Biomedical subjects

David W Pennington

Publications and source records attributed to David W Pennington.

3 recordsLinked to original sources

PubMed2005

Multimedia fate and human intake modeling: spatial versus nonspatial insights for chemical emissions in Western Europe.

Multimedia fate and multipathway human exposure models are widely adopted in assessments of toxicological risks of chemical emissions at the regional scale. This paper addresses the question of how much spatial detail is necessary in such models when estimating the intake by the entire population in large, heterogeneous regions such as Europe. The paper presents a spatially resolved multimedia fate and multipathway exposure model for Western Europe, available as IMPACT 2002. This model accounts for relationships between the location of food production and drinking water extraction as well as where population cohorts live relative to where chemical emissions occur. The model facilitates estimation of environmental concentration distributions, related levels of contaminants in foods, and the fraction of a chemical release that will be taken in by the entire human population (the intake fraction) at the regional scale. To evaluate the necessary spatial resolution, the paper compares estimates of environmental concentrations and the intake fraction from the spatially resolved model with the results of a consistent clone without spatial resolution. An evaluation for disperse emissions of PeCDF (2,3,4,7,8-pentachlorodibenzofuran, CAS# 5120731-4) suggests reasonable agreement with monitoring data for most impact pathways with both versions of the model, but that the generic vegetation models for estimating contaminant concentrations in agricultural produce require improvement. A broader comparison for a range of organic chemicals demonstrates that the nonspatial models are likely to be appropriate in general for assessing dispersed sources of emissions. However, it is necessary to include generic compartments in such nonspatial models to account separately for emissions that enter lakes with long residence times versus rivers that feed directly into seas. For assessing an emission source in a specific location, using models that are not spatially resolved can result in underestimation, or overestimation, of the population's intake by at least 3 orders of magnitude for some chemicals.

Benzofurans↗

PubMed2004

Aquatic ecotoxicological indicators in life-cycle assessment.

This paper compares available options for the aquatic ecotoxicological effect factor component in life-cycle assessment (LCA). The effect factor is expressed here as the change in risk per unit change in cumulative exposure, delta effect/delta exposure. The comparison is restricted to approaches linked, implicitly as well as explicitly, to species-sensitivity distributions (SSDs). This draws on recent insights for chemical mixtures and identifies the implications of different model choices. In spite of the many options, assumptions, and areas for further research, it is concluded that a single effect factor basis represents the best available practice for use in LCA at this time, deltaPAF(ms)/deltaC = 0.5/HC50, where deltaPAF(ms) is the change in the (potentially affected) fraction (PAF) of species that experiences an increase in exposure above a specified effect level, accounting for the presence of complex background mixtures (ms), deltaC is the change in cumulative exposure concentration of the chemical of interest, and HC50 is the median, chronic hazardous concentration for regional, multiple-species systems. The resultant aquatic effect factors are risk-based and can be estimated readily for many chemicals using available methods, without the need to describe the entire SSDs and without the need for additional data. For example, the octanol-water partitioning coefficient provides a sufficient estimation basis for about 50% of existing chemicals that have a narcosis mode of action. This also is relevant in LCA for chemicals that are at low concentrations in the environment, concentrations below the biological thresholds at which more specific modes of action would be of relevance.

Animals↗

PubMed2003

Extrapolating ecotoxicological measures from small data sets.

Risk screening is commonly conducted using multiple-species ecotoxicological measures such as the HC5, the hazardous concentration at which 5% of species in a specified (eco-)system are assumed to be stressed. This article demonstrates that the estimate of HC5 will not vary significantly among commonly adopted parametric models of species sensitivity distributions. Uncertainty is highly dependent on the number of species tested (sample size) and the relevance of the measurement to the assessment endpoint (e.g., acute measures for assessing chronic endpoints). This article cross-compares estimates of these uncertainties using different empirical and theoretical methods to propose sample to population extrapolation factors. Some theoretical parametric methods for estimating the confidence intervals on the HC(5) can result in large overconservatism, particularly if positive bias reduces uncertainty. The 95th percentile confidence interval on the HC5 estimate given only three chronic test results varies from 5 to 8 x 10(8), depending on the estimation method adopted.

Animals↗