The effect of water deficit stress on the growth, yield and

composition of essential oils of parsley

S.A. Petropoulos

a

, Dimitra Daferera

b

, M.G. Polissiou

b

, H.C. Passam

a

,

*

a

Agricultural University of Athens, Laboratory of Vegetable Production, 75 Iera Odos, 11855 Athens, Greece

b

Agricultural University of Athens, Laboratory of General Chemistry, 75 Iera Odos, 11855 Athens, Greece

Received 6 July 2007; received in revised form 21 September 2007; accepted 4 October 2007

Abstract

Three parsley cultivars (plain-leafed, curly-leafed and turnip-rooted) were grown under conditions of 35–40% and 45–60% water deficit in

order to evaluate the effect of this form of stress on plant growth, essential oil yield and composition. Plant growth (foliage and root weight, leaf
number) was significantly reduced by water stress, even at 30–45% deficit. Water stress increased the yield of essential oil (on a fresh weight basis)
from the leaves of plain-leafed and curly-leafed, but not turnip-rooted, parsley. However, on a m

2

basis foliage oil yield increased significantly only

in curly-leafed parsley. Water stress also caused changes in the relative contribution of certain aroma constituents of the essential oils (principally
1,3,8-p-menthatriene, myristicin, terpinolene + p-cymenene), but these changes varied between cultivars. The oil yield of roots was low and water
deficit stress had relatively little effect on the root oil composition. It is concluded that because the biomass of plants subjected to water deficit is
reduced, it is possible to increase the plant density of plain-leafed or curly-leafed parsley, thereby further increasing the yield of oil per m

2

.

However, the application of water deficit stress to parsley essential oil production must also take into account likely changes in oil composition,
which in turn relate to the cultivar.
#

2007 Elsevier B.V. All rights reserved.

Keywords:

Plain-leafed; Curly-leafed; Turnip-rooted parsley; Drought

1. Introduction

Parsley (Petroselinum crispum [Mill.] Nym. ex A.W. Hill) is

cultivated both as a herb for consumption in the fresh and dried
state and as a source of essential oils. Whereas plain-leafed and
curly-leafed parsley are cultivated for their foliage, turnip-
rooted or ‘‘Hamburg type’’ parsley, is cultivated for its
enlarged, fleshy, edible roots (

Petropoulos et al., 2006

).

The optimisation of irrigation for the production of fresh

leaves and roots of parsley is essential since, as in other
horticultural crops, water is a major component of the fresh
produce and significantly affects both weight and quality (

Jones

and Tardieu, 1998

). Water deficit in plants may lead to

physiological disorders, such as a reduction in photosynthesis
and transpiration (

Sarker et al., 2005

), and in the case of

aromatic crops may cause significant changes in the yield and
composition of essential oils. For example, water deficit

decreased the oil yield of rosemary (Rosmarinus officinalis L.)
and anise (Pimpinella anisum L.) (

Singh and Ramesh, 2000;

Zehtab-Salmasi et al., 2001

). By contrast, water stress had a

positive effect on pepper (Capsicum annuum L. var. annuum)
by increasing the phenolic capsaicinoids (capsaicin and
dihydrocapsaicin) and thereby increasing pungency (

Estrada

et al., 1999

). Moreover, although water stress caused a

significant reduction in the growth and oil yield of citronella
grass (Cymbopogon winterianus Jowitt.) per acre, oil yield
expressed on the basis of plant fresh weight increased, with the
severity of the water stress response varying with cultivar and
plant density (

Fatima et al., 2000

).

Plain-leafed parsley is widely grown throughout the

Mediterranean region, usually under irrigation, whereas the
turnip-rooted form has been proposed as a suitable alternative
crop within the framework of the European common
agricultural policy (

Petropoulos et al., 2006

). In Greece, the

principal areas of cultivation are situated close to large urban
markets, such as Athens and Thessaloniki. Because of the
relatively hot, dry climate, plants are susceptible to water stress,
especially during late spring and early summer; hence, the

www.elsevier.com/locate/scihorti

Available online at www.sciencedirect.com

Scientia Horticulturae 115 (2008) 393–397

* Corresponding author. Tel.: [phone]; fax: [phone].

