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



265 



Fourth, we investigated habitat selection by indi- 

 vidual species by comparing species' variances on 

 each factor with variance of the environment. Envi- 

 ronmental variance on a factor was determined by 

 calculating factor scores for each 1-minute interval. 

 First, the value for each environmental variable in a 

 1-minute interval was standardized by using the 

 appropriate "mean of means" and "standard devia- 

 tion of means" noted above. Then the scoring func- 

 tion was applied to each 1-minute sampling unit. 

 Environmental variance was then determined as the 

 variance of sampling unit scores. The score of a 1- 

 minute interval was assigned to all individuals of 

 all species seen in the interval. A species' variance 

 was then calculated for each species as the variance 

 of these scores. The procedure of investigating both 

 species and locality scores on multivariate axes cor- 

 responds to Rotenberry and Wiens' 4 "synthetic ap- 

 proach" to the study of communities. 



We compared a species' variance with environmen- 

 tal variance for each factor by using Levene's test 

 (Levene, 1960; Van Valen, 1978). Bonferroni correc- 

 tions for multiple statistical tests were made by us- 

 ing Rice's (1989) method for investigating tables of 

 statistical test results. Rice's method is a correction 

 for inflated type-I error in situations where several 



4 Rotenberry, J. T., and J. A. Wiens. 1981. A synthetic approach 

 to principal component analysis of bird/habitat relationships. 

 In D. E. Capen (ed.), The use of multivariate statistics in stud- 

 ies of wildlife habitat, p. 197-208. USDA Forest Serv. Gen. Tech. 

 Rep. RM-87, Rocky Mountain Forest and Range Experiment 

 Station, Fort Collins, CO. 



different tests of significance are made for a particu- 

 lar null hypothesis. Such a series of tests constitute 

 a "table of statistical tests." In this study, a "table" 

 was considered to be all significance tests made rela- 

 tive to a factor, the corresponding null hypothesis 

 being "species variances are not significantly differ- 

 ent from environmental variance with respect to this 

 factor." Active habitat selection by a species was in- 

 ferred when a species' variance was significantly 

 smaller than the observed environmental variance 

 (1-tailed test). This implies that the species was ac- 

 tively selecting a subset of the available environment 

 with respect to that factor. See Felley and Felley 

 ( 1987) and Felley et al. ( 1989) for more details. 



Results 



Environments and biota 



Cape Hatteras — Table 2 presents means of environ- 

 mental variables for flat areas traversed by dives on 

 the slope off Cape Hatteras. Holes and mounds were 

 common environmental features: one to several holes 

 and mounds were in view at almost all times. In gen- 

 eral, intervals where holes were dense also had a 

 large number of mounds. Tubes were variable in oc- 

 currence. Many were seen in dive 2627 but relatively 

 few were seen in dives 2629 and 2630. Grass detri- 

 tus was very common in dives 2627, 2629, and 2630, 

 occurring in almost every interval. Sargassum de- 

 tritus was infrequent in upper slope intervals (dives 



Table 2 



Environmental variables measured on each interval, with means and standard deviations (in parentheses) for each dive at Cape 

 Hatteras and Cape Lookout, North Carolina. Holes, mounds, and tubes were coded as follows: 0=none in the interval; l=no more 

 than 1 or 2 seen in an interval; 2=always 1 or 2 visible throughout an interval; 3=several always visible, but countable; and 4=too 

 many visible to count. Gastropod/echinoderm tracks, and grass and sargassum detritus were coded 1/0 for presence/absence in 

 the interval (only percentages are reported for these variables). Number of individuals were counted for small anemones, large 

 anemones, gastropods, and crinoids. 



Environmental variables 



Holes 



Mounds 



Grass Sargassum Small 

 Tubes Tracks detritus detritus anemones 



Large 

 anemones Crinoids Gastropods 



Cape Hatteras 



2623 

 2627 

 2629 

 2630 



Cape Lookout 



2619 

 2620 

 2621 



61 

 35 

 41 



7 



36 



7 

 32 



1.75 (0.830) 

 2.63 (0.490) 

 2.51 (0.506) 

 2.29 (0.756) 



3.36(0.529) 

 3.28 (0.488) 

 3.25 (0.440) 



1.36 (0.484) 

 1.66 (0.482) 

 1.78 (0.571) 

 1.14(0.378) 



2.22 10.485) 

 2.14 (0.378) 

 2.25 (0.508) 



2.41(1.321) 

 2.23 (0.942) 

 0.73 (0.633) 

 0.57(0.535) 



0.25 (0.439) 

 0.00 (-) 

 0.25 (0.440) 



100.0 

 45.7 

 85.4 

 14.3 



16.7 

 28.6 

 53.1 



47.5 

 88.6 

 82.9 

 100.0 



69.4 

 14.3 

 62.5 



18.0 6.87(6.711) 0.03(0.180) 5.90(15.367) 0.31(0.564) 



8.6 25.80(19.954) 0.17(0.382) 0.00 (-) 0.40(0.976) 



4.9 1.37(1.577) 4.63(3.006) 4.66(13.190) 0.27(0.449) 



0.0 0.00 (-) 0.00 (-) 0.00 (-) 0.00 (-) 



38.9 0.33 (0.676) 

 28.6 0.14 (0.378) 

 40.6 0.41 (0.560) 



0.53 (0.629) 

 0.43 (0.534) 

 0.44 (0.619) 



0.00 (-) 

 0.00 (-1 

 0.00 (-) 



0.03(0.167) 

 0.00 1 -) 

 0.00 (-) 



