Volatilome changes in brines along the spontaneous fermentation of Spanish-style cv. Chalkidiki green olives under high and low NaCl conditions

doi:10.21203/rs.3.rs-3122984/v1

Download PDF

Research Article

Volatilome changes in brines along the spontaneous fermentation of Spanish-style cv. Chalkidiki green olives under high and low NaCl conditions

https://doi.org/10.21203/rs.3.rs-3122984/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 21 Sep, 2023

Read the published version in Food and Bioprocess Technology →

You are reading this latest preprint version

Table olives are essential components of the Mediterranean diet and their consumption is constantly spreading worldwide. However, as there are special concerns regarding their contribution to salt intake, there is a trend to reformulate traditional products via safe interventions. The present work systematically examines the profile of volatile organic compounds (VOCs) in brines during spontaneous fermentation of Spanish-style cv. Chalkidiki green table olives under high (8% w/v) and 50% lower NaCl content using selected combinations of chloride salts (NaCl, KCl, CaCl₂) in pilot-scale for two consecutive periods. VOCs were determined by headspace solid-phase microextraction coupled with gas chromatography-mass spectrometry. Changes in population of the dominant microbial groups, pH and titratable acidity values were also investigated using appropriate methods. A total of 56 VOCs were identified, including acids, alcohols, carbonyls, esters, and phenols. A volatilomic approach through multivariate data analysis allowed to explore the links between volatile metabolites, microbial groups and fermentation stages under different NaCl conditions. The evolution of propanoic acid and 2-butanol was intriguing for monitoring a fermentation dominated by lactic acid bacteria under low NaCl conditions, while 3-methyl-1-butanol was indicative of yeast prevalence, arising from seasonal variation and/or high NaCl processing. Furthermore, a sharply increased content of acetic acid at a late stage of fermentation was typical in low NaCl brines. The comprehensive results aided to systematically evaluate the volatile variation in high/low NaCl content brines throughout spontaneous fermentation of Spanish-style cv. Chalkidiki olives and could be beneficial for marker-assisted monitoring of the fermentation processes in the future.

Spanish-style green table olives

Chalkidiki cultivar

volatile organic compounds

NaCl content

GC-MS

multivariate analysis

Table olives are essential components of the Mediterranean diet and their consumption is constantly spreading all over the world due to their nutritional and organoleptic characteristics. Recently published IOC data estimate a 7% increase of table olives global production in the 2022/23 period compared to the previous crop year, to reach 3,099,500 tons. The EU member states hold almost one-third of the global production. Among them, Greece is the second largest producer (24% of the EU production) and one of the most prominent exporters (IOC, 2023), justifying its role as a major force in the table olive industry.

Green table olives from cv. Chalkidiki cover 50% of the domestic table olive production with a five-year average (2017/18-2021/22) of 116,300 tons, 92% of which are exported (DOEPEL, 2023). These products are appreciated in the Greek and international markets for their specific texture and appearance. Since 2012, they are recognized with the Protected Designation of Origin (PDO) “Prasines Elies Chalkidikis” (EU, 2012). The traditional process refers basically to the Spanish-style debittering process that includes lye treatment with dilute aqueous NaOH solution (1.3–2.6%, w/v), followed by repeated washes with potable water for removal of the excess alkali. Then, the treated fruits are spontaneously fermented in NaCl brine (typically 8–11% w/v or even higher depending on local practices) for 2–4 months (Mastralexi et al., 2019). Although the process guarantees conformity with the current safety and quality specifications, there are serious concerns about the high NaCl content of the edible product (3.2-5.0 g per 100 g of edible flesh, or per 20–23 medium-sized olives) that does not align with international dietary guidelines (< 5 g/day salt from all sources for adults) (WHO, 2012).

Innovative approaches producing low NaCl products of Spanish-style green table olives involve the use of fermentation brines with reduced NaCl content (via total salt reduction or partial replacement by other chloride salts) (Anagnostopoulos & Tsaltas, 2022). Mantzouridou et al. (2020) pointed out that it is feasible to reformulate Spanish-style cv. Chalkidiki green table olives and reduce NaCl levels by safe interventions. The proposed process consisted of spontaneous (uninoculated) fermentation and 12-month storage in brines where NaCl was partially replaced with selected combinations of only permitted chloride salts (NaCl, KCl, CaCl₂, MgCl₂) (EU, 2011). Although incorporation of lactic acid bacteria (LAB) starter cultures (primarily Lactobacillus plantarum) could be effective in this regard (Bautista-Gallego et al., 2011, 2015; De Bellis et al., 2010; Pino et al., 2018, 2019; Vaccalluzzo et al., 2022), such an addition is not currently foreseen for any of the Greek table olive products.

Spontaneous fermentation of table olives takes place via a complex indigenous consortium of microbiota present in the raw material and the processing environment, principally constituted by LAB and yeasts. The reproducible formation of well-balanced microbiota together with the compositional characteristics of the olive fruit is crucial for ensuring that products comply with safety and quality standards. Taken that no thermal treatment is applied to the raw material, abnormal phenomena could occur (Vaccalluzzo et al., 2020). Furthermore, partial or complete removal of NaCl may favor the growth of pathogenic and/or spoilage microorganisms (Anagnostopoulos & Tsaltas, 2022) and modify the structure of microbiota (Mateus et al., 2016). So far, considerable efforts have been made toward early-stage detection of defective fermentation and estimation of the time needed for the process completion. In particular, the development of more sophisticated analytical and statistical techniques has stimulated numerous studies to monitor relevant biochemical markers throughout the process (Marx et al., 2017; Vaccalluzzo et al., 2020).

Headspace solid-phase microextraction (HS-SPME) coupled with gas chromatography-mass spectrometry (GC-MS) has been widely employed to investigate in detail the volatilome of brines and olives from Spanish-style green table olive fermentation in relation to cultivar, growing location (Cortés-Delgado et al., 2016; Garrido-Fernández et al., 2017; López-López et al., 2018; Mikrou et al., 2021; Montaño et al., 2003; Nanou et al., 2021), irrigation conditions (Cano-Lamadrid et al., 2015; Sánchez-Rodríguez et al., 2019) and operations during processing and storage (Alfonzo et al., 2018; Benítez-Cabello et al., 2019; Blana et al., 2014; de Castro et al., 2019; Garrido-Fernández et al., 2021a; Martorana et al., 2017; Nanou et al., 2021; Panagou & Tassou, 2006; Sánchez et al., 2000, 2017, 2018). Besides, the importance of the volatile organic compounds (VOCs) as possible markers of spoilage in brines from Spanish-style cv. Gordal and Manzanilla green table olive fermentations (de Castro et al., 2018; Garrido-Fernández et al., 2021b) and cv. Carrasqueña table olives (Sánchez et al., 2023) has been reported. Recently, Ruiz-Barba et al. (2023) successfully associated VOCs with the predominant microbial taxa and fermentation stages. Υet, few of these investigations included the volatilome of fermentation brines from Spanish-style cv. Chalkidiki green table olives traditionally processed under high NaCl conditions (≥ 8.0%) (Blana et al., 2014) or of the final products (Mikrou et al., 2021; Nanou et al., 2021). To our knowledge, there are no prior studies about the possible relationship between volatilome, microbiota and NaCl content of fermentation brines for this type of table olives.

The main objective of the present study was to investigate the volatilomic patterns of brines from spontaneous fermentation of Spanish-style cv. Chalkidiki green table olives under high and low NaCl conditions using selected combinations of chloride salts (NaCl, KCl, CaCl₂). Pilot-scale fermentations were run in parallel for two consecutive processing periods (2020/21, 2021/22). Special focus was given on exploring patterns of association between VOCs and the evolution of dominant microbial groups in each case. The goal was to identify biomarkers that could assist the control of the process under different NaCl conditions and the production of high-quality Spanish-style cv. Chalkidiki green table olives.

2.1. Materials

Edible commercial NaCl (Dadakarides Salt S.A., Thessaloniki, Greece) and food grade lactic acid (≥ 95% purity, Corbion Purac 80 company, Gorinchem, Netherlands) and citric acid (99.5-100.5% purity, Weifang Ensign Industry Co., Ltd, Shandong, China) were used by the local industry. Food grade CaCl₂ (95–98% purity) was purchased from Zirax (London, England) and KCl (99.2% purity) from K + S Minerals and Agriculture GmbH (Kassel, Germany). De Man–Rogosa–Sharpe (MRS) agar, Rose Bengal Chloramphenicol agar base (RBCA), Violet Red Bile Glucose agar (VRBGA), and chloramphenicol were from Lab M Limited (Heywood, UK). 4-methyl-2-pentanol (≥ 97.5% purity) and n-alkane standard solution (C8-C20) were purchased from Sigma–Aldrich (Steinheim, Germany). Cycloheximide (≥ 98% purity) and methanol (≥ 99.9% purity) were purchased from Glentham Life Sciences Ltd (Corsham, UK) and Honeywell (Seezle, Germany), respectively. All other reagents were of appropriate analytical grade.

