Projection of climate change impacts on beech and oak forests regeneration
Projet ANR REGEMAST
Supervision : CHUINE Isabelle, Directrice de recherche CNRS, CEFE, Montpellier, isabelle.chuine@cefe.cnrs.fr
Co-supervision : Nicolas Delpierre, Professeur Université Paris-Saclay, ESE, Saclay, nicolas.delpierre@universite-paris-saclay.fr
Collaboration : Nicolas Martin, Chargé de recherches INRAE, URFM, Avignon, nicolas.martin@inrae.fr
Place : CEFE, UMR CNRS 5175, Montpellier
Keywords: Tree ecophysiology models ; forest trees ; oak ; beech ; fecundity ; masting; phenology ; nutrient allocation ; climate change.
Working environment
You will be based at the CEFE laboratory in Montpellier. The CEFE is one of France’s leading research laboratories in ecology and evolution, focusing its activities on major societal concerns: biodiversity, global change and sustainable development. The aim is to establish scenarios for the evolution of ecological systems and strategies for their conservation, restoration or rehabilitation.
You will be working in the FORECAST team, which focuses on temperate and Mediterranean forests and is made up of ecophysiologists and modelers. In particular, the FORECAST team is developing various models used to project the geographical distribution, growth, survival and regeneration of Europe’s main forest species under future climatic conditions, as well as their capacity to adapt.
You will be supervised by Isabelle Chuine (DR CNRS, CEFE) and Nicolas Delpierre (Pr Paris Saclay). You will work for the ANR REGEMAST project and interact with project scientists from URFM (Avignon), LBBE (Lyon) and BIOGECO (Bordeaux) laboratories.
Project overview
Temperate forests have a pivotal role in global biogeochemical cycles and provide myriad ecological and economic benefits. Yet, their ability to withstand and adapt to the unprecedented pace and scale of climate change is a major concern. To ensure the long-term sustainability of forest ecosystems, strategic management decisions are urgently required. These decisions necessitate a deep understanding of the fundamental factors driving natural forest growth and regeneration. Yet, our progress in forest management practices is impeded by significant gaps in our comprehension of tree reproductive patterns, especially for species experiencing substantial inter-annual variations in fruiting intensity and population-wide synchronization, a phenomenon known as « masting ». This PhD project aims to provide projections of the fecundity, regeneration, growth and survival of beech and oak forests in France and in Europe using three ecophysiological models. It will be structured around 3 main questions: (i) Is the investment in reproduction each year related to source/sink dynamics of nutrients at flower initiation time? (ii) How does water availability affect carbon sequestration and allocation to reproduction versus growth (iii) is the effect of climate change stronger on fecundity than seedling recruitment? This project will be made part of the REGEMAST ANR project.
Objectives
The general objective of the PhD project is to produce a new version of the PHENOFIT5 model that will be able to predict the growth, fecundity, regeneration and survival of beech and oak forests in future climatic conditions. To reach this general objective, the PhD candidate will have to incorporate to the model (i) a daily allocation scheme of carbon, nitrogen and phosphorous based on the active sources and sinks and their phenology, (ii) a flower initiation as well as a growth onset and cessation phenology model and (iii) additional components of drought resistance. The model will then be used to generate projections of beech and oak forests growth, fecundity, regeneration and survival in France and Europe following climate scenarios SSP2(RCP4.5) and SSP5(RCP8.5) to address the above-mentioned questions.
Background
The approaches currently used to inform forest managers about the fate of forests in the following decades are statistical niche models. These models have different caveats 1, 2 among which a lower transferability in conditions not used to train them 3. Future climatic conditions are unlikely to be analogous to recent conditions, which raises questions about the reliability of simulations carried out under climate scenarios with such models4.Other more realistic ecophysiological models exist, and can provide much more information than a simple probability of presence of the species. Indeed, these models can simulate numerous ecophysiological processes which are regulated by the environmental conditions (temperature, light, photoperiod, water availability): photosynthesis, respiration, transpiration, nutrient allocation, growth, phenology. Therefore, they can provide predictions of primary productivity, tree growth, carbon sequestration. Some processes are however underrepresented in these models, especially those related to reproduction from flower initiation to seedling recruitment and to survival. Climate change has a strong impact on tree growth but also on tree fecundity and survival, but no model so far is able to provide projections of these three major components of fitness in future climatic conditions. We propose in this PhD project to build upon three existing ecophysiological models, each one having a detailed representation of one of the three components, to construct a model that will integrate all of them.
