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


     

Published online 1 May 2008
Published in J Environ Qual 37:944-955 (2008)
DOI: 10.2134/jeq2007.0219
© 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 Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Muñoz, G. R.
Right arrow Articles by Zhu, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Muñoz, G. R.
Right arrow Articles by Zhu, J.
Agricola
Right arrow Articles by Muñoz, G. R.
Right arrow Articles by Zhu, J.
Related Collections
Right arrow Best Management Practices
Right arrow Nutrient Management
Right arrow Soil Fertility and Productivity
Right arrow Animal Waste

TECHNICAL REPORTS

Waste Management

Field Evaluation of Nitrogen Availability from Fresh and Composted Manure

Gabriela R. Muñoza, Keith A. Kellinga,*, Karen E. Rylantb and Jun Zhua

a Dep. of Soil Science, Univ. of Wisconsin-Madison, 1525 Observatory Dr., Madison, WI 53706
b formerly TVA Environmental Research Ctr., Muscle Shoals, AL

* Corresponding author (kkelling{at}wisc.edu).

Received for publication May 3, 2007.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
Prediction of manure nitrogen availability to crops is key to ensuring adequate production while minimizing potential adverse environmental impacts. This field study estimated first-year and residual N availability from several manures subjected to various levels of processing, and evaluated the applicability of the presidedress soil N test (PSNT) in fields receiving the different manures using corn (Zea mays L.) as the test crop. Plots received several rates of fresh (FP), dried (DP), or composted (CP) poultry (Gallus gallus domesticus) manure, composted cow (Bos taurus) (CC) manure, ammonium nitrate (AN), or no N. Crop yields and N uptake from plots where CC was applied were undistinguishable from controls in most years, whereas poultry manures significantly increased corn production. Average apparent first-year N availability, as measured by fertilizer equivalence, was 57, 53, 14, and 4% for FP, DP, CP, and CC respectively. Apparent second-year N availability, as measured by relative effectiveness, was 18, 19, 12, and 7% for FP, DP, CP, and CC; however, for CC both first- and second-year estimates of apparent N recovery (ANR) could statistically not be separated from the controls. Apparent nitrogen avail-ability was greater for less processed manures and for CP compared to CC, emphasizing that producers should know the source and level of compost stability when these materials are used as a primary nutrient source. The PSNT successfully (87% correct) identified sites with a critical value of 24 mg kg–1 that were N sufficient across a variety of N amendments from those that would have benefitted from additional N input.

Abbreviations: AN, ammonium nitrate • ANOVA, analysis of variance • ANR, apparent N recovery • CC, composted cow manure • CP, composted poultry manure • DP, dried poultry manure • FE, fertilizer equivalence • FP, fresh poultry manure • GY, grain yield • PSNT, presidedress soil N test • RCBD, randomized complete block design • RelGY, relative grain yield • WPNU, whole-plant N uptake • WPY, whole-plant yield


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results and Discussion
 Conclusions
 REFERENCES
 
ANIMAL manures are widely used as N sources in agriculture. Manure N content and availability depend on several factors, including animal species, feed, bedding and confinement system, as well as type of manure storage and handling, including the amount of processing to which the manure has been subjected (Gilbertson et al., 1979; Smith and Peterson, 1982; Eghball et al., 1997).

Composting provides several advantages over fresh manure including reduction in volume, weight, odor, and fly-breeding potential, and most weed seeds and pathogens are destroyed (Rynk et al., 1992; Paul and Beauchamp, 1993; DeLuca and DeLuca, 1997; Larney and Blackshaw, 2003; Van Herk et al., 2004). Furthermore, during composting, although the concentration of some nutrients such as N and P may remain relatively constant, their form changes and typically the nutrients are stabilized, resulting in the potential for reduced nutrient losses during storage or after application (Hébert et al., 1991; Basso and Ritchie, 2005; Larney et al., 2006), and compost is approved in organic farming, whereas in some situations, fresh manure is not. A major shortcoming of composting is nutrient loss during composting, leading to reduced fertilizer value and possible point-source pollution (Hansen et al., 1993; Eghball et al., 1997; Tiquia et al., 2002; Hao et al., 2004). Of particular concern are N losses mainly through NH3 volatilization, although runoff, leaching, and denitrification can also occur. Measures of N volatilization during poultry manure composting range from 17 to 63% (Bonazzi et al., 1990; Mahimairaja et al., 1994; Kithome et al., 1999; Hao et al., 2004). Martins and Dewes (1992) measured losses of initial N from 47 to 77%, primarily as NH3, when mixtures of straw and swine, poultry, or cattle liquid manure were composted in containers. These and other studies observed NH4, P, and K losses in compost runoff to be high enough to potentially pollute surface waters, and Na and K to negatively affect soil structure (Ott et al., 1983; Nienaber and Ferguson, 1994; Eghball et al., 1997).

