OPCE-STR.vp Acta Bot. Croat. 69 (2), 259–274, 2010 CODEN: ABCRA 25 ISSN 0365–0588 Palynological classification of Onosma L. (Boraginaceae) species from east Mediterranean region in Turkey RIZA BI·NZET1*, I · RFAN KANDEMI·R2, NERMI·N ORCAN3 1 Adiyaman University, Faculty of Science and Art, Department of Biology, 02040 Adiyaman, Turkey 2 Ankara University, Faculty of Science, Department of Biology, Ankara, Turkey 3 Mersin University, Faculty of Science and Art, Department of Biology, Mersin, Turkey Twenty-five Onosma (Boraginaceae) taxa belonging to two subsections Haplotricha and Asterotricha from the east Mediterranean region in Turkey were studied by palynological analysis and numerical taxonomy. Application of discriminant function analysis to raw data obtained from the acetolysis and Wodehouse methods resulted in very good alloca- tion of species to their original groups. However the results obtained from acetolysis (99%) resulted in better discrimination than the Wodehouse (97%) method. A similar out- come was reached in principal component analysis and UPGMA. Higher percentage of phenetic variation was explained by the acetolysis method. The utility of palynological data in taxonomic classification with the use of using numerical methods is discussed. Keywords: Onosma, pollen, numerical taxonomy, classification, Turkey Abbreviations: DFA – discriminant function analysis, UPGMA – unweighted pair group method with arithmetic mean Introduction Recent studies and revisions have increased the number of species in the genus Onosma L. to over 230 species (BOISSIER 1897, DINSMOR 1932, HAYEK and MARKGRAF 1970, TUTIN et al. 1972, SHISHKIN 1974, RIEDL 1978, MEIKLE 1985, TEPPNER 1991, GE-LING et al. 1995). This genus has biennial and perennial members, and is generally suffruticose. Onosma species are recognized on the basis of indumentum characteristics along with flowers in terminal cymes, calyx accrescence, stamens inserted at the middle of the corolla. The genus Onosma (Boraginaceae) is represented by about 102 taxa (97 species) in Turkey and the endemism among native species is higher than 50 % (RIEDL 1978, YILDIRIMLI 2000, RIEDL et al. 2005, BI·NZET and ORCAN 2007). In the Flora of Turkey, the general classifica- tion of Onosma species was based only on indumentum characteristics, and palynological ACTA BOT. CROAT. 69 (2), 2010 259 * Corresponding author, e-mail : rbinzet@gmail.com U:\ACTA BOTANICA\Acta-Botan 2-10\282 Binzet.vp 11. listopad 2010 13:51:45 Color profile: Disabled Composite 150 lpi at 45 degrees data were not utilized. In addition, MAGGI et al. (2008) mentioned the lack of palynological data on the genus Onosma. Only a limited number of studies can be found in the literature; the first palynological study in the Boraginaceae was that of ERDTMAN (1952). The genera of Lithospermae com- prising 45 species of Onosma were palynologically studied with light microscopy (JOHN- STON 1954). The pollen morphology of the family Boraginaceae have been studied by ERDTMAN (1952), MARTICORENA (1968), HUYNH (1971, 1972), NOWICKE and RIDGWAY (1973), NOWICKE and SKVARLA (1974), DIEZ (1984), QURESHI and QAISER (1987) and PERVEEN et al. (1995). In recent years several Onosma species from Turkey have been ex- amined for pollen morphology by BI·NZET and ORCAN (2003a, b). Although palynological features can provide a wealth of taxonomic characters that have been important in inferring phylogenetic relationships and future classifications, there was not enough palynological data for the genus Onosma. In the present study we investigated the pollen morphology of the 25 taxa belonging to Onosma which is a difficult genus from systematic and taxonomic point of view. In order to resolve these difficulties and make use of palynological data in the genus Onosma we ap- plied the numerical taxonomic approach. This method has had a wide application to differ- ent plant taxa in Turkey (TOGAN et al. 1983, KENCE 1988, DOÐAN et al. 1992, DOÐAN and TOSUNOÐLU 1992, DOÐAN 1997, TÜTEL et al. 2005). Thus, we have attempted to classify 25 Onosma taxa on the basis of palynological characters obtained by two different methodolo- gies (acetolysis and Wodehouse) and tried to find a future methodology to be used in palynological data collection in the future. Materials and methods Pollen grains obtained from herbarium and fresh materials were prepared using the methods described by Wodehouse (WODEHOUSE 1935) and acetolysis by Erdtman (ERDT- MAN 1952). The Wodehouse and acetolysis methods were followed by light microscopic (LM) study. Polar and equatorial axis, pollen shape, length of pores (pori) and colpus (colpi), width of pores (pori) and colpus (colpi), exine thickness, intine thickness and length of polar triangular edge were measured with an Olympus BX 40 microscope (x10; x 100). The terminology used is in accordance with ERDTMAN (1952) and FAEGRI and IVERSEN (1964). At least 30 pollen grains were measured from each of 25 Onosma taxa. The raw data obtained from both Acetolysis and Wodehouse were analysed with discriminant function analysis (DFA) using the statistical package SPSS (2004) separately in order to find out the palynological relationships between 25 taxa (Tab. 1). The second sets of Onosma data matrix were formed (the averages of the species used in DFA) from 9 palynological characters for acetolysis and 10 for Wodehouse methods (Tabs. 2, 3). The matrices obtained from the two different methods were standardised and the new character distribution has mean 0 and standard deviation 1. Principal component analysis was ap- plied to the standardized data sets using correlation matrices and finally UPGMA dendro- grams were constructed based on the average taxonomic distance using NTSYS-pc (ROHLF 2004). Statistical analyses were performed using STATISTICAL PACKAGE (2004). 260 ACTA BOT. CROAT. 69 (2), 2010 BI·NZET R., KANDEMI·R I · ., ORCAN N. U:\ACTA BOTANICA\Acta-Botan 2-10\282 Binzet.vp 11. listopad 2010 13:51:45 Color profile: Disabled Composite 150 lpi at 45 degrees Results The Acetolysis and Wodehose average values are listed in tables 2 and 3. Principal component analysis and UPGMA were applied to acetolysis and Wodehouse data in order to visualize the palynological relationships of Onosma genus in the eastern Mediterranean region. Discriminant function analysis of acetolysis and Wodehouse data All characters used both in acetolysis and Wodehouse were significantly different among Onosma taxa (p<0.01) based on univariate ANOVA. A scatterplot of 25 Onosma ACTA BOT. CROAT. 69 (2), 2010 261 PALYNOLOGICAL CLASSIFICATION OF ONOSMA Tab. 1. Onosma species and collection localities along with grid square information Species Locality (grid square) Onosma albo-rosea Fisch. et Mey. Kahramanmaraþ, Kayseri: B6; Mersin: C4 O. angustissima Hausskn. et Bornm.* Antalya: C3; Mersin: C4; Adana: C5 O. armena DC.* Mersin, Karaman: C4 O. aucherana DC. Mersin: C4 O. auriculata Aucher ex DC. Mersin: C4 O. bornmuelleri Hausskn.* Mersin: C5 O. bracteosa Hausskn. et Bornm.* Karaman, Mersin: C4; Mersin: C5 O. caerulescens Boiss. Kahramanmaraþ: C6 O. cassia Boiss. Hatay: C5 O. frutescens Lam. Antalya: C3; Mersin: C4 O. gigantea Lam. Adana, Kahramanmaraº: B6; Mersin: C4; Mersin: C5; Osmaniye, Adana, Osmaniye, Kahramanmaraº: C6 O. inexspectata Teppner * Osmaniye: C6 O. isaurica Boiss. et Heldr.,* Mersin, Karaman: C4; Kilis: C6 O. lycaonica Hub.-Mor.* Mersin: C4 O. mersinana H. Riedl, Binzet et Orcan * Mersin: C5 O. mutabilis Boiss.* Kahramanmaraþ: B6; Mersin, Adana: C5 O. papillosa H. Riedl * Adana: B6 O. rascheyana Boiss. Kahramanmaraþ: C6 O. riedliana Binzet et Orcan * Mersin: C5 O. roussaei DC. Mersin: C4 O. rutila Hub.-Mor.* Mersin: C4 O. sericea Willd. Hatay, Kilis, Gaziantep, Kahramanmaraþ, Osmaniye: C6; Adana: B6 O. sieheana Hayek * Karaman, Mersin: C4 O. stenoloba Hausskn. Ex H. Riedl * Kahramanmaraþ: B6; Mersin: C5 O. taurica Pallas ex Willd. Mersin: C4; Hatay: C6 *endemic species U:\ACTA BOTANICA\Acta-Botan 2-10\282 Binzet.vp 11. listopad 2010 13:51:45 Color profile: Disabled Composite 150 lpi at 45 degrees 262 A C T A B O T .C R O A T .69 (2),2010 B I ·N Z E T R .,K A N D E M I ·R I ·.,O R C A N N . Tab. 2. Palynological characters of the Onosma species for Wodehouse (measurements in mm). Species A P E plg plt clg clt ex i ect/end t O. albo-rosea