Feeding preference of Plutella xylostella for leaves treated with plant extracts

Plutella xylostella L. is one of the main agents to cause damages to plants of Brassica genus, provoking negative impacts in cultures. The use of botanical extracts in plants protection has been related in literature, however, their use in the species analyzed in this study is not yet reported. We assessed the effect of aqueous and methanolic extracts of the species: Schinus terebinthifolius Raddi (Pink Pepper), Annona coriacea Mart. (Araticum), Duguetia furfuracea (A. St.-Hil.) Benth. & Hook. (Pindaúva do campo) and Trichilia silvatica C. DC. (Catiguá-branco), occuring in the state of Mato Grosso do Sul and whose feeding preference of P. xylostella larvae of 3rd instar. We intend to answer the following questions: (1) Are the plant species analyzed fagodeterrentes? (2) what type of extract produces the least food preferrence? To answer these questions, we treated cabbage disks with aqueous extracts stored in a refrigerator in periods of 0, 7, 14 and 21 days and the methanolic extracts were treated at concentrations of 0.5 mg/mL, 1.0 mg/ mL, 2.0mg/mL. The aqueous and methanolic extracts of T. silvatica presented the lowest values of feeding preference, 0.113 and 0.06, respectively, compared to other extracts.


INTRODUCTION
The wide-scale use of agrotoxic compounds and synthetic insecticides highlights the need to develop new technologies that will minimize their usage.
Ecologically aware consumers and producers prefer organic crops that are free of agrotoxic compounds and drive research on products that do not harm the environment (Krinsk et al. 2014).
The control of pests through alternative control methods, specifically with plant extracts, is being increasingly studied in order to minimize the chemical impact of insecticides (Mazzonetto et al. 2013).
Plutella xylostella (Lepidoptera: Plutellidae), commonly known as a cabbage moth, is considered to be the main agent responsible for damage to brassicas in commercial plantations worldwide, due to its high feeding rate during the larval period that causes substantial damage to crops, up to 100% IRYS F.S. COUTO et al. yield loss (Hamilton et al. 2005).Furthermore, the average annual cost of pest control exceeds one billion dollars (Talekar 1992, Yang et al. 1994, Haseeb et al. 2004).
For the control of Plutella xylostella (Linnaeus, 1758) (Lepidoptera: Plutellidae) the unselective use of synthetic insecticides has been reported (Boiça Júnior et al. 2005), and has led to the emergence of resistant populations (Talekar andShelton 1993, Furlong et al. 2013).Recent studies have focused on the control of insects using plants (Girão Filho et al. 2014, Trindade et al. 2011, Boiça Júnior et al. 2013).Previous studies have analyzed plant species and brassica cultivars that are resistant to insects (Sarfraz et al. 2010, Niu et al. 2014), and the contents of their aqueous and ethanolic extracts, which interfere with insect feeding and development (Trindade et al. 2008, Bandeira et al. 2013, Boiça Júnior et al. 2013).
Therefore, research on alternative control measures is underway and includes the use of plants, which are important for the sustainable control of insects.Plants tend to have a wide spectrum of activity (such as toxicity, repellence, feeding and oviposition inhibition, and insect growth regulation), and a relatively specific mode of action (Torres et al. 2006, Maia and Moore 2011, Conceschi et al. 2011, Liang et al. 2012, Niu et al. 2013, Amoabeng et al. 2014), ie therefore, each plant species has a specific mode of action on certain target insects according to the chemical compound they have, being innocuous to non-target organisms.Furthermore, their use is safe for larger animals as well as the environment, and the formulations that use plants as the primary material can be easily produced by farmers and small industries (Talukder and Howse 1994).According to Viglianco et al. (2006), most plant-based insecticides do not contribute to the development of resistance or to the emergence of pests.Furthermore, they do not have negative effects on non-target organisms or affect plant growth, seed viability, or the quality of products.
Taking into account the economic importance of brassicas and the negative impact caused by P. xylostella, we aimed to assess the following questions: (1) which of the analyzed species are phagodeterrents and (2) which plant extract is least preferred by the larvae for feeding?
Therefore, we focused on assessing the feeding preference of P. xylostella larvae for samples of cabbage leaves treated with aqueous and methanolic extracts of botanical species widely present in forests and Cerrado of Mato Grosso do Sul.The species include: Schinus terebinthifolius Raddi (pink pepper), Annona coriacea Mart.(araticum), Duguetia furfuracea (A.ST.-Hill.)Benth.& Hook.(pindaúva-do-campo) and Trichilia silvatica C. DC (white catiguá).Until the present moment, such species have not been evaluated according to their role in P. xylostella feeding preference.We hope that, in the future, we be able to use the agricultural properties of these species as an alternative method of plague control and are widely spread in the state.

