The California sea lion population, which occurs from the offshore islands of Mexico north to Vancouver Island, British Columbia, has increased dramatically in this century. In the late 1920s, 1,000-1,500 California sea lions were counted on land in California (Bonnot 1928, Cass 1985). Commercial harvest in the 1800s and early 1900s likely reduced the numbers of California sea lions at the turn of the century. With curtailment of commercial hunting in the early 1940s, the population gradually began to increase. Following passage of the MMPA in 1972, the California sea lion population off the West Coast of the United States has increased steadily at an average annual rate of more than 5% since the mid-1970s (Barlow et al. 1995) as indicated by pup counts (Fig. 5). Although the population is now very large and may be greater than any historical level (Low 1991), there is no evidence that it has reached its optimal sustainable population (OSP) level which is the management goal mandated by the MMPA (NMFS 1992).

In the United States, California sea lions breed primarily on the California Channel Islands of Santa Barbara, San Nicolas, San Miguel, and San Clemente. Large numbers of California sea lions also breed in Baja California, Mexico. California sea lions give birth to a single pup from late May through late June each year and breeding occurs in July. After the breeding season, most subadult and adult males (from both the Baja and the Channel Islands rookeries) migrate northward to central and northern California, Oregon, Washington, and British Columbia (Bartholomew 1967, Mate 1975). The peak of the northward migration occurs in September through October on the Oregon coast, in December in Washington, and in January and February in British Columbia. In the spring, most subadult and adult males migrate south, returning to the breeding rookeries in southern California and western Baja California. The southward migration peaks in Washington in March and April, and in Oregon in April and May (Brown 1988); most sea lions have left the Northwest by June (Gearin et al. 1986, 1988b). Some subadult males, adult males, and juveniles remain at haul-outs in central and northern California during the breeding season (Bartholomew 1967, Huber 1991,NMFS-SWFSC, unpubl. data.). Most of the year-round population in southern California consists of adult females, their pups, and juveniles.

Population abundance estimates are based on pup counts during the breeding season. To estimate total abundance, a correction factor (based on an estimate of the proportion of pups in the population) is applied to the number of pups counted in a given year (Boveng 1988, Lowry et al. 1992). From 1975 to 1994, pup counts of California sea lions in southern California (Fig. 5) increased at an annual rate of 5.2% (Lowry et al. 1992, Barlow et al. 1995, NMFS-SWFSC/AFSC unpubl. data). El Niño events have been shown to decrease the number of births and pup survival, especially the 1983 and 1992 events (DeLong et al. 1991, DeLong et al. 1993, NMFS-SWFSC/AFSC unpubl. data). The California sea lion population off the West Coast of the United States in 1994 was estimated at between 161,066 and 181,355 (Barlow et al. 1995).

In the last 15 years, counts of California sea lions at Everett, Washington (in Puget Sound) have increased from 108 in 1979 (Everitt et al. 1980) to 1,113 sea lions in 1995 (NMFS 1996a). They are present in Washington waters primarily during the nonbreeding season (September to May) and are concentrated in Puget Sound, particularly near Everett (Fig. 2). Counts of sea lions hauled-out in the Everett area are used as an index of the number of sea lions in the inland waters (Strait of Juan de Fuca/San Juan Islands, Hood Canal, Puget Sound). Preliminary analysis of mark-recapture data collected in 1995 indicates that counts at Everett may represent only 50-55% of the animals in the inland waters (NMFS-AFSC unpubl. data). Counts of sea lions in the inland waters of Washington averaged 300-500 animals from 1986 to 1994, and then peaked in 1995 at more than 1,100. Approximately 200-500 California sea lions have been observed during surveys in the 1990s on the offshore rocks and islands on the outer coast of Washington. The majority of these animals are found in the more northern portion of the coast. Sea lions are now also reported far upstream in rivers such as the Nisqually and Chehalis Rivers. Appendix D lists the population counts for regions in Washington.

Peak counts of California sea lions in Oregon have increased from 1,000-2,000 in the late 1970s to 5,000-7,000 in the early 1990s (ODFW unpubl. data). Counts made during the early 1990s for the Oregon regions are in Appendix D. California sea lions are found in greatest abundance on the south coast of Oregon at Rogue Reef and Orford Reef (800-1,000), at Cape Arago and Sea Lion Caves on the central coast (2,000-3,000), and at Cascade Head and the south jetty of the Columbia River on the north coast (2,000-3,000). From October to April, California sea lions are found in the Columbia River (300-500) from Astoria to the Bonneville Dam. They congregate in-river at Astoria at the east mooring basin and near fish processing plants (100-300), near the mouths of the Cowlitz and Lewis Rivers (50-100), and in the Multnomah Channel at the mouth of the Willamette River (10-50). In the Willamette River, small numbers (4-6) are found as far inland as Willamette Falls in Oregon City (see Fig. 3 for location). Increasing numbers of California sea lions are found in a number of coastal bays and rivers, including the Rogue River (10-20), Coos Bay (30-50), Yaquina Bay (50-100), and Tillamook Bay (6-12). Sea lions congregate at the mouths of many rivers primarily during salmonid runs, or during herring or smelt spawning returns.

