HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (10) Citing Articles via Google Scholar Google Scholar Articles by Fleming, K. L. Articles by Westfall, D. G. Search for Related Content PubMed Articles by Fleming, K. L. Articles by Westfall, D. G. Agricola Articles by Fleming, K. L. Articles by Westfall, D. G. Related Collections Spatial Variability Nutrient Management Production Agriculture Site-Specific Analysis Spatial Distribution

Site-Specific Management

Evaluating Soil Color with Farmer Input and Apparent Soil Electrical Conductivity for Management Zone Delineation

^a Agriliance, LLC, 4545 Wheaton Dr., F190, Fort Collins, CO 80525
^b USDA-ARS-WMU, 2150 Centre Ave., Bldg. D, Suite 320, Fort Collins, CO 80526
^c Dep. of Soil and Crop Science, Colorado State Univ., Fort Collins, CO 80523

^* Corresponding author (klfleming{at}agriliance.com)

Received for publication December 11, 2003.

	ABSTRACT

TOP ABSTRACT INTRODUCTION DEVELOPING MANAGEMENT ZONES MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Variable rate fertilizer application technology (VRT) can provide an opportunity to more efficiently utilize fertilizer inputs; however, accurate prescription maps are essential. Researchers and farmers have understood the value of dividing whole fields into smaller, homogeneous regions or management zones for fertility management. Management zones can be defined as spatially homogeneous subregions within a field that have similar crop input needs. Delineating management zones that characterize the spatial variability within a field may provide effective prescription maps for VRT. The objective of this research was to compare and evaluate management zones developed from soil color (SC) and farmer experience with management zones developed from apparent electrical conductivity (EC_a). These two methods of developing management zones were compared with soil nutrient levels, texture, and crop yields collected on two fields in 1997. The soil and yield parameters followed the trends indicated by both management zone methods at Field 1 with the highest values found in the high productivity zones and the lowest the low productivity zones. Significant differences were found among the management zones. However, at Field 2 the high and medium productivity zones were generally not significantly different using the SC approach, whereas the EC_a approach was effective in identifying three distinct management zones. Both methods of developing management zones seem to be identifying homogeneous subregions within fields.

Abbreviations: EC, electrical conductivity • EC_a , apparent electrical conductivity • GPS, global positioning systems • OLS, ordinary least squares model • SC, soil color • SOM, soil organic matter • SPA, spatial auto regressive model • VRT, variable rate fertilizer application technology

	INTRODUCTION

TOP ABSTRACT INTRODUCTION DEVELOPING MANAGEMENT ZONES MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
ACCURATE PRESCRIPTION MAPS are essential for effective VRT fertilizer application (Sawyer, 1994; Ferguson et al., 1996). Researchers and farmers have understood the value of dividing whole fields into smaller, homogeneous regions or management zones for fertility management. A management zone for VRT application can be defined as a subregion of a field that expresses a homogeneous combination of yield limiting factors for which a single rate of a specific crop input is appropriate (Doerge, 1999). Delineating management zones that characterize the spatial variability within a field may provide effective prescription maps for VRT.

Earlier studies proposed the division of fields by soil type (Carr et al., 1991). Their work indicated that applying different fertilizer treatments to contrasting soils in a field can generate greater returns than applying a field average fertilizer treatment. Mausbach et al. (1993) concluded National Cooperative Soil Survey is not adequate for variable rate application. Soil surveys were not intended for precision farming and tend to map at scale too coarse to be effective. Franzen et al. (2002) compared Order 1 and Order 2 soil surveys with maps from grid sampling and topography based zone maps. Order 2 soil surveys were seldom useful in determining zones for site specific management of NO₃–N. Order 1 soil surveys were more highly related to soil NO₃–N than Order 2 surveys, but their consistency was not as high as when topography based zones or maps from grid sampling were used. Landscape position also has been used to divide fields (Fiez et al., 1994). They found that landscape position alone was not effective in dividing fields into units for variable rate N management. However, Franzen et al. (2001) concluded that topography-based zone soil sampling may be useful in semiarid environments. Long et al. (1998) proposed using the spatial variation in grain protein levels to identify N management zones in spring wheat (Triticum aestivum L.). Currently, the procedure is applicable for dryland fields that are cropped annually to wheat. Practical implementation of this procedure requires that an appropriate sensor be made available to producers that can continuously read the protein concentration of grain from combine harvester. Ostergaard (1997) developed management zones for VRT N application based on soil type, yield, topography, aerial photos, and producer experience. Five fields were divided into 12 to 17 management zones. The zones were soil sampled to determine N rate. They found a $15 to $35 per hectare economic advantage using variable rate N application. Fleming et al. (2001) describe the application of soil color and farmer knowledge to define management zones for variable rate fertilizer application. Initial analysis indicates that this method is effective in identifying different management zones. However, ground truthing is needed to develop accurate VRT maps from the zones.

