Potential of bacteria isolated from chitin‐enriched soil against Euschistus heros (Hemiptera: Pentatomidae)

Abstract

BACKGROUND

Biological control using entomopathogenic microorganisms is a more sustainable alternative to chemical pesticides for managing insect pests in agriculture. The isolation of novel entomopathogenic microorganisms is essential to broaden the genetic base of bioinsecticides, which will enable their effective application under diverse cropping conditions. This study aimed to isolate entomopathogenic bacteria from chitin‐enriched soil with potential activity against E. heros.

RESULTS

From soil amended with chitin for 90 days, isolates CC1, CC2, and CC3 were obtained and identified as Bacillus thuringiensis, Bacillus altitudinis, and Priestia megaterium, respectively. Bioassays for control of E. heros indicated that the tarsal contact inoculation method was the most effective compared with dorsal contact and ingestion methods, with the tested isolates. Insect mortality increased with higher concentrations of inoculated bacteria. Isolates CC1 and CC2, using the tarsal inoculation method at the highest cell concentration (2.5 × 10⁹  CFU mL ⁻¹), caused stink bug mortalities of 79% and 61% within 10 days, respectively. Isolate CC3 showed the lowest insecticidal activity.

CONCLUSION

The bacteria CC1 (B. thuringiensis) and CC2 (B. altitudinis), isolated from chitin‐enriched soil, demonstrate promising potential as bioinsecticides against E. heros under laboratory conditions. To our knowledge, these findings represent the first report of the use of B. altitudinis in the control of E. heros. © 2025 The Author(s). Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

B. thuringiensis and B. altitudinis, isolated from chitin‐enriched soil, causing mortality in Euchistus heros under laboratory conditions. First report of B. altitudinis in the control of E. heros.

1. INTRODUCTION

Currently, arthropods are primarily controlled in agriculture and livestock through the use of synthetic pesticides, which are expensive, polluting, and originate from nonrenewable sources. ¹ , ² Biological control is a strategy that aligns with growing societal expectations for the production of healthier foods with less environmental impact. ³ In the 2024/2025 season, there was a 13% increase in the use of bioproducts in Brazil, and in the last 3 years, the growth rate was four times higher than the global average. ⁴ Approximately 60% of Brazilian farmers use some form of biocontrol. ⁵

The brown stink bug, Euschistus heros (Fabricius, 1798) (Hemiptera: Pentatomidae), is one of the main pests that affect soybean crops. Both adults and nymphs pierce the stems, leaves, and pods with their sucking mouthparts, feeding on the grains, which leads to wilting and pod abortion. ⁶ , ⁷ Owing to the significant economic importance of soybeans and the severe damage caused by E. heros, farmers have increasingly relied on chemical insecticides for their control. Consequently, resistance has developed to several active ingredients, including organophosphates (methamidophos, acephate, chlorpyrifos, and monocrotophos), cyclodienes (endosulfan), pyrethroids (lambda‐cyhalothrin and imidacloprid), and neonicotinoids (thiamethoxam). ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³

In Brazil, 57 commercial biological products are recommended for E. heros control. ¹⁴ However, the diversity of microorganisms used as active ingredients in these products is low: only the fungi Beauveria bassiana (Bals.‐Criv.) Vuill. and Metarhizium anisopliae (Metschn.) Sorokin, and the bacteria Pseudomonas chlororaphis, P. fluorescens, and B. thuringiensis. ¹⁴ Considering the wide variety of climate, crop, and agricultural management practices worldwide, expanding the genetic base of bioinsecticides in terms of species and strain/isolate diversity is necessary to enable their effective use under different cultivation conditions. ¹⁵ , ¹⁶ To expand the pool of entomopathogenic agents, additional isolations from environmental matrices are necessary, a laborious but essential endeavor for advancing the application of biological inputs in agriculture. Conventional isolation approaches rely predominantly on recovering microorganisms from diseased insects collected in the field. ¹⁷ However, soil constitutes one of the planet's most microbially diverse habitats and represents a prolific reservoir for organisms with biotechnological potential, including candidates for biopesticide development. ¹⁸

Bacillus spp. is an important genus of soil bacteria that are in the select group of endospore formers, which allows biological products based on them to have a longer shelf life. ¹⁹ , ²⁰ Bacillus species have several mechanisms that are used for the control of agricultural pests, including the production of phospholipase C enzymes, chitinases, proteinases, Cry and/or Cyt proteins, and paralytic sphingomyelinase toxins. ²¹ , ²² , ²³

