Grantee Research Project Results

 

Field monitoring studies were conducted to evaluate hormones in subsurface tile drainage and the ditch network receiving drainage from fields at an EPA-designated CAFO in Northern Indiana, which has approximately 600 ha of tile drained cropland treated with various types of animal wastes (beef, dairy, and poultry lagoon effluent, dairy solids, and subsurface injection of swine manure). Seven sampling stations (four in tile drains and three in the receiving ditch network) were installed and used to monitor flow continuously and collect water samples during storm events and baseline flow for a 17-month study period. Natural hormones 17α- and 17β-estradiol (E2), estrone (E1), estriol (E3), testosterone, and androstenedione and the synthetic androgens 17α- and 17β-trenbolone and trendione were monitored. Additional parameters such as pH, EC, nitrate, ammonia, ortho-P, selected pesticides, and dissolved organic carbon (DOC) also were measured on a subset of stream and tile drainage samples. The effects of CAFO impacted streams and ponds sites compared to reference sites were evaluated for abundance and diversity of fish (fish community study) and growth and reproductive conditions. Sexual differentiation and reproduction effects also were evaluated by exposing early life-stages of fish to hormones in controlled laboratory microcosms paired with field studies examining similar endpoints in native populations of fish in streams down gradient from CAFOs and in turtles inhabiting manure retention lagoons.

Our analysis of the monitoring work conducted from January 2009 to May 2010 focused on assessing hormone concentrations, fluxes and exports during storm events, base flow, thawing/snow melt events and effluent irrigation in subsurface tile drains and the ditch network receiving drainage from fields treated with various types of animal wastes. Hormones were detected in ~60% of the samples collected at each station during the 17-month study, with estrone being detected the most frequently and estriol the least. Testosterone and androstenedione were detected more frequently than synthetic androgens, which were detected in fewer than 15% of the samples. Hormone concentrations in subsurface tile drains increased during effluent irrigation and storm events. Hormones also appeared to persist over the winter, with increased concentrations coinciding with early thaws and snow melt from fields applied with manure solids. The highest concentration of synthetic androgens (168 ng/L) observed coincided with a snow melt. The highest concentrations of hormones in the ditch waters (87 ng/L for total estrogens and 52 ng/L for natural androgens) were observed in June, which coincides with the early life stage development period of many aquatic species in the Midwest.

 

Hormone fluxes were highest during storm events shortly following animal waste applications with flow-weighted concentrations generally decreasing during subsequent storm events prior to additional applications. The total estrogen fluxes from fields were highest during large spring storm events between May and June 2009 following dairy effluent irrigation and dairy solids applications. Hormone loads exported from the receiving ditch network ranged from 16 to 58 mg/ha for total estrogens, 6.8 to 19 mg/ha for natural androgens, and 4.2 to 44 mg/ha for synthetic androgens. High temporal variability in hormone export led to the majority of hormone loads occurring when flow rates were in the 80th percentile. High export during the spring months has the potential to negatively impact sensitive aquatic ecosystems downstream, as it coincides with the sensitive early life stages of many aquatic species. Therefore, our findings suggest that the short periods of time during which high-flow events occur must be targeted to effectively reduce hormone loads exported to downstream aquatic ecosystems.

 

From our sorption and degradation studies, isomeric-specific degradation was not evident. However, sorption is isomeric-specific with the α isomer sorbing less by approximately a factor of two. Therefore, α and β isomers sorbing the same is not a conservative decision-making approach. The lower sorption affinity of the α-isomer increases its likelihood to be transported easily to streams and rivers from agricultural fields, which is significant given it is the dominant TBA metabolite in manure and has been observed to have similar aquatic reproductive effects as 17β-trenbolone and likewise for 17α-E2 for some animal species.

 

Fish species diversity, intolerant species, and the index of biological integrity (IBI) showed significantly lower species richness and IBI for the two CAFO impacted agricultural ditches compared to the reference site. Dissolved oxygen was more consistent at the reference site compared to the CAFO impacted sites, but the means were not statistically different; however, water temperatures were 4 to 5 °C higher at the reference site. During field sampling for the fish community study in late April/early May and mid-June, 12 creek chub were randomly collected from each of the three sites to assess a gonadosomatic index (GSI), reproductive conditions, and reproductive stage. Somatic growth appeared faster at one of the CAFO impacted sites compared to the reference site as well as significant differences in reproductive conditions. Throughout the summer months, water levels declined and temperatures increased in the agricultural ditches with production of algal blooms, which severely reduced levels of dissolved oxygen throughout the summer. This resulted in an even higher decline in fish species richness and abundance. In comparison, the control site retained higher dissolved oxygen levels, higher total numbers of fish, and greater species richness. These results imply that current ditch management practices have negative impacts on aquatic fauna.

 

Fathead minnow embryos exposed to CAFO-impacted ditch water in spring-summer 2009 showed a higher % of males (~58 %) relative to the control populations, but not of statistical significance. In collaboration with USEPA Cincinnati and Duluth laboratories, we conducted in situ caging studies deploying adult fathead minnows in our CAFO and reference sites for a week in early June, which showed no significant differences in their relative expression of vitellogenin (VTG, egg yolk protein). Differences in plasma VTG concentrations in plasma samples collected from common snapping turtles (32 were recorded as male and 19 were recorded as female) over the three summers from 2007-2009, also were not statistically significant between CAFO impacted and reference sites. However, male turtles at the CAFO impacted poultry pond did have significantly higher testosterone than those at the reference pond.

Hormones appeared to persist over the winter, with increased concentrations coinciding with early thaws and snow melt from fields applied with manure solids. However, hormone fluxes and loads were highest during large spring storm events between May and June 2009 following animal waste applications. The majority of hormone export occurred when flow rates were in the 80th percentile of the flows experienced throughout the year. Due to this significant temporal inequality exhibited by hormone export, best management practices must be aimed at reducing export during high-flow events to achieve significant reductions in hormone loads. If these “windows of opportunity” are missed, then any efforts to mitigate loads during all other periods of time will have minimal impact on load reduction. Therefore, it is unlikely that best management practices effective at reducing loads during low-flow events (e.g., wetlands, no-till) will be effective best management practices for reducing hormone loads. The best management practices most likely to effectively reduce hormone loads are those that can intercept exported hormones during high-flow events (e.g., buffer strips) and those that reduce the input of hormones to agroecosystems (e.g., reducing the hormone to macronutrient ratio of applied animal wastes).

 

Although lower species richness and indices of biological integrity (IBI) appeared significantly lower in agricultural ditches impacted by drainage water from fields receiving animal wastes, variations in natural environmental parameters complicate a clear assessment of chemical exposure impacts. In situ exposure studies with fathead minnows and embryos, and snapping turtles revealed no statistical differences in reproductive conditions relative to controls. Limited knowledge exists on molecular pathways associated with sex differentiation in early life stages of fathead minnows. Therefore, we developed a method to determine genetic sex in early life stage fathead minnows using a sex-linked DNA marker for which changes in the expression of genes important in sex differentiation are being exploited to more definitively evaluate effects of endocrine disrupting chemicals on early life stage fish.