Prediction of manure or manure compost N availability to crops is key to ensuring adequate nutrient supply to maximize yields while avoiding over-application. Several researchers have shown that composted manure first-year N availability is typically about one-third to one-half that of fresh manure (Castellanos and Pratt, 1981; Brinton, 1985; Schlegel, 1992; Paul and Beauchamp, 1993; Wen et al., 1995) due to some losses of the readily available N before and during composting and much of the remaining N converting into chemical forms that are more stable than those originally present. The overall result is that N is less available in compost than fresh manures with the estimates of composted manure first-year N availability ranging from 2 to 35% (Kirchmann, 1990; Schlegel, 1992; Paul and Beauchamp, 1994; Hadas and Portnoy, 1994; Mahimairaja et al., 1995; Eghball and Power, 1999b; Preusch et al., 2002; Dao and Cavigelli, 2003; Miller et al., 2004; Larney et al., 2006). In spite of the number of studies that have evaluated composted manure N availability, only a few (Castellanos and Pratt, 1981; Kirchmann, 1990; Fauci and Dick, 1994; Mahimairaja et al., 1995; Warman and Cooper, 2000; Preusch et al., 2002) included poultry manure or broiler litter compost, and of these latter studies, only the Mahimairaja et al. (1995) and Warman and Cooper (2000) studies were done in the field, and none evaluated second-year or residual availability.

The soil presidedress soil N test (PSNT), measuring soil NO3–N in the 0- to 30-cm layer before the rapid growth period of corn, is a tool that has the potential to estimate N availability in a variety of soil environments (Magdoff et al., 1984, 1990). It provides for an in situ incubation, allowing consideration of all site-specific variables affecting N mineralization (Magdoff, 1991). This test can differentiate between sites likely to respond to sidedress N application, and sites with no need for further fertilization. It is less useful, however, for predicting the optimum N rate in sites requiring additional N (Klausner et al., 1993; Heckman et al., 1996). The generally lower N rates and the split N applications resulting from the use of PSNT translate into lower residual N and thus lower potential for N leaching after harvest (Durieux et al., 1995; Guillard et al., 1999). In addition to environmental benefits, PSNT use can increase profits in corn production (Babcock and Blackmer, 1992; Musser et al., 1995).

The objectives of this study were to estimate and compare first-year and residual N availability or N recovery by corn (Zea mays L.) from fresh, dried, and composted poultry manures, and composted cow manure in a 3-yr field experiment, and to evaluate the applicability of the PSNT in fields receiving various rates and kinds of manure.


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results and Discussion
 Conclusions
 REFERENCES
 
Experiment Layout and Sampling
A field trial was conducted from 1992 to 1995 at the Univ. of Wisconsin Agricultural Research Station, Arlington, WI (43°17' N, 89°22' E) on Plano silt loam soils (fine-silty, mixed, mesic, Typic Argiudolls). The fields had been in corn for 3 to 9 yr before the establishment of the plots, and were not part of any experiment the year before, when they received either a moderate N rate (90 kg ha–1) or no N fertilizer.

Treatments consisted of five N sources and four N rates. The N sources included NH4NO3 (AN) and four manure sources, namely fresh poultry (Gallus gallus domesticus) (FP), dried poultry (DP), composted poultry (CP), and composted cow (Bos taurus) (CC). Each N source was applied at four rates intended to provide approximately 67, 134, 202, and 269 kg available N ha–1 based on initially expected availability of 60% for the FP and DP manures and 25% availability for the composts (Kelling et al., 1998). These values were adjusted somewhat in subsequent years. Controls receiving no N were also included. Treatments were arranged in a randomized complete block design (RCBD) with four replications. Each year, a new portion of the field was used, and the treatments were re-randomized.

