untitled ACTA BOT. CROAT. 75 (1), 2016 67 Acta Bot. Croat. 75 (1), 67–73, 2016 CODEN: ABCRA 25 DOI: 10.1515/botcro-2016-0002 ISSN 0365-0588 eISSN 1847-8476 Anatomical characteristics of two Ornithogalum L. (Hyacinthaceae) taxa from Serbia and Hungary and their taxonomic implication Andrijana M. Andrić*, Milica M. Rat, Lana N. Zorić, Jadranka Ž. Luković Department of Biology and Ecology, Faculty of Science, University of Novi Sad, Trg Dositeja Obradovića 2, 21000 Novi Sad, Serbia Abstract – Anatomical characters of two morphologically similar Ornithogalum taxa, O. umbellatum and O. divergens, were investigated. An analysis of leaf and scapus cross-sections was performed on plants from ten populations from Serbia and Hungary, using light microscopy. The aim of this research was to give data about the qualitative and quantitative anatomical characteristics of these taxa, in order to evaluate their taxonomic signifi cance and single out distinctive anatomical features, as well as to contribute to the knowledge of the genus Ornithogalum in the studied region. On the basis of the variability of anatomical characters, similar populations formed two clusters, joining the plants previously determined as O. divergens and O. umbella- tum. The two taxa signifi cantly differed for most of the quantitative leaf and scapus characters. Since only quantitative differences were recorded in this research, anatomical characters could not be solely used to separate these two taxa. However, the results of anatomical investigations are consistent with the results of previous morphological and genetic analyses; therefore anatomical parameters could be useful as additional taxonomic characters. Key words: anatomy, leaf, Ornithogalum, scapus * Corresponding author, e-mail: andrijana.andric@dbe.uns.ac.rs Introduction Genus Ornithogalum L. is one of the most abundant in species and taxonomically the most interesting in the fami- ly Hyacinthaceae Batsch ex Borckhausen 1797 (Order As- paragales Bromhead, subfamily Ornithogaloideae) (Man- ning et al. 2009, Martínez-Azorín et al. 2011). The defi nition of taxa within this group has troubled taxono- mists for a long time (Martínez-Azorín et al. 2011); an ac- curate number of belonging species is still a controversial issue. Ornithogalum is native in Europe, Asia (as far as Af- ghanistan in the East) and Africa, where it is widely distrib- uted (Martínez-Azorín et al. 2010). The Mediterranean and South Africa are considered to be the centers of distribution (Zahariadi 1980, Cullen 1984). Natural habitats of these bulbous monocots are very heterogeneous; they are in bloom in spring and summer and the fl owers are entomophilous. These plants are important hosts for insects, especially for some groups of hoverfl ies, whose adults feed on pollen and nectar and larvae develop in the bulbs of these geophytes (Vujić et al. 2012). Some species of the genus are of importance in the fl ower trade, being grown as cut fl ower or potted plant crops, while some are poisonous to livestock and considered weeds (Ober- meyer 1978, Littlejohn and Blomerus 1997). Eleven subgenera and 34 species of the genus Ornitho- galum have been described in the Flora Europaea (Zaharia- di 1980) and Ornithogalum umbellatum L. 1753 and Orni- thogalum divergens Boreau 1887 are given as separate species. Thirteen Ornithogalum species were recorded in the Flora of Serbia (Diklić 1975), whilst eight were record- ed in Hungarian fl ora (Soó 1973). O. divergens was de- scribed as a subspecies of O. umbellatum in both of them. Ornithogalum umbellatum is a type species of the genus Ornithogalum. It is widespread in Europe, North Africa and Southwestern Asia and was introduced in North America (EuroPlusMed 2006, eFloras 2008). The species rank of this taxon is not questionable; however, the intraspecifi c systematics are often interpreted differently. The taxonomic position of Ornithogalum divergens has had lots of differ- ent interpretations (Martínez-Azorín et al. 2009); it was considered a hexaploid form of the polyploid complex of O. umbellatum (Moret et al. 1991, Moret 1992), a subspe- cies, a variety and it has also been widely misidentifi ed as ANDRIĆ A. M., RAT M. M., ZORIĆ L. N., LUKOVIĆ J. Ž. 68 ACTA BOT. CROAT. 75 (1), 2016 O. umbellatum. Nevertheless, Martínez-Azorín et al. (2009, 2010) presented its nomenclature and taxonomy as well as its ecology and distribution and described this species as clearly different from O. umbellatum s. s., primarily for its ploidy level, morphology of bulbils and infl orescence struc- ture. The need for redefi nition of the infrageneric taxa of the genus Ornithogalum was suggested by Garbari et al. (2003, 2008) in Italy. Herrmann (2002) deals with Ornithogalum umbellatum s. l. in Central Europe. A biometric study of O. umbellatum s. l. was done by Moret et al. (1991) in France and Moret (1992) investigated ploidy levels using numeri- cal taxonomy within this genus, concluding that species from O. umbellatum s. l. from France exhibit considerable variability. Systematic studies had been made previously on this complex by Van Raamsdönk (1986) who singled out O. umbellatum as the most widespread species among Euro- pean representatives of the genus and a variable wide-rang- ing taxon with populations showing different morphological character combinations. The nomenclature and taxonomy of O. umbellatum s. l. continue to be investigated in recent years in Spain (Martínez-Azorín et al. 2009), and