The Himalayan uplift and evolution of aquatic biodiversity across Asia: Snowtrout (Cyprininae: Schizothorax) as a test case

also among Schizothorax represented in Bhutan (S1 Appendix). Again, the BA-tree differs in placement of the YLTR and Bhutan-1 lineages, with the YLTR-West-A and YLTR West-B groups separated by the Bhutan-1A group, whereas Bhutan-1B and Bhutan-1C groups collapse and YLTR-East placing as sister to it (S1 Fig). Eastern QTP and Southeast Asia The E-QTP/ SEA-A species split into three significantly different clades (Fig 4, S3 Table), encompassing five species from the Irrawaddy/ Salween rivers and reflecting the deepest TIME- TREE estimate at 5.6 Ma (CI = 9.7–3.0). The E-QTP/ SEA-B diverged at 1.7 Ma (CI = 3.5–0.8) and consists of seven species from the Yangtze River, whereas the E-QTP/ SEA-C involves seven species from the Mekong/ Salween drainages. The Mekong divergence was seemingly driven by late Pliocene orogenesis [77], an estimate congruent with our timeline (2.4 Ma; CI = 5.0–1.2). Three species within the upper Salween (i.e., S. gongshanensis, S. lissolabiatus, S. nukiangensis) display low genetic divergence and frequent haplotype sharing, a potential by- product of Quaternary glaciations [79]. However, several E-QTP/ SEA-B species (such as S. davidi, S. koslovi, S. chongli, S. gongshanensis) have yet to be fully described or documented, other than in general terms [80]. Dispersal into Bhutan, Nepal, and western Himalaya The YLTR originates on the QTP, flows eastward then abruptly south through the Namche Barwa Syntaxis [5], with subsequent capture by the Brahmaputra River [6]. It represents an

important freshwater dispersal route from the QTP and its drainage history correlates with an hypothesized evolutionary transition between species on the QTP (sensu stricto) versus more recent divergences via the Lower Brahmaputra. In our analyses, samples from Nepal split into distinct clades, each representing separate drainages that are significantly different from Southeast Asian clades. Interestingly, a set of samples from Bhutan (Bhutan-2; N = 8) form a distinct clade sister to the easternmost Nepali drainage (Koshi River Basin; N = 10, S3 Table). This is consistent with an hypothesized east- to-west dispersal via the lower Brahmaputra. Species in the Koshi River group are tentatively referred to as S. richardsonii, S. esocinus, and S. progastus, whereas samples from Bhutan-2 are currently unidentified but link in the network with a haplotype of S. progastus from the Koshi River (Fig 5). The morphological similarities between S. richardsonii and S. progastus in the Koshi, Kar- nali, and Gandaki rivers are such that their discrimination can be problematic in the field [83], particularly at early life history stages. Despite morphological differences [29], the two species are genetically more similar within than among basins, which could be driven by ecological diversification along elevational gradients, allowing for emergence of similar phenotypes in each basin (i.e., homoplasy). This perspective stems from mtDNA data, but a recent study using ddRADseq indicated substantial admixture between the two in each drainage [84]. Our

study does not (and cannot) revise taxonomy, but it does identify regions (e.g., Nepal, Bhutan) where more focused research is needed to assess genetic and phenotypic diversity, and thus clarify taxonomy of Schizothorax in the Central and Eastern Himalaya. Schizothorax in the Western Himalaya are genetically less well defined and their clustering reflects a mosaic of historic and contemporary drainage arrangements (S1 Appendix). Samples from the Gandaki and the Karnali rivers cluster into distinct groups, with the Gandaki River haplotypes (N = 20) most distinct (separated by 26 mutations, Fig 5) and partitioned into nested clusters of S. richardsonii (three lineages), S. progastus (two lineages), and one repre- senting a mixture of the two (Fig 4). Samples from the Karnali River Basin (far western Nepal, PLOS ONE Himalayan uplift and aquatic evolution PLOS ONE | https://doi.org/10.1371/journal.pone.0289736 October 24, 2023 18 / 27