The long-term impact of pest and disease attacks on plant growth and production remains a major issue for the prediction of agricultural production. In order to address this issue, a modelling-based approach is used, mobilising models of pest development, plant growth and their interactions. In this thesis, we propose the development of a formalism enabling models to be coupled in order to integrate their interactions while minimising structural modifications to each model. The main application of this work concerns the evaluation of robusta coffee production in Uganda, exposed to attacks by pests such as CBB (Coffee Berry Borer), BCTB (Black coffee Twig Borer) and RBD (Red Blister Disease). Chapter One is devoted to formalising the coupling and implementing it in our context of biological population dynamics (plant organs, pests, pathogens). We adopt a cohort-based approach to structuring interactions between plants and pests. In this way, we limit the complexity that could arise from the unprecedented use of different types of models. Chapter Two deals with the modelling of pest dynamics. It proposes a modular framework for describing the dynamics of bark beetle-type pests, insects that spend their entire life cycle within a plant organ (except for the dispersal phase). This framework is based on two distinct modules: (a) one describes the life cycle, integrating the different developmental stages and viability of the pest, (b) the other describes interactions with the environment, taking into account the biotic and abiotic factors influencing the development and viability of the pest. Chapter three is devoted to the development of a plant growth model based on the GreenLab formalism. It models plant structure and function, and can be extended to the crop scale by factoring organs into cohorts. To take better account of interactions on plant structure, an extension of cohorts is introduced via a structural history. This development retains the advantages of factorisation while integrating the effects of pest attacks on new and existing cohorts. We introduce regulatory or control functions at the level of individual organs and the whole plant, enabling us to model a wide spectrum of plant responses to external effects. The following chapters focus on the application of the approaches developed to the Ugandan study field. First, the parameterisation of the models is discussed, based on the literature and on specifically developed experimental protocols. The analytical basis for model calibration is then detailed, taking into account growing conditions, in particular climatic parameters and factors linked to technical itineraries. Chapter Five uses simulations to analyse results and compare predictions with experimental data and literature. Emphasis is placed on the consequences of the modelled dynamics for the plant, in particular by assessing the effects of environmental and biotic constraints. Finally, a general discussion is offered on both the results obtained and the computational aspects. It also draws on the conclusions and partial discussions of the previous chapters.
- Interactions between pest attacks and plant growth using a model approach applied to Robusta coffee in Uganda : effects on production
- Houssem Triki - Université de Montpellier, AMAP - BotAnique et Modélisation de l'Architecture des Plantes et des végétations
- Marc JaegerJean-Noël Aubertot [Président]Jacques Gignoux [Rapporteur]Véronique Letort-Le Chevalier [Examinatrice]David Hill [Examinateur]Fabienne Ribeyre
- École Doctorale Information, Structures, Systèmes
- Doctoral
- 99259560309311
- AMAP - BotAnique et Modélisation de l'Architecture des Plantes et des végétations
- English
- Theses and HDR
- tel-05689288