"---
>-I-
t--t
L}20C)
I
N
---.J
.j:::o LU:>
Appendix Figure 0-10.Comparison of observed and predicted velocities from the IFG-2
hydraulic model at Island Side Channel,using two flows at the
transect 6 discharge site.
J I ).t J I J ,J ))J J el )o~I J
-
-
Application
For habitat simulation modelling purposes,the hydraulic simulation
model developed for Island Side Channel can simulate channel flows in
the mainstem discharge range of 35,000 to 70,000 cfs.
0-25
Mainstem West Bank Side Channel (RM 74.4)
Site Description
Mainstem West Bank Side Channel is located on the west bank of the main
channel Susitna River at river mile 74.4 (Appendix Figure 0-12).It is
approximately 2.2 miles in length.The mouth and two heads of this side
channel connect directly with the Susitna River.
The IFG modelling site in the lower portion of this side channel was 930
feet long (Appendix Figure 0-11).The study site is confined on the
west by a steep bank and on the east by a well vegetated island.The
portion of the side channel upstream of the study site is separated from
the mainstem by a network of side channels and well vegetated islands.
A minor channel is located within the study site on the east bank of the
side channel.During nonbreached conditions,the side channel primarily
consists of a series of pools and small riffles.Groundwater provides
the major contribution of flow prior to breaching of the head (Quane et
al.1985).
The two heads are both located approximately 1.5 miles upstream of the
study site (Quane et ale 1985).Breaching of Mainstem West Bank Side
Channel occurs when the mainstem overtops either of the two side channel
heads.The side channel has been estimated to be initially breached at
a mainstem discharge of 19,000 cfs (Quane et ale 1985).
Based on a review by Quane et ale (1985)of the stage versus mainstem
discharge rating curve (Appendix Figure 0-13),it has been determined
that at mainstem di scharges greater than 19,600 cfs,the hydraul i cs
within this side channel are directly controlled by mainstem discharge.
The site flow that occurs at 19,600 cfs was measured to be 5.7 cfs.
Hydraulic information was gathered from five transects (1,2,3,3A,4)
in the main channel and three transects (2A,3 in part,3B)in a minor
side channel of this study site (Appendix Figure 0-12).The corre-
sponding cross sections are presented in Appendix Figure 0-14 &0-15.
The two lower transects (1 &2)bisect primarily pool and run habitat,
the banks are gently sloping on both sides.On the upper three tran-
sects (3,3A,&4)the left bank consisted of an erosional bank and was
primarily bordered by alder.For modelling purposes,transects 3 and 3A
were ended on a finger-like gravel bar on the right bank which longitu-
dinally bisected the site with the main channel on the left and a minor
channel on the right which was free flowing at high flows,backwater at
median flows,and dry at low flows.This bar began downstream from
transect 4 and ended between transects 2 and 3.Transect 3A was placed
in order to obtain a better representation of the slow water debris-
strewn habitat along the left bank.The main channel habitat of these
three transects (3,3A,&4)consisted of run and riffle habitat.
Substrate at this site primarily consisted of rubble and cobble.The
thalweg gradient of the side channel is approximately 12.3 ft/mile
(Quane et al.1985).
D-26
""'"
)·····-1 )'1 1 1 1 ..)1 .,1 1
Cl
I
N
-..)
I
,~~...:+t
Appendix Figure 0-11.Overview of Mainstem West Bank Side Channel (RM 74.4).
C1
I
N
ex>
~'"
/),Staff Gage
Transect
o 260
I I
FEET
(Approximate Scale)
'r~~~
Appendix Figure 0-12.location of Mainstem West Bank Side Channel study site (RM 74.4).
)J J J )),J .)1 J ))I 'I
!
0•••10·Z••7)9 t",a..10''.0011_
.t •0."
Q=5.7 ch
on
lIllI;
I~
3t-,r-----.....--...L;:...--------
WSEL (+90 feet)
MAINSTEM WEST BAN K SIC TR 1
GAGE 74.451
3:o
..Ju.1
ow
ct::>
CIJ
<Cw
::aE
0••'10·,·J70S 'las 2.oem
.2 •0."
U.OOO 1 0 1 7'.700 .,.
lI$l\.•10·1•0617 o.0.J9lt •10
...z 11'04.
~-./COIITlllLl£O
~...U.JCIIlS 01 n.ooo.1s
...VSCL.10·t.lltO 00.Its1 •to
.•z •0.16
COIIYIllLLIP
1••IOt'S ,1 It.JOG .•''
t-..I_nOlI Oua.OPfO
OCT C_L(O
I',ran S ~~11.1110 ."
""[".."t"<1I(.ILOOfO
WSEL=92.85
~.~-MAINSTEM DISCHARGE,SUNSHINE (xlOOO cfsJ
f
-I~f
~I
if"II
1
-
~
Q)
.2
0
0).=.
..Jw
&-t
{/)
3:
-
-'0-~__._
MAINSTEM DISCHARGE,SUNSHINE ex 1000 cfs)
Appendix Figure 0-13.Compari son of rat;n9 curves for Mainstem West
Bank Side Channel transect l(Q site)(from Quane
et.a 1.1985).
0-29
.....-...__._--_.------------
105 105
104 CROSS SECTION I 104 ~CROSS SECTION 2
103 STATION 0+00 103 STATION 1+66-102 -102
.ii II
101 II 101II......100 -100
1111 Z 1111
Z liB 2 liB
0 117 I-117i=118 ~118ot>1111 \oJ 1111
\oJ 114 oJ 114oJ\oJ\oJ U 113
\oJ 82 \oJ 112
>III >III
5 110 !;i 110
oJ BII oJ BII
W BB W BB
It:B7 It:B7
B8 B8
B5 B5
0 100 200 300 400 0 100 200 300 400
DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet I
END
POINT
CROSS SECTION 3
STATION 5 +08
400300200100
450.11=====310.10
"v'«;.11\.....:;/
1011
104
0
103
I
-102
v)
-
a
II 101II...100
1111
Z liB011751111
>1111\oJ 114oJ
W 113
W 112
>III
i=110otBIIoJBBW
It:B7
Be
BII
0
DISTANCE FROM LEFT BANK HEADPIN (feet)
Appendix Figure D-14.Cross secti on of transects 1,2,and 3 at f~ainstem West Bank Side Channel
(adapted from Quane et ale 1985).
J J t J J I -I J ]t ,J I »J .J.-
)-]»j }1 J 1 ))1 1 -)-1 1
10:1 10:1
104 CROSS SECTION 3A 104 ]CROSS SECTION 4
103 STATION 5 +62 103 STATION 9+32..102 -102"..
<II 101 "101-Q)
100 I~-100
~
Z SS III
0 98 Z 1180t-117 ~j:97«98 98>450 eft ~95I&J 95 310 cft
-'94 l,/I&J 114I&J 6 cl.-'93 W 93
W 92 W 92>lit >III
t-90 t-110«
-'89 «89
I&J 88 -'88I&J0::87 0::87
88 88
85 ,.85
0 too 200 300 400 0 100 200 300 400
0
I DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet)w.......
Appendix Figure 0-15.Cross section of transects 3A and 4 at Mainstem West Bank Side Channel (adapted
from Quane et al.1985).
were
this
The
Calibration
Hydraulic data were collected for model calibration at three site flows:
6,310,and 450 cfs,the correspondi ng mean daily di scha rges for the
Susitna River were 19,600 cfs,30,500 cfs,and 32,000 cfs,respectively
(Appendix Table 0-4).Based on these data,an IFG-4 model was used to
forecast ·instream hydraulics.The streambed profile,stage of zero
flow,and observed and predicted water surface elevations for the study
reach are plotted to scale in Appendix Figure 0-16.All three data sets
were used to predict hydraulic information for side channel flows of 6
to 2,431 cfs (mainstem discharges of 18,000 to 75,000 cfs).
To evaluate the performance of the hydraulic model,observed and pre-
dicted water surface elevations,discharges,and velocity adjustment
factors were compared (Appendix Table 0-6).The 15 sets of observed and
predicted WSEL's for the five transects of the 3 calibration flows were
all within ±0.02 ft.of each other except for 2 sets which were within
±0.10 feet of each other.All the observed and predicted discharges
were within 10%of each other and all velocity adjustment factors were
within the good range of 0.9 to 1.1.Additionally,the stage infor-
mation of the model was compared to available rating curves (Appendix
Figure 0-13).
Transect (3A)was pl aced about 60 feet upstream from transect 3 to
represent the slackwater debris area along the left bank of the upper
portion of this study site.In order to complete this data set for
transect 3A for use in the model,the velocity information from transect
3 for the two site flows of 310 and 450 cfs were incorporated into
transect 3A cross sectional area and water surface elevations.After
incorporating this information into transect 3A,the discharge for the
310 cfs site flow,however,did not fall within 10%of the respective
discharge that was calculated at the discharge transect.As a result,
velocities for the 310 cfs site flow were adjusted upward by 17%.
At the low flow measurement of 6 cfs,the ve loci ty measu rements
made compl etely across transect 3A.The di scharge cal cul ated at
site was 18%higher than calculated at the discharge transect.
velocities at this transect were therefore reduced by 15%.
At transect 4 the water surface elevations were not similar across the
transect at the 6 cfs flow measurement.Therefore,a weighted average
water surface elevation was calculated for this transect.
At higher site flows several small side channel/backwater areas existed
which were not represented in the IFG-4 analysis.In order to evaluate
this potential habitat several transects were placed across one of these
areas,weighted usable area was to be determined by hand calculations.
However,this was not done because it was determined that this side
channel habitat was so small compared to the total area being hydrau-
lically modelled that it would not affect the total weighted usable area
response.
0-32
--
-
""'"
,~
-
-
j »)1 )1 ]-J }1 ))l'j )J
MAINSTEM WEST BANK SIDE CHANNEL
Thalwe"Profile with Observed and
Predlc;ted Water Surface ProfIJ e.
OOV~fi'////W "-AJr.;;;
EAST BANK MINOR CHANNEL
"..
..Thal••,Su,..,Dolo-841010
TIlal••,G.adi.nl-12.1I I••llmil.
Ob •••••d Wal.,So.Ia..EI••allon
SlllIulal.d Wal.,50,10••El ••olion
Exltapolohd Wal.,Su"o••EI ••ollon
EI ••allon of Z.,o Flo.
