Investigating the role of management and measurement technique on the temporal and spatial variability of carbon dynamics and nitrous oxide emissions from temperate grasslands
Citation:
Murphy, Rachael, Investigating the role of management and measurement technique on the temporal and spatial variability of carbon dynamics and nitrous oxide emissions from temperate grasslands, Trinity College Dublin.School of Natural Sciences, 2022Download Item:
Rachael Murphy Final Thesis Version.pdf (PDF) 4.365Mb
Abstract:
The Earth’s atmosphere consists primarily of nitrogen (N) in the form of dinitrogen (N2), and oxygen as well as greenhouse gas (GHG) molecules including water vapour, carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4). Anthropogenic activity through land use, land use change and intensive agricultural practices has contributed to the increase in ambient concentrations of these GHGs, reaching annual averages of 440 ppm for CO2, 332 ppb for N2O and 1866 ppb for CH4 in 2021. Consequently, this has resulted in changing climatic variables such as increases in global surface temperatures of 0.99 °C between 2001 and 2020 relative to the period between 1850 and 1900. In Ireland, the occurrence of extreme climatic events has increased, particularly in last decade, including heatwaves, droughts, storms, heavy precipitation, flooding and extreme cold spells. Emissions of CH4 and N2O are important in forcing such climatic events due to their respective global warming (GWP) potentials of 28 and 265 respectively, relative to CO2 over a lifespan of 100 years. In Ireland, agricultural landscapes are dominated by grasslands, accounting for approximately 58 % of the land surface area in Ireland, of which 40 to 279 kg N ha-1 yr-1 in the form of inorganic N is applied to grassland pastures depending on the stocking rate for dairy cows (1.0 to > 2.47 LSU/ha-1). Furthermore in 2020, 37.1 % of Irelands total GHG emissions were derived from the agricultural sector, and of this, 57.5 % was derived from enteric fermentation, followed by agricultural soils at 26.8 %, and to a lesser extent manure management, fuel combustion, liming and urea application at 10.3, 3.0, 1.9 and 0.5 %, respectively. The intensification of agriculture to meet the demands of a growing global population has altered the natural production and emission of CH4 and N2O. The formation of CH4 is catalysed by methanogenic bacteria during anaerobic metabolism where soil organic materials are broken down. In agricultural systems, CH4 is produced by enteric fermentation from ruminant livestock, accounting for 58 % of Ireland’s agriculture derived GHG emissions in 2020. Sources of N2O include the combustion of fossil fuels, waste management and industrial processes such as the formation of chemical N fertilizers. In Ireland, approximately 1 to 4 % of applied N to agricultural soils as chemical N fertilizers or animal excreta is emitted as N2O depending on the N loading rate of the inputs as well as environmental conditions which will influence the rate of N2O emissions produced, such as temperature and soil moisture. In Ireland, agriculture emits 90 % of the nation’s N2O emissions of which 38 % is derived from synthetic fertilizers, 23 % is derived from animal excreta during grazing and 14 % is derived manure management. Soil derived emissions of N2O are formed as either a by-product of the microbial process of nitrification under aerobic conditions, or as a transitional product of denitrification under anaerobic conditions. The spatial heterogeneity of agricultural soils facilitates the presence of both aerobic and anaerobic microsites existing in close proximity, and gradients of soil conditions that will influence the magnitude of microbial produced N2O, such as aeration, redox potential, temperature, moisture, substrate availability and N inputs. Following the application of N fertilizer to managed grasslands, N2O fluxes typically display a peak and decay pattern over time which is characterized by a log-normal distribution, normally lasting between 5 and 20 days. Due to the inherent spatiotemporal variability associated with N2O emissions from agricultural landscapes, it is still a difficult task to quantify field scale emissions of N2O with low uncertainties. To date the most commonly used method to quantify field scale emissions of N2O is the static chamber technique, which consists of taking manual gas samples, mainly once a day and over small spatial domains, generally less than 1 m2. Historically, these techniques has been used for investigating treatment effects on N2O emissions, but due to its limited spatial and temporal resolution, flux measurements are often attributed with high uncertainties. Conversely, micrometeorological techniques, such as eddy covariance (EC) are capable of making continuous, high frequency ecosystem scale (1 km2) flux measurements of N2O through recent developments in fast, high precision absorption spectrometers such as quantum cascade lasers (QCL). On the other hand, both high and low flux measurements by EC are integrated over a given area (i.e. the footprint), and therefore it can be challenging to disaggregate between different emission sources over a given spatial domain.
