Structural investigations of trichomes and essential oil composition of Salvia verticillata

Abstract

Krstic L., Malencic Dj. and Anackov G. 2006. Structural investigations of trichomes and the essential oil composition of Salvia verticillata. Bot. Helv. 116: 159 – 168.
The morphology of glandular and non-glandular trichomes and the essential oil composition of Lamiaceae play an important role in the ecology of these species as well as for their industrial use; they may also serve as taxonomic criteria. We studied the trichome anatomy and essential oil composition of three wild-growing populations of Salvia verticillata in Serbia to determine how strongly these traits vary among and within Salvia species, and to evaluate their potential taxonomic or economic significance. Microscopic investigations of leaf and calyx indumentum revealed that S. verticillata has the same types of peltate and capitate trichomes as other Salvia species. In combination, however, the trichome characteristics of S. verticillata differed from previously examined Salvia species and might therefore help in species identification. The density of glandular trichomes differed among the three populations and was lowest in a dry steppe habitat, but trichome density also varied substantially among individual leaves within each population. The essential oil content (determined gravimetrically in n-hexane) ranged from 0.40% to 0.42% of dry mass. In total, 39 different oil components were identified using GC-MS. Oil composition varied considerably among the three populations. Only three of the 39 compounds were well represented (>0.1% of oil) in all three populations: The main component was germacrene D in two populations (48.0% and 24.6% of oil, respectively) but (E)- caryophyllene (10.2%) in the third. Low oil content and high variability in oil composition may restrict the industrial use of wild-growing S. verticillata plants.

Full text

Structural investigations of trichomes and essential oil
composition of Salvia verticillata
Lana Krstic1, Djordje Malencic2and Goran Anackov1
1University of Novi Sad, Faculty of Sciences, Department of Biology and Ecology,
Trg D. Obradovica 2, Novi Sad, Serbia; e-mail: lana@ib.ns.ac.yu
2University of Novi Sad, Faculty of Agriculture, Novi Sad, Serbia
Manuscript accepted 6 September 2006
Abstract
Krstic L., Malencic Dj. and Anackov G. 2006. Structural investigations of trichomes
and the essential oil composition of Salvia verticillata. Bot. Helv. 116: 159 – 168.
The morphology of glandular and non-glandular trichomes and the essential oil
composition of Lamiaceae play an important role in the ecology of these species as well
as for their industrial use; they may also serve as taxonomic criteria. We studied the
trichome anatomy and essential oil composition of three wild-growing populations of
Salvia verticillata in Serbia to determine how strongly these traits vary among and
within Salvia species, and to evaluate their potential taxonomic or economic
significance. Microscopic investigations of leaf and calyx indumentum revealed that
S. verticillata has the same types of peltate and capitate trichomes as other Salvia
species. In combination, however, the trichome characteristics of S. verticillata differed
from previously examined Salvia species and might therefore help in species
identification. The density of glandular trichomes differed among the three populations
and was lowest in a dry steppe habitat, but trichome density also varied substantially
among individual leaves within each population. The essential oil content (determined
gravimetrically in n-hexane) ranged from 0.40% to 0.42% of dry mass. In total, 39
different oil components were identified using GC-MS. Oil composition varied
considerably among the three populations. Only three of the 39 compounds were well
represented (>0.1% of oil) in all three populations: The main component was
germacrene D in two populations (48.0% and 24.6% of oil, respectively) but (E)-
caryophyllene (10.2%) in the third. Low oil content and high variability in oil
composition may restrict the industrial use of wild-growing S. verticillata plants.
Key words: Epidermis anatomy, glandular trichomes, intraspecific morphological
variation, non-glandular trichomes, secondary compounds, secretory glands.
Bot. Helv. 116 (2006): 159 – 168
0253-1453/06/020159-10
DOI 10.1007/s00035-006-0767-6
 Birkhuser Verlag, Basel, 2006
Botanica Helvetica