Thai••,P,ofll.w#
/
9 2~O
~
(AwOl 5eol.)
M41NSTEM WEST lANK SIDE CHANNEL
TIlANSEn u.TA""ntt J TAAHn:CT 58
I...1811 ,...---tU'i'i ,iii ii'iii'iii iii
~+()O 10+00
STREAMBED STATION (100'I
""'_=m~;;Ml'9'#-%-...,.,/;/~
AOftO OAGEl
74.4"--1
la"NUn I
IAO'-.'G"".
14.411
T".NIICT I lIllANI!IECT''.....'lel IA
IAOf 60 GAG.£.11•."154 __
1T••"lun.
STREAMBED STATION 11001'
Appendix Figure 0-16.Comparison of observed and predicted water surface profiles from calibrated model
and surveyed thalweg at Mainstem West Bank Side Channel (adapted from Quane et
ale 1985).
Appendix Table 0-6.Comparison between observed and predicted water
surface elevations,discharges,and velocities for
1984 Mainstem West Bank side channel hydraulic -
model.
Strea'mbed Water Surface
Station Elevation Oi scharge Velocity
Observed Predicted Observed Predicted Adjustment """"(ft){ft}(ft)(cfs)(cfs)Factor
~
0+00 92.85 92.86 6.0 6.3 1.005
1+66 92.86 92.87 6.9 7.2 .991
5+08 93.25 93.26 6.9 7.2 1.004
5+62 93.51 93.52 5.8 6.1 .996
9+32 95.06 95.06 5.1 5.4 1.013
Qo =~Qp =~......
I
0+00 94.62 94.61 312.8 315.7 1.030
1+66 94.64 94.64 301.3 307.5 1.024
5+08 94.85 94.86 306.4 318.2 1.007 -5+62 94.93 94.99 292.8 288.6 .993
Qo =301.0 Qp =308.0
~
0+00 94.97 94.98 460.4 457.0 .974
1+66 95.00 95.00 446.1 438.2 .975
5+08 95.19 95.18 470.6 455.2 .994
5+62 95.29 95.23 409.6 415.3 1.001 -
9+32 96.54 96.45 473.9 451.9 .969
00 =452.0 Qp =444.0
Qo is the mean observed calibration discharge.
~,
Qp is the mean predicted calibration discharge.
""""
-
D-34
1'fIlI!!lIiI\
-
-
..-
"""
Verification
Based on the first level of verification by EWT&A,the model does an
excellent job of simulating channel hydraulics between 18,000 and 21,000
cfs mainstem discharge (6 and 20 cfs site flow)(Appendix FigureD-17).
Above 21;000 cfs,simulated water surface profiles deviate somewhat from
field observations.As a result,the model was rated good between
21,000 and 28,000 cfs mainstem discharge (20 and 200 cfs site flow),and
between 28,000 and 34,000 cfs mainstem discharge (200 and 500 cfs site
flow)the model again was rated excellent.Two calibration data sets
were collected within this range.Above 34,000 cfs,the quality of the
hydraulic simulations begins to deteriorate as the slope of the site
flow versus WSEL relationship flattens as a result of channel geometry.
The deviation between the regression line developed within the model and
that of the rating curve developed independently for the site increases
with discharge until the model simulations are no longer acceptable.
The model simulations were rated good between 34,000 and 41,000 cfs (500
and 727 cfs site flow),acceptable between 41,000 and 48,000 cfs (727
and 1000 cfs site flow),and unacceptable above 48,000 cfs mainstem dis-
cha rge.
At the second level of verification there is good agreement between the
predicted and observed values of depth and velocity (Appendix Figure
0-18).At the higher velocities (>2.5 ft/sec)they begin to spread
apart though.In Appendix Table D-7 the results of the statistical
tests are shown.There is again good agreement shown between the
observed and predicted values for both velocity and depth.The index of
agreement (d)is almost one,the total root mean square error (RMSE)is
largely composed of the unsystematic RMSE,and the y-intercept (a)is
close to zero with a slope (b)of almost one.?
Application
For habitat simulation modelling purposes,the hydraulic simulation
model developed for Mainstem West Bank Side Channel can simulate channel
flows in the mainstemdischarge range of 18,000 to 48,000 cfs.
0-35
Application Range of the Cali bra ted Hydraul ic Mode I
at Mainstem West Bank
RM (74.4)
Site Specific Flow,efs
o 13 307 690 1080 1555 2118 2431
I I ~I I I
II III ,.
~H ,'\',,\,
I I ,I Io10203040506070 75
CI
I
W
0'\MainstemDischarge at Sunshine Station,cts x 1000
_Excellent
_Acceptable
•Good
D Unacceptable
Appendix Figure D-17.Application range of the calibrated hydraulic model at Mainstem West Bank Side
Channel.
)J ),.J t )J J I ,9 )_J !).J
)I 1 ))1 )1 ]J 1 1 -J ))
MAINSTEM WEST BANK SIDE CHANNEL MAINSTEM WEST BANK SIDE CHA~~NEL
4-
lio o~~
rP S~
rdb;.0 a
o~C Cb~a
o 0 a a
a 0
n..EI 0 0 (jlIo 0
~'tl rP
f 0 8
2 3
OBSERVED VELOCITY (IT/SEC)
Observed v.s.Predicted VelocitiesObservedV.s.Predicted Depths
4-I I 4-
00
a 0
3.5 -l oettt I 3.5
fO~'"
3 0 3Iiw,....III
t t'-'
I:2.5 '-'2.5
I-~0-
W i3a2g 2a~w
t>aa1.5 w 1.5
w l-
n:~
0-a
1 w
CJ n:
I Q.
w.....0:111:I
0.5
0III I I I
0 1 2 :3 4-0
OBSERVED DEPTH (IT)
Appendix Figure D-18.Scatterplots of observed and predicted depths and velocities from
the calibrated IFG-4 hydraulic model at Mainstem West Bank Side
Channel.
·~.J !J J ,J J )cl )I J I ~I ):I ill
r
-
.-
r-
I
Circular Side Channel (RM 75.3)
Site Description
Circular Side Channel is located on the west bank of the Susitna River
at river mile 75.3 (Appendix Figure 0-19).It is approximately 0.9
.miles long and is separated from the mainstem by a large well vegetated
island.An extensive backwater area occurs in the lower portion of the
study site.A network of small channels at the head provide mainstem
flow into the site after breaching.Prior to breaching,flow is greatly
reduced and the channel is composed of 1a rge pools connected by small
riffles (Quane et a1.1985).'
Breaching of Ci rcu1ar Si de Channel has been estimated to occur at a
mainstem discharge of 36,000 cfs (Quane et a1.1985).It has been
determined that the hydraulics within this side channel are governed by
mainstem discharge at mainstem discharges exceeding 36,000 cfs.The
site flow that occurs at this mainstem discharge is estimated to be 26.8
cfs (Appendix Figure 0-20)(Quane et a1.1985).
Based on assessments by Quane et a 1.(1985),backwater does not occur
during non-breachi ng mai nstem di scharges.At breaching rna i nstem.di s-
charges of 55,200 to 66,700 cfs,however,an area of backwater was found
to occur upstream to a point approximately 90 feet above transect 2A.
At a mainstem discharge of 42,500cfs,backwater has been determined to
extend slightly past transect 2.
The IFG modelling study site in the upper half of Circular Side Channel
is 820 feet (Appendix Figure 0-21).The thalweg gradient of this study
site is 14.3 ft/mi1e (Quane et aL 1985).Riparian vegetation along
both banks consists mostly of alder and cottonwood.Substrate within
the lower reaches of the Circular Side Channel site consisted predomi-
nately of silts,sands,and gravels changing to rubbles at the upper
reaches.Hydraulic information .was gathered from six transects estab-
lished at this study site (Appendix Figure 0-21).The channel is
relatively straight and the cross sections are generally box shaped in
confi gurati on (Appendi x Fi gures 0-22 &0-23).Transects 1 and 2 were
located in shallow backwater.Transect 2A was located in a transitional
area which became run habitat at higher flows.Transect 3 was located
in riffle habitat.Transect 4 was located in run habitat at the end of
a pool,transect 5 bisected this pool.
Calibration
Hydraulic data were collected at two calibration flows:50 and 204 cfs
(Appendix Table 0-4).Mean daily discharges for the Susitna River on
the dates that calibration data were collected at the Circular Side
Channel study site were 42,500 and 55,200 cfs.An IFG-4 model was used
to forecast instream hydraulics based on these two calibration flows.
The streambed profile,stages of zero flow,and observed and predicted
water surface elevations for the study reach are plotted to scale in
Appendix Figure 0-24.The two data sets were used to predict hydraulic
information from side channel flows of 6 to 733 cfs (mainstem discharges
of 25,500 to 75,000 cfs)..
0-39
C),
.p-o
I
.,,~..~j,t
Appendix Figure 0-19.Overview of Circular Side Channel (RM 75.3).
J l J ~J J j J J )")]t J j J )
-!
-
-
.....
o:J)
(j....,
:s:.
O'......
u.
0.wa:
::len«w:a:
~
O,c •10~.'.1064 0..4.lOIi5
..2 •a.."
!
.....
<II
(j....,
::
0.....
LI..
0.wa:
::len«
UJ
:E
•
'•.•~",',_.0''.0'/
,1 ..o~./
0=26.8 cis •--------t
I
~
I~tg
l~
C1RCULAR SIDE CHANNEL TR4
GAGE 75.354
........_.i'-------------""'----....,...;..
MAINSTEM DISCHARGE,SUNSHINE (x 1000 Of5)WSEL (+85 feet),.
~Q=36,000 cll
Y ',,""'O<LEG
J6 ..00CI ~Q.$.".100 C-'\
WSEL=90.15 I llS/:L •10'1.1619 QO.<55:•85
- -..lvO~"~1IQt COllY_LEO
",JOO ~o!16.000 cr,
lISE<•100·110J QO.lIlSS •os
,.:t •0.'1....I
Wen:s:
"il __
MAINSTEM DISCHARGE.SUNSHINE (x 1000 cfsJ
Appendix Figure 0-20.Compari son of rating curves for Ci rcul ar Si de
Channel Transect 4 (from Quane et.al.1985).
