Wild Giant Panda Research Articles

A joint time and spatial attention-based transformer approach for recognizing the behaviors of wild giant pandas

Wild giant pandas, an endangered species exclusive to China, are a focus of conservation efforts. The behavior of giant pandas reflects their health conditions and activity capabilities, which play an important role in formulating and implementing conservation measures. Researching and developing efficient behavior recognition methods based on deep learning can significantly advance the study of wild giant panda behavior. This study introduces, for the first time, a transformer-based behavior recognition method termed PandaFormer, which employs time-spatial attention to analyze behavioral temporal patterns and estimate activity spaces. The method integrates advanced techniques such as cross-fusion recurrent time encoding and transformer modules, which handle both temporal dynamics and spatial relationships within panda behavior videos. First, we design cross-fusion recurrent time encoding to represent the occurrence time of behaviors effectively. By leveraging the multimodal processing capability of the transformer, we input time and video tokens into the transformer module to explore the relation between behavior and occurrence time. Second, we introduce relative temporal weights between video frames to enable the model to learn sequential relationships. Finally, considering the fixed position of the camera during recording, we propose a spatial attention mechanism based on the estimation of the panda's activity area. To validate the effectiveness of the model, a video dataset of wild giant pandas, encompassing five typical behaviors, was constructed. The proposed method is evaluated on this video-level annotated dataset. It achieves a Top-1 accuracy of 92.25 % and a mean class precision of 91.19 %, surpassing state-of-the-art behavior recognition algorithms by a large margin. Furthermore, the ablation experiments validate the effectiveness of the proposed temporal and spatial attention mechanisms. In conclusion, the proposed method offers an effective way of studying panda behavior and holds potential for application to other wildlife species.

Large-scale genome sequencing of giant pandas improves the understanding of population structure and future conservation initiatives

The extinction risk of the giant panda has been demoted from "endangered" to "vulnerable" on the International Union for Conservation of Nature Red List, but its habitat is more fragmented than ever before, resulting in 33 isolated giant panda populations according to the fourth national survey released by the Chinese government. Further comprehensive investigations of the genetic background and in-depth assessments of the conservation status of wild populations are still necessary and urgently needed. Here, we sequenced the genomes of 612 giant pandas with an average depth of ~26× and generated a high-resolution map of genomic variation with more than 20 million variants covering wild individuals from six mountain ranges and captive representatives in China. We identified distinct genetic clusters within the Minshan population by performing a fine-grained genetic structure. The estimation of inbreeding and genetic load associated with historical population dynamics suggested that future conservation efforts should pay special attention to the Qinling and Liangshan populations. Releasing captive individuals with a genetic background similar to the recipient population appears to be an advantageous genetic rescue strategy for recovering the wild giant panda populations, as this approach introduces fewer deleterious mutations into the wild population than mating with differentiated lineages. These findings emphasize the superiority of large-scale population genomics to provide precise guidelines for future conservation of the giant panda.

System evolution model of human disturbances in panda habitats

Anthropogenic associated habitat fragmentation has had a profound impact on great panda populations and their geographical distribution. Quantifying how panda disturbances manifest temporally and to what extent spatially can be challenging to model given the complexity of data structures and predictor variables involved. To inform future panda protection efforts and research in conservation ecology, we adopted a new approach to modeling the spatial patterns of simultaneously occurring panda disturbances over time using a system evolution model. Based on the 3rd (1999–2003) and 4th (2011–2014) National Wild Giant Panda field surveys, we analyzed the development and spatial trends of local and regional habitat disturbances affecting panda populations in China. Between 1999 and 2014, human disturbances in Minshan and Xiaoxiangling decreased by 16.15% and 36.62%, respectively, while increasing trends of 7.59%, 0.89%, and 34.49% were observed in Qionglaishan, Daxiangling, and Liangshan respectively. Ecological protection measures have also increased in Minshan, Xiaoxiangling, and Liangshan by 4.10%, 19.01%, and 4.56%, respectively, whilst a reverse trend was identified in Qionglaishan (-3.52%) and Daxiangling (-8.09%). The regional imbalance of human disturbances in Minshan, Xiaoxiangling, and Liangshan increased by 46.36%, 33.59%, and 18.82%, respectively, whilst decreasing in Qionglaishan (-4.81%) and Daxiangling (-36.39). Our findings demonstrate the utility of information entropy theory in examining cross-sectional data of multiple disturbances affecting wildlife populations. Application of this method in the context of panda habitat fragmentation has helped identify unique temporal and spatial characteristics of human disturbances impacting panda habitats over time.

