The effect of low plant density on response to selection for biomass production in switchgrass

Switchgrass (Panicum virgatum L.) is a model bioenergy species with a high biomass production from which renewable sources of fuel and electricity can be generated. The objective of this study was to perform divergent single-plant selection for biomass yield at low plant density, intermate the selected plants in polycrosses, and evaluate the performance of their half-sib (HS) progenies in sward and row plots. One thousand plants from Alamo and Kanlow populations were planted in unreplicated honeycomb trials with a plant spacing of 120 cm. Moving-ring selection was applied to select 15 high- and 15 low-yielding plants from each population. These were grown in separate polycrosses to create HS-families. Four HS-families from high- and 4 from low-yielding parents of Alamo and Kanlow, along with their bulks were evaluated for 3 years in sward plots with row spacing of 18 cm. Five HS-families from high- and five from low-yielding parents of Alamo and Kanlow were evaluated for 2 years in row plots spaced 76 cm along with their bulks. Overall, the row-plots had 20% higher biomass per unit area than the sward-plots. Across populations and plant densities, the highest-yielding HS-families produced between 2.0 and 9.3 t/ha more biomass than the lowest-yielding HS-families. The mean biomass of the HS-families from parents selected for high-yield was between 0.34 and 4 t/ha higher than the mean of the HS-families from the low-yielding parents. The annual response to selection for the mean and the bulk of the HS-families from the high yielding parents was 19 and 27% higher than the check. These results indicate that on average, high- and low-yielding parental genotypes were effectively selected from the two populations under low plant density.