Agronomy Journal Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Katsvairo, T. W.
Right arrow Articles by Glos, M.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Katsvairo, T. W.
Right arrow Articles by Glos, M.
Agricola
Right arrow Articles by Katsvairo, T. W.
Right arrow Articles by Glos, M.
Related Collections
Right arrow Maize
Right arrow Production Agriculture
Right arrow Site-Specific Analysis
Agronomy Journal 95:1012-1022 (2003)
© 2003 American Society of Agronomy

NITROGEN MANAGEMENT

Spatial Yield Response of Two Corn Hybrids at Two Nitrogen Levels

Tawainga W. Katsvairo, William J. Cox*, Harold M. Van Es and Michael Glos

Dep. of Crop and Soil Sci., 609 Bradfield Hall, Cornell Univ., Ithaca, NY 14853

* Corresponding author (wjc3{at}cornell.edu)

Received for publication July 1, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSION
 REFERENCES
 
The challenge for variable N rate management is to identify specific field areas that respond to specific N levels. We evaluated two corn (Zea mays L.) hybrids at two N rates (110–130 vs. 165–185 kg ha-1 at three sites and manure vs. manure + 55 kg ha-1 at two sites) to determine if corn responded differently to N and hybrids within fields. Spatial yield variability existed at all sites in dry years (1999 and 2001) but at only two sites in a wet year (2000). Spatial yield difference variability in response to N existed at only two of 13 site–year comparisons. Although late-spring soil NO3–N concentrations in the upper 30 cm were <25 mg kg-1 on 15 to 25% of the manured fields in the wet year, spatial yield difference variability in response to N did not exist. At a nonmanured site, spatial yield difference variability in response to N existed with temporal yield stability across dry years (r = 0.96). Surprisingly, corn responded to the higher N rate on 25% of this field where yields were least, but not where yields were greatest. Apparently, variable N rate management of corn requires more information than soil NO3–N concentrations and yield maps. Spatial yield difference variability between hybrids existed at only four of 15 site–year comparisons, despite hybrid interactions with sites. Adoption of variable hybrid selection is unlikely if hybrids that show interactions with sites do not show spatial yield difference variability within sites.

Abbreviations: dGPS, differentially corrected global positioning system • PSNT, pre-sidedress nitrogen test • Vn, nth leaf stage


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSION
 REFERENCES
 
CORN YIELDS often show significant spatial variability (Doerge, 1999) because of numerous factors including topography (Timlin et al., 1998; Kravchenko and Bullock, 2000), soil water content (Sadler et al., 2000), soil fertility (Mallarino et al., 1999), and weed control (Mallarino et al., 1999). Spatial variability of corn yield creates a significant challenge for N fertility management because excessive N can result in NO3–N contamination of surface and ground water and inadequate N results in yield and profit losses (Dinnes et al., 2002). Consequently, the application of variable N fertilizer rates has the potential to reduce NO3–N contamination of surface and ground water and increase profits by matching N fertilizer rates to the N fertility requirement of corn in different areas of a field (Doerge, 2002). Schmidt et al. (2002), however, reported that the N fertility requirement of corn was not consistently related to the soil organic matter content within a field. Also, Sogbedji et al. (2000) concluded that the N fertility requirement of corn was not consistently related to the soil drainage class within a field. The challenge for variable N rate management is to identify specific areas of a field that require more or less N for optimum corn yields.

Corn yields frequently show more temporal variability than spatial variability (Jaynes and Colvin, 1997; Lamb et al., 1997; Porter et al., 1998; Sogbedji et al., 2001). Consequently, the use of yield goal-based N applications, based on yield map data, may not be conducive to variable N rate management because high-yielding areas of a field are not consistent from year to year. Sogbedji et al. (2001) concluded that the application of sidedress N fertilizer rates should be based on late-spring weather conditions because temporal variability exceeds spatial variability of corn yield. They used data from field experiments and the LEACHM simulation model (Hutson and Wagenet, 1992) to demonstrate that economic and environmental benefits are gained from annual adjustments of N fertilizer based on late-spring weather conditions.

