Genome Wide AFLP Markers Support Cryptic Species in Coniophora (Boletales)

fungal biology xxx (2012) 1e7

journal homepage: www.elsevier.com/locate/funbio

Genome wide AFLP markers support cryptic species in Coniophora (Boletales)

Inger SKREDE*, Tor CARLSEN, Øyvind STENSRUD, Havard KAUSERUD

Microbial Evolution Research Group (MERG), Department of Biology, University of Oslo, P.O. Box 1066 Blindern, N-0316 Oslo, Norway article info abstract

Article history: Numerous fungal morphospecies include cryptic species that routinely are detected by se- Received 10 September 2011 quencing a few unlinked DNA loci. However, whether the patterns observed by multi-locus Received in revised form sequencing are compatible with genome wide data, such as ampliﬁed fragment length 29 February 2012 polymorphisms (AFLPs), is not well known for fungi. In this study we compared the Accepted 13 April 2012 ability of three DNA loci and AFLP data to discern between cryptic fungal lineages in the Corresponding Editor: three morphospecies Coniophora olivacea, Coniophora arida, and Coniophora puteana. The Joey Spatafora sequences and the AFLP data were highly congruent in delimiting the morphotaxa into multiple cryptic species. However, while the DNA sequences indicated introgression or hy- Keywords: bridization between some of the cryptic lineages the AFLP data did not. We conclude that AFLP as few as three polymorphic DNA loci was sufﬁcient to recognize cryptic lineages within Coniophora arida the studied Coniophora taxa. However, based on analyses of a few (three) sequenced loci Coniophora olivacea the hybridization could not easily be distinguished from incomplete lineage sorting. Hence, Coniophora puteana great caution should be taken when concluding about hybridization based on data from Hybridization just a few loci. Incomplete lineage sorting ª 2012 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. Multi-locus sequencing

Introduction (genealogical concordance phylogenetic species recognition, Taylor et al. 2000). Non-congruence of multiple gene trees The occurrence of cryptic species is a common phenomenon may reflect for example recombination, incomplete lineage in fungi as few morphological characters are available to dis- sorting or introgression. criminate among closely related taxa. Although the fruiting It has been debated how many DNA loci and gene trees that stage in most cases constitutes only a small part of fungi’s are necessary to recognize cryptic species (Rokas et al. 2003; life cycle, fruit body characteristics have often been used Rokas & Carroll 2005). Rokas et al. (2003) suggested that eight alone to define basidiomycete species. Multi-locus DNA se- to 20 sequenced loci were necessary to define a cryptic species. quencing have for the past decade been used to detect cryptic Nevertheless, in a study of yeasts including 106 genes, all clades lineages within fungal morphological species (Koufopanou were still not supported (Jeffroy et al. 2006). The number of loci et al. 1997; O’Donnell et al. 2000; Dettman et al. 2003; needed to define species will vary based on factors such as time Kauserud et al. 2006, 2007a,b; Jargeat et al. 2010; Carlsen et al. since divergence, effective population size, substitution rates, 2011). The recognition of the same phylogenetic groups across and whether partial or incomplete reproductive barriers exist. multiple gene trees indicates the presence of cryptic species The level of information (number of polymorphisms) per loci

* Corresponding author. Tel.: þ47 22854555; fax: þ47 22854726. E-mail addresses: [email protected], [email protected], [email protected], [email protected] 1878-6146/$ e see front matter ª 2012 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.funbio.2012.04.009

Please cite this article in press as: Skrede I, et al., Genome wide AFLP markers support cryptic species in Coniophora (Boletales), Fungal Biology (2012), doi:10.1016/j.funbio.2012.04.009 2 I. Skrede et al.

