Cats Vs. Birds: Researching The Research
In this reblogged post from his Vox Felina blog, feral cat advocate Peter J. Wolf demonstrates how to evaluate the validity of research with his in-depth critique of a study on bird predation. Policy decisions to promote the well-being of all animals should be based on sound research, not regurgitated assumptions. Animal advocates should be on the lookout for poorly documented assertions about animals in the press.
Reblogged with the kind permission of feral cat advocate Peter J. Wolf, from his blog, Vox Felina.
In “Estimated Number of Birds Killed by House Cats (Felis catus) in Canada,” published late last year in Avian Conservation and Ecology, Environment Canada research scientist Peter Blancher estimates that cats—owned and unowned—“kill between 100 and 350 million birds per year in Canada,” and suggests that this level of predation “is probably the largest human-related source of bird mortality in Canada.” 
I submitted the following comments to the journal in response to Blancher’s article, but retracted my submission upon learning that (1) the length is nearly twice as long as what is permitted, and (2) that I would be required to pay an “author fee” of about $340.
I’m not naïve enough to think that posting my comments here [Vox Felina] is comparable to having them published in ACE, but, given the considerable work involved—and, more important, the obvious policy implications of Blancher’s paper—I think it’s important that they be available.
• • •
The two factors most affecting the author’s predation estimates (i.e. the estimated number of feral cats, and the adjustments made for undetected prey) lack the sort of scientific rigor one expects from work with such obvious policy implications.
Canada’s Feral Cats
“Although it is known that there is a large and growing population of feral cats in Canada,” argues Blancher,  citing the Canadian Federation of Humane Societies website, “there are apparently no scientific estimates of their numbers.” In fact, a recent visit to the CFHS site revealed nothing to suggest that the population is growing, only that “feral cats are a growing concern in cities across Canada.” 
Lacking “scientific estimates,” Blancher turns to 15 “media reports” spanning seven cities and the province of Quebec. Of the 13 I’ve been able to track down, just one appears to be even remotely credible, relying not on “cat rescue experts”  or unattributed claims made by reporters, but on “an Ipsos-Reid poll conducted in Halifax in September 2008, which determined the number of cats per household, there are 109,362 cats in Halifax.” 
Although details of the poll results appear to be unavailable online, the Halifax example is useful to illustrate a larger point. In the story cited by Blancher, it’s suggested, based apparently on “research” conducted by Police Superintendent Robin McNeil, that “the housed cat population matches the feral and stray numbers cat-for-cat.”  In fact, it’s likely that McNeil’s “cat-for-cat” estimate, like so many others, can be traced to a frequently cited study from Alachua County, Florida, in which “the [unowned] free-roaming cat population was estimated to represent approximately 44 percent of the population of cats in the county,”  or, more generally, “suspected to rival that of pet cats.” 
But a careful review of the methods involved in the Alachua County research suggests that the resulting estimate may overstate the number of unowned cats in a particular community. In any case, and more to the point of the current discussion, the observed ratio of pet cats to unowned cats was, at best, a snapshot in time and space. The survey was conducted in 1999, and in a part of North America quite different climatically from Halifax. In Alachua County, located at 29.7º latitude, January temperatures average 12.7 C [55° F]; temperatures in Halifax, 20 degrees to the north, average -5.2 C [23° F] during the month.
Whatever the origins of the Ipsos-Reid estimate, climate differences alone suggest that Blancher’s estimated 295 feral cats for every 1,000 human inhabitants of Halifax—nearly twice the density (168 feral cats/1,000 humans) estimated for Alachua County—is quite implausible.
(As an aside: a 2009 Halifax Regional Municipality report  cites two different estimates of the city’s “owned domesticated cat population”: 39,248 (based on “the Calgary model”) and 93,600 (from “the Consultant report of 2005”), and, referring apparently to the Alachua County study, applies the same range to the unowned population because “it has been suggested that the stray/feral cat population is equivalent to the owned domestic cat population.” Unfortunately, I’ve been unable to obtain any additional details regarding either of the figures mentioned in the report.)