The presidedress soil nitrate test (PSNT), however, should reflect spring weather conditions in the northeastern USA by accounting for the net effects of weather conditions on mineralization of N, leaching, and denitrification of NO3–N (Magdoff, 1991). Durieux et al. (1995) reported that in Vermont the use of the PSNT instead of yield goal-based N recommendations reduced N rates by about 50 kg N ha-1, maintained corn yields, and reduced the leaching potential of NO3–N. Sogbedji et al. (2000) demonstrated in a lysimeter study that the use of the PSNT reduced fertilizer N use and lowered NO3–N concentrations in shallow ground water. Bundy et al. (1999), however, reported that in a regional study the use of the PSNT resulted in the overfertilization of N in some fields.

Variable N rate management has the potential to improve fertilizer N efficiency. The use of the PSNT as a guide should improve the success of variable N rate management, especially on manured fields, because it can identify specific areas of the field that should respond to additional N fertilizer. The first objective of the study was to determine if corn responded differently to N fertilization within individual fields that differ in soil NO3–N concentrations, based on the PSNT. A second objective of the study was to determine if corn responded differently to hybrid selection within individual fields. To date, no studies have documented the response of variable hybrid selection within individual fields.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSION
 REFERENCES
 
We formed farmer–research partnerships (Karlen et al., 1995) to conduct two field-scale studies on a dairy farm and three field-scale studies on two cash crop farms in central New York. At the dairy site, both fields (Onondaga 1 and 2) had been in alfalfa (Medicago sativa L.) from 1994 to 1997 and in corn in 1998. Consequently, the farmer planted second, third, and fourth-year corn in 1999, 2000, and 2001, respectively, which is consistent with the typical rotation of 4 yr of perennial forage followed by 4 yr of corn on dairy farms in New York. At the cash crop sites, the three fields (Seneca 1, 2, and 3) had been in a corn–soybean [Glycine max (L.) Merr.] rotation in the 1990s with soybean planted in each field in 1998. Consequently, the farmers planted first, second, and third-year corn in 1999, 2000, and 2001, respectively. In April of 1999, all fields tested medium to high in P and K with soil pH values above 6.5. Table 1 provides more detailed information on the five sites.


View this table:
[in this window]
[in a new window]
  		Table 1. Field size, dominant soils, and locations of the five sites in central New York, and the starter and sidedress N fertilizer rates for the low (L) and high (H) N levels in 1999, 2000, 2001.

 
The farmers plowed each field in the spring and performed their typical secondary tillage operations for corn planting. Two hybrids (Pioneer brand ‘3752’ and ‘37M81’) were planted at each site in each year in early to mid-May at about 80000 kernels ha-1 with 12-row planters at 0.76-m row spacing. We selected 3752 and 37M81 because they had shown hybrid x location interactions under New York growing conditions (Don Specker, personal communication, 1999). We used the split-planter technique (one hybrid assigned to the left-side hoppers and the other assigned to the right-side hoppers) for planting, which resulted in 12 rows of one hybrid alternating with 12 rows of another hybrid throughout the entire field. After planting, the farmers used recommended pest management practices to control insects and weeds. Insects and weeds were not a problem at any of the sites in any of the years.

