+Bioline International Official Site (site up-dated regularly)

¹International Centre of Insect Physiology and Ecology, P.O. Box 30772, Nairobi, Kenya;
²IACR-Rothamsted, Harpenden, Hertfordshire AL5 2JQ, UK;
³Kenya Agricultural Research Institute, P.O. Box 450, Kitale, Kenya

ABSTRACT

Maize is the principal food and cash crop for millions of people in the predominantly mixed crop-livestock farming systems in Kenya. Stemborers and striga (Striga hermonthica) are major constraints to increased maize production in eastern Africa. An intercropping and trap crop system has been developed, using a 'push-pull' strategy, for the control of stemborers in small scale maize farming systems. The 'push-pull' strategy involves trapping stemborers on highly susceptible trap plants (pull) and driving them away from the crop using repellent intercrops (push). Napier grass (Pennisetum purpureum Schumach) and Sudan grass (Sorghum vulgare sudanense Stapf.) are used as trap plants, whereas molasses grass (Melinis minutiflora Beauv.) and two species of desmodium (Desmodium uncinatum Jacq. and Desmodium intortum Urb.) repel ovipositing stemborers. The integrated 'push-pull' strategies were shown to increase parasitism of stemborers through attraction of parasitoids to one of the intercrops, molasses grass. The leguminous intercrop, silver leaf desmodium, drastically reduced damage to maize by the parasitic weed, striga. This aspect was further investigated and developed for integration with stemborer control. On-farm trials with farmers in Kenya have shown significant yield increases in maize farming.

Key Words: stemborers, maize, habitat management, 'push-pull', Kenya, Striga hermonthica

RÉSUMÉ

Le maïs est la principale nourriture et culture commerciale pour des millions de personnes pratiquant le système agricole prédominant au Kenya, associant la culture et l'élevage. Les foreurs des tiges et le striga (Striga hermonthica) sont les contraintes majeures de l'accroissement de la production de maïs en Afrique de l'Est. Un système de cultures associées et pièges, utilisant une stratégie de répulsion-attraction, à été développé pour contrôler les foreurs des tiges dans les petites exploitations agricoles. La stratégie de répulsion-attraction consiste à piéger les foreurs de tiges sur des plantes pièges fortement attractives (attraction) et à les éloigner de la culture en utilisant des cultures associées répulsives (répulsion). On utilise l'herbe à éléphants (Pennisetum purpureum Schumach) et l'herbe du Soudan (Sorghum vulgare sudanense Stapf) comme plantes pièges tandis que l'herbe à mélasse (Melinis minutiflora Beauv) et deux espèces de desmodium (Desmodium uncinatum Jacq. et Desmodium intortum Urb.) repoussent les foreurs de tiges prêts à pondre. La stratégie de répulsion-attraction permet d'augmenter le parasitisme des foreurs des tiges en attirant les parasitoïdes vers l'une des cultures associées, l'herbe à mélasse. L'association du desmodium à feuilles argentées (Légumineuse), réduit de façon drastique les dégâts causés au maïs par l'herbe parasite striga. Ce résultat a été approfondi et développé pour être associé avec la lutte contre les foreurs des tiges. Des essais au champ réalisés avec des fermiers au Kenya ont montré des augmentations de rendements significatives dans la culture du maïs.

Mots Clés: foreurs des tiges, maïs, gestion de l'habitat, répulsion-attraction, Kenya, Striga hermonthica

INTRODUCTION

Maize is the principal food and cash crop for millions of people in the predominantly mixed crop-livestock farming systems of eastern Africa. The stemborers Chilo partellus (Swinhoe) (Lepidoptera: Crambidae) and Busseola fusca Fuller (Lepidoptera: Noctuidae) and striga weed, (Striga hermonthica (Scrophulariaceae), are major biotic constraints to increased maize production in eastern Africa. Several national and international agricultural research centres continue their search for technologies that would lead to increased farm production due to improved stemborer and striga management (ECAMAW, 1998).

At least four species of stemborers infest maize in the region, causing reported yield losses of 20-40 %. Stemborers are difficult to control, largely because of the cryptic and nocturnal habits of the adult moths and the protection provided by the stem of the host crop for immature stages (Ampofo, 1986; Seshu Reddy and Sum, 1992). The main method of stemborer control that the Kenyan Ministry of Agriculture recommends is the use of chemical pesticides, which is often uneconomical and impractical to many of the smallholders.

