Variability in responses of bacterial communities ... - Semantic Scholar

3 downloads 111 Views 2MB Size Report
Feb 6, 2015 - 509–15. Berger S, Jang I, Seo J, et al. A record of N2O and CH4 emissions .... gan Tutorial. American. Community Ecology Package. R pack-.
FEMS Microbiology Ecology, 91, 2015, fiv013 doi: 10.1093/femsec/fiv013 Advance Access Publication Date: 6 February 2015 Research Article

RESEARCH ARTICLE

Variability in responses of bacterial communities and nitrogen oxide emission to urea fertilization among various flooded paddy soils Ning Wang1,2 , Long-Jun Ding1,2 , Hui-Juan Xu2,3 , Hong-Bo Li4 , Jian-Qiang Su3,∗ and Yong-Guan Zhu1,3 1

State Key Laboratory of Urban and Regional Ecology, Research Center for Eco–Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China, 2 University of Chinese Academy of Sciences, Beijing 100049, China, 3 Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, China and 4 State Key Laboratory of Pollution Control and Resource Reuse, School of the Environment, Nanjing University, Nanjing 210046, China ∗ Corresponding author: Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, China. Tel: 13859904149; E-mail: [email protected] One sentence summary: Fertilization affect bacterial communities and element biogeochemical cycling in flooded paddy soils and these effects might differ between soil types. Editor: Riks Laanbroek

ABSTRACT Fertilization affects bacterial communities and element biogeochemical cycling in flooded paddy soils and the effect might differ among soil types. In this study, five paddy soils from Southern China were subjected to urea addition to explore impacts of fertilization on nitrogen oxide (N2 O) emission and bacterial community composition under the flooding condition. 16S rRNA gene-based illumina sequencing showed no obvious shifts in bacterial community composition of five soils after urea addition. However, some genera were affected by fertilization addition and the influenced genera varied among soils. During the late period (day 8–19) of flooding incubation without urea addition, N2 O emission rates were elevated for all soils. However, urea effects on N2 O emission were different among flooded soils. For soils where nirS and nirK gene abundances increased with urea addition, N2 O emission was significantly increased compared to control treatment. Redundancy analysis showed that dissolved organic carbon, ammonium (NH4 + ), ferrous iron (Fe2+ ) and nitrate (NO3 − ) in pore water explained 33.4% of the variation in soil bacterial community composition, implying that urea regimes influenced the relative abundance of some bacterial populations possibly by regulating soil characteristics and then influencing N2 O emission. These results provided insights into soil type-dependent effect of fertilization on the overall bacterial communities and nitrogen oxide emission in flooded paddy soils. Key words: bacterial population; nitrous oxide; urea; functional gene

INTRODUCTION Agricultural management significantly influences bacterial community composition and biogeochemical cyclings of ele-

ments in paddy soils (Zhao et al., 2013; Guo et al., 2014). Fertilization and flooding managements are important anthropogenic activities for paddy soils to improve soil fertility, and thus to

Received: 24 October 2014; Accepted: 30 January 2015  C FEMS 2015. All rights reserved. For permissions, please e-mail: [email protected]

1

2

FEMS Microbiology Ecology, 2015, Vol. 91, No. 3

achieve high crop yields (Zhao et al., 2014). They can greatly alter soil physical, chemical, and biological properties and processes, influencing soil microbial communities and nitrogen (N) biogeochemical cycling (Glick 1995; Nacke et al., 2011). However, fertilization influences are complex in various types of flooded paddy soils, which need more studies to elucidate. Previous studies have shown the effect of fertilization (especially N fertilization) on soil microbial community composition and N-cycling (Chu et al., 2007; Islam et al., 2011). The response of bacterial communities to N fertilization varied under different conditions. Ruppel and Makswitat (1999) showed that N fertilization significantly altered microbial community composition of paddy soils, but Marschner et al. (2001) showed no shift in bacterial community composition with the addition of ammonium nitrate (NH4 NO3 ) to paddy soils. Similarly, no consensus has been reached in the effects of N fertilization on N2 O emissions. By altering functional microbial community composition which regulates N2 O emission, ammonium fertilization promotes N2 O emission (Flessa et al., 1996; Chen et al., 2012). However, there are also reports that N2 O emission from paddy soils is not affected by N fertilization (Zou et al., 2005). The effects of fertilization on N2 O emission are still unclear due to the complex N2 O formation pathways, including nitrification, nitrifier denitrification, denitrification, codenitrification, dissimilatory nitrate reduction to ammonium, nitrate assimilation and chemodenitrification, among which nitrification and denitrification are predominant (Braker and Conrad 2011). These processes are affected by many factors including soil characteristics (Maljanen et al., 2004). Vymazal (2007) and Faulwetter et al. (2009) showed that bacterial community composition and N2 O emission are affected by oxygen and nitrate content, organic matter quantity, quality, and availability, redox potential, temperature, pH and soil type. Among these factors, soil type determining soil physicochemical characteristics (e.g. soil texture, nutrient level) is important in shaping bacterial community composition and regulating N2 O emission. Thus, the inconclusiveness regarding the effect of fertilization on soil bacterial community composition and N2 O emission among paddy soils might be related to the differences in soil types. More studies are therefore needed to investigate the variability in the effects of N fertilization on the bacterial community and N-cycling among different types of paddy soils. It was hypothesized that paddy soil physicochemical properties would affect the responses of bacterial community and N-cycling to fertilization under flooding condition. To verify this hypothesis, five typical paddy soils with various physicochemical properties were collected from Southern China. Soils were flooded and amended with urea. The specific

objectives were to (1) investigate the responses of soil bacterial community and N2 O production to fertilization under the flooding condition; (2) determine the variability in the responses among different types of paddy soils; and (3) evaluate the role of soil properties in controlling bacterial community shifts and N2 O emission. In addition, N-cycling functional genes were determined to link soil bacterial community changes to N2 O emission to explain the variability in the effects of fertilization on bacterial community and N2 O production among flooded soils.