Giant panda face recognition based on PandaFaceNet

Individual recognition of animals via infrared camera trapping surveys is an important method for protecting and monitoring animals in the wild. However, several factors limit current survey methods used for individual animal recognition, such as the lack of accuracy and extensive time required to process data. Recently, new technologies and methods for individual recognition of animal images have been developed for rare wildlife species (e.g., giant pandas and lemurs). These new technologies require adequate and high-quality sampled images; however, it can be challenging for researchers to obtain an adequate sample size of wildlife images from the field. To overcome this problem, we proposed and tested a new small-sample individual recognition method adapted from FaceNet called PandaFaceNet, using data from a self-built giant panda (Ailuropoda melanoleuca) facial image database. We tested the proposed giant panda individual recognition method on unknown captive and wild giant panda datasets. The results showed that this method has 95.3% recognition accuracy for distinguishing among two captive giant panda facial images and 91% recognition accuracy for distinguishing among two wild giant pandas. Notably, PandaFaceNet achieves individual recognition through comparing two images and is an open-set identification method. Therefore, PandaFaceNet provides a novel method for giant panda research by opening up opportunities for analysis of small sample sizes of panda imagery data, while also providing new directions for research on rare wildlife more broadly.

Air Quality Impacts on the Giant Panda Habitat in the Qinling Mountains: Chemical Characteristics and Sources of Elements in PM2.5

The wild giant panda habitat is inaccessible and far away from the main areas of human activity, so environmental pollutants entering the habitat are mainly the result of external migration and spread through the atmospheric advection and diffusion processes and particulate matter deposition. To research the variation, transmission route, chemical characteristics, and source of PM2.5 in the habitat of wild giant pandas, we set up a PM2.5 sampling point near the Shaanxi Foping National Nature Reserve (SFNNR), which is the area with the highest population density of wild giant pandas in the Qinling Mountains. The 12-month average concentration of PM2.5 was 11.3 ± 7.9 μg/m3 from July 2021 to June 2022, and the 12-month average concentration did not exceed the limit value set in the standard. In the results of our analysis of element concentrations, As and Pb were much lower than the limit standard. Si, S, P, and Cl accounted for 99.60% of nonmetallic elements, while the proportion of the six metallic elements, Na, Mg, Al, K, Ca, and Fe, was 96.27%. According to the analysis results of enrichment factor (EF) and pollutant emission sources, there were four sources for the total elements in PM2.5, which were mainly distributed in the areas around the reserve. These included dust, coal combustion, biomass burning, and traffic-related emissions, which contributed 55.10%, 24.78%, 11.91%, and 8.22% of the total element mass in PM2.5, respectively. Additionally, Pb, Cu, Zn, As, Sc, Co, Ga, Mg, and, especially, Se were severely affected by human activities (coal burning, biomass burning, and traffic-related emissions). In the villages and towns around the wild giant panda habitat, the majority of energy for cooking and heating comes from coal and biomass burning, and older vehicles with high emissions are used more frequently. Therefore, to better protect the health of and reduce the impact of environmental pollution on wild giant pandas, we put forward relevant recommendations, including upgrading the energy structure of towns and villages near the habitat to increase the proportion of clean energy, such as photovoltaic power generation, natural gas, etc.; decreasing the combustion of coal and biomass; encouraging the upgrading of agricultural diesel machines and older vehicles used in these areas; and setting limits on vehicle emissions in areas surrounding the habitat.

Feeding habits and foraging patch selection strategy of the giant panda in the Meigu Dafengding National Nature Reserve, Sichuan Province, China.

