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


     

Published online 4 January 2008
Published in J Environ Qual 37:125-137 (2008)
DOI: 10.2134/jeq2007.0125
© 2008 American Society of Agronomy, Crop Science Society of America, and 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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Agricola
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Related Collections
Right arrow Phosphorus
Right arrow Runoff
Right arrow Animal Waste

TECHNICAL REPORTS

Surface Water Quality

Effect of Liquid Swine Manure Rate, Incorporation, and Timing of Rainfall on Phosphorus Loss with Surface Runoff

Brett L. Allen and Antonio P. Mallarino*

Dep. of Agronomy, Iowa State Univ., Ames, IA 50010. B.L. Allen was formerly a graduate research assistant and postdoctoral research associate and is currently at USDA-ARS, Northern Plains Agricultural Research Lab, Sidney, MT 59270

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

Received for publication March 10, 2007.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
Excessive manure phosphorus (P) application increases risk of P loss from fields. This study assessed total runoff P (TPR), bioavailable P (BAP), and dissolved reactive P (DRP) concentrations and loads in surface runoff after liquid swine (Sus scrofa domesticus) manure application with or without incorporation into soil and different timing of rainfall. Four replicated manure P treatments were applied in 2002 and in 2003 to two Iowa soils testing low in P managed with corn (Zea mays L.)–soybean [Glycine max (L.) Merr.] rotations. Total P applied each time was 0 to 80 kg P ha–1 at one site and 0 to 108 kg P ha–1 at the other. Simulated rainfall was applied within 24 h of P application or after 10 to 16 d and 5 to 6 mo. Nonincorporated manure P increased DRP, BAP, and TPR concentrations and loads linearly or exponentially for 24-h and 10- to 16-d runoff events. On average for the 24-h events, DRP, BAP, and TPR concentrations were 5.4, 4.7, and 2.2 times higher, respectively, for nonincorporated manure than for incorporated manure; P loads were 3.8, 7.7, and 3.6 times higher; and DRP and BAP concentrations were 54% of TPR for nonincorporated manure and 22 to 25% for incorporated manure. A 10- to 16-d rainfall delay resulted in DRP, BAP, and TPR concentrations that were 3.1, 2.7, and 1.1 times lower, respectively, than for 24-h events across all nonincorporated P rates, sites, and years, whereas runoff P loads were 3.8, 3.6, and 1.6 times lower, respectively. A 5- to 6-mo simulated rainfall delay reduced runoff P to levels similar to control plots. Incorporating swine manure when the probability of immediate rainfall is high reduces the risk of P loss in surface runoff; however, this benefit sharply decreases with time.

Abbreviations: BAP, bioavailable P • DRP, dissolved reactive P • STP, soil-test P • TPR, total runoff P • TS, total solids


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
PHOSPHORUS (P) is an essential crop nutrient, and P application is often necessary to increase soil-test P (STP) and/or to offset P removed in grain and forage. Swine manure often is applied to fields because it can be an effective source of plant nutrients. The plant availability of manure P and the proportion of organic and inorganic P forms vary greatly across manure sources, storage systems, and animal diets (Barnett, 1994). In some manures, such as liquid swine manure or poultry manure, inorganic P forms extractable with water or dilute acid or basic solutions may comprise 80 to 90% of the total P (Sharpley and Moyer, 2000). An increasing concentration of animal production in some parts of Iowa and the Corn Belt may necessitate the disposal of large amounts of manure over relatively small areas and consequent addition of P to many fields where no additional P is needed for optimum crop production (Whalen and Chang, 2001). Disposal of manure as a waste using excessive rates or inappropriate application methods can lead to eutrophication of freshwater resources (Correll, 1998).

The amount of P delivery to water bodies is influenced by various source and transport factors, including P source, rate and method of application, soil P level, field slope, soil erosion, surface runoff, subsurface drainage, and proximity to surface waters (Sharpley et al., 1994; Sims et al., 1998). The Iowa P Index is one example of an assessment tool developed to estimate the relative contribution of various source and transport factors to the risk of potential P loss from fields (USDA-NRCS, 2002; Mallarino et al., 2002). These P-assessment tools recognize that P can be delivered as particulate P (P bound to soil particles) or dissolved P in surface water runoff or subsurface drainage. Commonly measured P fractions in runoff include dissolved reactive P (DRP), which is primarily dissolved orthophosphate and is determined colorimetrically with the Murphy and Riley method (Murphy and Riley, 1962); bioavailable P (BAP), usually estimated as P extracted by Fe-oxide impregnated filter paper (Sharpley, 1993); and total runoff P (TPR). The DRP and BAP runoff fractions are considered the most critical P forms contributing to accelerated aquatic growth in surface waters. However, TPR can be a better estimate of long-term potential for eutrophication of surface waters because forms other than DRP or BAP can be used by algae after changes that occur at variable rates depending on water chemistry and other factors (Correll, 1998).

Soil-test P concentration is an important source factor of P assessment tools, and these tools include an estimate of the relationships between runoff P fractions and soil P. Routine soil P tests used to predict P sufficiency for crops are used for most P indices, although environmental P tests that could provide a soil P measurement better related to algal growth in water bodies correlate as well or better with runoff P concentrations (Pote et al., 1996; Kleinman and Sharpley, 2002; Maguire and Sims, 2002; Allen et al., 2006). Manure or fertilizer P application has a large impact on runoff P immediately after application and often eliminates otherwise good relationships between STP and runoff P (Sharpley and Tunney, 2000; Tabbara, 2003, Daverede et al., 2004). Incorporating manure P into the soil without significantly increasing soil erosion reduces runoff P concentration and loss. For example, Tabbara (2003) reported that incorporating liquid swine manure or fertilizer P into an Iowa soil reduced DRP and TPR concentrations and loads by as much as 30 to 60% depending on the application rate. However, Bundy et al. (2001) showed that tillage to incorporate dairy manure in a Wisconsin soil lowered runoff DRP concentration compared with surface application but increased TPR concentrations and loads due to increased sediment loss. The P source may also affect relationships between manure P rate and runoff P. In an indoor rainfall simulation study, Kleinman and Sharpley (2003) showed that application of dairy, poultry, and swine manures at similar P rates increased DRP less for dairy manure than for the other sources.

