Darwin's hypotheses to explain colonization trends: evidence from a <i>quasi</i>‐natural experiment and a new conceptual model

Abstract Aim We examined fish establishment trends in the aftermath of the connection of two Neotropical freshwater ecoregions. Darwin's conundrum was used to examine whether taxonomic and functional relatedness would explain establishment or whether the establishment of new populations was driven by ecological interactions ( Darwin's naturalization hypothesis , DNH ) or by shared environmental requirements ( adaptation hypothesis , AH ). Location Paraná River, south‐eastern South America. Methods Taxonomic relatedness was investigated using a one‐tailed Fisher exact test, considering all species and monophyletic groups. Functional traits were used to estimate niche dissimilarities among non‐native species and across the entire set of native species, generating distributions that were fitted to exponential or logarithmic curves, based on a new conceptual model. Results Of the 213 species from the Lower Paraná River that had the opportunity to reach the Upper Paraná River, 25% – 53 spp. – were successfully established. The ratios of unshared to shared genera for both successful and unsuccessful non‐native species were not different, but 7% of type I error probability for Characiformes and 8% for all species, would favour DNH. Our results indicate that both DNH and AH could explain successes but that DNH was more likely to explain failures in establishment. Main conclusion Darwin's naturalization hypothesis is more likely to explain the non‐native establishment in invasion processes. Both the DNH and AH explain successful colonization, but DNH provided the best explanation for failed non‐native establishments. Non‐native species that were functionally similar to native species (i.e. with similar traits) were more likely to fail, suggesting that biotic resistance was directly linked to ecological similarity. Biotic resistance could be the most important mechanism for explaining establishment or failure of non‐native species. The conceptual model proposed here can be used to predict which non‐native species are least likely to establish, based on the functional traits of species in the native community.