Perspectives in machine learning for wildlife conservation

Evaluating the effectiveness of mitigation measures in environmental impact assessments: A comprehensive review of development projects in Korea

Rapid urbanization and development projects in Korea have posed significant threats to biodiversity; thus, effective mitigation measures are required to preserve natural habitats. Nevertheless, the factors underlying variations in mitigation measure effectiveness according to the disturbance level and surrounding environmental conditions have not been clarified. This study evaluated the effectiveness of mitigation measures implemented in environmental impact assessments (EIAs) of development projects in Korea, with a focus on their effectiveness with respect to the disturbance level and surrounding environmental conditions. A review of 288 EIA reports from selected projects that implemented all 10 mitigation measures classified according to the Wildlife Conservation Comprehensive Plan was conducted. Using the biodiversity tipping point framework, the effects of mitigation measures on biodiversity were categorized into four levels and analyzed. Analysis of variance and redundancy analysis were then performed to discern the variance in mitigation measure effectiveness in terms of the disturbance level, surrounding environment, and species. The results revealed significant variations in the effectiveness of mitigation measures depending on the surrounding environment and disturbance level. Linear projects exhibited a clear impact on various species as the disturbance level increased, whereas area-based projects did not exhibit such pronounced effects. All species demonstrated a negative relationship with development duration, development area, and distance from urban centers. Notably, avian and amphibian species showed a strong negative correlation with the digital elevation model while reptiles and mammals exhibited a strong positive relationship with pre-development biodiversity and distance from protected areas, respectively. Mitigation measures play a key role in alleviating the adverse effects of development projects; therefore, our findings indicate the need for spatially tailored mitigation plans to augment their effectiveness.

Data from three camera trapping pilots in the Amsterdam Water Supply Dunes of the Netherlands

This paper presents the data (images, observations, metadata) of three different deployments of camera traps in the Amsterdam Water Supply Dunes, a Natura 2000 nature reserve in the coastal dunes of the Netherlands. The pilots were aimed at determining how different types of camera deployment (e.g. regular vs. wide lens, various heights, inside/outside exclosures) might influence species detections, and how to deploy autonomous wildlife monitoring networks. Two pilots were conducted in herbivore exclosures and mainly detected European rabbits (Oryctolagus cuniculus) and red fox (Vulpes vulpes). The third pilot was conducted outside exclosures, with the European fallow deer (Dama dama) being most prevalent. Across all three pilots, a total of 47,597 images were annotated using the Agouti platform. All annotations were verified and quality-checked by a human expert. A total of 2,779 observations of 20 different species (including humans) were observed using 11 wildlife cameras during 2021-2023. The raw image files (excluding humans), image metadata, deployment metadata and observations from each pilot are shared using the Camtrap DP open standard and the extended data publishing capabilities of GBIF to increase the findability, accessibility, interoperability, and reusability of this data. The data are freely available and can be used for developing artificial intelligence (AI) algorithms that automatically detect and identify species from wildlife camera images.

Applying XGBoost and SHAP to Open Source Data to Identify Key Drivers and Predict Likelihood of Wolf Pair Presence

Wolves have returned to Germany since 2000. Numbers have grown to 209 territorial pairs in 2021. XGBoost machine learning, combined with SHAP analysis is applied to predict German wolf pair presence in 2022 for 10 × 10 km grid cells. Model input consisted of 38 variables from open sources, covering the period 2000 to 2021. The XGBoost model predicted well, with 0.91 as the AUC. SHAP analysis ranked the variables: distance to the closest neighboring wolf pair was the main driver for a grid cell to become occupied by a wolf pair. The clustering tendency of related wolves seems to be an important explanatory factor here. Second was the percentage of wooded area. The next eight variables related to wolf presence in the preceding year, except at fifth, eighth and tenth position in the total order: human density (square root) in the grid, percentage arable land and road density respectively. Other variables including the occurrence of wild prey were the weakest predictors. The SHAP analysis also provided crucial added value in identifying a variable that had threshold values where its contribution to the prediction changed from positive to negative or vice versa. For instance, low density of people increased the probability of wolf pair presence, whereas a high density decreased this probability. Cumulative lift techniques showed that the model performed almost four times better than random prediction. The combination of XGBoost, SHAP and cumulative lift techniques is new in wolf management and conservation, allowing for the focusing of educational and financial resources.

Improving prediction of N2O emissions during composting using model-agnostic meta-learning

Nitrous oxide (N2O) represents a significant environmental challenge as a harmful, long-lived greenhouse gas that contributes to the depletion of stratospheric ozone and exacerbates global anthropogenic greenhouse warming. Composting is considered a promising and economically feasible strategy for the treatment of organic waste. However, recent research indicates that composting is a source of N2O, contributing to atmospheric pollution and greenhouse effect. Consequently, there is a need for the development of effective, cost-efficient methodologies to quantify N2O emissions accurately. In this study, we employed the model-agnostic meta-learning (MAML) method to improve the performance of N2O emissions prediction during manure composting. The highest R2 and lowest root mean squared error (RMSE) values achieved were 0.939 and 18.42 mg d-1, respectively. Five machine learning methods including the backpropagation neural network, extreme learning machine, integrated machine learning method based on ELM and random forest, gradient boosting decision tree, and extreme gradient boosting were adopted for comparison to further demonstrate the effectiveness of the MAML prediction model. Feature analysis showed that moisture content of structure material and ammonium concentration during composting process were the two most significant features affecting N2O emissions. This study serves as proof of the application of MAML during N2O emissions prediction, further giving new insights into the effects of manure material properties and composting process data on N2O emissions. This approach helps determining the strategies for mitigating N2O emissions.

