Limnol. Oceanogr., 44(2), 1999, 431–435

Because of photoacclimation, frequency distributions of phytoplankton single-cell fluorescence reflect variation in light histories among individual cells. As a result, phytoplankton can be used as biological tracers for vertical mixing. A time series of Prochlorococcus single-cell fluorescence distributions in the Sargasso Sea revealed changes that were qualitatively consistent with changes in water-column physical dynamics. Applying a photoacclimation-diffusion model to Prochlorococcus fluorescence from the time series yields a vertical diffusivity near 30 cm2 s21 for a remnant mixed layer beneath a diurnal thermocline, which is consistent with estimates obtained by other means. This estimate results from an integrated view of how Prochlorococcus responds to changes in light intensity induced by vertical mixing and therefore, may be an appropriate measure of vertical diffusion rates for productivity models that include photoacclimation dynamics. This approach provides a novel way to examine vertical mixing on time scales relevant to phytoplankton physiological responses. Understanding vertical mixing processes in the surface oceans and consequent phytoplankton photoacclimation has represented a major challenge for oceanographers (Denman and Gargett 1983; Falkowski 1984; Lewis et al. 1984a). This understanding is necessary for interpreting and modeling profiles of primary productivity (e.g., Lewis et al. 1984b; Gallegos and Platt 1985; Mallin and Paerl 1992; Franks and Marra 1994), for predicting the effects of ultraviolet radiation on phytoplankton (Kullenberg 1982; Smith and Baker 1982), and for interpreting ocean color observations. Measurements of average or bulk values of light-sensitive properties of phytoplankton have been used to estimate vertical mixing rates (Falkowski 1983; Lewis et al. 1984a; Therriault et al. 1990); such estimates will be improved by considering properties of individual cells since the within-population distributions of light-sensitive properties contain more information indicative of mixing processes than mean values alone (Lande and Lewis 1989). We examined the potential to use flow cytometrically (Olson et al. 1993) determined optical properties of the ubiquitous picophytoplankter Prochlorococcus as an indicator of mixing processes. Prochlorococcus is a small (0.6-mm diameter) and thus effectively neutrally buoyant phytoplankton cell that occurs at densities of 104–105 cells ml21 over large regions of the world’s oceans (Chisholm et al. 1988; Olson et al. 1990). Like all phytoplankton, these cells photoacclimate to changing light regimes. Photoacclimation can be observed in situ as changes in single-cell chlorophyll fluorescence, which can be measured using flow cytometry (Olson et al. 1990). Photoacclimation-diffusion model—Notations used here are defined when first used and are summarized in Table 1. Prochlorococcus chlorophyll fluorescence (normalized to the 0.33 power of light scatter to remove diel periodicity due to cell growth [Dusenberry 1995]), G, and its distribution within a population from a specific depth is determined by both vertical mixing and photoacclimation and can be modeled with a one-dimensional diffusion-photoacclimation equation (Lewis et al. 1984a; Cullen and Lewis 1988) (Fig. 1A): ]G ] ]G G 2 G ` 5 K 1 gG · (1) v 1 2 1 2 ]t ]z ]z G` The first term on the right-hand side parameterizes vertical mixing in terms of a diffusivity, KV, and the second term represents photoacclimation, modeled using a logistic formulation (Cullen and Lewis 1988), where g is the photoacclimation rate constant. For chlorophyll fluorescence, the fully acclimated value of single-cell fluorescence, G`, is a decreasing function of light intensity; photoacclimation establishes a gradient of increasing fluorescence with depth. Using a numerical analog to the analytical single-cell model of Lande and Lewis (1989), we have extended the bulk property model of mixing and photoacclimation to a single-cell implementation of Eq. 1. From the individual cell’s point of view, Eq. 1 can be rewritten in terms of the cell’s normalized fluorescence, G, and position, z: dG G 2 G ` 5 gG (2) 1 2 dt G`