Breeding maize (Zea mays) for Striga resistance: Past, current and prospects in sub‐saharan africa

Abstract Striga hermonthica, causes up to 100% yield loss in maize production in Sub‐Saharan Africa. Developing Striga‐resistant maize cultivars could be a major component of integrated Striga management strategies. This paper presents a comprehensive overview of maize breeding activities related to Striga resistance and its management. Scientific surveys have revealed that conventional breeding strategies have been used more than molecular breeding strategies in maize improvement for Striga resistance. Striga resistance genes are still under study in the International Institute for Tropical Agriculture (IITA) maize breeding programme. There is also a need to discover QTL and molecular markers associated with such genes to improve Striga resistance in maize. Marker Assistance Breeding is expected to increase maize breeding efficiency with complex traits such as resistance towards Striga because of the complex nature of the host‐parasite relationship and its intersection with other environmental factors. Conventional alongside molecular tools and technical controls are promising methods to effectively assess Striga in Sub‐Saharan Africa.


| INTRODUC TI ON
Maize is one of the most important cereal crops grown worldwide.
In Sub-Saharan Africa (SSA), it is regarded as the most important staple crop with huge potential for addressing the challenge of food insecurity (Abdoulaye et al., 2018). However, its productivity remains relatively low across SSA countries when comparing to the global average production (FAO, 2018). Amongst the major constraints that affect maize productivity, drought, low fertility and the parasitic weeds known as Striga hermonthica, have been recognized by farmers as the most widespread stresses (Atera et al., 2013;Edmeades, 2013;Das et al., 2019).
Striga, is a parasitic weed belonging to the Orobanchaceae family.
It infests and reduces yields of many cereal crops including maize by up to 100% (Atera et al., 2013;Chemisquy et al., 2010;Parker, 2012;Teka, 2014). Across the globe, more than 50 species belonging to the Orobanchaceae family are identified and known as crop pests. In SSA, S. hermonthica (Del.) Benth. and S. asiatica (L.) Kuntze are the most economically important species affecting maize production Teka, 2014). According to Parker (2012), the tropical semi-arid climatic conditions have allowed rapid development of the Striga and even its adaptation to context. Unfortunately, S. hermonthica infestation appears to be worsening due to the current intensive land use, mono-cropping practices and human demographic pressure. All these factors lead to a continuous decline in soil fertility, which greatly favours the Striga occurrence (Rich & Ejeta, 2008). In West Africa, Striga is widely found across the region where maize yield losses due to infestation can vary from 20% to 80% ( Ejeta, 2007;Kim et al., 2002).
In the last few decades, efforts have been made to develop methods for Striga control, including agronomic cultural practice, biological control, chemical, host plant resistance and genetically modified crops. However, these strategies are only moderately effective, because Striga are still expanding its natural range by causing more yield losses. From existing strategies, the most effective and sustainable control seems to be an integrated approach that uses resistant cultivars (Chitagu et al., 2014;Hearne, 2009;Yoder & Scholes, 2010). adding each year more Striga seeds into the soil after each growing season. Therefore, additional genes or sources of Striga resistance need to be found for introgression into maize elite varieties in order to develop varieties that support little or no Striga emergence. This review intends to give a brief update on current work towards Striga resistance emphasizing breeding methods for Striga resistance in Africa and the use of integrated Striga control mechanisms on maize.

| Economic impact of Striga infestation on maize production
Striga parasitism is a limiting factor to maize (Zea mays L.) cropping in the savannah zones of Sub-Saharan Africa (SSA) which constitutes the maize belt of the sub-region (Runo & Kuria, 2018). About 75% of cultivated land with maize in SSA is endemic to S. hermonthica (Akaogu et al., 2019). Maize yield losses under severe Striga infestation can be as high as 100% ( Figure 1) and are economically estimated to $7 billion in the SSA alone (Spallek et al., 2013). The Striga problem has been worsened by the increasing mono-cropping practice instead of rotation and intercropping systems, human demographic pressure on available land where up to 300 million farmers were exposed to the Striga infestation in SSA (Badu-Apraku & Fakorede, 2017). Challenges in managing Striga infestation lead to agricultural land abandonment in several West African countries including Benin, Burkina Faso, Niger, Nigeria and Togo (Atera & Itoh, 2011;Badu-Apraku, 2010;Badu-Apraku et al., 2014). Consequently, this has threatened food security and livelihoods of millions farmers in most countries in this region (Menkir et al., 2020).

