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 (5) Citing Articles via Google Scholar Google Scholar Articles by Clay, D. E. Articles by Dalsted, K. Search for Related Content PubMed Articles by Clay, D. E. Articles by Dalsted, K. Agricola Articles by Clay, D. E. Articles by Dalsted, K. Related Collections Water Stress Nitrogen Remote Sensing Maize Management

Remote Sensing

Characterizing Water and Nitrogen Stress in Corn Using Remote Sensing

^a Plant Science Dep., South Dakota State Univ., Brookings, SD 57007
^b Engineering Resource Center, South Dakota State Univ., Brookings, SD 57007

^* Corresponding author (david.clay{at}sdstate.edu)

Received for publication July 6, 2005.

	ABSTRACT

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

 
Interactions between water and N may impact remote-sensing-based N recommendations. The objectives of this study were to determine the influence of water and N stress on reflectance from a corn (Zea mays L.) crop, and to evaluate the impacts of implementing a remote-sensing-based model on N recommendations. A replicated N and water treatment factorial experiment was conducted in 2002, 2003, and 2004. Yield losses due to water (YLWS) and N (YLNS) stress were determined using the ¹³C discrimination () approach. Reflectance data (400–1800 nm) collected at three growth stages (V8–V9, V11–VT, and R1–R2) were used to calculate six different remote sensing indices (normalized difference vegetation index [NDVI], green normalized vegetation index, normalized difference water index [NDWI], N reflectance index, and chorophyll green and red edge indices). At the V8–V9 growth stage, increasing the N rate from 0 to 112 kg N ha^–1 decreased reflectance in the blue (485 nm), green (586 nm), and red (661 nm) bands. Nitrogen had an opposite effect in the near-infrared (NIR, 840 nm) band. At the V11–VT growth stage, reflectance in the blue, green, and red bands were lower in fertilized than unfertilized treatments. At the R1–R2 growth stage, YLWS was highly correlated (r = 0.58, P = 0.01) with red reflectance and NDVI (r = –0.61, P = 0.01), while YLNS was correlated with all of the indices except NDVI. A remote sensing model based on YLNS was more accurate at predicting N requirements than models based on yield or yield plus YLWS. These results were attributed to N and water having an additive effect on yield, and similar optimum N rates (100–120 kg N ha^–1) for both moisture regimes.

Abbreviations: , ¹³C discrimination • C_green, chlorophyll green index • C_{red edge}, chlorophyll red edge index • GNDVI, green normalized vegetation index • MIR, mid-infrared • NDVI, normalized difference vegetation index • NDWI, normalized difference water index • NIR, near-infrared • NRI, nitrogen reflectance index • YLNS, yield loss due to nitrogen stress • YLWS, yield loss due to water stress

	INTRODUCTION

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

 
NITROGEN and water stress interact to influence yields in fields with complex landscapes. Remote sensing can be used as a tool to assess these stresses (Barnes et al., 2000); however, before implementation of a remote-sensing-based N management program, it is necessary to determine if interactions between water and N stress interact to influence reflectance, yields, and the resulting N fertilizer recommendations.

Nitrogen stress reduces the production of the chlorophyll that is involved in the production of the reduced compounds nicotinamide adenine dinucleotide phosphate and adenosine triphosphate. These compounds are used by ribulose-1-5-bisphosphate carboxylase to drive CO₂ fixation. Reduced chlorophyll content can result in increased reflectance of photosynthetically active light (Gitelson et al., 2005), resulting in N-stressed plants appearing yellow. A number of different indices have been proposed to assess N stress. The NDVI was developed to assess plant greenness (Rouse et al., 1974; Jackson et al., 2004). The index relates the reflectance in the red and NIR spectral bands. The red band is near chlorophyll A absorption maximum, while the NIR band is sensitive to canopy cover (Barnes et al., 2000; Shanahan et al., 2001). A problem with this index is that it saturates at intermediate leaf area index values (Barnes et al., 2000; Jackson et al., 2004).

