Four Botanical Extracts are Toxic to the Hispine Beetle, Brontispa longissima, in Laboratory and Semi—field Trials

The potential of botanical extracts such as Celosia argenea L. (Caryophyllales: Amaranthaceae), Ricinus communis L. (Malpighiales: Euphorbiaceae), Mikania micrantha Humboldt, Bonpland & Kunth (Astrales: Asteraceae), and Catharanthus roseus (L.) G. Don (Gentianales: Apocynaceae) for the control of Brontispa longissima Gestro was evaluated in a bioassay and semi—field trial. Dose—response bioassay showed no significant difference in oral—toxicity among the extracts of C. argenea, M. micrantha, and C. roseus to larvae and adult of B. longissima. All extracts tested decreased the hatchability of B. longissima eggs. In particular, the extract of M. micrantha showed higher activity than others at the concentration of 5 mg/mL. In an antifeedant bioassay, the extract of C. argenea showed higher activity against the 1st larvae than that of other extracts (AF50 0.03 mg/mL), and C. roseus showed higher antifeedant activity to the 2nd to 5th larvae and adult of B. longissima (AF50 0.34, 0.33, 0.11, 0.43, and 0.20 mg/mL, respectively). The semi—field trial indicated that all extracts used in this study might reduce the pest population. Extracts of C. argenea and M. micrantha showed higher activities than that of C. roseus and R communis, and the decrease in population was 75.56% and 80.00% (without Abbott's correction) after seven days of treatment, respectively, at a concentration of 20 mg/mL. Therefore, these active botanical extracts may possess potential for use in control of B. longissima.


Introduction
The coconut hispine beetle, Brontispa longissima Gestro (Coleoptera: Chrysomelidae), is a major pest of coconut palm. This pest is native to Indonesia and Papua New Guinea, and widespread in Southeast Asia, Australia, Solomon Islands, and Pacific regions (Hassan 1972;Stapley 1973Stapley , 1980Jones 1986). B. longissima is also a quarantined insect in China (Lu 2004), where it was introduced in 1975 (Cheng 1991;Lau 1991). In 2002, B. longissima was first discovered in Hainan province of China and caused considerable damage to palm plants, especially the coconut. The palms in Guangdong, Guangxi, Yunnan, and Fujian provinces of China were also threatened (Lu 2004). Chemical pesticides are presently the most widely used control method for B. longissima, causing widespread damage to the environment and natural enemies. Moreover, insecticide resistance has developed in B. longissima (Miao 2005), therefore, new methods should be investigated. Plants may serve as this critical resource. In addition to their production of defensive substances with co-evolutionary origins (Kessler 2002;Nina 2003;Consuelo 2004), their botanical extracts are generally considered to be safe for human and animals, and have a variety of control mechanisms (Sivagnaname 2004). Therefore, effects of four botanical extracts on B. longissima were studied in order to evaluate their control potential and provide a new approach of Integrated Pest Management (IPM) of B. longissima.

Materials and Methods
Insects Specimens of B. longissima were collected from the field in Wenchang and Haikou municipalities of Hainan province, China, and reared for at least five generations on core coconut leaves. The insects were maintained in environmental conditions of 25 ± 1 °C and 75-80% relative humidity, with 14:10 L:D photoperiod. Leaves were collected and dried at 60°C in an incubator for two days and then ground into a powder. Three grams of ground powders were placed in containers with 100 mL methanol. In order to prevent solvent evaporation, the containers were covered with parafilm and foil and held in a refrigerator for 24 hours. After being filtered through a 0.22 µm Millipore filter (Millipore, www.millipore.com), the extracts were dried to powder using a rotary evaporator under reduced pressure at 45 °C to remove the solvent (Fliniaux 2004). The extracted powders were stored in a refrigerator. Extract powders were diluted with acetone for laboratory bioassays or in distilled water for semi-field trials.

Oral toxicity bioassay
The oral toxicity of botanical extracts on larvae or adults of B. longissima were investigated using a leaf dipping method (Shukla 2010). The developing core leaves of coconut were cut (10 cm length) and immersed into 10 mL of extract solutions using acetone as the solvent (5-50 mg/mL) for 10 minutes, and then removed and dried for two hours under air flow in a chemical hood. After drying, sets of 30 different instar larvae or adults of B. longissima were placed onto each treated coconut leaf, and then placed in a glass tube (150 mm length, 15 mm diameter). Leaves dipped in 10 mL acetone were used as control. For each treatment, three replicates were conducted. Percentage mortality was recorded after 72 hours; insects that showed no response with head movements or exhibited peristaltic contraction when touched with a fine brush were scored as dead. The larvae that had molted into pupa were scored as alive.

