HomeMy WebLinkAboutA03_tarbooFishSurvey NW Watershed
Fish Distribution and Abundance in Shallow Intertidal
Habitats of Tarboo and North Dabob Bays
Prepared for: Jefferson County Marine Resources Committee
ATT: Pat Pearson, Water Quality Agent
WSU Cooperative Extension
201 West Patison Street
Port Hadlock WA 98339
Prepared by: Peter Bahls, Aquatic Ecologist
Northwest Watershed Institute
2215 SE 55th Avenue
Portland, Oregon 97215
February, 2004
INTRODUCTION
Background
The Tarboo and north Dabob Bay study area extends from the mouth of Tarboo
Creek at the head of Tarboo Bay to include Broad Spit and Camp Discovery Creek in
Dabob Bay to the south (Figure 1). The study area (hereafter referred to as Tarboo-Dabob
Bay) may represent one of the highest quality and biologically important estuarine
environments in Puget Sound. Three large saltmarsh spits extending from the sides of the
bay create a diversity of hydraulic conditions and habitats. The southern most spits are
considered the separation between Tarboo and Dabob Bay. Inside the spits, shallow
mudflats and saltmarsh lagoons predominate. Outside of the spits, in Dabob Bay proper,
the seafloor plunges to more than 500 feet deep and shorelines are generally comprised of
the gravel and sand beaches typical of Hood Canal, with eelgrass beds occurring
offshore. Numerous small streams along the shorelines enter Tarboo-Dabob Bay.
Figure 1. Tarboo-Dabob Bay study area looking south.
1
Unlike many other parts of Puget Sound, the Tarboo-Dabob estuary remains in
remarkably pristine condition, with an abundance of high quality habitat for juvenile
salmonids and a variety of marine fish. About 400 acres are protected as state-owned
Natural Area Preserves, including the lower mile of Tarboo Creek and the coastal spits
and adjoining upland forest. However, the existing preserves provide only piecemeal
protection for the Tarboo-Dabob Bay ecosystem as a whole. With increasing
development expected in the watershed in the decades ahead, a more comprehensive
approach is needed to ensure that the water quality and aquatic habitats are protected over
the long term.
In 2001, Northwest Watershed Institute initiated the Tarboo Watershed
Assessment with 14 project partner organizations as the first step in a comprehensive
conservation project. The purpose of the Tarboo Watershed Assessment was to conduct
an integrated and science-based analysis of both Tarboo Creek and Tarboo-Dabob Bay to
identify high priority protection and restoration projects that could be conducted with the
participation of willing landowners. The assessment also aimed to obtain the baseline
data necessary to monitor the health of the watershed over time.
A first step in the assessment was to gain an understanding of the distribution and
abundance of fish species in the marine portion of the project area. In Tarboo-Dabob Bay,
no previous fish surveys had been conducted, with the exception of surveys conducted by
the Port Gamble S’Klallam Tribe at a single sample site in the southeast corner of the
study area (Hirschi et al. 2003). In this survey, the mouth of Camp Discovery Creek was
sampled by beach seine from December 2001 through October 2002 as part of survey of
North Hood Canal tidal creeks and independent marshes. Hirschi et al. (2003) reported
finding several Chinook salmon at this site. In addition, at many of their sample sites in
Hood Canal they caught juvenile chum salmon in January and February. The unusually
early occurrence of these chum along the shorelines suggested that they were the
offspring of summer chum, which spawn one to two months earlier than fall chum. These
findings further underscored the importance of investigating fish use within Tarboo-
Dabob Bay, an area that may be especially important as non-natal rearing areas for
juvenile salmonids.
2
Funding for this study was provided in part by the Jefferson County Marine
Resources Committee (JCMRC) of Washington State. The JCMRC is composed of
citizen representatives that were organized to address local marine issues and recommend
protection and restoration actions to Jefferson County and other appropriate
organizations. In 2003, the JCMRC identified Tarboo Bay-Dabob Bay as one of the three
priority areas for conservation in Jefferson County.
Objectives
The purpose of this study was to better understand fish use, distribution, and
abundance in Tarboo-Dabob Bay during the spring and early summer seasons when
salmonid were expected to be present in the estuary. This baseline information on fish
use can then be used in regional and watershed-specific conservation and restoration
efforts.
