Submitted to: Phytopathology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/9/1998
Publication Date: N/A
Citation: Interpretive Summary: Plant diseases do not happen in a vacuum. Pathogens live in complex communities of other microorganisms that also infect their host plants or that live harmlessly on or in the plants, sometimes synergistically. Scientists usually try to avoid the confusion of multiple interactions by studying only one disease at a time. However, research on ecology of disease in crops is growing in importance, particularly in relation to biological control of diseases using non-pathogenic microorganisms to suppress pathogens on plant surfaces. This type of research is handicapped, however, because up to now scientists have not had a sound mathematical foundation for characterizing competitive and synergistic interactions among microbes. We modeled the processes of competition through infection and reproductive stages of the wheat stem rust fungus, an important plant pathogen. Then we used the model to show that the research hdesign most commonly used to analyze interactions between organisms is inadequate. We showed that this design, which is called the de Wit replacement series analysis, can produce results that appear to indicate competitive inhibition of one microbe by another when there is no real competitive inhibition between them. By using our model to identify the features of microbial life cycles that cause erroneous conclusions from the de Wit analysis, we help scientists to avoid the mistakes and to design superior research approaches that will provide a more accurate understanding of how communities of microorganisms on plants interact with each other. Understanding how to accurately measure and exploit interactions between pathogens and microbes that minimize disease will lead to better control of plant diseases with reduced need for pesticide use.
Technical Abstract: The de Wit replacement series is a widely used design to study competitive interactions in ecology. In the de Wit design, two organisms are varied in frequency in mixed populations at constant density. Relative yields of each organism and the mixture are calculated in relation to that of each organism separately at the same population density as the mixture. The null expectation is that relative yield of each organism will equal its proportion in the mixture and the sum of individual relative yields will equal total yield of the mixture at all frequencies. Deviations from the de Wit null expectations are interpreted as evidence of unequal intra- and inter-species competition. We generated de Wit curves from a competition model developed from observed competition between strains of wheat stem rust. Our competition model has parameters for infection efficiency, maximum number of sporulating infections supported per leaf, sporulation efficiency, and maximum amount of sporulation supported per leaf as well a inter-strain competition coefficients for infection and sporulation. Our results showed that deviations from the null expectations occurred not only when inter- and intra-strain competition was unequal, but also when the strains differed in infection efficiency or maximum numbers of infections supported by leaves. Without first measuring each of the parameters for each competing strain by itself, it is impossible to deduce from a de Wit analysis how much of the observed deviations from null expectations are from differences in inter- and intra-strain competition and how much is due to differences in other traits.