Phytoplankton growth and diel variations in beam attenuation through individual cell analysis

Abstract

A number of investigators have observed diel variations in the bulk water inherent optical property beam attenuation, with a minimum near dawn and a maximum near dusk, and have assumed them to be caused by the phytoplankton. In an attempt to understand these patterns, mean forward light scatter of different populations of phytoplankton, from flow cytometric analysis of individual cells, was determined and found to show similar diel patterns for populations of pico- and nanophytoplankton measured in the equatorial Pacific and the Sargasso Sea. The cell concentration patterns do not generally correspond to those of beam attenuation. Laboratory experiments, combined with theoretical calculations, were undertaken to attempt to account for the diel variations in beam attenuation observed at sea and to estimate in situ phytoplankton group-specific growth and loss rates. To investigate how cell growth and division affect the optical properties of phytoplankton, cultures of the chlorophyte Nannochloris sp. were sampled over a diel cycle to measure cell size and concentration, light attenuation and absorption, flow cytometric light scattering (in forward and side directions), and carbon content. In addition, the refractive index was calculated using the anomalous diffraction approximation of Mie theory. At six different light levels ranging from 60 - 1500 μmol photons m-2 s-1 (specific growth rates from ~0.2 to ~0.7 d-1), cell division was tightly phased to the light:dark cycle, occurring soon after dark. There were pronounced diel patterns, with a minimum near dawn and a maximum near dusk, in cell size and in cell-specific beam attenuation, absorption, flow cytometric light scatter and carbon. The diel variations in the attenuation cross section were primarily influenced by the changes in cell size due to growth and division, while changes in refractive index had only a small effect. Because eukaryotic cells in the size range of Nannochloris are major constituents of many phytoplankton communities, these results have important consequences for the interpretation of diel variations in optical properties observed in the ocean. To interpret field data of diel variations in red beam attenuation, the relationships between cell optical properties determined in the laboratory were applied to three diel sampling experiments in the equatorial Pacific and four in the Sargasso Sea. In the equatorial Pacific in April and October 1992, the phytoplankton biomass was dominated by picophytoplankton (Prochlorococcus and Synechococcus), and by mixed populations of ultraphytoplankton (1-2 μm diameter) and nanophytoplankton (2-20 μm, mostly 2-3 μm). Flow cytometric measurements of mean forward light scatter of each of these populations showed typical diel patterns which were similar to those of bulk beam attenuation due to particles, whereas cell concentration changes were not. Using a combination of empirical calibrations relating beam attenuation to flow cytometric measurements of pure cultures of phytoplankton in the laboratory, and Mie theory, the contributions of different groups of phytoplankton to the diel variations in beam attenuation were estimated. The results indicate that the phytoplankton assemblage measured by flow cytometry can account for essentially all of the diel variation in the beam attenuation signal. In mo.st instances the nanophytoplankton were the largest contributor to total beam attenuation due to phytoplankton, but the ultraphytoplankton usually were more important in determining the diel variations in this property. Prochlorococcus were a smaller but appreciable contributor to beam attenuation changes, and Synechococcus were much less important. A similar analysis was performed for diel sampling experiments in the Sargasso Sea in January 1992, July 1993, and May 1994. The bulk beam attenuation due to particles was strongly correlated with calculated beam attenuation due to phytoplankton. During the July 1993 diel sampling, when pico- and nanophytoplankton populations were analyzed, in most instances the nanophytoplankton were the largest contributor to total beam attenuation due to phytoplankton, but Prochlorococcus were equally important at 70 m for some time points over the diel cycle. In the upper 40 m, Prochlorococcus were a smaller contributor to beam attenuation changes than the nanophytoplankton, and Synechococcus were even less important. These findings emphasize the need to characterize the composition of the phytoplankton community in order to use beam attenuation to monitor productivity. Flow cytometric measurements of phytoplankton light scattering and cell concentration over the diel light cycle were used to estimate in situ phytoplankton group-specific growth and loss rates in the equatorial Pacific and the Sargasso Sea. Measurements of forward light scatter were converted to cell volume and carbon using laboratory and theoretically derived calibration factors for specific groups of phytoplankton, including Prochlorococcus, Synechococcus, ultraphytoplankton, nan~phytoplankton, and coccolithophores. Assuming that division was phased, specific growth rates were estimated based on volume and carbon changes between minimum and maximum values over the day. Phytoplankton group-specific loss rates, and also separate day and night loss rates, were estimated from the calculated growth rates and measured cell concentrations over time. The method used to estimate growth rates works well for Prochlorococcus and appears to work for small eukaryotic phytoplankton, but leads to underestimates for Synechococcus and larger eukaryotes. Estimated growth rates for Prochlorococcus reached one division per day in the upper waters of the equatorial Pacific, but were about half that in the Sargasso Sea. For the eukaryotic phytoplankton, growth rates in the equatorial Pacific were near one division per day in the upper waters; in the Sargasso Sea, the growth rates approached that, but were more often lower. In general, phytoplankton growth rates were closely matched by in situ loss rates over the course of a day.