Management and Conservation Note Efficacy of Bear Deterrent Spray in Alaska TOM S. SMITH,1 Wildlife Sciences Program, Faculty of Plant and Wildlife Sciences, Brigham Young University, 451 WIDB, Provo, UT 84602, USA STEPHEN HERRERO, Environmental Science Program, Faculty of Environmental Design, University of Calgary, Calgary, AB T2N 1N4, Canada TERRY D. DEBRUYN, United States National Park Service, Alaska Support Office, 240 W 5th Avenue, Anchorage, AK 99501, USA JAMES M. WILDER, Minerals Management Service, 3801 Centerpoint Drive, Suite 500, Anchorage, AK 99503-5823, USA ABSTRACT We present a comprehensive look at a sample of bear spray incidents that occurred in Alaska, USA, from 1985 to 2006. We analyzed 83 bear spray incidents involving brown bears (Ursus arctos 61 cases, 74%), black bears (Ursus americanus 20 cases, 24%), and polar bears (Ursus maritimus 2 cases, 2%). Of the 72 cases where persons sprayed bears to defend themselves, 50 (69%) involved brown bears, 20 (28%) black bears, and 2 (3%) polar bears. Red pepper spray stopped bears��� undesirable behavior 92% of the time when used on brown bears, 90% for black bears, and 100% for polar bears. Of all persons carrying sprays, 98% were uninjured by bears in close-range encounters. All bear- inflicted injuries (n �� 3) associated with defensive spraying involved brown bears and were relatively minor (i.e., no hospitalization required). In 7% (5 of 71) of bear spray incidents, wind was reported to have interfered with spray accuracy, although it reached the bear in all cases. In 14% (10 of 71) of bear spray incidents, users reported the spray having had negative side effects upon themselves, ranging from minor irritation (11%, 8 of 71) to near incapacitation (3%, 2 of 71). Bear spray represents an effective alternative to lethal force and should be considered as an option for personal safety for those recreating and working in bear country. (JOURNAL OF WILDLIFE MANAGEMENT 72(3):640���645 2008) DOI: 10.2193/2006-452 KEY WORDS Alaska, bear deterrent spray, bear���human interactions, black bears, brown bears, polar bears, Ursus americanus, Ursus arctos, Ursus maritimus. Throughout North America, bear���human conflict periodi- cally results in serious, sometimes fatal, injuries to both bears and humans (Herrero 2002). These conflicts between bears and people include negative interactions that are aggressive, defensive, or nuisance in nature (Gore et al. 2006). A few studies have investigated bear���human conflict in North America (Herrero 1970 Middaugh 1987 Herrero and Higgins 1999, 2003 Miller and Tutterow 1999). Miller and Tutterow (1999) reported that brown bear (Ursus arctos synonymous with ������grizzly bear������ and hereafter brown bear) attacks resulted in 2.75 injuries and 0.42 deaths per year in Alaska, USA, from 1986 to 1996. Miller and Chihuly (1987) found that 72% of nonsport brown bear deaths in Alaska were the result of aggressive bear���human interactions. It is likely that some of these bear fatalities could have been avoided had nonlethal deterrents been available. On Alaska���s Kenai Peninsula, the number of brown bears killed in defense of life or property has increased more than 5-fold in recent years and presently exceeds population sustainability (Suring and Del Frate 2002). People rely on a variety of deterrents for protection from bears, including firearms, red pepper sprays, signal flares, incendiary screamers, and an assortment of noise makers (Herrero 2002). Red pepper spray repellants, hereafter bear spray, were initially developed in the 1960s as a defense against aggressive domestic dogs (Miller 2001). The active ingredients in bear spray, capsaicin and related capsaicinoid compounds, produce a nonlethal yet debilitating response, including coughing, sneezing, bronchoconstriction, apnea, retrosternal discomfort, laryngeal paralysis, and temporary blindness (Miller 2001). Miller (1980) tested dog repellent sprays on captive brown bears and found that charging bears were stopped when sprayed in the face. Spraying resulted in swift retreats to the farthest corner of the cage where bears rubbed their eyes and blinked vigorously (Miller 1980). Encouraged by these results, Miller (1980) advocated the development of red pepper spray���based repellents for bear defense. Initial tests of the improved formulation and packaging proved promising, so research trials were conducted involving captive bears (Hunt 1984). Rogers (1984) reported positive results when red pepper spray was used on free- ranging black bears (Ursus americanus). Importantly, none of these studies reported bears responding aggressively when sprayed. Herrero and Higgins (1998) analyzed 66 nonexperimental incidents in which bear spray was used on both wild brown and black bears and found that in aggressive encounters with brown bears bear spray ended the bears��� unwanted behavior in 94% (15 of 16) of incidents. However, in 6 cases the bear continued to act aggressively in 3 of these cases the bear attacked the person spraying. In 88% (14 of 16) of the cases the bear(s) eventually left the area after being sprayed. Results regarding black bears were more variable, but no humans were injured after spray use. Some people have been reluctant to rely on bear spray for protection. We believe several reasons contribute to their reluctance. Chief among these is the notion that bear sprays are too weak to dissuade curious or aggressive bears from approaching people. Additionally, some people believe that wind can easily render sprays ineffective and that wind- driven spray may incapacitate the user. We present data from Alaska bear spray incidents that address these concerns. Additionally, we present bear spray incidents involving polar bears (Ursus maritimus), the first reported in 1 E-mail: tom_smith@byu.edu 640 The Journal of Wildlife Management 72(3)

