Mapping the Response to Plastic Pollution in the Great Lakes: the full narrative synthesis. Every claim cited and confidence-flagged.
A proof-of-concept synthesis by Shay Cranmer, with AI research collaboration (see Methodology). Version 1, published July 2026.
Confidence flags: ⬤ Strong (peer-reviewed or government-published) · ◐ Moderate (single study, advocacy-reported, or dated) · ○ Extrapolated (assumption flagged). Every claim carries a numbered source; full list in Appendix A.
Every year, roughly 22 million pounds of plastic enter the Great Lakes.[1] ⬤
That number, about 10,000 metric tons from modeling by Matthew Hoffman's team at the Rochester Institute of Technology, is the one this report is named for, and it is worth sitting with before any other. Twenty-two million pounds is not an ocean statistic borrowed for local effect. It is ours: it enters through the rivers of Chicago and Toronto, off the beaches of Cleveland and Detroit, and it largely stays, because the Great Lakes do not flush the way oceans circulate. What comes in, settles in.
The load is not evenly shared. Lake Michigan alone receives roughly half the total, about 5,000 metric tons a year, followed by Erie (2,500 MT), Ontario (1,400 MT), Huron (600 MT), and Superior (32 MT).[1] ⬤ Hoffman's particle modeling shows the plastic does not stay where it enters: debris from Chicago and Milwaukee drifts to the eastern shores of Lake Michigan; Detroit's and Cleveland's washes to southern Lake Erie.[1] ⬤ One jurisdiction's discharge becomes another jurisdiction's beach. This single fact, that pollution crosses borders while policy does not, underlies most of what this report finds.
Forty million people drink Great Lakes water.[2] ⬤ The University of Toronto's Rochman Lab has found that 90 percent of water samples from the lakes exceed safe microplastic thresholds for wildlife,[3] ⬤ and that every single fish they sampled from the Great Lakes contained microplastics: 100 percent, compared with roughly 25 percent in Pacific Ocean sampling.[3] ⬤ Shedd Aquarium researchers found microplastics in 85 percent of fish in three Lake Michigan tributaries, averaging 13 particles per fish.[4] ⬤
The plastic is in the sediment too, and it has been accumulating for decades. Sediment cores show microplastic deposition increasing steadily since the 1960s.[6] ◐ Bottom samples have measured 431 particles per kilogram in Lake Erie sediment,[7] ◐ and on the lake's surface downstream of Detroit, Cleveland, and Erie, concentrations have reached 466,305 particles per square kilometer.[8] ◐
The Alliance for the Great Lakes' Adopt-a-Beach program, 25 years running and the region's longest-lived community cleanup effort, reports that 86 percent of the litter volunteers collect is partially or fully plastic, and about 40 percent is "tiny trash": fragments under 2.5 centimeters, too small to pick up faster than they arrive.[9] ⬤ In 2023 alone, 11,342 volunteers logged 28,312 hours and removed 573,608 pieces of litter.[9] ⬤ Since 2003, the running total exceeds ten million pieces.[9] ⬤
A distinct and under-watched stream is industrial: nurdles, the pre-production plastic pellets that ships and trains carry by the carload. When a Canadian Pacific train derailed near Rossport, Ontario in January 2008, four cars of polyethylene pellets went into Lake Superior; nurdles were still washing ashore near Nipigon eight years later.[10] ◐ A survey of 66 Great Lakes beaches found significant pellet accumulation on 42 of them, averaging 19.1 pellets per square meter.[11] ◐ No dedicated nurdle monitoring program exists anywhere on the lakes.[12] ⬤
Every figure above should be read as a minimum. As Section 4 details, some of the potentially largest inputs, among them tire-wear particles, atmospheric deposition, and nanoplastics, have barely been studied in the Great Lakes at all. The measured problem is the part we have thought to measure.
One more number, and then this report turns to the people doing something about it: the four cities contributing the most plastic to the lakes are Chicago, Toronto, Cleveland, and Detroit.[1] ⬤ Those same four cities are home to some of the most capable environmental organizations on the continent. The problem and the response live at the same addresses. Keeping the whole of that response in view, the map of it, is the subject of the rest of this report.
If Section 1 described the water, this section describes the room. More than 200 organizations work on Great Lakes protection across eight U.S. states, two Canadian provinces, multiple tribal nations, and two federal governments.[13] ⬤ On plastic pollution specifically, the room includes federal scientists, state regulators, tribal resource managers, academic labs, advocacy coalitions, and thousands of volunteers, each holding a piece of the picture that no one else holds, because it is exactly the granular, local, close-to-the-water knowledge a field like this runs on. It is not an empty field. It is a rich one, and with a population of working bodies this dense and this varied in mandate, geography, and research aim, keeping the whole picture in view is difficult for anyone. That is not a failing of the pieces; it is the nature of a puzzle this size, and there is no picture at all without them.
This section maps who is doing what. It is an attempt, the first we are aware of though we would be glad to learn otherwise, to gather the organizational layer of Great Lakes plastic pollution work in one place: not the science (well reviewed elsewhere[14]) and not the projects (tracked in GLRI's database[15]), but the actors themselves: their focus, their asks, and where they align without always having a way to see it.
