Impacts of land use and climate change on carbon and greenhouse gas dynamics of Irish peatland ecosystems

Abstract

Peatlands are wetlands that are found on every continent except Antarctica, covering around 3% of the global land area and are the largest terrestrial soil organic carbon store. They are estimated to store around 550 gigatons of carbon (C) which is equivalent to twice the amount stored in all of the world's forests. In Ireland, peatlands cover almost one-fifth of the total land area and are recognized as a major C sink that has been depleted but still represents a key climate mitigation tool. Unfortunately, many of Ireland's peatlands have been drained and degraded for peat extraction, or converted to forestry and agriculture. The Irish government has taken steps to protect and restore these peatlands as part of the national climate mitigation strategy. However, it is vital to understand the impacts of drainage along with changing climate on carbon and methane dynamics of these ecosystems and the key drivers of the emissions for effective management. Additionally, long-term monitoring can aid in assessing vulnerability and resilience of these ecosystems to climate change. The primary aim of this PhD research was to develop an interdisciplinary approach to investigate the impact of anthropogenic activities and climate change on the carbon and methane dynamics of a raised bog ecosystem across a drainage gradient by combining innovative experimental and observational techniques. Chapter 4 outlines the carbon dynamics and the key drivers influencing the uptake and emissions at a near-natural raised bog in Ireland using the Eddy Covariance technique. The measurement period included two drought years with different hydrological conditions (2018 and 2021) followed by humid/wetter normal climatic years (2019 and 2020) which provided insight into the response of such ecosystems to climate change. The reduction in soil moisture below 80% in 2018 was caused by a combination of a lowering water table for consecutive 109 days and higher temperatures compared to the other very dry year of 2021. The differing hydrological and climatic conditions between years resulted in the study area acting as a net source of carbon of 53.5 g C m-2 in 2018 and a net sink of -125.2 g C m-2, -75.7 g C m-2, and -49.8 g C m-2 in 2019, 2020 and 2021 respectively. The carbon dynamics in 2018 were primarily driven by ecosystem respiration however, the system went back to being a carbon sink in the following years displaying its resilience to drought events. Overall, this study highlights both the vulnerability and resilience of these ecosystems to climatic variability and emphasizes the need for long-term monitoring networks to enhance our understanding of the impacts of these events. This PhD research also focuses on using closed chamber measurement techniques (chapter 5) to understand methane dynamics at raised bogs under various management practices which included near-natural bog, under rehabilitation site, and degraded peatland. The average annual methane flux was ranging between 1.44 to 8.89 g C m-2 y-1 at a near-natural bog, -1.1 to 1.55 g C m-2 y-1 at an under-rehabilitation site, and 0 to 0.53 g C m-2 y-1 at the degraded site. Furthermore, this study proposes a framework to upscale these emissions from the chamber to ecosystem scale by testing the applicability of high-resolution multispectral satellite imagery with machine learning algorithms to provide fine-scale vegetation maps. These maps were used to upscale chamber-based methane fluxes at peatland sites across a drainage/management gradient. Our results provide evidence that upscaling is successful, as long as the dominant vegetation communities are appropriately mapped. Results from this study highlight the complex spatial heterogeneity of peatlands and their impact on methane fluxes. This study emphasis the need to include dominant plant species information in methane budget models and upscaling approaches. The framework proposed in this study was able to map all the ecotopes at the Lullymore (under rehabilitation) and Garryduff (degraded) sites with an overall validation accuracy above 95% in 2020 and 2021. However, Clara bog (near-natural) is an extremely challenging and complex site to map due to the heterogeneity of vegetation at the site, however, this framework was able to map all of the ecotopes present with an overall validation accuracy above 83%. This approach could be used to upscale fluxes at other sites and aid in refining site-specific emission factors for these systems. Chapter 6 highlights the potential of using remote sensing models to estimate Gross Primary Productivity (GPP) at sites where it is not feasible to install EC towers or conduct chamber measurements. GPP represents one of the key components of the peatland carbon cycle, and detailed knowledge of the spatial and temporal extent of GPP under changing management practices and climate variability is imperative to improve our predictions of peatland ecology and biogeochemistry. This research assessed the relationship between remote sensing and ground-based estimates of GPP for a near-natural peatland in Ireland using Eddy Covariance (EC) techniques and high-resolution Sentinel 2A satellite imagery. Hybrid models were developed using multiple linear regression along with six widely used conventional indices and light use efficiency model. Estimates of GPP using NDVI, EVI, and NDWI2 based hybrid models 2 and 4 performed well with a significant correlation over 0.89 between modelled GPP and EC-derived GPP. This study also reports site-specific light use efficiency parameters for both dry and hydrologically normal years based on the light response curves method (LRC). Overall, this research demonstrated the potential of combining EC techniques with satellite-derived models to better understand and monitor key drivers and patterns of GPP for raised bog ecosystems under different climate scenarios and has also provided light use efficiency parameters values for variable climatic conditions which can be used to estimate GPP using LUE models across various scales. Finally, this thesis emphases the importance of long-term measurements and monitoring at peatland sites across drainage gradients to better understand the impact of land use change and climate variability on carbon and methane dynamics. This information will enhance conservation, rehabilitation, and restoration activities and strategies and, will help to increase the carbon storage potential of these ecosystems. Additionally, this research shows the successful integration of innovative remote sensing and machine learning techniques with field measurements which could aid in refining tier 2 and tier 3 level emission factors for peatland systems in Ireland.