Authors: Nathaniel J. Hall, Mallory DeChant, Edgar Aviles Rosa, Paola Tiedemann

Detection dogs are frequently trained in target rich areas but must search operational areas where the probability of finding a target is low. Prior work has established that this can cause decrements in vigilance and detection performance. In this study, we developed a laboratory model of this phenomenon, where dogs worked on automated detection equipment in two nearly identical rooms. One room modeled a “training” environment with a rich rate of targets (90% of searches contained a target). A second room modeled an “operational” environment with a lean rate of targets (10% of searches contained a target). Dogs showed the anticipated drop in search behavior and performance in the operational environment while simultaneously maintaining robust search behavior and performance in the training environment. We then evaluated the use of “free” rewards and motivational stimuli (stimuli previously associated with a reward) to recover performance under operational/infrequent target conditions. No significant effect was found; although, “free” rewards increased search behavior but at the cost of increased false alerts. We then evaluated training dogs with a progressive schedule in the “training” room that slowly decreased target density daily until approximating the “operational” environment. Dogs that received this training significantly outperformed dogs that did not. This finding was then evaluated in a cohort of twenty-three certified operational explosive detection dogs across the US and found that dogs receiving the progressive schedule of reduced targets outperformed dogs receiving a control training, where the target density remained high and constant. The results show that the density of target odors dogs receive in training is an important variable to consider to improve dogs’ resilience when searching operational environments.