The sequence and time interval between manure application to soil and a runoff event also play key roles in the magnitude of P loss. Broadcast and unincorporated manure concentrates soluble P at the soil surface, and, therefore, the P may be easily removed by runoff water (Mueller et al., 1984; Eghball and Gilley, 1999). After manure application without incorporation into the soil, however, the potential for P loss declines over time as the more soluble P forms in the manure increasingly interact with soil (Edwards and Daniel, 1993). Mueller et al. (1984) reported sharp declines in DRP concentrations in runoff after about 2 mo from no-till plots broadcast with dairy manure. In their indoor rainfall simulation study, Kleinman and Sharpley (2003) showed that DRP and differences in DRP related to manure type diminished after repeated rainfall events. In a field study, Daverede et al. (2004) reported significantly lower DRP, BAP, and TPR concentrations and loads in runoff 6 mo subsequent to surface-applied liquid swine manure. One additional factor that affects P loss immediately after application of liquid manure without incorporation is partial soil surface sealing by the manure (Smith et al., 1998). However, Smith et al. (2001) reported that surface sealing was short lived and that in the long term, swine manure additions decreased runoff P by increasing soil aggregate stability and water infiltration. Incorporation of manure or fertilizer has reportedly decreased TPR concentrations and loads in runoff by placing P below the thin mixing zone of interaction between soil and surface runoff (Baker and Laflen, 1982; Kimmell et al., 2001; Withers et al., 2001; Sharpley, 2003). Other studies, however, found that manure incorporation or tillage decreased DRP concentrations but increased TPR concentrations and loads as a result of decreased infiltration and increased sediment loss (Eghball and Gilley, 1999; Bundy et al., 2001).

More information is needed concerning the effects of tillage, manure P rate, and timing of runoff events shortly after manure application on P loss with surface runoff. A better understanding of these relationships is necessary to maximize the effectiveness of comprehensive management tools such as the P index. Therefore, the objective of this field rainfall simulation study was to assess the impact of different application rates of liquid swine manure, its incorporation into the soil, and timing of simulated rainfall after manure application on concentrations and loads of DRP, BAP, and TPR in surface runoff.


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results and Discussion
 Conclusions
 REFERENCES
 
Sites, Treatments, and Manure Application
Phosphorus loss after liquid swine manure application to two Iowa fields was assessed using a simulated rainfall technique during 2 yr (2002 and 2003). The fields were selected to have low STP according to Iowa interpretations (Sawyer et al., 2002). One experiment was located near Boone on a field area with Clarion loam soil (fine-loamy, mixed, superactive, mesic Typic Hapludolls) with 3 to 5% slope, and the other was located near Marshalltown on an area with Tama silty clay loam soil (fine-silty, mixed, superactive, mesic Typic Argiudolls) with 8 to 12% slope. These sites hereafter are referred to as the Boone and Marshalltown sites. The Clarion soil is a moderately well drained, moderately permeable soil formed in glacial till on the uplands of north-central Iowa and south-central Minnesota with slopes ranging from 1 to 9% and medium surface runoff potential (USDA-NRCS, 2004). The Tama soil is a well and moderately drained, moderately permeable soil formed in loess on the uplands and high stream benches of Iowa and smaller areas of Illinois, Minnesota, and Wisconsin, with slopes ranging from 0 to 20% and medium to rapid surface runoff potential (USDA-NRCS, 2004). The fields had been managed with corn (Zea mays L.)–soybean [Glycine max (L.) Merr.] rotations without manure application during the last 20 yr and had not received P fertilizer since the last crop was harvested. The Boone site had been under chisel-disk tillage management; and corn residue was chiseled in the fall and disked in spring, whereas soybean residue was disked in spring. The Marshalltown site had been under no-till management for at least 10 yr.

Field plots measuring 3.8 by 9.1 m were laid out at each site to accommodate a randomized complete-block design with eight treatments at both sites and four replications at Boone and three at Marshalltown. Treatments were four rates of liquid swine manure P applied with or without incorporation into the soil by tillage. At Boone, treatments were first applied in April 2002 to soil with soybean residues, corn was planted in early May and harvested in late September, treatments were re-applied to the same plots in April 2003, and soybean was planted in early May and harvested in late September. At Marshalltown, treatments were first applied in October 2002 to soil with soybean residues, treatments were re-applied to the same plots in early April 2003, and corn was planted in late April and harvested in early October.

Manure was collected from underground storage pits of confined swine operations located near each site. Preliminary manure samples were collected from the pits 5 d before each application date and were analyzed for moisture and total P concentration to define the amount of manure for each treatment. For the first application, intended P rates were 0, 0.5, 1, and 2 times the suggested average grain P removal for corn–soybean rotations in Iowa (24 kg P ha–1 yr–1) (Sawyer et al., 2002). For the second application, intended P rates were twice the first-year amounts to better encompass the manure disposal rates used by most farmers. The manure was broadcast using a 1600-L capacity research applicator having extra-wide wheel separation (3 m); a pump to agitate the tank content; a flow meter; and two exit outlets, each with a pump and a splash pan to achieve the most uniform possible manure distribution. One full tank load was enough for each application date at each site. The applicator was calibrated to apply the lowest manure rate with one pass, and multiple passes with controlled traffic were made to apply the amounts needed for the other treatments. Because of known large P variability in manure, four 1-L samples were collected at different times while manure was being applied to better estimate moisture and P concentration in manure actually applied.