M 17.56 15.47 3.54 3.98 13.21 2.75 0.72 0.78 0.25 6.67 SD 0.46 0.52 0.47 0.58 0.49 0.24 0.12 0.07 0.3 Min–Max 16–18.5 14–16 2.5–4.5 3–5 12.5–13.5 2.5–3 0.5–0.9 0.6–1 6–7 O. angustissima M 14.99 12.72 3.01 3.63 12.38 3.36 0.37 0.68 0.66 6.4 SD 0.57 0.71 0.25 0.37 0.43 0.33 0.03 0.08 0.25 Min–Max 13.5–16 11.5–13.5 2.4–4.2 3–4.2 11.5–14 3–4 0.3–0.5 0.5–0.8 6–7 O. armena M 15.83 13.28 3.15 3.6 12.31 2.45 0.55 0.8 1.5 6.53 SD 0.51 0.71 0.46 0.43 0.5 0.28 0.07 0.07 0.29 Min–Max 14.5–16.5 12–14 2–4 2.5–4.5 11–13 1.5–3 0.4–0.7 0.6–1 6–7 O. aucherana M 14.63 11.32 2.65 3.08 10.62 2.87 0.44 0.62 0.5 5.7 SD 0.58 0.43 0.43 0.4 0.56 0.27 0.04 0.06 0.27 Min–Max 13.5–16 10–12 1.5–3.5 2–4 9–12 2.4–3 0.3–0.5 0.5–0.8 5–6 O. auriculata M 17.92 13.86 3.72 4.1 14.37 3.49 0.6 0.8 0.33 7.46 SD 0.47 0.46 0.46 0.5 0.64 0.31 0.08 0.06 0.32 Min–Max 16.5–18.5 12.5–14.5 2.5–4.5 3–5 13–15 3–4 0.4–0.8 0.6–1 6.5–8 O. bornmuelleri M 16.46 12.9 3.91 4.35 12.24 3.68 0.46 0.77 0.33 6.32 SD 0.49 0.56 0.54 0.51 0.53 0.3 0.06 0.08 0.38 Min–Max 15–17 11.5–13.5 3–4.5 3.5–5 11–14 3–4 0.3–0.7 0.5–1 4–7 O. bracteosa M 15.69 13.26 3.53 3.6 12.61 2.85 0.75 1.03 3 6.57 SD 0.58 0.62 0.53 0.64 0.71 0.23 0.04 0.12 0.32 Min–Max 14.2–16.5 12–14 3–3.9 2.5–4.5 11–13 2.5–3.2 0.5–0.8 0.8–1.2 6–7 O. caerulescens M 17.03 14.01 3.28 3.73 13.09 3.02 0.41 0.63 0.66 5.86 SD 0.62 0.24 0.22 0.3 0.61 0.28 0.07 0.07 0.29 Min–Max 15.5–18 13–14.5 2.8–3.6 3.2–4.2 12.5–13.5 2.5–3.5 0.3–0.5 0.5–0.8 5–6.5 O. cassia M 14.84 10.37 2.75 3.66 11.31 3.19 0.39 0.71 0.6 5.1 SD 0.5 0.47 0.57 0.61 0.52 0.32 0.05 0.09 0.29 Min–Max 13.5–15.5 9.5–11.5 1.5–4 2.5–5 10–12 2.5–3.5 0.3–0.5 0.6–0.8 4–6 U : \ A C T A B O T A N I C A \ A c t a - B o t a n 2 - 1 0 \ 2 8 2 B i n z e t . v p 1 1 . l i s t o p a d 2 0 1 0 1 3 : 5 1 : 4 6 C o l o r p r o f i l e : D i s a b l e d C o m p o s i t e 1 5 0 l p i a t 4 5 d e g r e e s A C T A B O T .C R O A T .69 (2),2010 263 P A L Y N O L O G IC A L C L A S S IF IC A T IO N O F O N O S M A Tab. 2. – continued Species A P E plg plt clg clt ex i ect/end t O. frutescens M 14.82 11.44 3.25 3.13 12.01 2.72 0.6 0.73 0.66 6.49 SD 0.51 0.55 0.59 0.52 0.57 0.25 0.05 0.08 0.35 Min–Max 13–15.5 10–12.5 2–4.5 2–4 10–13 2–3 0.4–0.8 0.6–0.9 5.5–7 O. gigantea M 17.56 13.64 3.76 3.76 14.21 2.83 1 0.41 1.5 7.04 SD 0.64 0.48 0.55 0.55 0.48 0.25 0.14 0.08 0.33 Min–Max 15–20 11–16 2.5–4.5 2.5–4.5 13–14.5 2.5–3.5 0.8–1.1 0.3–0.5 6–8 O. inexspectata M 15.81 12.75 2.78 3.07 11.77 2.83 0.36 0.75 0.5 5.9 SD 0.51 0.4 0.23 0.21 0.43 0.27 0.07 0.09 0.26 Min–Max 14.5–16.5 11.5–13.5 2.2–3.4 2.6–3.4 11.5–13 2.5–3 02–0.5 0.6–0.9 5–6.5 O. isaurica M 17.88 15.13 3.35 3.79 14.05 3.11 0.44 0.67 0.5 6.66 SD 0.65 0.5 0.32 0.28 0.51 0.3 0.06 0.07 0.37 Min–Max 16–19 13.5–16 2.8–4 3.2–4.4 12–15 2–3.5 0.3–0.6 0.5–0.8 5–8 O. lycaonica M 16.78 14.14 3.84 4.34 13.6 3.46 0.46 0.82 1.5 7.03 SD 0.81 0.65 0.61 0.5 0.62 0.31 0.06 0.08 0.4 Min–Max 15–18 13–15 2.5–5 3–5.5 12.5–14 3–4 03–0.6 0.7–1 4.5–8 O. mersinana M 16.47 11.7 2.82 3.49 11.87 2.81 0.36 0.72 0.5 7.59 SD 0.57 0.44 0.44 0.58 0.38 0.22 0.03 0.08 0.37 Min–Max 15–17 10.5–12.5 2–3.5 2.4–4.5 11–12.5 2–3.5 0.3–0.4 0.8–0.9 7–8 O. mutabilis M 16.03 14.28 4.18 4.15 13.06 3.25 0.52 0.89 2.5 8.05 SD 0.56 0.64 0.48 0.57 0.87 0.23 0.07 0.07 0.43 Min–Max 14.5–17 12.5–14.5 3.5–4.7 3–5 12–14 2.5–4 0.3–0.6 0.7–0.9 7–9 O. papillosa M 16.05 11.46 3.04 3.5 12.1 2.67 0.4 0.84 0.66 6.29 SD 0.63 0.52 0.47 0.67 0.54 0.26 0.05 0.08 0.33 Min–Max 14.5–17 10–12 2–4 2.5–4.5 10–13 2–3 0.3–0.5 0.7–0.9 5–8 O. rascheyana M 15.73 12.68 3.07 3.18 12.16 2.74 0.62 0.77 0.5 4.86 SD 0.43 0.48 0.19 0.25 0.65 0.23 0.04 0.05 0.22 Min–Max 14.5–16.5 11.5–13.5 2.6–3.4 2.6–3.6 11–13 2–3 0.4–0.8 0.6–1 4–6 U : \ A C T A B O T A N I C A \ A c t a - B o t a n 2 - 1 0 \ 2 8 2 B i n z e t . v p 1 1 . l i s t o p a d 2 0 1 0 1 3 : 5 1 : 4 6 C o l o r p r o f i l e : D i s a b l e d C o m p o s i t e 1 5 0 l p i a t 4 5 d e g r e e s 264 A C T A B O T .C R O A T .69 (2),2010 B I ·N Z E T R .,K A N D E M I ·R I ·.,O R C A N N . Tab. 2. – continued Species A P E plg plt clg clt ex i ect/end t O. riedliana M 16.32 12.72 2.79 3.59 12.61 3.26 0.41 0.65 0.66 7.75 SD 0.55 0.39 0.44 0.43 0.73 0.26 