Core samples were emptied individually into 3.8-liter Ziploc 

 plastic storage bags, sealed underwater, and then brought to the 

 surface. In the laboratory, samples were emptied into 3.8-liter jugs 

 containing 10 percent seawater formalin stained with rose bengal. Core 

 samples were later washed through a 1-millimeter sieve, and the 

 organisms retained were preserved in 70 percent ethanol. All animals 

 were identified to the lowest taxon possible. Voucher specimens of all 

 species collected were deposited in the zoological museum at Florida 

 Atlantic University, Boca Raton, Florida. 



Significant differences in numbers of species and individuals 

 between stations for each sampling period were tested according to the 

 methods in Sokal and Rolf (1969a) and compared to the statistical 

 tables in Sokal and Rolf (1969b). An F-max test was run on the raw 

 data, which was found to be heterogeneous and required a square-root 

 transforraation; a two-way analysis of variance (ANOVA) with replication 

 was performed on the transformed data. A priori (F-test) and a 

 posteriori (Student-Newman-Keuls) significance tests were then run. 



Species diversity was calculated by the Shannon-Weaver index, H' , 

 with the aid of a Univac 1106 computer: 



H' = - E P- log P. 

 i=l ^ ^ 



where the probability that one individual belongs to species 1 is P., 



and p. is n./N, where n. is the number of individuals of the 



i species, and N the total number of individuals in the sample. 



The equitability component of diversity (Pielou, 1966) was 

 calculated as follows: 



e = H'/log S 



where S is the total number of species. 



Faunal similarity among samples was tested using Czekanowski' s 

 coefficient weighted for abundance. The computer program for this 

 analysis is described in Bloom, Santos, and Field (1977). This 

 coefficient is calculated as follows: 



C = 2W/(A+B) 

 z 



where A is the sura of the measures of all species in one sample, B the 

 similar sum for the second sample, and W the sum of the lesser measures 

 of each species for the two samples being compared (Field and McFarlane, 

 1968). 



A matrix of coefficients was obtained, group average sorting was 

 performed (as recommended by Field and McFarlane, 1968), and a 

 dendogram was prepared. 



