| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Hortic. Sci., 305 Alderman Hall, 1970 Folwell Ave., Univ. of Minnesota, St. Paul, MN 55108
b Dep. of Natural Resources and Environmental Sci., 1102 S. Goodwin Ave., Univ. of Illinois, Urbana, IL 61801
* Corresponding author (bphorgan{at}umn.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Abbreviations: DAF, days after fertilization • LFN, labeled fertilizer N • MIT, mineralization-immobilization turnover • PET, potential evapotranspiration • PVC, polyvinyl chloride
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Miltner et al. (1996) studied the fate of 15N-labeled urea applied to Kentucky bluegrass turf and found that only 0.23% of the LFN was collected in the drainage water of lysimeters 1.2 m below the soil surface. The majority of the applied N was taken up by the plant or immobilized in the thatch, but recovery of LFN totaled 64 to 81%, suggesting volatile losses.
Starr and DeRoo (1981) also observed very little N leaching in a study involving application of 15N-labeled (NH4)2SO4 to turfgrass in the northeastern USA. Throughout their experiments, 15N-labeled NO3 was detected in leachate on only one occasion, and total recovery amounted to 64 to 76% of the LFN applied. These authors attributed the LFN loss to denitrification and NH3 volatilization.
Non-field and field studies to measure NH3 volatilization from turf show extremely variable results, depending on the source of N, application rate, temperature, thatch thickness, irrigation and rainfall following application, and soil moisture (Volk, 1959; Nelson et al., 1980; Torello et al., 1983; Bowman et al., 1987; Titko et al., 1987). Bowman et al. (1987) applied 58 kg N ha-1 as urea to a Yolo loam soil (pH 7.3) under Kentucky bluegrass turf and measured volatilization following different irrigation treatments. Without irrigation, 36% of the applied N volatilized, whereas volatilization was only 3% when 4.0 cm of irrigation was applied. One aspect of a turfgrass system that will dramatically affect volatilization is the presence of thatch. Significant urease activity, which hydrolyzes urea to NH3, occurs in the thatch layer (Bowman et al., 1987). Nelson et al. (1980) observed that within 8 d following urea application to a Flanagan silt loam (fine, smectitic, mesic Aquic Argiudolls), 39% of the applied N volatilized as NH3 from cores of Kentucky bluegrass containing 
5 cm of thatch, as opposed to only 5% volatilized from cores having 5 cm of soil and no thatch below the sod.
Another N loss mechanism is denitrification, which involves the reduction of N oxides to N gases. This process is carried out by facultative organisms that in the absence of O2 use N oxides as terminal electron acceptors (Broadbent and Clark, 1965). Denitrification is an important process in the soil, plant, and atmosphere continuum because it is the primary mechanism for return of N2 to the atmosphere (Stevenson and Cole, 1999). With plant productivity frequently limited by N supply, removal of inorganic N by denitrifying microorganisms could adversely affect plant growth and development. Moreover, one of the gaseous products of denitrification (N2O) is a greenhouse gas that has been implicated in stratospheric O3 destruction (Prather et al., 1995).
Relatively few quantitative estimates have been made of the N loss from turfgrass or grasslands through denitrification because of the difficulties associated with measuring N2 emission under field conditions (Steele and Vallis, 1987). Some researchers have used acetylene (C2H2) inhibition to estimate denitrification losses from grass under laboratory conditions (Mancino et al., 1988; Schwarz et al., 1994; Tenuta and Beauchamp, 1995). Mancino et al. (1988), for example, studied the effects of soil moisture content, soil texture, and soil temperature on denitrification losses from a Kentucky bluegrass sod. For silt and silt loam soil types, only 0.1 and 0.4% of the N applied as KNO3 was recovered as N2O when the soil moisture content was 80% at 22°C for 10 d. Above 80% saturation, N losses accounted for 5.4 and 2.2% of the N applied to a silt and silt loam soil, respectively. At 75% saturation, denitrification losses from the silt soil increased linearly with temperature between 22 and 30°C and accounted for 0.02 to 0.11% of the applied N. When the soils were at 100% saturation and the temperature was 30°C, maximal losses from the silt and silt loam soil were 94 and 46% of applied N, respectively. Thus, during periods of saturation in a soil, substantial losses of N could occur by denitrification.
