| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Crop and Soil Science, Oregon State Univ., Corvallis, OR 97331
b Dep. Crop and Soil Sciences, Washington State Univ., 7612 Pioneer Way E., Puyallup, WA 98371
c Dep. Horticulture, Oregon State Univ. North Willamette Research and Extension Center, Aurora, OR 97002
* Corresponding author (Dan.Sullivan{at}oregonstate.edu)
Received for publication February 14, 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: PAN, plant-available nitrogen • FNE, fertilizer nitrogen equivalency
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Correlations between C and N analyses of specific organic materials and the amount and timing of available N release have been reported for crop residues (Vigil and Kissel, 1991; VanLauwe et al., 1997; Trinsoutrot et al., 2000), animal manures (Castellanos and Pratt, 1981; Chae and Tabatabai, 1986), components of dairy manure (Van Kessel et al., 2000), and municipal biosolids (Gilmour et al., 2003). As C/N decreases, the percentage of lignin in a residue typically decreases, and the amount of residue that is rapidly decomposed in soil during the first 30 d in soil increases (Ajwa and Tabatabai, 1994; Van Kessel et al., 2000). While C/N is a good indicator of PAN released for fresh crop residues or manures, it not as useful as a general indicator for a set of organic materials that may include fresh organic matter such as crop residues, partially decomposed organic matter such as stored manure, and composts. Composting or storage of organic materials with low C/N typically reduces the decomposability of residue, but it often does not appreciably change C/N, because both C and N losses usually accompany decomposition (Hansen et al., 1993). Knowledge of amendment decomposition kinetics is especially helpful for organic materials with low C/N. For example, for biosolids with C/N of 5 to 10, Gilmour et al. (1985) and Gilmour et al. (2003) reported a linear relationship with a slope not significantly different than one between net amendment N mineralized (% of organic N) and decomposition in soil.
Computer simulation models have shown potential for predicting N mineralization under a wide range of conditions. With proper calibration and verification, many models have the potential to be used for data interpretation, education, and decision support in soil N management (Schaffer et al., 2001). One general criticism of models, however, is the need for further verification (Wagner et al., 1998), particularly using data independent of those used to develop and parameterize the model (McGechan and Wu, 2001).
The model DECOMPOSITION is a mechanistic computer simulation model first described by Gilmour and Clark (1988) and subsequently described in detail by Gilmour (1998). The model uses first-order kinetics to estimate PAN. Most other models that describe N mineralization processes also utilize first-order kinetics, modified for environmental variables (McGechan and Wu, 2001). DECOMPOSITION is a relatively simple model by today's standards. It simulates only the mineralization/immobilization of N following organic amendment application. The model has proved to be reasonably accurate in predicting PAN from municipal biosolids. In an evaluation of the model DECOMPOSITION using 30 different biosolids with field sites in Arkansas, Michigan, Virginia, and Washington, Gilmour et al. (2003) found good agreement between observed PAN vs. simulated PAN during one growing season using actual decomposition kinetics, weather, and amendment analytical data (r2 = 0.72). With three biosolids and one site, Gilmour and Skinner (1999) also found good agreement between observed PAN versus modeled PAN using DECOMPOSITION (r2 = 0.67). Findings from the model have been used to develop national guidance for biosolids application based on average air temperature and irrigation (Gilmour et al., 2000).
The current study was designed to provide data to support Extension recommendations for organic amendments. Specifically, our objectives were to (i) measure decomposition and N released from fresh and composted amendments and (ii) evaluate the performance of the model DECOMPOSITION, a relatively simple N mineralization/immobilization model as a predictor of N availability.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Broiler litter was obtained from a production broiler facility (Mossy Rock, WA) that routinely markets a compost product to farmers and landscapers. Although marketed as compost, the amendment production process includes only periodic turning of large windrows without moisture addition. The fresh broiler litter was dry-stacked for 56 d (2002 sample) or 14 d (2003 sample). The composted broiler litter samples were dry-stacked for 84+ d in both years.
Separated dairy solids were collected from a dairy (Woodburn, OR) that employed a flush manure handling system. A mechanical separator removed the coarse, fibrous fraction of the manure as it passed from the animal confinement area to a storage lagoon. Fresh dairy solids were stored less than 7 d at the dairy. Composted dairy solids were produced in windrows that were turned weekly over a 60-d period with a self-propelled windrow turner. At each turning, moisture was added (if needed) to maintain moisture of 500 to 600 g kg–1 in the compost windrow.