2.2. Olive samples and processing

Processing took place at Athos Olive SA (Ormylia, Chalkidiki, Greece) for two consecutive periods (October 2020-February 2021, period A and October 2021-February 2022, period B). Olives from the cv. Chalkidiki were harvested at mature greenish-yellow stage of ripening in early-mid October from olive orchards located in Ormylia (Chalkidiki, Greece). Immediately after harvesting, the fruits were transported to the processing unit for debittering according to industrial Spanish-style treatment (Mastralexi et al., 2019). The processed olives (150 kg) were transferred to pilot-scale food-grade plastic vessels (220 L) and covered with freshly prepared brines (70 L). Three different fermentation treatments were performed according to chloride salts composition of brines (w/v): 8% NaCl (C), 4% NaCl-4% KCl (T1) and 4% NaCl-3% KCl-1% CaCl₂ (T2) (Mantzouridou et al., 2020). Citric acid (0.2% w/v) and lactic acid (0.9% w/v) were added 48h after brining for acidification. Spontaneous fermentation took place at room temperature (ca. 20–22 ^oC). Each treatment was performed in duplicate and monitored for 120 days (October-February). For each treatment, fresh brine was being added periodically to maintain the initial olive/brine ratio, while total salt content was being adjusted to the initial level (8% w/v) by the addition of appropriate amounts of the respective mixtures of salts (in dry form). Representative brine samples from these three independent treatments were analyzed at specific time intervals.

2.3. Microbiological and physicochemical analyses of brines

The pilot-scale fermentation processes were monitored by measuring microbial populations, pH and titratable acidity in brines. Brine aliquots without or after appropriate dilutions in a sterile saline solution (9 g/L NaCl) were plated in triplicate on the selective media reported by Mastralexi et al. (2019) to enumerate Enterobacteriaceae, LAB, yeasts and molds. Results from plate counts were expressed as log₁₀ cfu/mL. pH values and titratable acidity (g lactic acid/100 mL) were measured as previously reported (Mastralexi et al., 2019).

2.4. VOCs determination in brines

2.4.1. Headspace Solid-Phase Microextraction (HS-SPME)

The brines that were sampled at early, middle and late fermentation stages (30, 60 and 120 days) were pre-concentrated in VOCs by solid phase microextraction (SPME) and the volatile fractions were analyzed by gas chromatography-mass spectrometry (GC-MS), according to Pino et al. (2018) with slight modifications. Briefly, an aliquot of brine (10 mL) was placed into 20 mL autosampler headspace vial to be analyzed by HS-GC-MS and spiked with 4-methyl-2-pentanol as internal standard (IS; 50 µL of 20 ppm methanolic solution). Also, a pooled QC (Quality Control) sample was prepared by mixing equal volumes from each brine and analyzed to examine reliability across the analytical study. Brines were equilibrated at 40 ºC with agitation for 20 min. A single SPME 50/30 µm (DVB/CAR/PDMS) fiber (Supelco Analytical, Bellefonte, PA, USA) was exposed to the headspace above the sample surface inside the vial, at a depth of 25 mm, for 20 min at 40°C. Following extraction, the SPME fiber was retracted and injected into the gas chromatograph inlet and desorbed for 1 min at 230°C. After 8 samples, the fiber was being conditioned at 270°C.

2.4.2. Gas Chromatography-Mass Spectrometry analysis (GC-MS)

The HS-GC-MS analysis was performed using an EVOQ GC-TQ Bruker triple quadrupole (Bruker Daltonics, Bremen, Germany) system with a CTC-PAL autosampler (CTC Analytics AG, Zwingen, Switzerland). The volatiles were separated on a HP-INNOWAX column (30 m×0.25 mm×0.25 µm) (Agilent Technologies, Santa Clara, CA, USA) with helium (purity 99.999%) as a carrier gas at a constant flow rate of 1 mL min^− 1. Initially the oven temperature was set at 40 ºC, held for 2 min, raised to 50 ºC with a 4 ºC min^− 1 rate, remained there for 5 min, and then increased at 230 ºC with a 4 ºC min^− 1 rate, and remained there for 10 min. The inlet temperature was set at 230 ºC and the injections were performed in splitless mode. For the MS analysis, electron ionization (EI) was performed at 70 eV in full scan spectra mode, with an m/z range of 33–400 amu and a 250 ms scan time. The ion source and transfer line temperatures were maintained at 230 ºC and 250 ºC, respectively. Analyses of blanks were conducted to ensure no carryover, while pooled QC samples were injected to monitor the instrumentation drift.

Compounds were identified by comparing the: (i) retention indices (RI) with those of National Institute of Standards and Technology (NIST) 17 Mass Spectral Library (NIST 17, mainlib library); (ii) MS spectra of peaks with those retrieved from NIST 17. RIs were calculated based on the homologous series of an n-alkanes mixture (C8-C20) analyzed under identical conditions as the samples. Three tiers of Reliability of Identification (RID) were implemented. Tier A: agreement of RI (ΔRI < 20) and MS (match > 900); tier B: ΔRI < 20 and MS similarity match > 800; tier C: ΔRI < 20 or MS similarity match > 800.

Semi-quantitative determination was conducted with the internal standard method (IS; 4-methyl-2-pentanol). Integration of the peak areas was conducted by MSWS 8 (Bruker Daltonics, Billerica, MA) software. Each sample was analyzed in duplicate and the results were expressed in µg of compound per liter of brine (µg/L).

2.5. Data processing and statistical analysis

One-way ANOVA and Duncan’s multiple range tests were applied to pH values, titratable acidity values, microbiological population and VOCs content data for statistical comparisons of the mean values using the SPSS software (v. 28.0, SPSS Inc., Chicago, USA). The level of significance was set at p < 0.05.

Principal Component Analysis (PCA) was performed using the SIMCA© 16.1 software (Umetrics, Umeå, Sweden). The input dataset comprised 36 rows (n = 36; 3 treatments×3 fermentation stages×2 vessels×2 processing periods) and 56 columns with normalized gas-chromatographic data per identified peak (%) to the total peak area (each VOC value was the mean per duplicate analysis of each vessel). The covariance of the new multidimensional data was assessed after the selection of the “significant” Principal Components (PCs) according to the total variance explained (> 95% cumulative variation, R²X(cum)) and/or fitness to the Kaiser criterion (components with eigenvalues greater than 1.0). Groups or patterns were explored in the PCA score (t) plots using the software visualization tools. Variables of statistical importance were identified with the aid of p loading values (absolute p ≥ 0.3).

3.1. Microflora composition of the brine

As a first step, pilot-scale spontaneous fermentations of Spanish-style cv. Chalkidiki green olives in periods A and B (2020/21 and 2021/22, respectively) were monitored by investigating changes in the population of the main microbial groups (LAB, yeasts and Enterobacteriaceae) in brines with different combinations of chloride salts (8% NaCl (C); 4% NaCl-4% KCl (T1); 4% NaCl-3% KCl-1% CaCl₂ (T2)) (Fig. 1).

A clear change in the microbiota profile from one period to the other regarding especially the relative abundance of LAB and yeasts is readily observed in these plots (Fig. 1).

In particular, in period A, the growth pattern of LAB was almost uniform during fermentation in different brines. Despite the differences in LAB population recorded immediately after equilibrium (3 days), LAB became the dominant microbial group in all cases. LAB population showed a rapid increase within the first 8 days of fermentation in all brines, attaining a maximum average value of around 8.1 log₁₀ cfu/mL after the first 30 days. This value is similar to those reported for Spanish-style cv. Chalkidiki green olives processed in brines of similar composition at lab scale (5 L) (Mantzouridou et al., 2020). Then, LAB counts remained stable until around 80 days. Yeasts coexisted with LAB but their population was maintained 1–3 log₁₀ units below those of LAB for the first 80 days. However, from this time point onward, LAB and yeast counts progressively decreased and increased, respectively, and finally, they reached the same level of around 6.2 log units at the end of the fermentation processes (Fig. 1). The changes in population of Enterobacteriaceae were very similar between the different brines. Specifically, they were initially detected at 3.9–5.2 log₁₀ cfu/mL but decreased progressively and their population values were below 2.2 log₁₀ cfu/mL after 30–40 days of fermentation.

In period B brines, the fermentation processes were dominated by yeasts instead of LAB, regardless of chloride salts composition. LAB were capable of growth after long term implantation, i.e., 40 days in C and 60 days in T1 and T2 brines (Fig. 1). At the end of the fermentation processes, LAB population reached 6.8, 5.2 and 5.2 log₁₀ cfu/mL in C, T1 and T2 brines, respectively. Yeast growth pattern was similar in all brines, with an average maximum population of around 6.2 log₁₀ cfu/ml after 1 month that was maintained stable throughout the fermentation processes. This count was comparable with that evidenced in period A. The undesirable Enterobacteriaceae showed prolonged persistence up to 80 days reaching maximum values of 3.3–3.6 log₁₀ cfu/mL before being reduced to < 2.0 log₁₀ cfu/mL at the end of the fermentation processes.

