• Energy Research

After modeling the large-scale climate response patterns of leaf unfolding, leaf coloring and growing season length of evergreen and deciduous French temperate trees, we predicted the effects of eight future climate scenarios on phenological events. We used the ground observations from 103 temperate forests (10 species and 3,708 trees) from the French Renecofor Network and for the period 1997-2006. We applied RandomForest algorithms to predict phenological events from climatic and ecological variables. With the resulting models, we drew maps of phenological events throughout France under present climate and under two climatic change scenarios (A2, B2) and four global circulation models (HadCM3, CGCM2, CSIRO2 and PCM). We compared current observations and predicted values for the periods 2041-2070 and 2071-2100. On average, spring development of oaks precedes that of beech, which precedes that of conifers. Annual cycles in budburst and leaf coloring are highly correlated with January, March-April and October-November weather conditions through temperature, global solar radiation or potential evapotranspiration depending on species. At the end of the twenty-first century, each model predicts earlier budburst (mean: 7 days) and later leaf coloring (mean: 13 days) leading to an average increase in the growing season of about 20 days (for oaks and beech stands). The A2-HadCM3 hypothesis leads to an increase of up to 30 days in many areas. As a consequence of higher predicted warming during autumn than during winter or spring, shifts in leaf coloring dates appear greater than trends in leaf unfolding. At a regional scale, highly differing climatic response patterns were observed.

A broad consensus has been reached on the need to adapt the management of our forests to the context of the rapidly changing climate, which resulted in the development of numerous models capable of simulating the impact of the climate change on the forest. The primary goal of this specific endeavor is to propose a novel framework of comparative analysis which could lead to the unique and universal description and mapping of these models. This framework is based on the reduction of the model output to the relatively simplistic information about the presence of the tree species suitable for the forest management i.e. - a binary classifier, making it comparable with the largely available tree presence observations. The framework we propose comes along with a new score, based on the joint use of the Principal Component Analysis and the Co-inertia Analysis, which evaluates the model vis-`avis the corresponding observations with the focus on its phase space dynamics i.e. its dependence on external environmental variables, rather than its spatial precision. The pertinence of the proposed multi-scale approach, suitable for the multi-scale analysis, is demonstrated by conjointly using prototype binary classifiers, designed for this purpose, and two different examples of binary classifiers used in the forest management - climate-dependent tree species distribution models. This work has the ambition to serve as the basis for a potential combination of different models at different spatial scales in order to improve the decision making process in the forest management.

Changes in temperature and rainfall linked to recent climate change increase the mortality rates of European temperate tree species. The economic importance of trees and the ecosystem services they provide differ according to their social status (dominant or suppressed trees) and their size. The extent to which climate change impacts these different categories in different ways remains little explored. Ecophysiological differences between tree size and status suggest different sensitivities to climate change. Dominant trees are exposed to more evapotranspiration than suppressed trees that benefit from buffered climatic conditions. Large trees are able to develop a network of fine roots that allow deeper water and nutrient uptake during water shortage periods, but that have higher water requirements and more physical constraints than small trees due to the fact that they must lift water to greater heights. We used 207,100 trees from the French forest inventory data (including 3,514 dead trees), representing eight common European tree species. For each species, we separated the tree population into three subsets of suppressed, small dominant and large dominant trees. For each subset, we modelled the mortality observed in a stand in the absence of disturbances (background mortality), with a focus on the differences in sensitivity to recent changes in temperature and rainfall. After having taken the main mortality drivers related to competition into account, as well as stand characteristics including logging intensity effect, we assessed the over-mortality linked to the recent changes in temperature and rainfall for each of the three subsets. When considering both changes in temperature and rainfall, the climate change related to over-mortality was greater for suppressed than for small or large dominant trees, for all the species. Over-mortality of suppressed trees was related to temperature increase, whereas a maximum vulnerability related to rainfall decrease was observed for large dominant trees. Over-mortality driven by climate change not only concerns large and dominant trees, but small and especially suppressed ones as well. These results suggest that in addition to wood production, forest renewal and ecosystem services associated with understorey vegetation are threatened by the recent changes in temperature and rainfall in European temperate forests.

