The Role of Ecotones in the Distribution of Andean Birds

This is the final paper of a five—part series on the distribution of birds in the Apurimac Valley of Peru. In one of the previous papers, I classified the upper and lower limits of avian distribution in the Cordillera Vilcabamba, Peru, as follows: (1) those that abutted on the range of a congener (presumably evidence of competitive exclusion), (2) those that approximately coincided with an ecotone, and (3) all others. A test of the limits presumptively enforced by competitive exclusion has previously been published. The present study involved limits in the second of these categories. For a set of 47 bird species whose limits (16 upper, 31 lower) coincided with the montane rain forest—cloud forests ecotone on the Vilcabamaba control transect, I determined whether, as predicted, the species expanded or contracted their distributions in localities in which the homologous ecotones were displaced upward or downward relative to the control elevation. Where the ecotone was displaced away from species' center of distribution (e.g., downward for species whose lower limits coincided with the ecotone on the control transect), 36 of 41 species (88%) were found to have expanded their distributions. This result upholds the provisional assessment of the ecotone as a distributional barrier to these species in the control locality. Where the ecotone was shifted toward species' centers of distribution, 43 out of 44 species (98%) had failed to contract fully in distribution; they were found on the other side of the ecotone, in what had been predicted to be alien habitat. This result was not anticipated. It is illustrative of a tendency of species to occupy a greater range of habitats near to vs. far from their centers of distribution. I also found that widespread Andean bird species had expanded distributions, in both upward and downward directions, in the biogeographically isolated Cordillera de la Costa of Venezuela. Reduced avian species diversity in this range relative to the main Andean chain appears to have led to a general relaxation of distributional restraints. Combining the results reported here with those presented in previous publications, I conclude that direct and diffuse competitive exclusion account for about two—thirds of the distributional limits of Andean birds, ecotones for about one—sixth, and unspecified factors that vary in parallel with the environmental gradient for the remaining one—sixth. Competitive exclusion emerges as the factor of overriding importance in the exceedingly diverse Andean fauna. In contrast, in temperature mountains, ecotones play a predominant role in limiting distributions, and competitive interactions are far less important.