Parasitic weeds in the genus Striga infest 40 % of arable land in the savanna region, causing an estimated annual loss of $7 to $13 billion (M'boob, 1989; Musselman et al., 1991; Lagoke et al., 1991). Infestation by Striga spp. has resulted in the abandonment of much arable land by farmers in Africa. The problem is more widespread and serious in areas where both soil fertility and rainfall are low. Weeding out striga is a time-consuming and labour-intensive activity, and the currently recommended control methods, including heavy applications of nitrogen fertiliser, crop rotation, use of trap crops and chemicals to stimulate suicidal seed germination, hoeing and hand pulling, herbicide application and the use of resistant or tolerant crop varieties (Berner et al., 1995), have limited acceptability to farmers, for both biological and socioeconomic reasons (Lagoke et al., 1991). No single method of control has so far provided a solution to both the stemborer and striga problems (Berner et al., 1995).

As part of our continuing efforts to manage cereal stemborers in eastern Africa (Khan et al. 1997a,b, 2000), we have investigated insect-plant interactions with the aim of identifying mechanisms by which stemborers colonise plants. Based on the information gathered on the interactions between stemborers and their host and non-host plants, we have developed the 'push-pull' or stimulo-deterrent diversionary strategy (Khan et al., 2000) to manage cereal stemborers in maize-based farming systems in eastern Africa. The strategy involves the use of both trap and repellent fodder plants, enabling stemborers to be simultaneously repelled from the maize crop and attracted to the trap plants.

Several plants have been identified which could be used as trap or repellent plants in a 'push-pull' strategy (Khan et al., 2000). Those that appear particularly promising are Napier grass (Pennisetum purpureum Schumach ), Sudan grass (Sorghum vulgare sudanense Stapf.), molasses grass (Melinis minutiflora Beauv.), silver leaf desmodium (Desmodium uncinatum Jacq.) and greenleaf desmodium (Desmodium intortum Urb.). Napier grass and Sudan grass have shown potential for use as trap plants, whereas molasses grass and the two desmodium species repel ovipositing stemborers. Molasses grass, when intercropped with maize, not only reduces infestation of the maize by stemborers, but also increases stemborer parasitism by a natural enemy, Cotesia sesamiae (Khan et al., 1997a). The leguminous intercrop of both species of desmodium significantly reduced striga damage to maize. This aspect has been further investigated and developed for integration with stemborer control (Khan et al., 2000).

This paper reports the results of on-farm 'push-pull' trials conducted with 50 farmers in the Suba and Trans Nzoia districts of Kenya during the long rains of 1998 and 1999.

MATERIALS AND METHODS

The participating farmers planted several combinations of attractant and repellent plants with stemborer-susceptible maize varieties in two locations during the long rainy seasons of 1998 and 1999. One location was Suba district, Kenya on the shores of Lake Victoria, which represents a low potential area of mid-altitude, hot and humid ecology. Maize in this area is infested with two major stemborer species, Ch. partellus and B. fusca, and with striga weed. Another location was the Trans Nzoia district representing a high potential, high altitude, wet and cool ecology. Maize in Trans Nzoia district is infested by B. fusca, but there is no striga infestation.

To establish a 'push-pull' trial, farmers planted a 1-m-wide perimeter of Napier grass trap plants around 900 m²maize plots with 1-m paths between maize and trap plants. Trap plants were established 2 weeks before the maize crop was planted. The maize variety used was a susceptible, medium maturity, commercial hybrid 511 in Suba District and a susceptible, long maturity, commercial hybrid 624 in Trans Nzoia District. Maize rows were spaced at 0.75 m while hills within the row were at 0.30 m. Maize was intercropped with desmodium in alternate rows and with molasses grass in the ratio of 10 rows of maize to one row of molasses grass. A control treatment of unprotected maize (maize mono) of equal size was planted 15 m away from the 'push-pull' treatment.

Sampling for stemborers and striga was done in random paired plots of 2 x 2 m along perpendicular transect lines bisecting both the treatment and control. The paired sample plots were replicated four times on the radii of the transect blocks and numbered from the perimeter to the centre of the field (Smart et al., 1989). Five hills were sampled in every sampling unit during each of the four sampling dates. Both destructive and nondestructive sampling of maize for the treatment and the control plots was done and data were collected for plants damaged by stemborers and number of stemborers recovered from the dissected plants. The borers were placed in vials labelled with block and plot numbers and reared in the laboratory on host diet for data collection on parasitisation. The number of striga plants per maize plant was recorded. Rating for striga was made on a 0-9 scale: 0, no striga; 1, 1-2 striga plants/maize plant; 3, 3-4 striga plants/maize plant; 5, 5-6 striga plants/maize plant; 7, 7-8 striga plants/maize plant; 9, more than 9 striga plants/maize plant. Yield was obtained as grain weight at 14% moisture in tonnes per hectare.