The U.S. EPA's Great Lakes National Program Office has coordinated binational water quality work since 1983 under the Great Lakes Water Quality Agreement, and operates GLENDA, the region's central water-quality database.[16] ⬤ The International Joint Commission, the 117-year-old binational body that oversees the Agreement, took its most consequential plastics step in November 2024, when its Science Advisory Board's Final Report on Microplastics proposed standard monitoring procedures, a risk-assessment framework, and formal designation of microplastics as a "Chemical of Mutual Concern" under the Agreement.[17] ⬤ That designation has not been acted on. The recommendation sits, fully drafted, waiting for governments to pick it up. That fact matters later in this report.
The money flows chiefly through the Great Lakes Restoration Initiative: $369 million in FY2026 across fifteen federal agencies and more than 2,000 tracked projects.[18] ⬤ NOAA contributes research capacity through its Great Lakes lab and funds cleanup and circular-economy partnerships through its Marine Debris Program, alongside the NCEI Marine Microplastics Database, one of the few fully public, machine-readable plastics datasets in the region.[19] ⬤
No two Great Lakes states regulate plastic alike, and the differences are not subtle.[20] ⬤ New York has the region's strongest policy stack: a statewide bag ban since 2020, and a foam ban since 2022 that expanded in January 2026 to cold-storage containers and hotel toiletries.[21] ⬤ Illinois led the original microbead fight and continues to tighten state-agency use of single-use plastics. At the other pole, Michigan, the state with more Great Lakes shoreline than any other, enacted a 2016 law that prohibits its own municipalities from banning bags at all.[20] ⬤ Wisconsin, Minnesota, and Indiana carry similar preemption laws. Ohio and Pennsylvania have simply not acted.
Two features of this patchwork deserve to be said plainly. First: half the Great Lakes states have laws preventing local plastic action, which means the level of government closest to any given beach is, in four of eight states, legally barred from responding to what washes up on it. Second, the Michigan paradox: the same state that preempts local action is investing $2 million in microplastic research through 2029.[20] ◐ Michigan is paying to learn things its own cities are pre-forbidden to act on. No one designed this contradiction. It becomes visible only when the funding ledger and the statute book are laid side by side, a view few people outside either system have much occasion to take.
The patchwork matters more because the plastic moves. As Section 1 noted, particle modeling shows Chicago's and Milwaukee's plastic drifting to Michigan's eastern beaches, and Detroit's and Cleveland's to southern Lake Erie.[1] ⬤ Discharge in a preemption state becomes cleanup in another state's jurisdiction. Pollution crosses borders; policy stops at them.
The Great Lakes Indian Fish & Wildlife Commission, Anishinaabe-led and representing tribes across ceded territory in Wisconsin, Michigan, and Minnesota, frames water contamination as a treaty-rights and food-sovereignty issue, and has pioneered integration of Traditional Ecological Knowledge into resource monitoring.[22] ⬤ Its position, increasingly echoed at the binational level, is co-management, not consultation: management actions can harm tribal resources when tribes are not at the table as governments.[22] ⬤ The IJC's own recent program review now recommends that First Nations, Métis, and Tribal governments participate as full partners in the Water Quality Agreement's review process.[23] ⬤ Indigenous-led research infrastructure is being built to match, including a new Haudenosaunee-led institute with data sovereignty as an explicit goal.[24] ⬤
A note on what this report does not contain, by design: no community-owned Indigenous datasets; information about tribal organizations is limited to what those organizations publish publicly about themselves. Indigenous data is governed by the CARE principles (Collective Benefit, Authority to Control, Responsibility, Ethics) and belongs in a synthesis like this one only through partnership, on tribal terms.[25] ⬤ This is a methodology commitment, not a gap apology; the architecture of this project was designed from the start to accommodate tribal data under tribal authority, if and when nations choose to bring it.
The advocacy layer is where the region's plastic work is most visible, and also where it is most fragmented. The Alliance for the Great Lakes runs Adopt-a-Beach, publishes the Plastic Free Great Lakes toolkit, and asks for extended producer responsibility (EPR), single-use phase-outs, preemption repeal, and sustained GLRI funding.[26] ⬤ The Ecology Center in Ann Arbor works the microplastic policy file, asking for the Chemical of Mutual Concern designation, EPR, and repeal of Michigan's preemption law.[27] ⬤ The Great Lakes Plastic Cleanup, a binational partnership of Pollution Probe and the Council of the Great Lakes Region, has removed more than 244,000 pieces of litter with cleanup technology across 140+ partner organizations.[28] ⬤ Circular Great Lakes pursues a zero-plastic-waste strategy; Shedd Aquarium runs education, cleanups, and coalition advocacy; Surfrider's Great Lakes chapters push the federal pellet bill; Beyond Plastics and PIRG work the federal file; the Michigan Environmental Council works the state one.[26] ⬤ Above them all, the Healing Our Waters Coalition aggregates more than 200 member organizations into the region's largest advocacy voice.[13] ⬤
Read that paragraph again, though, and something hopeful surfaces. EPR appears in the asks of at least three organizations, independently. The Chemical of Mutual Concern designation is asked for by the Ecology Center, and it has already been formally recommended by the IJC's science advisors.[17][27] ⬤ Preemption repeal appears on more than one list.
A word about what this alignment does and does not show. The people inside these organizations very likely know of one another's positions; this is a field that shares coalitions, conferences, and often staff. What does not exist is any public, durable artifact of the alignment. The shared asks live in professional networks and institutional memory, which means they are invisible to everyone outside those networks: to a legislator counting who supports a bill, to a funder looking for a coalition to back, to a new organization deciding where to add weight rather than duplicate it. Knowledge held in relationships is real, but it retires when people do, and it cannot be handed to a newcomer or laid on a decision-maker's desk. The region's plastic advocacy contains at least three points of policy alignment that function, to the outside world, as if they did not exist. Making the field's own knowledge durable, shareable, and legible beyond its personal networks is precisely the work this synthesis exists to do.
