Public Ground, Real Numbers: The Herd Data Buried in Management Plans

Ask a public-land hunter what the deer density is on the state forest they hunt and you will usually get a shrug: the number, they will tell you, simply isn't published. That belief is understandable and wrong. Across nearly 4,900 named state parks, wildlife-management areas, refuges, and federal installations, agencies do write down what they know about their deer: densities, controlled-hunt success, aerial counts. The catch is the form. It lives as sentences inside long management and area plans, not as the clean county tables hunters know how to read. The data isn't missing. It's buried, and the honest finding here is about data shape, not a leaderboard.

Start with the count, because it upends the premise. FDR's public-land layer holds 4,867 named properties (Wisconsin alone contributes over 1,600, with hundreds more each from Minnesota, Missouri, Michigan, and the federal estate) and attached to them are thousands of published deer figures. The problem public-land hunters describe is not that agencies say nothing about deer on these properties. It is that what agencies say is not shaped like the harvest tables they are used to. Illinois, uniquely, publishes clean per-site hunter-effort figures across hundreds of its public areas; most states do not. Elsewhere the deer information is there, but you have to go into the management plan to find it.

Sort the published figures by what they measure and the imbalance is immediate. The layer carries roughly 8,987 effort figures (permits, hunter days, participation) and 6,972 harvest figures (deer taken, by property and season), but only about 2,058 herd figures: the density and abundance numbers hunters actually want. Effort and harvest are easy to publish because they fall out of permit systems and check stations as clean counts. Density is hard: it requires a survey, and when a property does estimate it, the number tends to appear as a sentence in a plan rather than a cell in a table. So the one figure hunters most want is both the rarest and the most awkwardly stored.

The number a public-land hunter most wants (how many deer per square mile) is the rarest one published, and the one most likely to be a sentence instead of a cell.

It is worth showing that usable figures do exist, when a plan states one plainly. Wisconsin's Flambeau River State Forest plan reports forest-wide deer densities ranging from 23 to 31 deer per square mile: a measured range, exactly the kind of number hunters are told does not exist. On the federal side, Minnesota's Camp Ripley military installation runs a long-standing controlled deer hunt and publishes its detail: permits issued, hunter counts near 1,900 per season, and success rates around 15%, year by year. These are not goals or projections; they are observations, and they are quotable precisely because the source stated a clean value. The table below collects a handful.

If the figures are real, why does public ground feel like a data desert? Three reasons, and all three demand caution rather than enthusiasm. First, goal ≠ observation: management plans are full of target densities (the number a property is managing toward) and reading a goal as a measurement is the single most common way to be wrong about a property. FDR excludes flagged goals from any herd figure. Second, the data is prose: a density lives inside a paragraph about oak regeneration or songbird nesting, not in a table, so ordinary search never surfaces it. Third, and most insidious, is the OCR trap: pull hundreds of long PDFs through text extraction and a cross-reference like “Map 2.23” can be captured as if “2.23” were a deer density. That is why the honest move is not to chart the raw herd rows: most of the 2,058 are exactly these prose fragments and artifacts. The finding is not “here is a ranked list of public-land deer densities.” It is “the data exists, in the wrong shape, and it takes real work to separate a measured number from a goal or a mis-read map label.”

The reframe matters because “the data doesn't exist” and “the data is buried” call for opposite responses. Absence is final; a shape problem is an engineering problem. The deer information on public ground is written down across nearly 4,900 properties: it simply has to be found in the plan, verified as an observation rather than a goal, and cleaned of the artifacts that automated extraction introduces. Done carefully, that turns a folklore belief (“nobody knows the herd on public land”) into a bounded task. Done carelessly (charting every extracted number, presenting goals as measurements) it manufactures confident nonsense. FDR's position is the disciplined one: the public-land herd record is real, it is thin and awkward where it matters most, and it earns trust only one hand-verified figure at a time.