taxonom- ic revision of subgenus Ornithogalum was also done for the Spanish species including Ornithogalum divergens (Mar- tínez-Azorín et al. 2010). Many authors consider the genus Ornithogalum incon- venient from the standpoint of systematics stating that its morphology is insuffi ciently connected with variations in karyotype (Martínez-Azorín et al. 2010). Since this addi- tionally complicates taxonomic delimitations, anatomical data could make a particularly useful contribution to the so- lution of taxonomical problems (Meriç et al. 2011). Studies regarding anatomical characteristics referring to these taxa have not been numerous. A comprehensive view on leaf anatomy in relation to the systematics of the whole Hyacinthaceae family was given by Lynch et al. (2006). Morpho-anatomical variability of the leaves among differ- ent taxa of the genus Ornithogalum was also investigated by Peruzzi et al. (2007). Differences in leaf anatomy be- tween Ornithogalum nutans L. and Ornithogalum narbo- nense L. from Bulgaria were presented by Popova and An- astasov (1997). Morphological and anatomical analyses were conducted for leaf and scapus of O. nutans and Orni- thogalum boucheanum (Kunth) Aschers from Turkey (Mer- iç et al. 2011). Comparative anatomical studies of leaf and scapus features of twelve Ornithogalum species, belonging to the subgenera Ornithogalum and Beryllis, were recently done in Central Anatolia (Öztürk et al. 2014). Types of crystals are especially interesting from a taxonomic point of view and Tilton and Lersten (1981) described idioblasts with calcium oxalate crystals as a type of specialized cells in the gynoecium of Ornithogalum caudatum Aiton. Although the examined taxa have been research sub- jects in many fi elds worldwide, detailed investigations and publications on their anatomy in particular are very scarce, especially in this part of Europe. The main goal of this study is to contribute to the knowledge of the qualitative and quantitative anatomical characteristics of leaf and sca- pus of two Ornithogalum taxa, from several populations from Serbia and Hungary, with the aim of evaluating their taxonomic signifi cance and singling out distinctive anatom- ical features potentially useful as diagnostic characters. Materials and methods Plant sampling was carried out during the fl owering pe- riod, in April and May (2008–2013). Specimens for analy- sis were collected from ten different native populations in Serbia and Hungary (Tab. 1, Fig. 1). Plants were deter- mined and voucher specimens deposited in the Herbarium BUNS (2-1632 – 2-1641) at the Department of Biology and Ecology, Faculty of Science, University of Novi Sad. Ten specimens from each population were fi xed in 50% ethanol. Cross-sections of the middle parts of leaf and scapus, 60 μm thick on average, were made using a Leica CM 1850 cryostat at a temperature of – 20 °C. Anatomical characters were observed by light microscopy, using the image analyz- ing system Motic 2000. Ten cross-sections were examined for each population. In total, 31 quantitative characters were measured and compared, and relative proportions were calculated. The total cross-section area of leaf and scapus was measured and percentages of other parameters were determined considering total cross-section area as 100%. Palisade cell indices were calculated as the ratio of their height and width. Data were statistically processed using STATISTICA for Windows version 12.0 (StatSoft, 2014). The signifi - cance of differences in the analysed parameters among all Tab. 1. Analysed populations of Ornithogalum umbellatum and O. divergens in Serbia (SRB) and Hungary (HUN). Abb – abbreviation. Taxon Sampling locality Abb Coordinates Sampling date Voucher O. umbellatum Deliblatska Peščara, SRB DE N44°59’16” E20°56’42” 18.04.2008 2-1632 O. umbellatum Vršačke Planine, SRB VR N45°07’22” E21°19’38” 26.04.2009 2-1633 O. umbellatum Zlotska Klisura, SRB ZL N44°01’45” E21°57’28” 10.05.2009 2-1634 O. umbellatum Valjevo, SRB VA N44°25’43” E19°54’35” 29.04.2012 2-1635 O. umbellatum Mecsek, HUN ME N46°05’56” E18°13’07” 21.04.2013 2-1636 O. umbellatum Szeged, HUN SZ N46°15’10” E19°38’36” 22.04.2013 2-1637 O. divergens Titelski Breg, SRB TI N45°17’01” E20°14’29” 20.04.2012 2-1638 O. divergens Susek, SRB SU N45°13’51” E19°28’38” 19.04.2012 2-1639 O. divergens Novi Sad, SRB NS N45°14’51” E19°49’43” 26.04.2013 2-1640 O. divergens Hódmezövásárhely, HUN HO N46°50’42” E20°31’47” 22.04.2013 2-1641 ORNITHOGALUM ANATOMY ACTA BOT. CROAT. 75 (1), 2016 69 the populations was established using Duncan’s test (p ≤ 0.05), and between Ornithogalum umbellatum and Ornitho- galum divergens applying t-test (p ≤ 0.05). The general structure of sample variability was determined by principal component analysis (PCA), based on a correlation matrix. Multivariate discriminant function analysis (MDA) was done in order to test the hypothesis that the analysed sample was composed of groups which differed from each other according to anatomical features of leaf and scapus. Any characters that were not signifi cantly different between the two taxa, according to the results of t-test, were not includ- ed in MDA. Results Leaf anatomy Cross section of the leaf is U-shaped, slightly curved to- wards the dorsal side (Fig. 2). It is signifi cantly smaller in Ornithogalum umbellatum, with an average total area of 1.2–2.2 mm2, compared to 2.8–3.7 mm2 in Ornithogalum divergens. Adaxial side is smooth while the abaxial has 6–8 or 