Can soil amendments (zeolite or lime) - CiteSeerX

0 downloads 0 Views 213KB Size Report
Abstract. To determine the effects of soil amendments (lime or ammonium-sorbed zeolite) on ... In the pasture soil, the application of urine, urea, and soil amendments significantly affected daily and total N2O ... and wetland soils to meet the Kyoto protocol demand. ... floating glaucous sweet grasses (Glyceria declinata Breb.) ...
CSIRO PUBLISHING www.publish.csiro.au/journals/ajsr

Australian Journal of Soil Research, 2007, 45, 543–553

Can soil amendments (zeolite or lime) shift the balance between nitrous oxide and dinitrogen emissions from pasture and wetland soils receiving urine or urea-N? M. ZamanA,E , M. L. NguyenB , F. MathesonC , J. D. BlennerhassettA , and B. F. QuinD A

Summit Quinphos, PO Box 24-020, Royal Oak, Auckland, New Zealand. Soil and Water Management & Crop Nutrition, Joint FAO/IAEA Division of Nuclear Techniques in Food & Agriculture, PO Box 100, A-1400 Vienna, Austria. C National Institute of Water & Atmospheric Research Ltd, Gate 10, Silverdale Road, PO Box 111-15, Hamilton, New Zealand. D Quin Environmentals (NZ) Ltd, PO Box 125-122, St Heliers, Auckland, New Zealand. E Corresponding author. Email: zamanm [email protected] B

Abstract. To determine the effects of soil amendments (lime or ammonium-sorbed zeolite) on emissions of nitrous oxide (N2 O) and dinitrogen (N2 ) gases from pasture and wetland soils, a 90-day incubation experiment was conducted under controlled moisture and temperature conditions. Soil samples (0–0.10 m soil depth) collected from pasture and adjacent wetland sites were treated with 2 nitrogen (N) sources (cow urine or urea) at 200 kg N/ha with and without added soil amendments using 10-L plastic containers and then incubated at 25◦ C. Subsoil samples were taken out at different intervals to measure gaseous emissions of N2 O and N2 using the acetylene (C2 H2 ) inhibition method, ammonium (NH4 + ), nitrate (NO3 − ), soluble organic C, and pH. The anaerobic conditions (81% water-filled pore space) in wetland soils precluded nitrification, and therefore no increase in NO3 − , N2 O, or N2 was observed during the 90-day incubation period. In the pasture soil, the application of urine, urea, and soil amendments significantly affected daily and total N2 O and N2 emissions and their ratios over a 90-day incubation period. Total N2 O emission from urea-treated soil (48 kg N2 ON/ha) was significantly higher than from urine-treated soil (39 kg N2 O-N/ha) and the control soil (4.5 kg N2 O-N/ha). The application of zeolite significantly reduced N2 O emissions from urea and urine-treated soils by 45% and 33%, respectively, due to the sorption of NH4 + by zeolite. Liming had minor effect on N2 O emission. However, when lime was applied with zeolite, a significant reduction in N2 O emission was observed. Lime application alone was found to increase N2 emissions in urine and urea treated soils by 46% and 35%, respectively, and thereby lower N2 O : N2 ratios. The results indicate that zeolite reduced N2 O emission while lime increased N2 emissions and lowered N2 O : N2 ratios, and warranting further attention for mitigation of N2 O. Additional keywords: N2 O : N2 , urine, urea, lime, zeolite, pasture, wetland, N2 O : N2 ratio, mitigation.

Introduction Nitrous oxide, a long-lasting greenhouse and potential ozone (O3 ) depleting gas, makes up about 20% of New Zealand’s total greenhouse gas emissions inventory, and predominantly emits from urine patches in intensively grazed pastures (de Klein and Logtestijn 1994; de Klein et al. 2003). Urine patches in intensively grazed pastures cover about 20–25% of the area on an annual basis and contain significant amounts of N, ranging from 250 to 1000 kg N/ha (Haynes and Williams 1993; Jarvis et al. 1995). Riparian wetlands, natural or constructed areas of undrained land directly adjacent to a stream or river channel, can remove up to 90% of NO3 − -N and other nutrients entering riparian zones from adjacent upland agricultural catchments via surface runoff or seepage flows (Cooper et al. 1997) and may thus emit considerable amounts of N2 O. However, little is known about N2 O emissions from riparian wetlands. © CSIRO 2007

