HomeMy WebLinkAboutBradley Lake Salmon Escapement 1986"RECORD COPY"
RETURN TO BRADLEY O&M FILES
Alaska Power Authority ~
BRADLEY RIVER SALMON ESCAPEMENT
MONITORING STUDY 1986
BRADLEY LAKE
HYDROELECTRIC PROJECT
FEDERAL ENERGY REGULATORY COMMISSION
PROJECT NO. 8221-000
Prepared By
DAMES & MOORE
ANCHORAGE. ALASKA
December 1986
ALASKA POWER AUTHORITY
BRADLEY RIVER SALMON ESCAPEMENT
MONITORING STUDY
1986
BRADLEY LAKE HYDROELECTRIC PROJECT
Federal Energy Regulatory Commission
·Project No. P-8221-000
Prepared by
Dames & Moore
Anchorage, Alaska
December 1986
FINAL REPORT
BRADLEY RIVER SALMON ESCAPEMENT
MONITORING STUDY
1986
Prepared for:
BECHTEL CIVIL & MINERALS, INC.
AND
ALASKA POWER AUTHORITY
by
John Morsell
David Erikson
Anne DeToni
Mary Pears a 11
DAMES & MOORE
TABLE OF CONTENTS
INTRODUCTION •
METHODS • •
Study Area
Study Duration
Hoop Net Sampling •
Seine Sampling
Electroshocker
TABLE OF CONTENTS
Carcass Counts . • • •
Miscellaneous Observations
RESULTS
Overall Catch
Hoop Net Sampling •
Pink Salmon
Chum Salmon • • • • • •••
Chinook Salmon ••••••••••••••
Coho Salmon •••••••.
Sockeye Salmon .•••
Beach Seine Sampling
Electroshocking ••.•.•••
Carcass Counts
Tag Retention .••••••••••
Pink Salmon Stream Life Duration
Mark and Recapture Estimates
Pink Salmon
Chum Salmon
Pink Salmon Straying
Miscellaneous Observations
DISCUSSION •••••.•.••••
Comparison of 1986 Season with Prior Years
Abundance Indices • . • • • ••••.
Mark and Recapture Estimates .••.•
RECOMMENDATIONS FOR FUTURE YEARS •
REFERENCES . . • • • • . • . • • .
Page
1
1
1
3
3
6
6
7
7
7
7
7
9
9
9
15
15
15
15
18
18
18
19
19
21
21
22
22
22
23
24
26
29
LIST OF FIGURES
LIST OF TABLES
Figure
1
2
3
4
5
6
LIST OF FIGURES
Description
Lower Bradley River with Salmon Escapement Study Area
Bradley River Hoop Net Design ..... .
Hoop Net Sites and Beach Seine Locations ...... .
Hoop Net Catch Per Hour of Pink Salmon for all Nets
Combined Over the Study Period ...•.......
Hoop Net Catch Per Hour for Chum Salmon for all Nets
Combined Over the Study Period .
Distribution of Spawning Salmon .•.........
LIST OF TABLES
Page
2 . . . . 4
5
11
13 . . . . 17
1 Total Adult Salmon Catch for all Sample Methods Combined . 8
2 Hoop Net Catch Statistics for Pink Salmon . . . . 10
3 Hoop Net Catch Stati sties for Chum Salmon 12
4 Hoop Net Catch Statistics for Chinook Salmon . . . . . . . . . 14
5 Hoop Net Catch Statistics for Coho Salmon 16
6 Population Estimates for Pink Salmon Based on
Hoop Net Catches for Each Sample Week . . . . 20
i i
r
BRADLEY RIVER SALMON
ESCAPEMENT MONITORING
STUDY 1986
INTRODUCTION
The Federal Energy Regulatory Commission (FERC) license granted to the
A 1 ask a Power Authority for the Brad 1 ey Lake Hydroe 1 ectri c Project (Project
No. 8221-000) stipulates that a plan be developed and implemented to monitor the
abundance of salmon in the Bradley River. A Salmon Monitoring Plan was sub-
mitted to FERC in June 1986 (Alaska Power Authority 1986). The intent of the
monitoring program is to provide a yearly index of salmon abundance both during
the pre··operational and post-operational periods to allow an appraisal of pro-
ject impacts to the fish resources of the Bradley River. This report summarizes
the results of the first year of pre-operational studies of adult salmon escape-
ment to the Bradley River per the proposed scope of work described in the Salmon
Monitoring Plan.
The fish resources of the Bradley River have been delineated in con-
siderable detail through a series of studies (USFWS 1982; Woodward-Clyde
Consultants 1983, 1984; Northern Technical Services 1985}. These studies indi-
cated that pink salmon was the principal salmon species utilizing the river for
spawning although small numbers of chum, chinook and coho salmon may also spawn
in the system. The potential spawning habitat is limited to a short segment of
the river due to impassible waterfalls at one end and unsuitable substrate com-
bined with tidal flooding at the. other.
The primary objectives of the study were to:
1. Establish standard methods for comparing salmon abundance from year to
year with emphasis on pink salmon.
2. Provide an index or estimate of 1986 salmon escapement to the Bradley
River with emphasis on pink salmon.
METHODS
The study area consisted of a 2011 m (6600 feet) stream segment extending
from thE! downstream end of Riffle Reach to the upstream end of Bear Island
Slough (Figure 1). This study area encompassed most of the known pink salmon
spawning areas and was above the primary zone of tidal influence, a 1 though
Riffle Reach is affected at the higher tides.
