How Effective Is the Birdsbesafe® Cat Collar at Reducing Bird Mortality by Domestic Cats?

Many anthropogenic factors contribute to bird population decline. Researchers often cite collisions with buildings, wind turbines, and powerlines as major sources of mortality, resulting in the loss of millions of birds (Loss et al. 2015). Owned and feral domestic cats kill an estimated 2.4 billion birds each year in the United States alone (Loss et al. 2013, 2015); they pose a greater threat to songbird populations than any other anthropogenic factor save habitat degradation and destruction (Loss et al. 2013, 2015; Nogales et al. 2013; Marra and Santella 2016). In fact, domestic cats account for a significant percentage of the mortality of several bird species including northern mockingbird Mimus polyglottos, dunnock Prunella modularis, northern wren Troglodytes troglodytes, and great tit Parus major and create population sinks and areas of decreased population density for many other species (Loss and Marra 2017).

Loss and Marra (2017) note that domestic cats are one of the most widespread and ecologically harmful invasive species in the world. Cats are generalist predators, and will hunt birds, lizards, frogs, and small mammals even when provided with sufficient food by owners or caretakers (Loyd et al. 2013). Researchers have shown that hunting in domestic cats is an instinctive behavior, rather than one prompted solely by hunger (Adamec et al. 1976; Dickman and Newsome 2006). Although a recent study has shown that feeding cats a diet high in meat can reduce predation by 36% (Cecchetti et al. 2021), in general domestic cats act as subsidized predators, freed from normal ecological checks and balances by human intervention (Barratt 1998). Although hunting by domestic cats can greatly reduce prey populations, reduced prey populations will not in turn cause a reduction in cat populations (Kitts-Morgan 2015).

As subsidized predators, domestic cats have a significant impact on native wildlife populations (Marra and Santella 2016). Multiple conservation organizations advise that owners keep domestic cats indoors in an effort to protect birds (American Bird Conservancy 1997; Wildlife Society 2020). However, this solution is neither simple to carry out nor free of controversy. Many people are reluctant to keep their cats inside, and not all domestic cats can be easily managed indoors (Crowley et al. 2018). Some cat owners are both aware of and distressed by the potential impacts of their outdoor cats on wildlife (Crowley et al. 2018) and there are a variety of devices available to potentially reduce hunting success without confining outdoor cats.

Collar-mounted bells, the most-researched antipredation strategy, work by providing an auditory alert to potential prey. Several studies have shown belled collars to decrease hunting success compared to collars without a bell. For example, two studies, Ruxton et al. (2002) and Gordon et al. (2010), documented ≥ 50% reduction in prey items brought home by cats wearing belled collars compared to those wearing collars without bells. Ruxton et al. (2002) tracked 21 cats over 8 wk and Gordon et al. (2010) tracked 37 cats over 12 wk. Nelson et al. (2005) found that cats wearing belled collars showed a 41% reduction in hunting success for birds (n = 89 cats, over 12 wk).

Electronic devices such as CatAlert®, which produces a beep every 7 s to warn prey, are also available; a study by Nelson et al. (2005) showed them to decrease hunting success similarly to belled collars. Other methods include visual alerts and physical barriers. For example, the CatBib® is a colored flap of neoprene material that hangs from the cat's collar, acting both as a visual alert to potential prey as well as a physical barrier to the cat's paws (Cat Goods LLC). This device prevented 81% of normally successful hunters from catching birds and 45% from catching mammals (Calver et al. 2007). However, it physically restricts a cat's forward movement and may hinder escape from a predator or vehicle.

