2015-06-182015-05http://hdl.handle.net/2346/62349Predation is a strong selective pressure on prey due to the fact that as prey move around the landscape and perform their daily activities they must also avoid their predators. Predation risk, however, is not constant across a landscape, but varies greatly both spatially and temporally across the landscape. Thus knowledge of true (and instantaneous) risk is rarely perfect. As a consequence, prey must often set baseline levels of anti-predator behavior that are a compromise between the conflicting demands of vigilance and other daily activities. To reduce this uncertainty of risk and the ensuing trade off animals should use information available in the environment. Prey can directly gather information about risk from the environment (via vigilance behaviors), but this is often time-consuming and costly to perform. Information can also be gathered from con- and heterospecifics (i.e., social information), which can reduce time and cost. Information can be gathered from multiple modalities, such as chemical cues (olfaction), visual confirmation of physical actions of con- or heterospecifics, including predators, and vocalizations of predators or alarm calls of con- or heterospecifics (auditory). Information from social sources widens the pool of available information about risk and may be cheaper (in time and energy) than directly sampling the environment. My dissertation focuses on the use of auditory and visual social information cues of predation risk in two different systems. I examined if gray duikers (Sylvicapra grimmia), a small African antelope, respond to the heterospecific alarm calls of bushbuck (Tragelaphus scriptus). To test this I conducted a playback experiment and quantified perceived predation risk through giving-up densities (GUDs) from artificial food patches and three traditional vigilance measures (percent time spent vigilant, average scan rate, and average scan duration), collected from motion sensitive infrared video cameras. I also identified landscape characteristics (microhabitats and distance to vegetation), which may affect gray duiker perceived predation risk. GUDs, percent time spent vigilant and average scan rate were significantly higher during bushbuck alarm call nights than during silent control nights. Duiker perceived predation risk varied across microhabitat, which varies in terms of its efficacy as an escape substrate. Perceived predation risk was higher in fern microhabitats than in grassland habitats, whereas distance to nearest forest patch had no effect. I also examined if two dove species, mourning doves (Zenaida macroura) and Eurasian collared-doves (Streptopelia decaocto), respond with increased perceived predation risk to the visual cues of past predation events (i.e., bird carcasses). I quantified dove perceived predation risk using giving-up densities (GUDs) at artificial food patches where a simulated raptor predation events (i.e., dismembered dove wings), a wooden control object, or no object was present. Results showed that doves had higher perceived predation risk during simulated raptor predation events relative to controls. To my knowledge this is one of the first studies to test if cues of past predation events (i.e., bird carcasses) affect the perceived predation risk of conspecifics. Acoustic social information from predators, conspecifics, and heterospecifics can provide different information about predator identity, state, and location. Con- and heterospecific alarm calls encode similar information about predator identity state and location, but not all eavesdroppers may be able to decode these specifics. Additionally, differences the amount of predator overlap may affect the eavesdropper's response if the eavesdropper is unable to determine specifics about predator identity. Predators rarely vocalize when hunting which may limit the value of relying on predator vocalizations to update perceived predation risk. Thus, the differences in the types of information that can be gleaned from predator vocalizations and con- and heterospecific alarm calls may elicit differences in the strength of anti-predator responses. I conducted a literature review and quantified the differences in the number of significant and non-significant responses of these three different types of social information. Prey species perform significantly fewer antipredatory behaviors in response to predator vocalizations than to conspecific alarm calls. There were no differences in the strength of anti-predatory responses between heterospecific alarm calls and predator vocalizations or between heterospecific and conspecific alarm calls. Information about predation risk is not restricted to a single sensory modality and it would behoove prey to use cues about risk from multiple modalities, when available, to update their level of perceived predation risk. Social prey species have the benefit of gathering information about risk from conspecifics, but this information can also be gathered by and benefit heterospecifics, including those that are non-social and share the same predators. Understanding what sources of information prey use, how they respond, and what environmental factors (i.e., microhabitat, sightlines, escape substrate etc.) influence perceived predation risk will provide a clearer picture of prey behavioral decision-making and space use.application/pdfengPredation RiskEavesdroppingThesisUnrestricted.