Keith R. Solomon

Ecotoxicological Risk Assessment for Roundup® Herbicide

Glyphosate-based weed control products are among the most widely used broad-spectrum herbicides in the world. The herbicidal properties of glyphosate were discovered in 1970, and commercial formulations for nonselective weed control were first introduced in 1974 (Franz et al. 1997). Formulations of glyphosate, including Roundup® Herbicide (RU)1 (Monsanto Company, St. Louis, MO), have been extensively investigated for their potential to produce adverse effects in nontarget organisms. Governmental regulatory agencies, international organizations, and others have reviewed and assessed the available scientific data for glyphosate formulations and independently judged their safety. Conclusions from three major organizations are publicly available and indicate RU can be used with minimal risk to the environment (Agriculture Canada 1991; USEPA 1993a; WHO 1994). Several review publications are available on the fate and effects of RU or glyphosate in the environment (Carlisle and Trevors 1988;Smith and Oehme 1992 ; Malik et al. 1989;Rueppel et al. 1977; Sullivan and Sullivan 1997;Forestry Canada, 1989). In addition, several books have been published about the environmental and human health considerations of glyphosate and its formulations (Grossbard and Atkinson 1985; Franz et al. 1997). In addition, RU and other glyphosate formulations have been selected for use in a number of weed control programs for state and local jurisdictions in the United States. Many of these uses require that ecological risk assessments be conducted in the form of Environmental Impact Statements or Environmental Assessments. These documents are comprehensive and specific to local use situations. Documents are available for risk assessments in Texas, Washington, Oregon, Pennsylvania, New York, Virginia, and other states (USDA 1989;USDA 1992;USDA 1996;USDA 1997;USDI 1989; Washington State DOT 1993).

Aquatic ecotoxicology of fluoxetine.

Recent studies indicate that the pharmaceutical fluoxetine, a selective serotonin reuptake inhibitor, is discharged in municipal wastewater treatment plant effluents to surface waters. Few data on environmental fluoxetine exposure and hazard to aquatic life are currently available in the literature. Here, we summarize information on fluoxetine detection in surface waters and review research on single-species toxicity test, Japanese medaka (Oryzias latipes) reproduction and endocrine function, and freshwater mesocosm community responses to fluoxetine exposure. Based on results from our studies and calculations of expected introduction concentrations, we also provide a preliminary aquatic risk characterization for fluoxetine. If standard toxicity test responses and a hazard quotient risk characterization approach are solely considered, little risk of fluoxetine exposure may be expected to aquatic life. However, our findings indicate that: (1) the magnitude, duration and frequency of fluoxetine exposure in aquatic systems requires further investigation; (2) mechanistic toxicity of fluoxetine in non-target biota, including behavioral responses, are clearly not understood; and (3) an assessment of environmentally relevant fluoxetine concentrations is needed to characterize ecological community responses.

Sources, pathways, and relative risks of contaminants in surface water and groundwater: a perspective prepared for the Walkerton inquiry.

