FOLOV1NA: CATCHABILITY VEKS1UN OE LESLIE MODEL 



fishing (10-19 April) and the last 3 d (5-7 May), and 

 that the greatest increase in the catchability of P. 

 auricilla occurred after the time delay, it is possible 

 that the increase in catchability might have a time 

 lag component associated with it. However, given 

 the short time series of data, it would be difficult 

 to test the appropriateness of a more complicated 

 time lag model. 



Based on the fit of these two models the initial 

 exploitable population of the three species in the 

 150-275 m depth range at Pathfinder Reef is esti- 

 mated at 3,656 fish (Table 5). If we assume, based 

 on the species composition data (Table 1), that these 

 three species represent 90% of the exploitable 

 population then the total exploitable population at 

 the beginning of the intensive fishing is 4,062 fish. 



18 



1.6 



Q. 

 O 



Pristipomoides ouricillo 



J I I L 



25 50 75 IOO 125 150 175 200 225 250 275 

 CUMULATIVE CATCH 



Figure 3.— Daily catch per unit effort (CPUE) and predicted 

 CPUE based on the variable Leslie model as a function of adjusted 

 cumulative catch for Pristipomoides auricilla. 



From Figure 1 the length of the 183 m (100-fathom) 

 contour is estimated at 3.0 nmi, and the area in the 

 180-300 m depth range is estimated to be 0.4 nmi 2 . 

 With these area measures, density estimates of 

 1,354 fish per nmi of (183 m) 100-fathom contour 

 and 10,156 fish/nmi 2 , are obtained for Pathfinder 

 Reef. 



Estimates of bottom fish densities based on visual 

 observation from a submersible at Johnston Atoll 

 were 57,281 fish/nmi 2 for the 92-183 m (50-100 

 fathom) depth range and 66,199 fish/nmi 2 for the 

 1983-274 m (100-150 fathom) depth range (Ralston 

 et al. 1986). These figures are considerably larger 

 than both the point and interval estimates presented 

 here. Significantly, the study of Ralston et al. (1986) 

 also employed the Townsend Cromwell, and the 

 catch rates were comparable at Pathfinder and 

 Johnston (e.g., 3.18 bottom fish/line-hour for the 

 latter). Thus the difference between estimates of 

 standing stock is likely not due to differences in ab- 

 solute abundance but rather to differences between 

 exploitable population size and total population size. 

 For example, at Johnston Atoll at least 69 species 

 of fish were observed from the submersible, whereas 

 only 10 species were taken by fishing gear in the 

 same depth (Ralston et al. 1986). 



If the constant catchability Leslie model is applied 

 to the pooled data for the three species, an estimate 

 of exploitable population size of 2,689 is obtained, 

 about 71% of the estimate of the exploitable popula- 

 tion size for the three species when they are 

 estimated separately (Table 5). 



Size-specific behavior has been raised as a factor 

 which might affect catchability (Allen 1963). For all 

 three species, there is no evidence of intraspecies 

 size-specific behavior affecting catchability since for 

 two of the species the constant catchability model 

 fits well and for the third species, catchability 

 depends only on the population size of an interact- 

 ing species. Further, under the hypothesis that 

 within a stock catchability is size-specific across the 



Table 5.— Estimates of population size and catchability for three species. 



Species 



Model 



f? 



Catch- 

 ability 



SE 



Initial Confidence 



population interval 



size (95%) 



Pristipomoides Constant 



zonatus catchability 0.71 0.0038 0.0075 



Efefe Constant 



carbunculus catchability 0.35 0.0025 0.0010 



P. auricilla Variable 



catchability 0.89 0.00087 0.00031 



Three species Constant 



pooled catchability 0.66 0.0022 0.0047 



427 



