mycological research 113 (2009) 933–951 journal homepage: www.elsevier.com/locate/mycres Morphological and phylogenetic analyses of Pythium species in South Africa Adéle MCLEODa,*,1, Wilhelm J. BOTHAb,*,1, Julia C. MEITZa, Chris F. J. SPIESa, Yared T. TEWOLDEMEDHINa, Lizel MOSTERTa a University of Stellenbosch, Department of Plant Pathology, Private Bag X1, Matieland 7602, South Africa Agricultural Research Council, Plant Protection Research Institute, Private Bag X134, Pretoria, South Africa b article info abstract Article history: The genus Pythium is important in agriculture, since it contains many plant pathogenic Received 18 November 2008 species, as well as species that can promote plant growth and some that have biocontrol Received in revised form potential. In South Africa, very little is known about the diversity of Pythium species within 5 March 2009 agricultural soil, irrigation and hydroponic systems. Therefore, the aim of the study was to Accepted 21 April 2009 characterise a selection of 85 Pythium isolates collected in South Africa from 1991 through Published online 20 May 2009 to 2007. The isolates were characterised morphologically as well as through sequence and Corresponding Editor: phylogenetic analyses of the internal transcribed spacer regions (ITS) and the 5.8S gene of Gordon William Beakes the nuclear ribosomal DNA. Phylogenetic analyses showed that the isolates represented Keywords: and taxonomy of the genus Pythium. Mycological Research 108: 1363–1383]. Characterisation ten of the 11 published Pythium clades [Lévesque & De Cock, 2004. Molecular phylogeny Acanthicum of isolates in clade D and J suggested that the phylogenetic concept of Pythium acanthicum Chamaehyphon and Pythium perplexum respectively, needs further investigation in order to enable reliable Oomycetes species identification within these clades. Our phylogenetic analyses of Pythium species in Perplexum clade B also showed that species with globose sporangia group basal within this clade, and Violae are not dispersed within the clade as previously reported. The 85 South African isolates represented 34 known species, of which 20 species have not been reported previously in South Africa. Additionally, three isolates (PPRI 8428, 8300 and 8418) were identified that may each represent putative new species, Pythium sp. WJB-1 to WJB-3. ª 2009 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. Introduction The genus Pythium belongs to the phylum Oomycota and has a range of interactions with plants ranging from pathogenic to beneficial (Alexopoulos et al. 1996; Dick 2001a). Many Pythium species are important plant pathogens of several field and hydroponically grown crops, including deciduous fruit trees, vegetables, cereals and ornamentals (Alexopoulos et al. 1996; Chamswarng & Cook 1985; Gull et al. 2004; Ingram & Cook 1990; Martin & Loper 1999; Mazzola et al. 2002; Moorman et al. 2002). However, a substantial number of species within the genus are not pathogenic, and are saprophytic with some species even promoting plant growth and having potential as biocontrol agents (Martin & Loper 1999; Mazzola et al. 2002; Van der Plaats-Niterink 1981). Little is known about the diversity of Pythium species in South Africa. Denman & Knox-Davies (1992) and Crous et al. (2000) each published a review of the species present in South * Corresponding authors. Tel.:þ27 21 8084795; fax: þ27 21 8084956. E-mail addresses: adelem@sun.ac.za (A. McLeod), bothaw@arc.agric.za (W. J. Botha). 1 The first two authors contributed equally to the work. 0953-7562/$ – see front matter ª 2009 The British Mycological Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.mycres.2009.04.009 934 Africa. Altogether, these two publications listed a total of 20 species, considering only Pythium species names accepted by Dick (1990) and Van der Plaats-Niterink (1981). Subsequently, another twelve species were reported by other authors (Belbahri et al. 2008; Gull et al. 2004; Linde et al. 1994; Meyer & Van Dyk 2002; Moralejo et al. 2008). Thus, prior to our current study only 32 Pythium species have been reported from South Africa. Up until now, most of the Pythium species identified in South Africa were identified using morphological criteria, with the only exceptions being two newly described species, Pythium mercuriale and Pythium recalcitrans which used sequence data as well (Belbahri et al. 2008; Moralejo et al. 2008). Traditional identification of Pythium species based upon morphology is difficult due to the high variability observed, as well as intraspecific variation and overlap amongst species in their morphological characteristics (Dick 2001b; Martin 2000). Many Pythium isolates do not form sexual structures in culture, and can thus only be classified into groups based on their asexual structures (see Van der Plaats-Niterink 1981). The identification of heterothallic species, of which there are currently only a few described, often requires mating with compatible isolates to ascertain species identification (Van der Plaats-Niterink 1981). Molecular sequence data along with phylogenetic analyses are increasingly being used for species identification, as well as evolutionary inference in many living organisms, including Pythium. The first comprehensive molecular phylogeny of Pythium was published by Lévesque & De Cock (2004) and included analyses of the internal transcribed spacer regions (ITS) and the 5.8S gene of the nuclear ribosomal DNA of 116 accepted Pythium species. Their phylogenetic analyses divided the species into 11 major clades (Clades A–K). The ITS sequences of this study can be used for species identification, although a few Pythium species have identical or near identical ITS sequence data and require further morphological characterisation, and some may be con-specific (Lévesque & De Cock 2004). The mitochondrial cytochrome c oxidase subunit II (cox II) region and b-tubulin gene have also been used for the characterisation of a small number of Pythium species (Bailey et al. 2002; Martin 2000; Villa et al. 2006). The aim of the study was to investigate the diversity of Pythium species present in South African agricultural systems, through the characterisation of a selection of 85 isolates that were collected between 1991 and 2007. The isolates were characterised using morphological as well as ITS and 5.8S sequence data of the rDNA region, along with phylogenetic analyses. Materials and methods Pythium isolates and culturing Pythium isolates were obtained from agricultural systems including field soils, irrigation water and hydroponic systems (Table 1), using standard isolation techniques (Singleton et al. 1992). In total, 295 morphologically identified Pythium isolates were collected from 1991 to 2006. A subset of 85 isolates was selected from this collection that represented the morphological variation in the collection. Unfortunately some of the early deposited cultures did not survive storage and could A. McLeod et al. not be included. All isolates were hyphal tipped at least three times prior to morphological and sequence analyses. Isolates were routinely grown on cornmeal agar (Difco Laboratories, MI, USA) at 25  C, and were stored at 15  C in sterile distilled water (25 ml) in screw capped McCarthy bottles with several 2 cm pieces of sterile grass leaflets plus one 5 cm sterile tooth pick per bottle. All the cultures were submitted to the culture collection of the Plant Protection Research Institute (Agriculture Research Council, Plant Protection Research Institute, South Africa), which houses the national collection of fungi in South Africa and is internationally known as the PREM culture collection. Morphological characterisation and identification Colonies were sub-cultured on soft water agar (0.7 % agar and 30 mg ß-sitosterol per 1iter) and incubated for 3–5 d at 25  C. Plates were inverted and checked under a stereomicroscope for the formation of oospores, hyphal swellings, chlamydospores and appressoria. Sterile soil extract (20 g sandy soil suspension in 1 l distilled water, filtered and autoclaved at 121  C at 15 kPa for 20 min) in 65 mm-petri-dishes was inoculated with ten colonised plugs (5 mm2 diam) from soft water agar plates. Hemp seeds (Cannabis sativa L.) and sterile grass blades e.g. kikuyu (Pennisetum clandestinum Hochst. ex Chiov) or LM grass (Dactyloctenium australe Steudel), were floated on the sterile soil extract and incubated at 25  C under cool white fluorescent light until sporangial formation was observed. For species identification, original species descriptions and the dichotomous keys of Dick (1990), based on oogonial criteria and the revised key of van der Plaats-Niterink (Dick 1990), were consulted. In addition, species descriptions in the monograph of Pythium (Van der Plaats-Niterink 1981) were used. Species were identified by using all sexual and asexual morphological structures observed in culture including biometric data, as recommended by Dick (1990) and Shahzad et al. (1992). DNA isolation and ITS sequencing DNA was isolated from mycelium collected from 5 to 10 day old V8 agar plates (Miller 1955), using the WizardÒ SV Genomic DNA Purification system (Promega Corporation, Madison, WI, USA) according to manufacturer’s instructions. The ITS 1 and 2 regions and 5.8S gene were amplified using primers ITS6 (Cooke & Duncan 1997) and ITS4 (White et al. 1990). In a few isolates where amplification could not be obtained using these primers, the primer pair UN-UP18S42 