NITROGEN CHANGES AND DOMAIN BACTERIA RIBOTYPE DIVERSITY IN SOILS OVERLYING THE CENTRALIA, PENNSYLVANIA UNDERGROUND COAL MINE FIRE

 Dem Zweiten Weltkrieg (1939-1945)

 5 views
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Description
NITROGEN CHANGES AND DOMAIN BACTERIA RIBOTYPE DIVERSITY IN SOILS OVERLYING THE CENTRALIA, PENNSYLVANIA UNDERGROUND COAL MINE FIRE
Share
Tags
Transcript
  1910038-075X/05/17003-191–201March 2005Soil ScienceVol.170,No.3Copyright © 2005 by Lippincott Williams & Wilkins,Inc.Printed in U.S.A. T HE Centralia,Pennsylvania mine fire startedin 1962,when a trash fire ignited a surface-accessible coal seam (Trifonoff,2000).Since thattime,the mine fire has had a dramatic effect onthe near surface environment.Surface tempera-tures have been measured at over 400  C,andland collapses and the venting ofhot steam arecommon.The initial efforts offederal,state,andlocal governments to contain and stop the firefailed,and the town’s population has dwindledfrom over 1100 to approximately 20 at presentday (Jones,2000,personal communication).It isestimated that the Centralia mine fire will con-tinue to burn for well over 100 years and that itcould ultimately spread to include 1500 acres.While this fire continues to burn,it repre-sents a fairly unique opportunity to study the re-sponses ofbacterial communities to the environ-mental changes caused by coal mine fires. NITROGEN CHANGES AND DOMAIN BACTERIA RIBOTYPEDIVERSITY IN SOILS OVERLYING THE CENTRALIA,PENNSYLVANIA UNDERGROUND COAL MINE FIRE Tammy Tobin-Janzen 1 , Ashley Shade 1 , Leslie Marshall 1 , Kristina Torres 1 , Curtina Beblo 1 ,Christopher Janzen 2 , Jennie Lenig 2 , Amy Martinez 3 , and Daniel Ressler 3 1 Biology Department, Susquehanna University, Selinsgrove, PA. Dr. Tobin-Janzenis corresponding author. Current address: 514 University Avenue, SusquehannaUniversity, Selinsgrove, PA 17870. E-mail: tobinjan@susqu.edu 2 Chemistry Department, Susquehanna University, Selinsgrove, PA. 3 Department of Geological and Environmental Sciences, Susquehanna University,Selinsgrove, PA.Received June 23, 2004; accepted November 12, 2004.DOI: 10.1097/01.ss.0000160384.45625.b8 In this study, changes to soil chemistry and domain bacteria diversitywere analyzed in a near-surface environment recently impacted by theCentralia, Pennsylvania anthracite coal mine fire. As this undergroundfire expands into new areas, land collapses are common as hot gases ventto the surface, causing rapid changes in surface soil temperatures andchemistry. To determine how these environmental changes are affectingthe resident microbial populations, surface soil samples (at a depth of 0–20 cm) were taken from boreholes at eight locations that spanned bothnewly affected and unaffected areas. Soil temperature, pH, and chemicalcomposition were analyzed at each borehole. Terminal restriction frag-ment length polymorphism analysis of domain bacteria 16S rRNA geneswas utilized to monitor the associated changes in soil microbial diversity.Over a 1-year period, the maximum surface soil temperatures in this siteincreased from 47.0  C to 75.7  C. Whereas ribotype diversity did de-crease significantly within individual boreholes as temperatures in-creased, no significant correlation was observed between overall temper-ature increases and either soil chemistry or domain bacteria ribotypediversity. Unweighted pair group means analysis of Jaccard ribotype sim-ilarity coefficients indicated that soil bacterial community diversity clus-tered well with trends in temperature, ammonium, and nitrate levels.Since significantly elevated ammonium and nitrate levels were seen inseveral of the affected boreholes, polymerase chain reaction with primersspecific for ammonia-oxidizing bacteria and Tap–terminal restrictionfragment length polymorphism analysis of the ribotype profiles were per-formed. These analyses provided evidence that nitrifying bacteria werepresent throughout the site. (Soil Science 2005;170:191–201)Key words: Bacterial diversity, soil, fire, ammonia, nitrate. SS_03-05_160646_Tobin-Janzen 3/9/05 4:27 PM Page 191  Although other large mine fires exist worldwide(Peng et al.,1997;Prakash et al.,2001),they aregenerally located in active mines,and much of the research is understandably focused on con-tainment.Since the current Pennsylvania Depart-ment ofEnvironmental Protection