 







■■v:£*X 



FIGURE 1.— Differences between the bull rake (left) and pea digger (right) in number and spacing of teeth, and head size and mass. 



nearby seagrass bed (about 80% Zostera marina and 

 20% Halodule wrightii). Three replicate sediment 

 cores, taken in April 1981 at each site to a depth of 20 

 cm and analyzed by standard sieving techniques 

 (Ingram 1971;Folk 1974), revealed that sediments in 

 both habitats were predominantly medium, fine, and 

 very fine sands. In the seagrass bed, however, the 

 sediment size distribution shifted substantially to- 

 wards finer size classes: Average percent dry weights 

 in the three decreasing size classes (medium, fine, 

 and very fine sands) were 13.9, 44.7, and 18.3%, re- 

 spectively, as compared with 28.8, 64.1, and 2.9% in 

 the sand flat. Furthermore, the seagrass sediments 

 contained 20.6% silt and clay, whereas the sand-flat 

 sediments contained only 2.3% by weight within 

 these mud size classes. The sand flat held relatively 

 little shell debris, whereas buried empty clam shells 

 were common in the seagrass bed. Average density of 

 seagrass shoots in the seagrass habitat was 496 (±1 

 SDof 165)M 2 , ; based on eight 0.5 m~ 2 samples. We 

 selected all specific study plots in water < 1 m deep at 

 low tide, for ease of access. Our study plots were 

 located at about midrange of the depth occupied by 



430 



seagrasses in the Beaufort area. 



Sand-Flat Methods 



We chose 2 X 4 m plots in pairs, matching plots in 

 space, water depth, and surface appearance of the 

 substrate. We marked each plot by inserting a 1.6 m 

 stake at each corner. One plot (chosen at random) of 

 each of the 14 selected pairs was raked systematical- 

 ly for 6 min using a bull rake, while the other matching 

 plot in each pair was raked systematically for 6 min 

 with a pea digger. Prior to our use of the two rakes, we 

 had carefully observed the usage of each rake by 

 several professional clammers in the field so that we 

 could employ each device in a way that closely 

 resembled its customary usage. We used only back- 

 ward strokes in deploying the pea digger, which pene- 

 trated 3-6 cm into the sediments in both habitats. 

 During the 6-min raking period, variable proportions 

 of the 2 X 4 m plots were raked (including occasional 

 larger areas). In each plot the actual area raked was 

 marked in the field and recorded. We also recorded 

 the numbers of hard clams, collected from each trial, 



} 



