Forest inventories are essential for monitoring forest resources at national and regional scales. Most existing approaches estimate structural attributes from airborne LiDAR point clouds and species composition from multispectral imagery using hand-crafted features. However, joint exploitation of both modalities remains underexplored, and manual feature engineering can limit performance—particularly for 3D point clouds that are heavily simplified into one-dimensional metrics. For multispectral imagery, feature engineering is increasingly replaced by geospatial foundation model (GFM) embeddings (e.g., AlphaEarth and TESSERA), which provide richer representations of Earth observation data.We propose a unified deep learning framework, based on Point Transformer v3, that jointly estimates forest structural parameters and tree species composition, by combining airborne LiDAR with georeferenced GFM embeddings. We show that learned features from the LiDAR point cloud outperform hand-crafted features, and that mid-level fusion of those LiDAR features with GFM embeddings further improves species classification performance, while structural attributes are primarily driven by LiDAR data.The results highlight the potential of GFM embeddings in combination with learned LiDAR features, for scalable and generalizable forest attribute mapping.

Fine-scale, spatially explicit forest attribute maps are essential for guiding forest management and policy decisions. Such maps, based on the combination of National Forest Inventory (NFI) and remote sensing datasets, have a long tradition in the Nordic countries. Harmonizing the pixel size among national forest attribute maps would considerably improve the utility of the maps for users. However, the maps are often aligned with the NFI plot size, and the influence of creating these maps at different spatial resolutions (i.e. pixel sizes) is little studied. We assess the stand-level uncertainty (RMSE) of biomass, volume, basal area, and Lorey’s height estimates resulting from the aggregation of maps across varying spatial resolutions. Models fit at 16 m native resolution using more than 14 000 NFI plots were applied for predictions at pixels sizes (side lengths) of 1, 5, 10, 16, and 30 m. For independent validation, we used more than 600 field plots – that cover a total area of 24 ha and were clustered within 65 stands across Norway. For all attributes, the lowest RMSEs, ranging from 6.86% for Lorey’s height to 13.86% for volume, were observed for predictions at pixel sizes of 5 m to 16 m. The RMSE changes across resolutions were generally small (< 5%) for biomass, volume, and basal area. For Lorey’s height, changing the spatial resolution resulted in large RMSEs of up to 25%. Overall, our findings suggest that the main forest attributes can be mapped at a finer resolutions without complex adjustments.

The dataset includes Pan-European maps of timber volume (Vol), above-ground biomass (AGB), and deciduous-coniferous proportion (DCP) with a pixel size of 10×10 m for the reference year 2020. In addition, a measure of prediction uncertainty is provided for each pixel. The maps have been created using a combination of a Sentinel-2 mosaic, Copernicus layers, and National Forest Inventory (NFI) data.The mapping was done with the k-Nearest Neighbour (kNN, k=7) approach with harmonized data of species-specific Vol and AGB from 14 NFIs consisting of approximately 151 000 field plots across Europe. The maps cover 40 European countries, forming a continuous coverage of the western part of the European continent.A sample of 1/3 of NFI plots was left out for validation, whereas 2/3 of the plots were used for mapping. Maps were created independently for 13 multi-country processing areas. Root-mean-squared-errors (RMSEs) for AGB ranged from 53 % in the Nordic processing area to 73 % in the South-Eastern area. The maps are on average nearly unbiased on European level (1.0 % of the mean AGB), but show significant overestimation for small biomass values (53 % bias for forests with AGB less than 150 t/ha) and underestimation for high biomass values (-55 % bias for forests with AGB higher than 500 t/ha).The created maps are the first of their kind as they are utilizing a large number of harmonized NFI plot observations and consistent remote sensing data for high-resolution forest attribute mapping. While the published maps can be useful for visualization and other purposes, they are primarily meant as auxiliary information in model-assisted estimation where model-related biases can be mitigated, and field-based estimates improved. Therefore, additional calibration procedures were not applied, and especially high Vol and AGB values tend to be underestimated. We therefore discourage from summarizing map values (pixel counting) over areas in interest, as this may inadvertently result in biased estimates.

