Control of the tomato leaf miner, Tuta absoluta (Meyrick) (Lepidoptera: Gelechiidae) larvae in laboratory using entomopathogenic nematodes from subtropical environment

Abstract Tomato (Solanum esculentum) is one of the vegetable crops grown by both smallholder and commercial farmers in the Kingdom of Eswatini. Tomato leaf miner, Tuta absoluta M. is a major insect pest of tomatoes resulting in reduced tomato yields throughout the country. The study investigated the virulence of two sub-tropical EPN species on T. absoluta larvae. Steinernema yirgalemense and S. jeffreyense at different concentrations (0, 20, 40, 60 IJs/insect) were screened for efficacy (i.e. mortality rate) against larvae of T. absoluta in laboratory bioassays. Results obtained showed that S. yirgalemense and S. jeffreyense were able to kill T. absoluta larvae without significant differences between the two EPN species in 24-well bioassay trays. Significantly higher (p < 0.05) mortality was observed when 60 IJs/insect was used. The combination of S. yirgalemense at 60 IJs/insect (100%) and S. jeffreyense at 60 IJs/insect (98.3%) resulted in significantly higher (p < 0.05) mortality, compared with the other four combinations of EPN and concentration tested. In the leaf bioassays, S. yirgalemense (58.8%) resulted in significantly higher (p < 0.05) mean mortality compared to S. jeffreyense (46.4%). Steinernema yirgalemense at 60 IJs/insect resulted in significantly higher mean mortality compared to the other EPN and concentration combinations in the leaf bioassay. The results indicated that both EPNs tested were effective against T. absoluta larvae. Steinernema yirgalemense at 60 IJs/insect can effectively find T. absoluta larvae inside leaf mines, but large-scale field trials are recommended to demonstrate the potential use of the biocontrol agent within integrated pest management programs.

Tomato plants are among the world's most cultivated crops and they are cultivated by both smallholder and commercial farmers in the Kingdom of Eswatini (FAO, 2012). Tomatoes are targeted by a vast number of insect pests and diseases including bacterial wilt (Ralstonia solanacearum), fusarium wilt (Fusarium oxysporum) and tomato leaf miner, Tuta absoluta (Lepidoptera: Gelechiidae). Tomato leaf miner is one of the most consequential lepidopteran pests that affect tomato plants (Husin, 2017). Silva et al. (1998) reported that T. absoluta is a limiting factor for tomato production throughout the world, resulting in 70% losses in most tomato growing areas. The insect pest is able to attack fresh tomatoes and those grown for processing (Desneux et al., 2011). The newly introduced lepidopteran pest from South America has high reproduction potential; capable of producing 10 to 12 generations per year under the favorable conditions (Derbalah et al., 2012). With such high reproduction potential, they are likely to undergo genetic changes (mutation) which in turn leads to the development of pesticide resistant populations (Husin, 2017).
Tomatoes are attacked by T. absoluta at any developmental stage and the primary method for managing T. absoluta involves the use of synthetic insecticides (Tropea et al., 2012). The use of synthetic chemicals is not only detrimental to the environment and human health, but has other serious disadvantages, including reduced profits due to high insecticides cost, destruction of natural enemy populations, build-up of insecticide residue on the tomatoes and insect resistance (Lietti et al., 2005;Husin, 2017). The efficiency of chemical control on T. absoluta infestations has been reported to be poor due to the entophytic habit of the larvae, which are protected in the leaf mesophyll or inside fruits (Mallia, 2009;Terzidis et al., 2014). Herbert et al. (2001) reported that higher levels of resistance to Abamectin, Aartap and Permethrin were correlated with greater use of these compounds by tomato growers in Brazil. The finding suggested that the variation in insecticide use in Brazil resulted in variation in the susceptibility of T. absoluta (Herbert et al., 2001).
