ASSESSMENT OF DIEL CYCLING IN AGRICULTURAL STREAMS IN SOUTH-CENTRAL MICHIGAN- A FOCUS ON TURBIDITY
Two study sites in Rice Creek, an agricultural drain, show turbidity increases of 10 to 20 NTU at night with broad noisy peaks between 10 pm and 6 am and daytime saddles between 10 am and 7 pm. Downstream turbidity peaks lag upstream peaks by ~2 hours over a distance of 7.7km. Turbidity cycles also vary with depth in the stream. Measurements taken at 10 and 20 cm above the streambed exhibit in-phase diel cycles with the same periodicity and amplitude. Magnitudes are offset by 10 NTU, with higher turbidity closer to the streambed. A strong linear relationship (R2 of 0.89) between turbidity and TSS concentration suggests that turbidity is a good proxy for TSS. Loss-on-ignition of TSS samples indicates that suspended sediment is composed of about 40% organic and 60% mineral material. Measurements in the more natural North and South Branch tributaries of the Kalamazoo River exhibit similar diel turbidity cycles but with lower amplitudes and absolute values than in Rice Creek.
Turbidity cycles are roughly anti-phase with DO and pH, which may indicate a biological control. Simulation experiments of bioturbation result in short-lived and short-traveled pulses of turbidity. Turbidity varies in-phase with groundwater levels and specific conductivity, which may indicate a groundwater sapping effect. Leveloggers measured groundwater levels in four wells in adjacent wetlands. Water table levels, driven by evapotranspiration, varied cyclically with diel amplitudes from 1 to 9cm. We anticipated that stream stage levels would co-vary with groundwater levels but they did not. Stream stage cycles were less regular, out-of-phase, and about 2cm in amplitude relative to groundwater levels. Ongoing work is investigating both biological and physical controls on the diel turbidity cycles.