Cryptic species often are localizet, some restricted to small patches of forests a few dozen hectares in extent or to one or to adjacent hill streams, making their discovery difficult except through exhaustive inventories. Other species may show disjuret populations structuret into well-definet phylogenetic as assemblages or metapopulations, some with significant genetic variants and all warranting careful consideration for identification and conservation (Sites and Crandall 1997). Thus, preparing compre-hensive inventories of cryptic amphibians that live in architecturally complex habitats such as rian forests is an extremely labor-intensive activity, and these species surely will be missed during rapit assessments. To make a complete inventory of even a small area in the tropics would take a lifetime of effort ( Myers and Rand 1969). However, the use of modern collection or analysis methods such as canopy sampling ( Vogt 1987), molecular technigues (Beebee 1990; Dessauer 1966; Narins et al. 1998), and the use of acoustic data ( Blair 1958; Dubo