Evaluation of the Potential of Fungal Biopesticides for the Biological Control of the Seed Bug, Elasmolomus pallens (Dallas) (Hemiptera: Rhyparochromidae)

Elasmolomus pallens is a post-harvest insect pest of groundnuts which causes severe yield loss to farmers, particularly in Africa and Asia. Resistance to synthetic chemicals has been on the rise among insects and is a constraint on insecticides regulations. In view of the drive for alternative approaches to synthetic insecticides, this study evaluated the potential of biopesticides based on entomopathogenic fungi against E. pallens under laboratory conditions. Fungal isolates from the bug cadaver including Fusarium proliferatum F1, Aspergillus tamarii F2, A. flavus F3, Trichoderma atroviride F4, A. niger F4, and Metarhizium anisopliae (Meschn.) Sorokin, originating from the cadaver of Zonocerus variegatus were screened for virulence against the bug. Adult bugs were dipped briefly in conidial concentration 1 × 108 conidial mL−1 and observed at 25 ± 2 °C, 80 ± 10 RH and 14: 10 L:D for 10 days. The fungal isolates caused mortality ranging from 48 to 100% based on their potential to infect and kill the bug. Five conidial concentrations (1 × 104 to 1 × 108 conidia mL−1) were evaluated against adult bugs in the multiple-dose virulence bioassay. Lethal concentrations (LC50 and LC90) values of 6.75 × 106 and 4.42 × 109 conidia mL−1 were obtained for A. flavus F3 while M. anisopliae had 8.0 × 106 and 6.14 × 108 conidia mL−1 respectively. Lethal time (LT50 and LT90) values were 3.3 and 6.2 days for A. flavus F3 compared to 3.6 and 5.6 days for M. anisopliae, respectively. Thus, A. flavus F3 showed potential against E. pallens; and can be considered as an ideal isolate for incorporation into formulations for field applications.


Introduction
Elasmolomus pallens (Dallas) is a seed bug which belongs to the family Rhyparochromidae, within the superfamily Lygaeoidea and in the order Hemiptera. The family was formerly considered as a subfamily within the Lygaeidae until the recent reanalysis of the phylogenetic relationship of the taxon [1]. Members of the family are cosmopolitan in nature and are considered the largest of the lygaeoid bugs [2]. E. pallens, formerly known as E. sordidus (Fabricius, 1787), has a wider distribution cutting across tropical, subtropical, and parts of the temperate regions of Africa, Asia, and the Pacific Islands [3]. An important determinant in the dispersal of E. pallens is its association with the legume, Arachis hypogea L., also called peanuts or groundnuts [4].
In Sub-Saharan African countries where peanut farming is practiced and processing implements are lacking, harvested peanuts are left for weeks on the field to dry before stripping. When this aflatoxicosis which develop into cancers [26]. Because of these features of the fungus, researchers are pessimistic in recommending its application as a biological control agent [25]. However, many studies have found out that not all strains of A. flavus are naturally endowed to produce aflatoxins because, in nature, more than a few strains are atoxigenic and cannot produce aflatoxins [27][28][29]. Over the years, atoxigenic strains of A. flavus has been widely used for the biocontrol of aflatoxigenic strains in agricultural plantations. These strains are able to control the proliferation of the aflatoxigenic strains in plantations via competitive inhibition and bio-exclusion processes [25,30,31]. This implies that A. flavus has the potential to confer a dual advantage when applied as a biological control agent.
Considering that both adults and nymphs of E. pallens attack peanuts, causing both physical damage, loss of seed quality (Figure 1), and the attendant negative effects of synthetic insecticides on the environment, the need for an alternative and environment friendly approach for its control becomes imperative. This suggests the need to explore biological control agents effective for the control of E. pallens, other than being restricted to only known species as biocontrol agents. This will be fundamental to knowledge expansion on both the host insect and pathogens. In this study, the potential of EPF against E. pallens was evaluated for its biological control under laboratory conditions. Insects 2019, 10, x FOR PEER REVIEW 3 of 18 aflatoxicosis which is fatal and or chronic aflatoxicosis which develop into cancers [26]. Because of these features of the fungus, researchers are pessimistic in recommending its application as a biological control agent [25]. However, many studies have found out that not all strains of A. flavus are naturally endowed to produce aflatoxins because, in nature, more than a few strains are atoxigenic and cannot produce aflatoxins [27][28][29]. Over the years, atoxigenic strains of A. flavus has been widely used for the biocontrol of aflatoxigenic strains in agricultural plantations. These strains are able to control the proliferation of the aflatoxigenic strains in plantations via competitive inhibition and bio-exclusion processes [25,30,31]. This implies that A. flavus has the potential to confer a dual advantage when applied as a biological control agent.
Considering that both adults and nymphs of E. pallens attack peanuts, causing both physical damage, loss of seed quality (Figure 1), and the attendant negative effects of synthetic insecticides on the environment, the need for an alternative and environment friendly approach for its control becomes imperative. This suggests the need to explore biological control agents effective for the control of E. pallens, other than being restricted to only known species as biocontrol agents. This will be fundamental to knowledge expansion on both the host insect and pathogens. In this study, the potential of EPF against E. pallens was evaluated for its biological control under laboratory conditions.

