Closing the Loop

Every phase of the American food system has been mapped as a node in a largely linear flow: land to inputs to production to harvest to market to plate. That framing is accurate and also incomplete. Embedded within the line — and conspicuously absent from it in places where absence is a policy choice — are feedback loops. Where they function, they are among the most economically and ecologically efficient features of the entire system. Where they don’t, the waste is measurable and the gap between current state and possible state is not technical. It is a matter of incentive structure, market concentration, and political will.

What It Is

A cycle, in systems terms, is a closed-loop material or information flow: output from one phase becomes input to another, reducing net throughput, lowering cost, and often diminishing ecological impact. The food system is predominantly linear — extract, produce, ship, consume, discard — because linearity serves the interests of input suppliers who profit from repeat purchases and processors who profit from volume. Cycles reduce both.

That said, several significant cycles already operate at industrial scale. They exist not because they are virtuous but because they are profitable, and understanding why they closed — while others did not — reveals what it would take to close the rest.

Cycles That Work

These loops run continuously at industrial scale. They exist because the economics justified the infrastructure required to close them.

Cycles That Are Broken

These loops exist in structural form — the material and economic logic for closing them is sound — but market design or policy has degraded or blocked them.

Seed saving. For most of American agricultural history, farmers retained 10–15% of their harvest to plant the following year. This practice allowed crops to locally adapt, gave farmers price leverage against input suppliers, and created a genuine biological feedback loop between a specific farm’s soil conditions and its genetic material. Today, more than 90% of US corn, soybean, and cotton acreage is planted with patented genetically engineered varieties (USDA ERS, Recent Trends in GE Adoption). Saving and replanting these seeds violates technology use agreements enforceable by civil litigation. The loop was not outcompeted — it was made illegal.

Precision planting feedback. Every modern planting pass generates gigabytes of soil condition, seed placement, and germination data. This is the exact signal needed to close a feedback loop: observe what worked, adapt the next season’s inputs. But the data is retained by the equipment manufacturer, not the farmer. John Deere’s Operations Center and Climate Corporation’s FieldView (a Bayer property) hold the largest precision agriculture datasets in the world — and farmers cannot export their own data in formats usable outside the OEM ecosystem. The loop is technically present but informationally locked.

Biosolids cycling. Municipal wastewater treatment produces approximately 7 million dry tons of biosolids annually in the US — nutrient-rich organic matter that is currently applied to roughly 55% of US farmland as fertilizer (EPA Biosolids Basics). This is a functioning cycle. It is now threatened: EPA’s ongoing risk assessment of PFAS contamination in biosolids may result in restrictions that force municipalities to incinerate or landfill what is currently a productive nutrient stream, at an estimated annual cost increase of $2–4B for water utilities (Water Research Foundation, 2022).

Cycles That Don’t Exist Yet

These loops have no meaningful operational presence in the current system. The infrastructure, economics, or policy conditions to close them are absent. The gap is not technical.

Why It Matters

The working cycles in the American food system — DDGS, manure application, rendering, reusable crates — are not accidents of ecological consciousness. They closed because a private company calculated that the return on closing them was positive. That is the template for everything else.

The potential cycles catalogued above share a structural feature: their return is real, their barrier is not technical, and their benefit is distributed in a way that reduces the incentive for any single actor to bear the cost of infrastructure. Municipal AD facilities benefit ratepayers and farmland; neither group controls the utility budget. Backhaul routing benefits shippers and carriers; neither group controls the others’ routes. Gleaning benefits food banks and communities; the grower bears the coordination cost. These are coordination failures, not market failures, and they respond to policy instruments — mandates, subsidies, standardized data infrastructure — that other countries have applied with measurable results.

The food system will not close these loops on its own. The linear model is profitable for the firms that control it. But the cumulative cost of not closing them — in methane, in wasted nutrition, in depleted soil, in empty trucks burning diesel — accrues to everyone else.

What’s Being Done

The coordination failures and legal barriers described above are real — but so are the people working around and through them. Across right-to-repair law, phosphorus recovery technology, PFAS destruction pilots, and open-source seed systems, the distance between broken and functioning is shorter than it has been in decades.

Efforts Showing Results

John Deere $99M Right-to-Repair Settlement and FTC Case. In April 2026, Deere agreed to a $99 million class action settlement requiring the company to make digital repair tools available to all owners and independent repair providers for large agricultural equipment for ten years. Separately, the FTC and attorneys general of Illinois and Minnesota filed an antitrust suit that survived Deere’s dismissal motion in June 2025 and is proceeding toward discovery. Six states have enacted Right to Repair laws and over 33 additional bills were introduced in early 2026. The settlement creates a precedent and a decade-long implementation obligation; the FTC case could force industry-wide change beyond a single company. The next lever is enforcement: monitoring whether Deere actually delivers accessible, affordable diagnostic tools — and whether independent shops can use them without prohibitive licensing fees.

