UBC Undergraduate Research

Bird-window collision : a problem at UBC buildings Cavers, Gordon; Chien, Andy; Leung, Carmen; Nam, Tiffany Apr 12, 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52966-Cavers_G_et_al_ENVR_400_2015.pdf [ 8.29MB ]
JSON: 52966-1.0074569.json
JSON-LD: 52966-1.0074569-ld.json
RDF/XML (Pretty): 52966-1.0074569-rdf.xml
RDF/JSON: 52966-1.0074569-rdf.json
Turtle: 52966-1.0074569-turtle.txt
N-Triples: 52966-1.0074569-rdf-ntriples.txt
Original Record: 52966-1.0074569-source.json
Full Text

Full Text

  Bird­Window Collision: A Problem at UBC Buildings Final Report   In partnership with:  UBC SEEDS Program UBC Sustainability and Engineering UBC Building Operations Environment Canada                April 12, 2015 An Environmental Science 400 Project  Gordon Cavers Andy Chien Carmen Leung Tiffany Nam  Abstract Collision with building windows is one of the main causes of direct human­related bird deaths in North America, particularly during migration seasons. Vancouver has the highest densities of wintering birds of any Canadian city, so occupants on the University of British Columbia campus were interested in finding out whether or not bird collisions are a significant problem. As a result, for a period of eight months, four Environmental Sciences students, partnering up with The UBC SEEDS program, UBC Sustainability and Engineering, UBC Building Operations and Environment Canada, compiled and analyzed bird­collision data on the UBC Vancouver campus. The project mainly focuses on the problem of bird­building collisions and the severity of the situation. Ten buildings were chosen for their have high percentage of glass cover and are suspected to be the most problematic on campus. The number of bird collisions at each selected building were then recorded and examined for patterns. Since this study is one of the first bird­window collision studies on UBC, the results will help to inform future UBC bird­friendly building guidelines and future studies. .                   1 Table of Contents      ​Abstract​                                                                 1  Introduction​                                                           3  Methods​                                                                4  Results​                                                                  9  Discussion​                                                           13  Recommendations​                                               15  Conclusion​                                                           16  Acknowledgements                                              16  Team                                                                    17  References                                                           18    2 Introduction Studies have stated that bird­window collision (“strike”) cases on buildings have been escalating around the world and are results of birds not identifying windows in buildings, bus shelters, glass walkways, and glass partitions as barriers (Machtans et al., 2013). Birds may fly into windows, corner windows, glass walkways, glass partitions, and glassed lobbies when they see the reflection of clouds, sky or trees in the clear properties of glass (Klem, 2014). The reflection gives a mistaken impression to birds that they are flying into open air and it allows habitat or sky to be seen on the other side, creating the appearance of a clear passage.  Birds hitting windows is a ubiquitous problem; it is rare to find someone who hasn’t heard of or                                   witnessed a bird colliding with a window. The University of British Columbia Vancouver                         campus is no exception. The UBC SEEDS program, along with UBC Campus and Community                           Planning, UBC Building Operations and Environment Canada decided that it was appropriate                       to gather information on the bird­building collision situation on the UBC Vancouver campus. In                           the eight­month time period, the organizations worked closely with four environmental                     sciences students on this research project.  25 million birds are estimated to be killed annually in collisions with buildings in Canada                             (Machtans et al. 2013). Though the number is enormous, most of the data that had been                               collected were from Eastern North America, especially from cities with very high densities of                           birds during the migratory period, and not many documents have been established on the                           issue in areas west of the Rocky Mountains. However, from online peer­reviewed articles on                           the topic, it is found that windows in human residential and commercial structures are one of                               the main causes of bird deaths worldwide (Klem, 2014). It is difficult to exactly state the risk                                 for each bird species. Some birds may be more likely to encounter bird strikes, but study                               findings can disagree. In a study by Hager et al.(2014), it was found that bird­window                             collisions were highest in the least abundant species, such as red­eyed vireo, and lowest in                             species with high abundance values, for example the chipping sparrow (Hager & Craig,                         2014). Another study found adults and immatures equally likely to suffer a strike, but that the                               species with the majority of collisions would change depending on the season (Klem 1989). In                             disagreement with Hager et al., they cited unpublished data from Graber and Graber that                           showed it was birds with highest densities that had the highest number of strikes (Klem 1989).                               