Loss in monthly pasture production resulting from dryland conditions in the Waikato

Open Access
Publisher Website
Google Scholar
Abstract
This paper examines a simple model requiring monthly estimates of the maximum weather-dependent evapotranspiration and daily rainfall to predict the reduction in monthly dry land pasture growth relative to irrigated production.The basic assumption is that dry matter production is proportional to transpiration.The model is tested using results from two irrigation experiments on ryegrass-white clover pastures rotationally grazed by beef cattle.These trials took place in the central Waikato Valley on two soil types;an Aquic Hapludult (HAM) from 1953 to 1964 and a Typic Vitrandept (HOR) from 1962 to 1966.For this period dryland conditions reduced yields over the November through April period by 39 and 27% respectively on the two soils.Predicted monthly dry land yields using irrigated yields and the estimated water use showed no systematic bias compared with actual pasture production and accounted for 83% and 77% of the yield variation for the HAM and HOR soils respectively.Errors caused by the lack of fit of the model combine with those of measurement and the spatial variability of pastures so that the standard deviation of a predicted monthly yield is about 250 kg DM ha– 1 Errors attributable to the model are of a similar magnitude to those of conventional pasture yield measurements.The model is not sensitive to the details of the transpiration-soil water storage relationship assumed.Limitations and applications are discussed.