PACIFIC HAKE

General Biology

Geographical distribution

Pacific hake, Merluccius productus (Ayres, 1855), of the offshore stock range from Sanak Island in the western Gulf of Alaska to Magdalena Bay, Baja California Sur. They are most abundant in the California Current System (Bailey 1982, Hart 1973, Love 1991, NOAA 1990). There are three much smaller stocks with much smaller ranges: a Puget Sound stock, a Strait of Georgia stock, and a dwarf stock limited to waters off Baja California (Bailey et al. 1982, Stauffer 1985). The offshore stock of Pacific hake is migratory and inhabits the continental slope and shelf within the California current system from Baja California to British Columbia (Quirollo 1992). All life stages are found in euhaline waters at 9-15^oC (NOAA 1990).

Eggs and larvae of the offshore stock are pelagic in 40-140 m of water (Smith 1995), with eggs in the earlier stages being at the deeper depths (Moser et al. 1997). Pacific hake larvae tend to aggregate near the base of the thermocline or mixed layer (Stauffer 1985). This association with the thermocline or mixed layer may partially explain why Pacific hake in the Strait of Georgia and Puget Sound spawn near major sources of freshwater which would cause a stratified layer of low-salinity water on top of the well mixed marine waters common during the winter. Juveniles reside in shallow coastal waters, bays, and estuaries (Bailey 1981, Bailey et al. 1982, Dark 1975, Dark and Wilkins 1994, Dorn 1995, NOAA 1990, Sakuma and Ralston 1995, Smith 1995), and move to deeper water as they get older (NOAA 1990). Pacific hake school at depth during the day, then move to the surface and disband at night for feeding (McFarlane and Beamish 1986, Sumida and Moser 1980, Tanasich et al. 1991).

Adults are epi-mesopelagic (Bailey et al. 1982, NOAA 1990, Sumida and Moser 1980). Highest densities of Pacific hake are usually found between 50 and 500 m, but adults occur as deep as 920 m and as far offshore as 400 km (Bailey 1982, Bailey et al. 1982, Dark and Wilkins 1994, Dorn 1995, Hart 1973, NOAA 1990, Stauffer 1985). Spawning is greatest at depths between 130 and 500 m (Bailey et al. 1982, NOAA 1990, Smith 1995).

Smith (1995) recognized three habitats utilized by the offshore stock of Pacific hake: 1) a narrow 30,000 km² feeding habitat near the shelf break of British Columbia, Washington, Oregon, and California, populated 6-8 months per year, 2) a broad 300,000 km² open-sea area of California and Baja California populated by spawning adults in the winter and embryos and larvae for 4-6 months, and 3) a continental shelf juvenile rearing area of unknown size off California and Baja California.

Migrations

Offshore stocks spawn off Baja California in the winter, then mature adults begin moving northward and inshore, following the food supply and Davidson currents (Fig. 17) (NOAA 1990). Pacific hake reach as far north as southern British Columbia by fall. By early late fall, they begin the southern migration to southern spawning grounds and further offshore (Bailey et al. 1982, Dorn 1995, Smith 1995, Stauffer 1985) (see Fig. 17).

Stocks in the Strait of Georgia and Puget Sound undergo similar migration patterns, but on a greatly reduced scale (McFarlane and Beamish 1986, Shaw et al. 1990). In both areas, spawning occurs in locations proximate to major sources of freshwater inflow: near the Frazer River in the Strait of Georgia, and near the Skagit and Snohomish Rivers in Port Susan (McFarlane and Beamish 1985, Pedersen 1985). The Puget Sound and Strait of Georgia stocks spend their entire lives in these estuaries (McFarlane and Beamish 1986, Shaw et al. 1990).

Reproduction and development

Pacific hake may spawn more than once per season, so absolute fecundity is difficult to determine. Pacific hake are oviparous with external fertilization. Offshore stocks have 180-232 eggs/g body weight, but Puget Sound and Strait of Georgia stocks have only 50-165 eggs/g body weight (Mason 1986). Bailey (1982) estimated that a 28-cm female had 39,000 eggs, while a 60-cm female had 496,000 eggs.

Eggs are spherical, 1.14 to 1.26 mm in diameter with a single oil droplet, and are neritic and float to neutral buoyancy (Bailey 1981, Bailey et al. 1982, NOAA 1990). The pelagic eggs of Pacific hake off California are found at depths between 50 and 75 m over a bottom depth of at least 300 m (Moser et al. 1997). Pelagic eggs of Puget Sound Pacific hake are found at approximately the same depth, but Pacific hake eggs in Puget Sound are in the bottom 25 m of the water column over a bottom depth of about 110 m (Bailey 1982, Moser et al. 1997).

Embryonic development is indirect and external (NOAA 1990). Hatching occurs in 5-6 days at 9-10^oC and 4-5 days at 11-13^oC (Bailey 1982, Hollowed 1992). Larvae hatch at 2-3 mm total length (Stauffer 1985, Sumida and Moser 1984) with a yolk sac that is gone in 5-7 days (Bailey 1982). Larvae metamorphose into juveniles at 35 mm, typically in 3-4 months (Hollowed 1992). Juveniles range from 35 mm to 40 cm depending on sex (Bailey et al. 1982, Beamish and McFarlane 1986, Hollowed 1992).

In Puget Sound and the Strait of Georgia, female Pacific hake mature at 37 cm and 4-5 years of age (McFarlane and Beamish 1986). Females of the offshore stock mature at 3-4 years and 34-40 cm, and nearly all males are mature by age 3 and as small as 28 cm. Females grow more rapidly than males after 4 years; growth ceases for both sexes at 10-13 years (Bailey et al. 1982).

By age 3, most Pacific hake become available to the mid-water trawl fishery, although Pacific hake between ages 6 and 11 are most commonly caught. The maximum age of Pacific hake is about 20 years, but Pacific hake over age 12 are rare (Methot and Dorn 1995). The size-at-age of offshore Pacific hake has been declining since the 1960s (Methot and Dorn 1995). By the early 1990s, age-10 males were 47 cm, and age-10 females were 48 cm. McFarlane and Beamish (1985) reported a more rapid growth rate in Pacific hake from the Strait of Georgia compared to Pacific hake from offshore up to age 4, after which time their growth rate levels off. Moreover, the Strait of Georgia Pacific hake reach maximum mean lengths (approximately 44 cm) that are approximately 10 cm shorter than the length at maximum age for offshore Pacific hake. In Puget Sound, male Pacific hake rarely exceed a length of 40 cm, whereas females tend to be about 4 cm longer than males (Pedersen 1985). MacGregor (1971) noted a marked cline in size at maturity with latitude for Pacific hake. According to MacGregor (1971) Pacific hake grow to a larger size and mature at a larger size in the northern part of their range, when comparing Pacific hake from southern Baja California to Puget Sound. MacGregor (1971) noted that this same growth pattern is apparent in European hake (M. merluccius) with larger hake occurring in the north and smaller hake in the south.

Trophic interactions

Pacific hake larvae eat calanoid copepod eggs, nauplii, and adults (McFarlane and Beamish 1986, Sumida and Moser 1984). Juveniles and small adults feed chiefly on euphausiids (NOAA 1990). Large adults also eat amphipods, squid, Pacific herring, smelt, crabs, shrimp, and sometimes juvenile Pacific hake (Bailey 1981, Dark and Wilkins 1994, McFarlane and Beamish 1986, NOAA 1990).

Eggs and larvae of Pacific hake are eaten by walleye pollock, herring, invertebrates, and sometimes Pacific hake. Juveniles are eaten by lingcod, Pacific cod, and rockfish species. Adults are preyed on by sablefish, albacore, walleye pollock, Pacific cod, soupfin sharks, and spiny dogfish (Fiscus 1979, McFarlane and Beamish 1986, NOAA 1990). Another important group on predators of adult Pacific hake are marine mammals, including the northern elephant seal (Mirounga angustirostris), northern fur seal (Callorhinus ursinus), California sea lion (Zalophus californianus), and several species of dolphins and whales (Methot and Dorn 1995).

Size and age distributions

As was mentioned above in the "Reproduction and development section," Pacific hake in the Strait of Georgia tend to be shorter at age than Pacific hake in the offshore populations; in some cases up to 10 cm (McFarlane and Beamish 1985). In addition, Pacific hake from central Puget Sound appear to be 2 to 4 cm shorter at age than Pacific hake from the Strait of Georgia. Quinnell and Schmitt (1991) presented length/frequency data for Pacific hake from Puget Sound (Fig. 18) which demonstrated a trimodal length distribution, with most Pacific hake being 33 to 50 cm, and approximately similar numbers of Pacific hake being either 22 to 28 cm or 9 to 14 cm (see Table 2).

Phenetic and Genetic Information Relating to the Species Question

Phenetic and genetic information examined for evidence of DPS delineations of Pacific hake included presence of geographically-discrete and temporally-persistent spawning aggregations, and variation in seasonal migration patterns, year-class strength, parasite incidence, growth rate, size- and age-at-maturity, length frequency, fecundity, meristics and morphometrics, and genetic population structure.

Life History Information

In addition to the abundant migratory population of Pacific hake, that spawns offshore from Cape Mendocino, California to southern Baja California, several other stocks of Pacific hake have been identified including at least two that spawn in Puget Sound, several in the Strait of Georgia, several in the west coast inlets of Vancouver Island, and a small-bodied ("dwarf hake") off the west coast of southern Baja California (Nelson 1969, Bailey et al. 1982, Ermakov 1982, Bailey and Yen 1983, Beamish and McFarlane 1985, Pedersen 1985, Bollens et al. 1992a, Alados et al. 1993, Methot and Dorn 1995, Fox 1997).

The Pacific hake stocks from offshore (Baja California to the west coast of Vancouver Island), Strait of Georgia, and Puget Sound have been considered discrete from one another on the basis of differences either in: 1) allozyme frequencies (Utter 1969a, b; Utter and Hodgins 1969, 1971; Utter et al. 1970), 2) spawning locality (Alverson and Larkins 1969), 3) size- and age-at-maturity (Goñi 1988), 4) growth (Nelson 1969, Beamish et al. 1982, McFarlane and Beamish 1985), 5) year-class strength (McFarlane and Beamish 1985, Goñi 1988), 6) effective fecundity (McFarlane and Saunders 1997), 7) otolith morphology and annuli formation (McFarlane and Beamish 1985), or 8) the degree of infestation with the protozoan parasite Kudoa paniformis Kabata and Whitaker, 1981 (Kabata and Whitaker 1981, 1985; McFarlane and Beamish 1985).

