In spite of increasing application of presence-only models in ecology and conservation and the growing number of such models, little is known about the relative performance of different modelling methods, and some of the leading models (e.g. GARP and ENFA) have never been compared with one another. Here we compare the performance of six presence-only models that have been selected to represent an increasing level of model complexity [BIOCLIM, HABITAT, Mahalanobis distance (MD), DOMAIN, ENFA, and GARP] using data on the distribution of 42 species of land snails, nesting birds, and insectivorous bats in Israel. The models were calibrated using data from museum collections and observation databases, and their predictions were evaluated using Cohen's Kappa based on field data collected in a standardized sampling design covering most parts of Israel. Predictive accuracy varied between modelling methods with GARP and MD showing the highest accuracy, BIOCLIM and ENFA showing the lowest accuracy, and HABITAT and DOMAIN showing intermediate accuracy levels. Yet, differences between the various models were relatively small except for GARP and MD that were significantly more accurate than BIOCLIM and ENFA. In spite of large differences among species in prevalence and niche width, neither prevalence nor niche width interacted with the modelling method in determining predictive accuracy. However, species with relatively narrow niches were modelled more accurately than species with wider niches. Differences among species in predictive accuracy were highly consistent over all modelling methods, indicating the need for a better understanding of the ecological and geographical factors that influence the performance of species distribution models.

Explaining patterns of spatial variation in species composition is a major challenge facing ecologists. While many studies have focused on patterns of species richness and their causal factors, the spatial change in species composition has received far less attention (but see Harrison et al. ABSTRACTAim The aim of this paper is to evaluate the combined effects of geographical distance and environmental distance on patterns of species similarity (similarity in species composition between sites), and to identify factors affecting the rate of decay in species similarity with each type of distance.Location Israel.Methods Data on species composition of land snails and land birds were recorded in 27 sites of 1 · 1 km scattered across a rainfall gradient in Israel. Matrices of similarity in species composition between all pairs of sites were computed and analysed with respect to corresponding matrices of geographical distance and rainfall distance (defined as the difference in mean annual rainfall between sites, and used as a measure of environmental distance). Mantel tests were applied to determine the correlation between species similarity and each type of distance. Factors affecting the decay in species similarity were investigated by comparing different subsets of the data using randomization tests.Results Both rainfall distance and geographical distance had negative effects on species similarity. The effect of rainfall distance was statistically significant even after controlling for differences in geographical distance, and vice versa. The perunit effect of rainfall distance on species similarity decreased with increasing geographical distance, indicating that the two types of distances interacted in determining the similarity in species composition. Snails showed a higher rate of decay in species similarity with geographical distance than birds, and large snails showed a higher rate of decay than small snails, which are better passive dispersers. The per-unit effects of both rainfall distance and geographical distance on species similarity were higher in the desert region than in the Mediterranean region. Analyses focusing on a grain size of 10 · 10 m showed a lower similarity in species composition and a lower rate of decay in species similarity with rainfall distance than analyses carried out at a grain size of 1 · 1 km.Main conclusions Patterns of similarity in species composition are influenced by the combined effects of environmental variation, the position of the area along environmental gradients, the dispersal properties of the component species, and the scale (both spatial extent and grain size) at which the patterns are examined.

Summary 1.Species distribution models (SDM) are increasingly applied as predictive tools for purposes of conservation planning and management. Such models rely on the concept of the ecological niche and assume that distribution patterns of the modelled species are at some sort of equilibrium with the environment. This assumption contrasts with empirical evidence indicating that distribution patterns of many species are constrained by dispersal limitation. 2. We demonstrate that the performance of SDM based on presence-only data can be significantly enhanced by incorporating distance constraints (functions relating the likelihood of species' occurrences at a site to the distance of the site from known presence locations) to the modelling procedure. This result is highly consistent for a variety of niche-based models (ENFA, DOMAIN and Mahalanobis distance), distance functions (nearest neighbour distance, cumulative distance and Gaussian filter) and taxonomic groups (plants, snails and birds, a total of 226 species). 3. Distance constraints are expected to enhance the accuracy of niche-based models even in the absence of strong dispersal limitation by accounting for mass effects and spatial autocorrelation in environmental factors for which data are not available. 4. While distance-based methods outperformed niche-based models when all data were used, their accuracy deteriorated sharply with smaller sample sizes. Niche-based methods are shown to cope better with small sample sizes than distance-based methods, demonstrating the potential advantage of niche-based models when calibration data are limited. 5. Synthesis and applications. Incorporating distance constraints in SDM provides a simple yet powerful method to account for spatial autocorrelation in patterns of species distribution, and is shown empirically to improve significantly the performance of such models. We therefore recommend incorporating distance constraints in future applications of SDM.

In a review of recent challenges in conservation planning, Ferrier (2002) proposed the incorporation of models of similarity in species composition as a means for prioritizing areas for biodiversity conservation. A key assumption of this approach is that estimates of compositional similarity derived from models of similarity in species composition can be used as effective surrogates for real similarity data. We used data on snail distribution in Israel to test this assumption. We used two types of models to analyze patterns of similarity in species composition: one based on presence/absence data and the second based on abundance data. Both models accounted for large amounts of the observed variation in compositional similarity. Variation-partitioning analysis indicated that a considerable amount of the variation in compositional similarity could be separated into "pure" geographical versus "pure" environmental components, indicating that reserve selection procedures should take into account spatial considerations in determining priorities for conservation. The relative effects of geographical versus environmental factors varied between the two types of models, indicating that different indices of similarity should be used if one wishes to represent species composition per se or ecological communities including their relative species abundances. A comparison of distribution patterns of land snails and land birds in a subset of the study sites revealed a high degree of congruence in compositional similarity between the two groups. Moreover, compositional similarity in snails was a better predictor of compositional similarity in birds compared with all environmental and geographical distances taken together. Models calibrated based on data collected in small plots explained a considerable amount of the variation observed at larger scales, suggesting that sampling efforts required for conservation planning might be lower (and thus, more feasible) than assumed previously. Models of similarity in species composition may serve as an important tool for conservation planning.Resumen: En una revisión de los recientes retos de la planificación de la conservación, Ferrier (2002) propuso la incorporación de modelos de similitud en la composición de especies como una manera de priorizaŕ areas para conservación de la biodiversidad. Una suposición clave de este método es que se pueden utilizar las estimaciones de similitud en la composición derivadas de modelos de similitud en la composición de especies como sustitutos efectivos de datos reales de similitud. Utilizamos datos de la distribución de caracoles terrestres en Israel para probar esta suposición. Utilizamos dos tipos de modelo para analizar patrones de similitud en la composición de especies: uno con base en datos de presencia/ausencia y el segundo con base en datos de abundancia. Ambos modelos explicaron gran parte de la variación observada en la similitud en la composición. El análisis de partición de la variación indicó que una parte considerable de la variación en ...

No abstract