Poachers Alter Mammal Abundance, Seed Dispersal, and Seed Predation in a Neotropical Forest

Abstract

Abstract: We evaluated the hypothesis that poachers reduce the abundance of herbivorous mammals, and that this, in turn, alters seed dispersal, seed predation, and seedling recruitment for two palms ( Attalea butyraceae and Astrocaryum standleyanum ) in central Panama. Using physical evidence left by poachers and interviews with forest guards, we quantified poaching intensity for eight forest sites. We quantified mammal abundance using transect counts and small‐mammal traps. Abundance was inversely related to poaching intensity for 9 of 11 mammal species (significantly so for 5 species), confirming the first component of the hypothesis. The outcome of interactions among seeds, mammals, and beetles also varied with poaching intensity. Nonvolant mammals were the only seed‐dispersal agents, and rodents and beetles were the only seed predators. We quantified seed fate by examining the stony endocarps that encase the seeds of both palms. The large, durable endocarps were located easily on the forest floor and bear characteristic scars when a rodent or beetle eats the enclosed seed. The proportion of seeds dispersed away from beneath fruiting conspecifics was inversely related to poaching intensity, ranging from 85% to 99% at protected sites where mammals were abundant and from 3% to 40% at unprotected sites where poachers were most active. The proportion of dispersed seeds destroyed by beetles was directly related to poaching intensity, ranging from 0% to 10% at protected sites and from 30% to 50% at unprotected sites. The proportion of dispersed seeds destroyed by rodents was inversely related to poaching intensity, ranging from 85% to 99% at protected sites and from 4% to 50% at unprotected sites. Finally, seedling densities were directly related to poaching intensity. There was no single relationship between poaching intensity and the biotic interactions that determine seedling recruitment. The net effect of poaching on seedling recruitment can be determined only empirically. For these palms, seedling densities were directly related to poaching intensity.