The use of environmentally sound insect pest management strategies is important to minimum use of insecticides in tomato fields. Environmentally friendly strategies include cultural control (e.g. crop rotation, selective removal and destruction of infested plant materials), and the use of natural enemies (parasitoids, predators and entomopathogens) (Husin, 2017). Entomopathogenic nematodes (EPNs) are good alternatives to synthetic insecticides, are soil-dwelling organisms that attack insect pests that live in, on, or near the soil surface and can be used effectively to control important insect pests (Adams and Nguyen, 2002). EPNs in the families Steinernematidae and Heterorhabditidae do not affect non-target species, do not leave residues (Georgis et al., 2006) and are essential biocontrol agents used for controlling insect pests (Grewal and Georgis, 1999).
These EPNs are capable of penetrating and killing their hosts within 24-48 h of nematode invasion, which is caused by their mutualistic relationship with bacteria and of the genera Photorhabdus and Xenorhabdus that are carried in the intestine of Heterorhabditidae and Steinernematidae, respectively (Akhurst and Boemare, 1990;Husin, 2017). EPNs were reported to control T. absoluta but the use of sub-tropical EPN species is not yet known to farmers in the Kingdom of Eswatini. In this study, we tested the virulence of two sub-tropical EPN species on T. absoluta larvae.

Source of insects
Tuta absoluta larvae were collected from infested tomato fields around Mankayane (26 °44'58"S 31°02'56"E), kept in Perspex boxes measuring 15 × 20 cm and transported to the Entomology laboratory, Faculty of Agriculture, University of Eswatini. Tuta absoluta was mass-reared in rearing cages (50 × 50 × 50 cm) (Vermandel, Hulst, The Netherlands) on tomato plants at 25 ± 2 °C, 65 ± 5% RH, with a 16:8 L:D photoperiod. Rearing cages (60 × 60 × 90 cm) (Vermandel) with tomato plants on pots were used to mass-rear adult T. absoluta and were allowed to lay eggs for 48 h. The tomato plants with eggs were transferred back to the other rearing cages measuring 50 × 50 × 50 cm. After hatching, larvae were allowed to feed on potted tomato plants at ad libitum. Data on the developmental time and morphological descriptions was used to separate between the four T. absoluta instars (Cuthbertson et al., 2013;Vargas, 1970). Last instar larvae were then harvested by means of opening the mines and picking the larvae. Tenebrio molitor (Coleoptera: Tenebrionidae) (mealworm) larvae, used for culturing nematodes, were reared on a diet comprised of wheat bran, as reported by Van Zyl and Malan (2013).  Kaya and Stock (1997), using mealworm larvae kept at room temperature (±25°C). The IJs were harvested from the White trap during the first week of emergence, stored horizontally using 500-ml vented culture flasks, and used within one month after harvesting. The culture flasks were shaken on a weekly basis, to increase the amount of aeration, and the survival of the IJs, during storage. The origin of the two EPN species used in the study is indicated in Table 1.

Laboratory bioassay
Virulence experiments were conducted in 24-well bioassay plates (flat-bottom, Nunce, Cat. No.144530, Thermo Fisher Scientific (Pty) Ltd, Gauteng, Johannesburg, South Africa). The 24-well bioassay protocol was used to test the potential of Steinernema yirgalemense and S. jeffreyense to infect T. absoluta larvae under optimal laboratory conditions. Alternate wells were lined with a circular piece of 13-mmdiameter filter paper, to secure an even distribution throughout the well. Each alternate bioassay well with a T. absoluta larva (n = 60), was inoculated with 20, 40, 60 IJ/50 μ l water, while the control received distilled water only. Each piece of filter paper was inoculated with a predetermined concentration (Navon and Ascher, 2000) of IJs in filtered tap water. A control treatment comprised of 50 µl filtered tap water. One insect was added to each alternate well, and the trays were closed with the lid, placed in a closed plastic container lined with moistened tissue paper, and left in a growth chamber at 25°C for 48 h. For each treatment, five trays with 12 wells were used (n = 60). Mortality by nematode infection was confirmed after 48 h, by dissection under a stereo microscope to check for the presence of EPNs. The procedure was repeated on a separate date, with a different batch of nematodes.