In July and September of 1995, the California Department of Fish and Game (CDFG) conducted statewide aerial surveys to update counts of California sea lions on sites that were last surveyed on a statewide basis in 1980-82 by Bonnell et al. (1983). Summer counts in Bonnell et al. (1983) were 4,378 (1980) to 11,209 (1982), whereas counts in 1995 were 14,300, including 3,000 at the Farallon Islands (1994 estimates by W. Sydeman, Point Reyes Bird Observatory, 4990 Shoreline Highway, Stinson Beach, CA 94970. Pers. commun., July 1995). Fall counts in the early 1980s ranged from 10,334 to 24,348; in 1995, the total count was 16,900, including 1,000 at the Farallon Islands (Sydeman, pers. commun.).

Counts at Año Nuevo Island in 1992-94 ranged from 2,313 to 3,400 (mean 2,869) (Ono et al. 1993, NMFS-SWFSC unpubl. data) compared with 1,317 to 3,169 (mean 2,452) in 1980-82 (Bonnell et al. 1983). In the 1995 summer surveys, the largest concentration (6,745) of sea lions in central California was on Año Nuevo Island, comprising 60% of the total mainland count. Generally, the largest haul-out in central/northern California is on the Farallon Islands. Peak abundance usually occurs there during the spring or fall migration, but the highest numbers (6,000-7,000) have occurred during the summer breeding season in El Niño years (Huber 1991). Counts have also been made in the San Francisco Bay area since California sea lions first began hauling-out at Pier 39 in 1990, with peak counts occurring during winter. The highest number was 627 in February 1991 (K. Hanni, California Marine Mammal Center, Marin Headlands, Golden Gate National Recreation Area, Sausalito, CA 94965-2697. Pers. commun., August 1995).

Peak abundance of California sea lions occurs in southern California (south of Morro Bay) during the summer breeding season, when the majority of adults and a lesser proportion of subadults are present. In 1994, at the Channel Island breeding rookeries, the minimum count for the period of peak abundance was 81,300 sea lions (Barlow et al. 1995).

The Pacific harbor seal is distributed along the west coast of North America from Asuncion Island, off Baja California, northward to the Gulf of Alaska. Harbor seals are the most abundant pinniped in Washington and Oregon, and one of the most common pinnipeds in California. They are present year-round and pupping occurs in all three states. They occupy virtually all types of nearshore habitats (offshore rocks; sandy, gravelly, or rocky beaches; and estuarine mud flats) throughout the year and are found in most coastal bays and in many rivers. Populations of harbor seals have increased significantly since the MMPA was passed in 1972, but as with California sea lions, it has not been determined whether they have reached OSP (NMFS 1992).

The timing of harbor seal pupping occurs sequentially along the West Coast of the United States, with pups born earlier in the south and later in the north (Bigg 1973, Bigg and Fisher 1975). In Washington, there is additional variability in the timing of pupping, with considerably later (2 months) pupping in the inland waters (San Juans, Strait of Juan de Fuca, Eastern Bays, and Puget Sound) than on the coast and an extended pupping season from August to January in Hood Canal.

Females reach sexual maturity at 2-5 years of age, give birth for the first time at 3-6 years, and can live as long as 30 years (Bigg 1969). Most males reach sexual maturity at 3-5 years of age and few survive beyond age 20 (Bigg 1969). In many areas, pregnant females segregate into nursery aggregations, away from the main haul-out sites. Pups are weaned at about 4 weeks (Stein 1989), after which there is no further contact between the mother and pup. Once weaning is completed, nursery areas are abandoned and seals return to the main haul-out areas.

Harbor seals do not have extensive annual migrations. Based on studies of tagged seals, they usually remain within a 25-50 km area, although movements of up to 500 km have been recorded.

There are 319 harbor seal haul-out sites in Washington (Huber 1995). Numbers of harbor seals have increased by 7.7% annually between 1978 (when systematic counts began) and 1993 (Fig. 6). A correction factor was developed to account for harbor seals in the water during surveys in Washington and Oregon (Huber 1995). Using this correction factor (1.53) on the mean count of 22,310 seals (including pups) during the 1993 pupping season, the Washington total abundance in 1993 was estimated as 34,134 seals (Huber 1995). Recent counts (minimum population estimates) of harbor seals for each of the Washington regions are shown in Appendix D.

Some seals move seasonally from one area to another in response to locally abundant prey species such as eulachon in the Columbia River (Beach et al. 1985) or sockeye salmon in the Fraser River (Olesiuk 1993). Most information on harbor seal abundance in Washington is based on surveys conducted during the pupping season which occurs in May/June on the coast and July/August in the inland waters. The major exception is two studies on abundance and movements of harbor seals in the Columbia River and adjacent estuaries (Grays Harbor, Willapa Bay, Tillamook Bay) in 1980-82 and 1991-94 (Beach et al. 1985, Brown et al. 1995). Many of the seals which pup and breed in the coastal estuaries of Washington and Oregon in summer feed in the Columbia River in spring and fall (when salmonids are present) and in winter (when eulachon are abundant).

