This study investigated alarm calls given by Uinta ground squirrels (Spermophilus armatus) in the presence of a ground predator. I observed predator responses of 18 groups of three to four squirrels each for an average of three trials apiece. r~y objectives were: (1) to describe prey-predator interactions resulting in alarm calls, and (2) to test the following hypotheses: 1. Each Uinta ground squirrel (by sex and age) has an equal probability of giving an alarm call at any time of the season. 2. All Uinta ground squirrels are equally likely to call regardless of their distance to a burrow, closest conspecific, and the predator. 3. Alarm calls are as likely to occur in the search stage of predation as in the pursuit stage. 4. Callers and noncallers are equally vulnerable to predation. I found that: ( 1) each Uinta ground squirrel (by sex and age) in the experimental population had an equal probability of giving an alarm call in the presence of a predator through the season, (2) callers and noncallers were equally close to burrows at the time of the call, (3) the caller was typically located farther away from its closest conspecific than noncallers at the time of the cal l, (4) the caller was significantly closer to the predator than were noncallers at the time of the call, (5) alarm calls occurred significantly more often in the pursuit stage of predation than in the search stage, and (6) noncallers suffered significantly more predation than did callers. There appeared to be little risk and energetic cost associated with calling. Squirrels that called usually were being pursued by the predator and were very close to a burrow when they called. The callers had little to lose and could increase their inclusive fitness by warning relatives of the presence of danger. This study dealt only with responses to ground predators. Squirrels are likely to respond differently to avian predators. It is suggested that responses of animals to avian and terrestrial predators should vary with the potential threat that the predator poses. The apparent inhibition of secondary calls is discussed. Once animals are aware of the presence of danger, there is no need for another animal to repeat the message and reveal its location to the predator.

This research describes and analyzes the behavior of free-living Uinta ground squirrels (Spermophilus armatus) temporally, spatially and with reference to differences in sex, age, population density and habitat. In addition, the relationships between local meteorological parameters and time spent above ground are described and analyzed. The frequency of eight behaviors, feeding, moving, encounters, upright, motionless, grooming, nest gathering and calling, among ground squirrels was found to be labile, varying in frequency with time of day, period of the season, sex, age, population density and habitat, and varying in spatial distribution within the home range. The dominant behavior, feeding, was most frequent during the first and last two hours of the day and tended to occur within the home range core during these hours. During the midday periods home range use became more general and the frequency of the other behaviors increased as feeding frequency decreased. During periods of sexual activity (the beginning and end of the homeothermic season), male ground squirrels displayed high frequencies of encounters, moving and calling behaviors.