Specific objectives of the project were to:
1) Identify fish species using shallow intertidal shoreline areas during the February –
July sampling period, when juvenile salmonids were most likely to be present.
2) Better understand the distribution and abundance of salmonids and principle
forage fish in the study area in relation to general habitat types and seasonal
change.
METHODS
Sampling sites and timing
Twenty-five survey sites were selected in Tarboo-Dabob Bay with the intent of
sampling a diversity of shallow intertidal habitat types distributed throughout the project
area (Table 1). Sites were classified into habitat types based on a recent literature review
of the habitat preferences of salmonids (Anchor Environmental 2002; Simenstad 1983,
1985, 1991, 2000; Williams and Thom 2001). Saltmarsh-mudflats and stream deltas were
considered the most important habitats, whereas the sand and gravel beaches areas typical
of higher energy coastal spits and Dabob Bay shorelines were considered lower priority.
3
Sampling was conducted once a month for a one to three day period from
February 2003 through July 2003. Sampling generally ranged from 4 to 8 hours a day.
Because Tarboo Bay was too shallow to sample at moderate or low tides, we only
sampled during the high tide periods when tidal elevation was at least 8 feet. Seven
“standard” sites were sampled consistently every month, with the exception of two sites
that were not sampled in July. Twenty “supplemental” sites were sampled as time and
weather permitted (Table 1 and Figure 2).
Field procedures
We used field procedures and net specifications developed by the Skagit System
Cooperative Research Department. They developed a small net beach seine method for
sampling shallow intertidal shoreline areas (Skagit System Cooperative Research
Department 2003). We accessed sampling sites with a small skiff and outboard motor.
The areas seined were typically less than 4 feet deep and had relatively homogenous
habitat features such as: water depth and velocity, substrate, and vegetation. We
employed an 80-foot long, 6-foot deep net with 1/8-inch knotless mesh and weighted
bottom line and floating cork top line. To set the net, one person held one end of the net
on the beach while the other played out the net from a floating tote while wading out
from the beach (Figure 3). The net was set in a half circle “upstream” against the tide,
with both ends ending at the beach. Two to four persons then pulled the net in to the
beach (Figure 4). Small catches of fish were sampled directly from the net, while large
numbers of fish were held for sampling in a bucket or in a pocket of the net draped inside
a floating 3 by 4-foot frame of PVC pipe (Figure 5).
Three hauls were made at each sampling site. Each haul was located 20-60 meters
apart along the beach. Fish were identified to species if possible, and total length
measured for the first 10 to 20 fish of each species. Habitat variables were recorded for
each set, including salinity, sea surface temperature, substrate composition and shoreline
vegetation types. Data was entered onto a field data form (Appendix A), then into
computer Excel spreadsheet format. The location of each sample site was recorded in the
field using a handheld GPS and used to help map the site locations in GIS.
4
Table 1. Sampling sites listing habitat type and months sampled (dark boxes). Standard
sites were sampled from January through June or July (in bold type).
Month sampled
Site
ID Site Name Habitat Type FebMarAprilMayJuneJulyTotal1 Broadspit Point Spit 3
2 Broadspit Lagoon Saltmarsh 2
3 Broadspit Creek Beach-stream 4
4 West Dabob Rocks Beach 5
5 West Dabob Landslide Beach 5
6 Cabin Creek Beach-stream 6
7 South Spit Corner Spit 1
8 South Side of South Spit Spit 2
9 South Spit Point Spit 3
10 Spits Lagoon Saltmarsh 6
11 North Spit Point South Side Spit 6
12 North Spit Point North Side Spit 2
13 West Tarboo Bay Cove Saltmarsh 1
14 West Tarboo Floathouse Mudflat 3
15 Tarboo Creek Mudflat-stream 6
16 Spruce Creek Mudflat-stream 5
17 Boathouse Point Beach 6
18 Lagoon Creek Mudflat-stream 4
19 Long Spit East Side South End Saltmarsh 2
20 Long Spit East Side by Island Saltmarsh 4
21 Long Spit East Side Lobes Saltmarsh 5
22 Long Spit Point East side Spit 1
23 Long Spit Point West side Spit 1
24 Long Spit Beach Spit 3
25 East Dabob Landslide Beach-mudflat 3
Total Number of Surveys 1913161519 7 89
5
Figure 2. Survey site locations indicating number of surveys conducted at each site.
