Model helps correct detection errors in helicopter surveys
By Craig Nyhus, Lone Star Outdoor News
This story originally appeared in the April 23 issue of Lone Star Outdoor News.
Helicopter surveys are common on ranches throughout Texas, but the toughest thing to determine is the total amount of deer on the ranch.
Why?
Because of the numbers of deer that go unseen from the helicopter.
Pilots and observers usually estimate the percentage of the total number of deer they are seeing, based on habitat and cover conditions. Estimates too low can result in habitat issues from overpopulation, while estimates too high can result in overharvest.
Dr. Michael Cherry, the Stuart W.
Stedman Chair for White-tailed Deer Research at Caesar Kleberg Wildlife Research Institute, along with colleagues at the University of Georgia Warnell School of Forestry and Natural Resources, are trying to make the estimates more reliable.
The researchers developed a sightability model to correct counts of white-tailed deer observed during surveys, providing a framework for improving estimates derived from aerial surveys.
Surrogates, or 3-dimensional archery targets, were placed at assigned locations in different areas under different conditions, like varying vegetative obstructions, differing light conditions and different distances from the transect flown by the chopper. In 2019, they flew a study area on a large cattle ranch in Central Florida.
The model indicated that distance from the transect (the path the chopper is flying) and vegetative obstruction negatively affected detection of the 3-D targets.
At the study site, detection probability numbers were determined for the different conditions.
Using formulas developed by the model on three different units, population estimates were 26 percent higher than those derived from uncorrected counts.
Observers also recorded live deer during operational flights, and 60 percent of deer groups were observed while stationary.
How do you use this model in Texas?
“We would create a different model using the same approach,” Cherry said. “You could set out 3-D targets in different places and even cluster them together if you have deer hanging in groups. You would put them in open areas and in brushy spots. You have to record in what type of situation (in the open, etc.) you saw each deer.”
After the flight, the data would get reviewed.
“You would analyze the percentage of the surrogates you saw in the different areas,” Cherry said. “Then you could develop more informed count data and come up with a correction of the actual detections you got.”
Cherry also said deer behavior is a significant factor, especially in South Texas when it comes to helicopter flights.
“You have to consider whether the deer are habituated to helicopters,” he said.
While 3-D targets can be used to develop region specific models to account for imperfect detection of white-tailed deer, they aren’t one size fits all.
“They would be very regionally specific,” Cherry said. “Vegetation types change and detectability changes. Models are great on the ranch where you build them. They are pretty good on a similar ranch. They are not so good on a completely different ranch.”
Cherry said helicopter surveys excel in certain areas.
“When it comes to the age structure of the deer and the buck-to-doe ratios, they are really good,” he said. “When it comes to absolute counts of the number of deer on the ranch — that’s what we’re trying to improve.”
