Preprints
Filtering by Subject: Artificial Intelligence and Robotics
Large Language Models for Marine Protected Area Design and Management Planning
Published: 2026-10-12
Subjects: Artificial Intelligence and Robotics, Biodiversity, Bioinformatics, Natural Resources and Conservation
Achieving biodiversity targets requires advances in marine conservation planning and action. We make three arguments about large language models (LLMs): they add more value when embedded within decision-support systems; they integrate fragmented expertise, data, and tools; they can increase capacity in under-resourced settings through co-design. We illustrate these arguments with a conceptual [...]
On the Promise (and Practice) of Deep Comparative Methods: Expressivity, Limitations, and Design
Published: 2026-09-28
Subjects: Artificial Intelligence and Robotics, Biodiversity, Ecology and Evolutionary Biology, Evolution
Comparative biologists explain the diversity of organismal form and function by deciding which traits to measure and which evolutionary processes to model. These decisions make inference possible, but they also confine it to traits, relationships, and processes that can be specified a priori. Neural networks can expand this space, learning characters directly from complex observations and [...]
Aligning evidence and action: The Research-to-Decision (RED) Framework to address modern environmental and ecological challenges
Published: 2026-09-28
Subjects: Applied Statistics, Artificial Intelligence and Robotics, Biodiversity, Databases and Information Systems, Ecology and Evolutionary Biology, Environmental Health and Protection, Environmental Indicators and Impact Assessment, Environmental Monitoring, Longitudinal Data Analysis and Time Series, Models and Methods, Multivariate Analysis, Natural Resources and Conservation, Natural Resources Management and Policy, Statistical Methodology, Statistical Models
Ecological and biodiversity crises are intensifying amid rapid advances in sensing, computation, and artificial intelligence (AI). AI offers unprecedented capabilities to understand complex ecological systems, provided outputs are interpretable, applied responsibly, and evaluated by domain experts. When addressing key challenges, science-policy frictions arise through different requirements [...]
wildtag.ai: an open-access end-to-end offline AI-assisted desktop application for managing camera trap projects
Published: 2026-09-01
Subjects: Artificial Intelligence and Robotics, Bioinformatics, Computer Sciences, Databases and Information Systems, Ecology and Evolutionary Biology, Environmental Monitoring, Environmental Sciences, Life Sciences, Natural Resources and Conservation, Natural Resources Management and Policy, Physical Sciences and Mathematics, Terrestrial and Aquatic Ecology
1. Camera traps generate massive volumes of data, yet their decision-support potential is diluted by the bottleneck of manual processing. Automated classification has largely solved this, but classification alone does not remove the barriers to adoption: it must be coupled with the means to turn predictions into a verified, shareable dataset, which no existing tool does end-to-end, offline, and [...]
IUCN Red List Threatened Bird Classification from Citizen-Science Images using Fine-Tuned CNN Features and a Hierarchical Lightweight Classifier
Published: 2026-08-18
Subjects: Artificial Intelligence and Robotics, Biodiversity, Ecology and Evolutionary Biology, Ornithology
To prevent bird strikes at wind farms, an approaching protected bird must be identified so that the turbine can be halted before a collision. We present a bird-species classification module intended for that real-time, low-power edge setting: a hierarchical system assigning IUCN Red List categories to in-the-wild citizen-science bird images, combining a ResNet50 feature extractor fine-tuned on a [...]
Using large language models to identify reporting challenges for Target 6 implementation: a short validation against human assessment
Published: 2026-08-13
Subjects: Artificial Intelligence and Robotics, Biodiversity, Research Methods in Life Sciences
National reports and related biodiversity policy documents contain important information on implementation barriers, but extracting comparable evidence across many countries is timeconsuming. We tested whether large language models (LLMs) could provide a reliable firstpass synthesis of reported challenges relevant to Target 6 of the Kunming–Montreal Global Biodiversity Framework. Human assessor [...]
Deep Learning for Combating Wildlife and Environmental Crime: A Survey
Published: 2026-07-24
Subjects: Artificial Intelligence and Robotics, Natural Resources and Conservation, Natural Resources Management and Policy, Other Ecology and Evolutionary Biology
Wildlife and Environmental Crime (WEC) describes illegal activities that threaten biodiversity and harm the environment. In recent years, deep learning (DL) has shown strong potential in combating WEC by learning complex patterns in data through the extraction of high-dimensional features. DL also offers powerful capabilities for prediction, classification, and object tracking, and has become an [...]
