77 SOIL SURVEY HORIZONS S odium-afected soils cover more than 10 million acres in the North- ern Great Plains. Concerns have risen over the interpretation, classiication, and mapping of these soils, especially after tillage and implementation of newer management systems (no-till). Soil sodic- ity varies considerably across landscapes and within soil delineations. As a consequence, soil sodicity is diicult to measure, characterize, and manage. Soil sodicity is quantiied by laboratory measurements of exchangeable sodium percentage (ESP), sodium adsorption ratio (SAR), ield pH measurements, and observations of the physical appearance of soils. Laboratory measurements of ESP and SAR are relatively time- consuming and costly to complete and are therefore limited in number. Surrogate ield measures such as apparent or bulk electrical conductiv- ity (EC a ) derived from electromagnetic induction (EMI) measurements have the potential to improve the mapping and management of SAS. his approach is more expedient and economical than conventional laboratory determinations of ESP, SAR, pH, and electrical conductivity of the satu- rated paste (EC e ) (i.e., Wollenhaupt et al., 1986; Nettleton et al., 1994; Johnston et al., 1997; Lesch et al., 1998). It is generally reported that the EC a of soil is determined by a combi- nation of soluble salt content, clay content and mineralogy, soil water content, and soil temperature (McNeill, 1980). However, the physical con- tact between soil particles also contributes to soil EC a (Corwin and Lesch, 2003). Because of that, soil physical properties that inluence the amount of contact between soil particles, such as bulk density, will also inluence EC a (Jung et al., 2005; Brevik and Fenton, 2004; Corwin and Lesch, 2003; Malicki et al., 1989; Rhoades et al., 1989). Bulk density is inluenced by other properties, including soil aggregation (Brady and Weil, 2008; Wolf and Snyder, 2003), and therefore aggregation ends up inluencing EC a as well. Soil aggregation is inluenced by sodicity (Ruiz-Vera and Wu, 2006), and sodicity is typically measured using SAR or ESP. herefore, there is theoretical reason to believe that measurements of soil sodicity can be cor- related to soil EC a readings, even though sodic soils have low electrical conductivity values in a laboratory-based saturated paste test versus saline Sodium-affected soils (SAS) cover more than 10 million acres in the Northern Great Plains. Improving the classiication, interpretation, and mapping of SAS is a major goal of the USDA-NRCS as Northern Great Plains soil surveys are updated. Apparent electrical conductivity (EC a ) as measured with ground conductivity meters has shown promise for mapping SAS, but use of this geophysical tool needs additional evaluation. This study used an EM-38 MK2-2 meter (Geonics Limited, Mississauga, Ontario), a Trimble AgGPS 114 L-band DGPS (Trimble, Sunnyvale, CA) and the RTmap38MK2 program (Geomar Software, Inc., Mississauga, Ontario) on an Allegro CX ield computer (Juniper Systems, North Logan, UT) to collect, observe, and interpret EC a data in the ield. The EC a map generated on-site was then used to guide collection of soil samples for soil characteriza- tion and to evaluate the inluence of soil properties in SAS on EC a as measured with the EM-38MK2-2. Stochastic models contained in the ESAP software package were used to estimate the sodium adsorption ratio (SAR) and salinity levels from the measured EC a data in 30 cm depth intervals to a depth of 90 cm and for the bulk soil (0–90 cm). This technique showed promise, with meaningful spatial patterns apparent in the EC a data. However, many of the stochastic models used for salinity and SAR for individual depth intervals and for the bulk soil had low R 2 values. At both sites, signiicant variability in soil clay and water contents along with a small number of soil samples taken to calibrate the EC a values to soil properties likely contributed to these low R 2 values. Jeanne Heilig (jeanne.heilig@nd.usda.gov) and John Kempenich (john. kempenich@nd.usda.gov), USDA-NRCS, 2493 4th Ave. West, Rm. C, Dickinson, ND 58601-2623; Jim Doolittle, USDA-NRCS-NSSC, 11 Campus Blvd., Ste. 200, Newtown Square, PA 19073 (jim.doolittle@lin. usda.gov); Eric C. Brevik, Departments of Natural Sciences and Agricul- ture and Technical Studies, 291 Campus Dr., Dickinson State University, Dickinson, ND 58601 (Eric.Brevik@dickinsonstate.edu); Michael Ulmer, USDA-NRCS, 220 East Rosser Ave., P.O. Box 1458, Bismarck, ND 58502-1458 (mike.ulmer@nd.usda.gov). *Corresponding author. doi:10.2136/ssh2011-52-3-2 Published in Soil Surv. Horiz. 52(3):77–88 (2011). Evaluation of Electromagnetic Induction to Characterize and Map Sodium-Affected Soils in the Northern Great Plains Jeanne Heilig, John Kempenich, Jim Doolittle, Eric C. Brevik,* and Michael Ulmer Abbreviations: EC a , apparent electrical conductivity; EC e , saturated past conductivity; EMI, electromagnetic induction; ESP, exchange- able sodium percentage; HDO, horizontal dipole orientation; RSSD, Response Surface Sampling Design; SAR, sodium adsorption ratio; SAS, sodium-affected soils; VDO, vertical dipole orientation. Disclaimer: Trade names or commercial products are given solely for the purpose of providing information on the exact equipment used in this study, and do not imply recommendation or endorsement by the USDA- NRCS or Dickinson State University.