Pre-historical and historical persistence in Puget Sound

Tunnicliffe et al. (in press) examined fish remains in a complete Holocene sediment core sequence from Saanich Inlet, Vancouver Island, British Columbia. Pacific hake were one of the first fish species to occur in Saanich Inlet following glacial retreat from the region, after approximately 12,000 years before present (BP) (Tunnicliffe et al. in press). Fish abundance and species diversity peaked in Saanich Inlet between 7,500 and 6,000 BP, and the last 1,000 years have seen some of the lowest abundances of fishes in Saanich Inlet's marine history (Tunnicliffe et al. in press). The close proximity of Saanich Inlet to Puget Sound would suggest that Pacific hake were also likely established in Puget Sound by about 12,000 BP.

Pacific hake were identified in prehistoric fish skeletal remains from the Duwamish No. 1 archeological site (45-KI-23), located 3.8 km upstream from Elliott Bay on the Duwamish River, utilized by aboriginal humans between A.D. 15 and A.D. 1654 (Butler 1987). Gadiforms were present throughout the occupational history of this site, and were third and fourth in rank order of taxonomic abundance in two separate studies of fish bones performed at this site (following Salmonidae, Pleuronectiformes, and in one case Squalidae) (Butler 1987). Conversely, archaeological investigations of the West Point site on the north side of Discovery Park in Seattle (utilized by hunter-fisher-gatherers between 4,250 and 200 BP) found few remains of gadiforms, although some Pacific cod bones were identified at this site (Wigen 1995). Wigen (1995) postulated that differences in the frequency of gadiform remains found between the Duwamish and West Point sites may be related to the possible use of fish traps at West Point versus hook and line methods at the Duwamish site, or perhaps to differences in the season of human occupation between the two sites. In historic times, Pacific hake were reported as abundant in Puget Sound by Jordan and Starks (1895).

Spawning location and spawn timing

Within Puget Sound (including Hood Canal) Pacific hake are known to spawn in Port Susan (Nelson 1969, Pedersen 1985, WDFHMD 1992) and in Dabob Bay (Bailey and Yen 1983, Bollens et al. 1992a, Fox 1997) and there may be other spawning aggregations of Pacific hake in Puget Sound (Fig. 19) but only the Port Susan-Saratoga Passage population has been commercially exploited (Thorne et al. 1971, Kimura and Millikan 1977, Pedersen 1985). Smith (1936) stated that spawning Pacific hake of both sexes were taken in Hale Passage near Carr Inlet in southern Puget Sound in March of 1936. WDFHMD (1992) also lists Carr Inlet as a known Pacific hake spawning location. According to Nelson (1969) large numbers of Pacific hake eggs and larvae have been found in Puget Sound only at Port Susan, with small numbers of eggs and larvae occurring in southern Puget Sound, Hood Canal, and near Possession Sound. Miller and Borton (1980) summarized distribution records of Pacific hake in Puget Sound as found in published records, museum collections, and various boat logs. Centers of collection of Pacific hake in Puget Sound were heavily influenced by fishing effort and ease of access, and centered around Port Susan, Saratoga Passage, Possession Sound, the central Sound from Shilshole Bay to Port Madison, Port Orchard, Carr Inlet, Penn Cove and Holmes Harbor on Whidbey Island, and Dabob Bay in Hood Canal (Miller and Borton 1980). Pedersen (1985) stated that small groups of Pacific hake occur in other areas of Puget Sound, in addition to Port Susan, but he did not identify the areas specifically. Historically, commercial fisheries for Pacific hake in Puget Sound centered around the Port Susan, Saratoga Passage, Port Gardner, and southern Carr Inlet areas (Fig. 20, Pedersen and DiDonato 1982).

Table A-1 summarizes available data on spawn timing in various locations for Pacific hake. In Puget Sound, spawning occurs primarily during February through April, peaking in March (W. Palsson^[2]). Spawning aggregations begin to form up to a month before actual spawning. Within Puget Sound, peak spawning of Pacific hake occurs in mid-late-March in the Central Puget Sound population in Port Susan (Goñi 1988). Spawn timing of the Dabob Bay stock ranged from the beginning of February to the end of April in 1990 and from mid-January to the beginning of April in 1991 (Fox 1997). The mean back-calculated spawn date for Pacific hake in Dabob Bay was 14 March, in 1990, and February 20, in 1991 (Table A-1) (Fox 1997).

The main Pacific hake stock in the Strait of Georgia aggregates to spawn in the deep basins of the south-central Strait of Georgia (Fig. 19), with peak spawning occurring from March to May (Table A-1) (Goñi 1988, Shaw et al. 1990, Kieser et al. 1999). This area is bound by Halibut Bank and Gabriola Island, to the east and west, and Texada Island and Galiano Island to the north and south. Spawning aggregations of Pacific hake in south-central Strait of Georgia occur in two depth strata between 50-120 m and 150-330 m (Shaw et al. 1990). Beamish et al. (1976b) and McFarlane and Beamish (1985) stated that there is a second discrete stock of Pacific hake in the Strait of Georgia that has been found spawning northwest of Texada Island near Montgomery Bank (Fig. 19).

Foucher and Beamish (1980) reported that a third small stock of large Pacific hake has been observed spawning, 4-6 months prior to the main Strait of Georgia stock, in the Gulf Islands near Yellow Point in Stuart Channel (Fig. 19, Table A-1), suggesting this group is an additional discrete spawning stock (McFarlane and Beamish 1985). Likewise, Beamish et al. (1976a, c, 1978a) speculated that a stock of large Pacific hake may occur in Stuart Channel in the Gulf Islands that mature and spawn earlier than do Pacific hake in the open Strait of Georgia (Shaw et al. 1985a). Beamish et al. (1976c) stated that a small percentage of the presumed Stuart Channel stock appear to be in spawning condition year-round. An additional stock of Pacific hake was suggested to occur in Saanich Inlet by Beamish et al. (1978b) based on apparent different rates of growth and presence of larger than normal Pacific hake in this area.

Palsson et al. (1997) stated that the South Puget Sound Pacific hake, which spawn in the Port Susan area are distinct from the offshore migratory stock and probably distinct from the resident transboundary stock shared with Canada that spawns in the Strait of Georgia. This resident transboundary population is also considered distinct from the offshore migratory stock (Palsson et al. 1997). Although spawning of the stocks occurs in well separated areas, it is not clear to what degree precise homing to the spawning grounds occurs in the Strait of Georgia and Puget Sound (Goñi 1988). Alverson (1969) stated that the migration pattern and distribution of eggs and larvae indicate that the offshore migratory Pacific hake population is homogeneous. Alverson (1969) also stated that the evidence is good that Pacific hake in inshore waters of Puget Sound, and perhaps the Strait of Georgia, are distinct from the offshore migratory population.

Various Canadian publications provide evidence that two types of Pacific hake occur off the southwest coast of Vancouver Island. These two types consist of: 1) small numbers of resident Pacific hake that remain in the region year round, spending the summer in coastal inlets along the west coast of Vancouver Island, and 2) the much larger stock of migratory offshore Pacific hake that spawn off southern California and migrate north to feed in the spring and summer (Beamish and McFarlane 1985, Shaw et al. 1985b, Ware and McFarlane 1995). Separate resident stocks of Pacific hake apparently occur in Nootka Sound, Barkley Sound (Trevor Channel), Sydney Inlet, and Tahsis Inlet on Vancouver Island (Beamish and McFarlane 1985; Shaw et al. 1985b, 1989a, b; Ware and McFarlane 1995). Shaw et al. (1985b) stated that "it appears that each inlet contains a "resident" stock of hake which may have different spawning times assuming similar growth rates." Beamish and McFarlane (1985) cited unpublished data indicating that eggs and larvae of Pacific hake have been found in samples from January to April in the vicinity of Barkley Sound and Sydney Inlet "clearly indicating the presence of resident spawning stocks." Beamish (1981a) and Beamish and McFarlane (1985) also stated that since few Pacific hake have been observed in this region in winter, the putative resident stocks of Pacific hake off the west coast of Vancouver Island are likely small in size. Smith et al. (1990) speculated that as resident west coast Vancouver Island inlet Pacific hake mature, they may eventually mix with the offshore migratory population during summer months off southwest Vancouver Island. In addition, McFarlane and Beamish (1985) reported that small distinct local stocks of Pacific hake are suspected to occur in mainland inlets of the British Columbia coast north of the Strait of Georgia.

The offshore stock spawns off southern California, primarily from December to April, with peak activity occurring in January and February (Bailey 1981, Smith 1995)-although sometimes heavy spawning occurs in March (Fig. 17, Table A-1) (Bailey et al. 1982). Woodbury et al. (1995) provided evidence, based on back-calculated spawn dates of young-of-the-year Pacific hake collected in central California, that spawning occurred in some years from September to March but that the majority of survivors were spawned in January-February. Hirschberger and Smith (1983) reported on an anomalous group of over 180 Pacific hake collected in spawning condition in August 1980 along the coast of Oregon; a time of year and region where spawning Pacific hake had not been previously, or subsequently, reported.

Doyle (1992) and Hollowed (1992) reported the presence of Pacific hake eggs and larvae in ichthyoplankton samples collected offshore of Northern California, Oregon, and Washington in the spring of 1983 and 1984, but not in the spring of 1980, 1981, 1982, or 1985. Hollowed (1992) speculated that the 1983-84 El Niño may have caused a shift in Pacific hake spawner distribution to the north in the winter of 1983 and 1984, accounting for the finding of most eggs in those years between 40º and 44º N.

A stock of Pacific hake off the west coast of southern Baja California was identified as distinct from the main offshore stock by Vrooman and Paloma (1976) based on morphometry, meristics, and general protein electrophoresis. Vrooman and Paloma (1976) called this population "dwarf hake" and suggested that it does not interbreed with M. productus and may therefore be a separate species. Ermakov (1982) also differentiated between an "oceanic" and a "dwarf" Pacific hake off southern California and Baja California based on morphometrics and disjunct spawning localities. Bailey et al. (1982) regarded the separation of the dwarf and offshore stocks to be controversial and suggested the differences between the two units may not be genetic, but "are not inconsistent with changes caused by environmental effects in the different habitats." Mathews (1985) described the "dwarf hake" of Vrooman and Paloma (1976) off Baja California as M. hernandezi; however, the taxonomic status of this species is still uncertain (Cohen et al. 1990).

Tagging and distribution

In general, species in the Genus Merluccius do not survive capture and release well and therefore no tagging studies exist to infer patterns of migration (Fritz 1959). This generality also holds for Pacific hake, which are difficult to tag externally due to their fragility (MacLellan and Saunders 1995). Despite the lack of tagging data, Mason et al. (1984) and Mason (1986) thought it unlikely that offshore and Strait of Georgia Pacific hake stocks intermingle to any large degree, based on their distributional patterns; although, according to Mason (1986), there may be some interchange between the Strait of Georgia and Puget Sound stocks due to surface transport of larvae produced in the central Strait of Georgia. However, WDFW (2000) pointed out that since water leaves the Strait of Georgia primarily "through and west of the San Juans into the northern Strait of Juan de Fuca," direct exchange of larvae between the Strait of Georgia and Puget Sound would not be expected.

