campus m., et al. variability in composition, sensory profiles and volatile compounds of sardinian...

7/21/2019 Campus M., Et Al. Variability in Composition, Sensory Profiles and Volatile Compounds of Sardinian Monovarietal V…

http://slidepdf.com/reader/full/campus-m-et-al-variability-in-composition-sensory-profiles-and-volatile 1/12

237

L A R I V I S T A I TA L I A N A D E L L E S O S TA N Z E G R A S S E - V O L . X C - O T T O B R E / D I C E M B R E 2 0 1 3

27 olive oil samples coming from monovarietal olive orchards located in Sardinia (Italy)were collected and analyzed for basic chemical parameters (Free acidity, Peroxide value,

K232, K270,Δ

K),α

-tocopherol and phenolics content (HPLC), fatty acid profile (GC-FID), volatile compounds (HS-SPME-GC-MS) and sensory profiles (Panel test). All sam-ples resulted in them belonging to the extra virgin category. Chemical and sensory analy-ses highlighted inter and intra-cultivar differences between sample groups. Head spacevolatile compounds composition, lacking in the literature, was determined, and resultsshowed a prevalence of compounds coming from the Lipoxygenase pathway (LOX) actingmainly on Linolenic acid. The aldehyde E-(2)-hexenal was found the more abundant com-pound in the headspace, along with C6 and C5 compounds coming from the LOX path-way of linoleic and linolenic acids (Hexanal, 3-hexenal, 1-penten 3-one, 2-penten-1-ol, 1-penten-3-ol, and the ester 3-Hexen-1-ol acetate). The principal component analysis ofselected variables was used to discriminate samples from the same variety coming fromdifferent regions.

Keywords: olive oil, composition, volatile compounds, sensory analyses, sardinia.

La variabilità della composizione, profili sensoriali e composti volatili di olidi oliva vergini sardi monovarietali coltivati in diverse aree27 campioni di olio ottenuto da oliveti monovarietali situati in Sardegna (Italia) sono statiprelevati e analizzati nei parametri chimici merceologici (acidità libera, numero dei peros-

sidi, K232, K270,ΔK), contenuto inα-tocoferolo e sostanze fenoliche (HPLC), composi-zione acidica (GC-FID), composti volatili (HS-SPME-GC-MS) e profilo sensoriale (Paneltest). Tutti i campioni sono risultati appartenere alla categoria extra vergine. Le analisi chi-miche e sensoriali effettuate hanno messo in evidenza una variabilità significativa tra oliottenuti dalla medesima cultivar coltivata in zone differenti. La composizione in compostivolatili dello spazio di testa è stata determinata, mostrando una prevalenza dei composti

derivanti dalle reazioni della via della lipossigenasi, a carico principalmente dell’acidolinolenico. L’aldeide Trans-(2)-esenale è risultato il composto più abbondante nello spa-zio di testa, insieme ad altri composti C6 e C5 derivanti dalle reazioni della via LOX (esa-nale, 3-esenale, 1-penten-3-one, 2-penten-1-ol, 1-penten-3-ol, e l’estere 3-esen-1-olacetato). L’analisi delle componenti principali effettuata sulle variabili selezionate ha con-sentito di mettere in evidenza le differenze tra campioni provenienti da zone differenti.Parole chiave: Olio d’oliva, composizione, composti volatili, analisi sensoriale, Sardegna.

Variability in composition, sensoryprofiles and volatile compounds of

Sardinian monovarietal virgin oliveoils grown in different areas

M. Campus*

P. SeddaD. Delpiano

S. SecciG. Damasco

R. ZurruG. Bandino

Agris Sardegna, DiRArb - Cagliari

*CORRESPONDING AUTHOR Tel.: +39 070 60181

Fax: +39 6018204

e-mail address:[email protected]



INTRODUCTION

Virgin olive oil, the product of mechanical process-

ing of fruits of Olea europea L., has gained great

diffusion due to its sensory characteristics and

healthy properties [23, 30], with a growing con-

sumption all over the world in recent years. Sar-

dinia is a production region located in the Mediter-ranean basin, whose peculiar climate and soil

characteristics have favored the diffusion of both

wild forms (Olea europea oleaster, L.) and the

introduction of cultivated forms by the Phoenicians,

Romans and Spanish during their campaigns to

conquer [25, 26]. Throughout this 3000 years of

history, different cultivars differentiated, and the

actual forms are probably the result of cross breed-

ing of wild and cultivated forms. Sardinian olive cul-

tivars have been characterized both phenotypical-

ly [7, 8, 9, 10] and genetically [4, 5, 6] and produce

high quality virgin olive oils with peculiar character-istics, gaining national and international prizes in

recent years. Some varieties, classified based on

phenotypical traits, resulted in being genetically

homogeneous with others, based on DNA

microsatellite studies. The following varieties have

been identified so far: Bosana , Nera di Villacidro,

Tonda di Cagliari (Sin. Nera di Gonnos) , Pizz’e car-

roga , Semidana . There are few studies on Sardin-

ian monovarietal olive oils characterization. Some

authors marked the importance of the genotype on

the characteristics of monovarietal oils, pointing

the attention mostly to the fatty acid compositionand sensory profile and some chemical parame-

ters definition [15, 11]. In the literature, no studies

can be found on the characterization of volatile

compounds responsible for the aroma of Sardinian

monovarietal olive oils. Indeed, the evaluation of all

the quality characteristics, their development and

their preservation during the shelf life is of chief

importance both for producers and consumers.

