Wildlife Monitoring Research Articles

The Motion-Activated Picture Capture, Animal Detection, and Alert System

Abstract: The Motion-Activated Picture Capture, Animal Detection, and Alert System is an advanced wildlife monitoring project employing motion sensors and high- resolution cameras. When motion is detected, the system captures images and utilizes sophisticated algorithms for accurate animal detection and classification. Trained on diverse datasets, the machine learning algorithms continuously improve detection accuracy, adapting to various environmental conditions. The system includes a realtime alert mechanism, notifying authorities or conservationists when rare or endangered species are identified, enabling swift response to potential threats or conservation opportunities. Beyond conservation, the project has applications in research, education, and ecotourism, offering valuable data for biodiversity studies, educational engagement, and immersive wildlife experiences for visitors. This comprehensive solution leverages cutting-edge technologies to contribute significantly to wildlife protectionand ecosystem preservation.

Automatic Bird Species Recognition from Images with Feature Enhancement and Contrastive Learning

Accurate bird species recognition is crucial for ecological conservation, wildlife monitoring, and biological research, yet it poses significant challenges due to the high variability within species and the subtle similarities between different species. This paper introduces an automatic bird species recognition method from images that leverages feature enhancement and contrast learning to address these challenges. Our method incorporates a multi-scale feature fusion module to comprehensively capture information from bird images across diverse scales and perspectives. Additionally, an attention feature enhancement module is integrated to address noise and occlusion within images, thus enhancing the model’s robustness. Furthermore, employing a siamese network architecture allows effective learning of common features within instances of the same class and distinctions between different bird species. Evaluated on the CUB200-2011 dataset, our proposed method achieves state-of-the-art performance, surpassing existing methods with an accuracy of 91.3% and F1 score of 90.6%. Moreover, our approach showcases a notable advantage in scenarios with limited training data. When utilizing only 5% of the training data, our model still achieves a recognition accuracy of 65.2%, which is significantly higher than existing methods under similar data constraints. Notably, our model exhibits faster execution times compared to existing methods, rendering it suitable for real-time applications.

High-precision tracking and positioning for monitoring Holstein cattle.

Enhanced animal welfare has emerged as a pivotal element in contemporary precision animal husbandry, with bovine monitoring constituting a significant facet of precision agriculture. The evolution of intelligent agriculture in recent years has significantly facilitated the integration of drone flight monitoring tools and innovative systems, leveraging deep learning to interpret bovine behavior. Smart drones, outfitted with monitoring systems, have evolved into viable solutions for wildlife protection and monitoring as well as animal husbandry. Nevertheless, challenges arise under actual and multifaceted ranch conditions, where scale alterations, unpredictable movements, and occlusions invariably influence the accurate tracking of unmanned aerial vehicles (UAVs). To address these challenges, this manuscript proposes a tracking algorithm based on deep learning, adhering to the Joint Detection Tracking (JDT) paradigm established by the CenterTrack algorithm. This algorithm is designed to satisfy the requirements of multi-objective tracking in intricate practical scenarios. In comparison with several preeminent tracking algorithms, the proposed Multi-Object Tracking (MOT) algorithm demonstrates superior performance in Multiple Object Tracking Accuracy (MOTA), Multiple Object Tracking Precision (MOTP), and IDF1. Additionally, it exhibits enhanced efficiency in managing Identity Switches (ID), False Positives (FP), and False Negatives (FN). This algorithm proficiently mitigates the inherent challenges of MOT in complex, livestock-dense scenarios.

Occurrence and temporal activity pattern of Burmese Red Serow (Capricornis rubidus, Bovidae) in Baraiyadhala National Park, Bangladesh: insights from a camera trapping study

Abstract No single studies on the Burmese Red Serows, a little known bovid to date yet considered as nationally endangered species, have been conducted in Bangladesh. As part of a systematic wildlife inventory and monitoring project, this study utilized 48 camera traps to elucidate their occurrence, activity pattern and possible sympatric association with other artiodactyls species inhabiting Baraiyadhala National Park. The study found 25 independent serow events with relative abundance index (RAI) of 0.89. The number of individuals seems low, but this is by far the country’s stronghold population. The Red Serows are mostly nocturnal and showed moderate temporal activity with barking dear (∆1 = 0.59) and wild boar (∆1 = 0.62) in this area. Anthropogenic pressures due to tourists (RAI = 4.8) and poachers (RAI = 1.7) were highly evident and must have negative impact on overall wildlife of the national park. The study suggests similar studies in adjacent protected area (Hazarikhil Wildlife Sanctuary), habitat protection and restoration program, and awareness raising program targeting forest dependent communities for long term conservation of this species.

