This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy Fungal Genetics and Biology 46 (2009) 642–656 Contents lists available at ScienceDirect Fungal Genetics and Biology journal homepage: www.elsevier.com/locate/yfgbi A polyphasic approach to the taxonomy of the Alternaria infectoria species–group Birgitte Andersen a,*, Jens Laurids Sørensen a, Kristian Fog Nielsen a, Bert Gerrits van den Ende b, Sybren de Hoog b a b Center for Microbial Biotechnology (CMB), Department of Systems Biology, Building 221, Technical University of Denmark, DK-2800 Kgs. Lyngby, Denmark Centraalbureau voor Schimmelcultures (CBS-KNAW), Fungal Biodiversity Centre, P.O. Box 85167, 3508 AD Utrecht, The Netherlands a r t i c l e i n f o Article history: Received 20 December 2008 Accepted 28 May 2009 Available online 6 June 2009 Keywords: Chemical classification Fungi Haplotypes Molecular cladification Morphology Multivariate statistics Recombinants a b s t r a c t Different taxa in the species–group of Alternaria infectoria (teleomorph Lewia spp.) are often isolated from various cereals including barley, maize and wheat grain, ornamental plants and skin lesions from animals and humans. In the present study we made a polyphasic characterization of 39 strains morphologically identifiable as belonging to the A. infectoria species–group together with 12 strains belonging to closely related species: Alternaria malorum (syn. Cladosporium malorum), Chalastospora cetera (syn. Alternaria cetera) and Embellisia abundans. Morphological examination separated the 51 strains in three groups based on conidial appearance and arrangement: the A. infectoria species–group, E. abundans and a group containing C. cetera and A. malorum. The metabolite analyses on three different media showed two clusters, one containing all 39 A. infectoria species–group strains and one containing 10 strains of E. abundans, C. cetera and A. malorum. One E. abundans strain and one A. malorum strain were not included due to insufficient metabolite production. The separation of the A. infectoria species–group from E. abundans, C. cetera and A. malorum resulted mainly from the ability to produce altertoxins and novae-zelandins. The metabolite analyses also showed that all 51 strains were able to produce infectopyrones. The metabolite profiles of C. cetera and A. malorum were very similar with several metabolites of unknown structure in common. This is the first time that E. abundans, C. cetera and A. malorum have been reported as producers of infectopyrones. Sequence analyses of the internal transcribed spacer region (ITS), glyceraldehyde-3-phosphate dehydrogenase (gpd) and translocation elongation factor 1a (tef-1a) showed two clades: one with the 39 strains from the A. infectoria species–group and one with the 12 strains of E. abundans, C. cetera and A. malorum. The polyphasic approach in this study suggests that A. malorum var. polymorpha and the eight A. malorum strains do not belong in Alternaria, but in Chalastospora, however, as several distinct species. Splits Tree alignment of gpd sequences of 38 strains belonging to the A. infectoria species–group indicates that only three strains showed signs of recombination, while the remaining strains appeared to be clonal. Long term incubation at 7 °C in the dark showed that 12 out of 33 tested strains from the A. infectoria species–group were able to produce proascomata in axenic culture, but with no mature ascospores after 6 months. These findings suggest that Lewia/A. infectoria species–group must, at least in part, be homothallic. The results presented in this study show that ITS, tef-1a and gpd do not reflect ecology, secondary metabolism or morphology of the A. infectoria species–group and that molecular cladification and phylogeny cannot predict pathogenicity, host specificity or mycotoxin production. Ó 2009 Elsevier Inc. All rights reserved. 1. Introduction The Alternaria infectoria EG Simmons species–group sensu Simmons (Simmons and Roberts, 1993; Simmons, 2007) comprises more than 30 named anamorph taxa, among which Alternaria arbusti EG Simmons, Alternaria ethzedia EG Simmons, A. infectoria, Alternaria intercepta EG Simmons, Alternaria metachromatica EG Simmons, Alternaria oregonensis EG Simmons, Alternaria photistica EG Simmons, Alternaria triticimaculans EG Simmons, Alternaria tri* Corresponding author. Fax: +45 4588 4922. E-mail address: ba@bio.dtu.dk (B. Andersen). 1087-1845/$ - see front matter Ó 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.fgb.2009.05.005 ticina, Alternaria viburni EG Simmons are some (Simmons, 2007). It is the only group in Alternaria where some members have a teleomorph state, Lewia ME Barr and EG Simmons (Simmons, 1986). Morphologically, the A. infectoria species–group differs from other Alternaria species–groups in the three-dimensional sporulation pattern (Simmons and Roberts, 1993). Characteristic for the A. infectoria species–group is the production of small conidia (up to 70 lm in length) in branched chains with long, geniculate multilocus secondary conidiophores (up to 120 lm) between conidia (Simmons, 2007). Chemically, the A. infectoria species–group is very different from other Alternaria species, producing metabolites that are not found Author's personal copy B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 in other species–groups (Andersen and Thrane, 1996). None of the taxa in the A. infectoria species–group has ever been shown to produce alternariols or tenuazonic acid, which are common in other small-spored Alternaria species (Andersen et al., 2002) or altersolanols, common in some large-spored Alternaria (Andersen et al., 2008). On the other hand, taxa in the A. infectoria species–group produce infectopyrones and novae-zelandins (Christensen et al., 2005), which have never been detected in other Alternaria species–groups. However, infectopyrones have been found in other genera, such as Phoma Sacc., Stemphylium Wallr. and Ulocladium Preuss (Pedras and Chumala, 2005; Christensen et al., 2005; Andersen and Hollensted, 2008, respectively). Molecularly, taxa in the A. infectoria species–group have been analyzed at the sub-genus level using the ribosomal internal transcribed spacer region (ITS), glyceraldehyde-3-phosphate dehydrogenase (gpd) and translocation elongation factor 1a (tef-1a) sequences and the results showed that the A. infectoria species– group constitutes a quite distinct clade (de Hoog and Horré, 2002; Pryor and Bigelow, 2003). Another ITS sequence analysis creating an unrooted radial tree based on maximum likelihood calculation showed that Alternaria malorum (Ruehle) U. Braun, Crous and Dugan (Cladosporium malorum Ruehle), and the A. infectoria species–group comprised a single clade (Braun et al., 2003). An inquiry on Alternaria in the CBS database of ITS sequences showed that a strain of Chalastospora cetera (EG Simmons) EG Simmons (Alternaria cetera EG Simmons) and three strains of Embellisia abundans EG Simmons also clustered in the same clade as A. malorum and the A. infectoria species–group (unpublished results). Many taxa in the A. infectoria species–group are associated with various species in the grass family (Poaceae L.). They have been reported from stems, straw, leaves and grains of oat, barley, wheat and rye (Simmons, 1986; Andersen et al., 2002; Dugan and Peever, 2002; Perelló et al., 2008) and are also known to occur on maize (unpublished results). A. triticina Prasada and Prabhu, a known plant pathogenic species in the A. infectoria species–group, was first reported on wheat in India (Prasada and Prabhu, 1962) and later on the same host plant in Argentina (Perelló and Sisterna, 2006) and in Iran (Simmons, 2007). Other species isolated from discrete lesions of non-poaceae host plants, such as A. viburni and A. ethzedia, are presumed to have various degrees of pathogenicity (Simmons, personal commun.). In the last decade, taxa in the A. infectoria species–group have increasingly been isolated from human cutaneous infections, especially from immuno-compromised patients (de Hoog and Horré, 2002; Dubois et al., 2005; unpublished results). The objective of this work was to prove the hypothesis that taxa in Lewia/A. infectoria species–group are sexual fungi and that molecular sequence analysis and metabolite profiling will yield a number of clades and clusters that will correspond to the number of morphological species in the group. Previous research on other genera has shown that sequence analyses reveal cryptic molecular species (Taylor et al., 2000; O’Donnell et al., 2004). Studies on Penicillium and large-spored Alternaria have shown that results from metabolite profiling correlates with the morphological species concept (Andersen et al., 2008; Frisvad and Samson, 2004). Other research on Aspergillus and Stachybotrys has shown that molecular sequence analyses correlated with the morphological species concept and metabolite profiling (Samson et al., 2007; Andersen et al., 2003). Furthermore, controversy over whether the A. infectoria species–group comprises many species based on morphological differences or consists of only one species based on molecular sequence analysis has arisen. Therefore, this polyphasic study of the A. infectoria species–group, including molecular sequence analysis, metabolite profiling and mating tests, was set up. The study also included A. malorum, since this species had been reported to belong to the A. infectoria species–group (Braun et al., 2003), together 643 with C. cetera and E. abundans. One aim was to compare chemical and molecular findings with the current morphological classification and examine which factors correlate and which could resolve and segregate Alternaria from Chalastospora and Embellisia. Another aim was to test if other isolates than the human opportunists were able to grow at 37 °C. 2. Materials and methods 2.1. Fungal strains Fifty-one fungal strains belonging to the A. infectoria species– group (39), A. malorum (9), C. cetera (1), and E. abundans (2) were used. Identity, species–group affiliation, identification number, host plant, and geographic origin of all strains are given in Table 1. Strains are available from CBS collection, Fungal Biodiversity Centre, The Netherlands, and IBT collection, Department of Systems Biology, DTU, Denmark. 