MATERIALS AND METHODS Paddy soil sampling and basic properties Five types of paddy soils, including Jiaxing (JX) soil from Zhejiang Province, Yingtan (YT) soil from Jiangxi Province, Changde alluvium (CA) and Changde red clay (CR) soils from Hunan Province and Leizhou (LZ) soil from Guangzhou Province, China, were collected, respectively. Soil group, parent material and texture varied among soils (Table S1, Supporting Information). After collection, soil samples were air-dried, sieved (2 mm mesh), and homogenized for physicochemical property analyses and the incubation experiment. Soil total carbon (TC) and nitrogen (TN) were measured using an element analyzer (Vario MAX CNS, Elementar, Germany). Soil pH was measured in soil to deionized water suspension of 1:2.5 (w/v) using a Dual Channel pH/Conductivity Meter (XL60, Fisher Scientific, USA) (Lu 2000). Potential nitrification activity was measured using a chlorate inhibition method (He et al., 2007). Potential denitrification activity was determined using the acetylene (C2 H2 ) inhibition technique (Guo et al., 2011). Basic soil characteristics are listed in Table 1.

Incubation experiment For each soil, ∼2.4 kg dry weight (DW) of soil was mixed with deionized water at soil:water ratio of 1:1 for pre-incubation in dark at 25◦ C for 30 days. Then, each paddy soil was homogenized for incubation experiment. About 400 g DW of pre-incubated soil was transferred into wide-mouthed glass bottle (diameter 10 cm × height 15 cm), and then 400 mL of deionized water was added to flood soil. A soil porewater sampler (3S–10, Institute of Soil Science, Chinese Academy of Sciences, Nanjing, China) was vertically inserted into soil. Following this, two treatments were set in triplicates: urea treatment by adding urea solution into the flooding water at a rate of 200 mg N kg−1 dry soil and control treatment without urea addition. All bottles were incubated in

Table 1. Basic properties of five paddy soils used in this study. Paddy soil

JX YT CA CR LZ

pH

5.95 ± 0.06b 5.24 ± 0.04c 5.94 ± 0.03b 5.17 ± 0.01c 6.42 ± 0.01a

TC (%)

1.69 ± 0.02c 1.03 ± 0.03e 2.61 ± 0.04a 2.25 ± 0.02b 1.17 ± 0.02d

TN (%)

0.22 ± 0.04c 0.15 ± 0.03d 0.30 ± 0.02a 0.26 ± 0.03b 0.14 ± 0.02d

Nitrification activity (mg kg−1 d−1 )

Denitrification activity (mg kg−1 d−1 )

NH4 +

NO3 −

DOC

Fe2+

(mg L−1 )I

(mg L−1 )I

(mg L−1 )I

(mg L−1 )I

1.56 ± 0.17b 0.03 ± 0.02d 0.48 ± 0.01c 0.38 ± 0.11c 3.22 ± 0.06a

4.04 ± 0.21a 1.01 ± 0.02c 4.27 ± 0.01a 3.54 ± 0.10b 0.50 ± 0.22d

20.1 ± 1.56b 14.6 ± 1.51c 31.9 ± 2.53a 31.1 ± 2.45a 5.22 ± 0.35d

0.30 ± 0.02b 0.53 ± 0.11b 9.78 ± 1.22a ND ND

115 ± 12.4a 61.8 ± 0.71c 87.1 ± 7.41b 91.8 ± 6.57ab 16.4 ± 0.71d

38.8 ± 16.2a 51.6 ± 9.36a 8.18 ± 0.42c 12.8 ± 1.57b 2.79 ± 0.38d

I: concentration in soil porewater at day 0 of flooding incubation. a–d Different letters up values indicate significant (P < 0.05) differences among paddy soils. Data was given as mean and standard error of triplicate analyses. Abbreviations refer to: TC, total carbon; TN, total nitrogen; NH4 + , ammonium; NO3 − , nitrate; DOC, dissolved organic carbon; Fe2+ , ferrous ion; JX, Jiaxing soil; YT, Yingtan soil; CA, Changde Alluvium soil; CR, Changde Red soil; LZ, Leizhou soil; ND, not detected.

Wang et al.

dark at 25◦ C for 19 days. During the whole period of incubation, water was replenished to maintain a depth of overlying water at ∼2 cm. N2 O production and porewater samples were collected at different time intervals (0, 4, 6, 8, 13 and 19 days) of 19-day flooding incubation. Before gas sample collection, the pot was blocked with a rubber stopper inserted with a rubber plug. Gas samples were taken from the pot by a 50 mL air-tight syringe after blocking the pot for 2 h. After gas sampling, soil pore water was sampled using soil porewater sampler. At day 0, ∼1 g of soil samples were collected while soils samples at day 19 were destructively collected for bacterial community analysis. Soil samples collected at day 0 were defined as Control-0, while those collected at day 19 for control and urea treatments defined as Flood19 and Flood + Urea-19.

Porewater and soil gas emission measurement Porewater NH4 + and nitrate (NO3 − ) concentrations were measured using ion chromatography (ICS-3000, Dionex, USA), DOC was determined using total organic carbon (TOC) analyzer (Shimadzu TOC–Vcph, Japan), and dissolved ferrous ions (Fe2+ ) fixed with 0.5 M HCl was measured using colorimetry with a ferrozine solution (Lu 2000). Porewater parameters at day 0 of flooding incubation are shown in Table 1. N2 O concentrations in gas samples were analyzed within 1 day after sampling using gas chromatography (Agilent Technologies 7890A GC system, USA) with an electron capture detector.