Bridging the gap: bioinspired robotics as catalyst for interdisciplinary education

IntroductionIn contemporary educational philosophy, constructivist and constructionist theories emphasize active knowledge construction among learners. These paradigms advocate for learners as active participants, engaging in knowledge creation through interaction and experience. Problem-Based Learning (PBL) exemplifies these principles by placing students at the center of hands-on challenges that foster critical thinking and collaborative problem-solving. This study explores the integration of these educational theories through an innovative pedagogical framework, focusing on the development of bioinspired robotics.MethodsIn a six-stage educational process at Tecnológico de Monterrey, 24 undergraduates were immersed in bioinspired robotics through workshops on animal biomechanics, soft robotics, and neuroscience. Organized into four teams, students designed robots for wildlife observation and search and rescue. The curriculum integrated theoretical lectures, hands-on training, and practical applications, emphasizing interdisciplinary collaboration and real-world problem-solving. This approach blended traditional education with innovative, project-based learning tailored to local and ecological challenges.ResultsThe culmination of this initiative was marked by the creation of functional robotic prototypes, underscoring the effectiveness of the interdisciplinary approach. Post-project evaluations revealed marked enhancements in students' proficiency in both sciences and engineering. Furthermore, there was a pronounced strengthening of students' conviction regarding the importance of integrating bioinspired principles into engineering education.DiscussionThe study advocates merging traditional knowledge with modern technological innovation and bioinspiration to enhance learning outcomes. This integrated approach boosts analytical reasoning and scientific skills while also fostering empathy and emotional literacy, preparing students as holistic thinkers and versatile innovators. Combining biology, robotics, and education offers significant benefits, providing insights for educators and policymakers to equip students for future challenges. There is a strong consensus on the value of incorporating biological principles into engineering education, signaling a shift toward innovative, interdisciplinary curricula that enhance technical and broader cognitive skills.

Automated Detection and Counting of Wild Boar in Camera Trap Images.

Camera traps are becoming widely used for wildlife monitoring and management. However, manual analysis of the resulting image sets is labor-intensive, time-consuming and costly. This study shows that automated computer vision techniques can be extremely helpful in this regard, as they can rapidly and automatically extract valuable information from the images. Specific training with a set of 1600 images obtained from a study where wild animals approaching wild boar carcasses were monitored enabled the model to detect five different classes of animals automatically in their natural environment with a mean average precision of 98.11%, namely 'wild boar', 'fox', 'raccoon dog', 'deer' and 'bird'. In addition, sequences of images were automatically analyzed and the number of wild boar visits and respective group sizes were determined. This study may help to improve and speed up the monitoring of the potential spread of African swine fever virus in areas where wild boar are affected.

Rapid molecular species identification of mammalian scat samples using nanopore adaptive sampling

Rapid molecular species identification of mammalian scat samples using nanopore adaptive sampling

Integration of Indigenous Research Methodologies, Traditional Ecological Knowledge and molecular scatology in an assessment of mesocarnivore presence, diet and habitat use on Yurok Ancestral Lands.

Partnerships between Tribes and researchers in wildlife monitoring and application of Traditional Ecological Knowledge (TEK) have taken a variety of forms, and some scholars have noted a need for culturally sensitive approaches. Guided by Indigenous Research Methodologies, this research is coupled with Yurok TEK, or hlkelonah 'ue-megetohl ('to take care of the earth'), enabling an applied, culturally sensitive approach in partnership with the Yurok Tribe. We present results from a molecular scatology study of wildlife within the ancestral territory of the Yurok Tribe. Scats were collected opportunistically on road transects. All samples (N = 132) were analysed via DNA barcoding and results matched to documented 'Oohl 'we-toh (Yurok language) names to determine the depositor species (N = 8). Though there were four focal mesocarnivore species in our study, only bobcat (Chmuuek; Lynx rufus) and gray fox (Wergers; Urocyon cinereoargenteus) were detected as depositor species. Post hoc analyses were conducted to explore distribution, habitat use and selection in a use-availability context, and food habits of these two species. We found almost complete separation of bobcat and gray fox use of transects, as well as indication of partitioning of vegetation cover types and food. We demonstrate an integrated framework of Western and Indigenous sciences that allows the Indigenous researcher to transcend structured academic disciplinary boundaries. Our approach can be modified for partnerships between Tribes, agencies, academics and students for wildlife monitoring in broader geographic regions in various research applications.