Face au réchauffement climatique en cours, qui entraine un déclin des forêts dans différents biomes de la planète, un besoin important de connaissances et d’outils se fait sentir pour adapter les peuplements aux nouvelles conditions environnementales et à leurs évolutions à venir. La cartographie prédictive peut permettre la production d’une grande quantité d’informations, à fine résolution spatiale, pour différentes périodes de temps, et sur de vastes emprises géographiques. Nous détaillerons l’intérêt des cartes prédictives des facteurs du milieu décrivant les propriétés du sol et du climat, principalement réalisées à l’échelle nationale ou régionale, qui peuvent être combinées avec des observations de terrain et/ou des données issues d’images satellitales pour progresser dans la connaissance de l’écologie des espèces d’arbres. Nous montrerons également comment ces données peuvent être utilisées pour évaluer et cartographier la vulnérabilité des essences au regard des changements en cours, permettant d’identifier le niveau de risque en fonction du peuplement en place et du contexte environnemental. La dynamique spatiale et temporelle de la quantité d’eau disponible pour les plantes est un facteur essentiel pour évaluer ces risques, dont les effets sont fortement modulés en fonction de la composition et de la structure des peuplements en place. L’identification des marges de manœuvre avant que les limites écologiques des espèces ne soient atteintes est aujourd’hui un enjeu crucial pour aider les acteurs de la gestion forestière à adapter les différents types de peuplement de nos forêts au changement de climat. To cope with ongoing global warming, which is leading to a decline in forests in different biomes on the planet, there is a significant need for knowledge and tools to adapt stands to new environmental conditions and their future evolutions. Predictive mapping can enable the production of a large amount of information at fine spatial resolution, for different periods of time, over broad geographic areas. We’ ll detail the interest of predictive maps of environmental factors concerning soil and climate properties, mainly elaborated at national or regional scale, which can be combined with field observations and/or data from satellite images to improve our knowledge about the ecology of tree species. We’ll also show how they can be used to assess and map the vulnerability of the tree species to ongoing changes, identifting the level of risk based on information about the stands characteristics and the environmental context. The spatial and temporal dynamics of the water available to plants is a crucial factor for assessing these risks, with variations according to the stands composition and structure. Identifying available margins before the ecological limits of species are reached is today a crucial issue to help forest management stakeholders to adapt the different stands of our forests to climate change.

AbstractAimSoil water is essential for the physiological processes of plant growth and fitness. Owing to the difficulty of assessing wide variations in soil water reserves, plant distribution models usually estimate available water for plants through such climatic proxies as precipitation data (P) or climatic water balance (P minus potential evapotranspiration). We evaluated the ability of simple climatic proxies and soil water balance indices to predict the ecological niches of forest tree species.LocationFrance.MethodsSoil water content and deficits were computed and mapped at a resolution of 1 km × 1 km throughout France. The predictive abilities of these indices were compared with those of P and climatic water balance to model the distributions of 37 of the most common European tree species. We focused on two species with contrasting water tolerance, Quercus robur and Quercus pubescens, to illustrate the differences between climatic proxies and soil water balance in species response curves and distribution maps.ResultsThroughout France, soil water content was poorly correlated with P and climatic water balance, because low P in the lowlands can be compensated for by water provided by deeper soils, which is not the case in most mountainous areas. Soil water balance performed better than simple climatic water variables for explaining tree species distribution, improving 82% of the models for hygrophilous, meso‐hygrophilous, meso‐xerophilous and xerophilous species.Main conclusionsOur results showed that simple climatic values do not accurately represent available water for trees and that soil water balance indices perform better than do climatic proxies for most species. This point is crucial to avoid underestimating the importance of water in studies aimed at determine the ecological niches of plant species and their responses to climate change.

Increases in tree mortality rates have been highlighted in different biomes over the past decades. However, disentangling the effects of climate change on the temporal increase in tree mortality from those of management and forest dynamics remains a challenge. Using a modelling approach taking tree and stand characteristics into account, we sought to evaluate the impact of climate change on background mortality for the most common European tree species. We focused on background mortality, which is the mortality observed in a stand in the absence of abrupt disturbances, to avoid confusion with mortality events unrelated to long-term changes in temperature and rainfall. We studied 372 974 trees including 7312 dead trees from forest inventory data surveyed across France between 2009 and 2015. Factors related to competition, stand characteristics, management intensity, and site conditions were the expected preponderant drivers of mortality. Taking these main drivers into account, we detected a climate change signal on 45% of the 43 studied species, explaining an average 6% of the total modelled mortality. For 18 out of the 19 species sensitive to climate change, we evidenced greater mortality with increasing temperature or decreasing rainfall. By quantifying the mortality excess linked to the current climate change for European temperate forest tree species, we provide new insights into forest vulnerability that will prove useful for adapting forest management to future conditions.

Plant ecologists have recognized the importance of solar radiation for decades but have difficulty measuring it on plots. Proxies recorded on the ground or geographical information system (GIS) indices processed with a digital elevation model (DEM) have generally been used. Here we compare the efficiency of different methods of estimating spatially distributed topographic solar radiation, from the simplest ones (proxies based on slope, and sine or cosine transformed values of aspect) to more elaborate ones using a GIS program suited to calculations of monthly clear sky and overcast solar radiation. We used a 50-metre DEM to estimate solar radiation with these different methods for the whole of France (550 000 km²). Radiation indices were compared with ground measurements from meteorological stations and used to model the distribution of silver fir (Abies alba), sycamore (Acer pseudoplatanus), and downy oak (Quercus pubescens), forest species known to be sensitive to light. Results show that sine and cosine of aspect, combined or not with slope, are inefficient at simulating solar radiation over large areas. Solar radiation, calculated for clear sky and especially including cloud cover, is more relevant, leading respectively to an R² of 0.46 and 0.78 between measured and predicted annual radiation. Calculation with cloud cover appears to be the most efficient index for improving distribution models for the three species studied. Slope and aspect transformations are less efficient than the GIS calculations, but the difference between these proxies decreased on a local scale. Using both with GIS solar radiation, cosine of aspect, with or without interaction with slope, slightly improves distribution models on a local scale, but this effect attenuates with increase in area studied. We conclude that the effect of proxies studied is scale-dependent, but GIS-based calculation including cloudiness variability is more appropriate than topographic proxies or clear sky models in estimating solar radiation and improving the efficiency of plant distribution models.