A wetland mitigation credit clears in an office. The flood it was meant to compensate for still arrives at the same bend in the same creek, on the same afternoon, and it does not check whether a unit was debited twenty miles away.
If you searched wetland mitigation, you are probably holding a permit application, a compensation estimate, a conservation compliance determination, or an investment memo about credits. This post gives you the definition in the permittee's words, an honest read on what a credit does and does not prove, and what to hold instead if the thing you actually want is a floodplain that still works. Wetland mitigation is a real obligation settled with a real instrument. That instrument is not the wetland.
what you are actually buying
what is wetland mitigation?
Wetland mitigation is the compensation step in a regulatory sequence: avoid the impact, minimize what cannot be avoided, then compensate for what is left. Only the third step produces anything you can buy.
Two federal statutes drive almost all of it in the United States. Under Section 404 of the Clean Water Act, discharging dredged or fill material into waters of the United States requires a permit from the U.S. Army Corps of Engineers, and that permit usually carries a compensatory mitigation requirement — restoring, establishing, enhancing, or preserving aquatic resources to offset the permitted loss. On farm ground the second lane is wetland conservation compliance under the Food Security Act of 1985: a producer who converts a wetland can lose eligibility for USDA program benefits unless the conversion is mitigated.
The rule machinery behind all of this — the 2008 Compensatory Mitigation Rule, the preference order, credit release schedules, financial assurances, the RIBITS register — belongs to a habitat bank is a land deal. Read that one if you are underwriting the supply side. The policy goal the whole sequence serves, no net loss, gets its own honest treatment in no net loss is not a wet acre.
This post is about the question a buyer rarely asks out loud: where is the acre that answers for mine, and is it anywhere near the water I moved?
what is a wetland mitigation bank?
A wetland mitigation bank is a site where aquatic resources have been restored, established, enhanced, or preserved in advance of the impacts they will offset. A regulator approves the site, the plan, and the ledger. Credits release against milestones, and permittees inside an approved service area buy them to satisfy compensation obligations. The service area is usually a watershed boundary negotiated in the banking instrument — the geography inside which a regulator treats one wet acre as substitutable for another.
For a seller, the service area is a market. For a buyer, it is the entire question. It defines how far your compensation is allowed to travel from the acre you filled.
the credit moves. the water does not follow.
A wetland's flood-damage service is not only a property of the acre. Storage is on the acre; whose damage is avoided depends on where the acre sits relative to the water and to whatever is downstream.
Charles Taylor and Hannah Druckenmiller, writing in the American Economic Review in 2022, measured this directly against U.S. flood claims and land-cover data. They found flood-mitigation benefits concentrated in wetlands 500 to 750 meters from the nearest stream or river — with no detectable benefit from wetlands closer in, and none from wetlands farther out. They also found no detectable flood-mitigation benefit from wetlands in undeveloped areas, which is not a statement about ecological worth. It is a statement about arithmetic: flood damages avoided require something downstream that could have been damaged.
Now put that next to how compensation actually travels — and date it. J.B. Ruhl and James Salzman, studying Florida banks through about 2005, found mitigation landing on average more than 15 miles from the fill site, with losses concentrated in urban areas and replacement concentrated in rural ones. A later study of four Oregon counties (permits 1994–2008) found banks sited an average of 11 miles from the removal-fill site, in areas that were wealthier and less densely populated than the places the wetlands left. Work on 1,058 permitted mitigation transactions in the Chicago region between 1993 and 2004 found the same directional pull — high-density to low-density — while noting, usefully, that one mitigation method did not produce the effect. All three of those studies predate the 2008 rule's watershed approach; the Roychowdhury and Moore preprint below is the post-rule read, and it is modeled. The displacement is a tendency of the market, not a law of physics.
The economics are not mysterious. Development pressure is highest near existing development, which is exactly where land is most expensive and where a restored acre would sit closest to exposed property. Credits get supplied where land is cheap. Those two facts point in opposite directions.
A March 2026 preprint by Indumati Roychowdhury and Frances Moore at the University of California, Davis put a number on the gap across 915 U.S. wetland markets: wetlands lost to development between 1985 and 2021 delivered, on average, about 4.1 times the downstream flood protection of the wetlands created in compensation — and in some markets up to 78 times. Their own framing is careful, and worth keeping intact: wetland markets may succeed in preserving total wetland acreage while systematically failing to preserve downstream flood protection. That is a preprint, not yet peer-reviewed, and it is a modeled estimate rather than a measured flood. Read it as a direction, not a verdict.
This is not an accusation. It is a description of what the unit was built to trade: acres and assessed functions, fungible anywhere inside a watershed boundary. Flood protection is not distributed evenly inside that boundary, so a trade can be entirely valid and still move the service.
The wetland exists without the credit. Storage, filtration, and the slow release of a storm are things the ground does because of its soils, its water, and its position — and it did all of them for centuries before any register existed. No permit starts that. No sale stops it. What the credit determines is only who pays, and on which acre.
the four objects in this market
Permit files, pitch decks, and county meetings use these four as if they were one thing. Two of these are land; only one of them is at your address.
| what changes hands | what it settles | what holds the next flood | |
|---|---|---|---|
| wetland mitigation credit | An issued unit, released against milestones, usable inside a service area | A permittee's compensation obligation for a permitted impact | Whatever the bank site holds, at the bank site's position in the watershed |
| wetland mitigation bank | Land under an approved instrument, carrying construction, monitoring, and stewardship obligations that outlast the sales — underwritten in a habitat bank is a land deal | Nothing by itself; it is the supply side that issues the credits | The bank parcel, for as long as the easement and the endowment hold |
| water credit or certificate | A claim on a volume of water under an allocation regime — see a water certificate is not a wet acre | A user's entitlement to divert or to leave water instream | Nothing directly; a volume is not a landform |
| wet floodplain | Nothing. It is held by hydrology, by title, and by whoever controls the ditches and the pumps | Nothing. It has no obligation to discharge | Itself — which is the only answer that ever mattered |
Only one of those rows will be there when your creek comes up.
additionality, the service area, and the lag
Three questions decide whether a credit funded a real wet acre. They are not rhetorical, and the field has asked them of itself for twenty-five years.