Each plot contained four rows of corn and was 3.0 m wide by 10.7 m long, except in 1992 when they were 12.2 m long. Fresh and dried poultry manures were obtained from an egg-laying operation near Lake Mills, WI. The fresh manure was taken directly from the laying houses, whereas the dried manure was air-dried on the floor of a covered drying shed and then passed through a flash drier (48°C for 10 min) and a pellet-forming extruder. Composted poultry manure was produced by a covered windrow operation from manure obtained from local broiler operations in Muscle Shoals, AL; however, compost production problems prevented this material being available until 1993. The composted cow manure was from an uncovered windrowing commercial composting operation using straw-bedded dairy manure in Waupaca, WI, with a labeled fertilizer value of (0.5-0-0). The actual manure N and moisture concentrations were not known until after application; therefore, the N rates varied among N sources and years, and somewhat differed from the intended rates. Table 1 shows manure analyses and the lowest actual total manure N rates from each source used in each year. The higher rates were two, three, and four times these amounts. Treatments (manures and ammonium nitrate) were hand-applied after weighing by the first week of May each year, followed by disking twice within 24 h of applying the manure. Corn (DeKalb 547 in 1992 and Pioneer 3578 thereafter) was planted 5 to 10 d after treatment applications. Preplant soil test P and K were in the optimum or high range (Kelling et al., 1998) on all fields; therefore, only a uniform application of starter fertilizer (168 kg ha–1 of 6-24-24, 5 by 5 cm to the side and below the seed) was made to all plots at planting. Plots were thinned to uniform plant populations 40 to 60 d after planting. Final counts were 50,483, 47,606, and 56,121 plants ha–1 in 1992, 1993, and 1994, respectively. Standard pest management practices were used, including preemergence herbicide, in-row corn rootworm insecticide, and one cultivation when the corn was approximately 30 cm tall. Corn aboveground whole plants were harvested in late September or early October. Ten randomly selected plants from the two center rows of each plot were hand-cut 10 cm above the ground, chopped, and weighed. Corn grain was harvested using a plot combine on the two middle rows in late October or early November, with adjustment made for the removed whole plants. Whole-plant yields (WPY) are reported on a dry-weight basis, and grain yields (GY) at 15.5% moisture. Temperature and precipitation data for each growing season are provided in Table 2 .


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

  		Table 1. Chemical analyses and rates of manures applied in 1992 to 1994 to experimental plots at Arlington, WI at the lowest rate.{dagger}

 

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

  		Table 2. Growing season monthly average temperatures and precipitation and deviations from 30-yr averages for 1992 to 1995 at Arlington, WI.{dagger}

 
By first year, we refer to indices or properties measured the same year N fertilizer or manure was applied. By second year or residual, we refer to measurements done the year following N application. For the residual part of the study, from 1993 to 1995, corn (Pioneer 3578) was planted to the same plots that had received manure or AN the previous year. No additional N was applied, except the same amount of starter fertilizer was used at planting. Pest management, planting, and harvesting procedures were as described above. Plant populations after thinning were 54,896, 56,586, and 54,374 plants ha–1 in 1993, 1994, and 1995, respectively.

Soil samples to 30 cm (five cores per plot) were collected from all plots when corn plants were 15 to 30 cm tall to evaluate the effectiveness of the PSNT in predicting the need for N in addition to the pre-plant application.

Chemical Analyses
Subsamples of the harvested whole-plant and grain materials were oven-dried (55°C, 10 to 14 d) to determine dry matter, and ground to pass a 2-mm screen. Total N in the samples was measured in an automated colorimeter (Lachat Instruments, Milwaukee, WI) using the QuikChem method 13-107-06-2-D (Lachat Instruments, 1992a) following a semi-micro Kjeldahl digestion (Liegel et al., 1980). Manure subsamples were analyzed for dry matter, NH4–N and total N by the UW Soil and Forage Analysis Laboratory, Marshfield, WI (Combs et al., 2001). Soil samples collected for the PSNT were extracted with a KCl solution (Liegel et al., 1980) and analyzed for NO3–N in the Lachat colorimeter following the QuikChem method 12-107-04-1-B (Lachat Instruments, 1992b).

Statistical Analysis and Calculations
Whole-plant N uptake (WPNU), WPY, and GY were analyzed as a RCBD using proc mixed in SAS (SAS Institute, 1990). The blocking factors were the 3 yr and the four replications nested within each year, whereas the treatment factors were the five N sources, the four nominal N rates, and controls. The overall significance of the treatment effects on each of the three crop parameters (WPY, WPNU, or GY) was assessed using F-tests for first- and second-year measurements. If the overall treatment effects were significant at {alpha} = 0.10, further contrasts among the treatments were performed and least significant differences were used for all pairwise comparisons.

To evaluate corn responses to increasing rates of applied N, each year curves were fitted to WPNU vs. actual N rate applied the same year for each N source. Curves were chosen between first- and second-order polynomials, and an exponential function based on best fit. The same approach was also used for the residual year based on the N rates applied in year 1.

