A black bear opened the door of a Toyota in Gold Hill, Colorado, in the summer of 2026, got stuck inside, and spent the small hours leaning on the horn until someone let it out. On the other side of the world, a tusker called Chakkakompan broke into a ration shop at Anayirankal, in Kerala's Idukki district, and went for the sacks of rice. It was the twelfth time elephants had hit that shop.
Neither animal was behaving strangely. The bear was running a calculation in a year when the inputs changed. The elephant had learned, over three years and two tuskers, that the shop was part of the range. Same arithmetic, different clock — and the second one is what the first becomes if the pantry stays empty.
what climate change effects on animals look like at ground level
Ask what climate change does to animals and you usually get a range map — species belts sliding north and uphill across decades. That is real, and it is slow, and it is not what shows up in an incident report.
What a town meets is faster and much more specific. A thin winter. A late frost. A berry crop that failed. A waterhole that went dry in May instead of August. The animal that lived nearby last year still lives nearby. Its food does not.
When food or water fails in the wild, animals follow calories and taps. That is the whole mechanism. The dumpster, the maize field, the swimming pool, the ration shop — those are just where the nearest reliable calories happened to be.
the same raid, eight places
| where | who showed up | what failed | where they went | what the response costs |
|---|---|---|---|---|
| Colorado, New Mexico, Utah | black bears | worst snowpack on record across the Upper Colorado River Basin; late frost; chokecherry, serviceberry and Gambel oak crops thin | alleys, restaurant coolers, garages, parked cars — and onto the plains east of Colorado Springs, where no oaks grow | Colorado Parks and Wildlife logged over 8,600 bear reports through 31 August 2026 — more than all of 2025 — with 98 relocations and 294 bears killed (preliminary). New Mexico is at 187 killed, the most since 2011; Utah has passed three dozen, more than the previous three years combined |
| California | mountain lions, coyotes, bobcats, black bears | annual rainfall, and the forage, water and prey that track it | livestock pens, yards, trash, Tahoe kitchens | Seven years of state incident reports show conflict reporting rising 2.1% to 3.0% per 25 mm (about an inch) of rainfall lost, depending on species |
| Tucson and Phoenix, Arizona | javelina | long-term drought; surface water | pools, irrigation hoses, birdseed, pet food, moist soil | Arizona Game and Fish hauls water to more than 3,000 wildlife waters, year-round in the hottest years. Relocating javelina does not work — the animals die or return, and new ones move in |
| Churchill, Manitoba | polar bears | sea ice forming later, extending the fasting season — then, in April 2024, the building that stored the town's garbage burned down | the emergency landfill; when an ice storm cut power to the electric fence, more than thirty bears went through it | A holding facility locals call the polar bear jail, helicopter relocations, and a town of roughly 900 residents rebuilding a waste system sized for roughly 25,000 annual visitors |
| southern Zambia | elephants | drought dried water sources inside and around Mosi-oa-Tunya National Park; migration routes disrupted | maize fields, at night | Crop losses, injuries and deaths. Farmers report no compensation and no subsidy for protective measures |
| Munnar, Idukki, Kerala | elephants | rainfall far below normal through the first half of 2026; grassland gone unpalatable | tea estates, worker housing, sheds, ration shops | One tusker caused over ₹1.5 lakh of damage to a single estate shop. At Anayirankal, a ration shop has been raided twelve times across three years and two tuskers; the owner paid ₹45,000 to repair one of them and says compensation has not arrived |
| Cape Peninsula, South Africa | chacma baboons | habitat squeezed against a hard urban edge, with household waste as the reliable calorie | wheelie bins, kitchens, parked cars | 6,000 lockable baboon-proof bins, the first 2,200 delivered from mid-September 2026, plus a boundary fence from Zwaanswyk to Constantia Nek intended to secure habitat for roughly 250 baboons |
| Whyalla and the Mid North, South Australia | kangaroos, emus | record-low rainfall across the state's agricultural districts | town verges, farm pastures, stock water | Landscape board staff fielding reports of malnourished animals, farmers buying feed against grazing pressure, and public warnings not to put out food or water |
Eight places, one shape. Something that used to produce food or water stopped producing it, and an animal that could not wait walked toward the version of it that people maintain.
Not every failure is a drought, and the table should not be read as if it were. Churchill's problem is a fasting season stretched by late ice, made acute by an accident. Cape Town's baboons live against a boundary that was hard before this decade. The Anayirankal ration shop is a habituation story that started years before this dry season; the rainfall failure made it worse, not new. South Australia's kangaroo pressure is partly a population question, and the state's own aerial survey speaks to that better than a news cycle does. What generalizes is not the weather. It is the arithmetic: a wild pantry stops covering the animal's daily requirement, and a human pantry is closer.
the price of the same raid is not the same
In Aspen the receipt is a citation, an insurance claim, a hazing mat behind a barbecue restaurant, and — often enough — a dead bear. That is a real cost, and the people paying it did not ask for it.
