BioControl 44: 301–327, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands.

Pathogen incidence and their potential as microbial control agents in IPM of maize stem borers in West Africa A.J. CHERRY1,∗, C.J. LOMER2, D. DJEGUI2 and F. SCHULTHESS2 1 Natural Resources Institute, University of Greenwich, Chatham Maritime, Kent ME4 4TB, UK; 2 International Institute of Tropical Agriculture, 08 BP 0932, Cotonou, Benin; ∗ author

for correspondence: International Institute of Tropical Agriculture, 08 BP 0932, Cotonou, Benin (e-mail: [email protected]) Received 11 August 1998; accepted in revised form 3 September 1999

Abstract. A review of the existing basis for maize stem borer IPM is given and the role of pathogens in the system is evaluated. Survey work outlining the major groups of insect pathogens is described; fungi (Beauveria bassiana and Metarhizium anisopliae), bacteria (Bacillus thuringiensis and Serratia marcesens), and viruses (granuloviruses and cytoplasmic polyhedroviruses) were identified. The presence of other unidentified protozoans, nematodes, fungi and viruses was noted. The virulence of some of the more promising known insect pathogens was explored in preliminary bioassays. Considering the cryptic habits of the insects, and the low input agriculture practiced by the majority of maize farmers in sub-Saharan Africa, Beauveria bassiana isolates possessing the capacity to grow systemically in the maize plant are considered one of the more interesting candidates for development as microbial control agents despite limited control in preliminary trials. Further work should also investigate the potential of pathogens of moderate virulence, such as the protozoans and CPVs. Key words: Bacillus thuringiensis, entomopathogenic viruses, fungal endophytes

Introduction The most notorious insect pests of maize in Africa are a number of lepidopteran stem and cob borers such as Sesamia calamistis Hampson, Busseola fusca (Fuller) (Lep.: Noctuidae), Eldana saccharina (Walker) and Mussidia nigrivenella Ragonot (Lep.: Pyralidae) which have evolved with native grasses, sedges or other wild hosts (Schulthess et al., 1997) or indigenous cereals such as sorghum and millet. The only non-native stem boring pest is Chilo partellus (Swinhoe) (Lep.: Pyralidae) which was accidentally introduced from Asia into East and Southern Africa probably 100 years ago and has not yet reached West or most of Central Africa. Yield losses due to multiple-species attack vary greatly with season, ecozone and country. In the

302 Table 1. Oviposition site and feeding site damage caused by stemborer species S. calamistis (S.c.), B. fusca (B.f.), E. saccharina (E.s.), C. partellus (C.p.), M. nigrivenella (M.n.); +/− = yes/no Stemborer species

Oviposition site Inside leaf sheath Debris on soil, old leaves Leaf surface Husks Feeding sites, damage and symptoms Whorl feeding (windows) Leaf feeding Meristem feeding (dead hearts, plant loss) Stem tunneling (reduced grain filling; colonization of stems by stem rotting fungi and lodging; early leaf senescence) Grain feeding (loss of quality and quantity of grain; early colonization of grain by storage beetles and moulds; increased aflatoxin contamination of stored grain)

S.c.

B.f.

E.s.

C.p.

M.n.

+ + − −

+ − − −

− − − +

− − + −

− − − +

− + +

+ + +

− − −

+ + +

− − −

+

+

+

+

+

+

+

+

+

+

target ecozones in West Africa (see Schulthess et al., 1997), which includes the areas south of, and including, the forest-savanna transition zone, the midaltitude and highland region, losses range between 20 and 100% (Eijnatten, 1965; Usua, 1968; Endrödy-Younga, 1968; Girling, 1980; Bosque-Pérez and Mareck, 1991; Gounou et al., 1994; Sétamou et al., 1995). Devastating outbreaks of S. calamistis and E. saccharina occur in areas with bimodal rainfall during the second cropping season, often forcing farmers to abandon second season maize as a crop. The diapausing species B. fusca can cause similar economic damage in areas with one long cropping season (Schulthess et al., 1997). The damage caused by stem borers is complex because they occupy various niches within the plant and all stages of the plant are susceptible to attack by some species (Table 1). Recent studies at the International Institute of Tropical Agriculture (IITA), Benin, also showed that cob damage in the field promotes early colonization of maize by storage beetles and moulds, leading to increased contamination of stored grain by mycotoxins (Sétamou et al., 1997, 1998).

303 Control of maize stem borers in Africa presents a unique set of biological, technical and socio-economic challenges. Firstly, in ecozones such as the forest-savanna transition zone, with rapidly increasing human population pressure, deforestation and overlapping maize plantings, pest attacks can be highly variable between seasons and fields. Secondly, all species are cryptic, spending the majority of their life cycle hidden within the maize stalk or cob, protected from most mortality factors. The borer’s cryptic habits also makes damage inapparent. Thirdly, the same plant may be attacked by several species, each preferring different plant growth stages for oviposition, and so varying in their time of attack. Such multi-species infestations, which are common in West Africa, require several treatments over a cropping cycle and may complicate use of entomopathogens which tend to be species specific. Fourthly, even low levels of attack are extremely damaging to maize because of its inability to tiller. Finally, in most African countries, the price of maize is low, and the margin available for pest control expenditure is very small. Control of stem borers with chemical pesticides is neither affordable nor sustainable for the mostly resource-poor African maize farmers. Policy reforms with the removal of subsidies, the devaluation of local currencies, and the increasing cash constraint and poverty of small scale farmers, are all serious impediments to the purchase and use of pest management inputs. Neither effective culture practices, nor traditional control techniques are known. The International Agricultural Research Centers have traditionally focused on host plant resistance as the first-choice control option. IITA has developed lines with moderate levels of resistance to S. calamistis and E. saccharina, good levels of resistance to the major diseases, and desirable agronomic characteristics. The International Maize and Wheat Improvement Center (CIMMYT) in Mexico, and the International Center for Insect Physiology and Ecology (ICIPE) in Kenya, have developed genotypes with strong antibiosis to whorl feeding species such as B. fusca, Diatrea spp. and Chilo spp. (CIMMYT, 1989; ICIPE, 1991). Transgenic corn, containing the Bacillus thuringiensis (Berliner) toxin gene (Koziel et al., 1993), is currently proving a success with US farmers, providing good protection against the European corn borer, Ostrinia nubilalis (Hübner) (Lep.: Crambidae). However, it is not yet clear which toxins would be needed to provide protection against the full range of African stem borer species. The complex social, economic and environmental issues associated with the introduction of genetically engineered maize are currently being investigated by CIMMYT. Thus, at present African farmers do not have satisfactory control options available to them.

304 In the late 1980s an analysis of the stem borer ecosystem was initiated by IITA to look for other control options. The system analysis revealed opportunities for forms of biological control such as the exchange of species and strains of natural enemies between regions, and the use of non-co-evolved natural enemies, as well as habitat management solutions, namely the use of trap plants such as wild grasses on which larval mortality can be very high (Shanower et al., 1991; Sekloka, 1996; Schulthess et al., 1997; Sémeglo, 1997) and management of soil nutrients (Sétamou et al., 1993, 1995; Denké, 1995). Similar research is being carried by ICIPE (Khan et al., 1997). ICIPE has released and established Cotesia flavipes Cameron against C. partellus at the Kenyan coast (Mbapila and Overholt, 1997) from where it spread into Tanzania. IITA introduced an East African strain of Cotesia sesamiae (Cameron), provided by ICIPE, into Benin where it became established in some parts of the southern departments (Schulthess et al., 1997). One biotic factor not included in the system analysis was the role of entomopathogenic micro-organisms. The pest status of individual borer species varies between ecozones, countries and regions (Schulthess et al., 1997). Results from light trap catches in southern Benin showed that populations of S. calamistis crash long before the onset of the dry season when suitable hosts are still available (Kouame, 1995). In the same area, the most damaging borer species Africa-wide, B. fusca, is exceedingly rare. Some of the variation in the population dynamics of borers between regions was attributed to parasitoids (Kfir and Bell, 1993; Kfir, 1995; Schulthess et al., 1997), abundance and species composition of alternative hosts acting as trap plants or harboring natural enemies (Cardwell et al., 1997; Schulthess et al., 1997) and soil nutrients (Sétamou et al., 1995; Sétamou and Schulthess; 1995). These factors do not provide a complete explanation, and it is speculated that entomopathogenic micro-organisms could also play a role. In this paper we report the results of stem borer pathogen surveys, and of early field and laboratory assessments of pathogens from each of the major groups. The work contributes to the system analysis of maize stem borers by providing an indication of the role of pathogens in stem borer population dynamics. We discuss the potential contribution of the various pathogens to control strategies by linking what is known of pest and pathogen ecologies and suggest some future research and development activities necessary to assess the feasibility of these control options for resource-poor farmers in Africa.

