COS 99-1 - Burning does not shape grassland soil microbial diversity recovery from chronic nitrogen fertilization

Abstract: Global industrialization has increased nitrogen (N) availability across terrestrial ecosystems. N from atmospheric deposition or fertilization can shift soil microbial community composition, which in turn impacts nutrient cycling when functionally important microbial taxa change in abundance. However, fire disturbance could mitigate increased N availability and its consequences, because fire consumes plant litter, removing organic N from the ecosystem. Furthermore, since N deposition rates are now in decline, understanding the recovery of microbial structure and function from chronic N addition is increasingly important. We asked how prescribed fire affects the recovery of soil microbial community diversity and composition following the cessation of thirty years of N fertilization in a mesic tallgrass prairie. We predicted that annual burning would support greater soil microbial recovery from chronic N addition relative to co-located unburned areas. We used a long-term fertilization experiment at Konza Prairie Biological Station (Manhattan, KS, USA) where burned and unburned prairies were either unfertilized (control) or fertilized with N (10 g N m^-2 y^-1 as NH₄NO₃) for thirty years until 2017 when fertilization ceased. For the next five years (2017-2021), we measured soil bacterial and archaeal 16S rRNA gene composition twice each growing season in control and recovering plots, and continuously fertilized sub-plots.



Soil microbial community composition between control soils, and both previously and continuously fertilized soils, was well separated in NMDS space, and more strongly in the unburned than the burned treatment. However, there was no evidence of recovery in either fire treatment, and according to resistance and resilience indices based on Bray-Curtis similarity, burned prairies were neither more resistant nor resilient to fertilization than unburned prairies. Observed amplicon sequence variant richness varied strongly by year (P<0.001), and the temporal variability of both Shannon diversity and evenness differed among fertilization treatments (Fertilization×Year: P<0.05), but was unaffected by fire. Soil microbial community composition also changed interannually, explaining 16.8% of the variation (PERMANOVA: P=0.001), and the NMDS showed that community composition shifted along a similar trajectory for all field treatments, particularly between the dry and wet 2018 and 2019 growing seasons. Overall, while soil microbial community sensitivity to fertilization was greater in the absence of fire, prescribed fire did not increase recovery from chronic N-fertilization. Instead, there appears to be a fertilization legacy on soil microbial community composition that prescribed fire does not reverse. In conclusion, this study suggests chronic N-fertilization can push soil microbial communities to a new ecological state.