Removing one doe group held deer down 5 years in an unhunted herd

STEP 1 · The Idea

A 1991 model put the family-unit window at as many as 10 to 15 years

Porter and colleagues described the pattern in Environmental Management in 1991 and the version we read is the abstract. They call females highly philopatric and say female offspring remain near the dams across life. Home ranges overlap and spread outward like rose petals. Taking out a family unit of deer could ease local conflicts. The model put possible relief at as many as 10 to 15 years and that span is a model result.

Our rule treats that clock as a best case and checks the same August camera sites every year.

STEP 2 · The Test

One Adirondack removal left density lower across 5 years

The experiment sat in the central Adirondacks in New York. Comer and colleagues in the Journal of Wildlife Management describe a test McNulty and colleagues published in 1997 that removed a social group of 14 does. Nine collared deer from neighboring groups were watched and their home ranges did not change.

Oyer and Porter studied this same removal on Huntington Wildlife Forest. Their abstract in that journal says radiomarked deer from the surrounding country did not move into the removal area and density there remained lower for 5 years.

Deer that filled the area afterward were offspring of females still present inside the removal area or immigrants from neighboring social groups. Five years equals a third to a half of those 10 to 15 years allowed in the model.

Comer and colleagues add that this study area had not been hunted since 1932 and that the herd carried a very old age structure. Adirondack deer shift between summer and winter ranges while low recruitment and heavy losses in severe winters keep density low.

STEP 3 · Hunted Herds

Close relatives often used different groups in a hunted South Carolina herd

Comer and colleagues matched DNA relatedness to spatial groups on a 17,300-acre portion (7,000 hectares) of the Savannah River Site in South Carolina. They put radiocollars on 17 does and also used 21 dead does, most of them hunter kills.

The rose petal pattern expects mother and daughter pairs and full siblings to share overlapping home ranges. Only 60% of first-degree related doe pairs sat in the same social group by spatial association and the share for second-degree related pairs was 38%.

Heavy hunting had produced a young herd with average doe age under 2.5 years and with under 4% of does older than 4.5 years. Sitewide in 2002 the hunters took 1,318 deer from an estimated population of 5,500 (24%) and does accounted for 52% of that kill.

Does older than 2 years kept high range fidelity and genetic structure was weak. Higher young doe dispersal could make localized management less useful if vacant ground is recolonized quickly. Comer and colleagues also say the dispersal evidence in hand is not definitive so treat that line as a warning flag.

Comer and colleagues describe Savannah River Site density as low and in the same broad range as Adirondack density. They call harvest history the most striking contrast between the herds. Liberal antlerless harvests in most states may leave many herds with doe ages nearer the Savannah River Site pattern than the old Adirondack pattern.

STEP 4 · Dispersal Rates

These herds show young doe dispersal from 3.6% to 40%

Campbell and colleagues tracked 148 female deer in forested West Virginia and reported in the Wildlife Society Bulletin that 1 of 28 female fawns dispersed (3.6%). Dispersal did not occur once a female was older than a year. Deer density before harvest was estimated at about 31 to 52 per square mile (12 to 20 per square kilometer). The authors put the estimate at three to four times that Adirondack density. Does on the site still met the assumptions behind localized management. They urged experimental removals and did not run one.

Lutz and colleagues radio-marked 277 juvenile female deer in Pennsylvania. The abstract we read says 27 dispersed and our arithmetic puts the share at 9.7%. Dispersal largely fell at 1 year of age during the fawning season.

The same authors compared 12 populations in the Journal of Mammalogy in an abstract we read. Populations holding more deer per forested square kilometer had higher dispersal rates and longer dispersal distances. They call the result consistent with adult does driving subordinate yearlings out while adults seek isolation around fawning.

Comer and colleagues also list an earlier Illinois study where female fawns had 40% dispersal and yearling does had 20%.

A 2004 report by Porter and colleagues covers 59 females radiomarked from 1997 through 2000 in Irondequoit near Rochester. The abstract we read puts average annual dispersal under 15% for both yearlings and adults. At 8% dispersal the model said culling must reduce annual survival to 58% to keep numbers just under carrying capacity and annual survival of 42% to hold the population at half. Measured annual survival there was 64% and most deaths came from car collisions. The authors said movement data and modeling lend weight to localized doe management, though female dispersal could be critical for long-term success.

STEP 5 · Removal Scale

A Pennsylvania model held the thinner pocket reliably only once area reached 3,200 acres

Van Buskirk and colleagues built an agent-based model for Pennsylvania deer and published it in Ecological Modelling in 2021. The abstract we read says localized removal has not always been successful and names high deer density and female dispersal as possible causes. Most scenarios needed at least 5 years before density showed a substantial drop. The best case succeeded at the lowest surrounding density they tested of 30 deer per square mile when the localized antlerless harvest rate was 30% and the removal area was 5 square miles or larger. Under 5 square miles the ending density was highly variable and some scenarios ended above the starting density. At 40 or 50 deer per square mile the antlerless removal rate had to exceed 30% and the reduction needed more than 5 years at every area size.

Five square miles equals 3,200 acres (5 times 640) and our arithmetic makes a 160-acre farm one-twentieth of those acres. Campbell and colleagues estimated that the minimum removal area in their West Virginia study area could approximate 1.3 square kilometers or about 320 acres and the authors tied that size to Adirondack suggestions. Even the Pennsylvania best case is about ten times that patch. Campbell and colleagues sized an area for clearing one whole family group while the model put a 30% antlerless harvest on every acre inside the boundary.

STEP 6 · Field Rule

An August camera index shows whether one removal is still holding

Our read is that removal of a doe family gives no guaranteed stretch of quiet years on ordinary hunting ground. None of the sources we read tracked later doe numbers after a harvest on ordinary hunted ground so no year count is available for your place.

For 43 does on a lightly hunted property in South Carolina our analysis of collar records found median overlap of 91% between pre-rut ranges and post-rut ranges. The file cannot show departures and the overlap fits the Adirondack result that neighboring does kept their ranges after removal. Our read is that refill comes from fawns of does you left and from young does off nearby ground.

A yearly August camera count is our rule of thumb and no study we read tried this count. Start in the August before doe removal so a baseline exists. Use the same sites and the same three weeks each year and log photos of does per camera night plus the doe kill from the prior fall. Oyer and Porter examined relative abundance and the camera index is a rough hunter version of that. August keeps the sample outside the fawning season. In the Pennsylvania juvenile sample 52% of 25 GPS-collared deer made forays outside the natal range during the dispersal period and those forays did not result in dispersal. Mast, crop rotation or a hard winter can also move the index for one year.

When the camera index matches its earlier level across two straight Augusts a one-pass removal is failing to hold. The Pennsylvania model then points to a larger removal area and neighboring farms should remove does too. If the index remains lower across two straight Augusts the removal is still holding at the moment.

Our yearling buck working paper puts juvenile male dispersal from 46 percent to 80 percent across ten populations. Those male rates sit above every doe rate here so the August index says little about young bucks passing through.