| Biology and Striga spp. life cycle
Life cycle of Striga is synchronized to that of its host and involves mechanisms that coordinate lifecycles of both the parasite and the host (Bouwmeester et al., 2003). Striga life cycle generally involves: germination, host attachment, formation of haustoria, penetration and establishment of vascular connections, nutrients accumulation, flowering and seed production (Parker & Riches, 1993) (Figure 2). Germination of Striga seeds F I G U R E 1 Maize field devastated by S. hermonthica in the North of Benin Republic Source: Yacoubou (2018) depends on the presence of hormones known as strigolactones that are produced by the host and in other cases non-host species (Spallek et al., 2013). With the presence of strigolactones, parasite seedlings attach to the host and form vascular connections depriving it of its water, carbohydrates and minerals . Under stressful conditions plant roots exude strigolactone hormone to promote symbiotic relationship with soil microbes for mineral nutrient scavenging (Steven, 2014). Parasitic plants such as Striga hermonthica have exploited these strigolactone hormones as signals to stimulate the germination of their seeds (Runo et al., 2012) (Figure 3). During early stages of seed development, before emergence, the parasite depends totally on the host plant (Webb & Smith, 1996). At this stage of subterranean development, S. hermonthica inflicts maximum damage to the maize plant. The adverse effect of Striga on maize is manifested as stunting, chlorotic and necrotic lesions on the leaves and reduction of ear size and grain yield (Adetimirin et al., 2000). Striga spp. take about 4-10 weeks to complete its life cycle after emergence and this completion usually occurs after harvest of the host (Ramaiah et al., 1983).

