SOILS, SEC 3 • REMEDIATION AND MANAGEMENT OF CONTAMINATED OR DEGRADED LANDS •

RESEARCH ARTICLE

Effect of treated farm dairy effluents, with or without animal urine,on nitrous oxide emissions, ammonia oxidisers and denitrifiersin the soil

Siyu Chen1& Hong Jie Di1 & Keith C. Cameron1

& Andriy Podolyan1& Jupei Shen2

& Jizheng He2

Received: 17 October 2018 /Accepted: 16 December 2018# Springer-Verlag GmbH Germany, part of Springer Nature 2019

AbstractPurpose In New Zealand, the application of farm dairy effluent (FDE) on pasture soils is the third largest source of nitrous oxide(N2O) emissions from grazed grassland. Recently, new FDE treatment technologies have been developed to produce clarifiedwater (CW) and treated effluent (TE) to recycle water and reduce the volume of fresh water used at the farm dairy. The aim of thisstudy was to compare the effects of CW and TE with those of FDE on N2O emissions and the growth of ammonia-oxidisingbacteria (AOB), ammonia-oxidising archaea (AOA) and denitrifiers, when the effluents were applied to a grazed pasture soil.Materials andmethods Amicrocosm incubation study was carried out to determine the effects of applying CW, TE and untreatedFDE, with or without animal urine, on N2O emissions, and the abundance of AOB, AOA and the denitrifying functional genes,including nirS, nirK and nosZ. The soil used was a Templeton silt loam (Udic Haplustepsts). The effluents were applied atnitrogen (N) rates equivalent to 100 kg N ha−1 and the animal urine at 700 kg N ha−1. The soils were incubated at 12 °C tosimulate autumn/winter soil temperatures in New Zealand, and the soil moisture was maintained at field capacity.Results and discussion Results showed that the application of all the different effluents significantly increased the total N2Oemissions (0.21–0.28 kg N2O-N ha−1) compared with that in the control (0.18 kg N2O-N ha−1). However, there were nosignificant differences in total N2O emissions between the different effluent treatments. Similarly, although the application ofanimal urine together with the different effluents further increased N2O emissions (7.7–8.8 kg N2O-N ha−1) above that from theurine only treatment (5.8 kgN2O-N ha−1), there were no significant differences among the different effluent plus urine treatments.These N2O results corresponded with the changing trends of the abundance of AOB, AOA, nirS, nirK and nosZ, that is theapplication of the CW, TE and FDE, with or without animal urine, had a similar impact on the growth dynamics of thesemicrobial populations.Conclusions These results indicate that the application of the CW and TE to dairy pasture soils would have a similar effect onN2O emissions, ammonia oxidisers and denitrifiers as that of the untreated FDE, with or without animal urine. The treatedeffluent or clear water from the new effluent treatment technology would therefore not increase N2O emissions nor adverselyaffect the key microbial populations involved in N cycling in soil.

Keywords Ammonia-oxidising archaea . Ammonia-oxidising bacteria . Clarified water . Denitrifiers .

Effluent treatment technology . Farm dairy effluent . Nitrous oxide . Treated effluent . Urine

1 Introduction

Nitrous oxide (N2O) is one of the most significant non-CO2

greenhouse gases with a long-term global warming potentialaround 298 times greater than CO2 (IPCC 2007) and is alsobecoming the largest contributor to stratospheric ozone deple-tion (Ravishankara et al. 2009). Agriculture is the largestsource of N2O emissions in New Zealand with 93.7% beingemitted from agricultural soils (Ministry for the Environment

Responsible editor: Zhihong Xu

* Hong Jie Dihong.di@lincoln.ac.nz

1 Centre for Soil and Environmental Research, Lincoln University,Christchurch, Lincoln 7647, New Zealand

2 Research Centre for Eco-Environmental Sciences, Chinese Academyof Sciences, Beijing, China

Journal of Soils and Sedimentshttps://doi.org/10.1007/s11368-018-02229-8

2017). Nitrous oxide is produced from biological processes,nitrification and denitrification, as a part of the nitrogen (N)cycle in soil (Thomson et al. 2012; Mosier et al. 1998).Through nitrification, ammonia (NH3) is oxidised to nitrite(NO2

−) by ammonia-oxidising bacteria (AOB) andammonia-oxidising archaea (AOA). Nitrite is then oxidisedto nitrate (NO3

−) by nitrite-oxidising bacteria. The nitratecan then be denitrified, and the enzymes involved in the deni-trification processes include nitrate reductase (encoded by thenarG gene), nitrite reductase (encoded by nirS and nirKgenes), nitric oxide reductase and nitrous oxide reductase(encoded by nosZ).

In New Zealand, N2O emissions account for about 20.6%of agricultural greenhouse gases (Ministry for theEnvironment 2017). The dominant land use in New Zealandis pastoral agriculture (Di and Cameron 2017). In grazedgrassland soils, the N returned to the soil in animal excreta(urine and dung) is an important part of the N cycle (Saggaret al. 2013). Generally, pasture grazing occurs all year roundin New Zealand (Luo et al. 2013). The intensification in

livestock farming, i.e. the conversion of sheep and beef farm-ing to dairy farming over the past two decades, has led to agrowing volume of manure and animal excreta (includingfarm dairy effluent) being produced. Farm dairy effluent(FDE) is a mixture of dairy cattle excreta and washdownwaterin and around the farm dairy which is collected into a pond orsump. In New Zealand, FDE is the most common animalmanure collected and applied to grazed grassland and FDErepresents around one quarter of lactating dairy cattle excretaonNew Zealand dairy farms (Luo et al. 2013). The applicationof FDE onto pasture land can recycle nutrients which im-proves soil fertility and increases farming system sustainabil-ity (Luo et al. 2008). However, application of FDE onto landis also a source of NO3

− leaching and N2O emissions fromsoils because of the addition of available N and C and in-creased soil moisture and anaerobic conditions followingFDE application (Bhandral et al. 2007). In New Zealand, theapplication of FDE on pastures is the third largest source ofN2O emissions from agriculture (after emissions from urineand fertiliser) (van der Weerden et al. 2016).

