Copyright is owned by the Author of the thesis. Permission is given for a copy to be downloaded by an individual for the purpose of research and private study only. The thesis may not be reproduced elsewhere without the permission of the Author. Tillage-Induced Soil Nitrous Oxide Fluxes from Two Soils in the Manawatu A THESIS PRESENTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE IN AGRICULTURAL ENGINEERING INSTITUTE OF TECHNOLOGY AND ENGINEERING MASSEY UNIVERSITY, PALMERSTON NORTH NEW ZEALAND AKMALAKRAMKHANOV April, 2000 Abstract Enhanced greenhouse gas emissions of nitrous oxide (N2O) induced by agricultural practices is believed to be the major anthropogenic source. Studies conducted in New Zealand generally from pasture suggest low N2O emission, however, there is little information for arable farming systems. Therefore, there is a need for a site-specific assessment of the impact of tillage practices on N2O fluxes. This paper evaluates tillage system and land use effects on N2O emissions at two sites using a closed chamber technique. Sites included a Kairanga silt loam where maize/barley was grown continuously for either 17 (Kl 7) or 34 (K34) years, with a conventional tillage system (Kairanga), and an Ohakea silt loam where winter oats and summer fodder maize was double-cropped for five years with conventional (CT) and no-tillage (NT) systems (Massey). At both sites permanent pasture (PP) soil was used as a control. Spatial measurements for all treatments at Massey site showed large inherent variations in N2O fluxes (a mean CV=l 19%) which reflected natural soil heterogeneity , and perhaps the measurement technique used rather than the real differences due to the tillage and cropping systems evaluated. N 2O emissions measured from December 1998 to September 1999 from the PP were significantly lower (1.66 kg N 2O-N/ha/year) than the CT and NT plots at 9.20 and 12.00 kg N2O-N/ha/year respectively. However, there were no differences in N2O emission rates between the CT and NT treatments. Cumulative coefficient of variation (CV) of treatments ranged from 39 to 140%. Seedbed preparation using power-harrow which was done within few days of ploughing the CT plots reduced N 2O emissions by 65% within the first hour after power­ harrowing. However, N2O emission rates returned to the pre-power harrowing levels one month after power-harrowing. There was strong relationship between log-transformed values of soil moisture content (SMC) and N2O emissions in all treatments, PP (r = 0.73), CT (r = 0.75) and NT (r = 0.86). Seasonal variation in N 2O emission from the PP was in the order of winter=autumn>summer. Although fluxes in the CT were higher in winter than in the autumn season, there were no differences between the summer and autumn data. Similar 11 to the PP, the seasonal variations in N2O emission in the NT treatment were in the order of winter>autumn=summer. The estimated annual N2O emissions from the PP, Kl 7 and K34 (calculated as the mean of all individual closed cover chamber measurements between November 1998 and September 1999) from Kairanga site were similar at 3.24, 3.42 and 2.37 kg N2O­ N/ha/year, respectively. There were large variations in N2O emissions during the year with the mean flux rates ranging from 0.175 to 13.32, 0.175 to 16.91 and 0.088 to 30.05 kg N2O-N/ha/year in the PP, Kl 7 and K34 fields, respectively. Although overall comparison of treatment means did not show any discernible differences between management practices, there were signs that the K34 had lower emissions compared to the PP. N2O fluxes from the Kl 7 and PP field appeared to be influenced by SMC. There is clear indication that low or negligible emissions occur when gravimetric soil water content is less than 30% in the PP. Although N2O fluxes did not follow the rainfall patterns in the K 17 and PP, linear regression analyses indicated low but significant relationship r == 0.46 and 0.53 (0.72 when log-transformed), respectively. In the K34 field, SMC did not seem to govern fluxes which were especially apparent during wet months of April and May. The linear regression analysis using the measured data revealed no relationship (r = 0.12) between the SMC and N2O fluxes in the K34 treatment. Seasonal grouping of monthly log-transformed N2O emissions showed significant differences in all treatments. Summer season N2O emissions in the PP were the lowest than other seasons whereas no discernible differences were observed among other seasons. Although N2O fluxes during spring and summer were similar in the Kl 7 field , they were significantly lower than the winter and higher than autumn fluxes. There were considerably higher emissions in summer than in autumn in the K34 but seasonal variation between winter and spring was less profound. Spatial variability in N2O fluxes was large during the year with coefficients of variation (CV) ranging from 10 to 82%, 12 to 99% and 9 to 137% for the PP, Kl 7 and K34 fields, respectively. lll Acknowledgements I wish to express my sincere thanks to my chief supervisor Assoc. Prof. Ashraf Choudhary and co-supervisor Dr. S. Saggar of Landcare Research for their time, advice, supervision, encouragement and providing ideas during the course of this study. Many thanks to staff and post-graduate students in Agricultural Engineering, Massey University, for their help and friendship including Leo Bolter, Joane Brooks, Roanna Moore and my post-graduate colleagues Edmundo Viegas, Khurshid Anwar and Val de Villa. I also acknowledge assistance from Landcare Research, Palmerston North, for the use of laboratory facilities. Special thanks are due to Messrs Brian Daly, Graham Shepherd, Carolyn Hedley and Jaque for their help and guidance during laboratory experiment. I thank Mr. Duncan Hedderley for his assistance in computer analysis of the data. I am grateful to the New Zealand Official Development Assistance (NZODA) Programme, Ministry of Foreign Affairs and Trade, Wellington for the postgraduate student scholarship. Last, but certainly not least, I thank my parents, family and friends for continuing support and patience throughout my studies. Abstract Acknowledgements Table of Contents List of Tables List of Figures Table of Contents Chapter 1 - General Introduction 1.1 Research Objectives Chapter 2 - Literature Review 2.1 Introduction 2.2 Nitrous Oxide in the Atmosphere 2.3 Sources of Nitrous Oxide 2.4 Impact of Nitrous Oxide on the Environment 2.5 Global Nitrogen Perspective 2.6 Status of Nitrous Oxide Emissions in New Zealand 2. 7 The Role of Micro-Organisms 2.8 Mechanisms of Nitrous Oxide Production 2.8.1 Denitrification 2.8.2 Nitrification 2.8.3 Sinks 2.9 Agriculture as a Major Source of Nitrous Oxide Emissions 2.9.1 Effects of Tillage Practices IV i iii i v v i i i X 1 3 4 4 6 6 7 8 10 12 12 12 13 14 16 16 V 2.9.2 The Role of Fertilisers 17 2.10 Factors Controlling Nitrous Oxide Emission 18 2.10.1 Soil Properties 18 2.10.2 Soil Organic Matter 20 2.10.3 Soil Moisture Content 21 2.10.4 Soil pH 22 2.10.S Temperature 22 2.11 Soil Management Practices Impacting Nitrous Oxide Emission 23 2.12 Field Nitrous Oxide Emission Measurement Methods 24 2.12.1 Chamber Methods 25 2.12.2 Micrometeorological Methods 27 2.12.3 Ultra-Large Chambers with Long-Path IR Spectrometers 28 2.12.4 Nitrous Oxide Emissions Variability 29 2.13 Summary 30 Chapter 3 - Materials and Methods 31 3.1 Experimental Sites 31 3.1.1 Kairanga Site 32 3.1.2 Massey Site 34 3.2 Measurement of Field N20 Emission 37 3.2.1 Procedure 38 3.2.2 Analysis of N2O Concentration 42 3.3 Pilot Experiments 42 3.3.1 Exploratory Sampling 42 3.3.2 Spatial Variability 43 3.3.3 N2O Emissions During Seedbed Preparation 43 3.3.4 Prolonged Continuous Measurement 3.4 Ancillary Measurements 3.4.1 Soil Moisture 3.4.2 Soil Temperature 3.4.3 Soil pH, Total C and N 3.4.4 Rainfall Data 3.4.S Statistical Analysis Chapter 4 - Results and Discussion 4.1 Overview 4.2 Pilot Experiment 4.2.1 Exploratory Sampling 4.2.2 Spatial Variation 4.2.3 Pilot Experiment on Effect of Seedbed Preparation on N2O Emission 4.2.4 Effect of Prolonged Measurements on N2O Flux 4.3 Massey Experimental Site 4.3.1 Effects of Tillage on Soil pH, Total C and N 4.3.1.1 EffectofTillage Techniques on Soil pH 4.3.1.2 Effect of Tillage Techniques on Total C 4.3.1.3 Effect of Tillage Technique on Total N 4.3.2 Tillage Effect on Field N2O Emissions 4.3.3 Summary 4.3.3.1 Effect of Soil Moisture on N2O Emissions 4.3.3.2 Seasonal Variation 4.3.3.3 Spatial Variation 4.3.4 Summary VI 43 43 43 43 44 44 44 46 46 47 47 48 48 so so so 52 52 53 53 59 59 62 65 67 4.4 Kairanga Experimental Site 4.4.1 Effects of Tillage on Soil pH, Total C and N 4.4.2 Tillage Effect on Field N2O Emissions 4.4.2.1 Effect of Soil Moisture on N2O Emissions 4.4.2.2 Seasonal Variation 4.4.2.3 Spatial Variation 4.4.3 Summary Chapter 5 - Conclusions 5.1 General 5.2 Massey Experimental Site (Ohakea Silt Loam Soil) 5.3 Kairanga Experimental Site (Kairanga Silt Loam Soil) References Appendices vii 68 68 68 74 79 82 83 84 84 84 86 88 Table 2.1: Table 2.2: Table 2.3: Table 2.4: Table 3.1: Table 3.2: Table 4.1: Table 4.2: Table 4.3: Table 4.4: Table 4.5: Vlll List Of Tables Atmospheric concentrations of the major greenhouse gases, their 6 rise, residence time and contribution to the global warming. Biogeochemical distribution of Non Earth. 9 Global fluxes of nitrogen into and out of the terrestrial biosphere. 9 Comparison of flux measurements by flux gradient micro- 26 meteorological, combined megachamber/long-path IR, and small chamber gas chromatograph (GC) methods. Summary of monthly climatological observations data taken at 31 09.00 hrs at AgResearch and Aorangi Research stations close to study sites. Selected soil characteristics of the fields at the Kairanga site. 32 N20 emission rates using the exploratory samples in the pilot 47 experiment. Effect of time of analysis on N20 emissions following collection of 48 gas samples from the field. N20 emissions from 12 chambers measured to determine baseline 48 variation. Comparison of N20 fluxes before (March 17) and after (March 18) 49 power harrowing in the CT treatment at the Massey site. Ohakea silt loam soil pH, total C and total N of the Massey 52 experimental site . Table 4.6: The effects of tillage systems on the ranges of N20 emissions from 54 Massey plots. Table 4.7: The effects of tillage systems on the means of N20 emissions from 54 Massey plots. Table 4.8: Effects of tillage systems on the monthly means of N20 emissions 56 data at Massey site (December 1998 to September 1999). IX Table 4.9: Effects of tillage systems on the monthly means of log transformed 57 N2O emissions at Massey site (December 1998 to September 1999). Table 4.10: Effect of tillage techniques on soil moisture content at 0-10 cm 62 depth during N2O emission measurements. Table 4.11: Seasonal field N2O emissions as affected by the PP, CT and NT 65 treatments. Table 4.12: The effects of tillage systems on the ranges of N2O emissions from 68 Kairanga plots. Table 4.13: The effects of tillage systems on the means of N2O emissions from 69 Kairanga plots. Table 4.14: Effects of length of continuous cropping and permanent pasture on 70 the mean monthly N2O emissions at the Kairanga site (November 1998 to September 1999). Table 4.15: Effect of length of continuous cropping and permanent pasture on 71 the mean of log-transformed monthly N2O emissions at Kairanga site (November 1998 to September 1999). Table 4.16: Soil moisture content at 0-10 cm depth during N2O emission 76 measurements. Table 4.17: Correlation of soil moisture content with N2O emission rates at the 79 Kairanga field site. Table 4.18: Seasonal log-transformed field N2O emissions in the PP, CT and 80 NT treatments. X List Of Figures Figure 2.1 Simplified diagram of nitrogen transfers, removals and losses of the 5 agro-ecosystem. (Winteringham, 1984) Figure 2.2 Nitrous oxide measurements at Cape Grim, Tasmania (up to 1995), 10 and at Baring Head from 1995 onward. (NIWA report) Figure 2.3 Estimates of the relative net cumulative warming effect over 100 11 years of New Zealand's anthropogenic emissions of each of the major greenhouse gases as measured in 1995. (Ministry for the Environment, 1998) Figure 3.1 Kairanga site 34 years (K34) treatment field. Figure 3.2 Permanent pasture (PP) field treatment at Kairanga site. Figure 3.3 Massey site experimental treatment plots. 