Characteristics of Available Models for Ecological Risk Assessment

The characteristics of selected models of populations, ecosystems, and landscapes are summarized in Tables

1-3. Example applications of these three major types of models are presented in the next section.

To be useful in an ecological risk assessment, the output of ecotoxicological models must be linked to assessment endpoints and risk estimates. Models must also incorporate toxicological information either explicitly as exposure-response functions, or implicitly as key biological parameters capable of perturbation to mimic the effects of toxic chemicals (e.g., on processes such as survival, growth, and reproduction of organisms in a population model). In detailed risk assessments, it may be useful to link ecotoxicological models with fate/transport models or bioaccumulation models.

Linking model output with assessment endpoints and risk estimates

As discussed earlier, model output should match the assessment endpoints for an ecological risk assessment directly or be quantitatively related to the assessment endpoints. This objective is achieved by selecting the appropriate model. Ecotoxicological models may also provide risk estimates as output. A Monte Carlo analysis runs a model simulation multiple times with the input variables for each run selected from their prospective probability distributions. The output for a population model would show a series of population trajectories (e.g., graphs of abundance) over time or as a probability distribution. For example, the RAMAS Ecotoxicology software can be used to simulate multiple projections of population dynamics in a Monte Carlo analysis and the resulting output can be used to derive estimates of risk of extinction (or quasi-extinction), risk of a given percentage decline in a certain time period, and other estimates of population-level risk. An example of population trajectories derived from Monte Carlo analysis and the method for estimating quasi-extinction risk is shown in Figure 4. Essentially any projection of an output variable for an ecotoxicological model can be converted to a risk estimate using Monte Carlo analysis.

Incorporating toxicological information into ecological models

An ecotoxicological model includes consideration of toxicological effects on biological entities (i.e., organisms, populations, or communities). The most direct way to incorporate toxicological effects in an ecological model is to add an exposure-response function or a series of such functions for the various endpoints as part of the mathematical equations for the model or as a subroutine in a simulation. Figure 5 shows an example exposure-response relationship derived from laboratory testing of PCB-contaminated sediments with mummichog (Fundulus heteroclitus) and interpretation of effects on survival and reproduction using a population life-history model to estimate population growth rate. Alternatively, toxicity data may be incorporated into a

Table 1 Summary of biological variables in selected ecological models

Population models

Ecosystem models

Landscape models

Variablea

Scalar abundance (crucian carp)

Life-history matrix (fathead

Metapopulation ■

RAMAS-GIS

(California minnow) gnatcatcher)

AQUATOX

CASM

IFEM

ATLSS

LANDIS

JABOWA

Community characteristics Number of species Typical species

One Fish

Type of trophic interactions'3 Relative abundances in immigrant population

Individual population characteristics Abundance or biomass in native population Age- or stage-specific abundance or biomass Carrying capacity Density dependence Maximum age Population growth rate Overall mortality rate Age- or stage-specific mortality rate Age of first reproduction Overall frequency of reproduction Overall fecundity (offspring per reproduction) Age- or stage-specific fecundity Home range size Periodicity (seasonality) of presence in home range

Multiple Phytoplankton, zooplankton, benthic infauna, fish

Multiple Phytoplankton, zooplankton, benthic infauna, fish

Multiple Zooplankton, benthic invertebrates, fish

X Xc

X Xc

Multiple Aquatic vegetation, fish, seed-eating birds, piscivorous birds, deer FC

Multiple Multiple

Tree species Tree species

NO X

Xd Xd

Xe X

(Continued)

Table 1 (Continued)

Population models

Life-

Scalar history abundance matrix

Ecosystem models

Landscape models

Variablea

Metapopulation ■ RAMAS-GIS

(crucian (fathead (California carp) minnow) gnatcatcher)

AQUATOX

CASM

IFEM

ATLSS

LANDIS

JABOWA

Offspring dispersal distance Immigration or emigration probability or rate Habitat suitability coefficients (by habitat type)

Other characteristics

Spatially explicit population data Locations inhabited Abundances Distribution of competitors Distribution of predators

Shade tolerance, fire tolerance

Shade tolerance

aA blank cell indicates that the variable is not applicable to the model. bNone (NO), food chain or food web (FC), producer-consumer (PC). cTwo cohorts for fish species.

dAge-structured models are used only for aquatic species. eFor zero-age individuals only, representing sprouting ability. fBoth effective and maximum seed dispersal distance.

Adapted from Pastorok RA, Akcakaya HR, Regan HM, Ferson S, and Bartell SM (2003) Role of ecological modeling in risk assessment. Human and Ecological Risk Assessment 9(4): 939-972.

Table 2 Summary of environmental variables in selected ecological models

Population models

Ecosystem models

Landscape models

Variablea

Scalar abundance (crucian carp)

Life-history matrix (fathead minnow)

Metapopulation - RAMAS-GIS (California gnatcatcher)

AQUATOX

CASM

IFEM

ATLSS

LANDIS JABOWA

Environmental media, variables, and species categories modeled Soil

Surface water Sediment

Suspended particulates Soil biota Sediment in fauna Terrestrial vegetation Aquatic vegetation Plankton Fish

Small mammals Birds

Spatially explicit site data

Location of immigrant population source Distribution of habitat types or land types Distribution of food, water, nesting sites, etc.

aA blank cell indicates that the variable is not applicable to the model.

