Impact of urine and the application of the nitrification inhibitor DCD on microbial communities in dairy-grazed pasture soils

Soil Biology & Biochemistry 88 (2015) 344e353

Contents lists available at ScienceDirect

Soil Biology & Biochemistry journal homepage: www.elsevier.com/locate/soilbio

Impact of urine and the application of the nitrification inhibitor DCD on microbial communities in dairy-grazed pasture soils Sergio E. Morales a, *, Neha Jha b, c, Surinder Saggar b a

Department of Microbiology and Immunology, Otago School of Medical Sciences, University of Otago, Dunedin, New Zealand Ecosystems and Global Change Team, Landcare Research, Palmerston North 4442, New Zealand c Earlier Soil & Earth Sciences Group, Institute of Agriculture and Environment, Massey University, Palmerston North 4442, New Zealand b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 26 November 2014 Received in revised form 12 May 2015 Accepted 6 June 2015 Available online 20 June 2015

Tools to manage the emission of the greenhouse gas nitrous oxide (N2O), an intermediate of both nitrification and denitrification, from soils are limited. To date, the nitrification inhibitor dicyandiamide (DCD) is one of the most effective tools available to livestock farmers for reducing N2O emissions and minimizing leaching of nitrogen in response to increased urine deposition in grazed pasture systems. Despite its effectiveness in decreasing N losses from animal urine by inhibiting N processes in soils, the effect of DCD on the population structure of denitrifiers and overall bacterial community composition is still uncertain. Here we use three New Zealand dairy-grazed pasture soils to determine the effects of DCD application on microbial community richness and composition at both functional (genes involved in the denitrification process) and phylogenetic (overall bacterial community composition based on 16S rRNA profiling) levels. Results further confirm that the effects on microbial populations are minimal and transient in nature. The impact of DCD on microbial community structure was soil dependent, and a greater effect was attributed to intrinsic soil properties like soil texture, with community response to DCD in combination with urine being comparable to that under urine alone. Addition of DCD to cattle urine also reduced N2O emission between 23 and 67%. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Nitrous oxide denitrification Microbial functional genes 16S rRNA Next generation sequencing nitrification inhibitor

1. Introduction New Zealand's pastoral based agricultural system is a cornerstone of its economy. In 2013 a total of 7.84 million hectares of land were dedicated to grasslands (“Statistics New Zealand,” 2013), which support New Zealand's pasture based-livestock industries. It is estimated that approximately 42 million animals (30.79 million sheep, 6.48 million dairy, 3.70 million beef and 1.03 million deer) were supported in 2013 using a year-round grass fed outdoors approach (“Statistics New Zealand,” 2013). As a result, animal excreta in the form of dung and urine compose a major part of New Zealand's N input into soils (de Klein et al., 2003; de Klein and Ledgard, 2005; van der Weerden et al., 2011). Nitrogen deposited in this form is normally in excess of what is assimilated by plants resulting in a superfluous N pool lost via three major mechanisms: volatilization, leaching, and denitrification (Ball et al., 1979; Monaghan et al., 2007).

* Corresponding author. Tel.: þ64 3 479 3140; fax: þ64 3 479 8540. E-mail address: [email protected] (S.E. Morales). http://dx.doi.org/10.1016/j.soilbio.2015.06.009 0038-0717/© 2015 Elsevier Ltd. All rights reserved.

Managing N losses linked to urine deposition is difficult due to the myriad of factors controlling N transformations at farm (Saggar et al., 2013) or larger (Morales et al., 2015) scales. This has resulted in the reliance on simple approaches including the use of nitrification inhibitors (Di and Cameron, 2006). In New Zealand, dicyandiamide (also known as cyanoguanidine, dicyanodiamide or DCD) applied at 10 kg ha
1 had been used as an effective management tool to block nitrification, thus reducing denitrification substrates while limiting N loss through leaching and nitrous oxide (N2O) emission (Di et al., 2007; Zaman et al., 2007, 2009). Prior work suggests that DCD specifically inhibits bacterial ammonia oxidizer's by blocking the conversion of urine derived ammonia to hydroxylamine (Di et al., 2009). This inhibition is thought to be specific and to date no significant impacts on microbial communities have been observed (O'Callaghan et al., 2010; Wakelin et al., 2013). Once applied to soils, temperature dependent degradation of DCD occurs both in situ (Kelliher et al., 2008) and in vitro by common soil microorganisms (Hallinger et al., 1990). This degradation has been confirmed even after sustained application of the inhibitor over a 7 year period (Guo et al., 2013). However, previous methods used to assess the potential impacts of DCD on microbial

S.E. Morales et al. / Soil Biology & Biochemistry 88 (2015) 344e353

community structure have relied on low-resolution tools (e.g. DGGE) or focused solely on specific functional groups (e.g. ammonia oxidizers) (Di et al., 2014). The identification of vulnerable microbial populations is critical to maintaining soil health, and is particularly important given the recent detection of DCD in surface waters (Smith and Schallenberg, 2013). In this study we use soil microcosms, gas chromatography, soil chemical analyses, high throughput 16S rRNA gene amplicon sequencing, terminal restriction fragment length polymorphism (TRFLP) and quantitative PCR (qPCR) on three denitrifier functional genes (nirS, nirK and nosZ) to assess the impact of DCD on target denitrifier communities, as well as the overall microbial community by means of 16S rRNA gene analysis. The objectives were to determine: (1) the impact of urine and DCD application on total bacterial community composition and diversity, (2) the impact of urine and DCD application on specific target populations within the denitrifier guild (by assessing community abundance and diversity), (3) if microbial community response to DCD was influenced by soil type. 2. Materials and methods 2.1. Study site and sample collection Soils were collected from three different farms (Tokomaru [TM] silt loam from Massey University No.4 dairy farm in Palmerston North [40 220 58.2600 S, 175 360 31.0100 E], Manawatu [MW] fine sandy loam from a Longburn dairy farm [40 220 56.9900 S, 175 320 24.4900 E] and an Otorohonga [OH] silt loam from an AgResearch Ruakura dairy farm in Hamilton [38 11019.7000 S, 175120 35.6700 E]) that have been characterized previously and shown to exhibit contrasting denitrification enzyme activities and N2O emissions (Jha unpublished) (Table 1, Fig. S1). At each farm 25 soil cores (25 mm diameter and 100 mm long) were collected from each of the four randomly selected areas (100 m2 each). This resulted in 4 biological replicates per site, with each replicate consisting of 25 composited field cores. During sampling, areas around paddock entrances, water troughs and obvious urine or dung patches were avoided. The collected soil from each randomly selected area resulted in 4 field replicates of the soil on each farm. Soil samples for each replicate were sieved separately to 2 mm and immediately stored in plastic bags at 4  C for chemical analysis. Subsamples from each of the plastic bags were stored at
20  C for molecular analysis. 2.2. Microcosms Four microcosms were established for each farm using four field replicates to address the variability in soil characteristics identified across the grazed farm. Each microcosm consisted of 50 g (dry weight equivalent) subsamples of each soil placed in plastic containers (r ¼ 2.25 cm, h ¼ 7.4 cm, vol ¼ 117.63 cm3) with 1 mm holes (15 in number) on the walls of the container to allow for the acetylene (C2H2) gas to penetrate the soil and N2O to be released from the soil. Treatments were applied by bringing soils to saturation by gradually adding either deionized water in control treatments, urine (700 mg N kg
1 dry soil) or urine (700 mg N kg
1 dry soil) þ DCD (10 mg DCD kg
1 dry soil). In urine and urine þ DCD treatments the same amounts of deionized water, minus the volumes of bovine urine and DCD, were applied to respective containers to increase the soil water contents. The DCD application rate was based on prior work conducted in New Zealand (Singh et al., 2008; Kim et al., 2012; Zaman and Nguyen, 2012). To reduce N losses from grazed pasture 10 kg ha
1 of DCD in 800 L water using a fine particle suspension twice per year in late autumn and late winter is recommended (Di and Cameron,

