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 (12) Citing Articles via Google Scholar Google Scholar Articles by Heiniger, R. W. Articles by Clay, D. E. Search for Related Content PubMed Articles by Heiniger, R. W. Articles by Clay, D. E. Agricola Articles by Heiniger, R. W. Articles by Clay, D. E. Related Collections Spatial Distribution Remote Sensing Nutrient Management Site-Specific Analysis Soil Analysis Soil Methods/Instrumentation

SYMPOSIUM PAPERS

Using Soil Electrical Conductivity to Improve Nutrient Management

^a Dep. of Crop Sci., North Carolina State Univ., Raleigh, NC 27695
^b Dep. of Soil Sci., North Carolina State Univ., Raleigh, NC 27695
^c Plant Sci. Dep., South Dakota State Univ., Brookings, SD 57007

^* Corresponding author (Ron_Heiniger{at}ncsu.edu)

Received for publication June 21, 2001.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
While site-specific nutrient management has the potential for improving crop yields, the cost of intensive soil sampling is usually greater than the benefits gained. Apparent soil electrical conductivity (EC_a) has been used successfully to measure soil salinity, clay content, and, in the laboratory, nutrient concentrations. This study was initiated to determine if EC_a could be used to measure nutrient concentrations in the field. Fifteen field sites with 12 different soil series were studied in three topographic areas of North Carolina in 1997 and 1999. Soil samples and EC_a measurements were taken at the same locations in the field and analyzed for P, K, Ca, Mg, Mn, Zn, Cu, pH, cation exchange capacity (CEC), percentage humic matter (HM), and percentage sand, silt, and clay. Nutrient concentrations and soil properties were compared with EC_a using correlation and principal components (PC)–stepwise regression analysis. Few significant direct correlations were found between EC_a and the selected nutrient elements (R² < 0.50). Correlations improved when soil series were analyzed separately within a field. The results indicated that salinity, soil texture, or soil moisture were masking the response of EC_a to changing nutrient levels in the soil. While the PC–stepwise regression analysis found that EC_a was often a key loading factor, changes in soil texture, CEC, and HM resulted in field-specific relationships between EC_a and nutrient concentrations. The primary value of EC_a in measuring nutrient levels lies in its ability to identify small changes in soil texture, CEC, or HM that, in turn, indicate where differences in nutrient levels occur.

Abbreviations: CEC, cation exchange capacity • DGPS, differential global positioning system • EC_a, apparent soil electrical conductivity • EC_s, the electrical conductivity of the soil particles • EC_wc, the electrical conductivity of the mobile soil solution associated with large, continuous pores • EC_ws, the electrical conductivity of the soil solution associated with discontinuous pores • HM, humic matter • PC, principal component • _s, the volumetric content of soil particles • _w, the total volumetric content of water in the soil • _ws, the volumetric soil water content of the small, discontinuous pores

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
RESEARCH HAS DEMONSTRATED that site-specific nutrient management can increase crop yield and, in some cases, reduce the amount of fertilizer or lime applied (Kitchen et al., 1995; Bongiovanni and Lowenberg-Deboer, 1999). Unfortunately, the cost of intensive soil sampling required to accurately map soil nutrient content across a field is often higher than the economic benefits received from increasing crop yield, decreasing fertilizer cost, or both (Swinton and Mubariq, 1996; English et al., 1999). Measurements of EC_a have been used successfully to measure soil salinity, clay, and water content (Kachanoski et al., 1988; Williams and Hoey, 1987; Rhoades et al., 1976, 1989; Freeland, 1989). Current techniques using either the direct application of current to the soil (Veris 3100; Veris Technologies, Salina, KS) or electrical magnetic induction (Geonics EM38, Geonics Limited., Mississauga, ON, Canada) enable direct, instantaneous, in situ measurement of EC_a (Jaynes et al., 1995; Lund et al., 1999; Sudduth et al., 1999; Fritz et al., 1999). Because laboratory studies have shown that nutrients such as N and K directly affect the electrical conductivity of the isolated soil solution (Ouyang et al., 1998), it is reasonable to assume that EC_a could be used to measure the available nutrient content of the soil, eliminating the need for time-consuming and expensive soil sample acquisition and analysis.

