Every year, anywhere from 100 million to 1 billion American birds are thought to die after colliding with windows. Planners and builders have tried a variety of techniques to reduce collisions, but, according to a trio of researchers from the University of Alberta, these mitigation efforts are ultimately stymied by a poor understanding of what environmental conditions encourage collisions in the first place.

The Canadian team recently looked into this issue with the help of conservation biology undergraduate students who assisted with the study as part of their coursework. The main goal of the research was to identify the habitat characteristics associated with collisions; in particular, the research team investigated whether birds were more likely to be injured in urban versus rural environments, at houses versus apartments, near places with versus without bird feeders, and on older or younger properties (which potentially vary in the extent of surrounding vegetation). In addition to these environmental data, the research team also collected information on the number of birds hurt, the species of birds involved in collisions, and whether collision rates varied between seasons.

(The authors explored whether the presence of bird feeders, among other environmental factors, increased the likelihood of collisions with windows.)

The undergraduate students were responsible for drafting volunteers to provide all of these data. To this end, they created public-awareness pamphlets that were distributed to friends, family, and colleagues, both in paper form and via online social media. The pamphlets asked would-be participants to visit the conservation course website in order to read more about the project and enter information about avian collisions at their homes. The questionnaire covered eight "core" questions that collected information on participant demographics, observation and timing of bird collisions, length of home occupation, type of residence, presence of bird feeder, and location of home. This latter detail was provided in the form of a postal code, which was used to link the survey responses to census data (on neighborhood age) and to locate the home on Google Earth so that it could be classified as either urban or rural. The surveys also asked residents to identify which species were involved in collisions.

Of the ~1,800 people who participated in the survey over the two-year study period, approximately half remembered a bird-window collision at some point in the past; just over a third of individuals indicated that a collision had occurred within the past year alone. Although the mean collision rate was only 1.7 birds per year, this varied widely from one home to the next; several dozen collisions were reported at some residences, and the worst tally was 84 at a single home in a single year. Luckily, (known) mortality rates averaged less than 1 bird per year.


(Installation of decals may be one way to reduce the number of bird-window collisions. Image courtesy of Whispering Pines.)

Generally speaking, rural residences had more collisions and mortalities than urban residences; bird-window encounters were least common at apartments. Feeders tended to be associated with a higher number of collisions and mortalities, though dynamics varied a bit in urban and rural environments. There was some evidence that age of neighborhood influenced the likelihood of collision, but only in urban areas. Both the number of collisions and the number of deaths peaked in the summer and were lowest in the winter--perhaps because avian abundances vary greatly between these two seasons, since a large proportion of Canadian birds head south for the winter. Sparrows were the most common casualty of collisions, with chickadees and robins close behind; for all three species, a higher percentage of birds survived than died, though the difference was pretty slim for robins.

One of the reasons the authors were interested in pursuing this research was that there is a relative dearth of descriptive work on avian-window collisions. Results of the few studies that have been conducted have been used to create what the authors call a "social construction," or a widespread belief about something (in this case, the danger of windows) that is not actually based on scientific fact. Thus, the current study was developed not only to ascertain just how many birds truly are impacted by windows each year, but also to examine the environmental correlates of collisions--information that could be vital for developing useful conservation and management plans. Given the contributions of undergraduate students and members of the public, an additional benefit of the study is that it contributes to education and awareness.

(American robin, Turdus migratorius, a species that was particularly susceptible to both collisions, in general, and collision mortality. Image courtesy of Wikipedia.)

The researchers acknowledge some flaws in their work--in particular, the inaccuracies and biases associated with self-reporting--but they still feel that their results have made some useful contributions to the field. For example, their finding that feeders are associated with a higher likelihood of collision suggests that people should be careful to place seed and suet dispensers away from windows, or make their windows more obvious (e.g., by decorating them with decals and trimming back vegetation). By identifying that rural and older homes are more likely to experience collisions, the study helps show managers and planners which regions to focus on when attempting to mitigate the negative effects of windows. 

Despite these important findings, more work is needed for the future. Birds that appear to survive collisions may simply fly off and die elsewhere. This makes it difficult to accurately measure the effects of windows on individual health and fitness. Thus, better demographic data are needed, as are, potentially, tracking studies that follow injured birds after they fly off. This information will also be important for understanding the long-term (population-level) impacts of collisions; it may be that the benefits of supplemental feeding outweigh the drawbacks of increased exposure to windows, but it might also be the case that birds would be better off without either the bird seed or the windows. The authors recommend that future studies move away from opportunistic observations (as made in the current project), and instead involve either experimentation or  a "snapshot" approach, during which researchers would undergo an intensive period of comprehensive data collection. Given their positive experiences with mobilizing the troops via the Internet, they recommend the use of online social network sites for increasing participation in any future work on avian collisions.

---

If you enjoyed this article, you might want to follow Anthrophyis on Facebook!

Bayne, E.M., Scobie, C.A., and Rawson-Clark, M. 2012. Factors influencing the annual risk of bird-window collisions at residential structures in Alberta, Canada. Wildlife Research 39(7):583-592.

For more information on bird-window collisions, check out the website of the Acopian Center for Ornithology.