Dataset DOI: 10.5061/dryad.8pk0p2npj

Description of the data and file structure

The purpose of this data was to measure the cost and the efficacy of monitoring the pre- and post-bait abundance of feral cats via camera-traps or track counts.

Camera-Trap Monitoring

Two camera-trap arrays were used to monitor feral cat activity on Matuwa pre- and post-baiting. In 2018–2020, 120 camera-traps (Reconyx Hyperfire PC900 Professional camera; Reconyx, WI, U.S.A.) were installed using a stratified-random design based on the 20 most common geological types in the Wiluna region. The cameras were placed between 30 m and 200 m away from an ungazetted track. Camera-traps were, on average, 2.80 km from their nearest neighbour (min = 1.0 km, max = 5.9 km). Data from 70 of these cameras that were placed in an aerial baited site in 2018 and 2019 were used in subsequent analysis. In 2020 and 2021, the entirety of Matuwa was aerial baited, and data from all 120 cameras were used. In 2020–2021, a grid of 130 camera-traps with cameras spaced 1 km apart was installed on Bullimore sandplain in the south of Matuwa.

All cameras were mounted on a 30 cm high plastic sand peg, in a horizontal position, facing south, in a space with at least 3 m of open ground in front of the camera. Two olfactory lures (Canines-a-plenty and Catastrophic from Outfoxed Pest Control, Victoria, Australia) were placed on two natural sticks approximately 30 cm tall and 1 m apart, 3 m from the front of the camera, and refreshed at least 10 days before and after management. Herbaceous vegetation immediately in front of the camera was removed. Camera-traps captured three photos per trigger, with no quiet period. Timed photos were taken at 23:00 h to monitor the operation of the camera.  

We used data from the first 10 days after camera-traps and lures were set, pre- and post-baiting. Photos were stored in the Colorado Parks and Wildlife Photo Warehouse database (CPW) and viewed by at least two observers to confirm species identification. A histogram of time intervals between consecutive photos revealed that 99.6% of photos were captured either <5 min apart or >60 min apart. To minimise temporal autocorrelation, we grouped consecutive photos that were <5 min apart to create independent records for subsequent analysis. 

Track Counts

Track counts collect data that reflects the activity of feral cats on unsealed roads, referred to as the track activity index (TAI). Approximately two weeks pre- and post-baiting, two teams of experienced observers ran a single TAI transect at least 50 km in length each day for four consecutive days. Teams alternated transects each day to reduce observer bias.

TAI-transects occur on sandy 4WD tracks, which are initially cleared by towing a heavy iron drag behind a 4WD vehicle. Observers, driving all-terrain vehicles (ATVs) at a speed of 10–15 km/h, then inspect the transect for cat tracks, and clear new signs of animal activity by towing a chain iron drag. Cat tracks that occur within 1 km radius of one another on a daily survey are aggregated into one cat detection to minimise spatial autocorrelation. A histogram of distances between consecutive cat tracks revealed that 92.3% of tracks were <1 km apart and usually caused by cats that travelled along roads, leaving a continuous set of prints. The single 50 km transect is split at disused wells and intersections, with any observations recorded within 1 km of the well or intersection being discarded. The number of cats observed on each TAI-transect is scaled against the total length of the TAI-transect within each day and then averaged across sequential survey days. 

Files and variables

File: Camera_vs_TAI_2018-2021.xlsx

Description: Data analysed to compare the cost and efficacy of monitoring feral cats via camera traps or the track activity index.

Variables

• method: Fine scale method differentiating camera layout patterns, categorical
• Method: Coarse scale method listing only cameras vs TAI, categorical
• survey: survey period defined as pre-bait or post-bait period, categorical
• Survey: Label for graphing variables, categorical
• Year: Year, integer
• Treatment: Method used to apply management technique, aerial baiting or ground based baiting, categorical
• Day: Number of days since beginning of the survey period, integer
• LocationName: Name of transect or camera site, categorical
• Landsystem: Rangeland landsystem types used to describe a distinct combination of geology, drainage, elevation, slope, and vegetation, categorical
• LS_COMBO: Similar landsystem types were combined to reduce diversity in the covariate, categorical
• Geology: Geology present at a given camera site as per Geological survey of Western Australia, categorical
• DistanceBoundary: Distance between a camera trap and the boundary of the proprty in meters
• DistanceTrack: Distance between a camera trap and the nearest unsealed track in meters
• Scaled_cat_detections: Number of cats detected using track activity index on a given transect if scaled to 100km transect
• Integer: Integer of the number of cats detected using track activity index on a given transect if scaled to 100km transect