Journal of the Science of Food and Agriculture
J Sci Food Agric 85:1084–1090 (2005) DOI: 10.1002/jsfa.2066
Use of a lactic acid bacteria starter culture during green olive (Olea europaea L cv Ascolana tenera) processing Vincenzo Marsilio,1∗ Leonardo Seghetti,2 Emilia Iannucci,1 Francesca Russi,1 Barbara Lanza1 and Marino Felicioni2 1 Istituto 2 Istituto
Sperimentale per la Elaiotecnica, C da Fonte Umano 37, I-65013 Citta` S Angelo (PE), Italy Tecnico Agrario Statale ‘Celso Ulpiani’, Via della Repubblica 30, I-63100 Ascoli Piceno, Italy
Abstract: Among the Italian olive germplasm, ‘Ascolana tenera’ is one of the best varieties for table olive production. This research addressed the impact of different processing types (Greek-style and Spanish-style) on the fermentation and phenolic composition of olive fruit. In particular, the effects of a lactic acid bacteria (LAB) starter culture on the fermentation of naturally green olives processed according to the traditional Greek method were studied. Results revealed that Spanish-style processing produced a dramatic loss of total phenolics, while natural olive processing favoured a higher retention of biophenols. Oleoside 11-methylester, a phenol-related compound, and hydroxytyrosol, tyrosol, vanillic acid, 3,4-dihydroxyphenylglycol, oleuropein and oleuropein aglycons, as the main phenols, were detected in olive fruit. More interestingly, this research indicated that inoculation with LAB affected the pH, total acidity, microbial profile and palatability of olives. Olives fermented with the LAB starter culture were perceived by panellists to be less bitter and more aromatic than those spontaneously fermented. Thus the use of LAB inoculants during olive fermentation could be applied with the currently available technology. 2005 Society of Chemical Industry
Keywords: Olea europaea; Ascolana tenera; processing; fermentation; phenolics; sensory evaluation
INTRODUCTION Table olives are currently the most important fermented vegetable product in the developed world. Estimated world annual production is around 1.3 × 106 t, more than 80% of which comes from Mediterranean countries. About 44% of world production emanates from Europe, the main producing countries being Spain, Greece, Italy and Portugal. Either as an appetiser accompanying alcoholic and non-alcoholic beverages or in salads, pizzas and Mediterranean-style meals, table olives are highly appreciated for both their sensory characteristics and nutritive value, and their use is extensive in all markets. Italy’s annual table olive production amounts to about 8 × 104 t.1 Among the different olive varieties, ‘Ascolana tenera’ is one of the best for table olive production and, as a consequence, has been exported to many other countries such as the USA, Mexico, Israel and Argentina. ‘Ascolana tenera’, called ‘Picena’ by the ancient Romans and considered the best olive variety of the Empire, comes originally from the Marche region of central Italy and is mainly cultivated in the province of Ascoli Piceno on about 300 ha of land with more than 1000 t olives year−1
capacity.2 The cultivar produces large oval/spherical fruits with an average size of 121–140 fruits kg−1 and with a flesh-to-pit ratio higher than 6:1. The oil content is low, around 100–120 g kg−1 . The fruits have a soft texture with very delicate skin and pulp, so particular care must be taken during harvesting and processing. They are hand harvested from midSeptember to mid-October when they have reached a yellowish-green surface colour. Local authorities have asked the European Union to designate the variety as a denomination of protected origin (DPO), since ‘Ascolana tenera’ is known worldwide as a stoned olive fruit stuffed with assorted minced meats and vegetables and dipped in breadcrumbs before frying, a recipe also used to prepare other stuffed olive varieties referred to as ‘Ascolana-style olives’. However, this variety is also used for the preparation of Spanish-style whole green olives. The processing method consists of treating the olives with dilute NaOH solution to eliminate their natural bitterness, followed by several washings with water to completely remove excess alkali. The olives are then placed in a solution of sodium chloride (brine) where spontaneous lactic fermentation takes place. During fermentation,
∗
Correspondence to: Vincenzo Marsilio, Istituto Sperimentale per la Elaiotecnica, C da Fonte Umano 37, I-65013 Citta` S Angelo (PE), Italy E-mail:
