Fig 1.
Localization of the three study areas (Poptún, Sabaneta and La Romana) in Petén department, Guatemala.
Source: d-maps.com.
Fig 2.
Aerial pictures of the three study areas taken by a Mavic Pro in May-June 2018.
Left: Poptún, Center: Sabaneta, Right: La Romana.
Table 1.
Technical specifies of the Mavic Pro used for the UAV flights.
Table 2.
Characteristics of the UAV flights performed in Petén department, Guatemala in May-June 2018.
Front and side overlap are defined by the overlap area of two consecutive pictures along the flight line (front) and between transect lines (side). In Poptún, only three flight could be realized because of bad weather conditions during the last flight. The difference of flight time between sessions having the same settings could explained by differences in the light condition. The system automatically adjusts flight speed depending on ambient light conditions to avoid motion blur.
Fig 3.
Flow chart of the UAV pictures analysis process.
Fig 4.
Example of pictures excluded from the UAV picture analysis.
Left: blurred picture, Right: picture fully covered by forest canopy.
Fig 5.
Reviewing procedure of picture taken by the UAV in one of the rural sites.
The left picture shows the nine quadrants and the zigzag pattern used to review the picture. The right picture shows the zigzag pattern used to review each quadrant.
Fig 6.
Example of identifiable dogs and doubtful observations.
Left: dog, Right: doubtful observation.
Fig 7.
Picture taken during the foot-patrol transect walks where marked dogs are visible.
Table 3.
Prior information used in the capture-recapture model.
Table 4.
Number of dogs counted during UAV flights in comparison with the number of dogs spotted during the ground transect walks in three study areas of Poptún Municipality, Petén department, Guatemala in May-June 2018.
Table 5.
Percentage of dogs spotted in roads, yards and fields and forest in the UAV pictures and percentage of the study zone covered by each of type of land cover.
The percentage of the study zone covered by each type of land use was calculated using QGIS.
Table 6.
Dog population size estimation given by the foot-patrol transect survey, the human: Dog ratio and the total owned dog census collected during the “One Health Poptún” project.
Fig 8.
Structure of the dog population in the study region, Petén department, Guatemala.
Fig 9.
Study population of the methods used to estimate the size of dog populations.
a. Method using UAV, b. foot-patrol transect survey method and analysis using a CR model, c. dog census and human: dog ratio method. The colorful areas represent the part of the FRDD population covered by each method.