Fungal Genetics and Biology 49 (2012) 455–469
Contents lists available at SciVerse ScienceDirect
Fungal Genetics and Biology
journal homepage: www.elsevier.com/locate/yfgbi
Multigene molecular phylogenetics reveals true morels (Morchella) are especially species-rich in China ⇑ Xi-Hui Du a,c, Qi Zhao a, Kerry O’Donnell b, Alejandro P. Rooney b, Zhu L. Yang a, a Key Laboratory of Biodiversity and Biogeography, Kunming Institute of Botany, Chinese Academy of Sciences, Lanhei Road, No. 132, Kunming, 650201 Yunnan Province, PR China b Bacterial Foodborne Pathogens and Mycology Research Unit, National Center for Agricultural Utilization Research, US Department of Agriculture, Agricultural Research Service, 1815 North University Street, Peoria, IL 61604, United States c Graduate University of Chinese Academy of Sciences, Beijing 100049, PR China article info abstract
Article history: The phylogenetic diversity of true morels (Morchella) in China was estimated by initially analyzing Received 22 December 2011 nuclear ribosomal internal transcribed spacer (ITS) rDNA sequences from 361 specimens collected in Accepted 23 March 2012 21 provinces during the 2003–2011 growing seasons, together with six collections obtained on loan from Available online 6 April 2012 three Chinese herbaria. Based on the results of this preliminary screen, 40 Esculenta Clade (yellow mor- els) and 30 Elata Clade (black morels) were chosen to represent the full range of phylogenetic diversity Keywords: sampled. To investigate their species limits, we generated DNA sequences from portions of three protein- Ascomycota coding genes (RPB1, RPB2 and EF-1a) and domains D1 and D2 of the nuclear large subunit (LSU) rDNA for Biodiversity all 70 collections. To fully assess evolutionary relationships, previously published multilocus DNA Conservation biology DNA sequence sequence data representing all known Morchella species was included in this study. Phylogenetic analyses Fungi employing maximum parsimony and maximum likelihood frameworks resolved 30 species in China com- Range evolution pared with 22 in Europe and 19 within North America. Eleven novel phylogenetically distinct species were discovered in China, including two species within the Elata Clade and nine within the Esculenta Clade. Of the 30 species in China, 20 appear to be endemic, nine were also represented in Europe, and four putatively fire-adapted species have disjunct distributions in China, Europe and western North America. Although the diversification time estimates place the Esculenta Clade in China as early as the late Creta- ceous and the Elata Clade by the early Oligocene, 27 of the 30 species evolved between the middle Mio- cene 12 Mya and present. Ó 2012 Elsevier Inc. All rights reserved.
1. Introduction Although Morchella spp. are easily distinguished from other macrofungi by their sponge-like pileus, morphological species rec- True morels (Morchella spp., phylum Ascomycota) are largely ognition (MSR) within the genus is problematic due to their pheno- restricted to temperate regions of the Northern Hemisphere where typic plasticity, dearth of taxonomically useful characters, and they typically fruit for only a few weeks each spring. Due to their body plan which appears to have remained remarkably static over highly desirable flavor and short fruiting season, morels are among the past 100 million years (O’Donnell et al., 2011). Molecular sys- the world’s most prized edible fungi collected by mycophiles and tematic studies, based on analyses of nuclear ribosomal DNA se- gourmets. To meet the demand created by their growing popular- quence data (Hansen and Pfister, 2006; O’Donnell et al., 1997), ity, wild morels are harvested commercially and exported exten- have confirmed Morchellaceae monophyly as defined by synapo- sively from China, India, Turkey, Mexico, and the United States morphic eguttulate, multinucleate ascospores with a cluster of epi- (Pilz et al., 2007). In China, the annual export of dried morels in- plasmic granules at each pole. Although a number of molecular creased fivefold over the past 5 years to 900,000 kg, averaging systematic studies have been published on Morchella, employing $160 US dollars per kilogram. In addition to the export of commer- diverse molecular markers (reviewed in Pagliaccia et al. (2011)), cially harvested wild morels, efforts to meet the growing demand species limits have only recently been investigated (O’Donnell have included cultivation in a specialized indoor facility using pat- et al., 2011; Tasßkın et al., 2010, 2012) using multilocus DNA se- ented technology within the US (Ower et al., 1986). In addition, quence data and phylogenetic species recognition based on genea- preliminary efforts have been made at growing morels outdoors logical concordance and non-discordance (i.e., GCPSR; Dettman in Yunnan Province, China (Zhao et al., 2009). et al., 2003; Taylor et al., 2000). Collectively, the three published GCPSR studies resolved the Esculenta Clade (yellow morels) and ⇑ Corresponding author. Fax: +86 871 5150227. Elata Clade (black morels) as reciprocally monophyletic sisters, E-mail address: [email protected] (Z.L. Yang). comprising 18 and 31 species, respectively, and a basal monotypic
1087-1845/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.fgb.2012.03.006 456 X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469 sister clade represented by Morchella rufobrunnea, the only species To rigorously investigate species diversity represented by the that has been cultivated commercially indoors (Ower et al., 1986). 73 collections, we sequenced portions of the nuclear large subunit Results of the molecular phylogenetic studies also revealed that (LSU) rDNA, RNA polymerase largest (RPB1) and second largest most Morchella species appear to exhibit continental endemism subunit (RPB2), and translation elongation factor 1-alpha (EF-1a) and provincialism, which has greatly facilitated reconstructing genes. These sequences were added to published multilocus DNA their historical biogeography (O’Donnell et al., 2011). sequence datasets previously published for Morchella (O’Donnell Due to the limitations of MSR, field guides typically recognize et al., 2011; Tasßkın et al., 2010, 2012), and analyzed phylogeneti- fewer than five species and use European names for collections cally employing the criteria of genealogical exclusivity and non- made in Asia (Huang, 1998; Imazeki et al., 1988; Mao, 2000; Ying discordance (Dettman et al., 2003) under GCPSR. Phylogenetic spe- and Zang, 1994; Zang, 1996) and North America (Arora, 1979; We- cies were recognized as genealogically exclusive under GCPSR, if ber, 1988). To date, only four Morchella species and one subspecific they were supported by maximum parsimony (MP) and maximum taxon have been described from China, and these were based likelihood (ML) bootstrapping of one or more partition, and/or the exclusively on MSR (Chen and Liu, 2005; Li et al., 2006; Mou, combined dataset and no partition contradicted their monophyly 1987; Zang, 1987). Given the high level of cryptic speciation and (i.e., non-discordance; Dettman et al., 2003). In the absence of provincialism discovered within Morchella (O’Donnell et al., the ability to test the monophyly of two lineages within the Escu- 2011; Tasßkın et al., 2010, 2012), and the rich floristic diversity lenta Clade (Fig. 2) and five within the Elata Clade (Fig. 3) repre- within China (Liu, 1988; Ying, 2001), we hypothesized that sented by single collections, these seven lineages were GCPSR-based studies of geographically diverse collections of Mor- interpreted as putatively phylogenetically distinct because they chella might result in the discovery of multiple novel species lin- were significantly divergent from and not sympatric with their eages in heretofore unexplored regions of China, especially given putative sisters. that relatively few collections from eastern Asia were included in O’Donnell et al. (2011). Knowledge gained from such a survey is 2.2. DNA isolation essential for developing scientifically informed conservation prac- tices to enhance sustainability of morel harvests (Pilz et al., 2007) Prior to extracting total genomic DNA, all of the Chinese collec- and to advance our understanding of their genetic diversity, evolu- tions were dried over several exchanges of silica gel in tightly tionary relationships and geographic distribution. sealed plastic bags for one to several days. Approximately 10 mg Towards this end, we generated ITS rDNA sequences from 361 of dried pileus tissue was ground to a fine powder in a 1.5 ml collections of Morchella we collected from 21 provinces together microcentrifuge tube using a Kontes pellet pestle (Kaimu, China). with six collections obtained from three Chinese herbaria to ob- Once pulverized, the samples were suspended in 700 ll of CTAB tain an initial estimate of Morchella species diversity in China. extraction buffer (100 mM Tris–Cl pH 8.4, 1.4 M NaCl, 25 mM Based on our molecular phylogenetic analyses of the aligned ITS EDTA, 2% CTAB), and incubated for 1.5–2.0 h at 65 °C, during which rDNA sequences, 70 collections were chosen to represent the ge- time they were gently inverted 3–5 times. After the samples were netic diversity sampled for which portions of four additional nu- cooled to room temperature, 700 ll of chloroform-isoamyl alcohol clear genes were obtained. Phylogenetic analyses were conducted (24:1) was added to each tube. The mixture was vortexed briefly, on the individual and combined datasets, which included se- centrifuged at 12,000g for 10 min, and then 500 ll of the upper quences from Tasßkın et al. (2010, 2012) and O’Donnell et al. phase was carefully transferred to a new 1.5 ml microcentrifuge (2011) to: (i) investigate species diversity and geographic distri- tube. After a second chloroform-isoamyl alcohol (24:1) extraction bution of Morchella in China using GCPSR; (ii) assess for the first was performed, the supernatant was transferred to a new 1.5 ml time the utility of ITS rDNA sequence data for species-level phy- microcentrifuge tube and an equal volume of 100% isopropanol logenetics within the genus; (iii) estimate divergence times of at 20 °C was added to each tube. The tube contents were mixed Morchella species lineages in China; and (iv) advance our under- briefly by inversion to obtain a homogeneous solution and then standing of the global historical biogeography and range evolu- they were stored overnight at 20 °C to precipitate total genomic tion of Morchella. DNA. After the tubes were warmed to room temperature they were centrifuged at 12,000g for 10 min and the supernatant was dis- carded. The DNA pellet was washed consecutively with 70% and 2. Methods 100% ethanol, air-dried and then resuspended in 100 ll of sterile
double distilled H2O. All genomic DNA samples were stored frozen 2.1. Collections of Morchella studied at 20 °C until ready for use.
