What Collar Studies Show Hunting Pressure Does to Whitetail Movement

Collar studies of hunted whitetails disagree about how much deer change when hunters arrive, and the disagreement has an order to it. Where hunters were assigned at a known density or sat over the food deer needed, deer cut movement and took their food at night. Where hunts were short, the season sat on the post-rut, or the deer already lived away from the hunters, the change was small or absent. The papers point to how much of a deer's range hunters can see, more than to how many of them signed in.

The cleanest design is Oklahoma's rifle-season work on adult bucks. On a 1,861-hectare property, researchers followed 37 adult males (2.5 years and older) under three assigned levels of hunting: none, one hunter per 101 hectares, and one hunter per 30 hectares. They measured movement rate by day (06:00 to 18:00) and by night across 36 days (Little et al. 2016).

Movement rate fell over time in every group, including the unhunted control. The slope was similar for the control and the low-risk group and different for the high-risk group against either one, so part of the decline belongs to the calendar and part to the hunters. Relative displacement was highest at the start and three times lower after the season, which the authors read as deer using smaller areas more intensively.

A second 2016 paper from an overlapping research team, built on a controlled study with temporal and spatial risk levels, looked at distance traveled and small-range area over two-day windows (Marantz et al. 2016). Distance, area and exploratory movement all shrank, with the biggest decrease during the 16-day active hunting period. The deer did not expand their small ranges (microranges) and kept their site fidelity to previously established ranges.

Mississippi State's extension bulletin adds two numbers from an Oklahoma study by the university's deer lab, which it does not cite to a paper: by the second weekend of gun season, deer had raised their daytime use of cover 240 percent, and hunter observations of collared bucks known to be in the hunted areas had fallen 62 percent (MSU Extension P3927). The bulletin's authors conclude that hunted bucks stay around and change where and when they are present.

The same bulletin describes GPS collars on adult bucks on private land along Mississippi's Big Black River. It reports that hunters selected stand locations with feeders about two times more than natural vegetation, with summer food plots about five times more and with winter food plots about nine times more. As risk went from low to high, bucks used upland hardwoods, pine, herbaceous cover and bottomland less, and they avoided crop fields on high-risk days. Food plots looked safe in the daily models until day and night were separated. Then the pattern showed: winter food plots were selected five times more at night, and two-thirds of feeder visits fell between 6 p.m. and 6 a.m. The bucks did not change the size or complexity of their daily home ranges. Their most used bedding spots had twice the screening cover of spots in the same home range that they never visited.

A 2023 paper from the same Mississippi State group, whose abstract we read, reports the same mismatch on 42 collared adult males across the 2017 to 2018 and 2018 to 2019 firearm seasons (Henderson et al. 2023). The places hunters selected most were the places deer selected least by day, and deer selected some areas, food plots among them, as much as five times more at night. The bulletin does not cite that paper, so we treat the two as consistent findings from one research group, not as one data set.

South Carolina's Brosnan Forest adds the hunter-visibility side. Stewart and colleagues collared both sexes from 2009 to 2018 on a property where hunters were the main source of adult mortality (Stewart et al. 2022). They captured 111 adults and kept 70 with high fix success, 42 females and 28 males. Effort averaged 5 hours per 100 hectares per week, which the authors say could be considered low risk, and hunters rotated so that only about 10 percent of stands were occupied on a given day.

Food plots covered 6 percent of the study area and 15 percent of the ground around stands, but 23 percent of the ground within 100 meters of a stand where a deer could be seen. Hardwood drains, with acorns and woody browse, made up 4.5 percent of that viewable ground. By day, bucks used hardwood drains at a selection probability of 0.71 and food plots at 0.16. Does used hardwood drains at 0.61 and food plots at 0.41. At night both sexes preferred plots (bucks 0.72, does 0.65). The authors say the design cannot separate fear from an innate sex difference, and they note that hunting pressure was very low and that they had no data from unhunted periods.

Kilgo, Labisky and Fritzen followed 14 radio-collared adult females on Florida's Osceola National Forest from June 1990 to July 1991 (Kilgo et al. 1998; PDF). The equipment was VHF radio, not GPS, and the muzzleloader and general gun seasons were for males only, so the collared does were legal game only during the archery season.

