AI, imaging from the breeder's eye to satellites, and molecular markers, used to shorten the path from the first cross to a variety that farmers grow and fresh markets accept, in sweet corn and other crops of northern Vietnam.
Breeding programs in Vietnam weigh many traits at once, on small budgets and without dense genotyping, and now under a climate that shifts the seasons they were selected in. The work brings measurement, selection methods and prediction into those programs, and keeps testing whether a model still holds where it was not trained. The traits it selects on run from nitrogen use efficiency and tolerance of waterlogging to the eating quality a fresh market pays for, which is also what the Hanoi city grant on a sweet corn brand asks for. The next sweet corn hybrids from the VNUA program are being selected this way, on data that runs from what a breeder sees in the field to satellite images.
Sweet corn breeding
Sweet and waxy-sweet hybrids VNUA168, FC369 and MH81 are recognized for circulation. Parents and hybrids are chosen with multivariate analysis of field traits, SSR markers and combining-ability studies.
Read moreCrop breeding
Specialty corn hybrids for northern Vietnam, with parents and hybrids chosen by multi-trait indices and combining-ability analysis; the same selection tools applied to sweetpotato and taro germplasm.
Read moreMathematical modeling
Statistical and machine-learning models that predict crop traits, yield and production, and frameworks that tie those predictions to breeding decisions.
Read morePhenomics and genomics
Image-based, high-throughput phenotyping of seeds and plants, SSR markers for germplasm and sweetness genes, and genomic prediction for crops without dense genotyping.
Read moreRemote sensing and computer vision
Satellite and drone imagery with interpretable machine learning, to map and forecast production of a perennial crop across a province.
Read moreCrops and places
Most of the breeding work is on sweet corn and other specialty corn (supersweet, waxy-sweet and purple waxy types) at the Institute of Biology and Agricultural Technology, Vietnam National University of Agriculture, in Hanoi. Phenomics work on rice, wheat and soybean was done with colleagues at ICAR-Indian Agricultural Research Institute, New Delhi, where he completed his M.S. Remote sensing of Robusta coffee in Dak Lak is his Ph.D. research at Kyushu University. Sweetpotato and taro work is with colleagues at VNUA, and the quinoa study had co-authors in Vietnam, Japan and Argentina.
Working together
Collaboration is welcome on specialty maize germplasm and testing, multi-trait selection methods, image and satellite phenotyping, genomic prediction for crops with little genotyping data, and on-farm testing or seed multiplication with growers and seed producers. Students at VNUA looking for a thesis topic in these areas are welcome to write. Email ntduc@vnua.edu.vn.