APP下载

Long-time precipitation reduction and nitrogen deposition increase alter soilnitrogen dynamic by influencing soil bacterial communities and functional groups

2020-06-24GuoyongYANYajuanXINGShijieHANJunhuiZHANGQingguiWANGandChangchengMU

Pedosphere 2020年3期

Guoyong YAN,Yajuan XING,Shijie HAN,Junhui ZHANG,Qinggui WANG,∗and Changcheng MU

1School of Forestry,Northeast Forestry University,Harbin 150040(China)

2College of Agricultural Resource and Environment,Heilongjiang University,Harbin 150080(China)

3College of Life Sciences,Henan University,Kaifeng 475004(China)

4institute of Applied Ecology,Chinese Academy of Sciences,Shenyang 110016(China)

ABSTRACT The effects of precipitation reduction and nitrogen deposition increase on soil bacterial communities and functions impact soil nitrogen cycling.Seasonal changes could modify the effects of precipitation reduction and nitrogen deposition increase on bacterial communities and functions by changing soil environments and properties.Understanding soil microbial communities and the seasonal response of functions to precipitation reduction and nitrogen deposition increase may be important for the accurate prediction of changes in the soil nitrogen dynamics.Thus,a long-term field simulation experiment of nitrogen deposition increase and throughfall exclusion was established to investigate soil bacterial communities’response to nitrogen deposition increase and/or precipitation reduction,with no nitrogen deposition increase and no precipation reduction as a control,in a temperate forest.We examined soil bacterial communities(Illumina sequencing)under different treatments during the winter,freezing-thawing cycle periods(FTCs),and growing season.The bacterial functional groups were predicted by the FAPROTAX database.The results showed that nitrogen deposition increase,precipitation reduction,the combined effect of nitrogen deposition increase and precipitation reduction,and seasonal changes significantly altered the soil bacterial community composition.Interestingly,by combining the result of a previous study in which nitrogen deposition increase increased the nitrous oxide flux in the same experimental system,the loss of soil nitrogen was increased by the decrease in denitrification and increase of nitrification bacteria under nitrogen deposition increase,while ammonification bacteria significantly increased and N-fixing bacteria significantly decreased with precipitation reduction compared to the control.In relation to seasonal changes,the aromatic-degrading,cellulolytic,and ureolytic bacteria were lowest during FTCs,which indicated that FTCs might inhibit biodegradation.Nitrification and nitrite-oxidizing bacteria increased with nitrogen deposition increase or precipitation reduction and in FTCs compared to the control or other seasons.The interaction between treatment and season significantly changed the soil bacterial communities and functions.These results highlight that nitrogen deposition increase,precipitation reduction,seasonal changes,and their interactions might directly alter bacterial communities and indirectly alter the dynamics of soil N.

Key Words: copiotrophic bacteria,denitrification,multidimensional effect,nitrification,oligotrophic bacteria,seasonal changes,soil acidification,water stress

INTRODUCTION

Soil bacterial communities are an important part of the soil ecosystem,driving organic matter decomposition and soil biogeochemical processes that regulate soil carbon and nitrogen cycling(Bardgett and Wardle,2010).The responses of bacterial communities to environmental change will be important for the efforts to maintain ecosystem functions and to predict biogeochemical feedbacks to climate change(Evans and Wallenstein, 2014). Therefore, to understand how soil bacterial communities respond to is a particularly urgent and important need based on recent evidence of rapid global change.

Changes in global- and regional-scale precipitation patterns, along with altered water availability, have been observed in terrestrial ecosystems, while a significant decrease in precipitation during the 21st century has been predicted in some areas(IPCC,2013).Soil water influences the N-mineralization mediated by bacteria and plays a key role in the intracellular metabolism, transportation of nutrients, and serves as a medium for bacterial mobility(Drenovsky et al., 2004; Waring and Hawkes, 2018).Therefore,precipitation reduction can induce physiological stress and changes in the composition and functional characteristics of bacterial communities due to a decrease in water availability(Jurburg et al.,2018;Ochoa-Hueso et al.,2018).In general,precipitation reduction tends to decrease the diversity and abundance of bacterial communities(Fry et al., 2016; Ochoa-Hueso et al., 2018). However, some studies have found no significant changes in the abundance and composition of bacterial communities subjected to precipitation reduction(Griffiths and Philippot,2013).These inconsistent results may be caused by background precipitation variability (e.g., seasonal variability), which could affect bacterial communities more strongly than the direct effect of experimental precipitation treatments(Cregger et al.,2012;Gutknecht et al.,2012).Moreover,seasonal variation in soil moisture may result in microbial communities acclimated to fluctuations in precipitation,thus resulting in diminished responses to precipitation manipulation(Clark et al.,2009;Griffiths and Philippot,2013).Therefore,a better understanding of the responses of soil bacterial communities across seasons would enable researchers to better predict bacteria responses to precipitation reduction.

