Lethal and Sublethal Impacts of Acaricides on Tamarixia radiata (Hemiptera: Eulophidae), an Important Ectoparasitoid of Diaphorina citri (Hemiptera: Liviidae)

Abstract

The use of synthetic acaricides for management of pest mites may alter the efficacy of the ectoparasitoid Tamarixia radiata (Waterston) in biological control of Diaphorina citri Kuwayama, the vector of the bacteria associated with huanglongbing (HLB) in citrus orchards. We evaluated the toxicity of 16 acaricides that are recommended for the control of citrus-pest mites to T. radiata . Acrinathrin, bifenthrin, carbosulfan, and fenpropathrin caused high acute toxicity and were considered harmful (mortality >77%) to T. radiata . Abamectin, diflubenzuron, etoxazole, fenbutatin oxide, fenpyroximate, flufenoxuron, hexythiazox, propargite, spirodiclofen, and sulfur caused low acute toxicity and affected the parasitism rate and emergence rate of adults (F ₁ generation), and were considered slightly harmful to T. radiata . Dicofol and pyridaben did not affect the survival and action of the ectoparasitoid, and were considered harmless. In addition to its acute toxicity, carbosulfan caused mortality higher than 25% for >30 d after application, and was considered persistent. Acrinathrin, bifenthrin, fenpropathrin, fenpyroximate, propargite, and sulfur caused mortalities over 25% until 24 d after application and were considered moderately persistent; abamectin was slightly persistent, and fenbutatin oxide was short lived. Our results suggest that most acaricides used to control pest mites in citrus affect the density and efficacy of T. radiata in the biological control of D. citri . However, further evaluations are needed in order to determine the effect of these products on this ectoparasitoid under field conditions.

biological control, acaricide, mortality, parasitism rate, emergence rate

Huanglongbing (HLB) is among the most important and destructive citrus diseases worldwide ( Bové 2006 , Grafton-Cardwell et al. 2013 ). In Brazil, HLB is present in 6.9% of cultivated citrus trees (∼200 million trees) in the state of São Paulo, the main citrus-producing region of the country ( Fundecitrus 2013 ). Given the absence of a cure, HLB is managed by planting healthy seedlings, reducing inoculum by eliminating symptomatic trees, and mainly, control of the insect vector, the Asian citrus psyllid Diaphorina citri Kuwayama (Hemiptera: Liviidae). Currently, in Brazil, growers are spraying insecticides (organophosphates, pyrethroids, and neonicotinoids) from 18 to 25 times annually, to control D. citri ( Belasque Jr. et al. 2010 ). Despite their high efficacy, these products may alter the balance between pests and their natural enemies, causing outbreaks of secondary pests, resurgence of target pests, and selection of resistant populations ( Yamamoto and Bassanezi 2003 , Tiwari et al. 2011 , Guedes and Cutler 2013 ).

Therefore, conservation and augmentation of biological control agents are important strategies to reduce population levels of pests and the impacts caused by overuse of insecticides in citrus orchards. Among the natural enemies, the ectoparasitoid wasp Tamarixia radiata (Waterston) (Hemiptera: Eulophidae) has shown great potential for use in pest management programs to control D. citri ( Pluke et al. 2008 , Qureshi et al. 2009 , Hall and Nguyen 2010 , Williams et al. 2013 ). T. radiata is a specialized ectoparasitoid, developing preferentially in third- to fifth-instar D. citri nymphs ( Skelley and Hoy 2004 , Hall et al. 2013 ). In Brazil, this ectoparasitoid was found in 2006 in the municipality of Piracicaba, São Paulo, and is currently present in most citrus-producing regions of the country ( Paiva and Parra 2012 ). It can be mass-reared in the laboratory for inundative release in field conditions, and rapidly suppresses populations of D. citri ( Chen and Stansly 2014 ). In recent years, successive releases of T. radiata have been conducted in areas of HLB management programs, as well as in areas with orange jasmine [ Murraya paniculata (L.) Jack (Rutaceae)] infested with D. citri nymphs or in areas next to commercial orchards, where the ectoparasitoid is multiplying ( Parra et al. 2010 ). In São Paulo state, early estimates of parasitism rates ranged from 27.5 to 80% ( Gómez-Torres et al. 2006 , Parra et al. 2010 ). However, with increased use of chemical pesticides for control of the vector D. citri , parasitism rates were reduced to below 25.7% ( Paiva and Parra 2012 ).

In addition to D. citri , the flat mite Brevipalpus phoenicis (Geijskes) (Acari: Tenuipalpidae), vector of the Citrus Leprosis Virus ( CiLV ), and the citrus rust mite Phyllocoptruta oleivora (Ashmead) (Acari: Eryophyidae) are considered key pests of this crop ( Oliveira and Pattaro 2008 ). Although less important, mites of the family Tetranychidae, especially the citrus red mite Panonychus citri (McGregor) and the Texas citrus mite Eutetranychus banksi (McGregor), also require control because of the continual population outbreaks observed after successive applications of insecticides to control D. citri ( Yamamoto and Zanardi 2013 ). This increased use of acaricides can alter biological and behavioral parameters of the ectoparasitoid T. radiata and affect the action of this natural enemy in controlling D. citri . Some studies have demonstrated the acute and residual toxicity of certain acaricides to T. radiata ( Cocco and Hoy 2008 , Hall and Nguyen 2010 , Lira et al. 2014 ), but none has examined the sublethal effects of acaricides on its biological parameters and parasitism efficiency. Knowledge of sublethal effects contributes to the understanding of the impacts of pesticides on natural enemies ( Desneux et al. 2007 ). This information is important for the development of management strategies aimed at the conservation and augmentation of biological control agents, and also to guarantee the success of integrated pest management (IPM) programs in citrus. Considering the importance of T. radiata as a biological control agent of D. citri , this study assessed the lethal and sublethal impacts of 16 acaricides that are recommended for the control of citrus mites, on this ectoparasitoid.

