Caryologia. International Journal of Cytology, Cytosystematics and Cytogenetics 73(3): 133-151, 2020 Firenze University Press www.fupress.com/caryologia ISSN 0008-7114 (print) | ISSN 2165-5391 (online) | DOI: 10.13128/caryologia-569 Caryologia International Journal of Cytology, Cytosystematics and Cytogenetics Citation: N. Allah Raei Niaki, F. Attar, M. Mirtadzadini, K. Mahdigholi (2020) Pol- len and floral micromorphological Stud- ies of the genus Cotoneaster Medik. (Rosaceae) and its systematic impor- tance. Caryologia 73(3): 133-151. doi: 10.13128/caryologia-569 Received: July 18, 2019 Accepted: July 16, 2020 Published: December 31, 2020 Copyright: © 2020 N. Allah Raei Niaki, F. Attar, M. Mirtadzadini, K. Mahdigholi. This is an open access, peer-reviewed article published by Firenze University Press (http://www.fupress.com/caryo- logia) and distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All rel- evant data are within the paper and its Supporting Information files. Competing Interests: The Author(s) declare(s) no conflict of interest. Pollen and floral micromorphological Studies of the genus Cotoneaster Medik. (Rosaceae) and its systematic importance Nemat Allah Raei Niaki1, Farideh Attar1,*, Mansour Mirtadzadini2, Kazem Mahdigholi1 1Central Herbarium of Tehran University, School of Biology, College of Science, Univer- sity of Tehran, Iran 2 Deptartment of Biology, Faculty of Science, Shahid Bahonar University, PO Box 76169- 133, Kerman, Iran * Corresponding author. E-mail: fattar@khayam.ut.ac.ir) Abstract. The micromorphology of petal and pollen grain of 16 species of the genus Cotoneaster Medik., belonging to two subgenera (Cotoneaster and Chaenopetalum) has been studied by light and scanning electron microscopy. We used different multivariate statistical methods to reveal the species relationships. Results showed that in compari- son to most genera of the family Rosaceae, both tri- and tetracolporate pollen grains are observed in one specimen. Palynological observation revealed that shape of tricolpo- rate pollen in most species is prolate-spheroidal, but also subprolate and prolate pollen grains can be recognized. In the other hand, tetracolporate pollen is quadrangular. The main ornamentation type was mainly striate which in turn can be subdivided to several categories; however, psilate one is also recognized also (C. persicus). The results revealed that pollen traits are probably effective in separating the sections and using these traits for placing a species in a particular section is probably helpful. Apomixis is one of the reasons for not changing the ornamentation of the both surface of the petals and their similarity to each other in different species. Totally, the studied micromorphological characters of petal cannot be used as diagnostic tools for Cotoneaster in Iran. Keywords: Cotoneaster, Iran; pollen, petal, Rosaceae, SEM. INTRODUCTION The genus Cotoneaster Medik. which is mostly a