Search

International audience

International audience; Cette epopee doit beaucoup a la regrettee Nadine Brisson, premiere capitaine de« l' embarcation Sties ». Nous lui dedions avec emotion ce chapitre, dont les axes reprennent ceux de l'hommage qui lui a ete rendu le 16 octobre 2012 conjointement par le departement Environnement et Agronomie (EA) de l'Inra et !'Association frans;aise d' agronomie 123. Le modele Sties (Simulateur multidisciplinaire pour les cultures standard) a plusde vingt ans. Il fait maintenant partie des cinq modeles de culture les plus cites au monde. Cette success story s' est fondee sur un projet collectif dont le creuset a ete le departement EA naissant. Ce texte tente une analyse reflexive des interactions entre un groupe elargi de chercheurs modelisateurs et la necessite de mobiliser des moyens sur une tres longue duree, marquee par le renouvellement des questions posees face a l'evolution rapide des usages et des contextes d' application.

In Réunion, new sugarcane commercial cultivars selected for constrained environments, particularly with regard to low temperatures, have demonstrated their better adaptability by producing more biomass in the high-lying parts of the island. The germination dynamics of populations of single-bud setts were studied and characterized in climate chambers for three contrasting varieties at different constant temperatures. The emergence of setts as a function of time is faster and more complete for the variety R583, which is known to be better adapted to low temperatures than R570 and R577. Does the variety R583 selected under temperature-constrained environment perform better under such conditions as a result of its lower threshold temperature for germination? The objectives of our study were (i) to characterize the germination performance of three varieties under different temperature regimes, and (ii) to illustrate in the field the important simulated results developed with the help of mapping tools. A germination non-linear mixed model with random effects was developed from the degree-days data. This Weibull equation made it possible to determine a threshold temperature for the germination process for each variety. Using this germination threshold temperature, a simulation of the germination times was carried out for each zone on the west coast of Réunion based on weather data over 10 years. The two-degree difference between the base temperatures of varieties R583 and R570 shows the relative importance of this parameter on germination date, suggesting that the plant is able to establish more quickly. A comparison was made between the geographical limits obtained from the simulation of where the establishment of the variety R583 is possible given the time required for germination (in days), and the actual distribution of this variety.

In grassland modelling, understanding feedbacks between grassland ecosystems and the atmosphere in the context of regional scale climatic changes is essential for the accurate quantification of ecosystem water and carbon (C) fluxes. Different grassland models respond differently to environmental conditions and climatic circumstances. To test the sensitivity of different models to changes in input variables, ensemble modelling approaches are used because they generate an expanded envelope of possible systemic outputs. Here, an ensemble modelling approach was applied to explore water and C fluxes from grasslands in Europe. Seven grassland models were run at nine long-term grassland sites representing a broad gradient of geographic and climatic conditions. It was assessed the sensitivity to climate change factors including precipitation (P), temperature (T) and atmospheric CO2 concentration [CO2]. Baseline weather series (including [CO2]=380 ppm) were modified by changing T and P by -25%, -10%, -5%, +5%, +10%, +25% of the observed standard deviation and [CO2] by +5%, +10%, +15%, +25%, +50%, +100%. The obtained multi-model responses for each driver showed different levels of sensitivity. Soil temperature and gross primary production (GPP) displayed strong sensitivity to air temperature and precipitation. Based on the multi-model median of model responses, altered scenarios of precipitation had an important effect on modelled evapotranspiration from grassland swards. In general, yield biomass and GPP increased with elevated levels of [CO2]. Rising T and [CO2] had a fundamental effect on the C cycling of terrestrial ecosystems. This study demonstrates the use of ensemble modelling to address critical issues of uncertainty associated with individual model predictions, and provides increased understanding of water and C fluxes in grasslands under climate change.

This study developed Adaptation Response Surfaces and applied them to a study case in North East Spain on winter crops adaptation, using rainfed winter wheat as reference crop. Crop responses to perturbed temperature, precipitation and CO2 were simulated by an ensemble of crop models. A set of combined changes on cultivars (on vernalisation requirements and phenology) and management (on sowing date and irrigation) were considered as adaptation options and simulated by the crop model ensemble. The discussion focused on two main issues: 1) the recommended adaptation options for different soil types and perturbation levels, and 2) the need of applying our current knowledge (AOCK) when building a crop model ensemble. The study has been published Agricultural Systems (Available online 25 January 2017, https://doi.org/10.1016/j.agsy.2017.01.009), and the text below consists on extracts from that paper.

