Soil Science Society of America Journal 65:376-384 (2001)
© 2001 Soil Science Society of America
DIVISION S-4-SOIL FERTILITY & PLANT NUTRITION
Deep Banding Phosphorus and Potassium Fertilizers for Corn Managed with Ridge Tillage
Rogerio Borges and
Antonio P. Mallarino
Dep. of Agronomy, Iowa State Univ., Ames, IA 50011
Corresponding author (apmallar{at}iastate.edu)
 |
ABSTRACT
|
|---|
Broadcast fertilization leads to stratification of soil P and K in the ridge-till system, which may reduce fertilizer use efficiency. This study evaluated the response of corn (Zea mays L.) to broadcast or deep-band (15- to 20-cm depth) placements in 15 site–years. Fertilization rates were 0 to 56 kg P ha-1 and 0 to 132 kg K ha-1. Soil-test P (STP) and K (STK) were higher in the top 15-cm layer of the ridges. Phosphorus increased early plant growth (V5 stage) in five sites, early P uptake in nine sites, and grain yield in seven sites. Yield was increased by P when STP was <22 mg P kg-1 (Bray-1) in the top 15-cm layer of ridges or <18 mg P kg-1 in the top 15 cm of ridges and valleys. The P placements seldom differed (the deep-band P was better in one site). Potassium increased growth in 6 sites, K uptake in 14 sites, and grain yield in nine sites. The deep-band K increased yield over the broadcast K in four sites. The yield response to broadcast K across sites was not correlated with STK, but the response to deep-band K was negatively and linearly correlated with STK from various sampling positions. Corn responded to deep-band K in soils with above-optimum STK according to current soil-test interpretations. The results showed that both placements usually were similarly effective for P, and that deep banding often was superior for K.
Abbreviations: LSD, least significant difference • STK, soil-test K • STP, soil-test P
|
INTRODUCTION
|
|---|
THE RIDGE TILLAGE SYSTEM was rapidly adopted by Corn Belt producers during the 1980s. Producers were exploring various forms of conservation tillage to reduce costs, reduce soil erosion, reduce contamination of water supplies, and comply with soil-conservation requirements of government programs. The area in ridge-till corn in the North American Corn Belt, however, has remained constant or even decreased during the 1990s (CTIC, 1998). Although many reasons could explain this trend, a major one often mentioned by producers is the perception of serious problems with fertilizer management with ridge tillage. This is especially the case for K because K-deficiency symptoms are observed in corn leaves even when apparently soil test-based rates of broadcast K fertilizer are used.
Studies on fertilizer application for moldboard or chisel-plow tillage have shown that P fertilization seldom increases corn yield when soils test above 
20 mg P kg-1 when the Bray-1 extractant is used (Olson et al., 1962; deMooy et al., 1973; Rehm et al., 1981; Rehm, 1986; Walker and Raines, 1988; Mallarino et al., 1991a; Mallarino and Blackmer, 1992; Webb et al., 1992). Studies with K for moldboard or chisel-plow tillage showed that fertilization seldom increases corn yield when soils test above 130 mg K kg-1 when the ammonium acetate extractant is used (Hanway et al., 1962; deMooy et al., 1973; Rehm et al., 1981; Walker and Raines, 1988; Mallarino et al., 1991a, 1991b), although results are more variable than for P mostly due to differences in sample processing (i.e., use of field moist or dried samples). Levels above these values are considered high for corn in many Corn Belt states and no P or K fertilization is recommended. These interpretations may or may not apply to ridge tillage, however. Ridge tillage leads to accumulation of P and K in the ridges, mainly due to tillage operations and limited movement of these nutrients in soils (Karlen et al., 1991; Rehm, 1992). Vertical and lateral nutrient stratification may reduce crop early growth and yield. Limited research suggests that commonly used soil sampling techniques and the broadcast fertilizer application may not be appropriate for ridge tillage (MacKay et al., 1987; Rehm, 1992). Ridge tillage mixes soil, fertilizer, and residues more than no tillage. The common practice is to build ridges when corn is 15 to 30 cm tall (often between the V5 and V6 growth stages) and to leave the ridges undisturbed until planting time of the next season when the field is leveled by a planter device that opens the ridge and transfers soil to inter-row positions. Only the inter-row areas (i.e., inter-ridge areas or valleys) are cultivated for weed control. The ridge-building operation transfers soil mixed with fertilizers and residues from the inter-row area onto the seed rows several weeks after seedling emergence. Therefore, when P or K fertilizers are broadcast, the mixing takes place above the seeding depth and seedling roots grow in nonfertilized soil because broadcast fertilizer is positionally unavailable until the main root system develops.