345

2005). Single application of DCD was decided based on the results of (de Klein et al., 2011) suggesting repeated use of DCD does not impact its effectiveness as compared to single use. Urine used in the experiment was collected fresh from cows during milking (avoiding contamination from dung), and stored in tightly sealed plastic bottles at 4  C to avoid urea hydrolysis. Total C and N content in urine were determined and the amount of urine required for application to achieve desired N application was calculated based on this. Microcosms were set up in the presence and absence of acetylene (C2H2) (10% headspace volume of the glass jars i.e. 100 ml) resulting in 72 total microcosms (3 soils  3 treatments  4 replicates  2 ± C2H2). A separate set of four replicated soil samples (250 g each) for each treatment was prepared and incubated in glass jars for periodic measurements of changes in mineral-N, soil pH and microbial community structure. Once established, all microcosms were incubated at 25  C for 28 days. Gas sampling, chemical and molecular analysis were conducted on days 1, 3, 7, 15 and 28 days of application of treatments. Field moist soil samples were analyzed for soil pH, gravimetric water content, mineral N (NO3
and NH4 þ ), total nitrogen (TN), total carbon (TC), Olsen P, soluble C (K2SO4 extractable C from nonfumigated soils), microbial biomass carbon (MBC), denitrification enzyme activity (DEA), and denitrification rate (DR) before application of treatments. Periodic measurements of soil pH, MBC, soluble C, and mineral N contents were conducted following the application of the treatments during the entire incubation. Soil pH was measured in a 1:2.5 (w/w) soil to water mixture stirred vigorously then left to stand overnight before measurements using a PHM 83 Autocal pH meter (Blakemore, 1987). Soil water content (SWC) was determined gravimetrically by first weighing the wet soil samples, oven drying at 105  C for 24 h and re-weighing the dried soil. Soil NO3
eN and NH4 þ eN were determined by 1 h soil extraction with 2 M KCl solution at soil extract ratio of 1:5, and subsequent analysis of the filtrate colorimetrically using an automatic analyzer method (Downes, 1978). TN and TC were determined by combustion using a LECO CNS-1000 (Bremner, 1996; Nelson and Sommers, 1996). Olsen P was determined in 0.5 M NaHCO3 soil extracts (Olsen et al., 1954), by the phosphomolybdate method (Murphy and Riley, 1962) using a Spectrophotometer PU 8625 UV/VIS at 712 nm. Microbial biomass carbon (MBC) was determined using the chloroform fumigationextraction technique (Vance et al., 1987). Fumigated and nonfumigated soils were extracted with 0.5 M K2SO4 for 30 min (1:5 soil:extractant ratio), filtered and an aliquot was analyzed for organic C by acid-dichromate oxidation (Jenkinson and Powlson, 1976) in which an aliquot of soil extract was added to a mixture of sulphuric acid and orthophosphoric acid and boiled under refluxing condition for 30 min. Excess dichromate was titrated with ferrous ammonium sulfate. The additional oxidisable C obtained from the fumigated soils was taken to represent the microbial-C flush and converted to microbial-biomass C using the relationship: microbial CaC flush/0.41. Denitrification enzyme activity (DEA) was determined using the slightly modified method described by Luo et al. (2010). Soil samples (10 g equivalent dry weight) were placed in 125 ml flasks. Slurries were prepared by adding 25 ml of a solution containing 2.2 mg NO3
(35 mmoles NO3
) as KNO3
, 2.5 mg C (208 mmol C) as D-Glucose and 250 mg chloramphenicol. The flasks were sealed using Suba-Seal® septa (SigmaeAldrich) with air flushed from the flasks using N2 gas to create anaerobic conditions. Ten percent of the headspace volume (approx. 10 ml) of the flasks was replaced with purified (acetone-free) acetylene (C2H2). Gas samples (5 ml) for time 0 (T0) were taken immediately and replaced with an equal quantity of N2. The flasks were then placed on an orbital shaker (set at 125 rpm) and incubated at 25  C for 6 h. A 5 ml gas sample was

