First author: Department of Plant Pathology, University of Georgia, Athens 30602; and second author: USDA-ARS Forage & Turf Research Unit, University of Georgia Coastal Plain Experiment Station, Tifton 31793-0748
Pearl millet inbreds Tift 23DB, Tift 85DB, PS748BC, and Tift 89D2 were used to develop three categories of host mixtures (physical mixtures, random-mated populations, and mixtures of two-way and three-way crosses) representing different levels of complexity of resistance through increased heterogeneity within populations and through stacking of resistance genes within the heterogeneous populations. The potential of these mixtures to reduce rust epidemics was evaluated in the field. Area under the disease progress curves (AUDPCs) of all physical mixtures were less than the mean of the components in 1995 and were less than the mean of the components for five of the six mixtures in 1997. In 1996, AUDPCs of the physical mixtures were consistently greater than the mean of their components. AUDPCs of the random-mated mixtures and the mixtures of crosses were consistently less than the mean of the components in 1996 and 1997, with reductions ranging from 12 to 71%. Dry matter yield (DMY) of physical mixtures relative to the mean DMY of the components was inconsistent, ranging from 18% less to 50% more than the mean of the components. The random-mated populations and the mixtures of crosses yielded 18 to 40% more DMY than the mean yield of the pure stands of their components.