Approximately 20 mg of total lipid (from cold-pressed oil) were transferred to a Teflon-lined screw-top test tube with a glass pasteur pipette. Fatty acids were methylated to methyl esters by using 0.5 N sodium hydroxide (NaOH) in methanol and 14% boron trifluoride in methanol (Park and Goins, 1994). The fatty acid methyl esters (FAME) were quantified using a Varian 430 gas chromatograph with flame ionization detector with a fused silica capillary column (Chrompack CPSIL 88, 100 m length, 0.25 mm ID, 0.2 mm film thickness). Column temperature was 40-230 °C (held 2 minutes; 4 °C/minute; held 10 minutes). Fatty acid methyl esters in hexane (1ml) were injected into the column using a Varian CP8400 Auto-sampler with a split ratio of 100:1. The injection port as well as the detector were maintained at 250 °C. The carrier gas was hydrogen, at 45 psi. The makeup gas was nitrogen. Galaxy Star Chromatography Software was used to record the chromatograms. The peaks were identified by comparing the relative retention times of FAME peaks from samples with that of standards from Supelco (Supelco 37 Component Fame Mix 47885-U, Sigma-Aldrich Aston Manor, Pretoria, South Africa). Fatty acids were expressed as the relative percentage of each individual fatty acid as a percentage of the total of all fatty acids in the sample. The following fatty acid ratios and combinations were calculated from the fatty acid data: total saturated fatty acids (SFA), total mono-unsaturated fatty acids (MUFA), total polyunsaturated fatty acids (PUFA), total omega-6, total omega-3 fatty acid content and the PUFA/SFA (P/S) ratio.

Refractive Index (RI) was determined according to the Association of Official Analytical Chemists’ (AOAC), official method 921.08 (AOAC, 2000) with an Abbè programmed advanced refractometer from ATAGO® (Model RX 5000α) at 40 °C.

Hanus iodine Value (IV) was determined by the AOAC official method 920.158 (AOAC, 2000), which consists of adding a mixture of iodine and bromine in glacial acetic acid and measuring the excess of unused halogen by titration with sodium thiosulfate.

High performance liquid chromatography (HPLC) was used to determine the tocopherol amount in each sample using Shimadzu CR8A instruments (Champ sur Marne, France) equipped with a C18-Varian column (25 cm x 4 mm; Varian Incorporated, Middleburg, Netherlands). One gram of each oil sample was added to 5 mL of ethanol (EtOH), filtered (45μm) into a HPLC vial at 25 ºC. HPLC was carried out with an injection volume of 20 μL, a flow rate of 1 ml/min and estimated at UV of 296 nm. The mobile stages were: mobile stage A (Acetonitrile: Methanol: Isopropanol: Water (45:45:5:5, v/v/v/v)) and mobile stage B (Acetonitrile: Methanol: Isopropanol (50:45:5, v/v/v)).

For the Oxygen Radical Antioxidant Capacity (ORAC) analysis, the method of Prior et al. (2003) was used, where samples (10 g each) were diluted with 80% methanol in H₂O (1:1 w/v), and then vortexed for 2 minutes at room temperature and centrifuged for 10 minutes at 500 X g. The peroxyl radicals were generated with 2, 2’-Azobis (2-amidinopropane) dihydrochlorine (AAPH) and B-phycoerythrin (B-PE) was used as the detector of radical activity. The final reaction mixture was prepared in 10-mm-wide quartz cuvettes as follows: 1600 μL of 0.04 μM B-PE in 0.0075 M sodium (Na) - potassium (K) phosphate buffer at pH 7.0 and 200 μL of 50 μM Trolox (6-hydroxy-2,5,7,8-tetramethylchroman -2-carboxylic acid). Phosphate buffer was used as a blank (200 μL of methanol, diluted 1:20 (v/v) with 0.075 M Na-K, at pH 7.0 and Trolox (β and γ tocopherol analogue) as the standards during each run. Samples were pre-incubated at 37 ºC for 15 min. AAPH was added. Flourescence was measured and measurements were recorded every 5 minutes for 35 minutes at 570 nm until a steady state (plateau) of fluorescence decay was reached. The results were then calculated and expressed as μmol of Trolox equivalent (TE). The analyses were done in triplicate.

