1. Introduction

The researchesdesigned at identifying new toxic substances in order to control agricultural plagues are constant in the activities related to management and integrated control of plagues nowadays. This is partially a resultof the knowledge of the negative impactof the chemical products used in the ecosystem as well as the possibility of evolution of the resistance of the target insects to the products used, whether these are chemical and/or biological. A great number of these researches aredirected to the evaluation of toxicity of secondary metabolite of plants, such as phenols, alkaloids, glucosinolates, cyanogenic glycosides and saponins.¹More than 2000 species of plants are known to possess some insecticidalactivity. In many cases plants have a history as traditionalmedicines or are used to kill or repel insects. In spite of the various studies available about the entomotoxic effects of extracts of different species of plants, their application is still incipient.²

The saponins are glycosides with detergent properties, due to the presence of hydrophobic (aglycones) and hydrophilic (sugar) components with wide distribution among plants which may be toxic forherbivorous arthropods such as mites, beetles and Lepidoptera, among others.^3-7Their insecticidal activity may be associated to the ability of producing alterations in the feeding behaviour, in the molting process, of interacting with hormones that regulate the growth and causing death in the different stages of development.^8-11Their wide spectrum of action reaches a great number of herbivores and its amplitude of physiological impacts makes these substances an excellent model for the study of the effect of natural substances with insecticidal activities.⁸

The passion fruit,Passifloraalata, is one of the main species of economic importance of the gender cultivated in Brazil, which is used mainly as fruit and for juice and sweets production as well as a medicinal plant.^12-16Thesaponins are the main substances of its secondary metabolism. Five types of saponins were isolated and identified in its leaves: one is a steroid and four are triterpenic_(3-6).^17-18

The fall armyworm Spodopterafrugiperda (J. E. Smith) (Noctuidae) is a polyphagous insect, using more than 50 species of plants as food supplies, distributed in over 20 families,¹⁹ predominantly in grasses, among them corn and rice, which are important cultures for Brazilian and world economy. In rice,S. frugiperda is considered a pestof the initial phase, since it attacks the seedlings in the beginning of their development, feeding on the leaves and cutting the new stalk close to the ground,²⁰causing losses of 14 to 24% of the grains. Depending on the population level, the destruction of the crop can be total.²¹In maize, the attack can occur from the seedling stage until the tasseling and the formation of spikes. In the later attacks specimens between the stalk and the cob can be found, where they penetrate the female inflorescence destroying the grain.²²

The effect of saponins extract of some plants on the development of the Spodopteraspecieswas demonstrated in studies with S. litura.^2,4,23-27The results of these studies indicate alterations in the feeding behaviour, lengthening of the larvae and/or pupa stages, increased mortality and reduced fertility.

The study of plants with potential insecticidal activity is important to discover new active molecules, which may be isolated from plants or synthesized, or a molecule-prototype for structural changes to obtain a more active compound.

This work evaluated the effects of saponins extracts of PassifloraalataonSpodopterafrugiperda, identifying the sub lethal and lethal effects of these substances in different development stages of this species.

2. Material and Methods

2.1. Preparation and Characterization of the Saponin Extracts

To obtain saponin extract the dried leaves were submitted to maceration with 70% ethanol. The ethanol was filtered and placed in a rotary evaporator; the aqueous residue was extracted successively with chloroform, ethyl acetate and n-butanol. The evaporation of n-butanol fraction resulted in a fraction consisting mainly of saponins, which will be called henceforth saponin extract.¹⁷Thesaponin extract is characterized as a yellow-orange powder.