The science backbone runs through a handful of labs. RIT's Matthew Hoffman produced the loading estimates this report is named for, and now co-leads the $7.3M Lake Ontario Center for Microplastics and Human Health, the region's first sustained look at human health effects.[1][29] ⬤ The University of Toronto's Rochman Lab supplies the ecological-risk numbers.[3] ⬤ Michigan Tech, Central Michigan, and the University of Michigan–NOAA CIGLR partnership anchor broader Great Lakes research capacity.[30] ⬤
One project deserves its own paragraph. Wayne State University, with $1.19M from the Great Lakes Protection Fund, is building an open-source, AI-powered library for identifying the sources of microplastics, in partnership with three watershed organizations.[31] ⬤ Their tool is a microscope: where is this particle from? This project is a map: who is doing what about it? The two are complementary by construction: source-attribution findings only become action through the organizational layer a synthesis like this one makes visible. As of this writing, we could find no public, purpose-built channel to carry Wayne State's findings to the 200+ organizations who could act on them; absent one, those findings will travel the way most knowledge travels in this field, through personal and professional networks, at the speed and reach of whoever is in the room. That works remarkably well, until the person carrying it moves on. If a more durable channel would serve, this report and its companion map are offered as raw material toward one.
The region does not lack coordinating institutions: the Great Lakes Commission (an interstate compact since 1955), the Conference of Great Lakes Governors and Premiers, the Council of the Great Lakes Region, the binational Fishery Commission.[32] ⬤ Nor does it lack data platforms. The Water Quality Portal holds 430 million records; GLOS streams real-time sensor data; Great Lakes DataStream shares community monitoring; GLRI's database tracks 2,000+ projects.[33] ⬤
There is no lack of coordination here, and this report does not arrive to correct anyone. What it hopes to add is one further layer of synthesis: a map of the territory itself, showing who is working on what, where effort may be duplicated, and where the web of advocacy and knowledge runs thin.[33] ◐ The commissions coordinate governments; the platforms gather measurements; the coalitions gather voices. The picture of the whole board, the actors laid out in one view, mostly lives where hard-won field knowledge usually lives: in the memories and relationships of the people doing the work. In a field as wide as inland oceans, spanning the jurisdictions of two nations, keeping that picture current is genuinely hard, and few organizations have spare capacity to hold it in public form on top of their actual missions. This project's aim is to draw it from the field's own record and keep it in view: for the organizations working tirelessly on their own corners of the board, for the curious or worried newcomer who does not yet know the web exists, and for the legislator or funder who needs the whole picture at a glance.
A related note about reach, not existence: some of the region's richest data is, at present, also its hardest to get to. Adopt-a-Beach's twenty years of volunteer-collected records, a longitudinal dataset most fields would envy, are shared on request rather than posted online, and EPA's GLENDA asks for a free account.[34] ⬤ That is no failing of the data's stewards: sustaining twenty years of community collection is itself an achievement, and the infrastructure of open sharing costs money and staff time that mission budgets rarely spare. It is a fact about plumbing, and as Section 4 returns to, plumbing is fundable.
That is the space this project hopes to serve. The rest of this report walks through what the assembled map shows: first the good news (Section 3), then the open water (Section 4).
A report that opened with everything that is broken would be making a mistake that organizers and clergy learned to avoid a long time ago: people cannot act on a problem until they believe action is possible, and a field told only what is failing learns helplessness. So before this report maps what is missing, it owes an honest accounting of what is working, and the Great Lakes plastic response has more working parts than any one participant, deep in the demands of their own corner of it, usually gets the chance to see. This is not morale-boosting. It is evidence. Every category of intervention in which Section 4 will find gaps, whether policy, mobilization, coordination, funding, or data-sharing, already has at least one proven success on these lakes. The region is not waiting to learn whether these things can work here; it has already proven that they do.
The clearest proof is already law. In 2015, Congress passed the Microbead-Free Waters Act, banning plastic microbeads from rinse-off cosmetics; follow-up sampling measured reductions of up to 86 percent in polyethylene microbead discharge.[36] ⬤ A single, well-targeted federal rule pulled one plastic stream out of the water, and it began, years earlier, as a state fight that Illinois led before Washington followed.[20] ⬤ Policy on plastics is not theoretical here. It has a track record with a number attached.
The states are still adding to that record. New York remains the pace-setter: its bag ban has held statewide since 2020, and its foam ban keeps widening, most recently in the January 2026 expansion.[21] ⬤ Illinois, the original microbead pace-setter, continues to tighten state single-use plastic use.[20] ⬤ Together the two function as the region's proof-of-concept lab: whatever the other six states are told is politically impossible, New York or Illinois has often already done and survived.
Twenty-five years before this report was written, the Alliance for the Great Lakes began sending volunteers to the beach with data sheets. Adopt-a-Beach is now the longest-running community-science cleanup on the lakes, and its scale is its argument: in 2023 alone, 11,342 volunteers logged 28,312 hours and removed 573,608 pieces of litter, and the program's running total since 2003 exceeds ten million pieces.[9] ⬤ No agency mandated this. It is ordinary people, organized well, doing a hard thing repeatedly for a quarter century, which is proof that this region can sustain volunteer attention at a scale most environmental issues never reach. And there is a second, quieter piece of good news folded inside the first: for the very inputs science is currently missing (Section 4), that same network already solves two of the hardest parts of any new monitoring effort: recruitment and site access. New protocols, training, laboratory analysis, and quality control would all be real additional work; the people and the beaches are the parts already in place.
