USDA deregulated a genetically engineered forest tree — Darling 54 American chestnut, blight-tolerant, six years and seven months after the petition, and with the environmental impact statement abandoned (APHIS, 28 Aug 2026)

Why it matters (spine). The garden's non-toxic, human-health spine usually meets agricultural biotechnology as a chemistry question — what gets sprayed, on what, at what dose. This is the same regulatory instrument pointed at a different kind of organism: not a row crop with a herbicide-tolerance trait, but a native forest tree engineered for tolerance to the fungal pathogen that caused chestnut blight. It is the first card in the cabinet where the deregulated organism is a forest tree rather than a crop — a claim scoped to our own shelf, not to the world — and it is worth holding precisely because the honest reading is neither triumphant nor alarmed.

A framing boundary, set before the facts. Blight-tolerant American chestnut is widely discussed as a restoration project, and the petitioner is a forestry college. But the Federal Register notice read for this card states only that the line was "developed using genetic engineering for tolerance to the fungal pathogen." It contains no statement of restoration intent, no release plan, and nothing about the present ecological status of Castanea dentata. So this card treats restoration as reported context about the petitioner's objective — not as something the primary says, and not as a fact the tier vouches for. The intended-use language, if it exists, is in the petition, which was not opened.

What the notice says. "We are advising the public of our determination that Darling 54 American Chestnut (Castanea dentata), which was developed using genetic engineering for tolerance to the fungal pathogen, Cryphonectria parasitica, is no longer considered regulated." The determination rests on the petition data, available scientific data, a plant pest risk assessment, and public comments. The operative finding: "APHIS has determined that Darling 54 is unlikely to pose a greater plant pest risk than the nonmodified comparator and therefore is no longer subject to our regulations in 7 CFR part 340."

Document identity: Federal Register citation 91 FR 55510, FR document number 2026-17596, agency docket APHIS-2020-0030, two pages, publication date 28 August 2026, and "This change in regulatory status is recognized as of August 28, 2026." Signed "Done in Washington, DC, this 24th day of August 2026" by Sarah Helming, Acting Administrator, and filed on public inspection 27 August 2026 at 8:45 am. Petitioner is the State University of New York College of Environmental Science and Forestry (ESF), contact Mr Alan Pearson of APHIS Biotechnology Regulatory Services.

The naming correction is in the primary itself, and is worth reading. The petition was filed on 21 January 2020 for a line called Darling 58. The notice explains why the deregulated line is called Darling 54: "ESF's original January 2020 petition referred to Darling 58 instead of Darling 54. These two lines were produced at the same time, using the same transgenes in the same genetic background, so they express the same protein products. Multiple analyses in 2023 revealed that essentially all the offspring from the initial crosses were derived from Darling 54, due to a labeling error in 2016." ESF submitted a revised petition in August 2024 to correct the line name, and APHIS revised both the draft environmental impact statement and the draft plant pest risk assessment as a result. A regulatory file running six years and seven months from petition to determination, in which the identity of the organism under review had been mislabelled for seven years — 2016 to the 2023 analyses — discovered by the petitioner's own work and corrected on the record, is a data-integrity story as much as a biotechnology one. And the correction came from the primary, not from us.

The comment record inverted, and the notice states it without explaining it. Four rounds, as recorded in the notice:

Round	Date	Comments received / posted	Direction of the posted majority
Petition published	19 Aug 2020	4,337 / 4,321	"the majority were in favor of deregulation"
Notice of intent to prepare an EIS	6 Aug 2021	3,968 / 3,964	"a large majority were opposed"
Draft PPRA + draft EIS	10 Nov 2022	30,455 / 9,289	"the majority were in favor"
Revised petition, revised draft EIS and PPRA	6 Jun 2025	174,116 / 4,115	"the large majority were in opposition"

The notice also records that on the 2022 round "three were submitted by EPA stating that 'EPA did not identify significant environmental concerns to be addressed in the Final EIS.'" The gap between comments received and comments posted is explained in the notice's own footnote about multi-submission comment IDs at Regulations.gov; the reasons for the swings in direction are not explained, and this card does not speculate. What it does record is that the final round drew by far the most comment volume of the four and its posted majority opposed the action that followed.

The environmental impact statement was abandoned, not completed. "However, on August 18, 2025, APHIS published a withdrawal of the NOI (90 FR 40050; Docket No. APHIS-2025-0029). Therefore, consistent with this August 18, 2025 Federal Register notice, APHIS terminated work on the EIS." The determination therefore rests on the plant pest risk assessment alone. Given that APHIS had itself judged an EIS warranted in 2021 and then withdrew that judgement, the difference between a completed EIS and a terminated one is the difference between an ecological analysis and a pest analysis. Naming that is not an argument that the decision is wrong; it is a statement of what evidence exists.