Ovicidal bioassay
To determine the toxicity of different extracts on B. longissima eggs, males and females emerging from the laboratory colony were maintained for 10 days in a plastic mating cage (10 × 8 cm) with developing leaves of coconut as food. One female and two males were then removed from the mating cage and transferred to an ovipositor chamber consisting of a cylinder (15 cm length, 2 cm diameter) covered with a # 200 nylon screen. After eggs were deposited, coconut leaves with 30 eggs were dipped into extract solutions using acetone as the solvent (5 mg/mL) for five sec and then removed and placed under a chemical hood to dry for two hours. Coconut leaves with 30 eggs dipped into acetone were used as control. Three replicates of each treatment were conducted, and all replications were performed at the same time. The mortality ratio was calculated until no more eggs hatched.

Antifeedant bioassay
The bioassay was performed according to the method developed by Omar et al (2007). Extract was diluted by acetone and concentration ranged from 0.5 to 5 mg/mL, with acetone acting as the control. Tender coconut leaves (10 cm length) were immersed into 10 mL of extract solution for 10 minutes, then removed and dried for two hours under air flow in a chemical hood. Leaves immersed with acetone were used as controls. After drying, 10 larvae or 10 adults of B. longissima were placed onto each treated coconut leaf, and then placed in a glass tube (150 mm length, 15 mm diameter). The remaining leaves were removed after 24 hours to calculate the insect's food consumption on a dry weight basis. Initial dry weight of leaves was estimated from the regression of fresh weight to dry weight of additional leaf disks. For each treatment, three replicates were performed.

Semi-field trial
The semi-field trial was conducted at the Coconut Research Institute, Chinese Academy of Tropical Agricultural Sciences, China. The experimental designs were randomized complete blocks. All the plants were grown using the same cultural practices, including drip irrigation and fertilization as needed. The experiment started in early July 2010. There were five treatments (with one control) with six replications per treatment and six plants per replication. Coconut palms 2-3 meters high were chosen for use.
Botanical extracts at a dose of 20 mg/mL diluted with distilled water were used in the semi-field trial, whereas distilled water was used as the control. The treatments were sprayed by a manual sprayer (250 mL per tree). Once the applied extract solutions were dried, the core tender leaves were infested with 30 third instar larvae of B. longissima. In order to prevent the interference of escape and bodily injuries caused by experimental manipulations, the pest number was kept at 30 during the first six hours, as pre-test trials indicated that six hours was insufficient for Table 1. Lethal concentration values (LC50, mg·mL -1 ) for oral toxicity of different extracts to larvae and adults of Brontispa longissima, assessed in laboratory with dipping test.
Significant differences were determined by non-overlapping of 95% confidence limit. The data at the same column followed by the same letter indicate no significant difference at the level of 0.05 by Duncan's new multiple range test. The values were means ± SE. extracts to act. Finally, after 1, 3, 5, and 7 days, the treatment effects were assessed by counting the number of surviving specimens.

Statistical analysis
Egg, larvae, and adult percentage mortality were analyzed, and Abbott's formula (Abbott 1925) was applied to correct percentage mortality if control mortality was between 5% and 20%.
The Probit Regression program in Statistical Package for Social Science (SPSS) version 12 was used to determine LC-P line, lethal concentration values, and corresponding 95% fiducial limits of the upper confidence limit and the lower confidence limit. Significant difference was determined by nonoverlapping of the 95% confidence limit.
In the semi-field trial, the percentage mortality was corrected through Abbott's formula, and the results were analyzed with Duncan's new multiple range test using Statistical Analysis System (SAS) version 8.1. B. longissima that escaped were counted as alive.

Larvicidal and adulticidal bioassays
The response of B. longissima larvae and adults exposed to different botanical extracts is shown in Table 1. The results indicated that extracts of C. argenea, M. micrantha, C. roseus, were toxic to B. longissima except for the R. communis extract, which showed significantly lower toxicity.
No significant difference was shown in toxicities of C. roseus and R. commnuis extract to different instar larvae and adults of B. longissima, based on 95% confidence limits of LC 50 .

Ovicidal bioassay
All botanical extracts used displayed bioactivity toward B. longissima eggs by decreasing hatchability ( Table 2). The hatching rate of normal treatment (control) was 93.33% in six days. After treated with the extracts of C. argenea, M. micrantha, C. roseus, and R. communis, the hatchability decreased. Among the extracts used, M. micrantha showed the highest bio-activity.

