Introduction

The western flower thrips, Frankliniella occidentalis (Pergande), represents a significant global threat to agricultural and horticultural crops. Its adverse impact includes reduced crop yields due to direct feeding damage and the transmission of tospoviruses, such as impatiens necrotic spot virus and tomato spotted wilt virus, affecting a wide array of crops (Prins and Goldbach 1998). Managing F. occidentalis is particularly challenging because of its propensity to develop resistance to chemical insecticides (Gao et al. 2012; Demiroz and Kumral 2022; Yuan et al. 2023). Moreover, its thigmotactic behavior complicates the effective application of chemical insecticides (Sarkar et al. 2019). In light of the environmental concerns associated with chemical insecticides, the quest for alternative methods to manage thrips infestations has become imperative, with a growing focus on biological control agents.

Vietnam, like other nations in South-East Asia, aspires to become a prominent producer and exporter of high-quality agricultural products (Nguyen et al. 2020). Over the past few years, the Vietnamese government has put considerable effort into the promotion and implementation of various campaigns aimed at reducing pesticide usage (Pham et al. 2016). Thrips, including F. occidentalis, which are key agricultural pests known to quickly develop resistance to pesticides, are among the focal pests in this context (Poushkova and Kasatkin 2020). Notably, thrips management in Vietnam has predominantly relied on intensive insecticide use. However, rising awareness among consumers and authorities (Mai et al. 2019) about the potential health risks associated with pesticide residues on agricultural products has prompted a shift towards biological control methods as a more sustainable alternative to chemical approaches (Pham et al. 2013). Recent studies in Vietnam have demonstrated the potential of predatory mites from the Phytoseiidae family (Acari) to serve as effective biological control agents against F. occidentalis (Nguyen et al. 2019a, b).

The Phytoseiidae family encompasses over 2500 valid species worldwide (Demite et al. 2014, 2020), all of which are predators. Several species within this family hold economic importance as natural enemies, extensively employed in the biological control of various insect and mite pests (McMurtry and Croft 1997; McMurtry et al. 2013). In the present study, we set out to construct life tables of three phytoseiid species capable of preying on the early immature stages of F. occidentalis. These include two native species to Vietnam, Proprioseiopsis lenis (Corpuz and Rimando) and Amblyseius largoensis (Muma), along with a non-native species commercially available in the region, Amblyseius swirskii Athias-Henriot.

Amblyseius swirskii, initially identified in Israel in 1962, has since spread to 22 countries and has been available on the market since 2005, now being sold in over 50 countries (Calvo et al. 2015; Tixier et al. 2022). This versatile generalist predator feeds on various prey species and pollen (McMurtry et al. 2013; Alipour et al. 2016; Riahi et al. 2017) and has been observed on 48 plant families, with the highest incidence on Rosaceae, Rutaceae, and Solanaceae, including both cultivated crops and wild plants (Tixier et al. 2022). Amblyseius swirskii is an effective biological control agent of thrips (Messelink et al. 2006; Chow et al. 2010), spider mites (Momen 2009), and whiteflies (Nomikou et al. 2001, 2004; Messelink et al. 2008). The species can be mass reared cost effectively and in the field its populations can be supplemented with cattail pollen or astigmatid mites, enhancing its usefulness for augmentative releases (Goleva and Zebitz 2013; Leman and Messelink 2015; Pijnakker et al. 2016).

Amblyseius largoensis, on the other hand, has been identified in tropical regions across the Americas, on Indian Ocean islands, and more recently in Southeast Asia (Gallego et al. 2003; Liao et al. 2013, 2021; Da Silva et al. 2015; Nguyen et al. 2019a). It has been noted as the predominant predator linked to the red palm mite Raoiella indica (Hirst) infesting coconuts in Florida, Puerto Rico, and Trinidad and Tobago (Peña et al. 2009). Additionally, it exhibits a diverse diet in the field, preying on various arthropods such as Eriophyes mites (Kamburov 1971; Galvão et al. 2007), thrips (Nguyen et al. 2019a), rice moths (Nasr et al. 2015), fruit flies (Momen et al. 2016), and whiteflies (De Alfaia et al. 2018 ).

