BIOLOGICAL CONTROL OF THE LARGER GRAIN BORER, PROSTEPHANUS TRUNCATUS (HORN) (COLEOPTERA: BOSTRICHIDAE) IN BENIN AND TOGO

C. BORGEMEISTER^1,2, H. SCHNEIDER^1,2,3, C. ADDA¹, H. AFFOGNON^1,3, D. AGOUNKE⁴, A. BILIWA⁴, A. TCHABI^1,5, M. CAMARA^1,6, R.H. MARKHAM¹ & D. SCHOLZ^1,2

Introduction

The larger grain borer, Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae), was accidentally introduced into East and West Africa in the late 1970s and early 1980s, respectively (Dunstan & Magazini, 1981; Harnisch & Krall, 1984). Originating from Mexico and Central America (Markham et al., 1991), the presence of the beetle has been confirmed to date in a total of 13 African countries (Hodges, 1994; Adda et al., 1996; Sumani & Ngolwe, 1996). In many of the affected countries, P. truncatus has since become the most serious pest of farm-stored maize and dried cassava. In Tanzania, dry weight losses of maize of more than 30% over a short storage season have been recorded locally (Hodges et al., 1983; Keil, 1988; Henckes, 1992), and with the introduction of P. truncatus, dry weight losses of farm-stored maize in Togo were estimated to have risen from 7 to 30% (Pantenius, 1987). After six months of storage, mean dry weight losses of 20% were recorded for farm-stored cassava chips in Togo, and most of the losses were attributed to serious P. truncatus infestations (Wright et al., 1993). Chemical control strategies based on the application of a binary insecticide, combining a synthetic pyrethroid to control P. truncatus with an organophosphate to protect against other storage pests, especially Sitophilus zeamais Motschulsky (Coleoptera: Curculionidae), were introduced with considerable success in East Africa (Golob, 1988). However, such strategies have not been widely adopted in Benin (Agbaka, 1996) possibly because the insecticides are less effective under the more humid conditions in West Africa or because the socio-economic context is less favourable for their adoption (IITA, unpublished data). In addition, the desire to avoid the health and environmental dangers inherent in the application of pesticides to food and food crops in the rural sector, as well as the difficulties of distributing pesticides effectively in rural areas, has stimulated the search for biologically based alternatives (Böye et al., 1988). As part of the effort to develop classical biological control options, a natural enemy of P. truncatus, the predator Teretriosoma nigrescens Lewis (Coleoptera: Histeridae), was identified in Central America (Haines, 1981; Böye, 1988). Laboratory (Pöschko, 1993) and field (Camara, 1996) investigations of the prey specificity of T. nigrescens revealed that the Histerid, which locates its prey using the prey’s aggregation pheromone, is, due to this prey-finding behaviour, a highly specific predator, which may attack and prey on other insects, but largely prefers P. truncatus in its diet. Subsequently, the predator has been introduced to Africa, and it has so far been released in Togo (Biliwa et al., 1992), Benin (Anonymous, 1992), Ghana (Compton & Ofosu, 1994), Kenya (Giles et al., 1995), Guinea Conakry (IITA, unpublished data), and Zambia (IITA, unpublished data). Further releases are planned in Uganda, Burundi, Rwanda, Malawi, and Tanzania. We here report about studies monitoring and evaluating the establishment, spread and impact of T. nigrescens in Benin.

Materials and Methods

Surveys in south-western Benin

Data on flight activity of T. nigrescens and P. truncatus were recorded using sticky flight traps (Pherocon II, Trécé, Salinas, CA) baited with the two component synthetic aggregation pheromone of P. truncatus (1 mg of Truncall 1 and 1 mg of Truncall 2; AgriSense-BCS, Pontypridd, U.K.), which attracts both predator and prey (Böye et al., 1992). After the first detection of T. nigrescens in a pheromone trap in Benin in August 1992, in a site in the south-western part of the country close to the Togolese border, six surveys were carried out between October 1992 and February 1996, in order to monitor the spread of the predator in Benin. A total of 83 sampling sites in the Mono province of south-western Benin was used in the study (Fig. 1). Because both the pest and the natural enemy invaded Benin from Togo, sites were selected starting from the Togo - Benin border along a west to east gradient of approximately 80 km distance, with more sites in the border area. The locations were grouped in 9 strips of approximately 9 km width, in the following referred to as transects.

