Action: Combine trap and repellent crops in a push-pull system
Key messagesRead our guidance on Key messages before continuing
Parasitism: Two randomised, replicated, controlled studies from Kenya found that push-pull cropping systems increased parasitism of stem borer larvae. One of the studies found no effect on egg parasitism.
Natural enemies: Two randomised, replicated, controlled studies from Kenya and South Africa found push-pull systems had more natural predators, both in overall totals and the abundance of different predator groups.
Pests: Two of three studies (two randomised, replicated, controlled) in Ethiopia, Kenya and South Africa found fewer pests. One study found no effect on pest infestation, but pests were scarce throughout. Two replicated, controlled studies (one also randomised) found fewer witchweeds.
Crop damage: Two of three replicated, controlled studies (one also randomised) found less pest damage, but one study (where pest numbers were low) found effects varied between years and types of damage symptom.
Yield: Four of five replicated, controlled studies (two also randomised) found higher yields and one found no effect.
Profit and cost: Two studies in Kenya and a review found greater economic benefits. One study found higher production costs in the first year, but equal or lower costs in the following five years.
Crops studied were maize and beans.
Push-pull systems involve intercropping the main crop with plants that are repellent to pests (the ‘push’) while also growing plants (trap crops) that are attractive to pests around the main crop (the ‘pull’). This combination of repellent and attractive companion plants keeps invertebrate pests away from the crop and may provide additional benefits through improved habitat and resources for natural enemies. Push-pull systems can also be designed to suppress weeds at the same time as controlling pests. Ground-dwelling invertebrates are frequently surveyed using pitfall traps – small pots buried in the ground up to their rim and left empty or filled with liquid preservatives or water.
Here we present evidence from 10 of 13 studies testing this intervention.
Supporting evidence from individual studies
A replicated, paired, controlled study in 1998-1999 in western Kenya (Khan et al. 2001) found greater parasitism of stem borer (Crambidae and Noctuidae) larvae by wasps Cotesia spp. in a push-pull maize Zea mays system than in maize monoculture. On average, 12-43% of larvae were parasitised in a maize, Napier grass Pennisetum purpureum (trap crop) and desmodium Desmodium spp. (repellent crop) push-pull system (compared with 5-26% in a monoculture control) and 56-78% were parasitised in a maize, Napier grass and molasses grass Melinis minutiflora push-pull system (23-34% in controls). Fewer stem borers occurred in the push-pull systems, with 8-20 stem borers/40 maize plants in the maize-Napier-desmodium system vs. 39-57 stem borers in controls, and 8-10 stem borers in the maize-Napier-molasses system vs. 40-42 stem borers in controls. Maize yields were higher in the maize-Napier-desmodium (4-7 t/ha) and maize-Napier-molasses (7 t/ha) push-pull systems than controls (2-5 t/ha). Two push-pull systems were tested on 10 farms over two districts and two years. Napier grass was planted in 1 m-wide margins around 900 m² maize plots in both systems. In the first system maize and desmodium were planted in alternate rows. In the second system one row of molasses grass was planted for every 10 maize rows. A control was placed 15 m from each push-pull treatment.
A randomised, replicated, controlled study in 2001-2002 in western Kenya (Midega & Khan 2003) found more natural predators in push-pull maize Zea mays systems than in maize monocultures at three sites and in all maize growth stages. More predatory ants (Formicidae) occurred in push-pull than control plots (averaging 38-73 vs. 22-28 ants/maize growth stage in 2001, 38-109 vs. 23-59 ants in 2002) in all maize growth stages. More spiders (Araneae) occurred in push-pull than control plots in all stages in 2001 (14-35 vs. 10-20 spiders) and in the early and flowering (but not mature) stages in 2002 (11-30 vs. 7-13 spiders). More earwigs (Dermaptera) occurred in push-pull than control plots during the early (pre-flowering) stage (16 vs. 7 earwigs) but populations were similar or showed inconsistent differences in other stages. Two important predators of pest stem borers (Lepidoptera), including a ladybird Cheilomenes sp. and a lacewing Chrysopa sp., were only found in push-pull plots. Push-pull systems were tested in two long rainy seasons and comprised maize with a Napier grass Pennisetum purpureum or Sudan grass Sorghum vulgare sudanese trap crop and a silverleaf desmodium Desmodium uncinatum repellent crop. Fields were 30 x 30, 40 x 40 or 50 x 50 m.