Figure 1.
Effect of E. pallens attack on peanut kernels. Affected seed kernels become shriveled, losing quality and aesthetic value.

Bugs Collection and Rearing
Adult bugs were collected from groundnut fields in Menglembu, Perak (4°33′58.29′′ N and 101°2′53.95′′ E), Malaysia, from February to April 2017 in plastic containers and transported to the laboratory. The bugs were maintained in the laboratory as in Khosravi, Sendi [36] with some modifications. The bugs colonies were reared inside plastic cages (40 × 30 × 20 cm) at 25 ± 2 °C and 70 ± 10% relative humidity (RH) under a 14:10 h L:D photoperiods. Cotton balls soaked in sterile distilled water were placed in each container to maintain the humidity within the containers and provide water for the bugs to suck. Fresh groundnut seeds were used as a diet and were changed every 3 days. Gauze materials were placed in each container as a surface for oviposition. Gauze containing Figure 1. Effect of E. pallens attack on peanut kernels. Affected seed kernels become shriveled, losing quality and aesthetic value.

Bugs Collection and Rearing
Adult bugs were collected from groundnut fields in Menglembu, Perak (4 • 33 58.29 N and 101 • 2 53.95 E), Malaysia, from February to April 2017 in plastic containers and transported to the laboratory. The bugs were maintained in the laboratory as in Khosravi, Sendi [36] with some modifications. The bugs colonies were reared inside plastic cages (40 × 30 × 20 cm) at 25 ± 2 • C and 70 ± 10% relative humidity (RH) under a 14:10 h L:D photoperiods. Cotton balls soaked in sterile distilled water were placed in each container to maintain the humidity within the containers and provide water for the bugs to suck. Fresh groundnut seeds were used as a diet and were changed every 3 days. Gauze materials were placed in each container as a surface for oviposition. Gauze containing oviposited eggs were separated into new containers to hatch at 70 ± 10 (RH). Adult bugs, 2-3 weeks, were used for the virulence bioassay experiments.

Identification of Bug Species
The identification of the bug species was done via morphological examinations and molecular method. Genomic DNA from the legs of ethanol-preserved specimens were extracted and used for molecular identification [37], while morphological identification was done according to the method described by Katinka [38].

Fungal Isolates
Cadavers of E. pallens were collected from groundnut fields (as mentioned above) inside sterile plastic bags and transported to the laboratory for further analysis. The cadavers were surface sterilized by dipping in 1% (v/v) sodium hypochlorite (NaClO) solution for 5 mins and rinsed three times in sterile distilled water. The surface sterilized cadavers were placed on potato dextrose agar (PDA, Oxoid) plate amended with 0.5 g/L chloramphenicol and incubated in the dark at 25 • C for 3 days. Mycelia that developed were sub-cultured until pure cultures were obtained and kept on PDA slants at 4 • C until required. M. anisopliae was obtained from the stock culture collection of the Department of Microbiology, University of Ibadan (Nigeria) and used as a standard strain in this study.

