Effects of carbon substrate and irrigation on carbon dioxide emissions and denitrification for three grassland soils : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

Abstract

Conversion of non irrigated grassland to high−intensity farm systems with irrigation is a major land−use change to enable a higher yield year−round in New Zealand. Research is needed to better understand the links between soil carbon (C), water and nitrogen (N) and their dynamics in order to minimise the impacts of farm management practices on losses of soil C and N. The aim of the research was to investigate carbon dioxide (CO2) and nitrous oxide (N2O) emissions in relation to physical factors, and/or C and N substrate dynamics under irrigated grassland. Ultimately, the research will contribute to developing optimal irrigation management practices that promote the retention of soil organic matter (SOM) while minimising C and N losses. The key hypothesis of this study is that irrigation induced changes to soil water content will promote plant growth and thus root exudation, while also affecting relative gas diffusivity, CO2 emissions and N2O formation mechanisms. The research was based on experimental laboratory work oriented towards understanding soil C−N interactions in order to mitigate CO2 and N2O emissions in grazed grasslands. The aim of the first experiment was to determine the impacts of C substrates on soil CO2 and N2O emissions, under varying soil types and soil water contents. Three repacked Pallic grassland soils containing NO3--15N were held at three levels of matric potential (, −3, −5 and −7 kPa), while receiving daily substrate additions (acetate, glucose, water control) for 14 days. The daily CO2 and N2O emissions were monitored. Additionally, the N2O:(N2+N2O) ratios were determined using 15N methods on days 3 and 14. Results showed that across all soils, N2O peak emissions were higher for soils treated with glucose, with a range (± SD) of 0.1 ± 0.0 to 42.7 ± 2.1 mg N m−2 h−1. The highest cumulative N2O emission (2.5 ± 0.2 g N m−2) was measured in glucose-treated soil at a  of −3 kPa. In comparison with added glucose, acetate resulted in 2-fold higher N2 emissions in soils at low diffusivities. The N2O:(N2O+N2) emissions ratios varied with soil type (0.91-0.80) on day 3. Cumulative CO2 emissions increased with increasing soil diffusivity and soils amended with glucose had higher cumulative CO2-C emissions, ranging from 22.5 ± 1.3 to 36.6 ± 1.8, g C m−2. Collectively, I demonstrated that the increase of N2O, N2 and CO2 emissions in response to acetate or glucose addition depended on both soil type and soil matric potential. The findings indicate that non-fermentable substrates will enhance denitrification from soil. Using a similar setup and treatments, the aim of the second experiment was to determine the relationships between the priming effect and N2O emissions from the soil, in relation to N and C supply. I applied 13C−labelled substrates (acetate, butyrate, glucose; 80 μg C g−1, 6 atom% excess 13C), with water as a control, and 15N−labelled N as KNO3 (300 μg N g−1 soil, 40 atom% excess 15N) to three different soils and, after 3 days, measured the effects on the priming of SOM and sources of N2O emissions. I demonstrated that C substrate addition increased both CO2 and SOM derived N2O emissions in the presence of exogenous N. Emissions of CO2 and N2O from soils with added glucose (0.73 ± 0.13 μmol m−2 s−1 and 21.4 ± 12.1 mg N m−2 h−1) were higher than those from soils treated with acetate (0.64 ± 0.11 μmol m−2 s−1 and 10.9 ± 6.5 mg N m−2 h−1) or butyrate (0.61 ± 0.11 μmol m−2 s−1 and 11.0 ± 6.6 mg N m−2 h−1), respectively. Acetate addition induced a stronger priming effect (0.07 ± 0.09 μmol m−2 s−1) than that for glucose (0.02 ± 0.10 μmol m−2 s−1), while butyrate addition resulted in negative priming (-0.09 ± 0.05 μmol m−2 s−1). SOM derived N2O emissions were relatively low from soils with butyrate addition (1.4 ± 1.5 mg N m−2 h−1) compared with acetate (2.9 ± 2.3 mg N m−2 h−1) or glucose (9.2 ± 4.5 mg N m−2 h−1). However, I did not detect a clear relationship between priming effect and SOM derived N2O emissions. The findings highlight the need to consider the nature of the C substrate when interpreting processes regulating SOM decomposition and N2O emission source. In the third experiment, the components of net ecosystem C balance (FN) were partitioned for a C4 plant Bermuda grass (Cynodon dactylon L.), growing in mesocosms and irrigated with the same total quantity of water (15 mm day−1) applied at intervals of 1, 2, 3 days for 12 days (treatments I1, I2, and I3, respectively), whereafter treatment I2 was changed to watering every 6 days (treatment I6) and treatments I2 and I3 were continued for a further 18 days. Daily measurements of evaporation were made by weighing the mesocosms and chambers were used to measure rates of CO2 exchange to estimate FN, ecosystem respiration and respiration from leaves and soil plus roots (RS), and gross C uptake by the plants. Further, use of the C4 plant enabled partitioning of RS into the autotrophic and heterotrophic components of belowground respiration using a 13C natural abundance isotopic technique, requiring destructive sampling at the end of the experiment when differences in cumulative soil water deficit between the treatments were greatest. The findings showed that, over short periods with well−drained soil, irrigation frequency could be managed to manipulate soil water deficits to reduce net belowground respiratory C losses, particularly those from the microbial decomposition of SOM, with no significant effects on biomass production and N2O emissions. Thus, changes to the scheduling of irrigation could reduce CO2 emissions and SOM decomposition but not N2O emissions in conditions of moderate to high water deficits. This study showed that CO2 and N2O emissions are dependent on C availability and diffusivity that accounts for differences in both soil water content and soils, but there is no clear relationship between priming of SOM and N2O emissions. Changes to the scheduling of irrigation could be used to minimise soil C losses but that this is unlikely to affect N2O emissions with low N inputs on well-drained soils.