shrubby member of the subtribe Pyrinae, tribe Pyreae, subfamily Spiraeoideae, family Rosaceae (Campbell et al. 2007). Cotoneaster consists of about 90 species widespread in temperate Asia (except Japan), Europe and North Africa (Yü and Lu 1974; Lu and Brach 2003), although other authors consider the number of species 260 (Mabberley 2008) to 400 (Fryer and Hylmö 2009). The center of diversity for the genus is in Tibet and the Himalayas, though species are native across Asia, North Africa, and Europe (Bartish et al. 2001; Dickore and Kasperek 2010; Fryer and Hylmö 2009). In Iran, 19 species of the genus are mainly dis- 134 Nemat Allah Raei Niaki, Farideh Attar, Mansour Mirtadzadini, Kazem Mahdigholi tributed in Alborz Mts., elevations in NW (Azerbaijan province) and NE (Khorasan province) (Raei Niaki et al. 2009). Among these species, the C. assadii, C. esfan- diarii and C. persicus are endemic to Iran (Riedl 1969; Khatamsaz 1985; Khatamsaz 1992). The petal traits or the number of flowers in the cyme is the main characters used to determine interspecific relationships in Cotoneaster (Koehne 1893, Yu 1963). In some families, petal morphology is one of the most important diagnostic characters (Sharma et al. 2005, Campbell et al. 2007, Akcin 2009, Arianmanesh et al. 2016). The patterns of petal epidermis in angiosperms particularly Rosaceae family have important charac- ters for identification of close species (Christensen and Hansen 1998). Several researchers have focused on petal micromorphology of different genera of Rosaceae (Tahir et al. 2010, Sharifnia and Behzadi Shakib 2012, Omer et al. 2017). Regarding pollen morphology, it has been proved to be beneficial in systematic of the family Rosaceae (Hebda and Chinnappa 1990); however some others deny such an application (Moore et al. 1991) which is caused by easy hybridization among several species and even genera of the family. Regarding Cotoneaster, some a few studies (Kumar and Panigrahi 1995; Hsieh and Huang 1997; Perveen and Qaiser 2014) have reported some common features of pollen such as size, aperture number, exine thickness and ornamentation of surface. According to these studies, ornamentation of pollen sur- face including striate, sub-psilate and regulate ones is the most important feature in separating species. How- ever, generally they emphasized these characters only play a little role in separating a few species and pollen morphology is not a useful tool in classification of the genus. A comprehensive study on morphological and micro-morphological characters in Cotoneaster is almost lacking, moreover, the potential application of these characters in taxonomy of the genus has not been illus- trated yet. Therefore, the objectives of the present study were 1) to provide detailed morphological and micro- morphological information on petal and pollen mor- phology of Cotoneaster, and 2) to evaluate application of these characters to find out the inter species relation- ships and delimit the species taxonomically. MATERIAL AND METHODS Pollen sampling Totally 42 populations were collected and studied from 16 taxa of Cotoneaster from different habitats in Iran for study the pollen features (Table 1). 