33 33 35 Figure 3.4 Schematic layout of experimental treatment design plots at Massey 36 site. Figure 3.5 Closed cover chamber used to collect field nitrous oxide (N20) 41 emissions. Figure 3.6 Installed chamber fitted with 3-way tap for withdrawing air samples 45 . . usmg syrmge. Figure 3.7 Injecting 10 ml of gas sample into gas chromatography (GC) 45 instrument to measure nitrous oxide (N20) concentration. Figure 4.1 Increase in N20 concentration in a closed chamber over time in the 51 treatments: (a) conventional tillage (CT); (b) no-tillage (NT); (c) permanent pasture (PP) . (bars represent standard error) Figure 4.2 The effect of tillage treatment of the PP, CT and NT on the monthly 60 data of soil moisture (a), temperature (b) and log-transformed means of N20 emissions (c). Figure 4.3 Regression analysis between measured soil moisture content and 63 N20 emissions in the PP (permanent pasture), CT (conventional XI tillage) and NT (no-tillage) treatments. Figure 4.4 Regression analysis of log-transformed data between soil moisture 64 content and N2O emissions in the PP (permanent pasture), CT (conventional tillage) and NT (no-tillage) treatments. Figure 4.5 Monthly rainfall data for the Massey site from December 1998 to 66 September 1999 and 1996-97, 1997-98. Figure 4.6 The effect of tillage treatment of the PP (permanent pasture), K 17 7 5 (17 years continuous maize) and K34 (34 years of continuous maize) on the monthly data of soil moisture (a), temperature (b) and log­ transformed means of N2O emissions (c). (bars represent LSD) Figure 4.7 Regression analysis between measured soil moisture content and 77 N2O emissions in the PP (permanent pasture), K 17 ( 17 years of continuous maize) and K34 (34 years of continuous maize) treatments. Figure 4.8 Regression analysi s of log-transformed data between soil moisture 78 content and N2O emissions in the PP (permanent pasture), K 17 ( 17 years of continuous maize) and K34 (34 years of continuous maize) treatments. Figure 4.9 Monthly rainfall data for the Kairanga site from November 1998 to 81 September 1999, 1996-97 and 1997-98. (note: rainfall data for 97-98 was used from another station located in close proximity) Chapter 1 General Introduction 1 There is a growing concern world-wide about climate change. Atmospheric warming which is known to be caused by so-called "greenhouse gases" mainly include carbon dioxide (CO2), methane (CH4 ) and nitrous oxide (N2O) and to a lesser extent chlorofluorocarbons (CFCs) (IAEA, 1992). Presently, the increase in greenhouse gases other than CO2 in changing the climate is similar in importance as CO2. One such gas is N2O which despite its low concentration in the atmosphere, about 310 ppb (IPCC, 1995), on a molecule per molecule basis has a radiative force about 200 to 300 times that of CO2 (Jaques, 1992) and an average atmospheric lifetime of about 150 years (lAEA, 1992). It is widely accepted that the main source ofN2O is agriculture. Most N2O originates with soil processes, as intermediate product from microbial nitrification and denitrification (Delwiche, 1981 ). Increased emissions of N2O from soils are associated with fertilisation of soils with mineral nitrogen (N), animal manure, N derived from biological N2 fixation, and enhanced N mineralisation (MacKenzie et al., 1998). With fertiliser usage predicted to grow worldwide at 6-7% per annum (Peoples et al., 1995) and low N utilisation efficiency in agricultural systems, the potential of soils to form and emit N2O increases. This increased N2O emission to the atmosphere is of great concern and need quantification. Amounts of N2O emitted depend on complex interactions between soil properties, climatic factors and agricultural practices (Granli and Bockman, 1994). Main factors in the soil controlling N2O emissions are soil content of NH4 and NO3 (Ball et al. , 1997; Castaldi and Smith, 1998; Seneviratne and Van Holm, 1998); soil aeration status and soil water content (Carran et al., 1995; Teira-Esmatges et al. , 1998; MacKenzie et al., 1998); presence of degradable organic material which promotes microbial activity (Ineson et al., 1998; Kaiser et al., 1998); soil pH (Anderson and Poth, 1998; Sitaula and Bakken, 1993 ; Burth and Ottow, 1983) and soil temperature (Mahmood et al. , 1998). Although these are known interacting factors they are not always strongly correlated with N2O fluxes . Due to complexity of interactions between various factors , N2O emissions have very high spatial and temporal variations. Chapter 1 - General Introduction 2 Agricultural practices alter soil properties which influence the extent of N2O emissions. Intensive use of cultivation practices, both internationally and locally in New Zealand, hugely impact soil properties. In the Manawatu region heavier textured soils used for continuous maize production result in loss of soil organic matter (SOM) (Saggar et al., 2000) and deterioration in soil structure (Shepherd et al., 2000). Even short term tillage operations can affect SOM levels and microbial biomass (Aslam et al., 1999) which are of particular interest in nutrient transformations. Conversion of pastures to arable cropping in New Zealand results in depleting of SOM and soil fertility over time and additional N fertilisers are applied to compensate for the loss of organic N reserves. Since N fertilisation is considered as a major practice on increasing N2O emission from soil by providing an additional N source (Ryden and Rolston, 1983), careless N application on such soils may contribute to increased loss of N either as NO3 through leaching or as N2O emissions. An alternative to conventional tillage system is conservation tillage system which aims for sustainable agricultural production. Its growing acceptance is due to reduced soil erosion and runoff (Choudhary et al., 1993; Unger and Vigil, 1998; Myers and Wagger, I 996). enhanced moisture retention and infiltration (Baumhardt and Lascano, 1996), lower summer temperatures (Prihar et al., 1996) and possible increased net return to the farmer (Reicosky, 1994). However, the impact conservation tillage has on N2O emission is not known for these Manawatu soils. In the past 20 years, research of N2O emissions has concentrated on enhancing an understanding of N2O production processes and its controlling factors. Despite this it is not possible to predict the fate of a unit of N that is applied on a specific arable field (Mosier et al., 1996). Both short- and long-term in-situ measurements are needed to assess N2O emissions from soils. Studies by Ruz-Jerez et al. (1994) and Carran et al. (1995) from both poorly and well drained grazed pastures in the Manawatu region suggest low N2O emission from these low fertility hill lands. However, there is little information for arable farming systems. Therefore, there is a need for site-specific assessment of the impact of tillage practices on N2O fluxes. Chapter 1 - General Introduction 3 1.1 Research Objectives To evaluate long-term and short term impacts of different tillage systems and land use on N 2O emissions selected farming practices were chosen in the Manawatu region of New Zealand. The overall aim was to characterise land use practices and their effect on N2O emissions. This study is a part of a wider project on soil nitrogen recycling. The specific objectives of this study were as follows: • To quantitatively determine the rates of N2O emissions from fields sown with the conventional tillage (CT), no-tillage (NT) and compare these with permanent pasture (PP) fields throughout one management cycle. • To measure the response of soil N2O emissions to various cultural practices and selected environmental parameters such as changes in soil moisture and soil temperature . • To measure soil physical and chemical properties and asses their interactions with N2O emissions. 4 Chapter 2 Literature Review 2.1 Introduction This literature review addresses the global impact of changes in land use and soils as a presently known major source of anthropogenic N2O and discusses issues and possible ways of reducing N2O emission into atmosphere. Human interference in the land cover has brought changes in the environment. Modern agricultural practices, such as monocultures with limited return of crop residues to the soil and the use of chemical fertilisers, have been reported to cause severe soil degradation. Agricultural conversion disrupts the steady state conditions that exist in many natural conditions (Bouwman, 1990). The increase in concentration of a number of atmospheric gases has been having alarming global impact due to human activities. These gases include carbon dioxide (CO2). methane (CH4) , carbon monoxide (CO), nitrous oxide (N2O), nitrogen oxides ( o:-J and ammonia (NH3) . Possible causes of the increase in the concentration of these gases are the increasing emissions by the various sources and in some cases a reduced sink capacity (Bouwman, 1990). Bouwman (1990) has summarised data on the annual rise in the concentration of important gases, heat absorbing capacity and the contribution to the global temperature rise which occurred during the past 100 years (Table 2.1 ). N2O oxide is one of the "greenhouse gases" which forms part of the nitrogen cycling (Figure 2.1) which is involved in a number of atmospheric processes. Moreover, its reaction products are known to cause substantial changes in stratospheric processes. Soils are known major source of the enhanced emission of N2O while other sources have minor inputs to global emission. Release of N2O from soils is known to occur during biological denitrification, chemical denitrification and nitrification. Chapter 2 - Literature Review 5 Reported values of measured N2O emissions often reflect high variability. There are some already known factors which control N2O fluxes but they are not always primary ones. The potential of a soil to form and emit N2O increases with increasing availability ofN, but the actual amounts emitted depend on interaction between controlling factors. ~ "' .,, QJ u E ~ QJ :, E Cl "' QJ ~ ~ :, 0 C "' > aj .2 ~ tO N ia ., "' ~- u .qi -s Cl .! 0 0 £ C 0 ·-.,, "' C "' i5 "' C c· "O "O 0 ~ "O QJ QJ "' "O "O C X "tl "O "O - "' "' "O 4: z z 2 2 Culfr.tated or natural 1egumes, etc. Atmosphere / > .D .,, +! I E I"' I\ a: / \ \ \ \ u ~ -.,, QJ > .. L: ' \ \ \ ' \ • "' I "' ::: 1 "' ,_ u I M 0 a. I I CI 2 01-0 · - C ~ I (1) '::! I 0 ';:IN ~IZ o I · >IN >,z .D > :G I C ~I ·; 01~ ' ..J - ' ' Wildl ife, farm ' livestock, etc . ~ l Roo, '. 