18 



Wyss 1994), although it may alter the implied evolutionary pathway of selected characters. 

 However, given a reasonable degree of completeness (see Huelsenbeck 1991b), the overall 

 potential advantages of including fossil evidence cannot be discounted. 

 Although there are suggestions in the literature that molecular data may operate more 

 effectively than morphological data at lower taxonomic levels (e.g., Novacek 1993), 

 morphological data were used exclusively. This largely reflects the availability of such 

 data for all the desired species. Among phocids, both Mirounga spp. and Monachus spp. 

 are CITES-listed animals (Anonymous 1992), and all pinnipeds are subject to the Marine 

 Mammal Protection Act, making the acquisition of fresh samples as a source of molecular 

 data difficult. As well, the full potential of morphological data at low taxonomic levels 

 may not have been properly exploited yet, with the use of non-traditional or multistate 

 characters (see Character Analysis and Bryant 1989), possibly derived from techniques 

 such as morphometric analysis (see Discussion and Conclusion), hopefully improving the 

 effectiveness of this type of data in such cases. 



Data were obtained primarily from osteological specimens (see also Characters below). 

 This was necessitated by the tendency of museums to preserve mammals as skulls, 

 skeletons, and study skins. Furthermore, as phocids are fairly large mammals, many 

 specimens are represented by skulls alone. Generally, the best available (i.e., most 

 complete and undamaged) specimens for a given species were selected for study while 

 attempting to maintain an equal sex ratio. This latter point was especially important for 

 such grossly sexually dimorphic taxa as Zalophus californianus and Mirounga spp. 

 Damaged specimens were occasionally employed to view various internal characters of 

 the skull. Missing data were substituted by literature values wherever possible. 

 Numerous specimens of each taxon were examined in order to take account of intraspecific 

 variation. This was especially important for the pinnipeds, as they apparently display an 

 inordinate amount of intraspecific variation, primarily in their cranial characters (Mivart 

 1885; Doutt 1942; Davies 1958b; Ray 1976b). With respect to phocids, King (1966) has 

 also commented on how the large intergeneric differences of the skull confound 

 comparisons between the genera, and on the potential problems resulting from the high 

 intraspecific variability of the teeth. (She does add, however, that the tympanic region 

 seems to be relatively stable.) Unfortunately, however, postcranial material, and especially 

 the distal elements of the limbs, were typically only obtainable from a restricted number 

 of specimens. An extreme case is for Pusa caspica, where all postcranial observations 

 were derived from a single individual. 



Although Hennig (1966) notes that characters can be taken from any life stage of an 

 organism (i.e., any semaphoront), juvenile individuals were also largely excluded from 

 this study. This primarily reflects the very poor representation of juvenile specimens in 

 museum collections. However, this decision secondarily served to minimize the already 

 high intraspecific variation observed in phocids (see above) by avoiding comparisons 

 between vastly different age classes. 



Characters 



A total of 196 characters were examined in this study (see Appendix B). Characters were 

 selected so that they were at least theoretically observable from the material typically 