To limit harm to the bird community at UBC, it is necessary to be aware of the impacts that                                     our building structures have on the local bird community so that informed decisions can be                             made. Our project was built on previous work and can hopefully help shed light for future                               proposals.   The statistics collected for this project would be supplied for future reference and used for                             establishing bird­friendly guidelines and policy. The guidelines and policy can be taken into                         consideration when planning the layout and materials for a new building. This is particularly                           3 important as the University is continually investing in new and upgraded infrastructure,                       including the Engineering Student Centre to be completed in late 2015, Orchard Commons                         and the New Aquatic Centre, both to be completed by 2016 (Infrastructure Development,                         2014).   From all the data collection and analysis, the following two questions that had been proposed                             at the start of the project have also been answered: 1) What are the collision rates at selected buildings on the UBC Vancouver campus during the                               winter months? 2) If there is a bird strike problem, defined as five or more birds hitting a building in our time                                       frame, what recommendations and suggestions can be proposed to minimize the negative                       effects?   Methods 1) DATA COLLECTION – “BIRD SURVEY”   i.  Time Period: Total of 27 searched days between November ­ February Data collection took place from November 18, 2014 through February 26, 2015.                       Excluding the holidays, winter and reading breaks, we collected the data for three                         consecutive days per week for a four month period.  ii.Location of study: UBC Vancouver The ten buildings we have selected for our study are shown below in Figure 1., and the                                 names of the buildings are listed 1. through 10. The first group looked at the first five                                 buildings listed, the second group the next five buildings, respectively.   4  Figure 1. ​The ten selected buildings on UBC Vancouver Campus and their locations on                           the map of UBC. These buildings are known for having a relatively high percentage of glass                               cover on the exterior and were chosen under approval of our community partners.  1.​    ​Museum of Anthropology 2.​    ​Koerner Library 3.​   ​Earth Sciences Building (ESB) 4.​    ​Fred Kaiser 5.​    ​Totem Residence (hәm'lәsәm' House) 6.​     ​Irving K. Barber Learning Centre 7.​    ​Beaty Biodiversity Centre 8.​   ​Woodward Library  9.​    ​Centre for Brain Health 10.​  ​Pharmaceutical Sciences Building       5 iii.​              ​Tools   Figure 2a. The data collection sheet used throughout the study consisted of 16                         columns of various information such as the date, building ID, %glazing,                     distance from building to nearby vegetation.   We used the above chart to record data collected. The information we have collected as a                               part of our data set includes the date, building ID, observer, start time, end time, number                               of carcasses found, the state of the bird when spotted, comments, % glazing, distance                           from building to nearby vegetation as shown above in the chart. Other equipment we have                             used includes binoculars, clipboard, pen and/or pencil, and sanitary gloves to handle                       carcass and any other signs of bird collisions such as smears, feathers stuck on windows,                             and smudges.  6 Figure 2b. Sample photo of bird collision evidence. Carcass (left) was found on Nov                           18, 2014 by the S facade of Totem Residence while the small feathers stuck on                             windows (right) were found on Nov 25 by the same location.   ii.​            ​General methodology of “Bird Survey” ● Two teams of two each were assigned five of the ten buildings. ● Each building was divided into four different facades (N, S, E, W). ● For each building, two observers went around the building in opposite directions,                       observing windows and any other exterior surfaces for any signs of bird collision                         evidence and filled in the above chart for later reference and analysis. Splitting up was                             important to provide a different angle, allowing one member to catch something their                         partner missed. ● The data was then later uploaded on Google Doc into a combined data set.           7  2) DATA COLLECTION – Citizen Science   Figure 3. The digital version of the poster created to fulfill citizen science portion of our                               project. The posters were either printed and put up around the buildings or digitally                           displayed on the screens in various buildings.   