Effect of bovine urine and nitrification inhibitor DCD on ammonia-oxidising communities in three NZ soils as revealed by pyrosequencing of the amoA functional genes

Ammonia oxidising microorganisms (AOM) oxidise ammonia to nitrite and play a crucial role in the nitrogen cycling in all environments. In intensively-grazed pastoral soils AOM have been implicated as the key microorganisms responsible for adverse environmental impacts such as nitrate leaching, eutrophication and nitrous oxide (a potent greenhouse gas and ozone depleting substance) emissions. Application of nitrification inhibitors to pastures is an effective means to reduce the environmental impact of intensive agriculture. The objective of this work was to investigate the effects of high nitrogen (N) input in form of bovine urine and nitrification inhibitor (DCD) application on ammonia-oxidising bacteria (AOB) and ammonia-oxidising archaea (AOA) communities in three intensively-grazed pasture NZ soils. Soils were collected from Canterbury (CA), Waikato (WK) and Southland (SL) dairy soils and treated with bovine urine (U), urine + nitrification inhibitor DCD (U+DCD) and water as a control (C). After 40 days of incubation, total genomic DNA was extracted, PCR-amplified targeting AOB and AOA amoA genes and sequenced using 454 pyrosequencing (Roche). The reads were subsequently de-convoluted, QC-checked, frame-corrected (HMM-FRAME) and analysed using a set of bioinformatics and statistical tools. There were 43466 and 65470 reads, clustering at 95% sequence similarity into 213 and 65 operational taxonomic units (OTU) for AOB and AOA respectively. AOB communities in all three soils were represented almost exclusively by species with high similarity to Nitrosospira species, whereas taxonomic classification of AOA OTUs was not possible beyond Phylum Crenarchaeota phylum due to the lack of extensive reference databases. Bray-Curtis similarity matrices were constructed from the reduced and fourth root-transformed tables, followed by the non-metric multidimensional scaling (nMDS) analysis. Multivariate statistical test (ANOSIM) identified significant differences (p