Moreover, variability in oils coming from the same

cultivar but grown in different areas of Sardinia has

not been pointed out yet. In the present work,

chemical parameters, vitamin E, phenolics content,fatty acid profile, volatile compounds and sensory

characterization have been performed and cross

correlation among selected variables analyzed by

means of multivariate statistical analyses. The aim

of present work to add new insights on the factors

affecting the quality of Sardinian monovarietal olive

oils and eventually finding some of them useful to

highlight inter and intra-cultivar differences.

MATERIALS AND METHODS

SAMPLING27 freshly produced Sardinian mono-varietal olive

oils (SMO) were supplied by the producers. Har-

238

L A R I V I S T A I TA L I A N A D E L L E S O S TA N Z E G R A S S E - V O L . X C - O T T O B R E / D I C E M B R E 2 0 1 3

vest was carried out at a similar level of ripeness.

Oils, according to the producers, were obtained in

continuous operating plants.

Olives were processed within 24 hours from the

mechanical harvest, applying the following gener-

al work flow: washing and de-leafing, crushing,

malaxation, extraction with a 3 phase centrifuga-

tion system with low water added, centrifugation ina vertical separator, collection of the oil in steel

tanks, storage in a climatized room, then bottling.

The oils were representative of the range of mono-

varietal extra virgin olive oils obtained from Sardin-

ian cultivars available on the market. 16 samples

were from cultivar Bosana , 4 from Nera di Oliena , 3

from Semidana , 1 from Nera di gonnos , 2 from

Nera di Villacidro , 1 from Ogliastrina . Nera di Olie-

na and Ogliastrina are sinonyms of Nera di Villaci-

dro , according to Baldoni et al [6]. Nevertheless,

we kept the different names to differentiate the dif-

ferent areas of origin and their terroir . Orchardswere located in distinctive areas of the Sardinian

region, as shown in Figure 1. Once transported to

the laboratory, oils were transferred in darkened

glass bottles (250 ml for chemical analyses, 500 ml

for sensory analyses) and kept stored at 15°C in a

climatized room, in the dark, until the analyses. All

analyses were repeated 3 times (n=3). Along with

protocol numbers, samples were coded according

to the cultivar (B: Bosana; N.O: Nera di Oliena; N.V:

Nera di villacidro; Ogl.: Ogliastrina; N.G: Nera di

Gonnos; S: Semidana) and region of origin (N:

North; S: South; W: West; C: Center).

FREE ACIDITY, PEROXIDE VALUE, K232, K270, ΔKAnalytical parameters for commercial classifica-



tion were determined according to the methods

described in Regulation EEC/2568/91 of the Euro-

pean Commission [37]. Briefly: free acidity,

expressed as % of free oleic acid, was made by

titration of a solution of oil (filtered) dissolved in

ethanol/ethyl ether (1:1) with 0.1 N potassium

hydroxide in methanol, using phenolphthalein 1%

in ethanol as an indicator. Peroxide value, as mil-liequivalents of active oxygen per Kg of oil

(meq/Kg) was determined as follows: 5g of oil

were weighed in a 250 ml flask and were added

with a solution 1:1 of chloroform/acetic acid, then

left reacting with 1 ml of KI saturated solution, in

the darkness, for 5 minutes. Free iodine was then

titrated with sodium thiosulphate 0,01 N using

starch paste as an indicator. Molar extinction at

230 and 270 nm (K232 K270,) and ΔK were cal-

culated from absorption at 232 and 270 nm,

respectively, with a UV spectrophotometer (Jasco

V-560), using a 1% solution of oil in isooctane anda path length of 1 cm.

GC-FID FATTY ACIDS COMPOSITIONFatty acids methyl esters (FAME) were prepared

dissolving 0,5 g of oil in 6 ml of n-Hexane, adding

0,25 ml of KOH 2N in Methanol. After 10 seconds

of vigorous shaking, samples were centrifuged at

3000 rpm for 10 minutes and the supernatant col-

lected. FAME analysis was done with a Agilent

7890A Gas Chromatograph (Agilent, Palo Alto,

CA), equipped with a Flame Ionization Detector

(FID). Separation was carried out with a SupelcoEC-1000 capillary column (30 m, 0,25 µM film

thickness), using Helium as the carrier gas, at a

flow rate of 1.2 ml/min. The GC oven temperature

program began with the oven held at 185°C for 17

min, then increased to 220°C at 4°C/min, main-

tained at 220°C for 8 min, then to 230°C at 2°C/min,

for 6 minutes. The total run time was 45 min. Detec-

tor temperature was set at 300°C, H2 flow at 30

ml/min, air flow at 400 ml/min, make up gas (N2)

flow at 25 ml/min. Sample Injection (1µL) was

made in Split mode (50:1) at 60 ml/min. FAME

were identified comparing retention times withthose of authentic standards (Sigma Aldrich) and

expressed as area units (%) in respect to the total

TIC (Total Ion Current) area.

