| Popis: |
Despite the global importance in ecological processes, the Amazon rainforest has been subjected to high rates of deforestation, mostly for pasturelands, over the last few decades. In this study, we used a combination of deep shotgun metagenomics and a machine learning approach to compare physiological strategies of microbial communities between contrasting forest and pasture soils. We showed that microbial communities (bacteria, archaea and viruses), and the composition of protein-coding genes are distinct in each ecosystem. The diversities of these metagenomic datasets are strongly correlated, indicating that the protein-coding genes found in any given sample of these soil types are predictable from their taxonomic lineages. Shifts in metagenome profiles reflected potential physiological differences caused by forest-to-pasture conversion with alterations in gene abundances related to carbohydrate and energy metabolisms. These variations in these gene contents are associated with several soil factors including C/N, temperature and H++Al3+ (exchangeable acidity). These data underscore that microbial community taxa and protein-coding genes co-vary. Differences in gene abundances for carbohydrate utilization, energy, amino acid, and xenobiotic metabolisms indicate alterations of physiological strategy with forest-to-pasture conversion, with potential consequences to C and N cycles. Our analysis also indicated that soil virome was altered and shifts in the viral community provide insights into increased health risks to human and animal populations. |
