THE POPULATION-DYNAMICS OF ANNUAL PLANTS AND SOIL-BORNE FUNGAL PATHOGENS

1 Soil-borne fungi are a major group of economically important plant p athogens, yet they have rarely been studied in the context of host-pat hogen population biology. We develop general models of annual hosts an d soil-borne fungal pathogens to explore the conditions for host-patho gen coexistence in both agricultural and natural plant populations. We use empirical data from the literature to parameterize and simulate d ynamics with these models. 2 Initially we consider a simple system in which host density is assumed to be constant, as would be appropriate for agricultural systems. Model analysis shows that initial increase o f the pathogen population requires that host (crop) density be above a threshold, this threshold decreases with increases in the pathogen's over-winter survival rate and ability to grow saprophytically. 3 A mor e complex model, in which both host and pathogen populations can vary, is needed for natural populations. Results from this model show that stable coexistence is possible even when the pathogen has a positive i ntrinsic growth rate (and therefore it is also possible for the pathog en to persist in the complete absence of the host). 4 Model parameter estimates were obtained from the empirical literature for two common a nd important soil pathogens: Phytophthora spp. and Fusarium oxysporum; these pathogens differ in several life-history features. Computer sim ulation showed that for Fusarium, there were substantial ranges for wh ich coexistence or loss of the pathogen were predicted, while for Phyt ophthora most parameter estimates resulted in complete extinction when linear disease transmission was assumed; under the assumption of expo nential disease transmission, predicted dynamics were most likely to l ead to host persistence. 5 For both pathogens, within biologically rea listic regions of parameter space, small changes in parameter values c ould lead to qualitatively different outcomes, including deterministic chaos, suggesting that long-term dynamics may be difficult to predict .