Seasonal migrations

In autumn, the offshore stock of Pacific hake migrate from summertime feeding grounds (located between Queen Charlotte Sound in British Columbia and central California) to winter spawning areas (located between Cape Mendocino on the California coast and northern Baja California) (see Fig. 17). Spawning occurs from 60-1,655 km offshore at depths of from 120-400 m over bottom depths exceeding 1,000 m (Saunders and McFarlane 1997). Some Pacific hake may spawn as far south as off the southern tip of Baja California (Bailey 1982). The distribution of eggs and larvae and the migration pattern suggests that there is a single large offshore Pacific hake stock (Alverson and Larkins 1969). Adults migrate northward in the spring while juveniles remain off central and northern California (Bailey et al. 1982). The extent of northward migration is age-dependent, with older and larger fish migrating furthest north (Richards and Saunders 1990, Dark and Wilkins 1994, Saunders and McFarlane 1997). In warm years a greater proportion of the offshore Pacific hake stock moves into the Canadian fishery zone (Richards and Saunders 1990) and spawner distribution may shift further north as well (Hollowed 1992, Saunders and McFarlane 1997). Saunders and McFarlane (1997) summarized observations of latitudinal trends in biological characteristics such as age composition, sex ratio, mean size, and parasite prevalence for both summer-feeding and winter-spawning aggregations of offshore Pacific hake and propose processes that may explain these patterns.

Inshore Pacific hake that spawn in the Strait of Georgia, in Puget Sound at Port Susan and Dabob Bay, and in Nootka Sound, Barkley Sound, and Sydney Inlet on Vancouver Island are essentially resident stocks, although they may have relatively short spawning migrations (Ware and McFarlane 1995).

Year class strength

Strong year classes in offshore Pacific hake are not synchronous with those in Strait of Georgia Pacific hake (Beamish 1981a, McFarlane and Beamish 1985). According to Beamish et al. (1982), the dominant age-groups of Pacific hake in the Strait of Juan de Fuca and off the west coast of Vancouver Island were identical, and differed from Pacific hake in the Strait of Georgia. Analysis of age composition suggests that the differences between offshore Pacific hake and the inshore populations probably would be better characterized as differences in year-class variability rather than in year-class syncronicity (M. Dorn^[3]).

Goñi (1988) found "strong inequalities" between indices of year-class strength (YCI, calculated by adding up percent contributions of each particular year class at ages 4, 5, and 6) for Port Susan (Puget Sound) and Strait of Georgia Pacific hake. Although discrepancies between ageing methods employed for these two populations may have confounded correlations between year classes in this study, Goñi (1988) stated that the differing relative importance and lack of correlation between strong year-class abundances in Port Susan and Strait of Georgia Pacific hake could be interpreted as evidence of their physical isolation. However, the fact that ageing procedures for Pacific hake differed by agency for these two groups of fish (Goñi 1988) and that the YCI used by Goñi (1988) was sensitive to the exploitation level, suggests that apparent differences in the YCI can't be used as reliable evidence of stock separation. At the time that Goñi (1988) did her study, the exploitation level for Puget Sound Pacific hake was high, whereas Strait of Georgia Pacific hake had a low exploitation rate. A high exploitation rate would accentuate the variability in the YCI even with the same variability in year class strength. The observation that recruitment (as evident by strong year classes) is more variable in Puget Sound Pacific hake relative to Strait of Georgia Pacific hake isn't supportable (M. Dorn^[4]).

Parasite incidence

The softness and rapid deterioration of Pacific hake flesh following capture is generally considered to be due to two species of Kudoa, a genus of myxosporean protozoan parasites that infect the Pacific hake muscle fibers (Kabata and Whitaker 1981, 1985, 1986). The myxosporean parasite Kudoa paniformis was absent from Strait of Georgia Pacific hake (Kabata and Whitaker 1981, McFarlane and Beamish 1985) but was found in 57% of the large offshore migratory Pacific hake population (Kabata and Whitaker 1985). Kudoa paniformis was also absent in the putative resident Pacific hake stock in Tahsis Inlet on Vancouver Island, and was found in only one fish (11%) from the putative resident Pacific hake stock in Barkley Sound (Trevor Channel) (Shaw et al. 1989b).

Another less harmful but more widespread myxosporean parasite K. thyrsitis (Gilchrist, 1924) is found in Pacific hake from the Strait of Georgia, west coast Vancouver Island inlets, and offshore locations. This parasite is also prevalent in walleye pollock, some flatfish, and in several fish from Australia and South Africa (Kabata and Whitaker 1985). The presence of K. paniformis in the offshore stock but not in the Strait of Georgia or in Tahsis Inlet stocks indicates that this parasite likely infected the offshore Pacific hake stock subsequent to the separation of the inshore stocks (Kabata and Whitaker 1981, 1985). Distribution of parasites in the Genus Kudoa is further indication that resident Pacific hake stocks do not substantially intermingle with offshore migratory Pacific hake; Kudoa infection is spread either by release of spores from dead fish or via cannibalism.

Growth rate and body size

Due to the difficulty of visualizing scale annuli in Pacific hake, ageing of this species has typically occurred through analysis of the surface or internal annuli of otoliths (Etchevers 1971, Chilton and Beamish 1982). Due to difficulties in detecting growth zones in older, slower growing fish in the Strait of Georgia, Pacific hake in this area are aged by the "break and burn" method where the otolith is broken or sectioned through the nucleus and exposed to an alcohol flame, which enhances the contrast between the translucent and opaque zones (Chilton and Beamish 1982). Beamish (1979) stated that "age determinations using whole otoliths will not accurately determine the age of most older Pacific hake in some stocks." Puget Sound Pacific hake have routinely been aged by counting annuli on the surface of the otolith (Goñi 1988). Since growth zones on the otolith surface are difficult to identify in older, slower growing fish (Etchevers 1971), Beamish (1979) suggested that ages assigned to Pacific hake in the Puget Sound population by Kimura and Millikan (1977) may have underestimated the actual ages of older fish. Attempts to compare growth rates between stocks of Pacific hake are further compounded by apparent temporal changes in mean length-at-age and consequent interannual variations in mean growth rates within the offshore stock (Woodbury et al. 1995).

Hollowed et al. (1988) reported recent declines in mean length-at-age of offshore Pacific hake that may have been associated with the 1983 El Niño event or a density-dependent growth response to increased population abundance (Hollowed et al. 1988, Dorn 1992, Dark and Wilkins 1994). Despite differences in ageing methods applied to different stocks of Pacific hake, comparisons of growth parameters between stocks are routinely made.

Puget Sound Pacific hake have been reported to have a substantially slower growth rate than offshore Pacific hake (Alverson and Larkins 1969, Nelson and Larkins 1970). Likewise, Beamish et al. (1982) and McFarlane and Beamish (1985) noted that Pacific hake in the Strait of Georgia were considerably smaller than similar aged Pacific hake in the Strait of Juan de Fuca and off the west coast of Vancouver Island. Beamish et al. (1982) concluded that these differences supported the contention that Pacific hake in the Strait of Georgia are a separate stock from Pacific hake found in the western Strait of Juan de Fuca and offshore of Vancouver Island. The size of offshore and Strait of Georgia Pacific hake is reportedly similar up to the age at which they first mature, but offshore Pacific hake continue to increase in length, and reach larger sizes (Beamish 1979).

Goñi (1988) compared growth rate parameters from the literature for Puget Sound and Strait of Georgia Pacific hake and found between-stock differences in mean length-at-age that were significant for all cohorts examined. Comparison of growth plots of the two stocks revealed a consistent between-stock difference of about 5 cm in size-at-age. Puget Sound Pacific hake do not seem to grow as large overall as do Strait of Georgia Pacific hake (Goñi 1988).

Alverson et al. (1964) reported that mature Pacific hake taken off the Oregon-Washington coast averaged 52 cm in length with a range of from 22 to 71 cm. In the Strait of Georgia, the mean size of males was 52 cm and 54.5 cm for females between 1977 and 1981 (Beamish and McFarlane 1985). Between 1977 and 1981, the largest male and female Pacific hake reported from the Strait of Georgia were 77 and 84 cm, respectively, although very small percentages of either sex were greater than 60 cm in length (Beamish and McFarlane 1985). Most of the Pacific hake that occurred in the fishery in Port Susan in Puget Sound were from 32-45 cm in length (Pedersen 1985). Maximum lengths recorded by Pedersen (1985) for Puget Sound Pacific hake were 45 cm for males and 73 cm for females.

Nelson (1969) stated that for any given age, Pacific hake from inshore waters of Puget Sound and the Strait of Georgia are substantially smaller than the offshore migratory Pacific hake. For instance, the mean lengths of inshore Pacific hake at age 3 and 4 are 15 to 20 cm shorter than offshore Pacific hake of the same age (Nelson 1969). Pedersen (1985) stated that Puget Sound Pacific hake appear to be 2-4 cm larger at age 2 and 2-4 cm shorter at age 3 and older, than Strait of Georgia Pacific hake. Pedersen (1985) suggested that this relationship (and the fact that Puget Sound Pacific hake mature at a smaller size than do Strait of Georgia Pacific hake) may have been due to the intense commercial Pacific hake fishery in Puget Sound. The average sizes of Pacific hake in both Puget Sound and the Strait of Georgia are substantially smaller at the present time than they were in the 1980s. For example, very few Pacific hake larger than 30 cm are currently present in the Port Susan Pacific hake population (Figs. 21, 22).

Kautsky (1989) stated that "the coastal stock consistently attains larger sizes at age than the Puget Sound stock suggesting that the maximum attainable size for the Puget Sound stock is less than that for the coastal stock."

Shaw et al. (1989a) reported that mean length-at-age of Pacific hake in Trevor Channel in Barkley Sound on the west coast of Vancouver Island was significantly smaller than that for the migratory offshore Pacific hake from La Perouse Bank and Triangle Island off Vancouver Island.

Length and age at maturity

Table A-2 summarizes length at first maturity, at 50% maturity, and at 100% maturity for selected Pacific hake populations. Puget Sound and Strait of Georgia Pacific hake stocks appear to mature at a smaller size than the offshore migratory stock (McFarlane and Saunders 1997).

Historically, both male and female offshore Pacific hake matured at a length of about 40 cm (Best 1963), whereas male and female Pacific hake in the Port Susan population in Puget Sound matured at a length of about 30 cm (Kimura and Millikan 1977). Currently, length at 50% maturity for females in the Port Susan Pacific hake population is approximately 21.5 cm, compared to 29.8 cm in the 1980s (Fig. 23, Table A-2).