Total manure P was analyzed by the persulfate digestion method following USEPA Method 365.1 (USEPA, 1983) and by determining P in digests with the ammonium molybdate–ascorbic acid colorimetric method (Murphy and Riley, 1962). Total manure solids were determined gravimetrically following USEPA Method 160.3 (USEPA, 1983). Average total P and solids concentration of manure samples taken while manure was being applied at Boone in 2002 were 0.63 and 32 g L–1, respectively; in 2003, average total P and solids concentration were 0.94 and 43 g L–1, respectively. At Marshalltown, average manure total P and solids concentration in 2002 were 0.67 and 24 g L–1, respectively; in 2003 average manure total P and solids concentration were 1.3 and 58 g L–1, respectively. Therefore, the actual manure P rates for the first application were 0, 6.5, 13, and 26 kg P ha–1 at Boone and 0, 7, 14, and 28 kg P ha–1 at Marshalltown; for the second application, manure P rates were 0, 20, 40, and 80 kg total P ha–1 at Boone and 0, 27, 54, and 108 kg total P ha–1 at Marshalltown.

The manure spread to plots of the incorporated treatment was incorporated approximately 2 h after application (to let the liquid soak into the soil) by disking to a depth of 10 to 15 cm using a tractor and tandem disk-harrow with a similar wheel separation (3 m) used for the manure application. The disk-harrow had spring-teeth harrows in the rear to smooth the soil, and two passes were made in opposite directions for each plot. Similar methods were used for the second manure application. Controlled traffic was used, and only the nontrafficked area of each plot between the wheel tracks was used for the rainfall simulation study. During the period between the first manure application and the end of the study, the farmers planted and harvested corn or soybean as called for by the rotation without additional tillage and while avoiding equipment wheel tracks in the 3-m central section of each plot.

Simulated Rainfall Events
A portable rainfall simulator was built on a design by Miller (1987) with minor structural modifications suggested by the National Phosphorus Research Project (NPRP, 2002). Briefly, a Veejet HH-SS50 WSQ nozzle (Spraying Systems, Wheaton, IL) was supported by a cube frame made of Al pipes that measured 3 m on each side and was placed 3 m above and at the center of the rained-on area. Preliminary calibrations using collector pans showed that this nozzle applied a uniform volume of water over an area approximately 5 m in diameter. Before applying manure treatments or simulated rainfall, galvanized steel borders that were 15 cm in height were set into soil (7.5 cm depth) encompassing a 1.5 by 2 m area (referred to hereafter as a microplot) at the center of each plot so that no wheel track traffic affected the study area. At the down-slope end of each microplot and after applying manure but before applying rainfall, a v-shaped flume that measured 1.5 m in width and 0.5 m in length was installed with the upper edge level with the soil surface. The flume was equipped with a canopy to exclude direct input of rainfall, and a 10-cm diameter plastic pipe was connected to the flume to route runoff water away from the microplot to a plastic collecting vessel placed outside of the rainfall area and buried so that its surface was at ground level. Plastic curtains wrapped around the simulator frame eliminated potential effects of wind on the rainfall intensity and uniformity.

At both sites, simulated rainfall was applied three times for the first manure application and twice for the second manure application. At Boone, simulations were within 4 to 24 h and after 10 to 11 d and 5 mo of the first manure application and within 4 to 24 h and after 5 mo of the second manure application. At Marshalltown, simulations were within 4 to 24 h and after 15 to 16 d and 5 mo of the first manure application and within 4 to 24 h and after 6 mo of the second application. For simplicity, hereafter these rainfall events are referred to as 24-h, 15-d, and 6-mo events. Simulated rainfall for the 15-d runoff event was applied to a plot area adjacent to the area used for the 24-h event. Simulated rainfall for the 6-mo runoff event subsequent to the first manure application was applied to the same plot area used for the 24-h event. Simulated rainfall for the 24-h and 6-mo events subsequent to the second manure application was applied to different plot areas.

Water for the rainfall simulations was obtained from rural water sources close to the treatment plots. Water was analyzed once daily for DRP concentration and tested <0.005 mg P L–1 at Boone and <0.03 mg P L–1 at Marshalltown. Concentrations of Ca, Mg, and Na measured in two samples at each site and season averaged 220, 9, and 100 mg L–1, respectively, at Boone and 40, 2, and 43 mg L–1, respectively, at Marshalltown. Simulated rainfall was applied to each microplot in two sequential steps to reduce the time needed to produce runoff. For simulations within 24 h of manure application, rainfall was first applied at an intensity of 7 L min–1 to the point of runoff using a hose fitted with a nozzle attachment that was moved back and forth approximately 0.5 m above the microplot area. The amount of rainfall applied and time to the point of runoff varied between plots, mainly due to the different application rates of liquid manure and to the expected variation in soil physical properties across the plots. After 15 to 30 min, additional simulated rainfall was applied with the simulator described previously at 76 mm h–1 (energy of 0.278 MJ ha–1 mm–1) until 30 min of runoff occurred. This rainfall intensity (including time to runoff that on average was 5 min and varied little among plots) has a recurrence interval of approximately 13 yr in Iowa (Huff and Angel, 1992). In general, time to runoff for both sequential rainfall steps for 24-h events decreased as manure rates increased regardless of incorporation, but this effect was less consistent for the 15-d events and was not observed for 6-mo events (data not shown). For rainfall simulations 15-d or 6-mo after manure application, the high-intensity rainfall was applied the day before applying the low-intensity rainfall and collecting runoff. The total runoff collected during 30 min was weighed, and a 1-L sample was collected after vigorously stirring the container. A 20-mL subsample was filtered in the field using a 0.45-µm pore-size filter for each sample. The unfiltered sample was used for BAP and TPR analyses, and the filtered sample was used for DRP analysis. All runoff samples were kept in insulated boxes and taken to a cold storage room (4–5°C) until analysis.