Materials and Methods

Rearing of T.radiate

The culture was established from adults collected in citrus orchards in the municipality of Piracicaba, São Paulo, Brazil. The insect populations were maintained at the Insect Biology Laboratory of the Department of Entomology and Acarology of the “Luiz de Queiroz” College of Agriculture, University of São Paulo (ESALQ/USP), Piracicaba, São Paulo, Brazil. The culture was maintained in a climate-controlled room (25 ± 2°C, 70 ± 10% relative humidity [RH], and a photoperiod of 14:10 [L:D] h). For the rearing we used seedlings of orange jasmine ( M. paniculata ) cultivated in pots (2 liter). The seedlings were initially pruned to a height of 25 cm, and after producing sprouts (2–3 cm in length), they were placed in rearing cages (40 by 60 by 50 cm ³ ) and used as a substrate for oviposition of females and feeding by nymphs and adults of D. citri , according to the method proposed by Nava et al. (2007) . When D. citri nymphs reached the fourth and fifth instars, the seedlings were transferred to acrylic cages (90 by 50 by 50 cm ³ ) and T. radiata adults were released to parasitize the nymphs, according to the method described by Gómez-Torres et al. (2012) . Drops of honey were placed on the sides of the cages as food for the parasitoid adults.

Pesticides

Sixteen commercial acaricides registered for the control of pest mites (especially B. phoenicis and P. oleivora ) in citrus orchards were evaluated on T. radiata . All compounds were tested without adjuvant at the label rates recommended by the Brazilian Ministry of Agriculture, Livestock and Food Supply (MAPA; Agrofit 2014 ). The acaricides and concentrations (g a.i. liter ⁻¹ ) used in the bioassays are described in Table 1 .

Table 1.

Acaricides used in the bioassays

Common name	Trade name (MoA IRAC) a	Chemical group	Dosage used (g a. i. liter −1 )	Manufacturer	Target pest b	Other pests controlled c
Abamectin	Vertimec 1.8 EC (6)	Activators of chloride channels	0.005	Syngenta	CRM 1	CRM 2 , TCM, BM, CL, ACP
Acrinathrin	Rufast 5 SC (3A)	Sodium channel modulators	0.005	Bayer	FM	–
Bifenthrin	Talstar 10 EC (3A)	Sodium channel modulators	0.020	FMC	FM	CRM 1 , CFB, CO, ACP
Carbosulfan	Marshal Star 70 EC (1A)	Acetylcholinesterase inhibitors	0.350	FMC	CRM 1	ACP
Dicofol	Dicofol 18.5 EC (UN)	Compounds of unknown or uncertain MoA	0.370	Milenia	FM	CRM 1
Diflubenzuron	Micromite 24 SC (15)	Inhibitors of chitin biosynthesis, type 0	0.120	Chemtura	CRM 1	CL, CFB, ACP
Etoxazole	Borneo 4.6 SC (10B)	Mite growth inhibitors	0.021	Sumitomo	FM	–
Fenbutatin oxide	Torque 50 SC (12B)	Inhibitors of mitochondrial ATP synthase	0.400	Basf	FM	BM, CRM 1 , CRM 2 , TCM
Fenpropathrin	Danimen 30 EC (3A)	Sodium channel modulators	0.150	Sumitomo	FM	BCA, CFB, MFF, CO, ACP
Fenpyroximate	Ortus 5 SC (21A)	Mitochondrial complex I electron transport inhibitors	0.050	Arysta	CRM 1	BM, FM
Flufenoxuron	Cascade 10 EC (15)	Inhibitors of chitin biosynthesis, type 0	0.050	Basf	CRM 1	BM, FM
Hexythiazox	Savey 50 WP (10A)	Mite growth inhibitors	0.015	DuPont	FM	–
Propargite	Omite 72 EC (12C)	Inhibitors of mitochondrial ATP synthase	0.720	Chemtura	CRM 1	CRM 2 , FM, TCM
Pyridaben	Sanmite 20 EC (21A)	Mitochondrial complex I electron transport inhibitors	0.150	Iharabras	FM	CRM 1
Spirodiclofen	Envidor 24 SC (23)	Inhibitors of acetyl CoA carboxylase	0.060	Bayer	CRM 1	BM, CRM 1 , FM, TCM
Sulfur	Kumulus DF 80 WG	MAD	4.000	Basf	FM	BM, CRM 1

^a  Classification of the mode of action of pesticides according to Insecticide Resistance Action Committee (IRAC, 2014); MAD, mechanism of action not defined.

^b  Target pests: CRM ₁ , citrus rust mite [ Phyllocoptruta oleivora (Ashmead)]; FM, flat mite [ Brevipalpus phoenicis (Geijskes)].

^c  Other pests controlled: CRM ₂ , citrus red mite [ Panonychus citri (McGregor)]; TCM, Texas citrus mite [ Eutetranychus banksi (McGregor)]; BM, broad mite [ Polyphagotarsonemus latus (Banks)]; CL, citrus leafminer ( Phyllocnistis citrella Station); CFB, citrus fruit borer ( Gymnandrosoma aurantianum Lima); CO, citrus orthezia [ Praelongorthezia praelonga (Douglas)]; MFF, Mediterranean fruit fly [ Ceratitis capitata (Wiedemann)]; BCA, brown citrus aphid [ Toxoptera citricida (Kirkaldy)]; and ACP, Asian citrus psyllid ( Diaphorina citri Kuwayama).

Bioassays

All laboratory bioassays were conducted in a climate-controlled room (25 ± 2°C, 70 ± 10% RH, and a photoperiod of 14:10 [L:D] h), using a fully randomized design.

Bioassay 1: Acute Toxicity of Acaricides to T. radiata Adults

To assess the acute toxicity of acaricides to T. radiata adults, leaf discs (3.3 cm in diameter) of Valencia sweet orange [ Citrus sinensis (L.) Osbeck (Rutaceae)] were initially sprayed with 2 ml of solution, in a Potter tower (Burkard Scientific, Uxbridge, United Kingdom) adjusted to a pressure of 0.7 kg cm ⁻² , resulting in a fresh dry deposition of 1.8 ± 0.1 mg cm ⁻² , according to the criteria established by the Pesticides and Beneficial Organisms Working Group of the International Organization for Biological Control of Noxious Animals and Plants, West Palearctic Regional Section (IOBC/WPRS) for studies of pesticide toxicity to natural enemies ( Van de Veire et al. 2002 ). Distilled water was used as a control treatment.