Lipolytic enzymes, including certain phospholipases, may contribute to the degradation of the lipid‐rich epicuticle, the outermost layer of the arthropod exoskeleton, thereby increasing water loss and weakening the protective barrier of the exoskeleton. ²⁴ Once this barrier is compromised, entomopathogenic microorganisms can progress into the underlying cuticular layers, where chitin degradation occurs primarily through the action of chitinases and associated proteases. ²⁵ , ²⁶ Chitin is the main component of the exoskeleton of arthropods, ²⁷ and resistant to degradation by most soil microorganisms because of the strong β‐1,4 glycosidic linkages between N‐acetylglucosamine residues. ²¹ , ²⁸ Proteolytic and lipolytic enzymes can degrade key structural and physiological internal components of insects through complementary pathways. Proteases act on structural proteins of muscle tissues, the intestinal epithelium, and the protein‐rich components of the peritrophic membrane, compromising nutrient absorption and contributing to insect starvation. ²⁹

B. thuringiensis is the most studied entomopathogenic species in this genus, ³⁰ but there are also studies on B. subtilis, ³¹ , ³² B. pumilus, ³³ B. flexus, ³⁴ and B. cereus ³⁵ in the control of caterpillars, other stink bugs, acari, nematodes, flies, etc., demonstrating the great potential of this genus for isolating new microorganisms for the control of pests and diseases in agriculture. Additionally, these bacteria grow quickly, adapt to different culture media conditions, and are easy to cultivate in bioreactors. ³⁶ , ³⁷ To meet the demand for more sustainable pest control, the present study aimed to isolate entomopathogenic bacteria for the control of E. heros from chitin‐enriched soil.

2. MATERIAL AND METHODS

2.1. Isolation of bacteria

A sample of Haplic Acrisol (WRB) was collected from the 0–10 cm layer in an agricultural area (−29.718349, −53.707703) with a history of soybean, corn, oats, and wheat cultivation. The soil was placed in three 1 kg pots, and 8.33 g of ultrapure chitin (Sigma‐Aldrich) was mixed into each pot. The pots were incubated at room temperature in a greenhouse, and the soil was kept moist for 90 days. After this period, 1 g of soil from each pot was added to Erlenmeyer flasks containing 9 mL of saline solution (8.5 g L⁻¹ NaCl). The flasks were shaken at 80 rpm at 28 °C for 24 h. From each flask, 1 mL was transferred to test tubes and kept in a water bath for 12 min at 80 °C. The three suspensions were serially diluted to −6 and plated using the spread plate method on nutrient agar medium (NAM) (3 g L⁻¹ meat extract, 15 g L⁻¹ peptone, 15 g L⁻¹ agar, 1000 mL water, pH 6.8). The plates were then incubated for 36 h at 30 °C in the dark. Colonies were characterized as having Bacillus‐like morphology, and the three most abundant types were isolated by streaking on the same culture medium, resulting in three isolates (CC1, CC2, and CC3), which were examined under a microscope for Bacillus‐like cell morphology and Gram‐positive staining.

2.2. Identification of bacteria

For DNA extraction, pure cultures of these bacteria grown in nutrient broth were diluted in PBS 1× (phosphate‐buffered saline) and centrifuged at 8000 g for 3 min to obtain a cell pellet. Subsequently, 700 μL lysis buffer was added to the pellet, and DNA extraction was performed as proposed by Bag et al. ³⁸ and described by Moraes et al. ³⁹

Amplification of the 16S rRNA region was performed using the primers 8F (5′‐AGA GTT TGA TYM TGG CTC AG‐3′) and 907R (5′‐CCG TCA ATT CMT TTR AGT TT‐3′). ⁴⁰ The reactions consisted of buffer 1× (20 mM Tris–HCl pH 8.4; 50 mM KCl), 2 mM MgCl2, 0.2 mM dNTPs, 0.5 U Platinum Taq DNA polymerase (Invitrogen, Waltham, USA), 4 pmol of each oligonucleotide, 60 ng DNA, and sterile pure water to 20 μL. Amplification was conducted in a Nexus thermocycler (Eppendorf, Hamburg, Germany), programmed for one cycle at 95 °C for 3 min, 35 cycles at 94 °C for 30 s, 50 °C for 30 s, and 72 °C for 40 s, and finally, one cycle at 72 °C for 10 min.

The CC1 isolate was also subjected to PCR for the amplification of a gyrB gene fragment specific to B. cereus and B. thuringiensis, using the primers BCFW1 (5‐GTT TCT GGT GGT TTA CAT GG‐3′) and BCRW1 (5′‐CAA CGT ATG ATT TAA TTC CAC C‐3′) ⁴¹ following the same methodology as for 16S rDNA region amplification. PCR products were purified using the Wizard Gel Clean‐Up kit (Promega) and subjected to sequencing PCR using the Big Dye Terminator v3.1 kit (Applied Biosystems, Waltham, MA, USA) according to the manufacturer's instructions.

