Global Ecology and Conservation (Dec 2023)
Predicting population size at large scale: The case of two large felids
Abstract
Approaches that allow capitalizing on local population estimates to derive global population estimates with associated uncertainty are urgently needed, especially for naturally rare species of conservation concern. Here we used published population density estimates to predict large-scale density patterns and derive global population estimates for two species of large felids, the leopard and the tiger. We modelled population density for the leopard (n = 392) and the tiger (n = 547) as a function of environmental and anthropogenic variables, while controlling for differences in sampling method and sampling area, time of data collection, spatial autocorrelation, subspecies and political protection. We used Bayesian inference to generate a distribution of plausible population sizes. Both species showed higher densities in high productivity areas, the leopard being more abundant in high precipitation, high level of terrain roughness and agricultural areas, and the tiger in areas with low croplands and low roughness. Primary roads density showed a negative effect on both species. Secondary roads density was associated to higher densities for the leopard but lower densities for the tiger. Livestock biomass showed a humped relationship with tigers’ densities. Temporal trends in average density were negative for the tiger, experiencing an average decline of 34% (IQR: 11% − 53%). In contrast, the trend for leopards showed a marginal, yet uncertain, increase in recent years 21% (IQR: −5% − 57%). We predicted a global population estimate of 261,636 (IQR = 146,768 − 461,512) and 5201 tigers (IQR = 2596 − 10,460). Large-scale models of population density that rely on unstructured data can contribute to our understanding of species ecology, produce robust population size estimates for conservation assessment and inform large-scale conservation planning. At the same time, the uncertainty around these estimates highlights the limited knowledge available for these species which should be accounted for in conservation assessments.
