A reciprocal soil transplant experiment was initiated in 1994 in eastern Washington in which soil cores were transplanted between two elevations (310 m and 844 m); the lower site is warmer and drier, and the upper site is cooler and wetter. After 17 years, the bacterial community structure did not change significantly, although microbial function (enzymes, soil respiration) did. We resampled the transplanted cores the following year to characterize the soil organic matter chemistry (Fourier-Transform Ion Cyclotron Resonance Mass Spectrometry, “FTICR MS”), microbial community functional capability (metagenomics), and the biochemical potential of carbohydrate-active enzymes (assays).
Results/Conclusions
Rigorous statistical approaches were applied to extract distinctive features from the FTICR MS and metagenomic datasets to identify features that are differentially abundant. We found specific empirical formulae that discriminated soils by treatment, and developed methods to mine KEGG pathways to link those formulae with the enzyme assay data and metagenomic features using Trelliscope, an interactive statistical tool for exploratory data analysis. In this way, we identified the critical metabolic pathways that were significantly associated with treatment differences in both the metagenome (microbial capabilities) and in the chemical profile, indicating execution of these processes. For example, we identified impaired lysine biosynthesis in the lower site, through missing enzymatic capabilities in the metagenome in conjunction with absence of the coordinating reactions in the chemical data. The ability to infer biological and chemical bottlenecks from disparate sets of shotgun data presents new opportunities for exploring soil C persistence and vulnerability through the lens of microbial metabolism.