UAV equipped with infrared imaging for Cervidae monitoring: Improving detection accuracy by eliminating background information interference

Wild Cervidae(deer and their relatives) play a crucial role in maintaining ecological balance and are integral components of ecosystems. However, factors such as environmental changes and poaching behaviors have resulted in habitat degradation for Cervidae. The protection of wild Cervidae has become urgent, and Cervidae monitoring is one of the key means to ensure the effectiveness of wild Cervidae protection. Object detection algorithms based on deep learning offer promising potential for automatically detecting and identifying animals. However, when those algorithms are used for inference in unseen background environments, there will be a significant decrease in accuracy, especially in the situation that a certain type of Cervidae images are collected from single scene for algorithm training. In this paper, a two-stage localization and classification pipeline for Cervidae monitoring is proposed. The pipeline effectively reduces background interference in Cervidae monitoring and enhances monitoring accuracy. In the first stage, the YOLOv7 network is designed to automatically locate Cervidae in UAV infrared images, while implementing improved bounding box regression through the α-IoU loss function enables the network to locate Cervidae more accurately. Then, Cevdidae objects are extracted to eliminate the background information. In the second stage, a classification network named CA-Hybrid, based on Convolutional Neural Networks(CNN) and Vision Transformer(ViT), as well as Channel Attention Mechanism(CAM) enhances the expression of key features, is constructed to accurately identify Cervidae categories. Experimental results indicate that this method achieves an Average Precision (AP) of 95.9% for Cervidae location and a top-1 accuracy of 77.73% for Cervidae identification. This research contributes to a more comprehensive and accurate monitoring of wild Cervidae, and provides valuable references for subsequent UAV-based wildlife monitoring.

MULTIPLE FLYING OBJECT DETECTION USING MACHINE LEARNING ALGORITHM

Detecting multiple flying objects is a crucial task in various domains such as surveillance, wildlife monitoring, and airspace management. This paper presents an approach to detect multiple flying objects using object detection algorithms. The process involves several key steps, including data collection and annotation, preprocessing, model selection, training, evaluation, deployment, real-time detection, post-processing, and monitoring. Initially, a diverse dataset containing images or videos with various flying objects is gathered, and annotations are added to label each object with its corresponding class and bounding box coordinates. Preprocessing techniques like resizing, normalization, and augmentation are applied to enhance the dataset. Next, a suitable object detection algorithm is selected, considering factors like performance and computational efficiency. Common choices include YOLO model, depending on the specific requirements of the application. The chosen model is trained using the annotated dataset, fine-tuned, and evaluated using metrics like precision, recall, and mean Average Precision (mAP). Upon satisfactory performance, the model is deployed in the desired environment, integrated with appropriate hardware for real-time detection. Post-processing techniques such as non-maximum suppression (NMS) are applied to refine the detected bounding boxes, ensuring accurate identification of multiple flying objects. Regular monitoring and maintenance are conducted to keep the deployed model up-to-date and effective in dynamically changing environments

Environmental Monitoring in Internet of Things (IOT)

Abstract: The Internet of Things (IoT) connects physical objects to the internet, enabling autonomous data collection and analysis. In environmental monitoring, IoT plays a crucial role in gathering real-time data on parameters like air and water quality, weather, wildlife behavior, and urban heat islands. This abstract explores how IoT is used in environmental monitoring, covering standards, device types, and applications. IoT comprises interconnected devices that collect environmental data and transmit it for analysis, facilitating continuous observation and better resource management. Standards like MQTT and communication technologies like Wi-Fi govern IoT implementation. Environmental monitoring involves assessing air and water quality, weather conditions, wildlife behavior, and urban heat islands. Air quality monitoring measures pollutant levels in the atmosphere using IoT-based sensor networks. Water quality monitoring assesses chemical and biological characteristics of water bodies. Weather and climate monitoring observe atmospheric conditions using IoT-based sensor networks. Wildlife monitoring tracks animal movements and behavior using IoT-enabled devices. Urban heat island monitoring assesses temperature distribution in urban areas using IoT-based sensors. IoT revolutionizes environmental monitoring by enabling real-time, remote observation across various domains, contributing to better resource management and sustainability efforts.

Surveillance for highly pathogenic avian influenza A (H5N1) in a raptor rehabilitation center-2022.

