Trace Elements Induce Predominance among Methanogenic Activity in Anaerobic Digestion

Abstract

Trace elements (TE) play an essential role in all organisms due to their functions in enzyme complexes. In anaerobic digesters, control, and supplementation of TEs lead to stable and more efficient methane production processes while TE deficits cause process imbalances. However, the underlying metabolic mechanisms and the adaptation of the affected microbial communities to such deficits are not yet fully understood. Here, we investigated the microbial community dynamics and resulting process changes induced by TE deprivation. Two identical lab-scale continuous stirred tank reactors fed with distiller's grains and supplemented with TEs (cobalt, molybdenum, nickel, tungsten) and a commercial iron additive were operated in parallel. After 72 weeks of identical operation, the feeding regime of one reactor was changed by omitting TE supplements and reducing the amount of iron additive. Both reactors were operated for further 21 weeks. Various process parameters (biogas production and composition, total solids and volatile solids, TE concentration, volatile fatty acids, total ammonium nitrogen, total organic acids/alkalinity ratio, and pH) and the composition and activity of the microbial communities were monitored over the total experimental time. While the methane yield remained stable, the concentrations of hydrogen sulfide, total ammonia nitrogen, and acetate increased in the TE-depleted reactor compared to the well-supplied control reactor. Methanosarcina and Methanoculleus dominated the methanogenic communities in both reactors. However, the activity ratio of these two genera was shown to depend on TE supplementation explainable by different TE requirements of their energy conservation systems. Methanosarcina dominated the well-supplied anaerobic digester, pointing to acetoclastic methanogenesis as the dominant methanogenic pathway. Under TE deprivation, Methanoculleus and thus hydrogenotrophic methanogenesis was favored although Methanosarcina was not overgrown by Methanoculleus. Multivariate statistics revealed that the decline of nickel, cobalt, molybdenum, tungsten, and manganese most strongly influenced the balance of mcrA transcripts from both genera. Hydrogenotrophic methanogens seem to be favored under nickel- and cobalt-deficient conditions as their metabolism requires less nickel-dependent enzymes and corrinoid cofactors than the acetoclastic and methylotrophic pathways. Thus, TE supply is critical to sustain the activity of the versatile high-performance methanogen Methanosarcina.

Keywords: Methanoculleus; Methanosarcina; T-RFLP; amplicon sequencing; biogas reactor; mcrA; methanogenesis.

FIGURE 1

Depletion of trace elements in reactor R2 after omitting the trace element (TE) supply and decreasing the amount of the iron additive in week 72. Concentrations of cobalt, molybdenum, nickel, and tungsten (A), manganese and zinc (B) are shown. Error bars indicate minimum and maximum values of duplicate measurements.

FIGURE 2

Time course of process parameters of reactor R1 (green) and reactor R2 (orange). Depletion of TEs started in time period I (week 72-78, marked in light gray). Significant changes of some process parameters were observed in time period II (after week 78, marked in dark gray). (A) Average methane yields, error bars indicate standard deviation (n = 5). (B) Concentration of TAN in the reactor content, error bars indicate minimum and maximum values when measured twice per week. (C) Average hydrogen sulfide concentrations in the biogas, error bars indicate standard deviation (n = 5). (D) Average acetic acid concentrations in the reactor content, error bars indicate standard deviation (n = 6).

FIGURE 3

Non-metric multidimensional scaling (nMDS) plots of T-RFLP profiles of mcrA transcripts (A) and mcrA genes (B). Data from reactor R1 (green dots: period I, week 72–78; blue dots: period II, after week 78) and reactor R2 (orange triangles: period I, week 72-78; red triangles: period II, after week 78) are shown. Data points are labeled by week of sampling. Plots are based on the Bray-Curtis dissimilarity index. Environmental factors and T-RF shaping the community profiles most are shown as vectors (blue: p < 0.01; black: p < 0.001).

FIGURE 4

Non-metric multidimensional scaling plot of T-RFLP profiles of bacterial 16S rRNA genes of reactor R1 (green dots: period I, week 72–78; blue dots: period II, after week 78) and reactor R2 (orange triangles: period I week 72-78; red triangles: period II after week 78). Data points are labeled by week of sampling. Plots are based on the Bray-Curtis dissimilarity index. Environmental factors and T-RF shaping the community composition most are shown as vectors (blue: p < 0.01; black: p < 0.001).

FIGURE 5

Phylogenetic composition based on 454 amplicon sequencing of bacterial 16S rRNA genes from samples of R1 and R2 in weeks 76 and 80. Bacterial community composition is shown at the family level sorted according to the class level. Families with maximum abundances below 1% are summarized as rare.