Based on our results, we suggest that food web models should account for bacterial secondary consumption since this has important consequences for carbon fluxes with more carbon dioxide released by microbes and less microbial carbon available for the soil animal food web.