Microscopic Observation
The morphology of the isolates was observed following the conventional slide culture technique. Sterile PDA was cut into small squares of (approximately 1 cm) and a block was placed on a sterile glass slide inside Petri plates (100 × 15 mm) underlaid with sterile filter paper (Whatman 1). Each agar block was inoculated with a fungal colony on the four corners using sterile needles and a coverslip aseptically placed over it. The filter paper underlay was wetted with distilled water and incubated in the dark at 25 • C for 3 days. The coverslip was gently taken off and placed on a glass slide containing a drop of lactophenol cotton blue (LPCB) and examined under the light microscope (Leica DM500, Leica Microsystems, Wetzlar, Germany) with a 40× objective lens [39].

In Vitro Screening of Isolates for Aflatoxins Production
Putative A. flavus isolated from the bug cadavers were screened for their ability to produce aflatoxins on coconut agar medium (CAM). The screening was done according to the method described by Mamo and Shang [29].

Genomic DNA Extraction and Polymerase Chain Reaction (PCR)
Genomic DNA extraction for both bug and fungi were done using the DNeasy Blood and Tissue Kit (Qiagen, Hilden, Germany) following the manufacturer's protocol. For the bug DNA, PCR was done according to the protocol of Tembe, Shouche [37] using the primer sets LCO1490 (5 -GGTCAACAAATCATAAAGATATTGG-3 ) and HCO2198 (5 -TAAACTTCAGGGTGACCAAAAAATCA-3 ), while a fragment of the internal transcribed spacer (ITS) region of the fungal genomic DNA was amplified using the universal primer sets ITS1 (5 -TCCGTAGGTGAACCTGCGG-3 ) and ITS4 (5 -TCCTCCGCTTATTGATATGC-3 ). PCR protocol was conducted as in White and Bruns [40]. Sequences were aligned and compared to existing sequence data on the GenBank database using the Basic Local Alignment Search Tool (NCBI BLAST) [41].

Phylogenetic Analysis
The phylogenetic analysis involving the ITS sequence data of the isolates was done by comparing with registered sequences on the GenBank database. ITS sequences were aligned, and unnecessary portions removed before the tree was generated using MEGA 7.0 software, which gave the topology and length of the branches [42]. The phylogenetic relationship and the neighbor joining analysis were determined using the Kimura 2-parameter (K2P) model of base substitution in MEGA 7.0 [37].

Fungal Conidial Preparation
Fungal isolates were cultured on PDA (Oxoid Ltd., Basingstoke, UK) for 14 days at 25 ± 1 • C under 14:10 h photoperiods. Conidia were harvested by scraping the surface of the medium using sterile wooden spatula into 20 mL 0.05% (v/v) Tween 80 (polyoxyethylene sorbitan monooleate, Sigma Chemical Co., St. Louis, MO, USA) inside sterile 50 mL centrifuge tubes and homogenized by vortexing for 3 min. The conidial suspension was sieved through four layers of cheese cloth [20]. A ten-fold serial dilution of the spore suspension was done, and an aliquot estimated on a hemocytometer followed by a final adjustment to 10 8 . For the multiple-dose bioassay, five different conidial concentrations (1 × 10 4 to 1 × 10 8 conidia mL −1 ), were prepared.

Conidial Viability
The conidial viability for all isolates was determined by plating 100 µL of 10 6 conidia mL −1 on three replicate plates of PDA. The plates were incubated at 25 • C under 14:10 L:D regimes for 16-24 h. Germination was estimated by counting 300 conidia (100 conidia per area) under a brightfield microscope at 40× magnification. The formation of germ tubes at least half the size of the conidia, was indicative of viability [43].

Single-Dose Virulence Assessment Bioassay
Fungal isolates obtained from cadavers of the bugs and M. anisopliae were evaluated for virulence against E. pallens according to the method of Orduño-Cruz, Guzmán-Franco [32] with some modifications. Mixed sex adult bugs were used in five replicate groups of 10 each by dipping into 1 × 10 8 mL −1 of conidia for 10 s while control groups were treated with 0.05% (v/v) Tween 80. Treated bugs were placed in plastic containers (65 × 45 mm) containing damp sterile filter paper (Whatman 1). Cotton wool soaked in sterile distilled water was placed in each container to provide water for the insect to suck. Afterwards, the treated bugs and the controls were introduced into plastic containers with perforated lids to prevent the insects from escaping, provide ventilation and maintain the required humidity. Both the treatment and the controls were maintained at 25 ± 2 • C, 80 ± 10% relative humidity (RH) and 14:10 h photoperiod without diet. Mortality was scored after every 24 h for 10 days.