2.2. Morphological examination For morphological examination and DNA analysis, each strain was inoculated in three points onto a PCA plate (potato carrot agar; Simmons, 2007) and incubated under standardized conditions (Andersen et al., 2005). In brief: after inoculation the unsealed plates were incubated in one layer for 7 days at 23 °C under an alternating light/dark cycle consisting of 8 h of cool-white fluorescent daylight (tubes: TLD, 36 W/95, Philips, Denmark) and 16 h darkness. Slides for microscopy were made after 7 days using transparent tape preparations (Butler and Mann, 1959) mounted in lactophenol. All unidentified strains were compared with type cultures and descriptions according to Simmons (2007). For ascoma production under laboratory conditions, strains were transferred to another PCA plate that had been equipped with autoclaved wooden toothpicks to encourage the production of ascomata. In the first trial, each plate was divided into three sectors with three toothpicks. Three different strains were inoculated on the same plate, one in each sector. In the second trial, each plate was divided into two sectors with one toothpick. The same strain was inoculated in both sectors on the same plate. The plates were first incubated as ordinary PCA plates under alternating light at 23 °C for 2 weeks. Then the plates were sealed with para filmTM, placed up-side-down, and incubated at 7 °C in the dark for 6 months. For the ability to grow at high temperature, strains were transferred (three points) to PDA plates (potato dextrose agar; Difco, 213400). Plates were allowed to stand at room temperature for one day to ensure that all strains were viable and growing. The edge of the colonies was marked before incubation. After 12 days at 37 °C the colony edges were marked again and the plates were allowed to stand for two days at room temperature. Strains that had resumed their original growth characteristics were recorded as positive. 2.3. Metabolite extraction For metabolite analyses, each strain was transferred (three points) onto a DRYES plate (dichloran Rose Bengal yeast extract sucrose agar; Frisvad, 1983), a DG18 plate (dichloran 18% glycerol agar [31.5 g/l dichloran glycerol agar base (OXOID, CM0729); 220 ml/l anhydrous glycerol (JT Baker, 7044); 10 mg/l ZnSO47H2O (Merck, 8883); 5 mg/l CuSO45H2O (Riedel-de Haën, 12849); 50 mg/l chloramphenicol (Sigma, C-0378); 50 mg/l chlortetracycline hydrochloride (Sigma, C-4881)]), and a PDA + DN plate (potato dextrose agar [39.0 g/l potato dextrose agar (Difco, 213400), Author's personal copy 644 B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 Table 1 Host, origin and species–group affiliation of the Alternaria, Chalastospora and Embellisia strains and type cultures (T) used in this study. #a Genus species species–groupb Host Origin CBS # BA # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 43 44 45 46 47 48 49 50 51 52 A. infectoria sp.–grp A. triticina T A. infectoria T A. oregonensis T A. photistica T A. ethzedia T A. metachromatica T A. triticimaculans T A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. intercepta T A. viburni T A. arbusti T A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. C. cetera T A. malorum A. malorum A. malorum A. malorum A. malorum A. malorum var. polymorpha A. malorum A. malorum A. malorum E. abundans E. abundans T A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. Hordeum, grain Triticum Triticum Triticum Digitalis Brassica Triticum Triticum Hordeum Hordeum Hordeum Viburnum Viburnum Pyrus Triticum, grain Triticum, grain Hordeum, grain Hordeum, grain Hordeum, grain Elymus Soil – – – Soil Vitis Triticum, straw Triticum, grain Gossypium, seed Dianthus, seed Fragaria – Zea, fresh silage Zea, fresh silage Zea, fresh silage Zea, fresh silage Zea, fresh silage Zea, fresh silage Zea, fresh silage Zea, fresh silage Zea, fresh silage Zea, fresh silage Zea, fresh silage Homo, skin lesion Hordeum Paeonia Avena, straw Homo, skin lesion Homo, skin lesion Homo, skin lesion Zea, fresh silage DK, Jutland India UK USA, OR UK Switzerland Australia Argentina Italy NZ NZ Europe Europe USA, CA DK DK DK DK DK Australia Syria USA NZ NZ Lebanon USA, WA SA, Cape Prov. USA, OR Turkey UK NZ UK DK, Jutland DK, Jutland DK, N Jutland DK, N Jutland DK, NW Jutland DK, Jutland DK, S Jutland DK, Jutland DK, Jutland DK, N Jutland DK, E Jutland Germany UK NL – Italy NL Austria DK, N Jutland 120147 763.84 112250 542.94 212.86 197.86 553.94 578.94 704 1207 1209 1210 1211 1215 1218 1221 1228 1234 1239 1258 1259 1263 1286 1287 1294 1312 1315 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 JLS08 JLS09 JLS32 JLS33 a b inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. inf. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. sp.–grp. 119406 119407 596.93 120148 120149 120150 120151 120152 110898 173.80 148.66 114810 114809 900.87 112048 266.75 216.65 540.75 535.83 534.83 160.79 EGS # 17–061 27–193 29–194 35–172 37–143 38–132 41–050 42–086 43–070 43–160 49–137 49–147 91–136 41–072 34–063 29–159 102692 116001 106.52 308.53 109785 110803 110804 JLS38 Analysis number in this study. #42 is not included. Member of the A. infectoria species–group based on morphology. 10 mg/l ZnSO47H2O (Merck, 8883) and 5 mg/l CuSO45H2O (Riedel-de Haën, 12849)] with additional 30 g/l dextrose (BHD, 10117) and additional 3 g/l NaNO3 (Merck, A855537 811). After inoculation, the DRYES, DG18, PDA + DN plates were packed in perforated plastic bags and incubated in darkness at 25 °C prior to extraction. Metabolites were extracted separately from 14-day-old DRYES, DG18 and PDA + DN cultures. The extractions were done using a micro-scale extraction method modified for Alternaria metabolites (Andersen et al., 2005). In brief: three 6-mm agar plugs were cut from the center of the three colonies and the nine plugs were placed in a 2-ml vial. Plugs were extracted with 1.0 ml ethyl acetate containing 1% formic acid (v/v) and ultrasonicated for 60 min. Extracts were transferred to clean 2-ml vials, evaporated to dryness, re-dissolved ultrasonically in 400 ll methanol, and filtered through 0.45-lm PFTE filters (National Scientific Company, Rockwood, TN, USA) into clean 2-ml vials prior to HPLC analysis. Experiments were repeated once on DRYES a year after with all strains and for strains with low metabolite production a third time on DRYES a month later using 18 plugs. 2.4. HPLC-UV–VIS analyses Metabolite profiling was performed on an Agilent 1100 HPLC system (Agilent, Waldbronn, Germany) equipped a diode array detector collecting two ultraviolet–visible (UV–VIS) spectra per sec from 200 to 600 nm. Samples of 3.0 ll were separations on a 2  100 mm Luna 3 lm C18(2) (Phenomenex, Torrance, CA, USA) at 40 °C using a linear water–acetonitrile gradient and a flow of 0.4 ml/min. The gradient started at 15% acetonitrile, reached 100% in 20 min and was held for 5 min. Both eluents contained 50 ppm trifluoro acetic acid. A homologous series of alkylphenones Author's personal copy B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 was analyzed as external retention time references and used to calculate a bracketed retention index (RI) for each detected peak (Frisvad and Thrane, 1987). In this way each metabolite could be identified by its RI value and its UV–VIS spectrum and be recognized in other extracts. 2.5. HPLC-TOF-MS analyses Metabolite identification was done on the DRYES extracts using on a similar HPLC-DAD system as described above with minor changes: the column was 50 mm long, flow was 0.3 ml/min, and the eluents were buffered with 20 mM formic acid (Nielsen et al., 2005). High resolution MS detection was done on a Time Of Flight mass spectrometer (Water-Micromass, Manchester, UK) scanning m/z 60–900 and m/z 100–2000 in two separate scan functions at different skimmer settings (Nielsen et al., 2005). All samples were analyzed in both positive (ESI+) and negative electrospray (ESI–). Extracts were also analyzed using a gradient system optimized for A. infectoria metabolites, using a 2 mm  100 mm 3 lm Gemini Phenyl column (Phenomenex, Torrance, CA, USA) and a gradient starting at 10% acetonitrile, reaching 47% in 17 min and then 100% in 3 min and was held for 4 min. All samples were analyzed in ESI+ and ESI–. Metabolite standards (Nielsen and Smedsgaard, 2003) of 4Z-infectopyrone, AAL-toxins TA and TB, AK toxin I, altenuene, altenusin, alternariol, alternariol monomethylether, altersolanol A, altertoxin I, macrosporin, maculosin, tentoxin and tenuazonic acid were co-analyzed for verification. 