0 ;:, 0 , d;a,h aod docomposiHoo ...... J ~ of .,,;,1aod ,oo< "'''"" ~--,· ·.·-~ _;('. ·.· . . · ·" Soluble material Microbiological N pool of the soil Stored or 'immobilized' organic N-pool of the soil I • ••• I 1 Losses by ~le~ching Impermeable bedrock --..____;I Transfers within the system ~------ Additions to the system Losses or removals from the system / Figure 2.1 Simplified diagram of nitrogen transfers, removals and losses of the agro-ecosystem. (Winteringham, 1984) Chapter 2 - Literature Review 6 Table 2.1: Atmospheric concentrations of the major greenhouse gases, their rise, residence time and contribution to the global warming Type Residence Annual 1985 Radiative Contribution to time (years) Rise(%) concentration absorption greenhouse warming potential (%) CO2 100 0.5 345 ppm I 50 co 0.2 0.6-1 .0 90 ppb n.a. n.a. CH4 8-12 1 1.65 ppm 32 19 N20 100-200 0.25 300 ppb 150 4 03 0. 1-0.3 2.0 n.a. 2,000 8 CFCs 65-110 3.0 0.18-0.28 ppb >10,000 15 * Bouwman (1990) 2.2 Nitrous Oxide in the Atmosphere N20 is capable of absorbing infrared radiation, but it is inert in the troposphere. According to an IPCC ( I 995) report the mean estimated atmospheric concentration of N20 in I 990 was 3 IO ppb, corresponding to a reservoir of about 1500 TgN. Reports have put its increasing rate at approximately 0.2-0.3% per year and the observed rate of increase represents an atmospheric growth rate of 3 to 4.5 TgN per year. 2.3 Sources of Nitrous Oxide N20 oxide is formed from various sources but in general they can be divided into two categories: natural and anthropogenic. Natural sources include soils, oceans, fresh waters, fires, lightning, and vegetation. The fluxes from soils and marine waters are the dominant global sources of N20 (Jaques, 1992). Part of the N20 emission from water ecosystems originate from ground water which is supersaturated with N20 originating from soil microbial processes (Granli and Bockman, 1994 ). Soil microbial processes are generally regarded as the major global source of N20 . The increased emissions of N20 from natural sources are also partly caused by human activities which probably enhanced N20 formation in natural ecosystems. N20 oxide is produced directly from the burning of fossil fuels, although the mechanisms that cause N20 formation are not completely understood. Biomass burning Chapter 2 - Literature Review 7 including forest cleaning, savanna and sugar cane field fires , burning of agricultural waste and firewood is another possible source of N20 . Some N20 is formed during the processes involved in manufacturing of nitrogen fertilisers and production of nylon , plasticisers and industrial chemicals (Jaques, 1992). There are other minor sources which release N20 . They include anaesthetic usage, propellant usage, high voltage transmission lines and other minor miscellaneous sources. In general, all environments where nitrogen compounds are transformed by biological processes are potential sources of N20. Major agricultural practices influence N20 emissions (Mosier et al. , 1997; Kessavalou et al. , 1998b; Aulakh et al., 1984) and estimated that they may contribute approximately 70-90% of anthropogenic N20 emission. Fertiliser applied soils can emit N20 under either anaerobic or aerobic conditions. Studies show that soils have increased potential of emitting N20 depending on fertiliser types applied (Bremner et al., 1981 ). These authors have reported that the fertiliser-induced emissions of N20 observed after application of anhydrous ammonia greatly exceeded those detected in similar field studies using other N fertilisers and represented 4 - 6.8% of the fertiliser N applied. Chang et al. ( 1998) reported that under high N20 producing conditions, generally characterised by a high soil moisture content and concentrations of N20 in soil solution and groundwater, low gas diffusion rate, emissions from plants may even exceed those from the soil surface during periods of high transpiration. 2.4 Impact of Nitrous Oxide on the Environment Two main concerns with increasing emissions are the processes which enhance N 20 emissions. 1. N20 oxide contributes significantly to the atmospheric "greenhouse" effect by trapping outgoing terrestrial radiation (Crutzen, 1981 ). N20 is essentially transparent to incoming short-wave solar radiation but it absorbs and emits longwave radiation which influences the Earth's climate (Bolin et al. , 1986). Chapter 2 - Literature Review 8 2. Nitrogen oxides produced from N20 by photochemical reactions in the stratosphere are involved in destruction of stratospheric ozone. Stratospheric ozone layer shields the biosphere from harmful UV radiation and influences the vertical temperature profile and thus earth surface temperatures. Direct transport of NOx into the stratosphere from the earth's surface is unlikely because of the short atmospheric residence time of NOx, which is quickly converted to HN03 aerosols and thermally unstable organic nitrates and removed by wet and dry deposition (Crutzen, 1981). One of the major sinks of 0 3 is a reaction with NOx, which catalyses the destruction of 0 3 above 25 km in the stratosphere (Crutzen, 1981 ). However, Delwiche () 981) alerts that NOx also react with degradation products from volatile chlorinated carbon compounds and thus diminish their ability to cause ozone destruction . The influence of N20 on the stratospheric ozone shield is thus complex. However, slowly increasing atmospheric N20 concentration may have an over-all detrimental effect on the total ozone content of the stratosphere (Granli and Bockman , I 994 ). 2.5 Global Nitrogen Perspective Nitrogen (N) in the environment is reviewed from distribution and cycling of N which have been discussed in detail by Haynes ( 1986). Gaseous N in the atmosphere represents only 1.9% of earth's total N mass and the rest 98 % is held in rocks and minerals. Haynes (1986) summarised the biogeochemical distribution of N on Earth as shown in Table 2.2. The magnitude of the global flows of N due to some of the processes illustrated in Figure 2.1 is shown in Table 2.3. The major additions of N to the soil occur through the processes of wet and dry deposition and by the action of micro-organisms that fix atmospheric N2. Man is increasingly active in fixing N 2, by industrial processes and by use of internal combustion engines. In many areas, most nitrates in precipitation originate from sources related to man. The quantity of N fixed industrially and applied to agricultural lands is of the same order as that fixed by micro-organisms (Table 2.3). Chapter 2 - Literature Review 9 Losses of N occur through leaching of NO3, erosion and surface runoff, volatilisation of ammonia, gaseous losses of N2 and N2O, and, in agricultural ecosystems, plant or Table 2.2: Biogeochemical distribution of N on Earth Pool of nitrogen Atmosphere N2 N2 dissolved in oceans Biosphere Lithosphere Igneous rocks Sedimentary rocks Total N mass * compiled by Haynes, 1986 Total mass (TgN) 3.9 X 10 2.2 X 107 2.4 X 107 1.9 X 10 11 4.0 X 108 1.94 X 10 11 Percentage of total N mass (%) 1.9 0.01 0.01 97.8 0.2 Table 2.3: Global fluxes of nitrogen into and out of the terrestrial biosphere Inputs Wet and dry precipitation (NH]INH4) Wet and dry deposition (NOx) Wet and dry deposition (organic N) Atmospheric fixation (lightning) Biological fixation Industrial fixation Outputs Ammonia volatilisation Denitrification (N2 + N2O) Biogenic NOx production Fossil fuel burning (NOx) Fires (NOx) Leaching and runoff (inorganic) Leaching and runoff (organic) * compiled by Haynes, 1986 Process rate (Tg N yea( 1 ) 90-200 30-80 10-100 0.5-30 100-200 60 36-250 40-350 1-15 10-20 10-20 5-20 5-20 Chapter 2 - Literature Review IO animal removal. Although Table 2.3 shows the greatest N2O and N2 loss through denitrification, N2O emission can occur as a result of two separate processes: denitrification and nitrification. When a natural ecosystem is in a steady state (e.g., a mature forest) the rate of N input by precipitation and biological N2 fixation balances outputs by denitrification, volatilisation, and groundwater and storm loss. However, when natural ecosystems are disturbed the N cycle is interrupted because of removal of vegetation. Agricultural ecosystems often have a greater N requirement than natural ecosystems. While much of the N used by crop plants is harvested and removed from the ecosystem fertiliser additions are widely believed to have enhanced soil N2O emissions. 2.6 Status of Nitrous Oxide Emissions in New Zealand National Institute of Water and Atmospheric Research (NIWA) has made trace gas measurements at Baring Head (on the South Coast of the North Island near Wellington) 314 312 310 308 306 .c = = 304 302 • Cape Grim 300 • r:.1 Baring Head 298 ~ • 296 1975 1980 1985 1990 1995 2000 YEAR Figure 2.2 Nitrous oxide measurements at Cape Grim, Tasmania (up to 1995), and at Baring Head from 1995 onward. (NIWA report) Chapter 2 - Literature Review l l (Figure 2.2) since 1973. The measurements contribute to the global network for determining trends in greenhouse gas concentrations. Figure 2.2 shows a steady increase in N 20 levels at Cape Grim, Tasmania and Baring Head from 1970s to the present. Although N20 emission measurements are still rather uncertain due to high spatial and temporal variations they contribute almost 20% of New Zealand's total gree nhuu~c ga~ loadi ng (i n relation to carbon dioxide) and all of this is from land-based industries (Figure 2.3) (Ministry for the Env ironment, 1998). New Zealand's greenhouse anthropogenic gas emissions Methane 45% <1% SF6 0<1% Nitrous oxide 19% Carbon dioxide 36% Figure 2.3 Estimates of the relative net cumulative warming effect o, er 100 years of ew Zealand's anthropogenic emissions of each of the major greenhouse gases as measured in 1995. (Ministry for the Environment, 1998) In ew Zealand context, Carran (unpublished) states that nitrogen fertili ser use c learl y increases N20 emission, but it largely depends on how long soil s remain wet. In well­ drained soils and low rainfall areas, 20 production is low, but can increase up to 5-7 kg N/ha in poorly drained soils receiving moderate to high rainfall. The highest values recorded locally have been on wet, poorly drained soils during grazing by mobs of dairy cows that caused serious pugging (Carran et al., 1995). Although the steps to reduce emissions are not clear Carran (unpublished) suggests that drainage may have a positive effect in reducing N20 emission, whereas irrigation and nitrogen fertiliser can have a negative effect. Chapter 2 - Literature Review 12 2.7 The Role of Micro-organisms Bioti c so urces of N20 have been explained by Umarov (1990). Micro-organisms are the main bi o ti c sources of 20 in terrestrial and aquatic ecosystems. N umerous bacteria can den itrify in anaerobic and aerobic conditions while only highly specialised bacteria can nitrify in specific eco logical conditions (e .g. pH nearl y neutral , sufficient aeration. lack of organic matter, low concentration of ammonia). The fo rmation of N20 in natural conditions is caused by the ability of all organi sms for ox idation and reduction of nitrogen in different mineral and organic compounds. Micro­ organi sm s are capabl e of oxidation and reduction of all nitrogen compounds producing N20 amo ng other products. Formation and consumption of N20 are carried out by large groups of heterotrophi c and autotrophic micro-organi sms. Most of them are bacteri a. but there are some eucaryoti c organisms, especially fungi , which are able to produce N20 (U maro v. 1990). 