Multiple-Dose Virulence Bioassay
The virulence bioassay of entomopathogenic fungi against E. pallens was performed following the method of Resquín-Romero and Garrido-Jurado [44] with some modifications. A. flavus and M. anisopliae, which produced the highest mortality rates, were selected for the multiple-dose virulence bioassay. Five conidial concentrations were prepared, and the treated bugs and controls were maintained as above. Cadavers were dipped in 1% (v/v) sodium hypochlorite (NaClO) solution for 5 min to surface sterilize and rinsed 3 times in sterile distilled water. The disinfected cadavers were incubated at 25 ± 2 • C on sterile wet filter paper in Petri dishes for 5-7 days to activate conidial growth followed by examination under light microscope.

Data Analysis
Mortality data were corrected using Abbott's formula [45]. Statistical analyses of data were done using the IBM SPSS 21.0 software (USA). Analysis of conidial viability and bug mortality data was performed using one-way analysis of variance (ANOVA) and the means compared using the least significant difference (LSD) test. Probit analysis was used to estimate the lethal concentration (LC 50 and LC 90 ) and the estimated lethal time (LT 50 and LC 90 ) from infection-confirmed mortalities [46]. All statistical tests were significant at (α) 0.05.

Molecular and Morphological Identification of E. Pallens
The cytochrome oxidase I (COI) sequence of the bug sample was obtained and used for the molecular identification. The aligned sequence had a COI sequence data of 667 bp after incorporation into GenBank sequence data, thus representing members of the Rhyparochromidae as revealed by congeneric sequences obtained after BLAST. Morphologically, the bug is dark brown in color with small head in relation to the body, a long antenna (four segments) and a rostrum positioned on the head. The bug has two compound eyes, the maxillae and mandibles are fused into a needle-like stylet. Each pair of legs is positioned on a separate thorax and a well-established pronotum. The front wing pairs are positioned on the mesothorax while the metathorax bears the second pairs of wings while the abdomen contains 9 segments. BLAST confirmed the identity of the bug as E. pallens (GenBank accession number-MK024388) with 99% similarity with existing sequences on the GenBank database.

In Vitro Screening of Isolates for Aflatoxins Production
In the in vitro screening, positive plates containing aflatoxigenic strains appeared pinkish while negative plates with atoxigenic strains appeared colorless ( Figure 2). Out of 11 putative isolates of A. flavus screened on CAM, only 3 were aflatoxigenic while the remaining 8 were atoxigenic.

Molecular and Morphological Identification of E. Pallens
The cytochrome oxidase I (COI) sequence of the bug sample was obtained and used for the molecular identification. The aligned sequence had a COI sequence data of 667 bp after incorporation into GenBank sequence data, thus representing members of the Rhyparochromidae as revealed by congeneric sequences obtained after BLAST. Morphologically, the bug is dark brown in color with small head in relation to the body, a long antenna (four segments) and a rostrum positioned on the head. The bug has two compound eyes, the maxillae and mandibles are fused into a needle-like stylet. Each pair of legs is positioned on a separate thorax and a well-established pronotum. The front wing pairs are positioned on the mesothorax while the metathorax bears the second pairs of wings while the abdomen contains 9 segments. BLAST confirmed the identity of the bug as E. pallens (GenBank accession number-MK024388) with 99% similarity with existing sequences on the GenBank database.

in vitro Screening of Isolates for Aflatoxins Production
In the in vitro screening, positive plates containing aflatoxigenic strains appeared pinkish while negative plates with atoxigenic strains appeared colorless ( Figure 2). Out of 11 putative isolates of A. flavus screened on CAM, only 3 were aflatoxigenic while the remaining 8 were atoxigenic.