Insect detect: An open-source DIY camera trap for automated insect monitoring

Insect monitoring is essential to design effective conservation strategies, which are indispensable to mitigate worldwide declines and biodiversity loss. For this purpose, traditional monitoring methods are widely established and can provide data with a high taxonomic resolution. However, processing of captured insect samples is often time-consuming and expensive, which limits the number of potential replicates. Automated monitoring methods can facilitate data collection at a higher spatiotemporal resolution with a comparatively lower effort and cost. Here, we present the Insect Detect DIY (do-it-yourself) camera trap for non-invasive automated monitoring of flower-visiting insects, which is based on low-cost off-the-shelf hardware components combined with open-source software. Custom trained deep learning models detect and track insects landing on an artificial flower platform in real time on-device and subsequently classify the cropped detections on a local computer. Field deployment of the solar-powered camera trap confirmed its resistance to high temperatures and humidity, which enables autonomous deployment during a whole season. On-device detection and tracking can estimate insect activity/abundance after metadata post-processing. Our insect classification model achieved a high top-1 accuracy on the test dataset and generalized well on a real-world dataset with captured insect images. The camera trap design and open-source software are highly customizable and can be adapted to different use cases. With custom trained detection and classification models, as well as accessible software programming, many possible applications surpassing our proposed deployment method can be realized.

An integrated passive acoustic monitoring and deep learning pipeline for black-and-white ruffed lemurs (Varecia variegata) in Ranomafana National Park, Madagascar

The urgent need for effective wildlife monitoring solutions in the face of global biodiversity loss has resulted in the emergence of conservation technologies such as passive acoustic monitoring (PAM). While PAM has been extensively used for marine mammals, birds, and bats, its application to primates is limited. Black-and-white ruffed lemurs (Varecia variegata) are a promising species to test PAM with due to their distinctive and loud roar-shrieks. Furthermore, these lemurs are challenging to monitor via traditional methods due to their fragmented and often unpredictable distribution in Madagascar's dense eastern rainforests. Our goal in this study was to develop a machine learning pipeline for automated call detection from PAM data, compare the effectiveness of PAM versus in-person observations, and investigate diel patterns in lemur vocal behavior. We did this study at Mangevo, Ranomafana National Park by concurrently conducting focal follows and deploying autonomous recorders in May-July 2019. We used transfer learning to build a convolutional neural network (optimized for recall) that automated the detection of lemur calls (57-h runtime; recall = 0.94, F1 = 0.70). We found that PAM outperformed in-person observations, saving time, money, and labor while also providing re-analyzable data. Using PAM yielded novel insights into V. variegata diel vocal patterns; we present the first published evidence of nocturnal calling. We developed a graphic user interface and open-sourced data and code, to serve as a resource for primatologists interested in implementing PAM and machine learning. By leveraging the potential of this pipeline, we can address the urgent need for effective primate population surveys to inform conservation strategies.

Automatedly identify dryland threatened species at large scale by using deep learning

Dryland biodiversity is decreasing at an alarming rate. Advanced intelligent tools are urgently needed to rapidly, automatedly, and precisely detect dryland threatened species on a large scale for biological conservation. Here, we explored the performance of three deep convolutional neural networks (Deeplabv3+, Unet, and Pspnet models) on the intelligent recognition of rare species based on high-resolution (0.3 m) satellite images taken by an unmanned aerial vehicle (UAV). We focused on a threatened species, Populus euphratica, in the Tarim River Basin (China), where there has been a severe population decline in the 1970s and restoration has been carried out since 2000. The testing results showed that Unet outperforms Deeplabv3+ and Pspnet when the training samples are lower, while Deeplabv3+ performs best as the dataset increases. Overall, when training samples are 80, Deeplabv3+ had the best overall performance for Populus euphratica identification, with mean pixel accuracy (MPA) between 87.31 % and 90.2 %, which, on average is 3.74 % and 11.29 % higher than Unet and Pspnet, respectively. Deeplabv3+ can accurately detect the boundaries of Populus euphratica even in areas of dense vegetation, with lower identification uncertainty for each pixel than other models. This study developed a UAV imagery-based identification framework using deep learning with high resolution in large-scale regions. This approach can accurately capture the variation in dryland threatened species, especially those in inaccessible areas, thereby fostering rapid and efficient conservation actions.

Propagating variational model uncertainty for bioacoustic call label smoothing

Along with propagating the input toward making a prediction, Bayesian neural networks also propagate uncertainty. This has the potential to guide the training process by rejecting predictions of low confidence, and recent variational Bayesian methods can do so without Monte Carlo sampling of weights. Here, we apply sample-free methods for wildlife call detection on recordings made via passive acoustic monitoring equipment in the animals' natural habitats. We further propose uncertainty-aware label smoothing, where the smoothing probability is dependent on sample-free predictive uncertainty, in order to downweigh data samples that should contribute less to the loss value. We introduce a bioacoustic dataset recorded in Malaysian Borneo, containing overlapping calls from 30 species. On that dataset, our proposed method achieves an absolute percentage improvement of around 1.5 points on area under the receiver operating characteristic (AU-ROC), 13 points in F1, and 19.5 points in expected calibration error (ECE) compared to the point-estimate network baseline averaged across all target classes.

Novel community data in ecology-properties and prospects

New technologies for monitoring biodiversity such as environmental (e)DNA, passive acoustic monitoring, and optical sensors promise to generate automated spatiotemporal community observations at unprecedented scales and resolutions. Here, we introduce 'novel community data' as an umbrella term for these data. We review the emerging field around novel community data, focusing on new ecological questions that could be addressed; the analytical tools available or needed to make best use of these data; and the potential implications of these developments for policy and conservation. We conclude that novel community data offer many opportunities to advance our understanding of fundamental ecological processes, including community assembly, biotic interactions, micro- and macroevolution, and overall ecosystem functioning.

Embracing firefly flash pattern variability with data-driven species classification

Many nocturnally active fireflies use precisely timed bioluminescent patterns to identify mates, making them especially vulnerable to light pollution. As urbanization continues to brighten the night sky, firefly populations are under constant stress, and close to half of the species are now threatened. Ensuring the survival of firefly biodiversity depends on a large-scale conservation effort to monitor and protect thousands of populations. While species can be identified by their flash patterns, current methods require expert measurement and manual classification and are infeasible given the number and geographic distribution of fireflies. Here we present the application of a recurrent neural network (RNN) for accurate automated firefly flash pattern classification. Using recordings from commodity cameras, we can extract flash trajectories of individuals within a swarm and classify their species with an accuracy of approximately seventy percent. In addition to its potential in population monitoring, automated classification provides the means to study firefly behavior at the population level. We employ the classifier to measure and characterize the variability within and between swarms, unlocking a new dimension of their behavior. Our method is open source, and deployment in community science applications could revolutionize our ability to monitor and understand firefly populations.

