Echolocation and roosting ecology determine sensitivity of forest‐dependent bats to coffee agriculture

Abstract

Species differ in vulnerability to anthropogenic land use changes. Knowledge of the mechanisms driving differential sensitivity can inform conservation strategies but is generally lacking for species‐rich taxa in the tropics. The diverse bat fauna of Southeast Asia is threatened by rapid loss of forest and expanding agricultural activities, but the associations between species, traits, vulnerability to agriculture, and the underlying drivers have yet to be elucidated. We studied the responses of speciose insectivorous bat assemblages to robusta coffee cultivation in Sumatra, Indonesia. We compared abundance, species richness, and assemblage structures of bats between forests and coffee farms based on trapping data and evaluated the influence of vegetation complexity on assemblage composition and species‐level reactions. Bat abundance and species richness were significantly lower in coffee farms than in forests. Bat assemblage structure differed between land uses, and the overall variation can be largely explained by vegetation simplification. Species sensitive to coffee agriculture were associated with more complex vegetation structure, whereas tolerant species were associated with simpler vegetation structure. Sensitive and tolerant species differed in the type, frequency, and bandwidth of echolocation calls and roost use. Species sensitive to coffee use broadband and high‐pitched frequency‐modulated calls, which are efficient at detecting insects in complex vegetation, and roost in plant structures that may be lost as vegetation is simplified. In contrast, tolerant species used lower pitched constant‐frequency calls and roost in caves. We advocate for greater use of trait analyses in studies seeking to clarify the influence of agriculture on diverse tropical bat faunas. Abstract in Indonesian is available with online material.