346

S.E. Morales et al. / Soil Biology & Biochemistry 88 (2015) 344e353

taken after 2, 4, and 6 h of incubation from each flask, and each time the same amount of N2 gas was replaced in the flasks. 2.3. Total denitrification (N2O þ N2) and N2O emission Total denitrification and N2O emission rates were measured using the acetylene inhibition method described by Tiedje et al. (1989). In brief, collected soil samples were incubated in duplicate 1 L glass jars. Half of the jars were incubated without C2H2 and the other half with 10% headspace inside the jars replaced with C2H2. All jars were incubated at 25  C for 24 h. Gas samples were taken at time 0, 3, 6, 9, 12 and 24 h with the gas sample replaced each time by the same volume of air to keep the atmospheric pressure inside the jars constant. Gas samples were analyzed for N2O in a gas chromatograph (GC). The N2O emitted in non-C2H2 treated jars corresponded to N2O produced during both nitrification and denitrification in incubated soils. The N2O emitted from C2H2 treated jars corresponded to total denitrification since C2H2 inhibits conversion of N2O to N2 and it also inhibits nitrification making N2O as the sole product of denitrification. 2.4. DNA extraction, qPCR and T-RFLP analyses DNA was extracted from 0.25 g of each individual sample using a MoBio PowerSoil™ DNA Isolation Kit (MoBio, Solana Beach, CA) following the manufacturer's instructions. DNA quality and quantity was assessed using a Nanodrop Spectrophotometer (Thermo Fisher). No further DNA purification was performed, and all samples had a 260/280 ratio of 1.8e2.0. No inhibition of PCR reactions was observed when using working concentration diluted DNA. Real-time quantitative PCR (qPCR) was used to quantify bacterial nirS, nirK, nosZ, and rpoB genes as described previously (Deslippe et al., 2014). Primers used for q PCR were nirS Cd3aF, R3cd (Enwall et al., 2010), nirK Copper 583F, 909R (Dandie et al., 2011), nosZ 2F, 2R (Henry et al., 2006) and rpoB 1698 F (50 AACATCGGTTTGATCAAC-30 ), 2041 R (50 -CGTTGCATGTTGGTACCCAT-30 ) (Dahllof et al., 2000). All reactions were carried out in a LightCycler® 480 System (Roche) with SsoFast EvaGreen® SuperMix (Biorad). Reactions were carried out using 5 ng genomic DNA, 300 nM each primer and Ssofast supermix (Biorad) in a final volume of 10 ml. Each qPCR plate included relevant known template standards made from cloned PCR products. Standard curves were created from serial dilutions of linearized, insert-containing plasmids. Standards were generated by amplifying extracted DNA from a target soil and subsequent cloning of the purified PCR product. The cloned PCR products were sequenced and correct identity of the sequences were confirmed through BLAST. Product specificity and contamination for each plate were assessed using melt curves and negative (no DNA) controls for each primer respectively. Melt curve for each qPCR product was calculated at the end of each run using a continuous thermal gradient of 65e95  C. The cut off Ct value for negative controls for the nir and nosZ genes was set to 35. Terminal restriction fragment length polymorphism (T-RFLP) analysis was carried out as described in Deslippe et al. (2014). Primers for amplification of nirS, nirK and nosZ are the same as for qPCR except that fluorophore (carboxyfluorescein [FAM]) labeled reverse primers were used. PCR reactions were carried out using 20 ng genomic DNA, 1 mmol of each primer and 2X Thermo-Start PCR Master Mix (Thermo Scientific) in a final volume of 30 ml. All reactions were carried out using a Maxy Gene Gradient THERM1000 (Axygen) thermocycler. PCR products were digested using HhaI restriction enzyme (New England Biolabs), cleaned and analyzed using an ABI3730 Genetic Analyzer (Applied Biosystems). The threshold-normalized T-RFLP data was used in the calculations of gene richness and evenness.

2.5. 16S rRNA gene amplicon sequencing and analysis 16S rRNA gene amplification and amplicon sequencing on the Illumina MiSeq platform were done following the Earth Microbiome Project standard protocol (Caporaso et al., 2012) for each individual site replicate. Sequences were quality filtered using Qiime (1.7.0) default parameters (Caporaso et al., 2010). Open reference Operational Taxonomic Unit (OTU) picking was performed using the Greengenes reference library (McDonald et al., 2012) in order to bin sequences into clusters based on 97% sequence similarity. Multiple rarefactions (ten total) were carried out to rarify OTU tables (i.e. to randomly re-sample the pool of sequences) to an even depth of 22,700 sequences per sample. Samples that did not meet quality filtering or for which sequencing depth was too low were discarded (Table S1). Resulting OTU tables were used to calculate alpha diversity, beta diversity and for classification of OTUs using Qiime. Relative abundance was calculated from a merged (average) OTU table from the prior step. 2.6. Statistical analysis Analyses were performed in JMP (SAS Institute, Cary, NC, USA). Statistical significance for measured variables was calculated using ANOVA. Principal components analysis (PCA) consisted of datasets organized as data matrices (untransformed) composed of relative sequence abundance for OTUs clustered at 97% sequence similarity and classified to either the phylum level or to the lowest level of classification possible. Data was organized with rows representing different samples, and columns representing either relative abundance of phyla or non-redundant phylogenetic groups (e.g. Pseudomonas). PCA was chosen in order to keep the method of analyses between variables (chemical, microbial and molecular data) consistent. Patterns observed while analyzing community composition via PCA were corroborated via alternative pathways (PCoA, DCA and network analysis). To identify variables driving clustering of samples principal components were correlated to individual taxa and significantly (p < 0.05) correlated ones retained for further analysis. Identified taxa were clustered using hierarchical clustering (Mean distance) and compared to sample clustering based on predicted drivers of variance. Dominant OTUs were identified using Qiime's ‘make_3d_plots.py’ script for generating biplots. For statistical comparison of beta diversity Unifrac distances were calculated and significance was determined using ANOVA. Further processing and analysis was done using R and the phyloseq package (McMurdie and Holmes, 2013). 3. Results 3.1. Physicochemical response Application of urine with and without DCD transiently generated more alkaline conditions and increased MBC. Soluble C and mineral N contents in samples were also elevated, but sustained, when compared to the control (only water) and unamended (field moist soils) (Fig. 1 and Table S2). The effect of applied DCD was time dependent but evident from the significant (P < 0.05) differences in the NH4 þ eN contents of urine only and urine þ DCD applied treatments in the incubated soils (ANOVA, F ¼ 9.85). The NO3
eN content in the urine only treatment was significantly (P < 0.05) higher than the same in the DCD applied treatment (ANOVA, F ¼ 17.74). The differences in the mineral-N contents among the two urine treatments in the soils decreased with incubation time, until no difference was observed. Also the interaction of soil type and time of sampling was significant (NH4 þ eN, F ¼ 3.45; NO3
eN, F ¼ 38.86) suggesting differences in the mineral N contents in the

S.E. Morales et al. / Soil Biology & Biochemistry 88 (2015) 344e353

347

Fig. 1. Mean ± SEM of soil chemical properties for three dairy soils (MW ¼ Manawatu fine sandy loam, TM ¼ Tokomaru silt loam, and OH ¼ Otorohonga silt loam) in response to control (water), urine, and urine þ DCD application. MBC ¼ microbial biomass carbon.

two urine treatments varied in soil with time. The differences in the NH4 þ eN and NO3
eN contents among the urine treatments were observed on day 3 in the OH soil, but in the MW and the TM soils this difference was stronger at later times. The differences in NO3
eN contents in the two urine treatments were observed on day 15 in the TM and the MW soil. An increase in N2O and N2O þ N2 emissions was observed under urine conditions. The addition of DCD to urine reduced N2O emission from these soils by 23, 54 & 67% in the OH, MW, and the TM soil respectively. As compared to urine only treatments this decrease in N2O emission was significant only in the TM soil. Further, overall denitrification was significantly reduced under DCD conditions in the MW and the TM soil but not in the OH soil. The pattern of total denitrification was variable among the three soils. There was higher denitrification in the MW soil immediately after urine application than the TM and the OH soils, in which the increase in denitrification was gradual.

