Water, Air, and Soil Pollution

, 198:45

Fecal Sterol and Bile Acid Biomarkers: Runoff Concentrations in Animal Waste-Amended Pastures

Authors

    • Department of Plant and Soil Sciences, College of AgricultureUniversity of Kentucky
  • Dwayne R. Edwards
    • Department of Biosystems and Agricultural Engineering, College of AgricultureUniversity of Kentucky
  • Mark S. Coyne
    • Department of Plant and Soil Sciences, College of AgricultureUniversity of Kentucky
Article

DOI: 10.1007/s11270-008-9824-7

Cite this article as:
Tyagi, P., Edwards, D.R. & Coyne, M.S. Water Air Soil Pollut (2009) 198: 45. doi:10.1007/s11270-008-9824-7

Abstract

Nonpoint source pollution is the leading remaining cause of water quality problems. The extent of NPS pollution is often more difficult or expensive to monitor at the point(s) of origin, as compared to monitoring of point sources. This study evaluated the hypothesis that animal manure (chicken, cow, horse, and pig) applied to pasture contribute fecal sterols and bile acids to runoff. The study also assessed the potential benefit of fecal sterols and bile acids as biomarkers in distinguishing fecal pollution and its sources. Fecal sterol and bile acid concentrations were determined in flow-weighted composite runoff samples collected from 2.4 × 6.1 m plots (n = 3) amended with manure. Runoff was generated from simulated rainfall (152 mm.h−1). Runoff samples from manure-amended plots showed high concentrations of fecal sterol (ranged from 13 ± 1 to 1,287 ± 183) and bile acid (ranged from 24 ± 1 to 2,251 ± 248) biomarkers. The profiles of fecal sterols and bile acids in runoff samples were similar to those of fresh manure for all selected animals. For runoff and fresh manure, chenodeoxycholic acid, deoxycholic acid, epicoprostanol, and hyodeoxycholic acid were consistent biomarkers for chicken, cow, horse, and pig, respectively, suggesting that sterols and bile acids can be used to identify sources and occurrence of fecal matter in water and sediments.

Keywords

GC–MSLiquid–liquid extractionManureFecal bile acidsFecal sterolsRun-off

1 Introduction

Nonpoint source pollution introduces impurities into a surface-water body or an aquifer, usually through a non-direct route and from sources that are “diffuse” in nature. Discharges from nonpoint sources are usually intermittent, associated with water washing over the land, whether from rain or irrigation. This runoff picks up an array of contaminants including agricultural chemicals from farmland, and nutrients and toxic materials from urban and suburban areas. Pollution arising from nonpoint sources accounts for most of the contaminants in streams and lakes. To control nonpoint sources, it is important to identify the fecal pollution source so mitigation efforts can be effective. Understanding the origin of fecal pollution is paramount in assessing associated health risks.

Molecular markers (fecal sterols and bile acids) can be used as quantitative indicators of fecal presence (Venkatesan et al. 1986; Venkatesan and Kaplan 1990; Elhmmali et al. 1997, 2000; Maldonado et al. 2000; Bull et al. 2002, 2003; Tyagi et al. 2008). Because of their unique structure, fecal sterol and bile acids can often be linked to a single specific origin. However, such an approach has only rarely been applied to animal manure from the perspective of the origin of fecal matter in water or soil. Jarde et al. (2007) reported that sterol/stanol compounds could be potentially useful as specific molecular markers of pig slurry. Leeming et al. (1996) reported that the source-specificity of fecal sterols is caused by a combination of three main factors. First is the animal's diet. Each diet has a different sterol profile (i.e. different compounds present or in different amounts) so the proportions of sterol precursors entering the digestive tract are different. Second, even with low dietary intake of sterols, endogenous sterols are biosynthesized by higher animals and discharged to the digestive tract. Finally, and perhaps most importantly, anaerobic bacteria in the digestive tract of some animals bio-hydrogenate sterols to stanols of various isomeric configurations.

