12. 



Statistical techniques 



2 

 The species, their abundance and the m viewed were summed 



for consecutive 100 meter depth intervals for each camera run. 



This data was also standardized by calculating the number of 



individuals per 100 m^ , for each species and for the total 



fauna. The relative percent that each species contributed to 



the total fauna of that depth interval was also calculated. 



Additionally, the percent coverage of the different substrate 



types was calculated for each depth interval. 



Two methods of community analysis were used, hierarchial 



classification and principal components ordination. The 



hierarchial classification was based on overall species 



similarity between samples (depth intervals) . These samples 



were then consecutively combined into larger groups using the 



"unweighted" pair-group method (Sokal and Sneath, 19 63) . In 



this technique equal weight is assigned to each sample in a 



group, giving larger groups more weight than smaller ones. 



The species similarity index used was percent similarity 



(Whittaker and Fairbanks, 195 8) given by the formula: 



PS = 100 (1.0 - 0.5 21 p - P I ) = 100 min2(p , p ) 



ia ib ia ib 



where p is the proportion of sample "a" composed by species 

 ia 



"i" and p is the same for sample "b". These values were 

 ib 