α-TOCOPHEROL ANALYSIS0.1 g of oil was dissolved in 1.9 ml of acetone and

filtered (0.22 µm PVDF syringe filter). HPLC analy-

sis was carried out with a Waters HPLC system

equipped with a 600 Controller pump module, a

717 Plus autosampler and a 996 Photodiode Array

Detector (DAD). The separation was performed

with a Waters Spherisorb ODS 2 column(250x4.6x5µm) coupled with a Phenomenex guard

column using methanol: acetonitrile 50:50 (v/v) as

a mobile phase, eluting at a flow rate of 1.00

ml/min, in isocratic mode, at 25°C (total run time 18

min). α-tocopherol was detected at 295 nm and

identified comparing the RT with that of an authen-

tic standard. Concentration was calculated corre-

lating the area of the peak with the concentration,

referring to a standard calibration curve.

All reagents used were HPLC grade purchasedfrom Sigma Aldrich.

TOTAL PHENOLS DETERMINATIONTotal phenols were quantified according to the IOC

(International Olive Oil Council) method

COI/T.20/Doc No 29, November 2009 [17].

The same apparatus used for tocopherol quantifi-

cation was used for the analyses of the phenols.

Briefly, the method is based on direct extraction of

the biophenolic minor polar compounds from olive

oil by means of a methanol solution and subse-

quent quantification by HPLC using a water 0.2%H3PO4 (V/V) (A), methanol (B), acetonitrile (C) ter-

nary linear elution gradient. UV spectra were

acquired with a UV detector at 280 nm. Syringic

acid is used as the internal standard. The content

of total phenolic compounds is expressed in mg/kg

of tyrosol. All reagents used were HPLC grade

purchased from Sigma Aldrich.

ANALYSIS OF VOLATILE COMPOUNDSHS-SPME (Head Space-Solid Phase Micro Extraction)Fiber was obtained from the Supelco Company

(Bellefonte, PA). The fiber used for the extraction ofthe volatile components from the vial head space

was divinylbenzene/carboxen/polydimethylsilox-

ane (DVB/CAR/PDMS) 50/30 mm. Before use, the

fiber was conditioned, as recommended by the

manufacturer. 2 g of olive oil were placed in a 10 ml

vial closed by PTFE/silicone septum (Supelco).

Before extraction, the stabilization of the head-

space in the vial was accomplished by equilibra-

tion for 30 min at 40°C. The extraction was carried

out at 40°C, exposing the fiber within the vial head-

space, for 30 minutes.

GC–MS ANALYSESEach oil was analyzed by GC–MS using an Agilent

7890A/5975C system, with installed fused-silica

HP-5 capillary column (30 m x 0.25 mm, 0.25 µm

film thickness). After extraction of volatile com-

pounds, injections were performed using a SPME

manual device (Supelco). The fiber was thermally

desorbed into the GC and left in the injection port

(equipped with a 0.75 mm i.d. inlet liner) for 3 min.

The injector was set at 260°C and operated in the

splitless mode. All samples were analyzed 3 times

(n=3) working with the following temperature pro-gram: the initial oven temperature was maintained

at 40°C for 5 min, and then increased at 4°C/min

239

L A R I V I S T A I TA L I A N A D E L L E S O S TA N Z E G R A S S E - V O L. X C - O T T O B R E / D I C E M B R E 2 0 1 3



up to 220°C, and the final temperature was held for

10 min. The carrier gas was helium at a pressure of

15 psi on the column head. The mass spectra was

acquired at 70 eV ionization energy over a mass

range of 15-250 amu, ion source and transfer line

temperatures were 170 and 280°C, respectively.

Compounds identification was based on computer

matching against commercial (NIST 2011) libraryand MS literature data [18, 19, 20, 21]. Selected

standards were used to verify the identities of key

volatiles. All chemicals were purchased from

Sigma–Aldrich.

SENSORY ANALYSISSensory analyses were performed by the analytical

Panel of the Arboriculture Research Department

(DiRArb, Cagliari, Italy) of Agris, The Agricultural

Research Agency of Sardinia (Bonassai, Italy),

using twelve trained tasters. For classification pur-

poses, the sensory evaluation of olive oils was per-formed according to European regulations (Annex

XII of EC regulation n. 2568/91, 640/08).

A second Sensory Assessment sheet, developed

to allow a careful description of the peculiar senso-

ry profiles, was used. The sensory assessment

sheet was divided into 2 parts, one for the sensory

profile (oil description) and one for grading purpos-

es (oil quality evaluation). In the first, the presence

of defects and eleven sensory attributes were con-

sidered: Fruity (green olives), grass, artichoke,

green tomato, apple, almond, ripeness, fruitiness,

flowers, bitter, pungent, astringent and scalingintensity from 0 to 10. The sample presentation was

fully randomized. For quality grading purposes (oil

quality evaluation), an overall score was calculated

summing the scores for harmony, pleasantness-

complexity, bitterness and pungency, assessed on

a scale from 0 to 10, and olive fruitiness, on a scale

from 0 to 60, in terms of agreeableness. The sum

gives a measure of the overall sensory score of the

oil. Data were acquired and analyzed with the soft-

ware Sensory (Imedia, Cagliari, Italy), and results

reported as mean values.

STATISTICAL ANALYSESResults are reported as mean values. Significant

differences were determined by analysis of vari-

ance using Fisher’s least significant difference

(LSD) test at P = 0.05 to discriminate between

means. All analyses were carried out in triplicate

and the results were presented as means of three

repetitions. Principal Components Analysis (PCA)

was performed together with cross correlation

analysis on selected variables. Pearson’s correlation

between variables, where pertinent, were calculat-

ed. All the statistical analysis has been performedusing the statistic software Statgraphics plus v 5.1

(StatPoint Technologies Inc., Warrenton, VA).

