| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a USDA-ARS, 120 Keim Hall, Lincoln, NE 68583-0934
b Dep. of Crop and Soil Sci., Pennsylvania State Univ., 116 ASI Building, University Park, PA 16802
* Corresponding author (cjohnso2{at}bigred.unl.edu)
Received for publication November 14, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: ECa, apparent electrical conductivity • ECDP, deep apparent electrical conductivity • ECSH, shallow apparent electrical conductivity • GIS, geographic information system • SSM, site-specific management
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The implementation of SSM requires real-time and accurate global positioning system (GPS) equipment, geographic information systems (GIS) for spatial analysis and mapping, variable-rate applicators, and input prescription maps to define management zones and direct metering devices controlling input rates (Eliason et al., 1995). While the first three components are currently available, the last, an effective and economical basis for defining site-specific inputs, is lacking. In response to this need, significant research effort has been directed toward evaluating a variety of individual and combined GIS databases as frameworks for identifying stratified within-field management zones (regions of similar production potential). These include kriged soil test point data (Mulla, 1991); soil survey maps (Robert, 1989); topography (Kravchenko et al., 2000); remote sensing (McCann et al., 1996); topography and remote sensing (Tomer et al., 1995); topography, remote sensing, and farmer experience (Fleming et al., 1999); electrical conductivity sensors (Sudduth et al., 1997; Lund et al., 1999); and yield maps (Eliason et al., 1995; Stafford et al., 1999). These approaches to SSM have met with varying degrees of success that are often highly soil or region specific.
Because some factors affecting crop yields occur unpredictably, including weather, human error, and equipment malfunction (operator error, plugged spray nozzles or planters, herbicide drift, weed pressure, poor seed viability, etc.), the potential impact of SSM may be limited in some years. At best, it will optimize the interactions between soil and inputs of nutrients, seed, or pesticides by targeting soil indices related to production potential that are measurable, relatively stable, and manageable. The productivity of a given soil is determined by the cumulative effect of natural factors involved in its formation, including climate, topography, parent material, biological activity, and time (Jenny, 1941), and management history. Management history can significantly affect the range and spatial heterogeneity of soil chemical properties beyond that attributable to natural processes. This is particularly true in organic systems where input applications are typically less uniform than in conventional systems (Cambardella and Karlen, 1999).
While variations in individual soil factors have limited utility for SSM, their combined impact on water and nutrient use efficiency is highly relevant to both production potential and environmental concerns, such as NO3 leaching (Bouma and Finke, 1993) and soil acidification (Malhi et al., 1991). Fields can be mapped for multiple soil parameters using intensive grid sampling and interpolation. However, such techniques are often economically unfeasible, particularly in semiarid regions with predominately large-scale, dryland, low-input farms (McCann et al., 1996). For SSM to be cost effective in these regions, a surrogate measure is required, an external means for integrating and stratifying soil attributes associated with productivity. Whatever their derivation, prescription maps for SSM must satisfy two criteria. First, a strong relationship must exist between identified management zones and ground-truth soil test data, encompassing soil physical, chemical, and biological parameters underlying yield potential. Second, zones must be temporally consistent, given normal fluctuations in dynamic soil properties such as moisture and temperature. Ideally, farmers should be able to employ field-specific SSM prescription maps not only across seasons, but also for several years before re-evaluation.
One mapping option showing promise for SSM is soil ECa. Depending on the soil factor(s) dominating measured ECa and the strength of the relationship between the factor(s) and other soil characteristics, ECa may function as a direct and/or indirect indicator of multiple soil parameters (Sudduth et al., 1995; Doolittle et al., 1994; Jaynes et al., 1995b). For some soils, ECa mapping appears to integrate soil parameters related to productivity to produce a template of potential yield (Jaynes et al., 1993; Sudduth et al., 1995; Kitchen et al., 1999). Because most ECa research has been conducted in humid areas of the United States with high rates of precipitation, very little is known about the relationships among ECa, soil properties, and crop yields in semiarid regions.
