Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/7/2006
Publication Date: N/A
Citation: Interpretive Summary: Accurately measuring pasture yield is key to budgeting forage in grazing systems; however, producers often do not accurately measure pasture yields because they are not convinced that it pays. ARS scientists at University Park, Pennsylvania, in collaboration with farmers and extension specialists in Maryland, West Virginia, and Pennsylvania, developed generalized calibrations for a pasture plate meter broadly applicable across the northeast. General plate meter calibrations appear to have practical application across pastures when pastures are divided into proper botanical or sward structure groupings. Based on visual observations, these groupings appear to be due primarily to tiller size and density.
Technical Abstract: The ability to measure pasture herbage mass (HM) rapidly and accurately is needed in research and for on-farm pasture budgeting. The accuracy of plate meter estimates of HM is dependent on meter calibration. Robust general calibration equations would reduce the need to calibrate plate meters for each pasture and season combination. The purpose was to develop general calibrations for HM using a falling plate meter and determine the accuracy of the general calibrations. Rotationally stocked pastures were sampled using a falling plate meter, a rising plate meter, and a meter stick along transects with 30 or more heights taken at random with each unit. Fifteen paired falling plate meter height and herbage mass samples were taken per paddock for estimating mean herbage density (HD) and HM. Calibration equations of plate height to HD and HM were calculated using linear regression. Based on analysis of variance, pastures were divided into three ebotanical groups that differed in HD and HM. These groups differed in tiller density and size. Using the general calibrations to predict HD and HM for pastures in the test data set gave values within the confidence interval of clipped samples 54% and 68% of the time for 15 clipped samples and 63% and 74% of the time for 6 clipped samples. Based on validation regressions, the general calibrations were unbiased estimators for HD and HM having no constant or proportional bias, with CVs of 25 and 23 percent respectively. Pasture heights measured with plate meters or ruler were highly correlated. Cross calibration values were developed for comparing measurements made by the different techniques. General plate meter calibrations appear to have practical application across pastures when pastures are divided into proper botanical grouping that appear to be based