Proprioseiopsis lenis, originally discovered on Citrus nobilis (Rutaceae) in the Philippines by Corpuz and Rimando (1966), is naturally distributed in New Zealand (Schicha and Elshafie 1980). Subsequently, it has been recorded in Australia (Navasero and Corpuz-Raros 2005), Thailand (Oliveira et al. 2012), and more recently in Vietnam (Nguyen et al. 2019a). Researchers have suggested that P. lenis holds the potential to effectively control outbreaks of spider mites and thrips in South-East Asia.

Previous studies on the biological control potential of phytoseiid predators against F. occidentalis have predominantly focused on the mobile first instars of the prey. However, the larvae of this thrips species remain in the first instar for only a brief period, typically no longer than 24 h. Thereafter, they become more difficult to exploit as prey by most phytoseiid mites. However, compared to predation on thrips larvae, relatively less attention has been directed towards predation on thrips eggs. The significance of egg predation was suggested by Vangansbeke et al. (2018); Nguyen et al. (2019b), who demonstrated that phytoseiid mites can effectively locate and feed on eggs of F. occidentalis. Thus, the objective of the present life table study is to compare the development and reproduction of A. swirskii, A. largoensis, and P. lenis when preying on eggs versus first instars of F. occidentalis, with a view to assessing the potential roles of these phytoseiids in managing thrips outbreaks in South-East Asia.

Materials and methods

Colony of prey

Western flower thrips, F. occidentalis, were reared at the Department of Plants of Crops of Ghent University (Belgium) on bean pods (Phaseolus vulgaris Prelude) and fresh pollen of cattail, Typha latifolia L. (Nutrimite, Biobest Group, Westerlo, Belgium) in plastic boxes (30 cm × 20 cm x 8 cm). The rearing containers were kept in a climate chamber (Panasonic MLR 352 H) set at 25 ± 1 °C and 65 ± 5% relative humidity (RH) and a 16:8 h (L:D) photoperiod. Bean pods and pollen were replaced weekly. First instar larvae and adults of F. occidentalis were used for the experiments.

Colony of predatory mites

Proprioseiopsis lenis and A. largoensis were collected in Vietnam, and colonies were established at Ghent University in 2016. Amblyseius swirskii was provided by BioBest Group NV (Westerlo, Belgium). All predatory mites were reared on arenas made from bean leaves (P. vulgaris). These leaves were placed upside down on a 2 cm thick layer of wet cotton in plastic trays (20 × 13 × 5 cm). To prevent mites from escaping, a thin layer of tissue paper was placed on the edges of the leaves and kept moist. First instars of F. occidentalis were provided as the sole prey on a daily basis. Each leaf disc was utilized for two weeks, after which the mites were transferred to fresh leaf discs. All phytoseiids were reared in a Panasonic climate chamber (MLR 352) set at 25 ± 1 °C, 65 ± 5% RH and a 16:8 h (L:D) photoperiod.

Experimental setup

To investigate the development and reproduction of individual predatory mites, plastic dishes (5 × 1.5 cm) were used as rearing microcosms. The bottom of each dish was lined with a 5 × 0.3 cm moist cotton layer, on top of which a 4 cm diameter bean leaf section was placed. The edges of the leaves were covered with soft paper tissue, which was moistened daily to prevent mite escape. Approximately 12 h before the start of each experiment, small pieces of white cotton thread were placed in the stock colony of predatory mites to serve as oviposition substrates. Individual eggs deposited on the cotton were transferred to the rearing microcosms, with each treatment initiated using sixty eggs. For experiments involving larvae of F. occidentalis as prey, 15 first instars (less than 12 h old) were introduced into each rearing microcosm using a fine brush. This process was repeated every 24 h, at which point any surviving larvae from the previous day were removed. For experiments using eggs of F. occidentalis as prey, five female thrips were introduced into each rearing microcosm 24 h before introducing the predator to allow for egg laying. The mites were transferred daily with a fine brush to new microcosms containing fresh thrips eggs. To gather data on the duration of each developmental stage of the predatory mites, as well as on mortality and escape rates, observations were made every 24 h until all individuals had reached adulthood. Any escaped predatory mites were excluded from data analysis. The appearance of an exuvia in the rearing microcosm marked the first day of a new developmental stage for each individual predator. Upon completing immature development, each female phytoseiid was immediately paired with a male that had been reared on the same diet. This male was removed after 24 h. Only a singly mating was allowed in order to exclude variable effects of food competition and further mating events on the female’s reproductive output. Adults were observed daily to determine the preoviposition and oviposition periods, oviposition rate, and longevity. The oviposition rate was calculated by dividing the total number of eggs by the number of days during which oviposition occurred. The oviposition period was calculated as the time from the first egg laid until the last egg laid. Progeny from females of the same age were transferred to new rearing microcosms and provided with the same diet as their parents in order to determine the sex ratio. All experiments were conducted in a Panasonic climate chamber (MLR 352) set at 25 ± 1 °C, 65 ± 5% RH, and a 16:8 h (L: D) photoperiod.

