Short-Term Nitrogen Fertilization Affects Microbial CommunityComposition and Nitrogen Mineralization Functions in anAgricultural Soil

Yang Ouyang,a,b,c Jeanette M. Nortona

aDepartment of Plants, Soils, and Climate, Utah State University, Logan, Utah, USAbDepartment of Plant, Soil, and Microbial Sciences, Michigan State University, East Lansing, Michigan, USAcInstitute for Environmental Genomics, University of Oklahoma, Norman, Oklahoma, USA

ABSTRACT Soil extracellular enzymes play a significant role in the N mineralizationprocess. However, few studies have documented the linkage between enzyme activ-ity and the microbial community that performs the function. This study examinedthe effects of inorganic and organic N fertilization on soil microbial communitiesand their N mineralization functions over 4 years. Soils were collected from silagecorn field plots with four contrasting N treatments: control (no additional N), ammo-nium sulfate (AS; 100 and 200 kg of N ha�1), and compost (200 kg of N ha�1). Illu-mina amplicon sequencing was used to comprehensively assess the overall bacterialcommunity (16S rRNA genes), bacterial ureolytic community (ureC), and bacterial chi-tinolytic community (chiA). Selected genes involved in N mineralization were also ex-amined using quantitative real-time PCR and metagenomics. Enzymes (and markergenes) included protease (npr and sub), chitinase (chiA), urease (ureC), and arginase(rocF). Compost significantly increased diversity of overall bacterial communitieseven after one application, while ammonium fertilizers had no influence on theoverall bacterial communities over four seasons. Bacterial ureolytic and chitinolyticcommunities were significantly changed by N fertilization. Compost treatmentstrongly elevated soil enzyme activities after 4 years of repeated application. Func-tional gene abundances were not significantly affected by N treatments, and theywere not correlated with corresponding enzyme activities. N mineralization geneswere recovered from soil metagenomes based on a gene-targeted assembly. Under-standing how the structure and function of soil microbial communities involved withN mineralization change in response to fertilization practices may indicate suitableagricultural management practices that improve ecosystem services while reducingnegative environmental consequences.

IMPORTANCE Agricultural N management practices influence the enzymatic activi-ties involved in N mineralization. However, specific enzyme activities do not identifythe microbial species directly involved in the measured process, leaving the link be-tween the composition of the microbial community and the production of key en-zymes poorly understood. In this study, the application of high-throughput sequenc-ing, real-time PCR, and metagenomics shed light on how the abundance anddiversity of microorganisms involved in N mineralization respond to N management.We suggest that N fertilization has significantly changed bacterial ureolytic and chi-tinolytic communities.

KEYWORDS N fertilization, compost, microbial diversity, soil enzyme activity,N transformation rates, protease, urease, chiA, sub, npr, ureC, rocF, nitrogen cycleenzymes, nitrogen transformation rates

Citation Ouyang Y, Norton JM. 2020. Short-term nitrogen fertilization affects microbialcommunity composition and nitrogenmineralization functions in an agricultural soil.Appl Environ Microbiol 86:e02278-19. https://doi.org/10.1128/AEM.02278-19.

Editor Rebecca E. Parales, University ofCalifornia, Davis

Copyright © 2020 American Society forMicrobiology. All Rights Reserved.

Address correspondence to Jeanette M.Norton, jeanette.norton@usu.edu.

This article is journal paper 9157 from the UtahAgricultural Experiment Station, Utah StateUniversity.

Received 3 October 2019Accepted 11 December 2019

Accepted manuscript posted online 13December 2019Published

MICROBIAL ECOLOGY

crossm

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 1Applied and Environmental Microbiology

18 February 2020

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

Human input of chemical nitrogen (N) fertilizers to agricultural systems has in-creased more than 10-fold in the past 50 years to increase the yield of crops and

prevent food shortage for a growing human population (1). However, excessive andrepeated use of chemical N fertilizers may result in water and air pollution, soildegradation including reductions in soil organic matter and soil pH (2), and increasesin nitrate leaching and reactive N gas production (1, 3). Therefore, avoiding thecombination of high external N inputs with low N use efficiency remains a majorconcern for the sustainability of agroecosystems (4, 5). Application of organic Nfertilizers such as compost and manure is one effective strategy to improve soil qualityand functionality (6) while maintaining the N supply.

Soil microorganisms play a crucial role in the maintenance of soil fertility, and theyare often sensitive to N fertilization and management. Ammonium fertilizers contributea large but transient flush of inorganic N upon application, while organic N sourcesshow a slow inorganic release pattern due to N mineralization (7). Therefore, mineraland organic forms of N fertilization may exert different influences on soil microbialcommunities (2, 8, 9). Numerous field studies have shown that repeated mineral Nfertilization decreases bacterial diversity, while organic N fertilization increases bacterialdiversity (10–13). However, these studies were mainly conducted in long-term fieldfertilization experiments with only one sampling time and provided limited informationabout the temporal response of the soil microbial community in the field.

A wide variety of microbially derived extracellular enzymes mediate the depolymer-ization of the large N-containing polymers to monomers and ammonium (14). Mostprevious studies have focused on agricultural N management practices influencing theenzymatic activities involved in N mineralization (15–19). However, specific enzymeactivities do not identify the microbial species directly involved in the measuredprocess, leaving the link between the composition of the microbial community and theproduction of key enzymes poorly understood. For example, bacteria are assumed tobe the main degraders of urea and chitin (20, 21). It is still largely unknown how diverseare bacterial ureolytic and chitinolytic communities in soils and whether they areinfluenced by agricultural N management (22, 23). The development of real-time PCRand high-throughput sequencing could provide important information about how theabundance and diversity of microorganisms involved in N mineralization respond to Nmanagement (22–26).

Therefore, this study aimed to examine the short-term (�5 years) effects of inor-ganic and organic N management on soil microbial community composition, theabundance of functional genes involved in N mineralization, N transformation rates,and soil enzyme activities in replicated field plots. High-throughput sequencing ofmarker genes for bacterial ureolytic (ureC) and chitinolytic (chiA) communities was usedto identify specific urea and chitin degraders and examine their response to N fertil-ization. As a complementary resource, we also assembled several genes involved in Nmineralization from soil metagenomes. We asked whether soil microbial communityand soil enzyme activities would differentially respond to inorganic and organic formsof N fertilization. In addition, we also asked how N mineralization functions link with soilmicrobial communities in the context of contrasting N management practices. Under-standing how the structure and function of soil microbial communities change inresponse to different N fertilization practices is essential information for the selection ofsuitable agricultural management practices that improve the ecosystem services andreduce negative environmental consequences.

(This research was conducted by Y. Ouyang in partial fulfillment of the requirementsfor a Ph.D. from Utah State University, Logan, UT, 2016 [27].)

RESULTSSoil N transformation rates and enzyme activities. In August 2014, N treatments

significantly affected most measured soil N transformation rates and enzyme activitiesalthough there was no significant difference in the gross mineralization rate and grossammonium consumption rate (Table 1). Gross nitrification, net N mineralization, and

Ouyang and Norton Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 2

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

net nitrification had higher rates with ammonium sulfate (AS) and compost treatmentsthan the control, but these rates were not different between AS and compost treat-ments. Control and compost treatments produced higher gross nitrate consumptionrates than AS treatments. Compost treatment produced significantly higher soil respi-ration rates, dehydrogenase activity, acid phosphomonoesterase activity, and alkalinephosphomonoesterase activity than other treatments.

Both N treatments and sampling time showed significant effects on activity ofurease, arginase, protease and �-glucosaminidase (Fig. 1; see also Table S1 in thesupplemental material). Generally, these four enzymes had much higher activities incompost-treated soils than with the other treatments (Fig. 1). However, enzyme activ-ities showed different patterns in temporal variation over sampling time. Ureaseand arginase activities were lowest in July. Protease activity increased through thegrowing season and had higher activities in September and October. In contrast,�-glucosaminidase activity was relatively constant throughout the season.