Coordination is the thing the rest of this report finds most missing, which makes the places it already exists worth studying closely. The Great Lakes Plastic Cleanup, a binational partnership of Pollution Probe and the Council of the Great Lakes Region, has removed more than 244,000 pieces of litter using cleanup technology, meaning trash-capturing devices set in harbors and marinas, across a network of more than 140 partner organizations spanning both countries.[28] ⬤ Set the tonnage aside for a moment and look only at the structure: 140-plus organizations, two nations, one shared effort, actually coordinated. It is a working demonstration that the fragmentation this report documents is not a law of nature. When a coordinating body and a clear shared task both exist, the actors in this field cooperate readily, and at binational scale. The barrier is not willingness. It is the absence of the connective infrastructure that lets willingness find itself.
Money, the thing everyone in the nonprofit world is trained to assume is drying up, is on this issue moving. The Great Lakes Restoration Initiative was funded at $369 million for FY2026, sustained across a change in political weather that many expected to end it.[18] ⬤ And beyond the baseline, new and specific money has arrived for exactly the frontiers this report flags as thin: $7.3 million stood up the Lake Ontario Center for Microplastics and Human Health, the region's first sustained study of what this does to people;[29] ⬤ and $1.19 million from the Great Lakes Protection Fund is building Wayne State's open-source tool for identifying microplastic sources.[31] ⬤ The pattern matters more than any single grant. Funders are not merely holding the floor; they are betting on the specific unknowns, human health and source attribution, where the science is youngest. Capital is available for this work, including for the parts of it that no one has done yet.
Some of the best news is that a substantial share of the intellectual groundwork has already been done, and is sitting on a shelf, published. In November 2024, the International Joint Commission's Science Advisory Board published its Final Report on Microplastics, and it amounts to a ready-made regulatory roadmap: standard operating procedures for monitoring, a risk-assessment framework, and a formal recommendation to designate microplastics a "Chemical of Mutual Concern" under the binational Water Quality Agreement.[17] ⬤ This is good news carrying the asterisk that becomes Section 4's business: the roadmap is drafted but not yet adopted. Even so, the drafting is the expensive part. Governments deciding to act on microplastics would not be starting from zero; the framework their own binational body commissioned is written, public, and waiting. Implementation would still take real work, because adoption, funding, and coordination are never free. But the starting line is far closer than it was.
Finally, the region has already disproven the fear that data-sharing requires every organization to abandon its own tools. Great Lakes DataStream lets community and government groups publish water-quality data in a shared, open, standardized form, on the order of five million data points from more than 135 monitoring groups, without forcing any of them off the systems they already use.[38] ◐ The Cleveland Water Alliance has built comparable connective infrastructure around Lake Erie sensing and smart-lake technology.[39] ◐ The lesson these carry for everything this report proposes is exact: data-sharing works in this region when it meets organizations where they already are, instead of asking them to convert. That is the same principle, integrate rather than replace, on which a synthesis like this one must be built if it is going to last (a commitment this project makes explicit in its design; see Section 6).
Read together, these are not six anecdotes. They are six existence proofs. Policy has cut a plastic stream by as much as 86 percent. Volunteers have sustained a cleanup for twenty-five years. A hundred and forty organizations have coordinated across a border. Money is flowing toward the newest questions. The regulatory framework is already written. Data moves without forcing anyone to convert. Every mechanism this report will need in order to argue that its gaps are closable has already worked, at least once, on these lakes. What follows in Section 4 is therefore not a catalogue of failures but a map of unclaimed ground, and it should be read in the light of this one: in the Great Lakes, this work, when it is genuinely attempted, tends to succeed.
A synthesis that only celebrated would be a brochure. This section maps what is not happening, organized in three tiers, because different readers hold different levers. A working note on tone: every gap below is framed as it genuinely is, which is unclaimed territory. In a fragmented field, a documented gap is one of the most valuable things a report can hand you: it is the work you can take up without duplicating anyone, and in several cases, the chance to define a field by being first.
Thirteen substantive holes exist in the region's plastic science.[35] ◐ (per-item confidence varies; flagged individually) The largest, in likely order of consequence:
Tire-wear particles. Globally, tire abrasion is suspected to be among the largest single sources of microplastic, and there is almost no targeted Great Lakes research on it.[35] ◐ Millions of dollars and thousands of volunteer hours go to visible beach litter while a potentially larger, invisible input goes unmeasured. No regional lab has yet made it a central question; an early, well-designed study would shape how the whole field approaches it.
Methodology standardization. Different labs count different particle sizes with different methods; the literature is, in the field's own words, apples to oranges.[17] ⬤ The IJC's 2024 report proposed standard operating procedures which, like its designation recommendation, await adoption.[17] ⬤ This is the unlock gap: nearly every other knowledge gap is harder to close while this one stands.
