Diverse microbial communities hosted by the model carnivorous pitcher plant Sarracenia purpurea: analysis of both bacterial and eukaryotic composition across distinct host plant populations

Abstract

Background: The pitcher plant Sarracenia purpurea supplements nutrient acquisition through carnivory, capturing insect prey which are digested by a food web community of eukaryotes and bacteria. While the food web invertebrates are well studied, and some recent studies have characterized bacteria, detailed genetic analysis of eukaryotic diversity is lacking. This study aimed to compare eukaryotic and bacterial composition and diversity of pitcher communities within and between populations of host plants in nearby but distinct wetland habitats, and to characterize microbial functions across populations and in comparison with another freshwater community.

Methods: Pitcher fluid was sampled from the two wetlands, Cedarburg and Sapa Bogs, community DNA was extracted, and 16S and 18S rRNA amplicons were sequenced and data processed for community-level comparisons.

Results and conclusions: Bacterial diversity in the small pitcher volume rivaled that of larger aquatic communities. Between pitcher plant populations, several bacterial families (Kiloniellaceae, Acetobacteraceae, Xanthobacteraceae, Sanguibacteraceae, Oligoflexaceae, Nitrosomonadaceae, Chromatiaceae, Saprospiraceae) were significantly higher in one population. However, although predicted pitcher bacterial functions were distinct from other freshwater communities, especially for some amino acid metabolism, functions were similar across all the pitchers in the two populations. This suggests some functional redundancy among bacterial taxa, and that functions converge to achieve similar food web processes. The sequencing identified a previously under-appreciated high diversity of ciliates, Acari mites, fungi and flagellates in pitcher communities; the most abundant sequences from eukaryotic taxa were Oligohymenophorea ciliates, millipedes and Ichthyosporea flagellates. Two thirds of taxa were identified as food web inhabitants and less than one third as prey organisms. Although eukaryotic composition was not significantly different between populations, there were different species of core taxonomic groups present in different pitchers-these differences may be driven by wetland habitats providing different populations to colonize new pitchers. Eukaryotic composition was more variable than bacterial composition, and there was a poor relationship between bacterial and eukaryotic composition within individual pitchers, suggesting that colonization by eukaryotes may be more stochastic than for bacteria, and bacterial recruitment to pitchers may involve factors other than prey capture and colonization by eukaryotic food web inhabitants.

Keywords: Acari; Bacteria; Carnivorous plant; Ciliate; Eukaryotes; Fungi; Microbial diversity; Microbiome; Nutrient transformation; Wetland plant.

Figure 1. Bacterial and eukaryotic pitcher plant composition of samples in two populations.

Genetic analysis of pitcher plant communities from two wetland populations, Cedarburg and Sapa. Family phylotypes are based on 16S rRNA (bacterial) (A) and 18S rRNA (Eukaryotic) (B) taxa identified from sequencing of total community DNA in pitcher fluid samples. Samples were collected from Cedarburg (CB) and Sapa (Sp) wetlands from single plants (e.g., CB1-5) from individual pitchers (e.g., p1-4) or combined pitchers (e.g., p234). Taxa with only one representative sequence were removed, and any taxa representing < 0.01% of total in each sample were pooled as ‘Other’. AT, ambiguous taxa, IS, incertae sedis, unc., uncultured.

Figure 2. Heatmap of bacterial and eukaryotic families across the samples in two populations.

Heatmap of family-level taxa based on 16S rRNA (A) and 18S rRNA (B) sequences isolated from pitcher fluid samples (not-median adjusted). Color scale corresponds to the logarithmic transformation of the number of times a taxon was observed in each sample (green is lowest abundance, red is highest abundance). For bacteria in five samples from each population, taxa which were significantly higher in one population are shown with a p-value. Abbreviations are as for Fig. 1.

Figure 3. Pie charts of major eukaryotic groups and roles in microecosystem.

Composition of eukaryotic groups based on 18S rRNA sequences, showing most common Arthropod groups, ciliates, (ospithokont) flagellates, fungi, and other groups for each wetland population (A, B) (full details of taxa in each sample provided in Table S2). (C) Using data pooled for all 10 samples, taxa representing >1% of total sequences were scored as Food Web for aquatic taxa or those known to be pitcher inhabitants, Prey for non-aquatic Arthropods as likely captured prey, or Other for ambiguous taxa which may be present as incidentals (fungal spores blown into pitchers, taxa with unknown habitat or ecological role).

Figure 4. PCoA of pitcher plant bacterial composition and functions in two populations.

(A) Pitcher bacterial communities visualized with PCoA for pitcher bacterial composition, and (B) Non-metric multidimensional scaling plot of PICRUSt predicted bacterial metabolic functions, in five pitcher samples from each wetland, compared with five wastewater reference samples. Metabolic functions plot shows 15 selected function categories as vectors separating the communities. Convex hulls overlay the sample points for each group, showing overlap between the two wetland populations, which are distinct from the wastewater communities. Vector name abbreviations relate to metabolic processes associated with: b_Ala–b-Alanine, Lys–Lysine; Val_Leu-Ile–Valine-Leucine-Isoleucine; Trp–Tryptophan; Gly_Ser_Thr–Glycine-Serine-Threonine; TCA_cyle; S_met–Sulphur metabolism; Arg_Pro–Arginine-Proline; Glyc_lipid–glycerolipid; Sta_Suc–starch and sucrose; PS–photosynthesis; PS_Pig–photosynthetic pigments; Glycan_biosyn–glycan biosynthesis; Cyst_Met–cysteine-methionine; Phe_Tyr_Trp–Phenylalanine-Tyrosine-Tryptophan.

Figure 5. PCoA of pitcher plant eukaryotic taxa in two populations.

(A) PCoA score plot based on composition of eukaryotic taxa (18S rRNA sequences) identified in five pitcher samples from each wetland, compared with a curated freshwater Eukbase database from SILVA NR108. Convex hulls overlay the sample points for each group, showing overlap between the two wetland populations which are distinct from the Eukbase outgroup. (B) NMDS plot of community composition, using the most common 16 eukaryotic families as vectors to separate Cedarburg and Sapa samples, with correlation as the similarity measure.

Figure 6. Relatedness trees for samples in two populations based on 16S and 18S rRNA sequences.

Genetic diversity trees of samples from Cedarburg Bog and Sapa Bog based on 16S bacterial (A) and 18S eukaryotic (B) composition of taxa identified in five pitcher fluid samples from each population. Bootstrap values for jackknife trees generated in QIIME were based on 100 iterations with a minimum of 75% of the smallest sample sequence number. Cedarburg (CB) sample branches are black while Sapa (Sp) sample branches are grey. Bacterial composition of samples is compared with pooled wastewater outgroup (WW), and a curated freshwater EUKBASE database from SILVA.