Any natural rainfall occasionally occurring the 24-h or 15-d simulated rainfall events was excluded from the plots by covering them with plastic sheets until the risk of rain was over. The plots were never covered more than 12 h, and the natural rainfall never produced runoff outside the plots. No attempt was made to exclude natural precipitation from the plots between the 15-d and 6-mo rainfall simulation events. Recorded precipitation (rain or snow) during this period for 2002 and 2003 was 524 and 608 mm at Boone, respectively, and 39 and 578 mm at Marshalltown, respectively. Surface residue cover was measured for all microplots using a line-transect method described by Laflen et al. (1981). A 1.5-m-long rod with 10 points marked equidistantly across its length was placed across the width of each microplot. This was repeated at five locations within each microplot to account for variability of residue distribution.

Analysis of Soil and Surface Runoff Samples
Soil samples were taken from a depth of 0- to 15-cm from each plot before the first and second manure applications. This is the soil sampling depth suggested by Iowa State University and required by Iowa state agencies for agronomic and environmental soil P testing for all tillage systems (Sawyer et al., 2002; Mallarino et al., 2005). Soil samples were oven-dried at 65°C and crushed to pass through a 2-mm sieve and were used for all tests except total soil P and organic matter, for which subsamples were further crushed to pass a 0.5-mm screen. The total soil P was determined with the alkaline-oxidation digestion procedure (Dick and Tabatabai, 1977) adapted to an aluminum digestion block (Cihacek and Lizotte, 1990). Soil-test P by the Mehlich 3 P method was analyzed following procedures recommended for the North Central Region (Frank et al., 1998). Extracts were filtered through Whatman no. 42 filter paper, and extracted P was determined colorimetrically as described previously. Total C was determined by a combustion method (Wang and Anderson, 1998). Soil particle size distribution was determined by the standard pipette method (Walter et al., 1978). Soil pH was determined using a 1:1 soil/water ratio. Table 1 shows selected initial soil properties, and Table 2 shows STP before the second manure application.


View this table:
[in this window]
[in a new window]

  		Table 1. Selected initial soil properties for a 0- to 15-cm sampling depth.

 

View this table:
[in this window]
[in a new window]

  		Table 2. Soil-test P and total soil P (0- to 15-cm depth) before a second swine manure application in April 2003 at two sites.

 
Filtered runoff samples were analyzed for DRP with the ammonium-molybdate ascorbic-acid method (Murphy and Riley, 1962). Unfiltered runoff samples were analyzed for BAP and TPR. Procedures for BAP were those described by Sharpley (1993) for the Fe-oxide–impregnated filter paper extraction method adapted to filter paper disks (5.5 cm diam., Whatman no. 50), 50-mL centrifuge tubes, and 40 mL runoff. For TPR, runoff samples were digested using the alkaline-oxidation digestion method described previously as adapted to runoff by Laflen and Tabatabai (1984). The P in BAP extracts and TPR digests was determined colorimetrically by the Murphy and Riley (1962) method. Total solids (TS) in runoff were analyzed following the method of the American Public Health Association (APHA, 1998).

Statistical Design and Analysis
Treatment effects on runoff P for each runoff event were assessed by ANOVA and regression analyses using the GLM procedure of SAS (SAS Institute, 1999) for a randomized complete-block design assuming fixed treatment and block effects. The effects tested were manure, tillage (incorporation or not incorporation), and a manure x tillage interaction. Because the interaction was significant (P < 0.05) in most 24-h and 15-d runoff events, the average effects of manure or tillage are not shown or discussed. Linear and curvilinear (quadratic) effects of manure P were tested for each tillage treatment, and curvilinear trends are shown only when they were significant (P < 0.10) after the linear trends. In some instances, an exponential model fit the data significantly better than a quadratic equation (i.e., residual sums of squares for the quadratic model were larger than for the exponential model at P < 0.10). In these instances, the exponential model shown was fit using a nonlinear procedure included in Sigmaplot (SPSS Inc., 233 S. Wacker Drive, Chicago, Illinois).


   Results and Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
Conclusions
 REFERENCES
 
Runoff Phosphorus Concentrations
First Manure Application
Figure 1 shows effects of manure P rate and incorporation on runoff P concentration for the first manure application at Boone. When rainfall was applied within 24 h of manure application without incorporation, runoff DRP, BAP, and TPR increased linearly with increasing P rates. Incorporating the manure sharply decreased DRP, BAP, and TPR concentrations and the rate of increase as the P application rate increased. We cannot explain a smaller-than-expected TPR increase due to the highest manure P rate that explained a quadratic response. The incorporation effect was more obvious for DRP and BAP than for TPR even without considering this apparently anomalous TPR response to the high P rate. Runoff TS (not shown) followed a similar pattern to TPR for the incorporated treatments, and these two variables were significantly correlated (r = 0.77; P < 0.05). In contrast, TPR and TS were not significantly correlated for the nonincorporated treatment (not shown). Without manure incorporation, the fraction of TPR made up of DRP or BAP increased with the P rate (for example, DRP increased from 33% of TPR for the lowest rate to 62% for the highest rate). With manure incorporation, DRP and BAP fractions of TPR were much lower and increased less with increasing P rates (for example, DRP increased from 6% of TPR for the lowest rate to 16% for the highest rate). This result suggests that intensive rainfall immediately after applying high rates of manure without incorporation into the soil increases the risk of TPR loss and that the risk grows proportionally higher for forms of P that are readily available to algae. Other research has shown that incorporating manure places P below the zone of interaction between the soil and incoming rainfall or runoff and reduces DRP concentration in runoff (Withers et al., 2001; Tabbara, 2003; Haq et al., 2003; Daverede et al., 2004).


