DeMartini et al.: Fecundity comparisons of Panuhrus marginatus 



A significantly greater size-specific fecundity per- 

 sisted at Necker Island in the "after" period even if 

 the most-likely outliers from the "before" data were 

 excluded from the analysis. This conservative re-analy- 

 sis strongly suggests that period differences in the size- 

 specific fecundities of Necker Island lobster are real. 

 The magnitude of our best estimate of this difference 

 ( 16% ) further suggests that it represents a biologically 

 meaningful change. A much more extensive series of 

 before-and-after data would be necessary to resolve 

 this issue beyond a reasonable doubt. 



On balance, then, our data suggest that the size- 

 specific fecundity of NWHI spiny lobster has increased 

 at Necker Island, but likely not at Maro Reef, follow- 

 ing a decade of heavy exploitation at both fishing 

 grounds. Why and how can this be? 



The data of Polovina (1989) suggest one possible ex- 

 planation. The density-dependent decrease in mean cara- 

 pace length-at-onset of egg production has been propor- 

 tionately greater for female lobsters at Necker Island 

 (10-15%, 67.8 to 57.9-60.8 mm) than at Maro Reef (6- 

 9%, 74.8 to 68.2-70.5mm; table 3, Polovina 1989). In 

 1977, the pre-exploitation densities (catch per trap-haul) 

 of adult lobsters were higher at Necker Island (x±SEM 

 = 5.7610.75) than at Maro Reef (3.89±0.45); yet by 1986- 

 87, densities had declined more at Necker Island (to 

 2.14±1.24) than at Maro Reef (2.65±0.67; tables 1&2, 

 Polovina 1989). By decreasing densities to a greater 

 extent (thereby making more shelter or food resources 

 available per lobster), it is possible that the effects of 

 more intense cropping may now extend beyond decreased 

 size-at-maturity to greater size-specific fecundity for 

 Necker Island lobster. 



Other factors besides exploitation might be some- 

 how influencing the density and size-specific fecundity 

 of P. marginatus at Maro Reef. Relatively few small 

 lobsters were present at Maro Reef compared with 

 Necker island during either of the time-periods used 

 in our fecundity comparison (Polovina 1989; Fig. 1). 

 Even so, disproportionately low recruitment at Maro 

 Reef during 1986-88 might have recently exaggerated 

 long-term baseline differences in the body-size distri- 

 butions of exploitable P. marginatus at the two banks 

 (Polovina & Mitchum 1992). Of the several possible 

 reasons for higher pre-exploitation densities at Necker 

 Island (more productive bottom habitat, fewer adult 

 lobster predators), none is presently substantiated. 

 Studies are in progress to evaluate the quality and 

 distribution of bottom habitats among Maro Reef, 

 Necker Island, and other NWHI lobster grounds. 



Density-dependent reproduction in lobsters 



Density-dependent changes in egg production have been 

 described or suggested for several other species of spiny 



lobster. Examples include changes in the number of 

 spawnings (duration of spawning season) for individual 

 females (Chittleborough 1976, 1979) as well as in- 

 creases in size-specific fecundities (Thomas 1964, 

 Beyers & Goosen 1987, MacDiarmid 1989) in response 

 to lower population densities or otherwise greater food 

 availabilities. Body size at onset of sexual maturity 

 has responded in qualitatively different fashion to 

 changes in lobster densities among different popula- 

 tions and species of lobsters. Several studies of Jasus 

 lalandii (e.g., Pollock & Goosen 1991) have observed 

 smaller body sizes at first maturation in stunted or 

 physiologically stressed, high-density populations. As 

 mentioned previously, Polovina (1989) has demon- 

 strated the opposite response to reductions in adult 

 densities resulting from exploitation. A fundamental, 

 biological difference is whether size-at-maturity is an 

 age- or size-dependent trait for a particular species 

 and population. It is reasonable to expect species dif- 

 ferences in the determination of maturation, and com- 

 parative studies of the mechanisms regulating the on- 

 set of maturation in different spiny lobsters would be 

 informative. 



Statistical considerations 



Several studies have evaluated, but been unable to 

 detect, geographic variations in fecundity (Thomas 

 1964, Morgan 1972). The study by Morgan (1972) rep- 

 resents a case in which sample sizes were small and 

 power was low despite nontrivial effect sizes. Future 

 studies of lobster fecundity should report statistical 

 power, especially when tests are unable to reject null 

 hypotheses of no difference among populations. 



Acknowledgments 



We thank F. Parrish for providing the specimens used 

 for the "after" characterization; B. Lau for assisting in 

 the development of laboratory protocols; K. Landgraf 

 for help in locating files; W Haight for deciphering 

 historical catch records; J. Polovina for comments on 

 analysis design; and D. Aiken, J. Polovina, R.M. Smith, 

 and J. Uchiyama for constructive criticisms of draft 

 manuscripts. 



Citations 



Aiken, D.E., & S.L. Waddy 



1980 Reproductive biology. In Cobb, J.S., & B.F. 

 Phillips (eds.), The biology and management of 

 lobsters. Physiology and behavior, vol. I, p. 215- 

 276. Academic Press, NY. 