Haslewood (1967) reported that bile acids are excreted in the feces and varies from species to species for example; pigs produce 6-hydroxylated acids which is not present in human feces. Bile acids are characterized by a carboxylic acid group at the C23 position on the steroidal carbon side chain and a 3a- or 3b-hydroxyl group on the A ring (like sterols). Normal human feces contain more than 20 different bile acids that are formed from the primary bile acids, cholic and chenodeoxycholic acid (Bull et al. 2002). The primary bile acids, chenodeoxycholic acid (CDOCA) and cholic acid (CA) are formed in the liver from cholesterol and secreted with the bile to the intestine. The microorganisms present in the intestine transform primary bile acids (CDOCA and CA) to secondary bile acids. Most of the secondary bile acids are absorbed in the intestine and returned to the liver with only a small but significant fraction excreted in the feces. Bile acids can be used to distinguish a ruminant or nonruminant source of fecal contamination. This is possible because ruminant animals (e.g., bovines) produce predominantly deoxycholic acid whereas nonruminants (e.g., canines and humans) produce significant quantities of lithocholic acid. Lithocholic acid and deoxycholic acid are the major secondary bile acids excreted in the feces of humans and some higher animals. The absence of deoxycholic acid and the presence of hyocholic acids in porcine fecal material enable it to be distinguished from human and canine contamination (Setchell et al. 1983; Evershed and Bethell 1996; Simpson et al. 1999; Chaler et al. 2001).

To our knowledge, the contribution of sterol and bile acid biomarkers has not yet been studied for livestock runoff. The first objective of this study was to analyze the abundance of these molecular markers in simulated runoff samples collected from experimental plots treated with different animal manures (cow [Bos taurus], pig [Sus scrofa], horse [Equus caballus], and chicken [Gallus domesticus]). The second objective was to assess the potential benefit of these biomarkers by determining the fecal sterol and bile acid profiles for the simulated livestock runoffs.

2 Materials and Methods

2.1 Experiment-Design

To determine the runoff concentrations of fecal sterols and bile acids, a study was conducted in August 2006, using fifteen plots constructed on a Maury silt loam (fine, mixed, mesic Typic Paleudalf) at the University of Kentucky Agricultural Experiment Station. Plot dimensions were 2.4 × 6.1 m (major axis oriented up- and down-slope) with a uniform 3% slope along the major axis and cross-leveled across the minor axis. Each plot was bordered with galvanized iron (10 cm above and below ground surface) to isolate runoff. A gutter (made with two pieces of lightweight metal) was constructed with a 5% slope to ensure gravity flow and installed across the lower end of each plot to concentrate runoff for measurement and sampling. Runoff from the gutter entered a 5 cm inside diameter (ID) length of polyvinyl chloride pipe (PVC) and emptied approximately 45 cm above the bottom of a sump. Runoff was sampled as it exited the PVC pipe and before contact with the interior of the sump (Busheé et al. 1998). The vegetation for all plots was Kentucky-31 “tall” fescue (Festuca arundinacea Schreb).

The experiment consisted of five treatments with three replicates in a completely randomized design. The five treatment variables were cow manure, pig manure, horse manure, chicken litter, and a control (no manure). The fresh manure/litter was obtained from facilities located on the University of Kentucky Animal Research Center, near Lexington, Kentucky. Chicken litter was collected in plastic bags containing 15 to 20 kg of litter each, and manure (cow, horse, and pig) was collected in plastic buckets containing 15 to 20 kg of manure each. Samples of the fresh manure/litter were obtained from each bag or bucket and transported to the laboratory for analysis of water content and sterol and bile acid concentrations. The remainder of the manure/litter was applied to the surfaces of the plots manually with careful attention to the uniformity of application. The gross application rate of manure/litter was 10.09 Mg.ha−1, selected as roughly equivalent to typical N uptake (0.11 Mg N.ha−1) for fescue (Natural Resources Conservation Service (NRCS) 1992).

2.2 Runoff Sampling

Manure was allowed to sit for only 60 min on each plot before rain simulation. Three rainfall simulators were used to supply the water used to generate runoff. Each simulator was capable of applying 0 to 152 mm.h−1 simulated rainfall to one 2.4 × 6.1 m plot. A programmable logic controller interfaced with a computer controlled spray frequency. The simulated rainfall intensity was 152 mm.h−1, applied within 60 min of manure/litter application and maintained until 30 min of runoff had occurred from each plot. Total rainfall duration therefore differed between plots, but runoff duration was constant. Seven runoff samples were collected (approximately 1 L sample size) at 2, 4, 6, 8, 16, 24, and 32 min after the beginning of runoff. Runoff samples were collected by inserting a clean glass container (1 L volume) beneath the runoff exiting the gutter through the PVC pipe. Runoff entered the container for one min or until the container was filled, whichever came first. The time required to collect the samples was measured to compute runoff rate and volume. The runoff rate data were used to construct a single flow-weighted composite sample per plot from the associated seven discrete runoff samples. The runoff samples were then refrigerated at −20°C until analyzed for sterol and bile acid concentrations.

