Truffles have evolved from epigeous (aboveground) ancestors in nearly every major lineage of fleshy fungi. Because accelerated rates of morphological evolution accompany the transition to the truffle form, closely related epigeous ancestors remain unknown for most truffle lineages. This is the case for the quintessential truffle genus Tuber, which includes species with socio-economic importance and esteemed culinary attributes. Ecologically, Tuber spp. form obligate mycorrhizal symbioses with diverse species of plant hosts including pines, oaks, poplars, orchids, and commercially important trees such as hazelnut and pecan. Unfortunately, limited geographic sampling and inconclusive phylogenetic relationships have obscured our understanding of their origin, biogeography, and diversification. To address this problem, we present a global sampling of Tuberaceae based on DNA sequence data from four loci for phylogenetic inference and molecular dating. Our well-resolved Tuberaceae phylogeny shows high levels of regional and continental endemism. We also identify a previously unknown epigeous member of the Tuberaceae – the South American cup-fungus Nothojafnea thaxteri (E.K. Cash) Gamundí. Phylogenetic resolution was further improved through the inclusion of a previously unrecognized Southern hemisphere sister group of the Tuberaceae. This morphologically diverse assemblage of species includes truffle (e.g. Gymnohydnotrya spp.) and non-truffle forms that are endemic to Australia and South America. Southern hemisphere taxa appear to have diverged more recently than the Northern hemisphere lineages. Our analysis of the Tuberaceae suggests that Tuber evolved from an epigeous ancestor. Molecular dating estimates Tuberaceae divergence in the late Jurassic (∼156 million years ago), with subsequent radiations in the Cretaceous and Paleogene. Intra-continental diversification, limited long-distance dispersal, and ecological adaptations help to explain patterns of truffle evolution and biodiversity.

Summary• The spores of ectomycorrhizal fungi (EMF) play critical roles in the population and community development of EMF. Here, the germination and infectivity of EMF spores are examined with reference to the ecological traits of the EMF species.• Spores were collected from 12 EMF species, whose successional patterns have been studied in the volcanic desert on Mount Fuji, Japan. Spore germination experiments were conducted with host plants (Salix reinii), with nonhost plants (Polygonum cuspidatum), and without plants. The mycorrhizal formation ability of spores was also examined in seven EMF using spore inoculation experiments. To determine the effects of the spore preservation period, both experiments were repeated up to 1 yr after spore collection.• Spore germination was very low in the absence of host plants. In the presence of hosts, even 30 d after spore collection, spore germination was significantly enhanced in all pioneer EMF (c. 20%) but less so in late-stage EMF (< 5%), except in Hebeloma species. Mycorrhizal formation from spores was also greater in pioneer EMF but was significantly reduced by 1 yr of spore preservation.• High spore germination and infectivity of pioneer EMF should enable these species to colonize disturbed and isolated areas in accordance with their ecological traits.

PREMISE OF THE STUDY:Since mycoheterotrophic plants (MHPs) completely depend on their mycorrhizal fungi for carbon, selection of fungal partners has an important role in the speciation of MHPs. However, the causes and mechanisms of mycobiont changes during speciation are not clear. We tested fungal partner shifts and changes in mycorrhizal specifi city during speciation of three closely related MHPs-Gastrodia confusa ( Gc ), G. pubilabiata ( Gp ), and G. nipponica ( Gn ) (Orchidaceae)-and correlations between these changes and the vegetation types where each species grows. METHODS:We investigated the diversity of mycobionts of the three species by sequencing nrDNA ITS, and the sequence data were subjected to test changes in fungal specifi city and fungal partner shifts among the three species. Furthermore, we conducted multivariate analysis to test for diff erences in mycobiont communities of vegetation types where each species grows.KEY RESULTS: Two saprobic Basidiomycota, Marasmiaceae and Mycenaceae, were dominant fungal partners of the three species, and Gn was simultaneously associated with the ectomycorrhizal Russulaceae and Sebacinaceae. Although mycobiont composition diff ered among the three species, they also sometimes shared identical fungal species. Multivariate analysis revealed that mycobiont communities of the three species in bamboo thickets diff ered signifi cantly from those in other vegetation types.CONCLUSIONS: Fungal partner shifts are not necessarily associated with the evolution of MHPs, and fungal specifi city of Gc and Gp was signifi cantly higher than that of Gn , implying that the specifi city fl uctuates during speciation. Further, Gc exclusively inhabits bamboo thickets, which suggests that adaptation to particular fungi specifi c to bamboo thickets triggered speciation of this species.KEY WORDS ectomycorrhizal fungi; Gastrodia ; myceoheterotrophy; saprotrophic fungi; speciation; specifi city 208 • AMERICAN JOURNAL OF BOTANY Bruns, 2001 ), Corallorhiza of Orchidaceae ( Taylor and Bruns, 1999 ), Afrothismia of Th ismiaceae, Botrychium of Ophioglossaceae, Gymnosiphon of Burmanniaceae, Voyria of Gentianaceae, and Sciaphila of Triuridaceae ( Merckx et al., 2012 ). Th e fi rst two studies cited above revealed that fungal partner shift s occurred among closely related species within an MHP genus, whereas the third study confi rmed both patterns: partner shift s occurred in Gymnosiphon but not in the other four MHP genera examined. Th ese results indicate that partner shift s are not indispensable in the diversifi cation of MHPs. However, low sampling densities of plant species in some studies (e.g., Merckx et al., 2012 ) may have led to the diversity of mycobionts having been overlooked-and, thus, partner shift s may not be as common as currently thought. Meanwhile, changes in specifi city levels (i.e., phylogenetic breadth) of mycobionts among closely related MHP species have mostly been compared only for dominant fungal partners-for example, for the Corallorhiza striata complex ( Barrett ...

Host identity is among the most important factors in structuring ectomycorrhizal (ECM) fungal communities. Both host-fungal coevolution and host shifts can account for the observed host effect, but their relative significance in ECM fungal communities is not well understood. To investigate these two host-related mechanisms, we used relict forests of Pseudotsuga japonica, which is an endangered endemic species in Japan. As with other Asian Pseudotsuga species, P. japonica has been isolated from North American Pseudotsuga spp. since the Oligocene and has evolved independently as a warm-temperate species. We collected 100 soil samples from four major localities in which P. japonica was mixed with other conifers and broadleaf trees. ECM tips in the soil samples were subjected to molecular analyses to identify both ECM fungi and host species. While 136 ECM fungal species were identified in total, their communities were significantly different between host groups, confirming the existence of the host effect on ECM fungal communities. None of the 68 ECM fungal species found on P. japonica belonged to Pseudotsuga-specific lineages (e.g., Rhizopogon and Suillus subgroups) that are common in North America. Most of ECM fungi on P. japonica were shared with other host fungi or phylogenetically close to known ECM fungi on other hosts in Asia. These results suggest that after migrating, Pseudotsuga-specific fungal lineages may have become extinct in small isolated populations in Japan. Instead, most of the ECM fungal symbionts on P. japonica likely originated from host shifts in the region.

Fungal partner composition and specificity level changed with speciation in both leafy and leafless Neottia species. In particular, mycorrhizal associations likely shifted from Sebacinales Group B to Group A during the evolution from autotrophy to mycoheterotrophy. Partner shifts to Sebacinales Group A have also been reported in the evolution of mycoheterotrophy of other plant groups, suggesting that convergence to this fungal group occurs in association with the evolution of mycoheterotrophy.