Chapter 19 The biogeography of ectomycorrhizal fungi – a history of life in the subterranean

Kabir G Peay1 and P Brandon Matheny2 1 Department of Biology, Stanford University, California, USA 2 Ecology and Evolutionary Biology, University of Tennessee, Knoxville, TN, USA

19.1 Why study the The documentation of basic biogeographic biogeography of patterns is fundamental to understanding ectomycorrhizal fungi? any taxonomic group. Knowing the time or place that an organism first arose and diver­ The science of biogeography maps spatial sified provides a window into the ecological patterns of biological diversity as a means of and evolutionary processes shaping their understanding the evolutionary and ecolog­ diversity and tells us how they responded to ical processes that structure life on this past changes in the environment, or how planet (Lomolino et al., 2006). Understanding they might fare under new climates. Yet, for spatial patterns of biodiversity has given rise the organisms studied by microbiologists to some of the most important discoveries in (fungi, bacteria, viruses) and that constitute modern science. The unique composition of the majority of life on earth, basic details of flora and fauna in different parts of the biogeography remain poorly known, and world, coupled with the spatial and tempo­ are still a subject of debate. ral coincidence of allied species, led both Biogeography was one of the earliest Darwin and Wallace to the idea of evolution branches of plant and animal biology, but by natural selection (Darwin, 1859; Wallace, the development of microbial biogeography 1855). Similarly, Wegener’s theory of conti­ lagged, because the use of morphological nental drift was formulated in great part by systematics to differentiate microbes and the realization that the fossil record pro­ reconstruct their evolutionary histories was vided a window into the past arrangement a “fruitless search”, to quote Carl Woese of landmasses, precisely because of the (Woese, 1987). As a result of the paucity of s­patially restricted nature of species distribu­ morphological characters, until very recently tions. Biogeography may not be a field many microbes were thought to lack distinct familiar to most lay people or most micro­ biogeographical patterns (Finlay, 2002; biologists, but it is a powerful window into Peay et al., 2010c), and the importance of life on this planet. these organisms to the diversity and function

Molecular Mycorrhizal Symbiosis, First Edition. Edited by Francis Martin. © 2017 John Wiley & Sons, Inc. Published 2017 by John Wiley & Sons, Inc.

341 342 Molecular mycorrhizal symbiosis of earth systems was underappreciated 19.2 Ectomycorrhizal (Falkowski et al., 2008). The molecular revo­ biogeography in the lution in DNA sequencing (Horton and pre‐molecular era Bruns, 2001; Woese, 1987) has provided the tools necessary for accurate identification The limitation placed on organisms in their and mapping of microbial distributions, ability to disperse is, perhaps, the primary and is forcing reevaluation of microbial driver of biogeographic patterns. This limita­ biogeography. tion creates a historical tie between where a Ectomycorrhizal fungi (EMF) are obli­ species arises and its geographical distribu­ gate symbionts of dominant tree species in tion. Early biologists viewed fungi as organ­ both temperate and tropical biomes, and are isms with incredible dispersal potential an important part of the diversity of terres­ and cosmopolitan ranges. Wallace, Darwin, trial ecosystems. While ecological studies of Candolle, and von Humboldt, upon whose EMF have traditionally focused on local‐scale writings the field of biogeography developed, processes, understanding the large‐scale dis­ were focused primarily on the distributions tributional patterns of these organisms can of plants and animals and, where fungi shed light onto many key biological ques­ were noted, it was mostly to indicate their tions, such as the role of symbiosis in shaping uniqueness. species distributions, and how nutrient Candolle, an influential Swiss botanist, mutualisms are affected by climate and geol­ alluded to this in his observation that fungi ogy. Mushroom‐forming fungi are relatively would occur, “everywhere upon the earth rich in morphological features among … when the same circumstances propitious microbial groups. However, the study of their to their production occur” (de Candolle, biogeography has also progressed slowly. 1820). Similarly, the Reverend Miles Joseph In this chapter, we review the current Berkeley, one of the early founders of the state of EMF biogeography in light of new field of mycology, wrote in a letter to Charles data stemming from the use of DNA‐based Darwin, “that were not Fungi so much the molecular tools. We begin first with a look at creatures of peculiar atmospheric conditions, the pre‐molecular history of EMF biogeog­ there would seem to be no limit to the diffusion raphy, before turning to recent develop­ of their species” (Berkeley, 1863). ments. The remainder of the review moves While Darwin and Wallace could view from smaller to longer temporal scales, large‐scale distributional patterns of plants beginning with current species’ distribu­ and animals as a historical record of evolu­ tional patterns, and their effects on commu­ tion precisely because of their limited ability nity composition and diversity at large to move between regions, the presence of scales. From there, we take a phylogeo­ fungi seems to have been seen by contem­ graphic approach to examine the population porary biologists more as a reflection of the level processes that give rise to species distri­ “peculiar” environmental conditions in a butions over thousands of years, before particular place, rather than a record of finally turning to historical biogeography to evolutionary history. This perception was examine how the movement of continents perhaps self‐reinforcing; early fungal tax­ and long‐distance dispersal have shaped onomists arriving in North America expected modern distributions of EMF taxa. to find European species and, as a result, Chapter 19: The biogeography of ectomycorrhizal fungi – a history of life in the subterranean 343

(incorrectly) applied European names to the north from South American Nothofagaceae species they found – a problem that is still (southern beech) associates. In hindsight, being sorted out today. Consequently, however, none of these scenarios is likely m­aking biogeographic conclusions from dis­ correct, as these investigators were misled at tribution records based on morphological a more fundamental level by the fact that taxonomy can be highly misleading (Pringle the key morphological feature distinguish­ and Vellinga, 2006). ing Rozites as a genus (presence of a mem­ The early view that fungi were “unrelia­ branous veil) evolved multiple times ble as biogeographic markers”, coupled with independently in the Cortinariaceae a paucity of reliable taxonomic data, slowed (Peintner et al., 2002). the development of fungal biogeography Even for more conspicuous taxa, basic (Pirozynski, 1983) for nearly a century. biogeographic patterns were difficult to dis­ However, improved fungal taxonomy and tinguish. For example, Watling proposed a increased global migration of European or boreal origin for boletes (Watling, 2001), North American trained systematists led to rather than an origin centered on their cur­ some attempts to assemble geographic dis­ rent day diversity in Southeast Asia (Corner, tribution information for fungi, and to 1972). While these hypotheses were not determine the extent to which factors such directly testable at the time, these early as climate or host identity affected these dis­ s­cientists outlined the key questions that tributions (Bisby, 1943). Pirozynski (1983) continue to drive the field: was an early champion for the study of EMF (i) What are the centers of ectomycorrhizal as a window into biogeography, as he sus­ endemism? pected that their tight associations with host (ii) How important is long‐distance dispersal plants would more greatly limit their disper­ in shaping species distributions? sal and lead to distinct geographic pattern­ (iii) What are the key migration routes that ing, compared with other fungal groups. link regions? Still, he suspected much less endemism in (iv) When did this migration occur? fungi than in plants, at least in the context of These questions have proved fertile insular islands. grounds for the application of molecular Biogeographic hypotheses began to tools to assess biogeographic patterns among emerge, particularly for distinct groups with closely related species or populations (the clear centers of diversity. For example, the field of phylogeography), and more coarse strongly southern hemisphere‐biased distri­ global patterns at deeper taxonomic levels bution for the genus Rozites was thought to (the field of historical biogeography). suggest a Gondwanan origin (Bougher et al., 1994; Horak, 1981). However, distinguish­ ing amongst finer‐scale biogeographic sce­ 19.3 the advance narios was difficult. In examination of some of molecular phylogeography Central American Rozites, for example, Halling (Halling and Ovrebo, 1987) noted The development of DNA‐based tools for that it was not possible to tell if they were studying fungi has provided an unambigu­ northern co‐migrants with oaks from an ous way to discriminate fungal species Indo‐Malayan land bridge, or had migrated and reconstruct evolutionary relationships 344 Molecular mycorrhizal symbiosis

