Jump to main content.

Progress Report of the Ecological Committee on FIFRA Risk
The ECOFRAM Aquatic Exposure and Aquatic Effects Subgroups

On this Page


The Ecological Committee on FIFRA Risk Assessment Methods (ECOFRAM) was formed in June 1997. The Committee's purpose is to develop tools and processes within the FIFRA framework for predicting the magnitude and probabilities of adverse effects to non-target aquatic and terrestrial species resulting from the introduction of pesticides into their environment. An Aquatic Work Group was formed to identify and discuss probabilistic methods for aquatic risk assessments and develop recommendations for future use by EPA. In addition, we are identifying information that must be developed in order to validate the proposed methods in order to ensure that the proposed assessment process, if adopted by EPA, supports environmental decisions that are scientifically defensible.

In accompanying posters, the ECOFRAM Aquatic Exposure and Effects Subgroups are displaying reports on what has been achieved so far. Both groups have progressed to identifying some of the tools which will be needed to perform probabilisitic aquatic effects and exposure characterizations. In addition, both groups have started to consider a "tiered" framework within which the tools should be used along with the criteria that might determine when to move from one tier to the next.

Both teams recognize that one of the fundamental steps to success for ECOFRAM will be the way in which the Aquatic Exposure and Effects subgroups manage to combine their recommendations into an integral Aquatic Risk Assessment approach. The first joint meeting to integrate the two groups views on this will take place in December 1997. The resulting Aquatic Risk Assessment decision tree is expected to be the primary tool for making regulatory aquatic exposure assessments more predictable.

This poster shows the draft decision trees from both groups. The views of attendees will be sought on how the views could be combined. In addition, the poster displays lists of the issues raised so far as well as some key questions to prompt participation from SETAC attendees. The items being presented are "works in progress" and the teams are both requesting feedback from conference attendees especially via participation in the poster session.

Top of page


There are three primary issues facing the ECOFRAM Aquatic team in producing integrated tools for Aquatic Risk Assessment. The primary one is

Traditionally, this has been conceived as a "tier system"; however, it is possible that the process will become a "decision matrix" since it may not be a matter of simple progression from one level of complexity to a higher one. Another idea is for the process to consist of guidance on the criteria for selecting "tools" from the "Aquatic Risk Assessment Toolbox".

There is a need to avoid the use of "bright lines" in the process after the initial tier which decides whether there are any regulatory concerns. A mechanism is required to avoid sterile debates over differences in Risk Quotient (or its probabilistic equivalent) that are insignificant relative to the inherent uncertainty. One suggestion was to define broader "classes" of risk on a Red, Amber, Green basis.

During independent discussion in the workgroups, it is clear that the criteria for progression through the process are likely to be based on interim assessments of risk rather than independent estimates of effects or exposure - i.e. the process will require the efforts of the eco-toxicologist and the exposure scientist to be synchronized. This realization prompts two other key questions:



These points may require different answers depending upon where the current risk assessment falls within the overall process.

Top of page

EFFECTS Workgroup Draft Process

graphic of the tiers for aquatic effects assessment flowing from top to bottom:  Tier 1, Tier 2, Inter-Tier Risk Mitigation Feedback Loop, and Tier 3.  Each has more detailed explanation immediately following this diagram

Proposed Tiers for Aquatic Effects Assessment


Data generation and analysis for every chemical; results are used to determine what needs to be done next (if anything).

Decide on direction based on Tier 1 results. Any or all of the following pathways for further investigation may be applicable.
[We still need to determine criteria for each option.]

