Kathy R. Echols

Considerations involved with the use of semipermeable membrane devices for monitoring environmental contaminants.

Semipermeable membrane devices (SPMDs) are used with increasing frequency, and throughout the world as samplers of organic contaminants. The devices can be used to detect a variety of lipophilic chemicals in water, sediment/soil, and air. SPMDs are designed to sample nonpolar, hydrophobic chemicals. The maximum concentration factor achievable for a particular chemical is proportional to its octanol-water partition coefficient. Techniques used for cleanup of SPMD extracts for targeted analytes and for general screening by full-scan mass spectrometry do not differ greatly from techniques used for extracts of other matrices. However, SPMD extracts contain potential interferences that are specific to the membrane-lipid matrix. Procedures have been developed or modified to alleviate these potential interferences. The SPMD approach has been demonstrated to be applicable to sequestering and analyzing a wide array of environmental contaminants including organochlorine pesticides, polychlorinated biphenyls, polycyclic aromatic hydrocarbons, polychlorinated dioxins and dibenzofurans, selected organophosphate pesticides and pyrethroid insecticides, and other nonpolar organic chemicals. We present herein an overview of effective procedural steps for analyzing exposed SPMDs for trace to ultra-trace levels of contaminants sequestered from environmental matrices.

Use of the Semipermeable Membrane Device as an in Situ Sampler of Waterborne Bioavailable PCDD and PCDF Residues at Sub-Parts-per-Quadrillion Concentrations

Semipermeable membrane devices (SPMDs) were used to passively sample aqueous polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) in Bayou Meto, AR. The two sites were upstream and downstream from the confluence with a tributary that delivers PCDDs and PCDFs to the Bayou. Following dialysis, cleanup, and fractionation, four replicate 17-g SPMD samples from each site were analyzed by GC/MS, and four were evaluated by H411E bioassay. Traces of only OCDD and HpCDDs were detected in samples from the upstream site. The four samples from below the tributary contained averages of 1550 ± 80 pg of 2,3,7,8-TCDD, 1640 ± 80 pg of 2,3,7,8-TCDF, and lesser quantities of other congeners. The TCDD equivalents obtained by bioassay of replicate SPMD samples agreed well with results obtained by GC/MS. The quantities of 2,3,7,8-TCDD and 2,3,7,8-TCDF sequestered by SPMDs at the downstream site were used to estimate the aqueous concentrations for both compounds as 2 pg/L.

Polychlorinated biphenyls (PCBs) and organochlorine pesticides (OCPs) in harbor sediments from Sea Lots, Port-of-Spain, Trinidad and Tobago

Concentrations of polychlorinated biphenyls (PCBs) and organochlorine pesticides (OCPs) were determined in nearshore marine surficial sediments from three locations in Trinidad. Sediments were sampled at Sea Lots on the west coast, in south Port-of-Spain Harbor, south of Sea Lots at Caroni Lagoon National Park, and on Trinidads east coast at Manzanilla. Total PCB concentrations in Sea Lots sediments ranged from 62 to 601ng/g (dry weight {dw}), which was higher than at Caroni and Manzanilla, 13 and 8ng/g dw, respectively. Total OCP concentrations at Sea Lots were ranged from 44.5 to 145ng/g dw, compared with 13.1 and 23.8n/g (dw), for Caroni and Manzanilla respectively. The concentrations of PCBs and of some OCPs in sediments from Sea Lots were above the Canadian interim sediment quality guidelines. To date, this data is the first report on the levels of PCBs and other organochlorine compounds from Trinidad and Tobago.

Loading capacity and chromatographic behavior of a porous graphitic carbon column for polychlorinated biphenyls

Abstract A porous graphitic carbon column (Hypercarb) was used for the fractionation of polychlorinated biphenyls (PCBs) into classes of 2-4 ortho chlorines, 1 ortho chlorine and 0 ortho chlorine congeners. A method was developed that combined the fractionation of PCBs, polychlorinated dibenzo-p-dioxins and dibenzofurans in a variety of biotic environmental samples. Many of these samples have high concentrations of PCBs which cause fractionation problems as adsorption sites on the graphitic surface are occupied. The loading capacity of the column for PCBs was determined by injecting up to 1 mg of total PCBs and monitoring changes in chromatographic behavior of tetra-/di-ortho, mono-ortho and non-ortho substituted PCBs. Effective loading capacities were 1 mg for tetra-/di-ortho PCBs, but only 3–5 μg for non-ortho PCBs and about 2 μg for mono-ortho PCBs. Loading capacity of the PGC column for environmental fish and avian egg samples was determined to depend on the mono-ortho and non-ortho PCB levels found in these samples.