| Striga control methods
Striga control is essential to ensure food security in the SSA (Ejeta, 2007;Rodenburg et al., 2005). Several methods, ranging from agricultural practices to biological control exist and significant progress has been made in Striga control research within Africa (Table 1).
Cultural practices such as manual weeding, push and pull, crop rotation with non-host intercrops (trap crops), fertilizer application, soil and water management, and transplanting have been attempted, but they offered limited success in controlling Striga infestation (Oswald & Ransom, 2002;Fasil & Verkleij, 2007;Udom et al., 2007;Manyong et al., 2008;Ayongwa et al., 2010;Lagoke & Isah, 2010;Hailu et al., 2018). Inter-cropping cereals with legumes is another low-cost and viable strategy that has been reported to influence Striga spp. infestation (Carsky et al., 2000;Akanvou et al., 2006;Kanampiu et al., 2018). Legumes, through their roots, fix atmospheric nitrogen, add organic matter to the soil by contributing to soil conservation, preserving the streamline soil moisture and enhances soil biodiversity, thereby improving soil health and fertility, which directly contributes to Striga control. Intercropping legumes with cereals reduces S. hermonthica but does not eliminate the parasite (Khan et al., 2000(Khan et al., , 2007. Other methods for Striga control include biological control using herbicide-resistant maize variety (Imazapyr treatment), development of Striga-resistant germplasm, use of fungus Fusarium isolation by applying strigolactones (Kanampiu et al., 2002;Ejeta, 2007;Illa et al., 2010;Nzioki et al., 2016;Uraguchi et al., 2018;Zwanenburg, & Blanco-Ania, 2018;Kountche et al., 2019). All these approaches have been used with some degree of success to minimize the effect of Striga in maize production. The mode of action for each approach is different. For example, in the case of fungus, when F. oxysporum gets in contact with maize plants, there is a production of amino acids (L-leucine and L-tyrosine), that disrupt plant growth and development. These amino acids are toxic to Striga plants but innocuous to maize plants (Nzioki et al., 2016). The use of this biological F I G U R E 2 The life cycle of S. hermonthica on a susceptible host. Stages indicated: A = after-ripening and conditioning of S. hermonthica seed, B = germination of S. hermonthica seed, C = haustorial initiation and attachment of S. hermonthica to the host followed by a period of growth underground, D = emergence of S. hermonthica plants from the soil, E = flowering, insect pollination, seed set and dispersal. Duration of each phase of the life cycle is indicated. Source: Hearne (2001) [Colour figure can be viewed at wileyonlinelibrary.com] control tool allowed the increment of more than 45% maize yield in Striga endemic zones in Kenya (Nzioki et al., 2016). Strigolactones (SLs) reduce the accumulation of abscisic acid (ABA) in plant by up-regulating the ABA catabolic enzyme gene CYP707A1 (Lechat et al., 2015;Toh et al., 2012). The ABA is released by maize infected with S. hermonthica, that subsequently trigger stomatal closure to minimize water loss. SLs also increase the production of gibberellins (GA) hormones by up-regulating gibberellin3β-dioxygenase 1, which is involved in GA biosynthesis (Toh et al., 2015;Yao et al., 2016).
Although ABA and GA represent central plant hormones and are known to antagonistically regulate seed germination in non-parasitic plants, the effects of their exogenous application vary across parasitic plant species. Zehhar et al. (2002) and Toh et al. (2015), reported that neither GA nor ABA alone is sufficient to stimulate or inhibit seed germination in S. hermonthica, while Kannan and Zwanenburg (2014) and Zwanenburg et al. (2016) reported SLs application appears attractive owing to their decomposition in the soil within a short period. Nevertheless, the use of natural SLs for decomposition in soil does not seem a realistic alternative because the synthesis of these compounds is very labourious. More recently, genetic engineering has offered the promise of rapidly achieving resistance against Striga spp. Recent findings have shown that RNAs freely translocate between parasitic plants and their hosts (Kim & Westwood, 2015). This translocation suggests a possibility that RNA-interference (RNAi) could be used as a potential tool to interfere in vital processes within the parasite by transforming the host with an RNAi construct that targets gene sequences specific to the parasite (Shayanowako et al., 2017). This technique is constrained by the lack of genes to target for silencing as well as by the delivery of iRNAs into the parasite (Kirigia et al., 2014). This constrain can be overcome using viral induced gene silencing (VIGS). Using a Tobacco Rattle Virus (TRV) -VIGS system, Kirigia et al. (2014) have shown that this system works in S. hermonthica and has been proven as a useful system for candidate gene validation either in parasite development or parasitism, for the development of resistant transgenic maize. stimulants are insensitive to the strigolactone levels produced by the host (Lumba et al., 2017;Mutinda, 2018). Binding causes the degradation of an F-box protein, which in turn activates gene regulatory processes that lead to Striga germination (Lumba et al., 2017).

| Resistance mechanism to Striga in maize
It can also be due to the production of low haustorial initiation factors whose effect leads to a failure by Striga to develop haustorium effectively (Rich et al., 2004). Crop genotypes with preattachment resistance mechanism produce relatively low SLs, thereby inducing the germination of less parasitic seeds and consequently prevent the host plant from parasitism. Preattachment resistance has been shown in 'KSTP'94', an open-pollinated maize variety used by farmers in Eastern Africa for S. hermonthica management. This maize variety was shown to produce low amounts of sorgomol, a strigolactone that does not efficiently induce S. hermonthica germination (Karaya

Factors in favour of control options Setbacks for control options References
Manual  (Jamil et al., 2011;Robert, 2011). However, resistance associated with low production of Striga seeds germination stimulant may not be related to low production of total strigolactones, but rather to the types of strigolactones released (Yoneyama et al., 2010).
In contrast, postattachment mechanisms act after Striga has attached and attempted to penetrate the host (Figure 4b). These mechanisms result in physiological or biochemical barriers, that and consequently easy to multiply and readily available (Midega et al., 2016).
Although hybrids are known and desirable for their high productivity and quality, they have shown reduced pathogen resistance compared to the OPVs which have innate defence traits (Schroeder et al., 2013). It is, therefore, vital to understand the genetic make-up of the parents used to develop hybrids as this would be more useful for further development of improved maize germplasm with enhanced resistance to S. hermonthica.

