A molecular-based assay was employed to analyse and accurately identify various root-knot nematodes (Meloidogyne spp.) parasitizing potatoes (Solanum tuberosum) in South Africa. Using the intergenic region (IGS) and the 28S D2–D3 expansion segments within the ribosomal DNA (rDNA), together with the region between the cytochrome oxidase subunit II (COII) and the 16S rRNA gene of the mtDNA, 78 composite potato tubers collected from seven major potato growing provinces were analysed and all Meloidogyne species present were identified. During this study, M. incognita, M. arenaria, M. javanica, M. hapla, M. chitwoodi and M. enterolobii were identified. The three tropical species M. javanica, M. incognita and M. arenaria were identified as the most prevalent species, occurring in almost every region sampled. Meloidogyne hapla and M. enterolobii occurred in Mpumalanga and KwaZulu-Natal, respectively, while M. chitwoodi was isolated from two growers located within the Free State. Results presented here form part of the first comprehensive surveillance study of root-knot nematodes to be carried out on potatoes in South Africa using a molecular-based approach. The three genes were able to distinguish various Meloidogyne populations from one another, providing a reliable and robust method for future use in diagnostics within the potato industry for these phytoparasites.
Potato (Solanum tuberosum) is regarded as one of the most important vegetable crops in South Africa, with an average annual production of 2 million metric tons covering over 50 000 ha. Potato production in South Africa has been significantly affected by, among other factors, diseases, pests and plant parasitic nematodes, particularly root-knot nematodes (Meloidogyne spp.).
Root-knot nematodes are highly damaging phytoparasites which can cause significant yield and crop losses in potatoes (Powers et al., 2005). This loss cuts across the tropics, subtropics and through to temperate regions where susceptibility, tolerance of the cultivar and the population load of Meloidogyne species present in the soil during planting play a pivotal role in determining the level of loss (Viaene et al., 2007). Moreover, Meloidogyne species are capable of causing deformations in potato tubers in the form of galls, as well as brown spots which are characteristic of mature females residing just below the skin layer. Infected table and processing potato tubers are rejected in local or international markets, while infected seed tubers facilitate dissemination of these pathogens to new areas (Powers et al., 2005).
Several species of root-knot nematodes have so far been identified and characterized from infected potatoes and other plant hosts, but only six are currently considered to be globally destructive. These are: Meloidogyne chitwoodi, M. fallax, M. hapla, M. arenaria, M. incognita and M. javanica (Eisenback et al., 1981). Some of the regions where root-knot nematodes have been reported to cause damage to potatoes include Belgium (Waeyenberge & Moens, 2001), Florida (Chitwood, 1949), Malta (Vovlas et al., 2005), Saudi Arabia (Al-Hazmi et al., 1993), Turkey (Devran et al., 2009) and several other parts of the world as outlined by Viaene et al. (2007).
Identification of Meloidogyne spp. based only on classical approaches, such as use of morphology and morphometrics, isozyme profiles and the North Carolina differential host race test, is to some extent inaccurate, unreliable and laborious. To overcome these challenges, a polyphasic approach that integrates DNA-based diagnostics and classical methods is on the upward trend. When identification is supported by morphological diagnosis and molecular data from gene targets in a multilocus sequence analysis (MLSA) approach, there is greater confidence in the results obtained (Tigano et al., 2005). Furthermore, DNA-based methods can be applied to various stages of nematode development, discriminate individual species from mixed populations, and also use DNA voucher specimens that have been stored for several years.
To date, various molecular approaches have been adopted to identify Meloidogyne spp., including those which are closely related (Blok & Powers, 2009). These methods include the use of target regions such as the mitochondrial DNA (mtDNA; Tigano et al., 2005), intergenic spacer region (IGS; Blok et al., 1997; Wishart et al., 2002; Adam et al., 2007), external transcribed spacer region (ETS), and internal transcribed spacer regions (ITS; Palomares-Rius et al., 2007). Others include use of sequence characterized amplified region (SCAR) markers (Zijlstra, 2000; Randig et al., 2002; Tigano et al., 2010), amplified fragment length polymorphisms (AFLP; Semblat et al., 1998), randomly amplified polymorphic DNA (RAPD; Tigano et al., 2010), restriction fragment length polymorphisms (RFLP; Carpenter et al., 1992), satellite DNA probes (Castagnone-Sereno et al., 1999) and loop-mediated isothermal amplification (LAMP; Niu et al., 2011).
This study reports the distribution and genetic diversity of various Meloidogyne species across seven potato growing provinces in South Africa. Identification and phylogenetic analysis was based on sequences of three key regions: IGS and 28S D2–D3 expansion segments within the ribosomal DNA (rDNA) and the mitochondrial DNA region located between the 3′ region of the cytochrome oxidase small subunit II (COII) and the 5′ region of the 16S rRNA gene.
Materials and methods
Biological materials and nematode extraction
During the 2011/2012 potato growing season, potato tubers infected with root-knot nematodes from seven provinces (Table 1) were collected and submitted to the University of Pretoria, South Africa for nematode identification. Samples of symptomless potato tubers and those with nematode symptoms were collected in 2 kg bags which were clearly marked to indicate the name of the cultivar, name of the grower and the geographical origin. DNA reference samples (L15, L16 and L32) which were obtained from the James Hutton Institute (TJHI), UK were also included in this study.
Table 1. Geographic origin of Meloidogyne species in this study
The James Hutton Institute (TJHI) samples were used in this study as reference isolates for the closely related tropical species.
From each sample, nematodes were isolated using the centrifugal floatation method according to van Bezooijen (2006) with some modifications. In this method, 100 g of infected composite potato tuber peel was cut into pieces less than 1 cm and transferred into a domestic blender with 100 mL of 1% (v/v) bleach added to cover the sample, topped up with distilled water to reach the 250 mL mark, before being macerated for 35 s. Next, the suspension of root-knot nematodes and potato fragments was decanted on a set of nested sieves with decreasing mesh size of 710, 150, 45 and 38 μm. The potato pieces on the 710 μm-mesh sieve were thoroughly washed with running tap water before being discarded and the suspension on the other mesh sieves washed down thoroughly, and finally the residue collected on the 38 μm-mesh sieve transferred into a beaker. To separate potato fragments from root-knot nematodes, 1 teaspoon of kaolin was added to the collected residue, stirred well and centrifuged in 50 mL Falcon tubes at 1524 g for 7 min. The supernatant was then discarded gently and sucrose solution (at 450 g L−1) added to fill the Falcon tubes before centrifuging them at 1524 g for 3 min. Finally, the supernatant was decanted into a 38 μm-mesh sieve, rinsed well with tap water to remove the sucrose solution, and 50 mL of the residue collected in a sample bottle for examination and counting of the nematodes under a stereomicroscope.
Individual second stage juveniles (J2s) were used for DNA extraction using worm lysis buffer (WLB) which consisted of 50 mm KCl, 10 mm Tris pH 8·2, 2·5 mm MgCl2, 60 μg mL−1 proteinase K (Roche), 0·45% NP-40 (Fisher Scientific), 0·45% Tween 20 (Sigma) and 0·01% gelatine (Castagnone-Sereno et al., 1995). Five individual second stage juveniles (J2s) were picked from a sample using a small needle and transferred onto a microscope glass slide containing 15 μL WLB, where they were cut into small pieces under a stereomicroscope. The cut J2 pieces, suspended in 10 μL WLB on a microscope slide, were then transferred into another 10 μL WLB in a 0·5 mL centrifuge tube, centrifuged at 19 357 g for 2 min, transferred to −80°C for 15 min, before 7 μL of mineral oil was added to each tube. Thereafter, samples were incubated at 60°C for 1 h followed by a second incubation at 90°C for 10 min. The final step in this method involved decanting the mineral oil after the DNA extract had been frozen at −20°C. All DNA samples were stored at −20°C.
