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 Cox, M. S. Articles by Abshire, M. J. Search for Related Content PubMed Articles by Cox, M. S. Articles by Abshire, M. J. Agricola Articles by Cox, M. S. Articles by Abshire, M. J. Related Collections Soil Fertility and Productivity Other Soil Management Sustainable Agriculture

DIVISION S-8—NUTRIENT MANAGEMENT & SOIL & PLANT ANALYSIS

Variability of Selected Soil Properties and Their Relationships with Soybean Yield

Exp. Statistics Unit, Box 9653, Mississippi State Univ., MS 39762

^* Corresponding author (mcox{at}pss.msstate.edu)

	ABSTRACT

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Given the potential for high amounts of variability in yield-affecting soil factors, some soil property or properties may serve as a basis for site-specific soil management. The objectives of this study were to determine the variability of selected soil properties and to determine the relationships between these soil properties and soybean [Glycine max (L.) Merr.] yield. Soil samples were collected from the center point of 0.5-ha grids in three fields and were analyzed for soil-test-extractable Ca, Mg, K, P, pH in water, and texture. Relative elevation, slope, aspect, and soybean yield were also determined at these points. Coefficient of variation, Pearson correlation coefficients, and principal component (PC) analysis coupled with stepwise regression were used to analyze the data. Two of the three fields had medium to high P and K values but low yields while the third field had low P and K values but relatively higher yields, suggesting factors other than P and K levels were affecting yield. Soil variability, with the exception of pH, was highest in the North field. Potassium in this field exhibited evidence of high amounts of small-scale spatial or temporal variability. Across all three fields, pH had the lowest amount of variability while variability in soil fertility varied from year to year and field to field. Fertility parameters had to be considered with other soil factors to determine their relationship to yield. Topography-yield relationships varied from field to field. Areas with higher clay content in all three fields had higher yield, suggesting clay could be used as a basis for site-specific soil management.

Abbreviations: DGPS, differentially-corrected global positioning system • GIS, geographical information system • PC, principal component

	INTRODUCTION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
SITE-SPECIFIC SOIL MANAGEMENT is the process of managing soils based on localized conditions within field boundaries (Carr et al., 1991). Determining which soil factors to base management decisions on is often a complex process due to the interactions that affect crop yield. To be effective, management schemes must address both soil variability and soil properties limiting yield. Soil variability is caused by an assortment of different factors. Interactions among parent materials, topography, vegetation, tillage, fertilization, cropping history, and so on can influence the variability of the physical and chemical properties of fields. Thus, as fields may contain a variety of soil series and topographical features, variability in soil chemical and physical properties within field boundaries is much more common than are homogenous conditions.

In one study of nutrient variability and corn (Zea mays L.) yield in three Michigan soils, the CV for pH, P, K, Ca, and Mg ranged from 6% (pH) to 81% (Mg) (Pierce et al., 1994). Corn yield was highly variable, but little correlation was found between yield and soil fertility, indicating that other factors were influencing yield. A 1997 study had similar findings. Of eight soil fertility parameters tested, none were highly correlated with millet [Pennisetum glaucum (L.) R. Br.] yield (Stein et al., 1997). However, a negative correlation was found between yield and CEC, and the researchers attributed this correlation to higher clay and lower fertility soil exposed due to erosion.

Despite the variability in soil fertility parameters and the lack of correlation with yield, their importance to crop production cannot be overstated. Many researchers have proven the importance of soil fertility to soybean production (Barber, 1978; Hanaway and Olsen, 1980; Bharati et al., 1986; Marschner, 1995).

In addition to variability in soil chemical parameters, soil physical properties have also been found to vary widely. A wide range of variability was found in selected soil physical properties within two map units in east-central Illinois (Agbu and Olsen, 1990). In this study, CVs for slope and aspect were 93.5 and 83.6%, respectively, while the CV for pH in the B horizon was only 9.5%. Sand, silt, and clay contents also varied with CVs of 74.4, 22.9, and 23.6%. A combination of topographical features alone explained between 6 and 54% of the variability in corn and soybean yields in eight midwestern soils and, when combined with selected soil chemical properties, 10 to 78% of yield variability was explained (Kravchenko and Bullock, 2000). In this study, lower yields were found at higher slope positions while higher yields were found in lower slope positions. These findings supported the observations of Changere and Lal (1997) and McConkey et al. (1997). However, two studies found the opposite. Ciha (1984) found that higher wheat (Triticum aestivum L.) yields occurred on interfluve positions due to a deeper surface layer and better water retention than other slope positions. The second study found that eroded areas at higher elevations had higher yields due to additional plant available water being held by the higher clay levels in the eroded soils (Ebeid et al., 1995). Thus, the topographic effect on yield may be representative of plant-available moisture. Given the amount of variability in soil chemical and physical properties and their importance in determining crop yield, some property or combination of properties may serve as a basis for site-specific soil management.

