Tropical peatlands now emit hundreds of megatons of carbon dioxide per year because of human disruption of the feedbacks that link peat accumulation and groundwater hydrology. However, no quantitative theory has existed for how patterns of carbon storage and release accompanying growth and subsidence of tropical peatlands are affected by climate and disturbance. Using comprehensive data from a pristine peatland in Brunei Darussalam, we show how rainfall and groundwater flow determine a shape parameter (the Laplacian of the peat surface elevation) that specifies, under a given rainfall regime, the ultimate, stable morphology, and hence carbon storage, of a tropical peatland within a network of rivers or canals. We find that peatlands reach their ultimate shape first at the edges of peat domes where they are bounded by rivers, so that the rate of carbon uptake accompanying their growth is proportional to the area of the still-growing dome interior. We use this model to study how tropical peatland carbon storage and fluxes are controlled by changes in climate, sea level, and drainage networks. We find that fluctuations in net precipitation on timescales from hours to years can reduce long-term peat accumulation. Our mathematical and numerical models can be used to predict long-term effects of changes in temporal rainfall patterns and drainage networks on tropical peatland geomorphology and carbon storage.tropical peatlands | peatland geomorphology | peatland hydrology | peatland carbon storage | climate-carbon cycle feedbacks T ropical peatlands store gigatons of carbon in peat domes, gently mounded land forms kilometers across and 10 or more meters high (1). The carbon stored as peat in these domes has been sequestered by photosynthesis of peat swamp trees (2) and preserved for thousands of years by waterlogging, which suppresses decomposition. Human disturbance of tropical peatlands by fire and drainage for agriculture is now causing reemission of that carbon at rates of hundreds of megatons per year (2-5): Emissions from Southeast Asian peatlands alone are equivalent to about 2% of global fossil fuel emissions or 20% of global land use and land cover change emissions (6, 7). Because peat is mostly organic carbon, a description of the growth and subsidence of tropical peatlands also quantifies fluxes of carbon dioxide (1,4,8). Evidence from a range of studies establishes that accumulation and loss of tropical peat are controlled by water table dynamics (4, 9). When the water table is low, aerobic decomposition occurs, releasing carbon dioxide; when the water table is high, aerobic decomposition is inhibited by lack of oxygen, production of peat exceeds its decay, and peat accumulates. In this way, the rate of peat accumulation is determined by the fraction of time that peat is exposed by a low water table (Fig. 1).The water table rises and falls in a peatland according to the balance between rainfall, evapotranspiration, and groundwater flow. Water flows downslope toward the edge of each peat dome, where it is bo...

Emission of CO2 from tropical peatlands is an important component of the global carbon budget. Over days to months, these fluxes are largely controlled by water table depth. However, the diurnal cycle is less well understood, in part, because most measurements have been collected daily at midday. We used an automated chamber system to make hourly measurements of peat surface CO2 emissions from chambers root‐cut to 30 cm. We then used these data to disentangle the relationship between temperature, water table and heterotrophic respiration (Rhet). We made two central observations. First, we found strong diurnal cycles in CO2 flux and near‐surface peat temperature (<10 cm depth), both peaking at midday. The magnitude of diurnal oscillations was strongly influenced by shading and water table depth, highlighting the limitations of relying on daytime measurements and/or a single correction factor to remove daytime bias in flux measurements. Second, we found mean daily Rhet had a strong linear relationship to the depth of the water table, and under flooded conditions, Rhet was small and constant. We used this relationship between Rhet and water table depth to estimate carbon export from both Rhet and dissolved organic carbon over the course of a year based on water table records. Rhet dominates annual carbon export, demonstrating the potential for peatland drainage to increase regional CO2 emissions. Finally, we discuss an apparent incompatibility between hourly and daily average observations of CO2 flux, water table and temperature: water table and daily average flux data suggest that CO2 is produced across the entire unsaturated peat profile, whereas temperature and hourly flux data appear to suggest that CO2 fluxes are controlled by very near surface peat. We explore how temperature‐, moisture‐ and gas transport‐related mechanisms could cause mean CO2 emissions to increase linearly with water table depth and also have a large diurnal cycle.

No abstract