中国金钱豹（Panthera pardus）的分布、现状与监测策略

金钱豹(Panthera pardus)是世界上分布最广的野生猫科动物，但我们必须看到，它们在亚洲的种群数量已经显著减少。历史上，金钱豹曾广布于除干旱戈壁沙漠和海拔4'000米以上西部山地之外的中国全境。然而，近年的调查表明，金钱豹已经从很多地区消失，远没有人们曾经认为的那么常见。本论文在多重地理尺度下开展研究，旨在深化对中国金钱豹保护生物学的理解。论文采用多种数据源采集金钱豹在全国范围内的分布记录，以评估金钱豹的分布和生存现状。研究结果表明，金钱豹在中国急剧减少，只在11个省的44个位点发现了确凿的分布记录。现存金钱豹种群小(<50个体)且分散，主要存在于相互孤立的自然保护区中，使得它们面临着较高的局地灭绝的风险。值得注意的是，这些中国金钱豹的分布记录大多是在对其他旗舰物种进行调查和研究时所获得的，而很少直接源于关于金钱豹的项目和研究。因此，现有的金钱豹研究没有将那些不与其他旗舰物种重叠，尤其是金钱豹单独适应的栖息地和生态系统包括在内，那主要是一些离人类干扰较近而其他物种已经局地灭绝的区域。山西省就是一个这样的地区，该省位于华北地区黄河流域的中部，华北豹(P.p. japonensis)这一仅分布于中国的金钱豹亚种就生活在这里。在本论文中，我们建立物种分布模型以预测哪些环境条件适宜于金钱豹生存，进而确定哪些区域适于它们生存。首先，我们从2008到2013年间在全省范围内非系统安装的红外相机数据中提取出金钱豹的分布数据。然后，在1km2大小的栅格内选取分别代表地形、植被和人类活动的六个不相关变量，用来分析环境因子对雪豹分布的影响。MAXENT模型的分析结果表明，影响金钱豹分布的最重要的因素是森林覆盖率(贡献率为90%)，道路和人口密度的影响则比较小(分别为3.3%和2.6%)。研究发现，在山西省南部仍有大面积的适宜栖息地，包括了两个国家级保护区。选取其中面积最大的历山国家级自然保护区，作为进行第一次种群密度估算的区域。于是，我们在整个保护区(252 km2)内系统地安装了139台红外相机。通过对照片的个体鉴别，共识别出了5只金钱豹，并采用种群密度估算所能应用的最现代的统计建模技术进行了分析。通过标志重捕(MARK)和空间标志重捕(SPACECAP)模型所得的结果比较相近(分别为0.7和0.6±0.3只金钱豹／100km2)。这一密度相当于在中国东北极度濒危的东北豹(P.p. orientalis)，同时也是世界上金钱豹密度的最低记录之一。论文对应用痕迹调查作为红外相机的替代或补充手段进行种群监测的可行性进行了探讨，但找到的痕迹不多(粪便34个， 足迹3个， 抓痕5个， 毛发2处)。实际上，除非通过DNA分析验证，否则无法确认野外粪便样本所属的物种。研究发现，在所有判别为目标物种的粪便中，只有14%真正来自金钱豹，其它大多数都是犬科动物的粪便。因为豹的样本数量小，无法进行近一步的探究，所以我们应用雪豹(Panthera uncia)景观中的食肉动物群落进行分析判别，并得出了相似的结果：超过一半在野外被认为是雪豹的粪便实际上来自赤狐(Vulpes vulpes)。由此可见，因为错误识别率如此之高，依靠粪便样本所进行的物种调查和监测都应使用分子粪便学技术完成物种鉴定。同时，痕迹的低遇见率(0.1痕迹／km)提示我们，除非提高调查强度(加长样线距离)和／或增加发现率(在有积雪时进行调查)，否则无法应用痕迹调查达到预期效果。必须看到，我们对中国金钱豹的知识缺口依然很大，这种信息的缺失会降低保护工作的有效性。我们所进行的系统性调查适用于金钱豹这样难以发现的物种的监测和研究，即便在种群密度很低的情况下也依然适用。另一方面，痕迹调查无法提供分析所需的足够多的样本量，而且需要通过分子粪便学技术鉴定以避免误判。可以认为，监测方案以及管理实践必须得到优化并在整合国家、省和自然保护区各尺度基础上加以实施，以确保这些小而破碎化的种群得到长期有效的保护。