(Bruns et al., 1991). From the early discov­ similarities in gross morphology. However, ery of the polymerase chain reaction (PCR) recent sequencing of ITS and large subunit and development of fungal specific primers genes from western North American sam­ (Gardes and Bruns, 1993), the accessibility ples revealed monophyletic groups distinct of DNA sequence data has increased relent­ from Europe and Eastern North America lessly. The data provided by DNA sequences and, within the western groups, there were have advanced development of EMF bioge­ at least four cryptic species (Nguyen et al., ography in a number of critical dimensions. 2013). Notably, in this case, the use of DNA First, DNA sequence data allow a very sequence data is leading formal taxonomic rapid means of differentiating species. The description. Thus, distributional data can be internal transcribed spacer (ITS) regions of generated to support biogeographic research, the nuclear ribosomal RNA genes (nrDNA) even in advance of formal taxonomy, which are highly variable, and can often distin­ is often a much slower process. guish even closely related species. The gen­ Another important implication of DNA‐ erally accepted convention is that species are based study of EMF communities is that delineated at approximately 97% pairwise d­istributional data can be collected from sequence similarity, a level that corresponds vegetative structures growing in soil. Many well with established morphological species species fruit infrequently (Straatsma et al., (Hughes et al., 2009; Smith et al., 2007). 2001) and may be difficult to find, particu­ There are, of course, exceptions at which larly for species that grow in remote sites, this level of cutoff misses important details, not easily visited on a regular basis. For but it is useful enough that the ITS has example, Peay et al. (2010b) were able to served as the basis for molecular species identify > 100 species of EMF in less than identification, and has been selected as the five days of sampling, by sequencing colo­ official barcode locus for fungi (Kõljalg et al., nized root tips from a lowland rainforest 2013; Schoch et al., 2012). in Borneo while, over the same time period, The ability to confidently assess whether < 20 species of EMF were identified from two collections belong to the same species is fruiting bodies (Peay, unpublished). Increasing the basis for determining accurate distribu­ ease of data collection and international travel tional maps for EMF species. While morpho­ has led to a much greater influx of data from logical approaches can be highly accurate countries where little data were previously with a given site or season, their reliability available (Smith et al., 2011; Tedersoo et al., becomes less certain as the time and space 2007). between collections increases. For example, by sequencing both historical and modern collections, Pringle and Vellinga (2006) were 19.4 Ectomycorrhizal able to show that Amanita phalloides (the communities in space death cap) was introduced to North America, and that historical collections using this The documentation of broad spatial patterns name were misidentified. of biodiversity in plant and animal life is Similarly, the name Helvella lacunosa fundamental to current understanding of has been broadly applied to mushrooms ecology and evolution. The differences in across North American and Europe, based on species composition between North America Chapter 19: The biogeography of ectomycorrhizal fungi – a history of life in the subterranean 345

and Europe, or between tropical Southeast While Read proposed that these patterns Asia and South America, are indicators of are driven by nitrogen availability, recent the separate history of evolution in these studies have called into question whether places. Similarly, the convergence in form the nitrogen economy of ectomycorrhizal between regions with different evolutionary plants truly differs from other plants (Koele histories – for example, the Mediterranean et al., 2012). As such, it seems highly likely vegetation of California and southern that the differences in the abundance of Europe – points toward similar ecological ectomycorrhizal plants between the neo‐ constraints. These differences give rise to the and paleotropics is a product of the evolu­ biogeographic realms, and similarities to the tionary history of these two regions, rather biomes used in introductory textbooks (Cain than differences in environmental conditions et al., 2014; Lomolino et al., 2006). Our between the two. understanding of EMF distributional patterns lags far behind, both in terms of the distribu­ tion of the symbiosis itself and the distribution 19.5 Dispersal limitation of individual species or assemblages. Early attempts to map the distribution of Documenting the global distribution of EMF EMF symbiosis itself relied on our under­ species has been impossible until very standing of plant communities, as obligate recently. The ability to perform continental host associated fungi cannot grow without or global scale sampling of EMF communi­ their hosts. These maps showed an increase ties in situ became possible with the adapta­ in the frequency of EMF associations in tem­ tion of DNA sequencing approaches, such as perate and boreal forests, compared with 454 Pyrosequencing (Buée et al., 2009) or tropical forests or grasslands (Read, 1991). Illumina (Smith and Peay, 2014), to gener­ While it is true that EMF associations are ating ITS barcode sequence data. While we very common in temperate and boreal for­ are still in the early stages of analyzing ests (e.g., Pinaceae, Fagaceae), the pattern is these large datasets, some clear patterns are somewhat more complicated in tropical beginning to emerge. forests. The tropical rainforests of southeast Perhaps the clearest result is that EMF are Asia are dominated by the plant family not cosmopolitan. Evidence towards this end Dipterocarpaceae, which is ectomycorrhizal, had been accumulating from other types of and often account for over 40% of basal data. For example, studies of EMF from area. In neotropical rainforests, ectomycor­ poorly studied tropical regions (Peay et al., rhizal host plants are generally rare, but 2010b; Smith et al., 2011; Tedersoo et al., EMF‐hosting genera, such as Coccoloba 2010b) generally tended to show very little (Polygonaceae), Neea and Guapira (Nycta­ sequence similarity with well‐sampled ginaceae), Dicymbe (Caesalpinaceae), Aldina European or North American studies. (Fabaceae), and Pakaraimea (Diptero­ Similarly, Sato et al. (2012) matched ITS carpaceae) can vary wildly in abundance – sequences from fruiting bodies in a well‐ from minor components (Tedersoo et al., characterized site in Japan to global sequences 2010b), to forming large tracts of nearly in GenBank to map species’ ranges, and monodominant forest (Henkel, 2003; Smith found that most EMF had relatively restricted et al., 2013). ranges. Similarly, a recent attempt to create a 346 Molecular mycorrhizal symbiosis standardized database for ITS sequences using second‐generation sequencing. Talbot (Kõljalg et al., 2013) showed that approxi­ et al. (2014) sequenced hundreds of soil mately 80% of molecularly defined species‐ samples from EMF‐dominated pine forests level operational taxonomic units (OTUs) across North America. Soil samples were were restricted to a single continent. A simi­ separated into organic and mineral ­horizons lar meta‐analytic approach combined multi­ in order to look at community composition ple second‐generation sequencing studies, across large spatial gradients where environ­ and also concluded that there was limited ments vary, but also to contrast this with the taxon sharing between biogeographic realms effects of steep local environmental ­gradients. (Meiser et al., 2014). Tedersoo et al. (2014a) examined data from However, such approaches are limited, in sites across the globe, including all conti­ that they combine datasets collected using nents except Antarctica. Both studies used different methods, rely on the accuracy of a consistent sampling approach, molecular international sequence databases that are methodology, and included hundreds of notoriously error‐plagued (Bidartondo et al., samples. In both cases, the studies found 2008), and generally lack associated environ­ limited taxon‐sharing across study sites mental variables. From this perspective, two (see Figure 19.1). recent studies have attempted to examine Despite the fact that Talbot and colleagues continental‐scale fungal distribution patterns were looking at forests dominated by the

Geographic origin (color)