E-mail address:

[e-mail]

(H.C. Passam).

0304-4238/$ – see front matter # 2007 Elsevier B.V. All rights reserved.
doi:

10.1016/j.scienta.2007.10.008

application of a precise irrigation schedule is of major
importance. Since the effect of such stress on parsley growth
and essential oil composition has not been previously reported,
the experiments described in the present paper were under-
taken.

2. Materials and methods

Three parsley cultivars were used: (a) Petroselinum crispum

(Mill) Nym. ex A.W. Hill ssp. neapolitanum Danert cv. plain-
leafed (Geniki Fytotechniki, Greece), (b) P. crispum ssp.
crispum

(L.) cv. curly-leafed (Geniki Fytotechniki, Greece) and

(c) P. crispum ssp. tuberosum (Bernh.) Crov., turnip-rooted cv.
Fakir (Bejo Zaden b.v., Holland).

Seeds were sown in trays containing a compost of peat

(KTS2, Klasmann-Deilman Gmbh, Geeste, Germany) and
washed, riverbed sand in a ratio of 2:1 (v/v) on 12 January (year
1) and 17 January (year 2). At the 2–3 leaf stage, plants were
transplanted to 10 l plastic pots (3 plants per pot 8 pots per
treatment) containing the same compost and placed in rows so
as to achieve a plant density of 27 plants per m

2

. The pots were

retained in an unheated greenhouse until early spring, when
they were transferred outdoors. The substrate contained 150 g
superphosphate (0–46–0), 90 g potassium nitrate and 900 g
marble dust per m

3

.

Pots were irrigated with 0.3 l water for the first 2 weeks,

increasing gradually to 1.5 l by the end of the growth cycle,
regardless of treatment. Irrigation was applied once a week
during the early stage of growth increasing to up to three times a
week during the stages prior to harvest. Soluble fertilizer
(1.25 g l

1

of 20N-20P-20K) was dissolved in the irrigation

water once every 2 weeks throughout the cultivation. Water
deficit was achieved by withholding irrigation in relation to the
control until soil moisture reached the desired level. Two levels
of water deficit were maintained with the aid of tension meters
(Irrometer-Moisture Indicator, Irrometer, Riverside, CA), in
which irrigation was applied at soil water potentials of 30–45%
(level 1) and 45–60% (level 2), whereas the control was held at
0–10% (field capacity). These percent levels are those indicated
in terms of centibars by the tension meters, where 0%
represents soil at field capacity and 100% dry soil. Plants of all
treatments were hand-harvested at the stage of market
acceptability on 21–22 May (year 1) and 25 May (year 2).
At the time when water stress was applied, mean air
temperature was 16.0 and 20.8 8C, mean air temperature

max

was 20.1 and 25.2 8C, mean temperature

min

was 11.2 and

15.9 8C and relative humidity was 51 and 65% (years 1 and 2,
respectively).

In year 2, leaves and roots were separated and sliced into

small pieces immediately after harvest, placed in sealed,
airtight plastic food-bags and stored at

10 8C, in order to

determine the yield and composition of the essential oil.
Because the yield of oil from parsley is rather low, Clevenger
distillers, with 1000 ml round, heating flasks were used.
Samples of 100–150 g of sliced frozen leaves or roots were
boiled in distilled water in the flasks heated by thermo-mantles
(Barnstead Electrothermal EMV 1000, Barnstead International,

Southend, U.K.) for 3 h from the start of boiling. The volume of
the oil phase from each distillation was measured and the
product put into 30 ml glass bottles, sealed with parafilm and
stored at

20 8C until analyzed. In order to remove any traces

of water in the samples before gas chromatography analysis, the
essential oil was extracted from the water phase in 10 ml diethyl
ether, with a 50 ml extraction funnel.