1
ASH STUDY
N
~~\~' STUDY AREA
3.0 etc. • RIVER MILES
0 feet 500
SCALE
LOWER BRADLEY RIVER WITH SALMON ESCAPEMENT STUDY AREA
2 FIGURE 1
Study Duration
The study was conducted from July 14 through September 12, 1986, with
intensive sampling occurring on Tuesday, Wednesday and Thursday of each calendar
week. This 9 week sampling period coincided roughly with the duration of the
pink salmon spawning run based on the experience of previous investigators.
Hoop Net Sampling
Hoop nets were employed as a primary sampling technique. The principal
objective was to establish standard catch procedures that would allow the
development of a catch-per-unit-effort index of fish abundance.
Prior to the field investigation, hoop nets were fabricated according to
the dimensions shown in Figure 2. All netting was 2.5 inch stretch mesh nylon.
A number of net 1 eads were a 1 so prepared for attachment to the hoop nets in
various configurations depending on the needs of each net site. After initial
reconnaissance and some experimentation, seven hoop net sites were selected for
repeated use. The locations and configurations of the nets are illustrated in
Figure 3 and aerial photographs of each net site are included in the project
files. The permanent site for net 5 was not established until week 4 and net 7
was established in week 3; the other net sites were established in week 1. Hoop
net site selection was based on adequate water depth (at least 1.1 m), reason-
ably slow current velocity, and judgemental factors regarding the likelihood of
catching fish.
During most of the study weeks, the hoop nets were placed in the water on
Tuesday morning and fished until Thursday morning for a total of approximately
48 hours at which time they were removed from the water until the fo 11 owing
week. During week 7 high water prevented dep 1 oyment of a 11 the nets and only
nets 2, 4, 5 and 6 were set. During normal operation, nets were checked and
fish removed approximately every 4 hours during the day and were allowed to fish
undisturbed overnight. The nets were checked less frequently during week 7
because of equipment problems and difficult working conditions.
Pink salmon caught in the hoop nets were measured and tagged using sequen-
tially numbered Flay spaghetti tags. Other salmon species were measured and
marked by splitting the left pectoral fin. Sex and spawning condition {where
known) were recorded for all salmon.
3
....
~ Q c
~ m
~
1-c
2 112• STRETCH MESH NETTING THROUGHOUT
FLOATS
12" THROAT 18• THROAT ...-.. -
~ 0
>{'\
y
(S
6.
~ "'
. .\,.~HooL/ LEAD WEIGHTS
8' SQUARE FRAME OPENING
16' ... 1
BRADLEY RIVER HOOP NET DESIGN
~~--------------------------------------------------------------------~
I
I
.I
I
I
HOOP NETS _>zm>
SEINESITE ~
0
-
HOOP NET SITES AND BEACH SEINE LOCATIONS .
feet 300
SCALE FIGURE 3
5
Seine Sampling
Beach seining was also tested for possible use as a standard catch method
which might allow development of an abundance index. A beach seine was employed
at 2-4 sampling stations in each sample week. Only 2 stations were seined con-
sistently (Figure 3). Several other locations were tried during the early weeks
of the study but were eliminated as standard sample sites because of the hazar-
dous working conditions, presence of snags, high current velocity or other
pragmatic considerations.
Throughout most of the study, a 36 m ( 120 feet) bag seine constructed of
0.5 inch knotless nylon netting was employed. A 61 m (200 feet) seine was used
during the first 3 weeks of the study until it was lost on a snag. In all
cases, the seine was deployed by tying one end to shore and feeding the
remainder off the bow of a boat as it traveled in reverse making a broad sweep
across the river, down current, and back to the beach. The normal procedure was
to make 3 consecutive seine sets at each site during each sample week.
Fish captured in the seine were processed in the same manner as the hoop
net catch.
Electroshocker
Two electroshocking units were experimentally deployed during the study to
test their utility either as a standard capture technique or as a means of cap-
turing fish for mark and recapture estimates. The first consisted of a 220 volt
generator combined with a Coffelt VVP-3E electrofishing unit mounted in a
16-foot aluminum boat. An anode consisting of a 9.5 mm steel cable was mounted
from a fiberglass boom off the bow and a trailing cathode was mounted on the
side of the boat. The boat-mounted array was not camp 1 ete 1 y effective and was
replaced by a Smith-Root Type XI backpack electroshocker which was operated both
while on foot and from a boat.
After an initial period of experimentation, electroshocking was conducted
at irregular intervals throughout the study period. Its primary utility was as
a means of delineating the distribution of fish in shallow riffle areas.
6
Carcass Counts
Visible salmon carcasses were counted once per study week during low tide
by examining shorelines, snags and gravel bars between the fish camp and Fox
Farm Creek. As an experimental means of enhancing the carcass counts, low
chicken wire fences were constructed perpendicular to the current at two loca-
tions (Riffle Reach and Eagle Pool Reach) on gravel bars exposed at low tide.
The fences were 6 feet long and secured by rebar stakes.
Miscellaneous Observations
Water temperature and river stage (staff gauge reading) were recorded at
the camp in the morning and evening of each field day.
The presence of harbor seals and other events of biological interest were
recorded as appropriate.
RESULTS
Overall Catch
Adults of all 5 North American species of Pacific salmon were caught in the
Brad 1 ey River study area (Tab 1 e 1) with pink salmon (Onchorhynchus gorbuscha)
dominating the catch and chum salmon (~. keta) next in abundance. The percen-
tage of fish caught more than once was high for all species except sockeye
salmon (~. nerka); therefore, the methods used appeared to catch a significant
portion of the populations present in the study area.
