Superoxide and other reactive oxygen species (ROS) in the 2 0 environment shape microbial communities 1 and drive transformation of metals 2,3 2 1 and inorganic/organic matter 4,5 . Taxonomically diverse bacteria and 2 2 phytoplankton can produce extracellular superoxide during laboratory 2 3 cultivation 6-11 . Understanding the physiological reasons for extracellular 2 4 superoxide production by aerobes in the environment is a crucial question yet 2 5 not fully solved. Here, we showed that iron-starving Arthrobacter sp. QXT-31 2 6 (referred to as A. QXT-31 hereafter) secreted a type of siderophore 2 7 (deferoxamine, DFO), which provoked extracellular superoxide production by 2 8 carbon-starving A. QXT-31 when carbon sources were recovered. Several other 2 9 siderophores also demonstrated similar effects. RNA-Seq data hinted that DFO 3 0 stripped iron from iron-bearing proteins in the electron transfer chain (ETC) of 3 1 metabolically active A. QXT-31, resulting in electron leakage from the 3 2 electron-rich (resulting from carbons source metabolism) ETC and superoxide 3 3 production. Considering that most aerobes secrete siderophore(s) 12 and often 3 4 suffer from carbon starvation in the environment, certain aerobes are expected 3 5 to produce extracellular superoxide when carbon source(s) recover/fluctuate, thus 3 6influencing the microbial community and cycling of many elements. In addition, 3 7 an artificial iron-chelator (diethylenetriamine pentaacetic acid, DTPA) was 3 8 widely used in microbial superoxide quantification. Our results showed that 3 9