Phylogeny of the Fungal Isolates
The MEGA 7.0 software was used to construct the phylogenetic tree based on the sequence data of the ITS1-5.8S-ITS4 rDNA region. The fungal ITS sequence data were used for species identification and construction of phylogenetic relationship. The phylogenetic analysis showed that the 5 fungal genera had high level of similarities to existing sequences previously reported on the GenBank database ( Figure 3). However, the M. anisopliae used in this study was isolated from Zonocerus

Phylogeny of the Fungal Isolates
The MEGA 7.0 software was used to construct the phylogenetic tree based on the sequence data of the ITS1-5.8S-ITS4 rDNA region. The fungal ITS sequence data were used for species identification and construction of phylogenetic relationship. The phylogenetic analysis showed that the 5 fungal genera had high level of similarities to existing sequences previously reported on the GenBank database ( Figure 3). However, the M. anisopliae used in this study was isolated from Zonocerus variegatus cadavers. Isolates with ≥90% conidial viability (

Single-Dose Virulence Assessment Bioassay
All fungal isolates tested in the bioassay demonstrated the capacity to infect the bugs. The fungal isolates produced different mortality rates against E. pallens and were found to be significantly different (F 6,70 = 5.758; p < 0.0001). At 7 days after treatment, M. anisopliae caused 100% mortality of the bug after exposure while A. flavus (F3) caused 90% cumulative mortality of the bugs 10 days after treatment, followed by F. proliferatum (F1) 68%, A. niger (F5) 64%, A. tamarii (F2) 62%, and T. atroviride (F4) 48% compared to the control where mortality never reached 15% (Figure 4). A. flavus (F3) and M. anisopliae produced the highest mortalities and were selected for the multiple-dose virulence bioassay to further determine their virulence potential.
Other than corrected mortality, the mean mortality of E. pallens treated with a single concentration (1 × 10 8 conidia mL −1 ) was computed for all the fungal isolates tested in the single-dose virulence assessment test. The mean mortalities of the bug were higher between the 4th and 6th days of exposure to the fungal conidia of all isolates (Table 3)

Isolate
Name Key: * = is from cadaver of Zonocerus variegatus [49], while other isolates were obtained from an E. pallens cadaver. Percent conidial germination ± standard errors. Means followed by the same letter in a column are not significantly different (p > 0.05) according to the least significant difference (LSD) test.

Single-Dose Virulence Assessment Bioassay
All fungal isolates tested in the bioassay demonstrated the capacity to infect the bugs. The fungal isolates produced different mortality rates against E. pallens and were found to be significantly different (F6,70 = 5.758; p < 0.0001). At 7 days after treatment, M. anisopliae caused 100% mortality of the bug after exposure while A. flavus (F3) caused 90% cumulative mortality of the bugs 10 days after treatment, followed by F. proliferatum (F1) 68%, A. niger (F5) 64%, A. tamarii (F2) 62%, and T. atroviride (F4) 48% compared to the control where mortality never reached 15% (Figure 4). A. flavus (F3) and M. anisopliae produced the highest mortalities and were selected for the multiple-dose virulence bioassay to further determine their virulence potential.
Other than corrected mortality, the mean mortality of E. pallens treated with a single concentration (1 × 10 8 conidia mL −1 ) was computed for all the fungal isolates tested in the single-dose virulence assessment test. The mean mortalities of the bug were higher between the 4 th and 6 th days of exposure to the fungal conidia of all isolates (Table 3)
The LC and LT values which are measures of virulence among EPF was determined in terms of conidial concentration and time taken for mortality to occur. The LC 50 and LC 90 values obtained were 6.75 × 10 6 and 4.42 × 10 9 conidial mL −1 for A. flavus (F3) compared to 8.0 × 10 6 and 6.14 × 10 8 conidial mL −1 for M. anisopliae, respectively (Table 4), while the LT 50 and LT 90 values of the test isolates were 3.3 and 6.2 days for A. flavus (F3) and 3.6 and 5.6 days for M. anisopliae, respectively (Table 5).

