While in culture, specific growth rate (μ) can be easily determined through cell counts (N0 and Nt). This is not the case for incubations with natural assemblages, where grazers and other processes lead to cell loss and consequently an underestimation of Nt. Incubations with [2-(4-pyridyl)-5{[4-dimethylaminoethyl-aminocarbamoyl)-methoxy]phenyl}oxazole] (PDMPO) overcomes this issue by allowing the determination of N0 (Nt=N0×e^(μt) (Eq. 1)) based on the relative proportion of PDMPO stained cells (newly divided) and unstained cell at the end of an incubation as follows: Given Nt, the total cell number of a species from a subsample at a time t after PDMPO addition. Nt=nt+nt', with nt the total number of non-(PDMPO) stained cells of the species in the same subsample (cells that did not divide yet), and nt' the total number of cells with one PDMPO stained valve (cells issued from the first division after PDMPO addition to the culture media). All things being equal, the population (N0) at the time when PDMPO was added that gave rise to Nt should be N0=nt+((nt')/2)=Nt-((nt')/2) (Eq. 2). Growth rates from cells counts were estimated using (Eq. 1), while growth rates using PDMPO were calculated from the number of non-stained (nt), half-stained (nt') and fully stained (nt'') cells using μ[d-1]=ln((nt+nt')/(nt+(nt')/2))×(1/t) (Eq. 3) or μ[d-1]=ln((nt'+nt'')/((nt')/2))×(1/t) (Eq. 4). Cells attached to each other (potentially in the final phase of division) were also considered as non-stained, half-stained and fully-stained individuals, respectively, based on the presence/absence of a PDMPO signal on the valves.