Iron and sulfate reduction structure microbial communities in (sub-)Antarctic sediments

Abstract

Permanently cold marine sediments are heavily influenced by increased input of iron as a result of accelerated glacial melt, weathering, and erosion. The impact of such environmental changes on microbial communities in coastal sediments is poorly understood. We investigated geochemical parameters that shape microbial community compositions in anoxic surface sediments of four geochemically differing sites (Annenkov Trough, Church Trough, Cumberland Bay, Drygalski Trough) around South Georgia, Southern Ocean. Sulfate reduction prevails in Church Trough and iron reduction at the other sites, correlating with differing local microbial communities. Within the order Desulfuromonadales, the family Sva1033, not previously recognized for being capable of dissimilatory iron reduction, was detected at rather high relative abundances (up to 5%) while other members of Desulfuromonadales were less abundant (<0.6%). We propose that Sva1033 is capable of performing dissimilatory iron reduction in sediment incubations based on RNA stable isotope probing. Sulfate reducers, who maintain a high relative abundance of up to 30% of bacterial 16S rRNA genes at the iron reduction sites, were also active during iron reduction in the incubations. Thus, concurrent sulfate reduction is possibly masked by cryptic sulfur cycling, i.e., reoxidation or precipitation of produced sulfide at a small or undetectable pool size. Our results show the importance of iron and sulfate reduction, indicated by ferrous iron and sulfide, as processes that shape microbial communities and provide evidence for one of Sva1033's metabolic capabilities in permanently cold marine sediments.

Fig. 1. Sampling locations around South Georgia.

Core identifications are displayed, the red marked core was used for SIP incubations.

Fig. 2. Pore water concentrations of iron(II) (Fe²⁺), sulfate (SO₄^2‒), sulfide (H₂S), phosphate (PO₄^3‒), ammonium (NH₄⁺), dissolved inorganic carbon (DIC), and silicate (SiO₂) in surface sediments, South Georgia.

All missing values in Fe²⁺ and H₂S profiles were for data points below detection limit.

Fig. 3. Bacterial community composition and gene copy numbers in South Georgia surface sediments.

A Relative abundance of bacterial 16S rRNA genes in 10 depths of Annenkov Trough, Church Trough, Cumberland Bay and Drygalski Trough. B Bacterial 16S rRNA gene copies per gram wet sediment of samples displayed in (A) with error bars displaying SD of technical qPCR replicates.

Fig. 4. Distance-based redundancy analysis (dbRDA) ordination plot of bacterial communities in surface sediments of South Georgia.

Sample points are distinguished by site and core depth by shape and color, respectively. dbRDA1 (variation 47%) and dbRDA2 (variation 22%) axes are displayed, which constrain the Bray-Curtis distance matrix with geochemical parameters Fe²⁺, PO₄³⁻, NH₄⁺, SiO₂, and H₂S. The total model (F = 4.99, p < 0.01, Df: 5, 34) and each individual parameter (p < 0.05) was significant.

Fig. 5. Depth profile of contribution of sulfate and iron reducing microorganisms in Deltaproteobacteria to bacterial community and quantification of sulfate reducers (dsrA gene) in South Georgia surface sediments.

Relative abundance of 16S rRNA gene of taxa known for iron and/or sulfate reducing capabilities within Deltaproteobacteria (details in the text) was summed up per sediment depth. Fe²⁺ profile from Fig. 2 and dsrA gene copies per gram wet sediment  are displayed. Note the different scale for gene copies/g sediment for Annenkov Trough. Sequences of taxa known for iron reducing capabilities consisted of >78% Sva1033 in all depths of Annenkov Trough, Cumberland Bay, and Drygalski Trough.

Fig. 6. Results of SIP incubation with Cumberland Bay sediment.

A Visualization of RNA density separation with normalized RNA amount (ng RNA in fraction per total recovered ng RNA of sample). B Density separated 16S rRNA community composition. C dsrA transcript copies per ng cDNA from recovered RNA per fraction with SD of technical qPCR replicates as error bars. In the molybdate-amended treatment, the transcript copies were below detection limit (<100 copies) for the ultra-heavy fraction and highest in the light fraction with 3000 (¹³C) to 6700 (¹²C) copies per ng cDNA from recovered RNA. Legend of A corresponds to C.

Fig. 7. Abundance and activity of Sva1033.

Relative abundance of Sva1033 in South Georgia surface sediments (A) and in SIP incubations with Cumberland Bay (B) and Potter Cove sediment (C). A Separation between sites by color and shape. B Averaged relative abundance of Sva1033 across all SIP treatments displayed by bars while relative abundance of each treatment is individually displayed by data points distinguishable by shape and color.