| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Soil Science, Univ. of Wisconsin, 1525 Observatory Dr., Madison, WI 53706
b USDA-ARS Dairy Forage Research Ctr., 1925 Linden Dr. West, Madison, WI 53706
* Corresponding author (grmunoz{at}uwalumni.com)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Excessive soil nutrient accumulation and losses to surface and ground water are pressing environmental challenges facing the dairy and other animal industries. As dairy herds expand to remain economically viable, a larger percentage of the available cropland is devoted to corn silage. The noted expansion of corn silage production (Battaglia, 1999; Shaver, 2000) is due to this crop's ability to feed more cows (Bos taurus) than other forages per unit of cultivated area (Seglar, 1998), as well as favorable economics (Klemme, 1998) to the farmer. However, the effects of shifting more land to corn silage on other system components, such as N use, buildup and loss remains to be determined. Since only a relatively small amount of applied N is ultimately taken up by the crop, we wanted to track the fate of the unused portion to see whether it was lost or remained in the soil.
The objective of this study was to determine total and inorganic soil N and the N balance of a continuous corn silage cropping system receiving two fertilizer or dairy manure N rates of different application frequency across 3 yr. Unlabeled and 15N-enriched dairy manure were used, and the ability of each manure type to detect trends in soil N levels and account for applied N was compared. The use of 15N-labeled manure was an essential part of the study because it allowed direct N tracking in the cropping system, and provided more accurate measurements than unlabeled manure.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Before experiment initiation, the field was cropped to alfalfa (Medicago sativa L.) from 1994 to 1996 and to corn in 1997. No manure had been applied for at least 4 yr before the start of the trial. Treatments consisted of two inorganic fertilizer N levels (90 or 179 kg ha-1, as NH4NO3), two manure rates (estimated to provide 
90 and 180 kg available N ha-1 to corn the first year following application), a control receiving neither fertilizer nor manure, and three manure application intervals (every 1, 2, or 3 yr). The same manure N rate was used every year it was applied. Fertilizer N was applied every year to the same plots. The design of the field trial was a split plot, with fertilizer treatments and manure rates applied to whole plots. Application intervals were the subplots within the manure whole plots. Nitrogen rates were applied to subplots in the fertilizer whole plots. Whole and subplots (henceforth referred to as plots) were arranged in four randomized-complete blocks to give four replications of each treatment. The plots were 10.6 by 6 m, separated by 1.5-m alleys and contained eight corn rows that were 0.75 m apart. For the 15N experiment, microplots of 1.5 by 2.3 m containing three corn rows were established within each of the low manure rate plots, following the design proposed by Jokela and Randall (1987). Manure applied to these microplots in 1998, 1999, and 2000 had an atom % 15N of 1.47, 1.12, and 1.64, respectively.
Fresh dairy manure (composite of feces, urine, and straw bedding) was collected from a stockpile. Manure for the 15N microplots was labeled by feeding cows 15N-enriched silage and alfalfa following the procedure described by Powell and Wu (1999) to obtain labeled urine and feces having uniformly-labeled microbial and undigested feed N components. Fertilizer and manure were applied 5 d before planting. The field was disked twice within 3 to 20 h after manure application. Manure amounts and manure N and NH4–N applied each year are presented in Table 1. Manure dry matter content was 210, 260, and 240 g kg-1 for 1999 through 2000, respectively. Manure enriched with 15N had a dry matter content of 170 g kg-1 all 3 yr. Corn (cv. Lemke 6063) was planted every spring. Starter fertilizer (9-23-30, 224 kg ha-1 in 1998 and 1999, and 168 kg ha-1 in 2000) was band-applied to all plots at planting.
|
In fall 2000, all 15N microplots were systematically sampled 25 cm apart from the center of the microplot in every direction (i.e., two cores from the central row and two cores from between rows) using a plywood template. Samples were processed and analyzed similarly to those from the main plots. In addition, 15N enrichment of soil total- and inorganic-N fractions was determined. For 15N enrichment analysis, soil subsamples were hand-ground in a ceramic mortar and sieved to pass a 100-µm mesh screen.