	DEVELOPING MANAGEMENT ZONES

TOP ABSTRACT INTRODUCTION DEVELOPING MANAGEMENT ZONES MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Method 1: Soil Color with Farmer Input
The spectral properties of bare soil surfaces are largely governed by soil organic matter (SOM) and moisture (Schreier et al., 1988). The gray tone pattern in black and white aerial photographs is often a reflection of these soil properties and may be linked to productivity (McCann et al., 1996). Chen et al. (2000) found that remotely sensed imagery of a bare soil field could be quantified to describe the spatial variation of SOM. Predicted SOM was highly correlated with measured SOM. The technology and methodology were simple and accurate enough to be of practical use in agricultural production fields. Using LANDSAT images, Bhatti et al. (1991) reported SOM distribution highly correlated with the spatial distribution determined by grid sampling the surface soil. In addition, the semivariograms for the two methods of determining SOM were similar. More importantly, the SOM content was significantly correlated to winter wheat yield. These data suggest that SOM determined by aerial photographs or remote sensing may be used to determine site-specific management zones.

Scientists know that the experiences of farmers have been extremely important in the development of agriculture as we know it today (Crookston, 1996). If it were not for farmers' experiential decision-making, most of the modern agriculture that we take for granted today would be unknown. Dr. Crookston has spent several years working with agricultural researchers. It was his observation that the family farm has been a valuable research enterprise, each generation has studied its alternatives and made its decisions. Unfortunately, the potential contributions of farmer knowledge to soil and crop research is not being fully utilized today. Farmers generally know which areas of a field produce high yields and which areas are low in production. It is logical that nutrient needs are different for these areas. This allows identification of different management zones in a field using farmer knowledge as one of the predictors.

Method 2: Apparent Electrical Conductivity
Electrical conductivity (EC) is the ability of a material to transmit (conduct) an electrical current. Apparent electrical conductivity mapping has been useful in locating saline seeps in the northern Great Plains (Halvorson and Rhoades, 1974) and for diagnosing salinity-related problems in the irrigated Southwest (Rhoades and Corwin, 1981). Researchers also have used soil EC to measure or estimate many other chemical and physical properties of nonsaline soils. Williams and Hoey (1987) found that EC strongly correlates to soil texture. In addition to its ability to identify variations in soil texture, electrical conductivity has proven to relate closely to other properties that often determine a field's productivity (Lund et al., 1999). These include water content (Kachanoski et al., 1988), cation exchange capacity and exchangeable Ca and Mg (McBride et al., 1990), depth to claypans (Kitchen et al., 1999), soil organic C (Jaynes, 1996), and herbicide behavior in soil (Jaynes et al., 1994). Sudduth et al. (1995) found strong correlations between EC_a and depth to claypan and yield. Heermann et al. (1999) reported that EC_a was the best predictor of crop yield when compared with many other common soil and crop parameters. Soil classification using EC_a provides an effective basis for delineating interrelated physical, chemical, and biological soil attributes (Johnson et al., 2001). With the advent of global positioning systems (GPS), practitioners can now place EC_a measuring devices on GPS-equipped field vehicles and create EC_a maps for all types of agricultural soils. These maps can be the basis for developing management zones for VRT. The objective of this paper is to compare and evaluate prescription maps developed from soil color and farmer experience with prescription maps developed from EC_a.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION DEVELOPING MANAGEMENT ZONES MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Field studies were conducted in 1997 on two center pivot irrigated fields in northeastern Colorado to assess the technical and economic feasibility of precision farming. Data from these locations will be discussed and used to evaluate the management zones. The area of Field 1 was 71 ha and Field 2 was 52 ha. Soils in both fields included a Valentine sand (sandy, mixed nonacid, mesic Typic Ustipsamments), a Bijou loamy sand (coarse loamy, mixed, mesic Mollic Haplargids), and a Truckton loamy sand (coarse loamy, mixed, mesic Udic Argiustolls) (Soil Survey Staff, 1996). These are nonsaline soils. The fields were in continuous corn (Zea mays L.) throughout the study. Each field has been managed and operated by the same two farmers since 1977. Irrigation strategy was based on experience, no formal scheduling program was used. Fertility management was based on soil sampling and experience. All nutrients have been applied uniformly at these fields.

