Inoculation with acetic acid bacteria improves the quality of natural green table olives

1. INTRODUCTION

The olive tree (Olea europaea L.) constitutes an economic and social resource for the development of the populations of the zones where it is cultivated. The growing demand for olive oil and table olives has created favorable conditions for a more efficient expansion of its cultivation, accompanied in many countries by specific development programs (Fernandez Escobar et al., 2013Fernandez Escobar R, de la Rosa R, Leon L. 2013. Evolution and sustainability of the olive production systems. Options Méditerranéennes. Séries A Mediterr. Semin. 106, 11-41.). In Morocco, the olive sector is supported by the Green Morocco Plan launched in 2008. Recent statistics provided by the Moroccan agriculture department (2019) show that olive orchards cover about 1.1 million ha, distributed in irrigated lowlands at 39%, mountain areas at 36% and rainfed areas at 25%. Moreover, Morocco occupies one of the top positions among exporters of table olives in the world. According to the Ministry of Agriculture and Maritime Fisheries, during the 2013/2014 season, the domestic production was in the order of 1,6 million tons of olives, including 120,000 tons of table olives for the same period (7.6% total olive production) (Anon, 2016Anon. 2016. International olive council. Conclusions IOC Conf. COP22. Available at: http://www.internationaloliveoil.org/news/view/686-year-2016news/787-conclusions-of-the-ioc-conference-at-cop22?lang=fr_FR [Accessed July 20, 2017].).

Table olives, which are the most widespread fermented vegetables in Mediterranean countries, have a great economic significance as a food commodity. Their high nutritional value, the content in bioactive compounds, dietary fibers, fatty acid composition and antioxidants make table olives a valuable functional food (Campus et al., 2018Campus M, Degirmencioglu N, Comunian R. 2018. Technologies and trends to improve table olive quality and safety. Front. Microbiol. 9. https://doi.org/10.3389/fmicb.2018.00617 ). Green table olives are considered the most popular fermented vegetable commodity in Morocco. They constitute a major agro-industrial sector in the country’s economic development (Rokni et al., 2015Rokni Y, Ghabbour N, Chihib NE, Thonart P, Asehraou A. 2015. Physico-chemical and microbiological characterization of the natural fermentation of moroccan picholine green olives variety. J. Mater. Environ. Sci. 6, 1740-1751.).

Olives are usually harvested at different stages of maturity and then processed to eliminate the bitter taste due to the glucoside named oleuropein. The green table olive de-bittering process can be accomplished in two ways: a) The Spanish-style method, i.e. using a lye treatment prior to fermentation, and b) the natural method (Sánchez Gómez et al., 2006Sánchez Gómez AH, García García P, Rejano Navarro L. 2006a. Trends in table olive production Elaboration of table olives. Grasas Aceites 57 (1), 86-94. https://doi.org/10.3989/gya.2006.v57.i1.24 ). The Spanish style (the most common conventional preparation), consists of an alkaline treatment with sodium hydroxide, followed by several water washings of the fruits to remove residual lye. Then, the fermentation is continued in the brine from an initial salt concentration of 10-11% to reach the value of 5-6% of NaCl at equilibrium. The Californian style is quite distinct from that used for any other style of table olive, as there is no fermentation step. In this method, olives are subjected to a lye treatment (0,5 to 2% NaOH), followed by darkening with ferrous gluconate or ferrous lactate. Then, the olives are bottled or canned in brine and pasteurized or sterilized (Colmagro et al., 2001Colmagro S, Collins G, Sedgley M. 2001. Processing Technology of the Table Olive. in Jules J, (Ed.) Horticultural Reviews. John Wiley and Sons, Inc, pp. 235-242. https://doi.org/10.1002/9780470650783.ch5 ; Marsilio et al., 2001Marsilio V, Campestre C, Lanza B. 2001. Phenolic compounds change during California-style ripe olive processing. Food Chem. 74, 55-60. https://doi.org/10.1016/S0308-8146(00)00338-1 ). In the other hand, the natural method, also known as “Greek style”, does not require any lye treatment of the fruit, which is placed directly into the brine (8-14% NaCl and pH between 4.0 and 4.2), where fermentation takes place (Colmagro et al., 2001Colmagro S, Collins G, Sedgley M. 2001. Processing Technology of the Table Olive. in Jules J, (Ed.) Horticultural Reviews. John Wiley and Sons, Inc, pp. 235-242. https://doi.org/10.1002/9780470650783.ch5 ). In this case, the hydrolysis of oleuropein is attributed to the enzymatic reaction of ß-glucosidase and esterase produced by the microorganisms (Amer et al., 2017Amer A, Faiz ul-Hassan N, Kashfa B, Aasia B. 2017. Microbial β-Glucosidase: Sources, Production and Applications. J. Appl. Env. Microbiol 5, 31-46. https://doi.org/10.12691/JAEM-5-1-4 ).

