300 



Fishery Bulletin 102(2) 



fisherman, R. A. Cummings. Oysters and shell material col- 

 lected by dredges and tongs were separated aboard the boat 

 on a culling board, using the common culling techniques 

 (i.e., breaking apart oysters and shell with hammers, mal- 

 lets, and chisels). As mandated by law, oyster shell and 

 undersized oysters (<7 cm in height) were thrown overboard 

 above the reef from which they had been collected. 



Hand collections of oysters were conducted by scuba 

 divers (J. H. Grabowski and H. S. Lenihan). Unlike profes- 

 sional oyster divers in Chesapeake Bay and other areas, 

 who rake large quantities of shell and attached oysters 

 into baskets that are pulled aboard ship to be culled, the 

 divers in this trial adopted a different method designed to 

 preserve reef habitat. Instead of collecting shell and oys- 

 ters indiscriminately, the divers chose only those oysters 

 that appeared alive and of market-size. Selected oysters 

 were hand picked from the reef and placed in heavy plastic 

 mesh baskets that, when full, were subsequently pulled 

 aboard the boat with haul lines. 



To standardize fishing effort, each of the 12 harvested 

 reefs was harvested for 2 hours, regardless of the num- 

 ber of oysters collected. A 2-h harvest period for each 

 28.3-38.5 m 2 reef was considered to be a thorough but 

 not excessive level of harvesting by the professional 

 fishermen. The numbers of oysters collected in the final 

 three or four dredge hauls and oyster tongs were typically 

 lower (by -10-20%) than the preceding dredge hauls and 

 tongs. This reduction in the catch per unit of effort was 

 great enough that a fisherman foraging optimally would 

 normally cease harvesting at that time and move on to 

 another reef. Similarly, after 2 hours of diver-harvesting, 

 most of the clearly visible market-size oysters had been 

 harvested. 



Quantifying reef structure 



Measurements of oyster reef height and diameter were 

 conducted on all 16 experimental reefs both before and 

 after application of the three fishing methods. In Febru- 

 ary 1996, the preharvest height and radius of each oyster 

 reef were measured by scuba divers using a custom-made 

 "square angle," consisting of two pieces (2 m and 5 m 

 long) of 3-cm wide steel angle-iron, each with an attached 

 1-m long carpenter's level. Both pieces of angle iron were 

 marked at 1-cm intervals. The 5-m long (cross) piece was 

 attached to the 2-m long (upright) piece by a roller-joint. 

 The roller-joint allowed the cross piece to move up and 

 down the upright piece, thus providing a measure of reef 

 height, and to move horizontally in relation to the upright 

 piece, thus providing a measurement of reef radius. The 

 2-m long piece also had a 0.75-m long piece of angle iron 

 attached perpendicularly near its bottom so that it would 

 not sink into the seafloor when placed upright. 



One diver held the 2-m long angle iron perpendicular 

 to the seafloor at the edge of a reef, while the other diver 

 placed the 5-m long angle iron parallel to the seafloor, so 

 that one end rested on the highest point of a reef and the 

 other end met the upright angle iron at the reefs edge. 

 The height and radius of the reef were then measured 

 by recording the height at which the cross piece met the 



upright piece, and the distance at which the upright piece 

 met the cross-piece. For each reef, a mean diameter was 

 calculated by measuring three separate radii (oriented 

 at three compass bearings, all 120° apart), multiplying 

 the radii by two to estimate diameters, and then averag- 

 ing the three diameters. This averaging procedure was 

 undertaken because the reefs were not perfectly circular. 

 Measurements of reef height and radius were repeated in 

 March, two-five days after experimental harvests were 

 completed. 



Sampling oyster populations 



We sampled live and dead oysters on each treatment and 

 control reef before (late February 1996) and immediately 

 after (late March) experimental harvests to estimate 

 the proportion of oysters incidentally killed but not har- 

 vested by each harvesting treatment. Specifically, oyster 

 data was collected within 30 hours of the application of 

 the harvest treatment on each replicate reef. Densities of 

 live and dead oysters were quantified by divers who hap- 

 hazardly placed eight 0.5-m' 2 weighted PVC quadrats on 

 the reef surface at haphazard locations and recorded the 

 number of live and dead oysters greater >1 cm in height. 

 The density of dead oysters was measured by count- 

 ing the number of oyster shells that were articulated 

 and appeared relatively fresh (i.e., not black in color or 

 decayed), or oysters with somatic tissue exposed because 

 of cracked, broken, or punctured shells. Oysters with 

 exposed somatic tissue almost certainly die because of 

 predation by fishes and crabs in the Neuse River estuary 

 (Lenihan, 1999; and see Lenihan and Micheli, 2000). 

 Mean proportions of dead oysters were computed (dead 

 oysters/dead+alive oysters), as well as mean densities of 

 live and dead oysters on each reef. 



Catch per unit of effort 



The relative efficiency of each harvesting method was 

 determined by comparing the numbers of bushels (1 

 bushel=36.4 L) of market-size oysters taken per hour of 

 fishing. We quantified numbers of bushels for each har- 

 vesting method aboard the boat by placing oysters of legal 

 size in premeasured mesh baskets. After being counted, 

 and upon termination of the harvest trial, many of the 

 oysters were returned to other nearby reefs not involved 

 in the experiment. 



Statistics 



One-way analysis of variance (ANOVA) was used to com- 

 pare the following across harvest treatments and controls: 

 1) changes in mean reef height and diameter; 2) catch per 

 unit of effort; 3) the proportion of oysters found dead on 

 reefs before harvest; 4) the proportion of oysters found 

 dead on reefs after harvest; and 5) the absolute difference 

 in the proportion of oysters found dead before versus after 

 harvesting ([after minus before]). Data from all treat- 

 ment (dredging, tonging, and diver-harvesting; n = 4 for 

 each treatment) and the control Ui = 4) reefs were used in 



