Initial trail selection in mantrailing dogs under double-blind field conditions: a statistical evaluation of scent discrimination performance.

This study examined, under environmentally realistic yet strictly controlled field conditions, whether mantrailing dogs can reliably discriminate the individual human scent of a target person at the point of initial trail selection. A total of 70 operationally deployed dogs with heterogeneous training backgrounds participated in a randomised, double-blind field study comprising single-trail and scent-discrimination scenarios under standardised start conditions. Both positive and negative scent setups were included. Across all test conditions, directional decision success rates were consistent with chance expectation. All 95% confidence intervals fully overlapped with the chance range predicted by the corresponding binomial model for the respective task. Neither training level, organisational background, breed, age, nor recorded environmental parameters showed a measurable influence on the probability of correct trail selection. Repeated testing did not lead to performance stabilisation; success rates remained within chance expectations across all test rounds. Qualitative behavioural observations indicated frequent reliance on non-olfactory cues, including socially or contextually influenced search decisions, consistent with the presence of condition bias. Teams that appeared successful in individual trials did not demonstrate reproducible discrimination performance above chance upon retesting. Under controlled field conditions, mantrailing dogs did not show a reproducible above-chance ability to discriminate individual human scents during initial trail selection. These findings indicate that observed training or operational successes cannot be assumed to reflect a validated individual-scent discrimination capability and underline the need for rigorously controlled, reproducible study designs when evaluating mantrailing performance in applied contexts.