Stress-induced loss of social resilience in honeybee colonies and its implications on fitness

Stressors may lead to a shift in the timing of life-history events of species, causing a mismatch with optimal environmental conditions, potentially reducing fitness. In honeybees, the timing of brood rearing and nest emergence in late winter/early spring is critical as colonies need to grow fast after winter to prepare for reproduction. However, the effects of stress on these life-history events in late winter/early spring and the possible consequences are not well understood. Therefore, we tested whether (i) honeybee colonies shift timing of brood rearing and nest emergence as response to stressors, and (ii) if there is a consequent loss of social resilience, reflected in colony fitness (survival, growth and reproduction). We monitored stressed (high load of the parasitic mite Varroa destructor or nutrition restricted) colonies and presumably non-stressed colonies from the beginning of 2020 till spring of 2021. We found that honeybee colonies do not shift the timing of brood rearing and nest emergence in spring as a coping mechanism to stressors. However, we show that there is loss of social resilience in stressed colonies, leading to reduced growth and reproduction. Our study contributes to better understanding the effects of stressors on social resilience in eusocial organisms.

A landmark environmental law looks ahead

In late December 1973, the United States enacted what some would come to call "the pitbull of environmental laws." In the 50 years since, the formidable regulatory teeth of the Endangered Species Act (ESA) have been credited with considerable successes, obliging agencies to draw upon the best available science to protect species and habitats. Yet human pressures continue to push the planet toward extinctions on a massive scale. With that prospect looming, and with scientific understanding ever changing, Science invited experts to discuss how the ESA has evolved and what its future might hold. -Brad Wible.

Deep learning in terrestrial conservation biology

Biodiversity is being lost at an unprecedented rate on Earth. As a first step to more effectively combat this process we need efficient methods to monitor biodiversity changes. Recent technological advance can provide powerful tools (e.g. camera traps, digital acoustic recorders, satellite imagery, social media records) that can speed up the collection of biological data. Nevertheless, the processing steps of the raw data served by these tools are still painstakingly slow. A new computer technology, deep learning based artificial intelligence, might, however, help. In this short and subjective review I oversee recent technological advances used in conservation biology, highlight problems of processing their data, shortly describe deep learning technology and show case studies of its use in conservation biology. Some of the limitations of the technology are also highlighted.

Sensory collectives in natural systems

Groups of animals inhabit vastly different sensory worlds, or umwelten, which shape fundamental aspects of their behaviour. Yet the sensory ecology of species is rarely incorporated into the emerging field of collective behaviour, which studies the movements, population-level behaviours, and emergent properties of animal groups. Here, we review the contributions of sensory ecology and collective behaviour to understanding how animals move and interact within the context of their social and physical environments. Our goal is to advance and bridge these two areas of inquiry and highlight the potential for their creative integration. To achieve this goal, we organise our review around the following themes: (1) identifying the promise of integrating collective behaviour and sensory ecology; (2) defining and exploring the concept of a 'sensory collective'; (3) considering the potential for sensory collectives to shape the evolution of sensory systems; (4) exploring examples from diverse taxa to illustrate neural circuits involved in sensing and collective behaviour; and (5) suggesting the need for creative conceptual and methodological advances to quantify 'sensescapes'. In the final section, (6) applications to biological conservation, we argue that these topics are timely, given the ongoing anthropogenic changes to sensory stimuli (e.g. via light, sound, and chemical pollution) which are anticipated to impact animal collectives and group-level behaviour and, in turn, ecosystem composition and function. Our synthesis seeks to provide a forward-looking perspective on how sensory ecologists and collective behaviourists can both learn from and inspire one another to advance our understanding of animal behaviour, ecology, adaptation, and evolution.

A dataset for benchmarking Neotropical anuran calls identification in passive acoustic monitoring

Global change is predicted to induce shifts in anuran acoustic behavior, which can be studied through passive acoustic monitoring (PAM). Understanding changes in calling behavior requires automatic identification of anuran species, which is challenging due to the particular characteristics of neotropical soundscapes. In this paper, we introduce a large-scale multi-species dataset of anuran amphibians calls recorded by PAM, that comprises 27 hours of expert annotations for 42 different species from two Brazilian biomes. We provide open access to the dataset, including the raw recordings, experimental setup code, and a benchmark with a baseline model of the fine-grained categorization problem. Additionally, we highlight the challenges of the dataset to encourage machine learning researchers to solve the problem of anuran call identification towards conservation policy. All our experiments and resources have been made available at https://soundclim.github.io/anuraweb/ .

Identification of hybrids between the Japanese giant salamander (Andrias japonicus) and Chinese giant salamander (Andrias cf. davidianus) using deep learning and smartphone images

Human-mediated hybridization between native and non-native species is causing biodiversity loss worldwide. Hybridization has contributed to the extinction of many species through direct and indirect processes such as loss of reproductive opportunity and genetic introgression. Therefore, it is essential to manage hybrids to conserve biodiversity. However, specialized knowledge is required to identify the target species based on visual characteristics when two species have similar features. Although image recognition technology can be a powerful tool for identifying hybrids, studies have yet to utilize deep learning approaches. Hence, this study aimed to identify hybrids between the native Japanese giant salamander (Andrias japonicus) and the non-native Chinese giant salamander (Andrias cf. davidianus) using EfficientNetV2 and smartphone images. We used smartphone images of 11 individuals of native A. japonicus (five training and six test images) and 20 individuals of hybrids between A. japonicus and A. cf. davidianus (five training and 15 test images). In our experimental environment, an AI model constructed with EfficientNetV2 exhibited 100% accuracy in identifying hybrids. In addition, gradient-weighted class activation mapping revealed that the AI model was able to classify A. japonicus and hybrids between A. japonicus and A. cf. davidianus on the basis of the dorsal head spot patterning. Our approach thus enables the identification of hybrids against A. japonicus, which was previously considered difficult by non-experts. Furthermore, since this study achieved reliable identification using smartphone images, it is expected to be applied to a wide range of citizen science projects.

