Introduction

Plant-parasitic nematodes of the genus Meloidogyne are highly damaging pathogens, which are associated with low yield and quality losses in many crops worldwide, including potatoes (Solanum tuberosum). Damage caused by these phytoparasites is not only restricted to the tropical but also in sub-tropical and temperate regions (Wesemael et al. 2011).

Tropical species such as M. javanica, M. arenaria and M. incognita are the most dominant species affecting most of the crops in South Africa (De Waele and Elsen 2007). Due to use of morphological characters for many years, the three common tropical species; M. incognita, M. javanica and M. arenaria are well studied and better understood compared to other Meloidogyne spp. However, newly emerging Meloidogyne spp., potentially more damaging than the common tropical species, pose a new threat to crop production in these regions. Early and accurate identification of root-knot nematode species infesting crops is crucial for designing effective integrated pest management programmes. However, identification based on morphological features is time consuming and requires highly skilled personnel (Blok et al. 2002). Furthermore, perineal patterns of some species are highly similar making accurate identification based on morphology and morphometric traits challenging, even to an expert. This particularly applies to emerging or neglected Meloidogyne species such as M. enterolobii, M. hispanica and M. ethiopica, whose perineal patterns are often similar to those of the common tropical species, and thus accurate identification may be difficult or they can often be misidentified (Brito et al. 2004; Landa et al. 2008; Conceição et al. 2012). This is also compounded by taxonomic experience, which is often biased toward common tropical Meloidogyne species (Conceição et al. 2012). In this respect, the use of molecular techniques as outlined by Blok (2005) has become a useful method for distinguishing such species.

Of the emerging species, M. enterolobii is regarded as the most aggressive in comparison to other tropical root-knot nematode species (Brito et al. 2004). This is primarily due to its ability to overcome resistance genes, such as the Mi-1 gene in tomato (Kiewnick et al. 2009). The resistance breaking ability of this nematode species is an important factor that gives the nematode the ability to reproduce well and cause more galling than any other tropical species even in crops with root-knot nematode resistance (Cetintas et al. 2007). Currently, this nematode species is on the EPPO alert list (OEPP/EPPO bulletin 2011). In South Africa, it was first reported in Mpumalanga, in 1997 where it led to the decline and eventual death of infected but untreated guava trees (Willers 1997). This nematode species has also been identified in various parts of the world, such as France, the USA (Florida), two greenhouses in Switzerland, Brazil and China (Blok et al. 2002; Brito et al. 2004; Kiewnick et al. 2009; Tigano et al. 2010; Hu et al. 2011).

Potato production in South Africa is spread across 16 regions in upwards of 50 000 ha. Root-knot nematode damage is an important factor contributing to yield losses, tuber rejection and revenue loss. However, to our knowledge, there has been no comprehensive survey of root-knot nematode species infesting potatoes in South Africa. During the growing season of 2011/2012, 78 composite samples of root-knot nematode infected potato tubers (Solanum tuberosum) from different potato growing regions across South Africa were submitted to the University of Pretoria. From each sample, nematodes were isolated using the blender centrifugal flotation method (Bezooijen 2006). Five individual second stage juveniles (J2) per sample were picked and used for DNA extractions (Castagnone-Sereno et al. 1995).

Primers 194 (5′-TTAACTTGCCAGATCGGACG-3′) and 195 (5′-TCTAATGAGCCGTACGC-3′) were used to amplify the IGS region of the ribosomal DNA (rDNA) (Blok et al. 1997) while C2F3 (5′-GGTCAATGTTCAGAAATTTGTGG-3′) and 1108 (5′-TACCTTTGACCAATCACGCT-3′) were used to amplify the mitochondrial DNA region located between the 3′ region of the cytochrome oxidase small subunit II (COII) and the 5′ end region of the 16S rRNA gene (Powers and Harris 1993). Out of the 78 composite samples tested in this study, five samples produced an amplicon of 705 bp and 780 bp for COII and IGS, respectively. Both the IGS and COII amplification products obtained in this study agree with the expected size for M. enterolobii as previously reported (Brito et al. 2004; Tigano et al. 2005). The primers, 63VNL and 63VTH, targeting a 63-bp tandem repeat region of the mitochondrial genome produced a 322 bp fragment typical of M. enterolobii (Data not shown). Significantly, this fragment was absent from all other samples tested, further confirming the identity of these populations as M. enterolobii (Lunt et al. 1998; Brito et al. 2004).

Sequence and phylogenetic analyses has gained much popularity not only for identification but also for revealing genetic diversity of different Meloidogyne populations (McClure et al. 2012). Thus, we sought to compare the sequences of South African M. enterolobii with each other as well as to those obtained from GenBank for intra and interspecies variation, respectively. All PCR products obtained using COII and IGS primers were purified and cloned into CloneJET™ (Fermentas, Life Sciences). Three representative clones from each of the five samples were selected for bi-directional sequencing using the ABI3500xl model genetic analyzer (Applied Biosystems) at the University of Pretoria, South Africa. Consensus sequences obtained were compared for homology with those deposited in GenBank through BLAST search engine. The sequences of South African M. enterolobii were deposited in GenBank under accession numbers from JX522540 to JX522545.