In a farm-scale study in semiarid northeastern Colorado, Johnson et al. (2001) found that management zones based on ECa mapping (approximately 0–30 cm depth of measurement) provide a useful basis for soil sampling. Such zones also fulfill the first criteria for SSM prescription maps by effectively delineating within-field regions of varying production potential. Other published research indicates that spatial patterns in ECa do not change with temporal variation in soil moisture and/or temperature (Lund et al., 1999; Sudduth et al., 2001). These findings advance ECa mapping as a basis for SSM.
The primary objective of this study was to examine the relationships between ECa–based management zones and crop yields in a 4-yr crop rotation in the semiarid central Great Plains. Specifically, 2 yr of yield maps from two fields each of corn and winter wheat, two depths of ECa measurement (approximately 0–30 cm and 0–90 cm), and two methods for stratifying ECa into management zones (unsupervised and equal-size classification) were evaluated. A secondary objective was to consider the significance and potential application of ECa–based management zones for SSM in a semiarid cropping system. This second objective was supported by previous findings regarding the relationships between ECSH and soil physical, chemical, and biological properties (Johnson et al., 2001).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
In 1999, the experimental site was converted from a conventionally tilled wheat–fallow system to a no-tillage, high-intensity winter wheat–corn–proso millet (Panicum miliaceum L.)–fallow rotation. Each of the four crop treatments is applied to two fields (approximately 31 ha each) in a given year (Fig. 1) .
|
Yield maps were taken from corn and wheat fields in 1999 (Fields no. 3 and 6 for corn and Fields no. 5 and 8 for wheat) and 2000 (Fields no. 5 and 8 for corn and fields no. 1 and 4 for wheat) (Fig. 1). Data collected with a Micro-Trak grain yield monitor (Micro-Trak Syst., Eagle Lake, MN)1 were verified with grain weigh-ticket information (total bushels, moisture content, and test weight) and mapped using Farm HMS software (Red Hen Syst., Fort Collins, CO).1
Mapping and Classification of Apparent Soil Electrical Conductivity
The entire site was ECa–mapped in March 1999 by direct contact (15-m swath width at a speed of 4.5 m s-1), using a Veris 3100 Sensor Cart (Veris Technol., a division of Geoprobe Syst., Salina, KS)1 (Fig. 1). Calibrations were performed according to manufacturer instructions. Latitude, longitude, and ECSH and ECDP readings (mS m-1) were recorded at 1-s intervals by the Veris datalogger. Values of ECa were converted to dS m-1 for reporting. All fields were uncropped at the time of mapping, except for Fields 5 and 8 (Fig. 1), which were planted to winter wheat.
The ECa raw (point) data files were projected to UTM (Universal Transverse Mercator) coordinates in the NAD83 datum (North American Datum of 1983) in ArcInfo (ESRI, Redlands, CA).1 They were then interpolated by inverse-distance weighting using the nearest-neighbor technique and redefined as grid files (10-m grid cell resolution). Four management zones based on ranges of ECSH and ECDP (low, medium low, medium high, and high) were determined for each of the eight fields comprising the study site using a method called equal-size classification. The number of ECSH or ECDP grid cells in each field was tallied and divided by four to identify four zones approximately equal in area.
In addition to equal-size classes, ECSH maps were classified using a second method termed unsupervised classification (ERDAS, 1997). 1 Unsupervised classification uses an iterative process to group clusters of statistically similar data. The ECSH maps from each of the eight fields in the study site were individually interpolated by inverse-distance weighting and classified using unsupervised classification to form 12 classes within each field at a 10-m grid cell resolution. The 12 classes were then recoded into four ranges of ECSH: low, medium low, medium high, and high. Recoding is a highly subjective process wherein the 12 original unsupervised classes of ECSH were combined to mimic the dominant visible spatial patterns seen in the original gray-scale ECSH maps (Fig. 2) . Through this process, ECSH measurements (pixels) were aggregated into naturally occurring clusters that may reduce within-zone variance.