3

cler 480II, Switzerland). The primer pairs of Arch-amoAF/ArchamoAR (Francis et al., 2005), amoA-1F/amoA-2R (Rotthauwe, ¨ Witzel and Liesack 1997), cd3aF/R3cd (Throback et al., 2004), FlaCu/R3Cu (Hallin and Lindgren 1999) and nosZ-F/nosZ-1622R ¨ (Throback et al., 2004) were used to amplify and quantify archaeal and bacterial amoA, nirS, nirK and nosZ genes, respectively. Each individual PCR (total volume 20 μL) contained 2 μL of DNA (20 ng), 10 μL of 2 × SYBR Premix Ex Taq TM II (Takara, Japan), 0.2 μM of each primer, 0.2 μL of 0.1% BSA (Takara). For archaeal amoA gene, the condition was 95◦ C for 2 min followed by 40 cycles of 95◦ C for 30 s, 58◦ C for 30 s and 72◦ C for 45 s. For bacterial amoA gene, the condition was 95◦ C for 2 min followed by 40 cycles of 95◦ C for 30 s, 60◦ C for 30 s and 72◦ C for 32 s. For nirS and nirK genes, the condition was 94◦ C for 2 min, followed by 35 cycles of 94◦ C for 30 s, 57◦ C for 1 min and 72◦ C for 1 min. For nosZ gene, the condition was 95◦ C for 1 min, followed by 40 cycles of 95◦ C for 15 s, 61.5◦ C for 15 s and 72◦ C for 34 s. Florescence was recorded at 72◦ C in each cycle. Negative controls without DNA template were included in each gene amplification. Standard curves were obtained using gradient dilutions of standard plasmids containing archaeal and bacterial amoA, nirS, nirK and nosZ genes with known copy numbers. Every sample was quantified in three parallel qPCR reactions to ensure the correct amplification. However, inhibition was eliminated by highly diluting the DNA extracts. Only the reactions with efficiencies >90% and correlation coefficients r2 > 0.99 were accepted. Target gene copy numbers were calculated from the standard curves and presented per milligram of dry weight of soil (copies mg−1 DW).

Data processing and statistical analysis Bacterial 16S rRNA gene amplification and Illumina sequencing Genomic DNA in the soil samples collected at day 0 (Control–0), day 19 without urea addition (Flood–19) and day 19 with urea addition (Flood + Urea–19) were extracted using FastDNA SPIN Kit (MP Biomedicals, Solon, OH, USA). Bacterial 16S rRNA gene amplifications were completed utilizing primers that target V3 region of the bacterial 16S rRNA gene, 338F (5 –ACTCCTACGGGAGGCAGCAG–3 ), 533R (5 – TTACCGCGGCTGCTGGCAC–3 ). Each pair of primers was used to amplify a certain soil sample and barcoded with a unique error-correcting eight-base barcode on both forward and reverse primers (Hamady et al., 2008). Each sample was amplified using R the following conditions in a final volume of 50 μL: 25 μL SYBR Premix Ex TaqTM II (2×), 1 μL BSA (20 mg mL−1 ), 1 μL 338F (10 μM), 1 μL 533R (10 μM), 18 μL ddH2 O and 4 μL template DNA (100–200 ng). The thermal profile includes an initial denaturation step at 95◦ C for 5 min, 35 cycles of 95◦ C for 30 s (denaturing), 56◦ C for 30 s (annealing), 72◦ C for 30 s (extension), followed by a final extension step for 5 min at 72◦ C. PCR products were purified using a universal DNA Purification Kit (Qiagen, Germany). The DNA concentrations of the PCR products were determined using SpectraMax M5 (Molecular Devices) with PicoGreen. After purification and quantification, an equal amount of PCR products from different samples were mixed, purified and quantified. Combined PCR products were sent to Beijing Genomics Institute (Shenzhen, China) for Illumina sequencing.

Real-time PCR Quantification of archaeal and bacterial amoA, nirK, nirS and nosZ genes were performed using real-time PCR system (LightCy-

Emission rate of N2 O (ERt , mg kg−1 h−1 ) was calculated as following, ERt =

C ×V×P ×M , R×T×t×m × 1000

(1)

where C represents the concentration of N2 O in the gases collected from pot experiment (μL L−1 ); V represents the volume of the bottle (mL); P represents the atmospheric pressure (KPa, 101.325); M represents the relative molecular mass of gases; R represents the universal gas constant (8.314 L KPa mol−1 K−1 ); T represents the thermodynamic temperature (K, 273.15); t represents the close time of bottle before gas sampling (h); m represents the weight of dry soil (g). Sequencing data were processed using the Quantitative Insights into Microbial Ecology toolkit-Version 1.7.0 following a procedure similar to Caporaso et al. (2010). After removing any low-quality or ambiguous reads, the qualified sequences were clustered into operational taxonomic units (OTUs) at 97% similarity level by default. The most abundant sequence in each OTU was selected as the representative sequence and was assigned to taxonomy using an RDP classifier (version 2.2) (Wang et al., 2007). To assess the internal (within-sample) complexity of individual bacterial populations, PD whole tree, chao1, the observed species and Shannon index were calculated. Rarefaction curves were generated to compare the level of bacterial community diversity between the samples. The differences in overall community composition between the samples were determined using principle component analysis (PCA) based on the genus matrix (Lozupone, Hamady and Knight 2006). To find out which factors were important in shaping bacterial community, redundancy analysis (RDA) was conducted using R (2.14.0, http://www.r-project.org/) with the community ecology

4

FEMS Microbiology Ecology, 2015, Vol. 91, No. 3

package vegan (2.0–4) (Oksanen 2011). Envfit function (999 permutations) was used to remove environmental variables which insignificantly contributed to the total soil microbial community variance. Pearson correlations between the environmental factors and abundant genera in the five paddy soils were conducted. A one-way analysis of variance was conducted to test the differences in soil properties (pH, TC, TN, nitrification activity, denitrification activity and concentrations of NH4 + , NO3 − , DOC and Fe2+ ) among different types of soils. All statistical analyses employed the Statistical Analysis Systems (SAS, Version 9.1.3 for Windows, Cary, NC, USA). All data were processed using Microsoft Excel and all diagrams were plotted with SigmaPlot 10.0.

