Know more

PhD defended february 1st, 2013
Direction: Philippe Lucas - Joao AN Filipe - Thierry Doré

Abstract:

Nowadays it is still difficult to predict and control the spread of soilborne diseases that cause substantial damage in crop systems. The aim of this epidemiological interdisciplinary work is to propose models for the spatio-temporal spread of soilborne pathogens in order to point out key parameters for the control of soilborne diseases. This thesis considers the spread ofRhizoctonia solanion sugar beet as an example pathosystem and focuses on three main problems. First, we use experimental measures of the dispersal of the pathogen to parameterise a stochastic spatially explicit model and we show that host growth can trigger the development of epidemics by causing a switch from non-invasive to invasive behaviour. Second, using experimental data we build an age-varying model for the distribution of the incubation period that links hidden infections and above-ground observations of the disease. Then, we investigate the cryptic behaviour of epidemics by using a hierarchical model that considers a realistic incubation period. Third, we use a spatially-implicit model to estimate rates of infection from temporal disease data, and, to analyse the effects of biofumigation on epidemics. These parameters are integrated into an individual-based model to predict the stochastic development of epidemics. Our results confirm that biofumigation only permits a partial control and suggest that this biological control reduces uncertainty of the cryptic development of the disease. To finish with, we discuss the results of the thesis and we present the perspectives of this work.