Deep banding of P and K could increase fertilizer use efficiency, early plant growth, and grain yield. Work in Minnesota (Rehm, 1992) has shown marked corn-yield response to deep application of K. Furthermore, subsurface banding of P fertilizer has potential for reducing P losses with surface runoff and, consequently, for reducing contamination of surface water supplies. The objectives of this research were to study grain yield, early plant growth, and early P and K uptake of corn managed with ridge tillage as affected by broadcast and deep band fertilizer application methods.
|
MATERIALS AND METHODS
|
|---|
Fifteen short-term P and K response corn trials were established in farmers' fields in several regions of Iowa from 1995 through 1997. Corn was always planted after soybean [Glycine max (L.) Merr.]. Summarized information about management practices and soil characteristics are shown in Tables 1 and 2. Crop-management practices were those used by each farmer, except N, P, and K fertilization. Sites 4, 6, and 9 had histories of deep banding for P and K, whereas other sites had received only broadcast or in-the-furrow starter fertilization. Plots in each trial had a length of 18.3 m and the width varied from four to six rows. The row spacing was 96 cm except for Site 10, where it was 91 cm. The producers applied N fertilizer at or before planting (100–150 kg N ha-1, depending on the site and year) as liquid urea–ammonium nitrate solution or anhydrous ammonia injected into the shoulder of the old ridges or in the valleys. In addition, a rate of 50 kg N ha-1 (as urea) was broadcast by hand over all plots between planting and the ridge building operations.
Fourteen treatments consisted of four controls that received no P or K, the factorial combinations of two P rates and two placements, the factorial combinations of two K rates and two placements, and a P–K mixture applied with two placements. The fertilization rates were 14 or 56 kg P ha-1 and 33 or 132 kg K ha-1. The P–K mixture treatment consisted of 56 kg P ha-1 and 132 kg K ha-1. The high rates are similar to fertilization rates applied in alternate years by Iowa corn–soybean producers, and correspond to the estimated P and K removed in harvested grain. The low rates were chosen to apply fertilizer insufficient to maximize yield in low-testing soils, and correspond to one-half the estimated average yearly nutrient removal in grain. The placements were broadcast and deep-band applications. The broadcast fertilizer was spread onto the soil surface by hand. The bands were approximately 25 mm wide and were placed 15 to 20 cm below the surface of each ridge. In 1995, the coulter-knife combination used placed the fertilizer through a narrow vertical slit opened from the top and center of the ridge. In 1996 and 1997, the coulter-knife combination used placed the fertilizer through a slit opened in one shoulder of the ridge at a 45° angle and the band sometimes was as far as 7 cm off the center of the ridge. Both methods placed the band at 5 to 7.5 cm below the seeding depth. In two of the control treatments, the ridge was not disturbed before planting, whereas the other two received a coulter-knife pass without fertilizer application. For seven trials (Sites 1–7), the fertilizer treatments were applied in the fall of the previous year (in November or December), which is the most frequent practice for P and K fertilization in the Corn Belt. Six trials (Sites 8–13) evaluated residual effects of similar treatments that had been applied for the previous year's soybean crop (the soybean responses are not shown or discussed in this publication). Two trials conducted in 1997 (Sites 14 and 15) evaluated the residual effects of treatments applied in 1995 for corn grown at Sites 2 and 5 (a nonfertilized soybean crop was grown in 1996).
