Data from: Model-based comparisons of phylogeographic scenarios resolve the intraspecific divergence of cactophilic Drosophila mojavensis

The cactophilic fly Drosophila mojavensis exhibits considerable intraspecific genetic structure across allopatric geographic regions and shows associations with different host cactus species across its range. The divergence between these populations has been studied for more than 60 years, yet their exact historical relationships have not been resolved. We analysed sequence data from 15 intronic X-linked loci across populations from Baja California, mainland Sonora-Arizona and Mojave Desert regions under an isolation-with-migration model to assess multiple scenarios of divergence. We also compared the results with a pre-existing sequence dataset of 8 autosomal loci. We derived a population tree with Baja California placed at its base and link their isolation to Pleistocene climatic oscillations. Our estimates suggest the Baja California population diverged from an ancestral Mojave Desert/mainland Sonora-Arizona group around 230-270 Kya, while the split between the Mojave Desert and mainland Sonora-Arizona populations occurred one glacial cycle later, 117-135 Kya years ago. Although we found these three populations to be effectively allopatric, model ranking could not rule out the possibility of a low-level of gene flow between two of them. Finally, the Mojave Desert population showed a small effective population size, consistent with a historical population bottleneck. We show that model-based inference from multiple loci can provide accurate information on the historical relationships of closely related groups allowing us to set into historical context a classic system of incipient ecological speciation.嗜仙人掌果蝇莫哈韦果蝇（Drosophila mojavensis）在各异域地理区域中均展现出显著的种内遗传结构，且在其分布范围内与多种宿主仙人掌物种形成特异性关联。该类群种群间的分化研究已有六十余年历史，但其确切的演化历史关系仍未得到明确解析。我们基于隔离与迁移模型（isolation-with-migration model），对下加利福尼亚、索诺拉-亚利桑那大陆区域以及莫哈韦沙漠种群的15个X连锁内含子位点的序列数据开展分析，以评估多种分化情景。同时，我们将分析结果与已公开的8个常染色体位点序列数据集进行了比对。我们构建了以基部为下加利福尼亚种群的种群系统发育树，并将该种群的分化与更新世（Pleistocene）气候震荡相关联。我们的估算结果显示，下加利福尼亚种群约在23万至27万年前（Kya）从莫哈韦沙漠/索诺拉-亚利桑那大陆祖先类群中分化而来；而莫哈韦沙漠与索诺拉-亚利桑那大陆种群的分化则晚了一个冰期周期，发生在11.7万至13.5万年前。尽管我们发现这三个种群实际处于异域分布状态，但模型排序结果无法排除其中两个种群间存在低水平基因交流的可能性。最后，莫哈韦沙漠种群的有效种群规模（effective population size）较小，这与历史上的种群瓶颈效应（population bottleneck）相符。本研究证实，基于多位点的模型推断可精准获取近缘类群的历史演化关系信息，从而能够将这一经典的初期生态物种形成系统纳入历史演化背景之中。