The leaves of the C. ambrosioides plant were collected in the Medicinal Plants Garden of the Federal University of Lavras (UFLA), Minas Gerais (MG), Brazil. The roots of the P. bipinnatifidum species were collected on a farm in the municipality of Pains, MG, Brazil. The collection site for C. ambrosioides has the following coordinates: 21˚13′49.0476″S, 44˚58′27.4764″W and 933 m altitude. The location for the collection of P. bipinnatifidum was 20˚22′13″S, 45˚65′71″W and 923 m altitude. The ratification of the taxonomic species was performed in the ESAL Herbarium (Herbarium, Department of Biology, UFLA). A voucher specimen of each species was incorporated into the collection of the Herbarium under registration number ESAL 26769 for C. ambrosioides and registration number ESAL 27111 for P. bipinnatifidum.

The extraction of the essential oils was performed at the Laboratory of Organic Chemistry―Essential Oils, Department of Chemistry, UFLA. The material was subjected to hydrodistillation for two hours using a modified Clevenger apparatus coupled to a 5-liter round bottom flask. The hidrolact was collected and centrifuged in a horizontal crosshead centrifuge at 1100 g for 5 min. The essential oil was removed with the aid of a Pasteur pipette, packed in a glass flask that was wrapped with aluminum foil and stored under refrigeration [17] .

The GC-MS analyses were performed on a Perkin Elmer Autosystem XL gas chromatograph equipped with a fused silica column (30 m × 0.25 mm ID, DB-1 film thickness, 0.25 mM; J & W Scientific Inc.) coupled to a Perkin Elmer TurboMass mass spectrometer (software version 4.1). The oven temperature was programmed from 45˚C to 175˚C at a rate of 3˚C/min, and, subsequently, 15˚C/min to 300˚C, where the temperature was held constant for 10 min. The transfer line temperature was 280˚C, the temperature of the ionization chamber was 220˚C, and the carrier gas was helium at a linear velocity of 30 cm/s. The split ratio was 1:40.

The identities of the compounds were determined by comparison of their retention indices with those of the C9-C21 n-alkanes and by comparing the mass spectra with those of standard commercial and reference compounds present in existing oils in the laboratory, as well as by comparison with a mass spectral library developed at the laboratory of the Centro de Biotecnologia Vegetal, Faculdade de Ciências, Universidade de Lisboa [18] .

The essential oils were analyzed by gas-liquid chromatography on a Perkin Elmer 8700 gas chromatograph equipped with two flame ionization detectors, a system for processing data and an autoinjector. Two columns of different polarity were installed with the following characteristics: DB-1 methyl silicone immobilized phase in a fused silica column (30 m × 0.25 mm ID, film thickness 0.25 mm; J & W Scientific Inc.); DB-17HT phenylmethylsilicone stationary phase (30 m × 0.25 mm ID; film thickness 0.25 mm). The oven temperature was programmed from 45˚C to 175˚C at a rate of 3˚C/min, and, subsequently, 15˚C/min to 300˚C, where the temperature was maintained for 10 min. The temperature of the injector and detector ports was 290˚C and 280˚C respectively. The carrier gas was hydrogen, adjusted to a linear velocity of 30 cm/s. The split ratio was 1:50. The percentage composition of the oils was determined by integration of peak areas without using correction factors. The values given represent the average of two injections [18] .

Bioassays with D. speciosa were conducted at the Laboratory of Plant Resistance to Insects, Department of Entomology, UFLA, according to the method described by Assis et al. [19] , with modifications. The bean plants were grown in pots with a capacity of 3 kg using the C Horizon soil (Latossolo Vermelho Escuro) fertilized with 3 g of NPK fertilizer (8-28-16) per pot, equivalent to 450 kg∙ha^?1, as substrate. Four carioca beans seeds were planted per vessel. The pots were kept on benches inside the greenhouse. Thinning was performed 20 days after planting, leaving three plants per pot.

The collection of non-sexed D. speciosa adults was performed in a corn field with the aid of a plastic aspirator attached to a glass container. Subsequently, the beetles were taken to the laboratory, where they were kept for 24 h in an acrylic cage (30 × 30 × 80 cm) and fed with bean plants grown in pots.

A completely randomized design (CRD) with five treatments and four replications was employed. The statistical program used was Sisvar. Data were subjected to analysis of variance, and the average was compared by the Scott Knott Test at 5% probability [20] . The bean plants were sprayed with solutions of essential oils dissolved in aqueous Tween 80 at concentrations of 0.5%, 1.0%, 1.5% and 2.0% using a 10 mL plastic hand sprayer. A 2.0% solution of Tween was applied as the control treatment. The solutions were sprayed to the point that they ran down the leaves and stems of the plants.

The vessels were distributed randomly on benches inside the climate chamber (temperature equal to 30˚C during the day and 25˚C at night, relative humidity, 70% ± 10% and 12 h of light) and covered with organza fabric supported on two iron rods fixed to the substrate of the vessel, forming a cage of approximately 20 cm diameter and 60 cm high. Ten adults per cage were released after treatment [19] . Seven days after the application of the essential oils, the percentage of leaves with injury and the degree of defoliation were evaluated using the AM-300 portable meter (ADC BioScientific Ltd, England). The percent of insect mortality was also determined. In addition, the preference index (PI) with respect to consumption was also calculated according the method of Goeden and Kogan [21] using the following formula: PI = 2A/A + M, where A = degree of defoliation (plants with application of essential oils), M = consumed leaf area (control plant), PI = 1 (neutral), PI < 1 (phagodeterent), IP > 1 (phagostimulant).

