Correlations Between Water Quality Parameters and Levels of 4-NP in Water and Sediment of Stroubles Creek Watershed Aaron Bradner Project Report Submitted to the Faculty of the Virginia Polytechnic Institute and State University In Partial Fulfillment of the Requirements for the Degree of Master of Science In Crop and Soil Environmental Sciences Dr. Kang Xia, Co-chair Dr. Vinod Lohani, Co-chair Dr. Carl Zipper, Committee Member February 2013 Blacksburg, VA
28
Embed
Correlations Between Water Quality Parameters and - LEWAS Lab
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Correlations Between Water Quality Parameters and Levels of 4-NP in Water and Sediment of Stroubles Creek Watershed
Aaron Bradner
Project Report Submitted to the Faculty of the Virginia Polytechnic Institute and State University
In Partial Fulfillment of the Requirements for the Degree of
Master of Science In
Crop and Soil Environmental Sciences
Dr. Kang Xia, Co-chair Dr. Vinod Lohani, Co-chair
Dr. Carl Zipper, Committee Member
February 2013 Blacksburg, VA
1
ABSTRACT
This observational study examined the potential correlation of different water quality
parameters with concentration of the organic contaminant 4-nonylphenol (4-NP) in the bed sediment
and water column of Stroubles Creek, a tributary of the New River in southwest Virginia. Weekly sample
collection occurred over the course of approximately four months (Jun 25-Oct 15, 2012). On Monday
mornings two water samples and a sediment sample were collected from threes sites along Stroubles
Creek to be analyzed for 4-NP. Three water quality sondes maintained by Virginia Tech’s Biological
Systems Engineering Department on lower Stroubles were deployed continuously and recorded five
parameters (dissolved oxygen, pH, conductivity, turbidity, and temperature) every 15 minutes a few feet
upstream of the 4-NP sample collection sites. These data were compiled into a weighted weekly average
using an exponential decay curve, giving more weight to the most recent days before sample collection.
Analyzed along with the weekly average were the weekly minimum (min) and weekly maximum (max) of
each parameter. These max, min, and average values for all 5 parameters at each of the three sites were
compared to the concentration of 4-NP detected in the soil and water samples collected weekly from
each site to determine if there was any significant correlation (r > .30) between these parameters with
4-NP concentration.
The following observations were made for ln(4-NP water concentration): temperature
(maximum; positive correlation), turbidity (maximum and weighted average; both positive correlation),
and dissolved oxygen (minimum % saturation, weighted average % saturation, weekly minimum, weekly
weighted average; all negative correlation) were significantly correlated.
The following observations were made for ln(4-NP sediment concentration): temperature
Fig 2. Example table of data downloaded from sonde at site 3 on first day of deployment.
9
4-NP sample collection
Over the course of the summer and fall semester, two students in Dr. Xia’s lab drove through
the agriculture fields to the furthest sampling point downstream and collect two water samples
(~500mL) and one sediment sample (~500mL). The sediment samples consisted of finer organic
sediment near to the bank rather than gravelly sediment that clearly washed off the road. These
samples were stored in glass Mason jars with 5 drops of hydrochloric acid and stored on ice. The student
moved upstream to the next collection site, following the same sampling procedure, until all sampling
sites had been visited.
Prior to sampling, one student gathered 10 pre-cleaned Mason jars (2 water samples and 1
sediment sample for each of the 3 sites, then one overall field blank). Labels were prepared in lab and
applied to the jars. A second student collected ice from Latham hall and fill a cooler about 1/3 full with
ice. A small kit with necessary equipment (a trowel for collecting sediment samples, paper towels, a
solution of 50:50 methanol/water for decontaminating the trowel between samples, a small dropper-
topped bottle containing concentrated HCl for fixing the samples, and Nitrile gloves) was also taken to
the sampling site. The students then proceeded to Site 3, the furthest site downstream so as to prevent
false readings from disturbing upstream sediment.
At the site, three Mason jars and the kit of supplies were taken to the site from the truck, which
was parked along the dirt road through the agriculture field. One student took water samples from the
center of the creek. The samples were preserved immediately after acquisition by placing 5 drops of
concentrated HCl into each sample. A second student took sediment samples from the edge of the
stream bed but still under the water. This sample was also fixed with 5 drops of concentrated HCl. The
equipment and samples were carried back to the truck and the samples were put on ice in the cooler.