We have also incorporated Citizen Science into our study in which we set up a Gmail                               account for our group (bird.strikes.ubc@gmail.com) to encourage the public (though                   restricted to people on UBC Vancouver campus) to send us an email should they witness                             a bird collision. We put up posters around the buildings we were surveying, and also put                               up some digital advertisements. For the entire study period, we received two emails from                           the public and also received notifications regarding carcasses being spotted from our                       representatives from Environment Canada and UBC Building Operations.   3) DATA ANALYSIS Data collected from each building were analysed and fitted into a Poisson Regression                         Model(graph IV) by using statistical software R. Each facade was treated as a data point on                               the graph. Data from Koerner Library was not included due to the green roof ledge area in                                 front of the windows that would make the analysis invalid. The model assumes that each                             8 facade is independent of one another, follows a linear relationship with collision counts, and                           is identical in residual distribution and variance. Results Graph I. (a) ​Total number of bird/window collisions over all buildings. Data collected on 27                             days, across a fourteen­week period. The gaps represent the December exam period,                       holidays and reading break, when surveys were not conducted.      9 Graph I. (b) Total number of bird/window collisions for each building. Data collected on 27                             days, across a fourteen­week period. The gaps represent the December exam period,                       holidays and reading break, when surveys were not conducted. Graph II. ​Total recorded numbers of birds hitting windows (“collision counts”) at the ten                           buildings being monitored in this study. Data collected on 27 days, across a fourteen­week                           period at UBC Vancouver.             10 Graph III​. Total collision counts at each facade and their distance from the nearest vegetation.                             Three levels of facade distance from the vegetation were used. Level A is 0­2 meters, B is 2­10                                   meters and C is a building distance of 10 meters or more from vegetation. Data collected on 27                                   days, across a fourteen­week period at UBC Vancouver.  Graph IV. Glazing percentage versus total collision counts at each facade. Each data point is                             labeled with its respective facade of building. The Standard Error of estimated slope                         is 0.0052 which is small and the slope is statistically significant at 5% level with aȕ                                   P­value of 0.0294. A weak positive correlation of 0.19 was found between glazing percentage                           and collision counts. 11  Figure 4. Relative positions of the buildings and their total collision counts on campus                           map. In total, at our ten buildings we found evidence for 60 collisions (see graph III). The museum                                 of Anthropology(MOA) had the top collision count with 18 collisions, Beaty Biodiversity                       Museum (Beaty) and Earth Science Building (ESB) came in second with 11 counts each and                             Irving Library (Irving) came in third with 9 counts observed in 27 count days. Five out of ten                                   buildings observed, MOA, Beaty, ESB, Irving and Totem, had collision counts more than 5                           which is the number we defined to be a problem. These problematic buildings have many                             things in common. First is that they all have large and mostly clear windows. Second is that                                 there are large open space around the buildings. The sky and trees would be reflected in the                                 glass and birds would attempt to fly through and strike the window. Also the trees and                               vegetation around the buildings which attract birds would increase the chance of collisions. A                           possible contributing factor to the high number of strikes at the museum of anthropology could                             be that there is a pond combined with a forest patch in front of the 15m high window. The                                     collision rate seems to decrease at most of the buildings towards the end of January 2015                               (graph I and II). We observed that (graph IV) when the vegetation is very close to the window                                   (within 2m) the rate of collision is lower compared to when the vegetation is farther away                               (above 2m) from the glasses. A Poisson Regression analysis was used to fit on the data points of glazing percentage and                               total collision counts at each facade. From data points we found out that there is a weak                                 12 positive correlation of 0.19 between the glazing percentage of facades and the collision                         counts. The estimated regression model is:  expected # of collision= e^(­0.0925)+e^(0.0113*PercentGlazing)  For every percent increase in the glazing the collision increased by the rate e^0.0113 and it is                                 statistically significant with a p­value of 0.0294. However the intercept value,                     e^(­0.0925)=0.9116, is not significant with p­value of 0.80.   