Seasonal Variations in Soil Microbiota Profile of Termite

microorganisms

Article Seasonal Variations in Soil Microbiota Proﬁle of Termite (Syntermes wheeleri) Mounds in the Brazilian Tropical Savanna

Helena Ipe Pinheiro Guimaraes 1 , Renata Henrique Santana 2, Rafaella Silveira 3 , Otavio Henrique Bezerra Pinto 1, Betania Ferraz Quirino 3, Cristine Chaves Barreto 4 , Mercedes Maria da Cunha Bustamante 5 and Ricardo Henrique Krüger 1,*

1 Cellular Biology Department, University of Brasilia, Campus Universitário Darcy Ribeiro, Brasília D.F. 70910-900, Brazil; [email protected] (H.I.P.G.); [email protected] (O.H.B.P.) 2 Brazilian Federal Institute, Campus Planaltina, Brasília D.F. 70910-900, Brazil; [email protected] 3 Embrapa-Agroenergy, Parque Estação Biológica (PqEB), Genetics and Biotechnoloy Laboratory, PqEB s/nº, Brasília D.F. 70770-901, Brazil; [email protected] (R.S.); [email protected] (B.F.Q.) 4 Genomic Sciences and Biotechnology Graduate Program, Universidade Católica de Brasília, Brasilia D.F. 70790-160, Brazil; [email protected] 5 Department of Ecology, Biological Sciences Institute, University of Brasilia, Brasilia D.F. 70910-900, Brazil; [email protected] * Correspondence: [email protected]; Tel.: +55-61-3107-2977

Received: 31 July 2020; Accepted: 17 September 2020; Published: 27 September 2020

Abstract: Eusocial animals, such as the termites, often build a nest-like structure called a mound that provides shelter with stable internal conditions and protection against predators. Termites are important components of the Brazilian Cerrado biota. This study aimed to investigate the bacterial community composition and diversity of the Syntermes wheeleri termite-mound soil using culture-independent approaches. We considered the vertical profile by comparing two different mound depths (mound surface and 60 cm) and seasonality with samplings during the rainy and dry seasons. We compared the mound soil microbiota to the adjacent soil without the influence of the mound to test the hypothesis that the Cerrado soil bacterial community was more diverse and more susceptible to seasonality than the mound soil microbiota. The results support the hypothesis that the Cerrado soil bacterial community is more diverse than the mound soil and also has a higher variability among seasons. The number of observed OTUs (Operational Taxonomic Units) was used to express bacterial richness, and it indicates that soil moisture has an effect on the community distribution and richness of the Cerrado samples in comparison to mound samples, which remain stable across seasons. This could be a consequence of the protective role of the mound for the termite colony. The overall community taxonomic profile was similar between soil samples, especially when compared to the taxonomic composition of the Syntermes wheeleri termite’s gut, which might be explained by the different characteristics and functionality between the soil and the gut microbial community.

Keywords: termite; mound; soil; Cerrado; soil microbiology; environmental microbiology

1. Introduction Cerrado is the richest tropical savanna in the world [1]. It represents 24% of the Brazilian territory and it is the second largest biome after the Amazon rainforest [2,3]. The Cerrado soils are predominantly Oxisols characterized by their acidity, red color, low nutrient concentrations, and high clay content and aluminum concentration [3–5]. This biome is characterized by dry winters and rainy summers, with markedly seasonal rainfall having a great impact on the soil microbiology function [6,7].

Microorganisms 2020, 8, 1482; doi:10.3390/microorganisms8101482 www.mdpi.com/journal/microorganisms Microorganisms 2020, 8, 1482 2 of 14

Termite colonies are distributed in the Cerrado according to its niche preferences, such as wood and organic matter abundance [8]. Termite nests have complex structures, and may include subterranean galleries [9]. Termite mound walls are different from the subterranean nest, and provide stability to the community during the rainy season [10]. The presence of the termites in the Cerrado soil is associated with higher pH values, Calcium (Ca), Phosphorus (P), and Potassium (K) concentrations and lower Aluminum (Al) concentrations [11]. Soil physical and chemical parameters are significant determinants of the microbial community profile [12]. The portion of the ecosystem under the influence of the termite colony is called the termitosphere, and it has unique characteristics such as different soil texture and organic matter content when compared to the adjacent soil [13]. Some termite species are able to use the thermal proprieties of the mound to buffer temperature variations in hot and dry areas [14]. Microbial activity is more dependent on the vegetation and seasonality at soil surface (0–5 cm), while with increased profile depth, the microbial community responds more to soil characteristics [6]. The soil microbiota is responsible for 80% of the total processes that occur in the soil, such as carbon and nitrogen cycles [15,16]. Termites are eusocial animals, meaning they have a strict work distribution system according to castes [17]. In fact, the entire colony can be considered one superorganism divided according to its functions: reproduction (alates, kings, and queens), feeding and structure (workers), defense (soldiers), and protection and homeostasis (the mound) [17]. Workers are responsible for constructing the mound and foraging for food [18]. The use of soil for mound construction and consumption was the most important evolutionary event for the termites, and overlapped with the loss of protists which were substituted by bacterial symbionts in the gut of Termitidae termites [18]. The Termitidae family is part of the higher termite clade. It is the most prominent termite family in the tropical region, comprising most of the endemic species in the Cerrado [19]. The Syntermes genus includes termites from the subfamily Nasutermitinae, family Termitidae, which are widely distributed in the tropical savannas and are characterized by large size, subterranean mounds, and litter-feeding habits [8]. The absence of flagellates in the higher termite’s gut corresponds to the compartmentalization of the hindgut, which improves fermentation processes in this area [17]. Previous studies showed that the microbiota of the intestinal region and mound soil of the Syntermes termites are very different, indicating a higher prevalence of Actinobacteria and Proteobacteria in the mound. In contrast, Firmicutes and Spirochaetes are prevalent in the gut [18,20]. The main objective of this work was to characterize the soil’s bacterial community associated with mounds of the Syntermes wheeleri termite and compare it with the Cerrado soil microbiota. We collected the mound soil at two different depths (0–10 cm and 50–60 cm) and in different seasons (dry and rainy). We tested the following hypothesis: (i) the mound soil has lower bacterial diversity than the Cerrado soil, and (ii) seasonality does not affect bacterial richness of the mound soil.

2. Materials and Methods

2.1. Study Site and Sample Collection We conducted the study in a protected Cerrado area managed by the Brazilian Institute of Geography and Statistics (RECOR Ecological Reserve), Brasilia, Brazil (15◦56054.600 S; 47◦52011.700 W). The control soil and the mound soil were collected in a campo sujo, a Cerrado vegetation type characterized by the predominance of the herbaceous layer with scattered shrubs and trees (Figure1A). The Syntermes wheeleri termite species was identified by its soldier caste, and mound soil was collected at two different depths: at the mound surface (0–10 cm) and at a depth interval of 50–60 cm (mound soil) (Figure1B). Control soil was collected at a distance of 100 m from the mound at the depth interval of 50–60 cm. Samples were collected in triplicates at different periods to evaluate seasonal effects: late dry season (September 2017), late rainy season (March 2018), and early rainy season (transition) (November 2017) (Figure2). Soil was sieved using a 2 mm mesh in situ and stored in plastic bags on ice. After arrival at the laboratory, it was stored at 80 C until DNA extraction. − ◦ Microorganisms 2020, 8, x FOR PEER REVIEW 3 of 15

Microorganisms 2020, 8, x FOR PEER REVIEW 3 of 15 season (transition) (November 2017) (Figure 2). Soil was sieved using a 2 mm mesh in situ and stored seasonin plastic (transition) bags on ice.(November After arrival 2017) at (Figure the laboratory, 2). Soil was it was sieved stored using at −a80 2 mm°C until mesh DNA in situ extraction. and stored inMicroorganisms plastic bags2020 on, 8 ice., 1482 After arrival at the laboratory, it was stored at −80 °C until DNA extraction.3 of 14

Figure 1. Site overview and sample collection. (A) Campo sujo in the IBGE (Instituto Brasileiro de Figure 1. Site overview and sample collection. (A) Campo sujo in the IBGE (Instituto Brasileiro de Geografia e Estatística) Ecological Reserve during the dry season, and (B) Syntermes wheeleri termite FigureGeograﬁa 1. Site e Estat overviewística) Ecologicaland sample Reserve collection. during (A) the Campo dry season,sujo in andthe (IBGEB) Syntermes (Instituto wheeleri Brasileirotermite de mound in the campo sujo during the dry season. The red circle indicates the collection of the 0–10 cm Geografiamound in e the Estatísticacampo sujo) Ecologicalduring the Reser dryve season. during The the red dry circle season indicates, and (B the) Syntermes collection wheeleri of the 0–10termite cm mound surface soil samples, the purple circle indicates the collection of the 50–60 cm mound soil moundmound in surface the campo soil sujo samples, during the the purple dry season. circle indicatesThe red circle the collectionindicates the of thecollection 50–60 cmof the mound 0–10 soilcm samples, and the blue circle indicates the collection of the control samples, or Cerrado samples, also moundsamples, surface and the soil blue samples, circle indicates the purple the collectioncircle indicates of the controlthe collection samples, of orthe Cerrado 50–60 cm samples, mound also soil at at a depth of 50–60 cm within the soil. samplesa depth, of and 50–60 the cmblue within circle theindicates soil. the collection of the control samples, or Cerrado samples, also at a depth of 50–60 cm within the soil. 350