The trowel used for sediment sample collection was washed with the Methanol/Water solution to
remove residual debris. The students then proceeded to the next site and repeat the same method for
the three remaining sites. Once all samples were collected, the students returned to lab and stored the
samples in a freezer to prevent degradation before analysis.
4-NP Water Sample Extraction
Samples were thawed and analyzed within 48 hours to prevent degradation of analytes. To
extract 4-NP from the water samples, 200mL of each sample was measured in a graduated cylinder
washed with ultrapure water. The sample container was then washed out with ultrapure water and the
measured sample was poured from the graduated cylinder back into the rinsed container. The
graduated cylinder was washed again with ultrapure water and 200mL of the next sample was
10
measured. The second container was washed with ultrapure water and the sample was poured back
from the graduated cylinder into the rinsed container. This was repeated until 200mL of all the samples
had been measured.
The solid-phase extraction (SPE) filter was washed before filtering of samples with 3mL of 10%
methanol (MeOH) solution. Once the MeOH had run through the filter 3mL ultrapure water was run
through the filters. The SPE tubes were filled with 3mL of sample at a time and gravity filtered until the
rate of filtration slowed down enough to warrant using the vacuum filter (usually around 100mL). The
vacuum was turned on to filter the remainder of the sample for the sake of saving time.
Once the samples were completely filtered, the filters were rinsed again with 6mL of a 5%
MeOH solution and 6mL of ultrapure water. They were then dried with the vacuum on for 30 min. Once
dry, test tubes were placed under the filtration tubes and the filters were washed with a 70
dichloromethane: 30 acetone solution to wash the 4-NP in the filter into the tests tubes.
The eluted 4-NP was placed under a flow of pure nitrogen. This allowed for evaporation of liquid
while keeping the 4-NP in the test tube. Once dry, 500uL of MeOH was added and the tubes were
vortexed to dissolve the dry 4-NP in the MeOH. This mixture was placed in the gas chromatograph.
4-NP Sediment Sample Extraction
An arbitrary amount (>4g) of each sediment sample was placed in a vial and dried under vacuum
for 4 days. Once dry, 2 grams of each sample were weighed out and placed in a clean vial. 10 mL of a 1:1
hexane/acetone mixture was added to each sample and the vials were sonicated for 20 minutes.
Meanwhile, silica gel for dehydrating the samples was prepared by measuring 1g SiO2 and 1g Na2SO4 and
combining in one vial.
Once sonication was complete, the vials were centrifuged for 10 minutes at 3500 rpm and 25C.
At the end of centrifugation, the liquid in each vial was pipetted off the top and was processed through
the SPE. The same process for cleaning and filtering the water sample was followed for the sediment
sample.
Gas Chromatograph Analysis
Gas chromatograph-mass spectrometry analysis was performed using an Agilent 790A gas
chromatograph along with an Agilent 7000 series triple quadrupole GC/MS detector and an Agilent 7693
autosampler. The initial column of the GC/MS/MS was held at 60oC for 0.5 min, increased to 100oC at
15oC per minute, further increased to 200oC at 5oC per minute, and finally increased to 280oC at 25oC per
minute. The backflush time is 3.0188min making the whole time for a cycle 26.867min. The inlet
temperature was set at 200oC, transfer line temperature was set at 250oC, and the ion source
11
temperature was set at 200oC. The analytical column is HP-5MS (30 m × 0.250 mm i.d., 0.25um film
thickness, Agilent, USA) and with a backflush column of Agilent res. (0.78m×150 um ×0 um). Splitless
mode was used at a Helium gas flow rate of 2.25 mL per minute. The injection volume was 1 uL and the
collision gas used was nitrogen. The x compound was qualified by electron impact at 70 eV using
multiple reaction monitoring (MRM) mode. The MS/MS quantification and confirmation ions are m/z
(107+121+135+149). According to the elute pattern of the isomers, four time segments were set in the
MRM method.