A randomised, controlled trial in 2002-2003 in Potchefstroom, South Africa (Midega et al. 2007) found fewer pest spotted maize beetles Astylus atromaculatus in push-pull cropping systems of maize Zea mays, silverleaf desmodium Desmodium uncinatum and Napier grass Pennisetum purpureum (45-252 spotted maize beetles/plot) than in maize monocultures (453-649 beetles). The same effect occurred with Bacillus thuringiensis maize or conventional maize varieties. A greenhouse study found fewer spotted maize beetle catches in baited traps containing silverleaf desmodium (12% of captures) than baited control traps (27%), although similar tests in sorghum Sorghum bicolor fields found no effect of desmodium on beetle captures. Two push-pull plots (with different maize varieties) were compared with two monoculture controls. Push-pull plots comprised silverleaf desmodium planted between maize rows and Napier grass along plot margins. Plots were 38 x 35 m. Spotted maize beetles were counted on every maize plant in six 5 x 5 m areas/plot. In the greenhouse study, yellow water traps containing 2-phenylethanol lures were placed in cages with either a potted desmodium plant or a pot without desmodium. One hundred beetles were released in each cage and captures were monitored after 24 hours.
A replicated, controlled trial in 2003-2006 in 14 districts in western Kenya (Khan et al. 2008a) found 70-95% fewer purple witchweeds Striga hermonthica in a push-pull maize Zea mays cropping system (averaging 88 purple witchweeds/100 maize plants) than in maize monoculture (549 purple witchweeds), at 10 weeks after planting. Fewer maize plants were affected by stem borer (mainly maize stalk borer Busseola fusca and spotted borer Chilo partellus) damage in the push-pull system (averaging 6% of plants damaged) than in maize monoculture (23%). Maize yields were 37-129% higher in the push-pull (averaging 4.1 t/ha) than the monoculture system (2.2 t/ha) for all districts and seasons. In each district, the experiment took place on 20 randomly selected farms and for 3-7 cropping seasons. One push-pull and one monoculture plot was established on each farm. The push-pull system comprised silverleaf desmodium Desmodium uncinatum planted between maize rows, with three rows of Napier grass Pennisetum purpureum planted around the plot. The innermost row of Napier grass was planted 1 m from the maize crop. Stem borer damage was assessed for 100 maize plants/plot and purple witchweeds were counted in a 15 cm radius around each maize plant.
A replicated, controlled study in 1998-2004 in six districts in western Kenya (Khan et al. 2008b) found higher maize Zea mays yields in a push-pull cropping system (averaging 1.9-6.3 t/ha) than in maize-bean Phaseolus sp. intercrop (0.9-3.9 t/ha) or maize monoculture (1.0-3.9 t/ha) systems. Economic benefits were also higher in the push-pull system (averaging a US$47-880/ha gain) than in maize-bean (US$-25/ha loss to a US$491/ha gain) or maize monoculture (US$-113/ha loss to a US$156/ha gain) systems, in all but one district in one year. Total production costs were typically higher in the push-pull (US$236-394/ha) than the maize-bean (US$198-344/ha) or maize monoculture (US$172-266/ha) systems in the first study year at each site. Push-pull system costs (US$200-357/ha) were equal to or lower than costs in the maize-bean (US$221-332/ha) or monoculture (US$183-293/ha) systems in subsequent years. The push-pull system (designed to control stem borers (Lepidoptera) and weeds Striga spp.) comprised alternate rows of maize and silverleaf desmodium Desmodium uncinatum, with three rows of Napier grass Pennisetum purpureum planted around plots. Controls were intercropped maize and beans and monocropped maize. Ten farmers in each district implemented the three treatments on 600-2,225 m² plots. Yields were measured at harvest.
