1
|
García-Hidalgo M, García-Pedrero Á, Rozas V, Sangüesa-Barreda G, García-Cervigón AI, Resente G, Wilmking M, Olano JM. Tree ring segmentation using UNEt TRansformer neural network on stained microsections for quantitative wood anatomy. FRONTIERS IN PLANT SCIENCE 2024; 14:1327163. [PMID: 38259935 PMCID: PMC10800830 DOI: 10.3389/fpls.2023.1327163] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 10/24/2023] [Accepted: 12/14/2023] [Indexed: 01/24/2024]
Abstract
Forests are critical in the terrestrial carbon cycle, and the knowledge of their response to ongoing climate change will be crucial for determining future carbon fluxes and climate trajectories. In areas with contrasting seasons, trees form discrete annual rings that can be assigned to calendar years, allowing to extract valuable information about how trees respond to the environment. The anatomical structure of wood provides highly-resolved information about the reaction and adaptation of trees to climate. Quantitative wood anatomy helps to retrieve this information by measuring wood at the cellular level using high-resolution images of wood micro-sections. However, whereas large advances have been made in identifying cellular structures, obtaining meaningful cellular information is still hampered by the correct annual tree ring delimitation on the images. This is a time-consuming task that requires experienced operators to manually delimit ring boundaries. Classic methods of automatic segmentation based on pixel values are being replaced by new approaches using neural networks which are capable of distinguishing structures, even when demarcations require a high level of expertise. Although neural networks have been used for tree ring segmentation on macroscopic images of wood, the complexity of cell patterns in stained microsections of broadleaved species requires adaptive models to accurately accomplish this task. We present an automatic tree ring boundary delineation using neural networks on stained cross-sectional microsection images from beech cores. We trained a UNETR, a combined neural network of UNET and the attention mechanisms of Visual Transformers, to automatically segment annual ring boundaries. Its accuracy was evaluated considering discrepancies with manual segmentation and the consequences of disparity for the goals of quantitative wood anatomy analyses. In most cases (91.8%), automatic segmentation matched or improved manual segmentation, and the rate of vessels assignment to annual rings was similar between the two categories, even when manual segmentation was considered better. The application of convolutional neural networks-based models outperforms human operator segmentations when confronting ring boundary delimitation using specific parameters for quantitative wood anatomy analysis. Current advances on segmentation models may reduce the cost of massive and accurate data collection for quantitative wood anatomy.
Collapse
Affiliation(s)
| | - Ángel García-Pedrero
- Department of Computer Architecture and Technology, Universidad Politécnica de Madrid, Madrid, Spain
- Center for Biomedical Technology, Universidad Politécnica de Madrid, Madrid, Spain
| | - Vicente Rozas
- iuFOR, EiFAB, Universidad de Valladolid, Soria, Spain
| | | | | | - Giulia Resente
- Institute of Botany and Landscape Ecology, University Greifswald, Greifswald, Germany
- Department DISAFA, University of Torino, Torino, Italy
| | - Martin Wilmking
- Institute of Botany and Landscape Ecology, University Greifswald, Greifswald, Germany
| | | |
Collapse
|
2
|
Tonelli E, Vitali A, Malandra F, Camarero JJ, Colangelo M, Nolè A, Ripullone F, Carrer M, Urbinati C. Tree-ring and remote sensing analyses uncover the role played by elevation on European beech sensitivity to late spring frost. THE SCIENCE OF THE TOTAL ENVIRONMENT 2023; 857:159239. [PMID: 36208754 DOI: 10.1016/j.scitotenv.2022.159239] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 04/22/2022] [Revised: 09/03/2022] [Accepted: 10/01/2022] [Indexed: 06/16/2023]
Abstract
Extreme climate events such as late spring frosts (LSFs) negatively affect productivity and tree growth in temperate beech forests. However, detailed information on how these forests recover after such events are still missing. We investigated how LSFs affected forest cover and radial growth in European beech (Fagus sylvatica L.) populations located at different elevations at four sites in the Italian Apennines, where LSFs have been recorded. We combined tree-ring and remote-sensing data to analyse the sensitivity and recovery capacity of beech populations to LSFs. Using daily temperature records, we reconstructed LSF events and assessed legacy effects on growth. We also evaluated the role played by elevation and stand structure as modulators of LSFs impacts. Finally, using satellite images we computed Normalized Difference Vegetation Index (NDVI), Enhanced Vegetation Index (EVI) and LAI (Leaf Area Index) to evaluate the post-LSF canopy recovery. The growth reduction in LSF-affected trees ranged from 36 % to 84 %. We detected a negative impact of LSF on growth only during the LSF year, with growth recovery occurring within 1-2 years after the event. LSF-affected stands featured low vegetation indices until late June, i.e. on average 75 days after the frost events. We did not find a clear relationship between beech forest elevation and occurrence of LSFs defoliations. Our results indicate a high recovery capacity of common beech and no legacy effects of LSFs.
