gms | German Medical Science

Crop yield has increased during the last decades due to genetic gain and improved agronomic practices. In Germany, this trend can be seen in series of official value-of-cultivation-and-use (VCU) trials and on-farm. The genetic gain results from cultivating newly registered and enhanced cultivars selected by breeders and within VCU trials. Non-genetic changes can occur due to climate change, improved agronomic practices, or changes in government regulations. To distinguish between genetic and non-genetic gains and to estimate trends, two different approaches can be used. First, historical and recent cultivars can be tested together in a small set of locations and years. The alternative is to use historical data of official VCU trials. These data sets are large, but require a mixed model framework to distinguish between genetic and non-genetic trends. Mixed models with large data sets are computationally demanding. Thus, these data sets are most often reduced to cultivars tested for at least three years, dropping data on all cultivars that were culled after one or two years of testing. The drop-out is informative, but it is not clear whether this drop-out will affect predictions of genetic and non-genetic trends.

A simulation study was done mimicking the structure of German VCU trails in winter wheat (including the yield-dependent selection process). No trends were simulated.

Results showed unbiased trend estimates when using all available data. If data are restricted to cultivars tested for at least three years, a significant positive genetic trend of 0.11 dt ha^- and a negative non-genetic trend of the same size was observed.

The authors declare that an ethics committee vote is not required.