Life table parameters

Life table parameters were calculated for P. lenis, A. largoensis, and A. swirskii when presented with eggs or first instars of F. occidentalis. The intrinsic rate of increase (rm) was calculated using the formula developed by Lotka (1907) and Birch (1948):

$$ \sum {l}_{x}{m}_{x} {e}^{-{r}_{m}*x} =1$$

where x is the female age (in days), lx is the age -specific survival of females at age x, and mx is the number of female progeny produced per female at age x. This value is obtained by multiplying the mean number of eggs laid per female by the proportion of female offspring produced at age x. To estimate the standard error of rm, we utilized the Jackknife procedure following the methodology of Meyer et al. (1986) and Hulting et al. (1990). The generation time (T), net reproductive rate (R0), finite rate of natural increase (λ), and doubling time (DT) were calculated according to Maia et al. (2000). Generation time (T) signifying the time span (in days) between the first egg laid in one generation and the next, was formulated as:

$$ \text{T}=\frac{\sum x{l}_{x}{m}_{x}}{\sum {l}_{x}{m}_{x}}$$

The net reproductive rate (R0) measures the mean number of female offspring produced per female (females/female) and was computed using the following formula:

$$ {R}_{0}=\sum {l}_{x}{m}_{x}$$

Finite rate of natural increase (λ) quantifies the population growth rate over a unit of time and was assessed as:

$$ {\lambda }=\text{a}\text{n}\text{t}\text{i}\text{l}\text{o}\text{g} {e}^{{r}_{m}}$$

Doubling time (DT) represents the number of days required for a population to double in size and was calculated using the formula,

$$ \text{D}\text{T}=\frac{\text{l}\text{n}\left(2\right)}{{r}_{m}}$$

Statistical analysis

Data were subjected to statistical analysis (R_Studio, version 1.1.453, 23 April 2018) to analyse the effect of diet on the duration of the immature stages, preoviposition and oviposition period, daily and total oviposition, and adult longevity of P. lenis, A. largoensis and A. swirskii. Kolmogorov–Smirnov tests indicated that none of the data were normally distributed. Therefore, nonparametric Kruskal–Wallis ANOVAs were used to analyse the data and means were separated using the Mann–Whitney U test. Sex ratios of the progeny were compared by means of a logistic regression. This regression is a generalized linear model using a probit (log odds) link and a binomial error function. P-values smaller than or equal to 0.05 were considered significant.

Results

The immature survival rate of A. largoensis, P. lenis, and A. swirskii when provided with a diet of thrips eggs ranged from 91.67 to 96.67%, whereas when they were offered thrips larvae as prey, their survival rates showed a range from 70 to 83.33%. The survival rate of P. lenis on both thrips larvae and eggs (70 − 91.67%) was consistently lower than that observed in the other two species. Above survival rates take into consideration both dead and escaped individuals, so we further separated effects of diet on these categories. Death rate ranged from 13.33 to 20% when the phytoseiids were offered larvae of F. occidentalis, which was higher than that of their counterparts fed on thrips eggs, ranging from 0 to 1.67%. When the prey was thrips larvae, the escape rate was between 3.33% and 10%, which was similar to when thrips eggs were offered as food, with a range of 3.33–6.67%. When the food consisted of thrips eggs, the overall mortality rate of the phytoseiids (0.56%) was lower compared to the escape rate (5%). However, when the prey was thrips larvae, the overall mortality rate of the predatory mites (16.67%) was higher than the escape rate (7.22%).