The results from the GC-MS are displayed as chromatograms. Technical grade 4-NP was used as
external standard for the qualification and the quantification of total 4-NP. The graph is created as
counts vs acquisition time in minutes. In order to quantify the concentration of the 4-NP in each sample,
the peak area is determined using the program from 19 to 21.5 minutes. In figure 2, this is shown by the
shaded area under the peaks within this range. Since 4-NP has many isomers, the different peaks
represent the different isomers all of which are important to include when determining the
concentration of 4-NP in the samples.
The area of the peaks is used in conjunction with the calibration curve to determine the 4-NP
concentration in the final extract. The concentration of the final extractant in the 2mL GC vial multiplies
the extraction concentration factor of the corresponding result in 4-NP concentration in the water or
sediment sample. (Moutinho, 2012)
Fig 3. Chromatogram of the 400 μg/L standard run on July 3, 2012. Source: Moutinho, 2012
Data processing
12
Data was processed using Microsoft Excel and JMP. The water concentration data from each site
was averaged by site; the sediment concentration was not averaged as only one sample was collected.
The weekly parameter data from the three sites was averaged from one Monday through the following
Monday. The minimum, maximum, and weighted average was determined for each week as well in
order to coincide with the 4-NP sampling dates, which occurred on Monday mornings. Sonde data was
separated by site and week so that one sheet of data would contain one week (12PM Monday-11:45AM
the following Monday) of data from any given site (one, two, or three). This allowed for faster
processing of weekly average, weighted weekly average, weekly minimum, and weekly maximum.
Finally, the concentration data from each site was compared to the parameter data from each site.
The generation of a model for exponential decay was discussed with LISA so that each point
would have an increasingly greater factor in determining weekly averages, giving a total weight of one
for all points. The model that was produced was (y=9x10-6ex). This exponential decay, based on data
from Topp and Starratt (1999), gives a total weight of 1 to stream parameters collected over the entire
week while giving more weight parameters measured closer to the collection dates. Because 4-NP is
removed quite quickly from the stream in a matter of days, any parameters being compared to its
concentration were not expected to have as significant an impact as more recent parameter
measurements and were therefore weighted accordingly in the average. The minimum and maximum
values were also included to determine threshold values for partitioning and solubility.
Due to the strongly negative skewness of the raw data samples, a ln-normal plot was used to
provide a more normalized distribution for parametric analysis. Using JMP, the 4-NP concentration data
was converted into a ln-normal plot for both the water column and the bed sediment. Each value (min,
max, average, weighted average) of each parameter (dissolved oxygen, pH, conductivity, turbidity,
temperature) was plotted against the ln of both the water and sediment concentrations. Residuals were
also calculated.
13
Figure 4. Weighted average curve developed by to give more weight to more recently collected parameters, based off data from previous literature (Topp and Starratt 1999).
STATISTICAL ANALYSIS
Weekly analysis comparison: With the assistance of statistical collaborators from LISA
(Laboratory for Interdisciplinary Statistical Analysis), a program that assists students on the Virginia Tech
campus with statistical analysis of project data, the weekly sonde parameter weighted averages were
compared to the weekly 4-NP sample concentration and analyzed for correlation between any of the
parameters and 4-NP concentration.
Each measurement (maximum, minimum, average, and weighted average) of each parameter
(DO, % saturation DO, turbidity, conductivity, pH, and temperature) was then analyzed for correlation in
JMP using a scatterplot matrix. Correlation coefficient values greater than .30 were considered slightly
correlated, while any values greater than .50 were considered strongly correlated (personal
communication, LISA and Dr. Holtzman). This process was used as a preliminary test to determine which
parameters had at least some degree of correlation.
These parameters found to have a correlation coefficient greater than .30 were then analyzed
non-parametrically using Spearman’s rank correlation coefficient. Spearman’s coefficient was chosen to
analyze the correlation between ln(concentration) and parameter values for its ability to measure
statistical dependence between two continuous variables in addition to being less sensitive to strong
outliers that were prevalent in this study. These Spearman rho values were used to verify correlation.
Once non-parametric correlation between parameters and ln(concentration) had been verified, the
residual error of the linear correlation was plotted to show that it was evenly distributed against the ln-
Figure 8. Normalized ln(4-NP water concentration) distribution.
