Grays Harbor Juvenile Fish Use Assessment: 2011 Annual Report

Prepared for the Chehalis Basin Habitat Work Group

January, 2011

Prepared by: Todd Sandell, James Fletcher, Thomas Buehrens,

Andrew McAninch and Micah Wait

Sand Island, Grays Harbor Estuary

Table of Contents

Executive Summary ...... 1

Section 1: Introduction ...... 6 1.1 Purpose and Objectives ...... 6 1.2 Study Area ...... 7 Specific Hypotheses: ...... 9

Section 2: Methods ...... 11 2.1 Habitat Inventory/ Sample Site Selection...... 11 Table 2: Grays Harbor intertidal habitat types by estuary zone – Open water and mud flats removed (in hectares): ...... 14 2.2 Field Sampling Methodology ...... 14 Beach Seining ...... 14 Fyke Netting ...... 16 Data Recording/Water Quality Measures ...... 18 2.3 Catch calculations/ Fish densities ...... 18 Catch Data ...... 18 Density Calculations ...... 19 Section 3: Salmonid Catch Results ...... 19

(Seine Catches only) ...... 19 3.1 Salmon Growth/Age Class designations ...... 19 Juvenile Salmon Size Trends ...... 20 3.2 Salmonid Catch Totals – Beach Seining ...... 26 Salmon Catches by Zone and Site: Mapping ...... 28 3.3 Salmon Distribution and Timing (Seine Catches Only) ...... 33 Estuarine Zone Distribution by Species ...... 33 Estuarine Habitat Use by Salmon Species/Month ...... 44 Salmonid Estuarine Habitat Selectivity ...... 56 Coded Wire Tag (CWT) Recoveries/Analysis ...... 59 Salmon Origin Information from Coded Wire Tags...... 59 Section 4: Non-Salmonid Species (Seine Catches only) ...... 61 4.1 Non-Salmon Species Catch ...... 61 4.2 Non-Salmon Species Distribution and Abundance (Seine catches only) ...... 63 Baitfish ...... 63 Other Common Non-salmonids ...... 64 Section 5: Fyke Netting (only) Catch ...... 72

Trap efficiency ...... 72 Catch Trends ...... 73 Catch trends by fyke site ...... 74 Section 6: Fish Community Modeling ...... 78 6.1 Site Modeling – Physical Variables ...... 78 6.2 Salmon Modeling: Occurrence and Abundance ...... 82 6.3 Long Term Data Collection Goals: Escapement ...... 89

Section 7: Synthesis and Recovery Priorities ...... 92 7.1 Salmonid Spatial and Temporal Habitat Usage ...... 92 Chinook Salmon ...... 93 Coho Salmon ...... 94 Chum Salmon ...... 95 Steelhead trout ...... 96 Cutthroat trout ...... 96 Habitat Selectivity ...... 97 Coded Wire Tags (CWT) ...... 98 Predation ...... 98 Modeling summary ...... 99 7.2 Synthesis: Non-Salmonid/Baitfish Distributions and Abundance ...... 100 7.3: Grays Harbor Salmonid Habitat Recovery Priorities ...... 103 7.4: Response to Specific Review Panel Questions ...... 106

Additional Goals for 2012 ...... 108

Acknowledgements ...... 109

References...... 110

Appendices ...... 115

Executive Summary Grays Harbor (the Chehalis River estuary) is the second largest estuary in the state of Washington, covering 23,504 hectares at mean high high-water from the mouth at Westport to Montesano, and encompassing the tidally-influenced lower reaches of the Chehalis, Humptulips, Hoquiam, Wishkah, Johns and Elk Rivers as well as several smaller tributaries. The total drainage area, including all of the above tributaries, is 660,450 hectares, with 79% of the fresh water input from the Chehalis River.

Between 1900 and 1980, the Grays Harbor estuary had an overall net decrease in tidal wetlands due to extensive diking and filling activities, particularly in the upper portions of the estuary (Boule et al. 1983). A more recent analysis of historical estuarine habitat change detailed a 22% decline in tidal flats (3,493 hectares) due to upland conversion at the mouth of Grays Harbor and along the north channel, an increase in potential eelgrass habitat (1,793 hectares), and an increase in areas below extreme low- water (409 hectares), mainly due to a deepening of the channel for navigation near the mouth of the estuary (Borde et al. 2010). Within the Chehalis basin there are numerous distinct stocks of salmonids that are important to the overall biological diversity in Washington State. These include one stock of spring Chinook salmon (Oncorhynchus tshawytscha), one stock of summer Chinook, seven stocks of fall Chinook, seven stocks of coho salmon (O. kisutch,) two stocks of fall chum salmon (O. keta), two stocks of summer steelhead trout (O. mykiss), and eight stocks of winter steelhead (WDFW and WWTIT 1994). In addition, cutthroat trout (O. clarki) have previously been found throughout the drainage, and bull trout (Salvelinus confluentus) have been documented as present, but specific distribution data do not exist (Smith & Wenger 2001). All of these stocks have been in decline, as have most salmonid stocks in the Pacific Northwest.

Prior to the commencement of sampling, the intertidal areas of the Grays Harbor estuary were divided into six zones: the Estuary Mouth, Central Estuary, North Bay, South Bay, Inner Estuary, and [Chehalis River] Surge Plain. The habitats were classified using several geographic information systems (GIS) layers, with the National Wetlands

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 1

Inventory GIS database used as a basis for our classification (USFWS 2010). This project focused on juvenile salmonids, although abundance data on all other fish species caught were also recorded. The majority of the sampling was conducted using fine meshed beach seines which were deployed using a motorized skiff or set by hand; we used three different beach seines to sample the various habitats in Grays, each suited to particular habitats. Sampling began in March and continued into early October, 2011, and included 674 seine net sets and an additional 20 fyke net sets in tidal sloughs. The objective of the project was to develop a scientific basis for the evaluation of potential sites for future habitat restoration and protection projects. The specific goals were to document the distribution, abundance, habitat use, and timing of juvenile salmonids and other fishes in the Grays Harbor estuary, from riverine tidal waters through marine habitats.

Chum salmon were the most abundant juvenile salmonid captured in 2011 (N=6,528), followed by unmarked Chinook salmon (N=4,514), unmarked coho salmon (N=1,815), hatchery coho salmon (N=137), hatchery Chinook salmon (N=106), and one sockeye salmon. For Chinook and coho salmon, subyearlings dominated the catch. In addition, 99 cutthroat trout, 24 steelhead trout, 3 bull trout and one rainbow trout (parr) were captured. Wild subyearling Chinook salmon were captured throughout the sampling period from March to October, with the main “pulse” of hatchery fish occurring from July through September. A small number of yearling Chinook salmon were captured in April and May. Subyearling coho salmon were also captured from March to September, with a pulse of emigrating yearling coho salmon caught from April to June. Most juvenile chum salmon (all wild subyearlings) were caught from March through May, plus a few individuals in June. Other summary points on zone and habitat usage, as well predation, coded wire tag recoveries, and modeling results are highlighted below.

 Chinook salmon: wild subyearling Chinook salmon were widely distributed in all habitat types, but were most abundant in scrub/shrub habitats and lowest in beach and eelgrass habitats. Interestingly, only wild subyearling Chinook salmon were common at sand flat sites. Hatchery subyearling Chinook salmon were captured in highest densities at mud flat habitats and were less abundant

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 2

elsewhere, a result that may be attributed to Humptulips River inputs into North Bay, which has the highest concentration of mud flats. Few yearling Chinook salmon were caught in 2011.  Coho salmon: habitat usage by wild subyearling coho salmon was widespread, but not to the same degree as juvenile Chinook salmon; densities were highest in mudflats, scrub/shrub and forest habitats, and few were captured elsewhere. Some subyearling coho salmon were captured in tidal portions of the tributaries even in fall, suggesting that there is a diversity of coho life histories present in Grays Harbor. Few yearling coho salmon of either hatchery or wild origin were captured; these were in greatest abundance near the estuary mouth (North and South Bay zones).  Chum salmon: chum were found at highest densities at cobble beach habitats (Central Estuary) but were also common at forested, emergent marsh, eelgrass and aquatic vegetation sites, in that order. There was a striking difference in chum salmon densities between North (extremely low) and South Bay (abundant), suggesting that there may be a mechanism of exclusion from North Bay. The results of our modeling studies suggest that low salinities in North Bay early in the sampling year, likely as a result of elevated Humptulips River fresh water flows, may have discouraged chum salmon entry into this zone.  Trout: our catches of cutthroat trout demonstrate that this species is present in Grays Harbor through much of the year and in a variety of habitats, in contrast with previous studies. Catches of steelhead occurred in low numbers in all habitats, though too few steelhead or bull trout were captured to allow robust comparisons.  Zone Selectivity: wild subyearling Chinook salmon were most abundant in the North Bay, Surge Plain, and Upper Estuary zones, while hatchery subyearling densities were highest in the Central Estuary and North Bay, consistent with hatchery inputs from the Wishkah and Humptulips Rivers. Subyearling coho salmon densities were greatest in the Surge Plain, Inner Estuary and North Bay. Chum were present in the Surge Plain and Inner Estuary when sampling

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 3

commenced in March, but became increasingly more abundant towards the estuary mouth (with the exception of North Bay, where chum salmon were absent, as noted above). The South Bay zone, though receiving few juveniles from natural production in the Johns and Elk Rivers, was an important rearing area for juvenile Chinook and chum salmon, particularly from March-June. Though present in all zones, cutthroat trout were most common in the Surge Plain (particularly early in the year) and North Bay (often at the Humptulips River sites); densities were highest at scrub/shrub cover and emergent marsh sites. Too few steelhead or bull trout were captured to permit comparison by zone.  Habitat Selectivity: cobble beach and eelgrass habitats were used to a much greater degree than their proportional availability (especially by chum and wild subyearling Chinook salmon), which suggests they are especially important to salmon smolts in the Grays Harbor estuary. Aquatic vegetation beds scored low in all comparisons, a result that was surprising since these habitats are considered productive areas. Several theories might explain this: competition with non- salmon species could be higher since aquatic vegetation beds are also utilized by non-salmonids, aquatic vegetation beds may not be as productive in Grays Harbor in comparison with other systems because of turbidity, or aquatic vegetation beds may over-represented in the Shore Zone Inventory data.  Coded Wire Tag Recoveries (CWT): only 12 yearling coho salmon were recovered with CWT (7 hatchery, 5 from wild tagging studies). The recovery of juvenile coho salmon that originated from the Queets River indicates that the Grays Harbor estuary is utilized by out-of-basin hatchery fish for feeding, physiological maturation, or predator avoidance.  Predation: very few large non-salmonid fishes that may be considered salmon predators (particularly northern pikeminnow and staghorn sculpin) were captured in 2011, although our sampling methodology did not effectively sample deep, open water habitats where these may be found. There was evidence that other salmonids – namely steelhead and cutthroat trout – may be foraging upon

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 4

subyearling Chinook and coho salmon, based on similar spatial and temporal patterns of habitat use.  Presence and Abundance Modeling: regression models of subyearling coho suggested that their use of the estuary was primarily confined to the freshwater and slightly brackish areas, particularly sites with scrub/shrub or forest cover, which may provide cover from predators. Subyearling Chinook were the most widely distributed species across space and time. Salinity was negatively correlated with subyearling wild coho abundance and subyearling coho were essentially absent at salinities above 20 ppt. For chum and yearling wild and hatchery coho salmon, timing, as expected, explained much of the variability in abundance since these two species were only present in the spring and moved more rapidly through the estuary than Chinook salmon. Additionally, the models revealed that the North Bay and Upper Estuary zones had higher abundances of wild Chinook salmon than other zones, suggesting the potential importance of these areas to wild Chinook production.  Restoration Priorities: diking in the Johns River and two sloughs in the South Bay zone currently blocked to salmon immigration and rearing by “flap” style tide gates offer the best immediate opportunities for juvenile salmon restoration projects.  An additional goal of this project is to sample in years with “normal”, “low” and “high” escapement for the three main salmonid species (Chinook, coho and chum salmon) in order to analyze how the spatial and temporal distribution of juvenile salmon changes as the density of these species in the estuary fluxes. For wild subyearling Chinook salmon, 2011 fell within the range of a “normal” productivity year, while for both wild coho and chum salmon, 2011 was categorized as a “high” productivity year. Thus, a minimum of two additional sampling years will be required in order to sample these species in years where the escapement

values fall into the other ranges.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 5

Section 1: Introduction

1.1 Purpose and Objectives Pacific salmon that spawn in the rivers and streams of Water Resource Inventory Areas (WRIA) 22 and 23 all must pass through the nearshore habitats in the Grays Harbor Estuary as they migrate to the ocean. Estuarine environments are extremely productive habitats, and many life histories of juvenile salmon spend extended periods of time rearing in this environment. Understanding the extent, timing, and species composition of fish usage in these habitats is a critical component in the development of a salmon restoration strategy for the entire basin. The Grays Harbor Juvenile Fish Use Assessment project is designed to investigate current estuarine habitat use, particularly by juvenile salmonids, and to provide a scientific basis for the selection and prioritization of future salmonid habitat restoration and protection projects within the Grays Harbor estuary.

The project had four primary goals at the outset:

 Determine the abundance, distribution, emigration timing and habitat preferences of juvenile salmonids in the Grays Harbor estuary and tidally- influenced portions of its major tributaries  Gather information on the distribution, abundance and community structure of non-salmonid fishes in these same areas  Use the capture of coded wire tagged salmonids to identify the basin of origin, infer estuarine residence times, and estimate emigration speeds and growth following release from the hatcheries  Examine the underlying physical characteristics of the habitat types found in Grays Harbor to understand, and potentially model, the distribution of juvenile

fishes in the estuary

This project was proposed, developed, and conducted by the Wild Fish Conservancy under a grant from the Salmon Recovery Funding Board and the U.S. Fish and Wildlife Service. Sampling in the first year of this study (2011) began in mid-March and

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 6

continued through the beginning of October, and included 694 net sets (some additional sets were made to test net capture efficiency, but these catches are not included in this report). Of these, 608 were seine sets at the core sites, 58 were seine sets completed at secondary sites, and 20 were fyke net sets in tidal sloughs.

1.2 Study Area Grays Harbor (the Chehalis River estuary) is the second largest estuary in the state of Washington after the Columbia River estuary (Puget Sound is technically considered a “fjord”). The Grays Harbor estuary is a bar-built estuary that was formed by the combined processes of sedimentation and erosion caused by both the Chehalis River and the Pacific Ocean (Chehalis Basin Habitat Work Group, 2010). The estuary covers 23,504 hectares at mean high high-water (MHHW) from the mouth at Westport to Montesano, and encompasses the tidally influences lower reaches of the Chehalis, Humptulips, Hoquiam, Wishkah, Johns and Elk Rivers as well as several smaller tributaries. The total drainage area, including all of the above tributaries, is 660,450 hectares, with 79% of the fresh water input from the Chehalis River (Simenstad & Eggers 1981). The system flows are rainfall driven, with peak flows from December- January in an average year, and minimal input from snowmelt in the southern Olympic Mountains (surface drainage occurs primarily through the Satsop River basin). In the inner estuary (“Inner Harbor”), the main river channel splits into north and south channels; the north channel has been dredged for navigation. Vertical salinity stratification, with a salt water wedge typical of estuarine systems, occurs only in the south channel (Simenstad and

Eggers, 1981).

Between 1900 and 1980, the Grays Harbor estuary had an overall net decrease in tidal wetlands due to extensive diking and filling activities, particularly in the Inner portions of the estuary (Boule et al. 1983). A more recent analysis of historical estuarine habitat change detailed a 22% decline in tidal flats (3,493 hectares) due to upland conversion at the mouth of Grays Harbor and along the north channel, an increase in potential eelgrass habitat (1,793 hectares), and an increase in areas below extreme low- water (ELW) (409 hectares), mainly due to a deepening of the channel near the mouth of

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 7

the estuary (Borde et al. 2003). Upland logging may have led to an increase in sediment transport to the estuary, resulting in loss of tidal flats due to increased elevation. Dredging for navigation has dramatically deepened the area near the mouth of the estuary, as well as along the north channel, resulting in changes in circulation; wakes from large vessels may have increased channel border erosion and loss of tidal flats in the southern area of the estuary. Historically, the estuary received sediment input from the Columbia River, but the construction of jetties (first in 1900, then in the 1930s) may have reduced this input, which may also explain the loss of tidal flats in the lower estuary (Borde et al. 2003). The northern part of the estuary (“North Bay”) consists of extensive mud flats, most of which are submerged at mean high tide, and has experienced little change.