Introduction
Glandular trichomes are primary secretory organs of the Lamiaceae (Duke 1994;
Marin 1996; Gersbach et al. 2001). They are distributed on vegetative and reproductive
organs and produce bioactive lipophilic secondary products (essential oils), which may,
for example, contribute to the plants defence against herbivores and pathogens. Some
of these products are of interest to the pharmaceutical, food and pesticide industry. The
type and distribution of glandular trichomes can also be used for the differentiation of
closely related taxa, for example within the genus Salvia (Chakalova et al. 1993) or
Stachys (Demissew and Harley 1992). This is particularly valuable when small plant
parts (e.g. leaves) need to be identified (Chakalova et al. 1993). The anatomy of
glandular trichomes and their essential oil composition were also useful in solving
taxonomic problems within the Satureja complex (Bosabalidis 1990).
Two main types of glandular trichomes are found in the Lamiaceae – capitate and
peltate (Metcalfe and Chalk 1950 ; Marin 1996). Capitate glandular trichomes consist of
a basal cell of epidermal origin, a unicellular or multicellular stalk, and a round to pear-
shaped, uni- or multicellular secretory head. Peltate glandular trichomes consist of a
basal cell embedded in the epidermis, a short stalk cell, and a wide head made of several
secretory cells covered with a common cuticle,which can be arranged in one or two
circles, and with varying morphology. Capitate hairs are active on younger leaves, while
peltate hairs start with their secretory function once the secretion process of capitate
hairs is finished. The number of trichomes per unit plant surface varies among and
within species (Uphof and Hummel 1962; Bosabalidis and Kokkini 1997). High
trichome density tends to be associated with greater heat tolerance, light protection,
reduced transpiration and reduced predation by insects and herbivores (Dickison
2000).
Essential oils serve as a barrier against various external factors, including
herbivores, pathogens, UV-B radiation, extreme temperatures and drought, and they
may reduce transpiration or attract pollinators (Uphof and Hummel 1962; Corsi and
Bottega 1999; Valkama et al. 2004). The secretion of essential oils varies among and
within species and also depends on the density of glandular trichomes (Krstic et al.
2001). Intraspecific variation can be considerable and may be related to growth
conditions or to genetic differentiation. For example, Krstic et al. (2001) found a lower
essential oil content in Salvia glutinosa plants from a shaded forest edge (0.095%) than
in plants cultivated in full sun (0.152%); even lower contents were reported from South
Italy (0.03–0.05%) by Senatore and de Fusco (1997). Such variations may be relevant
to the commercial value of a species.
A Lamiaceae genus of particular economic relevance is the genus Salvia. In the flora
of Serbia, 15 species of Salvia have been described, some of which are well known for
their medicinal use and their aromatic and antioxidant properties (Diklic 1974). Salvia
species used in folk medicine include S. aegyptiaca,S. aethiopis,S. divinorum,S. plebeia,
S. sclarea and S. verbenaca, but the most important ones are S. officinalis,S. triloba and
S. miltiorrhiza. According to the German Pharmacopeia (Deutsches Arzneibuch DAB
10), only S. officinalis and S. triloba are widely used in Europe (Liu et al. 1995 ; Malencic
2001). Accordingly, most other wild-growing Salvia species have hardly been
investigated phytochemically or pharmacologically. One of these poorly investigated
species is Salvia verticillata L., a close relative of medicinally important Salvia species,
and therefore a species of potential pharmacological and commercial value.
160 Lana Krstic, Djordje Malencic and Goran Anackov