Abundance of genes involved in N mineralization. N treatment and samplingtime showed no significant effect on the abundances of the four functional genes

TABLE 1 Soil N transformation rates and enzyme activities in August 2014

Parametera

Value for the parameter by treatmentb

Control AS100 AS200 Compost

GMR (mg N/kg/day) 1.32 1.83 1.40 1.40GACR (mg N/kg/day) 2.51 2.88 2.14 2.71GNR (mg N/kg/day) 0.50 A 0.67 AB 1.00 B 0.99 BGNCR (mg N/kg/day) 0.60 B 0.25 A 0.23 A 0.80 BNMR (mg N/kg/day) 0.30 A 0.46 B 0.50 B 0.52 BNNR (mg N/kg/day) 0.36 A 0.49 B 0.54 B 0.53 BRR (mg [CO2-C]/kg/day) 7.21 A 7.81 A 7.32 A 11.45 BDehydrogenase (mg TPF/kg/h) 2.35 A 2.50 A 3.01 AB 3.82 BAcid phosphomonoesterase (mg p-nitrophenol/kg/h) 39.90 A 42.95 A 51.76 AB 63.05 BAlkaline phosphomonoesterase (mg p-nitrophenol/kg/h) 150.95 A 158.78 A 178.12 AB 189.50 BaGMR, gross N mineralization rate; GACR, gross ammonium consumption rate; GNR, gross nitrification rate; GNCR, gross nitrate consumption rate; NMR, netmineralization rate; NNR, net nitrification rate; RR, respiration rate; TPF, triphenyl formazan; AS, ammonium sulfate.

bValues are means (n � 4). Different uppercase letters indicate significant differences among treatments (P � 0.05).

FIG 1 Soil enzyme activities across four N treatments in 2014. Error bars represent standard errors (n � 4). Different letters above the barsindicate a significant difference among N treatments in a specific month (P � 0.05), based on repeated-measures ANOVA.

Fertilization Affects Community and N Mineralization Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 3

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

related to N mineralization in 2014 (Fig. S1 and Table S2). The mean values of theabundances of sub, npr, chiA, and ureC were 3.8 � 107, 2.7 � 105, 2.2 � 108, and9.7 � 107 copies per gram of dry soil, respectively. Pearson correlation analysis indi-cated that there was no significant correlation between the abundances of thesefunctional genes and the corresponding enzyme activity (Table S3).

The abundance of ureC was repeatedly measured in August from 2011 to 2014.Repeated-measures analysis of variance (ANOVA) indicated that year (P � 0.01) but notN treatment (P � 0.43) significantly changed the abundance of ureC, with an increasefrom 3.2 � 107 copies per gram of dry soils in 2011 to 1.08 � 108 copies per gram of drysoils in 2014 (Fig. S2).

Bacterial community composition. In total, 1,944,732 high-quality 16S rRNA gene

reads were obtained for 32 samples in 2011 and 2014, with 17,152 operationaltaxonomic units (OTUs). Across all soil samples, we detected 45 distinct prokaryoticphyla, although only 12 bacterial phyla and 1 archaeal phylum were among the mostprevalent (�1%) (Fig. S3). Proteobacteria, Actinobacteria, Acidobacteria, Bacteroidetes,and Gemmatimonadetes were the five most abundant phyla, which comprised morethan 75% of the relative abundance of the bacterial community. Two-way ANOVAindicated that the relative abundances of Proteobacteria, Actinobacteria, and Acidobac-teria were significantly changed by N treatments (Fig. 2a and Table S4). Actinobacteriaand Acidobacteria abundances were decreased by compost in 2011, while Proteobac-teria abundance was increased by compost in 2014. The abundances of these dominantphyla of the prokaryotic community were significantly influenced by sampling time(Table S4). For example, Acidobacteria, Bacteroidetes, and Verrucomicrobia abundancesincreased, while Actinobacteria and Firmicutes abundances decreased from 2011 to2014.

N treatment, but not year, significantly influenced the alpha diversity of the pro-karyotic community (Fig. 3). Compost treatment increased Chao1, the observed OTUs,and Shannon diversity in both 2011 and 2014. Bacterial community structure asrevealed by weighted UniFrac distance grouped differently in 2011 versus that in 2014,and compost treatment was distinct from AS and control treatments in both years (Fig.2b). Two-way permutational multivariate analysis of variance (PERMANOVA) furtherconfirmed that bacterial community structure was significantly affected by year(P � 0.001) and N treatment (P � 0.008).

Based on log2-fold change of the relative abundances of OTUs, many OTUs re-sponded significantly to inorganic and organic N fertilization (Fig. 4). Most of theseresponsive OTUs, mainly from Proteobacteria and Bacteroidetes, were enriched by Nfertilization. In 2011, there were only a few OTUs that were responsive to inorganic Naddition, while abundances of 41 OTUs were changed by compost treatment. In 2014,there were 33, 51, and 78 responsive OTUs in soils treated with AS with 100 kg of Nha�1 (AS100) and AS with 200 kg of N ha�1 (AS200) and compost, respectively. Soilstreated with AS100 and AS200 shared half of their responsive OTUs, while less than 4%responsive OTUs were shared between AS and compost treatments (Fig. S4a). In thecompost treatment, only seven responsive OTUs were shared between 2011 and 2014(Fig. S4b).

Ureolytic community composition. In total, 875,995 high-quality ureC reads were

obtained for 16 soil samples in June 2014, with 8,550 OTUs (95% amino acid identitycutoff). Based on the nearest match to reference taxonomy, we detected 10 distinctprokaryotic phyla although only seven bacterial phyla were among the most prevalent(�1%) (Fig. 5a). The majority of sequences were assigned to Proteobacteria (72%) andActinobacteria (12%). Thaumarchaeota were also detected, but their relative abun-dances were very low (0.4%). There was no significant difference in the relativeabundances of phyla among N-treated soils. The ureolytic community composition asrevealed by weighted UniFrac distance matrices was significantly changed by N treat-ments (P � 0.003) (Fig. 5b). Pairwise comparison demonstrated that results from com-

Ouyang and Norton Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 4

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

post treatment were significantly different from those with AS100 (P � 0.026) andAS200 (P � 0.031) treatments.

Top 50 ureC OTUs were used for detailed phylogenetic analysis (Fig. 6), whichaccounted for 24% of the total sequences. Most of these top OTUs were assigned toProteobacteria with families from the orders Burkholderiales, Rhizobiales, and Myxococ-cales. OTU 455 was closely affiliated with Rhizobiales and was not grouped togetherwith other Proteobacteria taxa. OTU 88 and OTU 436 were affiliated with the familiesNitrospiraceae and Myxococcales, respectively. Interestingly, several of the most abun-dant OTUs, such as OTU 54, OTU 100, and OTU 456, had no very close match in thecurrent reference database. Among the top 50 OTUs, abundances of 15 OTUs weresignificantly changed by N treatments. For example, abundances of OTU 313 and OTU668 were increased by compost treatment. However, abundances of several OTUs, suchas OTU 483 and OTU 233, were significantly reduced by AS200 treatment.

Chitinolytic community composition. In total, 53,709 high-quality chiA reads wereobtained for 16 soil samples in June 2014, with 3,572 OTUs (95% amino acid identity

FIG 2 (a) Relative abundances of Proteobacteria, Acidobacteria, and Actinobacteria, which are significantly changed by Ntreatment. Error bars represent standard errors (n � 4). Different lowercase letters indicate significant differences amongN treatments in a specific year (P � 0.05). (b) Nonmetric multidimensional scaling (NMDS) ordination (stress � 0.1) of theweighted UniFrac distance for bacterial communities under four N treatments in both 2011 and 2014.