RESULTS AND DISCUSSION

FREE ACIDITY, PEROXIDE VALUE AND UV SPECTROPHO-TOMETRIC INDEXES (K232, K270, ΔK).Table I. shows the physicochemical quality param-

eters for the SMO studied. Free acidity is related to

tissue damage prior to extraction, and its low value

reflects the effect of olive quality, harvesting, trans-port, and storage conditions on the overall quality.

Peroxide values and UV spectrophotometric index-

es reflect the degree of primary oxidative phenom-

ena occurring as a consequence of oxidation of

unsaturated fatty acids and the presence of conju-

gated dienes and trienes. All the analyzed oils

showed very low mean values (acidity ≤0.8%; per-

oxide index ≤20 meq O2 kg-1 ; K270 ≤0.22; K232

≤2.5 and DK ≤0.01), they belonged to the ‘‘extra

virgin’’ category, as stated by Regulation

EC/1989/2003 (EEC, 2003). The low values for

these parameters reflect the high commercial qual-ity of oils.

α-TOCOPHEROL AND TOTAL PHENOLS CONTENTAmong the different classes of compounds having

antioxidant activity, hydrophilic phenols and toco-

pherols are of major importance for the stability of

olive oil. It is well known that α-tocopherol is the

major tocopherol in olive oil [12]. The oil contents of

α-tocopherol for the studied samples ranged from

the 135,95 to 314 (mg/kg). Phenolic compounds

are closely associated with the nutrional and sen-

sory qualities of foods, contributing directly or indi-rectly to desirable or undesirable aromas and fla-

vors and to oxidative stability. The higher content of

phenolic substances and α-tocopherol were

detected in samples coming from Nera di Oliena

and Bosana , both coming from the center part of

Sardinia (see Tab. I). Previous studies emphasized

the high content of antioxidant compounds, such

as phenols, in Sardinian Cultivar Bosana [10], the

more extensively grown variety in Sardinia, which

make this oil particularly resistant to oxidation dur-

ing storage, although none had emphasized the

importance of the different growing areas on thevariability of its content, which has great influence

on the sensory characteristics of obtained oils, as

seen hereinafter in this article.

SENSORY ANALYSISFrom a sensory point of view all the examined sam-

ples belonged to the extra virgin olive oil class

(Tab. I), according to Regulation EEC/796/2002

(EEC, 2002) [36], without any defects (median of

the defects was equal to 0), characterized by a

panel score more than 6.5 and the median of the

fruitiness was above 0. As previously stated, a sec-ond sensory sheet was used to carefully describe

the sensory profiles; its results are given in Table II.

240




241

L A R I V I S T A I TA L I A N A D E L L E S O S TA N Z E G R A S S E - V OL . X C - O T T O B R E / D I C E M BR E 2 0 1 3

Briefly, the direct observation of the intensities of

attributes detected by tasters showed that the oils

studied were mainly characterized by medium (4-5

mean score), medium-intense (5-6 mean score)

perceptions of fruitiness (green olives), an equili-

brate taste of bitter and pungent, particularly

intense in samples of Bosana (coming from North,

South and Center) and Nera di Oliena , with grass

and artichoke sensations clearly detectable (Fig.

2). Light almond flavor was always present in all

samples. In some samples, notes of flowers, green

tomato and ripe fruit could be detected. Bitter and



242




243


Pungent taste were found significantly (P< 0,1) cor-

related with phenols content; Indeed, Pearson’s cor-

relation coefficient equals 0,557 (bitter) and 0,763

(pungent), indicate a moderately strong relationship

between variables. This is in accordance with previ-

ous works, that correlated bitter and pungent taste

with particular phenolic substances [38, 39].

The overall score of the oils was used as a variableto perform the Principal Component Analysis (data

not shown).

GC-FID FATTY ACIDS COMPOSITIONFatty acid composition of the oil may differ depend-

ing on the variety. Table III gives the fatty acids

composition for each fatty acid in the studied SMO

groups. The monounsaturated fatty acids have

great importance because of their nutritional impli-

cation and effect on oxidative stability of oils. Oleic

acid (C18:1) is the main monounsaturated fatty

acid and is present in higher concentrations (63.4-78.17%). The average level of palmitic acid

(C16:0), major saturated fatty acid in olive oil,

ranged from 11.26 for Ogl. E. to 15.72% for N.G.S.

samples. Concerning linoleic acid (C18:2), which

is much more susceptible to oxidation than

monounsaturated fatty acids, was observed to

have the highest percentage in N.G.S. (15.7%),

whereas the lowest one was found in Ogl.E

(6.39%). The other samples showed percentages

ranging between 6,39% (Ogl. E.) and 15,70%

(N.G.S.). For the other fatty acids: Miristic (C14:0),

Palmitoleic (C16:1); Margaric (C17:0), Margaroleic(C17:1), Stearic (C18:0); Linolenic (C18:3),

Arachidic (C20:0), Gadoleic (C20:1) and Behenic

(C:22), although having small amounts, they varied

between oil samples. The ratio C18:1/C18:2 can be

useful in olive cultivars characterization and stabil-

ity interpretation. Its values ranged from 4,04

(N.G.S.) to 12,23 (Ogl. E.). The same samples pre-

sented the higher (18,39%) and lesser (13,79)

average content of saturated fatty acids (SFA%).Regarding the total monounsaturated fatty acids

(MUFA), Ogl. E. contained the highest percentage

(79,22%) due to its high content in oleic acid.

