Introduction

The Chrysopidae comprises predatory species with high reproductive potential (Ono et al., 2017). Some Chrysopidae species show great potential for augmentative biological control of agricultural pests (Pasini et al., 2018). Lacewings, such as Ceraeochrysa cubana, Ceraeochrysa cincta, and Chrysoperla externa (Neuroptera: Chrysopidae), have been utilized for pest control, targeting aphids (Hemiptera: Aphididae), whiteflies (Hemiptera: Aleyrodidae), mites (Acari), and the coffee leaf miner (Lepidoptera: Lyonetiidae) (Pappas et al., 2011; Silva et al., 2021). In Brazil, studies on lacewings have been limited to a few species, primarily focusing on C. externa, due to its adaptability to diverse environmental conditions (Figueiredo et al., 2021). Adult C. externa feed on pollen grains, nectar, plant exudates, and honeydew (Albuquerque, 2009; Principi & Canard, 1984), while the larvae are obligate predators (Salamanca et al., 2011).

Integrated pest management (IPM) promotes the balanced utilization of biological control agents and other control strategies, such as chemical insecticides, semiochemicals, resistant plants, crop management, and genetic control (Jalali et al., 2009; Medina et al., 2009). However, in many cases, excessive insecticide applications are conducted in crops, which can significantly decrease populations of beneficial arthropods, including predators, parasitoids, and pollinators (Bueno et al., 2017; Carvalho et al., 2019). In IPM programs, the use of insecticides that are less harmful to natural enemies is crucial, as the excessive use of non-selective insecticides can disrupt the biological balance and lead to pest resurgence and outbreaks of secondary pests (Diamantino et al., 2014; Moura et al., 2010).

The insecticides thiamethoxam (neonicotinoid) and chlorantraniliprole (anthranilic diamide) (Bentley et al., 2010) are extensively utilized in agriculture and may have negative effects on beneficial insects, including increased mortality and sublethal impacts on non-target organs, which can compromise the insect longevity and reproduction. Despite being considered selective, these insecticides can still induce side effects on non-target insects (Catae et al., 2014; Munhoz et al., 2013). These chemicals suppress pest populations by interacting with their primary site of action, disrupting essential physiological processes, and ultimately causing insect mortality (Munhoz et al., 2013; Sâmia et al., 2018).

Thiamethoxam is a neurotoxic neonicotinoid that directly targets the nervous systems of insects by acting as an agonist on nicotinic acetylcholine receptors (nAChR) (Seifert, 2014). While its primary effect occurs in the nervous system, non-target organs such as the alimentary canal can also be affected (Fernandes et al., 2015).

Upon ingestion, chlorantraniliprole acts as a selective agonist for the ryanodine receptor (Dinter et al., 2009), an intracellular calcium channel. This results in abnormal muscle contraction, followed by rapid cessation of feeding, lethargy, regurgitation, paralysis, and ultimately the death of the insect (Chen et al., 2010; Cordova et al., 2006).

Both thiamethoxam and chlorantraniliprole have potential risks to non-target organisms such as lacewings due to their persistence in pollen grains, floral and extrafloral nectar (Gontijo et al., 2018; Oliveira et al., 2019), which are the main food resources for C. externa adults, as well as the prey consumed by the larvae (Cloyd & Bethke, 2011). Consequently, lacewing adults and larvae may be exposed to these insecticides during foraging activities.

Oral exposure of insects can lead to damage in the non-target alimentary canal, which serves as a potential gateway for various pathogens and xenobiotics (Denecke et al., 2018; Johnson et al., 2009). Although the peritrophic matrix and the midgut epithelium act as barriers against these harmful agents (Higes et al., 2013; Vachon et al., 2012), there is limited data on the side effects of these insecticides on the digestive tract of natural enemies.

Due to the lack of studies evaluating the effects of insecticides on C. externa, this work was conducted to assess the toxicity, midgut histopathological changes, and cytotoxic effects of thiamethoxam and chlorantraniliprole on both C. externa larvae and adults, aiming to uncover potential non-target effects of these insecticides on the predator. Understanding the cytotoxicity of insecticides might contribute to the use of less harmful compounds and the conservation of predatory C. externa populations.

