SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 2 June 2005
Published in Soil Sci Soc Am J 69:1118-1128 (2005)
DOI: 10.2136/sssaj2004.0279
© 2005 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
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 (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Dodd, J. R.
Right arrow Articles by Mallarino, A. P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Dodd, J. R.
Right arrow Articles by Mallarino, A. P.
Agricola
Right arrow Articles by Dodd, J. R.
Right arrow Articles by Mallarino, A. P.
Related Collections
Right arrow Nutrient Management
Right arrow Soil Fertility and Productivity

Nutrient Management & Soil & Plant Analysis

Soil-Test Phosphorus and Crop Grain Yield Responses to Long-Term Phosphorus Fertilization for Corn-Soybean Rotations

Jason R. Dodda,b and Antonio P. Mallarinoa,*

a Dep. of Agronomy, Iowa State Univ., Ames, IA 50011
b Formerly a graduate student, now at Pioneer Hi-Bred, Inc., 910 6th Street, Colo, IA 50056

* Corresponding author (apmallar{at}iastate.edu)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Farmers and nutrient management regulatory agencies are requesting better knowledge of P fertilization impacts on soil-test P (STP) and crop yield. This study evaluated STP and grain yield of corn (Zea mays L.) and soybean [Glycine max (L.) Merr.] as affected by long-term P fertilization in three trials evaluated from the 1970s until 2002 near Boone, Kanawha, and Nashua in central, northern, and northeast Iowa. Soils were Aquic Hapludolls and Typic Endoaquolls at Boone, Typic Endoaquolls at Kanawha, and Typic Hapludolls at Nashua. At Boone and Kanawha, treatments were the combinations of three initial STP levels (17–96 mg P kg–1, Bray-P1) and four annual rates (0–33 kg P ha–1). At Nashua, initial STP was 28 mg P kg–1 and treatments were 0, 22, and 44 kg P ha–1. Ten to twenty years of cropping were needed on soils testing 43 to 96 mg P kg–1 to observe yield response to P. Annual P rates that maintained near optimum STP (16–20 mg P kg–1) were 17, 14, and 13 kg P ha–1 yr–1 at Boone, Kanawha, and Nashua, respectively. Phosphorus required to increase STP 1 mg P kg–1 yr–1 were 23, 28, and 17 kg P ha–1 yr–1 at Boone, Kanawha, and Nashua, respectively. Critical STP concentrations (CC) identified across sites and years were 15 to 21 mg P kg–1 for corn and 12 to 18 mg P kg–1 for soybean. Observed grain yield and STP responses are useful to develop effective P management plans for corn–soybean rotations under approximately similar conditions to those in this study.

Abbreviations: CC, critical soil-test P concentration • STP, soil-test P


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
PHOSPHORUS is an essential nutrient for corn and soybean production, and management decisions concerning P fertilization affect crop yield and grower profitability. Many farmers of the U.S. Corn Belt have been applying P fertilizer during the last 50 yr to soils originally testing below STP requirements of corn and soybean. In a continued effort to increase profitability, however, many farmers have over-applied P resulting in very high STP levels in many fields. Soil-test summaries for samples submitted to Iowa soil testing laboratories show that the fraction of soils testing optimum or higher in P according to Iowa interpretations (Sawyer et al., 2002) increased from approximately one-fourth of the samples in the 1960s to more than one-half of the samples in recent years (Potash and Phosphate Institute, 2001; D. Wetterauer and A.P. Mallarino, Iowa State University Soil Testing Laboratory, unpublished data, 2003). Research has demonstrated corn and soybean yield response to P for various STP levels and has calibrated various soil P tests (Cope, 1981; McCallister et al., 1987; Mallarino et al., 1991; Webb et al., 1992; Cox, 1992, 1996; Mallarino, 1997). Based on this type of research, states have developed P recommendations, such as those used in Iowa (Sawyer et al., 2002).

