Learning-by-consequence foraging model of the Northern Mealy Amazon in a modified landscape of tropical moist forest

摘要

Understanding the dynamics of animal responses to heterogeneity of food resources in modified landscapes is crucial for the conservation of threatened species. We evaluated dietary stratégies of the Northern Mealy Amazon in a modified landscape of tropical moist forest in Los Chimalapas, Mexico. We established 30 phenology transects (200 X 6 m) to détermine food resource availability in primary evergreen, riparian, and secondary forests during the parrot breeding (March-April) and non-breeding (August-September) seasons, and determined parrot diet by observations of foraging individuals. We found spatiotemporal variation in food resource availability, with significantly greater food resource availability in the breeding season compared to the non-breeding season, and significantly fewer resources in secondary forest. Parrots consumed mainly seeds of a variety of plant species, and presented a medium dietary niche during the breeding season. Nevertheless, parrots foraged predominantly in primary forest where they demonstrated dietary selectivity, and rarely foraged in secondary forest where they consumed resources according to availability. We found évidence that Northern Mealy Amazons employ hier-archical criteria in habitat use and resource sélection, adjusting foraging stratégies according to resource availability. This learning-by-consequence foraging model may enable parrots to rapidly adapt to spatiotemporal variations in food resource availability. Primary tropical moist forest may bea key habitat type for Northern Mealy Amazons during the breeding season as this provides an abundance of food resources for parrots when they are raising young.