350 300

300 250 ) ) m

2m 50 m m (

( 200

)

n n ) m o i o m i t t m a m t ( i ( a 200

t p i i n n 150 c o p i o e i i t r t a c P t i a e t p r i i 150 c P p e i 100 r c P e r

P 100 50

50 0 jul/17 ago/17 set/17 out/17 nov/17 dez/17 jan/18 fev/18 mar/18 Jul/17 Aug/17 Sep/17 Oct/17 Nov/17 Dec/17 Jan/18 Feb/18 Mar/18 0 jul/17 ago/17 set/17 out/17 nov/17 dez/17 jan/18 fev/18 mar/18 Jul/17 Aug/17 Sep/17 Oct/17 Nov/17 Dec/17 Jan/18 Feb/18 Mar/18 FigureFigure 2.2. Monthly precipitation data from the Brazilian National Meteorology InstituteInstitute fromfrom JulyJuly 20172017 to March 2018 in Brasilia. Arrows indicate the months in which soil was collected: the dry season in Figureto March 2. Monthly 2018 in Brasilia.precipitation Arrows data indicate from the the Brazilian months National in which Meteorology soil was collected: Institute the from dry Julyseason 2017 in September 2017, transition in November 2017, and late rainy season in March 2018. toSeptember March 2018 2017, in Brasilia.transition Arrows in November indicate 2017, the months and late in rainy which season soil wasin March collected: 2018. the dry season in 2.2. Physico-ChemicalSeptember 2017, transition Soil Parameters in November 2017, and late rainy season in March 2018. 2.2. Physico-Chemical Soil Parameters Soil pH was measured after shaking it with distilled water using a HI 2221 Calibration Check 2.2. PhysiSoilco pH-C hemicalwas measured Soil Parameters after shaking it with distilled water using a HI 2221 Calibration Check pH/ORP calibrated potentiometer (HANNA instruments, Rhode Island, USA). Nitrate and ammonium pH/ORP calibrated potentiometer (HANNA instruments, Rhode Island, USA). Nitrate and measurementsSoil pH was were measured performed after using shaking a Shimadzu it with distilled UV-1203 water spectrometer using a (Shimadzu,HI 2221 Calibration Kyoto, Japan) Check at ammonium measurements were performed using a Shimadzu UV-1203 spectrometer (Shimadzu, pH/ORPwavelengths calibrated of 218, 228, potentiometer 254, and 280 nm(HANNA for the nitrate,instruments and with, Rhode Nessler Island reagent, USA for ammonium). Nitrate and [21]. Kyoto, Japan) at wavelengths of 218, 228, 254, and 280 nm for the nitrate, and with Nessler reagent ammoniumSoil gravimetric measurements water content were was performed determined using after a drying Shimadzu it for UV 72 h-1203 at 105 spectrometerC[21]. Available (Shimadzu, P and for ammonium [21]. Soil gravimetric water content was determined after drying◦ it for 72 h at 105 °C Kyoto,K were Japan extracted) at wavelengths with Mehlich of 1,218, while 228, Ca, 254 Mg, and (Magnesium), 280 nm for the and nitrate, Al were and extracted with Nessler with KClreagent 1M. for ammonium [21]. Soil gravimetric water content was determined after drying it for 72 h at 105 °C Organic matter (OM) was determined by the Walkley-Black method. These analyses were performed following the EMBRAPA (Brazilian Agricultural Research Corporation) protocols [22].

Microorganisms 2020, 8, 1482 4 of 14

2.3. Soil DNA Extraction and 16S rRNA Gene Library Genomic DNA was extracted according to the Griffiths et al. protocol [23] with 5% hexadecyltrimethylammonium bromide (CTAB) and quantified with Qubit BR. DNA quality was verified on 1% agarose gel. To describe the soil bacterial community, 27 DNA samples were sent to Macrogen, South Korea. The company constructed a library of the V3-V4 portion of the16S rRNA gene (Bakt_341F: 50-CCTACGGGNGGCWGCAG-30 Bakt_805R: 50-GACTACHVGGGTATCTAATCC-30), and sequencing was performed in an Illumina MiSeq platform (paired ended, 300 bp, 100 k reads/sample) and results were delivered in the fastq format. Sequencing quality was analyzed using the FASTQC program (0.11.8) (Babraham Institute, Cambridge, UK) [24]. Bioinformatics and statistical analysis. Raw sequences were previously treated by the bbduk program in the BBTools package for Nextera adaptors and Phix sequences removal (BBMap–Bushnell B.–sourceforge.net/projects/bbmap/). The treated sequences were then imported into the QIIME2 2019.7 program (University of Colorado, USA) [25], and the pipeline was executed for pair-ended sequences with the following steps: primer removal [26], quality filtering [27], dereplication, Operational Taxonomic Unit (OTU) 97% clusterization and chimera removal [28], and taxonomy assignment using the SILVA (2018) database for bacteria [29]. To compare the bacterial community among soils, principal component analysis (PCoA) based on Weighted Unifrac distance was calculated. We performed a rarefaction curve and evaluated microbial richness by observed OTUs number (Operational Taxonomic Unit), Pielou and Shannon indexes, which were obtained using the QIIME2 program. To examine differences between predicted gene abundances from functional inference analysis, an analysis of variance (ANOVA) or the Kruskal-Wallis test were performed with the Prism program (8.0.2) (GraphPad software, California, USA). Normality was tested with Shapiro test. A Permutational Multivariance Analysis of Variance (PERMANOVA) was carried out using Rstudio (1.456–© 2009–2018 RStudio, Inc., Boston, MA, USA) to evaluate the differences between bacterial community in the soil samples. Non-Metric Multidimensional Scaling (NMDS) based on Bray–Curtis dissimilarity and Weighted Unifrac PCoA were obtained in Rstudio with vegan [30] and phyloseq [31] packages. Additionally, functional inferences based on 16S rRNA gene data was performed with a PICRUST2 [32] in QIIME2. The enzyme-encoding gene abundances with their respective metabolic pathways were assigned by Enzyme Commission numbers (EC numbers) and the Kyoto Encyclopedia of Genes and Genomes (KEGG). Only genes assigned to the Nitrogen metabolism were evaluated. Each EC number was confirmed by the MetaCyc Metabolic Pathway database [33]. The functional prediction output was used as an input file on the Prism software. We performed ANOVA with Tukey’s post hoc test to test the differences in enzyme-encoding gene abundances between soil samples during the dry season. The dry season was particularly observed because we found higher microbial diversity during this period (see Section3). In addition, the majority of recent studies in Cerrado soils have been shown a higher microbial diversity in the dry season. Before statistical analysis, data were tested for normality by the Shapiro test.

3. Results

3.1. Comparison of Bacterial Community of Mound and Control Soil in the Dry Season The most abundant phyla, Acidobacteria, Actinobacteria, Chloroflexi, Planctomycetes, Proteobacteria, Verrucomicrobia, and WPS-2, were common in the control soil, mound surface and mound 60 cm soil samples in the dry season. Actinobacteria and Proteobacteria were significantly more abundant in the mound soil samples than in control samples. Rare phyla were defined as those that represent less than 1% of the total bacterial community, comprising 2% of the control soil total community and 1.5% of the mound soil total community (Figure3). Microorganisms 2020, 8, x FOR PEER REVIEW 5 of 15

Microorganismsthat2020 represent, 8, 1482 less than 1% of the total bacterial community, comprising 2% of the control soil total 5 of 14 community and 1.5% of the mound soil total community (Figure 3).

100 others (<1%) WPS-2 Verrucomicrobia Proteobacteria Planctomycetes GAL15 Firmicutes Chloroflexi

e Bacteroidetes c n

a Actinobacteria d

n Acidobacteria u

b 50 a

e v i t a l e R

0 o 0 0 d 6 a d rr n d e u n c o u m o m Samples

Figure 3. FigureMicrobial 3. Microbial phyla proﬁlephyla profile of the of drythe dry season season for for the the Cerrado Cerrado ( (50–6050–60 cm), cm), mound mound surface surface (0–10 (0–10 cm), and moundcm), soil and mound (50–60 soil cm) (50 samples–60 cm) samples in 2017. in 2017 The. The graph graph shows the the relative relative abundance abundance of the most of the most abundantabundant bacterial bacterial phyla. Valuesphyla. Values presented presented are the are meanthe mean from from the the Cerrado, Cerrado, mound mound surface surface,, and and mound soil triplicatesmound in soil the triplicates dry season. in the dry season.

