Denitrifying enzyme activity and carbon availability for denitrification following manure application

A field experiment was established which permitted the comparison of several methods to estimate carbon (C) availability for denitrifiers. Treatments consisted of a control (CONT), NH4NO3 (FERT), liquid dairy cattle manure (LCM), and solid (bedded) beef cattle manure (SBM) applied in early May and incorporated at rates of 255 kg N ha(-1) as total N (FERT) or ammoniacal N (LCM, SBM). Denitrification (dN) and N2O production rates were measured for soil cores taken from the 2.5-7.5 cm layer during a 49-d period following application. Gas production was related to denitrifying enzyme activity (DEA) and available C content. Available C content was estimated by several methods: water-soluble C (Csol), 0.5 M K2SO4-extractable C (Cext), aerobically respired C (Cresp), water-soluble phenolic acids (Cphen), and potentially available C (Cpot). All five indices of available C and DEA were strongly correlated with dN; Cresp, Cext, Cphen, and DEA were more strongly correlated with N2O production then were Cpot and Csol. The temporal patterns of dN and N2O production as well as C availability differed with the two manure treatments. These variables were relatively high early in the 49-d period with the LCM treatment but remained at relatively high levels for the SBM treatment throughout the period. These observations indicated a large quantity of readily available C applied in LCM, but a more sustained release of available C during the decomposition of bedding straw in SBM. Indices of available C were more useful than DEA in predicting the shortterm temporal pattern of dN rate. At this scale DEA was largely determined by available C supply. None of the methods used was superior for estimating C availability to denitrifiers. It is recommended that the easiest method be used for this purpose.