| Abstract: | Components of transmission for primary infection from soil-borne inoculum and secondary (plant to plant) infection are estimated from experiments involving single plants. The results from these individual-based experiments are used in a probabilistic spatial contact process (cellular automaton) to predict the progress of an epidemic. The model accounts for spatial correlations between infected and susceptible plants due to inhomogeneous mixing caused by restricted movement of the pathogen in soil. It also integrates nonlinearities in infection, including small stochastic differences in primary infection that become amplified by secondary infection. The model predicts both the mean and the variance of the infection dynamics of R. solani when compared with replicated epidemics in populations of plants grown in microcosms. The broader consequences of the combination of experimental and modelling approaches for scaling-up from individual to population behaviour are discussed. <br> |