Despite retaining a digestive system with carnivorous characteristics, the giant panda has now specialized to eat bamboo. Therefore, research on its feeding habits and foraging strategies has always been an important field in the study of giant panda population ecology. Located in the core distribution area of the giant panda population in the Liangshan mountains, the Meigu Dafengding Nature Reserve acts as the link between the major panda reserves in the Liangshan mountains and the key corridor zone connecting the giant panda populations in the east and west. This study was performed in 2017 and aimed to determine the feeding habits of giant pandas in this region, the heterogeneity of food resources in different seasons, and the selection strategy for foraging patches. To achieve this, we used the line transect method to investigate the giant panda population, collected fresh feces, analyzed the nutritional composition of bamboo samples, and set up a sampling grid of foraging patches in different seasonal distribution areas. It was found that wild giant pandas in the region spend most of the year at lower elevations(1939-3296m) foraging for Yushania maculata and Yushania ailuropodina, and only move to higher elevations(2844-3770m) in summer for Bashania fangiana. Their preferred foraging patches(n = 65) had certain topographical features, such as gentle slopes(< 25°) and closer proximity to water sources(< 500m), and the nutritional quality of bamboo resources in foraging patches was significantly better than in those they avoided or rarely used. The food habits and foraging patch selection strategies of giant pandas in the Liangshan and Qinling mountains are relatively similar, and the results of this study can help provide a scientific basis for the conservation and management of wild giant panda habitats in different mountain regions.

Large-scale genetic surveys for main extant population of wild giant panda (Ailuropoda melanoleuca) reveals an urgent need of human management.

There are only six isolated living giant panda populations, and a comprehensive understanding of their genetic health status is crucial for the conservation of this vulnerable species. Liangshan Mountains is one of the main distribution areas of living giant pandas and is outside the newly established Giant panda national park. In this study, 971 giant panda fecal samples were collected in the heartland of Liangshan Mountains (Mabian Dafengding Nature Reserve: MB; Meigu Dafengding Nature Reserve: MG; and Heizhugou Nature Reserve: HZG). Microsatellite markers and mitochondrial D-loop sequences were used to estimate population size and genetic diversity. We identified 92 individuals (MB: 27, MG: 22, HZG: 43) from the three reserves. Our results showed that: (1) genetic diversity of three giant panda populations was moderate; (2) several loci deviated significantly from the Hardy-Weinberg equilibrium and almost all these deviated loci showed significant heterozygote deficiencies and inbreeding; (3) three giant panda populations have substantial genetic differentiation with the most differentiation between MB and the two other populations; and (4) a large amount of giant panda feces outside the three reserves were found, implying the existence of protection gap. These results indicated that under stochastic events, the giant panda populations in Liangshan Mountains are at risk of genetic decline or extinction and urgent need of human management. This study revealed that high attention should be paid to the protection of these giant panda populations outside the Giant panda national park, to ensure their survival in their distribution areas.

Spatiotemporal evolution analysis of human disturbances on giant panda: A new approach to study cumulative influences with large spatial scales

The disturbance of wildlife habitats by human activities has become an important driving force in the evolution of ecological landscapes. However, human disturbances in panda habitats are highly diverse and complex. To study the changes in the temporal and spatial dynamics of human disturbances affecting panda populations, an analysis model for the evolution of human interference was designed using information entropy and statistical methods. Based on data from the third and fourth National Wild Giant Panda surveys in China, we analyzed the developmental trend, spatial distribution, and characteristics of the current state and regional distribution of human disturbances affecting panda populations. The results showed that shoot gathering, farming, and tourism were key local distribution factors, whereas hunting, road infrastructure, and logging had a more widespread impact on panda habitat. Livestock and herb gatherings, however, showed no clear distribution patterns. The influence of the main disturbance factors (logging, livestock, road infrastructure, and shoot gathering) tended to increase over time, whereas the development of the relative strength and weakness of the disturbances showed an evolutionary trend. Although logging, livestock, shoot gathering, and road infrastructure have historically been the main panda habitat disturbances, the impact of logging and shoot gathering has been diminishing over time, with road infrastructure emerging as the leading disturbance factor. Farming has shaped the distribution area of giant panda habitats over the past several hundred years. Additionally, logging has caused current fragmentation patterns for several decades. Moreover, it is estimated that road infrastructure will have a broad and profound impact in the coming decades. Our results indicate that the information entropy theory can be effectively used to analyze the temporal and spatial characteristics of human disturbances that affect wildlife. Our findings will provide a basis for ecological policies that aid the preservation of threatened panda populations and guide future research on habitat disturbance.

