Mimicking natural systems: Changes in behavior as a result of dynamic exposure to naproxen

Highlights

• Exposure to the pharmaceutical naproxen decreases crayfish aggression.

• Behavioral changes occur at sublethal concentrations found in freshwater systems.

• Dynamic exposure impacts aggressive behavior more than static exposure tests.

• Current toxicity assessments may not be sufficient for flowing freshwater systems.

Abstract

Animals living in aquatic habitats regularly encounter anthropogenic chemical pollution. Typically, the toxicity of a chemical toxicant is determined by the median lethal concentration (LC₅₀) through a static exposure test. However, LC₅₀ values and static tests do not provide an accurate representation of exposure to pollutants within natural stream systems. In their native habitats, animals experience exposure as a fluctuating concentration due to turbulent mixing, temporal variations of contamination (seasonal inputs), and contaminant input type (point vs. non-point). Research has shown that turbulent environments produce exposures with a high degree of fluctuation in frequency, duration, and intensity. In order to more effectively evaluate the effects of pollutants, we created a dynamic exposure paradigm, utilizing both flow and substrate within a small mesocosm. A commonly used pharmaceutical, naproxen, was used as the toxicant and female crayfish (Orconectes virilis) as the target organism to investigate changes in fighting behavior as a result of dynamic exposure. Crayfish underwent either a 23 h long static or a dynamic exposure to naproxen. Following exposure, the target crayfish and an unexposed size matched opponent underwent a 15 min fight trial. These fight trials were recorded and later analyzed using a standard ethogram. Results indicate that exposure to sublethal concentrations of naproxen, in both static and flowing conditions, negatively impact aggressive behavior. Results also indicate that a dynamic exposure paradigm has a greater negative impact on behavior than a static exposure. Turbulence and habitat structure play important roles in shaping chemical exposure. Future research should incorporate features of dynamic chemical exposure in order to form a more comprehensive image of chemical exposure and predict the resulting sublethal effects from exposure. Possible techniques for assessment include utilizing flow-through experimental set-ups in tandem with behavioral or physiological endpoints as opposed to acute toxicity. Other possibilities of assessment could involve utilizing fine-scale chemical measurements of pollutants to determine the actual concentrations animals encounter during an exposure event.

Introduction

The toxicity of a chemical is typically determined by its median lethal concentration (LC₅₀), which is the concentration at which half of a test population dies from exposure after a specified amount of time (Stephan, 1977). Traditionally, these exposure trials are conducted using static concentrations of pollutants (Anderson et al., 1980, Diamond et al., 2006, Lovern and Klaper, 2006; Williams and Dusenbury, 1990). A recent addition to this style of ecotoxicology has recognized that toxicant concentrations in natural systems are dynamic in time (Ashauer et al., 2006, Handy, 1994, Hoang et al., 2007, Wang, 2011). The exact dynamic nature of toxicant concentrations in freshwater systems is due to a number of factors including pulsed addition of a toxicant to an environment (Stoeckel et al., 2012), precipitation events that deliver toxicants to aquatic systems (Hangen et al., 2001), and the movement of toxicants through a system due to flow mechanics (Sanford, 1997). Because aquatic habitats vary considerably in structure, substrate, and the turbulent nature of flow, exposure in fluvial systems will be site specific and more challenging to replicate (Moore et al., 1994, Moore et al., 2000, Moore and Crimaldi, 2004). Newer studies have incorporated a pulsatile testing paradigm where model organisms are exposed for a shorter length of time to a single concentration multiple times (Cold and Forbes, 2004, Diamond et al., 2005, Earl and Whiteman, 2009, Zhao and Newman, 2006). Although an improvement in regard to a more naturalistic exposure paradigm compared to static trials, which do not account for the potential of multiple exposures to toxicants within a given time frame, toxicity values derived from static or pulsatile trials do not take into account environmental factors that alter the intensity, frequency, and duration of exposure.

