Back to October 2022

Three Genetics Tools to Improve Pecan Breeding

A view of rows of young pecan trees from above. The orchard floor is mowed neatly around the trees, creating a thin strip of dark green grass within each row.

An aerial view of the block of ‘Lakota’ x ‘87MX3-2.11’ trees at the USDA ARS research station in College Station, Texas. This cross between a Northern cultivar and a Southern Mexican native from Oaxaca was developed as part of the recently concluded Specialty Crop Research Initiative (SCRI) grant, and populations were planted in three locations: Tifton, Georgia (2017), Byron, Georgia (2020), and College Station, Texas (2020). (Photo by Warren Chatwin)

I’m sure many of you have had the experience of asking a scientist a question, only to feel like you’ve been smacked in the head with a dictionary! Scientists like myself often have a hard time separating themselves from the descriptive language we use with other scientists. Unfortunately, this communication failure often has the side effect that people don’t always understand what we do or how our efforts improve pecan breeding. But this communication is important because plant breeding is like steering a barge down a river. Things move slowly, and course corrections are difficult. I hope that by knowing where our metaphorical barge is going, you will be able to understand why focusing on using genetics to improve pecan breeding over the next five to ten years will create the best improvements for pecan breeding.

Before I introduce the three genetics tools our program will utilize, I want to acknowledge our outstanding successes with traditional pecan improvement approaches based on observable traits. Efforts by breeders, such as Dr. Tommy Thompson, have produced pecan cultivars with larger nuts, earlier harvest, and increased tolerance for scab disease. For those of you who are curious for details on our program’s pecan breeding process, my predecessor, Dr. L.J. Grauke, gave an excellent overview in his 2019 Pecan South article Family Trees: The Next Generation. The last decade has been a time of significant change for our program, with the retirement of Drs. Thompson and Grauke and the addition of new research staff like myself and Dr. Xinwang Wang. Throughout this transition, our program has focused on developing the resources needed to incorporate genetics into pecan breeding, and we are currently in the final stages of that effort. Our future success is directly attributable to the vision of my predecessors and the resources they established. To quote Sir Isaac Newton, “If I have seen further it is by standing on the shoulders of giants.” 

The biggest limiting factor in pecan breeding is the time it takes for a new seedling to mature and produce nuts. The “Holy Grail” of pecan breeding would be the ability to predict desirable mature traits (like nut size, shape, quality, or disease susceptibility) on young seedlings and discard trees that do not have the potential to be future cultivars. The techniques to do so have existed (and have been thoroughly tested) for over a decade in annual crop systems, but until recently, pecan didn’t have the essential tools in our genetic toolbox to use them. This change is due, in large part, to a USDA-NIFA Specialty Crop Research Initiative (SCRI) grant, “Coordinated Development of Genetic Tools for Pecan,” led by Dr. Jennifer Randall at New Mexico State University, which concluded in 2022. With the four pecan genome sequences (‘Pawnee,’ ‘Lakota,’ ‘Elliott,’ and a Mexican native pecan) and other tools developed in this grant, our focus as researchers can shift to applying these to pecan improvement. The first step in this process is connecting the traits we observe in our orchards with the genetics controlling them. 

Three types of tools let us associate observable traits with genetics: Family Mapping, Population Mapping, and Genomic Selection. Each tool is at different implementation stages in our Pecan Breeding Program. However, a core component of all three is that they take advantage of the natural genetic shuffling of chromosomes that occurs in each parent as they form the reproductive cells that will become the pollen and female flowers. After pollinating the female flower, an embryo is formed that contains one shuffled set of chromosomes from each parent that will develop into a seed. This genetic shuffling is why some siblings (who share the same parents) are very similar while others are very different. 