An initial estimate of Morchella species diversity in China was 2.3. PCR amplification and sequencing obtained by analyzing DNA sequence data from the nuclear ribo- somal internal transcribed spacer (ITS) rDNA region to screen PCR and DNA sequencing primers are listed in Supplemental 361 collections we made between 2003 and 2011 in 21 provinces Table S1. Each PCR reaction contained 1 ll of 20 ng/ll genomic together with six collections obtained from three Chinese herbaria DNA, 2.5 llof10 PCR reaction buffer, 0.5 ll dNTP mix (10 mmol), (Fig. 1 and Table 1). In addition, one Esculenta Clade collection 2 ll each of primer (5 lmol), 1.5 ll bovine serum albumin (20 mg/ from Israel and two Elata Clade collections from Germany were in- ml) and 1.5 U of Taq DNA polymerase (Biomed, China). The final cluded in this study. Unfortunately, efforts to obtain DNA sequence volume was adjusted to 25 ll with sterile distilled H2O. PCRs were data from the holotypes of Morchella bicostata and Morchella tibeti- conducted in an Applied Biosystems 2720 thermocycler (ABI, Fos- ca were unsuccessful; types of the three other morels described ter City, CA), using the following cycling parameters: 94 °C for from Chinese collections were unavailable for study. Based on 3 min, 35 cycles of 94 °C for 1 min, 50 °C for 30 s, 72 °C for 1 min, the results of this initial screen, 70 collections from China, two col- followed by a final extension of 10 min at 72 °C. Amplicons were lections from Germany and one from Israel were chosen to repre- electrophoresed in 1.2% agarose in 1 TAE, stained with Gold- sent the full range of genetic diversity represented in the 367- View™ (Guangzhou Geneshun Biotech Ltd., Guangdong, China), specimen ITS rDNA dataset (Table 1). All of the Chinese collections and then photographed over a ultraviolet transilluminator. PCR made in the present study are housed in the Herbarium of Crypto- products were purified using a Bioteke DNA Purification Kit (Bio- gams, Kunming Institute of Botany (HKAS). teke Corporation, Beijing, China), sequenced with ABI BigDye ver. X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469 457
Fig. 1. Map of China identifying the 21 provinces (bold font) where members of the Esculenta Clade (yellow morels = N) and Elata Clade (black morels = d) were collected. The dashed line and roman numerals identify the following four floristic subkingdoms modified from Wu and Wu (1996): (I) Eurasian, (II) Sino-Himalayan, (III) Sino-Japanese, and (IV) Malesian.
Table 1 Data for 75 collections analyzed via multilocus phylogenetics.
Speciesa Voucherb Locality GPS coordinate Dominant vegetation Month/year collected Mel-6 HKAS 62872 Yunnan, China 26°460N–99°380E Pinus yunnanensis (burned area) 07/2007 Mel-7 HKAS62863 Yunnan, China 26°460N–99°380E Pinus yunnanensis (burned area) 07/2007 Mel-7 HKAS62864 Yunnan, China 26°460N–99°380E Pinus yunnanensis (burned area) 07/2007 Mel-9 HKAS62865 Yunnan, China 26°440N–99°480E Pinus yunnanensis (burned area) 07/2007 Mel-9 HKAS62866 Yunnan, China 26°440N–99°480E Pinus yunnanensis (burned area) 07/2007 Mel-9 HKAS62867 Yunnan, China 26°440N–99°480E Pinus yunnanensis (burned area) 07/2007 Mel-10 HKAS62868 Yunnan, China 27°100N–99°270E Populus botanii 04/2007 Mel-10 HKAS62869 Yunnan, China 27°100N–99°270E Populus botanii 04/2007 Mel-10 HKAS62870 Germany Unknown Unknown 04/2010 Mel-10 HKAS62871 Germany Unknown Unknown 04/2010 Mel-13 HKAS62890 Shanxi, China 37°490N–111°210E–1619 m Populus sp., Betula platyphylla, Potentilla fruticosa 05/2009 Mel-13 HKAS62887 Yunnan, China 26°380N–99°490E Populus sp., Quercus sp. 04/2004 Mel-13 HKAS62888 Sichuan, China 31°520N–102°370E–3358 m Picea sp., Abies sp. 05/2009 Mel-13 HKAS62889 Sichuan, China 33°150N–105°530E Picea sp., Abies sp. 04/2004 Mel-13 HKAS62891 Xinjiang, China 44°220N–87°130E Picea sp., Abies sp. 06/2009 Mel-13 HKAS62892 Shanxi, China 37°490N–111°210E–1619 m Populus sp., Betula platyphylla, Potentilla fruticosa 05/2009 Mel-13 HKAS62893 Yunnan, China 27°130N–99°270E Populus sp., Quercus sp. 04/2005 Mel-13 HKAS62894 Shaanxi, China 35°520N–109°10E Populus sp., Robinia sp. 04/2004 Mel-14 HKAS62885 Sichuan, China 31°520N–102°370E–3358 m Picea sp., Abies sp. 05/2009 Mel-14 HKAS62886 Sichuan, China 31°550N–102°390E–3407 m Picea sp., Abies sp. 05/2009 Mel-16 HKAS62883 Jinlin, China 42°110N–128°100E–1152 m Picea sp., Populus sp. 05/2010 Mel-16 HKAS62884 Jinlin, China 42°110N–128°100E–1152 m Picea sp., Populus sp. 05/2010 Mel-19 HKAS62873 Gansu, China 33°450N–104°110E Coriaria nepalensis, Artemisia annua 04/2008
(continued on next page) 458 X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469
Table 1 (continued)
Speciesa Voucherb Locality GPS coordinate Dominant vegetation Month/year collected Mel-19 HKAS62875 Sichuan, China 31°550N–102°390E–3407 m Picea sp., Abies sp. 05/2009 Mel-20 HKAS62876 Yunnan, China 28°10N–99°460E–3128 m Populus botanii, Podophyllum hexandrum 05/2010 Mel-21 HKAS62878 Hubei, China 29°560N–109°150E–1425 m Rhododendron sp. 03/2010 Mel-21 HKAS62879 Sichuan, China 32°380N–105°180E Populus sp., Quercus sp. 05/2007 Mel-21 HKAS62880 Sichuan, China 32°380N–105°180E Populus sp., Quercus sp. 05/2007 Mel-31 HKAS62881 Gansu, China 33°230N–105°380E Artemisia annua, Populus sp. 04/2004 Mel-31 HKAS62882 Yunnan, China Unknown Unknown 05/2010 Mel-33 HKAS62874 Gansu, China 33°450N–104°110E Coriaria nepalensis, Artemisia annua 04/2008 Mel-34 HKAS62877 Yunnan, China Unknown Populus botanii, Rhododendron sp. 05/2010 Mes-6 HKAS25626 Heilongjiang, China Unknown Unknown 1992 Mes-6 HKAS56601 Shanxi, China 37°590N–111°420E Populus sp. 05/2009 Mes-6 HKAS59162 Liaoning, China 38°520N–116°10E–15 m Populus sp. 05/2010 Mes-6 HKAS59163 Liaoning, China 38°520N–116°10E–15 m Populus sp. 05/2010 Mes-6 HMJAU5334 Jilin, China 43°480N–125°540E Populus sp. 05/2007 Mes-6 HMJAU5454 Jilin, China 43°480N–125°540E Populus sp., Quercus sp. 05/2007 Mes-8 HKAS56676 Heilongjiang, China 44°500N–128°280E Populus sp. 05/2009 Mes-8 HKAS59167 Liaoning, China 40°400N–125°220E–214 m Larix sp., Betula sp., Pteridium sp. 05/2010 Mes-8 HKAS59168 Liaoning, China 40°400N–125°220E–214 m Larix sp., Betula sp., Pteridium sp. 05/2010 Mes-9 HKAS59118 Shandong, China 36°170N–117°90E–207 m Populus sp. 04/2010 Mes-9 HKAS59121 Shandong, China 36°170N–117°90E–207 m Populus sp. 04/2010 Mes-9 HKAS59123 Shandong, China 36°170N–117°90E–207 m Populus sp. 04/2010 Mes-9 HKAS59124 Shandong, China 36°170N–117°90E–207 m Populus sp. 04/2010 Mes-9 HKAS59128 Shandong, China 36°170N–117°90E–207 m Populus sp. 04/2010 Mes-10 HKAS25365 Yunnan, China Unknown Unknown 1992 Mes-10 HKAS59141 Yunnan, China Unknown Unknown 05/2010 Mes-13 HKAS25357 Yunnan, China Unknown Unknown 1992 Mes-13 HKAS55922 Shaanxi, China 35°530N–109°20E Robinia pseudoacacia, Populus simonii 05/2008 Mes-15 HKAS55894 Sichuan, China 31°310N–102°30E Populus sp. 04/2008 Mes-15 HKAS62913 Yunnan, China 23°20N–103°310E–2045 m Populus sp. 05/2011 Mes-15 HKAS62914 Yunnan, China 23°20N–103°310E–2045 m Populus sp. 05/2011 Mes-16 HKAS55839 Yunnan, China 27°120N–99°260E Populus sp., Quercus sp. 06/2003 Mes-16 HKAS55840 Yunnan, China 27°120N–99°260E Populus sp., Quercus sp. 06/2003 Mes-16 HAI-D-041 Israel Unknown Unknown 06/1997 Mes-19 HKAS55910 Shaanxi, China 33°110N–106°270E Catalpa ovata 04/2008 Mes-19 HKAS56568 Chongqing, China 29°20N–107°160E Bamboo 04/2009 Mes-19 HKAS56585 Henan, China Unknown Unknown 05/2009 Mes-20 HKAS55841 Yunnan, China 27°120N–99°260E Populus sp., Quercus sp. 04/2004 Mes-20 HKAS55842 Yunnan, China 27°120N–99°260E Populus sp., Quercus sp. 04/2004 Mes-21 HKAS55920 Shaanxi, China 35°530N–109°20E Robinia pseudoacacia, Populus simonii 05/2008 Mes-21 HKAS55921 Shaanxi, China 35°530N–109°20E Robinia pseudoacacia, Populus simonii 05/2008 Mes-22 HKAS55916 Zhejiang, China 27°400N–119°110E Bamboo 04/2008 Mes-22 HKAS55917 Zhejiang, China 27°400N–119°110E Bamboo 04/2008 Mes-22 HKAS55919 Zhejiang, China 27°400N–119°110E Bamboo 04/2008 Mes-23 HKAS56571 Chongqing, China 29°20N–107°160E Bamboo 04/2009 Mes-23 HKAS62911 Anhui, China 29°120N–117°150E–260 m Bamboo 04/2010 Mes-24 HMAS96865 Beijing, China 40°190N–116°370E Populus sp. 05/2002 Mes-25 HKAS62861 Sichuan, China 32°380N–105°180E Populus sp., Quercus sp. 04/2007 Mes-25 HKAS62862 Sichuan, China 32°380N–105°180E Populus sp., Quercus sp. 04/2007 Mes-26 HKAS55912 Hebei, China 38°520N–116°10E Phragmites australis 04/2008 Mes-26 HKAS55913 Hebei, China 38°520N–116°10E Phragmites australis 04/2008 Mes-27 HKAS55896 Sichuan, China 31°310N–102°30E Pyrus bretschneideri 04/2008 Mes-27 HKAS55897 Sichuan, China 31°310N–102°30E Pyrus bretschneideri 04/2008
a Species are identified by clade (Mel for Elata and Mes for Esculenta) followed by a unique Arabic number. b HKAS = Herbarium of Cryptogams, Kunming Institute of Botany, Chinese Academy of Sciences; HMAS = Herbarium of Mycological Institute, Chinese Academy of Sciences; HMJAU = Herbarium of Mycological Institute of Jilin Agricultural University; HAI = Herbarium, Institute of Evolution, Department of Biodiversity and Biotechnology of Fungi, University of Haifa, Israel.
