b
st
in
DE
TOMOHISA YUKAWA,? YUKI OGURA-TSUJITA? and TOSHIMASA HASHIMOTO?
*Odum School of Ecology, University of Georgia, 140 E. Green St., GA 30602, USA, ?Plant Ecology Lab, Smithsonian
Science, 1-1
Keywords: Ceratobasidiaceae, Ceratobasidium, mycorrhiza, parasitic fungi, specificity
Importantly, such analyses assume that the breadth of
The evolutionary history of broad interactions may
best be approached by quantifying specificity, or host
Correspondence: Richard P. Shefferson, Fax: +1 706 542 4819;
E-mail: dormancy@gmail.com
Molecular Ecology (2010) 19, 3008?3017 doi: 10.1111/j.1365-294X.2010.04693.xPhylogenetic patterns in the evolution of biological
interactions are often studied in relation to whether
they suggest cospeciation. In the simplest case, cospeci-
ation is observed as cophylogeny between suites of
interacting taxa, and can involve either a common evo-
lutionary response to external factors, or a reciprocal
the interaction is only one species. For example, mam-
mals may exhibit one-species-to-one-species relation-
ships with body lice, leading to phylogenetic patterns
in hosts and parasites (Hafner et al. 2003). Broad inter-
actions cannot be studied readily from a cophylogenetic
standpoint using contemporary methods, and so they
have rarely been studied phylogenetically.Received 22 February 2010; revision received 28 April 2010; accepted 29 April 2010
Introduction
evolutionary response (Brooks & McLennan 1991).Abstract
Host breadth is often assumed to have no evolutionary significance in broad interactions
because of the lack of cophylogenetic patterns between interacting species. Nonetheless,
the breadth and suite of hosts utilized by one species may have adaptive value,
particularly if it underlies a common ecological niche among hosts. Here, we present a
preliminary assessment of the evolution of mycorrhizal specificity in 12 closely related
orchid species (genera Goodyera and Hetaeria) using DNA-based methods. We mapped
specificity onto a plant phylogeny that we estimated to infer the evolutionary history of
the mycorrhiza from the plant perspective, and hypothesized that phylogeny would
explain a significant portion of the variance in specificity of plants on their host fungi.
Sampled plants overwhelmingly associated with genus Ceratobasidium, but also
occasionally with some ascomycetes. Ancestral mycorrhizal specificity was narrow in
the orchids, and broadened rarely as Goodyera speciated. Statistical tests of phylogenetic
inertia suggested some support for specificity varying with increasing phylogenetic
distance, though only when the phylogenetic distance between suites of fungi interacting
with each plant taxon were taken into account. These patterns suggest a role for
phylogenetic conservatism in maintaining suits of fungal hosts among plants. We stress
the evolutionary importance of host breadth in these organisms, and suggest that even
generalists are likely to be constrained evolutionarily to maintaining associations with
their symbionts.tal Research Center, PO Box 28, Edgewater, MD 21037, USA, ?Tsukuba Botanical Garden, National Museum of
Amakubo 4, Tsukuba 305-0005, Japan, ?Nippon Steel Kankyo Engineering Corp., Kisarazu, Chiba 292-0, JapanEnvironmenEvolution of host breadth in
mycorrhizal specificity in Ea
American rattlesnake planta
their fungal hosts
RICHARD P. SHEFFERSON,* CHARLES C. COWroad interactions:
Asian and North
s (Goodyera spp.) and
N,* MELISSA K. MCCORMICK,? 2010 Blackwell Publishing Ltd
breadth, in interacting clades. Phylogenetic approaches that phylogenetic measures of specificity are far more
could readily access geographically, and that were pos-
MYCORRHIZAL HO ST BREADTH EVOLUTI ON 3009are often utilized to understand host specialization
(Janz & Nylin 1998; Anderson 2006), and have recently
been adapted to the study of communities (Lozupone
et al. 2006; Hardy & Senterre 2007; Pommier et al.
2009). In broad interactions, the identities of interactors
as well as the breadth of the interaction may change,
resulting in the evolution of both parameters (Weiblen
et al. 2006; Shefferson et al. 2007), the latter being quan-
titative rather than nominal. Such patterns may be influ-
enced by different geographic and ecological
distributions, resulting in geographic mosaics of inter-
acting suites of species (Thompson 2009). When mea-
sured quantitatively, specificity may be mapped onto
the phylogenies of all interactors and common patterns
of evolution in specificity may be assessed.
This approach offers great promise in evolutionary
studies of the mycorrhiza, a symbiosis based on nutri-
ent exchange between terrestrial plants and soil fungi
involving a polyphyletic group of taxa in both king-
doms (Smith & Read 2008), as well as in other horizon-
tally transmitted microbial symbioses. The orchid
mycorrhiza is a form of this interaction found in family
Orchidaceae, the most species-rich family of flowering
plants, and is unique because of its morphology and
because it is thought to be typically parasitic?the plant
obtains nutrients from the fungus, but little evidence
exists of the reverse (Rasmussen 1995). The fungi form-
ing these associations typically make their livings in
other ways?many are saprotrophs living off organic
matter in the soil, and others are typically ectomycorrhi-
zal with other plants while others are plant parasites
(Roberts 1999; Yamato et al. 2005). Many orchid species
have evolved into purely parasitic, non-photosynthetic
forms, living entirely off of fungal carbon (Taylor &
Bruns 1997; Bidartondo 2005; Ogura-Tsujita et al. 2009),
although recent evidence suggests that some Goodyera
spp. may act more mutualistically (Cameron et al. 2008;
Hynson et al. 2009). This purported family wide para-
sitism has generated much interest in the mycorrhizal
specificity of orchids. DNA-based analyses of orchid
mycorrhizae have revealed that orchids can be special-
ists or generalists, and all shades in between (Taylor &
Bruns 1997; Wei? et al. 2004; Shefferson et al. 2005;
Abadie et al. 2006). However, although the specificity of
orchids for their fungi has been studied for many dec-
ades now, the macroevolutionary history of specificity
has been ignored. The reasons are twofold: first, the
fungi are unlikely to have evolved in response to the
orchids because of the likely rarity of the interaction
from the fungal standpoint, and second, both basal and
derived orchids typically associate with at least three
families of fungi, Ceratobasidiaceae, Sebacinaceae, and
Tulasnellaceae (Yukawa et al. 2009). However, we argue 2010 Blackwell Publishing Ltdsible to access given conservation concern for rare orch-
ids, and given the often difficult international politics
governing work with rare and endangered plants. Our
sampling thus reflects a balance between a need for
study material, the need to preserve extant populations
and species, and the difficulty of sampling the wide
geographic range and taxonomic diversity of the group.
