Microbial composition and function across an old-field chronosequence

Background/Question/Methods

Due to gradual shifts in plant species composition and litter addition through time, secondary succession can be used as a model to understand how plant communities shape soil microbial community composition and function. Because microbial communities in soil are structured by resource availability, we hypothesized that changes in plant community composition and litter inputs to soil during secondary succession should shape the composition and function of microbial communities. It follows that both plant and saprotrophic communities change concomitantly during secondary succession. To test this idea, we investigated changes in microbial composition and functional potential across a series of nine established old fields, ranging from 16 to 86 years since agricultural abandonment, as well as three forests representing potential late-successional ecosystems. Microbial composition was quantified according to fungal and bacterial species richness and β-diversity. We quantified functional potential using two approaches: shotgun metagenomics to calculate the composition of genes encoding enzymes involved in the decay of plant litter and extracellular enzyme assays.  Plant, soil, and root characteristics were collected to understand the primary factors contributing to differences in microbial function during secondary succession.

Results/Conclusions

Fungal communities and functional gene assemblages, as well as enzyme activity varied with soil and plant characteristics across the old field chronosequence, providing evidence that microbial composition and function shifts parallel changes in plant composition and litter inputs to soil during secondary succession.  Mantel results revealed that pairwise variation in site age correlated to fungal β-diversity and enzyme activity (R = 0.34 - 0.42, P = 0.014 - 0.015). Redundancy analysis indicated that soil organic matter alone composed the best model explaining fungal b-diversity and enzyme activity (F_1,7 = 1.2 – 13.7, P < 0.01). Further, variation in fungal functional gene composition was significantly predicted by the change in relative dominance of C3 grasses, the plant functional group with relatively low lignin litter (F_1,7 = 3.7, P = 0.02). However, no model significantly explained variation in bacterial community composition or functional gene composition. Together, these results suggest that the accumulation of soil organic matter and litter biochemistry shape fungal composition and function through succession, while having little impact on bacterial communities. Our results provide evidence that change in plant community composition during secondary succession modifies resource availability, which, in turn, appears to structure the composition and function of fungal communities in soil.