Wheat yield response to rainfall in a long-term multi-rotation experiment in the Victorian Wimmera

Abstract

Seventy-six years of wheat yield data from a long-term rotation experiment at Dooen in the Victorian Wimmera were analysed to describe the response of wheat yield to seasonal rainfall, crop sequence, and time. Wheat yields from 7 different 1- to 4-course rotations involving wheat, barley, oat, field pea, grass pasture and fallow were compared as a function of growing-season (May-November) rainfall. The field layout had no within-year replication, but each phase of each rotation was represented once in each year. An approximate quadratic response of wheat yield to both current year and previous year May-November rainfall was observed for each rotation. Previous year May-November rainfall boosted wheat yields grown on fallow, but decreased the yield of wheat grown on field pea or wheat stubble. Highest wheat yields followed fallow preceded by pasture, high yields followed fallow preceded by a cereal, moderate yields followed field pea, and low yields occurred for continuous wheat. Long-term trends in wheat yields adjusted for rainfall depended on crop sequence and fluctuated more in the non-fallow, 3-course rotations. Over the 76 years, average yield declined in all rotations except the continuous wheat, which was always low, but there was evidence that yield of all continuous cropped rotations had increased during the last 2 decades.