Materials and methods

Insects

Larvae and adults of C. externa were obtained from the Laboratory of Insect Biology at the Department of Entomology at the Universidade Federal de Lavras. The insects were maintained at 25 ± 2 °C, 60 ± 10% RH, and a 12 h photophase. Larvae were kept into cylindrical PVC vials (20 × 20 cm high × diam.) internally lid with towel paper as refuge and fed on eggs of Ephestia kuehniella Zeller (Lepidoptera: Pyralidae) provided ad libitum. Adults were kept in cylindrical PVC vials (20 × 20 cm high × dia) internally lid with paper filter. They were fed on a mixture of brewer’s yeast and honey (1:1) and water (Carvalho & Souza, 2009; Dantas et al., 2021; Farias et al. 2023).

Concentration-mortality bioassay

Insecticide solutions were obtained from the dilution of thiamethoxam (Actara 250 WG, 250 g kg−1 active ingredient, Syngenta Proteção de Cultivos, São Paulo, Brazil) and chlorantraniliprole (Altacor 350 WG, 350 g kg−1 active ingredient, DuPont do Brasil, São Paulo, Brazil) in distilled water. Toxicity tests were performed to estimate the concentrations causing 50% mortality of C. externa larvae (i.e., LC50). The concentrations used in the toxicity tests were 30, 16.66, 3.33, 0.66 and 0.59 ng a.i./µL for thiamethoxam, and 78.75, 52.75, 32.75, 19.62 and 7.87 ng a.i./µL for chlorantraniliprole. These concentrations were calculated based on the highest dose recommended for both insecticides to control coffee pests, which is a crop with natural occurrence of lacewing. The concentrations were obtained from stock solutions of thiamethoxam and chlorantraniliprole, and toxicity tests were carried out until reach of the sublethal concentration. Mortality was recorded after 24, 48, and 72 h of exposure.

Bioassay with larvae

Ephestia kuehniella eggs were dipped in the calculated LC50 of insecticides for five seconds and dried at room temperature for one hour. Control (untreated) eggs were immersed in distilled water for five seconds and dried at room temperature.

First-instar C. externa larvae were individualized in 20 wells microplates (8.54 cm wide × 12.7 cm length) to avoid cannibalism and competition and to ensure that all insects were fed ad libitum. The microplates with larvae were covered with voile tissue. Ephestia kuehniella eggs treated with the calculated LC50 every insecticide tested were onto the voile lids covering the microplates for larval feeding until they reached to the third instar. The insects were maintained at 25 ± 2 °C, 60 ± 10% RH, and a 12 h photophase. The experimental design was completely randomized with three treatments (thiamethoxam, chlorantraniliprole, and control) and four replicates, with five larvae each, which were used for midgut histopathological analysis.

Bioassay with adults

The bioassay was performed with 1 mL of diet (beer yeast and honey solution 1:1) containing the LC50 for thiamethoxam or chlorantraniliprole. The control treatment consisted in offering the untreated food source. Diet containing the insecticides were supplied for 24 h with a cotton piece placed on the bottom of the glass vials (3 cm high × 1.5 cm in diameter) (Carvalho & Souza, 2009). The insects were maintained at 25 ± 2 °C, 60 ± 10% RH, and a 12 h photophase. The design was completely randomized with three treatments (thiamethoxam, chlorantraniliprole, and control) and four replicates, each with ten 24 h-old adults, which were used for midgut histopathological analysis.

Light microscopy

For histopathological analyses, 10 larvae and 10 adults of C. externa from each treatment were dissected in 125 mM NaCl one, 12, and 30 days after exposure. The midguts were divided into anterior and posterior regions and transferred to 2.5% glutaraldehyde in 0.1 M sodium phosphate buffer (pH 7.2) for 12 h. Then the samples were dehydrated in a graded ethanol series (30, 50, 70, 80, 90, and 95%) and embedded in historesin Leica (Leica Biosystems, Germany) following manufacturer’s instruction. 3-μm thick sections were obtained on a rotatory microtome (RM 2155, Leica Microsystems, Vienna, Austria) with glass knives, stained with hematoxylin and eosin, and subsequently photographed under a light microscope with a digital camera (AxioCam ERc5s, Zeiss, Oberkochen, Germany) with the software AxioVision V. 4.8 (Zeiss, Oberkochen, Germany).

