As part of a three year study of the use of reed beds for the treatment of
wastewater, two pilot scale constructed wetlands (30m x 5m), one planted wi
th Phragmites australis and the other serving as an unplanted control, were
used to treat primary domestic sewage. This paper describes an experiment
conducted to test the hypothesis that increasing available oxygen by aerati
on of the wastewater in the treatment beds could increase nitrification and
hopefully nitrogen removal rates. Prior to aeration, the influent nitrogen
load in the wastewater passing through the beds was reduced by approximate
ly 45% and 10% in the planted and control beds respectively. With aeration
of the wastewater as it passed through each bed, nitrogen removal was appro
ximately 51% and 20% for the planted and control beds respectively. High ra
tes of nitrification were recorded for the planted bed with aeration, but r
emoval of nitrogen by denitrification was limited, probably because of the
absence of a suitable carbon source (indicated by low BOD5 concentrations).
High nitrification rates were not recorded in the unplanted control bed. T
he higher BOD5 concentration measured in this bed may have caused increased
competition for available oxygen between heterotrophic bacteria and the re
latively slow-growing nitrifying bacteria, resulting in lower rates of nitr
ification but more efficient denitrification.