Resilience of Arctic Mycorrhizal Fungal Communities after Wildfire Facilitated by Resprouting Shrubs Author(s): Rebecca E. Hewitt , Elizabeth Bent , Teresa N. Hollingsworth , F. Stuart Chapin III & D. Lee Taylor Source: Ecoscience, 20(3):296-310. 2013. Published By: Centre d'études nordiques, Université Laval DOI: http://dx.doi.org/10.2980/20-3-3620 URL: http://www.bioone.org/doi/full/10.2980/20-3-3620
BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.
BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research.
Resilience of Arctic mycorrhizal fungal communities after wildfire facilitated by resprouting shrubs1
5HEHFFD(+(:,772 (OL]DEHWK%(173,'HSDUWPHQWRI%LRORJ\DQG:LOGOLIH,QVWLWXWHRI$UFWLF%LRORJ\ 8QLYHUVLW\RI$ODVND)DLUEDQNV)DLUEDQNV$ODVND 86$HPDLO UHKHZLWW#DODVNDHGX 7HUHVD1+2//,1*6:257+ US Forest Service, Pacific Northwest Research Station, Boreal Ecology &RRSHUDWLYH5HVHDUFK8QLW)DLUEDQNV$ODVND 86$ )6WXDUW&+$3,1,,, '/HH7$
Abstract: Climate-induced changes in the tundra fire regime are expected to alter shrub abundance and distribution across the Arctic. However, little is known about how fire may indirectly impact shrub performance by altering mycorrhizal symbionts. :HXVHGPROHFXODUWRROVLQFOXGLQJ$5,6$DQGIXQJDO,76VHTXHQFLQJWRFKDUDFWHUL]HWKHP\FRUUKL]DOFRPPXQLWLHVRQ resprouting Betula nana shrubs across a fire-severity gradient after the largest tundra fire recorded in the Alaskan Arctic -XO\±2FWREHU )LUHHIIHFWVRQWKHFRPSRQHQWVRIIXQJDOFRPSRVLWLRQZHUHGHSHQGDQWRQWKHVFDOHRIWD[RQRPLF resolution. Variation in fungal community composition was correlated with fire severity. Fungal richness and relative DEXQGDQFHRIGRPLQDQWWD[DGHFOLQHGZLWKLQFUHDVHGILUHVHYHULW\ Introduction Fire frequency and severity play a strong role in regu- (Racine et al. /DQW]*HUJHO +HQU\ \HW lating shrub regeneration and distribution in the Arctic little is known about the response to fire of soil fungal com- munities critical to shrub growth. Shifts in the tundra fire 5HF DFF regime could influence vegetation change directly by open- $VVRFLDWH(GLWRU 7LP0RRUH ing new microsites for colonization and succession (White, 2 Author for correspondence. RULQGLUHFWO\E\DOWHULQJWKHDYDLODELOLW\RIFULWLFDO 3 Present address: School of Environmental Sciences, University of Guelph, Guelph, fungal symbionts, ectomycorrhizal fungi (EMF), that influ- 2QWDULR1*:&DQDGD 4 Present address: Department of Biology, University of New Mexico, Albuquerque, ence host plant performance (Hoeksema et al. ,Q 1HZ0H[LFR86$ general, fire effects on mycobionts are governed by burn '2, severity, which impacts the availability of host plants, and ÉCOSCIENCE, VOL. 20 (3), 2013 the depth of burning in the soil profile (Neary et al. VSRUHVDQG³ODWHVWDJH´IXQJLZKLFKFRORQL]HURRWVRI &HUWLQL ,WLVWKHUHIRUHOLNHO\WKDW(0)UHVSRQVH more mature trees and spread by mycelial growth but not to fire disturbance may be an important determinant of E\VSRUHV 'HDFRQ'RQDOGVRQ /DVW )R[ post-fire vegetation regeneration and expansion in the 1HZWRQ :KLOHWKLVFODVVLILFDWLRQV\VWHPLVOLNHO\ Alaskan Arctic. over-simplistic, the wide dispersal and rapid colonization High-latitude climate warming has altered landscape- of the early colonizers is suggestive of an r-strategy, while scale disturbance regimes in the Arctic (Hu et al. the later colonizers display some attributes consistent with a 7XQGUDILUHVRQWKH1RUWK6ORSHRI$ODVNDZHUHUDUHRFFXU- .VWUDWHJ\,QDGGLWLRQWR(0)KRVWSODQWVWKDWDUHSULPDU- UHQFHVRYHUWKHODVW\ +X et al. GXHWRFRRO ily ectomycorrhizal sometimes associate with dark septate moist conditions and low biomass of dry fuels (Wein, endophytes (DSE) and ericoid mycorrhizal fungi (ERM) ,QFUHDVHGIUHTXHQF\RIZDUPZHDWKHUWRJHWKHUZLWK 7HGHUVRR et al., 2009). Mycorrhizal composition has dif- ORZVXPPHUSUHFLSLWDWLRQ 6KXOVNL :HQGOHU ferential effects on host plant performance through nutri- however, has caused a non-linear increase in tundra area ent translocation, water uptake, carbon cost, and pathogen burned annually (Hu et al. ,QWKH$QDNWXYXN resistence, among other factors (Smith & Read, 2008). River Fire (ARF), the largest tundra fire recorded in the 7KXVLWLVOLNHO\WKDWVKLIWVLQP\FRUUKL]DOFRPSRVLWLRQ FLUFXPSRODU$UFWLFEXUQHG NP2 on the North Slope due to fire have an impact on mycorrhiza-mediated host RI$ODVND7KH$5)ZDVXQSUHFHGHQWHGLQERWKVL]HDQG plant performance and may influence shrub expansion into severity in the modern fire record and may be a harbinger of non-shrub tundra. Host plants that survive fire, such as greater landscape flammability in a future warmer climate resprouting shrubs, however, may maintain their pre-fire (Jones et al., 2009). Climate-sensitive changes in the fire mycorrhizal partners and facilitate mycorrhizal resilience regime are predicted to accelerate vegetation change attrib- to fire. uted to climate warming across the Arctic (Landhausser Here, we conducted the first investigation on the :HLQ LQFOXGLQJLQFUHDVHGVKUXEDEXQGDQFHDQG effects of fire on mycorrhizal community structure in Arctic shrub tundra expansion into graminoid tundra. tundra. We sampled roots from resprouting B. nana and Betula nana, one of the 4 dominant vascular plants used molecular tools to characterize fungal community in northern Alaskan tundra (Walker et al. H[KLELWV composition across a fire-severity gradient in the ARF. increased productivity in response to warming (Bret-Harte, 7KLVEXUQVFDULVLGHDOIRUVWXG\LQJILUHVHYHULW\HIIHFWVRQ 6KDYHU =RHUQHU DQGLVWKHRQO\GRPLQDQWVSHFLHV mycorrhizal communities because the large area burned that is obligately ectomycorrhizal (Molina et al. allowed us to examine landscape effects of fire on EMF While dwarf shrubs such as B. nana resprout from below- communities, and the wide spectrum of burn severities ground stems after fire, recovery to pre-fire productivity within the burn scar allowed us to investigate fire effects levels commonly takes over a decade (Fetcher et al. on a scale that would be logistically impossible to replicate ,IDILUHLVRIKLJKVHYHULW\LWZLOOQRWRQO\NLOOPRUHSODQW experimentally. Using this study site, we hypothesized that structures and likely lengthen the time required for post- plant-associated mycorrhizal community structure and ILUHVKUXEUHFRYHU\ :HLQ EXWDOVRFDXVHVXEVWDQ- FRPSRVLWLRQZRXOGFKDQJHZLWKEXUQVHYHULW\ )LJXUH WLDOVKLIWVLQ(0)FRPPXQLW\FRPSRVLWLRQ7KHSULPDU\ ,QSDUWLFXODUZHWHVWHGWKHIROORZLQJK\SRWKHVHV ILUH regulators of EMF composition are host plant diversity and severity has a greater effect on shrub mycorrhizal com- abundance (Molina et al. DQGHGDSKLFSURSHUWLHV munity structure than do abiotic factors such as soil pH and including soil pH and organic matter content (Kernaghan +DUSHU ERWKRIZKLFKDUHLPSDFWHGE\ILUHVHYHU- Only re-colonizing LW\ 'DKOEHUJ ,QIRUHVWHGHFRV\VWHPVVRPHILUH Surviving host-plant vegetation vegetation specialist fungi, primarily those in the Ascomcota, such as Wilcoxina sp. and Tuber sp. in the Ascomycota and Residual organic soil depth Rhizopogon sp. in the Basidiomycota, respond positively Total richness or abundance Remaining organic soil depth to fire (Baar et al. 7D\ORU %UXQV 2WKHU Richness of late-stage fungi fire-sensitive fungi, primarily those in the Basidiomycota, i.e. VSHFLHV LQ WKH 5XVVXODFHDH7KHORSKRUDFHDH DQG Amanitaceae, respond negatively to fire, are abundant in unburned stands, and are eliminated or reduced by fire dis- WXUEDQFH 7D\ORU %UXQV $IWHUILUHVHYHUDOGHFDGHV may be required for EMF composition, richness, and root WLSFRORQL]DWLRQWRUHWXUQWRSUHILUHOHYHOV 9LVVHU 7UHVHGHU0DFN &URVV of fungal diversity Measures 7KHSRWHQWLDOO\ORQJOLYHGILUHHIIHFWVRQ(0)FRPSRV- ition could result in alterations in host plant performance due to the shift from pre- to post-fire fungal communities. EMF are well known to undergo successional changes in Fire severity species composition in tandem with maturation of their FIGURE +\SRWKHVL]HGFKDQJHVLQPHDVXUHVRIIXQJDOFRPPXQLW\ KRVWV7KHVHSKHQRPHQDOHGWRWKHFDWHJRUL]DWLRQRI³HDUO\ composition across a tundra fire-severity gradient. Late-stage fungi are stage” fungi, which colonize young hosts effectively from likely K-strategists and predominantly basidiomycetes. HEWITT ET AL.: TUNDRA FIRE EFFECTS ON MYCORRHIZAL COMMUNITIES landscape position; 2) total richness of shrub mycorrhizal 7RPRQLWRUSRVWILUHYHJHWDWLRQUHFRYHU\WKH%XUHDX taxa decreases with increasing fire severity; and 3) distribu- of Land Management (BLM) established sites in the ARF tion of shrub mycorrhizal taxa varies along a fire-severity GXULQJDILHOGFDPSDLJQ,Q-XO\ JURZLQJ gradient (unburned to severely burned), with late-stage VHDVRQVDIWHUWKHILUHZHYLVLWHGEXUQHGVLWHVZLWKLQWKH taxa decreasing and fire-specialists increasing across the ARF burn scar corresponding to different fire severities and fire-severity gradient. XQEXUQHGVLWHQHDUWKHEXUQVFDU VLWHVWRWDO7DEOH, Sites were chosen to represent a continuous fire-severity Methods JUDGLHQWEDVHGRQWKHFRPSRVLWHEXUQLQGH[ &%, DILUH STUDY AREA AND FIELD SAMPLING severity metric, between 2 endpoints from unburned to %HWZHHQ -XO\ DQG 2FWREHU WKH$5) EXUQHG VHYHUHO\EXUQHG7KH$5)EXUQVFDULVDFFHVVLEOHRQO\ NP2 of upland shrubby tussock tundra underlain by by helicopter in the summer or snow machine in winter. continuous permafrost (Mack et al. 7KHGRPLQDQW 7KHUHIRUHZHFKRVHORJLVWLFDOO\DFFHVVLEOHVLWHVWRVDPSOH YHJHWDWLRQEHIRUHWKHILUHZDVPRLVWDFLGLFWXQGUD LQWHQVLYHO\ XQEXUQHGURDGDFFHVVLEOHVLWHQHDUWKH$5) ZLWKPRLVWQRQDFLGLFWXQGUD DQGVKUXEODQG EXUQVFDU VLWH DQGVLWHVZLWKKHOLFRSWHUDFFHVVPRG- covering smaller areas (Jandt et al. :HIRFXVHG HUDWHO\EXUQHGVLWH VLWH DQGKLJKEXUQVHYHULW\VLWH our sampling in the moist acidic tundra that is co-domin- VLWH 7DEOH, $WHDFKLQWHQVLYHVLWHZHKDUYHVWHG ated by sedges (Eriophorum vaginatum, Carex bigelowii), B. nanaUHVSURXWVDORQJSDUDOOHOîPWUDQVHFWV evergreen ericoid mycorrhizal shrubs (Ledum palustre PDSDUW7ZRVKUXEVZHUHVDPSOHGIURPHDFKWUDQVHFW and Vaccinium vitis-idaea), deciduous ectomycorrhizal with the first sampling location randomly determined and shrubs (Betula nana and Salix pulchra), mosses, and lichens WKHVHFRQGORFDWHGPIURPWKHILUVW7RLQFUHDVHWKHVSD- (Walker et al. tial extent of our study and potentially better represent the TABLE I. 6LWHGHVFULSWLRQVIRUVDPSOLQJVLWHVJURXSHGE\¿UHVHYHULW\FODVVHVZLWKLQRUDGMDFHQWWRWKH$QDNWXYXN5LYHU)LUHEXUQVFDU Fire Severity a Site b No. samples Location Site description c d Unburned 1 N, Unburned tundra site with Eriophorum vaginatum tussocks and Betula nana, W Ledum palustre, Rubus chamaemorus, sphagnum, and feather mosses in the inter-tussock space; 2C2A /RZ N, Moist tussock tundra site with B. nana, R. chamaemorus, Eriophorum sp. W tussocks, and some sedges; 2C2H N, Wet meadow with B. nana, L. palustre, R. chamaemorus, sphagnum mosses, W DQGQRQWXVVRFNIRUPLQJVHGJHV $, N, Moist lowland site with B. nana and Salix sp. shrubs, Carex bigelowii, and W Eriophorum sp. tussocks; 2C2A Moderate 104 N, Moist site with resprouting Eriophorum sp. tussocks and B. nana, W R. chamaemorus, and L. palustre in the inter-tussock space; 2C2A N, Moist site on a north hillslope with B. nana, R. chamaemorus, Eriophorum sp. W tussocks, and Polygonum bistortoides; 2C2A N, Moist site with B. nana, L. palustre, R. chamaemorus, and Eriophorum sp. W tussocks.; 2C2A N, Wet sedge meadow site with B. nana, R. chamaemorus, and Carex sp. in a W polygonated valley bottom; 3A3 N, Moist site with B. nana, L. palustre, Carex sp. sedges, and Eriophorum sp. W tussocks; 3A2D with patches of 3A3 polygons High 101 N, Moist tundra with exposed mineral soil and resprouting Eriophorum sp. W tussocks, R. chamaemorus, and B. nana shrubs; 2C2A N, Moist tundra with Eriophorum sp. tussocks, Saussurea angustifolia D.C., W B. nana shrubs, and combusted sphagnum mosses; 2C2A N, Moist ridgetop tussock site with exposed mineral soil, B. nana and Salix sp. W shrubs, and Eriophorum sp. tussocks ; 3A2D $ N, Moist site with exposed mineral soil, B. nana and Salix sp. shrubs, L. palustre, W Carex sp. sedges, and Eriophorum sp. tussocks; 3A2H N, High severity dry site with Equisetum sp., graminoids, Salix sp., B. nana W shrubs, and exposed mineral soil; 2C2C N, Hill top high severity moist tussock tundra site with B. nana, L. palustre, and W Eriophorum sp. tussocks; 2C2A a )LUHVHYHULW\FODVVHVDUHEDVHGRQ&RPSRVLWH%XUQ,QGH[ &%, 1RUPDOL]HG%XUQ5DWLR,QGH[ G1%5 YDOXHVDQGVLWHGHVFULSWLRQV b Bold site numbers denote intensively sampled sites, while extensively sampled sites are not bold. c :H FODVVL¿HG YHJHWDWLRQ FRPPXQLWLHV XVLQJ WKH9LHUHFN et al. YHJHWDWLRQ FODVVL¿FDWLRQ &$ RSHQ ORZ PL[HG