Increasing public and government awareness of eutrophication of surface water resources as a result of excessive P movement off fields is resulting in increased demands for more efficient P management and less tolerance for P application to high-testing soils. Previous Iowa research from two long-term sites (Mallarino et al., 1991; Webb et al., 1992) indicated that 8 to 9 yr of corn and soybean production without fertilization in high-testing soils (30–40 mg P kg–1, Bray-P1) were needed to observe statistically significant yield responses. Webb et al. (1992) reported that 16 kg P ha–1 yr–1 maintained initially optimum STP (18 mg P kg–1). Southern Minnesota research (Randall et al., 1997) showed that STP could be maintained at 18 to 20 mg kg–1 (Bray-P1) with 19 to 24 kg P ha–1 yr–1 for fields managed with a corn–soybean rotation. Researchers have found that more P fertilizer is needed to maintain high STP values than lower values (McCallister et al., 1987; McCollum, 1991; Webb et al., 1992). In Iowa, for example (Webb et al., 1992), 33 kg P ha–1 yr–1 were required to maintain 75 mg P kg–1 (Bray-P1) while 16 kg P ha–1 yr–1 were required to maintain 18 mg P kg–1 STP. Research on soils along the Atlantic seaboard has also demonstrated differences in the P rate required to maintain different STP values. For example, in a field study that included 24 sites, McCollum (1991) showed that annual P additions similar to average crop P removal with harvest nearly maintained STP when the initial value was near the CC for those soils (24 mg P kg–1, Mehlich-1 test) but not when the initial value was higher. Larger P requirements for maintaining higher STP values have been explained mainly by increased P removal (either through increases in yield and/or P concentration of harvested products) and conversion of plant-available P measured by soil tests to nonextractable forms via chemical reactions with soil constituents (McCollum, 1991; Webb et al., 1992).

Critical STP concentrations indicate values above which fertilization no longer results in yield responses or economic benefits. Reported CC values or ranges vary not only with the soil-test method, soil sampling depth, or region but also with the model fit to relationships between yield response and STP (Dahnke and Olson, 1990; Mallarino and Blackmer, 1992). Previous Iowa soil-test field calibration research has estimated CC for corn and soybean (Mallarino and Blackmer, 1992; Webb et al., 1992; Mallarino, 1997) at 12 to 20 mg P kg–1 by the Bray-P1 or Mehlich-3 tests. In eastern regions of the USA, research (Beegle and Oravec, 1990; Cox, 1992 and 1996) has identified Mehlich-3 CC for corn and soybean ranging from 18 to 41 mg P kg–1 depending on model used to identify the CC and the year.

Information about STP changes in long-term corn–soybean production with various P fertilization strategies can aid growers in managing nutrients more efficiently. Moreover, nutrient management plans being mandated by many states require estimates of the impact of P management practices on increasing or decreasing STP. Long-term trials provide data essential to study soil P trends over time as affected by cropping and P fertilization. Knowing how STP declines or increases when P rates smaller or larger than average crop removal are applied should aid in managing P fertilizer to maintain optimum STP over time and to determine how long a field will need to be in production without fertilization before STP decreases to levels that require fertilization. This research involved three Iowa experiments spanning 23 to 28 yr and objectives were to (i) study STP trends over time for different initial STP levels and annual P application rates, (ii) study corn and soybean yield as affected by P fertilization and STP, and (iii) determine CC for both crops.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soil-test P and crop grain yield data were collected from plots of three long-term field trials with corn–soybean rotations established in Iowa during the 1970s and evaluated until 2002. Two similar experiments were established at the Agronomy and Agricultural Engineering Research Center near Boone in 1975 and at the Northern Iowa Research Farm near Kanawha in 1976. Another experiment was established at the Northeast Iowa Research Farm near Nashua in 1979. Early results from Kanawha (1976–1989) were summarized by Webb et al. (1992) and early results from Nashua (1979–1989) were summarized by Mallarino et al. (1991). No data from Boone have been published. Soils were a Nicollet (fine-loamy, mixed, superactive, mesic Aquic Hapludoll)–Webster (fine-loamy, mixed, superactive, mesic Typic Endoaquoll) complex at Boone, a Webster–Canisteo (fine-loamy, mixed, superactive, calcareous, mesic Typic Endoaquoll) complex at Kanawha, and Kenyon (fine-loamy, mixed, superactive, mesic Typic Hapludoll) at Nashua. Corn and soybean were grown in rotation at the three sites by dividing the experimental areas into two halves, growing each crop each year by switching field half each year, and by applying identical treatments using identical experimental designs on each field half. The corn hybrids, soybean cultivars, herbicides, planting dates, and seeding rates were among those recommended for each region over time (not shown). Nitrogen was applied for corn at rates that varied from 168 to 201 kg N ha–1 over time. Potassium was applied as needed to maintain a soil-test K level > 150 mg K kg–1 (ammonium-acetate test, 15-cm soil depth). Plots with corn residues were chisel-plowed in the fall (October or early November) and disked in spring before planting soybean while plots with soybean residues only were disked in the spring. These tillage practices are used by the vast majority of Iowa farmers for corn–soybean rotations. The crop row width was 76 cm for both crops in all experiments.

