Selectivity of pesticides used in rice crop on Telenomus podisi and Trichogramma pretiosum 1

Rice is one of the cereals with greatest economic and social importance worldwide, and Brazil is the largest producer outside the Asian continent (FAO 2015). Despite the high yield, in Brazil, rice crops are subject to the action of numerous pest organisms that cause economic losses. The predominant method for pest control consists of applying pesticides (Pazini et al. 2015), which can exert a negative influence on the population of natural enemies of insect pests (Lou et al. 2013). ABSTRACT RESUMO


INTRODUCTION
Rice is one of the cereals with greatest economic and social importance worldwide, and Brazil is the largest producer outside the Asian continent (FAO 2015).Despite the high yield, in Brazil, rice crops are subject to the action of numerous pest organisms that cause economic losses.The predominant method for pest control consists of applying pesticides (Pazini et al. 2015), which can exert a negative influence on the population of natural enemies of insect pests (Lou et al. 2013).
Telenomus and Trichogramma species stand out as agents for the biological control in rice crops, and the main strategy for preserving them is the use of selective pesticides.This study aimed at evaluating the toxicity of pesticides used in irrigated rice crop on Telenomus podisi Ashmead (Hymenoptera: Platygastridae) and Trichogramma pretiosum Riley (Hymenoptera: Trichogrammatidae).Adults of these parasitoids were exposed to dry residues of pesticides, in a completely randomized experiment, with 25 treatments (24 pesticides + control) and four replications.The insecticides clorantraniliprole, flubendiamide and diflubenzuron and the biological insecticides based on Beauveria bassiana and Metarhizium anisopliae were harmless to T. podisi and T. pretiosum.The harmless herbicides were: 2.4-D amine, profoxydim, quinclorac, ethoxysulfuron and saflufenacil.The fungicide epoxiconazole + kresoxim-methyl was also harmless to these two biological control agents.Therefore, these pesticides are indicated for the integrated pest management, in flooded rice areas.
The use of pesticides in rice crops is a reality, and integrated pest management is an alternative required (Sosbai 2014).Therefore, one strategy for the preservation of the parasitoids is to use selective insecticides, herbicides and fungicides, as well as other chemical and biological agents more harmless to natural enemies (Biondi et al. 2012).However, there is a lack of information on the adverse effects of the main insecticides, herbicides and fungicides on eggs parasitoids used for biological control in Brazil, in rice crops.
This study aimed at evaluating the selectivity of pesticides (insecticides, herbicides and fungicides) used in irrigated rice crop on adults of T. podisi and T. pretiosum egg parasitoids.