2.3 Chemicals

Sterol reference standards (coprostanol, epicoprostanol, cholesterol, cholestanol, stigmasterol, and stigmastanol) and internal standards (5α-cholestane and nor-cholic acid) were purchased from Steraloids Inc. (http://www.steraloids.com). Bile acid standards (lithocholic acid, deoxycholic acid, ursodeoxycholic acid, cholic acid, chenodeoxycholic acid, hyodeoxycholic acid), derivatization agent N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA), HPLC grade hexane, chloroform, and methanol were purchased from Sigma (http://www.sigma-aldrich.com). The reference sterol and bile acid standard solutions of 100 mg.L−1 were prepared in DCM–isopropanol (2:1, v/v) and stored at 0°C. Anhydrous sodium sulfate (Na2SO4), baked at 500°C, was used in the extraction process for water removal.

2.4 Calibration

5α-cholestane and nor-cholic acid were used as internal standards because these compounds (5α-cholestane and nor-cholic acid) were not present in the feces of the tested animals. To optimize GC–MS conditions, 100 mg.L−1 of standard solution of the reference sterols and bile acids was prepared in DCM–isopropanol (2:1, v/v) and stored at 0°C.

2.5 Sterols and Bile Acids Extraction, Saponification, and Derivatization

2.5.1 Feces

Sterol and bile acid concentrations in the feces samples were determined using the method published earlier (Tyagi et al. 2008). 25–50 mg of sub-samples were taken for the total lipid extraction, to which 40 μg of internal standards (nor-cholic acid and 5α-cholestane) were added. Total lipids were extracted with 2:1 (v/v) chloroform: methanol and then the solvent were removed under a gentle stream of nitrogen. After saponification step, sterols and bile acids were extracted in the hexane and ethyl acetate respectively. The solvent was then evaporated to dryness and the residue was ready for derivatization by reacting with 100 μl of BSTFA: Pyridine (1:1 v/v). A 1 μl sample was injected into a gas chromatograph–mass spectrometer to perform the chromatographic analyses of sterols and bile acids.

2.5.2 Runoff

In this study, a slightly modified liquid–liquid extraction method (Nash et al. 2005; Peng et al. 2005) was used to analyze the fecal sterol and bile acid concentrations in runoff samples.

A 250 ml volume of runoff sample was transferred to a separatory funnel together with 40 mg.L−1 of 5α-cholestane and 40 mg.L−1 of nor-cholic acid internal standards. A 20 ml volume of chloroform–methanol (2:1, v/v) and 20 ml of DCM–isopropanol (2:1, v/v) were added to the sample, which was shaken for 30 min on a wrist action shaker at 600 rpm and then left overnight. An additional 20 ml of DCM–isopropanol (2:1, v/v) were added to separate the two phases, and the sample was shaken for another 30 min on a wrist action shaker at 600 rpm. The lower organic phase was collected in an Erlenmeyer flask on which a filtration funnel was mounted. The funnel contained glass wool and about 5 g of anhydrous sodium sulfate to remove any water content. The extraction procedure was repeated three times (without standing overnight) with 20 ml of DCM–isopropanol (2:1, v/v) solvent, and the pooled DCM–isopropanol phase was evaporated to dryness at 45°C under a gentle stream of nitrogen. The sample residue was then saponified with 5 ml of 2 mol.L−1 NaOH at 90°C for 1 h. After cooling at room temperature, sterols were extracted three times with 5 ml of n-hexane, and the pooled organic phase (upper phase) was evaporated to dryness at 55°C under a gentle stream of nitrogen. The residue was then ready for derivatization. After extraction of sterols, 10 ml of distilled water was added to the aqueous solution, which was then acidified to pH 2–3 with 2 mol.L−1 HCl, and the bile acids were recovered in ethyl acetate (upper phase). The solvent was removed under a gentle stream of nitrogen at 60°C, and the residue was ready for derivatization. Sterol and bile acid fractions were converted to their trimethylsilyl (TMS) derivatives by treating with 100 μl of BSTFA: Pyridine (1:1, v/v) and allowed to stand at 60°C for 30 min. A 1 μl sample was injected into a gas chromatograph–mass spectrometer (GC–MS) to perform the chromatographic analyses of sterols and bile acids. A laboratory blank (distilled water) and a calibration standard (preferably middle point of the curve) was always analyzed with the samples.

