Derivation of Values and Thresholds for Relative Cumulative Ranking System

Derivation of Values and EPA Guidelines

In order to relatively rank the 100 commonly used residential pesticides - including those use for pest management, right-of-ways, golf courses, algaecides, lawn and garden, and fungicides & fumigants (focusing on a common SC coastal crop, tomato farms) - for overall environmental safety within the tri-county area, we utilized the LD50 values (measures acute toxicity) and NOAEL values (measures chronic toxicity) derived and submitted to EPA by the registrant for evaluation for each taxonomic group tested. Environmental fate and transport characteristics (e.g., log Kow, Kocand half-life) were derived from Pubchem, TOXNET, and USEPA REDs Page. Organisms at different trophic levels within the food web are tested and include:

  1. Mammals (Acute and chronic rodent study)
  2. Avian species (terrestrial non-target vertebrate)
    • Mallard duck (Anas platyrhynchos)
    • Bobwhite quail (Colinus virginianus)
  3. Honey bee (Apis mellifera) (terrestrial non-target invertebrate species)
  4. Aquatic invertebrates (Freshwater: Daphnia magna)
  5. Fish (Freshwater: fathead minnow, rainbow trout; Saltwater: Sheepshead minnow)
  6. Terrestrial and Aquatic non-target plants (green algae)

The EPA testing guidlines for hazard are outlined in detail concerning the requirements of the bioassay and can be accessed from:

Reference site for toxicity guidelinesSome of the EPA OCSPP assays used in the cumulative ranking of residential compounds include:

  • Acute Toxicity: Acute Oral Rat Toxicity – updated in 1996; GLN #: 870.1100
  • Chronic Toxicity: Chronic Feeding Study – updated in 1998; GLN #: 870.4100
  • Acute Toxicity: Avian Acute Oral Toxicity Test – updated 2012; GLN #: 850.2100
  • Chronic Toxicity: Avian Dietary Toxicity Test – updated 2012; GLN #: 850.2200
  • Acute Toxicity: Honeybee Acute Contact Toxicity – updated 2012; GLN #: 850.3020
  • Acute Toxicity: Aquatic Invertebrate Acute Toxicity Test – updated 1996; GLN #: 850.1010
  • Chronic Toxicity: Daphnid Chronic Toxicity Test – updated 1996; GLN #: 850.1300
  • Acute Toxicity: Fish Acute Toxicity Test – updated 1996; GLN #: 850.1075
  • Chronic Toxicity: Fish Early Life-stage Toxicity Test – updated 1996; GLN #: 850.1400
  • Acute Toxicity: Algae Toxicity Test –updated 1996; GLN #: 850.5400

LD50 and NOAEL values were used for the relative cumulative ranking process as there is great variation in toxicological data availability for compounds due to the lengthy period between reregistration. In some instances, chronic values are not collected for certain ecological species, most likely due to the low acute toxicity. In the instances where no data yet exist for a certain parameter, then that particular parameter is given a null value. The next step for is to model development for specific geographic areas and for residential application scenarios rather than crop, ultimately developing EECs relevant to the tri-county area. For each toxicity test being considered (e.g., aquatic toxicity) the most sensitive surrogate species (i.e., the test species) that had the lowest LD50 value) for each test will be used in our final cumulative ranking system to provide conservative values for the relative cumulative ranking of safety.

For final ecological risk assessment, the EPA compares the Risk Quotient (RQ) for both acute and chronic toxicity (if possible) and compares it to the Level of Concern (LOC) set by the EPA. Each RQ is deterministic and is based on the peak estimated environmental concentration (EEC) from PRZM-EXAM models divided by the toxicological based benchmark dose (i.e., LC50, EC50, EC25). In re-registration Eligibility Decisions (REDs), a compound may exceed the LOC for several parameters, but the EPA also considers “mitigated use patterns or typical doses” that can drop a compound below the LOC. For a better understanding of LOCs and ecological risk assessment, please refer to the toolbox menu item, and click Technical Example and/or Useful Links (under the Risk category).

Cumulative Ranking and Derivation of Thresholds:

Thresholds for the toxicity values were set according to EPA hazard ranking system and environmental fate and transport values. It is important to note when looking at the LD/LC50 for toxicity values, that the lower the value the more toxic the compound. Furthermore, for chronic toxicity, Values used for each parameter, or compartment, being considered are the most conservative values (e.g., lowest LD50). Next, the most conservative values from each major category (acute and chronic toxicity for terrestrial and aquatic non target species) are placed in a second table. Relative cumulative rankings will be based on seven different equally weighted parameters. Due to the high degree of uncertainty concerning data used for analysis (e.g., estimated values for e-fate characteristics), and the lack of use data, each category will be divided into subcategories (low, moderate, and likely hazard) and given numeric values (1= low, 5= moderate, 10= high) depending upon its grouping based on set threshold values (please see the rationale and methodology for more details). For the toxicity and environmental fate parameters considered, thresholds were based on the EPA’s hazard rating system. Once numeric values were assigned to each category for a pesticide, then a summation was taken across categories and then averaged, and cumulative values were assigned. Cumulative frequency distributions were calculated to obtain a final cumulative hazard rating for each pesticide and each category. The cumulative frequency distribution will start from the lowest and go to the highest summed values - with the lowest values falling into the low hazard category and the highest grouping in to the likely to be hazardous category. Once normality and variance were checked, the three categories using SAS 9.3 (Cary, NC) by tertiles. Each tertile was checked against the others to confirm each was statistically significantly different (P<0.0001).

In using this ranking system, we are focusing on the overall potential for ecosystem hazard. The assumption being that if pesticides are applied according to the label, then they should not exceed the EEC established by the EPA. By using a hazard-based approach, our relative comparison for pesticides it gives users potential for ecosystem impact, not necessarily overall risk. A risk-based approach for residential scenarios in South Carolina will take efforts from many stakeholders.