Other species of fish caught during the study included all size classes of
Dolly Varden (Salvelinus malma) and slimy sculpins (Cottus cognatus). Juvenile
coho (~. kisutch) and/or chinook salmon (~. tchawytscha) were also caught inci-
dental to the seine sampling for adult fish.
Hoop Net Sampling
The hoop nets caught adult salmon consistently throughout the sample
period. Location and fishing effort were relatively easy to standardize.
Consequently, the hoop net catch record appears to provide a reasonable
indication of fish abundance and run timing.
7
Sample Week
1
July 15-17
2
July 22-24
3
July 29-31
4
Aug. 5-7
5
Aug. 12-14
6
Aug. 19-21
7
Aug. 26-28
8
Sept. 2-4
9
Sept. 9-11
Study Total
TABLE 1
TOTAL ADULT SALMON CATCH FOR ALL SAMPLE METHODS COMBINED
(Recapture Percentage in Parentheses}
Pink Chum Chinook Coho Sockeye
Salmon Salmon Salmon Salmon Salmon
5(0) 33(0) 4(0) 0 2(0)
36 ( 11) 65(29) 8(37) 0 1(0)
158(16) 149(42) 13(31) 0 1(0)
257(25) 102 ( 49) 13(61) 0 0
270(25) 69(70) 2(0) 17 (12) 3(0)
262(25) 20(65) 11 ( 91) 16 ( 37) 0
236 (27) 0 0 1 (100) 0
156 ( 27) 1 (100) 0 57 (14) 0
68(41) 2(0) 2(0) 38 ( 29) 1(0)
Total
Salmon
44(0)
110( 24)
321(29)
372(33)
361(32)
309(30)
237(27)
214(25)
111(35}
1448(25) 441 ( 44) 53 ( 47) 129( 22) 8(0) 2079(29)
8
Pink Salmon
Table 2 sulllllarizes the hoop net catch data for pink salmon for each
net and for each study week. The composite catch per hour (Figure 4) i nd i-
cates that the run began in mid-July, built to a peak in late August, then
declined rapidly. By September 13 there were few pink salmon left in the river
and those that were present were mostly spawned out. Figure 4 should be
interpreted cautiously because not all nets were employed in each study week
and, therefore, the effort from week to week was not consistently uniform. The
very high water conditions in week 7 prevented the use of some nets thus pre-
venting a full set of data during the probable peak of the run. Nevertheless,
it is felt that the general trends indicated in Figure 4 are a real indication
of the changes that occurred in the pink salmon population during the course of
the study.
Chum Salmon
Table 3 sulllllarizes the hoop net catch data for chum salmon and Figure 5
i 11 ustrates the composite catch per hour. Si gni fi cant numbers of chum salmon
were present in the river at the start of the study in mid-July. The peak abun-
dance appeared to occur in late July with the number of fish declining rapidly
throughout August. Few chum salmon were present after the third week in August.
Judging from the progressive condition of the fish from the beginning to the end
of the study, there appears to be little doubt that the chum salmon were
spawning in the Bradley River.
Chinook Salmon
The catch statistics for chinook salmon (Table 4) indicate that small num-
bers of Chinooks were present in the river through the third week in August with
greatest numbers occurring during the first 2 weeks of August. Many of the fish
caught in the hoop nets were caught more than once and net number 6 in Bear
Island Slough caught by far the most chinook salmon. Recapture data indicate
that 28 individual chinook salmon were captured of which 10 were females, 10
were mature males, and 8 were jacks (small, precocious males). Several of the
fish were recaptured more than once and spent at least several weeks in Bear
Island Slough. The progressive condition of the fish from pre-spawning to
spawned out through the course of the study suggests that a least some of the
chinook salmon spawned within the slough.
9
TABLE 2
HOOP NET CATCH STATISTICS FOR PINK SALMON
'
WEEK I
,,
1 2 3 4 5 6 7 8 9
Net
Number Total
CPH * Total Total Total Total Total Total Total Total
Catch Catch CPH Catch CPH Catch CPH Catch CPH Catch CPH Catch CPH Catch CPH Catch CPH
1 4 0.09 9 0.13 33 0.64 25 0.52 23 0.48 66 1.78 ----49 1.08 22 0.45
2 0 0 3 0.06 13 0.25 18 0.39 48 0.98 1 0.03 0 0 5 0.11 7 0.14
3 0 0 1 0.02 13 0.25 24 0.51 10 0.20 17 0.52 ----14 0.30 3 0.06
4 1 0.04 1 0.02 21 0.44 33 0.71 46 0.95 43 1.30 94 3.47 42 0.90 27 0.56
... 5 ----------38 0.82 10 0.20 35 1.09 131 4.16 20 0.43 3 0.06
0
6 0 0 4
'
0.08 8 0.15 6 0.13 1 0.02 2 0.06 11 0.25 4 0.09 0 0
I
!
7 ------7 0.14 8 0.17 5 0.10 26 0.77 ----11 0.23 1 0.02
All
Nets 5 0.03 18 0.07 95 0. 31 152 0.46 143 0.41 190 0.81 236 1. 51 145 0.44 63 0.18
-------.
* CPH -Catch Per Hour
... ...