2.6 Instrumentation and GC–MS Condition

Gas chromatographic analysis was performed with a Varian CP 3800 gas chromatograph (Varian Inc., CA, USA) equipped with a 30-m DB5 (0.25-mm i.d., 0.25-μm film thickness) fused silica capillary column and a split/splitless injector. The injector temperatures for sterol and bile acid compounds were 250°C and 280°C, respectively. A one μl sample was injected (autosampler—CombiPal autosampler, CTC analytics) in the splitless mode. The initial oven temperature was 150°C and 100°C for sterols and bile acids, respectively. Oven temperature for bile acids began at a lower temperature (100°C) than the sterols (150°C) to allow for better separation on the column. After one min, the oven temperature was increased to 250°C at 12°C per min, then to 300°C at 3°C per min. The latter temperature was maintained four min. Helium was the carrier gas at a flow rate of 1 ml.min−1.

The mass spectrometer (ion-trap mass spectrometer, Varian-Saturn 2200) (Varian Inc., CA, USA) was configured for electron impact ionization at 70 eV, 200°C source temperature, 0.49 s.scan−1, with a scan range of 210–600 amu. Data acquisition and processing were carried out using a Saturn-view workstation (5.52) data system. To identify the selected compounds, a combination of GC–MS data (namely retention time, ion ratio, and a unique ion (a quantitative ion per compound) was used. Verification of compounds of interest was accomplished by using two qualifier ions and the percentage of the relative ratios (±20%) compared to the quantification ion. Gas chromatographic and mass spectrometric details viz. retention time, quantitative ions, qualifier ions, ions ratio and percentage of the relative ratios of the qualifier ions for all interested compounds were determined (Table 1), which demonstrate suitable sterol and bile acid separation under the chromatographic conditions used in the analysis.
Table 1

Gas chromatographic–mass spectrometric details for sterols and bile acids

 

Retention time (min)

Relative retention timeb

Quantitative ion

Qualifier ion

Relative ratio of qualifier ionc (%)

1st

2nd

1st

2nd

Sterols

 5α-cholestane (IS)a

16.691

1.000

357

217

125.1

 Coprostanol

19.611

1.175

370

215

257

37.3

27.1

 Epicoprostanol

20.040

1.201

370

215

257

70.8

29.1

 Cholesterol

21.070

1.262

368

329

353

64.6

47.5

 Cholestanol

21.266

1.274

445

370

403

79.0

45.0

 Stigmasterol

23.216

1.391

394

484

379

77.7

39.4

 Stigmastanol

24.445

1.465

383

473

398

73.0

61.7

Bile acids

 Nor-cholic acid (IS)a

25.886

1.000

412

343

502

70.0

61.4

 Lithocholic acid

27.326

1.056

430

415

505

65.3

29.2

 Deoxycholic acid

27.632

1.067

593

428

266

68.7

39.6

 Cholic acid

27.814

1.074

426

343

516

118.3

43.8

 Chenodeoxycholic acid

28.136

1.087

428

323

413

15.3

41.5

 Hyodeoxycholic acid

28.514

1.102

428

255

413

289.5

63.4

 Ursodeoxycholic acid

28.879

1.116

518

428

413

48.7

31.8

aInternal standard

b(Retention time of the component)/(Retention time of the internal standard)

cRelative ratio of the qualifier ion = (area of the qualifier ion/area of the quantitative ion) × 100

