714 



Fishery Bulletin 103(4) 



ful in removing both roots and shoots in their entirety 

 (Peterson et al., 1983a). Shoots and roots, which were 

 collected in a 3-mm nylon mesh bag, were dried at 60°C 

 to constant weight to calculate total dry weight biomass 

 of seagrass. 



ANOVAs allowed us to test for a significant inter- 

 action between time (before versus after) and distur- 

 bance (dredge versus hand-harvest versus control) in 

 the biomass of seagrass and recruit density of bay 

 scallops (a basic BACI design; Green, 1979), indicative 

 of an impact of harvest. The cause of any significant 

 time x disturbance interactions was explored by using 

 Student-Newman-Keul (SNK) tests. Prior to each 

 analysis, Cochran's (1951) C-test was done to test for 

 heterogeneity of variances. Where variances were hetero- 

 geneous, data were In (x+1) transformed to remove 

 heteroscedasticity at a = 0.05. 



Results 



Of the two methods used to harvest adult scallops, hand 

 harvesting had by far the greater efficiency in these 

 shallow waters (ANOVA, P<0.0001). Over a period of 10 

 minutes, an average of 156 ±12 (1 SE) scallops within 

 each 25x8 m plot was harvested by hand as compared 

 to 26 ±1 scallops with the dredge. 



The two methods of harvesting differed significantly 

 in their impact on seagrass. Hand harvesting of scal- 

 lops did not increase dislodgement of seagrass above 

 the natural drift rate (Fig.l). Dredging, in contrast, 

 resulted in 127 times the export of seagrass. This ex- 

 traction did not, however, result in detectable reductions 

 in biomass per unit of area of seagrass within dredged 



plots when sampled one month later. There was no sig- 

 nificant temporal change in the biomass of seagrass in 

 any of the three treatments from before to one month 

 after harvesting (Table 1, Fig. 2). 



Fewer than 2% of the estimated total number of juve- 

 nile scallops in a plot were directly removed by dredg- 

 ing and none was removed by hand-harvesting. Never- 

 theless, sampling one month after harvesting indicated 

 depressed densities of juvenile bay scallops in dredged 

 plots (Table 2; Fig. 3). This difference could not be at- 

 tributed to natural change; small increases (16-55%) in 

 numbers of juvenile bay scallops in the hand-harvested 

 and control plots were documented over the same period 

 (Fig. 3). A comparison of size-frequency histograms of 

 juvenile bay scallops within each type of plot from be- 

 fore to after harvesting revealed that the decrease in ju- 

 venile scallop numbers in the dredged plots was primar- 

 ily due to losses of scallops in the smallest size classes 

 (<14 mm; Fig. 4). In the dredged plots, mean (±SE) 

 size of juveniles (<40 mm in shell height) increased 

 from 17.04 ±0.83 in January to 20.43 ±0.76 in February. 

 Over the same time period, mean size changed little in 

 the control (16.09 ±0.85 to 16.75 ±0.75 mm) or in the 

 hand-harvested (18.19 ±0.85 to 17.95 ±0.65 mm) plots. 



Discussion 



Previous research indicates that the implementation of 

 certain gear restrictions on estuarine bivalve fisheries 

 can minimize habitat destruction without sacrificing 

 harvesting efficiency (Peterson et al., 1983b; Lenihan 

 and Peterson, 2004). In our study, which successfully 

 mimicked the efficiency of commercial dredging and 