RESULTS Paddy soil basic characteristics Basic characteristics varied considerably among the examined paddy soils (Table 1). Soil pH was significantly (P < 0.05) higher for soils LZ, JX and CA (5.94–6.42) than CR and YT (5.17 and 5.24). TC and TN contents were significantly (P < 0.05) higher in CR and CA (TC: 2.25 and 2.61%; TN: 0.26 and 0.30%) than those in JX, YT and LZ (TC: 1.03–1.69%; TN: 0.14–0.22%). The nitrification activity of soil LZ was the highest, while its potential denitrification activity was the lowest among five soils. Before the incubation experiment (i.e. day 0), porewater NH4 + , DOC and Fe2+ concentrations were lowest in soil LZ, while NO3 − was not detected in soils CR and LZ. Concentration of NO3 − was highest in CA among five soils.

Effect of urea addition on soil porewater characteristics in flooded soils Compared to control treatments, urea amendments significantly increased porewater NH4 + concentration for five soils during the 19-day incubation under flooding (Fig. 1A). In the treatments with urea, porewater NH4 + concentrations peaked at day 4, followed by slight decreases to the end of the incubation. For the control and treatments with urea, porewater NO3 − concentration in soil LZ increased continually from ∼0 to 38.7 and 24.6 mg L−1 during 19-day flooding incubation, while the increase of porewater NO3 − concentration was not observed for the other four soils (Fig. 1B). During the whole period of flooding incubation, NO3 − was not detected in porewater of soil CR, while it maintained at low levels close to 0 mg L−1 in soils JX, YT and CA. For soil LZ, urea amendment significantly decreased NO3 − concentration, while the effect was not observed for the other soils. For both the control and treatments with urea, porewater DOC concentrations in five soils showed significant increases during the early period (0–4 day) of flooding, followed by slight decreases during 4–19 day (Fig. 1C). At the end of the incubation, porewater DOC concentrations were significantly lower for YT and LZ (∼50.0 and ∼18.0 mg L−1 ) than those for JX, CA and CR (>100 mg L−1 ) regardless of urea addition. Urea amendment slightly increased DOC concentrations for YT and LZ, while having no obvious effects for the other soils. During the 19-day flooding incubation, porewater Fe2+ concentrations for JX, CA and CR were significantly lower in urea treatments compared to control treatments, while those for the other two soils were not affected by urea addition (Fig. 1D). Concentration of Fe2+ slightly increased during day 0–4 of incubation and then decreased during day 4–19 in control and urea treat-

ments for JX, YT and CR, but continually increased for CA until the end of the incubation.

Effect of urea amendment on bacterial community composition in flooded soils A total of 45 soil samples, collected from each paddy soil at day 0 (Control-0) and day 19 of flooding incubation with (Flood + Urea-19) and without urea addition (Flood-19) in triplicates, were subjected to soil bacterial community composition analysis using Illumina sequencing. A total of 472 644 sequences were obtained. After filtering, 438 715 high-quality bacterial 16S rRNA gene sequences, ∼92.8% of the total sequences remained. Rarefaction of observed species showed that even at a sequencing depth of 13 240, the diversity of soil bacteria continued increasing rapidly with increasing sequencing depth (Fig. S1, Supporting Information), suggesting the high diversity of soil bacteria. No significant differences in PD whole tree, observed species, chao1 and Shannon index at the sequencing depth were found between Flood-19 and Flood + Urea–19 (Table S2, Supporting Information), suggesting that urea addition had no significant effects on internal bacterial community diversity. The major bacterial community composition in paddy soils were Clostridia, Actinobacteria, Anaerolineae, Alphaproteobacteria and Deltaproteobacteria at class level, with Clostridia being dominant for all five paddy soils (Fig. S2, Supporting Information). At genus level, Clostridium was dominant in all soils, while soil YT had a relatively higher portion of Symbiobacterium (Fig. S3, Supporting Information). PCA was performed in each type soil at genus level (Fig. 2). For all five paddy soils, soil samples collected at day 0 (Control-0) were separated from those collected at day 19 with (Flood + Urea-19) and without (Flood-19) urea addition (P < 0.05), which overlapped with each other according to Adonis analysis (P > 0.05). This indicated that flooding incubation without urea addition (Control-0 vs Flood-19) significantly changed soil bacterial community composition, while urea amendment (Flood-19 vs Flood + Urea-19) had no obvious effects on flooded soil bacterial community composition although it altered the relative abundance of some bacterial populations. At genus level, the relative abundances of Bacillus for JX, Candidatus Solibacter, Anaerolinea, and Bacillus for YT, Bacillus and Caloramator for CA, Caloramater for CR, Clostridium and C. Solibacter for LZ soil were significantly higher at day 19 (Flood-19) than at day 0 (Control-0) of flooding incubation without urea addition (Fig. S4A, Supporting Information). Comparison between soil samples Flood-19 and Flood + Urea-19 revealed that urea addition significantly increased the relative abundance of Clostridium and decreased that of Symbiobacterium for soil YT (Fig. S4B, Supporting Information). Different from soil YT, urea amendment had no obvious effects on the relative abundance of the dominant genus of Clostridium for the other four soils (Fig. S3, Supporting Information), but on low-abundant genera such as C. Solibacter and Anaerolinea (Fig. S4B, Supporting Information). By envfit function (999 permutations), porewater Fe2+ , NO3 − , NH4 + and DOC that significantly correlated with the bacterial community composition were selected in RDA analysis. These factors together explained 33.4% of the soil bacterial community composition variation (Fig. 3). Porewater NO3 − and Fe2+ have stronger effects (longer arrow along X-axis) on bacterial community composition. NO3 − was negatively related with the dominant genus of Clostridium, while Fe2+ was positively related with Clostridium (Table S3, Supporting Information).

