Pesticides are chemical substances that are used to kill, repel, or regulate the growth of biological organisms. This diverse group includes insecticides, herbicides, fungicides, nematicides, acaricides, rodenticides, avicides, wood preservatives, and antifoulants. The U.S. Environmental Protection Agency (EPA) recently estimated that > 1.2 billion pounds of pesticides are applied to crops, forests, residential areas, public lands, and aquatic areas in the United States each year (Kiely et al. 2004
). The release of these chemicals into the environment creates a potential for unintended adverse health impacts to both humans and nontarget wildlife.
The cumulative toxicologic impacts of pesticide mixtures is of particular concern for salmon and steelhead populations that are currently listed as either threatened or endangered under the U.S. Endangered Species Act (ESA 1973
). Many wild salmon stocks are in decline across much of the western United States [Nehlsen et al. 1991
; NOAA (National Oceanic and Atmospheric Administration) Fisheries 2008
]. Past salmon population extinctions (Nehlsen et al. 1991
) and current declines have been caused by decades of habitat degradation, overharvest, hydro-power operation, and hatchery practices (NRC 1996
). Major river systems that drain large agricultural and urban areas in California, Oregon, Washington, and Idaho provide freshwater habitat for ESA-listed salmon and steelhead (Figure 1
). Extensive surface water monitoring for pesticides, as part of the U.S. Geological Survey (USGS) National Water Quality Assessment program (NAWQA), has shown that current-use pesticides are frequently detected in these salmon-supporting river systems (Table 1
) (see also more recent monitoring studies by Carpenter et al. 2008
; USGS 2008
; Washington State Department of Ecology 2008
). Furthermore, pesticides almost always occur in mixtures with other pesticides. Analysis of NAWQA monitoring data found that > 90% of water samples from urban, agricultural, and mixed-use streams contained two or more pesticides (Gilliom 2007
). The toxicologic effects of these mixtures on the health of salmon are largely unknown.
In the years since the enactment of the FQPA, the U.S. EPA has identified several classes of pesticides that share a common mode of action (U.S. EPA 2002
). Among these are the organophosphate (OP) and N
-methyl carbamate (CB) insecticides. These two classes of chemicals inhibit the enzyme acetylcholinesterase (AChE), thereby interfering with cholinergic neurotransmission in both humans (Chambers 1992
) and fish (Fulton and Key 2001
). Because anticholin-esterase agents share a common mode of toxic action, the National Academy of Sciences recommended a dose-additive approach to assessing risks to human infants and children (NRC 1993
). Dose addition (or, for waterborne exposures to fish, concentration addition) assumes that the cumulative toxicity of the mixture can be estimated from the sum of the individual toxic potencies of each individual component chemical. This is how the U.S. EPA currently assesses the potential toxicity of mixtures of OP and CB insecticides in the context of the FQPA.
The assumption of dose addition or concentration addition for mixtures of anti-cholinesterase pesticides has also been extended to aquatic life (Junghans et al. 2006
). In salmon, concentration-additive inhibition of brain AChE activity by mixtures of OP and CB insecticides was recently demonstrated in vitro
(Scholz et al. 2006
). However, the in vivo
toxicity of anti cholinesterase mixtures may deviate from concentration addition if the individual chemicals in a mixture interact via toxicokinetic or toxicodynamic processes to produce either antagonistic or synergistic effects (Borgert et al. 2004
). Each of these possible outcomes (antagonism, addition, or synergism) has potentially important implications for the current regulatory paradigm, wherein risks of pesticides to ESA-listed salmonids are assessed based primarily on responses to single active ingredients. To define the extent to which OP and CB insecticides in mixtures interact, we exposed juvenile coho salmon (Oncorhynchus kisutch
) to all possible binary combinations of the OP insecticides diazinon, malathion, and chlorpyrifos and the CB insecticides carbaryl and carbofuran. We used the concentration–response curves for AChE inhibition by individual chemicals to statistically define concentration addition (i.e., no interaction within a mixture).
We have shown that in vivo
exposures to binary mixtures of OP and CB pesticides produced additive or synergistic AChE inhibition in the brains of juvenile coho salmon. The statistical departure from dose addition occurred for several chemical combinations at each of the three relative exposure concentrations, with a trend toward a higher incidence of synergism at the higher exposures. Where the degree of synergism was severe (e.g., for pairings of diazi-non and malathion), enhanced AChE inhibition (i.e., > 90%) corresponded to overt signs of anticholinesterase intoxication and death. This result is consistent with previous (single-chemical) OP and CB pesticide toxicity studies in other fish species (reviewed by Fulton and Key 2001
). At present, diazinon, chlorpyrifos, malathion, carbaryl, and carbofuran are some of the most extensively used insecticides in California and the Pacific Northwest (California Department of Pesticide Regulation 2008
). The frequency with which these chemicals are detected in some salmon habitats (Table 1
) and their combinatorial toxicity to juvenile salmon when they occur as mixtures suggest they may be limiting the recovery of several threatened and endangered populations.
