Exports of Hormones in Surface Runoff From Agricultural Fields Receiving Poultry Litter.

The use of poultry litter as a substitute for inorganic fertilizer is increasingly being encouraged, especially in states like Delaware, which have a surplus of poultry litter. However, in addition to nutrients, poultry litter also contains naturally occurring steroidal hormones which can be transported with runoff and sediment.  These hormones are also referred to as endocrine disrupting chemicals (EDCs) since they cause physiological and reproductive disorders in aquatic and wildlife species.  In this study, we evaluated the transport of estrone (E1), estradiol (E2) and estriol (E3) in surface runoff from agricultural plots subject to poultry litter application.  Treatments included raw poultry litter at two different levels and litter in pelletized form.  The experimental plots were 5m wide and 12 m long with reduced and no–tillage and with corn as the cover crop.  Surface runoff sampling was performed for 10 natural storm events that occurred during the summer (April-August) of 2008.  Initial screening of hormones in the water samples was performed using the enzyme linked immunoassay (ELISA) technique followed by confirmatory tests by LC/MS/MS.  Concentrations and mass exports of 17β–Estradiol (E2β), E1 and 17β–Estradiol–sulfate (E2β–17S) were significantly greater (p ≤0.10) from plots with litter application versus those without litter (control). Furthermore, mass exports and concentrations of the hormones were higher from plots with raw litter versus those receiving litter in pelletized form.  While the flow-weighted concentrations of hormones were low (<5 ng/L) the total mass exports varied between 5 to 133 µg/ha. Mass exports of hormones in runoff were significantly correlated (p ≤0.10) with the rainfall amounts but did not show any correlation with the timing of events (days after litter application). Although mass exports of hormones from no-till plots were lower than those from reduced tillage plots, the difference was not statistically significant.