ON VARIATION IN CROP YIELD BETWEEN AND WITHIN GRASSLAND FIELDS

A growth model for prediction of yield in grassland was validated. Based on weather data from Satenas (58-degrees-N) for the period 1975-1983, model sensitivity for two soil types on dry matter production was simulated. The difference in simulated yield between the soil types was 14%. The variation of forage yield within fields was estimated on a farm in northern Uppland (61-degrees-N). Prediction of forage yield was carried out using different numbers of field samples for calibration of the growth model. The effect of the size and number of samples, and of the time interval between calibration and harvest on the precision of the prediction is discussed.