Loading...

Pubblicazioni Scientifiche

Filtri di ricerca 3 risultati
Pubblicazioni per anno
Benchmarking tree species classification from proximally sensed laser scanning data: Introducing the FOR-species20K dataset
Mostra abstract
Proximally sensed laser scanning presents new opportunities for automated forest ecosystem data capture. However, a gap remains in deriving ecologically pertinent information, such as tree species, without additional ground data. Artificial intelligence approaches, particularly deep learning (DL), have shown promise towards automation. Progress has been limited by the lack of large, diverse, and, most importantly, openly available labelled single-tree point cloud datasets. This has hindered both (1) the robustness of the DL models across varying data types (platforms and sensors) and (2) the ability to effectively track progress, thereby slowing the convergence towards best practice for species classification. To address the above limitations, we compiled the FOR-species20K benchmark dataset, consisting of individual tree point clouds captured using proximally sensed laser scanning data from terrestrial (TLS), mobile (MLS) and drone laser scanning (ULS). Compiled collaboratively, the dataset includes data collected in forests mainly across Europe, covering Mediterranean, temperate and boreal biogeographic regions. It includes scattered tree data from other continents, totaling over 20,000 trees of 33 species and covering a wide range of tree sizes and forms. Alongside the release of FOR-species20K, we benchmarked seven leading DL models for individual tree species classification, including both point cloud (PointNet++, MinkNet, MLP-Mixer, DGCNNs) and multi-view 2D-based methods (SimpleView, DetailView, YOLOv5). 2D Image-based models had, on average, higher overall accuracy (0.77) than 3D point cloud-based models (0.72). Notably, the performance was consistently >0.8 across scanning platforms and sensors, offering versatility in deployment. The top-scoring model, DetailView, demonstrated robustness to training data imbalances and effectively generalized across tree sizes. The FOR-species20K dataset represents an important asset for developing and benchmarking DL models for individual tree species classification using proximally sensed laser scanning data. As such, it serves as a crucial foundation for future efforts to classify accurately and map tree species at various scales using laser scanning technology, as it provides the complete code base, dataset, and an initial baseline representative of the current state-of-the-art of point cloud tree species classification methods. © 2025 The Author(s). Methods in Ecology and Evolution published by John Wiley & Sons Ltd on behalf of British Ecological Society.
Managed forests are a stronghold of non-native beetles in Europe
Mostra abstract
The species richness of vascular plants in forests can have contrasting effects on the occurrence of non-native insects. The establishment of non-native insect populations may be facilitated by low plant species richness, which reflects the availability of few but easily accessible resources, or hampered by high plant species richness due to spatial dilution of resources or biotic resistance (i.e., resistance against biological invasions). The relationship between the species richness of plants and non-native insects is likely influenced by disturbance regimes, which, in European forests, mostly consists of timber harvesting. We investigated this relationship considering two major forest attributes: (i) species richness of non-native vascular plants and (ii) forest management. From 1101 forest plots in Europe, we gathered occurrences of 1212 vascular plant species, including 160 non-native species, and of 2404 beetle species, including 29 non-native species. We tested the relationship between the species richness of non-native beetles and plants using non-linear quantile regressions. We disentangled the effect of non-native plant species richness from that of management on the species richness of non-native beetles, while accounting for forest structural variables, using structural equation models. We found clear evidence of a hump-shaped relationship between non-native beetle and plant species richness. The general shape of the relationship persisted when considering only woody or non-woody plants, as well as only non-native plants. The relationship was also similar between managed and unmanaged forests. However, the proportion of non-native beetles in managed forests was higher than in unmanaged forests at the same plant species richness. Management had a direct negative effect on non-native beetle species richness, whereas non-native plant species richness had a direct positive effect. When considering all direct and indirect effects, management facilitated the occurrence of non-native beetles indirectly via non-native plants rather than directly. Synthesis and applications. Species richness of native and non-native vascular plants modulates the species richness of non-native beetles through relationships with opposite signs. The interplay with management regimes and forest structures determines whether non-native beetles are promoted. Forest management aimed at reducing the intensity of disturbance while encouraging native plant species richness could promote the dominance of dilution effects and biotic resistance and could moderate the establishment of non-native insects. © 2025 The Author(s). Journal of Applied Ecology © 2025 British Ecological Society.
Silvicultural regime shapes understory functional structure in European forests
Mostra abstract
Managing forests to sustain their diversity and functioning is a major challenge in a changing world. Despite the key role of understory vegetation in driving forest biodiversity, regeneration and functioning, few studies address the functional dimensions of understory vegetation response to silvicultural management. We assessed the influence of the silvicultural regimes on the functional diversity and redundancy of European forest understory. We gathered vascular plant abundance data from more than 2000 plots in European forests, each associated with one out of the five most widespread silvicultural regimes. We used generalized linear mixed models to assess the effect of different silvicultural regimes on understory functional diversity (Rao's quadratic entropy) and functional redundancy, while accounting for climate and soil conditions, and explored the reciprocal relationship between three diversity components (functional diversity, redundancy and dominance) across silvicultural regimes through a ternary diversity diagram. Intensive silvicultural regimes are associated with a decrease in functional diversity and an increase in functional redundancy, compared with unmanaged conditions. This means that although intensive management may buffer communities' functions against species or functional losses, it also limits the range of understory response to environmental changes. Policy implications. Different silvicultural regimes influence different facets of understory functional features. While unmanaged forests can be used as a reference to design silvicultural practices in compliance with biodiversity conservation targets, different silvicultural options should be balanced at landscape scale to sustain the multiple forest functions that human societies are increasingly demanding. © 2024 The Author(s). Journal of Applied Ecology published by John Wiley & Sons Ltd on behalf of British Ecological Society.