3.2. Functional community response Denitrifier community response as determined by qPCR of denitrification genes was initially driven by SWC, but a treatment effect was observed within urine treated samples (Fig. 2). There was a significant increase in SWC from field moist to saturation on nir (F ¼ 3.87, p < 0.005) and nosZ (F ¼ 3.39, p < 0.005) genes. An increase in SWC was correlated to increases in both nir and nosZ gene copies with a decrease, and return to original levels, in day 3. This effect was most pronounced within MW samples, with a smaller

effect on the other two sites. No consistent or significant effect of DCD application was observed. The response pattern of the denitrifier genes to the applied treatments was similar in the three soils except the MW soil that clearly shows lower gene abundance in the control sample versus the urine application (Fig. 2 and Table S3). There were little differences observed between the control and urine application in the TM and OH soils. Community structure as assessed by T-RFLP was consistent across all treatments, though variations over time were observed. The total number of both nir and nos gene T-RFs was originally 1.5e2 fold higher in the two nonallophanic soils (MW & TM) that are also rich in MBC, Olsen P and DEA than the allophanic soil OH. Although lower in number, the denitrifier phylotypes (T-RFs) in the allophanic soil were more stable to chemical changes (after addition of urine and urine þ DCD) than the other two non-allophanic soils. 3.3. Total community response Of the original 156 samples, a total of 149 samples were analyzed after removal of low quality reads and low coverage samples (samples where the total number of reads was lower than the chosen level of rarefication) (see Table S1). All remaining samples were analyzed at a sequencing depth of 22,700 sequences per sample. 3.3.1. Richness A significant decrease in microbial richness (assessed as observed OTUs97% based on 16S rRNA gene) and diversity was

348

S.E. Morales et al. / Soil Biology & Biochemistry 88 (2015) 344e353

Fig. 2. Mean ± SEM of microbial functional response for three dairy soils in response to control (water), urine, and urine þ DCD application. nosZ ¼ nitrous oxide reductase gene; nirS ¼ cytochrome cd1-type nitrite reductase gene; nirK ¼ copper-containing nitrite reductase; T-RFs ¼ terminal restriction fragments.

observed over all treatments concurrent with an increase in SWC (data not shown) when compared to field moist samples (Fig. 3 and Fig. S1). A treatment specific further reduction was observed for TM soils, where urine and urine þ DCD treated samples displayed a lower richness at all time points. Richness increased over time with increase in SWC from field moist to saturation with the rate of recovery being soil and treatment dependent. The only soil that exhibited a treatment effect (Tokomaru [TM]) did not recover fully, with both urine and urine þ DCD retaining a lower level of richness than in field moist samples.

3.3.2. Phylum and genus level response Microbial community changes at higher taxonomic levels (phylum) were observed primarily across different soils (Fig. 4 and Table S2). Multivariate analysis of 16S data showed that changes were linked to soil type rather than treatment with PC1 clustering sites based on soil texture. Clustering of samples based on phylum level community data indicated significant differences in microbial community structure present within all Tokomaru samples, with more variance in community structure (within PC2) across samples compared to other soils. This clustering of samples corresponding

Fig. 3. Mean ± SEM of observed operational taxonomic units (OTUs) in response to control (water), urine, and urine þ DCD application in three dairy soils. OTUs were determined based on 97% sequence similarity at the 16S rRNA gene.

S.E. Morales et al. / Soil Biology & Biochemistry 88 (2015) 344e353

349

Fig. S3-7). To further explore site-specific trends, each soil was analyzed separately. One soil (TM) did show a clear treatment effect, with MW displaying a weaker trend. However, no differences were observed in community composition between urine and urine þ DCD treated samples. No time effects were observed. It was noted that analyses at phylum or genus level (and as individual soils or all samples) resulted in very low principal component scores, indicating that very little of the variability within the data could be captured within any component. To further test community changes Beta diversity (as measured using unweighted Unifrac distance) was utilized to compare changes over time. All samples were compared to Day 1 samples within the control treatment (Fig. 6 and Tables S4e5). Small changes over time were observed for all samples. Response to DCD þ urine application was consistent across all soils, while response to both control and urine alone was variable both across time and soil type. All observed changes were small, and comparable in scale to changes within control samples over the time of the experiment. To identify dominant OTUs within samples, all OTUs identified more than 100 times in at least 25% of the samples were subsampled from the total dataset. Representatives of the Acidobacteria Group 5, Alphaproteobacteria, Bacilli, Betaproteobacteria, Deltaproteobacteria, Gammaproteobacteria, Spartobacteria and Sphingobacteria were identified. Bacilli, Alphaproteobacteria and Spartobacteria were consistently the most abundant taxa in all soils sampled independent of treatment (Figs. S6e7 and Table S6). Classification to genera was not possible for all identified OTUs, but within those with strong similarity to reference sequences, OTUs associated with the Acidobacteria Group 5, Geobacter and Spartobacteria were depleted within the well-drained silt loam (OH) samples.

Fig. 4. Community response to soil  treatment at phylum level. A) Principal component analysis representing the effects of soil origin and treatment on phylum level representation. The percentage of the variation in the samples described by the plotted principle components is indicated on the axes. B) Heat map (constructed using a two-way clustering analysis) comparing relative abundance of phyla contributing to the clustering of samples in (A). Color gradient indicates relative abundance with intensity of red representing high values relative to the sample mean and blue relatively low values. Samples are color coded on the left based on soil origin, treatment, drainage and texture class.

to individual sites was linked to an enrichment of OTUs assigned to the Armatimonadetes, Spirochaetes, Nitrospira, Other (unclassifieds), and WS3 for OH and MW samples, and the Chloroflexi, Cyanobacteria and Firmicutes within the TM samples. This community selection was mostly linked to texture class (Fig. 4b). To look at fine scale changes community composition was analyzed using genus rather than phylum level assignments. As seen at higher taxonomic levels, total community response at genus level was strongly linked to soil type and not treatment (Fig. 5 and

3.3.3. Response of target groups Key genera relying on nitrification pathway products were examined for treatment effects. No sustained pattern of enrichment or inhibition was observed across all soils in response to treatments (Fig. 7), although patterns were observed within certain time points. OTUs associated to the genera Pedobacter, Pseudomonas and Bacillus were consistently enriched under both urine and urine þ DCD conditions in all soils used within the 28 days in this study. Other taxa displayed inconsistent responses across different soils. The genus Geobacter showed an initial depletion in response to urine and urine þ DCD, but recovery was observed over time for two soils (MW and TM). The same genus was absent or in low abundance in all samples for OH. Key nitrifying taxa displayed no change in relative abundance based on treatments (Nitrobacter) or an increase in abundance under urine and urine þ DCD (Nitrosospira and Aminobacter). Potential denitrifiers (Bradyrhizobium, Paracoccus, Rhodococcus, Denitratisoma, Rheinheimera) displayed no clear difference between treatments. Other organisms involved in alterative nitrogen utilizing pathways like dissimilatory nitrate reduction to ammonium (DNRA) and nitrous oxide reduction (Anaeromyxobacter) or anaerobic ammonia oxidation (anammox) (Planctomyces) displayed variable and soil dependent responses. 4. Discussion Developing management strategies and tools for reducing greenhouse gas emissions and minimizing N loss from soils is crucial if countries are to reduce their agricultural impacts. Adopting mitigation options at large scale by farmers requires simple and effective tools, as has been the case for nitrification inhibitors. The mitigation of agricultural N loss by the implementation of nitrification inhibitors has been widely adopted