., -G) c
::0 m
~
2
n > -4 n
:::t
, 1
m
:0
5 c
:0
HOOP NET CATCH PER HOUR OF PINK SALMON
FOR ALL NETS COMBINED OVER THE STUDY PERIOD.
o+-----------4----------~----------~---------+----------~--------~~---------r----------+---------~
10 20 30 9 19 29 8 18 28
JULY AUGUST SEPTEMBER
.......
N
1
Net
Number Tot a I Total
Catch CPH Catch
1 13 0.28 26
2 3 0.06 7
3 2 0.09 5
4 3 0.13 12
5 -----
6 0 0 7
7 -----
All
Nets 21 0.14 57
• CPH -Catch Per Hour
2 3
Total
CPH Catch CPH
0.40 40 0.78
0.14 18 0.35
0.10 10 0.19
0.24 21 0.44
------
0.10 7 0.13
--35 0.69
0.22 131 0.42
TABLE 3
HOOP NET CATCH STATISTICS FOR CHUM SALMON
WEEK I
4 5 6 7 8 9
Total Total Total Total Total Total
Catch CPH Catch CPH Catch CPH Catch CPH Catch CPH _:
21 0.44 5 0.10 4 0.11 ----0 0 0 0
7 0.15 21 0.43 5 0.15 0 0 1 0.02 0 0
1 0.02 1 0.02 1 0.03 ----0 0 0 0
20 0.43 23 0.47 5 0.15 0 0 0 0 1 0.02
31 0.67 11 0.22 2 0.16 0 0 0 0 0 0
6 0.13 2 0.04 1 0.03 0 0 0 0 0 0
9 0.19 4 0.08 1 0.03 ----0 0 0 0
95 0.29 67 0.19 19 0.08 0 0 1 <0.01 1 <0.01
..
Co)
11 -G) c
:0 m
01
0.15
0 ,.
-t
0
::t ,
m
::tl
::r
0 c:
::tl
HOOP NET CATCH PER HOUR FOR CHUM SALMON FOR ALL NETS COMBINED
OVER THE STUDY PERIOD.
ot-----~------~------4-------~-----+----~~~====~==~-+----~
1 10 20 30 9 19 29 8 18 28
JULY AUGUST SEPTEMBER
......
+:»
1
Net
Number Tot a 1 Total
Catch CPH Catch
l 0 0 0
2 2 0.04 0
3 0 0 0
4 0 0 0
5 -- ---·
6 1 0.07 7
7 ·-----
All
Nets 3 0.02 7
•cPH -Catch Per Hour
2 3
Total
CPH Catch
0 0
0 1
0 1
0 0
----
0.06 8
--0
0.03 10
TABLE 4
HOOP NET CATCH STATISTICS FOR CHINOOK SALMON
"~----
WEEK I
4 5 6 7 8 9
Total Total Total Total Total Total
CPH Catch CPH Catch CPH Catch CPH Catch CPH Catch CPH
0 0 0 0 0 1 0.03 0 0 0 0 0 0
0.02 0 0 0 0 0 0 0 0 0 0 1 0.02
0.02 0 0 0 0 0 0 0 0 0 0 0 0
0 1 0.02 0 0 0 0 0 0 0 0 0 0
--1 0.02 0 0 4 0.12 0 0 0 0 0 0
0.15 11 0.23 2 0.04 6 0.19 0 0 0 0 1 0.02
0 0 0 0 0 0 0 0 0 0 0 0 0
0.03 13 0.04 2 0.01 11 0.05 0 0 0 0 2 0.01
'--
Coho Salmon
Coho salmon (Table 5) began to enter the catch during the fifth week
(August 12) with highest catches occurring during the first week in September.
Substantial numbers of fish were still present in the river during the final
study week; therefore, run duration and peak timing cannot be determined. Fresh
cohos (no spawning coloration) were still being caught in the second week in
September; consequently, it is likely that the coho salmon run continued well
past the final study week.
Sockeye Salmon
Eight sockeye salmon were caught in the hoop nets with 7 captured prior to
mid-August. All of these fish were fresh and none were recaptured.
Beach Seine Sampling
Beach seining was relatively successful at capturing pink and chum salmon
at some times and in some locations but was not sufficiently consistent to pro-
vide worthwhile index data. Changing conditions within the river and difficulty
in maintaining a consistent haul technique probably preclude seining as a method
for determining abundance. However, seining did show promise as a recapture
technique for mark and recapture estimates and allowed sampling of riffle
spawning areas. The seine catch is included within Table 1 and is discussed
further in the section on mark and recapture estimates.
Electroshocking
Electroshocking was largely unsuccessful as a capture technique for adult
salmon. The rapid current and turbid water made netting of stunned fish extre-
mely difficult. The few adult salmon caught with the electroshocker are
included in the Table 1 catch statistics.
The electroshocker did, however, allow momentary glimpses of salmon, espe-
cially in shallow areas, and, thus, was useful in delineating locations of fish
occurrence and likely spawning areas. As the stunned fish came to the surface,
it was also possible to see the colored tags of tagged fish.
Probable spawning areas based on electroshocker observations as well as
seine and hoop net data are delineated in Figure 6.