Chemical Analyses
Manure N content was analyzed following the procedures outlined by Combs et al. (2001). Total soil N was determined following a Kjeldahl digestion (Nelson and Sommers, 1972) with these modifications: 1.5 g of Kjeldahl mix (Na2SO4–CuSO4–Se in a ratio of 1000:32:5) and 5 mL concentrated H2SO4 were used. Our soil typically contained <1% of total N as NO3–N, therefore, a Kjeldahl digestion provided an adequate measure of total N (Bremner, 1996). For the topsoil (0 to 30 cm), a 1-g subsample was digested, and a 2-g subsample was used for soil taken from lower (30- to 60- and 60- to 90-cm) depths. The digest was diluted to 50 mL, filtered through acid-washed Whatman no. 2 filter paper and analyzed for NH4–N in an automated colorimeter (Lachat Instruments, Milwaukee, WI) using the QuikChem Method 13-107-06-2-D (Lachat Instruments, 1992b), with sodium phenate, and 5.2% sodium hypochlorite. Total N in plant tissue was determined following a similar procedure on 250-mg samples.
Soil NH4– and NO3–N were determined according to a modification of the procedure described by Liegel et al. (1980). The KCl extract was filtered through Whatman no. 2 paper and analyzed for NH4–N following the same procedure already described, and for NO3–N using the QuikChem Method 12-107-04-1-B (Lachat Instruments, 1992a).
For soil inorganic (NO3 plus NH4) 15N enrichment, KCl extracts were treated following the microdiffusion technique described by Stark and Hart (1996). Total N and 15N concentrations in soil, corn tissue, and manure samples from 15N microplots were determined using a Carlo Erba (Milan, Italy) elemental analyzer coupled with a mass spectrometer Europa 20/20 tracermass.
Statistical Analyses
Statistical analyses were performed using SAS software (SAS Institute, 1990). When there were soil N measurements repeated in time, ANOVA was conducted across sampling times and treatments. Given the differences known to exist from one cropping season to the next, which were usually reflected in the statistical analyses, ANOVA was performed by sampling time as well, to distinguish between treatments. Depth was acknowledged to have a major influence on soil characteristics and hence N levels; therefore, ANOVA was performed by depth. Polynomial models were fitted to soil N levels vs. sampling time (each time was assigned a number from 1 to 6), and the best fit was chosen between first- and second-order polynomials.
Calculations
Cumulative N recovery in corn was calculated by the difference method:
 | [1] |
Total soil N to a depth of 90 cm, harvested whole-plant N uptake, and manure N additions were used to compute soil–crop N balances. Nitrogen recovered in the soil (% N recovsoil) after 3 yr of fertilizer or manure applications was calculated according to:
 | [2] |
|
 | [3] |
Similarly, recovery of 15N in soil was calculated as:
 | [4] |
Total N recovered in soil plus harvested corn was calculated for both N and 15N as:
 | [5] |
Nitrogen losses were not directly measured in this experiment. Unaccounted for N was assumed to be lost by NH3 volatilization, denitrification, and/or leaching (mainly between cropping seasons). Where appropriate, soil total- or NO3–N concentrations in mg kg-1 were converted to kg ha-1 by assuming a soil bulk density of 1.3 g cm-3.
An estimation of soil NO3–N changes due to manure applications was obtained by monitoring soil 15N concentrations in the inorganic-N (NH4 plus NO3) fraction in subplots amended with 15N-labeled manure. For these calculations, 15N enrichment was assumed to be similar for NH4– and NO3–N, and therefore equal to that of the inorganic fraction. Soil 15NO3–N increase was entirely due to manure, and it equaled the 15NO3–N excess (over natural abundance) measured after applications, since initial 15N excess was zero. Percentage 15N recovered in soil by the end of the third year as 15NO3–N was calculated in a manner similar to Eq. [3] and [4]:
 | [6] |
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Soil NO3–N levels at each sampling depth and time are presented in Table 2 for the controls and plots receiving annual inorganic fertilizer or manure applications. Statistical analyses were done on each soil layer separately. There were significant effects of treatment, time, and treatment x time on soil NO3–N levels at all depths, except for the 30- to 60-cm layer, where the interaction was not significant. Probability values were <0.001 in all cases except for treatment x time at the deepest layer (P = 0.02). Statistical analyses of soil NO3–N levels by time and depth (Table 2) showed significant differences among treatments in the topsoil (0 to 30 cm) after the second cropping season (i.e., after two fertilizer or manure applications); hence, the following observations refer to those sampling times. The lowest and highest NO3–N levels always corresponded to the control and the high manure rate, respectively. Differences tended to become more pronounced with time. There was usually no difference in soil NO3–N levels in plots amended with the low manure rate or either fertilizer rate. Manure at the high rate significantly increased NO3–N compared with the low rate and the control.
|
Significant differences in soil NO3–N concentrations at the 60- to 90-cm depth were also detected after the first cropping season, except for spring 1999. Plots that received fertilizer at the high rate had significantly higher NO3–N levels than the control, which always had the lowest values. Soil NO3–N in plots receiving the high manure rate was higher than the controls in spring 2000 only. Plots fertilized at the low rate had higher NO3–N levels than the control in fall 1999.