and PY-LO28S22 (Bakkeren et al. 2000; Lévesque & De Cock 2004) was used. PCR reactions for both primer pairs consisted of 0.2 mM of each primer, 10 mM of each dNTP, 1 PCR buffer (Bioline USA Inc., Taunton, MA), 0.7 U BIOTAQ DNA polymerase (Bioline), 2 mg bovine serum albumin (BSA) Fraction V (Roche Diagnostics South Africa, Randburg), 5 mL DNA and 2 mM MgCl2 in a final volume of 40 mL. Amplifications were conducted in a 2700 Applied Biosystems (Foster City, CA) machine, starting with an initial denaturation of 5 min at 94  C, followed by 36 cycles of 30 s at 94  C, 30 s at 52  C and 60 s at 72  C and a final extension cycle of 7 min at 72  C. PCR amplification products were cleaned using the WizardÒ SV gel and PCR Clean-up System (Promega Corporation). PCR products Pythium species in South Africa 935 Table 1 – Morphological and sequence identification of South African Pythium isolates obtained from 1991 to 2007 from different regions and host/substrates in South Africa PPRI codea GenBank accessionb Region Isolation date Clade A 9009 FJ415895 Eastern Province 2006 9010 FJ415896 Limpopo Province Clade B 8587 8588 8589 FJ415897 FJ415898 FJ415912 8531 9011 8291 8292 8590 8591 8294 8592 8593 8297 8295 8296 8594 8759 8595 Host or substrate Morphological species name ITS sequence species namec P. aphanidermatum P. aphanidermatum 2007 Lycopersicon esculentum Cucumis sativus P. aphanidermatum P. aphanidermatum Gauteng Western Province North-West 2004 2005 1997 Lolium sp. Vitis vinifera Lactuca sativa P. torulosum*P. torulosum* P. tracheiphilum - FJ415900 FJ415899 FJ415904 FJ415905 FJ415901 FJ415902 FJ415903 FJ415913 FJ415914 FJ415909 FJ415910 FJ415911 FJ415906 FJ415907 FJ415908 Limpopo Gauteng North West Gauteng Mpumalanga Mpumalanga Mpumalanga Freestate Western Province Mpumalanga Mpumalanga Freestate Western Province Western Province Western Province 2007 2006 1991 1995 1993 1993 2005 2001 2005 1998 1998 1997 2005 2005 2005 Cucumis sativus Cucumis sativus Arum esculentum Arum esculentum Coffea arabica, soil Coffea arabica, soil Vitis vinifera, soil Triticum aestivum Vitis vinifera Triticum aestivum Triticum aestivum Triticum aestivum Vitis vinifera Vitis vinifera, soil Vitis vinifera, soil P. torulosum P. torulosum hyphal swelling-group with chlamydospores P. catenulatum P. catenulatum P. myriotylum P. myriotylum P. periilum P. periilum P. periilum P. vanterpoolii P. vanterpoolii P. aristosporum P. aristosporum P. aristosporum P. pyrilobum P. coloratum P. coloratum P. catenulatum*P. catenulatum* P. myriotylum* P. myriotylum* P. periilum*P. periilum* P. periilum* P. vanterpoolii P. vanterpoolii P. aristosporum*P. aristosporum* P. aristosporum* P. pyrilobum P. coloratum* P. coloratum* Clade D 8596 8597 8410 8598 8411 FJ415919 FJ415918 FJ415915 FJ415916 FJ415917 Western Province Mpumalanga Mpumalanga Western Province Mpumalanga 2001 1998 2005 2005 2005 Prunus armeniaca Avena sativa Macadamia sp. Vitis vinifera Macadamia sp. P. P. P. P. P. P. P. P. P. P. Clade E 8599 8600 7453 7472 9005 8633 8638 8601 8602 FJ415928 FJ415924 FJ415925 FJ415927 FJ415926 FJ415922 FJ415923 FJ415920 FJ415921 Western Province Western Province Mpumalanga Mpumalanga North West Western Province Western Province Western Province Western Province 2005 2005 2002 2005 2006 2006 2006 2005 2005 Vitis vinifera Vitis vinifera Zea mays, soil Zea mays, soil Brassica oleracea Malus sp. Malus sp., soil Vitis vinifera Vitis vinifera P. rostratifingens P. rostratifingens P. rostratifingens P. rostratifingens P. rostratifingens P. minus P. minus P. echinulatum P. echinulatum P. rostratifingens P. rostratifingens P. rostratifingens P. rostratifingens P. rostratifingens P. minus*P. minus* P. echinulatum*P. echinulatum* Clade F 8603 8604 8605 8608 8609 8610 8611 8612 8613 8634 8614 8635 8636 8637 8607 9008 9006 FJ415933 FJ415934 FJ415932 FJ415929 FJ415931 FJ415930 FJ415940 FJ415941 FJ415935 FJ415936 FJ415937 FJ415943 FJ415944 FJ415945 FJ415939 FJ415938 FJ415942 Limpopo Western Province Western Province Eastern Province Western Province Western Province Western Province Western Province Western Province Western Province Western Province Western Province Western Province Western Province Western Province Gauteng Western Province 1992 2005 2005 1994 2005 2001 2005 2005 2006 2006 2005 2006 2006 2006 2005 2007 2007 Citrus sp., soil Vitis vinifera Vitis vinifera Pisum sativum Vitis vinifera Agricultural soil Vitis vinifera Vitis vinifera Malus sp. Mauls sp. Vitis vinifera Malus sp., soil Malus sp., soil Malus sp., soil Vitis vinifera Hydroponic system Fragaria sp. P. spinosum P. spinosum P. kunmingense P. mamillatum P. mamillatum P. mamillatum P. paroecandrum P. paroecandrum hyphal swelling-group hyphal swelling-group P. violae hyphal swellings & appressoria hyphal swellings & appressoria hyphal swellings & appressoria P. irregulare or P. cryptoirregulare P. irregulare or P. cryptoirregulare hyphal swelling-group P. spinosum* P. spinosum* P. kunmingense*P. mamillatum P. mamillatum P. mamillatum P. paroecandrumP. paroecandrum P. sylvaticum P. sylvaticum P. species P. attrantheridium P. attrantheridium P. attrantheridium P. irregulare P. cryptoirregulare P. macrosporum - oligandrum oligandrum acanthicum acanthicum acanthicum oligandrum* oligandrum* acanthicum species species (continued on next page) 936 A. McLeod et al. Table 1 – (continued) PPRI codea GenBank accessionb Region Isolation date Host or substrate Morphological species name ITS sequence species namec Clade G 8300 FJ415946 Northern Cape Province 2005 Hoodia gordonii P. iwayamai P. species WJB-1 Clade H 9007 8508 FJ415947 FJ415948 Western Province Western Province 2006 2006 Diospyros sp. Malus sp. P. helicandrum P. helicandrum P. helicandrum P. helicandrum Clade I 7133 8616 8617 8419 8618 8619 8415 8620 8416 8621 FJ415954 FJ415955 FJ415957 FJ415956 FJ415958 FJ415951 FJ415953 FJ415952 FJ415949 FJ415950 Gauteng Western Province Western Province North West Western Province Mpumalanga North West Gauteng Mpumalanga Gauteng 1996 2005 2005 1993 2005 1995 1995 1993 1993 1996 Lactuca sativa Vitis vinifera Vitis vinifera Cucumis sativus Vitis vinifera Coffea arabica Brasssica oleracea Juniperus magnolia Citrus sp. Olea africana, soil P. P. P. P. P. P. P. P. P. P. Clade J 8424 FJ415960 Mpumalanga 1996 P. perplexum P. species 8403 8405 8622 8623 8428 8402 FJ415961 FJ415964 FJ415962 FJ415965 FJ415963 FJ415959 Limpopo Gauteng Limpopo Western Province Mpumalanga North West 2005 2003 2006 2005 2005 1994 Lycopersicon esculentum, soil Citrus sp. Olea africana, soil Citrus sp. Vitis vinifera Macadamia sp. Brasssica oleracea P. perplexum P. perplexum P. perplexum P. perplexum P. perplexum P. polymastum or P. jasmonium (nom. inval) P. P. P. P. P. P. Clade K 8408 8624 8625 8401 FJ415970 FJ415974 FJ415975 FJ415979 Gauteng Western Province Western Province Mpumalanga 1996 2005 2005 2003 P. oedochilum Pythium P-group sporangia Pythium P-group sporangia P. vexans species complex P. oedochilum P. chamaehyphon P. chamaehyphon P. vexans 9086 8632 8626 8423 8417 8627 8628 8629 8631 8418 FJ415977 FJ415978 FJ415972 FJ415973 FJ415971 FJ415966 FJ415968 FJ415969 FJ415967 FJ415976 Mpumalanga Western Province Western Province North West Kwazulu-Natal Western Province Western Province Western Province Western Province Mpumalanga 2007 2005 2005 2006 1991 2004 2005 2005 2006 1996 P. vexans species complex P. vexans species complex P. helicoides P. helicoides P. helicoides Pythium P-group sporangia Pythium P-group sporangia Pythium P-group sporangia Pythium P-group sporangia P. species P. P. P. P. P. P. P. P. P. P. Cynara cardunculus Vitis vinifera soil Vitis vinifera, soil Encephalartos middelburgensis, soil Persea americana Acacia xanthophloea Vitis vinifera Fragaria ananatis Soil, wetlands Angustum betulina Vitis vinifera Vitis vinifera Irrigation water Persea americana ultimum var. ultimum heterothallicum heterothallicum heterothallicum heterothallicum splendens splendens splendens splendens splendens P. P. P. P. P. P. P. P. P. P. ultimum var. ultimum heterothallicum*heterothallicum* heterothallicum* heterothallicum* splendens splendens splendens splendens splendens species perplexum species perplexum species WJB-2 jasmonium - vexans vexans helicoides helicoides helicoides litorale litorale litorale litorale species WJB-3 a Cultures of all isolates were deposited at the Plant Protection Research Institute (PPRI). b Accession numbers of the internal transcribed spacers (ITS) sequence of each isolate submitted to GenBank. c Species names followed by a ‘‘-’’ are being reported for the first time in South Africa. Species names followed by a ‘‘*’’ have identical or near identical ITS sequences to other Pythium species according to Lévesque & De Cock (2004). Only the species names that matched the morphological description are presented. The species to which these sequences have identical or near identical ITS sequence matches can be seen in the phylogenetic trees (Figs 1–8), or refer to Lévesque & De Cock (2004). were sequenced mainly using the ITS4 and ITS6 primers, but also primers UN-UP18S42, PY-LO28S22, OOM-UP5.8S01 and OOM-LO5.8S47B (Lévesque & De Cock 2004). Sequencing analyses were conducted by the Central Analytical Sequencing Facility at Stellenbosch University using the BigDye system (version 3.1 dye terminators, Applied Biosystem) and an ABI 3130XL Genetic Analyzer. Geneious Pro (Biomatters Ltd., Auckland, New Zealand) was used to view ABI trace files, and to obtain consensus double strand sequences