In this article, available N can be thought of as the portion of N that behaves as inorganic N, whereas recovered N is the amount of applied N actually taken up by the plant (measured in aboveground tissue) (Muñoz et al., 2004). First-year N availability was estimated using the fertilizer equivalence (FE) method (Klausner and Guest, 1981; Motavalli et al., 1989). In 1992, a straight line was adjusted to WPNU vs. AN rate. In 1993 and 1994, an exponential function, adapted from Klausner and Guest (1981), was chosen: WPNU = A – B exp (–C x Nrate). Using these curves, FE (the AN rate that would have produced the same WPNU as a given manure N rate) for each manure treatment was determined following the procedure outlined by Klausner and Guest (1981) and expanded by Muñoz et al. (2004). When using the exponential function, if WPNU for a given treatment was > A, it was equaled to 95 and 99.5% of A in 1993 and 1994, respectively, to calculate a FE. This approximation was used to overcome an inherent limitation of the FE approach where some WPNU values are too high to be handled by the equation. This is due in part to the fact that the equation used to calculate FE is based on AN treatments but, occasionally, manure results in higher crop production than that which can be achieved by inorganic fertilizer alone. This suggests that manure may have effects on corn production other than N supply, including supplying secondary and micronutrients, improving soil structure leading to better aeration and/or water retention, and/or stimulating soil microbial activity. Percent N availability was calculated as: N Avail = FE x 100/Applied N.

The resulting N availability values were then analyzed as a RCBD, in a similar manner as described above. The difference was that the treatments now consisted of the five N sources and four nominal N rates, but not the controls. Thus the treatments had a 5 x 4 two-factor structure. F-tests were used to assess the significance of the main effects and interactions.

First- and second-year apparent N recovery (ANR) was calculated according to the difference method as described by Motavalli et al. (1989):

Formula
where WPNUtmt denotes WPNU for a given treatment plot, WPNUctrl is the average control WPNU, and Applied N is the actual N rate (in kg ha–1) applied the same or previous year. Thus defined, ANR amounts are cumulative. The small amount of N in the starter fertilizer was not included in the actual applied N rate as this amount was also applied to the control plots.

For PSNT evaluation, relative grain yields (RelGY) were calculated as a percentage of non-N-limited GY, which was estimated as the average GY for the highest AN rate (269 kg N ha–1) in any given year (Paul and Beauchamp, 1993). Since in 1995 no experimental plot received N, we had no measure of non-N-limited GY. Therefore, data from this year were not included in PSNT analyses. The relationship between RelGY and soil NO3–N to 30 cm was described by either a linear-plateau model or a straight line.


   Results and Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
Conclusions
 REFERENCES
 
First-Year Yield and Nitrogen Uptake
The observed first-year WPY, WPNU, and GY values are shown in Table 3 . The RCBD ANOVA showed that the overall treatment effects were significant (P < 0.01) for all three crop parameters.


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

  		Table 3. First-year corn whole-plant yield, whole-plant N uptake and grain yield as affected by N source in 1992 to 1994 at Arlington, WI.

 
The results of the pairwise comparisons among the N sources are shown in Table 4 . Compared with the control, N amendments significantly increased corn yield and N uptake, with the exception of CC, which resulted in WPY and WPNU indistinguishable from the control. No significant differences were found between AN and FP. Ammonium nitrate increased WPNU significantly more than DP, but there were no differences between these in whole-plant or grain yield. Both yield and N uptake were significantly higher with FP than DP. Composted poultry manure resulted in significantly lower corn production than AN, FP, or DP manure. All other N sources resulted in higher WPY, WPNU, and GY than CC. In general, these results indicate that our initial estimates of FP apparent N availability were approximately correct since results were similar to those using AN; however, where the manure underwent more processing (i.e., DP and CP), we apparently underestimated equivalency. In the case of CC, little available N was being provided since these treatments could not be distinguished from the untreated controls.


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

  		Table 4. Effect of N sources on 1992 to 1994 corn whole-plant yield, whole-plant N uptake, and grain yield at Arlington, WI; probability values for pairwise comparisons.

 
Increasing rates of manure or inorganic N tended to increase corn yield and N uptake, as suggested by the significant regressions against actual N rate applied as shown in Fig. 1 with the specific regression equations given in Table 5 for WPNU for each study year. The regressions were significant (P < 0.01), except for CC in all years and CP in 1993. Separate regressions for WPY and GY were omitted since in general they followed similar trends to those of WPNU. In 1992, all relationships were straight lines, suggesting that WPNU was not maximized. Clearly, the assumptions used to establish the initial manure rates were too low, and these were increased in subsequent years. In 1993, the increase in WPNU with AN rate was exponential, with N uptake stabilizing between the 134 and 202 kg N ha–1 rates. Whole-plant N uptake increased linearly with DP rate. The response to FP rate was best described by a second-order polynomial with a maximum at about 350 kg N ha–1. The decrease in WPNU (also seen in WPY) at higher N rates suggests a possible toxic effect; however, in 1994, FP rates were only slightly lower, and no such effect was observed. Furthermore, in 1993, neither GY nor whole-plant N concentration (data not shown) exhibited a similar decrease at high rates. The apparent decrease is therefore likely due to anomalously low WPY in this year. In general, GY appeared to be a somewhat more consistent indicator of corn production than WPY, mainly because the samples were larger (two entire rows vs. 10 plants for WPY). This kind of sampling problem was also observed by Muñoz et al. (2004). In 1994, AN, FP, and DP showed a plateau in WPNU at the second or third N rate. For FP and DP, data were fit to second-order polynomials, whereas the relationship between WPNU and CP was linear.