Differences in means were analysed using one-way analysis of variance (ANOVA) and means were compared using Tukey's Studentised range test (SAS Institute, 1996). Data in Tables 1 and 2 were analysed using t-test. In all studies, a P value less than or equal to 0.05 indicated significance.

Cost-benefit analysis was applied to assess the economic returns to investments in habitat management strategies (Goletti and Govindani 1995; Nelson et al., 1996). The analyses combined agronomic, climatic, technical and socioeconomic variables to assess profitability of alternative habitat management strategies. The analyses incorporated sensitivity tests to factors that may change over time or space, such as crop yield and price variability. Data on various components of income (benefits) and costs were gathered and analysed from the 30 participating farmers in Trans Nzoia who planted maize with Napier grass, maize with Napier grass and desmodium, or maize with Napier grass and molasses grass. All these farmers had control plots of the same size as that of treatment plots. Economic data were also collected from 10 maize farmers, all of whom used some kind of commercial insecticide to control stemborers. Difference in means of benefit-cost ratio was analysed using one-way analysis of variance (ANOVA) and means were compared using Tukey's Studentised range test (SAS Institute, 1996).

RESULTS

The numbers of stemborer-damaged plants recorded and stemborer larvae and pupae recovered from maize mono and maize in 'push-pull' fields are presented in Table 1. Stemborer infestation on maize monocrop was significantly higher than that on the maize crop in a 'push-pull' system (P < 0.05, t-test). Reduction in striga infestation by intercropping maize with the two species of Desmodium 'push-pull' systems in Suba district was significant as compared to maize mono (P < 0.05, t-test) (Fig. 1). Parasitisation by Cotesia spp. of stemborers in 'push-pull' maize fields was significantly higher than in maize mono fields (P < 0.05, t-test) (Table 2). Increase in maize yields in 'push-pull' farms due to reduction in stemborer damage in Trans Nzoia district and stemborer and striga damage in Suba district was significant (P < 0.05, t-test) (Fig. 2).

Table 1. Stemborer damage and population on maize in 'push-pull' and control plots in Trans Nzoia and Suba districts of Kenya
during 1998 and 1999 long rainy seasons¹

Maize + Napier²

Maize + Napier + desmodium³

Maize + Napier + Molasses³

Stemborer damage (%)

Stemborer 40
maize plants

Stemborer damage (%)

Stemborers/40 maize plants

Stemborer damage (%)

Stemborers/40 maize plants

Year

Place

T

C

T

C

T

C

T

C

T

C

T

C

1998

Trans Nzoia

8.3

18.8 *

18.6

37.3*

4.8

20.2**

7.7

45.4**

6.5

18.9*

10.2

41.6**

Suba

14.9

25.7*

16.9

35.9*

6.7

21.6**

8

39.4**

-

-

-

-

1999

Trans Nzoia

11.7

23.1**

22.6

49.6*

9.7

18.5**

13.6

41.8**

12.1

19.9

7.6

39.9**

Suba

18.7

29.3*

22.7

425.8*

13.5

36.6**

19.7

57.4**

-

-

-

-

¹For each technology, data collected from 10 farmers; average of four samplings. T, treatment; C, control.
²Use of Napier grass as trap plant for stemborers; ³use of Napier grass as attractant to stemborers and desmodium or molasses
grass to repel stemborers from maize fields; *, significant difference at P = 0.05 (t-test); **, significant difference at P= 0.01(t-test)
between treatment and control; -, not planted.

Table 2. Percent parasitisation of stemborer larvae by Cotesia spp. in a 'push-pull' habitat management system

Maize + Napier

Maize + Napier + desmodium

Maize + Napier + molasses

Year

Location

T

C

T

C

T

C

1998

Trans Nzoia¹

39.7

21.9*

31.7

18.8*

55.7

23.3**

Suba²

9.8

4.8*

11.7

7.5*

-

-

1999

Trans Nzoia

51.8

36.1*

42.6

26.3*

77.9

33.5**

Suba

8.9

3.8*

11.5

5.1*

-

-

For each technology, data collected from 10 farmers; average of four samplings. T, treatment; C, control.
¹Main stemborer in Trans Nzoia was Busseola fusca and larval parasitoid was Cotesia sesamiae; ²in Suba,
stemborers were mainly mixtures of Chilo partellus and Busseola fusca and larval parasitoids as mixtures of Cotesia
flavipes and Co. sesamiae.*, significant difference at P = 0.05; **, significant difference at P = 0.01 between
treatment and control; -, not planted.