The total developmental time of P. lenis and A. swirskii females on either diet did not differ (Table 1). However, the protonymphal stages of these mites took comparatively more time to develop when feeding on thrips larvae than when provided with thrips eggs. Interestingly, when A. swirskii was fed on thrips eggs, its eggs took longer to hatch. Diet significantly influenced the immature development of A. largoensis females, with the exception of the larval stage. Females of A. largoensis that consumed thrips eggs completed the immature stages in 6.51 days, whereas it took them 8.61 days when fed on thrips larvae. Male phytoseiids maintained on larvae of F. occidentalis generally took longer to develop than their female counterparts. For A. swirskii males, there was no significant difference in development time when they were fed either thrips eggs or larvae throughout all stages of development. On the contrary, males of A. largoensis displayed significant differences in all stages of development except for the larval stage. Similarly, in P. lenis males, differences were evident in the deutonymph and total immature stages only. Males of A. largoensis and P. lenis required more time to reach the adult stage when feeding on thrips larvae (8.47 and 9.00 days, respectively) compared to when provided with thrips eggs (7.05 and 8.10 days, respectively) of F. occidentalis. After mating, the influence of diet extended to nearly all reproductive parameters and adult longevity (Table 2). When A. swirskii was fed on thrips eggs versus larvae, the preoviposition period, oviposition period, female longevity, total number of eggs, and female proportion differed significantly. On a diet of thrips larvae, A. swirskii exhibited a longer oviposition time (16.04 days), an extended female longevity (20.44 days), and a higher egg production rate (31.93 eggs per female) compared to when fed on thrips eggs.

Table 1 Developmental time of the immature stages of three predatory mites fed on the eggs or first-instar larvae of F. occidentalis at 25 ± 1°C and 65 ± 5% RH
Full size table
Table 2 Fecundity and longevity of A. largoensis, P. lenis and A. swirskii fed on eggs and first-instar larvae of F. occidentalis at 25 ± 1°C and 65 ± 5% RH
Full size table

In contrast, prey type had a less pronounced effect on the fecundity and longevity of A. largoensis. While females of this species displayed a shorter longevity (15.71 versus 17.77 days) when reared on thrips eggs compared to thrips larvae, they exhibited a higher oviposition rate (2.19 versus 1.83 eggs per female per day) when consuming eggs of F. occidentalis. Proprioseiopsis lenis was the least influenced by adult diet. Among all parameters, only the oviposition rate showed a significant difference when the predatory mites were fed on thrips eggs versus larvae, with rates of 1.78 eggs per female per day and 1.92 eggs per female per day, respectively. The sex ratio of offspring was significantly influenced by diet only in A. swirskii, which exhibited a female proportion of 65.65% on thrips eggs compared to 59.47% on thrips larvae. Life table parameters appeared to be more strongly affected by diet in A. swirskii and A. largoensis than in P. lenis (Table 3). A. largoensis stood out with the highest value of intrinsic rate of increase (0.209) when thrips eggs were provided as food to the phytoseiids. However, when mobile first instar thrips were the prey, A. swirskii exhibited a higher intrinsic rate of increase (0.190) than the other two predatory mites. Conversely, P. lenis consistently demonstrated the lowest value of intrinsic rate of increase on both prey types.

Table 3 Life table parameters of A. largoensis, P. lenis and A. swirskii fed on eggs and first-instar larvae of F. occidentalis at 25 ± 1°C and 65 ± 5% RH
Full size table

Discussion

In the present study, A. largoensis, A. swirskii as well as P. lenis successfully developed and reproduced on both eggs and first instars of F. occidentalis. This finding is not in line with the outcome of an earlier study by Nguyen et al. (2019b), in which adult females of P. lenis were rarely observed to feed on thrips eggs. We hypothesize that the lack of exposure to thrips eggs during the immatures stages and the high density of adult predators in a small experimental arena impacted the ability of P. lenis females to find and feed on thrips eggs in the Nguyen et al. (2019b) study.

The statistical analysis indicated significant differences among egg incubation times, although these differences were minor. This variation in egg development time is unlikely a result of treatment effects and may be linked to slight differences in egg collection times.