Although Blancher  ignores his Halifax estimate (the highest by far in his Table 2) when calculating “the number of feral cats per 1,000 persons in southern Canada,” his result of “between 50 and 150” remains questionable—again, in light of the Alachua County estimate of 168 unowned cats per 1,000 persons. And this one parameter had the greatest influence on Blancher’s estimate for birds killed annually by Canada’s feral cats, explaining approximately 31 percent of the total variance (as illustrated in his Figure 3).
While it’s not difficult to appreciate Blancher’s challenge in estimating the number of stray and feral cats in southern Canada, the importance of this figure to his claims that “predation by house cats is probably the largest human-related source of bird mortality in Canada” and “the majority [are] likely to be killed by feral cats”  warrants greater rigor.
Adjustment for Undetected Prey
No doubt Blancher  is correct in suggesting that “the number of birds returned and collected by owners is not a complete record of birds killed by their cats.” And he’s to be commended for acknowledging that “some animals brought home and assumed to have been killed by pet cats may have been killed by other means,” an obvious point, but one only rarely mentioned in the literature on the hunting habits of free-roaming cats. (On the other hand, Blancher fails to consider what’s actually implied: an adjustment factor less than 1.)
Nevertheless, the various factors he uses to adjust for undetected prey fail to stand up to careful scrutiny.
Blancher  is correct that “George  used a correction factor of 2 to offset undetected or scavenged prey, based on a comparison of prey returns when monitored continuously versus an average rate of monitoring,” but in doing so he was accounting for his own behavior, not the hunting behavior of his cats. Although this adjustment factor “has been reported widely… it is unfounded.” 
And the correction factor (3.3) used by Kays and DeWan (2004) was based on the observations of 12 cats successfully capturing “small mammals” rather than birds, which were, in fact, observed to avoid capture. As such, its use by Blancher (and the many others whose work he cites) is simply inappropriate.
As Blancher  suggests, “Loyd et al. (2013) used collar-mounted video cameras to increase samples of prey captured (n = 39) and found that pet cats brought home 23 percent of prey.” However, the implied correction factor of 4.3 is, again, inappropriate. Although Loyd et al.  fail to disclose exactly which of the 39 prey items were returned home, they do point out that “14 of 16 reptiles and amphibians (88 percent) were either eaten or left at the capture site.” Of the 7 additional items returned home, it seems quite reasonable to assume that birds (which made up 5 of the 39 total prey items) were about as likely as small mammals (10 of 39) to be included, and far more likely than the 8 invertebrates documented in the study.
And finally, to the highest of the adjustment factors used, 5.8, attributed to a study in central Poland by Krauze-Gryz et al. It seems clear from a careful reading of the original work that these cats have little in common with the pet cats to which the correction factor is applied. At least some of the 34 cats monitored were, for example, “classified as barn and farm cats,” which “were seldom fed and often received only milk.”  And in all cases, “the cats’ outdoor activity was not restrained, and all of the cats hunted regularly.”
Just as the estimate for the population of unowned cats was the greatest factor influencing the overall predation estimate for stray and feral cats, the adjustment for undetected prey is the greatest contributor to the predation estimate for pet cats (accounting for approximately 35 percent of the overall variance, as illustrated in Figure 5). Here too, therefore, greater rigor is warranted.
Even if we assume Blancher’s predation estimates are correct, his claim that “there may be consequences for bird populations”  resulting from the associated 2–7 percent annual mortality rate is questionable. Reviewing dozens of predation studies, Mike Fitzgerald and Dennis Turner concluded: “there are few, if any studies apart from island ones that actually demonstrate that cats have reduced bird populations.” 