In terms of net economic benefits, the results obtained from Trans Nzoia district showed highest returns from the 'push-pull' strategy using Napier grass and desmodium, followed by one using Napier grass and molasses grass (Fig. 3). Returns from maize and Napier grass alone were comparable to those fields where insecticide was applied to control stemborers. All treatments were significantly economically superior to the control.

DISCUSSION

A 'push-pull' strategy for controlling insect pests was first described by Pyke et al. (1987) to control Heliothis spp. in cotton. The pest was concentrated in a small area by the combined use of an attractant trap crop and a feeding deterrent. Later, Miller and Cowles (1990) devised the term 'stimulo-deterrent diversion' strategy for 'push-pull', and used the system to protect onions from the onion fly. They proposed to attract gravid females to onion culls and to protect the main crop with a combination of a deterrent and a toxin. However, in both cases, no consideration to natural enemies was given and a chemical deterrent or toxin was used to repel or kill the pest.

The present 'push-pull' strategy does not use any chemical deterrents or toxins, but uses repellent plants to push the pest from the main crop towards trap plants, which have inherent development-inhibitory properties against the trapped stemborers (Khan et al., 2000). The strategy also attempts exploitation of the pest's natural enemies through trap and repellent plants (Khan et al., 1997a, b).

The developed technology combines the 'push-pull' tactic to control stemborers, on the one hand, and in situ suppression of striga, on the other. The 'push-pull' tactic involves trapping stemborers on highly susceptible trap plants (pull) whilst driving them away from the maize crop using repellent inter-crops (push), whilst striga control is achieved through the use of intercrops that act through a combination of mechanisms, including abortive germination of seeds that fail to develop and attach on their hosts. By the end of 2000, 500 farmers in 6 districts in Kenya had confirmed that these approaches resulted in appreciable reductions in stemborers and striga infestations and increased maize yields (Z. R. Khan, unpublished data).

It has been our general principle that plants used in 'push-pull' pest management strategies must themselves have value for the communities involved. In the work described here, the trap crops and intercrops are all being used as forage for livestock in a zero grazing set up.

Cost-benefit analysis has shown returns to investment of over 2.2, with the maize monoculture returning less than 0.8 and pesticide intervention systems less than 1.8. A technology is perceived to be economically feasible if its net benefits (i.e. gross revenue less total costs) are positive. The higher the net benefit, the more economically viable is a technology. Similar data need to be collected from various agroecological and socioeconomic settings from different places and different types of farmers, including women farmers, in order to assess yield or income stability or variability over time. This is to assist in generating estimates of economic returns under different weather patterns in various agro-ecologies with different types of farming.

The habitat management approach, described here, is set to expand into Ethiopia, Uganda, Malawi and Tanzania. Pilot programmes have been initiated in Uganda and southern Africa, addressing stemborer and striga control in the arid and semi-arid areas. Each region has, in addition to varying climatic conditions and use of alternative cultivars, some differences in crops that must be taken into account, and considerable experience has been gained in this aspect by the pilot study in southern Africa (van den Berg et al. 2001). Whereas maize is the main crop in the farming systems in Kenya, sorghum, pearl millet and maize are important in other African countries. Pest management options in some of these regions are affected by low rainfall, the limited extent to which cattle are kept and the fact that the cattle are largely free-grazing. Whilst the increased planting of trap crops is justified by their significant roles as animal feed in eastern Africa, this may not be the case in other African countries and the 'added on' value of trap crops will have to be found in soil fertility (erosion) management or other novel uses of trap crops. However, wherever these approaches are developed for the specific needs of local farming communities, it is essential that the scientific basis of the modified systems be elucidated.

ACKNOWLEDGEMENTS

The Gatsby Charitable Foundation of UK funded this research. IACR receives grant-aided support from the Biotechnology and Biological Sciences Council of the UK. We thank N. Dibogo, D. Nyagol and A. Ndiege for technical assistance.

REFERENCES

The following images related to this document are available:

Photo images