2.8 Mechanisms of Nitrous Oxide Production Re lease of ox ides of nitrogen (N 20 , 0 and N0 2) from so il s is known to occur during bi o logical denitrifi cati on. chemical denitrification and nitrification . Denitrification and nitrifi cati on are two microbial processes which contribute most to the emissions of , 0 fro m so il s. Ini tia ll y. 1 20 was thought to o ri ginate largely through denitrificati on process. However, o ther research studi es have indicated that some of N20 may be produced thro ugh nitrification. Works reported by Bremner and Blackmer ( 1981 ) and Mosier et al. ( 1983) show that nitrification also can be a significant source of N20 . The re lati ve importance o f these processes va ri es with local circumstances. 2 .8.1 Denitrification Denitrification is a fo rm of anaerobic bacterial respiration dming which nitrogen (N) ox ides , principally nitrate and nitrite, are reduced sequentially through NO and N 20 to N 2 (Aulakh et al. , 1992). The commonly accepted reductive pathway for denitrification is as follows: Chapter 2 - Literature Review NO3 Nitrate • NO2 Nitrite • NO Nitric oxide • N 20 Nitrous oxide • I 3 N2 Di nitrogen Denitrification occurs when soil is starved of oxygen (e. g .. under conditions of temporary flooding by heavy rain, irrigation controlled flooding of rice paddy, etc.) (IAEA. 1984 ). During this anaerobic process, nitrate serves as an electron acceptor for the oxidation of organic (and sometimes inorganic) compounds, with the yield of energy and the release of gaseous N2. The majority of soil bacteria seem to be able to denitrify but they exhibit variety of incomplete reduction pathways (Umarov, 1990). According to Umarov ( 1990) ecological factors , such as nitrate concentration, level of aeration, temperature and pH have secondary significance. They affect on ly the denitrify ing potential. but the actual activity of the process is only affected in the presence of sufficient organic matter. Depending on so il conditions, intermediate products (NO and N 2O) can accumulate and eventually escape from the soil before being reduced. The ratio of N 2 to N 2O in the gases evolved from soil (approximately 16: 1) depends on such factors as soil pH, moisture content, redox potential (E1i), temperature. nitrate concentration and content of available C (lAEA. 1984). The N 2O emissions from denitrification are normally short-termed, episodic events occurnng during the initial development of soil anaerobiosis (Byrnes et al., 1990). Although denitrification may be discounted as a source of N 2O under the well-aerated soil conditions of the tropical dry season, it may be important during the rainy season (Anderson and Poth, 1998). 2.8.2 Nitrification itrification is classically defined as the process whereby NH 4 is oxidised via NO2 to NO3. The reactions are generally mediated in soi l by activities of two small groups of chemoautotrophic bacteria. One group, the NH4 oxidisers, init iates the process with the format ion of NO2, while a second group, the NO2 oxidisers, completes the process by Chapter 2 - Literature Review 14 converting NO2 to NO3 as promptly as it is formed (Haynes, 1986). Autotrophic nitrifiers use CO2 as a carbon source and obtain their energy by oxidation of NH4 (Umarov, 1990). There are two possi ble ways m which N2O gas is evolved during the process . The presumed intermediate NOH, or its dimer hyponitrite, may dismutate chemically under reduced 0 2 tensions to N2O or the dissimilatory enzyme system, nitrite reductase, may yield N 2O when 0 2 becomes limiting and NO2 replaces 0 2 as an electron acceptor (Schmidt, 1982). Alternatively NO and N2O can also be produced abiotically by reduction of NO2, especially when soil pH is low. In aerobic soils both NO and N 2O are most likely produced directly or indirectly by nitrification. N2O produced during nitrification has a greater probability of being lost from the soil than does N2O produced during denitrification. Because soil pores are generally open during nitrification and diffusion is allo~ed to the soil surface, there is less water present to dissolve N2O. In contrast to denitrification, emissions during nitrification in soils are relatively constant and may actually decrease temporarily following rainfall (Byrnes et al. , 1990). Davidson et al. ( 1993) reported that nitrification was the dominant source of N2O when the soi l was wetted at the end of the dry season, and that denitrification might be an important source of 20 during the wet season. These results support a conceptual model in which trace gas production is generally constrained by the rates of N mineralization and nitrification, while the specific ratios of NO and N 2O fluxes and the contributions from nitrifying and denitrifying bacteria are controlled largely by soil moisture. 2.8.3 Sin ks Reaction of N2O in the stratosphere is perhaps regarded as the only significant process for removal of N2O from the atmosphere. Crutzen ( 1981) explains in details that Chapter 2 - Literature Review 15 electronically excited O(' D) atom is produced by photolysis of ozone in the stratosphere. Approximately 10% of stratospheric N 20 is thought to be converted to nitric ox ide by the reaction NO then participates in further reactions with ozone and other reactive molecules. Two additi onal reactions in the stratosphere contribute to the removal of N?O from the atmosphere: So il s can act as a sin k too. In soil denitrification process, whether nitrate is converted to N 2 gas or N20, it is no longer readily available to plants, and only a limited number of mi cro-organi sms ( or micro-organisms in association with plants) are capable of "fixing" atmospheric nitrogen and bringing it back into the biological cycle (Delwiche, 1981 ). A number of studies observed negati ve emiss ions of N2O (Teira-Esmatges et al. , 1998). The o bse rved emiss ions of N20 were the result of two counteracting phenomena: the emi ssion of N20 fro m the soil and consumption of headspace N20 by the soil (Van Cleemput et al. , 1994 ). According to these authors the emission from the soil depends on: (i) the fo rmati on of N2O during denitrification and during nitrification, and its diffusion to the headspace. (ii ) the consumption of N2O through its reduction to N2 during the denitrification and the diffusion rate of 20 from the headspace into the soil. When N2O consumption was larger than its emission, the N2O concentration in the headspace decreased resulting in negative fluxes ofN2O. Similar findings were recorded by Mahmood et al. (1998) who observed N2O sink activity during the maize and wheat season on one third of sampling occasions which ranged between 0.01 and 0.94 g N ha- 1 h-1 . The generally low N 2O mole fraction of the gaseous N products of denitrification and frequently observed negative N2O fluxes Chapter 2 - Literature Review 16 indicated that the soil conditions under irrjgated maize and wheat were favourable for N20 reduction. Few studies have suggested that plants also acted as a sink. For example, Grundmann et al. ( 1993) demonstrated that 15N20 accumulation in the jars was lower in the presence than in the absence of plants. The authors also stated that 15N 20 was taken up by maize leaves and metaboli sed as a source of N . Acceptance of this process as a possible N20 s ink and N source in natural environment requires independent confirmation and quantificati on. 2.9 Agriculture as a Major Source of Nitrous Oxide Emissions Man has changed steady-state conditions of the environment in the last century. This in turn has negativel y affected the environmental processes. The extent of cultivation practices has increased nitrogen cycle through soil disturbance and the amounts of fertiliser put into soi ls. In terms of N 20 emission soil disturbance decreases flux es due to better aeration , however the amounts of fertiliser applied into cultivated so il s increases potential emissions of N 20 . 2.9.1 Effects of Tillage Practices Research has shown that conversion of ' native ' grass sod significantly increased net soil N 20 emi ss ions compared to wheat-fallow cropping (Kessavalou et al. , 1998b). More detailed study of the effects of conversion was conducted by Mosier et al. (1997) who also fo und that conversion of grassland to croplands typically increased the emission of N:P- Although N 20 fluxes were about 8 times higher for 18 months foll owing ploughing. the relative rates declined to 25-50% higher than the native site after 3 years. N20 emissions 2 years after ploughing were similar to the grass lands converted to a winter wheat-fallow production system about 70 years ago (Mosier et al., 1997). The authors found that reversion of cultivated soils that had been cultivated in the early decades of 20th century back to grasslands eventually lead to soils having similar N20 emissions to those of native soils of the same texture and parent material. Sites that were last ploughed in 1939 exhibited the same N20 flux rates as comparable never-tilled soils during 15-month measurement period. They concluded that it requires from 8 to 50 Chapter 2 - Literature Review 17 years following tillage to return to N 20 emission rates observed in the native state in shortgrass steppe. ln contrast, conservation practices aimed at reducing harmful effects of extensive conventional practices contributing to N20 emissions through improved soil conditions. such as increased so il organic matter, higher soil microbial biomass, and moisture content. all of which are favourable conditions for enhanced formation and emissions of 20. A num ber of reports indicate that no-till has higher em1ss1o ns rates compared to conventionally managed lands . The hypothesis that the extra C going below-ground may stimul ate N 20 production by soil s (as a consequence of enhanced root-derived avail ab le soi l C, acti ng as an energy source for denitrification) gave positive results (Ineson et al. . 1998). Colbourn ( 1988) observed denitrification in the undrained ploughed so il as one-fifth of that in undrained direc t-drilled soil. Cultivation restricted denitrification to a greater extent than drainage. ln drained soi l it was restricted to one-twelfth by ploughing. The authors attribute differences in denitrification to the denitrification in soil structure caused by the soil systems. In particular the direct-drilled soils both (drained and undrained) behaved similarly despite differences in drainage, but very differently to the ploughed so il s. T he lack of so il disturbance with reduced tillage leads to a reduction in large pores, increaseu so il aggregat ion and reduced so il aeration. Populations of bacteria responsible for 20 production may also be greater under no-till management. MacKenzie et al. ( 1998) associated higher N 20 emissions in the NT systems, with corn compared to other crops, to increased crop residue C and inorganic N from higher input from fertilisation. However, there are several alternatives that can be implemented to reduce N 20 emissions from the soil surface. 