Leaf trial
The experimental design consisted of eight tomato leaves for each treatment which was replicated four times in a completely randomized design. Tuta absoluta larvae were removed from infested tomato leaves collected from the field with the help of a needle and a camel hair brush under a light microscope. In total, 12 T. absoluta larvae were placed on the upper side of fresh tomato leaves and each leaf was placed in a petri dish. The lid of each petri dish was sealed with Parafilm (Pechiney Plastic Packaging, Neenah, WI, USA) to prevent T. absoluta larvae from escaping. The leaves containing the larvae were left for 24 h to mine and develop well-formed mines. Steinernema yirgalemense and S. jeffreyense at different concentrations (20, 40, 60 IJs/insect) were prepared and sprayed on the top side of the infested leaves, while the control treatments received distilled water only. Petri dishes, with the sprayed tomato leaves, were then placed in one plastic container per treatment, lined with wet paper towels to ensure high humidity, and kept in an incubator at 25°C. Tuta absoluta larval mortality was checked after 2 days and mortality was confirmed by dissecting the dead larvae under a light microscope to determine whether EPNs are present inside the dead larvae. The procedure was repeated on a separate date, with a different batch of EPNs to confirm results. Total numbers of dead larvae per EPN species at different EPN concentration were recorded.

Data analysis
Virulence assays were analyzed using analysis of variance (ANOVA); if the F-value was significant (p < 0.05), the means were differentiated by LSD MEANS (SAS Institute, 1985). The mortality data were corrected for the corresponding control mortality, using the formula: where CM is the corrected mortality, T is the percent mortality in treated insects, and C is the percent mortality in untreated insects (Abbott, 1925). If no significant interactions were found to occur between the main effect of the test dates and the treatments concerned, data obtained from the two test dates was pooled and analyzed, using a two-way ANOVA.
There was a significant effect (F (3, 28) = 17.50, p < 0.001) of the treatment on the percentage mortality. Steinernema yirgalemense at 60 IJs/ insect (79.2%) provided significantly higher (p < 0.05) mortality of the T. absoluta larvae, than all the other combinations applied. Steinernema jeffreyens at 60 IJs/insect (62.5%) gave larval mortality that was not significantly different from S. yirgalemense at 40 IJs/insect (58.3%) but significantly different (p < 0.05) from the other combinations. Steinernema yirgalemense at 40 IJs/insect and S. jeffreyense at 40 IJs/insect (46.7%) gave larval mortality not significantly different from each other but different (p < 0.05) from the other combinations. Steinernema jeffreyense at 40 IJs/insect and S. yirgalemense at 20 IJs/insect (39.1%) gave mortality not different from each other but significantly different (p < 0.05) from the other combinations. Steinernema jeffreyense at 20 IJs/insect (30.0%) gave larval mortality not significantly different from S. yirgalemense at 20 IJs/ insect but significantly different (p < 0.05) from the other combinations. The control gave T. absoluta larval mortality significantly lower (p < 0.05) compared to all the other combinations (Fig. 2).

Discussion
Two EPN species, S. yirgalemense and S. jeffreyense isolated from sub-tropical environment, were used for the first time against T. absoluta larvae with successful control obtained for both species. Results from the current study are consistence with Van Damme et al. (2016), who reported that S. carpocaspsae and S. bacteriophora are also capable of killing T. absoluta larvae. Both EPN species recorded higher mortality rates, confirming the nematode's potential to be used in an IPM program against T. absoluta. Husin (2017) also reported that three commercial EPN species were successful in killing the larvae and adults of T. absoluta after both growth stages were continuously exposed to the EPNs for 48 h. The author recorded 100% mortality when the EPNs were exposed to second, third and fourth larval instars while 96% was recorded when the first larval instar was exposed to the EPNs. In the current study, however, only the fourth larval instars were used resulting in 93% mortality in the laboratory bioassay and 58.8% when the EPNs where sprayed on a leaf infested with T. absoluta.
