24 



Fishery Bulletin 101(1) 



Table 1 



(A) Summary catch statistics {Townseiid Cromwell research cruise, Necker Bank, June 19991 and (B-D) fundamental linear-mass 

 interrelationships for body and egg sizes of female Hawaiian spiny lobster (Pa/iulirus marginatus) at Necker Bank, Northwestern 

 Hawaiian Islands. 



A Female catch statistics 



Total 



Berried 



TW„ 



TW,, 



median range 



median range 



834 



54.6 



350 



42.0 



50.1 



24-72 



51.1 



38-72 



B Relation of tail width to carapace length and vice versa; model: Y - aX + b 

 TW = 0.6087 CL + AAA and CL = 1.5772 TW - 4.00, 

 where TW = tail width in mm, CL = carapace length in mm, and a and h are fitted constants. 



[H-'=0.963, ?i=825,P<0.0011 



C Relation of body weight to carapace length; model: Y = aX'' 



SW = 0.00090 CL 2 9^=2 [range: 51.9-114.7 mm CL,n= 197,P<0.0011 



where BW = total body weight in g, and CL = carapace length in mm, for unbemed females (source: Uchida and Tagami [1984] ). 



D Relation of egg weight to egg diameter; model: Y = aX'' 



EW = 0.3985 ED2 2472 

 where EW = egg weight in 0.001 mg, and ED =egg diameter in mm. 



lr-'=0.833,;!=40, P<0.001] 



as the total number of pleopod-brooded eggs) was calcu- 

 lated as the product of mean RF and total weight of the 

 brooded egg mass. Pilot tests indicated that this procedure 

 estimated F with coefficients of variation (CV, SD/mean x 

 100%) consistently <5'7f. Subsamples of 25 eggs were ran- 

 domly taken from each female's total egg complement and 

 the diameter of individual eggs were measured (random 

 axis) at 500x magnification by using a dissecting micro- 

 scope and an optical micrometer. Average individual egg 

 weight was also independently derived as the ratio of the 

 weight to numbers of eggs present in the parent sample. 



Statistical analyses 



Relations of female body size to fecundity and body size 

 to egg size were evaluated for the 1999 samples by using 

 both linear and nonlinear least squares procedures (proc 

 REG, proc NLIN) of PC SAS for Windows v 6.12 (SAS 

 Institute, 1990a, 1990b). Analysis of covariance (ANCOVA; 

 proc GLM; SAS Institute, 1990c; Chubb, 2000) was used 

 to compare size-specific fecundity estimates of Necker 

 Bank P. ryjarginaliis among the three exploitation periods: 

 1978-81, 1991, and 1999. Subseasonal variation within 

 spawning seasons was controlled by the aforementioned 

 restriction on month of specimen collection, and single 

 collections were assumed to provide accurate character- 

 izations within exploitation periods. Fecundity data used 

 to characterize the 1978-81 and 1991 periods at Necker 

 Bank are listed in Appendix A of DeMartini et al. (1993). 

 Analogous comparisons of size-specific egg sizes were lim- 

 ited to the 1991 and 1999 periods because no data on vari- 

 ance of egg sizes were available for tiic 1978-81 samples 



(DeMartini et al., 1993). Body-size-fecundity relations 

 were allometric, hence log-linear (see Somers, 1991); natu- 

 ral logarithms were used for ANCOVAs and regressions of 

 log-linear relations. 



An index of reproductive potential (IRP; Kanciruk and 

 Herrnkind, 1976) was computed for the 1999 specimens 

 in order to determine the size classes of females that con- 

 tributed most to population egg production. The IRP was 

 constructed by using data for female P. marginatus caught 

 by the commercial fishery at Necker Bank during 1999, 

 collected by several contracted fishery observers. 



Results 



Fecundity and egg size of lobsters in 1999 



Fecundity A number of the female Panulirus marginatus 

 trapped on the research cruise at Necker Bank during 

 June 1999 were berried (Table 1). The estimated fecundity 

 of 40 females, spannmg 54.3 to 105.4 mm CL (39.3-67.4 

 mm TW, see CL-TW relation; Table 1), ranged more than 

 fivefold, from 109,865 to 590,530 eggs (Appendix A). 

 Fecundity was positively and nonlinearly related to TW 

 (Fig. 2) and CL and best described by the power equations 

 F = a TW' and F = a CL'', respectively, as 



F= 5.1743 rW''^ 7580^ 



|7-=0.889| 



and 



F= 7.9952 CL-^-^'i' 



lr2=0.900. both n=40. P<0.001. 



