Antimicrobial resistance (AMR) is the ability of microbes to survive exposure to antimicrobial agents. Even though AMR is not a disease per se, it is claiming more lives (directly and indirectly) than malaria and HIV combined. While a lot of research has been done in this area, most studies do not account for the fact that in nature, bacterial populations are genetically diverse and live in multispecies communities where they interact with each other and the environment. Importantly, the spread and evolution of AMR can be influenced by such interactions. This thesis focuses on two processes that can promote AMR in the environment: horizontal transfer of antibiotic resistance genes (ARGs) in polymicrobial communities and co-selection of resistance-like phenotypes by exposure to other biocides. In paper I, bioreactors were established to grow and maintain a stable polymicrobial community over longer timescales, enabling studies of how a promiscuous plasmid (pKJK5) disseminates ARGs among microbial members in the community. The plasmid was assimilated by less abundant community members, while in contrast one of the most abundant members remained plasmid-free. The composition and dynamics of the polymicrobial community changed during and after antibiotic treatment depending on the absence/presence of the plasmid. The presence of the plasmid also seemed to prevent a bloom of a pathogenic member of this microbial community. In paper II, we went beyond the taxonomy of plasmid hosts to identify and describe genome-encoded plasmid transfer barriers from metagenomes and single cell genomes. There was a clear barrier for plasmid transfer related to the biochemical Gram classification. Comparing genomes of transconjugants to non-transconjugants, there were multiple enriched genomic differences in genes involved in cell envelopes and post-translational regulations while canonical plasmid transfer barriers such as presence of other incompatible plasmids or bacterial defense systems did not seem to be a major constraint for the spread of pKJK5. In paper III, we explored co-selection of zinc and antibiotic resistance through a series of short term experimental evolution incubations with a Gram positive environmental isolate. There was an increased antibiotic tolerance in isolates previously grown in high concentrations of zinc. While no plasmid-encoded resistance genes were found (often causing co-selection), specific chromosomal mutations were distinct to either the zinc-evolved or control isolates. Our results also highlight the need for appropriate control lines that account for domestication, as the control line in our study increased susceptibility to antibiotics tested when compared to the originally isolated parental strain. The studies advances our understanding of evolution and spread of ARGs in natural polymicrobial communities and populations, and can in the long run help us forecast and model such processes in a more mechanistic way. While AMR can spread rapidly across communities, such transfer still encounters barriers that need further investigation. The cell envelope seems to be one important barrier to horizontal transfer of ARGs, but the studies also reveal ecological roles of resistance-plasmids in polymicrobial communities and a role for selective pressures other than antibiotics in fostering AMR.