go to previous section of this publication

Test organisms were selected according to the criteria of Paulus and Klein (1993) and van der Valk (1990), whereby suitable taxa should preferably be beneficials associated with insect pests, be easy to handle, be reliably identifiable and be available in sufficient numbers at least during the period of investigation. In all experimental sites tenebrionids dominated the ground dwelling arthropod fauna. As well as being abundant, larger tenebrionids are considered important antagonists of locusts and grasshoppers, since the hypogeal larvae may feed on egg pods and thereby contribute considerably to the natural control of acridids (Ashall and Ellis 1962, Niassy et al. 1993, Popov et al. 1990, Stower et al. 1958). Hence at least one tenebrionid species was chosen for testing in each study area. Additional tests were carried out with predatory bugs (Reduviidae) and spiders (Oxyopidae and Philodromidae). A total of 8 locust control agents and 13 different formulations (Table 3) were tested on 8 different species (Table 2) in prolonged trials (10-25 days). In addition, on the Cape Verde islands, the acute toxicity of the IGR triflumuron was tested on another 18 non-target species representing a wide range of taxonomic groups: Diplopoda, Chilopoda, Scorpiones, Araneae, Dermaptera, Tenebrionidae, Carabidae and Curculionidae (Müller 1992).

Table 2. Test species, locust control agents tested and kinds of test

Abbr.: AI = A. indica (oil); MV = M. volkensii (ethanolic extract); FC = fenoxycarb; TR = triflumuron; TE = teflubenzuron; BB = B. bassiana; D = dieldrin; PC = polytrin-c (profenofos + cypermethrin); C = contact action (toxicity test); S = stomach action (toxicity test); R = residual action (semi-field or cage trial).

Table 3. Specification of locust control agents tested (Trade names in italics)

Abbr.: A = azadirachtin; AF = active fraction.

In situ bioassays were carried out following the guidelines of Hassan (1985). Depending on the control agents, mortality, virulence (fungi) or the temporal course of larval development were chosen as parameters to assess insecticidal effects. In dose-response tests ten individuals were used for each dose. In collecting test organisms from the field only individuals of equal size were taken. A standardisation for size, age and sex similar to laboratory standards could of course not be established. In most cases five different doses plus one control were tested, in case animals were not readily available, at least three different doses were tested. For topical application, the intermediate dose (out of 3 or 5 concentrations) corresponded to the common field dose in locust control. This so-called surface-area equivalent SAE is calculated from an estimate of the dorsal surface-area of the individual test organism (Weyrich 1993). For example, the SAE for Coranus arenaceus (Walker) was 0.5 µg per individual given a surface-area of ca. 0.5 cm² and a field dose of 100 g a.i./ha. Control agents were applied topically (contact action) to the thorax (insects) or cephalothorax (spiders) using a hand-operated Burkhard-microapplicator. Alternatively, test animals were sprayed in an arena (16 m²) using ULV spinning disc-sprayers (Micro-Ulva). The calculation of doses in the feeding experiments (stomach action) was also based on an estimate of SAES.

Semi-field trials were carried out to investigate the residual action of various agents on tenebrionids. After ULV-application of plots in the natural habitat, cages covering an area of 4 m² were set up and stocked with 30 (Niger) or 50 (Sudan) individuals of each species per cage. In addition to mortality the relative activity RA was recorded daily as a sublethal parameter of effect (for definition, cf. Fig. 1 and Peveling and Weyrich 1992a). In Mauritania similar trials were set up with 100 individuals/cage. Here the objective was to study the relative activity undisturbed by insecticidal stress, i.e. no insecticides were sprayed. The number of replicates was 4 (Niger, Sudan) and 5 (Mauritania), the duration of the experiments was 15 (Niger), 10 (Sudan) and 17 (Mauritania) days. All insecticides were sprayed at recommended doses for locust control.

Figure 1. Mean relative activity RA in field-caged Pimelia angulata tschadensis over a 15 days post-spray period. The RA on day i is the percentage of surviving beetles on day i caught in a pitfall trap opened daily for one hour (18.30-19.30 h). Treatments: Neem oil (0.1 %) 10 l/ha; dieldrin (18%) 0.4 l/ha. Neem induces hyperactivity whereas dieldrin induces hypoactivity. Note that the phase pattern is principally the same.

Field trials to study the side-effects of neem oil and Beauveria bassiana were carried out in Niger only. Isolated patches of Schouwia thebaica (Forsk.) Schweinf. (Cruciferae) vegetation were chosen as trial sites. The plot-size was between 500 and 2500 m². Epigeal arthropods were sampled daily (24 hour catches) with four pitfall traps per plot for a period of 15 days. Malaise traps were also put up, but did not prove worthwhile to sample flying insects reliably because of strong daytime winds. The experimental layout was oriented to the so-called „Before-After-Control-Impact Design" (Stewart-Oaten et al. 1986, Underwood 1991) with a three day pre- and 12 day post-spray period (for details, see Peveling and Weyrich 1992a). Data analysis was carried out according to Everts et al. (1985).

go to next section of this publication