Assessment of the physicochemical characteristics, chemical and microbiological safety of two types of kilichi, a grilled meat produced in Niger

Abstract Production of kilichi, a grilled meat of West Africa, is a common method of meat preservation in Niger. Thirty samples of condiments‐coated kilichi and uncoated kilichi collected in Niger, were analyzed for microbiological contamination, as well as NaCl, protein and lipid contents, using standard methods. Contamination with Polycyclic Aromatic Hydrocarbons (PAHs) was also assessed using a HPLC–FLD technique. Highly significant differences (p < .001) were observed between coated kilichi and uncoated kilichi, for NaCl content (2.56% and 1.40%), for proteins (51% and 72%) and lipids (18% and 13%), respectively. Water activity was low in both kilichi, showing a potential microbial stability. Among the 15 European Union (EU) priority PAHs, 12 were detected in the samples. About 56.3% of coated kilichi samples exceeded the EU maximal limit for BaP, and 75% exceeded the EU maximal limit for the sum of 4 PAHs (PAH4). For uncoated kilichi, 28.6% of samples did not meet the standards for BaP and PAH4. About 6% of coated kilichi samples were not compliant with standards related to Staphylococcus aureus, Bacillus cereus, Clostridium perfringens, and 31%, 50% for yeasts and fungi, respectively. Escherichia coli and Enterobacteria were below the detection limit in both kilichi, but Salmonella and Bacillus cereus were detected only in one coated kilichi. The noncompliant samples of uncoated kilichi were in the proportions varying between 7% –86% for S. aureus, C. perfringens, yeasts, and fungi. This study showed potential risks associated with the consumption of traditionally produced kilichi in Niger due to both PAHs and pathogen bacteria contamination.

and providing substantial revenues for producers and animal proteins for populations.
Among these products, kilichi, a dried meat, coated with condiments or uncoated, then grilled, is well appreciated by consumers in local markets in Niger (Beidari & Mahamadou, 2014). Kilichi is a ready-to-eat meat product traditionally manufactured with bovine, camel, ovine, or goat fresh meat. Coated kilichi is produced by trimming meat, cutting into pieces of parallelepiped shape before slicing into flat thin sheets. The sheets are spread on millet or sorghum panicle mats for a first sun drying followed by marinating in a sauce made of complex blend of spices, before a second sun drying and briefly grilling at wood fire. For uncoated kilichi, the sun dried meat is slightly seasoned and grilled. Kilichi is generally packaged, only just before selling to consumers (Boubacar et al., 2019). The conditions of processing and distribution practices do not guaranty the safety of this product. Analyses conducted on this product in Tchad and Cameroun showed high levels of contamination by pathogenic bacteria, in particular Salmonella, presenting a risk of acute intoxication for consumer (Kimassoum et al., 2017;Mbawala, Daoudou, & Ngassoum, 2010). The problem of pathogenic bacteria seems to be the most worrying, but grilling step of the product shows risk of contamination by Polycyclic Aromatic Hydrocarbons (PAHs). In fact, PAHs are toxic chemical contaminants generated during the combustion of organic material. Food products mostly get chemical contamination through thermal processing such as smoking, roasting, grilling, frying, and drying where they are in direct contact with the combustion source (Farhadian, Jinap, Abas, & Sakar, 2010;Rose et al., 2015;Roseiro, Gomes, Patarata, & Santos, 2012).
At high concentration, PAHs could be carcinogenic and genotoxic. Some respiratory, cardiologic, immunologic, neurologic, reproductive, and genotoxic imperfections in human and animals are linked to the harmful effect of PAHs (EFSA, 2008;Olabemiwo & Ogunsola, 2014). The parameters like temperature, time, distance, biomass, relative humidity, and characteristics of the products have effect on the absorption and penetration of smoke components in product, and therefore, on its quality and stability (Akpambang et al., 2009;Santos, Gomes, & Roseiro, 2011 done on the characterization of kilichi in Niger, particularly on physicochemical, nutritional, and microbiological aspects, and to our knowledge, no study on the contamination by PAHs. Therefore, the aim of this work was to assess the nutritional, chemical, and microbiological characteristics of kilichi produced in Niger, and to estimate the risks associated with the consumption of PAHs contaminated kilichi for the purpose of improving its manufacturing process and quality.

| Sampling
The sampling was carried out in the regions of Agadez, Maradi, Niamey, Tahoua, & Zinder. About 30 samples of uncoated and coated kilichi were randomly bought at 17 manufacturing sites and markets of the regions as distributed in Table 1. The samples were composed of 14 uncoated samples (11 of beef and 3 of camel) and 16 coated samples (13 of beef and 3 of camel). Each collected sample was packaged in a sterile stomacher bag and stored in refrigerator for microbiological analysis and in freezer at −20°C for physicochemical analyses and determination of PAHs.

| Physicochemical analyses
The physicochemical parameters determined were: pH, moisture, water activity, protein, fat, ash, and NaCl contents. The pH was measured as described by Mgbemere, Akpapunam, and Igene (2011).
The moisture of samples was determined using the ISO 1442/1997 standard. The water activity (a w ) was measured according to the method described by Anihouvi, Ayerno, Hounhouigan, and Sakyi-Dawson, (2006)

| Polycyclic aromatic hydrocarbons analyses
The 15 Polycyclic Aromatic Hydrocarbons (PAHs) were determined in the kilichi samples using a High-Performance Liquid Chromatography coupled with a fluorescence detector (HPLC-FLD) according to the method described by Brasseur et al., (2007). A Model 600 E solvent delivery system, equipped with a Model 717 automatic injector, a Mistral TM oven, and 2,475 Fluorescence detector (WATERS Corporation), was used. A C18 Pursuit 3 PAH (100 × 4.6 mm, 3 µm) equipped with a ChromGuard (10 × 3 mm) precolumn, both for Varian (Agilent Technologies) were used to separate the PAHs. The PAHs were extracted from the kilichi samples using the method described by Veyrand et al., (2007). The kilichi samples were frozen in liquid nitrogen and then freeze-dried for 36-48H. For the extraction, one g of lyophilized kilichi, was homogenized with a mixture of hexane/acetone (50/50, v/v) by using the Accelerated Solvent Extraction system (ASE 200; Dionex). Then the solvent was evaporated until 1 ml, and then, reconstituted with 5 ml of cyclohexane. The reconstituted extract was then purified on a Chromatographic Column (Envi Chrom P Supelco), previously conditioned with 15 ml of ethyl acetate and 10 ml of cyclohexane. The extract was poured on the Column which was then washed two times with 3 ml of cyclohexane/ethanol (70/30, v/v) mixture. The PAHs were eluted three times using 4 ml of cyclohexane/ethyl acetate (40/60 v/v) mixture. The solvent was evaporated to dryness, then 90 µl of acetonitrile and 10 µl of the deuterated DiP-D14, used as internal standard (LGC Promochem, France), were added. Twenty-five (25) µl of this final extract was injected in the HPLC column. The limit of quantification (LOQ) corresponded to the first point of the calibration curve for each PAH and were 0.96 µg/kg of fresh weight for the Benzo (j) fluoranthene and Indeno [1, 2, 3-cd] pyrene and 0.24 µg/kg fresh weight for the rest of the PAHs.

| Microbiological analyses
Microbiological analyses were performed to determine the spoil-

| Statistical analyses
The software SAS version 8 was used for statistical analysis of data.
The means, medians, and standard deviations were calculated. The means comparisons were done with the test of Student at a threshold of 5% signification. The General Linear Model (GLM) was used for one-way analysis of variance (ANOVA). The Least Square Mean was used to compare the two types of kilichi.

