POPULATION-DYNAMICS OF BOTANICAL EPIDEMICS INVOLVING PRIMARY AND SECONDARY INFECTION

In this paper we study the dynamical properties of models for botanica l epidemics, especially for soil-borne fungal infection. The models de velop several new concepts, involving dual sources of infection, host and inoculum dynamics. Epidemics are modelled with respect to the infe ction status of whole plants and plant organs (the G model) or to lesi on density and size (the SW model). The infection can originate in two sources, either from the initial inoculum (primary infection) or by a direct transmission between plant tissue (secondary infection). The f irst term corresponds to the transmission through the free-living stag es of macroparasites or an external source of infection in certain med ical models, whereas the second term is equivalent to direct transmiss ion between the hosts in microparasitic infections. The models allow f or dynamics of host growth and inoculum decay. We show that the two mo dels for root and lesion dynamics can be derived as special cases of a single generic model. Analytical and numerical methods are used to an alyse the behaviour of the models for static, unlimited (exponential) and asymptotically limited host growth with and without secondary infe ction, and with and without decay of initial inoculum. The models are shown to exhibit a range of epidemic behaviour within single seasons t hat extends from simple monotonic increase with saturation of the host population, through temporary plateaux as the system switches from pr imary to secondary infection, to effective elimination of the pathogen by the host outgrowing the fungal infection. For certain conditions, the equilibrium values are shown to depend on initial conditions. Thes e results have important consequences for the control of plant disease . They can be applied beyond soil-borne plant pathogens to mycorrhizal fungi and aerial pathogens while the principles of primary and second ary infection with host and inoculum dynamics may be used to link clas sical models for both microparasitic and macroparasitic infections.