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 Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Inman, D. Articles by Westfall, D. G. Search for Related Content PubMed Articles by Inman, D. Articles by Westfall, D. G. Agricola Articles by Inman, D. Articles by Westfall, D. G. Related Collections Maize Maize Maize Remote Sensing Soil Fertility and Productivity Maize Management

SITE-SPECIFIC ANALYSIS & MANAGEMENT

Normalized Difference Vegetation Index and Soil Color-Based Management Zones in Irrigated Maize

^a Dep. of Soil & Crop Sciences, Plant Science Bldg., Colorado State Univ., Fort Collins, CO 80523-1170
^b Dep. of Forest, Rangeland and Watershed Stewardship, Colorado State Univ., Fort Collins, CO 80523-1170

^* Corresponding author (Raj.Khosla{at}colostate.edu).

	ABSTRACT

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

 
Spectral vegetation indices such as the normalized difference vegetation index (NDVI) have been shown to be useful for indirectly obtaining crop information such as photosynthetic efficiency, productivity potential, and potential yield. The objectives of this study were (i) to examine the relationships among NDVI determined early in the growing season, soil color-based management zones (SCMZ), and relative maize (Zea mays L.) grain yield and (ii) to determine if coupling soil color-based management zones with NDVI improves the accuracy of soil color-based management zone precision crop management strategy. Remotely sensed imagery was acquired by aircraft at approximately the eight-leaf crop growth stage (V8). Kappa statistics and percent areal agreement suggested a slight to substantial areal association among NDVI and relative grain yield (K = 0.10 to 0.63; % areal agreement = 13–67). Regression models were variable and explained among 25 to 82% of the variability in relative grain yield. Inclusion of soil color-based management zones in the regression models resulted in marginal improvements. When the affects of soil color-based management zones were removed, NDVI accounted for among 10 to 47% of the variability. The NDVI determined early does have potential to be useful in irrigated maize cropping systems. Coupling NDVI and SCMZs did not bring additional benefits to our soil color-based management zone strategy.

Abbreviations: DGPS, differentially corrected global positioning system • GIS, geographic information system • GPS, global positioning system • LAI, leaf area index • NDVI, normalized difference vegetation index • NIR, near-infrared • NUE, nitrogen use efficiency • PVI, perpendicular vegetation index • SAVI, soil adjusted vegetation index • SCMZs, soil color-based management zones

Normalized Difference Vegetation Index and Soil Color-Based Management Zones in Irrigated Maize

^a Dep. of Soil & Crop Sciences, Plant Science Bldg., Colorado State Univ., Fort Collins, CO 80523-1170
^b Dep. of Forest, Rangeland and Watershed Stewardship, Colorado State Univ., Fort Collins, CO 80523-1170

^* Corresponding author (Raj.Khosla{at}colostate.edu).

Received for publication January 13, 2007.
Spectral vegetation indices such as the normalized difference vegetation index (NDVI) have been shown to be useful for indirectly obtaining crop information such as photosynthetic efficiency, productivity potential, and potential yield. The objectives of this study were (i) to examine the relationships among NDVI determined early in the growing season, soil color-based management zones (SCMZ), and relative maize (Zea mays L.) grain yield and (ii) to determine if coupling soil color-based management zones with NDVI improves the accuracy of soil color-based management zone precision crop management strategy. Remotely sensed imagery was acquired by aircraft at approximately the eight-leaf crop growth stage (V8). Kappa statistics and percent areal agreement suggested a slight to substantial areal association among NDVI and relative grain yield (K = 0.10 to 0.63; % areal agreement = 13–67). Regression models were variable and explained among 25 to 82% of the variability in relative grain yield. Inclusion of soil color-based management zones in the regression models resulted in marginal improvements. When the affects of soil color-based management zones were removed, NDVI accounted for among 10 to 47% of the variability. The NDVI determined early does have potential to be useful in irrigated maize cropping systems. Coupling NDVI and SCMZs did not bring additional benefits to our soil color-based management zone strategy.

