progressively larger sample sizes of 4, 8, 16, and 28 cores. The average 

 number of species per core and the 95-percent confidence limits changed 

 little with increasing sample size (Table A-1). 



Thus, within relatively narrow confidence limits, each core is expected 

 to contain about 34 different species. Figure A-2 shows that 8 cores would 

 be necessary to accumulate about 90 species (75 percent of those present 

 in 28 cores) . 



2. Number of Individuals. 



In Table A-1, the common species have been ranked and three main groups 

 delineated on the basis of abundance. Means per core and their 95-percent 

 confidence limits are listed for the four sample sizes. The 20 species are 

 distinguished by the following characteristics: (a) They contain 87 per- 

 cent of the total number of individuals of all species (by group: 90-per- 

 cent mollusks, 90-percent crustaceans, and 87-percent worms); (b) variance 

 to mean ratios are greater than 1 in all cases and less than 7 with the 

 exception of P. civrifeva and Mediomastus aati-fomiensis ; and (c) the 

 mean abundance per core is greater than 1.5. The lower cutoff point in 

 Table A-1 is somewhat arbitrary, since abundance decreases gradually and 

 there is no clear-cut distinction between the last species included and the 

 first one excluded. Dominance of a few species can be seen by the fact 

 that 90 percent of the individuals are contained in about 20 percent of the 

 species. For most of the species, the confidence limits decrease sharply 

 from 4 to 8 cores and continued to decrease gradually to 28 cores. Excep- 

 tions are some of the species with the highest variance to mean ratios. 



Table A-1 also shows changes in the estimation of the total number of 

 individuals per core and individuals of major groups per core. 



3. Species Diversity and Evenness . 



Although this study was not concerned with estimating absolute species 

 diversity of the community (Brillouin, 1965; Pielou, 1966a), the Shannon 

 and Weaver (1963) formula was still used as the measure of diversity and 

 J (Pielou, 1966b) for evenness. The use of the parameters in a strictly 

 relative sense justified using these equations. 



Cumulative species diversity and evenness values are plotted in Figure 

 A- 3 from an average of five random orders of the 28 cores; diversity and 

 evenness appeared to stabilize between 4 and 8 cores. 



4. Choice of the Number of Replicates . 



In chosing the number of replicates taken at each station, the pract- 

 ical limitations of time and resources, the nature of the research objec- 

 tives, and the results of this analysis were considered. Further, as 

 mentioned previously, sampling was confined to permanently marked stations' 

 of a limited and well-defined area. 



69 