After the treatments were sprayed, the discs were kept in a climate-controlled room for 3 h to allow the residues to dry. Then, the discs were placed in individual Petri dishes (3.5 cm in diameter by 0.7 cm in height) containing a 2-mm layer of agar: water (2.5% w/v) and used as an experimental unit. Next, 10 parasitoid adults (5 females and 5 males) up to 24 h old were anesthetized with CO ₂ for 5 s and placed in each experimental unit. The experimental units were sealed with voile fabric to allow gas exchange and prevent the accumulation of excess moisture. A honey droplet (∼1 mm ³ ) was placed on the voile to serve as food for the parasitoids during the assay period. For each treatment, five repetitions were used.

The number of alive and dead insects was recorded 24 h after exposure to the residues. Moribund parasitoids and those that did not react to the touch of a fine brush were considered dead. The acute toxicity (mortality) of each acaricide was calculated using the formula of Abbott (1925) . Based on mortality data (M), the acaricides were classified according to the criteria defined by the IOBC/WPRS ( Van de Veire et al. 2002 ): class 1 = harmless (M < 25%), class 2 = slightly harmful (25% ≤ M ≤ 50%), class 3 = moderately harmful (51% ≤ M ≤ 75%), and class 4 = harmful (M > 75%).

Bioassay 2: Sublethal Effects of the Acaricides on T. radiata Adults

The acaricides considered harmless (class 1) in the evaluation of acute toxicity (bioassay 1) were used in the second bioassay. For this, 20 females of T. radiata up to 24 h old were released in each experimental unit made with Petri dishes (3.5 cm in diameter by 0.7 cm in height) containing a leaf disc (3.3 cm in diameter) of Valencia sweet orange treated with the respective acaricide or deionized water (control) and placed on a 2-mm layer of agar: water (2.5% w/v). Females were kept on product residues for a period of 24 h. For each treatment, five repetitions were used ( n  = 100). After this period, the moribund and dead females were counted (acute toxicity) and the surviving females were transferred to glass cages (90 by 50 by 50 cm ³ ) containing Rangpur lime seedlings (15 cm in height) infested with D. citri nymphs in the fourth and fifth instars, in a proportion of a female of the ectoparasitoid to 10 D. citri nymphs (daily parasitism rate of ∼7 D. citri nymphs per T. radiata female [ Gómez-Torres 2009 ]). The parasitism period was 48 h. After this period, the seedlings were transferred to new cages (90 by 50 by 50 cm ³ ) to evaluate the parasitism rate. The number of nymphs parasitized in each treatment was assessed 9 d after the seedlings were transferred to new cages. During evaluation, the branches with parasitized nymphs (mummified) were cut off and transferred to Petri dishes (9 cm in diameter by 0.7 cm in height) containing moistened filter paper, to assess the number of ectoparasitoid adults emerged (emergence rate). The adults obtained in each treatment were counted, separated by sex (sex ratio), and placed in individual glass tubes (2.5 cm in diameter by 8.5 cm in length) with a drop of honey, to evaluate the longevity of the insects. The insects (females and males) were observed every 24 h until they died.

Based on the mortality data (acute toxicity), on the parasitism (parasitism rate), and on the number of emerged parasitoids (emergence rate) of T. radiata , the total effect for each acaricide was estimated using the formula $E(\%)=100-(100-M)\times Er$ , proposed by Van de Veire et al. (2002) , where: M  = mortality corrected according to the formula of Abbott (1925) and Er  = sublethal effects, calculated as $Er=R_1\times R_2$ , where R₁  = ratio of the mean number of nymphs parasitized (mummified) by female in the acaricide treatment and the control, and R₂  = ratio of the mean number of emerged parasitoids (emergence rate) in the acaricide treatment and the control. Based on the total effect ( E ), each acaricide was classified according to the criteria established by the IOBC/WPRS for extended laboratory tests ( Van de Veire et al. 2002 ): class 1 = harmless (E < 25%), class 2 = slightly harmful (25% ≤ E ≤ 50%), class 3 = moderately harmful (51% ≤ E ≤ 75%), and class 4 = harmful (E > 75%).

Bioassay 3: Acute Toxicity and Sublethal Effects of Acaricides Applied to T. radiata Pupae

To evaluate the acute toxicity and sublethal effects of acaricides on pupae of T. radiata , acaricides that were considered harmless (class 1) to ectoparasitoid adults in the assay of acute toxicity (bioassay 1) were used. Our decision to use such acaricides was due to its mechanism of action (see Table 1 ). Acaricides that act as mite growth inhibitors and energy production can reduce the emergence, survival, and parasitism rates of offspring when these products are applied on immature stages of the parasitoids, due to ingestion and contact with the acaricide residues at the time of, or after, emergence. In this context, we decided to evaluate the possible effects of these acaricides on pupae of T. radiata to verify the compatibility of these products with the ectoparasitoid before recommending its use in integrated pest management programs. For this, Rangpur lime seedlings cultivated in 100-ml pots were pruned to a height of 25 cm to stimulate growth. After sprouts (2–3 cm in length), each seedling was infested with 20 D. citri nymphs of fourth and fifth instars and placed in a climate-controlled room for 24 h to allow the nymphs to acclimate on the host plant. After this period, the seedlings were placed in glass cages (90 by 50 by 50 cm ³ ) and T. radiata females were added to allow them to parasitize the nymphs (proportion of one female of ectoparasitoid to 10 D. citri nymphs). The exposure time for the parasitism was 24 h. Then, the seedlings with the parasitized nymphs were transferred to new glass cages (90 by 50 by 50 cm ³ ), where they remained for 9 d (period for development from egg to pupa), according to Rosa et al. (2012) . After this period, the seedlings with T. radiata pupae were sprayed with the insecticides, using a Guarany backpack sprayer (Guarany Indústria e Comércio Ltda., São Paulo, Brazil) equipped with a TXVS-4 conical nozzle (Teejet Technologies, São Paulo, Brazil) to the runoff point, and each seedling was used as the experimental unit. Distilled water was used as a control treatment. For each treatment, five repetitions with five seedlings per repetitions were used.

