Microbial and metabolic profiling reveal strong influence of water

15 Pages·2014·3.46 MB·English

Biogeosciences,11,1727–1741,2014 O
p
www.biogeosciences.net/11/1727/2014/ e
n
doi:10.5194/bg-11-1727-2014 Biogeosciences A
c
c
©Author(s)2014.CCAttribution3.0License. e
s
s
Microbial and metabolic profiling reveal strong influence of water
table and land-use patterns on classification of degraded tropical
peatlands
S.Mishra1,2,3,W.A.Lee3,A.Hooijer4,S.Reuben3,I.M.Sudiana5,A.Idris6,andS.Swarup1,2,3,7
1MetabolitesBiologyLaboratory,DepartmentofBiologicalSciences,NationalUniversityofSingapore(NUS),117543,
Singapore
2NUSEnvironmentalResearchInstitute,NUS,T-LabBuilding,5AEngineeringDrive1,117411,Singapore
3Singapore-DelftWaterAllianceBlockE1,No1EngineeringDrive2,NUS,117576,Singapore
4Deltares,P.O.Box177,2600MHDelft,theNetherlands
5CibinongScienceCenter,LIPI,Jl.RayaBogorKm46,CibinongBogor,16911,Indonesia
6FacultyofAgriculture,UniversityofJambi,Jambi,36122,Indonesia
7SingaporeCentreonEnvironmentalLifeSciencesEngineering(SCELSE),60NanyangDrive,NTU,637551,Singapore
Correspondenceto:S.Swarup([email protected])
Received:21July2013–PublishedinBiogeosciencesDiscuss.:26August2013
Revised:18January2014–Accepted:9February2014–Published:3April2014
Abstract.TropicalpeatlandsfromsoutheastAsiaareunder- could distinguish communities not only based on land-use
going extensive drainage, deforestation and degradation for types but also their geographic locations, thus providing a
agricultureandhumansettlementpurposes.Thisisresulting finer resolution than bacterial profiles. Agricultural inputs,
in biomass loss and subsidence of peat from its oxidation. such as nitrates, were highly associated with bacterial com-
Molecularprofilingapproacheswereusedtounderstandthe munitystructureofoilpalmplantations,whereasphosphates
relative influences of different land-use patterns, hydrologi- and dissolved organic carbon influenced those from mixed
calandphysicochemicalparametersonthestateofdegraded crop plantations and settlements. Our results provide a ba-
tropicalpeatlands.Asmicrobialcommunitiesplayacritical sisforadoptingmolecularmarker-basedapproachestoclas-
role in biogeochemical cascades in the functioning of peat- sify peatlands and determine relative importance of factors
lands,weusedmicrobialandmetabolicprofilesassurrogates that influence peat functioning. Our findings will be useful
of community structure and functions, respectively. Profiles inpeatlandmanagementbyprovidingabasistofocusearly
weregeneratedfrom230bacterial16SrDNAfragmentsand effortsonhydrologicalinterventionsandimprovingsustain-
145metabolicmarkersof46samplesfrom10sites,includ- ability of oil palm plantations by adopting mixed cropping
ingthosefromaboveandbelowwatertableinacontiguous practicestoincreasemicrobialdiversityinthelongterm.
areaof48km2 coveringfiveland-usetypes.Thesewerede-
gradedforest,degradedland,oilpalmplantation,mixedcrop
plantation and settlements. Bacterial profiles were most in-
fluenced by variations in water table and land-use patterns, 1 Introduction
followed by age of drainage and peat thickness in that or-
der.Bacterialprofilingrevealeddifferencesinsites,basedon Peatlands are formed by the accumulation of partially de-
the duration and frequency of water table fluctuations and cayed vegetation matter over over millennial timescales in
onoxygenavailability.Mixedcropplantationshadthemost low-lyingareasthatarefrequentlywaterloggedduetoheavy
diverse bacterial and metabolic profiles. Metabolic profil- rainfall or periodic inundation (Anderson, 1964). Peatlands
ing,beingcloselyassociatedwithbiogeochemicalfunctions, areahighlyvulnerablenaturalresourcethatcover50–70%
of global wetlands (Finlayson et al., 1999) and sequester
PublishedbyCopernicusPublicationsonbehalfoftheEuropeanGeosciencesUnion.
1728 S.Mishraetal.:Hydrologyandland-usechangeaffectsmicrobialecologyofpeatlands
one-thirdoftheworld’ssoilcarbon(Freemanetal.,2012).In In most ecosystems, changes in land-use patterns impact
southeastAsia,peatlandscoveranareaofnearly25million both microbial diversity and their activity (Wardle et al.,
haandstoreapproximately69Gtofcarbon,whichis77%of 1998;Ollivieretal.,2011;Putten2012).Watertabledepth,
allthetropicalpeatlandcarbonpool(88.6Gt),ofwhich65% which directly affects the oxygen levels in the peat layer
(57.4Gt of carbon) is in Indonesia itself, distributed within (Lahde, 1969), is an important factor in shaping bacterial
23.4millionha of peatland (Page et al., 2011). Carbon den- community structure. Water table depth also affects water
sity is relatively high in tropical compared to temperate or stress, which has been shown to have direct and indirect
borealpeatlands;thisislargelyaconsequenceofdeeperpeat influences on soil bacterial community composition (Fierer
layersintheformer,withpeatthicknessupto20m(Pageet etal.,2003).Whilegeochemicalconditionsaffectmicrobial
al.,2002). communities, they, in turn, affect their environment. There-
Southeast Asian peatlands are under threat from anthro- fore,itisimportanttostudymicrobialcommunitycomposi-
pogenicactivities,predominantlyduetodrainageanddefor- tion.Molecularprofilingapproacheshavebeenwidelyused
estationforagricultureandhumansettlementpurposes.For- to describe microbial diversity in soil, rhizosphere, extreme
est fires and biomass burning linked to land-use change ex- environments, freshwater and mangrove ecosystems (Bai et
acerbatethesethreats.ThesefireeventsinIndonesiacrossed al., 2012; Chan et al., 2013; He et al., 2012; Nold et al.,
allpastrecords,hitting371inthePollutantStandardsIndex 2000; Xiong et al., 2010). Microbial profiling approaches
(PSI)inSingapore,basedondailyaverage,duringthesmog thatarenowwidelyusedmainlyrelyonDNAfingerprinting
eventthatengulfedneighboringcountries,suchasSingapore (Zhou,2003;Nockeretal.,2007;Nazariesetal.,2013)oron
and Malaysia (Schmaltz, 2013). Such land-use changes and pyrosequencing of ribosomal DNA region (Chistoserdova,
hydrologicalinterventionshaveresultedinadrasticdecrease 2010;Heetal.,2010).Morerecently,metabolicprofilinghas
in peatland water tables, exposing the biomass sequestered alsoemergedasausefulapproachtoreportstatusofmicro-
in the peat to aerobic microbial oxidation. Of the total area bial functions in soil (Bundy et al., 2009) and rhizosphere
of peatlands in Indonesia, at least 2.2millionha have been (Leeetal.,2013).Tworecentstudiesofintact,forestedpeat-
converted to commercial oil palm plantations, which is ex- landsfromThailand(Kanokratanaetal.,2011)andMalaysia
pected to increase to 6.2millionha by 2020 (Miettinen et (Jackson et al., 2009) used pyrosequencing and fingerprint-
al.,2012a).Interspersedwithintheplantationsaresmallar- ingapproachestodescribetheirmicrobialdiversityandfunc-
easoftemporaryhumansettlements.InpeninsularMalaysia tional properties, respectively. They demonstrated the capa-
and the islands of Sumatra and Borneo, some 60% of peat bility of such techniques to show broad phylogenetic diver-
swamps had been partly or completely deforested by 2007, sityandgeneticpotentialtodegradebiomass,respectively.