In another study by Batalla-Carrera et al. (2010), the third and fourth larval instars of T. absoluta were reported to be highly susceptible to continuous exposure to S. carpocapsae, S. feltiae and H. bacteriophora for up to 48 h. The same was reported by Mutegi et al. (2017) when two EPN species (Heterorhabdities sp. and Steinernema karii) were used to test their effect on T. absoluta mortality, with both studies reporting that EPNs were able to infest and kill T. absoluta larvae under laboratory conditions. In another study by Ngugi et al. (2018), indigenous EPN isolates (TK1, S86, 97, 75 and R52) from Kenya were screened for infectivity on 2nd and 3rd larval stages of T. absoluta. All the five of the indigenous EPNs were effective against T. absoluta larvae as they caused larval mortality and infective juveniles were recovered. They explained that the ability of EPN to infect and reproduce within the host pest ensures persistence and recycling ability in the natural environment. This is critical for effective pest control and commercial production efficiency of EPNs (Kalia et al., 2014;Ngugi et al., 2018).
In the larvae bioassay, there was no significant difference on mortality caused by the two different EPN species. Both EPN species had equal capability of infesting and killing the larvae when T. absoluta larvae were exposed on the filter paper. In the leaf bioassay, S. yirgalemense was more effective compared to S. jeffreyense. Host cues including the release of carbon dioxide and other odorants released by host attract or repel EPNs which normally cause directional movement in laboratory studies (Lewis et al., 1993;Grewal et al., 1994;Bal and Grewal, 2015). The nematodes, using the cues, needed to first find the larvae inside the leaf mines and different nematodes respond differently on carbon dioxide cues. In the current study, S. yirgalemense is smaller in size compared to S. jeffreyense. With the narrow natural openings on T. absoluta larvae, the natural openings are capable of reducing the invasion of IJs making it difficult for the nematodes to enter (Renn, 1998). Such results are in line with Bastidas et al. (2014) who observed that the size of the EPN species plays an irrefutable role in limiting penetration of host insects with larger EPNs less capable of invading smaller hosts. Bal and Grewal (2015) reported that EPNs have to locate suitable hosts in order to complete their life cycle. Laboratory studies show differences in the host finding behavior of EPNs including cruisers which are active searchers, and ambushers which sit and wait foragers for the potential host (Lewis et al., 1992;Campbell and Gaugler, 1993;Bal and Grewal, 2015). The two EPNs used in the current showed differences in the way they located T. absoluta larvae in the leaf mines. A forager nematode actively moves through the soil or substrate to locate sedentary and slow-moving insects by detecting the carbon dioxide or other volatiles released by the insects (Campbell and Gaugler, 1997). Steinernema jeffreyense IJs were able to locate and infest find less T. absoluta larvae while S. yirgalemense may be an intermediate forager that employs both an ambush and a cruise foraging strategy which assisted in locating T. absoluta in the leaf mines (Campbell and Gaugler, 1997). Therefore, the intermediate forager S. yirgalemense is more effective at finding the larvae of T. absoluta inside the mines than the ambush forager S. jeffreyense.
The increase in mortality with an increase in concentration can be attributed to large number of EPNs and symbiotic bacteria released by the EPNs when they penetrate the larvae as reported by Eleftherianos et al. (2010). Mutegi et al. (2017) reported that Heterorhabditis sp. and Steinernema karii were able to kill T. absoluta larvae at various concentrations: 100, 300, and 500 IJs/ml. Tuta absoluta larvae mortality increased with increase in concentrations. The highest concentration (500 IJs/ ml) achieved the highest mortality. Goudarzi et al.