Impact response surfaces (IRSs) were constructed to depict the sensitivity of modelled spring and winter wheat yields to systematic changes in baseline temperature (between -2°C and +9°C) and precipitation (-50 to +50%) as simulated by a 26-member ensemble of process-based crop simulation models. The study was conducted across a latitudinal transect for sites in Finland, Germany and Spain. In spite of large differences in simulated yield responses to both baseline and changed climate between models, sites, crops and years, several common messages emerged. Ensemble average yields decline with warming (3-7% per 1°C) and decreased precipitation (3-9% per 10% decrease), but benefit from increased precipitation (0-8% per 10% increase). Yields are more sensitive to temperature than precipitation changes at the Finnish site while sensitivities are mixed at the other sites. Inter-model variability is highest for baseline climate at the Spanish site but is affected little by changed climate. Model responses diverge most under warming at the Finnish and German sites for winter wheat. The IRS pattern of yield reliability tracks average yield levels. Optimal temperatures for present-day cultivars are below the baseline at the German and Spanish sites suggesting that adoption of cultivars with higher temperature requirements might already be advantageous, and increasingly so at all sites under future warming. The study was conducted in the CropM component of the FACCE-JPI/MACSUR project.

Process based crop simulation models are widely used to assess crop production under current or future climate conditions. Most studies on climate impacts on crop growth are so far focussed on single crops and single-year simulations. However, it is known that the position of crops within a rotation can influence crop growth significantly due to carry-over effects between seasons. We compared crop models on crop rotation effects from five sites across Central Europe providing in total data of 301 cropping seasons and treatments. Treatments comprised irrigation, nitrogen (N) fertilisation, atmospheric [CO2], tillage, residue management, cover crops and soils. Crop rotations were simulated with 15 crop models as single-year simulations and/or continuous simulations over whole crop rotations in “restricted calibration” runs. Lower RMSE between observed and simulated crop yields were obtained for continuous runs as compared to single-year runs. Relatively low carry-over effects were observed due to equilibration of soil water over winter and high N fertilisation levels. Consistently, a sub-set of models applied to an additional rainfed Mediterranean site reproduced larger carry-over effects of soil water. Irrigation, N supply, cover crops and atmospheric [CO2] showed clearer effects than tillage and crop residue management. Model performance varied distinctly between crops showing the necessity to provide experimental data for model calibration also for less prominent crops.

Grassland models are valuable tools to test hypotheses on grassland ecosystem functioning. In the frame of FACCE MACSUR LiveM, a model intercomparison was conducted using a dataset from an observational and experimental network of nine multi-year grassland sites spread across Europe (France, Italy, Germany, Switzerland, The Netherlands, and United Kingdom) and Israel, and a suite of nine models to understand grassland functioning in the region. Grassland-specific approaches were compared to approaches mainly conceived to simulate crops and plant functional types. Model evaluation against actual measurements was performed before and after model calibration. The calibrated models were used to analyze their sensitivity to independent variations of temperature, precipitation and [CO2]. The results show to which extent calibration can accommodate model discrepancies. The sensitivity of simulated gross primary production to [CO2] and temperature is an important outcome, considering the fundamental effect of rising temperature and [CO2] on the C cycling of terrestrial ecosystems in the Euro-Mediterranean region. Overall, alternative models exhibit a different sensitivity to climate change factors, with different performances over different conditions. Explained by the basic processes of each model and also induced by different calibration methods, this difference is indicative that more models can be complementary and deliver greater insights than if they were applied individually.

The study ROTATIONEFFECT aims to compare the output of different models simulating field data sets with multi-year crop rotations including different treatments. Within the first Step (1a2a) data sets (comprising a total of 301 crop growth seasons) for 5 locations in Europe were distributed to 15 interested modeller groups. For this step only minimal information for calibration were provided to the modellers. In total 15 modelling teams sent their “uncalibrated” results as single-year calculations and/or continuous calculations of rotation depending on the capability of the model. Results have been evaluated and the paper submitted (European Journal of Agronomy). Now, within the 2nd step (1b2b) modellers were provided with more information on the crop for the calibration of models. Again, results of calibrated runs were collected. 6 models were capable to run the rotations as continuous runs and another set of 6 models provided single year simulations. A first overview of the improvement of predictions due to calibration has been produced. Result files have been uploaded to the web platform for CropM results at Aarhus University (Work package C2 – data management).