Since only small quantities of oil were required for

composition analysis, oil samples for this purpose were more
conveniently obtained by micro-steam distillation in a Likens–
Nickerson apparatus, which requires a shorter boiling time.
Samples of 10–12 g of sliced frozen leaves were boiled in
distilled water in 100 ml round flasks, heated by thermo-
mantles (Barnstead Electrothermal EMV 250), with 5 ml
diethyl ether for extraction. The process was carried out for 1 h
from the start of vapor condensation on the condenser walls.
The organic phase (containing the essential oil) was put into
30 ml glass bottles, sealed with parafilm and stored at

20 8C

until analyzed.

Essential oils extracted with hydro-distillation were ana-

lyzed by gas chromatography using a Hewlett Packard 5890 II
GC (Hewlett Packard, Waldbronn, Germany) equipped with a
FID detector and HP-5MS capillary column (30 m 0.25 mm,
film thickness 0.25 mm). Injector and detector temperatures
were set at 220 and 290 8C, respectively. Column temperature
was initially kept at 50 8C for 5 min, then gradually increased to
200 8C at a rate of 4 8C/min and maintained for 5 min. The flow
rate of helium was 1 ml/min. Quantitative data were obtained
electronically from FID area percent data without the use of
correction factors. Each extraction was replicated three times
and the compound percentages are the means of the three
replicates.

Gas chromatography/mass spectrometry (GC/MS) analysis

was applied for the samples obtained with micro-steam
distillation and performed under the same conditions as GC-
FID (column, oven temperature, flow rate of the carrier gas)
using a Hewlett Packard 5890 II GC equipped with a Hewlett
Packard 5792 mass selective detector (Hewlett Packard,
Waldbronn, Germany) in the electron impact mode (70 eV).
Injector and MS transfer line temperatures were set at 220 and
290 8C, respectively. The identification of components was
based on the comparison of their GC retention time and mass
spectra with authentic standards when possible. Additionally,
tentative identification was based on the comparison of their
relative retention time and mass spectra with those of the
NBS75K library data of the GC/MS system and literature data
(

Adams, 2001; Petropoulos et al., 2004

).

Leaf and root yields were subjected to analysis of variance

and means compared by the least significance difference test
using the statistical packages Statgraphics Plus 5.1 and JMP
4.0.2 v. Oil yield and composition statistics were carried out
with Microsoft Excel.

3. Results

Exposure of parsley plants to increasing levels of water

deficit caused a progressive decrease in the fresh weight of

S.A. Petropoulos et al. / Scientia Horticulturae 115 (2008) 393–397

394

foliage of curly-leafed and turnip-rooted parsley, whereas the
foliage weight of plain-leafed parsley decreased greatly when
plants were subjected to the lower level of water deficit (30–
45%) but not further at the higher water deficit level (45–60%).
The number of leaves per plant also decreased under the
influence of water stress (except curly-leafed parsley in year 2),
as did root fresh weight in year 1. In year 2, root fresh weight
was only reduced in plants of plain-leafed and turnip-rooted
parsley exposed to the higher level of water deficit (45–60%),
whereas the root weight of the curly-leafed cultivar was not
affected (

Table 1

).

The effect of water deficit stress on the yield of essential oils

extracted from parsley leaves (expressed as ml per 100 g fresh
weight) was variable. Thus, oil yield was higher in the leaves of
plants of plain-leafed and curly-leafed parsley subjected to
water stress, but not in the leaves of the turnip-rooted cultivar.
Overall, the yield of oil from the roots was lower than that of the
leaves and was not affected by water deficit stress, irrespective
of cultivar (

Table 2

). When expressed on an area basis (ml per

m

2

) oil yield of the leaves was found to be higher in curly-leafed

parsley subjected to water deficit stress, but not in the other two
cultivars. Water stress did not significantly affect the oil yield of
the roots on a fresh weight basis, and even decreased oil yield
on a m

2

basis in the plain-leafed cultivar (

Table 2

).