The distribution indicates that low concentrations are fairly common, with higher
concentrations becoming less concentration as concentrations increase. The ln-normal histograph
indicates that the ln of the concentration is slightly more normally distributed.
Below is an example of the graphs that were created during statistical analysis using the
comparison of ln(water concentration) vs. max temperature. Similar tests were run on all parameters;
following is a table of values for all parameters found to be significantly correlated in the scatterplot
matrix.
16
Figure 9. Scatterplot matrix correlation plot of ln(water conc) vs. temp max.
Figure 10. Plotted residuals of temp max vs. water concentration, showing normalized distribution. Mean residual error =1.015e-16.
17
Figure 11. Plot of ln(water conc) vs. temp max. The equation was found to be: ln (water conc)=-3.108683 + 0.0745811*tempMax
Figure 12. Correlation values, residual values, and equations for significantly correlated ln(water conc) parameters. Note that values determined to be not significant are not included in this table.
Figure 13. Correlation values, residual values, and equations for significantly correlated ln(sediment conc) parameters. Note that values determined to not be significant are not included in this table.
The weekly measured water parameters that were found to be correlated (r > .30) with ln(4-NP
water concentration) were maximum temperature (positive), maximum turbidity (positive), weighted
average turbidity (positive), minimum percent dissolved oxygen (negative), weighted average percent
Scatterplot correlation (p=.10) Spearman's rho Mean residual error Equation
Temp max 0.3158 0.7278 1.02E-16 ln (water conc)=-3.108683 + 0.0745811*tempMax
The weekly measured water parameters found to be correlated with ln(4-NP sediment
concentration) were minimum temperature (positive), maximum temperature (positive), average
temperature (positive), weighted average temperature (positive), maximum turbidity (positive),
minimum percent dissolved oxygen (negative), minimum dissolved oxygen (negative), maximum
dissolved oxygen (negative), and average dissolved oxygen (negative).
DISCUSSION
From reviewing previously published literature the observed correlations of ln(4-NP water
concentration)with temperature (positive), turbidity (positive), and rainfall (positive) were expected.
The increase in temperature would allow for better mixing between a lipophilic chemical and water
(Cornellisen et al., 1997), while greater rainfall would transport more 4-NP from its source into the
stream through surface runoff (Sole et al., 2000) and increased turbidity would allow for desorption of 4-
NP from fine sediment (Gao et al., 1998; Liber et al., 1999). The overall negative correlation with
dissolved oxygen was also expected as well based on the work of Faust and Holgne (1987). However, DO
is affected by many other stream parameters such as temperature and rainfall. Further testing should be
performed to rule out confounding variables.
Based on the aforementioned dissolved oxygen studies (Pellizetti et al., 1989; Faust and Holgne,
1987; Yu et al., 2003), the observed negative correlation with ln(4-NP sediment concentration) was
expected. However, the other significantly correlated parameters did not correlate with ln(4-NP
sediment concentration) in the expected manner. Temperature and turbidity both showed a positive
correlation with ln(4-NP sediment concentration), the same as ln(4-NP water concentration). This would
indicate that as temperature and turbidity increase 4-NP increases in both sediment and water, or
causes an overall net gain in the stream rather than partitioning from one matrix to another. From
previously published literature it would have been expected that as water temperatures increased, 4-NP
would desorb from fine particulate sediment more readily and decrease concentration within sediment
(Cornellisen et al., 1997). However, each temperature value measured indicated that increasing
temperature increased 4-NP concentration in sediment. Although this correlation may be caused by a
confounding extraneous variable such as runoff caused by rainfall, further research is suggested.
The positive correlation with turbidity may be explicable by surface runoff of 4-NP from its
source during a rain event (Sole et al., 2000) which would also coincide with increases in turbidity.
19
Although a source has yet to be determined for Stroubles Creek, it is safe to say that Stroubles Creek
itself does not produce synthetic chemicals like 4-NP on its own. Rather rain events cause washing from
the as-yet undetermined source into Stroubles, increasing overall net 4-NP concentration in the stream.
However, again further research would be suggested.