There are 101 harbor seal haul-out sites in Oregon (ODFW unpubl. data). As in Washington, counts of harbor seals have increased in Oregon since systematic counts were initiated in the 1970s, with an annual rate of increase of 7.4% from 1977 to 1993 (Fig. 7). In 1975, approximately 2,500 seals were counted statewide (ODFW unpubl. data); similar statewide counts of harbor seals in 1984-85 were 3,500-3,800 (Brown 1988). In 1993, the mean count of harbor seals in Oregon was 6,046 (ODFW unpubl. data). Most of the abundance information is from counts conducted during the pupping season in May/June, with the exception of the Columbia River as noted above. Using the correction factor (1.53) on initial counts, the Oregon total abundance estimate in 1993 was 9,251 harbor seals. Recent counts (minimum population estimates) of harbor seals for each of the Oregon regions are shown in Appendix D.

Since at least the 1960s, the abundance of harbor seals in California has been increasing and they have continued to occupy new haul-out sites. In 1982, there were 427 documented haul-out sites along the mainland coast. Currently, there are more than 860 documented haul-out sites in addition to haul-out sites on all eight of the Channel Islands in the southern California Bight.

The annual rate of increase in the harbor seal population in California was 5.6% from 1979 to 1995 (Fig. 8) based on counts by Hanan (1996). Systematic aerial photographic surveys have been conducted by CDFG specifically to count all harbor seals in California during their early summer molting period, which is considered to be the time of peak abundance onshore (Stewart and Yochem 1994). In 1995, the count of harbor seals in California was 23,336. Using a correction factor of 1.4 suggested by Boveng (1988) to account for seals in the water, the estimated abundance in California in 1995 was 32,699 harbor seals. Recent counts (minimum population estimates) of harbor seals for each of the California regions are shown in Appendix D.

This report reviews published information on California sea lion and harbor seal food habits, but because the data were collected for a variety of research programs that each address different research questions, much of the information is insufficient to address pinniped impacts on specific salmonid populations. There also are problems with applying available food habits information to the issue of impacts of pinniped predation on salmonids or other prey species in the ecosystems. There are caveats associated with all methods used to collect food habits data. These problems, described below, must be taken into account in any extrapolations or conclusions drawn from food habits data.

1. Food habits studies which involve the identification of fish by otoliths (fish ear bones) alone will underrepresent cartilaginous fish and fish with fragile or digestible otoliths. Harbor seals and other small pinnipeds may not always consume the heads (which contain otoliths) of larger fish such as salmonids (Pitcher 1980). For example, Riemer and Brown (1996) recently reanalyzed harbor seal food habits samples collected in the Columbia River in 1980-82 (Beach et al. 1985) using salmonid bones, gill rakers, and teeth, as well as otoliths for prey identification. Using these additional hard parts increased the occurrence of salmonids in all samples for both California sea lions and harbor seals (Appendices F and G). There are uncertainties concerning the appropriateness of relating the occurrence of hard parts from prey species to the actual occurrence of the prey species in the diet because it is not known if identifiable hard parts occur in the same proportion in food habits samples as they do in the actual diet.

2. Studies based on stomach contents of beachcast pinniped carcasses may not be representative of the food habits of healthy animals because beachcast animals are likely to have been sick or injured and may not have fed prior to stranding.

3. Food habits data based on pinnipeds taken incidentally in fisheries may be biased toward the diet of younger animals because they are often more susceptible to entanglement in fishing nets. Prey may be biased toward only those species which occur in or near the nets.

4. Studies which rely on observations of surface feeding by pinnipeds may overestimate the importance of larger prey because they are easier to observe. Smaller prey can be consumed under water. Other problems with interpretation include prey that are eaten too quickly to be identified or are ingested under water. Nonetheless, surface feeding observations are a good technique for quantifying pinniped predation on adult salmonids at sites where salmonid foraging occurs, such as river mouths.

5. Few studies involving shooting pinnipeds for food habits analyses have been conducted recently, so information from these studies is not current. Studies that have used this technique have generally provided direct and quantitative information on prey consumption by pinnipeds (Jobling and Breiby 1986, Jobling 1987).

Another source of information on pinniped feeding behavior is scars/wounds on salmonids that are attributable to predation attempts by pinnipeds. Salmonid scarring data, summarized below under "Salmonid Scarring by Pinnipeds," (see page 32) also have associated caveats that affect the interpretation of scarring data for food habits studies. The Working Group found that the quality of scarring data is inconsistent; differences exist between observers, sites, degree of interest, and diligence in reporting information. Estimates of scarring from fishway windows are biased downwards because only one side of the fish is observed. Scars are easier to observe on salmonids that have just returned to hatcheries. Because the skin darkens as salmonids get closer to spawning, making scars more difficult to detect, it is important to consider when scar data are collected relative to spawning time. In addition, there are non-pinniped sources of scars on salmonids which could be misinterpreted as pinniped marks; training to distinguish the two is not done at most locations nor done consistently.

