Using automated transparent chambers to quantify CO2 emissions and potential emission reduction by water infiltration systems in drained coastal peatlands in the Netherlands

Abstract. Worldwide, drainage of peatlands has turned these systems from CO2 sinks into sources. In the Netherlands, where ~7 % of the land surface consists of peatlands, drained peat soils contribute >90 % and ~3 % to the country’s soil-derived and total CO2 emission, respectively. Hence, the Dutch Climate Agreement set targets to cut these emissions. One potential mitigation measure is the application of subsurface water infiltration systems (WIS) consisting of subsurface pipes connected to ditch water. WIS aims to raise the water table depth (WTD) in dry periods to limit peat oxidation while maintaining current land-use practices. Here, we used automated transparent chambers in 12 peat pasture plots across the Netherlands to measure CO2 fluxes at high frequency and assess 1) the relationship between WTD and CO2 emissions for Dutch peatlands and 2) the effectiveness of WIS to mitigate emissions. Net ecosystem carbon balances (NECB) (up to four years per site, 2020–2023) averaged 3.60 and 2.69 t CO2-C ha-1 yr-1 for control and WIS sites, respectively. The magnitude of NECBs and slope of the WTD-NECB relationship fall within the range of observations of earlier studies in Europe, though they were notably lower than those based on campaign-wise, closed chamber measurements. The relationship between annual exposed carbon (defined as total amount of carbon within the soil above the average annual WTD) and NECB explained more variance than the WTD-NECB relationship. We found strong evidence for a reducing effect of WIS on CO2 emissions and no evidence for an effect of WIS on the WTD-NECB and annual exposed carbon-NECB relationships, meaning that relationships between either WTD or exposed carbon and NECB can be used to estimate the emission reduction for a given WIS-induced increase in WTD or exposed carbon. High year-to-year variation in NECBs calls for multi-year measurements and sufficient representative measurement years per site as demonstrated in this study with 35 site-years observations.