It was expected that decreases in pH during rain events would have a positive correlation with
water 4-NP concentration and a negative correlation with sediment 4-NP concentration. However, due
to the brevity of the changes (typically a .2 unit decrease over <2 hours before recovering) with respect
to the amount of time the stream was monitored (168 hours in a week), it is reasonable that no
correlation would have been noticed. Additionally, because no samples were taken during a rain event
but rather after the stream pH had sufficient time to return to its normal pH, it is reasonable that pH
would not appear to correlate. Finally, because the stream pH is constantly below the pKa of 4-NP, there
would not be much of a change in chemical properties between base flow and rain events. Regardless,
in future studies samples should be collected on shorter time scales, specifically during rain events, to
ascertain the effect acute decreases in pH may have on concentration.
Although runoff would bring more suspended sediment and higher DO, the correlation between
weekly averages of both at all three sites is negative. At high flow, DO is higher and turbidity is lower.
This may be due to the time of year during which data was collected: warm rain events would raise the
temperature from suburban runoff into the stream, decreasing DO rapidly.
Figure 14. A comparison of DO and turbidity shows a negative correlation, with higher DO during low flows rather than high as would normally be expected. This may be due to spikes in temperature during rain events in the summer months during which data were collected.
Concentration over course of collection did not appear to show any significant trends. There are
a number of outliers that could potentially be linked to rain events; however, due to lack of complete
20
rainfall data, this could not be determined, The outliers show no relevant correlation to other
parameters and may be caused by leaching from the source. Further investigation into the source of the
4-NP found in the stream would be informative.
Figure 15. Concentration of 4-NP in sediment samples over duration of sample collection.
Figure 16. Concentration of 4-NP in water samples over duration of sample collection.
Future Work
Because of a restriction of resources, samples were only collected on a weekly basis to ease
scheduling and reduce the use of resources. Future studies should include sampling during high flows to
determine short-term, acute effects as opposed to just chronic, week-long effects. Because all of the
parameters change to some degree for such a short time during rain events they have little impact on
the weekly average which may affect the correlation of some of the data. For example, pH may show
21
strong correlation during a rain event due to sharp decreases during rain events; because no samples
were collected during rain events, any correlation (or lack thereof) could not be observed. DO showed
significant negative correlation in both water and sediment samples; it may show stronger negative
correlation during cold winter rain events that raise DO, or weaker correlation during hot summer rain
events that lower DO. Although this is only conjecture it could be concluded more definitely in a future,
more thorough study.
This observational study occurred over the course of two different seasons (begin early summer,
end mid fall), which is significantly longer than any studies performed in the literature review. These
seasonal changes were unaccounted for in the final analysis that covered the entire study. These
seasonal changes can cause unaccounted variations such as: organic carbon input, water temperature,
rainfall amount, and DO, among others. Accounting for these seasonal differences would require an
increase in resources such as the ability to measure organic carbon. No calendar season was recorded in
entirety, which would make comparing one season to another inaccurate at best without a complete set
of coinciding data.
Rainfall data was obtained from Virginia Tech’s BSE department, but only ran for one month of
data collection and so was not analyzed. Few rain events occurred during the short period of available
rainfall data, which makes determination of correlation difficult. However, there are a number of
outliers present in the data that may well be due to rain events. At the collection points in the stream
rain events are the only likely cause for changes in parameters. In future endeavours it would be worth
obtaining continuous rainfall data, perhaps from the town of Blacksburg which would be more complete
than the one rain gauge at BSE site two.
There were a number of possible factors that could affect the concentration of 4-NP that were
not monitored in this study. As discussed in the literature review, organic carbon concentration in the
sediment was not accounted for nor was is standardized among the sites. The precipitation temperature
was also not recorded for the duration of this study; however, the temperature of the rain may
contribute to the transport of 4-NP from its source to the stream and therefore should be monitored in
the future.
It is possible that some equipment utilized in the acquisition of parameter data was not
functioning properly: for example, the BSE sondes located along the stream indicated a conductivity of
400 ms at the first site, a sudden drop to 0 ms at the second site, and an increase back to 390 ms at the
third site. This happened every week over the course of the study, at random intervals each week lasting
from a few hours to a few days. These readings were consistent, and the second site conductivity data
22
was not included in the final analysis. Additionally, one sonde reported an average turbidity of 200 NTU
which is unlikely except in the event of a weeklong storm, or with multiple readings of very high
turbidity. The rain data indicated there was only 13mm of rain during the week this average was
recorded, implying the possibility of obstructed equipment which is a possibility. These data points were
not excluded during analysis because they were infrequent, and multiple communications with BSE
assured that the equipment is maintained and calibrated regularly.