Apparent soil electrical conductivity consists of two components: (i) the contribution of the solid soil particles primarily associated with exchangeable cations and (ii) the contribution of the soil solution (Nadler and Frenkel, 1980; Shainberg et al., 1980). Rhoades et al. (1989) described EC_a using a two-pathway model, with one path for conductance via the discontinuous soil–liquid interface and the other path via continuous, large water-filled pores through the soil solution. They found that apparently, soil structure does not provide enough direct particle-to-particle contact to form a continuous pathway for current flow. Their model identifies five major factors that influence EC_a. These include: the electrical conductivity of the soil particles (EC_s); the electrical conductivity of the soil solution associated with discontinuous pores (EC_ws); the electrical conductivity of the mobile soil solution associated with large, continuous pores (EC_wc); the volumetric soil water content of the small, discontinuous pores (_ws); the total volumetric content of water in the soil (_w); and the volumetric content of soil particles (_s). Several studies have found that the major factors influencing EC_a are _w (Rhoades et al., 1976; Nadler, 1982); _s per unit of soil, influenced primarily by the texture and bulk density of the soil (Rhoades and Corwin, 1990); the CEC of the soil, which influences EC_s (Shainberg et al., 1980); and the amount of dissolved salts in the soil solution, which influences EC_ws and EC_wc (Malicki and Walczak, 1999).

The most important factor influencing EC_a is the total volumetric water content of the soil (Rhoades et al., 1976). Changes in volumetric soil water content tend to mask differences in the other factors influencing EC_a (Nadler, 1982). When the volumetric soil water content is high (near field saturation), the primary path for EC_a is through the larger continuous pores, and EC_wc is a major influence on electrical conductivity (Rhoades et al., 1989). However, when the volumetric soil water content is low (near wilting point), the primary path is through the soil particle–discontinuous soil pore pathway. In this situation, the EC_s and EC_ws are the major influences on EC_a.

Another major influencing factor is soil salinity (Malicki and Walczak, 1999). At high Na concentrations, the conductivity of the soil solution (EC_ws and EC_wc) is greater than that of the bulk soil (Shainberg et al., 1980). However, in nonsaturated, nonsaline soils, the conductivity of the bulk soil (EC_s) is greater due to the contribution of adsorbed ions. If we assume that salinity is not a major factor in most productive agricultural fields, it follows that in nonsaturated conditions, changes in soil nutrient levels will most likely influence EC_a by changing EC_s through differences in the type and number of cations held by the soil particles. Therefore, cations commonly associated with binding sites on the soil particles, such as Ca, Mg, or K, could influence EC_a by changing EC_s. However, the common assumption is that, in most field situations, the influence that changing levels of soil cations have on EC_s is minor compared with the influence associated with changes in soil bulk density and texture (Lund et al., 1999).

Differences in soil nutrient content could also potentially influence EC_ws and EC_wc through changes in the conductivity of the soil solution. Under saturated conditions, changes in nutrient levels could influence EC_a by changing EC_wc (Rhoades et al., 1989). Therefore, dissolved nutrients such as N and S should have more influence on EC_a under saturated conditions than when the soil is close to the wilting point. Unfortunately, large differences in volumetric soil water content across a field will likely mask the small differences that changing nutrient levels will have on EC_wc. Based on an analysis of the factors contributing to EC_a, it is clear that it will be difficult to directly measure site-specific changes in soil nutrient content using electrical conductivity.

An analysis of the factors that contribute to EC_a suggests that EC_a might be used along with other measurements of soil properties, such as CEC, texture, pH, and water-holding capacity, to determine nutrient levels in the soil. Studies have suggested that if spatial differences in water-holding capacity and salt concentration were taken into account, there would be a good relationship between soil nutrient levels and EC_a (Chang et al., 2001). Others have found that EC_a could be used to determine the depth to clay (Doolittle et al., 1994) and that these measurements identified areas where EC_a was related to soil nutrient content. Because soil properties such as texture and CEC are relatively static over time, a model that included these measurements along with EC_a could be developed to predict soil nutrient content.

Apparent soil electrical conductivity could also be used to indirectly determine field sites where soil nutrient levels differ. Depth of topsoil, soil water content, CEC, and texture are all good indicators of crop productivity (Jaynes et al., 1995; Sudduth et al., 1996). Because EC_a is able to directly measure these soil properties, it has the potential to identify management zones with differing productivity and nutrient requirements (Kitchen et al., 1999). Several researchers have reported that EC_a was useful in identifying field sites with different soil properties (Sudduth et al., 1999; Franzen and Kitchen, 1999). Sudduth et al. (1996) found that EC_a was correlated with depth to clay. Because depth to clay was the primary factor influencing yield levels at different field sites, EC_a proved to be a valuable tool in determining site-specific N requirements. These studies indicate that EC_a could, indirectly, provide a useful measure of nutrient levels.