305 Materials and methods Collection of samples in surveys During 1996, samples of living and dead stem and cob borer larvae and pupae were collected from maize and, to a lesser extent, from millet and sorghum in farmers’ fields in West Africa. During the maize growing period from April to December 1996 we conducted eight field surveys in Benin, four in Nigeria and two in Ghana. Each survey focused on different country regions. Maize fields were selected on the basis of plant growth stage and farmer consent at intervals of approximately 20 km along major roads. Within fields, plants were inspected at random for symptoms of stem borer damage (see Table 1) on stems and cobs. The cob borer M. nigrivenella was additionally sampled from the seed pods of its alternative host plants Parkia biglobosa, Mimosaceae, and Gardenia sokotensis, Rubiaceae. Samples of dead insects were also received from collaborators in Côte d’Ivoire, Togo, Cameroon and South Africa. Further samples were collected by soil baiting (Djaman, 1997) with stem borer larvae and from quantitative field monitoring experiments at IITA. One kg soil samples were collected from sites around Benin and returned to IITA. Up to 40 3rd and 4th instar S. calamistis or E. saccharina were placed with the soil in plastic basins and held at ambient temperature. After 1 week, live, dead and mycosed larvae were recovered and treated according to the following procedures. Preliminary screening procedures All dead larvae whether from field surveys, soil exposure or from quantitative field monitoring experiments were screened in a series of simple laboratory procedures to detect the presence of potential pathogens. For those samples where fungal sporulation was visible on cadavers, or where the cadaver was stiff and subsequent incubation under humid condition led to sporulation, spores were examined in lacto-phenol blue mounts, and cultured by plating onto Sabouraud dextrose agar (Merck Ltd., Darmstadt, Germany) (SDA) containing the antibiotic chloramphenicol at 0.05 g/l. After 4–5 days at 28 ◦ C, single colonies were re-isolated onto ordinary SDA or potato dextrose agar (PDA) (Merck Ltd.) slants, incubated until sporulation, and then refrigerated. For all other samples, larval smears were prepared on glass microscope slides and stained with Gurr’s Giemsa (Merck Ltd.) following acid hydrolysis (Poinar and Thomas, 1984), viewed under oil immersion objective at 1000×. All apparent pathogens were tentatively identified. Also for all samples, with the exception of mycosed larvae, where host larvae were identified and the homologous host larvae were available from IITA laboratory colonies

306 (S. calamistis and E. saccharina), 1 ml suspensions of triturated dead larvae were prepared in sterile distilled water and applied as 1 µl droplets to small discs (15–20 mm3 ) of artificial diet for a single dose bioassay of >20 third instar larvae per sample. Larvae were starved for 24 hours prior to dosing. Those larvae consuming the entire dose within 24 hours were transferred to clean artificial diet in glass tubes, and observed until death or adult emergence. Dead larvae were examined by stained smears following the procedure outlined above. Where host species were not available for bioassay, as in the case of B. fusca, samples were screened specifically for the presence of B. thuringiensis. Suspensions of dead larvae prepared as above, were treated with 50% ethanol for one hour to kill vegetative bacterial cells but leave resistant spores (Sneath, 1986). Suspensions were pipetted onto a standard peptone based nutrient agar (Merck Ltd.) and incubated for 48 hours at 30 ◦ C. Colonies conforming to B. thuringiensis/B. cereus morphology (Sneath, 1986) were re-isolated onto fresh nutrient agar, incubated for 48 hours at 30 ◦ C, and then identified as B. thuringiensis by vegetative cell morphology and the presence or absence of parasporal crystalline bodies (Thiery and Frachon, 1997) in Giemsa-stained smears viewed at 1000× under an oil immersion objective. A short list of candidate samples was compiled of those samples meeting one or more of the following criteria: (1) visible sporulation on cadaver; (2) pathogens visible in stained smears; (3) >50% mortality in larval bioassay; (4) isolation of B. thuringiensis. Identification of pathogens Samples containing suspected virus isolates, following examination of larval smears as described above, were examined by scanning electron microscope (SEM) at IITA Nigeria. The same samples were sent to the Natural Resources Institute (Chatham, UK) for identification by DNA/RNA RFLP profiles. Fungal isolates were identified to genus and species group on the basis of spore and colony morphology (Humber, 1997). B. thuringiensis was identified following the procedure outlined above. All other pathogens were identified using standard laboratory keys (Poinar and Thomas, 1984). Production and bioassay of isolated pathogens Beauveria bassiana production Selected strains of Beauveria bassiana (Balsamo) Vuillemin, isolated as above, were tested in the greenhouse for their potential to establish endophytically and provide systemic control of stemborers in maize (Bing and Lewis, 1991). Spores of locally isolated B. bassiana were grown on PDA

307 in Petri dishes at room temperature (24–28 ◦ C) and dried to <5% moisture content over silica gel. Spore viability was assessed using a standard 24 hour germination test on SDA (Hedgecock et al., 1995) and only batches with >85% viability were used. B. bassiana experiment I To confirm establishment of isolates in plants, maize seed (IITA var. TZESR-W) was sown in sterilized soil in black plastic bags in an airconditioned green-house at IITA station Benin during January 1997 and were thinned to one per bag after emergence. For each of five B. bassiana isolates (nos. 82, 184, 185, 478 and 479) 20 plants were inoculated by injection of 10 µl of aqueous spore suspensions (in 0.05% Tween 80) containing 1 × 109 spores ml−1 into the pseudo-stem (composed of early leaf sheaths) behind the sheath of leaf 3 at the V6 growth stage (Kling and Edmeades, 1997), 14 days after sowing. A further 20 control plants received an injection of Tween 80 suspension only. Systemic establishment of isolates within maize plants was checked 14 days post-inoculation. Stem pieces (10–20 cm) were washed in 10% bleach, rinsed in sterile distilled water, washed in 90% ethanol, and finally rinsed again in distilled water. For each of 8 plants per isolate a single 1–2 cm stem section was cut from the sterilised pieces. Sections were placed, cut face down, on SDA in Petri dishes at 28–30 ◦ C for 7 days to observe mycelial growth. Growth of B. bassiana from any of the sections was taken as indicative of systemic establishment. B. bassiana experiment 2 In a second experiment, maize plants were reared in the greenhouse as above. For each of 5 isolates (nos. 82, 184, 478, 479 and 480) 20 plants were inoculated into the space between the stem and sheath of leaf 3, at the V6 stage, by topical application with a suspension of 1 × 107 conidia in 100 µl 0.05% aqueous Tween 80. An additional 40 control plants received aqueous Tween 80 only. Inoculated plants were transplanted to a small uniform experimental field site at IITA-Benin 7 days post-inoculation. Isolates were randomized in a single block of 6 plots. Plants were irrigated as necessary from overhead sprinklers. Five 2nd instar S. calamistis were placed in the whorl of each plant 14 days after transfer to the field. After an additional 14 days all plants with dead hearts (plants killed by borer attack) were destructively sampled. For each isolate, the number of dead hearts was counted as well as the number of tunnels and mean tunnel length. Live insects were removed for incubation on sterile artificial diet, and dead insects were incubated in moist chambers to facilitate sporulation. Plants without dead hearts were left in the field until harvest when cobs were removed and weighed individually.