Water deficit stress affected the relative composition of the

essential oils of the leaves (

Table 3

). In plain-leafed parsley,

both levels of water stress caused a reduction in the relative
concentration of 1,3,8-p-menthatriene, but an increase in
myristicin. In this cultivar too, the percent concentration of
terpinolene + p-cymenene decreased at the higher stress level.
In the leaves of curly-leafed parsley, increasing water stress
caused a progressive decrease in the percentage of myristicin
which was offset by increases in a- and b-phellandrene,
terpinolene + p-cymenene. In contrast, the composition of the
essential oil extracted from the leaves of turnip-rooted parsley
was only marginally affected by water stress. In this cultivar, a
small, but significant decrease in the relative concentration of
b

-myrcene coincided with a small increase in apiole (

Table 3

).

Water stress had relatively little effect on the composition of

the essential oil extracted from the roots of turnip-rooted
parsley. A decrease in myristicin was observed at the higher
level of water deficit, as was a slight decline in the percentage of
terpinolene + p-cymenene at both levels of stress (

Table 4

).

Water stress did not appear to affect the relative composition of
the essential oil from curly-leafed or plain-leafed parsley,
except for a relative decrease in the apiole content of the latter,
coinciding with a small increase in b-pinene (data not shown).

4. Discussion

Essential oils derived from parsley are of value to the

cosmetic industry as well as for the synthesis of medicinal

Table 1
The effect of water deficit stress on the fresh weight of foliage and roots and the number of leaves per plant of three parsley cultivars

Cultivar

Level of water deficit

Year 1

Year 2

Foliage (g)

Root (g)

Leaf number

Foliage (g)

Root (g)

Leaf number

Plain-leafed

Control (0–10%)

62.3 a

17.8 a

8.2 a

80.7 a

27.8 a

9.1 a

Level 1 (30–45%)

32.8 b

11.3 b

5.9 b

55.6 b

24.9 ab

7.0 b

Level 2 (45–60%)

29.6 b

7.3 c

5.9 b

49.7 b

21.1 b

6.4 b

Curly-leafed

Control (0–10%)

52.6 a

14.8 a

6.5 a

86.3 a

14.3

6.7

Level 1 (30–45%)

28.7 b

9.0 b

4.5 c

58.1 b

14.5

6.5

Level 2 (45–60%)

24.4 c

5.8 c

5.3 b

46.3 c

14.1

6.0

Turnip-rooted

Control (0–10%)

52.2 a

42.3 a

6.9 a

62.2 a

44.4 a

6.7 a

Level 1 (30–45%)

33.7 b

27.8 b

5.1 b

43.7 b

41.4 a

5.7 b

Level 2 (45–60%)

29.5 c

20.8 b

4.9 b

35.5 c

25.8 b

4.8 c

Means for each cultivar within the columns that are not followed by a letter or are followed by the same letter are not significantly different at P = 0.05.

Table 2
The effect of water deficit stress on the essential oil yield of parley leaves and roots expressed as ml of essential oil per 100 g fresh plant material and ml per m

2

Level of water deficit

ml per 100 g fresh weight

ml per m

2a

Plain-leafed

Curly-leafed

Turnip-rooted

Plain-leafed

Curly-leafed

Turnip-rooted

Leaves

Control (0–10%)

0.04 b

0.05 b

0.04

0.87

1.17 b

0.67

Level 1 (30–45%)

0.06 a

0.07 b

0.05

0.90

1.09 b

0.59

Level 2 (45–60%)

0.07 a

0.11 a

0.05

0.94

1.38 a

0.48

Roots

Control (0–10%)

0.05

0.03

0.02

0.38 a

0.12

0.24

Level 1 (30–45%)

0.04

0.02

0.01

0.27 b

0.08

0.11

Level 2 (45–60%)

0.04

0.02

0.02

0.23 b

0.08

0.14

Means for each year within the columns that are not followed by a letter or are followed by the same letter are not significantly different at P = 0.05.

a

Plant density = 27 plants m

2

.