6
Figure 3. Setting the net in a semi-circle by playing it out from a floating tote.
Figure 4. Hauling in the net.
7
Figure 5. Sampling fish held within the PVC floating frame.
RESULTS AND DISCUSSION
General distribution, abundance, and seasonal change
A diversity of fish species was caught over the survey season. Table 2 lists the
fish species caught at all sites, the percent of sites at which a species was found, and the
total number of each species that was caught. Juvenile Chinook, coho, chum, and
cutthroat trout were identified. Juvenile forage fish, including surf smelt, sand lance, and
herring were caught. Six species of sculpin were identified. Because some sites were only
sampled one or two times and others were sampled as many as six times, a comparison of
the distribution and abundance of species using all data from all sites may be biased.
Nevertheless, this data is informative in showing general patterns of species distribution
and abundance.
8
Table 2. Total number of each fish species collected, and percent of the total sites at
which the species was found (percent occurrence).
Species
Percent
occurrence
Total
number
Coho salmon (Oncorhynchus kisutch) 48 37
Summer chum salmon (Oncorhynchus keta) 96 6821
Fall chum salmon (Oncorhynchus keta) 88 18005
Chinook salmon (Oncorhynchus tshawytscha) 16 5
Cutthroat trout (Oncorhynchus clarki clarki) 28 11
Pacific herring ( Clupea harengus pallasi) 12 4
Surf smelt (Hypomesus pretiosus) 36 136
Sand lance (Ammodytes hexapterus) 20 182
Bay pipefish (Syngnathus griseolineatus) 24 10
Flatfish (Psettichthys melanostictus) 4 1
Starry flounder (Platichthys stellatus) 36 18
Sand dab (Citharichthys sordidus) 4 1
Gunnel (Pholis sp.) 20 11
Shiner perch (Cymatogaster aggregata) 64 6446
Pacific snake prickeback (Lumpenus sagitta) 68 755
Three-spine stickleback (Gasterosteus aculeatus) 32 15
Buffalo sculpin ( Enophrys bison) 4 1
Great sculpin ( Myoxocephalus polyacanthocephalus)68 223
Fluffy sculpin (Oligocottus snyderi) 16 9
Saddleback sculpin (Oligocottus rimensis) 20 11
Staghorn sculpin (Leptocottus armatus) 100 772
Tidepool sculpin (Oligocottus maculosus) 8 6
Sculpin sp. 24 54
9
The abundance of fish caught changed dramatically over the season. Table 3
summarizes the total number of fish caught each month for selected species. This
summary included only the seven standard sites – five of which were sampled every
month, and two of which were sampled every month except July (when very few species
and no salmonids were caught at other sites). The consistent sampling at multiple sites
provided a reliable estimate of the change in abundance and use of shallow intertidal
habitats over the season. However, without more extensive sampling of deeper water
areas, we cannot determine whether the fish that were not sampled were in deeper water
or out of Tarboo-Dabob Bay completely.
Table 3. Seasonal changes in abundance for selected fish species at standard sampling
sites.
Month sampled Total
Species Feb Mar Apr May June July Catch
Coho 0 0 0 29 1 0 30
Chinook 0 1 1 1 1 0 4
Summer chum 357 232 255 1 0 0 845
Fall chum 0 5033 4795 160 0 0 9988
Cutthroat 0 0 4 3 0 0 7
Surf smelt 6 3 0 4 0 0 13
Pacific herring 0 0 0 1 2 0 3
Sand lance 0 0 5 0 0 0 5
Shiner perch 0 0 7 205 267 62 541
Staghorn sculpin 35 31 48 23 91 22 250
Snake prickleback 0 0 0 2 443 8 453
Chum salmon
Chum salmon were separated for identification purposes into summer and fall
chum stocks based on timing and size of the fish. All chum sampled in February were
considered for all intents and purposes to be summer chum (Figure 6). These chum were
approximately 40 mm in length and were found to be the same size as summer chum that
10
Figure 6. Summer chum salmon juveniles sampled in February, 2003.
were being reared at the Quilcene Hatchery at the time. The fish we caught were probably
wild summer chum since none of the Quilcene hatchery-reared summer chum were
released until several weeks after our mid-February sampling (L. Tallis, pers. comm.
