Skiba, Ute

Carvalho, Laurence

Heal, Kate

Rees, Bob

Harley, James F.

CEH Integrating fund

2013-07-08T10:03:30Z

2013-07-08T10:03:30Z

2013-07-01

River networks act as a link between components of the terrestrial landscape, such as soils and groundwater, with the atmosphere and oceans, and are now believed to contribute significantly to global budgets of carbon dioxide (CO₂), methane (CH₄) and nitrous oxide (N₂O). The idea of rivers being an inert conduit for carbon and nitrogen to reach the coast has been challenged recently, with considerable processing of carbon and nitrogen occurring in both the water column and bed sediments in the various aquatic components that make up a river network, including lakes, streams, rivers and estuaries. Although understanding of the cycling of carbon and nitrogen has improved markedly in the last 20 years, there is still much uncertainty regarding the production and emission of greenhouse gases (GHGs) linked to this processing across river catchments and few studies have quantified GHG fluxes from source to sea. Therefore this study aimed to a) understand the spatial and temporal saturations and fluxes of GHGs from both the freshwater River Tay catchment (Scotland) and the river Tay estuary, and b) understand what controls the production of GHGs within both a freshwater lake and across multiple sites in the freshwater river using laboratory incubations of sediment. Hotspots of in-stream production and emission were evident both in the freshwater catchment and the estuary, with significant temporal and spatial variability in saturation and emission (density) for CH₄, CO₂ and N₂O. CH₄ emission densities, across the freshwater river sites, ranged from 1720 to 15500 μg C m⁻² d⁻¹ with a freshwater catchment wide mean of 4640 μg C m⁻² d⁻¹, and in general decreased from upland to lowland sites along the main river stem, with notable peaks of emission in a lowland tributary and at the outflow of a lowland loch. This corresponds well with the main drivers of spatial variability which include allochthonous inputs from gas rich soil waters and in-situ production in fine grained organic rich sediments. CH₄ production was observed to be higher in the lowland tributaries (R. Isla 4500 μg C m⁻² d⁻¹) compared to main-stem river sites both in the lowland river (129 μg C m⁻² d⁻¹) and upland river which displayed an uptake of CH₄ (-1210 μg C m⁻² d⁻¹). The main driver of spatial variability in CH₄ production rates was the quality of the sediment, as production was higher in fine grained sediments rich in carbon compared to sand and gravels with a low carbon content. CH₄ production also varied seasonally, with temperature and seasonal variation in sediment quality as the predominant driving factors. CO₂ emission densities across the freshwater catchment ranged from 517 to 2550 mg C m⁻² d⁻¹ with a catchment mean flux density of 1500 mg C m⁻² d⁻¹. Flux densities on the whole increased along the main river stem from upland sites to lowland sites, with higher fluxes in lowland tributaries. Seasonally, CO₂ flux density was highest in late summer and autumn and lowest in winter at most sites, highlighting the importance in seasonal environmental controls such as temperature, light, and substrate availability. Production rates in the sediment increased from upland to lowland sites with highest production rates evident in the lowland tributaries, and in autumn sediment samples. N₂O emission density also showed considerable spatial and seasonal variation across the catchment with flux densities ranging from 176 to 1850 μg N m⁻² d⁻¹ with a mean flux of 780 μg N m⁻² d⁻¹. Mean fluxes were highest in the lowland tributaries and lowest in the upland river with sediment experiments finding similar spatial variation in N₂O production. On the whole, in-stream N₂O production and emission across the freshwater catchment was driven by increases in nutrient concentration (NO₃⁻, NH₄⁺) which in turn was related to the proportion of agricultural landuse. The saturation and emission of GHGs also varied substantially both spatially and temporally in the river Tay estuary, with a mean emission density of 2790 μg CH₄-C m⁻² d⁻¹, 990 mg CO₂-C m⁻² d⁻¹ and 162 μg N₂O-N m⁻² d⁻¹. The spatial variability of GHG concentrations and emission densities in the estuary were predominantly controlled by the balance between lateral inputs (from tidal flushing of surrounding intertidal areas), in-situ microbial production/consumption (both in the water column and bed sediments) and physical mixing/loss processes. Although emission densities of CH₄, CO₂ and N₂O appear low compared to the freshwater river, this is because the estuary is emitting large quantities of gas in the middle and outer estuary, for example net annual emission of N₂O increased from 84.7 kg N₂O-N yr⁻¹ in the upper freshwater section of the estuary to 888 kg N₂O-N yr⁻¹ in the middle estuary section, then decreased to 309 kg N₂O-N yr⁻¹ in the saltwater lower estuary. Overall, this study has shown that both dissolved and aerial fluxes of GHGs vary markedly both spatially and temporal from source to sea in a temperate river catchment, with hotspots of in-stream production and emission across the river catchment. The catchment (river, lake and estuary) was a smaller source of CO₂, CH₄ and N₂O emission (total emission and by area) compared to other highly polluted aquatic systems both in the UK and globally.

http://hdl.handle.net/1842/7527

en

The University of Edinburgh

carbon

nitrogen

greenhouse gas

river catchment

From source to sea: spatial and temporal fluxes of the greenhouse gases N₂O, CO₂ and CH₄ in the River Tay catchment

Thesis or Dissertation

Doctoral

PhD Doctor of Philosophy