and whether the loci are influenced by selection or evolutionary All samples of C. arida and C. olivacea were analyzed twice constraints might also be important. The comparison to ge- for the primer combination 6-FAM-EcoR1-AC Mse1-CA to nome wide data such as amplified fragment length polymor- check for scoring error and to calculate an error rate according phisms (AFLPs) can indicate to what degree a few sequenced to Bonin et al. (2004). The scoring error for C. olivacea and loci can be used to recognize cryptic species. Comparison of C. arida was 0.022 and 0.025, respectively. Unfortunately, no AFLPs and multi-locus sequence data in the basidiomycete samples were rerun for C. puteana or for the other primer com- Serpula himantiodes have suggested that AFLPs may detect ge- binations, thus no scoring error could be obtained for the total netic variation on a finer level than a multi-locus sequences datasets. dataset including only three nuclear markers (Kauserud et al. A parsimony tree was constructed in TNT (Goloboff et al. 2006). From other organisms we know that AFLP has been suc- 2008) with the fragments as ‘taxa’ to visualize their differ- cessfully used in combination with sequences to increase the ences. AFLP markers that appeared identical in the parsimony resolution of phylogenetic relationships (e.g. Despres et al. tree were removed before further analyses. One of the major 2003; Mitter et al. 2011). problems with AFLPs is that each fragment’s identity is un- The genus Coniophora DC. (Basidiomycota, Boletales) com- known and we do not know whether fragments are linked, ho- prises about 20 morphologically defined wood decaying spe- mologous or independent fragments supporting a group. cies (Burt 1917; Ginns 1982; Kirk et al. 2008). Numerous Thus, removing potentially linked fragments also automati- cryptic species were detected in the three widespread mor- cally removes support for genetic groups. However, as the phological species, Coniophora arida (Fr.) P. Karst., Coniophora five primer combinations used in this study have few nucleo- olivacea (Fr.) P. Karst., and Coniophora puteana (Schumach.) tide differences we expected some homologous fragments, P. Karst. using multi-locus DNA sequencing of three DNA and therefore considered it important to remove all poten- loci (Kauserud et al. 2007a,b). In addition, Kauserud et al. tially linked fragments. (2007a,b) suggested hybridization among lineages in C. arida In total, the 20 samples of C. olivacea resulted in 511 poly- and C. puteana. However, whether genealogies of only three morphic loci. After linked loci were removed, the final ana- nuclear loci are sufficient to distinguish hybridization from in- lyzed dataset had 358 polymorphic AFLP loci. For the 23 complete lineage sorting is unknown. samples of C. arida 445 polymorphic AFLP loci were scored. Re- Here we evaluate whether just a few DNA loci are suffi- moval of potentially linked loci resulted in a dataset of 360 cient to detect cryptic species in Coniophora,bycomparing polymorphic loci. Of the 48 analyzed individuals of C. puteana the multi-locus sequence data to genome wide AFLP data- 782 polymorphic AFLP loci were scored. After removing linked sets obtained from the same isolates of C. arida, C. olivacea, markers 589 polymorphic loci were ready for further analyses. and C. puteana as used in Kauserud et al. (2007a,b). We also test whether the AFLP data indicate ongoing hybridization Data analyses or whether incomplete lineage sorting is a more plausible explanation for the observed results in Kauserud et al. The AFLP datasets were grouped into lineages with the Bayes- (2007a,b). ian model-based clustering method implemented in structure 2.3.1 (Pritchard et al. 2000; Falush et al. 2007; Hubisz et al. 2009). All isolates were treated as dikaryons and we scored Material and methods the dataset as recessive alleles. We used the admixture model with correlated allele frequencies, with 106 Marcov chain Monte Samples Carlo (MCMC) iterations and a burn-in of 105 iterations, and re- peated the analyses ten times for each K.Wefurtheranalyzed We analyzed 23 isolates of Coniophora arida,20ofConiophora the structure results using the software CorrSieve, available as olivacea and 48 of Coniophora puteana. All of the isolates have an R package (R Development Core Team 2004; Campana et al. been analyzed earlier with multi-locus sequencing and 2011). The number of groups (K ) in the dataset were estimated grouped into cryptic lineages (Table 1)inKauserud et al. based four graphs obtained from CorrSieve. These four graphs (2007a,b). are: 1) the average log likelihood of the data for each K (lnPD). AFLP analysis was carried out according to Kauserud et al. 2) The correlation between runs ( p-value; Campana et al. (2006) using DNA extracts previously described in Kauserud 2011; Cockram et al. 2008), meaning whether the same results et al. (2007a,b). The five AFLP primer combinations 6-FAM- are obtained for each of the ten replicates for each K (Q matrix EcoR1-AC Mse1-CA, 6-FAM-EcoR1-AC Mse1-TA, 6-FAM-EcoR1- in CorrSieve). 3) Delta K measures of the rate of change in log AC Mse1-GA, 6-FAM-EcoR1-AT Mse1-CA, and 6-FAM-EcoR1- likelihood of the data between successive K’s (Evanno et al. AT Mse1-GA were analyzed for all samples. The fragments 2005). 4) DeltaF compares the Fst among obtained groups for were analyzed on an ABI 3730 DNA analyzer (Life Technolo- each K (Campana et al. 2011). After estimating the K with high- gies, Foster City, USA) at the ABI-lab at the University of est statistical support using a combination of the four graphs Oslo (http://www.mn.uio.no/bio/english/research/about/in- explained above, we analyzed each sub-group separately to vi- frastructure/abi-lab/index.html). The raw data was visual- sualize whether substructure was present within the groups. ized, aligned with the internal size standard (GeneScan 500 The sub-data analyses were run five times for each K,with Rox, Life Technologies), manually controlled and analyzed 205 MCMC iterations and a burn-in of 105 iterations. in GeneMapper 4.0 (Life Technologies). Fragment in the size Principal coordinate analyses (PCOs) were conducted in range 50e500 base pairs were scored as present (1) or absent NTSYS-PC (Rohlf 1990) using the similarity index of Dice (0). (Dice 1945). We chose the latter to prevent any biasing effect