A second index is the GNDVI (green normalized difference vegetation index; Gitelson et al., 1996). Shanahan et al. (2001) reported that GNDVI data acquired during midgrain filling can be used to produce relative yield maps. A third index is the NRI (N reflectance index; Bausch and Duke, 1996). This index is based on the ratios between reflectance in the NIR and green bands in N-deficient and well-fertilized plants. This index is based on the observation that vigorous plants have higher reflectance in the NIR band than stressed plants and that N-deficient plants have higher reflectance in the green band than non-N-stressed plants. Gitelson and Merzlyak (1994) reported that the ratio between the NIR and green bands were sensitive to chlorophyll content. Three other indices proposed by Gitelson et al. (2005) were C_green [(R_{800 nm}/R_{550 nm}) – 1], C_{red edge} ([R_{800 nm}/R_{red edge(700 nm)}] – 1), and C_NIR ([R_840–870/R_{red edge(720–740 nm)}] – 1), where R is reflectance.

Reflectance may also be impacted by water stress. In the plant, water stress can have a nonlinear impact on growth and development. Under moderate water stress, partial stomatal closure, which reduces H₂O transpiration and the CO₂ available for C fixation, may occur for only several hours a day. Plants experiencing moderate stress may not wilt or have photochemical activity impaired (Souza et al., 2004). As the water deficit increases, the plant increasingly appear wilted and the actual photochemical activity of chlorophyll can be reduced (Souza et al., 2004; Pettigrew, 2004). Reflectance may be impacted under these conditions. These findings suggest that a nonlinear response to water stress can delay, mask, or confound reflectance signals. The NDWI was proposed by Gao (1996) as a tool to assess water stress. This index relates reflectance in the NIR and MIR bands (Landsat bands 4 and 5) and has been used to estimate vegetative water content (Jackson et al., 2004). This approach is based on relationships between NIR reflectance and leaf area index, and MIR reflectance and vegetative water content.

Given that both water and N impact growth, development, and spectral reflectance, it is likely that the "best" index for assessing N stress will be different than the "best" index for assessing water stress. Barnes et al. (2000) proposed an approach for separating water and N stress. Their approach relied on using reflectance in the 790- and 720-nm wave bands to calculate the canopy chlorophyll content and differences in air temperature in watered and unwatered plants to calculate the crop water stress index. Potential interactions between N and water stress on reflectance were not investigated. The objectives of this study were to determine the influence of water and N stress on reflectance from a corn crop, and to evaluate the impacts of implementing a remote-sensing-based model on N recommendations.

	MATERIALS AND METHODS

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

 
Experimental Design
Soils at the research site were formed in wind-blown loess deposited over glacial outwash approximately 12 000 yr ago. The soil was classified as a Brandt silty clay loam (fine-silty, mixed, frigid Calcic Hapludoll). Organic matter content at the site was 45 g kg^–1, total N was 2.9 g kg^–1, and the soil had a saturated hydraulic conductivity of 0.72 m d^–1 (Clay, 1997). The site was chisel plowed in the fall. If required, monoammonium phosphate fertilizer was broadcast applied in the spring (Gerwing and Gelderman, 2005). Yield and crop reflectance differences were attributed to differential water and N stress because sufficient P fertilizer was applied to meet expected plant demand, soil K levels exceeded the recommended level, weeds were controlled, and insect or disease problems were not detected.

Longitude and latitude coordinates at the research site are 96°40' W, 44°18' N. Plot dimensions were 15 by 14 m. Treatments arranged in a randomized split block design were two water regimes (natural and natural plus supplemental) and four N rates (0, 56, 112, and 168 kg N ha^–1). Following the spring application of the N fertilizer (urea), the soil was disked. Each treatment was replicated four times. The DKC 44-46 RR Bt (Monsanto Co., St Louis, MO) corn hybrid was planted in 2002 and 2003 and the DKC 47-10 RR Bt corn hybrid was planted in 2004 at population levels of 80 000 plants ha^–1. The DKC 44-46 and DKC 47-10 hybrids required 1311 and 1344 growing degree days (base 10°C) to black layer.