3.2. Physicochemical characteristics of brines

Alongside the different LAB populations in the two consecutive periods, a significant variation in pH evolution was observed (Fig. 2). Thus, while all brines were acidified to pH values of 4.2–4.5 within the first 8–12 days, an unexpected raise of approximately 0.6 units was observed between 8 and 35 days only in brines of period B (Fig. 2). No such spikes were observed in LAB-dominated brines of period A. A similar observation regarding the brine pH evolution was also made in the past (2015/16), during industrial-scale processing of the PDO “Prasines Elies Chalkidikis” (Mastralexi et al., 2019). It seems that the adverse competence from yeasts and Enterobacteriaceae restricted LAB growth in this case. Nevertheless, at the end of all fermentation processes, the pH values were similar in all brines reaching minimum values of 4.0-4.2. Stronger acidification of period A brines compared to those of period B could be partially prompted because LAB outgrew yeasts causing a robust lactic acid production and increased acidity values (Fig. 2). However, a gradual increase of acidity until the end of fermentation was reported in both periods with final values ranging from 0.5 to 0.7 g lactic acid/100 mL brine.

Spanish-style cv. Chalkidiki green olives are fermented spontaneously via the indigenous microflora. Normally, LAB and yeasts govern the multispecies fermentation process. The ability of LAB to decrease the pH value of the brine via lactic acid production at the beginning of the process is crucial to inhibit the growth of Gram-negative bacteria (mainly Enterobacteriaceae) that may deteriorate the final product (Mastralexi et al., 2019). However, the results of this study showed that this consortium is uncontrolled, and, therefore, the outcome of the fermentation is unpredictable. The ratio between LAB and yeasts, as well as the dominance of either group during olive processing was strongly dependent on environmental factors such as seasonal variation, salt, pH etc. The dominance of yeasts during table olive fermentation may have both beneficial and harmful effects. Concerning the latter, some authors reported high pH value and low free acidity during yeast-dominated table olive fermentations, resulting in a product with milder taste and shorter shelf life (Arroyo-López et al., 2012). Nevertheless, unpublished data from sensory and microbiological quality assessment of the final products highlighted that, in both processing periods, the fermentation processes delivered safe and organoleptically acceptable products (data not shown).

In the next part of this study, the volatilome of the different brines in response to the microbial composition changes was systematically examined, aiming at better understanding of the metabolic processes that took place during table olive fermentations in brines with different chloride salts composition.

3.3. Analysis of Volatile Organic Compounds in brines

In total, 56 VOCs were systematically identified in the brines from different treatments regarding combinations of chloride salts (C, T1, T2) and different fermentation stage (30, 60 and 120 days) or processing period (A, B). They comprised acids, alcohols, carbonyls, esters and phenols (Tables 1 and 2).

Table 1

Concentration of Volatile Organic Compounds in brines with different composition in chloride salts, monitored during pilot-scale spontaneous fermentation of Spanish-style cv. Chalkidiki green olives in 2020/21 processing period (30, 60 and 120 days)
Code	Compounds^a	Brines^b
		CA			T1A			T2A
		30	60	120	30	60	120	30	60	120
	ACIDS
1	Acetic acid	702.5a	1025.5c,d	850.9b	800.8a,b	907.7b,c	1443.3f	691.0a	1044.9d	1169.8e
2	Propanoic acid	19.5a	31.0a,b	105.1d	27.4a,b	54.8b,c	347.5f	22.2a	76.2c	244.1e
3	2-Methylpropanoic acid	6.8b,c	8.4c,d	14.3f	1.6a	2.5a	10.0d,e	4.5a,b	2.3a	11.5e,f
4	Butanoic acid	4.6e	3.1a,b	4.1c,d,e	3.5b,c	2.6a	4.8e	3.6b,c,d	3.1a,b	4.4d,e
5	3-Methylbutanoic acid	13.5a,b,c	16.1c	26.9d	12.0a,b,c	9.4a	26.8d	9.7a,b	8.8a	14.5b,c
6	Hexanoic acid	14.5d	8.0c	4.5a,b	2.8a	6.3b,c	5.7b,c	5.4b	5.3b	5.7b,c
7	Octanoic acid	5.0e	2.3c,d	0.7a	1.0a,b	3.1d	2.0b,c,d	2.9d	1.7a,b,c	2.4c,d
8	Nonanoic acid	1.2b,c	0.5a	0.9a,b,c	0.8a,b	0.8a,b,c	1.3c	1.0a,b,c	0.7a	2.1d
	Total amount	767.6	1094.8	1033.1	849.7	987.2	1841.3	740.3	1142.9	1495.1
	ALCOHOLS
9	Ethanol	200.5b,c	394.2f	180.8a,b	240.9c,d	144.4a	305.6e	228.3c	387.1f	292.4d,e
10	2-Butanol	87.8a	302.3d	152.6b	192.2b,c	145.5b	335.9d	229.7c	408.2e	432.4e
11	1-Propanol	8.3a	3.6a	15.0a,b	9.9a,b	51.4c	212.0e	33.2a,b,c	151.4d	39.5b,c
12	2-Methyl-1-propanol	12.4c,d	17.1f	13.7d,e	9.5b,c	3.8a	8.0b	8.8b	9.5b,c	15.6e,f
13	2-Methyl-1-butanol	11.4a,b	22.9d	15.5a,b,c	28.2e	19.9c,d	18.2c,d	11.7a,b	16.0b,c	10.6a
14	3-Methyl-1-butanol	125.9b	294.3d	213.5c	93.1a	63.1a	130.5b	78.0a	87.8a	137.1b
15	3-Methyl-3-buten-1-ol	nd	9.3c	4.1a	6.7b	5.0a	4.4a	nd	nd	4.8a
16	3-Methyl-2-buten-1-ol	3.5a,b	3.1a,b	4.8c,d	2.7a	6.1e	8.6f	3.6a,b	4.2b,c	5.8d,e
17	1-Hexanol	56.4d	53.1c,d	42.9a	47.7a,b,c	53.4c,d	64.6e	51.8b,c,d	45.0a,b	48.4a,b,c
18	(Z)-3-Hexen-1-ol	63.5c,d	68.7c,d	59.9b,c	52.3b,c	71.5c,d	21.0a	54.1b,c	39.2a,b	82.8d
19	2-Ethyl-1-hexanol	4.3d	1.4a	2.6b,c	1.4a	1.2a	1.9a,b	1.6a,b	1.6a,b	3.6c,d
20	1-Octanol	nd	3.1c,d	2.7c	0.9a	nd	3.7d	6.1f	2.0b	4.9e
21	1-Nonanol	3.7d	1.4b	0.1a	nd	2.1c	1.5b	nd	1.4b	nd
22	Benzyl alcohol	21.6b,c	17.0a,b	14.3a	14.7a	18.1a,b	25.4c	14.4a	13.7a	19.5a,b,c
23	Phenylethyl alcohol	83.2b	60.6a	66.6a	47.8a	60.6a	83.6b	48.9a	49.9a	60.2a
	Total amount	682.4	1252.1	749.9	748.0	646.1	1225.0	770.2	1217.0	1136.7
	CARBONYLS
24	Octanal	nd	1.2a,b,c	1.3b,c	nd	1.0a,b	1.8c	nd	nd	0.8a
25	Benzaldehyde	27.6d	12.4c	13.1c	7.6b	8.1b	11.0c	1.8a	2.3a	3.8a
	Total amount	27.6	13.6	14.4	7.6	9.1	12.7	1.8	2.3	4.6
	ESTERS
26	Methyl acetate	5.7a	22.7e	13.5b,c	11.7b	12.6b,c	24.1e	4.8a	20.2d,e	16.3c,d
27	Ethyl acetate	78.8a	361.1e	151.0c	125.3b,c	96.7a,b	231.8d	89.4a,b	360.1e	201.3d
28	Ethyl propanoate	nd	4.3a	33.8b	9.8a	2.6a	96.8c	nd	31.9b	117.7d
29	Propyl acetate	5.8a	3.9a	14.3a,b	6.0a	27.0b	158.2e	6.4a	44.7c	88.6d
30	1-Methylpropyl acetate	nd	nd	5.7a	nd	nd	34.1b	nd	nd	67.5c
31	3-Methylbutyl acetate	36.7f	10.4c	14.6d	6.7b	14.1d	4.1a	20.9e	5.5a,b	22.3e
32	Methyl hexanoate	22.8e	9.5c	11.5d	1.3a	1.0a	8.8c	nd	0.6a	5.4b
33	Ethyl hexanoate	110.6d	16.6a	50.5c	14.2a	25.7a,b	28.5b	45.8c	23.6a,b	45.7c
34	Hexyl acetate	nd	0.8a	0.9a	nd	2.6c	1.4b	nd	nd	0.6a
35	Ethyl (E)-3-hexenoate	nd	0.9a	2.5c	nd	1.2a,b	nd	nd	nd	1.8b
36	(Z)-3-Hexenyl acetate	nd	1.3a	3.8b	nd	3.2b	2.0a	nd	nd	5.6c
37	Methyl lactate	nd	2.4b	4.7c	nd	2.9b	7.8d	nd	1.4a	nd
38	Ethyl lactate	25.6a,b	46.7d	51.2d	23.5a,b	28.4b	64.5e	20.4a	23.5a,b	35.3c
39	Methyl octanoate	30.4d,e	10.7a,b	24.0c,d	33.4e	22.6c	83.9f	18.3c	17.3b,c	5.4a
40	Ethyl octanoate	104.1c	24.7a	26.5a	32.4a	63.0b	17.6a	33.8a	27.2a	12.2a
41	Methyl nonanoate	15.7f	6.6c,d,e	5.3b,c	7.9d,e	8.4e	6.0b,c,d	4.0a,b	2.3a	7.3c,d,e
42	Ethyl nonanoate	7.9b	1.5a	1.8a	1.1a	10.0c	1.4a	6.2b	1.3a	2.8a
43	Methyl decanoate	6.6e	2.5c,d	nd	0.7a,b	3.1d	1.2b	1.6b,c	1.3b	1.7b,c
44	Ethyl decanoate	3.1d	0.9a,b	nd	0.5a	1.9b,c	5.3e	2.8c,d	nd	nd
45	Methyl 4-decenoate	nd	nd	nd	nd	0.3	nd	nd	nd	nd
46	3-Methylbutyl octanoate	nd	nd	nd	nd	nd	nd	nd	nd	nd
47	Ethyl 9-decenoate	5.2b	1.0a	nd	nd	1.4a	nd	nd	nd	nd
48	Benzyl acetate	nd	1.0b	1.1b	nd	nd	1.1b	nd	0.6a	1.5c
49	Methyl dodecanoate	1.7d	0.7b	nd	0.9b,c	1.1c	0.4a	nd	0.9b,c	nd
50	2-Phenylethyl acetate	3.4d	1.1a	3.0c	0.9a	1.3a	2.2b	nd	1.2a	3.1c
51	Methyl hydrocinnamate	2.8d	0.6a	1.0b,c	0.5a	1.0b,c	1.1c	nd	0.6a	0.5a
52	Ethyl dodecanoate	2.6b	nd	nd	nd	0.7a	nd	nd	nd	nd
53	Ethyl hydrocinnamate	7.0c	2.8a,b	5.8b,c	2.1a	3.1a,b	2.9a,b	3.0a,b	2.3a	2.9a,b
	Total amount	476.4	534.9	398.0	278.6	335.7	785.4	257.3	566.6	664.2
	PHENOLS
54	Guaiacol	1.6b,c	0.8a,b	1.9c	nd	nd	nd	3.6d	2.1c	5.0e
55	4-Methylguaiacol	23.2b,c	15.7a,b	35.9d	12.3a	17.8a,b	31.1c,d	17.8a,b	17.4a,b	35.9d
56	4-ethylphenol	nd	nd	0.5b	nd	nd	0.3a	0.6b	1.0c	2.1d
	Total amount	24.8	16.5	38.4	12.3	17.8	31.4	22.0	20.5	43.0
	Total	1978.8	2911.9	2281.2	1896.2	1996.3	3895.8	1791.6	2949.3	3305.0
^aMean values (n = 4; two vessels analyzed in duplicate); The concentration of each compound was expressed as µg/ L; Data in the same row with different lowercase letters are significantly different (p < 0.05); nd = not detected. ^bC: 8% NaCl; T1: 4% NaCl-4% KCl; T2: 4% NaCl-3% KCl-1% CaCl₂. A: 2020/21.