Among the main lactic acid fermented vegetable products, i.e. cucumbers, cabbages and olives, Spanish-style green-olives are the most economically important in Morocco. This method is the most widely used for making green olives on an industrial scale. It is based on the alkaline treatment of olives to eliminate oleuropein, the agent responsible for the bitterness of the fruits of the olive tree (Ramírez et al., 2017Ramírez E, Medina E, García P, Brenes M, Romero C. 2017. Optimization of the natural debittering of table olives. LWT - Food Sci. Technol. 77, 308-313. https://doi.org/10.1016/j.lwt.2016.11.071 ). After de-bittering, the olives undergo successive washes to remove residual lye. The treatment consists generally of (a) lye treatment 1.5 à 2% NaOH for 8-12 h at 24-25 °C; (b) washing by tap water for 14-15 h in two or three soakings, and (c) brining in 10-12% NaCl (El-Khaloui and Nouri, 2007El-Khaloui M, Nouri A. 2007. Procédés d’élaboration des olives de table à base des variétés Picholine marocaine et Dahbia. Transfert de Technologie en Agriculture 037, 77-80.; Chemonics International, 2007Chemonics International, Inc. 2007. Guide De Bonnes Pratiques De Fabrication Des Olives De Table, Agriculture and Agrobusiness Intégrés, USAID-Maroc, Repport n. 071, pp.1-42.).

