Effects of Drain Depth on Nitrogen Losses in Drainage in Shallow Water Table Soils

Abstract

A two-part study (a field study and a modeling study) was undertaken to investigate the effect of shallow subsurface drains on nitrate-nitrogen (NO3-N) losses in drainage effluent relative to deep subsurface drains. The field study was conducted in the Lower Coastal Plain of North Carolina over a two year period between May 12, 2001 and April 30, 2003. Crops grown on the site were corn in 2001 followed by wheat and soybean in 2001/2002. The field was nearly flat and the soils were Portsmouth sandy loam and Cape Fear loam, which, under natural conditions are very poorly drained. The site was subdivided into eight 1.8 ha plots, each drained by three parallel drains spaced 23 m apart. Drains in plots one through three were shallow with an average drain depth of 0.86 m. Drains in plots four through six were deeper, with an average depth of 1.20 m.

Precipitation, water table depth and subsurface drainage rates were measured continuously during the study. Subsurface drainage was sampled and analyzed to determine NO3-N concentration in the effluent. Measurements of water table depth and subsurface drainage quantity and quality were made within each plot.

Observed results indicate the average flow quantity from the shallow drains was reduced by 37% in Year 1 and by 26% in Year 2 relative to the average flow from the deep drains. However the difference in flow was not statistically significant. Average water table depth in the shallow plots was significantly more shallow than in the deep plots at the 10% level.

Drain depth had no statistically significant effect on drainage water quality. For Year 1, the average NO3-N losses from the shallow drains and deep drains were 21.7 and 28.0 kg ha-1, respectively. For Year 2, the average NO3-N losses from the shallow and deep drains were 28.9 and 23.5 kg ha-1 respectively. For the entire study period, total average NO3-N loss from the shallow drains was 50.6 kg ha-1 and the total average loss from the deep drains was 51.5 kg ha-1.

Observed behavior in shallow plot 2 was an anomaly. More subsurface drainage occurred and more NO3-N was lost from plot 2 than any other plot. Based on analysis performed in conjunction with predicted results from the modeling study, it was determined that plot 2 was not hydraulically isolated and was receiving surface runoff from an adjacent plot.

The second part of the study was a modeling study to further investigate the effect of drain depth on nitrate loss in a way that eliminated the plot to plot variation that was inherent in the field study. Hydrologic models of the research site were developed using DRAINMOD 5.1. Nitrogen dynamic models were developed using DRAINMOD-N II. Initial simulations were performed to calibrate the models using data collected during the field study. Then long-term simulations were performed for the period November 1991 to March 2003 using the calibrated models. A wheat-soybean-corn crop rotation was used. Simulations were performed for each plot with subsurface drains at 0.85 m and 1.2 m.

Based on DRAINMOD predictions for the plot 2 through 6, reducing drain depth from 1.2 to 0.85 m resulted in 17% reduction in subsurface drainage, 30% increase in surface runoff, and 4% increase in evapotranspiration over an 11 year period. Based on DRAINMOD-N II predicted results, averaged for plots 2 through 6, the 0.85 m drain depth resulted in 32% less NO3-N loss in subsurface drainage, compared to drains at a depth of 1.2 m.