A systematic unaligned sampling grid (Wollenhaupt et al., 1994) with a spacing of 50 by 36 m was superimposed over each experimental site in the spring of 1999 to establish sampling locations or stations. The location coordinates for each station were recorded with a differentially corrected global positioning system (dGPS). Soil samples were then collected from the 0- to 30-cm depth in mid-June of 1999 at the 4th leaf stage (V4) of corn growth (Ritchie et al., 1993) at each station. All soil samples were a composite of 10 soil cores taken on a 3-m grid centered on the georeferenced station. Subsamples from each 10-core composite were sealed in plastic bags, placed in a cooler in the field, transferred to cardboard cartons at the end of the day, and placed in a forced-air oven that was set at 55°C. After drying to constant moisture, the samples were analyzed for NO3–N colorimetrically with an autoanalyzer (Alpkem Corp, Clackamas, OR). Subsequent sampling was done at each station in 2000 and 2001. We used the dGPS at each site in 2000 and 2001 to navigate to the pre-established stations.

After soil sampling, the farmers sidedressed their fields with anhydrous ammonia in alternate rows at two N rates. The sidedressed N rates at the cash crop sites varied according to the amount of N applied in their starter fertilizer (Table 1). At each cash crop site, we aimed to apply about 30 kg N ha-1 above or below the recommended N rate (135 kg N ha-1 for corn following soybean and 160 kg N ha-1 for corn following corn for the specific soils at the cash crop sites; Cornell Univ., 1999). The farmers applied a total of 110 kg N ha-1 for the low N treatment and 165 kg N ha-1 for the high treatment in 1999 when corn followed soybean (Table 1). When corn followed corn in 2000 and 2001, the farmers applied a total of 130 kg N ha-1 in the low N treatment and 185 kg N ha-1 in the high N treatment.

At the dairy sites, the farmer followed his typical manure management practices. He surfaced-applied about 80000 L ha-1 of liquid manure in the fall followed by another 80000 L ha-1 of liquid manure in the spring, after which the fields were immediately plowed. The total N in the spring manure ranged from 2.9 to 3.2 g kg-1, depending on the year, with the NH4–N fraction ranging from 1.5 to 1.7 g kg-1 and the organic N fraction ranging from 1.2 to 1.5 g kg-1. We calculated that about 120 kg N ha-1 was available from the spring application but only 20 kg N ha-1 from the fall application because of volatilization, runoff, leaching, and denitrification losses in the fall, winter, and early spring (Cox and Cherney, 2002). Sidedress fertilizer N was not applied in 1999 at the dairy sites because the previous alfalfa crop provided a 55 kg N ha-1 N credit (Cornell Univ., 1999) and because the mid-June soil NO3–N concentrations averaged about 100 mg kg-1 in both fields. In 2000 and 2001, the previous legume credit from alfalfa was low (16 kg N ha-1 in 2000 and 0 in 2001), and the farmer sidedressed an additional 55 kg N ha-1 in the high N treatment at the Onondaga 1 and 2 sites. These two sites received about 140 to 150 kg N ha-1 in the low N treatment and about 200 to 210 kg N ha-1 in the high N treatment, 25 to 35 kg N ha-1 above or below the recommended N rate for the Honeoye–Lima soils at both sites (Cornell Univ., 1999).

The N treatments, which were six rows wide at each site, were randomized within the hybrids. Overall, there were a total of four strips (300 by 36 m) that contained the two hybrids and two N rates at the Onondaga 1 site, six strips (250–35 by 36 m) at the Onondaga 2 site, five strips (250 by 36 m) at the Seneca 1 site, two strips (800 by 36 m) at the Seneca 2 site and two strips (600 by 36 m) at the Seneca 3 site. The number of strips in each field was determined by the field dimensions.

The farmers harvested the corn in late October to mid-November in each year with six-row combines equipped with yield monitors that were connected to dGPS receivers. Yield measurements were taken every second by grain sensors while site coordinates were determined by the dGPS unit. The yield monitor was calibrated at the start of each field operation and intermittently throughout the harvest for both grain mass flow and grain moisture content. Weigh wagons, equipped with calibrated load cells, were used at each site to compare yields of each pass (one hybrid at one N level) in each strip with the average yield of the yield monitor. Yield differences between the two harvest methods were always <3%. Grain yield was determined at each station by calculating the mean value of the measurements recorded by the yield monitor along a 6-m length, which integrated three data points around each individual station. Yield maps were then developed for each site (ESRI, 2001).