Information Ecology: an integrative framework for studying animal behavior

Information is simultaneously a valuable resource for animals and a tractable variable for researchers. We propose the name Information Ecology to describe research focused on how individual animals use information to enhance fitness. An explicit focus on information in animal behavior is far from novel - we simply build on these ideas and promote a unified approach to how and why animals use information. The value of information to animals favors the theoretically rich adaptive approach of field-based research. Simultaneously, our ability to manipulate information lends itself to the strong methods of laboratory-based research. Information Ecology asks three questions: What information is available? How is it used (or not)? And, why is it used (or not)?

A high-resolution canopy height model of the Earth

The worldwide variation in vegetation height is fundamental to the global carbon cycle and central to the functioning of ecosystems and their biodiversity. Geospatially explicit and, ideally, highly resolved information is required to manage terrestrial ecosystems, mitigate climate change and prevent biodiversity loss. Here we present a comprehensive global canopy height map at 10 m ground sampling distance for the year 2020. We have developed a probabilistic deep learning model that fuses sparse height data from the Global Ecosystem Dynamics Investigation (GEDI) space-borne LiDAR mission with dense optical satellite images from Sentinel-2. This model retrieves canopy-top height from Sentinel-2 images anywhere on Earth and quantifies the uncertainty in these estimates. Our approach improves the retrieval of tall canopies with typically high carbon stocks. According to our map, only 5% of the global landmass is covered by trees taller than 30 m. Further, we find that only 34% of these tall canopies are located within protected areas. Thus, the approach can serve ongoing efforts in forest conservation and has the potential to foster advances in climate, carbon and biodiversity modelling.

Elephants and algorithms: a review of the current and future role of AI in elephant monitoring

Artificial intelligence (AI) and machine learning (ML) present revolutionary opportunities to enhance our understanding of animal behaviour and conservation strategies. Using elephants, a crucial species in Africa and Asia's protected areas, as our focal point, we delve into the role of AI and ML in their conservation. Given the increasing amounts of data gathered from a variety of sensors like cameras, microphones, geophones, drones and satellites, the challenge lies in managing and interpreting this vast data. New AI and ML techniques offer solutions to streamline this process, helping us extract vital information that might otherwise be overlooked. This paper focuses on the different AI-driven monitoring methods and their potential for improving elephant conservation. Collaborative efforts between AI experts and ecological researchers are essential in leveraging these innovative technologies for enhanced wildlife conservation, setting a precedent for numerous other species.

DJAN: Deep Joint Adaptation Network for Wildlife Image Recognition

Wildlife recognition is of utmost importance for monitoring and preserving biodiversity. In recent years, deep-learning-based methods for wildlife image recognition have exhibited remarkable performance on specific datasets and are becoming a mainstream research direction. However, wildlife image recognition tasks face the challenge of weak generalization in open environments. In this paper, a Deep Joint Adaptation Network (DJAN) for wildlife image recognition is proposed to deal with the above issue by taking a transfer learning paradigm into consideration. To alleviate the distribution discrepancy between the known dataset and the target task dataset while enhancing the transferability of the model's generated features, we introduce a correlation alignment constraint and a strategy of conditional adversarial training, which enhance the capability of individual domain adaptation modules. In addition, a transformer unit is utilized to capture the long-range relationships between the local and global feature representations, which facilitates better understanding of the overall structure and relationships within the image. The proposed approach is evaluated on a wildlife dataset; a series of experimental results testify that the DJAN model yields state-of-the-art results, and, compared to the best results obtained by the baseline methods, the average accuracy of identifying the eleven wildlife species improves by 3.6 percentage points.

Bioinspired robots can foster nature conservation

We live in a time of unprecedented scientific and human progress while being increasingly aware of its negative impacts on our planet's health. Aerial, terrestrial, and aquatic ecosystems have significantly declined putting us on course to a sixth mass extinction event. Nonetheless, the advances made in science, engineering, and technology have given us the opportunity to reverse some of our ecosystem damage and preserve them through conservation efforts around the world. However, current conservation efforts are primarily human led with assistance from conventional robotic systems which limit their scope and effectiveness, along with negatively impacting the surroundings. In this perspective, we present the field of bioinspired robotics to develop versatile agents for future conservation efforts that can operate in the natural environment while minimizing the disturbance/impact to its inhabitants and the environment's natural state. We provide an operational and environmental framework that should be considered while developing bioinspired robots for conservation. These considerations go beyond addressing the challenges of human-led conservation efforts and leverage the advancements in the field of materials, intelligence, and energy harvesting, to make bioinspired robots move and sense like animals. In doing so, it makes bioinspired robots an attractive, non-invasive, sustainable, and effective conservation tool for exploration, data collection, intervention, and maintenance tasks. Finally, we discuss the development of bioinspired robots in the context of collaboration, practicality, and applicability that would ensure their further development and widespread use to protect and preserve our natural world.

Monitoring Endangered and Rare Wildlife in the Field: A Foundation Deep Learning Model Integrating Human Knowledge for Incremental Recognition with Few Data and Low Cost

Intelligent monitoring of endangered and rare wildlife is important for biodiversity conservation. In practical monitoring, few animal data are available to train recognition algorithms. The system must, therefore, achieve high accuracy with limited resources. Simultaneously, zoologists expect the system to be able to discover unknown species to make significant discoveries. To date, none of the current algorithms have these abilities. Therefore, this paper proposed a KI-CLIP method. Firstly, by first introducing CLIP, a foundation deep learning model that has not yet been applied in animal fields, the powerful recognition capability with few training resources is exploited with an additional shallow network. Secondly, inspired by the single-image recognition abilities of zoologists, we incorporate easily accessible expert description texts to improve performance with few samples. Finally, a simple incremental learning module is designed to detect unknown species. We conducted extensive comparative experiments, ablation experiments, and case studies on 12 datasets containing real data. The results validate the effectiveness of KI-CLIP, which can be trained on multiple real scenarios in seconds, achieving in our study over 90% recognition accuracy with only 8 training samples, and over 97% with 16 training samples. In conclusion, KI-CLIP is suitable for practical animal monitoring.