Sequencing was performed using an ABI3130 sequencer (Applied Biosystems, Waltham, MA, USA). The resulting electropherograms were prepared by the DNA Sequencing Analysis Software Version 6.0 and analyzed using the Phred/Phrap/Consed package, ³⁹ , ⁴² to obtain the trimmed consensus sequences, considering only bases with Phred quality greater than 20. Qualified DNA sequences were compared with other sequences deposited in the GenBank by BLAST tool. ⁴³ Phylogenetic analysis was then carried out in MrBayes 3.2.3 software ⁴⁴ using the best fit model according to Akaike Information Criterion ⁴⁵ in four independent MCMC runs with five million generations, with sampling every 100 generations. At the end of analysis with a standard deviation less than 0.01, the initial 25% of the trees were discarded as burn‐in.

2.3. Screening for extracellular enzyme production

To evaluate the capacity of the bacterial isolates to produce proteases, culture medium Milk Agar (MA) was used, with this composition: peptone (0.5 g L⁻¹; Kasvi, Italy), yeast extract (0.3 g L⁻¹; Kasvi, Italy), skimmed UHT milk (10 g L⁻¹; Italac, Brazil) and agar (15 g L⁻¹; LabSynth, Brazil). ⁴⁶ The medium was autoclaved for 20 min at 121 °C without the UHT milk; the required volume of milk was added aseptically after sterilization. The isolates were streaked onto MA plates and incubated at 30 °C for up to 72 h. Hydrolysis halos were evaluated every 24 h. The assay was performed in quintuplicate (n = 5).

The ability of the bacterial isolates to produce lipase/esterase was detected by culturing the microorganisms on Tributyrin Agar (TA). The medium was composed of peptone (0.5 g L⁻¹; Kasvi, Italy), yeast extract (0.3 g L⁻¹; Kasvi, Italy), tributyrin (1 g L⁻¹; Sigma‐Aldrich, USA) and agar (15 g L⁻¹; LabSynth, Brazil). ⁴⁷ The isolates were streaked onto TA plates and incubated at 30 °C for up to 72 h. The presence of translucent halos was considered positive for lipolytic activity. The assay was performed in quintuplicate (n = 5).

The production of chitinases by bacteria was evaluated on solid medium enriched with colloidal chitin (MEQ). Colloidal chitin was prepared from chitin. ⁴⁸ Briefly, 5.0 g of ultrapure chitin powder (Sigma‐Aldrich, USA) was added to an Erlenmeyer flask containing 60 mL of concentrated HCl (LabSynth, Brazil), followed by shaking for 1 h in a flow hood (Nalgon, Brazil). The total volume was transferred to an Erlenmeyer containing 400 mL of ethanol (50 mL L⁻¹) and agitated for 30 min. The entire mixture was then poured into 1.0 L of chilled distilled water (4 °C). The solution was filtered using filter paper and a vacuum pump. The colloidal chitin was washed with distilled water and further filtered until pH 7.0 was reached. The chitin was then stored at 4 °C. The MEQ medium was composed of Na₂HPO₄ (0.6 g L⁻¹; LabSynth, Brazil), K₂HPO₄ (0.3 g L⁻¹; Merck, Germany), NH₄Cl (0.1 g L⁻¹; LabSynth, Brazil), NaCl (0.05 g L⁻¹; LabSynth, Brazil), yeast extract (0.005 g L⁻¹; Kasvi, Italy), agar (15 g L⁻¹; LabSynth, Brazil), nystatin (0.0005 g L⁻¹; Teuto, Brazil), produced colloidal chitin (0.1 g L⁻¹) and the fluorescent dye Calcofluor White M2R (0.1 g L⁻¹; Sigma‐Aldrich, USA) at pH 7.0. Isolates were streaked onto MEQ plates and incubated at 30 °C for up to 96 h. The presence of hydrolysis halo was evaluated every 24 h under UV light in a flow hood. The assay was performed in quintuplicate (n = 5).

