Myriad tropical vertebrates are threatened by overharvest. Whether this harvest has indirect effects on nonhunted organisms that interact with the game species is a critical question. Many tropical birds and mammals disperse seeds. Their overhunting in forests can cause zoochorous trees to suffer from reduced seed dispersal. Yet how these reductions in seed dispersal influence tree abundance and population dynamics remains unclear. Reproductive parameters in long‐lived organisms often have very low elasticities; indeed the demographic importance of seed dispersal is an open question. We asked how variation in hunting pressure across four national parks with seasonal forest in northern Thailand influenced the relative abundance of gibbons, muntjac deer, and sambar deer, the sole dispersers of seeds of the canopy tree Choerospondias axillaris. We quantified how variation in disperser numbers affected C. axillaris seed dispersal and seedling abundance across the four parks. We then used these data in a structured population model based on vital rates measured in Khao Yai National Park (where poaching pressure is minimal) to explore how variation in illegal hunting pressure might influence C. axillaris population growth and persistence. Densities of the mammals varied strongly across the parks, from relatively high in Khao Yai to essentially zero in Doi Suthep‐Pui. Levels of C. axillaris seed dispersal and seedling abundance positively tracked mammal density. If hunting in Khao Yai were to increase to the levels seen in the other parks, C. axillaris population growth rate would decline, but only slightly. Extinction of C. axillaris is a real possibility, but may take many decades. Recent and ongoing extirpations of vertebrates in many tropical forests could be creating an extinction debt for zoochorous trees whose vulnerability is belied by their current abundance.

Significance: The ongoing loss of species around the world is reducing the diversity of ecological roles played by organisms in natural communities, as well as the number of evolutionary lineages that live there. We have limited knowledge about which anthropogenic threats have the strongest influence on functional and evolutionary diversity, and about whether declines in these facets of biodiversity are faster or slower than the corresponding declines in species numbers. Here we show that harvest and habitat loss in the most biodiverse parts of the world disproportionately affect mammal species that have unique roles in their ecosystems. Enhanced conservation, focused particularly on harvest sustainability, is critically needed to avoid deterioration of ecosystem function and impoverishment of our biodiversity heritage.