In this study, we examined the effectiveness of another visual device, the Birdsbesafe® collar cover (BCC), which was developed specifically to help reduce domestic cat predation on songbirds (Birdsbesafe LLC 2021). The BCC (Figure 1) is a 5-cm-wide tube of brightly colored and patterned cotton fabric that fits around a standard breakaway collar. The fabric stands out in a ruff around the cat's head, making it more difficult for the cat to conceal itself while stalking prey. Songbirds have tetrachromatic color vision, made particularly acute by oil droplets attached to their cone cells, so the bright colors of the BCC are highly visible to them (Chen and Goldsmith 1986; Burns and Shultz 2012). Four previous studies conducted in Australia, the United States, and northern Europe have examined the effectiveness of the BCC (Hall et al. 2015; Willson et al. 2015; Pemberton and Ruxton 2020; Cecchetti et al. 2021). Each found significant reductions in the number of bird prey items brought home by domestic cats wearing the BCC compared to wearing no collar (Hall et al. 2015; Willson et al. 2015; Pemberton and Ruxton 2020; Cecchetti et al. 2021). These studies, as well as our own, used prey items brought to owners' doorsteps or porches as a general indicator of hunting success, as exact predation rates are impossible to measure without collar-mounted video cameras. Previous studies on BCC did not take factors such as cat demographics or hunting environment into consideration. Research on cats without predation-reducing devices determined that cats invested equal effort in hunting across age groups, sex, and hunting environments, but younger cats killed more prey than older cats (Loyd et al. 2013); however, previous research has not established whether the BCC is equally effective across these categories.

Our primary objective in this study was to determine if cat demographics have an impact on the effectiveness of the BCC. By combining data from two studies in New York (2014 and 2019) and one in Florida (2019), we tested the hypothesis that the BCC is equally effective at reducing bird mortality regardless of cat age, sex, or hunting environment. The 2014 New York data were also published by Willson et al. (2015). Using the two different study sites, we also tested the hypothesis that the BCC is more effective at reducing bird predation at higher latitudes. We base this hypothesis on differences in life history between temperate and subtropical latitudes, including vigilance levels of breeding birds (Ghalambor and Martin 2001; Willson 2016). Finally, we examined owner perceptions of cat comfort and their intentions regarding continued use of the BCC.

We followed the same methods of data collection as Willson et al. (2015), asking volunteer cat owners to test the devices on their outdoor cats. To obtain participants, we advertised through local news and internal email platforms in St. Lawrence County, New York, and Gainesville, Florida, and posted flyers in local community buildings. We specifically recruited cats that owners reported to be hunters of vertebrates and that regularly left their uneaten prey on their owners' doorsteps or porches.

Data collection for the New York (2014) study lasted a total of 12 wk (April–June), divided into six 2-wk periods (Willson et al. 2015). In contrast to the 2014 study, data collection for the studies in New York and Florida in 2019 lasted 8 wk, divided into four 2-wk periods (May–July). We chose the time periods in all three studies to correspond with local bird breeding seasons. In each location, we randomly divided the cats into two even groups (group 1 and group 2) to stagger the times cats had their collars on and off, with the exception that we sorted cats living in the same household into the same group. We made this exception because owners could not always be certain which cat in a multi-cat household brought home which prey item. Cats in group 1 spent the first 2 wk with the BCC and interior quick-release collar on, and group 2 began with no BCC or interior collar. Every 2 wk, both groups switched, so that each individual cat spent a total of half the study wearing the BCC.

We instructed cat owners to collect all vertebrate prey items brought home by their cats, bag and label them with the date and collar status, and freeze them. Cat owners also recorded any prey observed with their cats that subsequently escaped or was consumed. We recognize that cats may abandon or completely consume some of their kills (cats with collar-mounted video cameras only brought about 23% of prey items home; Loyd et al. 2013) and that this would have gone unrecorded in our study. At the end of the data collection period, we collected all frozen prey specimens from the owners and identified them to species when possible. We grouped two small mammal species, the white-footed mouse Peromyscus leucopus and the North American deermouse Peromyscus maniculatus, together as they are extremely difficult to distinguish morphologically. The 2014 and 2019 studies used BCCs with slightly different patterns, but both collars consisted of bright sections of red, blue, and yellow (Figure 1). At the end of the data collection period, each participant completed an exit survey. For all three studies, participants rated cat comfort level with the BCC and stated whether they planned to continue using the BCC. For the 2019 studies, participants also stated if their cats had previously worn any type of collar.

We performed randomization of means tests to determine the effectiveness of the BCC at reducing hunting success of domestic cats by comparing the amount of prey recovered per cat with and without the collar. Since this type of test uses resampling, it does not assume a normal distribution of data. We performed separate randomization of means tests for birds, mammals, and herpetofauna, using 5,000 randomly generated samples for each test. Only cats that brought home at least one bird were included in the analysis for birds, and likewise for small mammals and herpetofauna. We also used randomization of means tests to examine differences in the effectiveness of the BCC between latitudes. Here, we used mean reduction in prey recovered with the BCC on vs. off to measure BCC effectiveness at each latitude.

