Scatter-hoarding behavior of two Amazonian rodents: theory and application in forest fragments

Abstract

In the present work, I investigated what evolutionary and environmental factors motivate an animal to scatter-hoard. Scatter-hoarding occurs when animals cache food individually under the soil surface to harvest them later. In order to answer my question, I used two complementary approaches: a theoretical one and an empirical one. The theoretical approach was a two-time step model based on Holling's disc equation of harvest rate. It was intended to be simple and general, using general parameters to investigate caching dynamics: total foraging time, resource abundance, handling time, energetic content, and predation risk. The model depicted four general motives for an animal to cache when caching: 1) allows the animal to gain search time when search time is cheaper (search-time reallocation); 2) balances food intake over time (consumption reallocation); 3) decreases predation risk (predator avoidance); finally 4) when food quality increases while cached (food aging); The empirical study was conducted in forest fragments (one, ten and hundred hectares) and continuous forest of the Biological Dynamics of Forest Fragments, Central Amazon, Brazil. The objects of the study were: agoutis (Dasyprocta leporina), acouchies ( Myoprocta acouchy) and palm nuts (Astrocaryum aculeatum). Overall, the study showed that changes in scatter-hoarding behavior (levels of removal and caching, as well as caching distance) can be greatly explained by changes in resource abundance. Where there was more food, fewer seeds were removed. And when there was more food, more seeds were cached and they were cached farther away. But when looking specifically at effects of forest fragmentation, the relationship between resource abundance and removal or caching did not occur. More seeds were removed in smaller fragments; fewer seeds were cached in smaller fragments; and the ones cached were cached at shorter distances. Nevertheless, fragment size did not affect resource abundance. Factors such as a decrease in predation risk, decrease in pilferage, and the increase of the relative importance of agoutis as the agents of scatter-hoarding (which are the only scatter-hoarders capable of visiting the small fragments) may explain the changes observed in small forest fragments.