Forest ecosystems will play a critical role in achieving policy targets for biodiversity and conservation, such as those set out in the EU Biodiversity strategy for 2030. However, practitioners need to know where forests of high conservation value are to make the best-informed decisions about which forests to prioritize. Here, we combine airborne LiDAR (airborne laser scanning/ALS), optical satellite imagery, and gridded datasets on soil and water availability with machine learning models to predict forests' conservation value across Denmark. We then use change-detection algorithms to identify forests that had been disturbed since the collection of the LiDAR data to produce up-to-date estimates for the year 2020. Our models reached a high predictive capacity (82% accuracy) and suggested that 1982 km(2) (~31%) of Denmark's forests were of potential high conservation value. Our study demonstrates the utility of data fusion approaches to identify forest areas of high value for conservation at fine spatial resolutions (~10-100 m) and nationwide extents. However, uncertainties remain in our approach. Hence, our findings should be used to guide field-based assessments to confirm the in situ conservation value of the forests. Only in combination with such in situ data will approaches like ours enable decision makers to better protect forest biodiversity.

The Eurasian Water Shrew (Neomys fodiens) is one of the largest shrew (Soricidae) species in Eurasia. In Western Europe, this semiaquatic species often occurs in riparian and marshland habitats that have a high degree of naturalness, but is being threatened by habitat degradation and other anthropogenic factors. The species mostly occurs in low abundance and is elusive. Therefore, understanding its habitat use is challenging, yet imperative for establishing species-specific conservation measures. Technological developments in radio tracking and high-resolution remote sensing such as Light Detection And Ranging (LiDAR) now enable the quantification of ecological niches and provide insight into habitat requirements for a species. Here, we combined radio tracking and LiDAR to quantify habitat use by Eurasian water shrews. Alongside a lowland brook in the Netherlands, 20 individuals were tracked between September and October 2022, resulting in 332 unique locations of Eurasian water shrews. For each of these locations, 11 LiDAR-derived variables were calculated and subsequently analyzed in a species distribution model (SDM). The SDM yielded a model with a high accuracy (predictive performance AUC = 0.93). The variable of highest importance was dense and relatively short vegetation <1 m, which had a positive effect on Eurasian Water Shrew occurrence. Open areas seem to be avoided. Vegetation of heights between 1 and 15 m were found to be less important for the occurrence. The probability of occurrence decreased with increasing distance to water, indicating that the species occurs in the proximity of water, although vegetation-related variables were more important. The obtained detailed knowledge of fine-scale habitat use can be used to improve habitat conservation, restoration, and management for the species. Combining radiotelemetry data with LiDAR data is a promising approach to identifying species-habitat relationships of elusive species such as the Eurasian Water Shrew.

Background: The mapping of tree species within Norwegian forests is a time-consuming process, involving forest associations relying on manual labeling by experts. The process can involve both aerial imagery, personal familiarity, or on-scene references, and remote sensing data. The state-of-the-art methods usually use high resolution aerial imagery with semantic segmentation methods. Methods: We present a deep learning based tree species classification model utilizing only lidar (Light Detection And Ranging) data. The lidar images are segmented into four classes (Norway Spruce, Scots Pine, Birch, background) with a U-Net based network. The model is trained with focal loss over partial weak labels. A major benefit of the approach is that both the lidar imagery and the base map for the labels have free and open access. Results: Our tree species classification model achieves a macro-averaged F1 score of 0.70 on an independent validation with National Forest Inventory (NFI) in-situ sample plots. That is close to, but below the performance of aerial, or aerial and lidar combined models.

Aim The increasing availability of remote sensing (RS) products from airborne laser scanning (ALS) surveys, synthetic aperture radar acquisitions and multispectral satellite imagery provides unprecedented opportunities for describing the physical structure and seasonal changes of vegetation. However, the added value of these RS products for predicting species distributions and animal habitats beyond land cover maps remains little explored. Here, we aim to assess how metrics derived from different types of high-resolution (10 m) RS products predict the habitat suitability of wetland birds. Location North-eastern part of the Netherlands. Methods We built species distribution models (SDMs) with occurrence observations from territory mapping of two selected wetland bird species (great reed warbler and Savi's warbler) and metrics from a Dutch land cover map, country-wide ALS and Sentinel-1 and Sentinel-2 RS products. We then compared model performance, relative variable importance and response curves of the SDMs to assess the contribution and ecological relevance of each RS product and metric. Results Our results showed that ALS and Sentinel metrics improve SDMs with only land cover metrics by 11% and 10% of the Area Under Curve (AUC) for the great reed warbler and the Savi's warbler respectively. Assessments of feature importance revealed that all types of RS products contributed substantially to predicting the habitat suitability of these wetland birds, but that the most important variables vary among species. Main conclusions Our study demonstrates that metrics from different high-resolution RS products capture complementary ecological information on animal habitats, including aspects such as the proportional cover of habitat types, vegetation density and the horizontal variability of vegetation height. Land cover maps with detailed spatial and thematic information can already achieve high model accuracies, but adding metrics derived from ALS point clouds and Sentinel imagery further improve model accuracy and enhance the understanding of animal-habitat relationships.