The main purpose of this paper is to determine the utility of EC_a in mapping soil nutrient levels across a field for the purpose of making variable-rate nutrient applications. The specific objectives of this study are to examine the use of EC_a: (i) to directly measure soil pH, texture, CEC, and levels of selected elements (P, K, Ca, Mg, Mn, Cu, and Zn) across a field and (ii) in conjunction with measured soil properties to measure the concentrations of selected elements (P, K, Ca, Mg, Mn, Cu, and Zn) in a field.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Research was conducted over a 2-yr period, 1997 and 1999, on 16 fields chosen to represent the three major topographic regions (tidewater, coastal plain, and piedmont) of North Carolina. These field sites varied topographically, and the 12 soil series (Table 1) found across these sites had a wide range of soil textures and properties. All of the fields were managed using a standard corn (Zea mays L.)–wheat (Triticum aestivum L.)–soybean [Glycine max (L.) Merr.] crop rotation. Conventional tillage practices (chisel, disk, and field cultivator) were used at all of the field sites in the tidewater and coastal plain regions at some point in the rotation. In the piedmont, all of the fields were managed using continuous no-tillage practices. Previous fertilizer and lime applications had been uniformly applied based on soil test recommendations (Tucker et al., 1996).

1999 Data Collection
Nine locations were studied in 1999: four fields in the tidewater region, two in the coastal plain region, and three in the piedmont region (Table 1). At each location, soil samples were collected at two depths (0–30 cm and 30–90 cm) from 30- by 30-m grids. Each composite sample contained six cores collected within a circular area with a diameter of 5 m centered on the sample location recorded using a DGPS receiver. Soil samples were dried, ground to pass through a 2-mm sieve, and analyzed for CEC, HM, pH, active acidity, P, K, Ca, Mg, Mn, Zn, and Cu using the same testing procedures used in 1997. In addition, soil textural analysis was performed on all of the 30- to 90-cm samples and the 0- to 30-cm samples from the coastal plain and piedmont regions using the hydrometer method of Gee and Bauder (1986). The 0- to 30-cm samples from the fields in the tidewater region could not be analyzed because the organic matter content (>15%) of the surface horizon interfered with the testing procedure. On the same day that the soil samples were collected, EC_a was measured using an EM38 (Geonics Limited, Mississauga, ON, Canada). When taking readings, the EM38 unit was positioned both horizontally and vertically at ground level and at heights of 30, 60, 90, and 120 cm. At three representative sites in each field, soil samples were collected from depths of 15, 30, 45, 60, 75, 90, 105, and 120 cm and analyzed in the laboratory for electrical conductivity. Using the technique described by Rhoades and Corwin (1981), a calibration analysis was performed relating EC_a collected by the EM38 with electrical conductivity by depth as measured in the laboratory. This calibration was then applied to all EM38 measurements to obtain EC_a for 0- to 30- and 30- to 90-cm depths.

Statistical Analysis
Correlation analysis (SAS Inst., 1990) was conducted on each field and within fields on each soil series to determine if EC_a, CEC, HM, pH, soil type, sand, silt, or clay were related to soil test concentrations of P, K, Ca, Mg, Mn, Zn, or Cu. Significant results with a high Pearson correlation coefficient (>0.70) would indicate situations where the measured soil property could be used to estimate the concentration of that particular element in the soil.

Principal-components analysis was used to examine the relationship among the soil properties measured in this study (EC_a, CEC, HM, pH, soil type, percentage sand, percentage silt, and percentage clay) and to determine which soil properties were important influences on soil test concentrations of P, K, Ca, Mg, Mn, Zn, or Cu. To account for soil texture in 1997, a soil type variable was formed by assigning a number to the soil series found in each field. This number was then used in the PC analysis to represent the influence of soil texture on differences in the concentrations of the elements tested.

Due to the colinearity of the independent variables, correlation analysis could not be used to directly relate multiple soil properties to soil test levels of the elements tested. Principal-components analysis puts identified, correlated variables into groups. These groups then become new, independent, random variables that could then be used to identify which soil properties influenced nutrient levels. In this study, the objectives of using the PC–stepwise regression analysis were to identify the key soil properties that, in conjunction with EC_a, had significant relationships with element concentrations; determine the strength of that relationship; and determine the influence and role of each soil property in the relationship.