Figure 1
View larger version (23K):
[in this window]
[in a new window]

  		Fig. 1. Effect of manure application rate, incorporation, and time of simulated rainfall on runoff P concentrations for the Boone site in 2002. INC, manure incorporated; NOINC, manure not incorporated. NS indicates no interaction (P < 0.05) between P rate and incorporation.

 
Delaying simulated rain after manure application for approximately 15 d sharply reduced DRP and BAP (Fig. 1). The effect of delayed rainfall was proportionally greater for the nonincorporated treatment, which is shown by a substantial difference between the incorporation treatments only for the highest P rate that was much smaller than for the 24-h runoff event. In fact, a runoff event delay resulted in little or no reduction in DRP and BAP concentrations. The data (and especially the curvilinear trends) indicate that the effect of delaying runoff was proportionally greater for the low P rates. For example, DRP concentration with the high P rate for the 15-d event was 2.9 times lower than for the 24-h event but was 4.9 times lower for the lowest P rate. For nonincorporated treatments, the ratio of DRP to BAP decreased with delayed rainfall compared with the immediate rainfall (0.99–0.86), suggesting that the proportion of DRP in BAP decreased with time. Delaying simulated rainfall also decreased TPR but proportionally less than DRP and TRP. Moreover, the interaction between P rate and tillage was the opposite of results for DRP and BAP, and TPR for the two lower P rates was smaller without incorporation than with incorporation. The effects of delaying rainfall on DRP might be explained by proportionally more significant effects of applied P sorption by surface soil constituents for the nonincorporated and lower manure P rates. Lower TPR with delayed runoff and low, nonincorporated P rates compared with incorporated P rates can be explained by proportionally lower loss of particulate P without tillage.

Another significant result for the 24-h and 15-d runoff events at Boone was that DRP concentrations were only slightly lower than BAP concentrations (Fig. 1). Other Iowa research with poultry manure (Haq et al., 2003) and liquid swine manure (Tabbara, 2003) showed larger differences between DRP and BAP. The difference between DRP and BAP concentrations likely is affected by forms of P in the manure sources and by soil conditions affecting the proportion of dissolved and particulate P in the soils. Although a major proportion of BAP was accounted for by the DRP fraction in the manured plots, DRP was approximately one half of BAP in the nonmanured plots. This difference clearly shows the over-riding effect of recently applied manure. Although studies (e.g., Sharpley and Moyer, 2000) reported that about 90% of total P in swine manure may be inorganic, the results suggest that further study is required to investigate DRP and BAP relationships in runoff with various manure sources.

Delaying simulated rainfall for 6 mo at Boone (Fig. 1) reduced runoff P concentrations for the nonincorporated treatment to the point that manure P rates were not significantly related to the concentration of any runoff P fraction. This result is most likely explained by a combination of soil P sorption, plant P uptake and removal, runoff P loss with natural rainfall not measured during this interim period, and runoff P lost during the 24-h simulation (because simulated rainfall was applied to the same plot area for the 24-h and 6-mo events). The latter effect probably was not a significant contributing factor because calculations (not shown) indicated that the P lost during the 24-h event was <1% of the total P applied. The proportion of DRP and BAP in TPR also continued to decrease with time subsequent to nonincorporated manure application and was on average 7% for DRP and 15% for BAP for the 6-mo event. In similar fashion, the proportion of DRP in BAP for manure treatments decreased with time and was on average 51% (compared with 54% for controls) for the 6-mo event.

Figure 2 shows effects of manure P rate and incorporation on runoff P concentrations for the first manure application and subsequent rainfall simulations at Marshalltown. In general, the results at this site were similar to results at Boone with few exceptions that are noted herein. Runoff P concentrations (DRP, BAP, and TPR) increased linearly with increasing rates of manure P that was not incorporated for the 24-h runoff event. Incorporating the manure also decreased DRP, BAP, and TPR concentrations and decreased the impact of manure P rate. However, at this site the interaction between P rate and incorporation was significant (P < 0.05) only for DRP. There was slightly more variability at this site (models fit well but not as well as for the Boone site), and differences between incorporation and nonincorporation for BAP and TPR were substantial only for the highest P rate. Also, at this site the fraction of TPR made up of DRP or BAP for the nonincorporated or incorporated manure treatments did not follow the clear increasing pattern with increasing manure P rate observed at Boone. We found no clear explanation for this difference based on the measured manure and soil properties. However, this result might be explained by a greater residue cover at Marshalltown that determined a less intense manure P reaction with the soil. Soybean residue cover for the nonincorporated manure treatment was 97% for the Marshalltown site and 81% for the Boone site.


Figure 2
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 2. Effect of manure application rate, incorporation, and time of simulated rainfall on runoff P concentrations for the Marshalltown site in 2002. INC, manure incorporated; NOINC, manure not incorporated. NS indicates no interaction (P < 0.05) between P rate and incorporation.