usually accompanied by some form of drainage, and only Indegradedtemperatepeatlands,i.e.,thosepeatlandsthat
10% remained in pristine condition (Miettinen and Liew, haveundergonemassiveland-usechangeduetodrainageand
2010).Inrecentyears,rapidlyincreasingpeatcarbonlosses deforestation, much is known about the relationship of mi-
fromdrainedpeatlandsinsoutheastAsiahavebeenfoundto crobialdiversitytopeatlandfunctioningandgreenhousegas
contribute significantly to global greenhouse gas emissions emissions(Opeltetal.,2007;Ausecetal.,2009;Kimetal.,
(Melling et al., 2005; Furukawa et al., 2005; Couwenberg 2012; Tveit et al., 2013). In contrast, for tropical peatlands,
et al., 2010). Estimates of net carbon losses and resultant wehavearelativelypoorunderstandingoftherelationshipof
CO emissions from peatland drained for agriculture range microbialdiversityandfactorsinfluencingcommunitystruc-
2
from 30–40tCO ha−1yr−1 (Murdiyarso et al., 2010; Her- ture for intact as well as tropical peatlands under land-use
2
goualc’handVerchot,2011)toashighas70tCO ha−1yr−1 change.Givenboththeecologicalandeconomicimportance
2
(Couwenberg et al., 2010; Jauhiainen et al., 2012), exclud- ofthesepeatlands,itwillbeusefultounderstandthediffer-
ing forest biomass losses, fire losses and peat organic mat- ences among various land-use patterns in degraded tropical
ter losses in the initial years after drainage. Carbon losses peatlandwithrespecttomicrobialecology.Towardsthisdi-
from such emissions and through fluvial processes have led rection,itiscriticaltodevelopscientificmethodstomanage
totropicalpeatlandsbeingtransformedfromcarbonsinksto therapidchangeinlandusefrompristineconditionsofpeat-
carbonsources(Mooreetal.,2013).Theoxidationofdrained lands and monitor the progress as well as effectiveness of
peatiscausingrapidsubsidencebydisappearanceofthesur- ecosystemrestorationinterventions.Inconjunctionwithex-
face layers in the peat (Kool et al., 2006; Couwenberg and isting methods, such as field surveys of flora and fauna of
Hooijer, 2013). In a recent study from the same region of tropicalpeatlandsthatareundergoingland-usechange(Posa
Sumatra,peatwasreportedtosubsideatarateof5cmyr−1, etal.,2011),thesecombinedapproachesneedtobeadopted
ofwhich92%lossisduetooxidationandnotcompactionor toclassifypeatinordertoconserveremainingpristinepeat-
plantrespiration(Hooijeretal.,2012).Oxidationoforganic lands.Ourapproachisbasedontheabilityofmolecularpro-
matterhasbeenshownusingmesocosmsofborealpeattobe filingtocaptureshiftsincommunitystructureandmetabolic
due to stimulation of microbial growth, thereby causing the profilingtoreflectthefunctionaloutcomeofmetabolicactiv-
breakdown of organic matter and release of carbon dioxide itiesofmicrobes,plantrootsandtheirexudates,respectively.
inabiogeochemicalcascade(FennerandFreeman,2011). Using these two microbial profiling approaches, we report
Biogeosciences,11,1727–1741,2014 www.biogeosciences.net/11/1727/2014/
S.Mishraetal.:Hydrologyandland-usechangeaffectsmicrobialecologyofpeatlands 1729
theeffectsofwatertabledepthandoxygenavailability,land- tation”, depending on land cover in that area; and (5) small
usepatterns,ageofdrainageandpeatthicknessonbacterial holdermosaicandbuilt-upareas(188km2,or6%)wasclas-
diversity in degraded peatlands of Indonesia. We focus on sified as “settlements”. For this study, the contiguous land-
five land-use patterns from a contiguous study site: (a) de- usetypeswerechosenthatwerepresentinSiteAandSiteB.
gradedforest,whichincludesdrainedandheavilydeforested Intact PSF was not included in this study because of being
peatswampforest;(b)degradedland,whichincludesdefor- non-contiguous.