S.A. Petropoulos et al. / Scientia Horticulturae 115 (2008) 393–397

395

products. Oils are extracted mainly from the foliage or the
seeds, but yields are rather low in comparison with other
aromatic species, e.g. dill (Anethum graveolens L.) (

Callan

et al., 2007

). When grown in warm climates such as those of the

Mediterranean Basin, especially in late spring and summer,
parsley may easily be exposed to stress due to water deficit. As
shown in the present paper, such stress causes a reduction in
biomass, as expressed by mean foliage and root weight, as well
as leaf number per plant (

Table 1

), and probably results from a

disruption of photosynthesis, transpiration and other metabolic
processes (

Jones and Tardieu, 1998; Sarker et al., 2005

).

Water deficit stress does not reduce the essential oil yield of

parsley foliage on a fresh weight basis. Indeed, the oil yield of
both the plain-leafed and the curly-leafed cultivars increased
significantly under stress conditions (

Table 2

), thus compensat-

ing for the reduction in fresh biomass. This result contrasts with
those of

Zehtab-Salmasi et al. (2001)

, who reported that water

stress reduced oil yields from rosemary.

Singh and Ramesh

(2000)

also reported that water deficit stress reduced the oil

yield of rosemary on a hectare basis, but oil yield on a plant
fresh weight basis did not appear to be affected.

In curly-leafed parsley water deficit stress may be

considered beneficial since the oil yield of the foliage increases
with stress on an area (m

2

) basis. Indeed, this effect may be

further enhanced if planting occurs at a higher density, which is
permitted thanks to the smaller plant size under stress
conditions (

Petropoulos, 2006

;

Table 1

). The value of higher

plant densities under stress conditions has been noted for
citronella grass (

Fatima et al., 2000

). Theoretically, by

increasing the plant density of curly-leafed and plain-leafed
parsley from 27 to 36 plants per m

2

, an increase in net foliage

oil yield of 44 and 57%, respectively at a water deficit level of
45–60% may be achieved in comparison with unstressed plants
at the lower plant density, but the corresponding foliage oil
yield of turnip-rooted parsley is still 5% less than that of
unstressed plants. The value of increasing the plant population,
however, requires field testing since the increase in plant
density might further increase plant water stress.

The effects of water stress are not only confined to plant

growth and essential oil yield, but also extend to the quality of the
oil. Parsley aroma appears to be due primarily to 1,3,8-p-
menthatriene (

Jung et al., 1992; Masanetz and Grosch, 1998

).

Consequently, the reduction in the relative concentration of 1,3,
8-p-menthatriene in the oil extracted from plain-leafed parsley
subjected to water stress (

Table 3

) can be considered detrimental

to oil quality, even though this was partly offset by an increase in
myristicin, another important aromatic constituent (

Simon and

Quinn, 1988

). In curly-leafed parsley, the progressive decrease in

the percentage of myristicin due to increasing water stress may be
ameliorated by the increase in b-phellandrene, terpinolene and
p

-cymenene, which also contribute to parsley aroma (

Kasting

et al., 1972; Freeman et al., 1975

). Changes in the composition of

essential oils as a result of exposing plants to water stress have
also been reported for citronella grass (

Fatima et al., 2000

), but in

rosemary soil moisture levels had no effect on oil quality (

Singh

and Ramesh, 2000

).

