Background/Question/Methods The behavioral response race and the landscape of fear models predict that across a landscape there should be a negative relationship between the spatial distribution of a predator and its behaviorally active prey.  Additionally, prey have imperfect information on the whereabouts of its predator, which the predator should incorporate in its patch use strategy.  I used a one predator-one prey puma (Puma concolor)/mule deer (Odocoileus hemionus) system to test the following predictions regarding predator/prey distribution and patch use by the predator.  First, pumas will spend more time in high prey risk/low prey use habitat types while deer will spend their time in low risk habitats. Second pumas should select large forage patches more, 3) remain in large patches longer, and 4) visit large patches more often than smaller ones.  I tested these predictions with an extensive telemetry data set collected over 17 years in a study area of patchy forested habitat. 

Results/Conclusions   When active, pumas spent significantly less time (4.3 ± 1.7 (SE) %, n = 9) in low predation risk open areas than deer  (43.0 ± 11.5%, n = 10, t = 3.15, df = 17, P = 0.006). Pumas used large patches more than expected (Χ² = 5946.8 P < 0.001, df = 5).  Pumas visited large patches significantly more often than smaller patches (2.4 ± 0.23 vs 1.2 ± .08 visits/patch, paired t = 4.92, P < 0.001).  Pumas stayed significantly longer in larger patches than smaller ones (2.6 ± 1.27  vs 1.2 ± 0.07 days, paired t = 4.83, P < 0.001).  The results supported the prediction of a negative relationship in the spatial distribution of a predator and its prey and that the predator is incorporating the prey’s imperfect information about its presence.  These results indicate a behavioral complexity on the landscape scale that can have far reaching impacts on predator-prey interactions.