Limnol. Oceanogr., 44(2), 1999, 440–446

Cellular nutrient ratios are often applied as indicators of nutrient limitation in phytoplankton studies, especially the so-called Redfield ratio. For periphyton, similar data are scarce. We investigated the changes in cellular C : N : P stoichiometry of benthic microalgae in response to different levels and types of nutrient limitation and a variety of abiotic conditions in laboratory experiments with natural inocula. C : N ratios increased with decreasing growth rate, irrespective of the limiting nutrient. At the highest growth rates, the C : N ratio ranged uniformly around 7.5. N : P ratios ,13 indicated N limitation, while N : P ratios . 22 indicated P limitation. Under P limitation, the C : P ratios increased at low growth rate and varied around 130 at highest growth rates. For a medium with balanced supply of N and P, an optimal stoichiometric ratio of C : N : P 5 119 : 17 : 1 could be deduced for benthic microalgae, which is slightly higher than the Redfield ratio (106 : 16 : 1) considered typical for optimally growing phytoplankton. The optimal ratio was stable against changes in abiotic conditions. In conclusion, cellular nutrient ratios are proposed as an indicator for nutrient status in periphyton. The chemical composition of oceanic seston is known to be relatively constant at a C : N : P ratio of 106 : 16 : 1 (Redfield 1958; cf. Copin-Montegut and Copin-Montegut 1983). This biogeochemical ratio became widely known as the ‘‘Redfield ratio’’ and was subsequently physiologically interpreted for aquatic organisms. Droop (1974, 1975) investigated the nutrient content of phytoplankton within different limitation scenarios and developed the cell quota growth model. Internal nutrient ratios are equivalent to carbon-based cell quotas. Nitrogen and phosphorus supply supporting maximum growth rate was shown to lead to phytoplankton stoichiometry resembling the Redfield ratio (Goldman et al. 1979; Elrifi and Turpin 1985), and the internal nutrient ratios were proposed as an indicator of algal nutrient status (Healey and Hendzel 1980; Flynn 1990). Despite some criticism (Ryther and Dunstan 1971; Tett et al. 1985), biomass stoichiometry has been widely applied to assess nutrient supply to phytoplankton in marine (Paasche and Erga 1988; Burkhardt and Riebesell 1997) and freshwater studies (Sommer 1991a; Hecky et al. 1993). It should be noted that C : N : P ratios close to the Redfield ratio do not indicate the absence of light limitation (Tett et al. 1985). They just indicate that neither N nor P are limiting factors of growth (Goldman 1986). The use of nutrient ratios as an index of nitrogen or phosphorus limitation was also recommended for benthic microalgae (Borchardt 1996), because of the close phylogenetic relationship between benthic and pelagic microalgae. However, in benthic studies, it has seldom been applied to date (Engle and Melack 1993; Rosemond 1993; Rosemond et al. 1993; Hillebrand and Sommer 1997). The infrequent use may be because the relationship between cellular stoichiometry and growth rate for marine microphytobenthos has not yet been tested experimentally. Kahlert (1998) recently reviewed literature data from freshwater periphyton and found that C : N : P ratios are a reliable tool for the assessment of the nutrient status of benthic algae, proposing an optimum ratio of 158 : 18 : 1. In the present study, we wanted to answer the following questions: (1) Is there a consistent relationship between benthic microalgal growth rates and cellular stoichiometry, and (2) Is this relation independent of abiotic conditions? To investigate the response of internal nutrient ratios to changes in nutrient regimes, we used a semicontinuous dilution of culture media combined with sampling in intervals. We used natural inocula of algae to simplify comparison to natural assemblages. The algae for the inocula were scraped from an artificial substrate in the Kiel Fjord, Western Baltic Sea, 10 d before the experiments were started. The algae were cultured in unenriched filtered seawater under the same abiotic conditions as the experimental treatments (Table 1). At the beginning of the experiments, 1 ml of the inoculum was added to each treatment. The experiments were conducted in flat-bottom, transparent, polystyrene culture flasks with 30 ml total medium content. The algae grew as a biofilm on the bottom of the flasks, which were shaken once daily. This biofilm was a dense monolayer of cells lacking an overstory. It can therefore be assumed that CO2 was available in excess. The media used in the experiments consisted of organismfree–filtered seawater (0.2-mm cellulose-acetate filters) from the same location, enriched with nutrients and trace metals. The balanced medium (designated V) contained 80 mmol liter21 N (as NaNO3), 80 mmol liter21 Si (as Na2O3Si 3 5H2O), and 5 mmol liter21 P (as Na2HPO4 3 2H2O), resulting in a medium N : P ratio of 16. For the N-limited medium (designated Nlim), the nitrogen concentration was reduced to 10 mmol liter21. For the Si-limited (Silim) media, Si was reduced to 10 mmol liter21; for the P-limited medium (Plim), no phosphate was added, resulting in 1.0 mmol liter21 P. The experiments were conducted in autumn 1997 and spring 1998. In the autumn experiment, four different media were applied, and the treatments consisted of an alteration of dilution rate (Table 1). In spring 1998, three different media were used, and the temperature was altered (Table 1). The two experiments differed furthermore in the taxonomic composition of the inoculum and in the light intensity, which was measured with a LiCor LI 189 (Table 1). Each treatment was conducted in triplicate, resulting in 24 cultures in autumn (four media 3 two dilution rates 3 three replicates) and 27 cultures in spring (three media 3 three temperatures 3 three replicates). Thrice a week, the algae were counted alive employing an inverted microscope (Leitz DMIRB) at 3630 magnification. Up to 400 cells were counted per sample. To compare