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.

We investigated the effect of a flood on particulate organic matter (POM) dynamics in the floodplain and active channel of the Little Tennessee River in western North Carolina. We measured litterfall, leaf breakdown, and floodplain litter (before and after the flood) at 12 sites. Annual litterfall (256-562 g m 2 y 1) was typical of a temperate deciduous forest but lower than lowland floodplain forests in the eastern US. Leaf breakdown rates of 4 tree species (Acer rubrum, Carpinus caroliniana, Juglans nigra, and Platanus occidentalis) ranged from 0.001 to 0.010/d. We separated the 12 sites into 2 groups (inundated and non-inundated) based on the degree of flooding after a flood on 8 January 1998 and determined POM exchange between the active channel and floodplain. Significant transport of leaves from the floodplain to the river occurred at inundated sites, but transport of herbaceous or woody material did not occur. The flood increased leaf breakdown rates of all 4 tree species. Our measurements of litterfall, leaf breakdown, and floodplain litter standing crop suggest that leaves entrained from the floodplain of Little Tennessee River during floods are a source of POM to the active channel. However, flood input of leaves to the river were a small source of POM compared to direct leaffall.

No abstract

Summary1. Plant responses to nutrient heterogeneity may be influenced by spatial heterogeneity in other soil-based resources. We investigated how heterogeneity in two resources, nutrients and oxygen, affected root system morphology and biomass of three floodplain tree species, which differ in waterlogging tolerance from highly tolerant ( Nyssa aquatica L.) to intermediate ( Fraxinus pennsylvanica Marsh.) to less tolerant ( Liquidambar styraciflua L.). 2. We predicted that heterogeneity in soil oxygen would affect the well-known root proliferation response to nutrient-rich patches, and biomass, of the least tolerant species, L. styraciflua , but would not influence biomass or proliferation of the more tolerant species, F. pennsylvanica and N . aquatica . To test this prediction, we grew seedlings in pots where we manipulated the spatial distribution of oxygen (via waterlogging) and nutrients. 3. All species were precise foragers (i.e. they displayed preferential proliferation of roots in nutrient-rich patches) in uniformly drained and waterlogged conditions. When waterlogging and nutrients were both heterogeneous, roots of L. styraciflua avoided waterlogged patches and did not precisely forage for nutrients. In contrast, waterlogging had no negative effects on nutrient foraging of either F. pennsylvanica or N. aquatica ; however, these species did not precisely forage for nutrients when nutrient-rich and waterlogged patches were spatially separate. 4. Waterlogging reduced biomass of L. styraciflua , did not significantly affect biomass of F. pennsylvanica , and enhanced biomass of N. aquatica . In pots where waterlogging and nutrients were simultaneously heterogeneous, the spatial position of waterlogging relative to that of nutrients influenced the biomass response of F. pennsylvanica and N. aquatica to nutrient heterogeneity, but did not influence L. styraciflua . For F. pennsylvanica and N. aquatica , nutrients may have been easier to acquire in waterlogged patches due to faster diffusion rates. 5. Synthesis . Our results suggest that root system and biomass responses to nutrient heterogeneity can be affected by heterogeneity in another soil-based resource (i.e. soil oxygen availability). We believe ecologists should consider the effects of spatial heterogeneity in other soil-based resources when attempting to link nutrient heterogeneity to plant population or community attributes.

No abstract