3.1 (Sangon Co., Ltd., Shanghai, China), and then run on an ABI set (Fig. 2); and (ii) an 87-taxon, 4131 bp Elata Clade dataset 3730 DNA Analyzer. Raw sequence data were edited and aligned (Fig. 3). MP analyses were conducted with PAUP V 4.0b10 (Swof- with SeqMan (DNAStar Package, Madison, WI) and then they were ford, 2002); ML analyses were conducted with GARLI V 0.951 aligned automatically using MUSCLE ver. 3.8.31 (Edgar, 2004). (Zwickl, 2006). ML analyses of the concatenated datasets in GARLI Aligned sequences were visually inspected and manually adjusted used the GTR + I + C model of evolution based on analyses ob- using TextPad ver. 5.1.0 for Windows (http://www.textpad.com/) tained with ModelTest ver. 3.8 (Posada, 2006). One thousand MP or BioEdit ver. 7.0.9 (Hall, 1999; http://www.mbio.ncsu.edu/bioed- and ML bootstrap (BS) pseudoreplicates of the data were used to it/bioedit.html). Due to the presence of length variable indels with- assess clade support. Sequences generated in the present study in the ITS1 and ITS2 rDNA spacers, 321 and 254 ambiguously have been deposited in GenBank under accession numbers aligned nucleotide positions were excluded, respectively, from JQ321841–JQ322205 and JQ723016–JQ723151. the Esculenta (Table 2) and Elata (Table 3) Clade datasets. 2.5. Diversification time estimates 2.4. Phylogenetic analyses Divergence time estimates were generated using a Bayesian ap- The following two concatenated five-gene datasets were ana- proach as implemented in BEAST 1.6.1 (Drummond and Rambaut, lyzed via MP and ML: (i) a 73-taxon, 4098 bp Esculenta Clade data- 2007), using nucleotide sequences of the exonic regions of RPB1, X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469 459
Fig. 2. Morphological diversity of Morchella elata (A–D, black morels) and M. esculenta (E–P, yellow morels) Clade ascocarps from China: (A) Mel-10, (B) Mel-16, (C) Mel-19, (D) Mel-21, (E) Mes-6, (F) Mes-8, (G) Mes-9, (H) Mes-15, (I) Mes-19, (J) Mes-20, (K) Mes-21, (L) Mes-22, (M) Mes-23, (N) Mes-25, (O) Mes-26, and (P) Mes-27.
RPB2 and EF-1a. Calibration points for analysis were obtained by in Heckman et al. (2001) and Blair (2009) were used for the diver- including sequences of the following five species: Candida albicans, gence time for these five taxa. The Tamura-Nei plus gamma and Saccharomyces cerevisiae, Magnoportha grisea, Schizosaccharomyces equal frequencies model (Tamura and Nei, 1993) was identified pombe and Aspergillus flavus (see Supplemental Table S2 for Gen- by FINDMODEL (http://www.hiv.lanl.gov/content/sequence/find- Bank accession numbers for the calibration taxa). Dates reported model/findmodel.html) as the best fit to the data. Separate BEAST 460 X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469
Fig. 3. One of three equally parsimonious phylograms (MP) inferred from the Esculenta Clade (yellow morels) 73-collection 5-locus dataset comprising 4098 bp of aligned DNA sequence data and 675 parsimony-informative characters (PIC), after exclusion of 321 ambiguously aligned nucleotide positions from the ITS rDNA dataset. The 27 phylogenetically distinct species within this clade are identified by Mes followed by a unique Arabic number. The only species within this clade for which a binomial can be applied with confidence is Mes-1 (Morchella steppicola) from Hungary and the Czech Republic. The number above each internode represents the MP bootstrap value based on 1000 pseudoreplicates of the data. The ML bootstrap value is only indicated below an internode if it differed by P5% of the MP value. Sequences of M. rufobrunnea (Mruf) were used to root the phylogram. Bootstrap values obtained from analyses of the individual data partitions are provided in Supplemental Table S3. The NEXUS file is available as Supporting information dataset S1. analyses were run under an uncorrelated lognormal relaxed for 100 million generations, sampling every 1000 generations, to molecular clock and a strict molecular clock. In both analyses, estimate the posterior distribution of clade age and rates. Ten mil- the priors were identical and included the calibration dates taken lion generations were discarded as burn-in. Visualization of the from the literature (Blair, 2009; Heckman et al., 2001) and a BEAST output with TRACER version 1.5 (Rambaut and Drummond, birth–death speciation process as the tree priors. In addition, oper- 2007) indicated the MCMC has reached convergence, given that all ators were allowed to auto-optimize; the default was used for all of effective sample sizes were greater than 1200. Bayes factor calcu- the other priors. The Markov chain Monte Carlo (MCMC) was run lation in TRACER version 1.5 was used for model comparison. X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469 461
Table 2 Esculenta Clade 73-collection dataset statistics and summary sequence (see Fig. 3).
Locus # Charactersa # MPTsb MPT length CIc RId Aute Synf PIC/bpg(%) LSU rDNA 543 104 71 0.775 0.926 17 34 6.3 RPB1 718 16 172 0.884 0.941 61 80 11.1 RPB2 729 576 199 0.839 0.91 62 86 11.8 EF-1a 1059 6 328 0.777 0.914 69 154 14.5 ITS rDNA 1049 >2000 753 0.704 0.910 72 321 30.6 Combined 4098 3 1582 0.733 0.9 281 675 16.5
a Three hundred and twenty-one nucleotide characters were excluded as ambiguously aligned from the ITS rDNA partition and combined dataset. b MPTs, most-pasimonious trees. c CI, consistency index. d RI, retention index. e Aut, autapomorphic or pasimony uninformative character. f Syn, synapomorphic or parsimony informative character. g PIC/bp, parsimony-informative characters/base pair.
Table 3 Elata Clade 87-collection dataset statistics and summary sequence (see Fig. 4).
Locus # Charactersa # MPTsb MPT length CIc RId Aute Synf PIC/bpg(%) LSU rDNA 578 805 73 0.699 0.915 9 40 6.9 RPB1 768 >2000 372 0.677 0.922 46 172 22.4 RPb2 855 837 410 0.663 0.917 49 167 19.5 EF-1a 1089 18 513 0.67 0.93 46 227 20.8 ITS rDNA 841 >2000 686 0.726 0.912 51 287 34.1 Combined 4131 >2000 2139 0.662 0.91 201 893 21.6
a Two hundred and fifty-four nucleotide characters were excluded as ambiguously aligned from the ITS rDNA partition and combined dataset. b MPTs, most-pasimonious trees. c CI, consistency index. d RI, retention index. e Aut, autapomorphic or pasimony uninformative character. f Syn, synapomorphic or parsimony informative character. g PIC/bp, parsimony-informative characters/base pair.