for each isolate. Heterozygous or ambiguous sites were labelled using IUPAC codes. The consensus sequences obtained from the double strand sequence data of all isolates were deposited to the NCBI GenBank database as accession numbers FJ415895 to FJ415979. Cloning and sequencing of ITS region For some of the Pythium isolates ITS sequence data could not be obtained using direct sequencing of PCR products, since direct sequencing consistently yielded ambiguous, poor Pythium species in South Africa quality sequence data. The ambiguous sequence data was mostly due to single or two base pair indels within the ITS1 and ITS2 region. The species for which ITS sequencing data had to be obtained from cloned PCR products included Pythium mamillatum, Pythium litorale, Pythium helicoides, Pythium acanthicum, Pythium splendens, Pythium heterothallicum and Pythium rostratifingens. PCR products were cloned using InsTAcloneÔ PCR Cloning Kit (Fermentas Inc., Glen Burnie, MD) according to manufacturer’s instruction. Two clones were selected from each isolate, and consensus sequence data was obtained and analysed as described above. For all isolates the two selected clones yielded identical sequence data. Species identification using ITS sequence data The species identity of isolates was determined by first conducting a BLAST analysis for each sequence. The results of the BLAST analyses were used to select the sequence/s to which each South African isolate had the highest sequence similarity. These GenBank sequences along with all the extype and authentic Pythium species sequences deposited by Lévesque & De Cock (2004) and sequences of new species that were described subsequently were used in phylogentic analyses for each of the respective Pythium clades of Lévesque & De Cock (2004). For known Pythium species complexes such as Pythium vexans and Pythium irregulare, or species that are known to contain variation within the species, several GenBank sequences with high similarity to the South African isolates were selected, in order to determine whether the South African isolates still grouped within the species complex. Phylogenetic analyses The sequences of each clade were aligned using MAFFT sequence alignment program version 6 (Katoh & Toh 2008) followed by manual adjustments of the alignments in Sequence Alignment Editor v. 2.0a11 (Rambaut 2002). For each Pythium clade maximum parsimony analysis as well as Bayesian analysis was conducted. Maximum parsimony analysis was conducted using PAUP* (Phylogenetic Analysis Using Parsimony) v. 4.0b10. The analysis was performed using the heuristic search option with 10 random taxon additions. Tree bisection and reconstruction (TBR) was used as the branch swapping algorithm with the option of saving no more than 10 trees with a score greater than or equal to 5 (Harrison & Langdale 2006). Gaps were treated as missing data. All characters were unordered and of equal weight. Bootstrap support values were calculated from 1000 heuristic search replicates and 10 random taxon additions. Other measures calculated for parsimony included tree length (TL), consistency index (CI), retention index (RI) and the rescaled consistency index (RC) values. Bayesian analyses were conducted using MrBayes v. 3.0b4 (Ronquist & Huelsenbeck 2003). The program MrModeltest (J.J.A. Nylander, available from the inwww.ebc.uu.se/systzoo/staff/nylander.html) was ternet: used for selecting the optimal model of sequence evolution for each clade alignment. The likelihood and prior settings were changed in MrBayes according to the models found with MrModeltest for each partition. Markov chains were 937 initiated from a random tree and run for 1 million generations, keeping one out of every 100th generation. Convergence among chains was monitored by examining plots of log-likelihood values and observing when the values of the four chains have reached a plateau. The first 300 000 generations (burn-in) were discarded for the analyses of clades B, E, F, J and K, whereas the first 100 000 generations were discarded for the analyses of clades D, G, and I. The remaining samples were used to calculate the 50 % majority-rule tree and the posterior probability for the individual branches. Results and discussion This study, which identified 34 known Pythium species, is the first study where both molecular and morphological data were used for the identification of Pythium species in South Africa. Twenty of the reported species have not been reported previously in South Africa. Furthermore, three isolates were identified that possibly represent new species, Pythium sp. WJB-1, WJB-2 and WJB-3 (Table 1; Figs 5, 7 and 8). The South African Pythium species represented ten of the 11 major Pythium clades (designated clade A–K) identified by Lévesque & De Cock (2004) (Table 1). The only clade for which no isolates have been reported from in South Africa is clade C that includes Pythium grandisporangium and Pythium insidiosum (Lévesque & De Cock 2004). In addition to the 34 Pythium species that have been reported in our study, 18 additional species have previously been reported from South Africa. These species were almost all identified using only morphological characteristics, except for Pythium recalcitrans (Moralejo et al. 2008) and Pythium mercuriale (Belbahri et al. 2008) for which sequence data were also provided. In total, 52 Pythium species have now been reported from South Africa. The 18 species that have previously been reported in South Africa will be discussed under each clade below, except for Pythium hemmianum and Pythium hypogenum that were reported by Darvas et al. (1978) whose clade status in the absence of sequence data is unknown. Only isolates for which the species identification was unclear or not straightforward will be discussed, using the clade notation proposed by Lévesque & De Cock (2004). Clade A Clade A contained only two South African Pythium aphanidermatum isolates (PPRI 9009 and 9010), a species previously reported from South Africa by Wager (1941). The two South African sequences were very similar to the P. aphanidermatum CBS sequence (GI51235476). Clade B Eighteen South African isolates encompassing 9 species, including six previously unrecorded species, belonged in clade B (Table 1; Fig 1). Other morphologically identified clade B species that have previously been reported from South Africa include Pythium angustatum (Linde et al. 1994), Pythium arrhenomanes (Scott 1987), Pythium diclinum (Gull et al. 2004), Pythium dissotocum (Botha & Coetzer 1996) and Pythium 938 A. McLeod et al. Fig 1 – Phylogeny of Pythium species in clade B (Lévesque & De Cock 2004) based on the ITS1, 5.8S and ITS2 regions of nuclear rDNA. The tree presents one of the 670 equally parsimonious trees of a heuristic search. Numbers within the tree represent the bootstrap values followed by probability values in brackets. Branches with a 100 % bootstrap support and probability value of 1.00, are denoted by a star symbol. Bootstrap values lower than 60 % are not shown. Length [ 892, CI [ 0.643, RI [ 0.865, and RC [ 0.557. Numbers preceding species are the PPRI numbers of South African isolates. Sequences of species used by Lévesque & De Cock (2004) that are the ex-type, authentic strain or strains used for description in the monograph of Van der Plaats-Niterink (1981) are indicated in bold, as well as the three recently described species P. rhizo-oryzae (Bala et al. 2006), P. kashmirense (Paul & Bala 2008) and P. phragmitis (Nechwatal et al. 2005). Sequences that may represent incorrect GenBank submissions are indicated by a black dot. salpingophorum (Linde et al. 1994). Of the previously unrecorded species, Pythium tracheiphilum is considered as a serious pathogen of lettuce (Gonzalez et al. 2004; Tortolero & Sequeira 1978), from which it was also isolated in our study (Table 1). Both Pythium aristosporum and Pythium vanterpoolii are reported for the first time from the Southern hemisphere. The overall topology of our clade B phylogeny agreed, with a few exceptions with the phylogeny of Lévesque & De Cock Pythium species in South Africa (2004). Our analyses also only showed good bootstrap and probability support (96 %, 0.98) for one (subclade B2) (Fig 1) of the two previously identified subclades (B1 and B2). However, in contrast to the phylogeny of Lévesque & De Cock (2004), our subclade B2 was flanked by species representing their subclade B1. Lévesque & De Cock (2004) found that P. tracheiphilum, P. salpingophorum and Pythium conidiophorum clustered among several clade B species within subclade B1. In contrast, our analyses showed high bootstrap and probability support (94 %, 1.00) for a monophyletic basal cluster containing these species (Fig 1), which, except for P. conidiophorum whose sporangial type is unknown, all form globose sporangia, instead of the filamentous sporangia produced by the rest of the clade B species. Therefore, our clade B phylogeny supports the hypothesis of Lévesque & De Cock (2004) that the globose sporangial type is likely to be ancestral to the filamentous type. As noted by Lévesque & De Cock (2004), several of our South African Clade B isolates had ITS sequences that were almost identical to more than one Pythium species, and could therefore only be identified using additional morphological characteristics. These isolates included Pythium coloratum (PPRI 8759 and 8595), Pythium periilum (PPRI 8590, 8591 and 8294), Pythium myriotylum (PPRI 8291 and 8292) and Pythium