PETERSON ET AL.: IMPACT OF MECHANICAL CLAM HARVESTING 



which together formed a group of low-density bay 

 scallop matrices. About 84% of the variance in bay 



Table 7— For each habitat, total numbers of individuals found in 

 four randomly chosen cores from each of the two controls and the 

 two intense-kicking matrices on two dates, one before and one after 

 clam-harvest treatment. All species with total counts greater than 

 two are listed separately. 



scallop densities in fall 1983 is explained by seagrass 

 biomass in a simple linear regression. Figure 5 pre- 

 sents the relationship between average seagrass 

 biomass and bay scallop densities on a 1,225 m^ 

 scale, which suggests that the 2 intense-kicking 

 matrices contained even fewer bay scallops than 

 predicted from their reduced seagrass biomass. This 

 is similarly illustrated from calculations of the mean 

 numbers of bay scallops per 100 g of seagrass in each 

 matrix in fall 1983: control I (5.7), control II (5.1), 

 raking (4.7), light kicking (4.3), intense-kicking I 

 (2.0), and intense-kicking II (2.3). 



'These include molluscs, an amphipod, and additional polychaetes. 



10 20 30 40 50 60 



Av. seagrass dry moss (g/0.25m^) 



Figure 5.— Relationship between the average density of bay 

 scallops, Argopecten irradians, and the average biomass of 

 seagrass in fall 1983 samplings of each control and treatment 

 matrix of the clam harvest experiment in the seagrass matrix. 

 Clam harvest treatments had been applied in spring 1980 and again 

 in winter 1980-81. 



Table 8.— The effect of clam harvesting in the seagrass habitat on average bay scallop, Argopecten irradians, den- 

 sity per V4 m^ (±SE). Sample sizes per treatment matrix were 36 in fall 1980 and fall 1981 and 9 in spring 1981, 

 fall 1982, and fall 1983. No data are presented for the sand flat because of the rarity of bay scallops in that habitat. 

 Superscripts A-C indicate significant differences among matrices in Duncan's test at o = 0.05, with those means 

 sharing capital letter superscripts not differing significantly. 



1 '• - P < 0.01, *•* - P < 0.001, NS - P > 0.05 in one-way ANOVA on log (x + 1 )-transformed sample counts, comparing matrix 

 means on each date and in a two-way ANOVA over all dates (in the sums column). 



293 