The OP (oxon metabolites) and CB insecticides examined in this study do not interact in vitro
, where their combinatorial inhibition of salmon AChE can be explained by simple concentration addition (Scholz et al. 2006
). The departure from concentration addition for some pesticide pairings in vivo
is consistent with OP and CB insecticides acting on other biochemical targets. Although more work is needed to identify these targets, car-boxylesterases (CaEs) are candidate enzymes that may underlie the chemical interactions observed in this study. CaEs play an important role in the detoxification of many pesticides, including the OP and CB insecticides, via hydrolysis (Jokanovic 2001
; Wheelock et al. 2005a
). CaEs may also functionally protect AChE from insecticide toxicity by direct binding and sequestration, thereby preventing or delaying interaction between the insecticide and AChE (Jokanovic 2001
; Maxwell 1992a
). Mammalian studies spanning several decades have shown that anticholinesterase toxicity increases when CaE enzyme activity is inhibited (Casida et al. 1963
; Jokanovic 2001
; Maxwell 1992b
; Su et al. 1971
). Although few studies are documented in fish, exposures to OP and CB pesticides have been found to reduce liver CaE activity in salmonids (Ferrari et al. 2007
; Wheelock et al. 2005b
), with the OP chlorpyrifos acting as a more potent inhibitor of CaE activity than AChE activity (Wheelock et al. 2005b
). In another aquatic species (Daphnia magna
), pharmacologic inhibition of CaE significantly enhanced the toxicity of chlorpyrifos, malathion, and carbofuran (Barata et al. 2004
). Thus, although other biochemical targets may be involved in OP and CB synergism (Casida and Quistad 2004
), future mechanistic studies should give particular consideration to the role of CaEs in the pesticide interactions observed in this study.
To identify interactions between pesticides in mixtures, it was first necessary to normalize each concentration–response curve using the calculated EC50
concentration for that individual chemical. For all five insecticides, the concentrations that produce 50% brain AChE inhibition in salmon (Table 2
) are approximately 10- to 1,000-fold higher than the levels typically detected in surface water monitoring investigations (Hoffman et al. 2000
). However, we show here that many insecticide combinations produce additive toxicity at low, environmentally relevant concentrations (0.1 EC50
; Figure 3A
). Moreover, certain combinations showed a clear pattern of synergism even at these relatively low levels. For example, diazinon and chlorpyrifos were synergistic when combined at 7.3 μg/L and 0.1 μg/L, respectively. Surface water monitoring in the San Joaquin basin in California (Dubrovsky et al. 1998
) reported diazinon concentrations as high as 6.0 μg/L and chlorpyrifos levels up to 0.5 μg/L. The pairing of diazinon (7.3 μg/L) with malathion (3.7 μg/L) produced severe (> 90%) AChE inhibition as well as classical signs of anticholinesterase poisoning. Thus, for some chemical combinations, synergism is likely to occur at exposure concentrations below the lowest levels used in the present study. Although more work is needed to determine the lower bounds for pesticide interactions, this study indicates that synergism is likely to occur at concentrations that have been directly measured in habitats supporting threatened and endangered salmonids.
In quantitative terms, we have shown that an in vivo
screen for interactions between anticholinesterase insecticides is tractable in juvenile salmon. Although we examined only five pesticides, it would be straightforward to establish concentration–response relationships for AChE inhibition for the remaining OP and CB insecticides in current use. Given default assumptions of common mode of action and concentration addition (Lambert and Lipscomb 2007
), the relative potency of each insecticide could then be used to estimate the joint toxicity of chemicals in a mixture using a conventional toxic unit approach (Junghans et al. 2006
). Widely used insecticides and those with a relatively high toxic potency (e.g., the OP azinphos-methyl) could also be screened for interactions with other insecticides at low, environmentally realistic exposure concentrations. Where synergism occurs, additional safety factors could then be assigned to protect the health of threatened and endangered salmon. With the exception of safety factors for synergism, this process is similar to how the FQPA mandates evaluating the human health risks of OP and CB mixtures (FQPA 1996
Although habitat degradation is generally accepted to be a major causal factor in salmon declines (NRC 1996
), the specific contributions of current-use pesticides to the decline of salmon populations are not well understood. One key challenge to understanding this relationship is linking pesticide effects on individual fish to the intrinsic productivity of populations. Recent data by Sandahl et al. (2005)
began to address this challenge by showing that exposures to low, environmentally realistic concentrations of chlorpyrifos produced reductions in AChE activity that were closely correlated to reductions in swimming speed and feeding rates. Reductions in feeding are likely to lead to reductions in the size of exposed salmon at the time of their seaward migration, an end point that has been shown to be an important determinant of individual salmon survival (Higgs et al. 1995
; Zabel and Achord 2004
). By reducing survival rates, sublethal inhibition of AChE in juvenile salmon could potentially reduce the intrinsic productivity of salmon populations. Because mixtures of OP and CB insecticides produce dose-additive or synergistic AChE inhibition, they could magnify these population-scale effects.
The link to populations is important because most of the ongoing recovery planning for ESA-listed salmon is focused at this biological scale (Ruckelshaus et al. 2002
). Although many salmon habitats are affected by agrochemicals and urban runoff, restoration priorities are usually developed without the specific inclusion of toxics in quantitative analyses of limiting factors (Bartz et al. 2006
; Burnett et al. 2007
; Hoekstra et al. 2007
; Scheuerell et al. 2006
). In the larger context of salmon conservation, a future priority will be to establish a quantitative connection between the mixture toxicity observed in this study and higher biological scales via effects on growth and survival. This connection will help to bridge the disciplines of ecotoxicology and conservation biology (Hansen and Johnson 1999
) in their common goal of guiding the recovery of threatened and endangered species.