Juvenile Salmon use of Sinclair Inlet, Washington in 2001 and 2002

In 2001 and 2002, we studied the distribution, abundance, size, and trophic relationships of juvenile salmonids in the marine nearshore environment of Sinclair Inlet, Washington. This study was initiated to increase our understanding of how salmonids use shoreline environments in Puget Sound and how shoreline development may impact these species, information that is needed to help management agencies develop effective recovery strategies now required under the Endangered Species Act.

We focused our study on sub-yearling juvenile Chinook salmon (Oncorhynchus tshawytscha) because this species has been listed within Puget Sound as Threatened under the Endangered Species Act, and represents an important local economic, recreational and cultural resource. In 2001 and 2002, we conducted studies to increase our understanding of the use of nearshore habitat and food resources by juvenile salmonids in Sinclair Inlet, a narrow embayment located in central Puget Sound bordering Bremerton, Washington. Specific objectives of this study were to:

• Assess the spatial and temporal use of littoral (nearshore, shallow water) habitats by juvenile Chinook salmon and other juvenile salmonids during the time these fish occur in the Inlet;
• Assess the use of offshore (i.e., non-littoral) habitats by juvenile Chinook salmon;
• Determine how long cohorts of hatchery-produced juvenile Chinook salmon reside in Sinclair Inlet; and
• Examine aspects of the trophic ecology of juvenile Chinook in Sinclair Inlet by evaluating their diets and the diets of potential predators and competitors.

For purposes of this study, juvenile Chinook salmon were considered to be of hatchery origin if they had an identifying mark (clipped adipose fin, unclipped fish with coded wire tags (CWTs) and other identifying external marks). Juvenile Chinook salmon were classified as originating from natural production (i.e., �"wild”) if they had no identifying marks. Because not all hatchery juvenile Chinook salmon were distinctly marked in 2001 and 2002, the number of hatchery-produced fish obtained in our samples was underestimated.

In both 2001 and 2002, juvenile salmonids and other fishes occurring in littoral habitats of Sinclair Inlet were collected using beach seines. In both years, twenty-one sites were sampled throughout the three study areas to track spatial and temporal patterns of fish distribution in regular beach seine surveys (RBS) from February through September. Most of our analyses were based on a limited number of regularly sampled sites; eight in 2001 and thirteen in 2002. Both day and night sampling was conducted under various tidal conditions. In 2002, we conducted additional beach seining to recapture florescent pigment marked juvenile Chinook salmon in order to estimate their residence time in Sinclair Inlet. This sampling was referred to as Mark Recapture (MR) sampling. Catches of fish by beach seine were reported as numbers of fish in a beach seine haul, and represented as catch per unit of effort (CPUE).

A tow net (or two boat surface trawl) was used to sample the upper 3m of the water column of study sites within Sinclair Inlet in 2002 only. Tow net samples were collected monthly from May to August 2002 during day and night hours along both shorelines and offshore. Inshore tows generally followed the 5 m contour line while offshore tows were made in the deepest sections, which were generally 10 m or deeper. Tow net data is reported as catch per 10-minute tow (CPUE).

Juvenile chum salmon (Oncorhynchus keta) and several species of forage fish (Pacific herring- Clupea harengus pallasi, Pacific sandlance- Ammodytes hexapterus, and surf smelt- Hypomesus pretiosus) were present from March through the completion of sampling in September in both years. Few juvenile coho salmon (Oncorhynchus kisutch) were caught in either year. We did not capture pink (Oncorhynchus gorbuscha) or sockeye salmon (Oncorhynchus nerka). We observed no temporal pattern in the occurrence of forage fish. Juvenile chum catches peaked in littoral areas in April and May and in offshore areas in May and June (note that offshore areas were not sampled in April so the two data sets are not directly comparable). There was a steady increase in size of juvenile chum salmon from February through the end of sampling (September) indicating that juvenile chum salmon were rearing in Sinclair Inlet.

No bull trout (Salvelinus confluentus) were caught in either year. However, bull trout may have been present but not susceptible to our sampling methods.