Free fatty acids were determined according to the method of Pearson (1973), which involves the titration of the oil in an alcoholic medium against potassium hydroxide using phenolphthalein as indicator.

Peroxide value was determined by the official method of the AOAC 965.33 (AOAC, 2000). Peroxide value determines hydrogen peroxide, a primary oxidation product, and involves peroxides liberating iodine from potassium iodide by calculating the amount of sodium thiosulfate consumed.

The ρ-AV was determined using the technique of Hamilton and Hamilton (1992). It determines the amounts of aldehydes and ketones during secondary oxidation by reaction with ρ-Anisidine (4-methoxyaniline) using a spectrophotometer at 350 nm.

The oxidative stability index (OSI) was determined using the AOAC OSI official method Cd 12b-92 (AOAC, 2000) with a Rancimat 743 apparatus from Metrohm. Three samples were tested per cultivar.

The oxidative stability index (OSI) was determined using the AOAC OSI official method Cd 12b-92 (AOAC, 2000) with a Rancimat 743 apparatus from Metrohm. Three samples were tested per cultivar. Oxidative stability index (OSI) values were determined at 100, 110, 120 and 130 °C for each sample in duplicate. The extrapolation function of the Rancimat 743 apparatus was used to predict induction times at 25 and 30 ^oC (Rancimat Manual, 2009) according to the formula:

The fatty acid compositions of cold-pressed oil from a selection of cactus pear cultivars are shown in Table 1. According to the results presented, there are considerable variations in the fatty acid percentage of the selected cultivars. The dominating fatty acids identified (in decreasing order) were linoleic acid (C18:2 n-6) (C18:2c9,12); oleic acid (C18:1 n-9) (C18:1c9); palmitic acid (C16:0) and stearic acid (C18:0); while palmitoleic acid (C16:1 n-9) (C16:1c9), α-linolenic acid (C18:3 n-3) (C18:3c9,12,15), arachidic acid (C20:0) and lignoceric acid (C24:0) were notable in small amounts in all cultivars. C16:0, C18:0, C18:1c9 and C18:2c9, 12 were the only fatty acids detected at levels of more than 1%. This is in accordance with the results of various authors who reported that cactus pear seed oil contained primarily unsaturated fatty acids namely linoleic and oleic acid, and with a lower but significant content of palmitic and stearic fatty acids (Chougui et al., 2013; Ciriminna et al., 2017; Ortega-Ortega et al., 2017; Ramírez-Moreno et al., 2017, Regalado-Rentería et al., 2018; Loizzo et al., 2019). A completely different profile was found for the Greek O. ficus-indica which included butyric acid (C4:0), palmitic acid, stearic acid and oleic acid as the main fatty acids (Karabagias et al., 2020).

The C18:2c9,12 content varied from 58.56% (Nudosa) to 65.18% (Robusta) (Table 1). A slightly larger range for C18:2c9,12 between 57.75-67.32% was reported by de Wit et al. (2017) and de Wit et al. (2018) (56.86-65.21%). Labuschagné and Hugo (2010) also reported comparable results for fatty acid profiles with unsaturated fatty acids (C18:2c9,12) ranging between 57.75-67.32%. Linoleic acid (omega-6) is an essential fatty acid and a precursor of arachidonic acid (AA) biosynthesis which is the substrate for eicosanoid synthesis (Ghazi et al., 2013). Both present a hypocholesterolemic effect and inhibit colon cancer. Omega-6 PUFA is the source of inflammatory mediators’ prostaglandins (PGE) and leukotrienes. Consumption of oleic acid (n-9) and α-linolenic acid (n-3) act as AA antagonists and reduces the production of inflammatory mediators (Koshak et al., 2020). Higher levels were found in O. microdasys (Engelm) and O. macrorhiza as well as O. dillenii (> 70%) (Ghazi et al., 2013; Chahdoura et al., 2015; Ali Alsaad et al., 2019); while the current results are in agreement with the levels reported for O. ficus-indica varieties from Sicily, Saudi Arabia, Turkey and Tunisia (Ghazi et al., 2013; Ciriminna et al., 2017; Loizzo et al., 2019). Astiasarán and Candela (2000) reported a comparable study on PUFAs (C18:2c9,12) content of various vegetable oils, i.e. soy oil which obtained 48.7%, corn oils (47.7%), sesame oils (44.5%), sunflower oil (49.7%), olive oil ranging from 3.5-21% and cotton oils with 50%. These values are all significantly lower than the values obtained in this current study, implying that cactus pear is the best in terms of oil PUFA yield. PUFAs are able to alleviate symptoms of diseases such as coronary heart disease, stroke and rheumatoid arthritis.