Soil samples were collected from all sites. At Sites 1 to 7, samples were collected from each replication before applying the treatments. At Sites 8 to 15, which evaluated residual effects of treatments applied for previous crops, soil samples were collected from individual plots. Composite soil samples (16 cores, 2-cm diam. each) were collected from the ridges from two depths (0–15 and 15–30 cm) and from the valleys (0- to 15-cm depth only) for separate analyses. The samples were dried at 40°C and crushed to pass a 2-mm sieve for chemical analyses. Soil-test P, STK, pH, and organic matter were determined for each trial following procedures recommended for the North Central Region (Brown, 1998). Soil-test K was analyzed by the ammonium acetate method, and STP by the Bray-1 method. An additional P analysis by the Olsen extractant for soil of Site 6 (which had pH 7.4 and a weak reaction to acid) showed no evidence of reduced P extraction by the Bray-1 method (data not shown). Table 2 shows mean STP and STK values and corresponding Iowa State University interpretation classes for corn (0- to 15-cm depth and low subsoil P and K) for the experimental area of the seven sites that evaluated fresh fertilization and for the nonfertilized plots of the eight sites that evaluated residual fertilization for previous crops. The ranges for five STP classes (all units are mg P kg-1) are 0 to 8 for the very low class (VL), 9 to 15 for the low class (L), 16 to 20 for the optimum class (O), 21 to 30 for the high class (H), and 31 or higher for the very high class (VH). The ranges for five STK classes (all units are mg K kg-1) are 0 to 60 for the very low class (VL), 61 to 90 for the low class (L), 91 to 130 for the optimum class (O), 131 to 170 for the high class (H), and higher than 170 for the very high class (VH).
The aboveground part of corn plants were sampled at the V5 to V6 growth stage (15- to 30-cm height to the center of the whorl). Ten plants were collected randomly from rows that would not be harvested in each plot. Plant samples were dried in a forced-air oven at 60°C, weighed, and ground. Total P and K concentrations were determined by digesting 0.25 g from each sample with concentrated H2SO4 and H2O2 (Digesdahl Analysis System, Hach, Boulder, CO). Phosphorus was measured by colorimetry (Murphy and Riley, 1962) and K was measured by flame photometry. Total P or K uptake per plant was calculated for all plots. Nutrient concentrations in plant tissue are not shown to reduce the length of the article and because the study of P and K uptake is more useful to achieve the objectives of the study (concentrations can be calculated from data presented). Corn ears were picked by hand (7.6 m of two center rows), the grain was separated, and yield was adjusted to 155 g kg-1 moisture. Harvest plant population was measured on two rows of each plot, but data are not shown because population was not statistically affected by the treatments at any site. Precipitation and temperature data (not shown) were obtained from the nearest weather station for each trial (3 to 15 km from the sites).
Completely randomized block designs with three replications were used for all trials. Analyses of variance were performed for data from each site and across all sites. The treatments (14) sums of squares was partitioned into (i) a comparison between the two absolute controls and the two empty coulter-knife controls, (ii) a P rate by placement complete factorial excluding the P–K mixture (three rates, two placements, and an interaction) in which the rate and interaction main effects were further partitioned into single degree of freedom orthogonal comparisons, (iii) a similar K rate by placement complete factorial, and (iv) a set of nonorthogonal comparisons (one for each placement) between the P–K mixture and each equivalent P or K rate applied separately. The residual sums of squares (with 26 degrees of freedom) was used for all F-tests. Correlation and regression analysis were used to study relationships between relative yield increases and soil-test values. Relative yield was defined for each trial as the mean yield of the nonfertilized plots multiplied by 100 and divided by the P or K (as appropriate) treatment mean or means that produced the statistically highest grain yield.
|
RESULTS AND DISCUSSION
|
|---|
Fertilization and Placement Effects on Grain Yield
The coulter-knife pass without applying fertilizer did not differ (P 
0.1) from the absolute control in any trial. This result indicates that any response to deep banding was not due to a physical effect of the coulter and knife pass. Although grain yield was significantly increased by P and/or K treatments at several sites, no significant (P 0.1) interactions were found in any trial when the yield of the P–K mixture was compared with equivalent P or K rates applied separately. Therefore, means of the two controls are reported in tables and the yields for the P–K mixture (broadcast or deep banded) are not shown in tables or discussed.
One or more P fertilization treatments increased yield in seven sites (Table 3). Only the means of the two P fertilization rates are reported because the yield difference between the two P rates and the P rate by placement interaction were significant (P 0.1) only in one site. The four P fertilization treatments increased grain yield significantly in three sites that evaluated fresh fertilizer application (Sites 3, 6, and 7) and in three sites that evaluated residual effects of fertilization for the previous soybean crop (Sites 8, 9, and 12). A significant P rate by placement interaction revealed a yield response only to the deep-banded high-P rate at Site 4. This was the only site in which the P placements differed. The yield responses in Sites 8, 9, 12, and 13 demonstrated a significant residual effect of P fertilization for the previous year's soybean crop. The lack of response in Sites 14 and 15 should not be surprising because these sites evaluated residual effects of fertilizers applied for crops grown 2 yr earlier.