The leopard (Panthera pardus) is the most widely distributed wild felid species in the world. However, it is critical to note that their populations in Asia have declined sharply.
Historically, leopards were distributed across almost all of China, excluding the arid Gobi Desert and western mountainous areas above 4,000 meters above sea level. However, recent surveys have revealed that leopards have disappeared from many regions and are far less common than previously thought.
This study was conducted across multiple geographic scales with the aim of deepening our understanding of the conservation biology of leopards in China. We collected nationwide distribution records of leopards using multiple data sources to assess their current distribution and population status.
The study results showed that leopards have declined drastically in China, with confirmed distribution records only found at 44 sites across 11 provinces. The remaining leopard populations are small (<50 individuals) and fragmented, primarily occurring in isolated nature reserves, which exposes them to a high risk of local extinction.
Notably, most of these Chinese leopard distribution records were obtained during surveys and studies targeting other flagship species, rather than directly from leopard-specific projects or research.
As a result, existing leopard studies have failed to include habitats and ecosystems that do not overlap with other flagship species, particularly those specifically adapted to leopards—mainly areas close to human disturbance where other species have already gone locally extinct.
Shanxi Province is one such region: located in the central part of the Yellow River basin in North China, it is home to the North China leopard (P. p. japonensis), an endemic Chinese subspecies of the leopard.
In this study, we developed species distribution models to predict the environmental conditions suitable for leopard survival, thereby identifying areas that are viable for their habitation. First, we extracted leopard distribution data from camera trap records collected non-systematically across the entire province between 2008 and 2013. Next, we selected six uncorrelated variables representing topography, vegetation, and human activity within 1 km² grids, to analyze the impacts of environmental factors on snow leopard distribution.
Results from the MAXENT model showed that the most critical factor influencing leopard distribution is forest cover (contribution rate of 90%), while road density and human population density have relatively minor impacts (3.3% and 2.6%, respectively).
The study found that large areas of suitable habitat still exist in southern Shanxi Province, including two national nature reserves. We selected the largest of these, the Lishan National Nature Reserve, as the site for our first population density estimation.
We then systematically deployed 139 camera traps across the entire reserve (252 km²). Through individual identification of captured photos, we identified a total of 5 leopards, and analyzed the data using the most modern statistical modeling techniques available for population density estimation.
Results from the mark-recapture (MARK) and spatial mark-recapture (SPACECAP) models were relatively similar (0.7 and 0.6 ± 0.3 leopards per 100 km², respectively). This density is comparable to that of the critically endangered Amur leopard (P. p. orientalis) in Northeast China, and is also one of the lowest leopard density records globally.
The study also explored the feasibility of using sign surveys as an alternative or supplementary method to camera traps for population monitoring, but found very few signs: 34 scats, 3 tracks, 5 scratch marks, and 2 hair samples.
In fact, the species identity of wild scat samples cannot be confirmed without verification via DNA analysis. The study found that only 14% of scats identified as the target species in the field actually originated from leopards, with most others belonging to canids.
Due to the small sample size of leopard scats, further analysis was not feasible, so we analyzed carnivore communities in the snow leopard (Panthera uncia) landscape, and obtained similar results: over half of scats identified as snow leopard in the field actually came from red foxes (Vulpes vulpes).
These results demonstrate that due to such a high misidentification rate, species surveys and monitoring relying on scat samples should use molecular scatology techniques for species identification.
Additionally, the low detection rate of signs (0.1 signs/km) indicates that sign surveys cannot achieve the desired results unless survey effort is increased (by extending transect distances) and/or detection probability is improved (by conducting surveys during snow cover).
It is important to recognize that significant knowledge gaps still exist regarding leopards in China, and this lack of information reduces the effectiveness of conservation efforts.
The systematic surveys we conducted are suitable for monitoring and researching elusive species like leopards, even in areas with very low population densities.
On the other hand, sign surveys cannot provide sufficient sample sizes for analysis, and require molecular scatology identification to avoid misclassification.
It can be concluded that monitoring protocols and management practices must be optimized and implemented across national, provincial, and nature reserve scales to ensure the long-term effective conservation of these small, fragmented leopard populations.

AI搜集汇总

背景与挑战

背景概述

该数据集基于一篇研究论文，聚焦中国金钱豹的分布、现状与监测策略。研究发现金钱豹种群急剧减少，仅存于11个省的44个位点，种群小而分散，主要栖息在孤立的自然保护区中，面临较高的局地灭绝风险。研究采用红外相机和物种分布模型等方法，揭示森林覆盖率是影响分布的关键因素，并评估了监测策略，指出系统性调查适用于低密度种群，而痕迹调查需结合分子技术以提高准确性。