Introduction

Mango (Mangifera indica L.) fruit is only second to banana and pineapple among the most traded tropical fruit globally (RDA 2023, FAO 2024). It is native to India and is mainly cultivated in tropical and subtropical regions, including parts of India, China, Thailand, Mexico, Pakistan, and the Philippines, producing various varieties (Reddy et al. 2018; RDA 2023). India accounts for more than 40% of global production (FAO 2024). In Korea, a country entirely in the temperate zone, the cultivation of tropical and subtropical crops is limited, but the types and cultivation areas of such crops are increasing through greenhouse cultivation. This is due to the decreasing heating costs in winter associated with the continuous rise in temperature caused by global warming, as well as increasing consumer preference for and interest in domestically grown tropical crops (Kim et al. 2008; Jeong et al. 2020). The average temperature in Korea has risen by 1.9°C in the last three decades, from 11.8°C in 1993 to 13.7°C in 2023 (KMA 2024). This period covers the entirety of mango cultivation in Korea. Beginning on Jeju Island in 1993, mango was cultivated on 7.1 ha nationwide in 2001, and by 2022, 869.5 tons were produced on a cultivation area of ​​92.7 ha, representing an approximate 13.1-fold increase in cultivation area (Jeong et al. 2020; RDA 2023). Since 2005, cultivation has been carried out not only on Jeju Island but also in southern regions such as Gyeongsangnam-do and Jeollanam-do, including Yeonggwang, Tongyeong, and Haman (Kim et al. 2019; RDA 2023).

Various pests are known to damage mangoes. Peña et al. (1998) identified 260 economically significant, and minor pest species of mangoes, classifying them into 87 species that damage fruits, 127 that damage leaves, 36 that damage flowers, 33 that damage leaf buds, and 25 that damage the crown and branches. Pests that damage inflorescences and tender shoots mainly include leafhoppers of the family Cicadellidae. Thrips, mango leaf webbers, mealybugs, midges, scale insects, shoot gall psylla, and mites damage foliage and buds; mango shoot borers and mango stem borers damage shoots and stems; and fruit flies, stone weevils, mango pulp weevils, and fruit borers damage fruits (Reddy et al. 2018). Many species of thrips damage mango leaves, flowers, and fruit (Reddy et al. 2018). Among these, the chilli thrips (Scirtothrips dorsalis) are known as an serious pest of mangoes in major mango-producing countries in Asia including Malaysia, Thailand, Philippines, Bangladesh, Pakistan and Indonesia (Waite 2002; Aliakbarpour and Chesalmah 2010; Choi et al. 2013).

In Korea, two species of thrips, including the chilli thrips, along with the green peach aphid (Myzus persicae), the cotton aphid (Aphis gossypii), and the moth Adoxophyes honmai, are known to be the most serious pests in mango. These are known to cause severe damage on the leaves and fruits, including discolored veins of leaves, bumpy fruit surface caused by oviposition scars and discolored and corky fruit surface (Choi et al. 2013). Chilli thrips are positively phototactic and polyphagous, attacking not just mango but over 100 plant species, and they damage a wide range of economically important plants, including peppers, tea, strawberries, kiwifruit, tomatoes, and citrus fruits (Kumar et al. 2012; Dickey et al. 2015). In Korea, 54 plant species in 32 families are documented hosts, and chilli thrips are considered major pests of mangoes, citrus fruits, and grapes (Jeon et al. 2000; Hyun et al. 2012; Choi et al. 2013; Song et al. 2013).

