Research Paper

Distribution characteristics of the free-roaming dog population of the western area of Tongatapu, Kingdom of Tonga: Developing a GIS for biosecurity and animal welfare projects

Glenn D. Aguilar*, Vanya de Felice

Affiliation:
School of Environmental and Animal Sciences, Unitec | Te Pūkenga, Private Bag 92025, Victoria
Street West, Auckland 1010, New Zealand.
* Corresponding author: gaguilar@unitec.ac.nz

Received: 20 August 2023 | Accepted: 15 November 2023 | Published: 1 December 2023
Senior Editor: Dan Blanchon
https://doi.org/10.34074/pibs.00804

Download PDF

Abstract:
The free-roaming dog population of Tongatapu, Kingdom of Tonga, is a ubiquitous, well-known and persistent feature of the island. Information on the quantity, condition and spatial distribution characteristics of these dogs is needed for projects aimed at improving biosecurity conditions in different impacted sectors of society, while addressing the management of an animal associated with humans. Health and safety of the human population, agriculture and biodiversity are directly affected by the presence of free-roaming dogs in Tongan communities. A survey conducted over western Tongatapu recorded locations of free-roaming dogs sighted when driving along the roads of 12 towns and villages. Locations were recorded using an iPad and the mobile application Collector for ArcGIS. A total of 1,152 dogs was recorded, mostly within villages and residential roads. Patterns of spatial distribution of dog locations were described using kernel density, hotspot analysis, and cluster and outlier analysis tools available in ArcGIS Pro. Using the species distribution modelling algorithms Maxent and random forest (RF), the suitability of the entire island for free-roaming dogs was determined using as input the dog locations and environmental variables consisting of population, land use and road types. Results of both algorithms show that roads and residential areas consistently show the highest probability of free-roaming dog presence. The RF algorithm also produced maps that show predictions of the values of other variables collected during the survey, including body-condition scores, age group and sex types over the entire island. Overall, the maps produced and the database created contribute a baseline
biosecurity knowledge-base supporting rational decision-making and governance for the island.