VKUXE±VHGJH WXVVRFN WXQGUD &+ RSHQORZZLOORZ±VHGJHVKUXEWXQGUD$ VHGJH±ELUFKWXQGUD$+ VHGJH±ZLOORZWXQGUDDQG$ ZHWJUDPLQRLGKHUEDFHRXVWXQGUD d (QYLURQPHQWDOVLWHGHVFULSWLRQVPLQHUDOVRLOS+DFWLYHOD\HUGHSWKUHPDLQLQJVRLORUJDQLFOD\HUGHSWKDQG&%,GDWDIRUEXUQHGVLWHVKDYHEHHQVXP- marized in Jandt et al. DQG DUH DYDLODEOH RQOLQH DW ZZZIUDPHVJRY DQG ZLWK WKH$UFWLF /RQJ7HUP (FRORJLFDO 5HVHDUFK /7(5 GDWD DUFKLYH KWWS HFRV\VWHPVPEOHGXDUFGDWDFDWDORJKWPO 298 ÉCOSCIENCE, VOL. 20 (3), 2013 richness of taxa found across the fire-severity gradient, we was little visible morphological variation, i.e., few mor- obtained additional B. nana root systems from less inten- SKRW\SHVRQHDFKVKUXE7KHUHIRUHVHJPHQWVRIURRWV sively studied sites, i.e., extensive sites, through a collabor- were randomly selected from each B. nana root system and ative effort with the BLM during their helicopter-accessed, sampled for ectomycorrhizal root tips. Healthy ectomycor- SRVWILUHYHJHWDWLRQVXUYH\V$WWKHVHH[WHQVLYHVLWHV rhizal root tips from B. nana roots were identified using B. nana resprouts were sampled when they were present DGLVVHFWLQJPLFURVFRSH XSWRî PDJQLILFDWLRQ DQG GXULQJ%/0SRVWILUHYHJHWDWLRQVXUYH\VDORQJPWUDQ- individual root tips were removed with forceps. Criteria for sects. Shrubs occurred in low densities along these BLM selecting healthy root tips included a lack of root hairs, no WUDQVHFWVVR±VKUXEVZHUHVDPSOHGSHUVLWH 7DEOH, sign of necrosis, and turgid, intact tips, following both the Harvested shrub root balls were approximately 30 cm in workflow for root sampling and diagnostic features of EMF diameter, with the root crown in the centre and 30 cm deep. in Brundrett et al. ZLWKUHIHUHQFHWRPRUSKRORJLHV ,QWRWDOZHVDPSOHGURRWV\VWHPVZLWKLQWDFWURRWEDOOV LQ$JHUHU ± 3UHOLPLQDU\DQDO\VHVRQVKUXEV DWWKHH[WHQVLYHVLWHVDQGDWWKHLQWHQVLYHVLWHV/RZ GHPRQVWUDWHGWKDWWLSVZHUHVXIILFLHQWWRUHSUHVHQWWKH shrub density on the landscape and limited helicopter time richness of fungal amplicons present on a root system SUHFOXGHGVDPSOLQJRIODUJHUQXPEHUVRIVKUXEV7KHVKUXEV WLSVversusWLSV 7(6) ±P $SSHQGL[, were then transported to the University of Alaska Fairbanks, Hence, sampling more tips would not have drastically where they were gently washed with distilled water within changed community composition results. For that reason, ZHHNRIKDUYHVW5RRWVZHUHWUDFHGEDFNWRWKHURRW WLSVZHUHUDQGRPO\FKRVHQIURPDOOWKHHFWRP\FRU- FURZQDQGVWRUHGIRUPRQWKVDW&LQ51$ODWHU /LIH rhizal tips identified on the root system of an individual 7HFKQRORJLHV&RUSRUDWLRQ)RVWHU&LW\&DOLIRUQLD86$ B. nana shrub and pooled for DNA sequence analysis and which preserves sample morphology and nucleic acids $XWRPDWHG5LERVRPDO,QWHUJHQLF6SDFHU$QDO\VLV $5,6$ LQGHILQLWHO\XQWLOVDPSOHVFDQEHSURFHVVHG /DGHU of mycorrhizal fungi community structure. Pooled root tips %HQW 7D\ORU 1RWHWKDW51$ODWHUKDOWVDOOPHWD- were then placed in a single 0.6-mL Eppendorf tube, frozen bolic activity and protects extremely labile RNA, as well in a small amount of ultrapure water, and lyophilized. DV'1$ KWWS ZZZLQYLWURJHQFRP :HKDYHUHFRYHUHG equivalent DNA from fresh samples and samples stored MOLECULAR TECHNIQUES AND BIOINFORMATICS for over 2 y in RNAlater at room temperature. Mycorrhizal :HFRQGXFWHGDSUHOLPLQDU\DQDO\VLVXVLQJ$5,6$ PRUSKRORJ\DQGDQDWRP\ZHUHDOVRSUHVHUYHG '/7D\ORU to verify that we had adequately captured the fungal com- unpubl. data). PXQLW\DVVRFLDWHGZLWKHDFKLQGLYLGXDOVKUXE $SSHQGL[, 7R H[SORUH WKH UHODWLRQVKLS EHWZHHQ FRPSRVLWLRQ We used DNA sequencing to investigate mycorrhizal com- of mycorrhizal fungi, fire severity, and other potentially munity structure of B. nana resprouts. DNA was extracted controlling factors, each site was characterized in terms IURPO\RSKLOL]HGSRROHGURRWWLSVDPSOHVXVLQJWKH of latitude, vegetation cover (estimated in the field or 4LDJHQ'1(DV\3ODQW0LQL.LW 4,$*(1,QF9DOHQFLD by photographs), mineral soil pH (measured in the lab California, USA) according to the manufacturer’s instruc- post hoc), active layer (i.e., summer thaw in permafrost WLRQVDQGIROORZLQJWKHSURWRFRORI%HQWDQG7D\ORU soils) depth (measured in the field at 20 points and then averaged across site), and remaining soil organic layer Clone libraries were created for the pooled fungal depth (measured at 3 points and averaged across a site). DNA from each shrub (i.e.FORQHOLEUDULHV )XQJDO,76 )LUHVHYHULW\YDULDEOHVLQFOXGHG&%, PHDVXUHGLQWKHILHOG gene region sequences were obtained by PCR amplifica- in 30 m radius plots and calculated as an average rating WLRQZLWKWKH3&5SULPHUV86(5,76)DQG86(5,76 of consumption for fuel layers in tundra, including burn following Bent et al. 86(5SULPHUVKDYHDQDGGL- severity of substrate [litter, duff, and mineral soil expos- tional 8 bases, which enable them to work with the USER ure] and burn severity of vegetation from low vegetation )ULHQGO\&ORQLQJ.LW 1HZ(QJODQG%LRODEV,SVZLFK and tall shrubs) (Jandt et al. DQG1RUPDOL]HG%XUQ 0DVVDFKXVHWWV86$ 7KHUHVXOWLQJSULPHUVHTXHQFHVDUH 5DWLR,QGH[ G1%5FDOFXODWHGDVWKHFKDQJHLQUHIOHFW- as follows, with the additional bases underlined: USER- ance pre- to post-fire based on Landsat imagery) (Kolden, ,76)**$*$&$8&77**7&$777$*$**$$*7$$ Post hoc, we created fire-severity classes across our *DUGHV %UXQV 86(5,76 VHYHULW\JUDGLHQWIRUHDFKVLWHEDVHGRQ&%,G1%5YDOXHV ***$$$*87&&7&&*&77$77*$7$7*& :KLWH et al., DQGVLWHGHVFULSWLRQV 7DEOH, 'HWDLOHGPHWKRGVIRUGDWD (DFKUHDFWLRQPL[ / ZDVHOHFWURSKRUHWLF- collection and summarized fire, soil, and environmental DOO\VHSDUDWHGIRUPLQRQDQDJDURVHJHOLQ7%(EXIIHU site data are presented in Jandt et al. (QYLURQPHQWDO DJDURVH9PLQ WRUHPRYHDOOSULPHUVDQG data sets for the ARF have been archived with the Fire partially amplified PCR products from the amplicons of Research and Management Exchange System at www. LQWHUHVW2QHRIWKHVDPSOHVGLGQRWSURGXFHDPSOLFRQV IUDPHVJRYDQGZLWKWKH$UFWLF/RQJ7HUP(FRORJLFDO HLWKHULQWKH$5,6$H[SHULPHQWRULQWKLVRQHDQGZDVQRW 5HVHDUFK /7(5 GDWDDUFKLYH KWWS HFRV\VWHPVPEOHGX included in subsequent manipulations. DUFGDWDFDWDORJKWPO 7KH ± ES UHJLRQ FRQWDLQLQJ DPSOLFRQV RI each lane was excised from agarose gels with a scalpel. FUNGAL SAMPLING Amplicons were recovered from the gel slices using a ,Q0DUFKURRWV\VWHPVZHUHFXWLQWRFPVHJ- 4LDJHQ*HO([WUDFWLRQ.LW 4,$*(1,QF IROORZLQJWKH ments and placed in dishes of ultrapure