Trials at Boone and Kanawha were based on randomized complete-block, split-plot designs with three (Boone) or four (Kanawha) replications. Large plots measured 12 by 24 m and subplots measured 6 by 12 m. Treatments applied to large plots were initial P rates applied once to create three contrasting initial STP levels and treatments applied to subplots were four annual P rates. The initial P rates were 0, 145, or 291 kg P ha–1 (as triple superphosphate), which were applied before tillage and planting operations for the 1974 crops at Boone and the 1975 crops at Kanawha (crops were harvested but no yield was measured). These treatments are hereon referred to as IP1, IP2, and IP3, respectively. The annual P rates were 0, 11, 22, and 33 kg P ha–1 (triple superphosphate) applied in fall (October or early November) before any tillage and before soils froze. Annual P rates were first applied before the 1975 crop at Boone and the 1976 crop at Kanawha. At Kanawha, no annual P rate was applied for the 1997 crops, and beginning with the 1998 crops the two lower annual rates (11 and 22 kg P ha–1) were discontinued. At Boone, the annual P rates were discontinued beginning with the 1997 crops except for the highest annual rate (33 kg P ha–1) for the IP3 initial treatment.

At Nashua, treatments were the factorial combinations of six P and K annual rates, which were arranged as a randomized complete-block design with three replications. Nutrient rates were 0, 22, and 44 kg P ha–1 (triple superphosphate) and 0, 67, and 134 kg K ha–1 (potassium chloride). Fertilizer treatments were applied in fall (October or November) before tillage to plots 4.5 by 15 m in size. Only average results from plots that received the two high K rates are shown and discussed. Preliminary analyses of variance indicated no significant (P ≤ 0.05) yield response to K until 1998.

Soil samples were collected in fall of each year (October or early November) after crop harvest and before applying the annual P treatments. At Boone, all plots were sampled until the end of the study. At Kanawha, all plots were sampled until fall 1988 (before the 1989 crop), only plots with soybean residue were sampled from fall 1989 to fall 1995, and all plots were sampled since fall 1996 with the exception of plots of the highest annual P treatment in fall 1997 and 1998. At Nashua, all plots were sampled the first 2 yr, only plots with soybean residue were sampled from fall 1981 until the end of the study, and all control plots (no annual P application) with corn residue were sampled since 1997. Each sample consisted of 10 to 12 cores (1.9-cm diam.) taken from a 0- to 15-cm soil depth. Plant-available soil P was determined in duplicate with the Bray-P1 method following procedures recommended for the North Central Region (Frank et al., 1998). Soil organic matter and texture were measured in samples taken in fall 1996 from plots that received no annual P fertilizer. Soil texture was analyzed with the method of Walter et al. (1978) and was loam (257 g kg–1 clay) at Boone, clay loam (296 g kg–1 clay) at Kanawha, and loam (217 g kg–1 clay) at Nashua. Soil organic matter was analyzed with the method described by Wang and Anderson (1998), and was 49 g kg–1 at Boone, 57 g kg–1 at Kanawha, and 33 g kg–1 at Nashua. Soil pH was determined with a soil/water ratio of 1:1 (v/v) before applying the first annual P treatments and occasionally over time to assess lime needs. At Boone, mean initial pH was 6.4, decreased to 6.2 by 1989; limestone (6.7 Mg ha–1 of CaCO3) was applied to all plots in 1990, and remained 6.8 to 7.0. At Kanawha, mean initial pH was 7.4 and remained 6.6 to 7.5. At Nashua mean initial pH was 5.8, limestone (7.8 Mg ha–1 of CaCO3) was applied in 1981, and remained 6.7 to 7.0.

Analyses of variance for treatment effects on crop yield and STP were conducted for data from each year according to the experimental designs with the General Linear Models (GLM) procedure of SAS (SAS Inst., 2000). Orthogonal contrasts were used to compare means for the annual P treatments (within each initial P treatment at Boone and Kanawha). Trends of STP over time for each site and treatment combination were described by a linear model and an exponential model asymptotic to a maximum or a minimum using the Nonlinear Models (NLIN) procedure of SAS. The exponential model was the Mitscherlich equation expressed in the form suggested by Nelson and Anderson (1977), and is shown only when its residual sums of squares were significantly smaller (P ≤ 0.05) than for the linear model. The models were fit to STP means across replications and field halves for each site, initial P treatment (at Boone and Kanawha), and annual P treatment. Because no annual P rate exactly maintained STP, the gross maintenance P rate for each site (and initial P treatment at Boone and Kanawha) was calculated by interpolating from linear coefficients of equations fit to relationships between STP and time for the two annual P rates that encompassed the unknown maintenance rate (with STP increasing in one equation and decreasing in the other). The calculation is shown in Eq. [1], where Pdec is the P rate (kg P ha–1 yr–1) resulting in decreasing STP, Pinc is the P rate resulting in increasing STP, bdec is the linear coefficient of the equation for decreasing STP, and binc is the linear coefficient of the equation for increasing STP.