Yard trimmings and yard trimmings compost were produced at a large indoor commercial compost facility in Puyallup, WA. Yard trimmings consisted of approximately a 60:40 mixture (v/v) of woody tree and shrub trimmings mixed with grass clippings. Both yard trimmings and yard trimmings compost were ground with a hammer mill to reduce particle size on receipt at the facility. The yard trimmings then were placed in windrows, and allowed to heat for 7 d with periodic turning to kill weed seeds. The yard trimmings compost was prepared by placing yard trimmings in windrows and turning a minimum of 5 times over a 40-d period with a compost pile turner. Moisture was added at each turning, and windrows received forced aeration based on automated feedback from temperature probes placed in the windrows. After 40 d, the compost was placed in larger curing piles for at least 20 additional d. Before delivery to field experiments, large debris >11 mm (>7/16 in.) was removed from the yard trimmings compost by screening.
Fresh rabbit manure without bedding was obtained from a rabbit production facility near Corvallis, OR. The rabbit manure compost was produced in small outdoor windrows (50 m long, 1 m high and 2 m wide) in summer. Windrows were turned with a compost pile turner and moisture was added at least six times during 60 d of active composting. Some soil underlying the windrows was incorporated during composting. The finished rabbit manure compost was covered with plastic sheeting in the fall. The cover was removed just before transport to the field experiment locations the following spring.
The remaining organic materials used in the study (other broiler litter, other composts, specialty products; Table 1) came from a variety of sources and were included in selected laboratory incubations and field trials. Other broiler litters (n = 6) were obtained from broiler production facilities; information on storage and/or composting time and methodology was not available. Other composts (n = 6) included two samples of composted dairy solids from additional local facilities, one sample of solids from an anaerobic dairy manure digester, and three samples of composts prepared by a Puyallup, WA composting school. The anaerobically digested dairy solids were collected from a complete mix digester, operating at 36°C with a mean residence time of 28 d. The composts from the compost school were produced from a 40:20:15:5:10 v/v mixture of yard trimmings, separated dairy solids, cereal straw, horse manure, and laying-hen manure that was composted for approximately 180 d under passive aeration. Specialty amendments (n = 5) included two samples of pelleted organic fertilizer derived from fish byproducts, two feather meals, and one canola (Brassica spp.) meal.
Amendment Sampling, Handling, and Preparation for Analyses
Amendments were transported to the field experiment sites no more than 7 d before field application, and kept under cover. Immediately before amendment application at each field experiment, composite amendment samples for inorganic N analysis and for use in soil incubation experiments were collected, subsampled, placed into 1-L zippered polyethylene bags, then frozen within 24 h of collection. Each composite amendment sample was a subsample derived from a mixture of 20 grab samples collected from the amendment pile or bag.
Total solids in the fresh and thawed amendment samples were determined by drying at 55°C. There was no consistent difference between total solids in fresh samples or in samples thawed after freezing.
Field Studies
‘Jubilee’ sweet corn (Zea mays L.) was grown at both field sites located at the WSU Research and Extension Center at Puyallup, WA and at the OSU North Willamette Research and Extension Center near Aurora, OR. The Oregon site was located on a Willamette silt loam soil (fine-silty, mixed, superactive, mesic Pachic Ultic Argixerolls). The Washington site was located on a Puyallup fine sandy loam (coarse-loamy over sandy, mixed mesic Vitrandic Haploxerolls). The 2003 trials were conducted on acreage adjacent to the 2002 field trials (within the same field). The previous crop at the Oregon site in 2002 and 2003 was summer fallow. The previous crop at the 2002 Washington site was sweet corn followed by a winter triticale cover crop. The previous crop at the 2003 Washington site was sweet corn with no winter cover crop.
Pre-amendment soil samples collected in March or April of each year at each field site indicated that soil pH (>6.1) and soil test concentrations of P, Ca, Mg, and K were above established sufficiency levels for sweet corn production (Marx et al., 1999). At the Oregon sites only, blanket applications of soluble boron and sulfate of potash magnesia fertilizer (K2SO4.2MgSO4) were applied before seeding to supply 1 kg B ha–1 and 30 kg S ha–1. Soluble boron was sprayed on the soil surface with a boom sprayer. Granular sulfate of potash magnesia fertilizer was broadcast on the soil surface.