Multiple-Dose Virulence Bioassay
The bioassay utilized different conidial concentrations of the EPF which produced mortality rates proportional to increase in the conidial concentrations used. Here, A. flavus (F3) produced a significant effect on the mortality of the bugs (F5,6 = 5.644, p < 0.0001) with confirmed mortality ranging from 48% at 1 × 10 4 conidia mL −1 to 90% at 1 × 10 8 conidial mL −1 10 days after treatment ( Figure 5A). However, M. anisopliae also produced significant mortality of the bugs (F5,60 = 6.493, p < 0.0001) where a concentration of 1 × 10 4 conidia mL −1 produced 56% cumulative mortality 10 days after treatment while 1 × 10 8 conidia mL −1 caused a cumulative mortality of 100% 7 days after treatment, respectively, compared to the control ( Figure 5B).  The LC and LT values which are measures of virulence among EPF was determined in terms of conidial concentration and time taken for mortality to occur. The LC50 and LC90 values obtained were 6.75 × 10 6 and 4.42 × 10 9 conidial mL −1 for A. flavus (F3) compared to 8.0 × 10 6 and 6.14 × 10 8 conidial mL −1 for M. anisopliae, respectively (Table 4), while the LT50 and LT90 values of the test isolates were 3.3 and 6.2 days for A. flavus (F3) and 3.6 and 5.6 days for M. anisopliae, respectively (Table 5).

Discussion
Entomopathogenic fungi are known natural pathogens infecting insect hosts that can be collected from the field environment either infected or dead and incubated under laboratory conditions to isolate, document, and use the pathogens as biological control agents [56]. In this study, different species of fungi were isolated from the cadavers of E. pallens, some of which have been reported in previous studies. For example, the isolation of A. flavus from the cadavers of insects have similarly been reported in Lee and Kim [15], Assaf and Haleem [53], and [25], respectively; whereas the isolation of F. proliferatum from the asparagus beetle, Crioceris asparagi, was reported in [52].