Denitrification, the enzymatic reduction of NO3 − by aerobic heterotrophs in the presence of available C under anaerobic conditions, produces both N2 O and N2 (Tiedje 1988), while nitrification and dissimilatory NO3 − reduction to NH4 + (DNRA) produce only N2 O as a byproduct (Tiedje 1988; Firestone and Davidson 1989; Silver et al. 2001). The 3 microbial processes can occur simultaneously in soils and sediments across the landscape depending on physical and chemical conditions in their micro-sites. There is no evidence to suggest that nitrification or DNRA directly produces N2 . Nitrous oxide produced by various processes might form one pool before being reduced to N2 by N2 O-reductase enzyme. However, limited information is available about the bulk reduction of N2 O to N2 by N2 O-reductase, whose activity is regulated by various soil management and environmental factors (Stevens and Laughlin 1998; Bouwman et al. 2002). Unlike N2 O, emission of 10.1071/SR07034

0004-9573/07/070543

544

Australian Journal of Soil Research

N2 to the atmosphere is not associated with any environmental problems and it can be retuned back to the soil through symbiotic and non-symbiotic N2 fixation. The production of N2 O in soil is reported to be mainly controlled by the availability of NH4 + (substrate for nitrification) and NO3 − (substrate for denitrification and DNRA). Other known factors include soluble organic C, water-filled pore space (WFPS), O2 concentration, and soil pH (Linn and Doran 1984; Tiedje 1988; Firestone and Davidson 1989; Bremner 1997; Flessa et al. 1998; Wolf and Russow 2000; Sanchez et al. 2001; Cai et al. 2002). The Kyoto protocol, signed by New Zealand in December 2002, makes New Zealand legally bound to reduce its greenhouse gas emissions below the 1990 level during 2008–2012 to avoid the purchase of carbon credits. However, the sharp increase in fertiliser N inputs to grazed pastures since 1990 (New Zealand’s Greenhouse Gas Inventory Report, Ministry for the Environment 2005), coupled with the recent rapid intensification of dairy farming, has led to calculations that New Zealand will have to purchase approximately NZ$400m of carbon credits, even after allowing for oil-price induced reductions in petrol use. Some N2 O emission from soils is inevitable as it is part of the complex and dynamic N cycle; therefore, research work is needed to identify mitigation options from both pasture and wetland soils to meet the Kyoto protocol demand. Several mitigation options have been proposed to reduce N2 O emission from intensive grazed pasture systems, e.g. applying nitrification inhibitor (dicyandiamide or DCD) to retain N in NH4 + form (Di and Cameron 2003), applying both DCD and urease inhibitors (nBTPT, or trade name ‘Agrotain’) to slow urea hydrolysis and to further decrease nitrification (Singh et al. 2006; Zaman et al. 2006), or applying urea fertiliser with Agrotain and DCD (Dobbie and Smith 2003; Zaman et al. 2005), and zero or restricted grazing during the wet winter period to avoid soil compaction (Bhandral et al. 2005; Monaghan et al. 2005). However, it is not well known whether these measures can shift the N emission from predominantly N2 O to N2 . Zeolite comprises natural alumino-silicates, which are negatively charged and capable of sorbing NH4 + -N onto their surfaces (Nguyen and Tanner 1998), and it therefore has the potential to decrease N2 O emission. Similarly, lime is commonly applied to pasture soils every 3–5 years to enhance productivity and increase soil pH. High soil pH is known to affect the activity of N2 O-reductase (Flessa et al. 1998) and may therefore affect N2 production. The objective of this study was to identify whether soil amendments (lime or zeolite) have any impact on N2 O and N2 emissions and their ratios in pasture and wetland soils receiving different sources of N (urine v. urea). Materials and methods Site description The pasture and wetland sites were situated in the dairy catchment at Toenepi, (37◦ 44 S, 175◦ 35 E), which is located approximately 32 km from Hamilton, New Zealand. The permanent-pasture vegetation, oversown after burning of the original scrub at least 70 years ago, consisted predominantly of ryegrass (Lolium perenne L.) and some white clover (Trifolium