0.10 Eastern

Midwestern 0.05 Boreal

0.00 Western NMDS2 –0.05 Soil horizon (shape)

OH –0.10

–0.1 0.0 0.1 0.2 0.3 AH NMDS1

Figure 19.1 Regional differentiation of ectomycorrhizal fungal communities. The non‐metric multidimensional scaling (NMDS) ordination shows community similarity of ectomycorrhizal fungi detected in 561 individual soils cores through second generation DNA sequencing of the ITS region. Soil cores were collected from essentially monodominant natural Pine stands soils across North America. Geographic location (color) and soil horizon (symbol) are mapped onto soil cores according to the legend. Ectomycorrhizal fungi communities cluster primarily by geographic locality from which they were sampled. There was little effect of soil horizon on community similarity at the continental scale, despite strong, consistent changes in soil chemistry across organic (OH) and mineral (AH) horizons. These results demonstrate that biogeographic factors are the first order determinants of ectomycorrhizal community structure. (See insert for color representation of the figure.) Chapter 19: The biogeography of ectomycorrhizal fungi – a history of life in the subterranean 347 same plant family, and that soil samples One interesting approach to this is the showed a very strong vertical gradient in soil examination of spatial patterns within a sin­ chemistry, they found that EMF commu­ gle host group. The genus Alnus is interest­ nities, and soil fungal communities in gen­ ing in that it has a broad distribution (Europe, eral, were differentiated between regions of Asia, N. America, and S. America), and has North America, and that decay in commu­ a fairly unique habit, so often grows alone. nity similarity was best predicted by distance A few studies have compared Alnus EMF com­ between two samples. Tedersoo et al. (2014a) munities across large scales (Kennedy et al., also found that, at a global scale, biogeo­ 2011; Polme et al., 2013). These studies graphic regions contained unique species found regional differentiation of Alnus EMF, and, as a result, EMF communities clustered and that climate and soils explain some, but geographically – for example, New Guinea, not much variation. Polme found that the Australia and New Zealand in one cluster, greatest predictor of EMF community struc­ and Europe, West Asia, North America and ture was host phylogeny, even within this East Asia in another. closely allied group of species. While they found some evidence for Talbot et al. (2014) tried a similar taxon sharing across similar biomes, the approach, but also used the vertical soil d­ifferences between geographic regions are horizon to separate spatial and environmen­ consistent with the effectiveness of oceans tal variables. They also found that soil chem­ and mountains as dispersal barriers, and the istry explained only a small amount of historical legacy of such barriers appear to variability across these large scales. Similarly, play a major role in the current structure of Tedersoo et al. (2014a) found that climate fungal communities. Intriguingly, a recent and environmental variables explained only a global study of arbuscular mycorrhizal fungi fraction of global variance in EMF species com­ (AMF) showed very low levels of endemism position, in contrast with local scale studies of and little geographic structure (Davison EMF community composition that showed et al., 2015). Because these two groups have very strong effects of soil environment (Peay a similar trophic niche, it will be important et al., 2010b; Taylor et al., 2014; Toljander to understand whether the different patterns­ et al., 2006). Taken together, the preponder­ arise from differences in ecology (dispersal ance of evidence suggests that the pool of ability, host range), or evolution (clade age, available EMF in a community is set first by evolutionary rates, taxon delineation), or both. the biogeographic history of the region, in A major aim of these biogeographical which the community is embedded, and studies is also to identify predictable drivers that environmental sorting acts as an impor­ of EMF composition. This is particularly tant filter upon this pool at the local scale. challenging, because factors such as climate, soil chemistry and host species composition also tend to vary at these same scales. 19.6 Spatial patterns Because it is not possible to manipulate of diversity c­limate or host composition at such scales, innovative approaches are necessary in Diversity of EMF has been traditionally exam­ order to tease apart the main drivers of these ined through the lens of local processes – that spatial effects. is, that local conditions at the centimeter 348 Molecular mycorrhizal symbiosis or meter scale, such as pH or organic matter spores might disperse (i.e., the “mainland”). concentration, determine the number of They found that EMF diversity patterns EMF in a given sample (Bruns, 1995). corresponded well with patterns expected However, there are a number of biogeo­ under island biogeography theory. Species graphic hypotheses that predict local richness­ richness of EMF was highest on large on the basis of factors that operate and are host islands and host islands closer to the measured on the scale of kilometers and “mainland”. continents (Ricklefs, 1987). A handful of The increase in species richness with studies have now evaluated the importance increasing sample area is known as the spe­ of biogeographic hypotheses explaining cies area relationship (SAR), and has been diversity patterns in EMF. described as one of the few laws in ecology. The theory of island biogeography The shape of this relationship is normally (MacArthur and Wilson, 1967) postulates modeled as S = cAZ, where S = Species that richness is a function of the effects of Richness, c is a constant, A is the area sam­ island geography on rates of immigration pled, and Z describes the rate at which rich­ and extinction on that island. The theory is ness increases with area. There is no doubt based on population‐level processes, where that microbial communities are highly extinction rates are linked to population diverse, but there has been some debate sizes via island size, and immigration rates about whether the species area relationship are linked to island isolation from potential scales as rapidly as with macrobes. This is dispersal sources, such as the mainland. The important to our view of microbial diversity, theory predicts that larger islands should as a shallow SAR paints a picture of microbes house more species than smaller islands, that are incredibly diverse at small scales and that isolated islands should have fewer (e.g., single soil samples), but with few new species than those closer to the mainland. species added as new areas are explored. While no studies have examined EMF rich­ Some previous work on soil fungi and ness on actual islands in the light of island bacteria found SAR slopes (or z‐values) in biogeography theory, Peay and colleagues the range of 0.04–0.07, among the lowest (Peay et al., 2007, 2010a) used a “host island” values observed for any taxonomic group approach to study island biogeography of (Green et al., 2004; Horner‐Devine et al., EMF (Janzen, 1968). From the perspective 2004). By contrast, Peay et al. (2007) esti­ of obligate plant symbionts like EMF, non‐ mated a z‐value of 0.20 for EMF – more in host plants are inhospitable territory (much line with SAR slope values estimated for like an ocean is to most terrestrial organ­ macrobial organisms such as plants. The isms). Host plants can, thus, be thought of as nature of this relationship has important islands in a sea of non‐host vegetation. consequences for understanding diversity Using this approach, Peay and colleagues patterns in nature, and in understanding examined patterns of EMF diversity across how microbial communities will respond to a landscape in which EMF hosting pines processes like habitat fragmentation or occur patchily in a matrix of AMF shrubs. destruction. These patches varied greatly in size (1 to > There are still relatively few studies that 10 000 m2) and in their distance away from have estimated SAR for fungi (or EMF). a large tract of continuous forest from which However, our re‐analysis of EMF richness Chapter 19: The biogeography of ectomycorrhizal fungi – a history of life in the subterranean 349 from willow host islands on Mount Fuji h­undreds (Peay et al., 2012) of meters. (Nara et al., 2003) returned a z‐value esti­ Similar results have been reported as well mate of 0.23. This is probably somewhat of from saprotrophic fungi (Norros et al., 2012), an overestimate, though, as the size of these suggesting that dispersal limitation at the host islands increased with time, and so landscape scale is likely a common occur­ larger islands were also older. A very high‐ rence in fungi. Read in this context, the intensity sampling and sequencing effort for prolific spore production by EMF is not an all soil fungi across grasslands in the Swiss indicator of unlimited dispersal; rather, it is Alps also documented a fairly strong increase an indication of the difficulty of dispersal. in richness with sampling area, and esti­ Large sequence‐based datasets also allow mated that a minimum area of 16 km2 would for the exploration of major patterns of bio­ be needed to observe all fungi detected in geography to see if they are also consistent the study (Pellissier et al., 2014). in EMF. Perhaps the most well‐known spa­ As previously mentioned, dispersal is a tial diversity pattern is the latitudinal diversity critical process in most biogeographic mod­ gradient, where most groups of plants and els, and yet its role in the ecology of micro­ animals peak in diversity in tropical regions. organisms has been heavily debated. The Whether or not such patterns apply to dispersal process is not entirely well under­ microbial organisms has been unclear. For stood in EMF, in part due to the difficulty of example, marine microbes tend to peak in measuring dispersal and in part due to a diversity at mid‐latitudes, as do Bryophytes. poor understanding of the behavior of spores Work on EMF has suggested either no gradi­ in the life cycle of most EMF species. For ent (Peay et al., 2010b) or an inverse diversity example, critical details of spore biology, gradient (Tedersoo and Nara, 2010). such as germination cues (Nara, 2009), sur­ Current results seem to indicate that dif­ vival ability (Bruns et al., 2009; Peay et al., ferences in EMF diversity across latitudes 2009), and dispersal kernels (Galante et al., are fairly weak, perhaps with some diversity 2011), are unknown for most taxa. peaks at mid‐latitude northern temperate There are two parts to the dispersal forests (Tedersoo et al., 2014a). Explaining ­process – the movement of spores, and their this result is one of the more interesting subsequent colonization of host roots. Fungi challenges in EMF biogeography, and it is produce phenomenal numbers of spores, possible that deterministic factors, such as and single fruiting bodies may produce hun­ climate or soil chemistry, play an important dreds of millions of spores. However, a num­ role. However, if EMF diversity is highly ber of studies of local‐scale dispersal show dependent on host area, as predicted by that spore numbers decrease dramatically island biogeography (Peay et al., 2007), this with distance away from fruiting bodies, and pattern might result from the low abun­ that the vast majority of spores fall relatively dance of EMF hosts in many low‐latitude short distances from the fruiting body forests. (Galante et al., 2011; Li, 2005; Peay et al., Another possibility comes from the 2012). Experimental studies indicate that so‐called mid‐domain effect. The mid‐domain rapid decay in spore quantity can limit the effect hypothesis predicts that diversity will occurrence of EMF on seedlings, beginning be highest in areas where multiple species from tens (Dickie and Reich, 2005) to have overlapping ranges. A recent study 350 Molecular mycorrhizal symbiosis from Japan showed that this was true of related sister taxa – generally at the time­ EMF diversity in mid‐elevation montane scale of thousands to millions of years. The forests (Miyamoto et al., 2014). While this second, historical biogeography, attempts to would not explain the low diversity of reconstruct the evolutionary events that EMF in tropical regions, in combination influence the distribution and diversity of with SAR based‐theories it might explain large clades (families, orders) at the scale of the relatively high diversity in temperate tens to hundreds of millions of years, and versus boreal forests. with respect to key geological events such as the movement of continents and oceans (Lomolino et al., 2006). 19.7 eMF across time Despite the promise of this, field fungi and space are one of the most poorly studied major taxonomic groups from a phylogeographic Ectomycorrhizal fungi hail from a diverse perspective (Beheregaray, 2008; Lumbsch evolutionary background, with many inde­ et al., 2008). Phylogeoraphic studies of EMF pendent origins of this lifestyle within the have tended to focus on broadly distributed kingdom Fungi (Tedersoo et al., 2010a). species or species complexes, primarily in While contemporary geographic patterns in the northern hemisphere. A few key find­ community composition and biodiversity ings appear to be emerging from these stud­ are useful, by comparing all lineages at ies. The first is that broadly distributed once they ignore the valuable data on the “species” tend, in fact, to be complexes of historical processes that structure current species, often with non‐overlapping geo­ distributions, and which can be gained from graphic ranges. For example, the fly agaric, taking an explicitly evolutionary perspec­ Amanita muscaria, is widely distributed across tive. These kinds of historical events can be temperate and boreal forests in the north­ reconstructed using phylogenetic trees based ern hemisphere. Molecular phylogenetics on DNA sequences (Avise et al., 1987; showed that A. muscaria is actually a com­ Figure 19.2). plex of at least three distinct species (Oda Such analyses require a more specific et al., 2004) and that, despite the wide focus on particular lineages, and are not as variety­ of A. muscaria color morphs, these yet easily done using bulk environmental are not reliable indicators of phylogeny as sequences, generally requiring significant multiple color morphs occurred in all clades collections of fungal fruiting bodies. While (Geml et al., 2006). the results from any particular lineage reflect The second finding is that there is gener­ the chance events that make their evolu­ ally little geographic overlap in ranges tionary history unique, evidence from mul­ among these species complexes, and little or tiple independent lineages provides the basis no evidence for current dispersal between for strong generalizations. Here, we exam­ isolated geographic regions (Chapela and ine evolutionary evidence at two scales. The Garbelotto, 2004; Nguyen et al., 2013; Oda first, phylogeography (Avise et al., 1987), et al., 2004). Analysis of Tricholoma matsutake uses molecular evidence to reconstruct the (Chapela and Garbelotto, 2004) showed that historical events that influence modern dis­ matsutake allies can be clearly differentiated tributions of a single species or set of closely into species groups, and that these groups Chapter 19: The biogeography of ectomycorrhizal fungi – a history of life in the subterranean 351

