Appraisal of the selective properties of gill nets and implications for ...

L.E. Miranda et al. / Fisheries Research 45 (2000) 105±116 109where,P largeL rL lL meanL thproportion of large fish in the catchthe length at which fish are recruited to thegear (from equation in Table 1)the length at which fish are no longersusceptible to the gear (from equation inTable 1)mean length of fish captured by the gearthreshold length separating the small fromlarge fish recognized by the commercialmarketThe yield of large (Y large ) and small (Y small )fishwasestimated asY large ˆ Y…P large †;Y small ˆ Y…1 P large †:2.3.3. Commercial value per recruitThe commercial value of the catch was estimated asthe yield of ®sh in each size group times its wholesalevalue. Fish harvested by ®shers are bought by wholesalerswho sell them to merchants. Agostinho et al.(1999a,b) estimated that the prices obtained by wholesalerswere approximately US$1.5 kg 1 for large P.lineatus, $2kg 1 for large P. squamosissimus, and$0.9 kg 1 for small P. lineatus and P. squamosissimus,and all H. edentatus.2.3.4. Spawner biomass per recruitAlternative levels of harvest will in¯uence spawnerbiomass and thereby could affect recruitment.Because the spawner±recruit relation was unknown,we estimated the stock biomass per recruit and used itas a benchmark for comparing alternative mesh-sizeregulations (Norris, 1991). With this information,comparisons of yield from different mesh-sizes canbe limited to scenarios that produced the same spawnerbiomass per recruit.Spawner biomass (B) was estimated asB ˆ B1 ‡ B 2 ‡ B 3 if t m < t rB 3 ‡ B 4 if t m t rwhereB 1 ‰…N 1 e Fm W 1 †=kŠ‰H 3 H 1 Š for F ˆ 0B 2 ‰…N 1 e Fr W 1 †=kŠ‰H 1 H 2 ŠB 3 ‰…N 1 e Fl W 1 †=kŠ‰H 2 Š for F ˆ 0ift > t l‰…N 1 e Fm W 1 †=kŠ‰H 3 H 2 ŠB 4H 3b(X 3 ,P,Q)/(1/((G(P) G(Q))/G(P ‡ Q)))m years elapsed between age 1 and when 50% ofthe fish become maturet m time at maturity (Table 1)t r time when fish become susceptible to the geart l time when fish are no longer susceptible to thegearX 3 ˆe k(t m 1)2.3.5. Empirical estimates of mortalityEstimates of Z for each of the three study specieswere made with empirical data on ®sh abundancecollected with gangs of gill nets consisting of netshaving 3, 4, 6, 8, 10, 12, 14 and 16 cm stretch mesh.These gill nets were ®shed throughout Itaipu Reservoirduring 1984±1989. Age (t) was estimated fromlength (L) ast ˆ log e …1 L=L 1 †=k:Age was regressed on log e of catch of ®sh of age t.Regression was limited to the descending portion ofthe catch curve.3. ResultsEmpirical estimates of annual Z averaged 1.29(28% annual survival) for H. edentatus, 1.13 (32%survival) for P. squamosissimus, and 1.06 (35% survival)for P. lineatus (Fig. 2). The gill nets used toFig. 2. Catch curves for the three study species. Regression lines(solid) were fit to the descending portions of the curves. The lineswere extended (dashed lines) to extrapolate relative populationdensity at age 1. Values in parentheses next to the Z valuesrepresent 1 S.E.