Introduction Poultry is the second source of meat after cattle in Benin. The synthesis of work done in [1], indicate that the consumption rate is 22 % for poultry, 58 % for cattle, 13 % for sheep or goat and finally 7 % for pork. The availability of these meat products is provided through two areas: the import of frozen meat sector and the local poultry industry. The statistical study of modern poultry industry, performed in [2], shows that the proportion of import of these products is growing from one year to another. Thus, from 10.000 tons imported and consumed locally in 1996, the amount of imports increased to 80.000 tons in 2003 [1]. This implies an increase in the demand and the rate of consumption of imported poultry meat. The good storage of these frozen products requires refrigeration generally produced by appliances such as freezers, household refrigerators, refrigerated containers, cold rooms, which operate with electrical energy. But as in most of West African countries, Benin does not provide enough electricity. That reality conducts inexorably to energy shortage, voltage drop and selective power cut. This can lead to the breaking of the cold chain and finally to the loss of refrigerated products [3]. To this problem there is added the organization of the distribution chains, where the last link (retailers) does not have either adequate storage devices or cooling systems, so products are directly exposed to air, the influx of flies, dust and sun. Thus, dubious organoleptic quality of meat can be found in the markets. Despite numerous studies carried out in Benin in the field of food, very few have been devoted to quality assurance [4, 5] and specifically the quality of meat products and the imported poultry, although the work in [1, 6] have conducted microbiological traceability of turkey wings. Taking into account the lack of efficient service controls and the new impetus to these consumer products for economic reasons and services, it appears essential to characterize the physicochemical state of frozen products, like the microbiological quality already performed [1]. Thus, for good conservation and better preservation from the degradation of meat products, the evaluation of the physicochemical quality of turkey wings by descending traceability, through parameters such as temperature, total volatile basic nitrogen (TVB-N), moisture and volatiles becomes unavoidable. This work therefore aims to seek solutions to avoid deterioration of fats and proteins, which may cause allergies in consumers [7] and the loss of perishable goods [8]. Finally, another objective is to determine the physicochemical quality of imported meat products such as turkey wings all along the distribution chain in Benin. Study materials The study material consists of turkey wings, i. e. 30 samples of frozen raw meat taken during different marketing stages from wholesalers at the Autonomous Port of Cotonou (APC) in Benin. The latter is the landing place of the imported goods from the northern countries (France and the Netherlands (wholesaler A)) and Asia (wholesaler B) through the semi-wholesalers to the affiliates. Methodology The methodology adopted can be summarized into two phases: an investigation phase and an analysis one. Investigations. Two surveys were conducted in the city of Cotonou and its surroundings to collect the necessary statistical data for the work. The first survey aims to identify the different categories of imported frozen poultry and the most appreciated one in Benin. For the purpose, a list of questions was prepared then sent to retailers and consumers in Cotonou and around. The second survey aimed to trace the distribution channel (wholesalers, semi wholesalers and retailers) of frozen poultry products imported from each chain during the various sales phases. Sampling. Sampling was conducted along the distribution chain of the imported meat products starting from two wholesalers to arrive at their affiliated retailers. The basic principle in this sampling was the respect of the lots. A repetition has been carried out at each level. Thus, 72 hours after the first sampling at the wholesale level, other samples are taken from the concerned wholesaler and his related semi-wholesalers and retailers of type 1 and 2 (D type). Likewise, 72 hours after collection at the semi-wholesalers, we conducted sampling at retailers of type 3 and 4 and then sampled at the semi-wholesalers 72 hours later. The sample at retailer of type 3 was taken once more after. Figure 1 shows the organizational chart of sampling for each of the distribution channels. Table 1 provides the details on the sample code. Figure 1. Organizational chart of sampling in each of the distribution channels Table 1 The Codes of the 30 studied samples Channel A Channel B A = wholesaler channel A B = wholesaler channel B A’ = repetition wholesaler channel A B’ = repetition wholesaler channel B SGA1 =1st semi- wholesaler channel A SGB1=1st semi- wholesaler channel B SGA2 = 2sd semi- wholesaler channel A SGB2=2sd semi-wholesaler channel B SGA1’=repetition for 1st semi-wholesaler channel A SGB’ = repetition of 1st semi-wholesaler channel B SGA2’= repetition for 2sd semi-wholesaler channel A SGB2’ = repetition of 2sd semi-wholesaler channel B DI-A = retailer type I channel A DI-B = retailer type I channel B DI-A’ = repetition for retailer type I channel A DI-B’ = repetition for retailer type I channel B DIII-A1 = retailer type III affiliated to SGA1 DIII-B1 retailer type III affiliated to SGB1 DIII-A1’ = repetition 1st retailer type III affiliated to A channel DIII-B1’ = repetition 1st retailer type III affiliated to B channel DIII-A2 = retailer type III affiliated to SGA2 DIII-B2 = retailer type III affiliated to SGB2 DIII-A2’ = repetition 2sd retailer type III affiliated with channel A DIII-B2’ = repetition 2sd retailer type III affiliated with channel B DII-A = retailer type II affiliated to channel A DII-B = retailer type II affiliated to channel B DIV-A1 = retailer type IV affiliated to SGA1 DIV-B1 = = retailer type IV affiliated to SGB1 DIV-A2 = = retailer type IV affiliated to SGA2 DIV-B2 = = retailer type IV affiliated to SGB2 Note that the sampling was collected according to the standard method NFV04-501:1998. It involves the examination of five packages of the same batch. And after reporting the necessary information on the packaging, sampling of approximately 1 Kg is produced aseptically on every fifth package. The samples were kept in a cold regime and were quickly sent to laboratories for analysis. Rapid mixing was performed under aseptic conditions prior to the completion of three units (Unit 1, Unit 2 and Unit 3) per sample according to the quartering technique. One sample unit thus formed was used for liquid analysis. At the retail level, the samples were taken in the same way. At the market retailers’ level, the primary sample is obtained by mixing several piles. Determining the temperature, the freezer burn and the viscous layers. The temperature measurement was achieved aseptically, in situ, in the middle of samples using a sensitive thermometer that can take temperatures from - 40 to 60° C. Determination of the freezer burn and the viscous layers was performed by a sensory study. Determination of the Total Volatile Basic Nitrogen (TVB-N). The NF V 04-407 standard French method was used. The test sample weighing 10 ± 0.1 g was ground using a Moulinex before being introduced into clean and dry bottles; then 90 ml of perchloridric acid solution (6 g/100 ml) was added. The whole was mixed for 2 minutes at a speed of 9000 rounds/min before being filtered. The filtrate thus obtained is stored in a refrigerator at 2 °C. Then a few drops of phenolphthalein and silicone antifoam has been added to 50 ml of the filtrate. The resulting mixture was basified with 6.5 ml of NaOH solution (20 g/100 ml) and then distilled. The distiller was set to produce 100 ml of distillate in 10 minutes. The distillate was collected in 100 ml of boric acid (3 g/100 ml) contained in a conical flask supplemented with 3 to 5 drops of Tashiro indicator. Then, titration was carried out with a standard hydrochloric acid solution (0.05 mol/l). The pH value corresponding to the end point is around ± 5 0.1. A blank test is carried out using 50 ml of perchloric acid in place of the extract. The expression of TVB-N in mg per 100 g of sample is as follows: . With: V1 = volume of the solution of hydrochloric acid of 0.01 mole for the sample; V0 = volume of the solution of hydrochloric acid of 0.01 mole for the blank test; m = test sample. Determination of fat acidity. This study was performed according to standard NF V 04-403 applied to meat products. A 5g sample has been taken, crushed and then introduced into a clean and dry flask for 1 hour at 100 °C. The sample is then mixed with 50 ml of hexane and a few pumice stones are added. The whole was connected to a heating device and then heated. After 1 hour 30 minutes boiling, the mixture was filtered and the filtrate collected into a dry flask. Finally, using a rotary evaporator, the solvent was separated from the fat. Remaining traces of solvent were removed by baking at 105 ± 2 °C, and the sample is then cooled in a desiccator. The expression of the result is given by the following equation: Where FC is the fat content; m0 the mass of empty flask before filtration; m1 the mass of the flask after evaporation and finally, m is the mass of the test sample. The determination of the acidity of this fat was carried out with the 10th dissolution with 95° ethanol (w/v) previously neutralized with KOH solution (0.1 N). To this whole a few drops