Character Displacement in Evolutionary Time

Patterns of distribution of morphology in the fossil recordshould be examined more rigorously to evaluate the relativeeffectiveness of alternative evolutionary models and to integratethe dimension of true evolutionary time more directly with evolutionarytheory. Grant (1972) has discussed three main problems in the detectionof character displacement: we must assume that sympatry followedallopatry, that the state in allopatry represents the static,pre-contact state, and that there are no other reasons accountingfor differences within a species in allopatric and sympatriczones. These problems can be largely avoided with suitable paleoiitologicaldata. In an example involving the coincident 10 million-year historiesof two congeneric species of Devonian trilobites, the one knowninstance of sympatry reveals marked divergent shifts in onespecies and a "mixed" reaction of convergent, divergent, andneutral shifts in the other species. The characters undergoingthese shifts are largely those which (i) serve as specific differentia,and (ii) show the most change within one of the species overits entire history. This suggests that character displacementmay be simply a magnified microcosm of a general pattern ofinteraction between two species even when allopatric, providedthat allopatry is sustained through competitive exclusion.