Mycotoxin concentrations in rice from three climatic locations in Africa as affected by grain quality, production site, and storage duration

Abstract Information on the mycotoxin contamination of rice in Africa is limited although the risk of contamination is high. In this study, domestic milled rice processed by actors using suboptimal methods was purchased and total fumonisin (FUM), zearalenone, and aflatoxin concentrations determined at 0, 90, and 180 days after storage. Three different climatic locations, Cotonou (Benin) in the Guinea savanna, Yaoundé (Cameroon) in the Tropical forest, and N'diaye (Senegal) in the Sahel, were selected as storage sites. Subsets of the samples collected from Glazoue (Benin), Ndop (Cameroon), and Dagana (Senegal) were stored in plastic woven bags under room conditions in the respective sites with or without calcium oxide (burnt scallop shell—BSS, 0.1% w/w) treatment. Multivariance analysis showed that FUM concentration was positively influenced by the duration of storage only while zearalenone concentration was negatively influenced by relative humidity and head rice but positively by impurities. Zearalenone concentration was also influenced by sample collection/storage location, processing type, and duration of storage. Aflatoxin concentration was influenced negatively by storage room temperature and head rice but positively by impurities and chalky grains. In addition, aflatoxin concentration was influenced by collection/storage location and processing type. BSS treatment followed by storage for 6 months had no effect on the concentration of the three assessed mycotoxins. Strategies to reduce the risk of mycotoxin contamination in study sites will include the improvement of physical rice quality through better pre‐ and postharvest practices and proper packaging of both treated rice and untreated rice in hermetic systems before marketing and storage.


| INTRODUC TI ON
The production and processing of rice, which is considered a strategic staple in sub-Sahara Africa (SSA), mostly employ suboptimal methods often resulting to insufficiencies in quantity and quality (Mapiemfu et al., 2017;Ndindeng et al., 2015) of the end products.
Despite efforts toward improving rice production in sub-Saharan Africa (SSA) because of its strategic importance in diets of millions of people, controlling its fungal contamination remains challenging especially as most pre-and postharvest operations are rudimentary and manual. In addition, rice is mostly sold unpackaged or stored in plastic woven or jute bags, thus increasing the rate of fungal contamination and proliferation. Rice is also by itself a suitable culture medium for mycotoxigenic fungi especially when poorly stored as total AFLA, ZEA, ochratoxin A, deoxynivalenol, and citreoviridin are produced and accumulated (Almeida et al., 2012). In a recent review on worldwide occurrence of mycotoxins, the highest level of AFLAs was detected in polished rice from Africa (1,642 ppb) with an incidence of 50% compared to 850 ppb in corn from Asia with an incidence of 63% (Lee & Ryu, 2017).
In Sri Lanka, Bandara, Vithanege, and Bean (1991) detected levels of 185 and 963 ppb of AFLA B1 and AFLA G1 in parboiled rice. Sangare-Tigori et al. (2006) and Makun et al. (2011) reported the occurrence of AFLAs, FUMs, ochratoxin A, and deoxynivalenol in white rice from SSA. In SSA, rice is mostly consumed in the white and parboiled milled forms but there is no comparative report on their mycotoxin contents. It is however suspected that rudimentary parboiling conditions may favor mycofloral contamination and mycotoxin accumulation. Rudimentary artisanal parboiling, involving prolonged soaking (24-48 hr.) in water at atmospheric conditions and insufficient sun drying, produces inferior quality parboiled rice with moisture content greater than the recommended 12%-14% safe level for storage . Processed paddy samples from these technologies are therefore susceptible to toxigenic fungal colonization and mycotoxin production. When milled white and parboiled rice are poorly stored (inappropriate temperature and relative humidity conditions) for long durations, toxigenic fungal infestation, proliferation, and mycotoxin production are favored (Choi et al., 2015).
The use of resistant varieties, best practices in pre-and postharvest handling, physical, and chemical treatment have been reported as strategies to reduce postharvest losses by pest in stored grains and mycotoxin reduction to below safe levels (Sheahan & Barrett, 2017) especially in wheat (Cheli, Pinotti, Rossi, & Dell'Orto, 2013) and rice (Choi et al., 2015). Limited information exists on how physical grain qualities and processing type (parboiled or white milled) affect mycotoxin accumulation at different climatic locations in SSA.
This study sought to investigate the effects of sample collection/ storage location, relative humidity, and temperature in the storage room, physical grain qualities, processing type (white vs. parboiled), calcium oxide treatment (moisture absorber), and storage duration on the concentration of total FUM, ZEA, and AFLA in rice. As a contribution toward mycotoxin control, burnt scallop shell (BSS) powder at 0.1% was tested for their efficacy on mycotoxigenic fungal growth and mycotoxin accumulation in rice.
A 60 kg pooled sample of each rice type with initial moisture content of 14% was collected from rice mills in each zone, properly mixed, and stored in two replicates of 15 kg each as treated and untreated samples from April to September 2014. For treated samples, the rice was thoroughly mixed with 0.1% calcium oxide (BSS powder) as recommended by the manufacturer (Toyo SC Trading Co., Ltd, Tokyo, Japan). This product was expected to act as a desiccant to reduce the equilibrium moisture content of the samples. BSS has been approved in Japan as a food additive and safe for human consumption.