Fig. 1.

T. nigrescens survey sites and pheromone trap locations in the Mono province of south-western Benin, and T. nigrescens releases sites in southern Togo.

Continuous pheromone trapping

In addition to the surveys, continuous data on flight activity of P. truncatus and T. nigrescens were recorded in eight locations of the Mono province (Fig. 1). The same trap type and pheromone lures were used as in the surveys. Traps and lures were changed at weekly intervals. Flight data gathered between April 1992 and October 1997 are presented.

Storage experiments

Data from the untreated controls of four storage trials, carried out on the same farm (near Dogbo, Mono Province) in consecutive years, are presented here. The same maize cultivar, 'TZSR-W' (Tropical Zea mays, Streak Resistant-White), an improved, tropical adapted, floury 120-d IITA variety, was used in all trials and was planted on the farm during the respective major growing season (i.e., between April and August). The experiments were carried out between September and April of 1992/1993, 1993/1994, 1994/1995, and 1995/1996. The design and the amount of maize stored in the grain stores varied between the experiments, but in all trials cobs were stored with husks intact, in stores similar to traditional local stores. In the 1992/1993 experiment, 80 cobs per replicate (with a total of three replicates per treatment) were collected on each sampling occasion, and in the 1993/1994, 1994/1995 and 1995/1996 trials, 150 cobs were sampled per replicate (with a total of three and four replicates per treatment, respectively). The moisture content of the maize was determined using standard oven-drying techniques (ISO 1980). Sampled cobs were dehusked, shelled, and sieved to collect all adult insects. The data are expressed as mean numbers of P. truncatus and T. nigrescens per kilogram of dry-matter grain. Grain loss was measured using the count and weigh method (Harris & Lindblad, 1978), with 15 replicates per sampled store.

Extensive pheromone trapping and maize sampling

Large scale pheromone trapping of P. truncatus and T. nigrescens and parallel maize sampling in farmers’ maize stores were carried out for a period of 28 months, i.e., from May 1995 until August 1997 in southern Togo and the whole of Benin. A total of 124 trapping and sampling sites was chosen across the different agro-ecological zones from the humid-coastal region of Togo and Benin to the northern Guinea savannah in northern Benin (Fig. 2). Traps in the south were changed at two-week intervals. Due to the inferior P. truncatus presence in the northern part of Benin, in these regions traps were changed every month. During the storage period, in all pheromone trapping sites samples from farmers' maize stores (10 cobs) were collected every four weeks.

Fig. 2.

P. truncatus and T. nigrescens pheromone trap locations in Benin and Togo.

The damage level of the sampled cobs was assessed, using visual classification tools, and all insects per cob were identified and counted.

Results

Surveys

The numbers of trapped T. nigrescens considerably increased in the course of the 2.5 years of observation (Figs. 3 and 4). A comparison of the three October surveys revealed that the predator had spread between 1992 and 1994 throughout the whole area of investigation (Fig. 2). Rising numbers of T. nigrescens were accompanied by decreasing numbers of P. truncatus.

Fig.3.

P. truncatus (A, C, E) and T. nigrescens (B, D, F) trap catches during three different surveys in the Mono province of south-western Benin: October 1992 (A, B), October 1993 (C, D), and October 1994 (E, F). Traps were arranged in nine parallel transects running north to south placed at 10-km intervals from the border of Togo. Traps were examined weekly. Bars show mean trap numbers per location (error bars show + 1 SE).

Fig.4.

P. truncatus (A, C) and T. nigrescens (B, D) trap catches during two surveys in the Mono province of south-western Benin: May 1993 (A, B), and May 1995 (C, D). Traps were arranged in nine parallel transects running north to south placed at 10-km intervals from the border of Togo. Traps were examined weekly. Bars show mean trap numbers per location (error bars show + 1 SE).