The overarching aims of this thesis were:
1. To investigate the spatial and temporal variability and potential disparity between N2O flux measurements made using static chambers and EC techniques from a uniformly emitting surface (i.e. a grassland under silage and fertilizer management) and additionally, to assess the methodologies used to analyse and integrate log-normal chamber N2O flux data (arithmetic and Bayesian statistics) (Chapter 4).
2. To optimize the application and use of both static chambers and EC techniques to quantify N2O emissions under a more complex, heterogeneous emitting surface (i.e. a grazed managed grassland), where the EC technique provides high resolution, low uncertainty field scale flux measurements, and the static chamber technique assess the source contribution of various N sources from the system in order to upscale localized N2O flux measurements to the field scale (Chapter 5).
3. To assess the influence of cut and grazing management activities and their associated emissions (CO2, N2O and CH4) on the net carbon (C), net N and net GHG balance (NGHGB) (i.e. the net GHG exchange minus C exports) of the grassland at the field scale (Chapter 6).
Key findings presented in Chapter 4 showed that EC and static chamber N2O flux measurements were most comparable when N2O flux values were high (> 115 N2O-N µg m -2 hr -1) and showed spatial and temporal alignment when the chamber sample size was large (n ≥ 15) and the log-normal distribution of the dataset was accounted for using Bayesian statics. Conversely, when the chamber sample size was small (n ≤ 5), the Bayesian model produced large uncertainties due to the inability of the model to constrain an arithmetic mean from a log-normally distributed data set, thus suggesting that greater replication is necessary for constraining the spatiotemporal variability of static chamber flux measurements. Field scale N2O flux measurements using static chambers with Bayesian statistics (3.13 [± 0.24] kg N ha -1) were closer in magnitude with N2O flux measurements using EC techniques (3.35 [± 0.5] kg N ha -1), compared to the arithmetic approach (2.98 [± 0.17] kg N ha -1), highlighting the importance of accounting for the log-normal distribution of chamber N2O flux measurements for quantifying more realistic estimates of field scale missions of N2O.
In Chapter 5, the field site was under a grazing management regime, where different sources of N were applied in the form of fertilizer and animal excreta. As the EC technique is unable to disaggregate between emission sources, static chambers were used in tandem to quantify emissions from N sources that are characteristic of grazing systems including calcium ammonium nitrate (CAN), the combination of CAN and synthetic cow urine (SU+CAN), and the combination of CAN and dung (dung+CAN). Mean emission factors (EF) for CAN, SU+CAN and dung+CAN were quantified from four grazing events at 2.78 ± 0.90, 0.59 ± 0.12 and 0.64 ± 0.15 %, respectively, and used to upscale localised N2O flux measurements using static chambers to the field scale (FCH FIELD) for comparability with field EC based flux measurements. Similar to the cumulative flux findings observed in Chapter 4, total N2O-N emissions measured by EC were higher and with lower uncertainties relative to FCH FIELD at 6.62 ± 0.33 and of 5.16 ± 2.04 kg N ha-1, respectively. The seven-fold higher uncertainty attributed to FCH FIELD measurements relative to EC, was due to the low spatial and temporal resolution of the static chamber technique coupled with a low sample size (n = 5 per treatment per grazing), which collectively makes constraining the uncertainty in static chamber N2O flux measurements a difficult task. However using the static chamber technique in tandem with the EC technique provided valuable insights into the source contribution of field scale emissions of N2O from a grazing system. For instance, approximately half of the total N2O-N losses were derived from animal excreta, one third were derived from CAN and the remaining emissions represented background fluxes. Furthermore, approximately 20 % (1.01 kg N ha-1) of the total N2O-N flux calculated by FCH FIELD occurred during a spring grazing event, where this observation was further reinforced by statistical analysis showing a significant (p < 0.001) interaction between N2O emission and time and treatment. These findings show that the timing of grazing events can have a significant impact on the total annual N2O-N losses.