Three pairs of primers were used to carry out PCR assays to specifically identify Meloidogyne species as outlined below. Primers 194 (5′-TTAACTTGCCAGATCGGACG-3′) and 195 (5′-TCTAATGAGCCGTACGC-3′) were used to amplify the IGS region of the rDNA (Blok et al., 1997). To amplify the mitochondrial DNA region located between the 3′ region of the COII and the 5′ region of the 16S rRNA gene, primers C2F3 (5′-GGTCAATGTTCAGAAATTTGTGG-3′) and 1108 (5′-TACCTTTGACCAATCACGCT-3′) were used (Powers & Harris, 1993). Finally, the D2–D3 expansion segments located within the 28S region of the rDNA were amplified using primers D2A (5′-ACAAGTACCGTGAGGGAAAGTTG-3′) and D3B (5′-TCGGAAGGAACCAGCTACTA-3′) (Schmitz et al., 1998). All primers were sourced from Inqaba Biotechnologies. Amplifications were performed in a final volume of 25 μL containing 25 ng DNA, 200 μm dNTPs (Fermentas), 0·4 μm each forward and reverse primers, 0·5 U Taq DNA polymerase (Fermentas) and 10 × Taq DNA polymerase reaction buffer with 20 mm MgCl2 (Fermentas). For IGS amplification reactions, the following temperature profiles were used: 94°C for 2 min, followed by 45 cycles of 94°C for 30 s, 50°C for 30 s and 72°C for 90 s, with a final extension at 72°C for 10 min. The COII PCR reactions were set up at 94°C for 2 min, followed by 10 cycles of 94°C for 10 s, 48°C for 30 s and 68°C for 2 min. The next 25 cycles were set up at 94°C for 10 s, 48°C for 30 s with a final extension at 72°C for 10 min. All D2–D3 PCR reactions were set up at 94°C for 2 min, followed by 35 cycles of 94°C for 30 s, 57°C for 45 s and 72°C for 3 min, with a final extension at 72°C for 10 min. All PCR amplifications were carried out using a Biometra Analytica Jena thermocycler. From the amplified PCR products, an aliquot of 5 μL was stained with GelRed (Biotium) as an intercalating agent. The mixture was then loaded on 2% (w/v) agarose gels (Lonza) in 1 × TAE (Tris–acetate EDTA) buffer before the gel was run at 100 V for 60 min and observed under UV illumination (UVP Model M-15 UV transilluminator; Vilber Lourmat). All amplification band sizes were established by comparing them with a molecular ladder.
Cloning, sequence analysis and species identification
Samples that amplified consistently using IGS, COII and 28S D2–D3 primers were cloned using a CloneJET kit (Fermentas) according to the manufacturer's instructions. For those samples whose PCR products displayed multiple bands during PCR amplifications, all fragments were excised separately from the agarose gel and purified using the Wizard SV Cleanup System (Promega) according to the manufacturer's instructions. DNA was evaluated for purity and quantified using a NanoDrop spectrophotometer (NanoDrop Technologies). In each case, three representative clones from samples of different band sizes were selected, plasmid DNA isolated using a GeneJET plasmid Miniprep kit (Fermentas) and sequenced in both directions using the same amplification primers (for the three genes). Where a single clear PCR band was obtained, direct sequencing of purified PCR products was performed with a terminator cycle sequencing ready reaction kit (BigDye; Perkin-Elmer Applied Biosystems) according to the manufacturer's instructions. Sequencing was done using the ABI 3500xl model genetic analyzer (Applied Biosystems) at the University of Pretoria, South Africa.
To identify Meloidogyne species, raw sequences obtained were checked and edited manually using BioEdit v. 7.0.9 (Hall, 1999) to correct base mismatches. Consensus sequences obtained were compared to those deposited in the GenBank database through a blast engine search for sequence homology.
For phylogenetic analysis, unique sequences obtained for IGS, 28S D2–D3 and COII gene regions in this study and those retrieved from GenBank, NCBI database (Table 2) were aligned over the same length in clustalW (with gap opening penalty for multiple alignments of 10 and extension of 0·2) and muscle (with gap opening penalty for multiple alignments of −12 and extension of −1) using mega v. 5.0 (Tamura et al., 2011). This was done to reveal regions of similarity and dissimilarity between the sequences. Highly similar sequences from IGS-rDNA were then aligned over the same length using mafft v. 5.3 (Katoh et al., 2005) fitted into the jModelTest for a suitable model (Posada & Crandall, 1998) before generating phylograms using maximum likelihood (ML) and the phylip v. 4.0 software. During this analysis, all phylograms were constructed using 1000 bootstrap replicates to assess their support for each clade or phylogenetic branching (Landa et al., 2008).
