Forest soil metagenomes demonstrate chronic soil warming induces changes in microbial functional potential
An immediate response to chronically elevatedsoil temperatures is accelerated rates of carbon and nitrogen cycling. These processes are driven by microbial activity, and are known to be sensitive to the composition and functional diversity of the microorganisms present. We have found relatively subtle changes in the bacterial community in response to warming, indicating that changes in the relative abundance of key functional pathways may be responsible for observed changes in function. Here we generated 48 metagenomes from three long-term warming experiments to explore the effect of long-term warming on the genetic capacity and diversity of organisms present in soil. Reads were annotated using the Joint Genome Institute's IMG annotation pipeline, as well as Diamond against the CAZy database. We used DESeq to identify genes which differed significantly between heated and control plot soils, and then examined the complete pathways these genes are associated with using metabolic network analysis.
Bacteria accounted for 99% of the protein-coding reads, with Acidobacteria, Actinobacteria, and Alphaproteobacteria dominating the community. Warming treatment had a small but significant impact on both community structure and potential function. Genes identified as significantly enriched in heated mineral soil based on DESeq included peptidoglycan hydrolase and N-acetylglucosaminyl deacetylase, while those depleted by warming included phage-related genes and transposases. Fewer genes were affected by warming in the organic horizon, possibly as a consequence of greater within-treatment heterogeneity there. A separate analysis of complete metabolic pathways affected by warming treatment confirmed many of these results, and also identified typically-eukaryotic pathways such as N-glycan biosynthesis as underrepresented in heated plot metagenomes. Warming also affected the taxonomic affiliation of reads associated with the degradation of key biopolymers. This exploratory approach to identifying genes demonstrated that while some changes in the relative abundance of genes can be predicted by altered ecosystem function, others, such as horizontal gene transfer, cannot.