3.2. Seasonality3.2. Seasonality and Chemical and Chemical Parameters Parameters E ffEectffect on on BacterialBacterial Distribution Distribution in Mound in Mound and Control and S Controloil Samples Soil Samples The most abundant phyla among the samples of the late and early rainy season were: TheAcidobacteria, most abundant Actinobacteria, phyla among Chloroflexi, the samples Planctomycetes, of the Proteobacteria, late and early Verrucomicrobia, rainy season were: Acidobacteria,WD272 Actinobacteria,, and Thermotogae Chloroflexi, (Figure A1). Planctomycetes, The season’s influence Proteobacteria, on the bacterial Verrucomicrobia, community was WD272, and Thermotogaefurther explored (Figure through A1). The a Non season’s-Metric Multidimensional influence on the Scaling bacterial (NMDS) community graph, where was samples further explored through awere Non-Metric distributed Multidimensional according to the Bray Scaling–Curtis (NMDS) dissimilarity graph, matrix where calculated samples from the were OTU distributed (Operational Taxonomic Unit) table for the bacterial community (Figure 4A), although the accordingPERMANOVA to the Bray–Curtis analysis of dissimilarity the weighted Unifrac matrix matrix calculated did not fromsignificantly the OTU differentiate (Operational samples in Taxonomic Unit) tableregard for theto the bacterial site of collection community (Figure (Figure 4B). Verrucomicrobia4A), although phylum the PERMANOVA positively correlates analysis with of the humidity (Figure A2). weighted UnifracMicroorganisms matrix 2020, 8, didx FOR notPEER significantlyREVIEW differentiate samples in regard to the site6 of 15 of collection (Figure4B). Verrucomicrobia phylum positively correlates with humidity (Figure A2). A B

% Weighted UniFrac cerrado mound0 mound60 0.10

2 0.05 S D M N ] % 5 . 6 2

[ 0.00

2 . s i x A

−0.05

−0.10

−0.1 0.0 0.1 Axis.1 [27.3%]

Figure 4. Non-metricFigure 4. Non- dimensionalmetric dimensional scale scale analysis analysis from from the the Cerrado Cerrado soil samples soil samples (50–60 cm), (50–60 mound cm), mound surface (0–10surface cm), (0 and–10 cm) mound, and mound soil (50–60soil (50–60 cm) cm) at at the dry, dry, transition transition,, and rainy and seasons. rainy seasons. (A) Non-Metric (A) Non-Metric MultidimensionalMultidimensional Scaling Scaling (NMDS) (NMDS of) of thethe bacterial bacterial community community in regard in to regardsoil moisture to soilcontent. moisture Data content. from the bacterial Operational Taxonomic Unit (OTU) matrix was distributed according to the Bray– Data fromCurtis the bacterial dissimilarity Operational analysis and disposed Taxonomic with its Unit respective (OTU) soil matrix sample washumidity distributed measure. (B according) The to the Bray–Curtisweighted dissimilarity Unifrac matrix analysis analysis and displays disposed the sites with of the its Cerrado respective soil in soilgreen sample, mound surface humidity in measure. (B) The weightedorange, and Unifrac mound matrix soil in lavender analysis. displays the sites of the Cerrado soil in green, mound surface in orange, and mound soil in lavender. Values for N-nitrate, P, moisture, pH, K, Ca, Mg, Al, ammonium, total N, remaining-P, and organic matter are shown in Table S1. Since the seasons did not significantly affect parameter analysis, the PCA aggregates all of the samples for a greater statistical inference, which separates mound soil depths and Cerrado with a 63% of explanation related to soil chemical composition (Figure 5).

Figure 5. Principal component analysis of the soil chemical parameters for termite mound (mound 0 and mound 60) and Cerrado samples (cerrado) collected in 2017 and 2018. Parameter values are in Table S1. A previous analysis was performed to remove correlated variables, and the remaining parameters are pH, P, P-rem (remaining P), OM (Organic matter), Nitrogen, and Mg. The dry season is represented with a circle, the transition season with a square, and the rainy season with a dash.

Microorganisms 2020, 8, x FOR PEER REVIEW 6 of 15

A B

% Weighted UniFrac cerrado mound0 mound60 0.10

2 0.05 S D M N ] % 5 . 6 2

[ 0.00

2 . s i x A

−0.05

−0.10

−0.1 0.0 0.1 Axis.1 [27.3%]

Figure 4. Non-metric dimensional scale analysis from the Cerrado soil samples (50–60 cm), mound surface (0–10 cm), and mound soil (50–60 cm) at the dry, transition, and rainy seasons. (A) Non-Metric Multidimensional Scaling (NMDS) of the bacterial community in regard to soil moisture content. Data from the bacterial Operational Taxonomic Unit (OTU) matrix was distributed according to the Bray– Curtis dissimilarity analysis and disposed with its respective soil sample humidity measure. (B) The weighted Unifrac matrix analysis displays the sites of the Cerrado soil in green, mound surface in Microorganisms orange2020, 8,, and 1482 mound soil in lavender. 6 of 14

Values for N-nitrate, P, moisture, pH, K, Ca, Mg, Al, ammonium, total N, remaining-P, and Valuesorganic for N-nitrate,matter are P,shown moisture, in Tab pH,le S1. K, Since Ca, Mg, the Al,seasons ammonium, did not significantly total N, remaining-P, affect parameter and organic matter areanalysis, shown the in PCA Table aggregates S1. Since all the of seasonsthe samples did for not a signiﬁcantlygreater statistical aﬀ ectinference, parameter which analysis, separatesthe PCA mound soil depths and Cerrado with a 63% of explanation related to soil chemical composition aggregates all of the samples for a greater statistical inference, which separates mound soil depths and (Figure 5). Cerrado with a 63% of explanation related to soil chemical composition (Figure5).

Figure 5.FigurePrincipal 5. Principal component component analysis analysis of of thethe soil chemical chemical parameters parameters for termite for termite mound mound(mound 0 (mound 0 and moundand mound 60) and60) and Cerrado Cerrado samples samples (cerrado) collected collected in 2017 in 2017and 2018. and Parameter 2018. Parameter values are valuesin are in TableTable S1. AS1 previous. A previous analysis analysis was was performed performed to to remove remove correlated correlated variables, variables, and the and remaining the remaining parametersparameters are pH, are P, pH, P-rem P, P (remaining-rem (remaining P), P), OM OM (Organic (Organic matter),matter), Nitrogen Nitrogen,, and and Mg. Mg. The Thedry season dry season is representedis represented with a circle, with a the circle, transition the transition season season with with a square,a square, andand the rainy rainy season season with with a dash. a dash.

3.3. EcologicalMicroorganisms Parameters 2020, 8, x FOR and PEER Indexes REVIEW for Soil Samples and Seasons 7 of 15 The3.3. Shannon Ecological indexParameters indicates and Indexes that for the Soil CerradoSamples and soil Seasons samples are more diverse than the mound samples inThe the Shannon dry season index (Figureindicates6 A).that Thethe Cerrado OTU (Operationalsoil samples are Taxonomic more diverseUnit) than the number mound for the moundsamples and control in the samples dry season increases (Figure according6A). The OTU to the(Operational number Taxonomic of sequences Unit) in number the rarefaction for the curve. However,mound it does and not control reach samples a plateau, increases which according indicates to the that number the sequencing of sequences ofin samplesthe rarefaction does curve not cover. the However, it does not reach a plateau, which indicates that the sequencing of samples does not cover microbiome entirely (Figure A3). The Cerrado soil is the only sample that has a significant difference the microbiome entirely (Figure A3). The Cerrado soil is the only sample that has a significant in the observeddifference OTU in the number observed according OTU number to seasons, according showing to seasons, a showingdecrease a in decrease bacterial in bacterial richness as the rainy seasonrichness progresses, as the rainy while season seasonality progresses, while hasno seasonality statistical has di noff erence statistical for difference mound forsurface mound (0–10 cm) and moundsurface soil (0- (50–6010 cm) cm)and mound richness soil (Figure (50-60 6cm)B). richness The Pielou (Figure index, 6B). The a measure Pielou index, of equitability, a measure of did di ffer betweenequitability, Cerrado anddid differ mound between soil Cerrado samples. and However, mound soil the samples. dry season However, has the a higherdry season Pielou has a index in higher Pielou index in comparison to the early and late rainy seasons (Figure A4). comparison to the early and late rainy seasons (Figure A4).

