Groundwater quality and crop-yield responses to tillage management on a Sparta sand B. Lowery a,* , R.C. Hartwig a , D.E. Stoltenberg b , K.J. Fermanich c , K. McSweeney a a Department of Soil Science, University of Wisconsin-Madison, 1525 Observatory Drive, Madison WI 53706, USA b Department of Agronomy, University of Wisconsin-Madison, Madison, WI 53706, USA c UW Sea Grant Advisory Services, University of Wisconsin Center, Manitowoc, WI 54220, USA Abstract Groundwater beneath sandy soils in the Lower Wisconsin River Valley has more detections and larger concentrations of atrazine and nitrate±nitrogen (NO 3 ±N) than groundwater beneath other areas of sandy soils with similar genesis in Wisconsin, USA. A 2.8-ha ®eld study initiated in 1989 compared water quality and production potential of no-tillage (NT) and moldboard-plow tillage (MB) for corn (Zea mays L.) production with, and without, irrigation. Ridge-tillage (RT) under irrigation for corn and soybean [Glycine max (L.) Merr.] was added in 1991 to test the value of banding herbicide and placing N in the hydrologically inactive ridge. Soybean grown under NT and MB was added in 1993. Atrazine residues and NO 3 ±N were determined in soil and groundwater from 1989 to 1995. Soil samples were taken to a 1.5-m depth and groundwater samples collected from the surface of the aquifer. Atrazine and NO 3 ±N leaching were related to rainfall and/or irrigation. Leaching of atrazine and NO 3 ±N to groundwater below no-irrigation treatment was similar to that under RT with enough irrigation to meet evapotranspiration demands ± atrazine application in RT was only one-third of that used for NT and MB. Atrazine concentrations in soil samples were inconclusive because of limited atrazine detection. Corn yields for RT, MB and NT under irrigation to meet evapotranspiration averaged 8.6, 8.4, and 7.0 Mg ha 1 , respectively. However, there were no statistical differences between RT and MB, primarily because of a different yield rank between years. Soybean yields, averaged over the study period, for MB were 3.07 Mg ha 1 , which is signi®cantly greater than that for NT at 2.07 and RT at 2.19 Mg ha 1 . Atrazine loading from 1991 to 1994 under RT, where it was band applied, was 1.4110 3 kg ha 1 year 1 compared to 3.1710 3 kg ha 1 year 1 for broadcast applications in the MB/NT systems. Nitrate loading on a 4-year average for RT was 23.0 kg ha 1 year 1 compared to 45.2 kg ha 1 year 1 for MB/NT. The RT system, where atrazine was band applied over the top of the row and N was side-dressed, reduced atrazine leaching by 55% and NO 3 ±N leaching by 49% when compared with NT/MB where atrazine was broadcast and N not side-dressed. Selective placement of agrichemicals in hydrologically neutral zones of the RT system reduced leaching to less than one-half of that when placement was not selective. # 1998 Elsevier Science B.V. All rights reserved. Keywords: Ridge tillage; Herbicide; Fertilizer placement; Nitrate; Atrazine; Solute ¯ux 1. Introduction Agriculture now ranks as the primary source of non- point source pollution to streams and lakes in USA and Soil & Tillage Research 48 (1998) 225±237 *Corresponding author. Tel.: +1 608 262-2752; fax: +1 608 265- 2595; e-mail: blowery@facstaff.wisc.edu 0167-1987/98/$ ± see front matter # 1998 Elsevier Science B.V. All rights reserved. PII: S0167-1987(98)00148-2