Back

Microbial association networks reveal hidden keystone taxa and cross-kingdom interactions after dryland wildfires

Joukhajian, A.; Pulido Barriga, M. F.; Homyak, P.; Glassman, S. I.

2026-07-29 microbiology
10.64898/2026.07.28.741306 bioRxiv
Show abstract

How wildfires reorganize soil microbial interactions is a key knowledge gap, particularly for drylands that cover nearly 40% of Earths surface and face increasing wildfire frequency with global change. We compared bacterial, fungal, and cross-kingdom association networks across four timepoints from 2 weeks to 3 years post-fire in two California dryland systems: a high-intensity chaparral shrubland fire and a low-intensity Eastern Joshua tree desert fire, using nearly identical sampling designs and molecular workflows. Wildfire increased bacterial-fungal associations more than bacterial or fungal interactions in both systems, with burned plots consistently shifting toward cooperative over competitive associations. Bacterial-fungal interactions also increased in burned relative to unburned desert plots, suggesting fire promoted microbial associations in desert soils. Although microbial richness declined by up to 61% one year after chaparral wildfire but remained unchanged in the desert, network clustering declined in both systems, indicating reduced community resilience independent of richness loss. Pyrophilous bacteria, including Massilia and Noviherbaspirillum, emerged as keystone taxa after chaparral wildfire, while generalist bacteria and the putatively pyrophilous Pyronemataceae fungus Pseudotricharina structured desert burned networks. Cross-kingdom network analysis revealed shifts in post-fire microbiomes invisible to traditional diversity metrics, highlighting bacterial-fungal interactions and keystone taxa as drivers of dryland post-fire succession.

Matching journals

The top 11 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.