And Blancher’s evidence for “unsustainably high local predation rates or other impacts of cats on bird populations nesting or foraging on or near the ground”  is, upon careful inspection, rather lacking. Although Hawkins, for example, concluded that the differences in species abundance and diversity observed “were the results of the cat’s predatory behavior,”  his inability to account for the fact that 5 of the 9 ground-feeding species included in the study showed no preference for either his “cat area” or “no-cat area”  undermines any such claim.
And, while Baker et al.  claimed that their estimated predation rates were evidence of a dispersal sink, their predation estimates were greatly inflated due to the misapplication of Kays and DeWan’s correction factor (as described previously). And the claim made by Balogh et al.  that 2 of their 3 study sites are sinks, based largely on the basis of 19 fledgling mortalities observed over five months, is at best highly speculative.
The suggestion that these studies demonstrate population-level impacts, even at a local level, simply doesn’t hold up to careful scrutiny.
And finally, Blancher fails to acknowledge the fact that predation does not necessarily lead to population-level impacts. Like all predators, cats tend to prey on the young, the old, the weak, or unhealthy. At least two studies have investigated this in great detail, revealing that birds killed by cats are, on average, significantly less healthy that birds killed through non-predatory events (e.g., collisions with windows or cars). [14, 16] As the Royal Society for the Protection of Birds notes:
“Despite the large numbers of birds killed, there is no scientific evidence that predation by cats in gardens is having any impact on bird populations UK-wide… It is likely that most of the birds killed by cats would have died anyway from other causes before the next breeding season, so cats are unlikely to have a major impact on populations.” 
The policy implication of Blancher’s work are both obvious and serious. Similarly flawed work recently published in the U.S.  has, for example, been used by some [19, 20] as rationale for policy decisions that would likely result in the death of millions of domestic cats.
Less obvious, but just as serious, is the way such work tends to perpetuate misinterpretations and misrepresentations, thereby eroding the foundation upon which future research will be built. Blancher  is correct, for instance, that Kays and DeWans’  adjustment factor “has been used by others to adjust prey returns to prey killed.” [14, 22-24] But, as I’ve explained, the application of this factor to predation estimates of birds is simply inappropriate—regardless of its widespread misuse.
I agree with Blancher  that “studies are needed to assess impacts on bird populations and effectiveness of mitigative measures.” I would add, however, that such work requires rigor well beyond what’s been demonstrated here.
1. Blancher, P., Estimated Number of Birds Killed by House Cats (Felis catus) in Canada. Avian Conservation and Ecology, 2013. 8(2) http://www.ace-eco.org/vol8/iss2/art3/
2. CFHS Feral Cats. 2013. http://cfhs.ca/athome/feral_cats/
3. Ormsby, M. and J. Wilkes, Cats now housebound in Oakville — or owners face fine, in Toronto Star, 2011.
4. Lowe, L., Port of Halifax’s cat fix. The Coast, 2009. http://www.thecoast.ca/halifax/port-of-halifaxscat-fix/Content?oid=1102067
5. Levy, J.K., et al., Number of unowned free-roaming cats in a college community in the southern United States and characteristics of community residents who feed them. Journal of the American Veterinary Medical Association, 2003. 223(2): p. 202–205. http://avmajournals.avma.org/doi/abs/10.2460/javma.2003.223.202
6. Levy, J.K., D.W. Gale, and L.A. Gale, Evaluation of the effect of a long-term trap-neuter-return and adoption program on a free-roaming cat population. Journal of the American Veterinary Medical Association, 2003. 222(1): p. 42–46. http://avmajournals.avma.org/doi/abs/10.2460/javma.2003.222.42
7. Beazley, F.A., Trap, Neuter and Return (TNR) Program, 2009, Halifax Regional Municipality.
8. George, W., Domestic cats as predators and factors in winter shortages of raptor prey. The Wilson Bulletin, 1974. 86(4): p. 384–396. http://elibrary.unm.edu/sora/Wilson/v086n04/p0384-p0396.pdf
9. Fitzgerald, B.M. and D.C. Turner, Hunting Behaviour of domestic cats and their impact on prey populations, in The Domestic Cat: The biology of its behaviour, D.C. Turner and P.P.G. Bateson, Editors. 2000, Cambridge University Press: Cambridge, U.K.; New York. p. 151–175.
