Population structure in the Andaman keelback, Xenochrophis tytleri: geographical distance and oceanic barriers to dispersal influence genetic divergence on the Andaman archipelago

Abstract

Limited gene flow between populations due to geographic distance, presence of barriers or inherent low dispersal ability leads to the formation of genetically structured populations. Strong population structure indicates lowered levels or absence of gene flow which might lead to inbreeding and loss of genetic capacity to recuperate from anthropogenic stress and natural calamities. Terrestrial reptiles are generally known to have low dispersal abilities and few studies have explored drivers of their population structure on continental islands, where both anthropogenic stress and natural calamities are relatively common. We investigated the population structure and drivers of diversification of the Andaman keelback (Xenochrophis tytleri), an endemic, terrestrial and freshwater snake species in the Andaman archipelago, a continental group of islands in the Bay of Bengal. Data was collected from 86 individuals from seven islands and 78 individuals were sequenced for the gene Nuclear Dehydrogenase subunit 4 to identify the number of populations and distribution of genetic diversity across populations. We found 11 haplotypes on seven islands and observed high genetic differentiation between seven populations defined island-wise (F _ST = 0.82). We further tested the number of populations by incorporating spatial data into Bayesian Clustering Analysis (GENELAND) and identified six populations of the Andaman keelback. We tested for the influence of Isolation-by-distance on these populations. While the overall trend showed a positive correlation between geographic and genetic distance, a correlogram revealed that the positive correlation disappears beyond ∼20-40 km. We also tested for the presence of geographical barriers to gene flow using Monmonier's algorithm (SPADS), which identified five barriers to dispersal confirming that there are oceanic barriers to dispersal for some island populations of the Andaman keelback. As the Andaman Islands are arranged almost in a straight line from North to South, our data are insufficient to tease apart the roles of geographical distance and barriers to gene flow. We conclude that salt waters between near islands are weak barriers and as the geographical distance between islands increases, so does the strength of the barrier.

Keywords: Andaman Islands; Colubridae; Dispersal; Endemism; Gene flow; Phylogeography; Population genetics.

Figure 1. Study area map showing sampling locations from seven islands of the Andaman archipelago.

Sampling localities are indicated with white points. Islands with sampling have been labeled by name.

Figure 2. Median-joining haplotype networks.

(A) Median-joining haplotype network from 16S and CytB gene sequences. H 1–H 7: labels of each haplotype. Number of ticks on the network between haplotypes correspond to number of base pair differences between them. (B) Median-joining haplotype network from ND4 gene sequences. H 1–H 11: labels of each haplotype, number of ticks on the network between haplotypes correspond to number of base pair differences. Filled black circles correspond to median vectors. The circle size of each haplotype corresponds to the haplotype frequency.

Figure 3. Correlation between genetic and geographic distances.

(A) Scatter plot depicting the relationship between individual pairwise genetic distances (Tamura-Nei distance) and pairwise geographical distances; (B) Scatter plot depicting the relationship between population pairwise genetic distances (F_ST) and pairwise geographical distances.

Figure 4. Mantel correlogram for individual pair wise genetic distance vs. individual pair wise geographical distances.

All correlations are significant at α = 0.05 with Bonferroni-corrected p-values.

Figure 5. Map of population membership of individuals obtained from GENELAND.

Cluster membership result from GENELAND overlaid with the map of Andaman Islands in CorelDraw X7 (Map not to scale).

Figure 6. Maps of posterior probability of individuals to belong to clusters.

Sampled locations appear as black points on the maps. Maps are color coded based on posterior probability values–red, low probability (0.1) to white, high probability (1.0). (A) Map of posterior probability of individuals to belong to cluster Little Andaman, (B) Map of posterior probability of individuals to belong to cluster North Andaman; (C) Map of posterior probability of individuals to belong to cluster Rangat, (D) Map of posterior probability of individuals to belong to cluster Baratang; (E) Map of posterior probability of individuals to belong to cluster Mayabunder, (F) Map of posterior probability of individuals to belong to cluster South Andaman.

Figure 7. Geographical barriers between Island populations and GENELAND populations of the Andaman Keelback.

(A) Barriers detected between populations from seven islands-North Andaman, Middle Andaman, Baratang, South Andaman, Little Andaman, Havelock and Long Island; upper limit for the number of barriers was set to six. (B) Barriers detected between populations identified in GENELAND-North, Mayabunder, South, Baratang, Rangat and Little; upper limit for the number of barriers was set to five.