aristosporum (PPRI 8295, 8296, 8297) (Table 1). The morphological characteristics of P. coloratum are: i) bluish lilac colour of thick oospore wall ii) antheridia are mainly diclinous and stalked bearing 1 to 2 antheridial cells iii) stalks sometimes coiling around oogonium iv) some oogonia are monopapillate and v) aplerotic oospores. P. periilum possess inflated filamentous sporangial filaments in combination with plerotic oospores with several branched, antheridial stalks enveloping oogonium and vegetative hyphae coiling around oogonium and stalk. Typical characters for P. myriotylum are clusters of clavate, digitate appressoria adhering to bottom of petri dish, numerous branched antheridial stalks enveloping aplerotic oogonium and high maximum temperature tolerance. P. aristosporum can be distinguished from P. arrhenomanes by i) fewer number of antheridia per oogonium ii) mono- and diclinous antheridia and iii) aplerotic oospores. Phylogenetic analysis showed that the aforementioned isolates each grouped with high bootstrap and probability support (more than 97 %, 0.97) in clades containing the different species to which they are known to have almost identical ITS sequences (Lévesque & De Cock 2004; Fig 1). The species boundaries for P. periilum and Pythium graminicola should be redefined, together with a redescription of both species to validate con-specificity or maintaining separate species. Similarly P. coloratum shows minimal morphological differences compared with P. dissotocum, P. diclinum, P. marinum and Pythium lutarium (Lévesque & De Cock 2004). Pythium torulosum, Pythium folliculosum, Pythium catenulatum and Pythium rhizo-oryzae also have similar ITS sequences (Bala et al. 2006; Lévesque & De Cock 2004) and all clustered within a monophyletic clade with high bootstrap and probability support (90 %, 0.99). Two South African isolates (PPRI 8588 and 8587) that had been morphologically identified as P. torulosum, clustered with good bootstrap and probability support (86 %, 1.00) with the CBS sequences of P. torulosum (GI51235478) and P. folliculosum (GI51235530) (Fig 1; Table 1). The 939 characteristics of P. torulosum include the presence of globose plerotic oospores and mainly monoclinous antheridia, whereas P. folliculosum has elongated to pyriform oogonia with 1–3 oospores per oogonium. The two South African isolates (PPRI8531 and 9011) that were morphologically identified as P. catenulatum (Table 1; Fig 1) grouped with the reference P. catenulatum isolate (CBS 842.68; GI51235529) and P. rhizo-oryzae (GI28883554). There was no support for grouping of the P. catenulatum sequences in a cluster separate from the P. folliculosum and P. tolurosum cluster, indicating that these species are all clearly closely related at the ITS sequence level. P. catenulatum has morphological characteristics (catenulate hyphal swellings and strains that are usually heterothallic) that differ sufficiently from those of P. torulosum and P. folliculosum to allow accurate identification of this taxon. The morphology of P. catenulatum and P. rhizo-oryzae, a new species recently described by Bala et al. (2006), differ only with regard to the lack of lobulate inflated sporangia and release of zoospores, both of which may be influenced by cultural conditions. Therefore the possibility exists that these two species with near identical ITS sequences may be con-specific. Clade D Five South African isolates, morphologically representing Pythium oligandrum and Pythium acanthicum, fitted into clade D and have both been previously reported by Wager (1941) (Table 1; Fig 2). The ITS sequences of P. oligandrum, Pythium hydnosporum and Pythium amasculinum, that are known to be almost identical (Lévesque & De Cock 2004), all clustered in a monophyletic group with only 69 % bootstrap support, along with the two South African isolates (PPRI and 8596 and 8597) that were identified morphologically as P. oligandrum (Fig 2). These three species have minimal morphological differences and should be investigated for possible con-specificity (Lévesque & De Cock 2004). The three South African isolates PPRI 8410, 8411 and 8598 all morphologically resembled P. acanthicum (Table 1), but their ITS sequences showed a range of similarities (97–100 %) to the CBS sequence (GI51235471) of this species (Lévesque & De Cock 2004). There was good bootstrap and probability support (82 %, 0.96) for PPRI 8410 being included in the clade with the CBS P. acanthicum sequence (GI51235471). However, PPRI 8598 clustered with high bootstrap and probability support (98 %, 0.99) along with two Pythium acanthophoron GenBank sequences (GI30525735, GI6466723) (Fig 2) that were discussed by Lévesque & De Cock (2004) as possibly representing a new species. The identity of PPRI8411 and the GenBank sequence GI64429860 (KS121) are also uncertain, since both grouped with only 62 % bootstrap support within the P. acanthicum clade. This suggests that P. acanthicum may be a species complex that requires further study. Clade E The nine isolates that fitted into clade E encompass three species, all of which are new reports for South Africa (Table 1; Fig 3). In addition to these three species, Pythium hypogynum (Darvas et al. 1978) and Pythium rostratum (Linde et al. 1994) have previously been recorded in South Africa using morphological 940 A. McLeod et al. Fig 2 – Phylogeny of Pythium species in clade D (Lévesque & De Cock 2004) based on the ITS1, 5.8S and ITS2 regions of nuclear rDNA. The tree presents one of two equally parsimonious trees of a heuristic search. Numbers within the tree represent the bootstrap values, with bootstrap values lower than 60 % not shown. Length [ 356, CI [ 0.952, RI [ 0.922, and RC [ 0.878. Numbers preceding species are the PPRI numbers of South African isolates. Sequences of species used by Lévesque & De Cock (2004) that are the ex-type, authentic strain or strains used for description in the monograph of Van der Plaats-Niterink (1981) are indicated in bold. Sequences that may represent incorrect GenBank submissions are indicated by a black dot. characteristics. The South African Pythium minus isolates (PPRI 8638 and 8633) matched the morphological description of P. minus (Ali-Shtayeh & Dick 1985), a species which has rarely been isolated (Lévesque & De Cock 2004). These isolates grouped with high bootstrap and probability support (100 %, 1.00) with the representative CBS sequence (GI51235552) of this species, as well as with Pythium pleroticum and Pythium ramificatum that all have similar ITS sequences (Fig 3). Indeed, P. minus and P. ramificatum may be con-specific, since they also share virtually identical morphological characteristics. Studies on the con-specificity of these species will require a redescription of P. pleroticum, since the original description of P. pleroticum (Ito 1944) is vague and confusing with poorly described morphological characters. Isolates PPRI 8601 and 8602 were morphologically identified as Pythium echinulatum based on the presence of more than one Pythium species in South Africa 941 Fig 3 – Phylogeny of Pythium species in clade E (Lévesque & De Cock 2004) based on the ITS1, 5.8S and ITS2 regions of nuclear rDNA. The tree presents one of the 729 equally parsimonious trees of a heuristic search. Numbers within the tree represent the bootstrap values followed by probability values in brackets. Branches with a 100 % bootstrap support and probability value of 1.00, are denoted by a star symbol. Bootstrap values lower than 60 % are not shown. Length [ 1482, CI [ 0.664, RI [ 0.864, and RC [ 0.574. Numbers preceding species are the PPRI numbers of South African isolates. Sequences of species used by Lévesque & De Cock (2004) that are the ex-type, authentic strain or strains used for description in the monograph of Van der PlaatsNiterink (1981) are indicated in bold, as well as four other species P. ramificatum (Paul 1986), P. apiculatum (Paul 2006), P. longisporangium (Paul et al. 2005.), P. rostratifingens (De Cock & Lévesque 2004) not included by Lévesque & De Cock (2004). hypogenous or monoclinous antheridia and their more widely spaced oogonial projections with a broad base (Table 1). In contrast, Pythium erinaceum has a single diclinous antheridium and more densely packed slender (narrow-based) acuminate oogonial projections. The South African P. echinulatum isolates grouped with good bootstrap and probability support (86 %, 1.00) with the CBS sequence (GI51235493) for this species as well as with the CBS sequence for P. erinaceum (GI51235548) 942 and the type sequence of Pythium ornacarpum (GI6554173) (Fig 3). The validity of the different species in this morphologically similar subclade requires clarification. Five South African isolates (PPRI 8599, 8600, 7453, 7472 and 9005) were identified as Pythium rostratifingens (Table 1). The isolates showed a range of sequence similarities (97–100 %) with the type sequence (GI53830722) of this species, and grouped within a strongly supported (100 %, 1.00) monophyletic clade only containing isolates of P. rotstratifingens (Fig 3). All the South African P. rostratifingens isolates were initially morphologically identified as Pythium connatum (Yü 1973), a species which shares some of the morphological characteristics of P. rostratifingens. However, after P. rostratifingens was described by De Cock & Lévesque (2004), it was apparent that the morphological structures and colony growth patterns of the South African isolates correlated well with the newly