A new optical instrument, designed specifically to measure the natural or solar-induced fluorescence of chlorophyll a, was deployed in the tropical and temperate waters of the western South Pacific gyre. During the 12 days of transit from Pajpeete, Tahiti, to Auckland, New Zealand, measurements were made of the vertical distribution of temperature, salinity, upwelling radiance and downwelling irradiance at selected wavelengths, the beam attenuation coefficient at 683 nm, the scalar irradiance of photosynthetically available radiation (PAR), and the nadir radiance at 683 nm, which is the wavelength of peak emission by Chl a. In addition, water samples were collected at regular depth intervals and analyzed for the concentration of Chl a and for the spectral absorption coefficient of cells and associated detrital particles. To examine the relationship between natural fluorescence and the size and photosynthetic rate of the phytoplankton crop, we derived several simple equations and applied them to the analysis of the data. One relates the nadir radiance at 683 nm to the total fluorescence emitted by the phytoplankton crop within the field ofview, another relates natural fluorescence to the concentration of Chl a, and a third relates natural fluorescence to the gross rate of photosynthesis of the crop.We found that even in the extremely oligotrophic waters of the central South Pacific gyre, natural fluorescence was easily measured throughout the euphotic zone at depths >6 m. As found in previous studies, the value of natural fluorescence varied spatially and temporally with ambient scalar irradiance of PAR and the concentration of Chl .a. The quantum yield of fluorescence, which varied fivefold, averaged about 0.035 photons emitted to photons absorbed and generally decreased with increasing levels of exciting irradiance. Most importantly, we found natural fluorescence covaried closely with calculated rates of photosynthesis. A linear regression of calculated photosynthetic rate on fluorescence yielded a correlation coefficient of 0.84 and a slope of 2.0 atoms of carbon fixed per photon emitted as fluorescence.

No abstract

Values for the spectral absorption coefficient of particles sampled in the Sargasso Sea (34"N, 7O"W) in 1987 were subjected to a multiple regression analysis to assess contributions to light absorption by phytoplankton and detritus. Specifically, values for the spectral absorption coefficient at a given wavelength were treated as a function of the values for absorption at 675 and 570 nm. The spectrum for slopes for the 675-nm component has a slope like that of phytoplankton, and the slopes for the 570-nm component resemble that of detritus. Individual spectra were then partitioned into phytoplankton and detrital spectral absorption coefficients with an algorithm that minimized the difference between the original spectrum and sum of the two components. The twocomponent model accounted for >90% of the observed spectral absorption coefficients for all samples collected within the upper 300 m of the water column. Profiles of the two components indicated that there was little variation in mean absorption coefficient for detritus with depth. In contrast, the greatest variability was accounted for by changes in the phytoplankton component, both with depth and seasonally during 1987. This seasonal variation is interpreted in terms of photoadaptation by cells to changes in the light and vertical mixing fields.Measurements of the spectral absorption coefficient of natural waters are not easily obtained. Although particle backscattering