| Potential sources of Striga resistance in maize
Genetic improvement for Striga resistance depends on the availability of germplasm sources with different levels of resistance.
Therefore, resistance is prioritized in maize breeding programmes F I G U R E 4 Mechanisms of resistance to S. hermonthica in maize Source: Amusan et al. (2008) for regions where Striga is endemic and causes major yield losses to farmers. The sources of resistance to Striga have been identified in maize and other crops such as rice, sorghum and cowpea (Amusan et al., 2008;Haussmann et al., 2004;Mbuvi et al., 2017;Menkir, 2006;Yonli et al., 2006) (Table 2).

| Genetics resistance to Striga
Information on the genetic basis of resistance to Striga is critical for plant breeding and selection. Genes action for grain yield and other agronomic traits have been reported for maize under Striga infestation (Ejeta et al., 1997 (Kim, 1994;Berner et al., 995;Akanvou et al., 1997). As reported by Kim (1994) and Berner et al. (1995) (Sun et al., 2008). Evidence for strigolactones and strigolactone perception genes of the MAX-2-type in S. hermonthica, namely ShCCD7 and ShCCD8 has been provided (Liu et al., 2014). In tobacco, the silencing of CCD7 and CCD8 genes has delayed the virus parasite formation in the host, indicating that these two genes are a key in the parasitic life cycle (Aly et al., 2014).
Recently, some significant loci on chromosomes 9 and 10 of maize that are closely linked to ZmCCD1 and amt5 genes, respectively, and may be related to plant defence mechanisms against Striga parasitism have been identified (Adewale et al., 2020).
Availability of all this information on the type of gene action governing the inheritance of resistance to Striga in maize genotypes would, therefore, contribute to the introgression of resistance genes and dissemination of resistant genotypes (Akanvou & Doku, 1998).

| ME THODS FOR SCREENING Striga RE S IS TAN CE IN MAIZE
Development of Striga-resistant cultivars has been limited by the lack of dependable screening techniques (Yagoub et al., 2014).
Some of the screening techniques that have been used include field techniques, screen house and laboratory methods . Screen house technique has been used to screen maize genotypes for tolerance / resistance to Striga (Chitagu et al., 2014;Nyakurwa et al., 2018;Yohannes et al., 2016). In screen houses, screening for varietal resistance has been performed using pots and buried seed studies (Eplee & Norris, 1987;Rao, 1985;Sand et al., 1990). With regard to the pot screening techniques 'poly bag' and seed pan, and the 'Eplee bag' are used (Eplee, 1992;Rao, 1985).
The most important aspect in screen house evaluation is its compatibility with experiments on the efficiency in controlling the Striga vector (Kountche et al., 2019). Several studies have also demonstrated the validity of the Eplee bag technique as a good screening method (Ahonsi et al., 2002;Yonli et al., 2006). Previously, pot experiments were used to access the level of parasite variation in the attachment to the roots of diverse maize inbred lines alongside the plant host interaction . AGA is useful for screening maize genotypes with a high degree of success in identifying Striga-resistant varieties especially those emanating from the wild-species relatives such as Z. diploperennis and T.
Furthermore, the rhizotron screening system has been proposed as an ideal technique to circumvent the limits of field technique and initiate a reliable postattachment screening .

Rhizotrons are transparent root observation chambers that enable
Striga attached to the host plant to be counted. The AGA technique also allows the evaluation of resistance mechanisms phenotype and determination of the effect of Striga on host biomass over a period of time with minimal disturbance Runo et al., 2012). Rhizotron Perspex chambers have been extensively used to screen a variety of host species including maize (Mutinda et al., 2018).

| B REED ING APPROACHE S US ED FOR Striga RE S IS TAN CE IN MAIZE
Considerable efforts have been made in breeding for Striga resistance in cereals especially in maize and significant progress has been achieved in the development of improved varieties. After the identification of a potential source of resistance, the next critical step in the breeding programme depends on the breeder's ability to incorporate the resistance genes into the best-adapted varieties. This can be performed with several strategies, amongst which are the conventional and or classical breeding and the marker-assisted selection (MAS).