Fertilization Affects Community and N Mineralization Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 5

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

cutoff). Based on the nearest match to the reference taxonomy, most of the OTUs wereassigned to Actinobacteria and Proteobacteria (Fig. 5c). The relative abundance ofActinobacteria was lowest, but abundance of Proteobacteria was highest in compost-treated soils. The chitinolytic community composition as revealed by weighted UniFracdistance matrices was significantly changed by N treatments (P � 0.01) (Fig. 5d).Pairwise comparison demonstrated that results with the control treatment were sig-nificantly different from those with the AS100 (P � 0.029) and AS200 (P � 0.034)treatments.

Top 50 chiA OTUs were used for detailed phylogenetic analysis (Fig. 7), whichaccounted for 26% of the total sequences. Most of these top OTUs were assigned toStreptomyces (e.g., OTU 127 and OTU 265), Lentzea (e.g., OTU 70 and OTU 200), andActinoplanes (e.g., OTU 642) in the phylum Actinobacteria. OTU 155, the most abundantOTU, was closest to Xanthomonadaceae. Among the top 50 OTUs, abundances of 13OTUs were significantly changed by N treatments. For example, abundances of OTU

FIG 3 Alpha diversity of soil bacterial communities across N treatments. Error bars represent standard errors (n � 4). Differentlowercase letters indicate significant differences among treatments in a specific year (P � 0.05).

FIG 4 Log2-fold change in relative abundances of OTUs compared with those of the control treatmentin both 2011 and 2014. Each circle represents a single OTU with an adjusted P value of �0.1. Dashed anddotted lines indicate increases or decreases of 2-fold and 10-fold, respectively.

Ouyang and Norton Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 6

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

642, OTU 52, and OTU 42 were increased by compost treatment, while abundances ofOTU 155 and OTU 433 were significantly increased by AS treatments.

Gene-targeted assembly for N mineralization genes. Four of the selected Nmineralization genes (npr, chiA, rocF, and ureC) were recovered in our four soil metag-enomes, based on the gene-targeted assembly (Table S5); the sub gene was notrecovered by the same method. The number of OTUs recovered at 95% amino acididentity for these four selected N mineralization genes ranged from 4 to 280, and theirrplB normalized abundances ranged from 0.01 to 0.44 in four soil metagenomes. TherocF and ureC genes had higher OTU numbers and abundances than the other Nmineralization genes. The top 10 OTUs of these N mineralization genes and their bestmatches to reference databases are also summarized (Table S6). Top npr and chiA OTUrepresentatives often had relatively low similarity to reference sequences from Fun-Gene, with the average of 61%. Most of the top rocF OTUs were assigned to Acidobac-teria and Proteobacteria. Interestingly, most top ureC OTUs were assigned to Thaumar-chaeota.

Microbial community composition of steer waste compost. The bacterial com-munity composition of the steer waste compost in 2011, which was added to thecompost treatment, was also analyzed together with soils samples. There were six

FIG 5 (a) Relative abundances of the dominant phyla (�1%) for bacterial ureC. (b) Nonmetric multidimensional scaling (NMDS)ordination (stress � 0.09) of the weighted UniFrac distance for bacterial ureC under four N treatments. (c) Relative abundancesof the dominant phyla (�1%) for bacterial chiA. (d) Nonmetric multidimensional scaling (NMDS) ordination (stress � 0.05) ofthe weighted UniFrac distance for bacterial chiA under four N treatments.

Fertilization Affects Community and N Mineralization Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 7

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

prevalent phyla (�1%) in the steer waste compost (Fig. S5a), and Proteobacteria,Bacteroidetes, and Actinobacteria were the three most abundant phyla. Three monthsafter compost application, about 42% of OTUs from steer waste compost were detectedin the compost-treated soils. There were 20% of OTUs from steer waste compost thatwere detected only in compost-treated soil and not in the control soils; this accountsfor 6.6% of OTUs in compost-treated soils (Fig. S5b). For the OTUs shared only betweensteer waste compost and compost-treated soils, we found that 18 OTUs were alsopresent in the group of 41 responsive OTUs.

Ureolytic and chitinolytic community composition from steer waste compost used in2013 were also measured. Around 90% of ureC sequences were assigned to Proteobac-teria (Fig. S6). One year later, more than 50% of ureC OTUs from steer waste compostwere recovered in compost-treated soils, and 6.5% of ureC OTUs from compost-treatedsoils were present in both the steer waste compost and compost-treated soils. Exceptfor unclassified phyla, most of the chiA sequences in steer waste compost were

FIG 6 Maximum-likelihood tree of top 50 most abundant partial ureC OTUs. Different colors of branches and leaves indicate different phyla of ureC (color codingas shown in Fig. 5a). The heat map presents the relative abundances of OTUs of ureC among four N treatments in June 2014 (mean values; n � 4). Stars indicatea significant difference between values for controls and AS treatments, while circles indicate a significant difference between values for controls and composttreatment.

Ouyang and Norton Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 8

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

assigned into Actinobacteria. One year later, only 1.5% of chiA OTUs from compost-treated soils were present in both steer waste compost and compost-treated soils(Fig. S7).

DISCUSSION

In this study, the application of the organic N fertilizer (steer waste compost)significantly changed the structure of the bacterial community. This change wasdetected 3 months after the application. Compost strongly increased the richness anddiversity of the bacterial community. This observation is consistent with those of mostprevious studies (8, 9, 28–30). The elevated diversity of the bacterial community incompost-treated soils was partly due to the stimulation in the growth of native soilbacteria by high levels of available nutrients and diverse organic carbon fractions (6).In addition, we found that more than half of the OTUs from the steer waste compostwere recovered in compost-treated soils 3 months after application. The direct intro-

FIG 7 Maximum-likelihood tree of top 50 most abundant partial chiA OTUs. Different colors of branches and leaves indicate different phyla of chiA, as follows:purple, Actinobacteria; green, Proteobacteria; blue, Firmicutes; red, unclassified. The heat map presents the relative abundances of OTUs of chiA among four Ntreatments in June 2014 (mean values; n � 4). Stars indicate a significant difference between values for the control and AS treatments, while circles indicatea significant difference between values for the control and compost treatment.

Fertilization Affects Community and N Mineralization Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 9

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

duction of exogenous species to the soils may have contributed to the increasedmicrobial diversity (31) although the microbes originating from compost might be lesscompetitive with the native soil microbial community over the long term (8, 29).

Although previous studies reported that mineral N fertilization decreased the diver-sity of the bacterial community in agricultural soils (30, 32, 33), our results showed thatmineral N fertilization had no effect on the diversity and structure of the overallbacterial community. This observed stability may be related to the agricultural man-agement history. Our field plots were repeatedly planted with wheat and received ureaas an N source for the decade before 2011. This cultivation with crop monoculture andrepeated urea fertilization may homogenize the microbial community, favoring mi-crobes that are resistant to change due to mineral N fertilization (34).

We found that the overall bacterial community significantly varied between 2011and 2014. Soil microbial communities often show strong seasonal and interannualvariability (35). Duncan et al. (36) reported considerable changes in the bacterialcommunity over a 2-year period in fields planted with corn. Lauber et al. (37) found thatsoil community composition was variable over time in agricultural soils and that thechanges in communities were positively correlated with soil moisture and temperature.More interestingly, we found that the numbers of responsive OTUs increased signifi-cantly from 2011 to 2014. Most of these responsive OTUs were from Proteobacteria andBacteroides, which are generally considered copiotrophs (38). Furthermore, only a verysmall proportion of the OTUs was shared between 2011 and 2014. These resultssuggest that the temporal variability of the bacterial community was very high in theshort term after N fertilization.