[1]

Critical STP concentrations were determined with linear-plateau (Waugh et al., 1973), quadratic-plateau (SAS Institute, 2000), and Mitscherlich (expressed by Nelson and Anderson, 1977) models fit to relationships between STP and relative yield using the NLIN procedure of SAS. Soil-test P values used were from samples collected before planting each crop in plots that received no annual P over time and plots for which P application was discontinued in recent years. Therefore, fewer data were available for soybean at Nashua and Kanawha sites because plots to be planted with soybean (corn residue) were not sampled many years. Relative yield for a crop, site, year, and initial P treatment (at Boone and Kanawha) was defined as the mean yield of plots that received no P expressed as a percentage of the mean yield of plots that received the highest annual P rate. The CC determined with linear-plateau and quadratic-plateau models are STP values at which the linear or quadratic portions of each model joined the predicted plateau yield. Because the exponential model predicts an asymptote to a maximum, CC were calculated for 99% of the maximum predicted yield.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Trends of Soil-Test Phosphorus over Time
The initial and annual P fertilization treatments applied at Boone and Kanawha resulted in a wide range of STP values and different STP trends over time (Fig. 1 and 2) . The initial P application rates used to create contrasting STP levels (IP2 and IP3) were much greater than any one-time P rate used for corn or soybean and resulted in very high STP that declined sharply during the first 2 yr of the study. This very steep STP decline was obvious for the IP3 treatment at Kanawha and both IP2 and IP3 treatments at Boone, and may not represent decline in soils with naturally high STP or high levels resulting from a slow buildup over time. Although all data are shown in Fig. 1 and 2, STP data from the IP2 and IP3 treatments used for fitting regression models to relationships between STP and time (Table 1) do not include the first year at Kanawha nor the first 2 yr at Boone. Table 1 shows the initial STP data used for the models. All observed STP data are shown in Fig. 3 for the Nashua site.



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 1. Soil-test P change over time for three initial P treatments and four annual P rates at Boone. Open symbols indicate data not used for regression models in Table 1. A vertical arrow indicates when applications of the 11- and 22-kg P ha–1 yr–1 rates were discontinued.

 


View larger version (23K):
[in this window]
[in a new window]
  		Fig. 2. Soil-test P change over time for three initial P treatments and four annual P rates at Kanawha. Open symbols indicate data not used for regression models in Table 1. A dotted line and missing symbols in 2 yr for the 33-kg P ha–1 yr–1 rate indicate that soils were not sampled. A vertical arrow indicates when applications of the 11- and 22-kg P ha–1 yr–1 rates were discontinued.

 

View this table:
[in this window]
[in a new window]
  		Table 1. Regression models for relationships between soil-test P (STP) and cropping years for various initial and annual P treatments at two sites.{dagger}

 


View larger version (30K):
[in this window]
[in a new window]
  		Fig. 3. Soil-test P change over time for three annual P application rates at the Nashua.

 
Analysis of STP trends for plots that received neither initial P (IP1) nor annual P at Boone and Kanawha and no annual P at Nashua is particularly useful because initial STP was near values classified Optimum (16–20 mg P kg–1) for corn and soybean in Iowa (Sawyer et al., 2002). Initial STP was 22, 17, and 28 mg P kg–1 at Boone, Kanawha, and Nashua, respectively. The probability of crop response to P within the Optimum class is <25%, and only maintenance fertilization is recommended. At Boone and Kanawha, STP of plots that received no annual P application for the IP1 initial treatment decreased exponentially to a minimum asymptote (Fig. 1 and 2, Table 1). Calculations based on the exponential model for Boone and Kanawha indicate that STP declined into the Low interpretation class (9–15 mg P kg–1) after 2 or 3 yr and into the Very Low class after 10 yr at Boone and after 6 yr at Kanawha. During the first 10-yr period at Boone, STP declined 1.2 mg P kg–1 yr–1 on average, while during the first 6 yr at Kanawha STP decreased 1.5 mg P kg–1 yr–1. During the recent 15 yr at both sites, STP of those plots remained almost constant (declined only 0.1 mg P kg–1 yr–1). At Nashua, STP of nonfertilized plots decreased linearly (Fig. 3) at 0.67 mg P kg–1 yr–1. This rate of decrease was less steep than at Boone and Kanawha even though initial STP at Nashua was slightly higher. Therefore, the results indicate that discontinuing P application in Webster–Nicollet soils (Boone) and Webster–Canisteo soils (Kanawha) with STP near Optimum results in an STP decrease into the Low class in 2 to 3 yr of cropping and into the Very Low class after 6 to 10 yr of cropping, whereas in Kenyon soil (Nashua) there is gradual, linear, and less steep STP decline over time.