Because recovery of applied fertilizer N to turfgrass is typically less than quantitative, the objectives of this research were twofold: (i) to determine the fate of N applied to turfgrass, including direct measurement of denitrification; and (ii) to determine whether the completeness of recovery of 15N-labeled fertilizer applied to turfgrass is influenced by the presence of plants. The answers to these questions will provide further information on N-cycling dynamics in turf and on environmental impacts of fertilizer N use in turf.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Field Experiments
Two separate experiments were initiated in 1999 by inserting eight PVC cylinders into Kentucky bluegrass turf to a depth of 25 cm using a tractor-mounted hydraulic press. Sampling cylinders were constructed of 20-cm diam. PVC pipe cut to 30-cm lengths and equipped with a plastic flange to permit atmospheric sampling. A full description of the materials and methods used for constructing and inserting the modified PVC cylinders is provided by Horgan et al. (2001). Six PVC cylinders were selected after verifying that infiltration rates inside and outside the cylinder did not differ. At 0600 h on 5 May and 9 August 1999, KNO3 containing 98.5 atom % 15N (obtained from Isotec, Miamisburg, OH) was applied in solution to each plot at a rate of 4.88 g N m-2 (equivalent to 49 kg N ha-1) using a polyethylene wash bottle. To ensure complete transfer of the fertilizer solution, the wash bottle was rinsed three times with a total of 165 mL of deionized water.
Plots were irrigated twice a week to replace 80% of the potential evapotranspiration (PET) when rainfall totals did not exceed this value (PET values obtained from the Illinois State Water Survey). The turf was maintained at 5 cm using a pair of manual hand clippers, and a hand-held vacuum was held against the clippers in order to collect clippings as quantitatively as possible. Clippings were collected at biweekly intervals and were promptly dried in a forced air oven at 60°C for 72 h. Prior to analysis, the samples were ground to pass a 0.15-mm screen.
To accomplish atmospheric sampling as described by Horgan et al. (2001), brass lids were secured to the plastic flange on the PVC cylinder using a silicone gasket to create a gas- tight seal, and the lid was left in place for 3 h to trap the gases evolved from the soil and plants. Following this period of enclosure, a closed-loop circulating system was created by attaching a circulating pump and a gas sampling tube containing Ne to the brass lid. The air inside the closed chamber was thoroughly mixed by pumping for 20 min and a representative gas sample was collected in the sampling tube originally employed to introduce Ne. Analyses for 15N-labeled N2 and N2O were described by Mulvaney and Kurtz (1982), and for Ne as described by Horgan et al. (2001), using a Nuclide Model 3-60-RMS mass spectrometer (Spectromedix Corp., State College, PA).
To construct a mass balance of applied fertilizer N, intact PVC cylinders were extracted from the field upon completion of a 6-wk period in the spring and a 4-wk period in the summer. Each PVC cylinder was split longitudinally to expose the soil core within. The top segment of the core (verdure) was separated from the rest of the core at the soil-plant interface. Verdure samples were dried in a forced-air oven at 60°C for 72 h, and were then ground to pass a 0.15-mm screen prior to analysis. The remainder of the soil core was sectioned by depth into the following increments; 0 to 5 cm, 5 to 10 cm, 10 to 20 cm, and >20 cm. Each section was weighed and then transferred to deep-freeze storage (-20°C) until analysis.
Prior to analyses for N and 15N, soil sections were transferred to a forced-air oven at 60°C, and drying was carried out until no change in mass could be detected for a 24-h period. The sections were then weighed and pulverized with a hand-grinding mill, further ground and homogenized with a ball mill for 24 h, and finally, disk-milled so the entire sample passed through a 0.15-mm screen. In addition, the weekly clippings collected from a single PVC cylinder were combined with the corresponding verdure sample, and the mixture was thoroughly homogenized. Analyses for total N and 15N were performed on four 50-mg subsamples of plant tissue per PVC cylinder and on four 500-mg subsamples per soil section. These analyses involved Kjeldahl digestion by a semimicro method using a pretreatment with Fe and KMnO4 to recover (NO3 + NO2)-N (Bremner, 1996), followed by diffusion of the digest after treatment with NaOH in a Mason jar. The diffusion method employed was essentially that of Stevens et al. (2000), but with modifications to permit analysis of the entire digest. Following acidimetric titrations to determine total N, samples were processed as described by Mulvaney et al. (1997a) for N-isotope analysis with an automated Rittenberg system (Mulvaney et al., 1990, 1997b; Mulvaney and Liu, 1991).