Dynamic threats of nighttime light-represented human activities to giant pandas and their habitat

Human activities are one of the main factors threatening the long-term survival of wild giant pandas. Long-term and large-scale human activities can be quickly assessed by night light remote sensing data, which has been more and more widely used. In this paper, we analyzed the spatial patterns of nighttime light intensity and their dynamic changes from 2000 to 2020 in the entire giant panda distribution area using long-term nighttime light remote sensing data, and evaluated the dynamic threats of nighttime light-represented human activities to giant panda populations by proximity of light source and its intensity. We also assessed the relationships between the nighttime light dynamics and the trends of habitat fragmentation. Results showed that the intensity and range of the nighttime light around panda habitat had increased significantly from 2000 to 2020. The nighttime light intensity inside the natural reserves is significantly lower than outside, and it becomes more obvious over time. The intensity and range of nighttime light inside the natural reserves first increased and then decreased during the two decades, indicating that the protection inside the natural reserves effectively reversed the rapid increase in human activities in the previous decade. From 2000 to 2020, nighttime light sources became closer to local panda populations. In particular, they approached the pandas in the first decade, and stayed away in the second decade. This is mainly reflected in the weakened threats of nighttime light-represented human activities on the core large populations of the Qinling, Minshan and Qionglai Mountains from 2010 to 2020, but the threats on the peripheral populations continued to strengthen, where nighttime light became not only closer but also more intensified. The increase in nighttime light intensity and range were also significantly positively correlated with the increase in habitat fragmentation, especially during 2000–2010. Our study reveals the dynamic changes of nighttime light-represented human activities’ threats to wild giant panda and its habitat, gives advices for effectively protecting giant pandas and their habitats from human activities, and has implications for the assessment of human activities’ threats to other species in the world.

Automatic Recognition of Giant Panda Attributes from Their Vocalizations Based on Squeeze-and-Excitation Network.

The giant panda (Ailuropoda melanoleuca) has long attracted the attention of conservationists as a flagship and umbrella species. Collecting attribute information on the age structure and sex ratio of the wild giant panda populations can support our understanding of their status and the design of more effective conservation schemes. In view of the shortcomings of traditional methods, which cannot automatically recognize the age and sex of giant pandas, we designed a SENet (Squeeze-and-Excitation Network)-based model to automatically recognize the attributes of giant pandas from their vocalizations. We focused on the recognition of age groups (juvenile and adult) and sex of giant pandas. The reason for using vocalizations is that among the modes of animal communication, sound has the advantages of long transmission distances, strong penetrating power, and rich information. We collected a dataset of calls from 28 captive giant panda individuals, with a total duration of 1298.02 s of recordings. We used MFCC (Mel-frequency Cepstral Coefficients), which is an acoustic feature, as inputs for the SENet. Considering that small datasets are not conducive to convergence in the training process, we increased the size of the training data via SpecAugment. In addition, we used focal loss to reduce the impact of data imbalance. Our results showed that the F1 scores of our method for recognizing age group and sex reached 96.46% ± 5.71% and 85.85% ± 7.99%, respectively, demonstrating that the automatic recognition of giant panda attributes based on their vocalizations is feasible and effective. This more convenient, quick, timesaving, and laborsaving attribute recognition method can be used in the investigation of wild giant pandas in the future.

Automatically detecting the wild giant panda using deep learning with context and species distribution model

The giant panda is a flagship species in ecological conservation. The infrared camera trap is an effective tool for monitoring the giant panda. Images captured by infrared camera traps must be accurately recognized before further statistical analyses can be implemented. Previous research has demonstrated that spatiotemporal and positional contextual information and the species distribution model (SDM) can improve image detection accuracy, especially for difficult-to-see images. Difficult-to-see images include those in which individual animals are only partially observed and it is challenging for the model to detect those individuals. By utilizing the attention mechanism, we developed a unique method based on deep learning that incorporates object detection, contextual information, and the SDM to achieve better detection performance in difficult-to-see images. We obtained 1169 images of the wild giant panda and divided them into a training set and a test set in a 4:1 ratio. Model assessment metrics showed that our proposed model achieved an overall performance of 98.1% in mAP0.5 and 82.9% in recall on difficult-to-see images. Our research demonstrated that the fine-grained multimodal-fusing method applied to monitoring giant pandas in the wild can better detect the difficult-to-see panda images to enhance the wildlife monitoring system.