The highest amount of C18:1c9 was detected for American Giant with 16.37% compared with Gymno Carpo, which demonstrated the lowest amount with 12.22%. Robusta contained 13.46% C18:1c9. The highest value for vaccenic acid (C18:1 n-7) (C18:1c7) was observed for Meyers (6.53%) and the lowest for American Giant (5.14%). Robusta (O. robusta) obtained 5.48%. Loizzo et al. (2019) reported values ranging from 15.8 - 18.1% for C18:1c9 from two O. ficus-indica cultivars (red and yellow) extracted by two different methods (Soxhlet and Ultrasound-Assisted). These authors reported values ranging between 4.3% and 5.1% for C18:1c7. The values for C18:1c9 reported by these authors were higher than in the current study, while the C18:1c7 values were somewhat lower than the current study’s values. Ciriminna et al. (2017) obtained a higher content of vaccenic acid (6.29%) for yellow Sicilian Opuntia ficus-indica seed oil, while Chougui et al. (2013) reported no vaccenic acid in yellow Algerian fruit. A vaccenic acid content of 4.83% was reported by Tlili et al. (2011) for Tunisian cactus pears. Interestingly, seed oil from O. ficus-indica from Marocco had no oleic acid (Ghazi et al., 2013).

The saturated fatty acid (SFA) palmitic acid (C16:0) ranged between 10.97 and 15.07% and stearic acid (C18:0) ranged between 2.62 and 3.74%. American Giant obtained the lowest content of C16:0 (Table 1) and the highest oil percentage (Figure 1). These C16:0 values are slightly lower than those reported by de Wit et al. (2017) which ranged between 12.90 and 17.83% (Monterey and Nudosa, respectively) and the range reported by de Wit et al. (2018) namely 12.72-16.05%. The C18:0 content varied between 2.62% (Gymno Carpo) and 3.74% (American Giant) (Table 1), indicating only slight variation in C18:0 levels. Similar values were reported for chemically extracted oil as reported by de Wit et al. (2017) (2.21-3.39%) and de Wit et al. (2018) (2.4-4.01%). These results are in agreement to those of Loizza et al. (2019) who found that fatty acid content is not affected by extraction method but by variety. Interestingly, Regalado-Rentería et al. (2018) found that oil extracted by cold-pressing was richer in UFA and poorer in SFA than oils extracted with solvents. Regarding SFA, the palmitic acid (5.12%) and stearic acid (7.51%) found in O. dillenii presented the main saturated fatty acid (> 22%). These are even higher than the SFA found in soy (~3%) (Ali Alsaad et al., 2019). Stearic acid has a neutral effect on LDL.

According to Table 2 SFA ranged from 14.76 (Zastron) to 18.50% (Nudosa). The results for the fatty acid composition presented in Table 1 show that C16:0 and C18:0 were observed to be marginally lower than those reported in the publication of de Wit et al. (2017) (16.59-20.65%) and de Wit et al. (2018) (16.19-19.12%) yet higher than those obtained by Ennouri et al. (2005). These fatty acids will influence SFA ratios, since they are the main contributors to SFA. The stage of development of the fruit may be the cause of variation in fatty acid composition, but in this study the fruits were all picked at the same maturity stage (50% color break), although genotype, climate and weather, as well as environmental effects may be the reason behind this variation. According to de Wit et al. (2017), weather conditions, particularly were the main factors contributing to variations in fatty acid composition and oil percentage.

Total MUFA ranged between 19.12 and 23.10%, which corresponds to that reported by de Wit et al. (2017) (15.78-23.30%) and de Wit et al. (2018) (17.4-23.61%). Gymno Carpo and Zastron obtained the lowest MUFA contents (19.13% and 19,12%) (Table 2), while Meyers and Van As showed the highest contents of MUFA (23.14% and 23.10%), respectively. MUFA, such as C18:1c9, ranged between 12.22% (Gymno Carpo) and 16.07% (Van As) (Table 1). Oil containing high concentration of C18:1c9 has been reported to be more stable to oxidation, which is desired for improved shelf-life (Lajara et al., 1990).