Aerial 35-mm bare soil photographs obtained from USDA Farm Service Agency flights were used as an initial template in developing the soil color with farmer input management zones. The bare soil photographs were geo-registered and enhanced to contrast color differences using AgriTrak Professional software (Agritrak, 1998). The farmers then drew vector lines on the photographs using AgriTrak Professional based primarily on differences in soil color to establish the individual high, medium, and low productivity management zones. The decision to delineate three management zones was made by the farmers. Their logic was they felt in some areas of the field they had been applying adequate levels of crop inputs for optimum production while in others application rates applied were too high and too low for optimum production, leading to their choice of three management zones. A recent study using a quantitative clustering approach to develop management zones found that maximizing the variability between zones was achieved when the number of zones was three or less (Fridgen et al., 2004). After initially defining the soil color zones on their personal computers, each farmer drove over the fields and made qualitative adjustments in the zone boundaries on laptop computers based on topography changes and their production experience on the fields.

To develop the EC_a management zones, Fields 1 and 2 were mapped using a GPS equipped Veris model 3100 conductivity sensor (Lund and Christy, 1998). The Veris system identifies soil variability by directly sensing soil electrical conductivity. As the cart is pulled through the field, a pair of coulter electrodes transmit an electrical current into the soil, while two other pairs of coulter electrodes measure the voltage drop. The measurement electrodes are configured to measure EC_a over an approximate 0- to 30-cm depth (shallow reading) as well as a 0- to 90-cm depth (deep reading). The system georeferences the EC_a measurements using an external Trimble ag32 DGPS receiver, and stores the resulting data in digital form.

Data were collected in April 1998 on transects spaced approximately 10 m apart on a l-s interval, corresponding to a measurement every 3 to 5 m along the transects. This procedure resulted in a data density of 200 to 300 points per hectare.

Clustering was used to group the EC_a measurements into management zones. The objective of cluster analysis is to statistically minimize the within-group variability while maximizing the among-group variability to produce homogeneous groups that are definitive from one another.

The shallow, deep, and difference between the shallow and deep EC_a point readings were used for the cluster analysis; zones with the highest, intermediate, and lowest EC_a values were defined as high, medium, and low productivity zones, respectively. These data were standardized to a mean of 0 and standard deviation of 1 to bring all data to a standard measurement scale. The analyses were performed using Ward's Minimum Variance method using SAS statistical software (SAS Inst., 1990). This method is an agglomerative hierarchical clustering technique (initial clusters are formed from pairing of individual cases; subsequent clusters are formed from joining pairs of previous clusters and/or cases; clustering continues until all clusters/cases are joined into one cluster). Ward's method, as many clustering methods, is sensitive to outliers. Consequently, the algorithm was instructed to keep 10% of the most "different" cases out of clustering process. The clustered point data was then interpolated using proximity and smoothed using nine nearest neighbor majority in ArcView (ESRI, 1994) to create the final management zones.