Ostara Struvite Phosphorus Recovery at Chicago’s Stickney WWTP. Ostara Nutrient Recovery Technologies operates the world’s largest struvite crystallization facility at Chicago’s Stickney Water Reclamation Plant, producing approximately 9,000 tonnes per year of Crystal Green slow-release phosphorus fertilizer recovered directly from municipal wastewater. More than 35 full-scale commercial installations now operate across North America, Europe, and Asia. Struvite crystallization is technologically mature and economically viable at municipal wastewater scale — the main barrier to broader adoption is replication to agricultural AD systems at smaller scale, where cost-sharing programs and state nutrient trading credits could make the economics work.

CHAR Tech / Synagro Pyrolysis Pilot for PFAS Destruction. In May 2025, CHAR Tech and Synagro partnered with Baltimore City’s Department of Public Works to test a commercial-scale High-Temperature Pyrolysis system at the Back River facility. The process destroys PFAS to below detection limits in biochar output, with an average 97% removal rate, while producing syngas for energy recovery. A 2024 techno-economic analysis confirmed the process simultaneously eliminates microplastics (91–97% removal) and pharmaceuticals (over 99.9% removal). The Baltimore pilot is the critical commercial-scale proof point; the remaining barriers are high capital costs and the need for EPA interim guidance on classifying pyrolysis outputs so municipalities are not left in regulatory limbo.

Open Source Seed Initiative (OSSI). OSSI operates a copyleft-style pledge system for plant varieties, modeled on open-source software licensing. As of 2025, 550+ varieties from 52 plant breeders and 78 seed company partners are committed to never being restricted through patents or license agreements — farmers can save, share, trade, and breed from pledged seed indefinitely. OSSI operates primarily in vegetable and specialty crops, where its influence on market demand is already measurable. Expanding into commodity crop breeding — where patent restrictions are most economically damaging — would require public university breeding programs explicitly releasing varieties under the pledge, funded as a public good.

Where More Work Is Needed

Viable economics for municipal and small-scale anaerobic digestion. The financing model for AD broke when the Section 48 Investment Tax Credit expired for new projects after December 2024, and standalone RNG/biogas does not qualify for Section 48E. Transport costs make it uneconomical to aggregate food waste from dispersed sources to centralized facilities. Plastic contamination in source-separated organics degrades biogas yield and contaminates digestate. The most promising near-term approaches are co-location of AD with large food waste generators — hospitals, universities, food processors — to eliminate transport costs, combined with a legislative fix to allow standalone RNG projects to qualify for the Section 48E technology-neutral clean energy credit. Denmark’s network of over 100 centralized biogas plants demonstrates that regional logistics can be solved with policy coordination; South Korea’s mandatory food waste separation program has dramatically reduced plastic contamination. Neither model requires a new technology — only the administrative will to apply existing tools.

PFAS destruction infrastructure at scale. Pyrolysis can destroy PFAS at lab and pilot scale, but commercial-scale infrastructure does not exist at the capacity needed to handle US biosolids volumes. Incineration can destroy PFAS but faces air emissions permitting challenges. Meanwhile, EPA’s risk assessment is incomplete — leaving municipalities unable to safely land-apply biosolids but without approved alternative pathways. The result is a perverse default: regulatory uncertainty pushes biosolids toward landfills, which solves nothing. Scaling destruction infrastructure requires federal investment framed as environmental remediation (following the Superfund precedent), EPA fast-track approval for validated technologies, and liability frameworks that push PFAS costs upstream to manufacturers — the Maine model has already demonstrated this approach can be enacted at the state level.

Commodity crop seed saving access. Over 90% of corn and soybean acreage is planted with utility-patented varieties that carry no farmer seed-saving exemption — unlike protections under the Plant Variety Protection Act. The Open Source Seed Initiative has proven that copyleft mechanisms work for plants, but it has not yet penetrated commodity crop breeding where the economic stakes are highest. A federal farmer seed-saving exemption analogous to the PVPA research exemption, or sustained public university investment in open-germplasm commodity crop varieties, could shift this calculus without requiring the elimination of private seed patents — only a carve-out at the farmer end of the chain. The International Treaty on Plant Genetic Resources for Food and Agriculture provides multilateral models the US could adapt for domestic law.

The agricultural circular economy is not a utopian vision — it is a partially working system with identifiable bottlenecks, and the distance between where we are and where we could be is shorter than it looks. Phosphorus recovery from wastewater is already happening at commercial scale in Chicago. Over 191 farms are capturing methane from manure and selling it as vehicle fuel. Farmers just won a $99 million settlement against the company most responsible for locking them out of their own equipment. These are not pilot projects or press releases — they are functioning systems that can be replicated and expanded. The honest assessment is that the economic signals and policy infrastructure needed to scale these solutions are still incomplete, but they are closer to sufficient than they have been at any previous moment. Every bottleneck in the chain is either a regulatory failure, a market design problem, or an infrastructure gap that specific people with specific authority can fix. None of these fixes require waiting for a technological breakthrough — only the less glamorous work of changing procurement policies, updating tax credit language, building regional processing hubs, and demanding repair access to equipment you already own.