Phylogeny of Neotropical oryzomyine rodents (Muridae: Sigmodontinae) based on the nuclear IRBP exon

Sigmodontine rodents are the most diverse family-level mammalian clade in the Neotropical region, with about 70 genera and 320 recognized species. Partial sequences (1266 bp) from the first exon of the nuclear gene encoding the Interphotoreceptor Retinoid Binding Protein (IRBP) were used to infer the phylogenetic relationships among 44 species representing all 16 currently recognized genera of the largest sigmodontine tribe, the Oryzomyini. Monophyly of the tribe was assessed relative to 15 non-oryzomyine sigmodontine taxa representing all major sigmodontine lineages. Twelve taxa from seven muroid subfamilies were used as out-groups. The resulting matrix included 71 taxa and 386 parsimony-informative characters. Phylogenetic analysis of this matrix resulted in 16 equally parsimonious cladograms, which contained the following well-supported groups: (i) a monophyletic Oryzomyini, (ii) a clade containing all oryzomyines except Scolomys and Zygodontomys, (iii) a clade containing Oecomys, Handleyomys, and several species of forest-dwelling Oryzomys, and (iv) a clade containing the remaining oryzomyine taxa. The last clade is composed of two large subclades, each with lower nodal support, containing the following taxa: (i) Microryzomys, Oligoryzomys, Neacomys, and Oryzomys balneator; (ii) Holochilus, Lundomys, Pseudoryzomys, Nectomys, Amphinectomys, Sigmodontomys, and several species of open-vegetation or semiaquatic Oryzomys. Regarding relationships among non-oryzomyine taxa, sigmodontines, neotomines, and tylomyines do not form a monophyletic group; a clade containing Rheomys and Sigmodon is basal relative to all other sigmodontines; and the remaining sigmodontines are grouped in three clades: the first containing Thomasomyini, Akodontini, and Reithrodon; the second containing Abrothrichini, and Phyllotini, plus Wiedomys, Juliomys, Irenomys, and Delomys; and the third containing the oryzomyines. No conflict is observed between IRBP results and previous robust hypotheses from mitochondrial data, while a single case of incongruence is present between the IRBP topology and robust hypothesis from morphological studies. (C) 2003 Elsevier Science (USA). All rights reserved.