The diversity of ice algal communities on the Greenland Ice Sheet as revealed by oligotyping

Abstract

The Arctic is being disproportionally affected by climate change compared with other geographic locations, and is currently experiencing unprecedented melt rates. The Greenland Ice Sheet (GrIS) can be regarded as the largest supraglacial ecosystem on Earth, and ice algae are the dominant primary producers on bare ice surfaces throughout the course of a melt season. Ice-algal-derived pigments cause a darkening of the ice surface, which in turn decreases albedo and increases melt rates. The important role of ice algae in changing melt rates has only recently been recognized, and we currently know little about their community compositions and functions. Here, we present the first analysis of ice algal communities across a 100 km transect on the GrIS by high-throughput sequencing and subsequent oligotyping of the most abundant taxa. Our data reveal an extremely low algal diversity with Ancylonema nordenskiöldii and a Mesotaenium species being by far the dominant taxa at all sites. We employed an oligotyping approach and revealed a hidden diversity not detectable by conventional clustering of operational taxonomic units and taxonomic classification. Oligotypes of the dominant taxa exhibit a site-specific distribution, which may be linked to differences in temperatures and subsequently the extent of the melting. Our results help to better understand the distribution patterns of ice algal communities that play a crucial role in the GrIS ecosystem.

Keywords: Greenland Ice Sheet; Ice algae; Illumina; OTU; albedo; oligotyping.

Fig. 1.

Map showing the location of field sites in southwest Greenland (a) across a 100 km transect from the base camp to further inland (b). Sites 1a, 2 and 3 were sampled on 27 July and sites 1b, 2 and 3 were sampled on 5 August 2016. The samples around the base camp were collected over a 3-week period (c). Labels in bold type above the red dots represent site names and those below in parentheses indicate sample numbers. Full details can be found in Table 1. (Image source: Google Earth, 23.06.2017).

Fig. 2.

Algal community composition on the GrIS, with the eight most abundant taxa comprising >99 % of the total community at each sampling location. Full details can be found in Tables S3 and S4. The data presented here are based on the clustering of OTUs at 99 % similarity. The community composition based on the oligotyping approach can be found in Fig. 3. Locations are indicated above the histogram while sample types are shown below: CS, clean snow; DS, dirty snow; A, air; CI, clean ice; DC, dispersed cryoconite; all others represent dirty ice samples.

Fig. 3.

Graphs on the left show the most information-rich nucleotide position for the individual taxa, as revealed by the Shannon entropy analyses. Positions in the sequence alignment are on the x-axis, and the volume of Shannon entropy is on the y-axis. The entropy of each column in the nucleotide alignment was quantified [on a scale from 0 (none) to 1 (highest entropy)], and nucleotide positions with the highest entropy values were used for the oligotyping process. All the entropy peaks below ~0.2 can be regarded as sequencing noise and are likely not due to biological variation. The results of the Shannon entropy analyses were then used to derive distinct oligotypes, of which the distribution is displayed in the graphs on the right. The full details of the oligotype relative abundances can be found in Tables S5–S9. Oligotype names are derived from the nucleotide variations in the sequence alignment (Fig. S1).

Fig. 4.

Representative light micrographs of ice algae in samples GrIS17_18 (a) and GrIS17_10 (b) showing the more abundant Ancylonema (a) and Mesotaenium (b) cells, respectively. Bars, 10 µm.