BADCOCK and BAIRD: SYSTEMATICS OF STERNOPTYX 



Table 5. — Range of depths of maximum abundance of species of 

 Sternoptyx in sympatry at three locations (subadults and adults I 

 (Hopkins and Baird 1973; Badcock and Merrett 1976). 



Species 



Lat 2r N, 

 long, 86' W 



Lat. 18° N, 

 long 25° W 



Lai 11" N. 

 long 20 W 



S. diaphana 



S- pseudobscura 



S. pseudodtaphana 



600-750 m 

 850-1,000 m 

 Not present 



600-800 m 

 800-1.500 m 

 600-1,500 m 



500-700 m 

 800-1,000 m 

 600-1.000 m 



both congeners (Table 5). The shoaling oi Sternop- 

 tyx spp. between lat. 18° N and 11° N is not a 

 function of developmental state and is a feature 

 shown by many species of midwater fishes (Bad- 

 cock and Merrett 1977). 



In other areas of the eastern North Atlantic and 

 also in the Gulf of Mexico where S. pseudobscura 

 and S. diaphana share the water column, popula- 

 tions of S. pseudobscura are always centered 

 below those of S. diaphana. Data presented by 

 Baird (1971) indicated that S. pseudodiaphana is 

 usually taken in 800-1,200 m depth in the South- 

 ern Ocean. Evidence from recent collections from 

 the South Atlantic imply a similar pattern of ver- 

 tical distribution.^ In general, then, S. pseudodia- 

 phana and S. pseudobscura may be regarded as 

 deeper dwelling species of Sternoptyx while S. 

 diaphana is a shallower living form. In certain 

 areas of the Atlantic, discrete sampling has shown 

 S. diaphana to be centered deeper than indicated 

 above (Badcock 1970; Badcock and Merrett 1976; 

 Roper et al.^). Sternoptyx pseudobscura has been 

 shown to be of low abundance in these areas but 

 the deepening of S. diaphana is likely to be a 

 consequence of the sinking of isotherms relative to 

 other Eireas. The role of competitive interactions 

 among these species is yet undocumented and 

 these may also exert an effect on geographic pat- 

 terns of vertical distribution. Data on S. obscura 

 are not comprehensive, but a preliminary survey 

 of maximum depth of open trawl collections indi- 

 cate a depth range similar to S. diaphana (500- 

 1,000 m) in basins off southern California. 



An analysis of the vertical distribution of mid- 

 and late-metamorphic stages in the eastern North 

 Atlantic was possible only for S. pseudodiaphana 

 because of the problems in distinguishing between 

 such individuals of the other two species ex- 

 amined. As with subadults and adults, indi- 

 viduals of like developmental stage lay shallower 



'G. Krefft, Instutit fiir Seefischerei, Hamburg, West Germany, 

 pers. commun. 1976. 



'Roper, C.F.E.,R.H.Gibbs, Jr., and W.Aron. 1970. Ocean 

 acre: an interim report. Report to the U.S. Navy Underwater 

 Sound Laboratory. Contract No. N00140-69-C-0166. Smithson. 

 Inst., Wash., DC., 22 p. 



in the water column at lat. 1 1° N, long. 20° W than 

 at lat. 18° N, long. 25° W (400-800 m versus 

 500-900 m depth). Although the data are sparse, 

 there is evidence for ontogenetic vertical strat- 

 ification among metamorphic stages. At lat. 1 1°N, 

 long. 20° W, Stages 1-3 occurred only in 400-500 m 

 depth; Stage 4 in 400-600 m; Stage 5 in 500-700 m; 

 and Stage 6 in 500-800 m. A similar relationship is 

 implied for metamorphic stages from lat. 18° N, 

 long. 25° W, although stratification occurred 

 deeper in the water column. 



CONCLUSIONS 



The evidence presented shows Sternoptyx to 

 contain four closely related species. Morphological 

 distinctions between them are relatively slight, 

 but are consistent among the populations 

 examined. The four species have broad geographic 

 ranges and the limited data indicate the occur- 

 rence of geographic variation in S. pseudobscura 

 and S. pseudodiaphana at adult and postlarval 

 levels. Thus systematic difficulties arise in that 

 certain characters useful in distinguishing species 

 in sympatry may overlap when measurements 

 from other populations are included. 



Characters subject to allometric growth simi- 

 larly present systematic problems. Nevertheless, 

 most of the morphological criteria used here to 

 separate species are maintained irrespective of 

 population or developmental state. When found in 

 sympatry, distinctions are clear and species con- 

 sistently separable by many characters. 



While we distinguish two species pairs on the 

 basis of anal fin pterygiophore configuration, no 

 hypothesis of cladistic relationship among the 

 species is advanced. Considering the highly 

 specialized and peculiar morphology of the genus 

 (Baird 1971; Baird and Eckhardt 1972; Weitzman 

 1974; for discussion of family relationships), the 

 most parsimonious hypothesis advanced is that a 

 single ancestral species evolved which diverged 

 considerably from a more generalized hatchetfish 

 stock. Subsequent speciation in the genus proba- 

 bly involved the isolation of populations which 

 now show very slight morphological divergence, 

 exhibit various degrees of geographic variation, 

 and have distinct horizontal and vertical patterns 

 of distribution. 



ACKNOWLEDGMENTS 



Our thanks go to R. H. Gibbs, Jr. (Smithsonian 



817 