water. Sampling all PDQXIDFWXUHU¶VLQVWUXFWLRQVDQGOLJDWHGLQWRS1(%$ tips from the root ball was logistically unfeasible, and there (New England Biolabs). Ligation reactions were then used 299 HEWITT ET AL.: TUNDRA FIRE EFFECTS ON MYCORRHIZAL COMMUNITIES for the transformation of competent E. coli 2QH6KRW0$; substrate; and dNBR pixel value), mineral soil pH, latitude, (IILFLHQF\'+Į75&RPSHWHQW&HOOV,QYLWURJHQ*UDQG and elevation were all significantly correlated directly or ,VODQG1HZ 300 ÉCOSCIENCE, VOL. 20 (3), 2013 EXUQFDWHJRULHV )RUWKH,QGLFDWRU6SHFLHV$QDO\VLVZH EFFECTS OF FIRE SEVERITY AND SITE-LEVEL CHARACTERISTICS used 4999 permutations in the Monte Carlo test of sig- Variance in fungal community composition was more QLILFDQWREVHUYHGPD[LPXPLQGLFDWRUYDOXHVIRU278V strongly correlated with fire severity than with other abiotic ,QDGGLWLRQ278UHODWLYHDEXQGDQFHZDVYLVXDOL]HGXVLQJ IDFWRUV )LJXUH 7DEOH,,,D 7KHILQDOVROXWLRQRIWKH two-way cluster analysis dendrograms at the site level. RUGLQDWLRQIRUWKH278GDWDVHW\LHOGHGGLPHQVLRQVZLWKD For computation of the dendrogram we used the Sorensen ILQDOLQVWDELOLW\RI7KHILQDOVWUHVVIRUWKH278GDWD distance measure and the nearest neighbour group link- VHWZDVVXJJHVWLQJDJRRGRUGLQDWLRQZLWKORZULVNRI age method. All statistical analyses were performed in GUDZLQJIDOVHLQIHUHQFHV &ODUNH 0F&XQH *UDFH -03 6$6,QVWLWXWH,QF&DU\1RUWK&DUROLQD>86$@ 7KHD[HVUHSUHVHQWHG RIWKHYDULDQFHLQ with the exception of the multivariate analysis of mycor- IXQJDOFRPPXQLW\VWUXFWXUHZLWKD[LVFRQWULEXWLQJ UKL]DOFRPPXQLWLHVZKLFKZDVSHUIRUPHGLQ3&25' DQGD[LV$[LVZDVFRUUHODWHGZLWKYDULDEOHVWKDW 0-06RIWZDUH'HVLJQ*OHQHGHQ%HDFK2UHJRQ>86$@ UHIOHFWILUHVHYHULW\ &%,DQGG1%5 DQGODWLWXGH )LJXUH 7DEOH,,,D %\FRQWUDVWD[LVUHSUHVHQWVDFRPSOH[DFLG- ity gradient, as indicated by correlations with elevation Results DQGPLQHUDOVRLOS+ )LJXUH 7DEOH,,,D 7KLVLVOLNHO\ related to landscape position and the associated relationship MOLECULAR OPERATIONAL TAXONOMIC UNITS SAMPLED between topography and soil pH. Although fire severity 7RGHVFULEHWKHFRPSRVLWLRQRIP\FRUUKL]DOIXQJLRQ was an important correlating factor, we did not observe dis- UHVSURXWLQJVKUXEVDIWHUILUHZHVDPSOHGB. nana shrubs tinct clustering of sites by fire-severity class in ordination across a fire-severity gradient (30 shrubs from intensively space, nor a significant difference in fungal community VDPSOHGVLWHVDQGIURPH[WHQVLYHO\VDPSOHGVLWHV :H FRPSRVLWLRQDPRQJVHYHULW\FODVVHV 0533$ ± KDGD VXFFHVVUDWHZLWKVHTXHQFLQJ RIUHDGV P 2XUVXEVHTXHQWDQDO\VHVGHVFULEHGEHORZ >FORQHVîIRUZDUGDQGUHYHUVHVHTXHQFLQJ@ZHUHHOLP- showed that the effects of fire on the components of com- inated during clean up). We obtained 498 non-chimeric munity composition depended upon the scale of taxonomic sequences from clone libraries from 44 B. nana shrubs resolution. DFURVVDOOVLWHV ZHFRXOGQRWDPSOLI\'1$IURPRI Surprisingly, when we examined the effects of fire at WKHRULJLQDOVKUXEV 7KHVHVHTXHQFHVZHUHJURXSHG the coarsest taxonomic scale, we observed no shift to dom- LQWR278V $SSHQGL[,, 7DEOH, 7KLUW\RIWKH278V inance of pyrophilic, fire-specialist fungi in the Ascomycota nor decline in fire-sensitive, late-stage fungi in the occurred more than once, and 22 were singletons. Using Basidiomycota within study sites or across the burned study %/$67 VHDUFKHV DQG WUHH EXLOGLQJ ZH ZHUH DEOH WR area, contrasting with our third hypothesis. For example, GHVFULEHWD[DWRWKHVSHFLHVOHYHO WD[DWRWKH the proportion of taxa in the Basidiomycota did not decline JHQXVOHYHO DQGWD[DWRIDPLO\RUGHURUFODVV with increasing fire severity (F P QRU WD[RQZHFRXOGLGHQWLI\RQO\WRVXESK\OXP was it related to residual soil organic depth (F 7KHGRPLQDQW278VUHSUHVHQWHGGLIIHUHQWIXQFWLRQDO P 7KHUHZDVQRGLIIHUHQFHEHWZHHQWKHDYHUDJH JURXSV (0)'6((50DQGSDWKRJHQV 7DEOH,, $VLQ abundance of clone sequences for fungi in the Ascomycota other mycorrhizal studies, our rarefaction curves suggest and the Basidiomycota observed at each site (T that we would have recorded more taxa if we had sampled P $YHUDJHUDZDEXQGDQFHRIFORQHVHTXHQFHVIRU additional shrubs within each site, especially in the exten- IXQJLLQWKH%DVLGLRP\FRWDSHUVLWHZDV 6( VLYHO\VDPSOHGVLWHV2QWKHRWKHUKDQGDGGLWLRQDOVDPSOLQJ DQGIRUIXQJLLQWKH$VFRP\FRWD 6( 7KHUH was not feasible in many of these sites, as we sampled all was also no significant difference in the taxon richness available shrubs along the sparsely vegetated transects. of fungi in the Ascomycota and Basidiomycota for a Rare fungal species were likely underrepresented, given that site (T ± P ,QWRWDODFURVVDOOVLWHV RIWKH278VZHUHVLQJOHWRQV:HWKHUHIRUHXVHGWKH we observed 28 taxa in the Ascomycota and 24 in the rarefied taxonomic richness values for further analysis of Basidiomycota. Richness varied from 0 to 9 taxa for 278ULFKQHVV $VFRP\FRWDDQGWRIRU%DVLGLRP\FRWDDFURVVVLWHV TABLE II. &ODVVL¿FDWLRQDQGSUHVHQFHRIWKHPRVWDEXQGDQWRSHUDWLRQDOWD[RQRPLFXQLWV 278V DFURVV¿UHVHYHULW\FODVVHV 278 DEXQGDQFH %HVWPDWFKGHVFULSWLRQ 7\SHa Unburned Low Moderate High Russula decolorans EMF x x x Meliniomyces sp. ERM x x x x 39 Phialocephala fortinii DSE x x x 32 Sordariomycetidae sp. likely DSE x x x Russula sp. EMF x x x +HORWLDOHVVS OLNHO\(50 [ [ [ [ 22 Chalara sp. pathogen x Lactarius glyciosmus EMF x x Meliniomyces variabilis ERM x x Inocybe borealis EMF x a (0) HFWRP\FRUUKL]DOIXQJL(50 HULFRLGP\FRUUKL]DOIXQJL'6( GDUNVHSWDWHHQGRSK\WH HEWITT ET AL.: TUNDRA FIRE EFFECTS ON MYCORRHIZAL COMMUNITIES Unburned or genus level, the coarser taxonomic identities indicate that Low these fungi are not known pyrophilic ascomycetous fungi. Moderate When we considered both the presence and relative High DEXQGDQFHRIWD[DKRZHYHUZHREVHUYHGWKDW278DEXQ- dance declined across the fire-severity gradient for many of WKH278VVLJQLILFDQWO\FRUUHODWHGZLWKD[LVLQWKHRUGLQD- Elevation WLRQELSORW$OPRVWDOORIWKHGRPLQDQW278VKDGVWURQJ SRVLWLYHFRUUHODWLRQVZLWKD[LVUHSUHVHQWLQJVLWHVDORQJ dNBR pixel WKHGHFUHDVLQJILUHVHYHULW\JUDGLHQW 7DEOH,,,E 7KHWZR way cluster dendrogram highlights these patterns, showing Veg severity CBI DODFNRIYDULDWLRQLQ278SUHVHQFHDFURVVILUHVHYHULWLHV Sub severity EXWVWURQJYDULDWLRQLQDEXQGDQFHRI278VDFURVVDOOWKH Latitude VLWHVRQWKHILUHVHYHULW\VSHFWUXP )LJXUH 7KHVHUHVXOWV show that although late-stage fungi are present after fire, Complex acidity gradient