Fruits of the “Tonda di Cagliari” cultivar were handpicked from an irrigated orchard in north Sardinia in the last decade of October at a green ripe stage and soon transferred at our lab for processing. The cultivar has been chosen because it proved to be very suitable for this kind of preparation in previous trials carried out at our lab. Selection before brining was done by discarding olives with blemishes, cuts and insect punctures as well as under or over ripe fruits. Calibration was performed with laboratory callipers and olives ranging from 17 to 21 mm in transversal diameter were considered. Replicates of 12 kg were prepared for further processing.

After careful washing with tap water to remove field dust, the olives were transferred to 20 L previously sterilized plastic containers and then processed by brining, which was carried out by preparing sterile 4 and 7% (w/w) sodium chloride solutions which were used to completely fill the containers which were finally closed after a perforated cap was placed between the olives and the top part of the container to submerge the olives in the brine. The pH of the brine was previously adjusted to 4 with lactic acid to prevent the growth of gas pockets producing Enterobacteriaceae (Garrido-Fernández et al., 1997). Three containers per sodium chloride concentration were used. The containers were kept at 20°C. The sodium chloride concentration and pH were inspected at 5, 10, 15, 20, 30, 45 and 60 days and afterwards at monthly intervals up to 180 days in order to make corrections when necessary. The growth of aerobic oxidative yeasts and moulds on the brine surface was avoided by adjusting the brine level with freshly prepared sterile brine. We stopped determinations at 180 days as most of the chemical parameters had reached a steady state and the olives were ready for consumption as revealed by a sensory test.

Triplicate analyses of the following parameters for each container were carried out in the brine: pH with a glass electrode, free acidity (g of lactic acid per 100 mL of brine), sodium chloride (g 100 mL^-1), reducing sugars (g 100 mL^-1 of brine) and phenols (mg gallic acid 100 g^-1 of brine), according to Garrido-Fernández et al. (1997). The free acidity value obtained was corrected by subtracting the initial and following amounts of lactic acid added. Determinations on homogenised flesh were pH, sodium chloride, reducing sugars and total phenols. Total phenols were extracted according to Amiot et al. (1986) and determined as described by Garrido-Fernández et al. (1997). Sodium chloride was determined as for brines after extraction from flesh: in brief, 10 g of flesh were diluted with 30 mL of water and homogenized with an Ultra Turrax for 60 seconds, the resulting slurry was brought to 50 mL and centrifuged for 10 min at 5000 rpm. An aliquot of 1 mL was taken to make the analysis following the method proposed by Garrido-Fernández et al. (1997). Reducing sugars and polyphenol analysis in brine and flesh and sodium chloride in flesh were made at 0, 30, 60, 120 and 180 days.

The antioxidant activity was determined on the polyphenolic extract using the DPPH stable radical, following the method proposed by Ramadan and Morsel (2006). Briefly, 50 µL of methanolic extract were dissolved in 1950 µL of toluene and after vortexing for 1 minute an amount of 200 µL was made to react in a 1cm path cuvette at 22°C under continuous stirring with 1.8 mL of a 10^-4 M toluenic solution of DPPH. The decrease in absorbance at 515 nm was taken at 1, 5, 15, 30 and 60 minutes against a blank of pure toluene using a spectrophotometer (mod. 8453, Hewlett–Packard, Palo Alto, California). The antioxidant activity was considered as the difference in absorbance between each sample and a toluenic DPPH solution without the sample (control) and calculated as percentage of inhibition according to the following equation: % of inhibition = absorbance of control - absorbance of sample × 100 absorbance of control

Texture analyses in fresh olives and changes during processing were monitored with a texture analyzer (mod. TA-XT2, Stable Microsystems, Surrey, UK) fitted with a 5 or 50 kg load cell. The Texture Expert program version 1.21 was used for processing data. A puncture test and a Texture Profile Analysis (TPA) were performed using twenty olives for each determination and brine concentration.

For the puncture test the olives were placed on an HDP/90 platform and punctured in the lateral face. The test was performed using an SMS P/2 needle probe and a 5 kg load cell. The test speed was 1.0 mm s^-1 and the penetration depth was limited to 3 mm. The force-time curve was graphically acquired and elaborated. Three parameters were taken into account to evaluate the evolution of texture during fermentation (Figure 1): the olive peel break force (yield point), as the maximum force (N) reached during puncturing (F1); the hardness of the peel, as the area under the curve (N mm^-1) up to the yield point (from point 0 to F1, that is A1); the hardness of the flesh (A2), as the area (N mm^-1) calculated starting from the point of minimum force after the probe penetration (F2) and the maximum force (F3).

The TPA was carried out on the lateral face of whole olives in compression mode using an SMSP/75 plate. The probe speed was set at 0.5 mm sec^-1, time between compressions was 5 sec and the test was stopped at 15% deformation after the auto triggering signal. A 50 kg load cell was used. Hardness, cohesiveness, gumminess, chewiness and springiness, as reported by Szczesniak (1963) and Friedman et al. (1963), were evaluated and calculated by the instrument software. Texture determinations were carried out at 0, 30, 60, 120 and 180 days.

The table olives were assessed at the end of brining (180 days) by 30 untrained judges who were asked to express their preference on the two samples. We performed a paired preference test as reported by Stone and Sidel (2004a).