Transmission electron microscopy

Ten larvae and 10 adults of C. externa per treatment were dissected in 0.1 M sodium cacodylate buffer (pH 7.2) and 0.2 M sucrose at one, 12, and 30 days after experiment setup. Following dissection, midguts were divided into anterior and posterior regions and transferred to 2.5% glutaraldehyde in 0.15 M sodium cacodylate buffer pH 7.2 for 12 h. The samples were then post-fixed in 1% osmium tetroxide in the same buffer for two hours. Then, the pieces were dehydrated in a graded ethanol series (70, 80, 90, and 98%) and embedded in LR White resin (London Resin Company Ltd.) following manufacturer´s instructions. Ultrathin sections were obtained in ultramicrotome PowerTomes PT-X (RMC Boeckeler Instruments Inc., Tucson, USA) with glass knives and stained with 1% aqueous uranyl acetate for 10 min and lead citrate (Reynolds, 1963) for 10 min. The sections were analysed in a transmission electron microscope (EM 109, Zeiss, Oberkochen, Germany) at the Núcleo de Microscopia e Microanálise at the Universidade Federal de Viçosa.

Statistical Analysis

Lethal concentrations (LC25, LC50, LC75, and LC90) and confidence limits were determined by Probit analysis based on dose–response relationship (Finney, 1964) with the PROC PROBIT procedure of SAS v. 9.0 for Windows (Institute SAS Inc., 1997).

Results

Concentration-mortality

Pearson's Chi-square test statistic (P > 0.05) indicated that the data fitted the Probit regression models. The LC50 was estimated at 1.49 (95% confidence interval = 0.12–3.09) ng mL−1 (Chi-square = 26.68, Slope = 2.12 ± 0.34, P = 0.08) and 0.56 (95% confidence interval = 1.38–2.61) ng mL−1 (Chi-square = 78.70, Slope = 0.75 ± 0.37, P = 0.05) for chlorantraniliprole and thiamethoxam, respectively.

Histopathology

In the control group, the midgut of C. externa larvae and adults consisted of a single-layered epithelium composed of columnar digestive cells, along with nests of regenerative cells located in the basal region of the epithelium (Figs. 1a, 2a). The columnar cells exhibited similar characteristics in both the anterior and posterior midgut portions, displaying a well-developed apical brush border. The apical surface appeared irregular, with cytoplasm protrusions of various shapes and sizes extending toward the lumen. The nucleus was spherical, containing a few condensed chromatin clusters and a prominent nucleolus (Figs. 1b, 2b). The cytoplasm of the digestive cells contained vacuoles of different sizes in the basal and apical regions (Figs. 4b, 5b). The epithelium rested on a thin basal membrane surrounded by muscles. The midgut lumen exhibited a well-developed basophilic peritrophic matrix. The regenerative cells were clustered in nests located at the base of the epithelium throughout the entire organ, near the basement membrane. These cells were small, oval-shaped, or rounded, with intense basophilic cytoplasm (Fig. 2a, b).

Fig. 1
figure 1

Light micrographs of the midgut of Chrysoperla externa larvae. a) Anterior midgut portion of the control larvae, showing simple epithelium with digestive cells with spherical nucleus (Nu), cytoplasm with vacuoles (Va) and regenerative cells (Rc), well-developed muscle region (Mu), cytoplasmic protrusions (Pt) released into the lumen (L). b) Posterior midgut portion of the control showing bursh border (Bb), oval nucleus (Nu), and cytoplasm vacuoles (Va). c–d) Anterior and posterior midgut portions, respectively, exposed to thiamethoxam, showing digestive cells with brush border (Bb), nucleus (Nu), cytoplasm vacuoles (Va), and muscle layer (Mu). e–f) Anterior and posterior midgut portions, respectively, exposed to chlorantraniliprole, showing cell protrusions (Pt), cytoplasm vacuoles (Va), nucleus (Nu), regenerative cells (Rc), and intact muscle (Mu)

