Where we've been and where we're going: the importance of source communities in predicting establishment success from phylogenetic relationships

The last two decades have seen growing use of phylogenetic patterns to test hypotheses predicting the success of introduced species. Nearly all of these tests have focused on hypotheses pertaining to phylogenetic relatedness between introduced species and those of the recipient community, largely neglecting hypotheses regarding phylogenetic relationships in the source region. We synthesize hypotheses regarding how phylogenetic relationships of both recipient and source regions together influence establishment success. We also detail how best to account for differences in source communities within phylogenetic frameworks of invasion. Existing studies have predominantly focused on the environmental filtering and competition-relatedness hypotheses, which deal with relatedness to the recipient community. We discuss how these recipient-region hypotheses can be integrated with three hypotheses focused on the relatedness between an introduced species and the source community in which it originated: the evolutionary imbalance, universal tradeoff and competitive constraint hypotheses. We detail important issues that arise when testing alternative hypotheses and interpreting results. We highlight a lack of tests of synthetic phylogenetic hypotheses including both the source and recipient community phylogenetic structure, as well as important covariates such as propagule pressure. Such synthetic tests may be valuable for identifying general phylogenetic patterns in establishment success, predicting future invasions, and for stimulating further exploration of the underlying mechanisms of invasibility. We conclude with recommendations for future studies that use phylogenetic relationships to predict invasions: including source and recipient communities, using complete phylogenies and accounting for phylogenetic uncertainty, considering multiple stages of invasion and conducting analyses across spatial and phylogenetic scales where possible.