(2015) on a study of the effects of H. bacteriophora and S. carpocapsae as a biological control agents of Agrotis segetum (Lepidoptera) found that the highest larval mortality was achieved when EPNs were applied at a dose of 100 IJs/insect compared to 25, 50 and 75 IJs/insect. In another study by Dlamini et al. (2019), they reported that S. yirgalemense, H. noenieputensis and the exotic S. feltiae gave good control of banded fruit weevil (BFW), Phlyctinus callosus (Schönherr)(Coleoptera:Curculionidae) larvae at 100 IJs/insect. The results on the leaf bioassay experiment showed that S. yirgalemense and S. jeffreyense were able to kill T. absoluta larvae on tomato leaves, regardless of their relative size. Batalla-Carrera et al. (2010), on the study of efficacy of EPNs against T. absoluta in laboratory and greenhouse conditions, showed that EPNs are able to find and kill larvae on tomato leaves, despite their relative position (inside or outside the tomato leaf). Mortality caused by nematodes at a rate of 60 IJs/cm 2 varied from 76.3% by H. bacteriophora to 88.6% and 92% by S. carpocapsae and S. feltiae, respectively. The above results concurs with those reported by Gozel and Kasap (2015), who reported that EPNs are able to enter feeding canals in the leaves of tomatoes and kill T. absoluta larvae. Many larvae of T. absoluta died inside galleries, which show that IJs were able to find and infect them inside the mines.
The current results are not in agreement with those obtained by Sabry et al. (2016), who reported that T. absoluta mines per leaflets slightly decreased when S. carpocapsae was used alone against T. absoluta. The percentages of leaflets mines reduction in the first, second and third application were 10.5, 17.8 and 12.9%, respectively. The results showed that spinetoram and indoxacarb resulted in 99.3 and 80% mine reduction, respectively while S. carpocapsae alone was the least toxic. The results indicated that S. carpocapsae was not effective against T. absoluta due to the susceptibility of S. carpocapsae to some abiotic factors such as drought, light, UV radiation and high temperature. Also, the presence of the T. absoluta larvae in tunnels reduces the effectiveness of S. carpocapsae. In the current study, reduced mortality in the leaf bioassay was as a result of the fact that the EPNs were to look and find the insect in the mines while in the laboratory bioassay the T. absoluta larvae were exposed to the EPNs.
Steinernema yirgalemense at 60 IJ/insect was found to be more effective in controlling T. absoluta both in the laboratory and leaf bioassay. Such findings reveal that for the control of T. absoluta larvae in tomato leaves, S. yirgalemense would be more effective compared to S. jeffreyense. Steinernema jeffreyense is bigger in size compared to S. yirgalemense, making it to be easily limited by sieve plates present in opercules, stopping the nematode from penetration and causing more mortality (Ishibashi and Kondo, 1990). The smaller size of T. absoluta larvae coupled with smaller natural openings sizes hindered S. jeffreyense from entering through the normal infection routes or the larvae may produce smaller amounts of attractants such as carbon dioxide or kairomones, which makes it more difficult for nematodes to locate them (Husin, 2017).
If S. jeffreyense is the only nematode available, 40 IJs/insect is the maximum concentration that would be used to control T. absoluta larvae in leaf mines. This is because when the concentration of S. jeffreyense increases to 60 IJs/insect the same mortality of T. absoluta larvae is observed, hence increasing concentration will be a waste of the nematodes and resources. Van Damme et al. (2016) reported that mortality caused H. bacteriophora, S. carpocapsae and S. feltiae at a rate of 60 IJs/cm 2 varied from 76.3, 88.6 and 92% for H. bacteriophora, S. carpocapsae and S. feltiae, respectively. This also confirms that different EPNs applied at the same rate can result in different mortality, showing the virulence of each EPN species. The same results were observed in the current study where S. yirgalemense caused higher mortality compared to S. jeffreyense.

Conclusion
The two EPN species tested in the current study were effective in killing the larvae of T. absoluta under laboratory conditions regardless of their position and S. yirgalemense was more effective in killing T. absoluta larvae compared to S. jeffreyense. An increase in concentration resulted in an increase in mortality and among the evaluated nematodes concentrations, 60 IJs/insect was found to be more effective in killing T. absoluta larvae. The most effective nematode and EPN concentration combination to kill T. absoluta was 60 IJs/insect in both the lab and leaf bioassay. Steinernema yirgalemense at 60 IJ/insect was effective is recommended as it showed higher mortality rate. The study also showed the possibility of using the EPN species in IPM programs which reduce pollution, detrimental effects on the environmental, human health and reduce pesticide residues on tomato fruits while delaying the development of insect resistant to insecticides.