ABSTRACT
Background: Defining Andean anurans through their altitudinal limits has been a common practice in species lists, studies of responses to climate change among others, especially in the northern Andes. At least three proposals to differentiate Andean anurans from lowland anurans through elevation and at least one to differentiate Andean anurans from high mountain anurans have been formulated. However, the most frequently used altitudinal limits are not based on theoretical or numerical support, but on observations or practical definitions. Additionally, these proposals have been applied equally to different portions of the Andes, ignoring the fact that even between slopes of the same mountain, environmental conditions (and therefore the distribution of species) may differ. The objective of this work was to evaluate the concordance between the altitudinal distribution of anurans in the Colombian Andes and four different altitudinal delimitation proposals. Methods: We constructed our study area in a manner that allowed us to include species from the Andean region (as traditionally defined) and adjacent lowlands, because if the boundaries criteria were applied they would separate the species of the latter by themselves. We divided the study area into eight entities according to the watershed and the course of the most important rivers. We conducted a bibliographic search for all anurans in the cordilleras and inter-Andean valleys of Colombia and complemented the search with information on anurans for the region available in the GBIF. After curing the species distribution points, we generated elevation bands of 200 m amplitude for both the study area and for each Andean entity. Subsequently, we performed a cluster analysis to evaluate the grouping of the elevation bands according to their species composition. Results: In none of the cases (neither for the entire study area nor for any of the entities) we found a correspondence of any of the traditionally used boundaries and the altitudinal distribution of Anurans in the Andean region of Colombia. Instead, on average, the altitudinal delimitation proposals arbitrarily spanned the altitudinal distribution of about one third of the species distributed in the study area. Conclusions: We suggest that, although, based on our results, some Andean entities can be divided according to the altitudinal composition of the species that occur in them, we did not find any results that support the idea of a generalizable altitudinal limit for the Colombian Andes. Thus, to avoid biases in studies that may later be used by decision makers, the selection of anuran species in studies in the Colombian Andes should be based on biogeographic, phylogenetic or natural history criteria and not on altitudinal limits as they have been used.