The ureolytic community composition was significantly changed by N treatments inour soil. More specifically, inorganic N and organic N treatments harbored distinctureolytic community compositions. Other studies in agricultural soils have shown thatorganic matter from compost or manure may affect the soil ureolytic microbial com-munity (23). We also observed a significant difference in the amount of soil organic Cbetween AS and compost treatments in 2014 (39). In our study, we found that morethan 50% of OTUs from steer waste compost were present in compost-treated soilseven 3 months after application. These microorganisms inhabiting compost may alsoplay an important role in shaping the bacterial ureolytic community composition incompost-treated soils. We also found that abundances of 15 of the top 50 OTUs weresignificantly changed by the N treatments. Interestingly, five of those affected OTUswere enriched by compost application, while abundances of most of the affected OTUswere decreased by AS treatments. These results indicate that some ureolytic microor-ganisms may be repressed under ammonium-based fertilizers.

In this study, amplicon sequencing indicated that the ureolytic communities weremainly affiliated with Proteobacteria. This is consistent with the report of Collier et al.(21), which summarized that bacterial urease was most commonly found in Proteobac-teria. It is important to note that many ureC OTUs, even some top OTUs, had noidentified matches (�85% identity) to current references. This suggests that theprimer-based amplicon sequencing provides some information on previously uncul-tured ureolytic organisms. Interestingly, our gene-targeted assembly of soil metag-enomes showed that many of these recovered top ureC OTUs were affiliated withThaumarchaeota. We did detect ureC sequences closely related to Thaumarchaeota,based on amplicon sequencing, although their relative abundances were very low. Ourprevious studies found that ammonia-oxidizing archaea (AOA), which often containureC, were abundant in our soils (39, 40). The difference between metagenome andamplicon sequencing of ureC suggests a potential primer bias. However, both ampliconsequencing and the metagenome confirmed that Nitrospira were important potentialurease producers since several top ureC OTUs were from Nitrospira, some of which wereeven close matches to the complete ammonia-oxidizing (comammox) organism Nitro-spira inopinata (41).

Chitin is one of the most abundant organic N polymers in soil environments (42).Previous studies have found that bacterial chitinolytic communities were significantly

Ouyang and Norton Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 10

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

changed by chitin amendment (22). Since compost contains multiple organic N poly-mers, we hypothesized that compost application would also shape the chitinolyticcommunity. We did observe that compost treatment significantly increased abun-dances of several top chiA OTUs (Fig. 7). However, there was no significant differencein the composition of the chitinolytic community between control and composttreatment after 4 years of repeated compost application. This is partly due to the highvariability of chitinolytic community composition in compost treatment. Three of thereplicates in the compost-treated soils were highly separated from the control treat-ments, but one of the compost-treated plots was closely clustered with controls. Inaddition, we found that only 1.5% of chiA OTUs in steer waste compost were recoveredin compost-treated soil, indicating that chitinolytic microorganisms in compost are lesscompetitive than the indigenous chitinolytic community and weakly survive in soil.Interestingly, AS treatments significantly changed the chitinolytic community compo-sition. Several top chiA OTUs were strongly enriched by ammonium-based fertilizers(OTU 155 and OTU 433). These results indicate that the chitinolytic community in oursoil may be N limited.

In our study, the application of compost increased soil enzyme activities, which isconsistent with observations showing that organic amendments significantly increaseenzyme activities (15, 43, 44). This is possibly due to stimulation of microbial growthand related increases in the activity of the extracellular enzyme-organo complexes (45).However, we found that the abundances of functional genes involved in N mineral-ization were not affected by the N treatments and that the abundances of functionalgenes involved in N mineralization did not correlate with their corresponding soilenzyme activities. The lack of correlation between the abundances of functional genesand their corresponding enzyme activities may be attributed to several factors. First,primers used to target the functional genes did not cover all members of the microbialcommunity responsible for the specific enzyme function. For example, there are manydifferent groups of proteases (46, 47), but only limited proteolytic gene primers havebeen developed and identified for soil microbiome, including serine peptidase (sub)and neutral metallopeptidase (npr) (48). In addition, the primer pairs used in our studydid not cover the fungal community. Metagenomic analysis for the both prokaryoticand fungal communities may provide better coverage for the functional groups pro-ducing the specific enzymes although the depth of sequencing remains an issue.Second, DNA-based analyses do not differentiate inactive from active members of thesoil community. Proteomic or RNA-based techniques may be more appropriate to linkthe abundances of active functional groups with their corresponding enzyme functions(45). Third, production of extracellular enzymes is regulated by genes encoding thecorresponding enzyme, but once they are secreted out of the cells, their stabilizationand degradation are controlled by physical and chemical conditions of the environ-ment (16). Fuka et al. (25) reported that a significant correlation between sub and nprgenes and a potential protease was found only for sandy soils and not clay soil,suggesting that these relations may be soil specific.

In summary, the application of organic N fertilizer, but not inorganic N fertilizer,increased the diversity of the bacterial community and the activities of soil enzymes. Nfertilization significantly changed ureolytic and chitinolytic bacterial communities. Theabundances of selected functional genes involved in N mineralization were notaffected by the N treatments, regardless of the form of the inorganic and organicfertilizer used. The abundances of targeted functional genes were not correlatedwith the corresponding enzyme activities. Metagenomics or metatranscriptomicsanalysis associated with high-throughput sequencing targeting functional genesincluding those from fungi is needed to provide better coverage for the novelresponsible members of the microbial community. With this additional information,our ability to link microbial functional genes to their associated enzyme activityshould be strengthened.

Fertilization Affects Community and N Mineralization Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 11

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

MATERIALS AND METHODSSoil characterization. The details of the agricultural site (North Logan, UT, USA), experimental

design, treatments, soil sampling, and soil characteristics have been previously described (39, 40). Briefly,the experimental design is a randomized complete block with four blocks and four nitrogen treatments:control (no N fertilization), ammonium sulfate (AS at 100 and 200 kg of N ha�1, termed AS100 and AS200,respectively), and steer waste compost (200 kg of total N ha�1). Treatments were surface applied in Mayof each year and incorporated by tilling immediately after application. The soil is an irrigated, verystrongly calcareous Millville silt loam (coarse-silty, carbonatic, mesic Typic Haploxeroll). Soils weresampled in August from 2011 to 2014, and soils were also sampled monthly during the growing seasonof 2014. Six soil cores (0- to 15-cm depth; three cores in the intervals between rows and three cores inthe row between plants) were taken from each plot, composited, and thoroughly mixed, and a sampleof soil was stored at �80°C immediately after soils were brought to the laboratory.

Gross and net N transformation rates. Gross N transformation rates were determined in laboratoryincubations using a 15N pool dilution for soil sampled in August 2014. Three well-mixed 40-g dry-weight-equivalent subsamples were weighed into plastic specimen cups. Then, 1.6 ml of 15NH4

� solution[containing 1.69 mM (15NH4)2SO4 at 98 atom% 15N] or 15NO3

� solution (containing 3.33 mM K15NO3 at 99atom% 15N) was added to the soils and carefully mixed, creating a final soil water content of 0.18 kg kg�1.The quantity of 15N added approximately doubled the soil NH4

� or NO3� pool. Immediately following soil

mixing, one subsample was harvested and extracted with 2 M KCl to determine NH4� or NO3

�

concentration and 15N enrichment at time 0. The other subsample was placed in 1-liter Mason jars withlids containing butyl rubber septa and with 1 ml of water at the bottom of the jar to minimize loss ofmoisture from the soil. Jars were incubated for 48 h at 25°C before soil was extracted with 2 M KCl. SoilNH4

� or NO2� plus NO3

� was measured with a flow injection analyzer. The extracts were prepared for15N analyses using a diffusion procedure described by Stark and Hart (49), and the 15N enrichment wasmeasured by continuous-flow direct dry combustion and mass spectrometry with an ANCA 2020 system(Europa Scientific, Cincinnati, OH). Gross N transformation rates were calculated using the equation ofNorton and Stark (50).