S. verticillata (lilac, whorled sage) is a herbaceous perennial native to central and
eastern Europe and western Asia, where it grows under semi-arid, continental climatic
conditions. Stems are up to 80 cm heigh, erect and eglandular-hairy (Hedge 1972).
Leaves are mostly simple, ovate-triangular, with 1–2 pairs of small lateral segments,
cordate to truncate at the base, with an acute tip. Inflorescences consist of (8–)15–30
lilac-blue flowers. Sefidkon and Khajavi (1999) analysed the chemical composition of
the essential oil of this species in Iran, but no chemotaxonomic data are available from
Europe.
We investigated the glandular and non-glandular trichomes of S. verticillata leaves
and calyces to determine their anatomical structure and the chemical composition of
their products. Our aims were to determine (1) whether trichome characteristics of S.
verticillata differ from those of other Salvia species, and (2) how strongly trichome
density, essential oil content and essential oil composition vary among three S.
verticillata populations from contrasting habitats.
Materials and Methods
Plant material was collected during flowering from three wild-growing populations,
representing different types of habitats where S. verticillata occurs. Population 1 was
collected from Vrdnik (North Serbia, 400 m altitude) in a mesic meadow, alliance
Arrhenatherion elatioris. Population 2 was collected from Rimski Sanac (Pannonian
part of North Serbia, 80 m above sea level) in a xeric steppe habitat, alliance Festucion
rupicolae, and population 3 from Tara mountain (Southwest Serbia, altitude 1000 m), in
a mesic meadow, alliance Arrhenatherion elatioris. Voucher specimens were deposited
in the Herbarium of the Department of Biology and Ecology, University of Novi Sad
(BUNS).
Anatomical and morphological investigations were carried out on ten plants per
population using leaves from the middle part of the plants. For light microscopy, cross
sections of fresh leaves were made at the region of the main vein and at 1/4 of leaf width
using Leica CM 1850 cryostat, at temperature 18 to 20 8C, with cutting intervals of
25 mm. Sections were investigated using the Image Analyzing System Motic 2000. The
number of glandular trichomes was counted, and the diameter of peltate trichomes
measured on fifty leaves per population. The significance of differences between
populations was determined with Duncans test using STATISTICA, version 7.0. For
scanning electron microscopy (SEM), small pieces of dry leaves and calyces of ten
plants were sputter coated with gold for 180 seconds, 30 mA (BAL-TEC SCD 005), and
viewed with a JEOL JSM-6460LV electron microscope at an acceleration voltage of 20
kV.Essential oils were isolated from dry leaves and flowering tops in full blossom of 50
plants per population by the Europaean Pharmacopeia hydrodestillation methods,
using n-hexane as collecting solvent. The extracts were dried over anhydrous sodium
sulphate and decanted. Hexane was evaporated under reduced pressure to measure oil
yields. A suitable dilution of each oil in hexane (10 mg/ml) was then analyzed by GC-
FID and GC-MS. The GC conditions used were: column, HP-5 fused silica capillary
column, 30 m0.25 mm, film thickness 0.25 mm; column temperature 50 8C for 5 min.,
then heated to 2508C at a rate of 3 8C/min on a Perkin-Elmer 8500 gas chromatograph;
carrier gas, helium, constant flow (1ml/min); injector 280 8C; FID detector 2808C. Mass
spectra were obtained from a Hewlett-Packard 5973-6890 GC-MS system operating on
Botanica Helvetica 116, 2006 161

EI mode at 70 eV, equipped with HP-5 MS capillary column (30 m  0.25 mm, film
thickness 0.25 mm). The initial temperature of the column was 608C and then it was
heated to 2808C at a rate of 38C/min. The identification of individual compounds was
made by comparison of their retention times and mass spectra with those obtained from
authentic samples and/or the NIST/NBS, Wiley libraries spectra as well as with
literature data.
Results
The indumentum of S. verticillata leaves was made of non-glandular and glandular
trichomes, which were present on both lamina surfaces. Non-glandular trichomes were
long, uniseriate, multicellular (2–6 cells), with acute apical cells. On their surface, small
papillae were noticed. A ring of raised epidermal cells surrounded larger trichomes.
Non-glandular trichomes densely covered the whole leaf surface but were more
abundant on the abaxial epidermis and along the veins.
Both peltate and capitate glandular trichomes were observed. Peltate trichomes
were present on the abaxial epidermis only, in epidermal depressions (Fig. 1a,b; 2a).
They consisted of a basal cell, a short stalk cell and a broad, round head, composed of
four secretory cells. Above the secretory cells the thick cuticle elevated to form a large
subcuticular space for an accumulation of secreted material. Pores or cracks were not
recorded on cuticle using SEM, so it could be supposed that secreted material was
released after the cuticle had been broken. The number of peltate hairs per mm2of leaf
surface and their diameter were lower on the leaves of population 2 than on those of
populations 1 and 3 (Tab. 1). The number of peltate trichomes also varied strongly
within populations, whereas their diameter showed low variability (cf. variation
coefficients in Tab. 1).
Capitate glandular trichomes were present on both leaf surfaces but more numerous
on the adaxial epidermis and along the veins (Fig. 1c,d; 2b,c,d). Three types could be
distinguished. The hairs of type I had a short stalk cell and a bicellular head. Type II had
a short unicellular stalk and a unicellular secretory head. The hairs of type III were not
very common, and their density was always low. They were larger, with a long stalk
composed of one elongated stalk cell and one short, narrower neck cell and had a pear-
Tab. 1. Number of glandular trichomes per mm2of leaf surface area and diameter of peltate
trichomes. Data are meansstandard errors and coefficients of variation (%) of 50 leaves per
population. Letters in superscript point to significance of differences between the populations
(Duncan test); means without common letter differ significantly (p<^0.05).
Population 1 Population 2 Population 3
Capitate trichomes Nr. on adaxial epidermis 6.10.2a
(24.8) 4.30.2b
(28.0) 5.90.3a
(23.7)
Nr. on abaxial epidermis 3.90.2a
(28.5) 3.70.3a
(41.0) 3.50.3a
(42.6)
Peltate trichomes Nr. on abaxial epidermis 2.10.2a
(81.0) 0.90.1b
(69.8) 1.70.2a
(64.3)
Diameter (mm) 48.70.7ab
(9.0) 46.80.7b
(7.5) 50.70.9a
(9.3)
162 Lana Krstic, Djordje Malencic and Goran Anackov

