Rampant illegal hunting threatens wildlife populations inside many tropical protected areas, compromising their long-term effectiveness. A critical question concerns whether such harvest has indirect effects on non-hunted organisms that interact with the game species. For example many tree species are demographically reliant on seed dispersal by vertebrates that are threatened by hunting; the anthropogenic disruption of this animal-plant mutualism can severely alter the composition of tropical forests. Here I show that illegal poaching has reduced or extirpated several mammal species from national parks in northern Thailand. This, in turn, has negatively affected the demography of the canopy tree Choerospondias axillaris , which is dependent on the dispersal of its seeds to light gaps by gibbons ( Hylobates lar ), sambar deer ( Cervus elaphus ), and muntjac deer ( Muntiacus muntjak ). In parks where these mammals are heavily hunted, far fewer seeds are dispersed to light gaps and seedling abundance is significantly reduced. These results suggest that anthropogenic impacts such as overharvest can indirectly ramify through communities.
I also assessed the functional equivalence of the three seed-dispersing mammals in terms of their demographic impact on C. axillaris. Sambar and muntjac dispersed far more seeds than gibbons. Sambar deposited many seeds under female tree canopies; muntjac were the only disperser to deposit seeds in the most open habitats, which are beneficial for C. axillaris seed germination, seedling survival and growth. Using stage-based population models, I assessed how disperser-specific seed dispersal, variation in the frequency of canopy gap formation, and the interactive effects of these factors on plant demography influence the long-term population growth of C. axillaris. Large differences in dispersal quantity and small differences in dispersal quality, when placed in a biologically complex population-level context, resulted in only marginal variation in the impacts of these frugivores on tree abundance. Tree species more highly dependent on zoochorous seed dispersal will have more room for skewed interaction strengths among their dispersers. In measuring functional redundancy or in trying to predict the role of diversity in species interactions, we must explicitly account for variation in life-history traits.
