The role of geography, environment, and genetic divergence on the distribution of pikas in the Himalaya

Abstract

Pikas (Ochotona Link, 1795) are high-altitude specialist species making them a useful bioindicator species to warming in high-altitude ecosystem. The Himalayan Mountains are an important part of their range, supporting approximately 23%-25% of total pika species worldwide, yet we lack basic information on the distribution patterns. We combine field-based surveys with genetics-based identification and phylogeny to identify differences in species-environment relationships. Further, we suggest putative evolutionary causes for the observed niche patterns.

Location: Himalayan high-altitude region.

Methods: We sampled 11 altitudinal transects (ranging from ~2,000 to 5,000 m) in the Himalaya to establish occurrence records. We collected 223 species records using genetic analyses to confirm species' identity (based on some invasive and mostly noninvasive biological samples). Niche and geographic overlap were estimated using kernel density estimates.

Results: Most pikas in the Himalaya span wide elevation ranges and exhibit extensive spatial overlap with other species. However, even in areas of high species diversity, we found species to have a distinct environmental niche. Despite apparent overlapping distributions at broad spatial scales, in our field surveys, we encountered few cases of co-occurrence of species in the sampled transects. Deeply diverged sister-species pair had the least environmental niche overlap despite having the highest geographic range overlap. In contrast, sister-species pair with shallow genetic divergence had a higher environmental niche overlap but was geographically isolated. We hypothesize that the extent of environmental niche divergence in pikas is a function of divergence time within the species complex. We assessed vulnerability of species to future climate change using environmental niche and geographic breadth sizes as a proxies. Our findings suggest that O. sikimaria may be the most vulnerable species. Ochotona roylii appears to have the most unique environmental niche space, with least niche overlap with other pika species from the study area.

Keywords: Indicator species; Kernel density; niche breadth; noninvasive; pika.

Figure 1

The map shows sampled locations plotted on near 30 m resolution elevation data (source: Shuttle Radar Topography Mission, SRTM). Points and polygons are color‐coded to show species‐wise occurrences and their overall range (source: IUCN and Dahal et al., 2017). The map was made in QGIS version 3.0.0‐Girona (URL: http://qgis.org)

Figure 2

The map shows sampled locations plotted on near 30 m resolution elevation data (source: Shuttle Radar Topography Mission, SRTM) in Eastern, Central, and Western Himalaya. The photograph on the top left shows the pellet pile captured in field. Different transects in the three transect plots (Eastern, Central, and Western Himalaya) are color‐coded, and each transect shows searched sites and sampled species in different colors

Figure 3

Niche overlap of different pika species in the environmental space of the study area (PCA—env). The green color depicts the niche space of the first species, red of the second species and the overlapping range is shown in blue. Niche overlap (D) values are presented for comparisons of similarity of species 1 and 2. Paired overlap values which significantly (p < 0.05) differ from the associated null distribution. The significance test was done for both directions of species pairs (species 1 to 2 and species 2 to 1), and all comparisons were significant in both directions. BA and UDOI correspond to two indices of geographic overlap. Average pairwise genetic distances (GD) are the number of nucleotide differences in cytochrome b gene sequences. The correlation circle in the bottom shows variables contributing to the PC axes of the climatic niche of the study area. The arrow depicts the direction of correlation (same direction indicates a high correlation). Red to blue color (high to low) indicates variables contributing to the axes (for more clarity refer Supplementary Figure 2). All the bioclimatic variables are coded as numbers