Length frequencies

Figures 21 and 22 illustrate the temporal decline in the size of survey-caught Pacific hake in the Port Susan spawning population from the late 1980s to the present. A large proportion of the Pacific hake in Puget Sound sampled in the 1987 research-trawl survey (Quinnell and Schmitt 1991) were greater than 30 cm length (Fig. 18), indicating that this decline in average length and shift to smaller size frequencies occurred after this period of time. In the latter half of the 1990s, few Pacific hake larger than 35 cm were caught in the Port Susan acoustic-trawl surveys and by 1999 the majority were less than 25 cm in length (Figs. 21, 22).

Fecundity

Like hake species elsewhere, the Strait of Georgia Pacific hake stock shows evidence of resorption of unreleased oocytes following spawning (Foucher and Beamish 1980, Mason 1986, McFarlane and Saunders 1997). MacGregor (1966, 1971) also noted that small-yoked oocytes were resorbed following spawning of larger eggs in a sample of female Pacific hake collected off California in March and April.

The presence of oocytes of different maturity stages in pre-spawning Pacific hake and the retention of small-sized yoked oocytes in spent or partially spent Pacific hake have been interpreted differently by various researchers. In the case of Merluccius hubbsi, M. gayi, M. merluccius, M. capensis, and M. paradoxus multiple size classes of oocytes in different maturity states and retention of yoked oocytes in post spawners have been interpreted as evidence for serial or batch spawning (Osborne et al. 1999, and references therein). Similarly, Ermakov (1974) interpreted multi-modal oocyte diameters in Pacific hake as evidence for multiple spawning events in a single year. However, other researchers (MacGregor 1966, 1971; Foucher and Beamish 1980; McFarlane and Saunders 1997) reported that smaller yoked oocytes that remain after spawning in M. productus were completely resorbed and that a second spawning did not occur. Although other species of Merluccius may be batch spawners, it is currently assumed that Pacific hake spawn only once per year. The retention of some oocytes after spawning in Pacific hake suggests that traditional methods of estimating fecundity are not applicable to Pacific hake. Therefore, McFarlane and Saunders (1997) have defined "effective fecundity" in Pacific hake "as the number of yoked oocytes that are actually released to be fertilized."

McFarlane and Saunders (1997) reported that although total fecundity does not differ among Pacific hake stocks, effective fecundity differs between the migratory offshore stock and the smaller discrete stocks of Pacific hake in Puget Sound and the Strait of Georgia. All three stocks of Pacific hake retained and resorbed a portion of their oocytes (10-12% for the offshore stock, 32-44% for the Puget Sound stock, and 38-58% for the Strait of Georgia stock), but the Strait of Georgia stock retained a considerably higher percentage of eggs than the other stocks, ranging from 38% for the largest fish to 58% for the smallest (McFarlane and Saunders 1997).

Morphological Differentiation

Morphometric discrimination

Ehrich and Rempe (1980) examined morphometric differences (diameter of bony orbit, head length, precaudal length, and distances from the tip of snout to end of the pectoral and 2^nd dorsal fin) between four groups of Pacific hake found in the northern and southern regions of the Gulf of California, offshore of Baja California to Alaska, and in nearshore regions of the west coast of Baja California. The greatest differences were found between the offshore population and the southern Gulf of California population, while the offshore population was most similar to the southern nearshore population off the west coast of Baja California (Ehrich and Rempe 1980).

Shape and size of the otolith

McFarlane and Beamish (1985) reported that sagittal otoliths from offshore Pacific hake were more elongate and less concave in section than otoliths from Strait of Georgia Pacific hake, although no statistical analyses were published to test these observations. Anonymous (1968) also reported that otoliths from Puget Sound Pacific hake "vary" from offshore Pacific hake otoliths.

A number of studies have attempted to utilize interspecific and intraspecific size and shape variation in otoliths to identify species, populations and stocks of various hake species in the genus Merluccius (Lombarte and Castellón 1991, Torres et al. 2000, Bolles and Begg 2000). Lombarte and Castellón (1991) applied multivariate analysis to a numerical description of otolith outlines for four size classes of fish in six species of Merluccius. Analysis of otoliths from fish greater than 20 cm in length correctly classified individuals into a Euro-African group (M. merluccius, M. capensis, and M. paradoxus) and an American group (M. bilinearis, M. productus, and M. gayi). Lombarte and Castellón (1991) concluded that these morphological differences "are a reflection of genetic distance between species." Within Pacific hake (M. productus), "otoliths taken from individuals from different geographical areas [presumably from off west coast Vancouver Island and California] had no influence on otolith shape." Lombarte and Castellón (1991) did not apparently make a comparison of otoliths in offshore Pacific hake with otoliths from inshore Pacific hake.

Torres et al. (2000) demonstrated clear geographical differentiation between two groups of M. gayi (from Chile and Peru) and between Atlantic and Mediterranean samples of M. merluccius in morphometric measurements of otoliths. In both species, all otoliths could be correctly assigned to the appropriate geographical sample based on otolith analysis. However, two groups of M. hubbsi from off the southeast coast of South America could not be differentiated on the basis of otolith morphometrics. Likewise, Bolles and Begg (2000) successfully used whole sagittal otolith morphometrics, specific to fish age, to differentiate silver hake (M. bilinearis) stocks from the east coast of North America into a northern stock from the Gulf of Maine to Georges Bank and a southern stock from southern Georges Bank to the Middle Atlantic.

Otolith morphometrics related to length and width can be expected to reflect localized environmental variables. Although variation in otolith morphometrics can be used to differentiate stocks or management units of fish, the usefulness of these differences in the delineation of a DPS in a marine fish species is dependent on the degree to which otolith variability reflects environmental or genetic differences between groups of fish.

Genetic Information

Genetic population structure of hake species

Inada (1981) recognized 12 species of hake in the Genus Merluccius: 1) European hake M. merluccius, 2) Senegalese hake M. senegalensis, 3) Bengualean hake M. polli, 4) shallow-water Cape hake M. capensis, 5) deep-water Cape hake M. paradoxus, 6) silver hake M. bilinearis, 7) offshore hake M. albidus, 8) Pacific hake M. productus, 9) Panamanian hake M. angustimanus, 10) Chilean hake M. gayi, 11) Argentinian hake M. hubbsi, and 12) New Zealand hake M. australis.

Interspecific allozymic variation of hake has been investigated by Stepien and Rosenblatt (1996), Roldan et al. (1999), and Galleguillos et al. (1999), while Becker et al. (1988) and Quinteiro et al. (2000) examined between-species genetic divergence using mtDNA RFLP variation and comparison of sequence divergence in the control region of mtDNA, respectively (see "Glossary" for definitions). Intraspecific relationships have been studied using allozyme electrophoresis in M. merluccius (Pla et al. 1991, Lo Brutto et al. 1998, Roldan et al. 1998), M. capensis and M. paradoxus (Grant et al. 1987b, and references therein), M. hubbsi (Roldan 1991), and M. productus (Anonymous 1968; Utter 1969a, b; Utter and Hodgins 1969, 1971; Utter et al. 1970). Lundy et al. (1999) have investigated population structure in European hake through variation at six microsatellite loci.

The European hake, M. merluccius, is distributed along the eastern Atlantic coast from Norway to Morocco and throughout the Mediterranean Sea. Early efforts at detecting genetic population structure in European hake with protein electrophoresis revealed no significant variation at three allozyme loci among twelve samples ranging from Norway to the Bay of Biscay (Mangaly and Jamieson 1978). More recent genetic studies, using up to 21 polymorphic allozyme loci, have indicated a clear genetic difference between European hake in the Atlantic Ocean and the Mediterranean Sea, with the Straits of Gibralter acting as a geographic barrier (Pla et al. 1991, Roldan et al. 1998). Lo Brutto et al. (1998) detected insignificant levels of allozyme variation at four polymorphic loci among populations of M. merluccius along the coasts of Italy and Sicily. Despite the reported genetic homogeneity among Italian populations, Roldan et al. (1998) found significant allozyme genetic evidence of population substructuring in both Atlantic and western Mediterranean M. merluccius. Similarly, Lundy et al. (1999) found significant population subdivisions between Mediterranean and Atlantic European hake, but no substructure within the Mediterranean, using six polymorphic microsatellite loci. However, Lundy et al. (1999) did find significant differentiation in the same microsatellite loci between Bay of Biscay and Portuguese populations, which are currently managed as one stock, but no differentiation between southern Bay of Biscay and Celtic Sea populations, which are managed as separate stocks.

Grant et al. (1987b) detected only small amounts of genetic divergence by allozyme electrophoresis between stocks of both M. capensis and M. paradoxusoff Namibia and South Africa. More than 98% of the total genetic diversity in these species was found to occur within sampling locations for both species. Nei's genetic distances (D) between samples were generally less than 0.001. Although three widely separated spawning grounds have been identified for M. australis in New Zealand waters (Colman 1995), Smith et al. (1979) were unable to detect significant differences in allele frequencies at two polymorphic allozyme loci among four New Zealand sampling locations. Roldan (1991) found a complex structure to occur among M. hubbsi populations on the Argentinian continental shelf upon analysis of 4 polymorphic allozyme loci sampled at 10 locations. However, genetic heterogeneity among samples was primarily due to variation at a single locus (EST-1*) and sample sizes were relatively small (Roldan 1991). In general, species of Merluccius that have been investigated tend to show subdivided population structure around geographically complex coastlines (Roldan et al. 1998, Lundy et al. 1999), but not along linear coastlines (Smith et al. 1979, Grant et al. 1987b).

Pacific hake genetics

In a series of publications, Utter and coauthors (Utter 1969a, b; Utter and Hodgins 1969, 1971; Utter et al. 1970) compared protein electrophoretic variation in Pacific hake from various locations in Puget Sound, off the Oregon-Washington coast, and off southern California at four polymorphic loci (lactate dehydrogenase (LDH), transferrin, muscle protein, and esterase). Two alleles were detected at both the muscle protein and LDH loci, four at the transferrin locus, and five at the esterase locus (Utter and Hodgins 1971). No evidence of heterogeneity was found at LDH or esterase within or between the two sampling locales for offshore Pacific hake (off Oregon/Washington and southern California) (Utter and Hodgins 1969, 1971; Utter et al. 1970). Comparison between multiple samples of Pacific hake taken off the outer coasts of Oregon and Washington also revealed no heterogeneity at the transferrin or muscle protein loci (Utter 1969b, Utter and Hodgins 1971).