Figure 1
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 1. First-year whole-plant N uptake vs. actual N rate from several N sources at Arlington, WI, 1992 to 1994.

 

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

  		Table 5. Regression equations for first- or second-year whole plant N uptake (y) vs. actual N rate (x) from several N sources at Arlington, WI, 1992 to 1995.

 
In general, Fig. 1 shows that AN, FP, and DP tended to produce similar responses, when taking into consideration the actual N rates. Composted manures, on the other hand, resulted in substantially lower N uptake, in many cases with values similar to those from the controls. Although manure rates were intended to provide approximately similar amounts of available N, the actual N rates were considerably lower for CC than the other N sources. Several studies have reported higher corn production with fresh manures than composted manures (Ott et al., 1983; Brinton, 1985; Paul and Beauchamp, 1993). In a greenhouse study, Fauci and Dick (1994) found that corn dry matter yields with composted beef manure were similar to those obtained from the control, while CP produced significantly higher yields. In the same study, an 87-d incubation resulted in net immobilization of beef manure compost N in contrast with an estimated 36% mineralization of the organic N in CP.

Apparent First-Year Nitrogen Availability
Apparent first-year N availability was estimated using the FE method based on WPNU (Table 6 ). Nitrogen source (P < 0.01), rate (P = 0.01), and source x rate (P = 0.02) had a significant effect on apparent N availability. Pairwise contrasts of N availability among N rates were performed for each N source. For DP, N availability at the lowest rate was significantly higher (P < 0.01) than at any of the other rates. For FP, the two lowest rates had a higher apparent N availability than the highest rate (P = 0.07 and 0.06, respectively). This influence of rate on apparent availability was the consequence of the much greater increase in N loads compared to the amount of N taken up, and has been observed previously (Culley et al., 1981; Ma et al., 1999; Muñoz et al., 2004). In 1993, the rate effect for apparent N availability from FP was exacerbated by the lower than expected WPY, and therefore WPNU, measured at the highest rate. As discussed previously, this was likely due to non-representative whole-plant sampling. In 1992 and 1994, the calculation of N availability for DP at the low rate resulted in values significantly above 100% (144% in 1992 and 242% in 1994). This is a result of the very steep increase in WPNU from this N source, relative to the response from AN, with relatively small amounts of total N applied. Similar anomalies have been observed by others, especially at low rates of manure N applied (Motavalli et al., 1989; Muñoz et al., 2004). Since manure N availability is based on the expectation of 100% availability for the fertilizer, the manure N availability cannot exceed 100%, therefore these values were capped at 100% for the purposes of calculating the means shown in Table 6.


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

  		Table 6. Estimates of first-year N availability from several N sources based on fertilizer equivalence or apparent N recovery in 1992 to 1994 at Arlington, WI.

 
In 1993 and 1994, since the WPNU responses for AN, FR, and DP are curvilinearly rate-dependent, one might expect more accurate estimates of apparent N availability at the lower N rates (in the more responsive portions of the curves); however, since all three of the curves are generally shaped similarly, the availability estimates remain relatively consistent. It is only at the FP high rate in 1993 where WPNU was very low and the DP low rates discussed above that our estimates of availability are likely outliers.

In a few cases, especially with CC, apparent availability as determined by FE resulted in slightly negative values. Although it is possible, especially for immature compost, to result in net N immobilization (Shi et al., 1999), these results may be within experimental error for this field study.