California sea lions are opportunistic feeders, preying on a wide variety of fish and squid. Their diet is diverse, varying by location as well as seasonally and annually (Antonelis et al. 1984, Beach et al. 1985, Lowry and Folk 1987, Lowry et al. 1990, 1991, DeLong et al. 1993, Brown et al. 1995). Some of the more common prey within the breeding range in California are Pacific whiting, anchovy, market squid, and shortbelly rockfish (see Appendix H for genus/species of pinniped prey items) (Scheffer and Neff 1948, Fiscus and Baines 1966, Fiscus 1979, Antonelis et al. 1984). North of the breeding range, the diet shifts to those species which are locally and seasonally abundant. Sea lions move into specific areas during the nonbreeding season in response to local abundance of prey.

The results of food habits studies based on analysis of stomach contents, scat samples, and observational studies from Washington to California conducted since 1970 are summarized in Appendix F. The results of available information are discussed below starting from the north in Washington and moving south through California.

Washington

Important prey in Washington are Pacific whiting, herring, squid, spiny dogfish, gadids, and salmonids (Everitt et al. 1981, Gearin et al. 1986, Gearin et al. 1988b). Scat samples from California sea lions have been examined from two sites, Everett and Shilshole Bay, which are located about 20 miles apart in northern Puget Sound. In both areas, Pacific whiting and Pacific herring were the most frequent prey in the scats (Appendix F). Salmonids occurred in about 6% of the samples from Everett and in 25% of the samples from Shilshole Bay, at the entrance of the Lake Washington Ship Canal, and at the Ballard Locks (Gearin et al. 1988b). Sea lions congregate near Everett, apparently in response to a large spawning stock of Pacific whiting in Port Susan (Gearin et al. 1986). At the Ballard Locks, winter steelhead were the most frequently observed prey of sea lions based on surface feeding observations (NMFS 1995). Sea lions were also observed preying on downstream migrating juvenile salmonids (smolts) and on adult coho and sockeye salmon at the Ballard Locks (NMFS 1996a). Throughout Washington, California sea lions feed on steelhead and coho, sockeye, chum, and chinook salmon, both on free-swimming fish and on fish caught in gillnets and on hook-and-line gear (Gearin et al. 1986, Gearin et al. 1988a).

Columbia River

In the Columbia River, food habits studies utilizing beachcast sea lion carcasses (Beach et al. 1985), and studies on incidentally taken sea lions (Brown et al. 1995), indicate that the primary prey species are eulachon, salmonids, lamprey, herring, rockfish, and anchovy (Appendix F). About 13% of the beachcast samples contained salmonids, while 28% of the samples from incidental takes by the salmon gillnet fishery contained salmonids (Appendix F).

In the past decade, California sea lions have occurred seasonally with increasing frequency upriver in the lower Columbia River. Since 1990, California sea lions have been reported 128 miles from the river mouth near Willamette Falls on the Willamette River (a tributary of the Columbia). At least one to three sea lion males have been observed each spring near the fish ladders and smolt chute outfall in the area of Willamette Falls, consuming spring chinook salmon and winter and summer steelhead. In the spring of 1995, California sea lions were observed by ODFW and NMFS feeding on both adult salmonids and smolts. During limited observations (7 hours) over a 2-day period, one large California sea lion was observed to kill and consume an average of one salmonid per hour (ODFW unpubl. data). In 1996, at least five California sea lions were observed at Willamette Falls from April 2 through May 4. During limited observations in 1996 (155 hours), these sea lions were seen consuming 42 chinook salmon, 27 steelhead, and 20 unidentified salmonids (a total of 89 salmonids or about 0.6 salmonids/hour) (ODFW unpubl. data). ODFW and NMFS have implemented an observation/deterrence program at Willamette Falls to document the nature and extent of sea lion effects on salmonids at this site and to assess the effectiveness of various deterrence measures.

Oregon

In addition to the Columbia River area, food habits studies of California sea lions in Oregon have occurred at Cascade Head and the Rogue River (Appendix F). At Cascade Head, chub mackerel, Pacific herring, salmonids, unidentified skates, cephalopods, spiny dogfish shark, lamprey, Pacific whiting, sand lance, rockfish, and northern anchovy were the primary prey (Appendix F). Salmonid remains occurred in 29% of 82 sea lion scat samples taken in February 1994 and 8% of 32 samples collected in October 1994 (Riemer and Brown 1996).