Interpolation of coefficients of linear relationships between STP and time in Table 1 for Boone and Kanawha and in Fig. 3 for Nashua using Eq. [1] provided estimates of the annual P rate that would have maintained STP. At Boone, maintenance rates would have been 14 kg P ha–1 for IP1, 22 kg P ha–1 for IP2, and 29 kg P ha–1 for IP3. At Kanawha, maintenance rates would have been 17 kg P ha–1 for IP1, 26 kg P ha–1 for IP2, and 34 kg P ha–1 for IP3. At Nashua, STP would have been maintained with 13 kg P ha–1 yr–1. Increased maintenance P needs with increasing initial STP at Boone and Kanawha can be easily visualized in Fig. 1 and 2, and is in agreement with previous research (McCallister et al., 1987; McCollum, 1991; Webb et al., 1992). The methods and data collected in this study do not allow for determination of reasons for this result. Potentially higher P removal with harvest, higher P movement from the surface soil horizon to deeper soil layers, and higher P loss through erosion or leaching might explain the differences. Previous Iowa research with corn (Mallarino, 1996) showed that P fertilization can increase P concentration of grain at high STP levels but this effect would explain only a fraction of differences in maintenance rates observed in this study.

The maintenance P rate for Nashua was near values reported by Mallarino et al. (1991) for the early years of this experiment. The maintenance P rates for STP near Optimum values at Boone and Nashua were similar (14 and 13 kg P ha–1, respectively) and slightly lower (3 or 4 kg P ha–1 lower) than at Kanawha. However, initial STP at Nashua (28 mg P kg–1) was 6 to 11 mg P kg–1 higher than initial STP at Kanawha and Boone (17 and 22 mg P kg–1, respectively, for IP1). Therefore, based on increased maintenance rates for the higher STP observed for IP2 and IP3 treatments at Boone and Kanawha, the maintenance rate at Nashua would have been lower than at the other sites if initial STP had been similar. Different P removal with harvest did not explain lower maintenance needs at Nashua. Mean corn yield levels over time were similar at Boone and Nashua (within 100 kg ha–1 yr–1) and were approximately 200 kg ha–1 yr–1 greater than at Kanawha. Mean soybean yields were greater at Nashua, intermediate at Boone, and lowest at Kanawha (a difference of 250 to 300 kg ha–1 yr–1 between each site). A lower P maintenance need at Nashua might be explained by coarser texture of the Kenyon soil compared with soils at Boone and Kanawha.

Some annual P rates increased STP over time. At Boone and Kanawha, STP increased with the two highest P rates when no initial P was applied (IP1), only with the highest rate for IP2, and not even with the highest annual rate for IP3 (Fig. 1 and 2, Table 1). At Nashua, both annual P rates increased STP (Fig. 3). These increasing STP trends always were linear (Table 1). Linear STP increasing trends also were observed in other long-term studies (Cope, 1981; McCallister et al., 1987; McCollum, 1991). When no initial P was applied at Boone and Kanawha (IP1), the highest annual P rate increased STP by 1.42 and 1.18 mg P kg–1 yr–1, respectively. At Nashua, the highest annual P rate increased STP by 2.54 mg P kg–1 yr–1, which was higher than at other sites but at Nashua initial STP was 6 to 11 mg P kg–1 higher and the annual rate was 11 kg P ha–1 higher.

Because the same annual P rates were applied each year, STP increases over time expressed on a yearly basis from models in Table 1 can be easily transformed to an annual P rate basis. Calculations for the highest annual P rate (for IP1 at Boone and Kanawha) indicate that the annual P rate needed to increase STP 1 mg P kg–1 yr–1 was 23, 28, and 17 kg P ha–1 yr–1 for Boone, Kanawha, and Nashua, respectively. This result is important because nutrient management plans mandated by many state regulatory agencies require such estimates. A direct comparison of these values across sites is risky because the annual P rate used was similar for Kanawha and Boone but was higher for Nashua and initial STP was slightly different for these sites (17–28 mg P kg–1). Randall et al. (1997) reported that 26 kg P ha–1 yr–1 were needed to raise STP 1 mg P kg–1 yr–1 on a Minnesota Webster soil where continuous corn was grown for 7 yr followed by 11 yr of a corn–soybean rotation. The net annual P rate required to increase STP by a certain amount cannot be calculated in this study because grain P concentration was not measured and net applied P cannot be accurately calculated.