Abbreviations: DGPS, differentially corrected global positioning system • GIS, geographic information system • GPS, global positioning system • LAI, leaf area index • NDVI, normalized difference vegetation index • NIR, near-infrared • NUE, nitrogen use efficiency • PVI, perpendicular vegetation index • SAVI, soil adjusted vegetation index • SCMZs, soil color-based management zones

	INTRODUCTION

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

 
REMOTE SENSING has long been recognized as a tool to rapidly acquire crop information such as disease incidence and variety differences (e.g., Colwell, 1956; Hoffer, 1967). Leaves absorb approximately 40% of all incident solar radiation (Campbell, 2002) and photosynthetic pigments dominate the visible light reflectance spectrum of leaves. Chlorophyll-a absorption maxima are at 440 nm (blue) and 670 nm (red). Reflectance of the near-infrared (NIR) region is primarily affected by the internal structure of the leaf (i.e., lignin and cellulose present in the mesophyll cells) (Campbell, 2002). As plants develop normally, reflectance in the red wavelength region is reduced (i.e., increased absorption by chlorophyll) whereas reflectance in the NIR region is increased (i.e., more mesophyll cells being developed). Although chlorophyll-a has an absorption maximum at 440 nm, blue light reflectance is generally not used because of its susceptibility to Rayleigh scattering (Campbell, 2002). Using ratios that make use of chlorophyll absorption and NIR reflectance, the relative health of the plant and/or crop may be inferred. Chlorophyll density (µg chl cm^–3) of the leaf is directly affected by plant N content and availability.

Spectral vegetation indices such as the NDVI have been shown to be useful for indirectly obtaining crop information such as photosynthetic efficiency, productivity potential, and potential yield (Peñuelas et al., 1994; Thenkabail et al., 2000; Ma et al., 2001; Raun et al., 2001; Báez-González et al., 2002, Teal et al., 2006). The NDVI is a broadband index that is well correlated to leaf area index (LAI) and green biomass (Peñuelas et al., 1994), and is thus sensitive to photosynthetic efficiency (Aparicio et al., 2002). The NDVI has been shown in some studies to be useful for estimating grain yield in certain crops. Raun et al. (2001) showed expected yield determined from NDVI had a strong relationship with actual grain yield in winter wheat (Triticum aestivum L.), r² = 0.83, P > 0.0001. Ma et al. (2001) reported that NDVI could be used to reliably predict low and high yield in soybean (Glycine max L.).

One commonality in many studies that use NDVI, or similar vegetation indices to estimate grain yield, is that the estimates are better when LAI is low (i.e., during the early growing season or during senescence). The NDVI is more sensitive to changes in the crop canopy when the LAI is low, eventually becoming saturated as the crop canopy closes. Differencing indices such as the NDVI work well for early-season yield estimation in crops such as wheat because the row spacing minimizes soil background reflectance. Because of the wide row-spacing (i.e., 70 cm) used in maize crops, soil background can mask the reflectance of the crop canopy. Confounding this problem is the fact that when the maize canopy has closed enough that soil reflectance is no longer an issue, the NDVI is saturated and cannot be used. Soil reflectance is affected primarily by texture, mineral composition, organic matter, and moisture. In general, soil is a strong reflector of light in the red and NIR regions of the spectrum. Studies have suggested that vegetation indices should be adjusted to minimize the effects of soil reflectance (Baret et al., 1988; Huete, 1989; Rondeaux et al., 1996). The soil-adjusted vegetation index (SAVI) developed by Huete (1989) uses a single constant (L) to adjust for soil interference; this constant changes with changes in percent canopy closure. Use of a constant correction factor has limitations when applied to fields that have spatially variable soil albeto. Recent research suggests that spatial variability of several soil parameters (e.g., bulk density, soil texture, organic matter, and soil moisture) can be characterized using SCMZ (Mzuku et al., 2005).