Nine calibration standards were prepared and analyzed containing each of the sterol or bile acid compounds in a range of concentrations (0.5, 1, 5, 10, 20, 40, 60, 80 and 100 mg.L−1) with the appropriate internal standard added to each at 40 mg.L−1. Calibration curves were constructed by using the ratio of the peak area of the quantification ion of each compound to that of the internal standard (5α-cholestane or nor-cholic acid). Sample concentrations were determined by comparing the ratios of the quantification ion areas to that of the calibrated curves. The calibration curves for each sterol and bile acids demonstrated high linearity (r2 = 0.980). The lowest calibration standard (0.5 mg.L−1) was considered the limit of quantification (LOQ) for the analysis. Representative chromatograms illustrating sterol and bile acid profiles for runoff samples are in Figs. 1 and 2, which demonstrate suitable sterol and bile acid separation under the chromatographic conditions used in the analysis.
https://static-content.springer.com/image/art%3A10.1007%2Fs11270-008-9824-7/MediaObjects/11270_2008_9824_Fig1_HTML.gif
Fig. 1

Sterol representative ion chromatograms obtained by GC–MS analysis for the chicken, cow, horse, pig and control runoff samples. (Peak numbers are ISa α-cholestane, 1 coprostanol, 2 epicoprostanol, 3 cholesterol, 4 cholestanol, 5 stigmasterol, 6 stigmastanol)

https://static-content.springer.com/image/art%3A10.1007%2Fs11270-008-9824-7/MediaObjects/11270_2008_9824_Fig2_HTML.gif
Fig. 2

Bile acid representative ion chromatograms obtained by GC–MS analysis for the chicken, cow, horse, pig, and control runoff samples. (Peak numbers are ISb norcholic acid, 7 lithocholic acid, 8 deoxycholic acid, 9 cholic acid, 10 chenodeoxycholic acid, 11 hyodeoxycholic acid, 12 ursodeoxycholic acid)

3 Results and Discussion

3.1 Feces

Description of sterol and bile acid profiles for the selected animals (chicken, cow, horse, and pig) were explained in our earlier published work (Tyagi et al. 2008). In brief, the major sterols detected in pig feces were coprostanol, cholesterol, and stigmastanol; while for horse feces, epicoprostanol and stigmastanol were the dominant sterols. For cows, the major fecal sterols were coprostanol and cholesterol, while the chicken feces were distinguished by cholesterol and stigmastanol. The coprostanol concentration of chicken feces was comparable to other sterols. Our results are in contrast with a study of chicken feces (Leeming et al. 1996) that found coprostanol to be either absent or present at very low levels. Variations observed for the sterol profiles of chickens in present and other studies (Leeming et al. 1996) might be due to different diet compositions. In the USA, chicken diets are composed of mostly corn and soybean meal plus some minor ingredients and vitamins and minerals. While in some countries in Europe, ingredients such as barley, wheat, peas, rapeseed meal, sunflower meal, etc. are used in addition to corn and soybean meal. Therefore, a major difference is that some European diets may have a greater concentration of non-starch polysaccharides (which are not digested well). This, in turn, can affect the microflora in the intestinal tract of the chickens.

Major bile acids observed for pig feces were hyodeoxycholic acid and lithocholic acid, the high abundance of hyodeoxycholic acid was not observed for other chosen animals, enabling a clear distinction to be drawn between pigs and other animals. The prominent bile acids observed for cow and chicken feces were deoxycholic acid and chenodeoxycholic acid respectively. None of the bile acids was observed to be present in distinctive concentrations in horse fecal samples.

3.2 Runoff

To check the precision and accuracy of the method, nine parallel sets of runoff samples (three each for control, cow, and pig) were spiked by adding the 40 mg.L−1 of reference sterol and bile acid standards and analyzed by the method described in this paper. Representative chromatograms showing sterol and bile acid profiles for spiked and non-spiked runoff samples are shown in Figs. 3 and 4. Recovery data for the sterols and bile acids are shown in Table 2. The mean recovery was between 88.9% and 94.1%, and the variation was less than 8.7% for all sterols and bile acids.
https://static-content.springer.com/image/art%3A10.1007%2Fs11270-008-9824-7/MediaObjects/11270_2008_9824_Fig3_HTML.gif
Fig. 3

Sterol representative ion chromatograms obtained by GC–MS analysis for spiked and non-spiked control runoff samples. Runoff samples were spiked with 40 mg.L−1 of reference sterol standards. (where peak numbers are ISa α-cholestane, 1 coprostanol, 2 epicoprostanol, 3 cholesterol, 4 cholestanol, 5 stigmasterol, 6 stigmastanol)

https://static-content.springer.com/image/art%3A10.1007%2Fs11270-008-9824-7/MediaObjects/11270_2008_9824_Fig4_HTML.gif
Fig. 4