Human health effects. Forty million people drink this water, and the human health science is only beginning; the Lake Ontario Center, funded in 2024, is the first sustained regional effort.[29] ⬤ Its climate-plastic interaction work also stands alone; no equivalent exists for the other four lakes.[35] ◐
The unmeasured inputs. Atmospheric deposition of microplastics over the lakes: unstudied. Nanoplastics: almost no Great Lakes-specific research. Bottom-sediment distribution: described in the literature as "essentially unknown." Microfiber source proportions: unknown. Degradation behavior of "biodegradable" plastics in cold freshwater: unknown.[35] ◐
Economic impact. No serious regional quantification exists of what plastic pollution costs Great Lakes fisheries, tourism, water treatment, or public health.[35] ◐ For advocates who must make budget arguments to legislators, this absence is itself expensive: the case for action is currently made in particles per liter, when decision-makers budget in dollars.
No Great Lakes state has extended producer responsibility for plastics.[20] ⬤ EPR, which makes producers financially responsible for their packaging's end of life, is the policy centerpiece the region's own advocates converge on (Section 2.4), and the vacuum is total: zero of eight states.
Preemption blocks the local level in four states.[20] ⬤ Michigan, Wisconsin, Minnesota, and Indiana legally prevent their municipalities from restricting plastics, including every lakeside town watching litter arrive on the current from jurisdictions it cannot influence.
The federal file is pending, not dead. The Microbead-Free Waters Act of 2015 remains the proof that federal plastic policy can work: follow-up sampling measured up to an 86% reduction in polyethylene microbead discharge.[36] ⬤ Its successors wait: the Plastic Pellet Free Waters Act, reintroduced in the 119th Congress with a Michigan co-sponsor, would direct EPA to prohibit pellet discharge under existing Clean Water Act authority.[37] ⬤ The Break Free From Plastic Pollution Act has been reintroduced repeatedly without passage.[20] ⬤
The designation sits unclaimed. The IJC's science advisors have formally recommended microplastics be designated a Chemical of Mutual Concern under the binational Agreement, the step that would trigger coordinated binational management.[17] ⬤ The recommendation is drafted, published, and waiting. Adopting it would not start from zero: the scientific groundwork was commissioned by governments themselves and is already in hand.
Nobody monitors nurdles. Pre-production pellets contaminate 42 of 66 surveyed beaches; the Rossport derailment demonstrated they persist in the system for nearly a decade; a federal bill specifically about them is pending, and there is no dedicated nurdle monitoring program on any of the five lakes.[10][11][12] ⬤ We found no public evidence of any agency or organization with a dedicated program on it. Of every gap in this report, this one has the shortest distance between open and addressable: the survey methodology exists, the advocacy vehicle exists, and the monitoring infrastructure (Adopt-a-Beach's volunteer network) already walks those beaches.
The deepest gaps are not in the science or the statutes but in the connective tissue, and they are the reason this report exists.
The field's knowledge of itself has no commons. The map of who is doing what exists, but it lives in people: in institutional memory, professional relationships, and conference hallways, not in any shared, public, durable form (Section 2.6). A methodological note in fairness to the field: when this report says a channel or map "does not exist," it means we found no public evidence of one. Insider knowledge always exceeds public artifacts, and part of this project's purpose is to put these inferences in front of the field and be corrected. But the consequences of knowledge without a commons are visible from the outside: policy alignments that function as invisible to legislators and funders (2.4); a $1.19M source-attribution tool with no purpose-built public channel to the actors who could use its findings (2.5); a field whose accumulated understanding must be re-learned relationship-by-relationship by every newcomer, and which walks out the door with every departure. Siloed knowledge makes the field smaller than it actually is, a real cost when the thing being protected is a collection of inland oceans.
The richest data is the hardest to reach. Twenty years of Adopt-a-Beach records, shared on request; GLENDA, behind a free account. Community science data, largely unintegrated into state and federal systems.[33][34] ⬤ None of this is a failing of the data's keepers; these records exist at all because organizations did the patient work of collecting and safeguarding them, usually with no budget line for the machinery of sharing. The region does not have a data scarcity problem; it has an unfunded plumbing problem, and plumbing is buildable.
Effort concentrates where visibility is, not where need is. Beach cleanups, visible and photogenic and participatory, are the region's best-organized plastic response. Tire-wear particles, atmospheric deposition, and sediment loads, all invisible and unphotogenic, go essentially unaddressed (4.1). This is not a criticism of cleanups, which built the region's largest volunteer network and its longest dataset. It is an observation about what happens in a field with no shared map: attention pools where light already falls.
Section 5 turns these observations toward application: what a director, a legislator, a funder, and a researcher can each do with the map now that it exists.
A synthesis earns its length only if different readers can put it down knowing what to do next. The research on why tools and reports fail in the nonprofit sector is unambiguous on one point: one view built for everyone serves no one.[40] ◐ So this section is written four times, once for each of the readers this report most hopes to reach. Find yours, and skip the others without guilt.
The map's first gift to a director is relief from a specific kind of loneliness: the sense that your organization carries its issue alone. On the evidence of Section 2, you almost certainly do not. At least three organizations independently ask for extended producer responsibility; more than one asks for preemption repeal; the Chemical of Mutual Concern designation your policy staff may be building arguments for has already been formally recommended by the binational Agreement's own science advisors.[17][26][27] ⬤ Alignment that already exists is the cheapest coalition you will ever assemble. It requires a joint letter, not a campaign, and this map is offered as one public place from which to count your co-signers.