Management Zones
Doerge (1999) defined management zones as being homogeneous subregions of a field that have similar yield-limiting factors. Several management zone delineation techniques have been proposed in the literature for various cropping systems. Regardless of the method used to delineate management zones, they are used to classify a field into manageable units that differ in productivity potential. For example, a field may be classified into three zones: high, medium, and low productivity potential management zones. Using the management zones approach, agricultural inputs are envisaged to be applied variably across the field in accordance with the productivity potential of the management zones. However, within a management zone agricultural inputs are applied uniformly at a constant rate. The SCMZs are one of many techniques to characterize in-field spatial variability to facilitate precision nutrient management. Studies have shown SCMZs to be a simple and effective way to characterize spatial variability and increase N use efficiency (NUE) (Khosla et al., 2002; Hornung et al., 2006; Fleming et al., 2004; Inman et al., 2005a).

Although SCMZs have been shown to be an effective tool for developing precision nutrient management strategies in irrigated maize, using in-season remote sensing to characterize the spatial variability of the crop could lead to improvements in our existing SCMZ precision nutrient management strategy. The NDVI has been shown to be negatively affected by soil reflectance, however because SCMZs are delineated based on bare-soil imagery, coupling the two may reduce and/or account for the soil influence on the NDVI. We hypothesize that using SCMZs in conjunction with NDVI could improve the accuracy of the SCMZ precision nutrient management strategy. The objectives were to (i) examine the relationships among NDVI determined early in the growing season, SCMZ, and relative maize yield; and (ii) determine if coupling SCMZ with NDVI compliments and improves the SCMZ strategy.

	MATERIALS AND METHODS

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

 
Study Sites
This study was conducted over six site-years on three irrigated continuous maize fields in eastern Colorado. Site-years I and II were furrow irrigated fields located on a small experimental farm over two consecutive growing seasons (2002 and 2003, respectively). Site-years III and IV were center-pivot irrigated and located on a commercial maize field over two consecutive growing seasons (2001 and 2002, respectively). Site-years V and VI were center-pivot irrigated and located on a commercial maize field over two nonconsecutive growing seasons (2001 and 2003, respectively).

Site-years I and II were mapped as having Fort Collins loam (fine-loamy, mixed, superactive, mesic Aridic Haplustalf) (Soil Survey Staff, 1980). The Fort Collins loam is characterized as being a very deep, well-drained soil that forms in mixed eolian sediments and alluvium. The Fort Collins series occurs on terraces, hills, plains, and alluvial fans.

Site-years III and IV were mapped as having Bijou (coarse-loamy, mixed, superactive, mesic, Ustic Haplargid), Truckton (coarse-loamy, mixed, superactive, mesic, Aridic Argiustoll), and Valentine (mixed, mesic, Typic Ustipsamment) soil series (Soil Survey Staff, 1968). These soils are characterized as being very deep. The Truckton is well drained, Bijou is somewhat excessively drained, and the Valentine is excessively drained. Both Truckton and Bijou soils are derived from arkose parent material and occur on terraces, fans, and uplands. Valentine soils are eolian derived and occur on uplands.

Site-years V and VI were located on a field that was mapped as having Albinas (fine-loamy, mixed, superactive, mesic Pachic Argiustoll), Ascalon (fine-loamy, mixed, superactive, mesic, Aridic Argiustoll), and Haxtun (fine-loamy, mixed, superactive, mesic Pachic Argiustoll) soil series (Soil Survey Staff, 1981). These soils are characterized as being very deep, well drained, and have accumulated carbonates in the soil solum. The Ascalon series occurs on upland positions and is formed from calcareous parent material. The Haxtun series consist of eolian deposits that overlay buried soil, occurring in drainages and depressions. The Albinas series is alluvial and occurs on fans and terraces.

