Acute toxity and sublethal eff ects of pymetrozine on the whitefl y parasitoid Encarsia formosa Gahan

SUMMARY Sublethal eﬀ ects of a pymetrozine-based product (commercial product Chess 50 WP) on life history traits and population growth of one commercialized strain (“Dutch” strain) and two local populations (Bujanovac and Negotin) of the whiteﬂ y parasitoid Encarsia formosa Gahan (Hymenoptera: Aphelinidae) were evaluated in laboratory bioassays. All trials were carried out at 27±1°C temperature and under 60±10 % relative humidity and 16/8 h daylight/darkness photoperiod in four replications. Longevity of wasps exposed for 48 h to residues of the pymetrozine insecticide (LC 50 , 280 mg a.i./l) was shorter (by 2.7-3 days) than that of control wasps. Total parasitism of Negotin wasps was signiﬁ cantly reduced (by 8.2 %), as well as total parasitism and adult emergence of the Dutch strain (by 7.3 and 8.2 %, respectively), compared to control wasps. The instantaneous rate of increase ( r i ) of surviving adult wasps was also signiﬁ cantly reduced (by 6.6, 6.3 and 7.6 % in populations Negotin, Bujanovac and Dutch strain, respectively). Direct treatment of wasps at their pupal stage (LC 50 , 300 mg a.i./l) reduced total parasitism of Negotin wasps (by 8 %), and reduced r i levels, but the reduction was signiﬁ cant only for the Bujanovac (by 6.7 %) and Negotin (by 4.6 %) populations. Juvenile development of the parasitoid in treated pupae was signiﬁ cantly extended (by 0.3-1.1 days), compared to control wasps. The implications of these results on integrated control of the greenhouse whiteﬂ y are discussed.


INTRODUCTION
Th e parasitic wasp Encarsia formosa Gahan (Hymenoptera: Aphelinidae; Coccophaginae) has been used for many years for biological control of the cosmopolitan and polyphagous species of greenhouse whitefl y, Trialeurodes vaporariorum Westwood (Hemiptera: Aleyrodidae), as one of the most successful biological agents in greenhouse and ornamental crops around the world (Enkegaard & Brdsǿgaard, 2006;Pilkington et al., 2010). In order to achieve sustainable integration of biological and chemical measures, evaluation of the eff ect of pesticides, especially at population level, on any parasitoid used in integrated control programmes is needed Stark & Banken, 1999;Stark & Banks, 2003). Despite many restrictive factors, natural enemies can be eff ectively integrated through adequate knowledge of the pesticides to be used (selective pesticides and/or rates or timing of applications) and their eff ects (lethal or sublethal) on natural enemy populations (Croft , 1990;Greathead, 1995).
Pymetrozine is a relatively new insecticide, the only representative of pyridine azomethine, which is highly eff ective and specifi c in controlling insects that suck plant sap (Flückiger et al., 1992a,b;Nicholson et al., 1996). Various authors have demonstrated the effi cacy of pymetrozine in controlling whitefl ies, primarily T. vaporariorum and Bemisia tabaci biotype B (Ferron & Deguine, 2005;Kobayashi, 2007). Pymetrozine is also considered a valuable tool in the integrated protection management (IPM) of greenhouse whitefl y and other pest species (Acheampong & Stark, 2004;Ferron & Deguine, 2005). Pymetrozine is a modulator of chordotonal organs and its mechanism of action diff ers from the mechanism of action of insecticides in other chemical groups (IRAQ, 2017). Th is insecticide is not directly toxic to insects that suck plant sap, but it has an irreversible eff ect on their ability to feed, i.e. it acts by interfering with the nervous regulation of behavior related to insect nutrition. By causing inhibition of the neuromuscular system, which participates in food assimilation, pymetrozine can also cause a so-called "knock-down" eff ect in exposed insects, similar to most conventional neurotoxic insecticides (Harrewijn & Kayser, 1997).