Grain Yield Response to Annual Phosphorus Fertilization
The grain yield response to annual P application varied among initial treatments and over time. Initial STP ranged from values currently considered within the optimum interpretation class to values more than three-fold higher. At Boone, corn yield (Table 2) was significantly (P ≤ 0.05) greater with P than without P in 15 yr for the IP1 treatment, 11 yr for the IP2 initial treatment, and only 4 yr for the IP3 treatment. Annual P rates higher than 11 kg P ha–1 increased yield further only in 2 yr for IP1, 2 yr for IP2, and 1 yr for IP3. The 33-kg rate significantly increased yield over the 22-kg rate only twice (in 1 yr for IP1 and 1 yr for IP3). Ten years of cropping were needed to observe corn response to P for IP1 (for which STP was near optimum), 12 yr for IP2, and 15 yr for IP3. Annual P application increased soybean yield (Table 3) in 10 yr for IP1, 8 yr for IP2, and 4 yr for IP3. Rates higher than 11 kg P ha–1 increased yield further in 2 yr for IP1 and only 1 yr for IP2 or IP3. The 33-kg rate never produced greater (P < 0.05) soybean yield than the 22-kg rate, not even since the lower rates were discontinued. Eight years of cropping were required to observe soybean response to P for IP1, 10 yr for IP2, and 19 yr for IP3.


View this table:
[in this window]
[in a new window]
  		Table 2. Corn yield for various initial and annual P (AP) fertilizer treatments at Boone (yields for 1977, 1989, and 1990 were not available).

 

View this table:
[in this window]
[in a new window]
  		Table 3. Soybean yield for various initial and annual P (AP) fertilizer treatments at Boone (yields for 1977 were not available).

 
At Kanawha, corn yield response (Table 4) to annual P fertilization was significant (P ≤ 0.05) in 23 yr for IP1, 12 yr for IP2, and 6 yr for IP3. Four years of cropping were needed to observe corn response to annual P for IP1, 10 yr for IP2, and 17 yr for IP3. The two high annual rates increased corn yield further than the low rate in 4 yr for the IP1 initial treatment, 2 yr for IP2, and never for IP3. Only in some of the last 5 yr of the experiment, when the lower rates were discontinued, did the 33-kg increase yield over the 22-kg rate and only lowest initial P treatment (IP1). Soybean yield responses to annual P (Table 5) were observed in 24 yr for the IP1 initial treatment, 12 yr for IP2, and 6 yr for IP3. The number of crop-years required to observed soybean response to P were 4 yr for IP1, 11 yr for IP2, and 17 yr for IP3. The 22-kg annual P rate increased yield further than the 11-kg rate in 9 of the last 16 yr for IP1, 4 of the last 5 yr for IP2, and never for IP3. The 33-kg annual rate significantly increased yield over the 22-kg rate only once (and for IP1 when the lower rates were discontinued).


View this table:
[in this window]
[in a new window]
  		Table 4. Corn yield for various initial and annual P(AP) fertilizer treatments at Kanawha.

 

View this table:
[in this window]
[in a new window]
  		Table 5. Soybean yield for various initial and annual P (AP) fertilizer treatments at Kanawha.

 
At Nashua, where initial STP level was in the High interpretation class, crop response to annual P application was observed only in recent years as a result of STP decline in control plots (Table 6). Significant corn yield response to P (P ≤ 0.05) occurred only in 3 of the 24 yr of the experiment, and the high rate never increased yield more than the low rate. Soybean was more responsive to P than the corn, and yield increases were observed in 8 yr, mainly last years of the study. The more frequent soybean response to P compared with corn at this site was not observed at Boone and Kanawha. This difference might be explained by consistently higher soybean yield at Nashua than at the other sites. For comparable initial STP levels, mean corn yield was similar (within 140 kg ha–1) at Boone and Nashua but approximately 300 kg ha–1 greater than at Kanawha. However, mean annual soybean yield was 300 to 500 kg ha–1 greater at Nashua than at the other sites.


View this table:
[in this window]
[in a new window]
  		Table 6. Corn and soybean yield for three annual P (AP) fertilizer rates at Nashua.

