Illinois
Fertilizer Conference Proceedings
January 26-28, 1998
| Home | 1998 Index | Search |
| Home | 1998 Index | Search |
J.S. Schepers, D.S. Hagopian, and G.E. Varvel1
The concept of precision agriculture encompasses a number of assumptions and expectations on the part of producers. The assumptions involve the premise that spatial variability in crop yields can be explained in terms of inadequate fertility, water or temperature stress, insect infestations, weed pressures, or cultural practices that affect plant population and vigor. The expectations are that if producers understand the causes of spatial variability, they will have the option to take action to address the problems.
Fallacies in the assumptions are revealed when producers or others are not able to identify the causes of spatial variability in yields. Even when causes of yield variability are hypothesized, such thoughts are frequently expressed at harvest with little or no quantitative data to support the contention. When explanations are not readily forthcoming, concerns regarding crop nutrient availability frequently come into question.
Nutrient availability questions are more complicated than they may seem because crop uptake is a season-long process. Not only knowing which nutrients are limiting, but when they become limiting plays a part in designing site-specific management practices to reduce spatial variability. Nitrogen (N) is probably the most difficult nutrient to manage because it can exist in several forms, it can be lost through denitrification or nitrate leaching, its availability to crops is influenced by soil microbial activity, and it is used in relatively large quantities by crops.
Meeting crop N needs can be a major challenge considering all of the factors that can affect its availability to crops. Nitrogen fertilizer costs and environmental concerns place additional constraints on N management considerations. Because optimum economic yields for corn require only slightly less N fertilizer than for maximum yields, there isn't a great incentive to manage N closely. This is why producers find it disturbing any time they observe signs of N deficiency. Fortunately, most N fertilizers are soluble in water and it is possible to get N into the root zone any time during the growing season. The limitations are equipment to apply N fertilizer to tall growing crops and detecting when additional N is required.
Characterizing the N status of crops can be accomplished using traditional tissue testing procedures or by using a chlorophyll meter to indirectly measure photosynthetic activity (Blackmer et al., 1993; Blackmer and Schepers, 1995a and 1995b; Schepers and Blackmer, 1997). Chlorophyll meters operate by sequentially directing red and NIR light onto a small section of a leaf blade and measuring the relative amounts of each waveband (i.e., color) that passes through the leaf (Peterson et al., 1993). Red light is used by plants during photosynthesis. The more red light that passes through the leaf, the lower the photosynthesis rate. The linkage to N status is that increased levels of N availability usually correspond to higher photosynthetic rates, at least up to a point.
Assessing crop N status using chlorophyll meters works well for research plots, but is not practical for an entire field. As a result, canopy sensors were developed to rapidly monitor crop N status. The premise is that healthy and adequately fertilized corn plants are relatively dark green in color compared to N deficient plants. Chlorophyll meter readings increase with fertilizer N rate up to a point when something besides N begins to limit photosynthetic activity. Even though chlorophyll meters measure photosynthesis, they indirectly measure leaf greenness which is an indication of crop N status. The relationship between the amount of green and NIR reflected light from a crop canopy and N status is the basis for the Li-Cor sensor system.
The objective of this research was to develop techniques to monitor crop N status for the purpose of scheduling supplemental N fertilizer applications.
Minolta chlorophyll meters and Li-Cor canopy sensors were used to characterize the N status of irrigated corn. The study consisted of four hybrids fertilized at five N rates (0, 50, 100, 150, and 200 kg/ha). Chlorophyll meter readings were collected throughout the growing season on a weekly interval. Prior to silking, the uppermost fully expanded leaf was monitored and thereafter the ear leaf was monitored (Peterson et al., 1993). Li-Cor sensor readings were collected after silking by driving a high-clearance sprayer through the plots. Five sensors were positioned approximately four feet above the canopy and directed at a 45/ angle perpendicular to the rows. This positioning maximized the viewing of the crop canopy and minimized the bare soil influences. Data recorded by the Li-Cor sensors included readings for a broad band in the green portion of the spectra (centered at 550 nm) and another band in the near infrared (NIR) portion of the spectra (centered at 850 nm). Another detector was pointed upward to measure changes in light intensity caused by variable cloud cover during the time measurements were taken. Data from the upward-pointing sensor were used to adjust data from the down-pointing sensors.
Chlorophyll meter readings showed that adequately fertilized plots (i.e., highest N rates) consistently had the highest values throughout the season (Figure 1). Even plots receiving reduced amounts of N fertilizer tended to recover from the early season N stress as the season progressed. This recovery is partially attributed to irrigation with high nitrate water that supplies 7 lb N/acre-inch of water applied. Also, crop N needs decline after silking while N mineralization continues to supply N to the crop.