Several studies have examined the potential toxicity of this relatively new insecticide to parasitoids of the order Hymenoptera. Th ese studies focused mainly on direct eff ects of pymetrozine on adults and/ or pupae aft er its topical application (residual contact bioassay or direct treatment of adults and pupae). In general, pymetrozine and its commercial formulations are considered to be either slightly toxic or non-toxic to parasitoids (Tran et al., 2005;Medina et al., 2007).
Studies evaluating the risk of using the commercial pymetrozine-based formulation Chess WP for E. formosa and/or other parasitoid and predator species were mainly based only on the use of its recommended dose, and on acute toxicity and persistence tests performed according to the IOBC phase-up scheme (van de Veire & Tirry, 2003;Abu-Tara et al., 2008). Th ere are a number of studies that have classifi ed the eff ects of other pymetrozine formulations on other species of parasitoids (Hoddle et al., 2001;Torres et al., 2003;Acheampong & Stark, 2004) and predators (Torres et al., 2002;Yoshizawa & Aizawa, 2007;Cabral et al., 2008Cabral et al., , 2011. A small number of studies can be found in literature that address the profi le of pymetrozine through quantifi cation of life parameters of predator and parasitoid population growth (Rezaei et al., 2007;Kheradmand et al., 2012), and none of them concerns the parasitoid E. formosa.
As implementation of the principles of integrated control of whitefl y is in its initial stage in Serbia at present, the goal of this study was to evaluate the eff ects of pymetrozine on life history traits and population growth of two local populations and a commercial Dutch strain of the parasitoid. Th e results of this study are discussed in terms of T. vaporariorum integrated management strategies.

Parasitoid and whitefl y populations
Two local populations of E. formosa were started from pupae collected from tunnel greenhouses for vegetables and ornamentals in locations in Serbia with no known use of commercial parasitoid strains for biological control of greenhouse whitefl ies. Population Bujanovac was collected in Bujanovac (Southern Serbia; 42°30´27˝ N, 21°48´30˝ E) from Solanum nigrum L., and population Negotin was collected in Negotin (Easthern Serbia; 44°13´00˝ N, 22°31´00˝ E) from Hibiscus sp. Th e emerged female wasps of each population were identifi ed as E. formosa using the key given by Polaszek et al. (1992). Th e Dutch strain of E. formosa was purchased from the company 'Zeleni hit' , which is the local agent of Koppert Biological Systems Inc., Th e Netherlands, and it was successfully cultured as a reference strain. Th e Dutch strain of E. formosa and the two local populations of this parasitoid wasp were reared on T. vaporariorum at 27 ± 1°C, and under 60 ± 10 % RH and 16/8 h light/darkness photoperiod. All whitefl ies were reared on tobacco plants, cv. Samsun, in ventilated muslin cages and according to the recommended European Plant Protection Organisation (EPPO, 2004) methodology.

Insecticide
Th e commercial product Chess 50 (manufactured by Syngenta, Germany) is formulated as water dispersible granules (WG). Its content of pymetrozine as the leading product ingredient is standardised to 500 g/l.

Bioassays
All bioassays were conducted in four replications in a climate chamber at 27±1°C and under 60±10 % RH and 16/8 h light/darkness photoperiod. Th e bioassays were performed in Petri dishes (12 cm diameter), each having four lid openings (1 cm diameter) with muslin covers on top to provide ventilation and prevent internal condensation. Each Petri dish contained a 1 % agar layer onto which a tobacco leaf was settled. Parasitoid adults, pupae or whitefl y nymphs were released into Petri dishes, as required for each experiment. Th e pesticide was diluted in distilled water and applied by spraying onto the entire area of each Petri dish (i.e. the lid and lower dish containing a tobacco leaf on top of agar medium). Th e insecticide was applied by a Potter spray tower (2 ml of spray liquid, 100 kPa air pressure, aqueous deposit 2.7 ± 0.2 mg/cm 2 ).