308 B. bassiana experiment 3 In a third experiment, maize was sown in the greenhouse as above and transplanted to the field at the V6 stage. Inoculation of 20 plants for each of 7 isolates (nos. 82, 184, 185, 313, 478, 497 and 480) took place in the field 7 days later by topical application using the method described in experiment 2. Plants were infested with S. calamistis larvae and sampled as in experiment 2. Bacillus thuringiensis production Locally isolated samples of B. thuringiensis, along with the B.t. standard HD1 (= subsp. kurstaki) (supplied by the University of Greenwich, potency not given) were cultured and assayed using the following procedures. Seed cultures of each isolate were initially grown in 125 ml Erlenmeyer flasks containing 50 ml nutrient broth (Merck Ltd.) inoculated from single colony subcultures on nutrient agar, and incubated at 28–30 ◦ C on a rotary shaker at 150 rpm for 24 hours. Three ml of seed cultures were used to inoculate 150 ml BG medium (Johnson and Bishop, 1996) in 250 ml flasks. Flasks were incubated at 28–30 ◦ C and 150 rpm for 96 hours at which point the majority of cells had lysed to liberate crystals and spores. Spores and crystals were pelletted by centrifugation at 4500 rpm (2500 G), washed in sterile distilled water, re-centrifuged and finally resuspended in a sterile 15% (w/w) glycerol solution before storage at −15 ◦ C until required for use. The concentration of colony forming units (c.f.u.) per ml was determined from 10-fold serial-dilutions plated onto nutrient agar. The c.f.u. concentration of a commercial sample of DiPel DF (B.t. subsp. kurstaki) (Abbott Laboratories, North Chicago, USA) was calculated similarly. B. thuringiensis bioassays Initial bioassays were carried out as follows to determine median lethal concentrations (LC50) of the standard HD1 strain for 3rd instar S. calamistis, E. sacharina, M. nigrivenella and B. fusca. B.t. spore/crystal suspensions of a range of concentrations from 5 × 104 to 1 × 108 c.f.u. ml−1 were applied as 1 µl droplets via micropipette to discs of artificial diet (15–20 mm3 ) which were placed in small glass tubes along with a single larva, and plugged with cotton wool. Thirty larvae were treated per dose with a minimum of 3 replicates over time to account for natural variation in response (Robertson et al., 1995). Control insects were dosed with sterile distilled water. After 24 hours larvae which had consumed >50% of treated diet discs were supplied with a 0.5 cm3 plug of clean diet and held until 6 days post incoulation. LC50s were computed from day 6 mortality data using the probit analysis routine of the MLP computer program (Lawes Agricultural Trust, Rothamsted Experimental Station, UK) and a mean LC50 calculated for each insect

309 species from 3 replicates for which dose-response slopes satisfied the criteria of parallelism. The calculated HD1 LC50 was thereafter used as the standard dose for all of the test isolates against each of the host species. The same assay procedure as described above was used, and mortality rates compared to those obtained for the HD1 strain. Virus production Of the 16 isolates of local virus listed (Table 2), 8 selected from S. calamistis, one from E. saccharina and one from B. fusca were propagated in their respective host as follows: suspensions of each virus isolate were prepared by triturating the original host larva in sterile distilled water to a volume of 1 ml and then filtering. Concentrations were defined as 1 larval equivalent (l.e.) ml−1 . Suspensions were fed to 10 3rd instar larvae per isolate by applying 1 µl droplets via micropipette to small discs of artificial diet (15–20 mm3 ). Larvae were held in small glass tubes plugged with cotton wool. After 24 hours all larvae which had consumed the treated diet were supplied with a 0.5 cm3 plug of clean diet and held until death or pupation. Preliminary assays with propagated virus Subsequent bioassays with propagated virus isolates followed an identical procedure except for isolate 2 which was prepared at 0.2 l.e. ml−1 . Twenty larvae were dosed per isolate, mortality was recorded daily until death or pupation, and all assays were replicated 3 times. Survival data was analysed by Kaplan Meier survival analysis (SPSS, 1995) to determine average survival times (AST).

Results Surveys From the 1996 surveys, 577 dead larvae and pupae from field collected samples were passed through the primary screening procedure outlined above, of which 66 causal agents or pathogens were identified on the basis of indicated criteria. These are listed in Table 2 with tentative identifications. The distribution of pathogens among the host species was diverse. All three principal groups of entomopathogens, bacteria, viruses and fungi, were found in S. calamistis, E. saccharina and B. fusca. Of the techniques employed for pathogen detection, bioassays were the most productive although time consuming. Stained slides revealed little that would not have been detected by simple wet mounts, therefore this method will be used for future screening. It is important to note that even a thorough survey such as the one described

310 Table 2. Shortlist of selected samples based on preliminary screening, their origin and host Sample accession number

Isolated pathogens

Host species

Country and locality

2 5 33 35 60 61 64 78 82 83 84 85 86 87 88 91 93 96 99 113 116 134 135 156 178 179 180 181 182 183 184 185 190 194 206 204

B.t. & GV B.t. CPV in S.c. assay

B.f. B.f. B.f. B.t. B.f. B.f. B.f. M.n. E.s. M.n. M.n. M.n. M.n. M.n. S.c. E.s. E.s. S.c. S.c. E.s. M.n. S.c. S.c. B.f. E.s. S.c. S.c. E.s. E.s. S.c. E.s. A.i. E.s. B.f. S.c. E.s.

Ni Ni Ni Ni Be Be Be Be Be Be Be Be Be Be Be Be Be Be Be Be Be Be Be Gh Be Be Be Be Be Be Be Be Be Ca Be Be

B.t. B.t. B.t. B.t. Beauveria sp Beauveria sp Beauveria sp Beauveria sp Beauveria sp Beauveria sp Virus B.t. B.t. Bacilliforms B.t. S. marcescens S. marcescens CPV CPV B.t. Metarhizium sp Metarhizium sp Metarhizium sp Metarhizium sp Metarhizium sp Metarhizium sp Beauveria sp B. bassiana Metarhizium sp Unidentified fungus Mermithid nematode Mermithid nematode

Ngag Nde Aliade Ife Betekoukou Tori bossito Attogon Avlame IITA Agbog. Avlame Agbog. Kotokpa Avogbana IITA IITA IITA IITA IITA IITA Sahoue Avogbana Modani Ab. Bassila Bassila Bassila Bassila Bassila Bassila IITA Malanville Djougou Etoud Foret de Pene. Foret de Pene.

% Mortality in conspecific bioassay

60.00 50.00

100.00

35.00 50.00 100.00 50.00 50.00 20.00 10.00 90.00 60.00

311 Table 2. Continued Sample accession number

Isolated pathogens

Host species

Country and locality

220 242 253 255 263 279 285 286 292 295 296 313 321 323 386 393 429 478 479 480 481 482 492 493 495 497 540 543 544 569 574 575 860

B.t. CPV Mermithid nematode Mermithid nematode B.t. Unidentified fungus Mermithid nematode B.t. Unidentified fungus Unidentified fungus Unidentified fungus Beauveria CPV CPV CPV CPV B.t. Beauveria Beauveria Beauveria M. anisopliae M. anisopliae B.t. B.t. B.t. B.t. CPV CPV CPV Unidentified virus CPV CPV Hirsutella

B.f. S.c. S.c. S.c. B.f. A.i. S.c. B.f. B.f. A.i. A.i. S.c. S.c. S.c. S.c. S.c. B.f. A.f. A.f. A.f. E.s. E.s. E.s. B.f. E.s. B.f. S.c. S.c. S.c. E.s. S.c. S.c. B.f.

Be Ni Ni Ni Ni Be Ni Ni Ni Be Be Be Be Be Be Be Ni Ni Ni Be Be Be SA Be Be Be Be Be Be Be Be Be Ca

Barerou Minna Amakama Amakama Ikom-G Massin Amawbia Lasmoto Mbakor Bussokani Malanville Soclogbo Hlassame Hlassame Agbodjedo Davougon Lokoja Ikom Ikom Tchaouron Tchaouron Kwa-Zulu Natal Podji-Lemou Issaba Ahwium Zinvie Hopital Zinvie Agassa-Godo IITA Itadjebou Ketou Etoundi

% Mortality in conspecific bioassay

60.00 5.26

35.00 65.00 70.00 65.00 75.00

100.00 90.00 60.00 60.00 60.00 15.00 60.00 65.00

Busseola fusca (B.f.), Sesamia calamistis (S.c.), Eldana saccharina (E.s.), Mussidia nigrivenella (M.n.), Coniesta ignefusalis (A.i.), Auricula forficularis (A.f.), Bacillus thuringiensis (B.t.), nucleopolyhedrovirus (NPV), cytoplasmic polyhedrosis virus (CPV), granulovirus (GV), Nigeria (Ni.), Benin (Be.), Ghana (Gh), South Africa (S.a.), Cameroon (Ca).

312 Table 3. Results of experiment 2: effect of systemic B. bassiana establishment in maize on S. calamistis tunneling and mortality, and on maize cob weight Isolate

No. dead heart per 20 plants

Mean tunnel length per plant (cm + s.e.)

No. dead larvae per 20 plants

Mean cob weight in surviving plants (g + s.e.)

82 184 478 479 480 Pooled isolates Control

7 5 6 0 n.a.

2.18 4.60 4.62 0.00 1.55 3.17 10.50

6 5 6 0 n.a.

