In pursuit of a better world: crop improvement and the CGIAR

Abstract

The CGIAR crop improvement (CI) programs, unlike commercial CI programs, which are mainly geared to profit though meeting farmers’ needs, are charged with meeting multiple objectives with target populations that include both farmers and the community at large. We compiled the opinions from more than thirty experts in the private and public sector on key strategies, methodologies and activities that could the help CGIAR meet the challenges of providing farmers with improved varieties while simultaneously meeting the goals of: (i) nutrition, health, and food security; (ii) poverty reduction, livelihoods, and jobs; (iii) gender equality, youth and inclusion; (iv) climate adaptation and mitigation and (v) environmental health and biodiversity. We review the crop improvement processes starting with crop choice, moving through to breeding objectives, production of potential new varieties, selection and finally adoption by farmers. The importance of multi-disciplinary teams working towards common objectives is stressed as a key factor to success. The role of the distinct disciplines, actors and their interactions throughout the process from crop choice through to adoption by farmers is discussed and illustrated.

Return to Agrobiodiversity: Participatory Plant Breeding

“…The adoption of varieties by farmers in marginal areas has been, and continues to be, a problem [46,47]. The CGIAR is no exception, with a large number of varieties released but never adopted by farmers [48], arguably as a consequence, at least partly, of GEI.…”

Investigations into the emergent properties of gene-to-phenotype networks across cycles of selection: A case study of shoot branching in plants

“…Due to an increasingly harsh and unpredictable climate, improving the consistency and scope of predictions for crop performance is crucial for global agriculture to meet the challenge of feeding a global population of 10+ billion people (Reynolds et al, 2021). Current prediction methods used in crop breeding assume a simplified linear relationship between genotype and phenotypes (Falconer and Mackay, 1996; Meuwissen et al, 2001; Cooper et al, 2014; Walsh and Lynch, 2018; Gianola, 2021), thus limiting the realized selection response achieved by many crop improvement programs (Kholová et al, 2021). Although this simplified genotype-to-phenotype relationship (G2P map) is sufficient to successfully model the average selection trajectory of large populations (Cooper et al, 2014), this approach captures only a subset of all performance outcomes, potentially leading to misalignments between predicted performance and realized performance in the field.…”

“…Although this simplified genotype-to-phenotype relationship (G2P map) is sufficient to successfully model the average selection trajectory of large populations (Cooper et al, 2014), this approach captures only a subset of all performance outcomes, potentially leading to misalignments between predicted performance and realized performance in the field. Such simplified G2P relationships can hinder accurate predictions for crop performance in specific management and environment combinations (Kholová et al, 2021).…”

Machine Learning-Based Plant Detection Algorithms to Automate Counting Tasks Using 3D Canopy Scans

“…There is an increasing demand from plant-related research disciplines to access the specific plant parameters in large numbers of plants with enhanced throughput. This has become possible with the development of novel technologies, including sophisticated sensors and state-of-the-art computer vision algorithms (i.e., plant phenomics) [ 1 , 2 , 3 , 4 , 5 ]. These technologies provide powerful scanning tools (e.g., [ 6 , 7 , 8 ]) but usually lack the development of algorithms that can detect individual plants in complex canopies or calculate complex plant features (reviewed, e.g., [ 1 , 2 , 3 ]).…”

“…These technologies provide powerful scanning tools (e.g., [ 6 , 7 , 8 ]) but usually lack the development of algorithms that can detect individual plants in complex canopies or calculate complex plant features (reviewed, e.g., [ 1 , 2 , 3 ]). Therefore, more flexible approaches are required to meet the growing demand for detailed plant observations, notably, automated pipelines that extract complex plant features (e.g., to develop climate-ready crops, among others) [ 1 , 4 , 9 , 10 , 11 , 12 ].…”