5-8 individu- als of each location were studied and examined for 2 qualitative and 13 quantitative features (Table 2 and 3). Voucher specimens were deposited in TUH and FUMH (Table 1). Pollen obtained from flower buds at anthesis were prepared for light microscope (LM) using meth- ods described by Harley (1992) with some modifications, mounted in glycerol jelly on glass slides and sealed. For LM measurements, at least 20‒25 pollen grains were measured by Nikon light microscope model 200 M with aid of a × 100 eyepiece. For scanning electron micros- copy (SEM) examinations, pollen grains were not ace- tolysed according to the method of Erdtman (1960). The pollen were suspended in a drop of water for a while, and then directly transferred to a metallic stub by a fine pipette, and double sided cello tape were used and then the pollen were sputtered in chamber coated with gold (Sputter Coater BALTEC, SCDOOS). Coating with gold by the physical vapor deposition method (PVD) was restricted to 100 Å. The SEM examination was carried out on a TESCAN microscope. For detailed examination of sculpturing, the classification presented in Ueda and Tomita (1989) was used. For estimation of pollen fertil- ity, the pollen from fresh collected herbarium materials were stained by acetocarmine glycerin jelly, as described by Radford et al. (1974). Petal sampling In the current study, the micromorphological char- acteristics of petals of 16 species belonging to two sub- genera of the Cotoneaster (Cotoneaster and Chaeno- petalum) were studied for the first time. The collected specimens were deposited in the herbarium of Tehran University (TUH), Ferdowsi University of Mashhad Her- barium (FUMH) (Table 1). The voucher specimens are listed in Table 1. 122 specimens from 42 different popu- lations of Cotoneaster taxa were collected from their nat- ural habitats in Iran. Five micromorphological charac- ters were examined; among the five characters, four were qualitative and one character was quantitative (Table 4). The materials for SEM observation were directly mounted on stubs without any treatment, and sputter coated with gold-palladium. Petals and sepals of fully opened f lowers were investigated using a HITACHI model SU 3500 electron microscope at 15 kV accelerat- ing voltage; After a number of specimens had been com- pared under SEM. Both petal surfaces were examined. The epidermis of the petals was classified based on cell ornamentation, shape of cell (the primary sculpture), visibility of the boundary between cells using the termi- nology of Barthlott (1981) and Kay et al. (1981). 