More Acute Toxicity Tests
  • Focus on most sensitive groups from Tier 1.
  • More toxicity data would justify a smaller safety factor.
  • With enough species, can determine distribution of species sensitivities. Characterize using point estimate (e.g., 10th percentile); even better, use whole distribution in probabilistic risk characterization [need to develop techniques for this].
  • Provides more detailed ecological information for ecosystem-specific or regional analysis.
Chronic Toxicity Tests
  • Early life-stage tests with other fish (e.g. marine species).
  • Full life-cycle with other invertebrates (e.g. marine species).
  • Full life-cycle with fish (but need to improve protocol).
  • Use population models to extrapolate to long-term population impact
Pulsed Exposures
  • Laboratory exposure corresponding to PRZM/EXAMS output. Use results to characterize exposure-response relationship in terms of pulse height, duration, or other exposure metrics (instead of constant concentration as in Tier 1).
  • Pharmacokinetic (PK) models to explore effects associated with complex exposure patterns (including multiple pulses).
  • Cumulative hazard models.
  • Population models using PRZM/EXAMS (or other exposure model) output.
Sediment Toxicity Evaluation
  • For compounds with Kow > x, half-life > y, chronic toxicity < z.
  • Equilibrium Partitioning (EqP) assumptions to calculate sediment toxicity; compare against PRZM/EXAMS predictions of sediment concentrations.
    • Does EqP apply?
    • Are water col. species good surrogates for benthic species?
  • Sediment toxicity tests (e.g., midges, amphipods, mayflies, oligochaetes, polychaetes...)
    • Is EqP approach confirmed experimentally?
    • Are benthic species equally sensitive as water column species?
  • If yes to both, use water-column toxicity data for sediment risk assessment.
  • If no to either, conduct sediment toxicity tests with more benthic species.

Inter-Tier Risk Mitigation Feedback Loop

Risk assessors and risk managers consider mitigation options. Loop back through Tiers 1 and 2 to evaluate effectiveness, risk/benefit of different options.

Specialized studies to address specific areas of uncertainty in the risk assessment, or refine assessment for specific habitats, regions, or use patterns. Tools could include:

Top of page

Assessment Methods: IV Aquatic Risk Assessment

2 human-like drawings shaking hands with one representing Exposure and the other representing Effects so that together they represent the agreed aquatic risk assessment process


Subgroup team members and their affiliations are listed in adjacent posters.


The two schemes below reflect the initial thinking of the ECOFRAM Aquatic Effects and Exposure Subgroups. We will be trying to blend it into a common approach in December.

Please pass all comments to: Exit EPA DisclaimerPaul Hendley (Paul.Hendley@agna.Zeneca.com) or Exit EPA DisclaimerJeff Giddings (jgiddings@compuserve.com)

Top of page

EXPOSURE Workgroup Draft Process

4 steps: 1 is Are there any regulatory issues; 2 is exposure characterization; 3 is refining understanding of exposure; 4 is validation and performing compound and use specific studies or monitoring

Steps in the Exposure Workgroup Draft Process

  • Define the system at risk and help define scenarios for future modeling
  • Uses simple models to evaluate possibility of risk to Avian, Terrestrial Mammals, Aquatics, Humans/FQPA or "the Environment"
Result"Pass" or "Do further work" decision
ScenariosSemi-specific to crops and use patterns. E.g. row crops may be grouped but uses like rice, cranberries, forest, turf etc will have independent scenarios. It is likely that as the process develops, more and more crop specific scenarios will be incorporated.
  • Water column concentrations will be provided for a Static pond (as per current system)
  • Single point values for instantaneous, 4 day, 21 day, 56 day - should include simple error estimate
  • Should describe "shape" of exposure with time
  • Sediment concentrations - no output provided

  • Characterize exposure sufficiently to define risk.
  • Examine variation of exposure regionally and temporally
  • Perform FQPA risk assessment??
Result "Pass" or "Do further work" or "Modify label"
  • Most likely to be a "toolbox" with many options to select from.
  • The first one may be a simple PRZM/EXAMS run to provide definitive information not available from GENEEC on
    • temporal exposure profile distribution vs. chronic tox
    • sediment toxicity
    • impact of degradative processes other than soil aerobic (e.g. photolyis)
  • If this does not remove regulatory concerns, other tools will examine regional variability and the impact of "real world landscapes".
  • Representative sites across likely label regions such that variables remain appropriately correlated (e.g. using linked sets of regional rainfalls & watershed area & soil types & sustainable pond volumes)
  • Distribution of likely agricultural landscapes and permutations (slopes, cropping densities, proximities etc)
  • "Multi-aquatic-receiving-body" watershed to provide info on exposures in relevant habitat(s) (includes estuarine where appropriate)
  • tool needed to define which habitat(s) to concentrate on during exposure characterization
  • 50 years of real weather appropriate to region
  • No mitigation measures are incorporated at this stage
  • Chemical properties to be distributions where known (even regional values??) - if not known, "nominal" variations may be used.
  • Multi-dimensional probability surface showing regional and temporal variation
  • Provides water columns and sediment concentration data (instantaneous and intervals) for appropriate combination of static water bodies, constructed wetlands, canals and estuaries where appropriate, 1st, 2nd 3rd order streams or reservoir "habitats"

