Side-effects of insecticides used in tomato fields on Trichogramma pretiosum (Hymenoptera, Trichogrammatidae)

The side-effects of lufenuron (0.4g a.i./L), triflumuron (0.15g a.i./L), imidacloprid (0.28g a.i./L), cyromazine (0.11g a.i./L), methoxifenozide (0.12g a.i./L), pirimicarb (0.25g a.i./L) and abamectin (0.18g a.i./L) on different developmental stages of the egg parasitoid Trichogramma pretiosum Riley, were investigated. Abamectin was the only insecticide to affect parasitoid emergence and sex ratio, regardless of the developmental stage and parasitoid generation exposed. Abamectin, lufenuron and pirimicarb also decreased the lifetime of F1 females exposed during the egg-larva stage. The capacity of parasitism was significantly reduced by all the products when females were treated in pupal stage.


Introduction
Tomato (Lycopersicon esculentum Mill.) is cultivated in about 40.000ha throughout Brazil and the production is up to two million tons (Lopes and Santos, 1994).There are about 200 arthropod species associated with the tomato crop (Lange and Bronson, 1981), and the intensive and widespread use of pesticides required for the pest control in this crop has led to the development of pesticide resistance in key tomato pests, such as the tomato leafminer Tuta absoluta (Meyrick) (Siqueira et al., 2000a, b;2001).The social-economic importance of this crop and the environmental risks, due to the management practices employed, support an ever-increasing research interest in the phytosanitary problems of this crop (Michereff Filho and Vilela, 2001).
The genus Trichogramma shows different species responsible for natural control of several insectspest, mainly lepidopterous.Faria et al. (2000) related that the species Trichogramma pretiosum Riley, has been studied in Brazil for the control of several pests in corn, cassava and cotton crops.Furthermore, in the tomato crop, the control of the moth T. absoluta has been studied in some countries, through inundative releases of T. pretiosum (Amaya-Navarro, 1988;Faria Jr., 1992;Haji, 1996Haji, , 1997)).The use of parasitoids of the genus Trichogramma has also been evaluated for the control of other pests in this crop.Blackmer et al. (2001), accomplishing experiments in tomato crop, verified natural parasitism rates of Neoleucinodes elegantalis (Guenée) eggs for T. pretiosum about 28.7%, with the reduction of pesticide utilization.
The utilization of selective insecticides is a reasonable strategy in pest management, because it favors the conservation of natural enemies in the agroecosystem (Carvalho et al., 1999).Acta Scientiarum. Agronomy Maringá, v. 25, no. 2, p. 275-279, 2003 Works aiming to study the physiologic selectivity of different chemical groups of pesticides to Trichogramma spp.were accomplished.Cônsoli et al. (1998), evaluating the effects of several insecticides on the immature stages of T. pretiosum, verified that phenthoate and cartap were harmful, lambdacyalothrin and abamectin were intermediate, tebufenozid and teflubenzuron were harmless to slightly harmful.Brunner et al. (2001) using Potter's tower, sprayed insecticides on adults of T. platneri Nagarkatti up to two days old and observed that the compounds oxamyl, imidacloprid and Bacillus thuringiensis kurstaki caused 100% mortality 48 hours after spraying, being observed selectivity of the products diflubenzuron, fenoxycarb and tebufenozide for that species.
This paper intends to study the impact of some pesticides, at dosage rates currently used in Brazilian tomato fields, to immature stages of the egg parasitoid T. pretiosum.

Material and methods
Fifteen newly-emerged females of T. pretiosum were individualized in glass tubes (8.5cm x 2.5cm) and, subsequently blue paper cards (3.0cm x 0.5cm) with eggs of Anagasta kuehniella (Zeller) previously UV-killed (±125 eggs) were offered.The parasitism period was 24h.Afterwards, the females were removed and the paper cards were maintained under controlled conditions at 25±2ºC, 70±10% RH and 14h of photophase for T. pretiosum development.
To evaluate the pesticide effects, 15 newlyemerged females from treated eggs (F 1 generation) were randomly taken from each treatment, individualized in glass tubes (8.5cm x 2.5cm) and provided with non-treated and inviabilized eggs of A. kuehniella for 48 hours, allowing the assessment of testing the pesticide side-effects in individuals of the next generation.
The biological parameters evaluated were: 1) the number of eggs parasitized by F 1 generation insects; 2) the parasitoid emergency of F 1 and F 2 generations; 3) the parasitoid of sex ratio in the egg-larval, prepupal and pupal stages and also 4) the F 1 female longevity.
The experimental design used was completely randomized with eight treatments (seven pesticides and a control treatment in which only water was applied) and 15 replicates, analyzing the pesticide effects on three parasitoid developmental stages and in two generations.A paper card containing about 125 parasitized host eggs was the experimental unit used.
The number of parasitized eggs and the longevity were transformed to 5 .0 + x , and the parasitoid emergency was transformed to 100 / arcsen x before subjected to Anova.The comparisons between treatments were carried out using Scott and Knott's groupment analysis test at 5% probability (Scott and Knott, 1974).