Data on reducing sugars and polyphenol content on brines and olives and on the NaCl contents in the olives were subjected to one-way analysis of variance (ANOVA), with sodium chloride concentration as the group variable. Texture and antioxidant activity data were submitted to two-way ANOVA, where the factors were the sodium chloride and the days of storage. Data on sensory analysis were evaluated using the critical values table for duo-trio test and paired preference test (Stone and Sidel, 2004b). The software Statistica 6.0 for Windows was used to process data. Means, when required, were separated according to the Tukey test, with a significance level P≤0.05.

An analysis of the chemical-physical parameters in the brines and olives were carried out mainly to check the appropriate evolution of processing and to correct, at each inspection time, pH and sodium chloride values. Data obtained, in fact, were as expected and as reported in the literature (Garrido-Fernández et al., 1997; Piga et al., 2001; Piga and Agabbio; 2003) (Table 1). The pH and sodium chloride concentration in brines showed a major shift from the fixed values during the first 60 days of processing, and after that were more stable (data not shown). The values for the corrected free acidity revealed the development of values not exceeding 0.5 g of lactic acid per 100 mL of brine, which is evidence of fermentative yeast growth and an absence of lactobacilli (data not shown). The diffusion of polyphenols and reducing sugars from the flesh into brines was noticeable and significantly higher in olives processed with a 4% sodium chloride concentration. On the contrary and as expected, the sodium chloride content in the olives was significantly higher in the samples brined at a 7% salt concentration.

As a result of the penetration test, the particular structure of the olive fruits generates a particular graph which shows a progressive increasing force during penetration up to the breaking of the peel. This maximum force (F1) has already been considered as a hardness index of the peel (Romeo et al., 2012). This part of the curve looks like a type C one, as reported by Bourne (1979). In addition, the area under the curve up to F1, explains the work required to break the peel and thus represents its hardness. After this yield point a sudden force loss is registered due to the penetration of the first layers of flesh, which opposes a lower resistance to probe progress, and an increasing force up to a maximum value is registered. The area between the minimum force F2 and the maximum force F3 is a good index for flesh hardness. The evolution of the structure of the peel, as revealed by the puncture test, shows significant differences between olives brined with the two NaCl concentrations. In fact, the peel of olives brined with 7% NaCl showed higher values for the breaking force and hardness, with respect to the 4% processed olives (Table 2). These differences were significant after 60 days (Figures 2–3). Tassou et al. (2007) did not find differences in skin strength following brining with different sodium chloride concentrations, but they worked with black olives and performed the texture analysis on a peel strip and in tension mode. The most important result, however, is that peel break force and hardness increased during brining. Romeo et al. (2012) found similar results by brining turning color “Biancolilla” olives at 12%; while Georget et al. (2003) and Tassou et al. (2007) found a different behavior, but, as stated before, both the different degrees of ripening or mechanical testing may have influenced the results. This peel hardening may be explained by the fact that the peel shrivelled due to the contemporary effects of sodium chloride absorption and water loss, thus resulting in the development of a crust.

On the other hand, the different NaCl concentrations did not influence the flesh hardness, which, in turn, showed a significant decrease during the 180 days of brining (Table 2; Figure 4). This time, the results agree with those reported by Georget et al. (2003) and only in part with those reported by Romeo et al. (2012), and may be explained both by the loss of turgor and degradation of pectic polysaccharides during brining (Coimbra et al., 1996).

Figure 5 shows a typical TPA graph where the two peaks of the two compression cycles can be seen. This kind of test was not able to show texture differences between the two sodium chloride concentrations used, but gave important information about the evolution of the different textural attributes during brining (Table 3). All the considered parameters, in fact, changed significantly at the first sampling time and then remained statistically constant. Hardness, which is the measurement of the maximum force of the first bite, decreased, thus indicating a softening of the flesh, as already determined from the puncture test. The same trend was noticed for cohesiveness, which accounts for the strength of the bonds making up the body of the olive flesh, and for the springiness which explains the retained elasticity of the product. The decrease in these parameters influenced the evolution of chewiness and gumminess, calculated from the above first three cited parameters. The already cited work of Coimbra et al. (1996) may explain again the loss in texture of the product.

To our knowledge, these are the very first data on texture evolution using the TPA test during the brining of green table olives using two different sodium chloride concentrations.

The DPPH test is an in vitro radical scavenging (RS) measurement and its evolution during the 180 days of brining is reported in Figure 6 as percent absorbance decrease, with respect to a blank of DPPH without sample, after 60 minutes of reaction. It is worth pointing out that the major loss in RS activity has been recorded after the first month of brining (22–28%), thus mirroring the polyphenol loss, and that the total loss was close to 40% (34–42%) at the end of processing. A loss in antioxidant activity during brining has been already reported and was found to be well correlated to polyphenol loss (Ben Othman et al., 2009). In our case, the polyphenol loss was higher, accounting for 78–82% at the end of the brining period, with respect to the decrease in antioxidant activity reported before. Compounds other than polyphenols or polyphenols themselves in an intermediate state of oxidation may be present in the extract, thus explaining the lower decrease in RS, with respect to % loss in polyphenolic content. Olives brined with 7% of NaCl showed a higher RS activity, with respect to the 4% brined samples, probably due to the higher polyphenol content of the extract.

As stated for texture evolution, no other data are present in the literature concerning the antioxidant evolution during the brining of green table olives by means of two different sodium chloride concentrations.

The panellists positively judged both samples, however, 23 judges preferred the olives brined with 4% sodium chloride, as they found the olives less salty and bitter, with respect to the 7% brined ones. This result is significant at P≤0.01.