化学通讯是哺乳动物个体之间进行交流的主要通讯方式之一。对于营独居型生活的大熊猫，化学通讯在维系其社会关系、繁殖和交配发挥着关键作用。然而，以往有关大熊猫化学通讯的研究大都集中在圈养个体中，然而野生个体生活环境及其特殊、难于直接跟踪观察、样本很难获得，对于其化学通讯行为策略及其气味标记物化学成分尚缺系统研究。本文首先对秦岭佛坪大熊猫的尿液和肛周腺分泌物标记行为进行了研究，并探究其反标记行为策略。其次，借助红外相机技术，通过对王朗和佛坪两个国家级自然保护区的大熊猫化学通讯行为进行比较研究，探讨了不同生境类型中大熊猫的化学通讯行为策略。再次，采用气相色谱-质谱联用分析技术，对秦岭地区大熊猫肛周腺分泌物进行了分析，探究其化学信息的编码方式及其化合物成分在交配期和非交配期的季节性变化特征，并推断与繁殖相关的化学成分。同时，结合粪便分子生物学分析结果和红外相机资料，对遗留气味标记的个体进行了身份鉴定，并探讨雌性个体间化学通讯的性别差异。主要研究结果如下： 1．野生大熊猫气味标记的高度与体型显著正相关。体型越大，标记高度越高。在进行尿液标记时，个体都尽量地将腿抬高，最大化尿液标记的高度。同时，还发现野生大熊猫也存在“覆盖标记”和“邻近标记”两种反标记模式。这可能与大熊猫气味反标记的策略相关，大熊猫可以通过气味标记的高度来评估竞争对手的实力。 2. 环境差异影响大熊猫化学标记方式的选择。王朗和佛坪的大熊猫都会基于环境中标记树的特征来选择适当的标记方式，使得信号传递效率最大化。两种环境中大熊猫都偏好在胸径较大、有苔藓覆盖且非常粗糙的树干上进行尿液标记，而选择在较为粗糙、无苔藓覆盖的树干上进行肛周腺分泌物标记。王朗保护区总体海拔较高，气候更加湿润、凉爽，适合针叶树和苔藓的生长，乔木特征非常适合尿液标记气味的存留；而佛坪保护区主要为针阔混交林，以阔叶林为主，树木胸径较小、粗糙程度一般且苔藓覆盖较少，这些特征都非常适合大熊猫在树上进行肛周腺分泌物标记。这种“潜在标记树”特征差异导致王朗大熊猫尿液标记行为比佛坪更加频繁，而肛周腺标记行为相对比佛坪大熊猫更少。尿液标记和肛周腺标记在大熊猫信号传递中发挥着不同的功能，这种标记行为模式的改变可能会引起大熊猫繁殖和社会行为的变化。 3. 大熊猫肛周腺分泌物标记行为存在明显性别差异。DNA分子鉴定和红外相机数据表明，雌雄两性个体在标记频率上存在明显差异，雄性全年都进行肛周腺气味标记，交配季节标记更加频繁，而雌性个体很少进行肛周腺气味标记且主要分布在繁殖季节。 4. 雄性个体肛周腺分泌物中化学成分呈显著季节性变化。在雄性大熊猫的肛周腺标记物中共鉴定出104种化合物成分，包括27种醛类、22种脂肪酸、12种脂肪酸酯、10种类固醇、10个酮、5种杂环芳香族化合物、5种烯烃、4种酰胺类、4种醇类和角鲨烯等化合物。野生大熊猫呈现明显得季节性繁殖模式，在交配季节其标记物中也包含了更多的化学物质，存在32种特有的气味化合物，而所有被鉴定出的化合物中仅2种支链脂肪酸在交配季节的样本中缺失。推测这些气味化合物的季节性变化在大熊猫繁殖交配过程中发挥着重要作用，如对潜在异性配偶进行定位、性吸引和雄性个体间的竞争。 5. 肛周腺分泌物在野外环境中比较稳定，两周内的肛周腺气味标记无论是在物质数量还是主要化合物的相对丰富度均无明显变化。气味标记中相对分子质量较小的挥发性气味成分（如大量的短链醛类、酮类和脂肪酸）在交配季节物质的种类和含量都会增加，这些物质沸点较低，易挥发，适合于远距离传播化学信号；相对分子质量偏大的非挥发性物质，沸点较高，不易挥发，在环境中容易存留，更多的是作为固定剂或者用于近距离传播个体身体状况和社会地位等信息。研究还发现，在肛周腺气味标记中同一类气味化合物之间存在明显的正相关。气味信号间存在协同作用，通过两个或两个以上的分子共同编码传递同一个化学信号，这可能比任何一个单分子要产生更大的反应，能更大的限度的发挥化学信号的效果。 本论文研究结果为野生大熊猫化学通讯和繁殖特征提供了重要的数据背景支持。首次探究了野生大熊猫反标记行为；利用样线监测和红外相机对王朗和佛坪大熊猫气味标记行为进行了比较研究，在一定程度上揭示了大熊猫如何通过改变气味标记策略来适应不同的栖息地环境；初步解析了野生大熊猫肛周腺气味标记的信号密码，为野生大熊猫种群及其栖息地的保护和管理提供了重要的科学依据。同时，将野生大熊猫标记物的化学物质与圈养大熊猫进行比较，表明化学信息的缺乏可能是圈养大熊猫自然繁殖能力低下的重要原因之一，为圈养大熊猫的繁殖和管理提供了重要的理论依据。

Chemical communication is a dominant mode of communication in many mammals. The giant panda (Ailuropoda melanoleuca), a solitary species, forages most of the time in the dense bamboo forest and rarely confront with other conspecifics, thus most of the social interaction occurs in the mating seasons. Chemical communication, which is a dominantly communication in pandas, play an important role in maintaining their social relations and breeding. It’s been a hotspot research field about the giant panda chemical communication in animal behaviour and reproductive ecology. Due to topographic complexity, rigorous field work, a systematic research on the scent-marking behaviour strategies and chemical composition analysis of scent-marking in wild pandas is lack. To understand the over-mark strategies of the wild giant panda, we conducted our field study in Foping to collect information about the height of the urine and anogenital gland secretions (AGS) scent-mark of wild pandas. To determine whether the environments influence the scent-marking behaviours of giant pandas or not, we conducted a comparative study of giant panda scent-marking in Foping and Wanglang reserves in China by combining the line transect surveys and camera trapping. To evaluate the seasonal variation of the chemical signals associated with the reproductive status of the scent-marking in wild pandas, we deployed gas chromatography mass spectrometry (GC-MS) analysis of the chemical composition of the scent-marking, and coupled with the reproductive behavioral observations of wild pandas. We collected 88 AGS samples from marking trees in Foping Nature Reserve, China. We categorized these samples according to the reproductive period (Mating season: February-April; Non-mating season: other month) of giant panda. Specifically, we tested whether the chemicals in scent-marks of wild pandas were varied with the seasons and reproductive states, as theirs scent-marking behavioral patterns. We used DNA analysis of feces collected near the scent site and combined the camera–traps monitored in mark transects to identify each individual panda depositing the scent-marking. At the same time, we investigated the sex difference of marking frequency in wild pandas. The main results are as follows: 1. The height of the scent-marking is positively associated with the body size of the wild giant panda. An individual raises their leg up as much as possible when urinates. The larger the body size of individual, the higher leg lift for depositing urine scent-marking. This study also identified two different over-mark patterns in wild giant pandas: over-marking and adjacent-marking as in captivity. We speculated that those behaviours may be associated with the over-marking strategies of giant panda, indicating individuals strength of the opponent by the height of the scent marks. 