go to previous section of this publication

There was no evidence of adverse side-effects of neem on epigeal arthropods. In a prolonged feeding trial, in which high-dose treated food (dipped in 1 % neem oil solution) was offered ad libitum for 10 days, decreased fitness and reduced food uptake was observed in Adesmia cothumata orientalis as well as in Pimelia raffrayi raffrayi Senac, but there was no increase in mortality. In semi-field trials at a common application rate of 1 l/ha neither unclarified neem oil (0.04% azadirachtin, field dose 0.4 ml a.i./ha) nor enriched oil (5% azadirachtin, field dose 50 ml a.i./ha) had any effect on the activity or mortality of the same species. In contrast, Pimelia angulata tschadensis Koch showed a pronounced hyperactivity (Fig. 1) at a 10-fold higher application rate (10 l/ha) of 0.1% oil (field dose 10 ml a.i./ha). Likewise, the field trials (Table 4) revealed a significant increase in the activity abundance of some tenebrionid species and other arthropods. However, due to an over proportional catch during the pre-spray period in the control plot, the effect was exaggerated (Fig. 2). This bias becomes evident for instance in the case of the litter-dwelling anthicid beetle Notoxus sp. (Fig. 3). As a consequence, even significant effects cannot be interpreted as being definitely insecticide induced (Table 5).

Table 4. Small scale field trials on the effect of neem oil (0.1%, 10 l/ha) and B. bassiana oil concentrate (1.25 x 10¹³ conidia/ha) on epigeal arthropods in Niger. Total catches of 4 pitfall traps in each treatment over 15 days and frequency of taxonomic groups. Due to irregular sampling Psychodidae and Trichogrammatidae have not been included in frequency calculations

Figure 2. Total arthropod catches in untreated, neem-treated and B. bassiana-treated plots before and after spraying (arrow). For treatment details see legend in Fig. 1

Figure 3. Total catch of Notoxus sp. (Anthicidae) before (days 1-3) and after spraying (days 4-6, 7-9, 10-12, 13-15), and treatment/control ratio. For treatment details see legend in Fig. 1. Type of presentation adapted from van der Valk (1990)

Table 5. Effect of neem oil (N) and B. bassiana (B) on selected epigeal arthropods. Pooled pre-spray catches (days 1-3) are compared to post-spray catches (days 4-6 [E1], 7-9 [E2], 10-12 [E3], 13-15 [E4]) according to Everts et al. (1985). All data 10 log (n+1) transformed prior to processing. For treatment details see legend in Fig. 1

*Abbr.: A = agent; SE = standard error; * = differences significant at p < 0.05, Mann-Whitney test.

The question of sublethal effects of neem, especially possible interference with fecundity, can evidently not be answered on the basis of the available data. The same applies to the active ingredient extracted from Melia volkensii, which was even more effective than azadirachtin in semi-field trials against desert locust hoppers (Wilps et al. 1993). In a 15 day trial with M. volkensii-extract no contact action was observed in adult Coranus arenaceus. With respect to mortality, the same was found for larvae treated as last instars, although a clear retardation of development was detected, i.e. the imaginal moult was delayed with increasing dose. The ED₅₀ (50% inhibition at 15 days) of the ethanolic extract was only 0.0025 nl/individual (95% confidence limits: 0.00001-0.0530 nl). This dose clearly lies below the corresponding SAE of 0.05 nl/individual (10 ml a.i./ha). A retardation of development was also observed for S. gregaria by Wilps et al. (1993).

Following high-dose uptake of fenoxycarb-dipped food (2 g a.i./l), Adesmia cothurnata orientalis and Pimelia raffrayi raffrayi showed no response in a four day feeding test. Food uptake was also not reduced in comparison to the controls. In supplementary laboratory tests on S. gregaria, after topical application of up to 15 µg a.i. to third instar hoppers (SAE=1500 g a.i./ha!), no effect was detected except for a temporary green coloration (cf., Capinera et al. 1991).