We performed multiple linear regression analysis to determine if cat age, sex, or hunting environment explained the variation in the effectiveness of the BCC. We used mean reduction of prey recovered with the collar on vs. off as the metric for effectiveness. All three variables were treated as categorical, with age separated across young (1–6 y), middle (7–10 y), and old (11+ y), and hunting environment separated into rural and village, as described by cat owners. The term village encompasses environments with more buildings, such as suburbs and neighborhoods. We used second-order Akaike information criterion (AICc) values to rank the linear regression models of all possible combinations of the three variables, with prey type as a covariate (Burnham and Anderson 2002). Finally, we performed Pearson's χ² tests for independence to determine if there was a relationship between collar comfort and previous experience wearing a collar. For all tests, we considered P values below 0.05 to be significant. All statistical analyses were performed in R (R Core Team 2020) using the packages reshape, ggplot2, dplyr, AICcmodavg, and splitstackshape (Wickham 2007, 2016; Mahto 2019; Mazerolle 2019; Wickham et al. 2019).

For all years, we recruited 118 cats living in 81 households for the study. Owners from 17 households who were either noncommunicative or uncomfortable with the collars removed 24 cats from the study. Six of the removed cats (two households) were from the 2014 New York study, 8 (five households) were from the 2019 New York study, and 10 (10 households) were from the 2019 Florida study. We analyzed data on 94 cats living in 64 households: 19 cats in 10 households (2014) and 43 cats in 28 households (2019) in New York, and 32 cats in 26 households (2019) in Florida (Table 1). Seventy-nine (84%) of the study cats brought home at least one animal over the course of the study. Of those 79 cats, 67% captured birds, 75% captured mammals, and 30% captured reptiles or amphibians. (The sum of percentages exceeds 100% because 49 cats returned home with prey from more than one taxonomic category.) All but four of the herpetofauna killed were in Florida. We collected a total of 373 prey specimens, including 55 reported by Willson et al. (2015): 92 adult birds representing 35 species, 172 small mammals representing 14 species, 99 reptiles, 4 amphibians, 1 insect, and 5 unidentifiable animal remains (Tables 2 and 3).

The randomization of means tests showed that the BCC was effective at reducing mortality of birds but not small mammals. Cats brought home 61% fewer birds when wearing the BCC compared to no collar (P = 0.006). Cats also brought home fewer small mammals (37% fewer) and herpetofauna (47% fewer) while wearing the BCC, but the differences were not significant (P = 0.21 for small mammals, and P = 0.31 for herpetofauna; Figure 2). We found a difference in the effectiveness of the BCC between latitudes for birds (P = 0.047), but not for small mammals (P = 0.21). While cats were wearing the BCC, their predation on birds was 69% lower in New York and 47% lower in Florida. As we hypothesized, the BCC was a better tool at decreasing predation on birds in a temperate latitude compared with a subtropical latitude.

We found no differences in the effectiveness of the BCC based on cat age, sex, or hunting environment. Hunting environment was present in all the top multiple linear regression models analyzed by AICc (Table 4), but the effectiveness of the BCC was not appreciably different between categories (Figure 3). According to randomization of means tests, there was some difference in hunting success (measured as total prey recovered per cat) based on age; middle-aged cats brought home more prey than old cats (2.7 more prey items recovered per cat, P = 0.042). The hunting success for young cats was not different from middle-aged (P = 0.31) or old cats (P = 0.12; Figure 4). There was no difference in hunting success between sex (P = 0.089) or hunting environment categories (P = 0.11; Figure 4).

We received collar comfort survey responses from 41 households between the 2019 studies, representing 60 study cats. Thirty-three cats (55%) displayed no adverse reaction to the BCC, while eight (13%) showed strong negative reactions to the BCC. The remaining 19 cats (35%) showed some discomfort, but quickly became accustomed to the BCC. At least 8 of the 24 cats that owners removed from the study also displayed negative reactions to the BCC. There was no relationship between collar comfort rating and previous experience with any type of collar (χ² [3, n = 59] = 2.63, P = 0.45). Only 20 owners (49%) stated that they planned to continue using the BCC.