As another widespread global change factor,global anthropogenic nitrogen deposition has increased dramatically(Galloway et al.,2014).Nitrogen is a fundamental element that regulates the soil bacterial community composition and function(Zhou et al.,2018),particularly in temperate forests,where nitrogen is generally considered to be a major limiting factor for microbial growth and development(Wang et al.,2012).Nitrogen deposition increase can alleviate the nitrogen limitation of soil bacteria,which results in an increase in soil bacterial activity and a change in bacterial communities and functions(Li et al.,2016;Zhou et al.,2018).Previous studies have shown that nitrogen deposition increase tends to stimulate copiotrophic bacteria(bacterial taxa that have higher nitrogen demands,such as Proteobacteria and Bacteroidetes)(Ramirez et al.,2012;Ho et al.,2017).Another study about the effects of nitrogen deposition increase on bacterial communities also found negative or neutral effects(Hesse et al., 2015; Zeng et al., 2016). Although the responses of soil bacterial communities to nitrogen deposition increase have been studied extensively,the effect of nitrogen deposition increase on bacterial communities remains unclear(Ramirez et al.,2012;Nguyen et al.,2018).In this study, we used multipoint measurements to find common responses of bacteria to nitrogen deposition increase.

Moreover,soil water affects bacterial-mediated nitrogen processes physically,and thus the effects of nitrogen deposition increase on soil bacterial communities may strongly depend on soil water regimes(Bi et al.,2012;Li et al.,2016;Shi et al.,2018).Previous studies have shown that sufficient water can enhance the effects of nitrogen deposition increase on soil bacterial communities(Bi et al.,2012;Zhang et al., 2015), which suggests that precipitation reduction may alleviate the effects of nitrogen deposition increase on soil bacterial communities.In addition,environments that have greater seasonal and temporal variation, particularly in temperate and boreal forests with cold winters and long spring freezing-thawing periods(FTCs),may regulate the effects of nitrogen deposition increase and precipitation reduction on soil bacterial communities because a wide range of physiological tolerances may already exist within the community(Cruz-Martínez et al.,2009).Alternatively,long-time nitrogen deposition increase and precipitation reduction may also increase the severity of this variation,thereby resulting in new dynamics of bacterial community composition(Waldrop and Firestone,2006).Therefore,a better understanding of the responses of soil bacterial communities across seasons(especially during FTCs)would enable researchers to better predict bacteria responses to long-time and multiple climate changes.

In this study, in order to elucidate how soil bacterial communities and functional groups respond to nitrogen deposition increase,precipitation reduction,and their combined effects,we established a field experiment to simulate precipitation reduction(−30%throughfall)and nitrogen deposition increase(+5 g N m−2year−1)in a temperate forest.Previous studies in this experimental system had examined the effects of these treatments on litter decomposition,soil nitrous oxide (N2O) flux, and plant nitrogen uptake, but not its effects on the bacterial communities (Geng et al.,2017;Zheng et al.,2017;Zhou et al.,2019).Therefore,soil samples were collected,and soil bacterial communities were evaluated under different treatments during the winter,the FTC,and the growing season.We hypothesize that nitrogen deposition increase would alter soil bacterial communities and functional groups by increasing the relative abundance of copiotrophic bacteria,while precipitation reduction would change bacterial communities and functional groups by decreasing the movement and diffusion of substrates.Seasonal water variability strongly regulates the response of soil bacterial communities to precipitation reduction,while soil water variability could also alter the effects of nitrogen deposition increase on soil bacterial communities(Clark et al.,2009;Cruz-Martínez et al.,2009).Thus,we further hypothesize that the combined effects of precipitation reduction and nitrogen deposition increase and the interaction between treatments and seasonal changes would have a significantly different effect on soil bacterial communities compared with a single treatment.

MATERIALS AND METHODS

Site description

This study was conducted in the Changbai Mountain,northeastern China(42◦24′N,128◦06′E,and 738 m above sea level),which has a temperate-continental climate.The mean annual temperature is 3.6◦C. The average annual precipitation is 760 mm,more than 80%of which falls between May and August.The soil is classified as Eutric Cambisol according to the FAO classification(Nachtergaele et al.,2000).The slope is less than 30 degrees at the study site.The mean soil bulk density is 0.35 g cm−3at 0—10 cm soil depth and 0.68 g cm−3at 10—20 cm soil depth.The vegetation type is an old broad-leaved Korean pine mixed forest with an age exceeding 300 years(Zheng et al.,2017).The dominant four tree species are Pinus koraiensis,Quercus mongolica,Tilia amurensis, and Fraxinus mandshurica. The mean canopy height and diameter at 1.3 m breast height for these tree species are 26.5 m and 33.8 cm,respectively(Zheng et al.,2017).The tree density on the site is 435 trees per hectare.The main shrub species include Philadelphus schrenkii,Lonicera japonica,Euonymus alatus,Deutzia scabra,and Corylus mandshurica,while the main herbaceous species include Anemone cathayensis,Anemone raddeana,Funaria officinalis,Cyperus microiria,Brachybotrys paridiformis,Adonis vernalis,and Filipendula palmata.

Annual precipitation has declined in North China over the last 50 years(Wang et al.,2006).In the drought years of 1985,1997,1999,2001,and 2003,annual precipitation dropped to approximately 550 mm,which is approximately 30% less than the mean annual precipitation of 760 mm in the Changbai Mountains area in the last 30 years(these data were obtained from the Chinese Ecosystem Research Net(CERN)).In this study,the snow cover lasted for 146 d,with a maximum thickness of 45 cm during the experimental period of 2016/2017.

