Effects of including non-breeding bird species on predicted bird distributions for conservation planning in New Mexico

There has been uncertainty in the national gap analysis program about inclu ding non-breeding birds in distribution models because of concerns that dis tributions of migrant and winter birds are difficult to predict and are not necessary to assess biodiversity patterns. New Mexico gap analysis include d migrant and non-breeding birds assuming that distributions could be predi cted using habitat associations, and that excluding non-breeding bird habit at from avian richness projections potentially underestimates community typ es important to birds. We compared biodiversity estimates including non-bre eding birds (inclusive estimate-324 species) to estimates including only br eeding birds (breeding bird estimate-257 species) in terms of estimated pat terns of species richness. Inclusive and breeding bird richness estimates a greed about general locations of some species-rich areas and the most speci es-poor areas in the state, but were less comparable for intermediate al:ea s of bird occurrence. We found < 50% agreement between the two estimates ab out areas with highest species richness. When non-breeding birds were inclu ded, over 2,000,000 ha of short grass steppe shifted to a higher richness c ategory and another 11,600 ha moved into the highest richness category. Gra minoid wetlands, playa lakes, and waters in eastern New Mexico also showed elevated richness levels in the inclusive estimate. Our analyses indicate t hat only assessing breeding distribution does not reliably predict relative importance of areas used by birds throughout New Mexico and should not be used exclusively to identify potential gaps in conservation for land-use ev aluation and planning. (C) 2001 Elsevier Science Ltd. All rights reserved.