Large-scale patterns in biodiversity of microbial eukaryotes from the abyssal sea floor

Abstract

Eukaryotic microbial life at abyssal depths remains "uncharted territory" in eukaryotic microbiology. No phylogenetic surveys have focused on the largest benthic environment on this planet, the abyssal plains. Moreover, knowledge of the spatial patterns of deep-sea community structure is scanty, and what little is known originates primarily from morphology-based studies of foraminiferans. Here we report on the great phylogenetic diversity of microbial eukaryotic communities of all 3 abyssal plains of the southeastern Atlantic Ocean--the Angola, Cape, and Guinea Abyssal Plains--from depths of 5,000 m. A high percentage of retrieved clones had no close representatives in genetic databases. Many clones were affiliated with parasitic species. Furthermore, differences between the communities of the Cape Abyssal Plain and the other 2 abyssal plains point to environmental gradients apparently shaping community structure at the landscape level. On a regional scale, local species diversity showed much less variation. Our study provides insight into the community composition of microbial eukaryotes on larger scales from the wide abyssal sea floor realm and marks a direction for more detailed future studies aimed at improving our understanding of deep-sea microbes at the community and ecosystem levels, as well as the ecological principles at play.

Fig. 1.

Map of the southeastern Atlantic with sampling locations: western Guinea Abyssal Plain (1), eastern Guinea Abyssal Plain (2), Angola Abyssal Plain (3), and Cape Abyssal Plain (4).

Fig. 2.

Midpoint rooted maximum likelihood tree of all clones (general eukaryotic and group-specific primers). Because of their very long branches, the euglenozoans were removed and are displayed separately (Fig. 3). Their branching position is marked by an arrow. Bootstrap support values ≥ 70 are highlighted by black circles. Branches in red are clones; branches in black are named sequences retrieved from GenBANK. For clarity, clones were grouped into OTUs (5.21% threshold; Fig. S1), and only 1 representative clone from each OTU was included in the phylogenetic tree. These clones are identified by the last 4 numbers of their GenBANK accession numbers (GU218701-GU219463) followed by the lowest p-distance to the first BLAST hit and, after the slash, the lowest p-distance to the first named BLAST hit within each OTU. When the first BLAST hit was already a named sequence, the first value was omitted. Bar charts represent the number of clones per abyssal plain obtained for each OTU; pie charts represent the relative number of clones obtained from each abyssal plain for the major phylogenetic groups. Sequences in bold represent the nearest named neighbor for each OTU. Phylogenetic branches are extended toward their labels by gray dotted lines and by red dotted lines if the lowest p-distance to the first named BLAST hit within the respective OTU was > 5.21%. These branches are also highlighted by long dashes if the lowest p-distance to the first BLAST hit was already > 5.21%. Species names and GenBANK accession numbers of the reference sequences are listed in Table S6.

Fig. 3.

Maximum likelihood tree of Euglenozoa. For more details, see Fig. 2.

Fig. 4.

Dendrogram of pairwise F_ST values. Clone libraries of microbial eukaryotes from depths ≥ 750 m constructed using general eukaryotic primers were compared. Clone libraries from this study are highlighted in bold. For comparision, two clone libraries from non--deep-sea environments were included in the analysis (denoted by §). Significant P values for F_ST, UniFrac, and P-tests are indicated by an asterisk (P < .01). Nonsignificant P values are indicated by bold lines, and the respective tests are denoted by F (F_ST test), U (UniFrac test), and P (P-test). When only some pairwise tests within a cluster were insignificant, the respective tests are given in parentheses.