Experimental design

In this study,six plots(50 m×50 m)surrounded by a buffer strip(>20 m wide)were randomly established in an old broad-leaved Korean pine mixed forest in the Changbai Mountains during May 2009 in order to examine the effects of precipitation reduction on soil bacterial communities.Three plots were treated with 30%precipitation reduction,which was equivalent to a drought year’s precipitation rate,and the other three plots were not. As shown in Fig. S1(see Supplementary Material for Fig. S1), polycarbonate V-shaped translucent(95%transparency)panels were used to intercept throughfall, covering 30% of the plot area in order to intercept approximately 30%of the natural throughfall during the growing season.The panels were removed in winter (especially during the snowfall period) to allow snow to fall onto the forest floor and to avoid the effects of the uneven snow distribution on the soil system.The panels were fixed on an aluminum frame approximately 1 m above the soil surface to ensure normal airflow. In addition, the average annual atmospheric nitrogen deposition(as nitrate and ammonium,excluding nitrogen in snow)in the study area was approximately 2.3 g N m−2year−1(Lü and Tian,2007).To investigate the effect of nitrogen deposition increase on the soil bacterial communities,each of the six plots was split into two 25 m×50 m subplots using a PVC sheet inserted 50 cm deep into the ground to avoid nutrient mobilization between adjacent subplots.One of the subplots of each plot received additional nitrogen, and the other did not. The selected nitrogen addition level(5 g N m−2year−1)was approximately two times the mean annual atmosphere nitrogen deposition in the Changbai Mountains area.In each nitrogen deposition increase subplot,ammonium nitrate(NH4NO3)was weighed,mixed with 40 L deionized water,and sprayed with a backpack sprayer onto the forest floor monthly using six equal applications at the beginning of each month over the entire growing season(May to October),beginning in May 2009.The control plots(CK)simultaneously received 40 L of deionized water without nitrogen addition to avoid differences in water application.By October 2016,the total nitrogen application rate was 40 g m−2.

Soil sampling and analyses

The soil sampling schedule and corresponding climatic factors(e.g.,soil temperature,air temperature)are shown in Table I. The bacterial communities were evaluated in two extreme periods(winter and summer(i.e.,the growing season))and a strong fluctuation period(the FTC).We took measures once in winter(January,the coldest month)and once in the growing season(July).Moreover,the FTC was divided into three periods for measuring,the initial stage,mid-term,and termination of FTC,based on the snowpack thickness, soil temperature at 5 cm depth (T5cm), and air temperature (Ta), each of which had distinct biophysical characteristics.The initial stage of FTC(FTC_S)was defined as the period that started when the daily maximum Tawas above 0◦C(the start of snowmelt)and Tafluctuation caused status changes in water and soil. The mid-term of FTC(FTC_M)was defined as the period that started when the daily maximum Tawas above 0◦C,the daily minimum T5cmwas below 0◦C, and the snowpack had been completely melted (the end of snowmelt). The termination of FTC (FTC_E) was defined as the period that ends whenthe daily minimum T5cmwas above 0◦C (the end of soil freezing at 5 cm depth).During the five specified periods,six soil cores (diameter 5 cm) were randomly taken from the top layer(0—10 cm)of each subplot using a handheld auger and were pooled into one composite sample.All the visible extraneous materials(such as stones and roots)were removed by hand.Each composite soil sample was divided into two subsamples.The first one was sieved(2 mm)for the analyses of basic soil properties. The second one was immediately flash-frozen in liquid diatomic nitrogen(N2)and stored in cooling dry-ice boxes for transportation to the laboratory and then stored at −80◦C until it was processed for DNA extraction.

TABLE I Soil sampling schedule and climate factors at the time of sampling

The soil properties of all samples were measured following Yan et al.(2017).Soil pH was determined in a 1:2.5(soil:water,dry weight/volume)slurry using a pH meter(Sartorius PB-10,Germany).Soil total carbon(TC)and nitrogen(TN)were determined using an automated TC/TN analyzer(multi N/C 3100,Analytik Jena AG,Germany).Soil total phosphorus(TP)was measured by high-performance liquid chromatography after wet digestion with HClO4(Parkinson and Allen,1975).Dissolved organic carbon(DOC)and dissolved total nitrogen(DTN)were extracted from soil using distilled water (soil:water of 1:10, weight/volume) (Jones and Willett,2006)and determined using a Multi N/C 3100 Analyzer(multi N/C 3100,Analytik Jena AG,Germany).

DNA extraction,PCR,and sequencing

Total soil DNA was extracted from 0.25 g of soil using a DNeasy PowerSoil Kit(Mobio Laboratories,USA)according to the manufacturer’s protocol, and was stored at −20◦C until PCR amplification.The extracted DNA was confirmed by agarose gel electrophoresis.The ribosomal region(V3 and V4)of bacterial 16S rRNA genes was amplified with the primers 515F(5′-GTGCCAGCMGCCGCGGTAA-3′)and 806R (5′-GGACTACHVGGGTWTCTAAT-3′) (Zielinski et al., 2006). The purification of the PCR products was performed using the Agencourt AMPure XP(ThermoFisher Scientific, Finland) and sequencing was performed using Illumina Miseq according to the manufacturer’s protocol.The purified samples were analyzed using the Fragment Analyzer(Advanced Analytical,USA)and amplicons was sequenced on an Illumina MiSeq platform (MiSeq v3.1 software)at the Majorbio Bio-Pharm Technology,Shanghai,China.

