Site-specific management of fertilizer depends upon accurate knowledge about the source of spatial variability in crop yield. Variations in landscape characteristics and soil properties are thought to be responsible for much of the variation in crop yield. Landscape characteristics, soil properties, and corn/soybean yield data were collected from four transects (320 to 365 m long) at 30 m intervals. Corn/soybean yields were less at eroded slopes than at the nearly level summit or at the foot/toeslope positions. Corn yield was positively correlated with A horizon thickness and negatively correlated with surface pH. A horizon thickness, surface pH, tillage system, and growing season precipitation explained 72% of the variability in corn yield. Slope gradient, depth to free CaCO3, soil profile water storage, surface available K and growing season precipitation explained 44% of the variability in soybean yield.

Within-field nutrient variability causes some areas of a field to be more or less responsive to fertilization. The best soil sampling and fertilization strategies are those that best estimate and apply economic optimum fertilizer rates across a field. Although current site-specific management practices could achieve this goal, questions remain concerning the cost-effectiveness of alternative sampling strategies. This study compared various soil sampling schemes in eight fields using soil test P, K, pH, and organic matter. The schemes were grid sampling, sampling by digitized and detailed soil survey map units, sampling by elevation, and a targeted sampling based on various layers of information. The soil-test variability patterns varied markedly across the fields. The schemes varied greatly in reducing the within-unit variability and in the recommended fertilizer rate, but no scheme was superior across all fields and nutrients. The efficacy of all sampling schemes was lower for P and K, probably because of the larger impact of fertilization on small-scale variability. Over all fields, only schemes based on field averages or digitized soil survey maps resulted in significantly lower correctly fertilized areas than other schemes. For P, the most highly variable nutrient, the grid and targeted schemes usually were similarly effective, although the latter sometimes required fewer sampling units. Consideration of costs, field fertilization history, and likely response to fertilization is needed to select among various similarly effective sampling strategies.

Research was conducted to investigate the effect of N (N) fertilizer rate on corn grain yield and ethanol yield at three sites in Minnesota. A field-scale split plot design was used at each site with hybrid as the sub plot factor and a randomized complete block design containing four replications of four N treatments (0, 112, 168 and 201 kg ha-1) as the whole plot factor. Hybrid had a significant effect on grain and ethanol yields. The effect of N rate on protein content, starch content, and ethanol yields was dependent on hybrid and growing season weather conditions. As N fertilizer rate increased, protein contents increased, while starch content and ethanol yield decreased. Ethanol yield was optimized at the economically optimum N rate determined from corn yield response. Spatial patterns in corn protein content, starch content, and ethanol yields were significant and exhibited strong spatial structure. Precision management of N fertilizer rate across the landscape did not show significant effects on grain quality or ethanol yield. These results indicate that selection of hybrid is the most important factor affecting corn and ethanol yield. N rate is of secondary importance.

This study was conducted in Paris, IL, from 2001 to 2003 involving three corn ( Zea mays L.) hybrids, five N rates (0, 112, 168, 224, and 336 kg ha −1 ), and six site‐year comparisons to determine the significance of within‐field variation in corn yield and quality responses to N fertilization, differences between hybrids in yield and quality, and the feasibility of within‐field variable hybrid selection. On average, N fertilization significantly increased corn yield, protein content, and test weight, but decreased corn oil and starch content. The overall economically optimum nitrogen rate (EONR) was 125 kg ha −1 , but EONR varied from 93 to 195 kg ha −1 in different environments. The N rates that would maximize protein content and test weight (MAXN) varied from 143 to 303 kg ha −1 and 0 to 235 kg ha −1 in different environments, respectively. Significant within‐field variability in N response was detected in five of six environments for yield, but not in more than two environments for any quality parameter. Hybrid differences were significant in all six environments for test weight, followed by oil content (five), protein and starch content (four), and yield (three). Hybrid differences between 33G26 and 33J24 in test weight response to N were consistent across environments, showing the potential of hybrid‐specific N management for this quality parameter. However, hybrid differences in yield and quality did not vary significantly over space in most environments, showing limited potential of within‐field variable hybrid selection. Further studies involving more diverse within‐field soil–landscape conditions and hybrids are needed.