2.6. Data treatment of metabolite profiles Metabolite profile data from HPLC-UV–VIS were first treated with the Chemical Image Analysis (CIA) program using an algorithm described by Hansen (2003) as stated by Andersen et al. (2008). In brief: the raw HPLC data files, which are quantitative 2-D matrices (x-axis: time, y-axis: wave length, value in matrix: UV–VIS absorbance), were transferred from the HPLC to a standard PC and analyzed by an in-house written chemical image analysis (CIA) program (Hansen, 2003). No manipulations or peak selections were made before processing. Each HPLC file was processed first by a log10 scaling (to account for concentration differences among extracts), then a baseline correction and finally an alignment (to account for drift in baseline and retention time among identical metabolites in different runs) (Hansen, 2003). Each HPLC file was then compared to the other 50 HPLC files, pair-wise, using an algorithm described by Hansen (2003) giving a similarity value for each pair, which was entered into a new matrix. The resulting 51  51 similarity matrix was then used to calculate a dendrogram using WARD clustering method. Based on the result of the CIA, a binary matrix was made manually by scoring each metabolite from four printed HPLC chromatograms as present or absent (137 metabolites for the 51 fungal strains) and subjected to multivariate statistics using Unscrambler version 9.2 (CAMO ASA, Oslo, Norway). The matrix was analyzed using Partial Least Squares Regression (PLS-R) and Principal Component Analysis (PCA). A reduced matrix (124 metabolites and 49 strains) was subjected to cluster analysis using NTSYS-pc version 2.11 N (Exeter Software, Setauket, NY, USA) without standardization using Yule as correlation coefficient and Unweighted PairGroup Method using arithmetic Averages (UPGMA) as clustering method. The matrix was also analyzed by simple matching and Jaccard similarity coefficients in NTSYS. For metabolite identification and peak comparison, 12–20 of the largest peaks in the data files from HPLC-TOF-MS were inspected in UV–VIS, ESI+, ESI– modes were compared to peaks registered in the Quanlynx 4.1 software (Water-Micromass) using a ±m/z 0.02 and a retention time limit of ±0.3 min. If possible, a qualifier ion was 645 used for confirmation. Peaks of the metabolite standards mentioned above were also inserted along with predicted ions calculated from the masses of known Alternaria metabolites. In the latter case a ±3 min window was used, with the retention time predicted using the log D of the compound (calculated using ACD v.10, Advanced Chemical development Inc., Toronto, Canada), which was correlated to retention time of 50 representative secondary metabolites. All extracts were then analyzed by the Quanlynx software. For metabolite identification, each peak was matched against an internal reference standard database (800 compounds) as well as tentatively identified by searching the accurate mass in the 34,392 compounds in Antibase 2008 (Laatsch, 2008), comparing UV–VIS data, fragmentations, ionization efficiency in ESI– versus ESI+ and the retention time to information in the databases. 2.7. DNA extraction, PCR amplification and sequencing For DNA analysis, each strain was inoculated in three points onto a PCA plate and incubated as mentioned above. DNA was extracted from 7-day-old PCA cultures with UltraClean Microbial DNA isolation kit (Mo Bio Laboratories, Solana Beach, CA, USA) according to manufacturer’s protocol and stored at 20 °C. Amplifications of the three target genes were performed with the following primer combinations: ITS: V9G (de Hoog and Gerrits van den Ende, 1998) and ITS4 (White et al., 1990): gpd: gpd1 and gpd2 (Berbee et al., 1999), tef-1a: EF1–645F: (TCG TCG TYA TCG GMC ACG TCG A) and EF1–1190R (TAC CAG TGA TCA TGT TCT TGA TGA). EF1–645F and EF1–1190R were designed by aligning sequences of Gibberella zeae (XM388987), N. crassa (D45837), Aspergillus fumigatus (XM745295) and Ustilago maydis (XM751978) from GenBank. The numbers of the primers refer to the nucleotide position in N. crassa (D45837) at the 30 -end as done previously (Carbone and Kohn, 1999). ITS and gpd PCR reactions were performed in 12.5 ll volumes containing 0.5 ll DNA, 1 NHþ 4 -buffer [160 mM (NH4)2SO4, 670 mM Tris–HCI (pH 8.8 at 25 °C), 0.1% Tween-20] (Bioline, London, UK), 1 mM MgCl2, 0.04 mM dNTPs, 0.2 pmol of each primer and 0.5 U BIOTAQTM DNA Polymerase (Bioline, London, UK). Tef-1a PCR reactions were performed with 0.8 pmol of each primer and 2.5 mM MgCl2 and with the same concentrations of the remaining ingredients as for ITS and gpd. All PCR amplifications were performed after the same scheme with an initial denaturation at 94 °C for 5 min followed by 40 amplification cycles of 94 °C for 30 s, annealing for 30 s and 72 °C for 1:20 min, and a final extension at 72 °C for 7 min. Annealing temperatures were: 48 °C (ITS), 52 °C (tef-1a) and 59 °C (gpd). Amplicons were run on 1% agarose gels and visualised with UV after ethidium bromide staining. Sequence reactions were performed with the BigDyeÒ Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA). All sequence-PCR reactions were performed with the same protocol: 94 °C for 1 min, followed by 30 cycles of 94 °C for 10 s, 50 °C for 5 s and 60 °C for 4 min. DNA was purified with SephadexÒ G-50 (Pharmacia-Amersham) and sequenced. All sequences determined in this study have been submitted to GenBank and accession numbers are listed in Table 2. 2.8. Data treatment of molecular sequences Sequence electrophorograms of forward and backward runs were combined, analyzed, edited using DnaStar SeqMan II (LaserGene). Sequence data were saved and aligned with BioNumerics (Applied Maths, Kortrijk, Belgium). The alignments of tef-1a, ITS and gpd were concatenated into one alignment to construct a phylogenetic tree. The program RaxML (http://www.phylo.org/portal/ Home.do) was used to create the best tree using maximum likelihood and to calculate bootstrap values (Stamatakis et al., 2008). The same program and conditions were used to create individual trees of all three markers used in this study. Genetic diversity of Author's personal copy 646 B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 Table 2 Ascoma production on PCA and growth at 37 °C on PDA together with gpd haplotypes and GenBank accession numbers. #a Genus (species/group) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 43 44 45 46 47 48 49 50 51 52 Alternaria infectoria sp.–grp. A. triticina T (inf–grp) A. infectoria T (inf–grp) A. oregonensis T (inf–grp) A. photistica T (inf–grp) A. ethzedia T (inf–grp) A. metachromatica T (inf–grp) A. triticimaculans T (inf–grp) A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. intercepta T (inf–grp) A. viburni T (inf–grp) A. arbusti T (inf–grp) A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. Chalastospora cetera T A. malorum A. malorum A. malorum A. malorum A. malorum A. malorum var. polymorpha A. malorum A. malorum A. malorum Embellisia abundans E. abundans T A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. A. infectoria sp.–grp. a b Ascoma production 1. Trial 2. Trial +  + +  +   + + + + +  + + + + + ntb nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt +    + +         + + +  +    + +   +       + +  + +    nt    + nt nt nt nt nt Growth at 37 °C gpd ITS gpd tef-1a FJ214885 AY762948 FJ214897 AY762947 FJ214900 AY278833 AY762946 AY762949 FJ214890 FJ214886 FJ214859 FJ214882 FJ214876 FJ214857 FJ214887 FJ214863 FJ214865 FJ214866 FJ214868 FJ214864 FJ214888 FJ214870 FJ214894 FJ214861 FJ214860 FJ214883 FJ214884 FJ214895 FJ214896 FJ214898 AB120848 FJ214872 FJ214873 FJ214862 FJ214880 FJ214867 FJ214871 FJ214875 FJ214881 FJ214877 FJ214879 FJ214892 FJ214878 FJ214856 FJ214899 FJ214858 FJ214889 FJ214869 FJ214874 FJ214891 FJ214893 FJ214836 FJ214846 FJ214850 FJ214849 FJ214854 FJ214855 FJ214835 FJ214834 FJ214841 FJ214837 FJ214808 FJ214831 FJ214825 FJ214806 FJ214838 FJ214812 FJ214814 FJ214815 FJ214817 FJ214813 FJ214839 FJ214819 FJ214845 FJ214810 FJ214809 FJ214832 FJ214833 FJ214847 FJ214848 FJ214851 FJ214852 FJ214821 FJ214822 FJ214811 FJ214829 FJ214816 FJ214820 FJ214824 FJ214830 FJ214826 FJ214828 FJ214843 FJ214827 FJ214805 FJ214853 FJ214807 FJ214840 FJ214818 FJ214823 FJ214842 FJ214844 FJ214932 FJ214942 FJ214946 FJ214945 FJ214950 FJ214951 FJ214931 FJ214930 FJ214937 FJ214933 FJ214904 FJ214927 FJ214921 FJ214902 FJ214934 FJ214908 FJ214910 FJ214911 FJ214913 FJ214909 FJ214935 FJ214915 FJ214941 FJ214906 FJ214905 FJ214928 FJ214929 FJ214943 FJ214944 FJ214947 FJ214948 FJ214917 FJ214918 FJ214907 FJ214925 FJ214912 FJ214916 FJ214920 FJ214926 FJ214922 FJ214924 FJ214939 FJ214923 FJ214901 FJ214949 FJ214903 