In 2002, the natural water regime received 59 cm of precipitation, while the natural plus supplemental irrigation received an additional 9.9 cm of irrigation on four dates (30 May, 19 June, 15 July, and 19 August). In 2003, the natural water regime received 38 cm of precipitation, and, in the natural plus supplemental water treatment, 3.8, 1.6, 1.9, 1.9, 2.5, and 3.2 cm of irrigation was applied on 15 May, 29 May, 10 June, 19 June, 1 July, and 11 July, respectively. The total natural precipitation plus irrigation in 2003 was 52.9 cm. In 2004, corn growing under the natural water regime received 49 cm of precipitation, and under the natural plus supplemental water regime, 5 cm of irrigation was applied twice (26 July and 31 August). Total precipitation plus irrigation in 2004 was 59 cm. Growing degree days in 2002, 2003, and 2004 were 1390, 1392, and 1171, respectively. Daily temperatures and precipitation was measured with a weather station at the site.

At physiological maturity, grain from a 9.3-m² area was harvested. Grain was dried, weighed, ground, and analyzed for total N and ¹³C/¹²C ratio on a 20-20 Europa ratio mass spectrometer (PDZ Europa, Cheshire, UK) (O'Leary, 1993; Farquhar and Lloyd, 1993; Clay et al., 2005). The ¹³C/¹²C ratios were used to calculate (¹³C discrimination; Clay et al., 2005).

Quantifying Nitrogen and Water Stress
Carbon-13 discrimination provides an indirect measure of a plant's physiological response to water and N stress. The value provides an index of the relative amount of ¹³C discrimination that might occur from a variety of events. A value of 0 indicates that ¹³C discrimination did not occur, whereas positive values indicate that ¹³C discrimination occurred (O'Leary, 1993; Farquhar and Lloyd, 1993; Clay et al., 2001a, 2001b; Smeltekop et al., 2002; Clay et al., 2005).

The -based approach to quantifying yield losses due to N and water stress requires the development of equations that define the relationships between yield and , YLNS and , and YLWS and . These equations are then solved to determine YLNS and YLWS for each plot. A graphical representation of these values and complete discussion of the approach is available in Clay et al. (2005). This approach for quantifying N and water stress have been tested in wheat (Triticum aestivum L.; Clay et al., 2001a), corn (Clay et al., 2001b, 2005), and soybean [Glycine max (L.) Merr.; Clay et al., 2003]. The approach requires the development of two independent equations, one based on yield and the other based on :

[1]

[2]

where maximum yield is defined as 15 000 kg grain ha^–1, d is the difference between the value of a well-fertilized plant grown under conditions where water did not limit yields and the measured value for a given plant or plot where both water and N did impact yield, /yield WS is the partial derivative of the line relating and yield under conditions when water limits yield and N stress is constant, /yield NS is the partial derivation of the line relating and yield when N limits yield and water stress is constant Clay et al. (2001b, 2005). The /yield WS value is the slope of the upper boundary line relating yield and (Webb, 1972). The upper boundary approach has been used by numerous scientists to evaluate biological and ecological systems (Webb, 1972; Schmidt et al., 2000; Kitchen et al., 2003). The upper boundary line for grain yield in 2002, 2003, and 2004 were: yield (kg ha^–1) = 45 679 – 11 262, yield (kg ha^–1) = 31 237 – 6383, and yield (kg ha^–1) = 62 690 – 16 000, respectively. Based on these equations, the yield WS/ values for 2002, 2003, and 2004 were 8.9 x 10^–5, 1.6 x 10^–4, and 6.3 x 10^–5 ha kg^–1 grain, respectively. The /yield NS value was 3.03 x 10^–5 ha kg^–1grain for all years. This value was calculated by comparing yield and values for the different N rates within the natural soil moisture regime for data collected in 2002. By solving Eq. [1] and [2] simultaneously, YLWS and YLNS were determined (Clay et al. (2001a), and 2005).

