A test of the acoustic adaptation hypothesis in four species of marmots

Acoustic signals must be transmitted from a signaller to a receiver during which time they become modified. The acoustic adaptation hypothesis suggest s that selection should shape the structure of long-distance signals to max imize transmission through different habitats. A specific prediction of the acoustic adaptation hypothesis is that long-distance signals of animals in their native habitat are expected to change less during transmission than non-native signals within that habitat. This prediction was tested using th e alarm calls of four species of marmots that live in acoustically differen t habitats and produce species-specific, long-distance alarm vocalizations: yellow-bellied marmot, Marmota flaviventris; Olympic marmot, M. olympus; h oary marmot, M. caligata; and woodchuck, M. monax. By doing so, we evaluate d the relative importance the acoustic environment plays on selecting for d ivergent marmot alarm calls. Representative alarm calls of the four species were broadcast and rerecorded in each species' habitat at four distances f rom a source. Rerecorded, and therefore degraded alarm calls, were compared to undegraded calls using spectrogram correlation. If each species' alarm call was transmitted with less overall degradation in its own environment, a significant interaction between species' habitat and species' call type w ould be expected. Transmission fidelity at each of four distances was treat ed as a multivariate response and differences among habitat and call type w ere tested in a two-way MANOVA. Although significant overall differences in the transmission properties of the habitats were found, and significant ov erall differences in the transmission properties of the call types were fou nd, there was no significant interaction between habitat and call type. Thu s, the evidence did not support the acoustic adaptation hypothesis for thes e marmot species. Factors other than maximizing long-distance transmission through the environment may be important in the evolution of species-specif ic marmot alarm calls. (C) 1998 The Association for the Study of Animal Beh aviour.