Demographic and genetic structure of a severely fragmented population of the endangered hog deer (Axis porcinus) in the Indo-Burma biodiversity hotspot

The analysis of the combined period 2003–2018 showed a significant increase in the popula- tion (p<0.05, R2 = 0.608) at the rate of 10% annum-1. The population structure was skewed towards adult females. The number of females was higher as compared to males, juveniles and fawns (Fig 3A and 3D). The overall numbers of adult males, adult females and juveniles fluctuated, whereas the number of fawns declined steadily (Fig 3A–3C). The adult sex ratios were 34.2, 34.5 and 39.9 males per 100 females and the doe-to-fawn ratios were 16.4, 17.2 and 15.8 fawns per 100 females respectively, during the study period. Overall, for the three study years, the male-to-female ratio was 36.2 ± 1.9 males per 100 females, and the doe-to-fawn ratio was 16.5 ± 0.4 fawns per 100 females respectively (Table 2). Genetic diversity in the wild Twenty-seven unique individuals were identified from the 38 biological samples collected, and used for further analysis (Table 3). Of the 25 microsatellites screened, 23 were polymorphic, and two loci were non-amplified. The genetic variability study of the hog deer population (n = 27) showed that the number of alleles observed across the 23 loci varied from 2 to 5, whereas the mean number of alleles per locus was 2.70 ± 0.18. The average observed heterozy- gosity (Ho) was estimated at 0.42 ± 0.02, and the expected heterozygosity (He) was estimated at 0.51 ± 0.03. The mean PIC value was 0.43 ± 0.03 (Table 3). Out of the 23 microsatellites, no deviation from the HWE was observed (Table 3). Bottleneck detection

Of the 23 loci found to be polymorphic in the wild population, only seven loci showed a signif- icant heterozygosity excess “S3 Table”. Overall, the Wilcoxon test showed no evidence of a bot- tleneck in the past (p = 0.894 for IAM; p = 0.993 for TPM; p = 0.999 for SMM). The global analysis indicated no significant heterozygosity excess (He < Heq): 0.252 ± 0.025 < 0.284 ± 0.018 (under IAM); 0.252 ± 0.025 < 0.323 ± 0.021 (under TPM); 0.252 ± 0.025 < 0.359 ± 0.026 (under SMM). The hog deer population did not encounter any genetic bottleneck in the recent past as the allele frequency distribution had an approximately L-shaped curve (as expected under a mutation–drift equilibrium). Although this population is small and isolated, and has gradually declined in recent years, the selected microsatellite loci indicates that it has retained moderate level of heterozygosity. Table 1. Density estimate of hog deer (2006–08) in Keibul Lamjao National Park, India. Parameters Estimates % CV 95% CI Minimum Maximum 2006 2007 2008 2006 2007 2008 2006 2007 2008 2006 2007 2008 p� 0.38 ± 0.58 0.31 ± 0.29 0.30 ± 0.25 15.6 9.59 8.48 0.27 0.25 0.25 0.52 0.37 0.38 n/k� 0.36 ± 0.33 0.22 ± 0.20 0.24 ± 0.20 9.33 10.3 10.5 0.29 0.18 0.18 0.44 0.27 0.28 EDR� 245 ± 19.2 240 ± 11.5 241 ± 10.2 7.83 4.8 4.24 208.9 217.8 220.8 287.8 265.6 263 DS� 1.92 ±0.35 1.19 ± 0.17 1.25 ± 0.17 18.2 14.1 13.5 1.33 0.90 0.95 2.76 1.59 1.63 ES� 1.53 ± 0.10 2.29 ± 0.18 2.02 ± 0.18 6.74 8.2 8.87 1.33 1.93 1.68 1.75 2.72 2.42

D� 2.94 ± 0.57 2.75 ± 0.44 2.51 ± 0.40 19.46 16.3 16.2 1.99 1.99 1.82 4.32 3.80 3.46 N� 65 ± 12.6 61 ± 9.9 57 ± 9.2 19.46 16.3 16.2 44 44 41 96 84 79 �p, probability of detection under the curve; n/k, encounter rate; EDR, effective detection radius; DS, group density; ES, group size; D, individual density; N, population. https://doi.org/10.1371/journal.pone.0210382.t001 Demographic and genetic structure of hog deer in the Indo-Burma biodiversity hotspot PLOS ONE | https://doi.org/10.1371/journal.pone.0210382 February 6, 2020 6 / 16