After the spraying treatments, the experimental units were kept in a climate-controlled room for 3 h to allow the residues to dry. Afterward, branches with T. radiata pupae were cut off from each seedling, and placed on moistened filter paper in Petri dishes (9 cm in diameter) to evaluate the acute toxicity of the acaricides. The number of emerged adults in each experimental unit was evaluated every 24 h for 8 d. The mortality of each treatment was calculated based on the number of pupae (parasitized nymphs) and the number of adults emerged in each experimental unit. The acute toxicity (reduction in the emergence rate of the F ₀ generation) of each acaricide was calculated based on mean mortality observed in each acaricide treatment and the control. Adults surviving in each experimental unit were counted, separated by sex (sex ratio), and placed in glass cages (90 by 50 by 50 cm ³ ) containing seedlings infested with D. citri nymphs of fourth and fifth instars to evaluate the effects of the acaricides on the parasitism rate (F ₀ generation) and emergence rate (F ₁ generation) of the ectoparasitoid. For this, the D. citri nymphs were exposed to T. radiata females for parasitism for 48 h. After this period, surviving T. radiata adults were transferred to individual glass tubes (2.5 cm in diameter by 8.5 cm in length) to evaluate the longevity. During this period, the adults were fed with drops of honey as described for bioassay 1.

The parasitism rate (F ₀ generation) of T. radiata was determined based on the number of parasitized (mummified) D. citri nymphs in each experimental unit, 9 d after the T. radiata adults were removed from the cages. The emergence rate (F ₁ generation) of the ectoparasitoid was determined based on the number of T. radiata adults emerged in each experimental unit 15 d after the removal of T. radiata adults from the cages. Based on the acute toxicity (reduction in the emergence rate of the F ₀ generation) and sublethal effects (parasitism rate in the F ₀ generation, and emergence rate in the F ₁ generation) on T. radiata , the total effect was estimated for each acaricide, using the same procedure and criteria for analysis described in bioassay 2.

Bioassay 4: Duration of Harmful Effects of Acaricides on T. radiata Adults

The duration of the harmful effects of the acaricides was assessed on T. radiata adults, using the acaricides that were considered slightly harmful or harmful to ectoparasitoid adults in the bioassay of acute toxicity (bioassay 1), using the method proposed by Brunner et al. (2001) . Five Valencia sweet orange seedlings ∼80 cm tall and with 20 mature leaves, for each treatment, were sprayed with acaricides and distilled water (control treatment) until the runoff point, using a Guarany backpack sprayer equipped with a TXVS-4 conical nozzle. After the treatments were applied, the seedlings were placed in a greenhouse. At 1, 3, 7, 10, 17, 24, and 30 d after the treatments, one treated leaf of each plant was randomly removed, brought to the laboratory, and discs (3.3 cm in diameter) were cut for use in preparing the experimental units as described in bioassay 1. Then, 10  T. radiata adults up to 48 h old were anesthetized with CO ₂ for 5 s and transferred into each experimental unit. The experimental units were sealed with voile fabric and placed in a climate-controlled room. For each treatment, five repetitions were used ( n  = 50).

The mortality of insects was recorded after 24 h of exposure to the residues. Insects that did not react to the touch of a fine brush were considered dead. The mortality data for each acaricide and dates of assessment were corrected by the formula of Abbott (1925) . The products that reduced adult survival by <25% compared to the control treatment (distilled water) were classified according to the persistence scale proposed by the IOBC/WPRS: class 1: short life (<5 d), class 2: slightly persistent (5–15 d), class 3: moderately persistent (16–30 d), and class 4: persistent (>30 d; Van de Veire et al. 2002 ).

Data Analysis

Generalized linear models ( Nelder and Wedderburn 1972 ) with Gaussian distributions were used for the data analysis of the longevity of insects. The quality of the adjustment was verified with a half-normal plot with simulated envelope ( Hinde and Demétrio 1998 ). Multiple comparisons by the Tukey–Kramer test ( P  ≤ 0.05) using the “ glht ” function of the “ DTK ” package, with adjustment of the P values, were used to compare the means of different treatments. Sex ratio data for treatments were compared using the Chi-square test ( P  ≤ 0.05). All analyses were performed with the statistics software R version 2.15.1 ( R Development Core Team 2012 ).

Results

Acute Toxicity and Sublethal Effects of Acaricides on T. radiata Adults

Residues of acrinathrin, bifenthrin, carbosulfan, and fenpropathrin were harmful (class 4) to T. radiata adults (mortality greater than 77% after 24 h of exposure; Table 2 ). Abamectin, fenbutatin oxide, fenpyroximate, propargite, and sulfur acaricides caused mortality rates from 29.1 to 42.9% and were considered slightly harmful (class 2). Dicofol, diflubenzuron, etoxazole, flufenoxuron, hexythiazox, pyridaben, and spirodiclofen caused <20% mortality and were considered harmless ( Table 2 ). Despite their low acute toxicity, the acaricide diflubenzuron, etoxazole, flufenoxuron, and spirodiclofen reduced the parasitism rate and the number of emerged insects of the F ₁ generation, and were therefore considered slightly harmful (class 2) to T. radiata ( Table 3 ). Diflubenzuron, etoxazole, and flufenoxuron decreased the emergence of females (sex ratio) in the F ₁ generation compared to the other treatments. The longevity was lower in adults exposed to residues of diflubenzuron, flufenoxuron, and spirodiclofen, than in the control treatment ( Table 3 ). Dicofol and pyridaben did not affect the parasitism rate, emergence rate, sex ratio, or longevity, and were considered harmless (class 1) to adults of the ectoparasitoid ( Table 3 ).

Table 2.

Mean number of dead insects and standard error mean ( N  ± SEM) and corrected mortality (%) of T. radiata exposed to residual contact acaricides

Treatment	Dosage used (g a. i. liter −1 ) a	Acute toxicity		IOBC/WPRS class b
		N  ± SEM	%
Control	–	0.1 ± 0.22	–	–
Abamectin	0.005	2.8 ± 0.83	27.3	2
Acrinathrin	0.005	8.0 ± 0.64	79.8	4
Bifenthrin	0.020	10.0 ± 0.00	100.0	4
Carbosulfan	0.350	10.0 ± 0.00	100.0	4
Dicofol	0.370	2.4 ± 0.92	23.2	1
Diflubenzuron	0.120	0.4 ± 0.44	3.0	1
Etoxazole	0.021	0.6 ± 0.23	5.0	1
Fenbutatin oxide	0.400	4.0 ± 1.56	39.4	2
Fenpropathrin	0.150	9.8 ± 0.23	97.8	4
Fenpyroximate	0.050	3.8 ± 0.75	37.4	2
Flufenoxuron	0.050	1.5 ± 0.87	14.1	1
Hexythiazox	0.015	0.2 ± 0.22	1.0	1
Propargite	0.720	5.0 ± 0.64	49.5	2
Pyridaben	0.150	2.5 ± 1.0	24.2	1
Spirodiclofen	0.060	2.2 ± 0,8	21.2	1
Sulfur	4.000	5.0 ± 1.6	49.5	2

^a  Field dosage recommended for the control of mites pest in citrus.

^b  IOBC/WPRS class based on parasitoid mortality: class 1 =  harmless (M < 25%), class 2 = slightly harmful (25% ≤ M ≤ 50%), class 3 = moderately harmful (51% ≤ M ≤ 75%), and class 4 = harmful (M > 75%).

Table 3.

Sublethal and total effects of acaricides applied to adults of T. radiata

Treatment	Dosage used (g a. i. liter −1 )	M (%) a	R1b	R2c	E (%) d	IOBC/WPRS class e	Sex ratio f	Longevity (d) g
Control	–	4.0	6.30	6.25	–	–	0.74 ± 0.07a	6.2 ± 0.21a
Dicofol	0.370	19.4	8.06	5.56	3.6	1	0.79 ± 0.02a	4.7 ± 0.48b
Diflubenzuron	0.120	4.0	6.49	4.06	37.0	2	0.59 ± 0.03b	4.0 ± 0.18b
Etoxazole	0.021	4.0	5.42	3.75	48.4	2	0.64 ± 0.01b	6.3 ± 0.58a
Flufenoxuron	0.050	7.4	6.24	4.38	33.1	2	0.61 ± 0.04b	4.3 ± 0.13b
Hexythiazox	0.015	14.0	0.91	5.31	30.4	2	0.77 ± 0.05a	6.2 ± 0.78a
Pyridaben	0.150	24.0	7.94	5.38	13.3	1	0.87 ± 0.01a	6.0 ± 0.81a
Spirodiclofen	0.060	23.9	7.62	4.25	34.1	2	0.80 ± 0.03a	4.5 ± 0.68b
							χ 2  = 0.02178	F  = 3.367
							df = 7	df = 7, 32
							P  = 0.010	P  = 0.0464

^a M —Mortality corrected by formula of Abbott (1925) .

^b R₁ —Effect on parasitism (mean number of parasitized nymphs).

^c R₂ —Effect on emergence (mean number of adults emerged in F ₁ generation).

^d E —Total effect of acaricides calculated by formula of Van de Veire et al. (2002) .

^e  IOBC/WPRS class based on parasitoid mortality: class 1 = harmless (M < 25%), class 2 = slightly harmful (25% ≤ M ≤ 50%), class 3 = moderately harmful (51% ≤ M ≤ 75%), and class 4 = harmful (M > 75%).

^f  Means followed by the same letter in a column do not differ significantly by Chi-square test ( P  ≤ 0.05).

^g  Means followed by the same letter in a column do not differ significantly (GLM with Gaussian distribution, followed by post hoc Tukey test; P  ≤ 0.05).

Acute Toxicity and Sublethal Effects of Acaricides on T. radiata Pupae

The application of acaricides to T. radiata pupae indicated that diflubenzuron, etoxazole, flufenoxuron, and hexythiazox did not affect the emergence rate of adults or the sex ratio and longevity of T. radiata in the F ₀ generation, but did reduce the parasitism rate in the F ₀ generation and emergence rate in the F ₁ generation, and were therefore considered slightly harmful (class 2) to T. radiata ( Table 4 ). Dicofol and pyridaben caused mortality rates of 24.8 and 28.9%, respectively, but led to increases in the parasitism rate and did not affect the emergence rate, sex ratio, or longevity of the ectoparasitoid, and were considered innocuous (class 1) to T. radiata . Spirodiclofen did not significantly influence any of the variables evaluated, and was also classified as harmless (class 1) to T. radiata ( Table 4 ).

Table 4.

Total effects of acaricides applied to pupae of T. radiata

Treatment	Dosage used (g a. i. liter −1 )	M (%) a	R1b	R2c	E (%) d	IOBC/WPRS class e	Sex ratio (F 1 generation) f	Longevity (d) (F 1 generation) g
Control	–	4.0	6.82	6.80	–	–	0.69 ± 0.06a	7.9 ± 1.61a
Dicofol	0.370	24.8	8.32	7.48	15.2	1	0.82 ± 0.14a	6.7 ± 1.32a
Diflubenzuron	0.120	13.2	6.82	4.08	45.5	2	0.75 ± 0.19a	7.8 ± 1.17a
Etoxazole	0.021	6.0	6.17	4.76	40.8	2	0.80 ± 0.06a	7.6 ± 1.61a
Flufenoxuron	0.050	12.6	6.34	4.96	40.5	2	0.70 ± 0.14a	7.3 ± 1.13a
Hexythiazox	0.015	19.7	6.14	5.54	38.6	2	0.79 ± 0.10a	6.0 ± 1.36a
Pyridaben	0.150	28.9	7.18	7.48	13.3	1	0.77 ± 0.05a	6.5 ± 0.51a
Spirodiclofen	0.060	9.2	7.20	6.12	9.6	1	0.76 ± 0.08a	8.6 ± 1.31a
							χ2  = 0.222	F  = 0.468
							df = 7	df = 7, 32
							P  = 0.236	P  = 0.826

^a M —Effect on emergence rate (F ₀ generation) calculated by formula of Abbott (1925) .