When offered larvae of F. occidentalis in the present study, A. swirskii exhibited the highest intrinsic rate of increase, the highest fecundity, and the shortest juvenile developmental time, followed by A. largoensis and P. lenis. In our laboratory experiments, death and escape rates of A. largoensis, P. lenis, and A. swirskii preying on thrips larvae ranged from 13.33 to 20% and 3.33–10%, respectively. In previous studies using thrips larvae as prey reported survival rates varied considerably depending on the phytoseiid and thrips species involved, averaging only 13% for Iphiseius degenerans (Berlese) with F. occidentalis (Vantornhout et al. 2004), 38% for Neoseiulus californicus (McGregor) with F. occidentalis (Walzer et al. 2004), 62–88% for Neoseiulus cucumeris (Oudermans) with onion thrips Thrips tabaci (Lindeman) and F. occidentalis, respectively (Marisa and Sauro 1990), and 91% for P. lenis with F. occidentalis (Nguyen et al. 2019a).

The developmental time of A. swirskii in our study closely aligns with that reported by Wimmer et al. (2008) when the phytoseiid was offered larvae of F. occidentalis and T. tabaci under similar experimental conditions. The developmental time of A. swirskii males and females when fed F. occidentalis larvae (7.6–7.7 days) was longer than that observed when preying on the two-spotted spider mite Tetranychus urticae Koch (5.83–6.03 days) (Hosseininia et al. 2020) at 25 °C, but the intrinsic rate of increase and fecundity did not differ significantly. This may be related to the lower nutritional value of F. occidentalis larvae compared to other food sources such as pollen, whiteflies, or mixed diets (Nomikou et al. 2001, 2004; Wimmer et al. 2008), potentially influencing the life history parameters of A. swirskii. The duration of the A. largoensis juvenile stages in the current study (8.47–8.61 days) was longer than that reported on mixed stages of the red palm mites R. indica (5.92 days) and of Tetranychus gloveri Banks (7.11 days) on coconut (Carrillo et al. 2010) at 26.5 °C and 70% RH, and was shorter than that of those fed on the armored scale Aonidiella orientalis Newstead (11.43 days), or the mealybug Nipaecoccus nipa Maskell (13.5 days). On the other hand, the intrinsic rate of increase of A. largoensis (rm = 0.186) was significantly higher in the present study than when it was fed on R. indica (rm = 0.127) or T. gloveri (rm = 0.102). Conversely, when consuming mixed life stages of T. urticae (4.00-4.31 days and 11.34 days, respectively) or the first and second instars of F. occidentalis (4.09–4.34 days and 13.47 days, respectively), P. lenis completed the immature stage faster and had shorter longevity (Nguyen et al. 2019a) than when preying on thrips larvae (7.86-9.00 days and 16.55 days, respectively) in this study. As temperature is a key factor affecting growth and development of predatory mites, it may be in part responsible for the differences among studies (Kolokytha et al. 2011; Stavrinides and Mills 2011).

Jacobson et al. (2001) were the first to suggest that the phytoseiid N. cucumeris could prey on both thrips eggs and first instars of F. occidentalis, although they did not provide concrete evidence of this predation on eggs. Subsequent studies by Vangansbeke et al. (2018) proposed that A. swirskii could effectively control thrips eggs, substantiated by the reduced number of F. occidentalis larvae hatching in leaf arenas with predatory mites compared to arenas without them. This hypothesis gained further support from Nguyen et al. (2019b), who detected thrips DNA in the guts of phytoseiid mites that foraged for 2 h on bean leaves containing thrips eggs.

In our study, A. largoensis had the highest intrinsic rate of increase value, produced the highest total number eggs and had the shortest juvenile developmental time when offered thrips eggs, followed by A. swirskii and P. lenis. Death and escape rates of all predatory mites offered thrips eggs ranged from 0 to 1.67% and from 3.33 to 6.67%, respectively. Survival rates reported in the literature for phytoseiid mites preying on non-mobile foods such as insect eggs or pollen averaged 92% for A. largoensis with eggs of the coconut whitefly Aleurodicus cocois Curtis (De Alfaia et al. 2018 ), 70% for I. degenerans with Ricinus communis L. pollen (Vantornhout et al. 2004), and 93%, 80%, 97%, and 40% for P. lenis when fed Typha latifolia (Linnaeus), Luffa aegyptiaca (Miller), Zea mays (Linnaeus), and R. communis pollen, respectively (Nguyen et al. 2019a).The egg-to-adult development time of A. swirskii when preying on F. occidentalis eggs (7.19 days) was slightly longer than that observed when consuming eggs of T. urticae (6.58 days) and eggs of the cotton whitefly Bemisia tabaci (Gennadius) (6.01 days) (Seiedy et al. 2017) under similar experimental conditions. Additionally, the intrinsic rate of increase of (A) swirskii in our study (rm = 0.188) was slightly lower than that reported when preying on eggs of T. urticae (rm = 0.220) or eggs of (B) tabaci (rm = 1.25) (Seiedy et al. 2017). Notably, A. largoensis exhibited a shorter juvenile developmental time when preying on eggs of the coconut whitefly A. cocois compared to thrips eggs but had a significantly lower intrinsic rate of increase (rm = 0.04 versus 0.21). Nguyen et al. (2019a) reported that the intrinsic rate of increase of P. lenis was higher on T. latifolia pollen and L. aegyptiaca than on thrips eggs but not on Z. mays pollen, whereas the total fecundity of P. lenis was lower when feeding on pollen from T. latifolia, L. aegyptiaca, and Z. mays compared to F. occidentalis eggs.

Various phytoseiid mites have been tested against the first and second instars, as well as the pupal and adult stages of thrips (Ghasemzadeh et al. 2017; Khaliq et al. 2018; Ahmed and Lou 2018). However, their control effects on adult thrips and second instars have been reported to be limited (Khaliq et al. 2018). This limitation in subduing older thrips stages may arise from challenges related to the size (Saito 1986; Choh et al. 2012) and defence behaviours of the prey (Faraji et al. 2001, Magalhães et al. 2005, Wimmer et al. 2008, Choh et al. 2012, de Almeida and Janssen 2013). For instance, larvae of F. occidentalis have been noted to swing their abdomen to defend themselves from predation attempts, and to produce anal droplets as an alarm signal for conspecifics (Bakker and Sabelis 1989; Teerling et al. 1993, de Bruijn et al. 2006). In the present laboratory trials, immature mortality rates were higher when the studied phytoseiids were given thrips larvae as prey compared to when they were provided with thrips eggs. Higher mortality rates on F. occidentalis larvae can be explained by the difficulty of the larval and protonymphal stages of the three phytoseiids to attack larval thrips, demonstrating the ability of this prey to resist attack and cause harm to the predator. On the other hand, higher overall escape rates as compared to death rates on thrips eggs (5% versus 0.56%, respectively) whereas the inverse was noted for thrips larvae, may suggest that thrips eggs are a less preferred type of prey. Whether the studied phytoseiids prefer thrips larvae over eggs should be the subject of further experiments in which both prey types are offered simultaneously.

In summary, both eggs and larvae of F. occidentalis proved to be suitable prey to support development and reproduction of A. largoensis, P. lenis, and A. swirskii. The intrinsic rate of increase of A. largoensis was higher on F. occidentalis eggs than on thrips larvae, whereas in P. lenis the opposite trend was observed. Amblyseius swirskii, on the other hand, showed little or no difference in its development time and intrinsic rate of increase when feeding on either type of prey. These findings confirm the potential of the studied phytoseiid mites as biological control agents against both eggs and larvae of the Western flower thrips, extending the window for effective predation on F. occidentalis populations. Nevertheless, further research is warranted to explore the performance of these phytoseiid species under diverse temperature regimes, which is crucial for their practical implementation in South-East Asia. Future studies conducted on whole plants and under more field-realistic conditions will provide a more comprehensive understanding of the field establishment and predation potential of these phytoseiid mites. In regions like Vietnam, where pesticide usage often prevails as the primary method of pest control, these findings present an encouraging alternative that can safeguard the health of farmers, consumers, and the environment. Conserving populations of predatory mites in agricultural ecosystems or incorporating them into augmentative biological control programs could thus have a positive impact on agricultural production in South-East Asia and beyond.