M. Zaman et al.

repens L.) grazed by dairy stock (3 cows/ha). The seepage wetland (6817 m2 ) is located at a footslope of the pasture site and intercepts both surface runoff and seepage flow from adjacent pasture site. This wetland has been excluded from grazing animals for over 2 years. The wetland vegetation consists mainly of soft brome (Bromus hordaceus L.) with some floating glaucous sweet grasses (Glyceria declinata Breb.) and soft rush (Juncus effuses L.) and wiwi (Juncus edgariae L.). The climate at the site is humid–temperate with mean annual temperature of 13.7◦ C and rainfall of 1614 mm. The pastoral soil is predominantly Topehaehae silt loam (i.e. silt loam in topsoil with blocky clay loam at 0.30–0.75 m depth) derived from volcanic ash alluvium (an Orthic gley soil, NZ soil subgroup; Aeric Haplaquent, USDA Soil Taxonomy). The wetland soil is composed of loose organic material of approximately 0.50–0.75 m deep, which has probably been formed from sediments and organic matter from the pastoral site as well as organic matter decay from wetland vegetation. Consolidated dense bluish-grey clay of low permeability exists at a depth of 0.75–1.00 m. Soil sampling and chemical analyses In October 2002, 12 composite soil samples (0–0.10 m depth), each composite soil sample comprising 30 randomly collected soil cores, were collected from each pasture and wetland sites, bulked, and passed through a 4-mm sieve to remove visible plant litters and roots. Soil bulk density was measured using 4 soil cores. Four sieved soil samples from each site were analysed for WFPS, pH, mineral N, total N, total C, available Olsen P and other cations (Blakemore et al. 1987), nitrification potential (Schmidt and Belser 1994), and denitrification enzyme activity (DEA) (Tiedje 1982). To measure nitrification potential, sieved soil samples were first extracted with 2 M KCl (1 : 10 ratio) to determine the initial NO3 − concentrations. Two sets of sieved soil samples (100 g each) were then incubated with or without 2 mL of solution of 25 mg N/mL as (NH4 )2 SO4 in the dark at 25◦ C. Soil subsamples were taken out at different intervals, extracted with 2 M KCl, filtered, and analysed for NO3 − -N concentration on a flow injection analyser (FIA). DEA was measured by incubating 25-g sieved soil samples with 25 mL of solution of 1 mM glucose, 1 mM KNO3 , and 1 g/L of chloramphenicol in gas jars. Purified C2 H2 was injected into gas jars through a septum to achieve the final concentration of 10% (10 kPa) in the gas phase. The gas jars were placed on a rotary shaker in the dark and gas samples were collected at different timings for N2 O analysis. Cow urine was collected from Frisian cows during milking and stored below 4◦ C before application to avoid urine decomposition. Four urine samples were analysed for total N, mineral N, urea-N, and total C. Cow urine contained 4.69 g/L of urea-N (79% of the total N), 26 mg/L of NH4 + -N, 0.75 mg/L of NO3 − -N, 5.92 g/L of total N, and 10.7 g/L of total organic C. Incubation procedure Sieved soil samples providing the equivalent of 3 kg oven-dry soil for each soil type were transferred to a plastic bucket (10 L) with a lid. Cow urine at 104.5 mL and 194.9 mL was added to each bucket of pasture and wetland soils, respectively, to achieve an N application rate of 200 kg N/ha. A total of 72 plastic buckets

