Predator playback, foraging height, and phylogeny affect gap crossing behavior in tropical forest birds

Abstract

The effects of fragmented landscapes on tropical bird movements remain poorly understood, partly due to a dearth of experimental tests of gap crossing behavior. To learn more about the factors involved in gap crossing behavior, we used predator-playback experiments on four species of understory-insectivore birds (Chestnut-backed Antbird Poliocrania exsul, Black-crowned Antshrike Thamnophilus atrinucha, White-breasted Wood-wren Henicorhina leucosticta, Stripe-breasted Wren Cantorchilus thoracicus) at La Selva Biological Station, Costa Rica. We sought to understand species- and family-level differences in gap crossing behavior, specifically: (1) the effect of forest-opening (gap) width, and (2) the effect of perceived risk of predation. A total of 39 treatment (predator playback) and 39 control (silent playback) trials were conducted at gap sites ranging in width from 2.8 to 12.6 m. Predator playback decreased the number of times that birds crossed study gaps, increased latency time in all but one species, and increased the closest distance that birds approached the playback speaker. Gap width affected only latency time: as gap width increased, latency time increased. We observed strong differences in behavioral response between families and species of different foraging heights. In the statistical models, family (Thamnophilidae and Troglodytidae), foraging height (low or mid-story), and trial type (silent control or predator treatment) consistently emerged as strongest predictors of bird behavior. Our results indicate that small linear gaps (< 12.6 m) do not impede movement for these four focal species, but that risk of predation May hinder bird movement even at small gaps.