The major fluxes of organic carbon associated with physical transport and biological metabolism were compiled, analyzed and compared for the mainstem portion of Chesapeake Bay (USA). In addition, 5 independent methods were used to calculate the annual mean net ecosystem metabolism (NEM = production -respiration) for the integrated Bay. These methods, which employed biogeochemical models, nutrient mass-balances and summat~on of individual organic carbon fluxes, yielded remarkably similar estimates, with a mean NEM of +50 g C m-2 yr.' (i SE = 7.51, which is approximately 8 % of the estimated annual average gross primary production. These calculat~ons suggest a strong cross-sectional pattern in NEM throughout the Bay, wherein net heterotrophic metabolism prevails in the pelagic zones of the rnaln channel, while net autotrophy occurs in the littoral zones which flank the deeper central area. For computational purposes, the estuary was separated ~n t o 3 regions along the land-sea gradient: (1) the oligohaline Upper Bay (1 1 "L of total area); (2) the mesohaline Wd Bay (36% of area); and (3) the polyhallne Lower Bay (53% of area). A distinct regional trend in NEM was observed along this salinity gradient, with net heterotrophy (NEM =-87 g C m-' yr-') in the Upper Bay, balanced metabolism in the Mid Bay and net autotrophy (NEM = +92 g C m-' y r ' ) in the Lower Bay. As a consequence of overall net autotrophy, the ratio of dissolved inorganic nitrogen (DIN) to total organic nitrogen (TON) changed from D1N:TON = 5.1 for riverine inputs to D1N:TON = 0.04 for water exported to the ocean. A striking feature of this organic C mass-balance was the relative dominance of biologically mediated metabolic fluxes compared to physical transport fluxes. The overall ratio of physical TOC inputs (I) to biotic primary production (P) was 0.08 for the whole estuary, but varied dramatically from 2.3 in the Upper Bay to 0.03 in the Mid and Lower Bay regions. Similarly. ecosystem respiration was some 6-fold higher than the sum of all physical carbon sinks. This general negative correspondence between 1:P ratio and NEM, which occurred among Bay regions, was also evident in data available for organic C fluxes in other coastal ecosystems. An inverse relationship between NEM and P, postulated in a previous study, did not apply to Chesapeake Bay, and closer examination of available data revealed the importance of the loading ratio of DIN:TOC as a key control on coastal NEM. It is proposed here that the general global trend of coastal eutrophicatlon will lead to increasing values of NEM in estuaries worldwide. The management implications of this trend are complex, involv~ng both increased potential fisheries harvest and decreased demersal habitat. K E Y WORDS: Net ecosystem metabolism Production . Respiration Organic carbon . Inorganic nutrients Estuaries. Chesapeake Bay INTRODUCTIONRates of organic production in estuaries and other coastal ecosystems are among the highest in the bio-

Although experimental ecosystems a r r bdsic and versatile tools w~d e l y used In coastal research periphyt~c glo\vth on container walls IS an ~n t n n s l c artifact that must be considered when lnterprctlng results To bettel understand holv this a1 tifact may confound extrapoldtion of results f~o m controlled exper~ments to conditions In natural estudr~ne ecosystems w e examined ho\v wall perlphyton varied wlth container size and shape in summer and dutunin expenments Kepllcate ( n = 3) cvlind11ca1 mesocosms ot 3 volumes (0 1 1 0 10 m') werr establ~shed In both constant-depth (depth -1 m) and conbtant-shape (radiuddepth = 0 56) serles \ l e~o c o s n~s were in~tlated with unfiltered estudrine water and homogenized s e d~m e n t s Pel~phyton b~ornass and gross primary production (GPP) per unlt of wall area ~n c l e a s e d wlth incredslng r a d~u s ( I ) or decreasing ratio of \vall area (A,,,) to water volume ( V ) for mesocos~ns In both s e r~e s (A,,/V = 2/r) As a consequence per~phyton biornass and metabol~sm expressed per u n~t of water volume Increased as a quadratic functlon of increasing A,, / V ratlo Results also suggest a secondary s c a l~n g effect whereby wall perlphyton qrowth may he directly reldted to mesocosm depth although mechanisms for th15 effect I enialn uncledr Slgnlflcant correlations between perlphyton biomass (per m2 wall area) and 3 rnv~ronmental f a c t o~s (11ght a t t r n u a t~o n c o e f f~c~e n t nutnent concentration and zooplankton dbundance) suqqest that these factors may have played Important roles In r e g u l a t~n g wdll grolvth Add~tlonally, effects of wall penphyton growth on plankton community dynamics were also indicated by the signlflcant n e g a t~v e relat~ons between penphyton biomass and measures of both phytoplankton and zooplankton abundance The overall effect of periphyton on the e\perlmental ecosystems was evldent In the fact that perlphyton accounted for over 5 0 " ) of total ecosystem GPP and b~o m a s s after 2 to 4 ~v k of these expenments For mesocosm experimc.nts d e s~g n e d to examlne dyndmics of planktonic-benth~c ecosystems, our ~e s u l t s lmply that growth of wall periphvton which 1s controlled b\l fact015 s c a l~n g to the ~a d i u s of expenmental ecosystems tends to domintlte major b~otic pools and ratcs withln weeks