Frequently cited estimates for birds killed by cats in the U.S. actually exceed the number of birds estimated to be in the country. Documents obtained via Freedom of Information Act raise as many questions as they answer.
As I pointed out recently, the annual mortality estimates proposed in the 2013 paper, “The impact of free-ranging domestic cats on wildlife of the United States,” don’t add up. Or, to be more precise, they do add up—and up and up. Indeed, the authors’ “conservative” estimate of birds killed by outdoor cats appear to exceed the total number of land birds estimated to be in the country.
According to the Partners In Flight Population Estimates Database—which, given its intended use for “bird-conservation planning,” would seem to be the go-to source for the best estimates available—that total is 3.2 billion. That’s only 33 percent greater than the median estimate (2.4 billion) developed by Scott Loss, Tom Will, and Peter Marra—leaving very little room for the many other sources of mortality,  including the 365–988 million birds they’ve estimated are killed annually as a result of building collisions. 
And the high-end of their “conservative” estimate of annual cat-caused mortalities (4.0 billion) actually exceeds the PIF estimate by a significant margin—raising serious questions about the validity of the work portrayed by Marra, in Cat Wars, as the culmination of a century’s worth of evidence implicating cats in the decline of birds and other wildlife. 
As documents obtained via Freedom of Information Act reveal, though, this isn’t the weirdest part of the story.
In Search of 10–20 Billion Birds
To be clear, nobody should expect the PIF estimates to be accurate with 100 birds or so. As the project’s website notes, there are “many assumptions and uncertainties involved with estimating population size.” And the range alone—1.3 to 4.0 billion—says a lot about the accuracy of the estimates proposed by Marra and his co-authors.
But again, this isn’t really the most interesting part of the story. What’s curious is how the authors actually put their estimates into context by referring to the PIF estimates.
“For all North American land birds, the group of species most susceptible to mainland cat predation, existing estimates range from 10–20 billion individuals in North America.” 
But the estimate from the Guide to the Partners in Flight Population Estimates Database, the source cited by Marra and his co-authors, is considerably less than 10–20 billion: “approximately five billion breeding birds” (emphasis mine). 
Granted, the Guide’s authors qualify their estimate, based on Breeding Bird Survey data, as “likely a conservative total… as densities from Breeding Bird Censuses suggest the total could be two to three times higher in some regions.”  But their comparison of BBS and BBC data focused on just two of North America’s 66 Bird Conservation Regions, hardly sufficient to justify a broad-based adjustment upward. Moreover, they showed that BBS data compare quite favorably with Breeding Bird Atlas data “in the Ontario portion of BCR 13,”  suggesting that perhaps no adjustments are in order after all.
In any case, no mention of 10–20 billion birds—at least not in the source cited by Marra and his co-authors in the published version of the paper. But an earlier version of the paper, submitted to Science, provides some important clues:
“The estimated impact of bird mortality from cat predation represents 5–30 percent of all U.S. birds based on population estimates of between 10 billion to 20 billion (Banks 1979, USFWS 2002).”
Sure enough, the U.S. Fish and Wildlife Service estimated, in its two-page brochure Migratory Bird Mortality, published in 2002, “that a minimum of 10 billion birds breed in North America. Fall populations may be on the order of 20 billion.”  But that’s it—no citation and no additional information.
To find out where that 10–20 billion figure comes from, one needs to dig a little deeper, turning to the other source provided: Richard Banks’ Human Related Mortality of Birds in the United States (which happens to be a report for USFWS):
“A report by the American Ornithologists’ Union (1975) presents figures leading to estimates of a total breeding season bird population in the United States (exclusive of Alaska and Hawaii) of nearly 10 billion, which is probably doubled to 20 billion with the addition of young in the autumn. If the total breeding population remains stable from year to year, it follows that some 10 billion birds must succumb to one or another mortality factor each year.” 
At last, in the 1975 AOU report, the origins of the mysterious 10–20 billion figure become clear:
“In 1973 the Breeding Bird Survey from the U.S. Department of the Interior, Fish and Wildlife Service yielded an average of about 3,325 birds per square mile or 9,975 million for the U.S. exclusive of Alaska and Hawaii. With the addition of young each year the autumn population is probably about double that figure, or around 20 billion.” 
That was 1973. Forty years later, Marra and his co-authors were using this same BBS data to claim “cat predation represents 5–30 percent of all U.S. birds.”
Had they adjusted this important bit of context to reflect the more recent BBS data, and more sophisticated analysis, of the PIF estimates—which, it’s worth remembering is the source they actually cite—the authors would have found themselves claiming (based on estimates they insist are conservative) that outdoor cats kill 41–125 percent of this country’s birds, year in and year out.
Which, I think it’s safe to say, would have raised some eyebrows.
• • •
How does something like this happen? It’s not entirely clear, but again, the documents obtained through FOIA request provide some clues.
There’s a note, for example, on the version of “The impact of free-ranging domestic cats” that was submitted to Science, from co-author (and longtime TNR opponent) Tom Will, referring to the 10–20 billion figure and its original two references (i.e., Banks 1979 and USFWS 2002):
“We could leave it as is… although no mention of how this accounts for breeders + young of year + migrants to Canada. OR we could ask Pete Blancher to help us come up with an extimate [sic] based on PIF calculations.”
Precisely what happened next remains a mystery. But in subsequent versions of the paper, the 2004 PIF Guide, for which Blancher* was lead author, replaced Banks’ Human Related Mortality of Birds in the United States and USFWS’s Migratory Bird Mortality as the source for the claim that North America is home to 10–20 billion land birds.
The paper was, of course, eventually published—and attracted loads of media attention. And it’s now been given additional attention (and presumed credibility) via Cat Wars. And all the while, the authors were given a pass on this critical point (and a host of other flaws in their methodology).
And, just like that, several billion birds gone.
Note: I e-mailed Peter Marra, asking him to explain the considerable discrepancy between his estimates and those from the PIF database, but never received a response.
* Blancher’s name will be familiar to regular readers—he’s produced his own badly flawed estimates regarding predation by Canada’s outdoor cats.
1. Loss, S.R., T. Will, and P.P. Marra, The impact of free-ranging domestic cats on wildlife of the United States. Nature Communications, 2013. 4 http://www.nature.com/ncomms/journal/v4/n1/full/ncomms2380.html
2. Loss, S.R., et al., Bird-building collisions in the United States: Estimates of annual mortality and species vulnerability. The Condor, 2014. 116(1): p. 8–23. http://www.aoucospubs.org/doi/abs/10.1650/CONDOR-13-090.1
3. Marra, P.P. and C. Santella, Cat Wars: The Devastating Consequences of a Cuddly Killer. 2016: Princeton University Press.
4. Blancher, P.J., et al., Guide to the Partners in Flight Population Estimates Database. Version: North American Landbird Conservation Plan 2004, in Partners in Flight Technical Series No 5.2007.
5. USFWS, Migratory Bird Mortality, 2002, U.S. Fish and Wildlife Service: Arlington, VA.
6. Banks, R.C., Special Scientific Report—Wildlife No. 215, 1979, United States Department of the Interior, Fish and Wildlife Service: Washington, DC.
7. A.O.U., Report of the American Ornithologists’ Union Ad Hoc Committee on Scientific and Educational Use of Wild Birds, 1975, American Ornithologists’ Union: Supplement to The Auk, Vol. 92 No. 2 (July). p. 1A–27A.