2.4. Preparation of the inoculum

Initially, growth curves for the three bacterial strains were obtained. For this, they were separately inoculated into 125 mL Erlenmeyer flasks containing 50 mL of nutrient broth and incubated at 30 °C with shaking at 150 rpm. Samples were taken every 6 h for 72 h, one for measuring optical density at 540 nm and another for plating on Petri dishes to determine colony‐forming units (CFUs). For inoculum preparation, the three bacteria were grown separately in the same culture medium and under the same conditions described above until reaching a cell density of 2.5 × 10⁸ CFU mL⁻¹, when they were inoculated onto the insects as described below. In addition to the pure bacterial cultures, a further inoculum was prepared by mixing the three bacteria in equal proportions, hereafter referred to as the ‘Mix’.

2.5. Control of the Brown stink bug

Adult brown stink bugs of E. heros were obtained from the Insect Integrated Pest Management Laboratory at Universidade Federal de Santa Maria (UFSM). The five treatments evaluated in the stink bug control test were the pure inocula (at a concentration of 2.5 × 10⁸ CFU mL⁻¹) of each of the three bacteria, the mixture of all three bacteria (Mix), and the control (sterile deionized water), all prepared as described above. The bacteria were inoculated onto stink bugs using three methods: tarsal contact, dorsal contact, and ingestion. Each method used five replicates consisting of Petri dishes containing five stink bugs. All inoculum preparation and inoculation procedures were performed inside a laminar flow chamber.

For the tarsal contact method, 1 mL of the bacterial suspension was applied to a 90 mm diameter filter paper, sterilized by autoclaving, and placed at the bottom of sterile Petri dishes. Five stink bugs were then placed in each Petri dish and incubated, as described below. ⁴⁹ In the dorsal contact method, five stink bugs were placed in a sterile Petri dish, and 2 μL of the suspension was inoculated onto the dorsum of each bug with the aid of a micropipette. In both the tarsal contact and dorsal contact methods, two green bean pods, previously washed in sodium bicarbonate solution (12 g L⁻¹) and vinegar (15 mL L⁻¹), air dried by laminar flow chamber ventilation, and sterilized with UV light for 15 min, were added to each Petri dish. The pods were replaced every 2 days. ⁴⁹ For the ingestion method, after washing and superficial sterilization of the pods as described above, they were submerged for 5 min in each bacterial suspension or in sterile deionized water (control) and then allowed to dry under laminar airflow. Afterwards, they were placed in pairs, together with five stink bugs, in sterilized Petri dishes. All Petri dishes containing the stink bugs received a small amount of sterile cotton soaked in sterile water to provide moisture to the insects. The cotton was replaced every 2 days. The Petri dishes were incubated at 26 ± 2°C with a 14‐h photoperiod, and mortality was assessed daily for 10 days. ⁴⁹

Another test was performed to assess stink bug mortality caused by inoculum concentrations of 2.5 × 10⁶, 2.5 × 10⁷, 2.5 × 10⁸, and 2.5 × 10⁹ CFU mL⁻¹, using the tarsal contact method. Afterwards, they were inoculated onto stink bugs, which were then incubated and evaluated as previously described. For all methods, the colonies were re‐isolated from the insect body and incubated on nutrient agar medium to confirm that the recovered bacterium corresponded to the originally inoculated strain, according to the Koch's postulates.

2.6. Data analysis

Stink bug mortality data were corrected using Abbott's formula. ⁵⁰ Corrected mortalities were subjected to analysis of variance, and when significant, were separated by Tukey's test (α = 0.05) using the SISVAR software (version 5.3). ⁵¹ Probit analysis was used to estimate the mean lethal time (LT₅₀ = time to cause mortality of 50% of the population) for each inoculum concentration using the Ecotoxicology package in R software. ⁵²

3. RESULTS

3.1. Identification of the bacteria

Partial sequences (~820 bp) of 16S rDNA from the CC1, CC2 and CC3 bacterial isolates were obtained and compared with sequences available in GenBank. The 16S rDNA sequence of the CC1 isolate showed 100% identity with B. cereus and B. thuringiensis with the sequences deposited on GenBank. To better discriminate between the species, partial sequencing of the gyrase subunit B gene (gyrB) from the CC1 isolate was performed, and the similarity with GenBank sequences suggested that the isolate was B. thuringiensis, with 100% identity. The sequences of the CC2 and CC3 isolates showed 100% identity with B. altitudinis and Priestia megaterium (formerly Bacillus megaterium), respectively. The corresponding accession numbers of the analyzed 16S rDNA sequences from bacterial isolates CC1, CC2, and CC3 were deposited in GenBank under the codes PV867501.1, PV867502.1, and PV867503.1, respectively. In addition, the gyrB sequence of isolate CC1 was deposited under accession number PX501550. Phylogenetic analysis, based on the selected evolutionary model, supported and confirmed the identification of the species described above (Fig. 1).