0.05 0.07 0.33 Min–Max 15–17 12–13.5 2–3.5 2.5–4.5 11.5–13.5 2.5–3.5 0.3–0.5 0.5–0.8 7–8 O. roussaei M 14.84 10.5 2.79 2.82 10.71 2.45 1 0.51 2 5.96 SD 0.51 0.44 0.27 0.27 0.33 0.2 0.16 0.07 0.24 Min–Max 13–16 9–11.5 2–3 2–3 10–11 2–3 0.8–1.2 0.4–0.6 5.5–6 O. rutila M 14.88 10.68 2.66 2.78 12.98 2.3 0.43 0.67 0.33 5.68 SD 0.58 0.45 0.18 0.27 0.43 0.28 0.07 0.08 0.27 Min–Max 13.5–15.5 9.5–11 2.2–3 2.2–3.2 12–13.5 1.5–3 0.2–0.6 0.4–0.9 4–7 O. sericea M 17.43 14.7 3.43 4.13 13.9 3.62 0.29 0.71 0.5 6.49 SD 0.58 0.58 0.29 0.27 0.56 0.28 0.06 0.07 0.32 Min–Max 16–18 13–16 2.8–4 3.4–4.6 13–15 3–4 0.2–0.4 0.5–0.9 3–8 O. sieheana M 15.63 13.85 3.25 3.72 12.71 3.37 0.47 0.8 0.66 6.2 SD 0.72 0.5 0.62 0.55 0.63 0.31 0.08 0.06 0.3 Min–Max 14–16.5 12.5–14.5 2–4.5 2.5–5 11–14 2.5 4 0.3–0.7 0.7–0.9 5–7 O. stenoloba M 15.78 12.36 3.13 3.59 10.94 3.11 0.41 0.71 0.66 6.78 SD 0.45 0.48 0.43 0.44 0.38 0.29 0.03 0.08 0.28 Min–Max 14.5–16.5 11–13 2–4 2.5–4.5 9–12 2.5–3.5 0.3–0.6 0.6–1 6–7 O. taurica M 16.32 15.28 3.64 4.12 13.16 4.03 0.64 0.93 0.66 6.79 SD 0.55 0.54 0.47 0.47 0.5 0.3 0.1 0.1 0.3 Min–Max 14.5–17 14–16 2.5–4.5 3–5 12–13.5 3.5–4.5 0.5–0.8 0.7–1.1 6–8 M – mean, SD – standard deviation, Min–Max – minimum and maximum values), P – length of polar axis, E –width of equatorial axis, plg – length of pores (pori), plt – width of pores (pori), clg – length of colpus (colpi), clt – width of colpus (colpi), ex – exine thickness, i – intine thickness (only for Wodehouse method), ect./end – ectexine to endexine ratio, t – length of polar triangular edge U : \ A C T A B O T A N I C A \ A c t a - B o t a n 2 - 1 0 \ 2 8 2 B i n z e t . v p 1 1 . l i s t o p a d 2 0 1 0 1 3 : 5 1 : 4 6 C o l o r p r o f i l e : D i s a b l e d C o m p o s i t e 1 5 0 l p i a t 4 5 d e g r e e s A C T A B O T .C R O A T .69 (2),2010 265 P A L Y N O L O G IC A L C L A S S IF IC A T IO N O F O N O S M A Tab. 3. Palynological characters of the Onosma species for acetolysis (measurements in mm). Species A P E plg plt clg clt ex ect/end t O. albo–rosea M 22.51 18.82 2.33 10.82 15.3 2.8 0.99 0.25 9.12 SD 0.92 1.08 0.27 0.81 0.8 0.24 0.1 0.55 Min–Max 20–24 16–21 1.8–2.8 9–12 14–16 1.5–2.5 0.8–1.1 8.5–10 O. angustissima M 15.2 13.21 1.29 7.07 11.06 2.82 0.89 0.33 5.4 SD 0.81 0.76 0.26 0.79 0.64 0.43 0.08 0.38 Min–Max 13.0–16.5 11.5–15.0 0.8–1.8 5.5 –9 10–12 2.4–3.2 0.7–1 5–6 O. armena M 16.51 14.18 1.76 6.39 13.26 1.94 0.98 1.5 6.4 SD 0.89 0.65 0.26 0.66 0.77 0.18 0.09 0.47 Min–Max 14.5–18 12.5–15 1–2.2 5–7 12–14 1.5–2.5 0.8–1.2 6–7 O. aucherana M 19.59 14.54 2.13 7.51 13.42 2.27 0.98 0.5 6.56 SD 0.99 0.95 0.21 0.8 0.75 0.25 0.12 0.32 Min–Max 17–21 13–16 1.6–2.6 6–9 12–14 1.8–2.5 0.8–1.1 6–7 O. auriculata M 27.13 20.83 2.59 10.78 17.65 2.49 1.24 0.33 9.97 SD 1.76 1.71 0.28 1.49 1.2 0.21 0.15 0.76 Min–Max 22–29 16–23 2–3.2 8–14 13–21 1.5–3.5 1–1.6 8–11 O. bornmuelleri M 20.82 16.12 2.45 9.39 14.83 2.58 1.11 0.33 7.7 SD 1.01 0.71 0.33 0.5 0.8 0.26 0.09 0.68 Min–Max 18–23 14.5–17 1.8–3.2 7.5–11 13–15 2–3 1–1.2 6.5–9 O. bracteosa M 15.94 13.39 0.79 7.07 13.09 3.28 0.66 0.33 7.56 SD 0.67 0.45 0.21 0.66 0.58 0.25 0.07 0.41 Min–Max 14.5–17 12–14 0.4–1.2 5.5–8.5 9–15 2.5–4.5 0.5–0.9 6–8 O. caerulescens M 15.55 13.2 1.59 7.07 10.36 2.64 0.76 0.66 5.68 SD 0.91 0.88 0.18 0.95 0.4 0.32 0.08 0.48 Min–Max 13–17 12–15 1–2 5–9 9.5–11 2–3 0.6–1 5–6.5 O. cassia M 17.3 4 13.19 1.62 6.63 14.79 4.78 0.84 0.66 7.14 SD 0.95 0.54 0.18 0.63 0.76 0.43 0.08 0.48 Min–Max 15–19 11.5–14 1.2–2 5–8 13–16 4–5.5 0.6–1 6–8 U : \ A C T A B O T A N I C A \ A c t a - B o t a n 2 - 1 0 \ 2 8 2 B i n z e t . v p 1 1 . l i s t o p a d 2 0 1 0 1 3 : 5 1 : 4 6 C o l o r p r o f i l e : D i s a b l e d C o m p o s i t e 1 5 0 l p i a t 4 5 d e g r e e s 266 A C T A B O T .C R O A T .69 (2),2010 B I ·N Z E T R .,K A N D E M I ·R I ·.,O R C A N N . Tab. 3. – continued Species A P E plg plt clg clt ex ect/end t O. frutescens M 18.27 14.07 1.08 8.89 14.79 3.43 0.89 0.33 9.28 SD 1.18 0.57 0.3 0.69 0.7 0.34 0.09 0.76 Min–Max 15–21 12.5–15 0.6–1.8 7–10 13–16 3–4 0.7–1 8–10 O. gigantea M 19.33 15.91 1.98 9.11 17.51 2.41 0.77 0.33 8.9 SD 0.82 0.7 0.29 0.87 0.97 0.2 0.11 0.57 Min–Max 17.5–20 14–17 1.4–2.4 7–10.5 16–18 2–3 0.6–1 8–10 O. inexspectata M 14.46 11.89 1.3 6.91 10.2 2.49 0.7 0.5 5.22 SD 0.47 0.45 0.25 0.71 0.57 0.23 0.06 0.52 Min–Max 13.5–15 11–12.5 0.8–1.8 5.5–8 9.5–10.5 2–3 0.5–1 4.5–6 O. isaurica M 16.39 14.21 1.66 7.04 15.5 2.07 0.99 0.5 11.83 SD 0.74 0.72 0.34 1.13 0.92 0.33 0.1 0.67 Min–Max 14.5–17.5 12.5–15 1–2.2 5–9 14–16 1.8–2.4 0.8–1.2 10–13 O. lycaonica M 20.31 16.8 2.26 9.01 15.75 3.12 1.07 0.25 8.76 SD 1.09 0.65 0.32 0.77 0.8 0.31 0.13 0.54 Min–Max 17–22 15–18 1.6 –3 7.5–10.5 15–16 2.5–4 0.9–1.2 7–10 O. mersinana M 16.1 13.14 1.02 7.01 14.23 3.06 0.77 0.5 7.07 SD 0.78 0.72 0.2 0.79 0.87 0.3 0.08 0.5 Min–Max 14–17 11.5–14 0.6–1.4 5.5–8.5 13–14.5 2–4 0.6–1 6.5–7.5 O. mutabilis M 15.83 13.7 1.14 7.78 13.76 2.99 0.65 0.33 8.83 SD 0.54 0.45 0.26 0.99 0.73 0.2 0.06 0.73 Min–Max 14.5–16 12 –14 5 0.8–1.4 6.5–8.5 12.5–14.5 2–4 0.5–0.8 8–10 O. papillosa M 18.41 13.3 2.09 6.93 12.98 2.24 1.19 0.33 6.63 SD 0.83 0.63 0.48 0.74 0.62 0.32 0.9 0.39 Min–Max 15–20 11–14.5 1–3 4.5–9 12–14 2–2.5 0.8–1.2 5–8 O. rascheyana M 15.55 13.21 1.33 6.85 10.3 2.29 0.71 0.33 5.19 SD 0.62 0.47 0.26 0.66 0.6 0.25 0.08 0.38 Min–Max 14–16.5 12–14 0.8–2 5.5–8 9–11 1.5–3 0.6–0.9 4.5–5.5 U : \ A C T A B O T A N I C A \ A c t a - B o t a n 2 - 1 0 \ 2 8 2 B i n z e t . v p 1 1 . l i s t o p a d 2 0 1 0 1 3 : 5 1 : 4 7 C o l o r p r o f i l e : D i s a b l e d C o m p o s i t e 1 5 0 l p i a t 4 5 d e g r e e s A C T A B O T .C R O A T .69 (2),2010 267 P A L Y N O L O G IC A L C L A S S IF IC A T IO N O F O N O S M A Tab. 3. – continued Species A P E plg plt clg clt ex ect/end t O. riedliana M 19.38 14.05 2.08 6.92 14.99 3.61 0.95 0.66 7.84 SD 1.52 0.86 0.46 0.98 0.87 0.4 0.11 0.5 Min–Max 17–21 13–16.5 1–3 6–9 12–15.5 3–4 0.8–1 7–9 O. roussaei M 14.8 12.13 1.75 6.9 12.34 3.46 0.69 0.5 7.09 SD 0.87 0.71 0.68 0.25 0.75 0.33 0.08 0.57 Min–Max 13–16.5 10.5–14 1.2–2.2 5.5–8 11.5–13 2.5–4 0.6–0.8 6–8 O. rutila M 20.45 14.88 2.38 6.57 13.89 2.5 1 0.33 7 SD 1.35 0.65 0.5 0.97 0.8 0.24 0.13 0.45 Min–Max 18–23 13–16 1.5–3.5 5–8.5 12.5–15 2–3 0.8–1.2 6–8 O. sericea M 16.25 13.73 1.56 7.12 11.37 2.52 0.87 0.5 6.7 SD 0.86 0.62 0.3 0.3 0.58 0.22 0.09 0.52 Min–Max 15–18.5 12.5–15 1–2 5.5–8.5 10.5–14 2.2–3 0.7–1 5–7 O. sieheana M 19.25 16.41 2.05 9.33 14 2.73 0.98 0.66 7.69 SD 0.95 0.93 0.21 0.72 0.76 0.2 0.1 0.61 Min–Max 17–21 14–18 1.6–2.6 7.5–10.5 13–16 2–3.5 0.8–1.1 7–8 O. stenoloba M 16.34 13.57 1.41 7.11 14.32 3.64 0.9 0.66 8.03 SD 0.54 0.62 0.27 1.12 0.67 0.3 0.09 0.5 Min–Max 14.5–17 12–14.5 0.8–2 5–10 13–16 2.5–5 0.6–1.2 6.5–9 O. taurica M 22.85 19.09 2.64 9.43 19.53 1.61 1.02 0.66 9.55 SD 1.24 0.7 0.29 1.2 1.3 0.15 0.09 0.72 Min–Max 20–25 17–20 2–3.2 7–12 15–21 1–2 0.8–1.2 9–10 M – mean, SD – standard deviation, Min–Max – minimum and maximum, P – length of polar axis, E – width of equatorial axis, plg – length of pores (pori), plt – width of pores (pori), clg – length of colpus (colpi), clt – width of colpus (colpi), ex – exine thickness, i – intine thickness (only for Wodehouse method), ect./end – ectexine to endexine ratio, t – length of polar triangular edge U : \ A C T A B O T A N I C A \ A c t a - B o t a n 2 - 1 0 \ 2 8 2 B i n z e t . v p 1 1 . l i s t o p a d 2 0 1 0 1 3 : 5 1 : 4 7 C o l o r p r o f i l e : D i s a b l e d C o m p o s i t e 1 5 0 l p i a t 4 5 d e g r e e s taxa analyzed with the two preparation methods, provided similar groupings. The correct classifications obtained by acetolysis and Wodehouse were 99.0% and 96.9 % respectively. The data collected after acetolysis resulted in a better resolution in the genus Onosma. That is why only the scatterplot obtained from acetolysis data is shown in figure 1. The first three canonical variates explained 82.7% of the variation in acetolysis and 78.5% of the variation in the Wodehouse pollen preparation method (Tab. 4). Thus the data collected after acetolysis resulted in higher explanation of variation than the Wodehouse method. Such a difference was also seen in character correlations to the vectors in both methods. The char- acters P, E and plt were highly correlated with the first axis in the data collected after acetolysis and P, E and clg were highly correlated with the first axis in the data collected af- ter Wodehouse. In the second axis, however, totally different sets of characters were corre- lated with this axis: t, clg and plg in acetolysis and P, ex, and i in Wodehouse. Similarly to the second axis, totally different character sets were highly correlated in two different data 268 ACTA BOT. CROAT. 69 (2), 2010 BI·NZET R., KANDEMI·R I · ., ORCAN N. Fig. 1. The DFA scatterplot obtained from the analysis of acetolysis data. Tab. 4. Eigenvalues, % of variance and cumulative % comparison between Acetolysis and Wode- house. Function Acetolysis Wodehouse Eigenvalue % of Variance Cumulative % Eigenvalue % of Variance Cumulative % 1 53.809 48.5 48.5 32.801 43.2 43.2 2 23.659 21.3 69.8 16.144 21.3 64.5 3 14.241 12.8 82.7 10.679 14.1 78.5 U:\ACTA BOTANICA\Acta-Botan 2-10\282 Binzet.vp 11. listopad 2010 13:51:47 Color profile: Disabled Composite 150 lpi at 45 degrees collection methods; plg, plt and clt were highly correlated with the third axis in Acetolysis data whereas P, t and ex were highly correlated with the third axis in Wodehouse data. Principal component analysis (PCA) and UPGMA phenogram: acetolysis and Wodehouse data Two different data sets obtained from acetolysis and Wodehouse were subjected to PCA and UPGMA clustering using NTSYS-pc. The results were more or less similar to the re- sults of DFA. Although the numbers obtained for % variation explained the eigenvalues, which were close to each other, the scatterplot drawn after PCA shows differences. Simi- larly the character correlations to PC vectors show differences. The first three axes ex- plained 82.1% of total variation in acetolysis data and 76.3% in Wodehouse, which is close to the DFA result in which acetolysis resolves variation better than Wodehouse. P, E and plt characters showed high correlations to the first axis in acetolysis, whereas E, plt and plg characters were highly correlated with the first axis in Wodeouse data. On the second axis t, clt and ect/end in Acetolysis and ex, clg and ect/end in Wodehouse data have high correla- tions. The characters t, clg and ect/end in acetolysis data and P, i and ex in Wodehouse data were highly correlated with the third axis. The PCA scatterplot obtained for both type of methods showed different groupings and affinities between species. The scatterplot obtained from acetolysis data after PCA did not form any kind of strong groupings in the genus Onosma; however the formation of weak clusters did not correspond to any close affinity between species or subsections: Haplo- tricha and Asterotricha (Fig. 2). Also, the presence of weak affinities between species did ACTA BOT. CROAT. 69 (2), 2010 269 PALYNOLOGICAL CLASSIFICATION OF ONOSMA Fig. 2. PCA scatterplot of acetolysis data after standardization U:\ACTA BOTANICA\Acta-Botan 2-10\282 Binzet.vp 11. listopad 2010 13:51:48 Color profile: Disabled Composite 150 lpi at 45 degrees not correspond to any relationships in Onosma key. The scatterplot obtained from Wode- house data after PCA resulted in a very similar outcome, that the relationship between spe- cies could not be seen in the species key to the genus Onosma (Fig. 3). 270 ACTA BOT. CROAT. 69 (2), 2010 BI·NZET R., KANDEMI·R I · ., ORCAN N. Fig. 3. PCA scatterplot of Wodehouse data after standardization Fig. 4. UPGMA constructed from acetolysis data based on the general distance U:\ACTA BOTANICA\Acta-Botan 2-10\282 Binzet.vp 11. listopad 2010 13:51:53 Color profile: Disabled Composite 150 lpi at 45 degrees Acetolysis and Wodehouse data are visualized by PCA and also by UPGMA dendro- grams (Figs. 4, 5). The UPGMA dendrograms constructed do not match each other or the morphological relationships in the species of the genus Onosma. Discussion In this study, twenty-five Onosma (Boraginaceae) taxa belonging to 2 subsections (Haplotricha (Boiss.) Gürke. and Asterotricha (Boiss.) Gürke.) from the east Mediterra- nean region in Turkey were studied by palynological data approached by numerical taxo- nomic methods. It seems that none of the palynological methods (acetolysis and Wode- house) actually resulted in a good classification, similar to that given in the Flora of Turkey, for the species belonging to the genus Onosma. Numerical taxonomy uses a number of characters to construct a classification of Onosma. The Boraginaceae is a eurypalynous family (CLARKE 1977, DIEZ 1984) in which a large number of species can be recognized by their pollen characters (DIEZ and VALDES 1991). Pollen morphology may be a useful diagnostic tool in Onosma taxonomy. For example O. stenoloba Hausskn.ex Riedl and O. mersinana Riedl, Binzet et Orcan are very close to each other and pollen shape can help to distinguish these two species. Interestingly these two species grouped together in all ordination methods (DFA, PCA) except in UPGMA. However in cluster analysis (UPGMA) these two species were not very close to each other but were found in the same large group. The main outcome of this study is that the acetolysis methods resulted in a better expla- nation of palynological relationship within the genus Onosma than the Wodehouse method. This is due to some of the characters being not correctly measured in Wodehouse method. ACTA BOT. CROAT. 69 (2), 2010 271 PALYNOLOGICAL CLASSIFICATION OF ONOSMA Fig. 5. UPGMA constructed from Wodehouse data based on the general distance U:\ACTA BOTANICA\Acta-Botan 2-10\282 Binzet.vp 11. listopad 2010 13:51:55 Color profile: Disabled Composite 150 lpi at 45 degrees However in acetolysis, the fossilization of the pollen grains resulted in a better visualiza- tion of the edge of the characters and resulted in a better measurement. On the other hand some of the characters like P, E, plt, clg and t have a fairly good positive correlations be- tween acetolysis and Wodehouse, whereas, plg and ex had negative correlations and clt has a small correlation within two methods (data not shown). These characteristics resulted in differences in numerical taxonomic outcomes and the better results obtained by Acetolysis method. Thus as clearly shown in the present study, acetolysis methods would be more useful in further palynological, phylogenetical and numerical taxonomic studies. Also data from karyology and molecular markers should be included for a more complete visualization of the relationships within the genus Onosma. Acknowledgements This study was supported by Mersin University Research Fund (Project No. BAP-FBE BB (RB) 2004-1 DR). References BI·NZET, R., ORCAN, N., 2003a: Morphological, Anatomical and palynological study of Onosma bracteosum Hausskn. et Bornm. and Onosma mutabile Boiss. (Boraginaceae). Phytologia Balcanica 9, 97–111. BI·NZET, R., ORCAN, N., 2003b: Morphological and palynological studies on Onosma roussaei DC. and Onosma giganteum Lam. (Boraginaceae). The Herb Journal of Sys- tematic Botany 10, 57–76. BI·NZET, R., ORCAN, N., 2007: A new species of Onosma L. (Boraginaceae) From Southern Turkey. Novon: Journal for Botanical Nomenclature 17, 8–10. BOISSIER, E., 1879: Flora Orientalis sive enumeratio plantarum in Oriente a Graecia et Aegypto ad Indiae fines hucusque observatarum, 6: Coroliflorae et Monochlamydeae. H. Georg, Genevae and Basileae, Switzerland. CLARKE, G. C. S., 1977: The Nortwest European Pollen Flora, 10. Boraginaceae. Review of Paleobotany and Palynology 24, 50–101. DIEZ, M. J., 1984: Contribucion al Atlas Polinico de Andalucia Occidental, I. Boragina- ceae. Lagascalia 13, 147–171. DIEZ, M. J., VALDES, B., 1991: Pollen morphology of the tribes Eritrichieae and Cynoglos- seae (Boraginaceae) in the Iberian Peninsula and its taxonomic significance. 101, 49–66. DINSMOR, J. E. 1932: Flora of Syria, Palestine and Sinai. Beirut Press. Beirut. DOÐAN, M., 1997: Numerical taxonomic study on the genus Alopecurus L. (Gramineae). The Herb Journal of Systematic Botany 4, 71–76. DOÐAN, M., KENCE, A., TIÐIN, Ç., 1992: Numerical taxonomic study on Turkish Lathyrus (Leguminosae). Edinburgh Journal of Botany 49, 333–341. 272 ACTA BOT. CROAT. 