To validate the ¹³C-based approach described above, ¹³C-based YLNS and YLWS values were compared with values calculated using the difference approach (Clay et al., 2005). For the water stress comparison, only data from the well-fertilized treatments in the natural and natural plus supplemental treatments were compared (these calculations assumed N stress did not limit yields in these treatments). Yield losses due to N stress were calculated by comparing unfertilized with well-fertilized treatments within a soil moisture regime during the 3 yr. Difference calculations assumed that, within a moisture regime, N did not influence water stress. The ¹³C-based YLNS and YLWS values were highly correlated with difference-calculated values (Fig. 1 ). For a corn study conducted in Nebraska, Clay et al. (2005) had similar results.

View larger version (22K): [in this window] [in a new window]   Fig. 1. Validation of the 13C discrimination approach for estimating yield losses due to water (YLWS) and N stress (YLNS).

Broad-band reflectance indices were calculated using the equations: NDVI = (NIR – red)/(NIR + red); GNDVI_s = [(NIR – green)/(NIR + green)]/[(NIR_r – green_r)/(NIR_r + green_r)]; NDWI = (NIR – MIR)/(NIR + MIR); and NRI = (NIR/green)/(NIR_r/green_r), where GNDVI_r, NIR_r, and green_r were values from well-fertilized and watered control plots (Bausch and Duke, 1996; Shanahan et al., 2001; Jackson et al., 2004), and GNDVI_s was a standardized GNDVI value (divided by reflectance values in the well-fertilized control). At the V8–V9 growth stage, (i) the GNDVI_r for well-fertilized and -watered plants was 0.80, 0.75, and 0.80 in 2002, 2003, and 2004, respectively; and (ii) the NDVI_r (well-fertilized reference) values were 0.90, 0.85, and 0.88 in 2002, 2003, and 2004, respectively. At the V11–VT growth stage, (i) the GNDVI_r values for well-fertilized corn plants were 0.79, 0.82, and 0.82 in 2002, 2003, and 2004, respectively; and (ii) the NDVI_r values for well-fertilized corn plants were identical (0.90) in 2002, 2003, and 2004. At the R1–R2 growth stage, the GNDVI_r values for well-fertilized plants were similar (0.82) in 2003 and 2004.

Narrow-band reflectance values were used to calculate C_green (810/568 nm) and C_{red edge} (810/710 nm) (Gitelson et al., 2005). The red edge has been defined as the point of maximum reflectance slope in the area between 680 and 750 nm.

Statistical Analysis
The experiment was analyzed using ANOVA, correlation, and backward stepwise regression approaches. In the ANOVA analysis, year x treatment interactions were not detected, and therefore the reported values were averaged across years. Based on the F values from the ANOVA, P values were calculated. For significant P values (<0.05), LSD values (0.05 level) were calculated to separate means for factors with more than two treatments. Correlation analysis was used to determine the strength of the relationships between reflectance and crop parameters (yield, YLNS, and YLWS). Data from the 3-yr study were not aggregated for correlation analysis. For the backward stepwise regression, a significance level of 0.05 was used for including parameters in the model.

Fertilizer Model Testing
Remote sensing YLNS, YLWS, and yield models for the three growth stages were developed using a backward stepwise analysis. Nitrogen recommendation models were developed for data collected between V8 and V9 using three conceptually different approaches. The first approach (N deficiencies only) was based on the YLNS remote-sensing predictive model. The recommendation model was: N recommendation (kg N ha^–1) = (remote-sensing-predicted kg YLNS ha^–1)(0.021 kg N kg^–1 YLNS). This equation assumes that 0.021 kg N kg^–1 of grain were required (Gerwing and Gelderman, 2005).

The second approach was based on remote-sensing-estimated yield values (yield approach). The recommendation model for this approach was: N requirement = (0.021 kg N kg^–1 grain)(optimum yield [10 500 kg grain ha^–1]) – remote-sensing-predicted yield).

The third approach (yield and YLWS approach) considered yields and water regime information. The recommendation model was: N recommendation (kg N ha^–1) = (0.021 kg ha^–1)(optimum yield [10 500 kg grain ha^–1]) – YLWS – remote-sensing-predicted yield). The YLWS was defined as 850 and 0 kg grain ha^–1 for the natural precipitation and natural plus supplemental irrigation treatments, respectively. Predicted and measured values were compared and root mean square errors (RMSE = [(measured N requirement – calculated N recommendation)²/n]^0.5, where n is the number of comparisons used in the analysis) were calculated. Potential yield losses were determined for plots where the measured N requirement was greater than the calculated N recommendation using the equation (measured N requirement – calculated N recommendation)/0.021).

