Let me work with what I already have and do one more targeted search. Let me review the detailed content of my key results before writing. Now I have enough information to write a well-sourced synthesis. Let me also check the Virginia Tech genome study and the R3-5 more fully. Now I have comprehensive, well-sourced material. Let me compose the synthesis.
---
The single most promising near-term path to field-deployable blight-resistant American chestnuts is The American Chestnut Foundation's recurrent genomic selection (RGS) breeding program, which has pivoted from the original Burnham backcross plan toward a marker-assisted strategy that can accelerate gains per generation. Starting in 2023, TACF began making controlled crosses between selected parents, and by 2026 the program reported 115 crosses made, 6,878 trees genotyped, 2,150 seedlings inoculated for blight screening, 5,403 trees planted, and 3,810 seeds sown in the greenhouse. The program pursues four parallel breeding tracks: three using backcross hybrid trees to maximize blight resistance, balance blight and Phytophthora root rot (PRR) resistance simultaneously, or maximize PRR resistance — all while maintaining a minimum of 70% American chestnut ancestry — and a fourth track that maximizes blight resistance among offspring of large surviving wild American chestnuts. For the hybrid tracks, selected offspring are expected to carry between 70% and 85% American chestnut ancestry with a mean of 75%, indicating that substantial gains in resistance can be balanced with selection for high American identity. This multi-track approach is more scientifically robust than any single-gene strategy because it draws on polygenic, quantitative resistance from Chinese chestnut, and it integrates PRR resistance — a pathogen that is increasingly recognized as a co-equal threat in the southern portion of the species' range.
The most significant recent scientific development is the landscape genomics work on *Castanea dentata* that is now being leveraged directly for restoration planning. A 2024 PNAS paper provided a genome-guided framework for incorporating climate-adaptive diversity into blight-resistant restoration populations. A Virginia Tech–led study used deep learning to predict geographic origins from genome sequences. Using deep learning software, researchers were able to predict the geographic origins of a particular genome sequence by training the model using trees with known origins, and results showed that TACF was doing a good job producing trees with adaptive diversity, although attention must be paid to not losing this diversity through further breeding for blight resistance. Moving forward, this information can help the foundation restore specific blight-resistant American chestnut families according to regions in which their genome is best matched. This matters because blight resistance alone is necessary but insufficient — restored populations must harbor enough standing adaptive variation to respond to future climatic shifts and novel pathogens. A genomically informed deployment strategy prevents the genetic bottleneck that could doom even blight-resistant plantings to long-term failure.
The biggest remaining bottlenecks are both institutional and scientific. On the transgenic front, Darling transgenic American chestnuts with enhanced blight tolerance are still undergoing federal regulatory review by the EPA, USDA-APHIS, and the FDA, with timing not predictable for any of the agencies, and the EPA specifically may involve a multi-phased approval process. ESF's 2024 Progress Report states that they believe "Darling is suitable for release and further study as part of the restoration effort," yet the Center for Food Safety has challenged the adequacy of research and safety studies in its comments to USDA. On the breeding side, fundamental scientific uncertainty persists: after 30 years of backcross breeding under the Burnham plan, the number of loci that control blight resistance is not known with confidence, and it is uncertain whether alleles that confer blight resistance are segregating at the same or different genetic loci across sources of resistance. Meanwhile, Phytophthora root rot resistance is now explicitly integrated as a co-equal objective in TACF's backcross breeding program, reflecting growing recognition that blight resistance alone is insufficient for restoration in southern portions of the species' range — but PRR screening infrastructure and knowledge of its genetic architecture lag far behind work on blight.
New funding would have the highest impact-per-dollar in three areas. First, scaling TACF's genotyping and phenotyping capacity is the rate-limiting step; thirty years of backcross breeding involving over 50,000 trees has resulted in approximately 500 trees that contain on average only 12% Chinese chestnut DNA but have inherited at least some blight resistance and exhibit desired American chestnut character. After artificially inoculating seed orchards with the chestnut blight fungus and culling individuals with significant canker expansion, 5,000 trees remain from which to make the final selections of the 500 most resistant trees. Increasing throughput of genomic marker assays and inoculation trials would compress the timeline to identify elite parents by years. Second, supporting ESF's next-generation transgenic pipeline — particularly the DarWin constructs with tissue- and wound-specific promoters — would accelerate the development of a potential regulatory-ready successor to Darling 58. Third, investment in wild germplasm conservation is urgent: wild American chestnut germplasm conservation is a core restoration component, with genomic studies emphasizing the need to capture local adaptive diversity across the former range to buffer restored populations against future environmental change, and this living material is being lost as surviving root-sprout colonies senesce without reproducing. Funding directed at these three leverage points — genomic-selection throughput, next-generation transgenics, and germplasm banking — would compound across all restoration tracks and substantially shorten the path to landscape-scale reintroduction.