The spatiotemporal response of mammal communities to pulse-driven resource availability in the Kalahari data

The objective of this study was to detect pulse events at a forage patch scale and to identify the subsequent responses from various mammal functional groups of mammals in the arid Kalahari Desert of South Africa. The aim was to, therefore, understand how the Kalahari functions on a finer scale, particularly with increased rainfall variability in the drier years. We predicted the green-up (pulse event) to be synchronized over large areas with extreme rainfall, while we expected low synchronicity and very patch green-up in dry years. As ephemeral plants respond to rain showers within 2-3 weeks, we predicted the following responses from the animal functional groups would be detectable within ~8 weeks after the green-up: 1. There would be an increase in abundance of large antelopes. 2. There would not be an increase in abundance of small antelopes after the pulse event. 3. There would not be an increase in abundance of carnivores after the pulse event. 4. There would be an increase in the meso-carnivore abundance. 5. There would be an increase in rodent and lagomorph abundance. 6. As our cameras are in the dunes, we predict that during a period of normal rainfall, there would be a decrease in observations as animals may move to more attractive areas. We gathered the data from 42 cameras located in Tswalu Kalahari Reserve (n=18), Kgalagadi Transfrontier Park (n=12) and Khomani San Community Land (n=12). We used greenness index values obtained from the phenopix package in R, and subsequently used relative abundance index to compare functional group abundance 4 weeks before the pulse event and 8 weeks after the pulse event, to determine if there was a response. We used two survey seasons, October to May in 2018-2019 and 2019-2020. In this study we developed a novel way to detect pulse events in arid systems using bycatch from camera trap data. We found clear patterns of green-up and a clear difference in the green-up/resource availability between season 18-19 and 19-20. We also found some differences in the mammals RAI between season 18-19 and 19-20. We were able to link the pulse events to the short-term response of large antelopes, in Kgalagadi and the Khomani San. We did, however, not see any other responses to the green-up, which with the limitations of our data suggest that the responses are likely to be longer term or occurring at a larger spatial scale. This has allowed for a deeper understanding of animal responses in unpredictable arid environments. Climate change threatens systems such as the Kalahari with rising air temperatures and further precipitation unpredictability. Droughts are to increase in magnitude and frequency. With this increase in system variability, it is crucial that arid landscapes are managed effectively, both short-term and long-term. To be able to manage the system, it is of utmost importance to further our knowledge of the Kalahari system. This study has provided a starting point to do so.本研究旨在以草料斑块为尺度检测脉冲事件（pulse event），并识别南非干旱喀拉哈里沙漠中不同哺乳动物功能群的后续响应。本研究旨在更精细地解析喀拉哈里生态系统的运作机制，尤其是在干旱年份降雨变异性加剧的背景下。我们假设，极端降雨事件可驱动大范围区域同步出现植被返青（即脉冲事件），而在干旱年份，植被返青的同步性较低且呈现显著斑块化特征。由于短命植物会在降雨后2-3周内响应，我们预测在植被返青后约8周内，可观测到以下哺乳动物功能群的响应： 1. 大型羚羊的丰度将出现上升。 2. 脉冲事件发生后，小型羚羊的丰度不会出现上升。 3. 脉冲事件发生后，食肉动物的丰度不会出现上升。 4. 中型食肉动物的丰度将出现上升。 5. 啮齿类与兔形目动物的丰度将出现上升。 6. 由于本研究布设的相机位于沙丘区域，我们预测在正常降雨时段，动物可能迁移至更具资源吸引力的区域，导致相机观测记录量下降。 本研究的数据来自分布于茨瓦鲁喀拉哈里保护区（n=18）、卡拉哈里跨境公园（n=12）及科曼尼桑族社区土地（n=12）的42台红外相机。我们采用R语言phenopix包获取的绿度指数值，并通过相对丰度指数（Relative Abundance Index, RAI）对比脉冲事件发生前4周与发生后8周的功能群丰度，以验证是否存在响应。调研时段为2018-2019年及2019-2020年的10月至次年5月。 本研究开发了一种利用红外相机陷阱附带监测数据识别干旱生态系统脉冲事件的全新方法。研究清晰观测到植被返青模式，并发现2018-2019与2019-2020两个调研季的植被返青/资源可获得性存在显著差异。同时，两个调研季的哺乳动物相对丰度指数也存在一定差异。我们成功将卡拉哈里跨境公园与科曼尼桑族社区区域内的脉冲事件与大型羚羊的短期响应建立关联。但未观测到其他哺乳动物功能群对植被返青的响应，结合本研究的数据局限性，这提示相关响应可能属于长期效应，或发生在更大的空间尺度上。本研究加深了我们对不可预测的干旱环境中动物响应机制的理解。 气候变化正通过气温上升与降水不确定性加剧威胁喀拉哈里等干旱生态系统，干旱事件的强度与发生频率亦将提升。随着生态系统变异性加剧，对干旱景观的短期与长期有效管理至关重要。而要实现有效管理，深化对喀拉哈里生态系统的认知是重中之重。本研究为该领域的后续研究提供了重要起点。