A randomised, replicated, controlled study in the 2002-2004 at three sites (Midega et al. 2008) found more wolf spiders (Lycosidae) in push-pull maize Zea mays cropping systems than in maize monoculture in western Kenya (averaging 31-141 vs. 19-71 wolf spiders/plot) and Potchefstroom, South Africa (15-16 vs. 6 wolf spiders). Overall spider numbers (Araneae) were also higher in push-pull than monoculture systems in Kenya (52-187 vs. 30-101 spiders/plot) and South Africa (21-28 vs. 9-11 spiders). Spider diversity was similar between cropping systems in Kenya (21-60 species/plot) but higher in push-pull than monoculture systems in South Africa (21-31 vs. 9-14 species). Wolf spider diversity was similar between systems at all sites. Each cropping system was replicated four times at two sites in Kenya (using 40 x 40 m plots) and one site in South Africa (35 x 38 m plots). The push-pull system comprised silverleaf desmodium Desmodium uncinatum grown between maize rows and Napier grass Pennisetum purpureum planted around the plots. Spiders were sampled by pitfall traps and soil samples. Five pitfalls were placed in four 15 x 15 m areas/plot and monitored weekly. Five soil samples/plot (20 x 20 x 20 cm) were taken fortnightly.
A randomised, replicated, controlled study in three seasons between 2007 and 2008 in western Kenya (Khan et al. 2009) found fewer purple witchweeds Striga hermonthica in push-pull cropping systems (1-27 plants/plot) than in control plots of intercropped maize Zea mays and beans Phaseolus vulgaris (139-269 plants) or maize monoculture (259-460 plants), 12 weeks after planting. Damage to maize plants by cereal stem borers (Lepidoptera) was lower in push-pull cropping systems (0.4-6.7% plants damaged/plot) than in maize-bean intercrop (11-18% plants) and maize monoculture (10-28%) controls at 12 weeks after planting. Maize yields were higher in push-pull systems (4.6-5.6 t/ha) than intercropped (2.6-3.1 t/ha) and monoculture (2.8-3.5 t/ha) controls. Economic benefits were also greater in the push-pull system (US$639-1,532/ha) than in intercropped (US$45-129/ha) and monoculture controls (US$-176/ha loss to a US$91/ha gain). Push-pull systems of maize and beans provided similar weed and stem borer control, as well as similar yields and benefits, to push-pull systems of maize only. The push-pull systems comprised silverleaf desmodium Desmodium uncinatum grown between rows of maize or rows of mixed maize and beans. Three rows of Napier grass Pennisetum purpureum were planted around the plots. Treatments were replicated four times at two sites in 6 x 6 m plots.
A randomised, replicated, controlled study in 2002-2003 at two sites in western Kenya (Midega et al. 2009) found proportionately greater parasitism of young stem borers (Lepidoptera) in a push-pull cropping system (19% of larvae and pupae parasitised/plot) than in maize Zea mays monoculture (9-11% parasitised). Mortality caused by other factors (such a microbial disease) was similar between the push-pull system (range of 13.0-15.2% larvae and pupae killed) and the monoculture (9.6-11.4%). Similar proportions of stem borer eggs were parasitised in the push-pull and monoculture systems (21 vs. 18-25% eggs parasitised). Push-pull and monoculture treatments were tested in 40 x 40 m plots. Push-pull plots contained silverleaf desmodium Desmodium uncinatum between rows of maize and Napier grass Pennisetum purpureum trap crops around plot margins (spaced 1 m from the crop). Treatments were replicated four times at each site. Stem borer eggs, larvae and pupae were sampled from 10 maize plants in each of four 15 x 15 m areas/plot. Samples were assessed for parasitism by parasitoid wasps (Hymenoptera) in a laboratory. A separate laboratory study found that the common parasitoid wasp Cotesia sesamiae was attracted to silverleaf desmodium flowers.
A replicated, controlled study in 2004-2005 in Sibu-Sire, Ethiopia (Belay & Foster 2010) found similar stem borer infestation in maize Zea mays grown in a push-pull system (averaging 10-14% plants infested) and a monoculture control (10-19%) at harvest. Stem borer (Noctuidae and Crambidae) larvae densities were low, but fewer occurred in the push-pull system (0.05 borers/plants) than control (0.18 borers/plant) in 2005. Numbers were similar between treatments (0.3-0.4 borers/plant) in 2004. Cob damage was also similar between push-pull (0.3-0.7% cob surface damaged) and control (0.4-0.8%) systems in both years. Stem tunnelling by stem borers was scarcer in the push-pull (0.8%) than control (1.9%) system in 2005, but similar in 2004 (0.3 vs. 0.5%). Yield was similar between the push-pull (2.4-3.3 t/ha) and control (2.0-4.6 t/ha) systems in both years. The push-pull system used greenleaf desmodium Desmodium intortum between maize rows and three rows of Napier grass Pennisetum purpureum (of 50 cm width) along plot margins. The control comprised maize only. The push-pull system and control were tested at seven sites (0.5 ha) in 0.25 ha plots each. Infestation and yield were measured in four 4 x 4 m areas/plot, damage was assessed for 20 randomly selected plants.