The PC groups were identified from the correlation matrix using the FACTOR procedure in SAS (SAS Inst., 1990). Any PC group with an eigenvalue greater than 1 was selected because it explained a significant amount of the variance present in the soil properties at each location. In no case did the PC groups with eigenvalues >1 explain <66% of the cumulative variance. More commonly, these PC groups had a cumulative variance of 75%. The PC groups with eigenvalues >1 were then used in a stepwise regression procedure (SAS Inst., 1990) to determine if there was a significant relationship between the PC group and soil test concentrations of P, K, Ca, Mg, Mn, Zn, or Cu. When PC groups remaining in the regression model accounted for >50% of the variability in the concentration of that particular element, the eigenvectors (loading factors) were examined and the soil properties in the PC groups ranked according to the amount of variability explained by the PC group and the loadings (C|j|). For instance, a soil property that was a component of the PC group that accounted for most of the variability in the regression model and had the highest loading factor in that PC group was ranked first. Soil properties with loading factors <0.5 were not considered key latent variables and were not included in the ranking because they did not substantially influence the relationship between the PC groups and the nutrient concentration being examined. The ranking of the soil properties, strength of the loading factor, and sign (positive or negative) of the loading factor were used to determine the influence and role that each soil property had in explaining the variability in the concentration of each element.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Using Apparent Soil Electrical Conductivity to Directly Measure Nutrient Content
1997
In examining the direct correlation between EC_a and soil test levels of P, K, Ca, Mg, Mn, Zn, and Cu, only 24 significant relationships were found at the seven field sites measured in 1997 out of the 114 comparisons made (Table 2). The strongest correlations when soil series were not separated were between EC_a and P on fields B534 and F1 (R² = 0.26 and 0.37, respectively), EC_a and K on fields B534 and F1 (R² = 0.13 and 0.32, respectively), EC_a and Mg on field F1 (R² = 0.33), EC_a and Mn on fields B533 and B534 (R² = 0.31 and 0.16, respectively), EC_a and Zn on fields B533 and B534 (R² = 0.30 and 0.15, respectively), and EC_a and Cu on field B534 (R² = 0.13). Unfortunately, none of the relationships were strong enough to accurately predict soil test levels of any of the nutrients from measured values of EC_a. Furthermore, there were situations where negative relationships were observed between EC_a and nutrient concentrations that could not be caused by the influence of these ions on electrical conductivity.

Because volumetric soil particle density and volumetric soil moisture content have a large influence on EC_a and are closely related to soil texture as are CEC, pH, and HM, fields B531, B532, B533, B534, and B535 were divided by soil series, and correlations between EC_a and element concentrations were examined on each soil series to determine if there was an improvement in predicting concentrations of particular elements based on EC_a. In 40 out of 75 cases, there was an improvement in the strength of the relationship between EC_a and the concentration of the element measured (Table 2). Figure 1 shows the relationship between EC_a and soil P concentration. While the overall relationship was weak and showed a negative trend, the comparisons on each individual soil series were stronger and had a positive trend, indicating that EC_a was responding to increasing ion concentrations in the soil.

View larger version (22K): [in this window] [in a new window]   Fig. 1. Relationship between apparent soil electrical conductivity (ECa) and soil test concentrations of P on field B533 in the tidewater region of North Carolina in 1997. Individual regressions represent the three different soil series present in the field and the relationship between ECa and P concentrations within each soil series.

View larger version (46K): [in this window] [in a new window]   Fig. 2. Maps of apparent soil electrical conductivity (ECa) and soil test concentrations of P on field GR1 in the tidewater region of North Carolina in 1999. Sites denoted as feeding areas and a fence line were the points in the field where manure deposition was the greatest. The feeding bunks and fence line were removed several months before ECa and nutrient measurements were taken.

1997
From the seven fields studied in 1997, there were eight regression models where PC groups accounted for >50% of the variability in the measured nutrient level. Of these eight regression models, EC_a was identified as a key latent variable in six of the models (Table 5). In field B533, the PC–stepwise regression analysis found that both Mn and Zn were related (R² = 0.63 and 0.75, respectively) to two PC groups. In PC Group 1 (PC1), pH had the highest loading factor (Table 6), indicating it was negatively related to soil test levels of Mn and Zn. Cation exchange capacity, HM, and EC_a had similar loading factors and were positively related to soil test levels of Mn and Zn. Soil type dominated PC Group 2 (PC2), indicating a clear relationship between soil type and soil test levels of Mn and Zn. The change from heavy organic to lighter-textured loam soils across this field and the association of pH, CEC, HM, and EC_a with soil series (lower pH and higher CEC, HM, and EC_a on the organic Belhaven and Ponzer soil series and higher pH and lower CEC, HM, and EC_a on the lighter Wasda and Deloss soil series) clearly influenced the relationship with Mn and Zn.