shaped unicellular secretory head. The total number of capitate hairs did not differ
among the examined populations, but their number on the adaxial epidermis was sig-
nificantly lower in population 2 than in the two other populations. The number of
capitate trichomes varied strongly within populations (Tab. 1), though less than the
number of peltate trichomes.
The calyx indumentum was composed of the same types of glandular and non-
glandular trichomes, but compared to leaves, they were more abundant (Fig. 3). The
non-glandular trichomes had larger papillae on their surfaces and were more numerous
between the calyx ribs.
The essential oil content in dry herb (leaves and flowers) of S. verticillata ranged
from 0.40% to 0.42%. In total, 39 different oil components were identified, but only 23
to 27 components in individual populations (Tab. 2). Oil composition varied conside-
rably among the three populations: only three of the 39 compounds were well re-
presented (>0.1% of oil) in all three populations (Tab. 2). For populations 1 and 2, only
eight compounds were identified in an amount higher than 0.1%, compared to 27
compounds in population 3. The dominant components in populations 1 and 2 were
germacrene D (48.0% and 24.6%, respectively) and (E)-caryophyllene (13.4% and
19.0%, respectively), whereas the dominant components in population 3 were (E)-
caryophyllene (10.2%), b-cubebene (8.6%) and eicosane (8.5%). Using GC-MC only
Fig. 1. Salvia verticillata scanning electron micrographs. A. peltate hair, four secretory cells
visible; B. peltate hair; C. capitate hair, type II; D. capitate hair, type III. Magnification scale
bar: A – C: 10 mm; D: 50 mm
Botanica Helvetica 116, 2006 163

75% of volatile compounds could be identified in population 3, suggesting that these
plants were also rich in other, non-identified volatile carbohydrates, besides hy-
drocarbons and sesquiterpenes.
Fig. 2. Salvia verticillata light micrographs, population 1. A. peltate hair; B. capitate hair, type
I; C. capitate hair, type II; D. capitate hair, type III. Magnification scale bar: A, D : 20 mm;B,C:
10 mm.
Fig. 3. Salvia verticillata scanning electron micrographs, population 1. A. calyx; B. calyx non-
glandular hairs-detail. Magnification scale bar: A: 500 mm; B: 10 mm.
164 Lana Krstic, Djordje Malencic and Goran Anackov