A B

dry season 8.0 transition * rainy season 4000 7.5 * cerrado mound 0 s 3500 U

n 7.0

T mound 60 o n O

n a d h e 3000 S

6.5 v r e s

b 2500 6.0 O

2000 5.5 dry rain dry rain dry rain

cerrado cerrado cerrado mound 0 mound 0 mound 0 mound 60 mound 60 mound 60 transition transition transition

Figure 6.FigureDiversity 6. Diversity and and richness richness measures measures forfor the microbial microbial community community for the for Cerrado the Cerrado (50–60 cm), (50–60 cm), mound surface (0–10 cm), and mound soil (50–60 cm) samples collected in Brasilia, 2017 and 2018. (A) mound surface (0–10 cm), and mound soil (50–60 cm) samples collected in Brasilia, 2017 and 2018. Shannon index values for microbial diversity. The dry, transition, and rainy seasons are shown by the (A) Shannongrey scale index bar. values Significant for microbial values were diversity. considered The for dry,p value transition, < 0.05 and and are rainy indicated seasons with arean shown by the greyasterisk scale‘*’. ( bar.B) Analysis Signiﬁcant of the valuesobserved were OTUconsidered number for each for psoilvalue sample< 0.05at the and dry, are early indicated, and late with an asteriskrainy ‘*’. ( seasons.B) Analysis Statisti ofcally the significant observed values OTU were number considered for eachfor p value soil sample< 0.05 and at are the indicated dry, early, with and late rainy seasons.an asterisk Statistically ‘*’. signiﬁcant values were considered for p value < 0.05 and are indicated with an asterisk ‘*’. 3.4. Functional Inference from Taxonomy Regardless of taxonomical difference, functional prediction analysis allows us to infer the influence of the site or season in microbial processes. Genomic prediction from 16S rRNA gene data was performed for each sample, and the output is a table with all the functional pathways that can be used for a dissimilatory analysis. The final result indicates a possible functional redundancy between samples (Figure A5). Moreover, functional analysis of soil metataxonomic data enables annotation of nitrogen metabolism enzymes (Figure 7A). All 27 samples can potentially fix nitrogen (EC: 1.18.6.1), reduce nitrate to nitrite (EC: 1.7.7.1), perform nitric oxide reduction (EC: 1.7.2.1), produce glycerol precursors (EC: 1.1.5.3), oxidize ammonia to nitrite (EC: 1.14.99.39), and produce glutamine from ammonia and glutamate (EC: 1.4.1.14) (Table S2). The glutamine synthetase (GS, EC: 6.3.1.2) is also present in the samples and possibly more abundant in the mound soil (50–60 cm) during the dry season (Figure 7B).

Microorganisms 2020, 8, 1482 7 of 14

3.4. Functional Inference from Taxonomy Regardless of taxonomical difference, functional prediction analysis allows us to infer the influence of the site or season in microbial processes. Genomic prediction from 16S rRNA gene data was performed for each sample, and the output is a table with all the functional pathways that can be used for a dissimilatory analysis. The final result indicates a possible functional redundancy between samples (Figure A5). Moreover, functional analysis of soil metataxonomic data enables annotation of nitrogen metabolism enzymes (Figure7A). All 27 samples can potentially fix nitrogen (EC: 1.18.6.1), reduce nitrate to nitrite (EC: 1.7.7.1), perform nitric oxide reduction (EC: 1.7.2.1), produce glycerol precursors (EC: 1.1.5.3), oxidize ammonia to nitrite (EC: 1.14.99.39), and produce glutamine from ammonia and glutamate (EC: 1.4.1.14) (Table S2). The glutamine synthetase (GS, EC: 6.3.1.2) is also present in the samples and possibly more abundant in the mound soil (50–60 cm) during the dry season (FigureMicroorganisms7B). 2020, 8, x FOR PEER REVIEW 8 of 15

Figure 7. TermiteFigure 7. mound Termite soil mound in the soil dry in the season. dry season. (A) mound(A) mound samples samples are are shown shown in in orange orange (0 (0 cm, cm, surface) and red (60surface) cm depth), and red and (60 cm the depth), Cerrado and soilthe Cerrado samples soil are samples in blue are in in blue the in dry the season. dry season. The The enzymatic enzymatic reactions that possibly occur in greater abundance at the surface are: nitrogen reactions that possibly occur in greater abundance at the surface are: nitrogen metabolism—from metabolism—from its fixation until nitrite, nitrate, and glycerol production and the GOGAT pathway. its fixationThe until reverse nitrite, reaction nitrate, for glutamine and glycerol production production via GS pathway and occurs the GOGATin greater abundance pathway. in Thethe reverse reaction formound glutamine 60 cm samples, production whereas via th GSis pathway pathway is directed occurs to in glutamate greater abundanceproduction in in the the Cerrado mound 60 cm samples, whereassamples thisvia GOGAT. pathway (B) is Enzyme directed-encoding to glutamate genes abundance production comparison in the between Cerrado soil samples (50–60 c viam) GOGAT. (B) Enzyme-encodingand mound soils genes (50– abundance60 cm; 0–10 cm) comparison collected in betweena Cerrado soilarea (50–60during the cm) dry and season mound of 2017. soils The (50–60 cm; functional prediction based on 16S rRNA gene taxonomic inference in PICRUST2 using the Kyoto 0–10 cm) collectedEncyclopedia in a of Cerrado Genes and area Genomes during (KEGG the dry) database. season Differences of 2017. in The mean functional proportions prediction (ANOVA, based on 16S rRNA genep< 0.05) taxonomic are indicated inference by asterisks. in PICRUST2 using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Differences in mean proportions (ANOVA, p < 0.05) are indicated by asterisks. 4. Discussion 4. Discussion We compared the bacterial community of termite-mound soil in different seasons to campo sujo Cerrado soil, used as a control. The most abundant bacterial phyla found in the soil samples in this We compared the bacterial community of termite-mound soil in different seasons to campo sujo study were Proteobacteria, Acidobacteria, Actinobacteria, Verrucomicrobia, Chloroflexi, and Cerrado soil,Planctomycetes. used as a control. These The phyla most are abundant commonly bacterial found in phyla soil samples found in as the previously soil samples reported in this study were Proteobacteria,[12,16,24,34– Acidobacteria,36]. Most bacterial Actinobacteria, phyla in soil are rare Verrucomicrobia,; nonetheless, this “ Chloroflexi,low abundance and” part Planctomycetes. of the These phylacommunity are commonly represents found 41% inof total soil 16S samples rRNA gene as previously sequences and reported are found [12 in, 16a great,24, 34variety–36]. of Most soils bacterial [35]. Commonly found in the mounds, Actinobacteria have a symbiotic relationship with termites. phyla in soil are rare; nonetheless, this “low abundance” part of the community represents 41% of These bacteria recycle nutrients and produce antimicrobial peptides protecting the nest against total 16S rRNAforeign gene pathogens sequences [37,38]. Actinobacteriaand are found are in also a considered great variety to be of the soils second [35 most]. Commonly abundant found in the mounds,bacterial Actinobacteria phyla in the gut haveof higher a symbiotictermites and relationship are mostly prevalent with termites.on those that These feed on bacteria humic recycle nutrients andmatter produce and soil antimicrobial[39]. However, that peptides is not what protecting was observed the for nest the againstSyntermes foreignwheeleri termite pathogens gut, [37,38]. which has Firmicutes, Bacteroidetes, and Spirochaetes as the most abundant phyla [20]. Actinobacteria are also considered to be the second most abundant bacterial phyla in the gut of higher Similarly to the termite nest soil, the termite’s gut has been object of study as a distinctive termites andecological are mostly niche prevalent for thousands on of those different that bacterial feed on species humic and matter phylotypes and soilwith [ a39 broad]. However, set of that is not what wasmetabolic observed activities for [40] the. TheseSyntermes include lignocellulolytic wheeleri termite degradation, gut, which methanogenesis, has Firmicutes, CO2-reduced Bacteroidetes, and Spirochaetesacetogenesis, as the nitrogen most fixation abundant, and phylaother activities [20]. [40,41]. The termite’s nest taxonomic content is usually significantly different from that of the termite’s gut [18,42]. This could be explained by the different physiological conditions in the termite’s gut (i.e., anoxygenic and alkaline) and the termite’s nest soil (i.e., oxygenic and acidic). Regarding this study specifically, the pH of the S. wheeleri termite gut was reported as 9–10 [20] and the Cerrado soil between 5–6. Along with that, structural variations between gut and soil confer different microbial niches [42]. Nonetheless, microbial biomass is inevitably transferred through feces from the termite’s gut to the nest soil, directly impacting the

