Applying the Joint Probability Distribution Analysis for Pacific Northwest Salmonid Risk Assessment Qingli Ma, Rick Reiss and Cliff Habig Exponent, Inc. Paul Whatling Cheminova, Inc. OUTLINE OF THE PRESENTATION Overview of EPA and NMFS approaches for endangered species risk assessments The joint probability distribution analysis Assessing the potential risks of dimethoate to Pacific Northwest salmonids using the joint probability distribution analysis Comparison of estimated potential risks of dimethoate to Pacific Northwest salmonids Conclusions EPA APPROACH FOR PESTICIDE RISK ASSESSMENTS – RISK QUOTIENT (RQ) METHOD Exposure profile – estimate environmental concentration (EEC) Tier-1 GENEEC model estimate Tier-2 PRZM/EXAMS model estimate: 90th percentile values Effect profile – effect level (e.g., LC50) Risk characterization – RQ calculation RQ= Exposure/Toxicity For most risk assessments, the RQ is calculated by dividing a point estimate of exposure by a point estimate of effect Level of concern (LOC): Various risk presumptions for comparison with RQ to determine potential concern. EPA RISK PRESUMPTIONS FOR AQUATIC ANIMALS Risk presumptions RQ LOC Acute high risk EEC/LC50 or EC50 0.5 Acute restricted use EEC/LC50 or EC50 0.1 Acute endangered species EEC/LC50 or EC50 0.05 Chronic risk EEC/NOEC 1.0 NFMS APPROACH FOR DEVELOPING BIOLOGICAL OPINIONS (BIOPS) ON PACIFIC NORTHWEST SALMONIDS A similar assessment process to the EPA approach, with additional species population modeling for developing the “jeopardy” opinions. It integrates the status of the species, the environmental baseline, the exposure and the effects in developing BiOps. In the most recent BiOps (the third batch of six chemicals), NFMS used EPA’s Tier-1 GENEEC model to estimate potential cumulative exposure (EECs), which tends to significantly over-estimate exposure because of the conservative nature of the model. JOINT PROBABILITY DISTRIBUTION ANALYSIS In this approach, both the exposure and effect are treated as probabilistic distributions instead of point values Probabilistic distribution of exposure Probabilistic distribution of effect Joint probability distribution of exposure and effect, which describes the probability that an effect exceeding any given magnitude will occur under the range of exposure scenarios SAP and ECOFRAM recommend for probabilistic assessment of risks PROBABILISTIC DISTRIBUTION OF EXPOSURE Exposure assessment for dimethoate was undertaken with the linked PRZM/EXAMS model Major model input parameters that EPA used for the California red-legged frog assessment were used (for consistency with EPA RQ method) Parameter Value Parameter Value Water solubility (mg/L) 3200 Aerobic soil (d) 6.2 Kd (L/kg) 0.3 Aerobic aquatic (d) 16.4 Photolysis half-life (d) 353.0 Application rate (kg/ha) 0.56 Neutral hydrolysis (d) 6.8 Number of applications 3 Foliar half-life (d) 2.9 Application method aerial PROBABILISTIC DISTRIBUTION OF EXPOSURE Seven EPA standard PRZM/EXAMS scenarios were simulated that cover the major dimethoate use patterns, with considerations of rate and method of applications, use regions, and crops: CA-alfalfa CA-citrus MS-cotton OH-corn CA-lettuce OR-pears CA-cotton These scenarios represent the high-end exposure for the selected crops and are expected to produce runoff > 90% of the sites where the crop is grown. PROBABILISTIC DISTRIBUTION OF EXPOSURE The model was run for 30 years to generate statistically meaningful distributions The model-generated probability distributions of daily peak concentrations were used to construct the joint probability distributions for acute risk assessment. Probablity of Exceedance Probability of exceedance of dimethoate concentrations for CA-lettuce scenario 1 0.8 0.6 y = 1.086e-98.8x R² = 0.9749 0.4 0.2 0 0 0.005 0.01 0.015 0.02 0.025 Concentration (mg/L) 0.03 0.035 0.04 PROBABILISTIC DISTRIBUTION OF EFFECT For toxicity assessment, the concentration-effect relationship was derived from two effect endpoints: Percent of species affected, expressed as species sensitivity distribution The dose-response relationship SPECIES SENSITIVITY DISTRIBUTION (SSD) Invertebrate species were used as they serve as salmonid prey. Besides, fish is less sensitive than prey 9 freshwater and marine invertebrate species 19 of 48-hr toxicity data were selected for acute risk assessment The EPA Species Sensitivity Distribution Generator, which assumes a log-normal distribution, was used to generate SSD Species sensitivity distribution (SSD) DOSE-RESPONSE RELATIONSHIP The concentration-effect relationships between dimethoate and daphnia from three studies were used to generate the