Net in situ growth rates of phytoplankton in an oligotrophic, tropical shelf ecosystem

In situ growth rates of phytoplankton in tropical shelf waters of the Great Barrier Reef were measured with diffusion chambers. A broad disparity was observed between the growth responses of diatom species or assemblages (p,,,. = 2-5 doublings d-l) and assemblages of microflagellatcs, nonmotile ultraplankton, or phototrophic picoplankton (p,,,, < 2 doublings d-l). Several diatom species or groups consistently grew faster than the community Chl doubling rate (H,,, = 1.6 doublings d-r). Most dominant species and groups, regardless of taxonomic affiliation, exhibited the potential for high relative growth rates (r/a,.,> 0.5) at ambient concentrations of dissolved inorganic N ~0.2 pm01 liter- ’ and the time-averaged diurnal irradiances in situ>50 pEinst m-2 SK’. If relative constancy in phytoplankton standing crop and composition is to bc maintained, grazing rates must vary widely at the species level, but be broadly similar to growth rates over time scales of days. Although the full 40-80-m water column in the central Great Barrier Reef lies the euphotic zone, phytoplankton growth rates near the bottom seem to bc light limited. Primary production by phytoplankton is translated into population growth through increases in cell numbers by binary fission. Historically, four general approaches have been taken to measure or, more usually, es-timate in situ growth rates of phytoplankton species or communities. First, recognizable morphological features or events in mitosis have been used to monitor frequencies of dividing cells in individual species and from that population growth rates (e.g. Swift and Durbin 1972; Weiler 1980). Second, growth rates also have been inferred from biochem-ical indices or reaction rates to yield esti-mates of biomass augmentation (Falkowski and Owens 1982; Redalje and Laws 198 1) or relative (JI*,,,) growth rate (Goldman 1980). Unfortunately, most biochemical as-says aggregate individual species growth rates into an average community rate. Third, growth rates have been modeled mathe-matically from experimental and field dat