However, Utter and Hodgins (1971) stated that allelic frequencies of all four polymorphic loci differed significantly between offshore and Puget Sound Pacific hake and indicated that these populations were reproductively isolated. The average and range of frequencies of the most common allele for the four loci for the two regions were as follows: 1) esterase, 0.603 (range 0.577-0.655) in offshore samples (n=358) and 0.828 (range 0.733-0.904) in Puget Sound (n=903); 2) transferrin, 0.564 (range 0.536-0.583) in offshore (n=203) and 0.696 (range 0.672-0.750) in Puget Sound (n=115); 3) skeletal muscle protein, 0.982 (range 0.969-0.992) for offshore samples (n=225) and 0.730 (range 0.705-0.823) for Puget Sound (n=250); and 4) LDH, 0.980 in offshore samples (n=355) and 0.745 (range 0.695-0.794) in Puget Sound (n=762) (Utter 1969b; Utter and Hodgins 1969, 1971; Utter et al. 1970). Many of the Pacific hake samples used in the above allozyme studies of Utter and coauthors were collected in Puget Sound outside of the spawning season and distant from known spawning grounds; however, several collections (particularly for esterase and LDH) were made of fish in or near the spawning grounds (Port Susan) and during the spawning season and these samples did not differ significantly from any of the other Puget Sound samples (Utter 1969b; Utter and Hodgins 1969, 1971; Utter et al. 1970).

Utter et al. (1970) included analysis of esterase variation of one sample of 80 Pacific hake juveniles collected in Hood Canal (Dabob Bay in Hood Canal is a known Pacific hake spawning ground). The frequency of the most common allele in this sample (0.831) did not differ significantly from that of other samples taken in Puget Sound (average frequency of 0.828 for 12 samples) (Utter et al. 1970).

Prior to the recent decrease in body size of inshore Pacific hake (see "Length and age-at- maturity" section), Puget Sound fish averaged approximately 35 cm and offshore fish averaged about 50 cm. However, observations of large-sized (greater than 60 cm) Pacific hake have been made in both Puget Sound and the Strait of Georgia and speculation as to whether these large fish are from the offshore population has been made. Anonymous (1968) addressed this question and stated that: Hake of oceanic size have occasionally been caught in Puget Sound, which raised the question of whether the larger fish were migratory or indigenous. ... The gene frequencies of the large and normal fish in Puget Sound agreed with those of smaller fish from the same area. This indicated that the larger fish are indigenous to Puget Sound.

Goñi (1988) examined restriction fragment length polymorphism (RFLP) variation of mitochondrial DNA (mtDNA) in Pacific hake collected from California (four individuals pooled, collected off Cape Mendocino, California in August), Puget Sound (four individuals pooled, collected off West Point, Washington in August), and the Strait of Georgia (two separate individuals, collected in the central Strait of Georgia in November). Goñi (1988) observed four composite mtDNA haplotypes amongst these samples and stated that "The geographical distribution of these genotypes seems to reveal a certain degree of mixture between populations." Goñi (1988) also stated that "The apparent absence of high diversity in the mtDNA molecules might indicate that the three stocks either intermingle to a certain extent, or are units that have recently formed." However, several factors make the interpretation of Goñi's (1988) mtDNA study difficult. Homogenization of both the California and Puget Sound samples was done by Goñi (1988) with the assumption that within-sample variation was nonexistent. However, within-sample variation was found in pooled California and separate Strait of Georgia samples, leading Goñi (1988) to conclude that homogenization was inappropriate and may have masked the true results. Another factor that complicates the interpretation of Goñi's (1988) results is that all samples were collected outside of the spawning season and a considerable distance away from known spawning grounds of Pacific hake. The small sample sizes used in this study would also indicate that Goñi's (1988) study should be considered inconclusive.

Information Relevant to the Pacific Hake DPS Question

As stated in the previous "Approaches to the Species Question and to Determining Risk" section, four broad types of information were analyzed by the BRT in its determinations of whether Pacific hake in Puget Sound represent a "discrete" and "significant" population and therefore qualifies as a DPS under the ESA. These are: habitat characteristics, phenotypic and life-history traits, mark-recapture studies, and analysis of neutral genetic markers. As such data can only be properly evaluated in relation to similar information for the biological species as a whole, Puget Sound Pacific hake data were compared with data from Pacific hake from throughout the species' range.

As detailed in the previous sections on "Environmental History and Features of Puget Sound" and "Phenetic and Genetic Information Relating to the Species Question," specific information in the following categories was available for Puget Sound Pacific hake: physical habitat, spawning time and location, year-class strength, growth rate and body size, size and age at maturity, length frequency, fecundity, and protein electrophoretic variation. Data on migration patterns, tagging, parasite incidence, meristics and morphometrics, and genetic population structure using contemporary techniques were largely unavailable for Pacific hake in Puget Sound. A similar assemblage of data was available for Pacific hake from the Strait of Georgia, although protein electrophoretic data were lacking and studies on the incidence of the parasite Kudoa paniformis were available. With the exception of tagging and a contemporary study of genetic population structure, all categories of information mentioned above were available for offshore Pacific hake. The previous section on "Approaches to the Species Question and to Determining Risk" should be consulted for a general discussion of the relative usefulness of the various categories of data for DPS delineation. Issues of biological data quality for Pacific hake are addressed for each category in the preceding section on "Phenetic and Genetic Information Relating to the Species Question."

Discussion and Conclusions for Pacific Hake DPS Determinations

The BRT considered several possible DPS configurations for populations of Pacific hake in the northeastern Pacific Ocean in its attempt to identify a "discrete" and "significant" segment of the biological species that incorporates Puget Sound Pacific hake. After careful consideration of the available information, the BRT concluded that inshore resident Pacific hake from Puget Sound and the Strait of Georgia are part of a separate DPS from offshore (coastal) migratory Pacific hake that are seasonally distributed from southern California to as far north as southeastern Alaska. These inshore Pacific hake will hereafter be identified as the Georgia Basin Pacific hake DPS (Figure 1). Pacific hake that spawn occasionally off the west coasts of Oregon, Washington and Vancouver Island were considered to be opportunistic spawners belonging to the offshore Pacific hake stock and not part of the Georgia Basin DPS. Lack of biological information precluded the BRT from drawing any firm conclusions about the affinities of Pacific hake from west coast Vancouver Island inlets. At the present time, Pacif hake from west coast Vancouver Island inlets are not considered to be part of the Georgia Basin DPS.

The BRT identified a variety of evidence to support their conclusion that Georgia Basin Pacific hake constitute a separate DPS relative to offshore Pacific hake: 1) Differences in annual migration behavior; 2) significant allozyme frequency differences between Puget Sound and offshore Pacific hake; 3) absence of the protozoan parasite Kudoa paniformis in inshore populations compared to its common occurrence in offshore Pacific hake; 4) differences in otolith morphology between Strait of Georgia and offshore Pacific hake; 5) distinctiveness of the habitats of inshore Pacific hake (they spawn in deep, inshore basins that receive large freshwater inputs and are the only populations of Pacific hake that inhabit fjord-like environments); 6) wide geographic separation of inshore and offshore spawning locales; and 7) demographic data showing inshore Pacific hake are generally smaller for a given age, mature at a smaller size, and reach a smaller maximum length than offshore fish.

The BRT expressed several concerns about the available data; for example: 1) it is not clear to what degree demographic differences between Georgia Basin and offshore Pacific hake are driven by environmental or genetic differences, 2) some of the allozyme loci that show differences between the Puget Sound and offshore Pacific hake have been shown to be under selection in other animals, and 3) there is no obvious physical barrier preventing mixing of offshore and Georgia Basin Pacific hake, especially during the June-August period when offshore Pacific hake may occur near the mouth of the Strait of Juan de Fuca.

The Georgia Basin DPS encompasses at least five geographically-discrete spawning aggregations in deep-water basins, including Dabob Bay and Port Susan in Puget Sound and south-central Strait of Georgia, Stuart Channel, and Montgomery Bank in the Strait of Georgia (Figs. 1, 19). Therefore, the BRT considered whether there is evidence for multiple populations or stocks of Pacific hake within this DPS and, perhaps, multiple DPSs within the Puget Sound/Strait of Georgia area. Such information is limited. The majority of the BRT felt that good evidence that stock structure may exist within the Georgia Basin DPS includes: 1) the presence of geographically-discrete and temporally-persistent spawning aggregations, and 2) demographic differences between Strait of Georgia and Puget Sound fish. Tagging and genetic data for within Georgia Basin comparisons are unavailable or incomplete. Data showing apparent asynchronous year class strength between Puget Sound and Strait of Georgia Pacific hake were viewed as technically flawed (see above "Year class strength" section). Although the BRT could not with any certainty identify multiple populations or DPSs of Pacific hake within the Georgia Basin, the majority of the BRT acknowledged the possibility that significant structuring may exist within the proposed DPS and that such structure might be revealed by new information in the future.

Offshore Pacific hake migrate annually between summer feeding areas in waters off Oregon, Washington, British Columbia, and occasionally as far north as south central Alaska to spawning areas off southern California. The BRT did not attempt to determine whether offshore Pacific hake are composed of more than one DPS.

Assessment of Extinction Risk

Introduction

The petition discussed decline in abundance (Palsson et al. 1997), decline in average size, and predation by marine mammals (Schmitt et al. 1995) in its proposal to list Pacific hake in South Puget Sound. South Puget Sound was defined in the petition as the Sound east of Deception Pass and to the south of and east of Admiralty Point and south of Point Wilson on the Quimper Peninsula. Although the petition only discussed the spawning population of Pacific hake in the Port Susan area, it is known that Pacific hake also spawn in Dabob Bay (Fox 1997). The BRT concluded that Puget Sound populations of Pacific hake are part of the Georgia Basin DPS.

This section presents results of review and analysis of available information on abundance, evaluation of risk of extinction of the Port Susan population, and evaluation of the risk of extinction of the DPS as a whole. Hydro-acoustic estimates of the Port Susan population were revised under assumptions that are more appropriate for the risk analysis than those originally used. Also, new target strength estimates based on recent developments in hydro-acoustic technology were used for the revision. Risk assessment of the Port Susan population used two models to analyze the impact of pinniped predation under a wide range of assumed levels of predation. There were insufficient data available to evaluate the status of the Dabob Bay population. There were also insufficient data to perform more than a semi-quantitative analysis of the risk of extinction of the Canada portion of the DPS or of the DPS as a whole.

Information on Abundance and Composition

Port Susan

Biomass estimates of Pacific hake in Port Susan were given by Palsson et al. (1997)(Table 3). The Washington Department of Fish and Wildlife (WDFW) produced the estimates from annual hydro-acoustic surveys (Lemberg et al. 1990). After examination of available data and consultations with Wayne Palsson (W. Palsson^[5]) and Martin Dorn (M. Dorn^[6]) it was decided that analysis of the data shown in Table 3 could be improved in several ways for the risk analysis.

