Quantifying the mutualism-antagonism continuum for seed removal by a granivorous disperser

Abstract

Seeds removed by animals have one of two mutually exclusive fates – they are either predated or dispersed and still alive. The quality of seed dispersal by animals and the number of predated seeds will therefore determine net interaction outcomes for plants. Yet, it is poorly understood what proportion of removed seeds animals can predate before benefits of dispersal no longer outweigh costs of seed loss.

Here, we calculated the mutualism-antagonism continuum for seed removal of the fleshy-fruited tree Frangula alnus by the seed-predating bird Coccothraustes coccothraustes in Białowieża Forest. We integrated effects of the bird during seed dispersal (fruit handling, seed predation, and seed deposition) into microhabitat-structured tree population models.

Results of our models showed that the probability of a seed of F. alnus reaching maturity after seed removal by C. coccothraustes decreased from 0.0028% to 0% as seed predation increased from 0% to 100%. Seed removal was beneficial when less than 63.7% of seeds were predated, and antagonistic when more than 72.0% of seeds were predated. Modifying key model parameters (here, the negative effect of fruit pulp on seedling recruitment and the frequency of forest gaps) decreased and increased rates of seed predation, at which costs of seed loss outweighed benefits of seed dispersal (from 37.9% to 80.7%).

Our findings highlight that benefits of animal seed dispersal can largely outweigh costs of seed predation in a fleshy-fruited tree. Yet, the mutualism-antagonism continuum of seed removal depends on intrinsic factors (e.g. variation in interactions among individuals) and extrinsic factors (e.g. the environment) of seed dispersal and plant demography. Because C. coccothraustes was observed predating at least 80% of removed seeds, it appears to be an antagonist of animal-dispersed plants and exploiter of the seed dispersal mutualism.