From Scalable Biodiversity Measurement to Credible Biodiversity Metrics
Published: 2026-07-09
Subjects: Artificial Intelligence and Robotics, Biodiversity, Business, Ecology and Evolutionary Biology, Environmental Indicators and Impact Assessment, Environmental Monitoring, Environmental Sciences, Environmental Studies, Life Sciences, Natural Resources and Conservation, Natural Resources Management and Policy, Other Economics, Physical Sciences and Mathematics, Social and Behavioral Sciences, Statistical Models, Sustainability, Technology and Innovation
Governments struggle to develop effective policies to counter the decline of species and ecosystems. An obstacle to command-and-control and incentive-based mechanisms is that biodiversity is costly to measure, creating an information asymmetry in which firms and governments are incentivised to withhold information on adverse impacts. Using a principal-agent model, we show that credible reporting [...]
Foundational AI could usher in a new era for models of all life on Earth
Published: 2026-03-24
Subjects: Artificial Intelligence and Robotics, Ecology and Evolutionary Biology
Deep-learned foundational AI could usher in a new era for the simulation of whole ecosystems, argue Joseph Millard and colleagues.
Echo-dash: Keeping ecologists in the loop with an open source, online ecoacoustic dashboard for interactive exploration of spatiotemporal soundscape data
Published: 2026-03-08
Subjects: Artificial Intelligence and Robotics, Biodiversity, Databases and Information Systems, Environmental Monitoring, Population Biology, Software Engineering, Terrestrial and Aquatic Ecology
Passive acoustic monitoring (PAM) is being adopted in a range of contexts. Emerging methods facilitate analysis of large-scale data sets, but ecological interpretation of acoustic indices is not straightforward. In addition, the technical and logistical requirements of using emerging AI methods for big data mean that conservation actors increasingly adopt third-party analysis solutions. We argue [...]
Insect monitoring without pitfalls: seven steps for robust insect sensing systems
Published: 2026-02-13
Subjects: Artificial Intelligence and Robotics, Biodiversity, Computer Sciences, Ecology and Evolutionary Biology, Entomology, Life Sciences
Data shortages fuel controversy about an ongoing insect biodiversity crisis. Insects are immensely diverse and functionally critical for ecosystems, yet data on insect trends remain patchy and biased. Sensors, ranging from camera-equipped light traps to weather radar stations, are set to transform entomological data collection. Meanwhile, AI models that extract biological information from sensors [...]
Three Decades of Artisanal Gold Mining Reveal Contrasting Outcomes for Conservation Areas and Indigenous Lands
Published: 2025-12-09
Subjects: Artificial Intelligence and Robotics, Biodiversity, Earth Sciences, Ecology and Evolutionary Biology, Environmental Monitoring, Environmental Sciences, Geography, Life Sciences, Natural Resources and Conservation, Remote Sensing, Sustainability, Terrestrial and Aquatic Ecology
Artisanal gold mining (AGM) is a critical driver of tropical forest disturbance, yet is challenging to quantify. Using deep learning and three decades of satellite imagery (1995-2024), we mapped AGM expansion across three contrasting countries in northern Amazonia. We evaluate mining trends, and extent across differing land designations and associated impacts on carbon stocks. AGM area increased [...]
IQ-NET: Fast and Accurate Quartet Phylogenetic Inference Using Deep Learning Trained on Empirical DNA Alignments
Published: 2025-11-25
Subjects: Artificial Intelligence and Robotics
Phylogenetic inference is fundamental to modern biology, with many applications including evolutionary biology, epidemiology, and comparative genomics. While maximum likelihood and Bayesian methods remain the gold standard for phylogenetic analysis, they rely on simplifying assumptions and are computationally intensive. Recent machine learning approaches for phylogenetics offer speed advantages, [...]
GhostNetZero: AI for Detecting Marine Ghost Nets
Published: 2025-09-23
Subjects: Artificial Intelligence and Robotics, Biodiversity, Databases and Information Systems, Environmental Monitoring, Marine Biology, Research Methods in Life Sciences, Sustainability
Abandoned, lost, or otherwise discarded fishing gears (ALDFG), commonly referred to as ghost nets, pose a persistent global threat to marine biodiversity. Constructed from durable synthetic polymers, ghost nets remain intact for decades, continuing to entangle and kill marine organisms while damaging habitats and imposing economic burdens on fisheries and coastal communities. Despite their [...]
Vendi Information Gain for Active Learning and its Application to Ecology
Published: 2025-09-14
Subjects: Artificial Intelligence and Robotics, Biodiversity
While monitoring biodiversity through camera traps has become an important endeavor for ecological research, identifying species in the captured image data remains a major bottleneck due to limited labeling resources. Active learning—a machine learning paradigm that selects the most informative data to label and train a predictive model—offers a promising solution, but typically focuses on [...]