WDFW designed the surveys to produce estimates of biomass available to the fishery in each year. Their information indicated that peak abundance usually occurs in March. Since the fishing season often began in the preceding fall, WDFW usually added catches up to the time of the survey to the survey results to obtain a biomass estimate at the beginning of the fishing season. The fishery ceased in 1991. WDFW used one to three surveys taken in late February through mid March. Also WDFW and the industry desired that immature fish not be harvested. Pacific hake matured at about 30 cm during the early years of the survey. In most years WDFW used catch compositions of trawl surveys to first convert acoustic biomass estimates to estimates of Pacific hake biomass and then to convert Pacific hake biomass estimates to estimates of biomass of greater than 29 cm. However, WDFW included smaller Pacific hake in estimates for the earlier years. WDFW did not conduct trawl surveys in 1994 or 1995, but made biomass estimates from hydro-acoustic surveys (biomass estimates not in Table 3).

Wayne Palsson (W. Palsson^[7]) provided biomass estimates from 1982 through 1999 (data for the year 2000 were received subsequent to the analyses) and information about the quality of the surveys. It was decided not to use the 1994 and 1995 estimates, because WDFW did not conduct trawl surveys, and their1995 acoustic survey was in early February which is before the time of normal peak abundance. While WDFW's decision to add catch to the survey estimates and estimating biomass of Pacific hake greater than 29 cm were appropriate for fishery management, it was decided to use estimates of biomass of all Pacific hake in the Port Susan area at the time of surveys for risk assessment. The surveys occurred during the spawning season, which seemed the appropriate season for examination of productivity of the population. Catch was not added to the

survey estimates. Examination of data provided by WDFW (W. Palsson^[8]) revealed that size of maturity decreased since the early 1980s (Fig. 23). Recent surveys captured Pacific hake that were smaller than Pacific hake captured in earlier years, but mature fish comprised most of the biomass in both time periods. In addition, fish less than 30 cm comprise a significant proportion of Pacific hake consumed by pinnipeds (P. Gearin^[9]).

Martin Dorn (M. Dorn^[10]) reviewed the first draft of this document and noted that both NMFS (Traynor 1996) and Canada Department of Fisheries and Oceans (DFO) (Kieser et al. 1999) now use target strength relationships dependent on length for hydro-acoustic estimates of Pacific hake biomass rather than the constant target strength procedure used by WDFW. The length-dependent target strength method is considered more accurate and after consultation with WDFW (W. Palsson^[11]), the biomass estimates for Port Susan were revised (M. Dorn^[12]) Average weights and Pacific hake length frequency data needed for the revision were compiled from data supplied by WDFW. Length-frequency samples from trawls taken during the surveys (Fig. 21) were weighted equally in terms of weight rather than numbers of sampled fish to avoid bias towards larger fish. There were changes in trawls used for the surveys during the time span. These changes were assumed to not have significant impacts on the composition of the catch. Data from Kautsky (1989) were used to estimate that target strength = 20 log length - 73.5 (M. Dorn^[13]). New estimates of Pacific hake biomass under both the constant and length-dependent target strength models are shown in Table 4. Estimates made under the length-dependent target strength model were used for the following analyses.

Biomass estimates (Table 4) made under the length-dependant target strength assumption were higher than estimates made under the constant target strength assumption until 1997. The 1999 biomass declined to 12% of the 1983 estimate under the length-dependent target strength assumption compared to 19% under the constant target strength assumption (Fig. 24). Although catches were not added to survey biomass estimates, the new estimates (Table 4) were similar to or higher than the old estimates during the 1983-1993 period (Table 3). Average weight decreased from 0.298 kg in 1982 to 0.072 kg in 1999. There does not appear to be a trend in numbers of Pacific hake in the survey area (Fig. 25).

Preliminary results from the March 7, 2000 WDFW Port Susan Pacific hake survey were received subsequent to the above analyses (W. Palsson^[14]). Pacific hake biomass estimates were calculated using the length dependent target strength methodology described above (M. Dorn^[15]). Results are shown in Table 4. Reliable acoustic data were not available for the Possession Sound portion of the 2000 survey, because of equipment problems, and WDFW estimates that 15-20% of the total stock may have been missed (M. Dorn^[16]). The new estimates indicate that both biomass and numbers are at the lowest level since the surveys were started in 1982. If the survey missed 20% of the total biomass, the corrected biomass would be 1,240 mt, which would be the lowest on record, 52% of the 1999 biomass, 6% of the peak biomass in 1983, and represent an 85% decrease during the past 15 years. Average weight increased from 0.072 kg in 1999 to 0.091 kg in 2000. Compared to recent years there were relatively few fish smaller than 20 cm and relatively more fish larger than 30 cm.

Palsson et al. (1997) presented estimates of mid-water trawl catch per effort (Table 5) and Pacific hake biomass estimates from bottom trawl surveys (Table 6). Catch-per-effort data were not used in this analysis because of the difficulties in adjusting the data for undocumented changes in gear and fishing strategies. Bottom-trawl survey estimates were not used because there were not enough to serve as an index, and bottom-trawl surveys are not suitable for estimates of absolute abundance of Pacific hake because of the semi-pelagic behavior of Pacific hake.

Canadian portion of the Strait of Georgia

The DFO conducts periodic hydro-acoustic estimates of biomass of Pacific hake in the Canadian portion of the Strait of Georgia using length-dependent target strength (Saunders and McFarlane 1999). Timing of the surveys has changed. There was concern that March-April estimates included signal from the spring plankton bloom, particularly in 1981 and 1993 (Saunders and McFarlane 1999, Kieser et al. 1999). Since 1993, the surveys have been conducted in February. There was concern that February surveys occurred before peak in spawning and may have underestimated the biomass (Saunders and McFarlane 1999, Kieser et al. 1999). The estimates are shown in Table 7. Saunders and McFarlane (1999) stated that  "At the present time we do not have an adequate absolute or relative index of stock size and the recent biomass estimates should be considered a conservative minimum. Based on the information briefly stated above and reported in detail in Kieser er al. (1999) we believe the biomass of Pacific hake in the 1990's to be stable at approximately 50-60,000t."

Data in Saunders and McFarlane (1999) also revealed that, as in Puget Sound, average size of Pacific hake in the Strait of Georgia has decreased. Size-at-age data indicated that growth between ages 2 and 3 years considerably decreased between 1976 and 1999. Age-composition data indicated that the 1991-1992 year classes were strong and persisted in the samples through 1999. The 1995 and 1998 year classes were also strong compared to adjacent year classes, but do not appear to be as strong as the 1991-1992 year classes.

Risk Assessment

Port Susan

It was more difficult to estimate Pacific hake exploitation by pinnipeds than by humans. California sea lions and harbor seals are known to consume Pacific hake (Olesiuk 1993, Schmitt et al. 1995). Schmitt et al. (1995) estimated Pacific hake consumption by California sea lions in Puget Sound for the 1986-1994 period. However consultation with knowledgeable marine mammal experts, including the two junior authors of Schmitt et al. (1995) (S. Jeffries^[18] and P. Gearin^[19]), revealed that these estimates were not acceptable to the marine mammal research community. In addition, researchers have not estimated Pacific hake consumption by harbor seals in Puget Sound. Also, researchers have not attempted to understand functional relationships between Pacific hake consumption by pinnipeds and the abundance of Pacific hake and other potential prey.

Because of the uncertainty, Pacific hake consumption by pinnipeds in Puget Sound was treated as hypothetical values in what-if risk assessments of the Port Susan Pacific hake population. After consultation with experts at the NMFS's National Marine Mammal Laboratory (NMML), ranges of values were used that were consistent with published and unpublished information in the sense that the ranges were likely to include the real levels of consumption. There was insufficient knowledge to conclude that the actual levels were likely to be close to the center of the ranges.

Patrick Gearin (P. Gearin^[20]) indicated that estimates of consumption of all food items in Puget Sound by California sea lions and harbor seals given in NMFS (1997) are consensus estimates by the marine mammal research community and thus acceptable to them as the best available. They estimated that California sea lions on the average consumed 830 mt per year between 1986 and 1994, which is close to the lower estimate of Schmitt et al. (1995). They did not use the upper estimate of Schmitt et al.(1995), because they believed that it was not justified by research information. NMFS (1997) estimated that in 1993 harbor seals consumed 3,209 mt in Eastern Bays and 1,649 mt in Puget Sound proper (Fig. 26). They also provided an estimate for Hood Canal, but it was assumed that harbor seals in Hood Canal prey on the Dabob Bay rather than the Port Susan population. The peak count of sea lions in Puget Sound was 444. Population abundances of harbor seals were 3,479 in Eastern Bays, and 1,787 in Puget Sound proper.

Jeff Laake (J. Laake^[21]) provided estimates of predicted annual monthly counts of sea lions at Everett, Washington for 1986-1998. Year was defined in the same manner as for fishing year, which is March through February of the following year. Actual counts were available for about half of the possible year-month combinations. He used a generalized additive model containing spline-smoothed functions for year, season, and year-season to predict the average monthly counts. A Poisson error structure with over dispersion was assumed. Patrick Gearin (P. Gearin^[22]) provided peak count data for 1982-1999. Peak counts usually occurred in about March. A regression between peak count and average monthly count was used to estimate average monthly count for 1982-1985 and 1999. Average counts were then doubled because Schmitt et al. (1995) indicated that counts probably represented about 50% of the total Puget Sound population as was done for consumption estimates in NMFS(1997). Sea lion counts increased from 1982 to 1986, decreased from 1986 to 1989, increased from 1989 to 1995, and decreased from 1995 to 1999 (Table 9).

The literature details difficulties in estimation of pinniped diet composition (see Olesiuk 1990). These difficulties center around questions concerning prey specific digestion and retention rates. Variation in digestion and retention rates are also a source of uncertainty in studies of diet composition of fish, but compositions of stomach contents are usually used for fish studies, while compositions of scat contents are the predominant data source for pinniped studies. Different rates of digestion and retention are likely to produce less severe problems for stomach contents than for scat contents.

