The Interplay of physical and community complexity in soil systems

Data obtained in pure cultures of the different microbial groups offers a wide range of information of each specific strain, yet how many of those patterns are present in natural environments remains unclear. On the other hand, measurements of soil functions in field or laboratory settings are packed with unhandled parameters that are likely to explain most of the unexplained variation encountered. There is hence the need for approaches that handle at the same time the control and resolution permitted in pure culture settings but with the parameters present in natural microbiomes at the relevant scale. With the use of microfluidics, fluorescence microscopy, and genomic tools, we explored two examples of the implications of soil characteristics on bacterial interactions in pairwise and in community level.

We tested how does the interaction between the two mutually exclusive soil bacterial strains Pseudomonas putida and Bacillus subtilis, holds in microenvironments of various levels of complexity. In low-complexity environments both species showed lower growth than when growing by themselves, which differed from a well-mixed liquid environment where Pseudomonas putida outperformed and inhibited Bacillus subtilis. Fragmented mazes, however, allowed not only the coexistence of both strains, but in the right frequencies permitted them to reach higher yields than when growing separately, thus turning competition into collaboration. Spatial analysis of the space within the mazes indicates that complex mazes allowed colocalization and that the level of such colocalization was linked to the yield of both strains.

Extrapolating patterns from pairwise studies to entire communities can be challenging yet necessary. What we intended in the next set of experiments was to evaluate how does community function depends on its diversity and how these two are linked to spatial characteristics of the microenvironment. A natural soil inoculum was subjected to a series of dilutions to obtain an array of inoculums with decreasing levels of diversity. Each community was then incubated within microenvironments with different levels of complexity where their capacity for substrate enzymatic degradation was measured. We expected variance between replicates of each maze to increase in inoculums of lower levels of diversity, as the founder effect would become more important than in robust entire communities. We found, however, that the enzymatic degradation of the inoculum decreased below detection limits after the third dilution (0.01X). Also, enzymatic degradation of the entire community and the 0.1-fold dilutions depended on the maze type but was consistent in the trend of higher complexity leading to higher degradation. Metagenomic quantification revealed that diluting the initial inoculum effectively reduced the diversity of it and its composition resembled more the one from the last day of incubation experiments. Hence it is apparent that the most abundant bacteria are not the ones responsible for the evaluated function, which complement recent findings which show that high abundant taxa grow slower than low abundant ones by adding that low abundance and fast grower taxa might be the drivers of the entire soil functions.