Although much of the basin has been degraded by a combination of logging, channelization, gravel mining, water diversion, road building, diking, dredging, aquaculture, small-scale coal mining, mill effluent, sewage release and pesticide use for aquaculture and cranberry farming (Hiss et al. 1982; Wood & Stark 2002; Smith & Wenger 2001) , the area of the lower mainstem Chehalis River, the surge plain (river km 1-17, just east of Aberdeen to the confluence of the Wynoochee River), contains high- quality rearing habitat for juvenile salmon, particularly coho, and has been well studied (Moser et al. 1991; Simenstad et al. 1992; Team 1997; Henning et al. 2006; Henning et al. 2007). The area contains numerous sloughs and tidal channels, a relatively undeveloped floodplain with seasonal inundation, and a riparian forest dominated by older stands of conifers and hardwoods (Ralph et al. 1994).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 8

Specific Hypotheses:

1) There is differential use of estuarine habitat type by juvenile salmonids in the Grays Harbor estuary 2) Juvenile salmon utilize certain intertidal habitats in greater numbers than would be predicted by the abundance of those habitats alone within the habitat matrix of the Grays Harbor estuary (“habitat selectivity”) 3) Juvenile salmon from upstream tributaries will utilize habitats in South Bay (Johns and Elk River estuaries), even though natural production in these systems is low, because of the presence of large areas with excellent rearing habitat 4) Few smolts emigrating the Humptulips River will utilize “upstream” habitat (East of the Humptulips River delta), instead travelling to areas in South Bay or directly to sea (based on the results of Schroder and Fresh 1992) 5) Eelgrass beds will have higher densities of juvenile salmon, based on the results of previous research (R M Thom 1993; Thorpe 1994).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 9

Figure 1: Location of Grays Harbor (the Chehalis River estuary) in Washington State, U.S.A.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 10

Section 2: Methods

2.1 Habitat Inventory/ Sample Site Selection The intertidal areas of the Grays Harbor estuary were divided into six zones: mouth of the estuary, central estuary, north bay, south bay, inner estuary (referred to in the literature as the “inner harbor”), and Chehalis River surge plain (see Figure 2). The habitat classification was done using several geographic information systems (GIS) layers. The National Wetlands Inventory (NWI) GIS database was used as a basis for our classification (USFWS 2010). The NWI classes were used to group the polygons into subtidal, intertidal, emergent wetlands, shrub/scrub, and forested classes. Then soil data, Shore Zone Inventory data from the Washington Department of Natural Resources (WADNR 2001), and 2009 aerial imagery (USDA 2009) were used to manually subdivide the intertidal polygons into aquatic vegetation bed, mud flat, sand flat, and beach classes. During this step a comprehensive inspection of the classification result was conducted to verify the accuracy and make corrections as needed. Finally, the Shore Zone Inventory data ((WADNR 2001) was used to delineate some preliminary eelgrass habitats (a specific type of aquatic vegetation bed) (Figure 3). Additional locations of eelgrass habitat in the Grays Harbor estuary that were not in any current databases were noted during the 2011 summer sampling, when flows and turbidity decreased, allowing field identification and coordinate fixing via GPS. These areas are included in the habitat maps (Figure 3) and tables used to calculate the percentage of each habitat type (and area) in the Grays Harbor estuary. As a result, some of the habitat totals have changed in this final report.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 11

Figure 2: A map of the Chehalis River estuary (Grays Harbor), showing the six zones and

locations of the core, secondary, and fyke net sites sampled in 2011

The area (in hectares) of each habitat type, by zone, is shown in Table 1, below. These data were used to guide our choices of sampling sites; areas of open water were excluded (although in some cases we sampled the channel margins of some of these areas, which are typically aquatic vegetation beds or sand flats). Mud flats in north bay and the central estuary were sampled at very low or negative tides, when the exposed areas provide a barrier against which the net was drawn, allowing fish to become entrapped. These areas contain large quantities of woody debris, which limited our ability to sample with nets; our efforts to sample mud flats therefore focused on sampling adjacent aquatic vegetation beds (see Table 2 for a breakdown of habitat acreage with open water and mud flats removed).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 12

Figure 3: A map of the Chehalis River estuary showing the six zones and the distribution

of the eight habitat classification types

Table 1: Summary of Grays Harbor intertidal habitat types by zone (in hectares):

% of Habitat Type Mouth Central Inner North BaySouth Bay Surge PlainGrand Total Grand Open Water/Channel 8,614.0 26,350.7 17,234.8 17,260.2 9,132.3 8,625.9 87,217.9 64.5 Aquatic Vegetation Bed 21.5 9,687.6 604.9 8,886.6 1,900.5 0 21,101.0 15.6 Eelgrass 0 14.3 36.7 0.8 56.2 0 108.0 0.1 Mud Flat 0 8,105.1 5,618.5 3,696.5 306.0 0 17,726.1 13.1 Sand Flat 0 1,549.7 797.6 38.6 552.5 0 2,938.5 2.2 Cobble/Gravel/Sand Beach 82.7 0 0 0 0 0 82.7 0.1 High Emergent Marsh 90.3 231.2 365.1 353.5 1,155.8 336.4 2,532.4 1.9 Scrub/Shrub Cover 15.4 58.0 439.2 146.2 132.1 317.8 1,108.7 0.8 Forested 0 3.3 560.9 157.8 137.9 1,598.5 2,458.5 1.8 Total 8,823.8 46,000.0 25,657.8 30,540.2 13,373.4 10,878.7 135,273.7 100

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 13

Table 2: Grays Harbor intertidal habitat types by estuary zone – Open water and mud flats removed (in hectares): % of Grand Habitat Type Mouth Central Inner North BaySouth Bay Surge PlainGrand Total Total Open Water/Channel Aquatic Bed 21.5 9,687.6 604.9 8,886.6 1,900.5 0 21,101.0 69.6 Eelgrass 0 14.3 36.7 0.8 56.2 0 108.0 0.4 Mud Flat Sand Flat 0 1,549.7 797.6 38.6 552.5 0 2,938.5 9.7 Cobble/Gravel/Sand Beach 82.7 0 0 0 0 0 82.7 0.3 High Emergent Marsh 90.3 231.2 365.1 353.5 1,155.8 336.4 2,532.4 8.3 Scrub/Shrub Cover 15.4 58.0 439.2 146.2 132.1 317.8 1,108.7 3.7 Forested 0 3.3 560.9 157.8 137.9 1,598.5 2,458.5 8.1 Total 209.83 11544.1 2804.44 9583.53 3935.127 2252.7275 30329.79 100

To ensure coverage of the 6 zones and various habitat types, we utilized a two- tier sampling system consisting of primary sites (N=28; sampled bi-weekly), secondary sites (N= 9; sampled whenever possible, with a goal of at least once per month), and 3 fyke sites (for specific site locations and GPS coordinates, see Appendix 1). Some sites, particularly near the estuary mouth (Half Moon Bay), were sampled less frequently due to the restrictions of tides, swell, and weather. In addition, a number of sites that we originally intended to sample were dropped due to the presence of large numbers of snags, minimal water flows after the spring runoff season ended, other logistical issues, or denial of access by landowners. The locations of these sites are listed in Appendix 2.

2.2 Field Sampling Methodology Beach Seining Sampling was conducted using fine meshed beach seines which were deployed using a motorized skiff, or set by hand. One end of the net was fixed to shore; the net was then pulled offshore perpendicular to the beach and towed (or walked) back to shore in a 90° arc. We used three different beach seines to sample the various habitats in Grays Harbor and the tidally influenced portions of its major tributaries, each suited to particular habitats. The large seine has 1/8” mesh, is 120’ long and 12’ deep in the middle of the net and had the heaviest lead lines, allowing use at the deepest sites and in strong currents

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 14

(intertidal/subtidal fringe). The wing of the net that was tied to the beach tapered to 6’, while the other wing had no taper; the large seine sampled an area of 1050.71 m2 (0.1051 hectares) when deployed with this method. Tow lines (bridles) were attached to each end to facilitate deployment. The medium net also has 1/8” mesh and is 100’ long and 6’ deep uniformly, sampling an area of 729. 7 m2 (0.0723 hectares). We doubled the lead lines to allow this net to fish better in shallow areas of the main estuary, where stronger currents were common, but soft sediments made the use of the large seine undesirable (the large seine lead line sank into the mud, making retrieval difficult and damaging the fish by rolling them in large clumps of mud upon recovery). The small set net has 1/8”mesh, and is 80’ long, with no taper, and sampled an area of 466.9816 m2 (0.0467 hectares). This smaller net was rigged with 90’ of net along 80’ of lead and float line, creating a pucker, or pocket, in the net for holding fish. This net was primarily used in tributaries, the surge plain, or in North Bay (which is extremely shallow).

At each large seine sample site two consecutive seine hauls were conducted, with the net anchored to the same spot for each of the sample hauls but the net deployed in opposite directions, to provide a measure of catch variability. Shallow water habitats were sampled using the small seine protocol. At each site three consecutive hauls were conducted whenever possible, moving along the shore so that the same habitat was not sampled twice. Once the nets were closed they were brought into shore for catch processing. All fish captured were enumerated, identified to species, and visually scanned for marks and tags. Juvenile salmonids were also scanned for coded wire tags (CWT) and passive integrated transponder (PIT) tags; all CWT tagged salmon were kept (N=12) in order to determine basin of origin and release date. The first 20 individuals of each species/age class/mark status captured at a site were measured for fork length

(mm).

Core sites included: Beardslee Slough, Chehalis river (near bridge restoration site), Chenois Creek Flats, Cow Point, Damon Point, East Fork Hoquiam River, Elk River Flats, Goose Island Flats (NW corner), Humptulips Right River Channel, Humptulips River Mouth, John’s River Boat Launch (adjacent to the launch), John’s River Bar, John’s River

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 15

Channel, Lower Elliot Slough, Mallard Slough, Moon Island, Preacher Slough Restoration (bridge), Sand Island Channel (north side), Sand Island East, Sand Island Flats (South side), Sculpin Cove, Stearn's Bluff, West Fork Hoquiam River mouth, West Fork Hoquiam River, Westport Marina (South Bay Channel), Wishkah River Channel, Wynoochee River, and Wynoochee River Delta. The secondary sites included: Campbell Slough (near Humptulips), Half Moon Bay (Westport), Little Hoquiam River, Lower Wishkah River, North Bay Flats (Humptulips channel), Ocean Shores Flats, Preacher's Slough (corner),

South Bay Bridge, Withcomb Flats (north side).

Several of these secondary sites were sampled in June and July due to the presence of seals and their pups at our core sample sites on Goose and Sand Islands; due to the restrictions of the Marine Mammal Act, we could not resume sampling at the core sites until the seals and their pups had moved on.

Fyke Netting Some shallow water sites were sampled using fyke nets deployed at high tide and fished through the ebb. The fyke nets sites were: Jessie Slough, John’s River Slough, and Wishkah River Slough. To monitor salmon abundance in shallow tidal marshes and blind channels, we established three trapping sites on selected tidal sloughs and blind channels. From April through August these sites were sampled once or twice per month with a large 1/8 inch fyke trap net and cod end that was set across each channel at high slack tide (overall net dimensions varied based upon the dimensions of the channel) to completely block the channel. Fish were collected until low tide when the channel would drain no further (often, water remained in the channel, even at low tide). Fyke traps were placed on the Johns River at river km 3.0, Jessie Slough near the mouth of the Humptulips River, and the Wishkah River at river km 7.7 (Figure 4). All fish captured, including non-salmonids, were enumerated, identified to species, and scanned for marks or coded wire tags. At each site we collected basic water quality parameters

(temperature and salinity) using an electronic meter (YSI model 30).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 16

Figure 4: Aerial photograph of Grays Harbor showing fyke net locations in 2011

We made an effort to standardize data by both habitat area and by trapping efficiency, using a mark recapture method at the fyke sites to inform a modified Lincoln- Peterson model (Chapman 1951). We determined that the most appropriate mark recapture method was to isolate the slough by setting the fyke at high tide, then capture salmon above the fyke trap via seining over the next two hours, which would allow ample time for fish to resume normal behavior before low tide (approximately 6 hours later). The juvenile salmon were captured at several places above the fyke net to better represent the spatial distribution of fish, since where salmon reside may be related to whether they are resident within the slough, or roaming in and out of these habitats with the tide. Using an 80 X 6 ft. beach seine we captured, marked with a caudal fin clip, and released juvenile salmon with the minimum of handling.

The decision to use this approach rather than the conventional method of releasing fish collected and marked prior to setting the fyke, stems from the question of residency versus roaming behavior. Clearly, there is some degree of residency in these habitats among Chinook and especially coho salmon, whereas salmon captured at nearby main stem seine sites are likely to have a different degree of roaming versus resident behavior, presumably mostly roaming. If we use the conventional approach, we are likely to recapture more “roaming” fish than is representative of the slough habitat,

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 17

thus inflating our catch efficiency estimates. Our method also negates the logistical issues caused by tide and daylight that would require us to transport and hold fish for long periods (often overnight), which could increase mortality and further alter behavior.

Data Recording/Water Quality Measures For each sampling set, data on date, time of day, net type, percent of net haul utilized in the set (with few exceptions, 100% of the net was utilized; in some cases, such as when currents collapsed the net prematurely, a smaller percentage was “fished”), current (ebb, flow, low slack, high slack), habitat, duration of the net set (used in reviewing the data to determine if the net was fished for an unusually long time due to snags, resulting in that particular set’s catch being excluded from quantitative analysis) and weather were recorded. At each site temperature and salinity were measured using a water meter (Yellow Springs Instrument Co., Yellow Springs, Ohio); unfortunately our dissolved oxygen meter was inoperable for most of the sampling season despite several attempts at repair. Tidal heights were back-calculated from the ProTides website (http://www.protides.com/map/state-map.php?Washington), using date and time of day to estimate the tide height from the curves (for a complete list of the tide stations used, see Appendix 3).

2.3 Catch calculations/ Fish densities

Catch Data All data were originally recorded on a standardized data form in the field; subsequently data from the field forms were entered into Microsoft Access spreadsheets for analysis. Data were summarized into catch densities organized by sample site and date. To calculate the catch during each set, the total catch of each species was multiplied by the % of the net used in the set (usually 100%). For example, if only 90% of the net was utilized, the calculation was

(total catch for species Y) X (0.9) = adjusted total catch.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 18

These adjusted total catch numbers are reported in sections 3.2 (salmonid seine total catch, 4.2 (non-salmonid total seine catch), and, 5 (fyke net total catch).

Density Calculations To calculate the density of each species for each set, we then divided the above adjusted catch by the area of the net used in that set to get the catch per meter squared. When two or three sets were made at the same site on a given date, the adjusted catches were summed and divided by the total area sampled (the number of sets made at that site with a net of particular area) to generate an average density by site on that date. To calculate the density of a given species (or age class of a species) in hectares (a hectare is 10,000 m2), the formula was:

Density = (adjusted number of target species caught/total net area (m2)) X(10,000)

To compare catch densities between sampling zones or habitat types, the above densities in hectares were summed by the appropriate factor (zone or habitat type, by month) to account for biases created by unequal sampling efforts across space or time and the use of different nets at different sites. This is equivalent to the concept of catch per unit effort (CPUE), with the number of sets made and the area of the net used taken into account, normalizing the catch. These data are presented in sections 3.3 (salmon distribution) and 4.2 (non-salmon distribution).

Section 3: Salmonid Catch Results (seine catches only)

3.1 Salmon Growth/Age Class designations In general, smaller salmonids (age 0+, “ocean type”, or subyearlings) tend to spend more time in estuarine waters and are thus more dependent on estuarine habitats than larger juveniles (age 1+, “stream type”, or yearlings), which typically reside in streams for their first year of life prior to smolting. These classifications apply mainly to Chinook and coho salmon, who have the most diverse patterns of estuarine usage

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 19

(Hering et al. 2010; Bottom et al. 2005; Zaugg et al. 1985; Moser et al. 1991); chum salmon migrate directly to the sea in early Spring and are thus all subyearlings. No pink salmon were captured in 2011 sampling. We did capture one sockeye salmon (on 5/25/2011), at the Damon Point site, near the estuary mouth); it is likely that this fish originated from outside of the Chehalis River basin/estuary, although the specimen was released without a fin clip taken for genetic analysis. Steelhead trout, which typically rear in freshwater for 1-3 years (Quinn 2005), were all considered yearlings or older (subadult or adult), as were cutthroat and bull trout.

To differentiate between subyearling and yearling Chinook and coho salmon, we examined the fork length (mm) distributions of catch by species and month, as well as comparing hatchery (marked) and “wild” (unmarked) salmon. The delineations between the subyearling and yearling fish were quite clear in 2011, with few fish lengths falling in borderline length ranges. Based on our sampling protocol, and time limitations in the field to process large catches, only the first 20 of each species/marking (hatchery or wild, i.e. marked or unmarked) were measured for fork length. To estimate the age class designations of the remaining fish, we calculated the percentage of the catch that were subyearling salmon (based on fork length) and used that result to assign age classes to the remaining fish that were not measured in the field. These numbers are presented in the figures below and tables that follow.

Juvenile Salmon Size Trends Chinook Salmon

We caught wild juvenile Chinook salmon throughout the sample period from March to September, with the main “pulse” of hatchery fish occurring July through September. A small number of yearling Chinook salmon were captured in April and May, three of which were adipose clipped hatchery origin fish. The mean length of both hatchery origin and wild Chinook subyearling salmon steadily increased during the study from approximately 44 mm FL in March to approximately 100 mm FL in September (Figure 5, Table 3). The mean length of hatchery subyearling juvenile Chinook salmon was greater than wild subyearlings in July (Welch modified, 2-tailed, unpaired t-test: t=

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 20

7.01, p < 0.001), but not significant in August and September (t= -1.29, p =0.20; and t= 0.31, p =0.76 respectively). The mean fork length of hatchery subyearling juvenile Chinook salmon in July was 86.4 mm (SD= 6.3, N=170) and the mean fork length of wild juvenile Chinook salmon was 81.9 mm (SD= 10.3, N=582). All Chinook salmon released from hatcheries in the basin were adipose fin clipped (AdClip) in 2010 and 2011.

Figure 5: Chinook salmon fork lengths, by month, showing age class designations

Coho Salmon

Subyearling coho salmon were captured throughout the sample period from March to September, and a “pulse” of outmigrating yearling coho salmon were caught in April, May and June (Figure 6). None of the subyearling coho salmon released from hatcheries in the Chehalis River basin were adipose clipped in 2011; of the yearling fish, roughly 17% were marked in 2011. In 2010, no subyearling coho released from hatcheries were adipose clipped, meaning that our 2011 catches of yearling coho salmon may include approximately 14 yearling coho of hatchery origin that were not marked

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 21

(assuming 100% survival, which is unrealistic). For this reason, we did not attempt to estimate the number of unmarked yearling coho salmon captured in 2011 that may actually have been of hatchery origin.

The mean length of subyearling coho salmon steadily increased during the study from approximately 42 mm FL in March to approximately 80 mm FL in September (Figure 6, Table 4). In contrast, average fork length of yearling coho salmon decreased from April through May in both marked and unmarked fish. This trend shows the inclination for smaller yearling coho salmon to migrate through the estuary later than larger yearling coho salmon. This tendency presumably reflects the need for small individuals to grow a bit more to improve their odds of survival at sea.

Figure 6: Coho salmon fork lengths, by month, showing age class designations

The mean fork length of marked hatchery yearlings was significantly greater than unmarked “wild” yearlings in both April and May (Welch modified, 2-tailed, unpaired t- test: t= 2.48, p= 0.018; and t= 7.41, p= <0.001 respectively). The mean fork length of marked yearling coho salmon in April was 150.1 mm (SD= 13.3, N=32) and the mean

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 22

fork length of unmarked yearlings was 138.0 mm (SD= 19.6, N=26). In May, the mean fork length of marked yearling coho salmon was 139.8 mm (SD= 16.1, N=38) and the mean fork length of unmarked yearlings was 119.0 mm (SD= 14.9, N=199) (Table 4).

Chum Salmon

Most juvenile chum salmon were caught in March, April and May, plus a few individuals in June. The mean length of juvenile chum salmon remained the same in March and April, but increased considerably (18 mm) in May (Figure 7). The mean fork length of juvenile chum salmon in March and April was 42.1 mm (SD= 5.1, N=250) and 43.7 mm (SD= 6.0, N= 913) respectively; but increased to 60.1 mm (SD= 7.9, N=161) in May (Table 6). This seems to indicate that most fish entering the estuary in March continued directly to the ocean, but many chum salmon that entered in April took up residence through May. This pattern of estuarine use may reflect the inherent changes in productivity and the relative growth opportunities of the estuary versus the ocean early in the year. Several studies have pointed out the especially large contribution of harpacticoid copepods in the diets of juvenile salmon (Naiman & Sibert 1979; Mayama & Ishida 2003) and the residence period of chum salmon may be driven by the relative abundance epibenthic prey (Wissmar & Simenstad 1988) during the late Spring.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 23

Figure 7: Chum salmon fork lengths, by month (All chum salmon are subyearlings)

Table 3: Average fork lengths for juvenile Chinook salmon by month. Pair-wise comparisons of wild and hatchery subyearling in July, August and September were made using Welch’s modified unpaired t-tests.