Comparison of genetic characteristics between captive and wild giant pandas based on 13 mitochondrial coding genes.

Research on genetic diversity based on mitochondrial DNA of giant pandas mainly focused on a single marker or a few genes. To provide a more comprehensive assessment of the genetic diversity on giant pandas based on 13 mitochondrial protein coding genes. We assembled 13 protein coding genes in the mitochondrial genome of the giant panda based on the whole genome sequencing data, including ND1, ND2, COX1, COX2, ATP8, ATP6, COX3, ND3, ND4L, ND4, ND5, ND6 and Cyt b. We successfully obtained long sequence of 11,416 base pairs with all 13 genes for 110 giant panda individual, accounting for 67.93% in length of the mitochondrial reference genome. Haplotype diversity was 0.9518 ± 0.009 and nucleotide diversity (π) was 0.00157 ± 0.00014. We detected three new haplotypes, including GPC10 and GPC21 for the CR sequence and GPB12 for the Cyt b gene. These multi-gene sequences provided more genetic variable information to compare captive and wild giant panda population.

Complex effects of habitat amount and fragmentation on functional connectivity and inbreeding in a giant panda population

AbstractThe relationships between habitat amount and fragmentation level and functional connectivity and inbreeding remain unclear. Thus, we used genetic algorithms to optimize the transformation of habitat area and fragmentation variables into resistance surfaces to predict genetic structure and examined habitat area and fragmentation effects on inbreeding through a moving window and spatial autoregressive modeling approach. We applied these approaches to a wild giant panda population. The amount of habitat and its level of fragmentation had nonlinear effects on functional connectivity (gene flow) and inbreeding. Functional connectivity was highest when approximately 80% of the surrounding landscape was habitat. Although the relationship between habitat amount and inbreeding was also nonlinear, inbreeding increased as habitat increased until about 20% of the local landscape contained habitat, after which inbreeding decreased as habitat increased. Because habitat fragmentation also had nonlinear relationships with functional connectivity and inbreeding, we suggest these important responses cannot be effectively managed by minimizing or maximizing habitat or fragmentation. Our work offers insights for prioritization of protected areas.

Genome of the Giant Panda Roundworm Illuminates Its Host Shift and Parasitic Adaptation

Baylisascaris schroederi, a roundworm (ascaridoid) parasite specific to the bamboo-feeding giant panda (Ailuropoda melanoleuca), represents a leading cause of mortality in wild giant panda populations. Here, we present a 293-megabase chromosome-level genome assembly of B. schroederi to infer its biology, including host adaptations. Comparative genomics revealed an evolutionary trajectory accompanied by host-shift events in ascaridoid parasite lineages after host separations, suggesting their potential for transmission and rapid adaptation to new hosts. Genomic and anatomical lines of evidence, including expansion and positive selection of genes related to the cuticle and basal metabolisms, indicate that B. schroederi undergoes specific adaptations to survive in the sharp-edged bamboo-enriched gut of giant pandas by structurally increasing its cuticle thickness and efficiently utilizing host nutrients through gut parasitism. Additionally, we characterized the secretome of B. schroederi and predicted potential drug and vaccine targets for new control strategies. Overall, this genome resource provides new insights into the host adaptation of B. schroederi to the giant panda as well as the host-shift events in ascaridoid parasite lineages. Our findings on the unique biology of B. schroederi will also aid in the development of prevention and treatment measures to protect giant panda populations from roundworm parasitism.

Spatial models of giant pandas under current and future conditions reveal extinction risks.

In addition to habitat loss and fragmentation, demographic processes-the vagaries of births, deaths and sex ratio fluctuations-pose substantial threats to wild giant panda populations. Additionally, climate change and plans for the Giant Panda National Park may influence (in opposing directions) the extinction risk for wild giant pandas. The Fourth National Giant Panda Census showed pandas living in 33 isolated populations. An estimated 259 animals live in 25 of these groups, ~14% of the total population. We used individual-based models to simulate time series of these small populations for 100 years. We analysed the spatial pattern of their risk of extinction under current conditions and multiple climate change models. Furthermore, we consider the impact of the proposed Giant Panda National Park. Results showed that 15 populations face a risk >90%, and for 3 other populations the risk is >50%. Of the 15 most at-risk populations, national parks can protect only 3. Under the Representative Concentration Pathway 8.5 climate change scenario, the 33 populations will probably further divide into 56 populations. Some 41 of them will face a risk >50% and 35 face a risk >90%. Although national parks will probably connect some fragmented habitats, 26 populations will be outside national park planning. Our study gives practical advice for conservation policies and management and has implications for the conservation of other species in the world that live in isolated, fragmented habitats.