For IGS sequence dataset, highly similar sequences were aligned over the same lengths using MAFFT 5.3 (Katoh et al. 2005), fitted into the jModel test for a suitable model (Posada and Crandall 1998) before generating phylograms using Maximum likelihood (ML) and the Phylip 4.0 software. All phylograms were constructed using 1,000 bootstrap replicates to assess their support for each cluster or phylogenetic branching (Landa et al. 2008). The phylogenetic analysis of the COII sequence data set was performed using maximum parsimony (MP) (Tigano et al. 2005). For each data set, both the un-weighted and weighted MP analyses were performed using PAUP* 4.0b 10 software and support for each cluster assessed by using MP analysis with 1,000 replicates.

In this study, sequence analyses based on IGS and COII indicated that the sequences of South African M. enterolobii populations shared a 100 % sequence homology. The lack of variation could be due to the fact that all five M. enterolobii populations were from the same potato growing region. However, Tigano et al. (2010) observed similar homogeneity displayed by M. enterolobii populations from different geographic regions in Brazil, thus, it is likely that the lack of variation within the South African population is indicative of the homogeneous nature of M. enterolobii observed in other populations. No significant difference between the sequences of the South African population and those deposited in GenBank for M. enterolobii /mayaguensis (FJ555695.1 and AY635613.1) was observed confirming the lack of diversity between M. enterolobii populations from different geographic regions. Phylogenetic analysis using ML for the IGS sequences of these samples showed that the South African M. enterolobii populations and those from GenBank formed an independent cluster with a high bootstrap support value of 90 % (Fig. 1). This cluster, containing our populations and M. enterolobii from GenBank, was distinct from others and was sandwiched between tropical species and that of temperate adapted Meloidogyne species clusters. However, the cluster was slightly closer to that of the tropical Meloidogyne species than to the temperate adapted one. The two adjacent clusters, one consisting of mostly the tropical species (M. arenaria, M. incognita and M. javanica) and the other consisting of mainly temperate Meloidogyne species were well supported with bootstrap support values of 99 % and 100 %, respectively. Meloidogyne hapla, which is slightly a facultative parthenogenetic species, was also clearly separated during phylogenetic analysis and positioned in between the apomictic and automictic species but closer to the automictic species. This was in agreement with studies carried out previously, which 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).

Fig. 1
figure 1

Maximum likelihood (ML) analysis of the IGS-rDNA sequences of Meloidogyne populations in this study with other reference sequences. All populations in this study have designations beginning with Melo. Analysis was done using 1,000 bootstrap replicates. The bootstrap support value for each cluster is indicated on the nodes

Full size image

Using the mtDNA, sequences of the South African Meloidogyne populations were again shown to be identical to one another as well as to M. enterolobii sequences deposited in GenBank (Fig. 2). This cluster had a high bootstrap support value of 100 %. Although the topologies of the two trees were different (Figs. 1 and 2), both consistently showed that the South African and GenBank M. enterolobii populations clustered together with a high bootstrap support value. Furthermore, this cluster appeared more closely related to the tropical Meloidogyne species than to M. hapla, M. fallax and M. chitwoodi as previously observed with IGS generated trees. This was evidenced by the high bootstrap support (86 %) for M. enterolobii and the tropical species clusters. Using mtDNA sequences, McClure et al. (2012) were also able to group M. enterolobii closer to the tropical Meloidogyne species. The close relationship of M. enterolobii can be attributed to the mode of reproduction, since both M. enterolobii and most of the tropical Meloidogyne species are mitotically parthenogenetic (Tigano et al. 2005).

Fig. 2
figure 2

Maximum parsimony tree that has been rooted after an alignment of mtDNA sequences of Meloidogyne populations in this study. All populations in this study have designations beginning with Melo. Bootstrap support for each clade is indicated at the nodes. Bursaphelenchus xylophilus was used as an out-group

Full size image

In conclusion, M. enterolobii was identified in five samples of root-knot nematode infested potato tubers originating from the KwaZulu-Natal potato growing region in South Africa. None of the samples tested from the other regions were positive for M. enterolobii. Sequences of the South African M. enterolobii were highly similar to one another and to those obtained from the GenBank. To our knowledge, there are no current data available for M. enterolobii in potatoes within South Africa. Although first reported in guava in 1997, to date, there has been no data investigating genetic diversity of M. enterolobii in South Africa or indeed how the South African population compares to others from different parts of the world. This report contends that the presence of M. enterolobii in potato growing regions in South Africa is a potential threat to potato production and alternative methods of control will have to be investigated. The high reproduction rate and capacity of this nematode species to overcome root-knot nematode resistance genes and potential transmission through seed potato tubers can have a significant impact on potato production. Given the fact that this nematode has morphological characters such as the perineal patterns, which are highly similar to those of M. incognita, it is imperative, as with other emerging phytoparasites, to employ alternative and robust methods to accurately identify this highly virulent pathogen (Brito et al. 2004; Conceição et al. 2012).