MATERIAL AND METHODS
The study was conducted at the Universidade Federal de Pelotas, in Capão do Leão, Rio Grande do Sul State, Brazil, between 2014 and 2015.The procedures followed standards adapted from the International Organisation for Biological and Integrated Control of Noxious Animals and Plants (IOBC) (Hassan et al. 2000, Carmo et al. 2010).
A total of 24 pesticides that are commonly used in irrigated rice crop were evaluated on adults of T. podisi and T. pretiosum (Table 1).Selectivity bioassays were conducted with insecticides, herbicides and fungicides and a control treatment (distilled water).The doses used followed the registered maximum doses for rice and/or irrigated rice (Agrofit 2015) (Table 1), adjusted to a mix volume of 200 L ha -1 .The experimental design adopted was completely randomized, with 25 treatments (24 products + control) and four replications.
For the selectivity bioassays on adults of T. podisi, eggs of E. heros parasitized by T. podisi (± 50 eggs) were deposited in emergence tubes (glass bottles of 12 cm long x 2 cm in diameter in one end and 0.7 cm in another) with a few drops of pure honey.The tubes were placed in an acclimatized room (temperature: 25 ± 1 ºC; relative humidity: 70 ± 10 %; photophase: 14 h), until the parasitoids emergence.The pesticides were sprayed on glass plates (13 cm x 13 cm), in a Potter tower, calibrated to deposit 1.75 ± 0.25 mg of mix per cm 2 .The edges of the plates were protected by a square plastic structure, so that only the central area measuring 10 cm x 10 cm was sprayed.After the drying period, the plates were fixed in contact cages, in a circulating air system (Hassan et al. 2000).The outer surfaces of the plates that were not treated were covered with black paper, with a square-shaped cut in the center (7 cm x 7 cm), in order to concentrate the parasitoids in the treated area, attracted by the light.Emergence tubes with adult parasitoids (24 h of age) were attached to the cages, which were kept in an acclimatized room (temperature: 25 ± 1 ºC; relative humidity: 70 ± 10 %; photophase: 14 h).
Twenty hours later, the emergence tubes were uncoupled.Eggs of E. heros (cards with ± 100 eggs) were offered to the parasitoids in the cages, at 24 h, 48 h and 72 h after the pesticides spraying.The experiment was terminated after 96 h of parasitoids exposure to pesticide residues.After this period, the cards of eggs were removed from the cages and stored under the same test conditions to check for parasitism.
Eggs of A. kuehniella on cards containing 450 ± 50 eggs each were offered to the parasitoids in cages at 24 h (three cards), 48 h (two cards) and 96 h (one card) after the pesticides spraying.The experiment was terminated after 168 h of parasitoids exposure to pesticide residues.After this period, the cards of eggs were removed from the cages and stored under the same test conditions to check for parasitism.
The mean number of parasitized eggs per female in each treatment was used to estimate parasitism.The parasitism reduction, compared to the control treatment, was calculated by the equation PR (%) = [(1 -Vt/Vc) * 100], where PR is the percentage of parasitism reduction, Vt the mean parasitism for the treatment and Vc the mean parasitism in the control treatment.Thereby, the pesticides were classified according to the IOBC standards: class 1: harmless (PR < 30 %); class 2: slightly harmful (30 % ≤ PR ≤ 79 %); class 3: moderately harmful (80 % ≤ PR ≤ 99 %); class 4: harmful (PR > 99 %).
The data on the mean number of eggs per female parasitized were subjected to the Shapiro-Wilk normality test and homogeneity of variances by the Bartlett test.If these assumptions were not met, the Kruskal-Wallis non-parametric analysis was carried out, and the average of the treatments were compared by the Dunn test, at 5 %.In case of normality and equality of variances, the means were compared by the Scott-Knott test, at 5 %.The R software 3.2.0(R Development Core Team 2015) was used to carry out the tests.
* Pesticides unregistered for rice and/or irrigated rice crop -registration dose in soybean crop (Agrofit 2015).I Formulation and concentration (mL L -1 or g kg