The BCC reduces domestic cat predation of birds but not of small mammals or herpetofauna (Hall et al. 2015; Willson et al. 2015; Pemberton et al. 2020; Cecchetti et al. 2021). Our estimate of a 61% reduction in avian mortality is within the range found by previous studies of avian mortality and the BCC in other areas—a 50% reduction in Australia and a 78% reduction in England (Hall et al. 2015; Pemberton et al. 2020). Our results, along with the results of the aforementioned studies, suggest that the BCC could ameliorate the negative impacts of domestic cats in areas where they prey on vulnerable populations, such as those on islands or in areas with heavily managed songbird populations (Bell and Merton 2002; Bocetti et al. 2012; Marra and Santella 2016).

Our study supports a potential explanation for the differences in mortality reductions observed at different latitudes. Several studies agree that during breeding season, testosterone levels increase to higher concentrations compared to baseline values in higher-latitude birds than in tropical birds (Goymann et al. 2004; Hau et al. 2008, 2010). These higher testosterone levels allow temperate birds to secure territories, nest, and raise young during shorter breeding seasons, but they may also cause more distraction in temperate birds compared to tropical birds, hence making them more vulnerable to predation (Willson 2016). Results from studies on the effectiveness of the BCC support this hypothesis. We found that the BCC was more effective at higher latitude in the United States, while Pemberton et al. (2020) found higher effectiveness in England than Hall et al. (2015) observed in Australia.

The BCC can be a potential tool for outdoor cat owners who wish to decrease bird mortality in their own yards and neighborhoods. We note that our results—a 61% reduction in bird mortality with the BCC—are less dramatic than the figure of 95% obtained by Willson et al. (2015) from the 2014 New York study. This could be related to sample size, as mean bird mortality per cat was lower in the 2014 study (1.36 birds per cat) compared with the two 2019 studies (2.08 and 1.64 birds per cat; Table 5).

The BCC was effective regardless of cat age, sex, or hunting environment. However, for it to be successful, cat owners must be motivated to fit their cats with the BCC. Crowley et al. (2018) found that many cat owners have concerns over the safety of collars for outdoor cats, and are not convinced of the effectiveness of collar-mounted antipredation devices. Physical barriers such as the CatBib may be more effective at reducing bird predation than the BCC, but cat owners consider their use even less palatable. Although our study addressed both safety and efficacy, less than half of the 2019 study participants said they planned to continue using the BCC. Some of those who did plan to continue commented on the attractiveness of the BCC, calling it “cute,” “dashing,” and “adorable.” Owner perceptions are key to the successful implementation of any strategy for reducing domestic cat predation on birds.

One participant stated that they would consider using the BCC if their cat killed more birds. The view that an individual cat does not kill enough to threaten wildlife populations is common; cat owners generally do not accurately perceive their cats' predation rates (McDonald et al. 2015; Crowley et al. 2018). Given these findings, as well as the observation that cats only bring about 23% of their prey home, a cat owner's view that their cat does not kill a problematic number of birds is a potentially inaccurate and harmful one (Loyd et al. 2013). A recent study by Kays et al. (2020) found that owned, inside–outside domestic cats stay within 1 km² of their homes, resulting in very high cat densities near human residences; thus local predation levels by domestic cats may be 2–10 times those of wild predators. More than 57 million households in the United States place bird feeders outside their homes, attracting birds to these areas of heightened predation risk (U.S. Department of the Interior et al. 2016).

Besides cat owner perceptions, the biggest obstacle to the use of the BCC is the effectiveness of the interior collar. In this study we used breakaway collars in order to assuage owners' fears about collar safety. Over the 4 wk that cats in the 2019 studies were wearing the BCC, 44 cats lost their collars at least once. Seven cats lost their collars four or more times, with a mean of two collar losses per cat. Many of our study participants expressed frustration with the easy release of the collars, and 20 cat owners modified the interior collars in some way to prevent collar loss. Most of these modifications removed the breakaway function of the collars. With combined BCC and interior collar costs nearing U.S. $20 per unit, few cat owners may have interest in repeatedly buying replacement units for their prowling cats.

