Assessing the sensitivity and efficiency of conservation dogs: a case study on Tasmanian masked owls (Tyto novaehollandiae castanops)

Abstract

Elusive fauna are often understudied, as researchers often struggle to collect sufficient data on such species for rigorous analysis. Such species are often studied indirectly, via the observation of traces they leave behind, or via other non-invasive survey methods such as cameras or acoustic monitoring, but for many species, conservation detection dogs can provide a superior additional method to gather information on cryptic species.

This thesis looks at the standards to which conservation detection dogs are held globally, and then it examines the role conservation dogs might play in collecting information on rare and elusive species via field trials with a trained detection dog, with the Tasmanian masked owl (Tyto novaehollandiae castanops) as the focal species.

Although conservation detection dogs are recognized as an increasingly valuable tool, there are often no clear standards by which their performance is routinely assessed. I review the literature around conservation detection dog testing, with reference to similar testing requirements within other detection dog fields by examining 15 sets of guidelines/standards, used by conservation detection dog practitioners. I classified these guidelines/standards by practitioner type (government, industry, or university), and searched within them for information on i) pre-testing requirements and training guidelines; ii) trial design and implementation guidelines; iii) detection evaluation criteria, and iv) field safety and welfare behaviours. I found that guidelines for conservation detection dogs were scarce, rarely published, and inconsistent in key aspects of content, approach and detail. While blanket mandated standards for conservation detection dogs are not necessarily desirable, there is potential to enhance current standards through collaboration both within and between sectors. Furthermore, developing publicly available, non-mandatory, scientifically designed, flexible standards for working conservation detection dogs was identified as a priority. I recommend that these approaches be applied and adopted by conservation dog practitioners to improve and safeguard works being undertaken in this critically important area of biodiversity conservation.

Like many forest owls, critically endangered Tasmanian masked are both rare and elusive, which has led to data deficiencies which hamper conservation efforts, and makes them an excellent candidate species for exploring the potential for novel survey techniques. Traditional indirect surveys often have low success rates in accurately determining presence/absence of Tasmanian masked owls. Using indirect survey techniques for traces such as owl pellets can provide a more reliable source of data. However, pellets can be difficult to find in the field. I evaluate whether detection dogs can be effective in finding owl pellets in two steps: an experimental trial comparing a dog team to a trained ecologist; and in a pilot field study where the dog team searched potential owl habitat in Tasmanian forests. During the experimental trials, the dog team outperformed the human ecologist. The dog was more sensitive, finding 89% vs. 40% of pellets present, and took half as long to search an equivalent sized area, the combined efficiency of the dog team being 4.5 times greater than that of an ecologist working alone. During the pilot study, searching areas of potential masked owl habitat, no pellets were found. However, when sites were visited where owls were known to have been recently sighted, the dog team were able to find large numbers of owl pellets suitable for collection and potential data extraction. These trials suggest that dogs can find owl pellets with considerably greater sensitivity and efficiency than ecologists working alone, and that they would be best used in combination with targeted surveys, especially when the collection of as many pellets as possible is desirable for DNA extraction or other analysis. Show more