ested and drained peatlands that have yet to undergo con- The coordinates of sampling locations distributed across
version for agricultural use; (c) oil palm plantation, which two broad areas, referred to as Site A and Site B,
includes peatland area under palm plantations; (d) settle- were 103◦53052.5800E, 1◦43012.4700S and 103◦49032.2300E,
ments,whichincludespeatlandareaunderpalmplantations 1◦40058.2400S,respectively(Fig.1).Thetotalareascovered
interspersed with human settlements; and (e) mixed crop by the sites were 42 and 6km2, respectively. Out of five
plantation, which includes peatland area under palm plan- land-use patterns, degraded land with similar peat physical,
tations,pineappleandtapioca.Inordertodetermineimpor- geological and hydrological conditions was present in both
tanceofvariousmanagementpracticesofthesepeatlandson SiteAandSiteB(Fig.1).Allstudysites(Fig.1)havebeen
the structure and functioning of the bacterial communities, affected by fire in the past. The only exception is the de-
westudiedtheinfluencesof11physicochemicalparameters. graded forest (DHFN: Deep peat depth, High water table,
Basedonourfindingsreportedhere,wemakerecommenda- Degraded Forest and New drainage <10years; please re-
tions that will help in the classification, improved manage- fer to legends of Fig. 1 for acronyms of site description).
mentandsustainabilityoftropicalpeatlands. The fire events have occurred in the past in degraded land
(DHAN) in Site A and B in 2004 and 2005, respectively.
Burningoccurredinoilpalmandmixedcropplantationsites
2 Materialsandmethods inSiteB(MHPN,DHPN,MHXNandSHXN)in2004,sites
withoilpalmplantationsinSiteAinlowwatertabledepth
2.1 Sitedescriptionandsampling (MLPO) in 2001, and sites in oil palm plantations (DHPO)
andinsettlements(DHTO)in2000.Aspartofroutineman-
The study area is located in peatlands of the eastern part of agement practices, fertilizers are applied to the sites that
Jambiprovince,Sumatra,Indonesia(Fig.1).Forestedtropi- fall within oil palm and mixed crop plantations. The main
calpeatlandsareextensiveinthisareaandavarietyofland- categories of fertilizer are nitrogen–phosphorus–potassium
use patterns are present due to land intensification activi- (NPK 16:16:16) and urea, which are applied three times
ties. Land-cover classification was performed using visual ayear,andpotassiumchloride(KCl),whichisappliedonce
image interpretation and manual on-screen delineation of ayear.
land-cover polygons. The classification scheme was mainly Inordertomonitorthehydrologicalparameters,bothrain-
basedonvariationinphysicalvegetationcharacteristics(e.g., fall and water table depths were measured periodically us-
height,sparseness,etc.)andincludedthemainphasesofthe ing rain gauges at strategic locations and dipwells in each
tropicalpeatlandconversionanddegradationprocesses(Mi- transect.ThedipwellsconsistedofperforatedPVCpipesan-
ettinen et al., 2012b). The Landsat image and base maps chored into peat, reaching the mineral subsoil. These dip-
of field sites were obtained from the University of Jambi, wellswereusedtomonitorwatertabledepthchangesevery
Jambi, Indonesia. The coordinates of monitoring sites were two weeks, since 2009. Rainfall measurements were moni-
recorded using handheld global positioning system (GPS) tored on a daily basis. Average water table depths and total
devices.Thecoordinatesofthesesiteswereinsertedontothe rainfall were calculated for every month. Data from August
base maps using ArcView and ArcMap programs (ArcGIS- 2009 to August 2010 are being reported in this study. Sam-
Esri,CA,USA). pling was performed in March 2010, preceding which the
The overall mapped area in the eastern part of Jambi highestmonthlyrainfall(370±25mm)intheperiodstudied
comprised a total of 3390km2 (Fig. 1), out of which wa- wasrecordedinFebruary2010.
ter/seasonal water comprised 11km2, or 0.3% of total At each sampling location, a 1m3 pit was dug. Three
mapped area. The land-cover classes used in Miettinen et equidistant pits were used for sampling in each transect.
al. (2012b) were regrouped and reclassified for this study. These transects ranged between 120 and 550m at differ-
The land cover comprising (1) slightly and moderately de- entsamplinglocations.Wetooksamplesatapredetermined
graded peat swamp forest (PSF) (1656km2, or 49% of to- distance from the water table along the wall of the pit. At
tal mapped area) was classified as “intact PSF”; (2) heav- thisdistance,wereachedhorizontally10cmintothepeatto
ily degraded PSF and secondary forest (543km2, or 16%) collect least disturbed samples. This process was repeated
was classified as “degraded forest”; (3) shrubs, fern/grass for each of the four walls of the pit at the same distance
andclearancearea(712km2,or21%)wasclassifiedas“de- from the water table. A composite was then prepared us-
graded land”; (4) commercial plantations (279km2, or 8%) ing these four samples. The number of specified distances
wasclassifiedas“oilpalmplantation”or“mixedcropplan- in sampling varied according to the water table depths in
www.biogeosciences.net/11/1727/2014/ Biogeosciences,11,1727–1741,2014
1730 S.Mishraetal.:Hydrologyandland-usechangeaffectsmicrobialecologyofpeatlands
Fig.1
Fig. 1. Map showing Sumatra Island in Indonesia (top left) and land cover present in the eastern part of Jambi province (right). Study
◦ 0 00 ◦ 0 00 ◦ 0 00
sitesinthisregionofJambi,SiteAandSiteB,arelocatedatgeographicallocations103 5352.58 E,1 4312.47 Sand103 4932.23 E,
1◦40058.2400S,respectively.Abbreviationsareasfollows:D,deeppeatthickness(>7m);M,mediumpeatthickness(3–7m);S,shallowpeat
thickness(<3m);H,highwatertable(0–45cm);L,lowwatertable(>45cm);F,degradedforest;A,degradedland;P,oilpalmplantations;
T,settlements;X,mixedcropplantations;N,drainage(<10years);O,drainage(>10years).Land-usepatternsdescribedinthisstudyare
indicatedwithinbracketsinitalicsinthelegendboxabove.