the literature. Our goal was to provide data regarding the effectiveness of bear spray over a 20-year period. Given the overall lack of evaluation of the efficacy of bear���human conflict interventions, including bear spray, analysis of bear spray effectiveness is needed (Gore et al. 2006). Insight about bear spray efficacy may contribute to more informed decisions regarding its use and reduce human injury and nonsport loss of bears. METHODS We collected bear spray incident records from 1985 to 2006 from state and federal agencies, newspaper accounts, and anecdotally. We included all Alaska records (31) previously analyzed by Herrero and Higgins (1998) so we could present a comprehensive, updated assessment of bear spray incidents from Alaska. Bear spray incident variables of interest included date, time, location of incident, number of persons involved, person���s activity before interaction, bear species and age-sex class, bear���s activity before being sprayed, manufacturer of spray used, wind effects, effects on humans, dosage of spray administered, dosage of spray received, distance to bear when sprayed, bear���s response to spray, mechanical problems, and whether the bear returned after being sprayed. Whenever records were incomplete (n �� 10), we interviewed individuals involved. We regrouped values for the variable distance to bear when sprayed into broader categories to aid analysis (e.g., 0���5 m, 6���10 m, and 11��� 20 m). Subjectivity of incident records, presence of confounding factors (e.g., multiple manufacturer���s products having been used), and small sample sizes limited statistical analyses. We pooled bear spray incident data by bear species and bear behavior, consistent with Herrero and Higgins (1998). Data included incidents involving black, brown, and polar bears. We labeled bears curious if they were exploring the environment in a nonaggressive manner. We deemed bears aggressive when the encounter included behaviors such as charging, agonistic vocalizations, or persistent following (Herrero and Higgins 1998). In some instances, we could not infer the bear���s behavior and we classified those behaviors as unknown. We pooled data by behavior of the bear before being sprayed into 2 categories, food motivated and nonfood motivated, consistent with Herrero and Higgins (1998). Bears in the first category were perceived to be searching for human food or garbage. If aggressiveness was involved in these incidents, it was with respect to acquiring food or garbage. Bears in the second category were acting aggres- sively, and they were not attempting to acquire food or garbage. We defined successful outcomes as bear spray having stopped the undesirable behavior of the bear. A bear that no longer pursues a person, breaks off an attack, abandons attempts to acquire food or garbage, or turns and leaves the area are examples of successful outcomes. We deemed failures spray incidents in which the bear continued its pursuit, persisted in attempts to acquire food or garbage, or showed no change in its undesirable behaviors. A bear not leaving an area after being sprayed, however, was not deemed a failure so long as threatening behaviors, rummaging through trash, or direct risks to people ceased. To address wind effects on spray, we tested the velocity of bear spray issuing from canisters at the actuator, or nozzle, using a Kestrel wind meter (Nielsen-Kellerman, Inc., Sylvan Lake, MI). We held the meter approximately 5 cm from the actuator and released a 1-second burst of spray. We recorded maximum wind speed attained. We replicated this proce- dure 5 times to calculate a mean exit velocity for bear spray. We used the G test for goodness-of-fit for differences between observed and expected frequencies (Dytham 2003). We selected the G test because we were dealing with observed frequencies of various categories and expected proportions for those categories that we did not derive from the data. We set significance at P �� 0.05. RESULTS We analyzed 83 cases involving the use of bear sprays in Alaska (Table 1), of which 72 incidents involved persons spraying menacing bears, and the remainder (n �� 11) are examples of spray misuse or bear attraction to residues. We address instances of bear spray misuse separately. From 1985 to 2006, our sample of bear spray incidents showed that Alaska averaged 3.1 6 0.7 reported bear spray incidents per year. Of the 83 incidents we examined, brown bears were involved in 61 (74%), black bears in 20 (24%), and polar bears in 2 (2% G1 �� 96.6, P , 0.001). Of the 72 cases where persons defensively sprayed bears, 50 (69%) involved brown bears, 20 (28%) black bears, and 2 (3%) polar bears (G1 �� 73.0, P �� 0.000). All instances of spray misuse (n �� 11), or of spray residues attracting bears, involved brown bears. In 92% (46 of 50 G1 �� 41.4, P , 0.001) of close-range encounters with brown bears, spray stopped undesirable behavior in which the bear was engaged. In 90% (18 of 20 G1 �� 14.7, P �� 0.001) of close-range encounters with black bears, spray stopped the bear���s undesirable behavior. All bear-inflicted injuries (n �� 3) involved brown bears and were relatively minor (i.e., no hospitalization required). During 1985���1995, Herrero and Higgins (1998) found bear spray use in Alaska 94% effective overall (30 of 32 incidents G1 �� 31.3, P , 0.001) we found that in the decade following bear spray, efficacy was 90% (36 of 41 cases G1 �� 33.4, P , 0.001). Bear spray incidents for which time of day was known (65%, 47 of 72) show that none occurred between 0100 hours and 0600 hours, 14 (30%) occurred between 0600 hours and 1200 hours, 14 (30%) occurred between 1200 hours and 1800 hours, and 18 (38%) occurred between 1800 hours and 2400 hours only one (2%) occurred between 2400 hours and 0100 hours (Fig. 1). In 96% (69 of 72) of bear spray incidents the person���s activity at the time was reported (Fig. 2). The largest category involved hikers (35%), followed by persons engaged in bear management activities (30%), people at their home or cabin (15%), campers in their tents (9%), Smith et al. Efficacy of Bear Spray 641