On May 9, 2016, Mexico's national protected areas commission closed Playa del Amor.
The beach sits inside Parque Nacional Islas Marietas, an hour off the resort coast of Bahía de Banderas. Daily peaks near 3,000 visitors sat well above the existing 625-a-day management-plan cap. The coral underneath could not carry that. So CONANP shut it, ran a restoration program, and reopened it on August 31 under a hard cap: about 116 visitors per day, groups of 15, thirty minutes each.
Every hotel on the bay was still standing the whole time. Flights still landed in Puerto Vallarta. The thing people had flown in to photograph was simply off the shelf.
That is what happens when nobody protects tourism in advance: the product gets pulled and the distribution keeps running.
what "protect tourism" usually means
Say protect tourism in a tourism board meeting and everyone knows the agenda. Air service retention. Crisis communications after a bad headline. Shoulder-season campaigns when the numbers soften. Source-market diversification. Defending the lodging tax that pays for all of it. Fighting a short-term rental ordinance. Occasionally, a resilience plan that is mostly a communications plan.
That is real work, and the people doing it are usually the only ones in the building tracking the destination as a whole. But look at what it protects: demand, reputation, access, jobs, tax base. Every line item sits on the distribution side of the business.
None of it protects the thing demand is for.
what does it mean to protect tourism?
To protect tourism is to keep the living product a destination sells — snowpack, reef, whale season, lagoon, smoke-free airshed — in a condition that can still be sold next season. Marketing protects demand. A fee protects capacity. Insurance pays after a loss. Only funding the living system protects the inventory itself.
You do not protect tourism by protecting demand. You protect tourism by protecting the thing demand is for.
Tourism does not sell rooms. It sells a living place — and a living place can fail. That distinction is the spine of this series, and the pillar post works it through product family by product family: the destination is a living product.
This post takes the next question: what actually happens when the living product hits its limit, and what protects it.
closure is the honest zero
Every destination that has run into a limit has run the same sequence. Load rises. The living thing degrades. A public agency rations access.
Rationing is inventory management, and the agencies that do it are usually right and usually late. But it is worth being precise about what a cap buys, because "we manage carrying capacity" gets offered in a lot of rooms as if it were the same thing as protection.
| place | the living product | the cap | what the cap does | what it does not do |
|---|---|---|---|---|
| Playa del Amor, Islas Marietas (CONANP, Mexico) | coral, seabird island, beach | closed May 2016; reopened at about 116 visitors/day, groups of 15, 30 minutes | stops the bleeding; makes the visit survivable for the reef | fund coral restoration or water quality at bay scale — a headcount is not a budget |
| Hanifaru Bay, Baa Atoll (Maldives) | manta and whale shark feeding aggregation | marine protected area since 2009, managed since 2012: snorkel only, five vessels, 45-minute tokens, capped swimmers (80 originally, 45 under current guidance), certified guides, on-site rangers | rations pressure at the exact moment the animals are feeding; token fees route to the Baa Atoll Conservation Fund | change the water temperature that bleaches the coral the mantas depend on, or fund condition across the rest of the atoll |
| Venice historic centre (Comune di Venezia) | intact lagoon, walkable centre, acqua-alta-resilient access | 2026 access fee: 60 days, €5 or €10, day-trippers, historic centre only, still experimental | smooths peak-day crowding and gives the city a count it never had | fund the barene — the salt marshes that keep the lagoon a lagoon. MOSE protects floors. Neither protects the product |
| Protected-area entry fees generally (CONANP and peers) | parks, reefs, reserves | per-visitor fee | pays a floor of ranger time, moorings, enforcement | restoration and monitoring at the scale the surrounding destination economy operates |
A cap is a claim on volume. It decides how much damage gets done. It does not fund the recovery, and it cannot manufacture the thing being rationed.
That is not a criticism of the agencies. Hanifaru's token fee is the closest thing on that table to a hold, and it exists because a rationing regime was designed with a conservation fund attached. It is a small, good example of the pattern this post is arguing for — and it is the exception.
the snow proof: the year the cap was the weather
Colorado's 2025–26 ski season had no visitation cap at all. The mountain did the rationing.
Skier visits came in at about 10.5 million, down roughly 24% from the 13.8 to 13.9 million of recent seasons, and the lowest since 1991–92. Vail Resorts' Rocky Mountain segment was down about 25%. Lift revenue held up better than traffic, because passes were pre-sold — the weather risk had already been transferred to the guest. That is good treasury management. It is not a snowpack policy, and mountain towns felt the difference on the lag when winter tax collections came in.
Here is the honest limit: no certificate makes snow. Nothing in this protocol changes a storm track, and anyone who tells a resort CFO otherwise is selling something.