An ongoing, severe outbreak of highly pathogenic avian influenza virus (HPAI) A H5N1 clade 2.3.4.4b has been circulating in wild and domestic bird populations throughout the world, reaching North America in 2021. This HPAI outbreak has exhibited unique characteristics when compared to previous outbreaks. The global distribution of disease, prolonged duration, extensive number of species and individual wild birds affected, and the large impact on the global poultry industry have all exceeded historical impacts of previous outbreaks in North America. In this study, we describe the results of HPAI surveillance conducted at The Raptor Center, a wildlife rehabilitation hospital at University of Minnesota (Saint Paul, MN, U.S.A.), from March 28th-December 31, 2022. All wild raptors admitted to the facility were tested for avian influenza viruses using polymerase chain reaction (PCR) testing. All non-negative samples were submitted to the United States Department of Agriculture (USDA) Animal and Plant Health Inspection Service (APHIS) National Veterinary Services Laboratories for confirmatory HPAI testing and genetic sequencing. During the study period, 996 individual birds representing 20 different species were tested for avian influenza, and 213 birds were confirmed HPAI positive. Highly pathogenic avian influenza surveillance conducted at The Raptor Center contributed 75% of the HPAI positive raptor detections within the state of Minnesota, located within the Mississippi flyway, significantly augmenting state wildlife surveillance efforts. The viral genotypes observed in birds sampled at The Raptor Center were representative of what was seen in wild bird surveillance within the Mississippi flyway during the same time frame. Wildlife rehabilitation centers provide an opportune situation to augment disease surveillance at the human, wildlife and domestic animal interface during ongoing infectious disease outbreaks.

Identification of Rare Wildlife in the Field Environment Based on the Improved YOLOv5 Model

Research on wildlife monitoring methods is a crucial tool for the conservation of rare wildlife in China. However, the fact that rare wildlife monitoring images in field scenes are easily affected by complex scene information, poorly illuminated, obscured, and blurred limits their use. This often results in unstable recognition and low accuracy levels. To address this issue, this paper proposes a novel wildlife identification model for rare animals in Giant Panda National Park (GPNP). We redesigned the C3 module of YOLOv5 using NAMAttention and the MemoryEfficientMish activation function to decrease the weight of field scene features. Additionally, we integrated the WIoU boundary loss function to mitigate the influence of low-quality images during training, resulting in the development of the NMW-YOLOv5 model. Our model achieved 97.3% for mAP50 and 83.3% for mAP50:95 in the LoTE-Animal dataset. When comparing the model with some classical YOLO models for the purpose of conducting comparison experiments, it surpasses the current best-performing model by 1.6% for mAP50:95, showcasing a high level of recognition accuracy. In the generalization ability test, the model has a low error rate for most rare wildlife species and is generally able to identify wildlife in the wild environment of the GPNP with greater accuracy. It has been demonstrated that NMW-YOLOv5 significantly enhances wildlife recognition accuracy in field environments by eliminating irrelevant features and extracting deep, effective features. Furthermore, it exhibits strong detection and recognition capabilities for rare wildlife in GPNP field environments. This could offer a new and effective tool for rare wildlife monitoring in GPNP.

Research on Prediction of Illegal Wildlife Trade Based on PCA and Multiple Linear Regression

Illegal wildlife trade is a long-standing issue, exerting negative impacts over national economies, regional security, and even global ecosystems. Governments worldwide and numerous environmental and animal protection organizations have been deeply engaged in addressing this issue for many years. Historically, there has been a lack of a quantifiable model to analyze the impact of implementing wildlife monitoring and conservation projects using drones on illegal wildlife trade.In this paper, we primarily propose an Illegal Wildlife Trade Prediction Model. By integrating four secondary indicators related to wildlife trade with the aforementioned 16 indicators, we reduced them to four primary indicators. The methodological tests confirmed their strong correlation, and multiple regression analysis was used to predict illegal wildlife trade, thereby verifying the project's effectiveness. Innovatively, we established a model to measure the probability of project completion, combining data from model predictions, resulting in an approximate completion probability of 60%. In conducting a project sensitivity analysis, we also developed a new model to perturb the three primary indicators, predicting the project's final outcome under random fluctuation of indicators, with variation rates as low as 0.34%, 0.98%, and 1.22%, indicating the project's stability.

