There is a growing awareness that the bacterial interactions should follow a highly nonlinear pattern in reality. However, it is challenging to tract the varying bacterial interactions using the pair-wise correlation analysis, which fails to explore their potential effects on the behavior of microbes. Here, we utilize the regularized S-map to capture the varying interspecific interactions from the time-series data of bacterial community under the exposure to nitrite. Our results show that the bacterial interactions are highly variable and asymmetric interactions dominate the interaction pattern in community. Furthermore, we propose a Jacobian coefficient-based statistical method to predict the harmony level of a bacterial community at each successive ecosystem state. The result shows that bacterial community exhibits higher harmony level in nitrite-treated samples than the control group. We show that the community harmony level is positively associated with the specific endogenous respiration rates and biofilm formation of the culture. In addition, the community tends to process lower diversity and structural stability under zero and high nitrite stresses. We demonstrate that harmony level, other than structural stability, is a useful index to unveil the underlying mechanism of bacterial performance. Overall, the regularized S-map can help us to understand the bacterial interactions in eco-systems more accurately than previous approaches. Importance It has long been acknowledged that bacterial interactions play important roles in the community structure and function. Revealing the interaction variability can allow an understanding of how bacteria response to perturbation and why bacterial community performance changes. Such information should improve our skills to engineer the bacterial community (e.g., wastewater treatment plant) and achieve better removal performance and lower energy consumption.

In biofilm-based engineered ecosystems, the reactor performance was closely linked to interspecies interactions within a biofilm ecosystem, whereas the ecological processes underpinning such linkage were still unenlightened. Herein, the principles of community succession and assembly were integrated to capture the ecological laws of biofilm development by molecular ecological networks and assembly model analysis based on the 16S rRNA sequencing analysis and metagenomics in a well-controlled moving bed biofilm reactor. At the initial colonization phase (days 0−2, driven by initial colonizers), interspecific cooperation (74.18%) facilitated initial biofilm formation, whereas some pioneers, and keystone species disappeared at later phases. At the accumulation phase (days 3−30, rapid biofilm development), interspecific cooperation (81.41 ± 5.07%) contributed to rapid biofilm development and keystone species were mainly involved in quorum sensing or positively correlated with extracellular polymeric substance production. At the maturation phase (days 31−106, a well-adapted quasi-equilibrium state), increased interspecific competition (32.74 ± 4.77%) and higher small-world property facilitated the rapid information transportation and pollutant treatment, and keystone species were positively correlated with the removal of COD and NH 4 + -N. Homogenizing dispersal diminished the contemporary community dissimilarities, while turnover but rather nestedness governed the temporal variations in the biofilm succession period. This study highlighted the specificity of ecological processes at distinct biofilm development phases, which would advance our understanding on the development-to-function linkages in biofilm-based treatment processes.

No abstract

No abstract

Substantial attempts have been made to control microbial communities for environmental integrity, biosystem performance, and human health. However, it is difficult to manipulate microbial communities in practice due to the varying and nonlinear nature of interspecific interaction networks. Here, we develop a manifoldbased framework to investigate the patterns of microbial interaction variability in wastewater treatment plants using manifold geometric properties and design a simple control strategy to manipulate the microbes in nonlinear communities. We validate our framework using the readily available and nonsequential microbiome profiles of wastewater treatment plants. Our results show that some microbes in the activated sludge and anammox communities display deterministic rival or cooperative relationships and constitute a stable subnetwork within the whole nonlinear community network. We further use a simulation to demonstrate that these microbes can be used to drive a microbe in a target direction regardless of the community dynamics. Overall, our framework can provide a time-efficient solution to select effective control inputs for reliable manipulation in varying microbial networks, opening up new possibilities across a range of biological fields, including wastewater treatment plants.