Bioinformatics

The sequence data(.fastq)were analyzed using mothur version 1.39.5(https://mothur.org/wiki/Main_Page).The raw data files were assembled,and any sequences with ambiguous bases,two nucleotide mismatches or ambiguous characters in primer matching,or homopolymers longer than 8 bp or any without a minimum overlap of 50 bp were removed.We aligned the bacterial sequences to a SILVA reference alignment and removed erroneous reads.Then,the sequences were screened for chimeras using UCHIME(version 4.2.40)and the chimeric sequences were removed. Nonchimeric sequences were classified into bacterial taxa using the Naïve Bayesian Classifier.Unclassified sequences were removed.Sequences were clustered into operational taxonomic units(OTU)at 97%similarity using the nearest neighbor-joining method. The OTUs were assigned to bacterial functional groups using the FAPROTAX database(Louca et al.,2016).Functional profiles were generated by associating individual organisms with metabolic functions of particular ecological relevance,such as photoautotrophy and nitrate respiration,using an annotation database based on the FAPROTAX database(Louca et al.,2016).Then,by comparing our data with the database data,we explored the functional potential of the corresponding functional community.

The diversity indices of bacterial communities,the observed OTU richness,the complement of Simpson’s diversity index,and Simpson’s evenness index,were calculated for each sample using the mothur software. The bacterial Sørensen dissimilarity index(βSOR),the Simpson dissimilarity index(βSIM),and the nestedness-resultant dissimilarity index(βNES)were also calculated for each sample using the R software(version 3.2.5)(Baselga,2010).

Statistical analysis

The homogeneity of the variances was determined using Levene’s test, and the normal distribution was analyzed using the Kolmogorov-Smirnov test. The experiment was a split-plot design. Therefore, a general linear mixed model(GLMM)with Tukey’s multiple comparisons was used to evaluate the effects of the treatments and the sampling time on the soil properties,bacterial diversity indices,and bacterial functional groups. The sampling time, nitrogen deposition increase,precipitation reduction,nitrogen deposition increase×precipitation reduction,and treatment×sampling time were considered to be fixed factors, while blocks were used as random effects to determine the effects of nitrogen deposition increase,precipitation reduction,season,and their interaction on the soil properties,bacterial diversity indices,and bacterial functional groups.Nonmetric multidimensional scaling(NMDS)ordinations were performed to determine the total bacterial communities.Permutational multivariate analysis of variance(PERMANOVA)was performed to evaluate the effects of the treatments and sampling time on bacterial communities. The linear discriminant analysis(LDA)effect size method(LEfSe)was used to identify the significantly different abundant taxa of bacteria in all treatments or seasonal changes(http://huttenhower.sph.harvard.edu/galaxy/root?Tool_id=lefse_upload).When the anthropogenic treatments(nitrogen deposition increase and precipitation reduction)were used as the classes of the LEfSe analysis,the sampling times were used as subclasses of the LEfSe analysis,and vice versa.The threshold for the logarithmic LDA score was set to 2 (Segata et al., 2011). In addition to the LEfSe analysis,all statistical analyses were performed in R(version 3.2.5).

RESULTS

Effects of nitrogen deposition increase

Increase in nitrogen deposition increased the soil TN and DTN concentrations compared with those of CK,but the differences were not significant(Table II).Increase in nitrogen deposition significantly decreased the soil pH value and TP concentration compared with those of CK(Table II).Increase in nitrogen deposition tended to decrease bacterial diversity, but the difference was not significant compared with CK(Table III).Moreover,the Sørensen dissimilarity index and the Simpson dissimilarity index were lower in the nitrogen deposition increase treatment than in CK.

Increase in nitrogen deposition produced a significant(P <0.05)effect on the bacterial community composition(Fig.1).Compared with CK,increase in nitrogen deposition decreased the relative abundance of the Proteobacteria phylum by 6.99%,but it increased the relative abundance of the Acidobacteria,Chloroflexi,Verrucomicrobia,and Planctomycetes phyla by 2.25%,12.05%,12.88%,and 24.23%,respectively(Fig.2).The LEfSe is an algorithm for highdimensional biomarker discovery and provides an explanation that identifies genomic features (genes, pathways,or taxa)that characterize the differences between different treatments or different periods.Based on the LEfSe analysis at the genus level,the relative abundances of Oligoflexus,Gemmata, and Novosphingobium significantly increased with nitrogen deposition increase when compared with CK(Fig.3).

Nitrogen deposition increase had different effects on the different bacterial functional groups (Fig. 4). The relative abundances of cellulolysis,nitrite-oxidizing,and nitrification bacteria significantly increased when compared with CK,while the relative abundances of denitrification,nitrate-reducing,nitrate-respiring,N-fixing,N-respiring,and ureolysis bacteria significantly decreased.

Effects of precipitation reduction

The DTN concentrations increased,while the pH value decreased with precipitation reduction compared with CK(Table II).The DTC concentrations significantly increased in the precipitation reduction treatment when compared withCK(Table II).There was an increasing trend in all diversity indices with precipitation compared with CK(Table III).

