Feeding Inhibition by Chitosan on Larvae of Spodoptera litura (Lepidoptera: Noctuidae)

The high usage of synthetic insecticides for controlling Spodoptera litura could be detrimental to the environment, especially on non-target organisms. Therefore, more environmentally friendly pest management techniques should be used, for example, using a natural product such as chitosan. The objective of this study was to understand the effect of feeding inhibition by chitosan on the feeding activity of the third instar larvae of S. litura. The feeding inhibition test was carried out using the choice methods in a factorial completely randomized design (CRD) with seven treatments: i.e. chitosan 5×10, 15×10, 25×10, 35×10, 45×10 ppm and profenofos 0.18×10, 0.34×10 ppm, and one control treatment. The test of the choice methods was carried out by three larval laying positions: (1) between the control and treatment feeds (Position A), (2) above the treated feed (Position B), and (3) above the control feed (Position C), and were replicated three times. The non-choice test was done in a completely randomized design (CRD) with seven treatments plus control and were replicated four times. The results showed that the chitosan in the concentration range of 5×10‒45×10 ppm reduced feeding by S. litura larvae by 2.587 to 34.974% in the choice method, and 11.610 to 50,712% in the non-choice method. This feeding inhibition increased significantly with the increment of chitosan concentration. However, the inhibition effects by chitosan was weaker than the inhibition by profenofos LC50 in both tests at a concentration of 0.34×10 ppm: 44.331% and 62.491% respectively. In conclusion, chitosan with a concentration of 45×10 ppm at all larval laying positions showed the highest value of feeding inhibition activity on the third instar larvae of S. litura compared to other chitosan concentrations in both methods.


INTRODUCTION
The army worm, Spodoptera litura Fabricius (Lepidoptera: Noctuidae), is an important horticulture pest on approximately 120 types of plants (CABI, 2017). To this date, syntetic insecticides are still the main option to manage S. litura in the field. However, continuous and overuse of insecticides have detrimental effects on the environment (Untung, 1984). As reported by Irawan (2017) and Setiawati, et al. (2015), the application of λ-sihalotrin 25 EC (25 g/L) and carbofuran (Furadan 3 G) on long beans had negative effects on the predatory spider, Pardosa pseudoannulata, and reduced soil arthropod species richness by 30.67-34.32%. Thus, alternative management techniques that are environmentally friendly are required.
Beside having insecticidal properties, it is confirmed that at concentration of 0.9%, chitosan affects feeding preference of Aphis craccivora on long bean (Megasari et al., 2014). In addition, if it is modified with metals, such as nickel (Ni) and mercury (Hg), it has feeding inhibition activity on S. littoralis larvae (Badawy & El-Aswad, 2012).
Current research on the use of chitosan have heavily focused on its effects on plant pathogens and sucking-type insects, while chewing-type insects have been less studied. Therefore, the goal of this study was to test the feeding inhibiton effects of chitosan against 3 rd S. litura larvae, the most voracious stage.

Spodoptera litura Rearing
S. litura were collected from sunflower plants located in Sinduharjo Village, Ngaglik District, Sleman Region, Special Region of Yogyakarta and also obtained from the Indonesia Sweetener and Fiber Crops Research Institute (Balittas), Malang. Both colonies were reared at the Laboratory of Entomology, Department of Plant Protection, Faculty of Agriculture, Universitas Gadjah Mada, Yogyakarta since 2017.
Larvae of S. litura were kept in plastic containers covered with mess clothes (18×11×10 cm) and fed with artificial diets (Singh & Moore, 1985; with slightly modification; Table 1). Containers were cleaned daily and diets were change every 2 days or based on larvae needs. Pupae were collected and placed in a petri-dish (diameter = 9 cm; height = 2 cm) with a wetted cotton to maintain humidity of the container. Petridish was placed in a transparent plastic container (diameter = 17 cm; height = 19.5 cm) that was closed using mess clothes, and enabled pupae to hatch into adults. Adults were fed with 10% honey solution until they laid eggs.

Chitosan Preparation
Pure chitosan were obtained from Black tiger shrimp shells (90% of purification) provided by CV. Bio Chitosan Indonesia (Cirebon, Indonesia). Chitosan were disolved into 1% acetate acid before used in tests. Ningsih et al.: Feeding Inhibition

Choice Test of Feeding Inhibition Activity
Choice-test of feeding inhibition was set as a Factorial Complete Randomized Design (CRD) to determine feeding activity of S. litura in similar condition to field condition. Diets were made into cubicle shapes (1× 1× 1 cm; weight ±3 g). The diets were than immersed for 10 seconds into chitosan solutions at 5 different concentration, including 5×10 3 , 15×10 3 , 25×10 3 , 35×10 3 , and 45×10 3 ppm, and profenofos solutions at concentrations of 0.1810 3 ppm (LC 20 ) and 0.34×10 3 ppm (LC 50 ) as a comparison, while diets used as controls were immersed into 1% acetate acid solutions. Diets were then air dried on filter paper, and noted as weight before larval feeding. Treated and control diets weres placed respectively on the left and right of plastic containers (diameter = 10.5 cm; height = 7.5 cm) and labelled based on treatments.
Tests were done by placing ten 3 rd instar larvae in 10 plastic containers (diameter = 10.5 cm; height = 7.5 cm) that were separated by treatment. Larvae were placed at 3 positions, which were between control and treated diet ( Figure 1A), on top of treated diets ( Figure 1B), and on top on control diets ( Figure 1C). Each treatment were replicated 3 times and as much as 630 3 rd S. litura larvae were used in this choice test. Feeding inhibition activity was calculated using the formula from Hassanali & Bentley (1987). Feeding inhibition activity using the choice test: Feeding inhibition activity (FI) (%) = C-T × 100 C+T C = Diet weight before larvae feeding-diet weight after larvae feeding on the control diet T = Diet weight before larvae feeding-diet weight after larvae feeding on the treated diet

