Data from: Mesopredator release among invasive predators: controlling red foxes can increase feral cat density and alter their behaviour

The mesopredator release theory predicts that the density of subordinate predators will increase as dominant predators decline. Persistent debate around mesopredator release in part reflects the lack of robust, replicated experiments to test this theory, and the use of population indices which confound changes in mesopredator density and detectability. This uncertainty has immediate impacts for conservationists who are faced with managing sympatric invasive predators. We used replicated experimental designs and spatially-explicit models to examine whether mesopredator release of the feral cat Felis catus occurs in response to targeted control of the introduced red fox Vulpes vulpes. We surveyed three Control-Impact paired landscapes in a region with long-term fox control (1080 poison baiting), and conducted a Before-After Control-Impact Paired-Series experiment in another region. We used fox occurrence as a simple metric of fox populations and estimated feral cat density with spatial mark-resight models. Lethal fox control had varying effects on fox occurrence, consistent with variation in the duration and intensity of poison baiting. Correspondingly, responses in feral cat density ranged from negligible to a 3.7-fold higher density in fox-baited landscapes. At a fine spatial scale (200 m2), feral cat density was negatively associated with fox occurrence probability across both regions. These results were consistent with mesopredator release, although uncertainty was high in the region where fox control had only recently commenced. Feral cat detectability also varied across the (artificially-manipulated) gradients of fox occurrence probability. In one region, nonlinear models indicated that feral cats had lower detection and increased movement rates when foxes were uncommon, giving way to density suppression at high fox occurrence probabilities. Synthesis and applications. Our study provides replicated, experimental evidence that dominant predator suppression can be associated with a higher mesopredator density. Mesopredator release can manifest as changes in both behaviour and density, distorting inference if these processes are not distinguished. Our results may help explain why fox control does not consistently improve native prey persistence, suggesting integrated pest management may be necessary to improve conservation outcomes.

中级捕食者释放理论（mesopredator release theory）预测，随着顶级捕食者种群下降，次级捕食者的种群密度会随之上升。当前围绕中级捕食者释放的长期争议，在一定程度上源于缺乏可靠的重复实验以验证该理论，同时也与研究中使用了会混淆中级捕食者密度与可检测性变化的种群指数有关。这种不确定性对面临同域入侵捕食者管理工作的保护工作者造成了直接影响。
我们采用重复实验设计与空间显式模型（spatially-explicit models），探究针对引入种赤狐（Vulpes vulpes）的针对性防控是否会引发野化家猫（Felis catus）的中级捕食者释放效应。我们在长期开展赤狐防控（1080毒饵投放）的区域内，对3处对照-影响配对景观开展了调查，并在另一区域开展了前后对照影响配对序列实验。我们以赤狐出现率作为赤狐种群的简易衡量指标，并通过空间标记重捕模型（spatial mark-resight models）估算野化家猫的种群密度。
致死性赤狐防控对赤狐出现率的影响存在差异，这与毒饵投放的持续时长与强度差异相一致。相应地，野化家猫种群密度的响应幅度从可忽略不计到赤狐防控景观内密度提升3.7倍不等。在200平方米的精细空间尺度下，两个区域内的野化家猫种群密度均与赤狐出现概率呈负相关关系。尽管在近期才开始开展赤狐防控的区域内不确定性较高，但上述结果与中级捕食者释放效应的理论预期相符。野化家猫的可检测性也随赤狐出现概率的（人工调控）梯度发生变化：在其中一个区域，非线性模型显示，当赤狐较为稀少时，野化家猫的检测率更低、移动速率更高，而当赤狐出现概率较高时，则会出现野化家猫密度受抑制的情况。
综合与应用。本研究提供了重复实验证据，证明顶级捕食者的抑制作用可伴随中级捕食者密度升高。中级捕食者释放效应可同时体现在行为与种群密度的变化中，若无法区分这两类过程，会导致推论出现偏差。我们的研究结果或有助于解释为何赤狐防控无法持续提升本土猎物的存续能力，这提示采用综合虫害管理或许能改善保护成效。