Phenotyping is a severe bottleneck in plant breeding, variety testing, crop protection and G x E x M (genotype x environment x management) research. Phenotyping means optical, precise and partly artificial-intelligence-supported determination of plant size, architecture and ‘performance’. Establishing, calibrating and validating new phenotyping methodologies requires large and diverse sets of training and validation data. It also requires statistical learning methods to improve target traits such as yield and quality. Relevant training data need to be sampled on diverse sets of genotypes under diverse environmental conditions. The project aims to utilize the experiments and measurements that are being done in the context of variety testing (Agroscope) and breeding research (ETH) to build up such training data of relevance for Swiss agriculture. The aim is to harmonize methods and to identify those that hold great promise to be implemented as routine procedures at core field sites of ETH and Agroscope. Within the European phenotyping infrastructure network EMPHASIS, the project seeks to develop the basis for a pan-European network of sites with similar phenotyping capabilities, able to evaluate the specific suitability of different crop ideotypes to different climatic scenarios. Core efforts will be on the improvement of wheat breeding workflows for the Swiss situation, but phenotyping methods to be further developed will also be beneficial for soybean and many other crops.