The map's second gift is the opposite one: where you are alone, that fact is now legible, and being first has value. We found no public evidence of dedicated nurdle monitoring anywhere on the five lakes (Section 4.2), despite an existing survey methodology, a pending federal bill, and a volunteer network that already walks the beaches in question. Unclaimed gaps of that shape, where the method, the vehicle, and the labor all exist but no one has connected them, are listed throughout Section 4, and they are invitations.
One practical note on why a public map matters to your operations specifically. Sector research finds that nonprofits discover tools and partners overwhelmingly through peer networks, and that 62 percent of organizations say they lack the time to properly vet anything new.[40] ◐ A durable public map converts partner-finding and duplication-checking from a conference-season activity into a staff-meeting one. That is the whole intent: value inside fifteen minutes, because fifteen minutes is what the sector actually has.[40] ◐
Nothing in this subsection requires commissioning new science. The actions below are ordered by distance from done, shortest first.
Adopt the designation. The IJC's Science Advisory Board has formally recommended designating microplastics a Chemical of Mutual Concern under the Great Lakes Water Quality Agreement, and published the monitoring procedures and risk framework to go with it.[17] ⬤ The framework your own binational body commissioned is drafted, public, and waiting. Implementation would still be real work, but the preparatory science is already on the table.
Pass the pellet bill. The Plastic Pellet Free Waters Act, reintroduced in the 119th Congress with a Michigan co-sponsor, directs EPA to prohibit pellet discharge under authority the Clean Water Act already grants.[37] ⬤ Section 1 documented what pellets do here: contamination on 42 of 66 surveyed beaches, and a single 2008 derailment still washing ashore eight years later.[10][11] ◐
Reexamine preemption. In four of eight Great Lakes states, the level of government closest to any given beach is legally barred from acting on what washes up on it.[20] ⬤ Whatever position one takes on statewide uniformity, the Michigan paradox stands as this report found it: a state simultaneously funding microplastic research and forbidding its own cities to act on the results.[20] ◐
Know that the record is on your side. The one major federal plastics law this region has produced, the Microbead-Free Waters Act, which began as an Illinois state fight, cut polyethylene microbead discharge, with follow-up sampling measuring reductions of up to 86 percent.[36] ⬤ New York's policy stack has been in force since 2020 and survived.[21] ⬤ Plastic policy on the Great Lakes is not an experiment. It has a track record with numbers attached, and an EPR vacuum, zero of eight states, waiting for whoever writes the first template.[20] ⬤
The research behind this project included a deliberate study of platforms and tools that failed, and the failure modes repeat with almost embarrassing regularity: funding that stopped after launch, complexity that outran its users, tools that forced organizations off their existing workflows, and the pull of the novel over the necessary, what the sector calls shiny object syndrome.[40] ◐ The successes, among them Global Forest Watch, the Humanitarian Data Exchange, and this region's own DataStream, share the inverse profile: they invested in adoption infrastructure and user research, not just software, and they met organizations where they already were.[38][41] ◐
Read against that pattern, Section 4 is a funding target list, and the highest-leverage items on it are conspicuously unglamorous:
And one lesson from the graveyard applied to whatever you fund next, including efforts like this one: sustained boring money beats episodic exciting money. Tools die when champions leave and maintenance was never a line item.[40] ◐ Fund the maintenance.
The thirteen knowledge gaps in Section 4.1 are a research agenda in rough order of consequence, and three are field-defining openings: tire-wear particles (potentially the largest input, essentially unstudied here), atmospheric deposition (unstudied), and bottom-sediment distribution (in the literature's own words, "essentially unknown").[35] ◐
The unlock is methodological. The IJC's standard operating procedures exist;[17] ⬤ standardization happens by use, not decree. Adopting them in your next study design, and saying so, is the single act that makes your results, and everyone else's, count double.
And the under-used instrument: this region maintains one of the longest-running community-science networks on any freshwater system anywhere: a quarter century of trained volunteers who already walk, on schedule, the same beaches that need pellet counts and microfiber sampling.[9] ⬤ At present that network is pointed almost entirely at visible litter. A research partnership that handed it validated protocols for the invisible inputs would join the region's longest dataset to its newest questions with recruitment and site access already solved. The added costs are real (protocols, training, analysis, quality control), but they are the fundable kind. The mobilization exists. It is waiting for a protocol and the resources to run it.
This report is the product of an explicit division of labor between a human researcher and an AI collaborator, and we state it plainly because demonstrating that this collaboration can produce rigorous, verifiable public work, in a sector historically last to adopt new leverage, is part of the project's purpose. The AI side (research synthesis, drafting, data structuring) performed the wide search: organizational websites, government databases, peer-reviewed literature, press records, compiled April 2026 and re-verified against live sources in June 2026. The human side (direction, domain judgment, field instinct, verification) decided what mattered, caught what the AI got wrong, and brings something no search can: years as a volunteer water monitor on these systems, with the field knowledge and the scars that come with it.
What we could not access, we name: Adopt-a-Beach's granular twenty-year dataset (request pending; public summaries used instead),[34] the GLENDA database (free account required; access route confirmed, dataset not yet incorporated),[34] the Rochman Lab's raw data (available via repository; summary findings used),[3] and tribal data, the last excluded by design and not by barrier, per the CARE principles discussed in Section 2.3.[25] Before this version was published, its central inferences were also sent to the organizations best positioned to falsify them, the needs-assessment correspondence described in the project record, and the standing offer is permanent: where the field corrects this map, the map changes.