Chilli thrips populations rapidly increase in hot and dry environments. At 30°C, growth from egg to adult takes 20 days, with females laying 50–200 eggs (Reddy et al. 2018). Larvae and adults tear the surface of new leaves or fruits and suck their sap, mainly attacking along veins on the abaxial face of unhardened leaves. Affected leaves and fruits turn brown, and in severe cases, growth is stunted, leaves turn black and dry out, fruits turn black, and fruit deformities occur (Munj et al. 2020; Reddy et al. 2020). This causes direct and indirect damage to mangoes, thereby reducing plant productivity and product marketability (Munj et al. 2020). Chilli thrips can also act as a vector for topoviruses and non-viral diseases (Lahiri et al. 2022). The chilli thrips began to be recognized as a major pest in India, a major mango producing country, in the early 21st century with the emergence of insecticide-resistant species due to increased use of synthetic pyrethroid and neonicotinoid insecticides (Munj et al. 2020). Currently, chemical insecticides are the main control method used for chilli thrips, and in mango cultivation systems, neonicotinoids, organophosphates, and spinosyns are commonly used (Patel et al. 2013). However, the excessive use of chemical pesticides has led to the development of resistance in chilli thrips, making control difficult (Patel et al. 2013; Bana et al. 2015). With rising temperatures due to global warming promoting the continued occurrence of the chilli thrips, the chances of acquiring insecticide resistance, and thus the development rate of resistant populations, are expected to increase (Aristizábal et al. 2017).

In India, resistance of chilli thrips to organophosphate (monocrotophos, acephate, phosalone, dimethoate and triazophos) and carbamate (carbaryl) insecticides has been reported, with LC₅₀ values more than three times higher in resistant populations than in susceptible populations for all tested agents. In Florida, USA, resistance to acetamiprid and cyantraniliprole was confirmed, and resistance to spinetoram was particularly high, with 269-fold higher LC₅₀ values in resistant populations (Vanisree et al. 2011; Kaur et al. 2023). In Korea, research on chilli thrips has mainly focused on occurrence patterns and their ecology with each host. To date, no studies have reported on their insecticide resistance.

Because early response to pesticide resistance is important, continuous monitoring is a necessity. There are several experimental techniques for assessing insecticide resistance, including bioassay methods, immunological methods, and molecular methods. By assessing insect responses to insecticides through bioassays and conducting molecular experiments, the level of insecticide resistance and the mechanism of resistance can be clearly identified (Collins and Schlipalius 2018). Therefore, this study was conducted to evaluate the insecticide resistance of chilli thrips occurring in mango cultivation areas where domestic cultivation is expanding, aiming to obtain basic data for effective management.

Materials and Methods

Collecting chilli thrips

Chilli thrips were collected in major mango growing areas in Korea, including the cities of Seogwipo and Jeju on Jeju Island, Haman and Tongyeong in Gyeongnam, and Haenam and Jangheung in Jeollanam-do. To collect the insects, blackened and dried new leaves damaged by the chilli thrips before unfolding and severely damaged mature leaves that had turned black were collected and shaken by hand to separate the thrips. The species were identified based on their morphological traits using an optical microscope (SMZ1000, Nikon, Japan) (Chandra and Verma 2010; Kumar et al. 2011; Gopal et al. 2018).

Experimental insecticide

There are 141 pesticides registered for use against mango chilli thrips in Korea (RDA 2022), and eight single-agent pesticides from different IRAC (Insecticide Resistance Action Committee) mode-of-action groups were used in the experiment (Table 1).

Concentration range calibration bioassay

The chilli thrips insecticide response experiment was conducted in two stages. In the first stage, a concentration calibration bioassay was conducted to secure basic resistance information to determine the optimal concentration range for calculating LC₉₀ values, and in the second stage, an experiment was conducted by serially diluting the insecticides to produce test concentrations based on the results of the first experiment. The concentration range calibration bioassay was performed by diluting each insecticide to half the manufacturer’s recommended concentration. The bioassay was performed using the leaf immersion method (IRAC, Test method video No. 10), which is suggested by the IRAC as an appropriate bioassay method for the yellow flower thrips (Frankliniella occidentalis).

Young mango leaves that were free of pest damage and had not been exposed to pesticides within the past 3 weeks were collected, cut into 3 cm pieces, soaked in distilled water, for the control group, or distilled water containing the specified pesticide concentration for 30 s, and then dried in the shade in the laboratory for 30 min. Five sheets of filter paper (90 mm, Model No. 2, ADVANTEC, Japan) soaked in distilled water were placed on the bottom of an insect breeding dish (100 × 40 mm; Model No. 310102, SPL, Korea), and a 5 × 5 cm sheet of parafilm was placed in the center. The prepared leaves were then placed on the parafilm with the back of the leaves facing upward using tweezers, and 20 adult chilli thrips were placed in the dish. The dish was covered with a hand towel (Model No. 47201, Yuhan-kimberly, Korea), the lid was affixed over the towel, and the dish was incubated in a growth chamber (HB-303DH-O, Hanbaek scientific co., Korea) at 25 ± 1°C and 60% humidity with a 16:8 light–dark cycle for 72 hours. The experiment was performed three times for each pesticide and concentration. After 72 hours, mortality was assessed, and individuals that failed to respond when touched with a brush were judged to be dead.