Soil amendments and nitrous oxide and dinitrogen emissions

were used for both soil types. Urea was first dissolved in water and then applied in a volume of solution equivalent to that of the urine. The control (no N) treatments were treated with an equivalent volume of water only. Soil amendments (lime, zeolite, and lime plus zeolite) were then added to appropriate treatments and thoroughly mixed to achieve uniformity. Lime was applied to raise the soil pH to 7. The amounts of lime applied to each pasture and wetland soil were 47 g (15 t/ha) and 118.5 g (20 t/ha), respectively, as determined from a pre-incubation experiment. The amounts of zeolite applied to each pasture and wetland soils were 108.5 and 134.9 g, respectively. This application rate was based on our previous findings that 1 kg of clinoptilonite can sorb 5.7 g NH4 + -N (Nguyen and Tanner 1998). Zeolite had a density of 1.2 g/cm3 , porosity 60%, slurry pH 5–6 for 20% w/v, cation exchange capacity (CEC) 80 cmol/kg, internal surface area 32 m2 /g, and pore size/diameter ∼6A. WFPS in pasture and wetland soils after addition of N inputs and soil amendments remained unchanged at 58% and 81%, respectively. The soil mixtures in each plastic bucket were incubated at 25◦ C for 90 days. To allow exchange for accumulated gases in plastic buckets, the lid from each bucket was removed twice a day for 5 min. The soil moisture losses during incubation were minimal; however, each bucket was weighed every week to correct the soil moisture content by adding de-ionised water. Acetylene inhibition, gas sampling, and chemical analyses The C2 H2 inhibition technique described by Tiedje et al. (1989) was used to measure N2 O and N2 emissions from each treatment. This technique has some biological implications, for example it needs paired soil samples (with or without C2 H2 ), meaning that it is laborious. A small amount of C2 H2 (1%) can block nitrification and thus can underestimate denitrification in NO3 − -limited environments. Denitrifiers after repeated or long exposure to C2 H2 could adapt themselves to added C2 H2 and use it as a source of C, which is likely to stimulate denitrification rates. Acetone, which is added to C2 H2 as a stabiliser, could stimulate denitrifiers by providing an extra source of C. The most challenging step is the uniform distribution of the desired concentration of C2 H2 in a soil profile and in microsites inhabited by relevant microorganisms; however, this may be overcome by using sieved soils. Two sets of soil subsamples providing the equivalent of 25 g (oven-dry each) were weighed from each plastic bucket at different timings over a 90-day period and transferred to 1-L glass jar. The lid of each jar had a rubber septum for injecting C2 H2 /air and gas sampling. Fifty mL purified C2 H2 was injected into one set of gas jars using a 60-mL syringe to achieve the final concentration of 10% (10 kPa) in the gas phase. After injecting C2 H2 , the syringe was pulled forward and backward a few times to ensure the diffusion of C2 H2 . Air was injected into the other set of soil samples in a similar way. After venting, the 2 sets of soil samples were then incubated at 25◦ C for 24 h. A 20-mL gas sample was taken from each jar after 30 min of C2 H2 /air injection and transferred into a pre-evacuated 12-mL exetainer. After collecting the first gas sampling, 20 mL of air was injected back into each jar to maintain uniform pressure. After 24 h, gas samples were taken from each jar in a similar way. Gas samples were analysed for N2 O concentration on a gas chromatograph (Shimadzu GC 17A