The trap catches of the two May surveys revealed a similar scenario, with a notable increase of T. nigrescens numbers in 1995 compared to 1993 and, at the same time, decreasing numbers of P. truncatus (Fig. 4). Again T. nigrescens was present in all transects during the second survey, showing a spread of the predator throughout the whole province.

Continuous monitoring

When the flight activity of P. truncatus and T. nigrescens in the Mono province was monitored between April 1992 and October 1997 (Fig. 5), the flight activity of P. truncatus was strongly seasonal, with two peaks in trap catches, especially in the early years of the study. The first peak occurred between the end of December and the end of February, during the major dry season in southern Benin, and the second one between April and June, during the first part of the rainy season. The lowest flight activity of P. truncatus was always recorded in October. The December to February peaks of 1994-1995 and especially of 1995-1996 and 1996-1997 were notably lower than in previous years. Moreover, the April to June peak in 1996 and the flight activity in April and May 1997 were strikingly low compared with the previous years.

Fig.5.

P. truncatus (single line) and T. nigrescens (solid line) pheromone trap catches in the Mono province of south-western Benin from April 1992 to April 1997. Traps were examined weekly. Data points represent mean of seven to nine traps.

Teretriosoma nigrescens was first recorded at these trap sites in August 1992. Thereafter, the numbers of trapped T. nigrescens increased considerably. Highest trap catches of T. nigrescens were recorded in 1994 and 1995. With the decrease in flight activity of P. truncatus in 1996 and 1997, fewer numbers of T. nigrescens were recorded in the pheromone traps. The seasonal flight cycle of T. nigrescens exhibited only a single peak, occurring four to six weeks after the April to June peak of P. truncatus (Fig. 5).

Fig.6.

Number of P. truncatus (+ ) and T. nigrescens (6 ) in three consecutive on farm storage experiments in Ex-Soprova, Mono province of south-western Benin (i.e., 1992 until 1995) (A-C), data from the untreated control.

Field trials

Prostephanus truncatus densities and accumulated grain losses during three consecutive storage experiments in the Mono province showed that, in the 1992-1993 trial, the initial P. truncatus infestation was detected in November (i.e., six weeks after loading the maize stores) (Figs. 6A and 7A). From February onwards (i.e., after five months of storage), P. truncatus numbers started to increase in a nearly exponential fashion, reaching densities of more than 2,000 P. truncatus per kilogram of maize in April and May (i.e., after eight to nine months of storage) (Fig. 6A). High densities of P. truncatus were accompanied by extremely high grain losses (Fig. 7A). No T. nigrescens were recorded throughout this field experiment. In the 1993-1994 trial, the population build-up of P. truncatus commenced earlier but reached a lower maximum density, with approximately 800 P. truncatus per kilogram of maize at the end of January (i.e., six months after loading the stores) (Fig. 6B). Thereafter, P. truncatus numbers sharply decreased, reaching a level of approximately 400 beetles per kilogram maize. T. nigrescens was observed for the first time in the course of the 1993-1994 storage trial, with high densities from January onward. The decline of the P. truncatus population occurred during a period when T. nigrescens numbers were rising (Fig. 6B). Compared with the 1992-1993 experiment, the accumulated grain losses increased faster, reaching nearly 40% in February after five to six months of storage (Fig. 7B). However, no major increase of losses was observed subsequently. P. truncatus densities were extremely low in the 1994-1995 and 1995-1996 trials (Fig. 6C and D). Peak numbers were more than 10-fold lower than in the previous year and nearly 30-fold less than those in the 1992-1993 experiment. In the 1994-1995 experiment population density estimates of T. nigrescens during the last three months of the storage trial (i.e., between March and April) were nearly as high as those of P. truncatus. Higher numbers of P. truncatus were recorded in the course of the 1995-1996 storage trial. However, in both experiments maximum grain losses of 12% were recorded in May after nine months of storage (Fig. 7C and D).

Large scale trapping and sampling

The monitoring of P. truncatus and T. nigrescens over a period of two years revealed that both species were migrating from west to east, and from south to north, with T. nigrescens moving much faster than its prey (Fig. 8-11). The west and the middle of southern Benin had been colonised by P. truncatus first (Fig. 8). However, T. nigrescens was catching up with P. truncatus. The south-eastern part of Benin had been colonised approximately at the same time by both prey and predator (Fig. 9).

