Integrated management of Striga hermonthica, stemborers, and declining soil fertility in western Kenya

Abstract

Striga hermonthica (Delile) Benth., stemborers, and declining soil fertility are serious threats to sustainable food production in the Lake Victoria zone of Kenya. To address these constraints, promising integrated crop management technologies were evaluated, using a multi-locational design in four sub-locations in Siaya and Vihiga district (western Kenya) for six cropping seasons. Technologies evaluated consisted of the traditional maize (Zea mays L.) – bean (Phaseolus vulgaris L.) intercrop, maize – Desmodium (Desmodium uncinatum (Jacq.) DC.) push–pull intercrop, Crotalaria (Crotalaria ochroleuca G. Don) – maize rotation, and soybean (Glycine max (L.) Merr) – maize rotation. Within each of these systems, imazapyr-coated herbicide-resistant maize (IR-maize) and fertilizer were super-imposed as sub-plot factors. The push–pull system was observed to significantly reduce Striga emergence and stemborer damage from the second season onwards. IR-maize reduced and delayed Striga emergence from the first cropping season. Differences in Striga emergence and stemborer damage between the other systems were not significantly different. After five cropping seasons, the Striga seedbank was significantly higher in the maize-bean intercrop system than in the push–pull system under both maize varieties while the rotational systems had intermediate values not different from the day zero values. Under IR-maize, the Striga seedbank was significantly lower than under local maize for all cropping systems. Maize yields varied between seasons, districts, and cropping systems. Yields in the push–pull system were higher than in the maize-bean intercrop after two seasons and in the absence of mid-season drought stress. Both maize and soybean responded significantly to fertilizer application for both districts and for most seasons. The various interventions did not substantially affect various soil fertility-related parameters after five seasons. In the short term, IR-maize integrated in a push–pull system is the most promising option to reduce Striga while the rotational systems may need a longer timeframe to reduce the Striga seedbank. Finally, farmer-led evaluation of the various technologies will determine which of those is really most acceptable under the prevailing farming conditions.

Introduction

Maize (Zea mays L.) is one of the most important cereal crops in eastern Africa, where it serves as both a staple food and cash crop for millions of people. Grain yields under farmers’ conditions in the Lake Victoria Basin in Kenya (1.0 ± 0.5 t ha⁻¹) were observed to be less than 25% of the potential yield of 4–5 t ha⁻¹ (Tittonell et al., 2005). During so-called Rapid Rural Appraisals, which are short, informal surveys with farmer groups, farmers systematically ranked Striga spp. (witchweed), stemborers, and declining soil fertility as three major constraints affecting maize production in western Kenya (Odendo et al., 2001).

Striga spp. has infested about 212,000 hectares or about 15% of the arable land in the Lake Victoria Basin of Kenya alone (www.fao.org; CEPA, 2004), causing yield losses of between 30–50%, although losses of up to 100% have been reported (Hassan et al., 1995). Increased incidence of Striga has been attributed to cereal mono-cropping and declining soil fertility (Ransom, 2000). Of the 23 species of Striga spp. prevalent in Africa, Striga hermonthica (Delile) Benth. is by far the most socio-economically important in East Africa (Emechebe and Ahonsi, 2003). In western Kenya, this weed infests about 76% of the total area under maize and sorghum (Sorghum bicolor (L.) Moench), which are the main staple crops for the people in target area, causing annual losses estimated at US$ 41 million (Hassan et al., 1995, Kanampiu et al., 2002). Adoption of recommended control methods to reduce Striga infestation has been limited, partly by farmers’ reluctance to adopt such methods, accentuated by unfavourable biological and socio-economic conditions (Kanampiu et al., 2003, Oswald, 2005).

Cereal stemborers are important injurious insect pests of maize, with Chilo partellus Swinhoe (Lepidoptera: Crambidae) and Busseola fusca Fuller (Lepidoptera: Noctuidae) being the most important in the region. Farmers’ estimates of crop loss due to stemborers in western Kenya were 12.9% (De Groote, 2002), while direct observation of damage was 13.5% (De Groote et al., 2004). Although, several insecticides are able to effectively control stemborers, their use in western Kenya is limited (De Groote, unpublished data). The effectiveness of some of the recommended cultural control methods (e.g., burning of crop residues, manipulation of planting dates, removal of infested plants) is questionable (Van den Berg et al., 1998) and as a result, most smallholder farmers do not make any conscious attempt to control stemborers (Grisley, 1997).