A review in 2010 (Khan et al. 2010) described two studies that found significant control of stem borers (Lepidoptera) and purple witchweed Striga hermonthica when maize Zea mays was grown in a pull-pull system (Khan et al. 2000, the same study as Khan et al. 2001, and Khan et al. 2008a, summarised above). Napier grass Pennisetum purpureum margins acted as a trap crop for stem borers and greenleaf desmodium Desmodium intortum or silverleaf desmodium Desmodium uncinatum intercrops acted as weed- and pest-repellent plants. One study (Khan et al. 2008a) found that the push-pull system improved maize yields by approximately 2 t/ha/season compared to maize monocultures. The push-pull system also provided higher monetary benefits than maize monocultures (Khan et al. 2008b, summarised above, and De Groote et al. 2008).
De Groote H., Vanlauwe B., Rutto E., Odhiambo G.D., Kanampiu F. & Khan Z.R. (2010) Economic analysis of different options in integrated pest and soil fertility management in maize systems of Western Kenya. Agricultural Economics, 41, 471-482
- Khan Z.R., Pickett J.A., Wadhams L. & Muyekho F. (2001) Habitat management strategies for the control of cereal stemborers and striga in maize in Kenya. Insect Science and its Application, 21, 375-380
- Midega C.A.O. & Khan Z.R. (2003) Impact of a habitat management system on diversity and abundance of maize stemborer predators in Western Kenya. Insect Science and its Application, 23, 301-308
- Midega A.O., Van den Berg J. & Khan Z.R. (2007) Habitat management in control of Astylus atromaculatus (Coleoptera: Melyridae) in maize under subsistence farming conditions in South Africa. South African Journal of Plant and Soil, 24, 188-191
- Khan Z.R., Midega C.A.O., Amudavi D.M., Hassanali A. & Pickett J.A. (2008) On-farm evaluation of the 'push-pull' technology for the control of stemborers and striga weed on maize in western Kenya. Field Crops Research, 106, 224-233
- Khan Z.R., Midega C.A.O., Njuguna E.M., Amudavi D.M., Wanyama J.M. & Pickett J.A. (2008) Economic performance of the 'push-pull' technology for stemborer and Striga control in smallholder farming systems in western Kenya. Crop Protection, 27, 1084-1097
- Midega C.A.O., Khan Z.R., Van den Berg J., Ogol C., Dippenaar-Schoeman A.S., Pickett J.A. & Wadhams L.J. (2008) Response of ground-dwelling arthropods to a 'push-pull' habitat management system: spiders as an indicator group. Journal of Applied Entomology, 132, 248-254
- Khan Z.R., Midega C.A.O., Wanyama J.M., Amudavi D.M., Hassanali A., Pittchar J. & Pickett J.A. (2009) Integration of edible beans (Phaseolus vulgaris L.) into the push-pull technology developed for stemborer and Striga control in maize-based cropping systems. Crop Protection, 28, 997-1006
- Midega C.A.O., Khan Z.R., Van den Berg J., Ogol C.K.P.O., Bruce T.J. & Pickett J.A. (2009) Non-target effects of the 'push-pull' habitat management strategy: parasitoid activity and soil fauna abundance. Crop Protection, 28, 1045-1051
- Belay D. & Foster J.E. (2010) Efficacies of habitat management techniques in managing maize stem borers in Ethiopia. Crop Protection, 29, 422-428
- Khan Z.R., Midega C.A.O., Bruce T.J.A., Hooper A.M. & Pickett J.A. (2010) Exploiting phytochemicals for developing a 'push-pull' crop protection strategy for cereal farmers in Africa. Journal of Experimental Botany, 61, 4185-4196