Both fertilizer and manure additions tended to increase soil NO3–N compared with the control. Although the fertilizer effect on topsoil NO3–N levels was lower than that of manure, fertilizer increased NO3–N concentrations in lower soil depths indicating that more fertilizer- than manure-N moved downward as NO3–N during the growing season. This difference in behavior was probably due to the fact that NO3–N applied as inorganic fertilizer is immediately solubilized in soil and therefore more susceptible to downward movement within the soil profile if there is no crop to utilize it. More than half of manure N, on the other hand, is in organic forms, and virtually all of the rest is present initially as NH4–N. Hence, manure organic N has to be mineralized and nitrified before it becomes susceptible to leaching.
Another interesting trend observed in Table 2 is the almost invariably higher NO3–N concentrations observed in the spring than in the fall, at the 30- to 60- and 60- to 90-cm soil depths. This was likely due to downward movement of some N during the previous winter and early spring, whereas the fall sampling reflects a decrease in the NO3–N pool during the cropping season by plant uptake, and perhaps leaching; however, previous research on similar soils has shown little leaching during the growing season (Olsen et al., 1970; Kelling et al., 1977). Topsoil (0 to 30 cm) NO3–N concentrations did not follow the same pattern as at lower soil depths, probably because of continual mineralization and subsequent nitrification in the upper soil profile.
There was a clear trend, described by linear regression, toward increased NO3–N concentrations across time in the topsoil for both manure rates and for the high fertilizer rate. Equations and statistics are presented in Table 3. Regressions were not significant for the deeper soil layers. The slope for the high manure rate was significantly greater than those of the other treatments (P < 0.0001), indicating that NO3–N tended to accumulate to a greater degree in manured than fertilized plots. This also appears to support the argument for greater short-term leaching potential from the fertilizer. Comfort et al. (1987) found that high manure rates did not significantly increase inorganic N below 30 cm in the first application year, whereas fertilizer N did, and Kimble et al. (1972) measured more NO3–N available for leaching in fertilized than manured plots. Jokela (1992) found that manure had the same, or slightly lower leaching potential than an agronomically equivalent fertilizer rate.
|
|
Total Soil Nitrogen
Total soil N concentrations by depth at each of the six sampling times are shown in Table 4. As would be expected, total N concentrations decreased with soil depth. Regression analyses were performed to describe the effect of repeated N applications on total soil N (Mg ha-1) across time for the 0- to 90-cm depth. Soil N levels in control plots did not change during the 3-yr study period (P = 0.89), averaging 13.5 Mg ha-1. Fertilizer and manure treatments significantly increased total soil N during the study. Soil N increases due to fertilizer were best described by linear regression, whereas those due to manure were best described by second order polynomials. Regression equations, statistics, and graphics are presented in Fig. 1.
Net total soil N increases in plots that received the low and high manure rates, based on regression analyses, were 2.0 and 2.9 Mg ha-1, respectively, across the 3-yr study period. For fertilizer, these values were 1.9 and 2.0 Mg ha-1. It is clear that these measurements are not sufficiently accurate, since they predict soil N increases much higher than the total N applied (0.7 and 1.4 Mg ha-1 for manure, and 0.3 and 0.5 Mg ha-1 for fertilizer). It should be noted that, for most treatments, the last soil sample taken showed much higher N levels than the previous five. This greatly influenced our estimation of total soil N at the end of the third study year. If the fall 2000 soil samples were anomalously high, it is possible that additional observations may rectify our estimations.