Averaged across years and rates, N availability was 57, 53, 14, and 4% for FP, DP, CP, and CC, respectively, the latter not being statistically different from the controls. Several other studies have also shown higher N availability for fresh than composted manures including 40 vs. 15% (Eghball and Power, 1999a), and 38 vs. 20% (both for beef feedlot) (Eghball and Power, 1999b). In a greenhouse study, averaged over years, soils, rates, and methods, first-year chicken manure N availability was 60% for fresh (similar to our estimate of 57%) vs. 35% for composted (higher than ours), while for dairy manure it was 23 vs. 11% for fresh and composted, respectively (Castellanos and Pratt, 1981). The same trend was found in estimates of ANR comparing fresh vs. composted manure: 28 vs. 9% (dairy) (Brinton, 1985), 30 vs. 21% (dairy) (Castellanos and Pratt, 1981), 18 vs. 5% (liquid dairy vs. composted beef) (Paul and Beauchamp, 1993), and 8 vs. 2% (beef) (Paul and Beauchamp, 1994). Further data supporting the relatively lower availability of compost N are provided by incubation studies. In a field incubation, N mineralization of the organic fraction of composted beef cattle feedlot manure was estimated to be about half that of fresh manure, namely 11 vs. 21% (Eghball, 2000). In a pot experiment with ryegrass (Lolium multiflorum) and 15N-labeled dried and composted poultry manure, the ANR was 15 and 14%, respectively, after 106 d, but use of 15N showed that 26% of the N in the dried manure was taken up by the plant compared to 4% of the N from the composted manure (Kirchmann, 1990). Our mean ANR estimate is higher for DP and lower for CP at 30 and 7%, respectively.

Although it is widely acknowledged that N availability is lower for compost than for fresh manure, there is great variability in the actual estimates. For example, dairy or beef cattle manure compost ANR has been reported from 5 to 21% (Castellanos and Pratt, 1981; Brinton, 1985; Paul and Beauchamp, 1993; Eghball and Power, 1999b), while in our study it was not significantly different from zero. Castellanos and Pratt (1981) found that a very old compost had a much lower ANR (6%) than a more recently produced one (21%). Characteristics of the composting process, such as moisture and turning management, also play a key role in its plant availability (Shi et al., 1999). Details of the processes are often unknown when purchasing commercially available compost, but these factors clearly have an impact on its use as a nutrient source, and partly explain the wide differences in reported compost N availability and the lack of response to CC in our study. In addition, given the lower N mineralization observed in compost, some research has shown that compost should be continuously applied for several years until available N from the current plus previous years is enough to produce optimal yields (DeLuca and DeLuca, 1997). Meanwhile, additional N inputs may be needed until compost is able to sustain productivity by itself, particularly in fields that have low initial soil fertility.

Second-Year Yield and Nitrogen Uptake
Corn WPY, WPNU, and GY, measured 1 yr after N application, are presented in Table 7 . The overall treatment effects were significant (P < 0.01) for all three crop parameters. The pair-wise comparisons among the N sources (Table 8 ) showed that plots that had received N the previous year produced significantly higher yields and N uptake than controls, except for CC, where WPNU was the same as the control. Previous applications of AN, FP, and DP increased crop production more than composted manures, except WPNU was the same for DP and CP. In general, the effect of residual FP or DP N was not different from inorganic N, with the exception of WPY, which was higher for FP, and for GY, which was higher for DP. Residual FP N produced higher WPNU and GY than DP. Residual CP N increased WPNU more than CC, although yields were statistically the same.


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

  		Table 7. Second-year (residual) corn whole-plant yields, N uptake, and grain yield in 1993 to 1995 at Arlington, WI as affected by several N sources applied the previous year.

 

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

  		Table 8. Effect of residual N sources on 1993 to 1995 corn whole-plant yield, whole-plant N uptake, and grain yield at Arlington, WI; probability values for pairwise comparisons.

 
The substantial yield and N uptake responses to previous applications of AN, FP, and DP emphasizes the impact that previous N applications can have on crop performance. Even on this highly productive, prairie-derived soil (38 g kg–1 OM), yields were clearly negatively affected where significant amounts of carryover N were not present.

Curves fitted to the increases in WPNU vs. actual N rate applied in the previous year are presented in Fig. 2 with the regression equations for the second-year N uptake given in Table 5. Regressions were not significant in 1993 and are therefore not presented. It is likely that the cool and moist conditions especially in the first half of the 1993 growing season (average April to July temperatures 1.3°C below 30-yr norms and total rainfall 365 mm above average for this same period) (Table 2) resulted in leaching of NO3–N, greater denitrification, and lower mineralization causing the generalized absence of response to residual N. Responses of WPY and GY to AN, FP, and DP were similar. Residual CP produced increased WPY and WPNU in 1995 only. No significant effect of residual CP or CC was observed in GY in any year. Most significant relationships were best described by straight lines, indicating that, as expected, corn production was not maximized with the available amounts of N carried over from the applications made the previous year. However, residual FP in 1995 caused WPY and WPNU to peak at a previous year's N rate of 320 to 390 kg ha–1, with a clear drop at higher rates. It is again likely that the apparent drop in WPNU observed in 1995 is a result of the variation in our measurement of WPY as GY did not show a similar drop and were rather uniform for the residual FP for the four rates of application (Table 7).


