Few ecological studies have examined the effect of multiple simultaneous stressors on individual species and ecosystems . The statistical approach used here does not allow the separation of individual stressor effects, although some inference in terms of the most critical factors can be made from the model metrics (L-ratios) and from the model visualisations. We found complex interactions between the effects of temperature, CO2 concentration and macrofaunal species richness/species assemblage composition on ecosystem response, measured as MPB biomass. The MPB biomass provides a proxy estimate of the productive potential of mudflat systems  and has been used to model net system productivity. Thus, changes in MPB biomass are a crucial element of system performance and will have reverberating effects through the higher trophic levels [8, 38], just as changes in infaunal diversity will affect resource utilisation .
In the present study, MPB biomass was higher at low temperature across all CO2 regimes, and did not increase as CO2 increased, indicating that CO2 was not limiting to MPB under these conditions and that increasing temperature was detrimental. The models suggested that species richness/composition and temperature were more influential on MPB than CO2, but the interactions between the explanatory variables were significant. The interactive effects of species composition and temperature led to a significant reduction in MPB biomass. The implication of this for temperate estuaries may be quite profound. The overall metabolic balance of estuarine systems between net autotrophy or heterotrophy is under debate [28, 36] but the role of autotrophic production by MPB is clear. Reduction of this contribution to the carbon balance will push the system toward a more heterotrophic condition. A shift from autotrophic to heterotrophic conditions, or a shift to more extreme heterotrophy, is likely to have significant, but as yet undetermined, implications for ecosystem goods and services. Unless the net import of allochthonous carbon changes, then overall productivity may be expected to decline with potential effects on resource utilisation. Variation in temperature has already been shown to affect the carbon metabolism of coastal systems  and this supports our hypothesis that autochthonous productivity may be reduced through interaction between the climate change variables and species diversity.
In this study, the presence of macrofaunal species resulted in substantial decreases in MBP biomass, and there appeared to be a general decline in MPB with increasing macrofaunal species richness, specifically at the lowest temperature. This general trend is partly to be expected as all three macrofaunal species are known consumers of MPB [34, 60]. However, the presence of C. volutator had a disproportionately strong effect in reducing MPB biomass, consistent with previous research [34, 42, 61]. Rather than being attributed to consumption, the mechanism for this dominant effect is likely to be due to the constant resuspension of sediment [42, 62] during grazing and bioturbation (Additional file 1, Figure S8), leading to inhibition of photosynthesis by MPB, and also the prevention of MPB biofilm formation on the sediment surface .
Individual species responses to climate change are often highly uncertain , and environmental change could alter the balance between the functional groups present (through extinction, invasion or changes in abundance or behaviour), as well as the number and identity of species present in an ecosystem . This makes it very difficult to predict how an ecosystem may respond based on diversity alone . MPB utilise nutrients from the water column and pore waters whilst bioturbation by invertebrates is known to increase the flux rates and concentrations (NH4-N, PO4-P) available . Therefore, species-specific responses to climate change will affect more than one trophic level, and the nature of interactions between species will change as a consequence [8, 11, 20]. In the present study, the decline in MPB was driven by complex interactions between environmental variables and diversity effects, and mediated through infaunal grazing activity. In this case, there was no apparent compensation through an increase in nutrients caused by bioturbatory activity. The functional importance of species is also context-dependent, and functional impact may alter as conditions change. Thus, while functional diversity is important, it may be overshadowed by species identity as different species take up more prominent roles under changing scenarios. In our experiment, the dominant effect of one species (C. volutator) illustrates how the extinction of influential (rare or common) species may have profound effects on the ecosystem - and that these effects may be direct or indirect [67, 68].
Many studies focusing on the ecological consequences of altered biodiversity within the context of specific drivers of environmental change have concentrated on single variables and few species [11, 64, 65, 69]. Whilst these studies are informative in understanding the mechanisms behind ecosystem response, care must be taken in making predictions based on simplistic assumptions, such as additive and linear relationships. Studying the next level of complexity is problematic, and while a mesocosm approach may help provide conceptual advances, we recognise the limitations of any artificial system in providing realistic interpretations of natural ecosystem response . However, we have shown that interactive effects can have a fundamental influence on MPB biomass and since the balance between autotrophic and heterotrophic status in transitional systems may be delicate , then there is a real possibility that climate change may force an overall change metabolism in coastal systems. Given that coastal systems will be at the forefront of climate change effects, they may undergo profound changes in the near future with associated implications for ecosystem services.