Laboratory Experiment
A laboratory experiment was initiated to determine whether complete recovery of fertilizer 15N could be achieved 24 h after fertilization using the soil and plant preparation and analysis procedure previously described. Three bare soil cores and three cores containing soil under turfgrass were extracted from the field using a 91.5-cm2 cup-cutter 12.5 cm deep. Each core was fertilized with KNO3 (2.14 atom % 15N) at 44.69 mg N plot-1 (equivalent to 8.0 kg N ha-1). The fertilizer was applied by pipetting 10 mL of an aqueous solution, and the pipette was then rinsed twice with deionized water (20 mL total) to ensure a complete transfer. After 24 h, cores were frozen at -20°C. Plant and soil sample preparation and total N and 15N analyses followed the procedures previously described. No atmospheric samples were collected.
Greenhouse Experiment
A greenhouse experiment was initiated by inserting six PVC cylinders into the field at a location adjacent to the site of the experiments previously described, of which three were inserted into bare soil and three into soil under Kentucky bluegrass turf. Two cylinders of each type were selected after verifying that infiltration rates inside and outside the cylinder did not differ. The intact cylinders were removed from the soil, and the bottoms were sealed by inserting modified PVC end caps equipped with a 0.95-cm i.d. by 4.4 cm in length stainless steel male-hose connector (cat. no. 6-HC-1-4, Swagelok Co., Solon, OH) to permit leachate collection. The sealed cylinders were transported to the greenhouse, and the plants and/or soil inside the cylinders were treated at 0800 h with a solution of KNO3 containing 98.5 atom % 15N and supplying 4.88 g N m-2 (equivalent to 49 kg ha-1). Fertilization was performed as previously described for the field studies.
Atmospheric sampling commenced following fertilization and occurred daily from 1100 to 1400 h during the period of 24 May to 13 June 2000. Irrigation was applied with a polyethylene wash bottle at least once a week for 3 wk to maintain adequate turfgrass health with 14-h days (185 mmol sec-1 m-2 plus ambient sunlight) at 22 ± 2°C and 10-h nights at 18 ± 2°C. Turfgrass was maintained biweekly at 5 cm using manual hand clippers, and clippings were collected as previously described. In addition, leachate was collected twice a week. Total volume of leachate was recorded, and 100 mL was transferred to a 125-mL polypropylene screw-cap bottle and frozen at -20°C.
At the conclusion of the experiment, the PVC cylinders were split longitudinally to expose the soil core within. The top segment of the cores containing turfgrass (verdure) was separated from the rest of the core at the soil-plant interface. Verdure samples were dried in a forced air oven at 60°C for 72 h and ground to pass a 0.15-mm screen prior to analysis. The remainder of the soil core was sectioned by depth into the following increments: 0 to 5 cm, 5 to 10 cm, 10 to 20 cm, and >20 cm. The entire section was then weighed and frozen (-20°C). Plant and soil preparation, and analyses for total N and 15N were performed by the procedures previously described. The same analyses were performed with four replications on 10-mL aliquots of leachate samples.
Statistical Analysis
For the spring and summer field experiments, data obtained by mass spectrometric analysis for N2 and N2O were extrapolated to a 24-h period assuming uniform emission rates, and means for replicate plots are reported with a corresponding standard deviation. Total N for plant, soil, and leachate was determined in the laboratory with four replications followed by duplicate isotope-ratio analyses of each replicate sample. Data for replicate plots were averaged, and are reported with a corresponding standard deviation.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
|
Recovery of LFN in clippings ranged from 27 to 32% for the spring and from 20.5 to 22.3% for the summer (Table 1). Differences in fertilizer N recovery in clippings between seasons occur as plant growing conditions at the time of fertilizer application will affect plant uptake, which may explain why lower LFN recoveries were observed in the summer because of higher soil temperatures (Fig. 1). Plant stress during hot summer months can have an effect on fertilizer N recovery in clippings (Petrovic, 1990) and with numerous factors influencing N uptake by a plant, comparing research from various experiments is somewhat difficult. However, similar results were reported by Miltner et al., (1996) and Starr and DeRoo (1981), in which case recovery in clippings across a 2-yr period totaled 35 and 30% of LFN, respectively.