Uniting Experiments and Big Data to advance ecology and conservation

Many ecologists increasingly advocate for research frameworks centered on the use of 'big data' to address anthropogenic impacts on ecosystems. Yet, experiments are often considered essential for identifying mechanisms and informing conservation interventions. We highlight the complementarity of these research frameworks and expose largely untapped opportunities for combining them to speed advancements in ecology and conservation. With nascent but increasing application of model integration, we argue that there is an urgent need to unite experimental and big data frameworks throughout the scientific process. Such an integrated framework offers potential for capitalizing on the benefits of both frameworks to gain rapid and reliable answers to ecological challenges.

Digital twins: dynamic model-data fusion for ecology

Digital twins (DTs) are an emerging phenomenon in the public and private sectors as a new tool to monitor and understand systems and processes. DTs have the potential to change the status quo in ecology as part of its digital transformation. However, it is important to avoid misguided developments by managing expectations about DTs. We stress that DTs are not just big models of everything, containing big data and machine learning. Rather, the strength of DTs is in combining data, models, and domain knowledge, and their continuous alignment with the real world. We suggest that researchers and stakeholders exercise caution in DT development, keeping in mind that many of the strengths and challenges of computational modelling in ecology also apply to DTs.

Systematic review of machine learning methods applied to ecoacoustics and soundscape monitoring

Soundscape ecology is a promising area that studies landscape patterns based on their acoustic composition. It focuses on the distribution of biotic and abiotic sounds at different frequencies of the landscape acoustic attribute and the relationship of said sounds with ecosystem health metrics and indicators (e.g., species richness, acoustic biodiversity, vectors of structural change, gradients of vegetation cover, landscape connectivity, and temporal and spatial characteristics). To conduct such studies, researchers analyze recordings from Acoustic Recording Units (ARUs). The increasing use of ARUs and their capacity to record hours of audio for months at a time have created a need for automatic processing methods to reduce time consumption, correlate variables implicit in the recordings, extract features, and characterize sound patterns related to landscape attributes. Consequently, traditional machine learning methods have been commonly used to process data on different characteristics of soundscapes, mainly the presence-absence of species. In addition, it has been employed for call segmentation, species identification, and sound source clustering. However, some authors highlight the importance of the new approaches that use unsupervised deep learning methods to improve the results and diversify the assessed attributes. In this paper, we present a systematic review of machine learning methods used in the field of ecoacoustics for data processing. It includes recent trends, such as semi-supervised and unsupervised deep learning methods. Moreover, it maintains the format found in the reviewed papers. First, we describe the ARUs employed in the papers analyzed, their configuration, and the study sites where the datasets were collected. Then, we provide an ecological justification that relates acoustic monitoring to landscape features. Subsequently, we explain the machine learning methods followed to assess various landscape attributes. The results show a trend towards label-free methods that can process the large volumes of data gathered in recent years. Finally, we discuss the need to adopt methods with a machine learning approach in other biological dimensions of landscapes.

Progress toward the sustainable development of world cultural heritage sites facing land-cover changes

The quantification of the extent and dynamics of land-use changes is a key metric employed to assess the progress toward several Sustainable Development Goals (SDGs) that form part of the United Nations 2030 Sustainable Development Agenda. In terms of anthropogenic factors threatening the conservation of heritage properties, such a metric aids in the assessment of achievements toward heritage sustainability solving the problem of insufficient data availability. Therefore, in this study, 589 cultural World Heritage List (WHL) properties from 115 countries were analyzed, encompassing globally distributed and statistically significant samples of "monuments and groups of buildings" (73.2%), "sites" (19.3%), and "cultural landscapes" (7.5%). Land-cover changes in the WHL properties between 2015 and 2020 were automatically extracted from big data collections of high-resolution satellite imagery accessed via Google Earth Engine using intelligent remote sensing classification. Sustainability indexes (SIs) were estimated for the protection zones of each property, and the results were employed, for the first time, to assess the progress of each country toward SDG Target 11.4. Despite the apparent advances in SIs (10.4%), most countries either exhibited steady (20.0%) or declining (69.6%) SIs due to limited cultural investigations and enhanced negative anthropogenic disturbances. This study confirms that land-cover changes are among serious threats for heritage conservation, with heritage in some countries wherein the need to address this threat is most crucial, and the proposed spatiotemporal monitoring approach is recommended.

The Internet of Animals: what it is, what it could be

One of the biggest trends in ecology over the past decade has been the creation of standardized databases. Recently, this has included live data, formal linkages between disparate databases, and automated analytics, a synergy that we recognize as the Internet of Animals (IoA). Early IoA systems relate animal locations to remote-sensing data to predict species distributions and detect disease outbreaks, and use live data to inform management of endangered species. However, meeting the future potential of the IoA concept will require solving challenges of taxonomy, data security, and data sharing. By linking data sets, integrating live data, and automating workflows, the IoA has the potential to enable discoveries and predictions relevant to human societies and the conservation of animals.

SMART-BARN: Scalable multimodal arena for real-time tracking behavior of animals in large numbers

The SMART-BARN (scalable multimodal arena for real-time tracking behavior of animals in large numbers) achieves fast, robust acquisition of movement, behavior, communication, and interactions of animals in groups, within a large (14.7 meters by 6.6 meters by 3.8 meters), three-dimensional environment using multiple information channels. Behavior is measured from a wide range of taxa (insects, birds, mammals, etc.) and body size (from moths to humans) simultaneously. This system integrates multiple, concurrent measurement techniques including submillimeter precision and high-speed (300 hertz) motion capture, acoustic recording and localization, automated behavioral recognition (computer vision), and remote computer-controlled interactive units (e.g., automated feeders and animal-borne devices). The data streams are available in real time allowing highly controlled and behavior-dependent closed-loop experiments, while producing comprehensive datasets for offline analysis. The diverse capabilities of SMART-BARN are demonstrated through three challenging avian case studies, while highlighting its broad applicability to the fine-scale analysis of collective animal behavior across species.

