As an analogue of phosphorus, arsenic (As) has a biogeochemical cycle coupled closely with other key elements on the Earth, such as iron, sulfate and phosphate. It has been documented that microbial genes associated with As biotransformation are widely present in As-rich environments. Nonetheless, their presence in natural environment with low As levels remains unclear. To address this issue, we investigated the abundance levels and diversities of aioA, arrA, arsC and arsM genes in estuarine sediments at low As levels across Southeastern China to uncover biogeographic patterns at a large spatial scale. Unexpectedly, genes involved in As biotransformation were characterized by high abundance and diversity. The functional microbial communities showed a significant decrease in similarity along the geographic distance, with higher turnover rates than taxonomic microbial communities based on the similarities of 16S rRNA genes. Further investigation with niche-based models showed that deterministic processes played primary roles in shaping both functional and taxonomic microbial communities. Temperature, pH, total nitrogen concentration, carbon/nitrogen ratio and ferric iron concentration rather than As content in these sediments were significantly linked to functional microbial communities, while sediment temperature and pH were linked to taxonomic microbial communities. We proposed several possible mechanisms to explain these results.