46 



hinted at by Arnason et al. (1993)] which is the sister taxon to the remaining Phocini (plus 

 Erignathus) here. Phoca (sensu stricto) is also paraphyletic, despite the view of some 

 authors that Phoca largha is merely a subspecies of Phoca vitulina (Scheffer 1958; Burns 

 1970; Shaughnessy 1975; Baram et al. 1991). The relatively basal position of P. largha 

 initially appears weak, as the taxa internal to it are only united by three to five synapo- 

 morphies. However, such a position for P. largha (with respect to P. vitulina at least) is 

 indicated by McLaren (1975) and possibly by Mouchaty et al. (1995). As well, the low 

 number of synapomorphies may be an artifact of the polytomy in this region. With respect 

 to a strictly dichotomous branching arrangement, the polytomy requires that putative 

 synapomorphies for this node (#32; see Fig.5B) satisfy an increased number of descendent 

 lineages (four here). Presumably fewer synapomorphies exist to fulfill this more difficult 

 condition, than for dual descendent lineages (e.g., as in node #30) with the possibility of 

 further changes or reversals further along in the tree. In any case, a clade of P. largha and 

 P. vitulina is not formed in either of the two most parsimonious solutions. 

 The polytomy within the Phocini also prevents a clear assessment of the status of Pusa. 

 The consensus solution is equivocal; however, one of the two equally most parsimonious 

 solutions does reveal a monophyletic Pusa (Fig.5A). At the very least, a sister taxon 

 relationship for Pusa caspica is indicated with respect to the remaining Pusa spp., as 

 hinted at by Chapskii (1955b). However, the sister taxon status of Pusa to the remaining 

 Phocini, as put forth by McLaren (1975), is not supported here. 



Histriophoca and Pagophilus form a reasonably well supported monophyletic group. This 

 substantiates the long standing, but historically poorly tested claim that the two genera are 

 closely related, or at least very similar to each other (Chapskii 1955a; Davies 1958b; 

 McLaren 1975; de Muizon 1982a; Arnason et al. 1995; Mouchaty et al. 1995). De 

 Muizon's (1982a) hypothesis that any similarity may be exclusively due to symplesiomor- 

 phies is not supported. 



Summary of ingroup relationships 



Overall, the internal relationships of the phocines are comparatively weakly supported in 

 terms of numbers of synapomoiphies, especially for those taxa internal to Halichoerus 

 (Fig.5C). In fact, the internal relationships of the monachines are generally much better 

 (and more uniformly) supported, despite the weaker support for the subfamily as a whole. 

 The new relationships proposed within each subfamily are somewhat difficult to reconcile 

 with previous opinion. This is especially true for the phocines, which have been well 

 studied for the most part, and especially for the novel position of Erignathus advocated 

 here. The apparent polytomy within the Phocini (plus Erignathus here) is suggested by 

 the numerous conflicting taxonomic assessments and biogeographic or systematic 

 hypotheses for this group (e.g., compare Chapskii 1955a; Davies 1958b; McLaren 1966, 

 1975; Burns & Fay 1970; Ray 1976a; Repenning et al. 1979; de Muizon 1982a; Arnason 

 et al. 1995; Mouchaty et al. 1995; Perry et al. 1995). This lack of resolution is probably 

 traceable to the rapid adaptive radiation of this group in the post-early Pliocene and/or 

 Pleistocene (Ray 1976a), allowing insufficient time for its members to become clearly 

 differentiated (at least morphologically; see Arnason et al. 1995). Evidence from this study 