Experimental Procedure
Site-years I and II were planted with Garst hybrid 8802 at a row spacing of 76 cm and a planting density of 76,500 plants ha^–1. Site-years III and IV were planted with Pioneer hybrid 34G81 at 83,000 plants ha^–1 with a row spacing of 76 cm. Site-years V and VI were planted with Pioneer hybrid 33B50 at 84,000 plants ha^–1 with a row spacing of 76 cm. The SCMZs were delineated using the commercially available AgriTrak Professional software (Fleming et al., 1999). This program relies on three geographic information system (GIS) data layers: (i) bare soil aerial imagery on conventionally tilled land; (ii) farmer's perception of field topography; and (iii) farmer's past crop and soil management experience. These data layers were incorporated into a MapInfo (GIS) data-base (MapInfo Corp., Troy, NY) to run mathematical interpolation surfaces to develop three management zones (Khosla et al., 2002). Traits such as dark color, low-lying topography, and historic high yields were designated as a zone of potentially high productivity or high zone. Details of this technique are provided in Fleming et al. (1999), Khosla et al. (2002), and Fleming et al. (2004).

For each study site, remotely sensed imagery was acquired by aircraft at the six and eight-leaf crop growth stage (V6–V8) (Ritchie et al., 1993). Imagery recorded radiance in three bands (green, red, near-infrared) using a DuncanTech MS 3100 (Redlake MASD Inc., San Diego, CA). Geometric correction was performed in ERDAS Imagine 8.6 (Leica Geosystems GIS & Mapping LLC, Atlanta, GA) using image-to-image registration; root mean square error was <1 pixel for all images. Radiance was converted to apparent reflectance by using the histogram minimization method of Chavez (1975) in ERDAS Imagine software.

Grain was harvested at physiological maturity for all site-years. Because site-years I and II were located on a small experimental field and we did not have access to a yield monitor equipped small plot combine, grain was therefore hand harvested. Grain was hand harvested from geo-referenced locations. Each harvest sample consisted of two 1-m long sections of a maize row (1.52 m²) and grain samples was replicated four times, (i.e., four grain yield observations were obtained within each management zone). Samples were air-dried to a constant weight and analyzed for grain yield at 15.5 g kg^–1 moisture.

Site-years III, IV, V, and VI were harvested using a combine equipped with a yield monitor and a differentially corrected global positioning system (DGPS) unit. Yield monitor data were cleaned by removing erroneous data points, correcting for grain moisture, adjusting for the speed of the combine and the width of the combine header, as well as the grain flow rate (Hornung et al., 2006).

Using ArcView 3.2, remote sensing images were separated into individual spectral bands (green, red, and near-infrared) and overlain with the geo-referenced yield monitor data. Data from the individual bands were extracted for each yield monitor pixel. After data extraction, a random sample of 2,000 points was collected for analysis (Inman et al., 2005b). Each data point contained grain yield and corresponding green (G), red (R), and NIR apparent reflectance, as well as the precise geographic coordinates. The NDVI was calculated using Eq. [1]. The relative yield for a given observation was calculated by dividing the observed values by the maximum value for the site-year (Brouder et al., 2000).

[1]

Data Analysis
Objective 1
Analysis of variance (ANOVA) was used to determine if NDVI and grain yield were significantly different among management zones. When ANOVA was significant at P 0.05, Fishers LSD mean separation was performed. The relationship among NDVI determined early in the growing season, soil color-based management zones, and relative maize yield was quantitatively compared using K-means clustering, percent areal agreement, and Kappa statistics. K-means clustering was used to group the data into three NDVI clusters and three relative yield clusters using Splus 6.2 (Insightful Corp. Seattle, WA). Three clusters were chosen to correspond to the number of management zones. Clusters were then quantitatively compared to management zones using percent areal agreement and Kappa statistics (Campbell, 2002). Percent areal agreement and Kappa are calculated using an x-by-x "error matrix" in which x is the number of categories (Campbell, 2002). In our study, we wanted to determine how management zones (high, medium, and low productivity potential) compare to high, medium, and low NDVI clusters as well as high, medium, and low relative yield clusters. Percent areal agreement is an overall measure of how well the categories of an error matrix correspond to each other. Campbell (2002) stated that even a chance assignment of pixels to categories can produce very high percent areal agreement. The Kappa statistic is a more robust measure of the success of a classification because it adjusts for chance agreement (i.e., a random classification). The Kappa statistic is interpreted as how well a classification performs as compared to a chance classification of the same data set. More detailed information regarding the calculation, use, and interpretation of the Kappa statistic is provided in Campbell (2002) and Landis and Koch (1977).