N Mean FL (mm) SD Statistic Significance df

March subyearling wild 156 43.9 3.91 April subyearling wild 456 46.1 7.00 subyearling hatchery 3 59.0 5.6

yearling wild 3 111.7 4.9

yearling hatchery 2 133.5 16.3

May subyearling wild 610 54.2 10.6 yearling wild 8 119.3 13.3

yearling hatchery 1 140

June subyearling wild 496 62.8 11.8 subyearling hatchery 8 85.5 8.6

July subyearling wild 582 81.9 10.3 subyearling hatchery 170 86.4 6.3 t= 7.01 p< 0.001 459.8

Aug. subyearling wild 342 89.2 9.6 subyearling hatchery 96 90.4 7.6 t= -1.29 p= 0.198 187.9

Sept. subyearling wild 147 98.9 13.4 subyearling hatchery 40 99.9 17.7 t= 0.31 p= 0.7571 51.8

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 24

Table 4: Average fork lengths for juvenile coho salmon by month. Pair-wise comparisons of wild and hatchery yearlings in April and May were made using Welch’s modified unpaired t-tests.

N Mean FL (mm) SD Statistic Significance df

March subyearling 36 39.5 2.6 April subyearling 235 41.9 4.9 yearling wild 26 138 19.6 t= 2.48 p= 0.0175 40.6 yearling hatchery 32 150 13.3

May subyearling 189 47.3 6.2

yearling wild 199 119 14.9 t= 7.41 p< 0.001 49.9 yearling hatchery 38 139.8 16.1

June 134 55.2 9.5 subyearling

yearling wild 32 112.1 9.5 yearling hatchery 1 118

July subyearling 47 66.0 8.9 Aug. subyearling 77 71.5 5.8 Sept. subyearling 54 79.4 4.3

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 25

3.2 Salmonid Catch Totals – Beach Seining Data presented in this section refer to actual salmon catch numbers for 2011; note than in the following sections that deal with salmon densities by zone and habitat type, catch densities are presented by catch per hectare, and are thus increased by the multiplication factor required since our nets sample only a fraction of a hectare (ha). See sections 2.3 for more information. In 2011, 13,228 salmonids were captured; the majority of these were chum salmon that were present when sampling began in March, peaked in April, and had exited the estuary by then end of May (note that all chum salmon resulted from natural production) (Table 5; Figure 8c). Chinook salmon were the second most common salmonid captured in 2011 and were present during all months of the study, peaking in May but present in large numbers from April through July (Figure 8a). Coho salmon were most common from April through June, peaking in May (Figure 8b). As shown in the table below, unmarked fish dominated the catch for Chinook and coho salmon, but hatchery coho salmon were most commonly caught in April and May, while hatchery Chinook salmon were most common in July and August (Table 6).

Table 5: 2011 Salmonid catch totals, by species and unmarked/marked designations

Species/origin 2011 Season Total Chinook unmarked 4514 Chinook hatchery 106 Coho unmarked 1815 Coho hatchery 137 Chum 6528 Sockeye 1 Steelhead wild 21 Steelhead hatchery 3 Rainbow trout 1 Cutthroat trout 99 Bull trout 3 Total 13228

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 26

Figure 8a: Chinook salmon catch by month in Grays Harbor, 2011

1400 1200 1000 800

600 Chinook hatchery 400 Chinook wild 200 0

Figure 8b: Coho salmon catch by month in Grays Harbor, 2011

1400 1200 1000 800

600 coho unmarked 400 coho wild 200 0

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 27

Figure 8c: Chum salmon catch by month in Grays Harbor, 2011 (note the different Y-axis scale)

6000

5000

4000

3000 chum 2000

1000

0

Table 6: Monthly catch totals for the three most common salmon species captured in 2011

Month Chinook Coho Chum unmarked hatchery unmarked hatchery March 157 0 36 0 1062 April 897 5 444 81 5387 May 1124 1 688 51 247 June 927 8 365 4 0 July 989 247 170 4 0 August 549 100 106 0 0 September 149 40 61 0 0

Salmon Catches by Zone and Site: Mapping To further visualize our catches of salmonids in Grays Harbor in 2011, the following plots show actual catch data (not densities) for each site, with one map per zone. They are presented in order from upstream (Surge Plain) to the ocean, in Figures 9a-e. As there was only one site in the estuary mouth zone (Half Moon Bay, these data are combined into the plot for the Central Estuary zone.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 28

Figure 9a: Surge Plain

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 29

Figure 9b: Inner Estuary

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 30

Figure 9c: North Bay

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 31

Figure 9d: South Bay

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 32

Figure 9e: Central Estuary

3.3 Salmon Distribution and Timing (Seine Catches Only)

Estuarine Zone Distribution by Species Chinook salmon

Juvenile salmon were caught in all months sampled, from March through November, but the estuary zones showed differences in density and seasonal distribution. Mean juvenile Chinook salmon densities ranged from approximately 20 fish per hectare at the estuary mouth to over 290 fish/ha in North Bay (Figure 10a). There was a progressive shift in wild subyearling densities over the season from peak concentrations in the surge plain early in the season, to the Inner and central estuary in May, June and July, and finally to littoral habitats in the lower estuary later (Figures 9e, 11a). Peak density diminished through this progression as the juvenile Chinook salmon

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 33

dispersed, from 650 fish/ha in the surge plain in May to 114 fish/ha at the estuary mouth in July. The patterns of outmigration in North Bay need to be considered separately since the Humptulips River at the northern end of the bay is the largest contributor to Chinook salmon of the Grays Harbor tributaries, producing an estimated escapement of 5,761 fish in 2010 (data provided by staff at the WDFW Aberdeen office). Density of wild subyearling Chinook salmon was high throughout April, May and June in North Bay (approximately 500 fish/ha), before falling dramatically in July as fish dispersed toward the ocean. The patterns of juvenile Chinook salmon density in South Bay probably reflect the relative contributions from North Bay and the Chehalis River basin because South Bay doesn’t have any rivers that contribute significantly to Chinook production. Hatchery Chinook salmon densities were highest in the central estuary and North Bay (Figures 9c, d and 10a) which is consistent with hatchery inputs from the Wishkah River and Humptulips River.

Coho salmon

Mean juvenile coho salmon densities were several times smaller than Chinook and ranged from approximately 11 fish per hectare in the central estuary (yearling and subyearling combined) to over 60 fish/ha in North Bay (Figures 9c, 10b). Subyearling coho salmon densities were greatest in the surge plain, inner estuary and North Bay, largely due to the contribution of freshwater sites, which yearling coho salmon seek out until they can osmoregulate in a saline environment. Yearling coho smolt densities were greatest in the lower estuary in South Bay, North Bay and the estuary mouth, which may reflect a preference for the more productive littoral habitats and therefore better feeding opportunities. Yearling coho salmon smolts migrated rapidly through the estuary, entering in mid-April, peaking in May and leaving the estuary by the end of June (Figure 11b). The subyearlings appeared to migrate earlier, with peak density (126 fish/ha) occurring in the inner estuary in March (Figures 9b, 10b), and over a longer period.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 34

Chum salmon

Chum salmon fry typically enter an estuary from early February to late May, with the peak typically occurring in late March. Sampling in 2011 didn’t begin until late March; therefore the first half of the juvenile chum migration was likely missed by the 2011 sampling effort. Nonetheless, we likely captured the juvenile chum salmon migration somewhere close to its apex in March, when peak densities as high as 7000 fish per hectare were recorded in the central estuary (Figures 9e, 11c). The chum salmon migrated rapidly out of the estuary and although some of the fish spent a short period of residence in South Bay in April, most of the fry had left by May. Mean densities in North Bay were notably low, 0.35 fish/ha, especially compared to South Bay which had a mean density of over 300 fish/ha (Figures 9d, 10c). Since juvenile chum salmon from upriver must pass by the entrance of both these zones to access the estuary mouth, there must be either a mechanism of exclusion from North Bay or a strong preference for South Bay. The overwhelming difference in chum salmon densities between the two zones suggests the former; and prevailing water currents driven by river, tidal and landscape features may preclude juvenile fish from entering North Bay. It is also possible that strong tidal currents and a deep channel encourage smolts to move into South Bay if they approach the mouth of the estuary during a flood tide.

Steelhead trout

Just 24 steelhead trout were captured during the course of the sampling season; too few to recognize any patterns of distribution or timing. Mean densities by zone were predictably low, ranging from 0.06 fish per hectare in South Bay, to 3 fish/ha in North

Bay (Figure 10d).

Cutthroat trout

Mean cutthroat trout densities by zone were low, ranging from 0.7 to 7 fish per hectare (Figure 8d), but catches were frequent enough to observe some patterns of estuary use. In general, peak abundance within a zone coincided with peak abundances of Chinook salmon and coho fry, but not juvenile chum salmon (Figures 10 a-d); although this may

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 35

be a reflection of our sampling period, as we did not capture the early part of the chum salmon migration. However, the general pattern of cutthroat trout density within each zone appears to match changes in the density of subyearling Chinook and coho salmon that progressively make their way down the estuary. In the main estuary the seasonal pattern of cutthroat densities are most closely associated with subyearling Chinook salmon, but in North Bay, the pattern is less clear and there are strong similarities between cutthroat densities and subyearling coho salmon (Figures 11 a - d).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 36

Figure 10a: Chinook salmon density (in hectares) by zone in Grays Figure 10c: Chum salmon density by zone (in hectares) in Grays Harbor, 2011 Harbor, 2011

Figure 10b: Coho salmon density (in hectares) by zone in Grays Figure 10d: Steelhead and cutthroat trout density by zone in Grays Harbor, 2011 Harbor, 2011 (note the different Y-axis scale versus salmon)

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 37

Figure 11a: Chinook salmon density by zone/month

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 38

Figure 11a continued: Chinook salmon density by zone/month

Figure 11b: Coho salmon density by zone/month

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 39

Figure 11b continued: Coho salmon density by zone/month

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 40

Figure 11c: Chum salmon density by zone/month (note the changing scales on the Y-axes)

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 41

Figure 11c continued: Chum salmon density by zone/month

Figure 11d: Steelhead and cutthroat trout density by zone/month

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 42

Figure 11d continued: Steelhead and cutthroat trout density by I zone/month

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 43

Estuarine Habitat Use by Salmon Species/Month [Note: Not all salmonid species were captured in all habitats, so the number of graphs per habitat type in the following section varies (e.g. in sand flat habitat, only Chinook salmon were captured). Note also that the Y-axis scales vary between species due to large differences in the densities encountered; the gridlines in the graphs differ to draw the reader’s attention to this point]

Overall habitat use by species

Overall, wild subyearling Chinook salmon were widely distributed in all habitat types, highest in scrub/shrub (281.5 per hectare) and lowest in beach (40.9/ha) and eelgrass (30.5/ha). Hatchery subyearling Chinook salmon were captured in highest densities in mud flat habitats (47.4/ha), and were at low densities everywhere, most likely because a majority of the population emerged from the Humptulips River. Not enough yearling wild Chinook salmon were captured to allow generalizations (Figure

12a).

Wild subyearling coho salmon were also widespread, but not to the same degree as Chinook salmon; densities were highest in mudflats (140.7/ha), then shrub (51.2/ha), and forest (39.3/ha) habitats; few were captured elsewhere. Few yearling coho salmon of either hatchery or wild origin were captured, making it difficult to draw conclusions, though wild yearlings had their highest densities at cobble beach, eelgrass and emergent marsh (Figure 12b).

Chum (all wild subyearlings) were found at highest densities off cobble beach sites (409.1/ha) (particularly Damon Point, just North of the estuary mouth), but were also common at forested, emergent marsh, eelgrass and aquatic vegetation sites, in that order. Chum salmon were very rare in sand, mud flat or scrub/shrub cover habitats

(Figure 12c).

Steelhead densities were low in all habitat types (highest at mud flat sites, 12.2/ha), and Cutthroat trout were also rare, but densities were highest at scrub/shrub cover sites (9.3/ha) (Figure 12d).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 44

Specific salmonid habitat usage by habitat type and month

Cobble beach habitats: Chum salmon were by far the most abundant salmonid at cobble beach habitats, with densities peaking in March (7,280.8/ha) but also high in April (1,270.6/ha) before this species exited the estuary (Figure 13a). Wild and hatchery subyearling Chinook salmon were the next highest in abundance at cobble beach habitats, peaking in July (60.3/ha and 19.0/ha, respectively) and August (240.3/ha and 19.0/ha, respectively), presumably as these species emigrated to sea. Both were captured at low densities in other months. Not enough yearling Chinook salmon were captured to permit generalizations. Yearling wild coho salmon were also found at cobble beach sites in May (80.9/ha), though were much less abundant than chum or Chinook salmon (Figure 13a).

Sand Flat habitats: Only wild subyearling Chinook salmon were common at sand flat sites, peaking in July (453.0/ha). Other types of Chinook salmon and other salmonid species were not present at sand flat habitats (Figure 13b).

Mud Flat habitats: Wild subyearling Chinook and coho salmon were both common at mud flat habitat sites, peaking in August (179.9/ha and 184.2/ha, respectively) and September (157.0/ha and 164.2/ha, respectively). Wild subyearling Chinook salmon were also common at mud flats in July (142.8/ha), but hatchery subyearling Chinook salmon were only present in high densities in September (74.9/ha). The only other salmonids captured at mud flats were wild steelhead trout, also peaking in September, though at much lower densities (25.0/ha) (Figure 13c).

Aquatic Vegetation Bed habitats: Chum salmon were found at the highest densities in aquatic vegetation bed habitats, peaking in March (3,440.5/ha) and April (552.3/ha) as they emigrated to sea. Surprisingly, only wild subyearling Chinook salmon were also abundant at aquatic vegetation bed sites (though at an order of magnitude lower than chum salmon), and were common from April through August, peaking in June (246.8/ha). Wild subyearling and yearling coho salmon were also present in these habitats from March through June, though at much lower densities (Figure 13d).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 45

Eelgrass habitats: Chum salmon had the highest catch densities at eelgrass habitats, peaking in April (568.7/ha). The next most abundant species was wild subyearling Chinook salmon, with peak densities occurring in August (130.9/ha), followed by wild yearling coho salmon, peaking in May (76.1/ha) (Figure 13e).

Emergent Marsh Habitats: Chum salmon also had the highest densities at emergent marsh sites, and were present at these sites from March through May, peaking in April (905.0/ha). Wild subyearling Chinook salmon were also found in these areas from April through September, with relatively high densities occurring in May (173.2/ha), June (217.7/ha), and July (111.8/ha). Wild subyearling coho densities at eelgrass sites were highest in April (100.7/ha), and wild yearling coho densities peaked in May (63.3/ha). Cutthroat trout were also found at emergent marsh sites, though at very low densities, from April through August (Figure 13f).

Scrub/Shrub Habitats: Catch densities at scrub/shrub sites were dominated by wild subyearling Chinook salmon, and were elevated from March (488.1/ha) through May (1,070.4/ha), with some catches occurring in all months. Wild subyearling coho were also captured in these areas in all months, peaking in May (200.4/ha), but at lower densities than Chinook salmon. Early in the year, chum were present in scrub/shrub habitats, peaking as sampling began in March (71.2/ha) and slowly declining through April and May (34.7/ha and 2.6/ha, respectively). Both steelhead and cutthroat trout were also captured in these habitats, though at lower densities; densities of both peaked in April (5.5/ha for steelhead; 40.2/ha for cutthroat) (Figure 13g).

Forested habitats: Chum salmon also had the highest catch densities at forested sites, peaking in April (1,114.1/ha) but also elevated in March (428.5/ha). Wild subyearling Chinook salmon were also common in forested habitats from March through June, peaking in April (228.0/ha), as were wild subyearling coho salmon, which peaked in May (142.6/ha). Although cutthroat trout were captured at these sites in April, May, and July-August, densities were low in all months. Steelhead were captured in low densities at forested habitats only in April (Figure 13h).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 46

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 47

Figure 12a. Chinook salmon density, by habitat type Figure 12c. Chum salmon density, by habitat type

Figure 12b. Coho salmon density, by habitat type Figure 12d. Steelhead/Cutthroat trout density, by habitat type

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 48

Figure 13a. Salmonid density by month, cobble/gravel/sand beach habitat

Figure 13b. Salmonid density by month, sand flats

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 49

Figure 13c. Salmonid density by month, mud flat habitat

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 50

Figure 13d. Salmonid density by month, aquatic vegetation bed Figure 11d. d density by month, aquatic vegetation bed habitat habitat

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 51

Figure 13e. Salmonid density by month, eelgrass habitat

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 52

Figure 13f. Salmonid density by month, emergent marsh habitat

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 53

Figure 13g. Salmonid density by month, scrub/shrub habitat

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 54

Figure 13h. Salmonid density by month, forested habitat

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 55

Salmonid Estuarine Habitat Selectivity To compare the habitat usage preferences of the different species and age classes of juvenile salmonids captured in Grays Harbor in 2011, we utilized a non- parametric ANOVA procedure because the densities between habitats were not normally distributed. The Kruskal-Wallis method tested the null hypothesis that salmon densities among habitat types originate from the same distribution. Results for subyearling hatchery Chinook were significant at the alpha level (p<0.05), which indicates that median fish density within this group was different for at least one habitat type (Table 7). Results for subyearling wild coho, and subyearling wild Chinook were close to significant at the alpha level.

Table 7: Kruskal-Wallis non-parametric ANOVA comparing salmonid densities between habitat types

Species/age/origin Statistic (Χ2) df p-value Chinook subyearling wild 13.4126 7 0.0627 Chinook subyearling hatchery 16.4387 6 0.0116 Coho subyearling wild 10.8291 5 0.0549 Coho yearling wild 2.6428 6 0.8522 Chum subyearling wild 7.1706 5 0.2083

To determine which habitat types were different for subyearling hatchery Chinook densities we conducted a series of pairwise comparisons using the Mann–Whitney U test. The results revealed that the median densities of subyearling hatchery Chinook were significantly lower (p<0.05) in forested habitats than in mud flats, sand flats, aquatic vegetation beds and beach habitats. Subyearling hatchery Chinook densities were also significantly lower (p<0.05) in eelgrass and emergent marsh habitats than mud flats.

Another way of visualizing habitat use is by using an index that is weighted by available habitat to help reduce any bias derived from variation in availability. Habitat use indices

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 56

modified from Jacobs (1974) were calculated for the most common species/age/origin of salmon using the following formula:

Where Fa is the proportion of fish using habitat variable a, and Ha is the proportion of habitat variable within the estuary. Average density (fish/ha) for each habitat type was used to calculate Fa. Index values range from +1 to -1: positive values indicate that habitat use is greater than proportional availability; negative values indicate that habitat use is less than proportional availability; and zero indicates that habitat use is equal to its availability. The results are presented in Figures 14 a-e.