What funding a headwaters basin can do is hold everything the snow lands on and runs through — forest condition, soils, melt timing, stream temperature, the water supply the town and the snowmaking system both draw from. The hydrology of a bad snow year is its own subject, and we have already written it: when there's no snow, want snow? invest in the water cycle, and protect the top of the watershed.
The point for a destination is narrower. A season that misses on condition is an inventory event. It shows up in the P&L as weather, in the DMO budget as a marketing problem, and nowhere at all as a line item that funds the inventory.
the people who live there are not a branding problem
Overtourism is not a branding problem, and residents are not an obstacle to solve around.
The first costs of an overloaded destination land on the people who live in it: housing priced by nightly rates, water and waste systems sized for a population that triples in season, traffic, and the slow loss of access to your own beach, trail, or square. Venice's historic centre has lost more than half its residents since the 1950s while day-trip volume climbed. In a resort bay, the housekeepers, guides, and boat captains who make the product work usually live upstream of it, in the same watershed, with less water security than the properties they serve.
A cap that preserves the guest experience while a town cannot house its own staff has protected the wrong inventory. And a destination that degrades the living system degrades the place people live first and the place people visit second — same system, different exposure.
Two design consequences follow, and they are not optional:
The payor is whoever collects on the product. Flags, operators, DMOs, ports, airports, developers — the parties whose ADR, tax capacity, and CapEx already assume the living thing shows up. Not residents buying back access to their own place, and not a guilt line on a folio at checkout.
The hold has to be legible to the people who live there. Governance is not ours to supply; it already exists on the ground, in agencies, councils, ejidos, and communities that have stewarded these places far longer than any protocol. Ensurance is funding architecture, not a governance protocol. We fund; local people govern. If the condition reporting is not something a resident can look at, it is not protection — it is another report for the industry that commissioned it.
so what actually protects tourism
Four different instruments get described in destination meetings as "protecting the destination." They are claims on four different things.
| the spend | it is a claim on | when it moves | what it leaves behind |
|---|---|---|---|
| destination marketing, campaigns | attention | before the season | arrivals |
| access fee, carrying-capacity cap | volume | at the gate | a queue and a floor of park operations |
| insurance, parametric payout | a loss event that already happened | after | cash to respond, fast |
| ensurance certificate or line on a named place | condition | now, continuously | a funded living system and a path to permanence |
All four can coexist. A destination needs arrivals, needs rationing, and should carry insurance. The gap is the fourth row, and it is the only one that is a claim on the thing being sold. Earmarking a park fee to CONANP or the Baa Atoll Conservation Fund is necessary operations. A certificate is a different claim: it prices condition continuously, sits on a named account the destination can audit, and can end when the place is permanently protected. A fee-to-agency line funds the ranger. It does not, by itself, hold the living system outside the gate.
In plain terms: specific ensurance is a certificate tied one-to-one to an agent — an account representing a named place, like a bay, a basin, or a headwaters — so the money routes to that place and its stewards rather than to a general fund. General ensurance coins fund protection broadly, the way a thematic fund does. That is the whole crypto content of this post; the rest is a funding decision. See certificates on named places.
If the instinct is to insure the reef instead, read a payout is not a reef first — parametric payouts are genuinely good at speed and structurally incapable of holding condition. And how bed taxes, access fees, and climate levies could split selling the place from keeping it is the next post in this series.
the honest stage
Some of this is live and some of it is scaffolding, and it matters that we say which.
bahia-de-banderas.basin and ameca.basin are minted agents with accounts. So are colorado-headwaters.syndicate, eagle-river.basin, and roaring-fork-river.basin. A Venice lagoon agent does not exist yet. Two relevant coins are live and very thin: $manta and $snowpack together carry a few thousand dollars of market cap and single-digit holders.
That is the accurate picture. The accounts, the routing, and the condition accounting work. The tranches are small because destinations have not funded them yet. We would rather say that than let anyone read "protocol" as "already solved."
what to do with this
1. Name the SKU. One sentence, and it is a real test: if the living thing disappeared and the building still stood, would anyone book? If the answer is no, the room was never the product.
2. Size the hold to the book, not to the grant. Size the number against the revenue that assumes the mountain, the bay, or the lagoon. A $25,000 watershed grant next to a multi-billion nature-tied book is a different order of magnitude — not a different kind of care.
3. Put one tranche on one named place. Not a pledge to sustainable tourism in general. One basin, one bay, one headwaters, with condition reporting the CFO and the county commissioner can both read.
If your dependency is generic rather than destination-specific — water, pollination, soil — the how-to for a first check lives here: how to start paying for the nature you depend on. If it is a place you sell by name, start with the destination:
what this looks like for a hospitality business →