doseresponse curve: Anderson et al. (2006) Hertl et al. (2002) Song et al. (2007) The concentrations ranged from 0.8 to 9.95 mg/L and the mortality ranged from 0 to 100% The exposure time was 48 hrs and the temp. was 20-21 oC All three datasets were combined and fitted to a logarithm function to derive the effect distribution Dose-response relationship between dimethoate and daphnia Daphnia mortality (%) 100 80 y = 30.671ln(x) + 33.057 R² = 0.4952 60 40 20 0 0 2 4 6 8 10 Dimethoate concentration (mg/L) 12 GENERATION OF JOINT PROBABILITY DISTRIBUTION FOR RISK DETERMINATION The exposure distribution was integrated with the effect distribution to develop the joint probability distribution (JPD) A risk product (RP) was calculated from the JPD and used for risk determinations RP = Exceedance probability X Magnitude of effect The following four levels of 90th percentile RPs were used to categorize the risks after Giesy et al. (1999) and Giddings et al. (2005): If the RP < 0.25%, the risk is characterized as minimal If the 0.25% < RP < 2%, the risk is low If 2% < RP < 10%, the risk is intermediate, and If RP > 10%, the risk is characterized as high - 1 0.1 Joint Probability Risk Product 0.8 0.08 Max RP = 0.085% 90th percentile RP =0.071% 0.6 0.06 0.4 0.04 0.2 0.02 0 0 0 20 40 60 Species Affected (%) 80 100 Risk Product (%) Probability of Exceedance Joint Probability Derived from Acute Species Sensitivity Distribution (SSD) and Exposure Distribution for CA-Lettuce - Probablity of Exceedance 1 0.8 1E-08 1E-09 1E-10 1E-11 1E-12 1E-13 1E-14 1E-15 1E-16 1E-17 1E-18 1E-19 1E-20 Joint Probability Risk Product 0.6 Max RP << 0.25% 0.4 0.2 0 0 20 40 60 Mortality (%) 80 100 Risk Product (%) Joint Probability Derived from Daphnia Exposure and Dose-Response Relationship for CA-Lettuce SUMMARY OF THE RISKS DETERMINED FROM THE JOINT PROBABILITY DISTRIBUTION ANALYSIS The calculated risk product from the joint probability distribution derived from species sensitivity distribution and exposure was significantly less than 0.25% under one of the worst-case use scenarios (CA-lettuce) for dimethoate. Likewise, the calculated risk product from dose-response curve and exposure under the same use scenario was also significantly below 0.25%. Thus, it can be concluded that the risks imposed by dimathoate uses to salmonid preys are minimal. RISKS DETERMINED USING THE EPA RISK QUOTIENT METHOD The 90th percentile peak daily concentration predicted by PRZM/EXAMS from the same exposure scenario was 0.024 mg/L. This is the EEC EPA used for calculating RQ for acute risks The lowest acute toxicity endpoint for invertebrate species (marsh mosquito) is 0.031 mg/L The resulting RQ is 0.77 The EPA LOCs for aquatic animals are 0.5, 0.1, and 0.05 for acute high risk, acute restricted use, and acute endangered species, respectively. Thus, according to the EPA RQ method, dimethoate would trigger the acute risks for all three categories. COMPARISON OF RISKS DETERMINED BY JPD AND RQ METHODS The joint probability distribution analysis showed minimal risks from dimethoate use: The maximum and 90th percentile risk products were << 0.25% based on SSD analysis The maximum and 90th percentile risk products were << 0.25% based on dose-response analysis The RQ method calculated a RQ value of 0.77 and indicated potential risks for all three risk categories for aquatic animals The bias of the RQ method lies in its overly conservative assumptions and single point/species estimates for RQ calculations CONCLUSIONS The JPD analysis shows that the risks imposed by dimethoate uses to salmonid preys are negligible The JPD approach is more appropriate for risk determinations for the Pacific Northwest salmonids because they consume more than one invertebrate species It is a better tool for EPA and NMFS for jeopardy and habitat modification determinations The assessment is still highly conservative because it uses the stagnant PRZM/EXAMS farm pond scenario to estimate exposures in salmonid habitat REFERENCES Anderson, T.H., Tjornhoj, R., Wollenberger, L., Slothuus, and T., Baun, A. 2006. Acute and chronic effects of pulse exposure of Daphnia magna to dimethoate and pirimicarb. Environ. Toxicol. Chem. 25: 1187-1195 Song, M.Y., Stark, J.D., and Brown, J.I. 1997. Comparative toxicity of four insecticides, including imidacloprid and tebufenozide, to four aquatic anthropods. Environ. Toxicol. Chem. 16:2494-2500. Hertl, J. 2002. Acute toxicity of dimethoate to Daphnia magna in a 48-hour immobilization test. IBACON-Institut fur Biologische Analytik und Consulting. IBACON GmbH Study No. 10591220, 10591220A; Unpubl. Report, CHA Doc. No. 482 DMT.
© Copyright 2025 Paperzz