 
Delaying rainfall at Marshalltown 15 d also decreased runoff P (Fig. 2), but the effect for the nonincorporated manure treatment was less pronounced than at Boone. The difference between the two sites cannot be explained with certainty, but it might be explained by decreased potential for soil P sorption due to much greater residue cover at Marshalltown. We expected P loss at Marshalltown to decrease more than the Boone site because soil P sorption is usually greater in finer texture soils (Udo and Uzu, 1972; Parffit, 1978; Mozaffari and Sims, 1994) and because higher organic matter reportedly coats P sorption sites (Huang and Schnitzer, 1986; Daly et al., 2001; McDowell and Condron, 2001). Additionally, simulated rainfall at Marshalltown was delayed 5 to 6 d longer than at Boone. The additional delay should have resulted in increased P sorption and larger differences between immediate and delayed simulated rainfall events for the Marshalltown site. As noted for the Boone site, DRP concentrations were nearly similar to BAP concentrations at both rainfall events. However, the ratio of DRP to BAP or of DRP and BAP to TPR for nonincorporated treatments at Marshalltown did not decrease with delayed rainfall, as was the case at Boone. This is most likely related to the relatively smaller effect of delayed rainfall on decreasing DRP and BAP at this site and to greater residue cover.

Delaying simulated rainfall for 6 mo at Marshalltown (Fig. 2) further reduced runoff P concentrations for the nonincorporated treatment compared with the 15-d runoff event. However, in contrast to the Boone site, manure P application rates were still correlated to DRP (P < 0.04) and BAP (P < 0.06) but not with TPR. The apparent contrast with the 6-mo event at Boone might be explained by the different interim season (summer at Boone and winter at Marshalltown) affecting soil temperature and lack of plant removal processes during winter at Marshalltown. The proportion of DRP and BAP in TPR decreased during this time subsequent to nonincorporated manure application and was on average 22% for DRP and 34% for BAP for the 6-mo event. Similarly, the proportion of DRP in BAP for manure treatments decreased during this time and was on average 64% (compared with 33% for controls).

Second Manure Application
Results from a rainfall simulation event within 24 h of a second manure application conducted in 2003 at Boone (Fig. 3 ) and Marshalltown (Fig. 4 ) in general were similar to results discussed for the first application. However, concentrations of runoff DRP, BAP, and TPR for the nonincorporated manure treatment were much higher than for the first manure application because the manure P rates were higher and were applied to the same plots. Manure P application sharply increased DRP, BAP, and TPR concentrations in runoff compared with the nonmanured plots when manure was not incorporated. When the manure was incorporated, runoff P concentrations were increased only at Marshalltown (Fig. 4), and the increases were very small. When there was a manure P effect, runoff P increased linearly except for DRP at Marshalltown, when the increase was better described with a curvilinear model. The proportion of DRP or BAP in TPR increased with increasing manure P rates but less than for the first manure application (not shown), and the difference was substantial at Marshalltown, where DRP was 31% of TPR for the lowest P rate and 61% for the highest rate. The proportion of BAP in TPR was approximately similar at both sites (not shown). We have no clear explanation for the observed inconsistency of manure P effects on the proportion of DRP in TPR across manure applications and sites. However, the high DRP proportion for the high manure P rate at Marshalltown is the result of a sharp (curvilinear) DRP increasing trend that was not observed for BAP or TPR. The TPR concentrations were highly related to TS concentrations in runoff when manure was incorporated at Marshalltown (P < 0.01; r2 = 0.49) and Boone (P < 0.06; r2 = 0.31). When the manure was not incorporated, TPR was not correlated with TS at either site, probably because of much less and inconsistent sediment loss compared with the incorporated treatment.


Figure 3
View larger version (18K):
[in this window]
[in a new window]

  		Fig. 3. Effect of manure application rate, incorporation, and time of simulated rainfall on runoff P concentrations for the Boone site in 2003. INC, manure incorporated; NOINC, manure not incorporated. NS indicates no interaction (P < 0.05) between P rate and incorporation.

 

Figure 4
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 4. Effect of manure application rate, incorporation, and time of simulated rainfall on runoff P concentrations for the Marshalltown site in 2003. INC, manure incorporated; NOINC, manure not incorporated. NS indicates no interaction (P < 0.05) between P rate and incorporation.

 
Delaying simulated rainfall for 6 mo after the second manure application reduced runoff concentrations substantially at both sites (Fig. 3 and 4). For example, compared with concentrations from the 24-h runoff event, DRP decreased by a factor of 25 at Boone and by a factor of 14 at Marshalltown. At both sites, DRP and BAP concentrations were linearly related to manure P rate, although the linear coefficients were smaller than for the previous runoff event. The TPR at both sites was much smaller compared with the 24-h event, and the manure P rate effect was not significant. Calculations from data in Fig. 3 and 4 indicated that the proportion of DRP and BAP in TPR decreased with time subsequent to application of nonincorporated manure. For example, compared with the 24-h event, the proportion of DRP and BAP decreased by factors of 1.6 and 1.9 at Boone and 1.2 and 1.3 at Marshalltown, respectively. The significant reduction in manure P effect for this 6-mo event might be explained by a combination of soil P sorption, plant P uptake and removal, and runoff P loss with natural rainfall that was not measured. In contrast to the first manure application, simulated rainfall for the 24-h and 6-mo events was applied to different plot areas. One notable result at Marshalltown not observed at Boone was that concentrations of DRP and BAP for the 6-mo event subsequent to the second application were lower than after the first application, even though the P rates were higher for the second application. This result might be explained by the different interim season for the 2 yr. There was crop P uptake, much higher precipitation (578 mm), and probably greater P sorption during the warmer summer during the second 6-mo period compared with no crop uptake, almost no precipitation (39 mm), and probably little soil P sorption during the cold winter and early spring of the first 6-mo period.

