What does render plant communities vulnerable to biological invasions? Testing Darwin's Naturalization Hypothesis in the Canarian pine forest

Abstract

Invasive species have been identified as one of the main threats to biodiversity. Oceanic islands are particularly vulnerable to invasions because their biota are characterized by a suite of syndromes, such as low competitive ability and loss of dispersal capacities, which have evolved in the island setting. C. Darwin (1809-1882) was among the firsts to draw hypotheses about the invasion process. He observed that alien species are more prone to establish in an ecosystem where no close native species are present. Indeed, closely related species would either share the same niche, and hence, have competitive interactions, or the pathogens and/or predators of a native species would be unlikely to shift onto a phylogenetically distant alien species (Enemy Release Hypothesis). Subsequent empirical support for this assumption has been equivocal, generating a conundrum known as Darwin’s Naturalization Conundrum, which has opened a path for research on the mechanisms by which species become successful in a new environment. This study aims at determining whether biological invasions can, in line with Darwin’s hypothesis, be predicted from environmental features and/or characteristics of the native flora. We first determine whether environmental variation has an impact on the naturalization of alien, non-invasive and invasive species, i.e., whether some habitats are more prone to facilitate naturalization than others. We then determine whether the probability of invasion can be predicted from the taxonomic composition and life-history traits of native communities, i.e., whether some communities are, due to their taxonomic composition and/or life-history traits, more prone to biological invasions than others. Community composition and life-history traits of spermatophyte species were recorded from 40 10x10m plots in the pine forest of Tenerife (Canary Islands). In total, 135 species were recorded, including 16 alien species, 8 of which are invasives. The data were analyzed using multivariate statistics and General Linear Models. The relative abundance of alien species in Canarian pine forests was significantly related to environmental variables, supporting mounting evidence that environmental filtering is one of the strongest constraints determining variation in the vulnerability to introduced species of some communities as compared to others. Life-history traits, but not taxonomic composition of native species, were significant predictors of the presence of invasive species. This supports the idea that, while the taxonomic composition of the native community may not necessarily determine the opportunity of alien species to establish, the life-history traits of native species may be more important in the context of the niche pre-emption hypothesis as species sharing the same traits, regardless of their taxonomic identity, may be prone to competitive interactions. Despite retaining fairly high AUC values (>0.70) following cross-validation, the model predicting the presence of invasive species was characterized by high rates of false positives. This means that plots displaying the features that are characteristic for invaded plots, are not invaded yet, suggesting that the alien invasion of the Tenerife pine forest is an ongoing process.