FLEMINGER and HULSEMANN: FOUR SIBLING SPECIES OF PONTELLINA 



2 mm 

 I 1 



a -c 



Figure 34. — ThIV-V and urosome of female PonielUna sohrina with attached spermatophore: a. dorsal view; b. lateral 



view; c. ventral view. 



The program calculates an index of affinity 

 for all possible pairs of species as the geometric 

 mean of the proportion of common character 

 states corrected for the number of characters 

 used in the analysis: i.e., Jl\/A X B — Viy/W , 

 where A and B are the total number of charac- 

 ters recorded for the two species, B ^ A, and 

 J is the number of shared character states. 

 Values of the index above 0.5 have been found 

 empirically to provide objective repeatable 

 groupings of related sets of values. 



A nuniber of characters listed in Table 9 were 

 omitted from the recurrent group analysis to 

 avoid biasing the computations with redundant 

 information. Characters 1, 2, and 4 were not 

 scored since their morphometric states are 

 determined largely by the value of character 3. 

 Characters 11 and 14 were deleted since they 

 parallel character 10 in showing a direct re- 

 lationship to TL and to furcal length. In the 

 absence of a fossil record the distributional 

 characters 16 and 17 were not scored on the 

 intuitive grounds that they are complex deriv- 

 atives of both 1) overall genetic adaptation and 

 2) fortuitous abiotic historical events that 

 might obscure essential phylogenetic patterns. 

 All of the scored characters were weighted 

 equally and disregard the preliminary results 

 from SEM observations. 



The recurrent group analysis reveals only 

 one grouping with an index higher than 0.5, 

 that of morix and sobnna (Table 10). Similar- 

 ity between phonata and platychela falls well 

 below the acceptable level of significance. The 

 other possible pairings are dissimilar in most 

 to virtuallv all of the 10 characters used in the 



analysis. Assuming equal rates of evolution 

 the results indicate that the divergence of 

 pliimata, platychela, and the morii-fiobiiua 

 lineage are likely to be divisions of greater age 

 than that of morii and sohrina. 



GEOGRAPHICAL VARIATION 

 AND SYMPATRY 



In the course of this study two conspicuous 

 and parallel instances of geographical varia- 

 tion were encountered in the fifth legs of phauata 

 females. This variation was expressed in the 

 number of spines at the distal end of the endopod 

 and the length of the exopod relative to that of 

 the endopod. 



As noted above the number of spines on the 

 ' endopod of the female's fifth leg is polymorphic 

 throughout Pontellina. The bilateral two-spined 

 morph is overwhelmingly dominant in morii 

 and sobri)ia. However, four morphs are common 

 in platychela and plumata. Comparison of ran- 

 domly selected samples of Atlantic specimens 

 of pluniata with specimens from the Indian and 

 Pacific Oceans indicate significant differences 



Table 10. — Values of the recurrent group affinity index 

 (Fager, 1969) and the probability of obtaining this or high- 

 er values by chance for all possible pairs of Pontellina spp. 

 Further discussion in text. 



Species pair 



Affinity index 



103 