4 steps: 1 is Are there any regulatory issues; 2 is exposure characterization; 3 is refining understanding of exposure; 4 is validation and performing compound and use specific studies or monitoring

Steps in the Exposure Workgroup Draft Process

  • Refine uncertainties
  • Define key "drivers" of exposure
  • Examine whether model assumptions in exposure characterization step are applicable for chemical/use pattern.
  • Identify what could provide mitigation
ResultPass/Label Modification &/or recommendations for agreement with EPA Risk Managers
  • Incorporate more "real world" assumptions. E.g. distribution of likely usage scenarios (% adoption, likely local use frequency, local use rates etc)
  • Probably does more work on fewer sites representing key regions/habitats at risk
  • Incorporates mitigation technology and approaches
  • Provides spatial and temporal distribution of mitigated EEC's
  • Provides "categorized" risks to assist risk managers
  • Provides a suggested course for regulatory action
  • additional data needed

  • Cost Effective (generally less expensive than monitoring)
  • Time involved is days to months
  • Ability to evaluate "what-if" scenarios (e.g. climate, soil, application date)
  • Ability to evaluate effectiveness of some mitigation measures
  • Ability to predict concentrations over a continuum in space and time
  • Comparative exposure assessments are possible
  • Ability to simulate concentrations below analytical limits of quantification
  • Costly
  • Time involved is weeks to years
  • Difficult to design cost effective AND technically viable sampling programs
  • May require many years of monitoring and/or paired studies to evaluate effectiveness
  • Handling non-detects is difficulty

  • Results are accepted as "true" values
  • Sampling represents discrete points in space and time that can only be put into context with modeling
  • Study only represents one unique combination of conditions
  • Can be constrained by analytical precision and LOD
  • Results can be misleading if one year is a 1 in 100 event year!!
  • Difficult to interpret results in a probabilistic fashion
  • Cause and Effect may be difficult to assign (especially Biological)
  • Uncertainty in model predictions due to mathematical representation of complex governing processes, programming and user errors and uncertainty in input parameter values.
  • Simplifications required in representation of prototype systems

  • There is general public reluctance to accept predicted data
  • Needs calibration to see how closely predicted values match reality
  • Many input values have high uncertainties associated with them
  • Selected input parameters may not be environmentally feasible
  • Provides an actual measurement of chemical residue concentration, hydrologic response etc

  • Avoids conservatism from compounding conservative assumptions
  • When done well it is an excellent tool
  • Accounts for the inherent heterogeneity of the system

  • There is a greater acceptance of measured data
  • There is public confidence in monitoring data


Monitoring may be thought of as another "model" with definable uncertainties - particularly for Surface Water where sampling timings can be critical to capturing an "event".

The most powerful approach would be a combination using thoroughly planned monitoring data across several years to calibrate validated models in which regulators have confidence. The modeling will provide probabilistic estimates of exposure

Top of page

Issues for Joint Discussion

  1. Pond Parameters

    Need to define pond volume to watershed variation by region

  2. Pond/runoff mixing issues and Refugia.

    What is the biological significance of runoff entry areas experiencing higher residues while the remainder of the water body gets relatively low exposure??

    PRZM/EXAMS predicts immediate mixing!!

  3. What are we trying to protect?? Which habitats are of key interest

    • Flowing/static
    • Constructed wetlands
    • Drainage canals
    • Local "features"

  4. Sediment Concentrations

    Do we need step I sediment values and water column concns.?

  5. Exposure Duration Step I

    How to average data - arithmetic or geometric??

    Study Duration vs. Intervals reported - 56 days or 60 days

  6. Exposure Duration Step I

    Meaning of 21 day concs if caused by a 1 day spike??

    Impact on chronic study design by compound??

    Short generation time organisms - what is value of reporting by "annual return frequencies"??

    Value of minimum modeling time step of 1 day??

  7. Expression of Combined Probabilities

    Use of risk categories

  8. Linking Exposure Estimates to Existing Monitoring data


  9. Model Development, Validation and Research Needs

Top of page

List of Issues and Questions from ECOFRAM Exposure Team



Top of page

Publications | Glossary | A-Z Index | Jobs

Local Navigation

Jump to main content.