different sites. Samples using the above mentioned strategy plantationwasflooded;hence,onlyonesamplewascollected
were collected at a distance of 20–30cm above water table BWTandnonefromtheAWTposition.Peatwatersamples
(AWT)andfrom20–30cmbelowwatertable(BWT)inster- formetabolicprofilingwerecollectedfromdipwellsadjacent
ile50mLtubesfromallsites,withexceptionsat4sites.At toeachpit.Allsampleswereshippedonicetothelaboratory
threeoilpalmplantationsitesintheMLPOtransect(Fig.1), and processed immediately. In order to analyze the oxygen
the water table was extremely low (80cm below peat sur- availabilityateachsamplingpoint,anOX-NClark-typeoxy-
facelevel);hencesampleswerecollectedfrom20–30cmand gensensor(Unisense,Aarhus,Denmark)wasusedanddata
50cmAWT,respectively.Onelocationwithinamixedcrop wererecordedmanually
Biogeosciences,11,1727–1741,2014 www.biogeosciences.net/11/1727/2014/
S.Mishraetal.:Hydrologyandland-usechangeaffectsmicrobialecologyofpeatlands 1731
2.2 Bacterialcommunitystructure(terminal calcium using an ion chromatography analyzer (ICS-5000,
restrictionfragmentlengthpolymorphism Dionex).
(T-RFLP)analysis) PeatwatersampleswererunthroughaSolidPhaseExtrac-
tion cartridge using an Oasis® HLB cartridge (1cc/30mg;
Bulk peat gDNA was extracted using a ZR Soil Mi- 30µm, Waters, USA) in order to extract, concentrate and
crobe DNA MidiPrep™ extraction kit (Zymo Research clean up the metabolites (Parab et al., 2009). The samples
Corporation, Irvine, USA) based on the manufacturer’s were analyzed using ultra-performance liquid chromatogra-
protocol, with minor modifications. The extracts were phy (UPLC) in a Waters ACQUITY UPLC™ system (Wa-
quantified spectrophotometrically (Nanodrop ND-1000, ters Corp., MA, USA), equipped with a binary solvent
Nanodrop Technologies, Wilmington, DE, USA). Bacterial delivery system and an autosampler. The chromatography
16S rRNA genes were amplified using universal primer, was performed on a Waters ACQUITY C 1.7µm column
18
BSF517-GCCAGCAGCCGCGGTAA and BSR1541/20- (100×2.1mm).Massspectrometrywasperformedbasedon
AAGGAGGTGATCCAGCCGCA (Wilmotte et al., 1993). MSconditionsusingamassspectrometerequippedwithan
For T-RFLP analysis, forward primer was labeled with electrospray ionization source (UPLC-TOF-Bruker Dalton-
6-carboxyfluorescein (FAM) at the 50 end and reverse ics).DatawereextractedusingBrukerDaltonicsprofileanal-
primer was labeled with photo-induced electron transfer ysissoftware.
(PET)atthe30 end.PCRwasperformedintriplicate(50µL
reaction) using 50ng of template DNA and the following
parameters:initialdenaturationat95◦Cfor10min,followed 2.4 Dataanalysis
bydenaturation95◦Cfor1min,annealingat58◦Cfor30s,
extensionat72◦Cfor1minandfinalextensionat72◦Cfor Toanalyzethevariationinbacterialcommunitystructureas
7 min. Agarose gel electrophoresis, followed by staining well as difference in metabolic functions, due to influence
of products with SYBR Green (Invitrogen, USA) was per- fromanalyzedparameters(namely,watertable,land-usepat-
formed to check amplified product size and concentration. terns, age of drainage and peat thickness), multivariate sta-
Amplicons from three replicates of PCR were pooled and tisticaltechniques(PRIMER6,PRIMER-E,Ltd.,Plymouth,
cleaned up using NucleoSpin® Extract II according to United Kingdom) were used to calculate distance matrices
the manufacturer’s instructions. Five hundredng of each usingBray–Curtissimilarityindicesandone-wayANOSIM
amplicon were digested with restriction enzymes Alu I and (Analysis of Similarity) coefficients (Legendre and Legen-
Bsu RI (Fermantas) at 37◦C for 16h. Digests were then dre, 1998). Unconstrained ordination plots with 100 iter-
purifiedusingtheNucleoSpin® ExtractIIkit,andanaliquot ations using nonmetric multidimensional scaling (nMDS)
of 1µL was mixed with 8.5µL HiDi formamide (Applied based on Bray–Curtis similarity were used to represent the
Biosystems, Foster City, CA, USA) and 0.5µL of internal outcome (Kruskal 1964, Shepard 1962). To analyze the rel-
size standard (Applied Biosystems) for T-RFLP reactions. ative influence of different parameters over bacterial com-
The labeled terminal-restriction fragments (TRFs) were munities, two-way ANOSIM (Clarke, 1993) was used. The
detected on an ABI 3730XL automatic DNA sequencing global R statistic value (generated using one-way or two-
machine (Applied Biosystems) in the GeneScan mode. way ANOSIM) indicates the degree of separation between
For data collection from the DNA sequencing machine, the two communities, with values close to unity indicating
GeneMapper software (Applied Biosystems) was used to moreseparationandazerovalueindicatingnodifferencebe-
comparerelativelengthsofTRFswiththeinternalsizestan- tweenthegroups.
