The investigations were carried out into three villages: Eganga, Ethiolo, and Ebarack, located in the department of Salemata in Kedougou-Senegal in June 2018 (Figure 1). The process monitoring was done at the level of 04 women: 02 in the Eganga’s village, 01 in the Ethiolo’s village, and 01 in the Ebarack’s village.

The survey consisted of recording the oil process made by women. The objective is to watch the women making the oil process as they do in their daily lives and then identify all the involved steps. After each procedure, oil and oilcake samples were taken for further analysis. Six oil samples were obtained and distributed as follow:

1) In Eganga, two oil samples were obtained prepared by two women: one sample per woman;

2) At Ebarack, two oil samples were obtained made by the same woman. The first one was prepared following the normal diagram process, and the second one was prepared without modifications;

3) At Ethiolo, two oil samples were obtained prepared by the same woman: one of the oil extraction was made in 2017, and the second one was prepared in our presence.

The six samples were subjected to chemical analysis. All the analysis was done in duplicate, and the results were given as mean ± standard deviation. The acid index was determined according to the standard method NF ISO 660, while the Iodine index was conducted according to the standard method NF ISO 3961. The standard method NF ISO 3657 method was used to determine saponification index and NF ISO 3960 for peroxide index. The refractive index was determined on the perfectly anhydrous and filtered sample according to standard method NF ISO 6320 with a refractometer.

The oil production began with the crushing step. The seeds of Quassia undulata, also called “Agnarka” in the Bassari language, were separated from their hulls (Figure 2(a)). This step was complicated and took a lot of time because the crushing is done with the teeth. After that, seeds were crushed, pounded, and sieved. During the third step, a decoction of P. thonningii leaves was produced. P. thonnintigi was poured into the water, and as soon as the boiling’ began, the leaves were removed (Figure 2(b)). The seed’s powder was then added to the decoction (Figure 2(c)). The Boiling continued until oil appears (fourth step). As soon as the quantity of oil was large enough, it was recovered using a spoon (Figure 2(c)). The last step was optional and consisted of a slight heating (Figure 2(d)) of the oil to eliminate traces of water in order to allow better conservation.

After oil recovery, the oilcake was processed to obtain pasta named “Enangue”. The oilcake was first poured into a large basket with holes (Figure 3(a)). It

was washed with a decoction of P. thonningii and then with water. During the second step, the oilcake was poured into small baskets with holes having Lanea velituna leaves at the bottom (Figure 3(b)). Water was then gradually poured onto the oilcake and was drained (dewatering step). Finally, after water draining, the baskets were left in the open air to dry the oilcake at normal environment temperature. When the oilcake was completely dewatered, the obtained dough called “Enangue” (Bassari language) was consumed with sugar or salt.

The monitoring of manufacturing processes by the four women from Eganga, Ethiolo, and Ebarack’s villages gave rise to the diagram in Figure 4.

In Africa, different vegetable oils are extracted in traditional mode. Many of these methods include a mix of the seed powder with water at room temperature; for example, in Morocco, argan oil is traditionally obtained by crushing the seeds in a stone wheel, then the almond paste is added with water and mixed with the hand to express the oil [11]. Elsewhere the traditional oil extraction consists of oilseeds boiling with water to separate the oil from the ashes. It’s the case of Elaeis guineensis and Butyrospermum parkii oil extraction [12]. The same

method is used for the traditional production of the Carapa (Carapa spp. Meliaceae) oil produced in West Africa, known to have a bitter taste [4]. However, Weber et al., 2010, during this survey, show that some plant particles were added by some women (optional) during the boiling step [4]. The Bassaris women used the same process in Kedougou to extract the Q. undulata seed oil. Mirailles et al., 1988 during their study about the Q. undulata fatty acid and quassinoids contents, used the Soxhlet mode with hexane as a solvent to extract oil from the seed. They obtain oil with a slightly bitter taste contrary to the oil obtained in this study using traditional extraction in the presence of P. thonnintigi leaves. Therefore, many questions arise about the P. thonnintigi key role during oil extraction. Are P. thonnintigi leaves responsible for the absence of bitter taste? Quassinoids are known as bitter principles [13], and Mirailles et al., 1988, in their study, show that Q.undulata oil is rich in quassinoids compounds (Klaineanone, Glaucarubolone, and Chapparinone). Therefore in the future, more analysis can focus on the research of quassinoids contents of Q. undulata oil obtained after this traditional extraction using P. thonnintigi leaves.

Oil samples obtained during the survey were subject to different physicochemical analyses. The results are shown in Table 1.

Table 1 shows that the peroxide indexes are lower than 10 mEq/kg for all the oil samples, indicating low oxidation according to the WHO/FAO limits that considered that rancid oil peroxide value is between 20 and 40 mEq/kg [14]. The refraction index and saponification value are quite identique. The refraction index is about 1.46, showing that the Quassia oil is non-siccative oil. The saponification value is between 197 mg/g and 198 mg/g for Eganga’s and Ebarack’s samples and between 190 mg/g for the oil sample from Ethiolo. These differences

can be explained by the boiling step used by the woman in Ethiolo after the oil recovering because the saponification value can decrease with the oil refinery step [15]. The acid values of samples are low and quite similar (1.22 - 4.21 mg/g). The small iodine value (from 30.04 to 38.07) may be due to Quassia oil small unsaturated acid contain [16]. This agrees with the results of Mirailles et al., 1988 in their study about the Q. undulata oil fatty acid content where it is showed that the oil is rich in saturated fatty acid [10]. The Quassia oil sample in this survey is solid at room temperature of 20˚C - 24˚C.

During the survey, it appears during the women’s interview that the oil was used as a skin ointment after oil extraction. So, we compared Quassia oil’s physico-chemical composition to shea butter, known as a high-quality skin ointment. The results are given in Table 2. Quassia undulata oil and shea butter are both solid at room temperature, and the iodine value and saponification value are similar. However, the peroxide value and the acid value of Quassia undulata oil are lower than the shea butter ones. According to that, Quassia undulata oil stability can be higher than shea butter.

On the other hand, Mirailles et al., 1988 after their study on the fatty acid and quassinoid content, concluded that Quassia oil could be used as cooking oil (only after removing the Quassinoids from the oil). So, Quassia oil’s physicochemical oil composition was compared with the peanut and corn oil (Table 2). It is noticed that Quassia oil iodine value is lower than the peanut and corn oil one. Besides, Quassia oil is solid at room temperature, contrary to corn and peanut oil at room temperature. It means that Quassia oil’s unsaturated fatty acid content is lower than corn and peanut oil. The three oils have almost the same refractive index. The Quassia oil saponification value is higher than the peanut and corn oil ones. The 03 oil peroxide values are lower than 10 mEq/kg, showing that Quassia oil stability can be compared to the corn and peanut oil one.

Table 3 gives a comparison between Q. undulata produced in Senegal and Q. undulata made in Nigeria. It is noticed a difference between the results obtained in the 02 studies. The saponification value, iodine value, and refractive index are very different except for the peroxide and the acid value. This can be due to the production location; in the study of [20], the Moringa stenopetala oil saponification value and iodine value varied according to the sampling location. Besides, the peroxide value and the acid value were similar. On the other hand, the differences can be explained by the extraction mode. The Q. undulata oil characterized by Oko et al. [21] was extracted using a soxhlet extractor and hexane as

the solvent. During the survey, the oil production was done by boiling the seed in the P. thonningii decoction. It appears then the P. thonningii leaves play a key role during the oil extraction.