In a prolonged feeding test (25 days) triflumuron and teflubenzuron applied at a dose of 0.4 µg a.i./individual per day had no effect on P. subquadrata chudeaui Koch (Fig. 4). Likewise, in semi-field trials with A. cothurnata orientalis and P. raffrayi raffrayi, no side-effects by IGRs were observed. Coranus arenaceus, in contrast, showed a greater sensitivity 10 days after topical application (Table 6). The SAES corresponding to the adult LD₅₀ values are, however, substantially higher than the nominal dose of 50 g a.i./ha (SAE = 1600 g a.i./ha for triflumuron and SAE = 1200 g a.i./ha for teflubenzuron). Coranus arenaceus larvae reacted, as was expected, more sensitively than adults. For spiders no clear-cut dose-response relationship could be demonstrated, although in Thanatus sp. moulting problems appeared in individual cases, and teflubenzuron at doses > 0.5 µg/individual obviously caused increased mortality (Table 7). Even a repeated oral uptake through prey insects (house flies), which had been injected with up to 150 u g triflumuron (SC-formulated) prior to being fed, had no effect on Thanatus sp. and Peucetia sp. These findings stand in contrast to toxicity tests on Capeverdian spiders in which low LD₅₀ values similar to those for locusts and grasshoppers were obtained (Table 8), while all other taxa tolerated very high doses. To clarify this divergence, additional tests on European spiders, Xysticus cristatus (Clerck) [Thomisidae] and Pisaura mirabilis (Clerck) [Pisauridae], are currently being carried out.

Figure 4. Cumulative Mortality in a prolonged feeding trial with P. subquadrata chudeaui. Figures are the mean of 5 replicates per treatment with 10 individuals each. Insecticides were sprayed on flat wafers and fed to the test organisms. The maximum daily insecticide intake was 0.4 µg a.i./beetle for IGRs, and 2 µg a.i. for polytrin-c which corresponds to a field dose of 100g a.i./ha (IGRs) and 500g a.i./ha (PC) assuming a „worst case application"

Table 6. LD₅₀ (10 days, topical) for triflumuron (50 ULV) and teflubenzuron (50 ULV) on mixed sexes of larval (fifth instar) and adult Coranus arenaceus. Values given in µg a.i./individual. The acute toxicity (48 hours) for polytrin-c on adults is given as a reference

Table 7. Mortality in Thanatus sp. ten days after topical treatment with triflumuron (50 ULV) and teflubenzuron (50 ULV). Each group consisted of 10 individuals, most of which were females (control - = untreated control, control + = solvent-treated control)

Table 8. LD₅₀ (96 hours, topical) for triflumuron (250 OF) on various Capeverdian grasshoppers and spiders (after Müller, 1992)

As in the case of botanicals, the present results do not permit conclusions about sublethal effects on non-target organisms. Further laboratory testing on locusts revealed that sublethal dosages may significantly affect fecundity. In S. gregaria females which had been treated topically with 0.625 ng triflumuron (SAE = 62.5 mg a.i./ha) as third instar hoppers, the number of eggs/pod was reduced by 30% and the number of egg pods/individual by 50% (Stolz 1993). Moreover, life expectancy was reduced by 35%.

In the semi-field trial with B. bassiana in Niger no infection in Pimelia angulata tschadensis could be detected at a dose of 1.25 x 10¹³ conidia/ha. Likewise, in the field trials either there were no effects or they were not pathogen induced (cf., Tables 4, 5 and Figures 2, 3). Only in Trachyderma hispida (Forskal), after topical application of ca. 10⁵ conidia/individual (= SAE), sporadic infections were observed, indicating that the spore formulation was indeed virulent. Screening tests in Mauritania gave diverging results. After topical application at a nominal dose of 2.5 x 10¹³ conidia/ha, infections were detected in all test species. All Coranus arenaceus died after 10 days (Fig. 5), and 90% of Peucetia sp. died after 25 days. The strongly sclerotised tenebrionids were less sensitive, their mortality not exceeding 35% in T. hispida and 20% in P. subquadrata chudeaui after 23 days (Fig. 6). The same spore formulation also proved very effective against locusts (Wilps 1993), indicating both a wide host spectrum and a remarkable pathogenicity. In Coranus arenaceus an infection via food was also detected. In these cases the invasion started from the sucking tube (proboscis). Tenebrionids, on the other hand, were not even infected after being fed moistened wafers which contained ca. 2.5 x 10⁵ conidia/cm².

Figure 5. Cumulative mortality in a sensitivity test on the effect of B. bassiana on Coranus arenaceus. Figures are pooled data for 3 replicates (groups of 7 to 8 bugs) in each treatment. For topical application (contact action), bugs were treated at a dose of = 1.25 x 10⁵ conidialindividual, for oral application (residual action), treated flies ( 1.25 x 10⁵ conidialindividual) were fed to the bugs for one day

Figure 6. Cumulative mortality in a sensitivity test on the effect of B. bassiana on Trachyderma hispida and Pimelia subquadrata chudeaui. Figures are pooled data for two replicates (groups of 12 beetles) in each treatment. For topical application (contact action), individuals were treated at a dose of = 2.5 x 10⁵ conidia/individual. No mortality was observed after ingestion of contaminated food (residual action)

go to next section of this publication