MATERIALS AND METHODS
Plutella xylostella were reared in a laboratory under 25 ± 1°C, 55 ± 5% RH with a photoperiod of 12 h, using larvae and pupae collected from cabbage fields.
The adults were kept in a plastic cage and fed on a 10 mg/mL honey solution, provided on cotton.To obtain the eggs, a cabbage disc over a damp filter paper was placed inside the cage.
After oviposition, the leaves with eggs were placed in sterilized 30 × 15 × 12 cm plastic vessels to be used as pupation substrate.First, second, third, and fourth instar larvae were fed on organic cabbage leaves, initially sterilized using a solution of 5% sodium hypochlorite and then washed in running water before being placed into the recipients.
The healthy cabbage leaves were placed with the adaxial side towards the plastic recipient and the free abaxial side was used to place the larvae.Next, another cabbage leaf was placed with the abaxial side towards the larvae.The cabbage leaves were replaced daily or as soon as they wilted, ensuring that the larvae were always provided with the best leaves.This process was repeated daily until pupae were formed (Barros et al. 2012).
The species were identified by comparison with exsiccates in the herbarium at UFGD (DDMS).The plant material was kept in the herbarium at the Faculty of Biological and Environmental Sciences at UFGD with the number Annona coriacea: DDMS 4891, Duguetia furfuracea: DDMS 4890, Schinus terebinthifolius: DDMS 4889, and Trichilia silvatica: DDMS 4662.
The leaves were dried in a forced air circulation incubator for three days at a maximum temperature of 40ºC (±1ºC).After this period, the dried leaves were ground to obtain a fine, dry powder.
The aqueous and methanolic extracts were prepared by maceration.The aqueous extracts (AE) were prepared, separately, from 10-g of plant material with 100-mL of distilled water.Next, the material was filtered through a filter paper to remove the solid material.The extracts were stored at -4°C and then tested on days 0, 7, 14, and 21.
To prepare the methanolic extract (ME), a 100g sample of the powder was placed into a beaker with 1000-L of solvent (methanol) and macerated for 7 days.Filtering was performed every two days, for 30 days.The filtered extract was concentrated using a rotavapor set at 60°C, at a reduced pressure.The product obtained in this process was dissolved in distilled water at concentrations of 0.5 mg/mL, 1.0 mg/mL, and 2.0 mg/mL for subsequent use in the tests.
The free-choice tests were performed at 25 ± 1°C, 55 ± 5% RH, and a photoperiod of 12 h.The cabbage fragments (4 cm²) were placed in a Petri dish and distributed in the form of a cross, equidistant to one another, with two fragments immersed in the extract and two immersed in distilled water.Five third instar larvae of P. xylostella were placed on each dish, where larval instar was identified by the length of the larval cephalic capsule (0.33-0.44 mm).
After 24 h, the insect was removed, the leaf area was scanned, and the measurements were made from the images using the software ImageJ (Shneider et al. 2012).Leaf consumption was determined as the difference between the initial leaf area and the remaining leaf area after larval feeding.
The experiment was carried out in a completely randomized factorial design for the aqueous extract (four plants × four periods) and the methanolic extract (four plants × 3 concentrations), both with five repetitions, and each containing 10 subsamples.Analyses of variance were performed on the data; for the aqueous extract, regression analysis at 5% probability was carried out, using the software SISVAR 4.2.For the methanolic extract, the means were compared using Tukey's test at 5% probability using the software SANEST (Zonta 1984).
The response of the larvae to the plant extract was assessed using the feeding preference index (PI) (Kogan and Goeden 1970), being classified as phagostimulant with an index greater than 1, neutral if equal to 1, and phagodeterrent if less than 1, using the formula: PI = 2A/(M + A), where: A = area consumed from the treated discs; M = area consumed of untreated discs.