2003). In March, we observed two distinct size classes of chum, a newly emerged chum
of about 35 mm, and chum about 55 mm. All chum over 50 mm were assumed to be
summer chum due to their much greater size. During the April sampling, the distinct size
differences began to break down. We continued to consider all chum over 50 mm as
summer chum. However, a few noticeably larger chum of 60 to 70 mm were also
observed and may have represented the true summer chum. In May, only one large 70
mm chum was found and was assumed to be a summer chum. If our identification of
summer chum based on size and timing is correct, then summer chum inhabited Tarboo-
Dabob Bay for over three months, from late January through early May.
11
Figure 7. Marine survey results for salmonid distribution and abundance.
12
Summer chum had a surprisingly wide distribution. In the first month of sampling
in February, summer chum salmon were found at all sites sampled except the mouth of
Tarboo Creek (Figure 7). However, the catch was highly variable between hauls and
between sites. A few fish were caught at most sites, with substantially higher numbers
caught at a several sites. The relationship between chum abundance and habitat type was
not obvious and the variability may simply have reflected the clumped distribution of fish
that were traveling in schools. However, the highest numbers of fish caught in February
were from saltmarsh lagoon sites that fish may have preferred. Especially considering the
small size of these fish, it is striking to note that at low tide most of Tarboo Bay is
completely de-watered mud flats and that at high tides fish move in on the high tide to
populate the inner bay. We are planning to conduct statistical analysis to determine if
summer chum were found in significantly higher numbers in certain types of habitat than
others. This may help us answer the key question of whether summer chum are
preferentially selecting areas such as Tarboo Bay for extended rearing.
On January 28, 2004, we started a second season of sampling at the seven
standard sites (data not included in this report) and caught a few newly emerged chum
with belly slits barely healed at sites 7,15,16. These were all sites located at the mouths of
freshwater streams, two of which were inside Tarboo Bay. Either these fish were
summer chum that traveled significant distances soon after emergence from natal streams
such as the Big Quilcene River or alternatively, these fish may have represented an
unusually early emergence of fall chum that spawned in local streams. Genetic testing is
needed to confirm whether these chum salmon inhabiting the estuary in January and
February are indeed summer chum. That finding would not be surprising according to the
leading expert on Hood Canal summer chum salmon (C. Simenstad, personal
communication 2003).
13
Coho salmon
Thirty-seven coho juveniles were caught at 12 sites over the survey season (Table
2 and Figure 7). The fish ranged from 38 –133 mm in length, with 8 coho under 41 mm.
These results corroborate the finding of young-of-the-year coho in estuarine areas by
Hirschi et al. (2003) and others.
Most of the coho were sampled during the May survey (Table 3). The timing
corresponded to the timing of smolts out-migrating from Tarboo Creek. In a separate
beach seine survey of the lower portion of Tarboo Creek, 115 coho smolts ranging in size
from 76 to 152 mm were found on April 28; 33 coho were found in a subsequent survey
on May 11; and no coho were sampled on May 25. In addition to a Tarboo Creek origin,
two fish sampled in marine waters had clipped adipose fins, and may have come from the
Quilcene hatchery or netpen rearing operation.
The broad size range of the coho sampled in lower Tarboo Creek and in Tarboo-
Dabob Bay is remarkable, and may reflect a third life history strategy; to out-migrate
after two winters in freshwater, instead of just one. This theory is supported by our
observations of the sizes of coho juveniles in freshwater. During summer snorkel surveys,
we found a small fraction of coho juveniles that were significantly larger than most of the
other coho and were probably into their second year in freshwater.
Chinook salmon
Five Chinook salmon were found at four sites, three of which were located within
Tarboo Bay (Table 2 and Figure 7). Chinook were caught for four months, from March to
June (Table 3). They ranged in size from 37 mm (caught in March) to 80 mm (Figure 8).
The 37 mm Chinook, and perhaps the others, probably out-migrated from Tarboo Creek.
A small number of Chinook adults have been observed spawning in the lower mile of
Tarboo Creek every season that the stream was surveyed between 1994 and 2003 (7 of
ten years) (Bahls 2003). This population may have originated from a release of hatchery
smolts from a rearing pond project in the Tarboo valley in the early 1990s. Sampling
these small Chinook juveniles in Tarboo Bay adds further weight to the theory that these
fish are naturally reproducing at some level.