policy is to letthe Centralia mine fire expand and burn until itruns out offuel and extinguishes naturally,it pro-vides an ideal location in which to study the ef-fects that coal mine fires have on the near-surfaceenvironment in general.The Centralia mine firearea may also serve as a powerful model environ-ment in which to study the response ofnaturalbacterial communities to catastrophic thermalevents in situ .Such responses have been studiedafter forest fires (Mabuhay et al.,2003;Vasquez etal.,1993) and geothermal events (Norris et al.,2002).However,unlike the rapid environmentalchanges in those studies,the progression oftheCentralia mine fire is relatively slow and easy topredict.Therefore,experiments can be designedin which environmental conditions and soil bac-terial communities in the same relative locationcan be compared before and after impact by themine fire.To determine the feasibility ofsuch futurestudies,a 27  100 m site was chosen in an area lo-cated at the northwest edge ofthe Centralia minefire area.This site spanned a newly emerging ventbut also contained unaffected areas with normalsurface temperatures and healthy vascular plantcommunities.In the spring of2001 and again in thespring of2002,soil samples were collected fromsites adjacent to the vent and analyzed for temper-ature and chemical and vascular plant compositionto determine ifthe fire was indeed expanding aspredicted and to get baseline information about itseffect on the near-surface environment.Terminalrestriction fragment length polymorphism (T-RFLP) analysis ofsmall subunit RNA genes (Liu etal.,1997;Marsh et al.,2000) was used to estimatedomain bacteria diversity in these samples.Addi-tionally,polymerase chain reaction (PCR) withprimers specific for the16S rRNA genes ofam-monia-oxidizing bacteria (Hermansson,2001) wasused to determine ifthese bacteria were present inthe fire-impacted soils.MATERIALS AND METHODS Field Site  The mine fire is located in Centralia,Penn-sylvania,and is underlain by two near-surface coalseams,the Seven-Foot and the Skidmore beds.These beds lie in the Llewellyn Formation,which,along with the Pottsville formation,makeup the exposed materials on the Locust Moun-tain Anticline,(Arndt,1971) the ridge just southofCentralia and the site ofthe mine fire.Theseformations are predominantly conglomeratesandstones and sandstones.The soils are mappedas the Hazelton-Clymer Association,sandy loamsoils on ridgetops and sideslopes that formed inplace from brown to gray sandstone (Eckenrode,1985),though mining in this area has resulted inextremely heterogeneous materials resultingfrom excavation and fill.Soils in this area arenoted to be strongly acid to extremely acid,withpH  5.5 (Eckenrode,1985).The 27  100 m study site is located in a for-mer single-family dwelling neighborhood,andthe entire area srcinally consisted ofturf-basedlandscaping with native and exotic trees andshrubs.Between 5 and 10 years ago,the houses atthe study site were condemned by the state gov-ernment and demolished.Excavations were filledheterogeneously and leveled to grade.Two activevents emerged in the site during the study period. Soil Sample Collection On May 29,2001,eight surface soil samples(0–20 cm) were aseptically collected from eightboreholes by advancing a soil core to obtain rel-atively undisturbed samples.Boreholes 1–6 werecollected from a 12-m rectangular grid refer-enced to stationary site markers.Boreholes 7 and8 were located 9 and 18 meters west ofthe first 6boreholes,respectively.Samples were chilled onice in sterilized polyethylene bags for transport tothe laboratory,before being transferred to  80  C for long-term storage.All samples were trans-ferred to  80  C within 6 h ofcollection.Soiltemperatures were measured by thermocouplethermometer at the bottom ofthe borehole by aninsertion probe.This process was repeated on June 3,2002. Soil Chemistry Analysis In the laboratory,soil samples were mixedand a subsample was removed for water contentdetermination by oven drying at 105  C.Othersubsamples were used for nutrient determina-tions by mixing moist soil with 0.01 M  KCl (1:4),which was shaken at 250 orbits/min for 30 min-utes and separated with Whatman No.5 filter pa-per.Nutrient determinations were completedwith an Astoria Pacific International (Clackamas,Oregon) Astoria Analyzer continuous flow spec-trophotometer,using procedures based onUSEPA Methods 350.1 and 363.2 (USEPA,192T OBIN -J ANZEN , S HADE , M ARSHALL , ETAL .S OIL S CIENCE SS_03-05_160646_Tobin-Janzen 3/9/05 4:27 PM Page 192  1984),for ammonia and nitrate,respectively.SoilpH was determined in 1:1 soil/distilled waterslurry.Total sulfur was analyzed using a methodadapted from Tabatabai et al.