PUFA was established as the most prominent fatty acid ratio ranging between 59.23 (Nudosa) and 66.13% (Zastron). PUFA from de Wit et al. (2017) was at the lower range and ranged from 55.98 - 67.62%. C18:2c9,12 was established as the dominating PUFA with the lowest level of 58.56% for Nudosa and the highest level of 65.18% for Robusta (Table 1), with C20:3c9,12,15 (0.22 to 0.35%) and C20:3c8,11,14 (0.13 to 0.22%) occurring at very low levels.

Omega-6 (n-6) fatty acids ranged from 58.78-65.84% (Nudosa and Zastron, respectively). All the cultivars, with the exception of Nudosa, contained n-6 contents at above 60%. Nudosa attained the lowest concentration of n-6 fatty acid ratio with 58.78% (Table 2). Zastron had the highest contents of PUFA and n-6 but obtained the lowest content of MUFA. These contents were found to be slightly higher compared to results from de Wit et al. (2017) (55.98 - 67.45%). Pardo et al. (2009) concluded that cactus pear seed oils are good sources of omega-6 fatty acids.

The omega-3 (n-3) fatty acid ratio occurred at very low levels (Table 2). The contents ranged from 0.20 (Meyers) - 0.46% (Gymno Carpo), respectively. Omega-3 (n-3) fatty acids, such as C18:3c9,12,15 (C18:3 n-3) occurred at a very low level and ranged from 0.18 - 0.36% (Meyers and Morado in Table 2).

The PUFA/SFA ratio is used to measure the level of unsaturation and is taken as an instrument to measure the oil’s tendency to undergo autoxidation. The PUFA/SFA ratio ranged from 3.20 - 4.48% with Zastron occurring in the highest amount and Nudosa at the lowest level (Table 2). American Giant, Ficus-Indice, Morado, Robusta, Tormentosa, van As and Zastron also attained PUFA/SFA contents above 4%. Zastron was also amongst those that had the highest content of C18:2c9,12, while Nudosa occurred at the lowest level (Table 1). O. dillenii presented a ratio of 3.22.

The results for the physicochemical properties of the oil from cactus pear cultivars are presented in Table 3.

The RI values of the oils increase as a result of autoxidation. The RI depends on the chemical composition and temperature of the oil. It also increases with degree of saturation (Brahmi et al., 2020) and secondary functions on fatty acid chains. As indicated by the RI values presented in Table 3, little differences were observed among the RI values for all cultivars. The RI values ranged from 1.4659-1.4668 (Nudosa and Zastron). The higher the level of unsaturation, the higher the levels of RI and IV, for example, Zastron obtained the highest RI value (1.4668) and high PUFA content (66.08%) as well as high IV (127.43). Oils with a high unsaturation degree are more likely to undergo autoxidation and tend to oxidize easily when used at high temperatures, for example, PUFA (C18:2c9,12) oxidized ± 50 times faster than C18:1c9 (MUFA). Therefore, an increased RI value could be expected as a result of exposure to heat and light.

American Giant and Gymno Carpo were also amongst the cultivars with high PUFA, RI value and IV value (Tables 2 and 3). Furthermore, the higher the MUFA content (C18:1c9), the lower the RI and IV-values, for example, Van As attained a higher content of MUFA (23.14%), a low RI value of 1.4664 as well as a low IV value of 120.59 (Tables 2 and 3). These results fall within a similar range (1.464 RI) to those reported by Gharby et al. (2011). Previous literature reported a higher RI value of 1.4831 in cactus pear seed oil, 1.473 in rape seed oil, and 1.475 in Opuntia ficus-indica (Ennouri et al., 2005). Other studies reported slightly lower RI values ranging between 1.4658 and 1.4676 (de Wit et al., 2017).

Valuable information can be obtained by determining the density of the oils. Density gives information on the nature of fatty acids, such as chain length, unsaturation degree, and functional groups. Density will also be influenced by the composition and temperature of the oil. Together with RI, it can also indicate the purity of the oil (Brahmi et al., 2020). It is therefore recommended to include density analyses in future studies.

