Co-evolutionary relationships associated with biogeographical context mediate the response of native prey to introduced predators, but this effect has not yet been demonstrated for domestic cats. We investigated the main factors influencing the vulnerability of prey species to domestic cat Felis catus predation across Australia, Europe and North America, where domestic cats are introduced. In addition to prey data from empirical records, we used machine-learning models to compensate for unobserved prey in the diet of cats. We found continent-specific patterns of predation: birds were more frequently depredated by cats in Europe and North America, while mammals were favoured in Australia. Bird prey traits were consistent across continents, but those of mammalian prey diverged, notably in Australia. Differences between prey and non-prey species included mass, distribution, and reproductive traits, except in Australian mammals where there was no evidence for a relationship between mass and the probability of being prey. Many Australian mammal prey also have a high extinction risk, emphasizing their vulnerability compared to European and North American counterparts. Our findings highlight the role of eco-evolutionary context in assessing predation impacts and also demonstrate the potential for machine learning to identify at-risk species, thereby aiding global conservation efforts to reduce the negative impacts of introduced predators.

中文翻译：

---

各大洲之间自由放养猫科动物的捕食模式差异


与生物地理环境相关的协同进化关系介导本地猎物对引入的捕食者的反应，但这种影响尚未在家猫中得到证明。我们调查了影响猎物物种对家猫 Felis catus 捕食的脆弱性的主要因素，这些因素分布在引入家猫的澳大利亚、欧洲和北美。除了来自经验记录的猎物数据外，我们还使用机器学习模型来补偿猫饮食中未观察到的猎物。我们发现了特定于大陆的捕食模式：在欧洲和北美，鸟类更频繁地被猫捕食，而哺乳动物在澳大利亚更受欢迎。鸟类猎物特征在各大洲之间是一致的，但哺乳动物猎物的猎物特征存在差异，尤其是在澳大利亚。猎物和非猎物物种之间的差异包括质量、分布和生殖特征，但在澳大利亚哺乳动物中除外，没有证据表明质量与成为猎物的可能性之间存在关系。许多澳大利亚哺乳动物猎物也有很高的灭绝风险，与欧洲和北美的猎物相比，它们更加脆弱。我们的研究结果强调了生态进化背景在评估捕食影响中的作用，并展示了机器学习识别高危物种的潜力，从而有助于全球保护工作减少引入捕食者的负面影响。