Goodyera foliosa (Lindl.) Benth. is found throughout
Japan extending to Okinawa, and on the Korean Penin-
sula. G. hachijoensis Yatabe is found primarily in central
Japan. G. macrantha Maxim. is found from central to
southern Japan and on the Korean Peninsula. G. oblongifolia
is found in western North America. G. pendula Maxim.
is found in northern and central Japan. G. procera
(Ker-Gawler) Hook. is found in southern Japan, China,
India, and Malaysia. G. repens (L.) R.Br. is found
throughout the Northern Hemisphere, even extendinginformative than simple counts of host families (Taylor
et al. 2004; Shefferson et al. 2007).
Here, we assess the evolution of mycorrhizal specific-
ity in the rattlesnake plantain orchids (genus Goodyera,
with genus Hetaeria used as an outgroup). This genus
presents an interesting case study in the ecology of the
orchid mycorrhiza because one species, Goodyera repens,
has long been a subject of experimental research into
the nature of the orchid mycorrhiza (Downie 1943; Had-
ley & Purves 1974; Alexander et al. 1984; Cameron et al.
2006). First, we identify the fungi mycorrhizal with
these orchids. We then estimate mycorrhizal specificity
as the mean pairwise phylogenetic distance among
suites of fungi interacting with each orchid. Next, we
map this quantity onto a phylogeny of the genus Goody-
era, and assess whether specificity is partially deter-
mined by phylogeny.
Materials and methods
Study system
The genus Goodyera is a member of family Orchidaceae
that includes approximately 80?100 species distributed
primarily throughout the Northern Hemisphere, with
highest diversity in tropical eastern Asia (Satake et al.
1985; Ormerod & Cribb 2003). These species are rhizo-
matous, terrestrial perennials, but can grow onto the
bark of trees and rock faces in parts of East Asia. Ten
Goodyera and two species of the closely related genus
Hetaeria were sampled for this study (Table 1). Like
most orchids, these species are typically rare, with
small, disparate populations even in species with wide-
spread distributions. Our choices in study species repre-
sents the suite of species within the genus that we
Table 1 List of surveyed Goodyera species, regions and locales sampled, years harvested, and numbers of populations and individu-
e nu
d fu
ido
ushu
rego
rnia
G. pendula Japan Kochi, Shikoku
ushu
ginia
ki
42 94
3010 R. P . SH EFFERSON ET AL.G. procera Japan Amami Oshima, Ky
G. repens* USA Nelson County, Vir
G. schlechtendaliana Japan Chiba prefecture
Izu Archipelago
Mt. Tsukuba, Ibara
G. tesselata USA Massachusetts
G. velutina Japan Izu Archipelago
Hetaeria cristata Japan Izu Archipelago
H. agyokuna Japan Izu Archipelago
Totalsals sampled at each locale. Numbers in parentheses refer to th
mtLSU primers. Asterisks (*) indicate species to which we adde
for low sampling in our study
Species Country Region
G. foliosa var. laevis Japan Asahikawa, Hokka
Izu Archipelago
Kyoto City
var. maximowicziana Chiba Prefecture
Tochigi prefecture
G. hachijoensis var. hachijoensis Japan Izu Archipelago
var. izuhsmensis Izu Archipelago
var. matsumurana Amami Oshima, Ky
G. macrantha Japan Tochigi prefecture
G. oblongifolia USA Columbia Gorge, O
Klamath NF, Califo
Priest Lake, Idahoto northern tropical Africa. G. schlechtendaliana Reichb.
fil. is found throughout Japan, the Korean Peninsula,
and in eastern China. G. tesselata is found in eastern
North America. G. velutina Maxim. is found in southern
Japan and the Korean Peninsula. Hetaeria cristata Blume
is found in central and southern Japan, Indonesia, and
Taiwan. Hetaeria agyokuna (Fukuyama) Nackejima is in
southern Japan and in Taiwan.
Field methods
Sampling occurred from spring 2002 until summer
2007. We obtained locations of target populations from
local experts, landowners, and land managers, and vis-
ited sites throughout Japan and the USA. At each site,
we chose plants representing a range of life stages, from
small, vegetative sprouts to large, flowering individuals.
Between two and six roots were sampled per plant,
including 408 root samples from 94 individuals in 42
populations (Table 1). The total number of plants sam-
pled was kept at no more than 10% of each sampled
population due to conservation concern. All root sam-
ples were kept on ice in the field, and were transported
to the laboratory for microscopy and DNA extraction
within four days of field sampling.mber of plants yielding PCR product with fungal nucLSU or
ngal haplotype data from other studies in order to compensate
Year sampled No. Pops sampled No. plants sampled
2005 1 1
2005 4 12
2005 1 1
2005 2 11
2005 1 14
2005?2006 6 10
2005?2006 1 2
2005 1 1
2005 1 2
n 2003 2 2
2008 2 6
1998 1 1
2005 1 1
2005 5 8
2001 1 2
2005 1 1
2005 1 1
2005 1 1
2002 3 5
2005 3 8
2005 2 2
2005 1 2Laboratory methods
All roots were surface-sterilized using 20% bleach solu-
tion (Taylor & Bruns 1997). Light microscopy was used
to identify mycorrhizal samples, and four to five sam-
ples of roughly 0.5?1.0 cm in length were chosen per
plant. Characterization of mycorrhizal fungi involved:
(i) extraction of fungal and plant DNA from mycorrhi-
zal plant tissue; (ii) amplification of fungal genomic
regions useful in determining fungal identity; (iii)
assessment of basic patterns in fungal diversity within
roots, individuals, populations, and species; (iv) DNA
sequencing of unique strains; and (v) phylogenetic anal-
ysis for identification of mycorrhizal fungi and assess-
ment of specificity. Details of laboratory methods are
provided in Shefferson et al. (2005, 2007). We included
root tissue samples not colonized by mycorrhizal fungi
to provide negative controls. We tested each sample
with each of the following sets of primers targeting the
internal transcribed spacer (ITS): ITS1F-ITS4 (White
et al. 1990; Gardes & Bruns 1993), ITS1F-cNL2F (White
et al. 1990), ITS1-ITS4B (Gardes & Bruns 1993), and
ITS1OF-ITS4OF (Taylor & McCormick 2008). Some sam-
ples were also tested with ITS1?ITS4Tul (Taylor 1997),
although this was limited to samples that failed to
 2010 Blackwell Publishing Ltd
amplify via other primers and a few others, and did tered fungi with strong BLAST support were not phylo-
each plant taxon as a function of the plant phylogenetic
MYCORRHIZAL HO ST BREADTH EVOLUTI ON 3011Phylogenetic analysis
Sequences were edited in ChromasPro 1.5 for Windows
(Technelysium Pty. Ltd, Tewantin, Queensland, Austra-
lia) and analyzed with BLAST (Altschul et al.
1997) against the NCBI sequence database (National
Center for Biotechnology Information, GenBank:
http://www.ncbi.nlm.nih.gov) to detect similar sequ-
ences of known phylogenetic placement. We then con-
firmed BLAST designation via phylogenetic analysis in
a fungal ITS alignment representing the major groups
of basidiomycetes and ascomycetes (Taylor & Bruns
1997). Further analyses involved adding sequences to
alignments representing narrower phylogenetic breadth,
with reference sequences imported from GenBank.
Sequences were aligned using ClustalX 2.0.11 for Win-
dows XP (Thompson et al. 1997; Larkin et al. 2007). The
appropriate model of DNA evolution was determined
using FindModel (http://www.hiv.lanl.gov/content/
sequence/findmodel/findmodel.html; Posada & Cran-
dall 1998). Phylogenetic analysis involved maximum
likelihood searches in PhyML for Windows XP (Guin-
don & Gascuel 2003; Guindon et al. 2005; Ansimova &
Gascuel 2006), using the best model of DNA evolution
as chosen by FindModel. Branch support was estimated
via 1000 maximum likelihood replicates. Rarely encoun-not yield any PCR product not reported for other pri-
mer sets. PCR involved 35 cycles with an annealing
temperature of 55 C using an Eppendorf Mastercycler
epGradient S Thermocycler (Eppendorf AG, Hamburg,
Germany), and all species yielded fungal PCR product
except G. macrantha. Although we attempted to amplify
the mitochondrial large subunit (mtLSU) using primers
ML5?ML6 (White et al. 1990), these PCRs were unsuc-
cessful. Representative samples were chosen for each
plant via RFLP analysis of ITS PCR product using the
restriction enzymes DdeI, HinfI, and either MboI or
NlaIII (Gardes & Bruns 1996). The ITS and rbcL regions
from each plant species were also amplified via the
primers ITS1P?ITS4 (White et al. 1990; Taylor & Bruns
1997) and rbcL1F?rbcL1367R (Kores et al. 1997), respec-
tively. PCR cloning was performed with Stratagene XL-
10 Gold Ultracompetent cells (Stratagene Inc., La Jolla,
CA, USA) and the pDrive cloning vector (Qiagen Inc.)
when RFLP analysis suggested the presence of multiple
fungi. Clones representative of the major RFLP-types
were chosen for sequencing. We cycle sequenced
unique PCR samples with BigDye v. 3.1 chemistry
(Applied Biosystems Inc., Foster City, CA, USA), and
electrophoresed each sample on an ABI 3730 Genetic
Analyzer (Applied Biosystems Inc.) at the DNA Synthe-
sis and Sequencing Facility (University of Georgia). 2010 Blackwell Publishing Ltddistance, with phylogenetic distances estimated in Arle-
quin 3.11 for Windows (Excoffier et al. 2005). The latter
analysis differed from the former in that the former
tested whether the quantitative value of specificity itself
varied with plant phylogenetic distance, while the latter
tested the degree to which the phylogenetic distance
between the suites of hosts associating with each plant
taxon varied with plant phylogenetic distance.genetically analyzed, though they are presented with
BLAST results in this paper. Plant ITS and rbcL
sequences were also analyzed as above, with phyloge-
netic analysis proceeding on both loci together.
Sequences generated in this study have been deposited
in GenBank under accessions HM140988?HM141077,
and HM151401?HM151402. Phylogenetic trees and
alignments have been deposited on TreeBASE.
Analysis of specificity
Per Taylor et al. (2004), we quantified specificity as the
mean pairwise phylogenetic distance, p (Nei & Tajima
1981), among fungal haplotypes corresponding to
unique species or major clades identified in phyloge-
netic analysis, using Arlequin 3.11 for Windows (Excof-
fier et al. 2005). All fungal haplotypes found within
each orchid taxon were pooled to estimate p, and we
did not treat haplotypes originating from the same sam-
ple differently than we treated haplotypes from other
samples within the same taxon. We used only the fun-
gal 5.8S region in order to include the broadest assem-
blage of fungi for each plant species, including
ascomycetes and basidiomycetes. We added p for
G. pubescens from fungal data taken from McCormick
et al. (2004). We also added a Ceratobasidium cornigerum
haplotype to G. repens, based on previous reports
suggesting it to be a common symbiont of that orchid
species (Alexander & Hadley 1985; Cameron et al.
2006). We mapped these quantities via least squares
onto the plant phylogeny using the ape package in R
(Paradis 2006; R Development Core Team 2007). We ran
this analysis twice, with p for G. macrantha equalling 1
(narrow specificity) or 15 (broad specificity) to compen-
sate for the lack of successful PCR from this species,
but found no difference in evolutionary patterns so only
present the former result. We assessed whether plant
phylogeny determines mycorrhizal specificity in two
ways. First, we tested for phylogenetic autocorrelation
in specificity using Geary?s randomization approach to
Moran?s autocorrelation index using the ape package in
R (Gittleman & Kot 1990; Thioulouse et al. 1995; Paradis
2006). Second, we regressed the mean pairwise phyloge-
netic distance between suites of fungi associating with
The number of sampled plants per population and
input in Arlequin 3.11 for Windows (Excoffier et al.
potentially identified as C. albasitensis, and a fungus
share of the variation in p (populations: F = 0.421,
3012 R. P . SH EFFERSON ET AL.2005). Specificity for each replicate in each dataset was
estimated as before, and we estimated the mean p and
associated standard error for each dataset. We then
assessed the minimum number of individuals needed
to accurately assess specificity in that taxon as the point
at which mean p no longer increased with increasing
number of sampled individuals.
Results
Fungal identification
FindModel suggested that the most appropriate model
of DNA evolution in our phylogeny of the largest clade
of Goodyera mycorrhizal fungi (Ceratobasidiaceae) was
the HKY + C model. In our phylogeny of the next most
common fungal associates, within the Ascomycota, it
was the GTR + C model. FindModel further suggested
that the most appropriate model of DNA evolution in
our phylogeny of Goodyera and Hetaeria species was the
HKY model. Bootstrap support was low deep within
our main phylogenies, but fairly strong closer to the
tips (Figs 1 and 3).
Goodyera species associated overwhelmingly with spe-
cies in the fungal family Ceratobasidiaceae, but also
with occasional fungi in other families (Table S1, Sup-
porting information). G. foliosa associated with Ceratoba-
sidium papillatum or a close relative, as well as unnamed
Ceratobasidium taxa sister to C. angustisporum (Fig. 1). G.
hachijoensis associated with C. cornigerum, a funguspopulations per taxon varied in this study (Table 1).
We first assessed whether these inequalities may have
affected our results via regression analyses of mean fun-
gal p per orchid taxon as a function of the number of
populations per species and mean individuals per pop-
ulation. All analyses were conducted as general linear
models in PASW Statistics 17.0 for Windows (SPSS Inc.,
Chicago, IL, USA). Because a number of taxa could only
be sampled in low quantities (e.g., < 3 individuals), we
also tested whether this low sampling may have biased
our specificity towards narrow host breadth. To do so,
we characterized the frequencies of the number of fun-
gal haplotypes found per individual of the most widely
sampled taxon that exhibited wide specificity, G. foliosa
var. maximowicziana. We then created bootstrapped
datasets representing random draws from the fungal
haplotypes found in this taxon, with the number of
haplotypes per individual chosen according to the fre-
quencies of fungal haplotypes per individual in this
taxon. Each dataset corresponding to each number of
sampled individuals included 100 replicates. This boot-
strapped dataset was created in C++ and was used as3,3
P = 0.752; plants per population: F6,3 = 1.759, P = 0.344).
Further, bootstrap analysis of the G. foliosa var. maxi-
mowicziana dataset suggested that samples of two indi-
viduals were the minimum needed to maximize
estimated specificity in this broadly associating orchid
(Fig. 2), most likely due to the tendency for this species
to be colonized by multiple mycorrhizal fungi (mean
number of mycorrhizal fungi per individual = 1.85 ?
0.15 haplotypes).
Assessment of the phylogenetic contribution to
mycorrhizal specificity was equivocal but suggestive.
Geary?s randomization test yielded a low Moran?s I,
and was not statistically significant (I = )0.094,
P = 0.318). Mean pairwise phylogenetic distance
between the suites of fungi associating with each plant
species was significantly determined by phylogenetic
distance among plant taxa (F68,35 = 791.1, P < 0.001).
The ancestral condition appears to have been narrow
specificity (Fig. 3). A broadening of host breadth
occurred after the speciation of G. oblongifolia, with
extremely broad specificity observed in G. procera andnear C. angustisporum (Fig. 1). G. oblongifoilia associated
with fungi near C. albasitensis, C. bicorne, and C. angusti-
sporum (Fig. 1). G. pendula associated with C. cornigerum
and fungi near C. angustisporum (Fig. 1). G. procera asso-
ciated with C. cornigerum. G. repens, G. tesselata, and G.
velutina associated with fungi falling near C. angustispo-
rum. G. schlechtendaliana associated with these same
groups, as well as fungi falling near C. papillatum and
C. oryzae-sativae (Fig. 1). Of these three, G. repens? asso-
ciate was surprising given its occurrence away from C.
cornigerum, which was previously noted to be its main
symbiont. Hetaeria cristata and H. agyokuna both associ-
ated only with Ceratobasidium cornigerum (Fig. 1). Addi-
tionally, G. foliosa, G. hachijoensis, G. procera, and G.
velutina had sporadic associations with potentially
mycorrhizal ascomycetous endophytes falling near Phi-
alophora finlandia and Chalara dualis (Fig. S1, Supporting
information; Table S2, Supporting information). G. velu-
tina also rarely associated with potentially ectomycor-
rhizal associates falling into genera Russula and
Clavulina (Table S2, Supporting information).
Mycorrhizal specificity
Assessed as the mean pairwise phylogenetic distance
among all fungal haplotypes, including Ceratobasida-
ceae, ascomycetes, and all other potentially mycorrhizal
fungi, specificity did not vary with sampling effort. A
general linear model of p as a function of the number
of populations and plants per population sampled sug-
gested that both factors did not account for a significant 2010 Blackwell Publishing Ltd
Fig. 1 Phylogenetic placement of fungal taxa in the family Ceratoba
mined with sequences from the fungal internal transcribed spacer (I
Bank. Analysis was via maximum likelihood in PHYML for Window
1000 bootstrap replicates. Phylogeny is midpoint-rooted, due to the la
Ceratobasidiaceae.
9
8
7
6
5
2 4
Number of sampled individuals
8 106
M
ea
n 
?
Fig. 2 Assessment of bias in specificity estimates as a function
of the number of Goodyera individuals sampled. Data from
sampled G. foliosa var. maximowicziana were used. We boot-
strapped ?individuals?? of this taxon using a probabilistic
assessment of the number of fungal haplotypes per individual
(0.30 probability of one fungal haplotype, 0.55 of two, and 0.15
of three), and random draws with replacement from the pool
of all sampled fungal haplotypes discovered in this taxon. One
hundred such replicates were bootstrapped for each number of
sampled individuals, and the mean p and associated standard
error for each group of 100 replicates was estimated in Arle-
quin 3.11 for Windows (Excoffier et al. 2005).
MYCORRHIZAL HO ST BREADTH EVOLUTI ON 3013
 2010 Blackwell Publishing Ltdsidiaceae mycorrhizal with Goodyera species. Phylogeny deter-the G. foliosa clade also evolving relative generalization.
Specificity then renarrowed in the G. hachijoensis group
(Fig. 3).
Discussion
Mycorrhizal specificity appears correlated with phylog-
eny in this system, and so macroevolutionary history is
an important consideration determining the observed
pairing of plant and fungus in the orchid mycorrhiza.
This study is among the first to suggest that suites of
symbiotic hosts evolve to differ more with increasing
phylogenetic distance. These patterns indicate a role for
phylogenetic conservatism in determining which fungal
species form mycorrhizas with plants, as it does in
determining food webs (Cattin et al. 2004). Theoreti-
cally, this may stem from the fact that symbiotic hosts
often form a kind of habitat or niche for their partner
taxa, and phylogenetic conservatism is typically thought
of in the determination of niche (Wiens & Graham
2005; Lovette & Hochachka 2006).
Quantitatively, the mycorrhizal specificity we
observed in the plant hosts appears typical of orchid
mycorrhizal associations. For example, previous assess-
TS) region, and includes reference sequences from NCBI Gen-
s (Guindon & Gascuel 2003; Guindon et al. 2005), and involved
ck of agreement on the evolution of the members of the family
wing
ng fu
e no
in P
eny i
ensis
in R
pled.
3014 R. P . SH EFFERSON ET AL.ments of mycorrhizal specificity from the plant stand-
point in terrestrial and tropical orchid systems have
identified typically narrow suites of hosts, with some
species exhibiting fairly broad associations (Otero et al.
2002, 2004; McCormick et al. 2004). Ecological determi-
Fig. 3 Phylogeny of Goodyera and Hetaeria species sampled, sho
quantified as the mean pairwise phylogenetic distance, p, amo
Values at nodes include the estimated specificity ? 1 SE above th
isks. Phylogenetic analysis was via maximum likelihood analysis
don et al. 2005), and involved 1000 bootstrap replicates. Phylog
No rbcL sequences were obtained for G. hachojoensis var. izuhsm
acter evolution was inferred via least squares in the ape package
are followed by (S) if only one individual of that taxon was samnants have rarely explained these trends, although
sometimes host shifts occur with ontogeny or stress
(McCormick et al. 2006), and tropical species may be
more generalist than temperate species (Roy et al. 2009).
A phylogenetic assessment of specificity in another
orchid system, the lady?s slipper genus Cypripedium,
revealed that phylogeny is an important determinant of
mycorrhizal specificity in plants (Shefferson et al. 2007).
These orchids typically exhibit narrow host breadth,
with expansions known to have evolved only twice in
the genus, each time leading to one species (Shefferson
et al. 2007). Such patterns were repeated here: a narrow
interaction with fungi within the genus Ceratobasidium
appears to be ancestral in this group, supporting other
evidence that interactions with this fungal genus may
be as old as the orchid mycorrhiza itself (Yukawa et al.
2009). However, low bootstrap support in our phyloge-
nies reinforces the need for further work on this system
in order to strengthen inference about ancestral states.
The distribution and life history of genus Goodyera
may be determined in part by the combined ecology of
its mycorrhizal fungal hosts. Goodyera forms mycorrhi-
zae overwhelmingly with the basidiomycete genus
Ceratobasidium and occasionally associates with otherfungi, including other basidiomycetes such as Clavulina
sp. and Russula sp., and some ascomycetes, such as the
ectomycorrhizal Phialophora finlandia. Although Tulasnel-
la spp. have been noted to form mycorrhizas with this
orchid group, this association was phylogenetically rare
the evolution of mycorrhizal specificity. Here, specificity was
ngal 5.8S haplotypes found mycorrhizal with sampled plants.
de. Clades with bootstrap support ? 50% are noted with aster-
HYML for Windows Windows (Guindon & Gascuel 2003; Guin-
s rooted with Hetaeria cristata and H. agyokuna as the outgroup.
and var. matsumurana, and for G. pendula and G. tesselata. Char-
(Paradis 2006; R Development Core Team 2007). Taxon namesin our dataset, occurring commonly only in G. pubescens
(McCormick et al. 2004). An expanded sampling may
find more.
Fungi in the genus Ceratobasidium are basal hymeno-
mycetes that live saprotrophically in the environment,
parasitize plant tissues, and sometimes form ectomy-
corrhizae (Downie 1943; Roberts 1999; Yagame et al.
2008). They are often economically and ecologically
important pathogenic fungi. Ceratobasidium cornigerum is
a major pathogen of grasses and cereal crops (Roberts
1999), although it is mycorrhizal with some other orchids
(Otero et al. 2002). C. anceps parasitizes fern leaves (Gre-
gor 1935), C. bicorne is a root parasite of Pinus spp., and
C. calosporum is a free-living saprotroph (Roberts 1999).
Some previous studies have suggested that Goodyera
may commonly parasitize carbon resources from its
mycorrhiza (Hadley & Purves 1974; Alexander & Hadley
1985), although more recently carbon donation has also
been observed (Cameron et al. 2006, 2008). Although
ectomycorrhizal fungi are typically thought to be better
carbon donors for parasitic plants than saprotrophic
fungi (Bruns et al. 2001; but see Ogura-Tsujita et al.
2009), the potentially pathogenic nature of Ceratobasidium
species likely makes them excellent sources of energy for
 2010 Blackwell Publishing Ltd
plants that can tap into their nutrient flows. If orchids
2008; Hynson et al. 2009).
and quantitative framework for assessing these patterns,
Alexander C, Alexander IJ, Hadley G (1984) Phosphate uptake
MYCORRHIZAL HO ST BREADTH EVOLUTI ON 3015particularly in situations where species designations are
unclear [e.g. certain other fungal groups, including fam-
ily Tulasnellaceae, per Shefferson et al. (2007)]. Further
research should also focus on whether evolution in these
orchids and their mycorrhizal fungi occurs as in ways
predicted by the geographic mosaic theory of coevolu-
tion, in which coevolution may be initially rare in an
interaction and yet eventually dominate it due to chance
events and the dynamics of interacting populations.
Acknowledgements
We would like to thank Dr Ono at the Chiba Prefectural Natu-
ral History Museum for help in collection of orchid root sam-
ples, and A. Gross and H. Warner for laboratory assistance.
We are grateful to Dr John Gittleman and the Gittleman lab,
whose discussions helped provide direction to the analytical
methodology. We would also like to thank Elizabeth Farn-
worth and the New England Wildflower Society for help in
locating and sampling plants in New England. Previous drafts
of this manuscript have greatly benefited from reviews from
Dr Scott Nuismer, Dr Dirk Redecker, Dr Marc-Andre Selosse,
and three anonymous reviewers.
References
Abadie J-C, Pu?ttsepp U?, Gebauer G et al. (2006) Cephalanthera
longifolia (Neottieae, Orchidaceae) is mixotrophic: a
comparative study between green and nonphotosynthetic
individuals. Canadian Journal of Botany, 84, 1462?1477.
Alexander C, Hadley G (1985) Carbon movement between host
and mycorrhizal endophyte during development of the
orchid Goodyera repens Br. New Phytologist, 101, 657?665.In summary, we have shown evidence supporting
phylogenetic conservatism in the evolution of host
breadth in a broad interaction, that of the orchid mycor-
rhiza in the rattlesnake plantains (Goodyera spp.).
Although our results corroborate existing patterns
observed in genus Cypripedium, we argue that a broader
sampling within the genus, in particular extending to
species in more difficult to access portions of the Earth,
will be essential to generalizing our inferences to further
systems. We also argue that the most beneficial future
direction for research on the evolutionary ecology of
broad interactions focus on the expansion of the theorysometimes specialize on good sources of organic carbon,
as can be said of myco-heterotrophs (Bruns et al. 2002),
then specialization on these parasitic fungi may also cre-
ate a stable source of carbon, water, and potentially other
nutrients in times when the habitat is harsh. Even in this
case, carbon flow from orchid to fungus has been
observed often enough in Goodyera to warrant suspicion
that it may not be a parasitic group (Cameron et al. 2006, 2010 Blackwell Publishing Ltdby Goodyera repens in relation to mycorrhizal infection. New
Phytologist, 97, 401?411.
Altschul SF, Thomas LM, Alejandro AS et al. (1997) Gapped
BLAST and PSI-BLAST: a new generation of protein
database search programs. Nucleic Acids Research, 25, 3389?
3402.
Anderson B (2006) Inferring evolutionary patterns from the
biogeographical distributions of mutualists and exploiters.
Biological Journal of the Linnean Society, 89, 541?549.
Ansimova M, Gascuel O (2006) Approximate likelihood-ratio
test for branches: a fast, accurate, and powerful alternative.
Systematic Biology, 55, 539?552.
Bidartondo MI (2005) Tansley Review: The evolutionary
ecology of myco-heterotrophy. New Phytologist, 167, 335?352.
Brooks DR, McLennan DA (1991) Phylogeny, Ecology, and
Behavior: a Research Program in Comparative Biology.
University of Chicago Press, Chicago, Illinois.
Bruns TD, Bidartondo MI, Taylor DL (2001) Interactions of
ectomycorrhizal fungi and ectomycorrhizal epiparasites.
American Zoologist, 40, 956.
Bruns TD, Bidartondo MI, Taylor DL (2002) Host specificity in
ectomycorrhizal communities: what do the exceptions tell
us? Integrative and Comparative Biology, 42, 352?359.
Cameron DD, Leake JR, Read DJ (2006) Mutualistic mycorrhiza
in orchids: evidence from plant-fungus carbon and nitrogen
transfers in the green-leaved terrestrial orchid Goodyera
repens. New Phytologist, 171, 405?416.
Cameron DD, Johnson I, Read DJ, Leake JR (2008) Giving
and receiving: measuring the carbon cost of mycorrhizas in
the green orchid, Goodyera repens. New Phytologist, 180, 176?
184.
Cattin M-F, Bersier L-F, Banas?ek-Richter C, Baltensperger R,
Gabriel J-P (2004) Phylogenetic constraints and adaptation
explain food-web structure. Nature, 427, 835?839.
Downie DG (1943) Source of the symbiont of Goodyera repens.
Transactions of the Botanical Society of Edinburgh, 33, 383?390.
Excoffier L, Laval G, Schneider S (2005) Arlequin ver. 3.0: an
integrated software package for population genetics data
analysis. Evolutionary Bioinformatics Online, 1, 47?50.
Gardes M, Bruns TD (1993) ITS primers with enhanced
specificity for basidiomycetes?application to the
identification of mycorrhizae and rusts. Molecular Ecology, 2,
113?118.
Gardes M, Bruns TD (1996) ITS-RFLP matching for
identification of fungi. In: Species Diagnostics Protocols: PCR
and Other Nucleic Acid Methods (ed. Clapp JP), pp. 177?186.
Humana Press, Inc., Totowa, New Jersey.
Gittleman JL, Kot M (1990) Adaptation: statistics and a null
model for estimating phylogenetic effects. Systematic Zoology,
39, 227?241.
Gregor MJF (1935) A disease of bracken and other ferns caused
by Corticium anceps. Phytopathologische Zeitschrift, 2, 401?418.
Guindon S, Gascuel O (2003) A simple, fast, and accurate
algorithm to estimate large phylogenies by maximum
likelihood. Systematic Biology, 52, 696?704.
Guindon S, Lethiec F, Duroux P, Gascuel O (2005) PHYML
Online?a web server for fast maximum likelihood-based
phylogenetic inference. Nucleic Acids Research, 33, W557?559.
Hadley G, Purves S (1974) Movement of 14C from host to
fungus in orchid mycorrhiza. New Phytologist, 73, 475?482.
Hafner MS, Demastes JW, Spradling TA, Reed DL (2003) Rasmussen HN (1995) Terrestrial Orchids: from Seed to
3016 R. P . SH EFFERSON ET AL.Cophylogeny between pocket gophers and chewing lice.
In:Tangled Trees: Phylogeny, Cospeciation, and Coevolutio (ed
Page RDM). pp. 195?220, University of Chicago Press,
Chicago, Illinois.
Hardy OJ, Senterre B (2007) Characterizing the phylogenetic
structure of communities by an additive partitioning of
phylogenetic diversity. Journal of Ecology, 95, 493?506.
Hynson NA, Preiss K, Gebauer G (2009) Is it better to give
than to receive? A stable isotope perspective on orchid-
fungal carbon transport in the green orchid species
Goodyera repens and Goodyera oblongifolia New Phytologist,
182, 8?11.
Janz N, Nylin S (1998) Butterflies and plants: a phylogenetic
study. Evolution, 52, 486?502.
Kores PJ, Cameron KM, Molvray M, Chase MW (1997) The phy-
logenetic relationships of Orchidoideae and Spiranthoideae
(Orchidaceae) as inferred from rbcL plastid sequences.
Lindleyana, 12, 1?11.
Larkin MA, Blackshields G, Brown NP et al. (2007) Clustal W
and Clustal X version 2.0. Bioinformatics, 23, 2947?2948.
Lovette IJ, Hochachka WM (2006) Simultaneous effects of niche
conservatism and competition on avian community
structure. Ecology, 87, 14?28.
Lozupone C, Hamady M, Knight R (2006) UniFrac?an online
tool for comparing microbial community diversity in a
phylogenetic context. BMC Bioinformatics, 7, 371.
McCormick MK, Whigham DF, O?Neill J (2004) Mycorrhizal
diversity in photosynthetic terrestrial orchids. New
Phytologist, 163, 425?438.
McCormick MK, Whigham DF, Sloan D, O?Malley K,
Hodkinson B (2006) Orchid-fungus fidelity: a marriage
meant to last? Ecology, 87, 903?911.
Nei M, Tajima F (1981) DNA polymorphism detectable by
restriction endonucleases. Genetics, 97, 145?163.
Ogura-Tsujita Y, Gebauer G, Hashimoto T, Umata H, Yukawa
T (2009) Evidence for novel and specialized mycorrhizal
parasitism: the orchid Gastrodia confusa gains carbon from
saprotrophic Mycena. Proceedings of the Royal Society of
London. Series B, Biological Sciences, 276, 761?767.
Ormerod P, Cribb PJ (2003) 143. Goodyera. Distribution. In:
Genera Orchidacearum. Volume 3: Orchidoideae (Part two),
Vanilloideae (eds Pridgeon AM, Cribb PJ, Chase MW,
Rasmussen FN), P. 96. Oxford University Press, Oxford.
Otero JT, Ackerman JD, Bayman P (2002) Diversity and host
specificity of endophytic Rhizoctonia-like fungi from tropical
orchids. American Journal of Botany, 89, 1852?1858.
Otero JT, Ackerman JD, Bayman P (2004) Differences in
mycorrhizal preferences between two tropical orchids.
Molecular Ecology, 13, 2393?2404.
Paradis E (2006) Analysis of Phylogenetics and Evolution with R.
Springer Science LLC, New York.
Pommier T, Canback B, Lundberg P, Hagstrom A, Tunlid A
(2009) RAMI: a tool for identification and characterization of
phylogenetic clusters in microbial communities. Bioinfor-
matics, 25, 736?742.
Posada D, Crandall KA (1998) Modeltest: testing the model of
DNA substitution. Bioinformatics, 14, 817?818.
R Development Core Team (2007) R: A Language and
Environment for Statistical Computing. R Foundation for
Statistical Computing, Vienna, Austria.Mycotrophic Plant. Cambridge University Press, Cambridge,
UK.
Roberts P (1999) Rhizoctonia-forming Fungi: a Taxonomic Guide.
Kew Royal Botanic Gardens, Surrey, UK.