Microorganisms 2020, 8, 1482 8 of 14

Similarly to the termite nest soil, the termite’s gut has been object of study as a distinctive ecological niche for thousands of different bacterial species and phylotypes with a broad set of metabolic activities [40]. These include lignocellulolytic degradation, methanogenesis, CO2-reduced acetogenesis, nitrogen fixation, and other activities [40,41]. The termite’s nest taxonomic content is usually significantly different from that of the termite’s gut [18,42]. This could be explained by the different physiological conditions in the termite’s gut (i.e., anoxygenic and alkaline) and the termite’s nest soil (i.e., oxygenic and acidic). Regarding this study specifically, the pH of the S. wheeleri termite gut was reported as 9–10 [20] and the Cerrado soil between 5–6. Along with that, structural variations between gut and soil confer different microbial niches [42]. Nonetheless, microbial biomass is inevitably transferred through feces from the termite’s gut to the nest soil, directly impacting the microbial community structure. The work of Minkley and collaborators [43] showed that the bacterial community of the gut of the Hodotermes mossambicus termite species is nest-specific, and it is used to discriminate between nestmates and non-nestmates—an ability that is disrupted in antibiotic-treated termites [44]. Proteobacteria is currently the largest bacterial group, and it is also the most diverse. A lot of representatives in this phyla, such as the Rhizobia, are known for their ecological function in the nitrogen cycle, particularly for their ability to interact symbiotically with the plant root system and fixate nitrogen [45]. Ammonia is the primary product of the nitrogen fixation reaction via symbiosis [46] and also a product from glutamate and glutamine reduction via glutamine synthetase and glutamate synthase pathways [47]. Glutamine synthetase is responsible for glutamine production, which is associated to bacterial growth [48,49]. The glutamine synthetase enzyme appears to be more abundant in the mound surface (0–10 cm) soils and mound soil (50–60 cm) than the control soil (50–60 cm) during the dry season. In contrast, the glutamate synthase enzyme, present in all samples in the dry season, is responsible for executing the reverse pathway for production of glutamate [50], an amino acid associated with stress responses by bacteria in acidic environments [51]. Chemical processes and microbial activity are connected in the soil environment [52]. Simultaneous analysis of the bacterial community and soil parameters demonstrated that pH is the most significant factor for microbial composition, diversity, and biomass [53]. Soil is usually more acidic on the surface than deeper within [54], and moisture can alter the chemical composition of Cerrado soil [55]. Silva and collaborators also showed that Verrucomicrobia was the phyla most associated to the rainy season in Cerrado, corroborating results in this work. The dry season has the greater equitability (Pielou index) and bacterial richness (OTU number) in comparison to the other seasons. This phenomenon might be related to the fact that the environment is hostile during the dry season, and no specific group is favored. The more homogenous the community, the more resilient it is to external changes [56]. However, low richness communities can be as productive as high richness communities since the species composition and metabolic activity are important variables to be considered [13,57]. Sequencing techniques allow us to identify bacterial species, revealing the microbiota composition and its relationship with abiotic elements without the need to cultivate microorganisms [12]. Functional inferences based on taxonomy have been able to reconstitute the main functional aspects of a community [58–61], despite some limitations [32]. As the scientific community produces more information regarding 16S rRNA gene data and metagenomic sequences, a pattern of two gene cores emerged: a large set of non-essential genes or pseudogenes, as well as a set of specific genes selected by the environment [62]. Metagenomic analysis of mound soil demonstrated that it was functionally similar to that of the adjacent soil, with the exception of virulence, pathogenicity, and defense sequences, which are found in the mound and are responsible for nest protection [38]. The abundance of defense genes is often associated to the extensive reservoir of Actinobacteria found in these mounds [37].

5. Conclusions Termites are known as ecosystem engineers due to their ability to modulate the environment through its activities. In this work, the most relatively abundant bacterial phyla are common to soil samples of both Cerrado and mound soils, and they are: Acidobacteria, Actinobacteria, Chloroﬂexi, Microorganisms 2020, 8, 1482 9 of 14

Planctomycetes, Proteobacteria, Verrucomicrobia, and WPS-2, which differ from the abundant phyla present in the Syntermes wheeleri gut. Interactions between these microbial groups and abiotic elements of the ecosystem have not been studied yet. Diversity and functional redundancy are important aspects of a community that enable the coexistence of different groups that can perform the same process, thus contributing to the ecosystem’s stability. The principal component analysis shows three different soil samples: the Cerrado (50–60 cm), the mound surface (0–10 cm), and the mound soil (50–60 cm). Cerrado soil is considered to be the most diverse among all samples analyzed, according to the Shannon index, confirming the first hypothesis: “The mound-associated soil has a smaller bacterial diversity than the surrounding Cerrado soil”. Seasons are a relevant variable in the Cerrado biome, having great influence on the soil microbiome. Moisture interferes in the bacterial community structure and distribution in Cerrado and mound soil samples. This can be observed by the number of Operational Taxonomic Units (OTUs), which is significantly reducedMicroorganisms as 2020 the, 8, x FOR seasons PEER REVIEW progress for the Cerrado samples (50–60 cm),10 of 15implying that the bacterial richness in this soil is more susceptible to seasonal changes than mound soil samples, Seasons are a relevant variable in the Cerrado biome, having great influence on the soil which corroboratesmicrobiome. the Moisture second interferes hypothesis: in the bacterial “seasonality community does structure not and aff distributionect the mound-associated in Cerrado soil bacterial richness”.and mound The soil results samples. presented This can be in observed this work by the relate number microbial of Operational ecology Taxonomic elements Units of the termite mound soil to(OTUs) abiotic, which elements is significantly present reduced in the as the Brazilian seasons progress Cerrado for the biome. Cerrado Future samples studies(50–60 cm), should address implying that the bacterial richness in this soil is more susceptible to seasonal changes than mound the dynamics ofsoil termite samples, whichactivity corroborates and soil the microbiota. second hypothesis: “seasonality does not affect the mound- associated soil bacterial richness”. The results presented in this work relate microbial ecology Supplementaryelements Materials: of the termiteThe following mound soil itemsto abiotic are elements available present online in the Brazilian at http: //Cerradowww.mdpi.com biome. Future/ 2076-2607/8/10/ 1482/s1, Table S1:studies Chemical should address parameters the dynamics of the of soil termite from activity the Cerradoand soil microbiota. (50–60 cm), mound surface (0–10 cm) and mound (50–60 cm depth) samples in the dry, transition and rainy seasons; Table S2: Predicted enzymes of the Nitrogen cycle fromSupplementary the Cerrado Materials: (60 The cm), following mound items surface are available (0 cm) online and mound at www.mdpi.com/xxx/s1, (60 cm) soil samples. Table S1: Chemical parameters of the soil from the Cerrado (50–60 cm), mound surface (0–10 cm) and mound (50–60 cm Author Contributions:depth) samplesH.I.P.G., in the dry, R.H.S., transition R.S., and andrainy O.H.B.P.seasons; Table performed S2: Predicted theenzy experiments,mes of the Nitrogen data cycle collection,from and data analysis. H.I.P.G.,the Cerrado B.F.Q., (60C.C.B., cm), mound and surface O.H.B.P. (0 cm) and performed mound (60 cm) data soil samples. analysis. H.I.P.G., C.C.B., B.F.Q., M.M.d.C.B., and R.H.K. wroteAuthor the Contributions: manuscript. H.I., R.H.K. R.H.S, R.S. provided, and O.P. performed the funds the experiments, and permissions data collection for, and data data collection analysis. and analysis. All authors haveH.I., read B.F.Q., and C.C.B., agreed and toO.P. the performed published data versionanalysis. H.I., of the C.C.B., manuscript. B.F.Q., M.C.B., and R.H.K. wrote the manuscript. R.H.K. provided the funds and permissions for data collection and analysis. All authors have read Funding: Thisand research agreed to was the published funded version by grants of the manuscript. from Brazilian funding agencies CNPq (National Council for Scientific and TechnologicalFunding: This research Development), was funded by grants Fundaç fromã o Brazilian de Amparo funding agencies a Pesquisa CNPq (National do Distrito Council Federal for (FAP-DF), and CoordinationScientific for the and ImprovementTechnological Development), of Higher Fundação Education de Amparo Personnel a Pesquisa (CAPES).do Distrito Federal (FAP-DF), and Coordination for the Improvement of Higher Education Personnel (CAPES). Acknowledgments: We thank the Brazilian Institute of Geography and Statistics (IBGE RECOR Ecological Reserve) for access to theAcknowledgment sampling sites.s: We thank the Brazilian Institute of Geography and Statistics (IBGE RECOR Ecological Reserve) for access to the sampling sites.

Conﬂicts of Interest:ConflictsThe of Interest: authors The authors declare declare no conﬂictno conflict ofof interest interest.. The Thefunders funders had no role had in the no design role of in the the design of the study; in the collection,study; in the analyses, collection, analyses, or interpretation or interpretation of of data; data; in in the the writing writing of the manuscript, of the manuscript, or in the decision or into the decision to publish the results.publish the results.

Appendix A Appendix A

150 Transição Chuva a) b) 150 others Thermotogae WD272 Actinobacteria Planctomycetes Chloroflexi Acidobacteria Verrucomicrobia Proteobacteria 100 1% 1% e 2% 100 1% c 1% 1% 1%

n 1% 1% 1% a a 10% 1% v i d Relatit a n l 18% 15%

e 13%

u 6%

r 20% 23% b ve a i c a 5% 5% 4%

n 10% â

e 5%

abund v n 9% 4%

i 11% u t

b 14%

dana c 9% A 11% 8% l

e 11% 12% e R 10% 9% 8% 50 15% 50 12% 10% 17% 13% 16%

41% 36% 38% 37% 38% 35%

0 0 Cerrado Mound 0 Mound 60 Cerrado Mound 0 Mound 60

campo sujo campo sujo

cupinzeiro 60 cm cupinzeiro 60 cm cupinzeiro superfície Figure A1. Relative abundance of the bacterial phyla in the rainycupinzeiro season superfície and transition for the Cerrado Figure A1. Relative abundance of the bacterial phyla in the rainy season and transition for the Cerrado (60 cm), mound(60 surfacecm), mound (0 surface cm), and(0 cm) mound, and mound soil soil (60 (60 cm) cm) samples. samples. The samples The samples are shown arein the shown x axis in the x axis and the phylaand in the phylay axis in the for y axis the for transition the transition (a ()a) andand rainy rainy season season (b). ‘Others’ (b). ‘Others’refer to groups refer that to groups that represent less than 1% of the total bacterial community. represent less than 1% of the total bacterial community.