2.9.2 The Role of Fertilisers Increased nitrogen fertiliser use is considered to having substantial impact on N 20 Chapter 2 - Literature Review 18 despite small percentage of it being lost as a N20. The main reason is that fertiliser application potentially enhances N20 fluxes from the soils. Swensen and Bakken (1998) demonstrated that urea enhanced growth of nitrifying bacteria in mineral subsoil. The authors hypothesised that bacteria recovering from starvation or restricted activity after adding urea had higher production of N20. Similar findings by Seneviratne and Van Holm ( 1998) were made where urea application resulted in a 4-fold increase in N20 emission from the so il. Studying the effects of three N fertilisers (urea, ammomum sulphate, and calcium nitrate) on emissions of N20 from soil Breitenbeck et al. (1980) showed that the emissions of 20 induced by application of urea or ammonium sulphate markedly exceeded those induced by application of calcium nitrate. The emissions, however, represented <0.2% of the fertiliser N. Similar study with anhydrous ammonia by Bremner et al. ( 1981) found that this fertiliser-induced emissions of N20 greatly exceeded those detected using other N fertilisers and represented 4.0-6.8% of the fertiliser N . Gaseous losses of fertiliser applied N are generally less than 5% (Kaiser et al., 1998. MacKenzie et. al., 1998) but other studies have reported losses of up to 13 .6% (Teira­ Esmatges et. al., 1998). This could be due to the kinds of soil micro-organisms involved in the transforma ti ons of soi l N and subsequent N20 emissions. These were demonstrated by Castaldi and Smith (1998) where different forms of had very different impacts on N20 emissions ih the two light textured soils, suggesting the involvement of different microbial communities in the N20 production in arable and woodland so ils. The arable soil had a very low potential for N20 emissions derived from nitrifiable N. as compared with the N20 which was produced when the soil was provided with nitrite and nitrate and a carbon source. That is the likely reason of different fertiliser gaseous losses. 2.10 Factors Controlling Nitrous Oxide Emission 2.1 0.1 Soil Properties N20 emissions are reported to be strongly influenced not only by land use and Chapter 2 - Literature Review 19 management practices but also by a wide range of soil factors. For example. experiments conducted by Byrnes et al. (1990) indicated that emissions of N20 during nitrification may be more dependent on the soil to which fertiliser is applied than on the fertiliser type. N20 flux es generally depend on physical, chemical , and biological attributes of soil and on complex interactions between them. Fine-textured soils seem to emit more N 20 than coarse-textured ones , but this tendency can be masked or reversed by other factors, especially the soil water content. The clayey soils can have a higher potential for sustained N20 formation than sandy soils, since clayey soils can maintain a higher water filled pore space (WFPS) for longer periods than can light, easily drained sandy soils. Howeve r, 20 can escape more easily from coarse-textured soils. The resulting N 20-N emission at the soil surface depends on the intensity of both the nitrification and denitrification processes (Teira-Esmatges et al. , 1998). Inherent soil properties very strongly influence the magnitude of annual N 20 emissions, masking the influences of management or climatic variability (Lemke et al., 1998b ). These authors have suggested that large-scale spatial variability in the annual estimates of N20-N loss could be explained by differences in clay content, suggesting that this parameter could be an important criteria for stratification. Christensen and Tiedje ( 1988) observed stibstantial nitrogen losses for a few days only on sandy so il s, because they are flooded only for very brief periods. They also suggested that even in more clayey soils where flooding occurs for longer periods, excessive N losses might be of short duration due to consumption of the easily available carbon source. The influence of soil texture on N20 emission was generally studied in cases where denitrification prevails (Granli and Bockman, 1994). Denitrification of nitrate and nitrite is generally considered as the dominant source of N20 in heavy textured soils, where oxygen diffuses more slowly than in light textured soils . Introducing the concept of ion exchange or fixation in kinetic models on N20 formation can increase knowledge on these processes in soils (De Visscher et al., 1998). The Chapter 2 - Literature Review 20 authors suggest that N 20 formation can only be compared in soils with the same cation exchange capacity (CEC). 2.10.2 Soil Organic Matter Denitrifying organisms use organic C compounds as e lectron donors fo r energy and for synthes is of ce llular constituents, therefore, denitrification is strongly dependent on the ava il ability of organic compounds such as native soil organic matter. crop residues. root ex udates, and green and farmyard manure (Aulakh et al. , 1992). Organic matter in soil s is represented by plant debris or litter in various stages of decompos ition through to humus and also inc ludes microbial biomass (Bouwman, 1990) . Concentrations of biomass C and organic C in the so il are highly correlated with denitrification acti vity ( Drury et al. , 199 1 ), and there is evidence that increas ing carbon availability decreases the N20 frac tion (Firestone, 1982). Since organic matter in so il s is represented by plant debri s crop species have a significant infl uence on the N20 emi ssions (Kaiser et al. , 1998). The authors reported that the to tal 20 losses during the winter increased with decreasing dry matter-to-N­ content rat io of the plant residues incorporated into the soil by ploughing. Soil freez ing and thawing greatl y enhance the rate of 6.5 at micros ites so DNRA as process fo r 20 production may be much more important than presently realised. Jn the study of four deni trifying bacteria and soil fungus Burth and 0ttow ( 1983) concluded that the quantity rather than the quality of gas production shows a marked dependence on the pH of the environment. The composition of the denitrifying gases is specific to the organi sms in question (under given set of conditions) . In soil or water. denitrification by fungi such as Fusarium spp. should not be underestimated, the more s ince the strai n iso lated proved to be highl y effective in producing 20 from nitrite at reduced oxygen pressure. 2.10.5 Temperature N20 emissions are highl y variable and generally the amplitude of variation is greater at higher temperatures and with larger fertiliser doses (Bouwman, 1990). The effect of increasing soi l temperature on denitrification rate has been investigated in many laboratory studies . Temperature has been found to have positive correlation with N20 emission rates . The optimum temperature for the denitrification process is 25°C and Chapter 2 - Literature Review 23 above, while the process is slow at 2°C. Denitrification is still rapid at elevated temperatures and will proceed to about 60 to 65°C but not at 70°C. In nitrification the optimum temperature lies between 30 and 35°C, while below 5°C and above 40°C the acti vity is very low (Alexander, 1977). As temperature increases, the N20/ N 2 ratio declines. Important losses may also occur at lower temperatures. Keeney et al. ( 1979) have reported, that while the rate of denitrificat ion was low at temperatures below l 5°C, the amount of 20 ( 44-50% of the total gas production) was equivalent to that evolved at 25°C. Thus denitrification during the late autumn and early spring in temperate climatic zones could account for a signifi cant portion of the N20 released over the year, particularly as soils often tend to be saturated or at least have high moisture contents and consequently lower oxygen leve ls. over these periods as well. 2.11 Soil Management Practices Impacting Nitrous Oxide Emission The fertiliser-induced emissions have been estimated as contributing from 5% to 25% of the total N20 efflux on a global scale (Bolle et al. , 1986). These emissions might be altered by technology or by policy decisions that could promote use of one fertiliser over another or use of a particular fertiliser management practice to reduce N20 product ion (Byrnes et al., 1990) . Certain management practices can be encouraged to reduce or eliminate N20 emissions. For example. in a study by Seneviratne and Van Holm (1998) emissions of N 20 was almost completely el iminated by the application of wet surface mulches, presumably ei ther because of a decrease in the N20 production in the soil with the mulch or due to an increase of N20 consumption in the mulch or soil. The authors postulated that when a mulch is applied to a moist soil, micro-anaerobic conditions in the mulch can be formed due to moisture present and also due to the presence of a high amount of easily degradable organic matter resulting in the mulch forming a large oxygen sink. Those conditions may lead to the production of N20. In addition, the soil surface mulch may act as a barrier for N20 emitted during soil denitrification and CH4 produced in the moist soil. However, in this study, all plant materials used as mulches reduced the contribution to the enhanced greenhouse effect from the soil. Chapter 2 - Literature Review 24 No-till systems are reported to have higher N 20 emission rates (MacKenzie et al., 1998) but benefits of this management practice cannot be compromised for detrimental effects of conventional tillage practices. Modelling assessment of US agriculture by Mummey et al. ( 1998) suggested that no-till management in areas that are relativel y warm and wet may result in N 20 emissions similar to or less than conventional tillage and that no-till management may be a viable means to reduce N20 emissions while increasing so il qual ity. However, the authors argue that even though initial conversion to no-till had greater N20 flux it maybe short lived and simply the result of the non-equilibrium of the changing system. Over time as the crop surface residues increased the active fraction of the SOM. pool the cycling may become ti ghter and less available to be converted to gas. MacKenzie et al. (1998) suggested that corn system using conventional tillage, legumes in rotation. and reduced N fertiliser would decrease N 2O emission from agricultural fields . Low N2O fluxes fro m grassland compared to tillage system (Mosier et al.. 1997) could al so contribute to 20 mitigation with additional benefits of restoration of soil properties. These would be especially suitable for unsustainable and deteriorated crop lands. New research into different fo rms of fertiliser N with coating or release rate is another direction of mitigation alternatives to control N 2O emissions. 2.12 Field Nitrous Oxide Emission Measurement Methods Methods for fie ld measurements of N2O emissions from agriculture have been reviewed by Mosier ( 1990) and are described in a manual from the IAEA ( 1992). The study of N2O emissions from soils is complicated by experimental difficulties. The experimental methods can broadly be classified into field and laboratory studies. Comprehensive fi eld studies provide the best basis for emission estimates, whereas laboratory studies can give valuable insight into the relative importance of factors affecting emissions (Granli and Bockman, 1994). Smith et al. ( 1994) summarised the 20 fluxes measured by the various methods over the six consecutive days (Table 2.4). Chapter 2 - Literature Review 25 Usually gas samples are taken and analysed in a laboratory. N2O production can be precisely measured using gas clu·omatography (GC) and easily identified over relatively low atmospheric background concentrations (about 310 ppb) . The precision of GC is sufficient fo r chamber methods and can also be used for micrometeorological stations when 20 flux is adequate. As a general rul e. field N2O flux samplings should preferably be done without any disturbance to the measured system. Currently, chamber methods are more readily avail able and inexpensive and used most often in studies reviewed. However. other methods such as micrometeorological , ultra-large chambers with IR spectrometer and soi l air analysis are available. 