Post: Genomic Selection Indaba: Strategies to Revolutionize Tree Breeding in South Africa

October 11, 2024

Genomic Selection Indaba: Strategies to Revolutionize Tree Breeding in South Africa

Fikret Isik¹, Nanette Christie², Alexander A Myburg²

¹Department of Forestry & Environmental Resources, North Carolina State University, Raleigh, North Carolina, USA

²Forest Molecular Genetics (FMG) Programme, Department of Biochemistry, Genetics & Microbiology, Forestry and Agricultural Biotechnology Institute (FABI), University of Pretoria, South Africa

Genomic selection (GS) is a molecular breeding technique used in plant and animal breeding. Its application is expected to fundamentally change tree breeding over the next decade. This method uses a large number of DNA markers to predict the genetic merit of trees at the seedling stage for traits of interest.

A GS Indaba was organized to bring together key stakeholders in the South African tree breeding community. The aim was to explore the use of genomic applications and plan for future precision tree breeding in South Africa. The event took place on 26-27 October 2023 at Mbulwa Estate in Sabie, Mpumalanga. The Indaba was the culmination of the “Quantitative Genetics and Molecular Breeding” course led by Fulbright Visiting Professor Fikret Isik, in collaboration with Prof Zander Myburg and Dr Nanette Christie from the University of Pretoria’s Forest Molecular Genetics (FMG) Programme.

Before the Indaba, preparatory events were held, including an online course and two workshops. The first workshop in Pietermaritzburg earlier in 2023 focused on genetic data analysis, while the second, combined with the GS Indaba, explored hands-on genomic selection practices followed by strategic discussions on the application of GS in South African breeding populations. Prof Isik’s interactions with forestry companies in 2023 provided crucial insights into the state of conventional and molecular breeding efforts in the country.

Considering recent advancements such as DNA marker array development, large-scale genome sequencing projects for eucalypt trees and sequencing of a reference genome for tropical pines, the GS Indaba offered a timely assessment of South Africa’s readiness for molecular breeding approaches. A number of challenges relating to the history and design of local tree improvement programmes were identified.

Challenges in Implementing Genomic Selection in SA

South African forest tree improvement programmes have a long history especially for species such as Eucalyptus grandis. However, local tree breeding populations have not been designed for the implementation of GS. Implementing GS technology require well-structured, closed breeding populations with at least two generations for which accurate phenotypic and genotypic data is available. Current breeding efforts, especially for interspecific hybrids, often rely on crossing first-generation parents or even wild unimproved parents, followed by large-scale testing of clonal progeny from a handful of crosses. From these trials, a small number of superior genotypes are selected for commercial deployment. This means that most local tree improvement programmes are essentially “testing and selection” efforts. The feedback from the testing and selection phase to the breeding phase (which parents to cross for the next generation) is mostly lacking or neglected.
Genome-wide genotyping of forest tree species is still costly at approximately $25 per tree or seedling. Even though the current eucalypt and pine genotyping arrays have proven to be robust, cheaper DNA sequence-based technologies will be available in the near future. The FMG Programme led the development of a 50,000 SNP marker array (Pitro50K, Thermo Fisher Scientific) for tropical pine species. Similarly, a multi-species Eucalypt SNP array developed by EMBRAPA in Brazil has been used in eucalypt breeding programmes. Next, the FMG Programme will be working in collaboration with Prof Isik on a low-cost ($10-$12 per tree) AgriSeq SNP genotyping platform (Thermo Fisher Scientific) for pines, while for eucalypts, whole-genome resequencing will soon become a cost-efficient alternative.
The power of GS lies in being able to predict the genetic merit of trees when the candidates are still seedlings, thereby circumventing the lengthy and expensive field testing needed to assess the traits of those trees. The challenge for tree breeders is then to stimulate flowering in the juvenile plants and cross the GS selections in order to reduce the breeding cycle time. For cold-tolerant Eucalyptus species in SA, reproductive (late and uneven flowering) and propagation hurdles are critical limitations to developing GS approaches. Further research is needed to shorten the reproductive cycle and by stimulating early flowering of cold-hardy Eucalyptus species and pines.
The lack of skilled staff, databases, computational infrastructure and detailed GS plans are barriers to implementing GS in most organizations. A long-term GS plan is needed for successful implementation of GS. The plan should include a timeline, milestones, resources, and clear deliverables. Senior management support is crucial to ensure successful implementation and sustained progress in a breeding strategy that may span decades.
Highly skilled workforce turnover is a prevalent challenge, particularly as South Africa contends with the allure of opportunities abroad, leading to the departure of some of its finest talent.
Pedigree errors, loss of seedling identity and pollen source errors are particularly prevalent in tree breeding programmes and nurseries. Misidentification of pollen, seed, and seedlings can have severe and negative long-term implications in forest tree breeding.

