Ninety-four cats spent a breeding season prowling New York and Florida wearing what can only be described as a clown ruff. Picture a fat tube of bright cotton, red and blue and yellow, slipped over a breakaway collar so the fabric stands up in a fan around the head. Their owners bagged, labelled and froze every dead bird, mouse and lizard that turned up on the doorstep, then handed the whole grim collection over to be identified by species.

Results came out lopsided. Writing in the Journal of Fish and Wildlife Management, Margaret Jensen, Susan Willson and Abby Powell found that cats brought home 61% fewer birds while wearing the cover. Small mammal kills fell 37% and reptile and amphibian kills 47%, but neither of those results was statistically significant, so both could plausibly be noise.

Designed to catch a sparrow’s eye, not a mouse’s

The idea behind the collar is straightforward: songbirds carry four types of colour receptor, sharpened further by oil droplets on their cone cells, so bright reds and yellows should stand out to them roughly the way a fluorescent safety vest stands out to a person. Mice and voles carry none of that receptor equipment and get by mostly on smell, hearing and touch, so by the same reasoning, a cat in a party collar still looks to them like an ordinary cat.

Nothing else about the ambush changes: the fabric does not weigh a cat down or block its paws as a bib might. All it is meant to do is make the cat harder to miss, for whichever prey can actually see it coming in colour.

Latitude changed the answer

New York cats managed a 69% drop in birds killed; Florida cats, only 47%.

The researchers point to breeding hormones as a possible reason. Temperate songbirds carry higher testosterone levels during a compressed nesting season, which may leave them distracted and easier to catch, giving a rainbow collar more room to help. That remains a hypothesis rather than a demonstrated mechanism, though it does fit the geography of earlier trials.

The prey nobody counted

How much does a doorstep tally actually represent? Every figure above describes what cats carried home, which is a fraction of what they kill. A study using collar-mounted KittyCams on 55 suburban cats in Georgia, led by Kerrie Anne Loyd at the University of Georgia, found roughly a quarter of captured prey made it back to the house. The rest was eaten or abandoned where it fell.

So the 61% figure describes a sample, not a headcount. The collar also does nothing for nestlings and fledglings, which cannot escape a cat they have just been warned about. Sublethal costs go unmeasured too, since a bird that spends its morning fleeing has less time to feed its chicks.

The failure mode is the owner

One participant told the researchers they would consider using the cover if their cat killed more birds, which is its own small study in human nature.

Practical problems piled up faster than biological ones. Owners withdrew 24 cats partway through, mostly because they had stopped responding to the researchers or grown uncomfortable with the collar, and at least eight of those cats had also reacted badly to it themselves. Separately, 44 cats lost their collars at least once over four weeks, and 20 owners modified the collar to stop it happening, mostly by disabling the breakaway safety catch that was there to keep the cat from being strangled. Fewer than half the participating owners said they intended to keep using the thing, and replacements run about US$20 a set.

Australia and England got there first

Earlier trials had already tested the same idea, an ocean or two away. A two-year Western Australian study of 114 pet cats, published in Applied Animal Behaviour Science by Catherine Hall and colleagues, recorded roughly halved bird captures. A much smaller English trial of 19 cats by Pemberton and Graeme Ruxton, which the Journal of Zoology carried in 2020, put the figure at 78%.

Different hemispheres, different bird communities, same lopsided outcome: colour registers with birds, mammals shrug. That consistency is more persuasive than any single number, and it is also the reason the individual percentages should be held loosely. The samples behind these studies remain modest: 19 cats in England, 114 in Australia, 94 across the two American trials.

Other options tested in the same trial

A University of Exeter team led by Martina Cecchetti tested the cover alongside several rival ideas and published the comparison in Current Biology. Switching cats to a grain-free, high-meat food cut prey brought home by 36%. Five to ten minutes of daily play with a toy trimmed it by 25%. Bells did nothing measurable. Puzzle feeders made things 33% worse.

Those two winners work on a different part of the problem: the cat’s appetite for hunting, rather than its ability to sneak. That distinction matters for anyone weighing options: a cat that hunts less to begin with needs no equipment, and there is nothing to lose under the neighbour’s hedge.

Scale keeps the whole exercise humble. Free-ranging cats in the United States kill somewhere between 1.3 and 4.0 billion birds a year, by the Smithsonian estimate published in Nature Communications, and most of that is done by unowned cats that nobody is dressing in cotton. A ruff on a house cat is a local fix for a local garden. The real test comes later, in whether the collar survives on the cat past week three.