Net mineralization and nitrification were determined by a 21-day incubation. Fifteen grams of moistsoil (0.18 kg kg�1 water content) in a plastic specimen container was placed in a 1-liter Mason jar witha lid containing butyl rubber septum and 1 ml of water at the bottom. Soil was extracted with 2 M KClbefore and after incubation. Headspace CO2 was measured at 3 days, 7 days, 14 days, and 21 days by agas chromatograph with a thermal conductivity detector to determine the soil respiration rate.

Soil enzyme activities. We measured the activities of protease (EC 3.4.21), arginase (EC 3.5.3.1),urease (EC 3.5.1.5), and �-glucosaminidase (EC 3.21.30), dehydrogenase (EC 1.1.1), acid phosphomonoes-terase (EC 3.1.3.2), and alkaline phosphomonoesterase (EC 3.1.3.1). The details of the protocol formeasurement of these enzyme activities have been previously described (26). Briefly, for protease assay,soil samples were incubated at 37°C with 0.6% casein. Protease activity was calculated from the differencebetween amino acid concentrations over 2 h. The arginase activity was measured as reported by Bonde et al.(51). Soil slurries were incubated with final concentration of 1.0 mM L-arginine at 37°C for 1 h. Urease activitywas determined as shown by Gianfreda et al. (52). Fresh soil was incubated at 37°C with 0.2 M urea solutionfor 2 h. �-Glucosaminidase activity was determined by the method of Parham and Deng (53). Fresh soil wasmixed with sodium acetate buffer (pH 5.5) and p-nitrophenyl-N-acetyl-�-D-glucosaminide solution in 50-mlcentrifuge tubes and kept at 37°C for 1 h. Activities of dehydrogenase, acid phosphomonoesterase, andalkaline phosphomonoesterase were measured at 37°C as previously described (54).

Soil DNA extraction and real-time quantitative PCR. Soil DNA was extracted using an MoBioPowerSoil DNA isolation kit (MoBio Laboratories Inc, Carlsbad, CA). DNA extracts were quantified by usinga Quant-iT PicoGreen dsDNA BR assay kit (Molecular Probes, Inc., Eugene, OR) according to themanufacturer’s protocol. Quantitative PCR of genes encoding enzymes involved in soil N mineralizationwas performed using SsoAdvanced SYBR green Supermix and a CFX Connect real-time PCR detectionsystem (Bio-Rad Laboratories, Hercules, CA). We measured the abundances of genes encoding subtilisin(sub), neutral metalloprotease (npr), chitinase (chiA), and urease (ureC) for soil sampled in June andAugust 2014. Primers, amplification conditions, efficiencies, and calibration standards are summarized inTable 2. Standard curves were constructed with plasmids containing cloned gene products from genomicDNA of bacterial isolates (ureC, sub, and chiA) or from environmental DNA (npr), and R2 values rangedfrom 0.990 to 0.999 for all genes targeted. Duplicate assays for each gene and calibration standard serieswere measured in a single run.

Soil metagenome processing and gene-targeted assembly. Metagenomes were also obtainedfrom soil samples in June 2014. DNA samples from four replicates of each N treatment were pooled with

TABLE 2 Real-time PCR amplification conditions, efficiencies, calibration standard, and primers

Targetgene

Primerconcn(�M)

Size(bp)

No. ofcycles

Length of PCR step (s)

Efficiency(%)

Calibrationstandard

Primers(reference)

Denaturationat 95°C

Annealing(°C)

Elongationat 72°C

npr 0.75 233 40 20 30 (55) 30 108 Environmental DNA clone Fp nprI, Rp nprII (48)sub 0.75 319 40 20 30 (55) 30 88 Bacillus subtilis ATCC 6051 Fp subIa, RP subII (48)chiA 0.5 417 40 60 60 (55) 60 110 Stenotrophomonas rhizophila

ATCC BAA 473GA1F, GA1R (70)

ureC 0.5 317 35 60 60 (60) 120 92 Pseudomonas chlororaphis O6 ureC1F, ureC2R (71)

Ouyang and Norton Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 12

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

an equal amount of DNA. DNAs were then sequenced on an Illumina HiSeq 2500 platform with a2- by 150-bp paired-end format at the Joint Genome Institute. Quality-filtered metagenomes weredownloaded and used for gene-targeted assembly (55). Five genes involved in N mineralization (sub, npr,chiA, ureC, and rocF) and rplB were included for the assembly. For each gene of interest, seed sequences,hidden Markov models (HMMs), and nucleotide and protein reference sequences were downloaded fromFunGene (56). Default assembly parameters were used, and sequences were clustered at 95% amino acidsimilarity. A representative sequence from each cluster was searched against the reference genedatabase and the nonredundant database (nr) from NCBI using BLAST (57). Overall, the top hit of theserepresentative sequences to the reference gene database had a similarity higher than 49% and an E valuehigher than 1.5E�46.

Illumina sequencing and data analysis for ureC and chiA. Sequencing of the ureC and chiAamplicon libraries was accomplished for steer waste compost used in 2013 and soils sampled in June2014. The same ureC and chiA primers described above were used for high-throughput sequencing.Linkers were added to the primers for the ureC and chiA genes, while tags were added to separatedifferent soil samples (58). The same amounts of soil DNA were used for ureC and chiA amplifications, andthen the PCR products were further purified using size selection (Agencourt Ampure XP PCR purification).Pooled purified products were sequenced on an Illumina MiSeq platform (Illumina, Inc., San Diego, CA)using V3 chemistry (2- by 300-bp paired-end reads).

For a comparison, the bioinformatic analysis of ureC and chiA amplicons was performed using thesame Xander postassembly processes based on the Ribosomal Database Project (RDP) pipeline (59). Rawreads were split based on the tags (soil samples), and then forward and reverse reads were merged usingthe USEARCH workflow (60). High-quality ureC and chiA sequences were extracted from merged reads ineach sample using the RDP SeqFilters with a read Q score cutoff of 25. Chimera sequences were removedusing UCHIME (61) with the ureC and chiA nucleotide reference databases downloaded from theFunGene (56). The obtained sequences were further processed using the FrameBot tool (62) to fix frameshifts and translate DNA to protein. The remaining quality-screened protein sequences in each samplewere aligned based on ureC and chiA hidden Markov models using HMMER3 (63). The aligned sequencesfrom each sample were merged together using the RDP AlignmentTools. Sequences were furtherdereplicated, and singletons were removed. Operational taxonomic units (OTU) were clustered at 95%amino acid similarity using the RDP Clustering tool. The longest sequence from each OTU was chosen asa representative sequence. To obtain the phylum-level classification of representative sequences, thetaxonomy from the closest match to the protein reference downloaded from FunGene was used. If thepercent identity to the reference sequences was less than 80% (percent alignment of �90%), we definedthe phylum as unclassified. A maximum-likelihood phylogenetic tree was constructed from representa-tive sequences using FastTree with default parameters (64). An OTU table and taxonomy file were furtherorganized for diversity analysis using the R package phyloseq (65).

Illumina sequencing of 16S rRNA. The variable V4 region of the 16S gene was amplified with 515Fand 816R universal primers for the bacterial community (66). The 16S amplicon sequencing wasperformed on an Illumina MiSeq instrument (Illumina Inc., San Diego, CA). The Illumina raw reads wereprocessed using a custom pipeline developed at the Joint Genome Institute (https://bitbucket.org/berkeleylab/jgi_itagger). Briefly, raw reads were first quality filtered, and then the high-quality sequenceswere clustered into operational taxonomic units (OTUs) based on 97% identity for a prokaryotic data setusing the USEARCH pipeline (60). Taxonomies were assigned to each OTU using the RDP Classifier witha confidence threshold of 0.60 (67). All data files were then organized using the R package phyloseq (65).