168.5 146.9 154.3 146.1 165.8

(6.14) (7.93) a (5.85) (6.47) a (8.68)

6

176.5

(6.14) b

18

(0.19) (0.79) (0.48) (0.22) (0.37)∗ (1.01)∗

n.a.: not assessed. ∗ Significantly different by one-tailed t test (critical value = 1.94, p < 0.001). a & b: means followed by different letters are significantly different at p = 0.05.

will only detect known pathogens; cryptic and unknown pathogens inevitably go undetected through the screening process. From a consideration of the geographic origin of selected sampled it is apparent that pathogen biodiversity in Benin is not uniform. In particular, the IITA site on the southern coastal strip, was a rich source of both B. thuringiensis, despite no previous history of use, and of the S. calamistis virus. Two other zones in southern Benin which provided rich biodiversity were targeted for further studies in 1997 and 1998. Pathogen bioassays Experiments with B. bassiana The first experiment to assess establishment, in which fungus was injected into plant stems, gave the following rates of establishment; isolate no. 82: 4/8; isolate no. 184: 7/8; isolate no. 185: 5/6; isolate no. 478: 4/7; isolate no. 479: 5/6; control 0/6. In experiment 2, a large number of plants were destroyed in the field, either by the artificial infestation with S. calamistis or by natural stem borer attack; as a consequence few statistical analyses were possible. However, there are some interesting indications, shown in Table 3. Firstly, there appears to be a reduction in the length of larval tunnels in the plants treated with B. bassiana, and there were no tunnels in plants treated with isolate 479. The reduction in length was significant when data for all fungal isolates were pooled and compared with the control by one tailed t-test (critical value

313 Table 4. Reduction in tunnel number and lengths due to S. calamistis in maize plants with systemic infection by B. bassiana in experiment 3 Isolate

Mean tunnel length per plant, (cm ± s.e.)a

No. dead hearts per 100 plants

82 184 185 313 478 479 480 Pooled Control

0.88 1.60 2.90 1.17 0.67 0.50 0.75 1.40b 2.88b

(0.13) (0.70) (0.90) (0.44) (0.17) (0.00) (0.25) (0.28) (0.97)

4 3 5 3 3 2 2 3.14c 8

a In this experiment, number of tunnels = number of dead hearts. b Significantly different by one-tailed t test (critical value, 1.86, p =

0.043). c Mean of observations on treated plants.

= 1.94, p < 0.001). Secondly, there was a non-statistical trend indicating that dead hearts were less frequent in treated plants than in the control. In particular, none of the plants treated with isolate 479 were lost, while 18 out of 20 control plants were killed. Interestingly, as many dead larvae were found in the control as in the treated plants. Since the cause of death was not established, no conclusion can be drawn from this observation, but it might be speculated that death was not directly related to treatment. Finally, analysis of cob weight data showed a significant difference between means (F = 3.58, p = 0.008). Means comparison (Tukeys HSD) showed that at p = 0.05, the mean weights of cobs from plants treated with isolates 184 and 479 were significantly less than the mean cob weight for control plants. The third experiment was conducted during the dry season; plant growth was poor, there were no natural borer attacks, and it was difficult to establish artificial infestations. There was some apparent reduction in the number of dead-hearts due to treatment, but this could not be analyzed statistically (Table 4). There was a reduction in the number and length of tunnels, and this reduction in length was significant when all treatments were pooled and compared with the control by one-tailed t test (critical value, 1.86, p = 0.043). Dead larvae were recovered from the plants in experiments 2 and 3, but neither mycosis nor sporulation was observed during incubation.

314 Table 5. Mean percentage mortality (± s.e. mean) of 3rd instar Eldana saccharina and Sesamia calamistis 5 days after inoculation with a 1 µl dose of 3.5 × 106 c.f.u. ml−1 of local and standard B.t. isolates

Isolate

Mean % mortality ± s.e. mean Eldana saccharina Sesamia calamistis

2 5 61 64 78 91 93 99 156 220 263 286 429 492 493 495 497 HD1 DiPel

77.78 10.00 11.11 12.22 18.89 4.44 5.55 10.00 10.00 4.44 8.89 6.67 6.67 5.55 88.89 41.11 31.11 88.89 95.56

2.94 5.09 8.01 4.84 13.92 1.11 2.22 1.92 8.39 4.44 5.88 3.33 5.09 4.01 2.94 9.87 9.10 5.88 2.94

34.45 1.11 0.00 0.00 2.22 4.44 4.44 7.78 3.33 4.45 1.11 5.55 1.11 1.11 42.22 3.33 5.56 86.67 92.22

6.19 1.11 0.00 0.00 2.22 4.44 1.11 6.19 3.33 2.22 1.11 4.01 1.11 1.11 2.22 3.33 2.94 1.93 2.94

B. thuringiensis bioassays LC50s (c.f.u. × 106 + fiducial limits) for B.t. HD1 in 3rd instar S. calamistis, E. sacharina, M. nigrivenella and B. fusca were 3.18 (2.43–4.35), 1.68 (1.33–2.15), 4.48 (3.31–6.51) and 72.0 (54.2–105), respectively. Given the similarity in mean LC50 values for S. calamistis, E. sacharina, and M. nigrivenella, local isolates to be tested in these species were all prepared at 3.5 × 106 c.f.u. ml−1 . The LC50 for B. fusca was an order of magnitude higher than for the other species. A shortage of larvae from IITA colonies precluded any B.t. assays in B. fusca or M. nigrivenella in the present study. Of the 17 local B.t. isolates tested in S. calamistis and E. sacharina, only 2 (isolates 2 and 493) caused mortality in host larvae which approached (and none which exceeded) that caused by either HD1 or DiPel (Table 5). Given these results, none of the local B.t. isolates were selected for further study.

315 Table 6. Average survival times (AST) in days for larvae treated with S. calamistis and B. fusca viruses Isolate

Host

Virus

2 88 134 242 321 386 393 544 569 574 575

B.f. S. c. S. c. S. c. S. c. S. c. S. c. S. c. E.s. S. c. S. c.

GV CPV CPV CPV CPV CPV CPV CPV Unknown CPV CPV

AST ± s.e. (days) 9.14 35.46 32.15 26.90 37.98 33.12 36.45 33.41 — 29.33 28.02

1.01 1.16 1.13 1.78 1.05 0.85 0.85 0.71 — 1.47 1.06

Virus identity and bioassays Examination of progeny virus derived from sample 2 from B. fusca by SEM showed high density of particles with granulovirus (GV) morphology, and results of REN analysis of viral DNA confirmed the presence of a single GV isolate. The AST calculated from pooled bioassay data from 3 replicates was 9.14 ± 1.01 (s.e.) days (Table 6). Cumulative mortality is shown in Figure 1. Examination of S. calamistis virus samples by SEM revealed the presence of three virus morphologies, one distinctly cuboid with sides measuring 2.47 ± 0.11 µm (s.e.), a second, polyhedral in shape with mean diameter 1.33 ± 0.05 µm and a third, also polyhedral in shape, but with mean diameter 0.730 ± 0.04 µm. Electrophoresis of undigested RNA revealed a fragmented genome of 10 or more pieces, typical of cytoplasmic polyhedroviruses (CPV) (Tanada and Kaya, 1993), and the presence of multiple isolates was indicated by the presence of sub-molar bands in gel profiles. Significant quantities of uncut material other than from CPV were also present in the 50 Kb range, but restriction enzyme digests of identical samples could not detect the presence of an NPV despite observing polyhedra in distinct nuclear groups in wet mounts of larval hemolymph by phase contract microscopy. Larvae treated with CPV ceased growing shortly after ingesting the inoculum but there was a lag of 2 to 3 weeks before the first deaths (Figure 1). Final mortality varied between 63 and 93% with an AST range from 26.9 to 37.9 days (Table 6) but since virus concentration was not confirmed, little can be concluded at this stage. The E. saccharina isolate did not cause adequate mortality for

316

Figure 1. Cumulative mortality curves of B. fusca larvae treated with GV isolate ( = 2) and of S. calamistis larvae treated with CPV isolates, ● = 88, ❍ = 134, ■ = 242, 2 = 321, ▲ = 323, + = 386, ◆ = 393, 3 = 544.

Kaplan-Meier survival analysis and was too dilute or either EM studies or REN analysis.

