DAILEY and RALSTON: BIOLOGICAL DEVELOPMENT OF HETEROCARPUS LAEVIGATUS 



diately following than during the latter half of the 

 reproductive season. 



The total sample CL-frequency distribution of 

 males and females combined was analyzed by the 

 regression method of Wetherall et al. (in press) to 

 estimate L m and Z/K. When all shrimp are pooled 

 (N = 7,368), an estimate of L x = 61.7 mm CL 

 results. Further, the ratio of total mortality rate to 

 von Bertalanffy growth coefficient (Z/K) is esti- 

 mated to be 2.6. Calculations were repeated for 

 separate male and female subgroups, where it was 

 found that L = 57.9 and 62.5 mm CL and ZIK = 



00 



4.3 and 2.9 for males and females, respectively. 

 These results indicate that males generally grow to 

 a smaller size than females. 



The results of analyzing the progression of CL size 

 modes in frequency distributions of male and female 

 H. laevigatus provided preliminary estimates of K 

 = 0.35 yr -1 for males and 0.25 yr -1 for females. 

 The former result must be viewed with caution, 

 however, because two "solutions" were detected by 

 the computer search algorithm (Pauly 1982) which 

 differed little in fit. One of these, K = 0.70 yr- 1 , 

 we believe to be unjustifiably high in light of the 

 minor difference (8%) between the L^ of males and 

 the L^ of females obtained from the regression 

 analysis. Note that estimates of K and L k typical- 

 ly show a strong inverse correlation (Gallucci and 

 Quinn 1979). These results, in conjunction with the 

 estimates of ZIK for male and female shrimp pre- 

 sented earlier, provide the basis for preliminary 

 estimation of total mortality rate. We estimate Z 

 = 1.51 yr -1 for males and 0.73 yr -1 for females, 

 corresponding to annual survivorship fractions of 

 22% and 48% per year, respectively. These data in- 

 dicate that males grow faster while experiencing a 

 substantially greater total mortality rate than 

 females. 



DISCUSSION 



Earlier it was assumed that, aside from depth, all 

 shrimp samples were drawn from locations which 

 are dynamically homogeneous; i.e., the behavior of 

 shrimp populations through time does not vary from 

 site to site. This is clearly a restrictive and simpli- 

 fying assumption and is without doubt the major 

 limitation on the results presented here. Nonethe- 

 less, it was a necessary simplification for us to 

 analyze the commercial fishing data upon which this 

 study was based. Consequently, we view those 

 results which rely upon this assumption as tentative 

 and in need for further validation. 



Examination of the seasonal trend in the relative 



abundance of ovigerous females showed that in 

 Hawaii over 50% of H. laevigatus females bear eggs 

 from October to January, with a peak between 

 August and February. Wilder (1977) found a similar- 

 ly timed but more narrowly defined breeding season 

 for H. laevigatus in Guam, where the percentage of 

 ovigerous females in trap catches reached a max- 

 imum during December, but was not particularly 

 high in any other month. Clarke (1972) reported that 

 H. ensifer in Hawaii also reproduces in the winter. 

 The breeding season of these shrimps is unusual 

 among Hawaiian crustaceans and fishes, which 

 typically reproduce during the spring and summer 

 and uncommonly during the winter (Watson and 

 Leis 1974; Lobel 1978; Uchida et al. 1980; Uchida 

 and Tagami 1984; Walsh 1984). 



Our data also indicate that in Hawaii sexual 

 maturity of female ono shrimp occurs at approx- 

 imately 40 mm CL, a size similar to that reported 

 by King (1983) for shrimp from Fiji, Vanuatu, West 

 Samoa, and Tonga and by Moffitt and Polovina (fn. 

 5) for samples from the Marianas. Based upon the 

 estimated parameters of the von Bertalanffy growth 

 equation derived here, this corresponds to an age 

 of first maturity of 4 yr. Although we have no data 

 on the maturation of males, we believe they prob- 

 ably mature earlier and at smaller size, perhaps at 

 age 3 when they are 37-38 mm CL. Such a result 

 is consistent with the findings of Moffitt and Polo- 

 vina (fn. 5) who found that male H. laevigatus in the 

 Marianas mature at a smaller size than do females. 



Wilder (1977) speculated that both H. ensifer and 

 H. laevigatus in Guam are protandrous hermaphro- 

 dites, as did Clarke (1972) for Hawaiian populations 

 of H. ensifer. However, the results presented in 

 King and Moffitt (1984) tend to contradict this con- 

 clusion. These authors studied the morphometry and 

 sexuality of five deep water pandalids, including//. 

 laevigatus, in Fiji and the Marianas. Using the 

 relative length of the appendix masculina expressed 

 as a proportion of CL, they found no tendency 

 toward protandrous hermaphroditism. Moreover, 

 the sex ratio reported in their study was approx- 

 imately 1:1. 



Our results also indicate that for Hawaiian popula- 

 tions of H. laevigatus, and we speculate for most 

 tropical pandalids, a sex transition does not occur. 

 Wenner (1972) has termed the pattern exhibited in 

 Figure 4 the standard sex ratio pattern, as distin- 

 guished from one of reversal. Due to the large 

 numbers of females in small size classes, these data 

 are generally inconsistent with a protandric herma- 

 phroditic life history, as has been hypothesized by 

 previous workers on Heterocarpus spp. (Clarke 



923 



