Search for: All records

Milldams and their legacies have significantly influenced fluvial processes and geomorphology. However, less is known about their effects on riparian zone hydrology, biogeochemistry, and water quality. Here, we discuss the potential effects of existing and breached milldams on riparian nitrogen (N) processing through multiple competing hypotheses and observations from complementary studies. Competing hypotheses characterize riparian zone processes that remove (sink) or release (source) N. Elevated groundwater levels and reducing soil conditions upstream of milldams suggest that riparian zones above dams could be hotspots for N removal via denitrification and plant N uptake. On the other hand, dam removals and subsequent drops in stream and riparian groundwater levels result in drained, oxic soils which could increase soil nitrification and decrease riparian plant uptake due to groundwater bypassing the root zone. Whether dam removals would result in a net increase or decrease of N in riparian groundwaters is unknown and needs to be investigated. While nitrification, denitrification, and plant N uptake have typically received the most attention in riparian studies, other N cycle processes such as dissimilatory nitrate reduction to ammonium (DNRA) need to be considered. We also propose a novel concept of riparian discontinuum, which highlights the hydrologic and biogeochemical discontinuities introduced in riparian zones by anthropogenic structures such as milldams. Understanding and quantifying how milldams and similar structures influence the net source or sink behavior of riparian zones is urgently needed for guiding watershed management practices and for informed decision making with regard to dam removals.

 

The compounding effects of anthropogenic legacies for environmental pollution are significant, but not well understood. Here, we show that centennial‐scale legacies of milldams and decadal‐scale legacies of road salt salinization interact in unexpected ways to produce hot spots of nitrogen (N) in riparian zones. Riparian groundwater and stream water concentrations upstream of two mid‐Atlantic (Pennsylvania and Delaware) milldams, 2.4 and 4 m tall, were sampled over a 2 year period. Clay and silt‐rich legacy sediments with low hydraulic conductivity, stagnant and poorly mixed hydrologic conditions, and persistent hypoxia in riparian sediments upstream of milldams produced a unique biogeochemical gradient with nitrate removal via denitrification at the upland riparian edge and ammonium‐N accumulation in near‐stream sediments and groundwaters. Riparian groundwater ammonium‐N concentrations upstream of the milldams ranged from 0.006 to 30.6 mgN L⁻¹while soil‐bound values were 0.11–456 mg kg⁻¹. We attribute the elevated ammonium concentrations to ammonification with suppression of nitrification and/or dissimilatory nitrate reduction to ammonium (DNRA). Sodium inputs to riparian groundwater (25–1,504 mg L⁻¹) from road salts may further enhance DNRA and ammonium production and displace sorbed soil ammonium‐N into groundwaters. This study suggests that legacies of milldams and road salts may undercut the N buffering capacity of riparian zones and need to be considered in riparian buffer assessments, watershed management plans, and dam removal decisions. Given the widespread existence of dams and other barriers and the ubiquitous use of road salt, the potential for this synergistic N pollution is significant.

 

Large runoff, sediment, and nutrient exports from watersheds could occur due to individual extreme climate events or a combination of multiple hydrologic and meteorological conditions. Using high-frequency hydrologic, sediment, and turbidity data we show that freeze–thaw episodes followed by intense winter (February) rainstorms can export very high concentrations and loads of suspended sediment and particulate organic carbon (POC) and nitrogen (PN) from mid-Atlantic watersheds in the US. Peak suspended sediment (> 5000 mg L−1), POC (> 250 mg L−1) and PN (> 15 mg L−1) concentrations at our 12 and 79 ha forested watersheds for the February rainfall-runoff events were highest on record and the fluxes were comparable to those measured for tropical storms. Similar responses were observed for turbidity values (> 400 FNU) at larger USGS-monitored watersheds. Much of the sediments and particulate nutrients likely originated from erosion of stream bank sediments and/or channel storage. Currently, there is considerable uncertainty about the contribution of these sources to nonpoint source pollution, particularly, in watersheds with large legacy sediment deposits. Future climate projections indicate increased intensification of storm events and increased variability of winter temperatures. Freeze–thaw cycles coupled with winter rain events could increase erosion and transport of streambank sediments with detrimental consequences for water quality and health of downstream aquatic ecosystems. 

Dam removals are on the increase across the US with Pennsylvania currently leading the nation. While most dam removals are driven by aquatic habitat and public safety considerations, we know little about how dam removals impact water quality and riparian zone processes. Dam removals decrease the stream base level, which results in dewatering of the riparian zone. We hypothesized that this dewatering of the riparian zone would increase nitrification and decrease denitrification, and thus result in nitrogen (N) leakage from riparian zones. This hypothesis was tested for a 1.5 m high milldam removal. Stream, soil water, and groundwater N concentrations were monitored over 2 years. Soil N concentrations and process rates andδ¹⁵N values were also determined. Denitrification rates and soilδ¹⁵N values in riparian sediments decreased supporting our hypothesis but no significant changes in nitrification were observed. While surficial soil water nitrate‐N concentrations were high (median 4.5 mg N L⁻¹), riparian groundwater nitrate‐N values were low (median 0.09 mg N L⁻¹), indicating that nitrate‐N leakage was minimal. We attribute the low groundwater nitrate‐N to denitrification losses at the lower, more dynamic, groundwater interface and/or dissimilatory nitrate reduction to ammonium (DNRA). Stream water nitrate‐N concentrations were high (median 7.6 mg N L⁻¹) and contrary to our dam‐removal hypothesis displayed a watershed‐wide decline that was attributed to regional hydrologic changes. This study provided important first insights on how dam removals could affect N cycle processes in riparian zones and its implications for water quality and watershed management.

 

Centuries‐long intensive land‐use change in the north‐eastern United States provides the opportunity to study the timescale of geomorphic response to anthropogenic disturbances. In this region, forest‐clearing and agricultural practices following EuroAmerican settlement led to deposition of legacy sediment along valley bottoms, including behind mill dams. The South River in western Massachusetts experienced two generations of damming, beginning with mill dams up to 6‐m high in the eighteenth–nineteenth century, and followed by construction of the Conway Electric Dam (CED), a 17‐m‐tall hydroelectric dam near the watershed outlet in 1906. We use the mercury (Hg) concentration in upstream deposits along the South River to constrain the magnitude, source, and timing of inputs to the CED impoundment. Based on cesium‐137 (¹³⁷Cs) chronology and results from a sediment mixing model, remobilized legacy sediment comprised% of the sediment load in the South River prior to 1954; thereafter, from 1954 to 1980s, erosion from glacial deposits likely dominated (63 ± 14%), but with legacy sediments still a substantial source (37 ± 14%). We also use the CED reservoir deposits to estimate sediment yield through time, and find it decreased after 1952. These results are consistent with high rates of mobilization of legacy sediment as historic dams breached in the early twentieth century, and suggest rapid initial response to channel incision, followed by a long decay in the second half of the century, that is likely dependent on large flood events to access legacy sediment stored in banks. Identifying sources of sediment in a watershed and quantifying erosion rates can help to guide river restoration practices. Our findings suggest a short fluvial recovery time from the eighteenth–nineteenth century to perturbation during the first half of the twentieth century, with subsequent return to a dominant long‐term signal from erosion of glacial deposits, with anthropogenic sediment persisting as a secondary source. © 2020 John Wiley & Sons, Ltd.