Discussion and conclusions Entomopathogens have been successfully tested against stemboring species elsewhere in the world, and some have been developed into commercial products; Ostrinil, a granular formulation of B. bassiana from NPP, Pau, France, and DiPel G, a granular B.t. formulation from Abbott, Chicago, USA, are examples. B. thuringiensis has been used as a biopesticide for 30 years (Bohorova et al., 1997). B. thuringiensis has been tested against many lepidopteran pests of maize (Broza et al., 1991; Bohorova et al., 1997) including stem borers (McGuire, 1990; Medvecky and Zalom, 1992). Differences in the levels of activity to C. partellus larvae demonstrate the importance of screening a number of B.t. strains when considering their use in a microbial control program (Brownbridge and Onyango, 1992b). In the current study B.t. was found regularly in dead B. fusca larvae. However, none of the local B.t. isolates were as active against S. calamistis or E. saccharina as either commercially available DiPel or the standard HD1 strain on which this product is based. Nevertheless, our experiments show

317 that susceptibility of African stem-borer species to B.t. toxins is highly variable, and the frequency with which new isolates can be found indicates the potential for discovering more active strains. Whilst other insect pathogenic bacteria were isolated from stem borer larvae, none showed the potential for development of B.t. Beauveria bassiana has been used experimentally to suppress populations of O. nubilalis on maize for many years (Bartlett and Lefebvre, 1934; Riba, 1984; Bing and Lewis, 1991, and this has led to the production of a commercial product, Ostrinial, by NPP for control of O. nubilalis. Similarly, Maniania (1993) reports a reduction in damage due to C. partellus on maize in Kenya following application of B. bassiana isolate ICIPE 35. B. bassiana is not the only fungal species to infect stem borers. Chiou and Hou (1993) demonstrated the susceptibility of O. furnicalis Guenee to Metarhizium anisopliae in Taiwan. In Kenya locally occurring isolates of B. bassiana and M. anisopliae have been evaluated in topical application assays against C. partellus and B. fusca. Results varied between isolates and species. The most active strains were of the species M. anisopliae although both species may have potential as microbial control agents (Maniania, 1992). In our surveys, isolates of M. anisopliae, B. bassiana and Hirsutella sp. were found infecting S. calamistis, E. saccharina and C. ignefusalis. Preliminary bioassays with M. anisopliae and B. bassiana indicated a high degree of virulence towards S. calamistis, E. saccharina and M. nigrivenella following topical application of spore suspensions (Djaman, 1997). The recent finding that B. bassiana is able to grow endophytically within maize plants and reduce borer damage added a new dimension to the use of fungal entomopathogens against stem borers (Lewis and Cossentine, 1986; Lewis et al., 1996). Bing and Lewis (1992) showed that one particular B. bassiana isolate, ARSEF 3113, in some maize varieties could invade the plant via the cuticle following foliar application and persist season-long in the plant and reduce tunneling by O. nubilalis. The penetration process was examined in more detail by Wagner (1997), and this is generally perceived to be the most problematic part of the invasion and protection process. In our current work, isolates collected on insect were tested for their ability to invade maize plants. Preliminary assays were encouraging in demonstrating the existence of systemic action amongst a range of B. bassiana isolates. All isolates tested were able to grow inside the plant and provided some indications of reduction in damage and attack levels. In the case of isolate 479, where there was a heavy natural attack by indigenous stem borers, damage reduction was dramatic. However, the suggestion of a slight cob weight reduction due to some isolates, including 479, will dictate the need for further careful evaluation.

318 Among entomopathogenic protozoa, the microsporidia, and particularly members of the genera Nosema and Vairimorpha, have received considerable attention as insect control agents. However, other than a notable success in grasshopper control using Nosema locustae Canning in conjunction with an insecticide (Henry, 1981) there have been few if any instances of the successful use of microsporidia in agricultural pest control (Canning, 1993). Death is usually caused by sheer numbers of parasites, not by the presence of toxins and mortality does not generally occur quickly (Canning, 1993). Many instances of Nosema species’ infections have been documented in cereal stemborers (Lewis and Lynch, 1978; Andreadis, 1987; Bordat et al., 1993) and some have been tested in field trials (Odindo et al., 1992). However, there are concerns over possible deleterious effects of these parasites on parasitoid larvae within the infected host (Walters and Kfir, 1993; Bordat et al., 1994; Orr et al., 1994). In the current studies, although several samples containing protozoan spores were examined, resources were insufficient to culture the pathogens throughout their life cycle, thus identification beyond family level was not possible nor were assays conducted. Insect pathogenic viruses are known to occur in cereal stemborers (Vago, 1956; Raccah et al., 1975; Morel et al., 1981; Lewis and Johnson, 1982; Godse and Nayak, 1983; Jacquemard et al., 1985; Odindo et al., 1989; David and Easwaramoorthy, 1990; Neupane, 1990; Fediere et al., 1993; Jones, 1993; Fediere et al., 1994; Drolet et al., 1998) but despite the advantages offered by these pathogens, few field trials have been undertaken to explore their potential. Viruses have the advantage of being suitable for low-technology local production and can also cause disease epizootics resulting in long-term control. Local production of viruses has been shown to be cost-effective in many countries including Brazil, China, Egypt and Thailand (Jones, 1988; Jones et al., 1993). Our own surveys in West Africa have revealed the presence of a S. calamistis CPV, which may be similar to that occurring in Reunion (Jacquemard et al., 1985). However, CPVs typically cause chronic disease (Tanada and Kaya, 1993) whilst laboratory assays showed significant, although delayed, mortality. Unpublished field data (Cherry) indicated that this virus accounted for up to 25% mortality in S. calamistis larvae in an IITA-Benin maize plot in 1997. A granulosis virus in B. fusca, and a possible NPV in E. saccharina have also been found. Control strategies with microorganisms The full range of IPM control options including host plant resistance, cultural control, chemical and biological control were discussed in the introduction; here we focus on insect pathogens as microbial control agents. Microbial

319 control agents might be deployed as classical biological control agents, in inundative or inoculative release strategies, or in novel ways. Classical biological control The establishment of an entomopathogen as a classical biological control agent in a population for long-term control or suppression of a seasonal pest requires effective horizontal transmission within the population, transmission to and from an intra-seasonal reservoir, such as the soil or wild host plants, as well as good persistence in that reservoir. Alternatively, vertical transmission between generations would allow for long term persistence in a population but may not lead to epizootics. Hochberg (1989) identifies two distinct survival strategies for pathogens: they may be highly virulent and persist in reservoirs, or be less virulent and persist largely within the host population. The short cropping cycle of maize does not provide ideal conditions for establishment of persistent entomopathogens, particularly those whose principal means of transmission is horizontal, e.g. NPV, or those relying on air-borne transmission, e.g. entomophthoralean fungi. Limited mobility of stemborer larvae in cultivated cereals restricts the opportunity for horizontal transmission within a population. Horizontal transmission is further restricted within maize plants since multiple tunnels may be isolated one from another, and within individual tunnels larvae occur singly or in small numbers only. Infected larvae tend to remain and die within the maize stem. After harvest, stems are either left in the field, or used as fodder, fuel or building materials. These uses limit the flow of inoculum to the soil. On the positive side, since maize is sown in the rainy season there is ample chance for inoculum to be splashed up onto emerging plants. Cloud cover during this period is increased and UV levels correspondingly reduced. Opportunities for classical biological control of stem borers might therefore be greater among those pathogens exhibiting vertical transmission. Both CPVs and GVs are known to be capable of vertical transmission transovum and it is not surprising therefore that it is these viruses which have been found infecting stem borers. Reports that the granulosis viruses of some stem borers exhibit vertical transmission (Melamed-Madjar and Raccah, 1979; Easwaramoorthy and Jayaraj, 1989) may explain why infection rates in larvae of C. infuscatellus as high as 31% have been observed in sugarcane in India (Easwaramoorthy and Jayaraj, 1987). Other pathogens too, notably some of the microsporidia are known to exhibit vertical transmission (Henry, 1981; Andreadis, 1987) and there is potential for their use as classical biological control agents providing long term population suppression as part of an IPM strategy.