Vicariance

(a)

Long-distance jump dispersal

(b)

Overland migration

(c)

Time

Figure 19.2 The use of molecular phylogenies to infer past events shaping current day distribution patterns in ectomycorrhizal fungi. The phylogeny depicts three different scenarios to explain the modern‐day disjunction between species found in the grey and white areas, as indicated by colored symbols at the tips of the phylogeny. Gaps between areas indicate large barriers (oceans, mountains) across which species cannot easily move. Different scenarios depend primarily on the timing of connections between the grey and white areas at the node marked with a*. In scenario (a), the split between grey and white taxa is driven by a vicariant event that breaks up the ancestral range, followed by speciation. In scenario (b), the separation of the two areas predates the origin of the white lineage, and so must result from dispersal (arrow) across the barrier, followed by speciation. In scenario (c), the origin of the white lineages corresponds with the occurrence of a temporary land bridge, suggesting overland migration followed by speciation. can be clearly differentiated into clades with non‐overlapping distributions in many parts distinct geographic ranges, including a cir­ of the A. muscaria range (see Table 19.1). cumboreal clade, a clade in western North The third major finding from many of America, eastern North America and then these studies is that current patterns of Central America. For A. muscaria, the picture genetic diversity and distribution appear is slightly more complicated. All three linked to the recent history of glaciation and ­phylogenetic groups currently coexist in recolonization. The few EMF lineages that Alaska and are inferred to have a common have been examined appear to have north­ origin in Beringia. However, different migr­ ern origins during the Neogene. Both T. mat­ ation histories­ from Beringia have led to sutake and A. muscaria originated in humid Table 19.1 A summary of recent ectomycorrhizal biogeography studies based on molecular data. Mean or median age (million years) of crown groups are reported for each clade, with 95% highest posterior density (HPD) in parentheses. Where multiple estimates are available only the most recent is given.

Taxon Ancestral range Clade age MYA (HPD) Geographic structure Primary dispersal modes Sources

Tuber Eurasia 142 (122–162) Yes – many endemic clades in Europe. Primarily intra‐continental migration (Asia, Bonito et al. (2013) Distinct sister lineage present in Europe) and via landbridge migration to Jeandroz et al. Southern hemisphere. N. America (2008) Inocybaceae Paleotropics 60 (40–80) Yes – Some likely vicariance events Some boreotropical migration, possible jump Ryberg and (Africa‐India), limited clade diversity in dispersal for Australian and temperate Matheny (2011) S. America. S. American lineages. Matheny et al. (2009) Amanita sect. Paleotropics 56 (39–73) Yes – most species occur in clearly Primarily overland consistent with boreotropical Sánchez‐Ramírez Caesareae differentiated regional groups. migration. et al. (2015) Sclerodermatineae Asia/North 82 (60–115) Yes – at the individual species level but Very broad ancestral range, with some Wilson et al. America most clades old enough for broad pantropical lineages consistent with boreotropical (2012) distribution. migration. Frequent long‐distance dispersal. Hysterangiales Gondwana Not dated Yes – endemism at species level but Initially broad ancestral range, frequent Hosaka et al. geographic split only evident at deep long‐distance dispersal. (2008) nodes between northern and southern hemisphere. Austropaxillus‐ Western N.A. 22 (32–12) Yes – Australasian clade nested within Long‐distance dispersal; in situ diversification in Skrede et al. Gymnopaxillus (Serpulaceae) north temperate clade Australasia (2011) (Serpulaceae) Porcini clade Palaeotropical 42 (23–62)* Yes – lineages rare in southern Episodic long‐distance dispersal Dentinger et al. hemisphere (2010) Cortinarius Australasia 14 (18–10) Yes – north and south American lineages Long‐distance dispersal and founder‐event Harrower et al. violaceus group nested within an Australasian clade speciation (2015) Clavulina Tropical Not estimated (relative Yes – temperate clades derived from Not addressed Kennedy et al. rates used) tropical groups (2012) /Sebacina North America 57–45 MYA Yes – four major biogeographic groups Strong impact of dispersal limitation on large Tedersoo et al. (Sebacinales) detected scale; recent dispersal in northern Holarctic and (2014b) to Mediterranean and southern hemisphere

