MODEL EVALUATION FOR PERENNIAL GRASSES IN THE SOUTHERN UNITED-STATES

Models can be used to estimate yield response of grasses to applied N. This analysis was conducted to show variation of model parameters among grasses at the same location and differences among locations. The logistic equation was used to relate annual dry matter yield to applied N for four locations (Blairsville, GA; Fayetteville, AR; Thorsby, AL; Jay, FL) and five perennial grasses [bermudagrass (Cynodon dactylon L., Pers.), bahiagrass (Paspalum notatum Flugge), tall fescue (Festuca arundinacea Schreb.), dallisgrass (Paspalum dilatatum Poir.), and orchardgrass (Dactylis glomerata L.)]. The model contained three parameters (A, b, c). It was shown by analysis of variance that all grasses exhibited common exponetial coefficients (b, c) for each location, with variation among grasses and with water availability assigned to the linear coefficient (A). Applied N, N1/2, to reach one-half maximum yield appeared to be inversely related to clay content of the soil. Maximum efficiency of conversion of applied N to dry matter at N = N1/2 was highest for Coastal bermudagrass (approximately 40). The logistic equation provides an excellent model for yield response to applied N.