Tree‐level information is increasingly required to reduce uncertainty in biodiversity assessment and species‐aware carbon accounting, yet individual tree crown (ITC) delineation and species mapping remain challenging in compositionally heterogeneous forests at landscape scales. We present a scalable multimodal workflow for crown‐resolved, species‐level mapping that fuses very high‐resolution (VHR) WorldView‐3 imagery with airborne lidar. The approach uses lidar‐guided prompting to drive a minimally fine‐tuned Segment Anything Model (SAM) for high‐recall canopy detection, followed by shape‐aware instance separation with StarDist to delineate individual crowns. Across benchmarks on Hawaiʻi Island, the proposed sam‐stars framework achieves strong delineation performance with limited annotation, outperforming an end‐to‐end Detectron2 (Mask R‐CNN) baseline most strongly at small training sizes. In an object‐based species‐mapping workflow, crown‐level multispectral and structural features are classified with gradient‐boosted trees, yielding the highest species‐aware semantic panoptic quality (PQ = 0.47) among evaluated methods. Applied at landscape scale using a delineation model trained on only 254 crowns and a larger independent species inventory, the workflow produced a wall‐to‐wall map of 333,204 crowns. When aggregated to 30 m resolution, crown‐derived canopy cover was 19.0% compared with 14.9% in a publicly available land‐cover product, indicating substantial under‐detection in existing maps. By enabling label‐efficient, crown‐resolved, species‐aware mapping from widely available lidar and VHR imagery, this framework supports applications ranging from invasive species detection to restoration planning and species‐informed carbon accounting in structurally heterogeneous forest landscapes.
Heathland ecosystems have undergone substantial degradation over the past 150 years, mainly driven by land‐use changes, agricultural intensification and climate change. As they support significant biodiversity, their conservation is a major concern for both scientists, practitioners and public authorities. Heathland monitoring and conservation can be addressed through field‐based ecology and remote sensing, which rely on distinct but overlapping concepts, data and methodologies. The two approaches primarily differ in their spatial scale and analytical tools, ranging from fine‐scale and site‐specific field ecology to regional or national remote sensing observations. This paper compares the respective contributions of field ecology and remote sensing to heathland monitoring and conservation, with a particular focus on the potential of unmanned aerial vehicles (UAVs). The Essential Biodiversity Variables (EBVs) framework is used as a common reference to assess which variables are addressed by each approach. The study pursues three objectives: (1) comparing field ecology and remote sensing contributions, (2) examining the capabilities of remote sensing technologies and (3) evaluating the potential of UAVs for heathland monitoring. The analysis is based on the review of 160 peer‐reviewed articles (76 field ecology and 84 remote sensing studies). Results reveal a broadly similar distribution of EBVs between the two approaches. Field ecology studies more frequently address the Species traits , Species populations and Ecosystem function EBV classes, whereas remote sensing studies predominantly inform on the Ecosystem structure EBV class. Within remote sensing, satellite and aircraft platforms are the most commonly used, primarily relying on RGB and multispectral imagery. UAV‐based studies remain rare (13 articles), but recent multi‐platform approaches combining UAVs with LiDAR, multispectral or hyperspectral data at high spatial resolution highlight their strong potential. The paper also provides a synthesis of key EBVs relevant to heathland conservation for which UAVs appear particularly promising.