2.2. Obtention and Rearing of Insects

The establishment of S. frugiperda began with immaturelarvae collected in rice plantations, with the collaboration of researchers at the Estação Experimental do Arroz (EEA) of the InstitutoRiograndense do Arroz (IRGA), located in Cachoeirinha, RS. The insects were reared in the room for the creation of insects of the Laboratory of Microbiology of the University of Vale do Rio dos Sinos (UNISINOS), heated to 26°C, 75% RH and 12h photo phase. The adults were kept in plastic cages, fed with a glucose solution of 10%. Eggs were collected three times a week and put in gerbox with diet. Five or six days after hatching, the larvae were individualized in plastic cups, and maintained with artificial Poitoutdiet,²⁸ where they remained until they became pupae. As a routine procedure adopted, the pupae were identified, separated by sex, kept in sleeves with moistened filter paper and covered with tulle, until emergence of adults.

2.3. Bioassay

For the test with S. frugiperda nine solutions with increasing concentrations of saponin extract were used due to the 2x (312, 625, 1250, 2500, 5000, 10000, 20000, 40000 and 80000ppm) and a control of sterile distilled water. A solution of 20ml was prepared for each concentration, which, after use, was stored at 4°C. In this case, the extracts were diluted in sterile distilled water. The second instar larvae of S. frugiperda were maintained for seven days in mini acrylic plates (35mm diameter) containing Poitout diet, where 100μL of the solutions were applied on the diet. After thisthey were transferred to plastic cups containing only artificial diet, where they were observed until emergence of adults for the evaluation of possible sub lethal effects (Figure 1). Damages were evaluated under stereoscopic microscope and classified according to intensity (low, medium and high) (Figure 2).

Bioassays were kept at 26°C, 75% RH and 12h photo phase. Five repetitions of 30 larvae/concentration, totalizing 1500 insects were performed.

2.4. Data Analysis

Data were tested for normality using the chi-squared and Shapiro-Wilks and the homogeneity with the tests of Hartley and Bartlett. When these criteria were satisfied, they were compared by an ANOVA followed by multiple tests of Tukey. Otherwise, they were compared by Kruskal-Wallis followed by Dunn's multiple test. Statistical analysis was performed with the aid of TOXTAT Software version 3.3.²⁹To compare mortality rates between larvae and pupae of S. frugiperda the Student t test was conducted for different means and to analyse the deformation of the same species a simple linear regression was made.³⁰

3. Results

Figure 1. Characterization of the sublethal effects observed in Spodopterafrugiperda. A) Larva with abdominal distension; B) Larva with the point of necrosis in the integument; C) larva with incomplete molting process; D) Pupa with globular evagination in the integument; E) dead adults emerging from pupae; F) Pupa with failure in the process of sclerotinization of the integument; G) Pupa with morphological features of larvae; H) Pupa with the point of necrosis in the integument; I) Adult with the wings unfolded e J) Adult with wings slightly distended and pupae attached on the wing

Figure 2. Characterization of the intensity of deformation in Spodopterafrugiperda. A) Larva without deformation; B) Pupa without deformation; C) Adult without deformation; D) Larva with deformation of low intensity; E) Pupa with deformation of low intensity; F) Adult with deformation of low intensity; G) Larva with deformation of medium intensity; H) Pupa with deformation of medium intensity; I) Adult with deformation of medium intensity; J) Larva with deformation of high intensity; L) Pupa with deformation of high intensity e M) Adult with deformation of high intensity

Table 1. Summary results of the experiments indicating mortality levels by treatments and stage of development Treatments Larvae Mortality Pupae Mortality Cumulative Mortality Larval Period - days Pupal Period - days total % total % total % mean mean Control 5 3 0 0 0 0 17.61 10.2 Treatment 1 58 37 26 17 84 56.0 29.18 10.69 Treatment 2 44 29 23 15 67 44.7 20.23 9.79 Treatment 3 59 39 25 17 84 56.0 26.41 9.92 Treatment 4 59 39 22 15 81 54.0 21.1 10.51 Treatment 5 54 36 39 26 93 62.0 25.59 8.87 Treatment 6 40 27 31 21 71 47.3 18.78 9.57 Treatment 7 48 32 39 26 87 58.0 20.19 9.48 Treatment 8 44 29 24 16 68 45.3 18.47 9.31 Treatment 9 30 30 39 26 69 46.0 18.78 9.75