135Pollen and floral micromorphological Studies of the genus Cotoneaster Medik. (Rosaceae) and its systematic importance Data analysis The characters of pollen grains of the studied species are summarized in Tables 2 and 3. Multi-state qualita- tive characters converted into presence-absence descrip- tions. 13 pollen grain quantitative data were noted and treated statistically to determine average values for each species. PCA analysis were performed to check the similarity and dissimilarity between different taxa of the tribe, after linear standardization by range of each variable of the original data set. Principal Components Analysis (PCA) was performed to check the dissimilar- ity between the studied species based on palynological features useful for the delimitation of the species. To calculated Euclidean and taxonomic distance between different species of the genus, PCA ordination plot was performed (Podani 2000). PAST version 2.17 (Hammer et al. 2012) was used for multivariate statistical analyses of morphological data. RESULTS In the present investigation different micro mor- phological characters of the petal and pollen grain of Cotoneaster have been studied in detail. The investigat- ed specimens are given in Table 1. The petal and pollen morphological characters are summarized in Table 2, 3 and 4. Pollen morphology The most important characters are given in table 2 and 3. Selected micrographs are presented in (Figures 1, 2, 3). Pollen grains are shed as monad, medium-sized (P= 29.26–35.13). One important and interesting fea- ture observed frequently in most of the studied species, is the presence of tri- and tetracolporate pollen grains in the same specimen (e.g. Figures 1. e, o; Figures 2. e, k, q), while most species of family Rosaceae consist of tricolporate pollen. Percentage of this character is vari- able in different species, so that tetracolporate pollen in some species such as C. melanocarpus Fisch. ex A.Blytt , C. kotschyi (C.K.Schneid.) G.Klotz and C. multiflorus Bonge. is frequent, while the percentage of tricolporate pollen in some others such as C. tytthocarpus Pojrk. and C. morulus Pojark. constitute the major percentage of pollen grains. Regarding symmetry, both tricolporate and tetracolporate pollen grains are isopolar (e.g. Fig- ures 1. i, k, m; Figures 2. c, e, f, g, m) and heteropolar (e.g.Figures 1. c, e, q). In the case of the shape of tri- Table 1. List of the investigated taxa including origin of voucher specimens. Taxon Voucher information C. subgen. Cotoneaster C. melanocarpus ( Ledeb.) Lodd., G. Lodd. & W. Lodd. ex M. Roem. Mazandaran: Firoozkooh road, 30 km after Veresk village to Tehran, after Dogol station. Raei Niaki & Mahdigholi. 46888-TUH C. subgen. Chaenopetalum C. multiflorus Bunge. Kurdistan: Nushoor olia village; Attar, Raei Niaki & Maroofi, 46870-TUH C. suavis Pojark. Khorasan: Gifan, Misino mountain, 20790-FUMH C. hissaricus Pojark. Azerbayjan: after Peygham village to Kaleybar; Attar, Zamani & Raei Niaki, 37261-TUH. C. morulus Pojark. Azerbayjan: Orumiyeh, Marmishu lake, Attar & Zamani, 40614-TUH C. tytthocarpus Pojark. Mazandaran: Siah-bishe, 7 km after Pole-Zanguleh to Chalus. Raei Niaki & Mahdigholi, 46887-TUH C. luristanicus G. Klotz Luristan: Aleshtar. Ghahraman, Attar & Ghaffari. 