15
TABLE 5
HOOP NET CATCH STATISTICS FOR COHO SALMON
WEEK
1 2 3 4 5 6 7 8 9
Net
Number Total Total Total Tot a 1 Total Tot a 1 Total Total Tot a 1
Catch CPH * Catch CPH Catch CPH Catch CPH Catch CPH Catch CPH Catch CPH Catch CPH Catch CPH
l 0 0 0 0 0 0 0 0 2 0.04 2 0.05 ----1 0.02 2 0.04
2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0 1 0.02 1 0.03 ----1 0.02 2 0.04
4 0 0 0 0 0 0 0 0 2 0.04 5 0.15 1 0.04 13 0.28 6 0.13
5 0 0 0 0 0 0 0 0 2 0.04 3 0.09 0 0 4 0.09 10 0.20 ...
0> 6 0 0 0 0 0 0 0 0 5 0.10 0 0 0 0 17 0.36 8 0.16
7 0 0 0 0 0 0 0 0 2 0.04 5 0.15 ----20 0.43 10 0.20
All
Nets 0 0 0
I
0 0 0 0 0 14 0.04 16 0.07 1 0.01 56 0.17 38 0.11
--··· -······----·--·'--------·-···--··--·--~···---·-· ~-~···-
* CPH -Catch Per Hour
SCALE
SPAWNING AREAS FOR DISTRIBUTION OF SPAWNING SALMON.
PINK AND .
CHUM SALMON KING SALMON PINK SALMOND FIGURE 6
17
Carcass Counts
Carcass counting as an indexing technique was unsuccessful. Only 10 pink
salmon {2 with tags) and 3 chum salmon were observed during the course of the
study. Rapid current and generally rising water levels during the peak of pink
salmon spawning apparently removed most carcasses from the study area.
Downstream reaches within the intertidal zone also retained few carcasses {only
1 pink salmon was observed) because of the flushing action of the tides. The
experiment a 1 chicken wire fences were marginally successful in trapping car-
casses. Four pink salmon and 1 chum salmon were caught by the fences during a
trial run on August 8-14. Immediately after this period, the water level
dropped leaving 1 fence out of the water. The subsequent rise in water level in
late August prevented checking of the fences for the remainder of the study.
Tag Retention
Only 1 probable lost tag was noted among the pink salmon recaptures from
hoop net and seine sampling, suggesting that retention of the Flay tags was
generally good. However, a survey of Fox Farm Creek, a clearwater tributary to
the Bradley River, on September 10 (after the peak of the pink salmon run)
resulted in the recovery of 15 tags that were lying loose on the bottom gravels.
This unexpected occurrence suggested that the mechanical action of digging ~ay
cause the salmon to shed their tags during redd construction. During the latter
stages of spawning, when the fish are near death, the muscle tissue in the upper
back becomes soft and the tags can pull out relatively easily.
Pink Salmon Stream Life Duration
The recapture of tagged fish throughout the study period provided substan-
tial information on the duration of freshwater life for individual pink salmon.
Among the recaptures, 89.6 percent were fish that had been marked during the 2
weeks prior to the recapture week suggesting that a relatively small proportion
of fish present in the river had been there for longer than 14-17 days. Only 4
percent of the recaptures had been present in the river for more than 3 wee~s.
One fish was recaptured 42 days after it was originally marked and 4 fish were
present in the river for periods ranging from 23-31 days.
18
Mark and Recapture Estimates
Pink Salmon
Simple Peterson population estimates (Seber 1973) were computed for each
study week based on the combined hoop net and seine catch (Table 6). These
estimates were based on the following assumptions:
1. The number of pink salmon within the study area remained stable during
the 3-day sample period for each week.
2. All fish marked during the previous 2 sample weeks were still present
in the study area.
3. Fish marked 3 weeks or more prior to the sample week were no longer
present.
4. Marked and unmarked fish were equally vulnerable to capture.
The Table 6 estimates imply that the number of pink salmon reached a peak
of about 2800 fish in week 7 (August 26-29). It should be emphasized that this
method is based on probability statistics and is subject to random variation.
Therefore, the abundance estimates should be viewed as a range of values (the
95% confidence limits) with the most likely values centering around the single
estimates provided in Table 6. As the number of recaptures increases the sta-
tistical reliability increases. However, if one or more of the assumptions is
seriously violated then the reliability of the estimates decreases independent
of the statistical reliability. As is evident from the assumptions, these esti-
mates need to be treated with caution (see Discussion).
Two part i cu 1 ar ly successfu 1 sets of seine hau 1 s (one in week 5 and one in
week 6) provided an opportunity to corroborate the Table 6 estimates:
Tot a 1 Number Population
Captured Marked Recaptures Estimate
Week 5 (Aug. 12) 51 308 9 17 45 ( 918-3927) *
Week 6 (Aug. 19) 42 381 5 3200 (1367-10001)*
*95% Confidence Limits
19
TABLE 6
POPULATION ESTIMATESa FOR PINK SALMON
BASED ON COMBINED HOOP NET AND SEINE CATCHES
FOR EACH SAMPLE WEEK
I Sample Total Total Population
Markedb Week Captured Recaptures Estimate
1 4 0 0
2 36 4 1 72
3 158 31 8 612
4 259 156 29 1418
5 270 308 51 1630
6 262 381 43 2321
7 236 426 36 2793
8 156 397 33 1876
9 68 277 14 1345
a Estimates were based on the Peterson method (Seber 1973)
Population =Marked Fish x Total Captures
Recaptures
---
( 25 -1440)C
( 310 -1440)
( 963 -2066)
(1240 -2144)
(1724 -3210)
(2019 -4005)
(1337 -2728)
( 801 -2446)
b Number marked during the 2 week period prior to the sample week.
c 95% Confidence Limits.
