Combining experimentation and modelling to estimate primary and secondary infections of take-all disease of wheat

Primary and secondary infections are important processes in the epidemiology of plant diseases but can be difficult to quantify experimentally as they often occur at the same time. This problem is all the more challenging in the case of soil-borne diseases, as most processes are hidden in the soil and destructive sampling is time-consuming and makes it difficult to obtain enough observations of disease progress. Here we show how a combination of experimentation and modelling can be used in order to obtain parameters for primary and secondary infections for take-all disease of wheat. First, an experiment with one infected seedling and varying numbers of target seedlings allowed us to estimate the probability of secondary infection by growth of the mycelium through the soil and by growth via the crown of the plant. Several equations were tested for the contact term between susceptible and infectious roots. Secondly, an experiment with primary inoculum placed at different depths allowed us to estimate the probability of primary infection, taking into account secondary infections and the time needed for the roots to reach inoculum depth. In both experiments, the use of simple models was effective in isolating the desired effect from uncontrollable effects occurring in the soil. The probability of secondary infection through the crown was higher than the probability of infection through soil, and the contact term following the mass action or Reed-Frost equation gave a better fit to the data than the other equations tested. The probability of primary infection was higher when inoculum was placed just below the soil surface than when it was placed deeper in the soil.