and greatest diversity in the Nearctic (Wilson and Ruff
1999). The mountain beaver, family Aplodontidae, occurs
exclusively in the Pacific Northwest of North America. The
hystricognath rodents are represented by a single spe-
cies, Erethizon dorsatum(New World porcupine); jumping
mice (family Dipodidae) and beaver (family Castoridae) are
shared between the Nearctic and Palearctic biogeographic
regions.
Convergence
Changes in climate, exchanges between biogeographic re-
gions, and rapid faunal turnover on many continents have
resulted in local adaptive radiations characterized by eco-
morphological specializations. As a result, a frequent theme
associated with rodent radiations is convergent evolution
in terms of life-history strategies, behavior, and anatomy
and physiology, and with evidence of convergence emerg-
ing when one compares unrelated rodent lineages occupy-
ing different biogeographic regions. For instance, special-
izations for a subterranean lifestyle have evolved multiple
times in rodents, as seen by the degree of convergent mor-
phologies and physiologies displayed by the families Bath-
yergidae (Africa), Ctenomyidae (South America), Muridae
(Palearctic and Oriental), Octodontidae (South America),
and Geomyidae (North America). These specializations in-
clude a fusiform body, reduction in limb length, dorsally
flattened skulls, and behavior associated with burrow con-
struction (Stein 2000). Subterranean rodents also show
physiological convergence, characterized by specializations
for living in a high CO 2 environment (Buffenstein 2000).
Members of several rodent families (Heteromyidae, Dipo-
didae, and Pedetidae) show similar adaptations for life in
arid environments, especially as they relate to structure
of the kidney (Schmidt-Nielsen 1964; Al-kahtani et al.
2004) and changes in the postcranial skeleton associated
with saltatorial locomotion. Convergence for group liv-
ing in response to similar environmental constraints is a
common explanation for the evolution of complex social
systems, and several behavioral studies of rodents have
used comparative methods to investigate ecological features
shared by unrelated species of rodents that form social
groups and have similar mating systems (Lacher 1981;
Faulkes et al. 1997; Blumstein and Armitage 1998a; Armi-
tage 1999a; Ebensperger and Cofré 2000; Lacey and Wiec-
zorek 2003). Finally, convergence and parallelism of the
zygomasseteric system and dentition of rodents have ex-
acerbated efforts to identify monophyletic groups of ro-
dent families, making the classification of rodents difficult
(Wood 1955).
Evolutionary HistoryRodents first appear in the fossil record during the Paleo-
cene approximately 55 to 60 mya, with one of the oldest
families being the Paramyidae (Vianey-Liaud 1985; Har-
tenberger 1998). Most modern families of rodents under-
went an adaptive radiation in the Paleocene /Eocene, and
were well established by the Late Eocene to Early Oligocene
(Vianey-Liaud 1985; Jaeger 1988). Therefore, the rodent
tree of life has received little pruning in terms of its major
lineages.Issues related to divergence times
Rodents are well represented in the fossil record, and pale-
ontologists have established divergence times for several
rodent lineages. Nevertheless, recent divergence times de-
rived from “molecular clocks” often conflict with presum-
ably well-established paleontological dates. For instance,
earliest fossils place the divergence for the murid genera
Mus/Rattusbetween 12 and 14 mya (Jacobs and Pilbeam
1980; Jacobs et al. 1989). Using a molecular clock based on
albumin immunology, Sarich (1985) provided an estimate
of the Mus/Rattusdivergence at 22 to 24 mya, and a date of
41 mya was obtained by Kumar and Hedges (1998) from
their comparisons of eutherian mammals and estimates of
amino acid replacement differences across 658 nuclear gene
loci. When one considers new molecular dates for orders of
mammals, the discrepancy between fossils and molecules
becomes more controversial. The fossil record suggests a
divergence date no older than 65 myr for most orders of
placental mammals (Foote et al. 1999; Archibald 2003),
whereas molecules place the origin of major placental line-
ages in the Cretaceous at around 100 mya (Springer 1997;
Kumar and Hedges 1998; Eizirik et al. 2001). Therefore,
the molecular estimates suggest gaps in the fossil record
that are 50 to 60 myr earlier than the first appearance of
fossil evidence for particular orders (Foote et al. 1999;
Archibald 2003). How do these new molecular clocks im-
pact estimates of rodent divergence? Based on these molec-
ular dates, the origin of Rodentia would be 100 to 110 myr
or older (Springer 1997; Cooper and Fortey 1998; Kumar
and Hedges 1998), extending divergence time estimates
within Rodentia beyond any known fossils. According to
the estimate of Kumar and Hedges (1998), this results in a
divergence time of 110 myr (Middle Cretaceous) for the
separation of the two rodent suborders, Hystricognathi and
Sciurognathi, and this date is 50 myr prior to the first fossil
rodent or rodent ancestor.
There is at least one issue that tends to complicate recent
dates derived from a molecular clock of eutherians. Rates ofRodent Evolution, Phylogenetics, and Biogeography 11