968 



Fishery Bulletin 97(4), 1999 



Seasonal variation in the nearshore fish community 

 of Kachemak Bay 



To examine seasonal variation in nearshore fish 

 abundance, we used only data collected in 1996 

 ( 130,325 fish, 46 species, 283 beach seine sets in Feb- 

 ruary-December). Most species were present in 

 greatest numbers during summer months, appar- 

 ently-moving inshore as water temperatures in- 

 creased above 4°C in May (Fig. 2), and moving off- 

 shore as temperatures declined in October (Fig. 3). 

 Almost no fish were caught or observed during the 

 winter months (December-March; Fig. 3); however, 

 CPUE increased markedly for all species in June and 

 declined dramatically in mid-July. This pattern was 

 also reported by Blackburn et al. ( 1980; Fig. 4). This 

 change largely resulted from changes in nearshore 

 sand lance abundance but was compounded by move- 

 ments of pink salmon and Pacific cod. Large catches 

 of juvenile pink salmon increased the CPUE in June 

 as they migrated along shorelines (Orsi and 

 Landingham, 1985), but salmon CPUE declined rap- 

 idly in July as fish moved offshore (Fig. 3; Blackburn 

 et al. , 1980). Catches of gadids (particularly Pacific 

 cod and pollock) increased markedly in late summer 

 (Fig. 3) and contributed to the large August increase 

 in CPUE. However, most of the August-September 

 increase in CPUE was due to recruitment of first- 

 year sand lance (Fig. 3), which dominated late sum- 

 mer seine catches after adults moved out of nearshore 

 environments in late July (Blackburn et al. , 1980). 

 In October, adult sand lance returned to inshore 

 waters to spawn (Dick and Warnei', 1982; Robards, 



u 



E 



M A M J 



Figure 2 



Seasonal variation in 1996 sub.surface temperaturc-s (5 m 

 below low tide |0 m|) at Kachemak Bay i 1 ). Chisik Island 

 (2l, and the Barren Islands (3). 



1999). CPUE began declining rapidly in October, in 

 conjunction with a decline in diversity from 34 spe- 

 cies in August to only 3 by December. 



We also observed strong seasonal trends in use of 

 nearshore habitats by other species. Dolly Varden 

 moved into nearshore waters in April, where they 

 remained throughout July. After July they followed 

 salmon into natal freshwater systems to overwinter 

 (Isakson et al., 1971, Orsi and Landingham, 1985). 

 Juvenile rock sole {Pleuronectes bilineatus) and great 

 sculpins were consistently the most abundant fish 

 found in nearshore habitats during February-March. 

 Flatfish (Pleuronectidae) were caught at low rates 

 throughout spring and summer but were not present 

 after October (Fig. 3). Numbers of juvenile great 

 sculpins increased rapidly in spring owing to an in- 

 flux of small juveniles (<20 mm). Catches of these 

 juveniles fell markedly by late June, and small but 

 regular numbers of second-year individuals were 

 caught during summer. Only one capelin was caught 

 in Kachemak Bay (26 July 1995) prior to October 

 1996, when 1586 first-year fish were captured in 

 three seine sets. Several large capelin schools were 

 also observed at the same time. Capelin were also 

 captured in three of eight seine sets in December. 

 Although herring were present throughout the sum- 

 mer (Fig. 3), almost all of them (999< ) were captured 

 in five June and four August sets at one area (Hali- 

 but Cove ) historically known for its aggregations of 

 herring (Rounsefell, 1930). 



Species abundance, diversity (H'), and richness (D) 

 in the nearshore Kachemak Bay fish community in- 

 creased steadily from April to June, peaked in July, 

 and declined rapidly in September. Although sand 

 lance dominated the community in summer, more 

 than 30 species of fish were present during June- 

 August. Species evenness (J') declined throughout 

 summer, when sand lance dominated nearshore habi- 

 tats, but increased again in fall as numbers of spe- 

 cies and individuals declined. 



Variability among consecutive beach seine sets and 

 tidal states in Kachemak Bay 



To assess variability in catches among sets and tidal 

 states within Kachemak Bay, we made two sets im- 

 mediately adjacent to each other at each site during 

 consecutive high and low tides (i.e. 4 sets per site, 17 

 samples, 68 sets). CPUE declined markedly between 

 first and second sets at both high (42''/f decline) and 

 low (50'^ decline) tides, although numbers of species 

 caught remained similar (Table 4). Jaccard's simi- 

 larity coefficient for species composition indicated a 

 high degree of similarity between the two adjacent 

 high (75*^ ) and low (74^7^ ) tide sets. However, species 