A report about water, produced with AI, owes its readers a straight answer about AI's own water footprint. Data centers consume water, the buildout is fast, and some of it will draw, directly or through the grid, on the watersheds this report cares about. We do not dodge that, and Version 2's economic layer should treat AI infrastructure as what it is: one more industrial demand on the basin, to be measured like the others.
Here is the argument for proceeding anyway, and it is an argument about power. Actors organized around capital, including the supply chains this report counts in pellets per square meter, will adopt AI without hesitation, to extend reach and produce more, faster. The plastic will be made with AI whether or not the response to it is. Mission-driven organizations that abstain on principle pay a real price for it: they concede a lasting speed and scale advantage to the interests they answer.
And there is a second argument, quieter and maybe more important. Within this decade, people will sit in rooms deciding how the Great Lakes' water is used by and for AI infrastructure. The people in those rooms who understand the technology tactilely, who have used it, been failed by it, corrected it, and know from experience where its confidence outruns its knowledge, will govern better than people working from received opinion in either direction. You cannot regulate what you have never touched. This project is, among other things, one water person acquiring that standing, and documenting the acquisition so others can follow.
Every claim in this report carries one of three marks, and the system exists because of a specific, honest problem in AI-human work: an AI collaborator can misplace confidence, and no human collaborator can independently check everything. Our answer is to declare confidence per claim rather than let tone imply it.
⬤ Strong: peer-reviewed, government-published, or independently replicated. ◐ Moderate: a single study, a dated figure, or a number self-reported by the organization it describes. ○ Extrapolated: an inference we made ourselves, flagged as such. The rule beneath the marks: a claim that cannot be traced to its original source gets cut, not kept for being useful. That rule has already been applied: a widely repeated tap-water microplastics figure was removed from Section 1 of this report when its original study could not be traced. The report's single riskiest claim (that no organizational synthesis of this field exists) is marked as our own analysis, stated as such, and standing in front of the field for correction.[33] ◐ We consider this trust architecture a defining feature of the method, not a disclaimer, and we explain it here the way you explain something you are proud of, because we are.
There is a difference between counting an organization and knowing it, and this map contains both. The 200-plus figure is real, anchored in coalition membership rolls and platform registries.[13][38] Version 1 of the public dataset held 64 verified records: 14 profiled in depth (programs, data, current asks), 49 listed with core details from public sources, and 1 roster-level entry. Version 1.1 (July 12, 2026) added 191 roster-tier entries from the Healing Our Waters coalition's public membership list, after coalition staff pointed us to it in correspondence; the wider ecosystem of more than 200 organizations is now on the board by name. The long tail, meaning watershed councils, friends-of groups, and community monitoring organizations, enters at lighter depth by design. Every entry in the companion site declares which tier it is.
This is deliberate, and not only because of time. The deeper reason is the project's own ethic: small organizations should get to describe themselves, not be summarized from their newsletters by an outsider before a single conversation has happened. Every entry carries a standing invitation to add, correct, or deepen its own record. The map is community science applied to the field itself: the instrument improves in the hands of the people it measures.
Stated without cushioning: the research base is an April 2026 snapshot, re-verified in June, requiring a final verification pass at publication. There is no Canadian provincial deep-dive yet, a real hole in a binational synthesis. There is no economic modeling. The small-organization tier is thinner than the institutional tier, for the reasons and by the design just described. Where a figure is single-study or self-reported, it carries its ◐ grade in place rather than a borrowed ⬤. And the map's most important claims about coordination gaps are outside-view inferences awaiting the field's correction, which is not a flaw in the method; it is the method. But the reader deserves to know V1 carries it unfinished.
Version 1 of this project is a channel: an attempt to gather the field's knowledge of itself into one durable, public, legible place. That is the infrastructure whose absence lets monitoring reports die in agency file drawers and lets a field's accumulated understanding walk out the door with every departure. But a channel is not yet a discovery engine, and honesty requires saying which one this is. Nearly everything in Sections 1 through 4 was knowable, in pieces, by insiders. The genuinely new knowledge lies in joins this report has not yet performed: data layers that have rarely been laid on top of each other, in part because there has been no shared surface to lay them on. Three joins are ready and waiting:
Policy × community data. New York's expanded foam ban took effect January 1, 2026.[21] ⬤ Volunteers have counted foam on New York's Great Lakes beaches for two decades.[9] ⬤ Joining the statute's timeline to the beach counts would measure whether the ban is working, the same shape of analysis that gave the microbead act its 86 percent,[36] ⬤ performed on community-collected data. We found no published analysis of this kind.
Loading × presence. Federal tributary monitoring identifies which rivers deliver the plastic;[42] ⬤ this project's map identifies where the organizations stand. Overlaid, they answer a question neither dataset can ask alone: which inflows pour into stretches of the basin where no one is stationed. That overlay turns Section 4's "Open Water" from a list into a set of map coordinates.
Community monitoring × agency monitoring. DataStream carries 135-plus community groups' measurements;[38] ◐ GLENDA carries the agencies'.[16] ⬤ Joined, they yield a coverage analysis no single entity can currently see: where volunteers measure what agencies do not, and where nobody measures at all.
Beyond the joins, Version 2 needs what Version 1 was built to receive: tribal data under tribal authority, if and when nations choose partnership under CARE;[25] the Canadian provincial layer; the economic impact study Section 5.3 asks funders for; and the field's own corrections, gathered through the needs-assessment threads and the map's standing invitation. The architecture was designed for all of it from the first commit: every record carries its issue field, its source, and its confidence mark, so that new layers drop in as data, not as rebuilds.