In addition to the survey data, 2 emails had been received from our citizen science collection                               method. One reported witnessing a collision at Irving Library and the other at the Earth                             Sciences building (ESB).  Reports of bird carcass at buildings that are not in our study were also received during the                                 study period: Cooper’s Hawk being found at UBC Botanical Gardens, Varied Thrush at Liu                           Centre and Red­breasted Sapsuker at West Mall Annex.    Discussion As shown in the results section, five of the ten buildings we targeted in our study had more                                   than five incidences of bird­window collisions (strikes), which was our metric for determining                         whether a problem was present at a building. In terms of total strikes over our survey period,                                 the foremost was the Museum of Anthropology, followed by Irving Library, Beaty Museum,                         ESB, and  Totem residence hәm'lәsәm' House.  Irving (max. 85% glazing), Beaty (max. 95 % glazing), ESB (max. 85% glazing), and                           hәm'lәsәm' House (max. 60% glazing) all have large panes of glass in their windows                           exceeding at least 60% glass cover, along with other aggravating factors.  Note: max. % glazing indicates the maximum value of glazing cover percentage out of all 4 facades surveyed for each building.  The buildings have reasonably large forest patches (5­10 trees with shrubs) nearby which are                           attractive to birds. We also noticed (Figure 4) that buildings closer to the forest along Wreck                               Beach had higher incidences of strikes.The forest and forest patches are attractive because                         they provide food and shelter. The presence of these patches increases the number of birds                             present in the area, and thus increases the frequency of strikes. We measured distances                           from each building to nearby categories of vegetation. Our categories were shrubs, grass                         13 (lawn), trees (sparse) and forest patches. ESB has a patch to the east side, about 30m away,                                 shared with Beaty. The patch is to the south of Beaty within 10 m distance. Irving has a large                                     patch nearby to the south and west, as well as having trees across the street to the east. For                                     each of these patches of interest, we see (graph II) that the facade has a higher number of                                   strikes than facades that point away from a patch. As well as attracting birds to an area, a                                   patch of trees can also be reflected in the glass of the building. A bird will attempt to fly into                                       the seemingly clear sky, only to strike the window.   The Museum of Anthropology (MoA) is the most dangerous building to birds in our study                             group, and possibly at UBC. It has a maximum % cover glazing of 80 %, and is close to a                                       forest habitat. In our analysis, we have not included strike evidence found during our first                             week, because we are unable to tell the age of the evidence. However, it is interesting to note                                   that if we include our first surveys, we have evidence for the Museum of Anthropology as                               having had over a hundred bird­window collisions. The Museum is a significant problem                         building due to several factors.  First, it is a large building, and so will have more strikes proportionately, but this is not enough                                   to explain the numbers of strikes. The presence of water at the north end, the “Yosef Wosk                                 Reflecting Pool” installed in August 2010 (UBC Project Services, 2010), serves to attract birds                           to the area. There are also trees that some species of bird would use for habitat, also making                                   this area attractive.   The most important hazard is the large, 15m high windows present on the north end of the                                 Museum of Anthropology, facing the water and trees. We know from the literature (Klem,                           1989), (Bayne, 2012), that clear, large windows are hard to see from a bird’s perspective.                             Birds are trying to fly into the large display area of the MoA, and are colliding with the window.                                     Reflections in windows also pose a hazard to birds, as mentioned earlier: a bird can see the                                 reflection in the window, mistake it for open sky, and collide with the window.  In summary, we observed that the distances between the building to the nearby vegetation, in                             which is the most problematic, range from 2 to 10+ m. This is because if the nearby                                 vegetation is too close to the building then it cannot be reflected off the glass windows, and                                 instead blocks the window from view. For vegetation further from the window (2m+), the birds                             are relatively more easily tricked into seeing reflections of the nearby vegetation ­ increasing                           the number of bird collision occurrences as described above. As well, we also notice that %                               glazing cover of at least 60% as in the case of hәm'lәsәm' House, Museum of Anthropology,                               14 ESB, Beaty Museum, and Irving Library indicates the minimum % of glazing that makes the                             birds more prone to strike windows of the buildings.   