MOFFITT ET AL.: DISTRIBUTION AND YIELD OF DEEPWATER SHRIMP 



pressed in percent males) for the three major species 

 were 52.8% for H. ensifer (N = 3,302), 55.2% for 

 H. laevigatus {N = 12,555), and 24.2% for//, longi- 

 rostris {N = 1,408). These ratios tended to hold true 

 for all size classes except the very largest of each 

 species which were indeed nearly all females. Sex 

 ratios did differ by depth showing that the two sexes 

 tend to occupy different areas. In all three major 

 species, females were more abundant at the shal- 

 lower end of the depth range and males at the 

 deeper end (Fig. 3). This relationship is much more 

 obvious for H. laevigatus and H. longirostris than 

 for H. ensifer. For H. longirostris, this may explain 

 the small percentage of males taken in this study 

 since the deeper end of the depth range of this 

 species was not sampled. 



100 



80 



[2 60 



-I 

 < 



s 



v9 40 



20 



100 

 80-1 



(A 



uj 60 



< 

 S 



S? 40 



20 



100 

 80 



ffi 60 



-I 

 < 

 S 



a? 



40 - 



20- 



H. ensifer 



■',' 



II I 



H. laevigatus 



I 1 



."I 



H. longirostris 



400 500 600 700 800 900 1000 1100 1200 

 DEPTH (m) 



Figure 3.— Sex ratio by depth for three species of Heterocarpus 

 with 95% confidence limits. 



REPRODUCTION 



Pandalid shrimp are typically considered to be pro- 

 tandrous hermaphrodites. This is indeed the case for 

 the Pandalus species taken in the subarctic areas 

 (Butler 1964). In these species the shrimp spend the 

 first few years of life as functional males, trans- 

 forming into functional females for the last year or 

 two of life. Clarke (1972) and Wilder (1977) sug- 

 gested that the tropical Heterocarpus shrimp also 

 are protandrous hermaphrodites. The sex ratios ob- 

 tained in our studies, particularly the near even ratio 

 for the smaller individuals led us to believe that this 

 was not the case. King and Moffitt (1984) examined 

 males of several species of Heterocarpus and 

 Plesionika for relative growth of the appendix 

 masculina on the second pleopod (a secondary sex 

 characteristic). If these species were indeed pro- 

 tandrous hermaphrodites, the relative size of the 

 male appendage should decrease with increasing 

 carapace length (as the shrimp transforms from 

 male to female). This was not so for any of the 

 tropical pandalids examined. Instead, the relative 

 size of the male appendage increased with increased 

 carapace length indicating maturation as a male. 

 The Marianas data for the ratio (R) of the appen- 

 dix masculina to the appendix interna versus 

 carapace length (CL) was fit to the logistic model 

 with three parameters, a, b, and c (Gunderson et al. 

 1980). Table 3 lists the parameters obtained when 

 fitted to the model 



R X 100 = 



a 



I + g-6(CL-c) 



The fit to the nonlinear regression for H. longi- 

 rostris was not particularly good due to the lack of 

 small males in our collection. When the data for this 

 species are fit to a linear regression, however, the 

 slope of the regression is positive indicating rela- 

 tive growth of the secondary sexual characteristics 



Table 3. — Parameter values for the nonlinear regression 

 of the relative length of the appendix masculina versus 

 carapace length. 



Species a' b c^ 



Plesionika longirostris 

 Heterocarpus ensifer 

 Heterocarpus laevigatus 

 Heterocarpus longirostris 



^Asymptotic value for the ratio of ttie lengths of the appendix 

 masculina and the appendix interna. 



^Carapace length at the inflection point at 50% of the asymptotic 

 ratio. 



343 



