Probabilistic risk assessment of cotton pyrethroids: I. Distributional analyses of laboratory aquatic toxicity data.

This is the first in a series of five papers that assess the risk of the cotton pyrethroids in aquatic ecosystems in a series of steps ranging from the analysis of effects data through modeling exposures in the landscape. Pyrethroid insecticides used on cotton have the potential to contaminate aquatic systems. The objectives of this study were to develop probabilistic estimates of toxicity distributions, to compare these among the pyrethroids, and to evaluate cypermethrin as a representative pyrethroid for the purposes of a class risk assessment of the pyrethroids. The distribution of cypermethrin acute toxicity data gave 10th centile values of 10 ng/L for all organisms, 6.4 ng/L for arthropods, and 380 ng/L for vertebrates. For bifenthrin, cyfluthrin, lambda-cyhalothrin, and deltamethrin, the 10th centile values for all organisms were 15, 12, 10, and 9 ng/L, respectively, indicating similar or somewhat lower toxicity than cypermethrin. For tralomethrin and fenpropathrin, the 10th centiles were <310 and 240 ng/L, respectively. The distribution of permethrin toxicity to all organisms, arthropods, and vertebrates gave 10th centiles of 180, 76, and 1600 ng/L, respectively, whereas those for fenvalerate were 37, 8, and 150 ng/L. With the exception of tralomethrin, the distributions of acute toxicity values had similar slopes, suggesting that the variation of sensitivity in a range of aquatic nontarget species is similar. The pyrethroids have different recommended field rates of application that are related to their efficacy, and the relationship between field rate and 10th centiles showed a trend. These results support the use of cypermethrin as a reasonable worst-case surrogate for the other pyrethroids for the purposes of risk assessment of pyrethroids as a class.

Probabilistic risk assessment of cotton pyrethroids: II. Aquatic mesocosm and field studies.

Results of mesocosm and field studies with cypermethrin and esfenvalerate were analyzed and interpreted to support an ecological risk assessment of cotton pyrethroids in aquatic ecosystems. A core group of seven mesocosm studies conducted on two continents over the course of a decade were examined, and additional observations from mesocosm and field studies with these and other cotton pyrethroids were also brought to bear. The results for cypermethrin and esfenvalerate were remarkably consistent. They revealed a trend in sensitivity from amphipods, isopods, midges, mayflies, copepods, and cladocerans (most sensitive) to fish, snails, oligochaetes, and rotifers (least sensitive). With few exceptions, populations affected by pyrethroids in the mesocosms recovered to normal levels before the end of the year of exposure; most populations recovered within weeks. Factors presumed responsible for population recovery included internal refuges (areas of low exposure), resistant life stages, rapid generation times, and egg deposition by adults from outside the treated systems. Indirect effects on fish (which have been hypothesized to occur when invertebrate food sources are reduced) were not observed. The lowest-observed-adverse-effect concentrations for the overall ecosystems for cypermethrin and esfenvalerate corresponded to the 54th and 41st centiles of acute toxicity endpoints (LC50s) for arthropods measured in laboratory studies with these compounds, implying that a risk characterization based on 10th centiles would be highly conservative.

Probabilistic risk assessment of cotton pyrethroids: III. A spatial analysis of the Mississippi, USA, cotton landscape.

Estimates of potential aquatic exposure concentrations arising from the use of pyrethroid insecticides on cotton produced using conventional procedures outlined by the U.S. Environmental Protection Agency's Office of Pesticide Programs Environmental Fate and Effects Division seem unrealistically high. Accordingly, the assumptions inherent in the pesticide exposure assessment modeling scenarios were examined using remote sensing of a significant Mississippi, USA, cotton-producing county. Image processing techniques and a geographic information system were used to investigate the number and size of the water bodies in the county and their proximity to cotton. Variables critical to aquatic exposure modeling were measured for approximately 600 static water bodies in the study area. Quantitative information on the relative spatial orientation of cotton and water, regional soil texture and slope, and the detailed nature of the composition of physical buffers between agricultural fields and water bodies was also obtained. Results showed that remote sensing and geographic information systems can be used cost effectively to characterize the agricultural landscape and provide verifiable data to refine conservative model assumptions. For example, 68% of all ponds in the region have no cotton within 360 m and 92% of the ponds have no cotton within 60 m. Only 2% of ponds have cotton present in all directions around the ponds and within 120 m. These are significant modifications to conventional pesticide risk assessment exposure modeling assumptions and exemplify the importance of using landscape-level risk assessments to better describe the Mississippi cotton agricultural landscape. Incorporating spatially characterized landscape information into pesticide aquatic exposure scenarios is likely to have greater impact on the model output than many other refinements.

Probabilistic risk assessment of cotton pyrethroids: V. Combining landscape-level exposures and ecotoxicological effects data to characterize risks.

Since their introduction, synthetic pyrethroid insecticides have generated regulatory concerns regarding their toxicity to fish and aquatic invertebrates. In this paper we assess the potential for risks to aquatic ecosystems in cotton-growing areas, focusing on cypermethrin as a suitable representative of the pyrethroid class and static water bodies (ponds and lakes) as worst-case water bodies because of low levels of dilution. Reviews of cypermethrin effects under laboratory and field conditions have characterized the potential aquatic effects of the chemical. Also, a landscape-level exposure characterization has been conducted in a worst-case cotton-growing county, Yazoo County, Mississippi, USA, to provide a more realistic exposure characterization than is possible using standard model scenarios. Risks were characterized using the standard tier I and II approaches of the U.S. Environmental Protection Agency. In addition, a probabilistic risk assessment was conducted by comparing landscape-level exposure calculations for ponds and lakes in Yazoo County (modified tier II analysis) with distributions of laboratory effect concentrations and with data from field studies. Risk characterization using tier I and tier II models demonstrated a level of concern for certain aquatic organisms. However, modified tier II analysis showed that exposure concentrations are unlikely to exceed concentrations that might cause ecologically significant effects. Indeed, in the vast majority of cases, concentrations in the modified tier II analysis were several orders of magnitude lower than those at which effects would be predicted on the basis of laboratory and field data. The conclusion of minimal potential for adverse ecological effects was also supported by field studies, which showed that impacts on aquatic systems were negligible, even at concentrations many times higher than the modified tier II exposure concentrations.