described P. rostratifingens (De Cock & Lévesque 2004). Clade F The 17 clade F isolates in our study represented nine known species, five of which had not previously been recorded from South Africa (Table 1). In addition to these nine species, Pythium debaryanum (Wager 1941), Pythium intermedium (Linde et al. 1994) and Pythium recalcitrans (Moralejo et al. 2008) from this clade, have all been previously reported from South Africa. The topology of our clade F was generally in good agreement with that of Lévesque & De Cock (2004). However, the analyses of our Pythium violae sequences differed from theirs that showed that four P. violae sequences only grouped into two clades (F and G). We found that seven P. violae sequences (PPRI 8614; GI105301047, 6468690, 51235571, 51235560, 102993956 and 51235569) grouped into three clades (F, G and I) (Figs 4–6) with one of Lévesque & De Cock’s (2004) clade G sequences of this species (GI51235569) falling into our clade I (Fig 7). The P. violae sequences included in this study were from diverse hosts including Viola (GI6468690, Matsumoto et al. 1999; GI105301047, Villa et al. 2006; GI51235571, CBS132.37), carrot (CBS 178.76, GI51235569 and GI10299395, Klemsdal et al. 2008) and grapevines (PPRI 8614). Furthermore, as discussed by Lévesque & De Cock (2004), it is clear that several genetically distinct Pythium isolates have the morphological identity of P. violae. The South African P. violae clade F isolate (PPRI 8614) and the three GenBank P. violae sequences (GI105301047, GI6468690, GI51535571) formed a monophyletic clade with high bootstrap and probability support (95 %, 1.00), and were closely related to P. debaryanum and Pythium viniferum (Fig 4). P. viniferum and P. debaryanum are highly similar at the ITS sequence level (99 %) (Paul et al. 2008). The original species description for P. debaryanum is imprecise and confusing (Braun 1925) and the validity of this species remains doubtful (Van der Plaats-Niterink 1981). Pythium irregulare is known as a species complex that contains a substantial amount of genetic, as well as morphological variation (Garzón et al. 2007; Matsumoto et al. 2000; Van der Plaats-Niterink 1981). Within this species complex, Matsumoto et al. (2000) have identified four genetic groups (I–IV), with group III and IV possibly representing a new species A. McLeod et al. (Garzón et al. 2007; Lévesque & De Cock 2004). Recently, Garzón et al. (2007) described Pythium cryptoirregulare as a new species within the P. irregulare species complex using AFLPs, ITS and cox II sequence data (Garzón et al. 2007). Although Pythium regulare, Pythium cylindrosporum and Pythium cryptoirregulare have near identical ITS sequences, they each have distinct morphological characteristics which enable them to be separated (Garzón et al. 2007; Lévesque & De Cock 2004). Our phylogenetic analyses showed that the P. irregulare complex consisted of one major clade containing two subclades with good bootstrap (75 %, 98 %) and probability support (0.93, 1.00) (Fig 4). The South African isolates PPRI 8607 and 9008 grouped within the P. cryptoirregulare and P. irregulare s.s. subclusters respectively (Fig 4) and therefore could be identified as these species. Isolate PPRI8605 clustered within a strongly supported (100 %, 1.00) monophyletic group containing the CBS sequences of Pythium kunmingense (GI51235554) and Pythium spinosum (GI51235555), as well as PPRI8603 and 8604 (Fig 4). It seems likely that P. spinosum may be con-specific with P. kunmingense, since the two species share many morphological characters except for the number and arrangement of oogonial projections and the plerotic state of the oospores (Yü, 1973; Van der Plaats-Niterink 1981). The sexual structures of PPRI8605 fitted the description of P. kunmingense, and the isolate was therefore identified as this species. However, P. kunmingense, which has only been isolated rarely, was considered a doubtful species by Van der Plaats-Niterink (1981) but as a valid species by Dick (2001b). This species requires re-examination. The South African isolates that were identified through sequence analysis as Pythium macrosporum (PPRI9006) and Pythium attrantheridium (PPRI8635 to 8637) did not form sexual structures that allowed their morphological identification (Table 1). Both of these species are heterothallic (De Cock & Lévesque 2004; Van der Plaats-Niterink 1981). The sequences of the South African isolates grouped with 100 % bootstrap and at least 0.99 probability support with the representative CBS sequences of each species (Fig 4). Clade G Our study included only one clade G isolate PPRI8300 that represents a putative new species. Previously, only Pythium violae has been reported from South Africa (Linde et al. 1994). However, the true identity of this isolate remains uncertain in the absence of molecular data, due to the morphological species concept of P. violae being uncertain (discussed under clade F). Isolate PPRI 8300 was identified morphologically as possibly being Pythium iwayamai based on the descriptions for this species by Iwayama (1933), Ito & Tokunaga (1935) and Hirane (1960). However, these species descriptions were confusing and imprecise with regard to the plerotic state of oospores, mode of antheridial attachment, as well as variable dimensions that were recorded for asexual structures. The sequence of isolate PPRI 8300 grouped basal to the P. iwayamai type strain, along with the GenBank Pythium sp. 6 sequence (GI154762338) in a cluster that has 64 % bootstrap support and a probability value of 1.00 (Fig 5). The ITS sequences of Pythium species in South Africa 943 Fig 4 – Phylogeny of Pythium species in clade F (Lévesque & De Cock 2004) based on the ITS1, 5.8S and ITS2 regions of nuclear rDNA. The tree presents one of the 1000 equally parsimonious trees of a heuristic search. Numbers within the tree represent the bootstrap values followed by probability values in brackets. Branches with a 100 % bootstrap support and a probability value of 1.00, are denoted by a star symbol. Bootstrap values lower than 60 % are not shown. Length [ 886, CI [ 0.695, RI [ 0.906, and RC [ 0.630. Numbers preceding species are the PPRI numbers of South African isolates. Sequences of species used by Lévesque & De Cock (2004) that are the ex-type, authentic strain or strains used for description in the monograph of Van der Plaats-Niterink (1981) are indicated in bold, as well as three recently described species P. spiculum (Paul et al. 2006), P. cryptoirregulare (Garzón et al. 2007) and P. viniferum (Paul et al. 2008). Sequences that may represent incorrect GenBank submissions are indicated by a black dot. 944 A. McLeod et al. Fig 5 – Phylogeny of Pythium species in clade G (Lévesque & De Cock 2004) based on the ITS1, 5.8S and ITS2 regions of nuclear rDNA. The tree presents one of the 26 equally parsimonious trees of a heuristic search. Numbers within the tree represent the bootstrap values followed by probability values in brackets. Branches with a 100 % bootstrap support and a probability value of 1.00, are denoted by a star symbol. Bootstrap values lower than 60 % are not shown. Length [ 685, CI [ 0.756, RI [ 0.785, and RC [ 0.594. Numbers preceding species are the PPRI numbers of South African isolates. Sequences of species used by Lévesque & De Cock (2004) that are the ex-type, authentic strain or strains used for description in the monograph of Van der Plaats-Niterink (1981) are indicated in bold. Sequences that may represent incorrect GenBank submissions are indicated by a black dot. Pythium species in South Africa 945 Fig 6 – Phylogeny of Pythium species in clade I (Lévesque & De Cock 2004) based on the ITS1, 5.8S and ITS2 regions of nuclear rDNA. The tree presents one of three equally parsimonious trees of a heuristic search. Numbers within the tree represent the bootstrap values followed by probability values in brackets. Branches with a 100 % bootstrap support and a probability value of 1.00, are denoted by a star symbol. Bootstrap values lower than 60 % are not shown. Length [ 685, CI [ 0.822, RI [ 0.918, and RC [ 0.755. Numbers preceding species are the PPRI numbers of South African isolates. Sequences of species used by Lévesque & De Cock (2004) that are the ex-type, authentic strain or strains used for description in the monograph of Van der Plaats-Niterink (1981) are indicated in bold. PPRI 8300 and Pythium sp. 6 only had 89 % similarity, and each of these two sequences may thus represent a new species. Therefore, isolate PPRI 8300 is designated here as Pythium sp. WJB-1. The main morphological characters for WJB-1 are: hyphal swellings globose, intercalary or terminal (11–19 mm); oogonia spherical, smooth, mainly terminal, some intercalary (av. 17 mm diam); oospores plerotic or nearly so, 1–2 per oogonium, (av. 16 mm diam), wall up to 3.0 mm thick; antheridia monoclinous or diclinous, 1–2 per oogonium, stalks unbranched, antheridial stalk origin from oogonium 7–20 mm, antheridial cells (av. 6  8 mm), attached broad apical to oogonium; sporangia and chlamydospores not observed. The origin of the isolates of which the sequences (GI133919360, Bridge & Denton 2007; GI23092305, Hughes 946 A. McLeod et al. Fig 7 – Phylogeny of Pythium species in clade J (Lévesque & De Cock 2004) based on the ITS1, 5.8S and ITS2 regions of nuclear rDNA. The tree presents one of the 30 equally parsimonious trees of a heuristic search. Numbers within the tree represent the bootstrap values followed