320 In the current surveys, the S. calamistis CPV, the B. fusca GV and an unidentified microsporidian from S. calamistis are potential classical biological control agents for the maize stem borer ecosystem. However, no pathogen was identified which was clearly and specifically responsible for S. calamistis population crashes referred to in the introduction. Innundative and inoculative biological control with entomopathogens The cryptic feeding habit of stem borer larvae may be a specific adaptation allowing them to avoid many natural enemies and pathogens. This is especially effective in the thick-stemmed cultivated graminaceous plants like maize, sorghum and millet where natural larval infection rates by pathogens in these crops are usually low. Their cryptic nature makes them a very difficult target for any non-systemic topically applied pesticide, whether biological or chemical. To have a direct effect on crops, inundative control must target the phase when the pest is exposed or migrating, and this varies with species. Since each species would only be available for treatment with topical biopesticides during a specific window, the farmer would be required to monitor his crop and ensure complete and persistent coverage during each window. This, in turn, may require more than one application due to growth dilution and pathogen inactivation in the field. Although B.t. has a wide host range, baculoviruses and to a lesser extent the microsporidia and fungi, have narrow host ranges, which would dictate the use of several biological control agents to treat a species complex in one crop. There may be scope for the development of applications of B. bassiana or M. anisopliae to the whorl, as recommended for Ostrinil. In this case, one would be relying on the plant providing good UV protection, and rain splash continuously spreading the inoculum down the plant. Even slight systemic activity through endophytic, or even epiphytic growth, would greatly enhance its effect. The maize ecosystem in Africa is a harsh environment requiring robust or specialized formulations, which may be beyond the capacity of small scale producers. For a microbial pesticide to be a feasible option within this system, it would have to give good control of the key stem-borer species, be deliverable at low cost, and compatible with existing practices. Currently, the best options for inundative microbial control of stem borers are probably represented by existing B.t- based commercial products or by B. bassiana, either as granules or as aqueous emulsions. Despite a common preconception, microbial pesticides are not necessarily cheaper than conventional pesticides (Cherry et al., 1999), especially where the market is small. Market size may be small because of the restricted host range of the biopesticides, or limited farmer resources, and applies even where agents are mass produced in labour

321 intensive cottage style units. There are additional factors associated with distribution and storage of microbials, which can make them inappropriate for use by small-scale or resource poor farmers, not only in Africa. Currently, a microbial pesticide suitable for application against the complex of stem borer species is not readily available to the majority of resource-poor farmers in Africa. Novel control strategies with microbial control agents From a consideration of the above factors, an alternative method of deploying micobial control agents will be necessary. Incorporating the pathogen into the plant could be achieved either by genetic engineering or by the use of an endophytic microbe. Expression of B.t. endotoxins in maize plants following insertion of protein genes into the plant genome can provide a high level of protection from insect attack including stem borers (Vaeck et al., 1987; Perlak et al., 1990; James, 1998). However, there are many obstacles to the implementation of this technology in Africa (CIMMYT, 1994; Krattiger, 1997). An alternative to genetic engineering is the use of endophytic entomopathogenic micro-organisms. The capacity of B. bassiana to grow systemically inside maize plants was first reported by Bing and Lewis (1992), who demonstrated the reduction in yield loss in Beauveria-infected plants. In our current work, the B. bassiana isolates used were collected on insects and tested for their ability to invade the maize plant. All isolates tested appeared able to grow inside the plant and provided some indications of reduction in damage and attack levels. However, many questions remain and most of the described effects need more thorough confirmation. If the research issues can be solved, the use of systemic Beauveria might provide the most appropriate and available option for resource poor farmer in Africa. The simple technology, either as a seed dressing or application to the whorl should provide season long suppression of stem borers. Conclusions Various IPM technologies are being developed against stem borers. It is unlikely that host plant resistance will achieve the levels of suppression required, and chemical pesticides are beyond the means of most African maize farmers. Genetically engineered maize gives good control of European corn-borer in the USA, but the socio-economic, safety, regulatory and environmental issues involved are likely to delay the deployment of this technology in Africa for some years. Classical biological control with Hymenoptera offers good prospects, as does the use of wild host grasses as trap crops. Insect pathogens with the capacity for vertical transmission may also have poten-

322 tial as classical biocontrol agents. The use of Beauveria isolates, particularly those with systemic properties, would fit well within the broad IPM context being developed. Many research issues remain to be explored, principally the interaction with other micro-organisms in the maize plant, and any possible negative impact on yield or consumer health. The interaction with hymenopterous biological control agents needs to be explored, but this technology would fit well with cultural controls based on trap plants. As the amounts of inoculum required would be small, the cost should be low and within reach of maize farmers of any scale.

Acknowledgements Thanks to François Onikpo, Eliane Dossoumou, Flavien Zinsou, Agnassim Banito, Janvier Housou and Mathieu Balogoun for their technical assistance, to Margaret Brown (NRI, Chatham, UK) and Jenny Cory (IVEM, Oxford, UK) for REN analysis of virus isolates, to Alistair Bishop (UoG, Chatham, UK) for supply of B.t. HD1 and to Sam Korie (Biometric Unit, IITA) for statistical advice. This paper is an output from a project funded by the UK Department For International Development (DFID) [CRF/R6400H] for the benefit of developing countries. The views expressed are not necessarily those of DFID. The authors gratefully acknowledge additional support from the aid agency of Austria and donors supporting IITA through core funding.

References Andreadis, T.G., 1987. Horizontal transmission of Nosema pyrausta (Microsporida: Nosematidae) in the European Corn Borer, Ostrinia nubilalis (Lepidoptera: Pyralidae). Environ. Entomol. 16: 1124–1129. Bartlett, K. and C. Lefevre, 1934. Field experiments with Beauveria bassiana (Bals.) Vuill., a fungus attacking the European corn borer. J. Econ. Entomol. 24: 1147–1157. Bernhard, K., P. Jarrett, M. Meadows, J. Butt, D.J. Ellis, G.M. Roberts, S. Pauli, P. Rodgers and H.D. Burges, 1997. Natural isolates of Bacillus thuringiensis: worldwide distribution, characterisation, and activity against insect pests. J. Invert. Pathol. 70: 59–68. Bing, L.A. and L.C. Lewis, 1992. Temporal relationships between Zea mays, Ostrinia nubilalis (Lep.: Pyralidae) and endophytic Beauveria bassiana. Entomophaga 37(4): 525–536. Bing, L.A. and L.C. Lewis, 1991. Suppression of Ostrinia nubilalis (Hubner) (Lepidoptera: Pyralidae) by endophytic Beauveria bassiana (Balsamo) Vuillemin. Environ. Entomol. 20: 1207–1211. Bohorova, N., M. Cabrera, C. Abarca, R. Quintero, A.M. Maciel, R.M. Brito, D. Hoisington and A. Bravo, 1997. Susceptibility of four tropical lepidoteran maize pests to Bacillus thuringiensis Cry1-type insecticidal toxins. J. Econ. Entomol. 90: 412–415.

323 Bordat, D., A.E. Goudegnon and G. Bouix, 1994. Relationships between Apanteles flavipes (Hym: Braconidae) and Nosema bordati (Micospora, Nosematidae) parasites of Chilo partellus (Lep.: Pyralidae). Entomophaga 39: 21–32. Bordat, D., A.E. Goudegnon and G. Bouix, 1993. Influence of Nosema bordati (Microsporida: Nosematidae) on the biological cycle of Chilo partellus (Lepidoptera: Pyralidae), a stem borer of cultivated Graminae. Entomophaga 38: 551–564. Bosque-Pérez, N.A. and J.H. Mareck, 1991. Effect of the stem-borer Eldana saccharina (Lepidoptera: Pyralidae) on the yield of maize. Bull. Entomol. Res. 81: 243–247. Bronbridge, M. and T. Onyango, 1992. Laboratory evaluation of four commercial preparations of Bacillus thuringiensis (Berliner) against the spotted stem borer Chilo partellus (Swinhoe) (Lep., Pyralidae). J. Appl. Entomol. 113: 159–167. Brownbridge, M. and T. Onyango, 1992. Screening of exotic and locally isolated Bacillus thuringiensis (Berliner) strains in Kenya for toxicity to the spotted stem borer, Chilo partellus (Swinhoe). Trop. Pest Manage. 38: 77–81. Broza, M., M. Brownbridge, M. Hamal and B. Sneh, 1991. Control of the African armyworm Spodoptera exempta Walker (Lepidoptera: Noctuidae) in Kenyan fields with highly effective strains of Bacillus thuringiensis Berliner. Biocontrol Sci. Technol. 1: 127–135. Canning, E.U., 1993. Microsporidia. In: J.P. Krier (ed), Toxoplasma, Crytosporidia, Pneumocystis and Microsporidia. Academic Press, Inc., New York. pp. 299–370. Cardwell, K.F., F. Schulthess, R. Ndemah and Z. Ngoko, 1997. A systems approach to assess crop health and maize yield losses due to pests and disease in Cameroon. Agric. Ecosyst. Environ. 65: 33–47. Cherry, A.J., N.E. Jenkins, G. Heviefo, R.G. Bateman and C.J. Lomer, 1999. Operational and economic analysis of a West African pilot scale production plant for aerial conidia of Metarhizium spp. for use as a mycoinsecticide against locusts and grasshoppers. Biocontrol Sci. Technol. 9: 35–51. Chiuo, W.C. and R.F. Hou, 1993. Infection of the Asian corn borer, Ostrinia furnacalis Guenee (Lep., Pyralidae), with entomopathogens under screen house conditions. J. Appl. Entomol. 15: 246–253. CIMMYT, 1994. Biosafety at CIMMYT; Internal Guidelines. International Maize and Wheat Improvement Center (CYMMIT), Mexico City, Mexico. CIMMYT, 1989. Towards Insect Resistant Maize for the Third World. Proceedings of the International Symposium on methodologies for developing host plant resistance of maize insects. International Maize and Wheat Improvement Center (CIMMYT), Mexico City, Mexico. David, H. and S. Easwaramoorthy, 1990. Biological control of Chilo spp. in sugar-cane. Insect Sci. Appl. 11: 733–745. Denké, D., 1995. Influence de lafumure potassique sur la bioecologie des principaux foreurs du maïs (Zea mays L.) Cas de Sesamia calamistis Hampson (Lepidoptera: Noctuidae) & Eldana saccharina Walker (Lepidoptera: Pyralidae). Ing. Agr. thesis, University du Bénin, Lomé, Togo. 83 pp. Djaman, K., 1997. Le role des maladies des insectes dans l’ecosysteme des foreurs de tige et mineurs d’épi du mais: inventaire et efficacité des entomopathogènes: Cas du Bénin. Ing. Agr. thesis, Ecole Superieure D’Agronomie, Lomé-Togo. Drolet, J., I. Gerke, and H. Rasamimanana, 1998. Controle microbiologiques de borer blanc du riz à Madagascar par un virus d’insecte. Antennae 5: 19. Easwaramoorthy, S. and S. Jayaraj, 1989. Vertical transmission of granulosis virus of sugarcane shoot borer, Chilo infuscatellus Snell. Trop. Pest Manage. 35: 352–353.

