Abstract
Visual assessments of pasture dry matter (DM) are used by dairy farmers to assist feeding decisions and planning of feed supplies. Methods of visually estimating pasture DM were researched in experiments on typical dairying pasture in Northland during 1981–83. Regression analyses of visual estimates on cut DM showed a simple linear relationship. If their estimates were calibrated, inexperienced operators provided DM assessments of standardised pastures similar to those of experienced operators. Estimates obtained by physical contact with the pastures differed from those by visual assessment only. Imaginary quadrats could be used as effectively as fixed quadrats. Cuts taken from ‘field’ estimates differed less than 7% from cuts taken from the more carefully assessed ‘standard’ calibration. Comparing fixed quadrat positions in variable pastures in the morning and afternoon showed no significant difference between estimates, but the relationship probably changes over longer periods. Accuracy was improved by cutting more quadrats at any calibration, and by cutting quadrats near the extremes of the DM range rather than throughout the range. When estimating DM on a paddock scale, a single estimate was significantly less precise than the average value from estimating imaginary quadrats along a representative transect. On a farm scale, individual operators provided estimates of average DM on the farm within 200 kg DM/ha.