FJ214936 FJ214914 FJ214919 FJ214938 FJ214940 Haplotype + +  +  +   +   + + + +   +   + + + + + + + + +    + + + + +  + + + nt +  +  nt nt nt nt nt 1 3 2 4 nt 5 6 7 21 22 22 8 9 10 11 12 2 13 9 nt nt nt nt nt nt nt nt nt nt nt nt 6 23 24 15 15 15 9 15 19 9 2 20 14 15 6 16 17 15 18 15 Analysis number in this study. #42 is not included. Not tested. aligned sequences and the standardized Index of Association ðIsA Þ was calculated with LIAN 3.5 (Haubold and Hudson, 2000) by combining clustering information of the individual trees. The phylogenetic network was created with SplitsTree v4.8. The network structure was based on the neighbornet algorithm with a threshold set to 104 and applying the LogDet transformation. LogDet is a distance transformation that corrects for biases in the base composition (Wägele and Mayer, 2007). The phi-test incorporated in the SplitsTree software (Huson and Bryant, 2006) was used to test signals of recombination (p < 0.05, significant evidence of recombination). The test is proven to be a robust calculation and no previous knowledge about population history, recombination rate, mutation rate and rate heterogeneity across sites (Bruen et al., 2006) is necessary. Although large splits in networks do not necessarily imply recombination, split decomposition networks in conjunction with the phi-test can easily detect which sequences in a given data set contribute the most to the recombination signal (Salemi et al., 2008). The phi-test is repeated after possible recombinants are deleted from the alignment until p > 0.05 (no evidence of recombination). DnaSP v4.5 (Rozas and Rozas, 1995) was used to find the different haplotypes in the gpd alignment. Gaps and missing data were not considered during calculation. 3. Results 3.1. Morphology Examination of the sporulation patterns on PCA after seven days of growth in alternating light showed three different and very distinct morphologies as shown in Fig. 1. A. malorum, A. malorum var. Author's personal copy B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 647 Fig. 1. Morphology plates. (A): Chalastospora cetera (#20), (B): Alternaria malorum (#25), (C): Embellisia abundans (#31) and six strains belonging to the A. infectoria species– group: (D): A. oregonensis (#04), (E): A. infectoria (#03), (F): A. arbusti (#14) and (G–I): three A. infectoria species–group strains (#32, #44 and #17, respectively). Bar in G = 50 lm. polymorpha and C. cetera all had cylindrical didymo- and phragmoconidia produced in very long branching chains (Fig. 1A and B). Both strains of E. abundans had solitary, ovoid phragmoconidia (Fig. 1C), while all strains of the A. infectoria species–group shared the same morphology: ovoid, obpyriform or obclavate phragmoand dictyoconidia with secondary conidiophores of varying length produced in branched chains (Fig. 1D–I). Type cultures fitted the descriptions, except A. arbusti (#14), which, under the growth conditions in this study, showed a more branched three-dimensional structure than the original description depicted (Simmons, 2007). Conidial sizes, shapes, ornamentation, color etc. varied greatly between strains, whereas conidial appearance was quite consistent within a strain: e.g. conidia of A. infectoria (#03) were smooth, narrow-obpyriform and sallow colored (Fig. 1E), while conidia of A. infectoria species–group (#44) were narrow-obpyriform to broadovoid, finely rough and bronze colored (Fig. 1H). None of the strains morphologically identified as belonging to the A. infectoria species–group could be assigned to any of the type cultures and no two strains, except A. infectoria species–group strains (#10 and #11), could be classified as belonging to the same taxon. Table 2 shows the production of proascomata (ascomata without mature ascospores) after 6 months of incubation on PCA at 7 °C. In the first trial, where three strains were inoculated on the same PCA plate, no mating between strains was observed. Proascomata (Fig. 2) were formed in the center of each colony and in areas in the agar furthest away from the two other colonies and there was a clear demarcation line between colonies. In the second trial, where the same strain had been inoculated twice on the same plate, proascomata were again formed at the center and furthest away from the other colony. Fourteen out of nineteen A. infectoria species–group strains produced proascomata in the first trial, and only seven of these in the second trial. Additional 14 strains were tested in the second trial, out of which five produced proascomata. In the second trial, neither C. cetera nor E. abundans produced proascomata, but three out of the nine A. malorum did. In contrast to the strains of the A. infectoria species–group, the three A. malorum strains produced their proascomata on the toothpicks and not in the agar. When the experiment was terminated after 6 months, none of the proascomata had yielded any matured ascospores. 3.2. Growth on different media The result of the experiment on PDA at 37 °C given in Table 2 showed that 28 out of the 45 tested strains were able to grow at this temperature. Unfortunately, it was not possible to test all 51 strains in the set, since six strains were not viable after the first chemical experiments. Table 2 shows that all nine A. malorum strains were able to grow at 37 °C, but not C. cetera. During the incubation period, the colonies of A. malorum became dark brown and the mycelium thinner and more thread-like. Of the 33 tested strains of A. infectoria species–group, 20 were able to grow at 37 °C. The colonies lost their aerial mycelium, became waxy in their growth, and only produced filamentous mycelium after they were taken out of the 37 °C incubator. The only viable strain of A. infectoria species–group from a human infection (#45) was not able to grow on PDA at 37 °C. Examination of the colony appearance on DRYES incubated at 25 °C showed that most of the 51 strains produced hairy to granu- Author's personal copy 648 B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 Fig. 2. Proascomata in 3-week-old PCA cultures. (A): Alternaria infectoria species–group #19 (bar = 500 lm) and (B): A. infectoria species–group #16 (bar = 200 lm). lar mycelium, whitish to grayish in color. Some of the A. infectoria species–group strains produced only wet, yeast-like colonies on DRYES, especially the strains isolated from human lesions, which, however, did not affect their metabolite production. Filamentous mycelium was produced when the strains were transferred to PDA or PCA. Sporulation of the A. infectoria species–group was only seen on PCA and sometimes scarification and prolonged incubation were needed, while A. malorum, C. cetera and E. abundans sporulated well on PCA as well as on PDA + DN and PDA. DRYES and DG18 did not accommodate any sporulation. As a curiosity, it should be mentioned that A. metachromatica (#07) on PDA + DN produced an extracellular pigment that turned the reverse of the plate dark blue. 3.3. Chemical classification The chemical HPLC-UV–VIS analyses of the three growth media, DRYES, DG18 and PDA + DN, showed that the metabolite production, qualitative as well as quantitative, was greatest on DRYES and that metabolite production seemed to be inhibited on PDA + DN. Fig. 3 shows six selected HPLC chromatograms made from the DRYES extracts. Automated and unbiased chemical image analyses (CIA) of the HPLC-UV–VIS files were made with extracts from all three growth media to aid the selection of species-specific metabolites (dendrograms not shown). All three dendrograms gave the same grouping, but the one made with extracts from DRYES gave the most detailed dendrogram. It showed that the 51 strains grouped in four major clusters. One cluster contained all eight A. malorum strains, the C. cetera (#20) strain, and the type culture of E. abundans (#31). A second cluster contained 13 A. infectoria species–group strains, including the type cultures of A. infectoria (#03), A. ethzedia (#06), and A. metachromatica (#07), together with E. abundans (#30) and A. malorum var. polymorpha (#26). A third cluster contained 10 A. infectoria species–group strains, including the type cultures of A. intercepta (#12) and A. viburni (#13), while the last cluster contained 12 A. infectoria species–group strains, including type cultures of A. photistica (#05), A. arbusti (#14), A. oregonensis (#04), A. triticina (#02) and A. triticimaculans (#08). Visual examination of the HPLC chromatograms of E. abundans (#30) and A. malorum var. polymorpha (#26) showed very few peaks, meaning a very low metabolite production. The location of these two stains in the dendrogram was therefore questionable and they where subsequently removed from further analyses. A binary matrix was made by visual examination of each peak and its associated UV–VIS spectrum in the HPLC chromatograms and the CIA dendrograms. On no occasion were there metabolites produced on DG18 or PDA + DN that were not found in DRYES, however, some metabolites were easier to detect on DG18 due to a qualitatively simpler metabolite profile. The matrix consisted of 137 recognizable metabolites for the 51 strains and was subjected to a Partial Least