Full size image
Fig. 2
figure 2

Light micrographs of the midgut of Chrysoperla externa adults. a) Anterior midgut portion of the control, showing simple epithelium composed of digestive cells (Dc), brush border (Bb), strongly acidophilic brush border (Bb), lumen (L), nucleus (Nu), muscle (Mu), and basophilic granules (arrows). b) Posterior midgut portion of the control showing apex of the cells with apical protrusions (Pt) released into the lumen (L), cytoplasm weakly basophilic, nests with regenerative cells (Rc) at the base of the epithelium, nucleus with decondensed chromatin (Nu), nucleolus (arrows), cytoplasm vacuoles (Va), and muscle (Mu). c) Anterior midgut portion exposed to thiamethoxam for 24 h, showing digestive cells (Dc), irregular brush border (Bb), dilation of the apical cell region (asterisk), cytoplasmic protrusions with basophilic granules (Pt) released in the lumen (L). d) Posterior midgut portion exposed to thiamethoxam for 24 h, showing peritrophic matrix, nucleua with decondensed chromatin (Nu), and muscle (Mu)

Full size image

Larvae and adults exposed to the thiamethoxam LC50 exhibited midgut epithelium with digestive and regenerative cells that were similar to the control larvae in all post-treatment periods (Fig. 1c-1d, 3a, 3b).

Fig. 3
figure 3

Light micrographs of the midgut of Chrysoperla externa adults. a) Anterior midgut portion exposed to thiamethoxam, evaluated on the 12th day, showing intact brush border (arrow), digesive cell (Dc), nucleus (Nu), cytoplasm vacuoles (Va), regenerative cells (Rc), and muscle (Mu). b) Posterior midgut region exposed to thiamethoxam, evaluated on the 12th day, showing digestive cell (Dc) vacuolization (Va), basophilic nucleus (Nu), regular brush border (Bb), regenerative cells (Rc), and muscle (Mu). c) Anterior midgut portion exposed to thiamethoxam, evaluated on the 30th day, showing stomodeal valve (Sv), regular acidophilic peritrophic matrix (PM), nucleus (Nu), regenerative cells (Rc), and digestive cell (DC) with cytoplasm vacuoles (Va). d) Posterior midgut portion exposed to thiamethoxam, evaluated on the 30.th day, showing highly acidophilic bush border (Bb), apical cytoplasm protrusions (Pt) released in the lumen (L), vacuoles (Va), regenerative cells (Rc), and muscle (Ms)

Full size image

Similarly, C. externa larvae and adults exposed to the chlorantraniliprole LC50 showed no signs of histopathology in the midgut and peritrophic matrix, regardless of the duration of exposure (Fig. 4, 5).

Fig. 4
figure 4

Light micrographs of the midgut of Chrysoperla externa adult control and treated with chlorantraniliprole. a-b) Control epithelium (Ep) of anterior and posterior midgut portions, respectively, showing peritrophic matrix (PM), thick basal membrane (Mb), digestive cells (Dc) with cytoplasm protrusions (Pt), vacuoles (Va), nucleus (Nu), nucleolus (arrows), regenerative cells (Rc), muscle (Mu), and lumen (L). c-d) Anterior and posterior midgut portions, respectively, exposed to chlorantraniliprole for 24 h, showing regular cellular organization, with many regenerative cell nests (Rc), digestive cells (Dc) with brush border (Bb), nucleus (Nu) rich in decondensed chromatin (Nu), nucleolus (arrows), peritrophic matrix (PM), and muscle (Mu)

Full size image
Fig. 5
figure 5

Light micrographs of the midgut of Chrysoperla externa adult treated with chlorantraniliprole. a–b) Longitudinal section of the anterior and posterior midgut portions, respectively, exposed to chlorantraniliprole on the 12th day, showing), regenerative cell nests (Rc), digestive cell (Dc) with regular brush border (Bb), nucleus rich in decondensed chromatin (Nu), nucleolus (arrows), and muscle (Mu). c–d) Longitudinal section of the anterior and posterior midgut, respectively, exposed to chlorantraniliprole, evaluated on the 30th day, showing peritrophic matrix (PM), digestive cell (Dc) with brush border (Bb), cytoplasm vacuoles (Va), nucleus strongly basophillic (Nu), thick basal membrane (Bm), muscle (Mu), and regenerative cell (Rc). Bars = 20 μm. Figure d Bar = 1 μm