2.12.1 Chamber Methods ( 'losed Chambers 120 in cl osed chambers is determined directly by measuring the short term buildup or decrease in concentration in a sealed enclosure placed over the land surface. Normall y, 0ux chambers are s imply inverted containers forming a trap for gases emitted from the soi l ( or water) surface, and from which gas samples can be removed at intervals to determine the rate of flux . They are usually relatively easy to construct from a variety of readily ava ilable material s which are inert fo r the gas of interest (IAEA, 1992). Closed chambers can be divided into two types: fixed-base and removable. • Fixed-base chambers are constructed with separate collars which are inserted into the soil for the duration of the experiment, and gas-tight chamber is attached for short period (Am bus and Christensen, 1994 ). • Removable chambers are constructed as a one unit which are installed in the so il for a sampling time and then removed until next measurement date (Mosier and Hutchinson, 1981 ). Both types are widely used. Fixed-base type offers less disturbance to the soil because once inserted collars can stay until the end of experiment. Also pressure change in the soi l while inserting chambers is overcome by using collars. However, they do not Chapter 2 - Literature Review 26 provide randomisation of chamber positions as opposed to the use of removable type and also can interfere with agricultural operations such as tillage. Table 2.4: Comparison of flux measurements by flux gradient micro­ meteorological, combined megachamber/long-path IR, and small chamber gas chromatograph (GC) methods. Methods N 20 Flux , ng N m s April 8 April 9 April 10 April 11 Eddy correlation ND 75-113 ND ND microm eteoro logical TDL Flux grad ient ND ND 20-70* S 1-59 microrn eteoro logical TD L FTIR 4 7- 103 0-1 27 34 -55 ND GC 45 35 ND 65 Megachamber/long-path ND ND 247 295 IR (222 -32 1 )1' (266-384)-\ 0. 126 m- chambers with in ND ND 61 -230 ND tent area1GC analys is ( 128)t 0.126 m- and 0.49 m- 149-824 13-4 14 29-337 24-488 cha111bers~ /GC analysis (355H ( 160)! ( 177):;: (205):;: 0.0079 m- chambers in ( 19 1)! 74-854 ND ND 49x49 m grid/GC analys is (292)t l\'D. 1101 de/ermined. *.J./-6./ during 1he afiernoun and 20-70 under inversion conditions. /Single intergrated value and uncertainty range from -IO to + 30% of value . . / Range and 111ean. §Ungra:ed area only. Open Chamhers April 12 April 13 ND 38--1 8 ND ND I ND ND 73 -1 05 ND ND ND ND ND 21 -406 22-528 ( 142)! (2 I 8)t ND ND Open chambers are coupled to the atmosphere via an air inlet through which outside air is continuously drawn into the chamber and forced to flow over the enclosed soil surface . The gas flux from the soil is calculated from concentration difference, flow rate, and area covered by the chamber (IAEA, 1992). The main advantages of open compared to closed chambers are: Chapter 2 - Literature Review 27 • effects of N 2O accumulation are reduced or eliminated. This permits longer periods of measurements • conditions outside and inside the chamber can be more similar than for closed chambers. However. open types are sensitive to pressure deficits inside the chamber caused by the induced air flow which may cause artificially high fluxes. N2O concentrations are lower than in closed chambers, thus the method is less sensitive. 2.12.2 Micrometeorological Methods The basic concept of rnicrometeorological approaches, described in Denmead (1983), is that gas transport is accomplished by the eddying motion of the atmosphere which displaces parcels of air from one level to another. Three different approaches are used: eddy correlation, flux gradient and mass balance: • eddy correlation technique is obtained by conelating the instantaneous vertical wind speed at a point with the instantaneous concentration of that gas. In the natural environment the eddies which are important in the transport process occur with frequencie s extending up to 5 to 10 Hz. Therefore a rapid response detector is required; • flux gradient theory is based on the concept of diffusion of the gas along its mean concentration gradient. The method requires measurement of mean gas concentrations at a number of heights above the surface and knowledge of the appropriate diffusion coefficients; • in mass balance method, gas flux density is related to the horizontal distance from the upwind edge of the field and the top of the air layer influenced by the emission of the gas. The method assumes that the mean horizontal turbulent flux is much smaller than the mean, horizontal advective flux. The top of the air layer influenced by the emission of the gas is a function of stability and surface roughness but can generally be simply estimated. Chapter 2 - Literature Review 28 The advantages of the micrometeorological methods are that they do not disturb the environmental or soil processes which influence gas exchange; they allow continuous rapid measurement, thus facilitating the investigation of environmental effects; and they provide a measure of the average flux over a large area, thereby minimising the sampling problem created by point to point variation. However. compared to soil cover method rnicrometeorological techniques could be used only during periods of high N20 flux, because only then were the differences in N20 concentrations between sampling heights greater than the minimum detectable difference of the analytical method (Mosier and Hutchinson, 1981 ). Other di sadvantages are the need for large uniform fields with a m1111mum of air turbulence and appropriate weather conditions as periods of turbulence can give en-oneous results (Granli and Bockman, 1994). 2.12.3 Ultra-large Chambers with Long-path IR Spectrometers The ultra-large chambers. described in IAEA (1992) , is a new promising technique but not yet ready for routine flux measurements. Infra-red absorption spectrometers are now available which can give an average value for the concentration over distances of tens or hundreds of meters. These instruments are potentially very useful for measurements of average emi ssions from a whole experimental plot, by covering the plot temporarily with a large canopy. to act as a chamber and retain N20 emitted from the soil. Such a procedure also has the advantage, compared with the flux-gradient method and other rnicrometeorological methods mentioned above, of not requiring the associated measurements of windspeed etc., and of allowing the comparison of emissions from different treatments in a replicated field experiment, without any interference being caused by emissions from adjacent plots. Two such systems are available: a Fourier Transform Infra Red (FTIR) spectrometer with a mirror system which allows multiple reflections, and thus a total path of up to l km, that is capable of measuring N20 concentration changes down to a fraction of 1 ppb; and simpler, less sensitive IR spectrometer (the Siemens-Plessey 'Hawk' system), with the capacity to detect a concentration change of ca. 25 ppb of N20 (IAEA, 1992). Chapter 2 - Literature Review 29 2.12.4 Nitrous Oxide Emissions Variability A problem in measuring N20 emissions from soils is the large spatial and diurnal variations. In some cases the variation is considerable within only few metres on the same experimental plot. There are number of studies which report high spatial variation and also there is seasonal variation. Seosonol Varia tion Most studies report that seasonal pattern of N20 emissions occur during winter time, wh ich is not surpri sing due to soil moisture factor which tends to be higher in winter. For example. Kessavalou et al. ( 1998b) had 3 to 4 7% of nitrous emissions accounted fo r ,vinte r fluxes. Kaiser et al. ( 1998) reported 50% of annual N20 emiss ions during w inter (October to February) in Northern hemisphere, which resulted from both physical release of subsurface -prod uced N20 during soil freezing and microbial 20 production during daily thawing and freezing cycles. Moreover, the authors stated that the total ' 20 losses during the winter increased with decreasing dry matter-to-N-content ratio of the plan t residues incorporated into the soi l by ploughing. This suggests that plant spec ies can influence N20 emissions . While tha,ving of the soil surface dictates timing, N availability strongly governs the magnitude of the N20 em issions (Lemke et al., 1998a). The authors reported 16 to 60% 20 emi ss ions occurred during and just following snow melt in the spring. These mentioned above emi ss ions were related to wet season. Spatial I 'ariat ion Magnitude of N20 emissions is highly variable due to complex interactions between controlling factors, and high coefficient of yariations are often reported. Even within the sampling boundary 20 fluxes can differ significantly. For instance, Lemke et al. ( 1998b) reported large-scale variability of up to 92% which the authors related to differences in clay content. Factors such as topsoil nitrate, ammonium, and water contents and air permeability may contribute to the variability of N20 emissions (Ball et al. , 1997). However, sometimes variations are hard to explain without detailed sampling and measurements . Chapter 2 - Literature Review 30 Unfortunately, often reported coefficients of variation range from 8% to 325% (Malm1ood et al., 1998) to up to 6001 % (Teira-Esmatges et al.. 1998) are common in these kinds of studies. The spatial variability of denitrification could be attributed to ·'hot-spots" associated with high local concentrations of organic matter. The formation of these hot spots was governed by the creation of anaerobic conditions due to increased resp iration. following the introduction of a source of decomposable organic matter to the soil microbes (Christensen and Tiedje, 1988). 