Advanced breeding methodologies for wheat improvement in Bangladesh

Abstract

Crop improvement is a central objective to address global food security of the increasing population. Breeders and geneticists around the world are trying to find out the best ways and means that can select the superior lines of any crops. A 2% genetic gain is needed to keep up with the increasing global population and increasing food demand. To accelerate rapid genetic gain conventional breeding methods of crop selection should be complemented with the advanced molecular selection methods that encompasses with the genotyping technology. Rapid advances in technology like next generation sequencing that resulted in many sequenced genomes and the ability to quickly genotype thousands of individuals are providing the datasets to match genotyping to phenotyping. Here, we will describe the advanced breeding methodologies that can be used to improve any crop with specific focus on wheat improvement for the heat stressed environments of Bangladesh. Advanced breeding methodologies includes – predicting yield with the secondary traits, genome wide association studies to identify the significant genomic region for a specific trait and using whole-genome prediction models to calculate the genomic estimated breeding values to make genomic selection. There are two ways of predicting important traits of any crops – phenotypic prediction and genotypic prediction. Yield prediction is the final target of any breeding program but selection for yield is limited by the extent of field trials, fluctuating environments, and the time needed to obtain multi-year assessments. Proximal sensing data collection is increasingly implemented with high-throughput platforms that provide powerful and affordable information, while efficiently using this data is challenging. The objective of this study was to monitor wheat growth and predict grain yield in wheat breeding trials using high-density proximal sensing measurements under extreme terminal heat stress that is common in Bangladesh. We used several models and different secondary traits for this purpose. Our results showed that optimized phenotypic prediction models can leverage secondary traits to deliver accurate predictions of wheat grain yield, allowing breeding programs to make more robust and rapid selections. A genome wide association study (GWAS) was conducted for grain yield, yield components and other secondary traits in elite spring wheat germplasm grown in natural heat stressed environment in Bangladesh to identify genomic regions that control component traits and contribute to yield potential. A total of 2682 unique advanced wheat lines from the CIMMYT bread wheat program were planted in cohorts of ~540 lines in each of the five wheat growing seasons with measurement of important traits including grain yield and yield component traits and proximal sensing data including normalized difference vegetation index (NDVI) and canopy temperature (CT). To understand the genetic architecture of these traits, genome-wide association study (GWAS) was conducted using 39,912 SNPs from genotyping-by-sequencing. GWAS result were insignificant and variable for CT and NDVI supporting a hypothesis of highly polygenic genetic architecture. In contrast, large effect loci associated with days to heading and days to maturity were found on chromosomes 5A, and 5B at the Vrn-A1 and Vrn-B1 loci and the frequency and impact of these alleles was observed to vary over successive cohorts. We were able to find significant association in chromosome 3B and 4A for grain yield that colocalized with loci identified for thousand grain weight. Overall, this study highlights the utility of secondary traits including sensor based NDVI and CT to identify chromosome regions that contribute to yield and stress tolerance in South Asian spring bread wheat and better understand the genetic architecture, particularly for heading date and maturity which are critical targets of selection to avoid extreme terminal heat stress. By matching the dense genotyping data with the phenotyping data, we can successfully predict and select the best performed cultivar. Predicting crop performance and selecting them using genetic information is a major challenge for 21st century plant breeders. This is because a complex trait is controlled by thousands of genes and their interactions with the environment where the crops are grown. We have developed a genomic selection model for the heat stressed environment in South Asia. With the advanced wheat lines collected from CIMMYT, Mexico, a training population was created, and genomic selection was done for the breeding population. We found low to high prediction accuracy across the years and how to moderate prediction accuracy across trials. Days to heading and maturity showed the highest and consistent prediction accuracy while thousand grain weight and grains per spike had good predictability among the yield components. This genomic selection approach can be used in any unbalanced dataset that are common to any breeding program. It will ultimately accelerate the rate of crop improvement that is important to secure the global food security.