8–10 more or less prominent ribs, in O. umbellatum and O. divergens, respectively. Epidermis is single layered, cov- ered with a thin cuticle. Epidermal cells are round in shape, or slightly elongated, smaller and more elongated on the abaxial side. The leaves are amphistomatic, with fewer sto- mata on the adaxial side, positioned at the same level as epidermal cells. There are no trichoma on the epidermis. Mesophyll is composed of chlorenchyma differentiated into palisade and spongy tissue. Palisade tissue is single layered and present on both adaxial and abaxial sides. However, it is partly missing adaxially due to presence of large lacunae. Chlorenchyma cells are absent in the region of the main vein, and broad lacunae appear subepidermally, inter- spersed with irregularly shaped parenchyma cells. Palisade tissue cells are elongated, narrower towards the the adaxial side, perpendicular to lamina surface (Fig. 3). Chloroplasts are more numerous closer to the cell walls that surround intercellulars. Spongy tissue cells are round or ir- regular in shape, with fewer chloroplasts than palisade cells. Intercellulars are larger in spongy tissue (Figs. 2, 3). The cells in the middle part of the mesophyll are larger, and do not contain chloroplasts. Some of the large mesophyll cells contain bundles of needle-shaped CaOx raphides (Figs. 3 A, C). Vascular bundles are arranged in two rows, the bigger ones centrally located, and the smaller ones towards abaxial side. Ornithogalum umbellatum has 7–10 vascular bundles of both types, whilst they are more numerous in Ornithoga- lum divergens (10–11 abaxial and 10–12 central vascular bundles). Vascular bundles are surrounded with one to two layers of parenchyma cells (Figs. 2, 3). Large lacunae are present in mesophyll, between the bundles (Figs. 3 A, B). Scapus anatomy Cross section of the scapus is round to oval in shape, with slightly wavy edges and larger total cross-section area in Ornithogalum divergens (8.7–10.4 mm2) than in Orni- Fig. 1. Localities of the analysed populations in Serbia and Hun- gary. For abbreviations of sample localities see Tab. 1. Fig. 2. Leaf cross section: A) Ornithogalum umbellatum (Vršačke Planine), B) O. umbellatum (Zlotska Klisura), C) O. divergens (Susek), D) O. divergens (Novi Sad). Abbreviations: ri – rib. Scale bars = 100 μm. Fig. 3. Leaf cross section: A) Ornithogalum umbellatum (Delib- latska Peščara), B) O. divergens (Titelski Breg), C) O. umbellatum (Szeged), D) O. divergens (Novi Sad). Abbreviations: ade – adax- ial epidermis, abe – abaxial epidermis, sto – stomata, pt – palisade tissue, st – spongy tissue, cvb – central vascular bundle, avb – ab- axial vascular bundle, la – lacunae, cc – calcium oxalate crystals. Scale bars = 100 μm. ANDRIĆ A. M., RAT M. M., ZORIĆ L. N., LUKOVIĆ J. Ž. 70 ACTA BOT. CROAT. 75 (1), 2016 thogalum umbellatum (3.0–7.7 mm2) (Figs. 4 A, B). Epider- mis is one-layered, with round shaped cells in cross section, glabrous, with thin cuticle and a few stomata (Figs. 4 C, D). Cortex consists of 4–5 layers of spherical parenchymatous cells, some of which contain chloroplasts. Idioblasts with CaOx raphide crystals are sometimes observed in cortex, beneath epidermis. Pith is predominantely composed of pa- renchymal tissue. In its peripheral part 1 to 3 layers of scle- renchyma occur. Small vascular bundles are located within sclerenchyma and just below it. Their number is very vari- able, (12–27 in O. umbellatum and 17–27 in O. divergens), depending on total cross-section area and the area of scle- renchyma, in sense that they are more numerous if these areas are larger. Larger, colateral vascular bundles, 10 to 20 of them in O. umbellatum and 17–24 in O. divergens, sur- rounded with parenchyma sheath, are randomly distributed in the pith. Their number is usually smaller than the number of small sclerenchyma bundles. The bundles become larger towards the central part of the scapus. Interpopulation variability of the anatomical characters Duncan’s test and t-test showed that the plants previ- ously determined as Ornithogalum divergens (populations HO, NS, TI and SU) signifi cantly differed from those deter- mined as Ornithogalum umbellatum (DE, VR, ZL, VA, ME and SZ), for most of the quantitative characters (Tab. 2, On- line Suppl. Tabs. 1, 2). Considering leaf characters, the two taxa had a similar number of stomata, percentage of vascu- lar tissue and size of adaxial epidermal cells. O. umbella- tum had s signifi cantly smaller leaf cross-section area, with better developed palisade tissue, composed of signifi cantly larger and more elongated cells, than O. divergens. Al- though O. umbellatum had a signifi cantly lower number of vascular bundles, the total percentage of vascular tissue was not statistically different between the two taxa. Compara- tive analysis of scapus characters revealed that O. umbella- tum had a signifi cantly smaller scapus cross-section area, with better developed sclerenchyma and vascular tissue, than O. divergens. However, the percentage of pith paren- chyma was lower in O. umbellatum and parenchyma tissue was composed of signifi cantly smaller cells. Signifi cant dif- ferences between the two taxa were not recorded in per- centages of scapus cortex and pith, size of epidermal cells and the number of small vascular bundles. The variation of the anatomical parameters was exam- ined by PCA (On-line Suppl. Tab. 3). The fi rst principal component accounted for 25.86% of total variation. Param- eters