Late Miocene and early Pliocene events provide the best supported explanation of the deepest bifurcation in Ascaphus. Under this scenario, the major divergence in the lineage emerged during the late Miocene in response to the rise of the Cascade Mountain range. The Miocene (28–10 million years ago) and earlier Eocene of the Pacific Northwest were characterized by a mesic equable climate that supported broadleaf deciduous and evergreen forests with an increasing conifer component (Axelrod 1968; Wolfe 1969, 1978); such forests were likely similar to those that modern tailed frogs inhabit. The effect of the rain shadow created by the high Cascade range became increasingly apparent in the paleoflora through the Pliocene (Wolfe 1969); some plant lineages at-tained a distribution restricted to the east or west of the early Cascades (approximately 1000 m in elevation) during the late Miocene (Wolfe 1969). If the climatic influences of the early Cascades did not split contiguous populations of Ascaphus into eastern and western groups, they were in all likelihood divided by the rain shadow of the high Cascade Mountains, which currently partitions their range (Fig. 1).
Metter and Pauken (1969) and Pauken and Metter (1971) suggested that inland and coastal populations of Ascaphus were connected by gene flow through the mountains of central Oregon and did not diverge until the close of the Pleistocene (10,000 years ago). There is substantial evidence, however to refute a Pleistocene interpretation of the deepest diver-gence in Ascaphus. Assuming clocklike evolution (which can-not be rejected for our data), the magnitude of divergence observed between coastal and inland haplotypes would re-quire a minimum rate of 0.0616 substitutions per site per million years to have occurred during the Pleistocene. This is a much faster rate of evolution than the 0.0278 substitutions per site per million years typically estimated for mammals (e.g., Arbogast and Slowinski 1998). We strongly doubt that Ascaphus sequences are evolving at such a rapid rate. A gen-eration time of six or more years and a cold-water habitat have probably slowed the evolutionary rate of Ascaphus even below that of other ectotherms, which are hypothesized to evolve four to five times slower than mammals (Kocher et al. 1989; Thomas and Bechenbach 1989; Adachi et al. 1993; Martin and Palumbi 1993; Rand 1994). Thus, a Pleistocene or post-Pleistocene divergence between inland and coastal populations is unlikely.
This view is also supported by the paloebotanical record.