350

S.E. Morales et al. / Soil Biology & Biochemistry 88 (2015) 344e353

Fig. 5. Community response to soil  treatment  time based on 16S rRNA gene OTU97% relative abundance. Top) Principal component analysis representing the effects of soil origin and treatment on all sites [All sites] or individual soils [MW ¼ Manawatu fine sandy loam, OH ¼ Otorohanga silt loam, TM ¼ Tokomaru silt loam]. For individual soil plots, sampling date is labeled. Color indicates treatment (Control ¼ red; Urine ¼ green; Urine þ DCD ¼ blue) and shape origin of soils (^ ¼ MW; C ¼ OH; △ ¼ TM). The percentage in variance accounted for in each plotted principle component is indicated on the axes. Bottom) Heat map (constructed using a two-way clustering analysis) comparing relative abundance of taxa (columns) contributing to the clustering of all samples (rows) in the PCA analysis. Color gradient indicates relative abundance with intensity of red representing high values relative to the sample mean and blue relatively low values. Samples are color coded on the right based on day of sampling, soil origin, treatment, drainage and texture class.

though the impact of these inhibitors on target and non-target organisms is not well understood. This is crucial to the understanding of any hidden ecological costs that may decrease longterm soil productivity. Our results support prior findings, and suggest that DCD impacts are minimal, transient in nature and comparable to community disturbances triggered by urine. It also confirms that DCD is pathway specific, having a major impact on nitrification as evident from differences in mineral-N contents in DCD treated urine samples than urine only samples especially in the OH and the TM soils. The differences in the mineral N contents in the two urine treatments were short lived and variable among the three soils. DCD effect was not clear in this study, and the

Fig. 6. Changes in Beta-diversity over time (shown as unweighted UniFrac distances). Plot represents the mean (±SEM) Unifrac distance when comparing samples to Day 01 under control conditions. Results for ANOVA and Tukey HSD test in Tables S4e5.

differences in the pH and mineral N contents of the soil could be due to differential urea hydrolysis in the soils. DCD effectiveness could have been influenced by limited nitrification and increased denitrification occurring in the saturated soils incubated in this study. Studies in New Zealand (incubations under unsaturated conditions) have found significant differences in NH4 þ eN and NO3
eN contents for relatively longer time with application of DCD compared to urine only treatment (Singh et al., 2008; Zaman and Nguyen, 2012). Since no difference was observed in the community composition between two urine treatments, our data suggests that microorganisms are likely to have alternative pathways for energy generation and survival allowing them to persist in the presence of inhibitors. Since the persistence of DNA in soils is possible after microbial death, the presence of alternative pathways (while likely) is not the only possible explanation. However, to date, DCD has escaped the rise of resistant strains that commonly happen when inhibitors are utilized. This is only possible if A) the inhibitor results in the death of ALL targeted organisms, or B) the organisms are not truly being affected, rather the pathway is, and targeted organisms can utilize alternative pathways to continue their metabolic processes supporting the former interpretation. Our results support prior findings of reduction in N2O emission with DCD application (Singh et al., 2008; Luo et al., 2010; Zaman and Nguyen, 2012). However, our data demonstrates that reduction in N2O emissions is soil specific, with a clearly observed phenotype in the poorly drained TM soil when compared to the other two (MW and OH) soils tested (Fig. 1). These soils differ in their origin, geographical location, climatic condition and management practices performed on farms, as well as their capacities to denitrify probably due to their variations (Morales et al., 2015). To examine the hidden ecological costs associated to DCD use we examined its effect on: 1) the overall microbial community by means of 16S rRNA gene analysis, and 2) specific target populations

S.E. Morales et al. / Soil Biology & Biochemistry 88 (2015) 344e353

Fig. 7. Specific response (change in relative abundance) of representative taxa linked to nitrogen transformation (nitrification, denitrification, dissimilatory nitrate reduction to ammonium and/or nitrous oxide reduction).

within the denitrifier guild (by assessing community abundance and diversity) across three soils representing two texture classes. Total community analysis based on 16S rRNA gene amplicon sequencing corroborated prior findings suggesting little impact on total microbial community based on DCD exposure. Data supports prior conclusions that urine deposition is having a greater impact on community structure than exposure to DCD (Di et al., 2014; O'Callaghan et al., 2010; Uchida et al., 2011; Wakelin et al., 2013). A modest, but significant, decrease in richness was observed here with both the Urine only and DCD þ Urine treatments suggesting that urine is the major driver in community response, a trend previously reported using other approaches (O'Callaghan et al., 2010; Sanford et al., 2012; Wakelin et al., 2013). It further suggested that community changes across sites were primarily driven by soil texture class, rather than any given treatment (Table S6). However, limited replication (three soils and two texture classes) reduces our ability to confirm this. Changes in community structure were observed using both whole community Beta diversity metrics (Unifrac, Fig. 6) and data for individual phylogenetic groups

351

(Figs. 4e5 and 7). This data suggests that community response is site specific and prone to be affected by temporal scale as the community changed over time in all treatments and soils tested. Across all soils (within all treatments and time points), dominant taxa (at phylum level) were conserved and OTUs classified to the lowest taxonomic rank associated with commonly identified groups in soils including Acidobacteria Group 1, Acidobacteria Group 6 and Spartobacteria (Jones et al., 2009; Vogel et al., 2009; Bergmann et al., 2011). In addition, novel lineages (classified to the lowest rank) within the Bacteroidetes, Bacillales, Chitinophagaceae, Betaproteobacteria, Rhizobiales, Proteobacteria, and Bacteria compromised some of the most abundant OTUs within all sites. Clustering of sites, associated with changes in texture class, was driven by changes in certain lineages (Fig. 4) including novel OTUs from the newly formed phylum Armatimonadetes. Some of these lineages have been suggested to belong to a superphylum of terrestrial bacteria termed ‘Terrabacteria’, including the Actinobacteria, Cyanobacteria, Thermi (Deinococcus-Thermus), Chloroflexi and Firmicutes (Battistuzzi and Hedges, 2009; Rinke et al., 2013), indicates that texture might be a strong variable selecting for community structuring within these groups. OTUs with close similarity to phylogenetically defined organisms, were amongst the most dominant, and in some cases (Acidobacteria Group 5, Geobacter and Spartobacteria) demonstrated soil specific patterns (Figs. 6e7 and Table S6). These three groups were depleted within the well-drained silt loam (OH) samples. Geobacter are known for their metabolic flexibility and are normally associated with anaerobic environments that enable them to exploit such traits (Mahadevan et al., 2011), suggesting that oxygen status as mediated through drainage could negatively impact them. The other two groups (Spartobacteria and Acidobacteria Gp5) are known to be dominant members of soils but the information to date does not allow us to speculate as to the reason for their reduction in certain soils (Bergmann et al., 2011; Naether et al., 2012). The absence of a strong DCD response is not unwarranted given its mode of action. This inhibitor is known to be bacteriostatic, and capable of only an incomplete and reversible inhibition in pure cultures of Nitrosomonas under high levels (Zacherl and Amberger, 1990). A closer look at changes over time for specific microbial lineages involved in nitrogen cycling shows that even for targeted groups, response can be inconsistent and contrary to what is expected. Some nitrifiers were not reduced in abundance under DCD, or increased under urine. This can be a result of the fact that as analyzed these taxonomic groups and their relative abundance are determined by combining the response of all OTUs classified to that lineage. With this method if some OTUs respond but others do not, the mean response would be averaged across all OTUs, reducing the ability to detect changes. If this is happening to the nitrifying community it also suggests an alternative interpretation. Nitrification has been thought to be restricted to a small group of monophyletic organisms where the process is used for growth (autotrophic nitrification) or balancing redox within cells (heterodard and Knowles, 1989; De Boer and trophic nitrification) (Be Kowalchuk, 2001). Due to their slow growth rates and difficulty in culturing it was assumed that nitrifiers must have narrow metabolic or energy acquiring capabilities. The clear reduction in AOB numbers and inhibition of nitrification concomitant with reduction in N2O emissions further supported this (Di et al., 2009, 2010, 2014; O'Callaghan et al., 2010; Wakelin et al., 2013). However, our results suggest that not all organisms within known nitrifier lineages are inhibited, with some increasing in abundance under urine conditions. This observation could be explained by the presence of alternative pathways of energy generation, and hidden diversity of organisms carrying out the processes. Recent studies have demonstrated that alternative energy sources are used by