Wang et al.

5

Figure 1. Temporal changes in porewater ammonium (NH4 + ), nitrate (NO3 − ), DOC and ferrous iron (Fe2+ ) concentrations during 19 days of flooding incubation with (Urea) and without (Control) urea addition of 200 mg N kg−1 dry soil for five paddy soils (JX, YT, CA, CR and LZ). Each point represents the mean and standard error of triplicate analyses.

6

FEMS Microbiology Ecology, 2015, Vol. 91, No. 3

Figure 2. PCA of bacteria community composition in soil samples collected at day 0 (Control-0) and day 19 of flooding incubation with (Flood + Urea-19) and without (Flood-19) urea addition for each paddy soil. Each treatment had three replicates.

Figure 3. RDA of bacterial community composition at genus level in soil samples collected at day 0 (Control-0) and day 19 of flooding incubation with (Flood + Urea-19) and without (Flood-19) urea addition for five paddy soils (JX, YT, CA, CR and LZ). Left and bottom coordinate axis (black color) showed species variables values; right and top coordinate axis (red color) showed the arrows of environmental variables values.

Effect of urea amendment on N2 O emission in flooded soils Emission rates of N2 O in control treatments were significantly elevated at day 13 from all soils and were significantly higher in soil CA than other soils during the flooding period (Fig. 4). Follow-

ing urea amendment, N2 O emission at day 13 from soils JX, YT and CA was increased by 10.2-, 6.9- and 2.2-folds, while urea addition had no obvious effects for soils CR and LZ. The significant differences in N2 O emission between control and urea treatments for soils JX, YT and CA only occurred after 8 days of the incubation. Before day 8, N2 O emission rates in urea treatments

Wang et al.

7

Figure 4. Nitrous oxide (N2 O) emission rate for five paddy soils (JX, YT, CA, CR and LZ) during 19 days of flooding incubation with (Urea) and without (Control) urea addition of 200 mg N kg−1 dry soil. Each point represents the mean and standard error of triplicate analyses.

were similar to those in control treatments. After day 13, N2 O emission rates in the urea treatments decreased. However, at the end of the incubation, they were still significantly higher in urea treatments than control treatments for soils JX and CA.

Effect of urea amendment on N-cycling functional genes in flooded soils Archaeal amoA was the most abundant in soil LZ, while this soil had the lowest nirS, nirK and nosZ gene abundances among the studied five soils (Fig. 5). Flooding incubation without urea addition (Flood-19 vs Control-0) significantly increased the abundances of archaeal amoA, bacterial amoA, nirS and nirK genes in soils JX. Urea amendment (Flood + urea-19 vs Flood-19) significantly increased the abundances of nirS and nirK genes for soils JX, YT and CA, while it had no obvious effects for CR and LZ. Urea addition did not significantly shift the abundance of nosZ gene in all soils.

DISSCUSION Soil bacterial community shifts and N2 O emission during the flooding period without urea addition Previous studies have shown that bacterial communities in paddy soils could be shifted during flooding incubation (Noll et al., 2005; Kikuchi et al., 2007). We found similar results that 19-day flooding incubation without urea addition (control treatment) significantly changed bacterial community composition of five soils (Fig. 2), possibly through altering soil oxygen content (Noll et al., 2005). Aerobic organisms are favored at the beginning, while facultative anaerobic or anaerobic bacteria are better adapted in the later stages, when oxygen availability is limited (Ludemann, Arth and Liesack 2000). ¨ In addition to oxygen, shift in bacterial communities after 19-day flooding incubation might be related to changes in soil chemical characteristics (Fig. 1). Previous studies showed that the microbial transformations of iron and nutrient (carbon and nitrogen) were the dominant processes in flooded paddy soils (Schutter, Sandeno and Dick 2001; Weber et al., 2006; Ligi et al., 2014). The temporal changes in DOC, NH4 + , NO3 − and Fe2+ concentrations in porewater during the flooding period (Fig. 1) which could serve as available substrates of microbes (Chu et al., 2007) might influence the bacterial community shift (Ludemann, ¨

Arth and Liesack 2000). Significant correlations between the relative abundance of specific genera and NO3 − and Fe2+ as the dominant factors (Table S3, Supporting Information) further indicated that flooding influenced soil bacterial communities via altering soil chemical characteristics. However, the microbial populations affected by flooding incubation varied among the five soils (Fig. S4, Supporting Information), suggesting that the effects depended on soil type, which determined amounts and kinds of available substrates. Flooding could accelerate microbial C- and N-turnover in the soil and thus enhance N2 O emissions from soils (Wang et al., 2011; Berger et al., 2013). In this study, for the flooded soil without urea addition, N2 O emission rates were significantly increased for all soils at day 13 (Fig. 4), suggesting that the late period of flooding incubation might provide a suitable anaerobic environment and supplied C and N substrates for growth of N2 O-producing bacteria (Kimura 2000; Liesack, Schnell and Revsbech 2000). However, N2 O emission during flooding varied among soils. At day 13 of control treatments, N2 O emission rate was significantly higher for soil CA than other soils. The differences among soils might be attributed to different C and N substrates and indigenous bacterial community composition. For example, Bacillus, several members of which are known as denitrifying bacteria (Tiedje 1982), were enriched in CA after flooding incubation without urea addition (Fig. S4A, Supporting Information), probably contributing the elevated N2 O emission. Another reason might be related to the continual increase in porewater Fe2+ in CA during flooding incubation (Fig. 1D), which suggested that Fe(III) oxides might be continually reduced and then accelerated the oxidation of organic matter (Weber et al., 2006) which could supply more substrates to promote N2 O production.