On a global scale, pathogenic contamination of drinking water poses the most significant health risk to humans, and there have been countless numbers of disease outbreaks and poisonings throughout history resulting from exposure to untreated or poorly treated drinking water. However, significant risks to human health may also result from exposure to nonpathogenic, toxic contaminants that are often globally ubiquitous in waters from which drinking water is derived. With this latter point in mind, the objective of this commission paper is to discuss the primary sources of toxic contaminants in surface waters and groundwater, the pathways through which they move in aquatic environments, factors that affect their concentration and structure along the many transport flow paths, and the relative risks that these contaminants pose to human and environmental health. In assessing the relative risk of toxic contaminants in drinking water to humans, we have organized our discussion to follow the classical risk assessment paradigm, with emphasis placed on risk characterization. In doing so, we have focused predominantly on toxic contaminants that have had a demonstrated or potential effect on human health via exposure through drinking water. In the risk assessment process, understanding the sources and pathways for contaminants in the environment is a crucial step in addressing (and reducing) uncertainty associated with estimating the likelihood of exposure to contaminants in drinking water. More importantly, understanding the sources and pathways of contaminants strengthens our ability to quantify effects through accurate measurement and testing, or to predict the likelihood of effects based on empirical models. Understanding the sources, fate, and concentrations of chemicals in water, in conjunction with assessment of effects, not only forms the basis of risk characterization, but also provides critical information required to render decisions regarding regulatory initiatives, remediation, monitoring, and management. Our discussion is divided into two primary themes. First we discuss the major sources of contaminants from anthropogenic activities to aquatic surface and groundwater and the pathways along which these contaminants move to become incorporated into drinking water supplies. Second, we assess the health significance of the contaminants reported and identify uncertainties associated with exposures and potential effects. Loading of contaminants to surface waters, groundwater, sediments, and drinking water occurs via two primary routes: (1) point-source pollution and (2) non-point-source pollution. Point-source pollution originates from discrete sources whose inputs into aquatic systems can often be defined in a spatially explicit manner. Examples of point-source pollution include industrial effluents (pulp and paper mills, steel plants, food processing plants), municipal sewage treatment plants and combined sewage-storm-water overflows, resource extraction (mining), and land disposal sites (landfill sites, industrial impoundments). Non-point-source pollution, in contrast, originates from poorly defined, diffuse sources that typically occur over broad geographical scales. Examples of non-point-source pollution include agricultural runoff (pesticides, pathogens, and fertilizers), storm-water and urban runoff, and atmospheric deposition (wet and dry deposition of persistent organic pollutants such as polychlorinated biphenyls [PCBs] and mercury). Within each source, we identify the most important contaminants that have either been demonstrated to pose significant risks to human health and/or aquatic ecosystem integrity, or which are suspected of posing such risks. Examples include nutrients, metals, pesticides, persistent organic pollutants (POPs), chlorination by-products, and pharmaceuticals. Due to the significant number of toxic contaminants in the environment, we have necessarily restricted our discussion to those chemicals that pose risks to human health via exposure through drinking water. A comprehensive and judicious consideration of the full range of contaminants that occur in surface waters, sediments, and drinking water would be a large undertaking and clearly beyond the scope of this article. However, where available, we have provided references to relevant literature to assist the reader in undertaking a detailed investigation of their own. The information collected on specific chemicals within major contaminant classes was used to determine their relative risk using the hazard quotient (HQ) approach. Hazard quotients are the most widely used method of assessing risk in which the exposure concentration of a stressor, either measured or estimated, is compared to an effect concentration (e.g., no-observed-effect concentration or NOEC). A key goal of this assessment was to develop a perspective on the relative risks associated with toxic contaminants that occur in drinking water. Data used in this assessment were collected from literature sources and from the Drinking Water Surveillance Program (DWSP) of Ontario. For many common contaminants, there was insufficient environmental exposure (concentration) information in Ontario drinking water and groundwater. Hence, our assessment was limited to specific compounds within major contaminant classes including metals, disinfection by-products, pesticides, and nitrates. For each contaminant, the HQ was estimated by expressing the maximum concentration recorded in drinking water as a function of the water quality guideline for that compound. There are limitations to using the hazard quotient approach of risk characterization. For example, HQs frequently make use of worst-case data and are thus designed to be protective of almost all possible situations that may occur. However, reduction of the probability of a type II error (false negative) through the use of very conservative application factors and assumptions can lead to the implementation of expensive measures of mitigation for stressors that may pose little threat to humans or the environment. It is important to realize that our goal was not to conduct a comprehensive, in-depth assessment of risk for each chemical; more comprehensive assessments of managing risks associated with drinking water are addressed in a separate issue paper by Krewski et al. (2001a). Rather, our goal was to provide the reader with an indication of the relative risk of major contaminant classes as a basis for understanding the risks associated with the myriad forms of toxic pollutants in aquatic systems and drinking water. For most compounds, the estimated HQs were < 1. This indicates that there is little risk associated with exposure from drinking water to the compounds tested. There were some exceptions. For example, nitrates were found to commonly yield HQ values well above 1 in- many rural areas. Further, lead, total trihalomethanes, and trichloroacetic acid yielded HQs > 1 in some treated distribution waters (water distributed to households). These latter compounds were further assessed using a probabilistic approach; these assessments indicated that the maximum allowable concentrations (MAC) or interim MACs for the respective compounds were exceeded <5% of the time. In other words, the probability of finding these compounds in drinking water at levels that pose risk to humans through ingestion of drinking water is low. Our review has been carried out in accordance with the conventional principles of risk assessment. Application of the risk assessment paradigm requires rigorous data on both exposure and toxicity in order to adequately characterize potential risks of contaminants to human health and ecological integrity. Weakness rendered by poor data, or lack of data, in either the exposure or effects stages of the risk assessment process significantly reduces the confidence that can be placed in the overall risk assessment. (ABSTRACT TRUNCATED)

Probabilistic risk assessment of agrochemicals in the environment

Concern for the environment has resulted in greater scrutiny of both old and new plant protection products and increased e!orts have been directed to developing more rigorous but more realistic procedures for the ecotoxicological risk characterization of these agrochemicals. These techniques include probabilistic analysis of toxicity and exposure data and better understanding of the relationship between structure and function in populations of wildlife and the role of keystone species in maintaining ecosystem functioning. The ecological risk assessment method described here is centered on the use of probabilistic distribution functions that independently describe exposure concentrations and toxicological responses of organisms to the chemical of concern. The distributions are transformed to permit calculation of linear regression parameters. The regression parameters for the two distributions are then used to determine joint probabilities which interrelate the exposure and toxicology data. For ease of presentation the results are presented as an exceedence plot which depicts, based on the exposure data the percent of species likely to be a!ected and the percent of observations likely to cause this level of e!ect. In this paper, the use of the method is illustrated using data for chlorpyrifos in North American aquatic environments. These probabilistic risk assessment methods are being assessed for incorporation into assessment procedures in a number of regulatory jurisdictions. ( 2000 Elsevier Science Ltd. All rights reserved.