Real-time biodiversity analysis using deep-learning algorithms on mobile robotic platforms

Ecological biodiversity is declining at an unprecedented rate. To combat such irreversible changes in natural ecosystems, biodiversity conservation initiatives are being conducted globally. However, the lack of a feasible methodology to quantify biodiversity in real-time and investigate population dynamics in spatiotemporal scales prevents the use of ecological data in environmental planning. Traditionally, ecological studies rely on the census of an animal population by the "capture, mark and recapture" technique. In this technique, human field workers manually count, tag and observe tagged individuals, making it time-consuming, expensive, and cumbersome to patrol the entire area. Recent research has also demonstrated the potential for inexpensive and accessible sensors for ecological data monitoring. However, stationary sensors collect localised data which is highly specific on the placement of the setup. In this research, we propose the methodology for biodiversity monitoring utilising state-of-the-art deep learning (DL) methods operating in real-time on sample payloads of mobile robots. Such trained DL algorithms demonstrate a mean average precision (mAP) of 90.51% in an average inference time of 67.62 milliseconds within 6,000 training epochs. We claim that the use of such mobile platform setups inferring real-time ecological data can help us achieve our goal of quick and effective biodiversity surveys. An experimental test payload is fabricated, and online as well as offline field surveys are conducted, validating the proposed methodology for species identification that can be further extended to geo-localisation of flora and fauna in any ecosystem.

A Year at the Forefront of Hydrostat Motion

Currently, in the field of interdisciplinary work in biology, there has been a significant push by the soft robotic community to understand the motion and maneuverability of hydrostats. This Review seeks to expand the muscular hydrostat hypothesis toward new structures, including plants, and introduce innovative techniques to the hydrostat community on new modeling, simulating, mimicking, and observing hydrostat motion methods. These methods range from ideas of kirigami, origami, and knitting for mimic creation to utilizing reinforcement learning for control of bio-inspired soft robotic systems. It is now being understood through modeling that different mechanisms can inhibit traditional hydrostat motion, such as skin, nostrils, or sheathed layered muscle walls. The impact of this Review will highlight these mechanisms, including asymmetries, and discuss the critical next steps toward understanding their motion and how species with hydrostat structures control such complex motions, highlighting work from January 2022 to December 2022.

Disentangling spatiotemporal dynamics in metacommunities through a species-patch network approach

Colonization and extinction at local and regional scales, and gains and losses of patches are important processes in the spatiotemporal dynamics of metacommunities. However, analytical challenges remain in quantifying such spatiotemporal dynamics when species extinction-colonization and patch gain and loss processes act simultaneously. Recent advances in network analysis show great potential in disentangling the roles of colonization, extinction, and patch dynamics in metacommunities. Here, we developed a species-patch network approach to quantify metacommunity dynamics including (i) temporal changes in network structure, and (ii) temporal beta diversity of species-patch links and its components that reflect species extinction-colonization and patch gain and loss. Application of the methods to simulated datasets demonstrated that the approach was informative about metacommunity assembly processes. Based on three empirical datasets, our species-patch network approach provided additional information about metacommunity dynamics through distinguishing the effects of species colonization and extinction at different scales from patch gains and losses and how specific environmental factors related to species-patch network structure. In conclusion, our species-patch network framework provides effective methods for monitoring and revealing long-term metacommunity dynamics by quantifying gains and losses of both species and patches under local and global environmental change.

Conservation tools: the next generation of engineering-biology collaborations

The recent increase in public and academic interest in preserving biodiversity has led to the growth of the field of conservation technology. This field involves designing and constructing tools that use technology to aid in the conservation of wildlife. In this review, we present five case studies and infer a framework for designing conservation tools (CT) based on human-wildlife interaction. Successful CT range in complexity from cat collars to machine learning and game theory methodologies and do not require technological expertise to contribute to conservation tool creation. Our goal is to introduce researchers to the field of conservation technology and provide references for guiding the next generation of conservation technologists. Conservation technology not only has the potential to benefit biodiversity but also has broader impacts on fields such as sustainability and environmental protection. By using innovative technologies to address conservation challenges, we can find more effective and efficient solutions to protect and preserve our planet's resources.

Camera trapping expands the view into global biodiversity and its change

Growing threats to biodiversity demand timely, detailed information on species occurrence, diversity and abundance at large scales. Camera traps (CTs), combined with computer vision models, provide an efficient method to survey species of certain taxa with high spatio-temporal resolution. We test the potential of CTs to close biodiversity knowledge gaps by comparing CT records of terrestrial mammals and birds from the recently released Wildlife Insights platform to publicly available occurrences from many observation types in the Global Biodiversity Information Facility. In locations with CTs, we found they sampled a greater number of days (mean = 133 versus 57 days) and documented additional species (mean increase of 1% of expected mammals). For species with CT data, we found CTs provided novel documentation of their ranges (93% of mammals and 48% of birds). Countries with the largest boost in data coverage were in the historically underrepresented southern hemisphere. Although embargoes increase data providers' willingness to share data, they cause a lag in data availability. Our work shows that the continued collection and mobilization of CT data, especially when combined with data sharing that supports attribution and privacy, has the potential to offer a critical lens into biodiversity. This article is part of the theme issue 'Detecting and attributing the causes of biodiversity change: needs, gaps and solutions'.

Using machine learning to decode animal communication

New methods promise transformative insights and conservation benefits.

Multi-Object Tracking in Heterogeneous environments (MOTHe) for animal video recordings

Aerial imagery and video recordings of animals are used for many areas of research such as animal behaviour, behavioural neuroscience and field biology. Many automated methods are being developed to extract data from such high-resolution videos. Most of the available tools are developed for videos taken under idealised laboratory conditions. Therefore, the task of animal detection and tracking for videos taken in natural settings remains challenging due to heterogeneous environments. Methods that are useful for field conditions are often difficult to implement and thus remain inaccessible to empirical researchers. To address this gap, we present an open-source package called Multi-Object Tracking in Heterogeneous environments (MOTHe), a Python-based application that uses a basic convolutional neural network for object detection. MOTHe offers a graphical interface to automate the various steps related to animal tracking such as training data generation, animal detection in complex backgrounds and visually tracking animals in the videos. Users can also generate training data and train a new model which can be used for object detection tasks for a completely new dataset. MOTHe doesn't require any sophisticated infrastructure and can be run on basic desktop computing units. We demonstrate MOTHe on six video clips in varying background conditions. These videos are from two species in their natural habitat-wasp colonies on their nests (up to 12 individuals per colony) and antelope herds in four different habitats (up to 156 individuals in a herd). Using MOTHe, we are able to detect and track individuals in all these videos. MOTHe is available as an open-source GitHub repository with a detailed user guide and demonstrations at: https://github.com/tee-lab/MOTHe-GUI.