Table 1 e Isolates included in the study and their origin. Information of analyzed samples of Coniophora olivacea, C. arida and C. puteana. Column a indicates which lineage the samples belong to in Kauserud et al. (2007a,b), column b gives the number of heterozygote sites (from Kauserud et al. 2007a,b) and c the number of AFLP fragments (after correcting for missing data). Isolate Taxon Origin ab c Isolate Taxon Origin ab c

MUCL1000 C. puteana DEU 1 4 134 MUCL20566 C. olivacea DEU 6 1 80.8 MUCL30744 C. puteana BEL 1 1 142.4 P154 C. olivacea DEU 6 0 64 MUCL31020 C. puteana BEL 1 10 108 P161 C. olivacea CAN 3 3 114 MUCL20565 C. puteana GBR 2 0 110 82e34/3 C. olivacea NOR 3 1 109 MUCL28146 C. puteana NLD 1 2 134 DAOM145707 C. olivacea CAN 4 0 90 MUCL30396 C. puteana BEL 2 0 114 DAOM189127 C. olivacea AUS 5 1 64.3 MUCL30566 C. puteana BEL 1 8 125 FP100279 C. olivacea USA 3 0 63 P1 C. puteana DEU 1 3 135 FP100334 C. olivacea USA 3 0 81 P155 C. puteana DEU 1 2 118.5 FP104386 C. olivacea USA 2 3 102 P159 C. puteana SWE 1 2 125.7 L-9083 C. olivacea USA 3 0 51.6 P160 C. puteana SWE 1 0 137 L-9712 C. olivacea USA 3 4 94 P167 C. puteana DEU 1 3 139 MD-204 C. olivacea USA 1 0 89 P168 C. puteana DEU 1 1 137 MD-264 C. olivacea USA 1 0 99 BamEbw-109 C. puteana DEU 2 0 97 Mad4705 C. olivacea USA 1 0 89 59-1913/4 C. puteana NOR 1 1 130.5 RLG-6096 C. olivacea USA 3 0 87.1 52-1094/8 C. puteana NOR 1 7 137 FP105383 C. olivacea USA 2 0 93 82-97/3 C. puteana NOR 1 1 116 FPL1 C. olivacea GBR 4 0 69.2 81 C. puteana NZL 2 0 109 L-11386 C. olivacea USA 2 0 79.9 8 C. puteana USA 1 0 139 DAOM216045 C. olivacea CAN 4 0 79 76-77/2 C. puteana NOR 1 0 123.3 MUCL45999 C. olivacea CUB 1 0 97 CCBAS 524 C. puteana CZE 2 0 100 MUCL14244 C. arida BEL 1a 0 110 MUCL30793 C. puteana 2 0 105 MUCL30714 C. arida BEL 3 1 120 NBRC6275 C. puteana JPN 1 3 144 MUCL40342 C. arida ZIM 1b 0 126 FP 94445 C. puteana USA 3 1 97 MUCL30844 C. arida GBR 3 4 128 MD189 C. puteana USA 1 4 120 472 C. arida NZL 3 4 68.2 DAOM194147 C. puteana CAN 2 8 128 MUCL31038 C. arida BEL 3 0 89 # 309 C. puteana CAN 3 5 75.4 DAOM176418 C. arida CAN 1a 8 137 FP100258 C. puteana USA 1 0 121 RLG6926 C. arida USA 1b 0 81 FP104403 C. puteana USA 1 11 132 RLG-6888 C. arida USA 1a 25 117.4 RLG-5410 C. puteana USA 1 2 119 FP97436 C. arida USA 1b 0 110 S31 C. puteana USA 1 3 128.5 FP94481 C. arida USA 1a 7 120 DAOM137404 C. puteana CAN 1 0 98 FP90087 C. arida USA 1a 0 119 DAOM21055 C. puteana CAN 1 10 112.5 FP104424 C. arida USA 1b 2 128 DAOM31271 C. puteana CAN 3 18 103 FP103793 C. arida USA 1b 0 121.3 DAOM52883 C. puteana IND 2 0 98 FP100307 C. arida USA 2 5 109 DAOM138703 C. puteana CAN 2 1 106 DAOM52839 C. arida CAN 1a 5 113 DAOM137693 C. puteana CAN 3 0 104 DAOM138685 C. arida CAN 2 0 93 DAOM145623 C. puteana CAN 1 0 124 DAOM137396 C. arida CAN 1a 3 112 DAOM137908 C. puteana USA 1 8 113.1 FP103970 C. arida USA 1b 0 110.7 DAOM137779 C. puteana USA 3 6 116 FP56488 C. arida USA 1a 2 113.7 DAOM147445 C. puteana CAN 1 0 105 FP105382 C. arida USA 1a 8 104.8 DAOM196038 C. puteana RUS 1 4 128 DAOM196973 C. arida CAN 2 5 88.4 DAOM198027 C. puteana CAN 3 18 89 DAOM175779 C. arida CAN 1a 4 118.7 DAOM17535 C. puteana CAN 1 0 108 DAOM137697 C. puteana CAN 1 0 108 DAOM216046 C. puteana CAN 2 2 117 DAOM102773 C. puteana CAN 3 18 100 FP105938 C. puteana USA 1 19 112