Insecticide resistance bioassay

Based on the results of the concentration range calibration bioassay, the pesticides were serially diluted to test concentrations designed to produce lethalities ranging from 0% to 20% for concentration 1, 20% to 40% for concentration 2, 40% to 60% for concentration 3, 60% to 80% for concentration 4, and 80% to 100% for concentration 5. Aside from the concentrations tested, the experimental method was identical to that used in the calibration bioassay, with three replicates for each concentration. With no studies reported on insecticide resistance of S. dorsalis in Korea, there is no established susceptible strain of the pest. Thus, resistance was evaluated using resistance level (RL) values. The RL value (based on the same concept as the CEI [control efficacy level]) was calculated for each pesticide by dividing the 90% lethal dose (ppm) by the recommended dose (ppm): RL = LC₉₀ / recommended concentration (see Kang et al. 2023, 2024).

Statistical analysis

Mortality data from the calibration assay were arcsine transformed and subjected to an analysis of variance (SAS/STAT® 9.3 user’s guide 2011), with collection region, pesticide, and concentration as main factors and interactions involving the collection region. Tukey post-hoc tests were conducted to compare individual means among the main factors and interactions, with significance set at p < 0.05. To evaluate the RL, the mortality results for each concentration were used to obtain the LC₉₀ value using Probit analysis (SAS/STAT® 9.3 user’s guide 2011), which was then used to calculate the RL.

Results

Concentration range calibration bioassay

The insecticide responses of mango chilli thrips varied with region, insecticide, and concentration, and there were interactions among factors (df = 143, 288, F = 45.01, p < 0.0001), with the collection region affecting responses to insecticides, concentration, and the insecticide-dependent response to concentration (i.e., region × insecticide × concentration) (Table 2). Depending on the type of insecticide, efficacy varied. In the case of abamectin, efficacy differed by region, and insecticide efficacy by concentration also differed by region (df = 17, 36, F = 31.78, p < 0.0001) (Table 3). When treated with the recommended dose, the Haenam region population showed the highest mortality rate, 93.1%, and the Jeju region showed a mortality rate of 86.1% (Table 3). In other regions, the mortality rate was below 80%, with the Haman region population showing the lowest, 12.3% (Table 3). Overall, abamectin induced mortality in a concentration-dependent manner, but concentration had no statistically significant effects in the Seogwipo population (Table 3).

The effects of bifenthrin also differed depending on the region and concentration (df = 17, 36, F = 13.34, p < 0.0001). All regional strains exhibited mortality rates below 50% at the recommended dose. Notably, no mortality was recorded in the Haman region population even when treated with double the recommended dose (Table 3). The Haenam, Seogwipo, and Tongyeong populations experienced similar mortality rates under the recommended and double-dose treatments, but with the half-dose treatment, no mortality was recorded in the Seogwipo region population, a lower mortality rate than seen in the other two regions (Table 3). Similarly, region significantly affected mortality rates in the chlorfenapyr treatments. At the recommended dose, only the Jeju population experienced 100% mortality, while the Tongyeong population showed the lowest mortality rate at 17% (df = 17, 36, F = 26.33, p < 0.0001) (Table 3). In Haenam and Jeju populations, 100% mortality was observed when the recommended dose was doubled, but at the recommended dose, the Haenam population showed a lower mortality rate of 84.9% (Table 3). In the Haman population, there was no significant difference in mortality between the double-dose, recommended-dose, and half-dose treatments (Table 3).

