Selecting for multiple traits in complex production systems: A case study of sugarcane in China

Abstract

Genetic improvement of crops has been critical in reducing costs of production and improving quality and reliability of supply of food worldwide. Genetic improvement depends on testing, ranking, and selecting large numbers of candidate genotypes for predicted commercial value, based on data obtained in selection trials. However, this is often difficult because genotypes can vary for many different traits affecting costs across the value chain, including growing, harvesting, transport, processing, and sales revenue.Selection index methodology can be applied to help resolve this problem. Here we illustrate this methodology through a case study of sugarcane improvement in China. Sugarcane is a major crop, harvested and processed globally in a greater mass than any other crop, and China is the third largest producer in the world. Sugarcane industries in China and many other countries are also changing from labour intensive to mechanised systems of growing and harvesting. Accordingly, two contrasting production systems were considered: (i) current systems, including small farms and harvesting using manual labour (S1); and (ii) likely future production systems with larger farms with mechanised farming and harvesting operations and modern transport and processing sectors (S2). The economic impact of varying a range of crop traits was determined for each system, and economic weightings of each trait defined accordingly. Using example data from a selection trial in southern China, ranking of a set of sugarcane families (crosses) based on selection indices for each system was shown. Important differences in ranking of sugarcane families were observed for indices for S1 and S2, indicating how changes in selection criteria are appropriate for developing cultivars for future production systems compared with currently applied selection criteria.While the cropping production system and economic parameters in the case study were based on data from China, the approach used is applicable generally. Suggestions were made regarding ongoing general application of this methodology in commercial breeding programs. Using the results and recommendations from the case study, advantages and disadvantages of application of selection indices versus other methods to crop improvement programs are discussed.