of combining monospore cultures with dikaryotic mycelium used in Kauserud et al. (2007a) were included in the AFLP in the same analysis. This approach has proven successful data and the phylogenetic trees were therefore adopted from in earlier studies of individuals with varying karyotypes Kauserud et al. (2007a). (Jørgensen et al. 2008; Carlsen et al. 2010). Phylogenetic analysis of the multi-locus sequence data for C. puteana from Kauserud et al. (2007a) were performed in phy- Results and discussion logenetic analysis using parsimony (PAUP) (Swofford 2002) using only isolates included in the AFLP analyses to make sure AFLPs and multi-locus sequences gave largely a consistent the groupings in the parsimony can be compared to the pattern of dividing the three morphospecies into cryptic spe- AFLP data. For C. olivacea and Coniophora arida all isolates cies (Figs 1 and 2). However, the number of inferred genetic

groups from the STRUCTURE analyses was different than the (Jeffroy et al. 2006). Our analyses indicate that a few sequenced multi-locus phylogenies (Fig 1; Supp. Fig 1). According to the loci, in our case three, might be sufficient to separate between structure analysis the optimal delineation of the Coniophora cryptic species. However, this obviously depends on how re- olivacea AFLP dataset was four genetic groups (Fig 1A; Supp. cently the cryptic lineages diverged. Fig 1). The same four groups can easily be recognized in the Based on the DNA sequence data it was earlier hypothe- principal coordinate analysis (PCO; Fig 2). Based on the sized that hybridization might have occurred between various multi-locus DNA phylogeny, Kauserud et al. (2007a) recognized lineages in the C. arida and C. puteana species complexes six cryptic species in C. olivacea. The structure analyses of the (Kauserud et al. 2007a,b). In a few dikaryotic isolates of C. arida sub-datasets identified five groups, but could not place the iso- and C. puteana, highly divergent alleles with affinity to differ- late representing group 5 in Kauserud et al. (2007a; Fig 1A; ent sub-groups or presumed cryptic species co-existed. These Supp. Fig 2). When K ¼ 6 was applied on the full dataset in observations were interpreted as introgression of alleles structure, the same six groups was found in the AFLP dataset across lineages. However, the AFLP data did not support as in the multi-locus DNA phylogeny (Fig 1A). These six groups this, as the same putative hybrid isolates neither appeared were also separated in the PCO plot (Fig 2A). as admixed multi-locus AFLP genotypes in the structure anal- In Coniophora arida, two groups and a third admixed group yses nor in the PCO plots. In addition, we did not find more were recognized in structure when analyzing the full dataset AFLP fragments in the putative hybrids as expected for hy- (Fig 1B, Supp. Fig 1). After analyzing the sub-datasets separately, brids (Table 1). Notably, a recombinant ITS type was detected these two groups (both including the admixed group) were sep- in C. puteana (PS3) that combined sequence motifs from differ- arated into altogether five groups (Fig 1B, Supp. Fig 2). These five ent lineages (PS1 and PS3), but this isolate did not constitute were also apparent in the PCO analyses (Fig 2B). The three a mixed AFLP genotype as could be expected. In light of the groups recognized using K ¼ 2 correspond well with the multi- new genome wide data we find it more likely that the dikary- locus sequence phylogeny (Fig 1B). Results from the five groups otic isolates previously interpreted as hybrids includes an- recognized using the sub-datasets further subdivide ‘Group1’ cient allelic variants not sorted out by genetic drift into three groups. ‘Group 1a and 1b’ in the multi-locus sequence (i.e. incomplete lineage sorting). phylogeny was not recognized by the AFLP data. In