Progress 10/01/01 to 09/30/05
Outputs
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 to 1800 nm) collected at three growth stages (V8-V9, V11-VT, and R1-R2) were used to calculate six different remote sensing indices (NDVI, GNDVI, NDWI, NRI, Cgreen,, and Cred edge). 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 NIR (840 nm) band. At the V11-VT growth stage, reflectance in the blue,
green, and red bands were lower in fertilized than the unfertilized treatments and YLNS was negatively correlated to all of the indices except the normalized difference water index [NDWI= (NIR-MIR)/(NIR+MIR)], which was negatively correlated to YLWS (r = -0.28**). At the R1-R2 growth stage, YLWS was highly correlated (0.58**) to red reflectance and NDVI (-0.61**), while YLNS was correlated to 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 (between 100 and 120 kg N ha-1) for both moisture regimes.
Impacts
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. However, prior to 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. Many current techniques for using remote sensing for assessing N stress are based on research conducted in irrigated fields. Results from this research show that models based on this research may misdiagnosis water for N stress. Incorrect diagnosis can reduce profitability and increase the potential transport of N from agricultural fields to non-target areas.
Publications
- Ellsbury, M.M., S.A. Clay, D.E. Clay, and D.D. Malo. 2005. Within-field spatial variation of northern corn rootworm distributions. P 145-154, S. Videl et al. (ed.). Western Corn Rootworm: Ecology and Management. CAB International. Oxfordshire UK.
- Paz, J.O., W.D. Batchelor, D.E. Clay, S.A. Clay, and C. Reese. 2003. Characterization of Soybean Yield Variability Using Crop Growth Models and 13C Discrimination. ASAE meeting presentation # 033044.
- Chang, J., D. E. Clay, K. Dalsted, S.A. Clay, M. O Neill. 2004. Use of spectral radiance at multiple sampling dates to estimate corn Zea mays) yield using principal component analysis. Agron. J. 95:1447-1453.
- Sudduth, K.A., N.R. Kitchen, W.J. Wiebold, W.D. Batchelor, G.A. Bollero, D.G. Bullock, D.E. Clay, H.L. Palm, F.J. Pierce, R.T. Schuler, K.D. Thelen. 2005. Relating apparent electrical conductivity to soil properties across the north-central USA. Computers and Electronics in Agriculture 46:263-283.
- Clay, D.E., Z. Zhen, Z. Liu, S.A. Clay, and T.P. Trooien. 2004. Bromide and nitrate leaching in undisturbed soil columns collected from three landscape positions. J. Environ. Qual. 33:338-342.
- Chang, J., and D.E. Clay. 2006. Identifying factors for yield prediction models and evaluating model selection methods. Korean Crop Sci. Soc. J. Korean J Crop Science 50:268-275.
- Clay, D.E., Ki-In Kim, J. Chang, S.A. Clay, and K. Dalsted. 2006. Characterizing water and N stress in corn using remote sensing. Agron. Journal. (In press)
- Clay, D.E., S.A. Clay, D.J. Lyon, and J.M. Blumenthal. 2005. Can 13C discrimination in corn grain be used to characterize intra-plant competition for water and nitrogen? Weed Sci. 53:23-29.
