Moffitt and Parrish: Assessment of exploitable biomass of Heterocarpus laevigatas in Hawaiian Is , Part 2 



479 



times covering the entire 10 m^ quadrant, as well as 

 areas of exposed limestone often forming undercut 

 cliffs, and areas of small weathered volcanic rocks. 



The distribution ofH. laevigatus over the bottom was 

 evaluated in several ways. No significant values were 

 found for the x^ goodness-of-fit tests (P>0.10) of the 

 density observations recorded at each dive site, sug- 

 gesting that the shrimp were randomly distributed (in 

 a Poisson manner) instead of clumped at each dive site. 

 The mean and confidence intervals of H. laevigatus 

 density observed at each dive site are presented in 

 Table 2. Pooling data from all dive sites for the indepen- 

 dent variables (dive site, substrate material, substrate 

 particle size, and observer), only dive site and substrate 

 material were significantly correlated with shrimp den- 

 sity (ANOVA, P<0.05; Table 3). The distribution of 

 residuals did not differ significantly from a normal 

 distribution. Independent review of the dive video by 

 each observer yielded complete agreement on H. laevi- 

 gatus counts and substrate classification, indicating a 

 lack of observer bias in shrimp density estimation and 

 substrate associations. 



Comparisons of trap landings 

 and density estimates 



Mean observed densities were regressed on trap 

 catches for each dive site (Fig. 1) (r^ 0.97, P 0.0003). 

 The least-squares regression equation is 



CPUE, = 0.2896 (Dj), (SE 0.02364). 



An estimate of catchability is obtained directly from 

 the value of the slope (0.2896ha/trap-night) with 

 a confidence interval calculated as 0.2144-0.3648/ 

 trap-night. This estimate of catchability is <y3oth 

 that reported by Ralston and Tagami (1992) and is 

 reflected in the differences between observed and 



Q- 



400 800 1200 



Density (No. /ha) 



1600 



Figure 1 



Regression of trap catch per unit effort (CPUE; 

 number per trap-night) and observed density of 

 Heterocarpus laerigatus (number/ hectare) at 

 five sites in the Hawaiian Islands. 



expected H. laevigatus densities for Kona sites 2 and 

 3 (Table 2). 



Discussion 



Differences in behavior between H. laevigatus and 

 H. ensifer, the two species with greatest commercial 

 potential, may lead to some practical applications for 

 fishermen. The high activity level noted for /f. ensifer 

 in the presence of a baited container has also been 

 reported by Gooding et al. (1988) and Saunders and 

 Hastie (1989). The rapid attraction and entry of this 

 species into traps, even during daylight hours, indicate 

 that a short soak time may be adequate for commer- 

 cial harvesting. The lower activity level of H. laeviga- 

 tus observed in our study and reported by Saunders 

 and Hastie (1989) may indicate that a longer soak time 

 is more appropriate for this species. If so, a small vessel 

 with a limited number of traps could maximize total 

 catch by making two short sets during daylight hours 

 on H. ensifer grounds, followed by an overnight set on 

 H. laevigatus grounds, assuming that suitable concen- 

 trations of both species are present within a reasonable 

 proximity. 



