The monophyly of dipodids is strongly established (Ellerman, 1940; Klingener, 1964, 1984; Shenbrot, 1992; Shenbrot et al., 1995; Stein, 1990; Vinogradov, 1930). Authors have usually recognized either a single family, Dipodidae (Ellerman, 1940, 1961; Ellerman and Morrison-Scott, 1951; Holden, 1993; Hugueney and Vianey-Liaud, 1980; Klingener, 1964, 1984; Kowalski, 2001; McKenna and Bell, 1997; Ognev, 1948; Qiu and Storch, 2000; Savinov, 1970; Thomas, 1896; Vinogradov, 1930, 1937; Wang and Qiu, 2000; Winge, 1887), or two families, Zapodidae (including sicistines and zapodines), and Dipodidae (Corbet, 1978c; Corbet and Hill, 1992; Daxner-Höck, 1999; Lyon, 1901; R. A. Martin, 1994; Miller and Gidley, 1918; Pocock, 1922b; Schaub, 1958; Simpson, 1945; Vinogradov, 1925; Wang, 1985; Wilson, 1949; Wood, 1955). Whether one family or two, modern arrangements of subfamilies stems from Vinogradov’s (1925, 1930, 1937) classic research and classification, which provided the foundation for systematic research of dipodids. Phylogenetic studies have supported much of Vinogradov’s original classificatory arrangement, though the proposed relationships among some of the taxa have changed (Ellerman, 1940; Pavlinov and Shenbrot, 1983; Shenbrot, 1992; Stein, 1990; Zazhigin and Lopatin, 2000a).

Vinogradov went from accepting two families in 1925 to one in 1930 when he recognized Zapodinae, Euchoreutinae, Cardiocraniinae, Allactaginae, and Dipodinae, which was followed by Ellerman (1940), who also listed Sicistinae. Pavlinov and Shenbrot (1983) surveyed anatomy of male reproductive tracts among dipodines and recognized Euchoreutinae, Allactaginae, Cardiocraniinae, Salpingotinae, and Dipodinae within a Dipodidae. Later, Shenbrot (1992) incorporated male reproductive morphology, coronal structure of molars, and bullar anatomy that resulted in an eclectic classification formed from both cladistic analysis and degree of morphological divergence (phenetic distance). He recognized Allactagidae, Dipodidae (containing Cardiocraniinae, Paradipodinae, and Dipodinae), Sminthidae (with Sminthinae [= Sicistinae] and Euchoreutinae) and Zapodidae within a larger Dipodoidea, an arrangement also used by Pavlinov et al. (1995) and Shenbrot et al. (1995). In a phylogenetic study based on limb myology, Stein (1990) proposed retention of only two families, a primitive Sicistidae and derived Dipodidae (including zapodines, Euchoreutinae, and all other dipodids); in her scheme, Euchoreutinae is a sister group to zapodines and allactagines. Shenbrot’s arrangement differed significantly from that of Steins’ in that Sicista was not shown to be the most primitive dipodoid (Zapodidae and Sicistidae form an unresolved dichotomy), and Euchoreutinae was hypothesized to be most closely related to Sicistinae and united with Sicistinae in Sminthidae. From a paleontological perspecitive, Zazhigin and Lopatin (2000a) proposed a classification listing Zapodidae (containing Sicistinae and Zapodinae), Allactagidae (with Allactaginae and Euchoreutinae) and Dipodidae (which includes Cardiocraniinae, Dipodinae, and the extinct Lophocricetinae).

Sicistines are generally regarded as representing the most primitive dipodids morphologically (e. g., Dawson and Krishtalka, 1984; R. A. Martin, 1994; Miljutin, 1999; Stein, 1990). This cladistic position is supported by the few published analyses of gene sequences (nuclear LCAT, IRBP, and vWF) where in all phylogenetic reconstructions, Sicista is basal to different combinations of Zapus, Napaeozapus, Dipus, Allactaga, and Jaculus (DeBry, 2003; DeBry and Sagel, 2001; Jansa and Weksler, 2004; Michaux and Catzeflis, 2000; Michaux et al., 2001a), endorsing Stein’s (1990:310) results and conclusion "that sicistines are, in fact, the sister group to all other zapodid and dipodid rodents." In his review of early Tertiary North American rodents, Wilson (1949:128) wrote that the Sicistinae "may have been ancestral to the Zapodinae on the one hand, and to the Dipodidae on the other," a suggestion repeated by R. A. Martin (1994:7), who opined that the "sicistines form a recognizable stem group from which all later zapodids and probably dipodids evolved." This possibility was echoed by Miljutin (1999:119): "Although recent Sicistinae have their own trends of adaptive evolution, which are inconsistent or even opposite to those of Dipodoidea as a whole, their morphological characters and relatively low level of specialization make them suitable candidates for the role of ancestral stock for all other dipodoids." Sokolov et al. (1987b) and Vorontsov (1969), however, hypothesized that zapodines possess the most primitive karyotype of dipodids, from which those of sicistines and other dipodids are derived.