|
Progress 01/01/04 to 12/31/04
Outputs
Misdiagnosis of N and water stress in crop plants can have negative financial consequences. The objective of this study was to develop remote sensing-based models for predicting yield losses due to water (YLWS) and N (YLNS) stress in corn (Zea mays L.). A replicated N and water treatment factorial experiment was conducted in 2002 and 2003. YLWS and YLNS were determined using the 13C discrimination approach. Reflectance in the blue (486 nm), green (560 nm), red (660 nm), NIR (830 nm), and MIR (1650 nm) were measured at the V-8-V9, V11-VT, and R1-R2 growth stages. Crop yields were increased by adding N and supplemental irrigation. At the V8-V9 and V11-VT growth stages, crop reflectance was primarily influenced by N stress. At the V8-V9 growtg stage YLNS was negatively correlated to GNDVI (r = -0.46) and positively correlated to green (r= 0.570), red (r = 0.501), and MIR (r = 0.407). Similar results were observed at the V11-VT growth stage where GNDVI was negatively
correlated to YLNS (r = -0.62)and positively correlated to green (r = 0.617), red (r = 0.357), and MIR (r= 0.527). At the R1 to R2 growth stage NDVI was negatively correlated to YLNS (r= -0.409) and YLWS (r= -0.424). However, important differences existed; YLWS was positively correlated to blue (r = 0.518), red (r = 0.394), and MIR (r = 0.361) while YLNS was positively correlated to green (r = 0.605) and negatively correlated to NIR (r = -0.437) and GNDVI (r = -0.689). Remote sensing models, based on GNDVI, blue, green, and MIR bands, explained between 61 and 65% of the observed yield variability. For YLNS, the remote sensing models, based on GNDVI, blue, green, NIR, and 1650 nm bands, explained between 54 and 61% of the observed variability. At the R1-R2 growth stage, a remote sensing model that relied on the NDVI and GNDVI indexes explained 39% of the observed YLWS variability. Results from this study suggest that remote sensing models can be used to develop treatment maps for N
deficiencies.
Impacts
Predicting corn yields is important for on-farm planning, evaluating food security needs, estimating food shortages, and developing input data sets for C sequestration and watershed-based water quality models. A number of different approaches and band combinations have been proposed for estimating yields. To date there is no universally accepted approach. This research indicates that the best time to collect remote sensing for predictive purposes is between R3 and R4.
Publications
- Chang, J. D. E. Clay, C. G. Carlson, C. L. Reese, S. A. Clay, and M.M. Ellsbury. 2004. The Influence of different approaches to define yield goals and management zones on N and P fertilizer recommendations errors. Agron. J. 96:825-831.
|
Progress 01/01/03 to 12/31/03
Outputs
Producers commonly ask, when is the best time to collect remote sensing for identifying management zones and estimating crop yields? The goal of this project is to determine the influence of precipitation and Landsat collection date on corn (Zea mays) yield whole field yield predictions. Landsat data (Row and path 29) was collected (bare soil, R1, and R3-R4) from fields harvested with calibrated combines equipped with yield monitors and global positioning systems in 2000, 2001, and 2003. Landsat data was georectified, converted to a reflectance basis, and whole field reflectance values determined. Whole field reflectance values were merged with cleaned yield monitor data files and whole field yield prediction models were developed using the backward regression approach. Remote sensing collected early in the growing season provided information about soil drainage, color, and previous erosion, while remote sensing collected later in the growing season provided
information about crop maturity. Yield models based on rainfall and remote sensing collected at R1 and R3-R4 explained 75 and 93% of the whole field variability, respectively, while models based on bare soil only explained 45% of the yield variability. For remote sensing data collected at R1 and R3-R4, including during the growing season rainfall into the model improved yield predictions. These results indicated that information collected from two different scales (regional weather station network and yield monitor data) can be combined to develop whole field yield estimates. This research demonstrated that producer collected yield monitor data can be used to develop and test yield models based on remote sensing data. Based on these findings, the best time to collect remote sensing for predicting corn yields and identify productivity zones is late August or early September (R3-R4).
Impacts
Predicting corn (Zea mays L.) yields is important for on-farm planning, evaluating food security needs, estimating food shortages, and developing input data sets for C sequestration and watershed-based water quality models. A number of different approaches and band combinations have been proposed for estimating yields. To date there is no universally accepted approach. This research indicates that the best time to collect remote sensing for predictive purposes is between R3 and R4.
Publications
- Clay, D.E., C.G. Carlson, and J. Chang. 2004. Identifying the Best Approach to Identify Nutrient Management Zones: A South Dakota Example SSMG 41. Clay et al. (Ed) Site Specific Management Guidelines. Potash and Phosphate Institute. Norcross, GA.
- Ellsbury, M.M., S.A. Clay, D.E. Clay, and D.D. Malo. 2004. Within-field spatial variation of northern corn rootworm distributions. In p. (ed), at the 1994 Gottingen Rootworm Symposium.
- Gaspar, P., C.G. Carlson, and D.E. Clay. 2003. A Cookbook approach for determining the point of maximum economic return. SSMG 39. Clay et al. (Ed) Site Specific Management Guidelines. Potash and Phosphate Institute. Norcross, GA.