Iodine value is an indication of the total number of double bonds. The iodine value (IV) ranged from 120.58-127.43 (Tormentosa and Zastron). Zastron obtained the highest IV (number of gram iodine absorbed by 100g of oil) of 127.43 followed by American Giant (125.45), Morado (123.32) and Nudosa (123.92). Tormentosa recorded the lowest IV of 120.58 (Table 3). According to Table 2, Zastron and American Giant attained the highest PUFAs (high unsaturation level) as well as high IV and low MUFA content while Nudosa obtained the smallest content of PUFA. The values in this study are higher than those reported by Karleskind and Wolff (1992) ranging from 105.5-107.38 (cactus pear seed oils). De Wit et al. (2017) reported slightly lower iodine values ranging between 110.68 (Nudosa) and 126.82 (Robusta).

Since IV measures the unsaturation of oils, it can also be an indication of the stability of the oil since a high degree of unsaturation suggests a high susceptibility to oxidation. Low IV therefore indicates more resistance to oxidation because of increased saturation and vice versa (de Wit et al., 2017).

Vegetable oils are the most important dietary source of tocopherols. Tocopherols are phenolic compounds that are naturally occurring antioxidants that present biological activity (Ali Alsaad et al., 2019). Fernández-Martinez et al. (2004) indicated that fatty acids, natural antioxidants, tocopherols and sterols are the key constituents that define oil quality. Normand et al. (2006) stated that oils with high tocopherols have greater stability to oxidation. Tocopherols are natural antioxidants in plant foods, particularly those that contain high levels of PUFA as they are able to scavenge lipid peroxyl radicals of unsaturated lipid molecules, inhibiting the propagation of lipid peroxidation.

β and γ tocopherols ranged between 61.47 and 51.69 mg/100g (Table 3). Ofer obtained a higher content of β and γ tocopherols (51.69 mg/100g) followed by Morado (61.31 mg/100g), which means that Ofer and Morado will be more stable to oxidation compared to Nudosa (51.69 mg/100g). The values obtained in the current study are comparable to those reported by Chu et al. (2002) with cold-pressed cactus pear seed oil at 94.6 mg/100g, argan oil at 85.0 mg/100g, olive oil at 22.0 mg/100g, soybean at 65.0 mg/100g and sunflower oil at 49.0 mg/100g. Morales et al. (2012) found values of 140-220 µg/100 g.

Tocopherol contents are influenced by the extraction method (Loizzo et al., 2019). Red cultivars were found to contain higher contents than the yellow cultivars and were mainly in the form of γ-tocopherol. In Moroccan cactus pear seed oil, vitamin E was only present as γ-tocopherol (1.23%) (Ghazi et al., 2013). Regalado-Rentería et al. (2018) found a tendency for metabolites such as γ-tocopherol to be higher in cold-pressed oils than oils extracted with solvents. These authors reported γ-tocopherol values between 1.71 and 13.86 mg/100 g oil. During the analysis of anti-inflammatory compounds of O. ficus-indica seed oils from Saudi-Arabia, β-tocopherol contents of 1.56% were found. This oil was extracted with hexane (Boshak et al., 2020).

Lipid oxidation is a process which is initiated by free radical reactions at the double bonds of unsaturated fatty acids and has been reported to be the key factor in the quality deterioration of edible oils since it modifies the chemical, sensory and nutritional properties of the oils. Márquez-Ruiz et al. (1996) pointed out that autoxidation is another key factor that leads to the development of quality loss in refined oils during storage. The 2,2-diphenyl-1-picrylhydrazyl (DPPH) assay is normally used to predict the stability of cold-pressed oils against oxidation. Most of the literature cited included the use of the DPPH and 2,2’-azino-bis-3-ethylbenzthiazoline-6-sulphonic acid (ABTS) methods. Cao et al. (1996) reported the determination of total antioxidant capacity using the ORAC method, which is currently being used to determine the total antioxidant capacity and measures the plant extract’s ability to scavenge peroxyl radicals.

