The Presence of 17-β Estradiol and Fluoroquinolones in Kentucky s Surface Water

Size: px
Start display at page:

Download "The Presence of 17-β Estradiol and Fluoroquinolones in Kentucky s Surface Water"

Transcription

1 The Presence of 17-β Estradiol and Fluoroquinolones in Kentucky s Surface Water E.G. Beck 1, R. DeJaco 2, S. Webb 1, A. Fogle 3, D. Williams 1, M. Silitonga 4, and B. Lee 5 1 University of Kentucky, Kentucky Geological Survey 2 University of Kentucky, Department of Chemical and Materials Engineering 3 University of Kentucky, Department of Biosystems and Agricultural Engineering 4 Kentucky State University, College of Agriculture, Food Science and Sustainable Systems 5 University of Kentucky, Department of Plant and Soil Sciences Kentucky Water Resources Research Institute Symposium March 18, 2013 Lexington, Kentucky

2 17-β Estradiol One of 3 main estrogens (steroids) produced by the human body Predominant and most potent sex hormone in females Produced by animals (dairy cows, etc) Synthetic estrogens (17 alpha EE2) Also called E2 or estradiol 18 carbons 2 hydroxyl Potential Sources to the Environment Residential straight piping Runoff from fields where animals graze Wastewater treatment plant (WWTP) outfalls Runoff from fields where wastes (animal/human) are applied

3 Fluoroquinolones A subset of the synthetic broad-spectrum antibacterial drugs known as quinolones. Fluoroquinolones have a fluorine atom attached to the central ring system. Fluoroquinolones are used to treat humans and animals. Human Veterinary Ciprofloxacin Ofloxacin Enrofloxacin Fleroxacin Pefloxacin Danofloxacin Enoxacin Norfloxacin Marboloxacin Levofloxacin Ciprofloxacin Potential Sources to the Environment WWTP outfalls, runoff, and straight piping

4 Environmental Concerns 17-β Estradiol (most potent) Is released to the environment (Kolpin and others, 2002, etc) Half-life in river water of 0.2 to 9 days (Jurgens and others, 2002) Endocrine disruptor (Purdom and others, 1994; Jopling and others, 1998) 1 ppt shown to affect reproduction in trout (Hansen and others, 1998) Fluoroquinolones Is released to environment (Andreozzi and others, 2003, etc) Bacterial resistance from use in poultry production U.S. FDA banned use in poultry production (2005) Affects on wildlife are unknown Fate in environment is unknown Continuous input Not regulated

5 Two Pilot Studies (Screenings) Sampling Round 1 (June-July 2012) USDA-NIFA Southern Regional Water Program 10 states (Kentucky, Tennessee, Mississippi, Oklahoma, Arkansas, Texas, North Carolina, Georgia, Florida, and Alabama) 19 sites sampled in the Floyds Fork watershed KyWRRI Student Enhancement Grant 32 sites sampled in 5 watersheds (Bayou de Chien, Clarks River, South Elkhorn Creek, Banklick Creek, and Licking River) Sampling Rounds 2-5 (November-December 2012) USDA-NIFA Southern Regional Water Program 20 sites in 4 watersheds (80 samples) (Bayou de Chien, Clarks River, Floyds Fork, and South Elkhorn Creek)

6 Watersheds (HUC 11) Round 1: All Rounds 2-5: Bayou de Chien, Clarks River, Floyds Fork, and South Elkhorn Creek

7 Watershed Sampling Sites From bridges In stream (bridge accesses and WWTP property)

8 Field Methods Recorded field measurements using a Horiba U-22 water-quality meter* Measured discharge using a Marsh-McBirney flow meter* Grab samples collected in 40-ml amber vials USGS stream gage data when possible

9 17-β Estradiol Analytical Method Enzyme-Linked Immunosorbent Assay (ELISA) kit from Abraxis LLC* Cross-reactivity (n) 17-β Estradiol 100% (s) 17 alpha EE2 0.36% (n) Estrone (E1) 75% E2-3 Glucoronide < 1% (m) 16-Keto E2 27% E2-3 Sulfate < 1% (n) Estriol (E3) 4.1% Cholesterol < 1% ChroMate 4300 Microplate Reader by Awareness Technology, Inc.* Samples kept at 4 C and out of direct sunlight Samples analyzed within 7 days of collection 3 reps per sample bottle