Method
The model will be developed from the PHENOFIT5 model (unpublished), derived from the PHENOFIT4 5-7 model, the CASTANEA 8-10 model and the SurEau 11 model. PHENOFIT and SurEau are developed in Java on the CAPSIS modeling platform. The PHENOFIT5 model simulates the precise development over time of the tree’s various vegetative and reproductive organs over the seasons, the ability to resist frost and water stress, photosynthesis, respiration and transpiration, and allocates net productivity between the leaf, flower-fruit and wood compartments. It also takes into account intra- and interspecific competition. All processes are simulated on a daily time step, with the exception of carbon allocation. The CASTEANA model is an ecophysiological model that simulates the cycles of carbon, water and nutrients of monospecific forests, with a strong focus on the seasonality of the processes. CASTANEA has proved able to simulate realistic time series of resource acquisition and allocation to growth in oak and beech forests, under contrasted environmental and silvicultural conditions 8-10. The SurEau model is a hydraulic model that explicitly represents water flows between the soil, the tree and the atmosphere, as well as the cavitation process that leads to tree desiccation and death 11.
PHENOFIT5 will be the pivotal model in the model coupling scheme. We will first incorporate three new phenology models in PHENOFIT5: a phenology model of floral induction, a phenology model of flowering time for beech that takes into account the interaction between temperature and photoperiod, and the wood phenology model of CASTANEA. The new phenology models will be calibrated and validated for beech and oak using existing phenology data from the TEMPO network and cambium activity data collected by other research groups that will be associated to the work. Second, we will leverage simulations from the CASTANEA model of the dynamics of carbon and nitrogen internal reserves to elucidate which of both resources is the rarest at the time of floral induction, and likely to serve as a trigger to determine the investment in flower production. Simulations from CASTANEA will be validated by comparing (1) the seasonal and interannual dynamics of tree growth (namely girth increment) with measurements from dendrometers, and (2) the seasonal and interannual dynamics of nonstructural carbon, nitrogen and phosphorous. Some measurements of nitrogen and phosphorous concentration will be realized to complete the scarce data available in the literature. Third, the effect of cuticular conductance on tree transpiration will be incorporated in PHENOFIT5 by coupling the SUREAU model. Fourth, a dynamic daily allocation of carbon and/or nitrogen will be added to PHENOFIT5 based on the developmental status of carbon sources and sinks (leaves, flowers/fruits and wood), and their water and carbon status. This is the water and carbon status of the various compartments, as well as their state of activity, all of which depend on daily temperature conditions, light, water availability, which will determine the daily allocation scheme of nutrients. The allocation scheme will build upon the daily allocation rules already existing in CASTANEA, as well as results of previous studies which have manipulated carbon sources (leaves) and sinks (fruits) in oak 13. This dynamic resource allocation scheme based on the daily states of the different sources and sinks will allow testing whether a tradeoff in daily resource allocation between vegetative and reproductive organs can generate masting patterns.
The current versions of PHENOFIT5, CASTANEA and SurEau are already parameterized for the main forest species. The calibration of the new version of PHENOFIT5 could use two distinct methodologies, depending on the data available. Some parameters may be measured directly in the field or laboratory, or calculated using data already available from other ongoing projects (DroughtForC, FAGRESCUE, REGEMAST, TAWTREE) and from the ONF-RENECOFOR forest monitoring network. Other parameters may need to be statistically inferred using data on integrative variables such as species presence/absence data and/or biomass or secondary growth time series using various inference methods. One method could be to use the CMA-ES optimization algorithm 12, another one could be to use a Bayesian approach which would 1) allow clean management of uncertainties, 2) flexibly to integrate the influences of all types of covariates of interest and 3) take into account, through the elicitation of priors, our knowledge of these parameters. Finally, the updated version of PHENOFIT5 will be validated by comparing the predictions of fruit biomass to observations of fruit biomass from the REGEMAST project and the MASTREE dataset, as well as leaf, wood and fruit biomass of the RENECOFOR network.
Results
The model developed, once validated, will be used to address the three questions presented in the general objective of the PhD project, but it could be used to address other questions, either theoretical or applied, that will be selected with the PhD candidate. For example, a sensitivity analysis of the model could enable identifying which physiological traits, and which trait values are most important to maintain the carbon storage capacity of trees under future climatic conditions; and which other ones are most important to maintain high fecundity and survival.
Collaborations
The PhD project will involve collaboration between three researchers, Isabelle Chuine, Nicolas Delpierre and Nicolas Martin (INRAE, URFM, Avignon). Isabelle Chuine (DR CNRS CEFE) has been developing the PHENOFIT model at the heart of the thesis project for some twenty years. Nicolas Delpierre (Pr UPS ESE) is a tree ecophysiologist and modeller, specialized in the seasonality of carbon acquisition processes and the influence of nutrient and water cycles on these processes. He develops the tree ecophysiology model CASTANEA. Nicolas Martin-St Paul (CR INRAE URFM Avignon) is a tree ecophysiologist and modeller, specializing in the hydric functioning of forest trees, and is developing the SurEau model. The project will be part of the ANR project REGEMAST supervised by Marie-Claude Venner (Associate professor at Univ. Lyon) which aims at elucidating the environmental and genetic determinant of masting in beech and oak. The PhD student will therefore also interact with the researchers involved in this project.