by probability values in brackets. Branches with a 100 % bootstrap support and a probability value of 1.00, are denoted by a star symbol. Bootstrap values lower than 60 % are not shown. Length [ 965, CI [ 0.826, RI [ 0.921, and RC [ 0.760. Numbers preceding species are the PPRI numbers of South African isolates. Sequences of species used by Lévesque & De Cock (2004) that are the ex-type, authentic strain or strains used for description in the monograph of Van der Plaats-Niterink (1981) are indicated in bold. Pythium species in South Africa et al. 2003; GI157734582/3, unpublished) grouped basal to PPRI 8300 is interesting. All these isolates were isolated from the Antarctic, and together with other clade G species (P. iwayamai, Pythium paddicum and Pythium okanoganense) are collectively known as snow mold fungi, which are pathogenic on monocot grasses (Iwayama 1933; Lipps 1980a,b; Van der Plaats-Niterink 1981; Bridge et al. 2008). These species are cold-adapted mesophiles rather than true psychrophiles, since they grow well at room temperature (Bridge et al. 2008). In contrast, the apparently related PPRI 8300 was isolated in the Western Cape that has a more moderate temperature, and was isolated from Hoodia gordonii (Masson) Sweet ex Decne, a leafless spiny succulent plant indigenous to South Africa and Namibia. Clade H Two isolates (PPRI 8508 and 9007) fitted into clade H. The only members of this clade reported from South Africa are Pythium prolatum (Botha & Crous 1992) and Pythium helicandrum identified from the present study. P. helicandrum is only represented by one GenBank sequence, and has previously only been isolated from North America. The South African isolates had high sequence similarity (98.7 % and 99.1 %) to the CBS P. helicandrum sequence. The sequence identification was confirmed by morphological analyses that showed that the isolates had morphological characters that clearly distinguished it from P. prolatum. Clade I Our study included ten clade I isolates representing three species (Table 1), including Pythium heterothallicum that had not previously been found in South Africa, whereas the other two species have been reported previously (Doidge 1950). The five South African Pythium splendens isolates (PPRI 8415, 8416, 8619, 8620 and 8621) showed numerous large (av. > 32 mm) hyphal swellings with granulated contents typical for the species. Some variation in sequence data was also observed for the isolates, but there was high bootstrap and probability support (100 %, 1.00) for the monophyletic clade that included these isolates as well as the CBS sequence (GI51235509) of this species (Fig 6). The South African P. heterothallicum isolates (PPRI8419, 8616, 8617 and 8618) also showed sequence variation, but grouped in a strongly supported (100 %, 1.00) clade containing the CBS sequence (GI51235508) of this species as well as the sequence (GI30525752) of the recently described species Pythium glomeratum (Paul 2003) (Fig 6). Sequence variation within P.heterothallicum was also noted by Lévesque & De Cock (2004). Although P. heterothallicum and P. glomeratum are similar at the ITS level, they are morphological distinct. Clade J The seven South African isolates in clade J were difficult to identify to the species level, since their morphology often was not supported by their phylogenetic position. In this clade, only two Pythium perplexum isolates (PPRI 8405 and 8623) could be identified with certainty as a known species. 947 In addition to this species, Pythium acanthophoron (Darvas et al. 1978), Pythium buismaniae (Linde et al. 1994) and Pythium polymastum (Botha & Coetzer 1996) have all previously been reported in South Africa. However, these morphological identifications would need sequence data to confirm in view of the difficulty in identifying clade J species. Six of the South African isolates (PPRI 8403, 8405, 8622, 8623, 8424 and 8428) were morphologically identified as P. perplexum (Table 1). However, only PPRI 8405 and 8623 were confirmed to be P. perplexum since they grouped with high bootstrap and probability support (91 %, 0.77) with the CBS P. perplexum sequence (GI51235512), as well as with the CBS P. nodosum sequence (GI6466904) (Fig 7). Morphological characters described for P. perplexum (Kouyeas & Theohari 1977) and Pythium nodosum (Paul et al. 1998) are largely similar except for the large number of antheridial stalks enveloping the oogonia in P. nodosum and these two species may be conspecific. Isolates PPRI 8403, 8424 and 8622 all formed typical P. perplexum sexual and asexual structures according to Kouyeas & Theohari (1977) (Table 1), but their ITS sequences grouped with good bootstrap and probability support (89 %, 0.97) with the CBS Pythium nunn sequence (GI51235563) (Fig 7). The morphology of the South African isolates in the P. nunn clade did not match the original P. nunn species description of Lifshitz et al. (1984). For example the antheridial stalk and antheridial cell features were not observed in any of the South African isolates. Since the morphology of the South African isolates did match that of P. perplexum, this may suggest that the morphological P. perplexum group contains several phylogenetic lineages or possibly species. The identity of isolates PPRI8424, 8403 and 8622 therefore remains uncertain (Table 1). It is clear that the exclusive use of sexual structure morphology is not adequate for phylogenetic species classification in this group. Isolate PPRI 8428 also exhibited the morphological characteristics of P. perplexum, but the ITS sequence of this isolate did not group with the CBS P. perplexum sequence (GI51235512) (Fig 7). The sequence of PPRI 8428 grouped basal from the P. nunn clade within a monophyletic cluster, with high bootstrap support (94 %), containing two GenBank sequences GI170962955 and GI169883411 (Fig 7) and may represent a new species, provisionally named here as Pythium species WJB-2. The main morphological characters for WJB-2 are: hyphal swellings globose, ellipsoid or obpyriform, terminal or intercalary, some germinate with germ tube (av. 16 mm diam); sporangia globose, terminal (av. 18 mm diam), short discharge tubes (3–5 mm); oogonia smooth, terminal or intercalary, (av. 18 mm diam); oospores plerotic or nearly so (av. 17 mm diam), wall 2.0 mm thick; antheridia monoclinous, 1–2 per oogonium, cells inflated, elongated, attached broad apical to almost bellshaped to oogonium, stalks unbranched, uninflated, originate short distance from oogonium. Isolate PPRI 8402 was morphologically identified as Pythium polymastum or Pythium jasmonium (nom. inval.). P. jasmonium (nom. inval.) was considered as an invalid species by Dick (1990) and Van der Plaats-Niterink (1981). The sequence of isolate PPRI8402 had the highest sequence similarity to three GenBank sequences (GI18091841, GI18091842, GI164600733), which included the CBS 101.876 sequence of P. jasmonium 948 A. McLeod et al. Fig 8 – Phylogeny of Pythium species in clade K (Lévesque & De Cock 2004) based on the ITS1, 5.8S and ITS2 regions of nuclear rDNA. The tree presents one of the 670 equally parsimonious trees of a heuristic search. Numbers within the tree represent the bootstrap values followed by probability values in brackets. Branches with a 100 % bootstrap support and a probability value of 1.00, are denoted by a star symbol. Bootstrap values lower than 60 % are not shown. Length [ 1581, CI [ 0.644, RI [ 0.889, and RC [ 0.573. Numbers besides species are the PPRI numbers of South African isolates. Sequences of species used by Lévesque & De Cock (2004) that are the ex-type, authentic strain or strains used for description in the monograph of Van der Plaats-Niterink (1981) are indicated in bold, as well as the recently newly described species P. litorale (Nechwatal & Mendgen 2006), P. sterilum (Belbahri et al. 2006), P. citrinum (Paul 2004) and P. mercuriale (Belbahri et al. 2008). GenBank sequences that were possibly submitted as the incorrect species are followed by a black dot. Pythium species in South Africa (nom. inval.) (GI18091841) used by Lévesque & De Cock (2004) (Fig 7). These three sequences along with PPRI8402 formed a monophyletic clade with high bootstrap and probability support (100 %, 1.00), indicating that these isolates possibly do represent a valid species, distinct from known species within clade J, and indicates that P. jasmonium should be formally validated as a species name. Isolate PPRI 8402 showed morphological characters and dimensions more typical for P. jasmonium (nom. inval.) than P. polymastum. For example the smaller size of oogonia and oospores, oospore aplerotic to nearly plerotic, oogonial spines not mammiform but conical, blunt with broad base and oospore wall thin. Clade K In total, six clade K species have now been reported in South Africa following our analyses of 14 isolates within this clade. The species include the five species described in our study (Table 1) and Pythium mercuriale that was reported by Belbahri et al. (2008). Pythium helicoides and Pythium litorale are new records for South Africa (Table 1). In clade K, there were many isolates that only produced asexual structures in culture (Table 1). Among these, were isolates that were identified as P. litorale based on ITS sequence data. The original species description of P. litorale also only identified asexual structures (Nechwatal & Mendgen 2006). The species description of P. litorale (Nechwatal & Mendgen 2006) was published almost