When treated with cyantraniliprole, different insecticide responses were again observed by region, with populations from the Haenam, Jangheung, and Seogwipo regions showing higher mortality rates than those from Haman, Jeju, and Tongyeong (df = 17, 36, F = 24.56, p < 0.0001) (Table 3). At the recommended dose, mortality rates of over 80% were observed only in Haenam and Seogwipo region populations, while the Tongyeong region population showed a low mortality rate of 8.3% (Table 3). Dinotefuran produced low mortality in all regional populations. At double the recommended dose, mortality rates of 60.7% and 33.3% were seen in the Haman and Jeju populations, respectively, with mortality rates below 20% in all other regions (df = 17, 36, F = 3.96, p = 0.0014) (Table 3). Emamectin benzoate’s efficacy was affected by both region and treatment concentration, but showed high overall mortality rates (df = 17, 36, F = 19.75, p < 0.0001) (Table 3). At the recommended dose, only the Jangheung and Tongyeong populations showed mortality rates below 95%, experiencing 85.0% and 86.3% mortality, respectively, while the Haenam, Haman, and Seogwipo populations showed 100% mortality (Table 3). The Haman and Seogwipo populations were particularly susceptible, with 100% mortality even at half the recommended concentration, and the Haenam and Jeju populations showed mortality rates above 95%, also demonstrating high susceptibility (Table 3).

Like Emamectin benzoate, spinetoram’s effects differed depending on the region and concentration but induced high mortality rates overall (df = 17, 36, F = 4.66, p < 0.0001) (Table 3). The Seogwipo region strains showed 100% mortality at all three tested concentrations, and the Haenam, Haman, and Jangheung region populations exhibited mortality rates over 91.5%, even under half-dose treatments (Table 3). For the Jeju population, treatment at double the recommended dose induced a mortality rate of 100%, but at the standard dose and half the standard dose, relatively low mortality rates in the 80% range were recorded (Table 3). Thiacloprid also differed in efficacy among regions and treatment doses, but showed low overall insecticidal activity (df = 1 7, 36, F = 6.64, p < 0.0001) (Table 3). All regional strains showed mortality rates below 30% at the recommended dose, and the Seogwipo and Tongyeong populations experienced no mortality at half the recommended dose (Table 3).

Insecticide resistance assessment against the chilli thrips

To quantify the insecticide resistance of chilli thrips, the LC₅₀ and LC₉₀ values for each pesticide in each region were calculated, and the RL (resistance level) value ​​was calculated by dividing the LC₉₀ value for each pesticide by its recommended concentration (Table 4). Chilli thrips showed high levels of resistance to abamectin in the Seogwipo and Haman regions, with RL values ​​of 13.89 and 22.16, respectively, and the Jeju populations showed the lowest RL, 0.8 (Table 4). For bifenthrin, although there were differences in the RL by region, the average LC₉₀ value was four times higher than the recommended concentration. In particular, the Haman region strain showed an extremely high resistance, with an RL of over 150,000, and the Seogwipo region population exhibited high resistance, with an RL value of 30.78 (Table 4).

For chlorphenapyr, the highest RL value, 10.02, was found in the Tongyeong population, but the Haman and Jeju populations showed LC₉₀ values lower than the recommended dose. With cyanthraniliprole, a very high RL value, 92.46, was seen in the Tongyeong population, but RL values ​​below 2 were found in the Seogwipo and Jeju area populations. Overall, chilli thrips showed high overall resistance to dinotefuran, and even the most susceptible regional population, that of Haman, exhibited an RL of 6.77. On the other hand, emamectin benzoate showed high efficacy against all regional strains, with RL values ​​below 1 in all populations except that of Jangheung. Similarly, chilli thrips showed very low levels of resistance to spinetoram, with RL values​​ below 1 in all regional populations. Lastly, thiacloprid performed poorly, with thrips showing high RL values, 10 or higher, in all regional populations, with the Haenam population exhibiting the highest at 427.88 (Table 4).

