Davis et al Population assessment of Limulus polyphemus 



219 



from other states were unavailable or unreliable. 

 We therefore expanded the Delaware harvest to 

 represent the Delaware Bay region by making the 

 assumption that regional landings were equal to 

 ten times the landings in Delaware, the approxi- 

 mate relationship from 1997 through 2003 when 

 reporting was mandatory. When applicable, we 

 converted numbers to metric tons (t) (Prager and 

 Goodyear, 2001) using the relationship of Gibson 

 and Olszewski'' (1.8182 kg/horseshoe crab). Com- 

 mercial harvest of horseshoe crabs has been below 

 the regional quota of 2595 t since 2000 (Fig. 3). 



Assumptions associated with 

 the production models 



There were a number of general assumptions asso- 

 ciated with production models (Quinn and Deriso, 

 1999). We assumed that productivity (change in 

 biomass over time) responded instantaneously to 

 changes in population size. Changes in the biotic 

 and abiotic environments were ignored, and /• (the 

 intrinsic rate of population growth) and K (the carrying 

 capacity) were assumed to be constant. Because produc- 

 tion models combine all age classes, it was assumed that 

 size or age structure of the population would not have 

 major effects on population dynamics. 



Specific assumptions about population values were 

 also required by the model. All starting values of 



MSY and K were based on the maximum harvest in 

 the Delaware Bay from 1991 through 2003. The initial 

 guess for MSY was 1850 t (half of the largest catch), 

 and the initial guess for K was 37,000 t (ten times the 

 largest catch). For fishery-independent model runs, 

 we had the model freely estimate initial biomass in 

 relation to carrying capacity (B^IK), with our start- 