Polybrominated diphenyl ether metabolism in field collected fish from the Gila River, Arizona, USA-Levels, possible sources, and patterns

Polybrominated diphenyl ethers (PBDEs) were determined in fish collected from the Gila River, Arizona, a tributary of the Colorado River in the lower part of the Colorado River Basin. Fish samples were collected at sites on the Gila River downstream from Hayden, Phoenix, and Arlington, Arizona in late summer 2003. The Gila River is ephemeral upstream of the Phoenix urban area due to dams and irrigation projects and has limited perennial flow downstream of Phoenix due to wastewater and irrigation return flows. Fifty PBDE congeners were analyzed by high resolution gas chromatography/high resolution mass spectrometry using labeled surrogate standards in composite samples of male and female common carp (Cyrpinus carpio), largemouth bass (Micropterus salmoides) and channel catfish (Ictalurus punctatus). The predominant PBDE congeners detected and quantified were 47, 100, 153, 49, 28, and 17. Concentrations of total PBDEs in these fish ranged from 1.4 to 12700 ng g(-1) wet weight, which are some of the highest concentrations reported in fish from the United States. Differences in metabolism of several PBDE congeners by carp is clear at the Phoenix site; congeners with at least one ring of 2,4,5-substitution are preferentially metabolized as are congeners with 2,3,4-substitution.

Patterns of metal composition and biological condition and their association in male common carp across an environmental contaminant gradient in Lake Mead National Recreation Area, Nevada and Arizona, USA

There is a contaminant gradient in Lake Mead National Recreation Area (LMNRA) that is partly driven by municipal and industrial runoff and wastewater inputs via Las Vegas Wash (LVW). Adult male common carp (Cyprinus carpio; 10 fish/site) were collected from LVW, Las Vegas Bay (receiving LVW flow), Overton Arm (OA, upstream reference), and Willow Beach (WB, downstream) in March 2008. Discriminant function analysis was used to describe differences in metal concentrations and biological condition of fish collected from the four study sites, and canonical correlation analysis was used to evaluate the association between metal and biological traits. Metal concentrations were determined in whole-body extracts. Of 63 metals screened, those initially used in the statistical analysis were Ag, As, Ba, Cd, Co, Fe, Hg, Pb, Se, Zn. Biological variables analyzed included total length (TL), Fultons condition factor, gonadosomatic index (GSI), hematocrit (Hct), and plasma estradiol-17β and 11-ketotestosterone (11kt) concentrations. Analysis of metal composition and biological condition both yielded strong discrimination of fish by site (respective canonical model, p<0.0001). Compared to OA, pairwise Mahalanobis distances between group means were WB<LVB<LVW for metal concentrations and LVB<WB<LVW for biological traits. Respective primary drivers for these separations were Ag, As, Ba, Hg, Pb, Se and Zn; and TL, GSI, 11kt, and Hct. Canonical correlation analysis using the latter variable sets showed they are significantly associated (p<0.0003); with As, Ba, Hg, and Zn, and TL, 11kt, and Hct being the primary contributors to the association. In conclusion, male carp collected along a contaminant gradient in LMNRA have distinct, collection site-dependent metal and morpho-physiological profiles that are significantly associated with each other. These associations suggest that fish health and reproductive condition (as measured by the biological variables evaluated in this study) are influenced by levels of certain metals in the Lake Mead environment.

Novel associations between contaminant body burdens and biomarkers of reproductive condition in male Common Carp along multiple gradients of contaminant exposure in Lake Mead National Recreation Area, USA

Adult male Common Carp were sampled in 2007/08 over a full reproductive cycle at Lake Mead National Recreation Area. Sites sampled included a stream dominated by treated wastewater effluent, a lake basin receiving the streamflow, an upstream lake basin (reference), and a site below Hoover Dam. Individual body burdens for 252 contaminants were measured, and biological variables assessed included physiological [plasma vitellogenin (VTG), estradiol-17β (E2), 11-ketotestosterone (11KT)] and organ [gonadosomatic index (GSI)] endpoints. Patterns in contaminant composition and biological condition were determined by Principal Component Analysis, and their associations modeled by Principal Component Regression. Three spatially distinct but temporally stable gradients of contaminant distribution were recognized: a contaminant mixture typical of wastewaters (PBDEs, methyl triclosan, galaxolide), PCBs, and DDTs. Two spatiotemporally variable patterns of biological condition were recognized: a primary pattern consisting of reproductive condition variables (11KT, E2, GSI), and a secondary pattern including general condition traits (condition factor, hematocrit, fork length). VTG was low in all fish, indicating low estrogenic activity of water at all sites. Wastewater contaminants associated negatively with GSI, 11KT and E2; PCBs associated negatively with GSI and 11KT; and DDTs associated positively with GSI and 11KT. Regression of GSI on sex steroids revealed a novel, nonlinear association between these variables. Inclusion of sex steroids in the GSI regression on contaminants rendered wastewater contaminants nonsignificant in the model and reduced the influence of PCBs and DDTs. Thus, the influence of contaminants on GSI may have been partially driven by organismal modes-of-action that include changes in sex steroid production. The positive association of DDTs with 11KT and GSI suggests that lifetime, sub-lethal exposures to DDTs have effects on male carp opposite of those reported by studies where exposure concentrations were relatively high. Lastly, this study highlighted advantages of multivariate/multiple regression approaches for exploring associations between complex contaminant mixtures and gradients and reproductive condition in wild fishes.