Recently, a new FDE treatment technology (commerciallyknown as ClearTech™, (Cameron and Di 2018) has beendeveloped to separate the solids out from the liquid to produceclarified water (CW) and treated effluent (TE), to reduce thevolume of water used for washing the yard by recycling theCW. Because this is a new effluent treatment technology, it isnot known if the CW and TE would affect N2O emissions, Ndynamics and microbial populations when applied to land,compared with the untreated FDE. Here, we report a micro-cosm study which investigated the effects of the new CWandTE produced from this new FDE treatment technology (withand without cow urine as would occur on farm), on N2Oemissions, nitrification, denitrification, nitrifiers and denitri-fiers in the soil, compared with untreated FDE. It ishypothesised that (a) the CW and TE would have similarN2O emission factors compared with the untreated FDE whenapplied to soil and (b) there would be similar interactive

Table 1 Properties of the soil used in the study

Property Value

Organic matter (g kg−1) 5.0

Total nitrogen (g kg−1) 2.7

Total carbon (g kg−1) 29.0

pH (H2O) 6.1

Olsen P (mg kg−1) 45.7

CEC (cmolc kg−1) 13.3

Exchangeable calcium (cmolc kg−1) 7.5

Exchangeable magnesium (cmolc kg−1) 1.2

Exchangeable potassium (cmolc kg−1) 1.0

Exchangeable sodium (cmolc kg−1) 0.3

Base saturation (%) 74.0

Sulphate sulphur (mg kg−1) 6.7

Table 2 Original properties of thethree different types of effluentused

Chemical property Untreated FDE Clarified water Treated effluent

Turbidity (NTU) 2277.0 10.7 4882.7

Total solid (g m−3) 4233.3 1706.7 11,266.7

pH 7.2 5.9 5.9

Total nitrogen (g m−3) 495.0 311.0 570.0

Ammonium (g m−3) 115.3 119.0 121.0

Nitrate + nitrite (g m−3) 0.1 4.0 2.8

Total phosphorus (g m−3) 42.0 0.8 77.7

Dissolved reactive phosphorus (g m−3) 23.3 0.0 0.0

Total carbon (g m−3) 1270.0 655.0 2933.3

Soluble carbon (g m−3) 713.3 596.7 670.0

Fe (g m−3) 9.3 10.7 794.0

J Soils Sediments

effects on N2O emissions and microbial growth between thetreated effluents (CW and TE) and untreated FDE and animalurine when they were co-applied to soil. The aim of this studywas to improve knowledge and fundamental understanding ofthe effect of different forms of FDE (including untreated FDE,CW and TE), with and without animal urine, on N2O emis-sions, mineral N dynamics, soil pH and the abundance ofAOB, AOA and denitrifying functional genes (nirS, nirKand nosZ).

2 Materials and methods

2.1 Soil and farm dairy effluents

A Templeton silt loam (NZ Soil Classification: ImmaturePallic Soil (Hewitt 1993); USDA Classification: UdicHaplustepsts (Soil Survey Staff 2014) was used in this study.Samples (0–10 cm depth) were collected from the LincolnUniversity Research Dairy Farm, about 20 km south ofChristchurch in Canterbury, New Zealand (43° 38′ S, 172°27′ E) (Table 1). The mean annual maximum and minimumtemperatures measured on the farm are 17 °C and 4 °C, re-spectively, with an average annual rainfall of 666 mm. Thesoil was thoroughlymixed, with the roots and stones removed,and sieved through a 5-mm sieve. The soil was acclimatised inthe incubator at 12 °C for 1 week before treatments wereapplied.

Both farm dairy effluent and cow urine were collected fromthe Lincoln University Demonstration Dairy Farm. A newfarm effluent treatment technology (Cameron and Di 2018)has been developed at Lincoln University to separate thesolids out from the effluent in order to purify and recycle thewater to wash the yard. The treatment technology involvesadding a ferric iron (Fe3+) compound to coagulate the colloi-dal solids in the FDE in a treatment tank. Once the solids aresettled at the bottom of the treatment tank, the CW, with aturbidity less than 50 nephelometric turbidity units (NTU),can be recycled for cleaning the farm yard. The TE (the moreconcentrated effluent that has settled at the lower part of the

tank) is emptied into the storage pond before being irrigated topasture when conditions are suitable. The original propertiesof the three types of effluent are shown in Table 2. The con-centration of total N in the untreated FDE, TCE and TE wasadjusted by adding 2.15 g, 2.55 g and 1.99 g urea to 1 Lvolumes of the different effluents, respectively, to producethe same N content of 1.5 g N L−1. This limited the volumeof the effluent applied to the pottles and jars so that the soilwould not become too wet. The concentration of total N in theurine was 5.65 g N L−1 and was standardised by adding 5.77 gurea to 2 L of urine to give a standard concentration of 7 gN L−1, similar to the N concentration in urine that has beenreported in other studies (Di and Cameron 2017).

2.2 Incubation experiment

Eight treatments, each with four replicates, were establishedfor the soil sampling (pottles) or N2O sampling (jars) experi-ments (Table 3). The treatments were (i) untreated FDE, (ii)CW, (iii) TE and (iv) control (water), all with and without cowurine. All the effluents were applied at the same equivalentrate of 100 kg N ha−1, and the urine was applied at the equiv-alent of 700 kg N ha−1 (equivalent to 91 mg N kg−1 and636 mg N kg−1 dry soil, respectively). Urine was applied tosimulate grazed pastures where the grazing animal depositsthe urine on to pasture soil during grazing.

The treated pottles (for soil sampling) or jars (for N2O gassampling) (see below for details) were placed inside an incu-bator in a randomised block design. The incubator was set to atemperature of 12 °C (simulating the autumn/winter soil tem-peratures in New Zealand). The soil moisture content wasadjusted to, and maintained at, field capacity during the incu-bation (30% gravimetric water content), which was equivalentto 56.4% water-filled pore space (WFPS).

