Bird-landscape relations in the Chihuahuan Desert: Coping with uncertainties about predictive models

During the springs of 1995-1997, we studied birds and landscapes in the Chi huahuan Desert along part of the Texas-Mexico border. Our objectives were t o assess bird-landscape relations and their interannual consistency and to identify ways to cope with associated uncertainties that undermine confiden ce in using such relations in conservation decision processes. Bird distrib utions were often significantly associated with landscape features, and man y bird-landscape models were valid and useful for predictive purposes. Diff erences in early spring rainfall appeared to influence bird abundance, but there was no evidence that annual differences in bird abundance affected mo del consistency. Model consistency for richness (42%) was higher than mean model consistency for 26 focal species (mean 30%, range 0-67%), suggesting that relations involving individual species are, on average, more subject t o factors that cause variation than are richness-landscape relations. Consi stency of bird-landscape relations may be influenced by such factors as pla nt succession, exotic species invasion, bird species' tolerances for enviro nmental variation, habitat occupancy patterns, and variation in food densit y or weather. The low model consistency that we observed for most species i ndicates the high variation in bird-landscape relations that managers and o ther decision makers may encounter. The uncertainty of interannual variation in bird-landscape relations can be reduced by using projections of bird distributions from different annual m odels to determine the likely range of temporal and spatial variation in a species' distribution. Stochastic simulation models can be used to incorpor ate the uncertainty of random environmental variation into predictions of b ird distributions based on bird-landscape relations and to provide probabil istic projections with which managers can weigh the costs and benefits of v arious decisions. Uncertainty about the true structure of bird-landscape re lations (structural uncertainty) can be reduced by ensuring that models mee t important statistical assumptions, designing studies with sufficient stat istical power, validating the predictive ability of models, and improving m odel accuracy through continued field sampling and model fitting. Uncertain ty associated with sampling variation (partial observability) can be reduce d by ensuring that sample sizes are large enough to provide precise estimat es of both bird and landscape parameters. By decreasing the uncertainty due to partial observability, managers will improve their ability to reduce st ructural uncertainty.