Figure 2
View larger version (15K):
[in this window]
[in a new window]

  		Fig. 2. Second-year (residual) whole-plant N uptake vs. actual N rate applied the previous year from several N sources at Arlington, WI, 1993 to 1995.

 
Second-Year Apparent Nitrogen Recovery
Table 9 shows second-year ANR values for all N sources and rates. Some ANR values were not significantly different from zero. However, these values were included in the calculation of an average ANR, since the ANOVA had previously indicated that, for all sources except CC, WPNU was significantly higher than the control. In this way, ANR provided a more realistic measure of the actual value of residual N. Averaged across rates and years, ANR was 18, 10, 8, and 10% for AN, FP, DP, and CP, respectively. Compared to the amount of N recovered in the year of application, these values were generally less than half as large (first-year ANR are 46, 25, 30, and 7% for AN, FP, DP, and CP, respectively). The average values for residual ANR may be skewed somewhat low since the measurements for the cool and wet 1993 season are significantly lower. Previously reported residual ANR or N availability varies widely regardless of manure source, although they commonly range from 30 to 50% of first-year values (Paul and Beauchamp, 1993; Sørensen et al., 1994; Eghball and Power, 1999a). For residual CP, ANR was slightly higher than first-year ANR. Residual CC, on average, had no significant effect on WPNU, although it showed a residual effect in 1994. Other researchers have reported low residual value for composted cow manure (beef) where they found 3% ANR (Paul and Beauchamp, 1993) or 8% apparent N availability (Eghball and Power, 1999a). Using 15N-enriched fresh sheep manure, N recovery in the year following application was 2 to 3% (Thomsen et al., 1997), 4% (Jensen et al., 1999), and 4 to 6% (Sørensen et al., 1994). Our ANR of fresh, dried, and even composted chicken manure is higher.


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

  		Table 9. Apparent residual N recovery measured the year following applications from several N sources at Arlington, WI, 1993 to 1995.

 
As shown by Muñoz et al. (2004), ANR can be used to provide a more realistic estimate of manure N availability by dividing manure ANR by AN ANR at an approximately similar rate of available N (termed relative effectiveness). Residual poultry manure N availability has been estimated to be 6 to 11% using FE based on GY (Bitzer and Sims, 1988), and second-year constants for poultry manure decay series have been estimated to be 0.10 (Pratt et al., 1973), 0.05 (Bary et al., 2000), and 0.20 (Sims, 1987). The average relative effectiveness for the residual years of our study for which this could be calculated (1993 and 1994 only) for FP, DP, CP, and CC are 18, 19, 12, and 7%, respectively, on the higher end of this relatively wide range.

Presidedress Soil Nitrogen Test
Average soil PSNT values for experimental plots are presented in Table 10 . The PSNT values measured the year following N applications were lower than those measured in the same year N was added (P < 0.01). In fact, all values in the residual year fell below 21 mg kg–1, the suggested critical level for Wisconsin by Bundy and Andraski (1993, 1995), suggesting that, although residual N increased yields compared to the control plots, yield was still limited by N supply. As expected considering the cool, wet season, soil NO3–N levels were by far the lowest in 1993, ranging from 5.4 to 10 mg kg–1, when residual N had no significant effect on corn production compared to the zero N control. The RCBD ANOVA showed that treatment effects were significant for PSNT values (P < 0.01). A linear contrast showed that increasing rates of N, applied either the current or previous year, tended to increase the PSNT value (P < 0.01). Pairwise comparisons among N sources are shown in Table 11 for first- and second-year plots. For first-year measurements, PSNT values in plots receiving CC did not differ from controls, and both were significantly lower than those of any other N source. The highest PSNTs were observed in plots receiving AN, followed by FP, then DP, and finally CP. Heckman et al. (1996) and Roth et al. (1992) also found higher PSNT values in manured fields than non-amended fields. The second study noted that among manured fields, those receiving poultry manure tended to have higher PSNT than fields receiving other manures.


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

  		Table 10. Soil NO3–N to 30 cm at 15 to 30 cm plant height (PSNT) as affected by several N sources for year of N application and the following year at Arlington, WI, 1992 to 1994.

 

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

  		Table 11. Effect of N sources on corn PSNT values at Arlington, WI, 1992 to 1994; probability values for pairwise comparisons.

 
Interestingly, residual PSNT values in our CC plots were significantly higher than controls and CP, and did not differ from the other N sources, although there was no increase in corn production due to residual CC. The trend for poultry manures was as expected, with PSNT values in FP plots being higher than in plots receiving either DP or CP, and DP being higher than CP. Soil NO3–N concentrations to 30 cm reflected the relatively low residual value of inorganic N, and although PSNT was higher for AN than for DP and CP, it could not be differentiated from the FP or CC treatments.