|
|
|
Both ECSH and ECDP were compared with winter wheat and corn yield maps for significant associations using ANOVA and regression techniques. Relationships between yield and ECa–classified data layers were assessed by ANOVA for a randomized complete block design with ECSH or ECDP zones (identified using equal-size and/or unsupervised classification) as a treatment factor. All statistical analyses were performed using SAS (SAS Inst., 1997),1 and differences were declared significant at the 0.05 level, unless stated otherwise.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
The relationship between ECSH and yield, while applicable to wheat, did not hold across crops for the 2 yr evaluated. Although ANOVA showed significant (P 
0.0001) associations between ECSH and yield for three of four corn fields evaluated (Field 3 in 1999 and Fields 5 and 8 in 2000), these relationships were not consistent (Table 2). When regressing mean corn yields within ECSH class against mean ECSH within ECSH class, only Field 8 showed significant linearity (r2 = 0.90 with equal-size classification) (Fig. 4).
|
The disparity in relationships between ECSH and yields of corn and wheat may simply reflect differences in crop response to both the soil factors contributing to measured ECSH and to other soil factors with which they are correlated. It has been documented that factors affecting yield variability may differ among crops (Vieira, 1999). Yet, different water stress levels between corn and wheat across the 2 yr studied and limitations in ECSH effectiveness for delineating soil factors associated with root-zone water availability may better explain the divergence in ECSH–yield relationships for corn and wheat. Because its growing season corresponds well with precipitation patterns for the region, wheat is a more suitable crop for the central Great Plains than is corn. In addition, greater soil water storage due to the rotational sequencing of wheat after fallow improves water availability to that crop. These factors reduce water stress on wheat crops to benefit yield and yield consistency across years. Previous experiments describing the soil characteristics of ECSH–delineated zones indicate that ECSH is highly correlated with soil water, organic matter content, and total C and N, all indicators of improved water-holding capacity (Johnson et al., 2001). These surface soil characteristics appear to be essential determinants of winter wheat yield in the typical and high-yielding years encountered during this study.
Conversely, the corn growing season spans a period of low precipitation and high evaporative demand that diminishes water availability and water use efficiency. Furthermore, corn follows wheat in the rotation under study, making yields more susceptible to annual variations in precipitation. For these reasons, corn yields in the central Great Plains are particularly vulnerable to water stress during a 6-wk period between 15 July and 25 August; precipitation rates within this period explain 70% of the variability in corn yields (Nielsen, 1996). Corn crops in both 1999 and 2000 were highly drought stressed during this critical time. Consequently, yields were low in 1999, averaging 2092 and 2654 kg ha-1 (Fields 3 and 6), and poor in 2000, averaging 1386 and 1621 kg ha-1 (Fields 5 and 8). Climatic influences, particularly variability in precipitation timing and quantity, appear to diminish or confound the impact that underlying surface soil characteristics, integrated by ECSH, have on corn yields. It is possible that, in a year with higher precipitation during July and August, corn yields would present the same negative association with ECSH identified for wheat. Further studies are required to investigate this possibility.
Other investigators have found highly variable relationships between corn yields and various methods for delineating management zones. A study in the southeastern USA by Sadler et al. (1995) examined correlations between corn yields and soil condition delineated by soil survey map units. They documented the variable response of corn yield, rainfall partitioning, and water use efficiency both among and within map units in dry years. In western Iowa, Jaynes et al. (1995a) found inconsistent correlations between corn and soybean yields and electrical conductivity (approximately 0–150 cm depth of measurement) in high- and low-precipitation years, both among fields and across years. They hypothesized that this resulted from opposing responses among electrical conductivity–delineated zones, to low vs. excessive soil moisture, causing heightened yield variability within fields. In semiarid regions, yield reductions from excessive precipitation are rare. Thus, electrical conductivity–delineated management zones may be a more reliable indicator of corn yield potential and, therefore, a more useful basis for SSM in these regions than in those receiving higher rates of precipitation.