14
Figure 8. Young-of-the-year Chinook salmon caught in Tarboo Bay.
Fall chum salmon
Fall chum were the most abundant fish sampled. During March and April, fall
chum were “swamping” Tarboo-Dabob Bay, with approximately 10,000 fish caught at
seven sites and about 20 times more fall chum than summer chum sampled during that
period (Table 3). These fish probably originated in the Quilcene hatchery and hatcheries
elsewhere in Hood Canal, as well as out-migrating from Tarboo Creek, where there is a
strong naturally spawning population in the lower Creek (Bahls 2003).
Shiner Perch, Snake Prickleback, and Staghorn Sculpin
Three fish were relatively abundant later in the season. Numbers of Shiner perch
peaked in May and June (Table 3). Although they were abundant, they were not as
widespread as might be expected. Apparently, they come in to shallow water beaches
during this time of year to bear live young, as most of the fish caught were young of the
year or pregnant. The Snake prickleback seemed to prefer silt substrates. It was relatively
scarce until June, when it was caught in abundance. This fish could easily be
misidentified as a blenny (C. Moffett, pers. comm. 2003). Staghorn sculpin was the most
15
ubiquitous species, sampled in moderate numbers at most sites throughout the season,
with a peak in June. By July, these three species were the only species sampled.
Forage fish
We sampled low numbers of post larval sand lance, surf smelt, and Pacific herring
at a low percentage of sites in Tarboo and Dabob Bays (Table 2 and 3). It is not
surprising to find these fish, given recent forage fish surveys. In a comprehensive survey
of the beaches of Eastern Jefferson County, Kevin Long of North Olympic Salmon
Coalition, documented that some of the highest densities of forage fish spawning
occurred along the shorelines of North Dabob Bay (K. Long, pers. comm. 2003). Adult
forage fish were not found.
Three spine stickleback
A total of 15 three-spine sticklebacks were caught over the survey season (Figure
9). Surprisingly, this fish has not been found in intensive surveys in Tarboo Creek itself.
Yet, the stickleback is quite common in slow, warm water sections of Chimacum Creek,
immediately to the North of Tarboo Creek.
Figure 9. A large three-spine stickleback caught in Tarboo-Dabob Bay.
16
SUMMARY AND RECOMMENDATIONS
This study provides the first comprehensive documentation of the use of Tarboo-
Dabob Bay by a diversity of marine and anadromous fish species. Juvenile chum salmon
were found along the shorelines at almost all sites as early as the first sampling in
February 2003 and again in January of 2004. This early timing and the size of the chum
indicate that they were probably summer chum. However, genetic analysis should be a
high priority to confirm the stock identity. Summer chum may have been more abundant
in salt marsh habitats than Dabob beaches, but catches were highly variable from site to
site and more statistical analysis is needed to evaluate habitat preferences. By March and
April, smaller fall chum seemed to swamp the system, perhaps with negative effects on
summer chum. Most coho smolts were caught in May and most probably originated from
Tarboo Creek, with a fraction originating from hatcheries. Low numbers of Chinook
salmon juveniles were also found in Tarboo Bay from March through June. These
Chinook may also have originated in Tarboo Creek, where small numbers of Chinook
spawners have been documented annually since the early 1990s. A variety of other fish
species was captured, with staghorn sculpin being the most ubiquitous. The survey results
document that Tarboo Bay, with its extensive and diverse saltmarsh, mudflat, coastal spit,
and beach habitats, is important nursery habitat for juvenile salmon and a diversity of
marine fish.
This study has important implications for conservation. If the identification of
summer chum is correct, regional restoration strategies for federal listed salmon, such as
the Hood Canal Coordinating Council’s Salmon Recovery Strategy, need to recognize
that summer chum may out-migrate from their natal stream and adjacent estuarine habitat
within a matter of days or weeks to range widely in Hood Canal and spend up to three
months in non-natal rearing areas. High quality non-natal estuaries like Tarboo-Dabob
Bay may be more important habitats for summer chum survival than impacted natal
estuaries. Furthermore, the findings here generally support the existing research that has
found that stream mouths, mudflats, saltmarshes, and coastal spits are especially
important habitats for salmonids (Anchor Environmental 2002). Finally, although many
biologists have pointed out the importance of protecting shoreline and estuarine
17
vegetation, protecting the water quality of the numerous small and large streams that
enter the shoreline and provide the variable mix of fresh and saltwater for juvenile
salmonids may be equally crucial.