suggesting continuity between pre- and post-fire fungal communities, variability in the relative abundance of com- PRQ278VIROORZVK\SRWKHVL]HGSDWWHUQVDFURVVWKHILUH Mineral soil pH severity gradient. When we examined the number of rare 278VWKDWZHUHH[FOXGHGIURPWKHRUGLQDWLRQDQDO\VLVZH observed an increase in the percentage of singletons in more Complex fire-severity gradient VHYHUHO\EXUQHGVLWHV XQEXUQHG ORZ PRGHUDWH DQGKLJK SHUKDSVUHIOHFWLQJ FIGURE 2. NMDS ordinations of mycorrhizal community structure: less competition from abundant taxa. biplot of important environmental variables significantly correlated (r2 ZLWK278P\FRUUKL]DOFRPPXQLW\VWUXFWXUHIRUXQEXUQHG ORZPRGHUDWHDQGKLJKVHYHULW\VLWHV7KHYHFWRUVVKRZWKHGLUHFWLRQ Discussion and strength of correlations of environmental variables in relation to each site represented by mycorrhizal community structure (presence and abun- 7KLVLVWKHILUVWVWXG\WRLQYHVWLJDWHWKHHIIHFWRIILUH GDQFH ³9HJVHYHULW\´LVEXUQVHYHULW\RIYHJHWDWLRQDQG³6XEVHYHULW\´LV on variation in mycorrhizal composition of a dominant burn severity of substrate from each site. tundra shrub, Betula nana7KH$5)ZDVDUDUHHYHQWLQ At the scale of community richness, we observed a terms of fire severity and magnitude, and exploration of GHFOLQHLQWKHHVWLPDWHG278ULFKQHVVDVILUHVHYHULW\ &%, post-fire processes offers an opportunity to infer how spe- 2 cies may respond to the projected increases in tundra fire increased (F P R )LJXUH 7KLVGHFUHDVHLQ278ULFKQHVVZDVLQGHSHQGHQWRIWKH IUHTXHQF\DQGVHYHULW\ZLWKKLJKODWLWXGHZDUPLQJ,QRXU study, fungal composition correlated with a hierarchy of UHVLGXDORUJDQLFVRLOGHSWKVSRVWILUH &KDR) P ZKLFK FRQWUDVWHG ZLWK RXU K\SRWKHVHV fire-related, edaphic, and landscape variables. However, dis- developed from post-fire temperate and boreal sys- entangling the effect of each predictor variable is difficult, tems (Neary et al. 'DKOEHUJ &HUWLQL because the ARF was a natural fire. Although fire severity )LJXUH affected several aspects of fungal composition, including 7KHGRPLQDQWP\FRUUKL]DOWD[DGLGQRWVKRZVWURQJ WRWDO278ULFKQHVVDQGDEXQGDQFHRIGRPLQDQWWD[DRXU affinities with fire severities across the gradient. We observations contrasted with some of our hypotheses based observed no increase in presence of fire-specialists or RQVWXGLHVIURPERUHDODQGWHPSHUDWHIRUHVWV 9LVVHU decrease in presence of late-stage fungi across the gradient. Baar et al. 7KHPRVWDEXQGDQW278VDQGWKHGLIIHUHQWIXQJDOJURXSV (50(0)'6(DQGSXWDWLYHSDWKRJHQ RFFXUUHGLQ DISTRIBUTION OF FUNGI IN RELATION TO FIRE DISTURBANCE PXOWLSOHILUHVHYHULW\FODVVHV 7DEOH,, LQGLFDWLQJWKDW 7KHSULQFLSDOILQGLQJVRIRXUVWXG\ZHUHWKHODFN none were highly specialized for a given fire severity. of fire-specialist fungi across the burn scar and the persis- Similarly, the indicator species analysis produced no sig- tence of late-stage taxa on shrubs regardless of fire sever- QLILFDQWREVHUYHGLQGLFDWRUYDOXHVIRU278VLQWKHEXUQ LW\,QFRQWUDVWZLWKVWXGLHVIURPPRUHILUHSURQHELRPHV severity classes, showing that the presence and abundance we did not detect a decline in fire-sensitive fungi in the of the common taxa were not specifically associated with Basidiomycota or the pronounced dominance of fire-spe- DILUHVHYHULW\FODVV&RXQWHUWRRXUK\SRWKHVLV )LJXUH cialists in the Ascomycota after fire (Grogan, Baar & Bruns, late stage-fungi were present across fire-severity classes &DLUQH\ %DVWLDV EXWVHH-RQVVRQ et al., 7DEOH,, 7KHPRVWDEXQGDQW278VLQWKH%DVLGLRP\FRWD D ,WLVSRVVLEOHWKDWZHGLGQRWFDSWXUHWKHSHDN were in the genera Russula, Lactarius, Inocybe, Clavulina, colonization of shrubs by post-fire fungi because we sam- and Tomentella, several species of which are known late- SOHGGXULQJWKHVHFRQGJURZLQJVHDVRQ7KHGRPLQDQFH successional fungi, e.g., Lactarius glyciosmus (Fleming, of pyrophilic ascomycetous fungi, mainly from the order DQGRussula decolorans 9LVVHU 7KHPRVW Pezizales (Fujimura et al. LVZHOOGRFXPHQWHGIRU DEXQGDQW278VLQWKH$VFRP\FRWDZHUHLQWKHJHQXV WKHPRQWKVWR\DIWHUILUH &DLUQH\ %DVWLDV Meliniomyces, the subclass Sordariomycetidae, and the However, other studies have observed the prevalence of RUGHU+HORWLDOHVDORQJZLWKSXWDWLYHSDWKRJHQLQWKH ascomycete fire-fungi in the second growing season after genus Chalara 7DEOH,, 7KRXJKZHFRXOGQRWLGHQWLI\ fire (Grogan, Baar & Bruns, 2000) and their persistence for many of the dominant fungi in the Ascomycota to species XSWR\DIWHUILUH 9LVVHU 302 ÉCOSCIENCE, VOL. 20 (3), 2013 TABLE III. Correlations between environmental variables or fungal taxa and NMDS ordination of mycorrhizal communities in the ARF burn: a) Environmental variables correlated with NMDS ordinations. b) Fungal taxa presence and relative abundance correlated with the axes of 10'6RUGLQDWLRQV1RWHWKDWDSRVLWLYHYDOXHRQD[LVLQGLFDWHVORZ¿UHVHYHULW\ D &RPSOH[¿UHVHYHULW\JUDGLHQW &RPSOH[DFLGLW\JUDGLHQW $[LVa Axis 2 a Environmental variables r r2 r r2 /DWLWXGH ± 0.398 ± 6ORSH (OHYDWLRQ ± 0.306 %XUQVHYHULW\VXEVWUDWH ± 0.270 0.244 0.060 Burn severity vegetation –0.494 0.244 &%, ± 0.237 0.283 0.080 0LQHUDOVRLOS+ ± 0.470 $YHUDJHRUJDQLFVRLOGHSWK Active layer depth 0.090 0.008 0.286 0.082 G1%5FODVV ± dNBR pixel –0.466 0.217 E &RPSOH[¿UHVHYHULW\JUDGLHQW &RPSOH[DFLGLW\JUDGLHQW $[LVa Axis 2 a Phylum b 2SHUDWLRQDOWD[RQRPLFXQLW r r2 r r2 B Russula decolorans 0.245 ± A MeliniomycesVS 0.317 –0.600 0.360 A Phialocephala fortinii $ 6RUGDULRP\FHWLGDHVS ± 0.544 ± 0.206 B RussulaVS 0.458 A Helotiales sp. 0.488 0.238 –0.684 0.468 A ChalaraVS 0.317 0.496 B Lactarius glyciosmus 0.279 0.553 A Meliniomyces variabilis 0.489 B Inocybe borealis 0.345 –0.494 0.244 B ClavulinaVS ± ± 0.275 % 5XVVXODFHDHVS 0.223 –0.648 0.420 % &DQWKDUHOODOHVVS 0.345 –0.494 0.244 B LactariusVS 0.345 –0.494 0.244 B Lactarius torminosus ± B LactariusVS % $JDULFRP\FRWLQDVS ± ± 0.247 B Pseudotomentella tristis 0.317 0.496 B RussulaVS ± 0.300 ± B Tomentella subclavigera 0.345 –0.494 0.244 A Cladophialophora chaetospira ± ± 0.255 A MeliniomycesVS 0.306 A Meliniomyces bicolor 0.317 0.496 $ +HORWLDOHVVS 0.345 –0.494 0.244 B HebelomaVS 0.345 –0.494 0.244 A Phialocephala fortinii 0.930 0.864 % 5XVVXODFHDHVS 0.223 –0.648 0.420 a Bold r2YDOXHVKDYHVLJQL¿FDQWFRUUHODWLRQVZLWK106D[HV b )XQJDOSK\ODDUHDVIROORZV $ $VFRP\FRWDDQG% %DVLGLRP\FRWD 7KH GRPLQDQW (0) LQ RXU VWXG\ ZHUH DOO 5 Basidiomycota in the genera Russula, Lactarius, and Inocybe, which are abundant across arctic habitats, are 4 well represented across several host species, and have a KLJKVSHFLHVGLYHUVLW\ 7LPOLQJ et al. ,QFRQFHUW 3 with the broad distribution patterns of these fungi across WKHILUHVHYHULW\JUDGLHQW 7DEOH,, VRPHRIWKHGRPLQDQW 2 fungi, such as Russula decolorans and Lactarius glycios- richness [Chao1]) 1 musDUHNQRZQWRRFFXULQODWHVXFFHVVLRQDOVHUHV7KLV OTU SQRT(estimated suggests that these fungi survived fire in a mycelial state. 