Microbes in sediments below fish farming cages are exposed to a variety of compounds used in aquaculture, such as antifoulants, antimicrobials and mercury. Here, we assessed the abundance of the mercury resistance gene merA in the sediment of four fish farms and four pristine sites using quantitative PCR, and at three fish farms and two pristine sites by constructing clone libraries and sequencing merA amplicons. The abundance of merA was consistently greater at fish farm sites than at pristine sites, and the merA phylotypes were different between fish farm and pristine sites. Mercury concentrations were too low to cause selection pressure and did not differ significantly among samples. The primers designed specifically for this study were capable of detecting two new potential merA clades.