Australian Journal of Soil Research

545

Japan) equipped with a 63 Ni-electron capture detector operating at column, injector, and detector temperature of 55, 75, and 330◦ C respectively. The amount of N2 was calculated from the difference in N2 O production between C2 H2 -treated and nonC2 H2 -treated soils. Both C2 H2 and non-C2 H2 treated soils were discarded after 24 h of incubation. Additional subsoil samples were also removed from each bucket at each sampling interval and analysed for mineral N (NH4 + , NO2 − , and NO3 − ), soluble organic C, and pH. To measure soil mineral N, a soil sample equivalent of 5 g oven-dry soil was extracted with 50 mL of 2 M KCl for 1 h on a rotary shaker followed by filtration. Soil extracts were either immediately analysed for NH4 + , NO2 − , and NO3 − using FIA, or frozen until analyses were performed. Soluble organic C was measured by shaking a soil sample equivalent of 5 g oven-dry soil with 20 mL 0.5 M K2 SO4 for 30 min on a rotary shaker followed by filtration (Burford and Bremner 1975). Soil extracts were either immediately analysed for organic C on a LECO CNS Analyser 2000 or frozen until analyses were performed later. Soil pH was determined by shaking 10 g of moist soil with 20 mL deionised water (soil : de-ionised water suspension of 1 : 2 ratio) on a rotary shaker for 30 min and then measured with an electrode (Blakemore et al. 1987). Statistical analyses The 3 N treatments (no N, cow urine, and urea each applied at 200 kg N/ha) and 4 soil amendment treatments (no amendments, zeolite, lime, and lime plus zeolite) were replicated 3 times in a 3 × 4 × 3 factorial experiment for each soil type. Repeatedmeasure analysis of variance (ANOVA) was carried out to determine the effect of time on different parameters. ANOVA was carried out for each individual sampling time by including N, zeolite, lime, lime plus zeolite, and their 2-way and 3-way interactions in the ANOVA model. Least significant difference (l.s.d.) values at P = 0.05 were calculated when the interaction effects were significant. All the analyses were performed using SYSTAT (1994). Results Physical and chemical properties of pasture and wetland soils Both pasture and wetland soils are located in the same catchment area but they exhibited substantial differences in most physical and chemical properties (Table 1). The wetland soil received seepage water from the adjacent pasture soil and was therefore flooded and had WFPS of 81%, compared with 58% in pasture soils. The total N and C, available Olsen P, magnesium (Mg) and CEC of wetland soil were higher than those of the pasture soil (Table 1). Wetland soil had higher DEA and low nitrification potential than the pasture soil; therefore, it was presumed that the 2 soils would behave differently in terms of N2 O emissions when treated with 2 N sources (urine and urea) and soil amendments (zeolite, lime, and lime plus zeolite). Mineral N in pasture and wetland soils Soil NH4 + concentrations of pasture soil differed significantly (P < 0.001) and were influenced by the 2 N inputs and soil amendments during the incubation (Fig. 1a–d). Ammonium accounted for 38–60% of the mineral N and its concentrations

546

Table 1.

Australian Journal of Soil Research

M. Zaman et al.

Some soil physical and chemical properties of wetland and pasture soils Mean values from 4 replicates are shown

Analyses type

pH NH4 + (mg N/kg soil) NO3 − (mg N/kg soil) Total N (%) Total C (%) Olsen P (mg/L) Exchangeable cations (cmol/kg) Potassium Calcium Magnesium Sodium CEC Base saturation (%) Soil bulk density (g/cm3 ) Water filled pore space (WFPS) (%) Lime required to raise soil pH by 1 unit (t/ha) DEA (mg N2 O-N/kg soil.day) Nitrification potential (mg NO3 − -N/kg soil/day)

Pasture soil

Wetland soil

5.6 6.8 5.6 0.44 4.50 17

5.3 62 0.6 0.50 6.1 20

0.91 5.0 1.4 0.11 12 63 0.97 58 11.6 18.7 10

0.80 4.9 2.50 0.35 16 53 0.52 81 13.2 205 0.9

lime plus zeolite treatments. At day 30 in urea-treated soil, soil NH4 + concentration was 185 mg N/kg soil with no zeolite compared with 67 mg N/kg soil with zeolite. In both urea- and urine-treated soils, urea hydrolysis was completed on day 1 as evident from their NH4 + concentrations. In pasture soils, NO3 − concentrations in urine- and urea-treated soils followed a similar trend, increased slowly during the first week of incubation, reached their maximum at day 30, and then decreased afterward (Fig. 2). Nitrate concentration in urine- and urea-treated soil accounted for 40–62% of the mineral N. Urea-treated soils produced significantly (P < 0.01) higher NO3 − concentrations than urine-treated soils. Compared to soil with no amendments, zeolite application significantly (P < 0.01) decreased NO3 − concentration, whereas lime application significantly (P < 0.01) increased NO3 − concentration. Lime plus zeolite resulted in intermediate NO3 − levels. In wetland soils, soil NH4 + concentrations increased after the applications of urine and urea (Fig. 3a–d) and remained the major form of mineral N (99% of the mineral N) in most treatments throughout the incubation. Zeolite application significantly (P < 0.001) lowered NH4 + concentrations (Fig. 3b) compared with soils with no amendments (Fig. 3a), whereas lime application had no such significant effect of lowering NH4 + concentrations (Fig. 3c) compared with soils with no amendments. Lime application appeared to reduce the effect of zeolite in the combined lime plus zeolite treatment (Fig. 3d). In wetland soil, NO3 − concentrations in all treatments were