21658-TUH C. turcomanicus Pojark. Mazandaran: Firoozkooh Road, 30 Km after Veresk village to Tehran; Raei Niaki & Mahdigholi. 46890-TUH C. nummularioides Pojark. Mazandaran: Chalus road, between Reyzamin and Asara village; Attar, Zamani & Raei Niaki; 37203- TUH C. kotschyi (C.K.Schneid.) G.Klotz Mazandaran: Haraz Road, Yush village. Raei Niaki & Mahdigholi, 46897-TUH C. assadii khat. Mazandaran: Siah-bishe, Gachsar village; Raei Niaki & Mahdigholi, 46898-TUH C. nummularius Fisch. & C.A.Mey. Mazandaran: Firoozkooh Road, Seleben Village; Raei Niaki & Mahdigholi, 46901-TUH C. ovatus Pojark. Mazandaran: Firoozkooh Road, Seleben Village; Raei Niaki & Mahdigholi, 46892-TUH C. esfandiarii khat. Mazandaran: Firoozkooh, Arjmand village, Attar & Raei Niaki, 46886-TUH C. discolor Pojark. Mazandaran: Firoozkooh Road, 30 km after Veresk village to Tehran, Raei Niaki & Mahdigholi, 46889- TUH C. persicus Pojark. Khorasan: Pivehjan village, 23394-FUMH 136 Nemat Allah Raei Niaki, Farideh Attar, Mansour Mirtadzadini, Kazem Mahdigholi Ta bl e 2. E va lu at ed c ha ra ct er s of p ol le n gr ai ns in C ot on ea ste r sp ec ie s st ud ie d (v al ue s M ± S D μm ). M - M ea n va lu e; S D - St an da rd d ev ia tio n. A bb re vi at io ns : P , p ol ar a xi s le ng th ; E , eq ua to ria l a xi s l en gt h; P /E , p ro po rt io n of p ol ar a xi s t o eq ua to ria l a xi s l en gt h; M , M es oc ol pi um le ng th ; A , A po co lp iu m le ng th ; C , C ol pu s l en gt h; E , E xi ne th ic kn es s Ta xa P (µ m ) E (µ m ) P/ E( µm ) M (µ m ) A (µ m ) C (µ m ) E (µ m ) C. S ub ge n. Co to ne as te r C. S ec t.C ot on ea ste r C. S er . C ot on ea ste r C. in te ge rr im us 31 .0 (3 3. 73 ±2 .1 2) 39 .0 23 .0 (2 8. 20 ±3 .7 6) 35 .0 1. 3 12 .0 0( 18 .5 3± 4. 61 )2 5. 00 5. 00 (6 .6 0± 1. 40 )9 .0 0 27 .0 0( 29 .4 7± 2. 26 )3 5. 00 0. 70 (0 .9 7± 0. 12 )1 .1 0 C. S er . M ela no ca rp i C. m ela no ca rp au s 32 .0 (3 5. 00 ±1 .9 3) 39 .0 23 .0 (2 7. 47 ±3 .4 6) 34 .0 1. 3 13 .0 0( 16 .9 3± 2. 43 )2 3. 00 5. 00 (6 .7 5± 1. 04 )8 .0 0 28 .0 0( 30 .0 7± 1. 62 )3 3. 00 0. 70 (1 .6 9± 0. 42 )2 .0 0 C. S ub ge n. Ch ae no pe ta lu m C. S ec t.C ha en op et al um C. S er .R ac em ifl or i C. p er sic us 29 .0 0( 33 .4 0± 2. 85 )3 9. 00 18 .0 0( 25 .2 0± 4. 77 )3 5. 00 1. 37 11 .0 0( 17 .9 3± 5. 19 )3 0. 00 4. 00 (5 .8 7± 1. 12 )8 .0 0 25 .0 0( 29 .4 7± 2. 39 )3 4. 00 1. 00 (1 .7 1± 0. 32 )2 .0 0 C. d isc ol or 30 .0 0( 31 .1 0± 1. 28 )3 4. 00 25 .0 0( 28 .2 0± 2. 73 )3 3. 00 1. 11 17 .0 0( 21 .7 3± 4. 32 )3 0. 00 5. 00 (6 .3 3± 1. 18 )8 .0 0 22 .0 0( 22 .4 7± 1. 99 )3 0. 00 0. 70 (0 .9 7± 0. 12 )1 .1 0 C. a ss ad ii 30 .0 0( 32 .4 3± 2. 35 )3 5. 00 23 .0 0( 26 .0 0± 3. 67 )3 3. 00 1. 25 16 .0 0( 20 .5 3± 5. 73 )2 5. 00 4. 00 (6 .1 2± 1. 