This data repository contains the bird observation data used in Koma, Z., Seijmonsbergen, A.C., Grootes, M.W., Nattino, F., Groot, J., Sierdsema, H., Foppen, R. & Kissling, W.D. (2022): Better together? Assessing different remote sensing products for predicting habitat suitability of wetland birds. Diversity and Distributions 28: 685–699. The content of this directory is shared under Attribution-NonCommercial-NoDerivatives 4.0 International licence (CC BY-NC-ND 4.0, see https://creativecommons.org/licenses/by-nc-nd/4.0/). For accessing the bird occurrence data for further use then reproducing this article you can contact with Henk Sierdsema (Henk.Sierdsema@sovon.nl) and Ruud Foppen (Ruud.Foppen@sovon.nl) for further information.

The third Dutch national airborne laser scanning flight cam-paign (AHN3, Actueel Hoogtebestand Nederland) conducted between 2014 and 2019 during the leaf-off season (October- April) across the whole Netherlands provides a free and open-access, country-wide dataset with similar to 700 billion points and a point density of similar to 10(-20) points/m2. The AHN3 point cloud was obtained with Light Detection And Ranging (Li-DAR) technology and contains for each point the x, y, z co-ordinates and additional characteristics (e.g. return number, intensity value, scan angle rank and GPS time). Moreover, the point cloud has been pre-processed by 'Rijkswaterstraat' (the executive agency of the Dutch Ministry of Infrastructure and Water Management), comes with a Digital Terrain Model (DTM) and a Digital Surface Model (DSM), and is delivered with a pre-classification of each point into one of six classes (0: Never Classified, 1: Unclassified, 2: Ground, 6: Building, 9: Water, 26: Reserved [bridges etc.]). However, no detailed in-formation on vegetation structure is available from the AHN3 point cloud. We processed the AHN3 point cloud (-16 TB uncompressed data volume) into 10 m resolution raster lay-ers of ecosystem structure at a national extent, using a novel high-throughput workflow called 'Laserfarm' and a cluster of virtual machines with fast central processing units, high memory nodes and associated big data storage for manag-ing the large amount of files. The raster layers (available as GeoTIFF files) capture 25 LiDAR metrics of vegetation structure, including ecosystem height (e.g. 95th percentiles of normalized z), ecosystem cover (e.g. pulse penetration ra-tio, canopy cover, and density of vegetation points within defined height layers), and ecosystem structural complex-ity (e.g. skewness and variability of vertical vegetation point distribution). The raster layers make use of the Dutch pro-jected coordinate system (EPSG:28992 Amersfoort / RD New), are each-1 GB in size, and can be readily used by ecol-ogists in a geographic information system (GIS) or analyti-cal open-source software such as R and Python. Even though the class '1: Unclassified' mainly includes vegetation points, other objects such as cars, fences, and boats can also be present in this class, introducing potential biases in the de-rived data products. We therefore validated the raster layers of ecosystem structure using > 180,0 0 0 hand-labelled LiDAR points in 100 randomly selected sample plots (10 m x 10 m each) across the Netherlands. Besides vegetation, objects such as boats, fences, and cars were identified in the sam-pled plots. However, the misclassification rate of vegetation points (i.e. non-vegetation points that were assumed to be vegetation) was low (-0.05) and the accuracy of the 25 Li -DAR metrics derived from the AHN3 point cloud was high (-90%). To minimize existing inaccuracies in this country-wide data product (e.g. ships on water bodies, chimneys on roofs, or cars on roads that might be incorrectly used as vegetation points), we provide an additional mask that captures water bodies, buildings and roads generated from the Dutch cadaster dataset. This newly generated country-wide ecosystem structure data product provides new oppor-tunities for ecology and biodiversity science, e.g. for map-ping the 3D vegetation structure of a variety of ecosystems or for modelling biodiversity, species distributions, abun-dance and ecological niches of animals and their habitats.(c) 2022 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ )

Quantifying ecosystem structure is of key importance for ecology, conservation, restoration, and biodiversity monitoring because the diversity, geographic distribution and abundance of animals, plants and other organisms is tightly linked to the physical structure of vegetation and associated microclimates. Light Detection And Ranging (LiDAR) - an active remote sensing technique - can provide detailed and high resolution information on ecosystem structure because the laser pulse emitted from the sensor and its subsequent return signal from the vegetation (leaves, branches, stems) delivers three-dimensional point clouds from which metrics of vegetation structure (e.g. ecosystem height, cover, and structural complexity) can be derived. However, processing 3D LiDAR point clouds into geospatial data products of ecosystem structure remains challenging across broad spatial extents due to the large volume of national or regional point cloud datasets (typically multiple terabytes con-sisting of hundreds of billions of points). Here, we present a high-throughput workflow called 'Laserfarm' enabling the efficient, scalable and distributed processing of multi-terabyte LiDAR point clouds from national and regional airborne laser scanning (ALS) surveys into geospatial data products of ecosystem structure. Laserfarm is a free and open-source, end-to-end workflow which contains modular pipelines for the re-tiling, normalization, feature extraction and rasterization of point cloud information from ALS and other LiDAR surveys. The workflow is designed with horizontal scalability and can be deployed with distributed computing on different in-frastructures, e.g. a cluster of virtual machines. We demonstrate the Laserfarm workflow by processing a country-wide multi-terabyte ALS dataset of the Netherlands (covering-34,000 km2 with-700 billion points and -16 TB uncompressed LiDAR point clouds) into 25 raster layers at 10 m resolution capturing ecosystem height, cover and structural complexity at a national extent. The Laserfarm workflow, implemented in Python and available as Jupyter Notebooks, is applicable to other LiDAR datasets and enables users to execute automated pipelines for generating consistent and reproducible geospatial data products of ecosystems structure from massive amounts of LiDAR point clouds on distributed computing infrastructures, including cloud computing environments. We provide information on workflow performance (including total CPU times, total wall-time estimates and average CPU times for single files and LiDAR metrics) and discuss how the Laserfarm workflow can be scaled to other LiDAR datasets and computing environments, including remote cloud infrastructures. The Laserfarm workflow allows a broad user community to process massive amounts of LiDAR point clouds for mapping vegetation structure, e.g. for applications in ecology, biodiversity monitoring and ecosystem restoration.

This data repository contains country-wide data products for the ecosystem structure metrics generated from Airborne Laser Scanning (ALS) data across the Netherlands (AHN3). Twenty-five ecosystem structure metrics (at 10-meter resolution, GeoTIFF format) were derived from AHN3 dataset (https://downloads.pdok.nl/ahn3-downloadpage/) using Laserfarm workflow (https://zenodo.org/record/5636773). Laserfarm is a free and open-source workflow that enables efficient, scalable, and distributed processing of multi-terabyte LiDAR point clouds from national and regional ALS surveys into LiDAR metrics of ecosystem structure. All code of Laserfarm is hosted and freely available on GitHub (https://github.com/eEcoLiDAR/Laserfarm). The Jupyter Notebooks for the processing of the AHN3 dataset are available on GitHub (https://github.com/eEcoLiDAR/AHN/tree/main/AHN3). The twenty-five LiDAR metrics are related to three key dimensions of ecosystem structure (ecosystem height, ecosystem cover, and ecosystem structural complexity), and one additional layer of point density is also provided. Each GeoTIFF layer represents one LiDAR metric at 10 m resolution covering the whole Netherlands (file name as "ahn3_10m_featrue_name.tiff"). An overview of all the listed metrics (maps) is also provided in the PDF version (AHN3.pdf). A detailed description of the dataset is available from the following data publication:Kissling, W. D., Y. Shi, Z. Koma, C. Meijer, O. Ku, F. Nattino, A. C. Seijmonsbergen, and M. W. Grootes. 2022. Country-wide data of ecosystem structure from the third Dutch airborne laser scanning survey. Data in Brief: 108798.https://doi.org/10.1016/j.dib.2022.108798 A detailed description of all the metrics can be found in the README file (README.docx).