Table 2

Concentration of Volatile Organic Compounds in brines with different composition in chloride salts, monitored during pilot-scale spontaneous fermentation of Spanish-style cv. Chalkidiki green olives in 2021/22 processing period (30, 60 and 120 days)
Code	Compounds^a	Brines^b
		CB			T1B			T2B
		30	60	120	30	60	120	30	60	120
	ACIDS
1	Acetic acid	168.5a,b	421.0e	364.8d	201.0b	304.0c	602.0g	154.8a	277.5c	558.9f
2	Propanoic acid	5.5a	26.1d	33.3e	4.5a	19.2b,c	22.3c,d	2.0a	16.2b	26.4d
3	2-Methylpropanoic acid	nd	1.7a	12.9c	nd	1.5a	9.4b	nd	1.9a	10.1b
4	Butanoic acid	3.4c	6.2e	6.1e	2.6b	4.6d	8.6g	2.1a	3.6c	7.0f
5	3-Methylbutanoic acid	6.1a,b	8.5c	28.0f	4.8a	7.6b,c	17.9d	4.4a	6.4a,b	20.4e
6	Hexanoic acid	24.2d	23.6c,d	22.0c,d	14.3a	24.3d	43.3e	15.2a,b	17.2a,b	19.5b,c
7	Octanoic acid	16.6d	6.0a	6.3a	13.3c	9.2b	14.6c,d	14.5c	7.4a,b	6.9a
8	Nonanoic acid	1.4b,c	0.6a	1.0a,b	0.4a	1.8c,d	2.1d	1.5c	1.8c,d	1.6c,d
	Total amount	225.4	493.8	474.3	241.0	372.2	720.2	194.4	332.1	651.3
	ALCOHOLS
9	Ethanol	372.5b	551.7c	225.3a	468.5c	514.1c	525.8c	474.2c	697.5d	278.0a,b
10	2-Butanol	nd	14.8b	nd	nd	nd	nd	nd	7.2a	nd
11	1-Propanol	nd	10.2b	nd	nd	nd	nd	nd	5.8a	nd
12	2-Methyl-1-propanol	16.7a	37.8b	33.4b	29.1b	34.3b	31.6b	27.7b	33.5b	31.9b
13	2-Methyl-1-butanol	5.1a	73.7e	19.4b	51.4d	102.3f	130.5g	13.6a,b	50.0d	35.9c
14	3-Methyl-1-butanol	524.0a	757.0d	534.2a,b	699.4c,d	780.7d	633.0b,c	494.7a	606.7b,c	540.7a,b
15	3-Methyl-3-buten-1-ol	nd	29.3b	nd	6.0a	1.6a	nd	nd	nd	nd
16	3-Methyl-2-buten-1-ol	nd	1.8b	nd	nd	nd	nd	nd	0.5a	nd
17	1-Hexanol	28.7b	35.5c	19.9a	29.5b	18.5a	30.2b	28.7b	19.1a	26.4b
18	(Z)-3-Hexen-1-ol	nd	12.9b	nd	18.1c	4.4a	23.1d	nd	6.1a	nd
19	2-Ethyl-1-hexanol	4.8e	0.8a	1.7b,c	2.2c	1.0a,b	1.1a,b	3.6d	1.5a,b,c	1.7b,c
20	1-Octanol	nd	3.2a	9.3c	nd	nd	12.5d	nd	4.0a	6.8b
21	1-Nonanol	6.6e	3.0b,c	0.5a	4.5c,d	1.9a,b	12.0f	4.3c,d	1.9a,b	6.0d,e
22	Benzyl alcohol	4.9b	7.2c,d	4.6b	6.8c	3.0a	8.67e	4.7b	3.2a	8.1d,e
23	Phenylethyl alcohol	36.3b,c	41.8c,d	45.9d,e	30.7b	20.3a	80.4f	29.3b	17.6a	53.0e
	Total amount	943.1	1580.7	894.1	1346.0	1482.1	1489.0	1080.8	1454.6	990.1
	CARBONYLS
24	Octanal	nd	1.0	nd	nd	nd	nd	nd	nd	nd
25	Benzaldehyde	5.1b	5.8b	nd	5.0b	1.4a	nd	38.5c	1.1a	nd
	Total amount	5.1	6.9	nd	5.0	1.4	nd	38.8	1.1	nd
	ESTERS
26	Methyl acetate	3.1a	7.3c,d,e	8.6d,e	11.7f	6..4c,d	11.6f	3.9a,b	5.9b,c	9.6e,f
27	Ethyl acetate	36.8b	111.8e	88.9d	44.7b	71.7c	149.6f	20.5a	150.3f	89.9d
28	Ethyl propanoate	nd	51.3c	54.6c	26.1b	83.9d	nd	nd	4.2a	27.9b
29	Propyl acetate	nd	2.5a	14.8b	nd	nd	nd	nd	nd	43.6c
30	1-Methylpropyl acetate	nd	nd	12.5	nd	nd	nd	nd	nd	nd
31	3-Methylbutyl acetate	66.1e	7.9a	21.0b	8.2a	16.8b	16.5b	28.1c	6.8a	43.7d
32	Methyl hexanoate	39.6g	8.7c	85.4h	3.0b	11.0d	14.2e	18.0f	0.6a	38.4g
33	Ethyl hexanoate	207.2d	53.7a,b	104.1c	181.6d	83.9b,c	163.6d	171.0d	31.0a	73.9b,c
34	Hexyl acetate	nd	0.9	nd	nd	nd	nd	nd	nd	nd
35	Ethyl (E)-3-hexenoate	nd	nd	nd	nd	nd	nd	nd	nd	nd
36	(Z)-3-Hexenyl acetate	nd	1.6	nd	nd	nd	nd	nd	nd	nd
37	Methyl lactate	nd	2.0b	nd	nd	1.9b	4.3c	nd	1.1a	4.0c
38	Ethyl lactate	11.6a,b	24.4c	74.5d	12.2a,b	20.7c	122.9f	10.3a	19.0b,c	114.3e
39	Methyl octanoate	174.1e	68.1b,c	53.0a,b	109.0d	39.4a,b	92.4c,d	194.2e	21.4a	65.7b,c
40	Ethyl octanoate	60.5a	138.2b	50.8a	166.5b	135.4b	120.6b	524.6c	59.2a	32.2a
41	Methyl nonanoate	22.4c	6.4a,b	12.2b	7.8a,b	12.3b	24.7c	7.9a,b	1.2a	9.6b
42	Ethyl nonanoate	68.1e	5.4a,b	6.2a,b	8.3b,c	19.7d	23.4d	12.7c	0.6a	4.5a,b
43	Methyl decanoate	26.1c	6.8a,b	9.9b	23.0c	6.0a,b	23.1c	27.8c	2.2a	10.1b
44	Ethyl decanoate	70.0d	6.0a,b	3.9a	42.7c	3.5a	12.6b	39.3c	2.7a	4.3a
45	Methyl 4-decenoate	7.4d	2.0b	nd	4.8c	2.4b	nd	7.2d	0.6a	nd
46	3-Methylbutyl octanoate	3.1c	nd	nd	0.9a	nd	nd	1.5b	nd	nd
47	Ethyl 9-decenoate	28.7e	3.3b	nd	15.8d	6.2c	nd	16.7d	2.3a,b	nd
48	Benzyl acetate	nd	0.4	nd	nd	nd	nd	nd	nd	nd
49	Methyl dodecanoate	nd	2.1b	nd	2.9c	1.9b	3.5c	nd	0.8a	nd
50	2-Phenylethyl acetate	10.4b	0.8a	nd	1.0a	nd	nd	1.0a	nd	nd
51	Methyl hydrocinnamate	0.3	nd	nd	nd	nd	nd	nd	nd	nd
52	Ethyl dodecanoate	6.7c	0.4a	nd	2.9b	0.9a	nd	nd	0.7a	nd
53	Ethyl hydrocinnamate	1.0c	1.7d	nd	0.6a,b	0.5a	nd	0.9b,c	nd	nd
	Total amount	816.2	513.9	600.4	673.6	524.3	783.0	1085.7	310.6	572.2
	PHENOLS
54	Guaiacol	23.3c	0.7a,b	nd	1.3b	nd	nd	nd	nd	nd
55	4-Methylguaiacol	3.9a	6.9b	2.8a	21.0d	2.2a	12.4c	3.1a	3.0a	14.3c
56	4-ethylphenol	nd	nd	nd	nd	nd	nd	nd	nd	nd
	Total amount	27.2	7.6	2.8	22.3	2.2	12.4	3.1	3.0	14.3
	Total	2102.5	2602.9	1971.7	2288.7	2382.6	3004.6	2403.8	2101.3	2225.6
^aMean values (n = 4; two vessels analyzed in duplicate); The concentration of each compound was expressed as µg/ L; Data in the same row with different lowercase letters are significantly different (p < 0.05); nd = not detected. ^bC: 8% NaCl; T1: 4% NaCl-4% KCl; T2: 4% NaCl-3% KCl-1% CaCl₂. B: 2021/22.

