Grassland soil microbial community turnover in response to long-term nitrogen management

Background/Question/Methods

Nitrogen (N) availability is a driver of soil microbial diversity and function, and is affected by land management practices like prescribed burning (N removal through volatilization) and fertilization (N addition). Because soil microbes control critical feedbacks to ecosystem function, including short-term N immobilization and sustained N mineralization, it is important to understand the growth and turnover responses of microbial populations under conditions of contrasting N availability. This study utilized a long-term field manipulation in which tallgrass prairie plots were either annually burned or unburned, and fertilized (10 g NH₄NO₃-N m^-2) or unfertilized since 1986. Composite surface soil samples (0-20 cm) were collected monthly (November 2014 - December 2015) to measure long-term management responses and seasonal turnover of soil microbial communities, with these questions in mind: 1) How do long-term manipulations of N availability alter the relative abundance of soil bacteria and fungi, and alter soil bacterial community composition? 2) Do long-term shifts in N availability impact the seasonal turnover of soil microbial communities, or modify communities beyond the seasonal range of variability? Bacterial 16S rRNA gene and fungal ITS population sizes were estimated by qPCR, and bacterial community composition (BCC) was measured using Illumina MiSeq sequencing of 16S rRNA genes.

Results/Conclusions

We predicted total microbial population sizes and diversity would be lower in soils with higher N availability (unburned and fertilized treatments), due to greater competitive dominance of nitrophilic taxa. However, bacterial and fungal population sizes varied significantly by sampling month, not with long-term treatment. Bacterial populations were at least 10x greater in summer (June-August) than other months (1.52 x 10¹⁰ > 8.08 x 10⁸, P < 0.05). In contrast, BCC was not significantly seasonally variable (P = 0.71), but was impacted by both long-term fire and fertilization treatments on BCC, such that 11.0% of variation in bacterial heterogeneity was explained by either factor or the interaction between the two (P < 0.001 for each, permutational ANOVA). However, neither bacterial richness nor evenness varied with long-term treatment (P > 0.05), so nitrophilic dominance may not explain the BCC response. These results reveal that while long-term grassland management changes BCC beyond typical levels of seasonal variability, total bacterial populations change coherently month-to-month, potentially due to plant activity supplying similar levels of labile carbon during summer. Additionally, soil microbial community turnover appears to be affected by long-term differences in management via mechanisms not necessarily mediated by seasonal cellular responses to substrate availability.