| Conventional breeding for Striga resistance
Conventional plant breeding aims at increasing the chances of selecting individuals from populations generated from genetic mating designs. Selection has usually been carried out at the whole-plant level thereby, representing the net result of the interaction between genotype and environment . However, identification of potential sources of resistance is the first step of all Striga breeding programmes. To access the genes for resistance and incorporate them into well-adapted varieties, conventional breeding relies on techniques such as recurrent selection, half-sib or fullsib selection, S1 family and F1 family (hybrid) selection schemes.
Conventional breeding techniques were predominantly used in conferring superior combinations of Striga resistance alleles amongst susceptible cultivars (Menkir et al., 2004). It is, therefore, relevant to explore the applicability of many conventional breeding techniques generally used in various Striga resistance-breeding programmes.
Recurrent selection is designed to increase the frequency of favourable alleles in a population (Hallauer, 1992;Hallauer & Carena, 2012;Badu-Apraku & Fakorode, 2017). This procedure has been used effectively in maize to improve quantitatively inherited traits (Badu-Apraku, 2010; Menkir & Kling, 2007). Few studies have been conducted on the effectiveness of recurrent selection in improving the level of Striga resistance in maize (Menkir & Kling, 2007).  (Menkir & Meseka, 2019). The authors observed that on average, hybrids developed after the 1990s yielded 64% more and displayed 61% less parasite emergence and 30% less parasite damage at 10 weeks after planting compared with hybrids developed before the 1990s.
The half-sibling selection scheme is also one of the easiest ways in developing composite populations with at least moderate resistance to S. hermonthica (John & Sleeper, 1995). The full sib and selection from S1 progeny tests allows for an increased scope of variability in progeny from source populations and greater control over pollen, and should translate into an increased frequency of favourable alleles for Striga resistance in populations under selection (Hallauer, 1992;Menkir et al., 2004).
The backcross breeding procedure is straight forward if a source population or donor, with a high frequency of desirable alleles for Striga resistance is available. Therefore, rapid progress can be achieved in building resistance to Striga if a donor exhibiting high dominance for Striga resistance genes is identified. Under such condition, ideal recurrent parents would be genotypes combining early maturity and high yield .

Marker-assisted selection (MAS) is an indirect selection process
where a trait of interest is selected based on a marker linked to the trait, rather than on the trait itself (Ribaut et al., 2001). This breeding method allows the performance of a selected phenotype to be predicted based on the use of molecular markers at early generation.
Application of molecular markers has provided significant opportunities for breeders to characterize, evaluate and select maize germplasm widely used by public and private sectors. Molecular markers are also used for screening crop genotypes for tolerance to biotic or abiotic stress. Using SSRs and SNPs markers, some elite genotypes for the breeding of Striga resistance are selected and new makers have been identified, which significantly contributed to the differentiation of Striga tolerant and susceptible genotypes (Bawa et al., 2015;Shayanowako et al., 2018). Molecular markers can better help in the assessment of relatedness in isogenic lines to determine families that can be bulked or discarded, which in turn can reduce maintenance costs (Dean et al., 1999).
Several researchers have reported the efficiency and superiority of MAS and its effective integration into mainstream maize breeding programmes. Efforts deployed with the use of molecular tools can be utilized in determining families that can be bulked or discarded. Those families could also help in the selection of parental lines for Striga-resistant hybrids development with high yields and stable across many agroecologies (Akinwale et al., 2014;Mengesha et al., 2017).
Molecular marker technologies and the construction of genetic linkage maps have made it possible to detect genetic loci associated with complex traits (Kang et al., 1998;Sibov et al., 2003). Genetic linkage maps and quantitative trait loci (QTL) mapping technology have enhanced the efficiency of estimating the number of loci controlling genetic variation in a segregating population and the characterization of the map positions in the genome (Xiao et al., 1996). In maize, QTLs identification was focused mainly on abiotic and biotic stresses such as drought tolerance (Semagn et al., 2015;Tuberosa et al., 2002), low soil nitrogen (Mandolino et al., 2018;Ribeiro et al., 2018), pests (Jiménez-Galindo et al., 2017) and foliar diseases (Gowda et al., 2018). In SSA, little progress has been reported on the detection of QTLs or genes for Striga resistance in maize.
However, QTLs for resistance to S. hermonthica have been identified from local populations including wild relatives and successfully transferred through backcross breeding into adaptable maize populations (Rich & Ejeta, 2008

CO N FLI C T O F I NTE R E S T S
The authors declare that there is no conflict of interests.