SPATIAL DATA SCIENCE FOR PRIVATE LAND CONSERVATION

Abstract

My vegetable garden sits 500 feet west of my grandparents' farmhouse in Cross Plains, Wisconsin, surrounded by mowed lawn, forest, wetland, and—to the south—a hillside of corn and soy. It's a sunny spot, perfect for vegetables and flowers. One of my favorites, milkweed, appeared immediately after my grandparent’s neighbor first tilled the garden plot three years ago, popping up confidently from rhizomes that had survived beneath my grandmother's flower garden. I was immediately attached to how it stood in the cleared soil. I made it clear to anyone who entered: leave the milkweed. Everyone had opinions. The neighbor’s ten-year-old son (our mowing crew) nervously asked if I wanted his dad to till the garden one more time before I planted. My dad mulched around it. My uncle, an organic farmer, suggested I raze everything that wasn't a vegetable. My grandmother told me numerous times that another neighbor had great success covering his garden with cardboard to keep weeds down. Eventually, milkweed popped up all over the garden, unaware of the many opinions on its existence. Standing in the garden facing south, I see the back 40—a forty-acre parcel my grandfather rents to a neighbor. Until 2017, that view was of a hillside of prairie wildflowers and milkweed interspersed with large oaks. My grandfather had enrolled it in the Conservation Reserve Program. The federal government paid him to take the land out of production and plant it with native seeds to prevent erosion and promote biodiversity on his Driftless Area hills and valleys. When his contract expired, he decided he was too old to maintain it and didn’t want whoever took over his property to have to deal with it. He chose to rent the field to a neighboring farmer for a corn and soy rotation. A study from 1997, the year I was born, found that half of the overwintering monarch population in Mexico came from the Midwest, and many used milkweed in agricultural fields to lay their eggs. Now, each year, the closest field to the garden has one crop and no milkweed (Wassenaar & Hobson, 1998). People have differing opinions about the fenced-in garden. Now my partner and I dutifully tend to it, and each character in this story sees it as ours to work. But is it? It's the soil that groundwater flows through. It's the sanctuary that holds queen bees over winter. It's the breeding habitat of monarchs. We are the ones with the power to alter the land, but our choices affect all the creatures who depend on it. While I don't know if one garden plot's scattered milkweed plants have a substantial effect on the durability of the monarch population, I know that whenever I see a monarch in it, I am grateful. The questions raised by my garden aren't unique to Cross Plains. As an undergraduate, I encountered them repeatedly in class projects and internships: in Door County, where absentee owners-controlled shoreline access; along the Great Lakes, where public rights-of-way persisted in narrow road-ends between private properties; in the Lower Wisconsin Riverway, where I helped community members map the tight boundaries between conservation and development. Each project used parcel data to understand who owned land and how ownership shaped access and use. This thesis builds on that foundation, applying those same spatial methods at a broader scale to examine the relationship between land ownership and conservation. Chapter 1 develops a novel parcels-to-properties framework that links parcel-level data to ownership units, enabling consistent analysis across the rural Midwest. Chapter 2 uses this property-level dataset to map and model informal natural land–the unmanaged or semi-managed natural areas that persist outside of formal conservation programs–and to identify the social, economic, and ecological factors associated with their distribution. Instead of mapping individual access points, I examine 7.4 million rural properties. Instead of asking "can the public reach the water?", I ask "where does natural land persist on private land?" The methods are familiar to me—overlaying property boundaries with land cover data, analyzing ownership characteristics, modeling spatial patterns. But the stakes are different. Informal natural land doesn't have the legal protections of public access; it persists only as long as economic conditions, personal values, and land characteristics align to keep it there.