Figure 14a: Habitat selectivity of wild subyearling Chinook salmon

Figure 14b: Habitat selectivity of hatchery subyearling Chinook salmon

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 57

Figure 14c: Habitat selectivity of wild subyearling coho salmon

Figure 14d: Habitat selectivity of wild yearling coho salmon

Figure 14e: Habitat selectivity of chum salmon (all wild subyearlings)

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 58

Coded Wire Tag (CWT) Recoveries/Analysis A few coded wire tags (CWT) were recovered from the snouts of juvenile salmon collected in Grays Harbor in 2011. CWTs were extracted from fish in the lab and codes were read under magnification (40X). Codes were matched to the Pacific States Marine Fisheries Commission Regional Mark Information System (RMIS) database, and to the WDFW Wild Salmon Production/Evaluation Unit database. The codes were further verified by contact with individual hatcheries, fish biologists within WDFW, and biologists at the Quinault Tribe Department of Natural Resources (QDNR).

In total, 12 CWT were recovered juvenile salmon during the course of the 2011 field season, all of which were yearling coho salmon (Table 8). No patterns in distribution or timing could be determined due to the low sample size. Seven of the recovered fish were from hatcheries and five were part of the Chehalis Basin wild smolt trapping & tagging operations conducted in spring 2011 by the WDFW Wild Salmon

Production/Evaluation Unit.

Salmon Origin Information from Coded Wire Tags Two coho juveniles with CWT originated from the Quinault tribe’s Salmon River Fish Culture facility on the lower Queets River and migrated approximately 111 km to their point of capture in the south bay of Grays Harbor. We are awaiting release data from the Quinault Tribal DNR to determine travel time for these fish. The other 10 fish were tagged and released within the Chehalis watershed from sites on the Satsop River, Skookumchuck River and mainstem Chehalis River (Table 8). The CWT releases occurred over a period of several days to three weeks; therefore travel times couldn’t be estimated precisely for these fish. Nonetheless, minimum rates of migration were consistent with other studies (Berggren & Filardo 1993; Dawley et al. 1986) and indicated a rapid dispersal of yearling coho following their release.

The recovery of juvenile coho salmon that originated from the Queets River indicates that the Grays Harbor estuary is utilized non-basin hatchery fish for feeding, physiological maturation, or predator avoidance. Many of Washington’s coastal streams

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 59

have relatively small estuaries and Grays Harbor may be an important habitat for juvenile salmon from other systems along the coast.

Table 8: Summary of Coded Wire Tag (CWT) information from 2011 captures

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 60

Section 4: Non-Salmonid Species (Seine Catches only) 4.1 Non-Salmon Species Catch

During 2011, we caught 46 species of non-salmonid from mid-March to early September including six species of baitfish (Table 9). The most abundant species were surf smelt (Hypomesus pretiosus pretiosus) and three-spine stickleback (Gasterosteus aculeatus), accounting for 29% and 27% of the non-salmon catch respectively (Figure 16). The next most abundant species was shiner perch (Cymatogaster aggregate) representing 13%. Together, these three species made up roughly 70% of the non- salmon catch for 2011. Common species typically occurred in more than one third of the collections and had average density values greater than 100 fish per hectare.

Figure 16: Common non-salmon species shown as percentage of non-salmon catch for

Grays Harbor 2011

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 61

Table 9: Total non-salmonid catch by species for 2011 Species: catch total: Species: catch total:

Surf Smelt Plainfin Midshipman Hypomesus pretiosus pretiosus 33692 Porichthys notatus 31 Threespine Stickleback Buffalo Sculpin Gasterosteus aculeatus 31404 Enophrys bison 29 Shiner Perch Black Rockfish Cymatogaster aggregata 15029 Sebastes melanops 25 Northern Anchovy Whitefish Engraulis mordax 8303 Prosopium williamsoni 10 Pacific Staghorn Sculpin Tube-snout Leptocottus armatus 6514 Aulorhyncus flavidus 9 Pacific Herring Speckled Dace Clupea harengus pallasi 4556 Rhinichthys osculus 8 Saddleback Gunnel Pacific Ocean Perch Pholis ornata 4135 Sebastes alutus 7 English Sole Unidentified juvenile sculpin Parophrys vetulus 4042 7 Peamouth Arrow Goby Mylocheilus caurinus 3672 Clevelandia ios 6 Unidentified juvenile minnow Rock Greenling 1975 Hexagrammos lagocephalus 6 Pacific Sand Lance Striped Seaperch Ammodytes hexapterus 746 Embiotoca lateralis 5 Starry Flounder Crescent Gunnel Platichthys stellatus 729 Pholis laeta 4 Bay Pipefish Reticulate Sculpin Syngnathus griseolineatus 685 Cottus perplexus 4 Sand Sole Cabezon Psettichthys melanostictus 418 Scorpaenichthyes marmoratus 4 Unidentified post-larval flatfish Torrent sculpin 414 Cottus rhotheus 3 Prickly Sculpin White Sturgeon Cottus asper 229 Acipenser transmontanus 2 Pacific Snake Prickleback Pacific Tomcod Lumpenus sagitta 199 Microgadus proximus 2 Northern Pikeminnow Kelp Greenling Ptychocheilus oregonensis 114 Hexagrammos decagrammus 2 Largescale Sucker Pacific Sardine Catostomus macrocheilus 109 Sardinops sagax 1 Red-sided Shiner Largemouth Bass Richardsonius balteatus 97 Micropterus salmoides 1 Silver Surfperch Pumpkinseed Hyperprosopum ellipticum 97 Lepomis gibbosus 1 Bay Goby High Cockscomb Lepidogobius lepidus 52 Anoplarchus purpurescens 1 Unidentified larval fish Lingcod 35 Ophiodon elongatus 1 Pile Perch Sharpnose Sculpin Rhacochilus vacca 33 Clinocottus acuticeps 1 American Shad Coastrange Sculpin Alosa sapidissima 31 Cottus aleuticus 1

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 62

4.2 Non-Salmon Species Distribution and Abundance (Seine catches only)

Baitfish Of the six species of baitfish, surf smelt, northern anchovy (Engraulis mordax), and Pacific herring (Clupea harengus pallasi) were commonly captured. Surf smelt were the most ubiquitous of the bait fish, occurring in every month of sampling, every zone, and almost every habitat type (Figure 17a, 19a). Surf smelt and Pacific herring were represented in adult, juvenile and postlarval stages, while northern anchovy were present as adult and postlarval fish, and Pacific sand lance (Ammodytes hexapterus) as juvenile and postlarval fish. Of the six zones, the central estuary, South Bay and to a lesser extent North Bay, produced high densities of all the major baitfish species (Figure 20), while the peak abundance of surf smelt occurred at the estuary mouth (Figure 19a). All the baitfish shared common patterns of association with beach, sand flat and eelgrass habitat types (Figure 21), which partially reflects the availability of these habitats in the lower estuary. Only three species of baitfish (adult and juvenile surf smelt, juvenile Pacific herring and adult anchovy) were consistently abundant over the sampling period to indicate residence and actual utilization of the estuary. Pacific sand lance were caught infrequently although occasionally in great abundance in the central estuary, particularly near the mouth. The two other species of baitfish, American shad (Alosa sapidissima) and Pacific sardine (Sardinops sagax) were rarely encountered and never abundant. Surf smelt were at their peak abundance during March (almost 4000 fish/ha) primarily as post-larval fish (Figure 17a). By April, smelt densities fell considerably, but the juvenile fish continued to be present in the catch in substantial (although diminishing) numbers until August. Juvenile Pacific herring appeared to begin entering Grays Harbor in April and peaked in abundance by August (Figure 17b). There was no obvious relationship between the occurrence of post-larval/juvenile herring and the adults to suggest that these influxes were associated with spawning in the estuary. Patterns of abundance for northern anchovy were unclear: adult fish began entering the estuary in April, but average densities were low until August when they peaked at about 600 fish per hectare

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 63

(Figure 17c). Too few juvenile anchovy were captured to detect any patterns of distribution or abundance.

Other Common Non-salmonids Of the 40 or so other species of non-salmonids, six were considered common: three-spine stickleback, shiner perch, Pacific staghorn sculpin (Leptocottus armatus), saddleback gunnel (Pholis ornate), English sole (Parophrys vetulus), and peamouth (Mylocheilus caurinus). Three-spine stickleback and shiner perch were particularly abundant and they occurred in almost every month, every zone and every habitat type (Figures 18a, 20a, 22a). Both species reproduced in the estuary and both species were present in various life history stages, which we recorded as adult and juvenile stages in shiner perch.

Three-spine stickleback were ubiquitous throughout the sampling period, but recruitment peaked in August and September (Figure 18a) when many tiny juveniles became ever-present in our catch. Although stickleback were present in all the sampling zones, the greatest densities occurred in the central estuary, North Bay and South Bay (Figure 20a); and of the habitat types, eelgrass emerged as a preferred habitat. Adult shiner perch began to appear in our samples in April and densities steadily increased until July (over 100 fish/ha), the peak month of reproduction for this viviparous species (Figure 18a). The new cohort took over as the dominant life stage and the average density of juvenile shiner perch in July jumped to 866 fish per hectare. Shiner perch were present throughout the estuary, particularly in South Bay (Figure 20a).

Staghorn sculpin were another common catch at nearly every site except for those that were fully freshwater. Both staghorn sculpin and saddleback gunnel shared common patterns of abundance, by zone and by habitat type (Figures 20b, 22b), although the distribution of staghorn sculpin extended a little higher in the estuary because they can tolerate lower salinities than saddleback gunnel.

English sole and starry flounder (Platichthys stellatus) were the two most common flatfish collected in 2011. Grays Harbor is an important rearing area for juvenile English

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 64

sole, which utilize the extensive shallow sublittoral and lower littoral habitats (Simenstad & Eggers 1981). English sole were most abundant in South Bay (over 200 fish/ha), and their relative abundance diminished from the lower estuary to the surge plain (Figure 20c). Like several other species, North Bay had a relatively low abundance of English sole compared with South Bay and central estuary. In contrast, starry flounder were most abundant in the surge plain. Densities of juvenile English sole peaked in May (approximately 250 fish/ha) and then diminished toward the tail end of the season. Peak densities of starry flounder occurred a month later, in June (Figure 18c).

Several species of the minnow family (Cyprinidae) were captured in Grays Harbor, but only peamouth was considered common. They were most abundant in the surge plain and inner estuary, and in forested and shrub cover habitats (Figures 20d, 22d). Peamouth appeared in the catch May and increased in abundance throughout the rest of the sampling season (Figure 18d). Presumably, adult fish reside further upstream in winter and move downstream to the inner estuary in spring to reproduce: postlarval and juvenile peamouth became an increasingly large portion of the catch as the season progressed. Catches of pikeminnow were restricted almost exclusively to the surge plain, reflecting the freshwater nature of this species (Figure 20d). Average monthly densities were low, ranging from 0.2 to 13 fish/ha; but it should be noted that our shoreline sampling efforts targeted mainly juvenile pikeminnow, whereas mature fish tend to reside in the deeper channels: only once did we catch an individual of significant size

(508 mm in late June).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 65

Figures 17 a-d: Baitfish densities (in hectares) by month in Grays Harbor, 2011 c: Northern Anchovy a: Surf Smelt

d: Pacific Sand Lance b: Pacific Herring

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 66

Figure 18: Common non-salmonid densities (in hectares) by month in Common non- densities (in hectares) by month in Grays Harbor, Grays Harbor, 2011 2011 a: Three-spine Stickleback and Shiner Perch c: English Sole and Starry Flounder

b: Staghorn Sculpin and Saddleback Gunnel d: Peamouth and Pikeminnow

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 67

Figure 19: Baitfish densities (in hectares) by zone in Grays Harbor, 2011 a: Surf Smelt c: Northern Anchovy

b: Pacific Herring d: Pacific Sand Lance

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 68

Figures 20 a-d: Common non-salmonid densities (in hectares) by zone Common non- densities (in hectares) by month in Grays Harbor, in Grays Harbor, 2011 2011 a: Three-spine Stickleback and Shiner Perch c: English Sole and Starry Flounder

b: Staghorn Sculpin and Saddleback Gunnel d: Peamouth and Pikeminnow

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 69

Figure 21: Baitfish densities (hectares) by habitat in Grays Harbor, 2011 a: Surf Smelt c: Northern Anchovy

b: Pacific Herring d: Pacific Sand Lance

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 70

Figures 22 a-d: Common non-salmonid densities (in hectares) by Common non- densities (in hectares) by month in Grays Harbor, habitat in Grays Harbor, 2011 2011 a: Three-spine Stickleback and Shiner Perch c: English Sole and Starry Flounder

b: Staghorn Sculpin and Saddleback Gunnel d: Peamouth and Pikeminnow

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 71

Section 5: Fyke Netting (only) Catch

Trap efficiency In order to compare catches of salmonids in fyke nets between sites, dates, and with beach seine catches, a methodology was needed to transform fyke net catch per unit effort (unit-less fish per fyke haul) into fish density (fish per unit area). We intended to generate fyke haul abundance estimates for salmonids by species and age class, and divide these by the surface area of the tidal channel being sampled. Because some fish may not emigrate from tidal channels and thus be captured in a fyke net during the outgoing tide, it was necessary to estimate the efficiency of the fyke net in order to generate an unbiased abundance estimate for the tidal channel. Further, the efficiency of fyke nets may vary between sites, species, age classes, and across times and these factors may affect the proportion of fish leaving or remaining in tidal channels during the outgoing tide.

We attempted to generate unbiased abundance estimates for fyke net sets using a mark- recapture methodology and the Chapman (1951) modification of the Lincoln-Petersen closed population abundance estimate. To accomplish this, we first set the fyke across the tidal channel mouth to restrict fish movement and satisfy the closure assumption of our abundance estimator. We then spent the first two hours of the outgoing tide using beach seines to capture and individually mark salmonids upstream of the fyke net by removing a small (2 mm diameter) caudal fin clip. We sampled throughout the habitat available to them to ensure that each fish had a roughly independent and identical capture probability, and that the tagged group was representative of the population in the tidal channel as a whole. At low tide when we processed the contents of the fyke net, we then made note of the proportion of marked and unmarked salmon by size class and species. The unbiased abundance estimate, N^ was calculated according the equation below, with the following inputs: the number of fish marked upstream, M; the total number of marked and unmarked fish captured in the fyke net, C; and the total number of marked fish that were recaptured in the fyke net, R.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 72

Unfortunately, this methodology proved to be highly time-consuming and failed to generate satisfactory results. We had considerable difficulty capturing fish with seines upstream of the fyke net, resulting in small marked fish sample sizes. Further, the proportion recaptured was typically very small (~10 %), suggesting that in many cases a considerable proportion of salmonids were remaining in tidal channels, often in small isolated pools, throughout the tidal cycle leading to very imprecise estimates of abundance. It was therefore impossible to generate unbiased abundance estimates and thus densities of salmon in fyke sites. Our reporting of fyke net catch data is therefore limited to comparisons of raw catch data, with unknown and likely, variable sampling efficiency, and is therefore reported separately from beach seine data.

Catch Trends We observed juvenile salmon throughout the study period at two of the three fyke sites, with the exception of Jessie Slough, where no salmon were caught in early June and late August (Figure 23). At two locations, Johns River slough and Jessie Slough, coho were most abundant followed by Chinook; but at Wishkah River slough, Chinook were the most abundant salmon (Table 10). Juvenile chum densities were low, which reflects the earlier timing and direct pattern of outmigration to the ocean. Cutthroat trout and bull trout were also found to utilize some of the sites (Table 10).

Table 10. Fyke trapping season totals for salmonids at each of the three sites: Jessie Slough, Johns River slough and Jessie Slough. Adipose clipped hatchery fish in parentheses.

Johns River slough Wishkah River slough Jessie Slough

Chinook 27 (2) 305 30 (2) coho 122 198 (1) 211 chum 2 1 0 cutthroat trout 4 6 0 bull trout 1 0 0

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 73

Figures 23: Catch trends for all wild Chinook and coho salmon by sampling period at the three fyke trap sites: Wishkah River slough, Johns River slough and Jessie Slough.

Catch trends by fyke site Johns River slough

Johns River slough was sampled eight times in 2011. Sampling occurred once a month from April through September, with an additional sampling in May and August. The slough has an area of 8,126 m2 and is subject to significant tidal inundation. This resulted in greater and more variable surface salinity measurements than either of the other fyke sites (Figure 24) and a different species composition (Figure 25).

Juvenile salmon were observed from April through September, with peak catches in May and June (Figure 23a). Coho salmon were the most abundant salmonid, followed by Chinook salmon. Unmarked juvenile coho were observed at the site from April through September, with

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 74

peak catches in mid-May. Four cutthroat trout smolts were trapped from May through August, and one bull trout with a fork length of 385 mm was caught in April.

Non-salmonids in Johns River slough were primarily euryhaline fish (Figure 25). The most common were northern anchovy and juvenile Pacific herring, captured in August and September. Simenstad and Eggers (1981) reported these species present in Grays Harbor from mid-May (anchovy) and mid-June (herring) through September. They also found strong evidence that Pacific herring were spawning and rearing in the estuary, with spawning thought to occur near eelgrass beds. The high occurrence of herring and anchovy in Johns River slough suggests that tidal sloughs are an important habitat for both these species, especially toward the tail end of their estuary residence.

Figure 24: Trends in surface salinity (A) and surface temperature (B) from April through September at the three fyke sites: Wishkah River slough, Johns River slough and Jessie Slough.

a 30

Wishkah River slough

25 Johns River slough 20 Jessie Slough 15 10

Surface salinity Surface (ppt) 5 0

b

20

)

̊C

(

e 15

r

u

t

a

r e

p 10 Jessie Slough m

e Johns River slough T Wishkah River slough 5

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 75

Wishkah River slough

Wishkah River slough was sampled six times in 2011. Sampling occurred once a month from April through August, with an additional sampling in June. The slough is the largest of the three fyke sites at 10,572 m2 and also the least saline, remaining essentially fresh water throughout the sampling period (Figure 24a).

Juvenile salmon were observed from April through August, with peak catches in April and May (Figure 23b). Chinook salmon were the most abundant salmonid, followed by coho salmon and then cutthroat trout. Wild juvenile Chinook salmon were observed at the site from April through August, with the highest catch in April, declining through August. Coho salmon were also captured throughout the sampling period, peaking in May and declining through August. In addition to the salmon, five smolts and one adult cutthroat trout were trapped from

April through July.

Non-salmonids in Wishkah River Slough included several exclusively fresh water species, including peamouth, large scale sucker and whitefish. Three-spine stickleback and peamouth accounted for about 80% of the non-salmon catch (Figure 25).

Jessie Slough

Jessie Slough was sampled six times in 2011. Sampling occurred once in May and July and twice in May and August. The site is a small blind channel (2,011m2) off the main slough and near it’s confluence with the Humptulips River. Salinity was low in May and June, but became increasingly brackish as the river levels dropped throughout summer (Figure 24a).