Collapse
Affiliation(s)
- Enrico Tonelli
- Department of Agricultural, Food and Environmental Sciences, Marche Polytechnic University, Ancona, Italy
| | - Alessandro Vitali
- Department of Agricultural, Food and Environmental Sciences, Marche Polytechnic University, Ancona, Italy.
| | - Francesco Malandra
- Department of Agricultural, Food and Environmental Sciences, Marche Polytechnic University, Ancona, Italy
| | - J Julio Camarero
- Instituto Pirenaico de Ecología (IPE, CSIC), Apdo. 202, 50192 Zaragoza, Spain
| | - Michele Colangelo
- Instituto Pirenaico de Ecología (IPE, CSIC), Apdo. 202, 50192 Zaragoza, Spain; School of Agricultural, Forest, Food and Environmental Sciences (SAFE), University of Basilicata, 85100 Potenza, Italy
| | - Angelo Nolè
- School of Agricultural, Forest, Food and Environmental Sciences (SAFE), University of Basilicata, 85100 Potenza, Italy
| | - Francesco Ripullone
- School of Agricultural, Forest, Food and Environmental Sciences (SAFE), University of Basilicata, 85100 Potenza, Italy
| | - Marco Carrer
- Universitá degli Studi di Padova, Dipartimento Territorio e Sistemi Agro-Forestali (TeSAF), Viale dell'Università 16, 35020 Legnaro, Italy
| | - Carlo Urbinati
- Department of Agricultural, Food and Environmental Sciences, Marche Polytechnic University, Ancona, Italy
| |
Collapse
|
3
|
García‐Hidalgo M, García‐Pedrero Á, Colón D, Sangüesa‐Barreda G, García‐Cervigón AI, López‐Molina J, Hernández‐Alonso H, Rozas V, Olano JM, Alonso‐Gómez V. CaptuRING
: A
Do‐It‐Yourself
tool for wood sample digitization. Methods Ecol Evol 2022. [DOI: 10.1111/2041-210x.13847] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/30/2022]
Affiliation(s)
| | - Ángel García‐Pedrero
- Department of Computer Architecture and Technology, Universidad Politécnica de Madrid, E‐28660 Madrid Spain
- Center for Biomedical Technology, Universidad Politécnica de Madrid, E‐28223 Madrid Spain
| | - Daniel Colón
- Department of Computer Architecture and Technology, Universidad Politécnica de Madrid, E‐28660 Madrid Spain
| | | | - Ana I. García‐Cervigón
- Biodiversity and Conservation Area, Universidad Rey Juan Carlos, c/Tulipán s/n, E‐28933 Móstoles Spain
| | | | - Héctor Hernández‐Alonso
- EiFAB‐iuFOR, Universidad de Valladolid, Campus Duques de Soria, E‐42004 Soria Spain
- Area of Ecology, Faculty of Biology, Universidad de Salamanca, E‐37007 Salamanca Spain
| | - Vicente Rozas
- EiFAB‐iuFOR, Universidad de Valladolid, Campus Duques de Soria, E‐42004 Soria Spain
| | - José Miguel Olano
- EiFAB‐iuFOR, Universidad de Valladolid, Campus Duques de Soria, E‐42004 Soria Spain
| | - Víctor Alonso‐Gómez
- Department of Applied Physics‐EiFAB, Universidad de Valladolid, Campus Duques de Soria, E‐42004 Soria Spain
| |
Collapse
|
4
|
Sangüesa-Barreda G, Di Filippo A, Piovesan G, Rozas V, Di Fiore L, García-Hidalgo M, García-Cervigón AI, Muñoz-Garachana D, Baliva M, Olano JM. Warmer springs have increased the frequency and extension of late-frost defoliations in southern European beech forests. THE SCIENCE OF THE TOTAL ENVIRONMENT 2021; 775:145860. [PMID: 33631566 DOI: 10.1016/j.scitotenv.2021.145860] [Citation(s) in RCA: 13] [Impact Index Per Article: 4.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 10/23/2020] [Revised: 02/10/2021] [Accepted: 02/10/2021] [Indexed: 06/12/2023]