There were 32 glass jars (1 L) used in the N2O emissionsampling trial. Each jar was packedwith 600 g soil (dry equiv-alent) to a bulk density of 1 g cm−3. The effluent and cow urinetreatments were applied evenly over the soil surface. The totalweight of each jar (without lid) was recorded. The soil mois-ture content (30%) was maintained twice a week, after taking

Table 3 Description of thetreatments Treatment number FDE type (100 kg N ha−1) Urine (700 kg N ha−1) Replicates

1 Control (water) 0 4

2 Untreated FDE 0 4

3 Clarified water 0 4

4 Treated effluent 0 4

5 Control (water) 700 4

6 Untreated FDE 700 4

7 Clarified water 700 4

8 Treated effluent 700 4

J Soils Sediments

Table4

The

prim

erpairsandPCRconditionsused

inreal-tim

eqP

CR

Target

group

Prim

ername

Sequence(5′–3′)

Lengthof

amplicon

(bp)

Primer

final

concentration(nM)

Therm

alprofile

Amplificationefficiency

(R2>0.99)(%

)References

Bacterial

amoA

amoA

1F5′-G

GGGTTTCTA

CTGG

TGGTGGT-3′

491

250

95°C

for2min

×1cycle;

95°C

for20

s,57

°Cfor30

s,72

°Cfor30

s,and

85°C

for10

s×40

cycles

96–98

(Rotthauweetal.

1997)

amoA

2R5′-CCCCTCKGSA

AAGCCTTCTTC-3′

Archaeal

amoA

Arch-am

oAF

5′-STA

ATGGTCTGG

CTTA

GACG-3′

635

250

95°C

for2min

×1cycle;

95°C

for20

s,55

°Cfor20

s,72

°Cfor30

s,and80

°Cfor

10s×40

cycles

92–94

(Francisetal.2005)

Arch-am

oAR

5′-G

CGGCCATCCATC

TGTA

TGT-3′

nirS

cd3af

5′-G

TSA

ACGTSA

AGGA

RACSG

G-3′

410

750

95°C

for2min

×1cycle;95

°Cfor45

s,55

°Cfor45

s,72

°Cfor45

s,and

85°C

for20

s×40

cycles

93–95

(Michoteyetal.2000)

R3cd

5′-G

ASTTCGG

RTGSG

TCTTGA-3′

(Throbäcketal.2004)

nirK

FlaC

u5′-ATCATGGTSCTG

CCGCG-3′

474

780

95°C

for2min

×1cycle;95

°Cfor20

s,55

°Cfor30

s,72

°Cfor30

s,and85

°Cfor10

s×40

cycles

98–100

(Hallin

andLindgren

1999)

R3C

u5′-G

CCTCGATCAGR

TTGTGGTT-3′

nosZ

(I)

nosZ-F

5′-CGYTGTTCMTCG

ACAGCCAG-3′

424

750

95°C

for2min

×1cycle;95

°Cfor20

s,55

°Cfor30

s,72

°Cfor30

s,and

85°C

for15

s×40

cycles;

94–99

(Kloos

etal.2001)

nosZ1622R

5′-CGSACCTTSTTGCC

STYGCG-3′

(Throbäcketal.2004)

nosZ

(II)

nosZ-II-F

5′-CTIG

GICCIY

TK

CAYAC-3′

698

1000

95°C

for2min

×1cycle;95

°Cfor30

s,50

°Cfor30

s,72

°Cfor45

s,and

80°C

for10

s×40

cycles

76–81

(Jones

etal.2013)

nosZ-II-R

5′-G

CIG

ARCARAA

ITCBGTRC-3′

J Soils Sediments

gas samples and in between gas sampling, by the addition ofdeionised water until the jar reached the recorded weight.There were two 1-cm-diameter holes in the lid of the samplingjar to allow aeration.

Gas samples were taken twice a week for the first 70 daysand then taken weekly for the remainder of the study. The

method used to measure N2O gas emissions in this studywas similar to that of Hutchinson and Mosier (1981). Duringgas sampling, the gas jars were taken out of the incubator andthe lids were removed and replaced with gas sampling lids,which contained a rubber septum, tap and needle. Two sam-ples (12 mL) were taken 30 min apart (one at time 0, and one

Fig. 1 Daily N2O-N emissions. aEffluent treatments. b Effluentsplus urine treatments. The errorbars represent the standard errorof the mean (n = 4)

Table 5 Emission factors (EF1)for effluent-N in effluent onlytreatments (A) and for urine-N(EF3) (B)

Treatments Total emissions (kg N2O-N ha−1) N applied (kg N ha−1) EF (%)

Control (water) 0.18 0

Untreated FDE 0.26 100 0.09a

Clarified water 0.21 100 0.03a

Treated effluent 0.28 100 0.10a

LSD (5%) 0.103

Urine 5.8 700 0.8a

Untreated FDE + urine 8.0 700 1.1b

Clarified water + urine 8.8 700 1.2b

Treated effluent + urine 7.7 700 1.1b

LSD (5%) 0.178

The EF values with different lowercase letters are significantly different (P < 0.05)

J Soils Sediments

30 min later), into a pre-evacuated 6-mL glass vial for analy-sis. A preliminary study showed a linear increase in N2Oduring the 30-min sampling interval. The concentration ofN2O in the gas was measured using a gas chromatograph(SRO8610 linked to a Filson 222XL autosampler) fitted withan electron capture detector (ECD) (SRI Instruments, USA).

Accompanying the N2Omeasurements, a total of 32 pottleswere established for soil sampling to determine nitrificationrate dynamics and the population abundance of ammoniaoxidisers and denitrifiers. Each pottle contained 500 g of soil(dry equivalent). After the effluent and urine treatments wereapplied, the soil was thoroughly mixed. There were two 1-cm-diameter holes in the lid to allow aeration. The weight of eachpottle (without lid) was recorded after each subsampling inorder to maintain soil moisture content by adding deionisedwater twice a week.