Table 2. GenBank accession numbers for reference sequences used for phylogenetic analysis
The COII and D2–D3 sequence data sets were analysed using maximum parsimony (MP; Tigano et al., 2005; Landa et al., 2008). For each data set, both the unweighted and weighted MP analyses were done using paup* v. 4.0b10 software. Heuristic searches were performed using: an addition of 100 random replicates, tree bisection–recombination branch swapping (TBR), multiple trees retained and uninformative characters excluded. Support for each clade was finally assessed by using MP analysis with 1000 replicates (Landa et al., 2008). Unique sequence data obtained from Meloidogyne populations in this study was submitted to GenBank under accession numbers JX522540–JX522545, JX987322–JX987334, KC287187–KC287213 and KC295536–KC295537.
Distribution of Meloidogyne species
Of the 78 composite potato tuber samples collected in this study, 81% were found to be infected with various Meloidogyne species. These species were identified from nine different cultivars: Mondial, Sifra, Van der Plank, Up-to-date, Bufflespoort, Argos, Valor, Fiana and BP1, that are mainly grown in South Africa. Mondial showed the highest incidence (61%) followed by Up-to-date and Sifra cultivars with 11 and 4%, respectively. Valor and Bufflespoort constituted 3% each of infected tuber samples while Argos, Fiana, BP1 and Van der Plank were each found to average 1% of infected samples. Infected root-knot nematode samples were obtained from most of the major potato growing regions in the country with the exception of the Eastern and Western Cape, where samples were not able to be obtained (Table 1). For identification, second stage juveniles (J2) were extracted from potato peel, DNA extracted and subjected to PCR amplification using primers specific for IGS, D2–D3 and COII. PCR products of South African populations were compared to reference samples from TJHI, UK for identification (results not included). Based on PCR amplicon product size and blast algorithm analyses of IGS, D2–D3 and COII sequences, the majority of the potato tubers collected were found to be infected with the tropical Meloidogyne species M. incognita, M. javanica and M. arenaria. Meloidogyne javanica and M. incognita were the most prevalent, with 24 and 23% occurrence, respectively, recorded in all the regions sampled (Fig. 1). The other samples were identified as M. arenaria (17%), M. enterolobii (14%), M. chitwoodi (3%) and M. hapla (1%) (Fig. 2). The identity of the remaining 19% of samples could not be established. It is possible that these were plant-parasitic nematodes from other nematode genera because they did not amplify with the root-knot nematode specific primers (194/195). Meloidogyne enterolobii was identified in samples collected from the KwaZulu-Natal region, whereas M. chitwoodi was recorded mainly in the Free State (Fig. 1).
Species identification and phylogenetic analysis based on IGS, D2–D3 and COII
The unique sequences obtained in this study were used to construct phylogenetic relationships as shown in Figures 3–5. Using the IGS-rDNA, 16 new sequences were generated from different Meloidogyne species isolated in this study. These sequences were aligned together with seven IGS sequences for various Meloidogyne species retrieved from GenBank resulting in a total of 23 sequences. Alignment and phylogenetic analysis of these sequences resulted in several clades which were separated with varying bootstrap support (BS) values in the ML analysis as follows (Fig. 3): (i) M. javanica, M. incognita and M. arenaria populations (BS = 100%); (ii) all M. enterolobii populations (BS = 90%); (iii) M. hapla populations (BS = 100%); and (iv) M. chitwoodi populations (BS = 100%). In this analysis, the tropical Meloidogyne species were distinguished by a bootstrap support of 83% from the automictic species.
The 28S D2–D3 alignment was made up of 28 sequences, 16 of which were new sequences from the study populations. Maximum parsimony analysis of the 28S D2–D3 sequences demonstrated varying bootstrap support values for various clades and subclades (Fig. 4): (i) M. chitwoodi populations (BS = 76%); (ii) M. chitwoodi populations and M. fallax (BS = 91%); (iii) M. enterolobii populations (BS = 99%); (iv) M. incognita populations (BS = 98%); and (v) M. hapla populations (BS = 94%), M. javanica populations (BS = 100%) and M. arenaria populations (BS = 100%).
For the COII sequences of the 16S-rRNA, 33 sequences were aligned, 16 of which were from the populations in this study. All the study populations and those deposited in the GenBank database formed clades and subclades of varying bootstrap support values during maximum parsimony analysis as follows (Fig. 5): (i) M. graminicola, M. minor and M. naasi (BS = 77%); (ii) M. chitwoodi populations and M. fallax (BS = 98%); (iii) M. hapla populations and M. partityla (BS = 88%); (iv) automictic Meloidogyne species (M. chitwoodi, M. fallax, M. minor, M. naasi, M. marylandi, M. partityla and M. hapla) formed a major clade with 83% bootstrap support; (v) M. enterolobii populations (BS = 92%); (vi) M. javanica populations (BS = 92%); (vii) M. arenaria populations and M. morocciensis (BS = 94%); (viii) M. incognita populations (BS = 100%); and (ix) all tropical Meloidogyne species (including M. enterolobii) formed a major clade with 86% bootstrap support.