2. The variation of ecological context can influence the chemo-signaling strategies in the giant panda. Our results indicated that the giant pandas in Foping and Wanglang, whether urine or AGS scent-marking, selected the scent-marking trees based on specific characteristics which can promote signals efficacy. For urine-marking, pandas prefer large, mossy trees consisting very rough bark trees, whereas for AGS marking, pandas prefer rough-bark and moss-free trees in both reserves. Hardwood trees with smooth bark are dominated in the habitat of giant panda in Foping, in contrast coniferous trees with rougher bark are dominated in Wanglang. Consequently, urine marking is significantly more prevalent in Wanglang than Foping. Because urine and AGS marking are used in different functional and motivational contexts, the variation in marking patterns likely has consequences for reproductive and social behaviour. 3. There is a significant difference in AGS scent-marking patterns between males and females in wild giant pandas. Both DNA and camera photo identification indicated that AGS scent-marking patterns were remarkably different. Males deposited scent-mark thought out the year and more frequent during mating season, whereas females rarely were seen depositing AGS scent-mark. 4. Seasonal changes of the chemical composition are strikingly different in males AGS scent-mark. We tentatively identified 104 compounds in 88 AGS samples, which were comprised of twenty-seven aldehydes, twenty-two fatty acids, twelve fatty-acid esters, ten steroids, ten ketones, five heterocyclic aromatic organic compounds, five alkenes, four amides, four alcohols, squalene and other compounds. There was a seasonal reproductive pattern in wild giant panda with more compounds in AGS during mating season. Here, we identified 32 compounds exclusively in the samples of mating season. Those compounds are four alcohols (1-Octanol; 1-Hexadecanol; 1-Docosanol and Tetracosanol), ten fatty acid esters, and five ketones (2,5-Hexanedione; 2-Octanone; 2-Undecanone; Heneicosanone and Tricosanone). There were only two types of branched fatty acids (Pentadecanoic acid and Heptadecanoic acid) missing in mating season. This seasonal variation in chemical composition of AGS most likely plays an important role in governing giant panda reproduction, including mate location, attraction and male-male competition. 5. The chemical compounds of the AGS is stable. We found that there is no obvious change in AGS composition neither in terms of the number of chemical compounds nor the relative abundance of the major compounds in the samples. This indicates that the duration of fade-out of AGS scent-mark is at least two weeks in the wild. Lower molecular weight composition, such as the short-chain aldehydes, short-chain ketones and short-chain fatty acids, which increased in mating season have low boiling point and easily fade out, were suitable for long distance transmission. However, the semi-volatile substances, higher boiling point, and easy to persist in the environment, were act as fixatives or encode information on individual identity and competitive status in wild giant pandas. Furthermore, our study found the synergistic effects of the major compounds in wild pandas. Many chemosignals were blended, which consist of two or more molecules to synergistically. Those blends, which act as biologically signals, can elicit a greater response than any other component. This study provides first evidence for chemical communication and reproduction in wild giant pandas. We explored the over-mark behavior strategies in wild pandas, and revealed how giant pandas alter scent-marking behaviour in different ecological contexts, preliminarily uncovered the password of the AGS scent-mark in wild giant pandas. It provides an important scientific basis for habitat protection and management for the wild giant panda. At the same time, the result of comparing the chemical compounds in AGS of wild and captive pandas suggests that inadequate chemical signals and communication might lead to reproductive failure in captive pandas. This result also provides a theoretical basis for captive breeding and management.