Wilson: Effects of year and region on abundance and size of Theragra chakogiamma 



831 



tion near shore, and possible directed movements by juve- 

 niles all complicate "typical" southwestward dispersal of 

 pollock eggs and larvae from Shelikof Strait. For example, 

 the high age-0 densities around Kodiak Island during 1985 

 may be explained by high larval abundances upstream of 

 the main spawning area (Kendall and Picquelle, 1990), 

 which in turn may reflect northeastward transport of 

 eggs at depth in the countercurrent. On the other hand, 

 the high age-0 densities around Kodiak during 1988 are 

 more difficult to explain. Countercurrent flows have been 

 detected off the Gulf-side of Kodiak Island (Musgrave et 

 al., 1992) but historically this is not an area of high egg or 

 larval abundance. Furthermore, widespread occurrences 

 of high age-0 density suggest that dispersal was both 

 downstream and upstream, or that some fish originated 

 from elsewhere. 



The existence of multiple spawning areas further com- 

 plicates our understanding of pollock recruitment path- 

 ways. Sporadic sui-vey effort by the NMFS indicates 

 that spawning occurs at a number of different locations 

 throughout the western GOA between Prince William 

 Sound and Unimak Pass. The percentage of pollock bio- 

 mass in non-Shelikof spawning areas was 10, 21, 19, 29, 

 and 14'7f of the total in all areas sui-veyed in 1989, 1990, 

 1994-96, respectively (Williamson-^; Karp''; Wilson et al.", 

 and references therein). These estimates are conservative 

 because not all known spawning areas were surveyed each 

 year and survey timing may not have coincided with 

 peak spawning. In addition, other surveys of pollock, sum- 

 marized by Bechtol,^ indicate that the biomass of pre- 

 spawning aggi'egations in Prince William Sound ranged 

 from 28,85.5 to 114,344 metric tons during 1995, 1997, 

 and 1998. Although these surveys occurred after 1988 

 and conditions may have changed during the intervening 

 period, it seems likely that at least some age-0 pollock 

 were spawned in areas outside of Shelikof Strait. Esti- 

 mating relative production, however, is difficult because 

 it is impossible to identify where individual recruits were 

 spawned and because of spatial variations in the pro- 

 cesses that remove prerecruits from the GOA. As indi- 

 cated by studies of other species (Pulliam and Danielson, 

 1991; Frank, 1992), the relationships among these differ- 

 ent spawning groups could be very complex. 



The use of age-0 densities in the Kodiak region in fore- 

 casting recruitment suggests that "atypical" dispersal of 

 young from Shelikof Strait and production of fish from 

 other spawning areas may be relevant factors in under- 

 standing GOA pollock recruitment. Admittedly, this use 

 is based on only a few years of data. However. Smith et 

 al. (1984) reported age-0 densities for the 1980-82 year 

 classes in some bays within the Kodiak region that cor- 

 responded with year-class strength estimates. Further- 

 more, if sampling method is ignored, a time series of mean 



age-0 density near northeast Kodiak Island is available 

 for the 1980-88 year classes, except 1983, and it corre- 

 sponds reasonably well with recruitment (Pearson, r=0.74, 

 P=0.04, «=8) (author's unpubl. data). The Shumagin and 

 Unimak regions were comparatively less promising for 

 predicting recruitment, although the relative strength of 

 the 1985 and 1988 year classes was indeed evident among 

 the Shumagin densities. Age-0 fish from both the Kodiak 

 and Shumagin regions may move into Shelikof Strait 

 during winter as indicated by large aggregations of age- 1 

 fish observed there during early spring (McKelvey, 1996). 

 This movement probably involves a relatively constant 

 proportion of the Kodiak and Shumagin age-0 popula- 

 tions because their abundance as age-1 fish in Shelikof 

 Strait continues to be indicative of recruitment (McKelvey, 

 1996; Guttormsen and Wilson"'). Farther southwest, in 

 the Unimak region, juveniles may leave the GOA some- 

 time after their first summer. This movement may explain 

 why age-0 densities in the Unimak region during 1987 

 and 1988 had no relation to recruitment despite a 7-fold 

 variation in subsequent year-class strength. 



Given their relevance to recruitment, age-0 pollock 

 around Kodiak Island warrant closer consideration. As 

 previously discussed, some may come from Shelikof Strait. 

 Others may move downstream from spawning areas such 

 as those in or near Prince William Sound and Resurrection 

 Bay (Nelson and Nunnallee, 1985; Muter and Norcross, 

 1994; Norcross and Frandsen, 1996; Karp"). Downstream 

 drift is probable given the distance, currents (Schumacher 

 and Reed, 1980; Stabeno et al., 1995), spawning time 

 (Muter and Norcross, 1994), and estimated age-0 ages.^^ 

 Drift from eastern areas may be relatively important to 

 establishing age-0 populations along the upstream- and 

 Gulf-side of Kodiak Island when the influx of Shelikof fish 

 is low. Retention of young along this part of Kodiak Island 

 may be facilitated by shoreward flows and vortices cre- 

 ated by topographical influences on the prevailing current 

 (Lagerlof, 1983; Dunn et al.^). Local spawning is another 

 possible source of age-1 pollock as evidenced by adults in 

 prespawning condition in Marmot Bay, which is on north- 

 east Kodiak Island (Karp''; Williamson^-). 



The relatively large size of many age-0 fish may reflect 

 favorable conditions in the Kodiak Island vicinity because 

 size is generally regarded as having a positive effect on 

 survival. Walters et al. ( 1985 ) first reported a southwest-to- 

 northeast increase in age-0 size; however, a more rigorous 



Bechtol. W. R. 1998. Prince William Sound walleye pollock: 

 current assessment and 1999 management recommendations. 

 Regional Information Report No. 2A98-4 1,2.5 p. Alaska Depart- 

 ment of Fish and Game, Division of Commercial Fisheries Man- 

 agement and Development, .3.33 Raspberry Rd., Anchorage, AK 

 99.518. 



'" Guttormsen, M., and C.Wilson. 1998. Echo integration-trawl 

 survey results for walleye pollock in the Gulf of Alaska during 

 1998. In Stock assessment and fishery evaluation report, p. 

 509-530. Prepared by the Gulf of Alaska Groundfish Plan 

 Team, North Pacific Fishery Management Council. P.O. Box 

 103136. Anchorage, AK 99510. 



" Mean age of age-0 pollock in the Kodiak region during 1985-88 

 ranged from 127 to 143 d. These estimates are based on age-0 

 length and length-age relationships for the same age-0 popula- 

 tions (Bailey et al., 1996bi, to which mortality-adjusted density 

 weights were applied. 



'- Williamson, N. 1999. Personal commun. National Marine 

 Fisheries Service, 7600 Sand Point Way NE, Seattle, WA 

 98115. 