Mapping and quantifying 3D vegetation structure is essential for assessing and monitoring ecosystem structure and function within wetlands. Airborne Laser Scanning (ALS) is a promising data source for developing indicators of 3D vegetation structure, but derived metrics are often not compared with 3D structural field measurements and the acquisition of ALS data is rarely standardized across different remote sensing surveys. Here, we compare a set of Light Detection And Ranging (LiDAR) metrics derived from ALS datasets with varying characteristics to a standardized set of field measurements of vegetation height, biomass and Leaf Area Index (LAI) across three Hungarian lakes (Lake Balaton, Lake Fertő and Lake Tisza). The ALS datasets differed in whether the recording type was full waveform (FWF) or discrete return, and in their point density (4 pt/m2 and 21 pt/m2). A total of eight LiDAR metrics captured radiometric information as well as descriptors of vegetation cover, height and vertical variability. Multivariate regression models with field-based measurements of vegetation height, biomass or LAI as response variable and LiDAR metrics as predictors showed major differences between ALS recording types, and were affected by differences in spatial resolution, temporal offset and seasonality between field and ALS data acquisition. Vegetation height could be estimated with high to intermediate accuracy (FWF ALS data only: R2 = 0.84; combination of ALS datasets: R2 = 0.67), demonstrating its potential as a robust indicator of 3D vegetation structure across different ALS datasets. In contrast, the estimation of biomass and LAI in these wetlands was sensitive to variation in ALS characteristics and to the discrepancies between field and ALS data in terms of spatial resolution, temporal offset and seasonality (biomass: R2 = 0.20–0.22; LAI: R2 = 0.08–0.30). We recommend the use of FWF ALS data within wetlands because it captures more vegetation structural details in dense reed and marshland vegetation. We further suggest that ecologists and remote sensing scientist should better coordinate the simultaneous and standardized acquisition of field and ALS data for testing the robustness of quantitative descriptors of vegetation cover, height and vertical variability within wetlands. This is important for establishing operational and spatially contiguous ALS-based indicators of 3D ecosystem structure across wetlands.

Light Detection And Ranging (LiDAR) is a promising remote sensing technique for ecological applications because it can quantify vegetation structure at high resolution over broad spatial extents. Using country‐wide airborne laser scanning (ALS) data, we test to what extent fine‐scale LiDAR metrics capturing low vegetation, medium‐to‐high vegetation and landscape‐scale habitat structures can explain the habitat preferences of threatened butterflies at a national extent.

Numerous organisms depend on the physical structure of their habitats, but incorporating such information into ecological niche analyses has been limited by the lack of adequate data over broad spatial extents. The increasing availability of high‐resolution measurements from country‐wide airborne laser scanning (ALS) surveys – a light detection and ranging (LiDAR) technology – now provides unprecedented opportunities for characterizing habitat structure. Here, we use country‐wide ALS data in combination with presence–absence observations of birds from a national monitoring scheme in the Netherlands to quantify niche filling, niche overlap and niche separation of three closely‐related wetland birds (great reed warbler, Eurasian reed warbler and Savi's warbler). We developed a workflow to derive LiDAR metrics capturing different aspects of vertical and horizontal vegetation structure and used a principal component analysis (PCA), niche equivalency and niche similarity tests to analyse the fine‐scale breeding habitat niches of these warbler species in the Netherlands. The widespread Eurasian reed warbler almost completely filled the available wetland habitat space (93%) whereas the two other species showed considerably less niche filling (64% and 74%, respectively). Substantial niche overlap occurred among all species, but each species occupied a distinct part of the habitat space. The great reed warbler mainly occurred in tall and vertically complex wetland vegetation and was absent in areas with large proportions of reedbeds. The Eurasian reed warbler occupied all parts of the wetland habitat space, whereas the Savi's warbler mainly occurred in large homogenous reedbeds with low vegetation height. Our results demonstrate that broad‐scale ecological niche analyses can incorporate the fine‐scale 3D habitat preference of species with unprecedented detail (e.g. 10 m resolution), and thus go much beyond quantifying the climate niche and 2D habitat information from land cover maps. This is important to identify habitat features and priorities for biodiversity conservation in wetlands and other habitats.

