Choanoflagellates alongside diverse uncultured predatory protists consume the abundant open-ocean cyanobacterium Prochlorococcus

Abstract

Prochlorococcus is a key member of open-ocean primary producer communities. Despite its importance, little is known about the predators that consume this cyanobacterium and make its biomass available to higher trophic levels. We identify potential predators along a gradient wherein Prochlorococcus abundance increased from near detection limits (coastal California) to >200,000 cells mL^-1 (subtropical North Pacific Gyre). A replicated RNA-Stable Isotope Probing experiment involving the in situ community, and labeled Prochlorococcus as prey, revealed choanoflagellates as the most active predators of Prochlorococcus, alongside a radiolarian, chrysophytes, dictyochophytes, and specific MAST lineages. These predators were not appropriately highlighted in multiyear conventional 18S rRNA gene amplicon surveys where dinoflagellates and other taxa had highest relative amplicon abundances across the gradient. In identifying direct consumers of Prochlorococcus, we reveal food-web linkages of individual protistan taxa and resolve routes of carbon transfer from the base of marine food webs.

Keywords: choanoflagellates; heterotrophic nanoflagellates; microbial food webs; picocyanobacteria; trophic transfer.

Fig. 1.

Phytoplankton prey communities from the coast to open ocean and predatory protists feeding on Prochlorococcus. (A) Stations sampled and overall Pacific region (Inset). (B) Chlorophyll a concentrations (2009 only for simplicity) show the development of a DCM offshore. (C and D) Mean cell abundances (±SD) of Prochlorococcus, Synechococcus, and eukaryotic picophytoplankton in the (C) Surface Mixed Layer (SML) and (D) DCM (if present), generally from 6-y annual sampling. Transect mid-region data are variable due to interannual shifts (13). (E) Relative contributions of protistan groups identified as feeding (maroon) and those not feeding on Prochlorococcus (gray) to total 18S-V9 rRNA amplicons in the RNA-SIP experiment at 67-80. Note: log scale. (F) Relative enrichment in amplicon abundance in heavy fractions of the density gradient due to isotope incorporation in ¹³C/¹⁵N-labeled treatments compared to the same fractions from the controls (¹²C/¹⁴N). Shown is the average enrichment across all five heavy density windows per ASV (see also Dataset S2).

Fig. 2.

Predatory protist distribution patterns. (A) Ordination of predatory heterotrophic protist communities by principal component analysis, overlain with vectors of environmental parameters. Coastal and open-ocean habitats differed in community composition (PERMANOVA: F₆ = 4.61, P = 0.001), as did the SML and DCM (PERMANOVA: F₁ = 4.86, P = 0.001). (B) Relative abundance contributions of all protists putatively functioning as heterotrophic predators collapsed to broad groups, based on 18S-V4 amplicon analyses. (C) Habitat distributions of RNA-SIP-identified Prochlorococcus predators from the open-ocean shoreward. Shown are 18S-V4 amplicon abundances averaged across years with Aitchison distance and Ward-linkage based hierarchical clustering of ASVs. Information following taxon names indicates membership in known phylogenetic clades followed by amplicon identifiers (Dataset S1). All dictyochophytes were Pedinellales.