Runoff Phosphorus Loads
First Manure Application
Site properties that affect surface water flow may result in different P loads even when runoff P concentrations could be similar for some treatments or sites. Study of treatment effects on P loads within a site and rainfall event showed similar rankings for manure P rate and tillage effects to those shown for runoff P concentrations. This result is reasonable because differences in P loads across treatments within a site would parallel differences in runoff P concentrations when flow is similar across plots within a site. However, the treatment effects sometimes were different for P loads, and the statistical significance of sometimes large effects often was lower because of large flow variation. Large flow variation, even for a runoff event within a site, has been observed before, especially in rainfall simulation studies based on microplots, although Bundy et al. (2001) reported flow for 60 min of simulated rainfall with no pre-wetting.

The results of treatment effects on P loads for the four simulations conducted after the first manure application at Boone are shown in Fig. 5 . When manure was not incorporated, P loads in runoff for the rainfall event within 24 h of manure application increased linearly. Upon incorporation, P loads were much lower and increased at a much lower rate, especially for DRP and BAP. These trends mirror those observed for runoff P concentrations. Runoff quantity varied greatly across plots (not shown), but the variation was not consistent across treatments, and differences were not statistically significant (P < 0.05) between incorporated and nonincorporated manure or between application rates.


Figure 5
View larger version (23K):
[in this window]
[in a new window]

  		Fig. 5. Effect of manure application rate, incorporation, and time of simulated rainfall on runoff P loads for the Boone site in 2002. INC, manure incorporated; NOINC, manure not incorporated. NS indicates no interaction (P < 0.05) between P rate and incorporation.

 
When simulated rainfall was applied 15 d after manure application at Boone (Fig. 5), P loads decreased for all P rates compared with the 24-h event and did not differ (P < 0.05) for incorporated and nonincorporated treatments. However, data for DRP and BAP for the highest manure P rate suggest larger P loss for the nonincorporated treatment, which follows the trends shown for runoff P concentrations. In contrast to results for DRP and BAP loads and for TPR concentrations, neither tillage nor manure P rate influenced TPR loads. Delaying simulated rainfall for 6 mo resulted in smaller runoff P loads compared with previous runoff events for reasons suggested before (Fig. 5). Indeed, runoff volumes in manured plots were lower than control plots (1.1 times) for the 6-mo rainfall event, supporting similar reports that manure application (in the long term) can reduce surface runoff volume (Hensler et al., 1970; Mueller et al., 1984; Bundy et al., 2001; Gessel et al., 2004). Manure P and tillage effects were no longer significant.

Results for P loads after the first manure application at Marshalltown (Fig. 6 ) were somewhat different from results for runoff P concentrations at this site and from results for P loads at Boone. Two obvious differences observed for the 24-h and 15-d rainfall events were strong curvilinear trends for the P effects DRP, BAP, or TPR loads without incorporation of the manure and nonsignificant or very small linear manure P effects with incorporated manure. For the 24-h event, the incorporation effect on runoff P differed only for the highest manure P rate. Another obvious difference was that a 15-d delay in rainfall decreased runoff P concentrations significantly (Fig. 4) but did not decrease P loads for the nonincorporated manure P treatment compared with the 24-h event. Delaying simulated rainfall for 6 mo at Marshalltown resulted in substantially lower DRP, BAP, and TPR loads in runoff (Fig. 6). In contrast to results at Boone, runoff volume in manured plots at this site was not lower than for control plots. This could be due to the comparatively much lower quantity of runoff (7.5 times lower versus Boone) and the long-term history of no-till at Marshalltown.


Figure 6
View larger version (37K):
[in this window]
[in a new window]

  		Fig. 6. Effect of manure application rate, incorporation, and time of simulated rainfall on runoff P concentrations for the Marshalltown site in 2002. INC, manure incorporated; NOINC, manure not incorporated. NS indicates no interaction (P < 0.05) between P rate and incorporation.

 
The strong curvilinear trends and sustained high P loads for manure application without incorporation at Marshalltown for the 24-h or 15-d runoff events (Fig. 6) might be explained by a combination of less manure contact with soil due to a high residue cover and partial sealing of the soil, as was suggested by other research with liquid cattle or swine manure (Ross et al., 1979; Haraldsen and Sveistrup, 1996; Smith et al., 1998; Smith et al., 2001). Bundy et al. (2001) reported that surface sealing due to liquid swine manure application was short lived and that in the long term, manure additions decreased surface runoff volume by increasing aggregate stability and soil hydraulic conductivity. Ginting et al. (1998) reported that annual particulate P and total P losses were similar or lower from plots receiving solid beef manure than nonmanure plots.

Second Manure Application
Figure 7 shows treatment effects on P loads for the second manure application at Boone. Differences in P loads between nonincorporated and incorporated manure for the 24-h runoff event followed trends shown for concentrations. Although runoff P concentrations at a given P rate were comparable to those observed for the first manure application the previous year, the overall P load level and differences between incorporated and nonincorporated manure were larger than for the first application. This year, manure P application increased (P < 0.05) runoff P loads only when the manure was not incorporated. Runoff quantity for nonincorporated manure was on average 7.4 times greater than incorporated treatments (not shown) and increased with increasing rates of nonincorporated manure but not with incorporated manure. This result helps explain the higher P loads and higher rates of increase with increasing P rate this year. Also, this result supports the surface sealing effect of the very high liquid manure rate discussed previously. Delaying simulated rainfall for 6 mo reduced DRP, BAP, and TPR loads to the point that there were no significant tillage or manure P effects.


Figure 7
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 7. Effect of manure application rate, incorporation, and time of simulated rainfall on runoff P loads at the Boone site in 2003. INC, manure incorporated; NOINC, manure not incorporated. NS indicates no interaction (P < 0.05) between P rate and incorporation.

