Tropical Legumes I, led by the Generation Challenge Program (GCP) based at CIMMYT had two phases 2007-2010 and 2010-14). Partners included ICRISAT (chickpea and groundnut), CIAT (common bean) and University of California-Riverside (cowpea). Phase I developed critical genomic resources needed for modern breeding of BMGF focus legumes, while Phase II focused on applying these resources in pilots of several modern breeding schemes in each of the legume crops. Pilot marker-assisted selection (MAS) in a pedigree breeding scheme, marker-assisted backcrossing (MABC) and marker-assisted recurrent selection (MARS) were undertaken to identify key challenges of broader implementation at both CGIAR and NARS institutions.
Capacity building and institutional strengthening were a central theme of the project with considerable investment in training programs and physical infrastructure including phenotyping facilities; a total of 35 PhD and MSc students from 16 countries have been supported, either directly or indirectly, in their thesis work. The TL I made a significant contribution to developing high throughput genotyping and phenotyping platforms, identifying genomic regions (markers) associated with resistance to key biotic and abiotic constraints, including: Resistance to major foliar diseases, such as early and late leaf spots and rosette virus in groundnut; Fusarium wilt, bacterial blight, aphid and Striga resistance in cowpea; General improvements in pod harvest index and resistance to mosaic virus resistance in common bean; Genotyping platforms developed by the TL I were employed by TL II breeders, as were few molecular markers. In chickpea, for example, a marker for drought tolerance discovered in the TL I was transferred via Marker-Assisted Backcrossing to several locally adapted genetic backgrounds in Kenya and Ethiopia and these derived lines showing substantial promise in multi-location tests.
One of the major opportunities for the TL III is to more holistically integrate the outputs of TL I into its mainstream breeding pipelines. Despite these advances, several key challenges remain. Few CGIAR or NARS legume breeders can effectively utilize the genomic resources developed by the project without significant mentoring to interpret genotyping data and make selection decisions. Thus in order to fully leverage these genomic resources and technology toward applied goals, substantial training in this area will be needed. For groundnut development of breeder friendly genotyping platforms has proceeded slowly due to the difficult genetics of this crop (allotetraploidy). Fortunately, a new public/private groundnut genomics initiative Peanut Bioscience Initiative, in partnership with TL 1/TL III scientists, promise to make rapid gains in this crop that can be leveraged by the TL III.