 
Observed grain yield response and STP trends over time for the different P treatments and sites indicate that in soils testing 43 to 96 mg kg–1 STP, 10 to 20 yr of cropping without P additions were required before STP declined to levels where P fertilization increased (P < 0.05) yield. The maintenance P rates identified in this study (13–17 kg P ha–1) for STP near the Optimum class are lower than currently recommended maintenance rates for this class and comparable yield levels (Sawyer et al., 2002). Mean yields across the three sites for the highest annual P rate were 9900 and 3040 kg ha–1 for corn and soybean, respectively. Recommended maintenance rates for these yield levels are 29 and 18 kg P ha–1 for corn and soybean, respectively. These recommendations are based on assumed grain P concentrations of 2.91 g P kg–1 for corn and 5.80 P kg–1 for soybean. The P concentration used for corn is in the upper 20% of concentrations in grain from 24 P fertilization trials conducted during 1989 and 1990 (Mallarino, 1996) and 47 P fertilization or manure trials conducted in Iowa from 1994 until 2003 (unpublished data, A.P. Mallarino, 2004). The P concentration used for soybean is in the upper 30% of concentrations in grain from 30 P fertilizer experiments conducted in Iowa from 1994 to 1999 (unpublished data, A.P. Mallarino, 2004) but is near the average from 13 manure P fertilizer or manure trials conducted from 1999 to 2003 mainly in high-testing soils (unpublished data, M.M. Barbazan and A.P. Mallarino, 2004). Therefore, our results also indicate that current Iowa P maintenance recommendations for comparable yield levels of corn and soybean may result in a small STP build-up over time.

Soil-Test Phosphorus Critical Concentrations for Corn and Soybean
Exponential, linear-plateau, and quadratic-plateau models fit to relationships between STP and relative grain yield across all years of each site are shown in Table 7. This table also shows CC for each model, site, and crop. Data for soybean at Nashua are not shown because few years were available (plots with corn residue were not sampled from 1981 to 1991) and no model fit was significant at P < 0.05. However, all available data were included in calculations of CC across all sites shown in Table 8.


View this table:
[in this window]
[in a new window]
  		Table 7. Regression models for relationships between relative grain yield and soil-test P for corn and soybean at three sites and determined critical soil-test concentrations.

 

View this table:
[in this window]
[in a new window]
  		Table 8. Regression models for relationships between relative grain yield and soil-test P for corn and soybean across three sites and determined critical soil-test concentrations.

 
At Boone (Table 7), corn was not likely to respond to P at STP > 13 to 18 mg P kg–1 (depending on the model used) and soybean was not likely to respond when STP > 11 to 16 mg P kg–1. At Kanawha, the crops responded to P application up to slightly higher STP. Corn was not likely to respond to P at STP > 19 to 28 mg P kg–1 soybean was not likely to respond when STP > 17 to 24 mg P kg–1. At Nashua, the relationship between STP and relative corn yield was poorer than at the other sites as a result of higher STP and both small and infrequent yield response. Although model fits for corn were statistically significant for this site, the R2 were very low, observed STP concentrations were mainly in the nonresponsive range, and relationships cannot be reasonably used to establish site-specific CC.

There were large differences in CC determined by the models, which is a known fact. Dahnke and Olson (1990), Mallarino and Blackmer (1992), and Cox (1996) discussed implications of these differences for fertilizer recommendations and the profitability of fertilization. A CC range defined by values determined with the linear-plateau and quadratic-plateau models has been used before to compare results for different soil tests and locations (Mallarino, 1997, 2003). The CC ranges for both crops were 6 to 10 mg P kg–1 higher at Kanawha (with Webster–Canisteo soils) than at Boone (with Nicollet–Webster soils). This difference between soils cannot be explained with certainty. If the Bray-P1 test were underestimating plant-available P in the Webster–Canisteo soil complex at Kanawha (because of slightly alkaline pH due to CaCO3) as was suggested by previous research (Mallarino, 1997) in soils with pH > 7.3, the CC range should have been lower than for the Nicollet–Webster complex at Boone. Crop yields were higher at Boone than at Kanawha and soils at Boone are better drained than at Kanawha, which suggests better overall plant growth and perhaps root growth at Boone. Therefore, higher yield levels and better conditions for plant growth might explain lower STP requirements at Boone than at Kanawha. Critical STP concentrations calculated across all sites are shown in Table 8 and Fig. 4 . The CC range defined by the linear-plateau and quadratic-plateau models was 15 to 21 mg P kg–1 for corn and 12 to 18 mg P kg–1 for soybean.



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 4. Relationship between soil-test P and relative yield of corn (Zea mays L.) and soybean [Glycine max (L.) Merr.] across all years of experiments at three Iowa locations (LP, linear-plateau; QP, quadratic-plateau; and EXP, exponential).