N.G.S. were rich in total polyunsaturated fatty acids

(PUFA) (16,42%) because of its high contents in

linoleic acid representing the major fatty acid of

that fraction. The fatty acid composition has a rela-

tively wide variability due to the genetic (variety)

and environmental factors. It has previously been

used by a number of authors as a parameter for oil

classification [22, 28]. It is an essential aspect of

the qualitative assessment of olive oil.

ANALYSIS OF VOLATILE COMPOUNDSSolid-phase microextraction coupled to GC–MS

was used to characterize the volatile fraction of

SMO. (E)-2-hexenal was found to be the main

volatile component found in all analyzed samples,

its abundance (expressed as TIC, Total Ion Current

chromatogram) ranging from 16.16 to 48.44% of

TIC. This is in accordance with other authors,

which indicate that (E)-2-hexenal is the dominant

volatile of European virgin olive oils [13, 14, 37].

Indeed, the isolated and identified compounds forthe studied oil samples are mainly aldehydes



(36.68-69.13%); other classes of compounds includ-

ed alkenes (5.85-27.92%), alcohols (2.45-22.18%),ketones (1.56-13.95%), esters (1–13.65%), terpenes

(0.28–8.66%), alkanes (0.4–7.29%), acids (0.45–

8,16%), and other components (hydrocarbons)

(0–3.04%) (Fig. 3). All these volatile compounds,

whether major or minor, are crucial to olive oil qual-

ity. The products of Lipoxygenase cascade (LOX)

were the major components of volatile fraction of

the studied oils and the sum of their peak areas

ranged between 52.39 and 79.98% of the total

area. Mean % composition values for the different

groups of analyzed samples are shown in Table IV.

C6 compounds were the major components in the

oil headspace, ranging between 51,47% and

86.18% of the whole of the LOX products. The

headspace of tested olive oils showed a good

amount of C5 compounds, between 5.65 and

31,83% of total LOX products, in addition to C6

compounds, suggesting the presence of a addi-

tional branch of the LOX pathway leading to theirproduction, as indicated by Angerosa et al. [2].

The LOX pathway became active at the moment of

the olive crushing, when cells lose their inner com-

partmentation bringing substrates and enzymes in

physical proximity. C5 and C6 volatile compounds

formed, because of their solubility, are quickly

incorporated in the oily phase and accumulate dur-

ing the malaxation of the resulting pastes [3]. The

compounds identified are responsible for the posi-

tive aroma properties of high-quality oil, mainly

contributing to its green notes [40, 27]. Therefore,

changes of the concentrations of each C5 and C6compound can notably modify the sensory percep-

tions. From a quantitative point of view C6 aldehy-

des, E-2-hexenal, Hexenal, 3-hexenal and C5 com-

pounds 1-penten 3-one, 2-penten-1-ol, 1-penten-3-

ol and the ester 3-Hexen-1-ol acetate (3-Hexenyl-

acetate), are the most abundant LOX products

present in the headspace of analyzed samples.

Aldehydes are usually characterized by intense

sensory descriptions associated with green, fruity

and bitter sensory notes. C6 esters were identified

in all samples and their presence is associated

244




with fruity sensory notes [23]. The high level of (E)-

2-hexenal in olive oils shows the pre-eminence of

the (E)-2-hexenal/(E)-2-hexenol branch of LOX pat-

way, compared to the hexanal/hexanol one in all

the considered varieties. The volatile fraction of oils

coming from Bosana (B.N., B.S., B.C., B.E.), Nera

di Oliena (N.O) and Semidana from the west region

(S.W.) were characterized by a higher (E)-2-hexe-nal amount. The oils with the highest 2-hexen-1-ol

were S.W. Hexanol was detected in 2 samples,

namely 132 (Bosana, N) and 201 (SW). Moreover,

acetic acid hexyl ester (hexyl acetate) was present

in almost all aromas but in very low amounts, espe-

cially in Semidana , Nera di gonnos (N.G.) and Nera

di Villacidro (N.V.), although the differences were

not statistically significant among samples, except

for B.N., thus indicating a low presence of alcohol

acetyl transferase (AAT) activity [34] (<0,1%, of

total aroma, or not detected in the latter samples)

acting on Linoleic acid derived Lipoxygenaseproducts. This fact could be explained by a pro-

gressive inactivation of esterases due to the malax-

ation time, since esters undergo a considerable

decrease after 30 min of malaxation while C6 alco-

hols increased [1]. While Hexyl acetate was low in

concentrations, 3-hexenyl acetate was found in

good concentrations in several samples, showing

a good AAT activity on linolenic acid degradation

compounds. Enzymatic activity toward 3-hexen-1-

ol by olive oil AAT explains the scarcity of the short

chain alcohol in some of the analyzed samples

although the concentration of the product (3-hex-enyl acetate) is dominant among volatile esters.

Taken together, the results showed that the

amounts of compounds arising from linolenic acid

(LnA) are far greater than that of compounds from

linoleic acid (LA) and this agrees with previous

work [23, 14, 41]. Samples coming from the same

variety showed no statistical differences in the total

amount of LnA derived compounds, independently

of their geographical origin, so highlighting a major

influence of genotype over environmental factors.