RESULTS
There was variation in the preference index of P. xylostella during the storage period of the tested aqueous extracts (Figure 1).IRYS F.S. COUTO et al.
The aqueous extract of A. coriacea after days 0 and 7 showed a phagostimulant effect, with the PI gradually decreasing after 7 days changing from a phagostimulant to a phagodeterrent.
For S. terebinthifolius, there was a gradual decrease in the PI during the experimental period, with the aqueous extract acting as a phagostimulant with a PI greater than 1 (1.16) at period 0, and this number reduced by half (0.56) after 21 days.
The aqueous extracts of D. furfuracea and T. silvatica acted as phagodeterrents during the whole storage period, with the minimum storage period of D. furfuracea being 15 days (y = 15.18), and 11 days for T. silvatica (y = 11.19).The aqueous extract of T. silvatica after 21 days had greater PI values when compared to days 7 and 14.
A lower preference index (PI = 0.113) was achieved for all tested aqueous extracts from T. silvatica after 11 days of rearing.
For the methanolic extract, PI was significantly affected by the concentration of the extract (0.7714514; P ≤ 0.05), the plant species (0.7034860; P ≤ 0.05), and the interaction between the extract concentration and plant species (Table I).
There were no significant differences between the feeding preference indices at concentrations of 0.5 mg/mL and 1.0 mg/mL for the methanolic extracts of A. coriacea, S. terebinthifolius, and D. furfuracea.
At a concentration of 2.0 mg/mL, there was no variation between A. coriacea and S. terebinthifolius; however, D. furfuracea had a lower PI than the other species used in this study.
The methanolic extract from T. silvatica showed the best results at all concentrations, and therefore, led to the lowest leaf consumption by P. xylostella.
The aqueous and methanolic extracts of T. silvatica had the lowest PIs, at 0.113 (Figure 1) and 0.06 (Table I), respectively.
With regard to the types of extract, the leaf samples treated with water were preferred by the insects and the methanolic treatment for all treatments was least preferred by the insects.The methanolic extract of T. silvatica resulted in less leaf area being consumed by P. xylostella.

