Perennial grass-based agroecosystems are under consideration as sustainable sources of bioenergy feedstocks. Establishing these systems on land previously used for conventional agricultural production is expected to dramatically alter the composition and functional capacity of their associated soil bacterial communities, but the rate at which these changes will occur is unclear. Methods for characterizing bacterial communities are both varied and useful for documenting different aspects of the soil microbiota and their dynamics during this transition. Here, we studied the soil-associated bacterial communities of continuous corn and restored prairies systems within a cropping systems experiment 2-4 years after establishment using 1) phospholipid fatty acid (PLFA) profiling, 2) shotgun metagenomic sequencing, 3) amplicon sequencing of the 16S rRNA gene and 4) sequencing of the nitrogen-cycling gene nosZ. All characterization methods discriminated the bacterial communities between the two cropping systems, but the largest differences were observed with PLFA profiling. Differences between the two cropping systems did not significantly increase during the study period. The community compositions described by sequence-based methods were mutually correlated, but were only weakly correlated to the composition described by PLFA profiling. Shotgun metagenomics detected a much higher abundance of Actinobacteria than amplicon sequencing and revealed more consistent changes between cropping systems over time. Cropping system and interannual effects on the ratios of biomarkers associated with Gram-negative and Gram-positive bacteria were entirely different for PLFAs, rRNA amplicons, and shotgun sequenced 16S rRNA. Our findings highlight how soil bacterial community characterization methods differ in their detection of microbial community composition as a result of recent land use change. (C) 2016 Elsevier Ltd. All rights reserved.