Squares Regression (PLS-R) (result not shown), which gave metabolites specific for the A. malorum strains, the C. cetera strain, the A. infectoria species–group strains and the E. abundans strains, respectively (see Table 3). All 51 chromatograms were analyzed for the production of known Alternaria metabolites and tested negative for AAL-toxins, alternariols, altersolanols, altenuenes, tentoxin and tenuazonic acid. On the other hand, HPLC-UV– VIS as well as HPLC-MS analyses showed that all 51 strains, including E. abundans (#30) and A. malorum var. polymorpha (#26), were able to produce infectopyrone and 4Z-infectopyrone. Altertoxin derived metabolites were restricted to the A. infectoria species–group, but not produced consistently throughout the species–group, while macrosporin was found in three A. malorum strains. Novaezelandin production was shared by E. abundans and the A. infectoria species–group, but again not produced consistently throughout the species–group. A. malorum and C. cetera had a number of metabolites of unknown structure in common (e.g. RI value 694), but also produced unknown metabolites specific to each species (e.g. RI values 715 and 894, respectively). E. abundans did not produce any known metabolites, except for the infectopyrone, 4Z-infectopyrone, and novae-zelandin A, and only few species specific metabolites (e.g. RI values 691 and 779a) were detected. Based on the combined results of the CIA and the PLS-R, a matrix that consisted of 49 strains [excluding E. abundans (#30) and A. malorum var. polymorpha (#26)] and 124 known and unknown metabolites (excluding infectopyrones and other consistently produced metabolites) was constructed. The matrix was subjected to cluster analysis and the resulting dendrogram is shown in Fig. 4. The dendrogram shows a division of the 49 strains into two clusters, A (in light grey) and B. Cluster A contains the type culture of E. abundans (#31), the type culture of C. cetera (#20) together with six A. malorum strains in sub-cluster A1 and two A. malorum strains (#22 and #28) in sub-cluster A2. Cluster B, holding all the 39 strains of the A. infectoria species–group, could be divided into two sub-clusters with type cultures of A. photistica (#05) and A. triticina (#02) as outliers. Sub-cluster B1 contained six A. infectoria species–group strains and the type cultures of A. triticimaculans (#08), A. intercepta (#12) and A. viburni (#12) (marked with ), while the majority of the A. infectoria species–group strains clustered closely together in sub-cluster B2. Both B sub-clusters could be further divided based on production of different metabolites. The color-coding of strains in Fig. 4 is referring to different haplotypes. The result of a principal component analysis of the 39 A. infectoria species–group strains and 79 metabolites (given as their RI Author's personal copy 649 B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 A Embellisia abundans (#31) 100 75 50 25 0 0 B 5 10 15 20 25 Chalastospora cetera (#20) 100 75 50 25 0 0 C 5 10 15 20 25 Alternaria infectoria sp-grp (#33) 100 75 50 25 0 0 D 5 10 15 20 25 Alternaria oregonensis (#04) 100 75 50 25 0 0 E 5 10 15 20 25 Alternaria infectoria sp-grp (#01) 100 75 50 25 0 0 F 5 10 15 20 25 Alternaria triticina (#02) 100 75 50 25 0 0 5 10 15 20 25 Fig. 3. HPLC chromatograms (wave length 210 nm). (A): Embellisia abundans (#31), (B): Chalastospora cetera (#20) and four strains belonging to the Alternaria infectoria species group; (C): A. infectoria species–group (#33), (D): A. oregonensis (#04), (E): A. infectoria species–group (#01) and (F): A. triticina (#02). values) is shown in Fig. 5. It shows the association between metabolites and the strains that produce them. Metabolites common to most strains are located in the center, while metabolites specific to a few strains are located around the perimeter together with Author's personal copy 650 B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 Table 3 Production of known metabolites and selected, unknown metabolites, given by their retention index (RI) value, from Alternaria infectoria species–group, A. malorum, Chalastospora cetera and Embellisia abundans. Metabolite (RI)a b Altertoxin derived (855) Altertoxin derived (820) Altertoxin derived (846) Infectopyrone (839)c 4Z-Infectopyrone (824a) Infectopyrone derived (706) Infectopyrone derived (713b) Macrosporin (1068) Novae-zelandin B (892) Novae-zelandin A (680a) Novae-zelandin derived (726) 569 950d 691 779a 715 894 694 1010 642 1120 1076e 1047 752 a b c d e f A. infectoria sp–grp (39) A. malorum (9) C. cetera (1) E. abundans (2) 31 9 4 39 39 30 23 f 29 19 6 19 13         39 35 26    9 9 2 1 3        5  7 4 7 6 7 5 5    1 1            1 1 1 1 1 1 1     2 2     2    1 1     1 1 2 1 1 RI: retention index value, calculated by the HPLC from retention time. Same as metabolite 3 in Andersen and Thrane (1996). Same as metabolite 2 in Andersen and Thrane (1996). Same as metabolite 5 in Andersen and Thrane (1996). Same as metabolite 6 in Andersen and Thrane (1996). Not detected. the strains producing them. From Fig. 5 it can be seen that strains in sub-cluster B1 in Fig. 4 (marked with ) produced a large number of metabolites including the altertoxins and novae-zelandins, whereas strains in sub-cluster B2 produce fewer metabolites and not altertoxins or novae-zelandins. On the other hand, strains in B2 produce metabolites of unknown structure (e.g. RI values 569, 706, 713b, 752, 813), which are not produced by strains in B1. In general, many individual metabolites of unknown structure were found to be specific to only one or a few strains in the A. infectoria species–group, which hampered a clear grouping. 3.4. Molecular cladification The obtained sequences of gpd were 444–446 bp for A. photistica and the A. infectoria species–group and of 424 bp for A. malorum, E. abundans, and C. cetera. The aligned gpd sequences contained one intron of approximately 114 bp. The alignment dataset of all the strains contained 456 bp with 131 variable sites of which 85 were parsimony informative. Sequences of tef-1a were 437–440 bp containing two introns of approximately 250 bp in total. The tef-1a alignment dataset consisted of 443 bp containing 113 variable sites of which 65 were parsimony informative. The obtained ITS sequences were 490 bp for A. photistica, 519–525 bp for the A. infectoria species–group strains, 533–534 bp for the A. malorum strains and 523 bp and 525 for the E. abundance and C. cetera strains, respectively. The ITS sequence for A. photistica was smaller than those of the remaining strains due to a major deletion in ITS1. The ITS alignment dataset of all the strains contained 544 bp with 106 variable sites of which 60 were parsimony informative. The ITS and tef-1a dendrograms gave the same major division as the gpd dendrogram, but with lower resolution (data not shown). Strains that were identical in one gene sequence were nearly always different in another. Molecularly, all the A. infectoria species–group strains were mutually similar, but never identical, except for A. infectoria species–group strains (#10 and #11), which have identical sequences in all three tested genes. Fig. 6 shows an unrooted dendrogram for all 51 strains of the concatenated ITS, tef-1a and gpd sequences using maximum likelihood in RaxML. It shows two major clades, one with 38 A. infectoria species–group strains and with A. triticina (#02), three strains of A. infectoria species–group (#09, #48 and #51) and A. photistica as outliers and another clade with all A. malorum, C. cetera and E. abundans strains (in light grey). Within the latter clade, A. malorum var. polymorpha (#26) could not be distinguished from the remaining eight A. malorum strains. Nearest neighbors of A. malorum were C. cetera and E. abundans. The degree of variability within the A. infectoria species–group proved to be limited in all genes. Fig. 7 shows the nucleotide differences between 38 strains in the A. infectoria species–group [excluding A. photistica (#05)]. As seen in Fig. 7, most nucleotide differences in the three genes were observed in the spacers and introns, although there were some mutations in the coding region of gpd, which all occurred on the third codon position often with a silent C to T substitution. Table 4 shows alignment data for the three genes. Internal ITS alignment of the 38 strains in the A. infectoria species–group resulted in 40 variable sites of which 35 were parsimony informative and located in either the ITS1 or ITS2. The tef-1a alignment resulted in 113 variable sites of which 53 were parsimony informative. The gpd alignment showed 109 variable and 59 parsimony informative sites of which 18 were located in the 114 bp intron. Using DnaSP on the gpd alignment data of the 38 strains in the A. infectoria species–group, haplotypic groups were defined and are given in Table 2. Most haplotypic groups contained only one strain except for haplotype 2 (#3, #17 and #43), haplotype 6 (#7, #32 and #47), haplotype 9 (#13, #19, #38 and #41), haplotype 15 (#35–37, #39, #46, #50 and #52) and haplotype 22 (#10–11), resulting in 24 distinct haplotype groups. Fig. 8 shows the haplotype