We used a combined analysis of variance (ANOVA) model with years as random and sites as fixed variables to analyze soil NO3–N and yield data with General Linear Model (GLM) procedures of the SAS Statistical software package (SAS Inst., 1999). The Bartlett test (for unequal sample size), however, indicated nonhomogeneous variances for both soil NO3–N and yield data so the combined analyses were disregarded. We then analyzed each individual site–year comparison, and included strips as replications or random variables in the ANOVA model. The strips, however, were not significant at any of the sites and therefore removed from the model.

The soil NO3–N and yield data were then analyzed with stations (main plot) as a random variable, and hybrids (subplot) and N levels (subplot) as fixed variables in a split-split block experimental design. Station was tested against the station x hybrid plus station x N level minus station x hybrid x N level error term. We considered a significant station effect to be an indicator of significant spatial variability. Hybrid main effects were compared against the station x hybrid error term. A systematic plot layout, however, does not allow for accurate statistical analysis of hybrid main effects so the results were interpreted qualitatively. Nitrogen levels were tested against the station x N level error term, and all two-way interactions were tested against the three-way interaction error term. All effects were considered significant if P values were <=0.05. Means between N levels were separated by Fishers protected LSD (P = 0.05). Anhydrous ammonia averaged $0.33 kg-1 and the marketing year average price for corn was $0.086 kg from 1999 through 2001. Consequently, a yield increase >0.2 Mg ha-1 would be an economic response to the higher N rate at the cash crop sites. At the dairy site, however, where the grower does not routinely add sidedress N, a yield increase of >0.4 Mg ha-1 would be an economic response to the higher N rate because of the additional application costs ($17.50 ha-1) for anhydrous ammonia.

Yield data had homogeneous variances when combined across sites within individual years so yield data, averaged across sites will be presented. We used the regression (REG) procedure in SAS to determine linear and quadratic relationships between soil NO3–N concentrations in the low N treatment or averaged across both N treatments with corn yields for each site–year comparison. Regression coefficients were included in the equations if significant (P = 0.05). We also calculated simple correlations of individual station yields among the 3 yr to determine if there was a correlation for yields across years within each individual field.

Geostatistics were also used (ESRI, 2001) to analyze spatial variability and create spatial maps of soil NO3–N concentrations and yield differences [station x hybrid and station x N rate interaction LSD values (P = 0.05) were used if station x hybrid or station x N rate interactions existed at individual sites]. Sample variograms were fitted with spherical (best fit) variogram models using the following equation:

where {gamma}(h) is the spatial structure of the variogram, h is the distance between sampling locations, co is the nugget component of the variogram, c is the positive variance component, and a is the variogram range. The variogram range is the distance beyond which spatial correlation no longer exists. The nugget value represents unsampled variation or the random component of the variation. The ratio of the nugget to sill (co/co + c) indicates the degree of randomness in the spatial variability of the data.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSION
 REFERENCES
 
Weather conditions differed markedly among growing seasons (Table 2). For New York growing conditions, the 1999 growing season can be characterized as hot and dry, 2000 as cool and wet, and 2001 as warm and dry. Corn showed visible leaf wilting in most areas of the field at all sites during late June and early July in 1999 because of dry conditions. Also, in late June in 2000, corn showed leaf yellowing in some areas at all sites because of excessive precipitation. In 2001, corn showed some leaf wilting in August in some areas of the field at all sites because of the very warm conditions. At all sites, corn attained black layer formation in mid-September in 1999 and 2001 and in early October in 2000.


View this table:
[in this window]
[in a new window]
  		Table 2. Precipitation and growing degree days (GDD, 30–10°C) at five sites in New York during the 1999, 2000, and 2001 growing seasons.

