Comparison of Space and Sward Planted Switchgrass for Traits Associated With Biomass and Ethanol Yield.

Switchgrass(Panicum virgatum L.) is an important potential biofuel crop. Switchgrass breeding nurseries are typically space-planted; however, production is in dense swards. This disconnect may impact selection. The objectives of this research are to compare space-planted to sward-planted switchgrass for i.) correlations between morphological and yield traits, ii.) estimates of general and specific combining ability(GCA, SCA), heritability, and heterosis of yield and iii.) yield rankings among and within families. Two Kanlow selections, five Alamo selections, and Miami were crossed in a complete diallel. Progeny were divided into two clonal propagules and planted in adjacent nurseries in Knoxville, TN. Nurseries were planted in a randomized complete block design containing twenty replications of each cross and eight clonal replications of each parent. The space-planted nursery(HSP) contained single-plant plots on 1m centers. The sward-planted nursery(HSW) contained plots of 4x7 plants on .33m centers with 1m alleys. Fall evaluations included rust, height, tiller number, and yield. Spring evaluations included height, tiller number, tiller diameter, leaf diameter, color, and leaf angle. Data were analyzed using SAS. Within nursery correlations of yield with morphological traits were similar in both nurseries for all traits except rust and height. Between nurseries, correlations were moderate for all traits(r=0.31-0.53) except height and color(r=0.21-0.25). In both nurseries, GCA was not significant, SCA was significant, and one family showed significantly higher yield; however, this family differed by nursery. Narrow-sense heritability estimates for yield also differed(HSP:h²=0.08, HSW:h²=0.02). Mid-parent heterosis was observed in both nurseries but families exhibiting mid-parent heterosis differed between nurseries. High parent heterosis was observed only in HSP. An ANOVA showed significant interaction for family*location(p=<0.0001). Among family rankings, 53% of the top 30% and 65% of the bottom 30% disagreed between nurseries. Results indicate selection in an evaluation environment different from the production environment could impact selection and rate of crop improvement.