| Physical grain quality characterization of rice samples
The

| Isolation and identification of fungi
For fungal isolation, samples (200 g) were randomly collected from several locations in each bag from top to bottom using a grain collector after 0, 3, and 6 months. For the identification of mycoflora from rice grains, 10 g of 10 randomly selected subsamples of rice grains from each type was surface-disinfected using 10% sodium hypochlorite solution for 2 min, rinsed thrice with sterile distilled water, and then dried under a laminar flow hood on sterilized Whataman paper. For each sample, 15 whole grains were plated in triplicates (45 grains) on 90 mm in diameter Petri dishes containing Nutrient Broth Yeast Extract Agar or NBY (8 g nutrient broth, 2 g yeast extract, 0.5 g KH 2 PO 4 , 2 g K 2 HPO 4 , 2 g glucose, 15 g agar powder, and 1,000 ml distilled sterile water).
After incubation for seven days at 25°C (12-hr fluorescent light and 12-hr darkness), the resulting fungal colonies were individually subcultured onto potato dextrose agar and diagnosed by mounting conidia and mycelium on glass slides in water, examined under a light microscope, and identified using the keys of Barnett and Hunter (1972

| Isolation and identification of bacteria
Enumeration of bacterial strains was as described by Dossou and Silue (2018). It included subculturing and purification of single colonies on Peptone Sucrose Agar (PSA) medium (10 g peptone, 10 g sucrose, 1 g glutamic acid, 15 g Agar powder, and 1,000 ml sterile distilled water), DNA extraction for Multiplex PCR to diag-

| RE SULTS AND D ISCUSS I ON S
3.1 | Temperature, relative humidity in the storage room, and equilibrium moisture content  In Cotonou, the mean temperature, relative humidity, and estimated equilibrium moisture content of the grains during the storage period were 79%, 26°C, and 14.7%, respectively. These values indicate that the storage of grains in woven plastic bags or other systems that allowed moisture reabsorption was unsafe. In N'diaye, the mean temperature, relative humidity, and estimated equilibrium moisture content of the grain during the storage period were 62.8%, 27.8°C, and 12.4%, respectively, indicating that the storage of grains in that location during the study period in woven plastic bags or other systems that allow moisture reabsorption was safe. In Yaoundé, the mean temperature, relative humidity, and estimated equilibrium moisture content of the grain during the storage period (1) FUM predicted conecntration (ppm) = exp (4.66−3.24 * Absorbance) (2) ZEA predicted concentration (ppm) = exp (9.17−6.26 * Absorbance)