D-41
CJ
I
+:>
N
Appendix Figure 0-21.Location of Circular Side Channel study site (RM 75.3).
'uk,.".
,£It;__
U'~',
~(;,q.
q.'~:r,,,-\y~t-',u"""
))1 ~J l 1 }J J l ]J t ]))
]J J j 1 J ]1 1 1 )J }))1
CROSS SECTION I
STATION OtOO
'c ~204 ell
~.-----J !So 010
CROSS SECTION 2
STATION H98
,
2110240200
I
1601208040
105
104
103..102....101-100.....
Z 99
0 1111
~1J7«1J6>95I&J
..J 1J4I&J 93
I&J 1J2>IJ1
I-IJO«88..J
I&J 1111It:117
1111
1I11
021102402001801208040
105
104
103....102....101~.....100
Z 99
0 911
~97
~98
I&J 1J5
..J 1J4I&J 93
I&J 1J2>1J1
I-110«119..J
I&J 1111
It:117
1111
115
0
DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet)
Cl
I :~:1CROSS SECTION 2A-Pow
~103 STATION 2+65..102GO..101-.....100
Z 88
0 88~87«86>
1&.1 85
..J 841&.1 83
W 1J2>IJ1t=«80
..J 88
I&J 8110:87
811
115
0 40 80 120 1110 200 240 2110
DISTANCE FROM LEFT BANK HEADPIN (feet)
Appendix Figure D-22.Cross section of transects 1,2,and 2A at Circular Side Channel (adapted from
Quane et al.1985).
DISTANCE FROM LEFT BANK HEADPIN (feet)
2802402001601208040
--,..,204 cr..........~50c"...
CROSS SECTION 4
STATION 6+63
lOS lOS
104 CROSS SECTION 3 104
103 STATION 4+33 1-03....102 ....102GlGl..101 Gl 101......
100 ~100
Z 99 Z 99098098
I-97 I-97et«98>9a >ILl 9S ILl 9S..J 94 .J 94ILlILl
93 93
ILl 92 ILl 92>>j:91 204 c"~91
et 90 50 cfa 90
..J 89 ..J 89
ILl 88 ILl 880::0::117 87
116 86
liS liS
0 40 110 120 180 200 240 2110 0
DISTANCE FROM LEFT BANK HEADPIN (feet)
CROSS SECTION 5
STATION 8+20
lOS
104
CJ
103
I
-102
~
....
~
Gl 101Gl...100
99
Z 98097j:98~9SILl94..J
ILl 93
ILl 92
>91
5 90
89.J 118ILl
0::117
118
liS
0 40 110 120 180 200 240 2110
DISTANCE FROM LEFT BANK HEADPIN (feetl
Appendix Figure 0-23.Cross.section of transects 3,4,-and 5 at Circular Side Channel.
I I )J J
"'
J J j I -]I J J •_.~_J J
)1 'I )1 1 1 )~i 1 J )1 ]
CIRCULAR SiDE CHANNEL
Thaiwill Proflll wllh Ob.erVld and
Prldlclld Water SurfaCI Pro'''"
.~/
~
tAW::,·~~I.1
o
WI/)
Tholweo Grodl.nt·14.1 'ti.llmlle
Ob ....ed Wate,Su,'a..Ele.atlon
Simulated Wat.,Su,'a••Ele.ol'on
EIl,apalaUd Wale,Su,'o.e EI ••allon
Elavollon 0'Ze,o Flow
Thol.eg P,oflle
50+00u+oo10"'001&....00
t"ANSEcT S
10+00
,.AN,leT 4 "'..filIHCT I
I-811..-••1 -I
'''ANlE'eT -I
w
otoo
t •
..
a ..
I ;;
..j:::o ;;
U1 Z H
0;::..>t.W
..J
W
W::-tf;::..
..J
W
0:o•
..
ro
0';'00
STREAMBED STATION II..,)
Appendix Figure 0-24.Comparison of observed and predicted water surface profiles from calibrated model
and surveyed thalweg profile at Circular Side Channel (adapted from Quane et al.
1985).
To evaluate the performance of the hydraul ic model,observed and pre-
di cted water surface el evati ons,di scharges,and vel oci ty adjustment
factors were compared (Appendix Table 0-8).Because of the 2 cali-
bration flows only a 2 point rating curve was formulated.In evaluating
the performance of the model,observed and predicted WSELls and dis-
charges were the same because of this rating curve.Velocity adjustment
facto rs we re a11 with in the good ra nge of 0.9 to 1.1.Add iti ona 11 y,.the
stage information of the model was compared to the rating curves estab-
lished by Quane et al.1985 (Appendix Figure 0-20).
At the high flow measurement of 204 cfs,the original field measured
discharge at transect 2 was 34%lower than that calculated at the
discharge transect.In order to use this information in the model,the
individual velocity measurements were all adjusted upwards by 52%.Why
there was such a large discrepancy between flows at this particular
transect when the four other transect flow measurements were within 9%
of the discharge transect measurement is unknown.
At transect 5 there was a change in the channel cross section from when
the actual cross 'section survey was done and when the two calibration
flows were made.Between the cross section survey of September 5,1985,
and the two calibration flow measurements July 24 and August 17,1984,a
flood event occurred on August 26,1984.After this flood,the right
side of the channel at transect 5 was scoured out.In order to avoid
violating one of the underlying assumptions of the model,(i .e.,that a
rigid stream channel exists)the cross section determined from the two
calibration flows was used in the model.
During the 50 cfs calibration flow measurement a water surface elevation
was not surveyed for transect 5.In order to obtain a water surface
elevation for the model,a value was calculated from the average of the
depth measurements added to the corresponding cross section elevations
of the 50 cfs flow measurement.
Verification
Based on the first level of verification by EWT&A,the model does an
excellent job of simulating channel hydraulics between 39,000 and 57,000
cfs,mainstem discharge (38 and 213 cfs site flow).Above 57,000 cfs,
the simulated depth and velocity distributions begin to deteriorate in
quality.The model simulations were therefore rated good between 57,000
and 60,000 cfs (213 and 268 cfs site flow),acceptable between 60,000
and 63,000 cfs (268 and 334 cfs site flow),and unacceptable above
63,000 cfs mainstem discharge.Below 39,000 cfs,the model simulations
were also rated less than excellent as forecasted velocity and depth
distributions deteriorated in quality.The model simulations were rated
good between 36,000 and 39,000 cfs mainstem discharge (27 and 38 cfs
site flow)(Appendix Figure 0-25).Below 36,000 cfs mainstem (con-
trolling discharge),insufficient information is available to evaluate
the model.
At the second level of verification there is excellent agreement between
the observed and predicted velocities and good agreement between the
0-46
~,
-
-
-
-
Appendix Table 0-8.Comparison between observed and predicted water
surface elevations,discharges,and velocities
for 1984 Circular Side Channel hydraulic model.
....
Streambed Wa ter Su rface
Station Elevation Discharge Velocity
Observed Predicted Observed Predicted Adjustment
(ft)(ft) (ft)(cfs) (cfs)Factor
0+00 89.28 89.28 44.4 44.4 1.000
1+98 89.30 89.30 47.9 47.9 .998
2+65 89.41 89.41 56.0 56.0 1.000
4+33 90.20 90.20 43.7 43.7 1.000
6+63 90.60 90.60 50.9 50.9 .997
8+20 90.62 90.63 53.6 53.6 1.000
Qo =49.0 Qp =49.0
0+00 90.29 90.29 202.8 202.8 .998
1+98 90.27 90.27 203.1 203.1 .987
2+65 90.31 90.31 198.4 198.4 .999
4+33 90.66 90.66 176.9 176.9 .998
6+63 91.29 91.29 199.9 199.9 1.000
8+20 91.32 91.32 194.2 194.2 1.000
00 =196.0 Qp =196.0
Qo is the mean observed calibration discharge.
Qp is the mean predicted calibration discharage.
0-47
Application Range of the Calibrated Hydraulic Model
at Circular Side Channel
RM (75.3)
Site Specific Flow,cfs
o
o
10
2
20
12
30
43
40
118
50
268
60
537 733
70 75
Cl
I
~
00 Mainstem Discharge at Sunshine Station I cfs x 1000
mI Excellent II Good
-Acceptable y Unacceptable
Appendix Figure 0-25.Appli~ation range of the calibrated hydraulic model at Circular Side Channel.
_I !J I ],I J J .1 J J 1 I I I
-
.-
-
observed and predicted depths (Appendix Figure 0-26).The results of
the statistical tests also indicate good agreement between the predicted
and observed values for both velocity and depth (Appendix Table 0-7).
The index of agreement is near one,the total RMSE is mostly composed of
the unsystematic RMSE,and the y-intercept is close to zero with a slope
of almost one.
Application
For habitat simulation modell ing purposes,the hydraul ic simulation
model developed for Circular Side Channel can simulate channel flows in
the mainstem discharge range of 36,000 to 63,000 cfs .
0-49
CIRCULAR SIDE CH.ANNEL CIRCULAR SIDE CHAt'-JNEL
Observed vs.Predicted Depths Observed vs.Perdieted Velocities
4.I 4
(]
(]0
3.5 -l (](]3.5
~Da (]~...rP
3 u 3 (]
"
(]~D~BtP
W
tJ (I)cf"w t tp/LL
'--'2.5 Ire ~£J 'V 2.5
:r o aD ~(][Jl-I!D-ow20 20..J
0 W ,
W >
I-001.5 w 1.5
0 I-
w ·2
It 0
D-1 wC1ItID-Ol
C>(]
0.5
o pw=I oV (]
I I I I I I I
0 1 2 :5 4 0 1 2 3 4
OBSERVED DEPTH (FEET)OBSERVED VELOCITY (FT/SEC)
Appendix Figure 0-26.Scatterplots of observed and predicted depths and velocities from
the calibrated IFG-4 hydraulic model at Circular Side Channel.
,1 1 I _J J '1 )l I J )J I })I
-
,.....
-
,.....
!
,-
.....