Machine Learning-Based Sensor Data Fusion for Animal Monitoring: Scoping Review

The development of technology, such as the Internet of Things and artificial intelligence, has significantly advanced many fields of study. Animal research is no exception, as these technologies have enabled data collection through various sensing devices. Advanced computer systems equipped with artificial intelligence capabilities can process these data, allowing researchers to identify significant behaviors related to the detection of illnesses, discerning the emotional state of the animals, and even recognizing individual animal identities. This review includes articles in the English language published between 2011 and 2022. A total of 263 articles were retrieved, and after applying inclusion criteria, only 23 were deemed eligible for analysis. Sensor fusion algorithms were categorized into three levels: Raw or low (26%), Feature or medium (39%), and Decision or high (34%). Most articles focused on posture and activity detection, and the target species were primarily cows (32%) and horses (12%) in the three levels of fusion. The accelerometer was present at all levels. The findings indicate that the study of sensor fusion applied to animals is still in its early stages and has yet to be fully explored. There is an opportunity to research the use of sensor fusion for combining movement data with biometric sensors to develop animal welfare applications. Overall, the integration of sensor fusion and machine learning algorithms can provide a more in-depth understanding of animal behavior and contribute to better animal welfare, production efficiency, and conservation efforts.

An Efficient Method for Monitoring Birds Based on Object Detection and Multi-Object Tracking Networks

To protect birds, it is crucial to identify their species and determine their population across different regions. However, currently, bird monitoring methods mainly rely on manual techniques, such as point counts conducted by researchers and ornithologists in the field. This method can sometimes be inefficient, prone to errors, and have limitations, which may not always be conducive to bird conservation efforts. In this paper, we propose an efficient method for wetland bird monitoring based on object detection and multi-object tracking networks. First, we construct a manually annotated dataset for bird species detection, annotating the entire body and head of each bird separately, comprising 3737 bird images. We also built a new dataset containing 11,139 complete, individual bird images for the multi-object tracking task. Second, we perform comparative experiments using a state-of-the-art batch of object detection networks, and the results demonstrated that the YOLOv7 network, trained with a dataset labeling the entire body of the bird, was the most effective method. To enhance YOLOv7 performance, we added three GAM modules on the head side of the YOLOv7 to minimize information diffusion and amplify global interaction representations and utilized Alpha-IoU loss to achieve more accurate bounding box regression. The experimental results revealed that the improved method offers greater accuracy, with mAP@0.5 improving to 0.951 and mAP@0.5:0.95 improving to 0.815. Then, we send the detection information to DeepSORT for bird tracking and classification counting. Finally, we use the area counting method to count according to the species of birds to obtain information about flock distribution. The method described in this paper effectively addresses the monitoring challenges in bird conservation.

Animal Species Recognition with Deep Convolutional Neural Networks from Ecological Camera Trap Images

Accurate identification of animal species is necessary to understand biodiversity richness, monitor endangered species, and study the impact of climate change on species distribution within a specific region. Camera traps represent a passive monitoring technique that generates millions of ecological images. The vast numbers of images drive automated ecological analysis as essential, given that manual assessment of large datasets is laborious, time-consuming, and expensive. Deep learning networks have been advanced in the last few years to solve object and species identification tasks in the computer vision domain, providing state-of-the-art results. In our work, we trained and tested machine learning models to classify three animal groups (snakes, lizards, and toads) from camera trap images. We experimented with two pretrained models, VGG16 and ResNet50, and a self-trained convolutional neural network (CNN-1) with varying CNN layers and augmentation parameters. For multiclassification, CNN-1 achieved 72% accuracy, whereas VGG16 reached 87%, and ResNet50 attained 86% accuracy. These results demonstrate that the transfer learning approach outperforms the self-trained model performance. The models showed promising results in identifying species, especially those with challenging body sizes and vegetation.

Bridging clinic and wildlife care with AI-powered pan-species computational pathology

Cancers occur across species. Understanding what is consistent and varies across species can provide new insights into cancer initiation and evolution, with significant implications for animal welfare and wildlife conservation. We build a pan-species cancer digital pathology atlas (panspecies.ai) and conduct a pan-species study of computational comparative pathology using a supervised convolutional neural network algorithm trained on human samples. The artificial intelligence algorithm achieves high accuracy in measuring immune response through single-cell classification for two transmissible cancers (canine transmissible venereal tumour, 0.94; Tasmanian devil facial tumour disease, 0.88). In 18 other vertebrate species (mammalia = 11, reptilia = 4, aves = 2, and amphibia = 1), accuracy (range 0.57-0.94) is influenced by cell morphological similarity preserved across different taxonomic groups, tumour sites, and variations in the immune compartment. Furthermore, a spatial immune score based on artificial intelligence and spatial statistics is associated with prognosis in canine melanoma and prostate tumours. A metric, named morphospace overlap, is developed to guide veterinary pathologists towards rational deployment of this technology on new samples. This study provides the foundation and guidelines for transferring artificial intelligence technologies to veterinary pathology based on understanding of morphological conservation, which could vastly accelerate developments in veterinary medicine and comparative oncology.