Tab. 2. Composition of the essential oil of three populations of Salvia verticillata, expressed as
relative percentage of each compound as obtained from peak areas (t =trace, <0.1%). Con-
stituents are sorted according to retention time in GC.
Constituents Pop. 1 Pop. 2 Pop. 3
a-pinene t t 0.5
sabinene t t –
b-pinene t t –
myrcene t t –
b-phellandrene t 8.6 1.7
camphene – – 0.6
(Z)- b-ocimene – 6 –
(E)- b-ocimene t 7.5 –
linalool t t –
terpinen-4-ol t t –
a-terpineol t t –
a-cubebene t t 0.5
a-copaene t t –
b-bourbonene t t 0.4
b-elemene t t 1.1
(E)-caryophyllene 13.4 19 10.2
b-cubebene – – 8.6
a-elemene – – 1.1
a-humulene 7.2 10.2 4.8
a-muurolene – – 1.0
germacrene D 48 24.6 –
bicyclogermacrene 5.3 16.7 –
naphthalene, 1, 2, 3, 4, 4a, 5, 6, 8a-octahydroxyl – – 4.0
g-cadinene t t 1.3
d-cadinene 6.0 t 3.7
naphthalene, 1, 2, 3, 4, 4a, 7-hexahydroxyl – – 1.3
spathulenol 3.5 7.2 6.5
caryophyllene oxide t t 2.9
g-gurjunene – – 0.9
a-cadinol 10.4 t 3.7
eudesma-4(15),7-dien-1-beta-ol 6.0 t –
caryophyllenol-11 – – 0.6
aromadendrenepoxide – – 2.1
octacosane – – 0.5
nonahexacontanoic acid – – 0.5
cyclopentane – – 4.5
eicosane – – 8.5
2-pentadecanone, 6, 10, 14-trimethyl – – 2.5
1,2-benzenedicarboxylic acid – – 2.5
Total % of identified compounds 99.8 99.8 75.4
Botanica Helvetica 116, 2006 165

Discussion
Non-glandular and glandular trichomes of S. verticillata leaves were similar to the
trichome types described previously for Lamiaceae (Tab. 3). The number of head cells
in peltate trichomes varied from four to 16 in different Salvia species (Tab. 3). Accor-
ding to our findings, peltate trichomes of S. verticillata leaves had a four-celled secretory
head. Although the structure of trichomes should be of taxonomic value and constant
for the species (Uphof and Hummel 1962), this disagreed with the findings of Remenyi
(1997a), who reported 6-, 8- and 12-celled heads of sessile glandular trichomes. Four-
celled head had also been recorded for S. blepharophylla and S. glutinosa (Tab. 3).
Capitate glandular trichomes of three types were recorded for S. verticillata. Types I and
II, with a unicellular stalk and a bi- and unicellular secretory head, respectively, cor-
responded to trichomes described for S. blepharophylla,S. tomentosa and S. officinalis
(Tab. 3). Glandular trichomes of type III of S. verticillata were similar to those described
by Corsi and Bottega (1999) for S. officinalis, type III and IV. Despite the overall
similarity of trichome structures, the combination of a four-celled head on peltate
trichomes, together with the occurrence of types I, II and III for capitate trichomes, has
not yet been described for another Salvia species, suggesting that it might be a criterion
for the identification of S. verticillata. However, information about trichomes is still
missing for the majority of the wild-growing Serbian Salvia species, so that we cannot
exclude that other species have the same trichome characteristics as S. verticillata.
Another taxonomically important finding is that trichome morphology did not differ
between leaves and calyces.
Capitate glandular trichomes were more numerous than peltate trichomes. Com-
pared with S. glutinosa,S. tomentosa and S. officinalis (Tab. 3), S. verticillata seems to
have a lower total number of glandular trichomes on the abaxial epidermis and a higher
number on the adaxial epidermis of leaves. However, the number of glandular tri-
chomes is generally a very variable character, which is strongly influenced by external
factors such as temperature and light intensity (Uphof and Hummel 1962), so that
interspecific comparisons based on a small number of populations must be considered
with caution.
The yield of the essential oil (0.40–0.42%) permits the assignment of this species to
oil-poor representatives of Lamiaceae, as the widely accepted borderline between oil-
poor and oil-rich representatives of Salvia is 1% (Mathe et al. 1992; Senatore and de
Fusco 1997; Malencic et al. 2004). Sesquiterpenes were present in all populations,
consistent with previous reports that oil-poor species of Lamiaceae possess oil rich in
sesquiterpene compounds (Malencic et al. 2004). There were striking differences in the
essential oil composition of the three populations. Monoterpenes were well reACHTUNGTRENNUNGpresent-
ACHTUNGTRENNUNGed in populations 2 and 3, but only present in traces in population 1. Hydrocarbons were
missing in populations 1 and 2, from the North of Serbia, while population 3, from
southwest mountainous regions, was rich in them. The number of compounds found in
individual populations (23–27) was similar to the number of compounds found in a
wild-growing population in Iran (Sefidkon and Khajavi 1999). Two dominant compo-
nents in Iran, (E)-caryophyllene (24.7%) and a-humulene (7.8%) were also dominant
components in all three Serbian populations, suggesting that these are generally present
in S. verticillata. Of the two other components that were present insignificant amounts
in the Iranian population, a-muurolene (22.8%) and limonene (8.9%), only a-muu-
rolene was recorded in Serbia, and only in population 3.
166 Lana Krstic, Djordje Malencic and Goran Anackov