Adapted from Pastorok RA, Akcakaya HR, Regan HM, Ferson S, and Bartell SM (2003) Role of ecological modeling in risk assessment. Human and Ecological Risk Assessment 9(4): 939-972.

Table 3 Summary of chemical toxicity and other disturbance variables in selected ecological models

Population models

Variablea

Scalar abundance (crucian carp)

Life-history matrix (fathead minnow)

Metapopulation ■ RAMAS-GIS (California gnatcatcher)

Ecosystem models

AQUATOX CASM

Landscape models

IFEM

ATLSS

LANDIS JABOWA

Chemical toxicity or other disturbances data

Types of disturbances or stressors

Contaminant mass balance modeled Soil contamination modeled Surface water contamination modeled Sediment contamination modeled Terrestrial vegetation contamination modeled Aquatic vegetation contamination modeled Bioaccumulation Growth rates/effects Mortality rates/effects Reproductive effects Other effects

Exposure indicators Spatial location of disturbance Spatial extent of disturbance Time scale of disturbance^

Chemical, habitat reduction

Chemical (Mirex)

Risk of extinction

Cold, wet winters; fire

Risk of population decline

Dispersal of juveniles, subpopulation characteristics, spatial correlations

Chemical

Chemical Chemical Low water Fire, wind Mortality

Population Population size size

(biomass) (biomass)

Community composition, species range and density

aA blank cell indicates that the variable is not applicable to the model. ^Continuous (C), periodic (P), irregular or any pattern (I).

Adapted from Pastorok RA, Akcakaya HR, Regan HM, Ferson S, and Bartell SM (2003) Role of ecological modeling in risk assessment. Human and Ecological Risk Assessment 9(4): 939-972.

Interval extinction risk = 3 of 5 cases = 0.6

Interval extinction risk = 3 of 5 cases = 0.6

Figure 4 Example derivation of risk estimate from Monte Carlo analysis of a population model. Bold lines indicate abundances below threshold of concern. A small number of simulation runs are shown for clarity. In practice, >1000 runs could be used for a Monte Carlo analysis to derive a risk estimate based on frequency of runs in which abundance decreases below threshold. Reproduced by permission of Routledge/Taylor & Francis Group, LLc.

o = Initial condition

Figure 4 Example derivation of risk estimate from Monte Carlo analysis of a population model. Bold lines indicate abundances below threshold of concern. A small number of simulation runs are shown for clarity. In practice, >1000 runs could be used for a Monte Carlo analysis to derive a risk estimate based on frequency of runs in which abundance decreases below threshold. Reproduced by permission of Routledge/Taylor & Francis Group, LLc.

Figure 5 Example exposure-response relationship for PCBs and mummichog (Fundulus heteroclitus) derived from population modeling. Redrawn from Munns WR, Black DE, Gleason TR, etal. (1997) Evaluation of the effects of dioxin and PCBs on Fundulus heteroclitus populations using a modeling approach. Environmental Toxicology and Chemistry 16:1074-1081, with permission from Society of Environmental Toxicology and Chemistry.

Figure 5 Example exposure-response relationship for PCBs and mummichog (Fundulus heteroclitus) derived from population modeling. Redrawn from Munns WR, Black DE, Gleason TR, etal. (1997) Evaluation of the effects of dioxin and PCBs on Fundulus heteroclitus populations using a modeling approach. Environmental Toxicology and Chemistry 16:1074-1081, with permission from Society of Environmental Toxicology and Chemistry.

physiologically based or dynamic energy-budget model, which is linked with a demographic model.

Exposure data are summarized as chemical distributions in space and time for either environmental media relevant to toxic effects data (e.g., soil, sediment, water) or for tissue residues of chemicals in target species being modeled. Toxicity data may be based on dose- (or concentration-) response relationships, a toxicity threshold (e.g., no-observed-adverse-effect level (NOAEL)), or any field or laboratory data linking variation in toxic chemical concentrations with biological response endpoints. Data on toxicity and population or community parameters must be matched to the spatial-temporal scale of the ecotoxicological model.

Linking with fate/transport models

Models that predict the fate and transport of chemicals in the environment can provide the basic chemical distribution and concentration data needed for estimating exposure where measurements of exposure are not available (e.g., for a new chemical assessment) or impractical (e.g., for an endangered species assessment). Linking the fate and transport model with the ecotoxicological model provides a comprehensive risk assessment approach. Bioaccumulation models are a special class of fate/transport models that estimate chemical uptake and sometimes distribution in target tissues of organisms from chemical data in the environment, organism distribution, chemical uptake and elimination rates by organisms, and other factors. Bioaccumulation models may incorporate food web exposure models. Many ecotox-icological models that predict ecosystem and landscape endpoints also include submodels that describe environmental transport and fate of chemicals in the environment, as well as exposure of organisms to those chemicals. Most population models do not incorporate fate/transport models, although linking such models may be a trend for future development of methods for ecological risk assessment.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment