jozsef geml et al- evidence for strong inter- and intracontinental phylogeographic structure in...

8/3/2019 Jozsef Geml et al- Evidence for strong inter- and intracontinental phylogeographic structure in Amanita muscaria, a

1/8

Molecular Phylogenetics a nd Evolution 48 (2008) 694-701

Contents lists avallable .atSciencelriteot :

Molecular Phylogenetics and Evolution

jou rna l home page: www.elsevier.com / loca te / ympev

Evidence for strong inter- and intracontinental phylogeographic structurein Amanita muscaria,a wind-dispersed ectomycorrhizal basidiomycete

jozsef Gemla, *, Rodharn E. Tulloss b,1, Gary A. Laursen a, Nina A. Sazanovac, D.L.Taylor a

A lnstitue o f Arctic Biology U niversit y o f Alaska FairiJanks. 311 Irving I Building. 902 N. f(ayulwl( Drive, P.O. Box 757000, FairiJanlcs,AI( 9977.5-7000, USAh 1'.0. Bo x .17, Roosevelt, N) 0855.5-0057, USA" LoiJaml ' ()/ y of BatollY. Institute oJ Biological ProiJl ems of the North, For East Branch of Russiall Academy of Sciences, P ortovaya Street 18, Magadan 685000, Russia

ARTICLE INFO ABSTRACT

A growing number of molecular studie s show that many fungi have phylogeographic structures and thattheir distinct lineages are usually limited to different continents. As a conservative test of the extent towhich wind-dispersed mycorrhizal fungi may exhibit phylogeographic structure, we chose to studyAmallita muscaria, a host-generali st, widespread, wind-disper sed fungus. In this paper, we docum entthe existence of several distinct phylogenetic species within A mllscaria, based on multilocus DNAsequence data. According to our findings, A muscaria has strong int ercontinental genetic disjunction s,and, more surprisingly, ha s strong intracontinental phylogeographic structure, particularly within NorthAmerica, often corresponding to certain habitats and/or biogeographic provinces. Our results indicatethat the view ofA. muscaria as a common, widespread, easily identifiable, ecologicallyplastic fungus witha wide niche does not correctly represent the ecological and biological realities, On the contrary, thestrong associations between phylo genetic species and different habitats support the developing pictur eof ecoregional endemisms and r elatively narrow to very narrow ni ches for some lineages.

2008 Elsevier Inc. Allrights r eserved .

Article history:Received 20 Nove mber 2 007Revised 18 Apri l 2008Accepted 2 2 April 2008Avail able online 29 April 2008

K eywords:Amanit a muscaria

B -tubulin geneFu ngiInternal t ranscribed s pacer r egionPhyl og e og raph yRibosomal large subunit geneTra nslat ion e lougation f actor 1 -al ph a gene

1. Introduction

Mycorrhizal associations are abundant and widespread inalmost all ecosystems and approximately 80% of land plant speciesform associations with mycorrhizal soil fungi (Trappe, 1987). Insuch symbioses, fungi support plants with mineral nutrients, waterand other services and the fungi, in turn, receive photosynthatesfrom the autotrophic plants. Given their abundance and theiref f ects on plant growth, they are known to play important rolesin ecosystems (e .g., Read and Perez -Mor e no , 2003: Johnson andGeh ring, 200 7).

The sensitivity of mycorrhi zal fungi to climate change is essen -tially unknown. The ability of an individual fungal species to copewith the changing environment is lik ely to be related to their ge-netic diversity. According to the basic principles of conservation

genetics, populations possessing a small amount of genetic diver-sity are more susceptible to regional extinction during times of stress (e.g., rapid climatic change) than genetically diverse popula-tions (e.g., Avise , 2 00 0). Unveiling phylogeographic s tructures of ectomycorrhizal species, assessing their genetic diversity, andreconstructing their past responses to past climatic change s willhelp to fill this important void.

* Corresponding author. Fax: +1 9074746967 .E-ma il a ddress:jgeml@iab. alaska.edu(J.Gemi).

1 Res. Assoc. (hons.). New York Botanical Gard en, Bronx. NY, USA.

1055-7903 /$ - see front matter @ 2008 Elsevier Inc. All rights reserved.doi: 10.10 1 G /j.ympev.2008.04.029

As a conservative test of th e extent to which wind-dispersedmycorrhizal fungi may exhibit phylogeographic structure, w echose to study Amanita muscaria (L.: Fr.) Hooker. Amanita muscaria is native to temperate and boreal for est regions of the NorthernHemisphere, where it is an ectomy corrhizal (ECM) fungus with awide host range (Trappe, 1987). Although it is most commonlyassociated with various birch (Betula), pine (Pinus), spruce (Picea),fir (Abies), and larch (Larix) species, it is known to form ECM asso-ciations with representatives of other genera, particularly when itsprimary hosts are rare or non-existent in a certain area. Amanita muscaria has traditionally been reported as a single morphosp e-cies, although morphological variation has led to the publicationof several intraspecific varieties, such as A. muscaria var. muscaria (L.: Fr.) Hooker, A. muscaria va r .alba Peck, A. muscaria var. flavivolv- ata (Singer) Jenkins, A. muscaria var .formosa (Pers.: Fr.) Bertillon in

DeChambre, A. muscaria var. persicina Jenkins, and A. muscaria va r

.regalis (Fr.) Bertillon in DeChambre (J enkins, 19 86).This well known fungus is predicted to have little biogeo-

graphic structure for the following reasons: (1) it is widelydistributed and abundant, (2 ) its spores are largely wind-dis-persed and it produces copious abov e-ground fruiting bodies(mushrooms ), (3) it associates with a wide variety of both conif-erous and angiosperm host tr ees, and thus appears to have littlehost-specificity , and (4) it is considered to be an invasive specieswhere it has been introduc ed in the Southern Hemisphere (Bag-ley and Orlovich, 20 0 4).

This file was created by scanning the printed publication. Text errors identifiedb the software have been corrected: however some errors ma remain.
http://www.elsevier.com%21locafe%21ympev/http://www.elsevier.com%21locafe%21ympev/http://www.elsevier.com%21locafe%21ympev/http://www.elsevier.com%21locafe%21ympev/http://www.elsevier.com%21locafe%21ympev/http://www.elsevier.com%21locafe%21ympev/http://www.elsevier.com%21locafe%21ympev/http://www.elsevier.com%21locafe%21ympev/


2/8

J. G e ml et al. / M o l e cul ar Ph y lo g en e ti c s an d Ev o luti o n 48 (2 00 8 ) 69 4 - 7 0 1

Our findings, reported herein, ar e in sharp contrast to these the-oretical predi ctions. Her e, we document the existence of severaldistinct phylogenetic species within A. muscaria , based upon rnul-tilocus sequence data and various phylogeneti c and popula tion ge-netic analyse s. We find that A. muscaria has strong intercontinentalgenetic disjunctions , and, more surpri singly, shows strong phylog-eographic patterns within North Am erica. Our result s indicate thatthe view of A. muscaria as a common, wide spread, easily identifi-able, ecologi cally plast ic fungus with a wide niche does not cor-rectly repre sent the ecological and biological realities. On thecontrary, the strong associations between phylogenetic clades(both at species and intraspecific lev els) and different habi tats sup-port the dev eloping pic ture of ecoregional endemisms and rela-tively narrow to very narrow niches for some lineages .

2, Materials and methods

2 .1 . Isolates and DNA extraction

Ninety- eight specim ens were collected from various geographicregions spanning the known distribution of A. muscaria (Table I ).

DNA was extracted from small samples of dri ed specimen s usingthe E-Z 96 Fungal DNA Kit (Omega Bio-tek , Inc.. Doraville, GA) orthe DNeasy Plant Mini Ki t (QIAGEN , Inc., Valencia, CA ).

2 .2. Isolates and DNA sequencing

DNA sequence data were obtained for four loci: B-tubulin gene,translation elongation factor 1-alpha gene (EF 1 -a) , nuclear largeribosomal subunit gene (LSU), and the internal transcrib ed spacer(ITS) + 5.8S ribosomal subu nit gene region. The primer s, PCR, andsequencing protocols have been described previously ( Geml etal., 2 0 05 , 2006). The only excepti on was the B-tubulin gene, forwhich a new primer pair was con structed to specifically amplifyand sequence an appro ximately 180-bp fragm ent containing themost informative known region within B-t ubulin in A. muscaria:

AMBT-F (5 'C

AA AGC

GGA GCA GGT AAT AA) and AMBT-R (5'AGT ACCG CC ACCAAG CGA AT ).

2.3. Phylogenetic analysis

Sequence data obtain ed f or both strands of each locus were edi-ted and assembled for each isolat e using CodonCode Aligner v. 1.3.4(CodonCode lnc., Dedham, MA). N ewly generated sequences weredeposited in Genbank (EU07182 6-EU072015). Additional, previ-ously published (Oda et al., 2004 ; Geml et al. , 2006) A. muscariaDNA sequences were included in the analyses ( Table 1). Homolo-gous sequences of Amanita pant herina (isolate FB-30958 ) (Oda etal., 2004) were used to root all tree s. Sequence alignments were ini-tiated using Clustal W (Thomp son et al., 1997) and subsequentlycorrected manually . To test for phylog enetic conflict among the dif-f erent loci (i.e., if individual gene trees signif icantly differed fromeach other), the partition homo geneity test (PHT) was performedwith 1000 randomized dataset s, using heuristi c searches with sim-ple addition of sequenc es in PAUP * 4b10 (Swof f ord, 2002). Analyse swere conducted using maximum-par simony (MP) and maximum-lik elihood (ML) methods in PAUP * and Garli 0.94 (Zwi ck l, 2006),respectively , For the latter, the b est-fit evolutionary m odel wasdetermined by comparin g different evolutionary models with vary-ing values of base frequ encies, sub stitution types, a-parameter of the y-distribution of variable sites, and proportion of invariablesites via the Akaike inf ormation criterion (AIC) using PAU P* andModeltest 3 ,06 (Posada and Crandall, 1998), Gaps were scored as'missing data', Bootstr ap (B) test (Felsenstein, 1985) was used with100 0 replicates in both MP and ML , with the maximum number of

trees saved set to 10 for each replicate. To compare different tre etopologie s, Shimodair a-Hasegawa test s were used (Shimodairaand Hasegawa, 1999 ). The High Perform ance Computing clu stermaintained by the UAF Biotechnology C omputing R esearch Group(http: // biotech.inbre.a laska.edu/ ) was used to run Clustal W, Mod -eltest, and Garli.

2.4. Polymorphism and divergence

The number of p olymorphi c sites and their distribution amon gthe major clade s was determined for sequence data generatedfrom all loci (ITS, B-t ubulin, EF1-a , LSU).Within specie s, nucleotid ediversity was measur ed using rr, the average number of nucleotid edifferen ces among sequence s in a sample (Nei and L i, 1979).Between species, divergence was measured as D XY the averagenumber of nucleotid e substitutions per site between species pair s(Nei and Kumar, 2000). In addition, genetic differentiation ( F s t )(Huds on e t al.,1992), the number of fixed differenc es, and sharedmutation s were calculated for the specie s pairs, as were thenumb er of position s that were polymorphi c in one phylogen eticspecies but monomorphic in the other. Measures of v ariation anddifferentiation wer e performed with the computer pr ogra m DnaSP

v .4.10 .9 (Ro za s and Rozas , 1999),

2.5. Genetic differentiation among populations within ph ylogenetics pecies

Because the phylogenetic species mentioned ar e non- inter-breeding entities, the population-level analyses w ere condu ctedseparately for each species clade that cont ained enough specimen swith intraspecific variation : namely, clad es I and II. Identical se -quenc es were collapsed into haplotyp es using SNAP Map (Ayloret al., 2006) after excluding insertion or deletions (indels) and infi -nite sites violation s. The analyses present ed here assume an infi-nite sites model , under which a polymorphic site i s caused byexactly one mutation and th ere can be no more than two segregat-ing bases. Site compatibility matrices w ere generated from each

haplotype dataset u sing SNAP Clade and Matrix ( Bo

wd

enet al.,

2008) to examine compatibility / incompatibility among all variabl esites; with any resultant incom patible sites removed from the dataset. Taji ma's D (Tajim a, 19891) and Fu and Li's D * and F' (Fu and I.i,1993) test statistics were calcul ated with DnaSP v. 3.5 3 (R o zas andRoz a s ,1999) to test for departur es from neutrality. Genetic diff er-entiation among geographic population s was analyzed using SNAPMap, Seqtomatrix, and Perm test (Hudson et al., 1992) imple-mented in SNAP Workbench ( Price and Carbone, 2005 ). Permtestis a non-paramet ric permutati on method based on Monte Carlosimulation s that estimates Hudson's test statistics ( KST, KS, and KT)under the null hypothesi s of no genetic differenti ation. For thispurpose, specimens in clade I w ere assigned to the 'Al askan', 'East-ern North American ', 'We stern North Am erican', or 'Mexican'groups based on th e geographic regions they occupied. In cladeII, specimens wer e assigned to the 'Alaskan', 'Eur opea n', 'Asian',and 'Pacific NW North American' group s, the latter representingclade II /A in Fig. 1. Significance was e valuated by performing1000 permutations. If we found evidence for geographic subdivi-sion, MD IV (Niel sen and Wak eley. 2001 ) was used to determin ewhether there was any evid ence of migration between pairs of subdivided popul ations. MDIV implem ents both likelihood andBayesian methods using Marko v chain Mon te Carlo (MCMC) coa-lescent simulation s to estimate the migration rate (M ), populationmean mutation r ate (Theta), and diverg ence time (T), Ages weremeasur ed in coalescent units of 2N , where N is the population size.This approach a ssumes that all populations descend ed from onepanmicti c population that ma yor may not have been followedby migration.


3/8

69 6 J. G e m l e t al ., / M o l e cul a r Ph y lo g eneti cs and Evolution 48 (2008) 694-701


4/8

J . Geml et al. , / M o lec u lar Phyl o gene t ics and Evo l u t i on 48 (2008) 694-701

3. Results

3.1. Phylogenetic analysis

The ITS, B -tubulin, LSU, EF 1 -a, and the combined datasets con-sisted of 691, 195, 724, 439, and 2049 characters, respectively,including gaps. There were 35, 21, 18, 35, and 109 parsimony-informative characters, respectively. The PHT indicated that thephylogenetic signals present in the different loci were not inconflict (P = 0.1146). The General-Time-Reversible model, with cal-culated proportion of invariable sites (l = 0.6106) and estimated a-parameter (= 0.9005) of y-distribution (GTR+I+G),was selected asthe best-fit evolutionary model. One of the equally parsimonioustrees is shown in Fig. 1. The Shimodaira-Hasegawa test revealed that there were no significant differences among the ML and MPphylograms (P = 0.164). The ML phylogra m (-lnL = 4962.5256) ispublished in the Supporting Information (Fig. S1).

Eight major lineages receiving high support (I-VIII, f ig. 1 )were detected within A. muscaria . All clades were supported byall loci except that clades I , I I ,and II I were not monophyleticin the B - tubulin phylogra m. Nonetheless, the B- tubulin MP treedid not show significant conflict with MP trees generated fromthe other loci. When the clades were under monophyletic con-straint, the most parsimonious B -t ub u lin trees were only 0-2steps longer than the unconstrained trees described earlier.Apparently, this lack of conflict was not due to low phylogeneticsignal in B - tub ul in. A permutation tail probability (PTP) test (Ar-chie, 1989; Faith and Cranston, 1989 revealed that the B- t u bulin locus contributes phylogenetic signal to the combined dataset,because tree length of the original B - tu b ul in phylogram was sig-nificantly shorter (P < 0.01) than the length of the trees gener-ated based on randomly permuted B - t u b u lin datasets. Asexpected, clades I, II, III, IV, V, and VI were strongly supportedin analyses of the combined dataset with 77%, 93%, 81%, 100%,


5/8

698 J . G eml e t a l . / Molecul a r Phyloge n e tics a n d E vo lu t ion 48 (2 008) 69 4-701

100%, and 88 % MPB valu es, respectively (Fig. 1). Groupings with-in these major clades were generally not signi f icantly supportedwith the exception of two sub clades: I/A and II/A with 9 3% and90% MPB values, respectively. Phylogene tic relationship s among

the major clades r emained uncle ar. as none of the groupingswere signif icantly supported, excep t that lineage V I II representsa sister group to the rest of the A. muscar ia complex. the latterbeing monophyletic with 9 3% MPB.


6/8

J . Geml et al. / Molecular Phylogeneticsand Evolu tion 48 (2008) 694 - 701 699

to 0.07006 in B -tubulin, from 0.01789 to 0.1 0 308 in EF 1- x , a nd from0.00324 to 0.01004 in LS U.The Fs t values measuring the geneticdifferentiation between clade pair s were always high, between0.82496 and 1.00000 in ITS, 0.20565 and 1.00000 in B -tubulin,0.63681 and 1.00000 in EF 1- x , and 0.836 86 and 1.00000 in LSU.The number of fixed diff erences were in the range of 7-26, 0-13,3-53 , and 2-3 in ITS, B -tubutin, EF l -x , and LSU, respectiv ely. T henumber s of polymorphis m s s hared among clades were low (0 or

1) in all cases, and a varyin g numb er of sites polymorphi c in oneclade but monom orphic in the other were observed.

3.3. Genetic differentiation among populations within phylogeneticspecies

In clade I, all four geographic group s showed significant P-val-ues for Hudson's t est, indicating genetic differentiation (Table 3 ).MDIV indicated no gene flow (M = 0) and non- zero divergence

3.2. Polymorphism and diver gence

Intraclad e nucleotid e polymorphism i s summarized in Tab le 2.Nucleotid e diversity values (rr ) were generally higher in the pro-tein-codin g genes than in the ribosomal DNAregions. Wh en com-paring the values across lineages, clade I con sistently showed highnucleotid e diversity with th e greatest valu es in three out of fourloci, followed by clade II , In general, clades III -V I, with more re-

stricted geographic di stribution s, and hence smaller sampl e sizes,tended to have lower nu cleotide diver sity values. Sim ilarly, c ladeI consist ently had the highest number of haplotypes for all lo ci, fol-lowed by clade II , despite the fact that the number of sequenceswas usually greater in clad e II,

Interclade divergenc e is summarized in Table 51 (SupportingInformation ). Pairwise divergence value s between the clades var-ied widely . The average number of nucl eotide substitution per site(Dxy) values ranged from 0.00789 to 0.04639 in ITS,from 0.008 89


7/8

In addition to the species-level ecoregional endemism, we foundevidence for additional phylogeographic structure at the popula-tion-level in clades I and II as well . In clade I, coalescent analyses re-vealed lack of migration and consid erable divergence amon g thefour major geographic groups, i. e., 'Alaskan', 'Eastern North A meri-.can', 'Western North American', and 'Mexican', The divergence withthe weakest support was that found between populations of Easternvs. Western North America. In this case, the results were only mar-ginally significant, and som e current migration could not be ruledout. Future sampling in Canada and the northern Great Plains shouldprovid e evidence as to whether Eastern and Western North Americarepresent one or more populations . In clade II, t here may be a low tointermediate level of migration among most of the geographicgroups spanning Eurasia and Alaska . On the other hand, the 'Pacif icNorthwest North American' group (clade II/A in Fig. 1 ) showsunequivocal evidence for both genetic and ecoregional isolationfrom the rest of clade II. This latter group has only been found inthe maritime rainforests from Washington state to southeasternAlaska along the Pacific coast of North America.

Most of the major clades within A. muscaria may have divergedas a result of multiple fragmentations and geographic isolation of the ancestral population s due to climatic changes in the late Ter-tiary and Quaternary. In tim e, they likely have evolved in situ and

adapted to the specific plant communities inhabiting different bio-geographic regions. Although we cannot estimate the times of divergence of the clades with certainty due to the lack of fos silsand the great variance in nucleotide substitution rates in fungi, itseems very likely that the major lineages separated well beforethe Pleistocene glacial cycles (Gem l et al., 2006 ). On the other hand,the intraspecific phylogeographic gr oups, shown in Figs. 52 and 53and in Table 3, may represent groups that became isolated fromthe rest of their species in the Pleistocene and survived one ormore glacial maxima in local refugia .