California sea lion predation on salmonids in the lower Rogue River has been reported in several studies. Surface feeding observations have described lamprey (Jameson and Kenyon 1977) or salmonids (Roffe and Mate 1984, ODFW unpubl. data) as the most frequently identified prey. However, in 35 gastrointestinal tracts collected from California sea lions by Roffe and Mate (1984) in the same area, lamprey was the principal prey, occurring in 93% of the samples while steelhead occurred in 54% and chinook salmon in 11%. Spring chinook salmon return to the Rogue River primarily from late March to the end of May each year. In 1995, both Steller sea lions and California sea lions began appearing in increasing numbers in the mouth of the river in late April. From April to early June 1996, ODFW observed sea lions consume 9 chinook, 1 steelhead, 3 unidentified salmonids, and 3 Pacific lamprey in 55 hours of observation (ODFW unpubl. data).

California

California sea lion food habits have been studied at the Klamath River in northern California, at the Russian River, Pier 39 (San Francisco), the Farallon and Año Nuevo Islands in central California, and at the Channel Islands in southern California (Appendix F). At the Klamath River, Bowlby (1981) used direct observation of foraging behavior, examination of gastrointestinal tracts of beachcast sea lion carcasses, and scat analysis. Bowlby (1981) found lamprey was the main prey, with chinook salmon and steelhead occurring in 1-8% of samples depending on method of analysis (Appendix F). In spring 1978, Bowlby (1981) made 1,126 observations of sea lions foraging at the surface. Sea lions caught lamprey 96% of the time and salmonids less than 1% of the time. The major prey in gastrointestinal samples was lamprey. Scat analysis supported his conclusion that lamprey was the most frequent prey, followed by Pacific whiting and cephalopods (Appendix F).

Hanson (1993) studied the foraging ecology of California sea lions at the mouth of the Russian River in 1989-90. California sea lions were present during winter, spring, and early summer months, with a peak in abundance in the late winter and early spring. Pinniped abundance corresponded with peak salmonid returns. Hanson (1993) noted a 0.52 fish/hour catch rate for California sea lions when adult salmonids were in the river.

In central California, at the Farallon Islands, California sea lions consumed Pacific whiting and rockfish primarily (Appendix F). At Año Nuevo Island, California sea lions consumed Pacific whiting, shortbelly rockfish, and market squid (Appendix F) (NMFS-SWFSC/AFSC unpubl. data). No salmonid otoliths were found in the central California scat samples; however, other fish bones in the samples were not examined or identified. Scats collected at Pier 39 in San Francisco Bay indicate seasonal variability in the sea lion diet. Pacific herring was the most common prey during fall and winter, while spiny dogfish and a small number of salmonid smolts were also eaten (Appendix F) (California Marine Mammal Center, unpubl. data).

In southern California at the Channel Islands, based only on otolith and beak identification, the most commonly consumed prey were market squid, Pacific whiting, shortbelly rockfish, jack mackerel, chub mackerel, and northern anchovy (Appendix F) (NMFS-SWFSC/AFSC unpubl. data). Salmonid otoliths were found in only 2 out of the 9,513 scat samples examined during 12 years of sea lion food habits studies at the Channel Islands (Appendix F).

Harbor seals are opportunistic feeders, preying on a wide variety of cephalopods and benthic and epibenthic fish. Their diet varies as they take advantage of food that is seasonally and locally abundant. Food habits studies based on analysis of stomach contents, scat samples, or feeding observations in Washington, Oregon, and California since 1970 are summarized in Appendix G. These recent studies as well as some earlier studies are discussed below, starting from the north in Washington and moving south through California.

Washington

Harbor seal food habits in Washington were described in earlier years by Scheffer (1928), Scheffer and Sperry (1931), and Scheffer and Slipp (1944). Those studies indicated that Pacific whiting, tomcod, and walleye pollock were important prey, as well as flatfishes, Pacific herring, shiner perch, plainfin midshipman, and sculpins. More recent studies indicate that these species are still the most frequent prey of harbor seals in Puget Sound, the Strait of Juan de Fuca, and Hood Canal (Appendix G). At Gedney Island (near Everett) in northern Puget Sound, Pacific whiting dominate scat samples. In the coastal estuaries of Grays Harbor and Willapa Bay, northern anchovy, flatfish, crustaceans, smelt, and sculpin are the most frequent prey (Beach et al. 1985).

In most studies conducted before 1980, only otoliths were used to identify prey. Because harbor seals do not often eat the heads (where otoliths are found) of large fish such as salmonids, evidence of salmonid predation by harbor seals in the early studies was probably underestimated. However, even in early studies, predation on pink salmon in the fall, steelhead in the winter, and chinook salmon in the spring was reported in Puget Sound (Everitt et al. 1981, NMFS-AFSC unpubl. data).

In a reanalysis of samples collected in the period 1980-82, using bones and teeth as well as otoliths, the percentage of samples containing salmonid remains increased from 5% to 28% in Willapa Bay and from 4% to 10% in Grays Harbor (Riemer and Brown 1996). The data set containing the largest percentage of salmonid remains (50%) was from seals caught incidentally in salmon gillnet fishery in Grays Harbor (Appendix G). In addition to adult salmonid remains, smolt remains were also observed in scat samples from Grays Harbor (Brown et al. 1995) and Puget Sound (NMFS-AFSC unpubl. data).