that contributed the most to the total variability were percentages of palisade and spongy tissue, scapus cross- section area and percentage of scapus epidermis. The sec- ond component represented 18.91% of variation and was defi ned by the percentage of scapus cortex. The cumulative contribution percentage of the fi rst three PCs was 53.74%. The projection of the cases of the fi rst two components demonstrated that the investigated specimens could be sep- arated into groups according to the variability of anatomical parameters (Fig. 5). Although the groups representing pop- ulations were heterogeneous, Ornithogalum divergens and Ornithogalum umbellatum populations clustered, but were not completely distinctive from each other. Characters showing no signifi cant differences between the two taxa were not included in MDA. The results of the MDA (On-line Suppl. Tab. 4) indicated similarities in ana- tomical characters between the populations from the same group. Ornithogalum divergens and Ornithogalum umbel- latum populations were clearly separated along the fi rst dis- criminant axis (Fig. 6). The parameter that contributed most to the discrimination of the populations, as well as of the two taxa, was leaf cross-section area, which was signifi - cantly higher in O. divergens. Percentages of palisade and spongy tissue and percentages of scapus epidermis and pith parenchyma singled out as characters that also had a share in separation of populations and taxa. Fig. 4. Scapus cross section: A) Ornithogalum umbellatum (Valje- vo), B) O. divergens (Titelski Breg), C) O. umbellatum (Zlotska Klisura), D) O. divergens (Titelski Breg). Abbreviations: ep – epi- dermis, co – cortex, sc – sclerenchyma, svb – small vascular bun- dle, lvb – large vascular bundle, pp – pith parenchyma. Scale bars: 100 μm. Fig. 5. The projection of the specimens of the fi rst two compo- nents of the principal component analysis based on anatomical characteristics. Ellipses indicate populations of different taxa (right – Ornithogalum umbellatum, left – O. divergens) and repre- sent 95% of each sample. For abbreviations of sample localities see Tab. 1. ORNITHOGALUM ANATOMY ACTA BOT. CROAT. 75 (1), 2016 71 Discussion In order to contribute to currently insuffi cient knowl- edge about the genus Ornithogalum in the studied region, leaf and scapus anatomical characteristics of Ornithogalum umbellatum and Ornithogalum divergens were analysed. This type of research represents an addition to previous and a foundation for future Ornithogalum studies, especially when it comes to problematic taxa such as these; however there are different opinions on the potential taxonomical importance of this approach. Peruzzi et al. (2007) conclud- ed that leaf anatomical characteristics of Ornithogalum are useful for grouping similar species, while they are usually not suffi cient for the characterization of every taxon indi- vidually. Thus they are the most valuable for general phylo- genetic recapping and as a supplement to other types of analyses. Investigating leaf and scapus anatomy of species Fig. 6. The projection of the populations of the fi rst two factors of the multivariate discriminant analysis based on anatomical charac- teristics. For abbreviations of sample localities see Tab. 1. Tab. 2. Anatomical characteristics of Ornithogalum umbellatum and O. divergens: mean value ± standard error and coeffi cient of varia- tion % (in parenthesis). Asterisk (*) indicates signifi cant differences between the two taxa and “ns” stands for “not signifi cant”, according to t-test (p ≤ 0.05). Character / Taxon O. umbellatum O. divergens t-test total leaf cross-section area (mm2) 1.7 ± 0.1 (34) 3.2 ± 0.1 (27) * % leaf epidermis 11.9 ± 0.3 (20) 8.7 ± 0.3 (19) * % palisade tissue 20.7 ± 0.5 (20) 13.0 ± 0.7 (36) * % spongy tissue 64.3 ± 0.8 (9) 75.8 ± 0.8 (7) * % vascular tissue 2.7 ± 0.2 (44) 2.5 ± 0.1 (18) ns cross-section area of adaxial epidermis cells (μm2) 704 ± 34.5 (38) 784 ± 38.7 (31) ns cross-section area of abaxial epidermis cells (μm2) 504 ± 21.4 (33) 600 ± 25.2 (27) * cross-section area of adaxial palisade tissue cells (μm2) 1499 ± 49.2 (25) 1179 ± 95.4 (51) * cross-section area of abaxial palisade tissue cells (μm2) 1578 ± 40.4 (20) 1243 ± 100 (51) * leaf thickness (mm) 0.6 ± 13.2 (16) 0.7 ± 16.6 (14) * adaxial palisade cells index 2.5 ± 0.1 (26) 1.5 ± 0.1 (40) * abaxial palisade cells index 3.1 ± 0.1 (25) 1.9 ± 0.1 (40) * number of ribs on abaxial side 7 ± 0.2 (18) 9 ± 0.3 (23) * number of abaxial vascular bundles 9 ± 0.2 (14) 10 ± 0.3 (18) * number of central vascular bundles 9 ± 0.2 (15) 11 ± 0.3 (18) * number of stomata on adaxial epidermis on cross-section 7 ± 0.4 (45) 7 ± 0.4 (38) ns number of stomata on abaxial epidermis on cross-section 12 ± 0.5 (34) 13 ± 0.7 (38) ns total scapus cross-section area (mm2) 5.7 ± 0.3 (45) 9.6 ± 0.5 (32) * % scapus epidermis 4.8 ± 0.2 (31) 3.1 ± 0.2 (32) * % scapus cortex 20.3 ± 0.9 (33) 20.5 ± 0.6 (19) ns % scapus pith 74.8 ± 1.0 (10) 76.4 ± 0.7 (6) ns % scapus sclerenchyma 6.3 ± 0.2 (26) 4.4 ± 0.2 (22) * % scapus small vascular bundels 0.6 ± 0.03 (33) 0.3 ± 0.02 (29) * % scapus large vascular bundles 2.6 ± 0.1 (26) 2.2 ± 0.1 (24) * % scapus pith parenchyma 65.3 ± 1.1 (13) 69.5 ± 0.7 (7) * cross-section area of scapus epidermal cells (μm2) 725 ± 26.7 (29) 738 ± 38.2 (33) ns cross-section area of scapus cortex parenchyma cells (μm2) 1176 ± 39.2 (26) 1493 ± 55.1 (23) * cross-section area of pith parenchyma cells (μm2) 5001 ± 196 (30) 6652 ± 269 (26) * number of scapus cortex cell layers 4 ± 0.1 (17) 5 ± 0.1 (12) * number of small vascular bundles (in sclerenchyma) 20 ± 1.0 (37) 22 ± 1.0 (26) ns number of large vascular bundles (in parenchyma) 16 ± 0.7 (35) 20 ± 0.9 (30) * ANDRIĆ A. M., RAT M. M., ZORIĆ L. N., LUKOVIĆ J. Ž. 72 ACTA BOT. CROAT. 