Cats, Catbirds: Why the Smithsonian study doesn’t say what everyone thinks it does

There’s been a huge amount of publicity recently about a 2004 study, publicized by the highly-respected Smithsonian. It implies that cats are the main cause of death of songbirds, and has become a talking point for those advocating stronger measures against feral and outdoor cats, even to killing cats. The SF Weekly published an issue with a demonic cat on the cover; even the normally thoughtful New York Times took the press-release at face value. Views From the Thicket, a great blog about Golden Gate Park, repeated the SF Weekly’s story.

But did the study’s results actually merit that kind of spin? We think not.

First, a disclaimer. We’re not advocating here for any particular feline policy. But we do respect all animals and birds, and even more, we favor a thoughtful approach to data. So we went back to the actual study. (You can read it here as a PDF.)


Working in three Maryland suburbs filled with homes and gardens, the  researchers tagged a total of 69 baby catbirds with tiny radio-transmitters. They then tracked the birds alternate days until they found a dead bird, the detached transmitter, or lost the signal (when the birds presumably left the area i.e moved more than 5 km from the study sites).

At the end of the study, which was conducted between May and September 2004, they found that 42 of the birds died from various causes: predation, disease, glass panes, etc. Of that 42, they figured 33 were killed by some kind of predator. (They found the predator, or the bird’s remains with the transmitter attached, or they found the transmitter underground). In 19 of the cases, the researcher felt they could make a good guess at what kind of predator it was. Only 6 were killed by cats.

Another 3 were presumed to be killed by cats because of the state of the catbird corpse (missing a head). They thought this was a sure indicator of cat predation, but that’s been disputed. [ETA: A little websearching reveals birds of prey – hawks, owls, falcons, kestrels – seem to bite off the heads of their catch first. (There are slightly gruesome photographs.) If they are disturbed by people or competing birds, this could account for headless dead birds.]

We’ve shown this breakdown in the graphic above, where each circle’s area is proportionate to the number of birds it represents.

Unfortunately, the abstract of the study spun the same data differently:

Misleading phrasing

Is this true? Yes… but does it suggest a misleading conclusion? Also yes. It seems to suggest that around 80% of deaths were caused by predators, and nearly half of those by cats. That is not true. They had only 19 “known” predation events — meaning the ones where the researchers had a guess at Who Dunnit — compared to a total of 33 killed birds. [ETA2:  And 42 dead birds. Taking the guess-work out, here’s the story in graphic form.]


So let’s do a thought experiment: If Maryland had no cats, how many extra catbirds would have survived?

If we looked  only at the other 60 catbirds, (the ones that weren’t actually or putatively killed by cats) we find 33 died from other causes. They had a less than 50% chance of survival. So even assuming cats killed all the birds the scientists thought they had, only about 4 more birds would have survived.

If the birds killed by cats were in some way more vulnerable than the average — if they were louder, or weaker, or less wary, or already ailing — then they’d probably have been picked off by one of the many other predators. After all, that’s what predators do: They go after the weakest ones of a flock or herd. In that case, having zero cats in Maryland might have made no difference at all to the birds’ survival. Something else would have got them anyway.

So the effect of the cats seems to have been somewhere between 0 and 4 birds. Hardly the stuff of headlines. [ETA3: In any case, with such small sample sizes, statistical analysis is meaningless.]


What’s probably much worse for bird populations is habitat reduction. Responsible home-owners and dedicated park managers prune or remove trees, eliminate unsightly brush, trim bushes and plants, use pesticides to eliminate weeds and insects. All of those things destroys part of a bird’s world, maybe all of its world: cover; seeds; insects.

Here in San Francisco, and definitely in Mount Sutro Forest, habitat destruction is an issue. Thickets are being destroyed, because they’re “non-native” and “invasive” (the pictures here show an example the same place before and after this process).

Never mind that they’re the places where birds and animals can live and hide.

