STEVENS AND ARMSTRONG: CANCER MAGISTER IN GRAYS HARBOR 



descriptions see Stevens 1982). The area of each 

 stratum was determined by planimetry at the 

 level of mean lower low water (NOAA Chart No. 

 18502 Grays Harbor, 1979 edition). 



Sampling Design 



Crabs were sampled at stations 1-9 and 13 with a 

 4.9 m, 4-seam, semiballoon otter trawl net, having 

 38 mm stretch nylon mesh throughout and a 6 mm 

 cod end liner. Working width of the net was about 

 3.0 m. Distance towed was measured between 

 buoys placed at the beginning and end points of 

 each trawl, by compass triangulation to station- 

 ary objects whose positions were predetermined 

 and located on 7.5-min topographic maps (U.S. 

 Geological Survey). Distances were then con- 

 verted to area swept and catches expressed as 

 crabs/ha. At stations 10-12, underwater snags pre- 

 vented trawl operation so crabs were collected by 

 setting collapsible ring nets (76 cm diameter) cov- 

 ered with 12 mm mesh. A "set" consisted of 4 baited 

 nets set 50 m apart and fished for 20 min. Catches 

 were expressed as crabs/net. Trawls and ring net 

 sets were made within 1-2 h of slack low tide in 

 daylight. Occasional plankton tows were made 

 with a 0.5 m diameter conical net of 500 /xm mesh, 

 in the spring of 1980 and 1981, to determine if crabs 

 entered the bay as larvae. 



Stations 3, 6, 8, and 9 were sampled biweekly 

 from May through October 1980 and at intervals of 

 4-5 wk thereafter through July 1981. Other sta- 

 tions were generally sampled monthly except 

 when weather or boat problems precluded opera- 

 tions. Most stations were sampled on 13-19 occa- 

 sions during the 14-mo field study (May 1980 to 

 late June 1981), with the exception of stations 1 

 and 2 (6 and 10 samples, respectively). Station 13 

 was sampled quarterly on a diel basis, and com- 

 plete results from that diel study are reported 

 elsewhere (Stevens and Armstrong 1984). No sam- 

 ples were taken at stations 14 and 15 which were 

 used only to calculate crab populations based on 

 data from adjacent areas (see below). 



All crabs were measured to the nearest milli- 

 meter across the carapace between the notches 

 just anterior to the 10th anterolateral spines 

 ("carapace width" or cw), sexed, and released. Sub- 

 samples were used for width frequencies only in 

 May and June of 1980 and 1981, when early instars 

 were collected in large quantities. Surface and 

 bottom-water samples were collected during each 

 trawl with a modified Van Dorn bottle; tempera- 

 ture was measured to 0.1°C, and salinity deter- 



mined with a refractometer at room temperature. 



Growth Analysis 



Cumulative width frequencies of all crabs 

 caught during a given week were plotted on proba- 

 bility paper, and width limits were subsequently 

 defined as the curve inflection points (arbitrarily 

 nonoverlapping) to delimit the size range of each 

 year class through time, according to the method 

 of Cassie (1954). These were compared with fre- 

 quency graphs for verification. Values were inter- 

 polated during weeks in which too few crabs were 

 caught for accurate analysis. Each crab was then 

 assigned to an age-group on the basis of the width 

 limits for each sampling week. Age was defined as 

 the number of years since metamorphosis. Mean 

 widths were calculated for each age group (0+, 1+, 

 2+, and 3 + ) and plotted by sampling week. 

 Eighty-seven males (12-132 mm cw) and 74 female 

 crabs (15-115 mm cw) were frozen and returned to 

 the University of Washington where they were 

 opened at the epimeral line and dried to constant 

 weight at 60°C (48-72 h). Only hardshell intermolt 

 crabs were used. Log 10 dry weight (g) was plotted 

 against log 10 carapace width (mm) and regression 

 equations determined for each sex. Mean weights 

 for each age group of crabs were calculated at 

 monthly intervals from mean widths using the 

 regression equation (the 1977 year class was omit- 

 ted because the regression equation did not repre- 

 sent these larger animals). Weight-specific growth 

 rates {k) per month were calculated by use of the 

 equation 



W, = W ekt. 



The monthly percent weight increase was calcu- 

 lated as e k - 1. 



Crab Density Analysis 



Because counts of benthic invertebrates usually 

 show a contagious distribution (Elliott 1977), all 

 density data were transformed prior to analysis of 

 variance or regression by 



X t = Log 10 (density + 1), 



where X t is the transformed variable. 



Density was plotted against bottom-water salin- 

 ity, temperature, and estimated Chehalis River 

 flow by a stepwise multivariate procedure (SPSS 

 REGRESSION) for all trawl and ring net samples. 



471 