324 Easwaramoorthy, S. and S. Jayaraj, 1987. Cross infectivity of granulosis viruses infecting Chilo infuscatellus Snell. and C. sacchariphagus indicus (Kapur). J. Entomol. Res. 11: 170–174. Eijnatten, C.L.M.V., 1965. Towards the Improvement of Maize in Nigeria. H. Veeman and N.V. Zonen, Wageningen. Endrödy-Younga, S. 1968. The stem borer Sesamia botanephaga Tams & Bowden (Lep. Noctuidae) and the maize crop in central Ashanti, Ghana. Ghana J. Agric. Sci. 1: 103–131. FAO, 1991. FAO Production Year Book. Food and Agricultural Organization, Rome. Fediere, G., A.A. Taha, J.C. Veyrunes, X. Lery, J.L. Zeddam, S. Abol-Ela, M. El-Husseini and J. Giannotti, 1994. Characterization of a picorna-like virus isolated from the maize stem borer Sesamia cretica Led. (Noctuidae) in Egypt. Bull. IOBC/WPRS 17: 241–243. Fediere, G., A. Taha, S.A. Ela, X. Lery, J.L. Zeddam, J.C. Veyrunes and J. Giannotti, 1993. Account of a granulosis virus of Sesamia cretica Led. (Lepidoptera Noctuidae), a major pest of maize in north-east Africa: DNA characterization and viral diagnosis. Comptes Rendus de l’Academie des Sciences, Series 3, Science de la Vie 316: 1350–1354. Gerke, I. and J. Drolet, 1994. Protection Integrée en Riziculture au Lac Alaotra. Swiss Development Corporation, report 8A. Girling, D.J., 1980. Eldana saccharina as a crop pest in Ghana. Trop. Pest Manage. 26: 152– 156. Godse, D.B. and P. Nayak, 1983. Nuclear polyhedrosis of Sesamia inferens (Noctuidae: Lepidoptera) the pink stem borer of rice. Curr. Sci. 52: 682–683. Gounou, S., F.Schulthess, T. Shanower, W.H.O. Hammond, H. Braima, A.R. Cudjoe, R. Adjakloe, K.K. Antwi and I. Olaleye, 1994. Stem and Ear Borers of Maize in Ghana. Plant Health Management Research Monograph 4, International Institute of Tropical Agriculture, Nigeria. Hedgecock, S., D. Moore, P.M. Higgins and C. Prior, 1995. Influence of moisture content on temperature tolerance and storage of Metarhizium flavoviride conidia in an oil formulation. Biocontrol Sci. Technol. 5: 371–377. Henry, J.E., 1981. Natural and applied control of insects of protozoa. Annu. Rev. Entomol. 26: 49–73. Hochberg, M., 1989. The potential role of pathogens in biological control. Nature, 337: 262– 265. Humber, R.A., 1997. Fungi: Identification. In: L. Lacey (ed), Manual of Techniques in Insect Pathology. Academic Press, San Diego. pp. 153–85. ICIPE, 1991. Annual report of the International Center for Insect Physiology and Ecology. ICIPE, Nairobi, Kenya. Jacqumard, P., G. Crozier, A. Amargier, J.-C. Veyrunes, L. Croizier, D. Bordat and B. Vercambre, 1985. Présence de trois virus chez Sesamia calamistis Hampson (lépidoptère, Agrotidae) à l’ile de la Réunion. L’Agronome Tropicale 40: 66–71. James, C.M. 1998. Global Status of Transgenic Crops in 1997. ISAAA Briefs no. 2, International Service of the Acquisition of Agri-biotech Applications, Ithaca, New York. Johnson, C. and A.H. Bishop, 1996. A technique for the effective enrichment and isolation of Bacillus thuringiensis. FEMS Microbiol. Lett. 142: 173–177. Jones, K.A., 1993. Biocontrol of Rice Stem Borer in Madagascar: Consultancy Report and Recommendations. Natural Resources Institute, Chatham, UK. Jones, K.A., A. Westby, P.J.A. Reilly, and M.J. Jeger, 1993. The exploitation of microorganisms in the developing countries of the tropics. In: D.G. Jones (ed), Exploitation of Microorganisms. Chapman and Hall, London.

325 Jones, K.A. 1988. The use of insect viruses for pest control in developing countries. Aspects Appl. Biol. 17: 263–270. Kfir, R., 1995. Parasitoids of the African stem borer, Busseola fusca (Lepidoptera: Noctuidae), in South Africa. Bull. Entomol. Res. 85: 336–377. Kfir, R. and R.A. Bell, 1993. Intraseasonal changes in populations of the Africa stem borer Busseola fusca (Fuller) (Lepidoptera: Noctuidae) and its parasitoids in Natal, South Africa. J. Afr. Zool. 107: 543–553. Khan, Z.R., K. Ampong-Nyarko, P. Chiliswa, A. Hassanali, S. Kimani, W. Lwande, W.A. Overholt, J.A. Pickett, L.E. Smart, L.J. Wadhams and C.M. Woodcock, 1997. Intercropping increases parasitism of pests. Nature 388: 631–632. Kling, J.G. and G. Edmeades, 1997. Morphology and Growth of Maize. IITA/CIMMYT Research Guide 9, International Institute of Tropical Agriculture, Ibaden, Nigeria. Kouame, K.L., 1995. Seasonal abundance of the two maize stem borers Sesamia calamistis and Eldana saccharina and bionomics of the Sesamia egg parasitoid, Telenomus busseolae. Ph.D. Thesis, Simon Fraser University, British Columbia, Canada. Koziel, M.G., G.L. Beland, C. Bowman, N.B. Carozzi, R. Crenshaw, L. Crossland, J. Dawson, N. Desai, M. Hill, S. Kadwell, K. Launis, D. Maddox, K. McPherson, M.R. Meghji, E. Merlin, R. Rhodes, G.W. Warren, M. Wright and S.V. Evola, 1993. Field performance of elite transgenic maize plants expressing an insecticidal protein derived from Bacillus thuringiensis. Bio/Technology 11: 194–200. Krattiger, A.F., 1997. Insect Resistance in Crops: a Case Study of Bacillus thuringiensis (Bt) and its transfer to developing countries. ISAAA Briefs no. 2, International Service of the Aquisition of Agri-biotech Applications, Ithaca, New York. Lewis, L.C., E.C. Berry, J.J. Obrycki and L.A. Bing, 1996. Aptness of insecticides (Bacillus thuringiensis and carbofuran) with endophytic Beauveria bassiana, in suppressing larval populations of the European corn borer. Agri. Ecosyst. Environ. 57: 27–34. Lewis, L.C. and J.E. Cossentine, 1986. Season long intraplant epizootics of entomopathogens, Beauveria bassiana and Nosema pyrausta, in a corn agroecosystem. Entomophaga 31: 363–369. Lewis, L.C. and T.B. Johnson, 1982. Efficacy of two nuclear polyhedrosis viruses against Ostrinia nubilalis (Hubner) (Lep.: Pyralidae) in the laboratory and field. Entomophaga 27: 33–38. Lewis, L.C. and R.E. Lynch, 1978. Foliar applications of Nosema pyrausta for suppression of populations of European corn borer. Entomophaga 23: 83–88. Maniania, N.K., 1993. Effectiveness of the entomopathogenic fungus Beauveria bassiana (Bals.) Vuill. for control of the stem borer Chilo partellus (Swinhoe) in maize in Kenya. Crop Protect. 12: 601–604. Maniania, N.K., 1992. Pathogenicity of entomogenous fungi (Hyphomycetes) to larvae of the stem borers, Chilo partellus Swinhoe and Busseola fusca Fuller. Insect Sci. Appl. 15: 691–696. Mbapila, J.C. and W.A. Overholt, 1997. Oviposition, development and searching behaviour of Cotesia flavipes Cameron and Cotesia sesamiae (Cameron) (Hymenoptera: Braconidae) in relation to aestivating and non-aestivating Chilo spp. (Lepidoptera: Pyralidae) larvae in maize. Afr. Entomol. 5: 35–39. McGuire, M.R., B.S. Shasha, L.C. Lewis, R.J. Bartlett and K. Kinney, 1990. Field evaluation of granular starch formulations of Bacillus thuringiensis against Ostrinia nubilalis (Lepidoptera: Pyralidae). J. Econ. Entomol. 83: 2207–2210.