Use of 15Nitrogen-labeled Manure
Soil Nitrate-Nitrogen
According to soil 15NO3–N measurements, soil NO3–N increases due to 3 yr of manure application ranged from 0.1 to 6.7 kg ha-1 (Fig. 2)
. Most (74 to 94%) of this NO3–N was found in the upper 30 cm of soil. On average, more NO3–N was found in the 60- to 90-cm than in the 30- to 60-cm depth, but the difference was not significant and was likely due to high variability and/or the movement of N from previous year applications. Statistical analyses revealed that manure application interval had a significant effect on NO3 levels both in the 0- to 30-cm depth and the entire 0- to 90-cm sampled depth (P = 0.006 and 0.060, respectively). The greatest amount of soil NO3–N was found in plots receiving more frequent or recent manure applications (Fig. 2). This likely reflected the higher N loads resulting from repeated manure applications, and higher crop uptake and losses from manure that remained in the soil for a longer period. Treatment did not have an effect on 15NO3–N in the 30- to 60-cm depth. The only significant difference found at the 60- to 90-cm layer was for plots continuously manured, which had concentrations significantly higher than any other treatment. Manure increased NO3–N levels in the subsoil only in plots that received three consecutive manure applications, and only to a small degree. These data confirmed the large plot trends (Table 2), that manure applied at the low rate had a low leaching potential during this 3-yr period. Other researchers conducting direct 15N leaching measurements from manure showed leaching losses that ranged from <0.3 to 4% after 2 yr, with most of the leaching occurring during the year of application (Sørensen et al., 1994; Sørensen and Jensen, 1998). However, NO3–N leaching is likely to be underestimated by 15N experiments. Sørensen et al. (1994) measured much higher total NO3–N than 15NO3 leaching losses. Exchange of manure 15N with soil N will have an effect of lowering 15NO3–N concentrations until equilibrium is reached. In addition, although we did not follow 15NO3–N changes with time, it is clear that the more frequent the manure applications, the greater the soil NO3–N increase. It is realistic to infer that manure could impact N leaching in the future, especially on plots continuously manured.
|
|
Soil Nitrogen Balance
Manure and fertilizer N applications, crop N uptake, and differences in total soil N at the onset and end of the 3-yr trial were used to compute soil N balances for each treatment. According to the difference method (Eq. [1]), only 5 to 6% of total applied manure N was recovered in total corn dry matter by the end of the third cropping season, vs. 46 to 51% for applied fertilizer N (Table 5). Lower apparent recovery of manure than fertilizer N was mostly due to greater amounts of manure than fertilizer N applied, and to a lesser extent, somewhat lower corn N recovery in plots amended with manure than fertilizer. The apparent relative amount of fertilizer and manure N recovered in the soil (Eq. [2]) was two to seven times greater than N applications. Such gains in total soil N over and above N applications cannot be considered a rational result. The most logical explanation for these high estimates of soil N increases is that the fertilizer and manure N inputs represented only 1 to 10% of the basal soil N levels, which is less than the accuracy of the sampling and analytical methods used. The sampling and normal field variation likely further increased the uncertainty of the determinations.
|
|
Approximately one-half to two-thirds of applied manure N was recovered in soil (0 to 90 cm), with one exception. Only 24% of the manure 15N applied in 1998 was recovered in soil. Nitrogen recovered in the soil probably included slowly-decomposing and recalcitrant fractions of manure (undigested feed N in feces, which accounts for 20% of fecal N excreted by dairy cows; Powell and Wu, 1999), and manure N that was incorporated into new microbial biomass. Although our experiment did not allow us to corroborate this, it would be possible to label individual manure pools with 15N and follow their fate within the soil-crop system.
The effects of year and frequency of manure application on the amount of applied 15N recovered in the soil were not statistically significant. An average of 80% of the total 15N measured in the soil was present in the 0- to 30-cm depth, with 13 and 6% in the 30- to 60- and 60- to 90-cm depths, respectively. This suggests either relatively little downward movement of applied manure N, or that leached N may have moved out of the 0- to 90-cm layer. Nitrogen increases below 30 cm must have come from NO3–N movement to lower soil depths and subsequent immobilization, mixing by soil fauna, or possibly from roots. Depth differences in 15N recovery were statistically significant (P < 0.001), with highest recoveries obtained from the top 0- to 30-cm depth (38% of applied 15N). No differences in 15N recovery were observed between the 30- to 60- (6%) and 60- to 90-cm (2%) depths. Other studies have recovered a greater proportion of applied 15N in soil, such as 76 to 83% from the 0- to 10-cm depth, and 11 to 19% from the 10- to 45-cm depth (Sørensen et al., 1994). Sørensen and Jensen (1998) recovered 77.5% of applied manure (sheep) 15N from 0- to 15-cm depth and 1.6% from 15 to 30 cm. Jensen et al. (1999) recovered 39% of applied sheep manure 15N in the top 25 cm of soil. As discussed previously, higher N recoveries in these studies might have been due to rapid manure incorporation and subsequent lower volatilization losses. For all studies where 15N was measured at different depths, most of it was recovered in the top 10 to 15 cm. Sommerfeldt et al. (1988) also found manure to affect total N and organic matter to a depth of 30 cm only.
