Doyle et al.: Neustonic ichthyoplankton in the western Gulf of Alaska 



243 



and Zaprora silenus was 10-20 mm. Hexagrammos 

 stelleri, Pleurogrammus monopterygius, and Hemil- 

 epidotus jordani larvae were larger with predomi- 

 nant length ranges of 15-40 mm, 15-25 mm, and 

 15-30 mm, respectively. In contrast, larval sizes for 

 the cottids, Hemilepidotus spinosus and Hemil- 

 epidotus spp., were relatively small with a predomi- 

 nant length range of 5-15 mm. 



Daytime catches of larvae in the neuston were suf- 

 ficient to make diel comparisons in length distribu- 

 tions for only three of the dominant taxa. There was 

 no significant day-night difference in the length dis- 

 tribution of Hexagrammos decagrammus larvae (K- 

 S test; Z=0.07, P>0.05). Theragra chalcogramma lar- 

 vae caught at night (median length=6 mm) were 

 slightly, but significantly, larger (K-S test; Z=1.52, 

 P=0.02) than those caught during the day (median 

 length=5 mm). Day-night differences were much 

 greater for Ammodytes hexapterus larvae for which 

 the median length caught at night was 24 mm and 

 the median day length was 13 mm (K-S test; Z=4.99, 

 P<0.001). Migration of the larger larvae and juve- 

 niles to the surface at night may have been the cause 

 of this difference, but it is also likely that enhanced 

 sampler avoidance during daylight by large larvae 

 and juveniles reduced the daytime median larval 

 length significantly. 



Horizontal patterns of distribution 



Patterns of distribution illustrated here for total and 

 individual dominant taxa of neustonic larvae were 

 based on data combined for all cruises. The distribu- 

 tion maps therefore represent general patterns of 

 horizontal distribution for these species during spring 

 in this region and did not take into account day-night, 

 monthly, or interannual differences in catches. 



The pattern for total fish larvae in the neuston 

 indicated that highest concentrations generally oc- 

 curred to the southwest of Kodiak Island, in Shelikof 

 Strait, and off the northern tip of Kodiak Island (Fig. 

 6A). Southwest of the Shumagin Islands and north- 

 east of Kodiak Island, high densities of larvae were 

 more scattered. Despite the high intensity of sam- 

 pling seaward of Kodiak Island (Fig. 2), mean larval 

 concentrations tended to be low in this region. Based 

 on data which incorporated sampling during all sea- 

 sons, Kendall and Dunn (1985) and Rugen 3 fre- 

 quently recorded high concentrations of various spe- 

 cies of larvae in the neuston seaward of Kodiak Is- 

 land. The apparent scarcity of larvae here may there- 

 fore be characteristic of spring in the sampling area. 



Larvae of the osmerid Mallotus villosus were taken 

 primarily southwest of Kodiak Island along the 

 Alaska Peninsula as far southwest as the Shumagin 



Islands (Fig. 6b). They were scarce southwest of the 

 Shumagin Islands and seaward of Kodiak Island and 

 absent in the northeastern part of the sampling area. 

 This pattern is similar to that described by Rugen 3 

 except that the latter study plus Kendall and Dunn's 

 (1985) observations indicated a greater presence of 

 larvae seaward of Kodiak Island. These studies also 

 showed that M. villotus larvae were relatively scarce 

 both in bongo and neuston samples during spring; 

 the main spawning season seems to be late summer 

 through fall (Kendall and Dunn, 1985). 



Theragra chalcogramma larvae were usually most 

 abundant in the upper 50 m of the water column in 

 the southern Shelikof Strait area and along the 

 Alaska Peninsula during spring (Schumacher and 

 Kendall, 1991). Spawning takes place primarily in 

 the sea valley in Shelikof Strait during late March 

 and early April. Rugen 3 has also documented the 

 occurrence of pollock larvae on occasions in large 

 concentrations to the northeast of Kodiak Island. 

 These patterns were reflected in the distribution of 

 pollock larvae in the neuston documented during the 

 present study (Fig. 6C ). The scarcity of larvae within 

 Shelikof Strait may have been due to the low num- 

 ber of samples from that region. Pollock larvae were 

 absent or scarce along the outer shelf and slope indi- 

 cating that most of the larvae in the surface zone 

 were retained on the shelf. 



Neustonic larvae of Anoplopoma fimbria were most 

 abundant during late spring and summer in the west- 

 ern Gulf of Alaska where they were associated with 

 the shelf edge (Kendall and Dunn, 1985; Rugen 3 ). 

 The general distribution pattern documented here 

 for the spring months showed them to be most abun- 

 dant close to the shelf edge southwest and northeast 

 of Kodiak Island, as well as around the northern and 

 northwestern perimeter of Kodiak Island (Fig. 6D). 

 As with pollock, this species is a pelagic spawner in 

 deep water, and the distribution pattern of larvae 

 suggested that spawning occurred mainly in outer 

 shelf and slope waters, a pattern which is consistent 

 with what is known about the early life history of this 

 species (Kendall and Matarese, 1987; Doyle, 1992). 



The dominant hexagrammid species whose larvae 

 were abundant in the neuston of the sampling area 

 all spawn in coastal waters (Matarese et al., 1989). 

 In the Gulf of Alaska region, spawning of these spe- 

 cies seems to occur from fall through spring ( Kendall 

 and Dunn, 1985; Rugen 3 ). Larvae of the most abun- 

 dant species, Hexagrammos decagrammus, were 

 found to be dominant in the neuston during most of 

 the year; during the summer months there was a 

 large decrease in density. They were distributed 

 widely throughout the sampling area, but greater 

 concentrations were found in the southwestern re- 