Both sites had a typical maritime Pacific Northwest climate with cool, wet winters and mild, dry summers. Mean growing season temperatures were 18°C in 2002 and 20°C in 2003 at the Oregon site, and 16°C in 2002 and 17°C in 2003 at the Washington site. Seeding dates were 25 May for the OR 2002 site, 13 June for the OR 2003 site, 21 May for the WA 2002 site and 28 May for the WA 2003 site. Degree days (0°C base temperature) from amendment application to harvest averaged 2240 across sites (Table 2). Growing season precipitation was 72 mm in 2002 and 42 mm in 2003 at the Oregon site, and 81 mm in 2002 and 37 mm in 2003 at the Washington site. Irrigation was supplied by solid-set sprinklers on a schedule designed to limit leaching losses but provide adequate moisture for crop production. Approximately 2.5 cm irrigation was supplied every 7 d during peak crop evaporative demand.
|
Urea treatments (no organic amendment) supplied a total of 0, 56, 112, 168, or 224 kg N ha–1 during the growing season, split into starter and sidedress applications. Starter urea was broadcast 1 or 2 d before planting, with additional side-dress urea broadcast just before irrigation at the four- to six-leaf growth stage. The side-dress fertilizer application took place 1 to 3 d following the date of mid-season soil sample collection (Table 2). In Year 1, 28 kg ha–1 of urea-N was applied before planting with the remainder added at the six-leaf growth stage. In Year 2, 56 kg ha–1 of urea-N was applied before planting with the remainder at the six-leaf growth stage.
Organic amendments were weighed, and then broadcast on the soil surface. Amendment application rates (Table 1) were based on estimates of amendment PAN, targeted to supply 40 to 120 kg PAN ha–1 during the growing season, based on current Extension guidance for the Pacific Northwest (Bary et al., 2000). To calculate application rates, we estimated PAN (% of total N) of 45% for broiler litter, 10% for dairy solids, 30% for rabbit manure, 15% for yard trimmings, 10% for composts, and 50% for specialty products and assumed typical total N analyses and total solids concentrations. Urea was not applied to plots receiving organic amendments. Actual amendment application rates were determined after application by multiplying amendment application amount by measured total solids concentration. Amendments were incorporated into soil by rototilling to a depth of 15 cm. Broiler litter and rabbit manure were incorporated within 2 h of application to limit NH3 loss. Other materials with lower potential for NH3 loss were incorporated within 24 h.
Soil samples were collected at three times following organic amendment application: pre-plant (after 14 to 28 d), mid-season (after 44 to 75 d), and post-harvest (after 121 to 133 d; Table 2). Each sample was a composite of 8 to 12 soil cores (0- to 30-cm depth) collected per plot with a 25-mm-i.d. push probe. Soil samples were air-dried within 1 d to stop microbial activity.
Laboratory Incubations
Decomposition of amendment organic matter was estimated via incubation with soil. The incubation was performed at 22°C in sealed 0.95-L Mason jars, with three replicates per amendment sample. Soil for incubation was collected in May from each field site approximately 30 d before the start of the laboratory incubations, and held at field moisture at 4°C. At the start of each incubation, soil was allowed to warm to room temperature for 3 d, then sufficient moisture added to bring gravimetric soil moisture to 200 to 250 g kg–1. Thawed amendment samples equivalent to 1 g dry weight were incorporated into 50 g soil for an application rate of 20 g kg–1 in Year 1. In Year 2, we reduced amendment incorporation rates to bring laboratory incubation amounts closer to those used in the field experiments. Incorporation rates were 5 g kg–1 for broiler litter samples, feather meal, and pelleted fish; other amendments were incorporated at a rate of 10 g kg–1. Each incubation included soil-only control jars to allow calculation of net CO2 evolution associated with amendment incorporation.
We used a titration method to determine CO2 evolution in Year 1, and a gas chromatograph method to determine CO2 in Year 2. Measurements were made after 3, 7, 14, 21, 28, 35, 42, 49, and 70 d. In Year 1, CO2 was collected in vials containing 20 mL 1M NaOH that were placed inside the incubation jars. Vials were replaced inside the jars after each incubation interval. For CO2 determination in Year 1, the carbonate trapped by NaOH was precipitated with excess BaCl2, and the remaining NaOH was back-titrated with standardized 0.1 M HCl, using phenolphthalein as the indicator (Anderson, 1982). In Year 2, air was collected from sample jars by inserting a syringe through a rubber septum in the jar lid. Carbon dioxide was determined with a He-carrier gas chromatograph (Carle Series 100 AGC, Hach, Loveland, CO). A calibration experiment verified that the chromatograph CO2 method used in Year 2 yielded comparable data to the titration method used in Year 1. We incubated 15 amended jars (5 amendments x 3 replicates) and determined CO2 evolved after 3, 7, and 14 d. A regression between the two methods yielded an r2 of 0.98, a slope not significantly different than one, and an intercept not significantly different from zero.