I
Sauna Side Channel (RM 79.8)
Site Description
Sauna Side Channel is located on the west bank of the Susitna River at
river mile 79.8 (Appendix Figure D-27).It is approximately 0.2 miles
long.Both the mouth and head of the side channel are connected to a
larger side channel of the mainstem Susitna River.For the most part,
the side channel is confined on the west side by a high bank ~nd on the
east by a large sparsely vegetated gravel bar.A smaller side channel
enters just below the head of Sauna Side Channel on its west bank.This
side channel conducts flow to the study site during high mainstem
discharges,but dewaters before the head of Sauna Side Channel becomes
unbreached.Breaching flows result from overtopping of the side channel
that adjoins the head on the east bank of Sauna Side Channel.Prior to
breaching,the channel is composed of two large interconnected pools
whose water levels are maintained from ground water seepage originating
from the vicinity of the head.An extensive log jam at the head of
Sauna Side Channel influences the flow into this side channel.
Based on field observations and stage/discharge relationships,the
mainstem discharge estimated to initially breach Sauna Side Channel was
37,000cfs (Quane et ale 1985).A controlling discharge of 38,000 cfs
was determined for this side channel also based on this stage/discharge
relationship.A side channel flow of22.5 cfs was estimated to occur at
the 38,000 cfs mainstem discharge .as derived from the stage versus
streamflow rating curve (Appendix Figure 0-28).Quane et ale (1985)
determined that backwater does not occur in Sauna Side Channel during
non-breaching mainstem discharges.During breaching discharges of
54,600 to 56,700 cfs,however,backwater was observed to occur through-
out the Sauna Side Channel study site.
The IFG modelling site,located approximately 2,000 feet from the mouth
of this side channel,was 480 feet long (Appendix Figure 0-29).The
thalweg gradient at this site is 10.4 ft/mile (Quane et ale 1985).
Substrates throughout this site consist primarily of sands and silts.
The water is slow moving with velocities usually less than 1.0 ft/sec.
The left bank at the site is an erosional bank with a height exceeding
fi ve feet;ri pari an vegetati on along thi s bank consi sts of alder and
birch.In contrast;the left bank isa depositional bank with no
riparian vegetation.
Four transects were located within this study site (Appendix Figure
0-30).Transects 1 and 2 were located in shallow pool habitat whereas
transects 3 and 4 were located in deeper pools.
Calibration
Hydraulic data were collected at a calibration flow of 52 cfs corre-
sponding to a mainstem discharge of 52,000 cfs (Appendix Table 0-4).
Based on this single calibration flow,an IFG-2 model was used to
forecast instream hydraulics of this study site.The streambed profile,
stage of zero flow,and observed and predicted water surface elevations
for the study reach are plotted in Appendix Figure 0-31.This data set
D-51
FEET
(Approximate Scale)
-
-
RM79Ef)
500
!
River MUe
a
!
Appendix Figure 0-27.Overview of Sauna Side Channel (RM 79.8).
0-52
!
.....
<Il
:3
3:o
..J
I.L.
o
W0:.
:::>en«w
::!i:
!!
0'IC III 10-••1214 0...Z.:.""
,.!011 0.10
3:o
..J
I.L.C
o
W
0::::>en«w
::!i:
0...166 •11"11l5lL •lSI t._
"I •o.•a
Q =22.5 ctl
SAUNA SIDE CHANNEL TR2
GAGE 79.8S2
COIT1IGll [~
1I.GOCI ~0 S 17 .100 cf'
"'0.•1~·1.7_O·uSt •IS
.t 6.11
lIlT.a-r....lEl
15._~6 ~la.a c"
'"'£l1IJI.T 1111 1100.00£6
-;,._....,,10-.1.-.
MAINSTEM DISCHARGE,SUNSHINE (x 1000 cfsJ WSEL ("85 feet)
It +
I
I
--
-
-~03 _,-m
MAlNSTEM DISCHARGE.SUNSHINE (x 1000 cfsJ
Appendix Figure 0-28.Comparison of rating curves from Sauna Side
Channel transect 2 (from Quane et.at.1985).
0-53
Appendix Figure 0-29.Location of Sauna Side Channel study site (RM 79.8).
0-54 -,
-~J 1 }J ))})}1 )-J ]-}J »
r----.
20 40 150 110 100 120 140 160 1110 200
CROSS SECTION 2
STATION It81
'\,'-&2 Of.
10~
104-10;'
011 102......101-
Z
100
~1111
I-lUI
CI 117
>1115
ILl
.J II!!
ILl 114
ILl 113
>92
I-91
<r 110
.J 1111W
0::811
117
118
II!!
02040601101001201401110 1110 200
CROSS SECTION I
STATION 0+00
10~
104
103-102;101...-100
99
Z 11110117~1111
II!!I.Ll 114.J
I.Ll 113
ILl 112
>111
S 110
1111.J 1111ILl
0::117
116
II~
0
DISTANCE FROM LEFT BANK HEADPIN (feeO DiSTANCE FROM LEfT BANK HEADPIN (feet I
0
I 10!!10~
U1 104 CROSS SECTION 3 104U1103STATION3t71103--011 102 ;102..-101 ..101-100 -100-Z 1111 Z 1111211110116
I-117 I-117~118 <r 1115
ILl II!!>II!!.J ILl
W 114 .J 114
113 ILl 113
W 112 ILl 112>111 >11152oliI-110 I-110<r
.J 1111 <r 1111
ILl .J
0::1511 W 116
117 0::117
118 116
II~II~
0 20 40 80 110 100 120 140 1110 1110 200 0
DISTANCE FROM LEFT BANK HEADPIN (feet)
CROSS SECTION 4
STATION 4-t81
20 40 60 110 100 120 140 160 1110 200
DISTANCE FROM LEFT BANK HEADPIN (feet)
Appendix Figure 0-30.Cross section of transects 1,2,3,and 4 at Sauna Side Channel (adapted from
Quane et al.1985).
SAUNA SIDE CHANNEL
Thellwell Profile with Ob..rved ond
Predicted Wot.,Sur'oCi Pro'll ..
..
•
~
Thai...Sarv.,Dot.·84 100'
Thai •••Gracll.nt-10.4 ,..1111111.
OII..rv.cI Wal.r Sar'ace EI.vatlon
SllIIalat,d Wat.r Sur'ac.EI.vaIlOll
Ellropo'at.d wat.r Sur'"c,EI..all"n
Elavc"'an ,,'Z.ro F'o.
Tho'.,,,Pr,,'lI.
..
••,_._.._._._._..-._._.•._._._._._.._.._.._._._._...._._.-._11--.11:1:lE.ir.p.JoIIOlll ,e••._.._.._.._._._._._._._._.._._._._._._._._._._._._._._0_.-.-lec,..t .,..,.,,110 ••4.1
"-'~---~.~______~I':~~
..
'"II>5
.J
'""'lI:
CJ
t
U"l
('7)
II ,IIIA"'E"•TIllAMUCT I tu.aU,s JRA.'II;'•
I.......10"'00
801 ,iii i i • ; :i ;iii iii i I:,ii'iii I i 81:,ee;iii iii ,i"'},ii,I Ii'i D iii i,'.
0+00 8+~
STREAMBEO STATION IIllil
Appendix Figure 0-31.Compari son of observed and predi cted water surface profil es from calibrated model
and surveyed thalweg profile at Sauna Side Channel (adapted from Quane et ala 1985).
J J }j J !J ]J 1 ].D 1 1 1
-
-
-
was used to predict hydraulic information from side channel flows of 5
to 93 cfs (mainstem di scharges of 21,000 to 75,000 cfs).To eval uate
the performance of the hydraul ic model,observed and predicted water
surface elevations were compared (Appendix Table 0-9).Additionally,
the stage information of the model was compared to the rating curves
established by Quane et al.(1985)(Appendix Figure 0-28).
It was difficult to calibrate hydraulic'information at this site because
very limited field data were available.A site flow versus WSEL rating
curve could only be developed for transect 2 (Appendix Figure 0-28).
The IFG-2 model is essentially a water surface profile model and a
critical variable for calibrating it,is the water surface elevations of
simulated flows.Data,however,;s only available for transect 2 and
not for any of the other three transects.The actual velocity measure-
ments from other measured flows at the discharge transect,however,can
be compared to the model predicted velocities for those same flows.At
the discha,rge measurement for transect 2,however,there were only two
flows that were far enough away from the 52 cfs measurement to be usable
(38 and 68 cfs).Thus,the information available to hydraulically
calibrate the IFG-2 model for this site consists of the water surface
elevations and velocity measurements ,for all four transects at the
calibrating flow of 52 cfs,and water surface elevations and velocities
for the two other site flows of 38 and 68 cfs at transect 2.
This site is influenced by backwater and the effects are more pronounced
at the 68 cfs flow.From the field data,the observed top width is
greater by 20 feet,the water surface elevation is 0.93 feet higher,and
the average velocity is 0.20 ft/sec slower than predicted by the model.
At the 38 cfs flow,the effect seems to have reversed,with the observed
widths being similar,the WSEL 0.08 feet lower,and the average velocity
0.09 ft/sec faster than predicted by the model (Appendix Table 0-10).
In the calibration process,the original field WSEL was reduced by 0.1
feet.This adjustment was made in order to obtain water surface ele-
vations that agreed more closely to the lower site flows.It was felt
that this adjustment would make the model,in terms of predictability,
more sensitive at the lower site flows.By reducing the WSEL of tran-
sect 1 by 0.1 feet,the difference between the field and the model WSEL
at the 38 cfs flow was reduced from 0.18 feet,when the calibration
discharge WSEL was 90.71,to 0.08 feet,when the calibration discharge
WSEL was 90.61 feet (Appendix Table 0-10).
As a result of a flood on August 26,sediments were deposited in the
study site resulting in changes "in all the cross sections derived from
the calibration flow on July 23.As a result,the cross sections
obtained during the September 15 survey were used in the model until the
water's edge of the calibration flow was reached,then the cross
sections from the calibration flow were used.
When measuring the velocities and depths at each of the transects,the
discharge calculated at transect 4 was 16%lower than the 52 cfs site
flow calculated at the discharge transect.In order to utilize this
information in the model,the velocities were adjusted upwards by 16%.
0-57
Appendix Table 0-9.Comparison of field measured and model predicted
water surface elevations at the cal ibration flow
of 52 cfs for Sauna Side Channel.
*Field water surface elevations were reduced by 0.1 feet.
0-58
.....
