Evolution of Animal-Mediated Seed Dispersal Systems

Roman Yukilevich and Henry F. Howe Program in Ecology and Evolution, Department of Biology, University of Illinois-Chicago, Chicago, IL, 60607

 

We develop a model that describes the conditions for the evolution of seed dispersal systems involving animal dispersal agents. Six possible dispersal systems are identified, supplementing the two, which have previously been the foci of discussion: "specialized" and "generalized", for small and large dispersal assemblages, respectively. For purposes of discussion, these are contrasted with a non-interactive state, analogous to a null hypothesis. It states that in a specialized system, the plant specializes on a single, "exceptionally effective dispersal species", while in a generalized system, the plant generalizes on multiple dispersal species. We then identify environmental conditions that may be expected to draw the system out of the non-interactive state. New conditions are proposed to alter previous trends in specialization in a predictable fashion. Specialized system evolves under four distinct conditions: The 1) invasion of an equally effective dispersal agent, 2) heavy and constant infestation by seed or fruit destroying visitors, 3) an exceptionally effective dispersal agent evolves predatory feeding activities on fruits or seeds and 4) the extinction of an effective dispersal agent. Generalized system evolves only under one environmental condition: The invasion of an effective dispersal agent. In total, we predict three outcomes when Specialized and Generalized systems evolve: The Specialized system evolves into either "equally effective dispersal system" or into "reduced seed establishment system", while Generalized system only evolves into "effective invader system". After we discuss how environmental conditions are likely to alter the system of interest, we 1) predict novel selection pressures on the plant, 2) propose three hypotheses ("convergence, restriction, and attraction hypothesis") as to how a plant may evolutionarily respond, and 3) predict the dispersal-related plant traits ("dispersal syndromes") based upon the nature of each hypothesis. Finally, we discuss possible ecological and historical constraints on the evolution of seed dispersal systems.
Even though only five systems are likely in our model, each is further differentiated into plants that have depleting and non-depleting crops. Transitions from Specialized to Generalized systems and vice versa are influenced by the extent to which fruit crops are regularly depleted (depleting plants) or remain superabundant (non-depleting plants). In total, we argue that the break-up of a non-interactive state under certain environmental conditions and the subsequent evolutionary response by a plant drive the evolution of seed dispersal systems.