2.13 Summary This literature review has shown that N2O contribution to the global greenhouse gases is small but significant. Its contribution in New Zealand context is four times of that elsewhere. Detrimental impact of N2O in the environment is mainly trapping of outgoing terrestrial radiation and its involvement in destruction of stratospheric ozone. N 2O is known to result mainly from soil denitrification and nitrification processes. Although N2O is usually by-product of these processes magnitude of its release depends on several controlling factors and complex interactions between them . Emissions of 20 are best explained by soil moisture content, temperature. organic matter, nitrate level , physical properties and pH. However, they are not always correlated with those factors due to complexity of soil micro-organisms responsible for soil chemical processes. Measurement methods for N2O fluxes have evolved from simple chamber technique to microrneteorological and infra-red spectrometer methods capable of estimates from larger area fields. Despite new techniques N2O emissions are still characterised with high spatial as well as temporal variations making inventory assessments less confident. Despite such inherent variations in spatial variability it is important to provide a degree of quantification of these measurements. Hence, this study was undertaken to understand the likely quantifiable impacts of soils cropped with tillage, no-tillage and compare this with pasture soils N2O emissions. Chapter 3 Materials and Methods 31 3.1 Experimental sites Two field sites were chosen for this study, based on their suitability for comparison. The primary site Kairanga (40° 21' S, 175° 39' E) and the second site at Massey University (latitude 40° 23 ' S, 175° 38 ' E) were within the Manawatu region of the North Island of New Zealand. The summary of rainfall and soil temperatures for both sites is given in Table 3.1 . The cooperating farmers used management typical practices and the presence of our instrumentation in the fields did not interfere with the practices . Table 3.1: 1998-99 Month s November December January February March April May June July August September Summary of monthly climatological observations data taken at 09.00 hrs at AgResearch and Aorangi Research stations close to study sites. AgResearch Aorangi I Mean Air Soil Mean Soil Rainfall Temperature Temperature Rainfall Temperature (mm ) (QC) (QC) (mm) (QC) 111111 max m111 max 86 .0 16.4 19.0 46 .5 12.3 20.8 17.4 36.4 18 .0 20.7 39.4 14.9 25 .0 21.0 56.2 20.2 23.4 20.1 13 .1 24.9 20.2 43 .6 19.2 22.3 34.3 14.3 23 .9 18.6 45.4 18.6 20.6 65.3 10.1 19.3 13.7 77.8 15 .2 16.8 114.6 7.6 16.9 10.9 136.0 12.8 14.5 67.9 4.9 14.2 . 8.1 102.0 10.6 17.1 94.5 4.9 13.5 7.8 107.0 n/a n/a 86.5 3.4 13.4 6.5 58.0 n/a n/a 49.6 6.0 15.8 9.9 62.8 11.5 12.8 Chapter 3 - Materials and Methods 32 3.1.1 Kairanga Site These prime experimental sites are located about IO km northwest of Palmerston North. The soil has been classified as Kairanga silt loam with very poor drainage (Table 3.2) and soil characteristics described in Saggar et al. (2000). Table 3.2: Selected soil characteristics of the fields at the Kairanga site. Sites Soil classification pH Organic Total Microbial Grid C N C NZ us reference I :2.5 (%) (%) mg/kg soil so il:water Permanent Typic Typic NZMS 5.65±0.01 * 4.99±0.32 0.47±0.09 1585±5 Pasture Orthic Endoaquept 260 S24 (PP) Gley 272 609 17 years Typic Typic NZMS 5.58±0.04 3.30±0.49 0.28±0.05 474±11 cultivated Orthic Endoaquept 260 S24 (Kl7) Gley 272 609 34 years Typic Typic NZMS 6.20±0.02 2.03±0.30 0. I 8±0.09 261±15 culti vated Orthic Endoaquept 260 S24 (K34) Gley 272 609 * Values represem mean±standard error These farms representing a permanent pasture (PP), 17 years conventional cultivation (K 17) and 34 years conventional cultivation (K34) grown maize (Zea mays) crop (Figure 3.1) were included in this study. Permanent pas ture (Figure 3.2) was used as a control and had 160 cows strip-grazed on 3 acre ( 1.21 ha) sections of the farm for 9 hours during the day and 12 hours over the night. Maize was sown at both cultivated farms in late October 1998 . The 34 years farm received 306.5 kg/ha of (15: 10: 10:8) fertilizer at sowing and additional 300 kg/ha of urea as side dressing applied in early December 1998. Maize crop harvested for silage produced 16 t dry matter/ha. The Kl7 plot received 153 kg/ha (half of K34) of 15:10:10:8 starter fertiliser applied end of October and 300 kg/ha of urea side-dressed in early December 1998. The yield of maize grain was 6.9 t/ha. Ch apter 3 - Material s and Methods 33 Figure 3.1 Kairanga site 34 years (K34) treatment field . Figure 3.2 Permanent pasture (PP) field treatment at Kairanga site. Chapter 3 - Materials and Methods 34 The crop growth season was dry. Coupled with the very dry summer (which followed a ve ry wet October), the fertilizer added and especially the urea side-dressing would not have been utilized effectively, because of the lack of rainfall and very dry soils. Urea pellets were still visible two weeks after application. 3.1.2 Massey Site Massey experiment plots were established by the previous Department of Agricultural Engineering (now the Institute of Technology and Engineering) at Massey University Turitea campus, Palmerston North in 1995 (Figure 3.3) (Choudhary et al. , 1996). The soi l has been classified as Ohakea silt loam (Typic endoaqualf) and Gleyic Luvisol (F AO), v.: ith weakly clay-illuvial pseudomadenti-pallic representing youngest yel low­ grey earth with poor natural drainage. The experimental design involved: • Conventional Tillage (CT) - permanent pasture land was converted to a double crop rotation using conventional practices. Plots were moldboard ploughed and followed by rolling and two passes of a power harrow for seedbed preparation. • No-Till age (NT) - similarly , permanent pasture land was converted to a double crop rotation using no-tillage practice. There was no prior cultivation and weeds were controlled by using Roundup (360 g/1 glyphosate) at 4 I/ha. Sowing was done by direct drilling. • Permanent Pasture (PP) - was used as a control. Together with other treatment plots they were rotationally sheep grazed. Plots consisted of high fe11ility grass species including ryegrass (Lolium perenne), yorkshire fog (Ho/cw; lantus), poa (Poa spp) and cocksfoot (Dactyl is glomerata) represented 65% of sward, white clover (Tr if"olium repens) and other legumes ~20% (suckling clover (T. dubium Sibth). subterranean clover (T. Subterraneum) and lotus (Lotus spp)). Few weed species including catsea (Hypochaeris radicata) , hawkbit (Leontodon taraxacoides (villars) Merat), ribgrass (Plantago lanceolata), and chickweed (Cerastium glomeratum Thuill) constituted the remaining 15% of the pasture community (Aslam et al. , 1999). Chapter 3 - Materials and Methods 35 Figure 3.3 Massey site experimental treatment plots. Chapter 3 - Materials and Methods 36 10m NT MP 3.6m I p MP I . NT . I . I Sm p Sm .. ..... ,· NT MP p MP NT p Figure 3.4 Schematic layout of experimental treatment design plots at Massey site. CT - conventional tillage; NT - no-tillage; PP - permanent pasture Chapter 3 - Materials and Methods 37 Each treatment had four replicates arranged in a randomized block design. Each plot was 17 m long and 3.6 m wide with a 5 meter headland for machinery operation on both s ides of the fi eld (Figure 3.4). Summer fo dder maize was sown on the 8th of December 1998 at seed rate of 65 kg/ha. Aitchi son seed drill (Seedmatic 1112) was used for sowing and simultaneous Nitrophoska (12% N, 10% P, 10% K, 1% S) fertilizer was applied at the rate of 120 kg/ha. The maize was then grazed by a mob of sheep at maturity. The winter oats were sown at a seed rate of 120 kg ha-1 on 19 April 1999 with 200 kg/ha of Nitrophoska fe rtil izer with the Aitchison seed drill (Seedmatic 111 2). The crop at maturity was later sheep grazed. 3.2 Measurement of Field N2O Emission Nitrous oxide measurement techniques available nowadays include small to medium chambers, ultra-large chambers with IR spectrometer, micrometeoro logical method and so il air analysis. All the methods are widely used, however, in most of the studies N20 flu xes fro m so il and plant systems are measured using chamber methods involving gas chromatography analys is of N20 (Mosier and Hutchinsom, 1981 ; Ryden and Rolston, 1983; Goodroad et al. , 1984 ; Smith et al. , 1"994; Carran et al. , 1995; Kaiser et al. , 1998). Micrometeoro logical methods can measure gas emissions fro m the fields w ithout disturbing the gas exchange between the atmosphere and the soil/crop system. However, the usefulness of these method is limited by requirements such as : expensive equipment, extensive uniform surface areas with constant atmospheric conditions during each measurement period. In addition to those requirements small concentration fluxes are difficult to measure w ith micrometeorological methods. Few requirements of micrometeorological methods can be overcome by another promising technique such as ultra-large chamber with IR spectrometer which can measure N20 emissions from large areas and do not require uniformity of the fields and associated measurements of wind speed. The technique is limited to gas concentration measurements until the methodology for making the necessary accompanying Chapter 3 - Materials and Methods 38 measurements to estimate gas flux is perfected (Mosier, 1990). Small to medium size chambers are more suitable for comparing N20 emissions from treatments on a typical small plots as at Massey site. Other obvious advantages of this method are: • easy handling and randomization of chamber position over the duration of an experiment, as opposed to the use of fixed sampling positions • no interference with agricultural operations such as tillage particularly at farming sites used • l0,,1,• ratio rates of N20 fluxes can be measured • no extra equipment requiring electrical supply is needed • the chambers are simple and relatively inexpensive to construct Nitrous oxide flux measurements were made periodically each month at both sites for a year using closed chamber technique developed by Mosier and Hutchinson (1981) with s li ght modifications. Attempts were made to coordinate measurements at certain interval s, however, because of instrumental and logistics problems, constraints on time , and occasionally unsuitable weather conditions sampling times varied during experiment. Additional fortnightly measurements were also carried out. lAEA (1992) recommends to use at least four chambers if the coefficient of variation is less than l 00%, and at least 20 replicates for each treatment if CV is above l 00%. Mosier and Hutchinson ( 1981) estimated N20 flux density by micrometeorological method. and it always fell within the range of the four individual flux estimates made by the soil cover method. The authors interpreted this agreement as evidence that four soil cover determinations adequately sampled soil variability in N20 emissions. In this study due to the extent of the project and funds available only four chambers were employed for each treatment. 3.2.1 Procedure Nitrous oxide flux measurement procedure ·involved the following: 1. installation of chambers Chapter 3 - Materials and Methods 2. Applying high vacuum grease and closing the lid 3. Use .