Measured soil ECa is a function of salinity, clay type and percentage, water content, bulk density, and temperature (Rhoades et al., 1989; McNeill, 1980). Findings by Johnson et al. (2001) indicate that the primary drivers of ECSH at the Farm-Scale Intensive Cropping Study are clay content and CaCO3 salts (contributing to increased soil pH at high ECSH) (Table 1). Increases in these soil properties are characteristic of eroded parts of a field. Therefore, a negative correlation exists between ECSH and soil characteristics associated with yield potential. Conversely, decreases in surface clay content and CaCO3 correlate with other soil properties that improve water-holding capacity, nutrient exchange, and plant rooting depth (increase yield potential).
Because soil assessments corresponding to ECDP were not made, we must hypothesize the reason(s) for the reversal in relationship between (wheat and corn) crop yields and ECDP compared with that between wheat yields and ECSH. Because water is the greatest limiting factor to crop production in the central Great Plains, positive correlations between yield and ECDP must be linked to soil water content, particularly given that corn crops were highly drought stressed during both years examined. These findings indicate that ECDP is driven by soil water content, salts in the soil solution, and clay content, which at 0- to 90-cm depths likely both contributes to measured ECa and correlates with soil water-holding capacity. Probable factors controlling salinity of the soil solution are residual NO3 and NH4 and CaCO3. Measured ECDP may also reflect soil depth to lime present in the C horizon. Further research is needed to test associations among soil characteristics, crop yields, and ECDP.
Unsupervised versus Equal-Size Shallow Apparent Electrical Conductivity Classification
Classification is the partitioning of soil into regions of similar production potential as a means to describe within-field variability and create management zones. Clearly, for this study site, there exists a strong linear relationship between ECSH and wheat yields, allowing for the identification of ECSH–delineated management zones based on ranges of ECSH. Yet, how should these ranges be assigned? Classes may be defined using threshold values of either soil properties critical to productivity or regions of differing soil morphology (Lark, 1997). If these threshold values delineate levels of intrinsic soil fertility, they may present an ideal basis for SSM. Van Uffelen et al. (1997) applied a weighted distance measure to identify patterns in simulated yield maps for this purpose. Lark and Stafford (1997) used fuzzy multivariate clustering analysis to define heterogeneous management zones based on 3 yr of actual yield data. Yet, for farmers to adopt SSM, the development of management zones must be simple, functional, and economically feasible. Complex field assessments and data manipulation may not be justifiable in terms of time, benefit, or economics.
Unsupervised classification, one method for identifying ECa management zones, is based on the assumption that grouping ECa data points into naturally occurring clusters (ranges) will reduce within-zone yield variability. It represents a simplified approach for identifying threshold parameters related to yield potential. Table 3 compares unsupervised and equal-size classification methods. For three of four wheat fields evaluated, ECSH management zones derived from unsupervised classification showed increased F values over equal-size methods. Comparison of F values offers a rough estimate of the power of the two methods to separate yields. For the same three fields, unsupervised classification decreased variance within ECSH zones compared with equal-size methods. Fraisse et al. (2001) compared whole-field yield variance (one management zone) to yield variances calculated when fields were divided into one to six management zones using unsupervised classification. Yield variances associated with different numbers of management zones were expressed as a percentage of whole-field yield variance. Percentages were compared to determine which number of zones best reduced within-zone variance (increased between-zone variance) as a means to identify the optimal number of management zones for each field. Applying this approach to compare unsupervised and equal-size classification methods revealed a 0 and 5% reduction in yield variance with unsupervised classification (Table 3). Thus, differences in wheat yield partitioning due to classification method are subtle (Table 3 and Fig. 4), an indication that either unsupervised or equal-size classification methods may be acceptable bases for SSM.