Effects of 25 pharmaceutical compounds to Lemna gibba using a seven-day static-renewal test

Antibiotics are known to have antichloroplastic properties, but their effects on aquatic higher plants are virtually unknown. In order to address this issue, 25 pharmaceuticals, including 22 antibiotics, were assessed for phytotoxicity to the aquatic higher plant Lemna gibba. A 7-d static-renewal test was used, and plants were treated with 0, 10, 30, 100, 300, and 1,000 microg/L of pharmaceutical-containing growth media. Phytotoxicity was assessed using multiple growth and biochemical endpoints. Effective concentration (EC)50, EC25, and EC10 values as well as tests for significant differences between treatments and controls lowest-observed-effect concentration (LOECs) were calculated for each endpoint. Twelve different classes of antibiotics were assessed; however, only members of the fluoroquinolone, sulfonamide, and tetracycline classes of antibiotics displayed significant phytotoxicity. The most toxic members of each of these classes tested were lomefloxacin, sulfamethoxazole, and chlortetracycline, with wet weight EC25 values of 38, 37, and 114 microg/L, respectively. Injury symptoms were comparatively uniform and consistent among chemical classes while degree of phytotoxicity varied considerably. Both of these criteria varied markedly between classes. Wet mass was consistently the most sensitive endpoint above 100 microg/L; conversely, frond number was the most sensitive below 100 microg/L. Pigment endpoints were significantly less sensitive than growth endpoints.

Chlorpyrifos: Ecological Risk Assessment in North American Aquatic Environments

The objective of this risk assessment was to determine the probability and significance of effects of the organophosphate insecticide, chlorpyrifos, on aquatic ecosystems in North America. The assessment addressed both agricultural and nonagricultural uses. However, the primary focus of the risk assessment was agricultural ecosystems, especially row crops and, in particular, the “corn-belt” agroecosystems. The risk assessment also addressed potential effects from other agricultural uses as well as urban uses such as turf, termiticide, and home use. Exposure and effects in freshwater and saltwater environments were considered. Aquatic invertebrates and fish were included in the assessment, but amphibians, reptiles, birds, and mammals were not. The potential exposure of these organisms is small because of a lack of biomagnification of chlorpyrifos. Thus, if their prey are not affected, it is unlikely that organisms at higher trophic levels would be adversely affected. Measurements of chlorpyrifos residues in fish have shown both low probability and low concentrations of exposure (USEPA 1992b). Insufficient data on amphibians were available for a direct assessment of risks. A risk assessment of chlorpyrifos in terrestrial ecosystems was conducted in parallel with this aquatic risk assessment (Kendall et al., in manuscript). Chlorpyrifos is not used in isolation, and residues of other substances with the same mechanism of action may co-occur with chlorpyrifos and some of these may display additive toxicity (Bailey et al. 1997). Although the presence of these compounds could influence the overall conclusions of a risk assessment for the class of anticholinesterase insecticides, the extensive resources necessary to conduct a classwide review were not available and they were excluded from this evaluation.

Aquatic persistence of eight pharmaceuticals in a microcosm study

The persistence of eight pharmaceuticals from multiple classes was studied in aquatic outdoor field microcosms. A method was developed for the determination of a mixture of acetaminophen, atorvastatin, caffeine, carbamazepine, levofloxacin, sertraline, sulfamethoxazole, and trimethoprim at microg/L levels from surface water of the microcosms using solid phase extraction and high-performance liquid chromatography-ultraviolet (HPLC-UV) and liquid chromatography tandem mass spectrometry (LC-MS-MS). Half-lives in the field ranged from 1.5 to 82 d. Laboratory persistence tests were performed to determine the relative importance of possible loss processes in the microcosms over the course of the study. Results from dark control experiments suggest hydrolysis was not important in the loss of the compounds. No significant differences were observed between measured half-lives of the pharmaceuticals in sunlight-exposed pond water and autoclaved pond water, which suggests photodegradation was important in limiting their persistence, and biodegradation was not an important loss process in surface water over the duration of the study. Observed photoproducts of several of the pharmaceuticals remained photoreactive, which led to further degradation in irradiated surface waters.