The objectives of this study were to determine the variability of selected soil chemical and physical properties in three fields and to determine the relative influence of these soil properties on soybean yield.

	MATERIALS AND METHODS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
This study was conducted during 1998 and 1999 using grid sampling to determine how much variability existed in selected soil properties and how these properties related to soybean yield. Three soybean fields on the Mississippi Research and Extension Black Belt Branch Experiment Station located near Brooksville, MS, were chosen. The North field covered 8.4 ha, the South field covered 15 ha, and the East field covered 16 ha. The major soil series for all three fields was a Brooksville silty clay loam (fine, smectitic, thermic Aquic Hapluderts).

Each of these fields also contained small amounts of Okolona silty clay (fine, smectitic, thermic, Oxyaquic Hapluderts) and Demopolis silty clay loam (loamy, carbonatic, thermic, shallow Typic Udorthents) soils. These three soil series range from deep and somewhat poorly drained with very slow permeability to shallow and well drained. They are nearly level to gently sloping with slopes ranging from 0 to 5% and are formed in acidic silty clay or clay underlain by Selma Chalk (Brent, 1986). The climate is generally wet in the spring, with dry to drought-like conditions common during the late summer. Precipitation data for 1998 and 1999 are shown in Fig. 1. Before late fall 1997, the North field had a 25-yr history of tall fescue (Festuca arundinacea Schreb.) hay production with little or no fertilizer application. In the fall of 1997, the existing vegetation was incorporated to a depth of 15 cm to allow decomposition before planting. The South and East fields each had a 5-yr history of soybean production; however, highly eroded areas of the East field had exposed the underlying calcareous subsoil.

View larger version (25K): [in this window] [in a new window]   Fig. 1. Daily precipitation for the 1998 (yearly total = 122 cm) and 1999 (yearly total = 136 cm) growing years.

Yield data were collected with a properly calibrated commercial yield monitor and DGPS receiver mounted on a field combine. The yield monitor was configured to record yield data at 1-s intervals. After correction for the delay in the combine threshing mechanism, soybean yield at each soil sample location was determined by overlaying the raw data with the soil sample location in ArcView Geographical Information System (GIS) (ESRI, 1996). A 10-m radius buffer was created around each sampling point with the GIS software. Yield data for each soil sampling position were then calculated by selecting and averaging all of the yield points that intersected this buffer. This resulted in 8 to 12 yield data points being averaged for each soil sampling point.

Elevation data relative to a local benchmark were collected at each sample point using DGPS-equipped land surveying equipment with subcentimeter vertical accuracy. The elevation measurements were converted into cell-based digital elevation models using a spline function as an interpolation method. Slope values were derived from the elevation maps with ArcView Spatial Analyst (ESRI, 1996). Soil texture, elevation, and slope data were determined once in the fall of 1998.

The CV was used to describe the amount of variability for each soil parameter and yield for each year of this study. The yearly data for each field were then combined to remove the stochastic effect of years. Linear correlations among variables were calculated using PROC CORR of SAS (SAS Institute, 1996). Principal component analysis using PROC PRINCOMP procedure of SAS (SAS Institute, 1996) was used to address interpretation problems caused by correlation among some of the measured soil properties. Principal component analysis creates new, uncorrelated variables from highly correlated soil properties (Jolliffe, 1986). These new variables, or PCs, can often be interpreted through soil property variables that play a large role in their creation. Since the units of the variables were not identical, the correlation matrix was used to calculate the PCs. The resulting PCs were then used as independent variables in a stepwise regression procedure against yield. Since soybean yields could be due to a number of different soil conditions, PCs contributing up to 0.90 of the cumulative variance were included in the stepwise equation to avoid omitting possible yield limiting factors (Jolliffe, 1986). This criteria lead to 5 PCs being used in the North field, 4 PCs being used in the South field, and 5 PCs being used in the East field. The stepwise regression was then used to filter the PCs to result in a encompassing equation relating yield to measured soil parameters. Principal components had to meet a 0.15 significance level to be entered into the regression equation. It is also important to note that, despite having the same nomenclature (i.e., PC1, PC2...), the variables for each regression equation are independent of those for the other equations. Once the PCs that were significantly related to soybean yield were found, their meaning in terms of soil components had to be determined. Soil parameters with larger loadings in eigenvectors contribute more to the PC. To be included in the interpretation of the PC, loadings had to be greater than the value calculated with Eq. [1].

