As the number of non-native species introductions continues to increase, the need for tools to predict potential invaders is a central focus in invasion ecology. Trait-based models are a popular method used to predict non-native species success; however, they have many challenges to overcome and often fail to address the role of propagule pressure in failed invasions. Due to a lack of data, many studies cannot incorporate failed invasions into their models examining the success of invaders. Here, we analyzed the relationship between the functional traits of non-native species, both successful and failed, and native species in the Laurentian Great Lakes over three time periods between 1870 and 2010. We only examined failed invasions where there was an environmental match and sufficient propagule pressure for establishment as determined through records of authorized stocking, allowing us to directly test the contribution of traits to invasion success. We evaluated whether nearest neighbor functional distance (NNFD), mean functional distance (MFD), lake, time, and functional traits predicted invasion success and impact using logistic regression models. Our results indicate that species with functional traits associated with r-selected life-history strategies were more likely to establish and predators were more likely to have high impact. While smaller NNFD and MFD values predicted establishment success, higher MFD values predicted high impact. This study examined the role of functional traits in invasion success and demonstrated the importance that a change in community context contributes to regulating invasion success.

随着引进非本地物种的数量不断增加，对用于预测潜在入侵者的工具的需求已成为入侵生态学的重点。基于特征的模型是用于预测非本地物种成功的流行方法。但是，它们有许多挑战要克服，而且往往无法解决传播压力在入侵失败中的作用。由于缺乏数据，许多研究无法将失败的入侵纳入其研究入侵者成功的模型中。在这里，我们分析了劳伦山脉五大湖在1870年至2010年之间的三个时间段内，成功和失败的非本地物种的功能性状与本地物种之间的关系。我们仅检查了失败的入侵情况，这些情况通过授权的放养记录确定具有环境条件和足够的繁殖力来建立，因此我们可以直接测试特征对入侵成功的贡献。我们使用逻辑回归模型评估了最近邻功能距离（NNFD），平均功能距离（MFD），湖泊，时间和功能特征是否预测了入侵成功和影响。我们的结果表明，具有与r-选择的生活史策略相关的功能性状的物种更可能建立，掠食者更有可能产生高影响。较小的NNFD和MFD值可预测建立成功，而较高的MFD值可预测较高的影响。