DeMartini and Lau: Criteria for size at maturity in snappers 



455 



15 



20 



25 30 35 



Fork length (cm) 



40 



45 



Figure 3 



Scatterplots of mean percentage maturity by 2-cm length 

 class, for kalekale, Pnstipomoides siebotdii. Solid circles 

 inside hollow squares and solid line represent the con- 

 gruent observed values and fitted curve based on histo- 

 logical criteria of maturity and on maturity classified by 

 logistic regression on fork length, ovary weight, and oo- 

 cyte volume. The logistic curve was fitted by nonlinear 

 regression with each class mean weighted by the square 

 root of its count (see Fig. 1). 



and oocyte size can be used to classify maturity di- 

 rectly by using either discriminant analysis or logis- 

 tic regression, however. Logistic regression may be 

 more applicable in most studies because of its less 

 stringent assumptions (Press and Wilson, 19781. We 

 recommend using body length for these types of 

 analyses because length data are easier to obtain 

 than weight data. 



Relative merits of gross and histological evidence 



Histological inspection of stained ovary sections pro- 

 vides the most accurate assessment of sexual matu- 

 rity, but at the greatest cost. Gross morphometries 

 can be used to characterize maturity at a greatly re- 

 duced cost, but are less accurate predictors. The ben- 

 efit of a gross morphometric proxy depends on 

 whether the loss in accuracy is outweighed by the 

 cost savings. In our study, assignments of sexual 



maturity based on discriminant analysis or logistic 

 regression with body length or weight, ovary weight, 

 and oocyte volume were accurate (97-100% correct 

 classification) for ehu and kalekale. Analogous ma- 

 turity assignments based on only body size and ovary 

 weight (omitting oocyte volume) were less accurate 

 (93-97% correct), indicating a small but real contri- 

 bution of the egg size variable. Whether the benefit 

 of including an egg size variable merits the cost of 

 its measurement will vary with the particular study. 

 The egg size variable likely had no effect in the lo- 

 gistic regi-ession model for ehu because our specimens 

 included resting mature fish (undergoing oocyte atre- 

 sia) with small viable oocytes. Two of the three ehu 

 (32.5, 36.2 cm FL) that were misclassified as imma- 

 ture by discriminant analysis with body length, ovary 

 weight, and egg size were resting mature fish. Our 

 ehu data perhaps illustrate discrepancies between 

 gross and histological classifications that might arise 

 when histological criteria for maturity include evi- 

 dence such as atresia that is unrelated to oocyte 

 mass. Gross criteria likely are most accurate in cases 

 in which samples are collected during peak spawn- 

 ing periods when most fish are actively spawning. 

 Histological criteria may be less than 100% accurate 

 for cases (such as ehu, this study) in which some 

 adolescent fish may resorb yolked oocytes before first 

 spawning (De Veen, 1970; Ramsay and Witthames, 

 1996). The ability to determine maturity status with 

 reasonable accuracy with only gross metrics in this 

 study nonetheless suggests that such methods war- 

 rant further application and evaluation. 



Size at sexual maturity 



Ehu 



Our current best estimate of Lgp for ehu in 



Hawaiian waters (=28 cm FL) must be considered 

 preliminary because data for small immature fish 

 are relatively few. Everson ( 1984), in his pilot evalu- 

 ation of size at maturity for ehu from Hawaiian wa- 

 ters, also had few small specimens: only 2/185 fish 

 collected during the May-September spawning sea- 

 sons of 1977-81 were <30 cm FL. Prior to the esti- 

 mates provided in the present study, those provided 

 by Everson (1984) were the only size-at-maturity 

 estimates available for use in managing ehu in Ha- 

 waii (Kobayashi'^). 



Our paucity of specimens <25 cm FL might have 

 reflected hook-size selectivity (Ralston, 1990). How- 



^ Kobayashi, D. R. 1993. Effects of increasing the minimum 

 size limit or imposing fishing closures on three species of Ha- 

 waiian deepwater snappers. Honolulu Laboratory, Southwest 

 Fish. Sci. Cent., Natl. Mar. Fish. Serv., NOAA, Honolulu. HI 

 96822-2396. Admin. Rep. H-93-01. 18 p. 