UBC as an Ecological Trap In ecological trap theory, a trap is a low quality habitat which is preferably selected (Battin,                                 2004). Traps are problems because they can result in a population extinction, as ‘animals                           abandon superior habitats to settle in poor ones” (Battin, 2004). While we lack enough                           concrete data to state whether or not UBC is an ecological trap for birds, we can highlight                                 some areas for future study. More research is required for estimates on the total number of                               bird births and deaths on campus and the surrounding area. This is necessary to determine                             the severity of the ‘trap.’ If birds on campus are dying at only slightly higher rates than in                                   surrounding natural locations, then UBC is not that much a lower quality location than the                             (supposed) higher quality locations around it. This would make local population extinctions far                         less likely. There could also be benefits for birds on campus that result in higher nesting                               success or adult survival rates, excluding UBC from trap status.  Recommendations The important thing is to reduce the hazard of large windows. Mitigation methods are                           available to reduce the invisible clarity of windows already                 present on campus. At the Museum of Anthropology,               measures that could be taken would be the use of wind                     curtains (Figure 5a), window screens or decals (5b) to create                   a design that is unintrusive and gives a visual indicator to                     birds of the window’s presence. Wind curtains are cords hung                   from the top of the window which move with the wind and                       provide a visual reference for birds. Window screens are                 lengths of cloth screen which are installed to hang down                   several inches in front of a window. A bird will collide                     harmlessly with the soft screen instead of the window. If                   either of these methods are chosen to reduce the hazard of                     the windows, a pattern with gaps of only 5 cm vertically or 10                         cm horizontally is required, and be at least 0.32 cm wide.                     Importantly,decals or other markers must be applied to the                 15 outside of windows, or the reflective window will prevent them from being visible during much                             of the day.  We do not recommend the removal of water sources such as the Yosef Wosk Pool. While                               increasing bird presence in the area leads to more collisions, clean water will also support                             more diverse species. Melles (2003) found that some types of birds were three times more                             likely to be present in an area if clean water was available.   In future building planning, the avoidance of wide clear and tinted windows would do much to                               reduce collision numbers. A study by Klem (1989) showed that he “found or collected no                             records of strikes at opaque, translucent, or stained glass windows which present other visual                           effects.” The City of Toronto (2007) released guidelines on other options to provide a visual                             marker to birds. Their recommendation that gaps in patterns and grilles is out of date: they                               recommended gaps of “less than 28cm, with 10cm being optimal.” We now know that gaps of                               5cm vertically, or 10cm horizontally are necessary, and dimensions of 5cm by 5cm are                           needed for smaller birds such as kingfishers and hummingbirds (FLAP Canada, 2015). The                         City’s other suggestions, such as decorative grilles, films like those used for advertising on                           transit vehicles and multiple paned glass, are still examples of useful visual markers to birds.                             Sunshades to reduce reflections and angling the glass to reflect the ground (City of                           Vancouver, 2014) can also be beneficial.   Conclusion All in all, we determined that five of the ten buildings had a bird strike problem at the UBC                                     Vancouver campus: Museum of Anthropology, Irving Library, Beaty Museum, ESB, and The                       Totem hәm'lәsәm' House. In echo with the research questions, the collision rates have been                           fairly steady from November to February, as we did not observe any sudden spikes. We are                               unable to perform a comparison between collision counts in the winter and spring months,                           because it is invalid to relate three winter months of data collection with 1 month of spring                                 survey. As a result, referring to the definition of a bird strike problem with an identification of                                 five or more birds hitting a building, it can be summarized that the suspected bird strike                               problem can no longer be neglected and immediate actions need to be taken towards it.                             