Additionality asks the counterfactual: would this site have been restored anyway? On ditched and tiled ground that question is unusually tractable, because you can stand on the site and see the ditch. Preservation credits are the harder case: the rule only allows them for wetlands under demonstrated threat, at higher ratios, usually paired with restoration — because protecting existing wet ground produces a unit without a new wet acre.
The service area is the spatial version of the same question, and it is where the evidence above bites. A credit purchased at the far edge of a large service area can be entirely legitimate under the instrument and still leave the reach you filled without its storage.
The lag is the temporal version. The National Research Council's 2001 review for EPA, Compensating for Wetland Losses Under the Clean Water Act, found that restoration and creation sites seldom achieve functional equivalency with reference wetlands within the five-year monitoring window typical of permits; up to twenty years may be needed for some sites, and some components may never reach equivalency. The committee's recommendation followed: structural similarity — vegetation cover, most often — does not establish functional equivalence, and mitigation ratios should reflect functional loss rather than simply matching area.
You might be thinking this is building toward an argument that wetland mitigation is a scam. It is not, and that argument would be lazy. The 2001 review was commissioned by a regulator, and the 2008 rule was substantially the response to it: advance credit generation, the watershed approach, site protection, financial assurances, long-term management funding. Banks were placed at the top of the preference order precisely because a bank has already done the work a permittee promising future work has not. A regulated market that commissions a critical review and then rewrites its own rules around the findings is not the thing to be cynical about. Plenty of banks clear the bar that matters: ground that was drained is wet now, and it stayed wet after monitoring ended. Where that happened, the money did what it said it would do.
is a wetland credit the same as a restored wetland?
No. A wetland credit is a unit issued against a site's acres and assessed functions, released on milestones and spent when it discharges an obligation. A restored wetland is water held on a particular piece of ground long enough to grow the soils, plants, and storage that only wetness makes. Good banks deliver both, and the difference only shows up in two places: the credit travels the full width of the service area while the wet ground stays at one address, and the wet ground keeps sending a bill in every year after the last credit has been sold.
the cousin on the farm side
If you are a producer rather than a developer, the instrument reaching you is probably not a Section 404 bank.
On 24 July 2026 USDA announced $3 million through the Natural Resources Conservation Service Wetland Mitigation Banking Program to develop banks serving agricultural producers with wetlands subject to conservation compliance — grants of $100,000 to $1 million, no match required, applications closed 8 September 2026. Priority went to states with the most individual wetlands, wetland acres, and compliance requests: Georgia, Illinois, Indiana, Iowa, Michigan, Minnesota, Nebraska, Ohio, Pennsylvania, South Dakota, and Wisconsin. Award money pays for the unglamorous parts — site identification, the banking instrument, functional assessments, surveys, title work, design, and construction.
That is public money standing up supply so a farmer with an unavoidable impact has somewhere to buy. It is a cousin instrument, and a sensible one. A producer who inherited a century of tile drainage did not design the drainage district and is not the villain in this story. The point stands anyway: the credit that producer buys settles a compliance obligation, and the wet acre that used to sit in their section is still not in their section.
keep the credit. hold the place.
Nothing here argues for buying fewer credits. It argues for knowing which of your two problems each dollar solves. The permit is one problem. The floodplain is the other, and it has a street address.
Four questions separate them.
- Where is the compensating acre? Ask for the bank's position in the watershed relative to your impact, and relative to whatever is downstream of both. Distance to the stream network and distance to exposed property are not trivia — on the evidence above, they are most of the flood service.
- How many wet acres are you funding, by name? Credits retired is a compliance number. Named acres held wet this year is an operating number. Only the second one still means something after a functional assessment method is revised or the definition of a jurisdictional water moves again.
- Who pays the year after the last credit sells? A bank's stewardship obligation outlives its revenue curve. Endowments get sized once; costs do not stay fixed.
- If a specific reach is what you care about, who is funding that reach? Often nobody is, because no permit obliges anyone to.
Question four is where a certificate of specific ensurance fits. A certificate is issued 1:1 against a single agent — an onchain account standing for a named place, group, or purpose — and proceeds route to that account, funding present condition on named ground. It satisfies no Section 404 obligation and no conservation compliance determination, enters no mitigation ledger, and is not sold as an offset. That limitation is the feature: because it never claims a gain, it can fund the years after the endowment, or the reach no permit obliges anyone to fund, without anyone selling the same acre twice.
Stating our own position: the live doors are inland-wetlands.ensurance for the stock and marble-wetlands-preserve.avlt for a named place, certificates sit on /specific, and a handful of wetland-named coins trade in small volumes. There is no deep secondary market to exit into, and no third-party verifier stands behind our condition measurement yet — we have to earn that. We do put a number on condition, because a bill that nobody can read never gets paid. The number is a translation for capital, not a statement about what the floodplain is.
The floodplain is the asset. The credit is a receipt for a permit. The certificate is a standing payor for a place. Three different objects — and only one of them gets wet.
see specific ensurance certificates →