In both processing periods, brines with 50% lower NaCl content (T1, T2) were found enriched in total VOCs, compared to those with high NaCl content (8%) (C). In particular, the total VOCs’ content of T1 and T2 brines presented an increasing trend until the end of the LAB-dominated fermentation process (3895.8 µg/L and 3305.0 µg/L for T1 and T2, respectively) in period A. Whereas for C, the VOCs’ content increased up to 60 days of fermentation (2911.9 µg/L and 2602.9 µg/L for period A and B, respectively) and decreased afterward, probably due to microbial metabolism and decomposition.

After evaluating the changes in the relative content (%) of the VOCs according to their chemical class (Fig. 3), it was revealed that irrespective the fermentation stage, brines of period A were dominated by acids (~ 38–50%), alcohols (~ 31–43%) and esters (~ 14–24%), while brines of period B were dominated by alcohols (~ 44–69%) followed by esters (~ 15–45%) and lower levels of acids (~ 8–29%). It can be suggested that, in period B brines, alcoholic fermentation is mainly responsible for providing high content of alcohols, as a result of yeast dominance and metabolism (Sabatini & Marsilio, 2008). Phenols and carbonyl compounds were present at quite lower levels (< 2% each) at both processing periods.

A more-detailed inspection of the compositional changes in specific volatiles (Tables 1 and 2) can reveal that, regardless of the processing period and stage of fermentation, acetic acid (1) was the most abundant among the 8 identified volatile acids, representing ~ 75–94% of the total acid content. Acetic acid is a typical metabolite in Spanish-style green olive brines. Its formation has been related to lye treatment, most likely from the fragmentation of sugars found in olive pulp, and fermentation via different pathways in yeasts and bacteria, e.g., oxidation of ethanol or heterolactic fermentation (Blana et al., 2014; Sánchez et al., 2000). In addition, propanoic acid (2) was the second most important acid only in period A, with T brines containing significantly higher amounts than those in the C brines at the late stage of fermentation (Table 1). However, propanoic acid represented 17.4–19.4% relative to the sum of acetic and propanoic acids contents in T brines, slightly higher than in C brines (10.9%), but still scarce according to the scale suggested by Rejano et al. (1978), indicating that growth of Propionibacterium species (characteristic of the “fourth stage” of fermentation) was insignificant.

On the contrary, the profile of alcohols (15 identified compounds) differed significantly between the two processing periods. In particular, in period A brines, a complex profile that constituted mainly of ethanol (9), 3-methyl-1-butanol (14) and 2-butanol (10), along with 1-propanol (11), phenylethyl alcohol (23), (Z)-3-hexen-1-ol (18) and 1-hexanol (17) was found. On the other hand, volatile alcohols in period B brines constituted mostly of ethanol (9) and 3-methyl-1-butanol (14) (~ 78–90% of the total alcohol content). Variation in the level of individual alcohols is strongly related to the microbial activity and evolution of LAB and yeasts during spontaneous fermentation in brines. Ethanol is formed during table olive processing due to alcoholic and to a lesser extent, heterolactic fermentation (Sabatini & Marsilio, 2008). However, diffusion from lye treated olives may also contribute to its total content in brines (de Castro et al., 2019; A. H. Sánchez et al., 2000). 3-Methyl-1-butanol derives from the metabolic activity of yeasts and more specifically through the decarboxylation and deamination of the amino acid L-leucine in the Ehrlich pathway (De Angelis et al., 2015; Pires et al., 2014). These results are supported by microbiological analyses, which showed the proliferation of yeasts in period A (Fig. 1). Other researchers also reported higher contents of ethanol and 3-methyl-1-butanol in brines when yeast dominated the natural fermentation of Greek olives cv. Conservolea and Kalamata (Bleve et al., 2015), or Italian olives cv. Cellina di Nardò and Leccino (Bleve et al., 2014) and Sicilian green table olives cv. Nocellara Etnea (Randazzo et al., 2017). Besides, similar results were obtained in yeast-inoculated fermentations compared to spontaneous ones in brines of Spanish-style olives cv. Manzanilla (Benítez-Cabello et al., 2019) and in table olives from cv. Picual, Manzanilla and Kalamata (Tufariello et al., 2019). 1-Propanol (11) and phenylethyl alcohol (23) are considered products of yeast fermentation during Spanish-style table olive processing (Cortés-Delgado et al., 2016; Sabatini et al., 2009). The apparently higher level of 2-butanol (10) in period A brines strongly indicates that important microbial activity, in addition to or instead of LAB, occurred in these brines (Fleming et al., 1969). Increased contents of 2-butanol were also evidenced in Spanish-style cv. Nocellara del Belice table olives (Martorana et al., 2017) and in brines of Spanish-style olives cv. Moresca and Kalamata (Benítez-Cabello et al., 2019) during inoculated fermentation processes using selected LAB strains. 1-Hexanol (17) and (Z)-3-hexen-1-ol (18) can be formed enzymatically from polyunsaturated fatty acids either by the endogenous lipoxygenases or via a similar lipoxygenase-like metabolism, affected by microbial enzymes (Sabatini & Marsilio, 2008).

Esters were the largest chemical class consisting almost half of the total number of identified compounds (n = 28). Regardless of the processing period, ethyl and acetate esters were the most abundant, following the same trend as their alcohol and acid precursors, respectively. It is generally accepted that ethyl esters are biosynthesized via enzyme-catalyzed condensation reactions between fatty acids and ethanol, whereas the acetate forms derive from acetic acid and ethanol or higher alcohols (Pires et al., 2014; Sabatini & Marsilio, 2008). Many factors, e.g., pH, microbial and enzyme activity, NaCl content and precursors availability affect ester synthesis during fermentation process (Cristiani & Monnet, 2001; Mikrou et al., 2021). In this study, ethyl acetate (27), ethyl propanoate (28), propyl acetate (29), ethyl hexanoate (33) and ethyl octanoate (40) were the primary esters in period A brines. Ethyl acetate (27), ethyl hexanoate (33), ethyl lactate (38), methyl octanoate (39) and ethyl octanoate (40) dominated the ester profile of period B brines.

Considering the minor volatile phenols and carbonyl compounds, special emphasis was given to benzaldehyde (25) and 4-methylguaiacol (55) because of slightly higher relative contents in brines of the LAB-dominated fermentation process in period A. Benzaldehyde may be formed during fruit ripening and diffused in brines, and/or produced during the fermentation step by LAB species (Cortés-Delgado et al., 2016). As 4-methylguaiacol could represent a metabolic product of phenolic acids via LAB, e.g., L. plantarum (Cortés-Delgado et al., 2016), its content variation from one period to the other can be attributed to different phenolic dynamics of raw olive fruits. Indeed, the total polar phenol content of the drupes (as measured by Folin-Ciocalteu, data not shown) in period A was found almost two-fold higher than that in the next processing period. A large content variation is expected in their polar phenolic fraction, with oleuropein, the major representative constituent, ranging between 400 and 21,000 mg/kg (Mastralexi et al., 2019).