No-Choice Test of Feeding Inhibition Activity
This tests was done to confirm results from the choice test by forcing larvae to one type of diet. Procedure used were similar to the choice test and set as a Completely Randomized Design (CRD).
Diets treated with chitosan, profenofos, and control diets were placed in separated plastic containers (diameter = 6.5 cm; height = 4.8 cm). Each experimental unit consisted of 10 3 rd instar S. litura larvae placed into separated containers. Larvae placement only used one treatment, which was on the designated diet ( Figure 2). Each treatment was replicated 4 times resulting in 320 larvae used in this experiment. Remaining diets were then weighted. Diets consumed by larvae were determined by calculating the differences of weights before and after larval feeding. Observations were done several times over 48 hours.
To calculate water moist loss from diets due to evaporation was done by using a parallel diet that was not fed to larvae. The diet was then weighted and differences were calculated. Feeding inhibition activity was calculated using the formula from Hassanali & Bentley (1987). C = Diet weight before larvae feeding-diet weight after larvae feeding on the control diet T = Diet weight before larvae feeding-diet weight after larvae feeding on the treated diet Feeding inhibition activity (FI) (%) = C-T × 100 C

Statistical Analysis
Inhibition percentages of the choice tests were tested using an ANOVA as a Factorial CRD with two factors including solution type concentration and larvae placement position. Significant differences between treatments were then tested used a post-hoc Tukey HSD test at α = 5%. All statistical tests were done using R-3.4.4, 2018.

Choice-Test of Feeding Inhibition Activity
Results showed that feeding inhibiton was significantly affected by compound types and concentration, but not by the larva placement position and there were no interactions between these 2 factors (Table 2). Feeding inhition increased respectively to the concentration of chitosan and profenofos regardless of the position of the larva.

No Choice-Test of Feeding Inhibition Activity
Feeding inhibition using the no-choice test had similar results with the choice test. Higher chitosan concentrations had significantly higher feeding inhibition effects. Feeding inhibition of chitosan and profenofos showed similar patterns (Table 3).
Higher chitosan concentration created thicker film coatings that covered the diets and disrupted larval feeding. Decreased eating can occur due to  the deterrent effect and sublethal concentrations that reduce feeding rates. Chitosan is a natural biopolymer with semipermeable characteristics and creates coating films that have been reported to be used as seed coating that effectively repel Agrotis ipsilon, soybean pod borer, and soybean aphid feeding activity by 82.89, 87.24, and 80.21% respectively at concentrations of 5% (Zeng et al., 2012). However, this study was only able to decrease feeding rates by 35.0% at the choice test and 50.7% at the no-choice test when using chitosan concentrations of 45×10 3 ppm. These values were not higher than the profenofos LC 50 at 0.34×10 3 ppm treatment regardless with choice or no-choice tests. This implies that chitosan did not effectively inhibit feeding activity of 3 rd instar S. litura larvae. Antifeedant is a compound applicated at low concentration, but is able to highly inhibit feeding rates (Purrington, 2017). Polyphagous insects, including S. litura, have the ability to adapt to antifeedant compounds. In addition, this compound does not have high enough toxicity so that the adaptation by larvae occurs more quickly. Since chitosan is not able to reduce the feeding rate of S. litura, this also means that chitosan has no effect on its development and growth. In this phase S. litura will adjust to chitosan, so that its growth and development are not disturbed.
Results form this study were different from previous studies using pure chitosan (molecule weight 2.27×10 5 g/mol with concentration of 4 g chitosan/kg of artificial diet) and chitosan with nickel (Ni) and mercury (Hg) addition, which were able to inhibit S. littoralis feeding by 76, 90.2, and 86.8%. This feeding inhibition of pure chitosan was not significantly higher than chitosan with Ni and Hg addition (Badawy & El-Aswad, 2012). Ningsih et al.: Feeding Inhibition by Chitosan on Larvae of Spodoptera litura (Lepidoptera: Noctuidae) 185 a b According to Zhang et al. (2003), chitosan have insecticidal properties against lepidopteran and homopteran insects. Chitosan activity was higher in Plutella xylostella, about 72% compared to Helicoverpa armigera and S. exigua which were only 40% at a concentration of 3g/L, and Rhopalosiphum padi, Metopolophium dirhodum, and A. gossypii, i.e. 60-80% at a concentration of 600-6000 mg/L. In addition, application of chitosan at concentrations of 0.9% on long beans affected A. craccivora feeding preferences based on low populations and no lethal effects on this species is recorded (Megasari et al., 2014). This may be caused by chitosan's ability to induce resistant mechanisms of plants, such as the accumulation of phytoalexin, PR (pathogenesis related), proteinase inhibitors, and various enzymes including peroxidase. Peroxidase can induce lignification of plant's cell walls that act as a physical barrier that reduce herbivor's preferences (Hadrami et al., 2010).

CONLUSION
Feeding inhibition of 3 rd instar S. litura larvae due to chitosan increased respectively with the increase of concentration (5×10 3 -45×10 3 ppm). Chitosan concentration of 45×10 3 ppm inhibited feeding activity the most at all larval posistion placement compared to other concentrations regardless the choice or no-choice test.