Development of recombinase polymerase amplification assays for rapid and visual detection of canine distemper virus infecting giant panda

BackgroundCanine distemper virus (CDV) is an enveloped negative-strand RNA virus that exhibits a high mutation rate and continuously expands the range of hosts. Notably, CDV has infected giant panda with spill over from viral reservoirs in canines. Giant pandas (Ailuropoda melanoleuca), especially captive pandas, are known to be susceptible to natural infection with CDV. The high fatality rate of CDV poses a serious threat to the safety of the giant panda population. However, vaccines or drugs for canine distemper in giant pandas have not been developed to date. Therefore, a rapid test that can achieve accurate onsite detection of CDV is important to enable the timely implementation of control measures. In this study, we established a nucleic acid visualization assay for targeting the CDV N gene by using combines reverse transcription recombinase polymerase amplification with a closed vertical flow visualization strip (RT-RPA-VF).ResultsThe RT-RPA-VF assay does not require sophisticated equipment, and it was determined to provide rapid detection at 35 °C for 30 min, while the limit of detection was 5 × 101 copies/μl RNA transcripts and 100.5 TCID50 ml− 1 viruses. The results showed that the assay was high specific to CDV and had no cross-reactivity with other viruses infecting the giant panda. Compared with RT-qPCR, RT-RPA-VF assay had a sensitivity of 100% and a specificity of 100% in 29 clinical samples. The coincidence rate between RT-RPA-VF and RT-qPCR was 100% (kappa = 1), indicating that the RT-RPA-VF assay possessed good diagnostic performance on clinical samples.ConclusionsThe RT-RPA-VF provides a novel alternative for the simple, sensitive, and specific identification of CDV and showed great potential for point of care diagnostics for captive and wild giant panda.

Changes in the MicroRNA Profile of the Giant Panda After Canine Distemper Vaccination and the Integrated Analysis of MicroRNA-Messenger RNA.

Canine distemper (CD) is a significant threat to wild and captive giant panda populations. Captive giant pandas are inoculated with canine distemper virus (CDV) vaccination to prevent the infection with the CDV. As an important regulator, microRNA (miRNA) plays a crucial role in regulating gene expression, including in disease immunity. To understand the role of miRNA in immune response to CDV vaccination, we investigated the miRNA expression profile in five giant panda cubs after two inoculations, 21 days apart. A total of 187 conserved miRNAs and 96 novel miRNAs were identified. Among the 187 conserved miRNAs, 29 differentially expressed miRNAs were found postinoculation. The upregulation of miR-16, miR-182, miR-30b, and miR-101 indicated that the innate immune may be enhanced, whereas the upregulation of miR-142 and miR-19a are probably involved in the enhanced cellular immune response. However, the downregulated miR-155 and miR-181a might indicate the giant panda has weak ability to produce antibodies and memory B cells. Integrated analysis of miRNA-messenger RNA (mRNA) found 20 negatively regulated miRNA-mRNA pairs, where downregulated miR-204 might enhance giant panda cub innate immunity by increasing TLR6 expression, and downregulated miR-330 might activate macrophages and regulate the immune response by increasing TMEM106A expression. Our research provides key information for future development to enhance the immune response of giant pandas and potentially improve the survival of captive and wild giant panda populations threatened by CD.

