Poultry manure (PM) contains nutrients that are highly beneficial for crop production, but loss of nutrients, pathogenic bacteria, hormones, and other PM constituents to groundwater and surface waters can disrupt aquatic ecosystems, raise concerns for recreational water users, threaten drinking-water systems and human health, and have economic consequences for producers. While transport of manure-derived contaminants via runoff has been widely studied, transport of contaminants through artificial drainage systems in less well understood. The principle objective of this study was to compare the effects of poultry manure application rate and tillage practices (chisel-plow and no-till) on drainage tile-water quality. The secondary objective was to further understand the processes and pathways which impact the fate and transport these contaminants. While the focus of this dissertation is on fecal indicator bacteria (Escherichia coli and enterococci), the pathogen, Salmonella, and phosphorus, concentrations of nitrate and hormones in tile-water samples were also assessed. Results of this research can be used to inform farm-management decisions and watershed planning efforts.

Test plots used for this study are located at Field 5 at the Iowa State University¡¯s Agricultural and Biological Research Station, approximately 8 mile (13 km) west of Ames, Iowa. These plots contain calcareous loamy Canisteo-Clarion-Nicollet series soils derived from glacial till. Each plot is drained along its midline by a tile 1.2 meters deep, which outlets into a sump from which flow volume is determined. From 1998-2009, 9 plots were split in half, with each side receiving corn or soybeans in alternate years, and each plot was chisel-plowed annually, prior to planting. In 2010, these plots were transitioned to a continuous-corn rotation, and three established no-till plots were added to the study. From 2010 to 2012, PM was delivered to the site each spring, and broadcast to achieve target rates of 112 and 224 kg/ha nitrogen (PM1 and PM2). Actual PM application ranged from 5 ¨C 40 kg/ha depending on the nitrogen content of the manure. Control plots received no manure (PM0). Each manure application treatment was replicated on three chisel-plowed (CP) plots and one no-till (NT) plot.

To determine the effect of manure application rates and tillage on bacteria concentrations in tile-waters, water samples were collected weekly from drainage tile outlets and following precipitation events from 30 days before manure application to 100 days post application, when tiles were flowing. In 2011 and 2012, additional tile monitoring was employed to capture the hydrographic response to precipitation, and in 2012, smoke-testing was used to determine macropore densities. Manure and tile-water samples were analyzed for the pathogen, Salmonella spp. (SALM), fecal indicator bacteria (FIB), E. coli (EC), and enterococci (ENT). All three bacterial species were detected in tile-waters, with the highest plot-normalized concentrations observed under the NT PM2 plot in 2010 (3.7 ¡Á 104 cfu/100 mL EC, 3.6 ¡Á 106 cfu/100 mL ENT, and 1.6 ¡Á 104 cfu/100 mL SALM). Macropore density above the tile was also highest (13.9 pores/m) in the NT PM2 plot, in 2012. Individual and 30-day geometric mean ENT concentrations correlated more strongly to SALM than EC.

To determine whether over-wintering of bacteria had occurred, soil samples were obtained each spring at 0-15 cm and 15-30 cm depths. In 2012, soil samples were collected 21, 42, and 158 days after manure application (DAM), to assess the effects of time, manure application rates, and tillage on frequency of detection and concentrations of bacteria. Detection of all three bacteria species in spring soil samples indicated over-wintering of bacteria was possible. Despite dry conditions in 2012, all three bacteria species were detected 158 DAM. The highest SALM concentration (790 cfu/g dry weight) and detection rate (25%) was found in PM2 plots. SALM detection was higher in CP plots (20%) compared to NT plots (5%). In contrast, tillage practices had no apparent effect on EC or ENT survival, as indicated by both soil sample analyses, and decay rates estimated from tile-water bacteria concentrations. Decay rate constants (¦Ì) ranged from 0.044 to 0.065 day-1 for EC, and 0.010 to 0.054 day-1 for ENT.

Soil and tile-water samples were also used to determine the effects of manure application rate and tillage on orthophosphate (PO4-P) losses to tile drainage waters. Spring soil samples (0 ¨C 15 cm and 15 ¨C 30 cm) were analyzed for soil test phosphorus (STP) using the Bray¡¯s 1-P method. Water samples were obtained from drainage tiles following precipitation events in 2010 and 2011, and more frequent sampling throughout selected hydrograph events was conducted in 2011 and 2012. STP levels increased significantly under the PM2 treatment over the three year period, while significant differences were not observed under the PM1 treatment or the control plots (PM0). Despite the similarity between 0 ¨C 15 cm STP levels in chisel-plowed (CP) and no-till (NT) plots, orthophosphate (PO4) concentrations in drainage tile-waters were significantly higher under NT plots (>0.1 mg/L PO4-P) than under CP plots (¡Ü0.04 mg/L PO4-P) at P < 0.05. Event hydrographs from 2011 and 2012 showed short-lived peaks in PO4-P concentrations, which may be explained by macropore transport; however, elevated levels of PO4-P remain throughout the hydrograph in NT plots, suggesting that matrix flow accounts for much of the PO4-P transport to tiles. Further analyses of soils data indicates that NT soils had elevated pH levels relative to CP plots, indicating the potential for soil to retain phosphorus may be impacted by chemical conditions in the soil.

In general, this study shows that applying poultry manure at a low (~112 kg/ha N) application rate to chisel-plowed ground is likely to result in smaller bacteria and orthophosphate losses to tile-waters than application at a higher rate (~224 kg/ha N) on no-till plots. Even when heavy manure application and no-till practices were combined, tile-waters only exceeded recreational water quality standards (which do not apply directly to tile outlets) for E. coli under exceptionally wet conditions. However, results of this study also show that Salmonella can be present in waters containing little or no fecal indicator bacteria.

Producers and watershed managers can take away several important lessons from this research. This study confirms what is already widely accepted, that using lower application rates for poultry manure, and using fertilizers with lower levels of P to balance crop needs, would reduce contaminant losses, benefitting both producers and the environment. An increase in tillage, on the other hand, increases the risk of runoff and soil compaction, and may have other drawbacks; therefore, more modest changes to tillage, such as occasional disking, may be recommended to reduce bacterial water quality impacts by disconnecting macropores and helping to distribute phosphorus within the soil column. Control of pH, and other techniques for increasing the ability of soils to retain phosphorus, may prevent losses to tiles as well as surface-runoff. Although additional research is necessary, the results also suggest that adaptation of the Iowa P-Index to account for the potential for greater losses of P via the subsurface should be considered. Furthermore, transport of bacteria and phosphorus via tile drainage should be considered when designing watershed models and water quality improvement plans.