dard.Forprofilecomparison,minimalandmaximalcut-offs To analyze influences of geochemical traits on the bacte-
of 50 and 600bp, respectively, were set and fragments with rial community structure, CCA (Canonical Correspondence
peakheightbelow75wereremovedasfilternoise. Analysis) was performed using Canoco (version 4.5 for
Windows, PRI Wageningen, the Netherlands) (Lepš and
2.3 Chemicalanalysis Šmilauer, 2003). Presence/absence of TRFs was used as
“species”data.Geochemicaldata(anions,cations,DOCand
Five grams of peat were mixed with 40mL of analytical inorganic carbon) were included in the analysis as “envi-
gradewater.Themixturewasshakenat200rpmovernightto ronmental” variables. Ordination biplots approximating the
obtain a bioavailable extract for microorganisms from peat weighteddifferencesbetweentheindividualcommunities(T-
(Reynolds and Clarke, 2008). These extracts were used for RFLPpatterns)withrespecttoeachofthegeochemicalfac-
analysis of total dissolved organic carbon (DOC) and inor- tors(representedasarrows)weredrawn.Therelativeimpor-
ganic carbon using a total organic carbon analyzer (TOC-V tance of geochemical factors in explaining variation in the
CPH E200V 220V, Shimadzu). Aliquots from the same re- bacterialT-RFLPprofileswasexplainedbythelengthofthe
mainingextractswerealsoanalyzedforanionssuchasfluo- corresponding arrows, and the angle between arrows indi-
ride, chloride, nitrite, nitrate, phosphate, sulfate and cations cated the degree to which they were correlated. The impact
such as sodium, ammonium, potassium, magnesium, and ofgeochemicalvariablesoverbacterialcommunitystructure
www.biogeosciences.net/11/1727/2014/ Biogeosciences,11,1727–1741,2014
1732 S.Mishraetal.:Hydrologyandland-usechangeaffectsmicrobialecologyofpeatlands
was calculated using a Monte Carlo permutation test based 3 Results
on1000randompermutations(Rascheetal.,2011).
To predict the phylogeny of the bacterial species at the 3.1 Influencesofpeatcharacteristicsonbacterial
taxa level from the TRFs of 16S rDNA, Fragment Sorter communitystructure
Suite(FRAGSORT)(ver.5.0;AgriculturalResearchandDe-
velopment Center, Ohio State University) and Phylogenetic 3.1.1 Oxygenavailabilityandwatertable
AssignmentTool(PAT)(Kentetal.2003)wereused,adopt-
ingthemethodologydescribedinLefebvreetal.(2010).Mi- Oxygen availability was lower in the BWT samples com-
crobialCommunityAnalysis(MiCA)–avirtualdigestpro- paredtoAWTsamplesbyafactorofthreeormore(Fig.2a).
gram(Shyuetal.,2007)–wasusedtoconstructareference The BWT oxygen levels were similar across all land-use
databaseforeachsetofprimers. types.Basedonthis,henceforth,weusetheterms“oxic”and
“anoxic” conditions to refer to oxygen availability in AWT
2.5 Clonelibrarysequencingof16SrDNAgene and BWT zones, respectively. Effects of oxygen levels on
bacterialDNAprofileswereanalyzed(Fig.2b).Intheordi-
In order to validate the presence of predicted species eval-
nation plot, samples from the low-water-table sites (across
uated from FRAGSORT analysis, a clone library using
all oil palm plantations) tended to cluster together, regard-
16S rDNA from two randomly chosen sites that differed in
less of oxicand anoxic zones. The remaining samples clus-
water table, land-use pattern and oxic conditions was pre-
tered roughly into oxic and anoxic zones, although samples
pared.Thesiteschosenbelongtosettlementswithhighwater
from degraded land from anoxic zones were mixed within
tableandoilpalmplantationswithlowwatertable.Thesam-
oxic zones. BWT points found within AWT group are of
ples were taken from both oxic and anoxic zones of settle-
two types. One of these BWT groups is in a cluster that is
mentsandonlyfromoxiczonesofoilpalmplantationsites.
highly influenced by low water table (as shown in Fig. 2b).
ThecleanedPCRproduct(usingsamenon-labeledprimer
This explains why this subgroup has both AWT and BWT
sequences as described earlier) of 16S rDNA from set-
pointsallcomingfromlow-water-tablesites.Oftheremain-
tlements and oil palm plantation sites was cloned into
ing8BWTpointswithintheAWTgroup,3belongingtooil
3956bppCR® 4-TOPO® vector using TOPO TA Cloning
palmplantationsitesarerightattheperipheryoftheBWT-
kit for sequencing (Invitrogen, USA) according to the man-
specificgroup.Anotheronesample(markedbyarrow)isan
ufacturer’s protocol. One Shot® TOP 10 Competent Cells
outlierwhichwasduetoflooding.Thisleaves4BWTpoints
(Invitrogen, USA) was used in order to transform the re-
belongingtodegradedland,whichwebelievehavehighwa-
combinant plasmid. DNA Sequencing was performed on a
ter table fluctuations as the water table is not controlled by
DNA sequencer (ABI 3130 Genetic Analyzer) using for-
hydrologicalinterventions.Whileforestedareasarealsonot
ward or reverse M13 primers, on plasmid DNA extracted
managed,theyareunlikelytohavelargefluctuationsdueto
usingWizard® PlusSVMiniPrepDNAPurificationSystem
closedcanopy.Thus,thedegradedforestBWTpointfallsin
(Promega,USA)fromindividualclones.DNAsequencedata
theBWTspecificgroup.Anoxiczonessupportedmorecom-
wereanalyzedasdescribedpreviously(Reubenetal.,2012).
plex bacterial communities than oxic zones, although this
Briefly, sequences were trimmed and edited using MEGA5
was not the case for samples from degraded land (Table 1).
(Tamura et al., 2011). MAFTT (Katoh et al., 2009) was
ThepHvaluesmeasuredfromallsitesdidnotshowanysig-
used for aligning the sequences and identifying reverse ori-
nificantcorrelationwithShannondiversityindicesfromdif-
entation. Sequences were then reverse-complemented using
ferentland-usetypesatbothoxicandanoxiczones.
MEGA5. Vector contamination was checked using the vec-
Continuous monitoring of the water table and rainfall re-
tor screening tool in Sequin (http://www.ncbi.nlm.nih.gov/
vealed that variation in water table over the period (August
Sequin/sequin.hlp.html),andachimeracheckwasperformed
2009–August2010)wasinfluencedbyrainfallinthatperiod,
usingBellerophon(Katohetal.,2009)followedbyMallard withmaximalrainfallinFebruary2010of370±25mm,av-
1.2. (Ashelford et al., 2006). Sequences of 302 clones were
eragedoverallsites(Fig.3a).Samplingwasperformeddur-
submittedtoGenBank,andthenucleotidesequencedatare-
ing the time when water table was highest in the entire pe-
portedinthispaperarepublishedintheGenBanknucleotide
riod under study. Variation in the water table pattern was
databaseunderaccessionnumbersJF739556–JF739857.
similarforallsites(DHFN,DHPO,DHAN,DHTO,SHXN,
MHXN, DHPN and MHPN) except for five oil palm plan-
tation sites (MLPO) that had low water table (>50cm).
