Soil samples from a range of land cover classes were collected during the summer of
2015 within the city limits of Madison, WI (Ziter and Turner 2018). Twenty sites were
selected from each of five land cover classes: urban forest, urban grassland, open
space, low-density, medium-density (represented as F, G, O, L, and M, respectively, Fig.
1). These land cover types encompass areas such as city parks (open space), residential
backyards, highway roadsides, and prairies. Site selection and soil characteristics are
described in Ziter and Turner (2018). In Dane County, soils are mainly sandy-loam
glacial till, and in Madison, soils are predominantly Alfisols, with some Mollisols
being found in forest and oak–savanna vegetation and wet sedge meadows (Bockheim and
Hartemink 2017). In 2016, 20 road-terrace sites were sampled (represented as R). Sites
were distributed throughout the city and chosen based on accessibility to terraces along
major roads.
For all land types except road terraces, samples were collected from within a 30 m
x 30 m area in each of the five land types. At each site, four separate 5m x 5m plots
were developed in each quadrant, from which samples were taken. Four samples in total
were taken at each 30 m x 30 m site (n = 400, 100 sites x 4 plots). At road terrace
sites, samples were taken along a 40m transect at four 15 m intervals. Transects ran
parallel to the road within 3 feet of the curb, or at the center of the road median.
Soil from one forest plot was not measured for EC, and therefore total measured samples
equal 379. In both sampling years, soil samples were taken as a 10 cm deep soil core.
Samples were air dried for 24-72 hours, then put in a drying oven for 24 hours at
60 degC. Once dry, samples were sieved with a 200 µm mesh filter to remove gravel and
other large non-soil particles. If large solidified clumps had formed during the drying
process, they were broken up and re-sieved. A homogenized 5-gram sub-sample was then
taken from each replicate. To each sub-sample, 10 mL of deionized water was added in a
30 mL HDPE plastic vial. The mixture was allowed to saturate for at least 15 minutes,
with frequent re-suspension of the settled solids. Electrical conductivity (EC) of the
mixture was then measured with a Thermo Scientific Orion Star A222 portable conductivity
meter. A 1-point calibration was applied at the beginning of each day using a 500 µS/cm
standard. Bulk densities and hydraulic conductivities for F, G, O, L, and M sites were
obtained from Ziter and Turner (2018), but were not measured for road-terrace sites
(https://datadryad.org/resource/doi:10.5061/dryad.5pr17).
High-resolution impervious surface layers were not available for the city of
Madison, Wisconsin. To calculate the likelihood of a soil being impacted by road salts,
we manually digitized impervious surfaces (roads, sidewalks, driveways, and parking
lots) near sampling sites using a high-resolution Google Earth image overlain with city
GIS data to ensure correct georectification. We then calculated the distance from each
sampling plot to the nearest impervious surface using the sf package in R (Pebesma
2018). Given the error associated with GPS locations, any distance less than 1 m was set
equal to 1 m.
Bockheim, J. G., and A. E. Hartemink. 2017. Current and Future Soil Research. Pages
223–228. Springer, Cham.
Ziter, C., and M. Turner. 2018. Current and historical land use influence
soil-based ecosystem services in an urban landscape. Ecological Applications 28:643–654.