Despite the research by Pavlinov and Shenbrot (1983), Shenbrot (1992), Stein (1990), and Zazhigin and Lopatin (2000a), classification of dipodoid genera at either subfamily or family levels remains unresolved, reflecting different interpretations of analyzed morphological systems (teeth, head and postcranial skeletons, male reproductive anatomy, hind limb myology), lack of molecular data from a broader range of genera representing all subfamilies, and no phylogenetic reconstructions based on careful cladistic methodologies that employ character traits derived from a greater variety of morphological systems. Results from multi-trait (morphological and molecular) integrative cladistic analyses, which include appropriate outgroups, would test the hypotheses presented by Vinogradov, Pavlinov and Shenbrot, Shenbrot, Stein, and Zazhigin and Lopatin. Until such information become available, we maintain a modification of Vinogradov’s classification and recognize a single family, Dipodidae, with subfamilies Sicistinae, Zapodinae, Cardiocraniinae, Euchoreutinae, Allactaginae, and Dipodinae; hypotheses defended by Pavlinov and Shenbrot (1983), Stein (1990), Shenbrot (1992), and Zazhigin and Lopatin 2000a are discussed at the subfamilial level.

In this context, Klingener’s (1984:387) exposition is still relevant. He noted that dipodid genera show a gradation in hind limb osteology and myology (with cardiocraniines retaining many sicistine and zapodine traits); that arranging sicistines and zapodines in a separate family "is unwarrented on the basis of morphology of the hind limb and obscures the close relationship of these rodents to the jerboas"; that cardiocraniines exhibit extreme hypertrophy of the auditory bullae and extreme fusion of cervical vertebrae, which are derivations characteristic of dipodines, but allactagines have relatively small bullae and unfused cervical vertebrae; that "evolution of cranial and cervical structures has been subject to selective forces different from those acting on locomotor structures, and distribution of derived character states in different structural complexes in the spectrum of Recent genera has a mosaic pattern" and that treating "the dipodoid spectrum as a single family Dipodidae . . . is a better reflection of biological reality than dividing it."

Reviews of dipodid research and classification were contributed by Gambaryan et al. (1980), Heptner (1984), Klingener (1964, 1984), Shenbrot (1986, 1992), and Stein (1990); see also references cited in those reports. Cranial and dental characters were investigated by Vinogradov (1930). Comparative myology studied by Klingener (1964); facial myology by Gambaryan et al. (1980) and Meinertz (1941); myology of postcrania by Fokin (1971) and Stein (1990). Review of distribution and habitat of dipodids (excluding sicistines and zapodines) given by Kulik (1980). Chromosome numbers of members of each subfamily provided by Vorontsov (1969). Male genitalia studied by Vinogradov (1925) and Pavlinov and Shenbrot (1983). Comparative behavior and its taxonomic significance studied by Rogovin (1984). Significance of the relation between size of pinna and auditory bulla in allactagine and dipodine species chronicled by Pavlinov and Rogovin (2000). Geographic distributions, ecologies, and comparative data for Mongolian jerboas monographed by Sokolov et al. (1996, 1998). Taxonomy, characteristics, geographic ranges, ecologies, and other topics covering dipodoids occurring in Russia and adjacent regions reviewed in detail by Shenbrot et al. (1995). Panteleyev (1998) provided distribution maps for the Palaearctic species. Distributions of species occuring in the former USSR were verified and much enhanced by personal communication with G. I. Shenbrot, who also highlighted taxonomic problems; range and taxonomy of Chinese forms were illuminated by the efforts of Lin Yong-lie.

