Abstract

In this study, we estimated whether changes in hydrological pathwaysduring storms could explain the large temporal variations of dissolvedorganic carbon (DOC) and nitrogen (DON) in the runoff of threecatchments: a forest and a grassland sub-catchment of 1600m2 delineated by trenches, and a headwater catchment of 0.7km2. The average annual DOC export from the sub-catchments was 185 kg DOCha−1 y−1 for the forest, 108 kg DOCha−1 y−1 for the grassland and 84 kgDOC ha−1 y−1 for the headwatercatchment. DON was the major form of the dissolved N in soil and streamwater. DON export from all catchments was approximately 6 kg Nha−1 y−1, which corresponded to 60% ofthe total N export and to 50% of the ambient wet N deposition. DOC andDON concentrations in weekly samples of stream water were positivelycorrelated with discharge. During individual storms, concentrations andproperties of DOC and DON changed drastically. In all catchments, DOCconcentrations increased by 6 to 7 mg DOC l−1 comparedto base flow, with the largest relative increment in the headwatercatchment (+350%). Concentrations of DON, hydrolysable amino acids, andphenolics showed comparable increases, whereas the proportion ofcarbohydrates in DOC decreased at peak flow. Prediction of DOC and DONconcentrations by an end-member mixing analysis (EMMA) on the base ofinorganic water chemistry showed that changes in water flow pathslargely explained these temporal variability. According to the EMMA, thecontribution of throughfall to the runoff peaked in the initial phase ofthe storm, while water from the subsoil dominated during base flow only.EMMA indicated that the contribution of the DOC and DON-rich topsoil washighest in the later stages of the storm, which explained the highestDOC and DON concentrations as the hydrograph receded. Discrepanciesbetween observed and predicted concentrations were largest for thereactive DOC compounds such as carbohydrates and phenolics. Theyoccurred at base flow and in the initial phase of storms. This suggeststhat other mechanisms such as in-stream processes or a time-variantrelease of DOC also played an important role.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call