Breeding bird species richness in Spain: assessing diversity hypothesis at various scales Fernando Gonza ´lez-Taboada, Carlos Nores and Miguel A ´ ngel A ´ lvarez F. Gonza ´lez-Taboada (fgleztab@yahoo.com), A ´ rea de Ecologı ´a, Facultad de Biologı ´a, Univ. de Oviedo, C/Catedra´tico Rodrı ´guez Urı ´a s/n, ES-33071, Oviedo, Spain.  C. Nores and M. A ´ .A ´ lvarez, Indurot, Campus de Mieres, ES-33600, Mieres, Asturias, Spain. The variation of passerine species richness in Spain was studied at various spatial scales. Presence-absence data was resampled to construct three species richness maps in lattices of 10 10, 30 30, and 50 50 km UTM cells. The importance of habitat, species-energy, climatic variability, disturbance, history and geometric constraints hypotheses was assessed using geographical data. Stochastic, range-based models were used to simulate neutral colonization events from Europe or from Africa. The importance of small scale processes remained after the inclusion of environmental covariates, indicating a possible role of ecological interactions that was represented in the models by a conditional spatial autoregressive term. Historical effects and energy related measures explained most of the variation in regional species richness. Local and regional habitat structure measures explained the pattern only after large scale trends were considered. The differences when species richness was analyzed at each scale reveal the importance of spatial issues in diversity studies. The possible role of post glacial migration in shaping the observed patterns, and implications for conservation are discussed. The study of species diversity has a long tradition in ecology and biogeography. Most emphasis has been placed on the latitudinal diversity gradient, with the development of multiple non exclusive hypotheses (Pianka 1966, Rosenzweig 1995, Willig et al. 2003). Recent studies focused on the variation of the im- portance of these explanations according to changes of location and scale (Rahbek and Graves 2000, 2001, Whittaker et al. 2001, Blackburn and Gaston 2002, Willig et al. 2003, Cushman and McGarigal 2004, Rahbek 2005). Although a great variety of hypotheses have been proposed, historical factors, area, habitat, climate, and neutral theories remain among the most prominent. Given the hierarchical organization of ecological communities, it is expected that the main factors driving diversity gradients depend on both spatial scale and extent of analyses (Whittaker et al. 2001). Historical factors (speciation and dispersal) are frequently invoked to explain diversity variation at broad scales, from continents to specific regions (Rosenzweig 1995, Ricklefs 2004). On the other hand, present ecological factors could explain diversity from a bottom-up perspective, portraying a dynamic interpretation of the observed patterns (Whittaker et al. 2001). Area and climate-energy theories are the most supported by data and theoretical explorations (Rosenzweig 1995, Currie et al. 2004). An increase in area offers more opportunities to geographical speciation and supports a great variety of habitats (Rosenzweig 1995). That is, either though historical considerations, sampling artefacts or niche availability, larger areas contain more species. These considerations could be viewed as a sampling artefact; i.e. a greater area contains more individuals which may cause a false impression of higher richness (Gotelli and Colwell 2001). Therefore, area should be viewed as an un- informative covariate, being preferable to replace its effect on diversity by habitat measures (Whittaker et al. 2001). This is especially important if factors related to habitat loss and fragmentation are taken into account (Gotelli and Colwell 2001). Under the energy hypothesis (Wright 1983), ecolo- gical communities are assumed to be limited by resource availability and, therefore, zones with great energy inputs should allow the coexistence of more Ecography 30: 241  250, 2007 doi: 10.1111/j.2007.0906-7590.04824.x Copyright # Ecography 2007, ISSN 0906-7590 Subject Editor: Carsten Rahbek. Accepted 6 March 2007 241