Objective 2
Least squares regression analysis was used to regress relative grain yield on NDVI; indicator variables were included to determine if SCMZs can be used in conjunction with NDVI to provide early-season yield estimates (SAS Institute, 2001). Indicator variables are one means of quantifying categorical data within regression analysis (Neter et al., 1996). For this study, binary variables were assigned to each SCMZ and included as independent variables in the regression analysis (e.g., X₁ = 1 if high zone, X₁ = 0 otherwise; X₂ = 1 if medium zone, X₂ = 0 otherwise, etc.). Neter et al. (1996) provide a complete explanation of the use and interpretation of indicator variables in regression modeling. In addition, residual regression analysis was used to determine if the addition of remote sensing data (i.e., NDVI) could account for a significant amount of the variability in grain yield. This was accomplished by using least-squares regression analysis to regress grain yield on the SCMZs; residuals from this regression were then regressed on NDVI for each site-year. Regressing grain yield on SCMZ and analyzing the residuals from this regression removes the effect of soil color, as determined by the SCMZs. From this analysis we can determine to what extent soil color effects the relationship among NDVI and grain yield and if adding NDVI to the SCMZs accounts for more variability in grain yield than SCMZs alone.

	RESULTS AND DISCUSSION

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

 
Analysis of variance and mean separation for NDVI among management zones are presented in Fig. 1 . Among management zones, NDVI was found to be statistically different. In four out of six site-years, the NDVI increased with the productivity potential of the management zones. Mzuku et al. (2005) showed that soil color based management zones differ on the basis of soil texture, organic matter, and soil moisture. Comparing the NDVI results to the ANOVA and mean separation performed on grain yield among management zones (Fig. 2 ), the trends are similar. Grain yield was significantly different among management zones in four out of six site-years. When comparing plots of NDVI among management zones to plots of grain yield among management zones, statistical differences are nearly identical. These results suggest that the observed differences in NDVI among management zones are related to soil color differences, and that both NDVI and management zones provide similar information during the early growing season.

View larger version (31K): [in this window] [in a new window]   Fig. 1. Mean normalized difference vegetation index (NDVI) at crop growth stage V8, among management zones for all site years. Within a site-year bars with different letters are significantly different at P 0.05 level of significance. Site years are designated by roman numerals in the upper right-hand portion of each chart.

View larger version (35K): [in this window] [in a new window]   Fig. 2. Mean grain yield among management zones for all site-years. Within a site-year bars with different letters are significantly different at P 0.05 level of significance. Site years are designated by roman numerals in the upper right-hand portion of each chart.

View larger version (27K): [in this window] [in a new window]   Fig. 3. Scatter plots of normalized difference vegetation index (NDVI) at crop growth stage V8, vs. relative yield for site years (SY) I through VI.

View this table: [in this window] [in a new window]   Table 4. Best fit regression function and coefficient of determination (R2) associated with regressing normalized difference vegetation index (NDVI) at crop growth stage V8, and soil color-based management zones (SCMZ) on relative grain yield.

We removed the effects of SCMZs by regressing SCMZs on relative yield, the residuals were calculated and further analyzed. The NDVI was regressed on the residuals calculated from the previous regression analysis, results from these analysis are presented in Table 5 . Removing the affects of SCMZs, we were able to determine the extent to which the NDVI is capturing the variability of the growing crop. Among site-years, results were variable. Management zones alone accounted for 3 and 29% of the variability in relative grain yield (P < 0.01). However, NDVI accounted for 10 and 47% of the variability in the residuals (P < 0.01). In four of the six cases, NDVI accounted for more variability than management zones alone. If early-season NDVI were merely a surrogate measure of soil color, the NDVI would not have accounted for a significant amount of the variability in the residuals. These results demonstrate that early-season NDVI can provide valuable information and has the potential to be a useful tool in precision nutrient management for maize crops. As demonstrated in previous studies, soil color does affect early-season NDVI and yield relationship. However, this study shows that SCMZs are not necessarily the predominate explanatory factor in the observed relationship between NDVI and grain yield. More work should be done to further understand the relationship among SCMZs and NDVI in maize crops.