Discussion

Organophosphate, neonicotinoid, pyrethroid, spinosyn, avermectin, and pyrrole insecticides have been reported to be effective in controlling chilli thrips (Shibao 1997; Seal and Kumar 2010). In addition, acetamiprid, clothianidin, thiamethoxam, and imidacloprid were shown to be highly effective in preventing chilli thrips damage (Patel and Kumar 2017), and spinetoram, cyantraniliprole, tolfenpyrad, chlorantraniliprole, and thiamethoxam were found to be effective for control (Kumar et al. 2017). On the other hand, studies on chilli thrips in India and Florida, USA, reported resistance to neonicotinoid, diamide, spinosyn, and pyrethroid insecticides (Kumar et al. 2017; Kaur and Lahiri 2022). In line with the findings of previous studies conducted in other countries, the results of the current study showed high levels of resistance to neonicotinoids and pyrethroids. We found high levels of resistance against the neonicotinoids thiacloprid and dinotefuran and the pyrethroid bifenthrin in all chilli thrips populations from the six regions. Although there were differences in mortality responses among local populations, spinetoram and emamectin benzoate showed high overall control efficacy and low resistance indices. The other three tested pesticides, abamectin, chlorphenapyr, and cyantraniliprole, showed regional differences in efficacy, with resistance differing greatly among regions. In particular, cyantraniliprole showed large differences among regions, with the Tongyeong population showing a high RL value, 92.46, but the Jeju and Seogwipo populations showing low ones, 1.44 and 1.84, respectively, thus exhibiting over a 60-fold difference between regional populations. On the other hand, spinetoram, which belongs to the spinosyne family of pesticides, showed high and consistent control effects, producing mortality rates of over 90% at the recommended concentration in all regions.

Among chilli thrips collected from strawberry farms in Florida, USA, Kaur et al. (2023) reported spinetoram resistance ratios varying from 6 to 269 among farms. In Kaur et al. (2023), susceptible strains collected from fields where no pesticides had been applied exhibited LC50 and LC90 values ​​of 0.026 ppm and 8.64 ppm, respectively. In the current study, the LC50 values ​​for the regional populations were 0.73–3.83 ppm, which were higher than the LC50 values of the susceptible strains in Kaur et al. (2023), but the LC90 values ​​in the populations from Haman, Seogwipo, and Tongyeong—2.31, 4.61, and 7.55 ppm, respectively—were lower than the LC90 vlaues of the susceptible strains. The RL value of the chilli thrips against spinetoram was less than 1 for all eight cities and counties, indicating that insecticide resistance to spinetoram has not yet developed in the mango growing areas of Korea. However, as the application of these pesticides increases, as occurred in the Florida strawberry fields, the emergence of resistance is a serious concern. Therefore, while the eight pesticides used in this study are currently safe to use in local populations, cautious resistance management will need to be applied in the future.

The degree of pesticide resistance is estimated by comparing the responses of individuals collected in the field with those of susceptible populations reared in the laboratory or collected in the field, using the resistance factor (RF) or resistance ratio (RR) as a measure (Rao et al. 2019; Kaur et al. 2023). Thus, the RF and RR may be unreliable or impossible to calculate when an incorrect susceptible strain or population is used or none is available. This also raises concerns about erroneously meeting or failing to meet the resistance judgment criteria presented by the IRAC (Kang et al. 2024). Notably, categories for classifying the degree of resistance based on RR or RF values ​​differ among researchers, resulting in different resistance assessments (Ahmad et al. 2007; Stará et al. 2023). This has led to the development of alternative resistance assessment methods based on efficacy at recommended concentrations, including the control efficacy index (CEI), security index (SI), and pesticide efficacy index (PEI) (Sauphanor et al. 1998; Jeong et al. 2017; Stará et al. 2023; Kang et al. 2024). These indices are calculated based on the LC90 value of the field population and the recommended concentration for each pesticide. The CEI is derived by dividing the recommended concentration of the pesticide by the LC90 value of the field strain, and can be subdivided into sections to evaluate the degree of resistance, like the RR value. However, the SI index, which is the LC90 value of the field strain divided by the recommended concentration of the pesticide, has the limitation that this type of resistance subdivision cannot be applied (Kang et al. 2024). The RR and CEI are indices defined by an independent relationship rather than a linear relationship. The CEI is suggested to be suitable for monitoring resistance in field pest populations (Kang et al. 2024), and it has been used to do so with Frankliniella occidentalis, Aphis gossypii, and Bemisia tabaci (Choi et al. 2023; Lee et al. 2023; Kang et al. 2024; Kim et al. 2024). However, the term CEI frames the degree of resistance based on a derived idea, so it is considered more desirable to use the term RL, and so it is expressed as RL in this study. If the RL value is less than 1, the organism is considered to be sensitive, if it is greater than 1 and less than 5, the organism is considered to be moderately resistant, and if it is greater than 5, the organism is considered to be highly resistant (Kang et al. 2023).