The oxidation of oils happens mainly at the sites of unsaturation and the oxidation rate mainly depends on the number of double bonds and their positions. Table 3 shows that Zastron had the highest antioxidant capacity with an ORAC value of 5.37 μmol/g, followed by Algerian with 5.28 μmol/g. Meyers had the lowest antioxidant capacity (3.87 μmol/g). Kuti (2004) reported higher values which ranged from 1.6-15.8 mMTE/g in yellow-skinned fruits and 1.7-49.2 mMTE/g in the purple-skinned fruits.

It was found by Ramírez-Moreno et al. (2017) that extraction solvent had a significant effect on the free radical scavenging capacity of the seed oil and that the green O. ficus-indica variety had a higher antioxidant activity regardless of the solvent used. The antioxidant activity of O. ficus-indica seed oil from Saudi-Arabia (Brahmi et al., 2020) showed a lower DPPH value in cold-pressed oil than in oils extracted by the Soxhlet method. Antioxidant activity was strongly correlated with the amounts of phenolic compounds, including polyphenols and flavonoids. O. dillenii seed oil from Iraq demonstrated a strong antioxidant capability due to its ability to reduce oxidation as determined by the DPPH method (Ali Alsaad et al., 2019), which reinforced the effect of variety on antioxidant activity.

The level of free fatty acids (FFA) in oils is measured by the acid value (AV) generated upon the hydrolytic degradation of lipid particles and is therefore an indicator of hydrolytic activity, which consequently adds to the decrease in the time span of usability of the oil. The maximum FFA content limit recommended for cold-pressed oil is 8%, 0.3% for refined oils and 5% for virgin palm oils according to the Codex Alimentarius Commission standard (Codex Alimentarius Commission, 2009). According to the results presented in Table 3, the FFA content ranged between 1.15 and 4.30% (Morado and Van As) and the % FFA of all cultivars was within the recommended limit, implying that cactus pear seed oil is of good quality from a FFA point of view. In a previous study, % FFA ranged between 2.49 and 5.08% in chemically extracted oil (de Wit et al., 2017). The content in cold-pressed oil was thus lower. The production of FFA is a result of lipid hydrolysis which is triggered by chemical or enzymatic actions. Increased free fatty acid values are indicators of increased primary oxidation. Factors such as high temperature and humidity are accountable for an increase in FFA content in oils. An elevated acidity value of 21.2% was reported by Brahmi et al. (2020) for Algerian cold-pressed O. ficus-indica seed oil. This high value could be ascribed to the enzymatic hydrolysis of seeds during oil pressing, handling or processing, as well as elevated temperatures and the presence of water during the process. It was also mentioned by the authors that free fatty acids could be present since no refining of the oil took place, which could possibly remove acids as impurities.

Martín-Polvillo et al. (2004) reported that peroxide value (PV) measures the level of hydroperoxide in the oil and is a valuable tool for the indication of a commencement period of oxidation. Furthermore, PV reaches the highest level during the progression of oxidation and following this stage (secondary oxidation), the decomposition rate of hydroperoxides surpasses the rate of their development. Peroxide value is also used to test oxidative rancidity in oils and fats (Karleskind and Wolff, 1992). It was reported that the acceptable PV value for cold pressed oil is < 15 meq O₂·kg^-1 (Codex Alimentarius Commission, 1999). According to the results presented in Table 3, all cultivars attained PV values of < 15 meq O₂·kg^-1, therefore PV value results are in agreement with the recommended value for cold-pressed oil of good quality.

PV ranged between 3.02 and 5.87. Although Gymno Carpo and Tormentosa were amongst the cultivars with higher PV values (5.87 and 5.80), oxidation had not yet occurred in these cultivars since their PV values were still within the recommended limit of < 15 meq O₂.kg^-1, which indicates that a rancid taste was not yet noticeable (Table 3). Salvo et al. (2002) pointed out that a rancid taste in oils begins to be noticeable at PV values of 20-40 meq O₂·kg^-1. PUFAs are more susceptible to oxidation than MUFA and SFA. The higher PVs observed for Gymno Carpo and Tormentosa were probably a consequence of their higher unsaturation levels (Tables 1 and 2). American Giant had on average a PV of 3.02 meq O₂·kg^-1, which was the lowest compared to that of the other cultivars. Van As also had a relatively low PV value of 3.60 (Table 3). De Wit et al. (2017) reported PV values of chemically extracted cactus pear oils which were above the recommended limit and ranged between 9.5 and 23.30 meq O₂·kg^-1.

