This is the first part in a four-part series on the Empire Prairie.

For the past few years, much of my propagation research has been focused on the charismatic flora of the Hauser Street and Westport prairies. These two Dane County remnants have populations of some of my favorite plants: pasqueflowers, puccoons, phloxes, painted-cups, pussytoes—and those are just the P’s.

If you look up these prairies online, you’ll find webpages noting that they are remnants of the Empire Prairie, but the Wisconsin DNR website does little to explain the significance of this ancestral grassland. For years now, the mysteries of the so-called “Empire Prairie” have plagued me. Where were its borders? What about it was imperial? How was it different from the myriad other prairies of southern Wisconsin? And what is left of it?

I’ve got a lot to say about the Empire Prairie, so I’m doing this in multiple parts. This first post will be purely cartographic, focused on the extent of the prairie prior to White settlement. In the next posts, I’ll dive into historical and ecological considerations and discuss what remains of the Empire Prairie today.

The borders of the Empire Prairie at the time of White settlement

In the 1830s—when the land that would later become the State of Wisconsin was largely unsettled by White colonists—contractors of the United States government surveyed the area around the Empire Prairie. They laid out survey lines in a grid with one-mile spacing, and as surveyors walked their lines, they recorded when they entered and exited prairie segments (figure 1), expressed in terms of chains and links (figure 2).

The surveyors then sent their notes and sketches off to the office of the Surveyor General in Cincinnati, Ohio, where cartographers drew up official plat maps, with outlines of the prairie drawn in (figure 3). These outlines are not perfect, since measurements are only exact along the one-mile grid, and everything else is, at best, an educated guess.

Although the plat maps are not perfect, they certainly capture the general shape of the prairie. By digitizing those plat maps, we’re able to reconstruct the boundaries of the Empire Prairie as they were in the 1830s. I stayed as faithful to the original plat maps as possible, but at times I made small adjustments when I found inconsistencies between the surveyors’ notes and the official maps.

While cartographers left us with a nice, complete map of the prairie, they provided less information on what ecosystems surrounded the prairie; forests and savanna, for instance, are not mapped. We do, however, have good records of where wetlands were located, since surveyors meticulously noted when they entered and exited marshes, and these marshes are detailed in the official plat maps. Digitizing all the marshes is exceptionally laborious, so I’ll just show one section:

As for all the land that was neither marsh nor prairie, we have to make some educated guesses. Presumably, everything else had some amount of tree cover, ranging from sparse oak savanna to closed forest. Although surveyors rarely noted when they entered and exited forest or savanna areas, at each one-mile boundary post they would often describe the ecosystem type. Descriptions like “thinly timbered” or “scattered oak” or “scattered timber” all suggest savanna, while “grove” or “forest” suggest, well, forest.

So based on those once-per-mile descriptions, we can build a map of 1830s landcover in the Empire Prairie region with some basic spatial interpolation. There are nearly 6,000 survey points in the area around the Empire Prairie, so it’s lucky for me that somebody has taken all the survey notes and added them to a database. Using that database, we can create a coarse-scale mapping (figure 6).

For about half of the survey points, surveyors did not record a description of the ecosystem they were in. Most likely, these points are savanna. While I cannot be certain of this, there are two compelling lines of evidence.

1. Survey notes on witness trees. At each corner, surveyors would set a post and then describe a few witness trees nearby—along with distances and bearings—in case the post was lost. If no trees were around (as in the prairie), they would raise an earthen mound (figure 7). Environmental historians and ecologists have used the distances to these witness trees as a proxy for ecosystem type. So when witness trees are very close to the survey corner, we can guess that the area had higher tree density, and when witness trees were far from the survey corner, we can guess that tree density was lower.¹ This type of analysis suggests that savanna was far more common than forest in Dane and Columbia counties, so it’s likely that a large percentage of the undescribed corners were savanna.
2. Evidence from mathematical models of pre-settlement vegetation. LANDFIRE, a joint project of the U.S. government and The Nature Conservancy, produces a Biophysical Setting (BPS) map that models vegetation systems prior to White settlement based on physical site characteristics (e.g., soil, slope, aspect, climate) and past disturbance regimes (e.g., fire, flooding). The LANDFIRE BPS map shows an area modeled as prairie, with roughly the same borders as the 1830’s Empire Prairie, surrounded largely by savanna (figure 8).

That’s all I have for today on the historical extent of the prairie. Stay tuned for my musings on the social history and ecology of the Empire Prairie, as well as some discussion and mapping of its remaining fragments! ■

Have a question? If you have questions about Midwestern native plant propagation that you want answered in a future post, I’d love to hear from you! Shoot me an email at plantpropagationproject@gmail.com or leave a comment below!

Want to stay up to date on all my propagation protocols? Subscribe for free to receive new posts and support Practical Seed Ecology:

Want to support this work? Buy me a cup of coffee (or a bag of potting medium):

---

Notes

1. For more information on this type of methodology, see William Tans’ 1976 article “The presettlement vegetation of Columbia County, Wisconsin in the 1830’s” ↩︎