(1998),as follows.Twenty grams ofsoil was dried at 110  Covernight and then sieved through a 2-mmscreen,ground,and sieved again through a 150-  m screen.A 0.5-g portion ofthe finely siftedsample was combined with 0.5 g ofNaHCO 3 and 0.02 g ofAg 2 O.This mixture was then cov-ered with 0.5 g ofNaHCO 3 and heated in a fur-nace at 550  C for 3 h.The remaining residuewas quantitatively added to 15 mL of1.0 M  aceticacid and heated for 2 minutes at 200  C.The so-lution was cooled and diluted to 100 mL withdeionized water.Total sulfur was quantified fromthis mixture using a Dionex model DX500 ionchromatograph. DNA Extraction From Soil Samples DNA was extracted from three subsamples (toserve as technical replicates) ofthe soil collectedfrom each borehole using the MoBio® Soil Ex-traction Kit as per manufacturer’s protocols.Thecrude samples were further purified by runningthem through a 1% low-melting agarose gel (0.5 glow melting agarose,5  L ethidium bromide [5mg per mL] in 50 mL 1  TAE [40 mmol/L Tris-Acetate,2 mmol/L Na 2 EDTA  2H 2 O]).The high-molecular-weight DNA was excised from the gelby using a sterile razor blade and purified using thePromega Wizard® PCR Preps DNA PurificationSystem via manufacturer’s instructions. PCR Amplification of 16S rRNA Genes Each gel purified DNA sample was amplifiedin triplicate for T-RFLP analysis,using the uni-versal bacterial primers 1507–1492r (TAC CTTGTT AGG ACT T) (Wilson et al.,1990) and Cy5.5–labeled (IDT) 8–27f(AGA GTT TGA TCCTGG CTC) (Edwards et al.,1989).Each PCRreaction contained 200 ng gel-purified DNA,2.5  L ofeach primer suspended at 20 pmol/  L,and25  L Redi-mix TM Taq PCR Reaction Mixwith MgCl 2 (Sigma).A negative control contain-ing only Redi-mix,both primers,and steriledeionized water was run with every set ofPCRamplifications,and all reactions were discarded if any fragments were observed in the negativecontrol.The samples were amplified in an Ep-pendorfMastercycler Personal thermal cycler,us-ing 2 minutes ofdenaturation at 95  C followedby 35 cycles of94  C for 60 seconds,45  C for60 seconds,and 72  C for 120 seconds for theuniversal bacteria primers and 45 cycles of94  Cfor 60 seconds,57.5  C for 60 seconds,and 72  Cfor 60 seconds for the ammonia-oxidizing bacte-ria (AOB) primers.A final extension was per-formed at 72  C for 5 minutes for both primersets.Full-length 16S rDNA fragments were puri-fied from a low melting agarose gel as previouslydescribed. T-RFLP Analysis A 500-ng portion ofeach amplified 16SrDNA sample was digested to completion with10 units of  Rsa I restriction enzyme (New En-gland Biolabs) for 2.5 h at 37  C.These condi-tions were shown to digest an amplified Esche-richia coli  16S rRNA gene fragment tocompletion,generating the expected fragmentsizes.The digested fragments were then purifiedusing Amersham Biosciences AutoSeq TM G-50columns,according to the manufacturer’s proto-col.The resulting digestion mixtures were driedand centrifuged at 60  C,using a LABCONCOCentrivap Concentrator to remove all liquid.TheDNA was resuspended in 6.0  L offormamideloading dye (Amersham Biosciences ThermoSe-quenase Cy5.5 Terminator Cycle SequencingKit).The loading dye included 100,300,and 500bp internal Cy5.5-labeled size standards (VisibleGenetics).Each resuspended sample was split intothree 2.0-  L technical replicates and run at 1500V,30 mA,and 50  C on an Amersham Bio-sciences SEQ 4  4 Personal Sequencing Sys-tem.One lane containing external 50,100,200,300,450,and 500-bp standards (Visible Genetics)was run with each set oftechnical replicates tohelp control for gel variability.Standard curveswere calculated by using the migration ofboththe internal and external size standards and thenwere used to calculate individual terminal restric-tion fragment (TRF) sizes.The replicate valuesfor each TRF were averaged.TRF presence orabsence in each sample was converted to binarydata,and Jaccard similarity coefficients were cal-culated.Dendrograms ofthe Jaccard data weregenerated using the MultiVariate Statistical Pack-age v3.1 (Kovach Computing Services),usingboth an unweighted pair group mean analysis(UPGMA) and a minimum variance (Ward’s)method. Determining the Presence of Nitrifying Bacteria During the fall of2002,an additional threesoil samples were aseptically collected,as previ-ously described,from three new boreholes.Thefirst borehole was located in the southeast cornerofthe sample site,the second in the center oftheV OL . 