Soil subsamples were taken on days 1, 7, 14, 28, 63, 91,119, 154 and 210 after treatment application to determine the

concentrations of mineral N (including NH4+ and NO3

−), soilmoisture content, soil pH and the abundance of AOB, AOAand denitrifying functional genes (including nirS, nirK andnosZ). The mineral N, soil moisture content and soil pH wereanalysed immediately after sampling. For functional geneanalysis, a subsample of soil was stored in a − 80 °C freezeruntil analysed.

2.3 Analysis of mineral N and soil pH

Five grams of soil was taken, mixed with 25 mL of 2 M KCland shaken for 1 h at the speed of 120 oscillations per min.After centrifuging at 4000 rpm for 10min, the supernatant wasfiltered through 110-mm Advantec 5C filter paper. The ex-tracts were analysed for NH4

+ and NO3− using a flow injection

analyser (FIA) (Foss FIAstar 5000 Flow Injection Analyzerwith SoFIA software, version 2.00).

Fig. 2 Soil ammonium-Nconcentration. a Effluenttreatments. b Effluents plus urinetreatments. The error barsrepresent the standard error of themean (n = 4). Note the differencein the y-axis scale between a andb

J Soils Sediments

For soil pH measurement, 15 g of soil was mixed with25 mL deionised water and shaken well. Subsamples weresettled overnight (at least 12 h), and pH was measured by apH meter (Mettler Toledo SevenCompact).

2.4 Quantification of functional gene abundanceusing qPCR

DNA was extracted from 0.25 g soil subsamples using aNucleoSpin® Soil Kit (Macherey-Nagel, Düren, Germany)according to the manufacturer’s instructions. The abundanceof AOB amoA, AOA amoA, nirS, nirK, nosZ (I) and nosZ (II)was measured using qPCR on a Rotor-Gene™ 6000 (CorbettLife Science). A CAS-1200 Robotic liquid handling systemwas used to set up all PCR reactions (Corbett Life Science,Australia). All the soil DNA extraction samples were diluted1:10 with deionised water and used as a template in PCRreactions. The primer pairs used are shown in Table 4. A

typical reaction mixture contained 0.4–1.6 μL of each primer(Table 4), 8.0 μL of SYBR Premix Ex Taq (TaKaRa, NoriBiotech, Auckland, New Zealand), 1.5 μL template DNAand nuclease-free DI water to 16 μL as the final volume.The running of PCR was in accordance with the thermal pro-files shown in Table 4. After the amplification, a melting curveanalysis was done to confirm the specificity of the PCR prod-uct by measuring the fluorescence continuously during theincrease of temperature from 72 to 99 °C. Then, the datawas analysed by Rotor-Gene 6000 Series Software 1.7.

Standard curves for qPCR were performed using the fol-lowing process. The primers were used to amplify AOBamoA, AOA amoA, nirS, nirK, nosZ I and nosZ II from theextracted DNA. To purify the products of PCR, a clean-up kit(Axygen) was used and then cloned into the pGEM-T EasyVector (Promega, Madison, WI). The resulting clones werethen transformed into Escherichia coli JM109 competent cells(Promega) according to the manufacturer’s instructions. After

Fig. 3 The nitrate-Nconcentration. a Effluenttreatments. b Effluents plus urinetreatments. The error barsrepresent the standard error of themean (n = 4)

J Soils Sediments

the transformation, E. coli cells were grown on Luria-Bertani(LB) plates overnight at 37 °C. Then, 10 to 15 bacterial colo-nies from the LB plate were individually inoculated into a3 mL LB broth medium and incubated in an orbital incubatorshaker at 37 °C and 250 rpm overnight. The plasmids werethen extracted from overnight cultures using PureLink™Quick Plasmid Miniprep Kit (Life Technologies, Auckland,New Zealand). To generate the PCR amplicons containingeach gene of interest, the plasmids were then used as a tem-plate in the reactions of PCRwith T7 and SP6 primers. Furtherdetails can be found in Di et al. (2014).

2.5 Statistical analysis

All variables were statistically analysed using analysis of var-iance for a randomised complete block design. For N2O-Nemissions (and the emission factor), NH4-N and NO3-N, thedata values were an order of magnitude different between Bno

urine^ and Burine^ treatments, so these two sets of treatmentswere analysed separately (to avoid violating the essentialANOVA assumption of homogeneity of variance). In thecases of microbial functional genes, data values were logarith-mically transformed prior to analysis to ensure the homoge-neity of variance assumption was met.

3 Results

3.1 N2O emissions

The application of the CW, TE and the untreated FDE aloneresulted in significantly higher daily N2O emissions comparedto the control straight after application (P < 0.05, Fig. 1a).However, there were no significant differences in daily N2Oemissions between the different effluent treatments. Daily

Fig. 4 The soil pH. a Effluenttreatments. b Effluents plus urinetreatments. The error barsrepresent the standard error of themean (n = 4)

J Soils Sediments

N2O emissions decreased rapidly with time, reaching back-ground levels after about 30 days.

The application of the different effluents plus animalurine also significantly (P < 0.05) increased daily N2Oemissions compared to that in the control straight afterapplication (Fig. 1b). Daily N2O emissions reached peakvalues of between 68 and 109 g N2O-N ha−1 day−1 about75–100 days after treatment application (Fig. 1b). Thepeak daily N2O emissions recorded between 75 and100 days were significantly higher in the effluent plusurine treatments compared to those in the urine alonetreatment (P < 0.05). However, there were no significantdifferences among the untreated FDE plus urine, CW plusurine and TE plus urine treatments. Daily N2O emissionsgradually declined with time reaching background valuesabout 250 days after treatment application.

