Epsilon Open Archive

Abstract

The annual results of Swedish crop variety trials are presented in reports and on the internet for Sweden divided into seven regions (production areas) A-G covering southern Sweden. The yield results for test varieties are usually presented as ratios relative to the yield of a control variety. These ratios are presented per region, with the implicit assumption that differences in ratios may exist between regions. In this report, the division of agricultural districts into regions was investigated through cluster analyses. Districts that produced similar levels of yield or similar ratios were clustered into groups of similar districts. Cluster analyses were performed on regions, districts and soil types for spring barley, winter wheat and oats, based on a large data set of results from variety trials performed during the period 1997-2006. The study revealed that some regions, districts and soil types produce similar levels of yield or similar yield ratios. However, clusters of regions, districts or soil types that produce similar levels of yield do not always produce similar yield ratios.