Fig 1.
Study area map showing locations of five herd management areas (HMAs) across the western United States, in which we evaluated effects of GPS collars on free-roaming horses and burros, 2016–2020.
The Department of Interior, Bureau of Land Management manages each HMA for either horses or burros; HMA polygons are shown in brown along with labels, and red outline shows map area within the United States. HMAs with horses included Adobe Town, Wyoming, and Conger and Frisco, Utah. HMAs with burros were Lake Pleasant, Arizona, and Sinbad, Utah.
Table 1.
Study areas, sample sizes, dates of observations, and mean distance to horses and burros wearing radio collars from 2016 to 2020 in a study measuring potential effects of GPS collars on free-ranging feral equids in 5 BLM herd management areas in the western United States (WY = Wyoming; UT = Utah; AZ = Arizona).
Fig 2.
Proportion of effects observed by study area.
A. Proportion of observations where collars were observed upside down on the animal’s neck (spun); no effects were observed; or some effects were observed. B. Detailed breakdown of effects from panel A, including the proportion of observations of sweaty neck, indented fur, broken fur, chafe, scab, wound, or collar over ears. Observations were collected between 2016 and 2020, in 5 Herd Management Areas (HMAs) in the western United States. Horse HMAs were Adobe Town, Wyoming; Conger, Utah; Frisco, Utah. Burro HMAs were Lake Pleasant, Arizona and Sinbad, Utah. Although horses at Adobe Town HMA appeared to have a higher proportion of wounds and collar-over-ears, those effects were similar in proportion to other herds; the metric was proportion of effects/observation and Adobe Town had fewer observations than other study areas.
Table 2.
Candidate models, number of parameters (K), ΔAICc, Akaike weight (wi), and log-likelihood (LL) to evaluate the role of study area (Adobe Town, Conger, Frisco, Lake Pleasant, or sinbad herd management area), species (horse or burro), distance of observer to collared individual, observation quality score (obs code), body condition (BC), year of study, season (winter, spring, summer, fall), and which side of the animal’s neck was viewed (one side, both sides) on whether any physical effects (i.e., sweaty neck, indented fur, broken fur, chafe, scab, wound, or over the ears) were observed and recorded among feral horses and burros in 5 populations in the western United States between 2016 and 2020, USA.
Table 3.
Observations of the fit of radio collars on the necks of free-roaming feral horses and burros relative to body condition score, in five populations in the western United States, USA.
Data were collected between 2016 and 2020; results are pooled by species. Note that collars could be observed not fitting correctly on one observation, then fitting correctly the next time seen; all individuals were observed at least once with a correctly fitting collar and only a small subset were observed on a few occasions with an incorrect fit. Asterix (*) denotes inability to calculate standard deviation with only 2 observations of 2 horses, one of which had no body condition recorded.
Fig 3.
Box plots of body condition score of horses and burros and fit of telemetry collars from five populations of free-roaming feral equids in the western United States, 2016–2020.
Body condition scores range from 1 (emaciated) to 9 (obese). Black data points represent each observation in which collar fit and body condition of individual was recorded at the same time. Most collars were observed in correct position and tightness on the necks of horses and burros.
Fig 4.
The relationship between duration of wearing a collar and number of effects observed on horses and burros in the western United States, 2016–2020.
Regression lines for horses are shown in black, with lines for burros shown in grey, along with data points (triangles represent effects to fur and dots denote superficial effects), with shaded 95% confidence intervals also shown. The steeper slope for burros is a reflection of increased fur effects (indented fur, broken fur, sweaty neck) over time for burros with longer duration of wear. Superficial effects (chafe, wound, scab) also increased for burros with longer duration of wearing a collar, but the slope was lower than for fur effects. Horses had slight increases in fur effects over longer duration of wear but no increases in superficial effects.
Fig 5.
Frequency of a) feeding, b) moving, and c) standing behaviors between March and September 2016–2020, by burros at Sinbad Herd Management Area (HMA), Lake Pleasant HMA, and horses at Conger HMA, Utah, USA, comparing collared and uncollared individuals. Box plots show median (horizontal line) and interquartile ranges, and all data points are shown jittered to show detailed spread of points.
Table 4.
Number of remote collar removal attempts at each study site (using handheld ultra-high radio signal transmitters, Iridium web service, and timed-release).
Attempts resulted in either a failed attempt, where the collar failed to release after the drop command was sent, or a successful drop where the collar came off the animal. Number of failed and successful attempts are shown, with percentage of all attempts made at that site in parentheses. Note that repeat attempts were made to drop some horse collars. Collars that were not dropped successfully were removed when individuals came in during the gather at the end of the study.