some cases, the structure analyses indicated that other Also a good correspondence was observed in the Coniophora isolates (not hypothesized as hybrids by the DNA sequences) puteana complex between the AFLP data and the multi-locus were admixed, combining genetic components from different phylogeny (Fig 1C). For K ¼ 2 two groups and a third admixed groups. The isolate representing group 5 in the multi-locus group were recognized by structure (Fig 1C). These three phylogeny of C. olivacea is one such example. In the K ¼ 4 anal- groups separate Phylogenetic Species S1 into two groups, ysis in structure, this isolate came out as highly admixed and group PS2 and PS3 from Kauserud et al. (2007b) into one (Fig 1A), and the structure analyses of the sub-dataset could group (Fig 1C). Separate structure analyses of the two sub- not place the isolate in either group. However, using other set- groups (the admixed included in both subgroups) recognize tings of the full dataset (K ¼ 6) this isolate represents a distinct the admixed group as a separate group, PS2 and PS3 as sepa- non-admixed group. Hence, concerning putative admixed rate groups and further separate PS2 into two groups (Supp. AFLP genotypes, we would argue that results from various Fig 2; Fig 1C). These groups are identical as using K ¼ 6on runs of structure using different K’s must be evaluated in the full dataset (Supp. Fig 1; Fig 1C). parallel. Overall, the main genetic structure observed in Kauserud Also in other fungi, hybridization has been hypothesized et al. (2007b) with five cryptic species in C. arida, six in C. oliva- based on co-occurrence of divergent alleles. Hybridization cea and three in C. puteana were recognized in the PCO plots was suggested within the Tricholoma scalpturatum complex as and the structure analyses of the genome wide AFLP data, three loci (ITS, gpd, and tef ) were incongruent and a recombi- but the hierarchy of the groups were somewhat different. It nant ITS sequence was observed (Jargeat et al. 2010). Very sim- has been claimed that analyzes of a high number of indepen- ilar to our observation in C. puteana, a recombinant ITS dent loci are necessary to resolve the relationship among sequence was also found between Flammulina rossica and closely related species because of factors like incomplete lin- Flammulina velutipes (Hughes & Petersen 2001). It would be in- eage sorting (Gee 2003; Rokas et al. 2003). Recent acquirement teresting to see whether complimentary data would support if of many fungal genomes has opened the possibility to do phy- these specimens represent hybrids or whether incomplete lin- logenomics, and a study of 106 genes in yeast showed that eage sorting also may account for their observations. Because even when many markers were included, the phylogenetic many fungi have large population sizes, genetic drift may signal did not improve with the number of sequences work more slowly compared to other organisms. Ancestral

Fig 1 e STRUCTURE analyses of the AFLP datasets, and comparison to the multi-locus sequence datasets results of sTRUCTURE analyses based on AFLP data from the three species A) Coniophora olivacea,B)C. arida, and C) C. puteana, and comparison the previously published multi-locus sequence data. The columns to the left represent parsimony trees, as observed in Kauserud et al. (2007a,b). Branches with bootstrap support above 75 are indicated with black dots. *Indicates isolates suggested to be hybrids in Kauserud et al. (2007a,b). The mid column is the results from STRUCTURE analyses of the AFLP data using all individuals. The bar graphs represent the probability of assignment of each individual to a group. The column to the far right is the probability of assignment of an individual after analyzing subsets of the data. Bold text indicates the K with highest statistical support.

