Black Bear Attack Association and Agency Risk Management

Abstract

Attacks by bears on humans have increased in the United States as both human and bear populations have risen. To mitigate the risk of future attacks, it is prudent to understand past attacks. Information and analyses are available regarding fatal attacks by both black (Ursus americanus) and brown bears (U. arctos), and non-fatal attacks by brown bears. No similar analyses on non-fatal black bear attacks are available. Our study addressed this information gap by analyzing all agency-confirmed, non-fatal attacks by black bears in the 48 conterminous United States from 2000-2017. Government agencies across the country are responsible for species conservation, population management, and conflict control. State, federal, and tribal agencies are required to make decisions that communicate and mitigate the risk of an attack to the public. Agencies have been held legally responsible for those decisions, consuming time and money in litigation. This had led to a call for a more refined way to assess the risk of a bear attack with the creation of a risk management model (RMM). We used an email survey targeted at bear managers throughout the U.S. to 1) identify the primary challenges involved in mitigating the risk of a bear attack, 2) understand their perceptions of managing risk, and 3) assess their support for using an RMM. We further explored these objectives by using a focus group of bear managers at the 5th International Human-Bear Conflict Workshop. We identified 210 attacks by black bears on humans. Most (52%) were defensive, 15% were predatory, and 33% were of a new behavioral category we termed “other.” Of defensive attacks, 85% were by female bears, and 91% of those females had young. Ninety five percent of predatory attacks were by male bears, and 80% of other attacks were by male bears. Sixty-four percent of incidents had an attractant present during the attack, and 74% had prior bear damage or a food reward reported in the area prior to the attack. A classification tree model showed the highest proportion of severe attacks were among a victim with a dog, who was female, and who fought back during an attack. When compared with previous studies of fatal attacks by black bear, our results illustrate clear differences between fatal and non-fatal attacks. In the survey, managers identified human behavior and attractants as the biggest challenge associated with attack risk management. Managers showed moderate support for using a risk management model. Federal managers and less experienced managers were more likely to support a model. Of the three proposed theoretical models (quantitative, qualitative, or mixed) a qualitative model was preferred. The feedback during the focus group supported these results and provided further clarification. These results will help create a future RMM and have implications for risk assessment, attack mitigation, and how we advise the public to respond to an attacking bear.