(A) Acute toxicity bioassay with adults. (A) Acute toxicity bioassay with adults. Acute insecticide toxicity to E. formosa adults was assessed by spraying a series of pymetrozine concentrations, covering a range of 10-90 % mortality: 400,350,300,250,225,200 and 100 mg/l. Adults in a control treatment were sprayed with distilled water. Th e insecticide was applied to tobacco leaves placed onto agar and to the inner lid of each Petri dish. Petri dishes with tobacco leaves were left for 2 h to air dry at room temperature, and then parasitoid adults were released inside the dishes. Twenty adult wasps (12-24 h old) were released into each Petri dish that also contained a few droplets of honey on a piece of tinfoil (0.5 × 0.5 mm) fi xed to the lid of each Petri dish by Traganth-kit (a natural, non-toxic adhesive, manufactured by C.E. ROEPER, Germany). Honey was applied aft er the insecticide deposit had dried in order to avoid possible contamination of honey drops and ingestion of insecticide residues by parasitoid wasps. Mortality was calculated based on the number of live wasps in relation to the number of treated wasps 48 h aft er their release (EPPO, 2004).
(B) Acute toxicity bioassay with pupae. (B) Acute toxicity bioassay with pupae. Tobacco leaves with parasitised whitefl y nymphs (pupae) were fi xed to tinfoil with Traganth-kit. Aft er drying, the leaves were cut into pieces, each bearing about 25 parasitoid pupae (4 days old, i.e. 12 days aft er parasitoid oviposition in host nymphs) and placed on fi lter paper in plastic Petri dishes (fi lter paper was moistened with water to fi x them in place during exposure). Th en the pieces of tobacco leaves were treated with a series of pymetrozine concentrations: 500,400,350,300,250,200, and 100 mg/l. Two hours aft er treatment, the leaves were transferred to new Petri dishes and remained there until adults emerged from the pupae. Mortality was assessed as the emerged adult count in relation to the number of treated pupae nine days aft er treatment (EPPO, 2004).

Parasitism bioassays
(A) Parasitism bioassay with insecticide-exposed (A) Parasitism bioassay with insecticide-exposed parasitoid adults parasitoid adults was carried out by releasing 40 wasps into Petri dishes (10 adults aged 0-24 h per each of four dishes) already containing 200-250 third-or fourthinstars of whitefl y nymphs on tobacco leaves settled upon agar medium previously treated with 280 mg/l of pymetrozine. Th is concentration was within the 95 % confi dence limits (CLs) of the LC 50 value estimated in the previous acute toxicity bioassay with adults (Table  1). Approximately 12 whitefl y third-or fourth-instars were present per cm 2 (EPPO, 2004) because a higher number of nymphs would cause premature withering of tobacco leaves, while a lower number would lead to low parasitism. Th e Petri dishes with tobacco leaves were left for 2 h to air dry, and the wasps were then left to lay eggs over the next 48 h before they were transferred to new Petri dishes with hosts. Th e transfering of wasps to new leaves with host nymphs at 48 h intervals continued until the last wasp died. Control dishes were sprayed only with distilled water. Parasitised hosts were counted aft er turning black in appearance. Aft er counting, such pupae were transferred to new clean Petri dishes to monitor the survival of treated pupae (Stouthamer & Mak, 2002). Parasitism was calculated as the number of parasitised pupae per female alive during each 48 h period (parasitism/48 h) summed over the female wasp lifetime (total parasitism). Adult emergence was calculated as the total number of adults that emerged from parasitised whitefl y nymphs. Longevity was calculated as the total number of days that each wasp lived, assuming that wasps died at the midpoint of each 48 h interval.

(B) Parasitism bioassay with F (B) Parasitism bioassay with F 1 generation wasps
generation wasps was carried out in which 40 pupae (4 days old, i.e. 12 days aft er parasitoid oviposition) from each tested population were treated in Petri dishes with 300 mg/l of pymetrozine. Th e concentrations were within 95 % CLs for the LC 50 s calculated in the acute toxicity bioassay (Table 1). Aft er adults emerged from treated pupae, the development time, parasitism/48 h, total parasitism, adult emergence and longevity of the F 1 generation were calculated using the same method as described in the bioassay with exposed wasps. Development time was calculated as the total number of days that elapsed from parasitoid egg laying to adult emergence from pupae.
