Stand in your woods and it is fair to wonder what was here before. Before the farm, before the first harvest, before the roads. For most of Michigan, we have a surprisingly good answer. And it comes from people who were not thinking about ecology at all.
They were thinking about real estate.
Why anyone surveyed Michigan in the first place
After the Revolutionary War, the young United States was land-rich and cash-poor. The new western territory, Michigan included, was going to be sold off to raise money and settle the country. But you cannot sell what you have not measured. The Land Ordinance of 1785 set up the system still written into every deed today: the Public Land Survey System, a grid of townships six miles square, each divided into thirty-six one-square-mile sections of 640 acres.
Before any of that land could change hands, surveyors had to run the lines. From roughly 1816 to 1856, deputy surveyors for the federal General Land Office walked Michigan section by section, dragging chains through swamp and snow to mark the corners. Their job was boundaries, not botany. But to do that job, they had to write down what they saw.
Witness trees: marking a corner in a trackless forest
Picture setting a property corner in the middle of an unbroken forest, with no GPS, no road to reference, and a wooden post that will rot within a generation. How does the next person ever find it again?
The answer was the trees. At every section and quarter-section corner, the surveyor chose two to four nearby trees, called witness trees or bearing trees, and recorded each one's species, its diameter, and the compass bearing and distance from the corner. Then they blazed the trees so the marks could be found again. If the post was gone, you could still recover the exact corner by triangulating from the witnesses.
As they walked each line between corners, they kept running field notes: the trees the line passed through, the quality of the soil, where it crossed a stream or entered a swamp, where timber gave way to open prairie or oak barrens, where a recent fire or windstorm had been through. None of it was meant for science. It was the paperwork of selling land.
- Witness trees at each corner: species, diameter, bearing, and distance
- Line trees the survey line passed directly through
- Soil quality and the lay of the land
- Water: streams, rivers, lakes, marshes, and swamps
- Prairies, oak openings, and barrens
- Evidence of fire, windthrow, and beaver flooding
- Trails and signs of use
An accidental dataset
Here is the quiet genius of it. Because the corners fall on a regular grid, with a set of witness trees roughly every half mile, in every direction, across the whole state, those records add up to something no one designed: a systematic, located, dated sample of the forest itself. Hundreds of thousands of trees, identified and measured, at a moment before large-scale logging, land clearing, wetland drainage, and a century of fire suppression remade the landscape.
Pair that tree sample with the line notes about the prairies, the swamps, and the oak openings, and you can reconstruct not just which species grew where, but whole plant communities: mesic hardwoods here, oak savanna there, a tamarack swamp in the low ground. In Michigan, ecologists at the Michigan Natural Features Inventory digitized and interpreted these survey records into a statewide map of the vegetation circa 1800 (Comer and colleagues, 1995). That reconstruction is the layer behind our historical vegetation atlas.
The surveyors were drawing property lines. They had no idea they were also taking the last continuous inventory of a forest that had stood for ten thousand years.
What it can tell you about your land
Looked up for a single property, the circa-1800 reconstruction shows the ecological grain of the place: the community your ground most likely belonged to before it was farmed, cut, or drained. That is genuinely useful when you are deciding what to do next.
- A realistic restoration target. Was this oak savanna, mesic forest, prairie, or wetland? Working with a site's history is far easier than working against it.
- An explanation for what you see now: a closed, shady maple woods that the survey shows as open oak savanna is telling you a story about a fire regime that stopped.
- Context for habitat and management choices, from which mast trees to favor to where native seed has the best odds.
Where the record falls short
All of which makes it tempting to treat the map as a photograph of 1800. It is not, and it was never meant to be. A few honest limitations are worth keeping in mind.
- It is a sample, not a census. Surveyors picked witness trees by hand, and people have habits: favoring sound, conveniently sized, easily blazed trees, and sometimes particular species. Those biases are baked into the data.
- The names drift. Surveyors used common names that varied from one person and decade to the next; black oak, yellow oak, linn, and ironwood did not always mean what we would mean today, and some species were lumped together.
- The resolution is coarse. The detail clusters at corners about every half mile. Small wetlands, narrow inclusions, and gradual transitions get smoothed over or missed entirely.
- It is not a single moment. The surveys span forty years, and some districts were already changing by the time a surveyor arrived: cleared, burned, or beaver-flooded.
- Quality varied, and so did honesty. Some surveys were careless, and a few were outright fraudulent, with notes invented from a tavern rather than the field.
- Mapping it means interpreting it. Turning scattered points into colored polygons of plant communities takes modeling and judgment. The boundaries on the map were drawn, not surveyed.
Michigan learned this early. In 1815, the first crews sent to survey the interior reported it as a near-worthless tangle of swamp and sand. That verdict delayed settlement and saddled the state with a swampy reputation for years, until later surveyors walked the same ground and found rich farmland and timber. The notes always carried the eye, and the season, of the person holding the chain.
Why we still use it, and how
None of those caveats make the record less remarkable. It is the only systematic look we have at the landscape before we reshaped it, and for restoration it is one of the most valuable starting points there is. The key word is starting. We read the history, and then we ground-truth it.
Ground-truthing means checking the map against the things that do not lie: soils, slope and drainage, hydrology, and the living evidence still on site, like a few wide-crowned relict oaks that grew in the open, a remnant of prairie plants along an old fencerow, or the species quietly regenerating in the understory.
- Look up the property's likely community or communities circa 1800.
- Compare that to what is growing there today, and note what changed.
- Read the soils, topography, hydrology, and remnant plants on the ground.
- Set a realistic, site-specific target. Not a museum re-creation, but a healthy native community the land can hold.
- Sequence the work: removals, invasive control, seeding, planting, and fire in the right order.
Used that way, a two-hundred-year-old land survey becomes one of the most practical tools a Michigan landowner has. It is evidence to be weighed, not a blueprint to be obeyed.
FAQ
- What is a witness tree?
- A witness tree (or bearing tree) is a tree a surveyor used to mark a survey corner. At each corner they recorded two to four nearby trees by species, diameter, compass bearing, and distance, then blazed them so the corner could be relocated later. Those records are the backbone of the pre-settlement vegetation data.
- How accurate is the circa-1800 vegetation map?
- It is reliable as a general picture of the landscape and the communities present, because it draws on a statewide, systematic sample. It is far less precise at the scale of a single acre. It is a reconstruction built from a sample plus interpretation, so read it as a strong reference, not an exact survey of any one property.
- Does this tell me exactly what was on my property?
- Not exactly. The data points fall about every half mile, and the community boundaries are interpreted from those points. It tells you what your ground most likely belonged to, which is genuinely useful. But it should always be confirmed against soils, topography, and what is actually on the site.
- Can I use it to plan a restoration?
- Yes, as a starting point. The historical community is a sound first guess at a realistic restoration target. The next step is a site visit to ground-truth that target against current conditions before committing to a plan.