DISCUSSION
The results obtained from the feeding preference tests performed in this study revealed that the aqueous and methanolic extracts of T. silvatica were the most effective antifeedants on third instar P. xylostella larvae.Decreased feeding in insects has been reported in response to different Trichilia plant extracts (Ramírez et al. 2000, Wheeler et al. 2001, Rodriguez et al. 2003).Many species of Trichilia have effects on insects, including inhibition and regulation of growth (Koul et al. 2004), cytotoxic activity (Roel et al. 2000, Souza and Vendramim 2001, Cunha et al. 2008), and insecticidal activity (Matos et al. 2009, Nebo et al. 2010).
Limonoids, also known as meliacines, are the most abundant metabolites in Meliaceae (Viegas Junior 2003).The phagodeterrence of limonoids appears to occur due to the effect of their mouthparts (dissuasion), and also due to their post-ingestion toxicity (Mordue and Blackwell 1993).
Species of the genus Trichilia, studied by Medeiros et al. (2005), demonstrated a deterrent action towards oviposition, with T. pallida being 100% deterrent for oviposition of cabbage moths.Thus, it is possible that species belonging to the genus Trichilia produce the same class of secondary metabolites, and therefore, the deterrence of oviposition in T. silvatica should be assessed.
In this study, treatments using aqueous extracts of the four plant species showed that the production and immediate application of the extracts did not cause obvious antifeeding, with the PI varying between 0.9 and 1.1.The results demonstrated the efficient antifeeding effect of extracts on P. xylostella larvae for the time during which each aqueous extract was stored, with each species having a more efficient storage period, i.e., time during which their PI is the lowest.
All methanolic extracts tested conferred antifeeding effects, and were most efficient at the highest concentration tested.
The larvicidal potential of D. furfuracea has been reported by Rodrigues et al. (2006) and Silva et al. (2013) for other insects.For P. xylostella and D. furfuracea, the avoidance of feeding on discs treated with aqueous and methanolic extracts, may have occurred due to phagodeterrence, probably caused by the presence of secondary metabolites  (flavonoids, alkaloids, acetogenins in annonas, and terpenes) (Isman 2006).Duguetine, which belongs to the class of aporfinic alkaloids, is present in the genus Duguetia, which according to Silva et al. (2007Silva et al. ( , 2009) ) has insecticidal properties.
In a recent study, Krinski et al. (2014) showed the efficiency of annonas as insecticidal plants to different insects.However, for A. coriacea, there have been no reports in the literature for the control of P. xylostella.Acetogenin, a substance present in annonas, acts on insects by targeting the mitochondria and inhibiting NADH, leading to death of these insects (Zafra-Polo et al. 1996).
The bioactivity of acetogenins may vary significantly depending on the species, as well as solvent used in the extraction (Chirinos et al. 2007, Shaalan et al. 2005).With regard to the order Lepidoptera, a recent study by Freitas et al. (2014) showed the control of S. frugiperda using plant extracts by reducing the viability of the larval phase.Extracts of A. coriacea added to the diet of Anagasta kuehniella caused death in 50% of the larvae, and acted as an antifeedant (Coelho et al. 2007).
It is important to note that the aqueous extracts (7 to 21 days of storage) and methanolic extracts of S. terebinthifolius have a phagodeterrent effect on P. xylostella, which is supported by the presence of secondary metabolites, including tannins and flavonoids (isolated form leaves of S. terebinthifolius) (A. Salvi Júnior, unpublished data, Johann et al. 2010).Tannins inhibit digestion by inactivating digestive enzymes creating a tanninprotein complex that is difficult to digest, thereby affecting insect growth and survival (Mello and Silva-Filho 2002).
It was observed that the feeding deterrence of P. xylostella larvae was greater in response to methanolic extracts when compared to aqueous extracts.This may have occurred due to the solvents used during the extraction of secondary metabolites, with methanol extracting a larger number of secondary compounds.R.J.P. Moura (Unpublished data) extracted flavonoids and phenols using different solvents, including water, ethanol, and methanol.Methanol was more efficient for extracting these compounds than water, as it had a higher extractive capacity (Barros et al. 2011).
The using aqueous and methanolic extracts of plant species is an excellent alternative to prevent damage to crops caused by P. xylostella.The storage period of the extract, the solvent used during extraction, and the different concentrations used are factors that influence the feeding preference of P. xylostella.
The results of our study will facilitate future research on the identification of potential bioinsecticidal plant species as well as their chemical characterization, including identification of the compounds or class of compounds responsible for insecticidal activity.
It is concluded that all the species tested are phagodeterrents and the methanolic extract from T. silvatica shows the lowest leaf consumption by larvae of P. xylostella.

Figure 1 -
Figure 1 -Preference index (PI) of Plutella xylostella with regard to plant extract and storage period of each extract at a temperature of 25 ± 1°C, relative humidity of 55 ± 5% with a photoperiod of 12 h.Dourados-MS.(*significance level at 5% probability).

TABLE I Preference index (PI) and area consumed by Plutella xylostella per cabbage under different concentrations of (mg/mL) methanolic extracts.
Means followed by the same lowercase letters per column and uppercase per line, are significantly different as per the results of the Tukey's test at 5% probability.IRYS F.S.COUTO et al.