Although the Spanish-style method is a fast process, it presents some drawbacks as well for the producer (lye and washing water expenses), for the consumer (risk of the chemical residues and the deterioration of the nutritional value of the treated olives) (Shahidi and Kiritsakis, 2017Shahidi F, Kiritsakis A eds. 2017. Olives and Olive Oil as Functional Foods, John Wiley & Sons, Ltd, Chichester, UK. https://doi.org/10.1002/9781119135340 ) and for the environment (releases of high quantities of wastewater of high organic matter, high phenolic content, high salinity and conductivity). Indeed, this process causes a strong degradation of oleuropein, which constitutes a significant loss in the nutritional value of de-bittered olives (Rokni et al., 2015Rokni Y, Ghabbour N, Chihib NE, Thonart P, Asehraou A. 2015. Physico-chemical and microbiological characterization of the natural fermentation of moroccan picholine green olives variety. J. Mater. Environ. Sci. 6, 1740-1751.). Recently, several sources reported the possibility of use of unauthorized acids by producers such as sulfuric acid to quickly neutralize the lye residue after the de-bittering of olives. This is done in order to reduce the time and cost of washing and obtain more profit (Faid, 2018Faid M. (n.d.). Beware of eating red olives / Mohamed Faid - YouTube. Retrieved April 6, 2020, from https://www.youtube.com/watch?v=yJoUDDsy-c0&t=75s ). Hence, it could be important to further develop the sector by using natural table olives to overcome the drawbacks mentioned above. Natural table olives are harvested at the green-yellow stage, or fully ripened, previously washed and graded, then submerged into NaCl solutions (6-10% w/v), where fermentation takes place, mainly due to the metabolism of autochthonous microbiota (Campus et al., 2018Campus M, Degirmencioglu N, Comunian R. 2018. Technologies and trends to improve table olive quality and safety. Front. Microbiol. 9. https://doi.org/10.3389/fmicb.2018.00617 ). Yeasts and LAB control the progress of fermentation since they are more sensitive to salt concentration and the acidification of brines which are determined by metabolic activity exerted mainly by LAB (Campus et al., 2018Campus M, Degirmencioglu N, Comunian R. 2018. Technologies and trends to improve table olive quality and safety. Front. Microbiol. 9. https://doi.org/10.3389/fmicb.2018.00617 ). On the other hand, acetic acid bacteria (AAB) can also be associated with olives and olive products, in which they can play several roles (Kavroulakis and Ntougias, 2011Kavroulakis N, Ntougias S. 2011. Bacterial and β-proteobacterial diversity in Olea europaea var. mastoidis- and O. europaea var. koroneiki-generated olive mill wastewaters: Influence of cultivation and harvesting practice on bacterial community structure. World J. Microb. Biotechnol. 27 (1), 57-66. https://doi.org/10.1007/s11274-010-0426-3 ; Valero et al., 2017Valero A, Medina E, Arroyo-López FN. 2017. Microbial hazards and their implications in the production of table olives. in Singh O V, (Ed.) Foodborne Pathogens and Antibiotic Resistance. John Wiley & Sons, Inc, pp. 119-138. https://doi.org/10.1002/9781119139188.ch5 ). Acetobacter pasteurianus is one of the acetic acid bacteria (AAB) which have the capacity to incompletely oxidize ethanol as a substrate to produce acetic acid. For this reason, AAB are often involved, either directly or indirectly, in the fermentation of food and drinks (Cleenwerck and De Vos, 2008Cleenwerck I, De Vos P. 2008. Polyphasic taxonomy of acetic acid bacteria: an overview of the currently applied methodology. Int. J. Food Microbiol. 125, 2-14). It has been perfectly established that producers of table olives use organic acids, such as lactic, citric and also acetic acid, in order to favor the rapid start of fermentation and avoid the excessive increase in pH at the beginning of fermentation (Degirmencioglu, 2016Degirmencioglu N. 2016. Modern Techniques in the Production of Table Olives. in Boskou D, Clodoveo ML, (Ed.) Products from Olive Tree. IntechOpen, pp. 215-234. https://doi.org/10.5772/64988 ). It has also been reported that vinegar may be added to the brine solution to preserve the olives, adjust the pH, and impart a particular flavor (Colmagro, 2001Colmagro S, Collins G, Sedgley M. 2001. Processing Technology of the Table Olive. in Jules J, (Ed.) Horticultural Reviews. John Wiley and Sons, Inc, pp. 235-242. https://doi.org/10.1002/9780470650783.ch5 ). Thus, following our research, it seemed appropriate to consider three species of microorganisms: Lactobacillus plantarum S1, Saccharomyces cerevisiae LD01 and Acetobacter pasteurianus KU710511 CV01 isolated from Morocco (Mounir et al., 2016aMounir M, Belgrire M, Lahnaoui S, Hamouda A, Thonart P, Delvigne F, Ismaili Alaoui M. 2016a. Maîtrise de la fermentation alcoolique sous stress éthanolique, thermique et osmotique de la souche Saccharomyces cerevisiae YSDN1 en vue de la préparation du vinaigre de fruits. Rev. Marocaine Sci. Agron. Vétérinaires 4, 86-95.; Mounir et al., 2016bMounir M, Shafiei R, Zarmehrkhorshid R, Hamouda A, Thonart P, Delvigne F, Ismaili Alaoui M. 2016b. Optimization of biomass production of Acetobacter pasteurianus KU710511 as a potential starter for fruit vinegar production. African J. Biotechnol. 15, 1429-1441. https://doi.org/10.5897/AJB2016.15323 ; Mounir et al., 2016cMounir M, Shafiei R, Zarmehrkhorshid R, Hamouda A, Ismaili Alaoui M, Thonart P. 2016c. Simultaneous production of acetic and gluconic acids by a thermotolerant Acetobacter strain during acetous fermentation in a bioreactor. J. Biosci. Bioeng. 121, 166-171. https://doi.org/10.1016/j.jbiosc.2015.06.005 ; Mounir et al., 2018Mounir M, Fauconnier ML, Afechtal M, Thonart P, Ismaili Alaoui M, Delvigne F. 2018. Aroma profile of pilot plant-scale produced fruit vinegar using a thermo-tolerant Acetobacter pasteurianus strain isolated from Moroccan cactus. Acetic Acid Bact. 7, 1-11. https://doi.org/10.4081/aab.2018.7312 )

The overall objective of our work was to develop a method for the preparation of natural table olives using locally selected microorganisms and without resorting to the usual techniques. This objective involves optimizing the synergy conditions between the selected microorganisms. To do this, ten tests were carried out (on washed and sorted olives) so that we could study the effect of the de-bittering, the effect of inoculation by yeasts, the effect of the substitution of organic acids with vinegar, the effect of the substitution of organic acids with acetic bacteria and finally the effect of alternating aeration. The monitoring of physicochemical and microbiological parameters and organoleptic evaluation were carried out to judge the quality of the olives produced.