* Slightly older dates are also reported, based on a different calibration point. Chapter 19: The biogeography of ectomycorrhizal fungi – a history of life in the subterranean 353 temperate forests that occurred in the north, the evolutionary history of fungi over geo­ and then migrated southward as the climate logical time scales. The field has developed cooled. Some of the current genetic differ­ slowly, in part because the fossils that play entiation is interpreted as being the result of an important role in documenting the pres­ persistence of species in glacial refugia and ence or absence of particular forms as conti­ their separation by non‐host vegetation. nents joined and split are not suitably Murat et al. (2004) also showed that, for abundant for fungi to reliably document European Tuber melanosporum, genetic diver­ their distributions. However, the few fossil­ sity closely correlated with location of sus­ ized fungi that are available (e.g., Boyce pected glacial refugia, and that migration of et al., 2007; Hibbett et al., 1995) can be current genotypes can be tracked along sus­ incredibly useful as anchor points to cali­ pected host migration routes. The effects of brate the time scale in molecular phyloge­ glaciation are also evidenced by low allelic nies, and help provide more accurate dates diversity resulting from bottlenecks. The fact for key events. that gene flow seems to follow host migra­ With historical biogeography, the timing tion routes is also good evidence for the of splits between lineages (e.g., the origin of effectiveness of mountains or oceans as dis­ a new node in the tree) with respect to the persal barriers (Amend et al., 2010). position of the continents is critical in under­ One interesting exception may be the standing the manner in which current dis­ arctic fungi (both EMF and non‐EM), which tributions (and, in particular, disjunct appear to show relatively frequent transcon­ distributions) arose. If the nodes originated tinental dispersal, perhaps due to the prior to the breakup of a landmass or the extreme effects of climate change at high origin of a dispersal barrier (e.g., orogeny), latitudes and the need for very large disper­ the disjunction most likely represents sal events to track them (Geml, 2011). allopatric speciation; however, if the node Consistent with this, EMF communities on arises after the breakup, the disjunction the arctic island of Svalbard show evidence must result from long‐distance dispersal (de for frequent long‐distance, transoceanic col­ Queiroz, 2005). The timing of these events onization events (Geml et al., 2011). can provide insight into the biology and In general, the use of phylogeography to selection processes that are unique to EMF. identify current hotspots of genetic diversity For example, some key questions are how appears to match well with what is known common is long‐distance dispersal in EMF, is about locations of glacial refugia from plant the distribution of EMF determined primar­ studies. While it does seem as if these studies ily by the distribution of particular hosts, all show a concordance between host migra­ and where and when did most EMF lineages tions and EMF migrations, it is not clear if originate? the fungi are limited by or tracking the host Broadly speaking, it appears that most plants, or if both plant and fungal popula­ major fungal lineages (e.g., at the family tions are equally limited by the same disper­ or order level) are old enough to have sal barriers. global distributions. For example, the same As for phylogeography, there are still few EMF genera (Russula, Tomentella, Lactarius, studies that have attempted to reconstruct Inocybe) are dominant players in neotropical 354 Molecular mycorrhizal symbiosis rainforests and paleotropical rainforests, as conditions under which EMF symbiosis can well as in temperate and boreal forests, arise as a successful ecological strategy for despite the fact that few of these systems both plants and fungi. appear to share any biological species. While Historical biogeography studies also AMF have been around since plants first show that, for most large lineages of EMF, colonized the land, ectomycorrhizal symbi­ there are strong patterns of regional ende­ osis likely did not originate until ≈ 270–130 mism, with most species and many clades MYA, based on molecular dating of fungal restricted to single continents (Bonito and plant phylogenies (Hibbett and Matheny, et al., 2013; Hosaka et al., 2008; Matheny 2009). et al., 2009). The geographic range of indi­ The EMF habit has originated numerous vidual clades within EMF groups can also times (Tedersoo et al., 2010a) independently vary significantly. Within the Tuberaceae, from saprotrophic ancestors (Wolfe et al., the majority of clades are restricted to sin­ 2012). The transition from saprotrophy to gle continents, but a few, such as the black ectomycorrhizal symbiosis involves pro­ truffle\melanosporum clade, are widely found genomic changes, including the loss distributed across Europe, Asia and North of genes coding for plant cell wall degrad­ America (Bonito et al., 2013). The degree ing‐enzymes and an increase in genes for of endemism also varies, depending on secreted effector‐like proteins that mediate taxonomic scale. The Hysterangiales show host interactions (Kohler et al., 2015). Once clear evidence of separate Holarctic and evolved, though, the EMF habit generally Gondwanan lineages, but groups within appears to be stable, and reversion to the Holarctic appear to be broadly distrib­ s­aprotrophy is uncommon (Bruns and uted across Europe, North and Central Shefferson, 2004). America. Given the unique nutritional aspects of As suspected by early fungal biogeogra­ the symbiosis, understanding when and phers (Pirozynski, 1983), there does appear where the ectomycorrhizal habit evolved to be a unique set of southern hemisphere will give fundamental insight into the selec­ fungi. For example, within the Tuberaceae, tive pressures that lead to this lifestyle the genus Tuber is found primarily in the transition.­ This is an important question in Northern Hemisphere, while a distinct set of Earth’s history, as the presence of ectomyc­ sister taxa are found in Australia, New orrhizal symbiosis has strong impacts on Zealand and southern South America. In geochemical cycles (Averill et al., 2014; this case, the estimated divergence date Clemmensen et al., 2013). While it is easy to between northern and southern lineages, think of EMF as primarily a north temperate around 156–160 MYA corresponds with the association, based on phylogeographic stud­ breakup of Pangaea into the Gondwanan ies a number of EMF groups, including the and Laurasian landmasses (Bonito et al., Inocybaceae (Matheny et al., 2009), Amanita 2013). sect Caesareae (Sánchez‐Ramírez et al., 2015) Disjunct distributions are present in and Sclerodermatineae (Wilson et al., 2012), nearly every phylogenetic reconstruction show evidence for paleotropical origins. used in EMF historical biogeography. While The diversity of lineages with paleotropical there are some splits (e.g., Tuber) that match origins may provide important clues to the the timing of known vicariance events, such Chapter 19: The biogeography of ectomycorrhizal fungi – a history of life in the subterranean 355 as the breakup of Pangaea, there are many 19.8 Conclusions others that do not. These non‐vicariant d­isjunctions provide evidence that, over The field of EMF biogeography is still young, geological time frames, EMF are quite capa­ as it requires molecular tools that have only ble of long‐distance dispersal. For example, recently been developed. However, in the the timing of origin of some nodes associ­ last decade, we have already learned a great ated with Inocybe species in Australia and deal. The pre‐molecular view of unlimited New Zealand is incompatible with vicari­ dispersal has been rejected and replaced by a ance and, therefore, must have arisen from much more nuanced view of ectomycorrhi­ transoceanic dispersal (Matheny et al., zal ecology and evolution. Current patterns 2009). Neotropical species of the Cortinarius of EMF distribution reflect the importance violaceous group appear to have arisen from a of dispersal barriers and the unique timing long‐distance dispersal event from Australia and geological context in which individual within the last 13–7 million years (Harrower lineages arose, migrated and evolved. The et al., 2015). Molecular genetic evidence for presence of northern migration routes saprotrophic fungi is also consistent with between Asia and North America, and the episodic transoceanic dispersal events migration of host trees with climate and ice throughout the southern hemisphere ages, are evident in EMF phylogenies. (Moncalvo and Buchanan, 2008). While long‐distance dispersal events Along with the evidence for transoceanic have occurred, and are certainly important spore movement, many of the current in EMF, it is not clear that they are dispro­ d­isjunctions in EMF appear to have arisen portionately important, compared with the from overland migration and subsequent role of long‐distance dispersal in plant speciation. One scenario, the boreotropical (Higgins et al., 2003) or animal (Springer hypothesis, suggests that many groups et al., 2011) biogeography. In fact, it has reached their current distributions during been recently argued that long‐distance dis­ the late Miocene, when tropical forests persal events have been more common and extended north into areas that are currently important in the shaping the biogeography temperate. At higher latitudes, dispersal of plants and animals than previously sus­ between Europe, Asia and North America pected (de Queiroz, 2005, 2014). Similarly, was then possible through land bridges in the richness of EMF communities does not places such as Beringia. A very nice example seem to follow those of plants or animals at of the boreotropical hypothesis comes from the global or continental scale (i.e., latitudi­ divergence dating of EMF lineages within nal diversity gradient), but these exceptions Amanita sect. Caesarea (Sánchez‐Ramírez appear to be driven by similar underlying et al., 2015). Based on the timing and geo­ principles. For example, low diversity of graphic location of divergences, Sánchez‐ EMF in some tropical forests may be caused, Ramírez et al. were able to reconstruct an in part, by low host abundance and species overland migration route out of tropical area effects. Identifying these exceptions Africa ≈ 20–16 MYA, migration through may help elucidate the unique ways in southeast Asia to east Asia 8–4 MYA, and which EMF physiology responds to climate then migration across Beringia into the pacific or environment, and shed light into the northwest of N. America around 3 MYA. fundamental nature of this symbiosis. 356 Molecular mycorrhizal symbiosis