Schmitt et al. (1995) estimated that Pacific hake comprised 32% of the diet of California sea lions in Puget Sound during the 1986-1994 period. Their estimates were based on the estimated mass of individual prey items. Estimates in Schmitt et al. (1995) seem consistent with a more recent unpublished summary (Gearin et al. 1999), which showed that about 82% of sea lion scats contained Pacific hake parts, while the next two important items were dogfish parts, at about 22%, and salmon parts, at about 15%. Pacific hake parts are more likely to resist destruction by digestion than either spiny dogfish parts or salmon parts. However, since the major concentration of sea lions in Puget Sound overlaps both spatially and temporally with the major Pacific hake spawning activity, it would seem likely that Pacific hake comprise a significant portion of sea lion diets. Olesiuk et al. (1990) estimated boundaries on their point estimates of diet composition of harbor seals in the Canadian portion of the Strait of Georgia. Their gadiform contribution to the diet was 45.1%. Their lower limit was 28.0% (62% of point estimate) and upper limit was 60.9% (135% of point estimate). Schmitt et al. (1995) did not provide boundaries and used different methodologies in their study of California sea lions. It seemed reasonable to use a range that is broader than that used by Olesiuk et al. (1990) and to set the bounds at 50% and 200% of the Schmitt et al. (1995) estimates in an attempt to include the true value. The hypothetical range of consumption of Pacific hake by California sea lions in Puget Sound was calculated by multiplying total consumption by 0.16 (0.5 x 0.32) and 0.64 (2 x 0.32). Hypothetical estimates of Pacific hake consumption by California sea lions are shown for 10 levels within the above range in Table 9. The hypothetical estimates assume that consumption per sea lion was independent of Pacific hake abundance, and constant during the 1982-1999 time period.

Robert DeLong (R. DeLong^[23]) provided information on annual rates of change of populations of harbor seals based on WDFW/NMML data. Harbor seals were estimated to have increased by 3.3% annually in Puget Sound between 1985 and 1997. They were estimated to have increased by 2.7% annually in Eastern Bays between 1983 and 1998. It was assumed that the expansion rates applied to the entire 1982-1999 period for estimation of consumption of Pacific hake rates. The estimates of harbor seal abundance in 1993 by NMFS (1997) were used for the baseline population.

Researchers have developed less information on composition of the diet of harbor seals in Puget Sound than in the Canadian portion of Strait of Georgia or for California sea lions in Puget Sound. Pacific hake parts frequently occur in harbor seal scat samples (79%-Skokomish River, 84% - Hamma Hamma River, 100% - Duckabush River, 85% - Dosewallips River and 88% - Quilcene Bay) (S. Jeffries^[24]). These estimates are for the Hood Canal area and are shown here only to illustrate that Pacific hake apparently can comprise a significant portion of harbor seals in the general Puget Sound area. Pacific hake parts were estimated to occur in 32% of scat samples and Pacific hake comprised 5% of the diet by weight of harbor seals at Gertrude Island (South Puget Sound) from June 24, 1994 to October 23, 1995 (P. Gearin^[25]). Pacific hake also occurred in 80% of scat samples and comprised 83% of the diet by weight of harbor seals at Everett from January-April, 1989 and October-November, 1995. Olesiuk et al. (1990) estimated that Pacific hake comprised 42.6% of the diet of harbor seals in the Strait of Georgia. Since harbor seals are not as concentrated in the Port Susan area as sea lions are, it seems reasonable to set the bounds of Pacific hake contribution to harbor seal diet lower than used for sea lions. In the Eastern Bays, which includes Port Susan, it was set at 10-40%, a four-fold change from low to high as for sea lions. The low bound seemed reasonable, although it is two times higher than the estimate for Gertrude Island, which appears to be an extreme location. The high bound is about half of the Everett estimate, which also probably is an extreme location. At the suggestion of Robert DeLong (R. DeLong^[26]), the hypothetical Pacific hake contribution in the diet of harbor seals in Puget Sound was set at 5%, which is the estimate for Gertrude Island. The estimates are shown in Table 10. The hypothetical estimates assume that consumption per harbor seal was independent of Pacific hake abundance and constant during the 1982-1999 time period.

Population productivity--Two models were used to estimate the productivity of the Pacific hake population during the 1982-1999 period. The first model assumes that the annual consumption of Pacific hake by an individual pinniped is independent of Pacific hake abundance. The second model assumes that annual consumption of Pacific hake by an individual pinniped is described by the catch equation usually used to describe fish population dynamics, (i.e. it is dependent on abundance of Pacific hake, rate of natural mortality for Pacific hake, human generated fishing mortality, and number of pinnipeds). Both models assume that all estimated human and pinniped consumption is from the Port Susan population. While the commercial fishery and most observed sea lions occur in the Port Susan area, there are substantial occurrences of harbor seals in other areas of Puget Sound. The portion of Pacific hake from other populations consumed by pinnipeds is unknown. It is also not known if Hood Canal harbor seals or harbor seals west and/or north of Eastern Bays consume Pacific hake from the Port Susan population.

Under the first model, productivity in year i is
Prod(i) = (Bio(i+1) - Bio(i) + C_h(i) + C_sl(i) + C_hs(i))/Bio(i) (1)
Where,
Bio(i) = Biomass of Pacific hake in year i,
C_h(i)
= Catch by humans in year i, C_sl(i) = Consumption by California sea lions in year i, and
C_hs(i) = Consumption by harbor seals in year i.

Pacific hake biomass was estimated to increase considerably between 1993 and 1996 (Table 4). However missing biomass estimates for 1994 and 1995 preclude estimation of annual productivity estimates for 1993, 1994, and 1995. The missing data were approximated by assuming that productivity was constant for those three years and using iteration to estimate Prod(1993), Prod(1994), Prod(1995), Bio(1994), and Bio(1995).

Estimates of average annual productivity during the 1982-1998 period increased with pinniped consumption and ranged from 0.13 to 0.38 (Table 11). There was no obvious temporal trend in productivity at the higher assumed levels of predation, but productivity tended to decline over time (nonsignificant, r = -0.33) when pinniped predation was assumed to be low (Fig. 27). The lowest estimated annual value was -0.46 in 1996 under the hypothetical minimum pinniped predation. The highest value was 1.03 in 1982 under the hypothetical maximum pinniped predation. Estimates of productivity include impacts of migration to and from other populations of the DPS. It is not known what proportion of the estimated productivity is the result of migrations.

Under the second model biomass in year i+1 is

Bio(i+1) = Bio(i)e^-Z(i)(2)
Where,
Z(i) = M +F(i) - G(i),
M = Constant instantaneous rate of natural mortality,
F(i) = Instantaneous rate of exploitation mortality from all causes in year i,
F(i) = F_h(i) + F_sl(i) + F_hs(i),
F_h(i) = Instantaneous rate of mortality caused by exploitation by humans in year i,
F_sl(i) = Instantaneous rate of mortality caused by exploitation by sea lions in year i,
F_hs(i) = Instantaneous rate of mortality caused by exploitation by harbor seals in year i, and
G(i) = Instantaneous rate of productivity in year i. It includes migration to and from other populations.

M was assumed to be 0.23, which is the value used in Dorn et al. (1999a) to assess the offshore stock of Pacific hake. The offshore stock estimate included impacts of predation by pinnipeds, which are probably of minor importance compared to the Port Susan population, and was used to describe changes in numbers rather than biomass. The Port Susan population appears to be shorter lived than the offshore stock and thus probably has a higher value of M. However, M as used in the model does not include the impact of predation by pinnipeds.

The following constraints and relationships were used to solve iteratively for G(t). It was assumed that G(t) was approximately constant between 1993 and 1995 for middle levels of pinniped predation.

Total consumption in year i is
C(i) = Bio(i)F(i)(1-e^-z(i))/Z(i) and
C(i) = C_h(i) + C_sl(i) + C_hs(i).
Where
C_h(i) = human consumption in year i,
C_sl(i) = sea lion consumption in year i, and
C_hs(i) = harbor seal consumption in year i.
It follows from (2) that
Z(i) = -ln(Bio(i+1)/(Bio(i)).
Catchability coefficients q were estimated for sea lions and harbor seals where,
F_sl(i) = q_slN_sl(i),
N_sl((i) = number of sea lions in year i,
F_hs(i) = q_hsN_hs(i), and
N_hs(i) = number of harbor seals in year i.

The productivity estimates (Table 12) are similar to the results obtained using Model 1. Average productivity was greater than natural mortality, increased with increased hypothetical level of predation by pinnipeds, and ranged from 0.30 to 0.51. There was no obvious temporal trend in productivity at the higher assumed levels of predation, but productivity tended to decline over time (nonsignificant, r = -0.42) when pinniped predation was assumed to be low (Fig. 28). The lowest estimated annual value was -0.49 in 1996 under the hypothetical minimum pinniped predation. The highest value was 1.23 in 1982 under the hypothetical maximum pinniped predation.

Results of both models suggest that the Port Susan Pacific hake population would have increased between 1982 and 1999, if there had been no commercial exploitation and no pinniped predation, and either model held. It is likely that productivity would be lower if the population were approaching the carrying capacity of its habitat. Since the results did not indicate a positive trend in productivity as the population decreased, we are not able to estimate the carrying capacity. Population dynamics theory predicts that productivity would increase as biomass decreases. The lack of such a response for Port Susan Pacific hake suggests that productivity may have been impacted by natural or human related factors. One possible factor is the relatively warm climate conditions experienced since 1976. Average weight of Pacific hake decreased from 0.298 kg in 1982 to 0.072 kg in 1999. The decrease may have been partially caused by decreased growth as occurred for Pacific hake in the Canadian portion of the Georgia Basin (Saunders and McFarlane 1999). The possible decrease in growth may have been related to the relatively warm conditions or smaller size-at-maturity (Fig. 23) and may have had a negative impact on productivity. It is possible that the theoretically expected negative relationship between biomass and productivity would have been strong enough to significantly reduce the observed decline in Pacific hake biomass, if unknown factors had not affected the ability of the population to respond to decreased levels.

Both models have theoretical deficiencies in the description of predation by pinnipeds. For example, under the first model the consumption of Pacific hake per pinniped is constant until extinction of the Pacific hake population, and under the second model Pacific hake consumption per pinniped increases without bounds as the Pacific hake population increases. Both models ignore the effect of varying abundances of other prey.

There was a non-significant (r = -0.03) negative relationship between Pacific hake abundance and average sea lion count between 1986 and 1999, not including 1994 and 1995. Since California sea lion aggregations did not regularly occur in Puget Sound until 1979 (Schmitt et al. 1995), and sea lion abundance tended to increase until 1986, the year 1986 was chosen as the first year to examine the sea lion-Pacific hake relationship. There is no apparent trend in sea lion abundance in Puget Sound since 1986, although the coast-wide stock has continued to increase (NMFS 1997).

d) Projections-

Hypothetical projections (see Appendix B) indicated that uncertainty about rates of predation of Pacific hake by pinnipeds and the form of the relationships between Pacific hake predation by pinnipeds and commercial fishing precludes definitive conclusions concerning the risk of extinction of the Port Susan Pacific hake population.

Table 9. California sea lion counts and hypothetical consumption of Pacific hake in Puget Sound. California sea lion count data for 1986-1998 and estimates of ten levels of consumption of Pacific hake by California sea lions based on information provided by Jeff Laake (NMFS, F/AKC4, National Marine Mammal Laboratory, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.). California sea lion count data for 1982-1985 and 1999 data estimated from regression of mean counts and peak counts. Peak count data provided by Patrick Gearin (NMFS, F/AKC4, National Marine Mammal Laboratory, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.).