Juvenile salmon were observed from May through August, but sampling efforts in early June and late August captured no salmon (Figure 23c). A few juvenile Chinook salmon were observed at this site, but the most abundant catch was coho salmon, which contrasts with the nearby Humptulips River mainstem sites where juvenile Chinook were most abundant.

Non-salmonids in Jessie Slough were dominated by three species: shiner perch, staghorn sculpin and three-spine stickleback (Figure 25).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 76

Figure 25: Fyke trapping season totals for non-salmonids at each of the three fyke sites: Johns

River Slough, Wishkah River Slough and Jessie Slough.

a: Johns River Slough b: Wishkah River Slough

18 Anchovy 6 3 4 Threespine Stickleback

Peamouth 168 Herring (juv.) 14 202 658 17 Shad Shiner Perch 103 Largescale Sucker 291 Threespine Stickleback Speckled Dace Staghorn Sculpin 73 581 Prickly Sculpin Other Other

c: Jessie Slough

5 13 Shiner Perch

75 Staghorn Sculpin

Threespine Stickleback 204 79 Speckled Dace

Other

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 77

Section 6: Fish Community Modeling

6.1 Site Modeling – Physical Variables Due to the size of the Grays Harbor estuary and the number of tributaries entering near the estuary mouth, modeling the system presents interesting challenges in comparison with more traditional, “straight line” systems where one major river meets up with a small estuary before entering the ocean. As opposed to the straight line model, Grays Harbor can be visually described as a ”hand” (imagine looking at your left hand, where the wrist is the estuary mouth) where the fingers represent (in clockwise order) the Humptulips, Hoquiam, Chehalis, John’s and Elk Rivers (Figures 1, 2). In the absence of a good model of tidal flow for the estuary, the challenge is to relate the sampling areas along each of these fingers that, despite being relatively far apart in straight line distance in some cases (“as the crow flies”; e.g. Humptulips River sites and Elk River sites), are roughly equidistant from the estuary mouth due to the radial arrangement of the tributaries (Figure 2).

A Principal Components Analysis (PCA) of the core and secondary sampling sites’ physical variables was conducted to determine which characteristics were most influential in relating the sites (McCune & Grace 2002). The initial analysis was conducted using current, tidal height, salinity, and water temperature as the main factors, and revealed that the greatest variance in the data was explained by current (coded simply as flood, ebb, slack because actual velocity data were not available) while the second major axis was set by tidal height (Figure 26,

Table 11).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 78

Table 11: Results of the Principal Components Analysis (PCA) with 4 variables ------Eigenvector env 1 2 3 4 Current -0.5695 -0.3382 0.4922 0.5648 Height10 -0.1833 -0.8190 -0.4783 -0.2585 Salinity -0.4751 0.3973 -0.6953 0.3647 Temp -0.6453 0.2387 0.2135 -0.6936 ------

VARIANCE EXTRACTED, FIRST 4 AXES ------Broken-stick AXIS Eigenvalue % of Variance Cum.% of Var. Eigenvalue ------1 1.365 34.126 34.126 2.083 2 1.047 26.171 60.298 1.083 3 0.900 22.503 82.801 0.583 4 0.688 17.199 100.000 0.250

RANDOMIZATION RESULTS 999 = number of randomizations ------Eigenvalue Eigenvalues from randomizations from ------Axis real data Minimum Average Maximum p * 1 1.3651 1.0246 1.0906 1.2031 0.001000 2 1.0469 0.97291 1.0249 1.1007 0.132000 3 0.90011 0.89359 0.97375 1.0325 0.998000 4 0.68798 0.82225 0.91074 0.97800 1.000000 ------* p-value for an axis is (n+1)/(N+1), where n is the number of randomizations with an eigenvalue for that axis that is equal to or larger than the observed eigenvalue for that axis. N is the total number of randomizations.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 79

Figure 26: Plot of the Principal Components Analysis (PCA) with 4 variables

GH2011sitesPCA.12.13.PCA Axis2

Axis 1

However, it is clear while these two variables explain more than half of the variance (34.1% and 26.1%, respectively), salinity and temperature were also important, resulting in the “scatter” of the sampling sites in the figure, with no clearly defined principal axis. Current was the only significant factor in this model (p=0.001).

To test whether a measure of the distance between the sites was an important factor, and to test the hypothesis that tidal input was driving the interactions (tide influences flow, salinity, and temperature) we calculated the “distance from the estuary mouth” (in km) to each site, choosing as a reference point the center of the shipping channel between the jetties at the mouth of Grays Harbor. The distance to each site was then calculated along the thalweg of the major, and then minor in the case of tributaries, channel along which water would flow to reach the estuary mouth. The PCA output of this analysis, with 5 components, is shown below:

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 80

Table 12: Results of the Principal Components Analysis (PCA) with 5 variables

VARIANCE EXTRACTED, FIRST 5 AXES ------Broken-stick AXIS Eigenvalue % of Variance Cum.% of Var. Eigenvalue ------1 1.795 35.903 35.903 2.283 2 1.299 25.980 61.883 1.283 3 1.004 20.080 81.963 0.783 4 0.729 14.574 96.537 0.450 5 0.173 3.463 100.000 0.200

FIRST 5 EIGENVECTORS, scaled to unit length. These can be used as coordinates in a distance-based bi-plot, where the distances among objects approximate their Euclidean distances. ------Eigenvector env 1 2 3 4 5 Current 0.0633 -0.6840 -0.0296 -0.7261 -0.0062 Height10 -0.0665 -0.3238 -0.8770 0.3357 -0.0943 Salinity 0.7086 -0.0453 -0.0726 0.1016 0.6930 Temp 0.1415 -0.6186 0.4627 0.5781 -0.2214 Distance -0.6852 -0.2065 0.1028 0.1248 0.6796

RANDOMIZATION RESULTS 999 = number of randomizations ------Eigenvalue Eigenvalues from randomizations from ------Axis real data Minimum Average Maximum p * 1 1.7951 1.0394 1.1155 1.2302 0.001000 2 1.2990 1.0009 1.0488 1.1334 0.001000 3 1.0040 0.92738 0.99774 1.0755 0.348000 4 0.72868 0.87007 0.94990 1.0172 1.000000 5 0.17316 0.79464 0.88803 0.95693 1.000000 ------* p-value for an axis is (n+1)/(N+1), where n is the number of randomizations with an eigenvalue for that axis that is equal to or larger than the observed eigenvalue for that axis. N is the total number of randomizations.

Adding the distance to estuary mouth into the analysis did not improve the model, and only current and tidal height were significant components in relating the sites, explaining 35.9% and 26% of the variance in this model (both of these factors were significant in this model, with p=0.001 for both). The clustering of sampling sites in Figure 27 again shows a scatter pattern without a clear principal axis, indicating that there is no easily defined relationship between the physical characteristics of the sampling sites.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 81

Figure 27: Plot of the Principal Components Analysis (PCA) with 5 variables

GH2011sitesPCA.12.27.withestuarymouth Axis2

Axis 1

In 2012, we will create a new variable to relate the distance between sampling sites, utilizing the distance from the source river mouths (major tributaries) to the sampling sites to allow us to further investigate these data.

6.2 Salmon Modeling: Occurrence and Abundance In order to determine what factors influenced the occurrence and abundance of salmon across space and time in Grays Harbor, we constructed separate series of statistical models relating salmon abundance and occurrence to a series of temporal, spatial, and environmental variables. Models of abundance and occurrence were separated because of the inability to differentiate very low (undetected) abundance from cases where salmon were not present. The inclusion of data where salmon were not present in models of abundance could preclude obtaining reasonable results. Occurrence models were therefore used to understand what factors influenced whether salmon were present at a site on an occasion, and abundance models were used to understand what factors influenced abundance only when salmon were in fact present (e.g., (Mullahy 1986)). For occurrence models we used GLMs employing a logit-link function which assumes a binomial error distribution and relates a binary response

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 82

(presence/absence in this case) to linear combinations of predictor variables on a logit scale (Nelder & Wedderbu 1972). For abundance analyses we used GLMs employing a log-link function and assuming a negative binomial distribution, which is appropriate for over-dispersed count data, and functions similarly to a Poisson distribution. This analysis related catches of salmon, censored to remove occasions where no salmon were captured, to linear combinations of predictor variables on a log-scale. Separate occurrence and abundance models were constructed for each species, age class, and origin (hatchery or wild). For each species, age, and origin, all subsets of up to four predictor variables (Table 14), as well as the full model were tested in a series of additive main effects models.

We used a multi-model inference approach to evaluate how well models fit the data. Akaike’s Information Criterion for small sample sizes (AICc) was calculated to compare and rank the various models (AICc; (Burnham & Anderson 2002)). The difference between the AICc of a candidate model and the model with the lowest score (ΔAICc) was calculated and used to rank models. Models with ΔAICc less than or equal to 3 were considered to have substantial support, while those with values from 4 to 7 had some support and those greater than 7 had little support (Burnham & Anderson 2002). We also calculated Akaike weights (wi), indicating the strength of evidence supporting i as the best model, and relative likelihoods. Together these metrics were used to identify the best model or model group. All statistical analyses were performed in R (R Team 2011) unless otherwise noted.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 83

Table 13: Variables used in GLM models used to predict salmon occurrence across space and time in Grays Harbor, WA in 2011. Variable Definition Type Mean Range Date Day of Year Continuous -- 3/17 – 10/2 Month Month Factor -- 3 – 10 Zone Area of Grays Harbor Factor -- North, South, Central, Inner, Surge Plain Habitat Habitat Type Factor -- Aquatic Vegetation Bed, Cobble/Gravel/Sand, BeachEelgrass , Forested, High Emergent Marsh, Mud Flat, Sand Flat, Scrub/Shrub Cover Salinity Surface Water Salinity (ppt) Continuous 15.42 0 – 31.4 Temp Surface Water Temperature (C) Continuous 15.12 7.1 – 21.4 TideH Tide Height (ft) Continuous 3.67 -1.53 – 11.7 TideS Tide Stage Factor -- Ebb, Flood, High Water, Low Water Distance Distance to Harbor Mouth (mi) Continuous 20.33 5.92 – 49.59

The best model of subyearling wild Chinook occurrence included temperature, month, and zone, while the best model of abundance also included date (Table 14). Wild subyearling Chinook occurrence and abundance were negatively correlated with temperature, rose from March through June-July, before steadily dropping through October, and were greater in North Bay, South Bay, and Inner Estuary zones than in other zones. Wild subyearling Chinook abundance also declined over the season, but occurrence did not. Models of wild subyearling Chinook abundance were better differentiated than models of occurrence (ΔAICc of 5.82 vs. 0.32, respectively). The similar inputs in best models of both types lend confidence to the importance of the variable selected.

The best model of hatchery subyearling Chinook occurrence included water temperature and salinity, month, and habitat type, while models of abundance included all the same variables, substituting date for temperature (Table 14). Hatchery subyearling Chinook were generally not present until June and occurrence and abundance were greater thereafter, accounting for the importance of month, and potentially explaining the positive correlation with water temperature, which was generally greater later in the season. Hatchery subyearling Chinook occurrence was also positively correlated with salinity. Models of hatchery subyearling

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 84

Chinook were very poorly differentiated, however, with multiple models having ΔAICc < 3, whereas abundance models were modestly better differentiated.

Table 14: Model selection results for GLM models relating Chinook salmon abundance and occurrence to time of year, and environmental variables. The five best models including the full model, are listed from most plausible (ΔAICc=0) to least plausible. Models are separated by age class and by hatchery/wild origin.

Akaike Relative Cumulative Model1 AICc ΔAICc weight R2 likelihood Weight (wi) Subyearling Wild Occurrence Temp + Month + Zone 279.47 0.00 1.00 0.29 0.20 0.29 Temp + Month + Zone + TideS 279.79 0.32 0.85 0.25 0.22 0.54 Date + Temp + Month + Zone 280.79 1.32 0.52 0.15 0.21 0.69 Temp + Month + Zone + TideH 281.30 1.82 0.40 0.12 0.20 0.80 Temp + Salinity + Month + Zone 281.55 2.08 0.35 0.10 0.20 0.90

Subyearling Wild Abundance Date + Temp + Month + Zone 1440.35 0.00 1.00 0.80 0.35 0.80 Temp + Month + Zone 1446.17 5.82 0.05 0.04 0.32 0.84 Date + Temp + Month + Habitat 1446.65 6.30 0.04 0.03 0.34 0.87 Date + Temp + Salinity + Month 1447.51 7.16 0.03 0.02 0.30 0.90 Temp + Salinity + Month + Zone 1448.16 7.81 0.02 0.02 0.32 0.91

Subyearling Hatchery Occurrence Temp + Salinity + Month + Habitat 205.36 0.00 1.00 0.49 0.37 0.49 Date + Temp + Salinity + Month 206.74 1.38 0.50 0.24 0.44 0.73 + Habitat + TideH + TideS Temp + Month + Habitat + TideS 207.71 2.34 0.31 0.15 0.38 0.88 Salinity + Month + Habitat + TideS 211.01 5.65 0.06 0.03 0.37 0.91 Zone + Habitat + TideH + TideS Temp + Salinity + Month + Zone 211.72 6.35 0.04 0.02 0.33 0.93

Subyearling Hatchery Abundance Date + Salinity + Month + Habitat 288.12 0.00 1.00 0.48 0.60 0.48 Salinity + Month + Habitat 291.29 3.18 0.20 0.10 0.55 0.57 Date + Temp + Month + Habitat 291.79 3.67 0.16 0.08 0.57 0.65 Salinity + Month + Habitat + TideS 292.67 4.55 0.10 0.05 0.60 0.70 Salinity + Month + Habitat + TideH 293.13 5.01 0.08 0.04 0.56 0.74 1Yearling Chinook occurrence and abundance were not modeled since few were encountered.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 85

The best model of chum salmon occurrence included a date and salinity, while the best model of chum abundance included date, water temperature, and tide stage. Chum salmon occurrence was negatively correlated with date and positively correlated with salinity, while abundance was negatively correlated with date and temperature, and was greater during the flood tide than other stages. Catches occurred most frequently early in the sampling season in March, April, and May (Table 15). Models of chum occurrence were not well differentiated, with multiple models having ΔAICc < 3, while models of abundance were much better differentiated. All of the best models included date though, suggesting that timing was the dominate factor influencing chum occurrence. The inclusion of temperature and salinity in some models may result from the colder and lower salinity conditions which were generally prevalent during the early season when chum were present.

Table 15: Model selection results for GLM models relating chum salmon occurrence and abundance to time of year, and environmental variables. The five best models including the full model, are listed from most plausible (ΔAICc=0) to least plausible. Akaike Relative Cumulative Model11 AICc ΔAICc weight R2 likelihood Weight (wi) Subyearling Wild Occurrence Date + Salinity 86.07 0.00 1.00 0.31 0.70 0.31 Date + Temp + Salinity 87.51 1.44 0.49 0.15 0.70 0.46 Date + Salinity + TideH 87.78 1.70 0.43 0.13 0.70 0.59 Date + Temp 88.29 2.22 0.33 0.10 0.69 0.69 Date 89.14 3.07 0.22 0.07 0.71 0.76

Subyearling Wild Abundance Date + Temp + TideS 484.85 0.00 1.00 0.98 0.58 0.98 Date + Salinity + TideS 494.46 9.61 0.01 0.01 0.55 0.99 Date + Temp + Distance 494.82 9.97 0.01 0.01 0.60 0.99 Date + Temp + Zone 494.92 10.07 0.01 0.01 0.64 1.00 Date+ Temp 499.12 14.26 0.00 0.00 0.56 1.00 1All chum were assumed to be wild due to the inability to differentiate hatchery chum, which are not externally marked.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 86

The best model of wild subyearling coho salmon occurrence included date, salinity, zone, and habitat type, while the best model of abundance included the same variables substituting tide stage for zone (Table 14). Model selection results were well differentiated (no models within ΔAICc = 3) lending high confidence to the best models for both abundance and occurrence. Wild subyearling coho occurrence and abundance were negatively correlated with salinity, with occurrence declining quickly at salinities above 5 ppt, and coho essentially absent above 20 ppt. Occurrence of wild subyearling coho was greater in the inner estuary than other zones, and abundance and occurrence were greater in forested sites, shrub/scrub cover, and mudflats than other habitat types.

The best model of yearling wild coho occurrence included temperature, month, and zone (Table 16). Wild yearling coho were only present in April, May and June, corresponding to a rapid period of smolt emigration. The best model of hatchery coho abundance included date and month, and hatchery yearling coho were present during the same three month period as wild yearling coho. Both hatchery and wild yearling coho occurrence peaked in May. Models of ( wild?) yearling coho abundance were not well differentiated, complicating interpretation and inference, while models of hatchery yearling coho were not generated due to the small number of occasions in which they were encountered.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 87

Table 16: Model selection results for GLM models relating coho salmon occurrence and abundance to time of year, and environmental variables. The five best models including the full model, are listed from most plausible (ΔAICc=0) to least plausible. Models are separated by age class and by hatchery/wild origin. Akaike Relative Cumulative Model1 AICc ΔAICc weight R2 likelihood Weight (wi) Subyearling Wild Occurrence Date + Salinity + Zone + Habitat 225.33 0.00 1.00 0.88 0.43 0.88 Temp + Salinity + Zone + Habitat 230.87 5.54 0.06 0.06 0.41 0.93 Salinity + Month + Zone + Habitat 232.98 7.64 0.02 0.02 0.44 0.95 Date + Temp + Salinity + Month + 233.75 8.41 0.01 0.01 0.47 0.96 +Zone Zone + +Habitat Habitat + +TideH TideH + TideS+ TideS Date + Temp + Zone + Habitat 233.78 8.45 0.01 0.01 0.40 0.98

Subyearling Wild Abundance Date + Salinity + Habitat + TideS 484.85 0.00 1.00 0.98 0.58 0.98 Date + Salinity + TideS 494.46 9.61 0.01 0.01 0.55 0.99 Date + Salinity + TideH + TideS 494.82 9.97 0.01 0.01 0.60 0.99 Date + Salinity + Zone + TideS 494.92 10.07 0.01 0.01 0.64 1.00 Temp + Salinity + Month + TideS 499.12 14.26 0.00 0.00 0.56 1.00

Yearling Wild Occurrence Temp + Month + Zone 117.54 0.00 1.00 0.25 0.54 0.25 Temp + Month + Zone + TideH 118.82 1.28 0.53 0.13 0.54 0.38 Date + Temp + Month + Zone 118.89 1.35 0.51 0.13 0.54 0.51 Temp + Salinity + Month + Zone 119.21 1.66 0.43 0.11 0.54 0.62 Temp + Salinity + Month 119.25 1.71 0.43 0.11 0.50 0.72

Yearling Wild Abundance Date + Salinity + TideH + Distance 229.08 0.00 1.00 0.06 0.22 0.06 Temp + TideH 229.17 0.09 0.95 0.05 0.12 0.11 Temp + TideS 229.60 0.53 0.77 0.04 0.21 0.16 Temp + TideH + Distance 230.07 0.99 0.61 0.03 0.14 0.19 Date + Temp + TideH 230.29 1.22 0.54 0.03 0.14 0.22

Yearling Hatchery Occurrence Date + Month 74.89 0.00 1.00 0.12 0.33 0.12 Date + Temp + Month 75.03 0.15 0.93 0.11 0.34 0.23 Date + Salinity + Month 75.55 0.66 0.72 0.09 0.34 0.32 Temp + Month 76.00 1.11 0.57 0.07 0.31 0.39 Month 76.15 1.26 0.53 0.06 0.29 0.45 1Subyearling hatchery coho abundance and occurrence, and yearling hatchery coho abundance were not modeled since few were encountered.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 88

6.3 Long Term Data Collection Goals: Escapement The long term data collection goal for this project is to sample the Grays Harbor estuary in years that represent different levels of wild smolt productivity, including a “high”, “low” and “normal” year. Estimates of wild smolt productivity for the whole of Grays Harbor are unavailable. However, wild escapement estimates from WDFW (based upon index streams) are available, and these provide a coarse proxy to wild smolt production. It should be noted that using escapement as in index of smolt productivity relies on simplifying assumptions that disregard density-dependent effects and the considerable annual and stream-specific variation in productivity and survival.

