Correction to: The ISME Journal (2015) 9, 2153–2161; doi:10.1038/ismej.2015.26; published online 24 April 2015

Since the publication of this paper, the author has noticed an error in the third paragraph of Materials and Methods (under ‘Chemostat reactor operation’). It should read as follows (changes are underlined):

The culture media was autoclaved before use and sparged with a small flow of nitrogen gas while connected to the chemostat. Medium A contained per liter (day 0–271): 14.7 mmol KH2PO4, 0.8 mmol MgSO4.7H2O, 1.0 mmol NaOH, 10 mg yeast extract, 4 ml trace element solution (Vishniac and Santer, 1957), with only 2.2 g ZnSO4.7H2O, and NaNO3 and NH4Cl. The concentration of NaNO3 was 13.3 mM (day 0 untill 39), or 11.8 mM (from day 39). NH4Cl concentrations were 0.5 mM (day 26-68), 0.7 mM (day 1-26, 68-82 and 94–122), 1.5 mM (day 82–94 and 122–186) and finally omitted (from day 186). Medium B contained, per liter, initially 5.5 mM NaCH3COO.3H2O; this was gradually increased to 8.8 mM (day 26 untill 39), 10.3 mM (day 39 untill 47), 12.5 mM (day 47 untill 122), 16.2 mM (day 133 untill 140) or 19.8 mM (day 122–271).

The changed numbers are double of what was written in the manuscript. The initial numbers did not correspond to the medium concentrations, but the concentrations in the influent (while referred to as medium concentrations). The influent was a combined flow of equal amount of both media, hence, a dilution factor of 2.

The author would like to apologize for any inconvenience this may have caused.