To get a deeper view of those complex volatile profile changes in the brines that may denote the effects of different periods, combinations of chloride salts and/or fermentation stages, the results were analyzed with the aid of multivariate data analysis and explorative PCA, as elaborated in the next sections.

3.4. Principal Components Analysis

PCA was performed on the processed SPME-GC-MS data set (n = 36) to evaluate possible similarity patterns between the chemical data and those from microbiological analysis, as well as to identify possible markers of different pilot-scale fermentation processes. The results were evaluated taken the ratio of LAB to yeasts population in the brines under study.

The multivariate data analysis (56 variables x 36 brines) resulted in four principal components (PCs) that explain 95.3% of the total variance. The PCA scatterplot of the first two PCs (t1/t2 scores; Fig. 4a) clearly showed that brines were grouped separately across the PC1 axis according to processing period. Thus, period A brines tended to segregate at the right part of the plot with positive t1 values, in contrast to those of the next processing period that were more scattered and located in the left part of the plot with negative t1 values. It can be suggested that PC1 captures the seasonal variation in the population of the microbiota and compensates all the diagnostic information (80.0% of the variability) about the effects on the brines volatilome. Noticeably, all samples that were allocated with t[1] < 0 were characterized by relatively lower LAB/yeasts ratio (≤ 1). This means that distinct VOCs profiles were obtained when LAB were overpowered by yeasts, as discussed earlier in the text (Fig. 1).

An insight into the p1/p2 loadings plot of the corresponding PCA model (Fig. 4b) showed that apart from acetic acid (1), 3-methyl-1-butanol (14) and to a lesser extent ethanol (9) were the most important VOCs for the distinct grouping of brines from different processing periods across PC1. It could be that the overwhelming contribution of acetic acid in the brines of C and T processes (Tables 1 and 2) was mediated when yeasts dominated the microflora (i.e., period B) mainly due to enriched content of 3-methyl-1-butanol (Table 2). PCA results highlighted the interplay of those microbial metabolites as the most significant marker for microbiota profile changes in brines between A and B periods, regardless the brine composition in chloride salts or the fermentation stage. Hence, 3-methyl-1-butanol can be suggested as indicator of yeast prevalence during fermentation of Spanish-style cv. Chalkidiki green table olives, likewise its role in the spontaneous fermentation cv. Cellina di Nardò and Leccino table olives (Bleve et al., 2014). Hetero-fermentative LAB metabolism, mainly at the late stage of fermentation was probably the reason why acetic acid was found the major discriminant VOC of brines from the LAB-dominated processing period A (Table 1) (Bleve et al., 2014).

Other scattering patterns according to fermentation stage and chloride salts composition were exposed on PC3 and PC4 (Figure S1). The former brought information for period B brines, while the latter was more relevant with changes observed in period A brines.

To explore further these patterns, new rounds of PCA were performed independently on the two datasets for brines from the different processing periods. The data from period A were decomposed to 5 PCs that explained 95.5% of the total original variance while the data from year B resulted in a simpler model of 4 PCs that accounted for 96.2% of the total variance.

3.4.1. Evolution of VOCs profile during 2020/21 period

C and T (T1 and T2) brines of period A, at different fermentation stages (30, 60 and 120 days), were distributed on the PC1/PC2 (a) and PC1/PC3 (c) planes (Fig. 5a, c). At a first look, it seems that scattering across the PC1 axis patterns changes of VOCs during fermentation. PC2 and PC3 axes brought some information about the different VOCs profile of T brines (T1 and T2), compared with the C ones.

Specifically, C brines tended to segregate to the lower part of the PCA score plot with almost zero or negative t2 values (Fig. 5a, c). Variation in the contents of ethanol (9) and 3-methyl-1-butanol (14) along with certain ethyl esters, e.g., ethyl acetate (27) contributed highly to the different allocation of the C brines at the early and middle stages of the process (Fig. 5b, d), compared to the late stage. This finding suggests that the particular yeast-derived VOCs could be important markers of spontaneous fermentation of Spanish-style cv. Chalkidiki green olives in traditional, high NaCl brines. Differentiation among low NaCl (T1 and T2) and high NaCl brines (C) across PC2 was more pronounced at the late stage of fermentation (120 days). The p1/p2 loadings (Fig. 5b) signify that variance in propanoic acid (2) and 3-methyl-1-butanol (14) contents was the main reason behind the differential distribution of those low NaCl brines.

Further insight into the evolution of volatile profiles of C, T1 and T2 fermentation processes was attained by evaluating PCA results at 30 (Figure S2), 60 (Figure S3) and 120 days (Figure S4), separately. Within the T cluster, two sub-clusters were formed across PC2, according to the chloride salts composition (Figure S4a). Above all, 1-propanol (11) was more important for separate clustering of T1 at 120 days, while 2-butanol (10) accounted for separate grouping of T2 throughout the whole fermentation period (Figures S2b, S3b and S4b). Although the growth dynamics of the yeast population (Fig. 1) was not noticeably altered due to the different chloride salts composition, the former metabolite may be related to the changes in the diversity of species (Mateus et al., 2016). Noticeably, the principal role of 2-butanol (10) in further discriminating T2 brines (Table 1) can be associated with the presence of CaCl₂ that brings about slower alkali diffusion with concomitant faster acidification (Figs. 1 and 2) and, thus, easier LAB growth in those brines (Mantzouridou et al., 2020).

3.4.2. Evolution of VOCs profile during 2021/22 period

The next processing period, brines were also distributed over the PC1/PC2 plane, but more distinctively over different stages (30, 60 and 120 days) of fermentation, regardless of the differences in chloride salts composition (Fig. 6a). Ethyl octanoate (40) at 30 days, ethanol (9) and 3-methyl-1-butanol (14) at 60 days and, acetic acid (1) at 120 days, contributed greatly to the formation of the respective clusters (Fig. 6a, b). These results signify that yeasts were responsible for the early and middle stages of fermentation, while hetero-fermentation by LAB is probably evidenced at the late stage (Fig. 1).

In the early and middle stages of the yeast-dominated fermentation processes, C and T1 brines tended to group separately from those of T2 across PC3 (Fig. 6c). During that period, the volatile profiles of T2 brines evolved differently, probably due to the presence of CaCl₂. Nevertheless, both low NaCl conditions led to brines with distinct characteristics in comparison with C at the late stage of fermentation. C brines were distinguished from T1 and T2, owing to variance in 3-methyl-1-butanol (14) combined with those in ethanol (9) and acetic acid (1), respectively (Figs. 6d and S7). This is in line with observations of the preceding period emphasizing the particular yeast-derived metabolite (3-methyl-1-butanol) as diagnostic for high NaCl brines.

Overall and despite the complexity of VOCs patterns throughout the fermentation processes, volatile constituents associated with yeast (T1; 1-propanol in period A; ethanol in B) and LAB metabolism (T2; 2-butanol in A; acetic acid in B) indicated the effects of changes in the chloride salts composition of brines. This is reported for the first time for pilot-scale spontaneous fermentation of Spanish-style cv. Chalkidiki green olives under low NaCl conditions that delivered safe and organoleptically acceptable products.

In this study, data obtained during pilot-scale spontaneous fermentation of Spanish-style cv. Chalkidiki green olives under high and low NaCl conditions are a follow-up of the observations made in laboratory scale. Exploration of the correlation patterns of VOCs instead of comparing absolute contents of brines throughout the fermentations helped to draw some conclusive remarks about the effect of low NaCl processing and microbial composition. Overall, the evolution of the brines volatilome with emphasis on propanoic acid and 2-butanol is intriguing for monitoring a LAB-dominated spontaneous fermentation of Spanish-style table olives under low NaCl conditions. A sharp increase in acetic acid at a later stage of fermentation seems typical in low NaCl brines, indicating hetero-fermentative LAB metabolism. On the other hand, the role of 3-methyl-1-butanol was highlighted and suggests this molecule to be a promising candidate as marker of yeast dominance, as well as high NaCl processing. Further work is currently impending for the in-depth investigation of changes in microbial diversity during spontaneous fermentation of Spanish-style green table olives cv. Chalkidiki under high and low NaCl conditions, in order to unravel possible links between evolution of VOCs and microbial taxa.

Conflicts of Interest: The authors declare that they have no conflict of interest.

Authors’ Contributions

Conceptualization, Fani Th. Mantzouridou; methodology, Fani Th. Mantzouridou, Panagiota Alvanoudi, Andreana N. Assimopoulou, Alexandros Nakas, Stella A. Ordoudi; Investigation, Panagiota Alvanoudi, Alexandros Nakas, Stella A. Ordoudi, Fani Th. Mantzouridou and Andreana N. Assimopoulou; data curation, Fani Th. Mantzouridou, Andreana N. Assimopoulou, Stella A. Ordoudi, Panagiota Alvanoudi, Alexandros Nakas; writing—original draft preparation, Panagiota Alvanoudi; writing—review and editing, Fani Th. Mantzouridou, Stella A. Ordoudi, Andreana N. Assimopoulou and Alexandros Nakas; supervision, Fani Th. Mantzouridou; project administration, Fani Th. Mantzouridou; funding acquisition, Fani Th. Mantzouridou. All authors have read and agreed to the published version of the manuscript.