 
Soil NO3–N concentrations in the upper 30-cm soil depth at the V4 stage of corn growth differed across years and sites. Soil NO3–N concentrations were exceedingly high at the Onondaga sites in the dry 1999 and 2001 growing seasons (Fig. 1) . Onondaga 1, a field that has received manure for decades, had average soil NO3–N concentrations above 100 mg kg-1 in 1999 and 2001. Although spatial variability existed at the Onondaga 1 site for soil NO3–N concentrations (54–152 mg kg-1 in 1999 and 68–156 mg kg-1 in 2001), all stations within the field had concentrations above 25 mg kg-1, the estimated threshhold concentration at which corn does not respond to additional sidedress N fertilizer under New York growing conditions (Klausner et al., 1993). Onondaga 2 had average soil NO3–N concentrations of 95 mg kg-1 in 1999 and 60 mg kg-1 in 2001 with all areas of the field above 25 mg kg-1. Based on soil NO3–N concentrations, corn at the Onondaga 1 and 2 sites was not expected to respond to sidedress fertilizer N in any areas of the field in 2001.



View larger version (72K):
[in this window]
[in a new window]
  		Fig. 1. Soil NO3–N concentrations in the upper 30-cm soil depth in late-spring of 1999, 2000, and 2001 using ArcGIS at the Onondaga 1 (a, b, c) and Onondaga 2 (d, e, f) sites.

 
Soil NO3–N concentrations are greater in warm and dry compared with cool and wet springs because warm temperatures promote more rapid N mineralization and dry conditions reduce NO3–N losses via leaching and/or denitrification (Magdoff, 1991). Consequently, soil NO3–N concentrations averaged much less at the Onondaga 1 and 2 sites in 2000 (Fig. 1). More importantly, about 15% of the field at the Onondaga 1 site and 25% of the field at the Onondaga 2 site had soil NO3–N concentrations <25 mg kg-1. Based on the soil NO3–N concentrations, corn was expected to respond to sidedress fertilizer N in the southeastern and central western regions at Onondaga 1 and the very southern and northwestern regions at the Onondaga 2 site in 2000. The nugget/sill ratios of the variograms for soil NO3–N, however, were high at both sites in 2000, which indicate a high degree of randomness in the spatial variability of the data (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. The nugget value (co), the nugget/sill fraction (co/co + c), and the range from the variogram models for soil NO3–N concentrations in the upper 30-cm depth at five sites in mid-June of 1999, 2000, and 2001 and for yield differences between N rates at the Seneca 3 site and between hybrids at the Onondaga 1 site in 1999 and 2001.

 
Soil NO3–N concentrations averaged much less at the cash crop sites compared with the dairy site in all 3 yr. The Seneca 1 and 2 sites had the least spatial variability, whereas the Seneca 3 site had the most spatial variability for soil NO3–N concentrations (Fig. 2) . In 1999, the entire Seneca 3 site had soil NO3–N concentrations >25 mg kg-1 except for the very southern region, where concentrations ranged from 12 to 24 mg kg-1. In 2000 and 2001, soil NO3–N concentrations at the Seneca 3 site ranged from 12 to 24 mg kg-1 in the northern and from 0 to 12 mg kg-1 in the southern regions. The ranges in the variograms were large (>350 m) in all 3 yr, which indicate that soil NO3–N concentrations were spatially correlated for long distances. Also, the nugget/sill ratios were low, which indicate that most of the data variability was within the sampling interval. Spatially distinct soil NO3–N regions with low random variability increases the probability that the Seneca 3 site will respond to variable N rate management.



View larger version (55K):
[in this window]
[in a new window]
  		Fig. 2. Soil NO3–N concentrations in the upper 30-cm soil depth in late-spring of 1999, 2000, and 2001 using ArcGIS at the Seneca 1 (a, b, c), Seneca 2 (d, e, f), and Seneca 3 (g, h, i) sites.