Effect of urea amendment on soil bacterial communities Urea addition could alter bacterial communities in paddy soils, while there are also studies showing no change in bacterial communities with urea addition (Ruppel and Makswitat 1999; Marschner et al., 2001). In this study, urea amendment had no obvious effects on soil bacterial community composition shift (Fig. 2), but significantly increased the relative abundances of some genera (Fig. S4B, Supporting Information), which are

8

FEMS Microbiology Ecology, 2015, Vol. 91, No. 3

Figure 5. Abundances of N-related functional genes (archaeal and bacterial amoA, nirK, nirS and nosZ) in soil samples collected at day 0 (Control-0) and day 19 of flooding incubation with (Flood + Urea-19) and without (Flood-19) urea addition for five paddy soils (JX, YT, CA, CR and LZ). Different letters indicated significant (P < 0.05) differences among treatments for each soil. Each bar represents the mean and standard error of triplicate analyses.

related to the C- and N-cycling (Weber et al., 2006; Divya et al., 2011; Kanokratana et al., 2011). Anaerobic conditions prevail in the flooded paddy soils. However, oxic or partially oxic niches occur due to the diffusion of oxygen through the flooding water, providing a suitable environment for ammonium (NH4 + ) oxidation (Arth, Frenzel and Conrad 1998; Nicolaisen et al., 2004). Urea addition into soils significantly increased porewater NH4 + concentrations (Fig. 1A), which could be transferred by nitrification, denitrification, anammox and Feammox (Shrestha et al., 2009; Chen et al., 2012; Thamdrup 2012), and then influenced the relative abundances of some bacterial populations. In addition to NH4 + , we observed that urea addition also influenced DOC, NO3 − and Fe2+ concentrations (Fig. 1B–D), which played important roles in bacterial community variation among various soils (Fig. 3), implying that urea amendment influenced the relative abundance of some bacteria possibly by altering soil characteristics.

Different soil characteristics might explain why effects of urea amendment on specific genera varied among soils. NH4 + concentration was significantly correlated with genera of low abundance from JX, CA and CR, while it was correlated with high-abundance genera (Clostridium) from YT (Table S3, Supporting Information). It might be related to different DOC concentrations among soils. Urea addition increased NH4 + concentration, and then influenced the ratio of carbon and nitro¨ gen therefore shifting the soil bacterial community (Hogberg, ¨ Hogberg and Myrold 2007). During the flooding period, DOC was continually released into soil porewater, serving as the substrate of bacterial growth. However, DOC was gradually consumed with the increase of flooding time (Fig. 1C). At the end of the experiment, DOC concentration (100 mg L−1 ). The low DOC might be not enough to support the growth of the dominant bacteria in YT after 13 days. As a consequence, NH4 + could be transferred by

Wang et al.

various biological processes, which might directly or indirectly maintain the dominant bacterial growth (such as Clostridium) for soil YT. For soils JX, CA and CR, the decreased DOC (>100 mg L−1 ) might be still enough to support the dominant bacterial growth. Thus, the effect of NH4 + increase after urea addition on growth of the dominant bacteria was not apparent, but was obvious on some bacterial populations with lower abundances. DOC concentration in LZ was much lower than that in YT, while NH4 + concentration had no obvious correlation with the dominant bacteria. NO3 − seemed to be the main factor influencing the dominant bacterial community for LZ (Fig. 3), due to the accumulation of NO3 − in the late stage of flooding period (Fig. 1B). This was attributed to its highest nitrification activity but lowest denitrification activity (Table 1). The high NO3 − concentration possibly replaced low DOC, becoming the dominant controlling factor in shaping the bacterial communities. The significant correlation between the dominant genera Clostridium and NO3 − for soil LZ supported this hypothesis (Table S3, Supporting Information). NH4 + and NO3 − instead of DOC as bacterial growth substrate might be the reason why urea addition significantly increased porewater DOC concentrations for YT and LZ (Fig. 1C).

Effect of urea amendment on N-related functional genes Nitrification and denitrification are important processes in soil N-turnover, which are affected by urea amendment (Akiyama et al., 2013; Ligi et al., 2014). In this study, urea addition had no obvious effect in the abundances of archaeal amoA and bacterial amoA genes (Fig. 5). This is similar to Chen et al. (2010) suggesting that the abundances of archaeal amoA and bacterial amoA genes mostly depend on soil types regardless of urea addition. However, our difference from Chen et al. (2010) was that bacterial amoA gene abundance outnumbered archaeal amoA for all soils. This inconsistency might be attributed to high ammonium level, which could promote AOB growth (Di et al., 2009; Jia and Conrad 2009). Urea addition significantly elevated the nirS and nirK gene abundances for soils JX, YT and CA, suggesting input of N substrates stimulated N-turnover bacteria and accelerated their growth. Similar results have been reported (Chen et al., 2010). However, in soils CR and LZ, nirS and nirK gene abundances did not increase with urea addition, implying the response of the denitrifying genes to urea addition varied depending on soil properties. In soil CR, NO3 − was not detected in porewater during the whole flooding incubation (Fig. 1B). The lack of available NO3 − might explain why nirK and nirS gene abundances were not increased with urea addition. However, for soil LZ, high and increasing NO3 − concentration was found for urea treatment, but the nirK and nirS gene abundances were still low. The results implied that the dominant factor influencing the denitrifying bacteria abundance was not NO3 − but the other properties, which might be unaffected by urea addition.