The chemical components of essential oil of C. ambrosioides L. and P. bipinnatifidum followed by their calculated retention index and reported retention index with their contents expressed in percentage are in Table 1 and Table 2, respectively.

The principal component found in the essential oil from C. ambrosioides was α-terpinene (40.7%), followed by p-cymene (21.8%) and trans-ascaridol (12.5%) (Figure 1). A predominance of monoterpene hydrocarbons (62.8%) and oxygenated monoterpenes (13.2%) was observed.

The data obtained in this study confirm those of Monzote et al. [22] , who identified α-terpinene, p-cymene and ascaridol as the principal constituents of the oil from C. ambrosioides, with variations in chemical composition from 17.0% to 20.7%, from 20.2% to 21.1% and from 30.5% to 47.1%, respectively. Moreover, Borges et al. [23] encountered terpinolene (69.9%) and ascaridol (17.1%) in the essential oil from this species collected in Brazil. However, α-terpinene (51.3%), p-cymene (23.4%) and p-mentha-1,8-diene (15.3%) were the principal compounds encountered in C. ambrosioides plants harvested in Cameroon (Africa). This difference can be explained by the different locations where the plants were collected, among other factors [24] .

The essential oil from P. bipinnatifidum contained β-bisabolene (65.3%), trans-α-bergamotene (9.9%) (Figure 2), being composed principally of the sesquiterpene hydrocarbon compounds (91.6%) and monoterpene hydrocarbons (3.7%) (Table 2). Although the literature reports on the yield and composition of the essential oil from the roots of P. bippinatifidum are still scarce, α-pinene (13.3%), β-pinene (15.8%), limonene (15.5%), spathulenol (14.2%) and caryophyllene oxide (10.3%) have been found in the essential oils of plants of the Philodendron genus [25] . This composition differs from the data presented in the present study. The discrepancy of this information is probably related to the fact that the species studied are distinct and have a resemblance only at the genus level.

RI_cal = Calculated retention index, N. area = Normalization of the area, v = vestigial (less than 0.01%).

RI_cal. = Calculated retention index, N. area = Normalization of the area, v = vestigial (less than 0.01%).

It appears from the data presented in Table 3 that the use of the essential oils didn’t contribute to a reduction in the number and degree of injuries in bean leaves caused by the chrysomelid because there was no significant difference between the treatments.

With regard to the leaf area consumed, an influence on the feeding behavior of the insect pest was observed for both the essential oils (Table 4). There was a decrease in the consumption of the leaves proportional to the increase in the concentrations of the essential oils.

With respect to the preference index for consumption, antifeedant activity (PI < 1) was observed for the essential oils at all the concentrations tested, i.e., they inhibited feeding by the defoliator beetle (Table 5).

It is known that the eating behavior of insects depends on the integration of the central nervous system with the chemoreceptors located on the shanks, the mouth parts and the oral cavity and insecticides found in plants can act upon the chemoreceptors by stimulating deterrent cells or blocking phagostimulating cells to inhibit feeding [26] . The results are similar to those found by Seffrin et al. [27] regarding the use of natural products that exert antifeedant activity against D. speciosa on the common bean, although the plants studied are different from those of the present study.

The chromatographic analysis allowed the identification of terpenes in essential oils. According to Viegas - Júnior [28] , the ecological importance of these compounds as pesticides plant is well elucidated, as several monoterpenes have been isolated and proved its toxicity against the different insects, ensuring a mortality ranging from 40% to 100%. In addition, other classes of terpenes have been researched for a better understanding of their suppressive activities of feeding and repellency.

It is important to note that the efficiency of extracts or essential oils can be variable, depending on the plant species used and the species of insect pest to be controlled. Although the present study involved another kind of coleoptera, the mortality observed can be compared with that observed for the maize weevil Sitophilus zeamais by Mots, 1855; Coleoptera; Curculionidae) using powders of fruits and the whole C. ambrosioides. It had a highly toxic effect on the beetle [29] .

Moreover, the reduction in the survival of D. speciosa was proportional to the concentrations of the essential oils. There was a significant difference between treatments with regard to the percentage of insect mortality, proving that some insecticidal action of the essential oils existed (Table 6). The insecticidal activity of the essential oils against insects can be attributed to the principal chemical constituents. However, the occurrence of synergism with the minor components may also encourage this activity.

Means followed by the same letter in the same line do not differ at a 5% significance level by the Scott Knott Test.

Means followed by the same letter in the same line do not differ at a 5% significance level by the Scott Knott Test.

PI = 1(neutral); PI < 1 (phagodeterrent); PI > 1 (phagostimulant).

Means followed by the same letter in the same line do not differ at a 5% significance level by the Scott Knott Test.

Thus, the use of the essential oils from the leaves of C. ambrosioides and roots of the P. bipinnatifidum species can be a viable alternative for the control of D. speciosa in the common bean. However, field trials are needed to prove the efficacy of such oils in controlling this chrysomelid under natural conditions.