Among C5 compounds, 1-Penten-3-one was the

most abundant. It is identified in all but one oil sam-

ple (158, Nera di oliena, C) with mean values rang-

ing between 1,44 (135, Bosana, N) and 8,71 (164,

Bosana, S), with no statistical differences betweengroups. Its presence is associated with pungent,

mustard and strawberry sensory notes. Low

amounts of 1-penten-3-ol and 1,3-pentadiene also

affects the aroma of studied oils. Oil samples com-

ing from different varieties showed different volatile

profiles. It is reported that the cultivar is the domi-

nant factor in the formation of the oil aroma [23, 24].

PRINCIPAL COMPONENT ANALYSISPCA analysis performed on samples revealed 3

principal components, explaining 79,11% of vari-

ance. Variables are distributed in the PCA planbased on their relative contribution to the principal

components (Fig. 4). Variables taken into account

were Peroxide value (PRX), UV indexes (K232,

K270, ΔK), Free acidity (FA%), α-tocopherol con-

tent (TF), Total Phenols, (TP), Overall sensory score

(Score), C18:1-C18:2 ratio. We have taken into

account correlations between variables and com-

ponents (eigenvectors) >0,4. Analyzing the first

two component, accounting for the biggest vari-

ability of the data (63,114%), the first principal

component (38,6 percent of variance) is correlat-

ed mostly with oxidation indexes, such as the pres-ence of conjugated dienes (K232, 0,484) and

trienes (K270, 0,426), and Peroxide value, (0,438).

Moreover, it is slightly (0,358) correlated with enzy-

matic hydrolysis (Free acidity) and to C18:1/C18:2

ratio (0,327). Samples with higher oxidation index-

es plot in the negative part of the principal compo-

nent. We can define this first component in terms of

245




“Oxidation status”. Notably, first component

appear, although slightly, negatively correlated

(0,32) with C18:1/C18:2 ratio, which is reported as

a reliable relationship for stability interpretation and

varieties characterization. In this respect, only the

sample coming from Ogliastrina is significantly dif-

ferent from all other samples, altough Nera Oliena ,

Ogliastrina , and Nera di Villacidro , are geneticallyhomogeneous and up to date classified as Nera di

Villacidro according to DNA microsatellite studies,

we kept the different names to highlight the differ-

ent regions of origin. These latter samples grouped

all in the positive part of the first component.

The Second Principal component is highly and

positively correlated with total phenols (0,52) and

α-Tocopherol content (0,403) and slightly and neg-

atively correlated with free acidity (-0,364) and per-

oxide value (-0,363). Samples with higher phenols

and tocopherol clusterize in the positive part of the

second component axis. Oxidative stability of oliveoils is directly related to these parameters. This

second component can be interpreted in terms of

antioxidants content. In general, it is reported that

phenols content depends on the extraction system,

irrigation, and olive variety [31, 33]. The phenolic

content of samples ranged from relatively low (196)

to high (570). This second component differenti-

ates samples of Bosana coming from the north and

the south, which plot in the negative part, namely

with less phenols and tocopherols content (Tab. I),

from those coming from the center, all in the posi-

tive part of the component, with higher phenolsand α-tocopherol content. All samples but one of

B.N. (6 out of 7 samples) group in the positive part

of the first component and in the negative part of

the second component, formed a clear cluster.

Moreover, B.C. samples group in the negative part

of the first component and in the positive part of the

second (4 out of 5 samples), formed another clus-

ter. Also, Nera di Villacidro samples clusterize in

the negative part of the second component, while

Ogliastrina and Nera di Oliena , although genetical-

ly the same variety of Nera di Villacidro , clusterize

in the positive part. The variability in antioxidantscontent within samples coming from the same cul-

tivar, processed in the same way but cultivated in

different regions, reflects the high influence of

agronomic and pedoclimatic factors on this param-

eters magnitude. α-Tocopherol content values

ranged from 135,95 to 314,29 mg/Kg. In conjunc-

tion with phenols, they contribute to oxidative sta-

bility, thus affecting the shelf life.

CONCLUSIONS

There are few studies on Sardinian monovarietalolive oils characterization in international literature

and nothing can be found on their volatile com-

pounds composition. The results obtained provide

information on the sensory profile, fatty acid com-

position, phenol and α-tocopherol content, and

volatile compounds responsible for the aroma. The

analyzed olive oils showed chemical profiles corre-

sponding to extra virgin olive oil category. They are

also characterized by good sensory attributes

highly appreciated by the consumers. Majorvolatile compounds found were aliphatic C6 com-

pounds responsible for the green notes of olive

oils, mainly arising from the lipoxygenase pathway

of LnA. The aldehyde (E)-2-hexenal was present in

large amounts in the olive oil headspace of all

analysed oils and is the major volatile compound.

Other key compounds identified were mainly Hexa-

nal, 3-hexenal, 1-penten-3-one, 2-penten-1-ol, 1-

penten-3-ol, 3-Hexen-1-ol acetate. The results indi-

cate that LOX is the predominant pathway of

volatiles biogeneration in Sardinian virgin monova-

rietal olive oils, acting mainly on LnA. The principalcomponent analysis highlighted the importance of

the growing zone, for the majority of samples ana-

lyzed, on the oxidation status and antioxidant con-

tent of samples coming from the same variety.

More studies are needed to highlight the relation-

ships between olive variety, environment, process-

ing technology, and the resulting chemical compo-

sition, sensory profile and stability (shelf-life).