75 (1), 2016 belonging to two different subgenera (Ornithogalum and Beryllis), Öztürk et al. (2014) came to a similar conclusion. One of the latest anatomical researches of Ornithogalum was done by Meriç et al. (2011) on Ornithogalum bouchea- num and Ornithogalum nutans in Turkey. It is diffi cult to distinguish these two species morphologically, thus ana- tomical characters could be particularly useful. They dis- covered no differences in scapus cross-sections, except in total sectional area and in the number of vascular bundles, features that were not considered of anatomical importance (Meriç et al. 2011). Similarly, Öztürk et al. (2014) found no signifi cant variability on interspecies level, although they noted that plants belonging to subg. Beryllis had larger sca- pus diameter than members of subg. Ornithogalum, as well as more numerous vascular bundles. Our research has shown that O. umbellatum has a smaller scapus total cross- section area than O. divergens, with better developed scle- renchyma and vascular tissue, and a lower percentage of parenchyma tissue. However, no signifi cant qualitative dis- tinctions were found among O. umbellatum and O. diver- gens samples regarding scapus characteristics. On the other hand, it has been recorded that leaves differ in their anato- my in different species of this genus. Namely, leaves of O. nutans and Ornithogalum narbonense are presented as epi- stomatic (Popova and Anastasov 1997), while the O. um- bellatum and O. divergens studied here have amphistomatic leaves, as was shown for all 12 species investigated by Öz- türk et al. (2014), including O. umbellatum. Mesophyll of O. narbonense consists of two types of cells, which form palisade tissue and spongy parenchyma, while O. nutans mesophyll is composed of three cell types of palisade cells: elongated, rounded and irregularly shaped (Popova and An- astasov 1997). Meriç et al. (2011) have found mesophyll of O. nutans to be thinner, unifacial and to consist of mono- typic chlorenchyma cells, while O. boucheanum has thicker equifacial mesophyll differentiated into palisade and spongy tissue, with large lacunae between vascular bundles and spongy tissue cells (Meriç et al. 2011). Leaves of the O. umbellatum and O. divergens studied here, like those of O. narbonense (Popova and Anastasov 1997), O. boucheanum (Meriç et al. 2011) and various members of Ornithogalum and Beryllis subgenera (Öztürk et al. 2014), have chloren- chyma differentiated into palisade and spongy tissue. Pali- sade tissue is one-layered and present on both adaxial and abaxial sides of the O. umbellatum and O. divergens leaves analysed in our study. Uniseriate upper and lower palisade parenchyma was also shown in some species belonging to subg. Ornithogalum, including O. umbellatum, while mem- bers of subg. Beryllis had only abaxial palisade tissue layer (Öztürk et al. 2014). Well defi ned, large lacunae appear in spongy tissue. Popova and Anastasov (1997) noted that the adaxial and abaxial epidermis become contiguous margin- ally at the edges of leaf blade in O. narbonense and consid- ered this a typical feature of the species. O. umbellatum and O. divergens have cells of the upper and lower epidermis clearly separated by palisade tissue to the very edge, as shown in O. nutans (Popova and Anastasov 1997). In most of the representatives of the subfamily Ornithogaloideae vascular bundles in the leaf are distributed in two rows and bigger alternate with smaller (Lynch et al. 2006), which is also the case in our taxa. There is a wide range of variations regarding shapes of crystals in the Hyacinthaceae family. Prychid and Rudall (1999) noted that different calcium oxalate crystal types, their presence or absence in monocotyledons, may repre- sent useful taxonomic characters in certain groups, due to their specifi city and constancy. They observed the presence of raphides in this family. Lynch et al. (2006) found solitary styloides in leaves, which are considered to be a diagnostic feature for certain families; they noted raphides and transi- tional forms between styloids and raphids, as well as crystal druses. In Ornithogalum nutans and Ornithogalum bouche- anum, only raphides, immersed in the mucus of the paren- chyma cells of scapus and leaves, were detected (Meriç et al. 2011). The presence of needle-like crystals in parenchy- ma cells of leaves and scapus, singly as well as in small groups of a few, was observed in Ornithogalum umbella- tum and Ornithogalum divergens in our study. They are considered to be transitional forms between styloides and raphides, already mentioned in related taxa (Prychid and Rudall 1999, Lynch et al. 2006). It has been noted that raphides and styloids could be mutually exclusive, but if both types are present then intermediate forms occur, with two or three crystals per cell (Prychid and Rudall 1999). Chiappini (1962) recorded raphides in the parenchymatous tissues of the leaf and in several parts of the fl oral region of O. caudatum. Crystals, their presence in plants, their mor- phology and distribution, were noted to be very important features (Franceschi and Nakata 2005), thus constancy