A La Réunion, de nouvelles variétés sélectionnées pour des environnements contraints, notamment vis-à-vis de températures basses font la preuve de leur meilleure adaptabilité en produisant plus de biomasse dans les Hauts de l'ile. La dynamique de levée de populations homogènes de boutures d'un seul bourgeon a été étudiée et caractérisée en chambres climatiques à différentes températures constantes, pour trois variétés contrastées. La levée des boutures en fonction du temps est plus rapide et plus complète pour la variété R583, qui est connue pour être mieux adaptée aux basses températures que pour R570 et R577. Se basant sur l'hypothèse d'une température seuil de fonctionnement plus basse, une simulation des durées de débourrement-levée est réalisée par zone, sur la côte Ouest de La Réunion à partir de données climatiques des 10 dernières années. Un écart de 2 degrés sur le seuil montre l'importance relative de ce paramètre sur la date de levée, ce qui laisse supposer une faculté de la plante à s'implanter plus rapidement. Un parallèle est fait entre les limites géographiques obtenues, où l'implantation à dire d'expert est possible au vu du temps nécessaire en jour pour la levée et celles qu'illustrent les zones effectives récentes de plantations de la variété R583. La bonne implantation d'une variété, sans garantir une meilleure production finale en biomasse et en sucre, représente un atout indéniable dans l'implantation de la culture, notamment pour la compétition vis-à-vis de la flore adventice. L'outil de cartographie, associé utilisant les nombreuses données météorologiques disponibles, a permis d'illustrer l'effet de 2 valeurs de températures seuils sur la durée nécessaire à la levée.

International audience

Estimation des besoins en eau des cultures de maïs: application régionale ou parcellaire. Le modèle de culture STICS a été bâti pour estimer la production et la consommation des cultures, et d'abord des grandes cultures telles que le maïs et le blé. Il est construit pour être utilisé pour des unités homogènes telles que des parcelles, mais ses paramètres sont conçus et calibrés de telle sorte qu'il peut être appliqué à des domaines plus vastes, à condition que l'alimentation en données d'entrée soit possible. L'application parcellaire ou à des unités homogènes intra-parcellaires sera présentée dans l'exposé suivant, tandis que cet exposé a pour but de présenter une application de STICS à l'estimation des besoins en eau des parcelles dans une région. Ce travail, nommé Cogito, consiste en quelques adaptations de STICS aux problèmes spécifiques de l'irrigation, une base de données d'entrée parmi lesquelles on sélectionne les caractéristiques parcellaires et une interface qui facilite l'utilisation de l'outil par rapport à la version standard de STICS. Une validation des estimations de quantités d'eau consommée a été faite sur plusieurs années et types de sol. Cogito est utilisé pour faire des estimations stratégiques (taille de réserve d'eau), pronostiques (effet d'apports en cours de culture sur le déroulement de la culture) et diagnostiques (analyse a posteriori du comportement de la culture).

Variabilité intra-parcellaire et maîtrise de l'irrigation: cas d'un pivot dans l'Orléanais. La prise en compte de l'hétérogénéité du sol est une nécessité pour maîtriser les impacts environnementaux de l'irrigation par aspersion en particulier dans les parcelles de grande dimension sous pivot qui se rencontrent aussi bien en France que dans les autres pays du Nord comme du Sud. Il existe une demande aussi bien des agriculteurs pour définir les modes de conduites, en particulier de modulations locales des apports, qui seraient susceptibles d'améliorer la valorisation et de réduire les pertes d'eau, que des constructeurs de matériel pour identifier les développements technologiques nécessaires. La connaissance de la variabilité du sol est souvent insuffisante, en particulier au sein des unités pédologiques considérées comme homogène. Une approche méthodologique a été développée en s'appuyant sur l'expérimentation et la modélisation pour le cas d'un pivot couvrant une parcelle de 50 ha de maïs de la petite Beauce, dans le centre de la France. Lors de la campagne agricole 2000, les apports d'eau ont été réduits sur une partie de la parcelle, en supprimant certaines irrigations sur une travée du pivot. Sur les deux principales unités de sol, la variabilité des caractéristiques hydriques du sol, considéré en tant que "réservoir", est décrite à partir de sondages à la tarière d'une part et d'un prospection géophysique d'autre part. La variabilité des rendements a été obtenue par pesée en continu lors de la récolte mécanique, et est confrontée aux grandeurs précédentes. Le modèle STICS (Brisson et al., 1998) est ensuite utilisé pour représenter les transferts et la production suite à une campagne d'irrigation. Sa mise en oeuvre a été réalisée de manière classique sur les unités principales de la parcelle puis à l'échelle intra-unité; elle fait apparaître l'aptitude de la modélisation à rendre compte de l'hétérogénéité du sol, suivant les contextes et à analyser l'intérêt et les conséquences de la modulation les apports d'eau.