 
Corn yield had a significant station effect in 12 of 15 site–year comparisons with highly significant effects at all sites in 1999 and 2001 (Table 4). In 2000, only two of five sites had a significant station effect on corn yield, despite the excessive spring precipitation. Apparently, corn yields have more spatial variability in warm and dry compared with cool and wet growing conditions in New York (Fig. 3 and 4) . All sites have some type of subsurface drainage system, which evidently reduced the spatial variability of corn yields under the very wet spring conditions in 2000. Jaynes and Colvin (1997) also reported less spatial variability for corn yield in wet compared with dry years.


View this table:
[in this window]
[in a new window]
  		Table 4. Analyses of variances for corn grain yield for two hybrids at two N rates at five sites in New York in 1999, 2000, and 2001.

 


View larger version (87K):
[in this window]
[in a new window]
  		Fig. 3. Corn yields in 1999, 2000, and 2001 using ArcView at the Onondaga 1(a, b, c) and Onondaga 2 (d, e, f) sites.

 


View larger version (59K):
[in this window]
[in a new window]
  		Fig. 4. Corn yields in 1999, 2000, and 2001 using ArcView at the Seneca 1 (a, b, c), Seneca 2 (d, e, f) and Seneca 3 (g, h, i) sites.

 
All sites showed positive correlations for corn yields between the 1999 and 2001 growing seasons, which indicate temporal stability for the spatial relationship of corn yields at these sites in dry years (Table 5). Also, the Onondaga 1 and Seneca 3 sites had positive correlations between corn yields in 1999 and 2001 with yields in 2000, despite very different growing conditions. The temporal stability for the spatial relationship of corn yields at both sites increases the probability that they will respond to variable rate technology (Sawyer, 1994).


View this table:
[in this window]
[in a new window]
  		Table 5. Correlations between corn yields at individual stations within sites among the 1999, 2000, and 2001 growing seasons.

 
Corn yield had a significant N effect in the ANOVA model in only 5 of 13 site–year comparisons (Table 4). Corn yielded the same or greater at the lower compared with the higher N rate in five of six comparisons at the Seneca sites in 1999 and 2001 (Table 6). Furthermore, the 0.3 Mg ha-1 yield response at the Seneca 1 site in 2001 was only marginally economical. The results are consistent with the findings of Sogbedji et al. (2001), who suggested that N rates may be reduced about 35 kg N ha-1 below the recommended rate when dry conditions prevail in May and June in New York. Corn responded to the higher N rate in 2000 at only one of the three Seneca sites, in which case the response was economical. The results at the Seneca sites in 2000 are not consistent with the findings of Sogbedji et al. (2001), who suggested that N rates may be increased about 35 kg ha-1 above the recommended rate when wet conditions prevail in May and June. Overall, the response to N at the three cash crop sites was consistent across years, despite different corn yields, especially at the Seneca 2 site. This agrees with the findings of Vanotti and Bundy (1994), who reported that optimum N rates for corn are soil specific and should not be adjusted according to yield goal.


View this table:
[in this window]
[in a new window]
  		Table 6. Corn grain yield of two hybrids at a high (H) and low (L) N rate at five sites in New York during the 1999, 2000, and 2001 growing season.

 
Corn did not respond to additional fertilizer N at the Onondaga 2 site in 2000 and 2001 (Table 6). Corn had a positive response to additional fertilizer N at the Onondaga 1 site in 2000, but the 0.4 Mg ha-1 response was not economical because of the additional machinery costs for applying fertilizer N. Surprisingly, a hybrid x N rate interaction existed for corn yield at the Onondaga 1 site in 2001. The 0.7 Mg ha-1 yield response by 3752 at the Onondaga 1 site in 2001 was economical, despite soil NO3–N concentrations above 68 mg kg-1 throughout the field.