2. MATERIALS AND METHODS

3. RESULTS

From more than 60 isolates, three microbial strains were selected to carry out experiments aimed at the preparation of natural table olives. These strains, namely : Lactobacillus plantarum S1, Saccharomyces cerevisiae LD01 and Acetobacter pasteurianus KU710511 (CV01) were identified using biochemical and molecular techniques and characterized in previous studies (Mounir et al., 2018Mounir M, Fauconnier ML, Afechtal M, Thonart P, Ismaili Alaoui M, Delvigne F. 2018. Aroma profile of pilot plant-scale produced fruit vinegar using a thermo-tolerant Acetobacter pasteurianus strain isolated from Moroccan cactus. Acetic Acid Bact. 7, 1-11. https://doi.org/10.4081/aab.2018.7312 ; Mounir et al., 2016aMounir M, Belgrire M, Lahnaoui S, Hamouda A, Thonart P, Delvigne F, Ismaili Alaoui M. 2016a. Maîtrise de la fermentation alcoolique sous stress éthanolique, thermique et osmotique de la souche Saccharomyces cerevisiae YSDN1 en vue de la préparation du vinaigre de fruits. Rev. Marocaine Sci. Agron. Vétérinaires 4, 86-95.; Mounir et al., 2016bMounir M, Shafiei R, Zarmehrkhorshid R, Hamouda A, Thonart P, Delvigne F, Ismaili Alaoui M. 2016b. Optimization of biomass production of Acetobacter pasteurianus KU710511 as a potential starter for fruit vinegar production. African J. Biotechnol. 15, 1429-1441. https://doi.org/10.5897/AJB2016.15323 ; Mounir et al., 2016cMounir M, Shafiei R, Zarmehrkhorshid R, Hamouda A, Ismaili Alaoui M, Thonart P. 2016c. Simultaneous production of acetic and gluconic acids by a thermotolerant Acetobacter strain during acetous fermentation in a bioreactor. J. Biosci. Bioeng. 121, 166-171. https://doi.org/10.1016/j.jbiosc.2015.06.005 ) The determination of chloride, titratable acidity, pH, and counts of LAB, yeasts and AAB were carried out using the methods described in the Material and methods section. Then, samples were presented to tasters to evaluate their organoleptic quality.

The Figure 1 shows that the pH of the traditional test remained between 5 and 5.5 throughout the experiment, with three relatively high values observed between the 40th and 55th day of the experiment. On the other hand, the titratable acidity, expressed in grams of lactic acid per liter started with an acidity of 1.7 at the beginning of the experiment, and gradually decreased to 0.09 g/l, as noted in the 40^th day. Then it started to increase continuously to reach 1.75 in the last days of the experiment.

Figure 2 shows the evolution of the number of Lactic Acid Bacteria and Yeast in the brine during the spontaneous fermentation process of the traditional assay. This figure reveals that the evolution pattern of the concentration of the two types of microorganisms was similar. The two groups of microorganisms had an average value of 3.2·10⁵ CFU/ml during the first days of the experiment. The graph also shows that the concentration of the two groups changed by increasing to exceed 3·10⁶ CFU/mL on the 12^th day, and decreased to around 2·10⁶ CFU/mL between the 12^th and 23^th days. After that, the number of cells continued to increase slightly for the LAB strains; while the increase was much more significant for the yeast strains on the last 10 days to reach 4.7·10⁶ CFU/mL at the end of the experiment.

3.1. Effect of de-bittering and vinegar suppletion on the fermentation process

 ⌅

In this part of the study, de-bettering was performed with sodium hydroxide and two successive washes. After that, date vinegar, produced with the S. cerevisiae LD01 yeast strain along with the L. plantarum S1 LAB strain, was added to the 10.6% (w/v) NaCl brine, all anaerobically. Then, three flasks were inoculated with the LAB strain only and three others were inoculated both with the LAB and yeast strains. This test was not seeded by the AAB strain, so we did not count it.

Figure 3 shows the effect of de-bittering on fermentation process parameters with and without inoculation with the S. cerevisiae strain. As can be seen from the graphs in Figure 3.A, the titratable acidity for samples not seeded by the S. cerevisiae strain was 0.2 g/L at the beginning of the experiment, then gradually increased to 0.8 g/L lactic acid on the 27th day and then varied at around 0.67 g/l until the end of the experiment. For inoculated samples with the S. cerevisiae strain (Figure 3.B), the pH started with a value of 6.2 at the beginning of the fermentation cycle, then the pH stabilized at 5.8. The titratable acidity started with a value of 0.4 g/L lactic acid and it then gradually increased from the 8th day to around 1 g/L to be stabilized at this value afterwards. In this assay, the microbial count of the three repetitions monitored separately showed the same trend, especially during the first thirty days. The average of the initial values for the LAB strain in inoculated samples (Figure 3.B) was quite high (1.4· 10⁸ CFU/mL), and a very slight increase was observed up to the 18^th day followed by a sudden drop until the 30^th day to be close to a concentration of 10⁶ CFU/ml. A slight increase occurred during the last 30 days to reach a maximum value of 5.8 10⁷ CFU/ml for the third repetition and around 5.9·10⁶ CFU/mL for the other two replicates (Figure 3.B).