| Physicochemical characteristics of the kilichi
The results obtained from physicochemical analysis of both types of kilichi samples are presented in Table 2. The means of moisture content of kilichi are close to that obtained by Kalilou (1997) and Yacouba (2009), on both types of kilichi produced in Niger; and on coated kilichi produced in Nigeria and Cameroun (Apata, Osidibo, Apata, & Okubanjo, 2013;Jones, Tanya, Mbofung, Fonkem, & Silverside, 2001;Mgbemere et al., 2011;Olusola, Okubanjo, & Omojola, 2012). However, these values are lower than those obtained on Nigerian uncoated kilichi (19%-26% moisture content) (Raji, 2006)) and on dried Bitong (21.5%-25.3%) (Petit, Caro, Petit, Santchurn, & Collignan, 2014). They are also in agreement with the standard (at least 88% of dry matter) stipulated by the National Council of Normalization of Niger (Conseil National de Normalisation du Niger- CNNN, 2004). According to Prescott, Harley, and Klein (2002), the microbial growth would be impossible in food products with water activity values lower than 0.7. smoked Kitoza (0.94). Bitong is a traditional South African readyto-eat meat product made from raw meat by salting, curing, and drying. There are two types, moist beef Bitong with high moisture content (higher than 40%) and dried Bitong. Kitoza is a traditional meat product from Madagascar, manufactured with strips of pork or beef. The process includes salting and mixing with spices followed by sun drying or smoking. Uncoated kilichi presents a mean pH lower than that of coated kilichi. The values found are similar to the results reported by Kalilou (1997), with a pH ranging between 5.7 and 6.2 for both types of kilichi, by Jones et al., (2001) for coated kilichi (5.81), by Petit et al., (2014) for dried Bitong (5.5-6.26) and by Ratsimba et al., (2017) for dry Kitoza and smoked Kitoza (5.67 and 5.87, respectively). These mean pH levels are lower than average value of 6.33 obtained by Eke et al. (2013) on Dambu namma (fried ground beef). The pH is an important parameter to control the sensory quality of meat. The average salt content in coated kilichi is higher than that in uncoated kilichi; this is justified by the addition of salt and bouillon cube in the seasoning sauce used to coat kilichi. The salt should play a role of preservative at short and average term but could be a source of disease. Petit et al., (2014) and Ratsimba et al., (2017)  standard (50 à 70%) and the average coated kilichi protein content is similar to that obtained (49.8%) by Mgbemere et al., (2011).
However, these results are lower than those obtained by Kalilou (1997), which are 64% protein content for coated kilichi and 74% for uncoated kilichi, and lower than those obtained by Jones et al.,  These values on the other hand are higher than 3.70% obtained by Apata et al., (2013). The higher ash content of coated kilichi may be due to spices added.
The comparison of the two types of kilichi show a significant effect (p < .05) on moisture and fat, highly significant effects (p < .01) on pH and water activity, and very highly significant effects (p < .001) on salt, proteins, and ash contents (   There was a significant (p < .05) difference between both types of kilichi as far as DhA, BcF, and BaA concentrations were concerned.

| Contamination of Kilichi samples by PAHs
Also, significant (p < .1) differences of BkF and the sum of PAHs were observed between both types of kilichi (Table 5) smoked Gentile di maile in Italy, Suya, and smoked Shrimp (Manda et al., 2012;Farhadian et al., 2010;Rose et al., 2015;Carrabs et al., 2014;Akpambang et al., 2009 andKpoclou et al., 2013). The average concentration of PAH4 represents 57% of the total sum of PAHs for both types of kilichi. kilichi samples are not in accordance with the standard for BaP as well as for PAH4. Therefore, a consumer has 75% chance to consume a coated kilichi contaminated by the PAH4 and 56% by BaP against 29% for uncoated kilichi. Considering the Bench Mark Limit dose (BMDL 10 ), an adult of 60 kg is exposed to an intoxication risk by BaP if he consumes more than 210 g of uncoated kilichi per day, and for the PAH4, maximal ingestion level is 156 g. For coated kilichi, the daily maximal ingestion for an adult of 60 kg is 162 g and 90g for BaP and PAH4, respectively.
For all germs investigated, coated kilichi present higher loads than uncoated kilichi. This difference could be due to germs carried by seasoning spices. Shamsuddeen (2009), Frazier andWesthoff (2006), have reported that some spices do not have antimicrobial activity, so that the meat treated with spices could have high microbial load. In the same logic, Price and Schweigert (1971) stated that unless the spices are treated to reduce their microbial load, they can be a source of high number of undesirable germs in the product to which they are added. The presence of S. aureus, Enterobacteria, C. perfringens, E. coli, and Salmonella in the products is an indication of poor hygienic practices during processing. Edema, Osho, and Diala (2008) detected in spices used in the preparation of Suya (grilled meat product) the following microbial flora: AMB 3.42-3.50 × 10 5 cfu/g, yeast and fungi 1.27-1.43 × 10 5 cfu/g, Coliforms 0.37-0.47 × 10 5 cfu/g, Staphylococcus 0.37 × 10 5 cfu/g, Bacillus cereus 0.02 × 10 5 cfu/g, Salmonellas 0.37-0.43 × 10 5 cfu/g. These results confirm the high AMB load obtained in coated kilichi.

| CON CLUS ION
This study on the physicochemical characteristics and safety of kilichi produced in five regions of Niger reveals that the microbiological quality of the studied samples is generally not satisfactory.

ACK N OWLED G EM ENTS
The authors gratefully acknowledge the funding support for this research provided by the West African Agricultural Productivity Program (WAAPP/PPAAO Niger).

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

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