Entomopathogenic nematodes (EPNs) in the families Steinernematidae and Heterorhabditidae are of great interest because of their potential as commercial biological control agents against insects. Entomopathogenic nematodes are lethal, obligate parasites of soil-dwelling stages of insect pests (Burnell & Stock 2000). They occur naturally in all soils worldwide, with their control of the soil stages of insects acting as an ecosystem service for keeping insect populations at bay. Entomopathogenic nematodes have been successfully used worldwide as biological control agents to suppress over 200 economically important insect pests (Grewal, Ehlers & Shapiro-Ilan 2006; Kaya & Gaugler 1993; Shapiro-Ilan, Gouge & Koppenhofer 2002).

Entomopathogenic nematode surveys during the last decades yielded an increasing number of new species. Entomopathogenic nematodes are found globally, except for Antarctica, where they have yet to be found (Griffin, Downes & Block 1990; Hominick 2002). The study of natural EPN occurrence and distribution, as characterised by different ecological habitats in various geographical areas, provides baseline information for incorporating the pathogens in biological control programmes, and for understanding the complexities of their biodiversity. The factors such as geographical location, climatic conditions, soil properties and habitat type may determine the occurrence and distribution of the EPNs. To be effective as biological control agents, EPNs need to be adapted to the local environmental conditions of the site of application (Bedding 1990). However, with inundative application, with an immediate effect on the target organism, temperature and humidity are the most important environmental factors to take into consideration.

In South Africa, five major surveys have been conducted to isolate EPNs for use as biological control agents (Hatting & Malan 2017; Malan & Hatting 2015). A survey was conducted between 2004 and 2005, during which 498 soil samples were collected, with only 36 representing 7% of the total number of samples testing positive for EPNs (Malan, Nguyen & Addison 2006). This survey, aimed at obtaining EPNs for use as biological control agents against codling moth (Cydia pomonella Linnaeus), were restricted to the southern part of the Western Cape province. Another survey, aimed at determining the distribution and diversity of EPNs, was conducted between 2003 and 2005, during which 1508 soil samples were collected. This survey, in which only 79 (5%) of the total samples tested positive for EPNs (Hatting, Stock & Hazir 2009), was conducted in the Western Cape, Free State, Gauteng, Mpumalanga and KwaZulu-Natal provinces. The third survey, aimed at determining the potential of EPNs for the control of the soil stages of a false codling moth, Thaumatotibia leucotreta (Meyrick), consisted of 202 soil samples, of which 35 (17%) were found to test positive for the presence of EPNs (Malan, Knoetze & Moore 2011). Three provinces were targeted in the survey, namely the Western Cape, the Eastern Cape and Mpumalanga provinces. Recent surveys were conducted by Steyn et al. (2017b) from avocado, litchi and macadamia orchards in the subtropical regions of Mpumalanga, Limpopo and KwaZulu-Natal provinces, while Abate et al. (2018) surveyed Eucalyptus spp., Pinus spp. and Acacia mearnsii De Wild plantations in KwaZulu-Natal and Mpumalanga. Both surveys reported new Steinernema species, Steinernema fabii Abate, Malan, Tiedt, Wingfield, Slippers & Hurley (Abate et al. 2016) and Steinernema litchi Steyn, Knoetze, Tiedt and Malan (Steyn et al. 2017a), as well as new reports for South Africa of Heterorhabditis taysearae Shamseldean, El-Sooud, Abd-Elgawad & Saleh (Steyn et al. 2017b) and Heterorhabditis baujardi Phan, Subbotin, Nguyen and Moens (Abate et al. 2018; Steyn et al. 2017b). The above surveys have contributed to the existing knowledge of the geographical distribution and diversity of EPNs.

The natural vegetation type in the Western Cape consists of fynbos, with such vegetation consisting primarily of natural shrubland, which occurs in a small belt of the Western Cape, which has a Mediterranean climate, characterised by winter rainfall (Esler, Pierce & De Villiers 2014). The fynbos biogeography is known for its exceptional degree of biodiversity and endemism. The diversity of fynbos plants is extremely high, with over 9000 species of plants occurring in the area, of which about 6200 are endemic. The extremely high level of diversity is comparable to that of tropical rainforests or large islands, and it is unique to the relatively dry African continent (Esler et al. 2014). It is hypothesised that the diversity of EPNs in the fynbos will be as diverse, as different vegetation types it is associated with.

The objective of the current study was to determine the distribution and composition of indigenous EPNs in the natural vegetation of the Western Cape province, in terms of the fynbos found in the JS Marais Nature Reserve in the town of Stellenbosch. The information obtained should add to the knowledge of the occurrence and distribution of EPNs in mostly undisturbed soils in South Africa.