	RESULTS AND DISCUSSION

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

 
Crop Growth
An interaction between N and water was not observed (Table 1). The lack of an interaction was attributed to supplemental irrigation increasing N mineralization (Linn and Doran, 1984), which reduced the need for N fertilizer. Evidence supporting this hypothesis were higher mineralization rates in irrigated than nonirrigated plots and higher plant ¹⁵N values in the irrigated than the nonirrigated plots (Kim et al., 2004). Higher ¹⁵N values were attributed to fertilizer having a ¹⁵N value <0 while N derived from soil has ¹⁵N values ranging from 3 to 5. In this soil, the effects of management have previously been investigated. Clay et al. (1995) and Clay (1997) reported that (i) tillage influenced N mineralization in row relative to interrow areas; (ii) N mineralization peaks occurred in the early spring, while C peaks occurred in midsummer; and (iii) annual N mineralization can exceed 95 kg N ha^–1.

Nitrogen and Water Impacts on Crop Reflectance: ANOVA Analysis
Crop reflectance was influenced by the imposed treatment and wave band. A sample reflectance characteristics curve for the narrow-band data is provided in Fig. 2a . This curve shows the relative location of the broad bands. Broad-band data was used to calculate NDVI, GNDVI, and NRI values, while narrow-band data was used to calculate C_green and C_{red edge} index values.

View larger version (23K): [in this window] [in a new window]   Fig. 2. The influence of N and water stress on narrow-band reflectance measured on 28 July 2003. Values labeled as 0W and +W represent natural rainfall and supplemental irrigation, respectively; 0N and +N represent unfertilized and fertilized treatments, respectively. In (b), 0W N stress is the difference between the +N and 0N treatments in the 0W treatment. The +W N stress is the difference between the +N and 0N treatment in the +W treatment.

At R1–R2 in 2003 and 2004, the addition of N fertilizer increased all index values, reduced reflectance in the broad green band, and increased broad NIR band reflectance (Table 3). The impact of N treatments on relative narrow-band reflectance can be seen in Fig. 2b, where reflectance in the visual bands (blue, green, and red) was higher for the unfertilized than the fertilized treatment. In the NIR bands, the opposite results where observed. The narrow-band reflectance spectra had interesting results in the red edge region. At 710 nm, difference spectra (unfertilized – fertilized) in the natural (0W) and irrigated (+W) treatments had similar values and were both positive (reflectance in unfertilized higher than fertilized). At 760 nm, however, different results were observed, with the irrigated treatment (+W) becoming negative while the natural rainfall treatment (0W) remained positive. At 840 nm, the difference values in the 0W and +W treatments were both negative. These results suggest that changes in reflectance along the red edge provide valuable clues about interactions between water and N in the plant. Additional research using hyperspectral reflectance information is needed to explore these relationships.

Across the three growth stages, all indices were influenced by N rate. At the V11–VT growth stage, the index values were not influenced by supplemental irrigation. Different results were observed at the R1–R2 growth stage, where NDVI, GNDVI_s, and NRI were influenced by water stress. The inability to detect water stress at the V11–VT growth stage was attributed to a nonlinear physiological response to water stress that masked the expression of visual symptoms (Souza et al., 2004; Pettigrew, 2004). Barnes et al. (2000) had similar results and reported that water stress symptoms may not be detectable until 50% of the available water has been used.

Nitrogen and Water Impacts on Crop Reflectance: Correlation Analysis
The strength of the relationships among reflectance, N, and water stress were growth stage and band dependent (Table 4). The reflectance indices NDVI, GNDVI_s, NRI, C_green, and C_{red edge} were negatively correlated to YLNS at all sampling dates. These results were attributed to N stress reducing chlorophyll concentration, which increased reflectance in photosynthetically active wavelengths. The correlation coefficients for the nonstandardized indices (C_green and C_{red edge}) were generally lower than the correlation coefficients for the standardized indices (GNDVI_s and NRI). Other approaches for assessing N stress, such as the chlorophyll meter, have integrated this calibration procedure in the protocol (Francis and Piekielek, 1998).