The concentration of a toxicant in the environment will fluctuate with respect to intensity and duration as a result of the three major factors listed above that create temporal variation in toxicant concentrations. Intensity of a toxicant can change in response to an influx of a chemical into a lotic system or by chemical mixing (Miller, 1997). Frequency and duration can both alter toxicant concentrations as a result of habitat structure preventing continuous transport of toxicants downstream (Finelli, 2000). Given this level of understanding on the dynamic nature of toxicant concentration in natural freshwater systems (Murlis et al., 1992, Murlis and Jones, 1981), LC₅₀ values derived from static or even slow pulsatile tests do not accurately represent potential threats experienced by organisms under ecologically relevant conditions (Chapman, 2002). Given the disparity between previous LC₅₀ work with static concentrations over extended periods of time and the types of exposure (temporally and spatially dynamic) experienced by organisms in natural systems, a significant gap in understanding arises (Gordon et al., 2012). The current definitions and assessments of exposure are ill-formed when applied to a field setting. Exposure (as a set concentration of ppm or ppm over a time period) is not relevant to the spatial and temporal scales at which organisms experience toxicants, primarily as a result of the fluid mechanics of dispersal of chemicals (Murlis et al., 1992). Thus, to understand exposure from an ecological point of view, understanding the role of flow in dispersing chemicals in any singular habitat is crucial (Hart and Finelli, 1999, Sanford, 1997). Fortunately, the fluid dynamics (air or water) are fairly well understood and a theoretical concept can be developed.

Both terrestrial and aquatic systems are composed of fluid mediums that move toxicants through their environment (Denny, 1993). Fluid movement determines both the amount of time that toxic chemicals reside within a habitat and the fine-scale fluctuations in concentration that occur as a result of dispersal (Sanford, 1997, Vogel, 1994). Two processes determine the temporal dynamics of chemical dispersion in habitats: molecular diffusion and bulk advection (Denny, 1993). Given the time (<days) and size scales (>0.1 mm) under consideration for macroscopic organisms, diffusion can be ignored and the distribution of toxicants in an environment is due to bulk advection. Bulk advection can be subdivided into laminar and turbulent flow (Vogel, 1994). Based on the physical conditions of most habitats (wind or water speed and roughness elements), toxic exposure paradigms are almost universally turbulent and in turn, the distribution of toxicants within an environment are due to the mechanics associated with turbulence.

Because nearly all dispersion and transport of fluids in natural freshwater systems happens due to turbulence, flow can be considered of equal importance to other abiotic factors, such as light or temperature (Sanford, 1997). Additionally, the physical characteristics of aquatic habitats, including the topography of the stream bed and substrate composition impact the turbulent structures contained within the flow, creating unique landscapes over which odor and chemical plumes are dispersed (Fischer et al., 1979, Johnson and Rice, 2014, Moore and Crimaldi, 2004, Wolf et al., 2009). Given that each habitat is unique in these two fundamental characteristics, the residence time and concentration of toxicants will vary significantly across habitats (Edwards and Moore, 2014). Consequently, animals present within highly turbulent habitats will be exposed to chemical plumes that also vary significantly (Finelli, 2000, Moore and Atema, 1991, Weissburg, 2010). As motile aquatic organisms navigate various habitats, they may also encounter several chemical landscapes which consequently impacts the level of chemical exposure. Chemical landscapes that result from turbulent mixing have been shown to produce habitat-specific behavioral responses, indicating that aquatic organisms alter their behavior accordingly with the structure of the chemical plumes they encounter (Moore and Grills, 1999, Weissburg and Zimmer-Faust, 1994).

Thus, exposure in natural systems is more complex than static trials can model, since chemical plume structure has been linked with hydrodynamics (Edwards and Moore, 2014, Finelli et al., 1999, Finelli et al., 2000, Moore and Grills, 1999, Webster and Weissburg, 2009, Weissburg et al., 2002, Wolf et al., 2009). Chemical distributions will not be homogeneous as static tests suggest; they are patchy and change as they move downstream and interact with objects. As a result of the physical dispersion of chemicals by turbulence, two animals in different locations within any flowing habitat will be subjected to different and heterogeneous exposures from the same chemical plume (Weissburg, 2010). However, current toxicity assessment does not account for variations in exposure due to turbulence or habitat structure, despite ongoing discussion about the need for exposure tests that are more representative of fluctuating chemical concentrations (Diamond et al., 2005, Handy, 1994, Zhao and Newman, 2006). Rethinking chemical exposure becomes especially important when assessing toxicants that are both prevalent in flowing freshwater systems and potent in low doses. Therefore, construction of an exposure paradigm that mimics flowing habitats is necessary in assessing the sublethal effects of chemical exposure.