ACKNOWLEDGMENTS

Authors wish to thank Andrea Coni, Marco Serreli,Federico Corda and Sandro Cera for their technical

assistance. Marco Campus wishes to thank

Antonello Campus for providing graphic editing

support.

REFERENCES

[1] F. Angerosa, N. D’Alessandro, C. Basti, R.

Vito, Biogeneration of volatile compounds in

virgin olive oil: their evolution in relation to

malaxation time. J. Agric. Food. Chem. 46 ,

2940-2944, (1998).[2] F. Angerosa, R. Mostallino, C. Basti, R. Vito,

Virgin olive oil odour notes: their relationships

with volatile compounds from the lipoxyge-

nase pathway and secoiridoid compounds.

Food. Chem. 68 , 283-287, (2000).

[3] F. Angerosa, R. Mostallino, C. Basti, R. Vito,

Influence of malaxation temperature and time

on the quality of virgin olive oils. Food Chem.

72 , 19-28, (2001).

[4] A. Angiolillo, L. Baldoni, G. Bandino, M.

Mulas, Analisi molecolare con marcatori AFLP

delle risorse genetiche di olivo della Sarde-gna. Atti IV Conv. Naz. Biodiversità, Carlo Del-

fino Ed., 413-416, (1998).

246




[5] A. Angiolillo, M. Mencuccini, L. Baldoni, Olive

genetic diversity assessed using amplified

fragment length polymorphisms. Theor. Appl.

Genet. 98 , 411-421, (1999).

[6] L. Baldoni, N. Tosti, C. Ricciolini, A. Belaj, S.

Arcioni, G. Pannelli, M.A. Germanà, M. Mulas,

A. Porceddu, Genetic structure of wild and

cultivated olives in the central mediterraneanbasin. Ann. Bot. 98 , 935-942, (2006).

[7] G. Bandino, C. Moro, M. Mulas, P. Sedda,

Analisi dei caratteri varietali in una collezione

di germoplasma proveniente da popolazioni

di olivo reperite in diverse zone della Sarde-

gna. Atti del Congresso sul Germoplasma

Frutticolo. Salvaguardia e Valorizzazione delle

Risorse Genetiche. Alghero, 21-25 settembre,

211-216, (1992).

[8] G. Bandino, C. Moro, M. Mulas, P. Sedda, Bio-

logia fiorale e di fruttificazione delle cultivar di

olivo del germoplasma della Sardegna. Atti delConvegno su L'Olivicoltura Mediterranea: stato

e prospettive della coltura e della ricerca.

Rende (CS), 26-28 gennaio, 415-423, (1995).

[9] G. Bandino, M. Mulas, P. Sedda, C. Moro, Sur-

vey on olive genetic resources of Sardinia.

Acta Hort. 474 , 151-154 (1999).

[10] G .Bandino, S. Dettori, Manuale di Olivicoltu-

ra. 1st edn. Regione Sardegna, CIF, Cagliari-

Nuoro-Oristano, (2001).

[11] G. Bandino, P. Sedda, C. Moro, M. Mulas,

Monovarietal olive oils of Sardinia: trends and

prospects. In: I convegno nazionale dell’olivoe dell’olio, Claudio di Vaio (ed.), Hacta Italus

Hortus, Società Ortoflorofrutticoltura Italiana,

Firenze, 264-267, (2010).

[12] G. Beltran, M.P. Aguilera, C. Del Rio, S.

Sanchez, L. Martinez, Influence of fruit ripen-

ing process on the natural antioxidant content

of Hojiblanca virgin olive oils. Food Chem, 89 ,

207-215, (2005).

[13] B. Berlioz, C. Cordella, J.F. Cavalli, L. Lizzani-

Cuvelier, A.M. Loiseau, AND X. Fernandez,

Comparison of the Amounts of Volatile Com-

pounds in French Protected Designation ofOrigin Virgin Olive Oils. J. Agric. Food. Chem.

54, 10092-10101, (2006).

[14] J.F. Cavalli, X. Fernandez, L. Lizzani-Cuvelier,

M. Loiseau, Characterization of volatile com-

pounds of French and Spanish virgin olive oils

by HS–SPME of quality freshness markers.

Food Chem. 88 , 151-157, (2004).

[15] L. Cerretani, A. Bendini, A. Del Caro, A. Piga,

V. Vacca, M.F. Caboni, T.G. Toschi, Preliminary

characterization of virgin olive oils obtained

from different cultivars in Sardinia. Eur. Food.

Res. Technol. 222 , 354-361, ( 2006).[16] H. Guth, W. Grosh, Quantitation of potent

odorants of virgin olive oil by stable-isotope

dilution assays. J. Am. Oil Chemist’s Soc. 70 ,

513-518, (1993).

[17] COI/T.20/DOC.29, Determination of biophe-

nols in olive oils by hplc.

http://www.internationaloliveoil.org/estaticos/

view/224-testing-methods, (2009).

[18] W. Jennigs, T. Shibamoto, Qualitative Analysis

of Flavor and Fragrance Volatiles by GlassCapillary Gas Chromatography. Academic

Press, New York, (1980)

[19] D. Joulain, W.A. König, The Atlas of Spectra

Data of Sesquiterpene Hydrocarbone. E.B.-

Verlag, Hamburg, (1998).