Accurate identification and in-depth understanding of the population and genetic diversity of Meloidogyne species present in a given potato field is the first step in designing proper pest management programmes (Powers et al., 2005). This can only be achieved through a regular, comprehensive and accurate survey of Meloidogyne species across all the potato growing regions in South Africa. To this end, this study reports the identity of various Meloidogyne species collected from different potato growing zones across South Africa. The rDNA was chosen as the most appropriate target site because it has relatively conserved and highly variable regions which have been used in identification and in construction of phylogenetic relationships for root-knot nematodes and other nematode species (Landa et al., 2008). The mtDNA is also a well-conserved target, with a relatively fast rate of sequence polymorphism and rearrangements compared to the nuclear genome (Blouin, 2002).
In this study, a total of 78 composite potato tubers were collected from several potato growing regions within seven provinces of South Africa, of which 81% were found to be positively infected with root-knot nematodes. The higher incidence in cv. Mondial is probably due to the fact that it is the most popular commercially grown potato cultivar in South Africa (representing 61% of cultivars grown) and to the authors' knowledge it has not been reported as a resistant cultivar to root-knot nematodes. Various Meloidogyne species were identified and the information used to map the distribution of the different root-knot nematodes in potato growing fields in South Africa. This is the first comprehensive study to try and screen all Meloidogyne species infecting potatoes in South Africa at a molecular level.
Study findings indicated that most farms are dominated by the three common tropical species: M. incognita, M. javanica and M. arenaria. This is in line with common knowledge that the three tropical species are the most prevalent in vegetable production in South Africa (M. Marais, Agricultural Research Council, Pretoria, South Africa, personal communication). The highly damaging and resistance-breaking M. enterolobii was reported in 14% of potato tubers sampled. Most of the samples affected by this nematode were obtained from the KwaZulu-Natal potato growing region. It is not clear if M. enterolobii is distributed in other potato growing areas or if it is only restricted to this region. However, of concern is the fact that M. enterolobii was identified in seed producing farms, thus this is likely to have considerable consequences for seed and potato production in general.
The automictic Meloidogyne species were identified from a few areas where potatoes are grown in South Africa. The Free State was the only region from where M. chitwoodi was reported in this study. The Free State is generally cooler and thus perhaps this is indicative of the cool adaption of M. chitwoodi. Meloidogyne hapla was identified in samples from the Mpumalanga potato growing area. Mpumalanga is warmer and more tropical and this compares with other studies that have isolated M. hapla from tropical areas (Carneiro et al., 2000).
Molecular-based methods such as sequencing and phylogenetic analysis have been employed in some studies to resolve the identity of various Meloidogyne species (Landa et al., 2008; McClure et al., 2012). The IGS-rDNA region has been used in other studies as a diagnostic target to differentiate various Meloidogyne species (Blok et al., 1997; Wishart et al., 2002). As demonstrated by studies carried out by Holterman et al. (2012), analysis based on the IGS-rDNA sequences was capable of detecting and discriminating between individual Meloidogyne species. This study was also able to identify various Meloidogyne species based on IGS-rDNA and also on the 28S D2–D3 expansion segments and the COII. The IGS sequences from most of the Meloidogyne species identified were highly similar to the reference sequences deposited in GenBank. Based on sequence analysis of the tree gene regions, it was also observed that the individual South African Meloidogyne populations were, rather unexpectedly, highly homogeneous.