Total N2O emissions from the untreated FDE and TE weresignificantly higher than those from the control (P < 0.05), but

there was no significant difference in total N2O emissionsbetween the CWand control (Table 5). The differences in totalN2O emissions were not significant between the three differ-ent effluent treatments (P > 0.05).

The application of the three different effluents plus animalurine resulted in significantly higher total N2O emissions(8.8–7.7 kg N2O-N ha−1) than those in the urine alone treat-ment (5.8 kg N2O-N ha−1) (P < 0.05). However, there were nosignificant differences between the untreated FDE plus urine,CW plus urine and TE plus urine treatments (P > 0.05,Table 5).

The emission factors for the effluent-N (EF1) (the amountof N2O-N emitted as a % of the effluent N applied) in theuntreated FDE, CW and TE were similar, ranging from 0.03to 0.10% (Table 5). The emission factors for urine-N (EF3)were between 0.8 and 1.2% among the urine-applied treat-ments (Table 5). The EF3 values in the urine plus effluenttreatments were higher than those in the urine plus water

Fig. 5 AOB amoA geneabundance in the soil. a Effluenttreatments. b Effluents plus urinetreatments. The error barsrepresent the standard error of themean (n = 4)

J Soils Sediments

treatment (P < 0.05). However, there was no significant differ-ence on emission factors between the three different effluentsplus urine treatments (P > 0.05).

3.2 Soil ammonium-N dynamics

The application of the untreated FDE, CWand TE all resultedin significantly higher NH4

+-N concentrations in the soil,reaching between 77.9 and 63.1 mg NH4

+-N kg−1 soil, com-pared with that in the control straight after application (1.3 mgNH4

+-N kg−1 soil) (P < 0.05, Fig. 2a). The NH4+-N concen-

trations declined rapidly with time reaching backgroundvalues about 30 days after treatment application.

The application of animal urine resulted in a significant(P < 0.05) increase in NH4

+-N concentration in the soil wellabove those in the effluent only treatments (cf. Fig. 2a, b). TheNH4

+-N concentrations in the effluent plus urine treatmentswere higher than those in the urine alone treatment throughout

most of the incubation period. The NH4+-N concentrations

decreased rapidly with time (Fig. 2b). There was no signifi-cant difference in the NH4

+-N concentration among the un-treated FDE plus urine, CW plus urine and TE plus urinetreatments.

3.3 Soil nitrate-N dynamics

The application of the different effluents resulted in signifi-cantly higher NO3

−-N concentrations in the soil compared tothose in the control (water) treatment (P < 0.05, Fig. 3a).There was no difference in the NO3

−-N concentrations amongthe three different effluent treatments.

The application of urine resulted in significantly higherNO3

−-N concentrations compared to those in the effluent onlytreatments (cf. Fig. 3a, b). However, there was no significantdifference in NO3

−-N concentration between the different ef-fluent plus urine treatments (P > 0.05, Fig. 3b).

Fig. 6 AOA amoA geneabundance in the soil. a Effluenttreatments. b Effluents plus urinetreatments. The error barsrepresent the standard error of themean (n = 4)

J Soils Sediments

3.4 Soil pH

The application of the three effluents decreased the soil pHfrom 6.2 at the start of the incubation to between 5.0 and 5.5after 14 days of incubation (Fig. 4a). There was no significantdifference in soil pH between the effluent treatments(P > 0.05).

However, the application of animal urine in addition to theeffluents increased the soil pH to 7.6 at the start, which thendeclined to between 4.7 and 4.2 after 60 days of incubation(Fig. 4b). There was no significant difference in soil pH be-tween the different effluent plus urine treatments (P > 0.05).

3.5 Functional gene abundance

3.5.1 AOB

The application of the dairy effluents increased theAOB amoA gene copy numbers to between 6.75 × 107

and 7.65 × 107 copies g−1 soil after 14 days of incuba-tion (Fig. 5a). The AOB amoA gene copy numbers weresignificantly higher than those in the control for most ofthe sampling dates during the incubation (P < 0.05). TheAOB abundance then gradually declined over time, par-ticularly after 120 days of incubation. There was nosignificant difference in AOB abundance among thethree effluent treatments (P > 0.05).

The application of animal urine plus the differenteffluents increased the AOB amoA gene copy numbersto between 1.69 × 108 and 1.93 × 108 copies g−1 soilafter 28 days of incubation (Fig. 5b). These peakAOB amoA gene copy numbers were significantlyhigher than those in the effluent treatments withouturine (P < 0.05). The AOB abundance then declined rap-idly dropping to similar levels to those in the controlafter 210 days of incubation. There was no significantdifference in the AOB abundance among the differenteffluent plus urine treatments.

Fig. 7 nirS gene abundance in thesoil. a Effluent treatments. bEffluents plus urine treatments.The error bars represent thestandard error of the mean (n = 4)

J Soils Sediments

3.5.2 AOA

The AOA amoA gene copy numbers remained relativelystable between 0 and 150 days of incubation followingthe application of the FDE, TWC and FE (Fig. 6a). TheAOA abundance then increased after 210 days of incu-bation. There was no significant difference among theeffluent treatments throughout the entire incubation pe-riod (P > 0.05).

The AOA amoA gene copy numbers in the effluentplus urine treatments were similar to those in the effluentonly treatments between 0 and 90 days (Fig. 6b). TheAOA abundance then decreased after 90 days of incuba-tion to levels significantly below those in the control andall the effluent only treatments (P < 0.05, Fig. 6b). Therewas no significant difference in AOA amoA gene copynumbers among the different effluent plus urine treat-ments (P > 0.05).

3.5.3 nirS, nirK and nosZ

In general, similar dynamic trends with time were observed forthe denitrifier functional genes, nirS, nirK and nosZ (I and II)following the effluent and urine treatments (Figs. 7, 8, 9 and10). The application of the different effluents alone did notresult in major changes in the copy numbers of the denitrifierfunctional genes (Figs. 7, 8, 9 and 10). However, the urineapplication did reduce the copy numbers of these denitrifierfunctional genes. The denitrifier gene copy numbers in the ef-fluent plus urine treatments were similar to those in the effluentwithout urine treatments between 0 and 90 days of incubation(Figs. 7, 8, 9 and 10). The nirS, nirK and nosZ (I and II) geneabundances in the urine treatments decreased significantly tolevels below those in the control and effluent treatments be-tween 90 and 210 days (P < 0.05). However, there were nosignificant differences in the respective denitrifier gene copynumbers among the different effluent plus urine treatments.