Discussion
Entomopathogenic fungi are known natural pathogens infecting insect hosts that can be collected from the field environment either infected or dead and incubated under laboratory conditions to isolate, document, and use the pathogens as biological control agents [56]. In this study, different species of fungi were isolated from the cadavers of E. pallens, some of which have been reported in previous studies. For example, the isolation of A. flavus from the cadavers of insects have similarly been reported in Lee and Kim [15], Assaf and Haleem [53], and [25], respectively; whereas the isolation of F. proliferatum from the asparagus beetle, Crioceris asparagi, was reported in [52]. Gardezi [48] reported the isolation of A. tamarii from insect cadaver and testing its pathogenicity against several insect species. A. flavus was found to be the most abundant isolate among the fungal isolates identified. Similar finding has been reported in a study on the almond bark beetle, Scolytus amygdali [57]. However, all the fungal isolates demonstrated pathogenicity against E. pallens but at different degrees under the same experimental conditions. This may be attributed to established facts that sucking, forest, and soil-dwelling insect pests are very much susceptible to infection by EPF, because of the vulnerability of their cuticle to conidial adhesion, germination, and penetration contrary to control agents such as bacteria, viruses, parasitoids, and nematodes that must be ingested to initiate infection [58]. Their mechanisms of action are said to be due to inherent pathogenicity (the infective capacity of an entomopathogen resulting in disease) and virulence (the degree to which the host tissues are colonized by the pathogen with time) traits, which are gene-specified intrinsic features of EPF. However, these intrinsic features are largely dependent on host immune response, nature of formulations, growth medium composition, abiotic factors, and methods of application used [59].
Although there is an existing pessimism on the use of A. flavus as a biological control agent because of its potential for aflatoxin synthesis, this study confirmed that not all isolates are aflatoxigenic. This was based on the fact that, from the eleven A. flavus isolates obtained in this study, 8 (72.7%) were found to be atoxigenic while only 3 (27.3%) were aflatoxigenic (Figure 2). Similarly, Gupta and Gopal [25] observed that out of seven A. flavus isolates obtained from three insect groups-Stephanitis typica (lace bug), Opisina arenosella, and Proutista moesta-only two of the isolates were aflatoxigenic. Wicklow and Dowd [60] reported that atoxigenic strains of A. flavus were lethal to the maize corn insect Carpophilus hemipterus due to their ability to secrete certain sclerotial metabolites.
The single-dose virulence assessment bioassay shows that all isolates tested were pathogenic against E. pallens, although the rate of mortalities differed (p < 0.0001). A. flavus (F3) and M. anisopliae showed greater potential for virulence against E. pallens. A. flavus (F3) produced a mortality rate of 90% after 10 days compared to M. anisopliae which caused 100% mortality of the bugs 7 days after treatment. The killing ability of other fungal isolates were lower compared to the two isolates described (Figure 4). The two most effective entomopathogens were selected for virulence bioassay against the bug based on their lethal effects, which could be related to the assertion by Ferron [61], that both fungal species can secrete lipolytic enzymes during infection, which enables them to degrade the proteo-chitin complex. This may likely explain why the two isolates showed more virulence than the other isolates used. Studies have established that killing of insects by entomopathogenic fungi involves a series of successive steps that could lead to the death of the host depending on its ontogenic stage or immune response. These steps involve: (1) adhesion of the fungal conidia on the integument of the insect; (2) conidial germination under optimum conditions to form germ tubes; (3) degradation of cuticular structures by hydrolytic enzymes and mechanical effect to enable penetration; (4) conversion of the hyphae into blastospores to exploit nutrients in the host hemocoel; (5) blastospores utilize the available sugars and as well release toxins inside the hemolymph; (6) blastospores suppress the host immune system and release toxins that expedite killing of the host; (7) fungus exits the host through openings on the cuticle to produce spores on the cadaver surface [41,62]. Furthermore, Ferron [61] opined that once total invasion of the cadaver occurs, fungal conidiation on the surface of the mummified insect cadaver is dependent on the ambient environmental relative humidity. The mycelia develop from within the cadavers to the surface to produce conidiophores only when the atmosphere becomes saturated. Otherwise, the mummified cadaver remains dry and brittle. Under this condition, entomogenous agents become preserved in the form of chlamydospores.
In the multiple-dose virulence bioassay involving the two isolates ( Figure 5), A. flavus (F3) induced 90% cumulative mortalities of E. pallens in a fashion similar to Seye and Bawin [23] who reported high mortalities of the aphid bug Acyrthosiphon pisum caused by A. flavus and A. clavatus isolates, respectively. Gopal and Gupta [63] also reported 90% mortalities of Opsina arenosella larvae after 5 days when exposed to A. flavus AF2 (ITCC 5005) at a concentration of 10 6 conidia mL −1 . Similarly, Gupta and Gopal [25] reported 80% mortalities of nymphs of Stephanitis typica within 3 days of exposure to A. flavus (ITCC 5004). Generalist facultative EPF such as B. bassiana, M. flavoviride, and Paecilomyces spp cause a sigmoidal dose-mortality curve during the killing of susceptible hosts [64]. But Scully and Bidochka [24] evaluated two strains of the opportunistic pathogens A. flavus 9308 and A. flavus 6982 against Galleria mellonella larvae and found that the isolates demonstrated low virulence regardless of the dose used, such that instead of a sigmoidal curve, a rather flat curve was obtained. The study reported that no significant difference was observed with further increase in conidial concentration. However, Karthi and Vaideki [22] reported 71% and 63% mortalities of fourth and third instars of Spodoptera litura exposed to A. flavus at the highest conidial concentration of 4 × 10 6 conidia mL −1 .