Mapping 3D vegetation structure in wetlands is important for conservation and monitoring. Openly accessible country‐wide Airborne Laser Scanning (ALS) data—using light detection and ranging (lidar) technology—are increasingly becoming available and allow us to quantify 3D vegetation structures at fine resolution and across broad spatial extents. Here, we develop a new, open‐source workflow for classifying wetland‐related land cover types and habitats using fine‐scale lidar metrics derived from country‐wide ALS data. We developed a case study in the Netherlands with a workflow consisting of four routines: (1) pre‐processing of ALS data, (2) calculation of lidar metrics (i.e. 31 features representing cover, 3D shape, vertical variability, horizontal variability and height of vegetation as well as microtopography), (3) assessing feature importance of lidar metrics for classifying wetland habitats, and (4) applying a Random Forest algorithm for mapping and prediction. We used an expert‐based vegetation map for annotation and generated 100, 500 and 1000 annotation points for each class. Using a three‐level hierarchical approach, we differentiated at level 1 planar surfaces (e.g. roads and agricultural fields) from wetland vegetation with 82% mean overall accuracy, using predominantly height and horizontal variability metrics. At level 2, we classified wetland vegetation into four land cover types (forest, grassland, reedbeds, shrubs) with 71% mean overall accuracy, using lidar metrics related to vegetation height and horizontal and vertical variability. At level 3, we differentiated two types of land reed as well as water reed with 78% mean overall accuracy, using predominantly vertical variability metrics. Our results demonstrate that lidar metrics (related to vegetation height, cover, vertical and horizontal variability) derived from country‐wide ALS data can differentiate land cover types and habitats within wetlands at high resolution. Given appropriate annotation data, our workflow can be up‐scaled to a country‐wide extent to allow the comprehensive mapping and monitoring of wetlands at national scales.

LiDAR as a remote sensing technology, enabling the rapid 3D characterization of an area from an air- or spaceborne platform, has become a mainstream tool in the (bio)geosciences and related disciplines. For instance, LiDAR-derived metrics are used for characterizing vegetation type, structure, and prevalence and are widely employed across ecosystem research, forestry, and ecology/biology. Furthermore, these types of metrics are key candidates in the quest for Essential Biodiversity Variables (EBVs) suited to quantifying habitat structure, reflecting the importance of this property in assessing and monitoring the biodiversity of flora and fauna, and consequently in informing policy to safeguard it in the light of climate change an human impact. In all these use cases, the power of LiDAR point cloud datasets resides in the information encoded within the spatial distribution of LiDAR returns, which can be extracted by calculating domain-specific statistical/ensemble properties of well-defined subsets of points. Facilitated by technological advances, the volume of point cloud data sets provided by LiDAR has steadily increased, with modern airborne laser scanning surveys now providing high-resolution, (super-)national scale datasets, tens to hundreds of terabytes in size and encompassing hundreds of billions of individual points, many of which are available as open data. Representing a trove of data and, for the first time, enabling the study of ecosystem structure at meter resolution over the extent of tens to hundreds of kilometers, these datasets represent highly valuable new resources. However, their scientific exploitation is hindered by the scarcity of Free Open Source Software (FOSS) tools capable of handling the challenges of accessing, processing, and extracting meaningful information from massive multi-terabyte datasets, as well as by the domain-specificity of any existing tools. Here we present Laserchicken a FOSS, user-extendable, cross-platform Python tool for extracting user-defined statistical properties of flexibly defined subsets of point cloud data, aimed at enabling efficient, scalable, and distributed processing of multi-terabyte datasets. Laserchicken can be seamlessly employed on computing architectures ranging from desktop systems to distributed clusters, and supports standard point cloud and geo-data formats (LAS/LAZ, PLY, GeoTIFF, etc.) making it compatible with a wide range of (FOSS) tools for geoscience. The Laserchicken feature extraction tool is complemented by a FOSS Python processing pipeline tailored to the scientific exploitation of massive nation-scale point cloud datasets, together forming the Laserchicken framework. The ability of the Laserchicken framework to unlock nation-scale LiDAR point cloud datasets is demonstrated on the basis of its use in the eEcoLiDAR project, a collaborative project between the University of Amsterdam and the Netherlands eScience Center. Within the eEcoLiDAR project, Laserchicken has been instrumental in defining classification methods for wetland habitats, as well as in facilitating the use of high-resolution vegetation structure metrics in modelling species distributions at national scales, with preliminary results highlighting the importance of including this information. The Laserchicken Framework rests on FOSS, including the GDAL and PDAL libraries as well as numerous packages hosted on the open source Python Package Index (PyPI), and is itself also available as FOSS (https://pypi.org/project/laserchicken/ and https://github.com/eEcoLiDAR/ ).

