These data were collected in support of the sampling goals of the Aquatic Intermittency effects on Microbiomes in Streams (AIMS) Project. This study took place in Shane’s Creek (434 ha) at the Konza Prairie Biological Station, a long term ecological research station that has been in operation since 1980. Shane’s Creek is annually cattle-grazed and burned every three years. Konza is located in the Flint Hills of northern Kansas. In 2023, the outlet of the stream wet up in March and dried down in July; in 2024, the outlet of the stream wet up in March and dried down in September. Average annual precipitation for this site is 35.62 inches.
We constructed a wooden stream diversion structure halfway down a 380m reach containing four pools and three riffles above and below the structure. The structure had 4 1-foot diameter holes installed with PVC couplers. Four 130m corrugated flexible tubes were laid out from the diversion structure to the bottom of the impact reach. We deployed 20 STICs throughout the control and impact reach to quantify the extent of drying using rebar. Construction and preparation occurred in February and March, when the stream was dry, to minimize construction-related disturbances to the experimental sampling. Prior to experimental dry down, water was able to flow from upstream (control reach) to downstream (impact reach) through the holes. We obtained pre-dry down sampling to collect a reference point for both the control and impact reaches. These “pre-dry week X” samples were collected from April to mid-July due to multiple flooding events (April 25th, June 26th, and July 3rd) that disrupted the experimental setup and required redeployment of experimental structures and equipment. During the experimental dry down (“dry week X” samples), we attached the tubes to the couplers in the diversion structure. Flow was diverted around the impact reach into the downstream watershed for five weeks, during which we collected weekly samples. Following the final forced drying sampling, tubes were cut from their couplers to allow flow to resume in the impact reach (“re-wet week X” samples). After 6 hours, we performed day 0 re-wet sampling. Twenty four hours later, we performed day 1 re-wet sampling. One week later, the stream began to naturally dry down in both the control and impact reaches, and collected weekly natural dry down samples for two weeks (“natural dry down week X” samples).

We collected triplicate water samples for soluble reactive phosphorus (SRP; µg/L) and ammonia (NH4-N, µg/L) on each sampling occasion with a subset of samples also analyzed for nitrate/nitrite (NO3-N, NO2-N, µg/L). We collected samples at the specified location when water was present using a syringe in a well-mixed area of the stream. We then filtered water through sterile PVDF 0.45 µm syringe filters (VWR) into clean bottles following the AIMS Surface Water Chemistry SOP (Burgin 2024). For all analytes, we froze samples until analysis in the lab using colorimetric methods on an AQ300 Discrete Analyzer (SEAL Analytical, Mequon, Wisconsin, USA). We prepared and applied reagents to samples and standards of known concentrations. For SRP, reagents react to form a blue complex, the absorbance of which is measured at 880nm. Known standards were used to create calibration curves ranging from 0 μg/L to 75 μg/L with check standards run every 10. Detection limits were set at 6 μg/L. We used the phenol method for NH4-N with known standards used to create calibration curves ranging from 0 μg/L to 500 μg/L and check standards run every 10 samples. Detection limits were set at 11 μg/L. To determine NOx-N, we analyzed samples with and without cadmium reduction with known standards used to create calibration curves ranging from 0 µg/L to 1000 µg/L for NOx-N and 0 µg/L to 750 µg/L for NO2-N with detection limits set at 20 µg/L NO3-N. Differences between NOx-N and NO2-N were used to calculate NO3-N. Triplicates were inspected for outliers with mean and standard deviation reported in data.