However, for bugs treated with M. anisopliae, mortalities in E. pallens reached 100% after 7 days of treatment. Similarly, Loureiro and Moino-Jr [65] observed a 100% cumulative mortalities of the aphid (Hemiptera), Aphis gossypii, and Myzus persicae after 7 days of treatment with M. anisopliae. Also, a 100% cumulative mortality rate of the aphid Aphis craccivora treated with 10 8 conidial mL −1 of M. anisopliae was earlier observed by Saranya and Ushakumari [66] after 4 days of exposure to the fungus. Furthermore, Santos and Freitas [59] reported 100% cumulative mortality of the Hemiptera Thaumastocoris peregrinus exposed to 10 8 conidia mL −1 of M. anisopliae after 10 days. The differences between these findings may be attributed to factors such as isolates and strains variations, types of insect species, host immune responses, and the prevailing environmental conditions.
The LC and LT as yardsticks that measure virulence in terms of conidial concentrations and time length required to kill the test insect population exposed to a given concentration of a pathogen were determined. In the multiple-dose virulence bioassay against E. pallens, A. flavus (F3) showed lower median lethal concentrations (LC 50 ) and median lethal time (LT 50 ) compared to that of M. anisopliae (Tables 3 and 4). This shows that the isolate A. flavus (F3) was more virulent at killing 50% of the test E. pallens population faster and at lower conidial concentration compared to M. anisopliae. However, on the contrary, M. anisopliae progressed to achieve a lower LC 90 and LT 90 compared to A. flavus (F3) ( Tables 3 and 4). This could be attributed to the degree of persistence of the pathogens within the host hemocoel, nutrient exhaustion in the host, and the host's immune response to the pathogens over time. The ability of M. anisopliae to persist against the host's immune responses, exploit available nutrients, and withstand ambient environmental changes would have been responsible for the rise in virulence over A. flavus (F3) at the later stage of the bug infection. Seye and Bawin [23] had earlier reported higher virulence for A. flavus isolate against the aphid Acyrthosiphon pisum (Hemiptera: Aphididae) after 5 days of treatment, producing a lower LC 50 and LC 90 values (1.23 × 10 3 and 1.34 × 10 7 conidia mL −1 ) compared to M. anisopliae (3.67 × 10 3 and 9.71 × 10 7 conidia mL −1 ). Karthi and Shivakumar [22] did report higher virulence of A. flavus isolates against 3rd instar and 4th instar nymphs of Spodoptera litura, respectively. However, the observations in this study shows that A. flavus (F3) possess the capacity to perform better in virulence against some arthropod pest than M. anisopliae, though its insecticidal activities are rarely reported compared to the widely discussed M. anisopliae. But some studies have reported LC 50 values lower than what was obtained in this study. For example, FitzGerald and Hill [35] observed a lower LC 50 (5.29 × 10 5 conidia mL −1 ) for M. anisopliae compared to what was found in this study. Also, Ekesi and Akpa [67] described how four different strains of M. anisopliae differed in virulence against the aphid Aphis craccivora (Koch), with LC 50 values ranging between 3.1 × 10 5 to 7.4 × 10 6 conidia mL −1 . But Shah and Wang [68] asserted that fungal species and strains virulence, substrate composition, and culture methods determine to a large extent whether insecticidal compounds responsible for virulence are produced by conidia of EPF.
The lethal time (LT 50 and LT 90 ) obtained for both fungi used in this study agrees with the finding of Mweke et al. (2018) who observed 3.3 to 6.3 days for EPF tested against Aphis craccivora. But Saranya and Ushakumari [66] reported an LT 50 of 5.54 days for M. anisopliae against the Aphis craccivora, which is longer than obtained in this study. This therefore means that the isolates used in this study showed virulence potential against E. pallens, probably due to the host type, method of fungal conidial application, and the virulence of the strains used.
Death of the insects due to fungal parasitism was confirmed when the insect body becomes mummified and brittle. However, to validate whether death was due to infection by the fungal pathogens, the cadavers were cultured and examined for the development of fungal conidia. This is important such that where these fungal agents are applied as mycopesticides, they must confer the relative advantage of producing spores on the host cadavers to serve as a secondary source of inoculum for continuous propagation, natural host infection, reduction of insect menace, and costs of applications [69]. Since E. pallens attack peanuts by piercing the pods with their rostrum when they congregate under harvested groundnut on the field, secondary infection by EPF which is largely dependent on insect behavior becomes easier due to their susceptibility under such conditions. Treating this bug species with EPF exposes them to infection by the pathogens due to their social interactions and susceptibility of their cuticles under appropriate environmental conditions.

Conclusions
In this study, A. flavus was found to be more virulent against E. pallens compared to M. anisopliae based on their LC 50 and LT 50 values-killing half the population of the test bugs with lower conidial concentrations and at a reduced exposure time. Although it has often been less applied for the biological control of insect pests compared to the widely used M. anisopliae, yet, it has potential for the biological control of E. pallens and should be considered in myco-formulations for field applications. Further research on the field evaluation and development of these EPF into biocontrol agents of E. pallens and their incorporation into integrated pest management (IPM) systems should be explored.