174 



Fishery Bulletin 90(1). 1992 



purposes, volume was measured as thousands of cubic 

 yards (key) of dredged material (1 kcy = 765m3). 



To obtain crab loss due to dredging from these two 

 sets of information, we require crab entrainment rates 

 (e), measured as numbers of crab entrained per unit 

 volume dredged. Total entrainment (E) is 



El so- = D 



I s ' 6g " V 1 sg 



(2) 



Postentrainment mortality (m), expressed as a pro- 

 portion of those entrained, varies with gear type, age, 

 and season. Age-specific loss (L) of crab in a single 

 season, location, and gear combination will be 



■'-'al sg ~ •'^ 1 sg ' •''^al s ' tl^asg • 



(3) 



To compare the relative importance of losses from dif- 

 ferent age-classes, equivalent adult loss (E AL) for any 

 season-location-gear combination is calculated as 



EAL 



Isg 



^ '-' al sg ' "as I 



(4) 



Data and estimation 



Population abundance Crab population surveys 

 were conducted over a six-year period (1983-88) in 

 Grays Harbor and along the adjacent coast. Stratified 

 random sampling was done with a small beam trawl 

 at biweekly or monthly intervals during spring and 

 summer (May-September) with occasional sampling 

 during fall and winter. From these surveys, crab den- 

 sities were estimated for each stratum, and total 

 population estimates were computed separately for 

 Grays Harbor and the adjacent coast using the National 

 Marine Fisheries Service BIOMASS program (Alaska 

 Fish. Sci. Cent., 7600 Sand Point Way NE, Seattle, 

 WA 98115), which uses standard stratified random 

 survey statistical methods (Cochran 1962). Details of 

 the survey design and population estimates can be 

 found in Armstrong and Gunderson (1985) and Gunder- 

 son et al. (1990). In addition to the trawl surveys, 

 intertidal crab were sampled in 0.25 m^ quadrats at 

 several locations within the harbor, and total intertidal 

 population was estimated as described by Dumbauld 

 and Armstrong (1987). 



Growth and age-classes In general, age-class iden- 

 tification is difficult in crustaceans (Hartnoll 1982). The 

 lack of retained hard parts prohibits direct aging 

 techniques (such as scale analysis in fish), so age must 

 be estimated from size. We relied on visual separations 

 of age-classes in size-frequency plots from the popula- 

 tion surveys, but molting and individual variability in 

 growth obscure age-class modes except for young, 

 rapidly growing crab. In all cases, young-of-the-year 

 (age -I- ) crab were easily identifiable as a separate size- 

 group. Age 1 -I- size distributions sometimes overlapped 

 older ages; in these cases, visual estimates of the sep- 

 aration point were supplemented by projecting growth 

 from earlier observations. No reasonable separations 

 could be made for older ages. For this reason, our 

 analysis uses three age-classes: O-i-, l + , and >l + . 

 Within Grays Harbor, we believe that most crab leave 

 the estuary before their third year, so that almost all 

 crab within the estuary identified as > 1 -i- are actually 

 age 2 + , and this assumption is made in our analysis. 

 Proportions in each age-class were then calculated from 

 the total size-frequency distribution of each sampling 

 stratum. 



where Sag is the total natural survival to adulthood 

 from age-class i in season k (assumed equal in all 

 habitats). Total loss for the project is then 



EAL tot = Z EALug. 



Ug 



(5) 



Definition of model seasons Seasons were defined 

 to reflect important biological processes and major 

 changes in crab abundance through the year. The 

 spring season (April and May) reflects the start of 

 settlement of the + age-class and a period of migra- 

 tion into the estuary by age 1 + coastal crab; summer 

 (June-September) is a period of continued settiement. 