32 )9 .0 0 23 .0 0( 26 .0 5± 1. 25 )2 8. 00 0. 85 (1 .3 4± 0. 56 )1 .3 0 C. n um m ul ar iu s 31 .0 0( 33 .7 3± 2. 12 )3 9. 00 23 .0 0( 31 .0 0± 3. 82 )4 0. 00 1. 1 15 .0 0( 22 .6 7± 5. 33 )3 3. 00 4. 00 (7 .4 7± 1. 46 )1 0. 00 28 .0 0( 30 .1 3± 1. 96 )3 5. 00 0. 70 (1 .3 0± 0. 43 )2 .0 0 C. e sfa nd ia rii 28 .0 0( 30 .9 3± 2. 31 )3 5. 00 23 .0 0( 28 .2 0± 3. 76 )3 5. 00 1. 11 16 .0 0( 21 .9 3± 4. 88 )3 2. 00 4. 00 (9 .2 7± 1. 44 )8 .0 0 23 .0 0( 26 .0 0± 2. 03 )3 0. 00 1. 00 (1 .7 9± 0. 45 )2 .6 0 C. o va tu s 29 .0 0( 33 .4 7± 1. 76 )3 6. 00 25 .0 0( 30 .8 7± 3. 87 )3 6. 00 1. 1 11 .0 0( 20 .6 0± 4. 92 )2 8. 00 4. 00 (6 .0 7± 1. 39 )9 .0 0 26 .0 0( 29 .8 7± 2. 53 )3 6. 00 0. 80 (1 .3 9± 0. 48 )2 .0 0 C. S er .H iss ar ici C. h iss ar icu s 30 .0 0( 33 .8 0± 2. 01 )3 8. 00 21 .0 0( 27 .2 7± 3. 55 )3 4. 00 1. 26 12 .0 0( 18 .5 3± 4. 61 )2 5. 00 5. 00 (6 .9 3± 1. 03 )9 .0 0 27 .0 0( 29 .4 7± 2. 26 )3 5. 00 0. 90 (1 .0 9± 0. 20 )1 .7 0 C. tu rc om an icu s 31 .0 0( 35 .1 3± 3. 11 )4 3. 00 21 .0 0( 33 .2 0± 5. 72 )4 0. 00 1. 1 10 .0 0( 25 .8 0± 8. 21 )3 5. 00 5. 00 (7 .4 7± 1. 19 )9 .0 0 23 .0 0( 30 .2 0± 4. 34 )4 0. 00 0. 80 (1 .3 8± 0. 46 )2 .5 0 C. m or ul us 25 .0 0( 30 .2 0± 2. 98 )3 7. 00 13 .0 0( 21 .2 0± 3. 51 )2 6. 00 1. 47 7. 00 (1 3. 47 ±3 .5 8) 20 .0 0 5. 00 (6 .6 0± 1. 40 )9 .0 0 21 .0 0( 26 .1 3± 3. 02 )3 3. 00 1. 00 (1 .2 6± 0. 20 )1 .7 0 C. ty tth oc ar pu s 25 .0 0( 29 .2 6± 3. 01 )3 5. 00 18 .0 0( 22 .9 3± 3. 79 )3 2. 00 1. 29 9. 00 (1 3. 33 ±2 .7 7) 20 .0 0 5. 00 (6 .2 5± 2. 50 )1 0. 00 20 .0 0( 25 .3 3± 3. 11 )3 0. 00 0. 90 (1 .4 4± 0. 46 )2 .1 0 C. lu ri sta ni cu s 31 .0 0( 33 .5 3± 2. 20 )3 7. 00 27 .0 0( 33 .4 0± 3. 16 )3 8. 00 1. 01 17 .0 0( 26 .2 0± 5. 68 )3 7. 00 5. 00 (6 .9 7± 1. 26 )1 0. 00 25 .0 0( 28 .8 0± 2. 73 )3 3. 00 1. 20 (1 .7 5± 0. 33 )2 .5 0 C. k ot sc hy i 30 .0 0( 33 .2 0± 2. 40 )3 8. 00 21 .0 0( 24 .2 7± 3. 06 )3 2. 00 1. 39 11 .0 0( 16 .1 3± 4. 75 )3 0. 00 4. 00 (5 .1 0± 0. 96 )7 .0 0 26 .0 0( 29 .5 3± 2. 10 )3 4. 00 0. 80 (1 .2 3± 0. 33 )1 .9 0 C. n um m ul ar io id es 25 .0 0( 31 .2 7± 2. 68 )3 7. 00 21 .0 0( 23 .5 3± 2. 03 )2 8. 00 1. 34 12 .0 0( 15 .8 0± 2. 24 )2 0. 00 4. 00 (6 .5 3± 1. 36 )9 .0 0 21 .0 0( 27 .5 3± 2. 88 )3 3. 00 1. 00 (1 .4 0± 0. 38 )2 .0 0 C. S ec t.M ul tifl or i C. S er . M ul tifl or i C. m ul tifl or us 30 .0 0( 34 .7 8± 2. 82 )3 7. 00 20 .0 0( 26 .8 9± 5. 04 )3 6. 00 1. 32 8. 00 (1 4. 56 ±4 .1 9) 21 .0 0 5. 00 (7 .2 1± 1. 25 )9 .0 0 25 .0 0( 29 .6 7± 3. 08 )3 4. 00 1. 50 (1 .7 6± 0. 22 )2 .0 0 137Pollen and floral micromorphological Studies of the genus Cotoneaster Medik. (Rosaceae) and its systematic importance Ta bl e 3. C on tin ue e va lu at ed c ha ra ct er s o f p ol le n gr ai ns in C ot on ea ste r s pe ci es st ud ie d (v al ue s M ± S D μ m ). M - M ea n va lu e; S D - S ta nd ar d de vi at io n. A bb re vi at io ns : C /P , p ro po rt io n of co lp us to p ol ar a xi s l en gt h; S , S ha pe o f p ol le n; S c, Sc ul pt ur in g of p ol le n; F P, F er til ity p er ce nt ag e Ta xa C /P S Sc FP po re n um be r rid ge w id th in te r r id ge w id th po re w id th C. S ub ge n. Co to ne as te r C. S ec t.C ot on ea ste r C. S er . C ot on ea ste r C. in te ge rr im us 0. 