Historical escapement estimates were available from 1969 for chum salmon and 1970 for Chinook and coho salmon. Spring Chinook salmon constitute a very small portion of the Grays Harbor population and were therefore omitted from the index. Annual escapement varied widely over the historical period of record, especially for coho salmon and chum salmon (Figure 28). The distribution of the escapement data is right skewed for all three species

(Chinook, g1 = 0.965; chum, g1 = 1.674; coho, g1 = 0.979), so we decided to use quartiles to define productivity categories since it reflects the spread of data: we used the interquartile range (50% of observations) to define the lower and upper limits of a “normal” year for each species of salmon (Figure 29, Table 17); escapement numbers below the 1st quartile represent a “low” productivity year; and escapement numbers above the 3rd quartile represent a “high” productivity year.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 89

Figure 28: Annual escapement estimates for salmon in Grays Harbor from 1969-2010

Figure 29: Box and whisker plot summarizing annual escapement data for fall Chinook salmon, chum salmon and coho salmon from 1969 to 2010. The interquartile range (grey box) represents the range of escapement values that constitute a “normal” productivity year. Annual escapement values that fall above or below the interquartile range are considered “high” or “low” productivity years.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 90

Table 17: Escapement values defining the interquartile range, or “normal” productivity year, for fall Chinook, coho and chum salmon. Values that fall above or below the interquartile range are considered “high” or “low” productivity years. Escapement estimates for the brood years that produced most of the juvenile salmon captured in 2011 are shown in red.

Fall Chinook Coho Chum

1st Quartile 7,577 21,048 11,000

3rd Quartile 15,955 48,562 24,595

2010 escapement 14,531 59,096 33,537 2009 escapement 84,425

Wild escapement of fall Chinook salmon in 2010, the brood year that produced the subyearling Chinook salmon caught in 2011, fell within the range of a “normal” productivity year as defined by the interquartile range (Table 17). Wild escapement of coho salmon in 2009 and 2010, the brood years that produced the yearling and subyearling fish captured in 2011, were greater than the 3rd quartile and were categorized as “high” productivity years. Wild escapement of chum salmon in 2010 was also categorized as a “high” productivity year.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 91

Section 7: Synthesis and Recovery Priorities Estuaries have been shown to enhance the survival of juvenile salmon by providing:

(1) habitat for overwintering juvenile salmonids forced downstream during high river flows (Sommer et al. 2001; Simenstad et al. 1992; Henning et al. 2007);

(2) complex low-velocity refugia such as off-channel sloughs and large woody debris (LWD) (Gonor et al. 1988; Swales & Levings 1989; Wick 2002; Henning et al. 2006; Simenstad & Eggers 1981; Henning et al. 2007);

(3) time for migrating juveniles to adapt physiologically to sea water (Folmar & Dickhoff 1980; Healey 1980; Levy & Northcote 1982; Iwata & Komatsu 1984; Zaugg et al. 1985);

(4) opportune feeding conditions as drift insects and other prey items are trapped and concentrated due to flow reversals (Tschaplinski 1987; Eggleston et al. 1998; Simenstad & Eggers 1981);

(5) settling of suspended sediments and detritus, which can fuel soft-sediment habitat formation and detritus-based food webs exploited by salmon (Simenstad et al. 1982; Maier & Simenstad 2009; Thorpe 1994; Bottom et al. 2005).

(6) a refuge from piscivorous and avian predators, due in part to the often turbid water resulting from river flows and tidal action (Gregory & Levings 1998; De Robertis et al. 2003; Simenstad et al. 1982; Thorpe 1994).

The results from our 2011 sampling indicate a wide range of habitat utilization by the various species and age classes of salmonids found in Grays Harbor. Clear growth trends were observed for subyearling Chinook and coho and chum salmon (and to a lesser extent for yearling Chinook salmon), indicating that the estuary is functioning as a refuge for these fish as they physiologically mature and prepare to enter the ocean. Individual points are taken up below.

7.1 Salmonid Spatial and Temporal Habitat Usage Life history variation has been recognized as a critical factor in the resilience - the ability of native stocks to persevere and recover from perturbations in abundance, habitat and climate variability- among Pacific Northwest salmonid stocks (Greene et al. 2010; Bottom et al. 2005; Simenstad et al. 1982). In many systems with enhanced hatchery production, the

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 92

range of life histories has been reduced via the selection of a limited diversity of broodstock, particularly with regard to run timing. In many systems, this has resulted in “pulses” of emigrants reaching the estuarine rearing habitats at the same time, which may create artificially heightened competition for estuarine habitat. While meeting our objective of documenting temporal and spatial habitat usage by salmonids in Grays Harbor, a secondary goal is to identify alternative estuarine life history strategies that are still extant in the Chehalis basin and may be important both for the restoration of salmon populations in the

watershed.

Chinook Salmon Chinook salmon were the second most abundant salmonid captured in 2011 (after chum salmon) and were present during all months of the study, peaking in May but present in large numbers from April through July (Figure 8). A peak catch in May was also reported in a study by Deschamps et al. (1970), and Simenstad and Eggers (1981) also reported catches of subyearling Chinook salmon nearly year-round in Grays Harbor. This suggests a variety of Chinook salmon life histories are present in the basin, and, particularly for wild Chinook, estuarine residence times are longer than for other salmonid species (see below). In 2011, wild subyearling fish dominated the catch, but hatchery subyearling Chinook were present in the estuary from July through September. Wild subyearling Chinook salmon were most abundant in the North Bay, Surge Plain, and Upper Estuary zones, while hatchery subyearling densities were highest in the Central Estuary and North Bay (Figures 9 and 10), consistent with hatchery inputs from the Wishkah and Humptulips Rivers. Both wild and hatchery subyearling Chinook salmon steadily increased in fork length from March through September (Figure 5). Wild subyearling Chinook salmon were widely distributed in all habitat types, but were most abundant in scrub/shrub habitats and lowest in beach and eelgrass habitats. Interestingly, only wild subyearling Chinook salmon were common at sand flat sites (Figure 12); previous reports suggest that subyearling Chinook salmon were most commonly encountered in shallow water habitats, entering tidal channels and flats during periods of high flows and flooding tides (Levings et al. 1986; Levings et al. 1991; K. Fresh et al. 2005; Healey 1980; Zaugg et al. 1985). In 2011, hatchery subyearling Chinook salmon were captured in highest densities at mud flat

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 93

habitats and were less abundant elsewhere, a result that may also be attributed to Humptulips River inputs into North Bay, which has the highest concentration of mud flats. Very few yearling Chinook salmon were caught in 2011; this result echoes an earlier study in the Chehalis basin (Deschamps et al. 1970). Yearling Chinook salmon have previously been reported to reside in deeper dendritic tidal or river channels (Zaugg et al. 1985; Thorpe

1994; K. Fresh et al. 2005), which may explain our low catches.

Coho Salmon Coho were the third most abundant salmonid captured in 2011, peaking in May. Previous studies in Grays Harbor reported subyearling coho captures from March through October, peaking in mid-April through mid-June (Brix 1981; Tokar et al. 1970). In 2011, subyearling coho salmon were captured throughout the sampling period; the densities were greatest in the Surge Plain, Inner Estuary and North Bay. The mean fork length of subyearling coho salmon steadily increased during the study from March to September (Figure 6), as expected. Yearling coho salmon moved quickly through the estuary from April to June (Figure 11b), and densities were greatest near the estuary mouth (North and South Bay zones), which may reflect better feeding opportunities and/or different prey fields at more productive littoral habitats in those areas. However, the average fork length of yearling coho salmon decreased from April through May in both marked and unmarked fish, corroborating the evidence that yearling coho did not reside for long periods of time within the estuary. This finding was also consistent with studies elsewhere that have observed larger coho smolts, which may be more ocean ready, migrating before smaller smolts (Irvine and Ward 1989).

Habitat usage by wild subyearling coho salmon was widespread, but not to the same degree as juvenile Chinook salmon; densities were highest in mudflats, scrub/shrub and forest habitats, and few were captured elsewhere. Few yearling coho salmon of either hatchery or wild origin were captured, making it difficult to draw conclusions, though wild yearlings had their highest densities at cobble beach, eelgrass and emergent marsh sites; the former two results may be a result of yearling’s preference for deeper channels (these habitats typically occur near drop offs) (Healey 1982a; Moser et al. 1991). Some subyearling coho were captured in tidal portions of the tributaries even in fall, suggesting that there is a diversity of coho life histories

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 94

present in Grays Harbor. These fish may leave the estuary for the ocean in their first year of life or seek out freshwater habitats and leave as smolts the next spring, the “nomad” lifestyle alluded to in other studies (Koski 2009; Bell & W. Duffy 2007). In the late 1960’s, coho in their third year (by scale analysis) were also captured in the lower Chehalis River surge plain, again suggesting that historically, multiple life histories existed within the basin (Deschamps et al.

1970), although no coho salmon large enough to be three year olds were captured in 2011.

Chum Salmon Emigration of juvenile chum salmon in the lower mainstem Chehalis River begins as early as January and is certainly underway by late February and early March (Deschamps et al. 1970). Chum salmon (all wild subyearlings) were present when sampling began in March of 2011, peaked in April (at densities as high as 7,000 fish per hectare in the central estuary), and had exited the estuary by then end of June (Table 6; Figures 8c, 11c), closely in agreement with the timing reported in previous studies in Grays Harbor (Wright 1973; Tokar et al. 1970; Brix 1981; Simenstad & Eggers 1981). The mean length of juvenile chum salmon remained the same in March and April, but increased considerably in May (Figure 7). This may indicate that the chum entering the estuary in March continued directly to the ocean, but those entering in April took up residence and increased in size through May. This pattern of estuarine use may reflect the inherent changes in productivity and the relative growth opportunities of the estuary versus the ocean early in the year. Using mark-recapture methods, Simenstad and Eggers (1981) estimated chum residence in Grays Harbor to 2-4 weeks, which is roughly what we observed in April and

May of 2011.

The overwhelming difference in chum salmon densities between North and South Bay (Figures 9, 11c) suggests there must be either a mechanism of exclusion from North Bay or a strong preference for South Bay, which has a greater diversity of habitats. The results of our regression modeling suggest that low salinities in North Bay early in the sampling year, perhaps as a result of elevated Humptulips River fresh water flows, may have acted as just such a mechanism (Section 6.2). Prevailing water currents driven by river, tidal and landscape features may discourage juvenile chum from entering North Bay. An alternative hypothesis is that some of these chum salmon originate from outside Grays Harbor, although their pattern (as a species)

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 95

of short estuarine residence argues against this point (Mason 1974; Healey 1982b; Iwata &

Komatsu 1984; Quinn 2005).

Chum salmon were found at highest densities off cobble beach sites (particularly Damon Point, just North of the estuary mouth), but were also common at forested, emergent marsh, eelgrass and aquatic vegetation sites, in that order. Note that cobble beach habitats occurred rarely in the Grays Harbor estuary (only at Damon Point and Half Moon Bay), near areas with deep channels, strong currents and/or strong tidal exchanges and wave action. Chum salmon were very rare in sand, mud flat or scrub/shrub cover habitats (Figure 12c). The timing of chum salmon catches in 2011 (Figure 9) was similar to that reported in 1981 (Simenstad & Eggers 1981); chum were present in the Surge Plain and Inner Estuary when sampling commenced, but became increasingly more abundant towards the estuary mouth (with the exception of North

Bay, where chum salmon were absent, as noted above).

Steelhead trout Just 21 steelhead trout were captured during the course of the sampling season, precluding any conclusions about their habitat preferences; densities were highest at mud flat sites, though catches were low in all habitats. The timing of steelhead catch in 2011 was generally in agreement with that reported in previous studies (Brix 1974; Tokar et al. 1970; Simenstad & Eggers 1981), though our sampling methodology does not effectively cover deeper channels where steelhead are likely to be in greater abundance (Quinn 2005).

Cutthroat trout Cutthroat trout were encountered infrequently, but enough were captured to observe some patterns of estuary use. Although present in all zones, cutthroat were most common in the Surge Plain (particularly early in the year: Figure 11d) and North Bay (often at the Humptulips River sites). Densities were highest at scrub/shrub cover and emergent marsh sites (Figure 12d). Their densities closely aligned in space and time with those of subyearling Chinook and coho salmon, suggesting possible predation (see the section on predators, below, for more information). Our catches of cutthroat trout demonstrate that this species is present in Grays

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 96

Harbor through much of the year and in a variety of habitats, in contrast with previous studies

(Deschamps et al. 1971; Deschamps et al. 1970; Simenstad & Eggers 1981).

Habitat Selectivity The Kruskal-Wallis comparisons revealed significant differences in fish densities among habitat types for subyearling hatchery Chinook; whilst differences for subyearling wild Chinook and subyearling wild coho were close to the alpha level. A series of pairwise comparisons using the Mann–Whitney U test revealed that the median densities of subyearling hatchery Chinook were significantly lower (p<0.05) in forested habitats than in mud flats, sand flats, aquatic vegetation beds and beach habitats, which might reflect a rapid outmigration by hatchery Chinook following their release. Subyearling hatchery Chinook densities were also significantly lower (p<0.05) in eelgrass and emergent marsh habitats than mud flats. The Jacob’s Index, which is weighted by habitat availability, highlighted some interesting patterns in habitat use between species, age classes, and wild or hatchery origin. Wild subyearling Chinook salmon used beach habitat, sand flats, eel grass beds and shrub cover to a much greater degree than was available in the estuary (Figure 14a). Hatchery subyearling Chinook salmon used forested habitat, emergent marsh, and especially shrub cover much less than their wild counterparts (Figure 14b), which again might reflect a rapid outmigration by hatchery Chinook post-release. Subyearling coho salmon used those habitats most associated with freshwater residency: forest, shrub cover and mud flats (Figure 14c). By contrast, yearling coho salmon smolts demonstrated strong preferences for beach habitat and eel grass beds, as did chum salmon (Figures 14 d, e). The patterns of habitat use between chum and yearling coho smolts were remarkably similar overall.

The results also indicate some general patterns of habitat use. Beach and eel grass habitats were used to a much greater degree than their proportional availability, which suggests they are especially important to salmon smolts in Grays Harbor estuary. Aquatic vegetation beds scored low on all the indices, which is surprising since these habitats are considered productive areas. Several theories might explain this: competition with non-salmon species could be higher since aquatic vegetation beds are utilized by these species (Figures 21, 22); aquatic vegetation beds may not be as productive in Grays Harbor in comparison with other systems because of

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 97

turbidity, or aquatic vegetation beds may over-represented in the Shore Zone Inventory data

(WADNR 2001).

Coded Wire Tags (CWT) Few CWT salmon were recovered in 2011, making generalizations difficult. The recovery of juvenile coho salmon that originated from the Queets River indicates that the Grays Harbor estuary is utilized non-basin hatchery fish for feeding, physiological maturation, or predator avoidance. A 2006 report (Kurt L Fresh et al. 2006) observed a similar type of pattern for Chinook salmon (use of an area by non-local hatchery salmon) in Puget Sound. It is also reasonable to hypothesize that naturally produced coho salmon from out of basin are also utilizing the area, given that these hatchery fish are entering estuarine waters at the same time as some of the wild smolts. However, it is not possible to test this hypothesis without recovering the CWT of wild fish from out of basin or conducting a DNA analysis of unmarked salmon. Tissue collection for DNA analysis of wild Chinook salmon juveniles is planned as part of our

2012 sampling effort.

Predation Few large pikeminnow or staghorn sculpin were captured in 2012, suggesting that size classes of these predators big enough to consume smolts are not overly abundant in Grays Harbor, although our sampling methodology does not cover the deeper channels effectively. There was evidence that other salmonids – namely steelhead and cutthroat trout – may prey upon juvenile Chinook and coho salmon, and perhaps chum salmon, as has been found in Puget Sound (E. J. Duffy & D. a. Beauchamp 2008). Steelhead densities, while low overall, were highest at mud flat sites where subyearling Chinook and coho salmon were also common (Figure 12d). Cutthroat trout were also rare, but densities were highest at scrub/shrub cover sites, where wild subyearling Chinook and coho were very common. In general, cutthroat peak abundance within a zone coincided with peak abundances of Chinook salmon and coho fry. In addition, the general pattern of cutthroat trout density within each zone appears to match changes in the density of subyearling Chinook and coho salmon as they progressively make their way down the estuary (Figures 9 a-e). This, along with information on the timing of catch densities by month (Figures 11) suggests cutthroat (and potentially steelhead) may forage upon the smaller salmon

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 98

year classes, though we have no direct evidence that this is occurring; juvenile salmon were usually outnumbered by non-salmonid species in the catches. Only 3 bull trout were captured in 2011; the low abundance suggests that bull trout, while present in the estuary, are not likely present in large enough numbers to be considered a major predator of juvenile salmonids. The low densities of bull trout captured in this study in 2011 are similar to the results of previous studies (Jeanes & Morello 2006; Tokar et al. 1970; Wright 1973; Simenstad & Eggers 1981).

Modeling summary A principal components analysis (PCA) of the physical site characteristics identified current and tidal height as the variables that explained the most variance (34.1% and 26.1%, respectively). The inclusion of an additional variable, the “distance from the estuary mouth,” did not significantly improve the model. The scatter of sampling sites in the ordination showed a pattern without a clear principal axis, indicating that there was no easily defined relationship between the physical characteristics available for the sampling sites in 2011.

