Rethinking our approach to multiple stressor studies in marine environments

Abstract

Marine systems experience an unprecedented number of stresses caused by humans. Over the last 25 yr an increasing amount of attention has been given to examining the combined impacts of multiple stressors. Yet, existing studies reveal few patterns that facilitate predicting or understanding when multiple stressors should combine additively, synergistically, or antagonistically. One contributing factor to this lack of clarity may be the lack of a common framework that is based on a mechanistic understanding of stressor impacts. We adapt and advocate a general framework that is employed by the US EPA in terrestrial systems for use in marine systems. This framework involves 3 steps: (1) Mechanistically examine the impacts of multiple stressors on individual organisms. (2) Scale these impacts on individual organisms to population level responses to multiple stressors. (3) Examine context-dependent changes in stressor responses due to changes in community or ecosystem properties. We also argue that 3 specific aspects of previous studies hamper our ability to detect patterns in multiple stressor impact. First, a large number of studies have reported impacts on growth, survival, etc., without elucidating mechanisms. Second, the majority of studies provide insufficient data to determine whether threshold or nonlinear responses to stressors occur. Third, 32% of existing studies transformed data to meet model assumptions, but in so doing, they unknowingly altered the statistical model being tested. We argue that rectifying these 3 conditions will accelerate the detection of patterns in the way that multiple stressors combine to influence marine systems.

Implementing and Innovating Marine Monitoring Approaches for Assessing Marine Environmental Status

“…These stressors can have synergistic effects on marine ecosystems (Mora et al, 2013;Griffen et al, 2016), altering their functioning and ability to provide goods and services (Worm et al, 2006;Crain et al, 2008). Their impact is expected to be even stronger in enclosed and semi-enclosed basins with high population density, tourism flow and maritime activities (Danovaro, 2003).…”

Managing the Marine Environment, Conceptual Models and Assessment Considerations for the European Marine Strategy Framework Directive

“…For a single, specific Pressure, the relationship between Pressure and Impact varies according to the degree of Pressure (e.g., spatial extent, duration and/or frequency, intensity), the habitat type upon which the Pressure is acting, the component species and those species in the wider ecosystem which they support. This produces many potential Pressure-State change trajectories that increase in complexity with concurrent potentially synergistic or antagonistic combinations of Activities and Pressures (Griffen et al, 2016). Hence the need to move from a conceptual framework to "nested horrendograms" to encompass the interlinked complexity (e.g., Elliott et al, 2015).…”

“…The multiple Pressures will rarely be equal and will lead to cumulative and incombination effects which may be synergistic or antagonistic (Griffen et al, 2016). To indicate some of difficulties in assessing cumulative impacts, Crain et al (2008) analyzed 171 multiple stressors studies in marine and coastal environments and found effects to be 26% additive, 36% synergistic, and 38% antagonistic, while interaction type varied by response level, trophic level, and specific stressors.…”

“…In another meta-analysis of 112 experimental studies Darling and Côté (2008) found similar combined effects of two stressors with 23% additive, 35% synergistic and 42% antagonistic. Despite the lack of knowledge at the community and ecosystem level elucidating or predicting effects of combinations of individual Pressure impacts, we can measure the status of an ecosystem that is impacted by multiple Pressures (Griffen et al, 2016;Nõges et al, 2016;Teichert et al, 2016). Hence, while we can identify some elements, we are unclear regarding the precise changes at a sub-species, species, population or community level.…”

Coral responses to temperature, irradiance and acidification stress: linking physiology to satellite remote sensing

“…In a multiple-factor ANOVA on untransformed data, interaction terms refer to additive interactive effects between or among factors (Griffen et al 2016). To check for a multiplicative interactive effect of pCO 2 and temperature on symbiont cell densities, an ANOVA was performed on natural logarithm-transformed symbiont cell densities.…”