Thesis Abstract
Nitrous oxide (N2O) is a powerful greenhouse gas and ozone depleter, and it is produced in large quantities when a dry soil is rewetted in a phenomenon known as a hot moment. Therefore understanding this phenomena is important for tackling agriculture’s impact on climate change. A literature review and meta-analysis of N2O hot moments was conducted in Chapter 2, revealing that the amount of water added to the soil and how saturated the soil gets during rewetting are the most important controlling factors. For example, rewetting from 50 to 90% water filled pore space (WFPS), would produce more N2O emissions, than rewetting the same soil from 50% to 70%. However, it was clear that the current literature suffered from the lack of a standardised approach, as it was difficult to draw conclusions from experiments with different designs that had different rewetting strategies, controlled temperatures and soil core sizes. Moreover, drought length (i.e. the amount of time the soil stays dry before it was rewetted) had not been adequately investigated

It was still an open question as to how and why the soil’s microbial communities were responding in this manner to soil drought and rewetting, and so Chapter 3 was a lab experiment that aimed to investigate two of the key hypotheses suggested by the current literature. Firstly that drought and the resulting osmotic stress created a selection pressure which favoured the soil microbes that could rapidly denitrify upon rewetting. Secondly, that more carbon (C) and nitrogen (N) would become available from lysed microbes that could not survive the osmotic change. Neither of these were confirmed, as changes in the soil C and N and functional gene abundance could not explain differences in the sizes of the induced hot moments.

Chapter 4 repeated this experiment, with the same aims, and an additional measurement of messenger RNA abundance of key N cycling genes. While functional gene and transcript abundance failed to provide an explanation, ammonium (NH4+) in the driest treatment increased and then reduced in correlation with the largest hot moment. It was concluded that while the soil C and N pools can be enhanced by necromass, this is not the cause of the hot moment. This chapter also included a second experiment to investigate the effect of drought length, which was observed to have an inverted U shaped response, with the least amount of N2O being released when the drought phase was < 6 days or > 24 days. Overall, it was concluded that there are two stages of microbial quiescence that explain N2O hot moments and its response to changing antecedent conditions. The first stage is a state of semi-quiescence as the soil’s microbial communities prepare for the moisture conditions to change, during this stage they can rapidly respond to the changing WFPS by catabolising intercellular osmotic compounds, and the lower the WFPS the more of the soil microbial community are in this state, and therefore the greater the response once the soil is rewetted. However, once the drought becomes too extreme they enter a second stage of quiescence to survive the harsher conditions, which no longer allows a rapid response, explaining the quadratic effect of drought length.

This PhD also conducted a field trial on a long-term experiment based at Rothamsted’s Harpenden site, called Broadbalk, which is split into plots that have had different fertiliser types and quantities. It aimed to investigate how the legacy of fertiliser management, has impacted Broadbalk’s soil and its production of greenhouse gas (GHG) emissions (CO2, N2O, and CH4). This chapter compared differences between the farmyard manure (FYM), inorganic nitrogen and control plots (where only phosphorus and potassium was applied). The FYM treatment had more dissolved organic carbon (+160%) and triple the %C compared to the other treatments, and it had a significantly larger mean CO2 flux. The largest N2O emissions were recorded from the inorganic treatment post fertiliser application, but this did not result in significantly higher mean N2O emissions compared to the FYM treatment, which is likely due to the enhanced microbiological and biochemical activity from continuous organic fertiliser addition. CH4 fluxes were not affected by the different fertiliser regimes in this study, probably because the variables that typically affect CH4 emissions were not impacted by the selected treatments. The control treatment had the lowest GHG emissions, due to the lack of substrates in the soil.

In summary, this PhD quantified how the soil’s antecedent moisture conditions in a controlled environment effect the size of a hot moment, as well as providing new data on the effects of drought length, which due to its quadratic nature will change how researchers interpret the impact of more extreme droughts on N2O fluxes. It also conducted the first field trial on Broadbalk that measured all three of the key GHG greenhouse gases from soil. Concluding that the retention of soil C in the plots is the most important factor affecting CO2e, with CO2 making up the dominant portion of CO2e from all treatments.