[email protected] Contract/grant sponsor: Italian Ministry of Agriculture and Forestry (Received 21 June 2004; revised version received 22 September 2004; accepted 23 September 2004) Published online 26 January 2005
2005 Society of Chemical Industry. J Sci Food Agric 0022–5142/2005/$30.00
1084
Use of starters in olive processing
weak acidification rates are often registered, because most of the sugars and nutrients are lost by the effect of the lye treatment and strong washing system used, so the pH and final lactic acidity are very often unsuitable for safe storage of the end-product.3 Another problem is the softening of the fruits due to cell rupture and separation during industrial processing, resulting in a dramatic texture loss as a consequence of the lye degradation and solubilization of pectic polymers from the cell wall and middle lamella.4,5 In recent years, to meet the increasing demand for naturally/minimally treated foods and to preserve most of the original texture, naturally green olives in brine have also been produced. However, to our knowledge, no reports are available on the characteristics of this commodity. This prompted us to develop practical investigations to study the impact of processing conditions on the chemical, sensory and fermentation characteristics of olive fruit. Lactic acid bacteria (LAB) have long been employed in fermentation as a food preservation technique owing to their progressive acidification of the fermenting brine with a consequent pH decrease and the production of antimicrobial substances and bacteriocins.6 Over the past decade, phenolic compounds have attracted a great deal of attention in food quality because of their important role in human health and disease prevention.7 They contribute greatly to the flavour and colour of olives8,9 and, in addition, are important in the architecture of the fruit cell wall, since the formation of ether linkages with the cell wall polysaccharides has been hypothesised to play a role in controlling cell wall extensibility and texture properties.10 It has long been known that the composition of the phenolic fraction in olive fruit is very complex, depending on cultivar, maturity at harvest, preharvest agronomic conditions, season and environmental conditions.11,12 Many phenolic glycosides such as verbascoside, ligstroside, demethyloleuropein, cornoside and other phenolic compounds in free and bound forms have also been described in olive fruit.13 – 21 The present research addresses the impact of processing conditions on the levels of total phenols and evolution of each compound. Specific conditions assessed in this study include the evaluation of the effects of a selected LAB starter culture on olive fermentation.
EXPERIMENTAL Plant material and orchard management The trial was carried out in 2002 in a 30-year-old olive (Olea europaea L cv Ascolana tenera) orchard of the ‘Case Rosse’ Cooperative (Poggio di Bretta, Ascoli Piceno, Italy; 13◦ 39 E, 42◦ 53 N, elevation 250 m asl) planted on a sandy loam soil with trees spaced 6 m apart and trained using the traditional vase system without irrigation supply. In the experimental year the environment of olive growing was characterised by very low rainfall in spring, good precipitation J Sci Food Agric 85:1084–1090 (2005)
in May and July and frequent rain thereafter. The mean total rainfall in the period May–September was 66 l mq−1 , while the daily mean temperature increased from 10–12 ◦ C in April to 24–25 ◦ C in June and July and decreased to 18–20 ◦ C in September. Processing Olives were subjected to either Spanish-style (SSP) or Greek-style (GSP) processing. For SSP, 50 kg of olives were placed in a PVC container and covered with 30 l of 20 g l−1 NaOH solution to eliminate their natural bitterness. The alkali treatment lasted for 7 h at ambient temperature until the lye reached two-thirds of the flesh of fruits. The NaOH solution was then poured off and the olives were washed in water several times for a total of 30 h. After washing, the olives were covered with 40 g l−1 NaCl solution and left to develop spontaneous lactic fermentation. Another lot of 100 kg of olives from the same orchard was processed in parallel as GSP naturally green olives. After brining in 40 g l−1 NaCl solution for 1 week, the fruits were divided into two portions of 50 kg each and covered with 30 l of brine at the same concentration. The first portion was left to develop a spontaneous fermentation process (GSP-s), whereas the second portion was