      Year        Mean sea lion count                           Ten levels of Pacific hake consumption (mt) by California sea lions

Georgia Basin DPS

Saunders and McFarlane (1999) indicated that a conservative estimate of the biomass of Pacific hake in the Canadian portion of the Strait of Georgia during the 1990's was about 50,000 to 60,000 mt and that biomass was stable during this time. Biomass estimates for the Port Susan population ranged from 10,648 mt in 1990 to 2,365 mt in 1999 (Table 4). Using these estimates, the Port Susan Pacific hake population comprised from 3.8-17.6% of the combined Port Susan-Strait of Georgia population during the 1990's. If the Canadian portion of the Strait of Georgia population is maintained, extinction of the Port Susan population does not appear to pose a serious risk of extinction for the entire Georgia Basin DPS. However, the Canadian portion of the DPS has shown some signs of decline in the late 1990s so the situation warrants continued close monitoring.

Saunders and McFarlane (1999) did not recommend formal changes in the range of yield recommended for the Canadian population. However, because of concern about factors such as decreasing size-at-age and increasing predation by pinnipeds they suggested "that managers choose from the lower half of the yield range," which was 7,554 to 14,687 mt. Saunders and McFarlane (1999) also estimated that harbor seals consumed 11,000 mt of Pacific hake in the Strait of Georgia in 1996, ranging from 4,400 to 21,000 mt. They qualified the consumption estimate by observing that age composition and distribution of harbor seals had changed considerably since composition of the diet estimates were made in the 1980's. Thus the estimate and ranges may not be accurate.

If harbor seals consumed 11,000 mt and commercial catch was 7,554 mt of Pacific hake and the biomass was 60,000 mt, then the total rate of exploitation would have been 0.31. Average estimated total rate of predation of the Port Susan Pacific hake population was lower under low hypothetical values of predation by pinnipeds during the time that the population declined from 14,826 mt in 1982 to 2,365 mt in 1999 (Table 13).

Environmental risks to the Georgia Basin Pacific hake DPS

The above analyses examined the possible effects of human and pinniped predation on the population of Pacific hake in the Georgia Basin DPS. As previously mentioned, environmental factors could have been very important factors in the observed decreases in biomass and size.

Changes in migratory behavior and location specific size at age of the offshore population of Pacific hake appear to be related to environmental factors (Dorn 1995). In the discussion that follows, temperature is referred to with the understanding that temperature is just one parameter of what is probably a complex suite of environmental factors that fish encounter. During warm years, a greater portion of the offshore Pacific hake population is found off Canada during the summer feeding season (Dorn 1995), and during the very warm period of the late 1990's some Pacific hake apparently spawned off Washington and Canada, which is much further north than the typical spawning area off California and Mexico (Dorn et al. 1999a). The Port Susan population apparently has changed more than the Canadian portion of the DPS. It is possible that warm environmental conditions have caused the Port Susan area to be relatively less favorable for Pacific hake spawning than the Canadian portion of the Strait of Georgia. Some of the Port Susan population may have migrated to Canadian waters, or perhaps there has been less movement from Canadian waters than before. The warm period may be part of global warming that has occurred during the last century. There is evidence that anthropogenic increases in atmospheric CO₂ may cause global warming. However there is still considerable scientific debate on whether or not the observed increases have natural or anthropogenic causes. Continuation or perhaps even enhancement of the warm conditions observed in the Pacific Northwest could preclude improvement in the condition of the Port Susan population of Pacific hake unless the fish eventually adapt to these conditions.

There may be other anthropogenic changes in the environment that have adversely affected Pacific hake. As previously noted (see "Environmental History and Features of Puget Sound" section) there have been changes in kelp and eel grass beds. While kelp and eel grass beds are not an important habitat for Pacific hake, it is possible that reduced beds result in reduced detritus for detrital feeders which may be important sources of food for Pacific hake in Puget Sound. Anthropogenic changes in river flow patterns and increased turbidity could possibly cause changes in the ecosystem that are adverse to Pacific hake. There have been insufficient studies to determine if there have been impacts from anthropogenic sources of toxic chemicals.

Summary and Conclusions of Georgia Basin Pacific Hake Risk Assessment

In its deliberations concerning ESA risk assessment for Pacific hake in the Georgia Basin DPS, the BRT considered the status of the Port Susan and Strait of Georgia stocks, the relationships among stocks, and effects of potential risk factors.

The BRT identified several concerns about the status of the Port Susan stock. Biomass and numbers of fish surveyed during the spawning period in Port Susan are the lowest since the surveys began in 1992. Estimated biomass in 2000 was 992 mt, about half the biomass in 1999 and represents an 85% decrease in the past 15 years. The size composition of the stock also showed a marked shift to smaller fish. Consequently, recruitment appeared to be maintained through 1999 despite declines in spawning biomass. Numbers of Pacific hake fluctuated around 30 million fish between 1985 and 1999, except in 1996 when estimated numbers exceeded 60 million fish. However in 2000, estimated numbers fell below 11 million. The size, and presumably age, at maturity has also dropped substantially. Nearly all female Pacific hake over 20 cm sampled during the 1990s were mature, whereas in the early 1980s, none were mature until 24 cm and about half of the sampled females were mature by 30 cm.

In addition to concerns about the status of the Port Susan stock, the BRT identified several areas of uncertainty. The extent of any mixing of spawners or spawning products among stocks within Puget Sound or between Puget Sound and Strait of Georgia stocks is not known. Unlike in Port Susan, the abundance of Pacific hake in the Strait of Georgia has not markedly declined over the past 15 years, and recruitment of young fish to the Port Susan stock may be the result of migration from the Strait of Georgia and other areas. If so, the Port Susan stock measured during the spawning period may be a variable portion of a larger stock and its size may not be indicative of the size of the larger stock. Under this hypothesis, the BRT's concerns about the low abundance of Pacific hake observed in Port Susan may be considerably reduced, but the BRT did not reach a consensus on the likelihood or extent of potential mixing among stocks.

The effects of potential risk factors, such as pinniped predation, habitat alteration or loss, and environmental changes, are also poorly known. Environmental changes could contribute to the observed changes in the status of Port Susan stocks, such as decreased growth, size at maturity, and reduced survival. The effect of pinniped predation or other risk factors that may be contributing to the decline in Port Susan Pacific hake abundance is also inconclusive. For two hypothetical models of pinniped predation that were considered, uncertainties about predation rates and behaviors precluded definitive conclusions about the risk of extinction of the Port Susan stock. Predation by other fish on Pacific hake or reductions in prey abundance have not been evaluated. The potential effects of habitat loss or degradation are not known, although West (1997) speculated that juvenile survival could be reduced through loss or degradation of nearshore nursery habitats.

In contrast to Port Susan, Pacific hake abundance in the Canadian portion of the Strait of Georgia apparently has been stable during the 1990s. Estimated biomass ranged between 50,000-60,000 mt, much larger than the Port Susan stock. The status of the Pacific hake in Dabob Bay, its relation to stocks in other areas, or the potential existence of undetected stocks are all unknown. Similarly, it is not known if the factors contributing to the decline in Port Susan could similarly affect the Strait of Georgia stocks in the near future.

These uncertainties and the differences in stock status between Strait of Georgia and Port Susan Pacific hake made evaluation of the status of the DPS difficult. The BRT concluded that the Georgia Basin Pacific hake DPS was not presently in danger of extinction, but could with nearly equal likelihood fall into either of two categories: 1) not in danger of extinction, nor likely to become so in the foreseeable future, or 2) not presently in danger of extinction, but likely to become so in the foreseeable future. As a whole, the BRT gave slightly higher support to the first category. Over the next year much new information is expected to become available that will likely resolve many of the uncertainties about the status and relationship of stocks of Pacific hake within the Georgia Basin DPS. When it is available, the BRT urges that this new information be considered and extinction risk be reevaluated.

^[2] W. Palsson, WDFW, 16018 Mill Creek Blvd., Mill Creek, WA 98012-1296. Pers. commun. to B. McCain.
^[3]M. Dorn, NMFS, F/AKC3, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to R. Gustafson.
^[4]M. Dorn, NMFS, F/AKC3, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to R. Gustafson
^[5] W. Palsson, WDFW, 16018 Mill Creek Blvd., Mill Creek, WA 98012-1296. Pers. commun. to W. Lenarz.
^[6] M. Dorn, NMFS, F/AKC3, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[7] W. Palsson, WDFW, 16018 Mill Creek Blvd., Mill Creek, WA 98012-1296. Pers. commun. to W. Lenarz.
^[8] W. Palsson, WDFW, 16018 Mill Creek Blvd., Mill Creek, WA 98012-1296. Pers. commun. to W. Lenarz.
^[9] P. Gearin, NMFS, F/AKC4, National Marine Mammal Laboratory, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[10] M. Dorn, NMFS, F/AKC3, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[11] W. Palsson, WDFW, 16018 Mill Creek Blvd., Mill Creek, WA 98012-1296. Pers. commun. to W. Lenarz.
^[12] M. Dorn, NMFS, F/AKC3, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[13] M. Dorn, NMFS, F/AKC3, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[14] W. Palsson, WDFW, 16018 Mill Creek Blvd., Mill Creek, WA 98012-1296. Pers. commun. to BRT, July 26, 2000.
^[15]M. Dorn, NMFS, F/AKC3, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz
^[16]M. Dorn, NMFS, F/AKC3, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[17] W. Palsson, WDFW, 16018 Mill Creek Blvd., Mill Creek, WA 98012-1296. Pers. commun. to W. Lenarz.
^[18]S. Jeffries, WDFW, 600 Capitol Way N., Olympia, WA 98501-1091. Pers. commun. to W. Lenarz.
^[19]P. Gearin, NMFS, F/AKC4, National Marine Mammal Laboratory, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[20] P. Gearin, NMFS, F/AKC4, National Marine Mammal Laboratory, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[21] J. Laake, NMFS, F/AKC4, National Marine Mammal Laboratory, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[22] P. Gearin, NMFS, F/AKC4, National Marine Mammal Laboratory, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[23] R. DeLong, NMFS, F/AKC4, National Marine Mammal Laboratory, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz
^[24]S. Jeffries, WDFW, 600 Capitol Way N., Olympia, WA 98501-1091. Pers. commun. to T. Builder. September 17, 1999.
^[25] P. Gearin, NMFS, F/AKC4, National Marine Mammal Laboratory, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.
^[26] R. DeLong, NMFS, F/AKC4, National Marine Mammal Laboratory, 7600 Sandpoint Way NE, Seattle, WA 98115-6349. Pers. commun. to W. Lenarz.

Back to Table of Contents