Statistical analysis. For microbial diversity analyses, all samples were randomly rarefied to lowestnumber of reads per sample (27,366 reads for the 16S gene, 25,000 reads for ureC, and 1,278 reads forchiA) to compare differences between samples. Alpha diversity and beta diversity were then calculated.Nonmetric multidimensional scaling (NMDS) and PERMANOVA were conducted to visualize and assessthe distance matrices using the R package vegan. Fold change in relative abundances of OTUs under Nfertilization were determined using the R package DESeq2 (68). We removed OTUs that were sparselyrepresented across samples (baseMean of �1.7) and adjusted the P values with the Benjamini-Hochbergcorrelation method (68, 69).

Statistical analysis for seasonal dynamics of soil enzyme activities was analyzed using repeated-measures analysis of variance (ANOVA) with a PROC Mixed model. Treatment and year were used as fixedeffects, and block was used as a random effect. Data were log transformed as necessary to meetnormality assumptions. Two-way ANOVA was used to analyze the effects of treatment and time onfunctional gene abundances and alpha diversity of prokaryotic communities. One-way ANOVA followedby Tukey’s honestly significant differences (HSD) was performed to compare soil biological propertiesmeasured in August 2014. Pearson correlation coefficients were determined for the relationshipsbetween functional gene abundances and enzyme activities. ANOVA and Pearson correlation werecarried out with SAS, version 9.2, software (SAS Institute, Inc., Cary, NC).

Data availability. Illumina sequence data for ureC and chiA were deposited in NCBI under BioProjectaccession number PRJNA597781. Illumina sequence data for 16S rRNA are available under JGI project ID1061777. The following soil metagenomes are available in NCBI BioProject (accession numbers inparentheses): control soil (PRJNA375186), AS100 soil (PRJNA375187), AS200 soil (PRJNA375188), andcompost soil (PRJNA375189).

SUPPLEMENTAL MATERIALSupplemental material is available online only.SUPPLEMENTAL FILE 1, PDF file, 1.4 MB.

Fertilization Affects Community and N Mineralization Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 13

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

ACKNOWLEDGMENTSWe thank Cory Ortiz, Marlen Craig Rice, Henry Linford, and Lili Song for laboratory

and field assistance.This work was supported by the USDA National Institute of Food and Agriculture

under awards 2011-67019-30178 and 2016-35100-25091. The research was supportedby the Utah Agricultural Experiment Station, Utah State University and approved asjournal paper number 9157. Some of the work, including 16S Illumina sequencing andmetagenomics, was conducted by the U.S. Department of Energy Joint GenomeInstitute, a DOE Office of Science User Facility, and was supported by the Office ofScience of the U.S. Department of Energy under contract DE-AC02-05CH11231.

We declare that we have no conflicts of interest.

REFERENCES1. Robertson GP, Vitousek PM. 2009. Nitrogen in agriculture: balancing the

cost of an essential resource. Annu Rev Environ Resour 34:97–125.https://doi.org/10.1146/annurev.environ.032108.105046.

2. Geisseler D, Scow KM. 2014. Long-term effects of mineral fertilizers onsoil microorganisms—a review. Soil Biol Biochem 75:54 – 63. https://doi.org/10.1016/j.soilbio.2014.03.023.

3. Fowler D, Coyle M, Skiba U, Sutton MA, Cape JN, Reis S, Sheppard LJ,Jenkins A, Grizzetti B, Galloway JN, Vitousek P, Leach A, Bouwman AF,Butterbach-Bahl K, Dentener F, Stevenson D, Amann M, Voss M. 2013.The global nitrogen cycle in the twenty-first century. Philos Trans R SocLond B Biol Sci 368:20130164. https://doi.org/10.1098/rstb.2013.0164.

4. Spiertz J. 2010. Nitrogen, sustainable agriculture and food security. A review.Agron Sustain Dev 30:43–55. https://doi.org/10.1051/agro:2008064.

5. Tilman D, Cassman KG, Matson PA, Naylor R, Polasky S. 2002. Agriculturalsustainability and intensive production practices. Nature 418:671– 677.https://doi.org/10.1038/nature01014.

6. Diacono M, Montemurro F. 2010. Long-term effects of organic amend-ments on soil fertility. A review. Agron Sustain Dev 30:401– 422. https://doi.org/10.1051/agro/2009040.

7. Habteselassie MY, Stark JM, Miller BE, Thacker SG, Norton JM. 2006. Grossnitrogen transformations in an agricultural soil after repeated dairy-waste application. Soil Sci Soc Am J 70:1338 –1348. https://doi.org/10.2136/sssaj2005.0190.

8. Sun R, Zhang X-X, Guo X, Wang D, Chu H. 2015. Bacterial diversity in soilssubjected to long-term chemical fertilization can be more stably main-tained with the addition of livestock manure than wheat straw. Soil BiolBiochem 88:9 –18. https://doi.org/10.1016/j.soilbio.2015.05.007.

9. Marschner P, Kandeler E, Marschner B. 2003. Structure and function of thesoil microbial community in a long-term fertilizer experiment. Soil BiolBiochem 35:453–461. https://doi.org/10.1016/S0038-0717(02)00297-3.

10. van der Bom F, Nunes I, Raymond NS, Hansen V, Bonnichsen L, Magid J,Nybroe O, Jensen LS. 2018. Long-term fertilisation form, level and du-ration affect the diversity, structure and functioning of soil microbialcommunities in the field. Soil Biol Biochem 122:91–103. https://doi.org/10.1016/j.soilbio.2018.04.003.

11. Yeoh YK, Paungfoo-Lonhienne C, Dennis PG, Robinson N, Ragan MA,Schmidt S, Hugenholtz P. 2016. The core root microbiome of sugarcanescultivated under varying nitrogen fertilizer application. Environ Micro-biol 18:1338 –1351. https://doi.org/10.1111/1462-2920.12925.

12. Li F, Chen L, Zhang J, Yin J, Huang S. 2017. Bacterial community structureafter long-term organic and inorganic fertilization reveals importantassociations between soil nutrients and specific taxa involved in nutrienttransformations. Front Microbiol 8:187. https://doi.org/10.3389/fmicb.2017.00187.

13. Wang J, Song Y, Ma T, Raza W, Li J, Howland JG, Huang Q, Shen Q. 2017.Impacts of inorganic and organic fertilization treatments on bacterialand fungal communities in a paddy soil. Appl Soil Ecol 112:42–50.https://doi.org/10.1016/j.apsoil.2017.01.005.

14. Schimel JP, Bennett J. 2004. Nitrogen mineralization: challenges of a chang-ing paradigm. Ecology 85:591–602. https://doi.org/10.1890/03-8002.

15. Bowles TM, Acosta-Martínez V, Calderón F, Jackson LE. 2014. Soil enzymeactivities, microbial communities, and carbon and nitrogen availability inorganic agroecosystems across an intensively-managed agriculturallandscape. Soil Biol Biochem 68:252–262. https://doi.org/10.1016/j.soilbio.2013.10.004.

16. Burns RG, DeForest JL, Marxsen J, Sinsabaugh RL, Stromberger ME,Wallenstein MD, Weintraub MN, Zoppini A. 2013. Soil enzymes in achanging environment: current knowledge and future directions. SoilBiol Biochem 58:216 –234. https://doi.org/10.1016/j.soilbio.2012.11.009.