inoculated with a starter culture of lactic acid bacteria (GSP-i). A selected oleuropeinolytic Lactobacillus plantarum bacterial strain (LAB B1-2001) from the culture collection of the Istituto Sperimentale per la Elaiotecnica (Olive and Olive Oil Research Institute of the Italian Ministry of Agriculture and Forestry), Pescara, Italy was used as inoculant. The bacterial strain was propagated in MRS broth (Oxoid, Basingstoke, Hampshire, UK) for 18 h at 30 ◦ C, then the culture was reinoculated into MRS broth and incubated until the exponential phase of growth was reached. Cells were pelleted by centrifugation at 10 000 × g for 15 min at 4 ◦ C, washed twice with sterile water and resuspended at a concentration of ca 1010 colony-forming units (CFU) ml−1 . The culture was then inoculated into the fermenting brine at a ratio of 40 ml l−1 . All fermentation processes were carried out at a controlled temperature of 22 ± 2 ◦ C. The experimental trials were carried out in duplicate. Determination of pH and total titratable acidity pH was measured with a NeoMet 79P pH meter (Istek Inc, Seoul, Korea). Total titratable acidity (TTA) was determined as lactic acid by titration to pH 8.0 with 0.1 M NaOH. Sugar content Sugar content was determined by Fehling’s reagent according to the Italian official methods.22 Total phenolics Total phenolics were quantified by the Folin–Ciocalteu assay,23 a method based on the reduction of phosphomolybdic acid by phenols in aqueous alkali and currently used to determine the total free phenolic 1085
V Marsilio et al
groups in a sample. A 5 g sample of olive pulp was extracted three times with methanol. An 0.5 ml aliquot of olive methanolic extract was introduced into a test tube and mixed with 1 ml of 0.5 M Folin–Ciocalteu reagent. The mixture was allowed to stand for 5 min, then 3 ml of 200 g l−1 Na2 CO3 solution was added and the volume was made up to 10 ml with distilled water. The solution was stirred and allowed to stand for 30 min at room temperature. The mixture was then centrifuged for 10 min at 3500 × g and the absorbance of the supernatant was measured at 725 nm on a Perkin-Elmer Lambda 2 UV/VIS spectrophotometer (Norwalk, CT, USA). Total phenolics were expressed as caffeic acid equivalents. GC/FID and GC/MS analyses of biophenols Olive biophenols were analysed by the method reported by Patumi et al.24 Briefly, 5 g of olive pulp frozen in liquid nitrogen was extracted with aqueous ethanol (800 ml l−1 ) containing 5 g l−1 sodium metabisulfite, and 2 ml of an ethanolic solution of resorcinol (0.5 g l−1 ) was added as internal standard. The extract was concentrated under reduced pressure and washed with hexane to remove free fatty acids, oil and other lipid contaminants. The phenols were then extracted with ethyl acetate/water (2:1 v/v). The organic solution, filtered through anhydrous sodium sulfate and evaporated under vacuum at 30 ◦ C, was converted into trimethylsilyl ethers and analysed by GC/FID and GC/MS. A Carlo Erba (Milan, Italy) GC-5160 equipped with a flame ionisation detector (FID) and an HP1 capillary column (HewlettPackard, Palo Alto, CA, USA; 30 m × 0.32 mm (id), 0.1 µm film thickness) was used with hydrogen as carrier gas. The column temperature was programmed from 70 to 90 ◦ C at 20 ◦ C min−1 , from 90 to 300 ◦ C at 4 ◦ C min−1 and held at 300 ◦ C for 40 min. The sample (0.3 µl) was injected in ‘on-column’ mode. For GC/MS analysis an HP GC-5890 interfaced to an MSD-5970 was used under the same conditions as described above. Analyses were performed in electron impact mode at 70 eV using helium as carrier gas with a head pressure of 15 kPa, an interface temperature of 320 ◦ C and an ion source temperature of 200 ◦ C. Identification was performed by comparison of retention times and mass spectra with those obtained from authentic substances. The following reference substances were used: resorcinol (Merck, Darmstad, Germany); tyrosol, caffeic acid, vanillic acid, ferulic acid and p-coumaric acid (Sigma, St Louiss, MO, USA); oleuropein, rutin and luteolin-7O-glucoside (Extrasynth`ese, Genay, France). All other chemicals were of analytical grade. Microbiological assays At given fermentation times, brine samples were withdrawn from the containers and serial dilutions were prepared in sterile distilled water for microbiological counting by the standard plate method. From each dilution, 0.1 ml was spread on the media plates. Lactic 1086