| Physical quality of rice and microbial load in study sites
Physical grain qualities were evaluated in white and parboiled rice with the purpose of relating these qualities to mycotoxin levels at each site (Figure 2). The proportion of intact grains was higher in both parboiled and white milled rice, even though samples from Ndop-Cameroon and Dagana-Senegal recorded lower head rice ratios (46.1% and 60.8%, respectively) compared to those of Glazoue-Benin (85.5%). Head rice is predominantly influenced by parboiling, drying, and milling (Buggenhout, Brijs, Celus, & Delcour, 2013).
(2014) observed that while mold and yeast population increased at temperature and relative humidity of 21°C and 85%, respectively; there was no increase in bacteria regardless of temperature and relative humidity. When paddy contaminated with heat resistant bacterial spore is soaked in water at initial temperature of 70-85°C as commonly practiced in artisanal parboiling systems, the bacterial populations are bound to increase. Bacterial spores resistant to heat during steaming will survive and be reactivated during drying and prolonged storage at room temperature (Ankolekar et al., 2009). This explains why bacterial loads were comparable in parboiled and white rice in the three sites. Based on this work and other reports (Magan & Aldred, 2007), it is suggested that farmers and processors observe proper hygiene during harvesting, threshing, drying, parboiling, milling, and storage and to regularly disinfect equipment to limit pathogenic bacteria in processed rice. identified on rice and rice by-products from field, stored, and marketed samples in different regions around the world (Makun et al., 2007;Sales & Yoshizawa, 2005;Reddy et al., 2014). Penicillium spp., another important storage fungus, was detected at low levels in only white rice from Benin and Senegal compared to the high incidence reported by Makun et al. (2007). Neurospora spp. detected at high levels in white rice from Benin is reported to be common on starchyrich foods and in subtropical and tropical regions (Davis & Perkins, 2002;Turner, Perkins, & Fairfield, 2001).

| Modeling the effect of temperature, relative humidity, grain qualities, processing type, BSS treatment, and storage duration on FUM, ZEA, and AFLA concentrations in different sites
The concentrations of FUM, ZEA, and AFLA in relation to collection/ storage location, rice processing type, BSS treatment, and storage duration are presented as Supporting Information S1. The effect of temperature, relative humidity, physical grain quality, sample collection/storage location, processing type and BSS treatment, and duration of storage in plastic woven bags on FUM, ZEA, and AFLA accumulation is shown in Fumonisin concentration was not affected by room storage temperature and relative humidity, physical grain quality (head rice yield, impurities, and chalkiness), processing type (data not shown), collection/storage location, and BSS treatment. However, storage duration had a positive effect on FUM concentration with samples stored at 0 and 3 months recording 0.50 and 0.37 ppm lower concentrations respectively compared to those stored for 6 months. This suggests that there is a high risk of FUM to accumulate to significant levels when rice is stored for 6 months and above in plastic woven bags in the studied sites. The range in FUM concentrations from 0.13 to 1.48 ppm across the three study sites was higher than the 0.4-4.4 and 132.5 ppb respectively quantified for FUM B1 and B2 by Makun et al. (2011). Although Abbas et al. (1998) did not detect FUMs in polished rice from cultivars in the Texas and Arkansas commercial rice fields in the USA, they quantified levels of 14.5, 1.2, 3.4, and 3.5 ppm in hulls, brown rice, bran, and unpolished rice respectively.
The method used for quantifying these toxins was the competitive direct ELISA (CD-ELISA). In the same study using HPLC, higher accumulations of FUMs were reported. Sangare-Tigori et al. (2006) and Ghali, Ghorbel, and Hedilli (2009) did not detect FUMs in ten and in eleven rice samples respectively from Ivorian and Tunisian markets. and ZEA (by F. graminearum), respectively (Murphy et al., 2006). Sorenson, Hesseltine, and Shotwell (1967) had reported an optimum temperature of 28°C for AFLA B1 and G1 production on rice by Aspergillus flavus and a comparable production of AFLA B1 at 32°C but not G1. Properly dried grains inside woven plastic bags under the temperature and relative humidity conditions reported in this study slowly equilibrated with the temperature and humidity conditions of the macroenvironment over the storage period. This observation supports previous findings by Williams, Baributsa, and Woloshuk (2001) that showed that at low relative humidity (<29%), grains stored in nylon woven bags but not in Purdue Improved Crop Storage (PICS) bags showed a decrease in moisture content.
The studied physical grain qualities (head rice, impurities, and chalkiness) had no effect on FUM concentration whereas ZEA and AFLA concentrations were negatively influenced by head rice but positively by impurities ( fungi invasion than non chalky rice with a firm or hard endosperm as shown in this study. Oliveira et al. (2009)

