|Borchardt, Mark -|
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: July 7, 2009
Publication Date: N/A
Technical Abstract: Transport of P, N, sediment, and pathogens via runoff from crop fields, especially where manure has been applied, can contribute to degradation of surface waters, leading to eutrophication and potential health effects on humans and livestock. We used a paired-watershed design to evaluate field runoff losses of nutrients and pathogens from different manure/crop/tillage management systems for silage corn production. This design employs field-scale units that reflect the real-world landscape scale better than small conventional replicated plots. This central WI site has four drainage areas or “watersheds” of about 1.5 ha each, each equipped with a 60-cm H-flume, a flow meter, and an automated 24-bottle refrigerated sampler. Samples were analyzed for total P, dissolved reactive P, total N, NO3-N, NH4-N, suspended sediment (SS), and E. coli. Pathogens (protozoa, bacteria, and bovine viruses) were recovered separately using a novel flow-through glass wool filter system. During the 2-yr calibration period all watersheds were treated identically with fall dairy manure application and chisel plowing. That management was maintained as a control in one watershed, while treatments were initiated on the 3 other watersheds in Oct 2008: fall-seeded rye cover crop with spring manure and chisel plowing; fall surface-applied manure with spring chisel plowing; and fall manure and chisel plowing with permanent vegetative buffer strips. During the calibration period, nutrient and sediment loads varied by watershed, but all showed similar patterns over time. Runoff concentrations of TP ranged from 0.1 to 14 mg/L with an average of 3 mg/L. DRP was a small portion of TP for most events, averaging 10-17%, with the exception of 12 snowmelt runoff events for which DRP was 30 to 60% of TP. Linear regressions of runoff and water quality data (TP, DRP, SS) from 48 calibration-period events showed good linear relationships (R2 of 0.75-0.90 for most, lower for DRP), meeting the paired watershed design requirement as a baseline to which treatment period regressions will be compared.