Figure 3.  Fermentation process parameters in de-bittered samples using 2-3.5% w/v NaOH. Three replicates (L1V1, L1V2 and L1V3) were inoculated with 0.75% of S. cerevisiae LD01 strain (B), and three others were not (A) used as reference (L1VS1, L1VS2 and L1VS3). All samples were inoculated with 0.75% of L. plantarum S1 LAB strain. The samples were prepared according to the descriptions given in Table 1 (test L1V and L1VS, respectively). The tests were carried out anaerobically with one kg of olives brined in 2-liter bottles. Cell numbers were determined by the plate count method using MRS and YPD media, respectively, for LAB and Yeast.

On the other hand, Figure 3.B shows that the S. cerevisiae LD01 strain experienced similar growth for the 3 replicates, with an average concentration of 2.2·10⁷ CFU/mL, which remained more or less stable until the 20th day, followed by a decrease in concentration to arrive at 2.4·10⁶ CFU/mL. After that, the three repetitions experienced an increase in concentration with slopes that differed between each repetition to reach the end of the experiment at an average value of 3.6 10⁷ CFU/mL. However, Figure 3.A shows that the LAB strain had an average growth value of 10⁷ CFU/mL at the beginning of the experiment, with a concentration of 2·10⁸ CFU/mL on the 13^th day for the third repetition. The concentration remained approximately at an average of 5 10⁶ CFU/mL, then dropped slightly for the 3 samples to reach 9·10⁶, 1.2·10⁶ and 2.1·10⁶ CFU/mL for the three repetitions, respectively. Figure 3.A also shows that the yeast concentration of the first repetition showed an overall decrease from 1.65·10⁷ CFU/mL to reach 8.6·10⁶, 2.80·10⁶ and 8.2·10⁵ CFU/mL, repetitively, for repetitions 1, 2 and 3.

The second experiment was conducted with acidification of the brine with 10.6% (m/v) date vinegar produced in the lab using the selected yeast and AAB strains. The fermentation was carried out under anaerobic conditions. Three flasks were inoculated with the LAB strain (L2V1, L2V2 and L2V3) and three others were inoculated with the LAB and yeast strains (L2VS1, L2VS2, L2VS3). Microbiological counting of this test gave no results for lactic acid bacteria and yeasts during the first days of the experiment.

In samples not inoculated with S. cerevisiae (Figure 4.A), the starting pH was 3.9, which was adjusted by the addition of vinegar, then the pH increased to an average value of 4.5 on the 19^th day. Then, from the 20^th day, the pH fluctuated between 4.5 and 5 until the end of the experiment. For titratable acidity, the average starting value was 3.18 g/L lactic acid, and it then slightly decreased until the 28th day with 3.00 g/l and then stabilized at an average value of 3.50 g/l. For microbial growth, the curves obtained for the three repetitions did not keep the same pace. However, we noticed that there was a slightly positive effect on the growth of LAB but an insignificant effect on the growth of endogenous yeasts (Figure 4).

Figure 4.  Fermentation process parameters in non de-bittered samples with acidification of the brine with 10.6% (m/v) date vinegar. Three replicates (L2VS1, L2VS2, L2VS3) were inoculated with 0.75% of S. cerevisiae LD01 strain (B), and three others were not (A) used as reference (L2V1, L2V2 and L2V3). All samples were inoculated with 0.75% of L. plantarum S1 LAB strain. The samples were prepared according to the description provided in Table 1 (test L2VS and L2V respectively). The tests were carried out anaerobically with one kg of olives brined in 2-liter bottles. Cell numbers were determined by the plate count method using MRS and YPD media, respectively for LAB and Yeast.

3.2. Alternated aeration effect on microbial growth in the brine of naturally fermented olives

 ⌅

The olives of this test were prepared in brine at 6% (w/v) NaCl under the optimal conditions of A. pasteurianus KU710511 strain with alternating aeration at a rate of 0.2xV/h and 8 hours per day. Three flasks were seeded with LAB and AAB strains (L3V1, L3V2, L3V3) and three others with LAB, AAB and the yeast strains (L3VS1, L3VS2, L3VS3).

A shown in Figure 5A, the pH was around 3.9 during the first 28 days, and increased substantially to be stabilized at around 5.2 from day 40. The acidity started with an average value of 13.5 g/L of lactic acid and dropped significantly and stabilized at an average of 3 g/L. The pH was around 3.9 during the first 28 days for samples that were inoculated with the yeast strain (Figure 5.B), and increased significantly to about 5.5 by the end of the experiment. The acidity started with an average value of 16 g/L of lactic acid and dropped significantly and stabilized at an average value of 2.9.

Figure 5.  Fermentation process parameters of naturally fermented samples under alternating aeration and inoculation by AAB strain conditions. Three replicates (L3VS1, L3VS2 and L3VS3) were inoculated with 0.75% of S. cerevisiae LD01 strain (B), and three others were not (A) used as reference (L3V1, L3V2 and L3V3). All samples were inoculated with A. pasteurianus KU710511 (0.75%) and with L. plantarum S1 LAB strain (0.75%). The samples were prepared according to the description provided in Table 1 (test L3VS and L3V, respectively). Cell numbers were determined by the plate count method using MRS and YPD media, respectively, for LAB and Yeast.

We also noted that during the first days of the experiment, the yeasts showed an average growth of 8.35·10⁵ CFU/mL and then increased to around 8·10⁹ CFU/mL. Subsequently, the yeasts of L3V1 fell to 3·10⁷ CFU/mL on the 44^th day, increasing to 6.4·10¹¹ CFU/mL 13 days later. Those of L3V2 continued their progressive rise to arrive at 8.5 10¹⁰ UFC/mL towards the end (57^th day). L3V3 peaked on the 38th day at 8.7·10⁹ and dropped to 7.8·10⁷ CFU/mL on day 62. Whereas, the LAB strain exhibited a growth of approximately 5·10⁶ CFU/ml at the beginning of the experiment, then gradually increased to an average value of 2.21·10¹¹ CFU/mL towards the end of the experiment with an intermediate increase at around 5.8·10⁹ CFU/mL on day 31 for L3V1 and on day 38 for L3V3. On the other hand, the AAB strain started at very low concentrations (about 10³ CFU/mL), and subsequently increased to around 10⁸ CFU/mL by the 30th day. Subsequently, the concentration decreased to be stable at 5·10⁶ CFU/mL from the 42^nd day to the end of the experiment.

However, Figure 5B shows that the yeast strain showed a growth of 1.4·10⁴ CFU/mL for the L3VS1 test at day 6. As for the L3VS2 and L3VS3 repetitions, the first determination made on the 13th day gave an average of 7.2·10⁶ CFU/mL. Subsequently, the L3VS2 and L3VS3 replicates ranged from around 5·10⁹ CFU/mL; while L3VS1, after reaching a maximum of 9.3 10⁹ CFU/ml, gradually decreased to 3.7·10⁷ CFU/mL at the 57th day. The LAB strain had a cell concentration of approximately 5·10⁵ CFU/mL at the beginning of the experiment, then gradually increased to an average value of approximately 1.62·10¹⁰ CFU/mL towards the end of the experiment with an intermediate increase to around 1.5·10¹⁰ CFU/mL on the 36th day. Finally, the AAB strain started at very low concentrations (about 10³), and subsequently increased to around 10⁸ by the 30th day. Subsequently the concentration decreased to be stable at 3.5·10⁶ CFU/mL towards the end of the experiment.

3.3. Organoleptic evaluation of produced olives

 ⌅

The contingency table (supplementary file 2) contains the frequencies of the tasters who noted the different quality criteria of the samples by classifying them as “good”, “medium” or “bad”. Criteria were then subjected to Factorial Correspondence Analysis. The calculation of eigenvalues indicates that the variation shares explained by axis 1 and axis 2 were 50.1% and 22.6%, respectively. The main plane constituted by these two axes therefore alone accounted for 72.7% of the total variation of the cloud of points (Figure 6). The projection of the line points (10 samples) and the column points (quality criteria) in the main plane is shown in Figure 6. Examination of this figure indicates that:

The vertical axis F2 opposes the group formed by the samples C, D, A, G and H characterized by good appearance, good bitterness and quite normal firmness to the samples B, E and F, which have opposite characteristics, namely a normal to bad external appearance, a hard firmness and a quite pronounced bitterness. The horizontal axis F1 opposes the group formed by the samples G and H, characterized by a good smell and a normal salty flavor compared to the group formed by A, C and D, which is characterized by a normal to bad odor and a medium to bad salty flavor. Samples I and J were close to the center and are therefore not very well presented by this plan. However, it can be said that they have good appearance with a good smell to bad, fairly strong bitterness and hard firmness.

Figure 6.  Graphical representation of the results of a Factorial Correspondence Analysis (FCA) of the sensory evaluation of the fermented olives: projection in factorial axes 1 and 2 of fermented olives groups (A, B, C, D, E, F, G, H, I and J (Table 1)), regarding classes of the qualitative attributes (External appearance, firmness, odor, salty flavor, bitterness and altered flavor (Table 2)). The vertical axis opposes the group formed by the samples C, D, A, G and H, which are characterized by good appearance, good bitterness and normal firmness as compared to the samples B, E and F, which had opposite characteristics (dash-point). The horizontal axis opposes the group formed by the samples G and H, which are characterized by good attributes compared to the group formed by A, C and D (solid line). G and H samples were expected to be the most highly appreciated ones.