-
-
-
~,
-
)1 I 1 1 ]--..1
UI<A}T /F-Aut]11)
4/19/85
ANDY/Tables
1
Appendix Table 0-10.The effects of the backwater at Sauna Side Channel,information obtained from
transect 2.
Top Width (ft)
Field Model
Average Velocity (ft/sec)
Field Model
o
I
tJ'1
t..O
Original Modified
Site WSEl (ft)WSEl (ft)
Flow (cfs)Field Model Field Model
68 91.85 91.06 91.85 90.92
52 a 90.n b 90.74 90.61 c 90.62
38 90.24 90.42 90.24 90.32
a Calibration flow
b Original field WSEL input into model
c Field WSEl reduced by 0.1 ft
77 .0
53.5
50.5
55.0
53.0
52.0
0.32
0.53
0.51
0.52
0.49
0.42
No stage-site flow rating curve was developed for transect 1.When
inputting other flows into the model,the IFG-2 requires either the
associated WSEL for this flow or the slope.Because the WSEL could not
be obtained for other flows at this transect,a slope value of 0.00005
was input instead.This value was generated by the model from transect
1 at the calibration flow of 52 cfs.
Verification
The dominant influence of backwater on channel hydraulics makes the site
a poor candidate for application of IFG-2 modeling techniques.However,
because only one data set was collected,application of the IFG-4
hydraulic model was not possible.
Based on the first level of verification by EWT&A,the IFG-2 model for
this site does an excellent job of simulating channel hydraulics between
48,000 cfs and 58,000 cfs mainstem discharge (34 to 52 cfs site flow)
(Appendix Figure 0-32).Within this range,predicted WSEL's,depths,
and velocities are in close agreement with field information (evaluated
at 38 cfs by discharge.measurement made by Quane et a 1.(1985).The
predictive capability of the model within this range provides evidence
that the backwater influence within the study site is lessening with
decreasing discharge.
Below 48,000 cfs mainstem,there is increasing disagreement between the
WSEL IS predicted by the model and those extrapol ated from the rati ng
curve.At a 23 cfs site flow,the difference in predicted WSEL between
model and rating curve equation has increased to approximately one foot
at transects 1 and 2.Although there is evidence that suggests that the
model may be a more accurate predictor of WSEL's than the rating curve
equations below 48,000 cfs ma;nstem,insufficient information exists to
resolve the difference with confidence.Since depths become shallow
within this range,predictive errors in WSEL can result in significant
errors in predicted depths and velocities.For this reason,the recom-
mended extrapolation range ;s limited below 48,000 cfs.
Above a 48,000 cfs mainstem discharge,there is increasing,disagreement
between the WSEL's predi cted by the model and those observed i·n the
field.One of the premises of the hydraulic theory that is the basis of
the IFG-2 model is that the water surface profile of the study reach is
controlled by its slope.This premise is violated when the water
surface profile is influenced by mainstem backwater.From examination
of discharge measurements made at 48 and 68 cfs it is apparent that the
influence of backwater is increasing with stage above 58,000 cfs
mainstem.
Overall,the recommended extrapolation range is limited above 58,000
cfs.The model simulations were rated excellent between 48,000 and
58,000 mainstem discharge (34 to 52 cfs site flow).Good between 46,000
and 48,000 (31 to 34 cfs)and from 58,000 to 60,000 cfs (52 to 58 cfs).
Acceptable between 44,000 and 46,000 cfs (28 to 31 cfs)and 60,000 to
63,000 cfs (58 to 62 cfs).The model was rated unacceptab 1e below
44,000 cfs and above 63,000 cfs mainstem discharge (Appendix Figure
0-32)•
0-60
-
-
-
)1 --1 1 I )1 J -1
Application Range of the Calibrated Hydraulic Model
at Sauna Side Channel
RM (79.8)
Si te Specific Flow,cfs
5 12 22 37 56 80 93
I I I I I I
I==f.
I§t
I ====~=,I I I I Io10203040506070 75
CI
I
O'l
I-'
Mainstem Discharge at Sunshine Station,cfs x 1000
IillI Excellent Iii Good
-Acceptable D Unacceptable
Appendix Figure 0-32.Application range of the calibrated hydraulic model at Sauna Side Channel.
The velocity profiles produced by the IFG-2 model at transect 2 were
compared to the observed velocities at flows of 38 and 68 cfs (Appendix
Figure 0-33).Because this site is primarily a backwater area and the
IFG-2 hydraulic model is not a backwater model it was thought that
calibrating the model to more accurately predict at the lower flows
woul d be more criti ca 1 than at the hi gher flows.Thus at the 38 cfs
flow there is found a better correspondence between the observed and
predicted velocities.At the 68 cfs flow the backwater becomes more
apparent.A majority of the observed velocities are lower than the
predicted velocities and many of these values are lower than individual
38 cfs flow velocities.Because of the overa,.l low velocities,1.0
ft/sec,it was felt that this was the best compromise in applying this
model to the Sauna Side Channel site.
Application
For habitat simulation modelling purposes the hydraulic simulation model
developed for Sauna Side Channel can simulate channel flows in the
mainstem discharge range of 44,000 to 63,000 cfs.
D-62
~1
-
-
"""
~!
..,.
).1 .1 ).-1 ]_._))~)-]l-}1
SAUNA SIDE CHANNEL}Transect 2
..Observed ve lod ties (68 cts)
•Observed velod ties (38 cts)
A Predicted velocl ties (68 cts)
CI Predlcted velod t les (38 cts)
......
o .20 30 40
DISTANCE FROM LEFT BANK HEADPIN (ft)
Appendix Figure 0-33.Co~parison of observed and predicted velocities from the IFG-2
hydraulic model at Sauna Side Channel using two flows at the
discharge transect.
Sunset Side Channel (RM 86.9)
Site Description
Sunset Side Channel is located on the east bank of the Susitna River at
river mile 86.9 (Appendix Figure 0-34).It is approximately 1.1 miles
long and is separatea from the main channel of the Susitna River on the
west by a network of vegetated islands and side channels.The channel
is confined on the east by a high cut bank.Prior to breaching,the
side channel is composed of a sequence of pools and riffles.During
this period,flow is maintained in the main channel by groundwater
seepage and upwelling.After breaching,flows up to 3,900 cfs have been
measured (Quane et ale 1985).
Breaching of Sunset Side Channel results from the direct overtopping of
the head of the side channel by the mainstem Susitna River.Based on
assessments by Quane et ale (1985)the side channel initially breached
at 31,000 cfs and controlled at a mainstem discharge of 32,000 cfs.
The associated site flow at the controlling discharge has been esti-
mated to be 45.8 cfs while a flow of 41.1 cfs is derived from the flow
versus mainstem discharge rating curve (Appendix Figure 0-35).
Based on assessments by Quane et a1.(1985)a backwater area does not
occur in this side channel during unbreached conditions.But at breach-
ing mainstem discharges ranging from 56,000-66,700 cfs,an area of
backwater was observed to extend upstream approximately 1,100 feet to a
point between transects 1 and 2.
The IFG modelling site within Sunset Side Channel was located in the
lower portion of the side channel and was 1410 feet long (Appendix
Figure D-36).Hydraulic information was collected from seven transects
within this study site (Appendix Figures 0-37 &0-38).The channel
within the study site has a gradual bend.The right bank from transects
2 to 6 is erosional,becoming less steep and depositional at transects 0
and 1.On the left bank,transects 2 through 6 are primarily deposi-
tional in nature.In the areas of transects °and 1,the left bank
becomes steep and erosional.At transect 2 on the left bank a small
dewatered channel enters but water was never observed running in it
(Appendix Figure 0-36).The thalweg gradient within the study site is
9.5 ft/mile (Quane et ale 1985).·Riparian vegetation along the right
bank is primarily birch and spruce,whereas on the left bank it is
alder.
Transect °is located in a shallow pool habitat and has a substrate of
sand and small gravel.Transects 1 (the discharge site)and 2 are
primarily run habitat,and the substrate i~small gravel.At transect
3,the habitat changes to run and shallow pool habitat,the predominant
substrate is small and large gravel.The hydraulic control for tran-
sects 5 and 6 is transect 4.This transect represents riffle habitat,
with substrates composed mostly of small and large gravels.Transects 5
and 6 are located in deep pool habitat,with small and large gravel
substrate.
D-64
-
-
-
-
""'"
-
-
))})i l ))J ])}1 1 l
Staff Gage
Tr8.nsect
€a River Mile
o 1000
!I
FEET
(Approximate Scale)
...-.1'"~\'lE.0
SU 6 \1"'"d
(BRM 87
o
I
0'1
01
Appendix Figure 0-34.Overview of Sunset Side Channel (RM 86.4).
/
"/;
::o
...J
U.ll
Cl
Wa:
:Jen<w
~
Q.41.1 c fs
0...10-11•1*~1.-
PZ'.O.."::o...J
lL.
Cl
W
0:
:J
f/)
<W
~
0.<•10-o.sns rllSEL _'HIl '.]09'
.2 •0.99
Q=45.8 chi-----
/
I
/'"""
-
-
-.
SUNSET SIDE CHANNEL TRJ
GAGE 86.951
n.lll1I)S Q i 101.000 <f,
lfS[l •10-1•n66 Qo.n]3 •90
pI •O~"
lOT C1IIITtlII.UO
.11.11lO ~0 So J2.000 <'I
VSEL •lOO....,J ao.ors,..90
pl.0".57
J--__CIlllTIOQI.lEO
...J
Wen::
-II .._-;,+------.....---..1.-------_.-
MAINSTEM DISCHARGE,SUNSHINE (x1000 cfs)
•t
-51gl
~f
01
Cl J'0 WSEL=93.30 .a
0>
~
-L;:!:ID _
MAINSTEM DtSCHARGE,SUNSHINE (x 1000 cfs)
Appendix Fugure 0-35.Comparison of rating curves from Sunset Side
Channel at transect 1 (from Quane eta al.1985).
0-66
Staff Gage
Transect
FEET
(Approximate Scale)
"",~".....o
I
..J
W,Z.
"!Z-~;<::t:
i 0
w
Q
f ,(;
w
(I)-a:z
~-(I)
-
Appendix Figure 0-36.Location of Sunset Side Channel study site (RM 86.9).
