We show here that CO2 partial pressure (pCO2) and temperature significantly interact on coral physiology. The effects of increased pCO2 and temperature on photosynthesis, respiration and calcification rates were investigated in the scleractinian coral Stylophora pistillata. Cuttings were exposed to temperatures of 25°C or 28°C and to pCO2 values of ca. 460 or 760 μatm for 5 weeks. The contents of chlorophyll c2 and protein remained constant throughout the experiment, while the chlorophyll a content was significantly affected by temperature, and was higher under the ‘high‐temperature–high‐pCO2’ condition. The cell‐specific density was higher at ‘high pCO2’ than at ‘normal pCO2’ (1.7 vs. 1.4). The net photosynthesis normalized per unit protein was affected by both temperature and pCO2, whereas respiration was not affected by the treatments. Calcification decreased by 50% when temperature and pCO2 were both elevated. Calcification under normal temperature did not change in response to an increased pCO2. This is not in agreement with numerous published papers that describe a negative relationship between marine calcification and CO2. The confounding effect of temperature has the potential to explain a large portion of the variability of the relationship between calcification and pCO2 reported in the literature, and warrants a re‐evaluation of the projected decrease of marine calcification by the year 2100.

Summary Previous studies have demonstrated that coral and algal calcification is tightly regulated by the calcium carbonate saturation state of seawater. This parameter is likely to decrease in response to the increase of dissolved CO2 resulting from the global increase of the partial pressure of atmospheric CO2. We have investigated the response of a coral reef community dominated by scleractinian corals, but also including other calcifying organisms such as calcareous algae, crustaceans, gastropods and echinoderms, and kept in an open‐top mesocosm. Seawater pCO2 was modified by manipulating the pCO2 of air used to bubble the mesocosm. The aragonite saturation state (Ωarag) of the seawater in the mesocosm varied between 1.3 and 5.4. Community calcification decreased as a function of increasing pCO2 and decreasing Ωarag. This result is in agreement with previous data collected on scleractinian corals, coralline algae and in a reef mesocosm, even though some of these studies did not manipulate CO2 directly. Our data suggest that the rate of calcification during the last glacial maximum might have been 114% of the preindustrial rate. Moreover, using the average emission scenario (IS92a) of the Intergovernmental Panel on Climate Change, we predict that the calcification rate of scleractinian‐dominated communities may decrease by 21% between the pre‐industrial period (year 1880) and the time at which pCO2 will double (year 2065).

The effect of increased CO2 partial pressure (pCO2) on the community metabolism (primary production, respiration, and calcification) of a coral community was investigated over periods ranging from 9 to 30 d. The community was set up in an open‐top mesocosm within which pCO2 was manipulated (411, 647, and 918 µatm). The effect of increased pCO2 on the rate of calcification of the sand area of the mesocosm was also investigated. The net community primary production (NCP) did not change significantly with respect to pCO2 and was 5.1 ± 0.9 mmol O2 m−2 h−1. Dark respiration R increased slightly during the experiment at high pCO2, but this did not affect significantly the NCP:R ratio (1.0 ± 0.2). The rate of calcification exhibited the trend previously reported; it decreased as a function of increasing pCO2 and decreasing aragonite saturation state. This re‐emphasizes the predictions that reef calcification is likely to decrease during the next century. The dissolution process of calcareous sand does not seem to be affected by open seawater carbonate chemistry; rather, it seems to be controlled by the biogeochemistry of sediment pore water.

The enhancement of pico-and nanoplankton cell biomass by coral exudates was studied in the laboratory. Two types of mesocosms were used, the first one containing only a carbonate sand layer (control mesocosm) and the second one contaming a coral layer over the carbonate sand layer (coral mesocosm). During 10 h incubations, we followed the concentration of bacteria, cyanobacteria, and of auto-and heterotrophic flagellates, as well as the concentrations of inorganic (N and P) and organic (dissolved organic carbon, DOC) nutrients. There were no significant differences in inorganic nutrient concentrations between mesocosms. However, DOC concentrations in coral mesocosms exhibited peaks 5-to 13-fold hgher than control mesocosm levels; these peaks took lace between 13:OO and 17:OO h and lasted for ca 2 h. As a consequence, microbial growth was significantly enhanced in coral mesocosms. At the end of the incubations, bacterial biomass was 6-fold higher in coral relative to control mesocosms. Autotrophic biomass was 3 to 5 times higher in coral mesocosms. These results indicate that small amounts of coral exudates (0.5 to 10% of maximum DOC concentrations) are enough to greatly stimulate mcrobial growth.

Wild and managed bees are essential for global food security and the maintenance of biodiversity. At present, the conservation of wild bees is hampered by a huge shortfall in knowledge about the trends and status of individual species mainly due to their large diversity and variation in life histories. In contrast, the managed Western honey bee Apis mellifera is one of the best studied and monitored insects in existence. Since similar drivers may be relevant for the decline of wild bees and losses of managed honey bees, this raises the possibility that monitoring of honey bees may help to detect threatened regions for wild bees, thereby fostering urgently required conservation measures. However, this possible relationship has not yet been explicitly tested for. Moreover, research currently focused on honey bees as a model species may yield important insights into wild insect susceptibility to stressors and vice versa. Here we use the bees of Europe as a model to show that managed honey bees are not suitable surrogates for detecting declines in wild bees. A direct comparison of the response of wild bees and honey bees to the same threats (nutritional deficiencies, parasites and pathogens, pesticides, and a changing climate) shows that, whilst some of their responses may be similar at the individual level, when considered at the reproductive level (individuals versus colonies), many of their responses diverge. These results reinforce the need for basic research into wild bee biology, the need for national monitoring schemes for wild bee populations, and the call for conservation actions tailored to the individual ecologies of wild bee species.wild bees / indicator species / species specific / sociality / populations