Fig. 8 nirK gene abundance inthe soil. a Effluent treatments. bEffluents plus urine treatments.The error bars represent thestandard error of the mean (n = 4)

J Soils Sediments

4 Discussion

Results from this incubation study show that despite the differ-ences in the properties of the TE and CW compared with theuntreated FDE (Table 2), the N2O emissions from the threedifferent effluents were similar. This would indicate that theland application of the CW and TE produced from the neweffluent treatment technology (Cameron and Di 2018) wouldnot lead to increased N2O emissions compared with the landapplication of the untreated FDE. Thus, the first hypothesis ofthis study regarding the N2O emission potential from the treatedand untreated effluents was accepted. The N2O emission factorsfrom the three types of effluents, ranging from 0.03 to 0.1%,were lower than the standard 0.25% emission factor for effluentused for the 2017 New Zealand National Greenhouse GasInventory Report (Ministry for the Environment 2017). Theseresults from the laboratory incubation study are similar to thosereported from field lysimeter studies (Wang et al. 2018).

The emission factors of the animal urine-N appliedtogether with the different animal effluents, ranging from0.8 to 1.2%, were similar to the 1% EF3 value for animalurine used in New Zealand’s Greenhouse Gas Inventorycalculations (Ministry for the Environment 2017). Thesignificantly higher emission factors of the urine-N whenco-applied with all the different effluents compared withthe urine plus water treatment were probably because ofan interactive effect between the organic C in the effluentsand urine-N applied. It is known that organic C can en-hance denitrification, thus leading to increased N2O emis-sions (de Klein et al. 2001; Di and Cameron 2003;Cameron et al. 2013). However, it is important to notethat the emission factors of urine-N co-applied with theCW or TE were similar to those in the urine + the untreat-ed FDE (Table 5). This would indicate a similar interac-tive effect of the animal urine-N and the treated and un-treated effluents, when applied to the soil, in terms of

Fig. 9 nosZ I gene abundance inthe soil. a Effluent treatments. bEffluents plus urine treatments.The error bars represent thestandard error of the mean (n = 4)

J Soils Sediments

effects on N2O emissions. Thus, the second hypothesis ofthis study regarding the interactive effects of animalurine-N and the different effluents on N2O emissionswas accepted. The lack of a significant difference inN2O emissions among the three types of effluents, withor without urine, was probably because the difference inkey effluent properties (i.e. total N and total C) betweenthe effluent types was not large enough to lead to differentN2O emissions when the effluents are applied to the soil.More research is needed to verify these results under fieldconditions. The initial pulse of N2O emissions straightafter effluent and urine application was probably becauseof a priming effect following the application of the efflu-ents (Fig. 1a, b).

The similar ammonium and nitrate dynamics (Figs. 2 and 3)among the different effluents indicate that the nitrification rateswere similar in the soils treated with the different effluents.Therefore, the CW and TE from the new effluent treatment

technology would result in similar nitrification rates comparedto the untreated original effluent, when applied to soil.

During the process of nitrification, hydrogen ions (H+) arereleased; therefore, the soil pH decreases following the appli-cation of the effluents or animal urine (Fig. 4). The pH declinewas particularly significant where the animal urine was ap-plied because of the high rates of NH4

+ applied (Fig. 4b).Most of the N in animal urine is urea, which, upon hydrolysisin the soil, releases NH4

+, leading to increased nitrificationaccompanied by a pH decrease (Cameron et al. 2013; Di andCameron 2017). Following the urine application, the soil pHdid not decrease significantly but remained around 7.6 duringthe first 7 days of the incubation (Fig. 4b). This was becauseurea hydrolysis was still occurring and the H+ produced bynitrification was neutralised by the OH− produced by the ureahydrolysis. Further nitrification thereafter decreased the pH tobetween 4.7 and 4.2 after 60 days of the incubation (Fig. 4b).However, it is important to note that the pH decreased at the

Fig. 10 nosZ II gene abundancein the soil. a Effluent treatments.b Effluents plus urine treatments.The error bars represent thestandard error of the mean (n = 4)

J Soils Sediments

same rate among the different effluent treatments and amongthe different effluent plus urine treatments, again, indicatingsimilar nitrification rates among the effluent only treatmentsor effluent plus urine treatments.

It has been reported that the growth and activity of AOBand AOA may vary depending on soil and environmentalconditions (He et al. 2007; Dai et al. 2013; Robinson et al.2014; Di et al. 2014; Di and Cameron 2017). The AOB amoAgene abundance in the different effluent only treatments (in-cluding untreated FDE, CW and TE) showed similar trendsamong the different effluent treatments. This demonstratedthat the effluents treated by the new effluent treatment tech-nology had a similar effect on AOB growth in the soil com-pared with the untreated FDE. The NH4

+ contained in thedifferent effluents therefore stimulated AOB growth, increas-ing the AOB abundance above those in the control (P < 0.05).In contrast, the AOA population abundance did not changesignificantly until the end of the incubation study in the efflu-ent only treatments. In fact, the AOA abundance decreased inthe urine treatments, showing some inhibition effect on AOAabundance by the high rates of NH4

+ in the soil. These resultssupport those of Di et al. (2009, 2014, 2010) and Wang et al.(2011) who reported that the abundance and activity of AOBincreased in response to the addition of NH4

+, whereas AOAdid not grow or declined following high rates of NH4

+ appli-cation. The similar changing patterns of AOB or AOA amongthe respective effluent only or effluent plus urine treatments,again, demonstrated the similar effect of the three effluents onAOB or AOA populations.