TreeAnnotator version 1.6.1 was used to summarize the results as probabilities reported in Clayton et al. (2009) were derived on a maximum clade credibility (MCC) tree (Drummond and Ram- the basis of the appearance or disappearance of land bridges be- baut, 2007), including divergence times and highest probability tween the Holarctic and Nearctic as well as continental land mass density (HPD) values to assess the statistical uncertainty of the locations during these time periods. Constrained and uncon- divergence time estimates. FigTee 1.1.2 (Rambaut, 2008) was used strained models were employed for each set of dispersal probabil- to view the MCC tree. ities based on the area inferred for ancestral species (O’Donnell et al., 2011). Likelihood values for the constrained and uncon- 2.6. Historical biogeography strained models were compared directly (Clayton et al., 2009; Ree and Smith, 2008). Ancestral area reconstructions (AAR) were estimated from the MCC phylogeny using the program LAGRANGE version 2.0.1 (Ree 3. Results and Smith, 2008). The MCC phylogeny differed from the MP and ML phylogenies only in that the divergence dates from the BEAST 3.1. Phylogenetic diversity of Morchella in China analysis were mapped onto each node. Following detailed histori- cal biogeographic analyses of plants (Donoghue and Smith, 2004) We sequenced the ITS rDNA of 367 morel specimens collected and animals (Sanmartín et al., 2001), the four geologically persis- in 21 provinces to obtain an initial estimate of Morchella genetic tent areas recognized within the Holarctic included Europe, Asia, diversity in China. Due to the number and complexity of indels and eastern and western North America. Three additional areas and high nucleotide divergence among the ITS rDNA sequences, were delineated to describe the geographic distribution of Morch- separate alignments for members of the Esculenta and Elata Clades ella: the Dominican Republic, Venezuela + Ecuador, and the Canary were constructed with MUSCLE (Edgar, 2004). Based on MP analy- Islands. For the analyses of range evolution, a species was consid- ses of these two datasets, 40 Esculenta and 30 Elata Clade collec- ered to be endemic if it was restricted to Eurasia or North America. tions from China were chosen to represent the full range of Provincial species, by contrast, were defined as being restricted to genetic diversity sampled. Full GCPSR-based analyses of these 70 eastern or western North America, or Europe or Asia. Collections collections, together with two collections from Germany and one from Japan were coded as Asia because it was connected to China from Israel, were conducted by sequencing portions of the four loci 6 Mya (Sanmartín et al., 2001). Following Tasßkın et al. (2010, (EF-1a, RPB1, RPB2, LSU rDNA) employed in Tasßkın et al. (2010, 2012), collections from Turkey were coded as Europe. Models em- 2012) and O’Donnell et al. (2011). In addition, we also constructed ployed in the AAR varied on the basis of dispersal probabilities be- an ITS rDNA dataset, which included sequence data for members of tween areas as defined in Clayton et al. (2009). The dispersal the Elata Subclade reported in the latter two studies together with probabilities were used to evaluate biogeographic events over newly generated sequences for members of the Esculenta (N = 40) the Late Cretaceous (70 Mya), Middle Eocene (45 Mya), Early Oligo- and Elata (N = 30) Clades and M. rufobrunnea, to assess the utility of cene (30 Mya), and Early Pliocene (5 Mya) epochs. The dispersal this locus for phylogenetic reconstruction within Morchella. Se- 462 X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469 quences of M. rufobrunnea, the earliest diverging lineage within the of the species within seven derived lineages (indicated by bold genus, were used as the outgroup to root the phylogeny based on internodes in Fig. 3). The largest number of Esculenta Clade species more inclusive analyses (O’Donnell et al., 2011). Based on the re- was found in the Sino-Japanese (N = 12) and Sino-Himalaya (N =5) sults of 10 independent runs in GARLI (Zwickl, 2006), the best regions, compared with only two within the Eurasian and none ML tree inferred from the 87 taxon Elata and 73 taxon Esculenta within the Malesian regions (Fig. 1 and Table 1). Species within this Clade datasets received a lnL of 17541.57 and 14388.34, clade showed a strong preference for temperate deciduous forests, respectively. with approximately 70% of them apparently restricted to this The 5-gene Esculenta Clade dataset totaled 4098 bp of which biome. In contrast to several species within the Elata Clade, no 675 bp were parsimony informative. See Table 2 for a summary member of the Esculenta Clade was collected from a post-fire site. of tree statistics for the individual and combined partitions and Our results also strongly suggest that all of the Esculenta Clade Supplemental Table S3 for results of the bootstrap analyses. Of species in China appear to be provincial, except for the Eurasian the five partitions aligned with MUSCLE, the ITS rDNA was the only species Mes-8. Esculenta Clade species in North America are also one in which nucleotide positions were coded as ambiguously predominately provincial, except for Mes-4 and Mes-7 whose aligned. In contrast to the ITS2 which was relatively conserved in ranges are transcontinental. Although Mes-16 from Yunnan Prov- length across the breadth of the Esculenta Clade (i.e., 369– ince has a disjunct distribution represented by collections from Ha- 385 bp), indel number and placement was much more complex waii, Java, Turkey and Israel (Fig. 3), our working hypothesis is this within the ITS1 where it ranged in length from 268 bp in Mes-1 species evolved within continental Asia and that its current distri- to 586 bp in Mes-9. Therefore, to establish an alignment that re- bution outside of China represent relatively recent anthropogenic flected positional homology, 321 bp within the ITS1 and 6 bp with- introductions. in the ITS2 were coded as ambiguously aligned and excluded from With the exception of four putative species lineages repre- all subsequent analyses. Even with the exclusion of approximately sented by single collections (i.e., Mes-12 from Japan, Mes-18 from one-quarter of the nucleotide positions, the ITS rDNA still pos- Turkey, Mes-24 from Beijing, China, and Mes-? from Xinjiang), all sessed the largest number of phylogenetically informative charac- but one of the Esculenta Clade species were strongly supported ters (PIC = 321, Table 4), followed by EF-1a (154 PIC), RPB2 (86 PIC), as genealogically exclusive by MP/ML bootstrapping (Fig. 3, Sup- RPB1 (80 PIC) and the LSU rDNA (34 PIC). plemental Table S3). The single collection from Xinjiang Results of the present study substantially advance our under- (HKAS56660) was coded as Mes-? because it’s monophyly could standing of evolutionary relationships and species diversity within not be assessed by bootstrapping. Overall, evolutionary relation- the Esculenta Clade based on our discovery of nine novel species ships among terminal groups of species were generally strongly (Mes-19-to-27) within China. Moreover, two-thirds of the Escu- supported by bootstrapping. However, four nodes along the back- lenta Clade species were represented in Asia, including 16/27 in bone of the phylogeny received <50% bootstrap support. China and Mes-12 in Japan (Figs. 3 and 5). One of the most impor- We constructed and analyzed a 5-gene Elata Clade dataset (EF- tant findings of the present study was the discovery that the 16 1a, RPB1, RPB2, ITS + LSU rDNA) totaling 4131 bp of aligned se- Esculenta Clade species endemic to China comprised most or all quence data, comprising 893 parsimony informative characters
Table 4 Geographic distribution of 367 Morchella collections in China.
Speciesa # Collectionsb # Regions/# provincesc Region (# collections) Mel-6 1 1/1 Sino-Himalaya (1) Mel-7 19 1/1 Sino-Himalaya (19) Mel-9 6 1/1 Sino-Himalaya (6) Mel-10 (Mel-?) 38 + 1 (Mel-?) 2/2 Sino-Himalaya (38), Sino-Japanese (1) Mel-13 (= Mel-26) 57 3/6 Sino-Himalaya (10), Sino-Japanese (37), Eurasian (10) Mel-14 7 1/1 Sino-Himalaya (7) Mel-16 10 1/1 Sino-Japanese (7) Mel-19, Mel-20, Mel-34 99 4/9 Sino-Himalaya (73), Sino-Japanese (10), Eurasian (15), Malesian (1) Mel-21, Mel-33 33 2/4 Sino-Himalaya (3), Sino-Japanese (30) Mel-31 (Mel-23?, Mel-32?) 22 2/5 Sino-Himalaya (12), Sino-Japanese (10) Mes-6 10 2/4 Sino-Japanese (9), Eurasian (1) Mes-8 9 1/3 Sino-Japanese (9) Mes-9 10 1/1 Sino-Japanese (10) Mes-? 2 1/1 Eurasian (2) Mes-10 4 1/1 Sino-Himalaya (4) Mes-13 1 1/1 Sino-Japanese (1) Mes-15 3 1/2 Sino-Himalaya (3) Mes-16 2 1/1 Sino-Himalaya (2) Mes-19 3 1/3 Sino-Japanese (3) Mes-20 7 2/4 Sino-Himalaya (4), Sino-Japanese (3) Mes-21 3 1/1 Sino-Japanese (3) Mes-22 5 1/2 Sino-Japanese (5) Mes-23 5 1/4 Sino-Japanese (5) Mes-24 1 1/1 Sino-Japanese (1) Mes-25 4 1/1 Sino-Japanese (4) Mes-26 3 1/1 Sino-Japanese (3) Mes-27 2 1/1 Sino-Himalaya (2)
a Additional collections of Mel-? and Mes-? are needed to assess their species status. The geographic distribution of three sets of species (i.e., Mel-19, Mel-20 and Mel-34; Mel-21 and Mel-33; and Mel-23, Mel-31 and Mel-32) remains to be determined because the ITS rDNA sequence within each set was identical. Note that we have not confirmed the presence of Mel-23 and Mel-32 in China via GCPSR analyses. Also note that in our analyses Mel-26 appeared to be conspecific with Mel-13. b Because the ITS rDNA sequence of one collection from Shandong Province in the Shino-Japanese region identified by Mel-? differed from Mel-10 at 11 nucleotide positions, and we were unable to obtain sequence data from any of the protein coding genes from Mel-?, additional collections are needed to assess whether it is conspecific with Mel- 10. c Collections were obtained from four regions and 21 provinces (Fig. 1). X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469 463
(PIC), to assess the phylogenetic diversity of the Elata Clade species of contrast, was conserved at 156 bp. Visual inspection of the ITS within China. Tree statistics and bootstrap support for species rDNA alignment obtained with MUSCLE revealed that 231 bp with- monophyly are summarized, respectively, in Table 3 and Supple- in the ITS1 and 23 bp within the ITS2 rDNA regions were ambigu- mental Table S4. Of the five loci sequenced, the ITS rDNA partition ously aligned; these nucleotide positions were excluded from all was the most length variable. The ITS1 and ITS2 spacers ranged in subsequent analyses. The ITS rDNA with 287 PIC was the most length from 213 to 417 and 254 to 307 bp, respectively, within the informative partition within the Elata Clade dataset, followed by Elata Clade. The length of the 5.8S rDNA within Morchella, by way EF-1a (227 PIC), RPB1 (172 PIC), RPB2 (167 PIC), and the LSU rDNA