Funding: This research was carried out as part of the project «FILELIA – Development of edible olives friendly to a salt-reduced diet» (Project code: ΚΜΡ6-0079456) under the framework of the Action «Investment Plans of Innovation» of the Operational Program «Central Macedonia 2014-2020», that is co-funded by the European Regional Development Fund and Greece”.

Acknowledgments: Fani Mantzouridou sincerely thanks ATHOS OLIVE SA (Ormylia, Chalkidiki, Greece) for cooperation and support to this work.

Data Availability Statement

The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.

Alfonzo, A., Martorana, A., Settanni, L., Matraxia, M., Corona, O., Vagnoli, P., Caruso, T., Moschetti, G., & Francesca, N. (2018). Approaches to improve the growth of the starter lactic acid bacterium OM13 during the early stages of green Spanish-style table olive production. Grasas y Aceites, 69(3), https://doi.org/10.3989/gya.0103181.
Anagnostopoulos, D. A., & Tsaltas, D. (2022). Current Status, Recent Advances, and Main Challenges on Table Olive Fermentation: The Present Meets the Future. In Frontiers in Microbiology (Vol. 12). Frontiers Media S.A. https://doi.org/10.3389/fmicb.2021.797295.
Arroyo-López, F. N., Romero-Gil, V., Bautista-Gallego, J., Rodríguez-Gómez, F., Jiménez-Díaz, R., García-García, P., Querol, A., & Garrido-Fernández, A. (2012). Yeasts in table olive processing: Desirable or spoilage microorganisms? International Journal of Food Microbiology (Vol, 160(1), 42–49. https://doi.org/10.1016/j.ijfoodmicro.2012.08.003.
Bautista-Gallego, J., Arroyo-López, F. N., Romero-Gil, V., Rodríguez-Gómez, F., García-García, P., & Garrido-Fernández, A. (2011). Chloride salt mixtures affect Gordal cv. green Spanish-style table olive fermentation. Food Microbiology, 28(7), 1316–1325. https://doi.org/10.1016/j.fm.2011.06.002.
Bautista-Gallego, J., Arroyo-López, F. N., Romero-Gil, V., Rodríguez-Gómez, F., García-García, P., & Garrido-Fernández, A. (2015). Fermentation profile of green Spanish-style Manzanilla olives according to NaCl content in brine. Food Microbiology, 49, 56–64. https://doi.org/10.1016/j.fm.2015.01.012.
Benítez-Cabello, A., Rodríguez-Gómez, F., Morales, M. L., Garrido-Fernández, A., Jiménez-Díaz, R., & Arroyo-López, F. N. (2019). Lactic acid bacteria and yeast inocula modulate the volatile profile of Spanish-style green table olive fermentations. Foods, 8(8), https://doi.org/10.3390/foods8080280.
Blana, V. A., Grounta, A., Tassou, C. C., Nychas, G. J. E., & Panagou, E. Z. (2014). Inoculated fermentation of green olives with potential probiotic Lactobacillus pentosus and Lactobacillus plantarum starter cultures isolated from industrially fermented olives. Food Microbiology, 38, 208–218. https://doi.org/10.1016/j.fm.2013.09.007.
Bleve, G., Tufariello, M., Durante, M., Grieco, F., Ramires, F. A., Mita, G., Tasioula-Margari, M., & Logrieco, A. F. (2015). Physico-chemical characterization of natural fermentation process of Conservolea and Kalamàta table olives and developement of a protocol for the pre-selection of fermentation starters. Food Microbiology, 46, 368–382. https://doi.org/10.1016/j.fm.2014.08.021.
Bleve, G., Tufariello, M., Durante, M., Perbellini, E., Ramires, F. A., Grieco, F., Cappello, M. S., de Domenico, S., Mita, G., Tasioula-Margari, M., & Logrieco, A. F. (2014). Physico-chemical and microbiological characterization of spontaneous fermentation of Cellina di Nardò and Leccino table olives. Frontiers in Microbiology, 5(OCT), https://doi.org/10.3389/fmicb.2014.00570.
Cano-Lamadrid, M., Girón, I. F., Pleite, R., Burló, F., Corell, M., Moriana, A., & Carbonell-Barrachina, A. A. (2015). Quality attributes of table olives as affected by regulated deficit irrigation. LWT, 62(1), 19–26. https://doi.org/10.1016/j.lwt.2014.12.063.
Cortés-Delgado, A., Sánchez, A. H., de Castro, A., López-López, A., Beato, V. M., & Montaño, A. (2016). Volatile profile of Spanish-style green table olives prepared from different cultivars grown at different locations. Food Research International, 83, 131–142. https://doi.org/10.1016/j.foodres.2016.03.005.
Cristiani, G., & Monnet, V. (2001). Food micro-organisms and aromatic ester synthesis. In SCIENCES DES ALIMENTS (Vol. 21).
De Angelis, M., Campanella, D., Cosmai, L., Summo, C., Rizzello, C. G., & Caponio, F. (2015). Microbiota and metabolome of un-started and started Greek-type fermentation of Bella di Cerignola table olives. Food Microbiology, 52, 18–30. https://doi.org/10.1016/j.fm.2015.06.002.
De Bellis, P., Valerio, F., Sisto, A., Lonigro, S. L., & Lavermicocca, P. (2010). Probiotic table olives: Microbial populations adhering on olive surface in fermentation sets inoculated with the probiotic strain Lactobacillus paracasei IMPC2.1 in an industrial plant. International Journal of Food Microbiology, 140(1), 6–13. https://doi.org/10.1016/j.ijfoodmicro.2010.02.024.
de Castro, A., Sánchez, A. H., Cortés-Delgado, A., López-López, A., & Montaño, A. (2019). Effect of Spanish-style processing steps and inoculation with Lactobacillus pentosus starter culture on the volatile composition of cv. Manzanilla green olives. Food Chemistry, 271, 543–549. https://doi.org/10.1016/j.foodchem.2018.07.166.
de Castro, A., Sánchez, A. H., López-López, A., Cortés-Delgado, A., Medina, E., & Montaño, A. (2018). Microbiota and metabolite profiling of spoiled Spanish-style green table olives. Metabolites, 8(4), https://doi.org/10.3390/metabo8040073.
DOEPEL, Interprofessional Association for Table Olives (2023). Available online: http://www.doepel.gr (accessed June 11, 2023).
EU. Commission Implementing Regulation (EU) No 426/2012 of 22 May 2012 entering a name in the register of protected designations of origin and protected geographical indications (Πράσινες Ελιές Χαλκιδικής (Prasines Elies Chalkidikis) (PDO)). Off J Eur Union, L132, 1–7.
EU (2011). Commission Regulation (EU) No 1129/2011 of 11 November 2011 amending Annex II to Regulation (EC) No 1333/2008 of the European Parliament and of the Council by establishing a Union list of food additives. Off. J. Eur. Comm., L 295, 1.
Fleming, H. P., Etchells, J. L., & Bell, T. A. (1969). Vapor analysis of fermented Spanish-type green olives by gas chromatography. Journal of Food Science, 34(5), 419–422. https://doi.org/10.1111/j.1365-2621.1969.tb12794.x.
Garrido-Fernández, A., Benítez-Cabello, A., Rodríguez-Gómez, F., Jiménez-Díaz, R., Arroyo-López, F. N., & Morales, M. L. (2021a). Relating starter cultures to volatile profile and potential markers in green Spanish-style table olives by compositional data analysis. Food Microbiology, 94. https://doi.org/10.1016/j.fm.2020.103659.
Garrido-Fernández, A., Montaño, A., Cortés-Delgado, A., Rodríguez-Gómez, F., & Arroyo-López, F. N. (2021b). Multi-statistical approach for the study of volatile compounds of industrial spoiled manzanilla Spanish-style table olive fermentations. Foods, 10(6), https://doi.org/10.3390/foods10061182.
Garrido-Fernández, A., Montaño, A., Sánchez-Gómez, A. H., Cortés-Delgado, A., & López-López, A. (2017). Volatile profiles of green Spanish-style table olives: Application of compositional data analysis for the segregation of their cultivars and production areas. Talanta, 169, 77–84. https://doi.org/10.1016/j.talanta.2017.03.066.
International Olive Council (IOC) (2023). Economic Affairs & Promotion Unit. Available online: https://www.internationaloliveoil.org/what-we-do/economic-affairs-promotion-unit/ (accessed June 11, 2023).
López-López, A., Sánchez, A. H., Cortés-Delgado, A., de Castro, A., & Montaño, A. (2018). Relating sensory analysis with SPME-GC-MS data for Spanish-style green table olive aroma profiling. LWT, 89, 725–734. https://doi.org/10.1016/j.lwt.2017.11.058.
Mantzouridou, F. T., Mastralexi, A., Filippidou, M., & Tsimidou, M. Z. (2020). Challenges in the Processing Line of Spanish-style cv. Chalkidiki Green Table Olives Spontaneously Fermented in Reduced NaCl Content Brines. European Journal of Lipid Science and Technology, 122(4), https://doi.org/10.1002/ejlt.201900453.
Martorana, A., Alfonzo, A., Gaglio, R., Settanni, L., Corona, O., La Croce, F., Vagnoli, P., Caruso, T., Moschetti, G., & Francesca, N. (2017). Evaluation of different conditions to enhance the performances of Lactobacillus pentosus OM13 during industrial production of Spanish-style table olives. Food Microbiology, 61, 150–158. https://doi.org/10.1016/j.fm.2016.08.007.
Marx, Í. M., Rodrigues, N., Dias, L. G., Veloso, A. C., Pereira, J. A., Drunkler, D. A., & Peres, A. M. (2017). Assessment of Table Olives’ Organoleptic Defect Intensities Based on the Potentiometric Fingerprint Recorded by an Electronic Tongue. Food And Bioprocess Technology, 10, 1310–1323. https://doi.org/10.1007/s11947-017-1902-7.
Mastralexi, A., Mantzouridou, F. T., & Tsimidou, M. Z. (2019). Evolution of Safety and Other Quality Parameters of the Greek PDO Table Olives “Prasines Elies Chalkidikis” During Industrial Scale Processing and Storage. European Journal of Lipid Science and Technology, 121(3), https://doi.org/10.1002/ejlt.201800171.
Mateus, T., Santo, D., Saúde, C., Pires-Cabral, P., & Quintas, C. (2016). The effect of NaCl reduction in the microbiological quality of cracked green table olives of the Maçanilha Algarvia cultivar. International Journal of Food Microbiology, 218, 57–65. https://doi.org/10.1016/j.ijfoodmicro.2015.11.008.
Mikrou, T., Kasimati, K., Doufexi, I., Kapsokefalou, M., Gardeli, C., & Mallouchos, A. (2021). Volatile composition of industrially fermented table olives from Greece. Foods, 10(5), https://doi.org/10.3390/foods10051000.
Montaño, A., Sánchez, A. H., Casado, F. J., De Castro, A., & Rejano, L. (2003). Chemical profile of industrially fermented green olives of different varieties. Food Chemistry, 82(2), 297–302. https://doi.org/10.1016/S0308-8146(02)00593-9.
Nanou, A., Mallouchos, A., & Panagou, E. Z. (2021). Elucidation of the Volatilome of Packaged Spanish-Style Green Olives of Conservolea and Halkidiki Varieties Using SPME-GC/MS. Proceedings. 75. https://doi.org/10.3390/foods_2020-07629.
Panagou, E. Z., & Tassou, C. C. (2006). Changes in volatile compounds and related biochemical profile during controlled fermentation of cv. Conservolea green olives. Food Microbiology, 23(8), 738–746. https://doi.org/10.1016/j.fm.2006.02.005.
Pino, A., De Angelis, M. D., Todaro, A., Van Hoorde, K. V., Randazzo, C. L., & Caggia, C. (2018). Fermentation of Nocellara Etnea table olives by functional starter cultures at different low salt concentrations. Frontiers in Microbiology, 9(JUN), https://doi.org/10.3389/fmicb.2018.01125.
Pino, A., Vaccalluzzo, A., Solieri, L., Romeo, F. V., Todaro, A., Caggia, C., Arroyo-López, F. N., Bautista-Gallego, J., & Randazzo, C. L. (2019). Effect of sequential inoculum of beta-glucosidase positive and probiotic strains on brine fermentation to obtain low salt sicilian table olives. Frontiers in Microbiology, 10(FEB), https://doi.org/10.3389/fmicb.2019.00174.
Pires, E. J., Teixeira, J. A., Brányik, T., & Vicente, A. A. (2014). Yeast: The soul of beer’s aroma - A review of flavour-active esters and higher alcohols produced by the brewing yeast. In Applied Microbiology and Biotechnology (Vol. 98, Issue 5, pp. 1937–1949). https://doi.org/10.1007/s00253-013-5470-0.
Randazzo, C. L., Todaro, A., Pino, A., Pitino, I., Corona, O., & Caggia, C. (2017). Microbiota and metabolome during controlled and spontaneous fermentation of Nocellara Etnea table olives. Food Microbiology, 65, 136–148. https://doi.org/10.1016/j.fm.2017.01.022.
Rejano, L., Cancho, F. G., & Borbolla, J. M. R. (1978). La formacion de acido propionico durante la conservacion de las aceitunas verdes de mesa. II. Grasas y Aceites, 29, 203–210.
Ruiz-Barba, J. L., Sánchez, A. H., López-López, A., Cortés-Delgado, A., & Montaño, A. (2023). Microbial community and volatilome changes in brines along the spontaneous fermentation of Spanish-style and natural-style green table olives (Manzanilla cultivar). Food Microbiology, 113. https://doi.org/10.1016/j.fm.2023.104286.
Sabatini, N., & Marsilio, V. (2008). Volatile compounds in table olives (Olea Europaea L., Nocellara del Belice cultivar). Food Chemistry, 107(4), 1522–1528. https://doi.org/10.1016/j.foodchem.2007.10.008.
Sabatini, N., Perri, E., & Marsilio, V. (2009). An investigation on molecular partition of aroma compounds in fruit matrix and brine medium of fermented table olives. Innovative Food Science and Emerging Technologies, 10(4), 621–626. https://doi.org/10.1016/j.ifset.2009.05.001.
Sánchez, A. H., de Castro, A., López-López, A., Cortés-Delgado, A., Beato, V. M., & Montaño, A. (2017). Retention of color and volatile compounds of Spanish-style green table olives pasteurized and stored in plastic containers under conditions of constant temperature. LWT, 75, 685–691. https://doi.org/10.1016/j.lwt.2016.10.027.
Sánchez, A. H., De Castro, A., Rejano, L., & Montaño, A. (2000). Comparative study on chemical changes in olive juice and brine during green olive fermentation. Journal of Agricultural and Food Chemistry, 48(12), 5975–5980. https://doi.org/10.1021/jf000563u.
Sánchez, A. H., López-López, A., Cortés-Delgado, A., Beato, V. M., Medina, E., de Castro, A., & Montaño, A. (2018). Effect of post-fermentation and packing stages on the volatile composition of Spanish-style green table olives. Food Chemistry, 239, 343–353. https://doi.org/10.1016/j.foodchem.2017.06.125.
Sánchez, R., Pérez-Nevado, F., Martillanes, S., Montero-Fernández, I., Lozano, J., & Martín-Vertedor, D. (2023). Machine olfaction discrimination of Spanish-style green olives inoculated with spoilage mold species. Food Control, 147. https://doi.org/10.1016/j.foodcont.2022.109600.
Sánchez-Rodríguez, L., Cano-Lamadrid, M., Carbonell-Barrachina, Á. A., Sendra, E., & Hernández, F. (2019). Volatile composition, sensory profile and consumer acceptability of hydrosostainable table olives. Foods, 8(10), https://doi.org/10.3390/foods8100470.
Tufariello, M., Anglana, C., Crupi, P., Virtuosi, I., Fiume, P., Di Terlizzi, B., Moselhy, N., Attay, H. A. G., Pati, S., Logrieco, A. F., Mita, G., & Bleve, G. (2019). Efficacy of yeast starters to drive and improve Picual, Manzanilla and Kalamàta table olive fermentation. Journal of the Science of Food and Agriculture, 99(5), 2504–2512. https://doi.org/10.1002/jsfa.9460.
Vaccalluzzo, A., Celano, G., Pino, A., Calabrese, F. M., Foti, P., Caggia, C., & Randazzo, C. (2022). Metagenetic and Volatilomic Approaches to Elucidate the Effect of Lactiplantibacillus plantarum Starter Cultures on Sicilian Table Olives. Frontiers in Microbiology, 12. https://doi.org/10.3389/fmicb.2021.771636.
Vaccalluzzo, A., Pino, A., Russo, N., De Angelis, M., Caggia, C., & Randazzo, C. L. (2020). FoodOmics as a new frontier to reveal microbial community and metabolic processes occurring on table olives fermentation. In Food Microbiology (Vol. 92). Academic Press. https://doi.org/10.1016/j.fm.2020.103606.
World Health Organization (WHO). (2012). Guideline: Sodium intake for adults and children. Switzerland: Geneva.