|
|
The ECSH map collected from the experimental site used in this study documents the variability of ECSH among fields (Fig. 1). Measured ECSH is a reflection of both historic and recent management in each field. For example, the V-shaped patterns in each of the four corner fields are believed to be the result of around-and-around plowing of the east and west half sections of the site in the 1930s. Recent management history can be visually discerned by differences in the gray-scale map (differences in the magnitude of ECSH) among fields. These differences likely reflect variations in soil water and nutrient status due to uptake by the previous year's crop. For this reason, the application of equal-range classification to each field would preclude the association of ECSH zones across field boundaries. Conversely, the ECSH map classified by the unsupervised method (Fig. 3) shows a reasonable degree of continuity among ECSH zones in adjacent fields, indicating some normalization of measured ECSH.
For cropping systems using multiple-year rotations, such as the one studied at this experimental site, several years fall between times when a given crop is grown in a given field. This factor, in addition to climatic variability among years, requires the collection of data over several rotations (at least 10 or 12 yr) to build databases for the SSM of specific crops grown within individual fields. The use of unsupervised classification may allow data collected in multiple fields to be combined for extrapolation to many sites within and across farms.
Implications of Wheat Yield–Shallow Apparent Electrical Conductivity Relationships for Site-Specific Management
The ECa–yield relationships identified in this paper, coupled with ECa–soil attribute correlations identified in previous experiments (Johnson et al., 2001), provide information essential to the application of SSM in this semiarid cropping system. Effective SSM requires management zones that function: (i) to delineate variations in soil nutrient status and production potential; (ii) as a framework for metering inputs of fertilizer, pesticide, and seed; and (iii) as a basis for identifying yield goals or variations in yield potential across a field.
For a classification system to be effective, it should divide a field into zones delineating the same or similar yield-limiting factors (Lark and Stafford, 1997). Previous experiments at the Farm-Scale Intensive Cropping Study site reveal that management zones based on the unsupervised classification of ECSH effectively delineate soil characteristics related to productivity. Soil organic matter, total C and N, extractable P, water content, and microbial biomass were negatively correlated with ECSH and significantly different among ECSH zones. Because these zones partition both soil nutrients, N and P, and other soil characteristics related to nutrient and herbicide availability and production potential, they fulfill the first requirement for SSM, a useful basis for soil sampling to assess residual nutrients and soil attributes affecting herbicide efficacy. Second, the strong ECSH–yield–soil attribute relationships documented in this study support the employment of ECSH–based management zones as a framework for metering fertilizer, pesticide, and seed inputs.
Yield response curves can be used as a means to identify maximum crop yields, also known as the boundary line (Webb, 1972), by regressing yield against ECa. Kitchen et al. (1999) used this type of boundary-line analysis to estimate the magnitude of yield suppression for various crops given different weather and soil conditions; however, they stopped short of suggesting its use for yield goal setting in SSM. This is because ECa–yield relationships were highly variable for the claypan soils evaluated in north-central Missouri. Depending on the crop, weather conditions, and soil characteristics, yields increased with increasing ECa, decreased with increasing ECa, peaked at midrange values of ECa, or showed no relationship to ECa.
Although only 2 yr of data were evaluated, the strong and consistent negative correlations between ECSH and wheat yields and positive correlations between ECDP and yields of both corn and wheat support greater potential application for boundary-line analyses in this dryland cropping system. Yield data collected in a field in an above-average year may serve as an indicator of maximum potential yield, for that crop in that field, when expressed as a function of ECa. Given the 2 yr of available data for wheat, the regression of 1999 wheat yields against ECSH best portrays within-field wheat yield variability and yield potential (Fig. 6A) . Data from the two winter wheat fields in 1999 were combined for this analysis to increase the database size for future SSM application to the entire Farm-Scale Intensive Cropping Study site and potentially to nearby fields of similar soil type and topography. In highly heterogeneous soils, it may be important to collect yield data from specific fields for application to only those fields.
|
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
However, complementary data layers including an ECSH–classified map, ground-truth soil test information, and accumulated yield maps appear to address both actual yield and intrinsic soil productivity factors essential to establishing appropriate management zones for the SSM of winter wheat in this semiarid system. Weather influences on crop yield variability tend to be relatively straightforward in semiarid systems where yield variability largely reflects varying degrees of drought stress. For this reason, soil–yield–ECa relationships may be a more stable across years than is true for areas receiving higher rates of precipitation.