While much has been learned up to this or long‐distance migration? In addition, point, this is still an exciting time for study­ many c­lassic biogeographic patterns have yet ing EMF biogeography. We are in the lucky to be examined through the lens of EMF. For position of having to cope with a flood of example, to what degree do EMF respect incoming molecular data from all parts of traditional biogeographic barriers, such as the globe, and a variety of data streams Wallace’s line, the deep channel between (environmental samples, mushroom collec­ continental shelves that sharply demarcates tions, genomes). There are a number of the fauna of southeast Asian origin from emerging challenges that could be addressed those of Australian origin? with these data. Finally, biogeography has an important First, while biogeographic regions have role to play in the current era. In particular, been primarily delimited using macroorgan­ our understanding of the biogeographic isms, there is no comparable picture for soil history of EMF will help to identify inva­ fungi. The abundance of environmental sive species more quickly, and also to sequence data could quite feasibly be used understand the importance of EMF in to delineate major biogeographic regions for mediating the invasion of exotic tree EMF (Kreft and Jetz, 2010). species.­ While “microbes” are not generally Second, despite the early interest in at the top of conservation concerns, the Gondwanan fungi expressed in the pre‐ understanding that most EMF have spa­ molecular era, there still have been only a tially restricted ranges, and that loss of host handful of molecular EMF biogeography area could reduce their diversity, suggest studies, and there is a paucity of data exam­ that native EMF diversity may require con­ ining fungi with purported Gondwanan ori­ servation action to protect. These fungi gins. In addition, perhaps an inordinate have much to teach us about the history of amount of research has been focused on lin­ our planet and, with the new eyes provided eages of EMF that are edible or have human by modern molecular tools, all we need to cultural significance (e.g., Tuber, Boletus edu­ do is look. lis, Tricholoma matsutake, Amanita caesarea and Amanita muscaria). Biogeography of putatively old, dominant EMF lineages, such 19.9 References as the Thelephoraceae or Russulaceae, would shed a great deal of light into the Amend A, Garbelotto M, Fang Z and Keeley S. o­rigins and evolutionary history of the (2010). Isolation by landscape in populations of a symbiosis. prized edible mushroom Tricholoma matsutake. Conservervation Genetics 11(3), 1572–9737. As data from new lineages accumulate, it Averill C, Turner BL and Finzi AC. (2014). will be possible to search more rigorously Mycorrhiza‐mediated competition between plants for general patterns. How does the symbi­ and decomposers drives soil carbon storage. Nature otic lifestyle shape the biogeography of 505(7484), 543–5. EMF? How do these patterns differ from Avise JC, Arnold J, Ball RM, Bermingham E, Lamb T, saprotrophic fungal lineages that may be Neigel JE, Reeb CA and Saunders NC. (1987). Intraspecific phylogeogrpahy: the mitochondrial older or less constrained by host associa­ DNA bridge between population genetics and sys­ tions? What ecological features predispose tematics. Annual Review of Ecology and Systematics particular EMF to long‐range spore dispersal 18, 489–522. Chapter 19: The biogeography of ectomycorrhizal fungi – a history of life in the subterranean 357

Beheregaray LB. (2008). Twenty years of phylogeo­ Clemmensen KE, Bahr A, Ovaskainen O, Dahlberg graphy: the state of the field and the challenges A, Ekblad A, Wallander H, Stenlid J, Finlay RD, for the Southern Hemisphere. Molecular Ecology Wardle DA and Lindahl BD. (2013). Roots and 17(17), 3754–74. associated fungi drive long‐term carbon sequestration Berkeley MJ. (1863). Gardeners’ Chronicle & Agricul­ in boreal forest. Science 339(6127), 1615–8. tural Gazette, London. Corner EJH. (1972). Boletus in Malaysia. Singapore: Bidartondo MI, Bruns TD, et al. (2008). Preserving Lim Bian Han, Government Printer. accuracy in GenBank. Science 319(5870), Darwin C. (1859). On the origin of species by means of 1616–1616. natural selection, or the preservation of favoured races in Bisby GR. (1943). Geographical Distribution of the struggle for life. Fungi. Botanical Review 9(7), 466–482. Davison J, Moora M, Opik M, Adholeya A, Ainsaar L, Bonito G, Smith ME, et al. (2013). Historical biogeog­ Ba A, Burla S, Diedhiou AG, Hiiesalu I, Jairus T, raphy and diversification of truffles in the Johnson NC, Kane A, Koorem K, Kochar M, Tuberaceae and their newly identified southern Ndiaye C, Partel M, Reier U, Saks U, Singh R, hemisphere sister lineage. PLoS One 8(1), e52765. Vasar M and Zobel M. (2015). Global assess­ment of Bougher NL, Fuhrer BA and Horak, E. (1994). arbuscular mycorrhizal fungus diversity reveals Taxonomy and biogeography of Australian very low endemism. Science 349(6251), 970–973. Rozites species mycorrhizal with Nothofagus de Candolle A. (1820). Essai élémentaire de géogra­ and Myrtaceae. Australian Systematic Botany 7, phie botanique. In: Lomolino MV, Sax DF and 353–375. Brown JH (Eds.). Foundations of biogeography, pp. Boyce CK, Hotton CL, Fogel ML, Cody GD, Hazen 28–48. Chicago: University of Chicago Press. RM, Knoll AH and Hueber FM. (2007). Devonian de Queiroz A. (2005). The resurrection of oceanic landscape heterogeneity recorded by a giant fun­ dispersal in historical biogeography. Trends in gus. Geology 35(5), 399. Ecology & Evolution 20(2), 68–73. Bruns TD. (1995). Thoughts on the processes that de Queiroz A. (2014). The monkey’s voyage: how maintain local species‐diversity of ectomycorrhizal improbable journeys shaped the history of life. New fungi. Plant and Soil 170(1), 63–73. York: Basic Books. Bruns TD and Shefferson RP. (2004). Evolutionary Dentinger BT, Ammirati JF, Both EE, Desjardin DE, studies of ectomycorrhizal fungi: recent advances Halling RE, Henkel TW, Moreau PA, Nagasawa E, and future directions. Canadian Journal of Botany Soytong K, Taylor AF, Watling R, Moncalvo JM 82(8), 1122–1132. and McLaughlin DJ. (2010). Molecular phyloge­ Bruns TD, White TJ and Taylor JW. (1991). Fungal netics of porcini mushrooms (Boletus section molecular systematics. Annual Review of Ecology and Boletus). Molecular Phylogenetics and Evolution 57(3), Systematics 22, 525–564. 1276–92. Bruns TD, Peay KG, Boynton PJ, Grubisha LC, Dickie IA and Reich PB. (2005). Ectomycorrhizal Hynson NA, Nguyen NH and Rosenstock NP. fungal communities at forest edges. Journal of (2009). Inoculum potential of Rhizopogon spores Ecology 93(2), 244–255. increases with time over the first four years of a Falkowski PG, Fenchel T and Delong EF. (2008). The 99‐year spore burial experiment. New Phytologist microbial engines that drive Earth’s biogeochemi­ 181(2), 463–470. cal cycles. Science 320(5879), 1034–9. Buée M, Reich M, Murat C, Morin E, Nilsson RH, Finlay BJ. (2002). Global dispersal of free‐living Uroz S and Martin F. (2009). 454 Pyrosequencing microbial eukaryote species. Science 296(5570), analyses of forest soils reveals an unexpectedly 1061–1063. high fungal diversity. New Phytologist 184, Galante TE, Horton TR and Swaney DP. (2011). 95% 449–456. of basidiospores fall within one meter of the cap– a Cain ML, Bowman WD and Hacker SD. (2014). field and modeling based study. Mycologia 103, Ecology. Sinauer Associates, Inc, 596 pp. 1175–1183. Chapela IH and Garbelotto M. (2004). Phylo­ Gardes M and Bruns T. (1993). ITS primers with enh­ geography and evolution in matsutake and close anced specificity for basidiomycetes – application allies inferred by analyses of ITS sequences and to the identification of mycorrhizae and rusts. AFLPs. Mycologia 96(4), 730–741. Molecular Ecology 2, 113–118. 358 Molecular mycorrhizal symbiosis