The water table in sites with high water table depths fluc-
tuatedfrom−0.1to−0.7mintheperiodunderstudy.How-
ever,thesefluctuationsrangedfrom−0.75to−1.6minthe
sites with low water table depths in the same period, hence
exposing the low-water-table sites to different durations of
oxicandanoxicregimescomparedtohigh-water-tablesites.
Biogeosciences,11,1727–1741,2014 www.biogeosciences.net/11/1727/2014/
S.Mishraetal.:Hydrologyandland-usechangeaffectsmicrobialecologyofpeatlands 1733
Fig. 2
a) n=20 b) Above water table (AWT)- dominated group (oxic zones)
n=20 Above water table (oxic zones)
*** Below water table (BWT) group Low water table
200 *** Below water table (anoxic zones) (anoxic zones) subgroup
160
Hg) n=20 n=20
m n=20 *
m ***
( 120 **
O2
p
80
Similarity
20
40 2D Stress: 0.19 40
High oxygen availability (oxic zones) Low oxygen availability (anoxic zones)
AWT- Degraded forest BWT- Degraded forest
AWT- Degraded land BWT- Degraded land
0 AWT- Oil palm plantation BWT- Oil palm plantation
Degraded Degraded Oil palm Settlements Mixed crop
AWT- Settlement BWT- Settlement
forest land plantations plantations
AWT- Mixed crop plantation BWT- Mixed crop plantation
Fig.2.Oxygenlevels(pO2),mmHg(a)andnonmetricmultidimensionalscaling(nMDS)ordinationplotbasedonBray–Curtissimilarities
calculated from presence/absence data of 16S rDNA TRFs abundances (b) at above (oxic zones) and below (anoxic zones) water table
positionsindifferentland-usetypes.Levelofsignificancein(a)is∗ p<0.05,∗∗ p<0.01,and∗∗∗ p<0.001,andn=totalnumberof
independentmeasurements.
Table 1. Shannon diversity indices for different land-use patterns table differences for both oxic and anoxic zones were sig-
basedon16SrDNAandmetabolicdiversity. nificant across land use, peat thickness and age of drainage
(Table2:two-wayANOSIMvalues).
Above Below
water water 3.1.2 Land-usepattern,ageofdrainageandpeat
Land-use table table Metabolic thickness
patterns (oxic (anoxic diversity
zones) zones) Unlike the oxic zone, where water table was the predomi-
nantfactorinfluencingbacterialcommunitystructure,inthe
Degradedforest 3.14 3.53 4.07
anoxic zone, land-use pattern had equally as strong an in-
Degradedland 3.44 3.27 4.28
Settlements 3.19 3.38 4.24 fluenceaswatertableonbacterialcommunitystructure(Ta-
Oilpalmplantations 2.62 2.68 4.33 ble2).Insuchanoxiczones,thebacterialdiversitydecreased
Mixedcropplantations 3.48 3.62 4.47 in different land-use types in the following order: mixed
cropplantations>degradedforest>settlements>degraded
Averagediversityindicesfordifferentland-usepatternsbasedon16SrDNA
werewithin±0.17and±0.08intheoxicandanoxiczones,respectively. land>oil palm plantations (Table 1). In both oxic and
anoxiczones,highestdiversityofbacterialcommunitieswas
foundinmixedcropplantations,whereasleastdiversitywas
Between-sitecomparisonsofbacterialcommunitycomposi- presentinoilpalmplantations.
tionshowedstatisticallysignificantclusteringbasedonwater Ageofdrainageandpeatthicknesshadaweakeryetsta-
tablelevelinoxiczones(Fig.3b;Table2:one-wayANOSIM tistically significant effect as compared to water table and
values) land-usepatternsinshapingthebacterialcommunityprofiles
Water table depth greatly influenced bacterial diversity inbothoxicandanoxiczones(Table2:One-wayANOSIM
in both oxic and anoxic zones, with the influence being values). In the ordination plot of the bacterial communities
greater in the oxic zone. In order to determine the rela- basedonageofdrainage,twosub-groupsdominatedbylow-
tiveinfluencesofdifferentpeatcharacteristics,weanalyzed water-tablesitesandmixedcropplantations,respectively,re-
pair-wisecombinationsofpeatcharacteristicsusingtwo-way vealed that these two parameters had additional influences
ANOSIManalysis.Itrevealedmajorinfluenceofwatertable (Supplement Fig. S1). Similar to age of drainage, in the or-
and land-use pattern over other characteristics, namely, age dinationplotbasedonpeatthickness,therewasasub-group
ofdrainageandpeatthickness(Table2:two-wayANOSIM oflow-water-tablesitesinoxiczones(Fig.3b).
values).Thevariationsinbacterialcommunitiesduetowater
www.biogeosciences.net/11/1727/2014/ Biogeosciences,11,1727–1741,2014
1734 S.Mishraetal.:Hydrologyandland-usechangeaffectsmicrobialecologyofpeatlands
Table2.One-way(boldlettersindiagonalcells)andTwo-way(non-boldlettersinoff-diagonalcells)ANOSIMshowingvariationinbacterial
communitystructureduetoanindividualparameteracrossotherparameterstestedinoxicandanoxiczones.Thetopvalueineachwhite
cellreferstoinfluencebytheindividualparameteracrossothers(lefttoright),whereasthebottomvalueinthesamewhitecellshowsthe
converse(righttoleft).Valuesinwhitecellscanbedirectlycomparedwithinsamewhitecellsorbetweenwhitecellsofoxicandanoxic
zones.Asanexampleofwithin-whitecellcomparison:inoxiczones,theinfluenceofwatertablewasmuchhigheracrossdifferentland-use
patterns(globalR:0.702),peatdepth(0.946)andageofdrainage(0.609)thanviceversa(0.189,0.344,NS:−0.009),respectively.However,
inanoxiczones,land-usepatternhadhighinfluenceacrosswatertable(0.468),peatdepth(0.636)andageofdrainage(0.532),respectively.