Scope discipline: four things this card does not claim. It does not name the transgene or the protein product — the notice says the two lines "express the same protein products" and names neither, and nothing on this card fills that gap. It asserts nothing about human dietary exposure or food safety; chestnut is a food-bearing tree and the question is adjacent, but the document read here is a plant-pest determination and APHIS's finding covers plant pest risk only. It asserts nothing about restoration intent, release plans, or the ecological status of the species, per the boundary set at the top — the notice is silent on all three. And it makes no world-scoped priority claim: the notice does not describe this as a first of any kind, and the only "first" on this card is cabinet-scoped, that the deregulated organism here is a forest tree rather than a crop.

Money angle and metacrisis lever. Money angle: thin and honest. There is no product, no price and no market in this document; the petitioner is a state university, not a company. The commercial read, if any, is second-order — a deregulated engineered forest tree shows the 7 CFR 340 pathway can be completed for a tree species, which is a regulatory precondition for any future business in engineered forestry (timber, carbon, nut crops, resistant natives). That is a precondition, not a market, and the card claims no more. Metacrisis lever: in my judgement this is the agriculture shelf's most direct metacrisis lever, which is a judgement and is labelled as one. A regulator has cleared a genetically engineered tree of the species that chestnut blight removed from commercial and canopy significance — and whether engineered intervention in damaged ecosystems is a tool that gets used, and under what evidentiary standard, is the question sitting behind it. The ecological history in that sentence is reported context, not something this notice states.

What would prove it wrong (falsifier). Three limbs.

• The determination fails if the notice is corrected, withdrawn or stayed, or if litigation vacates it — an action taken on a PPRA after the EIS was terminated is a plausible target for a procedural challenge under NEPA, and the 174,116-comment final round shows an organised opposition exists. Nothing in the document indicates litigation, and none is asserted here.
• The scientific claim fails at the plant-pest boundary: APHIS found no greater plant pest risk than the nonmodified comparator, and would be wrong if Darling 54 turns out to pose plant pest risk the PPRA missed. It cannot be wrong about ecological, food-safety or gene-flow questions it did not ask.
• The restoration premise, which is reported context rather than a claim of this notice, fails on a timescale no regulatory document can settle: if blight tolerance does not translate into surviving, reproducing mature trees in the field over decades, then a deregulation is a permission and not a recovery. The test is field data on established plantings, and the garden holds none.

Garden-code note (L41, flagged not resolved). The toxicology-adjacent thread here — a genetically engineered food-bearing tree, with the dietary-exposure question named and left open — would take TKG in the Frontier Map's sense of Toxicology, while this same run's MCP card carries machine-layer content that the beat file would tag TKG for the machine layer. Tagging both senses in one run is the guess L41 forbids, so TKG is left out of this card's gardens list — as ebay-depop-acquisition-closed did on 30 July, and as the corn card did on 21 August — and the domains tag agriculture plus the body's stated toxicology boundary carry the link until the founder decides.

Provenance (what I actually read). The Federal Register document page for FR Doc. 2026-17596, fetched from the main thread on 28 August 2026. Dates are on the face of the document in three independent places — the publication date field ("Publication Date 08/28/2026"), the DATES section ("This change in regulatory status is recognized as of August 28, 2026"), and the signature block ("this 24th day of August 2026") — plus the public-inspection stamp of 27 August 2026 and the page metadata article:published_time: 2026-08-28. Every quotation, every comment count, every prior-notice date and the docket and citation numbers above are read from that page. The table is a restatement of counts printed in the notice's SUPPLEMENTARY INFORMATION; no count in it is derived or rounded.

Derived figures on this card, enumerated: three date intervals, all my arithmetic on dates the notice gives — six years and seven months (petition 21 January 2020 to determination 28 August 2026), seven years of mislabelling (2016 error to the 2023 analyses), and twenty-four days between the corn deregulation of 4 August 2026 and this one. There are no other derived figures.

I did not open the plant pest risk assessment, the response to comments, the signed determination, the original or revised petitions, the withdrawn notice of intent at 90 FR 40050, the Regulations.gov docket, the govinfo PDF, or any SUNY ESF publication — so this card describes what APHIS decided and on what record, and nothing about the underlying science beyond APHIS's own stated conclusion. Two things on this card are explicitly not from this primary and are labelled as such where they appear: the restoration framing (reported context about the petitioner's objective) and the ecological history of chestnut blight (background). Neither is vouched for by the tier. The cross-referenced corn card was opened in the repository to confirm its date and the instrument criticism attributed to it.