Translator Disclaimer
25 April 2018 Search strategies for conservation detection dogs
Alistair S. Glen, Clare J. Veltman
Author Affiliations +

Monitoring rare or cryptic species can be challenging, especially with limited time and resources. Dogs are often used for this purpose, but methods are highly variable. There is a need to optimise search methods for dog teams so that time and resources are used as efficiently as possible. Some degree of standardisation is also desirable so that search results are comparable between different times and places. The discipline of search theory has developed effective methods to maximise the probability of detecting a search object and/or maximise the efficiency of a search. However, these advances have not been explicitly applied to the use of dogs to search for plants and animals in the wild. Here, we provide a brief introduction to search theory, then discuss how ideas from search theory might be used to standardise and optimise the use of conservation detection dogs. We describe approaches that have been used, discuss their strengths and weaknesses, and suggest priorities for further research. Standardised methods based on search theory could increase the effectiveness of conservation detection dogs, and make search results more comparable across different locations and times.

A major challenge in wildlife management and conservation is monitoring the abundance and distribution of cryptic species. Many plants and animals can be difficult to detect, either because they are scarce (e.g. invasive species early in an incursion; Hauser et al. 2015), or because they are intrinsically cryptic (e.g. small seedlings; Patten and Milne 2008). Monitoring cryptic species requires methods that are affordable and repeatable (Reed et al. 2011). This means there is a need to optimise search strategies for cryptic species within the limitations of budget, time and human resources. Researchers and managers must decide how many sites to search, which ones, and how much effort to allocate to each (Baxter and Possingham 2011).

Failure to detect an organism does not necessarily mean it is absent, especially when density is low. In deciding what level of search effort to apply, managers have to weigh up the costs of surveillance against the risk of missed detections (Parkes and Nugent 2011).

Dogs are increasingly being used to search for cryptic animals and plants (Long et al. 2007a, b, MacKay et al. 2008, Woollett et al. 2013, Johnen et al. 2017). Searches are conducted for a wide variety of purposes; for example, to enumerate populations, to determine presence/absence of a species, to assist in capturing animals for research, or to locate and remove invasive plants or animals. However, search methods using conservation detection dogs are not standardised, and variation in their performance is not well understood or quantified (Clare et al. 2015a). Our aim was to collate published studies relating to conservation detection dogs (hereafter ‘conservation dogs’), review attributes of search theory relevant to conservation dogs, and synthesise lessons for future application by conservation managers.

Literature search

We searched Google Scholar and the ISI Web of Science database using the keywords: ‘wildlife detect* dog’; ‘scat detect* dog; conservation detect* dog; plant detect* dog; weed detect* dog’, and; (‘search theory’ AND (dog OR canine OR wildlife)). We scanned the resulting titles for references to the use of dogs to find wild plants or animals. We also searched for review articles on search theory in general. Additional articles were located from the reference lists of these publications. Our literature search yielded 152 publications (Supplementary material Appendix 1). There was an apparent taxonomic bias in published studies using conservation dogs; 63 studies related to mammals, 20 to plants, 9 to reptiles, 3 to invertebrates and 1 to birds. An additional 56 publications had no taxonomic focus; these related to methods (e.g. selection and training of dogs) or to search theory in general. A database with details of each publication is provided in Supplementary material Appendix 1.

Search theory

Initially developed by Koopman (1946, 1980) for naval applications, search theory helps determine an optimal search strategy, taking into account environmental conditions and resource constraints. It allows the user to estimate (and maximise) the probability of detecting a search object (e.g. a plant or animal) while explicitly considering characteristics of the search object, the physical environment and the capabilities of the searcher (Frost 2000).

Search theory has applications in such diverse fields as computer science (Geem et al. 2009), mineral exploration (Kolesar 1982), search and rescue (Cooper 2005), animal behaviour (Alpern et al. 2016) and invasive species management (Cacho et al. 2007), among others. Excellent syntheses are provided by Frost (2000), Frost and Stone (2001) and Washburn (2002). Cacho et al. (2007) provide an extremely useful introduction to search theory for ecologists.

Effective sweep width

Search theory allows the user to estimate the probability of detecting an object with a given search strategy and level of effort (Cacho et al. 2006, Perkins and Lovelock 2008). A fundamental concept in this process is ‘effective sweep width’ (R), which is a measure of the search objects detectability (Cooper et al. 2003, Cacho et al. 2007). It is defined as the distance either side of the searcher at which the number of missed detections inside the range equals the number of positive detections outside the range (Koopman 1946, 1980).

To design an optimal search pattern, it is necessary to estimate effective sweep width. This is influenced by characteristics of the search object (how conspicuous is it?), the capabilities of the searcher (e.g. speed, sensory acuity, ability to negotiate rough terrain), and environmental conditions such as weather and vegetation type (Frost 2000, Hauser et al. 2015). Robe and Frost (2002) describe a simple method to estimate effective sweep width experimentally. Search objects are placed at locations known to the observer, but not to the searcher, in an environment similar to that of a real search. Detections and missed detections are then recorded, along with the distance of each search object from the search path. This information is used to generate cumulative detection and non-detection curves. The intersection of the two curves indicates half the effective sweep width (Robe and Frost 2002). To obtain a reliable estimate, a minimum of 250 detection opportunities is recommended. This could be achieved, for example, by 10 searchers searching for 25 objects (Robe and Frost 2002).

Figure 1.

Two examples of lateral range curves, which describe the relati