In both parasitism bioassays (with insecticideexposed adult parasitoids and with insecticide-exposed F 1 generation wasps), parasitism and survival data were used to calculate the instantaneous rates of increase (r i ) using the equation: where N 0 is the initial number of individuals (i.e. 40 adult wasps per replicate), N f is the fi nal number of individuals (i.e. the number of surviving adult wasps, black parasitised pupae and adults emerged), and Δt is the number of days elapsed between the start and the end of a bioassay. Positive r i values indicate a growing population, negative r i values indicate a population in decline and r i = 0 indicates a stable population (Walthall & Stark, 1997). In order to standardise the infl uence of diff erent oviposition durations on r i values, N f values were determined at the end of the 14 th and 16 th days of oviposition in the fi rst and second parasitism bioassays, respectively, i.e. at the time intervals that corresponded to the shortest oviposition period achieved by tested females. In the parasitism bioassay with insecticideexposed adult parasitoids, Negotin and Dutch strain wasps had the shortest oviposition period (14 days), while Negotin wasps had the shortest oviposition period (16 days) in the parasitism bioassay with insecticideexposed F 1 generation wasps.

Statistical analysis
Concentration-mortality data from both acute toxicity bioassays were subjected to probit analysis using the POLO Plus soft ware (LeOra Soft ware, Berkeley, CA). A pairwise comparison of LC 50 s was performed using the lethal dose ratio test: when 95 % CLs for LC ratios included 1, the LCs were not signifi cantly diff erent (Robertson et al., 2007). Kaplan-Meier analysis was used to estimate wasp longevity (SPSS Statistics, Version 17), and survival curves were analyzed by the Log-rank test. Development, parasitism/48 h, total parasitism and adult emergence, and r i data were analysed by twoway ANOVA (insecticide treatment and population were factors), with means separated by Fisher's LSD test (p < 0.05). Parasitism/48 h was also analysed by repeated measures ANOVA. Means of all parameters for treatment and control, for each population, were separated by Student's t-test (p < 0.05). Parasitism and adult emergence data were transformed by √(x + 0.1) to normalise data and eliminate zero values.

Acute toxicity bioassays
Toxicity parameters of the insecticide Chess 50 WP aft er adult exposure to its residues and aft er direct treatment of E. formosa pupae are shown in Table  1. Pymetrozine demonstrated a signifi cantly higher toxicity to the adult than pupal stage of the parasitoid. Th e obtained LC 50 values for both developmental stages of the parasitoid, were close to the maximum recommended concentration of Chess 50 WP (0.06 %= 300 mg a.i./l) for use in protected cultivation systems. A LC 50 ratio test showed that female adults and pupae of all tested populations of E. formosa wasps were equally susceptible to pymetrozine treatment.

Parasitism bioassay with exposed adult wasps
Longevity of E. formosa wasps aft er exposure to pymetrozine residues are shown in Figure 1. Exposure to pymetrozine residues had a signifi cant impact on the longevity of E. formosa adult wasps (F 1,18 =142.06, p<0.001). Besides treatment, longevity reduction was also related to population (F 2,18 =8.17, p<0.01), unlike the interaction of these two factors, which caused no statistically signifi cant infl uence (F 2,18 =0.27, p=0.763). Th e exposed wasps of Bujanovac population lived 2.7 days shorter than control wasps, which was shown to be statistically signifi cant (F 1,6 =84.24, p<0.001). Pymetrozine residues shortened the lifetime of exposed Negotin wasps by 2.71 days (F 1,6 =30.392, p<0.01), compared to control wasps. A signifi cant diff erence (F 1,6 =52.249, p<0.001) was also revealed regarding the longevity of Dutch strain wasps from treatment, which lived 2.96 days briefer than control wasps. Th e survival curves for wasps from residual pymetrozine bioassay are shown in Figure 2. Wasps of all three populations survived for longer periods of time than those exposed only to distilled water: Bujanovac treatment vs. Bujanovac control (ww=60.766, p<0.001); Negotin treatment vs. Negotin control (ww=65.174, p<0.001) and Dutch strain treatment vs. Dutch strain control (ww=67.827, p<0.001).
