Uncertainty Analysis for an Equilibrium Partitioning–based Estimator of Polynuclear Aromatic Hydrocarbon Bioaccumulation Potential in Sediments

Butcher

Enviro Toxic and Chemistry

2011

The lower Passaic River, New Jersey, USA, a tidal tributary to Newark Bay and part of the New York and New Jersey Harbor Estuary, is contaminated with a variety of persistent organic pollutants (POPs) due to nearly two centuries of heavy industrialization and urbanization. Resident aquatic organisms are exposed to, and can bioaccumulate, a variety of chemical contaminants from sediments, water and other organisms. In the present study, the relationships between selected POPs detected in both surface sediments and aquatic organisms are examined statistically along with the efficacy of using empirical biota-sediment accumulation factors (BSAFs) to describe such relationships. Regression analyses were conducted on synoptic surface sediment POP data (0 to 15 cm in depth) and whole-body tissue POP data for three prominent organisms that reside in the river and which are important components of the food web: mummichog (Fundulus heteroclitus), blue crab (Callinectes sapidus), and white perch (Morone americana). Per the equilibrium partitioning theory on which the BSAF model is based, surface sediment data were normalized to total organic carbon and tissue data were normalized to lipid content for each organism. Normalized surface sediment concentrations were generally poorly correlated with normalized biota tissue concentrations. The BSAF model was not found to be a reliable means to predict concentrations of POPs in select lower Passaic River organisms, using surface sediment chemistry data.

Section: Discussionsupporting

confidence: 92%

Section: Introductionmentioning

confidence: 99%

Evaluation of potential relationships between chemical contaminants in sediments and aquatic organisms from the lower Passaic River, New Jersey, USA

Butcher

Enviro Toxic and Chemistry

2011

Enviro Toxic and Chemistry

“…This observed variation between individuals could be due to the differences of organisms' physiological conditions and heterogeneity of sediments [37]. Variability of lipid content has long been a concern in predicting risks associated with bioaccumulation of organic chemicals due to the hydrophobic nature of persistent organic compounds [39]. However, our study suggests that the uncertainty in measurements could also have significant impact on the prediction results.…”

Section: Discussionmentioning

confidence: 87%

“…Different degrees of uncertainty and variability of model parameters were observed between individuals. Variability of lipid content has long been a concern in predicting risks associated with bioaccumulation of organic chemicals due to the hydrophobic nature of persistent organic compounds [39]. Based on the allometric relationship listed in Hendriks et al [31], the variability of computed k d is relatively small compared with the predicted results.…”

Section: Discussionmentioning

confidence: 99%

A bayesian approach to parameter estimation for a crayfish (Procambarus spp): Bioaccumulation model

Lin

Berzins

Myers

et al. 2004

Bioaccumulation models are used to describe chemical uptake and clearances by organisms. Averaged input parameter values are traditionally used and yield point estimates of model outputs. Hence, the uncertainty and variability of model predictions are ignored. Probabilistic modeling approaches, such as Monte Carlo simulation and the Bayesian method, have been recommended by the U.S. Environmental Protection Agency to provide a quantitative description of the degree of uncertainty and/or variability in risk estimates in ecological hazards and human health effects. In this study, a Bayesian analysis was conducted to account for the combined uncertainty and variability of model parameters in a crayfish bioaccumulation model. After a 5-d exposure in the LaBranche Wetlands (LA, USA), crayfish were analyzed for polycyclic aromatic hydrocarbon concentrations and lipid fractions. The posterior distribution of model parameters were derived from the joint posterior parameter distributions using a Markov chain Monte Carlo approach and the experimental data. The results were then used to predict the distribution of chrysene concentration versus time in the crayfish to compare the predicted ranges at the different study sites.

Engineering in Life Sciences

“…Once input distributions are quantified, values for each input variable and model parameter are randomly drawn from their probability distributions using, for example, a Monte Carlo approach, preserving correlations among the inputs if applicable. The resulting combination of input values is used in the model (e.g., sediment fate and transport, contaminant flux) to generate a realization [45][46][47]. The process is repeated, while the nature of the sampling ensures that the realizations are equally probable, allowing a statistical summary of the model outputs.…”

Section: Uncertainty Modeling and Decision Making: Deterministic And mentioning

confidence: 99%

Scaling Methods of Sediment Bioremediation Processes and Applications

Adriaens

Michalak

2006

Bioremediation has been argued to be one of the most cost-effective remediation technologies available to reduce soil, sediment, or groundwater contamination, particularly because this approach may allow for the implementation of in-place strategies. Recent trends have advocated the application of innovative sediment stabilization strategies through placement of (reactive) capping material to allow long-term biodegradation of contaminants in these complex biogeochemical environments. The potential long-term risk reduction associated with this approach requires a demonstration of causal relationships between sediment or contaminant stability on the one hand, and microbial reactivity on the other. The spatial analysis needed to fully understand and quantify these correlations requires sensitive probabilistic techniques. Geostatistics has been used for the characterization of multi-scale spatial patterns for the last few decades, and the analysis of microbial attributes has shown significant spatial structures on microbial abundance and activity. However, there is a dearth of information on the applicability of geostatistics to quantitatively describe the interaction between the microorganisms and their environment. Using the Passaic River (NJ) dioxin data as a model dataset, multiple scaling models were applied to scale and interpolate sampled dioxin data and derive dechlorination signatures in sediments. Unlike conventional geostatistic tools that are based on the pointto-point spatial structures, the new multi-scale model (M-Scale) introduces a new framework for spatial analysis in which regional values at different scales are anchored by the correlations to each other. Spatial dioxin distributions and microbial dechlorination signatures were used as benchmarks for comparison of M-Scale to ordinary kriging. The results from cross-validation and jackknifing approaches applied to these datasets were analyzed and compared using Quantile-Quantile (Q-Q) plots and reproduction coefficients. These plots indicated that the M-Scale better preserves the local features of hotspots during data interpolation to a basin-wide scale. Current efforts focus on mapping microbial abundance and respiratory competence in the Anacostia River, based on measurements at three different scales. The outcomes of this work will be used to develop an uncertainty-based spatial decision tool for site remediation in this watershed using various capping strategies.