Ecological Risk Assessment for Aquatic Organisms from Over-Water Uses of Glyphosate

Although the herbicide glyphosate is most widely used in agriculture, some is used for the control of emergent aquatic weeds in ditches, wetlands, and margins of water bodies, largely as the formulation Rodeo. This article presents an ecological risk assessment (ERA) of glyphosate and some of the recommended surfactants as used in or near aquatic systems. Glyphosate does not bioaccumulate, biomagnify, or persist in a biologically available form in the environment. Its mechanism of action is specific to plants and it is relatively nontoxic to animals. As a commercial product, glyphosate may be formulated with surfactants that increased efficacy but, in some cases, are more toxic to aquatic organisms than the parent material. For this risk assessment, three model exposure scenarios--static or low-flow systems such as ponds, flowing waters such as streams, and systems subjected to tidal flows such as estuaries--were chosen and application rates from 1 to 8 kg glyphosate/ha were modeled. Additional measured exposure data from several field studies were also used. As acute exposures are most likely to occur, acute toxicity data were used as effect measures for the purposes of risk assessment. Toxicity data were obtained from the literature and characterized using probabilistic techniques. Risk assessments based on estimated and measured concentrations of glyphosate that would result from its use for the control of undesirable plants in wetlands and over-water situations showed that the risk to aquatic organisms is negligible or small at application rates less than 4 kg/ha and only slightly greater at application rates of 8 kg/ha. Less is known about the environmental fate and toxicology of the surfactants commonly used in combination with the Rodeo formulation of glyphosate. The surfactants used for this purpose were judged not to be persistent nor bioaccumulative in the environment. Distributional analysis of measured deposition concentrations of LI 700, suggest that this surfactant presents an insignificant acute risk to aquatic organisms. Assuming similar applications rates, significant ecological effects would not be expected from the use of some other surfactants such as Induce or X-77. Risks from the use of glyphosate +MON 0818 (Roundup) were slightly greater than those from glyphosate and surfactants such as LI 700; however, in over-water uses, risks were still considered small. Similar small risks were observed for measured concentrations of glyphosate in surface waters resulting from aerial application of Vision (a formulation equivalent to Roundup) to forestry areas in Canada. Concentrations measured after ground application presented a greater risk, but the data were sparse and the assessment is more uncertain.

Effects of Atrazine on Fish, Amphibians, and Aquatic Reptiles: A Critical Review

The herbicide atrazine is widely used in agriculture for the production of corn and other crops. Because of its physical and chemical properties, atrazine is found in small concentrations in surface waters—habitats for some species. A number of reports on the effects of atrazine on aquatic vertebrates, mostly amphibians, have been published, yet there is inconsistency in the effects reported, and inconsistency between studies in different laboratories. We have brought the results and conclusions of all of the relevant laboratory and field studies together in this critical review and assessed causality using procedures for the identification of causative agents of disease and ecoepidemiology derived from Kochs postulates and the Bradford–Hill guidelines. Based on a weight of evidence analysis of all of the data, the central theory that environmentally relevant concentrations of atrazine affect reproduction and/or reproductive development in fish, amphibians, and reptiles is not supported by the vast majority of observations. The same conclusions also hold for the supporting theories such as induction of aromatase, the enzyme that converts testosterone to estradiol. For other responses, such as immune function, stress endocrinology, parasitism, or population-level effects, there are no indications of effects or there is such a paucity of good data that definitive conclusions cannot be made.

Effects of pharmaceutical mixtures in aquatic microcosms.

Pharmaceuticals have a wide range of biological properties and are released into the environment in relatively large amounts, yet little information is available regarding their effects or potential ecological risks. We exposed outdoor aquatic microcosms to combinations of ibuprofen (a nonsteroidal anti-inflammatory drug), fluoxetine (a selective serotonin reuptake inhibitor), and ciprofloxacin (a DNA gyrase-inhibiting antibiotic) at concentrations of 6, 10, and 10 microg/L, respectively (low treatment [LT]); 60, 100, and 100 microg/L, respectively (medium treatment [MT]); and 600, 1,000, and 1,000 microg/L, respectively (high treatment [HT]). We maintained these concentrations for 35 d. Few responses were observed in the LT; however, effects were observed in the MT and HT. Fish mortality occurred in the MT (<35 d) and in the HT (<4 d). Phytoplankton increased in abundance and decreased in diversity (number of taxa) in the HT, with consistent trends being observed in the MT and LT. Zooplankton also showed increased abundance and decreases in diversity in the HT, with consistent trends being observed in the MT. Multivariate analyses for zooplankton and phytoplankton suggested interactions between these communities. Lemna gibba and Myriophyllum spp. showed mortality in the HT; growth of L. gibba was also reduced in the MT. Bacterial abundance did not change in the HT. All responses were observed at concentrations well below the equivalent pharmacologically active concentrations in mammals. Although the present data do not suggest that ibuprofen, fluoxetine, and ciprofloxacin are individually causing adverse effects in surface-water environments, questions remain about additive responses from mixtures.