Full size image

Cytotoxicity

In the control group, the digestive cells of both C. externa larvae and adults displayed a similar ultrastructure in the anterior and posterior regions of the midgut. The apical surface of the cells had long microvilli supported by actin filaments. The apical cytoplasm was abundant in mitochondria, vesicles with varying electron densities, and spherocrystals. The perinuclear cytoplasm contained rough endoplasmic reticulum, while the basal plasma membrane exhibited some infoldings associated with mitochondria. The lateral plasma membrane appeared smooth, without folds or interdigitations (Fig. 6a, b).

Fig. 6
figure 6

Transmission electron micrographs of the midgut digestive cells of Chrysoperla externa larvae. a) Anterior midgut portion of control, showing mitochondria (Mt), large vesicles (Ve), rough endoplasmic reticulum (Rer), basal membrane infoldings (arrow), myelinic figures (Mf), and basal lamina (Bl). b) Posterior region of control showing apex of digestive cells with microvilli (Mv), cell junction (Jc), abundant presence of mitochondria (Mt) in the apical region, vesicle (Ve), rough endoplasmic reticulum (Rer), vesicles with electron-dense content (Vd). c) Anterior midgut portion of the thiamethoxam treatement, showing cells with microvilli (Mv) towards the lumen (L), lipid droplets (Li), mitochondria (Mt) and granules (Gr). d) Posterior migut portion of the thiamethoxam treatement showing the presence of large vesicles (Ve), mitochondria (Mt) and lipid droplets (Li)

Full size image

Larvae and adults of C. externa that were fed the thiamethoxam and chlorantraniliprole LC50 concentrations exhibited digestive cells similar to those in the control group (Fig. 6c, d), except for the presence of some lipid droplets and autophagosomes (Fig. 7).

Fig. 7
figure 7

Transmission electron micrographs of the midgut digestive cells of Chrysoperla externa larvae exposed to chlorantraniliprole. a) Anterior midgut portion showing cells with regular microvilli (Mv) organization, cytoplasm with electron-dense content (Vd), vesicles with electron-dense concentric content (Vc), autophagic vacuoles (Au), lipid droplets (Li), mitochondria (Mt), and lysosomes (Ls). b) Posterior midgut portion showing cells with the presence of large autophagic vacuoles (Au), mitochondria (Mt), lipid droplets (Li), and vesicles with electron-dense content (Vd)

Full size image

Discussion

The toxicity tests reveal that the insecticides thiamethoxam and chlorantraniliprole induce mortality in both larvae and adults of C. externa. Megahed and El-Bamby (2020) demonstrated that thiamethoxam caused mortality in C. carnea larvae, resulting in a cumulative larval mortality of 83.40% after three days of exposure to the insecticide. This finding aligns with the results obtained by Gontijo et al. (2014), who reported a decreased survival time of C. carnea adults when orally exposed to thiamethoxam. Additionally, Mahmoudi-Dehpahni et al. (2021) reported that thiamethoxam induces mortality in the natural predator Orius albidipennis (Hemiptera: Anthocoridae). Also, chlorantraniliprole decreases the survival rate of the predator Coccinella septempunctata (Coleoptera: Coccinellidae) (He et al., 2019) and causes mortality in larvae and adults of both C. carnea and Chrysoperla johnsoni Henry, Wells, and Pupedis (Neuroptera: Chrysopidae) (Amarasekare & Shearer, 2013).

Our findings indicate that the larvae and adults of C. externa, when fed on eggs treated with LC50 thiamethoxam and chlorantraniliprole, do not exhibit histopathological alterations in the midgut epithelium. However, previous studies have reported their impact on the midgut of non-target species such as the honey bee Apis mellifera (L.) (Hymenoptera: Apidae) (Catae et al., 2014; Oliveira et al., 2013) and the stingless bee Scaptotrigona bipunctata (Lepeletier, 1836) (Apidae: Meliponinae) (Moreira et al., 2018). Specifically, thiamethoxam has been shown to induce cytoplasmic vacuolization in A. mellifera midgut cells (Oliveira et al., 2013). In contrast, our findings demonstrate that C. externa exposed to LC50 concentrations of these insecticides do not exhibit differences in vacuole formation compared to the control insects.

