Modelling resistance deployment and chemical applications to identify reactive, efficient, durable and cost-efficient strategies to control plant diseases.

Contexte et enjeux

The sector of banana for export relies heavily on the systematic use of chemicals despite their negative impact on human health and environment. This is mostly due to black sigatoka (black leaf streak) caused by the fungus Pseudocercospora fijiensis. The system structural vulnerability stems from the near-exclusive monoculture of Cavendish, a group with very low genetic diversity and high susceptibility. Therefore, the long-term viability of banana cultivation in a context of pesticide reduction and withdrawal of chemical molecules, now depends on the deployment of resistant cultivars.

In the short term, the challenge is to design epidemiologically effective strategies to introduce and deploy such resistant cultivars. However, thanks to its strong evolutionary potential, the pathogen can overcome varietal resistance. Thus, strategies must also be evolutionarily durable and economically viable to be relevant and adopted by farmers.

Given the large spatiotemporal scale of this problematic (multi-year, basin-scale), field experimentation is considerably limited to explore all possible deployment options. Modelling is a powerful investigation tool. In particular, the demographic-genetic, stochastic model landsepi simulates pathogen spread and evolution in explicit landscapes to evaluate epidemiological, evolutionary, and economic performance of resistance-deployment strategies. 

Objectifs

The main objective of this PhD project is to identify strategies that are simultaneously epidemiologically effective, evolutionarily durable, and economically viable. We tackle this challenge through the case study of black sigatoka of banana in the French West Indies, where banana is a key economic sector and the widespread of the disease depresses yields and fruit quality while raising labour costs.

For this, we will extend the spatiotemporal simulation model landsepi, initially developed on another pathosystem. Two aspects of this project are particularly innovative. Firstly, the joint modelling of pathogen adaptation to both varietal resistance and fungicides. Secondly, the evaluation of adaptive strategies – i.e. policies updated in time based on epidemiological and evolutionary feedbacks – still rare in pest management and co-designed with farmers in order to deliver practical alternatives for black sigatoka management.

This global objective will lead to test these four hypotheses:

• pathogens adapt to control methods;
• the most epidemiologically effective strategies are not necessarily the most durable or cost-efficient, so optimality depends on the evaluation criterion;
• each cropping cycle, the optimal strategy depends on the initial epidemiological-evolutionary context;
• pathogen adaptation may be penalised by fitness cost that can be leveraged to prevent or slow down its adaptation to resistance and chemicals.