Hamilton and Konar: Implications of substrate complexity and kelp variability for Alaskan nearsfiore fisfi communities 



193 



Table 1 



Partial correlation analysis (Pearson's correlation coefficient, r) between physical variables and kelp groups (n=473 transects). 

 Values are considered significant at P < 0.05. Rugosity provides a small-scale measure of habitat complexity and is the ratio of the 

 true distance contour along the bottom to a one-meter horizontal distance. Substrate size is based on the average diameter of the 

 substrate comprising the majority of a 0.25-m2 quadrat. Verticality was assigned to each site on the basis of overall vertical relief. 



Kelp group 



Canopy 



Annual understory kelp 



Perennial understory kelp 



were composed of much fewer individuals (at 

 most, tens of individuals/school). These juvenile 

 codfishes were not included in any analyses in 

 our study because of difficulties in accurately 

 quantifying them. 



Discussion 



Structural habitat complexity is important to 

 both fish and kelps in that greater physical 

 habitat complexity is associated with greater 

 overall densities of fish in these communities. 

 In particular, greenlings associated most consis- 

 tently with kelp beds that had a predominately 

 rocky (i.e., large cobble/bedrock) and structur- 

 ally complex bottom habitat. The association 

 of the dominant fish species in our study with 

 relatively larger substrate and higher rock relief 

 may indicate that south-central Alaska kelp-bed 

 fishes follow the same trend (of a strong associa- 

 tion offish with kelp and rocky substrate relief) 

 documented elsewhere. For example, in Cali- 

 fornia kelp beds, significant correlations exist 

 between fish density and bottom relief (Ebeling 

 et al„ 1980; Bodkin. 1986). However, rockfishes 

 in Puget Sound inhabited low-relief rocky kelp 

 beds during summer (Matthews. 1990) and in 

 Prince William Sound were positively associ- 

 ated with relief (Dean et al., 2000). Rockfishes 

 and codfishes in the present study were never 

 associated with any bottom structure, perhaps 

 because of the sporadic sightings of these species. The 

 lack of association of rockfishes and codfishes with any 

 physical habitat variables or kelp may reflect the tran- 

 sient nature and seasonal association of these fishes with 

 kelp. However, the rarity of any fishes observed higher 

 than one meter above the substratum in the present 

 study indicates that bottom structure may be important 

 to the fishes observed in our study, if perhaps indirectly, 

 by also being appropriate substratum for kelp habitat. 

 The variability of Alaska kelp communities and as- 

 sociated fish populations may be partially attributable 



Figure 3 



Monthly variation in the density (no. of plants/m^ ±standard error) 

 of understory kelp groups, n = 503 transects. Because surveys 

 were conducted during alternate neap tide cycles throughout the 

 17 month study, two surveys occurred, at least partly, during 

 October 2002. 



to the extreme seasonal nature of the northern environ- 

 ment. Increased sightings of fish during periods charac- 

 terized by warmer water indicate seasonal variability 

 in fish communities associated with kelp communities. 

 For example, although greenlings were observed in the 

 shallow, rocky nearshore sites during every month in 

 the present study, rockfishes and codfishes were rarely 

 observed except in summer. Healthy understory kelp 

 populations exist in this area on rocky substratum at 

 depths of up to 16 m (first author, personal observ.) 

 and fish populations may shift seasonally to similar 



