EVOLUTIONARY IMPLICATIONS OF CHROMOSOMAL HOMOLOGY IN FOUR GENERA OF STENODERMINE BATS (PHYLLOSTOMATIDAE: CHIROPTERA)

Phylogenies based on similarity of standard karyotypes of representatives of 17 of the 18 stenodermine (Stenoderminae: Phyllostomatidae) genera have been published (Baker, 1973; Greenbaum et al., 1975; Gardner, 1977). The phylogenies of Baker (1973) and Greenbaum et al. (1975) are based on the supposition that the primitive karyotype for the subfamily had a 2n = 30 or 32 and a fundamental number of 56-60. The phylogeny of Gardner (1977:313) is based on the assumption that the primitive karyotype for the subfamily had a "diploid number near or above 40 and a low fundamental number." A G~banding study (Patton, 1976) of 10 genera in the subfamily Phyllostomatinae (and representatives of the Noctilionidae and Mormoopidae as well) revealed that the karyotype which requires the least number of convergent events to derive karyotypes of living forms consisted of a 2n = 46; FN = 60. Although Gardner's proposed phylogenies and the study by Patton do not prove that the primitive karyotype for the common ancestor of living stenodermine genera is higher than 30, they do raise a valid question as to the degree of homology of the biarmed chromosomes of the respective karyotypes of those forms with 28 autosomes. Also see the discussion by Hecht and Edwards (1976) on the problem of convergence and systematics. To obtain an assessment of the degree of homology among stenodermine bats, we have examined the G-banding pattern of the chromosomes of representatives of the genera Sturnira, Artibeus, Enchisthenes and Uroderma. We chose Sturnira and Artibeus because these two genera represent living species of two lineages that are presumed by Smith (1976) (based on dental and osteological characters) to share a common ancestor that could have given rise to all stenodermine genera. Additionally, Artibeus and Sturnira have autosomal complements that are similar in shape and number (28). If for each Artibeus autosome there is an identically banded autosome in the karyotype of Sturnira, then the autosomes are almost certainly the same as those possessed by the last common ancestor of these genera. However, if the karyotypes of these genera were independently derived from a karyotype like that proposed as primitive for the family Phyllostomatidae (with a 2n = 46, Patton, 1976 and 2n = 40, Gardner, 1977) then the similarity in number and morphology of the autosomes would be the result of convergence and the degree of banding homology should be low. Specimens of the 2n = 44 cytotype of Uroderma bilobatum were examined because this is the karyotype most like that proposed by Gardner (1977) as primitive for the subfamily. Gardner proposes that such a karyotype underwent pericentric inversions (and possibly additions of heterochromatin) to produce a 2n = 30; FN = 56 karyotype. We examined Enchisthenes to determine how it differed from Artibeus, the genus to which it is most closely related. C-banded karyotypes from these species were studied to determine whether addition of heterochromatic arms was responsible for any of the au-