- Dalsted, D. J. Paris, D. Clay, S.A. Clay, C. Reese, and J. Chang. 2003. Selecting the Appropriate Satellite Remote Sensing Product for Precision Farming. SSMG 40. Clay et al. (Ed) Site Specific Management Guidelines. Potash and Phosphate Institute. Norcross, GA.
- Chang, J., D. E. Clay, K. Dalsted, S.A. Clay, M. O Neil. 2004. Use of spectral radiance at multiple sampling dates to estimate corn (Zea mays) yield using principal component analysis. Agron. J. 95:1447-1453.
- Clay, D.E., S.A. Clay, J. Jackson, K. Dalsted, C. Reese, Z. Liu, D.D. Malo, and C.G. Carlson. 2003. C13 discrimination can be used to evaluate soybean yield variability. Agron. J. 95:430-435.
- Chang, J., D.E. Clay, C.G. Carson, S.A. Clay, D.D. Malo, R. Berg, and W. Wiebold. 2003. The influence of different approaches for defining nutrient management zone boundaries on N and P recommendations. Agron. J. 95:1550-1559.
|
Progress 01/01/02 to 12/31/02
Outputs
Commonly accepted techniques to identify and evaluate management zone effectiveness are not available. The objectives of this study were to propose an approach for evaluating zone boundaries and to use this approach to determine the impact of different approaches for identifying zone boundaries on N and P fertilizer recommendations. Two statistics, pooled variances and mean square error (MSE), were used to assess the effectiveness of the different management zone boundaries. A F statistic was used to test these values. This study was conducted in two east central South Dakota fields between 1995 and 2000. The crop rotation was corn followed by soybean. Information used to identify zone boundaries were apparent electrical conductivity, elevation, aspect, order 1 soil survey, and the location of old homesteads. Corn and soybean yields had a significant amount of temporal and spatial variability. Yields were relatively stable in summit/shoulder, moderately stable in
backslope, and unstable in footslope areas. In footslope areas, yields were limited by too much water in wet-spring years, while in summit/shoulder areas yields were limited by too little water. The 4-ha block sampling had lower P and N fertilizer recommendation MSE values than the other techniques tested. Sampling the old homesteads separately from the rest of the field reduced the P recommendation MSE and did not influence the N recommendation MSE. Relative to whole field sampling and using a county average for the yield goal, the 4-ha block soil sampling and using season and landscape specific yield goals reduced P and N fertilizer MSE values 29 and 45%, respectively. The strategy with the lowest P and N MSE was the 4-ha block sampling combined with season/landscape specific yield goals.
Impacts
Fields are a mosaic of habitat types, each having unique characteristics that influence soil properties and crop yields. The effectiveness of identifying different productivity zones in production fields rests on the ability to identify the boundaries associated with different habitats. Information used to identify zone boundaries includes biological properties, farmer preferences, soil nutrients, soil maps, remote sensing, elevation, yields, and electrical conductivity information. A commonly accepted approach for testing the effectiveness of the zone boundaries is not available. In collaboration with producer collaborators, this research has been testing different approaches for identifying productivity zones in production fields.
Publications
- Clay, D. E., Kitchen, N., Carlson, C. G., Kleinjan, J. L., and Tjentland, W. A. 2002. Collecting representative soil samples for N and P fertilizer recommendations. Online. Crop Management doi:10.1094/CM-2002-12XX-01-MA.
- Clay, D.E., J.Chang, D.D. Malo, C.G. Carlson, C.Reese, S.A. Clay, M. Ellsbury, and B. Berg. 2001. Spatial variability of soil apparent electrical conductivity. Comm. Plant and Soil Analysis 32:1813-1827.
- Clay, D.E., S.A. Clay and C.G. Carlson. 2002. Site specific management from a cropping system perspective. Srinivasan, A. (ed) In Precision Farming- A Global Perspective.
- Clay, D.E., R. Engel, D. Long, and Z. Lui. 2001. Nitrogen and water stress interact to influence carbon-13 discrimination in wheat. Soil Sci. Soc. Amer. J. 65:1823-1828.
|
|