In field H08, the PC–stepwise regression analysis found that soil test levels of Mg were related to PC1 and PC2 (R² = 0.68) (Table 5). Principal-Component Group 1 was dominated by EC_a (C|j| = 0.912) followed by positive loadings from HM and CEC (Table 6). Principal-Component Group 2 was dominated by pH (C|j| = 0.934), with a much smaller contribution from CEC. The strong influence of EC_a in this regression model indicates a situation where EC_a along with other measured soil properties could be used successfully to determine element concentrations across a field.

Similarly, the PC–stepwise regression analysis found that soil test levels of Mg were related to PC1 in field F1 (Table 5). Principal-Component Group 1 was dominated by the negative loading factors CEC and HM (Table 6), with smaller positive loadings from pH and EC_a. Field F1 is located in the rolling hills of the piedmont region of North Carolina and has a clay loam soil texture. Changes in CEC in this field are associated with an increase in clay content and CEC on the sideslopes where runoff has eroded the surface horizon, exposing the B horizon where Mg levels are low. The positive relationship with EC_a is primarily the result of higher Mg levels in the toeslope areas of the field where eroded materials accumulate and where soil moisture is generally greater.

1999—0- to 30-cm Sampling Depth
From the nine fields studied in 1999, there were 20 regression models where PC groups accounted for >50% of the variability in the measured nutrient levels in the 0- to 30-cm depth. Of these 20 regression models, EC_a was identified as a key latent variable in 13 of the models (Table 7). The results in 1999 were similar to those found in 1997. In situations where EC_a was identified as a key latent variable in describing nutrient concentrations, changes in soil texture (Table 4) and CEC were also key factors in determining nutrient levels in the field.

1999—31- to 90-cm Sampling Depth
At the 31- to 90-cm sampling depth, there were 25 regression models where PC groups accounted for >50% of the variability in measured nutrient levels. Of these 25 regression models, EC_a was identified as a key latent variable in 18 of the models (Table 7). As was the case in the soil samples and EC_a measurements taken at the 0- to 30-cm depth, Ca and Mg were the two elements where PC groups comprised of the soil properties measured in this study often accounted for most of the variability in measured levels. In eight of the nine fields, the regression models accounted for >54% of the variability in Ca and Mg levels.

The dominant loading factors in fields B500, MR1, GR1, BOT1, BOT4, Fischer, and Graham were sand, silt, and clay content; CEC; HM; and EC_a (Tables 8 and 9). Except for CEC in the Fischer field and HM in the Graham field, CEC, HM, and EC_a were positively related to Ca or Mg levels. In fields B500, BOT1, and Graham, sand was negatively related to Ca and Mg levels while silt and clay were positively related. These results indicate that the positive relationship between EC_a and percentage clay is behind the positive relationship between EC_a and Ca or Mg soil test levels. However, for fields MR1, GR1, BOT4, and Fischer, sand content was a positive loading factor while silt and clay were negatively related to nutrient concentration. The results on these fields indicate that the positive relationship between EC_a and either Ca or Mg soil test levels must be influenced by other factors, such as CEC or HM. In the case of GR1, this could be the result of the positive association among salinity from the manure application, EC_a, and Mg concentrations.

Other elements in which the PCs accounted for >50% of the variability in the measured samples and EC_a was a key loading included Mn on fields B500, BOT1, and Hall and Zn and Cu on MR1 (Table 7). For Mn, CEC, EC_a, silt, clay, and pH were positive key loading factors while sand was negatively related to soil test levels (Tables 8 and 9). Although the PC–stepwise regression analysis of the data from MR1 indicated that these same key loading factors were related to Zn and Cu levels in this field (accounting for 58 and 51% of the variability, respectively) (Table 7), the relationship between the key loading factors and these elements differed. Apparent soil electrical conductivity, CEC, HM, pH, and sand were positive loading factors while silt and clay were negatively related to Zn and Cu soil test levels (Table 8). Based on the strong influence that EC_a and CEC had on these regression models (C|j| = 0.679 and 0.671, respectively) and the negative influence of silt and clay, it is reasonable to assume that EC_a was influenced by the relationship between CEC and changes in Zn or Cu concentrations.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Using Apparent Soil Electrical Conductivity to Directly Measure Nutrient Content
Measurements conducted on 16 fields over 2 yr found only a small number of cases where EC_a could be used to directly measure nutrient levels (Tables 2 and 3). Few significant linear relationships were found between EC_a and soil test levels of P, K, Ca, Mg, Mn, Zn, and Cu, and those that were significant were often weak (R² < 0.50). There were a few instances where strong significant relationships were found (R² ranging from 0.51–0.75). These occurred when the nutrient was closely associated with one of the four soil properties that directly influence EC_a: volumetric water content, volumetric content of soil particles, CEC, and dissolved salts in the soil solution. In field GR1, where EC_a was directly related to soil test levels of P, available P was closely linked to dissolved salts from animal manure applied to the field (Fig. 2). In the Graham field, where EC_a was directly related to soil test levels of Ca and Mg, these cations were associated with differences in CEC across the field and a strong relationship between EC_a and CEC (Table 4). In cases where elements were closely associated with soil properties that have a large influence on EC_a, it was possible to measure soil nutrient content through the use of EC_a.