^b R₁ —Effect on parasitism (mean number of parasitized nymphs in F ₀ generation).

^c R₂ —Effect on emergence (mean number of adults emerged in F ₁ generation).

^d E —Total effect of acaricides calculated by formula of Van de Veire et al. (2002) .

^e  IOBC/WPRS class based on parasitoid mortality: class 1 = harmless (M < 25%), class 2 = slightly harmful (25% ≤ M ≤ 50%), class 3 = moderately harmful (51% ≤ M ≤ 75%), and class 4 = harmful (M > 75%).

^f  Means followed by the same letter in a column do not differ significantly by Chi-square test ( P  ≤ 0.05).

^g  Means followed by the same letter in a column do not differ significantly (GLM with Gaussian distribution, followed by post hoc Tukey test; P  ≤ 0.05).

Duration of the Harmful Effect of Acaricides on T. radiata Adults

The application of acaricides that were considered slightly harmful or harmful to adults of T. radiata on citrus seedlings in the greenhouse indicated that carbosulfan maintained its harmful activity (≥25% mortality) for >30 d after spraying (DAS), and was therefore considered persistent (>30 d, class 4) according to the IOBC/WPRS study group criteria ( Table 5 ). However, acrinathrin, bifenthrin, fenpropathrin, fenpyroximate, propargite, and sulfur caused mortality exceeding 25% until 24 DAS, and were considered moderately persistent (16–30 d, class 3). Abamectin was slightly persistent (5–15 d, class 2) and fenbutatin oxide was evaluated as short life (<5 d, class 1; Table 5 ).

Table 5.

Mortality (%) of T. radiata adults 1, 3, 7, 10, 17, 24, and 30 d after spraying of the acaricides

Treatment	Dosage used (g a. i. liter −1 )	Mortality (%)—days after spraying (DAS)							IOBC/WPRS class a
		1	3	7	10	17	24	30
Control	–	6.0 ± 0.43	2.0 ± 0.32	4.0 ± 0.44	0.0 ± 0.00	2.0 ± 0.32	2.0 ± 0.32	0.0 ± 0.00	−
Abamectin	0.005	31.7 ± 3.75	37.8 ± 7.43	30.5 ± 7.13	29.5 ± 6.13	10.5 ± 4.73	5.2 ± 4.73	6.0 ± 0.40	2
Acrinathrin	0.005	87.5 ± 3.27	73.5 ± 4.44	68.3 ± 9.42	38.1 ± 3.33	15.8 ± 7.21	8.3 ± 4.55	4.9 ± 0.82	3
Bifenthrin	0.020	100.0 ± 0.00	100.0 ± 0.00	87.5 ± 2.17	70.0 ± 4.64	29.3 ± 4.25	22.2 ± 5.43	18.8 ± 9.93	3
Carbosulfan	0.350	100.0 ± 0.00	100.0 ± 0.00	92.5 ± 3.24	67.5 ± 8.37	43.2 ± 1.37	37.0 ± 3.57	31.1 ± 5.37	4
Fenbutatin oxide	0.400	33.5 ± 4.93	12.6 ± 6.23	17.5 ± 5.35	12.4 ± 7.75	18.9 ± 2.62	4.0 ± 0.58	3.8 ± 2.94	1
Fenpropathrin	0.050	93.6 ± 4.38	97.8 ± 2.24	87.5 ± 6.07	60.5 ± 6.91	37.4 ± 5.68	14.8 ± 8.22	38.3 ± 4.83	3
Fenpyroximate	0.150	40.4 ± 6.46	49.6 ± 8.76	43.0 ± 3.69	36.0 ± 4.65	27.4 ± 4.46	9.6 ± 6.83	18.3 ± 8.47	3
Propargite	0.720	50.2 ± 7.92	52.2 ± 8.14	39.5 ± 2.71	40.5 ± 2.48	33.1 ± 4.12	8.3 ± 4.56	48.9 ± 3.86	3
Sulfur	4.000	95.7 ± 4.30	95.7 ± 4.45	88.1 ± 3.56	88.1 ± 6.56	58.9 ± 6.94	19.1 ± 1.24	21.3 ± 7.23	3

^a IOBC/WPRS class based on the duration of the harmful effect of acaricides: class 1 = short life (<5 d), class 2 = slightly persistent (5–15 d), class 3 = moderately persistent (16–30 d), and class 4 = persistent (>30 d).

Discussion

Acute toxicity and sublethal effects of 16 acaricides recommended for the control of pest mites in citrus were evaluated on the ectoparasitoid T. radiata . The acaricides showed different toxicity levels according to the chemical group used in the bioassays. Acrinathrin, bifenthrin, and fenpropathrin pyrethroids and carbosulfan carbamate showed the highest acute toxicity to T. radiata adults, with >77% mortality 24 h after exposure. These results are similar to those obtained by Hall and Nguyen (2010) , who reported mortality rates above 91.3% for T. radiata adults, 24 h after direct spraying or exposed to dry residues of fenpropathrin. In addition to their effects on T. radiata , bifenthrin and fenpropathrin also caused high mortality to adults of Aphytis melinus Debach (Hymenoptera: Aphelinidae) ( Michaud and Grant 2003 ), Encarsia formosa (Gahan) (Hymenoptera: Aphelinidae), and Eretmocerus eremicus Rose & Zolnerowich (Hymenoptera: Aphelinidae) ( Prabhaker et al. 2007 ). However, only slightly acute toxicity (30–79% mortality) was observed in adults of Chrysocharis pentheus (Kamijo) and Sympiesis striatipes Ashmead (Hymenoptera: Eulophidae) ( Mafi and Ohbayash 2006 ) and Trichogramma chilonis Ishii and Trichogramma brasiliensis (Ashmead) (Hymenoptera: Trichogrammatidae) ( Shankarganesh et al. 2013 ) after 24 h of exposure to residues of bifenthrin and carbosulfan, respectively. The high acute toxicity of these acaricides is attributed to high absorption of the products by insects during the walking on the surfaces treated with acaricides, the rapid penetration capacity of products in the integument, and the low capacity of detoxification of active ingredients by enzymes oxidases and hydrolases ( Siegfried 1993 ). Therefore, our results indicate that the use of these acaricides in citrus orchards to control pest mites can severely affect the survival of T. radiata , and inhibit the action of this ectoparasitoid in controlling D. citri .

