The antifeedant activity of natural plant products towards the larvae of Spodoptera littoralis

Worms are becoming difficult pests to control since many insecticides are now forbidden and the present alternatives are insufficiently effective. Antifeedants of plant origin may have potential against such pests. In this work, several diterpenoids, flavones and coumarins extracted from plants were assayed as antifeedant agents against Spodoptera littoralis larvae. Among these compounds, four diterpenes (eriocephalin, salviacoccin, aethiopinone and oxocandesalvone) and four coumarins (oxypeucedanin, xanthotoxin, isoimperatorin and prangol) showed significant activity, whereas all the flavones tested were inactive. Some aspects of the structure-activity relationships of these compounds are briefly discussed. Additional key words: coumarins, diterpenes, flavones, higher plants, secondary metabolites.

Plants are important sources of natural products with the potential to be developed as commercial active ingredients for pest control.Antifeedants have been proposed as an alternative to synthetic insecticides (Belles et al., 1985).Clerodane diterpenoids are well known insect antifeedants isolated from plants of the genera Ajuga, Leonorus, Salvia, Scutellaria, Stachys and Teucrium (family Labiatae) (Merrit and Ley, 1992;De la Torre et al., 1994;Rodríguez-Hanh et al., 1994;Bremner et al., 1998).The most potent neoclerodanes are jodrelin and scutalpin, which are isolated from Scutellaria (Anderson et al., 1989).The coumarins have been much less studied although they too have been cited as insect antifeedants (Addor, 1994;Stevenson et al., 2003).In addition, a few papers document the antifeedant effects of flavones (Sharaby and Ammar, 1997) and other diterpenoids, for example those found in Plectranthus leaves (Wellsow et al., 2006).
This paper reports the antifeedant activity against S. littoralis of natural plant compounds belonging to different classes -clerodane, abietane, labdane and rosane diterpenoids, flavones and coumarins-with the aim of providing knowledge that may be of use in the development of new agrochemical products.
The phytochemicals used in this study are listed in Table 1 along with their plant sources.
Spodoptera littoralis was reared on a bean-based diet at 26 ± 1°C under a 16L:8D photoperiod.Larvae were originally collected several years ago from an alfalfa field in Totana (Province of Murcia); since then, this population has been reared in the lab to supply insects for research purposes.
Experiments were performed in 9 cm diameter Petri dishes each containing a glass fibre disc (diameter 2.1 cm) (Whatman GF/A).One hundred microlitres of a 50 mM sucrose solution plus 100 µL of a 100 ppm solution of the test compounds were applied to different discs (control discs received only the sucrose solution).A sixth instar larva was then added to each Petri dish.The discs (after the application of the products) were weighed before and after (18 h) the bioassay.Ten replicates were prepared for each compound assayed.
Following the methodology described by Simmonds et al. (1990), antifeedant indices (AI) were calculated according to the formula: where C = the weight (g) of the control disc consumed, and T = the weight (g) of the treated disc consumed.The AI ranged from -100 to +100.If a value of AI > 0 was returned, the compound was deemed to be an antifeedant; if a result of AI < 0 was returned, the compound was regarded as a phagostimulant.A mean AI was calculated for each assayed phytochemical.Comparisons between the eaten weight of the treated and control discs were made using the non-parametric Wilcoxon test.
The most active phytochemical was then assayed at different concentrations (4 doses, 2 replicates per dose, 100 larvae per replicate).Probit analysis (Robertson and Priesler, 1992) was then used to determine the doses with an antifeedant effect (AI ≥ 50) towards 10, 50 or 90% of the larvae.
Table 2 shows that the clerodane diterpenoids eriocephalin, salviacoccin and salvifarin, were active at 100 ppm, while the other compounds of this class had an AI value of around zero.The most active product (AI = 62.18) was eriocephalin.Overall, the abietane diterpenoids had less of an antifeedant effect, although aethiopinone and 2-oxocandesalvone A returned statistically significant results (AI = 29.82 and 23.02 respectively).In addition, larval feeding was reduced by the intake of 12-O-demethylcryptojaponol.As for the labdane and rosane diterpenoids tested, positive AI were obtained, although they were not much greater than those for the previously mentioned compounds (Table 2).
Eriocephalin, a compound isolated from Teucrium eriocephalum Wk, showed the greatest antifeedant effect.Table 3 shows the results of the probit analysis.The ED50 was 43 ppm, i.e., at this dose, an antifeedant effect (with AI ≥ 50) would be seen against 50% of larvae.European and Asian species of Salvia contain more abietane than clerodane diterpenoids (Rodríguez-Hanh et al., 1994).Aethiopinone is a strong anti-inflammatory substance (Hernández-Pérez et al., 1995) and possesses antimicrobial activity as well as cytotoxic activity towards cultured KB human carcinoma cells (Hernández-Pérez et al., 1999).Eriocephalin is known to be very active against S. littoralis (Simmonds et al., 1989;De la Torre et al., 1994).According to Geusken et al. (1983), substitutions in the decalin ring (4α,18-oxirane) and a β-substituted furan at C-12 are key structural features in this activity.
In contrast to the results reported here for S. littoralis, Ortego et al. (1995) reported that teucvin (a 19-nor   neoclerodane) at 100 ppm caused a significant reduction in feeding in Leptinotarsa decemlineata larvae.Salviacoccin and salvifarin have a saturated and α,βunsaturated γ-lactone group respectively.According to Simmonds et al. (1996), this explains the activity of these compounds extracted from Salvia.These authors obtained an AI of 59 (at a dose of 100 ppm) for salviacoccin -a stronger antifeedant effect than that seen in the present experiment (AI = 31.28).
The flavones had no significant effect on the feeding behaviour of the larvae (Table 2).
The coumarins were active against S. littoralis.Oxypeucedanin and xanthotoxin gave AI values of over 40 (Table 2).A remarkable antifeedant activity was recorded for isoimperatorin (AI = 31.89)and prangol (AI = 26.78).The coumarins are a class of naturally occurring antifeedants that have been less studied than the neoclerodanes.Calcagno et al. (2002) reported a deterrent effect of xanthotoxin on S. littoralis, with synergistic effects in mixtures with imperatorin.
When the structures of active and inactive coumarins, such as isoimperatorin and imperatorin or xanthotoxin and bergapten are compared, substitution groups at different positions (C-4 or C-9) account for the differences in antifeedant activity (Table 1).
In summary, the active diterpenoids isolated from species of Teucrium and Salvia (Labiatae) and a number of coumarins from Cachrys trifida (Umbelliferae), appear to have antifeedant potential with respect to S. littoralis.More research is required in this area to help find alternatives to banned pesticides.a,b,c Doses [ppm; solutions (100 µL) were applied to 2.1 cm diam.discs] with antifeedant effect (AI ≥ 50) towards 10%, 50% and 90% of larvae respectively.

Table 1 .
Phytochemicals tested a All species are members of the family Labiatae, except for Cachrys trifida Miller which belongs to the family Umbelliferae.