Iowa data for other tillage systems show a high probability of response when STP is below optimum and a lower probability of response when STP is optimum (Mallarino et al., 1991b; Mallarino and Blackmer, 1992; Webb et al., 1992; Bordoli and Mallarino, 1998). If current soil-test interpretations for the average STP in the top 15-cm layer of ridges and valleys are considered, four of the seven responsive sites tested very low or low and three tested optimum (Table 2). Soil-test P in the 15-cm top layer of the ridges was always higher than for the 15- to 30-cm layer or 15-cm top layer of valleys (Table 2). If STP in the top 15 cm of ridges were considered, four responsive sites would test in the very low or low classes, two would test optimum, and one would test high but within 1 mg P kg-1 of the optimum class. Six of the sites would be reclassified as very low or low if the top 15-cm layers of valleys were considered. Previous research has suggested that the STP from the ridges is more appropriate for predicting responses to P than STP from the valleys due to lesser root growth in soil of the valleys, especially in the trafficked ones (Bauder et al., 1985; Kaspar et al., 1991). The results of this study showed, however, that using STP data from ridges would improve very little the prediction of corn response to P. Use of STP from the top layer of ridges failed to correctly predict crop response in six sites whereas use of the average STP from the top 15 cm of ridges and valleys failed in seven sites. A prediction was considered incorrect when there was no response but the test predicted one, and when there was a response but the test predicted none. These estimates of failure were similar by assuming a response in soils testing lower than high or only in soils testing lower than optimum. Study of relationships between relative yield response and STP also showed little difference between these two sampling strategies. The relationship was not significant (P 0.1) for any soil-sampling position when all sites were included in the calculations. When data for Site 10, a site with extremely high STP, were not considered, the response decreased with increasing STP in the top 15-cm layer of ridges (r = 0.45, P 0.1) or with increasing average STP in the 15-cm top layer of ridges and valleys (r = 0.46, P 0.1).
One or more K fertilization treatments increased grain yield significantly (P 0.1) in nine sites (Table 3). Only the means of the two K fertilization rates are reported because the yield difference between the two K rates was significant (P 0.1) only in three sites. All K treatments increased yields in Sites 1 and 9, although in both sites the yield response was larger for the deep-band placement (the interaction rate by placement was significant). Only the deep-band placement increased yield in Sites 3 and 7 (which evaluated fresh fertilizer applications) and in Sites 10, 11, 14, and 15 (which evaluated residual fertilization effects). Only the broadcast placement increased yield in Site 6, a result that obviously departs from the general trend and cannot be explained satisfactorily. The two deep-band K rates did not increase or decrease yield at this site but increased early K uptake (K uptake responses are discussed later), and visual observations suggest no negative effects of the deep band (for example, possible shallower-than-planned application and salt effects). The significant higher response to deep-band K in several sites plus small, nonsignificant yield differences in favor of this placement in other sites probably explain a statistically significant response to both K fertilizer and placement across all sites. There was a difference of 340 kg ha-1 in favor of the deep-band placement over the broadcast placement, even considering the odd result at Site 6.
