Abstract
The initial colonization by Sinogastromyzon puliensis is hypothesized to be in southern Taiwan, followed by its northward dispersal. To unravel its phylogeographical history, the mitochondrial control regions of S. puliensis were sequenced. Hierarchical analysis of molecular variance (AMOVA) revealed that more genetic variation resided among populations (Φ ST = 0.879, p = 0) than within populations (Φ SC = 0.871, p = 0), as supported by significant pairwise F ST values showing high genetic differentiation and rare migrations among populations. High levels of genetic differentiation can be explained by either long-term isolation among populations or a consequence due to the founder effect. Both a Neighbor-joining (NJ) tree and maximum parsimony (MP) analysis divided 4 populations of S. puliensis into 3 major groups. Group I is composed of populations of the Wu and Choshui Rivers, in which reciprocal monophyly was not supported. The Tzengwen River, comprising group II, is sister to group I. Group III is the basal clade and consisted of only the Kaoping River in a phylogram rooted by S. wui. According to the topological pattern and character of haplotypes without a 16-bp deletion, the Kaoping River is proposed to be the first habitat colonized by the ancestor of S. puliensis, after which this fish dispersed northwards. This dispersal pattern is concordant with the previously proposed hypothesis.