ECHIMYINAE, Gray, 1825

Carvalho, Guilherme A. S. & Salles, Leandro O., 2004, Relationships among extant and fossil echimyids (Rodentia: Hystricognathi), Zoological Journal of the Linnean Society 142 (4), pp. 445-477 : 468-469

THE ECHIMYINAE AND DACTYLOMYINAE

This analysis does not support the monophyly of the extant echimyines; instead, they appear associated with the dactylomyines within a more inclusive Echimyinae /Dactylomyinae-clade (Clade 13). This result agrees with that from the molecular analyses of Lara et al. (1996) and Leite & Patton (2002) in which the members of the two subfamilies were also placed in the same monophyletic clade, although the relationships among the genera differ in the two analyses. This Clade 13 is supported by the absence of the metalophid in M 1 –M 3 [character 15: state 1; reversed (state 0) in Maruchito sp. ]; and the presence of a deep labial opening in the metaflexus of little-worn upper teeth [character 31: state 1; reversed (state 0) in Echimys semivillosus , polymorphic in Makalata didelphoides , uncertain in Isothrix and unknown in Paradelphomys ].

Within Clade 13, Callistomys appears as the most basal extant taxon in a polytomy that also includes a clade with both forms of the fossil genus Maruchito . Emmons & Vucetich (1998) had already called attention to the fact the Callistomys presented intermediate characteristics between eumysopines and echimyines, consistent with its basal position regarding the extant echimyines. These authors also suggested a close proximity between Callistomys and Maruchito .

The remaining extant echimyines do appear as a monophyletic clade (Clade 19), which is the sistergroup of a monophyletic Dactylomyinae (Clade 16). The association of these two clades in Clade 15 is supported by: the presence of a deep sulcus between the protoconid region and the anterolophid in little-worn dP 4 [character 6: state 1; reversed (state 0) in Isothrix and Paradelphomys ]; the presence of a deep sulcus between the lingual end of the protoloph and the posterior portion of the tooth in little-worn upper teeth (character 33: state 1; unknown in Paradelphomys , undetermined in Isothrix and polymorphic in Kannabateomys ); and the presence of four roots in the upper molar teeth (character 36: state 1; unknown in Paradelphomys ).

Clade 19, which includes all extant echimyines except Callistomys , is supported by the presence of a deep sulcus between hypolophid and the ectolophid/ protoconid region in little-worn dP 4 (character 7: state 1) and between the hypocone region and the posteroloph in little-worn upper teeth (character 34: state 1). Note, however, that in Isothrix the former (character 7) is polymorphic and the latter (character 34) is considered as uncertain.

Within Clade 19, Echimys semivillosus and E. chrysurus appear in a basal paraphyletic sequence, followed by a polytomy including Isothrix , Makalata didelphoides , Makalata sp. and a clade (Clade 22), in which M. grandis emerges as the sister-taxon of the Phyllomys – Diplomys clade (Clade 23). This topology, which differs considerably from that obtained by Lara et al. (1996) and Leite & Patton (2002) concerning the extant echimyines, does not corroborate the monophyly of the genera Echimys and Makalata . This lack of concordance between phylogenetic hypotheses and current taxonomy illustrates the urgent need for a thorough taxonomic revision for the Echimyinae .

The Dactylomyinae (Clade 16) comprise a welldefined group, supported in the present analysis by an inferior zygomatic root placed dorsally to the palatal region, which seems projected ventrally (character 46: state 1). Among dactylomyines, Kannabateomys appears as the most basal taxon, followed by Dactylomys and a clade uniting Olallamys and the fossil genus Paradelphomys .

The similarity in tooth morphology of Paradelphomys and the dactylomyines has already been noticed by Patterson & Wood (1982). These authors have questioned, however, whether this resemblance is indicative of affinity. Indeed, it is startling that such an ancient fossil taxon (from the Colhuehuapian – Early Miocene) was placed in such a derived position within echimyid phylogeny. Furthermore, the above-mentioned synapomorphy of Dactylomyinae (character 46: state 1) is unknown for Paradelphomys . Thus, it is possible that this association of Paradelphomys to the extant dactylomyines simply reflects convergence in molar crown structure, and that this fossil genus should group elsewhere as proposed by Vucetich & Verzi (1991). As suggested by Patterson & Wood (1982), this uncertainty will remain until more palaeontological information is gathered on Paradelphomys itself, and on the Tertiary fossil dactylomyines.

Seven fossil taxa are positioned basally to the Echimyinae – Dactylomyinae clade and may be considered as stem echimyines/dactylomyines. Although it is often taken that echimyines and dactylomyines have no known fossil record in the Tertiary ( Patterson & Wood, 1982; McKenna & Bell, 1997), the association of at least some of these taxa to extant echimyines and/ or dactylomyines has already been suggested by some authors (e.g. Vucetich & Verzi, 1991; Vucetich et al. 1993a; Emmons & Vucetich, 1998).

Indeed, Spaniomys , Adelphomys and Stichomys have been associated to the Dactylomyinae ( Vucetich & Verzi, 1991) and to Maruchito ( Vucetich et al. 1993a) , which itself has been associated to the echimyine Callistomys ( Emmons & Vucetich, 1998) . In the present analysis, all these genera do appear close to each other near the basis of the crown-group Echimyinae / Dactylomyinae , which agrees, at least partially, with the relationships suggested above.

The association of Protacaremys , Prospaniomys , Acarechimys and Caviocricetus has been previously suggested by Vucetich & Verzi (1991, 1996), who mention that these genera should be grouped with the fossil Deseadomys , Xylechimys and Platypittamys , as well as with the extant Mesomys and Lonchothrix . The present analysis does not corroborate such a group ( Deseadomys , Platypittamys and probably Xylechimys appear elsewhere), but it does show these four fossil taxa closely associated in a continuous paraphyletic sequence, which includes also the clade Mesomys – Lonchothrix .