Geml J. (2011). Coalescent analyses reveal contra­ into the black‐box. Molecular Ecology 10(8), sting patterns of intercontinental gene flow in arctic‐­ 1855–1871. alpine and boreal‐temperate fungi. In: Fontaneto D Hosaka K, Castellano MA and Spatafora JW. (2008). (Ed.). Biogeography of Microscopic Organisms: Is Biogeography of Hysterangiales (Phallomycetidae, Everything Small Everywhere? New York: Cambridge Basidiomycota). Mycological Research 112(Pt 4), University Press. 448–62. Geml J, Laursen GA, O’Neill K, Nusbaum HC and Hughes KW, Petersen RH and Lickey EB. (2009). Taylor DL. (2006). Beringian origins and cryptic Using heterozygosity to estimate a percentage speciation events in the fly agaric (Amanita mus­ DNA sequence similarity for environmental caria). Molecular Ecology 15(1), 225–239. s­pecies’ delimitation across basidiomycete fungi. Geml J, Timling I, Robinson CH, Lennon N, Nusbaum New Phytologist 182(4), 795–798. HC, Brochmann C, Noordeloos ME and Taylor DL. Janzen DH. (1968). Host plants as islands in evolu­ (2011). An arctic community of symbiotic fungi tionary and contemporary time. American assembled by long‐distance dispersers: phyloge­ Naturalist 104, 501–528. netic diversity of ectomycorrhizal basidiomycetes Jeandroz S, Murat C, Wang Y and Le Tacon F. (2008). in Svalbard based on soil and sporocarp DNA. Molecular phylogeny and historical biogeography Journal of Biogeography 39, 74–88. of the genus Tuber. Journal of Biogeography 35, Green JL, Holmes AJ, Westoby M, Oliver I, Briscoe 815–829. D, Dangerfield M, Gillings M and Beattie AJ. Kennedy PG, Garibay‐Orijel R, Higgins LM and (2004). Spatial scaling of microbial eukaryote Angeles‐Arguiz R. (2011). Ectomycorrhizal diversity. Nature 432, 747–750. fungi in Mexican Alnus forests support the host Halling RE and Ovrebo CL. (1987). A new species of co‐migration hypothesis and continental‐scale Rozites from oak forests of Colombia, with notes on p­atterns in phylogeography. Mycorrhiza 21(6), biogeography. Mycologia 79(5), 674–678. 559–68. Harrower E, Bougher NL, Henkel TW, Horak E and Kennedy PG, Matheny PB, Ryberg KM, Henkel TW, Matheny PB. (2015). Long‐distance dispersal and Uehling JK and Smith ME. (2012). Scaling up: speciation of Australasian and American species of examining the macroecology of ectomycorrhizal Cortinarius sect. Cortinarius. Mycologia 107(4), fungi. Molecular Ecology 21, 4151–4154. 697–709. Koele N, Dickie IA, Oleksyn J, Richardson SJ and Henkel TW. (2003). Monodominance in the ectomyc­ Reich PB. (2012). No globally consistent effect of orrhizal Dicymbe corymbosa (Caesalpiniaceae) from ectomycorrhizal status on foliar traits. The New Guyana. Journal of Tropical Ecology 19, 417–437. Phytologist 196(3), 845–52. Hibbett DS and Matheny PB. (2009). The relative Kohler A, Kuo A, Nagy LG, Morin E, Barry KW, ages of ectomycorrhizal mushrooms and their Buscot F, Canback B, Choi C, Cichocki N, Clum A, plant hosts estimated using Bayesian relaxed Colpaert J, Copeland A, Costa MD, Dore J, Floudas D, molecular clock analyses. BMC Biology 7(13). Doi: Gay G, Girlanda M, Henrissat B, Herrmann S, Hess 10.1186/1741‐7007‐7‐13. J, Hogberg N, Johansson T, Khouja HR, LaButti K, Hibbett DS, Grimaldi D and Donoghue MJ. (1995). Lahrmann U, Levasseur A, Lindquist EA, Lipzen A, Cretaceous mushrooms in amber. Nature 377, Marmeisse R, Martino E, Murat C, Ngan CY, 487. Nehls U, Plett JM, Pringle A, Ohm RA, Perotto S, Higgins SI, Nathan R and Cain ML. (2003). Are long‐ Peter M, Riley R, Rineau F, Ruytinx J, Salamov A, distance dispersal events in plants usually caused Shah F, Sun H, Tarkka M, Tritt A, Veneault‐ by nonstandard means of dispersal. Ecology 84, Fourrey C, Zuccaro A, Mycorrhizal Genomics 1945–1956. Initiative C, Tunlid A, Grigoriev IV, Hibbett DS and Horak E. (1981). Notes on taxonomy and biogeogra­ Martin F. (2015). Convergent losses of decay phy of Rozites Karsten. Sydowia 34, 94–108. mechanisms and rapid turnover of symbiosis Horner‐Devine MC, Lage M, Hughes JB and genes in mycorrhizal mutualists. Nature Genetics Bohannan BJM. (2004). A taxa‐area relationship 47(4), 410–5. for bacteria. Nature 432(7018), 750–753. Kõljalg U, Nilsson RH, Abarenkov K, Tedersoo L, Horton TR and Bruns TD. (2001). The molecular Taylor AFS, Bahram M, Bates ST, Bruns TD, revolution in ectomycorrhizal ecology: peeking Bengtsson‐Palme J, Callaghan TM, Douglas B, Chapter 19: The biogeography of ectomycorrhizal fungi – a history of life in the subterranean 359

Drenkhan T, Eberhardt U, Dueñas M, Grebenc T, relation to postglacial re‐colonization routes of the Griffith GW, Hartmann M, Kirk PM, Kohout P, Perigord truffle Tuber melanosporum. New Phytologist Larsson E, Lindahl BD, Lücking R, Martín MP, 164(2), 401–411. Matheny PB, Nguyen NH, Niskanen T, Oja J, Nara K. (2009). Spores of ectomycorrhizal fungi: Peay KG, Peintner U, Peterson M, Põldmaa K, ecological strategies for germination and dormancy. Saag L, Saar I, Schüßler A, Scott JA, Senés C, New Phytologist 181(2), 245–248. Smith ME, Suija A, Taylor DL, Telleria MT, Weiss Nara K, Nakaya H and Hogetsu T. (2003). M and Larsson K‐H. (2013). Towards a unified Ectomycorrhizal sporocarp succession and pro­ paradigm for sequence‐based identification of duction during early primary succession on Mount fungi. Molecular Ecology 22(21), 5271–5277. Fuji. New Phytologist 158(1), 193–206. Kreft H and Jetz W. (2010). A framework for delin­ Nguyen NH, Landeros F, Garibay‐Orijel R, Hansen K eating biogeographical regions based on species and Vellinga EC. (2013). The Helvella lacunosa distributions. Journal of Biogeography 37(11), ­species complex in western North America: cryptic 2029–2053. species, misapplied names and parasites. Mycologia Li DW. (2005). Release and dispersal of basidiospores 105(5), 1275–86. from Amanita muscaria var. alba and their infiltra­ Norros V, Penttilä R, Suominen M and Ovaskainen O. tion into a residence. Mycological Research 109, (2012). Dispersal may limit the occurrence of spe­ 1235–1242. cialist wood decay fungi already at small spatial Lomolino MV, Riddle BR and Brown JH. (2006). scales. Oikos 121(6), 961–974. Biogeography. Sunderland: Sinauer. Oda T, Tanaka C and Tsuda M. (2004). Molecular Lumbsch HT, Buchanan PK, May TW and Mueller phylogeny and biogeography of the widely distrib­ GM. (2008). Phylogeography and biogeography of uted Amanita species, A. muscaria and A. panthe­ fungi. Mycological Research 112, 423–424. rina. Mycological Research 108, 885–896. MacArthur RH and Wilson EO. (1967). The Theory Peay K, Bruns T, Kennedy P, Bergemann S and of Island Biogeography. Princeton: Princeton Garbelotto M. (2007). A strong species – area rela­ University Press. tionship for eukaryotic soil microbes: island size Matheny PB, Aime MC, Bougher NL, Buyck B, matters for ectomycorrhizal fungi. Ecology Letters Desjardin DE, Horak E, Kropp BR, Lodge DJ, 10, 470–480. Soytong K, Trappe JM and Hibbett DS. (2009). Peay K, Garbelotto M and Bruns T. (2009). Spore Out of the Palaeotropics? Historical biogeography heat resistance plays an important role in distur­ and diversification of the cosmopolitan ectomyc­ bance mediated assemblage shift of ectomycorrhi­ orrhizal mushroom family Inocybaceae. Journal of zal fungi colonizing Pinus muricata seedlings. Biogeography 36(4), 577–592. Journal of Ecology 97, 537–547. Meiser A, Balint M and Schmitt I. (2014). Meta‐ Peay K, Garbelotto M and Bruns T. (2010a). Evidence analysis of deep‐sequenced fungal communities of dispersal limitation in soil microorganisms:­ indicates limited taxon sharing between studies Isolation reduces species richness on mycorrhizal and the presence of biogeographic patterns. The tree islands. Ecology 91(12), 3631–3640. New Phytologist 201(2), 623–635. Peay K, Kennedy P, Davies S, Tan S and Bruns T. Miyamoto Y, Nakano T, Hattori M and Nara K. (2010b). Potential link between plant and fungal (2014). The mid‐domain effect in ectomycorrhizal distributions in a dipterocarp rainforest: commu­ fungi: range overlap along an elevation gradient nity and phylogenetic structure of tropical ecto­ on Mount Fuji, Japan. The ISME Journal 8(8), mycorrhizal fungi across a plant and soil ecotone. 1739–46. New Phytologist 185(2), 529–542. Moncalvo JM and Buchanan PK. (2008). Molecular Peay K, Schubert M, Nguyen N and Bruns T. (2012). evidence for long distance dispersal across the Measuring ectomycorrhizal fungal dispersal: mac­ Southern Hemisphere in the Ganoderma applana­ roecological patterns driven by microscopic prop­ tum‐australe species complex (Basidiomycota). agules. Molecular Ecology 16, 4122–4136. Mycological Research 112, 425–436. Peay KG, Bidartondo MI and Arnold AE. (2010c). Not Murat C, Díez J, Luis P, Delaruelle C, Dupré C, every fungus is everywhere: scaling to the biogeogra­ Chevalier G, Bonfante P and Martin F. (2004). phy of fungal‐plant interactions across roots, shoots Polymorphism at the ribosomal DNA ITS and its and ecosystems. New Phytologist 185, 878–882. 360 Molecular mycorrhizal symbiosis