Our results also point out that whil e most lineages occupyeither entirely allopatric or micro-allopatric habitat s, we knowvery little about the biology and ecology of the different clades thatlive in true sympatry, i.e.. growing within the same fore st stand inclose proximity. General ecologi cal theory predicts that sympat ric

taxa will minimize intersp ecific competition by niche partitioning(Hutchinson, 1957), In the clades under consideration, this maymean host-specificity, inhabiting different soil horizons, and/orfavoring sites with different micro climates. Future ecological stud-ies should include representatives of various phylogenetic spe ciesto elucidate the ecology of these important mycorrhizal fungiand, in particular, the evolution of host preference.

The implications of our re sults are not restricted to the A. mus-c aria comple x, but are important for biodiversity studies and con-servation of ectomycorrhizal taxa in general, and for a ssessingthese fungi's resilience and future responses to climate chang e. Be-cause mycorrhizal fungi, including A. muscaria , play key ecologic alroles in the decomposition, mineralization, immobilization, andthe transfer of nutrients to plants. and because the genetic div er-sity, distribution, population structure and abundance of particul ar

species are likely to affect the ra tes and patterns of activity of th eircommunities, knowing and pres erving fungal diversity is crucialf or sustaining overall functional biodiv ersity.

700 J. Geml et al . / Molecular Ph ylogenetics a n d Evolution 48 (2008) 694- 7 01

time ( T: 1,3-2.0) betwe en group pair s. The only exception wasbetween the 'Eastern North Am erican' and 'Western NorthAmerican' groups, where M DIV indicated non -zero levels of genef low, although the hypoth esis of M = 0 could not be rejectedgiven the shape of the posterior probability distribution (Fig.S2, Supporting Information).

Geographic groups in clade II also had significant P-values forHudson' s test (Table 3). In most case s, however, MDIV indicatedlow to intermediate levels of gene flow, although the hypothesis

. of M = 0 could not be rejected. As expected, the 'Pacific NorthwestNorth American ' group (clade II/A in Fi g. 1) always showed signif-icant results for no gene flow (M = 0) and non-zero divergence time(T: 3-4) when compared to oth er groups (Table 3 and Fig. S3, Sup-porting Information) .

4. Discussion

In recent years, molecular tools have revealed several examplesof phylogenetic speciation within complexes that were previouslytreated as morphological species . The vast majority of these exam-ples dealt with phylogenetic species that were allopatric on a con-tinental scale. For example, morpholo gical species complexes of fungi from the Northern Hemi sphere have generally been shown

to comprise two major lineag es, a Eurasian and a North Ameri can(e.g .. Shen et al., 2002: Taylor et al., 200 6 and the references there -in). In most studied fungi, the allopatric phylogenetic clades inha-bit similar environments in diffe rent continents, which implies aphylogenetic structure that has arisen a s a result of the lack of intercontinental dispersal.

On the other hand, while two such major clades (I and II) arealso present in A. muscaria, we found several other divergent lin-eages that occupy different habitats or regions in the same conti-nent, sometimes in relatively close proximity. In these cases,spore dispersal is unlikely to be a limiting factor, and adaptationto diff erent ecological niches is a more parsimonious explanation.Such ecoregional diversification i s particularly obvious in clade IIIthat is sympatric with the Eurasian clade (11 )over the former's en-tire range. but is predominantly found above treeline (rnicro-allo-

patric ) in Alaska. Also, clad e IV almost exclusively inhabits themixed pine-oak-hickory fores ts of the southeastern US and hasbeen collected infrequently as far north a s Long Island. LineagesV. VI. and VI II are newly discov ered species and have only beenfound on Santa Cruz Island off th e coast of California,