10 Fluoroquinolones Analytical Method Enzyme-Linked Immunosorbent Assay (ELISA) kit from Abraxis LLC* Cross-reactivity (a) Enrofloxacin 100% (h) Ofloxacin 11% (a) Danofloxacin 100% (h) Norfloxacin 7% (h) Ciprofloxacin 66% (a) Marbofloxacin 6% (h) Fleroxacin 25% (h) Levofloxacin 5% (h) Enoxacin 17% (a) Sarafloxacin < 1% (h) Pefloxacin 12% Oxolinic Acid < 1% ChroMate 4300 Microplate Reader by Awareness Technology, Inc.* Samples kept at 4 C and out of direct sunlight Samples preserved with 10% methanol within 24 hours of collection Samples analyzed within 14 days of collection 3 reps per sample bottle

11 Method Detection Limit (MDL) 17-β Estradiol Test sensitivity of 2.7 parts per trillion (ppt) (Abraxis LLC, 2011*) Calculated MDL is 3.0 ppt Kentucky Geological Survey Laboratory Services (40 CFR Part 136, Appendix B, Revision 1.11, 2011) Fluoroquinolones Test sensitivity of parts per billion (ppb) (Abraxis LLC, 2011*) Calculated MDL is ppb Kentucky Geological Survey Laboratory Services (40 CFR Part 136, Appendix B, Revision 1.11, 2011)

12 Confidence of Analytical Method for 17-β Estradiol Step 1. Standard Concentrations Rep 3.0 ppt 6.0 ppt 15.0 ppt Std. Dev Mean Bias 17% 6% 2% Step ppt 6.0 ppt 15.0 ppt Standard Error *Standard Error (CI) Lower Confidence Limit 3.2 ppt 5.9 ppt 14.9 ppt Upper Confidence Limit 4.0 ppt 6.9 ppt 15.7 ppt Results at or near the MDL of 3.0 ppt are suspect ELISA is an adequate screening method only

13 Statistical Analysis of Data Step 1: Averaged analytical reps with all detects Step 2: Used R (free statistical software) and maximum likelihood estimation (MLE) method to average analytical reps with nondects (Helsel, 2005) Step 3: Used R and Kaplan-Meier method to create boxplots of data (included detects and nondects) (Helsel, 2005). Step 4: Used MLE tests within R to determine if there are significant differences between sampling sites and watersheds (p < 0.05 indicates 95% confidence level).

14 Summary of Averaged Data Estradiol Number of Samples Number of Nondects (percent) Watershed Min Max Bayou de Chien < (75) Clarks River < (73) Floyds Fork < (60) South Elkhorn Creek < (53) Banklick Creek < (67) Licking River < (80) Fluoroquinolones Number of Samples Number of Nondects (percent) Watershed Min Max Bayou de Chien < < (100) Clarks River < (68) Floyds Fork < (67) South Elkhorn Creek < (72) Banklick Creek < < (100) Licking River < (90)

15 Southern Region Comparison Estradiol (ppt) Fluoroquinolones (ppb) State Min Max Min Max Kentucky < < Alabama < < < Arkansas < < < Florida < < < Georgia < < Mississippi < < North Carolina < < Oklahoma < < Tennessee < 3.0 < 3.0 < < Texas < <

16 Banklick Creek Land Cover (percent) Forest Developed Pasture Cultivated Other Sample Estradiol Fluoroquinolones Site (ppt) (ppb) < < 3.0 < < 3.0 < < < 3.0 < < 3.0 < sites 0 site To few data points to statistically compare sites within watershed

17 10 Licking River 9 Salyersville outfall Land Cover (percent) Forest Pasture Developed Cultivated Other Sample Estradiol Fluoroquinolones Site (ppt) (ppb) 1 < 3.0 < < 3.0 < < 3.0 < < 3.0 < < 3.0 < < 3.0 < < < 3.0 < < 3.0 < sites 1 site To few data points to statistically compare sites within watershed

18 4 Bayou de Chien 3 2 Land Cover (percent) Cultivated Forest Pasture Developed Other HUC Site poultry Graves houses County Sample Site Estradiol (ppt) 1 < 3.0 < < 3.0 to 3.6 < < 3.0 < < 3.0 to 3.5 < Fluoroquinolones (ppb) 2 sites 0 site To few detects to make statistical comparisons between sampling sites.