20
These estimates, although based on seine hauls at a single location, are reaso-
nably close to the combined hoop net and seine estimates.
There is no accurate way to estimate the total pink salmon escapement. If
it is assumed that the peak occurred in Week 7 and there is a complete turnover
of fish every 2 weeks, then a rough estimate can be made by adding the estimates
for weeks 4 and 7 together. Using these assumptions, the estimated total run
size is about 4200 pink salmon. Another approach, consistent with the method
used by A 1 ask a Department of Fish and Game to determine tot a 1 escapement from
weekly surveys (Pirtle and McCurdy 1980), is to sum the weekly population esti-
mates and divide by 2.5. Applying this method to the Table 6 estimates gives a
total escapement of 4826 pink salmon.
Chum Salmon
Population estimates for chum salmon using mark and recapture data from the
hoop nets are not meaningful. Chum ~almon were not tagged, consequently indivi-
dual fish could not be separated. It was, therefore, impossible to determine
whether marked fish had been marked prior to a sample week or within a sample
week. In addition, there was a tendency to catch the same fish over and over
again within any given hoop net in any given week. In the latter weeks of the
chum salmon run, a significant portion (50-70%) of the fish caught had been
marked previously (Table 1). During the study period, 246 unmarked chum salmon
were captured. The recapture data suggest that at least 50 percent of the fish
in the study area had been marked. The total number of chum salmon in the
Bradley River study area was very likely between 250 and 500 fish.
Pink Salmon Straying
Pink salmon with Bradley Lake Project tags were found at several locations
outside of the study area. A fish tagged in the Bradley River on August 14 was
recovered in Clear Water Slough, a tributary to the Fox River, on August 21
(Wunderlich, personal communication). Another tagged fish was seen in Clear
Water Slough but not captured. A third fish tagged in the Bradley River on
August 14 was caught by a fisherman in Mud Bay near Homer on August 22. Another
fish tagged on September 4 was recovered near Swift Creek on the north side of
Kachemak Bay on September 9. As mentioned earlier, 15 tags were recovered in
Fox Farm Creek, a tributary to the lower Bradley River, about 1.6 miles
downstream from the study area. It is evident from the above data that a
21
substantial number of pink salmon holding in the Bradley River study area even-
tually left the study area to spawn at other locations.
Miscellaneous Observations
Harbor seals (usually 1 or 2) were frequently observed in Eagle Pool and
Riffle Reach during the flood tide. Occasionally they ventured as far upstream
as the Fish Camp. On three occasions seals were caught in the hoop nets, but
were released unharmed. On one of the three occasions the seal had killed all
of the fish in the net, but in the other cases the fish were unharmed. Seals
obviously prey heavily on salmon in the Bradley River.
A family unit of four river otters was occasionally observed in Riffle
Reach. Signs along the river bank indicated that they frequently used the area
but no direct evidence of salmon predation was noted.
DISCUSSION
Comparison of 1986 Season with Prior Years
The apparent use of the Bradley River by pink salmon in 1986 indicates a
similar but somewhat higher abundance than seen in prior years. Study in 1983
(Woodward-Clyde Consultants) suggested a population of about 1000 fish and the
1985 field study (Northern Technical Services) indicated a population in the
range of 1000-3000 fish. Comparisons between this study and prior studies are
difficult because of varying methods and study timing, but the order of magni-
tude of fish numbers is similar and other observations regarding the biology of
pink salmon are consistent.
This study suggests that greater numbers of chum salmon utilize the Bradley
River than was previously suspected. Several hundred chum salmon spawned within
the study area in 1986.
About 30 chinook salmon were caught during the study, primarily in Bear
Island slough. The frequent recaptures and progressive condition of the fish
strongly suggests that spawning took place within Bear Island Slough and
possibly at other locations as well. Two previous studies also noted small num-
bers of chinooks in the slough (USFWS 1982, Woodward-Clyde 1983). The primary
question is whether these fish represent a small self-sustaining natural chinook
22
salmon run or whether these fish may be strays from the so-called 11 terminal
fisheries 11 in Kachemak Bay. Hatchery smolts are released in two locations, one
adjacent to Homer Spit and the other in Halibut Cove Lagoon; the adults return
to these locations but have no place to spawn. However, the timing of the
Bradley River run (early August peak) was several weeks later than the timing of
the terminal fishery or the timing of the natural chinook salmon returns to
Kachemak Bay streams (Oudiak, personal communication). Another possibility is
that wandering fish have recently established a reproducing population in the
Bradley River. However, such fish would likely take on the spawning phenology
of the parent stock (from the Crooked Creek hatchery) and spawn earlier in the
year. If the Bradley River chinook salmon run is a natural occurrence, it is
the only documented 1 ate run in Kachemak Bay (Hammarstrom, persona 1
communication). Juvenile chinook salmon were identified in one prior study in
the Bradley River (Woodward-Clyde 1983). A few suspected juvenile chinooks were
caught during this study in the beach seine, however identification was not con-
firmed. Laboratory examination is required to conclusively separate coho from
chinook juveniles.
The few sockeye salmon caught in this study were suspected to be strays
from other systems since there was no evidence of repeated capture or maturation
within 'the Bradley River. This finding is consistent with that of other
investigators.
Few conclusions can be drawn regarding coho salmon because the study period
only overlapped the early part of the run. Significant numbers were entering
the river and some were approaching spawning condition. Prior studies have not
sampled coho salmon extensively; consequently, coho use of the Bradley River is
essentially unknown.