ACKNOWLEDGMENTS
I thank my committee members Dr. Xia, Dr. Lohani, and Dr. Zipper for their support and
guidance through this project, and for taking the time to read revisions and offer suggestions to help
improve this report. I also thank the members of Dr. Xia’s lab: Terri Sosienski, Paul Parker, Christiana,
and Jennifer Moutinho for performing the sediment and water sample collection and analysis. I thank
my statistics team Andy Hoegh and Amy Till for their patience with me in explaining statistics and their
invaluable assistance with the data analysis. Thanks to Dr. Ingrid Lee for assistance with my literature
review and Dr. Mary Lee for providing suggestions on statistical analysis. Finally thanks to the BSE
department, in particular Dr. Cully Hession, Siavash Hoomehr, and the StREAM Lab for being willing to
work with me and for providing the water parameter data that was used in this observational study.
23
SOURCES
Brix R, Hvidt S, Carlsen L. 2001. Solubility of NP and NP ethoxylates on the possible role of micelles. Chemosphere 44:759-763. Cornellisen G, Van Noort PCM, Parsons JR, Govers HAJ. 1997. Temperature dependence of slow adsorption and desorption kinetics of organic compounds in sediments. Environmental Science and Technology 31:454-460. Dachs, J., D.A. Van Ry, and S.J. Eisenreich. 1999. Occurrence of estrogenic NPs in the urban and coastal atmosphere of the lower Hudson River estuary. Environmental Science and Technology 33:2676–2679. Ekelund R, Granmo A, Magnusson K, Berggren M. 1993. Biodegradation of 4-NP in seawater and sediment. Environmental Pollution 79:59-61. Faust, B.C., and J.Holgne . 1987. Sensitized photooxidation of phenols by fulvic acid and in natural waters. Environmental Science and Technology 21: 957–964. Gao JP, Maguhn J, Spitzauer P, Kettrup A. 1998. Sorption of pesticides in the sediment of the Teufelsweiher Pond (Southern Germany). I: Equilibrium assessments, effect of organic carbon content and pH. Water Research 32:1662-1672. Gao JP, Maguhn J, Spitzauer P, Kettrup A. 1997. Distribution of pesticides in the sediment of the small Teufelsweiher pond (southern Germany). Water Research 31:2811-2819. Kolpin, D.W., E.T. Furlong, M.T. Meyer, E.M. Thurman, S.D. Zaugg, L.B. Barber, and H.T. Buxton. 2002. Pharmaceuticals, hormones, and other organic wastewater contaminants in U.S. streams, 1999– 2000: A national reconnaissance. Environmental Science and Technology 36:1202– 1211. Leblanc GA, Xueyan M, Rider CV. 2000. Embryotixicity of the alkylphenol degradation product 4-nonylphenol to the crustacean Daphnia magna. Environmental Health Perspectives 108:1133-1138. Liber K, Knuth ML, Stay FS. 1999. An integrated evaluation of the persistence and effects of 4-NP in an experimental littoral system. Environmental Toxicology and Chemistry 18:357-362. Maguire, R.J. 1999. Review of persistence of NP and NP ethoxlates in aquatic environments. Water Quality Research Journal of Canada 34:37-78. McCormick SD, O’Dea MF, Moeckel AM, Lerner DT, Bjornsson BT. 2005. Endocrine disruption of parr-smolt ptransformation and seawater tolerance of Atlantic salmon by 4-NP and 17β-estradiol. General and Comparative Endocrinology 142:280-288. Miles-Richardson SR, Pierens SL, Nichols KM, Kramer VJ, Snyder EM, Snyder SA, Render JA, Fitzgerald SD, Giesy JP. 1999. Effects of Waterborne Exposure to 4-NP and NP Ethoxylate on Secondary Sex Characteristics and Gonads of Fathead Minnows (Pimephales promelas). Environmental Research 80:122-137.