TABLE II Soil physicochemical properties in different treatments in 2017 in the long-term experiment starting from 2009

TABLE III Bacterial diversity indices,Sørensen dissimilarity index(βSOR),Simpson dissimilarity index(βSIM),and nestedness-resultant dissimilarity index(βNES)in different treatments across seasons and in the different seasons in control(CK)in 2017 in the long-term experiment starting from 2009

Fig. 1 Nonmetric multidimensional scaling (NMDS) ordinations of bacterial community data by permutational multivarite analysis of variance(PERMANOVA)illustrating significant separation of the soil bacterial communities in different treatments across seasons(a)and in different seasons in control(CK)(b)(n=15 and 12 for treatment and season,respectively)in 2017 in the long-term experiment starting from 2009.More overlaps between different colors represent higher similarity.PR=precipitation reduction;NDI=nitrogen deposition increase.W=winter;FTC_S=the initial stage of the freezing-thawing cycle period(FTC);FTC_M=the mid-term of FTC;FTC_E=the termination of FTC;G=the growing season.

Fig.2 Relative abundances of the main soil bacterial phyla in different seasons in control(CK)(a)and in different treatments across seasons(b)(n=12 and 15 for season and treatment,respectively)in 2017 in the long-term experiment starting from 2009.CK=control;PR=precipitation reduction;NDI=nitrogen deposition increase;NP=nitrogen deposition increase and precipitation reduction.W=winter;FTC_S=the initial stage of the freezing-thawing cycle period(FTC);FTC_M=the mid-term of FTC;FTC_E=the termination of FTC;G=the growing season.

The precipitation reduction treatment also significantly(P <0.05) changed the bacterial communities (Fig. 1).Precipitation reduction increased the relative abundances of the phyla Chloroflexi(17.34%),Bacteroidetes(24.79%),and Planctomycetes(18.68%)compared with CK,but it decreased the relative abundances of the phyla Verrucomicrobia(16.62%)and Nitrospirae(21.35%)(Fig.2).At the genus level,the relative abundances of Undibacterium,Shimazuella,

Blastochloris,Alicyclobacillus,Cytophaga,Legionella,and Bdellovibrio significantly increased in the precipitation reduction treatment as compared with CK(Fig.3).

Regarding the bacterial functional groups, there were significant increases in the relative abundances of ammonification,cellulolytic,nitrification,and nitrite-oxidizing bacteria in the precipitation reduction treatment as compared with CK(Fig.4).The relative abundance of nitrogen-fixing bacteria significantly decreased in the precipitation reduction treatment when compared with CK(Fig.4).The reduction in precipitation did not significantly affect other bacterial functional groups.

Effects of nitrogen deposition increase and precipitation reduction combined

The combination of nitrogen deposition increase and precipitation reduction had a positive effect on the soil TN,DTN,and DTC contents(Table II).The combined nitrogen deposition increase and precipitation reduction significantly increased the bacterial diversity,richness,and evenness indices compared with those in CK(Table III).Likewise,the Sørensen dissimilarity index and Simpson dissimilarity index increased in the combined treatment when compared with those in CK, while the nestedness-resultant dissimilarity index decreased in the combined treatment.

The combined treatment significantly (P <0.05)changed the bacterial communities (Fig. 1). The relative abundances of the phyla Actinobacteria, Acidobacteria,Chloroflexi, Bacteroidetes, and Planctomycetes increased by 11.94%, 4.78%, 11.71%, 9.21%, and 19.51%, respectively, in the NP treatment as compared with CK, while the relative abundance of the Proteobacteria phylum decreased by 12.35%(Fig.2).In addition,at the genus level,the NP treatment significantly increased the relative abundances of Acidobacteria, Gaiella, OM190, Latescibacteria,Syntrophaceae,Vulgatibacter,Catecholicum,Defluviicoccus,Blastocatellaceae,and KD4-96 as compared with CK(Fig.3).

There were significant increases in the relative abundances of ammonification,aromatic-degrading,cellulolytic,and hydrocarbon-degrading bacteria in the NP treatment as compared with CK, while the relative abundance of nitrate-reducing bacteria significantly decreased in the NP treatment(Fig.4).

Seasonal effects

Sampling time had significant(P <0.05)effects on all the soil properties assayed.Bacterial diversity was the lowest in the growing season and highest in the winter(Table III).The bacterial richness also responded to seasonal changes,with the highest value occurring in the FTC_M period and the lowest in the FTC_S period. The nestedness-resultant dissimilarity index was the highest in the growing season and lowest in the FTC_S period.Moreover,the interactions between season and nitrogen deposition increase,between season and precipitation reduction, and between season,nitrogen deposition increase, and precipitation reduction had a significant effect on bacterial diversity and richness(Table IV).

The seasonal changes significantly changed the bacterial communities (Fig. 1, Tables V and VI). The interaction between season and treatment also significantly changed the bacterial communities (Table V). Specifically, at the genus level,the relative abundance of PeM15 significantly increased in the FTC_E period;the relative abundances of TK10, BIrii41, Phaselicystidaceae, Sphingorhabdus, and Solirubrobacterales significantly increased in the FTC_M period;the relative abundances of Solibacter,Bryobacter,and Methylobacter significantly increased in the FTC_S period; the relative abundances of H16, Gallionellaceae,and Rhodoblastus significantly increased in the growing season; and the relative abundances of Gemmatimonas,Jatrophihabitans,and Zixibacteria significantly increased in winter(Fig.3).

