Novel Approaches in the Assessment of Agricultural Tile-Drain Chemical Transport
Applying creative approaches to overcome the innate hurdles associated with complex study designs is another day at the office for the experts on our field studies team. Never was this more true than when one of our clients asked for help in satisfying the new EPA goals surrounding the effects of nutrient transport outlet water. Conventional study work would have unnecessarily wasted precious time and resources and was certainly not in our client’s best interest.
Excess nutrient transport from agricultural settings have contributed to a hypoxic zone in the Mississippi River basin and Gulf of Mexico, which, in turn led the EPA to form the Hypoxia Task Force with a goal to reduce the size of the hypoxic zone to less than 5,000 km2 by 2035. The Task Force also has an interim goal of a 20% reduction of nitrogen and phosphorus loading by 2025. Unsurprisingly, these environmental goals have created an urgent need for studies to better understand nutrient transport in agricultural landscapes.
At the client’s request, a Waterborne team led by our Senior Agricultural Engineer Greg Goodwin established a large-scale field study to observe the effects of treatments on the nitrate-nitrogen concentration in outlet water of a tile-drain corn and soy rotated field. The study design included 37 discrete tile monitoring stations, making for quite a large monitoring effort! Our team went to work to evaluate possible monitoring approaches that could be employed in this case. It didn’t take long to realize that a conventional water sampling approach with individual samples and physical sample collection would waste precious study resources in time and cost. A novel approach was certainly needed in this case!
After careful consideration of possible approaches, the team landed on an automated pass-through study design. This approach made use of dataloggers controlling water-level sensors and pumps at each of the 37 separate tiles. The team also set up two analysis stations capable of radio-communicating with the dataloggers. This setup allowed for individual tile station pumps to push water samples to the appropriate analysis station where it was then fed over a nitrate sensor set to automatically read and store sample results. The analysis stations were also equipped with full weather stations and tipping-bucket rain gauges. Data was transmitted via cellular modem and accessible to the project team in real-time.
Application of this automatic approach drastically improved sampling frequencies and gave the team the ability to collect data that wouldn’t have been feasible collecting individual physical water samples, including back-to-back flow events without the need for site visits. Improvement of the sampling frequency allowed our data scientists to see trends that we may not have otherwise captured with more infrequent sampling intervals. The approach also decreased time and costs associated with manual labor, sample transport, and analysis. The real-time availability of the analyte data helped the project team make faster study decisions.
In addition to the automated approach applied to this study, this design allowed for the use of sophisticated data analysis techniques. We used Aquarius as a means to identify and remove erroneous measurements using USGS-approved processes. This design also allowed us to apply a correlation matrix to each of the tiles and then used Pearson’s correlation coefficients to identify stronger and weaker correlation of tile flow measurements to each other. These correlations were then used to select the tiles for assignment of one of four application rates used in the study. Ultimately, the increase of replicates allowable by the automated design offered us a unique statistical power in this study to assign treatments from a baseline comparison of replicate responses. We’re continuously working to bring our clients options for novel study designs, whether it be in the form of new equipment, automated processes, out-of-the-box solutions, or assessment of statistical power within a study. If you have any questions of how a novel field study design could help with your specific needs, please feel free to contact Greg Goodwin at firstname.lastname@example.org.