Contaminants assessment in the coral reefs of Virgin Islands National Park and Virgin Islands Coral Reef National Monument

Coral, fish, plankton, and detritus samples were collected from coral reefs in Virgin Islands National Park (VIIS) and Virgin Islands Coral Reef National Monument (VICR) to assess existing contamination levels. Passive water sampling using polar organic chemical integrative samplers (POCIS) and semi-permeable membrane devices found a few emerging pollutants of concern (DEET and galaxolide) and polynuclear aromatic hydrocarbons. Very little persistent organic chemical contamination was detected in the tissue or detritus samples. Detected contaminants were at concentrations below those reported to be harmful to aquatic organisms. Extracts from the POCIS were subjected to the yeast estrogen screen (YES) to assess potential estrogenicity of the contaminant mixture. Results of the YES (estrogen equivalency of 0.17-0.31 ng/L 17-β-estradiol) indicated a low estrogenicity likelihood for contaminants extracted from water. Findings point to low levels of polar and non-polar organic contaminants in the bays sampled within VICR and VIIS.

Organochlorine contaminants in double-crested cormorants from Green Bay, WI: I. Large-scale extraction and isolation from eggs using semi-permeable membrane dialysis

A 41.3-kg sample of double-crested cormorant (Phalacrocorax auritus) egg contents was extracted, yielding over 2 L of egg lipid. The double-crested cormorant (DCC) egg extract, after clean-up and concentration, was intended for use in egg injection studies to determine the embryotoxicity of the organic contaminants found within the eggs. Large-scale dialysis was used as a preliminary treatment to separate the extracted contaminants from the co-extracted sample lipids. The lipid was dialyzed in 80×5 cm semi-permeable membrane devices (SPMDs) in 50-ml aliquants. After the removal of 87 g of cholesterol by freeze-fractionation, the remaining lipid carryover (56 g) was removed by 100 routine gel permeation chromatography (GPC) operations. A 41,293-g sample was thus extracted and purified to the extent that it could easily be placed at a volume of 5 ml, the volume calculated to be necessary for the egg injection study. Analyses were performed comparing contaminant concentrations in the final purified extract to those present in the original egg material, in the extract after dialysis and cholesterol removal, and in the excluded materials. Recoveries of organochlorine pesticides through dialysis and cholesterol ranged from 96% to 135%. Total polychlorinated biphenyls in the final extract were 96% of those measured in the original egg material. Analysis of excluded lipid and cholesterol indicated that 92% of the polychlorinated dibenzo-dioxins and-furans were separated into the final extract.

Metals in sediments and fish from Sea Lots and Point Lisas Harbors, Trinidad and Tobago.

Concentrations of heavy metals were determined in nearshore marine sediments and fish tissue from Sea Lots area on the west coast, at Caroni Lagoon National Park, and in the Point Lisas harbor, Trinidad. The most dominant metals found in sediments were Al, Fe and Zn with mean concentrations highest at Sea Lots (Al-39420 μg/g; Fe-45640 μg/g; Zn-245 μg/g), when compared to sediments from Point Lisas (Al-11936 μg/g; Fe-30171 μg/g; Zn-69 μg/g) and Caroni (Al-0400 μg/g; Fe-19000 μg/g; Zn-32 μg/g), High concentration of Cu, Al, Fe and Zn were also detected in fish tissue from Point Lisas and Caroni. Metal concentrations in fish tissue showed significant correlation with sediment metals concentration, which suggests that tissue levels are influenced by sediment concentration. Of the metals, only Zn, Hg and Cu had a bioaccumulation factor (BAF) greater than one, which suggests a high bioaccumulation potential for these metals.