Streams integrate biogeochemical processes operating at broad to local spatial scales and long term to short term time scales. Humans have extensively altered those processes in North America, with serious consequences for aquatic ecosystems. We collected data on Upper Tennessee River tributaries in North Carolina to: (1) compare landuse and landscape geomorphology with respect to their ability to explain variation in water quality, sedimentation measures, and large woody debris; (2) determine if landscape change over time contributed significantly to explaining present stream conditions; and (3) assess the importance of spatial scale in examining landuse influences on streams. Stream variables were related to both landuse and landscape geomorphology. Forest cover accounted for the most variation in nearly all models, supporting predictions of nutrient enrichment, thermal pollution, and sedimentation caused by landscape disturbance. Legacy effects from past catchment disturbance were apparent in sedimentation measures. Nitrogen and phosphorus concentrations, as well as stream temperature, were lower where riparian buffers had reforested. Models of stream physicochemistry fit better when predictors were catchment wide rather than more localized (i.e., within 2 km of a site). Cumulative impacts to streams due to changes in landuse must be managed from a watershed perspective with quantitative models that integrate across scales.

No abstract

Abstract. We investigated longitudinal patterns of ecosystem metabolism (primary production and respiration) at 4 sites along a 37-km segment of the Little Tennessee River (LTR), North Carolina. These sites corresponded to 4th-to 6th-order reaches in the LTR in an attempt to identify the transition from heterotrophic to autotrophic conditions in this river ecosystem. In addition, we compared autochthonous C production to supply of coarse organic material from direct litter fall and entrainment from the floodplain during floods to determine the contributions of each to river energetics on an annual basis. Metabolism was measured at several times of year at each site using the singlestation diel oxygen change method and reaeration estimated by the energy dissipation method. Gross primary production (GPP) ranged from 0.07 to 1.92 g C m Ϫ2 d Ϫ1 and increased ϳ3-fold from upstream to downstream. Respiration (R) ranged from 0.27 to 2.32 g C m Ϫ2 d Ϫ1 but did not change along the river continuum. Net ecosystem production (NEP) and P/R consistently showed that metabolism was heterotrophic in upstream sites and became autotrophic in the site furthest downstream. Calculated transitional P/R (i.e., where heterotrophic respiration is supported equally by autochthonous and allochthonous C sources) suggested that this heterotrophy-autotrophy shift occurred further upstream than where P/R ϭ 1. Annual rates of GPP were 3 times higher than litter fall and floodplain inputs of C, but R was higher than total C input suggesting that unmeasured C sources must be important for C dynamics in the LTR. The difference between measured C inputs and R decreased along the river continuum because of a 3-fold increase in GPP with little change in allochthonous input and R. Our results suggest that the LTR changes from heterotrophic to autotrophic along this stretch of river and that autochthonous C sources become more important for respiration and secondary production at downstream sites.

No abstract