A major source of both naturally produced and hatchery Chinook salmon in the study area was Gorst Creek, at the terminus of Sinclair Inlet. In addition, juvenile Chinook salmon originated from a large number of sources outside the study area.

In general, about 10% of the juvenile Chinook salmon collected each year and in each habitat type (i.e. offshore and littoral) were unmarked subyearlings and possibly the progeny of naturally spawning fish (�"wild”). There was little difference in patterns of distribution, abundance, and size of hatchery origin and �"wild” fish, suggesting: 1) hatchery and wild fish behave similarly, or 2) most of what we were calling �"wild” fish were actually unmarked hatchery fish. The presence of juvenile Chinook salmon < 50 mm FL (these are too small to be hatchery fish) in 2001 suggests that natural production is occasionally successful in the area. In 2002, juvenile Chinook salmon were not caught in the beach seine or in the tow net until after the hatchery releases into Gorst Creek began in late May. Juvenile Chinook salmon were caught in littoral areas of Sinclair Inlet from April to September, 2001 and May to September 2002. Chinook salmon were present on the last date of sampling in September each year. Peak catches in littoral and surface waters occurred in early summer (June-July) for both years. The size of juvenile Chinook salmon increased from June until September in both habitat types.

Abundance and size of fish can be a function of conditions under which samples are collected (e.g., tide, amount of floating vegetation, and time of day) as well as characteristics of the habitat at the collection site. Further, some of these variables may interact. In littoral habitats, there was not a clear effect of tide or time of day on CPUE of either wild or hatchery origin juvenile Chinook salmon. The size of juvenile Chinook salmon varied with time of collection (daytime vs. nighttime) of both hatchery and �"wild” origin in littoral (inshore) and neritic (offshore) habitats. In general, we captured larger fish during nocturnal sampling efforts, which may reflect avoidance of sampling gear during daylight hours. Alternatively, larger juvenile Chinook salmon may occupy deeper habitats during the day and move into the range of our sampling gears at night. However, differences in size and abundance of juvenile Chinook with respect to tidal stage or daytime/nighttime were not consistent and therefore, we did not separate these data in our analyses.

Habitat characteristics along Sinclair Inlet shorelines were assessed qualitatively for a variety of habitat factors including Area of capture (as defined in the study by three east-west regions in the inlet), north or south shoreline, type of substrate, amount of upland and submersed vegetation, shoreline modification, and slope. We evaluated the effect of the habitat characteristics on both CPUE and size of juvenile Chinook salmon. Of these variables, only Area of capture had a clear effect on abundance of juvenile Chinook salmon in littoral habitats. Highest abundances were found in the area closest to the terminus of Sinclair Inlet and generally declined with increasing distance from Gorst Creek. However, fish abundance was not significantly different in surface waters of the three areas of capture during tow net sampling. Area of capture also appeared to correlate with size of fish in each habitat type with larger fish generally occurring in the eastern end of the inlet (Area 3) compared to the other two areas. We could not detect an effect of any other habitat factor on size or abundance. There are several plausible explanations for these results. First, habitat factors may not influence fish abundance substantially at the site scale. At larger spatial scales (e.g., area of capture), our observations suggest that habitat is an important determinant is segregating chinook salmon life history stages. Second, our approach to measuring habitat may have been insufficient since we simply assessed habitat qualitatively. Third, other factors that we did not measure may have had an effect on where fish were found.

We marked and released juvenile Chinook salmon with three colors of fluorescent pigment and CWTs into Gorst Creek to estimate residence time in Sinclair Inlet. We developed an approach to mark juvenile Chinook salmon using various colors of fluorescent pigment that allowed us to follow cohorts of Chinook salmon during their migration through Sinclair Inlet. Residence time estimates were made for six groups of fish released into Gorst Creek using beach seine sampling in littoral zone habitats. Mean residence time (average of separate estimates of the six marked groups) in Area 1 (west end of inlet) and Area 2 (middle section of inlet) was 6.2 days and 8.3 days, respectively. The estimated maximum residence time for any group released into Sinclair Inlet was 59 days.