83 su bp ro la te Ty pe II -A 96 % 2. 00 (8 .8 7± 3. 00 )1 3. 00 0. 19 (0 .2 2± 0. 03 )0 .2 8 0. 04 (0 .0 8± 0. 02 )0 .1 4 0. 05 (0 .1 4± 0. 05 )0 .2 2 C. S er . M ela no ca rp i C. m ela no ca rp au s 0. 85 su bp ro la te Ty pe II -A 95 % 7. 00 (1 0. 33 ±2 .4 7) 16 .0 0 0. 18 (0 .3 1± 0. 08 )0 .4 7 0. 05 (0 .1 4± 0. 07 )0 .3 7 0. 05 (0 .2 1± 0. 09 )0 .3 8 C. S ub ge n. Ch ae no pe ta lu m C. S ec t.C ha en op et al um C. S er .R ac em ifl or i C. p er sic us 0. 88 su bp ro la te Ty pe V I 85 % 4. 00 (6 .1 1± 1. 80 )9 .0 0 0. 08 (0 .1 8± 0. 05 )0 .2 4 0. 07 (0 .1 3± 0. 04 )0 .2 2 0. 07 (0 .1 6± 0. 03 )0 .2 6 C. d isc ol or 0. 72 pr ol at e- sp he ro id al Ty pe V 97 % 1. 00 (4 .4 7± 2. 36 )9 .0 0 0. 12 (0 .2 1± 0. 05 )0 .2 9 0. 10 (0 .2 2± 0. 07 )0 .3 3 0. 08 (0 .1 6± 0. 07 )0 .2 9 C. a ss ad ii 0. 82 su bp ro la te Ty pe II I 92 % 7. 00 (9 .4 7± 2. 10 )1 5. 00 0. 20 (0 .2 6± 0. 04 )0 .3 5 0. 06 (0 .0 8± 0. 02 )0 .1 1 0. 07 (0 .1 6± 0. 04 )0 .2 6 C. n um m ul ar iu s 0. 89 pr ol at e- sp he ro id al Ty pe I 98 % 2. 00 (6 .9 3± 2. 41 )1 0. 00 0. 09 (0 .1 8± 0. 04 )0 .2 5 0. 06 (0 .1 3± 0. 04 )0 .2 3 0. 06 (0 .1 4± 0. 05 )0 .2 4 C. e sfa nd ia rii 0. 84 pr ol at e- sp he ro id al Ty pe I 98 % 4. 00 (6 .1 3± 1. 82 )9 .0 0 0. 08 (0 .1 5± 0. 04 )0 .2 4 0. 07 (0 .1 3± 0. 04 )0 .2 2 0. 13 (0 .1 8± 0. 03 )0 .2 5 C. o va tu s 0. 89 pr ol at e- sp he ro id al Ty pe II I 98 % 7. 00 (1 2. 53 ±3 .5 2) 17 .0 0 0. 12 (0 .1 7± 0. 03 )0 .2 4 0. 05 (0 .0 8± 0. 02 )0 .1 2 0. 07 (0 .1 0± 0. 03 )0 .1 7 C. S er .H iss ar ic i C. h iss ar icu s 0. 87 su bp ro la te Ty pe I 96 % 6. 00 (8 .2 7± 2. 41 )1 4. 00 0. 10 (0 .1 6± 0. 04 )0 .2 2 0. 07 (0 .1 6± 0. 07 )0 .3 0 0. 09 (0 .1 9± 0. 06 )0 .3 4 C. tu rc om an ic us 0. 85 pr ol at e- sp he ro id al Ty pe II -A 99 % 2. 00 (5 .9 3± 2. 29 )9 .0 0 0. 17 (0 .2 7± 0. 07 )0 .4 5 0. 13 (0 .2 0± 0. 04 )0 .2 9 0. 18 (0 .2 3± 0. 03 )0 .3 0 C. m or ul us 0. 86 pr ol at e Ty pe II -B 96 % 9. 00 (1 2. 40 ±3 .1 6) 21 .0 0 0. 15 (0 .2 2± 0. 05 )0 .3 3 0. 07 (0 .1 4± 0. 04 )0 .2 0 0. 08 (0 .1 5± 0. 07 )0 .3 3 C. ty tth oc ar pu s 0. 86 su bp ro la te Ty pe I 90 % 0. 00 (5 .7 3± 4. 12 )1 4. 00 0. 18 (0 .2 9± 0. 06 )0 .3 8 0. 06 (0 .1 0± 0. 02 )0 .1 3 0. 08 (0 .1 6± 0. 08 )0 .3 9 C. lu ri sta ni cu s 0. 85 pr ol at e- sp he ro id al Ty pe II I & IV 99 % 5. 00 (7 .7 3± 1. 65 )1 0. 00 0. 13 (0 .2 1± 0. 05 )0 .2 9 0. 07 (0 .1 4± 0. 05 )0 .2 3 0. 10 (0 .3 1± 0. 14 )0 .5 9 C. k ot sc hy i 0. 88 pr ol at e Ty pe V 95 % 3. 00 (5 .0 7± 1. 12 )7 .0 0 0. 13 (0 .1 8± 0. 03 )0 .2 6 0. 09 (0 .1 4± 0. 03 )0 .2 1 0. 08 (0 .1 9± 0. 06 )0 .3 0 C. n um m ul ar io id es 0. 88 pr ol at e Ty pe V 88 % 0. 00 (1 2. 20 ±5 .1 3) 20 .0 0 0. 14 (0 .2 4± 0. 05 )0 .3 4 0. 08 (0 .1 1± 0. 02 )0 .1 5 0. 