The denitrification, aromatic-degrading, cellulolytic,and ureolytic bacteria initially decreased and then increased with seasonal variation, with the lowest value occurring during the FTC_M period (Fig. 5). In contrast, nitriteoxidizing,nitrogen-fixing,and nitrification bacteria initially increased and then decreased with seasonal variation(Fig.5).The interaction between season and treatment significantlychanged the bacterial functional groups(Table VI).

TABLE V Results of permutational multivariate analysis of variance(PERMANOVA)on the effects of nitrogen deposition increase(NDI),precipitation reduction(PR),season,and their interactions on bacterial community composition in 2017 in the long-term experiment starting from 2009

TABLE VI Effects of nitrogen deposition increase(NDI),precipitation reduction(PR),season,and their interactions on relative abundances of bacterial functional groups

Fig.5 Relative abundances of bacterial functional groups,i.e.,plant pathogens(a),human pathogens(b),ammonia-oxidizing bacteria(c),ammonification bacteria(d), aromatic-degrading bacteria(e), cellulolytic bacteria(f), denitrification bacteria(g), hydrocarbon-degrading bacteria(h), nitrate-reducing bacteria(i),nitrate-respiring bacteria(j),nitrification bacteria(k),nitrite-oxidizing bacteria(l),nitrogen-fixing bacteria(m),nitrogen-respiring bacteria(n),and ureolytic bacteria(o),in different seasons in 2017 in the long-term experiment starting from 2009.Boxes show 25—75 percentiles,vertical lines show 10—90 percentiles,horizontal lines within boxes are median values(n=12),and points outside the boxes represent outliers.Different letters indicate significant(P <0.05)differences between seasons.W=winter;FTC_S=the initial stage of the freezing-thawing cycle period(FTC);FTC_M=the mid-term of FTC;FTC_E=the termination of FTC;G=the growing season.

DISCUSSION

Effects of nitrogen deposition increase

In the present study, the nitrogen deposition increase resulted in significant changes in the bacterial community composition,which is consistent with the results of previous studies(Zeng et al.,2016;Chen et al.,2018).Shifts in the bacterial composition following nitrogen deposition increase were previously explained by the copiotrophic hypothesis,in which copiotrophic bacterial groups(r-selected,e.g.,Proteobacteria)are more likely to increase under nutrient-rich conditions,while oligotrophic bacterial groups(K-selected,e.g.,Acidobacteria)would likely decline(Fierer et al.,2007;Koyama et al.,2014).However,in the present study,some copiotrophic organisms,such as Proteobacteria,decreased and some oligotrophic organisms, such as Acidobacteria,increased in abundances following nitrogen deposition increase, which is contrary to the copiotrophic hypothesis.These results indicate that the bacterial communities did not necessarily shift towards copiotrophic groups under longtime nitrogen enrichment conditions. Studies have found that short-time nitrogen deposition increase might stimulate copiotrophic groups, while long-time nitrogen deposition increase has negative effects on the bacterial community due to its side effects(such as soil acidification)(Treseder,2008;Ramirez et al.,2010;Yuan et al.,2016).In our study,the pH value was lower under nitrogen deposition increase than under control conditions.The optimal growth of soil bacterial communities has a narrow pH range,so the change in pH caused by nitrogen deposition increase may be an important factor affecting the bacterial community composition(Rousk et al., 2010; Zhang et al., 2017). Acidification makes the environment more restrictive,which may be the reason behind the change in the bacterial community.Previous studies showed that a low pH would lead to a decrease in the relative abundance of Proteobacteria and an increase in the relative abundance of Acidobacteria(Ramirez et al.,2010;Zhang et al.,2017),thus indicating that the shifts in the bacterial community composition could be largely attributed to the decrease in soil pH under nitrogen deposition increase in this study.

Changes in bacterial communities could result in altered rates and/or controls of the corresponding functions.In this study, we found that nitrogen deposition increase significantly increased the relative abundance of nitrification bacteria.A previous study of this experimental system found that the nitrogen deposition increase increased the N2O flux(Geng et al.,2017),indicating that the increased N2O flux might be caused by the increase in nitrification bacteria.The change in nitrification bacteria that we observed in response to nitrogen deposition increase may have resulted from the direct effect of NH4NO3application to the soil.The application of NH4NO3fertilizer would directly increase the concentration of soil ammonium,which may lead to an increase in the relative abundance of nitrification bacteria(Chen et al.,2013).However,nitrification produces hydrogen ions,and therefore the increase in nitrification bacteria abundance may explain why we observed a decrease in soil pH in response to nitrogen deposition inrease.Subsequently,the denitrification rate generally decreases with decreasing pH (Cavigelli and Robertson, 2000), which may support our results,which show that the relative abundance of denitrification bacteria significantly decreased with nitrogen deposition increase.Increases in nitrification bacteria and decreases in denitrification bacteria with nitrogen deposition increase would lead to soil nitrate accumulation.However,under nitrogen deposition increase,soil nitrate concentration did not increase significantly,and plant nitrate uptake did not increase significantly in the experimental system(Zhou et al.,2019).Although we did not have data regarding the response of nitrate leaching to nitrogen deposition increase,previous studies have reported increases in nitrate leaching with nitrogen deposition increase (Högberg et al., 2006;Fang et al.,2009).Thus,we speculate that the absence of an increase in soil nitrate concentration under nitrogen deposition increase could partly be attributed to the leachability of nitrates.

