Research engineer
As a behavior modification specialist, my primary focus is on developing olfactory detection tools designed to help ecologists gather data on elusive animal and plant species. Using animal training techniques grounded in learning principles and predominantly relying on positive reinforcement, one can train macrosmatic animals to behave as mobile, intelligent biological sensors - identifying the particular odor profile of a species or group of species by generalizing information contained in samples from a limited number of different individuals, searching for it using the most appropriate search technique, locating the source of the odor, and translating the olfactory information into a visual output signal through the display of a specific, non-ambiguous trained behaviour.
Domestic dogs are obvious candidates due to their wide availability, appropriate size range for most detection tasks, ease of care, and, for some individuals, strong motivation based on inherent drives. Although an approach relying on living organisms has unavoidable limitations, dogs have excellent capabilities for learning and remembering large numbers of different odors (
Waggoner et al. 2022), and thoroughly trained high-drive dogs can potentially conduct lengthy searches for rare targets with a high level of reliability.
I use adaptive protocols, precise criteria and clear communication methods to train dogs to detect specific biological odors and indicate their perceived sources. The alert behavior that I train is a "passive indication": in response to the conditioned olfactory stimulus, the dog sits with its nose pointing towards the perceived source of the odor and freezes until receiving a new cue. Because this behaviour - thoroughly trained in various contexts until it is stable - is incompatible with touching or interfering in any way with the targets, it ensures that the approach used is minimally invasive.
Depending on the objective and on feasibility concerns (e.g., dog safety, availability of training samples in quality/quantity commensurate with the objective), various types of odor sources can be targeted for each species or group of species - live individuals, dead individuals, or signs of presence (such as faeces, urine, hair, feathers, molts, tracks, nests...).
Past and current target species include amphibians (Lanza's alpine salamander Salamandra lanzai, common spadefoot toad Pelobates fuscus), reptiles (ocellated lizard Timon lepidus, European pond turtle Emys orbicularis, Loggerhead sea turtle Caretta caretta), insects (Crau plain grasshopper Prionotropis rhodanica), birds (calandra lark Melanocorypha calandra), mammals (carnivores) and fungi (black truffle Tuber melanosporum).
Paul Waggoner, Lucia Lazarowski, Bethany Hutchings, Craig Angle, Fay Porritt. Effects of learning an increasing number of odors on olfactory learning, memory and generalization in detection dogs. Applied Animal Behaviour Science, Volume 247, 2022.
Passive indication of a Lanza's alpine salamander, Pian del Re, Italy
HAL CEFE