Microorganisms 2020, 8, x FOR PEER REVIEW 11 of 15 Microorganisms 2020, 8, 1482 10 of 14 Microorganisms 2020, 8, x FOR PEER REVIEW 11 of 15

Figure A2. Correlation between the most abundant phyla in the soil and humidity. Linear regression analysisFigureFigure A2. A2between.Correlation Correlation bacterial betweenbetween phyla theandthe mostmostsoil moisture abundantabundant content phylaphyla inforin the thethe soil soilCerrado andand humidity.humidity. soil (60 cm LinearLinear depth), regressionregression mound surfaceanalysisanalysis (0 betweencm), and bacterial bacterialmound soilphyla phyla (60 and cm and soildepth) soil moisture moisture samples. content contentThe fornumbers the for Cerrado the shown Cerrado soil represent (60 soil cm (60 thedepth), cmR value depth),mound in themoundsurface linear surface(0 regression, cm), (0and cm), mound and and the moundsoil * symbol (60 soilcm in depth) (60 red cm are samples. depth) statistically samples. The significantnumbers The numbers shown values. represent shown represent the R value the Rin valuethe linear in the regression, linear regression, and the and* symbol the * symbolin red are in redstatistically are statistically significant signiﬁcant values. values.

A B A B

CCearmradpo sujo DSSreeyc csaeaason MCouupnindz 0eiro 0 cm TRCarhainnuy vseiaçaãsoon MCuopunindz e6i0ro 60 cm TCTrahrnausnivtisaoinção CCearmradpo sujo DSrey cseaason 3000 3000 Seca MCouupnindz 0eiro 0 cm TRCarhainnuy vseiaçaãsoon 2800 MCuopunindz e6i0ro 60 cm 2800 Transition 2600 2600 CThraunvsaição 3000 3000 2400 2400 2800 2800 2200 2200 2600 2600 s 2000 s U 2000 2400 U T 2400 T

O 1800

O 2200 d

1800 e 2200 d 1600 v e r s

1600 v 2000 e s r U s 2000 U e T 1400 b s T

O 1400 1800 O

b O d 1800 1200 O e d 1600 v 1200 e r

1600 v

e 1000 r s

1000 e 1400 b 1400 s O

b 800 1200 800 O 1200 600 600 1000 1000 400 400 800 800 200 200 600 600 0 0 400 400 -5000 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 55000 -5000 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 55000 200 200 0 Cobertura de sequenciamento 0 Sequencing coverage Sequencing coverage -5000 0 5000 10000 1500C0 ober2t0u00r0a de 2s5e0q00uenc3i0a0m00ento35000 40000 45000 50000 55000 -5000 0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 550 00 Cobertura de sequenciamento Sequencing coverage Sequencing coverage Cobertura de sequenciamento FFigureigure A A3.3. AlphaAlpha rarefaction rarefaction curv curvee of of the the observed observed OTUs. OTUs. (A (A) ) Cerrado Cerrado soil soil bacterial bacterial community, community, collectedcollectedFigure A at at3. different diAlphaﬀerent rarefaction seasons seasons (dry, (dry, curv rainy rainy,e of, and the and observedtransition) transition) OTUs. in in 2017 2017 (A and) and Cerrado 2018 2018;; (B soil) ( Bbacterial) bacterial bacterial community community community, in soilincollected soil samples samples at differentof ofthe the Cerrado seasons Cerrado (60 (dry, (60 cm cmrainy depth), depth),, and mound transition) mound surface surface in 2017 (0 (0 cm) and cm),, and2018 and mound; mound(B) bacterial soil soil (60 (60community cm cm de depth)pth) in collectedcollectedsoil samples in in 2017 2017 of andthe and Cerrado2018. 2018. (60 cm depth), mound surface (0 cm), and mound soil (60 cm depth) collected in 2017 and 2018.

Microorganisms 2020, 8, x FOR PEER REVIEW 12 of 15 Microorganisms 2020, 8, 1482 11 of 14 Microorganisms 2020, 8, x FOR PEER REVIEW 12 of 15 0.70

0.70 * 0.65 * 0.65 e

u o

l 0.60 e e

i u P o

l 0.60 e i P 0.55

0.55

0.50

ry n in d io a 0.50 it r s n ry a n in d tr io a it r s n a tr Figure A4. The Pielou number for bacterial equitability in the Cerrado soil (60 cm depth), mound Figure A4. The Pielou number for bacterial equitability in the Cerrado soil (60 cm depth), mound surfaceFigure A(04. cm) The, andPielou mound number soil for(60 bacterial cm depth) equitability samples collectedin the Cerrado for the soil dry, (60 transition cm depth),, and mound rainy surface (0 cm), and mound soil (60 cm depth) samples collected for the dry, transition, and rainy season seasonsurface in (0 2017 cm) ,and and 2018. mound Statistically soil (60 cmsignificant depth) valuessamples (p collected< 0.05) are for indicated the dry, with transition ‘*’. , and rainy in 2017 and 2018. Statistically signiﬁcant values (p < 0.05) are indicated with ‘*’. season in 2017 and 2018. Statistically significant values (p < 0.05) are indicated with ‘*’.

EC number Enzyme 1.18.6.1 nitrogenase CeSrrado EC1 n.7u.7m.1ber nitritEen rzeydmuectase CM0ou ncdm 0 1.18.6.1 nitrogenase CeSrrado nitrite reductase (ammonia CMo6u0n dc 6m0 1..7..27..21 nitriftoer rmeidnugc)tase CM0ou ncdm 0 nitrite reductase (ammonia CMo6u0n dc 6m0 1..7..2..12 nitrite redufcotramsein (gN)O forming) 1.7.2.4 nitrous-oxide reductase 1.7.2.1 nitrite reductase (NO forming) glycerol-3-phosphate 1..17..52..34 nitroduesh-oydxirdoeg erendausectase 1.14.99.39 amgmlyocneirao ml-3o-npohoxspygheantea se 1.1.5.3 dehydrogenase 1.1.47..929.6.39 hyadmromxyolanmiai nmeo dneohoyxdyrgoegneansaese 1.7.2.5 nitric oxide reductase 11.4.7.1.2.1.64 hgyldurtaomxyalatem sinyne tdheahsey d(GroOgGenAaTs)e 11.4.7.1.2.1.53 glutanmitraicte o sxyidneth raesdeu (cGtaOsGeAT) 16..43..11..124 glgulutatmamatinee s ysynnththaeseta (sGeO (GSA)T) 11..44..17..113 glutagmluattaem saytnet hsaysneth (aGsOeGAT) 1.164.3.1.13..225 gmluettahmaninee m soyntohoextyagsen (aGsSe) 1.4.7.1 glutamate synthase 1.14.13.25 methane monooxygenase Figure A5. Principal component analysis (PCoA) of the functional inference data from mound soil Figure(60 cm), A5 mound. Principal surface component (0 cm), and analysis Cerrado (PCoA) soil (60 of cm)the functional samples. The inference principal data component from mound coordinates soil (60 cm),Figurewere mound calculated A5. Principal surface with (0component the cm) Bray–Curtis, and analysisCerrado dissimilarity (PCoA)soil (60 ofcm) matrix the samples. functional andseparated The inference principal the data samples component from in mound red coordinates (Cerrado), soil (60 werecm),blue mound (moundcalculated surface 0), with and (0the orange cm) Bray, and (mound–Curtis Cerrado dissimilarity 60) (Axis–61.25%;soil (60 cm) matrix samples. Axis and 2–13.51%, separate The principaldAxis the samples 3–9.925%).component in red The coordinates (Cerrado), enzymes bluewereshown (mound calculated are common 0), withand toorangethe all Bray samples (mound–Curtis and 60)dissimilarity participate (Axis–61.25%; matrix in the Axis nitrogenand 2 –separate13.51%, cycle dAxis according the samples3– 9.925%). to in KEGG red The (Cerrado), database.enzymes shownblue (mound are common 0), and to orange all samples (mound and 60) participate (Axis–61.25%; in the Axis nitrogen 2–13.51%, cycle Axisaccording 3– 9.925%). to KEGG The database. enzymes Referencesshown are common to all samples and participate in the nitrogen cycle according to KEGG database. References 1. Klink, C.A.; Machado, R.B. Conservation of the Brazilian Cerrado. Conserv. Biol. 2005, 19, 707–713. [CrossRef] References 1.2. KlinkMarques, CA, J.J.G.; Machado Schulze, RB D.G.; (2005) Curi, Conservation N.; Mertzman, of S.A. the Trace Brazilian element Cerrado. geochemistry Conserv in Brazilian Biol 19:707 Cerrado–713. 1. httpsKlinksoils.://doi.org/10.1111/j.1523 Geoderma CA, Machado2004, 121 RB, 31–43. (2005)-1739.2005.00702.x [CrossRef Conservation] of the Brazilian Cerrado. Conserv Biol 19:707–713. 2.3. MarqueshttpsFurely,://doi.org/10.1111/j.1523 P.A.; JJ, Schulze Ratter, J.A.DG, SoilCuri- resources1739.2005.00702.x N, Mertzman and plant SA (2004) communities Trace element of the centralgeochemistry Brazilian in Brazilian Cerrado andCerrado their 2. soils.Marquesdevelopment. Geoderma JJ, SchulzeJ. 121:31 Biogeogr. DG,–43. Curi2009 https://doi.org/10.1016/j.geoderma.2003.10.003 ,N,15 ,Mertzman 97–108. [CrossRef SA (2004)] Trace element geochemistry in Brazilian Cerrado 3.4. Sourcesoils.Scalon, Geoderma PA. M.C.; F and Haridasan, 121:31 JA. R –(2009)43. M.; https://doi.org/10.1016/j.geoderma.2003.10.003 Franco,Soil Resources A. A comparative and Plant Communities study of aluminium of the Central and nutrient Brazilian concentrations Cerrado and in 3. TheirSourcemistletoes Development. PA. on F and aluminium-accumulating JA. J RBiogeogr (2009) Soil , Vol Resources 15 and , No non-accumulating and 1 , PlantBiogeogr Communities Dev hosts. HumidPlant of theTrop Biol. Central 15:972013,– 15Brazilian108, 851–857. Cerrado [CrossRef and] 4. ScalonTheir[PubMed Development. MC,] Haridasan J Biogeogr M, Franco , Vol 15 AC , No (2013) 1 , Biogeogr A comparative Dev Humid study Trop of 15:97 aluminium–108 and nutrient 4.5. concentrationsScalonTestoni, MC, S.A.; HaridasanDe in Almeida,mistletoes M, J.A.;on Franco aluminium Da Silva, AC L.;- (2013)accumulating Andrade, A comparative G.R.P. and non Clay- accumulating mineralogy study of aluminium ofhosts. Brazilian Plant and OxisolsBiol nutrient15:851 with– 857.concentrationsshrinkage https://doi.org/10.1111/j.1438 properties. in mistletoesRevista on Brasileira aluminium-8677.2012.00713.x de Ci-êaccumulatingncia do Solo 2017 and, 41 non, 1–16.-accumulating [CrossRef] hosts. Plant Biol 15:851– 5.6. Testoni857.Mendes, https://doi.org/10.1111/j.1438 SA, I.D.C.; de Almeida Fernandes, JA, M.F.;da Silva- Chaer,8677.2012.00713.x L, G.M.;Andrade Junior, GRP F.B.D.R. (2017) BiologicalClay mineralogy functioning of Brazilian of Brazilian oxisols Cerrado with 5. shrinkageTestonisoils under SA, properties. dideﬀ erentAlmeida vegetation Rev JA, Bras da Cienc types.Silva do L,Plant SoloAndrade Soil 41:12012– 16.GRP, https://doi.org/10.1590/18069657rbcs20160487359 (2017), 183–195. Clay [CrossRefmineralogy] of Brazilian oxisols with 7. shrinkageda Silva, M.R.S.S. properties.Produç Revã oBras de serapilheira, Cienc do Solo biomassa 41:1–16. e diversidade https://doi.org/10.1590/18069657rbcs20160487 de comunidades bacterianas do solo em á reas de Cerrado sob diferentes usos e manejos; University of Brasilia: Brasilia, Brazil, 2004.