These results show that the impacts of water and N stress on reflectance occurred at different growth stages. At the V11–VT growth stage, reflectance was primarily impacted by N stress, while at the R1–R2 growth stage, reflectance was influenced by both water and N. Correlations between yield and reflectance in the green and red bands were attributed to the effect of individual stresses on reflectance and yield. Green reflectance was sensitive to N stress while red reflectance was sensitive to water stress. The GNDVI_s values had stronger relationships with yield and YLNS than NRI, C_green, or C_{red edge}.

Predicting Yield and Yield Losses Due to Nitrogen and Water Stress Using Remote Sensing
Plant growth stage and crop parameter (YLNS, YLWS, and yield) influenced the amount of variability explained by the remote-sensing-based model. At the V8–V9 growth stage, the amount of variability explained during the 3 yr ranged from 63 to 68% (Table 6). For the data collected at the V8–V9 growth stage, N fertilizer models were developed. For the N deficiency model, the RMSE, YLNS, and average fertilizer needed to eliminate N deficiencies were 16.21 kg N ha^–1, 29000 kg grain, and 26.4 kg N ha^–1, respectively. For these calculations, YLNS were summed across all plots. Higher RMSE (21.61 kg N ha^–1), yield losses (32500 kg grain), and N recommendations (30.2 kg N ha^–1) were observed for N recommendations based on the yield model. For the third model (yield and YLWS), the RMSE, yield losses, and average additional N fertilizer needed to reach optimum yield were 21.71 kg ha^–1, 46400 kg, and 23.6 kg N ha^–1. Based on YLNS values, the actual amount of additional N fertilizer required to meet the plants' requirements was 26.5 kg N ha^–1. These results suggest that: (i) the relatively high RMSE for the yield plus water stress model resulted from applying less fertilizer than what was required; (ii) the model that "best" predicted the plants' requirement was the model based on N stress information only; and (iii) in landscapes with a range of yield potentials, a tool that assesses only N deficiencies may provide reasonably accurate N recommendations. In the above exercise, basing the recommendation on a combination of yield or soil moisture regime did not improve N recommendations relative to a model that only considered N deficiencies. These results were attributed to the lack of an interaction between N and moisture regime and adding supplemental irrigation causing an increase in N mineralization. Different results could be expected in a soil containing less labile organic N.

	CONCLUSIONS

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

	NOTES

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

 
South Dakota Agric. Exp. Stn. Journal Series no. 3503. Research supported in part by the United Soybean Board, South Dakota Soybean Research and Promotion Council, South Dakota Corn Utilization Board, NASA, USDA-CSREES, and South Dakota Agric. Exp. Stn.

	REFERENCES

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

 