CATER: Combined Animal Tracking & Environment Reconstruction

Quantifying the behavior of small animals traversing long distances in complex environments is one of the most difficult tracking scenarios for computer vision. Tiny and low-contrast foreground objects have to be localized in cluttered and dynamic scenes as well as trajectories compensated for camera motion and drift in multiple lengthy recordings. We introduce CATER, a novel methodology combining an unsupervised probabilistic detection mechanism with a globally optimized environment reconstruction pipeline enabling precision behavioral quantification in natural environments. Implemented as an easy to use and highly parallelized tool, we show its application to recover fine-scale motion trajectories, registered to a high-resolution image mosaic reconstruction, of naturally foraging desert ants from unconstrained field recordings. By bridging the gap between laboratory and field experiments, we gain previously unknown insights into ant navigation with respect to motivational states, previous experience, and current environments and provide an appearance-agnostic method applicable to study the behavior of a wide range of terrestrial species under realistic conditions.

Emerging technologies for behavioral research in changing environments

The first response exhibited by animals to changing environments is typically behavioral. Behavior is thus central to predicting, and mitigating, the impacts that natural and anthropogenic environmental changes will have on populations and, consequently, ecosystems. Yet the inherently multiscale nature of behavior, as well as the complexities associated with inferring how animals perceive their world, and make decisions, has constrained the scope of behavioral research. Major technological advances in electronics and in machine learning, however, provide increasingly powerful means to see, analyze, and interpret behavior in its natural complexity. We argue that these disruptive technologies will foster new approaches that will allow us to move beyond quantitative descriptions and reveal the underlying generative processes that give rise to behavior.

Identifying bird species by their calls in Soundscapes

In many real data science problems, it is common to encounter a domain mismatch between the training and testing datasets, which means that solutions designed for one may not transfer well to the other due to their differences. An example of such was in the BirdCLEF2021 Kaggle competition, where participants had to identify all bird species that could be heard in audio recordings. Thus, multi-label classifiers, capable of coping with domain mismatch, were required. In addition, classifiers needed to be resilient to a long-tailed (imbalanced) class distribution and weak labels. Throughout the competition, a diverse range of solutions based on convolutional neural networks were proposed. However, it is unclear how different solution components contribute to overall performance. In this work, we contextualise the problem with respect to the previously existing literature, analysing and discussing the choices made by the different participants. We also propose a modular solution architecture to empirically quantify the effects of different architectures. The results of this study provide insights into which components worked well for this challenge.

Through Hawks’ Eyes: Synthetically Reconstructing the Visual Field of a Bird in Flight

Birds of prey rely on vision to execute flight manoeuvres that are key to their survival, such as intercepting fast-moving targets or navigating through clutter. A better understanding of the role played by vision during these manoeuvres is not only relevant within the field of animal behaviour, but could also have applications for autonomous drones. In this paper, we present a novel method that uses computer vision tools to analyse the role of active vision in bird flight, and demonstrate its use to answer behavioural questions. Combining motion capture data from Harris’ hawks with a hybrid 3D model of the environment, we render RGB images, semantic maps, depth information and optic flow outputs that characterise the visual experience of the bird in flight. In contrast with previous approaches, our method allows us to consider different camera models and alternative gaze strategies for the purposes of hypothesis testing, allows us to consider visual input over the complete visual field of the bird, and is not limited by the technical specifications and performance of a head-mounted camera light enough to attach to a bird’s head in flight. We present pilot data from three sample flights: a pursuit flight, in which a hawk intercepts a moving target, and two obstacle avoidance flights. With this approach, we provide a reproducible method that facilitates the collection of large volumes of data across many individuals, opening up new avenues for data-driven models of animal behaviour.

Integrated Population Models: Achieving Their Potential

Precise and accurate estimates of abundance and demographic rates are primary quantities of interest within wildlife conservation and management. Such quantities provide insight into population trends over time and the associated underlying ecological drivers of the systems. This information is fundamental in managing ecosystems, assessing species conservation status and developing and implementing effective conservation policy. Observational monitoring data are typically collected on wildlife populations using an array of different survey protocols, dependent on the primary questions of interest. For each of these survey designs, a range of advanced statistical techniques have been developed which are typically well understood. However, often multiple types of data may exist for the same population under study. Analyzing each data set separately implicitly discards the common information contained in the other data sets. An alternative approach that aims to optimize the shared information contained within multiple data sets is to use a “model-based data integration” approach, or more commonly referred to as an “integrated model.” This integrated modeling approach simultaneously analyzes all the available data within a single, and robust, statistical framework. This paper provides a statistical overview of ecological integrated models, with a focus on integrated population models (IPMs) which include abundance and demographic rates as quantities of interest. Four main challenges within this area are discussed, namely model specification, computational aspects, model assessment and forecasting. This should encourage researchers to explore further and develop new practical tools to ensure that full utility can be made of IPMs for future studies.

A Lightweight Automatic Wildlife Recognition Model Design Method Mitigating Shortcut Learning

Recognizing wildlife based on camera trap images is challenging due to the complexity of the wild environment. Deep learning is an optional approach to solve this problem. However, the backgrounds of images captured from the same infrared camera trap are rather similar, and shortcut learning of recognition models occurs, resulting in reduced generality and poor recognition model performance. Therefore, this paper proposes a data augmentation strategy that integrates image synthesis (IS) and regional background suppression (RBS) to enrich the background scene and suppress the existing background information. This strategy alleviates the model's focus on the background, guiding it to focus on the wildlife in order to improve the model's generality, resulting in better recognition performance. Furthermore, to offer a lightweight recognition model for deep learning-based real-time wildlife monitoring on edge devices, we develop a model compression strategy that combines adaptive pruning and knowledge distillation. Specifically, a student model is built using a genetic algorithm-based pruning technique and adaptive batch normalization (GA-ABN). A mean square error (MSE) loss-based knowledge distillation method is then used to fine-tune the student model so as to generate a lightweight recognition model. The produced lightweight model can reduce the computational effort of wildlife recognition with only a 4.73% loss in accuracy. Extensive experiments have demonstrated the advantages of our method, which is beneficial for real-time wildlife monitoring with edge intelligence.

Current topics and challenges in geoAI

Taken literally, geoAI is the use of Artificial Intelligence methods and techniques in solving geo-spatial problems. Similar to AI more generally, geoAI has seen an influx of new (big) data sources and advanced machine learning techniques, but also a shift in the kind of problems under investigation. In this article, we highlight some of these changes and identify current topics and challenges in geoAI.