Spatial and temporal variation of dissolved organic carbon export from gauged and ungauged watersheds of Dee Valley, Scotland: Effect of land cover and C:N

Abstract

Dissolved organic carbon (DOC) concentration and export were quantified in 56 streams within the Dee Valley, Scotland. Percentage land cover was calculated for each of the watersheds. For a subset of 21 watersheds, organic and mineral soil layers were collected for carbon and nitrogen analysis. Peat cover derived from land cover data was a good predictor of annual DOC export for the selected 21 upland watersheds but not for the set of 56 watersheds. Using DOC values and mean watershed C:N ratio, we produced an empirical predictive model for estimation of annual DOC export for the 21 watersheds as well as a model of weekly fluxes in two of the larger watersheds to meet the needs of watershed managers. Annual exports for the 56 streams in the Dee Valley ranged from 1700 to 10,500 kg km−2 yr−1. Land cover was not related to annual DOC export for the 56 streams, but by season, arable land and grassland were both correlated to spring DOC export (R = −0.39 arable land and R = −0.33 grassland p<0.01). Summertime DOC export was correlated to heather, montane, and conifer land cover (R = −0.029 heather, R = −0.39 montane, and R = 0.33 conifer, p<0.05). Autumn and winter DOC export was best described by peat (R = 0.29 and 0.39, respectively, p<0.05). In the subset of 21 watersheds with soil information, watershed soil organic matter C:N ratio and peat had a higher correlation with annual DOC concentrations and export than any other land cover class. Source of peat distribution data may impact the relationship between DOC and peat cover, and this issue is examined. Modeled weekly DOC load compared well with observed DOC load in a time series evaluation.