[1]

where SC = selection criterion (Ovalles and Collins, 1988).

Varietal effects were included in the regression equation to determine if yield was dependent on variety. No varietal effects on yield were found to be significant in any field for either year of this study.

	RESULTS AND DISCUSSION

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
General Fertility and Yield
As a result of the lack of fertilization in the North field, average P was 24% lower and average K was 43% lower compared with the other fields (Table 1). Calcium in this field was high, averaging 6589 mg kg^-1, reflecting the influence of the underlying calcareous parent material. Other physical and chemical properties of this field were comparable with those found in the other two fields in this study (Table 1). Despite low P and K concentrations, soybean yield in 1998 was more than twice that in the South and East fields. In 1999, the differences in yield were not as great as in 1998, yet yield in the South field (77%) and East field (83%) were still lower than in the North field. These differences in yield, despite lower P and K values, suggest that some factor or factors other than P and K fertility were influencing soybean yield in this field; therefore, soil management based on these P and K fertility levels would not necessarily result in a yield increase.

The East field also had a 5-yr history of soybean production. Highly eroded areas of the field exposed the underlying, calcareous subsoil. These areas are reflected in the high pH, averaging 7.3, and Ca levels, averaging 7809 mg kg^-1 (Table 1). Phosphorus averaged 27.0 mg kg^-1, while K averaged 115.85 mg kg^-1, indicating adequate levels of fertility for soybean production. Zero yield for both years of this study occurred in highly eroded areas of this field and is attributed to the high pH, infertile subsoil that was exposed. Plants that grew in these areas exhibited visual stunting and chlorosis as might be expected from the high pH values in these areas. As the zero yield is attributed to these soil factors, these data points were included in the statistical analysis. As with the South field, yield was low in 1998, averaging only 452 kg ha^-1 but increased to 763 kg ha^-1 in 1999. These P and K yield conditions suggest that factors other than soil fertility are influencing crop production in this field as well as the other two.

Variability of Soil Chemical and Physical Properties
Soil variability is a key element in site-specific soil management. Variability in space and time for point data can give valuable insight into the dynamic nature of soil properties within a field's boundary. Management of this variability is worthwhile if the amount is high enough to justify the costs of obtaining the information or if this management will increase profit.

Variability of the soil chemical factors in the North field, with the exception of pH, tended to be high and changed yearly (Table 1). In 1998, pH had the lowest variability (11%) and the Ca content had the highest (67%). While in 1999, pH again had the lowest variability (12%) and K emerged as having the highest variability (85%), increasing roughly 60% despite the mean K content of the field remaining relatively stable. This difference in K variability is difficult to explain. Small-scale variability is one possible explanation for this large increase in the CV. While the minimum K value for this field did not change greatly from 1998 to 1999, the maximum K value increased roughly 108 mg kg^-1 (Table 1). As the soil samples were collected from the same localized area for each sample point, this increase in the K range indirectly implies that small scale or large temporal variability is present. The clay content (CV = 65%) of this field also varied widely while the sand (CV = 25%) and silt content (CV = 11%) were relatively stable (Table 1). The variability of the elevation (CV = 50%), slope (CV = 60%), and aspect (CV = 71%) in this field was high. These high CV values reflect the rolling topography typical of this resource area.

The South field was not as variable as was the North field. Coefficients of variation for the chemical properties were lowest for pH (1998, 11%; 1999, 12%) and highest for Mg (1998, 37%; 1999, 35%) for both years of this study (Table 1). Elevation (CV = 28%), slope (CV = 47%), and aspect (CV = 53%) in the South field were substantially less variable than were those of the North field indicating much more level topography (Table 1). Sand (CV = 25%) and clay contents (CV = 28%) were relatively stable in this field while silt content had a low variability (CV = 11%) (Table 1).

Variability in the East field was similar to the South field (Table 1), with the exception of Ca. Calcium levels 1998 and CV values (1998, CV = 27%; 1999, 57%) more than doubled in 1999 (Table 1). We believe this large increase in Ca levels was due to field management. Seed bed preparations exposed the calcareous parent material due to the shallow topsoil in the eroded areas of this field. Examination of the site-specific data revealed large increases in the Ca levels at six of the soil sampling locations. These six locations were located in the eroded areas of the field and the calcareous parent material was evident on the soil surface in 1999.

