de preferencias de hábitat y de distribución y abundancia invernal de aves en el centro de España. Análisis y predicción del efecto de factores ecológicos

Patterns of habitat preference and distribution and abundance of wintering bird fauna in central Spain. Analysis and prediction of the effect of ecological factors.— This paper analyses the effect of geographic, topographic, land use and habitat structure variables on the composition and structure of wintering bird communities in Central Spain (Iberian peninsula). Parameters describing the avifauna varied in a predictable way considering a small group of coarse–grained variables defining the geographical and altitudinal location of the censuses, and the basic characteristics of the structure and typology of habitats: 49–76% of variance accounted for total bird density and for abundance of four ecological groups, 37–63% accounted for species richness and diversity, and 65% explained the relative abundance of species with conservation problems according to the European scale (SPEC figures). Regarding the most widespread species in the study area, significant models were obtained by means of tree regression analysis for 50 species, with an average reduction of deviance of 39%. Altitude was the most important variable affecting bird community parameters and abundance of each bird species, showing a consistent and marked negative effect. Structural complexity of the vegetation and geographical location followed as the variables of importance explaining variability. The habitats with the lowest bird density, richness and diversity of birds were mountain grasslands/shrublands, young pine re–forestations, and Pyrenean oak forests at 1,200–1,600 m a.s.l. The habitats with the highest values on these parameters were riparian forests, agricultural mosaics, and holmoak "dehesa" parklands, mainly located at the southern and western part of the region and at intermediate altitudes. The total density of birds increased from east to west, was higher in intermediate altitudes than in the extremes of the altitudinal range, and increased with habitat structural complexity (i.e., vertical development and degree of vegetation cover), agricultural use of the land, and the presence of water (e.g., streams, rivers, flooded areas). Density of strictly wintering species in the study region decreased latitudinally from southeast to northwest, being higher at intermediate altitudes in localities with presence of water and woodlands dominated by coniferous trees. Abundance of facultative or obligate frugivorous species was very low. Richness of species was higher toward the western part of the study area, increased with habitat structural complexity, and was lower at higher altitudes. On the other hand, the less diverse bird assemblages were those that inhabit agricultural landscapes and/or areas located at higher altitudes. Habitats and areas with a higher relative abundance of species with conservation problems at the European scale (SPEC scores) were located at intermediate altitudes in the southwest and southeast of the study region. The relative abundance of this group of species was also associated to the presence of water, habitat structural complexity and agricultural use. The relationship between the European conservation status of species (SPEC scores) and the patterns of distribution, abundance, habitat preferences and ecological width of 72 species was also analysed. Species with more conservation problems on the European scale have a marked preference for structurally simpler habitats (e.g., agricultural and grasslands habitats) and have a broader altitudinal and among–habitats distribution. Results from the 44 census localities were extrapolated to the remaining region using a geographical information system in order to build predictive maps for density, species richness, species diversity and weighed European conservation status. This work shows that valuable knowledge can be obtained from fragmentary and dispersed data, in order to describe general patterns of distribution, abundance and habitat preferences of birds. This methodological approach could be a valid in environmentally heterogeneous, large regions, with few qualified bird observers and researchers.