Abstract Trophic cascade theories such as the ‘behaviourally-mediated trophic cascade hypothesis’ (BMTCH), have mainstreamed as ecological tools for conserving biodiversity and restoring ecosystems. The BMTCH relies on indirect negative effects of large carnivores through suppression of mesocarnivore activity and habitat use. Importantly, effects of top carnivores on mesocarnivores varies over time and space, is dependent on the species involved, and local context. In South Africa, there are very few free-ranging carnivores, as populations are often restricted to enclosed reserves. While predator-proof fences reduce human-wildlife conflict, they also influence space use within communities. We used an enclosed reserve with a relatively full complement of carnivores to test the generality of the BMTCH in the African context. Using single-species, multi-season occupancy models we investigated the spatial dynamics of multiple carnivores. We also investigated spatial partitioning by vegetation type and temporal partitioning. Our results revealed both support for and against the BMTCH. Lions and spotted hyaenas negatively influenced the detection probability of black-backed jackals and African wildcats, while leopards had a positive effect on these two mesocarnivore species. Additionally, lions positively influenced the detection probability of side-striped jackals. Although space use of carnivores in relation to vegetation type showed minimal evidence of spatial partitioning, each carnivore had a unique combination of abiotic and biotic factors influencing their spatial dynamics, which could facilitate co-existence. Temporal partitioning may also be promoting co-existence as activity patterns of smaller carnivores overlapped the least with sympatric carnivores, particularly lions. Extensive activity overlap between large carnivores does not promote co-existence. We suggest that applying the BMTCH as a universal law across ecosystems is unsupported and may lead to inappropriate conservation and management actions, and prevent protection of ecosystems.

中文翻译：

---

狮子不会改变河流：复杂的非洲大草原排除了大型食肉动物自上而下的强迫

摘要 营养级联理论，如“行为介导的营养级联假说”（BMTCH），已成为保护生物多样性和恢复生态系统的生态工具的主流。BMTCH 依赖于大型食肉动物通过抑制中型食肉动物活动和栖息地使用的间接负面影响。重要的是，顶级食肉动物对中型食肉动物的影响随时间和空间变化，取决于所涉及的物种和当地环境。在南非，自由放养的食肉动物很少，因为种群通常被限制在封闭的保护区内。虽然防捕食者围栏减少了人类与野生动物的冲突，但它们也会影响社区内的空间使用。我们使用了一个带有相对完整的食肉动物的封闭保护区来测试 BMTCH 在非洲背景下的普遍性。使用单一物种，在多季节占用模型中，我们研究了多种食肉动物的空间动态。我们还研究了植被类型和时间分区的空间分区。我们的结果显示支持和反对 BMTCH。狮子和斑点鬣狗对黑背豺和非洲野猫的检测概率产生负面影响，而豹子对这两种中型食肉动物产生积极影响。此外，狮子对侧条纹豺的检测概率有积极影响。尽管与植被类型相关的食肉动物的空间利用表明空间划分的证据很少，但每个食肉动物都有影响其空间动态的非生物和生物因素的独特组合，这可以促进共存。时间分区也可能促进共存，因为小型食肉动物的活动模式与同域食肉动物（尤其是狮子）的重叠最少。大型食肉动物之间广泛的活动重叠不会促进共存。我们建议将 BMTCH 作为跨生态系统的普遍规律应用是不受支持的，并且可能会导致不适当的保护和管理行动，并阻碍对生态系统的保护。