Principal Component Analysis
The PC-stepwise regression analysis revealed that soybean yield in the North field was only related to one PC. The regression equation was: Yield = 14.63 - 3.69 PC2. The R² for this equation was 0.30, indicating that only some of the variability in yield could be explained by the measured parameters. The second PC was dominated by the positive influences of elevation and slope and the negative influences of Mg and explained 15% of the total variance (Table 2). Since the coefficient of the PC is negative, soybean yield was negatively related to elevation and slope, meaning that lower, flatter areas of the field had higher yields. These findings are similar to those of Kravchenko and Bullock (2000), Changere and Lal (1997) and McConkey et al. (1997). We speculate that this is an indication of soil-water relations during the growing season. Droughty conditions were present with only 47 cm of precipitation in 1998 and 67 cm in 1999 (Fig. 1) during the growing season. The lower, flatter areas of the field would retain water longer, and thus provide an increase in plant-available water. As Mg is a macronutrient, the positive relationship between these two factors and yield is expected. The contrast between Mg and elevation and slope based on the eigenvectors may be due to erosion processes moving the Mg to lower, flatter areas of the field; however, there is no correlation between these factors to confirm this theory (Table 3.)

Elevation, clay content, and P contributed significantly to the third PC, which explained 18% of the total variance (Table 4). The influence of elevation and clay content would be related to plant available water as explained in the first PC while increasing P levels in the soil should lead to higher yields. In comparing the results from this field and the North field, topography and clay content are common factors influencing soybean yields.

As in the North field, only one PC was related to soybean yield for the East field. The regression equation was Yield = 8.47 - 1.03 PC1 with an R² = 0.20. The PC contrasted the positive effects of K, Mg, and P and the negative effects of clay and explained 40% of the total variance (Table 6). As in the other two fields, the coefficient of the PC is negative; hence, the opposite effect of the variables must be considered. The negative relation between yield and K, Mg, and P (Table 7) are again most likely due to low-yielding areas of the field allowing a buildup of nutrients across time and the positive relation with clay content reflecting plant-available water characteristics of the field. When comparing all three fields, fertility and topography did not appear to be consistent soil factors affecting yield in this analysis. In the South field, both a positive and a negative relationship was found between fertility and yield and the soil fertility parameters had to be taken in context with other factors in the PC to determine their influence, while topography had contrasting effects between the North and South fields. The commonality of the influence of clay content in all three fields would lead to the suggestion that this parameter should be used to determine management zones. However, the 2 yr this study was conducted were dry for a portion of the growing season. In years with normal or above normal rainfall, the effect of the clay content could be opposite of that found in this study and, hence, cause reduced yields as suggested by Lindstrom et al. (1986).

	CONCLUSIONS

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

Soil variability, with the exception of pH, was highest in the North field. Potassium in this field exhibited evidence of high amounts of small-scale variability with CVs increasing 60% from 1998 to 1999 despite soil samples being collected from the same areas indicating either small scale variability or large amounts of temporal variability. Across all three fields, pH had the lowest amount of variability while the highest amounts of variability of the chemical properties varied from year to year and field to field.

While soil fertility was related to yield in all three fields used in this study, its effects were not consistent. Fertility parameters that appeared in the PCs had to be considered with other soil factors to determine their relationship with yield. Topographic relationships to yield varied from the North field to the South field and were not related to yield in the East field. Clay content was a common factor affecting yields in all three fields. In all three fields, areas with higher clay content had higher soybean yield, which was attributed to more plant-available water during dry periods of the growing season. However, in years with normal or above normal rainfall, the effect of the clay content could be opposite of that found in this study and, hence, cause reduced yields.

	NOTES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 
Contribution of the Mississippi Agric. and Forestry Exp. Stn. Journal no. J9691.

Received for publication March 28, 2002.

	REFERENCES

TOP NOTES ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION CONCLUSIONS REFERENCES

 

