Hunters expect wildlife managers to determine how many deer there are in an area, and biologists strive to do so. Common approaches to examine deer population size and trends include harvest data reconstruction, pellet and track counts, and browse, camera, aerial, infrared and thermal-imaging surveys.
By John J. Ozoga,
Deer & Deer Hunting Contributor
Probably one of the most popular methods used for deer population estimation, especially on small private parcels, has been the spotlight survey. This system has been well studied during the past 50 years, with special emphasis placed on determining bias involved in locating individual deer during surveys.
Unfortunately, the precision of just about all deer population surveys tends to be arguable — even among trained deer biologists. Also, hunters usually contend there are fewer deer than survey results indicate, whereas biologists often insist there are probably more deer than casual observations or even detailed surveys suggest.
Despite high variability reported in deer observation (detection) rates, deer managers insist that errors in detection can be corrected, and that road-based deer spotlight surveys provide deer managers with reliable data that can be applied to management decisions.
The spotlight technique takes advantage of light reflected off the tapetums of a deer’s eyes at night. Earliest published data concerning the merits of deer spotlight surveys goes back to the early 1960s, when South Dakota researchers learned that bucks were not readily detected during summer.
One of the more extensive studies of the technique was reported by University of Michigan professor
Dale McCullough, who worked with the enclosed George Reserve deer herd in southern Michigan. He found results quite variable. Seldom were more than 50 percent of deer detected. Bucks were typically under-represented, with highest counts being in July. Fawns were also grossly undercounted and did not approach base counts until 10 months old.
Because dense vegetation was a special problem in southern Michigan, habitat open to good light penetration would probably have given better results. However, McCullough is quick to note that good deer range typically provides habitat diversity, with areas of dense escape cover that would present the same problem.
With regard to spotlight survey precision, McCullough concluded this: “The results demonstrate how seriously data may be biased. Results that seemed reasonable at the time were subsequently recognized as erroneous when the reconstructed population size and composition became known.”
Numerous other studies also have reported that spotlight deer counts are highly variable and frequently hampered by dense cover. However, some researchers suggest that repeated counts provide better data.
The Brosnan Forest Study
Given widespread use and popularity of deer spotlight surveys, researchers led by Bret Collier from Texas A&M University have raised these questions: Can spotlight surveys for whitetails “ … corrected for imperfect detection, provide a value that approximates or approaches a true (yet nearly always unknown) estimate of [deer] population size?”
In other words, are spotlight surveys really a deer management panacea or merely an exercise in futility? That question might be asked of numerous other procedures designed to determine deer density.
The research group led by Collier conducted a five-year study on a 22.5 square-mile forested (93 percent) tract in South Carolina called Brosnan Forest. They used standard nightly spotlight surveys, simultaneously coupled with use of thermal imager data, to tally deer. Two observers — one thermal imager, one spotlight — separated by a partition, independently looked for deer on each side of a truck. There were four observers per vehicle. Surveys were conducted on the same routes from late July to early August, annually, using the same observers.
Researchers identified 4,508 deer during 79 Bronson Forest surveys. Thermal imagers detected 85 percent (3,861 deer) of the total deer seen, and spotlights detected only 48 percent (2,174 deer) of the total. Likewise, of the deer observed, 33 percent were observed by the thermal imagers and spotlights, but 51 percent were only detected by thermal imagers, and 14 percent were only detected by spotlights.
This study demonstrated that the likelihood of detecting an individual whitetail during spotlight surveys is very low, averaging about 41 percent of the time.
Most deer missed by the spotlight were because of distance from the observer or heavy vegetation. Heavy vegetation also hindered thermal imager performance. Even using the same observers during each survey did not increase deer detection rates.
Collier and his co-workers used several models to test their data, but even the best failed to generate meaningful deer density figures, because the probability of deer detection via spotlight varied greatly depending upon the survey, observer, transect and year.
Therefore, it’s difficult, if not impossible, to generate correction factors necessary to evaluate long-term trends in deer population size using the spotlight method.
The researchers concluded this: “Simulations using our variance component analyses indicated that apparent trends in deer density over time are just as likely to be a function of random variability in the observation process as real changes in deer density, and little useful knowledge can be gained by examining or collecting these data.”
Because spotlight surveys are conducted across the United States on private and public lands as a method for deer population monitoring and harvest planning, estimating corrections for each location, observer and time frame are impractical.
However, according to Collier et al., this is not the primary issue involved. Their major concern is that spotlight surveys violate a basic sampling assumption: “That of random placement of transects.” That is, road-based sampling does not allow for random sampling of deer across the landscape.
As a result, those researchers suggest that most estimates of deer population size or trend based on spotlight surveys, regardless of what analytical method correcting for detection is used, “ … exhibit a considerable yet unknown amount of bias which is immeasurable given current sampling limitations.”
Given these research findings, it appears that spotlight surveys to determine deer density are an exercise in futility, not a panacea for deer management. In fact, Collier and his researchers questioned the general usefulness deer surveys that are intended to determine population size or trends.
The justification for use of deer spotlight surveys is that alternative methods are not available or would be too expensive to implement. However, researchers argue “ … that the decades of gathering spotlight data has provided little defensible information for what is likely the most important game animal in the United States.
“Biologically, the continued use of spotlight survey data to monitor deer and base management decisions represents a waste of resources and provides, at best, nothing more than a detection corrected estimate of the minimum number of white-tailed deer seen alive and near the road at the time of survey.”
Final Thoughts — Wow
As mentioned, Collier and his researchers consider spotlight surveys a waste of time and resources, and they also question the value of determining deer population size using current methods at any level.
“Finally, considering their shortcomings,” they wrote, “we believe that the relative value of deer density estimates for aiding in deer management decisions has been grossly overestimated. Rather, we suggest the perceived need to generate estimates of deer density has arisen from external pressure from non-professionals. In short, we as professional biologists are expected to know how many deer exist, because that is perceived to be what we do. Regardless of the accuracy, precision or intrinsic value of these estimates, the perception of the general public is that density estimates are a requisite for good deer management, and many biologists and agencies both knowingly and unknowingly perpetuate this fallacy.”
Wow, what a revelation. These comments are certain to rankle some wildlife managers who routinely estimate deer density, by whatever means, and staunchly defend their data as they attempt to apply it to deer population management at any level.
I applaud and welcome serious reappraisal of archaic procedures of questionable deer management value.