The potential for feedbacks between fire and soil fungi and in pyrogenic environments

Background/Question/Methods

In fire-frequented habitats plants producing pyrogenic fuels generate feedbacks on fire characteristics, thereby engineering fires and influencing vegetation. Soil fungi also affect fuels directly by altering litter decomposition and indirectly by affecting plant diversity and productivity, with effects varying among functional groups. Fungi also vary in susceptibility to fire, so fire may drive specific changes in fungal communities, producing predictable effects on fuels and the probability of future fires. We developed a conceptual model for “fire-fungal feedbacks” based on fire severity, then conducted a field experiment to compare differences in 1) fungal communities and 2) decomposition of new litter between recently burned and unburned (2 years) sites in an old-growth pine savanna (Wade Tract). Pine litter is more recalcitrant and produces hotter fires than savanna grasses, and pines host different symbiotic fungi than grasses. Therefore, experimental treatments also included sampling sites in the vicinity of pines or 10+ meters from pine overstory. Litter and soils were collected and high-throughput DNA sequencing of all fungi was used to analyze differences in fungal communities. Decomposition of new, site-specific litter (sterilized via gamma irradiation) was assessed using litter bags (30 µm mesh to exclude non-microbes) collected 3, 6, and 9 months after deposition.  

Results/Conclusions

Fire shifted soil fungal communities, eliminating some species and facilitating others. Sequencing of the 28S region produced 9407 OTU’s: 2809 unique to unburned sites, 2190 unique to burned sites, and 4408 shared. Community structure was significantly different between burned and unburned sites whether using OTU abundance or presence/absence only. Ongoing phylogenetic analyses will indicate if specific groups were favored or eliminated. Fungal communities were not distinct near or away from pines, possibly because fungi from pines and herbaceous plants overlapped at the spatial scale sampled. Data collection for decomposition indicated that new litter decomposition averaged 30% less in burned than unburned sites by 3 months after deposition. If these differences continue over time, fine fuels should accumulate much faster in burned sites, consistent with our fire-fungal feedback model. This research indicates that fire selects for specific fungi only present in recently burned sites and that fungi and other microbes breakdown new fuels much more slowly in recently burned than in not recently burned sites. Future research will be discussed to explicitly link fungi to decomposition differences, explore fungal effects on plant fuel production, and links these to the probability and/or severity of future fires.