(Conifer root rot) (Spiers & Hopcroft 1994; Brasier 2000; Newcombe et al. 2000; Linzer et al. 2008; Brasier & Kirk 2010). Introgression between Heterobasidion annosum into the invasive Heterobasidion irregulare species in Italy was recently confirmed using AFLP and multi-locus sequencing (Gonthier et al. 2007; Gonthier & Garbelotto 2011). Coniophora is also prone to human-mediated dispersal (through for example international trade of construction wood) resulting in large geographic distri- bution size of each cryptic species, but apparently not hybridization. Based on mating studies, the presence of four intersterility groups has earlier been observed in C. puteana (Ainsworth & Rayner 1990), which largely corroborate our observations. To our knowledge, similar barriers to gene flow have not been investigated in C. arida and C. olivacea by use of mating studies. Some of the presumed cryptic species of C. arida, C. olivacea,andC. puteana have overlapping geographic ranges, which would have enabled hybridization (Kauserud et al. 2007a,b). There are potentially numerous pre- and postzy- gotic barriers that hinder interspecific gene flow in fungi. Although interspecific crosses of homokaryons may result in dikaryons (e.g. Kauserud et al. 2007c), the dikaryons may be un- able to fructify, the resulting spores may be sterile or the F1 hybrid genome may not be able to survive due to e.g. lack of substrate specificity (Garbelotto et al. 1998). To conclude, our results suggest that a few numbers of sequenced loci can be enough to reveal the presence of cryptic species in fungi. However, caution should be taken when interpreting the co-existence of divergent and recombinant alleles as hybrids. Preferably to reveal hybridization, different types of DNA markers should be analyzed in parallel, representing both nuclear and mitochondrial DNA.

Acknowledgments

This study is supported by the Research Council of Norway (NFR186174/V4) and the University of Oslo. Jørn Henrik Sønstebø and two anonymous reviewers are thanked for help- ful comments on the manuscript.

Fig 2 e PCOs of the three species C. olivacea, C. arida and C. Appendix A. Supplementary material puteana based on AFLP data. The plots are colored based on the groups recognized in the STRUCTURE analyses for easy Supplementary data related to this article can be found comparison. The dotted lines deﬁne the statistically highest online at doi:10.1016/j.funbio.2012.04.009. supported groups using the full dataset.

references polymorphisms might then be maintained for a long time in closely related fungi. Another valuable approach would be to Ainsworth AM, Rayner ADM, 1990. Mycelial interactions and compare uni-parental (mtDNA) and bi-parentally (nDNA) outcrossing in the Coniophora puteana complex. Mycological inherited markers in parallel. Using both these types of Research 94: 627e634. markers may discern between introgression and incomplete Anderson JB, Kohn LM, 2007. Dikaryons, diploids and evolution. lineage sorting because divergence will affect both loci while In: Heitman J, Kronstad JW, Taylor JW, Casselton LA (eds), introgression may affect nuclear and mitochondrial markers Sex in Fungi: molecular determination and evolutionary implica- e differently (Anderson & Kohn 2007). tions. ASM Press, Washington DC, pp. 333 348. Bonin A, Bellemain E, Eidesen PB, Pompanon F, Brochmann C, Natural hybridization in fungi appears to be rare, and many Taberlet P, 2004. How to trace and assess genotyping errors in of the suggested fungal hybrids are plant pathogens prone to population genetics studies. Molecular Ecology 13: 3261e3273. human-mediated dispersal, including Ophistoma (Dutch Elm Brasier C, 2000. Plant pathology e the rise of the hybrid fungi. disease), Melampsora (Poplar leaf rust), and Heterobasidion Nature 405: 134e135.

