Canine detection for biodiversity protection: a multispecies approach to promote conservation and prevent wildlife trafficking.

The Colombian National Police (CNP) has canine teams specialized for detecting mammals, reptiles, and birds illegally trafficked by national and transnational criminal organizations. The objectives are to: (1) establish standards for the scent database, training protocol, and certification of canine teams for wildlife detection; and (2) determine dog sensitivity and specificity in discriminating between target and decoy scents through a double-blind test during the certification of animal scent detection dogs. The methodology initially involved 22 dogs; a structured screening process, based on behavioral suitability, trainability, and preliminary scent discrimination performance, reduced the cohort to four dogs that met the required operational criteria. These dogs conducted three tests daily for 3 days, to discriminate between target and decoy scents, including horsehair, dog and cat food, pig hooves, beef, and chicken feathers. A double-blind test was performed to assess evaluator agreement (Cohen's Kappa), as well as diagnostic performance, where sensitivity was defined as the proportion of correctly identified true positives (excluding false-negatives) and specificity as the proportion of correctly rejected true negatives (excluding false-positives), with false-positives and false-negatives recorded accordingly. The four dogs achieved near-perfect agreement (0.95-1.00) for the scents of feathers from the Ognorhynchus icterotis and Ara macao; hairs from the Saimiri sciureus; carapace of the Trachemys callirostris; skin from the Boa constrictor; and scales from Crocodylus fuscus and Iguana iguana. False positives for the pig-hoof scent as a decoy occurred on the first evaluation day for two of the dogs, but were not present on subsequent days. The established protocol was successful in training and certifying dogs for wildlife odor detection, with sensitivity and specificity ranging from 0.95 to 1.00 across all tested odors. Additional scents should be incorporated during training to achieve high success rates during certification and subsequent performance. Training materials should be species-specific to avoid contamination with other scents. The established protocol was effective for training and certifying canine teams for wildlife odor detection, demonstrating high levels of agreement, sensitivity, and specificity during controlled testing conditions. These findings support the use of standardized, species-specific scent materials and double-blind evaluation methods in certification processes for wildlife-detection dogs.