 
Figure 8 shows treatment effects on P loads for the second manure application at Marshalltown. Similar to results at Boone, manure P application without incorporation increased DRP, BAP, and TPR loads for the 24-h runoff event, and delaying runoff for 6 mo sharply reduced runoff P loads. However, there were some differences compared with trends observed for the previous year 24-h runoff event at this site and with trends at Boone. One difference was that curvilinear increasing trends for P loads with increasing rates of nonincorporated manure P were more obvious this year. Another difference was that this year, neither tillage (incorporation or not incorporation of manure) nor the interaction between manure P rate and tillage were significant (P < 0.05) for DRP and BAP loads, but the overall manure P rate effect was significant. Moreover, nonorthogonal comparisons (not shown) for both runoff fractions indicated a significant difference between manure incorporation vs. nonincorporation for the highest manure rate (larger for nonincorporated manure). Therefore, this year surface application of manure without incorporation resulted in higher P loss only for the very high manure rate, which was higher than rates producers usually apply when they base application of N or P needs for up to two crops. In contrast to results for Boone (Fig. 7), overall TPR loads were higher when manure was incorporated for the three lowest manure rates (P < 0.05) and were similar for incorporated and nonincorporated manure treatments for the highest P rate.


Figure 8
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 8. Effect of manure application rate, incorporation, and time of simulated rainfall on runoff P loads for the Marshalltown site in 2003. INC, manure incorporated; NOINC, manure not incorporated. NS indicates no interaction (P < 0.05) between P rate and incorporation.

 
Another interesting result from this year at Marshalltown relates to runoff quantity that greatly influenced runoff P loads, mainly TPR loads. As was the case with the 6-mo rainfall event after the previous year's manure application, runoff quantity was not lower in manured plots compared with controls (data not shown) probably due to reasons discussed previously. However, runoff quantity was on average five times greater in incorporated treatments than nonincorporated treatments (not shown), which was the opposite of observations previously discussed for the Boone site this same year. The large difference between the incorporation treatments at this site and the contrasting result compared with those at Boone might be explained by likely higher water infiltration for the nonincorporated (no tillage) treatment at Marshalltown determined by a history of no-till management. Also, average TPR loads for incorporated treatments compared with nonincorporated treatments were higher compared with those observed the previous year (Fig. 6). This is likely explained by greater residue cover in 2002 (64%) than in 2003 (40%), which led to 2.5 times greater sediment concentration in 2003.


   Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
REFERENCES
 
The results showed that P concentration and load in surface runoff within 24 h of applying liquid swine manure without incorporation into the soil increased linearly or exponentially with increasing P application rates. Incorporating the manure with tillage resulted in significantly lower runoff P concentrations and loads, and P rate effects were linear or sometimes not significant. On average for runoff events within 24 h of manure application across all P rates, sites, and years, DRP, BAP, and TPR concentrations were 7.6, 6.7, and 3.3 times higher, respectively, for nonincorporated manure than for incorporated manure, whereas runoff P loads for the same fractions were 13.8, 11.7, and 5.6 times higher. The proportions of DRP and BAP in TPR were greater for nonincorporated manure than for incorporated manure and became even greater as the P rate increased for nonincorporated manure but not for incorporated manure. On average for the 24-h runoff events, DRP and BAP concentrations were both 54% of TPR for nonincorporated manure, and 22 to 25% for incorporated manure, respectively. These results indicated that incorporation of manure decreased runoff P and the proportion of P forms readily available to algae in water bodies.

When runoff-causing rainfall was delayed 10 to 16 d after manure application without incorporation into the soil, runoff P concentrations and loads decreased sharply compared with runoff within 24 h of manure application. On average for all P rates, sites, and years, DRP, BAP, and TPR concentrations for the 10- to 16-d runoff events were 3.1, 2.7, and 1.1 times lower, respectively, whereas runoff P loads for the same fractions were 3.8, 3.6, and 1.6 times lower. These results indicate that even a short delay in runoff-causing rainfall can sharply decrease the risk of P loss from surface-applied liquid swine manure and that manure incorporation with tillage reduces P loss for runoff events immediately after application but not necessarily after 10 to 16 d. Delaying rainfall 6 mo, even at Marshalltown when no crop was grown during this period, reduced runoff P concentrations even further.

The results indicated a need for considering the probability of runoff-causing rainfall when addressing the risk of P loss from surface-applied liquid swine manure. Also, large variability for manure P rate and incorporation effects on short-term P loads within and across sites or seasons indicated the need for a better understanding of the influence of hydrologic factors at a small scale for developing effective management strategies to control P loss.


   ACKNOWLEDGMENTS
 
We thank Dr. James L. Baker for his advice for developing rainfall simulation methods, Carl H. Pederson for his assistance at applying manure and tillage treatments, and David and Brent Jacobson for allowing research work at their field and providing equipment. We also recognize partial funding support from the Agronomy Endowment (Iowa State University) and the Iowa Soybean Association.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
D. E. Kaiser, A. P. Mallarino, M. U. Haq, and B. L. Allen
Runoff Phosphorus Loss Immediately after Poultry Manure Application as Influenced by the Application Rate and Tillage
J. Environ. Qual., January 13, 2009; 38(1): 299 - 308.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
D. J. Wagner, H. A. Elliott, R. C. Brandt, and D. Jaiswal
Managing Biosolids Runoff Phosphorus Using Buffer Strips Enhanced with Drinking Water Treatment Residuals
J. Environ. Qual., June 23, 2008; 37(4): 1567 - 1574.
[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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Agricola
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Related Collections
Right arrow Phosphorus
Right arrow Runoff
Right arrow Animal Waste

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
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome