Pharmaceuticals are increasingly being detected in aquatic ecosystems worldwide. Particularly concerning are pharmaceutical pollutants that can adversely impact exposed wildlife, even at extremely low concentrations. One such contaminant is the widely prescribed antidepressant fluoxetine, which can disrupt neurotransmission and behavioural pathways in wildlife. Despite this, relatively limited research has addressed the behavioural impacts of fluoxetine at ecologically realistic exposure concentrations. Here, we show that 28-day fluoxetine exposure at two ecologically relevant dosages—one representing low surface water concentrations and another representing high effluent flow concentrations—alters antipredator behaviour in Eastern mosquitofish (Gambusia holbrooki). We found that fluoxetine exposure at the lower dosage resulted in increased activity levels irrespective of the presence or absence of a predatory dragonfly nymph (Hemianax papuensis). Additionally, irrespective of exposure concentration, fluoxetine-exposed fish entered the predator ‘strike zone’ more rapidly. In a separate experiment, fluoxetine exposure reduced mosquitofish freezing behaviour—a common antipredator strategy—following a simulated predator strike, although, in females, this reduction in behaviour was seen only at the lower dosage. Together, our findings suggest that fluoxetine can cause both non-monotonic and sex-dependent shifts in behaviour. Further, they demonstrate that exposure to fluoxetine at environmentally realistic concentrations can alter antipredator behaviour, with important repercussions for organismal fitness.

Contamination of aquatic habitats with pharmaceuticals is a major environmental concern. Recent studies have detected pharmaceutical pollutants in a wide array of ecosystems and organisms, with many of these contaminants being highly resistant to biodegradation and capable of eliciting sub-lethal effects in non-target species. One such pollutant is fluoxetine, a widely prescribed antidepressant, which is frequently detected in surface waters globally and can alter physiology and behaviour in aquatic organisms. Despite this, relatively little is known about the potential for fluoxetine to disrupt mechanisms of sexual selection. Here, we investigate the impacts of 30-day exposure to two environmentally realistic levels of fluoxetine (low and high) on mechanisms of pre- and post-copulatory sexual selection in the eastern mosquitofish (Gambusia holbrooki). We tested 1) male mating behaviour in the absence or presence of a competitor male, and 2) sperm quality and quantity. We found that high-fluoxetine exposure increased male copulatory behaviour in the absence of a competitor, while no effect was detected under male-male competition. Further, fluoxetine exposure at both concentrations increased total sperm count relative to males from the control group, while no significant change in sperm quality was observed. Lastly, low-fluoxetine males showed a significant reduction in condition index (mass relative to length). Our study is the first to show altered mechanisms of both pre- and post-copulatory sexual selection in an aquatic species resulting from environmentally realistic fluoxetine exposure, highlighting the capacity of pharmaceutical pollution to interfere with sensitive reproductive processes in wildlife.

The capacity of pharmaceutical pollution to alter behaviour in wildlife is of increasing environmental concern. A major pathway of these pollutants into the environment is the treatment of livestock with hormonal growth promotants (HGPs), which are highly potent veterinary pharmaceuticals that enter aquatic ecosystems via effluent runoff. Hormonal growth promotants are designed to exert biological effects at low doses, can act on physiological pathways that are evolutionarily conserved across taxa, and have been detected in ecosystems worldwide. However, despite being shown to alter key fitness-related processes (e.g., development, reproduction) in various non-target species, relatively little is known about the potential for HGPs to alter ecologically important behaviours, especially across multiple contexts. Here, we investigated the effects of exposure to a field-realistic level of the androgenic HGP metabolite 17β-trenbolone—an endocrine-disrupting chemical that has repeatedly been detected in freshwater systems—on a suite of ecologically important behaviours in wild-caught female eastern mosquitofish (Gambusia holbrooki). First, we found that 17β-trenbolone-exposed fish were more active and exploratory in a novel environment (i.e., maze arena), while boldness (i.e., refuge use) was not significantly affected. Second, when tested for sociability, exposed fish spent less time in close proximity to a shoal of stimulus (i.e., unexposed) conspecific females and were, again, found to be more active. Third, when assayed for foraging behaviour, exposed fish were faster to reach a foraging zone containing prey items (chironomid larvae), quicker to commence feeding, spent more time foraging, and consumed a greater number of prey items, although the effect of exposure on certain foraging behaviours was dependent on fish size. Taken together, these findings highlight the potential for exposure to sub-lethal levels of veterinary pharmaceuticals to alter sensitive behavioural processes in wildlife across multiple contexts, with potential ecological and evolutionary implications for exposed populations.