0 After fire, mycorrhizal fungi colonize roots from resident 0 0.5 1 1.5 2 2.5 3 3.5 mycelium and infected root tips that survive, dispersed Composite burn index spores, or fire-resistant propagules (Baar et al. ,QFRQWUDVWZLWKVWXGLHVIURPPRUHILUHSURQHWHPSHUDWH FIGURE 3. Fungal richness decreased with increasing burn sever- LW\5HJUHVVLRQRIWKHHVWLPDWHG278ULFKQHVV &KDR VDPSOHGDFURVV zones, our results indicate that the dominant fungi observed sites and rarified to an individual shrub (F P across the ARF were not part of a resistant propagule com- R2 &%,LQFUHDVHGZLWKLQFUHDVLQJILUHVHYHULW\DQGLVDFRPSRVLWH munity distinctive from the composition in the unburned measure of substrate and vegetation burn severity. 303 HEWITT ET AL.: TUNDRA FIRE EFFECTS ON MYCORRHIZAL COMMUNITIES VLWH 7D\ORU %UXQV 0DQ\RIWKHGRPLQDQWWD[D growth, not colonization from resistant propagules, to be observed across the fire-severity gradient were present the dominant process by which new roots are colonized at the unburned site, where we would expect mycelial (Jonsson et al. E +HQFHWKHXELTXLWRXVGLVWULEXWLRQ Matrix coding Fire severityHigh Moderate Low Unburned Min Max Site 63 37 13 101 41 32 30 20 104 51 27 5 1 Plants species Russula decolorans * Meliniomyces sp. * Phialocephala fortinii Sordariomycetidae sp. * Russula sp. * Helotiales sp. * Chalara sp. * Lactarius glyciosmus * Meliniomyces variabilis Inocybe borealis * Dominant OTUs Clavulina sp. Russulaceae sp. * Cantharellales sp. * Lactarius sp. * Lactarius torminosus Lactarius sp. Agaricomycotina sp. Pseudotomentella tristis * Russula sp. * Tomentella subclavigera * Cladophialophora chaetospira * Meliniomyces sp. * Meliniomyces bicolor * Helotiales sp. * Hebeloma sp. * Phialocephala fortinii * Russulaceae sp. * FIGURE 7ZRZD\FOXVWHUDQDO\VLVGHQGURJUDPRI278DEXQGDQFHDFURVVVDPSOHGVLWHVLQGLIIHUHQWEXUQVHYHULWLHV$EXQGDQFHRI278VLVVKRZQLQ JUH\VFDOHZKHUHWKHZKLWHVTXDUHVVKRZZKHUHWD[DDUHDEVHQWDQGWKHILOOHGEODFNVTXDUHVVKRZWKHKLJKHVWDEXQGDQFHRIWKH278 GHQRWHVWD[DWKDW ZHUHVLJQLILFDQWO\FRUUHODWHGZLWKD[LVRIWKHRUGLQDWLRQ5HODWLYHDEXQGDQFHIRUHDFKVLWHLVEDVHGRQWKHSRROHGQXPEHURIFORQHVSHUVKUXEZKLFKLVD FRPPRQZD\WRLQIHUSURSRUWLRQDOIXQJDOELRPDVVRUUHODWLYHDEXQGDQFH278GDWDZHUHWUDQVIRUPHGXVLQJWKH%HDO VVPRRWKLQJIXQFWLRQ 304 ÉCOSCIENCE, VOL. 20 (3), 2013 of dominant species, especially late-stage taxa, suggests on resprouting roots have a strong priority effect, rapidly that mycelial inoculum on the roots of surviving, resprout- colonize new short roots, and may be superior competi- LQJVKUXEVZDVPDLQWDLQHGRQWKHODQGVFDSHDIWHUILUH,Q WRUVIRUURRWWLSV2YHUDOOWKHILUHUHVSRQVHIRUP\FRUUKL]DO DGGLWLRQWR(0)ZHGHWHFWHG'6((50DQGSXWDWLYH communities associated with B. nana is different from that pathogen on the root systems of Betula nana. Similar to LQWKHDGMDFHQWERUHDOIRUHVW7ZRSRVVLEOHH[SODQDWLRQVIRU the distribution patterns of EMF taxa, the dominant taxa WKHVHELRPHGLIIHUHQFHVDUHWKDW WKHUDULW\RIWXQGUDILUHV in these fungal groups were ubiquitous root colonizers IRUDWOHDVWWKHODVW\PD\KDYHVHOHFWHGDJDLQVWILUH across the fire-severity gradient and showed no specialized specialist fungi, and 2) we sampled resprouting host plants fire response. that survived fire, whereas studies in other biomes have sampled soil, macrofungal fruit bodies, mycorrhizas in soil FIRE VERSUS ENVIRONMENTAL EFFECTS cores, and naturally occurring seedlings in sites where the Although we observed the persistence of late-stage host species was killed by fire. fungi and a lack of differentiation of fire-specialist and fire-sensitive fungi, wildfire severity was correlated with ECOLOGICAL SIGNIFICANCE OF FUNGAL LIFE HISTORY TRAITS variance in fungal community structure on resprouting 7KHSHUVLVWHQFHRIIXQJDOLQRFXOXPRQWKHODQGVFDSH B. nanaVKUXEV7KLVFRUUHODWLRQLVFRQVLVWHQWZLWKWKDW after an unprecedented wildfire event has strong ecological found in other post-fire studies (Dahlberg et al. implications for nutrient cycling and vegetation establish- Smith et al. +DPPDQ%XUNH 6WURPEHUJHU ment because of the life-history traits of the fungi that 7KHFRUUHODWLRQEHWZHHQIXQJDOFRPSRVLWLRQDQGODWLWXGH VXUYLYHGILUHRQWKHURRWVRIUHVSURXWLQJVKUXEV2QHRI in the ordination is also likely related to fire severity given the central findings from our study is the presence of late- the northward spread of the fire and increase in burn sever- VWDJHIXQJLUHJDUGOHVVRIILUHVHYHULW\,QJHQHUDOHQ]\PDWLF ity throughout the duration of the burn (Jandt et al. competency can be related to fungal successional stage While fire-severity variables were the primary factors cor- and associated declines in high-resource-quality substrates related with overall fungal community composition in (Dighton, 2003). For example, Russula decolorans, a Arctic Alaska, there was also a strong relationship between late-stage fungus, is preferentially found in low-nutrient, mineral soil pH, elevation, and fungal communities. late-successional-stage environments due to its affinity As with many observational studies of fire effects, the IRUORZTXDOLW\QLWURJHQVRXUFHV 7ROMDQGHU et al., 2006). effects of fire severity and other environmental variables Although the role of EMF as the primary conduits of were confounded because the ARF was a natural fire. We inorganic and organic forms of soil nutrients for host plants observed collinearity between fire severity, landscape, and in the Betulaceae (Smith & Read, 2008) is well under- VRLOYDULDEOHV,WLVOLNHO\WKDWHOHYDWLRQDQGPLQHUDOVRLOS+ stood, the ecological significance of associations with ERM greatly influence the structure of mycorrhizal communities DQG'6(UHPDLQXQFHUWDLQ 'HVOLSSH 6LPDUG pre-fire and that these factors also relate to the severity of 1HZVKDP 7KHSHUVLVWHQFHRIP\FRUUKL]DOWD[DPD\ VLWHOHYHOEXUQFKDUDFWHULVWLFV,WLVSRVVLEOHWKDWWKHGLVWDQFH provide resilience in the ecological function of a system between sampling sites may have influenced the observed by alleviating post-disturbance lag times in ecological pro- structure of the mycorrhizal communities. However, we cesses such as proteolytic activity, decomposition, and plant observed similar taxa across all the sites, indicating that dis- acquisition of soil resources, depending on the availability tance between sites is not the primary driver of differences of mycorrhizal and DSE fungi. in mycorrhizal community structure. FUNGAL RESILIENCE WITH A DYNAMIC TUNDRA FIRE REGIME COMPARISONS WITH THE ADJACENT BOREAL FOREST ,QRWKHUHFRV\VWHPVZKHUHP\FRUUKL]DODYDLODELOLW\LV ,QWKHERUHDOIRUHVWILUHVHYHULW\DIIHFWVWKHOHJDFLHV constrained after disturbance, early colonizing shrubs pro- RIP\FRUUKL]DOIXQJLLQPDLQZD\V E\DOWHULQJKRVW vide mycelial inoculum for establishing trees and shrubs, plant availability, which determines the potential for thus influencing successional trajectories (Nara & Hogetsu, mycelial growth, and 2) through the combustion and heat- 1DUDDE ,QWXQGUDWKHSUHVHQFHDQGGHQVLW\ ing of organic soil, which directly causes fungal mortality of resprouting B. nana host plants appear to mediate fire- 'DKOEHUJ ,QWKHWXQGUDHFRV\VWHPKRZHYHUUHVLG- severity effects on fungal composition and may provide ual soil organic depths did not correlate with fire severity a potentially important source of mycorrhizal resilience, in the ARF burn scar (Mack et al. DQGWKH\GLGQRW both in maintaining mycorrhizal diversity and in facilitating correlate with fungal richness or composition in our study. vegetation establishment under future scenarios of warm- 7KHVHREVHUYDWLRQVFRQWUDVWZLWKWKHLPSRUWDQWUROHRI ing and fire. After fire in the tundra, where resprouting organic soil depth in explaining the decline in fungal rich- shrubs are dominant components of the vegetation and fires ness after fire in the boreal forest (Dahlberg et al. $W are commonly of lower severity, we hypothesize that fire the scale of the entire ARF, host plant cover decreased with effects on fungal community composition will not likely increased fire severity (E. A. Miller, unpubl. data), which OLPLWVKUXESHUIRUPDQFHDQGYHJHWDWLRQFKDQJH,ISRVWILUH PD\PHDQWKDWWKHORZHU278ULFKQHVVRQVHYHUHO\EXUQHG host plant densities do affect resilience of fungal commun- sites can be explained through a species–area relationship ities to fire, it seems likely that fire will have the greatest (Peay et al. RUWKURXJKFRPSHWLWLYHLQWHUDFWLRQV effect on mycorrhizal shrub performance and vegetation among fungi for root tips (Kennedy, Peay & Bruns, 2009; change in ecosystems such as graminoid tundra that have .HQQHG\ :HK\SRWKHVL]HWKDWIXQJLWKDWVXUYLYHILUH low shrub density or in sites that burn so severely that host HEWITT ET AL.: TUNDRA FIRE EFFECTS ON MYCORRHIZAL COMMUNITIES SODQWVUDUHO\UHVSURXW,QWXQGUDSHUVLVWHQFHRIP\FRUUKL]DO Dahlberg, A., 2002. Effects of fire on ectomycorrhizal fungi in communities after fire disturbance could facilitate shrub fennoscandian boreal forests. Silva Fennica, 36: 69–80. expansion and range expansion of neighbouring ectomycor- 'DKOEHUJ$-6FKLPPHO$)67D\ORU +-RKDQQHVVRQ rhizal boreal forest tree species as warming continues at 3RVWILUHOHJDF\RIHFWRP\FRUUKL]DOIXQJDOFRPPXQLWLHV high latitudes. in the Swedish boreal forest in relation to fire severity and log- JLQJLQWHQVLW\%LRORJLFDO&RQVHUYDWLRQ ± 'HDFRQ - 6 'RQDOGVRQ ) /DVW 6HTXHQFHV DQG Acknowledgements interactions of mycorrhizal fungi on birch. Plant and Soil, 7KLVUHVHDUFKZDVVXSSRUWHGE\IXQGLQJIURPWKH$ODVND ± Experimental Program to Stimulate Competitive Research 'HVOLSSH- 66LPDUG%HORZJURXQGFDUERQWUDQVIHU 1DWLRQDO6FLHQFH)RXQGDWLRQ$ZDUG DQGWKHVWDWH among Betula nana may increase with warming in Arctic tun- of Alaska, a National Science Foundation Graduate Research GUD1HZ3K\WRORJLVW ± Fellowship (Division of Graduate Education - 0639280 and Dighton, J., 2003. Fungi in Ecosystem Processes. Marcel Dekker, DQGD&HQWHUIRU*OREDO&KDQJHVWXGHQWDZDUGWR 1HZ 306 ÉCOSCIENCE, VOL. 20 (3), 2013 +XDQJ; $0DGDQ &$3 $'1$VHTXHQFHDVVHPEO\ 0ROLQD5+0DVVLFRWWH -07UDSSH 6SHFLILFLW\SKH- SURJUDP*HQRPH5HVHDUFK ± nomena in mycorrhizal symbioses: Community-ecological ,QGH[ )XQJRUXP 3DUWQHUVKLS ,QGH[ )XQJRUXP /DQGFDUH FRQVHTXHQFHVDQGSUDFWLFDOLPSOLFDWLRQV3DJHV±LQ Research, Lincoln, New Zealand and RBG Kew : Mycology, 0)$OOHQ HG 0\FRUUKL]DO)XQFWLRQLQJ $Q,QWHJUDWLYH 6XUUH\ 2QOLQH >85/@ KWWS ZZZLQGH[IXQJRUXPRUJ 3ODQW±)XQJDO 3URFHVV &KDSPDQ DQG +DOO 1HZ HEWITT ET AL.: TUNDRA FIRE EFFECTS ON MYCORRHIZAL COMMUNITIES 7UHVHGHU..0&0DFN $&URVV5HODWLRQVKLSV :HLQ5 )UHTXHQF\DQGFKDUDFWHULVWLFVRI$UFWLFWXQGUD among fires, fungi and soil dynamics in Alaskan boreal forests. ILUHV$UFWLF ± (FRORJLFDO$SSOLFDWLRQV ± :KLWH36 3DWWHUQSURFHVVDQGQDWXUDOGLVWXUEDQFHLQ 9LVVHU6 (FWRP\FRUUKL]DOIXQJDOVXFFHVVLRQLQMDFNSLQH YHJHWDWLRQ%RWDQLFDO5HYLHZ ± VWDQGVIROORZLQJZLOGILUH1HZ3K\WRORJLVW ± 9LHUHFN/$&7'\UQHVV$5%DWWHQ .-:HQ]OLFN :KLWH77%UXQV6/HH -7D\ORU $PSOLILFDWLRQDQG 7KH$ODVNDYHJHWDWLRQFODVVLILFDWLRQ86'$)RUHVW direct sequencing of fungal ribosomal RNA genes for phylo- 6HUYLFH *HQHUDO 7HFKQLFDO 5HSRUW 31:*75 JHQHWLFV3DJHV±LQ0,QQLV'*HOIDQG-6QLQVN\ 3RUWODQG2UHJRQ 7:KLWH HGV 3&53URWRFROV $*XLGHWR0HWKRGVDQG Walker, D. A., M. K. Raynolds, F. J. A Daniels, E. Einarsson, $SSOLFDWLRQV$FDGHPLF3UHVV1HZ Appendix I APPENDIX,Verification of the adequacy of root tip richness associated with each shrub, we used a t-test to com- sampling and sequencing effort for each Betula nana shrub pare samples with different numbers of tips from the same XVLQJ$5,6$ shrub. Preliminary analyses on 6 shrubs demonstrated that WLSVZHUHVXIILFLHQWWRUHSUHVHQWWKHULFKQHVVRIIXQJDO ,Q RXU YLHZ VHTXHQFLQJ GDWD SURYLGH EHWWHU DPSOLFRQVSUHVHQWRQDURRWV\VWHP WLSVversus 36 tips, GLVFULPLQDWLRQ RI IXQJDO WD[D ZKLOH$5,6$ SURYLGHV T6 ± P 7KHUHZDVQRWDVLJQLILFDQWLQFUHDVH a sensitive, rapid, and cost-effective method to in the richness we observed for each shrub when we sam- estimate species richness, although it does not provide SOHGPRUHWKDQURRWWLSV phylogenetically explicit information (Liu et al. 7RYHULI\WKDWZHDGHTXDWHO\VHTXHQFHGWKHIXQJDO :HXVHG$5,6$WRGHWHUPLQHWKHDGHTXDWHQXPEHURIURRW FRPPXQLW\ZHFRPSDUHGRXU$5,6$SURILOHVWRWKH278 