A dog as a generalist plant detection tool. (15th June 2017)

• Record Type: Journal Article
• Authors: Mclean, I G; Sargisson, R J
• Editors: Westerman, Paula
• Is Part Of: Weed research. Volume 57:Number 4(2017)
• Page Start: 287
• Page End: 292
• Publication Date: 2017-06-15
• Subjects: weed management -- dogs -- match‐to‐sample protocol -- detection systems; Weeds -- Control -- Periodicals; Herbicides -- Periodicals; 632.5
• DOI: 10.1111/wre.12258
• Languages: English
• ISSNs: 0043-1737

Abstract

Summary: We tested the training of a dog as a generalist plant detector, to be used when plants are at low density or are difficult to find. Using a match‐to‐sample (MTS) protocol and a 10‐choice array, the dog found 11 plant species with 100% reliability from the first trial, and all 19 species tested with 100% reliability after three trials. The dog reliably found other parts of the same plant species when only one part was offered (e.g. flowers offered and leaves found). The methodology focussed on successful detection of an array of species, but false alarms (FAs) are also an issue when using any detection system. Most FAs given by the dog were of the plant used in the immediately preceding trial (proactive interference), which would not occur in an operational situation. Dogs have the potential to serve as generalist plant detectors, and training with the MTS protocol is a suitable and efficient method for producing such a dog.