Effects of precipitation reduction

In this study,precipitation reduction significantly altered the bacterial community composition,which is consistent with previous studies that showed that bacterial communities changed significantly with throughfall exclusions(Bouskill et al., 2013; Felsmann et al., 2015). In the precipitation reduction treatment, the soil water potential declines, the pore connectivity is reduced,and the solute concentrations increase,imparting physiological and physical stresses on all the bacterial communities and significantly altering them(Chowdhury et al., 2011; Bouskill et al., 2013; Yao et al., 2017). A previous study in this experimental system found that precipitation reduction altered the nutrient uptake strategy(increasing ammonium uptake and decreasing nitrate)(Zhou et al.,2019),which may lead to changes in root exudates and, in turn, affect bacterial communities.Nevertheless,based on bacterial characteristics like high rate of growth and turnover,the precipitation reduction may shift bacterial communities by selecting some taxa with tolerance to perturbation(Angel et al.,2010).In support of this,our results showed increased relative abundances of some bacterial phyla(e.g.,Planctomycetes)that could be considered to be tolerant to the decline in soil moisture(Bouskill et al.,2013).The results are consistent with that of a previous study that showed an elevated abundance of Planctomycetes under throughfall exclusion(Bouskill et al.,2013),which indicates that Planctomycetes are sensitive to changes in soil moisture and maintain adaptive traits that allow the group to increase competitiveness under precipitation reduction(Bacher et al.,2012). However, in this study, some bacterial phyla (e.g.,Verrucomicrobia)showed the opposite response due to the inability to adapt to precipitation reduction. These phyla,such as Verrucomicrobia,display a water-related opportunistic behavior;thus,their abundance may rapidly decline under precipitation reduction treatments (Barnard et al., 2013).Moreover, Verrucomicrobia perform a fast-response lifestrategy,while Planctomycetes follow a resistant life-strategy(Ramirez et al., 2012), which may also explain why they respond differently to precipitation reduction.However,our study indicates that long-time precipitation reduction creates new dynamics for bacterial communities.

In this study, some nitrogen transformation bacteria(ammonification, nitrification, and nitrite-oxdizing bacteria) significantly increased under precipitation reduction,which indicated that water was an important factor influencing nitrogen cycling. Precipitation reduction may have a significantly positive effect on the relative abundance of nitrification and nitrite-oxidizing bacteria by increasing the accumulation of ammonium ions,solute concentrations,and the proportion of oxygen-filled soil pores(Stursova et al.,2006;Szukics et al.,2012;Chen et al.,2013).Nitrification and nitrite-oxidizing bacteria control the conversion of soil inorganic nitrogen to nitrate.Increases in their relative abundances may lead to soil nitrate accumulation,which would increase the risk of soil nitrogen leaching(Wang et al.,2019).Moreover,in the present study,significant decreases in the relative abundance of nitrogen-fixing bacteria may be caused by nitrate accumulation and increased solute concentrations.This result is consistent with that of a previous study that showed that the nitrogen-fixing rates would decrease in soils with high nitrogen availability(Reed et al.,2010).

Effects of nitrogen deposition increase and precipitation reduction combined

Consistent with our expectations,there were significant combined effects of nitrogen deposition increase and precipitation reduction on the soil bacterial community composition,which is consistent with a previous study (Nguyen et al.,2018).As mentioned above,soil water regulates diffusion and the mobility of nutrients.Thus,the effects of nitrogen deposition increase on the bacterial community composition depend on the soil water availability(Zhang et al.,2015).A previous study showed that higher precipitation buffered the effects of nitrogen deposition increase on bacterial communities by increasing the diffusion and mobility of nitrogen(Li et al., 2016). In contrast, precipitation reduction may aggravate the effects of nitrogen deposition increase on bacterial communities.In this study,the combined effects had a stronger effect on the relative abundance of Proteobacteria than nitrogen deposition increase alone,which supports the above hypothesis.However,the precipitation reduction also relieved the effects of nitrogen deposition increase on some bacterial phyla(e.g.,Acidobacteria),which indicated that precipitation reduction may reduce the effect of nitrogen deposition increase by increasing plant ammonium absorption and decreasing the risk of soil pH reduction in the experimental system(Zhou et al.,2019).In summary,our study found that the combined effects of precipitation reduction and nitrogen deposition increase on soil bacterial communities may not be general synergistic effects or antagonistic effects but should be multidimensional effects.