Effect of urea amendment on N2 O emission Urea amendment had no obvious effects on soil bacterial community composition, but played significant roles in N2 O emission in soils JX, YT and CA. After 8 days, urea amendment significantly promoted N2 O production for JX, YT and CA (Fig. 4), possibly by increasing the abundances of nirK and nirS genes (Fig. 5). Another possibility is that urea addition increased NH4 +

9

concentrations, posing a toxic effect on enzymatic activity in N2 O reduction (Brons, Hagen and Zehnder 1991). However, this promoting effect of urea on N2 O production was not found in CR and LZ where increase in nirS and nirK gene abundances was not observed. The effect mechanisms of urea amendment on N2 O production in the two soils might be different. For CR, lower N2 O production might be due to the loss of NO3 − which was not detected during the whole period of flooding (Roy and Knowles 1994). Moreover, the relative abundance of Bacillus, several members of which are known as denitrifying bacteria (Tiedje 1982), was significantly lower in CR than other soils (Fig. S3, Supporting Information). In contrast, soil LZ had the highest nitrification activity and archaeal amoA, but its denitrification activity and nirS and nirK gene abundances were lowest (Table 1; Fig. 5), leading to a continual increase in porewater NO3 − (Fig. 1B). Although urea addition lowered NO3 − accumulation in porewater of LZ probably by suppressing growth of nitrifying bacteria (e.g. Nitrospira), NO3 − concentration was still significantly higher in LZ than other soils. The low denitrification activity was the reason why N2 O emission rate was not elevated with urea addition for soil LZ. In conclusion, agricultural managements played a vital role in shaping soil bacterial community and agricultural ecosystems. The present study showed that the effects of urea amendment on the relative abundances of specific genera and N2 O emission rates in flooded paddy soils varied depending on the soil types, which could deepen our understanding of impact of the soil management on soil microbial composition and function. Further quantitative analysis on the interaction between NH4 + and microbial communities and other nitrogen cyclingrelated functional genes in paddy soils might allow predictions of the microbial community change and N2 O emission from various paddy soils under different management strategies.

SUPPLEMENTARY DATA Supplementary data is available at FEMSEC online.

FUNDING This study was supported by the Strategic Priority Research Program of Chinese Academy of Science (Grant NO. XD15020302 and XD115020402) and the National Natural Science Foundation of China (41430858 and 41090282). Conflict of interest statement. None declared.

REFERENCES Akiyama H, Morimoto S, Hayatsu M, et al. Nitrification, ammonia-oxidizing communities, and N2 O and CH4 fluxes in an imperfectly drained agricultural field fertilized with coated urea with and without dicyandiamide. Biol Fert Soils 2013;49:213–23. Arth I, Frenzel P, Conrad R. Denitrification coupled to nitrification in the rhizosphere of rice. Soil Biol Biochem 1998;30: 509–15. Berger S, Jang I, Seo J, et al. A record of N2 O and CH4 emissions and underlying soil processes of Korean rice paddies as affected by different water management practices. Biogeochemistry 2013;115:317–32. Braker G, Conrad R. Diversity, structure, and size of N2 Oproducing microbial communities in soils—What matters for their functioning? Adv Appl Microbiol 2011;75:33.

10

FEMS Microbiology Ecology, 2015, Vol. 91, No. 3

Brons HJ, Hagen WR, Zehnder AJ. Ferrous iron dependent nitric oxide production in nitrate reducing cultures of Escherichia coli. Arch Microbiol 1991;155:341–7. Caporaso JG, Kuczynski J, Stombaugh J, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods 2010;7:335–6. Chen X, Shen JP, Xu ZH, et al. Soil type determines the abundance and community structure of ammonia-oxidizing bacteria and archaea in flooded paddy soils. J Soils Sediments 2010;10:1510–6. Chen Z, Liu J, Wu M, et al. Differentiated response of denitrifying communities to fertilization regime in paddy soil. Microb Ecol 2012;63:446–59. Chu H, Lin X, Fujii T, et al. Soil microbial biomass, dehydrogenase activity, bacterial community structure in response to longterm fertilizer management. Soil Biol Biochem 2007;39:2971–6. Di H, Cameron K, Shen JP, et al. Nitrification driven by bacteria and not archaea in nitrogen-rich grassland soils. Nat Geosci 2009;2:621–4. Divya B, Parvathi A, Bharathi PL, et al. 16S rRNA-based bacterial diversity in the organic-rich sediments underlying oxygendeficient waters of the eastern Arabian Sea. World J Microb Biot 2011;27:2821–33. Faulwetter JL, Gagnon V, Sundberg C, et al. Microbial processes influencing performance of treatment wetlands: a review. Ecol Eng 2009;35:987–1004. ¨ Flessa H, Pfau W, Dorsch P, et al. The influence of nitrate and ammonium fertilization on N2O release and CH4 uptake of a well-drained topsoil demonstrated by a soil microcosm experiment. J Plant Nutr Soil Sc 1996;159:499–503. Francis CA, Roberts KJ, Beman JM, et al. Ubiquity and diversity of ammonia-oxidizing archaea in water columns and sediments of the ocean. PNAS 2005;102:14 683–8. Glick BR. The enhancement of plant growth by free-living bacteria. Can J Microbiol 1995;41:109–17. Guo GX, Deng H, Qiao M, et al. Effect of pyrene on denitrification activity and abundance and composition of denitrifying community in an agricultural soil. Environ Pollut 2011;159:1886–95. Guo LJ, Zhang ZS, Wang DD, et al. Effects of short-term conservation management practices on soil organic carbon fractions and microbial community composition under a rice-wheat rotation system. Biol Fert Soils 2014:51:65–75. Hallin S, Lindgren PE. PCR detection of genes encoding nitrite reductase in denitrifying bacteria. Appl Environ Microb 1999;65:1652–7 Hamady M, Walker JJ, Harris JK, et al. Error-correcting barcoded primers allow hundreds of samples to be pyrosequenced in multiplex. Nat Methods 2008;5:235. He JZ, Shen JP, Zhang LM, et al. Quantitative analyses of the abundance and composition of ammonia-oxidizing bacteria and ammonia-oxidizing archaea of a Chinese upland red soil under long-term fertilization practices. Environ Microbiol 2007;9:2364–74. ¨ ¨ Hogberg MN, Hogberg P, Myrold DD. Is microbial community composition in boreal forest soils determined by pH, C-to-N ratio, the trees, or all three? Oecologia 2007;150:590–601. Islam MR, Chauhan PS, Kim Y, et al. Community level functional diversity and enzyme activities in paddy soils under different long-term fertilizer management practices. Biol Fert Soils 2011;47:599–604. Jia Z, Conrad R. Bacteria rather than Archaea dominate microbial ammonia oxidation in an agricultural soil. Environ Microbiol 2009;11:1658–1671.

