The ecology of a continental evolutionary radiation: Is the radiation of sigmodontine rodents adaptive?

Evolutionary radiations on continents are less well‐understood and appreciated than those occurring on islands. The extent of ecological influence on species divergence can be evaluated to determine whether a radiation was ultimately the outcome of divergent natural selection or else arose mainly by nonecological divergence. Here, we used phylogenetic comparative methods to test distinct hypotheses corresponding to adaptive and nonadaptive evolutionary scenarios for the morphological evolution of sigmodontine rodents. Results showed that ecological variables (diet and life‐mode) explain little of the shape and size variation of sigmodontine skulls and mandibles. A Brownian model with varying rates for insectivory versus all other diets was the most likely evolutionary model. The insectivorous sigmodontines have a faster rate of morphological evolution than mice feeding on other diets, possibly due to stronger selection for features that aid insectivory. We also demonstrate that rapid early‐lineage diversification is not accompanied by high morphological divergence among subclades, contrasting with island results. The geographic size of continents permits spatial segregation to a greater extent than on islands, allowing for allopatric distributions and escape from interspecific competition. We suggest that continental radiations of rodents are likely to produce a pattern of high species diversification coupled with a low degree of phenotypic specialization.     

A newly recognized family from the Horn of Africa,the Heterocephalidae (Rodentia: Ctenohystrica)

The Ctenohystrica is one of the three major lineages of rodents and contains diverse forms related to gundis, porcupines, and guinea pigs. Phylogenetic analyses of this group using mitochondrial and nuclear gene sequences confirm the monophyly of the infraorder Hystricognathi and most of its recognized subclades, including both the Neotropical caviomorphs and the African phiomorphs, which are recovered as sister groups. Molecular timetrees calibrated with 22 securely placed fossils indicate that hystricognath superfamilies originated in the Eocene and Oligocene and most families had appeared by the end of the Oligocene, ～23 Mya. Divergences leading to hystricognath genera took place in the Miocene and Pliocene, with a single exception. The naked mole‐rat (Heterocephalus) diverged from other African mole‐rats (Bathyergidae) in the early Oligocene (～31.2 Mya), when the four caviomorph superfamilies (Erethizonoidea and Cavioidea at 32.4 Mya, Chinchilloidea and Octodontoidea at 32.8 Mya) were first appearing in South America. The extended independent evolution of Heterocephalus suggested by this analysis prompted a closer examination of mole‐rat characters. Heterocephalus indeed shares many characters with bathyergids, befitting their joint membership in the parvorder Bathyergomorphi and superfamily Bathyergoidea as well as their shared exploitation of subterranean lifestyles. However, a diverse array of cranial, dental, postcranial, external, and ecological characters distinguishes Heterocephalus from other African mole‐rats. These differences equal or exceed those used to diagnose caviomorph families and justify recognizing the naked mole‐rat in its own family, Heterocephalidae Landry, 1957. This taxonomic arrangement poses questions for the inter‐relationships of fossil and extant mole‐rats and brings time equivalence to the ranks assigned to the major clades of hystricognaths. © 2014 The Linnean Society of London