network. The standardized Index of Association ðIsA Þ of the Author's personal copy B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 A1 A A2 B1 B2 B -0.20 0.10 0.40 0.70 651 E. abundans Fragaria 31 (T) C. cetera Elymus 20 (T) A. malorum Triticum 27 A. malorum Soil 25 A. malorum Gossypium 29 A. malorum Soil 21 A. malorum - 24 A. malorum - 23 A. malorum Triticum 28 A. malorum - 22 A. intercepta Viburnum 12 (T) A. viburni Viburnum 13 (T) A. infectoria Zea 52 A. infectoria Zea 36 A. infectoria Zea 34 A. infectoria Zea 33 A. infectoria Zea 37 A. infectoria Hordeum 46 A. triticimaculans Triticum 08 (T) A. infectoria Zea 44 A. infectoria Hordeum 18 A. infectoria Avena 48 A. infectoria Triticum 03 (T) A. infectoria Homo sapiens 51 A. infectoria Homo sapiens 49 A. infectoria Paeonia 47 A. arbusti Pyrrus 14 (T) A. infectoria Hordeum 09 A. oregonensis Triticum 04 (T) A. metachromatica Triticum 07 (T) A. infectoria Zea 43 A. infectoria Straw 32 A. infectoria Zea 40 A. infectoria Hordeum 19 A. infectoria Zea 38 A. infectoria Homo sapiens 45 A. ethzedia Brassica 06 (T) A. infectoria Zea 41 A. infectoria Zea 35 A. infectoria Hordeum 17 A. infectoria Triticum 15 A. infectoria Hordeum 11 A. infectoria Hordeum 10 A. infectoria Triticum 16 A. infectoria Homo sapiens 50 A. infectoria Zea 39 A. infectoria Hordeum 01 A. photistica Digitalis 05 (T) A. triticina Triticum 02 (T) 1.00 Yule/UPGMA Fig. 4. Dendrogram based on a cluster analysis of 49 metabolite profiles (1 Embellisia abundans, 1 Chalastospora cetera, 8 Alternaria malorum and 39 strains belonging to the A. infectoria species–group). Color-coding in the B cluster corresponds to haplotype groups given in Table 2 and Fig. 8. Strain labels: strain ID/host/strain number/type culture. Dendrogram calculated using the Yule correlation coefficient and UPGMA as the clustering method. Axis shows the correlation coefficient from 1 to 1. same A. infectoria species–group strains showed a tendency towards recombination events, ðIsA Þ = 0.1627. LIAN v3.5 was used to calculate the standardized Index of Association with 1,000,000 Monte Carlo samplings. The neighbornet splitstree of gpd alignment data (not shown) of the A. infectoria species–group showed mostly a treelike structure. The network also showed conflicting phylogenetic trees (histories) that can not be shown with a bifurcating tree. Conflicting phylogenetic signals can occur by recombination or by convergent substitutions and can not be distinguished by looking at the network alone (Salemi et al., 2008). However, a phi-test was able to detect the presence of recombination in aligned sequences. Repeated phi-test calculations after removing single sequences from the alignment showed the presence of recombinants. When the p-value increased till 0.05 or more, it was obvious that the recombinants were deleted from the alignment. Table 4 shows the p-values for the three genes in the phi-test of the 38 A. infectoria species–group strains [excluding A. photistica]. ITS and tef-1a had p-values > 0.05, which indicated Author's personal copy 652 1.0 B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 PC2 833a 752 713b 706 A. infectoria 39 1047 869 910 A. infectoria 19 813 0.5 A. metachromatica 07 (T) 792b A. infectoria 18 A. infectoria 17 A. infectoria 49 47 infectoria A.A. A.infectoria infectoria4803 (T)569 z-Inf-824a1118 681 A. infectoria 51 937a 685 A. infectoria 01 950729 -0.5 A. infectoria 34 Alx-820 A. infectoria 38 758 878 777 723 781b 977 1076 750c 936 757b 1279 871 A. infectoria 44 A. ethzediaA. 765c 749a 713c 06infectoria (T) 773605 1315 A.32 infectoria 15 602b 633 A. infectoria 16 A. infectoria 10 09 A. infectoria 753938b 712 Inp 766b A. infectoria 45 43 A. infectoria 937b 793 938a 11 822 808 765b 847 A. infectoria 563 Alx-846 Alx-855750b636 721 757a 788 952 1024 1011 588 765a NZ-B-892 979602a 1440713a 939 824b 738 749b A. infectoria 40 816b 816a NZ-A-680a A. arbusti 14 (T)A. infectoria 750a 41 1549 A. oregonensis 04 (T) 726 2047 0 A. viburni 13 (T) A. intercepta 12 (T) A. infectoria 36 A. infectoria 52 A. infectoria 33 A. infectoria 50 A. infectoria 35 A. photistica 05 (T) A. infectoria 37 46 A. infectoria Sub-cluster B1 A. triticimaculans 08 (T) A. triticina 02 (T) -1.0 -0.6 -0.4 -0.2 0 0.2 PC1 0.4 0.6 0.8 1.0 Fig. 5. Loadings plot based on a principal component analysis of 79 individual metabolites (39 strains belonging to the A. infectoria species–group). Strains are in black and metabolites in blue. Metabolites of known structure: Alx: altertoxin derivatives; NZ: novae-zelandin derivatives; z-Inf: infectopyrone derivate. Metabolites of unknown structure are only given by their Retention Index (RI) values calculated from their retention time on HPLC. Strain labels: strain ID/strain number/type culture. Strains in the grey box marked with  correspond to sub-cluster B1 in Fig. 4. Axes are score values. the absence of recombination. However, the p-value of gpd was 7.5  104 suggesting the presence of recombination events in this gene. After deleting strains #08, #45 and #51 (CBS 578.94, CBS 102692 and CBS 110804, respectively), the phi-test showed no significant evidence of recombination in the gpd data (p > 0.05) and therefore these three strains were considered to be recombinants. 4. Discussion At genus level, the Alternaria infectoria species–group could clearly be separated from the A. malorum/C. cetera/E. abundans group based on morphology as well as chemical classification and molecular cladification. The results show that strains morphologically identifiable as A. infectoria species–group produced altertoxins and novae-zelandins and yielded ITS, gdp and tef-1a sequences that were different from those of the A. malorum/C. cetera/E. abundans group (Figs. 1, 4 and 6 and Table 3). E. abundans, on the other hand, could only be segregated from the A. malorum/C. cetera group by morphology (Fig. 1), but not with any confidence by molecular or chemical means. Besides, the chemical similarity turned out to be too large and the number of species/strains used proved to be too few to speculate on the placement of Embellisia, compared to A. malorum/C. cetera group. Other studies based on ITS, SSU and gpd data, show species of Embellisia are scattered among genus Alternaria as well as genus Ulocladium despite its distinct morphology (Pryor and Bigelow, 2003). In contrast, C. cetera could not be segregated from the A. malorum/A. malorum var. polymorpha group by morphological, molecular, or chemical means (Figs. 1, 4 and 6 and Table 3). Two A. malorum strains (#23 and #24) were distinct and identical in all three methods. The other A. malorum strains yielded metabolite profiles that were similar, but not identical to each other and to that of C. cetera. Morphologically, A. malorum and C. cetera showed the same general sporulation pattern, but with some variation in conidial size and septation. The polyphasic data in this study shows that A. malorum var. polymorpha and the eight A. malorum strains, do not belong in the A. infectoria species–group as proposed by Braun et al. (2003), but suggest they belong to the same genus as Chalastospora cetera, however, as several distinct species. The production of infectopyrones and a pair of compounds with unknown structure (RI values 1047 and 1076) by all genera in this study corroborates the close relationship found in the molecular analyses, but on the other hand, some species of Ulocladium, which are phylogenetically more related to small-spored Alternaria (Pryor and Bigelow, 2003), also produce infectopyrones (Andersen and Hollensted, 2008) and these metabolites may be more widespread in Pleosporaceae. Furthermore, this is the first report on the production of infectopyrones by A. malorum, C. cetera, and E. abundans and the production of macrosporin and novae-zelandin A by A. malorum and E. abundans, respectively. At species level, the 10 Alternaria type cultures representing morphological species within the A. infectoria species–group were Author's personal copy B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 653 A. infectoria Triticum 03 (T) A. infectoria Hordeum 17 A. infectoria Zea 43 A. infectoria Homo sapiens 45 A. intercepta Viburnum 12 (T) A. infectoria Zea 41 81 A. viburni Viburnum 13 (T) A. infectoria Hordeum 19 A. infectoria Zea 38 98 A. infectoria Hordeum 01 A. infectoria Hordeum 11 71 A. infectoria Hordeum 10 A. infectoria Zea 44 A. infectoria Triticum 16 A. oregonensis Triticum 04 (T) A. arbusti Pyrus 14 (T) A. infectoria Zea 33 A. infectoria Zea 34 A. infectoria Paeonia 47 A. metachromatica Triticum 07 (T) A. infectoria Straw 32 A. triticimaculans Triticum 08 (T) 79 A. infectoria Zea 37 A. infectoria Hordeum 46 A. infectoria Zea 36 A. infectoria Homo sapiens 50 81 A. infectoria Triticum 15 A. infectoria Zea 52 A. infectoria Zea 39 A. infectoria Zea 35 A. infectoria Hordeum 18 A. ethzedia Brassica 06 (T) A. infectoria Zea 40 A. infectoria Homo sapiens 49 A. infectoria Avena 48 A. triticina Triticum 02 (T) 99 A. infectoria Hordeum 09 99 A. infectoria Homo sapiens 51 A. photistica Digitalis 05 (T) A. malorum Soil 21 100 70 A. malorum - 23 A. malorum - 24 99 A. malorum var. polymorpha Vitis 26 (T) A. malorum Triticum 28 