ACKNOWLEDGMENTS
This project was made possible by funding from the Jefferson County Marine
Resources Committee, Washington State Department of Ecology, and Northwest
Watershed Institute. The Port Townsend Marine Science Center (PTMSC) provided
invaluable assistance with the field surveys. I would especially like to thank Anne
Murphy, PTMSC Director, Cinamon Moffett, Volunteer Coordinator, and citizen
volunteers Rebekah Cadorette, Logan Cadorette, Dick Barrows, Erin Kenny, Jerry Proutt,
Kevin Long, Jeff Delia, Steve Blazina, Chad Witheridge, Merideth Barrett, and Sean
Gallagher. Generous scientific and technical guidance was provided George Pess, Casey
Rice, and Martin Liermann of NOAA’s Northwest Fisheries Science Center, and Eric
Beamer and Rich Henderson of the Skagit System Cooperative Research Department.
Phil Dinsmore, Dinsmore GIS Services, created the GIS maps.
LITERATURE CITED
Anchor Environmental L.L.C. 2002. Oakland Bay and Hammersley Inlet Nearshore
Habitat Assessment. Prepared for Squaxin Island Tribe and Taylor United
Shellfish Company by Anchor Environmental L.L.C, Seattle, WA.
Bahls, P. 2003. Salmon spawning ground counts in streams of upper Hood Canal, 1997-
1998 survey season. Point No Point Treaty Council Technical Report No. TR
03-02. Kingston, WA.
Hirschi, R., T. Doty, A. Keller, and T. Labbe. 2003. Juvenile salmonid use of tidal creek
and independent marsh environments in North Hood Canal: summary of first
year findings. Port Gamble S’Klallam Tribe, Kingston, WA.
Simenstad, C.A. 1983. The ecology of estuarine channels of the Pacific Northwest
coast: A community profile. FWS/OBS-83/05. U.S. Fish and Wildlife Service,
Olympia, WA.
Simenstad, C.A., K.L. Fresh, and E.O. Salo. 1985. The role of Puget Sound and
Washington coastal estuaries in the life history of Pacific salmon: an
18
unappreciated function. Pgs. 343-363, in: Kennedy, V.S. (ed.) Estuarine
comparison, Academic Press, New York.
Simenstad, C.A., C.D. Tanner, R.M. Thom, and L.L. Conquest. 1991. Estuarine Habitat
Assessment Protocol. Prepared for U.S. Environmental Protection Agency,
Region 10, Office of Puget Sound, Seattle, Washington.
Simenstad, C.A. 2000. Estuarine landscape impacts on Hood Canal and Strait of Juan de
Fuca summer chum salmon and recommended actions. Pgs. A3.111-132, in
WDFW and PNPTC Summer chum salmon conservation initiative, available at
www.wa.gov/wdfw.
Skagit System Cooperative Research Department. 2003. Estuarine fish sampling
methods. Skagit System Cooperative Research Department, La Conner, WA
Williams, G.D. and R.M. Thom. 2001. White Paper: Marine and Estuarine Shoreline
Modification Issues. Submitted to Washington Department of Fish and
Wildlife, Washington Department of Ecology, and Washington Department of
Transportation.
19
APPENDIX A. Nearshore Field Sampling Data Form
20
21
TARBOO WATERSHED PROJECT NEARSHORE FISH SAMPLING DATA FORM 2003
Northwest Watershed Institute, phone 503-235-2716
Location # ____ Name ________________________________________________ Date _________-03
Location Description ___________________________________________________________________
GPS ________________ , _________________ Wind _____, Air oC ____, Rain ____ %Cloud ______
Notes ________________________________________________________________________________
_____________________________________________________________________________________
Variable Set 1 Set 2 Set 3 Set 4
Time set
Time finish
Salinity (ppt)
Surface sea temp. (oC)
Max. depth of set (m)
Set size (100%=1.0)
Substrate types (OR,
SI, SA, GR, CO, BO,
Shoreline veg (Emerg.
SM, Spit, Forest)
Set#Species Total # Total length