Figure 1.

Bayesian phylogenetic analysis of 16S rDNA sequences of Bacillus and Priestia species, including the isolates CC1, CC2 and CC3 derived from soil. Posterior probability values are indicated at the nodes.

3.2. Enzyme production by bacteria

All bacteria grew on the three solid culture media used to evaluate enzymatic activity. Visible colonies were observed after the incubation period. The formation of halos indicating protease and lipase production was observed for all three bacteria, whereas halos resulting from chitinase activity were observed only in the culture medium inoculated with CC1 (Table 1).

Table 1.

Halo formation (+) in solid culture media indicating potential production of chitinases, proteases, and lipases by isolates CC1 (B. thuringiensis), CC2 (B. altitudinis), and CC3 (P. megaterium)

Isolate	Chitinases	Proteases	Lipase
CC1	+	+	+
CC2	−	+	+
CC3	−	+	+

3.3. Mortality of stink bugs using different inoculation methods

The isolate CC1 stood out by causing the highest stink bug mortality among the three inoculation methods evaluated (Fig. 2). In the tarsal contact method, the corrected mortality rate caused by this bacterium at the end of the experiment (10 days) was 68% (Fig. 2(a)). From the 4th day after inoculation, the performance of this isolate stood out from the other treatments and, by the 10th day, showed an average mortality rate 16% higher than that of the isolate CC2, which had the second‐best performance. However, a statistically significant difference between CC1 and CC2 was observed only on the 10th day post‐inoculation. In dorsal contact inoculation, the isolate CC1 caused a corrected mortality of 40% at the end of the experiment and was superior to CC3 and Mix from the 9th day onwards (Fig. 2(b)). However, no statistical differences were observed between CC1 and CC2 bacteria on any of the evaluated dates in dorsal contact inoculation. When the isolate CC1 was inoculated using the pod immersion method, it caused a corrected mortality of 47% in stink bugs on the 10th day (Fig. 2(c)). From the 5th day, it stood out from the other treatments; however, it showed statistically significant differences from all other treatments on the 7th and on 9th and 10th day post‐inoculation from CC3 and Mix. From the 5th to 10th day, the average mortality rate was 12% higher than that caused by the isolate CC2.

Figure 2.

Cumulative mortality corrected by Abbott's formula (±SD) of adult stink bugs (E. heros) after inoculation with three bacteria CC1 (B. thuringiensis), CC2 (B. altitudinis), and CC3 (P. megaterium) and with a mixture of the three bacteria (Mix) via tarsal contact (a), dorsal contact (b), and ingestion (c). The standard deviation is represented by the vertical bars shown for each data point, and when a bar is not present, it indicates that there was no standard deviation for that point. Means followed by the same letter do not differ from each other according to Tukey's test (P ≤ 0.05).

CC2 (B. altitudinis) caused mortality in stink bugs of 47%, 30%, and 32% after 10 days of inoculation by tarsal contact, dorsal contact, or pod immersion methods, respectively (Fig. 2). The mortality caused by CC2 was statistically lower than that caused by CC1 and statistically higher than that caused by CC3 only on the 9th day of the experiment with tarsal contact inoculation (Fig. 2(a)). When inoculated using this method, the average mortality from the 4th to the 10th day was 11% higher than that of the Mix treatment and 17% higher than that of the CC3 treatment. The dorsal inoculation method stood out from the two lower‐performing treatments starting from the 3rd day, and until the last day of the experiment, the average mortality was 9% higher than that of the treatments (Fig. 2(b)). With pod immersion, the isolate CC2 performed similarly to the CC3 and Mix treatments up to the 7th day of the experiment. From the 8th to the 10th day, it stood out prominently, with an average mortality rate of over 13% (Fig. 2(c)).

The treatments consisting of the isolate CC3 and a mixture of the three bacteria (Mix) showed the worst performance in controlling stink bugs across all three methods evaluated (Fig. 2). When inoculation occurred via tarsal contact, the Mix treatment stood out from the CC3 treatment between days 7th and 10th day, with average mortality rates 10% higher, although without statistical significance (Fig. 2(a)). In the dorsal contact and pod immersion methods, the mortality rates of the two treatments were the same (Fig. 2(b),(c)). Thus, considering the average across the three inoculation methods, stink bug mortality caused by bacteria at the end of the experiment (10 days) was 52% for CC1, 36% for CC2, 22% for Mix, and 19% for CC3.