In addition, acrinathrin, bifenthrin, and fenpropathrin pyrethroids were moderately persistent (≥25% mortality between 17 and 24 d after the products were sprayed), whereas the carbosulfan carbamate retained its harmful activity for >30 d, and was considered persistent for adults of T. radiata . Similarly, Prabhaker et al. (2007) found that, under laboratory conditions, the residual effect of fenpropathrin persisted longer than 21 d for adults of A. melinus , while in the field, Nguyen and Hall (2010) found up to 64.4% reduction in the population levels of adult T. radiata 22 d after the product was sprayed. The long residual period may be explained by high adherence capacity of these products in citrus leaves and by lower degradation by environmental factors ( Yadav et al. 2003 ). Thus, our results suggest that the successive use of these products can compromise biological control of D. citri exercised by ectoparasitoid T. radiata .

Despite their high acute toxicity and long residual period, these acaricides may be used in periods of low mobility and density of adults of T. radiata , to exploit the ecological selectivity of these products. In the main citrus-producing regions of Brazil, this period coincides with the dry months of the year (from mid-autumn to mid-spring), when the vegetative growth rate of citrus plants (sprouts) slows significantly, substantially decreasing infestation levels of D. citri nymphs ( Yamamoto et al. 2001 ) and the population of T. radiata adults in orchards ( Paiva and Parra 2012 ). It should be emphasized, however, that this period also coincides with the initial infestation of B. phoenicis (a main pest mite of citrus) and with the highest population levels of tetranychid mites (especially P. citri and E. banksi ) that cause considerable damage in plants and, in most cases, require chemical intervention for their control ( Oliveira and Pattaro 2008 ). Therefore, during this period, the use of these acaricides for management of these targets will not cause negative impacts on T. radiata adults.

The exposure of T. radiata adults to residues of abamectin, fenbutatin oxide, fenpyroximate, propargite, and sulfur, at recommended concentrations for the control of B. phoenicis and P. oleivora in citrus orchards ( Table 1 ), revealed that these acaricides caused low acute toxicity (<42.9% mortality) to T. radiata adults. However, our results were different from those obtained by Hall and Nguyen (2010) , who found high mortality (>79%) of T. radiata adults subjected to direct spray or exposed to residual contact of 0.016 g liter ⁻¹ abamectin. Likewise, Cocco and Hoy (2008) observed high acute toxicity (93.1% mortality) to T. radiata adults sprayed with abamectin in the lowest label rate (0.40 g liter ⁻¹ ). In addition to T. radiata , others parasitoids species demonstrated high susceptibility to abamectin. Luna-Cruz et al. (2011) and Liu et al. (2012) evaluating the acute toxicity of abamectin at concentrations of 0.75 and 0.0112 g liter ⁻¹ , respectively, on adults of Tamarixia triozae (Burks) (Hymenoptera: Eulophidae), found that these acaricides caused high mortality of the parasitoid after 24 h of exposure to residues. Likewise, Hernández et al. (2011) observed high mortality of Ganaspidium igrimanus (Kieffer) (Hymenoptera: Figitidae) adults when it was exposed to residual contact of 0.1123 g liter ⁻¹ abamectin, whereas Vanaclocha et al. (2013) found high susceptibility of A. melinus adults kept for 24 h on leaves of Clementine [ Citrus clementine Hort. (Rutaceae)] treated with 0.40 g liter ⁻¹ of acaricide. The lower acute toxicity observed in this study is attributed to lower concentration of abamectin used in bioassays. In addition, not adding oil at the time of product application may have intensified the degradation process of abamectin present in the foliar surface by environmental factors ( Vanaclocha et al. 2013 ).

Acute toxicity similar to that found in the present study was observed in adults of Trichogramma cacoeciae Marchal (Hymenoptera: Trichogrammatidae) ( Saber 2011 ) and E. eremicus ( Sugiyama et al. 2011 ) exposed to residues of fenpyroximate and sulfur, respectively. On the other hand, fenpyroximate was considered innocuous to T. triozae adults ( Liu et al. 2012 ); propargite was safe for E. formosa and A. melinus adults ( Gholamzadeh et al. 2012 , Vanaclocha et al. 2013 ); and fenbutatin oxide did not cause mortality of Aphidius colemani Viereck (Hymenoptera: Braconidae) adults ( Urbaneja et al. 2008 ). Based on the low acute toxicity, our results indicate that the use of these acaricides to control pest mites can reduce the population of T. radiata adults and the effectiveness of ectoparasitoid in the control of D. citri in citrus orchards.

