Eighty Percent Sterilized and Still Growing: The Anatomy of a Colony That Comes Back

Eighty percent is the number everyone repeats. Get a colony to eighty percent sterilized and it stops growing. Get there and hold, and it shrinks. The number is not made up, and the studies behind it are better than most rescue folklore. But the number is a rate, and it is quoted as a milestone, and that single category error is the mechanism behind the most common failure in community cat work: the colony you finished three years ago, back to its old count, with a new generation of intact cats in it.

Here is the mechanism.

Eighty Percent Is a Rate, Not a Finish Line

Start with the steelman, because the threshold is well grounded.

Miller and colleagues modeled free-roaming cat populations over fifty years in a stochastic individual-based simulation. Their finding: under trap-neuter-return, roughly 40% of the remaining untreated cats have to be sterilized every six months to produce consistent population decline, and "over a ten year period, sterilizing 40% of the naïve animals during each timestep resulted in a long-term cumulative sterilization rate of 75%."¹ Boone and colleagues, using the same modeling family, defined "high-intensity TNR" as sterilizing 75% of the intact cats present in each six-month step, which "generates a sterilization rate of over 80% throughout nearly all of the simulation trajectory, which reduces initial population size by about half over time."²

So: 80% standing sterilization does correspond to real population decline. That part of the folklore survives contact with the literature.

What does not survive is the tense. In both models the percentage is applied to the intact fraction, every six months, forever. The denominator refills. A colony that is 80% sterilized in June is a colony where four cats in twenty are intact, and those four are the ones producing next spring's denominator. Hitting eighty is not a state you achieve. It is a level you hold by continuing to trap at roughly the same tempo that got you there.

The failure begins the day the group stops counting the colony as active work.

The Vacuum Effect Does Not Care Why the Cats Left

This is the part experienced caregivers will want to argue with, so let me put it plainly.

The vacuum effect is the standard TNR rebuttal to catch-and-kill: remove a colony and unsterilized cats from the surrounding area move into the vacated territory, and within a couple of seasons you are back where you started.³ The claim is correct. It is also a claim about habitat, not about method. Territory does not know whether it emptied because a truck came or because a sterilized colony aged out. A niche opened by attrition is as open as a niche opened by removal.

Boone's model measured exactly this. At the ten-year mark, the proportion of living cats in the managed population that had been born outside it was above 90% under high-intensity TNR, against 30% under no action at all.² In raw counts, the mean number of living adults born outside the focal population was 23.86 under high-intensity TNR and 22.30 under low-intensity TNR, versus 14.52 under no management.² The most successful sterilization scenario produced the most immigrant-derived population in the simulation. The authors' own conclusion: "reducing abandonment and, where possible, immigration in conjunction with high-intensity TNR could improve outcomes more than for any other management option tested."²

Rescue people deploy the vacuum effect as a one-way weapon against removal. It is not one-way. It is the reason your finished colony refills, and it is the strongest available argument that TNR has to be geographically wide rather than locally thorough.

Your Own Success Makes the Remnant More Productive

The one long-term controlled field experiment on TNR ran twelve years across a 20 km² urban area, split into pre-intervention, mixed-intervention and full-intervention phases. Gunther and colleagues found that the cat population increased through the first two phases and "did not decline in highly neutered populations, presumably due to cat immigration." Only when high-intensity neutering was expanded to the entire city, above 70% neutering, did population growth reverse — reaching an annual reduction of about 7%. And even then: "This population reduction was limited by a rebound increase in cat reproduction and longevity."⁴

Read that last sentence carefully. Success compensates against itself: fewer breeding cats means less competition for food and territory, so the intact queens who remain raise more kittens successfully, and the sterilized cats you returned live longer. You are running against a system that partially undoes your work in proportion to how well the work went.

Worth noting what the models do not show here. Both Miller and Boone explicitly excluded differential longevity of sterilized cats from their simulations, Miller "because of the difficulty in interpreting the data in terms of actual changes in survival rate."¹² Gunther measured the compensation in the field; the models mostly did not include it. If anything, the simulation thresholds are optimistic.

The Arithmetic of the Four Cats You Missed

Take a forty-cat colony, half of them female, sterilized to 80%. That leaves eight intact cats, of which roughly four are queens.

Free-roaming queens produce a mean of 1.4 litters per year with a median of three kittens per litter, and 75% of those kittens die or disappear before six months of age.⁵ So:

• 4 queens × 1.4 litters × 3 kittens = 16.8 kittens born per year
• 16.8 × 25% surviving to six months = ~4.2 new adults per year

Four cats a year into a colony where the thirty-two sterilized residents are dying slowly, if at all. That alone is roughly replacement. Now add immigration: Boone's model used average immigration of 2% of the surrounding source population per six-month step, plus abandonment averaging one litter with about three surviving kittens per step.² Even at a fraction of those rates, the intact share of your forty cats is not 20% two springs later. It is 30%, then 40%, and the kittens are the ones you cannot catch because nobody has put a trap out since the year you finished.

That arithmetic is mine, applied to published parameters: an illustration of the mechanism, not a published result. Real colonies vary enormously in food supply, mortality and turnover. But the direction is not in doubt, and the order of magnitude is the point: a "finished" colony left alone regenerates a breeding core in about three years.

The Colony Is the Wrong Unit of Management

Here is the second thing worth disagreeing with, and it changes how a group should spend its surgery slots.

Most TNR groups work colony-by-colony, triaged by urgency: the worst situation, the loudest caller, the most kittens. Gunther's mixed-intervention phase is the direct test of that approach — neutering intensity high in some neighborhoods, low in others, across one continuous city — and the population went up.⁴ Contiguity, not local completeness, was what turned the curve.

The implication is unpleasant. At a fixed number of surgery slots per year, a group probably gets more population reduction by saturating one contiguous area and holding it than by taking the worst colony in each of six neighborhoods. Which means the correct answer to a caller three neighborhoods over is often "not this year, and possibly not next year either" — while you finish the area you are in. That is a hard thing to say to someone standing in their yard counting kittens, and I do not think most groups, including good ones, are willing to say it. But the alternative is a map of scattered sterilized colonies, each refilling from the untouched blocks between them, which is what the twelve-year data actually describes.

Where This Argument Is Weakest

Three honest objections.

One city, one study. Gunther is a single 12-year experiment in a dense urban environment with heavy public feeding. Contiguity should matter less where a colony is genuinely close to demographically closed — an island, a fenced industrial site, a rural property miles from the next one. If your colony is truly isolated, colony-by-colony work is defensible on its own terms.

The campus counterexample, read properly. The University of Central Florida program is the standard success story: the population fell from 68 cats to 23 over eleven years, with no kittens observed on site after 1995.⁶ A follow-up covering 1991 to 2019 reports an 85% decline from the 1996 census and 11 of 16 colonies eliminated.⁷ That is real. But the original paper also records that 47% of the cats were removed for adoption, against 15% remaining on site.⁶ A large share of that decline came from taking cats out, not from returning them. A campus is also access-controlled, geographically bounded, and staffed by people who live there. Most colonies have none of those properties, and citing UCF as evidence that returning cats shrinks populations elsewhere is reading past what the study says.

None of this is an argument for doing less. Boone's model is unambiguous on welfare: over ten years, an initial population of 50 cats produced a mean of about 1,000 preventable deaths under no action, 290 under low-intensity TNR, and 32 under high-intensity TNR.² Their finding, verbatim: "Lack of management (i.e., the no-action scenario) results in more cumulative preventable deaths, particularly of kittens, than any active management option. This includes lethal removal."² Scattered, low-intensity TNR still prevents an enormous amount of suffering. It cuts preventable deaths by roughly 70% while barely moving the population.

Which is the honest way to state the whole thing: most colony-by-colony TNR is a welfare intervention being sold as a population intervention. Both are worth doing. They are not the same claim, they need different amounts of money, and confusing them is how a group ends up defending a program against a city council using a population argument the program was never designed to win.

What Changes If You Take This Seriously

Three things, none of which are new procedures.

Treat a colony's sterilization percentage as a reading that decays, not a badge. If the last trapping night at a site was more than a year ago, the number you have is historical.

Draw a boundary on a map before you accept the next site, and prefer the colony that extends your treated area over the colony that starts a new island — even when the new island is worse.

And when you describe the program to a funder, a council or a skeptical neighbor, say which claim you are making. If you are running scattered TNR on a few hundred surgeries a year across a city, promise fewer kittens dying, because that is what the evidence supports and it is a serious thing to promise. Promise a smaller cat population only where you are actually saturating an area and intend to keep doing it.