Increases in aridity lead to drastic shifts in the assembly of dryland complex microbial networks

We have little information on how and why soil microbial community assembly will respond to predicted increases in aridity by the end of this century. Here, we used correlation networks and structural equation modeling to assess the changes in the abundance of the ecological clusters including potential winner and loser microbial taxa associated with predicted increases in aridity. To do this, we conducted a field survey in an environmental gradient from eastern Australia and obtained information on bacterial and fungal community composition for 120 soil samples and multiple abiotic and biotic factors. Overall, our structural equation model explained 83% of the variance in the two mesic modules. Increases in aridity led to marked shifts in the abundance of the two major microbial modules found in our network, which accounted for >99% of all phylotypes. In particular, the relative abundance of one of these modules, the Mesic Module #1, which was positively related to multiple soil properties and plant productivity, declined strongly with aridity. Conversely, the relative abundance of a second dominant module (Xeric Module #2) was positively correlated with increases in aridity. Our study provides evidence that network analysis is a useful tool to identify microbial taxa that are either winners or losers under increasing aridity and therefore potentially under changing climates. Our work further suggests that climate change, and associated land degradation, could potentially lead to extensive microbial phylotypes exchange and local extinctions, as demonstrated by the reductions of up to 97% in the relative abundance of microbial taxa within Mesic Module #1

Alexander von Humboldt Foundation; Ministerio de Economía y Competitividad, Grant/Award Number: CGL2013-44661-R; European Research Council, Grant/Award Number: 647038; Hermon Slade Foundation; Marie Sklodowska-Curie Actions of the Horizon 2020 Framework Programme, Grant/Award Number: 702057; Rutherford Discovery Fellowship; Marsden Fast-Start Grant; Australian Research Council, Grant/Award Numbers: DP170104634, DP190103714

Peer reviewed