4. DISCUSSION

In this study, four different combinations were applied to the “Picholine marocaine” olive variety using indigenous strains, namely Lactobacillus plantarum S1, Saccharomyces cerevisiae LD01 and Acetobacter pasteurianus KU710511 (CV01).

First, we noted that the greater or lesser variability between the repetitions of the same test at the level of the microbiological characteristics can be due to the moment of analysis. Since these were not carried out at the same time for the three repetitions (considering our human and material resources), we opted for the separation of repetitions over time by taking all the necessary precautions (operator, material and environment effects).

The traditional test was the only one where the brine was changed regularly (every three days), which may explain the variability in the microbiological results of the three repetitions, and also the fact that the pH alternatively varied from around 5.3 throughout the experiment. This behavior was also observed for titratable acidity. We also noted that the increase in microbial count, both for LAB and yeast strains, was not very pronounced since after a first value of about 2·10⁵ CFU/mL there was an increase to reach 10⁶ CFU/mL, then there was stagnation around this value throughout the experiment.

Results found by Zaragoza et al., (2017)Zaragoza J, Bendiks Z, Tyler C, Kable ME, Williams TR, Luchkovska Y, Chow E, Boundy-Mills K, Marco ML. 2017. Effects of Exogenous Yeast and Bacteria on the Microbial Population Dynamics and Outcomes of Olive Fermentations. MSphere 2, 1-14. https://doi.org/10.1128/mSphere.00315-16 revealed that the introduction of exogenous spoilage yeast and LAB into olive fermentations caused significant but distinct alterations in the numbers and diversity of microbes associated with the olives and brines. According to the microbial evolution of tests 1 and 2 (Table 1), we noted that the acidity had a more severe effect on the microbial concentration than the basicity, indeed this concentration was very weak or even null in the first days of experiment for test 2 (supplemented by vinegar). Whereas for test 1, where pH values were around 6 to 7 (because of NaOH residues), the population was relatively high and even reached 10⁸ CFU/mL at the beginning of the experiment but dropped gradually after a few days.

We also noted that alternate aeration increased the concentration of the LAB and the yeast strains, since for the anaerobic tests, it did not exceed 10⁷ CFU/ml; whereas for tests 3 and 4 (alternating aeration) it reached 10¹¹ CFU/mL. A possible relationship between LAB and yeasts during the fermentation of olives in brine was reported by (Heperkan 2013Heperkan D. 2013. Microbiota of table olive fermentations and criteria of selection for their use as starters. Front. Microbiol. 4, 1-11. https://doi.org/10.3389/fmicb.2013.00143 ). In addition to the LAB usually used in this kind of fermentation, the use of Saccharomyces cerevisiae yeast as a starter is justified by the role that it can play in the fermentation of olives by producing ethanol, ethylacetate, acetaldehyde and other compounds responsible for the development of olive flavors (Ciafardini and Zullo, 2019Ciafardini G, Zullo BA. 2019. Use of selected yeast starter cultures in industrial-scale processing of brined Taggiasca black table olives. Food Microbiol. 80, 103250. https://doi.org/10.1016/j.fm.2019.103250 ; Fernandez Escobar et al., 2013Fernandez Escobar R, de la Rosa R, Leon L. 2013. Evolution and sustainability of the olive production systems. Options Méditerranéennes. Séries A Mediterr. Semin. 106, 11-41.).

On the other hand, it was demonstrated that fermentation in altered aeration mode, served to modify the metabolic pathway of L. plantarum by promoting the conversion of lactate into acetate (Bobillo and Marshall 1991Bobillo M, Marshall VM. 1991. Effect of salt and culture aeration on lactate and acetate production by Lactobacillus plantarum. Food Microbiol. 8, 153-160. https://doi.org/10.1016/0740-0020(91)90008-P ). Indeed, acetic acid can inhibit the growth of gram-negative strains which are responsible for the degradation of table olives during storage (Makras and De Vuyst 2006Makras L, De Vuyst L. 2006. The in vitro inhibition of Gram-negative pathogenic bacteria by bifidobacteria is caused by the production of organic acids. Int. Dairy J. 16, 1049-1057. https://doi.org/10.1016/j.idairyj.2005.09.006 ). In addition, this process allowed us to use lower concentrations of NaCl (6% w/v) compared to the concentrations used in anaerobic fermentations (10 to 15% w/v). Reducing salt in the processing of olives is always a goal. In fact, a diet that is low in sodium and high in potassium and calcium is recommended to lower blood pressure and to protect against osteoporosis, colon cancer, and cardiovascular diseases (Degirmencioglu, 2016Degirmencioglu N. 2016. Modern Techniques in the Production of Table Olives. in Boskou D, Clodoveo ML, (Ed.) Products from Olive Tree. IntechOpen, pp. 215-234. https://doi.org/10.5772/64988 ).