Table 2. Summary results indicating the number of individuals with deformations by stage of development and treatments Single Deformations MultipleDeformations T1 T2 T3 T4 T5 T6 T7 T8 T9 T1 T2 T3 T4 T5 T6 T7 T8 T9 Larvae 1 1 3 4 9 7 8 8 10 0 0 0 0 0 0 0 0 0 Pupae 2 5 4 4 6 3 4 6 1 1 1 0 0 2 6 9 2 1 Adults 20 32 32 25 36 24 30 29 42 0 0 1 0 1 10 2 2 3 Total 23 38 39 33 51 34 42 43 53 1 1 1 0 3 16 11 4 4

The quantitative and qualitative results of the experiments are indicated on tables 1 and 2, respectively

Considering the total mortality (larvae and pupae) all treatments differed statistically from control but not among them (Figure 3) (p<0.05). Comparing the mortality of larvae with the mortality of pupae in each treatment, it was observed that the rate of larval mortality was higher than that of the pupae from the control concentration until the concentration 2500ppm. At higher concentrations (5000 to 80000 ppm), these rates showed no significant difference (Figure 4). There were 441 of larvae and 268 of pupae deaths, i.e. 8.1% and 29.4% of organisms exposed to saponin, respectively, totalizing 17.9% of the insects used in bioassays.

Of the 441 larvaethat died during the tests, 121 (27%) died with some deformity, of which 45 (37.2%) died during the process ofmolt, unable to release in whole or in part of molting, 8 (6.6%) had points of necrosis and 5 (4.1%) some parts of the body distended

Of the 268 dead pupae, 107 (39.9%) died with some deformity, of which 27 (25.2%) had retention of morphological characters of larvae, 88 (82.2%) died during the emergence of the adult failing to release in whole or in part, 5 (4.7%) had deformities in the form of globular evaginations of the cuticle, 23 (21.5%) had points of necrosis in the cuticle and 11 (10.3%) sclerotinization failures, leaving the internal organs exposed. Some specimens had more than one type of deformation, such as evaginations and points of necrosis (2.93%), or points of necrosis around the failures of the cuticle (0.97%). Often, when reviewing the experiments, the pupae with failure sclerotinization were found still alive, being dead in 70% alcohol immediately to avoid the suffering of the individual. The main deformation observed in 599 adults obtained after the bioassays was in the region of the wings, where 200 (33.4%) insects showed little or no distension of the wings.

A total of 1025 insects (larvae, pupae and adults) were observed withdeformations (76% of organisms) (Figure 5). These organisms were screened according to the intensity of their deformations: 612 (59.7 %) had deformities of low intensity, 144 (14.05%) of average intensity and 269 (26.2%) of high intensity (Table 3)

The linear regression obtained from the number of deformities observed in insects exposed to the saponin extracts, was a straight up line, which shows that there is a tendency to increase the number of deformed insects as the concentration of extract saponins increases (Figure 6)..

Figure 3. Total mortality (mean ± standard error) of Spodopterafrugiperda after bioassay of ingestion in different concentrations of saponin.Columns followed by different letters are significantly different (ANOVA, followed by Tukey's multiple tests α=0.05)

Figure 4. Mortality of larvae and pupae of Spodopterafrugiperda after ingestion test of saponin at different concentrations. Columns followed by different letters differ significantly at a given concentration (Student's t test, α = 0.05)

Figure 5. Deformations (mean ± standard error) observed in Spodopterafrugiperda, after ingestion test of saponin, divided into three categories (low, medium and high), by concentration of saponin

Figure 6. Linear regression of the number of deformations observed in Spodopterafrugiperda after ingestion test of different concentrations of saponin

