Chesapeake Ecotox Research Program: Research

Population Models

As the length of the life cycles of many commercially or ecologically important species preclude experimental approaches to documenting population level impacts of contaminant exposure, researchers will develop a range of population-level models to assess the impacts of contaminant exposure. The models will be used to explore the impacts of contaminant exposure to Chesapeake Bay fauna generally, and to interpret the outcome of mesocosm experiments involving exposure to contaminated sediments from the three Regions of Concern. The dynamics of populations exposed to contaminant stress will be examined in three model frameworks: (1) bioenergitics models; (2) Stage-based projection models and (3) individual-based models. Bioenergetics models will be developed using estimates derived from independent respirometry measurements of the surrogate species These models quantify the effects of changes in physiology resulting from contaminant stress on the partitioning of the energy ingested to maintenance, growth and reproduction (Rice 1990; Beyers et al. 1999a and b). In this application, the bioenergetic model will be used to predict growth of individual organisms, given estimates of their consumption.

Researchers will also develop stage-based projection models for all target organisms (Caswell 1989; Brewster-Geisz and Miller 1999) to quantify the relative sensitivity of population growth rate to exposure at different life stages. Stage-based models are demographic models that share similarities with the analysis of life tables, but represent the entire organism's life cycle by a small number of discrete stages that represent functional biological units (Caswell 1989). Thus, unlike the life table analysis that requires estimates for every year the organism lives, the stage-based model only requires estimates for each stage. Therefore, the realism of a many-staged model can be balanced with the precision of a simpler model when parameter estimation error is of concern. Caswell was one of the first to apply this model format to ecotoxicology (Caswell 1996). However, stage-based models have been widely used to explore the dynamics of marine and estuarine animals generally (polychaetes, Bridges and Heppell 1996, Levin et al. 1996; sea turtles, Crouse et al. 1987; Crowder et al. 1994; sharks, Brewster-Geisz and Miller 1999; blue crabs, Miller and Houde 1998; and anchovies, Pertierra et al. 1997). Central to the approach is the ability to formally estimate the sensitivity of the population growth rate to changes in the vital rates of life history stages within the organism's life cycle.

Individual-based models for fish species will use simulation approaches that track individual organisms throughout their life cycle (Van Winkle et al. 1993). By explicitly integrating the range of responses among individuals to exposure, these models allow quantification of the overall population level response (Madenjian et al. 1995). The model combines details of the bioenergetics with estimates of food availability and sources of mortality to predict both the number and characteristics of survivors. In the model, contaminant exposure can result from consumption of contaminated food items, or through direct contact with contaminated sediments. At each time step, and for every individual organism, the model assesses whether or not the organism feeds. If it feeds, the model calculates the net energy gain via simple bioenergetic rules. This surplus energy is partitioned between somatic and gonad growth. The model then evaluates whether the organism is vulnerable to sources of mortality and determines whether the it dies or survives. During these subroutines, the model will also track the body burden (total contaminants and specific concentration) in each individual organism. At the end of each time step, population level summary statistics are reported. The process is then repeated for the next time step. The Individual-based models will be used to estimate both the effect of sub-lethal concentrations of contaminants on population abundance and characteristics, and on the distribution of body-burdens within the population. Recently, researchers at Chesapeake Biological Laboratory have developed similar Individual-based models that track the fate of hydrophobic organic contaminants in a simplified food web in which fish are the top predators (Ashley 1999).