352

S.E. Morales et al. / Soil Biology & Biochemistry 88 (2015) 344e353

nitrifiers (Koch et al., 2014) which might also explain why resistance to nitrification inhibitors has not been observed to date. Alternatively this could be the result of classification errors due to misclassification of sequences based on: 1) lack of a better match, 2) clustering errors within the 3% OTU threshold, 3) strain level differences where functional traits are not conserved. One additional source of potential error is the use of real urine instead of a chemically defined artificial (synthetic) equivalent. One argument would suggest that a complex and variable mix of chemicals found in real urine can limit interpretation of results. However, prior work (Kool et al., 2006) has shown that no artificial urine can accurately mimic real urine. In this study we followed this guideline, and the decision was further justified due to the nature of the field sites. All soils used in this study were collected from established dairy grazed pastures receiving nitrogen deposition in the form of urine for the last 10e15 years. This study clearly showed that soil type and urine deposition are bigger drivers of microbial community structure than DCD. Typical response of nitrification and the response of functional groups linked to denitrification after DCD application were soil specific. At a total community level, urine deposition resulted in a decrease in richness both in the presence and absence of DCD, confirming that the inhibitor does not contribute to reducing richness further. Data did not support recent work where nirK gene abundance was reduced in the presence of DCD (Di et al., 2014), and further confirms that soil response might differ from location to location. However, the chemical response across multiple soils (and independent studies) taken together with the lack of impact on microbial community structure further supports the notion that the use of DCD as a tool for on farm N management is viable and ecologically safe option. One limitation of the study is that it does not account for changes in fungal groups that are known to be important for soil nutrient cycling so we do not know the impact DCD would have on this soil population. In addition, impacts beyond microbial population have not been accounted for. Although persistent negative effects were not observed for microbes, this might not be the case when larger organisms including invertebrates are included. Acknowledgements We thank Yi Wang for providing support with preparing samples for sequencing. We also thank Thilak Palmada from Landcare Research Palmerston North and Mike Sprosen from AgResearch Hamilton, for their assistance in collection of soil samples from Manawatu and Waikato. This research was funded and supported by a University of Otago Research Grant, as well as by the New Zealand Agricultural Greenhouse Gas Research Centre and Landcare Research Ltd. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.soilbio.2015.06.009. References Ball, R., Keeney, D.R., Thoebald, P.W., Nes, P., 1979. Nitrogen balance in urineaffected areas of a New Zealand pasture. Agronomy Journal 71, 309e314. http://dx.doi.org/10.2134/agronj1979.00021962007100020022x. Battistuzzi, F.U., Hedges, S.B., 2009. A major clade of prokaryotes with ancient adaptations to life on land. Molecular Biology and Evolution 26, 335e343. http:// dx.doi.org/10.1093/molbev/msn247. dard, C., Knowles, R., 1989. Physiology, biochemistry, and specific inhibitors of Be CH4, NHþ 4 , and co oxidation by methanotrophs and nitrifiers. Microbiological Reviews 53, 68e84.