Juvenile Chinook salmon with CWTs recovered from Sinclair Inlet in 2002 originated from 14 different watersheds and from as far away as the Fraser River in Canada. A total of 77% of the total recoveries originated from the Gorst Creek hatchery. Fish released into the Green River were recovered in Sinclair Inlet within 11 days of release, while fish released at Grovers Creek (approximately 25 km swimming distance) were recovered within 48 hours of their release. We found that Gorst Creek hatchery fish comprised nearly 100% of the CWT recoveries until midsummer, dropping to only 40% of the total recoveries after mid-summer through early fall. This is consistent with residence time estimates, suggesting that fish from Gorst Creek migrate rapidly through Sinclair Inlet and are subsequently replaced by non-natal fish. We believe it is reasonable to assume that because hatchery fish from outside of Sinclair Inlet migrate into the area, wild (i.e. naturally spawning) fish also exhibit similar behavior.

The diet of juvenile Chinook salmon in both years was similar with individual fish tending to eat a small number of relatively large prey. There was not evidence of consistent differences in diet of hatchery origin and wild juvenile Chinook salmon. In general, juvenile Chinook salmon appear to be primarily surface and mid water feeders while juvenile chum salmon were foraging primarily in mid waters to the bottom. The dominant prey of juvenile Chinook salmon in both years and in both habitat types (littoral versus neritic) consisted of a diverse mixture of aquatic and terrestrial insects, decapod crustaceans, amphipods, polychaetes, and barnacle larvae. Most decapods were either pinnotherid or porecellanid crab zoea. Crab zoea and other planktonic prey were generally more prevalent in diets in June and early July whereas polychaetes were more important in diets in late July and August. At least fifty insect families were identified in the stomach contents of juvenile salmon. We also noted that the type of insects consumed varied with time. While the origin of any prey and where it was consumed cannot be precisely determined in many cases, it is noteworthy that many of the insects that were eaten came from terrestrial sources and it is likely they were eaten off the surface of the water.

We inventoried intertidal habitat along the southern shoreline of the western portion of the inlet and both the northern and southern shoreline in the eastern portion of the inlet. The remaining shoreline is highly impacted by rock riprap and has limited intertidal habitat. Estimates based on aerial photographs indicate shoreline armoring along this unsurveyed shoreline at nearly 100%. Of the 10 km shoreline surveyed in the field, shoreline armoring was present along 78% of the shorline. We estimate that 91% of the entire 26 km of shoreline had armoring or was modified from natural conditions.

The findings of this study indicate that Sinclair Inlet is used by three major groups of juvenile Chinook salmon.

• The first group consists of hatchery origin fish released into Gorst Creek, typically in late May through the end of June. The fish disperse throughout the Inlet (appearing to use both inshore and offshore habitats), with most of the fish rapidly leaving the Inlet.
• Second, hatchery fish from sources outside the Inlet migrate into Sinclair Inlet. This group is present from July to September. Some of these fish may reside for an extended period of time in Sinclair Inlet, although we were unable to determine this from our data.
• Third, wild juvenile Chinook salmon use the Inlet. These fish could be naturally spawning fish from Gorst Creek or nearby local systems, or move into the Inlet from other river systems similar to hatchery fish. The only way to identify wild fish was by a lack of marks or tags identifying them as hatchery fish. It is possible that unmarked fish are of hatchery origin. We did not detect different patterns of distribution, growth patterns, or diet composition between hatchery and â€�"wild” Chinook. However, this may be due to the unmarked hatchery component of the â€�"wild” group or the low numbers captured of â€�"wild” fish overall. Alternatively, the two groups may behave similarly during their early life history in Sinclair Inlet.

The focus of these studies was on juvenile Chinook salmon, especially subyearling fish, because they are classified as Threatened under the Endangered Species Act. Juvenile Chinook salmon are present in Sinclair Inlet littoral habitats from early spring through early fall, at a minimum. Clearly, Sinclair Inlet shorelines are host to juvenile Chinook salmon from throughout the Puget Sound during late spring and summer months, and likely include both hatchery origin and natural origin. Therefore, proper management of nearshore habitats is important not only for local origin fish, but also for those that originate from a considerable distance.