Assessing the Relative Role of Environmental Factors That Limit the Distribution of the Yucatan Rattlesnake (Crotalus tzabcan)
We modeled the ecological niche of the Yucatan Rattlesnake (Crotalus tzabcan) to map its environmental suitability, estimate its potential distribution, and identify the environmental factors that limit its geographic range within the Yucatan Peninsula (YP). Niche models were built to evaluate, separately and together, the ability of different sets of environmental predictors to explain the distribution of this species. The relative contribution percentage from each variable was calculated from the models with better predictive ability. Response curves were also constructed to observe the influence of predictors on the estimated environmental suitability for C. tzabcan. According to our niche models, even though the distribution of this species covers a large part of the YP, the northwest portion is likely the most environmentally suitable region. By contrast, populations from the north of Guatemala and Belize (south of the YP) are placed within the limits of the species' niche, and the environmental conditions toward the base of the peninsula are not suitable for this species. Precipitation was consistently the most important factor to explain the distribution of C. tzabcan. Although unlikely to directly affect the physiology of C. tzabcan, precipitation affects other factors (e.g., predator, prey, and competitor densities; undergrowth coverage; and relative humidity) that influence survival. Our study contributes new and detailed knowledge of the distribution ecology of a species that was once the most culturally valued in the YP.Abstract
Study area (biogeographical region of the YP). Red circles represent localities where the presence of Crotalus tzabcan has been documented. Numbers indicate the administrative units that comprise this region: 1 = west of Tabasco, 2 = northwest of Chiapas, 3 = Campeche, 4 = north of Guatemala, 5 = Yucatán, 6 = Quintana Roo, and 7 = Belize.
Performance comparison (AUC ratio, OR, and AICc) of the niche models constructed with five different sets of environmental predictors (set1 = solar radiation variables, set2 = water vapor pressure variables, set3 = temperature variables, set4 = precipitation variables, and set5 = all variables but eliminating colinearity; |ρ| > 0.85). Boxplots represent differences across specific Maxent settings (combinations of regularization multipliers and features).
Ecological niche models of Crotalus tzabcan. Black points represent the presence records used to train models and crosses represent evaluation records. Environmental suitability estimated with Maxent is shown in the left panel. The small left panel represents the range of differences across Maxent replicates. Potential distribution determined from environmental suitability models according to three thresholds of permitted omission (E parameter) is shown in the right panel.
(Top) Biplot of the two variables with highest contribution to the ecological niche model of Crotalus tzabcan (combinations of these two available variables in the YP is shown in dark blue, whereas the combinations associated with the localities where the species has been observed is shown in light blue). (Bottom) Biplot of point density to identify the combinations of the two variables with better representativeness in the species' records.
Response curves of the two variables with highest contribution to niche models of C. tzabcan generated with Maxent.
Contributor Notes
