Developing morphological prediction equations for switchgrass cultivars in Indiana

Abstract Switchgrass ( Panicum virgatum L.) is an important warm‐season perennial grass in livestock systems and has been evaluated as an herbaceous energy crop. Switchgrass growth varies across environments. To accurately predict morphology, locally developed morphological development equations are needed. The objectives of this study were to compare the morphological development of a recently developed biofuel cultivar ‘Liberty’ to an improved forage cultivar ‘Shawnee’ in multiple environments in Indiana and to predict morphological development in response to growing degree days (GDD) and day of year (DOY). Pure stands of each cultivar were sampled weekly and biweekly at multiple locations in Indiana in 2016 and 2017. Morphological development was determined by the mean stage count (MSC) and mean stage weight (MSW) system. Prediction equations based on GDD and DOY for both MSC and MSW were developed from northern and central Indiana locations and validated using data from an independent site in west central Indiana. MSC and MSW were linearly related to GDD and had a quadratic relation to DOY. Significant main effects for GDD‐based prediction included location and cultivar ( P < .01), while DOY‐based predictions included location ( P < .01). Prediction r 2 on the independent validation dataset ranged from 96% to 98% for both MSC and MSW. When predicting MSW at the validation location, the GDD and DOY equations had r 2 values from 97% to 98%. Morphology predictions based on GDD and DOY resulted in similar r 2 values, suggesting that either measure could be used to accurately predict growth in Indiana.