The data in Fig. 2 also allow for a comparison of the efficiency of the three inoculation methods in causing stink bug mortality. For this calculation, only the two best‐performing bacteria (CC1 and CC2) were considered. The tarsal method was the most efficient, causing, on average, 58% mortality in stink bugs when inoculated with bacteria CC1 and CC2. The second most efficient method was pod immersion, which caused an average mortality of 40% in stink bugs, a value close to that observed in the dorsal inoculation method (35%).

3.4. Mortality of stink bugs at different inoculum concentrations

For this test, bacteria CC1 and CC2 and the tarsal inoculation method were selected, as they were the most effective in controlling stink bugs in previous tests. For both bacteria, the mortality increased as the cell concentration in the inoculum increased (Fig. 3). Thus, after 10 days of experiment and at the highest cell concentration (2.5 × 10⁹ CFU mL⁻¹), the highest stink bug mortality occurred, which was 79% for CC1 and 61% for CC2. The lowest mortality was 30% for CC1 and 17% for CC2 at the lowest cell concentration (2.5 × 10⁶ CFU mL⁻¹). The mortality caused by the concentrations 10⁷, 10⁸, and 10⁹ CFU mL⁻¹ were similar and did not differ statistically from each other from the 3rd day onward for CC1 and from the 6th day onward for CC2. On the other hand, the mortality caused by the lowest cell concentration was statistically different from the others from the 3rd day onward for CC1, but for CC2 the statistical differences between 10⁶ and 10⁷ CFU mL⁻¹ concentration occurred just in 10th day.

Figure 3.

Cumulative mortality corrected using Abbott's formula (±SD) of adult stink bugs (E. heros) after tarsal contact inoculation with bacteria CC1 (a) and CC2 (b) at concentrations of 2.5 × 10⁶; 2.5 × 10⁷; 2.5 × 10⁸ and 2.5 × 10⁹ CFU mL⁻¹. Means followed by the same letter do not differ significantly according to Tukey's test (P ≤ 0.05).

3.5. Estimation of lethal time

Using the data obtained from the mortality test of stink bugs, it was possible to estimate the LT₅₀, which is the estimated lethal time to kill 50% of the population (Table 2). There were no statistical differences in the LT₅₀ between the bacteria when inoculated at the same cell concentration, although there was a tendency for it to be smaller for CC1. A reduction in LT₅₀ was observed with an increase in inoculum concentration. Isolate CC1 had an LT₅₀ of 4.24 days at the highest concentration and 10.15 days at the lowest, while isolate CC2 exhibited an LT₅₀ of 5.42 days at the highest concentration and 11.38 days at the lowest. The data obtained from the mortality test of stink bugs at different inoculum concentrations did not fit the Probit model; therefore, it was not possible to estimate the LC₅₀ (lethal concentration to kill 50% of the population).

Table 2.

Estimated lethal time for mortality of 50% of E. heros adults (LT₅₀) inoculated by the tarsal contact method with different concentrations of bacteria CC1 and CC2

Concentration (CFU mL−1)	Bacteria	Angular coefficient(±a)	LT50 (days) (95%b)	χ2 (c)
2.5 × 106	CC1	2.99 (±0.317)	10.15 (8.87–15.78)	1.54 (8)
	CC2	1.73 (±0.208)	11.38 (9.06–19.23)	1.25 (8)
2.5 × 107	CC1	2.46 (±0.124)	5.21 (4.29–6.36)	0.97 (8)
	CC2	1.91 (±0.071)	7.33 (5.77–10.63)	0.35 (8)
2.5 × 108	CC1	3.17 (±0.160)	5.16 (4.39–6.02)	1.26 (8)
	CC2	2.75 (±0.158)	6.43 (5.42–7.91)	1.20 (8)
2.5 × 109	CC1	3.06 (±0.219)	4.24 (3.52–4.96)	2.66 (8)
	CC2	2.80 (±0.223)	5.42 (4.56–6.48)	2.67 (8)

^a = standard error; ^b= confidence interval; ^c = tabulated degrees of freedom (P = 0.05; df = 5). The number of individuals tested for each treatment was 25.

4. DISCUSSION

In this study, three bacteria of the genus Bacillus and one of the genus Priestia were isolated, two of which exhibited outstanding performance. The bacteria CC1 and CC2, when inoculated at the highest concentrations (2.5 × 10⁹ CFU mL⁻¹), caused mortalities of 79% and 66% of the stink bugs after 10 days, respectively. Considering that the stink bugs were inoculated with an aqueous suspension of bacteria, this mortality would likely be higher if the bacteria were mixed into formulations with additives that enhance their action, as is done in the bioinput industry.