Improvements in the correlation between EC_a and soil nutrient levels were often found when fields were subdivided by soil series. On field B533, the amount of variability explained by a linear model relating EC_a to soil P content increased from 3 to 24% by subdividing the field according to soil series (Table 2 and Fig. 1). The fractions of sand, silt, and clay in the soil sample were often significantly related to EC_a (Table 4). Because soil texture influences both volumetric soil moisture content and the volumetric content of soil particles, this would make sense. Therefore, differences in soil texture often influence EC_a much more than small differences in element concentrations. By dividing fields into areas of similar soil texture, one source of variation was removed, and this improved the accuracy of using EC_a to measure changes in soil concentrations of nutrient elements.

Relationships among Apparent Soil Electrical Conductivity, Other Soil Properties, and Soil Nutrient Content
In the 16 fields studied, the PC–stepwise regression analysis was able to identify 37 regression models out of a possible 165 comparisons in which PC groups accounted for 50% of the variability in nutrient levels and where EC_a was one of the key loading factors (Tables 5 and 7). Principal-component groups consisting of EC_a and other measured soil properties were not able to consistently account for changing levels of most of the elements measured in this study, in particular, P and K. However, as the initial correlation analysis indicated, EC_a along with measured soil properties, particularly CEC and soil texture, was able to account for a majority of the variability in soil test levels of Ca and Mg in 13 out of 22 comparisons made at the 0 to 30-cm depth (R² from 0.52–0.93) and in 13 out of 18 comparisons made at the 31- to 90-cm depth (R² from 0.54–0.88).

In most of these comparisons, EC_a, CEC, and clay were all identified as positive loading factors. Under these non-water-saturated conditions, _ws, _s, and EC_ws would be the most important factors influencing EC_a. As clay content increased, there would be an increase in _s and _ws and increased concentration of Ca and Mg cations in the soil solution of the discontinuous pores associated with increases in CEC would have increased EC_ws. While it is reasonable to assume that the influence of silt and clay content on _s and _ws were important influences on EC_a, the relationship between EC_a and Ca or Mg soil test levels was not always associated with an increases in silt and clay content. On six fields, clay was negatively related to Ca or Mg soil test levels even though EC_a had a positive relationship. This occurred most often in the piedmont where increases in clay content were associated with eroded areas of the field where the concentration of Ca or Mg was low. In these situations, increases in CEC contributed to the increased concentration of Ca and Mg in the soil solution and to increases in EC_a.

Unfortunately, even when describing Ca and Mg, the PC–stepwise regression analysis was not able to consistently identify models that accounted for most of the variability in soil test levels. Furthermore, the model parameters, PC groups, and key loading factors differed according to the field being studied. This means that a single relationship among EC_a, soil properties, and soil element concentrations cannot be developed. This doesn't mean that EC_a has no value in determining nutrient levels in the soil. Instead, this study shows that EC_a can be valuable tool when used in conjunction with multivariate statistical procedures in identifying soil properties and their relationship to nutrient availability. Because there are close relationships among CEC, soil texture, and EC_a, measurements of EC_a could be substituted for detailed soil textural sampling to develop a detailed field map showing where element levels are likely to change. This map could then be used to determine nutrient management zones. Franzen and Kitchen (1999) demonstrated the use of EC_a to delineate management zones. Apparent soil electrical conductivity could also be used in co-kriging to reduce the density of soil sampling without sacrificing the accuracy of the mapped data.

This study shows that it is unlikely that EC_a can be used to directly determine soil nutrient content across a field. However, EC_a can be used along with other measured soil properties in a multivariate analysis to describe the key factors influencing changes in nutrient concentrations and to establish nutrient management zones. To use EC_a in the process of establishing nutrient management zones, it is important to know and account for site-specific changes in volumetric soil moisture content, soil texture, CEC, and salinity (either from manure applications or on saline soils).

