Detection of standing retention trees in boreal forests with airborne laser scanning point clouds and multispectral imagery

Abstract

1. In a landscape consisting primarily of intensive forestry interspersed with some protected areas, multifunctional forestry with retention trees can play a crucial role in nature conservation. Accurate mapping of retention trees is important for guiding landscape-level conservation and forest management and improving landscape connectivity. Sizeable dead and living retention trees play a particularly important ecological role but even their large-scale inventory is often intensive through field work and/or inaccurate. We aimed to detect and classify retention trees using the novel nationwide Finnish airborne laser scanning (ALS) data (~5 pulses/m 2 ) in conjunction with unrectified colour-infrared (CIR) aerial imagery. 2. Applying photogrammetric principles, we added spectral information from the CIR imagery to the ALS-derived point cloud. For a training dataset of 160 retention trees from 19 stands and a geographically separate validation dataset of 79 trees from eight stands, we segmented trees via individual tree detection (ITD), removed most trees belonging to the regenerating vegetation layer, and classified trees into living conifers, living broadleaves and dead trees by linear discriminant analysis.3. The detection rate via ITD differed considerably for dead and living trees, with 41.7% of all dead and 83.8% of all living trees being detected with relatively low commission error rates. Dead trees with smaller diameters and heights were more likely missed, while grouping caused living tree omission. For classification into living conifers, living broadleaves and dead trees, an overall accuracy of 67.3% was achieved in training and 71.2% in validation using only ALS-derived metrics. When adding spectral metrics, the overall accuracies were 79.6% and 61.0% for training and validation respectively. 4. Our findings imply that wall-to-wall large-scale high density ALS data can be used to detect retention trees rather accurately-even larger dead trees-and that metrics derived solely from ALS data can accurately classify detected retention trees into living conifers, living broadleaves and dead trees. Considering the ecological value of retention trees, our results are promising and indicate that

Old trees in young forests : Biodiversity management in planted conifer forests in southern Sweden

“…Studies have demonstrated that the LiDAR measurement error for individual tree heights (of a given species) could be less than 1.0 m (Holmgren and Persson, 2004). Retention trees can be detected on clear-cut areas (Holmström et al, 2019, Hardenbol et al, 2022, Lindbladh et al, 2022 via visual cues. However, visual interpretation of orthophotos is a subjective way to find trees, and trees retained in clear-cut areas may not be present at thinning.…”

“…This advancement may thus provide an unparallel data source that can be used for biodiversity conservation. Many studies have successfully located and characterised retention trees using point clouds from ALS (Naesset and Økland, 2002, Holmgren and Persson, 2004, Hardenbol et al, 2022. However, the use of the canopy height model is sparser (Mustonen et al, 2008, Zielewska-Büttner et al, 2016 and not necessarily targeted for biodiversity conservation management.…”

Towards complex applications of active remote sensing for ecology and conservation

“…Parallel to testing such fundamental ecological hypotheses, efforts have also been moving towards directly deriving required metrics from active data sources, which can potentially be further applied as inputs for a wide range of studies in ecology and conservation. In this Special Feature, Hardenbol et al (2022) discuss how combining airborne active LiDAR and passive multispectral data can help in detecting retention trees, an important component for safeguarding biodiversity of fauna and flora in close‐to‐nature forestry (e.g. Fedrowitz et al, 2014).…”

“…This can be of practical value in, for example, Nordic Europe, with country‐wide aerial data regularly acquired by forest services. Hardenbol et al (2022) achieved detection rates ranging from 41.7% (dead trees) to 83.8% (living trees) by applying an individual tree detection algorithm, although higher omission errors occurred for dead trees with smaller diameters and heights. However, adding spectral metrics from colour infrared photography did not systematically enhance the overall accuracies of classifying living conifers, living broadleaves or dead trees, indicating that the sole use of LiDAR metrics may suffice for separating retention trees from other trees.…”

“…1. Stand-, canopy-and tree-level structural analysis by active RS data (Atkins et al, 2023;Blanchard et al, 2023;Coverdale & Davies, 2023;Hardenbol et al, 2022;Schlund et al, 2022;Singh et al, 2023;Tatsumi et al, 2022;Zhang & Liu, 2023).…”