\)Winge to take gas sample through the air vent.fitted on the lid cover (both. syringe and air vent, fitted with three way tap) -I. Repeat step 3 after I , 2 and 3 hours Ji-om.first gas sampling 5. Transport gas samples to the laboratory for analysis by electron capture detector (£CD-CC) 6. Calculate N20.fluxji·om the concentration increase 39 The sampling chamber consisted of two cylinders glued together (Figure 3.5). Main outer custom made cylinder was 0.1 m tall and have been designed to be closed with the lid. Cylinder and the lid were molded from PVC material and were 6 mm thick. The lid was designed to be screwed on the top of the cylinder. In order to make the chamber longer cylindrical pipe was glued to the inner part of the custom made cylinder. Inner cylinder was a 0.2 m tall polypropylene pipe of diameter 0.24 m. The bottom end of the chamber \\'.as machine-beveled to facilitate its insertion into the ground . Chambers were inserted into the soil depth of 0.1 m giving a head space volume of 4.6 1 1 on enclosure with a plastic lid (Figure 3.5). An area covered by chambers was 45 cm-. To assure that there was no leakage beneath the lid cover high vacuum grease was applied onto rubber band used as a seal. Chambers were removed after measurements until the next sampling. Quadruplicate measurements were made at approximately same locations within a small uniform area of the field on each sampling date at Kairanga site and chambers at Massey site were placed along the treatment plots. Chambers in cropped fields at Kairanga were placed between maize rows. At Massey, due to cropping for grazing, distance between rows did not allow placement of chamber in the interrow. However, plant interference during chamber installments were kept to a minimum. Adequate precautions were taken to minimize disturbance of the energy and mass transfer processes normally operating at and above the soil surface. For example, temperature raise tends to enhance soil nitrous oxide flux. To avoid temperature changes Chapter 3 - Materials and Methods 40 111 the so il and atmosphere under the chamber, during sampling on sunny day , the chambers were covered with insulating material and basket over it. Initi ally the chambers were left for three hours with an hour interval for air sampling in order to detect if there was any leakage. Gas samples were taken fro m the headspace immediately after seali ng (t0) and at I hour time intervals thereafter over a period of 3 hours. This was done to check the linearity of concentration increase in the chamber. Since after several sampling occasions N20 concentration tended to have linear increase over three hour peri od, in later samplings it was reduced to only two hours. Head space gas samples were collected through the air vent on the lid cover using 60 ml po lypropylene syringes fi tted with 3-way taps (F igure 3.6) . Syringes with gas samples we re held in the plasti c bags in set of four and were kept in air conditioned laboratory with constant temperature fo r overnight. Because of concern over the pem1eability of po lypropylene to 20 , fi eld samples were analyzed as soon as possible after co llection, usuall y v;ithin 24 hours. Ambient air at t0, collected at the time of chamber install ation, was used as a reference fo r calculating N20 gas fluxes. Nitrous oxide flux was computed from the concentrat ion increase over the time. The follovving fo rmula was used fo r calculation: [ ( 28.0134 x 273 )x v x r i -TOX_ l _ ] 22.4 273 + t°C I 000 1(2;3) 2 .V ,0- N fl ux = -==---------------~ (~Lg N2O-N/m /h) - Area where: 28.0 I 34 and 22.4 - are constants; t°C - soil temperature at 10 cm depth; V - chamber vo lume (L) ; T 1-T0 - is the change in ppb measured; T 1 - N 20 measured after one hour interval ; T0 - N20 measured at the time of closing lid (ambient concentration); 2 A - area covered by the chamber (m ); 1/1(2;3)- is I/incubation time in hours, i.e. 1 (or 2 or 3) hour buildup Chapter 3 - Materials and Methods Figure 3.5 Clo ed cover chamber used to collect field nitrous oxide (N2O) emissions. 41 Chapter 3 - Materials and Methods 42 3.2.2 Analysis of N2O Concentration Analyses were done using Mosier and Mack (1980) method capable of precise determination of ambient N20 concentrations. Air samples were analyzed for N20 using a GC. Hewlett-Packard 3385A integrator recorded the sample chrnmatogram and performed peak area integration. Argon-methane carrier gas consisting methane ( 10±0.3%). oxygen (< l 0 ppm), water (<5 ppm) in argon was used in the columns. Gas samples from sy ringes were introduced into a 5 cm3 gas sampling loop through an inlet system shown in Figure 3.7. Each analysis took approximately 6 minutes and 10 samples per hour could be routinely analyzed. To calibrate and check for possible contamination of the sampling system, routine tests were made by running analysis of standard preparation. Standard gas was prepared with concentration of 914 ppb. Standard Preparation To prepare standard gas 1.067 litre jar was used. First the jar was flashed with pure nitrogen for about 20 seconds and then filled with nitrogen gas. A 50 ml sample of filled nitrogen gas from the jar was taken out and replaced with a 50 ml of standard gas (20 ppm N20 in nitrogen ( 19.2±0.4 ppm N20)) to give a final concentration of 914 ppb. 3.3 Pilot Experiments 3.3.1 Exploratory Sampling Initially a pilot study was undertaken to measure N20 flux using the chamber technique. Exploratory sampling was conducted on November 3, 1998 after maize planting. A field with a history having 34 years of continuous cultivation with maize/barley was chosen for this purpose. Three chambers were installed in the morning. Four gas samples were taken from each chamber, first at the time of closing the lid, and after 1, 2 and 3 hours. Due to the time constraints only one sample could be analysed next day on November 4. The rest of gas samples were analysed on November 5. Chapter 3 - Materials and Methods 43 3.3.2 Spatial Variability Due to high spatial variabi lity reported by other studies of N20 emissions (Carran et al., 1995; Ruz-Jerez et al.. 1994) twelve chambers per treatment at Massey site were used to determine spatial variability. 3.3.3 N2O Emissions During Seedbed Preparation Nitrous oxide emissions were measured during seedbed preparation 111 March. Chambers were installed in the CT plots at Massey site within few days after ploughing. Fluxes were compared to the emissions measured next day immediately after power harrowing operation. 3.3.4 Prolonged Continuous Measurement Concurrently, in March to examine effect of prolonged continuous measurement on 20 flux gas samples were collected over the longer time than usual 2-3 hours. Chambers \Yere installed in all treatment plots at Massey site and were left in the soil for 9 hours with periodic sampling at 0, 1, 2, 3, 5, 7, and 9 hours. 3.4 Ancillary Measurements 3.4.1 Soil Moisture Soi l moisture at IO cm depth was measured each time during N20 measurements. Field samples were co llected, weighed, oven-dried to constant mass at 105°C and then weighed again. The final mass M 5, and the difference between the wet and dry masses Mw were used to calculate gravimetric soi l water content (SMC): SMC= M ,.. x 100% M 3.4.2 Soil Temperature Soil temperature at IO cm depth was measured at each nitrous oxide sampling time at each site using handheld digital thermometer. Temperature probe used for this purpose was inserted into the soil next to each chamber. Temperature indications were recorded manuall y. Chapter 3 - Materials and Methods 44 3.4.3 Soil pH, Total C and N Soil pH, total carbon and nitrogen, and bulk density were measured once. Soil organic carbon content was measured using a Laboratory Equipment Corporation (Leco) high­ frequency induction furnace (Blakemore et al. , 1987). The measurements were done on 0.5 subsamples taken from air-dried soil collected from each plot. Samples were collected from two depths, 0-10 and 10-20 cm, from each plot after the winter oats being grazed. and from the adjacent pasture plots on September 1999. Soil from each plot was bulked before subsampling. Organic-N content of the 0-10 and 10-20 cm samples were determined by the Kjedahl method. 3.4.4 Rainfall Data Rainfall data were obtained for both sites from nearby stations. AgResearch (E05363) rainfall data was used for Massey site as it was located around Turitea campus. For Kairanga fields data from Aorangi Research Station on Lockwood road was obtained. 3.4.5 Statistical Analysis An exploratory analyses of the data was performed to determine the distribution of N20 emission rates. The results of the first step indicated whether or not the data needed to be transformed in order to satisfy the normal distribution assumption of the ANOV A analyses. Log-transformation was satisfactory for the majority of sampling dates . A general linear model procedure (GLM) was used for analyses of experimental data. An analyses of variance (ANOV A) using test of least significant difference (LSD) at 5% confidence level was used for comparisons of treatments. Regression analyses between measured so il moisture and N20 emission rates were completed using data analysis option of Excel. All the raw data collected during the period of study is appended in appendices I to II. Chapter 3 - Materials and Methods 45 Figure 3.6 Insta lled chamber fitted with 3-way tap for withdrawing air samples F igure 3.7 Injecting 10 ml of gas sample into gas chromatography (G C) instrument to measure nitrous oxide ( 20 ) concentration. 46 Chapter 4 Results and Discussion 4.1 Overview Nitrous oxide emissions measured with closed chamber technique in this study are believed to have wide soil variability (Mosier and Hutchinson. 1981 ). During field sampling all precautions were taken to minimise possible leakage or disturbance to processes occurring at and above soil surface. However, few problems arose during field samplings especially during dry and hot summer period of measurements. Forced insertion of chambers into the soil was rather difficult in dry soil despite the bottom of the chambers being sharp. IAEA ( 1992) guidelines have recommended that the chamber should only minimally affect the surface boundary layer resistance. In these samplings, however, it is unknown what effect direct forcing of chambers caused. The problems associated with the in-situ removable closed chambers technique for measuring N2O fluxes have been di scussed previously by Mosier (1990) and IAEA ( 1992). In thi s study. one major problem with the chambers employed rather than the methodology was evident. As the chambers were forced into the soil, it led to cracks and occasionall y breakage of chambers itself. However, there were suffi cient chambers to obtain four replicates from each treatment. It is worth noting that Kessavalou (1998b) recorded more N20 from the cropped row than inter row locations . The authors reported mean N20 flux from the row 0.5 to 2. 7 times greater than that from the interrow zone, probably due to more avai lable C and subsequently greater nitrification and denitrification activities in the vicinity of plant roots where the fertiliser N was placed. This study did not encompass the measurements fro m the interrow, therefore emissions from the management systems assessed may have underestimated actual N20 fluxes. Chapter 4 - Results and Discussion 47 4.2 Pilot Experiment 4.2.1 Exploratory Sampling Results obtained from a preliminary study on calibration of the chamber technique and assessment of field variability in N2O emissions are shown in Table 4.1 N2O emissions varied widely among the replicates. Replicate 1 gave over 4 times as much N2O as replicate 2 during the first hour. The low rate of N2O emission in replicate 2 may partly be due to some leakage. The soil at the time of this study was dry and the insertion of chambers probably did not allow tight seal between the chamber and the field soil. Table 4.1: 20 emission rates using the exploratory samples in the pilot experiment Replicates Hourly increase (µg 20- /ha/hr) R1 22 16 10 R2 5 2 2 R3 14 12 10 The results (Table 4.1) also show that the rate of N2O emission in all the replicates declined over time. This rate of decline in N2O emissions among the three chambers was not consistent. Because of these variations it was decided to employ a minimum of four chambers for future regular measurements. The other variable was the timing of analysis after collection of the sa