Biogeography of polymorphic phenotypes: Mapping and ecological modelling of coat colour variants in an elusive Neotropical cat, the jaguarundi (Puma yagouaroundi)

Abstract

The jaguarundi Puma yagouaroundi is a small Neotropical cat that presents two main coloration phenotypes (grey/dark vs. reddish). Although these coat colour variants have been known for decades, and historically speculated to be associated with different habitats, their exact geographical distribution has never been mapped. Moreover, their association to different habitats has so far not been tested statistically, so that their ecological relevance with respect to varying environmental features remains unknown. Based on 566 location records encompassing the entire historical range of the species obtained from camera‐traps, captures and skins held in scientific collections, we produced suitability models for both jaguarundi phenotypes using maximum entropy algorithms of niche modelling. The frequency of grey/dark jaguarundis is c. 80%, whereas reddish animals represent c. 20% of our overall sample set. However, there were marked differences in these frequencies across regions. Although the spatial distribution of grey/dark animals did not depart substantially from random expectations (as it encompassed the whole species range), the occurrence of the ancestral reddish form was strongly and significantly non‐random. In spite of their broad distribution across multiple habitats, grey/dark animals were significantly associated with moist and dense forests, whereas reddish forms were associated with dry and open areas such as deserts and xeric landscapes. Furthermore, there were clear spatial differences in the suitability models generated for these coat colour phenotypes. We also employed the distribution models to investigate whether particular environmental predictors could explain these different distributions. Predictors related to moisture were especially influential on the differences between the grey/dark and reddish models, and demonstrate an effect of natural selection on coloration traits, suggesting that a complex interplay of different ecological processes regulates this system over evolutionary time.

Mapping black panthers: Macroecological modeling of melanism in leopards (Panthera pardus)

“…To perform in-depth spatial analyses of the two phenotypes (non-melanistic and melanistic), we used the approach proposed by [38] to generate separate potential distribution models for each of them. We used 38 explanatory environmental variables and landscape data: 35 bioclimatic variables obtained from the Worldclim (http://www.worldclim.org) and Climond (http://www.climond.org) databases, a digital elevation model (obtained from the SRTM — http://www2.jpl.nasa.gov/srtm), landscape features (obtained from the ESA GlobCover Project 2009 — http://due.esrin.esa.int/page_globcover.php, and NASA NEO NDVI Modis — http://neo.sci.gsfc.nasa.gov).…”

Occurrence of leucism in Eira barbara (Carnivora, Mustelidae) in Brazil

“…The occurrence of the leucistic phenotype observed in the Oriximiná and Caraça populations may represent an opportunity to investigate the possibility of establishing a recessive trait in these populations, such as the near fixation of melanism, another recessive color anomaly, observed in leopards of the Malay Peninsula (Kawanishi et al 2010). On the other hand, the occurrence of different coloration phenotypes in tayras throughout its distribution area, provides an opportunity to investigate possible adaptive advantages of these phenotypes (Silva et al 2016) in relation to the ecological conditions present in their areas of occurrence. …”

A review of Gloger's rule, an ecogeographical rule of colour: definitions, interpretations and evidence

“…Following Booksmythe et al () I adopted a conservative approach and assigned an effect of 0 to missing effects and an effect opposing the predictions of Gloger's rule when only the sign of the effect was missing. For four effects (four studies) I could not extract a standardised effect due to the nature of the analyses (climate niche modelling, structural equation modelling, canonical correlations; Watson & Hanham, ; Schueler, ; da Silva et al , , ). But since in all cases the authors concluded that the results supported Gloger's rule, I decided to exclude these studies rather than assign them an effect of 0.…”