tips to sample to capture fungal richness for each shrub and GDWDVHWIRUHDFKVKUXE$5,6$SURILOHVZLOOUHSUHVHQWUDUH to verify that we sufficiently sequenced the taxa for each taxa, while sequences from the clone libraries could pot- shrub. HQWLDOO\FDSWXUHRQO\WKHGRPLQDQW278VDVVRFLDWHGZLWK HDFKVKUXE7KHGLVWULEXWLRQVRI278DQG$5,6$ULER- METHODS type abundances were examined for normality by visually )RU $5,6$ P/ UHDFWLRQ PL[HV ZHUH SUH- inspecting normal quantile plots, assessing skewness and SDUHG XVLQJ WKH IRUZDUG SULPHU )$0,76) kurtosis, and using the Shapiro–Wilks W test. We used a &77**7&$777$*$**$$*7$$ *DUGHV %UXQV paired tWHVWWRFRPSDUHWKHREVHUYHG$5,6$DQG278 ODEHOHGRQWKH¶HQGZLWK)$0DIOXRUHVFHLQDPLGLW richness on individual shrubs. We also used the Mantel test (Applied Biosystems, Carlsbad, California, USA), and 0DQWHO ZLWK0RQWH&DUORSHUPXWDWLRQVWRHYDOXDWH WKHUHYHUVHSULPHU,767&&7&&*&77$77*$7$7*& the null hypothesis that there was no significant relationship (White et al. XVLQJ0DS0DUNHU;UKRGDPLQH between the community structure for site-level commun- VL]HVWDQGDUG %LR9HQWXUHV0XUIUHHVERUR7HQQHVVHH86$ LW\SURILOHVUHSUHVHQWHGLQWKH278DQG$5,6$PDWULFHV and following the protocol of Bent et al. 6DPSOHV 7RIRUPDWV\PPHWULFDOPDWULFHVIRUWKHVLWHOHYHO$5,6$ ZHUH UXQ RQ DQ$%, *HQHWLF$QDO\]HU $SSOLHG DQG278PDWULFHVZHHOLPLQDWHGVLWHV VLWHVDQG %LRV\VWHPV)RVWHU&LW\1HZ-HUVH\86$ 3RSFP ZLWKRXWOLHUIXQJDOFRPPXQLW\SURILOHV7KHUHVXOWLQJ DUUD\75)/3BSURWRFRO 278PDWUL[KDGVLWHV URZV ZLWK278V FROXPQV BIOINFORMATICS DQGWKHVHFRQGPDWUL[KDGVLWHV URZV ZLWK$5,6$ ribotypes (columns). :H XVHG *HQH0DSSHU $SSOLHG %LRV\VWHPV )RVWHU&LW\1HZ-HUVH\86$ WRUHDGHDFK$5,6$HOHF- We determined that we had sequenced enough clones to WURSKHURJUDP7KHSHDNKHLJKWVIRUHDFKIUDJPHQWZHUH adequately represent fungal richness for each shrub because relativized by dividing the fluorescence height for each WKHQXPEHUVRI278VREWDLQHGIURPVHTXHQFHVHTXDOHGWKH peak by the total fluorescence for a sample profile (Fisher QXPEHUVRI$5,6$ULERW\SHV 743 P ,Q 7ULSOHWW $5,6$ULERW\SHVZHUHELQQHGDWES addition, there was a significant relationship between the ELQVL]H 'XQEDU7LFNQRU .XVNH :HXVHGES VWUXFWXUHRI$5,6$DQG278FRPPXQLW\GDWDVHWV 0DQWHO $5,6$ULERW\SHELQVDVVXUURJDWHVIRU³VSHFLHV´LQFHUWDLQ test site matrices r P )RUVXEVHTXHQW IXQJDOFRPPXQLW\DQDO\VHV%LQQHG$5,6$ULERW\SHDEXQ- DQDO\VHVZHXVHGWKH278GDWDVHWEHFDXVHLWDOORZHGXV dance data have been archived with the Bonanza Creek to comment on taxon identity in addition to community /7(5 KWWS ZZZOWHUXDIHGXGDWDBEFIP structure. All statistical analyses were performed in JMP 6$6,QVWLWXWH,QF&DU\1RUWK&DUROLQD>86$@ ZLWK STATISTICS AND RESULTS the exception of the multivariate analysis of mycorrhizal 7RGHWHUPLQHWKHQHFHVVDU\QXPEHURIURRWWLSVWKDWZH FRPPXQLWLHVZKLFKZDVSHUIRUPHGLQ3&25' 0-0 needed to sample in order to adequately capture the fungal 6RIWZDUH'HVLJQ*OHQHGHQ%HDFK2UHJRQ>86$@ 308 ÉCOSCIENCE, VOL. 20 (3), 2013 Appendix II APPENDIX II, TABLE,2SHUDWLRQDOWD[RQRPLFXQLWV 278V IURPFORQHGVHTXHQFHVIURPBetula nanaVKUXEVVDPSOHGDFURVV¿UHVHYHULWLHVLQ the Anaktuvuk Burn. 2SHUDWLRQDO %HVWPDWFK 7D[RQRPLF 278 1R 1R *HQ%DQN %HVWPDWFK *HQ%DQN 8QLW DEXQGDQFH VKUXEV VLWHV DFFHVVLRQ GHVFULSWLRQ ,GHQWLW\ a 2YHUODSb Bit score c accession d 278 )- Russula decolorans .& 278 () MeliniomycesVS .& 278 $< Phialocephala fortinii .& 278 +0 6RUGDULRP\FHWLGDHVS .& 278 $< RussulaVS .& 278 +0 +HORWLDOHVVS .& 278 +0 ChalaraVS .& 278 '4 Lactarius glyciosmus .& 278 +0 Meliniomyces variabilis .& 278 -1 Inocybe borealis .& 278 $0 InocybeVS .& 278 +4 TomentellaVS .& 278 (8 ClavulinaVS .& 278 $0 5XVVXODFHDHVS .& 278 $% &DQWKDUHOODOHVVS .& 278 '4 LactariusVS .& 278 '4 Lactarius torminosus .& 278 $% LactariusVS .& 278 $< $JDULFRP\FRWLQDVS .& 278 $- Pseudotomentella tristis .& 278 $< RussulaVS .& 278 +4 Tomentella subclavigera .& 278 (8 Cladophialophora chaetospira .& 278 +0 MeliniomycesVS .& 278 $< Meliniomyces bicolor .& 278 -1 +HORWLDOHVVS .& 278 $) HebelomaVS .& 278 +4 TomentellaVS .& 278 $< Phialocephala fortinii .& 278 '4 5XVVXODFHDHVS .& 6,1* '4 5XVVXODFHDHVS .& 6,1* )- 5XVVXODFHDHVS .& 6,1* $< Phialocephala fortinii .& 6,1* $0 +HUSRWULFKLHOODFHDH .& 6,1* $< MeliniomycesVS .& 6,1* $% +HORWLDOHVVS .& 6,1* -1 ArticulosporaVS .& 6,1* -) ChalaraVS .& 6,1* $< Rhyzoscyphus ericae .& 6,1* -1 LaccariaVS .& 6,1* $% TomentellaVS .& 6,1* $< +HORWLDOHVVS .& 6,1* (8 Articulospora tetracladia .& 6,1* $< PhialocephalaVS .& 6,1* )- +HORWLDFHDHVS .& 6,1* +4 MeliniomycesVS .& 6,1* $< PhialocephalaVSFRPSOH[ .& 6,1* () MeliniomycesVS .& 6,1* () Meliniomyces variabilis .& 6,1* () MeliniomycesVS .& 6,1* +4 MelinomycesVS .& 6,1* $- TomentellaVS .& a ,GHQWLW\ RYHUDOOIUDFWLRQRILGHQWLFDOSRVLWLRQVDFURVVKLJKVFRULQJVHJPHQWSDLUV b 2YHUODS OHQJWKRIWKHKLWSDUWLFLSDWLQJLQDOLJQPHQWZLWKRXWWKHJDSV c %LWVFRUHV KLJKHUVFRUHVLQGLFDWHWKHJUHDWHUVLJQL¿FDQFHRIWKHDOLJQPHQWEHWZHHQWKHTXHU\VHTXHQFHDQGWKHKLWVHTXHQFH d *HQ%DQNDFFHVVLRQ WKHDVVLJQHG*HQ%DQNDFFHVVLRQQXPEHUIRUWKHUHSUHVHQWDWLYH278VHTXHQFH 309 HEWITT ET AL.: TUNDRA FIRE EFFECTS ON MYCORRHIZAL COMMUNITIES Appendix III APPENDIX III, TABLE,. &RUUHODWLRQVEHWZHHQYDULDEOHV D HQYLURQPHQWDODQG¿UHYDULDEOHVE UHVSRQVHYDULDEOHV D ,QGHSHQGHQWYDULDEOHV 6SHDUPDQ¶VFRUUHODWLRQ p /DWLWXGH 0LQHUDOVRLOS+ (OHYDWLRQ 0LQHUDOVRLOS+ (OHYDWLRQ /DWLWXGH %XUQVHYHULW\RIVXEVWUDWH 0LQHUDOVRLOS+ %XUQVHYHULW\RIVXEVWUDWH /DWLWXGH %XUQVHYHULW\RIVXEVWUDWH (OHYDWLRQ %XUQVHYHULW\RIYHJHWDWLRQ 0LQHUDOVRLOS+ %XUQVHYHULW\RIYHJHWDWLRQ /DWLWXGH %XUQVHYHULW\RIYHJHWDWLRQ (OHYDWLRQ %XUQVHYHULW\RIYHJHWDWLRQ %XUQVHYHULW\RIVXEVWUDWH &RPSRVLWH%XUQ,QGH[ &%, 0LQHUDOVRLOS+ &RPSRVLWH%XUQ,QGH[ &%, /DWLWXGH &RPSRVLWH%XUQ,QGH[ &%, (OHYDWLRQ &RPSRVLWH%XUQ,QGH[ &%, %XUQVHYHULW\RIVXEVWUDWH &RPSRVLWH%XUQ,QGH[ &%, %XUQVHYHULW\RIYHJHWDWLRQ G1%5SL[HO 0LQHUDOVRLOS+ G1%5SL[HO /DWLWXGH G1%5SL[HO (OHYDWLRQ G1%5SL[HO %XUQVHYHULW\RIVXEVWUDWH G1%5SL[HO %XUQVHYHULW\RIYHJHWDWLRQ G1%5SL[HO &RPSRVLWH%XUQ,QGH[ &%, b) Response variables Spearman’s correlation p (VWLPDWHGULFKQHVV &KDR 2EVHUYHGULFKQHVV 0DR7DX (VWLPDWHGULFKQHVV &KDR 3URSRUWLRQRIEDVLGLRP\FHWHULFKQHVV (VWLPDWHGULFKQHVV &KDR 3URSRUWLRQRIEDVLGLRP\FHWHDEXQGDQFH 2EVHUYHGULFKQHVV 0DR7DX 3URSRUWLRQRIEDVLGLRP\FHWHULFKQHVV 2EVHUYHGULFKQHVV 0DR7DX 3URSRUWLRQRIEDVLGLRP\FHWHDEXQGDQFH 3URSRUWLRQRIEDVLGLRP\FHWHULFKQHVV 3URSRUWLRQRIEDVLGLRP\FHWHDEXQGDQFH