The results for K are remarkable in several ways, and are in sharp contrast to results for P. Data in Fig. 1
show the contrasting results observed for means across the seven P-responsive sites and the nine K-responsive sites. First, neither of the two broadcast K rates increased yields significantly in six responsive sites (Sites 3, 7, 10, 11, 14, and 15), even though the high rate applied an amount of K considered adequate for growing two crops in low-testing soils. Second, the response to deep-band K occurred at five of the eight sites that evaluated residual fertilization effects, and only in three of them (Sites 11, 14, and 15) the high-K rate produced significantly higher yields than the low-K rate (which suggests an expected higher residuality of the high rate). This result suggests that producers can achieve the yield benefits of K deep banding with a one-time application of the fertilizer needed for two or even three crops. Most Corn Belt producers use a one-time application of the P and K needed for a 2-yr corn–soybean rotation.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 1. Interaction of fertilization rate by placement method for grain yield for responsive P trials (top graph) and K trials (bottom graph). Data are means across seven sites for P and nine sites for K
|
|
The response to K occurred in soils that tested optimum or higher in STK when data from 15-cm samples from ridge and valley positions were averaged (no site tested low). Soil-test K in the 15-cm top layer of the ridges was higher than for the valleys or the 15- to 30-cm depth of ridges (Table 2). Correlations of relative yield response to broadcast K and STK were not statistically significant for any sampling position (not shown). In contrast, yield responses to deep-band K and STK were significantly (P 0.1) correlated (the response decreased with increasing STK). The highest coefficient was for the average STK in the top 15-cm layer of both ridges and valleys (r = -0.63, P 0.01) and the lowest for the 15- to 30-cm layer of ridges (r = -0.48, P 0.07). Coefficients for other sampling positions ranged from -0.56 to -0.60 (all significant at P 0.05). Curvilinear regression models (exponential or segmented polynomials) did not improve significantly (P 0.1) the fit of a linear model (not shown). No correlation should be expected according to current STK interpretations for corn in Iowa and the Corn Belt. The observed significant linear correlations suggest that responses of ridge-till corn to deep-band K may occur up to STK values outside the range of values observed in our study. The sampling depth did not help in explaining the occurrence of responses. Although nine sites would be classified one interpretation class lower if the 15- to 30-cm layer of ridges were used, only one responsive site (Site 1) would be reclassified from high to optimum. Moreover, the only two sites that would be reclassified as low (Sites 12 and 14) showed no significant yield response. In contrast to results for no-till corn reported by Bordoli and Mallarino (1998), responses to K placement in this study were not related with nearby rainfall in any period of the growing season.
Fertilization and Placement Effects on Residual Soil Test Values
Plots that received the P treatments were not sampled after the corn harvest. Individual plots were sampled before planting corn, however, in sites that evaluated residual effects of fertilization for the previous year soybean crop (Sites 8 to 13) and for corn planted 2 yr before (Sites 14 and 15). These samples provided estimates of the residual effect of the treatments on STP and STK values after 1 or 2 yr of application (Table 4). Data for the low P and K rates are not shown because (as was expected) these rates never increased (P 0.1) the residual STP or STK over that of the control plots. The STP of plots that received the high P rate was influenced by the placement and the position from where the soil samples were collected. Both P placements increased (P 0.1) STP in the top 15-cm soil layer of the ridges (compared with the control) in about one-half of the sites. However, the increase was more consistent and greater for the deep-band placement. The broadcast placement never increased STP in the 15- to 30-cm layer of the ridges, and the deep-band placement increased it only in one site. The broadcast placement increased STP in the top 15-cm layer of the valleys in four sites, but the deep-band placement increased it only in two sites. Although sampling error and differential P removal from different soil layers could confound the results, the residual STP values suggest that the actual banding depth was above a 15-cm depth in most sites (except Site 9); at least part of the band was above the tilled depth in some sites (because STP in the valleys was increased) and was below the tilled depth in other sites (because deep banding increased STP only in the top 15-cm layer of the ridges).
View this table:
[in this window]
[in a new window]
|
Table 4. Initial soil P and K in plots of eight trials that evaluated the residual effects of treatments applied 1 or 2 yr before the current corn crop
|
|
The high broadcast K rate never increased (P 0.1) STK in the top 15-cm layer of the ridges, but the high deep-banded K rate increased it in one-half of the sites (Table 4). This result coincides with results for STP, although the failure of the broadcast placement in increasing the nutrient concentration in the ridges was more marked for STK than for STP. This result for STK coincides remarkably well with the observed lack of yield response to broadcast K in most sites. Both placements seldom increased STK in the 15- to 30-cm layer of the ridges (a similar result was observed for STP). Both placements seldom increased STK in the top 15-cm soil layer of valleys, which is in contrast to results for STP because the broadcast placement increased STP in the valleys of one-half of the sites. The differences between results for P and K cannot be explained with certainty. Soil-test data for both nutrients (but especially for STK) show high sampling error (a known problem with conservation tillage, especially when banded fertilization is used), and differences between P and K uptake and recycling with crop residues likely were involved.
Fertilization and Placement Effects on Early Corn Growth
The treatments influenced early growth in several sites but there were no significant differences (P 0.1) between fertilization rates or between the absolute and empty-knife pass without applying fertilizer in any trial. Also, there were no significant interactions between nutrients or between rates and placements in any trial. Thus, means of the two fertilization rates and of all control treatments are shown in the tables and discussed.