170 ~ N O . 3B ACTERIAL D IVERSITY A BOVE A C OAL M INE F IRE 193 SS_03-05_160646_Tobin-Janzen 3/9/05 4:27 PM Page 193  site,and the third at the western edge ofthe sam-ple site.Soil temperature and ammonium and ni-trate levels were determined for each soil sample.Two hundred nanograms ofpurified genomicDNA from each ofthese boreholes was amplifiedwith the CTO189fA/RT1r primer pair (as de-scribed in Hermansson,2001).These primersspecifically amplify a 116-bp region ofthe 16SrRNA genes ofAOB.A positive control contain-ing Nitrosomonas europaea genomic DNA (ATCC)was run to verify the specificity ofthe primers,and a negative control containing only Redi-mix,both primers,and sterile deionized water was runto check for environmental contamination.RESULTS Study Site and Surface Soil Chemistry In 2001,the temperatures taken in boreholes1–6,which were adjacent to the emerging vents,ranged from 25  C to 47  C (Table 1),whereasthose from unaffected sites 7 and 8 were 17.0  Cand 13.2  C,respectively.The soils in both the affected and the unaffected boreholes weremedium to dark brown,moist,sandy loam inter-spersed with coal and ash.An initial characteriza-tion ofvascular plants in the sample site showedthat unaffected areas (below 30  C) were pre-dominated by Kentucky bluegrass ( Poa pratensis ),nimblewill ( Muhlenbergia schreberi  ),goldenrod( Solidago spp. ),black raspberry ( Rubus occidentalis ),and both native and exotic trees and shrubs thatwere planted when the area was a single-familydwelling neighborhood.Fall Panicum ( Panicumdichotomiflorum Michx ) was able to survive inmoderately affected areas (32  C to 38  C),whereas the only vascular plant capable ofsurviv-ing between 38 and 60  C was Purslane ( Portulacaoleracea ).Sites above 60  C were barren ofvascu-lar plant life.During 2002,the mine fire front moved fur-ther into the study site,as predicted,and thetemperature in boreholes 1–6 increased by anaverage of16.1  7.5  C (from an average of 36.4  C to an average of52.4  C),with site 2showing a striking increase of28.7  C.By con-trast,the average temperature increase in unaf-fected boreholes 7 and 8 was only 4.0  2.3  C(Table 1).Surface soil temperatures across the samplesite were not strongly correlated with changes inammonium ( r  2  0.31) or nitrate ( r  2  0.31).However,ammonium and nitrate levels in someaffected boreholes did fluctuate to significantlyhigher levels than those seen in unaffected bore-holes.For example,in 2001,boreholes 4 and 5had ammonium levels of12.63 and 13.79 mg/kg,respectively,whereas the nitrate level in borehole5 was 103.1 mg/kg (Table 1).The second sampling year occurred at the be-ginning ofa regional drought,and an average de-crease in soil moisture of0.10 g/g (0.24–0.14g/g) over the entire sampling area was seen.In-terestingly,borehole 2,which was the hottestborehole in the study,had higher soil moisture in2002 than in 2001,perhaps due to condensationofvent gases (Table 1).No correlation was seenbetween temperature and total sulfur ( r  2  0.37)in the surface soil samples.Soil pH was slightlyacidic to acidic throughout the study site,rangingfrom 6.35 to 4.11.Although the average pHacross the sample site increased from 2001 (4.56  0.56) to 2002 (5.55  0.55),this increase wasnot correlated to increases in soil temperature ( r  2  0.25). Bacterial Community Diversity Adjacent to the Vents During 2001,T-RFLP analysis ofdomainbacteria 16S rRNA genes was performed to de-termine ifthis technique would be an effectiveway to assay bacterial community diversity in thestudy site and to gather baseline data regardingmicrobial community structure around the newvent.Internal size standards were used to align thepeak profiles ofdifferent samples generated bythe sequencer and to standardize data betweengels.Three technical replicates were performedfor each borehole sample,and only peaks presentin all three replicates were used in subsequent