0-67
110 110
1011 CROSS SECTION 0 1011 JCROSS SECTION I
1011 STATION 0+00 1011 STATION 2+23...~
•107 ...107..•::lOll •1011
10'
...
10'-Z 104 1040Z
103 2 103
I-102«102 I-
>101 «101
\IJ 100 >100.J \IJ
UJ 1111 .J 1111
1111 \IJ 1111UJ117117>\IJ
1111 >1111I-115 --491.1.i=115«
..J 114 127 .t•«114
l&J 113 ..J 1130::\IJ
112 0::112
111 111
110 110
0 100 200 300 400 0 100 200 300 400
DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEfT BANK HEADPIN (feet)
0
I 110 1100'1 1011 CROSS SECTION 2 1011 ~CROSS SECTION 3CO
~loa STATION 4 +75 loa STATION 7+58~...107 ...107•••1011 4)1011.......
~105 -105
Z 104 Z 104
0 103 0 103
I-102 i=102«101 «101>>\IJ 100 \IJ 100
.J 1111 .J 1111\IJ IIIl \IJ 1111
UJ 117 l&J 117>1111 >1111 ~V\;;::--i=4111.tI 5 ......-::::I 498 eh
«115 127.h 115 --=127 el.
.J 114-.J 114
\IJ 113 \IJ 1130::112 0::112
111 111
110 110
0 100 200 300 400 0 100 200 300 400
DISTANCE FROM LEFT BANK HEADPIN «feet)DISTANCE FROM LEFT BANK HEADPIN (feet)
Appendix Figure 0-37.Cross section of transects 0,1,2,and 3 at Sunset Side Channel (adapted from
Quane et al.1985).,J ..._J ),I ,J/J I J -]
~D J 'D J B J J
))1 }1 )-])l ")1 )J )1 -J
110 110
1011 CROSS SECTION 4 10111 CROSS SECTION 5
1015 STATION 9"'10 1011 STATION II ~53
~
-;107 -107
1015 lD 1015..lD...105 -105~-104 1 Z 104
Z 10301030
5 102 ...102
101 ~et 101
>100 >100lIJW
..J 1111 ..J 99
lIJ 915 W 915
iii 97 W 87 ~,J 498 eft>911 498 eto >88 ---........127 eft--121 eft ,
i=95 ~~i=81S
et 94 et 84..J ..J
lIJ 93 W 113
0:112 0::92
111 91
110 110
a 100 200 300 400 a 100 200 300 400
DISTANCE FROM LEFT BANK HEADPIN (feetl DISTANCE FROM LEFT BANK HEADPIN (feet)
0
I 110en1011 ~CROSS SECTION 61..0 1011 STATION 14 ...10--101•J!1011....10e
Z 104
2 103...102
~101
W 100
..J 911
lIJ 911
lIJ 117 ~,
J 498 ell>811
,
127 ell"i=95et94..J
lIJ 113
0::lI2
91
90
°100 200 300 400
DISTANCE FROM LEFT BANK HEADPIN (feet)
Appendix Figure D-38.Cross section of transects 4,5,and 6 at Sunset Side Channel (adapted from
Quane et al.1985).
Calibration
Hydraulic data were collected at two calibration flows:127 and 496 cfs
(Appendix Table 0-4).Mean daily discharges for the Susitna River on
the dates that calibration data were collected at the Sunset Site
Channel study site were 42,500 and 57,800 cfs,respectively.Based on
these two calibration flows,an IFG-4 model was used to forecast
instream hydraulics at this study site.The streambed profile,stage of
zero flow,and observed and predicted water surface elevations for the
study reach are plotted to scale in Appendix Figure 0-39.Both cali-
bration data sets were used to predict hydraulic information from side
channel flows of 7 to 1,603 cfs (mainstem discharges of 21,000 to 75,000
cfs)•
To evaluate the performance of the hydraulic model,observed and pre-
dicted water surface elevations,discharges,and velocity adjustment
factors were compared (Appendix Table 0-11).The hydraulic model at
Sunset Side Channel is similar to Circular Side Channel.Because of the
2 calibration flows,only a 2 point rating curve was formulated.In
evaluating the performance of the model,observed and predicted WSEL1s
and discharges were the same because of this rating curve.Velocity
adjustment factors were all within the good range of 0.9 to 1.1.
Additionally,the stage information of the model was compared to the
rating curves established by Quane et ale (1985)(Appendix Figure 0-35).
In the model,the stages of zero flow are not the same as those deter-
mined from the thalweg survey by Quane et al.1985 (Appendix Table
0-12).The stage of zero flow values,input into the model,were
derived from the thalweg points of the model input cross sections of
transects 0,1,2,and 4.The reason for this change in thalweg eleva-
tions is likely the result of the flood event.All the points used in
the model were from measurements made,before the flood,whereas the
Quane et ale (1985)thalweg survey was done after the flood -event.
At transect 6,the velocities at the high calibration flow measurement
(496 cfs)were adjusted upwards by 15%and at the low calibration flow
measurement (127 cfs)adjusted downwards by 21%.Because this transect
bisects a deep pool with eddies,it is difficult to obtain an accurate
discharge measurement.The eddy effect was much more pronounced at the
high calibration flow measurement,as there was about a 40 foot a
section in which the velocities were negative.Because of its depth and
slow velocities,this area was considered valuable habitat for rearing
juvenile salmon.In order to facilitate using these negative velocity
values in the model these measurements were treated as positive.
At transect 3,there was a difference i nWSEL I s at the 127 cfs cali-
bration flow.WSEL at the left bank was 95.03 feet,whereas at the
right bank it was 94.90 feet.As the staff gage WSEL was 94.93 feet and
the majority of flow occurred along this right side,a WSEL of 94.93
feet was used in the model.
At transect 4,there was a large discrepancy (0.54 ft)in WSEL's across
the transect at the .calibration flow of 127 cfs.This occurred because
0-70
'"""
P'"
-
If/i!!!+,
-
1 1 1 '))»)1 j 'J J 1 )))1
SUNSET SIDE CHANNEL
Thalwl~Proflll with Ob.lrVldand
Predictld Water Surfaci Profl/..
•
~
Thai."Surv.,Del.'840828
ThaI ••,Gradi.nl'8.11 f••lImil.
Obatrv.d Waler Surfac.EI.varlan
Slmu'at.d Wal.r Surfac.EI.vallan
Eotrapalal.d Waf.r Surfac.EI.vallan
Eleval ion af Z.ra I'la.
Tho'••,Profile
••
TII&"'ICT •n .....fel tn......cT 4ffll....UGT •nANueT itltAttllC!ITU"KCTO
:_'_'_'_'_'_':'_'_'_'_'_'_'~:'_'_'_'_'_'_'';''::_'_'_'_'_._;~._:._._._...,._._.-1000 "I}_---=::::;.: ",.".
.-_.-._.-.-_.__._._.----_._-_._._._._._-_.,..~~~.--.,~~~~_.-._._......-.-.,...
_..._.~~~~wl!@~~~:~_~--~------~-."7JZ""W I'_I_~--W"~
••
..
..
!..o
~..
..J....
>~
..J..
III
CJ
I
-....J.....
,I..1410 '••1-.1
n ;i ;iii Iii iii i'iii iii.iii iii Ii'iii ,iii i I Iii i ,iii iii'i ,iii Iii
001-00 noD IMOO 'StOO 10+00 U ....OO ]Otoo
STREAMBED STATION 11001)
Appendix Figure 0-39.Comparison of observed and predicted water surface profiles from calibrated model
and surveyed thalweg profile at Sunset Side Channel (adapted from Quane et al.
1985).
Appendix Table 0-11.Comparison between observed and predicted water
surface elevations,discharges,and velocities
for 1984 Sunset Side Channel hydraulic model.
D-72
-
-
-
Appendix Table 0-12.Differences between stages of zero flow input into
the model and Quane et al.(1985)thalweg survey
at Sunset Side Channel.
Stage of Zero Flow (ft)
Transect Model Input Thalweg Survey
0 92.30 92.50
1 92.60 93.00
2 93.40 93.60
3 93.40 93.60
4 94.20 94.40
5 94.20 94.40
"""6 94.20 94.40
,."..
0-73
the section of the channel where a majority of the flow occurred was
higher in elevation and separated by a gravel berm from a lower eleva-
tion minor channel where the staff gage was located.In order to
utilize this cross section in the model,the channel cross section of
the minor channel was elevated upwards by 0.6 feet.
At a section of transect 3,because of channel configuration,the
individual velocity measurements for the 127 cfs site flow were greater
than the corresponding velocity measurements at the higher 496 cfs site
flow.If these original values were to be used in the model the simu-
lated velocities would decrease·with increasing site flows rather than
increase as expected under normal circumstances.In order to amend this
situation,the velocities were adjusted such that the relationship would
simulate a positive increase in velocities with corresponding increases
ins i te flow.
Verification
Based on the first level of verification by EWT&A,the model does an
excellent job of simulating channel hydraulics between 50,000 and 61,000
cfs,mainstem discharge(275 and 649 cfs site flow).Above 61,000 cfs,
the realiability of the simulated depth and velocity distributions begin
to decrease.The model simulations were rated good between 61,000 and
64,500 cfs (649 and 850 cfs site flow),acceptable between 64,500 and
67,000 cfs (850 and 1,000 cfs site flow),and unacceptable above 67,000
cfs mainstem discharge.Below 50,000 cfs,the model simulations were
also rated less than excellent,primarily because of reduced effec-
ti veness in predi cti ng water surface profi 1es as compared to fi el d
observations.The model simulations were rated good between 38,000 and
50,000 cfs (89 and 275 cfs site flow),acceptable between 32,000 and
38,000 cfs (41 and 89 cfs site flow),and unacceptable below 32,000 cfs
mainstem discharge (Appendix Figure D-40).
At the second level of verification there is excellent agreement for
ve loci ty and good agreement for depth between observed and predi cted
values (Appendix Figure 0-41).For a small number of depths there is a
deviation away from the expected one to one relationship and this maybe
attributable to the adjustments in the channel cross section at transect
4.The statistical tests show good agreement between these predicted
and observed values (Appendix Table 0-7).The index of agreement is
almost one,the total RMSE is mostly composed of the unsystematic RMSE,
and the y-intercept is essentially zero with a slope of 0.99.