Bile acid representative ion chromatograms obtained by GC–MS analysis for spiked and non-spiked control runoff samples. Runoff samples were spiked with 40 mg.L−1 of reference bile acid standards. (where peak numbers are ISb norcholic acid, 7 lithocholic acid, 8 deoxycholic acid, 9 cholic acid, 10 chenodeoxycholic acid, 11 hyodeoxycholic acid, 12 ursodeoxycholic acid)

Table 2

Percentage recovery of the sterols and bile acids from spiked runoff samples (%) (n = 9)

Sterols

Mean ± S.E.

Bile acids

Mean ± S.E.

Coprostanol

93.3 ± 5.0

Lithocholic acid

93.3 ± 2.8

Epicoprostanol

90.8 ± 3.8

Deoxycholic acid

93.0 ± 5.5

Cholesterol

88.9 ± 4.3

Cholic acid

94.1 ± 4.6

Cholestanol

90.5 ± 4.9

Chenodeoxycholic acid

89.3 ± 8.7

Stigmasterol

90.6 ± 6.8

Hyodeoxycholic acid

93.6 ± 3.5

Stigmastanol

92.1 ± 5.4

Ursodeoxycholic acid

92.0 ± 5.2

Three each for control, cow, and pig feces samples

Table 3 summarizes the reference sterols and bile acids mean concentrations for the runoff samples (chicken, cow, horse, pig, and control). Runoff from the control treatments showed very low concentrations of sterols (1–9 mg.L−1) and none of the bile acids (representative chromatograms are shown in Figs. 3 and 4). Runoff concentrations of sterols/bile acids from the control plots were very low or not detectable; it suggests the experimental plots used in the study were not contaminated with the referred fecal biomarkers. Results in Table 3 shows that the concentrations for some of the sterols and bile acids varied widely (by factors of 2–2,251) among treatments (chicken, cow, horse, and pig). Runoff concentrations of major sterols and bile acids from pig manure-treated plots were detected in the following order: hyodeoxycholic acid (1,827–2,685 mg.L−1)> stigmastanol (942–1,567 mg.L−1)> coprostanol (910–1,195 mg.L−1) > cholesterol (346–632 mg.L−1)> lithocholic acid (268–455 mg.L−1). The most abundant fecal sterols in the runoff for the horse manure-treated plots were epicoprostanol (154–191 mg.L−1) and stigmastanol (238–265 mg.L−1). Epicoprostanol was observed two–117 times higher for horse-treatment runoff than from the pig, cow, chicken, and control treatments.
Table 3

Sterol and bile acid content in the runoff samples (mg.L−1)

 

Control

Chicken

Cow

Horse

Pig

Coprostanol

1 ± 0

288 ± 19

353 ± 28

117 ± 12

1,088 ± 90

Epicoprostanol

1 ± 0

13 ± 1

34 ± 7

177 ± 12

119 ± 20

Cholesterol

9 ± 1.7

528 ± 26

303 ± 64

57 ± 3

476 ± 84

Cholestanol

1 ± 0.33

39 ± 1

142 ± 28

29 ± 3

290 ± 54

Stigmasterol

5 ± 0.33

151 ± 15

24 ± 6

136 ± 19

198 ± 88

Stigmastanol

2 ± 0.33

590 ± 6

86 ± 25

248 ± 9

1,287 ± 183

Lithocholic acid

ND

66 ± 4

67 ± 13

89 ± 7

364 ± 54

Deoxycholic acid

ND

35 ± 2

147 ± 15

95 ± 7

25 ± 0.33

Cholic acid

ND

56 ± 4

36 ± 1

26 ± 1

24 ± 1

Chenodeoxycholic acid

ND

214 ± 20

ND

32 ± 1

45 ± 2

Hyodeoxycholic acid

ND

ND

ND

ND

2,251 ± 248

Ursodeoxycholic acid

ND

34 ± 1

ND

ND

31 ± 2

Mean ± S.E. (n = 3)

