Principal investigator: Vikram Chhatre, research geneticist, USDA Forest Service (vikram.chhatre@usda.gov)
Co-authors: Keith Woeste, former PI, national program lead for genetics, research and development deputy area, USDA Forest Service; Richard Cronn, research geneticist, Pacific Northwest Research Station, Corvallis, Oregon; Denita Hadziabdic, associate professor, entomology and plant pathology, University of Tennessee; Mo Zhou, associate professor, Forestry and Natural Resources, Purdue University; James McKenna, operational tree breeder, USDA Forest Service (retired); James Warren, biological scientist, USDA Forest Service.
There is a long history of tree breeding and improvement in the United States. Pioneering tree improvement programs focused primarily on southern pines, which are the bread and butter of the construction timber industry. These programs implemented classical tree breeding and selection for economically important traits, a long and arduous process often requiring decades of work before any gains are realized, primarily due to long generation times of many forest tree species. Genomics-based methods have the potential to accelerate this process by incorporating genetic marker data from the family structure when estimating breeding values using statistical methods such as best linear unbiased prediction, or BLUP. The resulting method, G-BLUP, may help accelerate identifying superior parent trees in a breeding program.
The initial goal of this project was to apply this method termed “breeding without breeding” to the numerous black walnut provenance trials established by the HTIRC using phenotypic (growth and form) and genotypic (genetic variation) data to predict superior parents. Researchers anticipate, at least in theory, that well-performing trees can be identified and potentially used by stakeholders in commercial plantations.
Using prior knowledge of the familial relationships within HTIRC provenance trials, researchers collected leaf or bud tissues from 83 mother trees and more than 1,000 progeny. They recorded growth (height) and form (straightness and low branching patterns) from the same individuals.
The tissue is used for isolating genomic DNA, which is subjected to targeted genotyping on a MassArray panel to identify genetic variants in the coding regions of the genome. Resulting data is then used to verify the known mother-progeny relationship, potentially identify the pollen parent and to feed into statistical models to estimate G-BLUPs, which provide a measure of genetic superiority of the parents (breeding value).
Researchers are currently gathering data to estimate the G-BLUP breeding values. However, based on the genetic data available, they were able to verify parent x progeny relationship for at least some of the samples.
“Being able to corroborate field data using genomic data is an important step forward,” principal investigator Vikram Chhatre said. “While we were able to verify some of the parent x progeny relationships, it remains a work in progress. One unexpected result was the low degree of statistical confidence we saw in parental assignment analysis. We are currently reassessing this analysis and hope to provide more details in the summer of 2025.”
Future research inquiries would investigate whether this method is suitable for species that do not have a well-established breeding program. Focusing on an array of measurable traits also would broaden the genetic variability to be studied using this statistical model.
Goals:
The overarching goal of tree breeders is to develop planting stock with highly desirable traits (disease resistance, environmental adaptation, superior quality timber, to name a few) as rapidly as possible. If increasing the efficiency and speed of the selection and breeding process is the goal, then “breeding without breeding” is one of the potential tools available. This project aims to serve as a “proof of concept” that this technique could work in eastern hardwoods and will therefore be impactful for stakeholders. Depending upon the results of the project, it may be possible to develop the G-BLUP metric further with implications for the commercial sector.
This project is aimed at using statistical models that have been tested in a small number of cases. The “breeding without breeding” method (conceptualized by El-Kassaby et al., 2009) has so far been tested only to a limited extent (e.g., Douglas fir, Norway spruce and Scotch pine). From a theoretical standpoint, the method presents a conceptual advancement in the field.
A novel contribution of this project would be to demonstrate that the method could work in eastern hardwood species. It should be noted that among genomics-based methods in the area of forest tree breeding, genomic selection (GS) and genome wide association studies (GWAS) have been successful and broadly implemented. Those technologies, however, require considerably more financial investment for deep sequencing of individuals. “Breeding without breeding” does not require deep sequencing since it uses pre-developed MassArray genotyping method.
Methods:
Researchers collected leaf or bud tissues from 83 mother trees and over 1,000 progeny. They recorded growth (height) and form (straightness and low branching patterns) from the same individuals.
The tissue is used for isolating genomic DNA, which is subjected to targeted genotyping on a MassArray panel to identify genetic variants in the coding regions of the genome. Resulting data is then used to verify the known mother-progeny relationship, potentially identify the pollen parent and to feed into statistical models to estimate G-BLUPs, which provide a measure of genetic superiority of the parents (breeding value).
2024 Findings:
Researchers do not yet have all the data necessary to estimate the G-BLUP breeding values. However, based on the genetic data they had available, they were able to verify parent/progeny relationships for at least some of the samples. Being able to corroborate field data using genomic data is an important step forward.
One unexpected result was the low degree of statistical confidence that researchers saw in parental assignment analysis. They are currently reassessing this analysis.
Key Partners/Collaborators:
- Richard Cronn of the USDA Forest Service from the Pacific Northwest Research Station was instrumental in developing the MassArray genotyping panel that is being used in this study.
- James McKenna painstakingly planted many of the HTIRC trials that are part of this study.
- Other collaborators include: Denita Hadziabdic and Sarah Boggess of the University of Tennessee at Knoxville, Mo Zhou of Purdue University and James Warren of the USDA Forest Service.