87 A. malorum Gossypium 29 81 A. malorum Soil 25 A. malorum - 22 93 A. malorum Triticum 27 C. cetera Elymus 20 (T) 100 E. abundans Dianthus 30 E. abundans Fragaria 31 (T) 0.01 79 72 Fig. 6. Unrooted consensus dendrogram based on 51 strains (2 Embellisia abundans, 1 Chalastospora cetera, 39 Alternaria infectoria species–group, 8 A. malorum and 1 A. malorum var. polymorpha). Maximum likelihood tree of 3 partial genes (ITS, gpd and tef-1a) constructed using RaxML (Cipres webserver). Bootstrap values > 70% are indicated. Color-coding in the A. infectoria species-group clade corresponds to haplotype groups given in Table 2 and Fig. 8. Strains marked with  correspond to cluster B1 in the chemical analysis. Strain labels: strain ID/strain number/type culture. located in different sub-clades depending on the molecular sequence examined, but with A. photistica (#05) and A. triticina (#02) as outliers. With each individual gene, variability was largely random, judging from low bootstrap values and from obtaining dif- ferent groupings when different algorithms were used for tree reconstruction. When genes were concatenated, A. viburni (#13) clustered at 81% bootstrap support with three strains (#19, #38 and #41) identified as A. infectoria species–group sensu Simmons. Author's personal copy 654 B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 Fig. 7. Mutations at each position in the aligned ITS, tef-1a and gpd sequences of the 38 A. infectoria strains (except A. photistica #05) with the predicted exons and introns. The single type cultures were not unambiguously separated from the core of A. infectoria. The same pattern was seen in the chemical data with A. photistica (#05) and A. triticina (#02) as outliers. No distinct groups or clades were formed in the molecular analyses, while there was a certain grouping in metabolite profiles and metabolite families (Figs. 3–5 and Table 3). Some A. infectoria species–group strains were able to produce altertoxin derivatives, while others produced metabolites of unknown structures, but not altertoxins. No distinct morphological groups were seen either among the strains in the A. infectoria species–group. Morphology showed that basically each strain was a taxon in its own right. Lastly, no groupings or correlations could be found between proascoma formation, ability to grow at 37 °C, host or geographic origin and haplotypes, metabolite production or morphological identity. Our original hypothesis was that taxa in Lewia/A. infectoria species–group were sexual fungi and that molecular sequence analyses and metabolite profiling would yield a number of groups according to the genealogical concordance phylogenetic species recognition (GCPSR) (Taylor et al., 2000). Our data, however, indicate that only three strains in the A. infectoria species–group show evidence of recombination and that several isolates are able to produce proascomata in axenic culture. Since several taxa in the A. infectoria species–group have been shown to produce ascomata with viable ascospores in axenic cultures (Kwasna and Kosiak, 2003; Simmons, 2007; unpublished results), Lewia/A. infectoria species–group must, at least in part, be homothallic and the purpose of ascoma formation in nature could be a survival strategy. The high similarity in nucleotide sequence amongst the A. infectoria species–group strains (Fig. 6), suggests that most strains are clonal and may have derived via mutations from one common ancestor similar to the arbuscular mycorrhizal fungi (Rosendahl, 2008). Several studies (reviewed in Taylor et al., 2000; O’Donnell et al., 2004) show an increase in numbers of taxa, when going from morphological species recognition via biological recognition to GCPSR, which corresponded with either geographic origin or hosts. In our study we see the opposite: molecular cladification yields the lowest number of taxa in the A. infectoria species–group (A. photistica and one phylogenetic taxon), while the chemical classification gives more (A. photistica and A. triticina and two chemically different taxa) and with morphological appearance giving the highest number (38 morphologically different taxa). Applying GCPSR to the A. infectoria species–group would lead to synonymizing of all morphological species in the A. infectoria species–group under one name: A. infectoria Simmons. Alternatively, morphological species recognition could be applied and strains in the A. infectoria species–group would represent new ‘‘emerging” species that require a name and a formal description. But according to Taylor et al. (2000) and Rosendahl (2008), GCPSR can only be applied to sexual/heterothallic fungi, not to homothallic/clonal strains, so neither of the two approaches (one species or 38 species) is workable. In practice, however, there is a regular need for identification of Alternaria isolates, because they have acquired different abilities in nature, which affect us negatively. Some isolates have been encountered as opportunistic human pathogens, others as plant pathogens and others again are saprotrophic on cereals producing biologically active metabolites. Artificial identification systems based on any stable differentiation characters (e.g. PCR, AFLP, metabolite profiles, sporulation patterns obtained under standardized conditions) still play an important role in taxonomy. Strains in the A. infectoria species–group show characteristic phenotypical traits, which can be detected, recognized, and used for identification. Depending on the users needs, identification of taxa in the A. infectoria species–group can be done to different levels. In medical mycology, molecular identification using ITS is fast, wellknown and often the only method to obtain the correct diagnosis for isolates in the A. infectoria species–group, since strains from human lesions rapidly loose their ability to sporulate in vitro. Strains used in this study that originated from human skin lesions sporulated poorly, even under optimal conditions, and went sterile after one or two transfers. However, they still maintained their ability to produce all the characteristic metabolites in spite of their vegetative or yeast-like growth. Concerning alternarioses in humans or animals, generally only identification to species–group level is needed, since the same medical treatment (e.g. itraconazole) is likely to be applicable regardless of taxon identification within the A. infectoria species–group (Brasch et al., 2008; Dye et al., 2009). In plant pathology, phytosanitary, and quarantine, on the other hand, ITS sequencing is not enough to identify a known pathogen or discover a new pest that requires quarantine. With our current knowledge, described plant pathogens like A. triticina, A. viburni, and A. intercepta can be distinguished from other taxa of the A. infectoria species–group using morphology. In the cladistic analyses, A. triticina (#02) grouped with different taxa in the A. Table 4 Alignment data set for 38 strains in the Alternaria infectoria species–group, except A. photistica, of the three genes with number of mutations, parsimony informative mutations, sites and p-value in phi-test. gpd tef-1a ITS a b c Total number of sites (gaps/missing) Total number of mutations Number of parsimony informative sites Parsimony informative mutations (%)a Parsimony informative sites (%)b phi-Test p-valuec 456 (40) 443 (17) 546 (71) 109 113 40 59 53 35 54.1 46.9 87.5 12.9 12.0 6.4 7.5104 0.15 0.41 Percentage is calculated using the total number of mutations. Percentage is calculated using the total number of sites. p-Value < 0.05 shows presence of recombination. Author's personal copy 655 B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 19 20 3 8 12 3 10 6 23 17 14 8 3 22 1 18 17 13 23 3 7 16 8 12 13 16 21 5 2 9 7 14 15 7 6 8 1 24 13 6 3 8 3 15 2 11 7 2 5 7 9 5 4 Fig. 8. Haplotype network of A. infectoria species–group strains (except A. photistica #05) based on maximum parsimony tree of gpd sequences. The circles represent the 24 different haplotype groups, which are given in Table 2. The lines between the groups connecting the haplotypes show the number of nucleotides differing. Circles with solid colors (haplotypes 2, 6, 9, 15 and 22) contain more than one strain and circles with red lines (haplotypes 7, 14 and 18) show the position of the recombinant strains. infectoria species–group depending on the chosen DNA sequence, but was an outlier chemically, having a different metabolite profile. Further research may yield A. triticina specific metabolites that can be used to facilitate identification. In food safety, taxa in the A. infectoria species–group regularly contaminate cereal grain (Andersen et al., 1996; Pitt and Hocking, 1997; Kosiak et al., 2004; Perelló et al., 2008). The most urgent need is to know what secondary metabolites and other biologically active compounds are produced in the cereals like wheat, barley, and maize. Since current knowledge does not allow connections between metabolite profiles and morpho-species to be made, chemical analyses are needed. The results presented in this study show that these household genes (ITS, tef-1a and gpd) do not reflect ecology, secondary metabolism or morphology of the A. infectoria species–group and that molecular cladification and phylogeny cannot predict pathogenicity, host specificity or mycotoxin production. Concerning the classification and the systematic placement of the strains and morphospecies in the A. infectoria species–group, a polyphasic approach is needed, but there are inconsistencies between the different taxonomic features and we therefore refrain from recommending any taxonomic changes at this point in time. Acknowledgments The authors would like to thank EG Simmons for cultures and for suggestions to the manuscript and Jens C. Frisvad and Ulf Thrane for fruitful discussions. This project was a collaboration between CMB, DTU, Denmark and CBS-KNAW, Fungal Biodiversity Centre, The Netherlands and was supported in part by 1) a the SYNTHESYS Project (NL-TAF-1843), http://www.synthesys.info/, which is financed by European Community Research Infrastructure Action under the FP6 ‘‘Structuring the European Research Area” Programme”, 2) a grant from the Danish Directorate for Food, Fisheries and Agri Business (FFS05-3) and 3) the VILLUM KANN RASMUSSEN foundation. References Andersen, B., Hollensted, M., 2008. Metabolite production by different Ulocladium species. Int. J. Food Microbiol. 