19 Clarks River Land Cover (percent) Cultivated Forest Pasture Developed Other Sample Site Estradiol (ppt) Fluoroquinolones (ppb) 1 < 3.0 to 3.6 < < 3.0 to 2.8 < < 3.0 to 3 < to < 3.0 to 3.2 < < 3.0 < < 3.0 < Murray outfall 4 sites 1 site To few detects to make statistical comparisons between sampling sites.

20 Floyds Fork Land Cover Percent Forest 43 Pasture 32 Developed 18 Cultivated 5 Other 2 Jeffersontown outfall Sample Site Estradiol (ppt) Fluoroq (ppb) 1 < 3.0 < < 3.0 < < 3.0 < < 3.0 < < 3.0 < < 3.0 < < 3.0 to 2.9 < < 3.0 to 7.8 < < 3.0 < < 3.0 < < 3.0 < < 3.0 to 7.2 < to < 3.0 to 2.9 < < 3.0 to 8.2 < to < 3.0 < < 3.0 to 3.9 < to < 3.0 < < 3.0 < < 3.0 < sites 3 sites

21 Floyds Fork 17-β Estradiol Fluoroquinolones MDL MDL FF08 FF12 FF14 FF16 FF12 FF14 FF16 Kaplen-Meier method used to generate boxplots. Based on MLE test p = 0.2 Based on MLE test p = 0.08 Can state there are differences at the 90% confidence level. Therefore, cannot state that the 4 sites are statistically different.

22 South Elkhorn Creek Land Cover Percent Pasture 63 Developed Forest 8 Cultivated 4 Other Sample Site Estradiol (ppt) Fluoroq (ppb) 1 < 3.0 to 3.8 < < 3.0 to 15.3 < to < 3.0 to 2.9 < < 3.0 to 8.5 < to < 3.0 to 3.7 < <3.0 < < 3.0 to 5.2 < < 3.0 to 4.4 < to < 3.0 < < < sites 3 sites Lexington outfall

23 South Elkhorn Creek 17-β Estradiol Fluoroquinolones MDL MDL SE02 SE04 SE05 SE08 SE02 SE04 SE08 Kaplen-Meier method used to generate boxplots Based on MLE test p = Can state there are statistical differences between sites at 95% confidence level. Based on MLE test p = Can state there are statistical differences between sites at 95% confidence level.

24 e-04 1e-03 1e-02 1e-01 Watershed Comparison 17-β Estradiol Fluoroquinolones MDL MDL BC CR FF SE CR FF SE Based on MLE test p = 0.02 Can state there are statistical differences between watersheds at 95% confidence level. Based on MLE test p = 0.13 No statistical differences between watersheds.

25 Conclusions ELISA kit for 17-β Estradiol and Fluoroquinolones is an adequate screening method. Estradiol was detected in all 6 watersheds. Fluoroquinolones were detected in 4 of the 6 watersheds. Watersheds with the most developed land cover South Elkhorn Creek (24%), Banklick Creek (36%), and Floyds Fork (18%) had the highest estradiol concentrations 14.3, 11.1, and 7.8 ppt, respectively. To few detects in 4 of the 6 watersheds makes it difficult to analyze data statistically. Using the nondects, a handful of sites in the Floyds Fork and South Elkhorn Creek watersheds could be analyzed statistically.

26 Future Research Screen additional watersheds using ELISA method. Collect additional surface water samples and analyze for specific compounds. Collect samples from karst springs. Use passive samplers in streams and karst springs to calculate time-weighted averages. Collect groundwater samples from water wells and screen using ELISA method and analyze for specific compounds.

27 Acknowledgements Funding USDA-NIFA Southern Regional Water Program KyWRRI Student Enhancement Grant (USGS 104B Program) Kentucky Geological Survey Laboratory Services Jason Backus and Andrea Mitchell ArcMap Land Cover Data Dan Carey, KGS retired; Bethany Overfield, KGS

28 Karst