• Agbu, P.A., and K.R. Olsen. 1990. Spatial variability of soil properties in selected Illinois mollisols. Soil Sci. 150:777–786.
• Barber, S.A. 1978. Growth and nutrient uptake of soybean roots under field conditions. Agron. J. 70:457–461[Abstract/Free Full Text]
• Bharati, M.P., D.K. Whigham, and R.D. Voss. 1986. Soybean response to tillage and nitrogen, phosphorus, and potassium fertilization. Agron. J. 78:947–950.[Abstract/Free Full Text]
• Brent, F. 1986. Soil Survey of Noxubee County, Mississippi. USDA, U.S. Gov. Print. Office, Washington, DC.
• Carr, P.M., G.R. Carlson, J.S. Jacobsen, G.A. Nielsen, and E.O. Skogley. 1991. Farming soil, not fields: A strategy for increasing fertilizer profitability. J. Prod. Agric. 4:57–61.
• Changere, A., and R. Lal. 1997. Slope position and erosional effects on soil properties and corn production on a Miamian soils in central Ohio. J Sust. Agric. 11:5–21.
• Ciha, A.J. 1984. Slope position and grain yields of soft white winter wheat. Agron. J. 76:193–196.[Abstract/Free Full Text]
• Cox, M.S. 2001. The Lancaster soil test method as an alternative to the Mehlich 3 soil test method. Soil Sci. 166:484–489.
• Ebeid, M.M., R. Lal, G.F. Hall, and E. Miller. 1995. Erosion effects on soil properties and soybean yield of a Miamian soil in western Ohio in a season with below normal rainfall. Soil Technol. 8:97–108.
• Environmental Systems Research Institute. 1996. ArcView Spatial Analyst. ESRI, Redlands, CA.
• Gee, G.W., and J.W. Bauder. 1986. Particle size analysis. p. 383–411. In A. Klute (ed.) Methods of soil analysis: Part 1. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.
• Hanaway, J.J., and R.A. Olsen. 1980. Phosphate nutrition of corn, sorghum, soybeans, and small grains. p. 681–692. In F.E. Khasawneh et al. (ed.) The role of phosphorus in agriculture. ASA, CSSA, and SSSA, Madison WI.
• Jolliffe, I.T. 1986. Principal component analysis. Springer-Verlag. New York.
• Kravchenko, A.N., and D.G. Bullock. 2000. Correlation of corn and soybean grain yield with topography and soil properties. Agron. J. 92:75–83.[Abstract/Free Full Text]
• Lindstrom, M.J., T.E. Schumacher, G.D. Lemme, and H.M. Gollany. 1986. Soil characteristics of a Mollisol and corn (Zea mays L.) growth 20 years after topsoil removal. Soil Tillage Res. 7:51–62.
• Marschner, H. 1995. Mineral nutrition of higher plants. 2nd ed. Academic Press, San Diego, CA.
• McConkey, B.G., D.J. Ulrich, and F.B. Dyck. 1997. Slope position and subsoiling effects on soil water and spring wheat yield. Can. J. Soil Sci. 77:83–90.
• Ovalles, F.A., and M.E. Collins. 1988. Variability of northwest Florida soils by principal component analysis. Soil Sci. Soc. Am. J. 52:1430–1435.[Abstract/Free Full Text]
• Pierce, F.S., D.D. Warncke, and M.W. Everett. 1994. Yield and nutrient variability in glacial soils of Michigan. p. 133–150. In P.C. Robert et al. (ed.) Proc. 2nd Int. Conf. on Site-Specific Management for Agricultural Systems, Minneapolis, MN. 27–30 Mar. 1994. ASA, SSSA, and CSSA, Madison, WI.
• SAS Institute. 1996. SAS/STAT user's guide. Version 6. 4th ed. SAS Inst., Cary, NC.
• Stein, A., J. Brouwer, and J. Bouma. 1997. Methods for comparing spatial variability patterns of millet yield and soil data. Soil Sci. Soc. Am. J. 61:861–870.[Abstract/Free Full Text]

This article has been cited by other articles:

				M. S. Cox and P. D. Gerard Soil Management Zone Determination by Yield Stability Analysis and Classification Agron. J., September 10, 2007; 99(5): 1357 - 1365. [Abstract] [Full Text] [PDF]

				A. N. Kravchenko and G. P. Robertson Can Topographical and Yield Data Substantially Improve Total Soil Carbon Mapping by Regression Kriging? Agron. J., January 1, 2007; 99(1): 12 - 17. [Abstract] [Full Text] [PDF]

				H. Shahandeh, A. L. Wright, F. M. Hons, and R. J. Lascano Spatial and Temporal Variation of Soil Nitrogen Parameters Related to Soil Texture and Corn Yield Agron. J., April 27, 2005; 97(3): 772 - 782. [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 Cox, M. S. Articles by Abshire, M. J. Search for Related Content PubMed Articles by Cox, M. S. Articles by Abshire, M. J. Agricola Articles by Cox, M. S. Articles by Abshire, M. J. Related Collections Soil Fertility and Productivity Other Soil Management Sustainable Agriculture