Corn had station x N rate interactions at only the Seneca 3 site in 1999 and 2001, which indicated that corn responded similarly to N rates in most areas of the other sites (Table 4). Corn did not show station x N rate interactions at the Onondaga sites in 2000, when 15 to 25% of the field had soil NO3–N concentrations <25 mg kg-1. Corn did show quadratic relationships between yield and soil NO3–N concentrations in the low N treatment (no sidedress N) at Onondaga 1 in 2000, but the predictability was low (Table 7). Again, the nugget/sill ratios were high at both Onondaga sites in 2000, which indicate that a large portion of the variability existed at scales less than the sampling interval used. The use of the PSNT, even with relatively dense sampling, was not accurate enough to identify specific areas within manured fields that would or would not respond to additional fertilizer N.


View this table:
[in this window]
[in a new window]
  		Table 7. Regression equations describing the relationship between corn yield, averaged across N treatments and in the low N treatment, and soil NO3–N concentrations in the upper 30-cm soil depth at the four-leaf stage of corn growth at five sites in New York in 1999, 2000, and 2001.

 
The station x N rate interactions at the Seneca 3 site in 1999 and 2001 indicate that variable N rate management may be possible within this field in dry years. Furthermore, the spatial relationship of corn yields in this field had very high temporal stability (r = 0.96) between the 1999 and 2001 growing seasons. Corn yields had significant linear relationships with soil NO3–N concentrations in all 3 yr (Table 7) because soil NO3–N and corn yields were mostly greater in the northern compared with the southern regions of the field. Surprisingly, corn did not respond to the higher N rate in the northwestern region of the field where yields exceeded 9.0 Mg ha-1, even when soil NO3–N concentrations ranged from 12 to 24 mg kg-1 in 2001 (Fig. 5) . In contrast, corn responded to the greater N rate on about 15% of the southern region of the field in 1999 and about 30% in 2001, where yields were <6.0 Mg ha-1. If the grower used a yield goal-based approach for the application of N fertilizer, sole reliance on yield map data would have misled the grower into applying too much fertilizer N in the northwestern region and too little fertilizer on about 25% of the southern region of the field. Likewise, if the grower relied on the PSNT, the entire field would have received the recommended N rate of 160 kg ha-1 instead of 129 kg ha-1 (Klausner et al., 1993).



View larger version (42K):
[in this window]
[in a new window]
  		Fig. 5. Corn yield differences, based on LSD (0.05) interaction values, in 1999 and 2001 using ArcGIS between N rates at Seneca 3 (a = 1.0 Mg ha-1 LSD, b = 0.9 Mg ha-1) and between hybrids at Onondaga 1 (c = 1.0 Mg ha-1, d = 1.4 ha-1) sites.

 
Hybrid had a large mean square in the ANOVA model for corn yield at most sites in all 3 yr (Table 4). When combined across sites within a year, corn yield had site x hybrid interactions in all 3 yr (Table 6). 37M81 generally had greater quantitative yields at the Onondaga sites, whereas 3752 generally had greater quantitative yields at the Seneca sites, especially in the dry 1999 and 2001 growing seasons. Weather conditions and management practices (except for the use of manure vs. fertilizer N) were mostly the same across sites in all 3 yr, so different soil conditions presumably contributed to the site x hybrid interactions. Corn yield, however, had station x hybrid interactions in only 4 of 15 site–year comparisons, which indicate that the hybrids had similar relative yield differences throughout most fields. Corn yields thus had significant site x hybrid interactions, presumably because of different soil conditions across sites, but few station x hybrid interactions, despite spatial variability for corn yield at all sites.