Comparison of winter wheat yields and ECa showed strong correlations between yield and both ECSH and ECDP measurement depths in the average and above-average–yielding years encountered in this study. Yields were negatively correlated with ECSH and positively correlated with ECDP. Zone treatments based on ECSH showed better yield discrimination than those based on ECDP. Although ECa–yield relationships were not evaluated for a year with below-average yields, winter wheat yields tend to be reasonably consistent in the central Great Plains when wheat follows a fallow year. Field maps separated into management zones based on unsupervised ECSH classification showed slight improvement in zone partitioning (decreased within-zone variance) over equal-size classification. Thus, the unsupervised classification of ECSH appears to be the best basis for management zone identification in winter wheat. Future research is required to determine to what degree these results can be generalized to a regional scale.
Strong correlations among ECSH, soil properties, and wheat yields indicate that ECSH–delineated management zones provide an excellent framework for SSM of dryland winter wheat. Management zones based on the unsupervised classification of ECSH serve three functions essential to SSM. They provide: (i) a basis for soil sampling to assess nutrient levels and soil attributes affecting herbicide efficacy; (ii) a means to calculate nutrient inputs by using boundary-line analyses between ECSH and wheat yield to identify yield goals (variation in maximum yield potential across a field); and (iii) a prescription map for metering fertilizer, pesticide, and seed inputs. The first two functions are essential for determining fertilizer, herbicide, and seeding rates within management zones while the last delineates the zones to which they will be applied.
Similar potential exists for SSM of corn. Within-zone corn yield means for the below-average and low-yielding years studied showed no consistent relationship with ECSH but strong positive correlation with ECDP. Because corn yields are highly variable in the region studied, even if strong correlations are found between ECSH and corn yields in high-yielding years, ECDP may offer a more realistic basis for the establishment of SSM zones for corn. Further research is required to identify the soil factors driving measured ECDP and characterize soils falling within ECDP–based management zones. The continued collection of winter wheat and corn yield maps from this study site over a number of low-, average-, and high-yielding years, and the evaluation of these data in conjunction with ECSH and ECDP, will improve our ability to establish appropriate management zones for both wheat and corn in this semiarid setting.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
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]
|
|
|
|
|
|
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]
|
|
|
|
 |
|
 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]
|
|
|
|
|
|
A. Hornung, R. Khosla, R. Reich, D. Inman, and D. G. Westfall Comparison of Site-Specific Management Zones: Soil-Color-Based and Yield-Based Agron. J., March 2, 2006; 98(2): 407 - 415. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Flowers, R. Weisz, and J. G. White Yield-Based Management Zones and Grid Sampling Strategies: Describing Soil Test and Nutrient Variability Agron. J., May 13, 2005; 97(3): 968 - 982. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Inman, R. Khosla, D. G. Westfall, and R. Reich Nitrogen Uptake across Site Specific Management Zones in Irrigated Corn Production Systems Agron. J., January 1, 2005; 97(1): 169 - 176. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. R. Schepers, J. F. Shanahan, M. A. Liebig, J. S. Schepers, S. H. Johnson, and A. Luchiari Jr. Appropriateness of Management Zones for Characterizing Spatial Variability of Soil Properties and Irrigated Corn Yields across Years Agron. J., January 1, 2004; 96(1): 195 - 203. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. K. Johnson, K. M. Eskridge, B. J. Wienhold, J. W. Doran, G. A. Peterson, and G. W. Buchleiter Using Electrical Conductivity Classification and Within-Field Variability to Design Field-Scale Research Agron. J., May 1, 2003; 95(3): 602 - 613. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||
symbols represent selected soil-sampling sites.