International scope
The PhD project does not involve international collaboration. However, the model developed in the project is of interest to different teams modeling the impact of climate change on forests around the world. There will be several opportunities of discussions with foreign scientists collaborating with the supervisors. It is also expected that the PhD candidate presents the results of the project at international conferences such as the joint SFE-BES meeting, the EGU and the ESA annual meeting.
Valorization of the research
The model will be available on the CAPSIS modeling platform. The work will be promoted in publications submitted to high-impact journals that respect Open Science principles. At least three publications in peer-reviewed journals are expected. A first publication will describe the new model obtained, and in particular the dynamic daily allocation of nutrients in the different compartments, and its validation for beech and oak in different pedoclimatic contexts. A second publication will present the specific influence of the tree’s water status on carbon storage in the different compartments. The third will present the model’s projections according to different climate scenarios on a national scale, and will establish the evolution from 1970 to 2100 of carbon fluxes and stocks in forest ecosystems, according to different climate scenarios and different silvicultural scenarios. It will also be promoted through presentations at international conferences (e.g. BES, ESA, EGU). A fourth one could address the adaptive capacity of beech and oak forests to climate change.
Supervision and training
The thesis work will be monitored through weekly face-to-face meetings (1-2 hours) with the supervisor, and monthly meetings (1h-2h) with the three co-supervisors in a mixed face-to-face and visioconference mode. Face-to-face meetings with all the supervisors will be scheduled according to the progress of the work and the predefined timetable.
Profile
Student graduated of a master degree in ecology with solid knowledge in plant ecophysiology and forest ecology. Skills in R, in writing and oral presentation required. Experience and strong interest in programming and modeling. Ability to work in a team. English level B2 required.
How to apply
Send a cover letter, a letter of reference and a CV to Isabelle Chuine, isabelle.chuine@cefe.cnrs.fr
References
1 Kaitlin C. Maguire, et al. Controlled comparison of species- and community-level models across novel climates and communities. Proceedings of the Royal Society B: Biological Sciences, 283(1826):20152817, March 2016. https://doi.org/0.1098/rspb.2015.2817
2 Matthew C. Fitzpatrick, et al. How will climate novelty influence ecological forecasts? Using the Quaternary to assess future reliability. Global Change Biology, 24(8):3575–3586, 2018. ISSN 1365-2486. https://doi.org/10.1111/gcb.14138
3 Van der Meersch V., E. Armstrong, F. Mouillot, A. Duputié, H. Davi, F. Saltré & I. Chuine (2025) Paleorecords reveal biological mechanisms crucial for reliable species range shift projections amid rapid climate change. Ecology Letters 28(2) e70080. https://doi.org/10.1111/ele.70080
4 Medlyn, B.E., Duursma, R.A. and Zeppel, M.J.B. (2011) Forest Productivity under Climate Change: A Checklist for Evaluating Model Studies. Wiley Interdisciplinary Reviews: Climate Change, 2, 332-355. https://doi.org/10.1002/wcc.108
5 Chuine, I., & Beaubien, E. G. (2001). Phenology is a major determinant of tree species range. Ecology Letters, 4(5), 500-510.
6 Saltré F, Duputié A, Gaucherel C, Chuine I (2015) How climate, migration ability and habitat fragmentation affect the projected future distribution of European beech. Global Change Biology, 21 (2), 897-910. https://doi.org/10.1111/gcb.12771
7 Gauzere J, Teuf B., Davi H., Chevin LM., Caignard T., Leys B., Delzon S., Ronce O., Chuine I. (2020) Where is the optimum? Predicting the variation of selection along climatic gradients and the adaptive value of plasticity. A case study on tree phenology. Evolution Letters 4(2): 109-123. 10.1002/evl3.160
8 Guillemot, J. et al. The dynamic of the annual carbon allocation to wood in European tree species is consistent with a combined source–sink limitation of growth:implications for modelling. Biogeosciences 12, 2773–2790 (2015).
9 Guillemot, J. et al. Environmental control of carbon allocation matters for modelling forest growth. New Phytol. 214, 180–193 (2017).
10 Delpierre, N. et al. Temperate and boreal forest tree phenology: from organ-scale processes to terrestrial ecosystem models. Ann. For. Sci. 73, 5–25 (2016).
11 Martin-StPaul, N., Delzon, S. & Cochard, H. Plant resistance to drought depends on timely stomatal closure. Ecology Letter. 20, 1437–1447 (2017).
12 Van der Mersch & Chuine (2023) Estimating process-based model parameters from species distribution data. Method Ecol Evol 14(7): 1808-1820 10.1111/2041-210X.14119
13 Le Roncé I., Dardevet E, Venner S., Schönbeck L., Gessler A., I. Chuine, JM. Limousin (2023) Reproduction alternation in trees: testing the resource depletion hypothesis using experimental fruit removal in Quercus ilex. Tree Physiology, 43(6), 952–964. 10.1093/treephys/tpad025
Commentaires récents