Application
For habitat simulation modelling purposes the hydraulic simulation model
developed for Sunset Side Channel can simulate channel flows in the
mainstem discharge range of 32,000 to 67,000 cfs.
0-74
-
-
-
-1 1 )1 J J -]1 ].j 1 1
Application Range of the Calibrated Hydraulic Model
at Sunset Side Channel
RM (86.9)
Si te Specific Flow.cfs
o 7 31 107 280 614 1191 1603
I I I I I I
~~
~
~.
~~
~~~~
I I I I I
o 10 20 30 40 50 60 70 75
Cl
I
-.....J
01 Mainstem Discharge at Sunshine Station,cfs x 1000
_Excellent
_Acceptable
ill GoodoUnacceptable
Appendix Figure 0-40.Application range of calibrated hydraulic model at Sunset Side Channel.
SUNSET SIDE CHANNEL
Observed V.s.Predicted Depths
4 r lffll 4
ata°3.5 -I Df(§I 3.5
D~~D ..-.
3 o~,0
0 3w,....III
t ~mD "-
...."t2.5 'oDD'D
'-'2.5I
I-~lL ~DDW 0029 2
0 wwIfPdl0>I-
0 051.5 w 1.5wl-
n:U
lL 0
0 1 0 W
I 0::
.........Il.
O"l
0.5 D 0.5
o·0
0 1 2 :;4 0
OBSERVED DEPTH (FT)
SUNSET SIDE CHANNEL
Observed v.s.Predicted Velocities
D
~
2 :;
08SE~ED VELOCITY (fT /SEC)
o
#o
4
Appendix Figure D-41.Scatterplots of observed and predicted depths and velocities from
the calibrated IFG-4 hydraulic mode1 at Sunset Side Channel.
J J .~J J !!J J .~}J J 1
,.".
~.
I,
Trapper Creek Side Channel (RM 91.6)
Site Description
Trapper Creek Side Channel is located on the west bank of the Susitna
River and is approximately 5.0 miles long (Appendix Figure 0-42).It
has a relatively uniform,broad,·and flat bottomed alluvial channel
which is fed by multiple heads.It is separated from the mainstem
Susitna River by a complex of sand bars,small channels,and vegetated
islands.The head portion of this side channel is located in a complex
of small channels and vegetated islands making it difficult to identify
the origin of breaching flows (Quane et al.1985).
During unbreached conditions,flows in Trapper Creek Side Channel are
principally due to Cache Creek and groundwater from the upper reaches of
the side channel.Breaching of Trapper Creek Side Channel is the result
of the direct overtopping of the multiple heads of the side channel by
the mainstem Susitna River.Based on assessments by Quane et al.
(1985),the channel is estimated to be initially breached at a mainstem
discharge of 43,000 cfs.Based ·on the comparison of the stage versus
mainstem discharge rating curve for transect 4 (Appendix Figure D-43)by
Quane et al.1985,a discharge of 44,000 cfs was selected as the con-
trolling breaching discharge.This mainstem discharge corresponds to a
streamflow measurement of 31.4 cfs.
Based on assessments by Quane et a 1.(1985),backwater has not been
observed.But at mainstem discharges ranging from 15,700 to 22,700 cfs,
pooling was observed at transects 1,2,and 3 whi ch resul ted from the
control located about 370 feet downstream from transect 1.
The 790 foot long IFG modelling site at Trapper Creek Side Channel was
located "in the lower portion of the side channel in a broad open channel
.area (Appendix Figure 0-44).Four cross sections were surveyed within
this area to define channel geometry (Appendix Figure D-45).The upper
two transects were situated ina run,whereas the lower two transects
were in a pool influenced by a downstream control.Substrate consisted
primarily of rubble and gravels with some sand at the first transect.
The thalweg gradient of the side channel is 12.1 ft/mile (Quane et a1.
1985).
Calibration
Hydraulic data were collected at three calibration flows:16,32,and
389 cfs (Appendix Table 0-4).Mean daily discharges for the Susitna
River on the dates that cal ibration data were collected at the Trapper
Creek study site were 20,900 cfs,44,000 ·cfs,and 57,700 cfs respec-
tively.Based on these calibration flows an IFG-4 model was used to
forecast instream hydraul ics for this study site.The streambed pro-
file,stages of zero flow,and observed and predicted water surface
elevations for the study reach are plotted to scale in Appendix Figure
0-46.All three data sets were used to predict hydraulic information
for side channel flows from 9 to 1,351 cfs (mainstem discharges of
12,000 to 75,000 cfs).
0-77
Appendix Figure D-42.Overview of Trapper Creek Side Channel (RM 91.6).
0-78
-
-
-
;I'JSWJ',,
-
o
Icoa
']'--]1 ]j J 1 I 1 j 1 1
,;!....:il,;~,",;,@W"Y"
....~~
b.Staff Gage ~
___Transect
o 260
I 1
FEET
(Approximate Scale)
]g
Appendix Figure 0-44.Location of Trapper Creek Side Channel study site (RM 91.6).
!
-
_0 •10.3•5525 (lISEl •9DJ 11.6511
I Ie.%•D."
10
I~
tilI......
I~
I~
Q:31.4 ctl
0se •\D·%·D5l1 [VSD..1ll1 I.Z033--......
,.-2 ..0."
3:o
...I
LL.J
Cl
UJa::
:::l!J).«w::a:
'.c •1O-IS •ts,\0..3."541
,%•D.";
r-__q,..IO..Z•SIlI3 0..'."]1
,z •0."
3:o
....I
LL.!
o
Wcr
:::len«
UJ::a:
TRAPPER CREEK SIC TR4
GAGE 91.651
-.~;lfC ..,+--..J........_-..
MAINSTEM DISCHARGE.SUNSHINE (x1000 cfs)WSEL(+90 feet)'
.t
~~:!o~r.,.
~I '-COITIOI.UD
15 0 I..l!6~£Ql oS o.s:l).lOD ell::
2 ,o d I/S(t •IG"'''SS al •I 5&5 •to
~I ,Z·u
'"'"'"WSEL=92.10
....I
~"-.00.tD"'O.UID
u-.ooc i CO So ....000 cfs
wsn •IOD.DIllI OD.D'Y,•90
...'..O~&J
~,
-
-..t-------~-------___.
MAINSTEM DISCHARGE:SUNSHINE (x 1000 CfS)
Appendix Figure D-43.Comparison of rating curves from Trapper Creek
Side Channel transect 4 (from Quane et.al.
1985)•
0-79
j 1 I 1 )1 J J ·~l C~l i
10'105
104 CROSS SECTION I 104 j CROSS SECTION 2
103 STATION 0 ..00 103 STATION 2 ...89-102 102..ii..101:!:101 ..
100 ...100....
Z 88 Z 811081101111
I-117 j:117~>8S :!1111
UJ 115 II'.J UJ
UJ 114 .J 9"
113 UJ 113~"-I 389 010
l&J 112 l&J 112 ...........<7 32 cro l6 01.>5
i=81 >III
~110 t-110
.J 118 ~118
l&J .J 11110::liS UJ
111 ct:111
1111 1111
S5 115
0 100 200 300 400 0 100 200 300 400
DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (feet)
0 :~=j CROSS SECTION 4tlOll
CROSS SECTION 3CO104.......103 STATION 5 +76 103 STATION 7+90--102 ...102-....101 ..101........100 ....100....
811 Z 1111
Z III 0 8110117j:117~18 ~811
>85 UJ 85
l&J 84 .J 84.J UJUJ13 113,,~
;;Ie 'P.'""'"160fa
l&J 12 I&J 112
>81 >81
~110 ~80
..J 1111 .J elll
l&J 1111 I&J 1111
D:117 II:117
III III
I'115
0 100 200 300 400 0 100 200 300 400
DISTANCE FROM LEFT BANK HEADPIN (feet)DISTANCE FROM LEFT BANK HEADPIN (fset)
Appendix Figure 0-45.Cross secti on of transects 1,2,3,and 4 at Trapper Creek Si de Channel .
(adapted from Quane et al.1985).
TRAPPER CREEK SIDE CHANNEL
Thalwev Profile with ObServed and
Predicted Water Surface Protlles
..
/'L..!t0,.,1
IA"pro•.lco'"
~
Thalw.g Su,v.,Dol.'8401t11l
Thalw.g G,adi.nt.12,1 f••I"nll.
Ob ••rwed Wa'.r Surfao.Elevation
SIIIlulaled Wat.,surfa..EI.vallan
Elt,apalal.d Wal.,Surfa••EI.vaiion
EI.valla.of Z.ra flail
Thalw.g Profll.
..
LI100'.0 GAGtl
I'tISI IAO"'••O~Gf:"'51 AOF.'GAG£"DraB G.ol£
1t.lIsl 'I"~
TIU.",tCT I TUNIECT J:TR'NSECT'TaANSlECl4
1011 "e'HI -It~!oo I I iii i a::t:oo i'i ,.iii 10';'00 ,i iii i I fe:':'OO '
..
=-.._._._._._._._._._._._.-._.-...O.,.}_."_._D_._._"_"_·---·-"•319 etl E.llapolaliop r01'l1l1 III'Z ......_._._._._._._._.._._.._._.._.._.••'52:c:f.",draulu:mod.'
g .._._.._._....tach ~!c ta .:,.._._._._._._._._._._.._._._.._....._._.._._._.._.._._._._._._._._._.._....-.-._.-"'~".........-----*"'-
~.---------~------.----------------_.-------_._-~---------------------------_.-~~.J
:10 ~
>~~
.J ....
'"
o
I
00
N
STREAMBED STATION Uootl
Appendix Figure 0-46.Comparison of observed and predicted water surface profiles from calibrated model and
surveyed thalweg profile for Trapper Creek Side Channel (adapted from Quane et al.1985).
)-~I J __~t t .1 I ___.J --_.]_~_J
-
-
i
,~
To evaluate the performance of the hydraul i c model,observed and pre-
dicted water surface elevations,discharges,and velocity adjustment
factors were compared (Appendix Table D-13).Of the 12 sets of observed
and predicted WSEL1s,six sets were within ±0.02 feet of each other and
the other six sets were within ±0.05 feet of each other.All the
observed and predicted discharges were within 10%of each other except
for'one set in which there was an 11%difference.All velocity
adjustment factors were within the good range of 0.9 to 1.1.Addi-
tionally,the stage information of the model was compared to the rating
curves established by Quane et a1.(1985)(Appendix Figure D-43).