[20] D. Joulain, W.A. König, D.H. Hochmuth, Ter-

penoids and related constituents of essential

oils. Library of Mass Finder 2.1, (2001)

[21] K.L. Goodner, Practical retention index mod-

els of OV-101, DB-1, DB-5, and DB-Wax for

flavor and fragrance compounds. LWT Food.

Sci. Technol. 41, 951-958, (2008).[22] C.M. Lanza, C. Russo, F. Tomaselli, Relation-

ship between geographical origin and fatty

acid composition of extra-virgin olive oils pro-

duced in three areas of Eastern Sicily, Italy. J.

Food Sci. 4 , 359-366, (1998).

[23] G. Luna, M.T. Morales, R. Aparicio, Characteriza-

tion of 39 varietal virgin olive oils by their volatile

compositions. Food Chem. 98 , 243-252, (2006).

[24] H. Manai, F. Mahjoub-Haddada, I. Oueslati,

D. Daoud, M. Zarrouk, Characterization of

monovarietal virgin olive oils from six crossing

varieties. Sci. Hortic. 115 , 252-260, (2008).[25] A. Milella, Contributo allo studio delle cultivar

sarde di olivo. Indagini condotte in provincia

di Sassari. Studi Sassaresi, 5 , 3-29, (1957).

[26] A. Milella, Contributo allo studio delle cultivar

sarde di olivo. II. Indagini condotte in provin-

cia di Cagliari e Nuoro. Studi Sassaresi, 11, 1-

51, (1961).

[27] M.T. Morales, J.J. Calvente, R. Aparicio,

Infuence of olive ripeness on the concentra-

tion of green aroma compounds in virgin olive

oil. Flav. Frag. J. 11, 171-178 , (1996).

[28] M.J. Motilva, T. Ramo, M.P. Romero, Caracte-rizacion geografica de los aceites de oliva vir-

genes de la denominacion de origen protegi-

da ‘Les Garrigues’ por su perfil de aicidos

grasos. Grasas y Aceites 52 , 26-32, (2001).

[29] J.M. Olias, A.G. Perez, J.J. Rios, L.C. Sanz,

Aroma of virgin olive oil: biogenesis of the

green odour notes. J. Agric. Food Chem. 41,

2368-2373, (1993).

[30] D.B. Panagiotakos, C. Chrysohoou, C. Pit-

savos, K. Tzioumis, I. Papaioannou, C. Ste-

fanadis, The association of Mediterranean

diet with lower risk of acute coronary syn-dromes in hypertensive subjects. Int. J. Cardi-

ol. 82 , 141-147, (2002).

247




[31] M. Patumi, L. D’andria, V. Marsilio, G. Fonta-

nazza, G. Morelli, B. Lanza, Olive and olive oil

quality after intensive monocone growing

(Olea Europea L., CV. Kalamata) in different

irrigation regimes. Food Chem. 77 , 27-34,

(2002).

[32] S.M. Preziuso, G. Di Loreto, A. Biasone, Phe-

nolic compounds related to bitter and pun-gent organoleptic virgin olive oil attributes. In:

I convegno nazionale dell’olivo e dell’olio,

Claudio di Vaio (ed.), Hacta Italus Hortus,

Società Ortoflorofrutticoltura Italiana, Firenze,

280-283.

[33] M.P. Romero, M.J. Tovar, J. Girona, M.J. Motil-

va, Changes in the HPLC phenolic profile of

virgin olive oils from young trees (Olea euro-

pea L., CV. Arbequina) grown under different

deficit irrigation strategies. J. Agric. Food

Chem. 52 , 3649-3654, (2004).

[34] J.J. Salas Characterization of Alcohol Acyl-transferase from olive fruit. J. Agric. Food

Chem. 52 , 3155-3158, (2004).

[35] H. Schroder, Protective mechanisms of the

Mediterranean diet in obesity and type 2 dia-

betes. J. Nutr. Biochem. 18 , 149–160,

(2007).

[36] EC (1991) Regulation No. 2568/91. Office of

Joint Commission European Communities.

[37] S. Vichi, A.I. Castellote, L. Pizzale, L.S. Conte,

S. Buxaderas, E. Lopez-Tamames, Analysis of

virgin olive oil volatile compounds by head-

space solid-phase microextraction coupled to

gas chromatography with mass spectrometric

and flame ionization detection. J. Chromatogr.

A 983, 19-33, (2003).

[38] F. Gutiérrez, M.A. Albi, R. Palma, J.J. Rios,J.M. Olías, Bitter taste of virgin olive oil: corre-

lation of sensory evaluation and instrumental

HPLC analysis; J Food Sci 54 , 68-70, (1989).

[39] M.T. Morales, M. Tsimidou, The Role of volatile

compounds and polyphenols in virgin olive oil

quality. In: Handbook of Olive Oil: Analysis

and Properties, Harwood J. and Aparicio R.

(eds.), 393-458, (2000).

[40] R. Aparicio, M.T. Morales, Characterization of

Olive Ripeness by Green Aroma Compounds

of Virgin Olive Oil. J Agric. Food Chem. 46 ,

1116-1122. (1998).[41] J.J. Salas, C. Sánchez-Peña, R. Aparicio,

Impact of the Suppression of Lipoxygenase

and Hydroperoxyde Lyase on the Quality of

the Green Odor in Green Leaves. J Agric.

Food Chem. 53 , 1648-1655 (2005)

Received, October 2, 2012

Accepted, January 10, 2013

248


campus m., et al. variability in composition, sensory profiles and volatile compounds of sardinian...

Documents