Development of Giant Panda Nature Reserves in China: Achievements and Problems

This paper provides a review of the history and development of Giant Panda Nature Reserves in China over the past 56 years. Four challenging issues facing the Giant Panda Nature Reserves are discussed. (i) Although the total population of wild giant pandas is increasing, the local populations are declining in small isolated reserves. This is concerning because the current population structure of wild giant panda shows a trend of continuing fragmentation and isolation. These phenomena are driven by the combined effects of natural geography and human disturbance. (ii) The full range of giant panda habitat is not currently under the protection of the network of nature reserves. Although Giant Panda Nature Reserves cover a total area of 35,326 km2, only 25,765 km2 of panda habitat overlaps with the reserves, with large areas of panda habitat (with documented evidence of giant panda activity) not included in the reserves. Conservation regulations have limited effects on human activities in these areas, resulting in further habitat fragmentation and reduced connectivity among giant panda populations. (iii) The pressure of the management system of Giant Panda Nature Reserves. Because of the lack of regulation outside reserves, stress is being transferred to these areas from the reserves, resulting in their being exposed to continuously increasing human pressure and aggravating conflicts between giant panda conservation and economic development. (iv) There are ongoing conflicts between giant panda conservation and the livelihoods of people who live within or around the reserves. On one hand, resource utilization conflict exacerbates the change of land cover. The grain growing industry and deforestation place great pressure on efforts to conserve the giant panda. On the other hand, a shortage of funds for conservation and lack of balanced policies have exacerbated existing conflicts between the conservation of giant pandas and the livelihoods of people. The main conclusions related to the development of Giant Panda Natural Reserves are: (1) The management system and regulatory regime of Giant Panda Nature Reserves in China have continuously evolved and improved, (2) The level of socioeconomic development in the communities within and surrounding the Giant Panda Nature Reserves have progressively increased, and (3) A Giant Panda National Park is currently under development.

Adaptability of the structure and biomass of Arundinaria spanostachya clonal populations grazing by wild giant pandas.

To understand the defense mechanism of Arundinaria spanostachya clonal populations in response to grazing by giant pandas, dynamic variations in A. spanostachya clonal population structure and biomass allocation in a wild giant panda habitat at the Liziping Nature Reserve were evaluated, as well as whether the clonal populations would be continuously used by the wild giant pandas. The population density of each age-class in the grazed and control plots after grazing (2014a and 2015a) was similar to that before grazing (2013a). The effects of grazing on the size-class and height-class structures were relatively lower. Before and after grazing, the perennial individuals showed the highest total biomass, followed by the biennial and annual individuals, and the maximum dry matter content in each module was found in the culm, followed by the branch and leaf. The dry matter content of A. spanostachya individuals increased as the age class increased, whereas the total water content decreased. The maximum water content allocation in the modules was observed in the culm, and no significant differences were found between the shoot and leaf. Thus, foraging by the wild giant pandas had no impact on the size-class and height-class structures and biomass allocation of A. spanostachya clonal populations, and the clonal populations have established an adaptive mechanism against grazing by giant pandas. After grazing, the A. spanostachya clonal populations showed greater self-adjustment ability to restore the status to that before grazing and, thus, continuously supply food for the giant pandas. Further management intervention of A. spanostachya clonal populations after the foraging of wild giant pandas is not needed, which has implications for understanding the impact of co-evolutionary mechanisms between giant panda and its staple bamboo species.

The “wildness” of the giant panda gut microbiome and its relevance to effective translocation

Since 2005, 12 captive pandas have been translocated into the wild. Of these, three were killed by conditional pathogen infections after release. Given that the gut microbiome is important to host health and environmental adaptation, increasing the “wildness” of the gut microbiomes of captive pandas might improve translocation success. Here, we analyzed the long-term monitoring data (16S MISEQ data about 700 samples) of two wild giant panda populations (Xiaoxiangling and Minshan Mountains populations), five GPS-collared translocated giant pandas (TPs), and their living environment (dietary bamboos and soils) in the Liziping Reserve, China. We found that panda gut microbiome composition shifted dramatically after translocation: in the captive panda gut microbiome, Enterobacteriaceae dominated, while in the wild panda gut microbiome, Pseudomonadaceae dominated. More “wild” gut microbes were identified in the wild and semi-wild translocated panda guts than in the panda guts prior to release. However, under the current translocation strategy, the semi-wild stage lasts only 2–3 months, which might not be long enough to establish a stable “wild” gut microbial community. The observed changes in the translocated panda gut microbial communities (i.e., the proportional increase in Pseudomonadaceae abundance) and metagenomic function (i.e., the higher proportion of essential amino acid metabolic activity) might reflect a putative evolutionary adaptation to the local environment at the translocation site. We suggest that candidate pandas live with their mothers in a fenced area at the translocation site for an additional year prior to release. This is also more similar to the typical life cycle of the local wild giant pandas.