• Barnes, E.M., T.R. Clarke, S.E. Richard, P.D. Colaizzi, J. Hoberland, M. Kostozewski, P. Waller, C. Choi, R. Riley, T. Thompson, R.J. Lascano, H. Li, and M.S. Moran. 2000. Coincident detection of crop water stress, nitrogen status, and canopy density using ground-based multispectral data. In P.C. Robert et al. (ed.) Proc. Int. Conf. on Prec. Agric., 5th, Bloomington, MN. 16–19 July 2000 [CD-ROM]. ASA, CSSA, and SSSA, Madison, WI.
• Bausch, W.C., and H.R. Duke. 1996. Remote sensing of plant nitrogen status in corn. Trans. ASAE 39:1869–1875.
• Chang, J., S.A. Clay, D.E. Clay, D. Aaron, and D. Helder. 2005. Comparing spectral reflectance data collected from two different spectroradiometers for ground reflectance measurements. Commun. Soil Sci. Plant Anal. 37:1799–1807.
• Clay, D.E. 1997. Comparison of the difference and delta¹⁵N approaches for evaluating liquid urea ammonium nitrate utilization. Commun. Soil Sci. Plant Anal. 28:1151–1161.
• Clay, D.E., S.A. Clay, J. Jackson, K. Dalsted, C. Reese, Z. Liu, D.D. Malo, and C.G. Carlson. 2003. Carbon-13 discrimination can be used to evaluate soybean yield variability. Agron. J. 95:430–435.[Abstract/Free Full Text]
• Clay, D.E., S.A. Clay, D.J. Lyon, and J.M. Blumenthal. 2005. ¹³C discrimination in corn grain can be used to separate and quantify yield losses due to water and nitrogen stresses. Weed Sci. 53:23–29.
• Clay, D.E., S.A. Clay, Z. Lui, and C. Reese. 2001b. Spatial variability of C-13 isotopic discrimination in corn (Zea mays). Commun. Soil Sci. Plant Anal. 32:1813–1828.
• Clay, D.E., R. Engel, D. Long, and Z. Lui. 2001a. Nitrogen and water stress interact to influence carbon-13 discrimination in wheat. Soil Sci. Soc. Am. J. 65:1823–1828.[Abstract/Free Full Text]
• Clay, D.E., T.E. Schumacher, and K.A. Brix-Davis. 1995. Carbon and nitrogen mineralization in row and interrow areas of chisel and ridge tillage systems. Soil Tillage Res. 35:167–174.
• Farquhar, G.D., and L. Lloyd. 1993. Carbon and oxygen isotope effects in the exchange of carbon dioxide between terrestrial plants and the atmosphere. p. 47–70. In J.R. Ehleringer et al. (ed.) Stable isotopes and plant carbon–water relations. Academic Press, New York.
• Francis, D.D., and W.P. Piekielek. 1998. Assessing crop nitrogen needs with chlorophyll meters. SSMG-12. In Clay et al. (ed.) Site-specific management guidelines [Online]. Available at http://www.ppi-far.org/ssmg (accessed April 2005; verified 24 Jan. 2006). Potash and Phosphate Inst., Norcross, GA.
• Gao, B. 1996. NDWI—a normalized difference water index for remote sensing of vegetation liquid water from space. Remote Sens. Environ. 58:257–266.
• Gerwing, J., and R. Gelderman. 2005. Fertilizer recommendation guide. EC 750. Coop. Ext. Serv., South Dakota State Univ., Brookings.
• Gitelson, A., Y. Kaufman, and M. Merzlyak. 1996. Use of a green channel in remote sensing of global vegetation from EOS-MODIS. Remote Sens. Environ. 58:289–298.[CrossRef]
• Gitelson, A.A., and M. Merzlyak. 1994. Spectral reflectance changes associated with autumn senescence of Asculus hippocastanum and Acer platanoides leaves. Spectral features and relation to chlorophyll estimation. J. Plant Physiol. 143:286–292.[ISI]
• Gitelson, A.A., A. Viña, D.C. Rundquist, V. Ciganda, and T.J. Arkebauer. 2005. Remote estimation of canopy chlorophyll content in crops. Geophys. Res. Lett. 32:L08403 doi:10.1029/2005GL022688.
• Jackson, T.J., D. Chen, M. Cosh, F. Li, M. Anderson, C. Withall, P. Doriaswamy, and E. Hunt. 2004. Vegetative water content mapping using Landsat derived normalized water index for corn and soybean. Remote Sens. Environ. 92:475–482.
• Kim, K., D.E. Clay, S.A. Clay, and C.G. Carlson. 2004. Yield environment influence on N response. In Annual meetings abstracts [CD-ROM]. ASA, CSSA, and SSSA, Madison, WI.
• Kitchen, N.R., S.T. Drummond, E.D. Lund, K.A. Sudduth, and G.W. Buchleifer. 2003. Soil electrical conductivity and topography related to yield for three contrasting soil–crop systems. Agron. J. 95:483–495.[Abstract/Free Full Text]
• Li, H., R.L. Lascano, E.M. Barnes, J. Booke, L.T. Wilson, K.F. Bronson, and E. Segarra. 2001. Multispectral reflectance of cotton related to plant growth, soil water, and texture, and site elevation. Agron. J. 93:1327–1337.[Abstract/Free Full Text]
• Linn, D.M., and J.W. Doran. 1984. Effect of water-filled pore space on carbon dioxide and nitrous oxide production in tilled and non-tilled soils. Soil Sci. Soc. Am. J. 48:1267–1272.[Abstract/Free Full Text]
• O'Leary, M.H. 1993. Biochemical basis of carbon isotope fractionation. p. 19–28. In J.R. Ehleringer et al. (ed.) Stable isotopes and plant carbon water relations. Academic Press, New York.
• Pettigrew, W.T. 2004. Physiological consequences of moisture deficit in cotton. Crop Sci. 44:1265–1272.[Abstract/Free Full Text]
• Rouse, J.W., R.H. Haas, and D.W. Deering. 1974. Monitoring vegetation systems in the Great Plains with ERTS. p. 309–317. In ERTS Symp. 3rd, Goddard Space Flight Center, Greenbelt, MD. NASA SP-351. Vol. 1.
• Schmidt, U., H. Thoeni, and M. Kaupenjohenn. 2000. Using a boundary line approach to analyze N₂O flux data from agricultural soil. Nutr. Cycling Agroecosyst. 57:119–129.
• Shanahan, J.F., 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 to estimate corn grain yield. Agron. J. 93:583–589.[Abstract/Free Full Text]
• Smeltekop, H., D.E. Clay, and S.A. Clay. 2002. The impact of intercropping annual ‘Sava’ snail medic on corn production. Agron. J. 94:917–924.[Abstract/Free Full Text]
• Souza, R.P., E.C. Machado, J.A.B. Silva, A.M.M. Lagoa, and J.A.G. Silveira. 2004. Photosynthetic gas exhange, chlorophyll fluorescence, and some associated metabolic changes in cowpea (Vigna unguiculata) during water stress and recovery. Environ. Exp. Bot. 51:45–56.[CrossRef]
• Webb, R.A. 1972. Use of the boundary line in the analysis of biological data. J. Hortic. Sci. 47:309–319.