ND not detected

Runoff samples from the cow manure-treated plots showed concentrations of dominant fecal sterols and bile acids in the following order: coprostanol (297–388 mg.L−1) > cholesterol (199–421 mg.L−1) > deoxycholic acid (125–175 mg.L−1). However, runoff samples from the chicken litter-treated plots were distinguished by noticeable concentrations of cholesterol (495–579 mg.L−1), stigmastanol (582–601 mg.L−1), and chenodeoxycholic acid (184–252 mg.L−1). In the study, at least one major sterol or bile acid biomarker, viz. chenodeoxycholic acid, deoxycholic acid, epicoprostanol, and hyodeoxycholic acid, were observed from the runoff of chicken, cow, horse, and pig manure-treated plots, respectively. Chemical profiles (mean ± S.E. values) of the sterol and bile acid biomarkers from the runoff treatments (chicken, cow, horse, and control) are shown in Fig. 5. In the Fig. 5, the standard error values clearly describe that the sterol and bile acid concentrations were consistent (no significant variation observed) among single animal species when compared to the mean concentrations of the referred sterol or bile acid biomarker. It was observed that many of the biomarkers mentioned above are not unique to just one animal or another, but the overall distribution of sterols and bile acids is specific for each animal species.
https://static-content.springer.com/image/art%3A10.1007%2Fs11270-008-9824-7/MediaObjects/11270_2008_9824_Fig5_HTML.gif
Fig. 5

Chemical profiles of sterols and bile acids (mean values) for the runoff of their different treatments (chicken, cow, horse, pig and control). Error bars indicates the variation in the concentrations between samples, if multiple samples were analyzed. (where numbers are 1 coprostanol, 2 epicoprostanol, 3 cholesterol, 4 cholestanol, 5 stigmasterol, 6 stigmastanol, 7 lithocholic acid, 8 deoxycholic acid, 9 cholic acid, 10 chenodeoxycholic acid, 11 hyodeoxycholic acid, 12 ursodeoxycholic acid)

The sterol and bile acid biomarker profiles between the runoff and the feces samples for the selected animals (chicken, cow, horse, and pig) are shown in Fig. 6. A logarithm scale was used to visually accommodate the sterol and bile acid concentrations that may differ by orders of magnitude. Although fairly large concentration differences were observed between the runoff and the feces samples of their potential source animals, interestingly, the ratios among the sterol and bile acid biomarkers were reasonably consistent between the runoff and the feces samples for all animals. Abundance of the sterol and bile acid biomarkers in runoff samples were similar to the fresh manure of their source animals, which indicates that their detection could be a good surrogate indicator of water contamination by these animal wastes.
https://static-content.springer.com/image/art%3A10.1007%2Fs11270-008-9824-7/MediaObjects/11270_2008_9824_Fig6_HTML.gif
Fig. 6

Comparison of algorithm values for the sterol and bile acid concentrations between runoff and feces samples for the selected animals. (where numbers are 1 coprostanol, 2 epicoprostanol, 3 cholesterol, 4 cholestanol, 5 stigmasterol, 6 stigmastanol, 7 lithocholic acid, 8 deoxycholic acid, 9 cholic acid, 10 chenodeoxycholic acid, 11 hyodeoxycholic acid, 12 ursodeoxycholic acid)

4 Conclusions

Runoff samples obtained from manure applied plots showed high sterol and bile acid concentrations, even if the application rate of manure/litter was equal to typical N fertilization (0.11 Mg N.ha−1) for all plots. Detection of fecal sterol and bile acid biomarkers in the runoff samples permit us to consider the possibility of using them as a tool in determining fecal pollution sources, suggesting further utility in the context of agriculture runoff and streams. The overall distribution of sterol and bile acid biomarkers were specific for each animal species. However, additional study is still needed to investigate the transport of these biomarkers and others (e.g., 5β-campestanol, 5β-epicampestanol, and 5β-epistigmastanol which are common among herbivores) into water bodies.

Acknowledgement

We acknowledge the Kentucky Agricultural Experiment Station, University of Kentucky, USA for funding this work; the individuals who collected the fecal and runoff samples, and the associated departments for allowing the fecal samples to be part of this work. John May (lab technician) of ERTL, UKY is thanked for his technical assistance in using GC–MS.

Copyright information

© Springer Science+Business Media B.V. 2008