A number of technologies have been developed to alleviate stemborer and/or Striga constraints in smallholder farms. These include the ‘push–pull’ technology for stemborer and Striga control, and imazapyr-coated herbicide-resistant maize (IR-maize) and cereal-legume rotations for Striga control. The ‘push–pull’ technology is based on a stimulo-deterrent concept (Miller and Cowles, 1990). In this strategy, maize is intercropped with a stemborer moth-repellent plant, Desmodium uncinatum (Jacq.) DC., while an attractant host plant, Napier grass (Pennisetum purpureum Schumach.) is planted as a trap plant around this intercrop. Volatiles produced by the Desmodium repel the host-seeking moths while those produced by the Napier grass are attractive to them (Khan et al., 2000, Chamberlain et al., 2006). Studies have shown that Desmodium also significantly suppresses Striga, leading to enhanced grain yields (Khan et al., 2000, Khan et al., 2006). The root exudates of Desmodium contain blends of secondary metabolites with Striga seed germination stimulatory and post-germination inhibitory properties (Tsanuo et al., 2003). The first group of semiochemicals stimulates Striga seeds to germinate while the second group inhibits lateral growth thereby hindering the development of the haustorial root system and subsequent attachment to the host plant (Tsanuo et al., 2003).

Rotations between fast-growing, nitrogen-fixing herbaceous legumes and cereals have also been used to deplete the Striga seedbank and reduce Striga emergence, since certain legumes have the ability to trigger suicidal Striga germination (Carsky et al., 2000). Scientists at the International Institute of Tropical Agriculture, Nigeria, have developed dual purpose soybean (Glycine max (L.) Merr) germplasm that produces leafy biomass without sacrificing high grain yields, often resulting in substantial yield increases for a subsequent maize crop compared with less leafy soybean varieties (Sanginga et al., 2003). These soybean varieties were also bred for promiscuity or the ability to establish an effective N fixation symbiosis with the native Bradyrhizobium spp., thus reducing or eliminating the need for inoculation through application of external bacteria. Not surprisingly, maize growing after these improved soybean varieties had 1.2–2.3-fold grain yield increase compared to the control (Sanginga et al., 2002). Some of these soybean varieties also triggered suicidal germination of Striga (Sanginga et al., 2003). Herbaceous legumes, such as Crotalaria ochroleuca G. Don, have also been demonstrated to reduce the Striga seedbank (Gacheru and Rao, 2001).

IR-maize is resistant to imazapyr herbicide, which is used as a coating around the seeds. After absorption by the crop roots, the herbicide is exuded and kills attaching or attached Striga seedlings as well as its nearby non-germinated Striga seeds in the soil (Kanampiu et al., 2002). It has been demonstrated that seed dressings of IR-maize with small amounts of imazapyr can provide season long control of Striga while allowing for intercropping with legumes (Kanampiu et al., 2002, Kanampiu et al., 2003). Development of IR-maize germplasm used in this study and its commercialization in Kenya is outlined in Kanampiu et al. (2003).

Declining soil fertility is another major limitation to crop production in the target area (Vanlauwe et al., 2006). In the Lake Victoria Basin, nitrogen (N) and phosphorus (P) have been identified as the main limiting nutrients (Vanlauwe et al., 2006). Whereas biological N fixation in the push–pull or rotational systems can contribute to reducing soil nitrogen depletion, replenishment of the available soil P pool mainly happens through application of P fertilizer.

Clearly, while some of the above-mentioned technologies were developed with specific constraint(s) in mind, they all potentially alleviate other maize production constraints. Furthermore, some of the above technologies can be integrated, for instance, through inclusion of IR-maize in the push–pull or the rotational systems. Moreover, single technologies are often promoted by the research institute that developed these, thereby preventing farmer communities from evaluating a range of alternative technologies. The objective of this study was to evaluate a set of promising integrated technologies for controlling Striga, stemborers, and declining soil fertility in the Lake Victoria basin. This multi-institutional evaluation was implemented through multi-locational and multi-seasonal, on-farm trials encompassing all above technology components. The main hypothesis was that integrated technologies exist that can simultaneously alleviate various constraints to maize production in the target area.