Global synthesis of the ecological impacts of invasive non-native vertebrates

Abstract

Human-mediated redistribution of organisms has reshaped global ecosystems, leading to the increasing homogenization of biotic communities. Among invasive taxa, vertebrates, including mammals, birds, reptiles, amphibians, and fishes,are recognised as among the most harmful groups due to their high establishment rates and strong top-down pressures. Despite their known consequences, a systematic global synthesis explicitly linking these impacts to specific mechanisms across all vertebrate classes has been lacking. We addressed this gap by leveraging InvaPact v1.0, a global database containing over 99,000 impact records extracted via fine-tuned large language models. Ecological impacts were quantified using an expanded Environmental Impact Classification for Alien Taxa (EICAT) framework, which assigns impact scores based on the severity and level of biological organisation affected. We analysed the spatial and temporal distribution of these impacts, compared vertebrate impacts to those of invertebrates, and assessed taxonomic representativeness. Our analysis reveals that ecological harm is concentrated within a limited subset of taxa. The most impactful species identified include wild boars ( Sus scrofa ), domestic cats ( Felis catus ), rabbits ( Oryctolagus cuniculus ), and black rats ( Rattus rattus ). Predation emerged as the dominant mechanism of impact, particularly for Mammalia and Actinopterygii, primarily driving declines in native population sizes and changes in assemblage structure. Spatially, the highest impact magnitudes were recorded in Australia and the United States, with significantly higher impacts observed in island ecosystems compared to mainland ecosystems. Furthermore, we found that vertebrates generally have a greater impact across all habitat types than invertebrates. These findings highlight that invasive vertebrate impacts are uneven across taxa and space, often driven by a few high-impact "worst" invaders. The high vulnerability of island biota to predation-mediated extinctions underscores the need for prioritised management and eradication programs in these regions. Additionally, our representativeness analysis suggests research biases toward charismatic or well-known invaders, such as mammals, which may influence global reporting patterns.