Fig.7.

Dry weight grain losses in three consecutive on-farm storage experiments in Ex-Soprova, Mono province of south-western Benin (i.e., from 1992 until 1995) (A-C), data from the untreated control (error bars show + 1 SE).

Fig.8.

Results of trap catches and store samples in south-west of southern Togo and the Mono district of Benin. Trap catches are mean P. truncatus (Pt) ( ) and T. nigrescens (Tn) (· ) numbers in 32 traps placed in Togo and Benin. Traps were changed every two weeks. Store samples are mean P. truncatus numbers per cob in 32 stores in the vicinity of the trap sites; sampling interval four weeks, observation period for trap and store samples from July 1995 to January 1997.

In the northern part of Benin, P. truncatus was mainly found around market places, suggesting a dispersion of the pest through trade (Fig. 10 and 11). In these areas, T. nigrescens was yet detected only in low numbers. In the course of this study, grain losses due to P. truncatus attack were considerably decreasing in all sampling sites in the south of Benin and Togo (Fig. 8 and 9). In the northern part and in some sites of the central region of Benin, however, grain losses were increasing in areas with high P. truncatus trap catches (Fig. 10 and 11).

Fig.9.

Results of trap catches and store samples in the Atlantique and Quéme districts of Benin. Trap catches are mean P. truncatus (Pt) ( ) and T. nigrescens (Tn) (· ) numbers in 36 traps. Store samples are mean P. truncatus numbers per cob in 36 stores in the vicinity of the trap sites; observation period for trap and store samples from June 1995 to May 1997 (for more details see Fig. 8).

Discussion

The survey data clearly demonstrate the rapid establishment and spread of T. nigrescens throughout the Mono Province of south-western Benin. Within the 4 years of observation, the predator became abundant across the whole area of investigation, indicating a west-to-east spread of approximately 70-80 km. The faster spread of the predator in Benin, compared with that indicated by studies from Togo (Böye & Fischer, 1993) and Kenya (Giles et al., 1995), is probably related to the intensive trade of maize grain in coastal Benin. Maize from southern Togo is often exported, both on a local and on a regional level, to southern Benin. The recent outbreaks of P. truncatus in Burkina Faso and in Niger probably resulted from the importation of infested maize (Bosque-Pérez et al., 1991; Adda et al., 1996), and it is equally likely that the spread of T. nigrescens not only depends on the active dispersal of the predators but that the insects are also distributed in the region by maize trade.

Fig.10.

Results of trap catches and store samples in central Togo and the Zou district of Benin. Trap catches are mean P. truncatus (Pt) ( ) and T. nigrescens (Tn) (· ) numbers in 20 traps. Store samples are mean P. truncatus numbers per cob in 20 stores in the vicinity of the trap sites; observation period for trap and store samples from May 1995 to February 1997 (for more details see Fig. 8).

Results from the trap catches during the five surveys between 1992 and 1995 demonstrate a distinct increase of T. nigrescens accompanied by decreasing numbers of P. truncatus in the pheromone traps. Follow-up studies of the T. nigrescens releases in Kenya revealed that the numbers of P. truncatus in pheromone traps considerably decreased in the release areas, accompanied by rising trap catches of T. nigrescens (Giles et al., 1995).

Fig.11.

Results of trap catches and store samples in the Atacora and Borgou districts of Benin. Trap catches are mean P. truncatus (Pt) ( ) and T. nigrescens (Tn) (· ) numbers in 36 traps; traps were changed every four weeks. Store samples are mean P. truncatus numbers per cob in 36 stores; observation period for trap and store samples from May 1995 to March 1997 (for more details see Fig. 8).

Acknowledgements

The work was conducted within the framework of projects supported by the German Ministry for Economic Co-operation and Development (BMZ). The authors would like to thank the colleagues from the Service de Protection des Végétaux (SPV), Porto Novo, Benin, for their collaboration in the course of the T. nigrescens releases in northern Benin.

References