Regression analysis of factors affecting salmon occurrence and abundance revealed differential use of the estuary between species and age classes of salmon. Date and sampling month were the two most commonly selected variables across species and age classes, highlighting the importance of seasonality in the use of the estuary by salmon. For chum and yearling wild and hatchery coho salmon, timing, as expected, explained much of the variability in abundance since these two species were only present in the spring and moved more rapidly through the estuary than Chinook salmon. Additionally, the period of highest abundance for wild yearling coho (May) was also the period of highest abundance for hatchery yearling coho. Chum and yearling coho salmon models also included a correlation with temperature and salinity, which may be an artifact of when sampling was initiated; catches of both species were greatest early in the season when temperatures were low, potentially explaining the effect of temperature on their abundance. For salinity, if chum had already emigrated from the riverine sites by the time our sampling commenced, it is understandable that their abundance was greater in higher salinity sites lower in the estuary. Alternatively, another explanation for the positive correlation between chum and yearling coho abundance and salinity is that these two species may move quickly through the freshwater portions of the estuary, spending slightly

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 99

longer in the higher salinity areas while undergoing the osmoregulatory transition necessary to enter the ocean. In contrast to yearling coho, the salinity was negatively correlated with subyearling wild coho abundance and subyearling coho were essentially absent at salinities above 20 ppt. Models of subyearling coho suggested that their use of the estuary was primarily confined to the freshwater and slightly brackish areas, particularly sites with scrub-shrub or forest cover, which may provide cover from predators. Subyearling Chinook were the most widely distributed species across space and time. Although timing was also correlated with their abundance, with occurrence and abundance peaking during June and July, subyearling Chinook, unlike yearling coho and chum salmon, were present in all sampling months. Additionally, the North Bay and the upper estuary had higher abundance of wild Chinook than other zones, suggesting the potential importance of these areas to wild Chinook production. Importantly, the spring and early summer may offer wild Chinook some refuge from competition with hatchery Chinook, which were not present in large numbers until June. However, the peak in abundance for both hatchery and wild subyearling Chinook was June through August.

7.2 Synthesis: Non-Salmonid/Baitfish Distributions and Abundance In evaluating the role Grays Harbor estuary provides for salmon, it is important to remember that salmon comprise a minority of the total fish community. During 2011, we caught 46 species of non-salmonids from mid-March to early October (Table 9), nine of which were considered common in at least some zones and time periods: surf smelt, three-spine stickleback, shiner perch, northern anchovy, staghorn sculpin, Pacific herring, saddleback gunnel, English sole and peamouth. The most abundant species were surf smelt (Hypomesus pretiosus pretiosus) and three-spine stickleback (Gasterosteus aculeatus), accounting for 29% and 27% of the non- salmon catch respectively (Figure 16). Other studies examining the diets of these non-salmonids indicated very little predation on salmon, but potential for competition where diets overlap

(Bottom & Kim K Jones 1990).

Baitfish

Six species of baitfish were caught in 2011, three of which - surf smelt, northern anchovy and Pacific herring - were consistently abundant over the sampling period, indicating residence

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 100

and utilization of the estuary. Surf smelt were the most common species of baitfish, occurring in every month of sampling, every zone, and most habitat types (Figures 17a, 19a). Northern anchovy were the second most abundant, followed by Pacific herring. Pacific sand lance were caught infrequently. The lower intertidal zones - South Bay, the Central Estuary and North Bay – have higher salinities and produced high densities of all three common baitfish species (Figure 19), while the peak abundance of surf smelt occurred at the estuary mouth. The two other species of baitfish, American shad and Pacific sardine, were rarely encountered and never abundant. All the baitfish shared common patterns of association with cobble beach, sand flat and eelgrass habitat types (Figure 21), which partially reflects the availability of these habitats in the lower estuary. However, the relative densities of baitfish at eelgrass sites far exceeds the relative contribution of eelgrass to overall habitat complexity in the estuary, which suggests that these species/life stages have a strong preference for eelgrass habitats.

Surf smelt were at their peak abundance during March (almost 4000 fish/ha) primarily as post-larval fish (Figure 17a). By April, smelt densities fell considerably, but these juvenile fish continued to be present in the catch in substantial (although diminishing) numbers until August. Juvenile Pacific herring appeared to begin entering Grays Harbor in April and peaked in abundance by August (Figure 17b). There was no obvious relationship between the occurrence of post-larval/juvenile herring and adult herring to suggest that these influxes were associated with spawning in the estuary, although Simenstad and Eggers (1981), indicated there was evidence of estuarine spawning and rearing, with spawning thought to occur near eelgrass beds. Patterns of abundance for northern anchovy were unclear: adult fish began entering the estuary in April, but average densities were low until August when they peaked at about 600 fish per hectare (Figure 17c).

Other Common Non-salmonids

Of the 40 or so other species of non-salmonid, six were considered common: three-spine stickleback, shiner perch, Pacific staghorn sculpin, saddleback gunnel, English sole, and peamouth. Three-spine stickleback and shiner perch were the most abundant occurring in

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 101

almost every month, every zone and every habitat type (Figures 18a,20a, 22a). Both species reproduced in the estuary and both species were present in various life history stages.

Three-spine stickleback were captured throughout the sampling period, but recruitment peaked in August and September (Figure 18a) when the new cohort of juveniles became ever- present in our catch. Although stickleback were present in all the sampling zones, the greatest densities occurred in the Central Estuary, North Bay and South Bay (Figure 20a), preferentially in eelgrass habitats.

Shiner perch began to appear in our samples in April and steadily increased in abundance through July, the peak month of reproduction for this viviparous species (Figure 18a). The new cohort took over as the dominant life stage; and the average density of juvenile shiner perch in July increased from 19 fish/ha in May to 866 fish/ha in July. Shiner perch were present throughout the estuary, but South Bay was a particularly productive zone for this species (Figure 20a).

Staghorn sculpin were another common catch at nearly every site except for those that were fully freshwater. Both staghorn sculpin and saddleback gunnel shared common patterns of abundance, by zone and by habitat type (Figures 20b, 22b), although the distribution of staghorn sculpin extended a little higher in the estuary. Interestingly, peak abundance by month is quite separate for these species (Figure 18b), which implies that the timing of reproduction may have co-evolved to reduce competition for resources, or that competition between the two species drives the relative abundance.

English sole and starry flounder (Platichthys stellatus) were the two most common flatfish collected in 2011. Grays Harbor is an important rearing area for juvenile English sole, which utilize the extensive shallow sublittoral and lower littoral habitats (Simenstad & Eggers 1981). English sole larvae settle out in the estuary from March through May after a period of pelagic residence in nearshore marine waters. In 2011, densities of juvenile English sole peaked in May (approximately 250 fish/ha) and then diminished toward the end of the season (Figure 18c), which corresponds with the period of larval settlement.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 102

Several species of the minnow family (Cyprinidae) were captured in Grays Harbor, but only peamouth was considered common. This species is more freshwater oriented than the other euryhaline species in Grays Harbor, which explains why they were most abundant in the surge plain and Inner estuary, and in forested and shrub cover habitats (Figures 20d, 22d). Peamouth appeared in the catch May and increased in abundance throughout the rest of the sampling season (Figure 18d). Presumably, adult fish reside further upstream in winter and move down to the Inner estuary in spring to reproduce: postlarval and juvenile peamouth became an increasingly large portion of the catch as the season progressed.

Another species of minnow, pikeminnow (Ptychocheilus oregonensis), is considered an important predator of juvenile salmon in some systems. (Pikeminnow predation on juvenile salmon in the Chehalis is reviewed in detail by Fresh et al. (2003). Catches of pikeminnow were restricted almost exclusively to the surge plain (Figure 20d), which reflects the freshwater nature of this species. Average monthly densities were low, ranging from 0.2 to 13 fish/ha; but it should be noted that our shoreline sampling efforts targeted mainly juvenile pikeminnow, whereas mature fish tend to reside in the deeper channels.

We did not catch any spiny dogfish (Squalus acanthias), Pacific Tomcod (Microgadus proximus), or Pacific hagfish (Eptatretus stoutii) in 2011, although these species were reported in low abundance in previous studies of the Grays Harbor estuary (Tokar et al. 1970; Wright 1973).

7.3: Grays Harbor Salmonid Habitat Recovery Priorities The size of the Grays Harbor estuary provides many opportunities for restoration activities to enhance salmonid recovery, so one of our goals was to determine which areas are most exploited by juvenile salmon and thus should be restoration priorities. The following list consists of areas where juvenile salmonid densities were high and should be considered as top priorities:

1) A tide gate near Ocosta State Park (south side of the estuary, between the old Weyerhaeuser mill site and the south bay bridge), near the confluence with the estuary, blocks salmonid entry to this potentially productive rearing habitat. The area that would become tidally influenced if a modern tide gate was put in place (or if the old “flap” style

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 103

gate were simply removed) is estimated at 56 hectares; the estimated area that was historically inundated prior to development is in excess of 100 hectares (aerial photo, with

the area isolated by the tide gate highlighted in green, is shown below).

2) Another “flap” style tide gate is located at the slough between the south bay bridge and Westport (highlighted in green on the aerial photo below) and also excludes juvenile salmon from reaching high quality rearing habitat adjacent to the South Bay zone. Removal of this tide gate would result in access to an estimated 40 hectares of additional habitat for juvenile

salmon.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 104

3) We observed a natural break in one of the dikes on the East side of Johns River, due to tidal erosion. This area, still heavily diked, contains abundant rearing habitat for juvenile salmon (estimated to be in excess of 150 hectares) and should be considered for restoration projects. Although natural production of salmon from the Johns River is low, this area contains excellent rearing habitat and could aid salmonid recovery in the basin (picture below).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 105

4) The South Bay zone (Elk River estuary) was heavily utilized by juvenile salmon of all species, even though natural production is low in the Elk River drainage (similar to the Johns River). Fortunately, much of the land is protected through private ownership, although the eastern shore is owned by a timber company. In order to preserve this area, shoreline buffers from logging need to be maintained or expanded. There is the potential for water quality in this zone to be negatively affected by effluent from the cranberry farms located in the headwaters of both the Elk and Johns Rivers, as these effluents can contain excessive concentrations of insecticides that could harm juvenile salmon and/or their prey (Wood &

Stark 2002), and water quality monitoring to counter this potential threat is recommended.

7.4: Response to Specific Review Panel Questions

Hatchery vs. Wild Salmon Displacement

The degree to which wild and hatchery salmon use of Grays Harbor overlaps may be of importance since estuary rearing habitat is thought to be a constraining factor on production for salmon, particularly, Chinook. If hatchery fish cause density dependent displacement of wild conspecifics, this could negatively impact wild stocks. Although our study was not designed to explicitly test whether displacement occurred - a complex question that would likely entail telemetry, individual-based models, or other analytical approaches - we can report the degree to which occurrence and abundance of hatchery and wild conspecifics overlapped.

We are only able to report (with any confidence) on the overlap of hatchery and wild salmon subyearling Chinook, since large numbers of hatchery subyearlings are released and are mass-marked, enabling identification in the field. Although chum and yearling coho are also released, chum are not marked, precluding comparison, and a small proportion (~17%) of yearling coho were marked in the contributing brood year.

Wild subyearling Chinook were the most widely distributed species/age class across space and time, although there was a progressive shift in densities over the season from peak concentrations in the surge plain early in the season, to the upper and central estuary in May,

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 106

June and July, and finally to littoral habitats in the lower estuary later (Figures 9e, 11a, Table 14). Peak density diminished through this progression as the juvenile Chinook salmon dispersed, from 650 fish/ha in the surge plain in May to 114 fish/ha at the estuary mouth in July. In contrast, hatchery subyearling Chinook were not present until June, suggesting that the early season may provide a temporal refuge from potential competition for wild Chinook. Hatchery Chinook also appeared to move more quickly through the estuary, as demonstrated by the positive correlation between their abundance and salinity. This more rapid emigration may also minimize competition between the two species. Finally, in stark contrast to other areas in Washington, such as Puget Sound, densities of hatchery Chinook were consistently much lower than wild fish, including where they co-occurred, also potentially limiting the scope for competition (e.g., Duffy et al. 2011).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 107

Additional Goals for 2012

 Determine if Grays Harbor estuary habitats are being used by salmonids from other basins, particularly the Columbia River system, on their northward migration. Chinook salmon will be sampled (fin clips) in areas near the estuary mouth in 2012 and submitted for genetic analysis to determine if emigrants from other systems are utilizing Grays Harbor for rearing.  A community analysis via non-metric multi-dimensional scaling (NMDS) will be conducted on salmonid and non-salmonid fish communities to further investigate the relationships potentially governing fish distributions in the estuary. This will also allow us to determine the potential for competition between species in the various habitats.  2011 was an unusual year in terms of weather, with the wettest, coldest spring in 60 years and a relatively wet summer and dry autumn. A goal of this project is to sample in years with “normal”, “low” and “high” escapement (Section 6.3) for the three main salmonid species (Chinook, coho and chum salmon) in order to analyze how the spatial and temporal distribution of juvenile salmon changes as the density of these species in the estuary fluxes. For wild subyearling Chinook salmon, 2011 fell within the range of a “normal” productivity year, while both wild coho and chum salmon 2011 was categorized as a “high” productivity year (Table18). Thus, a minimum of two additional sampling years will be required in order to sample these species in years where the escapement

values fall into the other ranges.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 108

Acknowledgements A large number of people were involved in the 2011 sampling effort, which was blessed with the wettest, coldest spring in 60 years, and we are indebted to them for their assistance (and persistence) in the field: Frank Staller (Wild Fish Conservancy), Molly Gorman, Jessica Jang and Rebecca Gehri (WFC Interns), Christian Sartin, , Elanor Wolff, Robert Wadsworth, Megan Lang, Sarah Farley, Julia Comstock-Ross, Jordan Miller, Zach Andre, Matt Buss, Andrew Muchlinski, , Tanya Budiarto, Craig Zora and Jim Huinker (WA DNR), Craig Francis Peters, Kim Jones , Nicholas Pace, and Curtis Vincent. The project also benefitted from the able assistance of members of the Chehalis Tribe DNR: we wish to thank Glen Connelly, Mark White, and other members of the Chehalis Tribe for their time and the use of their airboat to reach and sample difficult sites in North Bay. We also want to acknowledge Bob Burkle (WDFW) for sharing his knowledge of the Grays Harbor estuary, Charmane Ashbrook (WDFW) for assistance in locating and compiling salmon escapement data, and the WDFW Aberdeen office for the loan of a CWT wand for the 2011 sampling season. Steven Ferraro, Pat Clinton and their colleagues at the U.S. Environmental Protection Agency (Newport, Oregon) kindly shared data on the distribution of eelgrass and other habitats in Grays Harbor. Thanks also to DNR staff from the Quinault tribe for assistance with determining the origin and time of release for the coded-wire tagged coho salmon released from their nation. We wish to thank the commercial fishing community in Westport for advice and knowledge on the distribution and historical trends of fisheries in the region, as well as hazards to navigation. Finally, we wish to thank the numerous private landowners, Grays Harbor Audubon, the Cascade Land Conservancy, Washington State Parks, WDFW, Washington DNR, the Port of Grays Harbor and the cities of Aberdeen and Hoquiam for making much of this work possible by allowing us to access sampling sites on their property. This project was funded by the Grays Harbor county Salmon Recovery Funding Board

(RCO #10-1412P) and the U.S. Fish and Wildlife Service and is available online as a pdf

(as of February, 2012).

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 109

References

Bell, E. & Duffy, W., 2007. Previously Undocumented Two-Year Freshwater Residency of Juvenile Coho Salmon in Prairie Creek, California. Transactions of the American Fisheries Society, 136(4), pp.966- 970.

Berggren, T.J. & Filardo, M.J., 1993. An analysis of variables influencing the migration of juvenile salmonids in the Columbia River basin. North American Journal of Fisheries Management, 13, pp.48–63.

Borde, A.B. et al., 2003. Geospatial habitat change analysis in Pacific Northwest Coastal estuaries. Estuaries, 26(4B), pp.1104-1116.

Borde, A.B. et al., 2010. Geospatial Habitat Change Analysis in Pacific Northwest Coastal Estuaries. Estuaries, 26(4), pp.1104-1116.

Bottom, D.L. & Jones, Kim K, 1990. Species composition, distribution, and invertebrate prey of fish assemblages in the Columbia River estuary. Progress in Oceanography, 25, pp.243-270.

Bottom, D.L. et al., 2005. Salmon at River’s End: The Role of the Estuary in the Decline and Recovery of Columbia River Salmon,

Boule, M.E., Olmsted, N. & Miller, T., 1983. Inventory of wetland resources and evaluation of wetland management in Western Washington.

Brix, R., 1974. 1973 studes of juvenile salmonids in rivers tributary to Grays Harbor, Washington, Olympia: Washington Department of Fisheries.

Brix, R., 1981. Data report of Grays Harbor juvenile salmon seining program, 1973 - 1980, Olympia: Washington Department of Fisheries.

Burnham, K.P. & Anderson, D.., 2002. Model selection and inference: a practical information theoretical approach, New York: Springer-Verlag.

Chapman, D.G., 1951. Some properties of the hypergeometric distribution with applications to zoological censuses, University of California Publications on Statistics.

Dawley, E.M. et al., 1986. Migrational characteristics, biological observations, and relative survival of juvenile salmonids entering the Columbia River Estuary, 1966-1983,

Deschamps, G., Wright, S.G. & Watson, R.E., 1970. Fish migration and distribution in the lower Chehalis River and upper Grays Harbor W. D. of Fisheries, ed. , Technical , pp.1-59.

Deschamps, G., Wright, S.G. & Watson, R.E., 1971. Fish migration and distribution in the lower Chehalis River and upper Grays Harbor W. D. of Fisheries, ed. , Technical , pp.1-59.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 110

Duffy, E.J. & Beauchamp, D. a., 2008. Seasonal Patterns of Predation on Juvenile Pacific Salmon by Anadromous Cutthroat Trout in Puget Sound. Transactions of the American Fisheries Society, 137(1), pp.165-181.

Duffy, E.J. et al., 2011. Ontogenetic Diet Shifts of Juvenile Chinook Salmon in Nearshore and Offshore Habitats of Puget Sound. Transactions of the American Fisheries Society, (January 2012), pp.37-41.

Eggleston, D.B. et al., 1998. Estuarine fronts as conduits for larval transport: hydrodynamics and spatial distribution of Dungeness crab postlarvae. Marine Ecology-Progress Series, 164, pp.73-82.

Folmar, L.C. & Dickhoff, W.W., 1980. The parr-smolt transformation (smoltification) and seawater adaptation in salmonids. Aquaculture, 21, pp.1-37.

Fresh, K. et al., 2005. Role of the estuary in the recovery of Columbia River basin salmon and steelhead: An evaluation of the effects of selected factors on salmonid population viability. , p.125.