That Muroidea (Platacanthomyidae, Spalacidae, Calomyscidae, Nesomyidae, Cricetidae, and Muridae) is the sister group to Dipodidae (Dipodoidea) has been consistently demonstrated by analyses of the following DNA sequences: nuclear LCAT gene (Michaux and Catzeflis, 2000); nuclear vWF (Huchon et al., 1999, 2000); LCAT and vWF combined (Michaux et al., 2001b); nuclear IRBP gene (DeBry and Sagel, 2001; Jansa and Weksler, 2004): nuclear CB1, IRBP, and RAG2 (DeBry, 2003); mitochondrial cytochrome b and ND4 (Conroy and Cook, 1999); mitochondrial cytochrome b and 12S rRNA (Montgelard et al., 2002); mitochondrial 12S rRNA combined with nuclear GHR, BRCA1, and vWF genes (Adkins et al., 2001); mitochondrial 12S rRNA (Nedbal et al., 1996); nuclear GHR and BRCA1 (Adkins et al., 2003); nuclear vWF, IRBP, and A2AB (Huchon et al., 2002); (samples are from Sicista, Zapus, Napaeozapus, Allactaga, Dipus, and Jaculus in different combinations). The molecular data strongly reinforces a significant body of past research based upon a range of morphological systems derived from extant (Alston, 1876; Forsyth Major, 1896; Hartenberger, 1985; Klingener, 1964; Méhely, 1913; Meinertz, 1941; Thomas, 1896; Tullberg, 1899; Winge, 1887) and fossil (Emry, 1981; Emry et al., 1998; Lindsay, 1977; Schaub, 1934; Walsh, 1997; Wang, 1985; R. W. Wilson, 1949) species that indicates dipodids to be a sister group of, or closely related to muroid rodents (see also the reviews by Klingener, 1964, 1984). Schaub (1958) and McKenna and Bell (1997) expressed this relationship in their classifications by arranging Dipodoidea and Muroidea in the infraorder Myodonta. Wang (1985) noted the close similarity in molar occlusal patterns between cricetines and early Oligocene "Zapodidae," and similarities in molar patterns between the middle Eocene Asian dipodid Aksyiromys and cricetines prompted Emry et al. (1998:222) to suggest that Aksyiromys "was close to a common ancestor of Cricetidae and Zapodidae; the Asian cricetine Phodopus was used as an outgroup in Stein’s (1990) phylogenetic study and Cricetinae in general (as defined by by Carleton and Musser, 1984) was implied as an outgroup by Shenbrot (1991a). Alternative alliances between dipodoids and non-muroid rodent groups proposed in the older literature (e. g., Dobson, 1882b; Miller and Gidley, 1918; Parsons, 1894; Zittel, 1893) and even more recent contributions (e.g., Hartenberger, 1998) are not supported by morphological, molecular, or paleontological data.

North American and Asian Eocene strata contain evidence for the evolutionary beginnings of dipodoids. Current consensus identifies the North American Elymys (middle Eocene), Simimys (middle to late Eocene), Simiacritomys (late Eocene), and possibly Nonomys (late Eocene), along with Asian Aksyiromys, Banyuesminthus, Blentosomys, Primisminthus, and Ulkenulastomys (all middle Eocene) as the oldest members of Dipodoidea (Emry and Korth, 1989; Kelly, 1992; Walsh, 1997; Wang and Dawson, 1994; Wang and Qiu, 2000). Whether any or all genera are ancestral to dipodids, belong in Sicistinae or Zapodinae, or represent the sister group to Dipodidae within a larger Dipodoidea has yet to be determined (see review of these Paleogene genera and analysis by Walsh, 1997). Armintomys tullbergi, from the early part of the middle Eocene of North America and the only recorded member of Armintomyidae (Dawson et al., 1990), is the oldest known rodent with hystricomorphous zygomasseteric structure and incisor enamel transitional from pauciserial to uniserial (Wang and Dawson, 1994). It is apparently a primitive myomorph, and its morphological traits (sciuravid-like molar occlusal patterns, hystricomorphy, and intermediate incisor enamel architecture) indicate that A. tullbergi "may be derived from the sciuravids" (Dawson et al., 1990:145) and also "may represent the closest known sister group to a unit including dipodoids and cricetids" (Wang and Dawson, 1994:251). In Walsh’s (1997) phylogenetic analysis, Armintomys, along with the more derived North American Pauromys (early to middle Eocene), formerly considered a sciuravid (McKenna and Bell, 1997), is placed in Myodonta (the infraorder also containing dipodoids and muroids; see McKenna and Bell, 1997) and basal to the branching point of Dipodoidea and Muroidea (see McKenna and Bell, 1997). The basic outline of this phylogenetic pattern was discerned more than 50 years ago by Wilson (1949;127), when in reviewing limited material compared to that now available, he suggested that "the muroids and dipodoids, although arising from a common (mid-Eocene?) ancestor, have been distinct groups since the late Eocene. Moreover, this common ancestor apparently arose from the sciuravines, if any stem is known at present."