These eight lineages app ear to be distinct phylogenetic specieswith no gene flow among th em, as indicated by the results of thePHT test and the tree topologies. Although. in theory , laboratorymating tests may provide information on whether or not in vitro interbreeding is possible among different phylogenetic groups ,we did not attempt crossing the r epresentatives of the distinctclades to test for increased reproductive isolation with increasedgenetic distance . Assessments of sporo carp and viable basidiospor eproduction of the 'hybrid' off springs, the ultimate measure of suc-cessful breeding, are virtually imp ossible to carry out in obligat e

ectomycorrhizae, such as A.m

us c ari a . Nonetheless, the concor-dance of gene trees fulfills th e phylogenetic species recognition cri-teria f ollowing T aylor e t al ., (2000) , and suggest reproductiv eisolation among the lineage s even when multiple lineages oc curin sympatry, e.g., in interior Ala sk a, the central Atlantic coastal re-gion of the US. and on Santa C ruz Island. The non-monophyly of some clades in the B -t ubulin dataset is lik ely due to incompletelineag e sorting at that locus. This is parti cularly plausible, becauseclades with the largest population sizes (occupying by far the larg-est geographic areas ) were the non-monophyletic ones, while thesmaller, regionally endemic clade s were monophyletic across allloci. Larger populations require longer periods of time to lose allancestral alleles.

Ac kn owledgments

This research was sponsored by th e Metageno mics of BorealForest Fungi project (NSF Grant No . 0333308 ) to D.L.T.and oth ersand the University of Alaska Presidential International Polar YearPostdoctoral Fellowship to J.G . The author s also thank Lisa Grub-isha for providing the Santa Cru z Island specimens, and Shaw nHouston and James Long at th e Biotechnology Computing Resear chGroup at UAF for the technic al support in phylogenetic analys es.


8/8

J . G e ml e t al . / Mole cular P hyl oge n e ti c s an d E v o lution 48 ( 2 0 08) 6 9 4 - 7 0 1

Finally, we are gratefu l to the anonymou s reviewers for th e helpfulcomments .

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi: 1 0.10 1 6/ j.ympev.2008.04.029.

References

Archle.j.w. , 1989. A randomi zation test f or phylo genetic information in systematicdata. Syst. Zool. 38, 239-252.

Avise , J.e .. 2000. Phylogeography: The History and Formation of Species. HarvardUniversity Press, Cambridge.

Ay lor. D.L.. Price. E.W., Carbone, I.. 2006. SNAP: combine and map modulesf or multilocu s population genet ic analysis. Bioinformatics 22, 1399-1401.

Bagley. S.J.. Orlo vich, D.A.,2004. Genet s ize and distr ibut ion o f Amanita mu scar ia in asuburban park, Dunedin, New Z ealand. N Z J. Bot. 42,939-947.

Bowden, L.C ., Price, E.W., Carbone, I. , 2008. SNAP Clade and Matrix. V ersion 2.Available from: , Department of Plant Path ology, North Carolina State University.

Faith, D.P., Cranston, P.S., 1991 . Could a cladogram this short have arisen by chancealone? On permutation test s f or cladistic structure. Cladist ics 7. 1-28.

Felsenstein. J .. 1985. Confiden ce limits on phylogenies: an approach USing thebootstrap. Evolution 39, 783-791.

Fu. Y.X.. Li, W.H .. 1993. Maximum lik elihood estimation of population parameters.Genetics 134, 1261-1270.

Geml, J .. Davis. D.D. . Geiser, D.M.. 2005. Phylogenetic analyses reveal deeplydivergent species lineage s in the genus S pho e robo lu s. Mol. Phylogener. Evol. 35.313-322.

Ceml, J.. Laursen, G.A.. O'Neill, K ., Nusbaum, H.C . , Taylor, D.L.. 200 6. Beringianorigins and cryptic speci ation events in the fl y agaric (Amanita muscari a). Mol.Ecol. 15, 225-239.

Hudson, R.R.. Boos, D.D., Kaplan, N .L., 1992. A s ta ti st ical t es t for d etectinggeographic subdivision. Mol. Bi ol . Evol. 9, 138-151.

