All of that anxiety in fish really does help their food chain function

Though the I'm late to the scene on this issue, it's become something of great interest to me in my studies of pharmaceuticals and personal care products (PPCP) in the environment. A confluence of analytical chemistry breakthroughs and environmental toxicology has produced concern regarding fluoxetine and other selective serotonin reuptake inhibitors in treated and untreated surface waters. The US Geological Survey began measuring prescription, nonprescription, and personal care products in streams nationally at sub-parts per billion concentrations (Kolpin et al., 2002). Emerging concern regarding the aquatic ecotoxicology of antidepressants discharged into effluents from municipal wastewater treatment plants began in 2003, when Bryan Brooks et al. at Baylor University was able to demonstrate effects of SSRIs on the survival and reproductive fitness of several species of aquatic biota (Brooks et al., 2003). Internationally, communities became concerned about environmental effects, as well as the potential for human exposures.

Only a few endpoints in few species are currently available for the effects of SSRIs on the behavior of fish. Hybrid striped bass (Morone saxatilis x M. chrysops) exposed to fluoxetine experienced concentration- and duration-dependent decreases in the ability to capture prey, evaluated as increased time to capture and eat prey (Gaworecki and Klaine, 2008). Interestingly, the bass were exposed for 6 days, and then allowed to recover for 6 days, and yet the effects persisted. The LOEC for decreased predator ability in the hybrid bass was 23.2 µg/L (Gaworecki and Klaine, 2008). These behavioral effects were anchored to brain concentrations of serotonin; serotonin levels were depressed to 24, 28, and 49% of control by 23.2, 51.4, and 100.9 µg/L fluoxetine by day 9 in the experiment, or half-way through the recovery period. By day 12, the levels of serotonin in the treated groups had recovered, but not completely to control levels. Thus, chronic, sublethal exposure to fluoxetine or other SSRIs may have lasting effects on serotonin and serotonin-regulated predator-prey activity in bass (Gaworecki and Klaine, 2008). The changes in behavior indicate a potential population effect, but behavior is not a regulatory endpoint (Gaworecki & Klaine, July 2008, PMID: 18547660; also see Scientific American article ).

Chronic 91-day exposure, from neonate to adulthood, of Western mosquitofish to fluoxetine resulted in behavioral changes including increased lethargy for concentrations greater than 0.5 µg/L fluoxetine. At 534 µg/L fluoxetine, mosquitofish stayed at the surface and layed on their sides with little to no swimming movement. At 53 µg/L fluoxetine, fish were less responsive, and less severe effects were observed in 5 and 0.5 µg/L fluoxetine. The LOEC for lethargy and decreased movement/responsiveness in this study is 0.5 µg/L fluoxetine, while the NOEC would be 0.05 µg/L fluoxetine (Henry and Black, 2008). It is difficult to decide if slightly decreased activity observed in this experiment is actually an adverse effect. Since moderate effects began at 53 µg/L fluoxetine, this appears to be the concentration eliciting an adverse effect.

Previously, increased endogenous serotonin levels in the brain of rainbow trout Oncorhynchas mykiss were demonstated to correlated with decreased competitition and increased submissive behavior (Overli et al., 2004). Changes in the brain serotonin levels of fish could potentially decrease this competitive ability, perhaps reflected in the “lethargy” noted by Henry & Black (2008) in mosquitofish. However, population effects of these behavioral changes are not currently understood and not easily extrapolated from this limited behavioral data set.

However, before anyone freaks about what SSRIs in surface waters might be doing to fish at both the individual and population level due to modulation of behavior, it's important to keep in context the concentrations being used. Fluoxetine, only one of five major SSRIs, has been measured at several national and international impacted sites, or sites suspected to have higher concentrations of wastewater pollutants. Further, the highest concentration of fluoxetine to date (published in the literature) was in groundwater, at 56 ng/L (Barnes et al 2008). Thus, the concentrations being used in the Gaworecki (2008) study and others are 1000 times greater than those currently present in the environment (as far as we know). The major uncertainty that could diminish this margin of safety is the behavior of the complex PPCP mixture. And at this point that's too complex to tackle.

But, one could, say, carry out a "concentration addition" of the highest reported environmental concentrations and use this to model the effects of all SSRIs in surface waters. This is an idea based on a study using daphnia and algae, where addition of all concentrations of the five SSRIs is appropriate to predict the effects of the mixture (Christensen et al., 2007; Henry and Black, 2007). Still, mixtures modeling is the most interesting area for SSRIs in water - how does the mixture act, and can low doses of all of the SSRIs add up to physiological impacts?