GlobalRstatistics
Above-water-table(oxic)zones Below-water-table(anoxic)zones
Water Land Peat Ageof Water Land Peat Ageof
table use depth drainage table use depth drainage
0.702b 0.946b 0.609c 0.541a 0.878a 0.66b
Watertable 0.527c Watertable 0.359b
0.189a 0.344a −0.009 0.468b 0.266a 0.247a
0.485b 0.361b 0.636c 0.532c
Landuse 0.267a Landuse 0.413b
0.358 0.76b 0.268 0.101
0.433c 0.563b
Peatdepth 0.214b Peatdepth 0.165a
0.574b 0.643b
Ageofdrainage 0.389c Ageofdrainage 0.192a
Levelofsignificanceisap<0.05,bp<0.01,cp≤0.001.
Fig.3
a) b) Above water table (oxic) zones
0 450
Sampling time
-0.2 400
-0.4 350
m)
-0.6 300m
Water table (m)-0-.18 220500onthly rainfall (
m
-1.2 150otal 2D Stress: 0.15
T
-1.4 100 Medium peat (3-7m) - Low water table
Deep peat (>7m) - High water table
-1.6 50 Medium peat (3-7m) - High water table
Low water table High water table Total monthly rainfall (mm) Shallow peat (<3m) - High water table
-1.8 0
Fig.3.(a)RainfallandwatertabledatafromAugust2009toAugust2010atallsamplinglocationsfromSiteAandSiteB.Thelocations
areshowninFig.1.Watertablelevelswereaveragedacrosslocationswithinhighwatertable(0–45cm)andlowwatertable(>45cm),
respectively.Theaveragedvaluesarerepresentedas“highwatertable”and“lowwatertable”,respectively.(b)Nonmetricmultidimensional
scaling(nMDS)ordinationplot,basedonBray–Curtissimilaritiescalculatedfrompresence/absencedataof16SrDNATRFsabundances,
showingvariationbetweenbacterialcommunityacrossdifferentwatertabledepthandpeatthickness,fromabove(oxiczones)watertable
positions.Low-water-table(LWT),high-water-table(HWT)anddifferentpeatthicknesssamplesarerepresentedbyopensymbols,closed
symbolsanddifferentshapes,respectively.
3.2 Distributionofbacterialtaxa datethetaxagroupidentified,aclonelibrarywascreatedthat
revealeddifferencesinabundanceoftaxabasedonwaterta-
ble,land-usepatternandoxygenavailability(Table3),which
In order to identify the dominant members of the bacterial
was consistent with our previous findings (Figs. 2 and 3b).
communities,wepredictedthetaxausingtheTRFs.Tovali-
Biogeosciences,11,1727–1741,2014 www.biogeosciences.net/11/1727/2014/
S.Mishraetal.:Hydrologyandland-usechangeaffectsmicrobialecologyofpeatlands 1735
Table3.Correspondenceofclonedataincomparisontopredictedspecies.“AWT”:above-water-tablesamples;“BWT”:below-water-table
samples.
MLPO-AWT(fromoxiczones) DHTO-AWT(fromoxiczones) DHTO-BWT(fromanoxiczones)
(Mediumpeat-lowwatertable-oil (Deeppeat-highwatertable- (Deeppeat-highwatertable-
palmplantations-drainage>10years) settlements-drainage>10years) settlements-drainage>10years)
Species No.of No.of % No.of No.of % No.of No.of %
distribution clones predicted coverage clones predicted coverage clones predicted coverage
(taxonomiclevel) (102) taxaby (104) taxaby (96) taxaby
FRAGSORT FRAGSORT FRAGSORT
Environmental 71 313 22.7 60 278 21.6 74 281 26.3
unclassified
Gammaproteobacteria 19 34 56.7 25 28 89.3 15 23 65.2
Alphaproteobacteria 5 45 11.2 10 41 24.4 3 41 7.3
Acidobacteria 5 5 100.0 73 3 100.0 2 2 100.0
Actinobacteria 0 53 0.0 1 49 2.0 0 22 0.0
Betaproteobacteria 2 44 4.5 2 37 5.4 2 41 4.9
Planctomycetes 0 12 0.0 1 9 11.1 0 14 0.0
Proteobacteria 0 4 0.0 2 3 66.7 0 1 0.0
Acidobacteria had 100% similarity with the predicted taxa mainly associated with DOC, ammonium and phosphates.
asmentionedaboveforallthesitessampledforclonelibrary. Salinity had some influence on bacterial communities from
The coverage between the predicted taxa and clones identi- sites with mixed crop plantations and settlements; the latter
fied for Gammaproteobacteria ranged from 56.7 to 89.3% corroborated the identified salinity-associated species from
between the sites sampled. The highest coverage between settlements.
thepredictedtaxaandclonesidentifiedwasfoundinthesite Metabolic profiling of peat water from different land-use
withsettlementsinoxiczones.Basedonsequencedatabase patterns was performed (Fig. 5) and was compared with
searcheswiththeclonesequences,twospecieswereidenti- bacterial profiling (Fig. 2b) to directly evaluate the effects
fiedbasedonidentitiestoknownentries.Outofwhich,Bre- of bioavailable organics that influence the bacterial com-
vundimonassp.,reportedinitiallyfromsalinesoils(Wanget munities. Bacterial communities from anoxic zones were
al.,2012),wasfoundabundantinsettlements. marginally separated based on land-use patterns (Fig. 2b).