Impact of pesticides on the F 1 and F 2 emergence of T. pretiosum
Abamectin was the pesticide which most affected parasitoid emergence when used at the egg-larval stage of T. pretiosum, independently of the generation.
Cyromazine, imidacloprid and methoxifenozide were toxic for individuals of the generations F 1 and F 2 .Lufenuron showed no effects at this stage (Table 1).
For the F 1 generation, when prepupae of T. pretiosum where exposed to pesticides, parasitoid emergence rates were smaller for abamectin, imidacloprid and lufenuron (Table 1).Despite the relatively high emergence rate observed with abamectin application, allowing 70% emergence of T. pretiosum, all the surviving insects died soon afterwards and subsequently there parasitism ability could not assessed (Table 2).This mortality may be due to the contact of the parasitoids with abamectin residue through the host egg chorion.The insects treated with cyromazine, methoxifenozide, pirimicarb and triflumuron showed emergence rates similar to those of the control treatment, suggesting their safety to parasitoids of the F 1 generation.Cyromazine, imidacloprid, lufenuron, methoxifenozide, triflumuron and pirimicarb allowed more than 80% emergence for parasitoids of the F 2 generation.Abamectin was highly toxic to T. pretiosum (Table 1).
Acta Scientiarum. Agronomy Maringá, v. 25, no. 2, p. 275-279, 2003  There were significant effects of the pesticides on parasitoid emergence when they were exposed during their pupal stage (Table 1).In F 1 , the highest emergence rate was observed for pirimicarb, differing from those obtained with cyromazine, imidacloprid, lufenuron, methoxifenozide and triflumuron, which showed intermediate toxicity.The drastic reduction in the emergence rate of T. pretiosum was observed with abamectin.This last pesticide also affected insect emergence at the following generation, allowing only 49.7% emergence.Lufenuron and triflumuron also reduced the emergence of T. pretiosum.No significant difference was observed between the treatments cyromazine, imidacloprid, methoxifenozide, pirimicarb and control, for the individuals of F 2 generation (Table 1).
The results reported here resemble those of Carvalho et al. (2001) and Cônsoli et al. (1998), who applied abamectin on eggs of A. kuehniella containing the parasitoid T. pretiosum in the egg-larval, prepupal and pupal stages, and verified the reduction of the emergence rate from 10 to 40%.Carvalho et al. (2001) did not observe significant reduction in emergence of parasitoids when T. pretiosum were treated with cyromazine, triflumuron and pirimicarb in the egg-larval stage.This different results may have occurred due to the different origin of the parasitoid populations used in the present study, which may influence the quality of the parasitoid population and the intensity of the insect exposition to pesticide residues (Prezotti et al., 1994;Pratissoli and Parra, 2001).
Regarding the prepupal and pupal stages, Cônsoli et al. (1998) also observed similar results to those reported in the present study with lufenuron and abamectin.Carvalho et al. (2001) also noted similar effect for the abamectin, cyromazine, pirimicarb and triflumuron on the same parasitoid species.Regarding the pupal stage, the results with triflumuron resemble those reported by Narayana and Babu (1992) for T. chilonis Ishii.

Pesticide impact on sex ratio of T. pretiosum
There was a significant effect of pesticides on sex ratio (Table 3).When the pesticides were applied during the egg-larval, prepupal and pupal stages, the smallest number of females was observed with abamectin (0.38, 0.37 and 0.41, respectively).For the egg-larval and prepupal stages, the sex ratio observed with cyromazine was not different from the control indicating its selectivity to the parasitoid based on this parameter (Table 3).Selective effect of cyromazine on the developmental stages (egg-larval, Acta Scientiarum. Agronomy Maringá, v. 25, no. 2, p. 275-279, 2003 prepupal and pupal) of T. pretiosum has also been previously reported (Carvalho et al., 2001).All the pesticides compromised the sex ratio of T. pretiosum when the host eggs containing the parasitoid in the pupal stage were treated; however, abamectin showed the highest distortions on sex ratio (Table 3).

