224 



Fishery Bulletin 104(2) 



trawl and the DE 16-ft 160-mm survey, both of which 

 were included in model FI-3, and possibly explain the 

 low population estimates. 



In other trend analyses conducted during the previous 

 assessment, a population decline in the Delaware Bay 

 was less evident. Using data from 1997 through 2003, 

 we found that only four of eight fishery-independent 

 surveys showed a significant decline, partially owing 

 to high variability and low power. By incorporating a 

 number of these surveys into a production model, we 

 also found that the decreasing biomass in recent years 

 becomes more apparent. These production model runs 

 provide the added benefit of estimating stock status 

 and management benchmarks, as well as the benefit of 

 evaluating future management options. 



Although interpretation of absolute biomass or popula- 

 tion size from a surplus production model can be some- 

 what problematic, our estimates are roughly comparable 

 to estimates from previous studies. Hata and Berkson 

 (2003) calculated a mean 2001 population size of 4.4 

 million adults (95% confidence intervals of 2.1 million 

 and 6.8 million) in the Delaware Bay from daytime 

 trawl survey data. Botton and Ropes (1987) estimated 

 2.3 to 4.5 million adults in this region. In the present 

 study, our estimates of the 2003 population size ranged 

 from 0.6 million to 3.6 million crabs, and the mean of 

 the seven model runs equaled 2.0 million crabs. Eighty 

 percent confidence intervals ranged from 0.3 million to 

 6.6 million crabs for all model applications. Although 

 within the range of results from the other studies, these 



wide confidence intervals provide little information for 

 management. It will therefore be more appropriate to 

 interpret relative biomass (B/B„j;j.) and relative fishing 

 mortality (i^/.F^/.sj) for use in management decisions 

 (Prager, 1994). 



It is important to understand the spatial scale and 

 population represented by these models and analyses. 

 This regional model is a compilation of a number of 

 localized fishery-independent surveys, most of which 

 encompassed a relatively small spatial area. However, 

 the model results should not be interpreted at a more 

 localized scale because landings data are combined for 

 the region. Similarly, interpretation of results should 

 not be expanded to represent other Atlantic horseshoe 

 crab populations outside the Delaware Bay region be- 

 cause neither survey nor harvest data in this model 

 extend to other regions. Nevertheless, the Delaware Bay 

 is believed to be the center of abundance and spawning 

 activity for Atlantic horseshoe crabs; therefore popula- 

 tion trends in this region may have significant implica- 

 tions for adjacent populations. 



In analyses conducted for ASMFC,- trend analyses 

 identified dramatic regional differences in horseshoe 

 crab population trends. Although the Delaware Bay, 

 eastern Long Island Sound, and New England popula- 

 tions have experienced declines in recent years, the 

 southeast population and the western Long Island popu- 

 lation have remained stable or have increased. Future 

 production models applied to these other regions will 

 hopefully clarify these trends and allow managers to 