Parasitism/48 h of wasps in all test populations exposed to pymetrozine residues was under a signifi cant infl uence of observation periods (F 8,144 =199.05, p<0.001). Between and within observation periods, all main eff ects and their related interactions were shown to be statistically signifi cant at the signifi cance level p=0.05, except the interaction of treatment and population between observation periods, which was not signifi cant (F 18,2 =3.41, p=0.055).
Parasitism/48 h of wasps from the test populations of E. formosa are presented in Figure 3. Bujanovac wasps and the Dutch strain achieved lower parasitism/48 h than control females, but not statistically signifi cant. Regarding Negotin population wasps, pymetrozine caused a signifi cant reduction only in the last observation interval (12-14 days of oviposition) (F 1,6 =12.400, p<0.05). In all test populations, the exposure of adult wasps to pymetrozine activity resulted in shorter oviposition periods by two days, compared to control wasps. Bujanovac population had the longest oviposition period, both treated (16 days) and control wasps (18 days). Wasps from Negotin and the Dutch strain laid eggs for two days less than Bujanovac females both under treatment conditions (14 days) and control (16 days).
Th e highest total parasitism was achieved by Bujanovac wasps (143.95 pupae/female), which parasitised the highest average number of whitefl y nymphs even under control conditions (151.28 pupae/female). Population Negotin suff ered the greatest reduction in total parasitism under treatment conditions and achieved the lowest value of that parameter. Total parasitism of the treated Bujanovac wasps was reduced by 4.8 % (F 1,6 =3.76, p=0.100), which proved to be statistically insignifi cant, while the parasitism of Negotin and Dutch strain wasps was signifi cantly reduced by 8.2 % (F 1,6 =8.67, p<0.05) and 7.3 % (F 1,6 =22.31, p<0.01), respectively, compared to control. Similar to total parasitism, the reduction in total emergence of adults from Dutch strain pupae in F 1 generation (  Total adult emergence of Negotin adults was reduced by 5.6 % (F 1,6 = 0.629, p=0.458), and Bujanovac adults by 5.3 % (F 1,6 = 1.707, p=0.239), but the reductions were not statistically signifi cant. Similar to total parasitism, the highest average adult emergence in both variants (treatment and control) was achieved by Bujanovac wasps.
As a consequence of signifi cantly reduced survival and/or parasitism 14 days aft er oviposition started (the duration of the shortest oviposition period of exposed Negotin and Dutch strain adult wasps), the instantaneous rate of increase (r i ) of surviving female wasp adults was also signifi cantly lower: 6.64 % (F 1,6 =51.41, p<0.001), 6.34 % (F 1,6 =14.73, p<0.01) and

Figure 5. Survival curves of E. formosa parasitoid wasps in F 1 generation from local populations Bujanovac (B) and
Negotin (N), and the commercial Dutch strain (D); c = control (distilled water); t = (pymetrozine 300 mg/l) Th e survival curves for parasitoid wasps of all test populations are shown in Figure 5. Wasps that ecloded from pupae treated with pymetrozine survived better than females that ecloded from pupae treated only with distilled water (Bujanovac treatment vs. Parasitism/48 h of female wasps that survived direct pupal treatment was signifi cantly aff ected by the observation period (F 7,126 =372.17, p<0.001). All main eff ects and related interactions were found to be signifi cant both between and within observation periods, except treatement between observation periods (F 7,126 =2.69, p=0.118), and the interaction of treatment and population, which had no signifi cant eff ect (F 7,126 =2.57, p=0.104).