One or more P treatments increased (P 0.1) early corn growth (as measured by plant dry weight) in five sites (Table 5). Four of the responsive sites evaluated fresh fertilizer application (Sites 3, 4, 6, and 7), and one (Site 14) evaluated residual fertilizer effects for previous crops. Except for Site 14, the soils tested optimum or lower when STP from the top 15-cm layer of ridges and valleys was averaged and there was also a grain yield response. In Site 14, STP of nonfertilized plots was high, and there was no grain yield response. The P placement influenced growth only at Site 7, where growth was greater for the deep-band placement. The broadcast P placement apparently increased growth in Site 9 (which evaluated residual fertilization effects), but this apparent response should not be considered. In this site, the difference between the broadcast placement and the control was very small, and the statistically significant placement effect in favor of the broadcast placement arises from an unreasonably low growth for the deep-band placement. It is remarkable that, independently of statistical significance, early growth was greater for the deep-band P placement at 10 sites.
Potassium fertilization increased (P 0.1) early corn growth in six sites (Table 5). Three sites evaluated fresh fertilizer application (Sites 1, 2, and 7) and three sites evaluated residual effects of previous applications (Sites 9, 10, and 12). The placements did not differ in Sites 1 and 2, and the deep-band K was better than the broadcast K in the other four sites. The deep-band K placement apparently increased growth in Site 5, but this apparent response probably should not be considered. In this site, the difference between the deep-band placement and the control was small, and the statistically significant placement effect likely arose from an unreasonably low growth for the broadcast placement (lower than for the control). These results suggest that K fertilization can increase the growth rate of ridge-till corn under some conditions difficult to predict. Soils at all responsive sites tested high or higher when STK from the top 15-cm layer of ridges and valleys was averaged, and there was also a grain yield response in Sites 1, 7, 9, and 10, but not in Sites 2 and 12. Trends in favor of the deep-band placement at several sites probably explain a significant placement difference across all sites.
Growth responses to P or K fertilization and placement were not correlated with STP or STK from any sampling position (not shown). Sites with early growth response to P or K seldom coincided with sites with grain yield response. Responses in both yield and early growth coincided only in threes sites for P and in two sites for K. This lack of agreement could be explained by changes in soil K availability and growing conditions during the growing season. The early growth responses to P or K fertilization and placement observed in this study are in contrast with observations made by Mallarino et al. (1999) for no-till corn in Iowa. Deep-banded and broadcast P usually increased early growth of no-till corn, even in soils testing high in P, and K seldom increased early growth. The difference in early growth response to P could be explained by warmer soil temperatures in the ridges than in no-till, which are not covered by crop residue. Warmer spring temperatures would increase soil P availability, root growth, and shoot growth during the early stages of the crop. The different responses for K cannot be explained satisfactorily with the methods used, although the answer may be in contrasting root growth patterns and water relations between the two tillage systems.
Fertilization and Placement Effects on Early Plant Nutrient Uptake
The fertilizer rates and placements had greater effects on early P and K uptake than on grain yield or early growth. Furthermore, the fertilizer placement influenced the P and K uptake differently in several sites. Phosphorus fertilization increased (P 0.1) early P uptake at nine sites (although increasing trends were obvious in most sites) and the high P rate increased uptake more than the low rate at four sites (Table 6). The P placement influenced the P uptake in Sites 6, 7, and 12, where the deep-band placement was better. The placement method influenced the difference in P uptake between the low and high P rates (P 0.1) only at Site 7 (i.e., the interaction placement by rate was significant). In this site, and for reasons difficult to explain, the high broadcast P rate did not increase P uptake. A statistical analysis across all sites revealed significant differences due to P rate and placement (the deep-band placement was better) and no significant interaction of P rate by placement. This overall advantage of the deep-band P placement is explained by a significant advantage in three sites (6, 7, and 12) and consistent trends in many other sites. Such an advantage of the deep-band placement was similar for the two P rates used and was not related to placement effects on grain yield. The deep-band P produced higher grain yield than the broadcast P only in Sites 4 and 13, but in these sites the P uptake was not significantly affected by the placement. On the other hand, the two sites in which the deep-band P produced higher early growth (Sites 6 and 9) also showed a P-uptake response to deep-band P.
The effects of K fertilization and placement on early plant-K uptake were more marked than for P (Table 7). Some or all K treatments increased early K uptake (P 0.1) in all sites except one, and the highest K rate (averaged over placements) increased K uptake over the low rate in eight sites. The deep-band K increased K uptake (P 0.1) more than the broadcast placement in 11 sites. Moreover, the broadcast placement increased K uptake over the control only at six of the responsive sites (Sites 5, 6, 8, 9, 12, and 15), and in four of these (Sites 5, 8, 9, and 15) only when the highest rate was applied. The statistical analysis across all sites revealed significant differences due to K rate, placement, and a placement by rate interaction. Figure 2
describes the interaction fertilizer rate by placement across all sites and also shows the marked difference between results for P and K. The broadcast placement was much less efficient in increasing corn K uptake than P uptake. This result was also observed by Mallarino et al. (1999) for early nutrient uptake by no-till corn.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 2. Interaction of fertilization rate by placement method for early plant P uptake (top graph) and K uptake (bottom graph). Data are means for nine sites with a P uptake response and 14 sites with a K uptake response
|
|
|
SUMMARY AND CONCLUSIONS
|
|---|
Phosphorus fertilization increased corn early growth and grain yield in several low-testing soils, and the deep-band placement was better than the broadcast placement only at one site. The deep-band placement tended to increase P uptake more than the broadcast placement in most sites, but the differences were statistically significant only in two sites that evaluated fresh fertilizer applications and in one site that evaluated residual effects of fertilization for a previous crop. Potassium fertilization increased yield in several sites and, in contrast to results for P, the deep-band K placement produced higher yields than the broadcast K in several sites. Moreover, in four responsive sites the broadcast K did not increase yield over the control. The greater efficiency of the deep-band K placement was even more marked for early K uptake. Occasional responses of early plant growth to P or K fertilization and to deep banding were poorly related to grain yield.
The results showed that current soil-test P interpretations and fertilizer recommendations for corn, which are based on chisel-plow tillage, can be used for ridge tillage when equal numbers of 15-cm deep samples are collected from ridges and valleys. Collecting samples only from ridges probably is equally appropriate or only slightly better in predicting crop response to P. The grain yield increases due to P fertilization occurred in some of the sites where STP was <22 mg P kg-1 in the top 15-cm layer of the ridges or <18 mg P kg-1 in the top 15 cm of ridges and valleys. In contrast to results for P, results for K showed that current soil-test interpretations may not apply to ridge-till corn. Yield responses to deep-band K were observed in several sites with STK levels that would result in no K fertilizer recommendation according to current interpretations. The yield response to deep-band K tended to decrease linearly with increasing STK measured at various depths and positions relative to the ridges, but no critical STK concentration could be identified.
The overall conclusion of this study is that deep-band K often will markedly increase the production efficiency of ridge-till corn compared with broadcast fertilization. The magnitudes of the responses showed that deep banding of K will be more cost-effective than broadcast K in many situations, especially because the results showed that producers can achieve the yield benefits of K deep banding by applying once the fertilizer rate needed for two crops. Because producers usually apply P and K together, the increased yield response over the broadcast placement will likely be due to the K in the mixture but deep-banded P may occasionally increase both early growth and grain yield.
|
NOTES
|
|---|
Iowa Agric. Home Econ. Exp. Stn. Journal Paper no. J-18523. Project 3233. Research supported in part by the Leopold Center for Sustainable Agriculture.
Received for publication January 18, 2000.
|
REFERENCES
|
|---|
- Bauder, J.W., G.W. Randall, and R.T. Schuler. 1985. Effects of tillage with controlled wheel traffic on soil properties and root growth of corn. J. Soil Water Conserv. 40:382–385.
- Bordoli, J.M., and A.P. Mallarino. 1998. Deep and shallow banding of phosphorus and potassium as alternatives to broadcast fertilization for no-till corn. Agron. J. 90:27–33.[Abstract/Free Full Text]
- Brown, J.R. 1998. Recommended chemical soil tests procedures for the North Central region. North Central Regional Publ. 221 (Rev.). Missouri Agric. Exp. Stn., Columbia, MO.
- Conservation Technology Information Center. 1998. National crop residue management survey. CTIC. West Lafayette, IN.
- deMooy, C.J., J.L. Young, and J.D. Kaap. 1973. Comparative response of soybeans and corn to phosphorus and potassium. Agron. J. 65:851–855.[Abstract/Free Full Text]
- Hanway, J.J., S.A. Barber, R.H. Bray, A.C. Caldwell, M. Fried, L.T. Kurtz, K. Lawton, J.T. Pesek, K. Pretty, M. Reed, and F.W. Smith. 1962. North central regional potassium studies. III. Field studies with corn. Iowa Agric. Home Econ. Exp. Stn. Res. Bull. 503. Iowa Agric. Home Econ. Exp. Stn., Ames, IA.