In addition, fenbutatin oxide, abamectin, fenpyroximate, propargite, and sulfur retained their harmful activity for up to 1, 10, 17, 24, and 24 d after the products were sprayed, respectively. Fenbutatin oxide also showed low residual persistence (short life) to adults of A. melinus parasitoid (<10% mortality until 6 d after spraying; Morse et al. 1987 ) and for females of predatory mites Iphiseiodes zuluagai Denmark and Muma (Acari: Phytoseiidae) (<30% mortality, 1 d after spraying) and Euseius alatus DeLeon (Acari: Phytoseiidae) (<30% mortality until 5 d after spraying) of product ( Reis and Sousa 2001 ), demonstrating that this acaricide caused less impact on the biological control agents present in citrus orchards. The low residual persistence of fenbutatin oxide on T. radiata adults is a desirable feature and important for integrated pest management programs, because it allows the establishment of ectoparasitoid in the production area in a short period of time after application. On the other hand, abamectin, fenpyroximate, propargite, and sulfur demonstrated a lower compatibility with the ectoparasitoid T. radiata . However, these results were obtained in laboratory, where the insects were continually exposed to residues of acaricides (maximum harmful potential); probably in field conditions, the acute toxicities and the residual effects of these products will be lower than those observed here. This hypothesis is supported by the study of Nguyen and Hall (2010) , who found a residual period of abamectin and sulfur of 8 d, whereas Morse et al. (1987) and Vanaclocha et al. (2013) reported a mortality of A. melinus adults of 10 and 38.3% at 6 and 7 d after spraying of abamectin, respectively. Therefore, in areas with high incidences of pest mites, where releases of T. radiata will be conducted to control D. citri , our results indicate that releases can be performed safely 1, 10, 17, 24, and 24 d after spraying of fenbutatin oxide, abamectin, fenpyroximate, propargite, and sulfur, respectively. Notably, however, abamectin is one of the most-used active ingredients in the control of citrus leafminer [ Phyllocnistis citrella Station (Lepidoptera: Gracillariidae)], which infests citrus plants during the vegetative growth phase ( Paiva 2011 ). Fenbutatin oxide, fenpyroximate, propargite, and sulfur are recommended for management of Polyphagotarsonemus latus (Bancks) (Acari: Tarsonemidae), which also infests plants during vegetative growth. This period also coincides with higher infestation levels of D. citri nymphs in citrus plants, as well as of its ectoparasitoid T. radiata . Therefore, these acaricides should be used with caution in order to avoid compromising the efficacy of this ectoparasitoid in biological control of D. citri in citrus orchards.

Dicofol, diflubenzuron, etoxazole, flufenoxuron, hexythiazox, and spirodiclofen did not cause significant mortality (<25%) to T. radiata adults. Hall and Nguyen (2010) found similar results for T. radiata adults sprayed directly or exposed to dry residues of diflubenzuron. Diflubenzuron and flufenoxuron were considered nontoxic to adults of Telenomus remus Nixon (Hymenoptera: Scelionidae) ( Carmo et al. 2010 ), and flufenoxuron and lufenuron did not cause mortality of adults of Eretmocerus mundus (Mercet) (Hymenoptera: Aphelinidae), E. eremicus , and E. formosa ( Sugiyama et al. 2011 ). Likewise, Vanaclocha et al. (2013) reported that the use of etoxazole and hexythiazox was compatible with the parasitoid A. melinus , demonstrating that these acaricides do not affect the survival of adult parasitoids. The low mortality of adults is attributed to action mechanism of products. Diflubenzuron and flufenoxuron act by inhibition of chitin biosynthesis at the time of ecdysis ( Willrich and Boethel 2001 ), whereas etoxazole and hexythiazox inhibit the growth of the arthropods by a mechanism similar to benzoylphenylureas ( Demaeght et al. 2014 ). Thus, these products present high activity while applied on immature stage of insects, but did not affect the survival of parasitoid adults.

Although they did not cause adult mortality, residues of diflubenzuron, etoxazole, and flufenoxuron led to lower longevity (except etoxazole) and parasitism rates in T. radiata females compared to the control. These acaricides also reduced the emergence rate and sex ratio (proportion of females) in the offspring (F ₁ progeny). Similarly, Schneider et al. (2004) observed lower longevity and parasitism rate in Hyposoter didymator (Thunberg) (Hymenoptera: Ichneumonidae) females exposed to residues of diflubenzuron, and Tabozada et al. (2014) found reductions in parasitism ability, emergence rate, sex ratio, and longevity of Trichogramma evanescens (Westw.) (Hymenoptera: Trichogrammatidae) and Bracon brevicornis Wesmael (Hymenoptera: Braconidae) treated with flufenoxuron. The lowest emergence rate in the F ₁ generation may be attributed to the mortality of parasitoid larvae within D. citri nymphs, caused by malformation of the parasitoid integument (lower chitin biosynthesis) before its emergence ( Nauen and Smagghe 2006 , Schneider et al. 2008 ). The reduction in longevity and the emergence rate and mainly in the sex ratio of offspring, further reduces the effectiveness of the ectoparasitoid in biological control of D. citri , because only the female wasps parasitize this psyllid. In addition to their sublethal effects on adults, these acaricides also reduced the emergence rate and parasitism rate of T. radiata females originated from pupae treated with acaricides. Although the pupal stage is considered to be more tolerant than adults to xenobiotic agents, our results showed that spraying these products may affect the biological and behavioral parameters of the ectoparasitoid. The harmful effects of acaricides sprayed on T. radiata pupae are attributed to their penetration ability and to direct contact with the acaricide residues when the parasitoid emerges. These hypotheses are supported by the studies of Schneider et al. (2003 , 2004) , who found that diflubenzuron easily penetrated the integument of H. didymator when applied on the parasitoid pupae. However, future studies should be conducted to elucidate the main forms of contamination of T. radiata pupae by active ingredients of acaricides.

Of the acaricides evaluated, only dicofol and pyridaben were harmless to T. radiata . However, Hall and Nguyen (2010) reported that the use of pyridaben in the concentration of 0.375 g liter ⁻¹ (2.5 times higher than in this study) caused low mortality of T. radiata adults exposed to contact with the product residues, demonstrating that the acaricide is compatible with IPM programs. Likewise, Yamamoto and Bassanezzi (2003) reported that dicofol was safe for Ageniaspis citricola (Hymenoptera: Encyrtidae) adults, an important parasitoid of P. citrella . Previous studies have shown that dicofol and pyridaben were harmless or only slightly harmful to Azya luteipes Mulsant, Coccidophilus citricola Brèthes, Cycloneda sanguinea (L.), and Pentilia egena Mulsant (Coleoptera: Coccinellidae), and lacewings (Neuroptera: Chrysopidae), important predators of arthropod pests in citrus orchards ( Yamamoto and Bassanezzi 2003 ). These results indicate that these acaricides are safe and can be used at times of high population density or in conjunction with inundative releases of T. radiata not affecting the survival and effectiveness of the ectoparasitoid for biological control of D. citri .

Acknowledgments

We thank the Brazilian Federal Agency for the Support and Evaluation of Graduate Education (CAPES) and the National Council for Scientific and Technological Development (CNPq - grant number 140651/2013-6) for financial support and award of scholarships. We also thank Janet W. Reid for revising the English text.