of phenolphthalein were added. This procedure was followed by titration with a KOH solution. The potassium hydroxide solution was calibrated with the hydrochloric acid solution 0.1 N in order to know its normality accurately. The acidity index (AI) is expressed in milligrams per gram of fat content according to the relation below: Vhere AI = acid value; V = volume of KOH used; N = normality of KOH in mol/l; M = molar mass of KOH in mg/mol; m0 = mass of the empty flask before filtration; m1 = mass of the ball after evaporation. Determination of moisture and volatile matters. The method to achieve this goal is the parboiling technique according to the NFV.04-461:2001 standard recommended for meat products. For this purpose, the test sample was 5 g. The sample is placed in clean, previously labelled empty carriers, dried at 100 °C and weighed. The tests are performed in triplicate in oven heated at 105 °C for 72 hours. After that, the carriers are removed and placed in a dryer for 30 minutes and successive weighing is made until the weight is constant. Moisture (H) and volatiles maters (HMV) are expressed as a percentage of fresh product. The result is given by the following equation: Where: % H is the percentage of moisture and volatiles; Pe - the weight of the sample; Pf - the weight of the whole (boat and sample) before baking; P0 - the weight of the whole (carriers and sample) after steaming. Statistic analysis. The analysis was performed in triplicate using Excel software. For physicochemical parameters, Simple Statistic Analysis was considered to get means and standard deviations. Results and discussion Data collected during investigations. According to the responses collected from the survey, there is nearly a dozen of frozen meat products from poultry served at the Benin population. Based on the preferences provided by distributors, it appears that turkey rumps are the most popular products, followed in second by turkey fins and gizzards, wings, thighs of poultry and whole chickens (Figure 2). Figure 2. Distribution of consumption of frozen meat products imported and marketed in Benin This study on turkey wings should have covered turkey rumps, but the services of the Directorate of Livestock had reported that these products are banned from sale in Benin by a recent ministerial decree due to potential diseases risks resulting [2]. Furthermore, the distribution network of two major importers (A and B) has been traced (Figure 3). These surveys have helped identify four types of retailers from wholesalers and semi-wholesalers: - retailers of type I: these are operators with a freezing device that are directly supplied by a wholesaler; - retailers of type II: operators who do not have freezing apparatus. There are supplied by a wholesaler; - retailers of type III: these are operators with a freezing device and whose supplier is a semi-wholesaler - retailers of type IV: operators who do not have freezing apparatus and whose supplier is a semi-wholesaler. Figure 3. Distribution network of two major importers A and B Actors of distribution channels have also provided information on the duration of conservation of products (from receipt to the total sales) This time is at least 72 hours and can even last 2 months for wholesalers, but semi-wholesalers stock up generally every 10 days, while the retailers of type I and III top up every week and retailers of types II and IV every day. All of these actors apply the FIFO (First In, First Out) method. The values of the temperatures recorded (Table 2) during the investigations show that the temperature of storage of turkey wings is not uniform either at the distribution chain A or at B. The highest temperature measured is + 21 ° C (at D IV-A2) and the lowest of -18 °C (A, DIII-A1, B). Based on the normative specifications available into [9], the temperature of frozen products must be at least -18 °C, which value is equivalent to a compliance rate of 13.33 % on the A channel and 6.66 % on B, meaning an average compliance rate of 10 % for both channels. This break in the cold chain can cause spoilage [10]. The results show that from one stage to another in the same distribution system, significant variations in temperature occur. Thus, moving from A wholesaler to its semi wholesalers, the variation of the temperature is over 12 °C. The investigation pointed out that at the same operators, there have been wide variations, all of this reflecting that the cold chain is not maintained. Table 2 Temperature measurement of freezer burn and viscous layers during sampling Sample Code Temperature, °C Freezer burn Viscous layers A -18 00 00 A’ -12 01 00 SGA1 -6 01 00 SGA2 -2 03 00 SGA1’ -2,5 05 00 SGA2’ -9 04 00 DI-A -18 04 00 DI-A’ 00 05 00 DIII-A1 00 00 00 DIII-A1’ 00 01 00 DIII-A2 +02 02 00 DIII-A2’ +01 00 00 DII-A +15 00 00 DIV-A1 +15 00 00 DIV-A2 +21 00 01 B -18 03 00 B’ -10 05 00 SGB1 -6 03 00 SGB2 -5 05 00 SGB’ -5 05 00 SGB2’ -9 00 00 DI-B -14 03 00 DI-B’ -1 03 00 DIII-B1 -5 00 00 DIII-B1’ -5 05 00 DIII-B2 -2 00 00 DIII-B2’ -3 01 00 DII-B +09 00 01 DIV-B1 +11 04 00 DIV-B2 +19 02 00 Maximum +21 05 01 Minimum -18 00 00 On Table 2, it is apparent that the presence of burning and viscous layer depends on the freezing temperature variations during storage; these are physical parameters of assessment to appreciate the start of alteration of a perishable product. Contents of total volatile basic nitrogen (TVB-N), moisture and volatile matter in the acidity of fat. All these analysis have led to the results shown in Table 3. Table 3 Contents of TVB-N, moisture (HMV) and acid index of fat (IAMG) Parameters Codes ABVT IAMG HMV Parameters Codes ABVT IAMG HMV A 12,6 0,14 79,88 B 12,56 0,20 79,46 A’ 12,6 0,16 79,13 B’ 12,56 0,20 79,40 SGA1 12,67 0,14 79,44 SGB1 12,58 0,20 79,09 SGA2 12,6 0,14 79,37 SGB2 12,62 0,22 79,02 SGA1’ 13,36 0,17 77,89 SGB’ 12,59 0,20 79,10 SGA2’ 13,01 0,15 69,71 SGB2’ 12,61 0,22 78,62 DI-A 15,00 0,27 79,12 DI-B 12,62 0,27 75,00 DI-A’ 17,84 0,28 71,00 DI-B’ 12,71 0,27 78,07 DIII-A1 15,00 0,23 69,71 DIII-B1 15,03 0,20 78,07 DIII-A1’ 16,00 0,18 79,12 DIII-B1’ 15,01 0,22 63,57 DIII-A2 17,65 0,14 71,00 DIII-B2 14,78 0,21 62,20 DIII-A2’ 18 0,14 69,55 DIII-B2’ 14,79 0,25 64,49 DII-A 19,20 0,18 65,06 DII-B 16,00 0,25 63,15 DIV-A1 13,45 0,41 63,35 DIV-B1 16,08 0,21 63,10 DIV-A2 23,05 0,28 61,63 DIV-B2 16,10 0,27 61,00 Mean values 15,46 0,20 72,99 Mean values 13,90 0,22 72,22 Minimum 12,6 0,14 61,63 Minimum 12,56 0,2 61 Maximum 23,05 0,41 79,88 Maximum 16,1 0,28 79,46 Standard deviation ± 2,74 ± 0,06 ± 5,78 Standard deviation ± 1,36 ± 0,02 ± 7,27 Figure 4 shows a variation of moisture and volatile matters. It appears clearly that the moisture content decreases during products distribution. This confirms that the samples begin thawing that causes slight exudation inducing moisture loss; these observations are in agreements with those reported by [11] on frozen foods. The general trend of humidity is a decreasing function and the highest water losses are observed at the retail level, mainly for those who display their products in the open air. Similarly, with the retailers of type I and III, the repackaging creates a release of water (in vapour form) inducing the important variations shown on Figure 4. Similar observations have been outlined by the literature [12]. A difference in the variations between the two channels is also noticeable, channel A being the one with the biggest oscillations. Figure 4. Change in humidity and volatile matters during the distribution of products Furthermore, Figure 5 shows a variation of the fat Acid Index (IAMG). This index is a factor of deterioration of the fat of animal products according to [7]. The guide value in [9] is 0.25 mg/g fat. By comparing the results obtained with this value, it results a non-compliance rate of 30 %. This non-compliance is mostly seen at retailers. That implies oxidation of the fat in the meat. As in the case of Humidity and Volatiles variation of the Index of Fats of acid depends on the chain and it's still on the channel A that its greatest changes are observable, indicating then the noticeable fluctuations in its cold chain. Figure 5. Acid index Change of Fats in milligrams per gram of fat during the distribution of products The variation of the content of Total Volatile Basic Nitrogen (Figure 6) is contrary to that of the humidity and volatile substances. The trend curves show an increase in this parameter on the two chains but chain A is higher. Figure 6. Variation of the TVB-N concentration (in mg per 100 g) during the distribution of products Note that the values obtained for this parameter do not exceed the guide value of 35 mg/100 g of sample. However, trends show a continuous deterioration of the products. This deterioration accelerated by the storage method could also lead to the decrease of the nutritional value of products. Thus, following the irregularities noted during this study, it is stated that traders and all those involved in this sector should adopt a traceable system [13, 14] and a quality management approach to their products. All retailers should use insulated boxes, possibly equipped with refrigerated gel packs or simply be storing ice packs [15]. Conclusion The levels of acidity of fat and values of the Total Volatile Basic Nitrogen (TVB-N) obtained indicate that the degradation and/or alteration are not significant at the wholesale level but the steps which follow. There is a death acceleration phenomenon, this could lead to loss of quality of turkey wings. An alternative solution remains to maintain the cold chain by using appropriate devices. It’s also adequate to implement the FIFO process.