Fresh, Kurt L, Schroeder, S.L. & Carr, M.I., 2003. Predation by Northern Pikeminnow on Hatchery and Wild Coho Salmon Smolts in the Chehalis River , Washington. North American Journal of Fisheries Management, (Seiler 1989), pp.1257-1264.

Fresh, Kurt L et al., 2006. Juvenile Salmon use of Sinclair Inlet, Washington in 2001 and 2002. Washington Department of Fish and Wildlife, (March).

Gonor, J.J., Sedell, J.R. & Benner, P.A., 1988. What we know about large trees in estuaries, in the sea, and on coastal beaches. In C. Maser et al., eds. From the forest to the sea: a story of fallen trees. Portland, OR: U.S. Forest Service, Pacific Northwest Research Station, pp. 83-112.

Greene, C.M. et al., 2010. Improved viability of populations with diverse life-history portfolios. Biology Letters, 6(3), pp.382-386.

Gregory, R.S. & Levings, C.D., 1998. Turbidity reduces predation on migrating juvenile Pacific salmon. Transactions of the American Fisheries Society, 127(2), pp.275-285.

Healey, M.C., 1982a. Juvenile Pacific salmon in estuaries: The life support system. In V. S. Kennedy, ed. Estuarine Comparisons . New York: Academic Press, pp. 315-342.

Healey, M.C., 1982b. Timing and relative intensity of size-selective mortality of juvenile Chum salmon (Oncorhynchus keta) during early sea life. Canadian Journal of Fisheries and Aquatic Sciences, 39(7), pp.952-957.

Healey, M.C., 1980. Utilization of the Nanaimo River estuary by juvenile Chinook salmon, Oncorhynchus tshawystcha. Fishery Bulletin, 79(3), pp.653-668.

Henning, J.A., Gresswell, R. E. & Fleming, I. a., 2007. Use of seasonal freshwater wetlands by fishes in a temperate river floodplain. Journal of Fish Biology, 71(2), pp.476-492.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 111

Henning, J.A., Gresswell, Robert E. & Fleming, Ian A., 2006. Juvenile Salmonid Use of Freshwater Emergent Wetlands in the Floodplain and Its Implications for Conservation Management. North American Journal of Fisheries Management, 26(2), pp.367-376.

Hering, D.K. et al., 2010. Tidal movements and residency of subyearling Chinook salmon (Oncorhynchus tshawytscha) in an Oregon salt marsh channel. Canadian Journal of Fisheries and Aquatic Sciences, 67(3), pp.524-533.

Hiss, J., Meyer, J. & Boomer, R., 1982. Status of Chehalis River salmon and steelhead fisheries and problems affecting the Chehalis Tribe. , p.61.

Iwata, M. & Komatsu, S., 1984. Importance of estuarine residence for adaptation of chum salmon (Oncorhynchus keta) fry to seawater. Canadian Journal of Fisheries and Aquatic Sciences, 41(5), pp.744-749.

Jacobs, J., 1974. Quantitative measurements of food selection. Oecologia, 14, pp.413-417.

Jeanes, E.D. & Morello, C.M., 2006. Native char utilization: Lower Chehalis River and Grays Harbor Estuary, Aberdeen, Washington. , p.81.

Koski, K.V., 2009. The Fate of Coho Salmon Nomads : The Story of an Estuarine-Rearing. Ecology And Society, 14(1).

Levings, C.D., Conlin, K. & Raymond, B., 1991. Intertidal habitats used by juvenile Chinook salmon (Oncorhynchus-tshawytscha) rearing in the north arm of the Fraser-River estuary. Marine Pollution Bulletin, 22(1), pp.20-26.

Levings, C.D., McAllister, C.D. & Chang, B.D., 1986. Differential use of the Campbell River estuary, British- Columbia, by wild and hatchery-reared juvenile Chinook salmon (Oncorhynchus-tshawytscha). Canadian Journal of Fisheries and Aquatic Sciences, 43(7), pp.1386-1397.

Levy, D.A. & Northcote, T.G., 1982. Juvenile salmon residency in a marsh area of the Fraser River estuary. Canadian Journal of Fisheries and Aquatic Sciences, 39(2), pp.270-276.

Maier, G.O. & Simenstad, C.A., 2009. The Role of Marsh-Derived Macrodetritus to the Food Webs of Juvenile Chinook Salmon in a Large Altered Estuary. Estuaries and Coasts, 32(5), pp.984-998.

Mason, J.C., 1974. Behavioral ecology of chum salmon fry in a small estuary. Journal of the Fisheries Research Board of Canada, 31(1), pp.83-92.

Mayama, H. & Ishida, Y., 2003. Japanese studies on the early ocean life of juvenile salmon,

McCune, B. & Grace, J.B., 2002. Analysis of ecological communities, Gleneden Beach, Oregon: MJM Software Design.

Moser, M.L., Olson, A.F. & Quinn, T.P., 1991. Riverine and estuarine migratory behavior of coho salmon (Oncorhynchus kisutch) smolts. Canadian Journal of Fisheries and Aquatic Sciences, 48(9), pp.1670- 1678.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 112

Mullahy, J., 1986. Specification and testing of some modified count data models. Journal of Econometrics, 33(3), pp.341-365.

Naiman, R.J. & Sibert, J.R., 1979. Detritus and juvenile salmon production in the Nanaimo Estuary. III. Importance of detrital carbon to the estuarine ecosystem. Journal of the Fisheries Research Board, Canada, 36, pp.497-503.

Nelder, J.A. & Wedderbu, R.W., 1972. Generalized Linear Models. Journal of the Royal Statistical Society, Series A-General, 135(3), pp.370-384.

Quinn, T.P., 2005. The behavior and ecology of Pacific salmon and trout, Bethesda, Maryland: American Fisheries Society and the University of Washington Press.

R Team, R.D.., 2011. R: A language and environment for statistical computing.

Ralph, S.C., Peterson, N.P. & Mendoza, C.C., 1994. An inventory of off-channel habitat of the lower Chehalis River with applications of remote sensing.

De Robertis, A. et al., 2003. Differential effects of turbidity on prey consumption of piscivorous and planktivorous fish. Canadian Journal of Fisheries and Aquatic Sciences, 60(12), pp.1517-1526.

Schroder, S. & Fresh, K., 1992. Results of the Grays Harbor coho survival investigations, 1987-1990. , p.413.

Simenstad, C.A. & Eggers, D.M., 1981. Juvenile salmonid and baitfish distribution, abundance, and prey resources in selected areas of Grays Harbor, Washington, Seattle.

Simenstad, C.A. et al., 1992. Ecological status of a created estuarine slough in the Chehalis River estuary: report of monitoring in created and natural estuarine sloughs.

Simenstad, C.A., Fresh, K L & Salo, E.O., 1982. The role of Puget Sound and Washington coastal estuaries in the life history of Pacific salmon: An unappreciated function. In V. S. Kennedy, ed. Estuarine Comparisons. New York: Academic Press, pp. 343-364.

Smith, C.J. & Wenger, M., 2001. Salmon and Steelhead habitat limiting factors: Chehalis basin and nearby drainages, WRIA 22 and 23. , p.448.

Sommer, T.R. et al., 2001. Floodplain rearing of juvenile chinook salmon: evidence of enhanced growth and survival. Canadian Journal of Fisheries and Aquatic Sciences, 58(2), pp.325-333.

Swales, S. & Levings, C.D., 1989. Role of off-channel ponds in the life-cycle of coho salmon (Oncorhynchus-kisutch) and other juvenile salmonids in the Coldwater River, British Columbia. Canadian Journal of Fisheries and Aquatic Sciences, 46(2), pp.232-242.

Team, W.E., 1997. A Comparative Assessment of a Natural and Created Estuarine Slough as Rearing Habitat for Juvenile Chinook and Coho Salmon. Estuaries, (4), pp.792-806.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 113

Thom, R M, 1993. Eelgrass (Zostera marina L.) transplant monitoring in Grays Harbor, Washington, after 29 months. , p.21.

Thorpe, J.E., 1994. Salmonid fishes and the estuarine environment. Estuaries, 17(1A), pp.76-93.

Tokar, E.M., Tollefson, R. & Denison, J.G., 1970. Grays Harbor: downstream migrant salmonid study. , p.93.

Tschaplinski, P.J., 1987. The use of esturies as rearing habitats by juvenile coho salmon. In T. W. Chamberlin, ed. Applying 15 years of Carnation Creek Results. Nanaimo, B.C.: Pacific Biological Station, pp. 123-141.

USDA, 2009. NAIP Aerial Imagery. , 2011(January).

USFWS, 2010. Classification of Wetlands and Deepwater Habitats of the United States. , 2011(January).

WADNR, 2001. Washington State shorezone inventory. , 2011(January).

Wick, A.J., 2002. Ecological function and spatial dynamics of large woody debris in oligohaline brackish estuarine sloughs for juvenile Pacific salmon. School of Aquatic and Fisheries Sciences, p.125.

Wissmar, R.C. & Simenstad, C.A., 1988. Energetic Constraints of Juvenile Chum Salmon (Oncorhynchus keta) Migrating in Estuaries. Canadian Journal of Fisheries and Aquatic Sciences, 45, pp.1555-1560.

Wood, B. & Stark, J.D., 2002. Acute toxicity of drainage ditch water from a Washington state cranberry- growing region to Daphnia pulex in laboratory bioassays. Ecotoxicology and Environmental Safety, 53(2), pp.273-280.

Wright, S.G., 1973. Resident and anadromous fishes of the Chehalis and Satsop Rivers in the vicinity of Washington public power supply system’s proposed nuclear project no.3. , p.26.

Zaugg, W.S., Prentice, E.F. & Waknitz, F.W., 1985. Importance of river migration to the development of seawater tolerance in Columbia River anadromous salmonids. Aquaculture, 51(1), pp.33-47.

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 114

Appendices

Appendix 1: Core (bold text), secondary and fyke sites sampled in 2011

Site Name latitude longitude habitat zone type Beardslee Slough 46.867158 -124.038720 Aquatic Vegetation Bed South Bay core site Campbell Slough (near Humptulips) 47.044676 -124.058826 Scrub/Shrub Cover North Bay secondary site Chehalis river near restoration site 46.947571 -123.654645 Scrub/Shrub Cover Surge Plain core site Chenois Creek Flats 47.009079 -124.033366 Aquatic Vegetation Bed North Bay core site Cow Point 46.961900 -123.847065 High Emergent Marsh Upper estuary core site Damon Point 46.945315 -124.105779 Cobble/Gravel/Sand Beach Central estuary core site East Fork Hoquiam River 47.025343 -123.875533 Forested Upper estuary core site Elk River Flats (South Bay Forest) 46.856100 -124.034283 Forested South Bay core site Goose Island Flats (NW corner) 46.981537 -124.073087 Aquatic Vegetation Bed Central estuary core site Half Moon Bay (Westport) 46.904246 -124.130385 Cobble/Gravel/Sand Beach Mouth secondary site Humptulips Right River Channel 47.049975 -124.046594 Forested North Bay core site Humptulips River mouth 47.035434 -124.052548 Aquatic Vegetation Bed North Bay core site Jessie Slough 47.043077 -124.053996 Scrub/Shrub Cover North Bay fyke Johns River Boat Launch Seine Site 46.899973 -123.996389 High Emergent Marsh Central estuary core site Johns River Bar 46.921800 -124.015204 Aquatic Vegetation Bed Central estuary core site Johns River channel 46.902084 -123.987036 High Emergent Marsh South Bay core site Johns River slough 46.889758 -123.978253 High Emergent Marsh South Bay fyke Little Hoquiam River 46.987709 -123.913718 Scrub/Shrub Cover Upper estuary secondary site Lower Elliot Slough 46.978491 -123.768695 Forested Surge Plain core site Lower Wishkah River 47.027065 -123.811352 Forested Upper estuary secondary site Mallard Slough 46.855625 -124.061887 High Emergent Marsh South Bay core site Moon Island 46.968083 -123.944742 Aquatic Vegetation Bed Upper estuary core site North Bay Flats (Humptulips channel) 47.010698 -124.091296 Aquatic Vegetation Bed North Bay secondary site Ocean Shores Flats 46.997688 -124.111703 Aquatic Vegetation Bed North Bay secondary site Preacher Slough Restoration (bridge) 46.944667 -123.651644 Forested Surge Plain core site Preacher's Slough (corner) 46.950953 -123.691820 Forested Surge Plain secondary site Sand Island channel (North side) 46.957501 -124.052176 Aquatic Vegetation Bed Central estuary core site Sand Island East 46.949635 -123.731544 Aquatic Vegetation Bed Surge Plain core site Sand Island Flats (South side) 46.966641 -124.070969 Aquatic Vegetation Bed Central estuary core site Sculpin Cove 46.837546 -124.023062 High Emergent Marsh South Bay core site South Bay Bridge 46.862984 -124.065418 Aquatic Vegetation Bed South Bay secondary site Stearn's Bluff 46.924029 -123.978258 Aquatic Vegetation Bed Central estuary core site West Fork Hoqiuam River mouth 46.996364 -123.896837 Scrub/Shrub Cover Upper estuary core site West Fork Hoquiam River 47.015359 -123.910941 Forested Upper estuary core site Westport Marina (South Bay Channel) 46.899119 -124.087932 Eelgrass South Bay core site Wishkah River Channel 47.005423 -123.808460 Forested Upper Estuary core site Wishkah River slough 47.005423 -123.805846 Scrub/Shrub Cover Upper estuary fyke Withcomb Flats (North side) 46.917876 -124.065249 Sand Flat Central estuary secondary site Wynoochee River 46.969991 -123.613493 Scrub/Shrub Cover Surge Plain core site Wynoochee River Delta 46.962568 -123.607227 Forested Surge Plain core site

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 115

Appendix 2: Sites dropped from the sampling list due to logistical issues or denial of permission by landowners

Site Name latitude longitude habitat zone Chehalis River Blue Slough 46.948947 -123.724940 Forested Surge Plain Chehalis River near FL launch 46.947578 -123.650114 Forested Surge Plain Chehalis River surge plain #1 46.948947 -123.724940 Forested Surge Plain Chehalis River surge plain #2 46.950105 -123.723101 Forested Surge Plain East Fork Hoquiam slough 47.021878 -123.869697 Forested Upper estuary Elk River Bridge 46.862940 -124.065218 Aquatic Vegetation Bed South Bay Grass Creek 47.004231 -124.001816 High Emergent Marsh North Bay Humptulips River bar 47.041356 -124.050687 Forested North Bay John's River channel near slough 46.890539 -123.984360 High Emergent Marsh South Bay Point Brown marsh 46.939013 -124.152277 High Emergent Marsh Mouth Sand Island East Alt. (mud beach) 46.952415 -123.725693 Mud Flat Surge Plain Upper Elk Flats 46.846530 -124.033540 Forested South Bay Westport slough hwy fyke site (off hwy) 46.858896 -124.084592 High Emergent Marsh South Bay Wynoochee River mouth 46.961321 -123.608950 Forested Surge Plain

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 116

Appendix 3: Online resources for estimating the tidal heights during sampling in 2011

Site name Zone Tide predictor toWebsite use Half Moon Bay [aka Westport ocean side] Mouth Westport http://www.protides.com/washington/2964 Westport Marina (South Bay Channel) South Bay Westport http://www.protides.com/washington/2964 Elk River Flats (aka South Bay Forest) South Bay Bay City http://www.protides.com/washington/164/ Beardslee Slough South Bay Bay City http://www.protides.com/washington/164/ Beardslee Slough Mouth South Bay Bay City http://www.protides.com/washington/164/ Mallard Slough South Bay Bay City http://www.protides.com/washington/164/ John's River channel South Bay Markham http://www.protides.com/washington/1592/ John's River slough South Bay Markham http://www.protides.com/washington/1592/ Ocean Shores Flats 1 North Bay Westport http://www.protides.com/washington/2964 North Bay Flats 1 (aka Humptulips flats 1) North Bay Westport http://www.protides.com/washington/2964 North Bay Flats 2 (aka Humptulips flats 2) North Bay Westport http://www.protides.com/washington/2964 Campbell Slough North Bay Westport http://www.protides.com/washington/2964 Humptulips River mouth North Bay Westport http://www.protides.com/washington/2964 Jessie Slough North Bay Westport http://www.protides.com/washington/2964 Chinois Creek Flats Alternate North Bay Westport http://www.protides.com/washington/2964 Chinois Creek Flats Alternate 2 North Bay Westport http://www.protides.com/washington/2964 Chinois Creek Flats 1 North Bay Westport http://www.protides.com/washington/2964 Chinois Creek Flats 2 North Bay Westport http://www.protides.com/washington/2964 Sand Island channel (North side) Central estuary Westport http://www.protides.com/washington/2964 Sand Island Flats (South side) Central estuary Westport http://www.protides.com/washington/2964 Goose Island Flats (NW corner) Central estuary Westport http://www.protides.com/washington/2964 North Bay Bar (Goose Island Alt.) Central estuary Westport http://www.protides.com/washington/2964 Damon Point Central estuary Westport http://www.protides.com/washington/2964 Bar off John's River Central estuary Markham http://www.protides.com/washington/1592/ Withcomb Flats Central estuary Westport http://www.protides.com/washington/2964 Stearn's Bluff Central estuary Markham http://www.protides.com/washington/1592/ Moon Island Upper estuary Aberdeen http://www.protides.com/washington/11/ Cow Point Upper estuary Aberdeen http://www.protides.com/washington/11/ West Fork Hoquiam River Upper estuary Aberdeen http://www.protides.com/washington/11/ West Fork Hoquiam River Mouth Upper estuary Aberdeen http://www.protides.com/washington/11/ Hoquiam River Upper estuary Aberdeen http://www.protides.com/washington/11/ Hoquiam River slough Upper estuary Aberdeen http://www.protides.com/washington/11/ East Fork Hoquiam River Upper estuary Aberdeen http://www.protides.com/washington/11/ Wishkah River Channel Upper estuary Aberdeen http://www.protides.com/washington/11/ Wishkah River Slough Upper estuary Aberdeen http://www.protides.com/washington/11/ Lower Elliot Slough Surge Plain Cosmopolis http://www.protides.com/washington/616/ Preacher's Slough (corner) Surge Plain Cosmopolis http://www.protides.com/washington/616/ Sand Island East Surge Plain Cosmopolis http://www.protides.com/washington/616/ Preacher Slough Restoration Surge Plain Cosmopolis http://www.protides.com/washington/616/ Wynoochee River Surge Plain Cosmopolis http://www.protides.com/washington/616/ Chehalis near Friends landing Surge Plain Cosmopolis http://www.protides.com/washington/616/ Chehalis restoration site Surge Plain Cosmopolis http://www.protides.com/washington/616/ Chehalis River surge plain Surge Plain Cosmopolis http://www.protides.com/washington/616/

Grays Harbor Juvenile Fish Use Assessment, 2011 Wild Fish Conservancy Page 117