More than half (39) of the samples collected the nature reserve contained EPNs. Two endemic species of Steinernema (29 isolates) were isolated and identified with molecular techniques viz. Steinernema khoisanae (24 isolates) and Steinernema nguyeni Malan, Knoetze and Tiedt (five isolates). Heterorhabditis safricana Malan, Nguyen, De Waal and Tiedt (Malan et al. 2008), the only Heterorhabditis found, was isolated from two samples. An unknown Oscheius sp. was also isolated from two soil samples (Table 1).

Sequences of the ITS region of S. khoisanae generated with the two primer pairs, TW81 (F) and AB28 (R) and 18S (F) and 26S (R), did not produce good-quality sequences. Only partial sequences of either the ITS1 with the forward primer, or ITS2 with the reverse primer, could be submitted to GenBank. Only in two cases was a full ITS (ITS1-2.8S-ITS2) region for S. khoisanae was obtained. However, sequencing of the D2–D3 regions of the 28S gene large subunit produced good-quality sequencing > 880 base pairs and confirmed their identity (100%) with the isolates of populations of S. khoisanae (Table 1) reported from other areas in the Western Cape province.

All species isolated were trapped using mealworm, except at two samples, where S. khoisanae was trapped with wax moth. In 11 samples, both the mealworm and the codling moth trapped both S. khoisanae and S. nguyeni. Entomopathogenic nematodes were trapped in 14 samples using wax moth. In two samples, EPNs were trapped with all three trapping hosts, whereas, in 10 samples, both wax moth and mealworm trapped EPNs (Figure 1). The Oscheius sp. was trapped using mealworm (Table 1). The soil analysis showed that S. khoisanae were found in all the soil types throughout the nature reserve (Table 2).

Of all the isolates, the body length of 10 IJs was measured to confirm the molecular identification. The body length of S. khoisanae specimens from the JS Marais Nature Reserve was found to be slightly shorter than those of the type specimens 1010 (916–1127) µm versus 1062 (904–1159) µm. In the case of S. nguyeni, the body length of the IJ correlated with that of the type specimen. For H. safricana, the IJ body length of the type specimen 697 (656–725) µm was found to be slightly longer, versus 600 (550–676) µm of the specimens found in this study. The sequences derived from the Oscheius sp. could not be matched to any known sequences in GenBank (Table 3); thus, possibly, it is a new, undescribed species of the genus.

The occurrence and diversity of the EPNs found in the JS Marais Nature Reserve were high (51%), with four species being identified, namely S. khoisanae, S. nguyeni, H. safricana and an Oscheius sp. The Steinernema spp. were more abundant than the others, which is found in 94% of the positive samples. Steinernema khoisanae was the most common species, which is found in 86% of the positive samples, with it being distributed all over the nature reserve. Of the positive soil samples, three contained S. nguyeni, three H. safricana and two with an unknown Oscheius sp. All four species are endemic to South Africa. The trend of more steinernematids being recovered than heterorhabditis as stated by Hominick (2002) was confirmed in the current survey.

In previous surveys undertaken to test for the presence of EPNs, the recovery rate was, in general, very low. A non-targeted survey by Malan et al. (2006) showed a 7% recovery rate, while, in a targeted citrus orchard survey, 17% recovery was obtained for a total of 119 sampling sites. Hatting et al. (2009) undertook a comprehensive survey (of 1500 samples), with a very low recovery rate of 5%. Most of these samples were taken from agricultural soil, which could be the reason for the low recovery rate obtained. In the current survey, less than four samples were collected per hectare, which is not regarded as intensive, as in orchards, 1 ha is usually divided into four quadrants, from which five subsamples are combined to give one sample, with a total of four samples per hectare (Steyn, Malan & Addison 2020). From the information obtained, it can be extrapolated that the occurrence of EPNs in the JS Marais Nature Reserve was exceptionally high.

Three susceptible hosts were chosen for the isolation of EPNs because some species or strains could prove to be host-specific (Malan et al. 2011). During the present study, the most effective trapping host was T. molitor, from which all positive samples were trapped, except for two species of S. khoisanae, which were isolated with wax moth larvae. Unfortunately, the natural insect host of none of the EPN species isolated are known (Table 1). Some EPN species were found to be highly specific, like Steinernema scapterisci Nguyen and Smart, which appears to reproduce only in adult mole crickets (Scapteriscus spp.) (Nguyen & Smart 1990). This indicates that many more EPNs might have been present in the soil but were not trapped by the three hosts used. Steinernema bertusi Katumanyane, Malan, Tiedt, Hurley, a newly described species for South Africa, was successfully isolated using wax moth and mealworm on two separate occasions and at two separate locations, from the KwaZulu-Natal and Mpumalanga provinces (Katumanyane et al. 2020).