of the type and characteristics of crystals might be considered potential taxonomic characters. Variety of shapes of calci- um oxalate crystals seems to be repeatable throughout the generations, illustrating consistency of genetic and physio- logical parameters controlling them (Prychid and Rudall 1999) and further studies of this aspect could be usefull for systematic analysis of this group. Statistical analyses pointed out that most of the anal- ysed anatomical characters have shown signifi cant differ- ences among observed populations. Observed differences were mostly quantitative, not qualitative. The most variable scapus characters among populations were total cross-sec- tion area and percentages of epidermis and cortex. Among the leaf characters, percentages of palisade and spongy tis- sue contributed most to the total interpopulational variabil- ity. The type of variability of these characters separated specimens that belong to two taxa, Ornithogalum umbella- tum and Ornithogalum divergens, into different groups. The size of leaf cross-section area proved to be the most important discriminative character between the two taxa. Moreover, the percentage of palisade tissue is signifi cantly higher in Ornithogalum umbellatum and that of spongy tis- sue in Ornithogalum divergens; these are the leaf characters that contribute to discrimination of taxa with somewhat lower proportions. According to the PCA and MDA results, the ten popula- tions could be classifi ed into two groups: one joining the plants previously determined as Ornithogalum divergens and the other one comprising Ornithogalum umbellatum populations. T – test showed signifi cant differences between ACTA BOT. CROAT. 75 (1), 2016 73 ORNITHOGALUM ANATOMY the two taxa in most of the analysed characters. These results are consistent with the studies of morphological charac teristics and ploidy levels (Garbari et al. 2003, 2008, Martí nez-Azorín et al. 2009, 2010). However, anatomical characters alone could not be used to separate these morpho- logically similar taxa, due to the fact that only quantitative, not qualitative anatomical differences were recorded be- tween them. These parameters could be a useful additional tool in resolving taxonomical problems, yet not for the char- acterization of a single species. They enabled a successful grouping of similar populations, but still remained insuffi - ciently reliable for precise delimitation of O. divergens from O. umbellatum. Acknowledgements The authors would like to thank Tijana Nikolić, Sonja Mudri-Stojnić and Dušanka Krašić, colleagues who helped us with this work. The study was supported by the Ministry of Education, Science and Technological Development, Re- public of Serbia, Grant No. 173002. References Chiappini, M., 1962: Osservazioni carioembriologiche in Ornitho- galum di Sardegna: Nota I. – Sporogenesi e sviluppo dei ga- me tofi ti in Ornithogalum caudatum Ait. Giornale Botanico Italiano 69, 91–102. Cullen, J., 1984: Ornithogalum L. In: Davis, P. H. (ed.), Flora of Turkey and the East Aegean Islands, 227–244. Edinburgh University Press, Edinburgh. Diklić, N., 1975: Genus Ornithogalum (Liliales – Liliaceae). In: Јosifović, М. (ed.), Flora SR Srbije VII, 544–559. Serbian Academy of Sciences and Arts, Belgrade (in Serbian). EuroPlusMed PlantBase, 2006: Distribution map. Retrieved Janu- ary 21, 2015 from http://ww2.bgbm.org/EuroPlusMed/PTax- onDetail.asp?NameId=3452&PTRefFk=8000000. Flora of North America, 2008: Distribution map. Retrieved Janu- ary 21, 2015 from http://www.efl oras.org/fl orataxon.aspx?fl ora_ id=11&taxon_id=242101813. Franceschi, V. R., Nakata, P., A., 2005: Calcium oxalate in plants: formation and function. Annual Review of Plant Biology 56, 41–71. Garbari, F., Giordani, A., Marcucci, R., Tornadore, N., 2003: The genus Ornithogalum L. (Hyacinthaceae) in Italy, XIV: to- wards a redefi nition of infrageneric taxa, with new proposals. Bocconea 16, 269–281. Garbari, F., Peruzzi, L., Tornadore, N., 2008: Ornithogalum L. (Hyacinthaceae Batsch) e generi correlati (subfam. Ornithoga- loideae Speta) in Italia. Atti della Società Toscana di Scienze Naturali, Memorie, serie B 114, 35–44. Herrmann, N., 2002: Biological Flora of Central Europe: „Orni- thogalum angustifolium” nom. prov., Syn. p.p. O. orthophyl- lum ssp. kochii = O. kochii Parl., O. gussonei Ten. Flora 197, 409–428. Littlejohn, G. M., Blomerus, L. M., 1997: Evaluation of Ornithog- alum genebank accessions for some characteristics of impor- tance for breeding cut fl owers or pot plants. Genetic Resourc- es and Crop Evolution 44, 227–234. Lynch, A. H., Rudall, P. J., Cutler, D. F., 2006: Leaf anatomy and systematics of Hyacinthaceae. Kew Bulletin – Royal Botanic Gardens 61, 145–159. Manning, J. C., Forest, F., Devey, D. S., Fay, M. F., Goldblatt, P., 2009: A molecular phylogeny and a revised classifi cation of Ornithogaloideae (Hyacinthaceae) based on an analysis of four plastid DNA regions. Taxon 58, 77–107. Martínez-Azorín, M., Crespo, M. B., Juan, A., 2009: Nomencla- ture and taxonomy of Ornithogalum divergens Boreau (Hya- cinthaceae) and related taxa of the polyploidy complex of Or- nithogalum umbellatum L. Candollea 64, 163–169. Martínez-Azorín, M., Crespo, M. B., Juan, A., 2010: Taxonomic revision of Ornithogalum subg. Ornithogalum (Hyacinthace- ae) in the Iberian Peninsula and the Balearic Islands. Plant Systematics and Evolution 289, 181–211. Martínez-Azorín, M., Crespo, M. B., Juan, A., Fay, M. F., 2011: Mo- lecular phylogenetics of subfamily Ornithogaloideae (Hyacin- thaceae) based on nuclear and plastid DNA regions, including a new taxonomic arrangement. Annals of Botany 107, 1–37. Meriç, C., Aksoy, Ö., Dane, F., 2011: Morphological and anatomi- cal contributions to the taxonomical identifi cation of two Or- nithogalum taxa (O. nutans and O. boucheanum) from Flora of Turkey. Biologia 66, 68–75. Moret, J., Favereau, Y., Gorenfl ot, R., 1991: A biometric study of the O. umbellatum complex in France. Plant Systematics and Evolution 175, 73–86. Moret, J., 1992: Numerical taxonomy applied to a study of some ploi- dy levels within the Ornithogalum umbellatum complex (Hya- cinthaceae) in France. Nordic Journal of Botany 12, 183–195. Obermeyer, A. A., 1978: Ornithogalum: a revision of the southern African species. Bothalia 12, 323–376. Öztürk, D., Koyuncu, O., Yaylaci, Ö. K., Özgişi, K., Sezer, O., Tokur, S., 2014: Comparative anatomical studies on twelve Ornithogalum (Asparagaceae) species (eleven nonendemic, one endemic) belonging to subgen. Ornithogalum and subgen. Beryllis, growing naturally in Eskişehir (Central Anatolia-Tur- key). Journal of Scientifi c Research and Reviews 3, 40 – 49. Peruzzi, L., Caparelli, K. F., Cesca, G., 2007: Contribution to the systematic knowledge of the genus Ornithogalum L. (Hyacin- thaceae): morpho-anatomical variability of the leaves among different taxa. Bocconea 21, 257–265. Popova, M., Anastasov, H., 1997: A morphological and anatomi- cal study of two Bulgarian Ornithogalum species (Liliaceae). Bocconea 5, 737–741. Prychid, C. J., Rudall, P. J., 1999: Calcium oxalate crystals in mono cotyledons: A review of their structure and systematics. Annals of Botany 84, 725–739. Soó, R., 1973: Systematical and geobotanical handbook of the Hungarian fl ora and vegetation V. Akadémiai Kiadó, Budapest (in Hungarian). Tilton, V. R., Lersten, N. R., 1981: Ovule development in Orni- thogalum caudatum (Liliaceae) with a review of selected pa- pers on angiosperm reproduction III. Nucellus and megag- ametophyte. New Phytologist 88, 477–504. Van Raamsdonk, L. W. D., 1986: Biosystematic studies on the umbellatum-angustifolium complex of the genus Ornithoga- lum (Liliaceae). II. Genome characterization and evolution. Nordic Journal of Botany 6, 525–544. Vujić, A., Radenković, S., Ståhls, G., Ačanski, J., Stefanović, A., Veselić, S., Andrić, A., Hayat, R., 2012: Systematics and tax- onomy of the rufi cornis group of genus Merodon Meigen (Diptera: Syrphidae). Systematic Entomology 37, 578–602. Zahariadi, C., 1980: Ornithogalum L. In: Tutin, T. G., Heywood, V. H., Burges, N. A., Moore, D. M., Valentine, D. H., Walters, S. M., Webb, D. A. (eds.), Flora Europaea 5, 35–40. Cam- bridge University Press, Cambridge. ORNITHOGALUM ANATOMY ACTA BOT. CROAT. 75 (1), 2016 O n- lin e Su pp l. Ta b. 1 . A na to m ic al c ha ra ct er is tic s of th e le af p re se nt ed a s m ea n va lu e, a nd c oe ffi c ie nt o f v ar ia tio n % (i n br ac ke ts ). C ha ra ct er s: 1 – to ta l c ro ss -s ec tio n ar ea (m m 2 ) , 2 – % le af e pi de rm is , 3 – % p al is ad e tis su e, 4 – % s po ng y tis su e, 5 – % v as cu la r t is su e, 6 – c ro ss -s ec tio n ar ea o f a d ax ia l e pi de rm is c el ls (μ m 2 ) , 7 – c ro ss -s ec tio n ar ea o f a ba xi al e pi de rm is c el ls (μ m 2 ) , 8 – c ro ss -s ec tio n ar ea o f ad ax ia l p al is ad e tis su e ce lls (μ m 2 ) , 9 – c ro ss -s ec tio n ar ea o f a ba xi al p al is ad e tis su e ce lls (μ m 2 ) , 1 0 – le af th ic kn es s (m m ), 11 – a da xi al p al is ad e ce lls in de x, 1 2 – ab ax ia l p al is ad e ce lls in de x, 1 3 – nu m be r of ri bs o n ab ax ia l s id e, 1 4 – nu m be r o f a ba xi al v as cu la r b un dl es , 1 5 – nu m be r o f c en tr al v as cu la r b un dl es , 1 6 – nu m be r o f s to m at a on a da xi al e pi de rm is o n cr os s- se ct io n, 1 7 – nu m be r o f s to m at a on a b- ax ia l e pi de rm is o n cr os s- se ct io n. D iff er en t s up er sc ri pt s in di ca te th at d iff er en ce s be tw ee n lo ca lit ie s ar e si gn ifi ca nt a cc or di ng to D un ca n’ s te st (p ≤ 0 .0 5) . O . u m be lla tu m O . d iv er ge ns V rš ač ke Pl an in e Z lo ts ka K lis ur a D el ib la ts ka Pe šč ar a V al je vo M ec se k Sz eg ed Su se k N ov i S ad Ti te ls ki B re g H ód m ez ö vá sá rh el y 1 2. 2c (2 9) 1. 6c d (2 9) 1. 9c (2 4) 1. 9c (3 3) 1. 2d (1 5) 1. 3d (1 5) 3. 4a b (1 9) 3. 7a (3 5) 3. 0b (1 6) 2. 8b (2 1) 2 12 .1 bc (1 7) 15 .0 a ( 11 ) 11 .4 bc d ( 16 ) 12 .5 b ( 17 ) 10 .8 cd e ( 10 ) 9. 5e fg (1 6) 8. 8f gh (1 2) 8. 2g h ( 15 ) 10 .3 de f ( 20 ) 7. 7h (1 4) 3 17 .3 c ( 17 ) 19 .6 bc (2 0) 20 .7 b ( 17 ) 20 .7 b ( 14 ) 25 .1 a ( 19 ) 20 .7 b ( 13 ) 18 .2 bc (1 8) 13 .3 d ( 17 ) 12 .3 d ( 33 ) 8. 3e (3 1) 4 66 .6 de fg (6 ) 59 .9 h ( 18 ) 64 .6 ef g ( 6) 64 .8 fg (6 ) 62 .3 gh (9 ) 67 .7 de f ( 4) 70 .7 cd (6 ) 76 .4 b ( 3) 74 .8 bc (5 ) 81 .2 a ( 4) 5 4. 0a (3 4) 3. 5a b ( 31 ) 3. 3b c (2 3) 2. 0f (3 5) 1. 8f (2 5) 1. 9f (2 3) 2. 3d ef (1 5) 2. 1e f ( 18 ) 2. 7c de (1 5) 2. 8b cd (1 0) 6 81 4a bc (3 7) 88 4a b (2 8) 82 4a bc (3 1) 74 6b c (2 4) 49 6d e (2 3) 45 9e (3 3) 10 08 a ( 30 ) 66 8b cd (2 1) 66 2c d ( 26 ) 79 8a bc (2 1) 7 60 6a b (2 2) 48 6b c (2 4) 62 8a b (3 0) 53 6a b (2 6) 39 4c (4 2) 37 8c (1 9) 66 5a (1 7) 65 8a (2 6) 49 8b c (2 0) 58 1a b (3 3) 8 18 83 a ( 18 ) 14 96 b ( 24 ) 17 15 ab (7 ) 15 92 ab (1 6) 97 2c (2 4) 13 37 b ( 9) 17 57 ab (2 7) 14 95 b ( 32 ) 61 3d (5 7) 85 2c d (2 4) 9 18 80 a ( 17 ) 14 30 b ( 17 ) 16 26 ab (1 8) 17 16 ab (1 7) 13 58 bc (1 7) 14 57 b ( 13 ) 18 56 a ( 16 ) 16 41 ab (3 5) 65 8c (5 1) 81 8c (2 4) 10 0. 