ABSTRACT-Community metabolism was investigated uslng a Lagrangian flow resplrometry technique on 2 reef flats at Sloorea (French l'olynesia) during austral winter and Yonge Reef (Great Barrier Reef) during austral summer. The data were used to estimate related air-sea CO, disequilibrium. A slne funct~on did not sdtisfactorily model the d~e l light curves and ov~rrstsmated the metabolic parameters.The ranges of community gross primary production and respirat~on (P,, and R, 9 to 15 g C m-? d-') were within the range previously reported for reef flats, and community net calcification (G; 19 to 25 g CaCO, m-2 d.') was higher than the 'standard' range. The molar ratio of organlc to inorganic carbon uptake was 6:l for both sites. The reef flat at hdoorea displayed a higher rate of organic production and a lower rate of calcification compared to previous measurements carried out during austral summer. The approximate uncertainty of the daily metabolic parameters was estimated using a procedure based on a Monte Carlo simulation. The standard errors of P,, R and R,IR expressed as a percentage of the mean are lower than 3 % but are colnparat~vely larger for E, the excess production (6 to 78'1;, INTRODUCTIONThe biological processes controlling the CO2 partial pressure in seawater ( p C 0 2 ) and, consequently, the air-sea CO2 fluxes are net organic production and calcification, which have opposite effects on pC02. Coccolithophorid blooms and coral reefs are the major marine photosynthetic and calcifying systems and are, therefore, well suited to studies of the interactive 'E-mail: gattuso@naxos.unice.fr "Present address: Un~versiti. de Liege, Mecanique des Flu~des Geophysiques, Unite d'oceanographie Chimique, Institut d e Physique (B5). 8-4000 Sart Tilman, Belglum effects of photosynthesis-respiration and precipitation-dissolution of calcium carbonate on the seawater CO2 system and air-sea CO2 fluxes. Coccolithophorid blooms are transient (Holligan et al. 1993), annually cover on average 14 X 105 km2 of the world's oceans (Brown & Yoder 1994) and have been shown to behave as sources for atmospheric CO, (Robertson et al. 1994). Coral reefs are permanent, long-lived ecosystems covering 6 X 105 km2 (Smith 1978a). There is a rather large set of data on reef community lnetabolism (Kinsey 1985, Smith 1995 but very few estimates of air-sea CO2 fluxes in coral reefs . In most reefs investigated, gross primary production (Pg) and respiration (R) are nearly balanced (P,/R = 1) and netResale of full article not perrnltted

Community metabolism (primary production, respiration and calcification) and air-sea CO2 fluxes of the 'Tiahura barrier reef' (Moorea, French Polynesia) were investigated in November and December 1991 Gross production and respiration were respectively 640.2 to 753 and 590.4 to 641.5 mm01 (02 or CO2) m-' d-l (7.7 to 9.0 and 7.1 to 7.7 g C m-2 d ') and the reef displayed a slightly negative excess (= net) production. The contribution of planktonic primary production to reef metabolism was negligible (0.15 ?4, of total gross production). Net calcification was positive both during the day and at night; its daily value was 243 mm01 CaCO, m-2 d -' (24.3 g CaC03 m m 2 d-l). Reef metabolism decreased seawater total CO2 by 433.3 mm01 m--d-' The air-sea CO, fluxes were close to zero in the ocean but displayed a strong daily pattern at the reef front and the back reef. Fluxes were positive (COz evasion) at night, decreased as irradiance increased and were negative during the day (COz invasion) Integration of the fluxes measured during a 24 h experiment at the back reef showed that the reef was a source of CO2 to the atmosphere (1 5 minol m-2 d-l).

We investigated the effects of changes in nutrient concentrations on phytoplankton biomass and community composition during 8 field trips performed during different seasons in the southwestern coral lagoon of New Caledonia. The lagoon is characterized by spatial variation in macronutrient concentrations, with locally elevated values in the bays bordering the city of Nouméa. Low DIN:DIP (dissolved inorganic nitrogen:dissolved inorganic phosphorus) and elevated Si:DIN ratios suggest that nitrogen is the macronutrient that likely drives phytoplankton community composition. Most of the microphytoplankton groups discriminated by inverted microscopy and the picoplankton groups distinguished by flow cytometry present significant and distinct relationships with inorganic nitrogen concentrations. Picophytoplankton-dominated assemblages are replaced by microphytoplankton-dominated assemblages with increasing DIN concentrations. Within the picophytoplankton, Prochlorococcus abundance dominates in the adjacent oceanic and southern lagoon shelf sites, and assemblages shift to Synechococcus-dominated populations in the bays, with an increasing proportion of picoeukaryotic phytoplankton. Within the microplankton, 142 species of microphytoplankton were identified, mainly represented by diatoms, dinoflagellates, and coccolithophorids. Nutrient enrichment in the bays favors large diatoms at the expense of coccolithophorids and dinoflagellates, which dominate in adjacent oceanic and southern shelf waters. Therefore, although moderate, the elevated nitrogen concentrations in the bays result in increased phytoplankton biomass, accompanied by important shifts in the phytoplankton community structure.

Community metabolism and air-sea carbon dioxide (COJ jluxes were investigated in July 1992 on a finging reef at Momea (Frach Polynesia). The benthic community was dominated by macroalgae (85 % substratum cover) and comprised of Phaeophyceae Padina tenuis (Bmy), Turbinaria ornata (Turner) J. Agardh, and Hydroclathrus clathratus Bmy (Howe); Chlorophyta Halimeda incrassata $ ovata J. Agardh (Howe); and Ventricaria ventricosa J. Agardh (Olsen et West), as well as several Rhodophyta (Actinotrichia fragilis Forskal (Bargesen) and several species of encrusting coralline algae). Algal biomass was 171 g dry weight-m-,. Community gross production pK), respiration (R), and net calcijication (G) were measured in an opentop enclosure. P and R were respectively 248 and 240 mmol (20,-m-,. &I, and there was a slight net dissolution of CaCO, (0.8 mmoEm-2.d-1). This site was a sink for atmospheric CO, (10 ? 4 mmol CO,.m-z*d-l), and the analysis of data j o m the literature suggests that this is a general feature of algaldominated reefs. Measurement of air-sea CO, Juxes in open water close to the enclosure demonstrated that changes in small-scale hydrodynamics can lead to misleading conclusions. Net CO, evasion to the atmosphere was measured on the fingzng reef due to changes I in the cuwent pattern that drove waterjom the barrier reef (a CO, source) to the study site. IGJ index words: cmal re& algal-dominated; calcijication; community metabolism; air-sea CO, jlux; pn'mary productionCoral reef ecosystems are dominated by photosynthetic and calcifylng organisms such as scleractinian corals and macroalgae. They are therefore well suited to investigating the effect of the metabolism of organic (photosynthesis and respiration) and inorganic (precipitation and dissolution of CaCO,) carbon of benthic communities on the seawater inorganic carbon system. These biological processes can induce large changes in the partial pressure of carbon dioxide (pCO,), which controls air-sea CO, fluxes.The chemistry of carbon dioxide (CO,) in natural water is quite complex because dissolved C 0 2 combines with water to generate bicarbonate (HCO,) and carbonate (COi-) ions. The concentrations of those three species of inorganic carbon are controlled by equilibrium constants that depend on temperature, salinity, and pressure. Production and degradation of 1 mole of organic matter by photosynthesis and respiration respectively decreases and increases dissolved inorganic carbon (DIC) (DIC = [CO,] + [HCO;] + [COq-1) by approximately 1 mole: 729