Some recommendations that have been proposed to minimize the negative effects of the                         issue are in addition to putting up visual markers, reducing large­sized glass windows, and                           possibly replacing them with tinted or opaque windows. Being the first project of its kind, it                               16 hopefully can add to the University’s plan of consideration when partaking in any                         infrastructure construction in the future.   Acknowledgements This project would not have been nearly as successful without the input of a great many                               people. We’d like to thank Mike Giannias of Building Operations, and Ildiko Szabo of the                             Beaty Museum. We’d also like to thank our very informative community partners Krista De                           Groot, Hannah Brash, and Penny Martyn, as well as our professors Bernardo, Tara, Sara and                             Kari, and our classmates for their feedback.     Team The data for the Bird Strike Study on selected UBC buildings was collected and analyzed by                               dedicated fourth year students majoring in Environmental Sciences. They are Andy Chien,                       Gordon Cavers, Carmen Leung and Tiffany Nam. Andy, Gordon and Carmen are currently in                           the land, air and water area of concentration, while Tiffany is in the ecology and conservation                               area of concentration.        17  References:  Battin, J. (2004). When good animals love bad habitats: ecological traps and the                         conservation of animal populations. Conservation Biology, 18(6), 1482­1491.  Bayne, E. M., Scobie, C. A., & Rawson­Clark, M. (2012). Factors influencing the                         annual risk of bird–window collisions at residential structures in Alberta, Canada. ​Wildlife                       Research​, ​39​(7): 583­592.  Butler, Robert W., Gary W. Kaiser, and G. E. J. Smith.(1987) Migration Chronology,                         Length of Stay, Sex Ratio, and Weight of Western Sandpipers,(Calidris mauri) on the South                           Coast of British Columbia (Cronología de la Migración, Periodo de Estadía, Proporción de                         Sexos y Peso de Calidris mauri, en la Costa sur de la Columbia Británica, Canada.​Journal of                                 Field Ornithology​, ​58​(2):103­111.  Butler, R. W., & Kaiser, G. W. (1995). Migration chronology, sex ratio, and body mass                             of least sandpipers in British Columbia. ​The Wilson Bulletin​, ​107​(3):413­422.  City of Toronto Green Development Standard. (2007). Bird­friendly development                 guidelines. Retrieved from http://www1.toronto.ca/city_of_toronto/city _planning/zoning         __enviro nment/files/pdf/development_guidelines.pdf on March 13th, 2015.   City of Vancouver. (2014). Bird Friendly Design Guidelines. Retrieved from                   http://vancouver.ca/files/cov/bird­friendly­strategy­design­guidelines­draft­2014­09­01.pdf on   March 13, 2015.   C. Sheppard. (2011) Bird­Friendly Building Design. ​American Bird Conservancy​, The                   Plains, VA, 58p  De Groot, K.. (2015) Personal communication.  18 Drever, M. C., Lemon, M. J., Butler, R. W., & Millikin, R. L. (2014). Monitoring                             populations of Western Sandpipers and Pacific Dunlins during northward migration on the                       Fraser River Delta, British Columbia, 1991–2013. ​Journal of Field Ornithology​, ​85​(1), 10­22.  FLAP Canada. (2015). Bird Building Collision Reduction. Retrieved from                 http://www.flap.org/commercial_new.php. on April 11, 2015.   Hager, Stephen B., Craig, Matthew E. (2014) Bird­window collisions in the summer                       breeding season. ​PeerJ​ 2​, e460. Online publication date: 1­Jan­2014.  Klem Jr., Daniel. Farmer,Christopher J. Delacretaz,Nicole. Gelb, Yigal. Saenger,                 Peter G. (2009) Architectural and Landscape Risk Factors Associated with Bird­glass                     Collisions in an Urban Environment. ​The Wilson Journal of Ornithology​, ​121​(1):126­134. DOI: http://dx.doi.org/10.1676/08­068.1  Klem Jr,Daniel. (1989) Bird: Window Collisions. The Wilson Bulletin, Vol. 101, No. 4                         (Dec., 1989), pp. 606­620 Published by: Wilson Ornithological Society   Article Stable URL: http://www.jstor.org.ezproxy.library.ubc.ca/stable/4162790  Klem Jr, D. (2014). Landscape, Legal, and Biodiversity Threats that Windows Pose to                         Birds: A Review of an Important Conservation Issue. ​Land​, ​3​(1): 351­361.  Macek, Michael. (2014). Introduction to Avian Challenges.​International Zoo Yearbook​,                   48(​1): 1­6.  Melles, S., Glenn, S., & Martin, K. (2003). Urban bird diversity and landscape                         complexity: species­environment associations along a multiscale habitat gradient.               Conservation Ecology​, ​7​(1), 5.  Reed, E. T., Cooch, E. G., Cooke, F., & Goudie, R. I. (1998). Migration patterns of                               black brant in Boundary Bay, British Columbia. ​The Journal of wildlife management​,                       62​(4):1522­1532.  UBC Vancouver Construction Update. (2014). Retrieved October 10, 2014 from http://www.infrastructuredevelopment.ubc.ca/infrastructure/projects/construction.htm  19 University of British Columbia Project Services ­ Yosef Wosk Reflecting Pond.                     Retrieved March 13, 2015 from http://www.projectservices.lbs.ubc.ca/portfolio/public­realm/yosef­wosk­reflecting­pond.htm  20 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items