Kanokratana P, Uengwetwanit T, Rattanachomsri U, et al. Insights into the phylogeny and metabolic potential of a primary tropical peat swamp forest microbial community by metagenomic analysis. Microb Ecol 2011;61:518–28. Kikuchi H, Watanabe T, Jia Z, Kimura M, et al. Molecular analyses reveal stability of bacterial communities in bulk soil of a Japanese paddy field: estimation by denaturing gradient gel electrophoresis of 16S rRNA genes amplified from DNA accompanied with RNA. Soil Sci Plant Nutr 2007;53:448–58. Kimura M. Anaerobic microbiology in waterlogged rice fields. Soil Biochem 2000;10:35–138. Liesack W, Schnell S, Revsbech NP. Microbiology of flooded rice paddies. FEMS Microbiol Rev 2000;24:625–45. Ligi T, Truu M, Truu J, et al. Effects of soil chemical characteristics and water regime on denitrification genes (nirS, nirK, and nosZ) abundances in a created riverine wetland complex. Ecol Eng 2014;72:47–55. Lozupone C, Hamady M, Knight R. UniFrac–An online tool for comparing microbial community diversity in a phylogenetic context. Bmc Bioinformatics 2006;7:371. Lu R. Analysis Methods of Soil Agricultural Chemistry. China: Agricultural Science and Technology Publishing House (In Chinese), 2000. Ludemann H, Arth I, Liesack W. Spatial changes in the bacte¨ rial community structure along a vertical oxygen gradient in flooded paddy soil cores. Appl Environ Microb 2000;66:754–62. ¨ Maljanen M, Komulainen V-M, Hytonen J, et al. Carbon dioxide, nitrous oxide and methane dynamics in boreal organic agricultural soils with different soil characteristics. Soil Biol Biochem 2004;36:1801–8. Marschner P, Yang C-H, Lieberei R, et al. Soil and plant specific effects on bacterial community composition in the rhizosphere. Soil Biol Biochem 2001;33:1437–45. Nacke H, Thurmer A, Wollherr A, et al. Pyrosequencing-based ¨ assessment of bacterial community structure along different management types in German forest and grassland soils. Plos One 2011;6:e17000. Nicolaisen MH, Risgaard-Petersen N, Revsbech NP, et al. Nitrification–denitrification dynamics and community structure of ammonia oxidizing bacteria in a high yield irrigated Philippine rice field. FEMS Microbiol Ecol 2004;49:359–69. Noll M, Matthies D, Frenzel P, et al. Succession of bacterial community structure and diversity in a paddy soil oxygen gradient. Environ Microbiol 2005;7:382–95. Oksanen J. Multivariate Analysis of Ecological Communities in R: Vegan Tutorial. American. Community Ecology Package. R package version 1(7), 2011. Rotthauwe JH, Witzel KP, Liesack W. The ammonia monooxygenase structural gene amoA as a functional marker: molecular fine-scale analysis of natural ammonia-oxidizing populations. Appl Environ Microb 1997;63:4704–12. Roy R, Knowles R. Effects of methane metabolism on nitrification and nitrous oxide production in polluted freshwater sediment. Appl Environ Microb 1994;60:3307–14. Ruppel S, Makswitat E. Effect of nitrogen fertilization and irrigation on soil microbial activities and population dynamics—a field study. J Plant Nutr Soil Sci 1999;162:75–81. Schutter M, Sandeno J, Dick R. Seasonal, soil type, and alternative management influences on microbial communities of vegetable cropping systems. Biol Fert Soils 2001;34:397–410. Shrestha J, Rich JJ, Ehrenfeld JG, et al. Oxidation of ammonium to nitrite under iron-reducing conditions in wetland soils: Laboratory, field demonstrations, and push-pull rate determination. Soil Sci 2009;174:156–64.

Wang et al.

Thamdrup B. New pathways and processes in the global nitrogen cycle. Annu Rev Ecol Evol S 2012;43:407–28. ˚ et al. Reassessing PCR primers ¨ Throback IN, Enwall K, Jarvis A, targeting nirS, nirK and nosZ genes for community surveys of denitrifying bacteria with DGGE. FEMS Microbiol Ecol 2004;49:401–17. Tiedje JM. Denitrification. In: Page AL, Miller RH, Keeney DR, et al. Methods of Soil Analysis. Part 2. Chemical and Microbiological Properties. Madison, WI: American Society of Agronomy, 1982, 1011–26. Vymazal J. Removal of nutrients in various types of constructed wetlands. Sci Total Environ 2007;380:48–65. Wang J, Jia J, Xiong Z, et al. Water regime–nitrogen fertilizer– straw incorporation interaction: field study on nitrous oxide emissions from a rice agroecosystem in Nanjing, China. Agr Ecosyst Environ 2011;141:437–46.

11

Wang Q, Garrity GM, Tiedje JM, et al. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microb 2007;73:5261–7. Weber KA, Urrutia MM, Churchill PF, et al. Anaerobic redox cycling of iron by freshwater sediment microorganisms. Environ Microbiol 2006;8:100–13. Zhao G, Bryan BA, King D, et al. Impact of agricultural management practices on soil organic carbon: simulation of Australian wheat systems. Global Change Biol 2013;19:1585–97. Zhao J, Ni T, Li Y, et al. Responses of bacterial communities in arable soils in a rice-wheat cropping system to different fertilizer regimes and sampling times. Plos One 2014;9:e85301. Zou J, Huang Y, Jiang J, et al. A 3-year field measurement of methane and nitrous oxide emissions from rice paddies in China: effects of water regime, crop residue, and fertilizer application. Global Biogeochem Cy 2005;19:1–9.