Is niche divergence more likely in parapatry? A test inSclerurus mexicanussensu lato (Aves: Furnariidae)

ABSTRACT Aim While species’ niches are understood to diverge through time, the mode of these divergences is not fully understood. The null hypothesis of niche divergence in the present is that of conservatism; that species are likely to occur in the same environment as their ancestors and, therefore, as allopatric sister species. Alternatively, species are expected to diverge when selective pressure, such as parapatry with a congener, occurs. Here, we analyse niche divergence in a mosaic of allo- and parapatric Sclerurus mexicanus sensu lato populations to determine if niche divergence is more likely in parapatry. Location Neotropics. Methods We created a dataset of 1,100 vetted localities for S. mexicanus sensu lato, and assigned each point to each of the seven described populations. We created individual dispersal areas for each species in Maxent following: 1) a broad approach in which training areas where allowed to overlap with parapatric congeners; and 2) a narrow approach in which species dispersal areas adhered to strict parapatry and there was no overlap in training areas. We complemented these ENMs with ‘random’ models drawn from points within each training area, thus creating ‘null’ ENMs against which we could test niche divergence using the metric Schoener’s D. Results There was no significant difference in the performance of broad and narrow training areas. Few significant divergences were found, and all of those that were found consisted of allopatric populations. Partial divergences were frequently recovered when comparing parapatric taxa. Main conclusions In contrast to our expectation, we found no significant evidence for increased niche divergence in parapatric versus allopatric species. Possible explanations for these findings include true conservatism even among parapatric lineages ( e.g ., additional environmental or biological factors may delimit distributions) or artefactual errors inherent to model creation. We discuss the implications of these findings, and discuss ways to improve upon tests of niche divergence in the future, especially when dealing with cases of parapatry.