Combined effects of multiple stressors: A field based response-surface experiment reveals variation in interactive effects across a range of concentrations and levels of organization

Background/Question/Methods

Ecosystems are affected by multiple anthropogenic stressors from a wide range of sources. This is problematic because multiple stressors can modify each others’ influences in ways that may vary depending on their respective intensities/concentrations. Such effects are poorly understood because they require large complex experiments that have hitherto been very difficult to run in the field. A novel system has recently been developed to deliver controlled doses of multiple stressors to simulate complex regimes of stress in numerous experimental plots at a marina near Dublin, Ireland.

Here we used this system to test whether copper and chlorpyrifos (a biocide) act independently or interactively on marine fouling communities and how their combined effects on organisms and ecosystem functions vary depending on their respective concentrations.

A response-surface design was used in which five environmentally relevant concentrations of each stressor were manipulated in factorial combinations for a total of 25 treatments, each replicated four times. Stressors were delivered for six weeks. Measurements were then made of chemical uptake (mass spectrometry) and impacts at several levels of biological organisation, from sub-cellular (cellular viability) to ecosystem (respiration and rates of water clearance).  Data were analysed using General Additive Models (GAMS).

Results/Conclusions

Combinations of stressors produced interactive effects that would not have been revealed without using the response-surface approach. For example, when copper and biocide were combined they interacted antagonistically, negating each others’ effects such that at intermediate concentrations of biocide and higher concentrations of copper, the cellular viability of mussels was no different from that in control plots. Concentrations of copper in the tissues of mussels were affected by a synergistic interaction with biocide, such that mussels exposed to low concentrations of biocide and high concentrations of copper accumulated twice as much copper than was accumulated with the highest concentration of copper alone.

Effects also varied among response variables, with ecosystem-level responses not always being predictable on the basis of sub-cellular responses. Although cellular viability was an effective predictor of ecosystem-level responses to individual stressors, its responses to multiple stressors did not match responses in terms of rates of clearance and respiration by the assemblage.

Our results highlight that non-additive effects of multiple stressors may be even more complex and more common than previously thought.  This field-based research also reveals that multiple stressors can reduce the ability of established biomarkers of pollution to indicate impacts at higher levels of organisation.