Antifeedant bioassay
The antifeedant effects of botanical extracts on larvae and adults of B. longissima are shown in Table 3. The results clearly indicated that the extract of C. argenea showed higher antifeedant effect on 1 st larvae with AF50 0.03 mg/mL, followed by M. Significant difference was determined by non-overlapping of 95% confidence limit. Table 4. Effect of different extracts on B. longissima larvae in the semi-field trial.
The data at the same column followed by the same letter indicate no significant difference at the level of 0.05 by Duncan's new multiple range test. The values were means ± SE. micrantha and C. roseus. However, the extract of C. roseus showed higher antifeedant activity to 3 rd instars and adults of B. longissima, and similar antifeedant effects occurred in 4 th and 5 th instars between extracts of C. roseus and C. argenea, as well as in 2 nd instars of B. longissima. There was a significant difference between the antifeedant effects of C. roseus and M. micrantha to different stages of B. longissim, except among 2 nd instars. Extract of R. communis showed the lowest antifeedant effects in instars of B. longissima compared with others.

Semi-field trial
The total pest population decreased after treatment with the botanical extracts at the concentration of 20 mg/mL (Table 4). Extracts of C. argenea and M. micrantha showed higher control efficiency than that of C. roseus and R. communis.

Discussion
An acceptable insecticide effective kills target organisms, but results in low mortality of non-target organisms. Phytochemicals may serve as suitable alternatives to synthetic insecticides in future because they are relatively safe, inexpensive, and readily available throughout the world. Qin et al. (2010) documented that alcohol extracts of Piper sarmentosum and Artocarpus heterophllus could slow down the development of B. longissima; the generation cycle of B. longissima was found to be 73.58 days when treated with 0.01 g/mL of P. sarmentosum extract, which was delayed compared with the control (43.34 days). Additionally, the fecundity of B. longissima declined in the same study. The essential oil from P. sarmentosum had strong antifeedant and oral toxicity on B. longissima, and our results in the test indicated that all four botanical extracts showed oral toxicity to B. longissima from 1 st instar to adult, especially extracts of C. argenea, M. micrantha, and C. roseus. All botanical extracts used may decrease hatchability; some may even reduce egg hatchability by half, notably the M. micrantha extract. This could prove extraordinarily effective in the control of B. longissima by reducing the pest population before they start to cause any damage. Furthermore, all extracts used showed antifeedant effects, which indicated the various uses of botanical extract in B. longissima control. These results confirmed the conclusions reported by Pang (2000) that plant secondary substances might have oral toxicity and antifeedant activity properties, and act to slow down development in pests. Though the required effective quantity of botanical extract is greater than that of synthetic insecticides, the preparation of suitable formulations using additives might enhance their efficiency. To our knowledge, these extracts have not been tested against other insects, and it is possible that the effects shown on B. longissima may extend to other insects. However, the full potential of the four extracts in reducing damage on the plant and controlling the pest remains unknown.
It is also worth noting that botanical extracts of C. argenea, M. micrantha, and C. roseus showed higher oral toxicity than that of R. communis, and extracts of C. roseus showed better antifeedant activity than others, and higher effects in hatchability inhibition appeared in C. argenea and M. micrantha extracts. Semi-field trial results also indicated that extracts of C. argenea and M. micrantha showed higher activity in decreasing the B. longissima population. All of these differences were perhaps caused by the various secondary substances in different botanical extracts.
Because of the many negative effects of using chemical insecticides in the control of B. longissima and the shifting trend towards to integrated control, our results emphasize the importance of exploring the effectiveness of using botanically-derived materials as supplements to insecticides in B. longissima control. Field evaluations and research into their mode of action and effects on non-target organisms are currently underway. Our study only represents a preliminary screening of target plants, and detailed bioassay-guided phytochemical studies are needed to identify active constituents. The relative environmental safety of identified constituents needs to be studied, since the biological activity of these plant extracts might be due to various extract components including phenolics, terpenoids, and alkaloids, which are present in plants (Park 2000). These compounds may jointly or independently contribute to produce bio-activity against B. longissima.
Further fractionation and compound isolation, particularly of the minor components of botanical extracts, will hopefully reveal a potent phytochemical or synergistic mixture that will be comparable to or even better than synthetic insecticides. In addition, only methanol was used as a solvent for extracting potential bioactive constituents, indicating that only the effects of relatively polar constituents were compared. The effects of B. longissima constituents extracted by other solvents needs to be studied.