Modernization of agricultural land use across Europe is responsible for a substantial decline of linear vegetation elements such as tree lines, hedgerows, riparian vegetation, and green lanes. These linear objects have an important function for biodiversity, e.g., as ecological corridors and local habitats for many animal and plant species. Knowledge on their spatial distribution is therefore essential to support conservation strategies and regional planning in rural landscapes but detailed inventories of such linear objects are often lacking. Here, we propose a method to detect linear vegetation elements in agricultural landscapes using classification and segmentation of high-resolution Light Detection and Ranging (LiDAR) point data. To quantify the 3D structure of vegetation, we applied point cloud analysis to identify point-based and neighborhood-based features. As a preprocessing step, we removed planar surfaces such as grassland, bare soil, and water bodies from the point cloud using a feature that describes to what extent the points are scattered in the local neighborhood. We then applied a random forest classifier to separate the remaining points into vegetation and other. Subsequently, a rectangularity-based region growing algorithm allowed to segment the vegetation points into 2D rectangular objects, which were then classified into linear objects based on their elongatedness. We evaluated the accuracy of the linear objects against a manually delineated validation set. The results showed high user’s (0.80), producer’s (0.85), and total accuracies (0.90). These findings are a promising step towards testing our method in other regions and for upscaling it to broad spatial extents. This would allow producing detailed inventories of linear vegetation elements at regional and continental scales in support of biodiversity conservation and regional planning in agricultural and other rural landscapes.

Vegetation structure is a key determinant of animal diversity and species distributions. The introduction of Light Detection and Ranging (LiDAR) has enabled the collection of massive amounts of point cloud data for quantifying habitat structure at fine resolution. Here, we review the current use of LiDAR‐derived vegetation metrics in diversity and distribution research of birds, a key group for understanding animal–habitat relationships.

Recent advances of Light Detection and Ranging (LiDAR) have facilitated the accurate mapping of ecosystem structure at local and landscape scales. Moreover, the increasing availability of openly accessible Airborne Laser Scanning (ALS) point clouds provides new opportunities for mapping the fine-scale extent and distribution of ecosystems and habitats across broad spatial extents. However, processing of such massive amounts of ALS data is still challenging because it requires computationally efficient software and tools for data handling. Here, we develop an open-source object-based approach for the fine-scale identification and delineation of reedbeds which should permit their mapping across broad spatial extents using country-wide ALS data. Reedbeds are wetlands dominated by the common reed (Phragmites australis) and constitute a major habitat for breeding and migratory birds, thus being of high conservation value in Europe. We apply a region growing segmentation algorithm to automatically classify ALS point clouds into structurally similar vegetation patches and apply it to selected wetland areas in the Netherlands (Giethoon and Naardermeer). Our workflow consists of four steps. In a first step, we derive the most commonly used neighborhood-related LiDAR metrics such as roughness of the vegetation top, descriptive statistics of vegetation height and echo ratio. In a second step, these calculated LiDAR metrics are used to classify the point cloud into ground, vegetation, buildings and water classes. In a third step, an object-based analysis is applied to the vegetation class using a region growing segmentation algorithm for delineating homogenous patches. In a fourth step, the segments are classified into reedbed and non-reedbed vegetation using additional object-related features such as distance from a water body, area and shape of the object. For validation, vegetation plots and available high resolution aerial orthophotos (e.g. DKLN-NL) are used. The vegetation plots come from the Dutch Vegetation Database which uses the Darwin Core Archive (DwC-A) as an open-access (meta)data standard. This allows access to standardized vegetation sampling inventories from various habitats across the Netherlands, and hence provides the opportunity to test the methodology at a country-wide scale. Our results demonstrate that an object-based approach allows the extraction of fine-scale boundaries and delineation of patches of reedbeds as well as their separation from woody vegetation elements such as stand-alone trees within reedbeds. Accuracy assessment indicates a good performance which allows for successful upscaling to broad spatial extents. We further outline how the developed workflow can be built into a new open-source software for processing country-wide ALS data using scalable computing. The fine-scale mapping of reedbed habitats will complement available land cover maps which usually lack information on small and scattered habitats such as reedbeds. Reedbed habitat maps will also provide the necessary spatially-explicit information for predicting species distributions of threatened wetland birds (e.g. Great Bittern and many reed-dwelling warblers) which are a prime target for nature conservation in Europe.