American beaver (Castor canadensis) have been translocated for population restoration, reduction of human-wildlife conflict, and enhancement of ecosystem function. Yet few studies have assessed dispersal of beaver, making it difficult to determine at what scale translocations are appropriate. Genetic studies can provide inferences about gene flow, and thus dispersal. We used a landscape genetic approach to evaluate whether landscape features influenced gene flow among beaver in the Coast Range of western Oregon, USA, using samples collected April–September 2014. We collected genetic samples from live-captured (n = 232), road-killed (n = 2) and trapper-provided (n = 58) tissue samples and genotyped them at 10 microsatellite loci. We mapped records of beaver translocations into or within the study area during the twentieth century to consider the effect of those movements on genetic structure. We used population assignment tests to delimit genetic clusters, evaluated correspondence of those clusters with watershed boundaries and translocation history, and then estimated differentiation between clusters and between watersheds using model-based and model-free approaches. We evaluated how individual genetic differences varied with geographic distance, and investigated related pairs within clusters. We developed landscape resistance models incorporating slope, distance to water, and watershed boundaries at 2 scales, and estimated effective distances between sample locations with least cost path and circuit theoretic analyses. We evaluated the correlation of individual genetic distances with effective distances using a pseudo-bootstrapping approach. Landscape genetic models did not explain spatial variation in genetic structure better than geographic distance, but hierarchical genetic structure corresponded with watershed boundaries and suggested influences from historical translocations. Pairwise individual genetic distances were positively correlated with geographic distances to 61 km; highly-related pairs mostly were detected <1 km apart (median = 1.0 km,  = 14.6 ± 2.3 [SE] km, n = 77). We concluded that slope and distance to water did not strongly limit dispersal and gene flow by beaver in this system, but concordance of genetic structure with watershed boundaries suggests that dispersal is more common within than between watersheds. Genetic differentiation of beaver within this topographically complex system was much greater than reported in a study at similar spatial scales in relatively flat topography. We recommend that translocation efforts of American beaver in topographically complex landscapes occur within watersheds when possible but conclude that dispersal can occur across watersheds. © 2021 The Wildlife Society. This article is a U.S. Government work and is in the public domain in the USA.

中文翻译：

---

俄勒冈州沿海美国海狸的景观遗传学

美洲海狸（Castor canadensis）已被迁移以恢复种群数量、减少人与野生动物之间的冲突并增强生态系统功能。然而，很少有研究评估海狸的扩散，因此很难确定什么规模的易位是合适的。遗传研究可以提供关于基因流动的推论，从而提供散布的推论。我们使用景观遗传方法来评估景观特征是否影响美国俄勒冈州西部海岸山脉海狸的基因流，使用 2014 年 4 月至 9 月收集的样本。我们从现场捕获的 ( n  = 232)、路被杀死 ( n  = 2) 和捕手提供 ( n = 58) 组织样本并在 10 个微卫星位点对它们进行基因分型。我们绘制了 20 世纪研究区域内或研究区域内的海狸易位记录，以考虑这些运动对遗传结构的影响。我们使用种群分配测试来划分遗传簇，评估这些簇与流域边界和易位历史的对应关系，然后使用基于模型和无模型的方法估计簇之间和流域之间的差异。我们评估了个体遗传差异如何随地理距离变化，并调查了集群内的相关对。我们开发了包含坡度、到水的距离和流域边界的 2 个尺度的景观阻力模型，并以最小成本路径和电路理论分析估计样本位置之间的有效距离。我们使用伪引导方法评估了个体遗传距离与有效距离的相关性。景观遗传模型并不能比地理距离更好地解释遗传结构的空间变异，但层次遗传结构与流域边界相对应，并暗示了历史易位的影响。成对个体遗传距离与61公里的地理距离呈正相关；高度相关的对大多被检测到相距 <1 公里（中位数 = 1.0 公里，景观遗传模型并不能比地理距离更好地解释遗传结构的空间变异，但层次遗传结构与流域边界相对应，并暗示了历史易位的影响。成对个体遗传距离与61公里的地理距离呈正相关；高度相关的对大多被检测到相距 <1 公里（中位数 = 1.0 公里，景观遗传模型并不能比地理距离更好地解释遗传结构的空间变异，但层次遗传结构与流域边界相对应，并暗示了历史易位的影响。成对个体遗传距离与61公里的地理距离呈正相关；高度相关的对大多被检测到相距 <1 公里（中位数 = 1.0 公里， = 14.6 ± 2.3 [SE] 公里，n  = 77）。我们得出结论，坡度和与水的距离并没有强烈限制海狸在该系统中的扩散和基因流动，但遗传结构与流域边界的一致性表明，流域内的扩散比流域之间更常见。在这个地形复杂的系统中，海狸的遗传分化比在相对平坦的地形中类似空间尺度的研究中报告的要大得多。我们建议美国海狸在地形复杂的景观中的易位努力在可能的情况下发生在流域内，但得出的结论是，可以跨流域传播。© 2021 野生动物协会。这篇文章是美国政府的作品，在美国属于公共领域。