Marine planktonic ciliates that prey on macroalgae and enslave their chloroplasts

We found two tide-pool ciliates, Strombidium oculatum and Strombidium stylifer, that ingest ulvaceous green macroalgae and retain their chloroplasts. Sequences of the form I ribulose bisphosphate carboxylase/oxygenase (rubisco) large subunit gene (rbcL1) from chloroplasts found in the ciliates cluster with the Ulvophyceae sequences on GenBank and with those of ulvaceous macroalgae from pools in which the ciliates were collected. In addition, we have cultivated S. stylifer in vitro on filtered seawater supplemented with pieces of Enteromorpha thalli that had been treated with light and temperature shock to maximize production of reproductive unicells (swarmers). An average growth rate of 1.08 6 0.07 SE [standard error] d21 was measured when S. stylifer was grown this way. Because both ciliates and the algal swarmers contain eyespots while vegetative cells in the Ulvophyceae do not, we speculate that these reproductive cells are the source of both the chloroplasts and the pigments used in the ciliates’ eyespots. This ciliate will not grow in the dark and is required to ingest fresh chloroplasts every few days, making it an obligate mixotroph. Ingestion of macroalgae by ciliates constitutes an upside-down trophic link, contrary to the usual pattern for planktonic food webs, in which production passes from very small organisms to successively larger ones. Our finding suggests that macroalgal production, heretofore believed to contribute predominantly to detrital or macroherbivore food chains, may be an important food source in the nearshore plankton and that ciliates may play an important role in this trophic pathway. Planktonic ciliates are often the dominant herbivores in coastal waters and members of the genus Strombidium usually comprise the most abundant taxon within this group (Heinbokel and Beers 1979; Capriulo and Carpenter 1983; Montagnes et al. 1988; Pierce and Turner 1992). Some of the Strombidiidae have a mixotrophic nutritional mode, retaining the chloroplasts of ingested algae and using them to supplement their nutrition with photosynthesis (McManus and Fuhrman 1986; Stoecker et al. 1987; Stoecker 1991; Bernard and Rassoulzadegan 1994). These ciliates are thus both producers and consumers, and models of marine food webs suggest that mixotrophy results in increased trophic efficiency in the sea (Stoecker 1998; Stickney et al. 2000). Until now, ciliate mixotrophy has only been found to result from ingestion of unicellular microalgae. Grazing of ciliates on macroalgae has not been reported. We studied the ingested chloroplasts of two ciliate species, Strombidium oculatum and Strombidium stylifer, isolated from tide pools in Europe and North America. Both are small (ø50–60 mm in length) and planktonic. S. oculatum has the interesting habit of encysting and excysting on an endogenous circatidal rhythm such that it is active and swimming in tide pools during low tide and encysted during high tide (Faure-Fremiet 1948; Jonsson 1994; Montagnes et al. 2002b). Both ciliates contain chloroplasts that are grassgreen in color, and partially digested green unicells can be seen within food vacuoles (Fig. 1), suggesting a chlorophyte origin. Both ciliates also have a prominent eyespot at the oral end and show strong phototactic behavior. Methods—Cloning and phylogenetic analysis: We collected S. oculatum from tide pools on the Irish Sea (Isle of Man) and Dublin Bay and Galway Bay (Ireland) and S. stylifer from pools on Long Island Sound (USA). Samples of green macroalgae (Enteromorpha-like members of the Ulvales) were collected from Dublin Bay and Galway Bay pools. Samples for ciliate DNA analysis were collected from pools isolated from the open sea at low tide. Several 10s of ciliates (typically 40–60) were picked with a drawn capillary pipette into microcentrifuge tubes containing DNA extraction buffer (100 mmol L21 NaCl, 50 mmol L21 TRIS-HCl [pH 8.0], 25 mmol L21 EDTA, 0.5% sodium dodecyl sulfate [SDS]). After treatment with proteinase K, DNA was extracted and the form I ribulose bisphosphate carboxylase/ oxygenase (rubisco) large subunit gene (rbcL1), which in all eukaryotic algae is encoded in the chloroplast genome (Tabita 1999), was amplified by PCR. Primer sequences were designed based on a conserved region of form I rubisco large subunits from various organisms: RubI-forward: GCTGCATTCCGTWTBACWCCWCAACCAGG; RubI-reverse: GTGRATACCRCCWGAAGCWACWGG. Polymerase chain reaction (PCR) was performed as follows: 35 cycles of 25 s of denaturation at 948C, 30 s of annealing at 588C, and 40 s of elongation at 728C. PCR products of the expected size (1,053 base pairs) were cloned into the TA vector and transformed into competent cells (Invitrogen). For each PCR product we obtained from the ciliate samples, 17–26 bacterial clones were randomly picked up and plasmid DNA was isolated. Macroalgal samples from Dublin Bay and Galway Bay were treated in a similar manner. All ciliate and green macroalgal clones were subjected to restriction fragment length polymorphism (RFLP) analysis using the restriction enzymes EcoRI and HindIII. We then selected 27 clones for sequencing based on differences observed in RFLP patterns (17 identical and 4 different clones for ciliates, 6 identical clones for green macroalgae) in order to maximize coverage of different clones. Sequencing was carried out using the BigDye Terminator Cycle Sequencing Kit (PE Applied Biosystems) and an ABI Prism 377 automated sequencer. Sequences were analyzed using the Basic Local Alignment Search Tool (BLAST) program against gene sequences in GenBank for similarity to known gene sequences. Alignment of sequences and subsequent phylogenetic analyses were carried out using CLUSTAL X1.8 (Thompson et al. 1994). The deduced amino acid (aa) sequences of rbcL1 from the ciliate and algal samples were analyzed with those published in GenBank. A phylogenetic tree was constructed using the neighbor-joining method (Saitou and Nei 1987) and was corrected using the Kimura method (Kimura 1980). The reli-