Data from: Counting cats: spatially explicit population estimates of cheetah (Acinonyx jubatus) using unstructured sampling data

Many ecological theories and species conservation programmes rely on accurate estimates of population density. Accurate density estimation, especially for species facing rapid declines, requires the application of rigorous field and analytical methods. However, obtaining accurate density estimates of carnivores can be challenging as carnivores naturally exist at relatively low densities and are often elusive and wide-ranging. In this study, we employ an unstructured spatial sampling field design along with a Bayesian sex-specific spatially explicit capture-recapture (SECR) analysis, to provide the first rigorous population density estimates of cheetahs (Acinonyx jubatus) in the Maasai Mara, Kenya. We estimate adult cheetah density to be between 1.22 ± 0.301 and 1.28 ± 0.322 individuals/100km2 across four candidate models specified in our analysis. Our spatially explicit approach revealed 'hotspots' of cheetah density, highlighting that cheetah are distributed heterogeneously across the landscape. The SECR models incorporated a movement range parameter which indicated that male cheetah moved four times as much as females, possibly because female movement was restricted by their reproductive status and/or the spatial distribution of prey. We show that SECR can be used for spatially unstructured data to successfully characterise the spatial distribution of a low density species and also estimate population density when sample size is small. Our sampling and modelling framework will help determine spatial and temporal variation in cheetah densities, providing a foundation for their conservation and management. Based on our results we encourage other researchers to adopt a similar approach in estimating densities of individually recognisable species.

诸多生态学理论与物种保护计划均以精准的种群密度估算为依据。针对种群正快速衰退的物种，精准的密度估算需采用严谨的野外调查与分析方法。然而，获取食肉动物的精准密度估算结果往往颇具挑战——这类动物天然种群密度较低，且通常行踪隐秘、活动范围广阔。本研究采用非结构化空间采样野外设计方案，结合基于性别的贝叶斯空间显式捕获-再捕获（spatially explicit capture-recapture, SECR）分析，首次对肯尼亚马赛马拉地区的猎豹（Acinonyx jubatus）开展了严谨的种群密度估算。经对本研究设定的四组候选模型分析，我们估算得到成年猎豹的种群密度介于1.22±0.301至1.28±0.322只/100km²之间。本研究采用的空间显式方法揭示了猎豹密度的‘热点区域’，表明猎豹在研究区域内的分布具有空间异质性。本次使用的SECR模型纳入了活动范围参数，结果显示雄性猎豹的活动范围为雌性的四倍，这一现象可能源于雌性的活动受到繁殖状态以及/或猎物空间分布的限制。本研究证实，SECR模型可应用于非结构化空间数据，成功刻画低密度物种种群的空间分布特征，且在样本量较小时仍可实现种群密度估算。我们的采样与建模框架将有助于解析猎豹密度的时空变化规律，为其保护与管理提供科学依据。基于本研究结果，我们呼吁其他研究者在估算可个体识别物种的种群密度时采用类似方法。