Restoring connectivity: the effect of riparian replanting on in-stream organic carbon dynamics in a degraded agricultural landscape

Streams and rivers are intrinsically linked to the terrestrial environment by the exchange of water, nutrients, organic matter and biota. Terrestrial-aquatic connectivity has been disrupted by the degradation and removal of riparian vegetation due to widespread agricultural development. Loss of terrestrial vegetation modifies channel shading, in-stream habitat, and the quantity and composition of organic carbon (i.e. energy) subsidies provided to stream food webs. Collectively, these changes result in biodiversity loss and altered ecosystem functioning. Replanting riparian vegetation aims to alleviate the adverse ecological effects of riparian clearance. Replanting is already commonly used for ecological restoration, but revegetation may become more widespread if restoration activities are driven by economic forces, such as payments for planting trees to mitigate climate change. However, replantings currently are often spatially limited and isolated, so the plantings may not have a large effect on halting or reversing ecological degradation. Riparian clearance and revegetation are likely to alter carbon dynamics, which is a critical process underpinning biodiversity and ecosystem functioning in streams. We have little knowledge of if, and when, riparian replanting will restore in-stream organic-carbon processes towards pre-clearance conditions, or of the broad-scale effects on carbon balances. The integration of aquatic fluxes into regional carbon budgets is an important component of regional, national and global carbon accounting. I sought to: (1) quantify the reach-scale effects of replanting on in-stream organic carbon dynamics; (2) assess the potential for organic matter properties to reflect the success of restoring terrestrial-aquatic connectivity; and (3) upscale organic carbon fluxes to project the consequences of revegetation for carbon balance and atmospheric feedback at watershed (= catchment) scales. I assessed the dynamics of aquatic organic carbon (input, standing stock, export and metabolism) in 15 streams (1st-3rd Strahler order) of an agricultural landscape in south-eastern Australia. Ten of the streams had reaches that were replanted with native woody vegetation 8 to 22 years prior to the study; I refer to these restored reaches as ‘replanted’ and other reaches in which there had been no plantings, pasture reaches, as ‘untreated’. Replanted stream reaches had greater inputs and accumulation of terrestrial organic carbon on the stream-bed than did untreated reaches. Replanting was correlated with a reduction in net ecosystem productivity and a shorter organic carbon turnover length. Within two decades of planting, metabolic rates in replanted reaches had values more typical of those in natural, forested streams, supporting the use of ecosystem metabolism as a functional indictor of restoration success at reach-scales. Metabolic measures could be combined with pattern-based measures, such as biodiversity, to demonstrate the ecological value of replanting. The export of organic carbon was governed by land-use and climatic variables at spatial scales larger than typical replanting projects. Watershed tree cover affected the composition of dissolved organic matter, but not its quantity. A greater proportion of the dissolved organic matter in agricultural streams was from within the stream and a reduced proportion was from terrestrial vegetation, compared to streams in forested watersheds. The characteristics of dissolved organic matter potentially provide an aggregate measure of aquatic and terrestrial connectivity over large spatial scales. The quantity of total organic carbon transported was controlled by discharge. Projected increases in rainfall variability will affect the timing and magnitude of storm-flows, altering the fluxes of energy subsidies among ecosystems in landscapes. The estimated organic carbon budget showed that replanted reaches potentially were a greater source of carbon than were untreated reaches (net change -0.52 g C m-2 day-1 ± 0.80 SD). At a watershed scale, this increased carbon loss per unit area of stream was small compared to organic carbon export from 3rd-order streams. Riparian restoration at reach spatial scales (i.e. 100s of m) can restore ecosystem processes towards pre-clearance condition, within two decades. However, the effects of replanting at small scales may be overwhelmed by changes to hydrologic processes arising from probable increased climate variability in the future. Addressing land-use effects requires a landscape perspective that incorporates spatial context and connectivity at multiple scales to guide restoration activities into the areas likely to provide the greatest ecological return for investment