Fig. 4. One of >2000 equally parsimonious phylograms inferred from the Elata Clade (black morels) 87-collection five-locus dataset, using sequences of M. rufobrunnea (Mruf) to root the phylogeny. The combined dataset totaled 4131 bp of aligned DNA sequence data and included 893 parsimony-informative characters, after exclusion of 254 ambiguously aligned nucleotide positions from the ITS rDNA partition. The 34 phylogenetically distinct species within this clade are identified by Mel followed by a unique Arabic number. Latin binomials can be applied confidently to Mel-1 (M. tomentosa), Mel-3 (M. semilibera), and Mel-4 (M. punctipes). The number above each internode represents MP bootstrap support from 1000 pseudoreplicates of the data. ML bootstrap support is only indicated when it differed by P5% of the MP value. Bootstrap support obtained from analyses of the individual data partitions are provided as Supplemental Table S4. See Supporting information dataset S2 for the NEXUS file. 464 X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469
(40 PIC). In contrast to the weakly supported backbone observed in the Mel-12-to-24 subclade; O’Donnell et al., 2011), which com- our phylogenetic reconstruction of the Esculenta Clade, MP and ML prises over one-third of Morchella (i.e., 21/61 species), was dated analyses of the combined 5-gene dataset recovered a nearly fully at 12.15 Mya [95% HPD interval: 9.46–15.48] in the middle Mio- resolved phylogeny of the Elata Clade (Fig. 4). However, additional cene. Results of the BEAST analysis indicate that 51/61 (83.6%) of collections of six putative species lineages represented by single the Morchella species lineages diversified as recently as the middle collections are needed to rigorously test their species limits via Miocene 612.15 Mya. Although the divergence time estimates bootstrapping (i.e., Mel-8 OR-USA, Mel-11 Canary Islands, Mel-17 place the Esculenta Clade in China as early as the late Cretaceous Bulgaria, Mel-30 Turkey, and Mel-33 and Mel-34 China). More and Elata Clade in Asia by the early Oligocene (Fig. 5), our results inclusive analyses of two collections of Mel-18 from the Dominican indicate that 27 of the 30 Chinese species lineages (i.e., 90%) diver- Republic and three collections of Mel-24 from northeastern United sified between the middle Miocene and present. The three older States strongly supported the monophyly of these two species species lineages represented in China are either obligate (i.e., (O’Donnell, unpubl.). Mel-6 and Mel-9) or facultative (i.e., Mel-10) post-fire morels GCPSR analyses of the individual and combined partitions re- whose evolutionary origins appear to have been within western vealed that 13/34 Elata Clade species were represented in China North America. (Figs. 4 and 6). Three of these (Mel-6, Mel-7 and Mel-9) were col- The constrained 70 Mya dispersal-extinction cladogenesis lected from burned coniferous forests in Yunnan province. A fourth (DEC) model was used for the ancestral area reconstructions post-fire species, Mel-10, appears to be a faculatively post-fire spe- (AAR) of the 75-taxon dataset because it received a higher likeli- cies in that has been collected from non-burned sites in Yunnan, hood value than the constrained and unconstrained DEC models China, Germany (Fig. 4), and Turkey (Tasßkın et al., 2010, 2012). at the four different time periods tested (i.e., lnL M70 = 160.6, The remaining 10 Elata Clade species from China were nested M45 = 163.2, M30 = 166.1 and M5 = 171.8). The optimal AAR ob- within the species-rich Elata Subclade. Six of the Elata Subclade tained with LAGRANGE version 2.0.1, using the 70 Mya DEC model, species present in China have been collected in other countries required 28 range expansions within Morchella (Fig. 5). Sixteen and these include Mel-13 from India, Mel-16 and Mel-19 from range expansions were intercontinental, nine were intracontinen- Europe, Mel-20 from Europe and Turkey, Mel-21 from Japan, and tal, and three appear to represent putative transoceanic long dis- Mel-31 from Turkey (Fig. 4; O’Donnell et al., 2011; Tasßkın et al., tance dispersals (LDD) or, alternatively, introductions by humans. 2010). Mel-13 from six provinces in western China was one of Two-thirds of the range expansions involved members of the Elata the most widespread species of Morchella in China. Results of the Clade (12 intercontinental, 5 intracontinental and 2 putative LDD) present study also suggest that Mel-26 from Turkey may be nested compared with only nine within the Esculenta Clade (4 interconti- within Mel-13 (Fig. 4). The largest number of Elata Clade species nental, 4 intracontinental and 1 putative LDD). Intracontinental was found in the Sino-Himalaya (N = 12) and Sino-Japanese range expansions were asymmetric with five of eight involving (N = 7) regions, compared with only three within the Eurasian migrations from Asia to Europe in the Palaearctic. Dispersals were and one within the Malesian regions (Fig. 1 and Table 1). Species equally common out of Asia (N = 9), Europe (N = 9), and western within this clade showed a strong preference for evergreen conifer- North America (N = 9), whereas only one out of eastern North ous forests, with approximately 70% of them apparently restricted America was observed. Eight of the 16 intercontinental range to this biome. Lastly, it is noteworthy that no representative of the expansions within Morchella involved nonadjacent areas, consis- Morchella semilibera clade (Mel-3, Mel-4 and Mel-5) was discovered tent with species extinctions in the intervening areas. Roughly in the present survey of China. three-quarters of the species (i.e., 48/61) appeared to be restricted to a single area (i.e., they exhibited provincialism). However, 10 3.2. Diversification time estimates and historical biogeography of species occupied two areas and three of the fire-adapted species Morchella (Mel-7, Mel-9 and Mel-10) were present in China and Europe in addition to their putative area of endemism in western North We estimated the origin of Morchella and the Morchellaceae America. Of the 13 non-provincial species, 8/13 exhibited conti- using a relaxed molecular clock in BEAST version 1.6.1 (Drummond nental endemism (Europe–Asia, N = 6; North America, N = 2) and and Rambaut, 2007) from 75 aligned, partial RPB1, RPB2 and EF-1a 4/13 occupied adjacent regions in the Old and New World. By con- nucleotide sequences, comprising 62 phylogenetically distinct spe- trast, Mel-6 occupied two nonadjacent areas (i.e., Asia and western cies of Morchella, eight sequences of the morchellaceous taxa Verpa North America). Of the 16 intercontinental range expansions with- and Disciotis, and sequences of five taxa used to calibrate the chro- in Morchella (Fig. 5), the earliest was the hypothesized dispersal of nogram against the geologic time scale (Walker et al., 2009). The the Mes-1 lineage into Europe via the Thulean North Atlantic Land MCC chronogram (Fig. 5) was topologically concordant with phy- Bridge between the late Cretaceous and early Eocene 68.67 Mya logenies of the Esculenta and Elata Clade inferred from the com- [95% HPD interval: 49.19–93.03]. This bridge may also have served bined 5-locus datasets (Figs. 3 and 4). These analyses indicate as the dispersal corridor in the range expansions of the M. semilib- Morchella split from its epigeous sisters Verpa and Disciotis in the era clade (i.e., Mel-3-to-Mel-5) and its sister clade from North middle Triassic 274.06 Mya [95% HPD interval: 213.96–347.77]; America to Europe between the early Paleocene to middle Eocene divergence of Verpa and Disciotis was dated to the early Cretaceous 61.40 Mya [95% HPD interval: 48.51–76.77]. By contrast, three 147.29 Mya [95% HPD interval: 106.58–196.72]. Evolutionary separate Beringian land bridges (BBI, BBII and BBIII sensu San- diversification of the three main lineages within Morchella was da- martín et al., 2001) at different geologic time periods appear to ted to the late Jurassic-to-early Cretaceous within western North have served as dispersal corridors for the remaining 12 paleocon- America. The earliest diverging branch of true morels was repre- tinental range expansions (Fig. 5). The earliest two of these are sented by the monotypic M. rufobrunnea lineage with an estimated hypothesized to have taken place during the Eocene via BB-I, divergence time of 154.15 Mya (95% HPD interval: 121.79–194.71) followed by migration of the ancestor of Mel-12 from the Old in the late Jurassic, followed by the split of the Esculenta and Elata to the New World via BB-II during the Miocene. The remaining Clades 123.46 Mya [95% HPD interval: 98.74–152.95] in the early nine intercontinental range expansions appear to have utilized Cretaceous. Subsequent evolutionary diversification of the Elata BB-III during the Pliocene-to-Pleistocene over the last 3 million Clade was dated at 73.75 Mya [95% HPD interval: 70.49–109.89] years. The 12 paleocontinental dispersals were evenly distrib- and the Esculenta Clade at 68.67 Mya [95% HPD interval: 49.19– uted between Nearctic-to-Paleoarctic and Paleoarctic–Nearctic 93.03]. The root node of the species-rich Elata Subclade (formerly migrations. X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469 465
V A V Mes-14 A W = Western North America A A Mes-15 A E = Eastern North America A Mes-16 U = Europe A A A Mes-13 A = Asia A A Mes-19 V = Venezuela & Equador A A Mes-24 D = Dominican Republic A A Mes C = Canary Islands A -22 A A A Mes-12 B = Beringian land bridge A A N = North Atlantic land bridge A A Mes-23 = dispersal resulting in A B A A Mes-10 A range expansion A E Mes-11 A E A A Mes-25 A A A Mes-21 U A U Mes-17 U A W E A W E Mes U B -7 A E A A Mes-6 A U A U Mes-5 A Mes A A -27 A A Mes-26 W E U Mes-18 U A Mes-9 A A B W E A A U AA Mes-8 Esculenta E A A Mes-20 A N Clade W E W E Mes-4 U E EE Mes-2 E E Mes-3 U Mes-1 W Mel-1 E E Mel-4 N W E W W Mel-5 U U Mel-3 W U U U Mel-23 W E U B EE Mel-24 U U AA Mel-31 UAA U U Mel-32 U W Mel U U W -22 U B U Mel-28 U U Mel-29 U U Mel-30 W U U A Mel W A U AA U A -19 Elata A A A Mel-33 A U Clade W A A Mel A -34 W A U D U AA Mel-20 U D Mel-18 A Elata A A Mel-21 W Subclade U U Mel-17 W U B U U Mel-27 U W W Mel-12 U U U Mel-26 U U U U AA Mel-13 U U A Mel W -14 W A B EE EE Mel-15 U A U A U A Mel-16 N U U U Mel-25 U C C Mel-11 B W W Mel-8 W W A UU W A UU Mel-7 W W W A UU Mel-10 B W B W W W A UU Mel-9 B WAA Mel-6 B Morchella rufobrunnea lineage WU Mel-2 W Mruf
145.5 65.5 55.8 33.9 23.0 5.3 1.8
Jurassic Cretaceous Paleocene Eocene Oligocene Miocene Plio P
150 125 10075 50 25 0
Fig. 5. Geographic range evolution of Morchella (true morels) under the constrained M70 DEC model (Clayton et al., 2009) using the maximum likelihood-based program LAGRANGE (Ree and Smith, 2008). The maximum clade credibility (MCC) chronogram, with divergence time given in million years before present (Walker et al., 2009), was based on the combined analysis of a four gene dataset using BEAST. The five calibration taxa used to estimate the divergence times (see Section 2) and the morchellaceous taxa Disciotis and Verpa are not shown in the MCC chronogram to focus on the range evolution of Morchella. Ancestral area reconstructions (AARs) with the highest likelihood are shown above and below each internode. Note that most ancestral areas and extant species are restricted to a single area. Red branches are used to identify 28 dispersals resulting in range expansions. Dispersals between the Old and New Worlds appear to have involved the Thulean North Atlantic (N) or Beringian (B) land bridges as the dispersal corridor. Phylogenetic species are identified as follows: Mes for Esculenta Clade and Mel for Elata Clade followed by a unique Arabic number; Mruf = M. rufobrunnea. Plio = Pliocene, P = Pleistocene. The NEXUS file is provided as Supporting information dataset S3.