Generally speaking, the application of the three differenteffluents alone or when applied together with animal urine didnot lead to different fluxes of the denitrifier abundance in thesoil among the effluent or effluent plus urine treatments. Thiswould indicate that despite the different compositions of thethree types of effluent, the application of the CW or TE isunlikely to lead to different denitrifier growth compared withthe application of the untreated FDE. Although the total or-ganic C content of the three different effluents were different(Table 2), the amounts of soluble organic C were similar, andit is the soluble organic C that is readily available to stimulatemicrobial activity. This may partly explain the similar denitri-fier responses to the application of the different effluents(Miller et al. 2009; Cameron et al. 2013). It was indicated byPaul and Beauchamp (1989) that denitrification was highlycorrelated with the concentration of water-soluble C in ma-nure and suggested that the water-soluble C might be con-sumed as a primary source of C by denitrifiers.

However, the denitrifier abundance decreased in the efflu-ent plus urine treatments compared with the Control at thelater stages of the incubation (Figs. 7, 8, 9 and 10). Soil pHmay have played a part in decreasing the denitrifier populationabundance in the urine treatments. Optimum soil pH fordenitrifying organisms was reported to be between 7 and 8

(Sherlock et al. 1992). In acidic soils, the denitrification rate isoften low (Šlmek and Cooper 2002; Fageria and Baligar 2008)particularly with pH less than 5 (McLaren and Cameron1996). Therefore, as soil pH decreased to below 5.0 after60 days of incubation and to around 4.4 after 90 days ofincubation, this may have limited denitrifier growth and activ-ity in the urine treatments. It has previously been reported thatthe copy numbers of denitrifiers in acidic soils were signifi-cantly lower than those in neutral pH soils (Čuhel et al. 2010).Therefore, soil pHmay have been an important factor in caus-ing the denitrifier abundance in the urine treatments to de-crease below that in the control.

5 Conclusions

Land application of farm dairy effluent to recycle the nutrientsis the preferred option for DFE management on New Zealanddairy farms. Results from this study demonstrated that the landapplication of the CWand TE produced from the new effluenttreatment technology, with or without animal urine, would notlead to different N2O emissions compared with the land ap-plication of the untreated FDE. Similarly, the application ofthese new effluents would also lead to similar changes in soilpH, nitrification rate dynamics and population abundance ofAOB, AOA and denitrifying functional genes in the soil com-pared with the untreated FDE. Further long-term studies underfield conditions are required to verify these findings from themicrocosm studies.

Acknowledgments Wewould like to thank Jie Lei, Carole Barlow, SteveMoore and Emily Huang of Lincoln University for the technical support.

Funding information The study received a catalyst fund from the NewZealand Ministry of Business, Innovation and Employment (MBIE) andfunding support from Ravensdown, Ltd., and the Programme ofIntergovernmental Cooperation in Science and Technology(2017YFE0109800).

Publisher’s note Springer Nature remains neutral with regard to jurisdic-tional claims in published maps and institutional affiliations.

References

Bhandral R, Bolan NS, Saggar S, Hedley MJ (2007) Nitrogen transfor-mation and nitrous oxide emissions from various types of farm ef-fluents. Nutr Cycl Agroecosyst 79(2):193–208

Cameron KC, Di HJ (2018) A new method to treat farm dairy effluent toproduce clarified water for recycling and to reduce environmentalrisks from the land application of effluent. J Soils Sediments (thisissue)

Cameron KC, Di HJ, Moir JL (2013) Nitrogen losses from the soil/plantsystem: a review. Ann Appl Biol 162(2):145–173

Čuhel J, ŠimekM, Laughlin RJ, BruD, ChènebyD,Watson CJ, PhilippotL (2010) Insights into the effect of soil pH onN2O and N2 emissions

J Soils Sediments

and denitrifier community size and activity. Appl EnvironMicrobiol76(6):1870–1878

Dai Y, Di HJ, Cameron KC, He JZ (2013) Effects of nitrogen applicationrate and a nitrification inhibitor dicyandiamide on ammonia oxi-dizers and N2O emissions in a grazed pasture soil. Sci TotalEnviron 465:125–135

de Klein CAM, Sherlock RR, Cameron KC, Van Der Weerden TJ (2001)Nitrous oxide emissions from agricultural soils in New Zealand—areview of current knowledge and directions for future research.Royal Soc NZ 31(3):543–574

Di HJ, Cameron KC (2003) Mitigation of nitrous oxide emissions inspray-irrigated grazed grassland by treating the soil withdicyandiamide, a nitrification inhibitor. Soil Use Manag 19(4):284–290

Di HJ, Cameron KC (2017) Ammonia oxidisers and their inhibition toreduce nitrogen losses in grazed grassland: a review. J Royal Soc NZ48(2–3):127–149

Di HJ, Cameron KC, Shen JP, Winefield CS, O’CallaghanM, Bowatte S,He JZ (2009) Nitrification driven by bacteria and not archaea innitrogen-rich grassland soils. Nat Geosci 2(9):621–624

Di HJ, Cameron KC, Shen JP, Winefield CS, O’CallaghanM, Bowatte S,He JZ (2010) Ammonia-oxidizing bacteria and archaea grow undercontrasting soil nitrogen conditions. FEMS Microbiol Ecol 72(3):386–394

Di HJ, Cameron KC, Podolyan A, Robinson A (2014) Effect of soilmoisture status and a nitrification inhibitor, dicyandiamide, on am-monia oxidizer and denitrifier growth and nitrous oxide emissions ina grassland soil. Soil Biol Biochem 73:59–68

Fageria N, Baligar V (2008) Ameliorating soil acidity of tropical oxisolsby liming for sustainable crop production. Adv Agron 99:345–399

Francis CA, Roberts KJ, Beman JM, Santoro AE, Oakley BB (2005)Ubiquity and diversity of ammonia-oxidizing archaea in water col-umns and sediments of the ocean. Proc Natl Acad Sci U S A102(41):14683–14688

