Research ArticleECOLOGY

The global-scale distributions of soil protists and their contributions to belowground systems

See allHide authors and affiliations

Science Advances  24 Jan 2020:
Vol. 6, no. 4, eaax8787
DOI: 10.1126/sciadv.aax8787
  • Fig. 1 Protistan communities, dominated by Cercozoa and Ciliophora, are composed of mostly consumers across the 180 global soils studied.

    (A) Boxplots of the percent relative abundance (percentage of all protistan reads) of major soil protistan groups and (B) the percent relative abundance of lineages within the Filosea-Sarcomonadea (percentage of all Filosea-Sarcomonadea reads), and (C) the relative abundance of each general protist life history strategy (consumer, parasite, and phototroph), with samples averaged within each broad biome category.

  • Fig. 2 Protistan parasite relative abundance is controlled by mean annual precipitation and temperature.

    (A to C) Top: For the three parasite groups—Apicomplexa, Oomycota, and Ichthyosporea—the percent increase in mean squared error (MSE) of predictions when a given predictor is permuted randomly (e.g., a measure of variable importance) for all environmental predictors that were significant (P < 0.01) in the random forest models for explaining variation in parasite abundance across the 180 soils studied. For all parasite groups, mean annual temperature (MAT) and mean annual precipitation (MAP) explained the greatest percent of variation. Bottom: All three groups show a significant positive relationship with mean annual temperature and mean annual precipitation. Points represent the 180 soils, and for each group, the relative abundance of all amplicon sequence variants (ASVs) detected was summarized per site.

  • Fig. 3 The phylogenetic distribution of protistan clades colored by their identified ecological preferences.

    (A) Phylogenetic tree of protist clades (from 924 protist ASVs that were detected in >10% of the 180 soils). The inner ring shading indicates protist clade membership clustered at 96% patristic distance for clades with predictable ecological distributions, and the outer ring colors identify the environmental cluster memberships for those clades (see table S2 for details). PSEA, precipitation and seasonality. We note that this phylogeny is based on short read data and not intended to accurately represent broad supergroups. (B) Associations (Spearman’s ρ) between the standardized relative abundance of all clades within a particular environmental cluster (e.g., mean z scores) and the environmental predictor.

  • Fig. 4 Many protistan and bacterial taxa are highly correlated.

    Pairwise correlation heatmap of co-occurrences between protistan and prokaryotic ASVs (those found at >10% of sites). The bacterial ASVs (x axis) are collapsed by order-level taxonomy for readability—the heatmap color indicates the number of bacterial ASVs that were correlated with a particular protistan ASV (y axis).

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/6/4/eaax8787/DC1

    Fig. S1. Site locations for the 180 unique soil samples for which we obtained 18S rRNA gene sequence data (green circles; 137 samples for which we also obtained 16S rRNA gene amplicon data) and 46 samples for which we obtained SSU rRNA gene data from shotgun metagenomic sequencing (blue circles).

    Fig. S2. Pairwise correlation heatmap for the climatic and edaphic characteristics associated with the 180 soils analyzed.

    Fig. S3. Detection of the top 50 dominant protistan lineages by 18S rRNA amplicon sequencing versus shotgun metagenomic sequencing.

    Fig. S4. Factors shaping protistan community (namely, climatic variables) are distinct from those factors shaping prokaryotic communities (namely, pH).

    Fig. S5. The top predictors of ubiquitous protist lineages (protistan ASVs were phylogenetically clustered into clades at 96% similarity).

    Fig. S6. A heatmap of the correlation coefficients based on semipartial correlations between the relative abundance of protistan clades (one clade per row) that had significant semipartial correlations to environmental predictors (n = 43 of 50 protist clades that were predictable in distribution).

    Table S1. Location and site information for the 180 soils for which we obtained 18S rRNA gene amplicon data.

    Table S2. Information on the 116 abundant protistan lineages (phylogenetic clades) including taxonomic affiliation, the percent variance explained (var.expl), the top environmental predictor from random forest models (predictor1), the number of ASVs in the lineage (freq.), and the ecological cluster affiliation.

    Table S3. Location and site information for the 46 soils for which we obtained shotgun metagenomic data.

  • Supplementary Materials

    This PDF file includes:

    • Fig. S1. Site locations for the 180 unique soil samples for which we obtained 18S rRNA gene sequence data (green circles; 137 samples for which we also obtained 16S rRNA gene amplicon data) and 46 samples for which we obtained SSU rRNA gene data from shotgun metagenomic sequencing (blue circles).
    • Fig. S2. Pairwise correlation heatmap for the climatic and edaphic characteristics associated with the 180 soils analyzed.
    • Fig. S3. Detection of the top 50 dominant protistan lineages by 18S rRNA amplicon sequencing versus shotgun metagenomic sequencing.
    • Fig. S4. Factors shaping protistan community (namely, climatic variables) are distinct from those factors shaping prokaryotic communities (namely, pH).
    • Fig. S5. The top predictors of ubiquitous protist lineages (protistan ASVs were phylogenetically clustered into clades at 96% similarity).
    • Fig. S6. A heatmap of the correlation coefficients based on semipartial correlations between the relative abundance of protistan clades (one clade per row) that had significant semipartial correlations to environmental predictors (n = 43 of 50 protist clades that were predictable in distribution).
    • Table S1. Location and site information for the 180 soils for which we obtained 18S rRNA gene amplicon data.
    • Table S2. Information on the 116 abundant protistan lineages (phylogenetic clades) including taxonomic affiliation, the percent variance explained (var.expl), the top environmental predictor from random forest models (predictor1), the number of ASVs in the lineage (freq.), and the ecological cluster affiliation.
    • Table S3. Location and site information for the 46 soils for which we obtained shotgun metagenomic data.

    Download PDF

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article