7b (1 8) 0. 6c (2 0) 0. 7b (1 4) 0. 7b (1 4) 0. 6c (8 ) 0. 6c (9 ) 0. 8a (9 ) 0. 8a (1 1) 0. 6c (1 0) 0. 7b (1 0) 11 2. 2b (2 1) 2. 4a b (3 2) 2. 8a (1 6) 2. 6a b (1 9) 2. 2b (2 5) 2. 6a b (3 5) 2. 1b (1 3) 1. 7b c (3 9) 1. 2c (3 0) 1. 0c (2 8) 12 2. 5c (1 5) 2. 5c (2 3) 3. 0b c (1 1) 3. 8a (1 6) 3. 2b (2 1) 3. 6a b (2 6) 2. 8b c (1 0) 2. 1c (3 1) 1. 5d (3 7) 1. 3d (4 6) 13 7b (1 8) 8b (1 9) 7b (2 2) 8b (9 ) 6b (1 0) 6b (1 3) 10 a ( 12 ) 10 a ( 28 ) 10 a ( 24 ) 8b (1 7) 14 9b c (1 0) 10 b ( 9) 10 b ( 9) 9b c (1 2) 8c (8 ) 7d (1 5) 10 b ( 15 ) 11 a ( 15 ) 11 a ( 19 ) 10 b ( 16 ) 15 10 bc (1 0) 10 bc (1 0) 9c (8 ) 8c d (1 1) 8c d (1 3) 7d (1 3) 10 bc (1 7) 11 b ( 16 ) 12 a ( 14 ) 10 bc (2 0) 16 8a b (6 0) 8a (3 6) 8a bc (3 3) 8a (3 1) 5c (2 5) 5b c (4 1) 7a bc (4 1) 8a b (3 3) 6a bc (4 4) 6a bc (3 4) 17 13 ab (3 2) 12 ab (2 8) 13 ab (3 5) 15 a ( 26 ) 8c (2 4) 10 bc (3 2) 14 ab (3 9) 14 ab (3 2) 12 ab (3 9) 10 bc (4 0) 1 ANDRIĆ A. M., RAT M. M., ZORIĆ L. N., LUKOVIĆ J. Ž. 2 ACTA BOT. CROAT. 75 (1), 2016 O n- lin e Su pp l. Ta b. 2 . A na to m ic al c ha ra ct er is tic s of th e sc ap us p re se nt ed a s m ea n va lu e, a nd c oe ffi c ie nt o f v ar ia tio n % (i n br ac ke ts ). C ha ra ct er s: 1 – to ta l c ro ss -s ec tio n ar ea (m m 2 ) , 2 – % s ca pu s ep id er - m is , 3 – % s ca pu s co rt ex , 4 – % s ca pu s pi th , 5 – % s ca pu s sc le re nc hy m a, 6 – % s ca pu s sm al l v as cu la r b un de ls , 7 – % s ca pu s la rg e va sc ul ar b un dl es , 8 – % s ca pu s pi th p ar en ch ym a, 9 – c ro ss -s ec tio n ar ea of s ca pu s ep id er m al c el ls (μ m 2 ) , 1 0 – cr os s- se ct io n ar ea o f s ca pu s co rt ex p ar en ch ym a ce lls (μ m 2 ) , 1 1 – cr os s- se ct io n ar ea o f s ca pu s pi th p ar en ch ym a ce lls (μ m 2 ) , 1 2 – nu m be r o f s ca pu s co rt ex c el l l ay er s, 13 – n um be r of s m al l v as cu la r bu nd le s (i n sc le re nc hy m a) , 1 4 – nu m be r of la rg e va sc ul ar b un dl es ( in p ar en ch ym a) . D iff er en t s up er sc ri pt s in di ca te th at d iff er en ce s be tw ee n lo ca lit ie s ar e si gn ifi ca nt a c- co rd in g to D un ca n’ s te st (p ≤ 0 .0 5) . O . u m be lla tu m O . d iv er ge ns V rš ač ke Pl an in e Z lo ts ka K lis ur a D el ib la ts ka Pe šč ar a V al je vo M ec se k Sz eg ed Su se k N ov i S ad Ti te ls ki B re g H ód m ez ö vá sá rh el y 1 7. 2c de (3 0) 3. 0g (4 8) 6. 6d e ( 29 ) 4. 1f g ( 25 ) 5. 7e f ( 30 ) 7. 7b cd e ( 42 ) 8. 7a bc d ( 43 ) 10 .1 ab (1 2) 10 .4 a ( 22 ) 9. 2a bc (4 8) 2 3. 7c d (2 0) 6. 3a (2 7) 4. 4b c (1 7) 6. 3a (1 4) 4. 6b (2 6) 3. 7c d (2 3) 4. 4b c (1 6) 3. 0d e (1 2) 2. 2e (2 8) 2. 8d e (2 2) 3 18 .9 bc (1 4) 27 .7 a ( 18 ) 19 .9 bc (1 5) 27 .4 a ( 16 ) 12 .8 e ( 23 ) 15 .4 de (3 1) 21 .7 b ( 11 ) 25 .0 a ( 11 ) 18 .3 cd (1 3) 17 .2 cd (1 5) 4 77 .5 b ( 4) 66 .0 e ( 9) 75 .7 bc (4 ) 66 .3 e ( 7) 82 .6 a ( 4) 80 .9 ab (7 ) 73 .9 cd (3 ) 72 .0 d ( 4) 79 .5 ab (4 ) 80 .0 ab (4 ) 5 6. 2a b (1 5) 6. 9a (2 3) 6. 5a b (1 9) 5. 6b (3 4) 7. 3a (2 4) 5. 2b c (2 7) 5. 5b (2 0) 3. 8d (1 6) 4. 1c d (1 4) 4. 2c d (1 8) 6 0. 8a (1 5) 0. 6a b (5 3) 0. 6a b (3 3) 0. 5b c (2 7) 0. 7a b (3 0) 0. 6a b (2 1) 0. 4c d (2 4) 0. 4c d (1 8) 0. 4c (2 8) 0. 3d (2 3) 7 3. 4a (1 8) 2. 8b (2 4) 2. 8b c (2 2) 2. 1d (1 9) 2. 1d (1 9) 2. 5b cd (1 3) 2. 3c d (3 6) 2. 0d (2 1) 2. 3b cd (1 9) 2. 1d (1 5) 8 67 .1 b ( 4) 55 .6 c ( 13 ) 65 .8 b ( 5) 58 .1 c ( 10 ) 72 .5 a ( 6) 72 .6 a ( 9) 65 .9 b ( 5) 65 .9 b ( 4) 72 .7 a ( 5) 73 .5 a ( 4) 9 57 0c (2 2) 71 1b c (2 9) 77 8b (2 2) 96 4a (2 5) 73 4b c (1 3) 59 4c (1 8) 97 4a (3 2) 77 3b (2 0) 61 3b c (1 8) 59 3c (1 9) 10 12 22 bc (2 1) 12 07 bc (2 9) 14 05 b ( 24 ) 13 49 bc (1 4) 94 8d (2 0) 92 6d (1 3) 17 51 a ( 14 ) 17 44 a ( 14 ) 13 27 bc (1 8) 11 50 cd (1 8) 11 40 35 ef (2 1) 36 09 f ( 26 ) 47 88 de (2 0) 56 38 cd (2 1) 54 58 cd (2 3) 64 81 bc (2 8) 53 34 cd (1 8) 83 86 a ( 20 ) 54 89 cd (1 3) 74 03 ab (1 2) 12 5a (1 8) 4b cd (1 4) 4a bc (1 9) 4c d (0 ) 4d (1 3) 4d (1 7) 4a bc (1 2) 5a (9 ) 4a bc (1 2) 5a b (1 5) 13 27 a ( 32 ) 12 e ( 33 ) 25 ab (2 4) 15 cd e (1 2) 19 cd (1 8) 24 ab c (2 8) 17 cd e (2 2) 21 bc (2 2) 27 a ( 12 ) 22 bc (2 6) 14 20 b ( 26 ) 10 d ( 26 ) 19 bc (2 7) 12 d ( 15 ) 14 cd (2 4) 18 bc (2 8) 17 bc (3 7) 18 bc (1 4) 24 a ( 10 ) 20 b ( 41 ) ORNITHOGALUM ANATOMY ACTA BOT. CROAT. 75 (1), 2016 On-line Suppl. Tab. 3. Principal component analysis: factor coor- dinates of the variables based on correlations. Only variables with factor values > 0.7, which signifi cantly contribute to total varia- tion, were presented. Character Factor 1 Factor 2 Factor 3 Percentage of palisade tissue 0.724 –0.213 0.363 Percentage of spongy tissue –0.779 0.113 –0.082 Total scapus cross-section area –0.775 –0.109 0.359 Percentage of scapus epidermis 0.829 0.260 –0.173 Percentage of scapus cortex 0.307 0.718 –0.365 % Total variance 25.86 18.91 8.97 On-line Suppl. Tab. 4. Multivariate discriminant analysis: stan- dardized coeffi cients for canonical variables. Only variables with root values > 0.7, which signifi cantly contribute to total variation, were presented. Character Root 1 Root 2 Root 3 Total leaf cross-section area 1.346 0.456 –0.416 Percentage of palisade tissue 0.071 0.132 –1.336 Percentage of spongy tissue 0.264 0.088 –1.105 Percentage of scapus epidermis –0.400 –0.190 –0.815 Percentage of scapus pith parenchyma –0.477 0.999 –0.058 Percentages of the vectors 41.89 18.97 16.48 3