GOPALAKRISHNAN: BIOLOGY AND TAXONOMY OF NEMATOSCELIS 



overlapping regions of their distribution. The fact 

 that the observed diagnostic feature hes on the 

 reproductive structure would suggest probable 

 genetic separation. Nonetheless, until more is 

 known about the ecology and behavior of these 

 forms, I do not wish to formally describe them as 

 species or subspecies and will consider them 

 ecophenotypes. The pattern of geographical dis- 

 tribution of these forms appears to be correlated 

 with differences in environmental characteristics, 

 particularly the distribution of dissolved oxygen 

 in the water column (Gopalakrishnan 1974). 



Allopatric populations are inferred to have un- 

 dergone reproductive isolation if they are 

 morphologically distinct and do not show any 

 overlap in their diagnostic features. Nematoscelis 

 diffieiUfi. and A'^. megalops are allopatric, occupying 

 the northern and southern transitional zones of 

 the Pacific respectively. Nematoscelis megalops 

 also occurs in the Atlantic and Indian oceans. On 

 the basis of similarities in the structure of the 

 petasma of A'^. difficUis and A^. megalops, Karedin 

 (1971) questioned the validity of A^. difficilis. 

 Brinton (1962) considered them a sibling species 

 pair evolved as a result of complete geographical 

 separation. Although closely related, there are 

 certain morphological features that distinguish 

 one species from the other. Both quantitative and 

 qualitative features of the reproductive system, 

 the petasma in this case, indicate significant 

 differences. The petasmae of A^. megalops in the 

 South Pacific, South and North Atlantic, and South 

 Indian oceans show no apparent structural 

 difference. These populations are probably in 

 continuum facilitating gene exchange. (Com- 

 munication between North and South Atlantic 

 populations of N. megalops appears possible from 

 the fact that only in this ocean does the 

 characteristically subtropical species A", atlantica 

 occur also in the tropics. This suggests that in the 

 Atlantic the low-latitude boundaries of distribu- 

 tion of subtropical species approach the equator, 

 permitting at least occasional north-south com- 

 munication). The validity of A", difficilis can be 

 supported by the fact that this species and N. 

 megalops do not overlap geographically in the 

 Pacific Ocean. Both species live in comparable en- 

 vironments (narrow mid-latitude zones) and 

 therefore are probably exposed to similar selection 

 pressures. In such a situation, even though 

 geographical isolation would be complete, 

 morphological differentiation might be slow. In 

 the absence of gene exchange, the populations 



would be expected to have diverged genetically. 

 The relative lengths and shapes of the median 

 lobe, proximal, terminal, and lateral spines of pe- 

 tasmae of the two species differ, supporting the 

 validity of A^. difficilis. In Hansen's words (1916) 

 these differences are "certainly so sharp, so im- 

 portant, and so constant that they are sufficient 

 for separating N. difficilis from A'^. megalops." 



The A'^. atlantica population in the North Pacific 

 is spatially separated from its counterpart in the 

 South Pacific Ocean. No morphological distinct- 

 ness was evident in this population, although it 

 may prove to be genetically separate from others. 

 Nematoscelis lobata is endemic to the Sulu and 

 Celebes seas in the Indo-Australian Archipelago 

 (Gopalakrishnan 1974). According to Hansen's 

 (1916) description, this species is very similar to N. 

 microps. The present study indicates that in many 

 morphological characters N. lobata is more related 

 to A'^. atlantica than to A", microps. Nematoscelis 

 lobata and A'', atlantica are allopatrically dis- 

 tributed and have probably acquired characters 

 which promote or guarantee their reproductive 

 isolation. 



The observed differences in the structure of the 

 first thoracic leg (maxilliped) of species of Nema- 

 toscelis indicate presumed specialization in feed- 

 ing habits. All species of this genus are recognized 

 to be omnivores. From a comparison of the first 

 thoracic legs, it appears that species would be ex- 

 pected to show different types of feeding. Nema- 

 toscelis microps, N. atlantica, and N. lobata, hav- 

 ing marginal setae (one row) on their propodus of 

 the first thoracic legs, may be better fitted for 

 filtering a large proportion of phytoplankton in 

 their food, whereas N. difficilis and A^. megalops, 

 having three rows of setae on the propodus of the 

 first leg, may select more animal food. In this re- 

 spect A^. gracilis and A'^. tenella are intermediate. 

 Existing information on the gut contents of 

 species of Nematoscelis, e.g., Nemoto (1967) and 

 Weigmann (1970), is inadequate to substantiate 

 this. 



ACKNOWLEDGMENTS 



This work was supported in part by National 

 Science Foundation Grant GA31783 and in part by 

 the Marine Life Research Program, the Scripps 

 Institution of Oceanogrpaphy's component of the 

 California Cooperative Oceanic Fisheries Inves- 

 tigations, a project sponsored by the Marine 

 Research Committee of the State of California. I 



813 