Chapter 6 synthesizes the impact that management activities described in chapters 4 and 5 have on the net N, C and overall NGHGB at the field scale by quantifying the N and C imports into and exports out of the grassland system. Findings from this study showed that N imports influenced the net N budget of the grassland, where under silage management the system had a net neutral N balance of 0.1 ± 6.0 g N m-2 yr-1. In contrast, under grazing management where N imports were higher (i.e. from both fertilizer application and animal excreta) and the system transitioned into a net N sink at -17.9 ± 5.5 g N m-2 yr-1. The net C balance showed that the grassland was a greater sink of C under a grazing management relative to the cut management at -311.5 ± 81.8 and 61.6 ± 26.7 g C m-2 yr-1, respectively. This was mainly due to both larger C exports from silage cuts reducing the C sink, while in comparison, biomass consumed by grazing livestock was recycled back into the system through excretion and additionally, due to a greater capacity for plants to assimilate C during grazing from ungrazed paddock strips while silage cuts removed all available biomass to approximately 4 cm. To assess the impact of emissions of non-CO2 gases on the NGHGB, budget components and emissions of N2O and CH4 were converted to CO2 equivalents (CO2eq), by multiplying GHGs by their respective GWPs. Under the cut management, N2O emissions reduced the net ecosystem exchange (NEE) sink (-2010.8 g CO2 m-2 yr-1) by 7 % (-1870.7 g CO2eq m-2 yr-1), however emissions of N2O and CH4 reduced the NEE sink under a grazing management (-1355.3 g CO2 m-2 yr-1) by 20 and 58 % (-296.5 g CO2eq m-2 yr-1), respectively. Overall, the grassland remained a sink of CO2 with a NGHGB of -86.0 ± 90.1 and -84.4 ± 319.4 g CO2eq m-2 yr-1, under a cut and grazed management respectively. These findings show that at the field scale, both management practices greatly offset CO2 sinks from temperate grasslands due large C exports from biomass removal from the system, thus limiting the capacity of the system to photosynthesize and also through emissions of non-CO2 gases and their potent GWPs.
The key results from this thesis offer the following recommendations to the research community: (i) For quantifying field scale emissions of N2O using the static chamber technique, a minimum of five chamber replicates should be used, but where practically feasible up to 15 chamber replicates should be considered to further reduce the uncertainty in flux measurements and to improve the statistical robustness of N2O flux datasets; (ii) The frequency of static chamber flux measurements for quantifying baseline emissions should be at least once a week, but should increase to once a day for one to two weeks in order to capture the ecosystem response to additional N inputs from management over time i.e. the peak and decay pattern; (iii) Temporal upscaling of N2O emissions for single management events can be achieved by using statistical methods which explicitly account for the log-normal distribution of N2O emissions, for example, Bayesian statistics, however further development of the Bayesian approach is necessary for application to multiple management activities (i.e. fertilizer application and grazing) and interacting emission sources; (iv) To best quantify field scale emissions of N2O over space and time from managed grasslands, using EC and static chamber techniques in a complimentary fashion is strongly recommended as it enables a more informed quantification of field scale emissions in response to management activities relative to utilizing both techniques in isolation; (v) At the field scale, agricultural practices greatly offset the C sink of the grassland system in this study, namely through C exports from biomass removal and GHG emissions from livestock production. In order to reduce the impact of these systems in forcing climate change, policy makers will be required to incentivise farmers to transition to more sustainable agricultural practices with the aim of preventing large C losses through grazing and harvest cuts, and to increase C inputs through enhanced organic fertilization or increase soil organic carbon (SOC) stocks for example, through biochar additions below the 30 cm soil horizon where C decomposition rates are low and/or the establishment of multi-sward species with deep and extensive root systems to grassland systems.
Sponsor
Grant Number
Department of Agriculture, Food and the Marine, Manipulation and Integration of Nitrogen Emissions
15S655
Author's Homepage:
https://tcdlocalportal.tcd.ie/pls/EnterApex/f?p=800:71:0::::P71_USERNAME:MURPHR32Description:
APPROVED
Author: Murphy, Rachael
Advisor:
Saunders, MatthewPublisher:
Trinity College Dublin. School of Natural Sciences. Discipline of BotanyType of material:
ThesisAvailability:
Full text availableKeywords:
Grasslands, Eddy covariance, Greenhouse gases, Climate changeLicences: