WILLIAMS: GROWTH AND SURVIVAL OF MANILA CLAM SPAT 



sumed that any predation on the clams before 

 they were able to bury within the substrate did 

 not substantially affect the magnitude of the 

 adult clam density, as compared with the other 

 treatments. 



I constructed Plots A, II, and III in July 1976, 

 and Plots R, S, and T in July 1977 at a tidal 

 height ranging from +0.25 m to +0.75 m (MLLW 

 datum), as determined by the height of the 

 nearby oyster dikes (Figure 1). For Plots A, R, S, 

 and T, wooden lath stakes were driven into the 

 ground to delineate 2 x 2 m squares, that were 

 then subdivided by stakes into four rows and four 

 columns. The result was sixteen 0.25 m^ subareas 

 in each plot. Four replicates of Treatments 1-4 

 were then constructed within each plot to form a 

 4x4 Latin squaire array of treatments. For Plots 

 n and ni, stakes were driven into the ground to 

 delineate a 1.5 m wide by 12.0 m long plot. 

 Treatments 1-4 were randomly assigned to sub- 

 areas within the plot so that the resulting 

 configuration consisted of four different 1.5 x 1.5 

 m treatments, separated by 1.5 m spaces between 

 each treatment, within each plot. 



A continuously recording thermograph was 

 buried 1 cm below the surface gravel next to Plot 

 A in October 1976. The recording tape was re- 

 placed monthly. 



The large plots were constructed to minimize 

 possible edge effects caused by the close proxim- 

 ity of treatments to each other. To further 

 minimize possible edge effects, samples were 

 taken only from the inside 0.25 m^ area of each 

 treatment in both the large and small plots. 



mesh size of 0.149 mm was necessary to insure 

 retainment of the smallest, newly settled spat in 

 freshly preserved samples. The residue from the 

 fiinest sieve was placed in a Petri dish under a 

 compound dissecting scope (50x) and examined 

 for clam spat. 



In 1976, when a large larval settlement was 

 detected from the weekly gravel samples, I sam- 

 pled all of the treatment areas in each plot. I con- 

 structed a sampling template from a piece of 1.91 

 cm thick plywood that had the inside 0.125 m^ 

 removed. Seine twine was stretched over the 

 opening to form a 5 x 5 grid with each square 

 5.72 cm on a side. 



For each individual treatment area in a plot, 

 five squares in the grid were randomly selected as 

 the sample sites from which to take cores. After 

 removing the cores, the holes were filled to beach 

 level with gravel taken from beside the plot, at a 

 depth of 4 cm, in order to avoid introducing newly 

 settled spat to the plots. 



To follow the grovi1,h and survival from settling 

 size, I took 10 cores per treatment in November 

 and December 1976 and in January, March, and 

 April 1977 from Plots A and II. The location of the 

 cores within each treatment was randomly 

 selected, excluding all core areas previously 

 utilized. In June 1977 the remaining core areas in 

 the different treatments were sampled. For each 

 sampling period, all cores in each treatment were 

 sieved in the laboratory, the clams were counted, 

 and a height and length measurement was taken 

 on the first 30 clams encountered per treatment. 



Sampling 



To check for newly settled clams, each week I 

 took two or three gravel samples beside each plot 

 by twisting a 20.28 cm^ clear plastic tube 2 cm 

 deep into the beach. A small hand trowel was 

 then shoved down beside and rotated under the 

 tube as it was removed from the gravel, prevent- 

 ing any material from falling out. The contents of 

 the tube were then transferred to a bottle or plas- 

 tic bag. A 10% formaldehyde solution with a con- 

 centration of 0.01% phloxine B dye was then added 

 to the container. 



In the laboratory, I washed the gravel samples 

 through a series of Tyler' sieves. Sieving down to 



^Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service. NOAA 



Sample Sizes 



Twenty core samples per treatment were taken 

 immediately after newly settled clams were ob- 

 served. Using methods given by Elliott (1971), I 

 determined that the number of clam spat per core 

 per treatment had a negative binomial distribu- 

 tion. The counts were transformed (logj^X) and a 

 mean and variance computed. These numbers 

 were utilized in Elliott's formula for the determi- 

 nation of sample sizes. A standard error equal to 

 20% of the mean and a 95% confidence were used. 

 The results showed that a sample size of 5 for 

 each treatment would have been sufficient. Based 

 on this and the amount of time necessary to pro- 

 cess each core, 10 cores per treatment were chosen 

 as the sample size for subsequent sampling 

 periods. 



893 