Bergmann, G.T., Bates, S.T., Eilers, K.G., Lauber, C.L., Caporaso, J.G., Walters, W.A., Knight, R., Fierer, N., 2011. The under-recognized dominance of Verrucomicrobia in soil bacterial communities. Soil Biology and Biochemistry 43, 1450e1455. http://dx.doi.org/10.1016/j.soilbio.2011.03.012. Blakemore, L.C., 1987. Methods for chemical analysis of soils. NZ Soil Bureau Scientific Report 80, 71e76. Bremner, J.M., 1996. Nitrogen e total. In: Methods of Soil Analysis. Part 3-Chemical Methods. Soil Science Society of America Inc., pp. 1085e1121. Caporaso, J.G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F.D., Costello, E.K., ~ a, A.G., Goodrich, J.K., Gordon, J.I., Huttley, G.A., Kelley, S.T., Fierer, N., Pen Knights, D., Koenig, J.E., Ley, R.E., Lozupone, C.A., McDonald, D., Muegge, B.D., Pirrung, M., Reeder, J., Sevinsky, J.R., Turnbaugh, P.J., Walters, W.A., Widmann, J., Yatsunenko, T., Zaneveld, J., Knight, R., 2010. QIIME allows analysis of highthroughput community sequencing data. Nature Methods 7, 335e336. http:// dx.doi.org/10.1038/nmeth0510-335. Caporaso, J.G., Lauber, C.L., Walters, W.A., Berg-Lyons, D., Huntley, J., Fierer, N., Owens, S.M., Betley, J., Fraser, L., Bauer, M., Gormley, N., Gilbert, J.A., Smith, G., Knight, R., 2012. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. The ISME Journal 6, 1621e1624. http:// dx.doi.org/10.1038/ismej.2012.8. Dahllof, I., Baillie, H., Kjelleberg, S., 2000. rpoB-based microbial community analysis avoids limitations inherent in 16S rRNA gene intraspecies heterogeneity. Applied and Environmental Microbiology 66, 3376e3380. Dandie, C.E., Wertz, S., Leclair, C.L., Goyer, C., Burton, D.L., Patten, C.L., Zebarth, B.J., Trevors, J.T., 2011. Abundance, diversity and functional gene expression of denitrifier communities in adjacent riparian and agricultural zones. FEMS Microbiology Ecology 77, 69e82. http://dx.doi.org/10.1111/j.15746941.2011.01084.x. De Boer, W., Kowalchuk, G.A., 2001. Nitrification in acid soils: micro-organisms and mechanisms. Soil Biology and Biochemistry 33, 853e866. de Klein, C.A.M., Ledgard, S.F., 2005. Nitrous oxide emissions from New Zealand agriculture e key sources and mitigation strategies. Nutrient Cycling in Agroecosystems 72, 77e85. http://dx.doi.org/10.1007/s10705-004-7357-z. de Klein, C.A.M., Barton, L., Sherlock, R.R., Li, Z., Littlejohn, R.P., 2003. Estimating a nitrous oxide emission factor for animal urine from some New Zealand pastoral soils. Australian Journal of Soil Research 41, 381e399. http://dx.doi.org/10.1071/ SR02128. de Klein, C., Cameron, K.C., Di, H.J., Rys, G., Monaghan, R.M., Sherlock, R.R., 2011. Repeated annual use of the nitrification inhibitor dicyandiamide (DCD) does not alter its effectiveness in reducing N2O emissions from cow urine. Animal Feed Science and Technology 166, 480e491. http://dx.doi.org/10.1016/ j.anifeedsci.2011.04.076. Deslippe, J.R., Jamali, H., Jha, N., Saggar, S., 2014. Denitrifier community size, structure and activity along a gradient of pasture to riparian soils. Soil Biology and Biochemistry 71, 48e60. Di, H.J., Cameron, K.C., 2006. The use of a nitrification inhibitor, dicyandiamide (DCD), to decrease nitrate leaching and nitrous oxide emissions in a simulated grazed and irrigated grassland. Soil Use and Management 18, 395e403. Di, H.J., Cameron, K.C., Sherlock, R.R., 2007. Comparison of the effectiveness of a nitrification inhibitor, dicyandiamide, in reducing nitrous oxide emissions in four different soils under different climatic and management conditions. Soil Use and Management 23, 1e9. Di, H.J., Cameron, K.C., Shen, J.P., Winefield, C.S., O'Callaghan, M., Bowatte, S., He, J.Z., 2009. Nitrification driven by bacteria and not archaea in nitrogen-rich grassland soils. Nature Geoscience 2, 621e624. http://dx.doi.org/10.1038/ ngeo613. Di, H.J., Cameron, K.C., Shen, J.-P., Winefield, C.S., O'Callaghan, M., Bowatte, S., He, J.Z., 2010. Ammonia-oxidizing bacteria and archaea grow under contrasting soil nitrogen conditions. FEMS Microbiology Ecology 72, 386e394. http:// dx.doi.org/10.1111/j.1574-6941.2010.00861.x. Di, H.J., Cameron, K.C., Podolyan, A., Robinson, A., 2014. Effect of soil moisture status and a nitrification inhibitor, dicyandiamide, on ammonia oxidizer and denitrifier growth and nitrous oxide emissions in a grassland soil. Soil Biology and Biochemistry 73, 59e68. http://dx.doi.org/10.1016/j.soilbio.2014.02.011. Downes, M.T., 1978. An improved hydrazine reduction method for the automated determination of low nitrate levels in freshwater. Water Research 12, 673e675. €ck, I.N., Stenberg, M., Soderstrom, M., Hallin, S., 2010. Soil reEnwall, K., Throba sources influence spatial patterns of denitrifying communities at scales compatible with land management. Applied and Environmental Microbiology 76, 2243e2250. http://dx.doi.org/10.1128/AEM.02197-09. Guo, Y.J., Di, H.J., Cameron, K.C., Li, B., Podolyan, A., Moir, J.L., Monaghan, R.M., Smith, L.C., O'Callaghan, M., Bowatte, S., Waugh, D., He, J.-Z., 2013. Effect of 7year application of a nitrification inhibitor, dicyandiamide (DCD), on soil microbial biomass, protease and deaminase activities, and the abundance of bacteria and archaea in pasture soils. Journal of Soils Sediments 13, 753e759. http://dx.doi.org/10.1007/s11368-012-0646-2. € fer, P.R., Goldbach, H., 1990. Several aspects of bacterial Hallinger, S., Wallno dicyandiamide degradation. Naturwissenschaften 77, 332e334. Henry, S., Bru, D., Stres, B., Hallet, S., Philippot, L., 2006. Quantitative detection of the nosZ gene, encoding nitrous oxide reductase, and comparison of the abundances of 16S rRNA, narG, nirK, and nosZ genes in soils. Applied and Environmental Microbiology 72, 5181e5189. http://dx.doi.org/10.1128/AEM.00231-06. Jenkinson, D.S., Powlson, D.S., 1976. The effects of biocidal treatments on metabolism in soildI. Fumigation with chloroform. Soil Biology and Biochemistry 8, 167e177.

S.E. Morales et al. / Soil Biology & Biochemistry 88 (2015) 344e353 Jones, R.T., Robeson, M.S., Lauber, C.L., Hamady, M., Knight, R., Fierer, N., 2009. A comprehensive survey of soil acidobacterial diversity using pyrosequencing and clone library analyses. The ISME Journal 3, 442e453. http://dx.doi.org/ 10.1038/ismej.2008.127. Kelliher, F.M., Clough, T.J., Clark, H., Rys, G., Sedcole, J.R., 2008. The temperature dependence of dicyandiamide (DCD) degradation in soils: a data synthesis. Soil Biology and Biochemistry 40, 1878e1882. Kim, D.-G., Giltrap, D., Saggar, S., Palmada, T., Berben, P., Drysdale, D., 2012. Fate of the nitrification inhibitor dicyandiamide (DCD) sprayed on a grazed pasture: effect of rate and time of application. Soil Research 50, 337. http://dx.doi.org/ 10.1071/SR12069. Koch, H., Galushko, A., Albertsen, M., Schintlmeister, A., Gruber-Dorninger, C., Lucker, S., Pelletier, E., Le Paslier, D., Spieck, E., Richter, A., Nielsen, P.H., Wagner, M., Daims, H., 2014. Growth of nitrite-oxidizing bacteria by aerobic hydrogen oxidation. Science 345, 1052e1054. http://dx.doi.org/10.1126/ science.1256985. Luo, J., De Klein, C., Shepherd, M., Ledgard, S., 2010. Effect of nitrification inhibitor on nitrous oxide emissions in pasture soils. In: 19th World Congress of Soil Science: Soil Solutions for a Changing World. International Union of Soil Sciences, pp. 1e6. Mahadevan, R., Palsson, B.Ø., Lovley, D.R., 2011. In situ to in silico and back: elucidating the physiology and ecology of Geobacter spp. using genome-scale modelling. Nature Publishing Group 9, 39e50. http://dx.doi.org/10.1038/ nrmicro2456. McDonald, D., Price, M.N., Goodrich, J., Nawrocki, E.P., DeSantis, T.Z., Probst, A., Andersen, G.L., Knight, R., Hugenholtz, P., 2012. An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea. The ISME Journal 6, 610e618. http://dx.doi.org/10.1038/ ismej.2011.139. McMurdie, P.J., Holmes, S., 2013. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS ONE 8, e61217. http:// dx.doi.org/10.1371/journal.pone.0061217. Monaghan, R.M., Hedley, M.J., Di, H.J., McDowell, R.W., Cameron, K.C., Ledgard, S.F., 2007. Nutrient management in New Zealand pasturesdrecent developments and future issues. New Zealand Journal of Agricultural Research 50, 181e201. Morales, S.E., Jha, N., Saggar, S., 2015. Biogeography and biophysicochemical traits link N2O emissions, N2O emission potential and microbial communities across New Zealand pasture soils. Soil Biology and Biochemistry 82, 87e98. http:// dx.doi.org/10.1016/j.soilbio.2014.12.018. Murphy, J., Riley, J.P., 1962. A modified single solution method for the determination of phosphate in natural waters. Analytica Chimica Acta 27, 31e36. http:// dx.doi.org/10.1016/S0003-2670(00)88444-5. Naether, A., Foesel, B.U., Naegele, V., Wust, P.K., Weinert, J., Bonkowski, M., Alt, F., Oelmann, Y., Polle, A., Lohaus, G., Gockel, S., Hemp, A., Kalko, E.K.V., Linsenmair, K.E., Pfeiffer, S., Renner, S., Schoning, I., Weisser, W.W., Wells, K., Fischer, M., Overmann, J., Friedrich, M.W., 2012. Environmental factors affect acidobacterial communities below the subgroup level in grassland and forest soils. Applied and Environmental Microbiology 78, 7398e7406. http:// dx.doi.org/10.1128/AEM.01325-12. Nelson, D.W., Sommers, L.E., 1996. Total carbon, organic carbon, and organic matter. In: Methods of Soil Analysis. Part 3-Chemical Methods. Soil Science Society of America Inc., pp. 961e1010. O'Callaghan, M., Gerard, E.M., Carter, P.E., Lardner, R., Sarathchandra, U., Burch, G., Ghani, A., Bell, N., 2010. Effect of the nitrification inhibitor dicyandiamide (DCD) on microbial communities in a pasture soil amended with bovine urine. Soil Biology and Biochemistry 42, 1425e1436. http://dx.doi.org/10.1016/ j.soilbio.2010.05.003.