The synthesis you have just read is the argument that the surface was worth building. The joins are the argument for what it becomes next.
Every numbered reference above resolves here, with citation detail sufficient to locate the source and an honest confidence grade wherever a figure is single-study or self-reported: declaring confidence per claim is the method, not a disclaimer. The twelve most load-bearing sources are already live as clickable chips across the site (see How We Know This).
[13] Landscape research compilation: 200+ org count; Healing Our Waters Coalition membership.
[14] ACS comprehensive review (free preprint: knowledge.uchicago.edu/record/5027), the science-review counterpart.
[15] GLRI Projects Database (2,000+ projects, public Excel/map).
[16] EPA GLNPO / GLENDA program history.
[17] IJC Science Advisory Board, Final Report on Microplastics (Nov 2024): SOPs, risk framework, Chemical of Mutual Concern recommendation. ijc.org PDF.
[18] GLRI FY2026 appropriation ($369M): ELPC statement on Senate passage; Senate-passed figure, final enacted amount may differ (◐).
[19] NOAA GLERL; Marine Debris Program; NCEI Marine Microplastics Database.
[20] State policy landscape table (Research - Plastic Pollution Data.md): bag/foam/preemption by state; MI 2016 preemption law; MI $2M research program; Break Free From Plastic reintroductions.
[21] NYSDEC press release, Dec 2025: foam ban expansion effective Jan 1, 2026.
[22] GLIFWC: treaty rights framing, TEK integration, co-management position.
[23] IJC program review: Indigenous governments as full partners recommendation.
[24] Haudenosaunee Environmental Research Institute / Ohneganos (Great Lakes Echo, Feb 2026): data sovereignty goal.
[25] CARE Principles for Indigenous Data Governance.
[26] Alliance for the Great Lakes: programs and policy asks; advocacy org table.
[27] Ecology Center (Ann Arbor): asks.
[28] Great Lakes Plastic Cleanup: 244K+ pieces, 140+ partners.
[29] Lake Ontario Center for Microplastics and Human Health ($7.3M, NIEHS/NSF, 2024).
[30] CIGLR, Michigan Tech GLRC, Central Michigan IGLR.
[31] Wayne State AI microplastic source library ($1.19M, Great Lakes Protection Fund) + watershed partners.
[32] Great Lakes Commission; Governors & Premiers; CGLR; Fishery Commission.
[33] Platform limitation analysis (Research - Landscape and Organizations.md): WQP, GLOS, DataStream, GLENDA, GLANSIS, GLIN, GLRI DB; the "no organizational synthesis" claim is OUR analysis, marked ◐ and stated as such.
[34] Adopt-a-Beach data access (email-request only); GLENDA access via free EPA Central Data Exchange account (cdx.epa.gov: ready-to-download files + query tool; access route confirmed July 3, 2026; account creation pending).
[35] Critical gaps list (compiled from the project's research files): per-item confidence varies; tire-wear and sediment characterizations are field-level assessments graded ◐.
[36] Microbead-Free Waters Act (2015): 86% PE discharge reduction study.
[37] H.R.7543 (119th Congress), Plastic Pellet Free Waters Act; congress.gov.
[38] Great Lakes DataStream (The Gordon Foundation): open, standardized water-quality data platform; ~5M data points, 135+ monitoring groups. Platform-reported totals as of retrieval date (◐).
[39] Cleveland Water Alliance: Lake Erie smart-lake / sensor and data infrastructure; program-reported (◐).
[40] Nonprofit technology adoption research compilation (Research - Tools and Nonprofit Adoption.md): sector survey statistics (62% lack vetting time, <3% tech budgets, 76% no data strategy), platform failure-mode analysis, champion model, "value in 15 minutes." Industry-survey-reported; ◐ throughout; per-stat originals to be chased for Appendix A.
[41] Precedent platforms: Global Forest Watch (WRI adoption infrastructure), Humanitarian Data Exchange (user-research lessons), HealthMap, Climate TRACE. Public program documentation; ◐.
[42] USGS Great Lakes tributary monitoring data (ScienceBase): tributary plastic/microplastic loading; basis for the proposed loading × presence overlay. ⬤ (public dataset; coverage varies by tributary and year).
[1] Hoffman & Hittinger, RIT: annual plastic loading estimates and particle transport modeling. rit.edu/news/researchers-study-plastic-pollution-great-lakes
[2] EPA / Great Lakes Commission: drinking water dependency figure.
[3] Rochman Lab, University of Toronto: water sample thresholds; fish sampling. Data via Borealis repository.
[4] Shedd Aquarium tributary study.
[5] Tap water particle figure: REMOVED from this report; the widely repeated statistic could not be traced to its original study. Entry retained to document the cut.
[6] FACETS journal: temporal sediment cores, 1964–1989 accumulation trend.
[7] Lake Michigan/Erie benthic study (65.2 vs 431 particles/kg).
[8] Lake Erie surface concentration study (downstream of Detroit/Cleveland/Erie).
[9] Alliance for the Great Lakes, Adopt-a-Beach 20-year analysis and 2023 annual summary.
[10] Rossport derailment records and 2016 shoreline reports.
[11] Great Lakes beach pellet survey (42 of 66 beaches).
[12] Absence-of-program claim: verified against IJC 2024 Microplastics Report and agency program lists, June 2026.