sta-tistical analyses.Since the same general trends incommunity diversity were seen across the entireribotype profile,and since much greater resolu-tion was present in the smaller fragments,onlypartial ribotypes,in the 1 to 300-bp size range,were used for subsequent analyses.T-RFLP analysis was initially performed onboreholes 2,4,and 5 in 2001.There was no sig-nificant difference in the diversity ofdomain bac-teria communities in these sites,as determined bythe number ofTRFs (Table 1).Additionally,boreholes 4 and 5 showed significant peak simi-larity,with a Jaccard similarity index of0.833(Table 2).By contrast,pairwise comparison of boreholes 4 and 5 with borehole 2 revealed Jac-card similarities ofonly 0.552 and 0.607,respec-tively.Boreholes 4 and 5 also shared slightly ele-vated temperatures and high ammoniumconcentrations,suggesting that similar popula-tions ofnitrifying bacteria might be capitalizing194T OBIN -J ANZEN , S HADE , M ARSHALL , ETAL .S OIL S CIENCE SS_03-05_160646_Tobin-Janzen 3/9/05 4:27 PM Page 194  on the high availability ofammonium.The Tap-TRFLP program at the Ribosomal DatabaseProject II site (Cole et al.,2003) was thus used todetermine ifpotential nitrifying bacteria couldbe identified in the TRF profiles.Unfortunately,most AOB were predicted to generate TRFslarger than 300 bp and thus were not resolvablein this experiment.However,samples 4 and 5both contained a 167-bp fragment,designatedwith an asterisk in Fig.1,which was consistentwith the presence ofmembers ofthe nitrite-ox-idizing genus Nitrospira .To follow up on these interesting observa-tions and to expand the study into a wider areathat included both emerging vents,T-RFLPanalysis was performed on boreholes 1 through 7in the 2002 season.As shown in Table 1,the av-erage number ofTRFs in boreholes 2,4,and 5decreased significantly (from an average of24.3  1.2 TRFs in 2001 to an average of17.7  1.5TRFs in 2002) as the temperatures in these bore-holes increased from generally mesophile-sup-porting temperatures of47  C or less to temper-atures more supportive ofthermophiles or ther-motolerant species (52.2–75.7  C).Additionally,although there was only a weak negative correla-tion between temperature and the number of TRFs across the sample site ( r  2  0.67),bore-holes with temperatures greater than 50  C hadsignificantly fewer TRFs (17.25  2.99) thanthose with temperatures below 50  C (26  5.43)(Table 1). Jaccard similarity analysis was expanded to in-clude boreholes 1–7 in 2002 (Table 2),and den-drograms were generated using both the UP-GMA and minimum variance (Ward’s) methods.Both dendrograms generated similar results,andonly the UPGMA results are shown in Fig.3.Thetwo most similar ribotype profiles,as determinedby Jaccard and UPGMA analysis,were seen inboreholes 4 and 5 during 2001.Boreholes 2 and4 also showed substantial similarity in 2002,V OL . 170 ~ N O . 3B ACTERIAL D IVERSITY A BOVE A C OAL M INE F IRE 195 TABLE 1Chemical and T-RFLP analysis ofsurface soil samples from boreholes 1 through 8 in 2001 and 2002.TempNH 4  NO 3  MoistureTotal sulfurBorehole(°C)(mg/kg)(mg/kg)(g/g)pHNo.TRFs(mg/kg)2001200220012002200120022001200220012002200120022002133.553.39.297.730.907.940.240.144.114.54ND16231247.075.73.635.602.308.910.280.304.245.602519166.5328.039.04.654.153.913.450.210.054.125.83ND23141.1437.652.212.631.022.1454.160.180.094.875.892316171.4547.062.313.793.23103.19.840.220.164.585.672518192.1625.032.02.700.772.010.450.230.085.806.35ND26157.7717.022.60.430.551.880.950.220.114.415.46ND25232.8813.215.63.351.823.920.590.290.134.325.06NDND61.03Boreholes 1 through 6 were located in a 12-m rectangular grid adjacent to two emerging vents;boreholes 7 and 8 were situ-ated in unaffected areas to the west ofthe vents.TABLE 2 Jaccard similarity coefficients for domain Bacteria ribotypes from boreholes 1 through 7 in 2001 and 2002Borehole,year122334556720022001200220022001200220012002200220021,20021.0002,20010.3851.0002,20020.5000.6091.0003,20020.4500.5000.5791.0004,20010.4580.5520.6360.6361.0004,20020.4210.4800.7500.5790.6361.0005,20010.4000.6070.6360.7270.8330.6361.0005,20020.4740.3100.4500.4760.6090.4500.5421.0006,20020.4090.4640.6000.7000.6670.6000.7390.5001.0007,20020.3810.3000.3040.2800.4070.2500.3570.3480.3601.000 SS_03-05_160646_Tobin-Janzen 3/9/05 4:27 PM Page 195
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks