A rainbow collar slipped over the one 94 hunting cats already wore sent the birds they brought home down by nearly two thirds, and the small mammals kept arriving at the same rate

The bird is on the doormat when you open the door.

Cat owners know the ritual. The animal is fed, well fed, and it goes out and hunts anyway, and it carries the results back to show you.

Then the same cat starts wearing a tube of loud striped fabric over its collar.

The doormat still gets used.

But what lands on it changes, and it only changes for one half of what the cat catches.

What a bird sees that a mouse does not

Birds are extraordinary at color. Most carry four types of color receptor where a mammal makes do with two or three, and a great many of them see into the ultraviolet on top of that.

So a saturated band of stripes moving through long grass does not read to a bird as an object. It reads as a flare.

A mouse sees almost none of it.

Which is the whole reason the rainbow collar works, and the reason it works on exactly one half of the problem. A Swiss trial of 31 cats found the colored cover cut the birds brought home by around 37 percent with no measurable effect on mammals at all. The same trial then added a bell, and the mammals dropped by 54 to 62 percent while the bird figure barely moved.

The number, and where it comes from

The headline figure comes from three separate trials pooled together, two in New York and one in Florida, run across bird breeding seasons.

Researchers slid the cover over the ordinary cat collar of 94 pet cats that were already known hunters. Each animal alternated two week blocks with the rainbow collar and without it, and everything it carried home was counted and identified.

Then the two piles were compared.

With the cover on, the cats brought back 2.7 times fewer birds, a cut of 61 percent. It is the largest pooled sample anyone has run on this device, and the effect held up across all three sites rather than resting on one good year in one good yard.

The size of the thing it is up against

Here is the number that gives the whole business its weight. A systematic review put annual bird mortality from free ranging cats in the contiguous United States at between 1.3 and 4.0 billion, and mammals between 6.3 and 22.3 billion.

The low end alone is nearly four times the human population of the country.

Those are estimates with wide bands rather than a census, and the authors say so plainly. Even taken at the bottom of the range, the review concluded that free ranging cats are likely the single greatest source of human caused mortality for American birds and mammals.

And most of it is not your cat. Roughly 69 percent of the bird toll and 89 percent of the mammal toll was attributed to unowned animals, the feral and barn and stray cats that get no bowl at anybody’s back door.

Where the effect gets weaker

This is the part that usually goes missing. The pooled work found the cover did better in New York than in Florida, 69 percent against 47, temperate against subtropical, and that gap was statistically real.

Season swings it harder still. In the original New York work the spring figure was startling, roughly nineteen times fewer birds recovered, while the autumn figure from the same cats was about three and a half times.

Both are true, and quoting only the spring number is how a plastic lawn gets sold as grass. The combined study is the one to quote.

What did not vary was the cat itself. Age, sex and the kind of ground it hunted made no measurable difference, so this is not a device that happens to suit one sort of animal.

What it does not fix

A collar only helps a cat that wears one. The unowned animals doing most of the damage wear nothing, and no piece of fabric reaches them, which is why this sits alongside the harder work rather than replacing it.

Islands are the sharpest version of that. A global review found feral cats responsible for at least 14 percent of recorded bird, mammal and reptile extinctions worldwide, and the principal threat to almost 8 percent of critically endangered species in those groups.

Those losses happened where the prey had no evolved defense, in the same way a displaced fox reshapes a place it was never part of. The national estimate is the reason anyone bothers with a rainbow collar at all.

The authors of the pooled study wrote that the collar cover could be one tool within a larger effort to decrease domestic cat predation of songbirds, and that sentence is doing real work. One tool, not a solution.

Still, for the specific cat in your specific yard, the arithmetic is hard to argue with. Two thirds fewer birds for a piece of fabric is an unusually good trade, and the cat does not appear to mind the outfit.