The Effects of Human Age, Group Composition, and Behavior on the Likelihood of Being Injured by Attacking Pumas

Documentation from the years 1890 to 2000 of 185 instances of pumas (Puma concolor) attacking humans in the United States and Canada has provided statistical evidence that pumas are less likely to kill or injure humans in certain circumstances. We identified incidents of fatal attacks, severe injuries, light injuries, and no injuries as a function of human age class, group size, body posture, and conspicuous action, such as noise making, running, or shooting. Ordinal multinomial regression revealed that age class (< 13 years old vs. older) was not a statistically reliable predictor of attack severity. This statistical method also revealed that there was no reliable association between the number of individuals present during the attack and attack severity. Nevertheless, examination of specific attack outcomes indicated that the likelihood of escaping injury increased when two or more people were present. The speed that individuals moved during the attack did not predict attack severity, but it was apparent that the lowest likelihood of escaping injury (26%) and greatest frequency of severe injuries (43%) occurred when individuals remained stationary. In contrast, half of the individuals who ran when they were attacked escaped injury, whereas running was associated with only a small increase in the frequency of fatal attacks (28%), compared with remaining stationary (23%). Evidence that half of the individuals who ran escaped injury suggests that pumas are assessing immobility in humans as they might with other prey, using it as an index of prey inattention or disablement and hence greater vulnerability.