WildCLIP: Scene and Animal Attribute Retrieval from Camera Trap Data with Domain-Adapted Vision-Language Models

Wildlife observation with camera traps has great potential for ethology and ecology, as it gathers data non-invasively in an automated way. However, camera traps produce large amounts of uncurated data, which is time-consuming to annotate. Existing methods to label these data automatically commonly use a fixed pre-defined set of distinctive classes and require many labeled examples per class to be trained. Moreover, the attributes of interest are sometimes rare and difficult to find in large data collections. Large pretrained vision-language models, such as contrastive language image pretraining (CLIP), offer great promises to facilitate the annotation process of camera-trap data. Images can be described with greater detail, the set of classes is not fixed and can be extensible on demand and pretrained models can help to retrieve rare samples. In this work, we explore the potential of CLIP to retrieve images according to environmental and ecological attributes. We create WildCLIP by fine-tuning CLIP on wildlife camera-trap images and to further increase its flexibility, we add an adapter module to better expand to novel attributes in a few-shot manner. We quantify WildCLIP’s performance and show that it can retrieve novel attributes in the Snapshot Serengeti dataset. Our findings outline new opportunities to facilitate annotation processes with complex and multi-attribute captions. The code is available at https://github.com/amathislab/wildclip.

Improved Re-Parameterized Convolution for Wildlife Detection in Neighboring Regions of Southwest China.

To autonomously detect wildlife images captured by camera traps on a platform with limited resources and address challenges such as filtering out photos without optimal objects, as well as classifying and localizing species in photos with objects, we introduce a specialized wildlife object detector tailored for camera traps. This detector is developed using a dataset acquired by the Saola Working Group (SWG) through camera traps deployed in Vietnam and Laos. Utilizing the YOLOv6-N object detection algorithm as its foundation, the detector is enhanced by a tailored optimizer for improved model performance. We deliberately introduce asymmetric convolutional branches to enhance the feature characterization capability of the Backbone network. Additionally, we streamline the Neck and use CIoU loss to improve detection performance. For quantitative deployment, we refine the RepOptimizer to train a pure VGG-style network. Experimental results demonstrate that our proposed method empowers the model to achieve an 88.3% detection accuracy on the wildlife dataset in this paper. This accuracy is 3.1% higher than YOLOv6-N, and surpasses YOLOv7-T and YOLOv8-N by 5.5% and 2.8%, respectively. The model consistently maintains its detection performance even after quantization to the INT8 precision, achieving an inference speed of only 6.15 ms for a single image on the NVIDIA Jetson Xavier NX device. The improvements we introduce excel in tasks related to wildlife image recognition and object localization captured by camera traps, providing practical solutions to enhance wildlife monitoring and facilitate efficient data acquisition. Our current work represents a significant stride toward a fully automated animal observation system in real-time in-field applications.

Wildlife as sentinels for emerging zoonotic diseases: A review of surveillance systems in the USA

Wildlife plays a crucial role as sentinels for emerging zoonotic diseases, serving as early indicators of potential threats to human and animal health. This review examines the surveillance systems in the USA that utilize wildlife as indicators for the presence of emerging zoonotic pathogens. The review focuses on the methodologies, challenges, and implications of these surveillance systems. In the USA, wildlife surveillance for zoonotic diseases is primarily conducted through passive surveillance, where wildlife carcasses are collected and tested for pathogens. Additionally, active surveillance programs target specific wildlife species known to host zoonotic pathogens or species that may be in close contact with humans or domestic animals. These surveillance efforts are complemented by the use of sentinel species, such as birds or bats, which are monitored for signs of disease that could indicate the presence of zoonotic pathogens. Challenges in wildlife surveillance include the vast geographic range of many wildlife species, making it difficult to sample populations comprehensively. Additionally, the diversity of wildlife species and habitats in the USA presents logistical challenges for surveillance efforts. Furthermore, there are challenges related to data sharing and coordination among agencies responsible for wildlife and public health. The implications of wildlife surveillance for public health policy and practice are significant. Early detection of zoonotic pathogens in wildlife can lead to timely public health interventions, such as vaccination campaigns or changes in land use practices to reduce human-wildlife contact. Moreover, wildlife surveillance can inform the development of predictive models for disease outbreaks, enabling more effective preparedness and response measures. In conclusion, wildlife surveillance plays a critical role in the early detection and monitoring of emerging zoonotic diseases in the USA. Continued investment in surveillance systems and research is essential to enhance our understanding of zoonotic disease dynamics and improve our ability to protect human and animal health.