17. Jian S, Li J, Chen J, Wang G, Mayes MA, Dzantor KE, Hui D, Luo Y. 2016.Soil extracellular enzyme activities, soil carbon and nitrogen storageunder nitrogen fertilization: a meta-analysis. Soil Biol Biochem 101:32– 43. https://doi.org/10.1016/j.soilbio.2016.07.003.

18. Bandick AK, Dick RP. 1999. Field management effects on soil enzymeactivities. Soil Biol Biochem 31:1471–1479. https://doi.org/10.1016/S0038-0717(99)00051-6.

19. Fernandez AL, Sheaffer CC, Wyse DL, Staley C, Gould TJ, Sadowsky MJ.2016. Associations between soil bacterial community structure andnutrient cycling functions in long-term organic farm soils followingcover crop and organic fertilizer amendment. Sci Total Environ 566:949 –959. https://doi.org/10.1016/j.scitotenv.2016.05.073.

20. Bai Y, Eijsink VGH, Kielak AM, van Veen JA, de Boer W. 2016. Genomiccomparison of chitinolytic enzyme systems from terrestrial andaquatic bacteria. Environ Microbiol 18:38 – 49. https://doi.org/10.1111/1462-2920.12545.

21. Collier JL, Baker KM, Bell SL. 2009. Diversity of urea-degrading microorgan-isms in open-ocean and estuarine planktonic communities. Environ Micro-biol 11:3118–3131. https://doi.org/10.1111/j.1462-2920.2009.02016.x.

22. Kielak AM, Cretoiu MS, Semenov AV, Sørensen SJ, van Elsas JD. 2013.Bacterial chitinolytic communities respond to chitin and pH alteration insoil. Appl Environ Microbiol 79:263–272. https://doi.org/10.1128/AEM.02546-12.

23. Wang L, Luo X, Liao H, Chen W, Wei D, Cai P, Huang Q. 2018. Ureolyticmicrobial community is modulated by fertilization regimes and particle-size fractions in a black soil of northeastern China. Soil Biol Biochem116:171–178. https://doi.org/10.1016/j.soilbio.2017.10.012.

24. Baraniya D, Puglisi E, Ceccherini MT, Pietramellara G, Giagnoni L, ArenellaM, Nannipieri P, Renella G. 2016. Protease encoding microbial commu-nities and protease activity of the rhizosphere and bulk soils of twomaize lines with different N uptake efficiency. Soil Biol Biochem 96:176 –179. https://doi.org/10.1016/j.soilbio.2016.02.001.

25. Fuka MM, Engel M, Gattinger A, Bausenwein U, Sommer M, Munch JC,Schloter M. 2008. Factors influencing variability of proteolytic genes andactivities in arable soils. Soil Biol Biochem 40:1646 –1653. https://doi.org/10.1016/j.soilbio.2008.01.028.

26. Ouyang Y, Reeve JR, Norton JM. 2018. Soil enzyme activities and abun-dance of microbial functional genes involved in nitrogen transforma-tions in an organic farming system. Biol Fertil Soils 54:437– 450. https://doi.org/10.1007/s00374-018-1272-y.

27. Ouyang Y. 2016. Agricultural N management affects microbial commu-nities, enzyme activities, and functional genes for nitrification and nitro-gen mineralization. PhD dissertation. Utah State University, Logan, UT.

28. Wang J, Xue C, Song Y, Wang L, Huang Q, Shen Q. 2016. Wheat and ricegrowth stages and fertilization regimes alter soil bacterial communitystructure but not diversity. Front Microbiol 7:1207. https://doi.org/10.3389/fmicb.2016.01207.

29. Hamm AC, Tenuta M, Krause DO, Ominski KH, Tkachuk VL, Flaten DN.2016. Bacterial communities of an agricultural soil amended with solidpig and dairy manures, and urea fertilizer. Appl Soil Ecol 103:61–71.https://doi.org/10.1016/j.apsoil.2016.02.015.

Ouyang and Norton Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 14

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from

30. Chaudhry V, Rehman A, Mishra A, Chauhan PS, Nautiyal CS. 2012.Changes in bacterial community structure of agricultural land due tolong-term organic and chemical amendments. Microb Ecol 64:450 – 460.https://doi.org/10.1007/s00248-012-0025-y.

31. Semenov AV, van Overbeek L, van Bruggen A. 2009. Percolation andsurvival of Escherichia coli O157:H7 and Salmonella enterica serovar Typhi-murium in soil amended with contaminated dairy manure or slurry. ApplEnviron Microbiol 75:3206–3215. https://doi.org/10.1128/AEM.01791-08.

32. Ding J, Jiang X, Ma M, Zhou B, Guan D, Zhao B, Zhou J, Cao F, Li L, Li J. 2016.Effect of 35 years inorganic fertilizer and manure amendment on structureof bacterial and archaeal communities in black soil of northeast China. ApplSoil Ecol 105:187–195. https://doi.org/10.1016/j.apsoil.2016.04.010.

33. Ramirez KS, Craine JM, Fierer N. 2010. Nitrogen fertilization inhibits soilmicrobial respiration regardless of the form of nitrogen applied. Soil BiolBiochem 42:2336 –2338. https://doi.org/10.1016/j.soilbio.2010.08.032.

34. Griffiths BS, Philippot L. 2013. Insights into the resistance and resilienceof the soil microbial community. FEMS Microbiol Rev 37:112–129.https://doi.org/10.1111/j.1574-6976.2012.00343.x.

35. Shade A, Gregory Caporaso J, Handelsman J, Knight R, Fierer N. 2013. Ameta-analysis of changes in bacterial and archaeal communities withtime. ISME J 7:1493–1506. https://doi.org/10.1038/ismej.2013.54.

36. Duncan DS, Jewell KA, Suen G, Jackson RD. 2016. Detection of short-term cropping system-induced changes to soil bacterial communitiesdiffers among four molecular characterization methods. Soil BiolBiochem 96:160 –168. https://doi.org/10.1016/j.soilbio.2016.02.002.

37. Lauber CL, Ramirez KS, Aanderud Z, Lennon J, Fierer N. 2013. Temporalvariability in soil microbial communities across land-use types. ISME J7:1641–1650. https://doi.org/10.1038/ismej.2013.50.

38. Fierer N, Bradford MA, Jackson RB. 2007. Toward an ecological classifi-cation of soil bacteria. Ecology 88:1354 –1364. https://doi.org/10.1890/05-1839.

39. Ouyang Y, Norton JM, Stark JM, Reeve JR, Habteselassie MY. 2016.Ammonia-oxidizing bacteria are more responsive than archaea to nitro-gen source in an agricultural soil. Soil Biol Biochem 96:4 –15. https://doi.org/10.1016/j.soilbio.2016.01.012.

40. Ouyang Y, Norton JM, Stark JM. 2017. Ammonium availability andtemperature control contributions of ammonia oxidizing bacteria andarchaea to nitrification in an agricultural soil. Soil Biol Biochem 113:161–172. https://doi.org/10.1016/j.soilbio.2017.06.010.

41. Daims H, Lebedeva EV, Pjevac P, Han P, Herbold C, Albertsen M, JehmlichN, Palatinszky M, Vierheilig J, Bulaev A, Kirkegaard RH, von Bergen M,Rattei T, Bendinger B, Nielsen PH, Wagner M. 2015. Complete nitrifica-tion by Nitrospira bacteria. Nature 528:504 –509. https://doi.org/10.1038/nature16461.

42. Beier S, Bertilsson S. 2013. Bacterial chitin degradation-mechanisms andecophysiological strategies. Front Microbiol 4:149. https://doi.org/10.3389/fmicb.2013.00149.

43. Mäder P, Fliessbach A, Dubois D, Gunst L, Fried P, Niggli U. 2002. Soilfertility and biodiversity in organic farming. Science 296:1694 –1697.https://doi.org/10.1126/science.1071148.

