Limited access to quality certified seeds has remained the stumbling block to increased agriculture production in sub-Saharan Africa (Barriga & Fiala, 2020). This has happened after governments and development partners spent substantial resources in the last half-century investing in developing the formal seed system to supply high-yielding varieties (HYV) and replace the informal seed system (Croft et al., 2017). The support started with the infrastructure and institutional development of public research that was later liberalised and privatised (Buanec & Heffer, 2002). After privatisation, there have been concerted efforts to develop appropriate policies to regulate the sector and improve regional trade as the private sector's role in breeding, multiplying and distributing seeds in sub-Saharan Africa increased (Louwaars & de Boef, 2013). However, evidence suggests a lack of progress in terms of seed quantities, quality, and crop diversity delivered by the private sector in line with farmers' expectations and needs (Langyintuo et al., 2010). Quality is one of the major factors that has affected the adoption and use of certified seeds and has been associated with challenges affecting the public regulatory system (Maereka, 2020). Studies have reported on the effect of seed quality on the adoption and use of certified seeds (Derwisch, Morone, Tröger, & Kopainsky, 2016; Lunduka, Fisher, & Snapp, 2012; Simtowe, Asfaw, & Abate, 2016; Sperling, Loevinsohn, & Ntabomvura, 1993; Sperling & McGuire, 2010). However, continued direct seed distribution programs by governments, non-governmental organisations and projects have offered a lifeline for the formal seed system (Chinsinga, 2010; Sperling, Cooper, & Remington, 2008). For instance, the Malawi government spent over US$1.4 billion between 2005 and 2015 on subsidised inputs with 4% of that spent on certified seeds where hybrid maize takes the lion's share (World Bank, 2015).
The seed quality challenges have come in different forms from counterfeit seeds, fake seeds, fraudulent labelling, regulatory offences, and sub-standard or of not acclaimed variety and quality (Bold, Kaizzi, Svensson, & Yanagizawa-Drott, 2015; Kilic, Ilukor, Kilian, & Stevenson, 2017).1 Studies using quality tests in Nigeria, Kenya, and Uganda have shown poor germination rates of certified maize seed varieties below the minimum requirement of 95% and seed packages containing half of the genuine seed (Ashour, Gilligan, Hoel, & Karachiwalla, 2018; Bold, Kaizzi, Svensson, & Yanagizawa-Drott, 2017). The counterfeiting has been reported to be done through manufacturing and labelling of incorrect seeds, imitation or adulteration, and packaging of sub-standard or diluted seeds (de Boef, Pierson, & Kim, 2014; Karingu & Ngugi, 2013). There is documented evidence of high levels of farm inputs being substandard than their labelling claims making farmers risk-averse and affecting their adoption and use (Bold et al., 2015; de Boef et al., 2014).
Counterfeiting involves tampering with intellectual property rights including patents, trademarks, copyrights, and designs affecting consumer goods including agricultural inputs (Karingu & Ngugi, 2013). The proliferation of counterfeiting represents a problem with the diffusion of HYV, assessment of on-farm yield potential, and impact on farmers' income as they are a misrepresentation of genuine seed aimed at gouging consumers (Herring & Kandlikar, 2009). Using a learning model from on-farm trials to understand why farmers in SSA have not adopted improved inputs including certified HYV other studies have shown that the farmers are aware of the existence of substandard products (Bold et al., 2017). They found huge variations in the quality of agricultural inputs that are not reflected in the pricing, limiting farmers' ability to infer quality due to a lack of price correlation (Mailath & Samuelson, 2001). Nevertheless, the willingness of some farmers to buy low-quality inputs has increased traders' practice of not stocking high-quality products at a premium price (Bold et al., 2015; World Bank, 2007). Although the farmers are aware of the problem of counterfeits government subsidies, discounts, and stockouts have continued to entice them to buy and test the products (Fairbairn, 2017; Herring & Kandlikar, 2009). However, this affects farmers' expectations due to associated risks and uncertainty (Munshi, 2004).