Even though the degree of early season N stress diminished with time, the effects of the deficiency were seen throughout the year in terms of reduced plant vigor and smaller plants. Shortly after silking, chlorophyll meter readings for the four hybrids followed similar trends, but had different values for the same N rate (Figure 2). Figure 1 and Figure 2 illustrate why it is risky to establish a "target" chlorophyll meter reading considering the temporal and hybrid differences. This is why calculating a sufficiency index is a wise approach when dealing with biological systems (i.e., meter value for plants in question divided by value from adequately fertilized plants).
Chlorophyll meter readings represent a "point-in-time" measurement and are not able to sense the cumulative effects of plant vigor such as smaller plants, reduced leaf area (Figure 3). In contrast, the Li-Cor sensor system monitors the crop canopy rather than a small area on individual leaves. Another difference between the two monitoring systems is that the chlorophyll meter approach measures the upper-most expanded leaf until silking, which is usually about four leaves below the newest leaf. Young corn leaves are typically somewhat striated and yellowish until they become fully expanded. These immature leaves are the ones that would be monitored with Li-Cor canopy sensors and are probably a better indicator of current N status than more mature leaves. The reason more mature leaves are monitored when using chlorophyll meters is to minimize variability caused by non-uniform color that is common in young leaves when taking a limited number of measurements from small areas on leaf blades.
The Li-Cor sensors used in this study were designed to measure reflectance of green and NIR portions of the spectra (Blackmer et al., 1994; Blackmer and Schepers, 1996, Blackmer et al., 1996). The major differences in NIR reflectance between hybrids in Figure 4 are due to architecture of the leaves. Normal (red) NDVI is used to measure green biomass and estimate changes in vegetative state, but is insensitive to all but low chlorophyll contents. Gitelson et al. (1997) found that reflectance near 550 nm is sensitive to a full range of chlorophyll contents. Therefore, an index that uses green instead of red reflectance will enable a more precise estimation of pigment content. Combining data from the green and NIR bands into what is called a green NDVI (normalized difference vegetative index) according to the following expression generates a graph that mimics the chlorophyll meter graph (Figure 4).
Green NDVI = (NIR - green)/(NIR + green)
Theoretically, the green NDVI value should integrate crop N status (greenness) and biomass (NIR) into something that is indicative of yield potential. While crop greenness certainly influences dry matter production, other factors like plant population and weed infestations influence the amount of biomass monitored by the NIR sensor. Perhaps this accounts for the less than earth shaking trends in Figure 5. Nonetheless, the sensors provide a reasonable comparison of yield potential within hybrids. One caution worth noting when using any kind of sensor is that interactions between different stresses may affect the various wavebands differently. In the case of water stress, the affect on reflectance in the green or red bands is much less than on the NIR band (Schepers and Blackmer, 1996). Therefore, just because one is using a sensor to monitor crop N stress doesn't mean you can turn off your brain.
1J.S. Schepers, D.S. Hagopian and G.E.
Varvel are Soil Scientists, USDA–Agricultural Research Service, Agronomy
Department, University of Nebraska, Lincoln, NE.
Blackmer, T.M. and Schepers, J.S. 1995a. Use of a chlorophyll meter to monitor
N status and schedule fertigation for corn. J. Prod. Agric. 8:56-60.
Blackmer, T.M. and Schepers, J.S.1995b. Use of a chlorophyll meter to monitor N status and schedule fertigation for corn. J. Prod. Agric. 8:56-60.
Blackmer, T.M. and Schepers, J.S. 1996. Aerial photography to detect nitrogen stress in corn. J. Crop Phys. 148:440-444.
Blackmer, T.M., Schepers, J.S. and Vigil, M.F. 1993. Chlorophyll meter readings in corn as affected by plant spacing. Comm. Soil Sci. Plant Anal. Vol. 24, No. 17-18, 2507-2516.
Blackmer, T.M., Schepers, J.S., and Varvel,G.E. 1994. Light reflectance compared to other N stress measurements in corn leaves. Agron. J. 86:934-938.
Blackmer, T.M., Schepers, J.S., Varvel, G.E., and Walter-Shea, E.A. 1996. Nitrogen deficiency detection capabilities of light reflection from irrigated corn canopies. Agron. J. 88:1-5.
Gitelson, A.A., Y.J. Kaufman, and M.N. Merzlyak. 1997. Remote sensing of pigment content in higher plants: principals and techniques. Proc. 3rd International Airborne Remote Sensing Conference and Exhibition. July 7-10, 1997, Copenhagen, Denmark.
Peterson, T.A., Blackmer, T.M., Francis, D.D., and Schepers, J.S. 1993. Using a chlorophyll meter to improve N management. Coop. Exten. Service, Univ. Nebr.-Lincoln, NebGuide G93-1171A.
Schepers, J.S. and Blackmer, T.M. 1997. Chlorophyll meter method for determining N. Handbook on Reference Methods for Plant Analysis. Soil Testing and Plant Analysis Council. (in press)
Schepers, J.S., Blackmer, T.M., Wilhelm, W.W., and Resende, M. 1996. Transmittance and reflectance measurements of corn leaves from plants with different nitrogen and water supply. J. Plant Phys. 148:523-529