4. Discussion species diversity in China from only nine reported in O’Donnell et al. (2011) to 30 in the present study. Notably, we have discov- 4.1. Phylogenetic diversity of Morchella in China ered that Morchella is considerably more species-rich in China than in Europe (N = 22) or North America (N = 19). Furthermore, our re- Phylogenetic analyses of 367 collections of true morels from 21 sults indicate that China has served as the primary center of Mor- provinces has significantly extended our knowledge of Morchella chella diversification since the Miocene. One of the most important 466 X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469
by 10 collections from the eastern costal province of Shandong Morchella A where Japan was connected to continental Asia 6 million years ago (Sanmartín et al., 2001). Our discovery of Mes-16 in Yunnan Europe Province adds support to the hypothesis that it’s transoceanic dis- 13 junct distribution in disturbed sites in Java and Hawaii (O’Donnell et al., 2011), a greenhouse containing exotic plants in Turkey 6 (Tasßkın et al., 2012), and Israel in the present study represent 3 anthropogenic introductions from its endemic area in continental 20 6 2 6 Asia. 1 Asia Western Eastern The present study is the first to use a DNA sequence alignment N. America N. America of the ITS rDNA to infer evolutionary relationships across the breadth of the Esculenta Clade that accurately reflects positional homology. Previously Kanwal et al. (2011) published analyses of an ITS rDNA dataset comprising sequences of several Esculenta B Morchella esculenta and Elata Clade species, but this analysis is seriously flawed in Clade two important respects. First, no ambiguously aligned nucleotide Europe positions were excluded from their genus wide alignment; how- 4 ever, we found that over half of the ITS1 was unalignable within our Esculenta and Elata Clade datasets. Secondly, because Kanwal 1 et al. (2011) did not study type collections, nor did most of the peo- ple who have deposited ITS rDNA sequences of Morchella in Gen- Bank, species names applied to true morels have been broadly 16 2 3 misapplied, as noted for other fungi (Nilsson et al., 2006; Vilgalys, Asia Western Eastern 2003). As a result, three species included in the Kanwal et al. N. America N. America (2011) phylogeny are paraphyletic and one is polyphyletic. Results of the present study build on the initial estimate of six Elata Clade species in Asia (O’Donnell et al., 2011), including five from China and one from Japan, to our present estimate of 13. C Morchella elata We also discovered that Mel-17 from the Beijing Herbarium Clade (HMAS), which was reported as having been collected in China in Europe the aforementioned study, was actually from Bulgaria. In contrast 9 to members of the Esculenta Clade, which exhibit high continental endemism and provincialism within the Northern Hemisphere 5 (present study and Tasßkın et al., 2010, 2012; O’Donnell et al., 3 2011), only 3/13 Elata Clade species collected in our survey appear 4 5 3 to be restricted to China. Not surprisingly, the majority of the non- Asia 1 Western Eastern provincial species (8/13) occupy adjacent areas in Europe. How- N. America N. America ever, three obligately (Mel-6, Mel-7 and Mel-9) and one putative facultatively fire-adapted species (Mel-10) in China have disjunct distributions in western North America, and three of these were also well represented in Turkey (Tasßkın et al., 2010, 2012) Fig. 6. Venn diagrams depicting the high level of continental endemism and provincialism within (A) Morchella, (B) the Esculenta Clade, and (C) the Elata Clade (Fig. 5). Previously, the disjunct distributions of Mel-7, Mel-9 and within the Northern Hemisphere. Mel-10 in western North America and Turkey were hypothesized to have been mediated by human activities, possibly linked to findings reported herein was the discovery of nine novel species importation of exotic forestry tree species into Turkey from wes- within Esculenta Clade (yellow morels) and two novel species tern North America (O’Donnell et al., 2011; Tasßkın et al., 2010). within the Elata Clade (black morels) in China. Another noteworthy However, with the discovery of four Elata Clade post-fire species finding is our discovery that the majority of the Esculenta Clade ap- within Yunnan Province, which are all thought to be endemic to peared to be adapted to mixed temperate deciduous hardwood for- conifer forests within western North America, we have to consider ests compared with the Elata Clade’s preference to an evergreen that relatively recent long-distance dispersal (LDD) may have con- coniferous biome, indicating that the phylogenetic niche conserva- tributed to their current disjunct ranges in Europe and Asia. We tism (Donoghue, 2008) may have played a significant role in con- further speculate that if LDD is involved, then their current distri- straining the geographic distribution of Morchella. butions in coniferous forests may reflect niche conservation, given The present study significantly extends our understanding of their putative evolutionary origins within similar Pinus, Abies and Esculenta Clade diversity in Asia from six reported in O’Donnell Picea dominated forests of western North America. et al. (2011) to 17. Remarkably, the Asian species comprised close In contrast to the Esculenta Clade where no post-fire species are to three-quarters (i.e., 17/23) of the derived lineages within Escu- known, six species putatively endemic to western North America lenta Clade whose origins date to the middle Miocene, based on within the Elata Clade appear to be fire-adapted (Mel-1, Mel-6, our diversification time estimates (Fig. 5). Another noteworthy Mel-7, Mel-8, Mel-9 and Mel-10). We speculate that this adaptive finding is all of the species within Esculenta Clade exhibited pro- shift evolved in situ in their putative area of endemism in western vincialism except for the Eurasian species Mes-8, which was previ- North America in response to periodic lightning-induced forest ously only documented from Europe where it is widely distributed fires. The available data suggests the former three species are obli- (O’Donnell et al., 2011). Of the six Asian species reported in the gate fire-adapted, whereas this adaptation in the latter two species aforementioned study, Mes-12 from Japan was the only Asian spe- may be facultative given that both have been collected on burned cies we failed to find in the present survey of China. However, Mes- and non-burned sites (present study; O’Donnell et al., 2011; Tasßkın 9, which was previously only known from Japan was represented et al., 2010, 2012). As previously reported (O’Donnell et al., 2011), X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469 467 our results indicate these adaptive shifts evolved convergently, sions (N = 9). We interpret this to reflect, in part, migration out of which suggests ecological speciation may have played an impor- the ancestral area in western North America, as is evident for the tant role in the evolutionary diversification of morels. earliest range expansions within the Esculenta and Elata Clades. By contrast, the majority of intracontinental range expansions 4.2. Diversification time estimates and historical biogeography of within Morchella appear to have taken place relatively recently Morchella (i.e., 615 Mya) between the middle Miocene and Pleistocene. These recent range expansions were very likely facilitated by the Our results support a late Jurassic origin of the Morchella rufob- disappearance of the Turgai Sea in the Palaearctic and Mid-Conti- runnea lineage 154.15 Mya [95% HPD: 121.79–194.71] and an nental Seaway in the Nearctic, which are thought to have func- early-to-mid-Cretaceous origin of the Esculenta and Elata Clades tioned as vicariant agents limiting intracontinental biotic 123.46 Mya [95% HPD: 98.74–152.95], all in western North Amer- exchanges of Morchella within the Old and New Worlds until the ica. These age estimates differ slightly from those reported previ- end of the Eocene (Sanmartín et al., 2001). ously (O’Donnell et al., 2011), in that they push the age of these Diverse biotas, including Morchella, appear to have been nega- lineages back in time approximately 22–24 million years; how- tively impacted by late Miocene climate changes and aridification. ever, the older dates fall within the Bayesian 95% HPD values re- Widespread desertification-related extinctions of diverse plants ported in the aforementioned study. Even though the 95% HPD within western North America, for example, have been proposed values for these lineages overlapped, the medium node heights to explain why eastern Asia-eastern North America represents indicate the M. rufobrunnea lineage represents the basal most the largest intercontinental disjunction category in plants (Donog- divergence of true morels, followed by the origin of the Esculenta hue and Smith, 2004). Widespread extinctions also appear to have and Elata sister clades approximately 30 million years later. As taken place within Morchella, given that species were missing from noted previously (Berbee and Taylor, 2010; Garcia-Sandoval Asia (N = 3) and western North America (N = 2) in five of the nine et al., 2011), molecular