In Idukki the receipt is a ration shop that now stocks less than it should because the owner is afraid, and repairs he paid for himself. In southern Zambia it is a season's maize eaten in a night, and sometimes a person who does not come home. In Zambia there is no compensation scheme, no adjuster, and no HOA. In Kerala there is a scheme on paper, and the shop owner is still waiting on it.
That difference matters more than the similarity. Flatten it and you get a genre of writing where a Colorado dumpster and a Zambian maize field are the same anecdote, and the family with the least capacity to absorb the loss quietly becomes a metaphor for someone else's argument. The mechanism generalizes. The consequence does not. Anywhere the loss falls on subsistence, coexistence is not a lifestyle choice — it is an unfunded liability assigned to the household with the least ability to carry it.
what the numbers can carry, and what they can't
The most useful measurement of this so far is not a bear count. It is a rainfall regression.
Researchers at UCLA and UC Davis took seven years of California Department of Fish and Wildlife incident reports (2017 to 2023) and tested them against the prior twelve months of precipitation. For every 25 mm — roughly an inch — of annual rainfall lost, reported conflicts rose about 2.11% for mountain lions, 2.21% for coyotes, 2.56% for American black bears, and 2.97% for bobcats. Carnivores as a group responded most strongly. Herbivores and omnivores did not — black bears being the notable exception.
A dry year is measurable in a wildlife agency's inbox before it is visible in a population survey.
The study's lead author is careful about what the data can prove, and so should anyone citing it be. Reports capture what people call in. It is not fully separable whether a dry year produces more encounters or makes people perceive and report the same encounters as worse. Both of those are still information about a landscape under stress — but they are different claims, and a serious reader will notice which one you made.
The Colorado figures deserve the same caution. Through 31 August 2026, they are preliminary, they come from a reporting system that also got busier, and they are a snapshot of a bad year, not a trend line. They should be read as a description of a food year, not a scoreboard.
two ends of the same shelf
We have written two of these before, from opposite directions.
polar bears and melting ice is what happens when the platform that produces the food disappears. A polar bear does not lose seals; it loses the ice it needs to hunt them from.
elk can't winter on a golf course is what happens when the food is still there and the ground is not. Elk winter range sits in low valleys — exactly where people build.
Melted, developed, burned, or simply too dry: the pantry fails several ways. The animal's response to all of them looks like the same walk toward town.
who ends up paying, and for what
| who | what they're already funding | what nobody is funding |
|---|---|---|
| towns and wildlife agencies | alley patrols, bins, hazing equipment, relocation, euthanasia, staff time around the clock | the forage and water that would have kept the animal out of the alley |
| land stewards and managers | fences, gleaning, monitoring, conflict response | the condition of the stands and grasslands they manage, in the years that condition fails |
| foundations | mitigation kits — bins, collars, solar fencing, rapid-response teams | the upstream food year, which rarely fits a grant line because nothing bills for it |
| tribes and rural communities nearest the animals | night watches, lost crops, damaged structures, the social cost of living with a species other people admire from a distance | any structure that pays them for carrying that cost |
Every item in the left column is competent work, and none of it is a mistake. It is also, almost entirely, downstream. The one clear upstream line anywhere in this post is Arizona hauling water into the backcountry — an agency deliberately keeping the wild option viable so animals have somewhere else to be.
the alley is managed. the pantry is not.
A berry crop has no maintenance budget. A shola-grassland mosaic has no operator. Winter range has an owner, but nobody pays that owner for the browse it produces. When these systems fail, the failure shows up as somebody else's line item — a claim, a citation, a helicopter, a dead animal — and never as a deferred-maintenance charge against the asset that actually failed.
That gap is what ensurance is built to close. The condition of a place is treated as the asset, and funding is routed to it before the failure, not to cleanup afterward. Two instruments do the routing: a coin funds protection broadly across the protocol, and a certificate funds one named place directly. A certificate's proceeds land in the account of that named place; a coin's spread across the protocol's places. Neither is a general fund.
We should say plainly where this stands. The accounts are live, the instruments trade, and the volumes are small. This is early market infrastructure, not a finished funding channel for six countries. The argument is that the pantry belongs on someone's balance sheet, and right now it is on nobody's.
Pricing the condition of land is a bridge to get capital moving toward it. It is not a statement about what the animal is worth. A bear is not the price of a totaled car, and an elephant is not a maize invoice.
where to go next
explore what's being funded, and where → — natural assets, places, and the accounts behind them.
If you want the mechanism at full length in one place, start with the pillar: you don't have a bear problem. you have a food-year problem.
For the two species stories this post leans on: polar bears and melting ice and elk can't winter on a golf course.