Soil sampling was completed in April 1997 on a 76-m grid over both fields. The surface 0 to 0.2 m was analyzed for NO₃, NH₄, P, K, Zn, pH, SOM, and texture. Subsoil samples from 0.3- to 0.6-, 0.6- to 0.9-, and 0.9- to 1.2-m increments were analyzed for NO₃–N.

The fields were yield mapped with two Case IH Axial Flow combine harvesters equipped with Micro-Trak yield monitors. Location data were collected on a 1-s interval using an Ashtech Super C/A receiver GPS system in differential mode. The GPS system was a 12 channel receiver. All the GPS positional data were based on the World Geodetic System of 1984 (WGS84). The GPS data was differentially corrected with a base station set near the experimental field. Yield data were processed and mapped using Farmers Software Harvest Mapping System (Redhen Systems, 1996).

An analysis of variance was performed on the soil and yield data from the management zones using S-PLUS statistical software (Mathsoft, 1995). Initially an ordinary least squares model (OLS) was used; when spatial autocorrelation between observations existed, a spatial auto regressive model (SPA) was used. Spatial autocorrelation was measured using Moran's I (Moran, 1950). The Moran's I is used to test for the presence of spatial autocorrelation in a two-dimensional plane. In this study, the null hypothesis is that the measurements of soil and yield data are independent of one another in space. Moran's I is a dimensionless statistic and can be interpreted as a Pearson product-moment correlation between variables x and y. The index generally ranges over the interval of –1 to +1, but can exceed these limits if an irregular pattern of weights has been used or extreme values are heavily weighted (Bonham and Reich, 1999).

The spatial weight matrix used in the spatial model was based on inverse distance squared. The null hypothesis of no significant differences between management zones was tested by the ANOVA model

where y.. is the grand mean of all observations, a(i) is the effect of the ith management zone, and e(ij) are independent random errors. Equality of means by management zone was tested with a restatement of the ANOVA model in the form of a regression model

where

a1 = 1 if high productivity management zone

a1 = 0 if medium productivity management zone

a1 = –1 if low productivity management zone

a2 = 1 if medium productivity management zone

a2 = 0 if high productivity management zone

a2 = –1 if low productivity management zone

When spatial autocorrelation exists in a data set the assumption of independent observations in a classical model is violated. This can underestimate the standard errors, resulting in increased overall significance between individual parameters or overestimate the standard error resulting in an overall lack of significance when one exists (Bonham and Reich, 1999). All of the soil sample data points were spatially joined to their respective management zones for the analysis of variance. For the spatially dense yield data the points were averaged in 30 by 30 m pixels and then spatially joined to their respective management zones. This was necessary for the spatial analysis in S-PLUS. The spatial algorithms are too complex to analyze very large data sets such as the complete yield and EC_a on personal computers available at this time. Data pixels that fell into more than one management zone were dropped.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION DEVELOPING MANAGEMENT ZONES MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Field 1
When comparing the two methods the management zones and results of the analysis of variance were similar on this field with the following common observations (Fig. 1).

View larger version (33K): [in this window] [in a new window]   Fig. 1. Field 1 management zones–soil color with farmer input management zones on the left and apparent electrical conductivity management zones on the right (low productivity = white, medium productivity = gray, high productivity = black).

These results are similar to the findings of Mulla and Bhatti (1997). They concluded that in terms of practicality, organic matter estimated from bare soil images offer the best feasibility in dividing fields into management zones

The Moran's I for the raw data was significant across all parameters, indicating the data are spatially auto correlated. In general the spatial model had slightly lower standard error and higher r² using both management zone approaches. However, the spatial model was not always significantly better than the OLS. With the SC method the Likelihood ratios were not significant for P, sand, silt, and clay while with the EC_a approach the Likelihood ratios were not significant for NO₃, Zn, and silt. The Moran's I for the residuals of these parameters was also not significant, indicating the management zones are accounting for the spatial autocorrelation in these parameters.

It is interesting how similar the SC management zones that were defined using a qualitative supervised classification with our cooperating farmers are to the EC_a management zones that were defined using a quantitative unsupervised classification. The two methods seem to be capturing similar information in this field.

Field 2
Results of the spatial analysis of variance for the SC method on Field 2 indicated that the soil and crop parameters were different across the management zones; however, they did not follow the trends indicated by the farmer-developed zones. The zone classified as medium in productivity was highest in SOM, K, yield, and clay, while the high productivity zones had intermediate values for these parameters. The low productivity zone had the lowest values for the parameters listed above and were highest in sand (Tables 4, 5, and 6). In addition, no significant difference was detected between the high and medium productivity zones for the soil parameters discussed above. This would seem to indicate this method is only defining two zones in this field.

View larger version (25K): [in this window] [in a new window]   Fig. 2. Field 2 management zones–soil color with farmer input management zones on the left and apparent electrical conductivity management zones on the right (low productivity = white, medium productivity = gray, high productivity = black).

Other studies have also shown that clustering algorithms can be effective in developing management zones. Stafford et al. (1998) used yield maps to regionalize fields into management units. The clustering procedure they used recognized subregions of the fields with distinct patterns of season to season variation in yields. Boydell and McBratney (1999) found that a number of consecutive years of remotely sensed cotton (Gossypium hirsutum L.) yield estimates clustered to generate regions of similarity for management zones. Shatar and McBratney (2001) used a k means cluster analysis on sorghum [Sorghum bicolor (L.) Moench] yields, soil organic C, and soil K to subdivide fields into homogeneous units. The algorithm presented showed some promise for identifying spatially contiguous zones, which are more homogenous in soil properties than the whole field. Yang and Anderson (1996) also had success using clustering to classify management zones. In this study color infrared digital video images of two grain sorghum fields in south Texas were acquired and classified into management zones using unsupervised classification. Their analysis of variance showed that plant biomass and yield differed significantly among their management zones.

The results from the Moran's I analysis on the raw data was similar to Field 1 with all of the parameters being spatially auto correlated using both methods. However, the Moran's I for the residuals of most of the soil parameters was not significant, indicating the management zones for the most part were accounting for the spatial correlation in the data in this field. The exceptions were NO₃ in both zone delineation methods and sand and clay with the EC_a method. The Likelihood ratios were significant for these parameters, indicating the spatial model was significantly better.

	SUMMARY AND CONCLUSIONS

TOP ABSTRACT INTRODUCTION DEVELOPING MANAGEMENT ZONES MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 
Both methods of developing management zones seem to be identifying homogeneous subregions within fields. However, in Field 2 the EC_a approach was more effective in identifying three distinct management zones. The Moran's I analysis indicated the data were spatially auto correlated. On Field 1 the analysis was mixed with certain soil parameters the OLS model was adequate; with others a spatial auto regressive model was needed. On Field 2 the OLS model was adequate; the MZ accounted for the spatial correlation for the most part in this field. This analysis has identified the large scale spatial variability between zones for each method. However, further analysis is needed to evaluate the small scale spatial variability within zones to determine which method is characterizing the fields most accurately. An intense small scale cluster sampling technique within each zone is needed to accomplish this. Based on this initial evaluation it appears EC_a is more consistent in identifying different management zones across fields. Differences in the effectiveness of soil color with farmer input management zones between fields may be due to the selective skills of the different farmers in recognizing zone boundaries. Further testing over a broader scope of fields, farmers, and crop production systems is needed to confirm these results. In the future, methods using the information from both EC_a and soil color along with farmer knowledge may provide the most effective management zones.

	ACKNOWLEDGMENTS

 
Appreciation is expressed to the Agricultural Experiment Station for its financial support of this project. Special thanks goes to Bob Geisick and Larry Rothe, our cooperating farmers in this research. Without their cooperation, this research would not have been possible.

	REFERENCES

TOP ABSTRACT INTRODUCTION DEVELOPING MANAGEMENT ZONES MATERIALS AND METHODS RESULTS AND DISCUSSION SUMMARY AND CONCLUSIONS REFERENCES

 

• Agritrak. 1998. Agritrak training manual. Ver. 2.5. Agritrak, Fort Morgan, CO.
• Bhatti, A.U., D.J. Mulla, and B.E. Frazier. 1991. Estimation of soil properties and wheat yields on complex eroded hills using geostatistics and thematic mapper images. Remote Sens. Environ. 37:181–191.
• Bonham, C.D., and R.M. Reich. 1999. Influence of special autocorrelation on a fixed effect model used to evaluate treatment of oil spills. Appl. Math. Comput. 106:149–162.
• Boydell, B., and A.B. McBratney. 1999. Identifying potential within-field management zones from cotton yield estimates. p. 321–341. In J.V. Stafford (ed.) Precision agriculture 99. Vol 1. Yield and quality: Location, sensing, and geostatistics, Oxford, UK.
• Carr, P.M., G.R. Carlson, J.S. Jacobsen, G.A. Nielsen, and E.O. Slogles. 1991. Farming soils, not fields: A strategy for increasing fertilizer profitability. J. Prod. Agric. 4:57–61.
• Chen, F., D.E. Kissel, L.T. West, and W. Adkins. 2000. Field-scale mapping of surface soil organic carbon using remotely sensed imagery. Soil Sci. Soc. Am. J. 64:746–753.[Abstract/Free Full Text]
• Crookston, R.K. 1996. Using decision cases to enhance technology transfer in precision agriculture. p. 1117–1122. In P.C. Robert et al. (ed.) Proc. 3rd Int. Conf. on Site-Specific Management for Agricultural Systems. ASA, CSSA, and SSSA, Madison, WI.
• Doerge, T. 1999. Defining management zones for precision farming. Crop Insight. Vol. 8. No. 21. Pioneer Hybrids, Johnston, IA.
• Environmental Systems Research Institute. 1994. Grid commands. ESRI, Redlands, CA.
• Ferguson, R.B., C.A. Gotway, G.W. Hergert, and T.A. Peterson. 1996. Soil sampling for site-specific nitrogen management. p. 13–22. In P.C. Robert et al. (ed.) Proc. of the 3rd Int. Conf. on Site-Specific Management for Agricultural Systems. ASA, CSSA, and SSSA, Madison, WI.
• Fiez, T.E., B.C. Miller, and W.L. Pan. 1994. Assessment of spatially variable nitrogen fertilizer management in winter wheat. J. Prod. Agric. 7:86–93.
• Fleming, K.L., D.G. Westfall, D.W. Wiens, and M.C. Brodahl. 2001. Evaluating farmer defined management zones for variable rate fertilizer application. Precis. Agric. 2:201–215.
• Franzen, D.W., A.D. Halvorson, and V.L. Hofman. 2001. Management zones for soil N and P levels in the northern great plains. In P.C. Robert et al. (ed.) Proc. of the 5th Int. Conf. on Site-Specific Management for Agricultural Systems [CD]. ASA, CSSA, and SSSA, Madison, WI.
• Franzen, D.W., D.H. Hopkins, M.D. Sweeney, M.K. Ulmer, and A.D. Halvorson. 2002. Evaluation of soil survey scale for zone development of site specific nitrogen management. Agron. J. 94:381–389.[Abstract/Free Full Text]
• Fridgen, J.J., N.R. Kitchen, K.A. Sudduth, S.T. Drummond, W.J. Wiebold, and C.W. Fraisse. 2004. Management zone analyst (MZA): Software for subfield management zone delineation. Agron. J. 96:100–108.[Abstract/Free Full Text]
• Halvorson, A.D., and J.D. Rhoades. 1974. Assessing soil salinity and identifying potential saline seep areas with field soil resistance measurements. Soil Sci. Soc. Am. Proc. 38:576–581.
• Heermann, D.F., J. Hoeting, H.R. Duke, D.G. Westfall, G.W. Buckleiter, P. Westra, F.B. Peairs, and K.L. Fleming. 1999. Interdisciplinary irrigated precision farming team research. p. 212–130. In J.V. Stafford (ed.) Precision agriculture ‘99. Vol 1. Soil and crop factors: Location, sensing, and sampling. Oxford, UK.
• Jaynes, D.B. 1996. Improved soil mapping using electromagnetic induction surveys. p. 169–179. In Proc. 3rd Int. Conf. on Site-Specific Management for Agricultural Systems. ASA, CSSA, and SSSA, Madison, WI.
• Jaynes, D.B., J.M. Novak, T.B. Moorman, and C.M. Cambardella. 1994. Estimating herbicide partition coefficients from electromagnetic induction measurements. J. Environ. Qual. 24:26–41.
• Johnson, C.K., J.W. Doran, H.R. Duke, B.J. Wienhold, K.M. Eskridge, and J.F. Shanahan. 2001. Field-scale electrical conductivity mapping for delineating soil condition. Soil Sci. Soc. Am. J. 65:1829–1837.[Abstract/Free Full Text]
• Kachanoski, R.G., E.G. Gregorich, and I.J. Van Wesenbeck. 1988. Estimating spatial variations of soil water content using noncontacting electromagnetic inductive methods. Can. J. Soil Sci. 68:715–722.