	CONCLUSIONS

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

	ACKNOWLEDGMENTS

 
Authors would like to acknowledge and thank the USDA-IFAFS, and CSU Agricultural Experiment Station for funding part of this research project.

	NOTES

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

 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.

	REFERENCES

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

 

• Aparicio, N., D. Villegas, J.L. Araus, J. Casadesús, and C. Royo. 2002. Relationship among growth traits and spectral vegetation indices in Durum Wheat. Crop Sci. 42:1547–1555.[Abstract/Free Full Text]
• Báez-González, A.D., P. Chen, M. Tiscareño-López, and R. Srinivasan. 2002. Using satellite and field data with crop growth modeling to monitor and estimate maize yield in Mexico. Crop Sci. 42:1943–1949.[Abstract/Free Full Text]
• Baret, F., G. Guyot, and D.J. Major. 1988. TSAVI: A vegetation index which minimizes soil brightness effects on LAI and APAR estimation. p. 1355–1358. In 12^th Canadian Symp. On Remote Sensing and IGARRS '90, Vancouver, BC, Canada. 10–14 July 1989.Geoscience and Remote Sensing Soc. of Inst. of Electrical and Electronics Engineers, Piscataway, NJ.
• Brouder, S.M., D.B. Mengel, and B.S. Hofmann. 2000. Diagnostic efficiency of the blacklayer stalk nitrate and grain nitrogen tests for maize. Agron. J. 92:1236–1247.[Abstract/Free Full Text]
• Campbell, J.B. 2002. Introduction to remote sensing. 3rd ed. The Guilford Press, New York.
• Chavez, P.S. 1975. Atmospheric, solar, and M.T.F. corrections for ERTS digital imagery. p. 724–932. In Proc., Am. Soc. of Photogrammetry Fall Conf., Phoenix, AZ. Soc. of Photogrammetry and Remote Sensing, Bethesda, MD.
• Colwell, R.W. 1956. Determining the prevalence of certain cereal crop diseases by means of aerial photography. Hilgardia 26:223–286.
• Doerge, T. 1999. Defining management zones for precision farming. Crop Insights 8:1–5.
• Fleming, K.L., D.G. Westfall, D.W. Wiens, L.E. Rothe, J.E. Cipra, and D.F. Heermann. 1999. Evaluating farmer developed management zone maps for precision farming. p. 335–343. In P.C. Robert et al. (ed.) Precision agriculture. Proc. of the 4th Int. Conf., St. Paul, MN. 19–22 July, 1999. ASA, CSSA, and SSSA, Madison, WI.
• Fleming, K.L., D.F. Heermann, and D.G. Westfall. 2004. Evaluating soil color with farmer input and apparent soil electrical conductivity for management zone delineation Agron. J. 96:1581–1587.
• Hoffer, R.M. 1967. Interpretation of remote multispectral imagery of agricultural crops. Res. Bull. 831. Purdue Univ. Agric. Exp. Stn., West Lafayette, IN.
• Hornung, A., R. Khosla, R. Reich, D. Inman, and D. Westfall. 2006. Comparison of site-specific management zones. Soil color based and yield based. Agron. J. 98:405–417.
• Huete, A.R. 1989. A soil-adjusted vegetation index (SAVI). Remote Sens. Environ. 25:295–309.[CrossRef]
• Inman, D., R. Khosla, M. Lefsky, and D.G. Westfall. 2005b. Early season grain yield prediction using remote sensing and soil color-based management zones. p. 835–841. In J.V. Stafford (ed.) Precision Agriculture '05. Wageningen Academic Publ., Wageningen, the Netherlands.