126, 172–179. Andersen, B., Thrane, U., 1996. Secondary metabolites produced by Alternaria infectoria and their use as chemotaxonomic markers. Mycotoxin Res. 12, 54–60. Andersen, B., Dongo, A., Pryor, B.M., 2008. Secondary metabolite profiling of Alternaria dauci, A. porri, A. solani, and A. tomatophila. Mycol. Res. 112, 241–250. Andersen, B., Hansen, M.E., Smedsgaard, J., 2005. Automated and unbiased image analyses as tools in phenotypic classification of small-spored Alternaria spp. Phytopathology 95, 1021–1029. Andersen, B., Krøger, E., Roberts, R.G., 2002. Chemical and morphological segregation of Alternaria arborescens, A. infectoria and A. tenuissima species– groups. Mycol. Res. 106, 170–182. Andersen, B., Nielsen, K.F., Thrane, U., Szara, T., Taylor, J.W., Jarvis, B.B., 2003. Molecular and phenotypic descriptions of Stachybotrys chlorohalonata sp. nov. and two chemotypes of Stachybotrys chartarum found in water-damaged buildings. Mycologia 95, 1227–1238. Andersen, B., Thrane, U., Svendsen, A., Rasmussen, I.A., 1996. Associated field mycobiota on malt barley. Can. J. Bot. 74, 845–858. Berbee, M.L., Pirseyedi, M., Hubbard, S., 1999. Cochliobolus phylogenetics and the origin of known, highly virulent pathogens, inferred from ITS and glyceraldehyde3-phosphate dehydrogenase gene sequences. Mycologia 91, 964–977. Brasch, J., Busch, J.-O., de Hoog, G.S., 2008. Cutaneous phaeohyphomycosis caused by Alternaria infectoria. Acta Derm. Venereol. 88, 160–161. Braun, U., Crous, P.W., Dugan, F., Groenewald, J.Z., de Hoog, G.S., 2003. Phylogeny and taxonomy of Cladosporium-like hyphomycetes, including Davidiella gen. nov., the teleomorph of Cladosporium s. str. Mycol. Prog. 2, 3–18. Author's personal copy 656 B. Andersen et al. / Fungal Genetics and Biology 46 (2009) 642–656 Bruen, T., Philippe, H., Bryant, D., 2006. A simple and robust statistical test for detecting the presence of recombination. Genetics 172, 2665–2681. Butler, E.E., Mann, M.P., 1959. Use of cellophane tape for mounting and photographing phytopathogenic fungi. Phytopathology 49, 231–232. Carbone, I., Kohn, L.M., 1999. A method for designing primer sets for speciation studies in filametous fungi. Mycologia 91, 553–556. Christensen, K.B., van Klink, J.W., Weavers, R.T., Larsen, T.O., Andersen, B., Phipps, R.K., 2005. Novel chemotaxonomic markers for the Alternaria infectoria species– group. J. Agric. Food Chem. 53, 431–9435. de Hoog, G.S., Gerrits van den Ende, A.H.G., 1998. Molecular diagnostics of clinical strains of filamentous Basidiomycetes. Mycoses 41, 183–189. de Hoog, G.S., Horré, R., 2002. Molecular taxonomy of the Alternaria and Ulocladium species from humans and their identification in routine laboratory. Mycoses 45, 259–276. Dubois, D., Pihet, M., Le Clec’h, C., Croué, A., Beguin, H., Bouchara, J.-P., Chabasse, D., 2005. Cutaneous phaeohyphomycosis due to Alternaria infectoria. Mycopathologia 160, 117–123. Dugan, F.M., Peever, T.L., 2002. Morphological and cultural differentiation of described species of Alternaria from Poaceae. Mycotaxon 83, 229–264. Dye, C., Johnson, E.M., Gruffydd-Jones, T.J., 2009. Alternaria species infection in nine cats. J. Feline Med. Surg. 11, 332–336. Frisvad, J.C., 1983. A selective and indicative medium for groups of Penicillium viridicatum producing different mycotoxins on cereals. J. Appl. Bacteriol. 54, 409–416. Frisvad, J.C., Samson, R.A., 2004. Polyphasic taxonomy of Penicillium subgenus Penicillium A guide to identification of food and air-borne terverticillate Penicillia and their mycotoxins. Stud. Mycol. 49, 1–173. Frisvad, J.C., Thrane, U., 1987. Standardized high-performance liquid chromatography of 182 mycotoxins and other fungal metabolites based on alkylphenone retention indices and UV–VIS spectra (diode array detection). J. Chromatogr. 404, 195–214. Hansen, M.E., 2003. Indexing and analysis of fungal phenotypes using morphology and spectrometry. Ph.D. Thesis. IMM, DTU, Denmark. ISSN 0909-3192. Haubold, B., Hudson, R.R., 2000. LIAN 3.0: detecting linkage disequilibrium in multilocus data. Bioinformatics 16, 847–849. Huson, D.H., Bryant, D., 2006. Application of phylogenetic networks in evolutionary studies. Mol. Biol. Evol. 23, 254–267. Kosiak, B., Torp, M., Skjerve, E., Andersen, B., 2004. Alternaria and Fusarium in Norwegian grains of reduced quality – a matched pair sample study. Int. J. Food Microbiol. 93, 51–62. Kwasna, H., Kosiak, B., 2003. Lewia avenicola sp nov and its Alternaria anamorph from oat grain, with a key to the species of Lewia. Mycol. Res. 107, 371–376. Laatsch, H., 2008. AntiBase 2008. The Natural Compound Identifier. Wiley-VCH GmbH & Co., Weinheim, Germany. Nielsen, K.F., Smedsgaard, J., 2003. Fungal metabolite screening: database of 474 mycotoxins and fungal metabolites for dereplication by standardised liquid chromatography-UV–mass spectrometry methodology. J. Chromatogr. A 1002, 111–136. Nielsen, K.F., Gräfenhan, T., Zafari, D., Thrane, U., 2005. Trichothecene production by Trichoderma brevicompactum. J. Agric. Food Chem. 53, 8190–8196. O’Donnell, K., Ward, T.J., Geiser, D.M., Kistler, H.C., Aoki, T., 2004. Genealogical concordance between the mating type locus and seven other nuclear genes supports formal recognition of nine phylogenetically distinct species within the Fusarium graminearum clade. Fungal Genet. Biol. 41, 600–623. Pedras, M.S.C., Chumala, P.B., 2005. Phomapyrones from blackleg causing phytopathogenic fungi: isolation, structure determination, biosyntheses and biological activity. Photochemistry 66, 81–87. Perelló, A.E., Sisterna, M.N., 2006. Leaf blight of wheat caused by Alternaria triticina in Argentina. Plant Pathol. 55, 303. Perelló, A.E., Moreno, M., Sisterna, M.N., 2008. Alternaria infectoria species–group associated with black point of wheat in Argentina. Plant Pathol. 57, 379. Pitt, J.I., Hocking, A.D., 1997. Fungi and Food Spoilage. Blackie Academic and Professional, London, UK. Prasada, R., Prabhu, A.S., 1962. Leaf blight of wheat caused by a new species of Alternaria. Indian Phytopathology 15, 292–293. Pryor, B.M., Bigelow, D.M., 2003. Molecular characterization of Embellisia and Nimbya and their relationship to Alternaria, Ulocladium and Stemphylium. Mycologia 95, 1141–1154. Rosendahl, S., 2008. Communities, populations and individuals of arbuscular mycorrhizal fungi. New Phytologist 178, 253–266. Rozas, J., Rozas, R., 1995. DnaSP, DNA sequence polymorphism: an interactive program for estimating Population Genetics parameters from DNA sequence data. Comput. Applications Biosci. 11, 621–625. Salemi, M., Gray, R.R., Goodenow, M.M., 2008. An exploratory algorithm to identify intra-host recombinant viral sequences. Mol. Phylogenet. Evol. 49, 618–628. Samson, R.A., Noonim, P., Meijer, M., Houbraken, J., Frisvad, J.C., Varga, J., 2007. Diagnostic tools to identify black Aspergilli. Stud. Mycol. 59, 129–145. Simmons, E.G., 1986. Alternaria themes and variations (22–26). Mycotaxon 25, 287– 308. Simmons, E.G., 2007. Alternaria. An Identification Manual. CBS Fungal Biodiversity Centre, Utrecht, The Netherlands. Simmons, E.G., Roberts, R.G., 1993. Alternaria themes and variations (73). Mycotaxon 48, 109–140. Stamatakis, A., Hoover, P., Rougemont, J., 2008. A rapid bootstrap algorithm for the RAxML web-servers. Syst. Biol. 75, 758–771. Taylor, J.W., Jacobson, D.J., Kroken, S., Kasuga, T., Geiser, D.M., Hibbett, D.S., Fisher, M.C., 2000. Phylogenetic species recognition and species concept in fungi. Fungal Genet. Biol. 31, 21–32. Wägele, J.W., Mayer, C., 2007. Visualizing differences in phylogenetic information content of alignments and distinction of three classes of long-branch effects. BMC Evol. Biol. 7, 147–170. White, T.J., Bruns, T., Lee, S., Taylor, J., 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis, M.A., Gelfand, D.H., Sninsky, J.J., White, T.J. (Eds.), PCR Protocols: A Guide to Methods and Applications. Academic Press, San Diego, pp. 315–322.