Karleskind and Wolff (1992) reported PV estimations of margarine containing Opuntia ficus-indica oil which were far below the levels permitted by the international benchmarks and ranged between 0.38 and 0.39 meq O₂·kg^-1. PV measures the extent of rancidity reactions during storage, and therefore indicates quality and stability. A PV of 12 meqO₂/kg was reported for Algerian O. ficus-indica cold-pressed seed oils (Koshak et al., 2020). PV is therefore affected by genetics, cultivars, growth conditions, soil geography, and harvesting routines as well as handling and storage.

The ρ-AV was taken as a tool used to quantify secondary oxidation by surveying the measurement of unsaturated aldehydes as a result of the decomposition of hydroperoxides (Shahidi and Zhong, 2005). The ρ-AV is an indicator of oxidative rancidity in oils. The results presented in Table 3 demonstrate that the ρ-AV ranged between 0.98-4.08 mmol/kg for Tormentosa and Algerian, respectively. Algerian had the highest ρ-AV of 4.08 followed by Van As with ρ-AV of 2.94 mmol/kg. Tormentosa acquired the lowest ρ-AV content of 0.99 mmol/kg. The results demonstrated that there were only small differences among the ρ-AV of American Giant (1.37), Morado (1.40), Ofer (1.64) and Nudosa (1.58). Algerian recorded the highest ρ-Anisidine value and PV value, which is indicative of secondary oxidation products. The high ρ-AV may be due to their higher level of unsaturation. The recommended ρ-Anisidine value (ρ-AV) limit is < 10.0 mmol/kg (Codex Alimentarius Commission, 1999). The ρ-AV values obtained were within the acceptable limit (Table 3). Some of these ρ-AV values were higher than those reported in the literature which ranged from 0.02 - 3.73 mmol/kg (de Wit et al., 2017). Aldehydes made up the highest percentage of cactus pear seed oil volatiles (Karabagias et al., 2020). Aldehydes are responsible for, among others, acrid, burnt-fat, fruit-like, fermented, nutty, deep fat, butter, chicken and fusty odors, flavors and smells. The butyric acid reported by Karabagias et al. (2020) is responsible for the buttery flavor of the cactus pear seed oil.

The OSI value gives an indication of the resistance of lipids to oxidation and is also used for quality control of the oils, as well as an indication of the shelf-life of oils. The shorter the reaction time, the more susceptible the oil is to oxidation and therefore higher values will imply more resistance to oxidation. The values ranged between 1.78 and 3.17 h and were lower than the values obtained by de Wit et al. (2017), which ranged between 1.79 and 4.15 h. Ficus-Indice had the highest OSI value of 3.17 hours at 110 °C followed by Nudosa (2.92 hours) and Ofer (2.92 hours), implying that these cultivars were more resistant to oxidation reactions. Van As, which attained the highest content of C18:1c9 (Table 1), was determined to be the most favored oil since it is rich in oleic acid and oils rich in C18:1c9 have potential for combining the hypocholesterolemic effect and greater oxidative stability. Oils which are high in MUFA (C18:1c9) such as Nudosa (Table 2) are less susceptible to oxidative degradation and in this manner indicate potential for applications requiring high oxidative stability. Gharby et al. (2011) reported OSI of 7 hours at 110 °C for Opuntia ficus-indica, 31.23 hours for argan, 7.5 hours for olive oil and 5 hours for soybean and sunflower oil.

Induction time extrapolated to 25 and 30 °C was done to predict the shelf-life of the oils based on OSI using Rancimat induction times (Rancimat Manual, 2009). The induction period is an indication of how long the oils will take to reach an end point or be oxidized. American Giant is presented as an example in Figure 2. Induction time is also described as the time taken before the rapid increase in oxidation occurs and is used to make relative assessments on oxidative stability (Gharby et al., 2011).

The accelerated testing can then be extrapolated to real time conditions. Algerian was found to have a shorter induction time of 5 months, 2 weeks when extrapolated to 25 °C and 3 months, 5 weeks when extrapolated to 30 °C, implying that Algerian will have a shorter shelf-life compared to Tormentosa, with an anticipated time span of usability of 9 years and 4 months when extrapolated to 25 °C and 5 years 3 months when extrapolated to 30 °C (Table 3).