Interactions between environmental factors shape the composition and function of microbial communities
Environmental factors often combine non additively to shape the functions and taxonomic composition of microbial communities and the traits and phenotypes of microbial populations (Dal Bello et al., 2021; Estrela et al., 2021a, 2021b; Fonte et al., 2013; Lázár et al., 2022; Pacheco et al., 2021; Pacheco and Segrè, 2021; Replansky and Bell, 2009; Smith et al., 2023; Wood et al., 2012). In what follows, we provide a few recent examples of environmental interactions in the context of community composition and function, with the aim of illustrating this point.
Nutrients interact non additively to determine community composition. Microbial communities assemble in complex environments that contain multiple resources. These resources can have different effects in different members of the community: a molecule that is discarded by a microorganism is often useful for others (D’Souza et al., 2018; Goldford et al., 2018; Mataigne et al., 2021). In a recent study (Estrela et al., 2021b), we have systematically investigated how nutrients combine to shape the taxonomic composition of microbial communities. Through an enrichment approach from soil and plant communities (described in detail elsewhere (Estrela et al., 2021a)), we studied the taxonomic composition of stable, serially passaged communities in media containing various sets of two growth limiting nutrients, and compared these with the community compositions that formed in each nutrient separately. We found that interactions between nutrients were often strong. Increasing the number of supplied resources from one to two or even three had no effect on the total diversity of our enrichment communities (Estrela et al., 2021a), and this was often due to an effect we termed “nutrient dominance”, where one of the two nutrients masked the ability of the other to recruit taxa into the community (Estrela et al., 2021b). For example, the taxonomic composition of enrichment communities where both glycerol and succinate were supplied externally as the only carbon sources were nearly indistinguishable from the composition of the communities assembled in glycerol alone. The majority of the taxa that were coexisting stably when succinate was the only supplied nutrient were excluded from the community when glycerol was also supplied in the culture medium at the beginning of every passage (Estrela et al., 2021b).
These results echo similar findings from other groups. Pacheco and Segre (Pacheco et al., 2021) found strong non-additive interactions between nutrients in serially passaged, defined communities, and Dal Bello et al (Dal Bello et al., 2021) found that community richness increased very weakly as we increase the number of supplied resources, barely doubling as we went from 1 to 32. Together with older evidence (Replansky and Bell, 2009) these studies support the idea that nutrients often interact non additively in their effect on community composition. It stands to reason that such non-additivity will extend to function also.
Environmental variables interact non additively to determine community functions. In addition to their non-additive effect on the taxonomic composition of a community, the effect of two or more environmental variables on community functions is also often non additive. For instance, Jimenez et al studied the combined effect of different environmental variables on the methanogenic output of an anaerobic digestor community (Jiménez et al., 2014). Specifically, these authors examined the combined effect of three substrates (pig manure, rice straw and clay residual) at different concentrations on the specific methanogenic activity of the digester community, and employed a fractional factorial design followed by a linear regression with interaction terms to predict out of sample. Their approach revealed the existence of strong pairwise interactions between the substrates, particularly for clay residual and rice straw under mesophilic conditions. The interactions were strong enough to change the sign effect of increasing clay concentration at high vs low manure concentrations: increasing clay concentration had a positive effect on the methanogenic activity when manure concentration was low, but a negative effect when manure concentration was high. Sign-effect interactions (akin to sign epistasis in genetics) are common in microbial consortia (Sanchez et al., 2023).
In another study, Molina-Barahona et al sought to optimize the environmental conditions for diesel removal in contaminated soils through biostimulation (Molina-Barahona et al., 2004). To this end, they investigated the combinatorial effects of various environmental factors that are known to individually have an effect on diesel removal rate: C:N ratio, moisture, and the amount and type of crop residue (CR). These factors were individually toggled between two values and combinations were generated through a fractional factorial design approach. The authors then found that interactions between moisture and CR type had a modest but significant effect on the amount of diesel removed in microcosms, as the amount of diesel removed in the presence of both factors at the same time was higher than expected by the product of each in isolation (Molina-Barahona et al., 2004) .
On a similar vein, Zhou et al, studied the combined effect of three binary variables: glucose addition, sodium dodecyl benzene sulfonate (SDBS) addition, and immobilization of the bacteria on biochar (or lack thereof) on the biodegradation of polycyclic aromatic hydrocarbons (PAH) by a microbial consortium (Zhou et al., 2023). While glucose alone stimulated PAH biodegradation, it had a negligible effect when combined with SDBS unless the consortia was also immobilized. In this instance, glucose had a modest (but seemingly significant) positive effect on the degradation of low molecular weight PAH. The authors propose a mechanistic explanation for this three-way interaction: Glucose had been previously found to stimulate the metabolism of PAHs and it is also known to promote microorganism to absorb more PAHs. In turn, SDBS is believed to increase the bioavailability of PAHs by enhancing their mobility and bringing them closer to the immobilized bacteria (Zhou et al., 2023).
All of these examples illustrate the potential effect of environmental interactions in the performance of single and multi-strain microbial consortia. On the one hand, the functional response of microbial consortia to environmental manipulations creates opportunities for their optimization. On the other hand, the combinatorial complexity of these environmental interaction effects may appear to be dissuasive. While it is indeed undoubtedly challenging, as we will see below there exists clear and convincing evidence that it is far from impossible.