simultaneously with that of Pythium sterilum (Belbahri et al. 2006) and it is likely that these two species are con-specific, since they have the same morphology and also form a well supported (100 %, 1.00) monophyletic clade (Fig 8). The name P. litorale was published first and would have priority according to the International Code of Botanical Nomenclature. The asexual South African isolates (PPRI 8625 and 8624) that had the sequence identity of the normally sexual Pythium chamaehyphon (Fig 8), may be sterile or heterothallic. The two South African isolates, unlike the original species description (Sideris 1932), produced papillate sporangia with nested internal proliferation. The production of proliferating sporangia seems to be an unstable character and may not be produced under all conditions, as has been reported in Pythium vexans (Van der Plaats-Niterink 1981). In contrast, papillation of sporangia is a more stable morphological character. Lévesque & De Cock (2004) did not comment upon the morphology of the CBS P. chamaehyphon isolate used in their phylogenetic study. Our isolates require further investigation alongside the CBS P. chamaehyphon isolate to determine whether this species has nonpapillate or papillate sporangia. It can also not be ruled out that isolates PPRI 8625 and 8624 may represent a different species. The two other GenBank sequences that were submitted as P. chamaehyphon (GI 6468664 & GI 52854084) are most likely incorrect submissions, since they clustered with P. vexans (Fig 8). The South African isolates PPRI 8401, 8632 and 9086, were all considered to be P. vexans since these isolates grouped in a large clade (bootstrap support 100 % and probability value 1.00) containing the CBS sequences of P. vexans (GI51235567), Pythium cucurbitacearum (GI51235521) and Pythium indigoferae 949 (GI51235568, CBS261.30) (Fig 8). The South African isolates all produced typical P. vexans morphological features (Kouyeas & Theohari 1977). Clade K is known as a clade that potentially contains many undescribed species (Lévesque & De Cock 2004; Nechwatal et al. 2008). Our study identified one Pythium isolate, PPRI 8418, that could possibly represent a new species designated here as Pythium sp. WJB-3. There was no probability support for Pythium sp. WJB-3 being included within the P. helicoides cluster, and the isolate grouped basal to the P. helicoides cluster, along with the GenBank sequences of Pythium sp. JN-8 (GI78217382) and Pythium sp. JN-5 (GI78100295) (Fig 8). The morphology of Pythium sp. WJB-3 is distinct from that of P. helicoides. Pythium sp. JN-8 (GI78217382), also known as Pythium species VIII (Nechwatal et al. 2008), showed a 92 % sequence similarity to Pythium species WJB-3. The most characteristic features of Pythium species WJB-3 are: main hyphae are 6 mm diam with short knob-like lateral branches; oogonia are smooth, mainly terminal or intercalary, borne on short side branches (av. 29 mm); oospores are aplerotic, yellow–brown colour and thick-walled (2–4 mm); antheridia diclinous, 2–6 per oogonium, cells elongated, unbranched, lateral attached to oogonium, furrowed with slight constrictions (5– 7  17–28 mm), antheridial stalks branched with lateral appendages encircling oogonium; sporangia terminal, av. 34  80 mm, (sub-)globose, ellipsoid, obovoid, reniform, lobed or distorted shapes, usually papillate or bipapillate (subterminal or lateral), non-proliferating, occasionally perpendicular attached to sporangiophore, short discharge tubes (av. 4  11 mm), also direct germination with a germ tube; daily growth at 25  C on CMA 18 mm, min. 5  C, opt. 24  C; max. 33  C. Colony growth on CMA, white aerial fluffy mycelium with numerous scattered tufts of aerial mycelium. Conclusions Although morphological descriptions and ITS sequence data are useful for Pythium species identification, there are still several uncertainties in Pythium taxonomy using these data, which may be attributed to various factors. Firstly, some isolates have been described under different species names based on small morphological differences even though the isolates cluster in the same clade or subclade with identical ITS sequences or only a few nucleotide differences. Secondly, the current Pythium dichotomous species key of Dick (1990) needs to be supplemented and updated with inclusion of all new valid species and exclusion of invalid species. Special attention should be given to those species of which the species descriptions are incomplete, confusing and or imprecise, since precise and complete original species descriptions are crucial when conducting morphological identifications. During our study we identified Pythium iwayamai, Pythium nunn, Pythium violae, Pythium cucurbitacearum, Pythium orthogonon and Pythium kunmingense as requiring revision. Thirdly, there are quite a number of incorrect GenBank ITS Pythium species submissions, including those from published manuscripts. Therefore, it is currently best to only consider those sequences submitted by Lévesque & De Cock (2004). Lastly, 950 several Pythium species including Pythium rostratifingens, Pythium spinosum, Pythium heterothallicum, Pythium helicoides, Pythium vexans, Pythium irregulare, Pythium perplexum and Pythium splendens all contain intraspecific ITS sequence variability that should be kept in mind when species identifications are made with ITS sequence data. Species boundaries in Pythium have not been completely resolved to date, but may be clarified as more gene regions are sequenced and comprehensive multi-gene phylogenies are created (Villa et al. 2006; Martin 2000). The concept of a species in Pythium is not yet clear with regard to the morphological, biological and phylogenetic species concepts as well as possible gene flow within and between populations. The new concept of Genealogical Concordance Phylogenetic Species Recognition (GCPSR), may help to resolve species boundaries, identify events such as speciation and gene duplication, as well as determine the concordance of different gene genealogies of species (Taylor et al. 2000). A comprehensive phylogenetic network and database for Pythium, as has been established for Phytophthora (Park et al. 2008) will assist with the identification of new and current species. Such a database is in the process of being established for Pythium (www.pythiumdb.org), but it currently only contains 31 species and needs to be expanded. Acknowledgements We would like to thank the National Research Foundation (NRF) of South Africa and Stellenbosch University for funding of the work. references Alexopoulos CJ, Mims CW, Blackwell M, 1996. Introductory Mycology, fourth edn. John Wiley & Sons, Inc., New York. Ali-Shtayeh MS, Dick WM, 1985. Five species of Pythium (Peronosporomycetidae). Botanical Journal of the Linnean Society 91: 297–317. Bailey AM, Mitchell DJ, Manjunath KL, Nolasco G, Niblett CL, 2002. Identification to the species level of the plant pathogens Phytophthora and Pythium by using of the ITS1 region of ribosomal DNA as capture probes for PCR ELISA. FEMS Microbiology Letters 207: 153–158. Bakkeren G, Kronstad JW, Lévesque CA, 2000. Comparison of AFLP fingerprints and ITS sequences as phylohenetic markers in Ustilaginomycetes. Mycologia 92: 510–521. Bala K, Gautam N, Paul B, 2006. Pythium rhizo-oryzae sp. nov. isolated from paddy fields: taxonomy, ITS region of rDNA, and comparison with related species. Current Microbiology 52: 102–107. Belbahri L, Calmin G, Sánchez-Hernández E, Oszako T, Lefort F, 2006. Pythium sterilum sp. nov. isolated from Poland, Spain and France: its morphology and molecular phylogenetic position. FEMS Microbiology Letters 255: 209–214. Belbahri L, McLeod A, Paul B, Calmin G, Moralejo E, Spies CFJ, Botha WJ, Clemente A, Descals E, Sánchez-Hernández E, Lefort F, 2008. Intraspecific and within-isolate sequence variation in the ITS rRNA gene region of Pythium mercuriale sp. nov. (Pythiaceae). FEMS Microbiology Letters 284: 17–27. Botha WJ, Coetzer RLJ, 1996. Species of Pythium associated with root-rot of vegetables in South Africa. South African Journal of Botany 62: 196–203. A. McLeod et al. Botha WJ, Crous PW, 1992. A wilt disease of Rhododendron caused by Pythium prolatum and Cylindrocladium scoparium. Phytophylactica 24: 75–78. Braun H, 1925. Comparative studies of Pythium debaryanum and two related species from Geranium. Journal of Agricultural Research 30: 1043–1062. Bridge PD, Denton G, 2007. Isolation of diverse viable fungi from the larvae of the introduced chironomid Eretmoptera murphyi on Signy Island. Polar Biology 30: 935–937. Bridge PD, Newsham KK, Denton GJ, 2008. Snow mould caused by a Pythium sp.: a potential vascular plant pathogen in the maritime Antarctic. Plant Pathology doi:10.1111/j/13653059.2008.01868.x. Chamswarng C, Cook RJ, 1985. Identification and comparative pathogenicity of Pythium species from wheat roots and wheatfield souls in the Pacific Northwest. Phytopathology 75: 821–827. Cooke DEL, Duncan JM, 1997. Phylogenetic analysis of Phytophthora species based on the ITS1 and ITS2 sequences of ribosomal DNA. Mycological Research 101: 667–677. Crous PW, Phillips AJL, Baxter AP, 2000. Phytopathogenic Fungi from South Africa. Department of Plant Pathology, University of Stellenbosch, South Africa. Darvas JM, Scott DB, Kotzé JM, 1978. Fungi associated with damping-off in coniferous seedlings in South African nurseries. South African Forestry Journal 104: 15–19. De Cock AWM, Lévesque CA, 2004. New species of Pythium and