69 (2), 2010 BI·NZET R., KANDEMI·R I · ., ORCAN N. U:\ACTA BOTANICA\Acta-Botan 2-10\282 Binzet.vp 11. listopad 2010 13:51:55 Color profile: Disabled Composite 150 lpi at 45 degrees DOÐAN, M., TOSUNOÐLU, C., 1992: A numerical analysis of leaf blade morphology and its possible implication over the infrageneric delimitation in the genus Helictotrichon s.I (Gramineae). Turkish Journal of Botany 16, 365–372. ERDTMAN, G., 1952: Pollen morphology and plant taxonomy. Almquist and Wiksells, Uppsala, 539. FAEGRI, K., IVERSEN, J., 1964: Textbook of pollen analysis, Munksgaard, Copenhagen, 263. GE-LING, Z., RIEDL, H., KAMELIN, R., 1995. Gentianaceae: Boraginaceae. In: ZHENGYI, W., RAVEN, P. H., eds. Flora of China, 16. 329–375. St. Louis and Beijing: Missouri Botani- cal Garden Press. HAYEK, A., MARKGRAF, F. R., 1970: Prodromus florae peninsulae Balcanicae. Verlag Des Repertoriums. HUYNH, K. L., 1971: Le Pollen de Genera Arnebia Forssk. et du genre, Macromedia DC. (Boraginaceae), et la position taxonomique particuliere du M. echioides (L.) Boiss. Candollea 26, 165–171. HUYNH, K. L., 1972 : The original position of the generative nucleus in the pollen tetrads of Agropyron, Itea, Limnanthes and Onosma, and its phylogenetic significance in the An- giosperms. Grana 12, 105–112. JOHNSTON, I. M., 1954: Studies in the Boraginaceae, 26. Further evaluations of the genera of the Lithospermae. Journal of the Arnold Arboretum 35, 1–81. KENCE, A., 1988: Numerical methods in biological clasifications (In Turkish). Proceedings 9 Ulusal Biological Congress, Sivas, 411–414. MAGGI, F. KOLARCIK, V., MARTONFI, P., 2008: Palynological analysis of five selected Onosma taxa. Biologia 63, 183–186. MARTICORENA, P. C., 1968: Granos de pollen de plantas chilenas. Gayana Botanica 17, 36–60. MEIKLE, R. D., 1985: Flora of Cyprus, 2. Bentham-Moxon Trust, Royal Botanic Gardens, Kew, London. NOWICKE, J. W., RIDGWAY, J. E., 1973: Pollen studies in the genus Cordia (Boraginaceae). American Journal of Botany 60, 584–591. NOWICKE, J. W., SKVARLA, J. J., 1974: A palynological investigation of the genus Tourne- fortia (Boraginaceae). American Journal of Botany 61, 1021–1036. PERVEEN, A., QURESHI, U. S., QAISER, M., 1995: Pollen flora of Pakistan-IV Boraginaceae. Pakistan Journal of Botany 27, 327–360. QURESHI, U. S., QAISER, M., 1987: Palynological Study of Onosma (Boraginaceae) from Pakistan. Pakistan Journal of Botany 19, 99–105. RIEDL, H., 1978: Boraginaceae. In: DAVIS, PH. (ed.), Flora of Turkey and the east Aegean is- lands, 6, 237–437. Edinburgh University Press, Edinburgh. RIEDL, H, BI·NZET, R., ORCAN, N., 2005: A New Species of Onosma (Boraginaceae- -Lithospermeae) from Southern Turkey. Edinburgh Journal of Botany 61, 127–130. ACTA BOT. CROAT. 69 (2), 2010 273 PALYNOLOGICAL CLASSIFICATION OF ONOSMA U:\ACTA BOTANICA\Acta-Botan 2-10\282 Binzet.vp 11. listopad 2010 13:51:55 Color profile: Disabled Composite 150 lpi at 45 degrees ROHLF, F. J., 2004: NTSYS-PC, Numerical Taxonomy and Multivariate Analysis System, version 2.2. Department of Ecology and Evolution, State University of New York, Exeter Software, Stony Brook, New York. SHISHKIN, B. K., 1974: Flora of the U.S.S.R., 19. Botanical Institute of the Academy of Scien- ces of the USSR, Keter Publishing House Jerusalem Ltd. Jerusalem. STATISTICAL PACKAGE FOR THE SOCIAL SCIENCES, 2004: SPSS for Windows, 13.0, Standard version. SPSS Inc., Chicago, Illinois, USA. TEPPNER, H., 1991: Onosma. In: STRID, A., TAN, K. (eds.), Mountain flora of Greece, 2, 26–29. Edinburgh University Press, Edinburgh. TOGAN, I., AYDEM, N., KENCE A., 1983: A numerical taxonomic study of Carthamus taxa in Turkey. In: FELSENSTEIN, J. (ed.), Numerical taxonomy, pp. 552–556. NATO ASI Se- ries, Springer Verlag, Berlin. TÜTEL, B., KANDEMIR, I., KUª S., KENCE, A., 2005: Classification of Turkish Plantago L. species using numerical taxonomy. Turkish Journal of Botany 29, 51–61. TUTIN, T. G., HEYWOOD, V. H., BURGES, N. A., MOORE, D. M., VALENTINE, D. H., WALTERS, S. M., WEBB, D. A., 1972: Flora Europaea, 3. Cambridge University Press, Cambridge. WODEHOUSE, R.P., 1935: Pollen grains. McGraf-Hill, New York. YILDIRIMLI, ª., 2000: The chorology of the Turkish species of Boraginaceae family. The Herb Journal of Systematic Botany 7, 257–272. 274 ACTA BOT. CROAT. 69 (2), 2010 BI·NZET R., KANDEMI·R I · ., ORCAN N. U:\ACTA BOTANICA\Acta-Botan 2-10\282 Binzet.vp 11. listopad 2010 13:51:55 Color profile: Disabled Composite 150 lpi at 45 degrees