of 26 taxa being identified in the photographs to date. In all 

 three areas the dominant invertebrates are the sabellid worm, 

 Myxicola infundibulum , burrowing anemones, Cerianthus sp . (most 

 likely C. borealis ) , and the sea scallop, Placopecten magel- 

 licanus . 



Scallop Density and Distribution 



Scallops occur in varying densities at the three sites. 

 They are most numerous on Fippennies Ledge where the mean density 

 is 0.98 + 2.19 scallops m" 2 . On Platts Bank the density drops to 

 0.31 + 0780 m~ 2 while on Jeffreys Ledge it is even lower at 0.21 

 + 0.50 m~ 2 . 



The actual density of scallops per photo frame (1.24 m 2 ) on 

 Fippennies Ledge (Figure 3) graphically demonstrates the extreme 

 variability in scallop densities over the course of three 

 transects. Chi-square analysis of this photographic data was 

 employed to compare the actual distribution of the population 

 with that of a calculated Poisson distribution. In all cases the 

 scallop distributions deviated significantly from the Poisson, 

 indicating that the scallop distribution was not random. A 

 negative binomial distribution could be fit to each of the data 

 sets and adequately described the populations at the three sites, 

 (Figure 4). These statistical techniques show that scallops are 

 not distributed randomly over the grounds, even at the lowest 

 density observed. In contrast to a random distribution, the 

 contagious distribution of the variants in our analysis indicates 

 that the scallops are located in clusters or patches over the 

 bottom. To further illustrate this, the data for Fippennies 

 Ledge and Jeffreys Ledge have been standardized to 500 photo- 

 graphs and the expected distribution of the number of animals 

 per frame calculated from the negative binomial and plotted in 

 Figure 5. In both cases the data follows a similar pattern with 

 a rapid decline in the numbers of pictures containing multiple 

 scallops. On Fippennies Ledge where the scallop density is the 

 highest, the distribution tails off much more gradually. 



Since the scallops show a clustered distribution the 

 question arises as to the effect of quadrat size on the analysis. 

 To evaluate this effect each photo from the highest density area 

 on Fippennies ledge was subdivided into approximately 1/4, 1/2 

 and full frame (0.33, 0.66, 1.24 m 2 ) and the number of scallops 

 counted per quadrat. Morisita's index of dispersion (Elliot, 

 1983) was then calculated for each of the data sets and the data 

 plotted in Figure 6. As quadrat size is decreased, Morisita's 

 index of dispersion should approach a value of 1. Such a value 

 indicates a random distribution, or in this case, a breakdown of 

 the contagious distribution actually observed in the 0.66 m and 

 1.24 m 2 quadrats. From this plot it can be inferred that 

 scallops are generally separated by a distance of at least 25 to 

 60cm. Observations from the submersible confirm that although 

 scallops are often clustered they do space themselves over the 

 bottom. Rarely were scallops observed close enough to actually 



247 



