Fishery Bulletin 91 1 



1993 



effect when sites were evaluated separately (F 152 =0.11, 

 P=0.75). However, female size significantly but weakly 

 (P 2 =0.08) affected egg size when data for the two sites 

 were pooled (CL effect: F I53 =2.104, P=0.04; median 

 egg diameter = 30.2 EPU+0.039 CL mm ; N=55; EPU= 

 0.0197 mm). 



Spawning frequency 



The relative frequency of berried/total adult females 

 collected at Maro Reef and Necker Island during the 

 summer of 1988 (iV=3085 adult females), 1990 (1198), 

 and 1991 (1165) was 0.246±3.401 (*±1SD, N=6 site- 

 year combinations). This index of the spawning fre- 

 quency of females did not differ (t'=0.01, P>0.9) from 

 0.230±0.510, the estimated frequency for 3037 females 

 collected during the summer of 1977 (N=2 site-years). 



specific fecundity might be expected to co-vary with 

 egg size and spawning frequency (Gadgil and Bossert 

 1970). However, in many organisms, offspring size and 

 number often do not track one another simultaneously 

 or to an equivalent extent (Capinera 1979, Roff 1982). 

 Therefore, our observation that egg size did not co- 

 vary with egg number in Panulirus marginatus should 

 not be surprising. Perhaps strong selection for plank- 

 tonic larvae of relatively invariant body size is typical 

 within particular populations of spiny lobster, even 

 though average egg sizes might differ among popula- 

 tions of some species. This speculation is consistent 

 with our observation that estimated egg volume var- 

 ied only about 50% among female P. marginatus of a 

 large range of body sizes from either site. This value is 

 low compared with those of most marine teleosts 

 (Bagenal 1971). 



Discussion 



Fecundity-body size relations 



Exponents of the curvilinear, F=aCL b , relations ob- 

 served in this study ranged from 2.16±0.241 (statisti- 

 cally equal to 2.0) to 2.50±0.175 (2.0<b<3.0; Fig. 1). 

 Perhaps both area of egg-bearing surface and volume 

 (female body mass) influence fecundity in this species. 

 Prior data on size-specific fecundity for an Oahu popu- 

 lation of P. marginatus allow us to estimate the expo- 

 nent in the equation F=a-CL b as 2.96±0.31SE (7V=11; 

 table XV, Morris 1968). 



A variety of linearxubic relationships are available 

 for other spiny lobster species. The following values of 

 the exponent b in the power equation are either known 

 or calculable for: Panulirus interruptus (1.0, Lindberg 

 1955), P. homarus (1.0, Berry 1971), P. cygnus (1.0, 

 Morgan 1972); Jasus verreauxi (1.0, Kensler 1967), 

 J. edwardsii (1.0, Kensler 1968; 3.01, McDiarmid 1989; 

 2.11-3.75, Annala & Bycroft 1987), J. lalandii (1.0, 

 Beyers & Goosen 1987 and Pollock 1987; 3.58, 

 Zoutendyk 1990), J. tristani (1.0, Pollock & Goosen 

 1991), Nephrops norwegicus (2.35, Thomas 1964 in 

 Aiken & Waddy 1980). Best fits thus vary from linear 

 to cubic both among species and among geographic 

 populations within species. The only apparent pattern 

 is the generally linear relationship between length and 

 egg number in other Panulirus spp. 



Parameter interrelations 



Although size-specific fecundity clearly changed be- 

 tween the "before" and "after" periods of exploitation 

 for Necker Island lobster, neither egg size nor spawn- 

 ing frequency appeared to differ between the two sam- 

 pling periods at either study site. Theoretically, size- 



Site and period comparisons 



The site (geographic population) differences observed 

 in this study of NWHI spiny lobster are not without 

 precedent in other spiny lobsters. Annala & Bycroft 

 (1987) and Beyers & Goosen (1987), for example, ob- 

 served geographic differences in the size-specific 

 fecundities of Jasus edwardsii and J. lalandii, respec- 

 tively. For J. edwardsii, differences in size-specific fe- 

 cundity were detected for populations ranging from 

 extreme southern to northern New Zealand (Annala & 

 Bycroft 1987). The spatial scale of the geographic pat- 

 tern noted by Beyers & Goosen (1987) for J. lalandii 

 was 600 km. Approximately 650 km separate Necker 

 island and Maro Reef. 



Unlike the Necker Island population, the fecundity 

 of lobsters from Maro Reef was indistinguishable be- 

 tween the two sampling periods. The absence of a pe- 

 riod effect at Maro Reef was not due to larger variance 

 at a similar effect size (Cohen 1988); rather, the stan- 

 dard deviation of the residuals of InF at Maro Reef 

 (0.217) resembled that for Necker Island (0.206). There 

 is the possibility, though, that size-specific fecundity 

 has increased at Maro Reef too, but we were unable to 

 detect it because of an artifact that particularly af- 

 fected the fecundity estimates of "after" specimens from 

 this site. Many of our "after" specimens from Maro 

 Reef had Stage-2 eggs. Greater egg loss might have 

 occurred during the handling of these specimens, be- 

 cause eggs at later stages of development may be more 

 prone to dislodgement (Annala & Bycroft 1987). Anec- 

 dotal observations of the integrity of egg masses of the 

 "after" females from Maro Reef suggest that, even though 

 some additional egg loss might have occurred, it is un- 

 likely to have been sufficient to obliterate an increase 

 in size-specific fecundity of a magnitude similar to that 

 observed for Necker Island lobster. 