Abundance Indices
The composite catch-per-unit-effort data from the hoop net sampling appears
to be a reasonable indication of salmon numbers within the study area that can
be compared from year to year as long as the river conditions remain reasonably
constant. Unfortunate 1 y, the extreme flooding that occurred in October after
completion of the study has already altered the course of the river signifi-
cantly and two of the hoop net sites will not be strictly comparable in future
years. Nevertheless, under average summer flow conditions hoop nets deployed in
23
a manner consistent with the methods of this study will provide an index of
abundance for pink and chum salmon and provide some information for the other
species as well. Abundance indices derived from hoop net catch per hour are
expected to be a significantly more reliable indicator of relative abundance
from year to year than mark and recapture estimates and should be relied on in
future years where possible.
During the operation of the Bradley Lake Hydroelectric Project summer flow
in the lower Bradley River will be substantially reduced and at times may reach
the required minimum of 100 cfs, or less than 10 percent of existing flow.
Continued use of hoop nets will have to be reviewed in light of the much reduced
flow, and the hoop nets, if used, may need to be redesigned for shallow water
conditions. Trap nets with rectangular frames 2-3 feet high may be adaptable
for use under the rev1sed flow regime at least in some locations.
The other capture methods utilized were not suitable as a means of
providing a reliable index of abundance under the river conditions that existed
in 1986. Beach seining showed some potential but changing flow conditions and
difficult deployment in rapid current conditions prevented a consistent level of
effort. A modified seine with coarse mesh (less water resistance) may improve
the utility of seining as an index technique under pre-operational conditions.
Beach seining may be a feasible indexing method during the lower, more con-
sistent flow regime that will be present during project operation.
Mark and Recapture Estimates
The dynamic population situation that exists within a stream during a
salmon spawning run does not provide good conditions for accurate estimates of
fish numbers using mark and recapture methods. Single census mark and recapture
estimates depend on the validity of several assumptions:
1. A closed population, i.e. there should be no additions or deletions to
the population during the estimate period.
2. Tagged and untagged fish are equally vulnerable to capture which
imp 1 i es that marked fish mix randomly with unmarked fish and that
sampling techniques catch a random sample of fish from the study area.
3. Marked fish are consistently recognized.
24
It is evident that assumption 1 is highly questionable since during the
early part of the run new fish are continually moving into the area and during
the latter part of the run fish are continually dropping out because of mor-
tality. The methods used in this study attempted to minimize the impact of
these factors by conducting a series of discrete estimates each covering a short
period of time. In the situation where there is recruitment to the population
since the fish were marked but no mortality of marked fish, the population esti-
mate will be a valid estimate of the number at the time of the recapture sample
(Seber 1973). Since the recapture period only encompassed 3 days it was assumed
that changes to the population due to immigration of new fish would be minimal
during the recapture period. The impact of mortality was minimized by only con-
sidering fish marked in the 2 weeks prior to the recapture period as being part
of the marked population. When both recruitment and mortality occur, the popu-
lation size will be overestimated (Seber 1973). It is likely that the 2-week
criteria applied to marked fish underestimated the number of marked fish present
in those weeks prior to the peak of spawning and overestimated the number after
the peak when mortality was highest. Consequently, the estimates for the early
weeks of the study (Table 6) may have underestimated numbers of pink salmon
while the estimates in the last two weeks may have overestimated numbers pre-
sent.
Another factor to consider, especially with regard to the earlier esti-
mates, is the fact that some tagged fish left the study area to spawn elsewhere
thus reducing the number of marked fish available for recapture. This effect
would tend to cause overestimation of actual numbers and may offset to some
extent any underestimation caused by the conservative 2-week criteria as applied
to the estimat~s in weeks 2-7. It may be worthwhile in future years to try and
estimate the proportion of tagged fish that leave the study area. At a minimum,
the number of tagged fish spawning in Fox Farm Creek should be observed.
Assumption 2 also presents some problems. The hoop nets tended to catch
fish moving upstream or fish that were milling in the vicinity of the nets.
Seine samples tended to catch fish that were spawning or holding within specific
riffle areas. Consequently, there may have been a tendency to catch marked fish
at a higher frequency than would be expected on a purely random basis because
the same stream areas were being repeatedly sampled. This tendency would cause
the population estimates to be lower than actual populations. However, the
genera 1 fee 1 i ng among the investigators in regard to pink salmon is that the
25
marked fish were well mixed within the total population at least during the
early part of the run prior to actual spawning.
Assumption 3 was true for pink salmon. Tags were easily identified and all
unmarked fish were tagged in the same place making it highly likely that fish
with lost tags were also identified as having been tagged by the presence of the
needle scar.
Although the mark and recapture estimates provide problems of interpreta-
tion, they do provide an order of magnitude indication of real population size
which will be useful in comparing the 1986 pink salmon escapement to future
years. These estimates may be especially important since the post-operational
flow regime will be very different from that which now exists making hoop net
catch index values difficult to compare between pre-operational and post-
operational study years.
RECOMMENDATIONS FOR FUTURE YEARS
1. Continue use of the hoop nets during the pre-operational monitoring
phase as deployed in 1986 as the most reliable means of indexing popu-
lation abundance. However, the existing hoop nets may not be appli-
cable after the project is in operation. Post-operational flows may
be as low as 10 percent of current summer flows, consequently working
conditions on the river will change after the dam is in place. Depth
of water may be too shallow to use the current net design. It may be
possible to use a modified shallow net design during project opera-
. tion or a combination of net designs depending on site charac-
teristics.