10. Loyd, K.A.T., et al., Quantifying free-roaming domestic cat predation using animal-borne video cameras. Biological Conservation, 2013. 160(0): p. 183–189. http://www.sciencedirect.com/science/article/pii/S0006320713000189
11. Krauze-Gryz, D., J. Gryz, and J. Goszczyński, Predation by domestic cats in rural areas of central Poland: an assessment based on two methods. Journal of Zoology, 2012: p. n/a. http://dx.doi.org/10.1111/j.1469-7998.2012.00950.x
12. Hawkins, C.C., W.E. Grant, and M.T. Longnecker, Effect of Subsidized House Cats on California Birds and Rodents. Transactions of the Western Section of the Wildlife Society, 1999. 35: p. 29–33. http://www.tws-west.org/transactions/Hawkins Grant Longnecker.pdf
13. Hawkins, C.C., Impact of a subsidized exotic predator on native biota: Effect of house cats (Felis catus) on California birds and rodents, 1998, Texas A&M University.
14. Baker, P.J., et al., Cats about town: Is predation by free-ranging pet cats Felis catus likely to affect urban bird populations? Ibis, 2008. 150: p. 86–99. http://www.ingentaconnect.com/content/bsc/ibi/2008/00000150/A00101s1/art00008
15. Balogh, A., T. Ryder, and P. Marra, Population demography of Gray Catbirds in the suburban matrix: sources, sinks and domestic cats. Journal of Ornithology, 2011: p. 1–10. http://dx.doi.org/10.1007/s10336-011-0648-7,
16. Møller, A.P. and J. Erritzøe, Predation against birds with low immunocompetence. Oecologia, 2000. 122(4): p. 500–504. http://www.springerlink.com/content/ghnny9mcv016ljd8/
17. RSPB, Are cats causing bird declines? 2011. http://www.rspb.org.uk/advice/gardening/unwantedvisitors/cats/birddeclines.aspx
18. 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
19. Fenwick, G.H., House cats: The destructive invasive species purring on your lap, in The Baltimore Sun, 2013.
20. Lynes, M., No. 1 bird killer is outdoor cats, in San Francisco Chronicle. 2013. http://www.sfgate.com/opinion/openforum/article/No-1-bird-killer-is-outdoor-cats-4250692.php
21. Kays, R.W. and A.A. DeWan, Ecological impact of inside/outside house cats around a suburban nature preserve. Animal Conservation, 2004. 7(3): p. 273–283. http://dx.doi.org/10.1017/S1367943004001489
22. Baker, P.J., et al., Impact of predation by domestic cats Felis catus in an urban area. Mammal Review, 2005. 35(3/4): p. 302-312. http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2907.2005.00071.x/abstract
23. Thomas, R.L., M.D.E. Fellowes, and P.J. Baker, Spatio-Temporal Variation in Predation by Urban Domestic Cats (Felis catus) and the Acceptability of Possible Management Actions in the UK. PLoS ONE, 2012. 7(11): p. e49369. http://dx.doi.org/10.1371%2Fjournal.pone.0049369
24. MacLean, M.M., et al., The usefulness of sensitivity analysis for predicting the effects of cat predation on the population dynamics of their avian prey. Ibis, 2008. 150(Suppl. 1): p. 100–113. http://dx.doi.org/10.1111/j.1474-919X.2008.00864.x
Peter J. Wolf is the Cat Initiatives Analyst for Best Friends Animal Society, one of the nation’s largest animal welfare organizations and a leader in the development and operation of community cat programs. Peter’s role involves the application of research to a range of communications- and policy-related projects. Peter is also the founder of Vox Felina, a blog featuring in-depth analysis of science and policy issues related to the management of free-roaming cats.