clock estimates such as the ones reported nonadjacent Nearctic–Palaearctic disjunctions (Fig. 5). We further here have several potential sources of error including calibration speculate that fragmentation of ancestral species ranges resulting and rate heterogeneity. In the absence of a fossil record, we cali- from late Miocene climate changes and aridification may have con- brated the Morchella chronogram against the geological time scale tributed to the explosive radiation of the species-rich Elata Subc- (Walker et al., 2009) using published diversification time estimates lade 12.15 Mya [95% HPD: 9.46–15.48]. This evolutionary for five phylogenetically diverse Ascomycota (Blair, 2009; Heck- dynamic subclade comprises roughly one-third of Morchella (i.e., man et al., 2001). In the unlikely event that a fossil morel is ever 21/61 species) and accounted for 10/28 range expansions. found to use as an internal calibration for Morchella, refinement Results of the AARs revealed several noteworthy patterns with- of the divergence time estimates will have to rely on identifying in Morchella (Fig. 5). As noted for range evolution of the rubiaceous a large set of phylogenetically informative orthologous genes that shrub Psychotria in the Hawaiian Islands (Ree and Smith, 2008), evolve in a clock-like fashion from the whole genomes of related ancestral ranges of Morchella spp. rarely occupied more than one Ascomycota. Because divergence times can be biased by rate heter- area. Range expansions within Morchella were typically accompa- ogeneity, as shown for angiosperms with different generation nied by allopatric cladogenesis, which may help explain in part times (Smith and Donoghue, 2008), we used an uncorrelated log- why approximately 80% of the extant species exhibit provincialism normal model of molecular evolution in our divergence time esti- (O’Donnell et al., 2011). However, our discovery of four fire- mates to account for potential lineage-specific rate differences adapted species thought to be native to western North America (Drummond et al., 2006). in Yunnan Province, China (Mel-6, Mel-7, Mel-9 and Mel-10), and As reported for diverse plants and animals within the Holarctic reports of three New World endemic post-fire morels in Turkey (Donoghue and Smith, 2004; Sanmartín et al., 2001), range evolu- (Tasßkın et al., 2010, 2012; Mel-7, Mel-9 and Mel-10), opens the pos- tion in Morchella involved numerous biotic exchanges within and sibility that these widely disjunct distributions might be due to rel- between the Palaearctic and Nearctic (O’Donnell et al., 2011). In atively recent transoceanic long distance dispersals (LDD). While contrast to many plants that evolved within and dispersed out of the role that LDD has played in range expansions within Morchella Asia (Donoghue and Smith, 2004), range expansions within Morch- remains an open question, it has been invoked to explain the cur- ella were nearly evenly divided between ones out of Asia (N = 9), rent distributions of diverse Basidiomycota (Hibbett, 2001; Hosaka Europe (N = 9) and western North America (N =9)(Fig. 5). Our et al., 2008; Matheny et al., 2009; Moncalvo and Buchanan, 2008; AARs indicate the M. rufobrunnea lineage and the Esculenta and Skrede et al., 2011) and Ascomycota (Liu et al., 2009; Takamatsu Elata Clades evolved initially within western North America in et al., 2006). However, Tasßkın et al. (2010, 2012) and O’Donnell the late Jurassic and middle Cretaceous, respectively. Consistent et al. (2011) rejected LDD in Morchella in favor of human introduc- with their theorized evolutionary origin in western Nearctic, the tions of horticultural and silvicultural plants and accompanying two earliest range expansions within the Esculenta Clade and the soil containing ascospores and/or sclerotia (Miller et al., 1994)to three earliest within the Elata Clade involved intercontinental explain widely disjunct species distributions. Our discovery of migrations out of western North America into the Palearctic. Three Mes-16 in China in the present study, coupled with the fact this of these are hypothesized migrations across the Thulean North species was collected on disturbed sites in Hawaii, Java, and Turkey Atlantic Land Bridge at different times between the Cretaceous to strongly suggests that its range outside of China was due to human middle Eocene. The remaining 12 intercontinental range expan- introductions. Therefore, we predict that future surveys may lead sions, as reported for true truffles (Jeandroz et al., 2008) and di- to the discovery of Mes-14 and Mel-18 in continental Asia. Addi- verse Basidiomycota (Garnica et al., 2011; Geml et al., 2008; tional studies are also needed to assess whether their putatively Hibbett, 2001), and most vascular plants (Donoghue and Smith, heterothallic reproductive mode, absence of a highly dispersive 2004), appear to have utilized different Beringian land bridges conidial stage, and niche conservation help explain their low vagil- (BB-I, BB-II and BB-III sensu Sanmartín et al., 2001) as the dispersal ity. Our analyses suggest possible niche conservation of the Escu- corridor between the Old and New World during the Eocene (BB-I), lenta and Elata Clades within temperate deciduous and Miocene-to-Pliocene (BB-II) or Pliocene-to-Pleistocene (BB-III) coniferous forests, respectively, in that approximately 70% of the over the past 30 million years. Range evolution in Morchella differs species within these two clades are only known from these two from plants and animals in that intercontinental exchanges biomes. Because both biomes were well established in the (N = 16) appear to be more common than intracontinental expan- Northern Hemisphere by the time the Esculenta and Elata Clade 468 X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469 began to diversify in the Cretaceous (Axelrod, 1960), these niches References may have been conserved for the past 100 million years. Arora, D., 1979. Mushrooms Demystified. Ten Speed Press, Berkeley. Axelrod, D.J., 1960. The evolution of flowering plants. In: Tax, D. (Ed.), Evolution 4.3. Conclusions and future studies after Darwin, vol. 1. Univ. of Chicago Press, Chicago, pp. 227–305. Berbee, M.L., Taylor, J.W., 2010. Dating the molecular clock in fungi – how close are Our analyses have provided the first detailed molecular phylo- we? Fungal Biol. 24, 1–16. Blair, J.E., 2009. Fungi. In: Hedges, S.B., Kumar, S. (Eds.), The Timetree of Life. Oxford genetic assessment of true morels (Morchella) in mainland China. University Press, New York., pp. 219–251. Herein we reported on the discovery of 11 novel phylogenetically Bonito, G.M., Gryganskyi, A.P., Trappe, J.M., Vilgalys, R., 2010. A global meta-analysis distinct species which establishes eastern Asia as the most tax- of Tuber ITS rDNA sequences: species diversity, host associations and long- on-rich region with 30 species. Ancestral area reconstructions distance dispersal. Mol. Ecol. 19, 4994–5008. Chen, J.Y., Liu, P.G., 2005. A new species of Morchella (Pezizales, Ascomycota) from and analyses of range evolution indicated that 90% of these species southwestern China. Mycotaxon 93, 89–93. diversified between the middle Miocene and present 615 Mya. Our Clayton, J.W., Soltis, P.S., Soltis, D.E., 2009. Recent long-distance dispersal maximum likelihood estimation of Morchella biogeographic history overshadows ancient biogeographical patterns in a pantropical angiosperm family (Simaroubaceae, Sapindales). Syst. Biol. 58, 395–410. required 28 range expansions and revealed a complex history of Dettman, J.R., Jacobson, D.J., Taylor, J.W., 2003. A multilocus genealogical approach biotic exchanges between the Old and New Worlds. Most of these to phylogenetic species recognition in the model eukaryote Neurospora. were interpreted as Beringian and Thulean North Atlantic vicari- Evolution 57, 2703–2720. Donoghue, M.J., 2008. A phylogenetic perspective on the distribution of plant ances, but several appeared to have involved rare transoceanic long diversity. Proc. Natl. Acad. Sci. U.S.A. 105, 11549–11555. distance dispersals or human introductions (Bonito et al., 2010; Donoghue, M.J., Smith, S.A., 2004. Patterns in the assembly of temperate forests Vellinga et al., 2009). around the Northern Hemisphere. Philos. Trans. Roy. Soc. Lond. B 359, 1633– 1644. Results of the present study should help inform taxonomic revi- Drummond, A.J., Ho, S.Y.W., Phillips, M.J., Rambaut, A., 2006. Relaxed phylogenetics sion, which is urgently needed, given that 16/20 species endemic and dating with confidence. PLoS Biol. 4, e88. http://dx.doi.org/10.1371/ to China appear to be unnamed. Our results also punctuate the journal.pbio.0040088. Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by need for future systematic surveys of Morchella in China and adja- sampling trees. BioMed. Cen. Evol. Biol. 7, 214. http://dx.doi.org/10.1186/1471- cent regions to more fully assess species diversity and their geo- 2148-7-214. graphic distribution, especially in poorly sampled biodiversity Edgar, R.C., 2004. MUSCLE: a multiple sequence alignment method with reduced hotspots (Myers et al., 2000; Zhang et al., 2009). Information time and space complexity. BioMed. Cen. Bioinform. 