Historic changes in sediment loading and sediment sources in a large agricultural river system
Rivers and streams in intensively row-cropped, agricultural watersheds are often impaired by high sediment loads and turbidity. As expected, erosion from agricultural fields is a significant source of this sediment. However several recent studies have shown that erosion of non-field, near-channel sources such as streambanks and bluffs can be a larger contributor to the sediment impairments. Sediment cores from Lake Pepin, a natural impoundment on the Mississippi River, integrate the erosion history of a large agricultural watershed in the corn-belt of the Midwestern USA. These cores show that not only have sediment loads increased by nearly 10 fold over the past 150 years, but that non-field, near-channel sources now contribute more than two-thirds of the annual load. The increasing dominance of near-channel erosion sources raises the question: have rivers in agricultural watersheds become more erosive, and if so why?
Fundamental to any increase in erosive power of rivers is an increase in flow. We quantified changes in flow, water yield, and runoff ratio for 21 rivers in agricultural watersheds of Minnesota over the period 1940-2009. Half of the rivers showed statistically significant increases of 50-100% in annual flow. Equally important, half of the rivers showed no changes in flow, despite being geographically nearby. Increased precipitation, crop-conversions, and artificial drainage, including ditching, wetland drainage and sub-surface tiling, all have the potential to increase river flow, and all three have coincident trends over the latter half of the 20th century. To disentangle the role of each driver, we compared changes in climate, cropping patterns and drainage to changes in flow in each watershed. By calibrating annual flow to an annual water budget over the first 35 years of record, were we able to quantify the increase in flow in the more recent 35 years that could be attributed to each of the drivers. Changes in precipitation and crop conversion accounted for less than half of the increase in flow with the majority attributed to artificial drainage. Rivers with increased flow exhibited channel widening of 10-40%, with the amount of widening strongly correlated to the increase in annual flow. This widening of the stream channels is consistent with the observed increasing trend in non-field, near-channel erosion inputs to Lake Pepin and highlights land-use change in general and artificial drainage specifically as contributors to turbidity impairments in these rivers.