Brasier CM, Kirk SA, 2010. Rapid emergence of hybrids between Kauserud H, Stensrud O, Decock C, Shalchian-Tabrizi K, the two subspecies of Ophiostoma novo-ulmi with a high level of Schumacher T, 2006. Multiple gene genealogies and AFLPs pathogenic fitness. Plant Pathology 59: 186e199. suggest cryptic speciation and long-distance dispersal in the Burt EA, 1917. The Thelephoraceae of North America. VIII. basidiomycete Serpula himantioides (Boletales). Molecular Ecology Coniophora. Annals of Missouri Botanical Garden 4: 237e269. 15: 421e431. Campana MG, Hunt HV, Jones H, White J, 2011. CorrSieve: soft- Kauserud H, Shalchian-Tabrizi K, Decock C, 2007a. Multilocus ware for summarizing and evaluating Structure output. sequencing reveals multiple geographically structured line- Molecular Ecology Resources 11: 349e352. ages of Conioiphora arida and C. olivacea (Boletales) in North Carlsen T, Elven R, Brochmann C, 2010. The evolutionary history of America. Mycologia 99: 705e713. Beringian Smelowskia (Brassicaceae) inferred from combined Kauserud H, Svegarden IB, Decock C, Hallenberg N, 2007b. Hy- microsatellite and DNA sequence data. Taxon 59:427e438. bridization among cryptic species of the cellar fungus Conio- Carlsen T, Engh IB, Decock C, Rajchenberg M, Kauserud H, 2011. phora puteana (Basidiomycota). Molecular Ecology 16: 389e399. Multiple cryptic species with divergent substrate affinities in the Kauserud H, Svegarden IB, Saetre GP, Knudsen H, Stensrud Ø, Serpula himantioides species complex. Fungal Biology 115:54e61. Schmidt O, Doi S, Sugiyama T, Hogberg€ N, 2007c. Asian origin Cockram J, White J, Leigh FJ, Lea VJ, Chiapparino E, Laurie DA, and rapid global spread of the destructive dry rot fungus Mackay IJ, Powell W, O’Sullivan DM, 2008. Association map- Serpula lacrymans. Molecular Ecology 16: 3350e3360. ping of partitioning loci in barley. BMC Genetics 9: 16. Kirk P, Cannon P, Minter D, Stalpers J, 2008. Dictionary of the Fungi. Dettman JR, Jacobson DJ, Taylor JW, 2003. A Multilocus genea- CABI, Wallingford. logical approach to phylogenetic species recognition in the Koufopanou V, Burt A, Taylor JW, 1997. Concordance of gene model eukaryote Neurospora. Evolution 57: 2703e2720. genealogies reveals reproductive isolation in the patho- Despres L, Gielly L, Redoutet B, Taberlet P, 2003. Using AFLP to genic fungus Coccidioides immitis. Proceedings of the National resolve phyogenetic relationships in a morphologically diver- Academy of Sciences of the United States of America 94: sified plant species complex when nuclear and chloroplast 5478e5482. sequences fail to reveal variability. Molecular Phylogenetics and Linzer RE, Otrosina WJ, Gonthier P, Bruhn J, Laflamme G, Evolution 27: 185e196. Bussieres G, Garbelotto M, 2008. Inferences on the phylo- Dice LR, 1945. Measures of the amount of ecologic association geography of the fungal pathogen Heterobasidion annosum, in- between species. Ecology 26: 297e302. cluding evidence of interspecific horizontal genetic transfer Evanno G, Regnaut S, Goudet J, 2005. Detecting the number of and of human-mediated, long-range dispersal. Molecular clusters of individuals using the software Structure: a simula- Phylogenetics and Evolution 46: 844e862. tion study. Molecular Ecology 14: 2611e2620. Mitter KT, Larsen TB, de Prins W, de Prins J, Collins S, Weghe GV, Falush D, Stephens M, Pritchard JK, 2007. Inference of population Safian S, Zakharov EV, Hawthorne DJ, Kawahara AY, Regier JC, structure using multilocus genotype data: dominant markers 2011. The butterfly subfamily Pseudopontiinae is not mono- and null alleles. Molecular Ecology Notes 7: 574e578. basic: marked genetic diversity and morphology reveal three Garbelotto M, Otrosina WJ, Cobb FW, Bruns TD, 1998. The Euro- new species of Pseudopontia (Lepidoptera: Pieridae). Systematic pean S and F intersterility groups of Heterobasidion annosum Entomology 36: 139e163. may represent sympatric protospecies. Canadian Journal of Newcombe G, Stirling B, McDonald S, Bradshaw HD, 2000. Botany 76: 397e409. Melampsora x columbiana, a natural hybrid of M. medusae and Gee H, 2003. Evolution e ending incongruence. Nature 425 M. occidentalis. Mycological Research 104: 261e274. 