In plots receiving manure in 1998 only, the highest amount of 15N (59%) could not be accounted for. Actual N losses were not measured, but a longer period of time elapsed between spreading and incorporation of the manure in 1998 (20 h) than during the other two study years (2 h), possibly allowing for greater NH3 volatilization. Our measurements do not seem to indicate much leaching (Table 2 and Fig. 2), although some losses via this pathway may have occurred. Unaccounted-for 15N (36% on average) was probably lost mainly through NH3 volatilization and denitrification. Denitrification losses have been estimated to range from 0.2 to 7.1% of incorporated dairy manure N (Goodroad et al., 1984; Lessard et al., 1996) with usually higher losses (up to 26%) for slurries (Thompson et al., 1987; Paul and Zebarth, 1997a,b). Ammonia volatilization losses as high as 61 to 99% of the N applied as NH4 have been measured for broadcast manure in 5 to 25 d (Lauer et al., 1976). Liquid dairy manure can lose 24 to 33% of its NH4–N in 6 to 7 d after being disked (Beauchamp et al., 1982). Up to 40% of NH4–N can be volatilized, even within a few hours (Meisinger and Jokela, 2000).
It is apparent that soil N balances based on 15N measurements are less variable and, therefore, likely more reliable than those based on unlabeled N. The use of 15N labeled dairy manure provided direct measurements of manure N in the crop and soil system. As long as manure 15N enrichment is high enough to be detected, analyses are more precise. The drawback of using 15N-labeled manure is that the experimental setup and the analyses are more expensive and time consuming, and require significantly more careful sample preparation.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Use of 15N-enriched manure confirmed the trends in vertical distribution of soil N observed with unlabeled manure. It also showed that increases in soil N were higher in plots receiving more frequent or recent manure additions. Clearly, 15N-labeled manure provided more accurate measurements of total soil N and estimates of soil N balance. Using 15N-labeled manure, 18% of applied N was recovered in harvested corn. Approximately 46% of applied manure 15N remained in soil, and 36% could not be accounted for. Most of these losses were probably due to NH3 volatilization between the time of manure field application and soil incorporation, and denitrification throughout the cropping season. Most of the N remaining in soil (an average of 82% of that recovered) was present in the top 30 cm, irrespective of frequency of manure application.
The results obtained in this experiment showed that, although more costly and time-consuming, 15N studies can be a valuable tool in dairy manure research. This method proved to be a better approach than unlabeled manure to study the fate of manure N within the soil-crop system.
|
ACKNOWLEDGMENTS |
|---|
Received for publication February 22, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Bosshard, E. Frossard, D. Dubois, P. Mader, I. Manolov, and A. Oberson Incorporation of Nitrogen-15-Labeled Amendments into Physically Separated Soil Organic Matter Fractions Soil Sci. Soc. Am. J., May 29, 2008; 72(4): 949 - 959. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Nyiraneza and S. Snapp Integrated Management of Inorganic and Organic Nitrogen and Efficiency in Potato Systems Soil Sci. Soc. Am. J., August 9, 2007; 71(5): 1508 - 1515. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Sorensen and I. K. Thomsen Separation of Pig Slurry and Plant Utilization and Loss of Nitrogen-15-labeled Slurry Nitrogen Soil Sci. Soc. Am. J., August 25, 2005; 69(5): 1644 - 1651. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. H. Misselbrook, J. M. Powell, G. A. Broderick, and J. H. Grabber Dietary Manipulation in Dairy Cattle: Laboratory Experiments to Assess the Influence on Ammonia Emissions J Dairy Sci, May 1, 2005; 88(5): 1765 - 1777. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Powell, Z. Wu, K. Kelling, P. Cusick, and G. Munoz Differential Nitrogen-15 Labeling of Dairy Manure Components for Nitrogen Cycling Studies Agron. J., March 1, 2004; 96(2): 433 - 441. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Vadose Zone Journal | |||||
Fertilizer rates were 90 and 179 kg N ha-1 yr-1 for the low and high levels, respectively. 
3-yr average manure rates were 236 and 459 kg total N ha-1 yr-1 for the low and high levels, respectively. Regression equations: Fertilizer high, y = 12.88 + 0.4004x, R2 = 0.743, P = 0.027; Fertilizer low, y = 12.76 + 0.3775x, R2 = 0.637, P = 0.057; Manure high, y = 14.48 - 1.241x + 0.2595x2, R2 = 0.912, P = 0.026; Manure low, y = 14.83 - 2.115x + 0.3579x2, R2 = 0.906, P = 0.029; Control, y = 13.53.