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 

• Bongiovanni, R., and J. Lowenberg-DeBoer. 1999. Economics of variable rate lime in Indiana. p. 1653–1665. In P.C. Robert et al. (ed.) Precision agriculture. Proc. Int. Conf., 4th, St. Paul, MN. 19–22 July 1998. ASA, CSSA, and SSSA, Madison, WI.
• Chang, J., D.E. Clay, C.G. Carlson, S.A. Clay, and C. Reese. 2001. Determining the impact of approaches to classify nutrient management zones. In P.C. Robert et al. (ed.) Precision agriculture [CD-ROM]. Proc. Int. Conf., 5th, Minneapolis, MN. 16–19 July 2000. ASA, CSSA, and SSSA, Madison, WI.
• Doolittle, J.A., K.A. Sudduth, N.R. Kitchen, and S.J. Indorante. 1994. Estimating depths to claypans using electromagnetic induction methods. J. Soil Water Conserv. 49:572–575.
• English, B.C., R.K. Roberts, and S.B. Mahajanashetti. 1999. Spatial break-even variability for variable rate technology adoption. p. 1633–1642. In P.C. Robert et al. (ed.) Precision agriculture. Proc. Int. Conf., 4th, St. Paul, MN. 19–22 July 1998. ASA, CSSA, and SSSA, Madison, WI.
• Franzen, D.W., and N.R. Kitchen. 1999. Developing management zones to target nitrogen applications. Site-Specific Management Guidelines. SSMG-5. Potash and Phosphate Inst., Norcross, GA.
• Freeland, R.S. 1989. Review of soil moisture sensing using soil electrical conductivity. Trans. ASAE 32(6):2190–2194.
• Fritz, R.M., D.D. Malo, T.E. Schumacher, D.E. Clay, and C.G. Carlson. 1999. Field comparison of two soil electrical conductivity measurement systems. p. 1211–1217. In P.C. Robert et al. (ed.) Precision agriculture. Proc. Int. Conf., 4th, St. Paul, MN. 19–22 July 1998. ASA, CSSA, and SSSA, Madison, WI.
• Gee, G.W., and J.W. Bauder. 1986. Particle-size analysis. p. 635–662. In A. Klute (ed.) Methods of soil analysis. Part 1. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
• Jaynes, D.B., T.S. Colvin, and J. Ambuel. 1995. Yield mapping by electromagnetic induction. p. 383–394. In P.C. Robert et al. (ed.) Site-specific management for agricultural systems. Proc. Int. Conf., 2nd, Minneapolis, MN. 27–30 Mar. 1994. ASA, CSSA, and SSSA, Madison, WI.
• Kachanoski, R.G., E.G. Gregorich, and I.J. Van Wesenbeeck. 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.J. Birrell. 1995. Comparison of variable rate to single rate. p. 427–439. In P.C. Robert et al. (ed.) Site-specific management for agricultural systems. Proc. Int. Conf., 2nd, Minneapolis, MN. 27–30 Mar. 1994. ASA, CSSA, and SSSA, Madison, WI.
• Kitchen, N.R., K.A. Sudduth, and S.T. Drummond. 1999. Soil electrical conductivity as a crop productivity measure for claypan soils. J. Prod. Agric. 12:607–617.
• Lund, E.D., P.E. Colin, D. Christy, and P.E. Drummond. 1999. Applying soil electrical conductivity technology to precision agriculture. p. 1089–1100. In P.C. Robert et al. (ed.) Precision agriculture. Proc. Int. Conf., 4th, St. Paul, MN. 19–22 July 1998. ASA, CSSA, and SSSA, Madison, WI.
• Malicki, M.A., and R.T. Walczak. 1999. Evaluating soil salinity status from bulk electrical conductivity and permittivity. Eur. J. Soil Sci. 50(3):505–514.
• McLean, E.O., M.R. Heddleson, and G.J. Post. 1959. Aluminum in soils: III. A comparison of extraction methods in soils and clays. Soil Sci. Soc. Am. J. 23:289–293.[Abstract/Free Full Text]
• Mehlich, A. 1984. Mehlich-3 soil test extractant: A modification of Mehlich-2 extractant. Commun. Soil Sci. Plant Anal. 15(12):1417–1422.
• Nadler, A. 1982. Estimating the soil water dependence of the electrical conductivity soil solution/electrical conductivity bulk soil ratio. Soil Sci. Soc. Am. J. 46:722–726.[Abstract/Free Full Text]
• Nadler, A., and H. Frenkel. 1980. Determination of soil solution electrical conductivity from bulk soil electrical conductivity measurements by the four-electrode method. Soil Sci. Soc. Am. J. 44:1216–1221.[Abstract/Free Full Text]
• Ouyang, D., A.F. MacKenzie, and M. Fan. 1998. Phytotoxicity of banded urea amended with triple superphosphate and potassium chloride. Agron. J. 90:734–739.[Abstract/Free Full Text]
• 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]
• Rhoades, J.D., and D.L. Corwin. 1990. Soil electrical conductivity: Effects of soil properties and application to soil salinity appraisal. Commun. Soil Sci. Plant Anal. 21:837–860.
• Rhoades, J.D., N.A. Manteghi, P.J. Shouse, and W.J. Alves. 1989. Soil electrical conductivity and soil salinity: New formulations and calibrations. Soil Sci. Soc. Am. J. 53:433–439.[Abstract/Free Full Text]
• Rhoades, J.D., P.A.C. Raats, and R.J. Prather. 1976. Effects of liquid-phase electrical conductivity, water content, and surface conductivity on bulk soil electrical conductivity. Soil Sci. Soc. Am. J. 40:651–655.[Abstract/Free Full Text]
• SAS Institute. 1990. SAS/STAT user's guide. Version 6. 4th ed. SAS Inst., Cary, NC.
• Shainberg, I., J.D. Rhoades, and R.J. Prather. 1980. Effect of exchangeable sodium percentage, cation exchange capacity, and soil solution concentration on soil electrical conductivity. Soil Sci. Soc. Am. J. 44:469–473.[Abstract/Free Full Text]
• Sudduth, K.A., N.R. Kitchen, and S.T. Drummond. 1999. Soil conductivity sensing on claypan soils: Comparison of electromagnetic induction and direct methods. p. 979–990. In P.C. Robert et al. (ed.) Precision agriculture. Proc. Int. Conf., 4th, St. Paul, MN. 19–22 July 1998. ASA, CSSA, and SSSA, Madison, WI.
• Sudduth, K.A., N.R. Kitchen, D.F Hughes, and S.T. Drummond. 1996. Electromagnetic induction sensing as an indicator of productivity on claypan soils. p. 671–681. In P.C. Robert et al. (ed.) Precision agriculture. Proc. Int. Conf., 3rd, Minneapolis, MN. 23–26 June 1996. ASA, CSSA, and SSSA, Madison, WI.
• Swinton, S.M., and A. Mubariq. 1996. Returns to farmer investments in precision agriculture equipment and services. p. 1009–1018. In P.C. Robert et al. (ed.) Precision agriculture. Proc. Int. Conf., 3rd, Minneapolis, MN. 23–26 June 1996. ASA, CSSA, and SSSA, Madison, WI.
• Tucker, M.R., J.K. Messick, and C.C. Carter. 1996. Crop fertilization based on North Carolina soil tests. Circular 1. North Carolina Dep. of Agric., Agron. Div., Raleigh.
• Williams, B.G., and D. Hoey. 1987. The use of electromagnetic induction to detect the spatial variability of the salt and clay contents of soils. Aust. J. Soil Res. 25:21–27.