Parasitism/48 h of wasps that had ecloded from treated and control pupae is presented in Figure 6. Pymetrozine caused no shortening of oviposition period in any test population (Bujanovac and Dutch strain females oviposited eggs for 16 days, and Negotin wasps for 14 days). Parasitism/48 h of the treated wasps of Bujanovac population diff ered signifi cantly from control wasps over all observation intervals except 10-12 (F 1,6 = 0.072, p=0.797) and 12-14 days of oviposition (F 1,6 =1.722, p=0.237), when the diff erences were not statistically signifi cant. In wasps of the Dutch strain population, pymetrozine caused no signifi cant reduction in parasitism/48 h in any observation interval during oviposition, as it did in Negotin wasps.
Fourteen days aft er oviposition began (the duration of the shortest oviposition period of population Negotin in treatment and control), the r i of treated Bujanovac (F 1,6 =133.4, p<0.001) and Negotin (F 1,6 =27.35, p<0.01) wasps decreased by only 6.7 and 4.6 %, respectively, which nevertheless turned out to be statistically signifi cant. Th e r i of the Dutch strain was not signifi cantly lower than the rate of control wasps (Table  3). Reduction in r i was caused by treatment (F 1,18 =62.8, p<0.001), population (F 2,18 =26.8, p<0,001), and their interation (F 2,18 =33.5, p<0,001). Considering treatment conditions, the highest r i value was achieved by wasps of the Dutch strain, while Bujanovac wasps had the highest rate in the control.

DISCUSSION
Th e pymetrozine formulation Chess 50 WP did not show high acute toxicity to adults or pupae of any of the three examined populations of E. formosa. Th e obtained LC 50 values for both parasitoid development stages were close to pymetrozine concentration recommended for use in protected cultivation systems (300 mg/l). In adult bioassays, pymetrozine applied at mean lethal concentrations obtained in acute toxicity studies signifi cantly reduced wasp survival, life longevity, total parasitism of Negotin and Dutch strain populations, total emergence of Dutch strain wasps, and instantaneous rate of increase of all three test populations. Applied directly to the parasitoid pupal stage, pymetrozine signifi cantly extended juvenile development, reduced total parasitism of Negotin wasps and signifi cantly reduced the instantaneous rate of increase of Bujanovac and Negotin populations.
A large number of studies have estimated pymetrozine as non-toxic to diff erent development stages of E. formosa, some other parasitoid species and predators, classifying it as compatible for use with these benefi cial organisms. In a laboratory residual test, van de Veire and Tirry (2003) showed that pymetrozine (Chess 25 % WP, 200 g a.i/1) did not cause any mortality of E. formosa adults (one, three and seven days aft er exposure), proving also to be harmless to another two benefi cials tested, the predators Macrolophus caliginosus Wagner and Amblyseius (Neoseiulus californicus) McGregor, while causing toxicity to the predator Orius laevigatus (Fieber). Laboratory research by Abu-Tara et al. (2008) showed that pymetrozine (applied at two recommended concentrations) reduced (<10 %) adult emergence from pupae of the endoparasitoids Bemisia tabaci (Gennadius), E. formosa and Eretmocerus mundus Mercet. Studying the compatibility of diff erent insect growth regulators in the control of B. argentifolii Bellows and Perring, Hoddle et al. (2001) noted that pymetrozine caused the lowest mortality of juveniles of the parasitoid Eretmocerus eremicus Rose and Zolnerowich, but also the lowest mortality rate of whitefl ies as their hosts fi ve days aft er parasitism. In a laboratory study conducted by Morales et al. (2006), in which the formulation of pymetrozine Plenum® WP150 (25 mg a.i./l) was tested on all development stages of Hyposoter didymator (Th unberg), a solitary endoparasitoid of Spodoptera littoralis larvae, there was no statistically signifi cant diff erence in eclosion between treated and untreated pupae. In a study by Medina et al. (2007) the recommended dose of pymetrozine was shown to be totally harmless to H. didymator pupae aft er its uptake with artifi cial food treated topically. Th e results of these authors are in congruence with a laboratory research by Krespi et al. (1991), who noted that there was no signifi cant diff erence in the emergence of adult parasitoids Aphidius uzbekistanicus Luzhetzki aft er direct treatment of parasitoid pupae.