CC1 showed the best performance in controlling the brown stink bug among all three inoculation methods (tarsal contact, dorsal contact, and ingestion). In terms of molecular identification, CC1 is part of the B. cereus sensu lato group and was identified as B. thuringiensis based on gyrB gene sequencing. The B. cereus sensu lato group includes the species B. anthracis, B. cereus sensu stricto, and B. thuringiensis, among others because they have close phylogenetic relationships even though with high ecological diversity. ⁵³ , ⁵⁴ Difficulty in distinguishing B. thuringiensis from B. cereus using only 16S rDNA sequencing has been previously reported. ⁵⁵ One common distinguishing feature is the production of Cry proteins by B. thuringiensis, which B. cereus typically lacks. However, some B. thuringiensis strains can lose the plasmids that encode crystal production and thus become phenotypically indistinguishable from B. cereus. ⁵⁶ Crystal formation in B. thuringiensis is plasmid‐encoded, and plasmid curing is a frequent phenomenon in bacteria. ⁵⁷ , ⁵⁸ , ⁵⁹ For these reasons, sequencing of the gyrB gene was performed to achieve a more reliable species‐level differentiation ⁴¹ , ⁶⁰ thereby clearly distinguishing the species in our study.

CC1 produced all tested enzymes (chitinases, proteases, and lipases), suggesting multifactorial insecticidal action against E. heros. These three enzymes are likely to act synergistically to breach the exoskeleton, after which proteases and lipases damage internal tissues. This sequence of events probably accelerates tissue destruction, dehydration, malnutrition, and deeper bacterial invasion, explaining the high mortality observed. ²⁴ , ²⁹ , ⁶¹ , ⁶² , ⁶³

CC2 belongs to the species B. altitudinis and shows the second‐best performance in controlling stink bugs. To our knowledge, this is the first report of the bioinsecticidal activity of B. altitudinis. However, recent studies have demonstrated that this species has significant potential for controlling other targets of agricultural interest. There are reports of the bionematicidal effects of this species against Meloidogyne incognita (Kofoid & White, 1919) ⁶⁴ and M. javanica (Treub, 1885). ⁶⁵ Ye et al. ⁶⁶ isolated eight metabolites from the B. altitudinis AMCC1040 strain, six of which showed nematicidal effects (caprylic acid, 2‐methylbutyric acid, isovaleric acid, 2‐isopropoxy ethylamine, acetic acid, and diacetyl). The strain BRHS/S‐73 showed a biofungicidal effect against Thanatephorus cucumeris (A.B. Frank) Donk (1956), the causal agent of root rot in beans. ⁶⁷ Li et al. ⁶⁸ reported a biobactericidal effect of the AMCC 101304 strain against potato scab, caused by Streptomyces spp.

CC3, identified as P. megaterium, exhibited the lowest insecticidal activity among the three strains tested. According to VARY et al., ⁶⁹ this bacterium is widely used in the pharmaceutical, food, and textile industries, as well as in agriculture. In the latter, there are reports of strains with biofungicidal, bionematicidal, and biobactericidal activities, as well as plant growth promoters. ⁷⁰ , ⁷¹ , ⁷² , ⁷³ The combination of the three isolates (Mix) did not lead to high insect mortality, which may be explained by the two‐thirds reduction in the concentrations of the best performing isolates CC1 and CC2, a factor previously shown in the concentration assays to significantly impair their efficacy. Additionally, the incorporation of isolate CC3, which exhibited limited effectiveness, may have further contributed to the reduced performance observed. CC2 and CC3 showed potential to produce proteases and lipases; however, CC3 exhibited the poorest insecticidal performance, indicating that in vitro detection of extracellular enzymes does not necessarily predict in vivo virulence. Virulence is a complex set of processes that depends on gene expression, secretion dynamics, and additional host–pathogen interactions. ⁷⁴

A comparison of the methods for inoculating bacteria into stink bugs showed that the tarsal method was the most efficient. During spraying under field conditions, only a minority of the droplets directly reached the stink bug, as in the dorsal contact method. This is because stink bugs usually occur in the lower third of soybean plants. ⁷⁵ Tarsal contact proved to be the most efficient inoculation method because of the significant movement of stink bugs throughout the different periods of the day. Walking over areas where droplet deposition occurs results in greater inoculation through tarsal (indirect) contact. ⁷⁶ It was observed that with the dorsal contact method, the droplet containing the bacterial suspension did not spread properly over the dorsum of the stink bugs because surface tension‐reducing additives were not used. Furthermore, studies have reported that stink bugs produce antimicrobial substances, which may inhibit the action of bacteria. ⁷⁷ , ⁷⁸ , ⁷⁹