Inorganic N accumulation was measured using incubation intervals of 14, 42, and 70 d, and the same amendment incorporation rates were used to measure CO2 evolution. For inorganic N measurement, moist soil (650 g dry wt.) + amendment were incubated in zippered 3.8-L polyethylene bags. The larger incubation samples provided a larger amount of soil for repeated subsampling, and allowed for ventilation while maintaining consistent soil moisture. The target soil moisture content used in this experiment (200 to 250 g H2O kg–1) allowed thorough mixing of the amendments with soil, and it allowed the soil aggregates to remain small. The small soil aggregates were easily subsampled, and porosity was maintained throughout the soil-amendment mixture. The zippered tops of the bags were left partially open during incubation to facilitate air exchange, thereby reducing the potential for denitrification. Incubation bags were placed within 20-L plastic tubs that contained moistened foam pads to increase humidity. At 14-d intervals, soil moisture in the bags was measured and replenished if necessary. Soil subsamples (15 g) for inorganic N analysis were collected from the bags on Day 0, 3, 7, 14, 42, and 70 in Year 1 and at Day 0, 14, 42, and 70 in Year 2, extracted with 50 mL 2M KCl, and refrigerated (4°C) until analysis. Additional soil samples were collected at each sampling date for determination of actual soil moisture. Soil moisture data were used to convert inorganic N concentrations in moist soil to a dry weight basis.
Soil and Amendment Analyses
Inorganic N in all soil samples and in amendment samples was determined by automated colorimetric methods after extraction with 2M KCl at the Oregon State University Central Analytical Laboratory. Ammonium-N was extracted from both fresh and dried amendment samples using a 1:5 solids/extractant ratio. Nitrate-N was extracted from only the fresh amendment samples, because nitrate is not volatilized when samples are oven-dried. After shaking for 30 min, KCl extracts were filtered through Whatman No. 42 paper and refrigerated before analysis. Extracts were diluted to prevent interference from dissolved organic matter in colorimetric N determinations. Ammonium-N was determined using a salicylate-nitroprusside method (Gavlak et al., 1994) and nitrate-N was determined by a cadmium reduction method (Gavlak et al., 1994). For total N and C determination, organic amendments were prepared by grinding to pass a 2-mm screen. Total N and C were determined using a combustion analyzer equipped with an infrared detector (LECO Instruments Model CNS 2000, LECO Instruments, St. Joseph, MI; Sweeney, 1989).
Calculations and Statistics
Total N concentrations (Nt) in amendment samples were calculated as:
 | [1] |
For each field trial site-year, we calculated full-season PAN supplied by amendments with a two-step calculation. First, we used linear regression to calculate the fertilizer N efficiency coefficient (eff) for the urea fertilizer treatments:
 | [2] |
Then Eq. [2] was rearranged and amendment PAN calculated using a fertilizer N equivalency approach:
 | [3] |
For field pre-plant soil samples, and for mid-season soil samples, we calculated PAN as the additional N recovered from soil following organic amendment application, as:
 | [4] |
Equation [4] was also used for laboratory PAN calculation, except that it was calculated using units of mg kg–1.
Amendment decomposition, as measured by CO2 evolution, was calculated as:
 | [5] |
Linear regression was performed using S-Plus (S-Plus 6.1 for Windows, Insightful). Regression equations for observed vs. modeled PAN were tested at P = 0.05 to determine whether the regression intercept was different than zero, and whether the slope of the regression equation was different than one.