The high abundance of S. khoisanae in the Western Cape province could suggest the presence of a wide range of suitable endemic hosts and favourable conditions for the species, which might contribute to its high observed frequency (Hatting et al. 2009; Malan et al. 2006, 2011). From the soil analysis (Table 2), S. khoisanae was noted to be present in all the soil types (LmSa, SaLm, Lm) of the nature reserve.

Intra-individual variability of the ITS region of most of the S. khoisanae individuals influenced the ability of standard primer pairs to produce high-quality sequences for this region that could successfully be aligned to identify the species. However, in this study, only two sequences were obtained by means of aligning the forward and the reverse sequences, both using the 28S and 26S primer pairs (Table 1). Using the D2D3 region of the 28S gene region, high-quality sequences were obtained with both the forward and the reverse primers, thus confirming their identity as S. khoisanae. The ITS region, which is among the most important molecular markers used in identification, taxonomy and phylogeny of EPNs, was used with for the species identified from South Africa, as well as from other parts of the world (Malan & Hatting 2015). Půža et al. (2015) found that, in Steinernema feltiae (Filipjev) Wouts, Mráček, Gerdin & Bedding, intra-individual variability was found in 65% of specimens, and in almost 90% of the glaseri group, sampled from different localities. However, in the JS Marais Nature Reserve, 93% of the individuals in the population showed intra-individual variation early in the process of sequencing the ITS region, with the generation of short sequences at the 3’ and 5’ ends, which could not be aligned. Půža et al. (2015) also suggested the importance of using the D2D3 region to confirm the new species status of an EPN isolate. In this case, the D2D3 was used to confirm the species identification of most of the populations found in the same area, which is very rare occurrence.

De Waal et al. (2010) tested the virulence of six EPN species against diapausing codling moth, an insect of agricultural importance. The results showed the high susceptibility of codling moth for S. khoisanae, causing 97% and 88% infection, respectively. During a field trail against the false codling moth in citrus, S. khoisanae did not perform as well as did the other species tested. As the IJ of S. khoisanae is relatively large (≈1 cm), it generally cannot be used as a biocontrol against such micro insects as Bradysia impatiens and Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae) (Dlamini, Allsopp & Malan 2019; Katumanyane, Ferreira & Malan 2018). This specific population of S. khoisanae from the JS Marais Nature Reserve with its different ITS region should be tested against pest insects to compare it to other S. khoisanae populations with a normal ITS profile.

Steinernema nguyeni was first isolated from soil collected from beneath Olea europaea subsp. africana L. (Miller), an indigenous tree, near Clanwilliam in the Western Cape. The nematode was trapped using both G. mellonella and T. molitor. Similarly, it was reported from a survey conducted by De Waal, Malan and Addison (2011), as being isolated from undisturbed soil in the Knysna area of the Western Cape, but not in any other surveys.

Heterorhabditis safricana (SF281), which was described from a peach orchard in the Western Cape, has been isolated by trapping, using Galleria as host. It seems to be a rare species in South Africa, as previous surveys have detected its presence only in a peach orchard (Malan et al. 2008). Just two isolates were recovered during the current survey. Some controversy surrounds the status of the Oscheius (= Heterorhabditidoides) in its role as an entomopathogen, but the species found in the JS Marais Nature Reserve could most probably be a new species.

In terms of the recovery rate and frequencies reported by previous surveys, caution must be taken because of the differences in the sampling design, the time of sampling and the insect bait used. The high number of positive samples observed in the current survey may be attributed to the use of three different trapping hosts, and the increased number of sampling sites for the area to compensate for the patchy distribution of EPNs, as reported by Stock et al. (1999). The use of different insect hosts has shown that some isolates prefer certain hosts, which may lead to an underestimation of the natural occurrence of EPNs. However, the current study confirms that the use of mealworms is very effective for the isolation of species.

The present study provides an understanding of the distribution and diversity of the EPNs found in the JS Marais Nature Reserve, which is a natural ecosystem surrounded by urban development, which has remained undisturbed for over a century, in the heart Stellenbosch, in the Western Cape province. The unique intraspecific variation found in the ITS region of the S. khoisanae populations of the JS Marais Nature Reserve should be further investigated, as well as its implications for the use of the ITS region as a molecular marker.