Accurate predictions of barley phenotypes using genomewide markers and environmental covariates

Authors: Neyhart, Jeffrey; SILVERSTEIN, KEVIN A - University Of Minnesota; SMITH, KEVIN - University Of Minnesota

Submitted to: Crop Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/2/2022
Publication Date: N/A
Citation: N/A

Interpretive Summary: Climate change is expected to cause more adverse conditions for crop production. Breeding new crop varieties that are more tolerant of these stresses will improve overall agricultural resiliency; however, determining what varieties are ideal under diverse growing conditions is an expensive process in both time and resources. Recent advances in genomics technology and climate data science may, in theory, allow breeders to predict the yield or quality of hundreds of plant varieties across different environments, but this has yet to be tested in the field. We measured yield and quality for 183 malting barley varieties in more than 41 different environments representing a range of climatic and soil conditions. These measurements were combined with genomic information of these varieties and local weather and soil data to build a prediction framework. We showed that this approach can accurately predict the yield and quality outcomes of an entirely different set of barley varieties grown in new environments. Our predictive method will allow breeders to more quickly identify crop varieties that excel under varying growing conditions, helping to ensure grower profitability and consumer food security under a changing climate.

Technical Abstract: Breeding high-yielding and high-quality crops for alternative growing conditions is critical for agricultural climate change adaptation, progress towards which could be accelerated by predicting new crop variety performance from genomic markers and climate data. Using phenotypic observations of a barley (Hordeum vulgare L) population in forty-one field environments, we developed and tested a generalizable approach for identifying abiotic factors predictive of phenotypic expression. Validated models enabled highly accurate predictions of the performance of untested breeding germplasm for adaptational, yield, and quality traits in new growing sites or environments. Our results support the use of a relatively simple framework for predicting and exploiting phenotypic variation in crops to ensure food security under changing climates.