Modelling the behaviour of crown rot in wheat caused by Fusarium pseudograminearum

Existing spatiotemporal models for disease progress are unsuitable for describing epidemics of residue-borne diseases of cereals like crown rot, caused by Fusarium pseudograminearum, which increase in incidence over successive years and where inoculum has a fixed location. Monte Carlo simulation was used to explore the behaviour of a simple mechanistic model for crown rot, in which inoculum from each infected plant in one season came into contact with a defined number of plants in the following season. Contacted plants became infected according to a probability of infection that was less than 1. Growth chamber experiments confirmed that contact between infested residue and plants was required for infection. Observed features of epidemics at a long term trial site, including disease progress and maximum incidence, were satisfactorily fitted by simulations with a starting incidence of 4 %, 12 plants contacted by inoculum from each infected plant, and probability of infection of 0.66. This confirms that our understanding of disease behaviour at moderate to high levels of incidence is basically correct. However, infection rate between seasons was sensitive to initial incidence, suggesting that dispersal methods such as ascospores or conidia may be important at low incidence. A simple descriptive model was developed for the disease in bread wheat in the northern grains region of eastern Australia. Inoculum potential was estimated from the square root of the product of crown rot incidence and yield (a surrogate for biomass) of the preceding crop. Incidence of disease was related to inoculum potential by an infection constant. Crown rot inoculum was assumed to decline exponentially with time between susceptible crops. These two relationships were combined to predict the behaviour of the disease in different rotation systems. Over time, crown rot incidence converged to an equilibrium level that was determined by the infection constant, average yield, and the survival rate of inoculum. Modelled epidemics behaved in a way that was consistent with field data for continuous wheat, wheat-chickpea and wheat-sorghum rotations. This descriptive model could be used in extension or as a way of understanding how management and environment affect the disease.