Their activity centers sat about 420 meters from a road during hunting seasons and about 367 meters outside them. During hunting seasons the does' nocturnal activity rate was 63.7 percent against 44.8 percent by day. Outside the seasons the rates were 57.1 and 53.5 percent, with no significant gap. Daily range size did not differ (about 26 hectares against 22). The does avoided clearcuts, young pine and other open habitat and preferred swamp and mature pine. The authors write that hunter activity was concentrated within 200 meters of roads, and that in the dense understory a small shift was probably enough. The study had no unhunted control, and the rut, changing food and stray hunting dogs overlapped the season.

At Chesapeake Farms in Maryland, Karns and colleagues GPS-collared 19 adult bucks and got data through both windows from nine (Karns et al. 2012). The pre-hunt window ran 5 to 24 November 2006 and 5 to 23 November 2007, defined as the breeding period; the hunt window was Maryland's two-week shotgun season. Movement between 20-minute fixes fell from about 120 to 89 meters and activity fell at dawn, day, dusk and night. Home range (306 to 261 hectares, P = 0.090) and core area (71 to 59 hectares, P = 0.160) did not differ statistically. Of 23 encounters within 100 meters of a known hunter, nine produced a turn or flight, and the seven pronounced flights averaged 257 meters (maximum 550). The deer returned to the same places, and their vulnerability to harvest was unchanged (0.08 before, 0.09 during). The authors attribute the drop in movement at least partly to a hunt that sat on the post-breeding period.

In north Georgia, Rosenberger and colleagues watched 20 collared does through seven firearms hunts for antlered deer on two wildlife management areas (Rosenberger et al. 2024). Day and night step lengths did not differ across the pre-hunt, hunt and post-hunt periods. The core area shrank by about a hectare, from 7.0 to 6.0, which the authors judge unlikely to matter. The authors' explanation is that the does already lived where hunters were not: the average relative probability of hunter selection inside a doe's broader use area was 27 percent, and 27 percent of that area held rhododendron or mountain laurel. A separate hunter-GPS study on the same areas found only 5 percent of the management area more likely to be selected by hunters than not, and its authors concluded that vast refuge areas for deer likely existed (Rosenberger et al. 2022).

In southwest Wisconsin, 188 collared bucks tracked hourly through the rut showed no significant effect of hunting season or of firearm-season opening weekend on movement rates (Hunsaker et al. 2025). The firearm season there opened after the peak of the rut, and the authors suggest rut behavior may have overridden any hunting effect. Our guide to sunrise and sunset movement covers that study's within-day findings.

One Alabama study found a response only after splitting the week into single days: daylight movement fell from Saturday to Sunday and stayed low into midweek (Wiskirchen et al. 2022). We break that down in Deer Move Less by Day on Sunday Than Friday.

Each study counted hunters in its own unit, so a ranking by hunter number would mislead. The conversions below are ours; 100 hectares is 1 square kilometer.

Alabama's public-land Saturday peak (0.72) sits below Oklahoma's low assignment, and Alabama deer still cut daytime movement by Sunday. Georgia's check-in totals ran up to 2.9 per 100 hectares, above both, and the does did not change. The designs and units differ, so we cannot rank the responses. The pattern is still informative. Georgia's authors explain their null result by exposure, meaning how much of the deer's range hunters used, and the South Carolina numbers say the same thing: plots were 6 percent of the ground and 23 percent of what a stand could see.

Our reading of the papers is that the pressure a tract puts on deer is the overlap between hunter sight lines and deer cover, more than any headcount. A workable test for a piece of ground:

For the hunter-side map, see our guide to reading public-land pressure by effort, not distance, and for how crowded Illinois public sites are, How Crowded Is That Public Land?. Our movement-score guide explains why a same-day rating cannot carry any of this.

We read the full text of Stewart 2022, Kilgo 1998, Karns 2012, Rosenberger 2024, Hunsaker 2025, Wiskirchen 2022 and the MSU bulletin. Little 2016, Marantz 2016, Henderson 2023 and Rosenberger 2022 were abstract-level reads, so their finer figures are not used here. Kilgo's 200 meters describes one Florida forest, and one farm in Maryland is not a rule for a state forest.

See what Scout maps: /scout.