Drivers of Spatiotemporal Variability in Evapotranspiration and Recharge from Irrigated Agroecosystems in the Wisconsin Central Sands.

Groundwater from glacio-fluvial aquifers in the Northern Great Lakes states (Wisconsin, Minnesota, Michigan) supports freshwater ecosystems and anthropogenic land uses, including irrigated agriculture. Irrigation in this setting increases landscape evapotranspiration (ET) and decreases net groundwater recharge, thereby lowering water levels in lakes and wetlands and reducing the flow of streams. In this context, we define net groundwater recharge as the difference between downward (percolation) and upward (pumping) fluxes to and from the water table. The Wisconsin Central Sands (WCS) is one region in conflict over irrigation and drying surface waters, as approximately 335 billion L of groundwater are pumped annually to irrigate 80,000 ha of potato, maize, pea, and snap bean. Successful water management in the WCS requires a nuanced understanding of crop ET and groundwater recharge (deep percolation) to support agriculture while sustaining prized headwater streams, 80+ lakes (> 32 ha), and abundant wetlands. We tested the hypothesis that differences in crop type would drive agroecosystem water budgets in the WCS. To this end, we inferred budgets using 24 passive capillary wick lysimeters and 96 soil moisture probes in six fields on Isherwood Farms in Plover, WI. We quantified ET and recharge (as percolation) on a weekly basis (monthly during deep frost) from November 2013-2014, a wetter than average period. Annual recharge rates were 211 ± 151, 307 ± 214, 443 ± 469, and 481 ± 382 mm for potatoes, peas, sweet corn, and field corn, respectively. Corresponding annual ET rates were 677 ± 151, 580 ± 214, 487 ± 468, and 448 ± 381 mm. The spatiotemporal variability of our results suggests that edaphic and topographic factors are often greater drivers of recharge and ET than crop type. Future work will relate soil electrical conductivity, particle size, topography, and plant phenology to spatiotemporal patterns in recharge and ET.

---