Between the time that the first two calibration flows (389 and 32 cfs)
were made and the last calibration flow of 16 cfs was made,the channel
cross section at transect 1 was scoured by a flood event.In order to
utilize this information in the model,the cross section determined from
the survey and the 16 cfs flow measurement were used,and the WSEL I s of
the two calibration flows (389 and 32 cfs)were then reduced by 0.37
feet.
Transect 1 was determined to be a poor site for measuring discharge
because it was a pool area affected·by a downstream control.The
velocities for the 32 cfs calibration flow were therefore adjusted
upwards by 27%,and at the 16 cfs calibration flow were also adjusted
upwa rds by 20%.
Verification
Based on the first level of verification by EWT&A the model does a good
job of simulating channel hydrau1 ics between 20,000 cfs and 54,000 cfs
mainstem discharge (15 and 220 cfs site flow)(Appendix Figure 0-47).
There are sufficient deviations in water surface elevation and discharge
between predicted and observed values within this range to preclude
attainment of the excellent rating.This occurs because the model is
approximating a portion of the rating curve described by two adjoining
linear relationships with a single line.
Between 54,000 cfs and 58,000,cfs mainstem (220 and 460 cfs site flow)
the model does an excellent job of simulating channel hydraulics.
Beyond 58,000 cfs mainstem,the quality of the simulations begins to
deteriorate as tne slope of the stage/discharge relationship for the
site flattens with a change in channel geometry.The deviation between
the regression l"ine developed within the model and that of the rating
curve increases with discharge until the model simulations are no longer
acceptable.The model simulations were rated good between 58,000 cfs
and 61,000 cfs (460 and 600 cfs site flow),acceptable between 61,000
cfs and 66,000 cfs (600 and 820 cfs site flow),and unacceptable above
66,000 cfs mainstem (Appendix Figure 0-47).
At the second level of verification there is good agreement between the
observed and predicted values for velocity and depth (Appendix Figure
0-48).The statistical tests al so show good agreement between the
predicted and observed values (Appendix Table 0-7).The index of
agreement is 0.99,the total RMSE is largely composed of the unsys-
tematic RMSE,and the y-intercept is almost zero with a slope near one.
0-83
Appendix Table D-13.Comparison between observed and predicted water
surface elevations,discharges,and velocities
for 1984 Trapper Creek Side Channel hydraul ic RI]
model.
-
0-84
J )1 J )'I 1 ]---]----1 1
Application Ranoe of the Cali brated Hydraul ic Mode I
at Trapper Creek Side Channel
RM (91.6)
Site Specific Flow,cfs
1&22 29 101 564 1030 1351
I I I I I I
11:::==
I
I I I I I ,
o 10 20 30 40 50 60 10 15
CJ
I
OJ
U1 Mainstem Discharoe at Sunshine Station.cfs x 1000
III Excellent
_Acceptable
•Good
D Unacceptable
Appendix Figure 0-47.Application range of the calibrated hydraulic model at Trapper Creek Side
Channel.
TRAPPER CREEK SIDE CHANNEL TRAPPER CREEK SIDE CHANNEL
Observed vs.Predicted Depths
4
,~
o~jo,0
o ~08
2 .3
OBSERVED VELOCITY (FT!SEC)
Observed vs.Predicted Velocities
4 I I
3.5
,....
0 .3w
lfl
"-t 2.5'-'
~
£3
20
..J
~
0 1.5wt-o
iswn:
0-
0.5
0
02.82.4
o 8
o 0
0008 0
IJD o~Q
DtB~D
°fQ1 Co
o
0.8 1.2 1.6 2.0
OBSERVED DEPTH (fEET)
0.4
3 i I
2.8
2.6
2.4
2.2
2
1.B
1.6
1.4
1.2
0.8
0.6
0.4
0.2
o
0.0
~w
W
IL.
'-'
:r
twa
owt-
U
iswn:
0.o
I
OJ
0'1
Appendix Figure D-48.Scatterplots of observed and predicted depths and velocities from
the calibrated IFG-4 model at Trapper Creek Side Channel.
.)J ~.J _J .J ,~_....,J _c.~)__I
Application
For habitat simulation modelling purposes the hydraulic simulation model
developed for Trapper Creek Side Channel can simulate channel flows in
the mainstem discharge.range of 20,000 to 66,000 cfs.
SUMMARY
A summary of the range of Illainstem discharges that the hydraulic models
can simulate for the rearing habitats of salmon at the six lower river
IFG modelling sites is presented in Appendix Table 0':'14.
I~
Appendix Table 0-14.
Site (RM)
Summarization of the range of mainstem
discharges that the hydraulic models can
simulate for the rearing habitats of salmon at
the six lower river IFG modelling sites.
Mainstem Discharge Range (cfs)
-
,....
Island Side Channel (63.2)
Mainstem West Bank (74.4)
Circular Side Channel (75.3)
Sauna Side Channel (79.8)
Sunset Side Channel (86.9)
Trapper Creek Side Channel (91.6)
0-87
35,000 to 70,000
18,000 to 48,000
36,000 to 63,000
44,000 to 63,000
32,000 to 67,000
20,000 to 66,000
..-
ACKNOWLEDGEMENTS
The authors express their appreciation to the following for their
assistance in preparing this report:
The other ADF&G Su Hydro Aquatic Studies Program staff who
provided their support to this study.For collection of field
data:Fred Metzler,Pat Morrow,Isaac Queral,Glenn Freeman,
and John McConnaughey.To Paul Suchanek for collection of the
cover information used in assessing the weighted usable areas
of the models.In reduction of the data,making the many
compu~er runs,and helping prepare this appendix:Fred
Metzler,Mary Shiffer,Dan Kingsley,and Kathy Sheehan Dugan.
To Tim Quane,Pat Morrow,and Isaac Queral for use of much of
their findings and figures from Task 36 support technical
report -Hydrological Investigations at Selected Lower Susitna
River Study Sites.To the editors:Doug Vincent-Lang,Tim
Quane,and Drew Crawford.For the cartography on the final
figures:Carol R.Hepler and Roxann Peterson.
To E.W.Trihey and Associates;particularly Bob Aaserude and
Diane Hilliard for their valuable expertise in the collection
of data,the calibration,and the verification of the
hydraulic models .
0-88
.....
1, j
.....
-
I.....
LITERATURE CITED
Acres American,Inc.(Acres).1982.Susitna hydroelectric project
draft FERC license application,volume 1,exhibit E,chapter 2.
Anchorage,Alaska •
Ashton,W.S.,and S.A.Klinger-Kingsley.1985.Response of aquatic
habitat surface areas to mainstem discharges in the Yentna River
confluence to Talkeetna reach of the Susitna River.Draft report.
R&M Consultants,Inc.and E.Woody Trihey and Associates.Prepared
for Alaska Power Authority.Susitna Hydroelectric Power Project.
Anchorage,Alaska.
Bovee,K.D.,and R.Milhous.1978.Hydraulic simulation in instream
flow studies:theory and techniques.Instream Flow Information
Paper No.5.Instream Flow Service Group.USFWS.Ft.Collins,
Colorado.
•1982.A guide to stream habitat and analysis using instream-----~f'low incremental methodology.Instream Flow Information paper No.
12.Coop.Instream Flow Service Group.USFWS.Colorado.
Buchanan,T.J.,and W.P.Somers.1969.Techniques of water resources
investigations of the United States Geological Survey.Chapter A8.
Discharge measurements at gaging stations.USGS.Washington DC.
Hilliard,N.D.1985.Extrapolation limits of the 1984 middle river IFG
models.Technical Memorandum.E.Woody Trihey and Associates.
Anchorage,Alaska.
Hilliard,N.D.,S.Williams,E.Woody Trihey,R.C.Wilkinson,and C.R.'
Steward III.1985.Summary of hydraulic conditions and habitat
forecasts at 1984 middle river study sites.Draft report.E.
Woody Trihey and Associates.Prepared for Alaska Power Authority.
Susitna Hydroelectric Power Project.Anchorage,Alaska.
Instream Flow Group (IFG).1980.The incremental approach to the study
of instream flows.USF&WS.W/IFG-8)W31.Ft.Collins,Colorado.
Main,R.1978.IFG-4 program user's manual.U.S.Fish and Wildlife
Service.45 pp.
Milhous,R.T.,D.L.Wegner,and T.Waddle.1984.User 's guide to the
Physical Habitat Simulation System (PHABSIM).U.S.Fish and
Wildlife Service.Instream Flow Information Paper No.11.
FWS/OBS-81/43 Revised.Fort Collins,Colorado.
Quane,T.,P.Morrow,and I.Queral.1985.Hydrological Investigations
at Selected Lower Susitna River Study Sites.Alaska Department of
Fish and Game.Su Hydro Aquatic Studies Task 36 Support Technical
Report.Alaska Department of Fish and Game.Anchorage,Alaska.
D-89
Suchanek,P.M.,K.J.Kuntz,and J.P.McDonell.1985.The relative
abundance,distribution,and instream flow relationships of
juvenile salmon in the lower Susitna River.Alaska Department of ~
Fish and Game.Susitna Aquatic Studies Report No.7,part 2.
Alaska Department of Fish and Game.Anchorage,Alaska.
Trihey,E.W.1979.The IFG incremental methodology.In G.L.Smith,
ed.Proceedings of the Instream Flow Criteria and Modeling
Workshop.Colorado Water Resources Research Institute,Colorado
State University.Pages 24-44.Information Series No.40.Fort
Collins,Colorado.
•1980.Field data reduction and coding procedures for use with
--"7""t'he IFG-2 and IFG-4 hydraul ic simul ati on model s.Instream Flow
Service Group,USFWS.Fort Collins,Colorado.
•and D.L.Wegner.---use with the physical
Flow Group.Instream
Colorado.
1981.Fiel d data collection procedures for
habitat simulation system of the Instream
Flow Service Group.USFWS.Fort collins,
Wilmott,C.J.1981.On the validation of models,physical geography 2.
V.H.Winston and Sons.p.184-194.~
-
.....
-
D-90