Pesticide impact on the parasitism by T. pretiosum
There was a significant effect of the pesticides on egg parasitism by T. pretiosum of the F 1 generation.Abamectin reduced in 66.54% the number of parasitized eggs when applied at egg-larval stage.Cyromazine, pirimicarb and imidacloprid showed an intermediate effect on parasitism, while triflumuron, lufenuron and methoxifenozide results did not differ from control (Table 2).
When the insecticides were applied on the insects during the prepupal stage, there was no significant reduction of the number of eggs parasitized by T. pretiosum.On the other hand, insects exposed to abamectin died soon after emergence, not allowing the parasitism of the females.This was possibly due to the contact with residues of this pesticide through the host egg chorion, as reported before.
When pesticides were applied on A. kuehniella eggs in which parasitoids were at the pupal stage, abamectin remained the most harmful pesticide.The other pesticides were less toxic, but they significantly reduced the number of parasitized eggs per female of the F 1 generation (Table 2).
The results found here confirm those obtained by Carvalho et al. (2001) with the growth regulator triflumuron when applied on the egg-larval and prepupal stages of the T. pretiosum.Cônsoli et al. (1998) also reported a reduction in the parasitism of females emerged from eggs of A. kuehniella exposed to these two insecticides.

Pesticide impact on the longevity of F 1 of T. pretiosum
Females of the F 1 generation of T. pretiosum exposed to abamectin during the egg-larval stage had their longevity reduced (42.2%).The other pesticides were less toxic, but they significantly reduced the females longevity of the F 1 generation (Table 2).
The longevity of females treated at the prepupal stage with methoxifenozide and pirimicarb was not affected.However, abamectin has proved extremely toxic, killing the insects still inside the host egg, or after their emergence.The remaining insecticides showed intermediary effect (Table 2).
Cyromazine, imidacloprid, methoxifenozide and triflumuron did not affect the longevity of parasitoids female when they were exposed during the pupal stage.The pesticides abamectin, lufenuron and pirimicarb significantly reduced the longevity of the females treated at the pupal stage (Table 2).
The results here reported agree with those of Carvalho et al. (2001). However, Carvalho et al. (2001) did not observe some harmful effects of triflumuron on T. pretiosum, diverging from this study, since it was observed here that triflumuron was slightly toxic to females of T. pretiosum.These divergences are probably associated with the differences between populations of T. pretiosum (Carvalho et al., 2001).Zaki and Gesraha (1987) also reported that diflubenzuron, another acylurea as lufenuron and triflumuron, reduced longevity of T. evanescens (Westwood) in about 44%.
The results showed that, regardless of the development stage of T. pretiosum and the generation, the pesticide abamectin is highly toxic to this species.Abamectin, lufenuron and pirimicarb reduced the longevity and the number of parasitized eggs when the parasitoids were exposed at the egglarval stage.The prepupal stage was the most tolerant to the studied pesticides, except for abamectin.2., 1986, Guangzhou. Anais... Paris: INRA, 1988. v.43, p.453-458. BLACKMER, J.L. et

Table 1 .
Emergence (%) (± SEM) of Trichogramma pretiosum, in the generations F 1 and F 2 treated with insecticides during the egglarval, prepupal and pupal stages when developing in Anagasta kuehniella eggs * Means followed by the same capital letter in the column and lower case in the line are not significantly different by the Scott and Knott test at 5%

Table 2 .
Number of parasited eggs/female (±SEM) and longevity (±SEM) of Trichogramma pretiosum in the F 1 generation, treated with insecticides during the egg-larval, prepupal and pupal stages when developing in Anagasta kuehniella eggs Means followed by the same capital letter in the column and lower case in the line are not significantly different by the Scott and Knott test at 5% *

Table 3 .
Sex ratio (± SEM) of Trichogramma pretiosum treated with insecticides during the egg-larval, prepupal and pupal stages when developing in Anagasta kuehniella eggs Means followed by the same capital letter in the column are not significantly different by the Scott and Knott test at 5% * al. Oviposition preference of Neoleucinodes elegantalis (Guenée) (Lepidoptera: Crambidae) and rates of parasitism by Trichogramma pretiosum Riley (Hymenoptera: Trichogrammatidae) on Lycopersicon