Complex competition interactions between Egyptian fruit bats and black rats in the real world

Description

Research Hypothesis: We hypothesized that bats perceive rats not merely as competitors for food, but as a significant predation risk. We predicted that this perceived risk would lead to distinct anti-predator behaviors (e.g., avoidance, increased vigilance) in bats when rats are present, differing from their behavior in the presence of only other bats (conspecifics). Furthermore, we hypothesized that these interactions would be dynamic, shifting with seasonal changes in resource availability and predator-prey encounter rates. Predation Risk vs. Competition: When rats were present, bats significantly reduced their landing rate on the food platform. For the landings that did occur, bats exhibited significantly higher vigilance (longer scanning durations before feeding) and had lower foraging success compared to when only other bats were present. This suggests the bats' response is driven by predation fear, not just competition. Behavioral Plasticity: The data reveals a remarkable seasonal shift in bat behavior. In winter, when alternative food sources were scarce and rat encounters were infrequent, bats primarily adopted a risk-averse strategy of avoidance. In spring, when food was more abundant and rat encounters were common, bats shifted towards a more confrontational strategy, including actively attacking rats to gain access to food. 2. Description of Data and Code Files The data was collected over a seven-month period (December 2017 - May 2018) in a semi-natural, open bat colony with a provisioned food platform. Interactions were monitored using surveillance video cameras. Description: This file contains the core behavioral data for individual bat landing events. Each row represents a single landing by a bat on the food platform and details the context and the bat's subsequent behaviors. Data Gathering for the table "data_rat_and_bat_duration.csv": Bat landings were manually detected from video. The behavioral variables (e.g., vigilance, foraging success) for each landing were then manually annotated and quantified from the video recordings. Data Gathering for the table "data_rat_arrival_number_30min.csv": Rat arrivals were manually annotated from video footage. The number of unique arrival events was then summed for each 30-minute interval of each night. Data on food depletion was estimated based on bat feeding activity. Description: These two MATLAB scripts contain the code used to perform the statistical modeling and AIC-based model selection for Table 2 and Table 4 in the manuscript (updated for the model validation on 06/08/2025). GLM_rat_arrival_number.m: Performs a GLM with a Poisson distribution to model the factors influencing the number of rat arrivals (data_rat_arrival_number_30min.csv). GLM_food_risk.m: Performs a logistic regression (a GLM with a binomial distribution) to model the factors influencing bat foraging success in the presence of a rat (data_rat_and_bat_duration.csv).

Institutions

Categories

Funding

Licence