Multiscale functional-structural plant modelling at the example of apple trees – FSPM Apple

WP2:
3D sugar and water transport model for the apple carrier branch: model-based fruit quality analysis
• Characterization of carbon allocation in fully developed carrier branches, under different fruit loads and with or without girdling
• Model development: Implementation of water and sugar transport model

WP3:
Analysis and modelling of feedback between plant C and water relationships and apple tree growth. Consequences for modelling genotype x environment interaction
• Modelling genotype and water deficit effects on tree architecture and growth using MAppleT
• Analysis of C/water fluxes and growth dynamics under drought
• Development of an integrated model coupling MAppleT and a C/water source-sink model, calibration and validation

WP4:
Analysis, parameterization and statistical evaluation of multi-scale functional-structural plant models: methods and application to apple tree model
• A priori parameterization and sensitivity analysis for the apple tree models
• Parameter Estimation Methods and Algorithms for Multiscaled FSPMs
• Parameter Estimation and Model Selection for the Apple Tree Models

WP2: Study of the evolution of the gas exchange at the leaf scale, under low load, with or without girdling. Development of a photosynthesis model to simulate treatment effects on sugar production within the fruit bearing branch. In 2016, C13 labeling of photoassimilates to track transport within the branch, on two leaf types (rosette and bourse shoot). Measurement of the topology and the geographic distribution of the sources and sinks within the fruit bearing branch. This work will be used to model in 3D the branch and its C transport. Experiments on the manipulation of carbon will allow calibrating the transport model (GroIMP).
WP3: A water stress experiment was set up in the orchard, with three cultivars, with soil water sensors. A first topological dataset in MTG format was acquired, and the differential irrigation treatments were launched. Analysis of growth dynamics of principal axes of 120 hybrid individuals from a cross between cv. ‘Starkrimson’ and ‘Granny Smith’ were conducted. A search for QTL was carried out, on functional parameters representing architectural traits. This prediction phase and its link with MAppleT aim at the integration of genetic variability in the FSPM. Qinqin Long (WP1 Göttingen) contributed to a working interface between OpenAlea and GroIMP.
WP4: Statistical methods and algorithms are devised that allow the parametric identification of models, while characterizing their uncertainty. With this in mind, our research is focusing on Bayesian parameter estimation and sensitivity analysis (SA). Specifically, we translate models into hidden Markov chains and are thus confronted by the problem of joint parameter estimation and hidden states for non-linear models. We develop in this project variance decomposition methods for SA, truly adapted to correlated input parameters. Objective: use results of SA to aid model parameterization. Bayesian estimation and SA, are developed using PYGMALION, a modelling platform developed by us.

Halfway into the project, the outstanding features of this project are the good collaboration among the three ANR-funded WPs, as well as with the DFG funded WP1. This is firstly represented by the development of a functional interface between the two main modelling platforms, OpenAlea and GroIMP. Secondly, the setting up of a common experimental setup for field experiments, using the same cultivars (even same nursery origin) and newly planted trees, with an emphasis on water stress at WP3 (Montpellier) and on fruit load at WP2 (Angers), will allow a high degree of comparability of results as well as of mutual data use in the models of each WP. Accordingly, the results obtained from fruiting trees in 2017 will feed a common model based on MAppleT (OpenAlea: Montpellier) and SugarFlow (GroIMP: Angers). This common model and its precursors/constituting parts will be tested and streamlined using the optimization procedures offered by WP4 (Paris). What is more, thanks to the activities of WP3 this joint model will represent one of the first apple models to explicitly consider and process genetic information in morphogenetic rules, thereby considerably improving the quality of model predictions. In doing so, management of model and data complexity will be a future challenge in this project which can be mastered by applying advanced sensitivity analysis techniques for parameterization (WP4).

• Bayol B, Cournède P-H, Sainte-Marie J, Viaud G, Costes E, Migault V, Pallas B, Chi F, Kurth W, Long Q, Merklein J, Streit K, Buck-Sorlin GH, Poirier-Pocovi M, Pradal C. Multiscale Functional-Structural Plant Modelling at the Example of Apple Trees: Project description. IEE Proceedings of FSPMA conference, Qindao November 2016.
• Streit K., Henke M., Bayol B., Cournède P.-H., Sievänen R. and Kurth W. Impact of geometrical traits on light interception in conifers: analysis using an FSPM for Scots pine. Proceedings of FSPMA 2016.
• Pallas B., F. Reyes, C. Pradal, Costes E.,Multi-scale simulation of carbon transport in trees. Annual meeting of FSPMApple projet, 4 - 5 April, 2016, Göttingen
• Migault V., Pallas B., Costes E., Accounting for genetic and soil water restriction effects on shoot growth and organ morphology in simulations of 1-year-old apple trees. Annual meeting of FSPMApple project, 4-5 April, 2016, Göttingen
• Cournede PH, Streit K, Bayol B, Sainte-Marie J. Sensitivity Analysis as a Tool for the Parameterization of Complex Systems.
• Poirier M., Buck-Sorlin G. 2016. Modelling temporal variation of different parameters used in photosynthesis models of Thornley and Marshall & Biscoe: influence of fruit load and girdling. Workshop «Modelling tree growth and physiological processes at organ, branch and whole-tree scale«, 4 - 5 April, 2016, Göttingen, Germany
• Poirier M., Buck-Sorlin G. Modelling temporal variation of parameters used in two photosynthesis models: influence of fruit load and girdling in fruit bearing branches of apple trees. IEE Proceeding of International Conference on Functional-Structural Plant Growth Modeling, Simulation, Visualization and Applications (FSPMA 2016). 7-11 November, 2016, Qingdao, China