This article has been cited by other articles:

				J. Sawchik and A. P. Mallarino Evaluation of Zone Soil Sampling Approaches for Phosphorus and Potassium Based on Corn and Soybean Response to Fertilization Agron. J., November 6, 2007; 99(6): 1564 - 1578. [Abstract] [Full Text] [PDF]

				M. S. Grigera, R. A. Drijber, K. M. Eskridge, and B. J. Wienhold Soil Microbial Biomass Relationships with Organic Matter Fractions in a Nebraska Corn Field Mapped using Apparent Electrical Conductivity Soil Sci. Soc. Am. J., August 3, 2006; 70(5): 1480 - 1488. [Abstract] [Full Text] [PDF]

				W. K. Jung, N. R. Kitchen, K. A. Sudduth, R. J. Kremer, and P. P. Motavalli Relationship of Apparent Soil Electrical Conductivity to Claypan Soil Properties Soil Sci. Soc. Am. J., May 6, 2005; 69(3): 883 - 892. [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 (12) Citing Articles via Google Scholar Google Scholar Articles by Heiniger, R. W. Articles by Clay, D. E. Search for Related Content PubMed Articles by Heiniger, R. W. Articles by Clay, D. E. Agricola Articles by Heiniger, R. W. Articles by Clay, D. E. Related Collections Spatial Distribution Remote Sensing Nutrient Management Site-Specific Analysis Soil Analysis Soil Methods/Instrumentation