24
Moutinho, J. 2010. Investigating the Occurrence and Fate of 4-NP in a Watershed Impacted by Urban Development. Pelizzetti, E., C. Minero, V. Maurino, A. Sclafani, H. Hidaka, and N. Serpone. 1989. Photocatalytic degradation of NP ethoxylated surfactants. Environ. Sci. Technol. 23:1380–1385. Pereira WE, Domagalski JL, Hostettler FD, Brown LR, Rapp JB. 1996. Occurrence and accumulation of pesticides and organic contaminants in river sediment, water and clam tissues from the San Joaquin River and tributaries, California. Environmental Toxicology and Chemistry 15:172-180. Pionke HB and Chesters G. 1973. Pesticide-Sediment-Water Interactions. Journal of Environmental Quality 2:29-45. Shang, D.Y., R.W. MacDonald, and M.G. Ikonomou. 1999. Persistence of NP ethoxylate surfactants and their primary degradation products in sediments from near a municipal outfall in the Strait of Georgia, British Columbia, Canada. Environmental Science and Technology 33:1366–1372. Sole ,M.,M.J.L.DeAlda,M. Castillo,C. Porte,K. Ladegaard-Pedersen, and D. Barcelo. 2000. Estrogenicity determination in sewage treatment plants and surface waters from the Catalonian area (NE Spain). Environmental Science and Technology 34:5076–5083. Thiele, B., K. Gunther, and M.J. Schwuger. 1997.Alkylphenol ethoxylates: Trace analysis and environmental behavior. Chemical Reviews 97: 3247–3272. Topp, E and Starratt, A. 1999. Rapid Mineralization Of The Endocrine-Disrupting Chemical 4-Nonylphenol In Soil. Environmental Toxicology and Chemistry 19:313-318. Turchi CS and Ollis DF. 1990. Photocatalytic degradation of organic water contaminants; mechanisms involving hydroxyl radical attack. Journal of Catalysis 122:178-192. Vazquez-Duhalt R, Marquez-Rocha F, Ponce E, Licea AF, Viana MT. 2005. NP, an integrated vision of a pollutant. Applied Ecology and Environmental Research 4:1-25. Vinten AJA, Yaron B, Nye PH. 1983. Vertical transport of pesticides into soil when adsorbed on suspended particles. Journal of Agriculture and Food Chemistry 31: 662-664. Voice TC, Weber WJ. 1983. Sorption of Hydrophobic compounds by sediments, soils and suspended solids I-theory and background. Water Research 17:1433-1441. Wauchope RD. 1978. The pesticide content of surface water draining from agricultural fields-a review. Journal of Environmental Quality 7:459-472. Weston DP, You J, Lydy MJ. 2004. Distribution and Toxicity of sediment-associated pesticides in agriculture-dominated water bodies of California’s central valley. Environmental Science and Technology 38:2752-2759.
25
Xia K, Jeong CY. 2004. Photodegradation of the Endocrine-Disrupting Chemical 4-NP in Biosolids Applied to Soil. Journal of Environmental Quality 33:1568-1574. Xie L, Thrippleton K, Irwin MA, Seimering GS, Mekebri A, Crane D, Berry K, Schlenk D. 2005. Evaluation of Estrogenic Activities of Aquatic Herbicides and Surfactants Using an Rainbow Trout Vitellogenin Assay. Toxicological Sciences 87:391-398. Yuan SY, Yu CH, Chang BV. 2003. Biodegradation of NP in river sediment. Environmental Pollution 127:425-430. Younos, T. Parece, T. DiBettito, S. Sprague, T. 2006. The Stroubles Creek Watershed: History of Development and Chronicles of Research. VWRRC Special Report No. SR48-2010. Personal communication, Dr. Golde Holtzman (October 2012). Personal communication, Andy Hoegh and Amy Till (November 2012).
26
APPENDIX
Figure A. Spearman rank correlation coefficient and associated p-values values for all ln(4-NP water conc) parameter correlations.
Figure B. Spearman rank correlation coefficient and associated p-values values for all ln(4-NP sediment conc) parameter correlations.
27
Figure C. Pearson rank correlation coefficient values for all parameters.
Figure D. Associated p-values of Pearson rank correlation coefficients for all parameters.