Why are some groups of bacteria more diverse than others? We hypothesize that the metabolic energy available to a bacterial functional group (a biogeochemical group or ‘guild’) has a role in such a group’s taxonomic diversity. We tested this hypothesis by looking at the metacommunity diversity of functional groups in multiple biomes. We observed a positive correlation between estimates of a functional group’s diversity and their metabolic energy yield. Moreover, the slope of that relationship was similar in all biomes. These findings could imply the existence of a universal mechanism controlling the diversity of all functional groups in all biomes in the same way. We consider a variety of possible explanations from the classical (environmental variation) to the ‘non-Darwinian’ (a drift barrier effect). Unfortunately, these explanations are not mutually exclusive, and a deeper understanding of the ultimate cause(s) of bacterial diversity will require us to determine if and how the key parameters in population genetics (effective population size, mutation rate, and selective gradients) vary between functional groups and with environmental conditions: this is a difficult task.

Hypothesis and theory-based studies in microbial ecology have been neglected in favour of those that are descriptive and aim for data-gathering of uncultured microbial species. This tendency limits our capacity to create new mechanistic explanations of microbial community dynamics, hampering the improvement of current environmental biotechnologies. We propose that a multiscale modelling bottom-up approach (piecing together sub-systems to give rise to more complex systems) can be used as a framework to generate mechanistic hypotheses and theories (in-silico bottom-up methodology). To accomplish this, formal comprehension of the mathematical model design is required together with a systematic procedure for the application of the in-silico bottom-up methodology. Ruling out the belief that experimentation before modelling is indispensable, we propose that mathematical modelling can be used as a tool to direct experimentation by validating theoretical principles of microbial ecology. Our goal is to develop methodologies that effectively integrate experimentation and modelling efforts to achieve superior levels of predictive capacity.

Is it possible to find trends between the parameters that define microbial growth to help us explain the vast microbial diversity? Through an extensive database of kinetic parameters of nitrifiers, we analyzed if the dominance of specific populations of nitrifiers could be predicted and explained. We concluded that, in general, higher growth yield (Y_XS) and ammonia affinity (a⁰_NH3) and lower growth rate (µ_max) are observed for ammonia-oxidizing archaea (AOA) than bacteria (AOB), which would explain their considered dominance in oligotrophic environments. However, comammox (CMX), with the maximum energy harvest per mole of ammonia, and some AOB, have higher a⁰_NH3 and lower µ_max than some AOA. Although we were able to correlate the presence of specific terminal oxidases with observed oxygen affinities (a⁰_O2) for nitrite-oxidizing bacteria (NOB), that correlation was not observed for AOB. Moreover, the presumed dominance of AOB over NOB in O₂-limiting environments is discussed. Additionally, lower statistical variance of a⁰_O2 values than for ammonia and nitrite affinities was observed, suggesting nitrogen limitation as a stronger selective pressure. Overall, specific growth strategies within nitrifying groups were not identified through the reported kinetic parameters, which might suggest that mostly, fundamental differences in biochemistry are responsible for underlying kinetic parameters.

In microbial communities, the ecological interactions between species of different populations are responsible for the spatial distributions observed in aggregates (granules, biofilms or flocs). To explore the underlying mechanisms that control these processes, we have developed a mathematical modelling framework able to describe, label and quantify defined spatial structures that arise from microbial and environmental interactions in communities. An artificial system of three populations collaborating or competing in an aggregate is simulated using individual-based modelling under different environmental conditions. In this study, neutralism, competition, commensalism and concurrence of commensalism and competition have been considered. We were able to identify interspecific segregation of communities that appears in competitive environments (columned stratification), and a layered distribution of populations that emerges in commensal (layered stratification). When different ecological interactions were considered in the same aggregate, the resultant spatial distribution was identified as the one controlled by the most limiting substrate. A theoretical modulus was defined, with which we were able to quantify the effect of environmental conditions and ecological interactions to predict the most probable spatial distribution. The specific microbial patterns observed in our results allowed us to identify the optimal spatial organizations for bacteria to thrive when building a microbial community and how this permitted co-existence of populations at different growth rates. Our model reveals that although ecological relationships between different species dictate the distribution of bacteria, the environment controls the final spatial distribution of the community.

The importance of microbial communities (MCs) cannot be overstated. MCs underpin the biogeochemical cycles of the earth's soil, oceans and the atmosphere, and perform ecosystem functions that impact plants, animals and humans. Yet our ability to predict and manage the function of these highly complex, dynamically changing communities is limited. Building predictive models that link MC composition to function is a key emerging challenge in microbial ecology. Here, we argue that addressing this challenge requires close coordination of experimental data collection and method development with mathematical model building. We discuss specific examples where model-experiment integration has already resulted in important insights into MC function and structure. We also highlight key research questions that still demand better integration of experiments and models. We argue that such integration is needed to achieve significant progress in our understanding of MC dynamics and function, and we make specific practical suggestions as to how this could be achieved.