Tool 1: Family Mapping (Linkage Mapping)

Family Mapping (or Linkage Mapping) is a tool that locates (or maps) chromosome break-points in a family of individuals when their DNA is compared to their parents’ DNA. These chromosome break-points are the natural product of the genetic shuffling process. These maps and observations of a trait of interest (like bud break or percent nut scab) can be used to locate the position of the genes controlling the observable traits. Scientists call these regions QTL (Quantitative Trait Loci), and locating one is like the first beeps from a metal detector. It lets you know something important is in that area (of the chromosome), but your precision depends on various factors. One major limitation of Family Mapping is you are limited to examining traits that are different between both parents of your population (such as round nuts vs. long nuts). You also get the best results when the traits are controlled by one or a few genes (which is uncommon). Our program has established two family mapping populations: ‘Elliott’ x ‘VC1-68’ (est. 2007-2008) and ‘Lakota’ x ’87MX3-2.11′ (est. 2016-2018). 

‘Elliott’ x ‘VC1-68’: This progeny population from two vigorous seedstock varieties was planted in 2010. Because of trait differences between ‘Elliott’ and ‘VC1-68,’ we expect this population to identify genetic-trait associations for leaf and nut scab susceptibility alongside various nut size and shape characteristics. Collaborative work with Dr. Patricia Klein at Texas A&M University on this population has resulted in identifying one major and two minor genetic-trait associations for budbreak and one minor genetic-trait association for leaf scab susceptibility (Bentley et al., 2020). The ‘Elliott’ x ‘VC1-68’ population is mature, and most trees are producing nuts. Research with Texas A&M is ongoing to identify genetic-trait associations for nut size, shape, and quality traits. Associating nut traits with genetics is critical for advancing pecan breeding.

‘Lakota’ x ‘87MX3-2.11’: This cross between a Northern cultivar and a Southern Mexican native from Oaxaca was developed as part of the recently concluded Specialty Crop Research Initiative (SCRI) grant, and populations were planted in three locations: Tifton, Georgia (2017), Byron, Georgia (2020), and College Station, Texas (2020). This strategy allows us to examine how similar genetics perform in different environments; none of these populations are mature enough to bear nuts. Because of differences between ‘Lakota’ and ‘87MX3-2.11,’ we expect these populations to identify genetic-trait associations for leaf and nut scab resistance, bud break timing, leaf retention, nut vivipary, Phylloxera insect susceptibility, and various nut quality traits. One genetic-trait association has already been identified for Phylloxera insect susceptibility, and collaborative research is ongoing with The University of Georgia and the USDA-ARS Southeastern Fruit and Tree Nut Research Laboratory to identify more associations (Lovell et al., 2021). 

Tool 2: Population Mapping (Genome-Wide Association Mapping)

Population Mapping also relies on using the historical natural genetic shuffling of each parent’s chromosomes that occurs when the reproductive cells (pollen and pistillate flowers) form to identify associations between genes and observable traits. The big difference, however, is instead of using a large family to identify chromosome break-points, we rely on the many generations of genetic shuffling that nature has already made for us! When picking trees for this Population Mapping tool, we want trees that are minimally related to each other and show diversity in our traits of interest. We then obtain DNA sequence and trait observations from this population and use math and statistics to help us identify which genetic-trait associations are more likely to be real. 

The multiple generations of historical genetic shuffling that naturally occurred in this population allow us to more accurately narrow down the chromosome positions of genes controlling traits (like smaller strokes of a metal detector). Another major advantage to Population Mapping is the ability to examine any trait we can observe in pecan (instead of just traits that are different between the parents of a Family Mapping population). However, it isn’t a genetic silver bullet for every situation because the unknown family relationships within the natural population can lead to false gene-trait associations. This tool is also best applied to traits controlled by single genes or regions.

NCGR-Carya: In addition to our Pecan Breeding Program, our USDA location maintains the National Collection of Genetic Resources for Pecans and Hickories (NCGR-Carya), a germplasm repository containing thousands of native pecans from across its entire geographic range alongside a diverse collection of pecan cultivars. This collection contains the ideal set of pecans to use for Population Mapping. Dr. Klein’s lab at Texas A&M has previously used our repository to obtain DNA sequence data for 100 diverse pecans, five hickories, and three hybrids. They used these sequences and field observations for flowering type (Type 1 or Type 2) to identify a genomic region controlling flowering type in pecan (Bentley et al., 2019). This region can now be sequenced to identify the flowering type in any pecan seedling. The collaborative SCRI research grant that established our most recent Family Mapping orchard also funded the DNA sequencing of a diverse set of pecans and hickories from our repository (713 pecans, 129 hickories, and 30 hybrids). A larger Population Mapping effort is being analyzed using this DNA sequence data and historical trait observations from our germplasm repository.