No competing interests reported.

Download PDF

Journal Publication

published 21 Sep, 2023

Read the published version in Food and Bioprocess Technology →

Editorial decision: Major revision
30 Jun, 2023
Editor assigned by journal
30 Jun, 2023
Submission checks completed at journal
29 Jun, 2023
First submitted to journal
29 Jun, 2023

You are reading this latest preprint version

Volatilome changes in brines along the spontaneous fermentation of Spanish-style cv. Chalkidiki green olives under high and low NaCl conditions

Status:

Journal Publication

Version 1

Abstract

Figures

1. Introduction

2. Material and Methods

2.1. Materials

2.2. Olive samples and processing

2.3. Microbiological and physicochemical analyses of brines

2.4. VOCs determination in brines

2.4.1. Headspace Solid-Phase Microextraction (HS-SPME)

2.4.2. Gas Chromatography-Mass Spectrometry analysis (GC-MS)

2.5. Data processing and statistical analysis

3. Results and Discussion

3.1. Microflora composition of the brine

3.2. Physicochemical characteristics of brines

3.3. Analysis of Volatile Organic Compounds in brines

3.4. Principal Components Analysis

3.4.1. Evolution of VOCs profile during 2020/21 period

3.4.2. Evolution of VOCs profile during 2021/22 period

4. Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1