Population genetics and habitat evaluation to inform conservation of Asian elephants in China

Abstract

China lacks a national conservation action plan for Asian elephants (Elephas maximus), and genetic information on remaining populations, which are confined to Yunnan Province, and a comprehensive evaluation of habitat suitability for these animals are lacking. I extracted DNA from dung samples from all five remaining elephant populations in China, used mitochondrial and microsatellite markers to investigate their genetic and demographic structure. I also documented the extent of genetic variability of elephants within China and compared it with elsewhere in Asia. In addition, I used multiscale multivariable species-distribution models to investigate the effects of anthropogenic variables (such as landscape fragmentation, roads and settlements) on habitat suitability for elephants throughout Yunnan Province, including places beyond the current range of elephants where legal protection from poaching and human persecution could allow elephants to become established. A total of 237 unique genotypes (153 females, 84 males) were identified from 497 dung samples obtained from five populations, representing 81% of wild elephants in China. The effective population size was small (28, range 25 – 32), but the total population showed no signs of being inbred (FIS = 0.04). However, a high level of inbreeding (FIS = 0.35) was observed in Nangunhe, where the population is small (22 – 28 individuals) and isolated. Historic demographic contraction likely accounted for low total population haplotype diversity (Hd = 0.235), but moderate nucleotide and nuclear diversity (π = 0.6%, He = 0.55) was attributable to post-bottleneck recovery involving recent population growth plus historical gene exchange with elephants in Myanmar, Lao PDR and Vietnam. The five populations fell into three clusters, with Nangunhe elephants differing consistently from the other four populations (FST = 0.23); elephants from Mengyang, Simao and Jiangcheng belonged to a single population (henceforth, MSJ) that differed from the Shangyong population (FST = 0.11). Habitat-suitability modelling indicated that elephant presence was sensitive to the scale at which each variable was measured, and multivariate models with variables at their optimum scales had higher discrimination than models with variables at uniform scales. Anthropogenic variables measured at coarse scales of 8-, 16- or 32-km had the greatest influence on elephant presence. A total area of 12,942 km2 of habitat, mostly along the Myanmar and Lao PDR borders, was predicted to be highly suitable for elephants, and 1,584 km2 of potentially suitable habitat was identified in three prefectures where there are currently no elephants. I estimate that Yunnan Province could support 588 (range 181 – 1,114) more elephants, twice the current population, and even more if habitat restoration and enhancement was undertaken. My genetic findings suggest that Chinese elephants should be managed as two distinct units: Nangunhe, and another combining Shangyong and MSJ. Long-term viability of both units will require transboundary initiatives to restore gene flow between elephants in China and neighboring countries through establishment of movement corridors. Furthermore, because 95% of habitat predicted to be suitable was outside protected areas, these lands should be incorporated within a new national park for elephants proposed by the state and Yunnan provincial governments in 2021.