Using the above criteria and based on the level of concern about resistance, each pesticide was categorized as either a pesticide of concern in all regions, a pesticide of concern only in specific regions, or a pesticide with no concern about resistance in any region. Bifenthrin, dinotefuran, and thiacloprid were pesticides of concern in all study areas, and high resistance was confirmed in all regions for dinotefuran and thiacloprid. With bifenthrin, intermediate resistance was found in two regions, but there were also areas, such as Haman, where extremely high resistance was detected. On the other hand, resistance to some pesticides showed highly regional patterns. For example, abamectin encountered high resistance in two regions, while chlorphenapyr encountered high resistance in only one. For pesticides facing high resistance regardless of the region, we believe it will be necessary to select other types of pesticides that are generally more effective for use in systemic treatments. For pesticides facing resistance in only certain areas, it is thought that management will be necessary to prevent further deterioration of their effectiveness, such as the use of emamectin benzoate.

Chemical means are effective in controlling target pests in the short term, but there is a persistent problem of resistance development (Seal et al. 2006; Chung and Park 2009; Rao et al. 2019). To prevent this, a method of alternating pesticide applications is being used, and recently, studies are being conducted to introduce an integrated pest management method that applies chemical control along with agronomic and biological control to reduce opportunities for resistance acquisition and slow the rate of its development (Chung and Park 2009; Rao et al. 2019). Seal and Kumar (2010) recommended the alternate spraying of 12 chemical and biological control agents, considering the system and duration of their effect, for the control of the chilli thrips, including abamectin, chlorfenapyr, dinotefuran, and spinetoram, which were used in this study. Among the selected agents, imidacloprid, which belongs to group 4A of the IRAC insecticide classification, and spinetoram, which belongs to group 5, have a residual period of 15 days against chilli thrips larvae or adults on leaves or in the soil, which are long-term effects compared to those of other types of insecticides (Seal and Kumar 2010), so it is thought that they will be relatively useful in alternating spraying programs for the control of chilli thrips.

For the effective management of chilli thrips, keeping their population below a threshold level is necessary, and for such management, systematic treatment with insecticides from different families, eliminating concerns of cross-resistance, is essential, and utilizing natural enemies or biological control agents in the control system can be an alternative. However, to effectively implement such control systems, it is essential to monitor the pest’s level of resistance to each pesticide at the currently used concentration (Rao et al. 2019). Particularly with pesticides whose resistance can increase rapidly from year to year. Additionally, continuous resistance monitoring is deemed necessary for pests with fast generational turnover and limited registered pesticides, which is exactly the case for chilli thrips in mango fields. In addition, molecular markers that can be practically applied to quickly and efficiently identify pesticide-specific resistance would be a great benefit, and research in this area should also be conducted.

Conclusions

We assessed for the first time the status of insecticide resistance in chilli thrips occurring in mango cultivation areas in Korea. Responses to eight pesticides varied depending on the region, and chilli thrips populations were found to be sensitive to spinetoram in all surveyed areas, with LC90 values below the recommended dose of the pesticide. Emamectin benzoate was effective in all regions except one, while high resistance to dinotefuran and thiacloprid was recorded in all areas. Intermediate to high resistance was recorded for Cyanthraniliprole in all regions, but an exceptionally high RL of 92.46 was seen in one region. Similarly, bifenthrin also encountered intermediate to high resistance in all regional populations, but the extreme RL of 151251 was recorded in one area. For abamectin and chlorphenapyr, susceptible, intermediate, and highly resistant populations were variably encountered among the regions. It is believed that the control effect of bifenthrin, dinotefuran, and thiacloprid, for which resistance is relatively high in all regions, should be replaced by insecticides of different types. For abamectin or chlorfenapyr, for which resistance varies by region, judicious use guided by continuous resistance monitoring is required. Ongoing resistance management will be necessary through systematic treatments employing insecticides with low local resistance development.