Roy M, Watthana S, Stier A et al. (2009) Two mycohetero-
trophic orchids from Thailand tropical dipterocarpacean for-
ests associate with a broad diversity of ectomycorrhizal
fungi. BMC Biology, 7, 51.
Satake Y, Ohwi J, Kitamura S, Watari S, Tominari T (1985)
Wild Flowers of Japan: Herbaceous Plants (Including Dwarf
Shrubs) (in Japanese), p. 300. Heibonsha Publishers, Ltd,
Tokyo, Japan.
Shefferson RP, Wei? M, Kull T, Taylor DL (2005) High
specificity generally characterizes mycorrhizal association in
rare lady?s slipper orchids, genus Cypripedium. Molecular
Ecology, 14, 613?626.
Shefferson RP, Taylor DL, Wei? M et al. (2007) The
evolutionary history of mycorrhizal specificity among lady?s
slipper orchids. Evolution, 61, 1380?1390.
Smith SE, Read DJ (2008) Mycorrhizal Symbiosis, 3rd edn.
Academic Press, New York, New York.
Taylor DL (1997) The Volution of Myco-heterotrophy and
Specificity in Some North American Orchids. PhD Thesis,
University of California at Berkeley.
Taylor DL, Bruns TD (1997) Independent, specialized invasion
of ectomycorrhizal mutualism by two nonphotosynthetic
orchids. Proceedings of the National Academy of Sciences, USA,
94, 4510?4515.
Taylor DL, McCormick MK (2008) Internal transcribed spacer
primers and sequences for improved characterization of
basidiomycetous orchid mycorrhizas. New Phytologist, 177,
1020?1033.
Taylor DL, Bruns TD, Hodges SA (2004) Evidence for
mycorrhizal races in a cheating orchid. Proceedings of the
Royal Society of London. Series B, Biological Sciences, 271, 35?43.
Thioulouse J, Chessel D, Champely S (1995) Multivariate
analysis of spatial patterns: a unified approach to local and
global structures. Environmental and Ecological Statistics, 2, 1?
14.
Thompson JN (2009) The coevolving web of life. American
Naturalist, 173, 125?140.
Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins
DG (1997) The ClustalX windows interface: flexible strategies
for multiple sequence alignment aided by quality analysis
tools. Nucleic Acids Research, 24, 4876?4882.
Weiblen GD, Webb CO, Novotny V, Basset Y, Miller SE (2006)
Phylogenetic dispersion of host use in a tropical insect
herbivore community. Ecology, 87, 62?75.
Wei? M, Selosse M-A, Rexer K-H, Urban A, Oberwinkler F
(2004) Sebacinales: a hitherto overlooked cosm of
heterobasidiomycetes with a broad mycorrhizal potential.
Mycological Research, 108, 1003?1010.
White TJ, Bruns TD, Lee SB, Taylor JW (1990) Amplification
and direct sequencing of fungal ribosomal RNA genes for
phylogenetics. In: PCR Protocols: a Guide to Methods and
Applications (eds Innis MA, Gelfand DH, Sninsky JS, White
TJ), pp. 315?322. Academic Press, New York.
Wiens JJ, Graham CH (2005) Niche conservatism: Integrating
evolution, ecology, and conservation biology. Annual Review
of Ecology Evolution and Systematics, 36, 519?539. 2010 Blackwell Publishing Ltd
Yagame T, Yamato M, Suzuki A, Iwase K (2008) Ceratobasi-
diaceae mycorrhizal fungi isolated from nonphotosynthetic
orchid Chamaegastrodia sikokiana. Mycorrhiza, 18, 97?101.
Yamato M, Yagame T, Suzuki A, Iwase K (2005) Isolation and
identification of mycorrhizal fungi associating with an
achlorophyllous plant, Epipogium roseum (Orchidaceae).
Mycoscience, 46, 73?77.
Yukawa T, Ogura-Tsujita Y, Shefferson RP, Yokoyama J (2009)
Mycorrhizal diversity in Apostasia (Orchidaceae) indicates
the origin and evolution of orchid mycorrhiza. American
Journal of Botany, 96, 1997?2009.
R. P. S. is an assistant professor studying life history evolution
and the mycorrhizal symbiosis in long-lived plants. C. C. C. is
a post doctoral scientist studying the evolution of fungal sym-
biont diversion within mycorrhizal symbioses. M. K. M. is an
ecologist studying the effects of environmental variation on
mycorrhizal associations and how those associations affect
plant and fungal diversity. T. Y. is a botanist studying specia-
tion patterns and mycorrhizal associations in East Asian flow-
ering plants. Y. O. is an ecologist and botanist studying plant
phylogeny from the mycorrhizal viewpoint. T. H. is a botanist
studying the propagation of orchids.
Supporting Information
Additional supporting information may be found in the online
version of this article:
Table S1 ITS haplotypes of basidiomycete root endophytes in
sampled Goodyera plants, likely to function as mycorrhizal
fungi. Here, the number of endophytes refers to the likely
number of species found per the Ceratobasidium phylogeny
(Fig. 1) or via BLAST results
Table S2 BLAST search results of ITS sequences of fungi out-
side of the Ceratobasidiaceae encountered in sampled Goodyera
roots
Fig. S1 Phylogenetic placement of ascomycetous taxa mycor-
rhizal with Goodyera species. Phylogeny determined with
sequences from the fungal ITS region, and includes references
sequences from NCBI GenBank. Analysis was via maximum
likelihood in PHYML for Windows (Guindon & Gascuel 2003;
Guindon et al. 2005), and involved 1000 bootstrap replicates.
Phylogeny is midpoint-rooted.
Please note: Wiley-Blackwell are not responsible for the
content or functionality of any supporting information sup-
plied by the authors. Any queries (other than missing material)
should be directed to the corresponding author for the article.
MYCORRHIZAL HO ST BREADTH EVOLUTI ON 3017 2010 Blackwell Publishing Ltd