ABSTRACT Floodplain restorations are an increasingly common nature‐based solution (NbS) to support flood risk reduction and enhance ecosystem services while protecting biodiversity. Monitoring vegetation structure and habitats in restored floodplains is critically important to support evidence‐based NbS implementation and biodiversity conservation interventions. Remote sensing (RS)‐based machine learning (ML) methods for floodplain vegetation mapping are complicated by hydrology driven, spatiotemporally dynamic habitat mosaics, high heterogeneity, spatial interspersion, and spectral mixing. Leveraging prevailing theory of floodplain ecology, ecohydrology‐informed ML modeling could improve habitat classifications. We tested the effectiveness of a novel RS and hydrological data fusion to classify vegetation and land cover types in floodplains along the lower Missouri River using supervised ML algorithms. Extreme Gradient Boosting model results showed higher classification accuracy from RS hydrological data fusion (overall accuracy, OA = 85.8%, kappa = 0.84) over multisensor (optical radar) RS data fusion (OA = 80%, Kappa = 0.77). Certain land cover classes, particularly those with structurally similar vegetative communities but ecologically important differences in hydrology (e.g., soil moisture, subcanopy inundation), showed greater classification uncertainty. The user's and producer's accuracies for palustrine emergent and upland herbaceous, two dominant land cover types associated with high uncertainty, increased by ≥ 10%, an improvement from 55% to 60% in base model to 70%–75% in ecohydrology‐informed model. Post hoc SHapley Additive exPlanations (SHAP) analysis further highlighted the importance of hydrological features in model predictions and distinguishing spectrally mixed hard‐to‐separate classes. Our results demonstrate how incorporating ecohydrological knowledge into ML modeling improves vegetation classification in heterogeneous landscapes such as large river floodplains. This methodological framework can support Natural Flood Management and NbS implementation monitoring, but further validation across river systems and years will be needed to assess broader transferability. Future work utilizing new and additional high‐resolution RS products for canopy height and subcanopy water could also make great contributions to habitat classification in floodplains.
Reliable estimation of fish abundance and demographic structure is critical for assessing ecosystem condition and informing conservation management. Non-invasive surveys, particularly remote video, are increasingly used for ecological monitoring; however, existing analytical approaches cannot distinguish individuals across video sequences unless they are observed simultaneously within the same video frame. Consequently, the number of individuals and their life-stage compositions may be underestimated relative to the footage captured. Here, we present MIDS (Multi-frame Individual Distinctions by Size), a novel analytical software that uses length measurements, precision error and statistical inference to distinguish individuals across multiple frames within a video sequence. We applied MIDS to 48 stereo-video deployments across two estuaries (Wallis Lake and Lake Macquarie, New South Wales, Australia) and two habitat types (rocky reefs and sandy beaches), and compared abundance and life-stage estimates to the single-frame MaxN metric. MIDS significantly increased abundance estimates across most stereo-camera deployments despite differences among habitats and estuaries. The largest increases occurred within juvenile cohorts of Yellowfin Bream (Acanthopagrus australis), Sand Whiting (Sillago ciliata) and Luderick (Girella tricuspidata), where the median difference between MIDS and MaxN ranged from one to six additional juveniles per camera deployment across most estuaries and habitats (all effect sizes r > 0.70). Increases among adult and harvestable life stages were generally smaller and based on fewer observations, reducing statistical power to detect consistent differences between methods. While MIDS provides an improved analytical framework for extracting additional abundance and demographic information from stereo-video surveys, further validation is required before its broader applications to conservation and management can be fully realised. Nevertheless, MIDS opens opportunities for improved data extraction from spatially separated camera arrays, longer sampling durations and emerging AI-driven length and identification workflows, by enabling the removal of repeat observations across time. Collectively, these advances position MIDS as a scalable tool for strengthening ecological inference in aquatic systems.
Remote sensing of colonial wildlife has largely focused on estimating population size, providing limited insight into internal demographic structure and spatial organization. We present a transferable, individual-based framework that integrates morphology-informed classification and spatial metrics to extract demographic composition and social spacing from high-resolution unmanned aerial vehicle (UAV) imagery. The workflow combines gray-value discrimination, geometry-based posture identification, aggregation area estimation, and centroid-based distance analysis within a unified object-level pipeline. Applied to a large Antarctic seabird colony, the framework quantified age structure, behavioral states, aggregation patterns, and inter-individual spacing, with overall extraction errors below 15%. As a case demonstration, this approach offers a replicable workflow that extends wildlife remote sensing beyond abundance estimation toward structural and spatial metrics, enabling investigation of population organization and ecological processes in dense animal aggregations.
