Perception of predation risk in tropical dry forest birds in central Mexico
Description
Individuals are expected to balance the costs and benefits of their escape decisions based on aspects of their biological condition and life history, for example, size, body mass or satiation level. Flight initiation distance (FID) is the distance at which animals escape when approached by humans, under standardized conditions. It is used as a proxy of animals´ perception of predation risk. Escape behavior is a proxy to measure anti-predator strategies and to monitor tolerance and human impacts on birds. This study analyzes the perception of predation risk in bird species within a tropical dry forest with continuous human presence. Our experiment evaluates the influence of body mass, activity, and residence status on the perceived predation risk based on FID. We predicted longer FID in larger than in smaller bird species (based on body mass), in perching birds relative to foraging individuals, and in migratory relative to resident species. The values of FID in 19 species (82% residents) during 76 trials ranged from 2.46 to 32.29 m. There was a quadratic relationship with size, in which birds with intermediate size had longer FID than lighter or heavier species. Perching individuals had longer FID than foraging individuals. FIDs were similar between resident and migratory birds. Behavioral mechanisms and life history of particular species may explain differences in escape strategies.
Files
Steps to reproduce
Fieldwork was conducted in April 2024 during the dry season within the northwestern portion of the Sierra de Huautla Biosphere Reserve (REBIOSH) in Morelos, Mexico, focusing on the rural ejido of El Limón de Cuauchichinola. This sampling period was deliberately chosen to observe bird communities experiencing high daily thermal variation and limited vegetation cover, while simultaneously avoiding the breeding season to ensure that parental care behaviors did not confound baseline escape decisions. To collect the flight initiation distance (FID) data, a single observer walked slowly through a diverse mosaic of tropical dry forest, secondary vegetation, and cattle paddocks to locate perched birds using binoculars. Once an individual bird was selected, the observer initiated a direct approach from a standardized starting distance of approximately 25 to 30 meters away from the target tree, walking at a constant pace of roughly 1 m/s while maintaining direct eye contact with the bird. The exact moment the bird took flight, the observer halted to record the height of the initial perch in the tree and the horizontal ground distance from the observer to the base of the tree. The true linear FID was subsequently calculated using the Pythagorean Theorem to account for the spatial positioning of the bird relative to the approaching human. Because the birds were not banded, consecutive trials involving the same species were separated by a minimum geographic distance of 50 meters to minimize the likelihood of sampling the same individual twice and causing pseudoreplication. Furthermore, all experimental trials were temporally standardized and conducted between 06:30 h and 11:00 h to eliminate the time of day as a confounding variable. For each unique trial, the bird's behavior at the exact moment prior to the human approach was categorized into one of two operational states: Perching (stationary and vigilant) or Foraging (actively handling or consuming food). Following field data collection, species identifications were verified and their migratory residency status was assigned as either Resident or Migrant. Species-specific average body mass values in grams were then extracted and compiled from the global AVONET bird trait database. The finalized dataset structures these variables into a unified matrix containing 76 unique approach trials across 19 distinct bird species, preserving the raw trial IDs, tracking dates, species identities, body mass values, baseline activities, and final FID calculations.
Institutions
- Universidad Autónoma del Estado de MorelosMorelos, Cuernavaca