5, 1–19. Garcia-Sandoval, R., Wang, Z., Binder, M., Hibbett, D.S., 2011. Molecular gained from these studies is crucial for advancing morel conserva- phylogenetics of the Gloeophyllales and relative ages of clades of tion genetics and for formulating informed policies directed at Agaricomycotina producing a brown rot. Mycologia 103, 510–524. insuring commercial harvests are sustainable. Lastly, with one Garnica, S., Spahn, P., Oertel, B., Ammirati, J., Oberwinkler, F., 2011. Tracking the evolutionary history of Cortinarius species in section Calochroi, with Morchella whole genome sequence in the pipeline, and the likeli- transoceanic disjunct distributions. BioMed. Cen. Evol. Biol. 11, 213. http:// hood of several others to quickly follow, access to these genomic dx.doi.org/10.1186/1471-2148-11-21. resources should provide a wealth of phylogenetically informative Geml, J., Tulloss, R.E., Laursen, G.A., Sazanova, N.A., Taylor, D.L., 2008. Evidence for strong inter- and intracontinental phylogeographic structure in Amanita loci needed to develop a fully resolved phylogeny needed to serve muscaria, a wind-dispersed ectomycorrhizal basidiomycete. Mol. Phylogenet. as a robust framework for understanding intrinsically and biotech- Evol. 48, 694–701. nologically important facets of their biology such as the evolution Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl. Acids Sym. Ser. 41, 95–98. of nutritional mode and mating systems. Hansen, K., Pfister, D.H., 2006. Systematics of the Pezizomycetes – the operculate discomycetes. Mycologia 98, 1029–1040. Acknowledgments Heckman, D.S., Geiser, D.M., Eidell, B.R., Stauffer, R.L., Kardos, N.L., Hedges, S.B., 2001. Molecular evidence for the early colonization of land by fungi and plants. Science 293, 1129–1133. We are very grateful to Prof. Y.J. Yao and Dr. T. Z. Wei, Institute Hibbett, D.S., 2001. Shiitake mushrooms and molecular clocks: historical of Microbiology of the Chinese Academy of Sciences (HMAS, biogeography of Lentinula. J. Biogeogr. 28, 231–241. Hosaka, K., Castellano, M.A., Spatafora, J.W., 2008. Biogeography of Hysterangiales Beijing), Prof. B. Tolgor, Mycological Institute of Jilin Agricultural (Phallomycetidae, Basidiomycota). Mycol. Res. 112, 448–462. University (HMJAU), Prof. G. Kost and Dr. K.-H. Rexer, University Huang, N.L., 1998. Colored Illustrations of Macrofungi of China. China Agricultural of Marburg, Germany (MB), and Dr. G.S. Barseghyan, University Press, Beijing. of Haifa, Israel (HAI) for sending us specimens on loan. Thanks Imazeki, R., Otani, Y., Hongo, T., 1988. Fungi of Japan. Yama-Kei Publishers Co., Ltd., Tokyo (in Japanese). are also due Drs. B. Feng, Y.C. Li and Z.W. Ge, Kunming Institute Jeandroz, S., Murat, C., Wang, Y., Bonfante, P., Le Tacon, F., 2008. Molecular of Botany (KIB) for kindly providing specimens and photos. We also phylogeny and historical biogeography of the genus Tuber, the ‘true truffles’. J. thank Stacy Sink, NCAUR-ARS-USDA, for excellent technical assis- Biogeogr. 35, 815–829. Kanwal, H.K., Acharya, K., Ramesh, G., Reddy, M.S., 2011. Molecular characterization tance and John Plischke III for the photo of Morchella rufobrunnea of Morchella species from the Western Himalayan region of India. Curr. used in Fig. 5. This study was supported by the Joint Funds from Microbiol. 62, 1245–1252. the National Natural Science Foundation of China and Yunnan Pro- Li, S.-H., Zhao, Y.-C., Chai, H.-M., Zhong, M.-H., 2006. Two new species in the genus Morchella (Pezizales, Morchellaceae) from China. Mycotaxon 95, 319–322. vincial Government (No. U0836604), the National Basic Research Liu, K.-B., 1988. Quaternary history of the temperate forests of China. Quart. Sci. Rev. Program of China (No. 2009CB522300), the Hundred Talents Pro- 7, 1–20. gram of the Chinese Academy of Sciences, and the Key Laboratory Liu, M., Milgroom, M.G., Chaverri, P., Hodge, K.T., 2009. Speciation of a tropical fungal species pair following transoceanic dispersal. Mol. Phylogenet. Evol. 51, of Biodiversity and Biogeography (KIB, No. KBB-200805). The men- 413–426. tion of trade names or commercial products in this publication is Mao, X.L., 2000. The Macrofungi in China. Henan Science and Technology, solely for the purpose of providing specific information and does Zhengzhou. Matheny, P.B., Aime, M.C., Bougher, N.L., Buyck, B., Desjardin, D.E., Horak, E., Kropp, not imply recommendation or endorsement by the US Department B.R., Lodge, D.J., Soytong, K., Trappe, J.M., Hibbett, D.S., 2009. Out of the of Agriculture. The USDA is an equal opportunity provider and Palaeotropics? Historical biogeography and diversification of the cosmopolitan employer. ectomycorrhizal mushroom family Inocybaceae. J. Biogeogr. 36, 577–592. Miller, S.L., Torres, P., McClean, T.M., 1994. Persistence of basidiospores and sclerotia of ectomycorrhizal fungi and Morchella in soil. Mycologia 86, 89–95. Appendix A. Supplementary material Moncalvo, J.M., Buchanan, P.K., 2008. Molecular evidence for long distance dispersal across the Southern Hemisphere in the Ganoderma applanatum–australe species complex (Basidiomycota). Mycol. Res. 112, 425–426. Supplementary data associated with this article can be found, in Mou, C.-J., 1987. Two species of Morchella, new to Xinjiang, and a variety, new to the online version, at http://dx.doi.org/10.1016/j.fgb.2012.03.006. science. Mycosystema 6, 122–123. X.-H. Du et al. / Fungal Genetics and Biology 49 (2012) 455–469 469
Myers, N., Mittermeier, R.A., Mittermerier, C.G., da Fonseca, G.A.B., Kent, J., 2000. Takamatsu, S., Matsuda, S., Niinomi, S., Havrylenko, M., 2006. Molecular phylogeny Biodiversity hotspots for conservation priorities. Nature 403, 853–858. supports a northern hemisphere origin of Golovinomyces (Ascomycota: Nilsson, R.H., Ryberg, M., Fristiansson, E., Abarenkov, K., Larsson, K.-H., Köljalg, U., Erysiphales). Mycol. Res. 110, 1093–1101. 2006. Taxonomic reliability of DNA sequences in public sequence databases: a Tamura, K., Nei, M., 1993. Estimation of the number of nucleotide substitutions in fungal perspective. PLoS One 1, e59. http://dx.doi.org/10.1371/journal.pone. the control region of mitochondrial DNA in humans and chimpanzees. Mol. Biol. 0000059. Evol. 10, 512–526. O’Donnell, K., Cigelnik, E., Weber, N.S., Trappe, J.M., 1997. Phylogenetic relationships Tasßkın, H., Büyükalaca, S., Dogan, H.H., Rehner, S.A., O’Donnell, K., 2010. A multigene among ascomycetous truffles and the true and false morels inferred from 18S molecular phylogenetic assessment of true morels (Morchella) in Turkey. Fungal and 28S ribosomal DNA sequence analysis. Mycologia 89, 48–65. Genet. Biol. 47, 672–682. O’Donnell, K., Rooney, A.P., Mills, G.L., Kuo, M., Weber, N.S., Rehner, S.A., 2011. Tasßkın, H., Büyükalaca, S., Hansen, K., O’Donnell, K., 2012. Multilocus phylogenetic Phylogeny and historical biogeography of true morels (Morchella) reveals an analysis of true morels (Morchella) reveals high levels of endemics in Turkey early Cretaceous origin and high continental endemism and provincialism in relative to other regions of Europe. Mycologia 104, 446–461. the Holarctic. Fungal Genet. Biol. 48, 252–265. Taylor, J.W., Jacobson, D.J., Kroken, S., Kasuga, T., Geiser, D.M., Hibbett, D.S., Fischer, Ower, R., Mills, G.L., Malachowski, J.A., 1986. Cultivation of Morchella. US Patent M.C., 2000. Phylogenetic species recognition and species concepts in fungi. 4594809. Fungal Genet. Biol. 31, 21–32. Pagliaccia, D., Douhan, G.W., Douhan, L., Peever, T.L., Carris, L.M., Kerrigan, J.L., Vellinga, E.C., Wolfe, B.E., Pringle, A., 2009. Global patterns of ectomycorrhizal 2011. Development of molecular markers and preliminary investigation of introductions. New Phytol. 181, 960–973. the population structure and mating system in one lineage of black Vilgalys, R., 2003. Taxonomic misidentification in public DNA databases. New morels (Morchella elata) in the Pacific Northwestern USA. Mycologia 103, Phytol. 160, 4–5. 969–982. Walker, J.D., Geissman, J.W., compilers, 2009. Geologic time scale. Geol. Soc. Am. Pilz, D., McLain, R., Alexander, S., Villarreal-Ruiz, Berch, S., Wurtz, T.L., Parks, C.G., http://dx.doi.org/10.1130/2009. McFarlane, E., Baker, B., Molina, R., Smith, J.E., 2007. Ecology and Management Weber, N.S., 1988. A Morel Hunter’s Companion: A Guide to the True and False of Morels Harvested from the Forests of Western North America. General Morels of Michigan. Two Peninsula Press, Lansing. Technical Report PNW-GTR-710. Portland: US Department of Agriculture, Forest Wu, Z.Y., Wu, S.G., 1996. A proposal for a new floristic kingdom (realm) – the E. Service, Pacific Northwest Research Station, 161pp. Asiatic kingdom, its delimitation and characteristics. In: Zhang, A.L., Wu, S.G. Posada, D., 2006. ModelTest Server: a web-based tool for the statistical selection of (Eds.), Proceedings of the First International Symposium on Floristic models of nucleotide substitution online. Nucl. Acids Res. 34, W700–W703. Characteristics and Diversity of East Asian Plants. China Higher Education Rambaut, A., 2008. FigTree. Institute of Evolutionary Biology, Univ. Edinburgh, Press/Springer-Verlag, Beijing, China/Berlin, Germany, pp. 3–42. Edinburgh. Ying, J.Z., Zang, M., 1994. Economic Macrofungi from Southwestern China. Science Rambaut, A., Drummond, A.J., 2007. Tracer v1.4.