If 1% of U.S. cat owners used the BCC with their cats, avian mortality could be reduced by approximately 4.9 million birds in the United States each year, using the estimate of 744 million birds killed per year by owned, inside–outside domestic cats (Loss et al. 2013) and non–breeding-season data from Willson et al. (2015). However, Loss et al. (2013) estimated that owned domestic cats account for less than one-third of cat-caused avian mortality in the United States. Feral cats, which are estimated to hunt more actively than owned cats, kill the majority of birds (Loss et al. 2013). Many feral cats live in colonies managed with trap–neuter–return methods; in some colonies, dedicated trap–neuter–return nonprofit volunteers often feed and interact with the feral cats on a daily basis. Fitting colony cats with the BCC is one potential tool for reducing their ecological impact. Fundraising for cat food, shelter, and neutering is a common approach to colony cat care (Trap–Neuter–Return Texas 2021) and could potentially assist in the purchase of BCCs and interior collars for these cats.

The BCC will not solve the devastating problem of songbird predation by domestic cats. It is ineffective for nestling and fledgling birds, which cannot easily escape when the BCC alerts them to a cat's presence. In addition, there is no reason to expect that the BCC reduces sublethal effects of cats on songbirds, such as reduced foraging efficiency due to vigilance or frequent escape flights. For example, the presence of a cat model decreased nest provisioning in the European blackbird Turdus merula, leaving nests more vulnerable to predation (Bonnington et al. 2013). For these reasons, the BCC could be one useful tool within a larger conservation initiative. Homeowners association covenants, or local ordinances such as permit mitigation requirements, could mandate the use of antipredation devices such as the BCC for outdoor cats. These measures would be most effective within 1 km of critical migratory stopover areas or threatened species' habitats, based on the 1-km² or less home ranges observed by Kays et al. (2020) for domestic cats.

Although we found that the BCC reduces avian mortality during the breeding season for songbirds, it is just one tool available for reducing bird mortality from anthropogenic sources. Since domestic cats account for more annual bird deaths than any other anthropogenic source apart from habitat loss, the most desirable conservation action is to bring all cats indoors (Loss et al. 2015). However, all methods of reducing avian mortalities by domestic cats will involve public education and outreach campaigns.

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author.

Data S1. Raw data on prey recovered from domestic cats across two 8-wk studies of the Birdsbesafe collar cover in northern New York and Florida (spring 2019) and a 12-wk study in northern New York (spring 2019). We submitted data as a CSV file showing age, sex, latitude, and hunting environment (defined as village or rural by owners) for each cat.

Available: https://doi.org/10.3996/JFWM-21-055.S1 (14 KB CSV)

Data S2. Demographic information on domestic cats participating in two 8-wk study periods of the Birdsbesafe collar cover in northern New York and Florida (spring 2019) and a 12-wk study period in northern New York (spring 2014). Submitted as a CSV file showing species, prey type, and whether or not the cat was wearing the collar for each specimen.

Available: https://doi.org/10.3996/JFWM-21-055.S2 (2 KB CSV)

Reference S1. U.S. Department of the Interior, U.S. Fish and Wildlife Service, U.S. Department of Commerce, and U.S. Census Bureau. 2016. National survey of fishing, hunting, and wildlife-associated recreation.

Available: https://doi.org/10.3996/JFWM-21-055.S3 (28.837 MB PDF) and https://www.fws.gov/wsfrprograms/subpages/nationalsurvey/nat_survey2016.pdf

The authors would like to thank all the cats and their owners for participating in the study. We thank Mackenzie Dorr from University of Florida GREBE (Gators Ready for Exceptional Birding Experiences) for logistical support. Thanks also go to Dr. Kristine Hoffmann for aid with statistical analysis. We are grateful to the Associate Editor and anonymous reviewers for valuable insights that improved earlier drafts of this manuscript. This work was supported by St. Lawrence University. The Florida portion of this study was performed under the auspices of the University of Florida IACUC approval number 201910591, and the New York portion under St. Lawrence University IACUC approval numbers F13-4 and S19-2. The survey described in this report was organized and implemented by Susan Willson and was not conducted on behalf of the U.S. Geological Survey.

Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.