• Kitchen, N.R., K.A. Sudduth, and S.T. Drummind. 1999. Soil electrical conductivity as a crop productivity measure for clay pan soils. J. Prod. Agric. 12:607–617.
• Long, D.S., R.E. Engel, and P. Reep. 1998. Grain protein sensing to identify nitrogen management some in spring wheat. Site-specific management guidelines no. 24. Potash and Phosphate Inst., Norcross, GA.
• Lund, E.D., and C.D. Christy. 1998. Using electrical conductivity to provide answers for precision farming. p. I-327–I-334. In Proc. 1st Int. Conf. on Geospatial Information in Agriculture and Forestry. ERIM Int., Ann Arbor, MI.
• Lund, E.D., C.D. Christy, and P.E. Drummond. 1999. Practical applications of soil electrical conductivity mapping. p. 771–779. In J.V. Stafford (ed.) Precision agriculture '99. Proc. of the 2nd European Conf. on Precision Agriculture. Oxford, UK.
• Mathsoft. 1995. S-PLUS guide to statistical and mathematical analysis. Ver. 3.3. Mathsoft, Seattle, WA.
• Mausbach, M.J., D.J. Lytle, and L.D. Spevey. 1993. Application of soil survey information to soil specific farming. p. 57–68. In P.C. Robert et al. (ed.) Soil specific crop management. ASA, CSSA, and SSSA, Madison WI.
• McBride, R.A., A.M. Gordon, and S.C. Shrive. 1990. Estimating forest soil quality from terrain measurements of apparent electrical conductivity. Soil Sci. Soc. Am. J. 54:290–293.[Abstract/Free Full Text]
• McCann, B.L., D.J. Pennick, C. van Kessel, and F.L. Walley. 1996. The development of management units for site-specific farming. p. 295–302. In P.C. Robert et al. (ed.) Proc. of the 3rd Int. Conf. on Site-Specific Management for Agricultural Systems. ASA, CSSA, and SSSA, Madison, WI.
• Moran, P.A.P. 1950. Notes on continuous stochastic phenomena. Biometrica 37:17–23.
• Mulla, D.J., and A.U. Bhatti. 1997. An evaluation of indicator properties affecting spatial patterns in N and P requirements for winter what yield. p. 145–153. In J.V. Stafford (ed.) Precision agriculture ‘97. Vol. 1. Spatial variability in soil and crop. Oxford, UK.
• Ostergaard, H.G.S. 1997. Agronomic consequences of variable N fertilization. p. 315–320. In J.V. Stafford (ed.) Precision agriculture ‘97. Vol. 1. Spatial variability in soil and crop. Oxford, UK.
• Redhen Systems. 1996. Farmers Software Harvest Mapping System. Ver. 3.1. Redhen Systems, Fort Collins, CO.
• Rhoades, J.D., and D.L. Corwin. 1981. Determining soil electrical conductivity–depth relations using an inductive electromagnetic soil conductivity meter. Soil Sci. Soc. Am. J. 45:255–260.[Abstract/Free Full Text]
• SAS Institute. 1990. SAS/STAT user's guide. Ver. 6. 4th ed. SAS Inst., Cary, NC.
• Sawyer, J.E. 1994. Concepts of variable rate technology with consideration for fertilizer application. J. Prod. Agric. 7:195–201.
• Schreier, H., R. Wiart, and S. Smith. 1988. Quantifying organic matter degradation in agricultural fields using PC-based image analysis. J. Soil Water Conserv. 43:421–424.
• Shatar, T.M., and A.B. McBratney. 2001. Subdividing a field into contiguous management zones using a k-zone algorithm. p. 115–120. In S. Blackmore and G. Grenier (ed.) Third European Conf. of Precision Agriculture. Vol. 1. New tools and methods for precision agriculture. Agro, Montpellier.
• Soil Survey Staff. 1996. Keys to soil taxonomy. 7th ed. USDA-NRCS.
• Stafford, J.V., R.M. Lark, and H.C. Bolam. 1998. Using yield maps to regionalize fields into management units. p. 225–237. In Proc. of the 4th Int. Conf. on Site-Specific Management for Agricultural Systems. ASA and SSSA, Madison, WI.
• Sudduth, K.A., D.F. Hughes, and S.T. Drummond. 1995. Electromagnetic induction sensing as an indicator of productivity on claypan soils. p. 671–681. In Proc. of the 2nd Int. Conf. on Site-Specific Management for Agricultural Systems. ASA, Madison, WI.
• Williams, B.G., and D. Hoey. 1987. The use of electromagnetic induction to detect the spatial variability of the salt and clay content of soils. Aust. J. Soil Res. 25:21–27.
• Yang, C., and L.G. Anderson. 1996. Determining within-field management zones for grain sorghum using aerial videography. p. 606–611. In Proc. 26th Int. Symp. on Remote Sensing of Environment. Remote Sensing of Environ., Tucson, AZ.

This article has been cited by other articles:

				M. Bridges, W. B. Henry, D. L. Shaner, R. Khosla, P. Westra, and R. Reich Spatial Variability of Atrazine and Metolachlor Dissipation on Dryland No-tillage Crop Fields in Colorado J. Environ. Qual., October 23, 2008; 37(6): 2212 - 2220. [Abstract] [Full Text] [PDF]

				D. L. Shaner, R. Khosla, M. K. Brodahl, G. W. Buchleiter, and H. J. Farahani How Well Does Zone Sampling Based on Soil Electrical Conductivity Maps Represent Soil Variability? Agron. J., September 8, 2008; 100(5): 1472 - 1480. [Abstract] [Full Text] [PDF]

				J. A. Taylor, A. B. McBratney, and B. M. Whelan Establishing Management Classes for Broadacre Agricultural Production Agron. J., September 11, 2007; 99(5): 1366 - 1376. [Abstract] [Full Text] [PDF]

				N. E. Derby, F. X. M. Casey, and D. W. Franzen Comparison of Nitrogen Management Zone Delineation Methods for Corn Grain Yield Agron. J., February 6, 2007; 99(2): 405 - 414. [Abstract] [Full Text] [PDF]

				Y. Miao, D. J. Mulla, W. D. Batchelor, J. O. Paz, P. C. Robert, and M. Wiebers Evaluating Management Zone Optimal Nitrogen Rates with a Crop Growth Model Agron. J., April 11, 2006; 98(3): 545 - 553. [Abstract] [Full Text] [PDF]

				M. Mzuku, R. Khosla, R. Reich, D. Inman, F. Smith, and L. MacDonald Spatial Variability of Measured Soil Properties across Site-Specific Management Zones Soil Sci. Soc. Am. J., August 25, 2005; 69(5): 1572 - 1579. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (10) Citing Articles via Google Scholar Google Scholar Articles by Fleming, K. L. Articles by Westfall, D. G. Search for Related Content PubMed Articles by Fleming, K. L. Articles by Westfall, D. G. Agricola Articles by Fleming, K. L. Articles by Westfall, D. G. Related Collections Spatial Variability Nutrient Management Production Agriculture Site-Specific Analysis Spatial Distribution