Columbia River

Food habits of harbor seals in the Columbia River have been reported in a number of studies conducted since the early 1980s (Appendix G). A variety of prey species including anchovy, herring, lamprey, starry flounder, and other flatfish, sculpin, gadids, smelt, and salmonids are taken by seals in this area (Beach et al. 1985, Brown et al. 1989, Brown et al. 1995, Riemer and Brown 1996). The prey item most frequently taken by harbor seals in the Columbia River in winter is eulachon, and the highest counts of seals in the river coincide with the winter spawning of eulachon. Brown et al. (1989) examined gastrointestinal tracts of harbor seals killed incidentally in winter salmon gillnet fisheries from 1986 to 1988 and found eulachon in 100% of the samples and estimated that 97% of the prey eaten during the sampling period was eulachon. No salmon remains were identified in the Brown et al. (1989) samples even though most were collected during salmon fisheries. The seals clearly targeted on the much more abundant eulachon, but three other factors also may have influenced the probability of finding salmon remains in the gastrointestinal tracts of the seals sampled. First, fish species identifications were based only on otoliths and presence of salmonid flesh, and therefore other salmonid bones would have been missed. Second, the samples were collected during a gillnet fishery that occurs early in the spring chinook migration, before peak numbers of spring chinook occur inriver. Third, the samples were collected primarily from young seals that may be too small to successfully forage on the large chinook salmon migrating upriver. Nonetheless, when the eulachon run is over, the 2,000-3,000 harbor seals in the river will switch to other abundant prey species which then would include salmonids.

The occurrence of salmonids in harbor seal food habits studies in the Columbia River has varied from none to up to 60% occurence in some samples. Salmonids were identified in only 3 of 436 scat samples collected in the early 1980s by Beach et al. (1985): 1 with sockeye salmon remains (April 1982) and two with steelhead remains (April and August 1981). These scat samples were reanalyzed for salmonid remains using new identification techniques by Riemer and Brown (1996), and the occurrence of samples with salmonid remains increased from 3 to 28, with frequencies of occurrence in the total sample increasing from <1% to 6% (Appendix G). Beach et al. (1985) also found salmonid remains in 12% of 50 harbor seal gastrointestinal tracts collected from beachcast seal carcasses. These samples were collected throughout the year, but the largest numbers of samples were obtained from February through May. In 1991 to 1994, Brown et al. (1995) found salmonid remains in 13% of 61 harbor seal gastrointestinal tracts collected in the Columbia River. Salmonids were most frequently found in samples collected in the spring (33% of samples in April and May) and fall (60% of samples in September and October). Similarly, salmonids were identified in 19% of 67 scat samples collected in the spring (April 1995) and 39% of 36 scat samples collected in the fall (September and October) of 1994 (Riemer and Brown 1996). No salmonid remains were found by Riemer and Brown (1996) in 51 harbor seal scat samples collected in the winter (February and March) of 1992-93, similar to the lack of salmonid remains in samples collected in the winters of 1986-88 (Brown et al. 1989). Preliminary results of recent food habits studies show salmonid remains (primarily juveniles) in about 20% of 186 harbor seal scats collected in March through May of 1996 (NMFS-AFSC unpubl. data). In summary, salmonids appear to be targeted as prey by harbor seals primarily in the spring and fall, possibly because they are abundant and available in the river at that time in contrast to the winter when eulachon are much more abundant.

Oregon

As in other areas, harbor seals in Oregon feed on a wide variety of prey species (Appendix G), including anchovy, smelt, herring, flatfish, cottids, gadids, sculpins, rockfish, sand lance, salmonids, and cephalopods (Beach et al. 1985, Brown and Mate 1983, Harvey 1987, Graybill 1981, Roffe and Mate 1984, Brown et al. 1995, Riemer and Brown 1996). In these studies, salmonids were reported as occurring in 1-30% of samples (scats or gastrointestinal tracts) or direct surface-feeding observations.

Beach et al. (1985) examined 38 harbor seal scats collected in September and October in Tillamook Bay on the north Oregon coast. While other prey species were more common, one sample contained remains from a minimum of 19 small steelhead. Salmonid remains were identified in only 1 of 150 harbor seal scat samples collected in Netarts Bay, just south of Tillamook Bay, by Brown and Mate (1983). However, the majority of these samples were not collected during times when salmonids were known to occur there. Based on direct feeding observations, Brown and Mate (1983) estimated that harbor seals consumed 6.1%, 7.2%, and 1.5% of the 1978, 1979, and 1980 chum salmon runs, respectively, in Whiskey Creek (a tributary of Netarts Bay). The number of seals feeding in the area was similar each year, but the impact of their predation was greatest when the annual chum salmon return was low. For example, in 1979, seals took more than 7% of just over 550 returning salmon, while in 1980 the estimated consumption of nearly twice as many fish represented less than 2% of a return of more than 5,000 salmon.