In addition, a previous study showed that most of the microbial functional groups significantly responded to nitrogen deposition increase only when there was a water stress(Zhang et al.,2015).This is contrary to our results,which showed that nitrogen deposition increase significantly affected the nitrogen transformation of bacterial groups under precipitation reduction conditions. Zhou et al. (2019)showed that nitrate and ammonium concentrations were significantly higher under the combined effects of nitrogen deposition increase and precipitation reduction than under control conditions or nitrogen deposition increase alone in this experimental system in July 2017, which indicates that precipitation reduction decreased the loss of available nitrogen leaching under nitrogen deposition increase.Simultaneously, precipitation reduction can alter the soil physical structure and plant physiological properties. All these factors may lead to increased soil nitrogen loss by increasing the relative abundance of ammonification bacteria and decreasing the relative abundance of N-fixing bacteria(Fig.4).These results indicated that the combined effects of precipitation reduction and nitrogen deposition increase played an essential role in nutrient status and soil properties,which,in turn,influenced bacterial functional groups.

Seasonal effects

The bacterial diversity was lower in the growing season and higher in the winter,which is consistent with the results of a previous study(Ladau et al.,2013).The growing season of low substrate availability(plant-microbe competition)may be associated with bacterial diversity reductions(Moreau et al.,2015).Moreover,the nestedness-resultant dissimilarity index was the highest in the growing season and lowest in the FTC periods, indicating that bacterial replacement is responsible for bacterial diversity in the growing season and bacterial turnover is responsible for bacterial diversity in the FTC periods. The results suggest that although the spring FTC may kill a large number of bacteria,it has a low impact on the bacterial community structure. During the growing season,the bacterial community is more affected by rhizodeposition, favoring copiotrophic organisms and reducing diversity,thereby increasing species replacement and altering the bacterial community structure(Preece and Peñuelas,2016).Furthermore,plant rhizodeposition might explain the increase in aromatic compound-related taxa during the growing season(Zhalnina et al.,2018).

In addition,seasonal changes also had a significant effect on the bacterial community composition,particularly an increase in Actinobacteria in winter and an increase in Proteobacteria in the growing season, which is consistent with the result of Kuffner et al. (2012). A previous study interpreted seasonal shifts of different phyla as reacting to fluctuations in the carbon substrate supply(Lipson,2007).Strickland et al.(2009)also proved the effects of the substrate quality and availability on soil bacterial communities.Thus,it is conceivable that the dynamics in the substrate supply may explain the seasonal shifts in the bacterial community composition observed in this study.

This shift in the bacterial community was characterized by the increase and decrease in relative abundances of certain bacteria, which led to altered bacterial functional groups(Fig.5).We found that the relative abundances of cellulolytic and ureolytic bacteria were the lowest during the FTCs,which indicates that FTCs might lead to a reduction in biodegradation by reducing the abundance of related bacteria.Previous studies have shown that FTCs can improve litter composition(Zhu et al.,2013).We speculate that the acceleration of litter decomposition in FTCs may not be due to biodegradation but to physical effects(mechanical fragmentation).Moreover,the physical effects of FTCs can breakdown soil aggregates,thereby increasing the availability of soil nitrogen(Song et al.,2017).Moreover,FTCs can also promote microbial cell destruction and stimulate the metabolism of surviving microbes via the internal release of carbon and nitrogen(Song et al.,2017).Furthermore,denitrification bacteria were reduced and nitrification bacteria were increased in FTCs when compared with the other periods(Fig.4),indicating that carbon and nitrogen are released by mechanical fragmentation and cell destruction could promote the conversion of soil nitrate. Sanders-DeMott et al. (2018) reported that FTCs reduced plant nitrogen uptake capacity and increased soil solution ammonium,which could also lead to an increase in the soil nitrate transformation.However,FTCs are often accompanied by high soil moisture, so increasing nitrate transformation may increase the risk of soil nitrogen loss(Risk et al.,2013).

Environments that have greater seasonal and temporal variation, particularly in temperate forests with long and cold winters,may regulate the effects of nitrogen deposition increase and precipitation reduction on soil bacterial communities because a wide range of physiological tolerances may already exist within the community(Cruz-Martínez et al.,2009).In this study,we found that seasonal changes and anthropogenic treatments had a significant interaction effect on bacterial communities.Likewise,nitrogen deposition increase and precipitation reduction may increase the seasonal variation severity of the bacterial community,thus resulting in new dynamics in the bacterial community composition(Waldrop and Firestone,2006).

CONCLUSIONS

Seasonal changes,precipitation reduction,and nitrogen deposition increase significantly affected the composition of the soil bacterial communities. Moreover, there were significant combined effects of nitrogen deposition increase and precipitation reduction on the soil bacterial community composition.Since changes in the bacterial community composition would result in altered corresponding functions,nitrogen deposition increase and precipitation reduction might increase the loss of soil nitrogen by decreasing denitrification bacterial groups and increasing ammonification bacterial groups.Furthermore,the interaction between treatment and season also significantly changed the soil bacterial communities and functions,indicating that the response of bacterial communities to nitrogen deposition increase and precipitation reduction was complex and could be highly dependent upon the underlying seasonal variability.

ACKNOWLEDGEMENT

This research was part of the project Global Change and Response which is supported by the National Key Research and Development Program of China(No.2016YFA0600800)and the National Natural Science Foundation of China(Nos.41773075,41575137,31370494,and 31170421).

SUPPLEMENTARY MATERIAL

Supplementary material for this article can be found in the online version.


登录APP查看全文