Chlorantraniliprole has been extensively investigated for its selectivity against natural enemies (Amarasekare & Shearer, 2013; Gontijo et al., 2014), including Doru lineare (Eschscholz) (Dermaptera: Forficulidae) (Stecca et al., 2014), Telenomus podisi (Ashmead) (Hymenoptera: Platygastridae) (Silva et al., 2018), C. externa, and Coleomegilla quadrifasciata (Schöenherr) (Coleoptera: Coccinellidae) (Armas et al., 2019). However, in the case of C. externa larvae and adults, that does not occur following ingestion of chlorantraniliprole. Conversely, sublethal concentrations of chlorantraniliprole exhibit cytotoxic effects on the non-targeted midgut cells of the silk worm Bombyx mori Linnaeus (Lepidoptera: Bombicidae) larvae, characterized by irregular protrusions in the microvilli, abnormal apocrine secretion, hypertrophy of regenerative cells, and apoptosis (Munhoz et al., 2013).

A plausible explanation for the absence of histopathological and cytotoxic effects of thiamethoxam and chlorantraniliprole in the midgut of C. externa larvae and adults might be attributed to the mutualistic association between chrysopids and gut symbiotic yeasts and fungi (Hemalatha et al., 2014; Principi & Canard, 1984; Woolfolk & Inglis, 2004). These symbionts are responsible for synthesizing essential nutrients and enhancing defense mechanisms against toxic substances (Dillon & Dillon, 2004; Nardi et al., 2002). Additionally, the activity of detoxifying enzymes, such as mono-oxygenases and esterases, may also contribute to avoid tissue damages in the midgut. Previous studies have reported that these enzymes play a role in determining the susceptibility of Chrysoperla zastrowi sillemi (Esben-Petersen) (Neuroptera: Chrysopidae) to certain insecticides (Shankarganesh et al., 2016). Chrysoperla carnea larvae, for instance, exhibit significant tolerance to pyrethroids due to esterase detoxification, and the presence of carboxylesterases has been reported in Chrysoperla spp. (Grafton-Cardwell & Hoy, 1985; Pree et al., 1989). Cytochrome P450 monooxygenases play a central role in the oxidative metabolism of xenobiotic compounds (Feyereisen, 2005) and are typically active in the insect midgut (Brun et al., 1996). Nevertheless, further investigations are required to ascertain the precise role of different enzymes in conferring C. externa resistance to various insecticides.

The presence of lipid droplets and autophagosomes in the digestive cells of C. externa exposed to LC50 thiamethoxam and chlorantraniliprole suggests a potential detoxification mechanism within the midgut cells. Lipids serve as a nutritional reserve and are metabolized by cells for energy production and as signaling molecules (Alberts et al., 2015). Therefore, the occurrence of lipid droplets in the digestive cells may indicate both energy conversion and initial signaling for detoxification processes (Arrese & Soulages, 2010). Autophagosomes, on the other hand, play a crucial role in the turnover of intracellular components, including damaged organelles (Alberts et al., 2015). In insects exposed to pesticide stress, autophagosomes have been observed in the digestive cells and are considered potential organelles involved in detoxification processes (Carneiro et al., 2022; Fiaz et al., 2018; Santos et al., 2018; Serra et al., 2021, 2023).

Conclusion

Larvae and adults exposed to the thiamethoxam and chlorantraniliprole LC50 exhibited midgut epithelium with digestive and regenerative cells that were similar to the control larvae in all post-treatment periods. Both LC50 of the insecticides showed no of histopathological effects on the midgut epithelium and peritrophic matrix, regardless of the duration of exposure. Our results indicate that the LC50 concentrations of thiamethoxam and chlorantraniliprole do not exhibit side effects on the midgut of C. externa larvae and adults. However, additional studies are necessary to evaluate potential side effects on the behavior and reproduction of this predator.