- Karlen, D.L., E.C. Berry, T.S. Colvin, and R.S. Kanwar. 1991. Twelve-year tillage and crop rotation effects on yields and soil chemical properties in northeast Iowa. Commun. Soil Sci. Pl. Anal. 22: 1985–2003.
- Kaspar, T.C., H.J. Brown, and E.M. Kassmeyer. 1991. Corn root distribution as affected by tillage, wheel traffic, and fertilizer placement. Soil Sci. Soc. Am. J. 55:1390–1394.[Abstract/Free Full Text]
- MacKay, A.D., E.J. Kladivco, S.A. Barber, and D.R. Griffith. 1987. Phosphorus and potassium uptake by corn in conservation tillage systems. Soil Sci. Soc. Am. J. 51:970–974.[Abstract/Free Full Text]
- Mallarino, A.P., and A.M. Blackmer. 1992. Comparison of methods for determining critical concentrations of soil test phosphorus for corn. Agron. J. 84:850–856.[Abstract/Free Full Text]
- Mallarino, A.P., J.M. Bordoli, and R. Borges. 1999. Phosphorus and potassium placement effects on early growth and nutrient uptake of no-till corn and relationships with grain yield. Agron. J. 91:37–45.[Abstract/Free Full Text]
- Mallarino, A.P., J.R. Webb, and A.M. Blackmer. 1991a. Corn and soybean yields during 11 years of phosphorus and potassium fertilization on a high-testing soil. J. Prod. Agric. 4:312–317.
- Mallarino, A.P., J.R. Webb, and A.M. Blackmer. 1991b. Soil test values and grain yields during 14 years of potassium fertilization of corn and soybean. J. Prod. Agric. 4:560–566.
- Murphy, J., and J.P. Riley. 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chem. Acta. 27:31–36.
- Olson, R.A., A.F. Dreier, C.A. Hoover, and H.F. Rhoads. 1962. Factors responsible for poor response of corn and grain sorghum to phosphorus fertilization: I. Soil phosphorus level and climatic factors. Soil Sci. Soc. Am. J. 26:571–574.[Abstract/Free Full Text]
- Rehm, G.W. 1986. Effect of phosphorus placement on early growth, yield and phosphorus absorption by irrigated corn. J. Fert. Issues 3:12–17.
- Rehm, G.W. 1992. Response of ridge-till corn to potash fertilization. p. 139–143. In Proc. 22nd North Central Extension-Industry Soil Fertility Conf., St. Louis, MO. 18–19 Nov. 1992. Minnesota Ext. Serv., Univ. of Minnesota, St. Paul, MN.
- Rehm, G.W., R.C. Sorensen, and R.A. Wiese. 1981. Application of phosphorus, potassium, and zinc to corn grown for grain or silage: Early growth and yield. Soil Sci. Soc. Am. J. 45:523–528.[Abstract/Free Full Text]
- Walker, W.M., and G.A. Raines. 1988. Effect of corn cultivar, phosphorus and potassium on yield and chemical composition. J. Plant Nutr. 11:1715–1726.
- Webb, J.R., A.P. Mallarino, and A.M. Blackmer. 1992. Effects of residual and annually applied phosphorus on soil test values and yields of corn and soybean. J. Prod. Agric. 5:148–152.
This article has been cited by other articles:
|
|
|
|
 
A. P. Mallarino and R. Borges
Phosphorus and Potassium Distribution in Soil Following Long-Term Deep-Band Fertilization in Different Tillage Systems
Soil Sci. Soc. Am. J.,
February 27, 2006;
70(2):
702 - 707.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. Borges and A. P. Mallarino
Broadcast and Deep-Band Placement of Phosphorus and Potassium for Soybean Managed with Ridge Tillage
Soil Sci. Soc. Am. J.,
November 1, 2003;
67(6):
1920 - 1927.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
A. P. Mallarino
Field Calibration for Corn of the Mehlich-3 Soil Phosphorus Test with Colorimetric and Inductively Coupled Plasma Emission Spectroscopy Determination Methods
Soil Sci. Soc. Am. J.,
November 1, 2003;
67(6):
1928 - 1934.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