Peintner U, Horak E, Moser MM and Vigalys R. Proceedings of the National Academy of Sciences of the (2002). Phylogeny of Rozites, Cuphocybe and United States of America 109(16), 6241–6. Rapacea inferred from ITS and LSU rDNA Skrede I, Engh IB, Binder M, Carlsen T, Kauserud H sequences. Mycologia 94(4), 620–629. and Bendiksby M. (2011). Evolutionary history of Pellissier L, Niculita‐Hirzel H, Dubuis A, Pagni M, Serpulaceae (Basidiomycota): molecular phylog­ Guex N, Ndiribe C, Salamin N, Xenarios I, Goudet eny, historical biogeography and evidence for a J, Sanders IR and Guisan A. (2014). Soil fungal single transition of nutritional mode. BMC communities of grasslands are environmentally Evolutionary Biology 11, 230. structured at a regional scale in the Alps. Molecular Smith D and Peay, K. (2014). Sequence depth, not PCR Ecology 23(17), 4274–90. replication, improves ecological inference from Next Pirozynski KA. (1983). Pacific mycogeography: Generation DNA Sequencing PLoS One 92, e90234. an appraisal. Australian Journal of Botany 10, Smith ME, Douhan GW and Rizzo DM. (2007). 137–159. Intra‐specific and intra‐sporocarp ITS variation of Polme S, Bahram M, Yamanaka T, Nara K, Dai YC, ectomycorrhizal fungi as assessed by rDNA Grebenc T, Kraigher H, Toivonen M, Wang PH, sequencing of sporocarps and pooled ectomycor­ Matsuda Y, Naadel T, Kennedy PG, Koljalg U rhizal roots from a Quercus woodland. Mycorrhiza and Tedersoo L. (2013). Biogeography of ecto­ 18(1), 15–22. mycorrhizal fungi associated with alders (Alnus Smith ME, Henkel T, Aime MC, Fremier AK and spp.) in relation to biotic and abiotic variables Vilgalys R. (2011). Ectomycorrhizal fungal diversity at the global scale. New Phytologist 198, and community structure on three co‐occurring 1239–1249. leguminous canopy tree species in a Neotropical Pringle A and Vellinga EC. (2006). Last chance to rainforests. New Phytologist 192, 699–712. know? Using literature to explore the biogeogra­ Smith ME, Henkel TW, Uehling JK, Fremier AK, phy and invasion biology of the death cap mush­ Clarke HD and Vilgalys R. (2013). The ectomycor­ room Amanita phalloides (Vaill. Ex fr. : Fr.) link. rhizal fungal community in a Neotropical for­ Biological Invasions 8(5), 1131–1144. est dominated by the endemic dipterocarp Read DJ. (1991). Mycorrhizas in ecosystems. Pakaraimaea dipterocarpacea. PLoS One 8(1), e55160. Experientia 47(4), 376–391. Springer MS, Meredith RW, Janecka JE and Murphy Ricklefs RE. (1987). Community diversity: relative WJ. (2011). The historical biogeography of roles of local and regional processes. Science Mammalia. Philosophical Transactions of the Royal 235(4785), 167–71. Society B: Biological Sciences 366(1577), 2478–502. Ryberg M and Matheny PB. (2011). Asynchronous Straatsma G, Ayer F and Egli S. (2001). Species rich­ origins of ectomycorrhizal clades of Agaricales. ness, abundance, and phenology of fungal fruit Proceedings of the Royal Society B: Biological Sciences bodies over 21 years in a Swiss forest plot. 279(1735), 2003–2011. Mycological Research 105(5), 515–523. Sánchez‐Ramírez S, Tulloss RE, Amalfi M, Moncalvo Talbot JM, Bruns TD, Taylor JW, Smith DP, Branco S, J‐M and Carine M. (2015). Palaeotropical origins, Glassman SI, Erlandson S, Vilgalys R, Liao HL, boreotropical distribution and increased rates of Smith ME and Peay KG. (2014). Endemism and diversification in a clade of edible ectomycorrhizal functional convergence across the North American mushrooms (Amanita section Caesarea). Journal of soil mycobiome. Proceedings of the National Academy Biogeography 42(2), 351–363. of Sciences of the United States of America 111(17), Sato H, Tsujino R, Kurita K, Yokoyama K and 6431–6346. Agata K. (2012). Modelling the global distribu­ Taylor DL, Hollingsworth TN, McFarland JW, Lennon tion of fungal species: new insights into micro­ NJ, Nusbaum C and Ruess RW. (2014). A first bial cosmopolitanism. Molecular Ecology 21(22), comprehensive census of fungi in soil reveals both 5599–612. hyperdiversity and fine‐scale niche partitioning. Schoch CL, Seifert KA, Huhndorf S, Robert V, Spouge Ecological Monographs 84(1), 3–20. JL, Levesque CA and Chen W. (2012). Nuclear Tedersoo L and Nara K. (2010). General latitudinal ribosomal internal transcribed spacer (ITS) region gradient of biodiversity is reversed in ectomycor­ as a universal DNA barcode marker for Fungi. rhizal fungi. New Phytologist 185, 351–354. Chapter 19: The biogeography of ectomycorrhizal fungi – a history of life in the subterranean 361

Tedersoo L, Suvi T, Beaver K and Kõljalg U. (2007). comparative phylogenetic analyses. Molecular Ectomycorrhizal fungi of the Seychelles: diver­ Ecology 23(16), 4168–83. sity patterns and host shifts from the native Toljander JF, Eberhardt U, Toljander YK, Paul LR and Vateriopsis seychellarum (Dipterocarpaceae) and Taylor AFS. (2006). Species composition of an Intsia bijuga (Caesalpiniaceae) to the introduced ectomycorrhizal fungal community along a local Eucalyptus robusta (Myrtaceae), but not Pinus nutrient gradient in a boreal forest. New Phytologist ­caribea (Pinaceae). New Phytologist 175(2), 170(4), 873–883. 321–333. Wallace AR. (1855). On the law which has regulated Tedersoo L, May TW and Smith ME. (2010a). the introduction of new species. Annals and Ectomycorrhizal lifestyle in fungi: global diversity, Magazine of Natural History 16, 184–196. distribution, and evolution of phylogenetic line­ Watling R. (2001). The relationships and possible ages. Mycorrhiza 20(4), 217–263. distributional patterns of boletes in south‐east Tedersoo L, Sadam A, Zambrano M, Valencia R and Asia. Mycological Research 105(12), 1440–1448. Bahram M. (2010b). Low diversity and high host Wilson AW, Binder M and Hibbett DS. (2012). preference of ectomycorrhizal fungi in western Diversity and evolution of ectomycorrhizal Amazonia, a neotropical biodiversity hotspot. The host associations in the Sclerodermatineae ISME Journal 4(4), 465–71. (Boletales, Basidiomycota). New Phytologist Tedersoo L, Bahram M, et al. (2014a). Global diver­ 194(4), 1079–95. sity and geography of soil fungi. Science 346(6213), Woese CR. (1987). Bacterial Evolution. Microbiological 1256688. Reviews 51(2), 221–271. Tedersoo L, Bahram M, Ryberg M, Otsing E, Wolfe BE, Tulloss RE and Pringle A. (2012). The irre­ Koljalg U and Abarenkov K. (2014b). Global versible loss of a decomposition pathway marks biogeography of the ectomycorrhizal/sebacina the single origin of an ectomycorrhizal symbiosis. lineage­ (Fungi, Sebacinales) as revealed from PLoS One 7(7), e39597.