During the study, an attempt was made to construct phylogenetic relationships based on consensus sequences derived from IGS-rDNA of various Meloidogyne species. Phylogenetic analysis of consensus sequences derived from IGS-rDNA using ML analysis was able to group the tropical adapted species into one clade which is completely distinct from the clade formed by the temperate species. Closely related apomictic species (M. javanica, M. arenaria and M. incognita) grouped together into one clade with 100% bootstrap support while automictic species (M. chitwoodi and M. fallax) also grouped together into another clade with 100% bootstrap support. The close relationship between M. fallax and M. chitwoodi was demonstrated by the high homology between the reference sequences of M. fallax and M. chitwoodi. Populations of M. enterolobii (which is a tropical species) formed an independent clade which is slightly closer to the tropical species than the temperate one. This was supported by a 94% bootstrap support using ML analysis. Meloidogyne hapla, which is a slightly facultative parthenogenetic species, was also clearly separated during phylogenetic analysis to form an independent clade in between the apomictic and automictic species, but closer to the automictic species. Other studies carried out previously also suggest that M. hapla is more closely related to the automictic species than to apomictic species based on percentage nucleotide base substitution using total genomic DNA (Castagnone-Sereno et al., 1993). The phylogenetic tree was constructed without an out-group because the primer pair (194/195) used in this study is only specific to the Meloidogyne genus. Furthermore, ML analysis was adopted for IGS sequence data because it gave the best tree topology compared to MP analysis.
The 28S D2–D3 expansion sequences in this study gave similar results to those obtained using the IGS region of the rDNA. This is despite the fact that D2–D3 primers used in this study can amplify the 28S region of rDNA associated with Meloidogyne species and that of other plant-parasitic nematode genera, for example Xiphinema (Gutierrez-Gutierrez et al., 2010). Amplifications of the 28S D2–D3 expansion segments was therefore used to indicate the presence of members from the phylum Nematoda and at the same time corroborate the results obtained by IGS-rDNA and COII sequences. Various Meloidogyne species identified in this study grouped correctly with those already identified in various parts of the world based on phylogenetic analysis of the 28S D2–D3 sequences. The 28S D2–D3 phylogenetic analysis did not resolve automictic and apomictic Meloidogyne species into separate and distinct clades, despite the fact that individual sequences were able to identify the species. This has also been witnessed in other studies which have used this gene target for phylogenetic analysis of various Meloidogyne species (Landa et al., 2008).
Use of COII in this study confirmed the diversity of Meloidogyne species which were also identified based on IGS and 28S D2–D3 sequences. The COII is a highly conserved region and it was therefore easy to discriminate different Meloidogyne species, as also demonstrated in other studies (Blouin, 2002; Tigano et al., 2005). Further phylogenetic analysis based on COII showed that all tropical Meloidogyne species grouped together to form a single clade with a bootstrap support of 86%, while the temperate apomictic species also grouped together to form a single clade with 93% bootstrap support. Meloidogyne enterolobii is more closely related to the tropical species than it is to M. hapla, M. fallax and M. chitwoodi. This is why the species grouped with other tropical Meloidogyne species, and this was supported by 86% bootstrap support for the clade containing M. enterolobii and the tropical species. This grouping was consistent with other studies that have been carried out elsewhere (McClure et al., 2012). The close relationship of M. enterolobii can be attributed to the mode of reproduction, because both M. enterolobii and most of the tropical Meloidogyne species are mitotically parthenogenetic (Tigano et al., 2005). The unique and high sequence homology for the three gene sequences (IGS, 28S D2–D3 and COII) for M. enterolobii easily differentiated M. enterolobii from other Meloidogyne species. The sequence analysis based on COII sequences in this study together with reference isolates in GenBank showed that they were highly rich in AT nucleotides (results not shown). McClure et al. (2012) observed the same while analysing COII sequences of various Meloidogyne species from 238 golf courses in the western United States.
In conclusion, this is the first time sequence data for various Meloidogyne species in potatoes from South Africa has been reported after carrying out a molecular-based survey. This study reports the presence of both apomictic and automictic Meloidogyne species in potato growing regions in South Africa. The information presented here can form a basis for formulating alternative methods of control. Findings presented here can also be used by seed producers who are integral in ensuring that various Meloidogyne species are not disseminated from one region to another. The high reproduction rate and capacity of M. enterolobii to break the Mi-1 resistance gene can significantly affect potato production.
This work was funded by the National Research Foundation (NRF) –Technology and Human Resources Programme (THRIP) and Potatoes South Africa (PSA). EMO received an NRF (Thuthuka) – Grant Holder Linked Bursary. The authors also thank Professor Piet Hammes (University of Pretoria), Dr Mariette Marais (ARC Biosystematics, South Africa) and Dr V. C. Blok (TJHI, UK) for technical assistance.