through correlations
between Fusarium colonization and mechanical properties reached the conclusion that soft endosperm maize landraces from Brazil were highly susceptible to contamination. Expanding knowledge on the differences in fungal colonization of chalky and non chalky rice grains will be of great application in the rice value chain.
Although it was expected that parboiling will result in the control of fungi and mycotoxin in milled parboiled rice, Kaushik (2013) indicated that parboiling seemed not to be a favorable method for AFLA control. This view is further supported by Dors, Pinto, and Badiale-Furlong (2009) who reported on the migration of mycotoxins from the outer layers into the endosperm during parboiling. In this study, parboiled samples recorded higher AFLA-confirming previous studies but lower ZEA concentrations compared to white milled rice suggesting a possible degradation of ZEA during parboiling.
In SSA, domestic milled rice (white and parboiled) is mostly of low quality and characterized by low head rice, high proportions of impurities, and chalky grains. In addition, both suboptimal pre-and postharvest practices are commonly used (Amponsah et al., 2017;Mapiemfu et al., 2017;Ndindeng et al., 2015). This rice is mostly sold in bulk or stored in plastic woven bags, jute bags, or bulk storage systems. In this study, the storage of grains in such systems was unsafe especially under Guinea savanna and Tropical forest climatic condition representing respectively subhumid and humid agroecological zones (HarvestChoice, 2009). Strategies to reduce the risk of mycotoxin contamination in SSA will include improvement of the physical rice quality (head rice ratio, proportion of impurities, and chalky grain). This can be achieved through the use of improved pre-and postharvest practices and proper packaging of rice in hermetic systems before marketing and storage at higher than ambient temperatures. Although relative humidity and temperature did not affect ZEA and FUM accumulation in this study, Choi et al. (2015) reported temperature and relative humidity of 21°C and 97% respectively, as suitable for Fusarium proliferation and mycotoxin production.
Burnt scallop shell powder neither affected bacterial load nor mycotoxin concentration in this study. However, bactericidal action of heated scallop shell powder (CaO) against Psuedomonas aeruginosa biofilms on eggshell has been reported at 0.05%-0.3% concentration (Jung et al., 2017). It is suspected that the 0.1% concentration of the BSS powder used in this study was insufficient because the storage system freely allowed the exchange of moisture between the stored samples and the environment. Thus, the need to further investigate the efficacy of various concentrations and dose application rate for the control of mycotoxins in stored grains using different storage systems.

| CON CLUS IONS
The environmental conditions of temperature and relative humidity in the Guinea savanna (subhumid), Tropical forest (humid), and Sahel (semi-arid) zones predisposes rice stored in plastic woven or jute bags to moisture re-absorption (re-wetting) or moisture loss (drying), thus compromising its quality during storage. Therefore, rice should be stored in hermetic systems especially for periods longer than 3 months. Poor physical qualities (low head rice, high proportions of impurities, and chalkiness) of rice produced in SSA enhance microbial colonization and consequently mycotoxin accumulation.
Processing (white and parboiled milled rice production) affected mycotoxin concentration differently as AFLA concentration was higher in parboiled samples while ZEA concentration was higher in white milled samples. Storage duration affected mycotoxin concentration differently as FUM and ZEA concentration increased with duration of storage while AFLA did not. The 0.1% concentration of the BSS powder used did not affect the microbial load and total FUM, ZEA, and AFLA concentrations. The BSS dose used was insufficient because the storage system facilitated the exchange of moisture with the environment. Strategies to reduce the risk of mycotoxin contamination in those sites will encompass the improvement of the physical rice qualities (head rice ratio, impurities, and chalky grain) through better pre-and postharvest practices and proper packaging of both treated and untreated rice in hermetic systems prior to marketing and/or storage.

Agri-Food System (RICE-CRP [CRP 15]) from the CGIAR System
Organization. Special thanks to Toyo SC Trading Co., Ltd for providing BSS powder.

CO N FLI C T O F I NTE R E S T
The authors declare that they do not have any conflict of interest.

E TH I C A L R E V I E W
This study does not involve any human or animal testing.