This article has been cited by other articles:

				A. E. Dellinger, J. P. Schmidt, and D. B. Beegle Developing Nitrogen Fertilizer Recommendations for Corn Using an Active Sensor Agron. J., October 21, 2008; 100(6): 1546 - 1552. [Abstract] [Full Text] [PDF]

				R. P. Sripada, J. P. Schmidt, A. E. Dellinger, and D. B. Beegle Evaluating Multiple Indices from a Canopy Reflectance Sensor to Estimate Corn N Requirements Agron. J., October 21, 2008; 100(6): 1553 - 1561. [Abstract] [Full Text] [PDF]

				E. M. Perry and D. A. Roberts Sensitivity of Narrow-Band and Broad-Band Indices for Assessing Nitrogen Availability and Water Stress in an Annual Crop Agron. J., June 23, 2008; 100(4): 1211 - 1219. [Abstract] [Full Text] [PDF]

				K.-I. Kim, D. E. Clay, C. G. Carlson, S. A. Clay, and T. Trooien Do Synergistic Relationships between Nitrogen and Water Influence the Ability of Corn to Use Nitrogen Derived from Fertilizer and Soil? Agron. J., May 7, 2008; 100(3): 551 - 556. [Abstract] [Full Text] [PDF]

				K. L. Martin, K. Girma, K. W. Freeman, R. K. Teal, B. Tubana, D. B. Arnall, B. Chung, O. Walsh, J. B. Solie, M. L. Stone, et al. Expression of Variability in Corn as Influenced by Growth Stage Using Optical Sensor Measurements Agron. J., February 6, 2007; 99(2): 384 - 389. [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 (5) Citing Articles via Google Scholar Google Scholar Articles by Clay, D. E. Articles by Dalsted, K. Search for Related Content PubMed Articles by Clay, D. E. Articles by Dalsted, K. Agricola Articles by Clay, D. E. Articles by Dalsted, K. Related Collections Water Stress Nitrogen Remote Sensing Maize Management