LONGITUDINAL STREAM SYNOPTIC (LSS) MONITORING TO EVALUATE WATER QUALITY IN RESTORED STREAMS

Abstract

Many kilometers of streams are being restored in the Chesapeake Bay watershed and elsewhere in efforts to stabilize streambanks, protect infrastructure, and improve water quality. Urban development and impervious surface cover increase peak flows, which degrade streams. Restoration strategies often employ engineering approaches to enhance stream-floodplain reconnection, dissipate erosive forces from urban runoff, and enhance contaminant retention. In this study, longitudinal stream synoptic (LSS) monitoring (sampling multiple points along flowpaths across both space and time) was conducted to assess the effectiveness of different forms of stream restoration in attenuating pollutants downstream. Spatial and temporal monitoring of carbon, nutrients, salt ions, and metals were conducted across five watersheds experiencing varying levels of stream-floodplain reconnection and stormwater management within the Chesapeake Bay region. Study sites included Sligo Creek (minimal floodplain reconnection), Paint Branch (streambank stabilization without significant reconnection), Scotts Level Branch (engineered stream-floodplain reconnection), Little Paint Branch (natural floodplain reconnection from sedimentation), and Campus Creek (regenerative stormwater conveyance with engineered floodplain reconnection). We investigated: (1) whether changes in water chemistry can be detected along longitudinal flowpaths in response to stream-floodplain reconnection, and (2) which monitoring scales across space and time can provide useful information regarding the effectiveness of restoration. Results from this work suggest that longitudinal synoptic monitoring can track the fate and transport of multiple contaminants and evaluate restoration strategies across high spatial-resolution scales. Along all five watersheds, stream water chemistry varied substantially across finer spatial scales (sometimes within hundreds of meters) in response to changes in landscapes, restoration features, or local hydrology. There were significant declining concentrations (p<0.05) or stable concentrations of nutrients, salts, and metals as streams flowed through restoration features. There were significant increasing trends in chemical concentrations (e.g. Na+, Ca2+, K+) in unrestored stream reaches with increasing impervious surface cover. Principal component analysis (PCA) also indicated that there were changes in the chemical compositions of mixtures of salts, metals, and nutrients in response to restoration projects, storm events, and seasons. Interestingly, dissolved Fe and Mn concentrations showed significant increasing trends along some stream reaches with hydrologically connected floodplains. Fe and Mn also showed significant decreasing trends along some unrestored stream reaches surrounded by increasing impervious surfaces. Increased concentrations of dissolved Fe and Mn may have been an indicator of increased hydrologic connectivity between groundwater and surface water and decreased redox potentials. Overall, longitudinal water quality changes over meters and kilometers can be useful in detecting effects of stream restoration on water quality at the watershed scale. Results suggest that water quality in urban streams can change locally in response to restoration projects for multiple chemicals, but the incremental changes associated with different forms of stream restoration and riparian conservation can also be overwhelmed across broader watershed spatial scales and during storm events.