In Coos Bay, Graybill (1981) found salmonid otoliths (steelhead and coho) in only 3 of 297 harbor seal scat samples (1%) collected primarily during May through September. Little evidence of smolts was noted even though two salmon-ranching (aquaculture) operations had released thousands of salmonid smolts into Coos Bay during the period of scat collection. It must be noted, however, that Graybill (1981) identified salmonids only from otoliths, and that the new prey identification techniques used by Riemer and Brown (1996) were not used to revise the occurrence of salmonids from these samples.

In the Rogue River, Roffe and Mate (1984) estimated from gastrointestinal tract examinations that steelhead comprised 5% and chinook salmon comprised 6% of the harbor seal diet in the late 1970s. However, based on numbers of seals present at that time, they concluded that salmonid consumption had a negligible impact on fish stocks in the Rogue River. More recently, ODFW found that Pacific lamprey, rex sole, rockfish, and Pacific tomcod occurred most frequently in scat samples (ODFW unpubl. data). However, salmonid remains (both adults and juveniles) were found in 13% of the scat samples. Harbor seal abundance peaked in March and April, but the occurrence of salmonids in scat samples was greatest in October (43%) and April to May (10-20%). During 55 hours of observations by ODFW from April to June 1996, harbor seals consumed 2 unidentified salmonids, 1 jack chinook, and 19 Pacific lamprey (ODFW unpubl. data).

California

Along the California coast, pinnipeds have been observed preying on adult salmonids and smolts at river mouths, estuaries, and the open ocean (Briggs and Davis 1972, Jones 1981, Herder 1983, Miller et al. 1983, Hart 1987, Hanson 1993, Stanley and Shaffer 1995, CDFG unpubl. data). In northern California, harbor seals have been observed consuming migrating salmonids (Stanley and Shaffer 1995, Hart 1987, Herder 1983, Miller et al. 1983).

Scats collected from the Klamath River contained remains of salmonids, lamprey, smelt, and flatfish (Herder 1983). In the lower Klamath River, substantial predation on released, tagged chinook salmon, coho salmon, and steelhead by harbor seals was observed each year between 1976 and 1990 in conjunction with seining operations conducted by CDFG to monitor upstream salmonid migrations during August, September, and early October. Because these salmonids had been held, handled, and tagged, they were likely more vulnerable to predation. In 1980-81 and 1984-88, studies were conducted to evaluate the feeding activity of harbor seals during these seining and tagging operations. The estimated percentage of seined/tagged fish taken by seals was relatively constant, ranging from about 3% to 8% (Hart 1987, Stanley and Shaffer 1995). Hart (1987) observed that a majority of the fish were consumed by as few as 12 seals.

Hanson (1993) observed the foraging behavior of harbor seals at the mouth of the Russian River in 1989-90. Harbor seals were present year-round; however, peak abundance corresponded with peak salmonid returns. Hanson (1993) also analyzed harbor seal scats and found that seals utilized a wide variety of prey, including flatfish, octopus, Pacific whiting, hagfish, and plainfin midshipman which were present during all seasons, while salmonids, smelt, lamprey, and rockfish occurred on a seasonal basis (Appendix G). Salmonid remains (primarily unidentified smolts) were most numerous during the winter (December to February) when they occurred in 20% of the scats. They were also found in 4% of the scats collected in the fall and 3% of the scats in the spring. Hanson (1993) observed harbor seals successfully capturing adult salmonids in shallow waters at catch rates of 0.27 fish/hour.

In central California, harbor seal scats contained shrimp, octopus, northern anchovy, plainfin midshipman, white croaker, and staghorn sculpin in San Francisco Bay (Appendix G). In the Monterey Bay area, based on otoliths and beaks only, harbor seals consume rockfish, octopus, spotted cusk-eel, white croaker, market squid, flatfish, staghorn sculpin, and plainfin midshipman (Appendix G). In Elkhorn Slough, Harvey et al. (1995) did not detect any evidence of salmonids in harbor seal scats even though steelhead may occasionally enter Elkhorn Slough.

In the Channel Islands off southern California, the most common prey were rockfish, octopus, spotted cusk-eel, and plainfin midshipman, based on otoliths and beaks only (Appendix G). On the southern California mainland, octopus and plainfin midshipman were the most commonly consumed food. Seals also consumed market squid, rockfish, flatfish, Pacific whiting, and spotted cusk-eel, based on otoliths and beaks (Appendix G).

Scars attributed to predation attempts by sea lions and harbor seals have been observed on adult salmonids at fish ladders and hatcheries, and in sport fish landings in Oregon and Washington since the early 1980s (Scordino 1993). These marks consist of descalings in the form of two overlapping arches caused by seal or sea lion canine teeth, two to three parallel scratches caused by seal claws, or puncture wounds caused by pinniped bites. The first quantification of scarring was at Bonneville Dam, where counters looked at one side of the salmonids as the fish moved rapidly past the fish-ladder windows. In 1980, fish counters observed 328,612 salmonids from four species in the fish ladder window and documented that at least 0.4% of the salmonids observed had evidence of pinniped-caused scars (Scordino 1993).

