Skip to main content
ARS Home » Southeast Area » Tifton, Georgia » Crop Genetics and Breeding Research » Research » Publications at this Location » Publication #164496


item Wilson, Jeffrey - Jeff

Submitted to: International Sorghum and Millets Newsletter
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/20/2006
Publication Date: 12/20/2006
Citation: Wilson, J.P., Tapsoba, H., Rajarajan, J. 2006. Mixture design analysis of rust infection in three-component pearl millet mixtures. International Sorghum and Millets Newsletter 47:143-145..

Interpretive Summary: Cultivar mixtures are a useful approach for slowing epidemics of some diseases. Mechanisms resulting in the lowered disease in two-component mixtures hve been most widely studied, however, three-component mixtures present particular difficulties in assessing host-pathogen interactions that result in their effectiveness. Some theories of the mechanisms resulting in lower disease in mixtures include indirect host-pathogen interactions which are difficult to quantify. If these effects do exist, it should be possible to identify three-component mixtures that would result in an optimum control of disease. These studies used mixture design analysis to evaluate rust development in three-component mixtures of pearl millet to determine whether optimum mixtures existed that resulted in a minimum of disease. Results of these studies suggest that direct effects of the resistance genes are the primary mechanisms reducing disease severity. Optimum three-component mixtures that resulted in a minimum level of disease development were not detected, indicating indirect effects did not contribute to the lower disease development in these experiments.

Technical Abstract: Mixture design analysis was used to evaluate disease development in mixtures composed of three components with different levels of resistance. Treatments consisting of three pearl millet (Pennisetum glaucum) inbreds (Tift 85DB, PS748BC, and Tift 89D2) with differential rust resistance were mixed in proportions defined according to a simplex centroid design model, and evaluated at Tifton, GA from 1995 to 1997. Final rust severity and area under the disease progress curve (AUDPC) data were fitted to either quadratic or special cubic regression models within and across years. Coefficients of determination (R2s) were > 95% for the models fitted by data within year, and > 0.86 for models fitted by data over years. Synergistic and antagonistic effects of mixtures were observed within years, but specific effects were inconsistent between years. When averaged over all years, disease control was primarily the result of the effectiveness of the resistance of the components and their proportions in the mixture. Disease responses were maximized at 100% Tift 85DB (the most susceptible inbred) and minimum at 100% Tift 89D2 (the most resistant inbred). The proportions of Tift 85DB and Tift 89D2 were the dominant influence on disease development, whereas PS748BC (the inbred with intermediate resistance) had the least effect on disease development when the proportions of the other two inbreds were equal to each other. We did not identify a three-component mixture in which disease was minimized, thus, consistent indirect effects of mixing due to host-pathogen interactions (induced resistance in the host, genetic disequillibrium among virulence alleles or virulence and fitness alleles in the pathogen, or cost of virulence) were not observed.