Oxygen availability had a strong influence on the abun- The communities were separated based on habitat, as re-
dance of bacterial taxa in high and low water table depths vealed by significant separation of the flooded site (indi-
(Supplement Fig. S2). In the oxic zones, the water table cated by arrow) from the non-flooded sites of mixed crop
didnotaffecttaxaabundance,whereas,intheanoxiczones, plantations (Fig. 2b). When comparing the functional data
all major taxa were more abundant in the low-water-table frommetabolicprofiling(Fig.5),distinctclustersofdifferent
sites. Among the five most abundant taxa (α-, β- and γ- land-usetypeswereformed.Metabolicprofilingnotonlydif-
Proteobacteria, Actinobacteria and Firmicutes Bacillales), ferentiatestheland-usepatternsbutalsoclearlydistinguishes
Actinobacteria had the largest difference in abundance be- samplesbasedongeographicalsamplingposition.Forexam-
tweenhighandlowwatertablesitesinbothoxicandanoxic ple,twositesfromdegradedlandinSiteBformedadistinct
zones. clusterfromtheothertwositesfromdegradedlandinSiteA
(DHAN in Fig. 1). Similarly, two oil palm plantation sam-
3.3 Relationshipofenvironmentalandgeochemical ples (extreme left of Fig. 5), though belonging to different
parameterswithbacterialcommunitystructure peatthickness(MHPNandDHPNinFig.1),wereclustered
verycloseastheywerefromthesametransect.
Canonical correspondence analysis was used to identify
the association of environmental and geochemical traits
with bacterial communities from different land-use patterns 4 Discussion
(Fig. 4a, b and Supplement Table S1). Bacterial communi-
ties from the oil palm plantations were associated with ni- Bacterial and metabolic markers that represent the complex
tratelevelsinbothoxic(Fig.4a)andanoxic(Fig.4b)zones. nature of bacterial communities and metabolic processes of
Nitrate levels were lower in the anoxic zones compared to diverse biota, respectively, provided the resolving power to
oxiczonesbyafactorof20–30,whichislikelytodrivedif- distinguish different habitats. This resolution ranged from
ferences in their community structures. Bacterial communi- centimeter scale in depth measurements to kilometer scale,
ties in the mixed crop plantations, on the other hand, were where sites were distributed within the 48km2 of the study
www.biogeosciences.net/11/1727/2014/ Biogeosciences,11,1727–1741,2014
1736 S.Mishraetal.:Hydrologyandland-usechangeaffectsmicrobialecologyofpeatlands
Fig. 4
0 0
.1A) .1 B)
Nitrate
Dissolved organic carbon (DOC)
Phosphate
Magnesium
Calcium
Dissolved inorganic carbon Potassium Sulfate
Chloride Ammonium
Chloride
Sodium Magnesium Sodium
Ammonium
Sulfate Calcium
Dissolved inorganic carbon
Dissolved organic carbon (DOC) Nitrate
Oxic zones Phosphate Anoxic zones
Settlements Settlements
Degraded land Degraded land
Degraded forest Degraded forest
0 Oil palm plantations 0 Oil palm plantations
.1 Mixed crop plantations .1 Mixed crop plantations
--1.0 1.0 --1.0 1.0
Fig.4.Canonicalcorrespondenceanalysisof16SrDNAgenebasedT-RFLPdatasetsandenvironmentaldataindifferentland-usepatterns
from oxic (a) and anoxic (b) zones. Geochemical data, represented with arrows, are nitrates, dissolved organic carbon (DOC), dissolved
inorganiccarbon,chloride,magnesium,ammonium,sodium,calcium,sulfateandphosphate.Testofsignificance(pvalue)ofallcanonical
axesis0.05and0.009inoxic(a)andanoxic(b)zones,respectively.
Fig.5
forunderstandingvariationatlargescalesofspatialdistribu-
2D Stress: 0.05
tion. Both sets of molecular markers distinguished different
land-usetypes,butwithdifferentlevelsofresolution.Com-
paredwithmicrobialprofiles,thoseofmetabolitesweread-
ditionallyabletodifferentiateland-usetypesfromlocations
thatareseparatedbynearly8kmdistance.Ourfindingsabout
bacterialprofilinghaveledustoidentifygeochemicalfactors
that influence the state of degraded peatlands. In addition,
Oil palm plantation sites
with low water table depth metabolic profiling, which relies on markers derived from
functions of diverse biotic communities and not just bacte-
ria,provideafinerclassificationofpeatlandsites.Metabolic
Settlement Degraded land Degraded forest Oil palm plantation Mixed crop plantation profiling can, therefore, be used in developing better prac-
tices for mapping peatlands, which can be a tool for both
Fig. 5. Nonmetric multidimensional scaling (nMDS) ordination
managementandpolicydevelopment.
plot, based on Euclidean distance calculated from intensity of
While there have been reports of effects of land-use
metabolitesextractedfrompeatwaterofdifferentland-usepatterns.
Thearrowrepresentsthefloodedsitewithmixedcropplantations. change and hydrology on CO2, N2O and CH4 emissions
(Jauhiainen et al., 2008, 2012a, b) and hydrology on sub-
sidence (Hooijer et. al., 2010, 2012), our approach allows
area.Thus,thesamesetofmolecularmarkersprovidedady- multiple parameters to be evaluated simultaneously using a
namic range of resolution at four orders of magnitude. Mi- singlemolecularprofilingapproach.Ourfindingsshowthat
crobial markers have been extensively used to study alter- microbialprofilesfrompeatlandsitesaremostinfluencedby
ation in community structures due to changes in land-use variationsinwatertableandland-usepatterns.Thesetwoare
patternsatlargescalesofspatialdistribution,suchasinPa- followedbyageofdrainageandpeatthicknessininfluencing
cificNorthwestmarinesedimentcommunities(Brakeretal., thebacterialcommunitystructure.Acrossdegradedpeatland
2000), in high levels of nuclear-waste-contaminated vadose thatareunderhydrologicalmanagement,watertablefluctu-
sediments at the Hanford Site in the US (Fredrickson et al., ates due to drainage, rainfall and other physical parameters
2004), in western Amazon soils (Jesus et al., 2009) and in (Jauhiainen et al., 2008; Hooijer et al., 2010). The ability
Antarctic dry valley (Chan et al., 2013), among other bio- ofmicrobialmarkerstodistinguishthelow-andhigh-water-
geographiclocations.Incomparison,therearerelativelyfew table sites shows their robustness to capture differences in
studiesthathaveusedmetabolitesasfunction-basedmarkers communitystructuresdespitethedifferencesintherangeof
Biogeosciences,11,1727–1741,2014 www.biogeosciences.net/11/1727/2014/