In 1990, an increased incidence of pinniped scarring (increasing from negligible numbers up to 19% of salmon examined) was noted by biologists handling Snake River spring chinook at Lower Granite Dam, raising concerns about impacts of pinnipeds on listed salmonids (Scordino 1993). From 1990 to 1993, steelhead and spring/summer chinook handled at the Lower Granite Dam were examined for evidence of attempted pinniped predation. Most predation scars are tooth marks and claw rakes that result in minor descaling of the fish and probably do not affect survival. However, some of the more severe bites and puncture wounds on fish may result in mortality. Scars were found on 7.8% of the steelhead and on 16.4% of the spring/summer chinook; this included the more severe puncture wounds on 2.1% of the steelhead and 5.7% of the chinook (Harmon et al. 1994). In 1994, Huber et al. (1995) handled steelhead and spring chinook at the Bonneville Dam and several hatcheries on the Columbia and Snake Rivers. They found 24% of the steelhead with pinniped marks (including 10% with bite wounds) and 16% of the spring chinook with pinniped marks (including 4% with bite wounds).

In more recent studies on pinniped scarring on spring/summer chinook salmon and steelhead in the Columbia and Snake Rivers, pinniped canine marks and claw rakes were found on 12% of the spring/summer chinook salmon (Huber et al. 1995a, Harmon et al. 1994). Many of the tooth marks and claw rakes examined consisted of minor descaling of the fish that probably did not affect their survival; however, up to 6% of the spring chinook salmon had more severe bites and marks that could have resulted in mortality (Harmon et al. 1994, Huber et al. 1995a). Harmon et al. (1994) hypothesized that pinniped-induced stress from scarring and injuries may also result in lowered spawning success and that considerable salmonid mortality, both direct and indirect, may result from confrontations with, and injuries from, pinnipeds.

Huber et al. (1995a) determined that most of the predation marks on spring/summer chinook salmon in the Columbia River were caused by harbor seals, but 10% were clearly inflicted by California sea lions. It is not clear if the preponderance of marks attributable to harbor seals is the consequence of more harbor seals feeding on the salmonids, or of harbor seals being less successful than sea lions at capturing and consuming chinook salmon. More data are needed on harbor seal predation on free-swimming adult salmonids before accurate estimates of losses can be generated.

In Oregon, there are few river systems where pinniped scarring has not been observed or reported. At most of these locations, seals and/or sea lions have been observed foraging in stretches of the river in the few miles just below hatcheries (e.g., Nehalem River, Tillamook Bay, Salmon River), and in some cases directly at the base of fish ladders, weirs, or fishways (e.g., Willamette River, Netarts Bay, Yaquina Bay, Coos Bay). In general, ODFW has noted the highest observed scarring rates occur on winter steelhead (6-53%), followed by coho (11-20%). Chinook salmon have the lowest frequency of predator scars (8-14%). Scarring rates for winter steelhead observed at various hatcheries in Oregon are shown in Table 1. Fewer data exist for scarring rates on coho and chinook, but frequencies at Alsea River and Salmon River have been reported from 8% to 14% of returning adult fish.

In California, Miller et al. (1983) reported that at an ocean-farming operation at Davenport, about 15% of the adult salmon and steelhead that entered the return structure had been bitten by harbor seals. During two spawning seasons from 1994 to 1996 in Scott Creek, pinniped tooth and claw marks have been found on 28-40% of the captured coho salmon and 31-50% of the captured winter steelhead (Monterey Bay Salmon and Trout Project, unpubl. data; D. Streig and M. McCaslin, Monterey Bay Salmon and Trout Project, 825 Big Creek Road, Davenport, CA 95017. Pers. commun., February 1997). On the San Lorenzo River, 15% of the returning winter steelhead had pinniped bite and claw marks in the 1991-92 season. The proportion of pinniped-scarred steelhead increased to 47% and 54% in the 1994 and 1995 seasons (Monterey Bay Salmon and Trout Project, unpubl. data; Streig and McCaslin, pers. commun., February 1997). The higher rate of pinniped scarring in steelhead may be a consequence of steelhead traveling close to the shoreline in the upper 2 meters of the water column, where they are more vulnerable to capture by pinnipeds than coho which occur in more open areas. In Scott Creek, coho salmon may be more susceptible to predation by harbor seals than other areas because they linger in the lagoons before they make their upstream migration to spawn (Streig, pers. commun., September 1995). Precocious male steelhead (those that return to fresh water to spawn after less than 1 year in the ocean) returning to Scott Creek do not exhibit any tooth or claw marks. These males are smaller (less than 22 inches) than adult steelhead and thus, harbor seals may be more efficient in capturing and killing (rather than scarring) these smaller steelhead (Streig, pers. commun., September 1995).