• Inman, D., R. Khosla, D.G. Westfall, and R. Reich. 2005a. Nitrogen uptake across site-specific management zones in irrigated corn production systems. Agron. J. 97:169–176.[Abstract/Free Full Text]
• Khosla, R., K. Fleming, J.A. Delgado, T. Shaver, and D.G. Westfall. 2002. Use of site specific management zones to improve nitrogen management for precision agriculture. J. Soil Water Conserv. 57:515–518.
• Landis, R.J., and G.G. Koch. 1977. The measurement of observer agreement for categorical data. Biometrics 33:159–174.[CrossRef][Web of Science][Medline]
• Ma, B.L., L.M. Dwyer, C. Costa, E.R. Cober, and M.J. Morrison. 2001. Early prediction of soybean yield from canopy reflectance measurements. Agron. J. 93:1227–1234.[Abstract/Free Full Text]
• Mzuku, M., R. Khosla, R. Reich, D. Inman, F. Smith, and L. MacDonald. 2005. Spatial variability of measured soil properties among soil color-based management zones. Soil Sci. Soc. Am. J. 69:1572–1579.[Abstract/Free Full Text]
• Neter, J., M.H. Kunter, C.J. Nachtscheim, and W. Wasserman. 1996. Applied linear regression models. 3rd ed. McGraw-Hill/Irwin Publ., New York.
• Peñuelas, J., J.A. Gamon, A.L. Fredeen, J. Merino, and C.B. Field. 1994. Reflectance indices associated with physiological changes in nitrogen- and water-limited sunflower leaves. Remote Sens. Environ. 48:135–146.[CrossRef]
• Raun, W.R., J.B. Solie, G.V. Johnson, M.L. Stone, E.V. Lukina, W.E. Thomason, and J.S. Schepers. 2001. In-season prediction of potential grain yield in winter wheat using canopy reflectance. Agron. J. 93:131–138.[Abstract/Free Full Text]
• Ritchie, S.W., J.J. Hanway, and O.G. Benson. 1993. How a corn plant develops. Publ. SR48. Iowa State Univ. Ext., Ames.
• Rondeaux, G., M. Stewart, and F. Baret. 1996. Optimization of soil adjusted vegetation indices. Remote Sens. Environ. 55:95–107.[CrossRef]
• SAS Institute. 2001. Statistical analysis software. Version 8. SAS Inst., Cary, NC.
• Shanahan, J., J.S. Schepers, D.D. Francis, G.E. Varvel, W.W. Wilhelm, J.M. Tringe, M.R. Schlemmer, and D.J. Major. 2001. Use of remote sensing imagery to estimate corn grain yield. Agron. J. 98:583–589.
• Soil Survey Staff. 1968. Soil survey of Morgan County, Colorado. USDA Coop. Soil Survey. U.S. Gov. Print. Office, Washington, DC.
• Soil Survey Staff. 1980. Soil survey of Larimer County, Colorado. USDA Coop. Soil Survey. U.S. Gov. Print. Office, Washington, DC.
• Soil Survey Staff. 1981. Soil survey of Yuma County, Colorado. USDA Coop. Soil Survey. U.S. Gov. Print. Office, Washington, DC.
• Teal, R.K., B. Tubana, K. Girma, K.W. Freeman, D.B. Arnall, O. Walsh, and W.R. Raun. 2006. In-season prediction of corn grain yield potential using normalized difference vegetation index. Agron. J. 98:1488–1494.[Abstract/Free Full Text]
• Thenkabail, P.S., R.B. Smith, and E. DePauw. 2000. Hyperspectral vegetation indices and their relationships with agricultural crop characteristics. Remote Sens. Environ. 71:158–182.[CrossRef]

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 Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Inman, D. Articles by Westfall, D. G. Search for Related Content PubMed Articles by Inman, D. Articles by Westfall, D. G. Agricola Articles by Inman, D. Articles by Westfall, D. G. Related Collections Maize Maize Maize Remote Sensing Soil Fertility and Productivity Maize Management