Microorganisms 2020, 8, 1482 12 of 14

8. Constantino, R.; Acioli, A.N.S. Termite diversity in Brazil (Insecta: Isoptera). In Soil Biodiversity in Amazonian and Other Brazilian Ecosystems; CABI Publishing: Wallingford, UK, 2006; pp. 117–128. 9. Lee, K.E.; Wood, T.G. Termites and Soils; Academic Press: London, UK, 1971; p. 251. [CrossRef] 10. Singh, K.; Muljadi, B.P.; Raeini, A.Q.; Jost, C.; Vandeginste, V.; Blunt, M.J.; Theraulaz, G.; Degond, P. The architectural design of smart ventilation and drainage systems in termite nests. Sci. Adv. 2019, 5, eaat8520. [CrossRef] 11. Fageria, N.K.; Baligar, V.C. Properties of termite mound soils and responses of rice and bean to nitrogen, phosphorus, and potassium fertilization on such soil. Commun. Soil Sci. Plant Anal. 2005, 35, 2097–2109. [CrossRef] 12. Araujo, J.F.; De Castro, A.P.; Costa, M.M.C.; Togawa, R.C.; Júnior, G.J.P.; Quirino, B.F.; Bustamante, M.M.D.C.; Williamson, L.; Handelsman, J.; Kruger, R.H. Characterization of soil bacterial assemblies in Brazilian savanna-like vegetation reveals acidobacteria dominance. Microb. Ecol. 2012, 64, 760–770. [CrossRef] 13. Lavelle, P.; Spain, A. Soil Ecology Part II, Chapter IV: Functioning of the Soil System; Springer: Dordrecht, The Netherlands, 2001. 14. Turner, J.S. Beyond biomimicry: What termites can tell us about realizing the living building. In Proceedings of the 1st International Conference on Industrialized, Integrated and Intelligent Construction, Loughborough, UK, 14–16 May 2008. 15. Nannipieri, P.; Ascher-Jenull, J.; Ceccherini, M.T.; Landi, L.; Pietramellara, G.; Renella, G. Microbial diversity and soil functions. Eur. J. Soil Sci. 2003, 54, 655–670. [CrossRef] 16. Dar, G.H. Soil Microbiology, Ecology and Biochemistry, 3rd ed.; Elsevier Inc.: Amsterdam, The Netherlands, 2010. 17. Bignell, D.E.; Roisin, Y.; Lo, N. Biology of Termites: A Modern Synthesis; Springer: Berlin/Heidelberg, Germany,2011. 18. Moreira, E.A.; Alvarez, T.M.; Persinoti, G.F.; Paixão, D.A.A.; Menezes, L.R.; Cairo, J.P.L.F.; Squina, F.M.; Costa-Leonardo, A.M.; Carrijo, T.F.; Arab, A. Microbial Communities of the gut and nest of the humus- and litter-feeding termite Procornitermes araujoi (Syntermitinae). Curr. Microbiol. 2018, 75, 1609–1618. [CrossRef] [PubMed] 19. Constantino, R. Padrões de diversidade e endemismo de témitas no bioma Cerrado. In Biodiversidade, Ecologia e Conservação do Cerrado; Ministério do Meio Ambiente: Brasilia, Brazil, 2005; pp. 319–333. 20. Santana, R.H.; Catão, E.C.P.; Lopes, F.A.C.; Constantino, R.; Barreto, C.; Kruger, R.H. The gut microbiota of workers of the litter-feeding termite Syntermes wheeleri (Termitidae: Syntermitinae): Archaeal, bacterial, and fungal communities. Microb. Ecol. 2015, 70, 545–556. [CrossRef][PubMed] 21. Catani, R.A.; Bittencourt, V.C.; de Cangiani, A.M. Extração e determinação do íon amônio do solo. An da Esc Super Agric Luiz Queiroz 1969, 26, 57–73. [CrossRef] 22. Claessen, M.E.C.; De Barreto, W.O. MANUAL de MÉTODOS de ANÁLISE de SOLO; EMBRAPA: Brasilia, Brazil, 1996. 23. Griﬃths, R.I.; Whiteley, A.S.; O’Donnell, A.G.; Bailey, M.J. Rapid method for coextraction of DNA and RNA from natural environments for analysis of ribosomal DNA- and rRNA-based microbial community composition. Appl. Environ. Microbiol. 2000, 66, 5488–5491. [CrossRef][PubMed] 24. Andrews, S. FASTQC A Quality Control Tool for High Throughput Sequence Data; Babraham Bioinforma: Cambridge, UK, 2010. Available online: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ (accessed on 25 September 2020). 25. Bolyen, E.; Rideout, J.R.; Dillon, M.R.; Bokulich, N.A.; Abnet, C.C.; Al-Ghalith, G.A.; Alexander, H.; Alm, E.J.; Arumugam, M.; Asnicar, F.; et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat. Biotechnol. 2019, 37, 852–857. [CrossRef] 26. Martin, M. Cutadapt removes adapter sequences from high-throughput sequencing reads. EMBnet J. 2011, 17, 10. [CrossRef] 27. Bokulich, N.A.; Subramanian, S.; Faith, J.J.; Gevers, D.; Gordon, J.I.; Knight, R.; Mills, D.A.; Caporaso, J.G. Quality-ﬁltering vastly improves diversity estimates from Illumina amplicon sequencing. Nat. Methods 2013, 10, 57–59. [CrossRef] 28. Rognes, T.; Flouri, T.; Nichols, B.; Quince, C.; Mahe, F. VSEARCH: A versatile open source tool for metagenomics. Peer J. 2016, 4, e2584. [CrossRef] 29. Quast, C.; Pruesse, E.; Yilmaz, P.; Gerken, J.; Schweer, T.; Yarza, P.; Peplies, J.; Glöckner, F.O. The SILVA ribosomal RNA gene database project: Improved data processing and web-based tools. Nucleic Acids Res. 2012, 41, D590–D596. [CrossRef] Microorganisms 2020, 8, 1482 13 of 14

