FISHERY BULLETIN: VOL. 72. NO. 1 



in the frequency of the four morphs (Table 11). 

 The bilateral one-spined morph is much more 

 frequent than the bilateral two-spined morph in 

 the Atlantic Ocean whereas in the Indian and 

 Pacific Oceans the frequencies of the two morphs 

 are more alike and the differences are not signi- 

 ficant. 



Notably, the frequencies of the morphs in 

 platyckela differ significantly from those of the 

 pluniata sample from the Atlantic but not from 

 those of the Indian and Pacific samples of phtiii- 

 ata. Of the j^airs of species sharing common 

 boundaries only niorii and sobrina do not appear 

 to have appreciably different morph frequencies 

 (Table 11). 



In the ratio exopod to endopod length for 

 the fifth legs in females, the distribution of the 

 Atlantic sample of plumata differed significantly 

 from those of the Indian and Pacific Oceans 

 (Table 12). As in the case of the endopodal spines, 

 pairs of species with common boundaries show- 

 ed significantly different distributions of the 

 exopod-endopod ratios. 



Extrapolating from the similarity of Indian 

 and Pacific Ocean samples, differences between 

 Atlantic and Indian-Pacific populations of 

 plumata appear to be geographically abioipt. 

 Thus they may be viewed as refiecting 1) local 



pressures on phi mat a within each geographical 

 population, 2) restricted gene flow between the 

 Indian and Atlantic Ocean populations, 3) or 

 both. P. plumata showed the highest frequency 

 of unique character states in PontelUna (Table 

 9). Furthermore it tends to occupy a conspicuous- 

 ly peripheral position relative to the other three 

 species in its dimensions of sexually modified 

 appendages in both sexes (e.g., Figures 27, 28, 

 30, 31). It is also the only species of the genus 

 sharing common boundaries with the other 

 three species of Poitt('lUna. Therefore, the geo- 

 graphical differences between Atlantic and 

 Indian-Pacific populations of plumata parallel 

 the extensive character divergence (Mayr, 1970: 

 51-53) otherwise distinguishing the species. Con- 

 sidering the fact that all morphological structures 

 involved are sexually modified it would appear 

 that we are witnessing reinforcement of pre- 

 mating barriers (Dobzhansky, 1970:376-382). 



Similar disjunct morphological differences 

 distinguishing Atlantic from Indian and Pacific 

 populations of epipelagic calanoids have been 

 reviewed (Fleminger and Hulseraann, 1973) 

 and the number of examples increased (Flemin- 

 ger and Hulsemann, 1973; Fleminger, 1973). 

 Similar patterns in the strength of divergence 

 in secondary sexual characters relative to 



Table 11. — X^ test of homogeneity in the distribution 

 spines on the endopod of the fifth legs in adult females. 



of 



Number of spines on endopod 

 (left leg-right leg): 



2-2 



1-2 



2-1 



1-1 



Population 



Number of specimens 



Total 



platychela. Atlantic Ocean 

 plumata, Atlantic Ocean 

 plumata. Pacific Ocean 

 plumata, Indian Ocean 

 morii, Indian Ocean 

 morii. Pacific Ocean 

 sobrina, eastern tropical 

 Pacific Ocean 



plumata Atl. vs. pliiinata Pac. 

 plumata Atl. vs. plumata Ind. 

 plumata Pac. vs. plumata Ind. 

 platychela vs. plumata Atl. 

 platychela vs. plumata Pac. 

 platychela vs. plumata Ind. 

 morii Pac. vs. sobrina 

 morii Ind. and Pac. vs. plumata 



Ind. and Pac. 

 sobrina vs. plumata Pac. 



X2 



d.f. 



84.34 



< 0.001 



104 