Tool 3: Genomic Selection

Family Mapping and Population Mapping are two approaches to the same end goal: identifying gene regions associated with traits that can eventually be applied in pecan breeding. Genomic Selection is a different approach focused on improving breeding and, in my opinion, is the most complicated of the three tools. Unlike Family Mapping or Population Mapping, which work best when traits are controlled by single genes or regions with big effects (uncommon), Genomic Selection works best on traits controlled by many genes with small effects (the more common scenario). It requires a specially designed population of crosses of multiple cultivars that represent the genetic diversity used in our breeding program. All these seedlings must have sequenced DNA and be monitored for important trait observations as they mature. All this information is used to train a mathematical model that creates a number (breeding value) that tells you how good each individual is for breeding a specific trait (like nut shape or scab susceptibility). A breeder can use these values to select the ideal parents to prioritize breeding pecans with targeted characteristics. This information can also be used to train a separate mathematical model to predict the traits of closely related individuals from only their DNA sequence data (Genomic Prediction).

This prediction ability is where we arrive at the “Holy Grail” I mentioned earlier in the article. This process is an especially powerful tool in long-lived species like pecan, where mature traits like nut production and quality cannot be evaluated for at least 5 to 10 years after planting. Identifying mature traits in immature seedlings could shorten the time it takes to release new pecan cultivars by up to a decade. We are currently in the planning stages to create the population needed to enable using Genomic Selection and Prediction in pecan breeding over the next decade. 

Closing Thoughts

I hope I didn’t lose you too badly in those complex concepts! If I did, the failing is mine, and I’ll happily entertain any questions you have over email. Through this article, you now have at least a passing understanding of the direction pecan breeding will be moving over the next decade. I am incredibly excited and optimistic about the future of pecan breeding. We are on the cusp of being able to use tools that were previously only available to the most well-provisioned and widely grown crops. 

Like all things pecan, the new support for the USDA program will take time to implement and enable us to fully realize the benefits of these tools in improving pecan breeding. Traditional pecan breeding practices are not being replaced. They will continue to be used alongside new layers of information created with these genetic tools, leading to greater accuracy, faster genetic improvement, and increased efficiency. All three tools have improved the quality and speed of breeding programs in other crop systems, and there is no reason to expect a difference for pecan. 

Bentley, N., Grauke, L. J., & Klein, P. (2019). Genotyping by sequencing (GBS) and SNP marker analysis of diverse accessions of pecan (Carya illinoinensis). Tree Genet. Genomes, 15(1), 403.
Bentley, N., Grauke, L. J., Ruhlman, E., Klein, R. R., Kubenka, K., Wang, X., & Klein, P. (2020). Linkage mapping and QTL analysis of pecan (Carya illinoinensis) full-siblings using genotyping-by-sequencing. Tree Genetics and Genomes, 16(6), 1–20.
Lovell, J. T., Bentley, N. B., Bhattarai, G., Jenkins, J. W., Sreedasyam, A., Alarcon, Y., Bock, C., Boston, L. B., Carlson, J., Cervantes, K., Clermont, K., Duke, S., Krom, N., Kubenka, K., Mamidi, S., Mattison, C. P., Monteros, M. J., Pisani, C., Plott, C., … Randall, J. J. (2021). Four chromosome scale genomes and a pan-genome annotation to accelerate pecan tree breeding. Nature Communications 2021 12:1, 12(1), 1–12.
Author Photo

Warren Chatwin

Dr. Warren Chatwin is a researcher part of the USDA ARS Crop Germplasm Unit in College Station, Texas. Find more info at