Phytophthora. Studies in Mycology 50: 481–487. Denman S, Knox-Davies PS, 1992. Overview of Pythium species and diseases recorded in South Africa from 1926 to the end of 1989. Phytophylactica 24: 79–84. Dick MW, 1990. Keys to Pythium. University of Reading, School of Plant Science, Department of Botany, 64 p. Dick MW, 2001a. The Peronosporomycetes. In: McLaughlin DJ, Mclaughlin EG, Lemke PA (eds), The Mycota VII Part A. Systematics and Evolution. Springer-Verlag, Germany, pp. 39–72. Dick MW, 2001b. Straminipilous Fungi: systematics of the Pereonosporomycetes, including accounts of the marine straminipilous protists, the Plasmodiophorids and similar organisms. Kluwer Academic Publishers, Dordrecht, The Netherlands. Doidge EM, 1950. The South African fungi and lichens to the end of 1945. Bothalia 5: 1–1094. Garzón CD, Yánez JM, Moorman GW, 2007. Pythium cryptoirregulare, a new species within the P. irregulare complex. Mycologia 99: 291–301. Gonzalez AJ, Tello JC, Herrero ML, 2004. First report of Pythium tracheiphilum causing wilt and leaf blight on lettuce (Lactuca sativa) in Spain. Plant Disease 88: 1382. Gull C, Labuschagne N, Botha WJ, 2004. Pythium species associated with wilt and root rot of hydroponically grown crops in South Africa. African Plant Protection 10: 109–116. Harrison CJ, Langdale JA, 2006. A step by step guide to phylogeny reconstruction. The Plant Journal 45: 561–572. Hirane S, 1960. Studies on Pythium snow blight of wheat and barley, with special reference to the taxonomy of the pathogens. Transactions of the Mycological Society of Japan 2: 82–87. Hughes KA, Newham KK, Lawley B, 2003. Solar UV-B radiation inhibits the growth of Antarctic terrestrial fungi. Applied Environmental Microbiology 69: 1488–1491. Ingram DM, Cook RJ, 1990. Pathogenicity of four Pythium species to wheat, barley, peas and lentils. Plant Pathology 39: 110–117. Ito S, Tokunaga Y, 1935. Notae mycologiae Asiae orientalis. I. Transactions Sapporo National History Society 14: 11–33. Ito T, 1944. Some aquatic species of Phycomycetes found in Kyoto. Journal of Japanese Botany 20: 51–60. Iwayama S, 1933. On a New Snow-rot Disease of Cereal Plants Caused by Pythium sp. Pamphlet Agricultural Experimental Station, Toyama-Ken, Japan, 20 pp. Pythium species in South Africa Katoh K, Toh H, 2008. Recent developments in the MAFFT sequence alignment program. Bioinformatics 9: 286–298. Kouyeas H, Theohari I, 1977. Notes on the taxonomy of Pythium vexans de Bary and related species. Annals of the Phytopathological Institute, Benaki 11: 274–283. Klemsdal SS, Herrero ML, Wanner LA, Lund G, Hermansen A, 2008. PCR-based identification of Pythium spp. causing cavity spot in carrots and sensitive detection in soil samples. Plant Pathology 57: 877–886. Lévesque CA, De Cock AWAM, 2004. Molecular phylogeny and taxonomy of the genus Pythium. Mycological Research 108: 1363–1388. Lifshitz R, Stanghellini ME, Baker R, 1984. A new species of Pythium isolated from soil in Colorado. Mycotaxon 20: 373–379. Linde C, Kemp GHJ, Wingfield MJ, 1994. Pythium and Phytophthora species associated with eucalypts and pines in South Africa. European Journal of Forest Pathology 24: 345–356. Lipps PE, 1980a. A new species of Pythium isolated from wheat beneath snow in Washington. Mycologia 72: 1127–1133. Lipps PE, 1980b. The influence of temperature and water potential on asexual reproduction by Pythium spp. associates with snow rot of wheat. Phytopathology 70: 794–797. Martin FN, 2000. Phylogenetic relationships among some Pythium species inferred from sequence analysis of the mitochondrially encoded cytochrome oxidase II gene. Mycologia 92: 711–727. Martin FN, Loper JE, 1999. Soilborne plant diseases caused by Pythium species: ecology, epidemiology, and prospects for biological control. Critical Reviews in Plant Sciences 18: 111–181. Matsumoto C, Kageyama K, Suga H, Hyakumachi M, 1999. Phylogenetic relationships of Pythium species, based on ITS and 5.8S sequences of the ribosomal DNA. Mycoscience 40: 321–331. Matsumoto C, Kageyama K, Suga H, Hyakumachi M, 2000. Intraspecific DNA polymorphisms of Pythium irregulare. Mycological Research 104: 1333–1341. Mazzola M, Andrews PK, Reganold JP, Lévesque CA, 2002. Frequency, virulence, and metalaxyl sensitivity of Pythium spp. isolated from apples roots under conventional and organic systems. Plant Disease 86: 669–675. Meyer L, Van Dyk K, 2002. Susceptibility of pasture crops to Rhizoctonia solani and other fungi associated with crater disease of wheat on the Springbok flats, South Africa. African Plant Protection 8: 57–63. Miller P, 1955. V-8 juice agar as a general purpose medium for fungi and bacteria. Phytopathology 45: 461–462. Moorman GW, Kang S, Geiser DM, Kim SH, 2002. Identification and characterization of Pythium species associated with greenhouse floral crops in Pennsylvania. Plant Disease 86: 1227–1231. Moralejo E, Clemente A, Descals E, Belbahri L, Calmin G, Lefort F, Spies CFJ, McLeod A, 2008. Pythium recalcitrans sp. nov. revealed by multigene phylogenetic analysis. Mycologia 100: 310–319. Nechwatal J, Mendgen K, 2006. Pythium litorale sp. nov., a new species from the littoral of lake Constance, Germany. FEMS Microbiology Letters 255: 96–101. Nechwatal J, Wielgoss A, Mendgen K, 2005. Pythium phragmitis sp. nov., a new species close to P. arrhenomanes as a pathogen of common reed (Phragmites australis). Mycological Research 109: 1337–1346. Nechwatal J, Wielgoss A, Mendgen K, 2008. Diversity, host, and habitat specificity of oomycete communities in declining reed stands (Phragmites australis) of a large freshwater lake. Mycological Research 112: 689–696. Park J, Park B, Veeraraghavan N, Jung K, Lee Y-H, Blair JE, Geiser DM, Isard S, Mansfield MA, Nikolaeva E, Park S-Y, Russo J, Kim SH, Greene M, Ivors KL, Balci Y, Peiman M, Erwin DC, Coffey MD, Rossman A, Farr D, Cline E, Grunwald NJ, Luster DG, Schrandt J, Martin F, Ribeiro OK, Makalowska I, Kang S, 2008. Phytophthora database: a forensic database supporting the identification and monitoring of Phytophthora. Plant Disease 92: 966–972. 951 Paul B, 1986. A new non-zoosporic species of Pythium from Algeria. Hydrobiologia 140: 233–236. Paul B, 2003. Pythium glomeratum, a new species isolated from agricultural soil taken in north-eastern France, its ITS region and its comparison with related species. FEMS Microbiology Letters 225: 47–52. Paul B, 2004. A new species of Pythium isolated from burgundian vineyards and its antagonism towards Botrytis cinerea, the causative agent of the grey mould disease. FEMS Microbiology Letters 234: 269–274. Paul B, 2006. A new species of Pythium isolated from a vineyard in France. FEMS Microbiology Letters 263: 194–199. Paul B, Bala K, 2008. A new species of Pythium with inflated sporangia and coiled antheridia, isolated from India. FEMS Microbiology Letters 282: 251–257. Paul B, Bala K, Gognies S, Belarbi A, 2005. Morphological and molecular taxonomy of Pythium longisporangium sp. nov. isolated from the Burgundian region of France. FEMS Microbiology Letters 246: 207–212. Paul B, Bala K, Lassaad B, Calmin G, Sanchez-Hernandez E, Lefort F, 2006. A new species of Pythium with ornamented oogonia: morphology, taxonomy, internal transcribed spacer region of its ribosomal RNA, and its comparison with related species. FEMS Microbiology Letters 254: 317–323. Paul B, Galland D, Bhatnagar T, Dulieu H, 1998. A new species of Pythium isolated from the Burgundy region in France. FEMS Microbiology Letters 158: 207–213. Paul B, Mathew R, Kanak B, Paul A, Henry M, Lefort F, Belbahri L, 2008. Morphology, taxonomy and phylogenetic analysis of a new species of Pythium isolated from France. Fungal Diversity 28: 55–63. Rambaut A, 2002. Sequence Alignment Editor Version 2.0. University of Oxford, Oxford. Ronquist F, Huelsenbeck JP, 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574. Scott DB, 1987. Identification and pathogenicity of Pythium isolates from wheatfield soils in South Africa. Phytophylactica 19: 499–504. Shahzad S, Coe R, Dick MW, 1992. Biometry of oospores and oogonia of Pythium (Oomycetes): the independent taxonomic value of calculated ratios. Botanical Journal of the Linnean Society 108: 143–165. Sideris CP, 1932. Taxonomic studies in the family Pythiaceae II. Pythium. Mycologia 24: 14–61. Singleton LL, Mihail JD, Rush CM, 1992. Methods for Research on Soilborne Phytopathogenic Fungi. APS Press, St. Paul, Minnesota. Taylor JW, Jacobsen DJ, Kroken S, Kasuga T, Geiser DM, Hibbett DS, Fisher MC, 2000. Phylogenetic species recognition and species concepts in fungi. Fungal Genetics and Biology 31: 21–32. Tortolero O, Sequeira L, 1978. A vascular wilt and leaf blight disease of lettuce in Wisconson caused by a new strain of Pythium tracheiphilum. Plant Disease Reporter 62: 616–620. Van der Plaats-Niterink AJ, 1981. Monograph of the Genus Pythium. Studies in Mycology, No. 21. Centrallbureau Voor Schimmelcultures, Baarn, Netherlands. Villa NO, Kageyma K, Asano T, Suga H, 2006. Phylogenetic relationships of Pythium and Phytophthora species based on ITS rDNA, cytochrome oxidase II and beta-tubulin sequences. Mycologia 98: 410–422. Wager VA, 1941. Descriptions of South African Pythiaceae with records of their occurrence. Bothalia 4: 2–35. White TJ, Bruns T, Lee S, Taylor J, 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ (eds), PCR Protocols, a Guide to Methods and Applications. Academic Press, San Diego, pp. 315–322. Yü YN, 1973. Five new species of Pythium. Acta Microbiologica Sinica 13: 116–123.