acid bacteria were enumerated on MRS agar (Oxoid) at 30 ◦ C for 72 h and yeasts on Malt Extract agar (Oxoid) at 28 ◦ C for 72 h. Coliforms were also enumerated on 20 g l−1 Brilliant Green Bile (2%) broth (Oxoid) at 37 ◦ C for 24 h using the most probable number technique. Micro-organism enumeration in each solution was done in duplicate. The number of colony-forming units in 1 ml of brine (CFU ml−1 ) was calculated in each case. Sensory evaluation Fermented products were assessed at the Istituto Tecnico Agrario Statale, Ascoli Piceno, Italy by a panel consisting of 17 students selected for their interest in and consumption of table olives. Four training and monitoring sessions were organised to familiarise the panellists with the commodity. Odour, bitterness and the textural attributes firmness and crispness were evaluated. The descriptive analysis25 used a 10-point intensity scale ranging from 0 (no perception) to 10 (extreme). All samples were tested at room temperature under normal daylight conditions. The SPSS Professional Statistic package (SPSS Inc, Chicago, IL, USA) was used for statistical processing of sensory data. Significant differences between attribute scores were determined by analysis of variance (ANOVA) with the student’s t test at p < 0.01. Multivariate approaches were also applied using principal component analysis (PCA) to recognise the sensory information able to discriminate between spontaneously processed and inoculum processed olives and to represent the relationships between the olive-processing systems examined and the attributes evaluated.
RESULTS AND DISCUSSION Changes in pH and titratable acidity Figure 1 shows the changes in pH and TTA at different stages of olive fermentation. The pH fell from 5.5 to 4.4 in SSP olives, from 4.7 to 4.2 in GSP-s olives and from 4.3 to 3.9 in GSP-i olives. This reduction may
Figure 1. Evolution of pH and TTA during olive fermentation: , pH SSP; , TTA SSP; ♦, pH GSP-s; , TTA GSP-s; , pH GSP-i; , TTA GSP-i.
J Sci Food Agric 85:1084–1090 (2005)
Use of starters in olive processing
be due to the formation of lactic acid, which is the main LAB metabolic product. However, it is known that other acids such as succinic, citric, malic and acetic also contribute to the overall pH. Table olives with high pH (>4.0) are considered to be perishable when they are not subjected to preservative processes that delay undesirable biological and biochemical changes.26 Titratable acidity reached 2.2 g l−1 in SSP olives, whereas GSP-s and GSP-i olives had acidity values of 2.5 and 3.5 g l−1 respectively. The acidity levels found were in accordance with the microbial populations detected in the fermenting brines. Effect of fermentation on sugar content The evolution of the sugar content (Table 1) was monitored because sugars play an important role during olive fermentation, acting as carbon sources for micro-organisms.27,28 The sugar content decreased by 90% from the initial value in SSP olives, of which 49% occurred during the debittering and washing steps, whereas in GSP-s and GSP-i olives it decreased by 73 and 81% respectively over 5 months of fermentation. Therefore the LAB used as inoculants increased the sugar consumption and consequently the acidification rate, thus ensuring that any change in olive composition did not compromise the quality of the finished product. Effect of processing on total phenolic content and phenolic fraction composition Data on the total phenolic content in raw and processed olive fruits of ‘Ascolana tenera’ variety are shown in Fig 2. The levels of total phenolics in unprocessed and SSP olives were found to be 5138 and 488 mg kg−1 (on a wet weight basis) respectively. These results indicate that dramatic changes in total phenolic content occurred in the olive flesh during debittering with lye, washing with water and fermentation in brine. After 5 months of fermentation the levels were 2073 and 2513 mg kg−1 in GSP-i and GSP-s olives respectively. The mild treatments performed during processing were responsible for the fourfold higher retention of phenols in naturally green olives than in Spanish-style green olives. The composition of the raw and processed olive fruit phenolic Table 1. Evolution of sugar content (g kg−1 olive pulp) during processinga
Sample Raw material After debittering After washing 30 days of fermentation 60 days of fermentation 150 days of fermentation
SSP olives
GSP-s olives
GSP-i olives
37 28 19 16 6 4
37 — — 33 26 10
37 — — 30 19 7
a
Data are mean values of three replications; the coefficient of variation is
