Abstract
In nature, organisms are continually exposed to multiple stressors and contaminants. However, most toxicity studies are performed in controlled laboratory settings with a single contaminant and do not account for interactions with co-occurring natural stressors, such as predation. Interactions can range from antagonistic to synergistic, and multiple stressor studies can therefore give a more realistic understanding of the effect of anthropogenic stressors on natural populations. Excess copper is toxic to aquatic organisms, and copper pollution is an increasing challenge along the Norwegian coast due to its use as an antifouling agent in aquaculture. In the copepod Tigriopus brevicornis, a perceived predation risk potentiated the effect of copper toxicity on development time. In this study I investigate the combined effect of predation risk and copper toxicity on survival in five other species of copepods: Acartia clausii, Pseudocalanus sp., Temora longicornis, Tisbe sp. and Triconia sp. I hypothesized that mortality from copper increases by simultaneous exposure to predator cues, and that the benthic Tisbe sp. is less sensitive to copper than the four pelagic species. I monitored individual survival during 48 h of exposure to copper (0, 150, 450 and 1350 µg/L), alone and in combination with predator cues from three-spined stickleback. Use of an automated imaging robot allowed me to estimate time of death with a time resolution of 25 min. The data set was analysed using the reduced General Unified Threshold model for Survival (GUTS) as well as with a traditional survival model. Contrary to my expectations, predation risk reduced mortality from copper in all species. Reduced bioavailability of copper appears to be an important part of this effect. When accounting for bioavailability, the GUTS models still suggested an additional effect of predation risk, but the direction and mechanism of this effect were inconclusive. Species sensitivity differences were best explained by different dominant rate constants in GUTS. As expected, Tisbe sp. was the least sensitive, and three of the pelagic species were the most sensitive. Triconia sp. was surprisingly tolerant within the tested time frame, but this appears to be due to slow kinetics and not intrinsic tolerance. This result shows the need for time-independent measures of toxicity like GUTS parameters for sensitivity comparisons. My results show that bioavailability may be a complicating factor in predicting multiple stressor effects and that mechanistic modeling with GUTS can provide deeper insights into the processes underlying sensitivity differences than traditional toxicity measures.