This entry was posted in Environment, Mount Sutro Stewards, Mt Sutro Cloud Forest and tagged , , , , , , . Bookmark the permalink.

15 Responses to Cats, Catbirds: Why the Smithsonian study doesn’t say what everyone thinks it does

  1. Pingback: Cats, Lies, and the Smithsonian Catbird Study « FOREST KNOLLS

  2. Dan says:

    Three problems with this blog post:

    – You didn’t link to the Smithsonian study, so I can’t so easily go and read the study to see for myself what it says. Bad form.

    – The article only implied that cats are the main cause of death of small birds? (It didn’t say so explicitly?) Are you sure then that you’re not just reading into it: afterall, even you conceded here that cats do in fact kill some birds.

    – I’d be astounded if the Smithsonian or any other respected conservation biologist said explicitly that habitat destruction isn’t the most significant threat to wildlife, including birds. I mean, sure, they’d discuss other serious threats – and there are other threats that are serious – but to say that habitat destruction is relatively unimportant is absurd. So I need to see the actual 2004 article itself to believe that one.

    • webmaster says:

      Hi Dan,

      We did indeed intend to link to the study, and somehow the link got dropped earlier. (A draft version of the article accidently got posted. We’ve corrected that, so you can check the study.)

      Incidentally, the study was made in 2004, but was published only recently. What the study said is shown in the yellow box. As far as we could tell, it didn’t go into the issue of habitat destruction because of the study design. For instance: If homeowners trimmed bushes or trees and removed cover for the birds, making them more vulnerable to predators, the study wouldn’t have a mechanism to evaluate that.

      • Dan says:

        Thanks for the link.

        And yeah, I see what you’re saying now about the study design. And yes, that’s a valid point – thanks for discussing it.

        *objection withdrawn*

  3. milliontrees says:

    Thank you for digging into this study. We also have analyzed a “study” by the University of Nebraska that reported $17 Billion of “damage” caused by cats killing birds to justify their conclusion that cats should be killed: As SaveSutro finds in this post, we also find that the Nebraska “study” does not stand up to scrutiny.

    We also invite you to visit a closely related post: In this post we report that efforts to exterminate animals on behalf of specific species of birds have not been effective and have had unintended negative consequences.

    Although we don’t advocate for a particular method of managing feral cat populations, we recommend that those who are sincerely concerned about this issue read the comments on the SFWeekly article on the subject. We learned a great deal from those comments. Advocates for trap-neuter-release (TNR) make a good case for the effectiveness of their effort and we were impressed by their dedication. Comments from those who advocate for killing cats are in sharp contrast.

  4. Charlie says:

    Good points. When my friend mentioned this study I had a similar point. The biggest killer of birds in suburbs is SUBURBS! After the suburbs are built, most habitat is destroyed and with it most birds are killed or driven away. It’s true that many suburban tolerant birds are killed by cats but since it’s an urban ecosystem with a huge mix of different species, whether this is a bad thing or not depends simply on what you are trying to manage for. Outdoor cats are detrimental to birds but they aren’t detrimental to all wild animals. For instance, they are great news for one of the top predators of the hidden suburban ecosystem – coyotes!

    I do think that in areas adjacent to habitat and open space there are really good reasons not to have an outdoor cat. It can affect the populations of less-tolerant bird species… and can affect your cat in a highly detrimental way by turning it into coyote food. Coyotes now occupy all habitats in the continental US, with very few exceptions, including suburban areas.

    Trap-Neuter-Release seems like the best way to manage feral cats. Killing them is obviously not a pleasant choice, but letting them overpopulate and then starve to death isn’t much better.

    • Dan says:

      “Trap-Neuter-Release seems like the best way to manage feral cats.”

      Maybe. I’m not sure. But do you see anyone interested in doing this on the scale required? I don’t.

      • webmaster says:

        Dan, on TNR — I don’t know. I think here in San Francisco there are groups working on it. Whether the same thing happens in the suburbs or in rural areas I don’t know.