In the tarsal inoculation test, the bugs remained in constant contact with the surface of the inoculated filter paper, increasing the contact time between the bacteria and bugs. The longer the entomopathogenic microorganisms remain in contact with the insect cuticle, the greater their control efficiency. ⁷⁷ , ⁸⁰ , ⁸¹ Furthermore, the isolation method used in the present study could have favored the production of chitinase‐producing microorganisms. The higher mortality rate observed in bugs inoculated via tarsal contact may be a consequence of the chitinolytic action of the bacteria on the tarsi, thus promoting the breakdown of the exoskeleton and microbial invasion of the insect's body. ⁸² , ⁸³ , ⁸⁴

The concentration of entomopathogenic bacteria that are inoculated into target organisms is one of the main factors determining the success of control. In the present study, the mortality of stink bugs increased as the concentration of cells in the inoculum increased. For the best‐performing bacterium (CC1), the highest concentration (10⁹ CFU mL⁻¹) was 11% and 53% more effective than 10⁸ and 10⁷ CFU mL⁻¹ concentrations, respectively, demonstrating the importance of inoculating adequate amounts of cells for effective biological control. Pu et al. ⁸⁵ also reported a difference in the corrected mortality of Rhynchophorus ferrugineus (Olivier, 1790) from 17% to 94% between B. thuringiensis concentrations of 10⁸ and 10¹¹ CFU mL⁻¹, respectively. Zimmer et al. ⁸⁶ demonstrated that two B. thuringiensis strains, israelensis and kurstaki, only achieved significant efficacy at concentrations above 10⁹ against Musca domestica (Linnaeus, 1758).

We did not find other published studies that estimated the LT₅₀ of Bacillus species for the control of E. heros and the number of studies on fungi is limited. Sousa ⁸⁷ evaluated the control efficiency of E. heros by two fungi (B. bassiana and M. anisopliae), and observed LT₅₀ values of 5.1 and 5.2 days, respectively, at a concentration of 5 × 10⁷ conidia mL⁻¹, similar to the value of 5.2 observed for the CC1 bacteria. In our study, LT₅₀ did not differ between CC1 and CC2 at the same cell concentration because interpolation occurred within the LT₅₀ ranges. However, the importance of increasing the bacterial concentration on LT₅₀ was evident, as the difference between the lowest and highest cell concentrations was significant for both bacteria. The LT₅₀ is an important parameter for assessing insecticide efficacy, as lower values indicate a faster reduction in crop damage and pest reproduction.

In this study, soil enriched with chitin for 90 days was used for bacterial isolation. In theory, this method favors the acquisition of chitinolytic microorganisms, since chitinase production is one of the mechanisms that microorganisms can use to break down the insect exoskeleton, thereby triggering events that lead to its death. Some studies have indicated that chitinolytic microorganisms can effectively control insects. ⁸⁸ , ⁸⁹ One of the main advantages of using chitinolytic microorganisms is their specific mode of action, as chitin is not present in plants or vertebrates. ⁹⁰ In addition, the bacteria produced lipases and proteases in the culture medium, which suggests that these enzymes (along with others not evaluated) likely act synergistically during infection and contribute to the internal damage that leads to insect death. Although promising, further studies are needed to confirm the efficiency of this method for isolating entomopathogenic microorganisms, evaluating different chitin concentrations, soil incubation periods, chitin sources, other microorganism genera, and so on.

The search for new microorganisms that can control pests and diseases in agriculture is ongoing. There is a need for new bioproducts to be made available on the market and to increase the genetic base of bioinsecticides, which will allow for their efficient use in diverse environmental conditions. In the present study, we isolated bacteria belonging to the genera Bacillus that showed preliminary entomopathogenic activity against the brown stink bug in laboratory assays. These microorganisms can be used for the development of new bioinsecticides, but further studies are needed to confirm the efficiency of this bacterium in causing mortality in the brown stink bug and other insects under laboratory, greenhouse, and field conditions.

5. CONCLUSION

The strains CC1 of B. thuringiensis and CC2 of B. altitudinis, isolated from chitin‐enriched soil, showed entomopathogenic activity against E. heros in laboratory assays. Stink bug mortality is more effective when the bacteria are applied to the stink bugs via the tarsal contact method than by pod immersion or dorsal inoculation. The higher the concentration of bacterial cells inoculated on the insect, the greater and faster the mortality. To our knowledge, this work is the first to report B. altitudinis as a bioinsecticidal agent against E. heros.

CONFLICT OF INTEREST

The authors declare that there are no conflicts of interest regarding the publication of this article.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.