To aid planning for future conditions, a reasonable post-operational
summer flow scenario should be predicted for the lower Bradley River
based on existing hydrological information. A visit to the Bradley
River study area in the fall or winter when flow is similar to the
predicted post-operational summer flow would provide a firsthand indi-
cation of what post-project river conditions will be like.
2. Continue use of beach seining as a supplemental index technique and as
a means of sampling for mark and recapture estimates. A custom beach
26
seine should be constructed for use during the 1987 season. The seine
should have 2.5 inch mesh netting, no bag, extra heavy lead line and
large floats. Dimensions should be about 120 feet by 6 feet.
Additional beach seining in the lower river below the study area
during peak migration should be considered to provide an indication of
the extent to which tagged fish are leaving the study area.
3. Discontinue carcass counts as a sampling methodology.
4. Discontinue use of the boat-mounted shocker but retain the backpack
shocker for use as a supp 1 ementa 1 capture technique and for de 1 i-
neating fish distribution.
5. Tag all salmon captured during the study (in 1986 only pink salmon
were tagged). Tagging all the salmon will entail no extra field
effort and will allow the recovery of substantially more biological
information on the other salmon species.
6. Survey salmon use of Fox Farm Creek once per week to determine the
proportion of tagged fish that move from the study area to the creek.
Visual observations can be easily conducted at low tide.
7. Catch and examine in the lab a sample of juvenile salmonids from Bear
Island Slough to confirm species indentification (Coho vs. Chinook)
and determine if young chinook salmon are present. This will help
determine if the Bradley River supports a self-perpetuating population
of chinooks.
8. Continue the use of mark and recapture estimates. Such estimates are
difficult to apply to dynamic populations of spawning salmon in a
stream; nevertheless, the proportion of marked to unmarked fish repre-
sents the only 1 ink to actua 1 numbers of fish present. These est i-
mates may be valuable when comparing pre-operational to
post-operational conditions since indexing methods may not be com-
parable because of the difference in flow regimes.
9. Develop a computerized data handling system for analyzing and tabu-
lating field data. The 1986 study generated substantially more data
than was anticipated based on the results of previous studies.
27
Standard fie 1 d data sheets designed for computer entry a 1 ong with
software to sort, analyze, and compute would save substantial time in
the long run and allow a more thorough data analysis on a smaller
budget.
10. Maintain the 3-person work crew as the basic work team. A crew of 3 ,,
persons is essential for the safe and efficient deployment of hoop
nets under the existing conditions in the river. Three days of inten-
sive field work during each calendar week appears to be a reasonable
approach to obtaining the needed information. During the much lower
flows that will be present during project operation it may be possible
to reduce manpower requirements.
11. Maintain the 1986 study timing and duration. The schedule of study
for 1986 coincided well with the pink salmon run. If additional
information is desirable regarding chum salmon escapement, then the
study should start about two weeks earlier. If additional information
is desirable regarding the coho salmon population, then the study
should continue at least until the end of September. Extension of the
study period into the fall could, however, entail risk to personnel
and sampling difficulties because of the erratic and often high water
levels associated with fall rains.
12. Study logistics and major equipment items as employed in the 1986
study were appropriate and should be maintained except for minor
changes. Use of a field camp within the study area is the least
expensive and most efficient arrangement and should be continued. A
larger and more stable jet-powered boat would be desirable.
28
REFERENCES
REFERENCES
Alaska Power Authority. 1986. Salmon monitoring plan for Fed. Energy Regulatory
Commission Proj. No. P-8221-000. By: Stone and Webster Engineering Corp.,
Anchorage, AK.
Dudiak, N. (Alaska Department of Fish and Game). 1986. Personal communication
with David Erikson, Dames & Moore.
Hammarstrom, S.L. (Alaska Department of Fish and Game). 1986. Personal com-
munication with David Erikson, Dames & Moore.
Northern Technical Services, Inc. 1985. 1985 Salmon Escapement Survey Report.
Prepared for Alaska Power Authority, Anchorage, Alaska. 29 pp.
Pirtle, R.B. and M.L. McCurdy. 1980. Prince William Sound general districts
1976 pink and chum sa 1 mon aeri a 1 and ground escapement surveys and con-
sequent brood year egg deposition and pre-energent fry index programs.
Alaska Department of Fish and Game Technical Data Report No. 51.
Seber, G.A. 1973. The estimation of animal abundance and related parameters.
Griffin and Co. Ltd., London. 506 pp.
U.S. Fish and Wildlife Service. 1982. Appendix B: Bradley Lake Hydroelectric
Project. Homer, Alaska. Final Coordination Report. USFWS Western Alaska
Ecological Services. Anchorage, AK. 131 pp. In: U.S. Army Corps of
Engineers. 1982. Bradley Lake Hydroelectric Project, Alaska. Final
Environmental Impact Statement. Alaska District, U.S. Army Corps of
Engineers. ·
Woodward-Clyde Consultants. 1983. Bradley River instream flow studies.
Prepared for Alaska Power Authority, Anchorage, Alaska. 77 pp.
Woodward-Clyde Consultants. 1984. Report on salmon fry sampling in the Bradley
River. Prepared for Alaska Power Authority, Anchorage, Alaska. 6 pp.
Wunderlich, M. (U.S. Fish and Wildlife Service). 1986. Personal communication
with John Morsell, Dames & Moore.
29