The combination of alternating aeration, the use of optimal growth conditions of the AAB strain and the chosen NaCl concentration served to create a synergy among the three strains which led to the elaboration of table olives with very good organoleptic quality and in a fairly short time (40 days).

On the other hand, for test 4, the use of alternating aeration with the same NaCl concentration as test 3 was not sufficient to obtain good olives after 60 days of fermentation. This finding suggests that this trial required more time, despite the high LAB and yeast populations (of the same order of magnitude as in test 3), suggesting that the AAB strain plays a vital role in the success of fermentation by creating the synergy necessary to obtain table olives with good sensory quality in a relatively short time. Indeed, aerobic fermentation is generally used to avoid the appearance of a deterioration called “gaseous pockets” which is manifested by the swelling of the skin caused by accumulation of gas under the epidermis of the olive (Lanza 2013Lanza B. 2013. Abnormal fermentations in table-olive processing: Microbial origin and sensory evaluation. Front. Microbiol. 4, 1-7. https://doi.org/10.3389/fmicb.2013.00091 ). In addition, AAB strain could adjust the acidity of the medium and thus promote the onset of fermentation by providing yeasts and lactic acid bacteria with favorable conditions for their development by continuously eliminating the ethanol produced by the yeast and thus maintaining an acceptable level of pH by producing acetic acid (Hammoucha and Taleb, 2017Hammoucha J, Taleb O. 2017. Contribution à l’amélioration des conditions de fermentation des olives de table. Institut Agronomique et Vétérinaire Hassan II, Rabat. Maroc).

Concerning the industrial test, lye de-bettering resulted in olives without bitterness and a good appearance, although the texture was judged as normal to soft. On the other hand, the L3V test led to both a very good taste, a normal texture and a salinity which was very much appreciated compared to the industrial test.

5. CONCLUSION

The objective of the work was to contribute to the development of a method for the preparation of natural table olives using locally selected microorganisms and without resorting to chemical additives. The effects of the parameters, essentially inoculation with Lactobacillus plantarum S1, Saccharomyces cerevisiae LD01 and Acetobacter pasteurianus KU710511 strains and alternating aeration were assessed. The samples were inoculated with pure suspensions of these microorganisms and then monitored over time using the plate count method. Only molecular analyses could provide accurate data on the survival of the selected starters during fermentation, but the inoculated tests by selected strains most likely contributed to reach those CFU/ml numbers as they were inoculated in a high number at the beginning of the fermentations.

It has been reported previously that acetic acid, the main product of oxidative fermentation by AAB in ethanol, is a compound that has a positive correlation with the growth of yeasts and could have an accelerating effect on LAB growth (Pino et al., 2018Pino A, De Angelis MD, Todaro A, Van Hoorde KV, Randazzo CL, Caggia C. 2018. Fermentation of Nocellara Etnea table olives by functional starter cultures at different low salt concentrations. Front. Microbiol. 9. https://doi.org/10.3389/fmicb.2018.01125 ). The results obtained served to confirm that the aerobic fermentation normally used for the elimination of the deterioration of table olives due to “gas pockets” can also be used to create the conditions necessary for the introduction of the acetic acid bacteria which are strict aerobes in the fermentation process of the olives. In addition, alternated aeration conditions allowed for increasing the cell biomass during fermentation and the use of a lower concentration of NaCl compared to anaerobic fermentation.

The microbiological, physicochemical and sensory analyses carried out demonstrated the possibility of introducing a new process for producing natural table olives. In this method, we used a starter composed of two strains, namely Acetobacter pasteurianus KU710511 (CV01) and Lactobacillus plantarum S1 under the optimal conditions of AAB growth as reported by Mounir et al., (2016b)Mounir M, Shafiei R, Zarmehrkhorshid R, Hamouda A, Thonart P, Delvigne F, Ismaili Alaoui M. 2016b. Optimization of biomass production of Acetobacter pasteurianus KU710511 as a potential starter for fruit vinegar production. African J. Biotechnol. 15, 1429-1441. https://doi.org/10.5897/AJB2016.15323 with 6% NaCl and alternating aeration of 8 hours/day.