 
Relationships between relative yield and STP and the determined CC ranges confirm some previous results from Iowa, but do not confirm others. The CC for corn approximately coincides with results based on numerous short-term (1 yr) experiments (Mallarino and Blackmer, 1992; Mallarino, 1997) and with the current Optimum Iowa STP interpretation class (16–20 mg P kg–1) for both corn and soybean (Sawyer et al., 2002). However, the CC range for soybean is slightly lower than the Optimum class. Further indications of slightly lower STP needs for soybean were provided by previous work with ridge-till and no-till soybean (Borges and Mallarino, 2000, 2003), although the number of sites and infrequent yield response did not allow for determining CC in either study.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Results of these experiments spanning nearly 30 yr showed that long-term STP trends are greatly affected by the initial STP value, cropping years, and annual P fertilization rates. With the only exception of the Nashua site (Kenyon soil), not applying P to soils testing 17 to approximately 70 mg P kg–1 resulted in a steep STP decline during the first 6 to 10 yr followed by a very small and gradual decline. Not applying P at Nashua resulted in a linear STP decline over time. Soil with high initial STP required higher annual P additions to maintain STP compared with soils testing near optimum for crops. Annual P rates required to maintain initial STP near the Optimum Iowa interpretation class (16–20 mg P kg–1) were 13 to 17 kg P ha–1 yr–1 but up to 33 kg P ha–1 yr–1 were required when STP was as high as three-fold higher. These maintenance rates were lower than currently recommended rates in Iowa. Use of current rates likely results in a small STP buildup over time. Study of STP trends over time for annual P rates that increased STP indicated that the P required to increase STP by 1 mg P kg–1 was 23, 28, and 17 kg P ha–1 yr–1 for sites with Nicollet–Webster (Boone), Webster–Canisteo (Kanawha), and Kenyon (Nashua) soils, respectively.

Ten to twenty years of cropping without P fertilization were needed before yield response to P was observed in soils testing 43 to 96 mg P kg–1. Corn and soybean responded to annual P fertilization 50 to 75% of the time when STP ≤ 20 mg P kg–1 and did not respond at higher STP levels. Crops seldom responded to P rates > 11 kg P ha–1 yr–1 unless STP < 16 mg P kg–1 (the low boundary of the current Optimum Iowa class). Critical STP concentration ranges across sites and years defined by linear-plateau and quadratic-plateau models were 15 to 21 mg P kg–1 for corn and 12 to 18 mg P kg–1 for soybean. The CC range for corn encompasses the current Optimum class by extending only 1 mg P kg–1 into the Low and High classes. The CC range for soybean encompasses the upper half of the Low class (9–15 mg P kg–1) and the lower half of the Optimum class.

Overall, this study provided useful information of long-term corn and soybean yield response to P and of STP decline or buildup over time for various initial STP levels and P fertilization rates. This information is needed for improved P management plans, which are useful for farmers and are required by state regulatory agencies. Also, results indicated that the current Optimum Iowa STP interpretation class is appropriate for corn but is slightly higher than needed for soybean.


   ACKNOWLEDGMENTS
 
We dedicate this paper to the memory of John R. Webb, former Professor of Agronomy at Iowa State Univ., who established two experiments summarized in this paper and collected data for many years. We thank the following persons for their valuable help at managing the experiments during many years. Kenneth Ross and Kenneth T. Pecinovsky, Superintendents (former and current, respectively), Iowa State Univ. Northeast Research and Demonstration Farm; David Rueber, Superintendent, Iowa State Univ. Northern Research and Demonstration Farm; and Regis D. Voss, Professor of Agronomy (retired), Iowa State University.

Received for publication August 19, 2004.


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




This article has been cited by other articles:


Home page
 Soil Sci.Home page
G. Rubio, M. J. Cabello, F. H. Gutierrez Boem, and E. Munaro
Estimating Available Soil Phosphorus Increases after Phosphorus Additions in Mollisols
Soil Sci. Soc. Am. J., October 30, 2008; 72(6): 1721 - 1727.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
M. Bermudez and A. P. Mallarino
Impacts of Variable-Rate Phosphorus Fertilization Based on Dense Grid Soil Sampling on Soil-Test Phosphorus and Grain Yield of Corn and Soybean
Agron. J., May 11, 2007; 99(3): 822 - 832.
[Abstract] [Full Text] [PDF]

Home page
 Soil Sci.Home page
B. L. Allen and A. P. Mallarino
Relationships between Extractable Soil Phosphorus and Phosphorus Saturation after Long-Term Fertilizer or Manure Application
Soil Sci. Soc. Am. J., February 2, 2006; 70(2): 454 - 463.
[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 (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Dodd, J. R.
Right arrow Articles by Mallarino, A. P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Dodd, J. R.
Right arrow Articles by Mallarino, A. P.
Agricola
Right arrow Articles by Dodd, J. R.
Right arrow Articles by Mallarino, A. P.
Related Collections
Right arrow Nutrient Management
Right arrow Soil Fertility and Productivity

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