Monitoring wildlife abundance through track surveys: A capture-mark-recapture inspired approach to assess track detection by certified trackers in the Kalahari, Botswana

The surveying of animal tracks is a cost-effective wildlife monitoring technique in conservation, provided the substrate is suitable for tracks to imprint and observers are skilled enough to detect and identify them. However, as with every method quantifying population abundance, it can be biased if track detectability is not accounted for. Track detectability has two components: probability p1ˆ for an animal movement path to intersect with the sampling unit at least once and probability p2ˆ of track detection. Here, we measured p2ˆ, often overlooked, and tested to what extent it is affected by observers and sampling conditions, by applying a capture-mark-recapture approach to a driven track survey of 12 wildlife species in the Kalahari, Botswana. We also characterized and quantified disagreements between pairs of observers on the identification and ageing of tracks. Except for fresh (≤24 old) tracks being slightly more detectable (p2ˆ= 0.77 vs 0.61) and reliably identified (probability of agreement in species identification = 0.92 vs 0.87) than older tracks, we demonstrated that p2ˆ is high and most importantly largely constant across sampling conditions and highly skilled observer pairs. Species identification mismatches between observer pairs were rare (8% for fresh tracks) and limited to specific couples of close species, materializing into confusion rather than directional bias. Where p1ˆ can be accounted for in the derivation of population abundance from track counts (e.g. Formozov–Malyshev–Pereleshin formula), our results relative to p2ˆ validate the use of track surveys for wildlife monitoring in the Kalahari. Our conclusion is potentially applicable to a large range of environments across the globe, especially where observer skills are present among local communities, and conservation can benefit both wildlife and people, fostering their coexistence.

BirdCollect: A Comprehensive Benchmark for Analyzing Dense Bird Flock Attributes

Automatic recognition of bird behavior from long-term, un controlled outdoor imagery can contribute to conservation efforts by enabling large-scale monitoring of bird populations. Current techniques in AI-based wildlife monitoring have focused on short-term tracking and monitoring birds individually rather than in species-rich flocks. We present Bird-Collect, a comprehensive benchmark dataset for monitoring dense bird flock attributes. It includes a unique collection of more than 6,000 high-resolution images of Demoiselle Cranes (Anthropoides virgo) feeding and nesting in the vicinity of Khichan region of Rajasthan. Particularly, each image contains an average of 190 individual birds, illustrating the complex dynamics of densely populated bird flocks on a scale that has not previously been studied. In addition, a total of 433 distinct pictures captured at Keoladeo National Park, Bharatpur provide a comprehensive representation of 34 distinct bird species belonging to various taxonomic groups. These images offer details into the diversity and the behaviour of birds in vital natural ecosystem along the migratory flyways. Additionally, we provide a set of 2,500 point-annotated samples which serve as ground truth for benchmarking various computer vision tasks like crowd counting, density estimation, segmentation, and species classification. The benchmark performance for these tasks highlight the need for tailored approaches for specific wildlife applications, which include varied conditions including views, illumination, and resolutions. With around 46.2 GBs in size encompassing data collected from two distinct nesting ground sets, it is the largest birds dataset containing detailed annotations, showcasing a substantial leap in bird research possibilities. We intend to publicly release the dataset to the research community. The database is available at: https://iab-rubric.org/resources/wildlife-dataset/birdcollect

Insights from 20 years of mammal population research in Indonesia

AbstractMammal populations are declining in biodiverse tropical regions. Global analyses have identified Indonesia as a hotspot of vertebrate decline, although relatively few data are available to substantiate these claims. We reviewed research articles published during 2000–2020 on 104 medium-sized to large terrestrial mammal species found in Indonesia to help inform conservation management and future research. We identified 308 peer-reviewed studies published in English or Bahasa Indonesia, with an increase in publication rate (articles published per year) over time. Studies of species distributions dominated the literature, followed by publications on abundance, species diversity and combinations of these topics. Most publications concerned single-species studies conducted at a single location and a single point in time. We identify four key issues that should be addressed by future research and conservation efforts: (1) disproportionate focus on a small number of species; (2) geographical bias towards west Indonesia (Sumatra, Kalimantan and Java–Bali), with few published studies from central (Sulawesi, Nusa Tenggara and Maluku) and east (Papua) Indonesia; (3) limitations to survey design, sampling effort and data analysis; and (4) lack of long-term wildlife population studies. We also note challenges local researchers face in publishing their studies in international journals because of language barriers and costs. Greater use of existing biodiversity data and continued capacity building for local researchers, particularly those in central and east Indonesia, are critical to effectively guide future wildlife monitoring and improve the conservation status of Indonesian mammals.