Evaluating the potential of winter beet in northern Germany by a simulation model

Abstract Autumn-sown sugar beets may achieve a faster canopy closure in spring compared to spring-sown sugar beets, thus increasing radiation interception and yield. However, to date yield benefits of autumn-sown beets cannot be quantified, as they start to bolt in spring. Therefore, simulation models for winter beet need to be developed. From 2009 to 2012, field trials in northern Germany were carried out with sugar beets sown in spring (conventional sugar beet) as well as sown in autumn (bolting winter beet). Based on these data, a dynamic simulation model was developed and calibrated for winter beets. The model is based on light use efficiency (LUE) approach and considered winter beet-specific aspects like leaf (area) losses due to frost, vernalisation and bolting. For conventional sugar beets, the validation revealed a RMSE of 164 g m -2 and a model efficiency (EF) of 0.93 for root dry matter. For the total dry matter (DM tot ) of bolting winter beets, the validation showed a RMSE of 216 g m -2 and an EF of 0.69. Implemented assumptions regarding the non-bolting winter beets allow the model to simulate this beet type as well. A simulation over 12 years (1998-2010) for the site Hohenschulen estimated 1921 g DM tot  m -2 , 971 g DM tot  m -2 and 2474 g DM tot  m -2 for conventional sugar beets, bolting winter beets and non-bolting winter beets, respectively. Our results suggest that developing non-bolting winter beets seems be a promising approach to increase beet yields.