782e782. O’Donnell K, Kistler HC, Tacke BK, Casper HH, 2000. Gene gene- Ginns J, 1982. A Monograph of the genus Coniophora (Aphyllo- alogies reveal global phylogeographic structure and repro- phorales, Basidiomycetes). Opera Botanica 61:1e61. ductive isolation among lineages of Fusarium graminearum, the Goloboff PA, Farris JS, Nixon KC, 2008. TNT, a free program for fungus causing wheat scab. Proceedings of the National Academy phylogenetic analysis. Cladistics 24: 774e786. of Sciences of the United States of America 97: 7905e7910. Gonthier P, Nicolotti G, Linzer R, Guglielmo F, Garbelotto M, 2007. Pritchard JK, Stephens M, Donnelly P, 2000. Inference of popula- Invasion of European pine stands by a North America forest tion structure using multilocus genotype data. Genetics 155: pathogen and its hybridization with native interfertile taxon. 945e959. Molecular Ecology 16: 1389e1400. R Development Core Team, 2004. A Language and Environment for Gonthier P, Garbelotto M, 2011. Amplified fragment length poly- Statistical Computing. R Foundation for Statistical Computing, morphism and sequence analyses reveal massive gene intro- Vienna. gression from the European fungal pathogen Heterobasidion Rohlf F, 1990. NTSYS-PC. Numerical Taxonomy and Multivariate annosum into its introduced congener H. irregulare. Molecular Analysis System. Exeter Software, Setauket, NY. Ecology 20: 2756e2770. Rokas A, Carroll SB, 2005. More genes or more taxa? The relative Hubisz MJ, Falush D, Stephens M, Pritchard JK, 2009. Inferring contribution of gene number and taxon number to phylo- weak population structure with the assistance of sample genetic accuracy. Molecular Biology and Evolution 22: group information. Molecular Ecology Resources 9: 1322e1332. 1337e1344. Hughes KW, Petersen RH, 2001. Apparent recombination or gene Rokas A, Williams BL, King N, Carroll SB, 2003. Genome-scale conversion in the ribosomal ITS region of a Flammulina (fungi, approaches to resolving incongruence in molecular phyloge- Agaricales) hybrid. Molecular Biology and Evolution 18:94e96. nies. Nature 425: 798e804. Jargeat P, Martos F, Carriconde F, Gryta H, Moreau PA, Gardes M, Spiers AG, Hopcroft DH, 1994. Comparative studies of the poplar 2010. Phylogenetic species delimitation in ectomycorrhizal rusts Melampsora medusae, M. larici-populina and their inter- fungi and implications for barcoding: the case of the Tricho- specific hybrid M. medusae-populina. Mycological Research 98: loma scalpturatum complex (Basidiomycota). Molecular Ecology 889e903. 19: 5216e5230. Swofford DL, 2002. PAUP. Phylogenetic Analyses using Parsimony. Jeffroy O, Brinkmann H, Delsuc F, Philippe H, 2006. Phylogenomics: Sinauer Associates, Sunderland. the beginning of incongruence? Trends in Genetics 22: 225e231. Taylor JW, Jacobson DJ, Kroken S, Kasuga T, Geiser DM, Jørgensen MH, Carlsen T, Skrede I, Elven R, 2008. Microsatellites Hibbett DS, Fisher MC, 2000. Phylogenetic species recognition resolve the taxonomy of the polyploid Cardamine digitata ag- and species concepts in fungi. Fungal Genetics and Biology 31: gregate (Brassicaceae). Taxon 57: 882e892. 21e32.

Please cite this article in press as: Skrede I, et al., Genome wide AFLP markers support cryptic species in Coniophora (Boletales), Fungal Biology (2012), doi:10.1016/j.funbio.2012.04.009