In northwestern Mexico, low-density cattle ranching is a common and culturally imbedded practice to extract economic value from large, undeveloped landholdings. Though domestic cattle (Bos taurus) may negatively impact tree diversity in the Mexican tropical dry forest (TDF), few researchers have quantified cattle foraging preferences for woody plant species, or tested hypotheses explaining potential impact patterns. In this study, I ask: are cattle selective browsers of woody plants? And if so, what variables can explain cattle preferences? To answer these questions, I designed and deployed custom-made animal-mounted time-lapse video and data logging equipment to record cattle feeding behavior and movements. The data collection system, which records detailed movement and foraging behavior over a 5 day sampling interval, gives an unprecedented window into the lives of free-ranging livestock as they interact with their environment. I collected 2,291 unique 10- and 20-second videos of cattle feeding behavior. Using GPS data on cow foraging paths, I returned to 165 feeding points, confirmed the identity of the species eaten, and censused woody plants within a 5m by 5m area surrounding the eaten plant. I then used an information-theoretic statistical approach to address questions about dietary selectivity in cows at different scales.
Results/Conclusions
The best predictor of frequency in diet for most woody plants is frequency in the environment, although some statistically significant exceptions suggest cattle preference for or avoidance of certain species. Selective behavior, however, is more apparent at smaller spatial scales, within 25m2 grids surrounding an observed feeding location. My results suggest that cows are not picky eaters— in terms of available species, they eat what is abundant, but they will select some and avoid others, given the immediate choices in front of them. However, they spend significantly more time in habitats with preferred species, suggesting that selection may take place at a landscape level.