326 Medvecky, B.A. and F.G. Zalom, 1992. Conventional and alternative insecticides, including a granular formulation of Bacillus thuringiensis var. kurstaki, for the control of Busseola fusca (Fuller) (Lepioptera: Noctuidae) in Kenya. Trop. Pest Manage. 38: 186–189. Melamed-Madjar, V. and B. Raccah, 1979. The transstadial and vertical transmission of a granulosis virus from the corn borer Sesamia nonagriodes. J. Invert. Pathol. 33: 259–264. Morel, G., B. Vercambre, M. Bergoin and C. Louis, 1981. A cytoplasmic polyhedrosis in the spotted borer, Chilo sacchariphagus Boj. (Lepidoptera: Pyralidae). Agron. Trop. 36: 378–381. Neupane, F.P., 1990. Status and control of Chilo spp. on cereal crops in Southern Asia. Insect Sci. Appl. 11: 501–534. Odindo, M.O., M. Opondo-Mbai and P.A. Amutalla, 1992. Effect of sorghum phenology on the control of Chilo partellus with Nosema marucae (Microspora: Nosematidae). Biocontrol Sci. Technol. 2: 307–314. Odindo, M.O., W.A. Otiengo, G.W. Oloo, J. Kilori and R.C. Odhiambo, 1989. Prevalance of microorganisms in field-sampled borers on sorghum, maize, and cowpea in western Kenya. Insect Sci. Appl. 10: 225–228. Orr, D.B., L.C. Lewis and J.J. Obrycki, 1994. Behaviour and survival in corn plants of Ostrinia nubilalis (Lepidoptera: Pyralidae) when infected with Nosema pyrausta (Microspora: Nosematidae) and parasitized by Macrocentrus grandii (Hymenoptera: Braconidae). Environ. Entomol. 23: 1020–1024. Perlak, F.J., R.W. Deaton, T.A. Armstrong, R.L. Fuchs, S.M. Sims, J.T. Greeanplate and D.A. Fihshoff, 1990. Insect resistant cotton plants. Biotechnology 8: 939–943. Poinar, G.O.J. and G.M. Thomas. 1984. Laboratory Guide to Insect Pathogens and Parasites. Plenum Press, New York. Raccah, B., V. Melamed-Madjar and J. Cohen, 1975. A granulosis virus from Sesamia nonagriodes. J. Invert. Pathol. 26: 145–146. Riba, G., 1984. Application en essais parcellaires de plein champ d’un mutant artificiel du champignon entomopathogene Beauveria bassiana (Hyphomycete) contre la pyrale du mais, Ostrinia nubilalis (Lep.: Pyralidae). Entomophaga 29: 41–48. Robertson, J.L., H.K. Preisler, S.S. Ng, L.A. Hickle and W.D. Gelertner, 1995. Natural variation: A complicating factor in bioassays with chemical and microbial pesticides. J. Econ. Entomol. 88(1): 1–10. Schulthess, F., N.A. Bosque-Pérez, A. Chabi-Olayé, S. Gounou, R. Ndemah and G. Georgen, 1997. Exchange of natural enemies of lepidoteran cereal stemborers between African regions. Insect. Sci. Appl. 17: 97–108. Sekloka, E., 1996. Contribution à l’étude bioécologique de Eldana saccharina Walker (Lep., Pyralidae) sur Zea mays L. et les plantes hôtes alternatives. Ing. Agr. thesis, University of Benin. Sémeglo, K.A., 1997. Etude comparative de la bioécologie de trois Noctuidae foreurs de tiges de céréales: Sesamia calamistis, Sesamia poephaga et Busseola fusca zur Zea mays L. et quelques poacées sauvages. Ing. Agr. thesis. University of Benin. Sétamou, M., K.F. Cardwell, F. Schulthess and K. Hell, 1998. Effect of insect damage to maize ears, with special reference to Mussidia nigrivenella (Lepidoptera: Pyralidae), on Aspergillus flavus (Deuteromyces: Monoliales) infection and aflatoxin production in maize before harvest in the Republic of Benin. Ecol. Behav. 91: 433–438. Sétamou, M., K.F. Cardwell, F.Schulthess and K. Hell, 1997. Aspergillus flavus infection and aflatoxin contamination of preharvest maize in Benin. Plant Dis. 81: 1323–1327. Sétamou, M. and F. Schulthess, 1995. The influence of egg parasitoids belonging to the Telonomus busseolae (Hymenoptera: Scelionidae) species complex on Sesamia calamistis

327 (Lepidoptera: Noctuidae) populations in maize fields in southern Benin. Biocontrol Sci. Technol. 5: 69–81. Sétamou, M., F. Schulthess, N.A. Bosque-Pérez and A. Thomas-Odjo, 1995. The effect of stem and cob borers subjected to different nitrogen treatments with special reference to Sesamia calamistis Hampson (Lepidoptera; Noctuidae). Entomol. Exp. Appl. 77: 205–210. Sétamou, M., F. Schulthess, N.A. Bosque-Pérez and A. Thomas-Odjo, 1993. Effect of plant nitrogen and silica on the bionomics of Sesamia calamistis (Lepidoptera: Noctuidae). Bull. Entomol. Res. 83: 405–411. Shanower, T., F. Schulthess and S. Gounou, 1991. Distribution and Abundance of Some Stem and Cob Borers in Benin. Plant Health Management Research Monograph 1. International Institute of Tropical Agriculture, Ibadan, Nigeria. Sneath, P.H.A., 1986. Endospore-forming gram-positive rods and cocci. In: P.H.A. Sneath, N.S. Mair, M.E. Shape and J.G. Holt (eds), Bergey’s Manual of Systematic Bacteriology. Wiliams and Wilkins, Baltimore, London, Los Angeles, Sydney. pp. 1104–1140. SPSS, M.J. Norusis (ed.), 1995. SPSS for Windows. Release 7.0, base system user’s guide. SPSS Inc., Chicago. Tanada, Y. and H.K. Kaya, 1993. Insect Pathology. Academic Press, Inc., San Diego. Thiery, I. and E. Frachon, 1997. Identification, isolation, culture and preservation of entomopathogenic bacteria. In: L. Lacey (ed), Manual of Techniques in Insect Pathology. Academic Press, San Diego. pp. 55–77. Usua, E.J., 1968. Effect of varying populations of Busseola fusca larvae on the growth and yield of maize. J. Econ. Entomol. 61: 375–376. Vaeck, M., A. Reynaerts, H. Hofte, S. Jansens, M. De Beuckeleer, C. Dean, M. Zabeau, M. Van Montagu and J. Leemans, 1987. Transgenic plants protected from insect attack. Nature 328: 33–37. Vago, C., 1956. Actions virusales indirectes. Entomophaga 1: 82–87. Wagner, B.L. 1997. The colonisation of maize, Zea mays L., by the entomopathogenic fungus, Beauveria bassiana (Balsamo) Vuillemin. Ph.D. Thesis, Iowa State University. Walters, H.S. and R. Kfir, 1993. Development of Nosema partelli sp. n. (Protozoa: Microsporida: Nosematidae) in the stem borer Chilo partellus (Swinhoe) (Lepidoptera: Pyralidae). Afr. Entomol. 1: 57–62.