Metagenomics reveals bacterial and archaeal adaptation to urban land-use: n catabolism, methanogenesis, and nutrient acquisition

Abstract

Urbanization results in the systemic conversion of land-use, driving habitat and biodiversity loss. The “urban convergence hypothesis” posits that urbanization represents a merging of habitat characteristics, in turn driving physiological and functional responses within the biotic community. To test this hypothesis, we sampled five cities (Baltimore, MD, United States; Helsinki and Lahti, Finland; Budapest, Hungary; Potchefstroom, South Africa) across four different biomes. Within each city, we sampled four land-use categories that represented a gradient of increasing disturbance and management (from least intervention to highest disturbance: reference, remnant, turf/lawn, and ruderal). Previously, we used amplicon sequencing that targeted bacteria/archaea (16S rRNA) and fungi (ITS) and reported convergence in the archaeal community. Here, we applied shotgun metagenomic sequencing and QPCR of functional genes to the same soil DNA extracts to test convergence in microbial function. Our results suggest that urban land-use drives changes in gene abundance related to both the soil N and C metabolism. Our updated analysis found taxonomic convergence in both the archaeal and bacterial community (16S amplicon data). Convergence of the archaea was driven by increased abundance of ammonia oxidizing archaea and genes for ammonia oxidation (QPCR and shotgun metagenomics). The proliferation of ammonia-oxidizers under turf and ruderal land-use likely also contributes to the previously documented convergence of soil mineral N pools. We also found a higher relative abundance of methanogens (amplicon sequencing), a higher relative abundance of gene sequences putatively identified as Ni-Fe hydrogenase and nickel uptake (shotgun metagenomics) under urban land-use; and a convergence of gene sequences putatively identified as contributing to the nickel transport function under urban turf sites. High levels of disturbance lead to a higher relative abundance of gene sequences putatively identified as multiple antibiotic resistance protein marA and multidrug efflux pump mexD, but did not lead to an overall convergence in antibiotic resistance gene sequences