Similar to the results of our study, a study conducted by Joseph et al. (2011) showed that the development of juvenile stages of the endoparasitoid Aphidius ervi Haliday (even at concentrations that were sublethal for the aphid host) was signifi cantly compromised by the use of pymetrozine. Pymetrozine had a negative eff ect on the development of A. ervi larvae, causing changes in the sex ratio of that species in male favor ( Joseph et al. 2011), and increased mortality and reduced life longevity, compared to control wasps (Tran et al. 2005). Similarly, although the pymetrozine formulation Plenum® 50 WG (500 mg a.i /l) has been known by its good reputation of being selective for insects that suck plant sap (Flückiger et al. 1992a, b;Sechser et al., 2002), Harrewijn and Kayser (1997), classifying the eff ects of this formulation on A. ervi juveniles, noted a signifi cant reduction in the growth and development success (40 %) of the parasitoid in the contaminated body of its host. Only the egg and larval stages of the parasitoid were signifi cantly aff ected by the presence of sublethal doses of the formulation in bodies of their hosts, while adult emergence and nymph development did not diff er from controls in uncontaminated hosts. Th is may be due to either direct or indirect eff ect of the contaminated host on the parasitoid. Insecticides can be directly absorbed by parasitoids that develop inside the contaminated bodies of their hosts (the insecticide is moved to the third trophic level) (Harrewijn & Kayser, 1997).
Adverse eff ects of pymetrozine on biological and/ or population parameters of benefi cial organisms have been documented in a small number of studies. Very little is still known about sublethal eff ects of this compound on the parasitoid/host and prey/predator interactions within the trophic system. In the study by Joseph et al. (2011), pymetrozine did not reduce the number of off spring of A. ervi females (Haliday) but, as in our study, it impacted negatively the development of parasitoid larvae. Similarly, Morales et al. (2006) noted that despite a low acute toxicity of pymetrozine to the parasitoid H. didymator, the parasitoid lifetime was signifi cantly reduced, regardless of the insecticide mode of action; in contrast, pymetrozine did not aff ect parasitoid survival, nor did it reduce parasitism.
In a demographic study conducted by Kheradmand et al. (2012), pymetrozine caused 27 % mortality of adults of the aphid parasitoid Diaeretiella rapae (McIntosh) (Hymenoptera: Braconidae), and signifi cantly reduced its life table parameters (intrinsic rate of increase -r m , net reproductive rate -R 0 , and fi nite rate of increaseλ), compared to the control. No mortality was observed in pupae treated with pymetrozine and this insecticide had no signifi cant eff ects on life table parameters of wasps emerged from treated pupae. In a laboratory study by Rezaei et al. (2007), pymetrozine caused a 34 % reduction in r m value of the common green lacewing, Chrysoperla carnea (Stephens).
Aft er exposing adult E. formosa wasps to pymetrozine residues, Bujanovac population was found to show more favorable values of all examined parameters than the Dutch strain population of wasps. Regarding treated parasitoid pupae, the surviving Dutch population wasps achieved the highest values of the instantaneous rate of increase, while Bujanovac wasps lived and oviposited longer, and achieved higher values of total parasitism and total emegence. Similar to our previous fi ndings, Bujanovac population wasps were again shown to be more promising for integrated control of whitefl y (Drobnjaković et al., 2018(Drobnjaković et al., , 2019. Considering the obtained lethal concentrations for both examined stages of the parasitoid, and taking into account the reduction in reproductive and demographic parameters, which is (lower than 10 %), the application of the recommended concentration of pymetrozine would show limited adverse eff ects on parasitoid population, so that pymetrozine can be used together with the parasitoid E. formosa without causing a great impact on its eff ectiveness in integrated whitefl y control. Disadvantages of this insecticide include its signifi cant impact on wasp lifetime (in bioassays with adults) and juvenile development (in bioassays with pupae).
Th ese results off er a starting point for further investigation of local E. formosa populations, in comparison to the commercial Dutch strain, as biological agents intended for integrated control of T. vaporariorum in vegetable and ornamental protected crops. Further research should focus on evaluating pymetrozine in greenhouse trials with an emphasis on population-level responses.