78 



Fishery Bulletin 104(1) 



converted catch curves were compared by 

 using a modified Ntest (Sokal and Rohlf, 

 1995 . 



Backwards extrapolation of the length-con- 

 verted catch curves was used to estimate 

 the probability of capture data. Selectivity 

 curves were generated by fitting a logistic 

 function to the plot of the probability of cap- 

 ture against size, from which values of the 

 parameters L^g, L^g, and the size at which 

 fish were fully recruited to the fishery (Ljqq) 

 were obtained. 



Estimates of the annual instantaneous 

 rate of natural mortality (M) were obtained 

 for each species with the empirical equation 

 derived by Hoenig (1983). Maximum age 

 estimates of 31 years for D. pictiim and 21 

 years for L. nebulosus from the literature 

 (Loubens, 1980; Edwards and Shaher. 1991) 

 were used because the maximum ages and 

 sizes obtained in our study were considerably 

 lower than other reported values. 



The annual instantaneous rate of fish- 

 ing-induced mortality (F) was calculated by 

 subtracting the natural mortality rate (M) 

 from the total mortality rate (Z) derived from 

 age-based catch curves (F=Z-M). The calcu- 

 lation was also made for the upper and lower 

 95% confidence intervals for estimates of Z 

 in order to derive a range of fishing mortal- 

 ity rate estimates. The exploitation rate (E) 

 was calculated as the proportion of the fish- 

 ing mortality in relation to total mortality 

 {E=FIZ). 



Assessment of the fishery 



Relative yield and biomass-per-recruit analyses were 

 used to assess the fishery. Growth (k and L^), mortal- 

 ity (M), and selectivity (Z-f^) parameters were used as 

 model inputs, and knife-edge selection was assumed. 

 The Beverton and Holt (1966) yield-per-recruit (YPR) 

 model modified by Pauly and Soriano (1986) was used to 

 estimate the sizes at maximum yield per recruit ^L^^^) 

 and to predict the effects of increasing the mean size 

 at first capture (Z-5q) to the mean size at first sexual 

 maturity (L^^^) and that at which yield per recruit 

 would be maximized (^,„ax*- Estimates of exploitation 

 rates representing 1) a marginal increase of relative 

 yield per recruit which is 0.1 of its value at the origin 

 (£gi) and 2) maximum yield iE^^^^) were also derived 

 from the model. The exploitation rates corresponding to 

 i^^pt and ^iin,,, (^upt and£|,|^|,) were calculated and used 

 to estimate the relative biomass per recruit for each 



,^J^ V 



^^^E> 



■■^/>---l 



Figure 2 



Photomicrographs of transverse sections through the sagittal oto- 

 liths of (A) Diagramma pictum. 56.0 cm Lj, and (B) Lethrinus nebu- 

 losus, 45.7 cm Lp. Dots show opaque zones (scale bar=l niml. 



and i'niax frorn relative biomass- 



species for Ljq, L^ 

 per-recruit curves. Precautionary target iF^^) and limit 

 ^^limit' biological reference points were calculated as 

 0,5 and 2/3 M, respectively, and used to infer resource 

 status by direct comparison with the fishing mortality 

 rates established for the study species. 



Results 



Age and growth 



Alternating translucent and opaque growth increments 

 were observed in transverse sections of the sagittal 

 otoliths of D. pictum and L. nebulosus when viewed 

 with transmitted light under low-power magnification 

 (Fig. 2). For both species, one growth increment consist- 

 ing of an opaque and translucent zone was formed on 

 an annual basis. Opaque bands formed in the summer 

 months between May and September in association with 

 increasing sea water temperatures (Fig. 3); conversely 

 translucent zones were deposited in the autumn and 

 winter (October to February) in association with decreas- 

 ing sea water temperatures. 



The maximum age estimates determined from counts 

 of opaque bands were 13 and 14 years for D. pictum 

 and L. tiebulosus, respectively. Size-at-age relationships 

 were asymptotic in form and there was considerable 

 individual variability in growth (Fig. 4) (parameters of 

 the von Bertalanffy growth function are given in Table 

 1). A comparison of the growth characteristics between 

 sexes revealed that there were no significant differ- 

 ences in parameter estimates for both species (P=0.125, 

 df=319 for D. pictum and P=0.878, df =324 for L. nebu- 