44. Moeskops B, Buchan D, Sleutel S, Herawaty L, Husen E, Saraswati R,Setyorini D, De Neve S. 2010. Soil microbial communities and activitiesunder intensive organic and conventional vegetable farming in WestJava, Indonesia. Appl Soil Ecol 45:112–120. https://doi.org/10.1016/j.apsoil.2010.03.005.

45. Nannipieri P, Giagnoni L, Renella G, Puglisi E, Ceccanti B, Masciandaro G,Fornasier F, Moscatelli MC, Marinari S. 2012. Soil enzymology: classicaland molecular approaches. Biol Fertil Soils 48:743–762. https://doi.org/10.1007/s00374-012-0723-0.

46. Vranova V, Rejsek K, Formanek P. 2013. Proteolytic activity in soil: a review.Appl Soil Ecol 70:23–32. https://doi.org/10.1016/j.apsoil.2013.04.003.

47. Rawlings ND, Barrett AJ, Bateman A. 2010. MEROPS: the peptidasedatabase. Nucleic Acids Res 38:D227–D233. https://doi.org/10.1093/nar/gkp971.

48. Bach H-J, Hartmann A, Schloter M, Munch JC. 2001. PCR primers andfunctional probes for amplification and detection of bacterial genes forextracellular peptidases in single strains and in soil. J Microbiol Methods44:173–182. https://doi.org/10.1016/s0167-7012(00)00239-6.

49. Stark JM, Hart SC. 1996. Diffusion technique for preparing salt solu-tions, Kjeldahl digests, and persulfate digests for nitrogen-15 analy-sis. Soil Sci Soc Am J 60:1846 –1855. https://doi.org/10.2136/sssaj1996.03615995006000060033x.

50. Norton JM, Stark JM. 2011. Regulation and measurement of nitrification

in terrestrial systems. Methods Enzymol 486:343–368. https://doi.org/10.1016/B978-0-12-381294-0.00015-8.

51. Bonde T, Nielsen T, Miller M, Sørensen J. 2001. Arginine ammonificationassay as a rapid index of gross N mineralization in agricultural soils. BiolFertil Soils 34:179 –184. https://doi.org/10.1007/s003740100395.

52. Gianfreda L, Sannino F, Ortega N, Nannipieri P. 1994. Activity of free andimmobilized urease in soil: effects of pesticides. Soil Biol Biochem 26:777–784. https://doi.org/10.1016/0038-0717(94)90273-9.

53. Parham JA, Deng SP. 2000. Detection, quantification and characteriza-tion of �-glucosaminidase activity in soil. Soil Biol Biochem 32:1183–1190. https://doi.org/10.1016/S0038-0717(00)00034-1.

54. Tabatabai M. 1994. Soil enzymes, p 778 – 833. In Weaver R, Angle S,Bottomly P, Bezdicek D, Smith S, Tabatabai A, Wollum A (ed), Methodsof soil analysis. Part 2—microbiological and biochemical properties. SoilScience Society America, Madison, WI.

55. Wang Q, Fish JA, Gilman M, Sun Y, Brown CT, Tiedje JM, Cole JR. 2015.Xander: employing a novel method for efficient gene-targeted metag-enomic assembly. Microbiome 3:32. https://doi.org/10.1186/s40168-015-0093-6.

56. Fish JA, Chai B, Wang Q, Sun Y, Brown CT, Tiedje JM, Cole JR. 2013.FunGene: the functional gene pipeline and repository. Front Microbiol4:291. https://doi.org/10.3389/fmicb.2013.00291.

57. Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos J, Bealer K,Madden TL. 2009. BLAST�: architecture and applications. BMC Bioinfor-matics 10:421. https://doi.org/10.1186/1471-2105-10-421.

58. Herbold CW, Pelikan C, Kuzyk O, Hausmann B, Angel R, Berry D, Loy A.2015. A flexible and economical barcoding approach for highly multi-plexed amplicon sequencing of diverse target genes. Front Microbiol6:731. https://doi.org/10.3389/fmicb.2015.00731.

59. Cole JR, Wang Q, Fish JA, Chai B, McGarrell DM, Sun Y, Brown CT,Porras-Alfaro A, Kuske CR, Tiedje JM. 2014. Ribosomal Database Project:data and tools for high throughput rRNA analysis. Nucleic Acids Res42:D633–D642. https://doi.org/10.1093/nar/gkt1244.

60. Edgar RC. 2013. UPARSE: highly accurate OTU sequences from micro-bial amplicon reads. Nat Methods 10:996 –998. https://doi.org/10.1038/nmeth.2604.

61. Edgar RC, Haas BJ, Clemente JC, Quince C, Knight R. 2011. UCHIMEimproves sensitivity and speed of chimera detection. Bioinformatics27:2194 –2200. https://doi.org/10.1093/bioinformatics/btr381.

62. Wang Q, Quensen JF, III, Fish JA, Lee TK, Sun Y, Tiedje JM, Cole JR. 2013.Ecological patterns of nifH genes in four terrestrial climatic zones ex-plored with targeted metagenomics using framebot, a new informaticstool. mBio 4:e00592-13. https://doi.org/10.1128/mBio.00592-13.

63. Eddy SR. 2011. Accelerated profile HMM searches. PLoS Comput Biol7:e1002195. https://doi.org/10.1371/journal.pcbi.1002195.

64. Price MN, Dehal PS, Arkin AP. 2010. FastTree 2—approximatelymaximum-likelihood trees for large alignments. PLoS One 5:e9490.https://doi.org/10.1371/journal.pone.0009490.

65. McMurdie PJ, Holmes S. 2013. Phyloseq: an R package for reproducibleinteractive analysis and graphics of microbiome census data. PLoS One8:e61217. https://doi.org/10.1371/journal.pone.0061217.

66. Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Huntley J, Fierer N,Owens SM, Betley J, Fraser L, Bauer M, Gormley N, Gilbert JA, Smith G,Knight R. 2012. Ultra-high-throughput microbial community analysis onthe Illumina HiSeq and MiSeq platforms. ISME J 6:1621–1624. https://doi.org/10.1038/ismej.2012.8.

67. Wang Q, Garrity GM, Tiedje JM, Cole JR. 2007. Naive Bayesian classifier forrapid assignment of rRNA sequences into the new bacterial taxonomy. ApplEnviron Microbiol 73:5261–5267. https://doi.org/10.1128/AEM.00062-07.

68. Love MI, Huber W, Anders S. 2014. Moderated estimation of fold changeand dispersion for RNA-seq data with DESeq2. Genome Biol 15:550.https://doi.org/10.1186/s13059-014-0550-8.

69. Whitman T, Pepe-Ranney C, Enders A, Koechli C, Campbell A, BuckleyDH, Lehmann J. 2016. Dynamics of microbial community compositionand soil organic carbon mineralization in soil following addition ofpyrogenic and fresh organic matter. ISME J 10:2918 –2930. https://doi.org/10.1038/ismej.2016.68.

70. Williamson N, Brian P, Wellington EMH. 2000. Molecular detection ofbacterial and streptomycete chitinases in the environment. Antonie VanLeeuwenhoek 78:315–321. https://doi.org/10.1023/A:1010225909148.

71. Koper TE, El-Sheikh AF, Norton JM, Klotz MG. 2004. Urease-encodinggenes in ammonia-oxidizing bacteria. Appl Environ Microbiol 70:2342–2348. https://doi.org/10.1128/aem.70.4.2342-2348.2004.

Fertilization Affects Community and N Mineralization Applied and Environmental Microbiology

March 2020 Volume 86 Issue 5 e02278-19 aem.asm.org 15

on February 15, 2021 by guest

http://aem.asm

.org/D

ownloaded from