Wainwright et al. Effects of dredging on a crab population 



175 



rapid growth, and steady mortality for + crab and 

 relative stability for older age-classes. Fall (October- 

 December) and winter (January-March) are periods for 

 which we have little sampling data, but both are periods 

 of general population decline, migration from intertidal 

 to subtidal areas within the estuary by + crab, and 

 emigration from the estuary by older age-classes. 

 Where data were lacking during fall and vdnter, values 

 were projected from late-summer populations accord- 

 ing to the trends in numbers observed in years for 

 which winter data were available. 



Definition of geograplnic strata The population 

 survey design had four strata within Grays Harbor: 

 Outer Harbor, North Bay, South Bay, and Inner Har- 

 bor (Fig. 1). The navigation channel passes through two 

 of these (Inner and Outer Harbor), and crab densities 

 within various reaches of the channel were assumed 

 to be the average densities for the corresponding 

 sampling strata. In fact, crab densities estimated within 

 the channel during entrainment studies are quite com- 

 parable with those estimated from the corresponding 

 strata of the regular surveys (Dinnel et al. 1986, Dum- 

 bauld et al. 1988, Wainwright et al. 1990). Thus calcula- 

 tions for Bar, Entrance, and South Reaches used crab 

 densities for the Outer Harbor, while Inner Harbor 

 values were used from Crossover Reach to Aberdeen 

 Reach. Crab densities decline upriver, and South Aber- 

 deen Reach was assumed to have no crab. 



Mortality Mortality estimates were calculated by 

 regressing logarithm of population abundance on age. 

 This method was applied separately for early juveniles 

 (age + ) and for older juveniles and adults (age 1 -i- and 

 older). Because substantial migration of -i- crab to or 

 from the estuary does not occur, mortality rates spe- 

 cific to Grays Harbor could be calculated for this age- 

 group. To estimate mortality, total estuarine -i- and 

 1 -I- populations were calculated from the six years of 

 trawl survey data. Estimates for 0+ subtidal popula- 

 tions were supplemented with intertidal estimates to 

 provide a complete representation of the estuarine 

 population. Direct calculation of mortality requires 

 analysis of a population with no recruitment or migra- 

 tion. Settlement had essentially ended by July of each 

 year, so we chose July of the + year as the starting 

 point for calculations. During the 1 + year, migration 

 begins near the end of the summer as crab leave the 

 estuary. Because of this, we chose June of the 1 -i- year 

 as the endpoint for estimating first-year survival. First- 

 year mortality estimates were calculated for each of 

 five cohorts (1983-87 year-classes). 



Estimation of mortality for older ages is more dif- 

 ficult for two reasons: age-class separation is difficult 

 and inacctu*ate, and migration to and from the estuary 



occurs. Because of these problems, a different approach 

 was used. To reduce problems of migration, population 

 estimates for the estuary and adjacent coast were 

 combined. Age-class separations were made using an 

 instar analysis technique (Armstrong et al. 1987, Oren- 

 sanz and Gallucci 1988) to identify instar composition 

 of the population. Instar abundances were then as- 

 signed to year-classes. To reduce errors from sampling 

 and age-class identification, monthly abundance esti- 

 mates were averaged over all year-classes, then aver- 

 aged over months within each survey season to give 

 a single estimate for each age-class (a): 



Na = mean(Namv). 



(6) 



where Na^y is the abundance estimate for age a in 

 month m of sample year y. Then survival from age a 

 to a -I- 1 was calculated as 



N, 



a+l 



-"a, a+1 



N, 



(7) 



Because a single strong year-class biases estimates 

 calculated in this way, the very strong 1984 year-class 

 was excluded. The calculated age-specific natural mor- 

 tality rates were then combined to produce the survival 

 schedule (Sas in Eq. 4) used to calculate equivalent 

 adult loss from unadjusted loss. 



Estimating entrainment rate Numerous studies 

 have been conducted to estimate the rate of entrain- 

 ment of crab by various kinds of dredges, and the 

 subsequent damage and mortality to entrained crab 

 (McGraw et al. 1988). Entrainment and subsequent 

 mortality are discussed separately below. 



A regression relationship was used to predict the en- 

 trainment rate (crab entrained/key dredged; e in Eq. 

 2) from trawl-based density estimates (crab/ha). This 

 approach was used by Armstrong et al. (1987) and 

 McGraw et al. (1988) to estimate entrainment rates for 

 a hopper dredge. More data have been collected since 

 those studies, so a new relationship has been calculated. 

 Sampling during the entrainment surveys consisted of 

 two parts: sampling of the dredged material stream 

 aboard a hopper dredge, and concurrent trawl surveys 

 within the channel section being dredged. During each 

 survey, sampling occurred over a two- to three-day 

 period and covered several stations wathin the naviga- 

 tion channel. For each survey, mean entrainment (crab 

 per key dredged) and mean density (crab per ha) were 

 calculated over all samples within each station. This 

 provided a total of 14 points which were used to 

 calculate the regression. Details of survey methods are 

 given in McGraw et al. (1988). 