353

Olsen, S.R., Cole, C.V., Watanbe, F.S., Dean, L.A., 1954. Estimation of Available Phosphorus in Soils by Extraction with Sodium Bicarbonate (No. 939). United States Department of Agriculture. Rinke, C., et al., 2013. Insights into the phylogeny and coding potential of microbial dark matter. Nature 499 (7459), 431e437. Saggar, S., Jha, N., Deslippe, J., Bolan, N.S., Luo, J., Giltrap, D.L., Kim, D.G., Zaman, M., Tillman, R.W., 2013. Denitrification and N2O: N2 production in temperate grasslands: processes, measurements, modelling and mitigating negative impacts. Science of the Total Environment 465, 173e195. http://dx.doi.org/ 10.1016/j.scitotenv.2012.11.050. Sanford, R.A., Wagner, D.D., Wu, Q., Chee-Sanford, J.C., Thomas, S.H., Cruz-García, C., , A., Krishnani, K.K., Ritalahti, K.M., Nissen, S., Rodríguez, G., Massol-Deya €ffler, F.E., 2012. Unexpected nondenitrifier nitrous oxide Konstantinidis, K.T., Lo reductase gene diversity and abundance in soils. Proceedings of the National Academy of Sciences 109, 19709e19714. http://dx.doi.org/10.1073/ pnas.1211238109. Singh, J., Saggar, S., Giltrap, D.L., Bolan, N.S., 2008. Decomposition of dicyandiamide (DCD) in three contrasting soils and its effect on nitrous oxide emission, soil respiratory activity, and microbial biomassdan incubation study. Australian Journal of Soil Research 46, 517e525. http://dx.doi.org/10.1071/SR07204. Smith, I., Schallenberg, M., 2013. Occurrence of the agricultural nitrification inhibitor, dicyandiamide, in surface waters and its effects on nitrogen dynamics in an experimental aquatic system. Agriculture, Ecosystems & Environment 164, 23e31. http://dx.doi.org/10.1016/j.agee.2012.09.002. Statistics New Zealand [WWW Document], 2013. Statistics New Zealand [WWW Document]. URL. www.stats.govt.nz. http://www.stats.govt.nz (accessed 08.01.14.). Uchida, Y., Clough, T.J., Kelliher, F.M., Hunt, J.E., Sherlock, R.R., 2011. Effects of bovine urine, plants and temperature on N2O and CO2 emissions from a sub-tropical soil. Plant Soil 345, 171e186. http://dx.doi.org/10.1007/s11104-011-0769-z. van der Weerden, T.J., Luo, J., de Klein, C.A.M., Hoogendoorn, C.J., Littlejohn, R.P., Rys, G.J., 2011. Disaggregating nitrous oxide emission factors for ruminant urine and dung deposited onto pastoral soils. Agriculture, Ecosystems & Environment 141, 426e436. http://dx.doi.org/10.1016/j.agee.2011.04.007. Vance, E.D., Brookes, P.C., Jenkinson, D.S., 1987. An extraction method for measuring soil microbial biomass C. Soil Biology and Biochemistry 19, 703e707. Vogel, T.M., Simonet, P., Jansson, J.K., Hirsch, P.R., Tiedje, J.M., van Elsas, J.D., Bailey, M.J., Nalin, R., Philippot, L., 2009. TerraGenome: a consortium for the sequencing of a soil metagenome. Nature Reviews Microbiology 7, 252. Wakelin, S.A., Clough, T.J., Gerard, E.M., O'Callaghan, M., 2013. Impact of shortinterval, repeat application of dicyandiamide on soil N transformation in urine patches. Agriculture, Ecosystems & Environment 167, 60e70. http:// dx.doi.org/10.1016/j.agee.2013.01.007. Zacherl, B., Amberger, A., 1990. Effect of the nitrification inhibitors Dicyandiamide, Nitrapyrin and Thiourea on Nitrosomonas-Europaea. Fertilizer Research 22, 37e44. Zaman, M., Nguyen, M.L., 2012. How application timings of urease and nitrification inhibitors affect N losses from urine patches in pastoral system. Agriculture, Ecosystems & Environment 156, 37e48. http://dx.doi.org/10.1016/ j.agee.2012.04.025. Zaman, M., Nguyen, M.L., Blennerhassett, J.D., Quin, B.F., 2007. Reducing NH3, N2O and NO
3 eN losses from a pasture soil with urease or nitrification inhibitors and elemental S-amended nitrogenous fertilizers. Biology and Fertility of Soils 44, 693e705. http://dx.doi.org/10.1007/s00374-007-0252-4. Zaman, M., Saggar, S., Blennerhassett, J.D., Singh, J., 2009. Effect of urease and nitrification inhibitors on N transformation, gaseous emissions of ammonia and nitrous oxide, pasture yield and N uptake in grazed pasture system. Soil Biology and Biochemistry 41, 1270e1280.