Data from: Integrative testing of how environments from the past to the present shape genetic structure across landscapes

Tests of the genetic structure of empirical populations typically focus on the correlative relationships between population connectivity and geographic and/or environmental factors in landscape genetics. However, such tests may overlook or misidentify the impact of such factors on genetic structure, especially when connectivity patterns differ between past and present populations because of shifting environmental conditions over time. Here we account for the underlying demographic component of population connectivity associated with a temporarily dynamic landscape in tests of the factors structuring population genetic variation in an Australian lizard, Lerista lineopunctulata, from 24 nuclear loci. Correlative tests didn't support significant effect from factors associated with a static contemporary landscape. However, spatially explicit demographic modeling of genetic differentiation shows that changes in environmental conditions (as estimated from paleoclimatic data) and corresponding distributional shifts from the past to present landscape significantly structures genetic variation. Results from model-based inference (i.e., from an integrative modeling approach that generates spatially explicit expectations that are tested with Approximate Bayesian Computation) contrasts with those from correlative analyses, highlighting the importance of expanding the landscape genetic perspective to tests the links between pattern and process, revealing how factors shape patterns of genetic variation within species.

景观遗传学（landscape genetics）领域中，针对自然种群遗传结构的常规检验研究，通常聚焦于种群连通性与地理及/或环境因子之间的相关关系。然而，此类检验可能会忽视甚至误判这些因子对遗传结构的影响，尤其是当种群连通性模式因环境条件随时间动态变化，而在历史种群与现存种群间出现差异时。本研究针对澳大利亚蜥蜴Lerista lineopunctulata的24个核基因座（nuclear loci），开展了塑造其种群遗传变异的因子的检验，并在检验中纳入了与随时间动态变化的景观相关联的种群连通性潜在种群统计组分。相关检验结果显示，与静态现存景观相关的因子并未对该物种的遗传变异产生显著影响。然而，针对遗传分化的空间显式种群统计模型分析表明，环境条件变化（基于古气候数据估算得到）以及由此引发的从历史到现存景观的分布范围偏移，显著构建了种群的遗传变异结构。基于模型的推断（即通过生成空间显式预期结果，并采用近似贝叶斯计算（Approximate Bayesian Computation）进行检验的整合建模方法）所得结果，与相关分析的结果相悖。这一结果凸显了拓展景观遗传学研究视角的重要性：通过检验模式与过程之间的关联，揭示因子如何塑造物种种群内的遗传变异模式。