Corn yield did have station x hybrid interactions at the Onondaga 1 site in 1999 and 2001, which suggests that variable hybrid selection may be possible within this field. Also, the spatial relationship of corn yields had high temporal stability (r = 0.88) in dry years, with yields >9.0 Mg ha-1 in western and northeastern regions and yields <6.0 Mg ha-1 in southeastern regions in 1999 and 2001. In 1999, 3752 compared with 37M81 yielded the same or greater in the northeastern region, whereas 37M81 yielded greater or the same in the central portion of the field (Fig. 5). In 2001, 37M81 compared with 3752 yielded greater in the southern central region of the field and the same in the remaining areas of the field. Although there was a station x hybrid interaction in this field in the dry years, variable hybrid selection would not be the appropriate management practice, because 37M81 vs. 3752 yielded the same or greater in most areas of the field.


   CONCLUSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSION
REFERENCES
 
Corn yields had significant spatial variability with temporal stability at all sites in the dry years. Significant spatial variability with temporal stability provides an opportunity to manage the yield variability via variable sidedress N application in mid- to late June. Furthermore, growers could use the PSNT and/or previous yield map data as a guide for variable fertilizer N application. Unfortunately, on a dairy farm, the use of the PSNT was not precise enough to identify specific areas within manured fields that would respond to variable N rate management. Also, on a cash crop farm, yield goal-based N applications, based on yield map data, would have resulted in overfertilization of N in about 25% of the field where corn yields were greatest and underfertilization on about 15% of the field where corn yields were the least. Apparently, the successful adoption of variable N rate management in the northeastern USA requires more information than late spring soil NO3–N concentrations and/or yield map data from previous years.

Hybrid x site interactions existed for yield in all years with the relative performance of the hybrids markedly different between the dairy and cash crop sites. Split-planter studies can help farmers evaluate hybrids and facilitate their selection of the best hybrids for their farm. The two hybrids evaluated in this study, however, had limited spatial yield differences within individual fields, despite strong interactions with locations. Site-specific hybrid selection within fields thus showed limited potential for successful adoption. Adoption of site-specific hybrid selection within fields will be a challenge if hybrids that show interactions with sites do not show spatial yield difference variability within individual sites.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSION
 REFERENCES
 




This article has been cited by other articles:


Home page
 Agron. J.Home page
J. P. Schmidt, N. Hong, A. Dellinger, D. B. Beegle, and H. Lin
Hillslope Variability in Corn Response to Nitrogen Linked to In-Season Soil Moisture Redistribution
Agron. J., January 1, 2007; 99(1): 229 - 237.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
Y. Miao, D. J. Mulla, P. C. Robert, and J. A. Hernandez
Within-Field Variation in Corn Yield and Grain Quality Responses to Nitrogen Fertilization and Hybrid Selection
Agron. J., January 5, 2006; 98(1): 129 - 140.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
A. N. Kravchenko, G. P. Robertson, K. D. Thelen, and R. R. Harwood
Management, Topographical, and Weather Effects on Spatial Variability of Crop Grain Yields
Agron. J., March 1, 2005; 97(2): 514 - 523.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
D. Inman, R. Khosla, D. G. Westfall, and R. Reich
Nitrogen Uptake across Site Specific Management Zones in Irrigated Corn Production Systems
Agron. J., January 1, 2005; 97(1): 169 - 176.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
W. J. Cox, W. A. Knoblauch, H. M. van Es, T. W. Katsvairo, and M. A. Glos
Economics of Purchasing a Yield Monitor for Split-Planter Corn Hybrid Testing
Agron. J., September 1, 2004; 96(5): 1469 - 1474.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
T. W. Katsvairo, W. J. Cox, and H. M. Van Es
Spatial Growth and Nitrogen Uptake Variability of Corn at Two Nitrogen Levels
Agron. J., July 1, 2003; 95(4): 1000 - 1011.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (8)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Katsvairo, T. W.
Right arrow Articles by Glos, M.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Katsvairo, T. W.
Right arrow Articles by Glos, M.
Agricola
Right arrow Articles by Katsvairo, T. W.
Right arrow Articles by Glos, M.
Related Collections
Right arrow Maize
Right arrow Production Agriculture
Right arrow Site-Specific Analysis

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Crop Science Vadose Zone Journal
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome