Seasonal Dynamics of Haptophytes and dsDNA Algal Viruses Suggest Complex Virus-Host Relationship

Abstract

Viruses influence the ecology and diversity of phytoplankton in the ocean. Most studies of phytoplankton host-virus interactions have focused on bloom-forming species like Emiliania huxleyi or Phaeocystis spp. The role of viruses infecting phytoplankton that do not form conspicuous blooms have received less attention. Here we explore the dynamics of phytoplankton and algal viruses over several sequential seasons, with a focus on the ubiquitous and diverse phytoplankton division Haptophyta, and their double-stranded DNA viruses, potentially with the capacity to infect the haptophytes. Viral and phytoplankton abundance and diversity showed recurrent seasonal changes, mainly explained by hydrographic conditions. By 454 tag-sequencing we revealed 93 unique haptophyte operational taxonomic units (OTUs), with seasonal changes in abundance. Sixty-one unique viral OTUs, representing Megaviridae and Phycodnaviridae, showed only distant relationship with currently isolated algal viruses. Haptophyte and virus community composition and diversity varied substantially throughout the year, but in an uncoordinated manner. A minority of the viral OTUs were highly abundant at specific time-points, indicating a boom-bust relationship with their host. Most of the viral OTUs were very persistent, which may represent viruses that coexist with their hosts, or able to exploit several host species.

Keywords: Haptophyta; Megaviridae; Phycodnaviridae; marine viral ecology; metagenomics; viral–host interactions.

Figure 1

Isopleth diagrams showing seawater density (σt) and chlorophyll a (Chl a) fluorescence (RFU = relative fluorescence units) at the sampling station in Raunefjorden, respectively.

Figure 2

Abundance of microbial plankton at the sampling station in Raunefjorden measured by flow cytometry. (A) Phototrophic picoeukaryotes (filled circles) and nanoeukaryotes (open circles); (B) Synechococcus (open circles) and total bacteria (filled circles); and (C) V1 (low fluorescence viruses), V2 (intermediate fluorescence viruses) and V3 (high fluorescence viruses).

Figure 3

(A) Relative abundance of operational taxonomic units (OTUs) in the different haptophyte orders or clades. The OTUs represented seven accepted haptophyte orders: Pavlovales, Phaeocystales, Coccolithales, Isochrysidales, Syracosphaerales, Zygodiscales and Prymnesiales. The latter is represented by the families Chysochromulinaceae and Prymnesiaceae. The reads that did not belong to any of these orders were assigned to class Prymnesiophyceae (unclassified) or Haptophyta (unclassified); (B) Number of unique OTUs (richness) in each sample after rarifying to lowest read abundance (i.e., subsampling to obtain equal sample size).

Figure 4

Cluster dendrogram illustrating Bray-Curtis dissimilarity in the haptophyte OTU compositions between the five samples from Raunefjorden. OTUs were defined as reads with ≥98% nucleotide similarity. Sequences were normalized to 100 in each sample and log-transformed prior to similarity calculations. Samples connected by red lines were not significantly differentiated (SIMPER permutation test). Black lines indicate significant differentiation (p < 0.05, SIMPER permutation test).

Figure 5

Cluster dendrogram illustrating Bray-Curtis dissimilarity between the virus OTU compositions in five samples from Raunefjorden. OTUs were defined as sequences with >95% amino acid similarity. The sequence data were normalized to 100 in each sample and log-transformed prior to similarity calculations. Samples connected by red lines could not be significantly differentiated (SIMPER permutation test). Black lines indicate significant differentiation (p < 0.05, SIMPER permutation test).

Figure 6

Relative abundances of 21 different OTUs containing more than 50 reads in the different samples.

Figure 7

Midpoint rooted phylogenetic tree of the most abundant OTUs (>50 reads, marked in red) with similarities to the Megaviridae and Phycodnaviridae families, respectively. The tree was calculated based on the DNA-sequences encoding partial MCP-genes (FastTree v2.1.8 with default parameters). Bootstrap values form 100 replicates and aLRT- likelihood-values >0.5 are shown on nodes. Abbreviations: CroV; Cafeteria roenbergensis virus, Moumou; Moumouvirus goulette, Mimi; Mimivirus, Mega; Megavirus chiliensis, AaV; Aureococcus anophagefferens virus, PoV; Pyramimonas orientalis virus, PkV; Prymnesium kappa virus, HeV; Haptolina ericina virus, HhV; Haptolina hirta virus, CeV; Chrysochromulina ericina virus, PgV; Phaeocystis globosa virus, PpV; Phaeocystis pouchetii virus, PBCV; Paramecium bursaria Chlorella virus, MpV; Micromonas pusilla virus, OsV; Ostreococcus sp. virus, OlV; Ostreococcus lucimarinus virus. OTU/P0605, OTU/P0606, OTU/P0604, OTU/P0607, OTU/P0601, OTU/M0501 are all sequences from an earlier study in Raunefjorden [32]. Reference strains marked in green are haptophyte-infecting viruses maintained in culture.