Within the last twenty years, pharmaceutical compounds, (i.e., prescription drugs, hormones, and over the counter medications) have been reported to contribute to freshwater pollution, primarily as effluent from sewage treatment plants (Arnold et al., 2014, Santos et al., 2010). Over 600 different pharmaceuticals have been detected in several aquatic systems including surface water, effluents from sewage treatment plants, groundwater, and drinking water (Hughes et al., 2013, Meredith-Williams et al., 2012, Segura et al., 2009). Pharmaceuticals are biologically active compounds designed to modify the physiology, and possibly the behavior, of their intended target without killing those targets (Boxall et al., 2012, Meredith-Williams et al., 2012, Monteiro and Boxall, 2010). Because they are designed to be therapeutically effective with little degradation, pharmaceuticals enter freshwater habitats pharmacologically active. Pharmaceuticals pose a significant threat to wildlife, both due the multiple pathways of release into freshwater habitats and a lack of knowledge about their uptake and effects in aquatic organisms (Brodin et al., 2014, Carlsson et al., 2006, Fick et al., 2009, Shore et al., 2014). Studies have found that aquatic organisms exhibit a variety of behavioral responses to pharmaceutical pollution and that these responses can occur at sublethal concentrations (Berninger et al., 2011, Melvin and Wilson, 2013, Perreault et al., 2003). This sublethal effect can have broader ecological consequences to aquatic ecosystems and should be further investigated as pharmaceuticals are an emerging class of contaminants.

The purpose of this study was to investigate how sublethal concentrations of chemicals dispersed by turbulence impact an organism's behavior. Naproxen, a non-steroidal anti-inflammatory drug (NSAID), commonly used for pain relief was selected for this study. Naproxen has been detected in varying concentrations in freshwater bodies across the globe and the intended therapeutic use, side effects, and detection in stream effluents make naproxen a relevant pharmaceutical to investigate (Carballa et al., 2008, Carlsson et al., 2006, Corcoran et al., 2010, Isidori et al., 2005, Nikolaou et al., 2007, Straub and Stewart, 2007, Ternes, 1998).

Crayfish were used due to their ecological importance in stream environments (Momot, 1995, Usio and Townsend, 2004). Crayfish are sensitive to pollution, making them important bioindicators of overall ecosystem health. Additionally, crayfish are found in a broad range of aquatic habitats, indicating a high probability of exposure to anthropogenic chemicals. Within stream communities, they play a significant role at multiple trophic levels, acting both as predators and detritivores (Dorn and Wojdak, 2004, Parkyn et al., 2001, Schofield et al., 2001). Consequently, the absence of crayfish from a stream has the potential to negatively affect the other organisms present. Crayfish behaviors, including aggression, have been extensively studied (Bergman and Moore, 2003, Blake and Hart, 1993, Davis and Huber, 2007, Lahman et al., 2015). Agonistic interactions among crayfish are ecologically important, as fighting is key in determining social hierarchies (Goessmann et al., 2000). This social behavior is an important component to acquisition of resources (food and shelter), as well as mate selection. Because a well-established fight ethogram exists (Bergman and Moore, 2003) and crayfish frequently engage in agonistic behaviors (Bergman et al., 2003, Moore and Grills, 1999, Wolf and Moore, 2002), they are a model organism to assess behavioral changes due to chemical exposure.

In order to assess behavioral changes due to chemical exposure, crayfish were exposed to various concentrations of naproxen under one of two exposure paradigms: static and dynamic. Our hypothesis was that crayfish in the dynamic exposures would show greater impairments to all concentrations of naproxen due to the large-scale fluctuations in exposure as compared to the stable concentrations in static conditions.