Both environmental factors (Mathe et al. 1992) and genetic differences (Werker et
al. 1985) have been found to cause variation in oil composition in other species of
Lamiaceae. The number of populations investigated in this study does not yet allow
conclusions about factors determining the essential oil composition of S. verticillata.
However, our results suggest that low oil content and high intraspecific variability in oil
composition may restrict the commercial use of wild-growing S. verticillata plants.
This work was financially supported by the Ministry of Science and Envir onmental Protection of Serbia and
Montenegro, Grant No. 143037. We would like to thank Mr. Milos Bokorov, University Center for Electron
Microscopy, Novi Sad for his technical assistance and SEM micrographs.
References
Bisio A., Corallo A., Gastaldo P., Romussi G., Ciarallo G., Fontana N., de Tommasi N. and
Profumo P. 1999. Glandular hairs and secreted material in Salvia blepharophylla Brandegee
ex Epling grown in Italy. Ann. Bot. 83: 441 – 452.
Bosabalidis A.M. and Kokkini S. 1997. Infraspecific variation of leaf anatomy in Origanum
vulgare grown wild in Greece. Bot. J. Linn. Soc. 123: 353 – 362.
Bosabalidis A.M. 1990. Glandular trichomes in Satureja thymbra leaves. Ann. Bot. 65: 71 – 78.
Tab. 3. Non-glandular and glandular trichome characteristics of various Salvia species (ade:
adaxial epidermis; abe: abaxial epidermis; –: not determined).
Species Non-glandular
trichomes:
nr. of cells
Peltate
trichomes:
nr. of head
cells
Capitate
trichome type Mean
number of
glandular
trichomes
mm2
References
ade abe
S. verticillata 3–6 4 I, II and III 5.7 5.3 This study
S. verticillata – 6, 8 and 12 – – – Remenyi (1997a)
S. aethiops 3 or more 16 – – – Remenyi (1997b)
S. aurea – 6–8 I* – – Serrato-Valenti et
al. (1997)
S. blepharo-
phylla 1, 4–5 or more 4 I and II – – Bisio et al. (1999)
S. glutinosa 2 –7 4 I 8.0 24.5 Krstic et al. (2001)
S. officinalis 3–4 12 I and II 3.0 6.3 Venkatachalam et
al. (1984)
Corsi and Bottega
(1999)
2–4 III and IV Chakalova et al.
(1993)
S. officinalis f.
aurea 2–4 12 4.8 8.7 Chakalova et al.
(1993)
S. tomentosa 2 –4 12 I and II 0.66 10.6 Chakalova et al.
(1993)
* plus trichomes with multicellular stalk and unicellular head
Botanica Helvetica 116, 2006 167