        By the way: Your Cyprus bird blog is amazing.
        (For other readers: Clicking on Dan’s name in his comments will take you to his bird blog about the Mediterranean island of Cyprus.)

        • Dan says:

          Well, I really hope that whatever it is, some sort of action takes place*. Although as you note in your post above, we still must maintain full attention to habitat preservation first and foremost.

          *Although I’ll freely admit that I’m basing this attitude on just being fed up with the number of cats over here. In the center of Nicosia there are an estimated 1000 cats/sq km, and while it may be less in the outskirts of the city where I live, I’m sure that it’s still at least a couple hundred, if not more. It’s a plague, and the native Cypriots still won’t stop feeding the strays, to say nothing of reducing their numbers. :-/ …. thus my interest in this discussion.

          Many thanks for the compliment. 🙂

  5. webmaster says:


    I understand your concern: Cats are a *visible* predator of birds. You have to be lucky to see a hawk stoop on a smaller bird, it’s often in the air and gone in seconds. Owls, many rodents, and other predators are nocturnal so much more difficult to observe. But this doesn’t mean that cats have a major impact on bird populations.

    Even if this study did say what it purports to do, there’s at least two important considerations (besides habitat):
    1. The study area was specific to a particular bird and particular neighborhoods, and the sample size was small. Do its findings have broader application? Probably not. And especially not in a different country, and an island setting.
    2. It doesn’t look at second-order effects of feral-cat presence. Cats are mainly predators on rodents. By keeping rodent numbers down, are they reducing competition for plant-food such as seeds, also eaten by birds? And, (because these are residential areas) by keeping rodents down, are they also reducing the probability that home-owners will resort to poisons to control rodent populations, thereby also poisoning other predators such as hawks and owls?

    • Dan says:

      Great points. I particularly like point #1, and I’d like to see a study like that done.

      And that’s true to a point about cats primarily predating on rodents, except that cats are opportunists. I’ve seen cats being quite successful against birds on specific occasions (anecdotally speaking): fledgling birds and exhausted migrants seem to be easy targets. You could argue that the exhausted birds are the weaker ones, but I’m not sure that one could argue similarly on the fledglings.

      Also the reduced competition issue would be very limited indeed to only certain bird species. Generally speaking, the only seed-eating ground birds here are House Sparrows – all of the finches and buntings here generally stick to the trees. Again anecdotally, the birds I see getting caught by cats here are thrushes and warblers: insect-eating birds.

      But every area is different, and these are just my anecdotal observations in gardens in residential neighborhoods here…

      • webmaster says:

        It’s true every area is different, and so is every species.

        In terms of attacking nestlings or fledglings: Though the young birds themselves are vulnerable, perhaps parent birds that build nests that are difficult to reach pass on their genes more effectively? Or fledglings that are cautious or quiet are less likely to be caught?

        I think the way second-order effects work in an eco-system is pretty complex. I just read, for instance, on the Cornell website that “Gray Catbirds sometimes destroy eggs and nestlings of woodland species including Eastern Wood-Pewee, Chipping Sparrow, and Song Sparrow.”

        So if cat predation actually reduced their numbers, maybe it would benefit other species…

        To look at the actual impact of cats on bird species in a particular ecosystem, you need a control. Maybe exclude cats from part of a specific habitat area and see if there’s a change in the composition and density of bird life, compared with the rest of it.

        • Dan says:

          Yeah, I agree the nestlings are inaccessible, but the fledglings just seem clumsy and inexperienced, generally speaking.

          And valid point on Catbird predation of other woodland species. I actually heard a similar study a few years ago, but with Bluejays that were reducing the number of successful nests among woodland bird species but only where the woodlands had been fragmented by housing developments (surprise, surprise).

  6. webmaster says:

    Yes… true. Scrub jays are predators on hummingbirds (birds and eggs) in San Francisco.

  7. Pingback: Cat Haters Out in Force Ignoring Damage Birds Do & Feral Dog Problems | The Pink Flamingo

Comments are closed.