Symbiotic colonies of the coral Acropora sp. were cultured in a factorial design of three temperatures (21, 25 and 28 degrees C) and two light intensities (200 and 400 mu mol photon m(-2) s(-1)), under constant conditions. A temperature of 25 degrees C and a light intensity of 200 mu mol photon m(-2) s(-1) was the starting culture condition. Metabolic (photosynthesis, respiration, calcification and surface expansion rate) and geochemical measurements (delta O-18, delta C-13, Sr/Ca and Mg/Ca) were conducted on 6 colonies for each experimental condition. Metabolic measurements confirmed that respiration, photosynthesis, calcification and surface expansion rate responded to the combined effect of temperature and light. Under each light intensity, mean calcification rate was linearly correlated with mean photosynthetic activity. Geochemical measurements were also influenced by temperature and, to a lesser degree, by light. All geochemical proxies measured on 6 nubbins showed a wide scattering of values, regardless of the environmental condition. Compared to the other proxies, delta O-18 exhibited a different behavior. It was the only proxy exhibiting temperature tracer behavior. However, while mean values of Sr/Ca, Mg/Ca and delta C-13 were well correlated, the correlation between the later and mean delta O-18 differed with light level. This suggests that both skeleton deposition and temperature oxygen fractionation differs according to light intensity. Overall, the effect of light on geochemical values seems to compromise the use of proxy calibrations solely based on temperature influence. Under high light conditions, the great amplitude shown by individual net photosynthesis is directly proportional to the highly variable zooxanthellae density. As light is affecting all of the proxies, we thus assume that the strong geochemical variability observed could be explained by various algae densities, each nubbin responding according to its zooxanthellae amount. Accordingly, we suggest that each symbiosome (the assemblage of few corallites with their symbionts) presents its own vital effect influence over time. Therefore, at a bulk sample scale, light could be considered as one of the major causes of what is commonly referred to as the 'vital effect'. The meaning of delta O-18 calibration versus temperature established from distinct colonies differs from calibration calculated from samples collected following the growth axis of a single coral head. Finally, in order to quantitatively reconstruct climatic condition, we suggest a new paradigm based on the statistical treatment of the combination of time-series information from several proxies, all measured on the same sample from a continuous symbiosome.