A field trial was done in the Upper Tiffin River Watershed, in southeastern Michigan, to determine the influence of liquid dairy manure effluent (LDME) management practices on the quality of agricultural subsurface-drain water. Samples from subsurface drains were analyzed for nutrients, fecal-coliform and Escherichia coli (E. coli) bacteria, antibiotics, chemicals typically detected in wastewater, and the occurrence of genes indicating the presence of shiga-toxin-producing E. coli, or of bovine-specific Bacteroidetes bacteria. Samples were collected from November 2, 2006, to March 20, 2007, from eight subsurface drains under field plots that received no LDME and no tillage (controls) or received 4,000 or 8,000 gallons per acre (gal/acre) of LDME and either no tillage or two different types of tillage. The two types of tillage tested were (1) ground-driven, rotary, subsurface cultivation and (2) rolling-tine aeration. Samples were collected before LDME application and at 4 hours, and 1, 2, 6, 7, and 14 days post-application.

Nutrient concentrations were high in subsurface-drain water throughout the field-trial period and could not be attributed to the field-trial LDME application. Of the 59 drain-water samples, including those collected before LDME application and control samples for each date, 56 had concentrations greater than the U.S. Environmental Protection Agency (USEPA), Ecoregion VI recommended surface-water criterion for total phosphorus, and all samples had concentrations greater than the recommended total nitrogen criterion. Nitrate + nitrite nitrogen concentration exceeded 20 milligrams per liter for every sample and contributed most to the total nitrogen concentrations. Substantial increases in drain-water concentrations of organic and ammonia nitrogen and total phosphorus were found for all treatments, including controls, at 14 days post-application after 0.84 inch of rainfall over 2 days.

E. coli concentrations exceeded the USEPA recreational-water-quality single-sample criterion of 235 colony forming units per 100 milliliters in only 3 of 56 samples. Of these three samples, two were collected within 1 day post-LDME application from the treatment receiving 8,000 gal/acre LDME with no tillage (NT8000). The third sample was from the rolling-tine aerator treatment with 4,000 gal/acre LDME application rate after the first significant rainfall.

Two wastewater chemicals and two bacterial genes (eaeA and stx1) detected in the LDME, but absent in field blank or pre-application samples, were detected in the 4-hour or 1-day postapplication NT8000 samples. No LDME-associated chemicals were detected in later samples from the NT8000 treatment, and none were detected in samples from other treatments after the first significant rainfall.

Results of this field trial were somewhat equivocal with respect to the influence of LDME concentration and tillage practices on subsurface-drain water quality, both immediately after LDME application and in the longer term, after significant rainfall. Interpretation of study findings is limited by the fact that treatments were not replicated, and flow rate or discharge from the subsurface drains was not measured. Nevertheless, study results provide useful information about nutrient and bacteria concentrations in subsurface drains during the non-growing season. In addition, study results demonstrate some potential for the use of chemical and microbiological indicators of LDME transport to subsurface drains.