Pooling data from all study years and experimental plots, we obtained a critical PSNT value of 23.8 mg kg–1 (Fig. 3 ). This is in good agreement with generally accepted values of 20 to 25 mg kg–1 (Blackmer et al., 1989; Fox et al., 1989; Meisinger et al., 1992; Bundy and Andraski, 1993; Klausner et al., 1993; Bundy and Andraski, 1995) across several states and soil conditions. The test was considered "successful" when either PSNT < 23.8 (critical value) and RelGY < 93%, or when PSNT > critical value and RelGY > 93%. Our cutoff value of 23.8 mg kg–1 was able to correctly predict the need for additional N 87% of the time in the 1992 to 1994 period. Other reported success rates are similar: 84% (Klausner et al., 1993), 82% (Jemison and Lytle, 1996), and 81% (Heckman et al., 1996). About 5% of the plots had a PSNT > critical value, but RelGY < 93%, representing potential losses in corn production, since no additional N would have been recommended, but addition of some N would have likely increased yields. The remaining 8% were plots where sidedressing would have been recommended but it apparently was not needed. These data confirm that the PSNT is a quite robust and useful tool to predict the need for further N fertilization across a range of N sources and conditions, provided caution is exercised in unusual years (extremely wet or dry). The PSNT is particularly successful in reducing unneeded fertilization in fields with high soil NO3–N from manure applications (Roth et al., 1992; Paul and Beauchamp, 1993; Sims et al., 1995).


Figure 3
View larger version (22K):
[in this window]
[in a new window]

  		Fig. 3. Relative grain yield (RelGY) vs. soil NO3–N to 30 cm (PSNT) for all experimental plots, except 1995 residuals at Arlington, WI, 1992 to 1994.

 
When RelGY were plotted vs. PSNT by N source (Fig. 4 ), critical PSNT values for AN and CP were close (22.8 and 23.2 mg kg–1, respectively) and slightly lower than for FP and DP (26.2 and 26.8 mg kg–1, respectively). However, the 95% confidence intervals for the critical PSNT for all N sources overlapped, in most cases substantially. Taking a conservative approach, we do not conclude that they differed. Moreover, from Fig. 3, all observations seemed to fall within a single population. This supports the use of a single PSNT cutoff value to detect the need for additional N supply regardless of the N source.


Figure 4
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 4. Relative grain yield (RelGY) vs. soil NO3–N to 30 cm (PSNT) by N source at Arlington, WI, 1992 to 1994. The plot shows averages by treatment, but equations were fitted using all observations, and controls were included in each regression.

 

   Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results and Discussion
 Conclusions
REFERENCES
 
This study estimated first-year apparent N availability of 57, 53, 14, and 4% from fresh poultry (FP), dried poultry (DP), composted poultry (CP), and composed cow (CC) manure, respectively, although based on whole-plant N uptake, the latter could not be distinguished from the untreated controls. These values are generally similar to those reported by others and emphasize the need for producers using these N sources to understand the amount and kind of processing to which the manure has been subjected. Use of composted manures as an N source will likely require substantially larger application rates to meet the crop N needs or may need to be supplemented with a more available source. Measurements of apparent N recovery from the various N sources in the year following applications allows for apparent availability estimates based on relative effectiveness. These residual year values of 18, 19, 12, and 7% for FP, DP, CP, and CC, respectively, tended to be somewhat higher than those reported previously, and provided some indication that our source of composted cow manure needed at least two seasons to start releasing its N. Although manure testing will determine the amount of N present in various manures, users need to understand that differences in apparent availability exist. Based on these data across the various N sources and growing years, the PSNT provided a relatively reliable (87% correct) measure if supplemental N would be needed or not. Use of the PSNT may in part substitute for gathering more precise manure source or rate information.


   ACKNOWLEDGMENTS
 
This research was funded by TVA Environmental Research Ctr., Muscle Shoals, AL and the College of Agricultural and Life Sciences, Univ. of Wisconsin-Madison and is gratefully acknowledged. Appreciation is expressed to R.P. Wolkowski, P. Wakeman, and P.E. Speth for help with the field portion of this study.


   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
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 Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Muñoz, G. R.
Right arrow Articles by Zhu, J.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Muñoz, G. R.
Right arrow Articles by Zhu, J.
Agricola
Right arrow Articles by Muñoz, G. R.
Right arrow Articles by Zhu, J.
Related Collections
Right arrow Best Management Practices
Right arrow Nutrient Management
Right arrow Soil Fertility and Productivity
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