Mineralization-immobilization turnover (MIT) of N occurs naturally in soil, one result being a progressive decrease in the 15N enrichment of mineral N following 15N fertilization. It is possible that the lower percentages of 15N found in the turf during the summer experiment may have been due to MIT, owing to active microbial cycling of mineral N. If so, an increased recovery of LFN would have been expected for the 0- to 5-cm soil section. This was not in fact observed; therefore, the smaller amounts of LFN found in turf plants during the summer experiment can probably be attributed to removal of soil NO3 through denitrification.
Although a significant amount of LFN was recovered as N2 or N2O, recovery of LFN was far from complete in our work (Table 2) and did not exceed 73.1%. Total recovery of LFN ranged from 65 to 73% for the spring experiment and from 57 to 68% for the summer experiment. As noted by Miltner et al. (1996), sampling and mixing of soil extracted from the field can contribute substantially to variation in recovery data. To minimize such variation, preliminary studies were conducted to compare the precision achieved by different methods of processing soil samples for recovery of LFN (data not shown). The results from these studies showed that greatest precision was achieved by extracting the entire soil core (several kilograms) from the field without subsampling, oven-drying at 60°C, hand grinding to pulverize large soil aggregates, ball milling for 12 h to homogenize and further grind the soil, and lastly, disk milling so that the entire soil core would pass through a 0.15-mm screen. With this technique, the coefficient of variation did not exceed 1% when total-N and 15N analyses were performed on 10 replicate 500-mg samples of soil.
|
25% of that applied. In the spring experiment, recovery of LFN decreased consistently with greater depth, whereas a larger percentage of LFN was found in the 10- to 20-cm soil section than in the 5- to 10-cm section during the summer experiment. The latter finding can likely be attributed to downward movement of labeled NO3, as a result of the 8.9-cm rainfall event 4 DAF. The risk of such movement cannot be avoided when using a N fertilizer source that is readily soluble, and one effect will be a decrease in plant N availability. In addition, recovery of LFN in the 0- to 5-cm soil section would have been reduced by denitrification, which was much more extensive in the summer than in the spring. It is generally accepted that microbial activity is higher in the surface soil than at greater depths (Speir et al., 1984) because C compounds released from plants serve as an energy source for microorganisms. For this reason, denitrification losses typically decrease with depth (Yeomans et al., 1992; Sotomayor and Rice, 1996; Luo et al., 1998; Richards and Webster, 1999), which helps to further explain why recovery of LFN was reduced in the 0- to 5-cm soil section during the summer experiment. Only 0.5 to 1.7% of the LFN was found below 20 cm, which suggests negligible leaching of fertilizer NO3, and is consistent with similar findings from numerous other studies (Nelson et al., 1980; Snyder et al., 1981, 1984; Starr and DeRoo, 1981; Petrovic et al., 1986; Gold and Groffman, 1993; Miltner et al., 1996; Owens et al., 1999). According to our original hypothesis, denitrification accounts for the incomplete recovery observed when LFN is applied to turfgrass. Although our research demonstrates the potential for substantial N loss as N2 or N2O, our recoveries were disconcertingly low and are not readily explained (Table 2). These finding suggests that denitrification does not account for the 15N deficits observed in the present or previous studies. To determine whether the methods employed in our work were adequate, a laboratory experiment was conducted to compare LFN recovery from a bare soil and a turfgrass system 24 h after fertilizer application. In both cases, recovery of LFN was essentially complete at 100.2 ± 1.9 and 98.7 ± 9.6%, respectively. However, the relatively high standard deviation for the plant-based system suggests the possibility that N loss may have occurred by other processes beside denitrification.
Since our efforts to obtain a mass balance of applied fertilizer N were not achieved in the field (Table 2), a greenhouse experiment was initiated to compare recovery of LFN for a bare soil and a turfgrass system. Unlike the field experiments, this experiment was designed to measure the loss of labeled N by leaching as well as by denitrification. Recovery of LFN for the bare soil was 84.2 ± 16.4%, as compared with 70.6 ± 9.9% for the turf.