Opportunities

Collaboration is key! The South African tree breeding community is tightly knit and cohesive. Its members demonstrate a willingness to share resources, collaborate on large-scale projects, and exchange material. Organizations are already collaborating under the Camcore (https://camcore.cnr.ncsu.edu/) umbrella to test species and provenances.
Despite the departure of some of its finest talent, South Africa boasts a cadre of proficient forest geneticists and tree breeders, demonstrating resilience in retaining a pool of well-trained professionals. These experts are well-positioned to spearhead advancements in molecular breeding, showcasing the nation’s commitment to fostering expertise in cutting-edge fields.
The South African tree breeding community is fortunate to have the full support and direct engagement of the FMG Programme at UP. This Programme, and the Precision Tree Breeding Platform hosted by it, has provided tremendous value and services to tree breeding programmes in SA, including high-throughput DNA isolation services, development of DNA markers, genotyping panels, and research projects addressing the challenges facing tree breeding and managed forests.
The DNA isolation service and the development of microsatellite markers for DNA fingerprinting and parentage analysis by the FMG is a great resource and opportunity for the SA tree breeding community. Arguably, the highest impact in SA breeding programmes has been the implementation of DNA fingerprinting to correct the misidentification of genotypes or species during all phases of breeding, propagation and deployment.

Conclusions and Recommendations

Based on the presentations, discussions during the GS Indaba, interactions of Prof Isik with tree breeding professionals, and site visits, the organizing committee reached the following conclusions and recommendations.

GS is not a silver bullet. It is an important tool in a tree breeders’ toolbox. Implementing GS in operational breeding programs is still expensive. It requires managing complex logistics and therefore a focus on species or hybrid combinations in which the benefits of GS can be realized. We recommend a collaborative approach under the Camcore umbrella with FMG support to start pilot projects focused on important species or hybrid combinations. This way, costs are shared, reducing the initial investment for each company. It will also help to create large and well-defined training populations to be used by the participants. This approach ensures a careful and effective introduction of GS, building a common foundation for success among different groups. If some well-resourced companies are interested in building their own GS programmes, we strongly recommend beginning with a pilot project to gain experience before expanding to multiple breeding populations.
A successful GS program requires well-managed, multiple-generation breeding populations with strong genetic linkage between generations. South African tree breeding programmes should allocate resources to improve well-suited landraces to local conditions rather than invest limited resources in experimenting with many new species and provenances. Breeding populations should be large enough to make genetic gain in the long term. Modern mating designs based on computer algorithms should be used to optimize multiple breeding objectives, such as increasing genetic gain and managing genetic diversity.
South African breeders should transition to modern experimental designs (e.g. incomplete block designs) to improve the quality of data collected from genetic field trials. Genetic tests need to collect high-quality data to predict the genetic merit of trees for selection. In South Africa, large randomized complete blocks with row plots are still the standard used in progeny testing. However, such large blocks at the test sites show high variation in elevation, soil fertility, and soil moisture, and are therefore highly inefficient for genetic tests.
Keep the genetic tests simple with one objective: separating noise from the genetic signal to select the winners. Genetic field tests should not be designed to accomplish multiple objectives, such as evaluating the effects of silvicultural practices (e.g., thinning, weeding, and fertilization), while also performing growth-yield modelling. Such practices reduce the efficiency of the progeny tests and complicate data analysis. Separate field tests for silviculture and growth-yield should be planned.

Genomic selection is an important new tool for recurrent selection strategies. We call it Recurrent Genomic Selection (RGS). If the infrastructure is in place, it could revolutionize forest tree breeding. Below, we provide an example from Pinus taeda breeding in the USA (Figure 1) that was presented at the GS Indaba.

In summary, the GS Indaba was pivotal for South Africa’s tree-breeding landscape, fostering critical reflection, strategic planning, and hands-on engagement to unleash the transformative power of genomics. Prof Isik’s (North Carolina State University, USA) leadership was instrumental in guiding the industry toward alignment with emerging trends and best practices, ensuring ongoing success in molecular breeding endeavors.

Classical selection results: The distribution of breeding values for tree height of the first (ACE1, pink density plot) and second generation (ACE2, blue density plot) populations of Pinus taeda. Classical selection based on the height data of the first generation moved the population mean for height in the second generation (from 20.5 m to 21.0 m). Selection works!

Genomic selection results: The same two genetically linked populations were genotyped with the same SNP markers to validate the GS approach. A GS model was developed based on the data of the first generation (ACE1, pink distribution). The height of the second generation (ACE2, green density plot) was predicted using DNA markers. Genomic selection moved the mean height of the population in the right direction (23.6 m to 24.2 m), even better than classical selection. Genomic selection works!

Figure 1: Comparison of classical and genomic selection approaches in Pinus taeda using a two-generation population (courtesy of the Cooperative Tree Improvement Program at North Carolina State University).

Source: Genomic selection (GS)

Back to NEWS:
BUSINESS l CLIMATE CHANGE l CONSTRUCTION l EDUCATION l ENVIRONMENT l FORESTRY ENGINEERING l FORESTRY l GOVERNMENT l INTERNATIONAL l LAND l RECYCLING l RESEARCH l ROOFING l SHORT HAUL l SILVICULTURE l SOCIAL RESPONSIBILITY l TRANSPORT l TREATMENT l TRANSPORT l VALUE ADDING