Human Health and Environmental Relative Risks of WNV Mosquito Control Products

METHODS

Product Identification Criteria

Due to the time constraints and limited staffing for this project, a detailed review of all the data registrants generate to support product registration was not undertaken. Instead, a strategy was developed to identify registered products (1999 registration year) along with their use rates, methods of application, appropriate sites and environmental exposures. This was followed by a search of secondary sources for toxicity information on each of the active ingredients. There was a large amount of information available for some of the compounds, and little information for others.

The Silver Platter Information Retrieval System (SPIRS) database at Purdue University was used to identify products registered in Maine in 1999 (Appendix I; Tables 1 and 2). Relevant site information was also obtained from the SPIRS database (57) (Appendix II; Tables 1 - 10). Next, an in-house label file review was conducted to check for changes in registration during the 2000 registration year. The labels and material safety data sheets (MSDS) were then reviewed for environmental and human health concerns as well as verification of sites, methods of application and use rates. In the event of a WNV outbreak in the year 2001, current registration of a product would be verified and a label consulted prior to use.

The labels and MSDS supply a wealth of information on the toxicity and risks associated with pesticide products. Twenty-seven products were included in this review which resulted in over 200 pages of labels and MSDSs. Since much of the information is not relevant to mosquito control, i.e. agricultural, forestry, ornamental and structural uses, a decision was made to extract the pertinent information and include it in label review tables. Information required by EPA on the pesticide labels is a result of a detailed federal risk assessment process. This applies both to human, nontarget mammals, insects (including bees), birds, wildlife and aquatic organisms. The information from the labels is presented in Appendix III; Tables 1 to 10. EPA's standards for precautionary label information are found in 40CFR156.10 (h)(i) and (ii) (Appendix III Section 1). Pesticide products, specifically the organophosphates, malathion and naled are in the process of re-registration under new standards created by the Food Quality Protection Act (FQPA) of 1996. The synthetic pyrethroids have undergone risk assessment using pre-FQPA methods and are in the queue for re-registration.

Risk Assessment; inputs

Risk of a toxic effect when one is exposed to a chemical at a certain dose level can be predicted mathematically. The acute or short term risk is related to the ratio of the toxicity factor (data from laboratory animals) to the exposure factor (from modeling or environmental monitoring). Mammalian toxicity data was obtained from secondary sources and is summarized in Appendix IV; Tables 1 and 2; adulticides and 3; larvicides. Cancer classification for the active ingredients are found in these tables. Interpretation of these rankings is found in Appendix V Section 1. Because WNV related treatments (and any associated exposure) are expected to be limited, quantitative carcinogenic risk assessment has not been performed. The information will be used qualitatively in the decision making process.

Toxicity data for aquatic species is grouped into warm water fish, cold water fish, estuarine and marine species and freshwater invertebrates. These data are found in Appendix IV tables 4; adulticides and 5 ; larvicides. There is a large quantity of data in a variety of species for certain compounds (malathion, naled resmethrin, Bti, methoprene and temephos) and less data on others (permethrin, phenothrin, B. sphaericus , and POE MMF). The more extensive the data the greater the comfort level with the risk assessment. A search of the primary peer reviewed public scientific literature would improve this database.

Toxicity data for birds and bees is located in Appendix IV table 6; adulticides and 7 for larvicides. Given the larvicide use patterns and lower levels of toxicity, risks would be less from larvicides than risks from adulticides. So much so, that EPA has waived the bee toxicity data requirement for the organophosphate temephos.

Environmental half life and solubility are summarized in Appendix V Tables 1; adulticides and 2; larvicides. This information will be incorporated when we obtain the software for modelling capabilities. At this point in time, this information will be used qualitatively in the decision making process.

Data regarding toxicity and exposure are expressed using the metric system of units (mg/kg). Use information is expressed in the English system (lbs ai/acre). The table of conversion factors for these units is found in Appendix VI; section 1. The use rate in the units mg/acre (mg/A) are found in Appendix VI tables 1 and 2; adulticides and Table 3; larvicides. The quantity mg/A is used in the relative risk determination as a surrogate for exposure.

Risk Assessment; assumptions

Risk assessment terms are defined in the definition section on page xi. There are two methods of evaluating the toxicity/exposure relationship for non-cancer toxic endpoints. These are the margin of exposure (MOE) and the risk quotient (RQ). The MOE method is used for human health risk assessments and the RQ is used for environmental risk assessment. Short term NOAELs are not available for the compounds in question so relative risks were estimated using subchronic (90 days to 6 months) dietary NOAELs (mg/kg/day).

The assumption is made that except for use rates, exposure parameters are comparable and that the EPA models used to estimate absolute risk are linear in the areas of single applications. This assumption is supported by the fact that the ratio of the EPA's MOEs for naled and malathion is similar to the ratio of the modMOEs for naled and malathion. The ratio of the EPA modelled MOE for naled (1500) (7) to that of malathion (10,000) (37) is 0.15. The ratio of the comparable modMOEs for naled (0.000022) and malathion (0.000184) (Appendix VII) is 0.12. When multiple applications and environmental fate enters the equation, the assumption of linearity is no longer valid (69).

The existence of an EPA approved label with use rates, application techniques and prohibitions implies that EPA's criteria for acceptable absolute risk has been met. Because of this, and given the lack of in-house modeling capabilities, the BPC's advisory committees examined relative risk issues. Where EPA's risk assessments are available, the absolute risks are included in the text along with the relative risk discussion.

In the evaluation of relative risks, a full exposure assessment is not necessary. Instead the application rate in mg/Acre was used as a modified exposure (modEXP) factor. This modExp was then used in the MOE calculation creating a modified MOE (modMOE), and in the risk quotient calculation to create a modRQ. By comparing the modMOE for each compound for humans and non-target mammals and the modRQ for each compound for each non-target group (warm water fish, cold water fish, marine and estuarine species, freshwater invertebrates, and birds), a selection may be made minimizing the potential human health and environmental impacts. EPA's acceptable risk cutoff points are not valid for the relative risk comparison.

NEXT: Results: Human Health Risks; Adulticides