Plant-based systems differ inherently from bare soil, in that roots are constantly aerating the soil surface, evapotranspiration is occurring, plant senescence supplies microorganisms with organic C as an energy source, and nutrients are removed from the soil via plant uptake. A further difference arises for high maintenance turfgrass, in that irrigation is typically applied daily. Because of these differences between turf and bare soil, LFN losses as N2 and N2O were greater from turfgrass than from bare soil (Table 3) , and totaled 19% for turf as compared with 7% for bare soil. These results are consistent with previous work by Larsson et al. (1998), who found that emission of N2O from a grass sward (6 kg N2O-N ha-1) greatly exceeded emission from a bare soil (0.2 kg N2O-N ha-1). With bare soil, the lower N emission rates observed in our work are likely due to infrequent wetting of the soil surface by irrigation, which kept the soil profile drier and therefore more aerobic; moreover, these plots would have contained less available C, as compared with soil under turfgrass. The experiment in the greenhouse only involved irrigation approximately once a week to maintain adequate turfgrass growth (vs. replacing 80% of PET for the field experiments twice a week). Rolston et al. (1982) have suggested that less frequent irrigation moves the fertilizer N deeper into the root zone, resulting in less NO3 in the upper part of the soil profile where high C and high water contents may occur simultaneously.
|
Recovery of LFN from turfgrass for the field and greenhouse experiments reported here ranged from 57 to 73% (Table 2), even when taking into account N losses due to denitrification and leaching. Results from our work suggest that turfgrass, or plants in general, lead to large deficits in recovery of LFN; however, a deficit was also obtained for bare soil, which may reflect errors associated with the sampling procedure for N2 and N2O. Data extrapolation from 3-h measurements of N2 and N2O emission to a 24-h period is a possible source of error in our recovery estimates (Velthof et al., 2000). However, because denitrification is temperature-dependent, emissions of N2 and N2O would tend to be greatest during the day when the sun provides radiative heating of the soil surface (one possible exception to this theory would be during the summer experiment when soil temperatures may be greatest following the last atmospheric sample collection period). Although data extrapolation no doubt led to some error in estimating these emissions, the effect was probably to slightly overestimate gaseous N loss because atmospheric sampling occurred from 0800 to 1100, from 1100 to 1400, and from 1400 to 1700 h. The methods employed in our work would not have detected gaseous loss of NH3, but such loss is unlikely to have been appreciable, given the fact that N was applied as NO3 to a fine-textured acidic soil. Some N loss as NH3 may have occurred from senescing plant tissue or from sod as previously noted by many investigators (Volk, 1959; Watkins et al., 1972; Torello et al., 1983; Bowman et al., 1987; Titko et al., 1987; Gooding and Davies, 1992). Moreover, the possibility cannot be excluded that other N compounds may have volatilized from leaves, such as amines, oxides of N, HCN, oximes, and some alkaloids (Wetselaar and Farquhar, 1980).
|
ACKNOWLEDGMENTS |
|---|
Received for publication June 5, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. C. Knight, E. A. Guertal, and C. W. Wood Mowing and Nitrogen Source Effects on Ammonia Volatilization from Turfgrass Crop Sci., July 30, 2007; 47(4): 1628 - 1634. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. W. Frank, K. M. O'Reilly, J. R. Crum, and R. N. Calhoun The Fate of Nitrogen Applied to a Mature Kentucky Bluegrass Turf Crop Sci., December 2, 2005; 46(1): 209 - 215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Engelsjord, B. E. Branham, and B. P. Horgan The Fate of Nitrogen-15 Ammonium Sulfate Applied to Kentucky Bluegrass and Perennial Ryegrass Turfs Crop Sci., July 1, 2004; 44(4): 1341 - 1347. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. M. Easton and A. M. Petrovic Fertilizer Source Effect on Ground and Surface Water Quality in Drainage from Turfgrass J. Environ. Qual., March 1, 2004; 33(2): 645 - 655. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Steinke and J. C. Stier Nitrogen Selection and Growth Regulator Applications for Improving Shaded Turf Performance Crop Sci., July 1, 2003; 43(4): 1399 - 1406. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
Values in parentheses are standard deviations.