30. Oksanen, J.; Blanchet, F.G.; Kindt, R.; Legendre, P.; McGlinn, D.; Minchin, P.R.; O0Hara, R.B.; Simpson, G.L.; Solymos, P.; Henry, M.; et al. Vegan Community Ecology Package. R. Package Vegan Vers. 2.2-1; World Agrofestry Centre: Nairobi, Kenya, 2015. 31. McMurdie, P.J.; Holmes, S. Shiny-phyloseq: Web application for interactive microbiome analysis with provenance tracking. Bioinform. 2014, 31, 282–283. [CrossRef] 32. Langille, M.G.I.; Zaneveld, J.; Caporaso, J.G.; McDonald, D.; Knights, D.; Reyes, J.A.; Clemente, J.C.; Burkepile, D.E.; Thurber, R.L.V.; Knight, R.; et al. Predictive functional profiling of microbial communities using 16S rRNA marker gene sequences. Nat. Biotechnol. 2013, 31, 814–821. [CrossRef] 33. Caspi, R.; Billington, R.; Fulcher, C.A.; Keseler, I.M.; Kothari, A.; Krummenacker, M.; Latendresse, M.; Midford, P.E.; Ong, Q.; Ong, W.K.; et al. The MetaCyc database of metabolic pathways and enzymes. Nucleic Acids Res. 2018, 46, D633–D639. [CrossRef][PubMed] 34. de Araujo, A.S.F.; Mendes, L.W.; Lemos, L.N.; Antunes, J.E.L.; Beserra, J.E.A., Jr.; de Lyra, M.D.C.C.P.; Figueiredo, M.D.V.B.; Lopes, A.C.D.A.; Gomes, R.L.F.; Bezerra, W.M.; et al. Protist species richness and soil microbiome complexity increase towards climax vegetation in the Brazilian Cerrado. Commun. Biol. 2018, 1, 135. [CrossRef][PubMed] 35. Delgado-Baquerizo, M.; Oliverio, A.M.; Brewer, T.E.; Benavent-González, A.; Eldridge, D.J.; Bardgett, R.D.; Maestre, F.T.; Singh, B.; Fierer, N. A global atlas of the dominant bacteria found in soil. Science 2018, 359, 320–325. [CrossRef][PubMed] 36. Rappé, M.S.; Giovannoni, S.J. The uncultured microbial majority. Annu. Rev. Microbiol. 2003, 57, 369–394. [CrossRef] 37. Sujada, N.; Sungthong, R.; Lumyong, S. Termite Nests as an abundant source of cultivable actinobacteria for biotechnological purposes. Microbes Environ. 2014, 29, 211–219. [CrossRef] 38. Enagbonma, B.J.; Aremu, B.R.; Babalola, O.O. Profiling the functional diversity of termite mound soil bacteria as revealed by shotgun sequencing. Genes 2019, 10, 637. [CrossRef] 39. Herve, V.; Liu, P.; Dietrich, C.; Sillam-Dussès, D.; Stiblik, P.; Šobotník, J.; Brune, A. Phylogenomic analysis of 589 metagenome-assembled genomes encompassing all major prokaryotic lineages from the gut of higher termites. Peer J. 2020, 8, e8614. [CrossRef] 40. Hongoh, Y. Toward the functional analysis of uncultivable, symbiotic microorganisms in the termite gut. Cell. Mol. Life Sci. 2011, 68, 1311–1325. [CrossRef] 41. Warnecke, F.; Luginbühl, P.; Ivanova, N.N.; Ghassemian, M.; Richardson, T.H.; Stege, J.T.; Cayouette, M.; McHardy, A.C.; Djordjevic, G.; Aboushadi, N.; et al. Metagenomic and functional analysis of hindgut microbiota of a wood-feeding higher termite. Nature 2007, 450, 560–565. [CrossRef] 42. Manjula, A.; Pushpanathan, M.; Sathyavathi, S.; Gunasekaran, P.; Rajendhran, J. Comparative Analysis of Microbial Diversity in Termite Gut and Termite Nest Using Ion Sequencing. Curr. Microbiol. 2015, 1–9. [CrossRef][PubMed] 43. Minkley, N.; Fujita, A.; Brune, A.; Kirchner, W.H. Nest specificity of the bacterial community in termite guts (Hodotermes mossambicus). Insectes Sociaux 2006, 53, 339–344. [CrossRef] 44. Kirchner, W.H.; Minkley, N. Nestmate discrimination in the harvester termite Hodotermes mossambicus. Insectes Sociaux 2003, 50, 222–225. [CrossRef] 45. Robertson, L.A.; Kuenen, J.G. The Prokaryotes; Springer: New York, NY, USA, 2006; Volume 5. 46. Brown, C.M.; Dilworth, M.J. Ammonia assimilation by rhizobium cultures and bacteroids. J. Gen. Microbiol. 1975, 86, 39–48. [CrossRef] 47. Kennedy, I.R. Primary products of symbiotic nitrogen fixation. II. Pulse-labelling of serradella nodules with 15N2. Biochem. Biophys. Acta Gen. Subj. 1966, 130, 295–303. [CrossRef] 48. WoolFolk, C.; Shapiro, B.; Stadtman, E. Regulation of glutamine synthetase. Arch. Biochem. Biophys. 1966, 116, 177–192. [CrossRef] 49. Fildes, P.; Gladstone, G.P. Glutamine and the Growth of Bacteria. Br. J. Exp. Pathol. 1939, 20, 334–341. 50. Vanoni, M.A.; Curti, B. Glutamate synthase: A complex iron-sulfur flavoprotein. Cell. Mol. Life Sci. 1999, 55, 617–638. [CrossRef] 51. Feehily, C.; Karatzas, K.-A. Role of glutamate metabolism in bacterial responses towards acid and other stresses. J. Appl. Microbiol. 2012, 114, 11–24. [CrossRef] Microorganisms 2020, 8, 1482 14 of 14

52. Baveye, P.C.; Otten, W.; Kravchenko, A.; Balseiro-Romero, M.; Beckers, É.; Chalhoub, M.; Darnault, C.; Eickhorst, T.; Garnier, P.; Hapca, S.; et al. Emergent properties of microbial activity in heterogeneous soil microenvironments: Different research approaches are slowly converging, yet major challenges remain. Front. Microbiol. 2018, 9, 1–48. [CrossRef] 53. Zhalnina, K.; Dias, R.; De Quadros, P.D.; Davis-Richardson, A.; Camargo, F.A.O.; Clark, I.M.; McGrath, S.P.; Hirsch, P.; Triplett, E.W. Soil pH determines microbial diversity and composition in the park grass experiment. Microb. Ecol. 2014, 69, 395–406. [CrossRef] 54. Reeves, J.L.; Liebig, M. Depth matters: Soil pH and dilution effects in the Northern Great Plains. Soil Sci. Soc. Am. J. 2016, 80, 1424–1427. [CrossRef] 55. Da Silva, M.R.S.S.; De Castro, A.P.; Kruger, R.H.; Bustamante, M.M.D.C. Soil bacterial communities in the Brazilian Cerrado: Response to vegetation type and management. Acta Oecologica 2019, 100, 103463. [CrossRef] 56. Wittebolle, L.; Marzorati, M.; Clement, L.; Balloi, A.; Daffonchio, D.; Heylen, K.; De Vos, P.; Verstraete, W.; Boon, N. Initial community evenness favours functionality under selective stress. Nature 2009, 458, 623–626. [CrossRef][PubMed] 57. Bell, T.; Newman, J.A.; Silverman, B.W.; Turner, S.L.; Lilley, A.K. The contribution of species richness and composition to bacterial services. Nature 2005, 436, 1157–1160. [CrossRef][PubMed] 58. Ferreira, R.M.; Marques, J.P.; Pinto-Ribeiro, I.; Costa, J.–L.; Carneiro, F.; Machado, J.C.; Figueiredo, C. Gastric microbial community profiling reveals a dysbiotic cancer-associated microbiota. Gut 2017, 67, 226–236. [CrossRef] 59. Sun, C.; Yuan, T.; Chen, L.; Xie, Z.; Shen, Z. Occurrence of potentially pathogenic bacteria on shared bicycles. Int. J. Hyg. Environ. Heal. 2020, 224, 113442. [CrossRef] 60. Boon, E.; Meehan, C.J.; Whidden, C.; Wong, D.H.-J.; Langille, M.G.I.; Beiko, R.G. Interactions in the microbiome: Communities of organisms and communities of genes. FEMS Microbiol. Rev. 2014, 38, 90–118. [CrossRef] 61. Bill, M.; Chidamba, L.; Gokul, J.K.; Labuschagne, N.; Korsten, L. Bacterial community dynamics and functional profiling of soils from conventional and organic cropping systems. Appl. Soil Ecol. 2020, 157, 103734. [CrossRef] 62. Falkowski, P.G.; Fenchel, T.; Delong, E.F. The Microbial Engines That Drive Earth’s Biogeochemical Cycles. Science 2008, 320, 1034–1039. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).