Maize in Malawi remains the most important crop as measured by cropland area covered and the large portion of both public and private sector investments made (Langyintuo et al., 2010; Smale, Byerlee, & Jayne, 2013). Maize covers more than 60% of all harvested area of staple crops in Malawi with an average per capita consumption exceeding 150 kg per annum (Denning, Kabambe, Sanchez, Malik, & Flor, 2009; FAOSTAT, 2015; Hassan, Mekuria, & Mwangi, 2001; Marechera, Muinga, & Irungu, 2016). But production has remained below optimum due to low use of improved inputs, land degradation, and lately climate variability (Barimah, Doso, & Twumasi-Ankrah, 2014; Kugbei & Bishaw, 2002; Louwaars & de Boef, 2013; Omoyo, Wakhungu, & Oteng’i, 2015). Availability, access and utilisation of improved and quality seeds have been identified as one of the most important limiting factors to increased production in order to achieve food security (Sperling & McGuire, 2012; Tripp & Rohrbach, 2001). This happened despite liberalisation and privatisation of the seed sector and the growth of the private seed companies who have only supplied enough certified maize seed for 35% of cropland area in Angola, Ethiopia, Kenya, Malawi, Mozambique, Tanzania, Uganda, Zambia, and Zimbabwe (Langyintuo et al., 2010). Hence, to date, SSA countries don’t have an effective seed delivery system characterised by lengthy development and delivery period of new varieties due to increased regulatory processes affecting the release, commercialisation, and certification (Auriol & Schilizzi, 2015; Marechera et al., 2016). This is despite the hybrid maize seed remaining an important factor in improving food security in Malawi where large investments through government subsidies have been made to improve access, usage, and productivity (Chinsinga, 2010).
Just as there is a need to invest in quality seed production there is equally the need to build trust among users of seed to increase their adoption (Kopainsky, Tröger, Derwisch, & Ulli-Beer, 2012). An important factor that has undermined the building of trust is globalisation which has complicated the enforcement of intellectual property laws in a rapidly evolving market place creating fraud opportunities (Spink, 2011). For developing countries, the certification function has remained largely a government responsibility, facing financial challenges and lacking objectivity (Tripp, 1995). To date, several technologies have been used to improve certification and the identification of seed batches, prevent wrong packaging, and identify counterfeit packages from production to end-users (Pedrini, Merritt, Stevens, & Dixon, 2017). For example, seed coating and tracers were tested on certified tobacco seeds in China and open data source data collection through mobile phones was used for tracking seed yam production in Nigeria (Guan et al., 2013; Ouma et al., 2019). In addition, end-user authentication, e-verification, and tamper-proof packaging were piloted in Ghana and Uganda (Ashour, Billings, Gilligan, & Karachiwalla, 2015; de Boef et al., 2014). The above complement public certification that largely uses quality tests for seed certification based on phenotypic traits (Bold et al., 2017), perception, (Derwisch et al., 2016) and beliefs (Ashour, Billings, Gilligan, Hoel, & Karachiwalla, 2016).
However, DNA fingerprinting was tested in Uganda, Malawi, and Ghana to identify crop varieties and was proven to be definitive and objective (Kilic, Moylan, Ilukor, & Phiri, 2016; Poets, Silverstein, Pardey, Hearne, & Stevenson, 2020). Other studies have used DNA fingerprinting to accurately identify crop varieties in wheat and sweet potatoes in Ethiopia, and cassava in Malawi (Jaleta et al., 2020; Kilic et al., 2016; Kosmowski et al., 2019). Using genetic distance, DNA fingerprinting is able to show genetic purity and variation among individual organisms or species by acting as signs or flag (Collard, Jahufer, Brouwer, & Pang, 2005). Genetic variation using DNA markers is possible at any growth stage of a plant including seed making it a powerful tool for conducting a rapid test of germplasm or species against known reference (Poets et al., 2020). The DNA markers are classified into three according to how they are determined: hybridisation-based; polymerase chain reaction (PCR)-based; and DNA sequence-based (Winter & Kalh, 1995). Polymorphic markers that can determine genetic variation between species, therefore, will be used in the study through simple sequence repeats (SSR), a codominant marker that can differentiate sizes and have several alleles (Nadeem et al., 2017). SSR is technically easy, agile, dependable, and can be used on different populations although it requires time and labour (McCouch et al., 1997). Other studies have used DNA fingerprinting to identify and assess the level of genetic purity of open-pollinated varieties in Malawi, Zambia, and Zimbabwe (Setimela, Warburton, & Erasmus, 2016). The study contributes to the literature by being the first to use DNA fingerprinting to determine the genetic variation of hybrid maize seeds on the market. This is important as current seed quality tests are based on value for cultivation and use (VCU) and standardised tests for distinctiveness, uniformity, and stability (DUS) using phenotypic traits only (Setimela et al., 2016; Tripp, Louwaars, & Eaton, 2007).
Therefore, this study investigated the existence of counterfeit certified seeds on the market in Malawi. First, based on study reports of high counterfeiting levels in the outlet-consumer interface (de Boef et al., 2014; Langyintuo et al., 2010) we used the Seed Trade Association of Malawi registration database to identify licensed and unlicensed agro-dealers in the selected districts. Second, using the mystery shopper approach, seed samples were bought from agro-dealers and reference seeds from supplying companies. Third, quality and genetic tests were done on the seed samples. The following sections detail the materials and methods, the results and discussion, and the conclusion.