RESULTS AND DISCUSSION
Significant differences were observed in the parasitism reduction of T. podisi and T. pretiosum between the treatments with insecticides (Table 2).Chlorantraniliprole, flubendiamide, diflubenzuron and the biological insecticides based on Beauveria bassiana and Metarhizium anisopliae did not differ from the control, in terms of number of eggs parasitized, and were classified as harmless (class 1), with a parasitism reduction of no more than 10 % for both parasitoids (  2007), which cause insect death by preventing the normal muscle contractions of the insect.These compounds are very selective, since they act more specifically on phytophagous insects, primarily the Lepidoptera order (Stecca et al. 2014).In addition, they are considered harmless to natural enemies (Lahm et al. 2009), because they act by contact primarily through ingestion, as evidenced by our contact test with dry pesticide residues (Table 2).
Diflubenzuron was also selective to Telenomus remus Nixon (Hymenoptera: Platygastridae) (Carmo et al. 2010) and T. pretiosum (Carvalho et al. 1994).Insecticides like diflubenzuron, which is an insect growth regulator, have the ability to kill specifically the target insect and preserve the agents of biological control.Hormones that trigger the physiological molting process have different efficacy among the taxonomic orders of insects (Carmo et al. 2010).Pests and their natural enemies are generally from different orders.Moreover, insect growth regulators affect immature stages of insects during the whole molting process (Reynolds 1987), and, thus, adults of non-target species, such as parasitoids and predators, are rarely affected (Bastos et al. 2006).Lufenuron, triflumuron and novaluron, belonging to the Benzoylurea chemical group, as well as diflubenzuron, were also described as harmless to T. pretiosum (Carvalho et al. 2010) and may I Mean + standard error (four replications) of parasitized eggs per female after 96 h (T.podisi) and 168 h (T.pretiosum) of parasitoids exposure to pesticide residues.
# Results by Kruskal-Wallis (H = 21.2010;p-value = 0.0007), followed by the Dunn test (p < 0.05).* Results by Kruskal-Wallis (H = 21.8216;p-value = 0.0006), followed by the Dunn test (p < 0.05).## Results by Anova (F = 17.4249; p-value = < 0.0001), followed by the Scott-Knott test (p < 0.05).** Results by Kruskal-Wallis (H = 17.5981; p-value = 0.0073), followed by the Dunn test (p < 0.05).II Parasitism reduction in comparison to the control treatment.III IOBC classes: 1 = harmless (< 30 %); 2 = slightly harmful (30-79 %); 3 = moderately harmful (80-99 %); 4 = harmful (> 99 %).Means followed by the same letter, in the columns, do not differ significantly.In line with the results obtained in this study, M. anisopliae was previously observed to be selective to adults of T. podisi, and did not affect their parasitism and viability, even when the parasitoids were exposed to direct contact with the fungus (Agüero & Neves 2014).Amaro et al. (2015) assessed the toxicity of entomopathogens on adults of T. pretiosum and found that biological insecticides with the basis of B. bassiana and M. anisopliae do not reduce parasitism and are classified as harmless, meaning that they can be used safely in combination with the parasitoid.In addition, there was no reduction in the number of eggs parasitized by T. pretiosum, even in sprays of entomopathogens in pre and post-parasitism (Potrich et al. 2009).
Neurotoxic insecticides, such as those belonging to the Pyrethroids and Neonicotinoids chemical groups, are generally classified as less-selective compounds to the egg parasitoids Telenomus spp.and Trichogramma spp.(Moura et al. 2006, Giolo et al. 2007, Stefanello Júnior et al. 2008a, Preetha et al. 2009, Carmo et al. 2010, Koppel et al. 2011, Goulart et al. 2012, Oliveira et al. 2013).As reported by Stefanello Júnior et al. (2008a), this toxicity originates from the similarity in the transmission mode of nerve impulses not only between the different orders of insects, but also between the various animal phyla.
There was no significant difference for the number of eggs parasitized by females of T. podisi and T. pretiosum between the herbicide treatments, therefore, all products were classified as harmless (class 1) (Table 3).
The results indicated that the herbicides had little effect on T. podisi and T. pretiosum populations.This is highly relevant, given that weeds are the main limiting factor for rice yield, and weed control is generally performed using herbicides (Sosbai 2014).Low toxicity to egg parasitoids in different developmental stages was also reported for herbicides registered for soybean (Carmo et al. 2009, Magano et al. 2013) and maize (Stefanello Júnior et al. 2008b and2011).
Fungicides toxicity showed no significant difference for the average number of eggs parasitized by females of T. podisi and were classified as harmless to the parasitoid (Table 4).Fungicide sprays in rice fields predominate in the late booting stage of plants (Counce et al. 2000, Sosbai 2014), and the use I Mean + standard error (four replications) of parasitized eggs per female after 96 h (T.podisi) and 168 h (T.pretiosum) of parasitoids exposure to pesticide residues.

Telenomus podisi
In general, adults of T. podisi seem to be more tolerant than adults of T. pretiosum to the toxic effects of insecticides, herbicides and fungicides (Tables 2, 3 and 4).Carmo et al. (2009) also observed this difference, with T. remus being more tolerant than T. pretiosum, both in the egg and pupa phases.
It is important to consider that, based on the methodology used in this study, the adults from the egg parasitoids were subjected to the maximum contact with pesticides, under laboratory conditions and no-choice tests.Under field conditions, pesticide effects may be minimized by their own degradation, or even because parasitoids may escape to untreated areas.Thus, additional tests under semi-field and/ or field conditions, with the pesticides classified as moderately harmful (class 3) and harmful (class 4), recommended by the IOBC (Hassan et al. 2000), should be carried out for the completion of the side effects classification.
Information obtained in this study constitutes a pioneering and important step to choose pesticides selective to egg parasitoids used for the integrated pest management of lowland rice crops.In addition, this information may be useful to integrate management practices for the preservation of these natural enemies in other crops also conducted in this agro-ecosystem, such as soybean and maize.

Table 1 .
Pesticides used in irrigated rice crop on Telenomus podisi and Trichogramma pretiosum.