Felley and Vecchione: Nekton habitat on the continental slope of North Carolina 



271 



Although the various areas appeared qualitatively 

 different, analyses of species' habitat preferences 

 showed us that species in both areas were distrib- 

 uted in relation to similar types of benthic features. 

 Some forms were typically seen in areas with dense 

 aggregations of holes and mounds; others were seen 

 in areas with fewer holes and mounds. Thus, at both 

 localities, nekton distributions were related to par- 

 ticular types of infaunal assemblages. Associations 

 with particular epifaunal assemblages were demon- 

 strated by those species found most often in dense 

 patches of small anemones and (at Cape Hatteras) 

 in areas with dense patches of large anemones or 

 crinoids. Analysis of species preferences suggested 

 that in different areas on the continental slope, nek- 

 ton species respond to similar sets of environmental 

 variables. 



While analysis of species preferences identified 

 broad patterns of habitat selection for a group of spe- 

 cies, analysis of species variances allowed identifi- 

 cation of habitat selection by individual species. Ac- 

 tive habitat selection (occurrence of a species in a 

 subset of available environments) was demonstrated 

 for 4 of the 13 species included in the Cape Hatteras 

 analysis. While habitat selection according to factors 

 1 and 2 was shown for only one species each, habitat 

 selection according to density of mounds and grass 

 detritus (factor 3) was shown for three species (hake, 

 lizardfish, sergestid shrimp). Note that sergestid 

 shrimp distribution was related to benthic features, 

 although shrimp were always seen hovering 1 to 2 m 

 above the bottom. Such a seemingly counterintuitive 

 result ( a species selecting habitat according to a vari- 

 able that it does not seem to monitor) suggests the 

 need for more study in this environment. 



For most species at Cape Lookout and for several 

 at Cape Hatteras, small sample sizes made tests of 

 variance equality quite weak. An uncommon species 

 restricted to a particular habitat might be sampled 

 in numbers too low to allow a statistically signifi- 

 cant test. Examples abound in this study. Our analy- 

 ses identified various environmental gradients. Lo- 

 cations with scores at the extremes of the gradients 

 were sampled in low numbers (e.g. Fig. 2). Forms 

 characteristic of these extreme environments were 

 sampled in correspondingly low numbers, unless they 

 happened to be extremely abundant in a few inter- 

 vals. Thus, we found it most difficult to show statis- 

 tically significant habitat selection for those species 

 that were tightly clustered in specific habitat types. 



Some examples of uncommon forms found in spe- 

 cific habitat types came from dive 2629 (Table 3). 

 The lizardfish was most abundant in this dive, and 

 Species A and the only eels observed on the upper 

 slope were visible during this dive. The dive area had 

 a high diversity of habitats. The submersible passed 

 over areas with many holes and few anemones or 

 other epifauna, areas dense with small anemones, a 

 dense patch of crinoids (from 5 to 50 individuals per 

 interval over 8 intervals), and an aggregation of large 

 anemones (from 4 to 12 individuals over 18 inter- 

 vals). The extreme scores of Species A, cancroid crab, 

 and lizardfish were strongly affected by their high 

 densities in particular habitats seen in dive 2629. 

 Detailed studies of such heterogeneous areas could 

 clarify the environmental variables used by slope 

 species to select their habitats. 



In this study, we saw patterns of habitat choice by 

 nekton at different scales, from differences along the 

 slope, to species preferences for different habitats 



