|Bobryk, C - University Of Missouri|
|Myers, David - University Of Missouri|
|Shanahan, John - Dupont Pioneer Hi-Bred|
|Sudduth, Kenneth - Ken|
|Gunzenhauser, Bob - Dupont Pioneer Hi-Bred|
|Gomez Raboteaux, N - Dupont Pioneer Hi-Bred|
Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/2/2016
Publication Date: 4/22/2016
Citation: Bobryk, C.W., Myers, D.B., Kitchen, N.R., Shanahan, J., Sudduth, K.A., Drummond, S.T., Gunzenhauser, B., Gomez Raboteaux, N.N. 2016. Validating a digital soil map with corn yield data for precision agriculture decision support. Agronomy Journal. 108:(3)957-965. doi: 10.2134/agronj2015.0381.
Interpretive Summary: Traditionally, agricultural producers have treated whole fields uniformly when applying agronomic inputs. One common way to improve on this approach has been to use publically available soil survey maps to partition fields into smaller areas that may behave differently and can be the basis of sub-field management. Management that accounts for within-field variation is often referred to as precision agriculture. Recently, newer high-resolution topography information along with advanced digital soil mapping techniques have provided new soil classification systems that need to be evaluated and compared to the older soil maps for precision agriculture applications. The purpose of this investigation was to evaluate a new classification system called “Environmental Response Units” (ERU) by examining how well it accounted for yield variation within farmers’ fields compared to the publically-available USDA Soil Survey Geographic Database (SSURGO). When compared on over 400 farmers’ fields scattered over four US Midwest states ERU accounted for 24% of the yield variation within fields, whereas SSURGO only accounted for 16% of yield variation. More yield variation was captured by ERU than by SSURGO in 86% of the fields. The yield variation calculation we employed was a reasonable method to test the performance of high-resolution digital soil map products. We conclude that the ERU digital soil maps classify similar soil and landscape characteristics in a way that allows improving precision agriculture management applications. Farmers will benefit from this research because it can help them optimize their seed and fertilizer inputs to match production potential within fields. Optimizing these inputs will be more profitable for farmers. Matching input applications to a better-characterized soil resource will also help minimize field losses of agrichemicals, and thereby benefit the general public with cleaner lakes and streams.
Technical Abstract: Variability in soil and landscape characteristics is known to challenge producers in implementing site-specific crop management strategies in precision agriculture (PA). There are growing numbers of digital soil mapping (DSM) procedures that build upon traditional soil survey information by employing high-resolution soil-landscape characteristics; however, reliable methods are needed to evaluate how well new DSM products improve upon traditional soil classifications for greater utility in PA management. This study demonstrates a methodology that quantitatively determines the efficacy of a high-resolution DSM by comparing its ability to capture corn (Zea mays L.) yield variability compared to USDA Soil Survey Geographic Database (SSURGO). An area-weighted variance scoring function (Rv) was employed to calculate reductions in corn yield variance from a new DSM, termed Environmental Response Unit (ERU), and from SSURGO compared to whole-field (WF) variance. Corn yield-map data were collected and corrected for common data collection errors from 409 fields across 4 states within the Midwest U.S. in 2010–2012. Yield variance reduction with SSURGO and ERU increased as WF variance increased; however, ERU yield variance reduction was greater than SSURGO for 86% of the fields. The average Rv across all site-years for SSURGO and ERU was 16 and 25%, respectively, which equated to a 57% higher median yield variance reduction with ERU over SSURGO. The Rv metric, evaluated here with corn yield, can help producers understand the quality of a sub-field classification framework. The ERU is an example of how a higher-resolution DSM may help producers gain confidence in making PA management decisions.