Typically, Indonesian oil palm plantations rest on rolling topography. There is limited information on how topography affects soil fertility and oil palm yields. A study was conducted to evaluate these relationships in a commercial oil palm plantation located in South Sumatra, Indonesia. Two sites with differing past management history and fertility regimes were each partitioned into three topographic positions. At each topographic position, yields were recorded at 10-day intervals over a period of 2 years. Leaf and soil samples were collected from corresponding points spaced at 36.4 m (x direction) and 8.7 m (y direction) using a systematic scheme. Leaf analysis was performed to quantify nitrogen (N), phosphorus (P), potassium (K), magnesium (Mg), and calcium (Ca), and soil analysis was carried out to determine pH, organic carbon (C), extractable P, exchangeable K, Mg, and Ca, effective cation exchange capacity (ECEC), and texture. The collected data were subjected to exploratory, univariate, and bivariate analyses, as well as analysis of variance. Empirical production functions based on measured variables were defined for each topographic position. Results showed that average yields at both study sites varied with topographic position. At site 1 (Sungai Pelepah Estate), the sideslope and the summit consistently gave higher yields than the toeslope. At site 2 (Sri Gunung Estate), a yield gradient was observed with the highest yield occurring at the toeslope and the lowest yield from the summit. Soil fertility varied across topographic positions at both sites. The measured leaf/soil variables showed varying levels of optimality/sufficiency across topographic positions. In most cases, leaf and soil variables showed comparable performance as yield predictors. Validation of the calibrated models showed reasonable accuracy for the toeslope of site I and all three positions at site 2.

Better understanding of within-field spatial variability of crop quality parameters and yield are needed for precision management of crops. This study was conducted to determine the magnitude of within-field variability in soil properties, corn ( Zea mays L.) quality parameters and yield and to characterize their spatial structures. Another objective was to compare the effects of hybrid on corn quality, yield, and the spatial structure of grain quality. Four Pioneer hybrids were planted side-by-side, two in each of the two study fields in eastern Illinois, USA. Coefficients of variation (CV%) for soil properties varied from 6.3 (pH) to 56.8% (soil test P). All the soil properties (except pH at Site 2) displayed well-defined spatial structures, with either strong or moderate spatial dependence. Variability in corn quality and yield (CVs < 10%) was smaller than variability in soil properties. Most quality parameters examined at Site 1 exhibited either moderate or strong spatial dependence, except that corn oil (both hybrids), kernel roundness and weight (hybrid 33Y18) did not show any spatial correlation. Hybrid 33G26 had significantly higher yield and quality for most quality parameters than 33Y18 at Site 1. At Site 2, hybrid 34W67 was significantly lower in oil and protein content, length, roundness and vitreousness than 34K77, but higher in other quality parameters. Significant differences in spatial structures were also observed across hybrids for some corn quality parameters. We conclude that hybrid selection is an important strategy for precision management of corn for optimum yield and quality.

This study evaluated an integrated approach to delineate site-specific management zones (MZ) using relative elevation, organic matter, slope, electrical conductivity, yield spatial trend map, and yield temporal stability map (ROSE-YSTTS) against two other approaches using only soil and landscape information (ROSE) or clustering multi-year yield maps (CMYYM) oil two no-till corn (Zea mays L.)-soybean [Glycine max (L.) Merr] rotation fields in eastern Illinois. It was found that the ROSE approach was least effective in accounting for crop yield variability, while the CMYYM approach was least effective in accounting for soil and landscape variability. The integrated ROSE-YSTTS approach was reasonably effective in accounting for both soil-landscape and yield variability. We conclude that the ROSE-YSTTS approach to MZ delineation procedure can overcome the weaknesses of approaches that are based only oil soil and landscape or yield information, and thus is more likely to be useful for management purposes.

Materials and Methods: This research was conducted in a 30-hectare field located in Central South Dakota at Dakota Lakes Research Farm in 2001. The field was located at 99.998534 W longitude and 44.357852 N latitude. The field was not irrigated and had a 23-meter (m) elevation change. The field was harvested (31 July 2001) with a combine equipped with a calibrated yield monitor. During harvest, grain samples were collected from the combine. Using Geographic Information Systems (GIS) software, yield in three different landscape positions (footslope 533-540 m; backslope, 541-548 m; and summit, 549-556 m) was determined. 'Russ' hard red spring wheat was planted on April 16 and 17, 2001 at 146 kilograms/hectare (kg/ha). Monoammonium phosphate (MAP) was placed with the seed at 78 kg/ha. The seed was treated with Vitavax RTU. Weed control was achieved using 0.47 liters Bronate (bromoxynil plus MCPA) per acre with 45 liters of water. Nitrogen treatments using urea (46-0-0) were applied on May 16. The nitrogen was surface banded using a modified Concord 110 air seeder cart and spray boom with outlets on 50.8-cm centers. A 12.7-cm band of fertilizer was applied using this method on 50.8-cm centers. Nitrogen was applied at three rates in strips that encompassed the entire length of the field. These strips were replicated 3 times across the field. The nitrogen treatments were randomized in the strips. The nitrogen rates were 0 kg N/ha, 100 kg N/ha and 200 kg N/ha.