178 



Fishery Bulletin 104(2) 



From the analysis of spatial associations between 

 fishes and large, individual structure-forming inver- 

 tebrates, six of 108 species were found more often ad- 

 jacent to colonies than predicted by their abundance 

 along transects. This result indicates that there may 

 be spatial associations that do not necessarily include 

 physical contact with the sponges and corals. However, 

 the median distances between these six fish species and 

 large invertebrates (1.0-5.5 m) were not particularly 

 small. Thus, it is likely that these fishes and inverte- 

 brates are present in the same types of habitats and 

 that there is not necessarily a functional relationship 

 between these two groups of organisms. 



Parrish (2004) reached similar conclusions on stud- 

 ies of black coral in Hawaii. Although fish densities 

 were higher in areas that included corals, when bot- 

 tom relief and depth were accounted for these densi- 

 ties were not higher than those for surrounding areas 

 without corals. Thus, there was no clear evidence that 

 corals served to aggregate fish. Rather, fishes and cor- 

 als co-occurred in areas with similar physical relief and 

 unique flow regime (Parrish, 2004). Auster (2005) also 

 reached similar conclusions by finding no significant 

 difference in the density of a common rockfish species 

 iSebastes fasiatus) between areas of rock and boulders 

 with dense coral cover and similar areas having dense 

 epifaunal cover (i.e., without coral). Auster concluded 

 that although dense coral and dense epifaunal habitats 

 were functionally equivalent, the epifaunal habitat was 

 more widespread in his study area, making that habitat 

 perhaps more important to the greater rockfish popu- 

 lation. Finally, Syms and Jones (2001) demonstrated 

 that removal of high densities of soft corals caused no 

 significant changes in the associated fish communities 

 and that the heterogeneity of habitat generated by soft 

 corals was indistinguishable from equivalent habitat 

 formed by rock alone. Thus, fish-invertebrate associa- 

 tions, by themselves, do not necessarily demonstrate 

 the functional importance of invertebrates as habitat 

 to benthic fish populations. 



One possible conclusion from our study is that ob- 

 served fish-invertebrate associations, like those reported 

 for many cold-water corals, can be overstated. In the 

 absence of quantitative information, observers may re- 

 member the few positive associations between fishes and 

 structure-forming invertebrates but ignore (or forget) 

 the more numerous observations of large invertebrates 

 with no associated fishes. Indeed, the general impression 

 of the authors after making submersible observations 

 was that there were higher numbers of fishes associated 

 with large invertebrates when in reality only five fishes 

 were observed lying directly on a large invertebrate in 

 the video transects. A more likely conclusion, however, 

 is that the continental shelf communities of southern 

 California are unique and that large black corals do 

 not have the high number of commensals as seen, for 

 example, on Primnoa in Alaska. An additional consider- 

 ation is the relatively low number and size of individual 

 sponges, gorgonians, and black corals observed in this 

 study. Primnoa in Alaska can form massive stands 3 m 



20 



10 







200 



100 







3000 



1500 







CO 



■g 



> 1000 



"D 



g 500 

 o 



d 

 z 



100 



50 







400 

 



800 



400 







Gorgonians 



X. 



15 20 25 30 40 



Black corals 



10 20 30 40 50 60 70 80 90 100 

 "=134 Shelf sponges 



10 20 30 40 50 60 70 80 90 100 

 n=i.262 Foliose sponges 



10 20 30 40 50 60 70 80 90 100 

 Height (cm) 



Figure 9 



Size distributions of large, structure-forming inver- 

 tebrates. Arrows indicate ma.ximum sizes observed in 

 each group. 



tall and 7 m wide in some areas (Krieger and Wing, 

 2002). Moreover, the majority of fishes were observed 

 on the largest individuals in their study (>15 m^ in 

 volume) Most of the corals in our study were <0.5 m in 

 height. Given the importance of this issue, we argue for 

 more rigorous quantitative studies on fish-invertebrate 

 associations that would include densities of fishes and 

 sizes of both fishes and invertebrates. 



Regardless of their associations with fishes, the struc- 

 ture-forming invertebrates described in this study are 

 very likely to be ecologically important on continental 

 shelf ecosystems and are certainly significant in their 

 own right. Observation on the health of the larger in- 

 vertebrates indicates few damaged (0.1%) or dead (0.2%) 

 individuals and a low incidence of fishing gear in the 

 areas surveyed (Tissot et al., unpubl, data). These ob- 

 servations are consistent with the absence of a signifi- 

 cant commercial bottom trawling fishery in our survey 

 area, which has been associated with negative impacts 

 on large invertebrates in other locations (Watling and 

 Norse, 1998; Freese et al. 1999: Krieger. 2001). Thus, 

 this study affords a unique view of what appears to be 



