Inferring microbial interactions in thermophilic and mesophilic anaerobic digestion of hog waste

Abstract

Anaerobic digestion (AnD) is a microbiological process that converts organic waste materials into biogas. Because of its high methane content, biogas is a combustible energy source and serves as an important environmental technology commonly used in the management of animal waste generated on large animal farms. Much work has been done on hardware design and process engineering for the generation of biogas. However, little is known about the complexity of the microbiology in this process. In particular, how microbes interact in the digester and eventually breakdown and convert organic matter into biogas is still regarded as a "black box." We used 16S rRNA sequencing as a tool to study the microbial community in laboratory hog waste digesters under tightly controlled conditions, and systematically unraveled the distinct interaction networks of two microbial communities from mesophilic (MAnD) and thermophilic anaerobic digestion (TAnD). Under thermophilic conditions, the well-known association between hydrogen-producing bacteria, e.g., Ruminococcaceae and Prevotellaceae, and hydrotrophic methanogens, Methanomicrobiaceae, was reverse engineered by their interactive topological niches. The inferred interaction network provides a sketch enabling the determination of microbial interactive relationships that conventional strategy of finding differential taxa was hard to achieve. This research is still in its infancy, but it can help to depict the dynamics of microbial ecosystems and to lay the groundwork for understanding how microorganisms cohabit in the anaerobic digester.

Fig 1. The fluctuation of COD and TS over time.

Ten time points were selected according to the changing rate of COD and TS in the MAnD (A) or TAnD (B) digesters as labeled in the bottom of the diagram. The gray region indicates the stable duration of COD/TS removal rates (i.e., RCOD or RTS).

Fig 2. Biogas production in the anaerobic reactors.

MAnD (A) or TAnD (B) digesters. (A) The MAnD digester had lower efficiency of gas (-●-) or methane (-○-) production than the TAnD one (B). The percentage of methane (CH₄%) in biogas was similar in the two reactors.

Fig 3. Different microbial compositions at MAnD or TAnD digesters.

Microbial communities shown at the family level under mesophilic (A) or thermophilic (B) conditions. 36 BRP microbial families related to five biogas generation pathways were denoted as p1 for hydrolysis, p2 for acidogenesis, p3 for acetogenesis, p4 for methanogenesis, and deS for desulfurization.

Fig 4. Microbes with different topological niches under mesophilic or thermophilic conditions.

The interaction networks identified at mesophilic (A) or thermophilic (B) temperatures. Nodes represent microbial families. The size of the nodes corresponds to their betweenness centralities. A larger node indicates higher betweenness centrality which has a large influence on the transfer of information through the network. The color of the nodes conveys the number of in-degree interactions, describing how many OTUs influence this node. The node shown in dark red connects the maximum in-degree nodes, i.e., 10. The node closest to the center of the network has the highest level of eigenvector centrality, a measure of the influence of a node in a network. Three nodes linked by a red dotted circle indicate the top three most influential OTUs. The top six are denoted inside an orange dotted circle. The pink arrow indicates an activate relationship, and the blue arrow indicates a repressive event. Nodes marked with an asterisk (*) have a very low level of out-degree interactions.

Fig 5. The association between metabolic trophic relations and microbial interactions.

(A) Metabolic complementarity index had a borderline association with positive interactive strengths. (B) Metabolic competition index showed a statistically significantly association with negative interactive strengths.