Abstract
Obligate 1-to-1 mutualisms between species pairs are rare in practice and anomalous in theory. A complex community holds the potential for thousands of such interactions; a negligible number exists. A brief discussion of pollination and seed dispersal suggests checks to mutualistic coevolution and indicates that reciprocal adaptation where it occurs is general rather than specific. Diversity diffuses selection from any 1 source, reducing the possibilty of direct reciprocal selection among species pairs. Normal processes of successsion, disturbance and geographic flux of species ranges continually change selective regimes over evolutionary time. Commonplace variation in such population attributes as age and spatial distributions of plants that provide flower and fruit resources alter the intensity of interaction between plants and animals and thereby select for general rather than narrowly specific relationships. Polygenic inheritance promotes uneven rates of evolution among mutualists. Asymmetrical coevolution leaves 1 member of an interacting pair with more potential for response to selection from organisms outside of the mutualism than the other. In a world of ecological flux, this hinders the evolution of finely tuned mutualism. What specificity exists is likely to be at taxonomic levels higher than the species and will reflect mutually compatible biological groundplans rather than pairwise reciprocal evolution of species pairs. Pairwise coevolution should be largely restricted to depauperate communities and will almost certainly be rare in highly diverse biotas.