Table 3. Summary results of the damage levels by each treatment Damage Level T1 T2 T3 T4 T5 T6 T7 T8 T9 Low 82 69 65 65 62 61 59 74 75 Medium 19 20 13 9 11 18 16 16 12 High 8 9 26 24 37 39 46 32 44 Total 109 98 104 98 110 118 121 122 131

In summary the deformities were observed in 76% of organisms exposed to saponin. These deformations were linked, primarily, to the process of molt, as 45 larvae of S. frugiperda died during molting process, failing to release in whole or in part of molted cuticle. Besides the problems related to the process of molt, the organisms that had points of necrosis in the larvae stage were located mainly in the thorax and the pupae stage and adults, in most cases, in the region of the wings. Moreover, problems were observed in sclerotinization of pupae, which opened up large holes in the region of the wings, leaving the interior of organisms exposed. In organisms of control treatment no deformities were observed.

4. Discussion and Conclusions

This study identified the impact of saponins extract on the different stages of development of S.frugiperda.Besides the mortality of larvae and pupa, the presence of structural deformities was observed in all stages. When considered the total impact of the treatments 47.4%of the adults emerged without structural problems identified by optical microscopy. The effects of the saponin extract on S. frugiperdamay be associated to a deterrent toxic action and/or by its interactions with substances responsible for the different stages of development of the specie.

The deterrent effect reduces the consumption of food producing nutritive deficiency causing deficient growth or deformities, which may lead to death or inhibit the progress for the next stage.^26,27,31,32Among the alterations observed, the presence of necrosis points stood out, sclerotinization problems and retention of larvae characters and tumors.These results are similar to the ones registered,²¹for the biotypes of corn and rice of S. frugiperdausing different extracts of different species of plants.It is important to highlight that alterations in the length of the larvae and pupal stages were not observedin this study, as registered for S.littoralis²³e for S. frugiperda.³³The differences in these results may be associated to the variation in the composition of saponins among the plant species used for the acquisition of the extracts.^3,34

The physiological basis of the toxicity of saponins has not been fully elucidated, but we know that there is a great interaction between them and the cell membranes, with effects on the hydrophobic-lipophilic balance and permeability of these because they are capable of forming complexes such as sterols, for instance, cholesterol.^10,11,35The toxicity of saponins on arthropods may derive from their ability of interacting with free steroids of the intestine and/or of inhibiting the digestive proteases, reducing the rates of digestion and absortion.^8,35,36

The mortality and deterrence caused by saponin were also observed in experiments that evaluated the effects of different types of saponin for aphids,^38,39,42nematodes,^40-42beetles^31,43,44 and Spodoptera sp.^2,5,6,25,27

A fact that drew attention was the appearance of specimensin the intermediary period between larvae and pupae (1.8% of exposed organisms). Apparently, the organisms began the process of becoming pupae and could not complete it, which led to death. The appearance of intermediate individuals between pre-pupae and pupae can occur when the activity of juvenile hormone, which controls the metamorphosis, is affected.³²

Furthermore, the metamorphosis may have been affected because of saponins form complexes with sterols, like cholesterol, both in the intestine in cell membranes of insects. These complexes formed by saponin make sterols unavailable, influencing the synthesis of ecdison, one of the hormones involved in the process of molt. It has been shown that insects are unable to produce these sterols, removing them entirely from the diet.^{10,11,34,36,39} The statements of these authors may account for the large number of deaths of larvae and pupae during the processes of molt and emergence of pupae malformed, beyond the failures of sclerotinization observed.

In conclusion the mortality of S. frugiperda showed significant difference between control and other treatments. It was observed that the death of the specimens was not immediate after the ingestion of saponin, but slow and gradual, which apparently occurred under the influence of saponin in the metabolism of the larvae, pupae and adults. These results indicate the potential of saponins present in P. alataas a control agent of S.frugiperda. This potential needs to be better evaluated, especially if the facility of raising and large scale production of the plant is considered.

ACKNOWLEDGMENTS

We acknowledge the CNPq by the financial support to do this research at Biology Postgraduate Program of UNISINOS