Computer Simulation Model
The model DECOMPOSITION is a mechanistic model first described by Gilmour and Clark (1988) and described in detail by Gilmour (1998). The model uses first-order kinetics to estimate the rates of C and N transfer between six pools: fresh organic amendment (rapid and slow fractions), microbial biomass (indigenous and new), and soil organic matter (decomposable and recalcitrant) using daily time steps. Initial soil organic matter is divided equally between decomposable and recalcitrant pools. Default decomposition rate constants (d–1) are 0.035 for fresh biomass, 0.00093 for indigenous biomass, and 0.00020 for decomposable soil organic matter. The recalcitrant soil organic matter pool does not decompose. The decomposition of amendments is modeled using a sequential first-order model where the rapid fraction is entirely decomposed before decomposition of the slow fraction begins. Amendment decomposition rate constants (kr for the rapid pool and ks for the slow pool) are determined from 25°C laboratory incubation data using a segmented regression of the natural log of the percent C remaining. Nitrogen mineralization and immobilization in the model are driven by C decomposition, microbial efficiency, and C/N of the amendment, biomass, and soil organic matter. Microbial efficiency is assumed to be 0.4. Carbon/nitrogen ratios are assumed to be 10 for soil organic matter and 8 for microbial biomass. As the amendment decomposes, if amendment C/N is >15, then soil inorganic N is immobilized in microbial biomass. The model assumes zero N loss via NH3 volatilization or denitrification processes. Amendment decomposition rates (measured at 25°C) are modified for temperature using a Q10 approach and modified for soil moisture using a soil water factor scaled from 0 to 1 (Gilmour, 1998). Soil temperature is assumed to equal air temperature.
John Gilmour (John Gilmour, Fayetteville, Arkansas) received input data from our project consisting of a numeric amendment sample ID, amendment analyses for total N, total C, and NH4–N (dry weight basis), and amendment application rate. Gilmour calculated fast and slow decomposition rate constants for use in the simulations (Table 3) using our data for cumulative amendment CO2 loss over 70 d (Fig. 1). For use in the model, he multiplied our measured decomposition rate constants by 1.28, assuming a Q10 of 2, to estimate the amendment decomposition rate constant at 25°C. For modeling of inorganic N release in the laboratory incubations, Gilmour received data on soil moisture during incubation, temperature, and inorganic N recovered from the soil-only control (no amendment) at each soil incubation sampling date. For the field experiments, we provided Gilmour with dates of amendment application, daily average air temperatures collected from weather stations located within 1 km of each field site, and daily precipitation and daily irrigation application amounts.
|
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Composting reduced cumulative decomposition (70 d in laboratory) from 62 to 29% of applied C for dairy solids, from 66 to 9% for rabbit manure, and from 24 to 11% for yard trimmings (Fig. 1; Table 3). Dry stacking of broiler litter (sold as "compost") did not consistently reduce amendment decomposition in soil. Amendments sold as broiler litter "compost" retained most of the characteristics of raw broiler litter: rapid decomposition during the first 7 d in soil, amendment C/N of 9 to 10, and NH4–N of 5 to 9 g kg–1 (Table 1). Most of the difference in cumulative decomposition between composts and raw materials occurred during the first 7 to 30 d of incubation in soil at 22°C. Anaerobically digested dairy solids had decomposition (39% in 70 d) that was intermediate between raw dairy solids and composted dairy solids. Other composts included in the study averaged 17% decomposition in 70 d. Specialty products (fish, feather, and canola meals) had very high cumulative decomposition (average = 76% in 70 d) as expected for raw amendments with C/N of 4 to 8. Decomposition rates observed in this study are similar to those reported by others for fresh and composted organic amendments (Ajwa and Tabatabai, 1994; Gilmour, 1998).
Values for rapid and slow decomposition rate constants also reflect differences among amendments (Table 3). Composted amendments generally had a single rate of decomposition using the rate constant fitting procedure described by Gilmour (1998). Averaged over all composts (n = 14; excluding broiler litter "compost"), the average decomposition rate was 0.003 C d–1 (0.3% of remaining C per d). Values for rapid pool decomposition (kr) demonstrated relative differences among materials in early decay rate (Table 3). For example, the first-order decay rate for the rapid pool (kr) for fresh broiler litter was, on average, 2.0 times the kr value for dairy solids. The kr values for composted dairy solids, composted yard trimmings, and composted rabbit manure were of the same order of magnitude as the ks values for their raw counterparts. This finding demonstrated that once the easily decomposable fraction of a raw material was broken down, the remaining material exhibited a decomposition rate similar to that of compost.
Plant-Available Nitrogen
We estimated fertilizer nitrogen equivalency (FNE) in the field trials using crop N uptake + post-harvest soil inorganic N as the measured response variable (Eq. [3]; Table 4). We chose crop N uptake + post-harvest soil inorganic N as the best indicator of FNE because it accounts for N mineralized late in the growing season, and it yielded a linear N response curve across all rates of applied fertilizer N. The FNE regression equations used to estimate PAN had reasonable fertilizer N recovery efficiency, averaging 68% across the four field site-years (Table 4). We also determined FNE based on other crop response variables: crop N uptake, fresh weight ear yield, and chlorophyll meter readings of leaves (Gale, 2005). All methods used for estimating FNE yielded similar values for the organic amendment treatments. Coefficients of determination (r2) between FNE determined using crop N uptake + post-harvest soil inorganic N (Eq. [3]) and other variables were: 0.80 for chlorophyll readings, 0.96 for crop N uptake, and 0.90 for fresh weight ear yield (Gale, 2005).
|
|
|
|
Model Verification: Laboratory and Field
Across all amendments, observed PAN was strongly correlated with PAN estimated by the DECOMPOSITION model, as demonstrated by regressions at 70-d in the laboratory (r2 = 0.74) and for the full season in the field (r2 = 0.78; Fig. 5). Although strongly correlated, modeled PAN was often greater than observed PAN. For the 70-d laboratory incubation, the slope of the regression line was not different than one, but the intercept was significantly greater than zero at P = 0.05. For the full season in the field, the regression line was not different than the 1:1 line in slope or intercept. Across all amendments, the average model overprediction of PAN determined by subtraction (modeled PAN minus observed PAN for each amendment) was 3% PAN at 14 d, 7% at 42 d, and 10% at 70 d in the laboratory (Table 5). In the field, the model overpredicted by an average of 3% PAN at preplant, 10% at mid-season, and 8% for the full season (Table 5).
|
|
The model was more accurate in predicting PAN from composts having a single decomposition rate constant than for uncomposted materials that had rapid and slow decomposition rate constants in the laboratory (Fig. 4) and in the field (Fig. 5). The composts had very low decomposition rates compared to fresh materials (Table 3). Compost PAN increased slowly with time in the laboratory (Fig. 4) and in the field (Table 5). Often the 70-d PAN for composted materials was close to that measured at the start of the incubation. Thus, the modeling of PAN for composts did not provide a strong test of model capacity to simulate N mineralization and/or immobilization processes.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This study demonstrated that field and laboratory measurements of PAN were strongly correlated with PAN modeled by DECOMPOSITION, a model that includes only the mineralization/immobilization portion of the nitrogen cycle. The model tended to overpredict PAN, particularly for amendments with high initial NH4–N concentrations, and for amendments with rapid decomposition rates in soil. The simplest explanation for this discrepancy is that in our trials, some PAN was lost as NH3. However, this explanation is not supported by the 1:1 relationship observed between field PAN (some potential for NH3 loss) and laboratory PAN (very low potential for NH3 loss). The strong correlations between measured and modeled PAN in the field trials suggested that the algorithms used by DECOMPOSITION to predict N mineralization from amendments are sufficiently detailed to use for planning purposes, when excellent N management practices are used. In situations where substantial PAN losses are expected, the mineralization/immobilization algorithms used in DECOMPOSITION must be integrated into a more complete model. Our study included only one amendment that demonstrated N immobilization (fresh dairy solids). Therefore, we did not obtain sufficient data to verify outputs from the N immobilization algorithms in DECOMPOSITION. The DECOMPOSITION model approach was most valuable for organic materials with C/N of 10 to 15. These materials had the widest range in N availability, depending on whether the organic fraction of the amendment was readily decomposable (e.g., fresh manure) or stable (compost).
This study demonstrated that traditional laboratory analyses of amendments (C/N and total N) and short-term laboratory incubations that determine decomposition and net available N released from amendments have value in providing improved estimates of field PAN for the first growing season. We suggest a minimum data set be used to characterize typical amendment characteristics in guidance publications for organic farmers and others that use manure to supply N, including: amendment total N, C/N, total solids, NH4–N and NO3–N, decomposition of the amendment determined after 7, 14, and 28-d incubation, and PAN released from the amendment after 28-d incubation at room temperature (approximately 22°C). Guidance should also include the number of independent amendment samples used to determine typical values. Having three decomposition data points (7, 14, and 28 d) is sufficient to demonstrate whether an amendment has been composted (relatively constant decomposition rate over time) or not composted (rapid decomposition in the first 7 d, followed by slower decomposition). In both our field and laboratory studies, a large fraction of amendment PAN was recovered in the first 28 d after application. Decomposition over 28 d is also suitable for determining whether a single decomposition rate constant, or two rate constants (rapid and slow decomposition) are appropriate representation of decomposition kinetics for use in computer simulation models.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. T. Gilmour Estimating Yield and Yield Response using Computer Simulation of Plant Available Nitrogen from Soil Organic Matter and Manure Soil Sci. Soc. Am. J., January 21, 2009; 73(1): 328 - 330. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