Hutchinson, G.E., 1957. Concludin g remark s. Cold Spring Harbor Symp. Qu an t. Biol .22.415-427.

jenk ins. D.T., 1986. Amanita of North Americ a, Mad River Press, Eureka . johnson. N.C .. Gehring. C .A., 2007. Mycorrhizas: symbiotic mediato rs of rhizosphere

and ecosystem processes. In: Cardon, Z.G., W hitbeck . J.L. (Eds.). The

Rhizosphere: An Ecological Perspective. Academic Press, Amsterdam. pp. 73-100.

Nei. M .. Kumar,S., 200 0. Molecular Evolution and Phylogenetics. Oxford UniversityPress.

Nei, M., Li, W.H .. 1979. Marhern arical model f or st udying genetic variati on in tenusof restriction endonucleas es. Proc, Natl. Acad. Sci. USA 76. 5269 -5273.

Nielsen, R., Wakeley, j., 2001. Distin guishin g migration from isolation: a M arkovchain Monte Carlo appr oach. Genetics 158, 885-896.

Oda, T.. Tanaka. c .. Tsuda . M .. 2004. Molecular phylogeny and bio geography of thewidely distributed Aman ita species. A m usccaria and A. panther ina. Mycol. Res.

108.885-896. 'Posada, D.. Crandall. K.A.. 1998. Modeltest: testing the model of DNA substitution.

Bioinforma tics 14.817-818.Price, E.W. . Carbone, I.. 2005. SNAP: work bench management tool f or evolutionary

population genetic analysis. Bioinform atics 21. 402-404.Read, D. j., Perez-Moreno.]., 2003. Mycorrhizas and nutrient cycling in ecosysrpms-

a Journey towards relevance? New Phytol. 157.475-492.Rozas. J.. Rozas. R.. 1999. DnaSP version 3: an integrated program f o r molecu la r

population genetic s and molecular evolution analysis. Bioinformatics 15. 174-175.

Shen, Q , Geiser. D.M., Roys e. D.J.. 2002. Molecular phylngenerir an alysis of Gri fola f r ondosa (m aitake) rev eals a species .parti tion separat ing e astern NorthAmerican and Asian isol ates, Mycologia 94,472-482.

Shimodaira, H., Hasegawa, M., 1 999. Multiple comparisons or log - lik elihood withapplications to phylogenetic inf erence. Mol. Biol . Evol. 16. 1114-111 6.

Swof f ord. D.L., 2002. PAUP . Phylogenetic Analysis Using Parsimon y (' And OtherMethods), 4.0b4a. Sinauer Associ ates. Sunderland. MA.

Ta jima, F.. 1989. Stati stical method lor testing the neutral mutation hypothesis byDNA polymorphi sm. Genetics 123 . 585-595. .

Taylor, j.W .. Jacobson. D.J.. Kroken, S.. I(asuga. T., Geiser, D.M .. Hibbett, 0.5 .. Fisher.

M.C ., 2000. Phylogenetic species recognition and species concepts in f ungi.Fungal Genet. Biol . 31, 21-32.

Tay lor. J.W., Turner , E., To wnsend. J .P., Dettman, .J .R.,Jacobson. D .. 2006. Eu k ar yoticmicrobes, species recognition a nd the geographic limit s of specie s: examplesf rom the kingdom Fungi. Philos. Trans. R. Soc. Lond. Ser. B 361. 1 947 -1 96 3.

Thompson. J.D., Gibson. T.].. Pl ewniak , F.. J eanmougin, F.. Higgins. D ,C .. 1997. TheCLUSTAL X window s interface: f lexible strategies for multiple sequencealignment aided by quality an alysis tools. Nucleic Acids Res. 25. 4876 -488L

Trappe,J .M., 1987. Phylo genetic and ecological aspects of mycorrophy in theangiosperms from an evolutionary standpoint . In: Saf ir , G.R. (Ed ),

Ecophysiology of VA My corrhizal Plants. CRC Press. Boca Raton, pp. 5-25.Zwick l, D.J .: 2006. Genetic algorithm approaches f or the phylogenetic analysis of

large biological sequence datasets under the maximum l ik elihood criterion.Ph.D. Dissertation. The University of Texas at Austin.

jozsef geml et al- evidence for strong inter- and intracontinental phylogeographic structure in...

Documents