Abstract
Climate change is increasing the frequency of extreme climate events, causing profound impacts on forest function and composition. Late frost defoliation (LFD) events, the loss of photosynthetic tissues due to low temperatures at the start of the growing season, might become more recurrent under future climate scenarios. Therefore, the detection of changes in late-frost risk in response to global change emerges as a high-priority research topic. Here, we used a tree-ring network from southern European beech (Fagus sylvatica L.) forests comprising Spain, Italy and the Austrian Alps, to assess the incidence of LFD events in the last seven decades. We fitted linear-mixed models of basal area increment using different LFD indicators considering warm spring temperatures and late-spring frosts as fixed factors. We reconstructed major LFD events since 1950, matching extreme values of LFD climatic indicators with sharp tree-ring growth reductions. The last LFD events were validated using remote sensing. Lastly, reconstructed LFD events were climatically and spatially characterized. Warm temperatures before the late-spring frost, defined by high values of growing-degree days, influenced beech growth negatively, particularly in the southernmost populations. The number of LFD events increased towards beech southern distribution edge. Spanish and the southernmost Italian beech forests experienced higher frequency of LFD events since the 1990s. Until then, LFD events were circumscribed to local scales, but since that decade, LFD events became widespread, largely affecting the whole beech southwestern distribution area. Our study, based on in-situ evidence, sheds light on the climatic factors driving LFD occurrence and illustrates how increased occurrence and spatial extension of late-spring frosts might constrain future southern European beech forests' growth and functionality. Observed alterations in the climate-phenology interactions in response to climate change represent a potential threat for temperate deciduous forests persistence in their drier/southern distribution edge.
Collapse
Affiliation(s)
| | - Alfredo Di Filippo
- Department of Agriculture and Forest Sciences (DAFNE), University of Tuscia, Viterbo, Italy
| | - Gianluca Piovesan
- Department of Agriculture and Forest Sciences (DAFNE), University of Tuscia, Viterbo, Italy
| | - Vicente Rozas
- EiFAB-iuFOR, University of Valladolid, Campus Duques de Soria, 42004 Soria, Spain
| | - Luca Di Fiore
- Department of Agriculture and Forest Sciences (DAFNE), University of Tuscia, Viterbo, Italy
| | | | - Ana I García-Cervigón
- Biodiversity and Conservation Area, Rey Juan Carlos University, c/Tulipán s/n, E-28933 Móstoles, Spain
| | | | - Michele Baliva
- Department of Agriculture and Forest Sciences (DAFNE), University of Tuscia, Viterbo, Italy
| | - José M Olano
- EiFAB-iuFOR, University of Valladolid, Campus Duques de Soria, 42004 Soria, Spain
| |
Collapse
|