Of the 4,629 species of mammals recognized by Wilson and Reeder, rodents represent 43% of species diversity within the class Mammalia. Several diagnostic characteristics associated with general morphology define rodents as a monophyletic group (e.g., group sharing a common ancestry). The primary characteristic is the pair of long incisors, resulting from the loss of canines and the creation of a diastema or gap between the incisors and cheek teeth, consisting of premolars and molars.
Other characteristics of the skull and skeleton, dentition, and basic soft anatomy tend to support monophyly. Although the monophyly of Rodentia appears well supported on the basis of examinations of morphological traits of living forms as well as fossil lineages, several studies based primarily on amino acid sequence data from nuclear genes and some limited analysis of nucleotide sequence data have suggested that the order Rodentia is not monophyletic. According to these molecular studies, the guinea pig and its relatives group closer with other orders of mammals than they do to families of rodents, especially rats and mice of the family Muridae. However, as more extensive molecular studies have been initiated and more sophisticated analyses that take into account complex patterns of variation within rodents have been performed, molecular support for rodent monophyly has been forthcoming.
The order Rodentia presents some serious problems with respect to classification, especially the recognition of suborders. As specializations in rodents tend to focus on the masticatory apparatus (e.g., structure of the lower jaw, shape of the infraorbital foramen on the anterior end of the skull, and placement of musculature associated with jaw action), early classifications were based on the origin and insertion of jaw muscles (e.g., masseter muscles) and the arrangement of changes in portions of the skull associated with this musculature. Therefore, the more basic classifications focus on either changes in the infraorbital foramen and the placement of masseter muscles or structure of the lower jaw. Using these characteristics, the number of recognized rodent suborders has varied, depending upon the proposed classification scheme, from three (Sciuromorpha, squirrels; Myomorpha, rats and mice; and Hystricomorpha, guinea pigs and relatives) to as many as 16.
Considering that the origin and insertion of jaw muscles and the structure of the skull relate directly to feeding, these features reveal high levels of parallel evolution (similar features independently derived). Molecular data, derived from nucleotide sequences, have confirmed the fallacy of using features that tend to converge in structure in response to similar functions. As a consequence of problems related to the origin and insertion of jaw muscles and modifications of the skull to accommodate such changes, the most consistent morphological feature is associated with the lower jaw. Based on this feature, two suborders of rodents can be identified, Hystrocognathi and Sciurognathi. The former suborder contains South American caviomorph rodents (guinea pigs and their relatives) and several Old World families including porcupines (Hystricidae), cane rats (Thryonomyidae), African mole-
In many cases, the phylogenetic relationships among families and genera of rodents are still controversial. Based on morphology and recent molecular data, families in the suborder Hystricognathi appear most closely related to the gundis of the family Ctenodactylidae. Several additional groups sharing a common ancestry appear to be well supported including: rats and mice of the superfamily Muroidea (families Muridae and Dipodidae); pocket gophers (family Geomyidae) and pocket mice and kangaroo rats (family Heteromyidae) of the superfamily Geomyoidea; squirrels (family Sciuridae) and mountain beavers (family Aplodontidae). The placement of other families like the Castoridae (beavers), Pedetidae (springhares), Anomaluridae (scaly-
Paleontologically, the order Rodentia dates to the Paleocene era at approximately 57 to 60 million years ago (mya). The earliest known family is the Paramyidae, containing sciurid-
One biogeographic enigma pertains to the origin of the South American caviomorph families of suborder Hystricognathi. According to fossil evidence, the family is no older than late Eocene to early Oligocene (36 to 40 mya), yet the closest relatives of this group occur in Africa, a continent separated from South America by thousands of miles (kilometers) during the period at which caviomorphs first appear in South America. Although some paleontologists suggest that caviomorph rodents evolved the characteristics shared with African hystricognaths in parallel, all recent morphological and molecular data suggest that these two groups share a common ancestry. Based on shared ancestry, other paleontologists and biogeographers have suggested an over-