Growth and grazing on Prochlorococcus and Synechococcus by two marine ciliates

The two most abundant marine autotrophic prokaryotes, Prochlorococcus and Synechococcus, often have different distributions in the ocean. For example, Synechococcus is restricted to the first 100 m, whereas Prochlorococcus extends much deeper in oligotrophic waters. This is in part explained by differences in adaptation to nutrient and light regimes. However, they could also be subjected to different predation rates. To explore this hypothesis, we compared the consumption of these two picoplankters by an algivorous ciliate, Strombidium sulcatum, and a bactivorous ciliate, Uronema sp. For both ciliate species, removal rates were higher, by a factor of 3 to 10, for Synechococcus compared to Prochlorococcus when prey items were presented alone or together. The growth of the two ciliates fed Synechococcus and/or Prochlorococcus also differed. S. sulcatum grew well on both prey items, whether alone or together, whereas Uronema sp. grew slowly when fed Synechococcus and very poorly when fed Prochlorococcus either alone or with Synechococcus. Our results suggest that Prochlorococcus may be less subject to ciliate predation than Synechococcus. Prokaryotic picoplankton often dominate phytoplankton assemblages in marine systems (Platt et al. 1983; Olson et al. 1985; Blanchot and Rodier 1996). For many open oceans, the contribution of one picophytoplankton group, Synechococcus, in terms of abundance and contribution to primary productivity has been recognized for nearly 20 yr (Johnson and Sieburth 1979; Waterbury et al. 1979; Morris and Glover 1981). The existence of Prochlorococcus was established relatively recently using flow cytometry, and it appears to have a significance, in terms of carbon fixation, comparable to that of Synechococcus (Chisholm et al. 1988). The relative importance of Prochlorococcus differs among oceanic regions and often seems to vary inversely with that of Synechococcus (Campbell and Vaulot 1993; Li 1995; Landry et al. 1996; Partensky et al. 1996). In oligotrophic open waters, Prochlorococcus populations are more abundant and extend deeper in the water column than Synechococcus throughout most of the year (Olson et al. 1985; Chisholm et al. 1988; Campbell and Vaulot 1993; Campbell et al. 1994). The distinct distributions of Synechococcus and Prochlorococcus are generally thought to reflect adaptations to different nutrient and light regimes. For example, maximal Prochlorococcus concentrations have been reported to occur in nitrate-depleted layers (Lindell and Post 1995; Blanchot and Rodier 1996), whereas Synechococcus can be abundant in transition areas where nitrate is present (Chisholm et al. 1988; Glover et al. 1988a,b; Campbell and Vaulot 1993; Campbell et al. 1994). Prochlorococcus appears better 1 Present address: National Centre for Marine Research, 16604 Aghios Kosmas, Hellinikon, Greece. 2 Station Biologique, CNRS, INSU et Université Pierre et Marie Curie, Place Georges Teissier, 29680 Roscoff, France Acknowledgments Financial support was provided by the Commission of the European Communities through grants ‘‘MEDEA’’ MAS3 CT95-0016 and ‘‘MATER’’ MAS3-CT96-0051. S.J. was supported by a doctoral fellowship form the French Ministry of Education and Research. We appreciate the efforts of two anonymous reviewers and D. Kirchman in helping us improve the manuscript. adapted for growth at low light intensities relative to Synechococcus (Moore et al. 1995). However, it is worthwhile to point out that the observed distributions, usually attributed to different growth capacities, are the sum of both growth and mortality. Chroococcoid cyanobacteria have long been observed in the food vacuoles of nanoplanktonic protists (Johnson et al. 1982), but their contribution to protist nutrition is uncertain. In culture studies, Synechococcus has been described as a poor food item for protists (Verity and Villareal 1986; Caron et al. 1991), while field populations of Synechococcus can apparently support rapid growth of some ciliates (Simek et al. 1995; Pérez et al. 1996; Simek et al. 1996). Data on the growth rate of Prochlorococcus are relatively abundant (e.g., Goericke and Welschmeyer 1993; Moore et al. 1995; Vaulot et al. 1995) compared to the little existing information on grazing losses (Liu et al. 1995; Reckermann and Veldhuis 1997). To our knowledge, there are no data on the food value of Prochlorococcus. The question arises then as to whether or not Synechococcus and Prochlorococcus are exploited similarly by protist grazers. There are reasons to suspect that, although Prochlorococcus and Synechococcus are roughly similar in size, the two may be removed at different rates. Selective ingestion of picoplankton-sized particles by flagellates (Epstein and Shiaris 1992; Sherr et al. 1992; Jürgens and DeMott 1995) and ciliates (Turley et al. 1986; Sanders 1988; Simek et al. 1994; Christaki et al. 1998) has been reported. The ingestion of picoplankton can be affected by quality and motility of prey as well as small differences in prey size and the physiological state of the grazer (Sanders 1988; Snyder 1991; Christaki et al. 1998). Furthermore, even if a prey type is removed efficiently by grazers, it may not experience high grazing pressure over extended periods of time if it is an inadequate food source for the grazer. Given these considerations, we thought it of interest to compare Synechococcus and Prochlorococcus as prey items for planktonic ciliates. We compared consumption of Prochlorococcus and Synechococcus by an algivorous ciliate, S. sulcatum, and a bactivorous ciliate, Uronema sp. Short-term 53 Grazing by ciliates Fig. 1. Strombidium sulcatum ingestion: changes in cell concentrations of Prochlorococcus, Synechococcus in experiments with S. sulcatum. S. sulcatum culture with addition of (A) Prochlorococcus SS120, (B) Synechococcus WH8103, and (C) mixed SS120 and WH8103. Open symbols show prey concentrations in control solutions. Error bars show the range of duplicate cultures. Where error bars are not shown, the range is smaller than the symbol. Fig. 2. Uronema sp. ingestion: changes in cell concentrations of Prochlorococcus, Synechococcus in ingestion experiments with Uronema sp. Uronema culture with addition of (A) Prochlorococcus SS120, (B) Synechococcus WH8103, and (C) mixed SS120 and WH8103. Open symbols show prey concentrations in control cultures. Error bars show the range of duplicate cultures. Where error bars are not shown the range is smaller than the symbol. experiments were used to estimate ingestion rates and possible differential removal of Prochlorococcus and Synechococcus. Long-term experiments compared Prochlorococcus and Synechococcus as food sources for the two ciliates. Materials and methods Culture conditions—Prochlorococcus SS120 (Chisholm et al. 1992), approximately 0.65 mm in diameter, and Synechococcus WH8103 (Waterbury et al. 1986), originally isolated from Sargasso Sea and approximately 1.0 mm in length, were grown in 500-ml sterile flasks in n K/10—Cu medium in aged seawater as described in Scanlan et al. (1996). The two well-characterized strains (Moore et al. 1995) are typical of oligotrophic provinces of the open ocean (Campbell and Iturriaga 1988; Goericke and Welschmeyer 1993). Cultures of both populations were acclimated for 3 weeks to experimental conditions. Cultures were grown at 20 6 0.58C in a temperature-regulated room under continuous light (15 mE m22 s21), provided by a pair of cool-white fluorescent bulbs wrapped in blue filter (Lee filter, band-pass at 475 nm). Neither Prochlorococcus nor Synechococcus cultures were axenic. The cultures used for the experiments were in exponential growth phase, with background heterotrophic bacterial densities of approximately 1 3 106 bacteria ml21 compared to 1 3 107 autotrophs ml21. S. sulcatum and Uronema sp., originally isolated from the bay of Villefranche-sur-Mer (Mediterranean Sea), were maintained in stock cultures on a bacterized wheat-grain medium (Rivier et al. 1985). To obtain exponentially growing cultures, protozoa inocula from stock cultures were transferred into bacterized yeast extract media (0.015–0.030 g liter21, see Christaki et al. 1998 for details). Ingestion experiments—In short-term experiments, we estimated ingestion rates of S. sulcatum and Uronema sp. cultures feeding on (1) Prochlorococcus SS120, (2) Synechococcus WH 8103, or (3) mixed Prochlorococcus and Synechococcus. Ciliates were removed from late exponential growing cultures when the concentration was 0.25 and 1.0 3 103 ml21 for S. sulcatum and Uronema sp., respectively. Fifty-milliliter aliquots of ciliate cultures were spiked with exponentially growing Prochlorococcus and/or Synechococcus cultures, yielding a final total concentration of prokaryotic autotrophs of approximately 5 3 105 ml21. The concentration of a particular picoautotroph was 5 3 105 ml21 when offered alone and 1.5–2.5 3 105 ml21 when offered with the other picoautotroph. In the ingestion experiment, heterotrophic bacteria from the ciliate and picoautotroph cultures were present in concentrations of about 7.5 3 106 ml21. Control solutions of picoautotrophs were prepared by adding the same concentration of autotrophs to 50 ml of 0.2-mm–fil-