Chakalova E.S., Genova E.M. and Thuy N.T. 1993. Comparative anatomical investigations on
leaves of Salvia officinalis L. and S. tomentosa Mill. Fitologija 46: 3 – 11.
Corsi G. and Bottega S. 1999. Glandular hairs of Salvia officinalis: new data on morphology,
localization and histochemistry in relation to function. Ann. Bot. 84 : 657 – 664.
Demissev S. and Harley M.M. 1992. Trichome, seed surface and pollen characters in Stachys
(Lamioidae: Labiatae) in tropical Africa. In: Harley R. M. and Reynolds T. (eds.). Ad-
vances in Labiate Science. Royal Botanic Gardens, Kew, 149 – 166.
Dickison W.C. 2000. Integrative Plant Anatomy. Academic Press, San Diego, 297 – 307.
Diklic N. 1974. Rod Salvia L. In: Josifovic M (ed.). Flora SR Srbije VI. SANU, Beograd, 432 –
453.
Duke S.O. 1994. Glandular trichomes – focal point of chemical and structural interactions. Int.
J. Plant Sci. 155: 617 – 620.
Gersbach P.V., Wyllie S.G. and Sarafis V. 2001. A new histochemical method for localization of
the site of monoterpene phenol accumulation in plant secretory structures. Ann. Bot. 88:
521 – 525.
Hedge I.C. 1972. Genus Salvia L. In: Tutin et al. (eds.). Flora Europaea, Vol. 3, Cambridge
University Press, London, p. 192.
Krstic L., Malencic Dj., Merkulov Lj., Popovic M., Adamovic D. and Boza P. 2001. A con-
tribution to the morpho-anatomical study of wild-growing and cultivated Salvia glutinosa L.
Med. Plant Rep. 8 (8): 38 – 47 (in Serbian, English summary).
Liu J., Zapp J. and Becker H. 1995. Comparative phytochemical investigation of Salvia mil-
tiorrhiza and Salvia triloba. Planta Med. 61: 453 – 455.
Malencic Dj. 2001. Biochemical investigations of selected wild-growing Salvia species from
Vojvodina province. PhD Thesis, University of Novi Sad, Faculty of Science (In Serbian,
English summary).
Malencic Dj., Couladis M., Mimica-Dukic N., Popovic M. and Boza P. 2004. Essential oils of
three Salvia species from the Pannonian part of Serbia. Flavor Fragr. J. 19: 225 – 228.
Marin P. 1996. Orasice i trihome u familiji Lamiaceae. Bioloski fakultet, Belgrade.
Mathe Jr.I., Olah L. , Mathe A., Miklossy V., Bernath J., Blunden G., Patel A. and Mathe I.
1992. Changes in the essential oil production of Salvia officinalis under climatic conditions
of the temperate belt. Planta Med. 58: 680 – 686.
Metcalfe C.R. and Chalk L. 1950. Anatomy of the dicotyledons, Vol II. Clarendon Press,
Oxford.
Pharmacop Europenne, vol. 1. 1983. Maisonneuve SA, Sainte-Ruffine.
Remnyi M.L. 1997a. Research on epidermal appendices of Salvia aethiops L. Magyar Nç-
vnyanatmiai Szimpzium, Abstracts, Szeged, 89.
Remnyi M.L. 1997b. Variations in tissue structures on the epidermis of Salvia verticillata L.
Magyar Nçvnyanatmiai Szimpzium, Abstracts, Szeged, 91.
Sefidkon F. and Khajavi M.S. 1999. Chemical composition of the essential oils of two Salvia
species from Iran: Salvia verticillata L. and Salvia santolinifolia Boiss. Flavor Fragr. J. 14:
77 – 78.
Senatore F. and de Fusco R. 1997. Essential oils from Salvia spp. (Lamiaceae). I. Chemical
composition of the essential oils from Salvia glutinosa L. growing wild in Southern Italy. J.
Essent. Oil Res. 9: 151 – 157.
Serrato-Valenti G., Bisio A., Cornara L. and Ciarallo G. 1997. Structural and histochemical
investigation of the glandular trichomes of Salvia aurea L. leaves and chemical analysis of
the essential oil. Ann. Bot. 79 : 329 – 336.
Uphof J.C.T. and Hummel K. 1962. Plant Hairs. Borntraeger, Berlin.
Valkama E., Salminen J-P., Koricheva J. and Pihlaja K. 2004. Changes in leaf trichomes and
epicuticular flavonoids during leaf development in three birch taxa. Ann. Bot. 94: 233 – 242.
Venkatachalam K.V., Kjonaas R. and Crotean R. 1984. Development and essential oil content
of secretory glands of sage (Salvia officinalis). Plant Physiol. 76: 148 – 150.
Werker E., Putievsky E. and Rivid V. 1985. The essential oils and glandular hairs in different
chemotypes of Origanum vulgare L. Ann. Bot. 55 : 793 – 801.
168 Lana Krstic, Djordje Malencic and Goran Anackov