Table 3
The effect of water deficit stress on the relative concentrations of the principal components of the essential oil extracted from parsley leaves

Component

Relative concentration (%) of essential oil components in the leaves of

Plain-leafed parsley

Curly-leafed parsley

Turnip-rooted parsley

Control

Level 1

Level 2

Control

Level 1

Level 2

Control

Level 1

Level 2

a

-Pinene

4.11

3.98

4.32

1.17

1.13

1.86

5.36

5.75

7.6

b

-Pinene

3.35

3.36

3.50

1.04 b

0.99 b

1.95 a

4.69

5.28

6.7

b

-Myrcene

6.76

5.76

4.28

6.35

6.54

7.91

26.94 a

24.44 b

22.0 b

a

-Phellandrene

1.07

0.96

0.99

0.29 c

0.61 b

0.83 a

0.97 a

0.37 b

1.1 a

p

-Cymene

nd

a

nd

0.6

0.28

0.44

nd

1.02

0.97

1.0

b

-Phellandrene

25.07

23.63

21.97

9.66 b

18.66 a

20.96 a

26.73

25.41

29.7

Terpinolene + p-cymenene

10.77 a

10.23 a

6.79 b

2.62 c

3.86 b

5.77 a

18.69 a

21.28 a

16.4 b

1,3,8-p-Menthatriene

5.49 a

2.63 b

1.95 b

0.15

0.34

3.18

3.16

nd

3.5

b

-Elemene

0.77

0.78

0.96

3.52

4.50

3.32

0.66 a

0.45 b

0.3 c

Myristicin

28.63 b

33.43 a

33.35 a

61.09 a

48.19 b

41.20 c

0.43

1.34

1.0

Apiole

2.91

1.90

6.35

2.96

3.04

2.95

nd

0.10 b

2.7 a

Total

88.93

86.66

85.06

89.13

88.30

89.63

88.65

85.39

92.00

Means for each cultivar within the rows that are not followed by a letter or are followed by the same letter are not significantly different at P = 0.05.

a

nd: not detected.

Table 4
The effect of water deficit stress on the relative concentrations of the principal
components of the essential oil extracted from the roots of turnip-rooted parsley

Component

Relative concentration (%) of essential oil
in roots from

Control

Level 1

Level 2

a

-Pinene

nd

nd

1.1

b

-Pinene

16.43 a

12.97 b

21.0 a

b

-Myrcene

9.88 a

5.67 b

7.7 ab

a

-Phellandrene

1.53

1.35

1.6

p

-Cymene

1.72 a

1.25 b

1.1 b

b

-Phellandrene

20.95 a

14.37 b

19.8 a

Terpinolene + p-cymenene

3.04 a

2.31 b

2.5 b

1,3,8-p-Menthatriene

nd

nd

nd

b

-Elemene

1.05

1.1

1.5

Myristicin

13.55 a

16.05 a

7.7 b

Apiole

23.59 b

33.92 a

25.7 b

Total

91.74

88.99

89.7

Means within the rows that are not followed by a letter or are followed by the
same letter are not significantly different at P = 0.05. nd: not detected.

S.A. Petropoulos et al. / Scientia Horticulturae 115 (2008) 393–397

396

Turnip-rooted parsley is the only one of the three parsley

types that is cultivated for its roots, which are fleshy tap roots as
opposed to the branching adventitious root systems of the plain-
leafed and curly-leafed cultivars (

Petropoulos, 2006

). Because

of the low yield of essential oil from parsley roots, only turnip-
rooted parsley may be considered for root oil production. Even
so, the yield is appreciably less than that of the foliage on a fresh
weight basis (

Table 2

). Although water deficit stress may affect

the quality of oil from the roots of turnip-rooted parsley, since
the percentage of myristicin was slightly reduced under stress
conditions, the main effect of water stress here is a reduction in
oil yield on a plant or m

2

basis. Even when applying higher

plant densities, the net root oil yield under stress conditions is
similar to the oil yield under normal conditions and plant
densities, rendering this practice of no practical value.

In conclusion, although water deficit stress reduces plant

biomass, in the case of curly-leafed parsley this is offset by an
increase in the essential oil yield per 100 g fresh tissue.
Although it may be possible to increase oil yield of both curly-
leafed and plain-leafed parsley per m

2

by exploiting the

reduction in plant size to use higher plant densities, the
feasibility of such a practice must first be evaluated in terms of
changes in oil quality, which are also related to the cultivar.

