Reassessing the role of milldams in Piedmont floodplain development and remobilization

• Erosion from streams with and without historical milldam influence were compared across different spatial scales.
• Average milldam deposits contribute 15% greater erosion compared to surrounding streambanks.
• At length-scales of 102 meters, percent variance in bank erosion explained by milldams and local gradient are comparable.
• Segment-averaged erosion is not significantly different for groups of streams with and without the influence of milldams.
• Milldams may be most effective at increasing bank erosion along streams draining 10 – 100 km2.
• Our results bridge conflicting research and provide a conceptual model to guide effective sediment reduction strategies.

Recent emphasis on milldams as a cause of colonial-era floodplain sediment deposits and source for modern remobilization has ignited debate in scientific and environmental policy forums, especially in the mid-Atlantic USA. Although milldams are regionally ubiquitous and visually striking, no investigation has demonstrated that they are a dominant driver of regional floodplain sediment accumulation and remobilization. Long-term analysis in Baltimore County, Maryland found that milldams were sufficient, but not necessary for historical sediment accumulation. We compared erosion from streams upstream and downstream of confirmed milldam locations, in paired segment comparisons with and without the influence of historical milldams, and pooled across 37 study segments. We found the scale of inquiry produced different interpretations regarding the role of milldams in regional erosion and sedimentation. Although local milldam sediment remobilization rates were 15% greater than those immediately downstream, neither paired nor pooled segment comparisons provided evidence that milldam deposits are contemporary dominant sources of sediment. Further comparison across drainage areas demonstrated that milldam deposits could raise local remobilization rates by augmenting sediment availability along supply-limited channels. We introduce a unifying conceptual model that explains apparently contradictory findings in the literature where, rather than the sole cause of legacy sediment deposits, some milldams are nonetheless responsible for local increases in sediment remobilization rates along portions of the stream network.

Scatterplots illustrating trends of stream power, floodplain sediment storage, and gross erosion rates in Mid-Atlantic stream segments. (A) The trends of gross erosion for milldam (MD) segments exhibit significantly larger erosion rates than non-milldam (NMD) segments at drainage areas between 12 and 80 km2 (vertical dashed lines). (B) Data illustrating how milldam deposits might produce the observed results above. Stream power - estimated as the product of channel slope and log-drainage area (Seidl and Dietrich, 1992) - peaks between 2 and 50 km2. Yet the trend in floodplain sediment storage (red), measured as the product of segment-averaged valley width and bank height, exhibits a notable increase beyond 40 km2. The gross erosion trend for segments without milldams in A corresponds to the trend in sediment storage in B (both red lines). The increase in gross erosion for milldam segments in A may occur when milldams increase local sediment storage (hypothetical dashed blue line in B) along small- to mid-basin channels with relatively high transport capacity.Figure optionsDownload as PowerPoint slide