Hallin S, Lindgren P (1999) PCR detection of genes encoding nitritereductase in denitrifying bacteria. Appl Environ Microbiol 65(4):1652–1657

He JZ, Shen JP, Zhang LM, Zhu YG, Zheng YM, Xu MG, Di HJ (2007)Quantitative analyses of the abundance and composition ofammonia-oxidizing bacteria and ammonia-oxidizing archaea of aChinese upland red soil under long-term fertilization practices.Environ Microbiol 9(9):2364–2374

Hewitt AE (1993) New Zealand soil classification. Reprint with correc-tions. Manaaki Whenua - Landcare Research New Zealand, Lincoln

Hutchinson G, Mosier A (1981) Improved soil cover method for fieldmeasurement of nitrous oxide fluxes. Soil Sci Soc Am J 45(2):311–316

IPCC (2007) Climate change 2007: mitigation: contribution of WorkingGroup III to the Fourth Assessment Report of the IntergovernmentalPanel on Climate Change: summary for policymakers and technicalsummary. Cambridge University Press, Cambridge

Jones CM, Graf DRH, Bru D, Philippot L, Hallin S (2013) The unac-counted yet abundant nitrous oxide-reducing microbial community:a potential nitrous oxide sink. ISME J 7(2):417–426

Kloos K, Mergel A, Rösch C, Bothe H (2001) Denitrification within thegenus Azospirillum and other associative bacteria. Funct Plant Biol28(9):991–998

Luo J, Saggar S, Bhandral R, Bolan N, Ledgard S, Lindsey S, Sun W(2008) Effects of irrigating dairy-grazed grassland with farm dairyeffluent on nitrous oxide emissions. Plant Soil 309(1–2):119–130

Luo J, Ledgard SF, Lindsey SB (2013) Nitrous oxide and greenhouse gasemissions from grazed pastures as affected by use of nitrification

inhibitor and restricted grazing regime. Sci Total Environ 465:107–114

McLaren RG, Cameron KC (1996) Soil science: sustainable productionand environmental protection. Oxford University Press, Auckland

Michotey V, Méjean V, Bonin P (2000) Comparison of methods forquantification of cytochrome cd 1-denitrifying bacteria in environ-mental marine samples. Appl Environ Microbiol 66(4):1564–1571

Miller M, Zebarth B, Dandie C, Burton D, Goyer C, Trevors J (2009)Influence of liquid manure on soil denitrifier abundance, denitrifica-tion, and nitrous oxide emissions. Soil Sci Soc Am J 73(3):760–768

Ministry for the Environment (2017) New Zealand’s Greenhouse GasInventory: 1990-2015. Wellington

Mosier A, Kroeze C, Nevison C, Oenema O, Seitzinger S, Van CleemputO (1998) Closing the global N2O budget: nitrous oxide emissionsthrough the agricultural nitrogen cycle. Nutr Cycl Agroecosyst52(2–3):225–248

Paul J, Beauchamp E (1989) Effect of carbon constituents in manure ondenitrification in soil. Can J Soil Sci 69(1):49–61

Ravishankara A, Daniel JS, Portmann RW (2009) Nitrous oxide (N2O):the dominant ozone-depleting substance emitted in the 21st century.Science 326(5949):123–125

Robinson A, Di HJ, Cameron KC, Podolyan A, He J (2014) The effect ofsoil pH and dicyandiamide (DCD) on N2O emissions and ammoniaoxidiser abundance in a stimulated grazed pasture soil. J SoilsSediments 14(8):1434–1444

Rotthauwe JH, Witzel KP, Liesack W (1997) The ammoniamonooxygenase structural gene amoA as a functional marker: mo-lecular fine-scale analysis of natural ammonia-oxidizing popula-tions. Appl Environ Microbiol 63(12):4704–4712

Saggar S, Jha N, Deslippe J, Bolan N, Luo J, Giltrap D, Kim D, ZamanM, Tillman R (2013) Denitrification and N2O: N2 production intemperate grasslands: processes, measurements, modelling and mit-igating negative impacts. Sci Total Environ 465:173–195

Sherlock R, Müller C, Russell J, Haynes R (1992) Inventory informationon nitrous oxide. Report for the Ministry of Environment,Wellington, pp 1–72

Šlmek M, Cooper J (2002) The influence of soil pH on denitrification:progress towards the understanding of this interaction over the last50 years. Eur J Soil Sci 53(3):345–354

Soil Survey Staff (2014) Keys to soil taxonomy. Soil taxonomy, 12th edn.U.S. Dept. of Agriculture, Natural Resources Conservation Service,Washington, D.C.

Thomson AJ, Giannopoulos G, Pretty J, Baggs EM, Richardson DJ(2012) Biological sources and sinks of nitrous oxide and strategiesto mitigate emissions. Philos Trans Biol Sci 367:1157–1168

Throbäck IN, Enwall K, Jarvis Å, Hallin S (2004) Reassessing PCRprimers targeting nirS, nirK and nosZ genes for community surveysof denitrifying bacteria with DGGE. FEMS Microbiol Ecol 49(3):401–417

van der Weerden T, Cox N, Luo J, Di H, Podolyan A, Phillips R, SaggarS, de Klein C, Ettema P, Rys G (2016) Refining the New Zealandnitrous oxide emission factor for urea fertiliser and farm dairy efflu-ent. Agric Ecosyst Environ 222:133–137

Wang S, Wang Y, Feng X, Zhai L, Zhu G (2011) Quantitative analyses ofammonia-oxidizing archaea and bacteria in the sediments of fournitrogen-rich wetlands in China. Appl Microbiol Biotechnol 90(2):779–787

Wang XM, Di HJ, Cameron, KC (2018) Effect of treated farm dairyeffluent on E. coli, phosphorus and nitrogen leaching and green-house gas emission: a field lysimeter study. J Soils Sediments (thisissue)

J Soils Sediments