Acknowledgements

We thank Bejo Zaden (Holland) for kindly providing the

seeds of turnip-rooted parsley, and S. Colovos, V. Tsagaraki and
V. Petroulea for technical assistance. The financial support of
the Greek National Foundation of Scholarships is gratefully
acknowledged.

References

Adams, R.P., 2001. Identification of Essential Oil Components by Gas Chro-

matography/Quadrupole Mass Spectroscopy. Allured Publ. Corp., Carol
Stream, IL.

Callan, N.W., Johnson, D.L., Westcott, M.P., Welty, L.E., 2007. Herb and oil

composition of dill (Anethum graveolens L.): effects of crop maturity and
plant density. Ind. Crop Prod. 25, 282–287.

Estrada, B., Pomar, F., Do˜A

ˆ az, J., Merino, F., Bernal, M.A., 1999. Pungency

level in fruits of the Padron pepper with different water supply. Sci. Hort. 81,
385–396.

Fatima, S.F., Farooqi, A.H.A., Srikant, S., 2000. Effect of drought stress

and plant density on growth and essential oil metabolism in citronella
java (Cymbopogon winterianus). J. Med. Aromat. Plant Sci. 22 (1B),
563–567.

Freeman, G.G., Whenham, R.J., Self, R., Eagles, J., 1975. Volatile flavour

components of parsley leaves. J. Sci. Food Agric. 26, 465–470.

Jones, H.G., Tardieu, F., 1998. Modelling water relations of horticultural crops:

a review. Sci. Hort. 74, 21–46.

Jung, H.P., Sen, A., Grosch, W., 1992. Evaluation of potent odorants in parsley

leaves [Petroselinum crispum (Mill) Nym. ssp. crispum] by aroma extract
dilution analysis. Lebensm.-Wiss. u.-Technol. 25, 55–60.

Kasting, R., Anderson, J., Von Sydow, E., 1972. Volatile constituents in leaves

of parsley. Phytochemistry 11, 2277–2282.

Masanetz, C., Grosch, W., 1998. Key odorants of parsley leaves (Petroselinum

crispum

(Mill) Nym. ssp. crispum) by odour-activity values. Flavour Frag. J.

13, 115–124.

Petropoulos, S.A., 2006. The effect of N-fertilization and stress on the devel-

opment and the chemical composition of three parsley types. Ph.D. Thesis.
Agricultural University of Athens, Greece.

Petropoulos, S.A., Daferera, D., Akoumianakis, C.A., Passam, H.C., Polissiou,

M.G., 2004. The effect of sowing date and growth stage on the essential oil
composition of three types of parsley. J. Sci. Food Agric. 84, 1606–1610.

Petropoulos, S.A., Akoumianakis, C.A., Passam, H.C., 2006. Evaluation of

turnip-rooted parsley (Petroselinum crispum ssp. tuberosum) for root and
foliage production under a warm, Mediterranean climate. Sci. Hort. 109,
282–287.

Sarker, B.C., Hara, M., Uemura, M., 2005. Proline synthesis, physiological

responses and biomass yield of eggplants during and after repetitive soil
moisture stress. Sci. Hort. 103, 387–402.

Simon, J.E., Quinn, J., 1988. Characterization of essential oil of parsley. J.

Agric. Food Chem. 36, 467–472.

Singh, M., Ramesh, S., 2000. Effect of irrigation and nitrogen on herbage, oil

yield and water-use efficiency in rosemary grown under semi-arid tropical
conditions. J. Med. Aromat. Plant Sci. 22 (1B), 659–662.

Zehtab-Salmasi, S., Javanshir, A., Omidbaigi, R., Aly-Ari, H., Ghassemi-

Golezani, K., 2001. Effects of water supply and sowing date on performance
and essential oil production of anise (Pimpinella anisum L.). Acta Agron.
Hung. 49 (1), 75–81.

S.A. Petropoulos et al. / Scientia Horticulturae 115 (2008) 393–397

397