08 (0 .1 2± 0. 04 )0 .2 4 C. S ec t.M ul tifl or i C. S er . M ul tifl or i C. m ul tifl or us 0. 85 su bp ro la te Ty pe I & IV 96 % 0. 00 (4 .1 3± 3. 01 )9 .0 0 0. 14 (0 .2 5± 0. 05 )0 .3 5 0. 08 (0 .1 2± 0. 02 )0 .1 6 0. 07 (0 .1 7± 0. 07 )0 .3 4 138 Nemat Allah Raei Niaki, Farideh Attar, Mansour Mirtadzadini, Kazem Mahdigholi colporate pollen in equatorial view, prolate- spheroidal (e.g. Figures 2. m), subprolate (e.g. Figure 2. i) and pro- late (Figure 1. i) shapes (column S in Table 3) are recog- nized, while in polar view triangular (e.g. Figures 1.a, q) and trilobate (e.g Figures 1. h, Figures 2. a, k) shapes can be recognized. In the other hand, tetracolporate pollen are quadrangular (e.g. Figures 1. c, e, Figures 2. e, q). Shape of apex varies from obtuse (e.g. Figures 1. c, g, m, Figure 2. e) to truncate (e.g. Figure 1. E; Figures 2. g, i). Colpi which occupy 72% in C. discolor Pojark. to 89% in C. nummularius Fisch. and C. ovatus Pojark. of length of the polar axis, are arranged meridionally (e.g. Figure 1. G; Figure 2. e) or parallel (e.g. Figures 1 i, k; Figures 2. g, i). Endopores which are located in the mid- dle of ectocolpi, consist of distinct (e.g. Figures 1. c, e, g) or indistinct (e.g. Figures 2. c, g, i) projections. The mean of polar axis length (column P in Table 2) var- ies from 29.26 µm in C. tytthocarpus to 35.13 µm in C. turcomanicus Pojark. while the mean of equatorial axis length (column E in Table 2) varies from 21.20 µm in C. morulus to 33.40 µm in C. luristanicus G. Klotz. The mean of mesocolpium axis length (column M in Table 2) varies from 13.33 µm in C. tytthocarpus to 26.20 µm in C. luristanicus. Regarding apocolpium axis length (col- umn A in Table 2), range is from 5.10 µm in C. kotschyi to 9.27 µm in C. esfandiarii. The mean of colpus length (column C in Table 2) varies from 22.47 µm in C. dis- color to 30.20 µm in C. turcomanicus. The thickness of exine (column E in Table 2) which is clearly composed of two layers (ectexine and endexine) varies from 0.97 µm in C. discolor to 1.79 µm in C. esfandiarii Khat. Results of fertility test showed that most species have high per- centage of fertility so that this character (column FP in Table 3) ranges from 85% in C. persicus to 99% in C. tur- comanicus. With regard to sculpturing, the prominent ornamentation is striate (e.g. Figures 1. d, f, l, p; Figure 2. h); however some others such as psilate (Figures 2. o, p) can be recognized. Also the sterile pollen grains have deformed shape (Figure 3. q). As illustrated above, main feature of several spe- cies (i.e. sculpturing) is very homogenous in different species. But type of sculpturing, number of perfora- tion and the perforation size is different in same spe- cies. For example, series Hissarici members represent a rather uniform group but different types of sculptur- ing are observed in these species. On the basis of this character, C. persicus is separated from other species with its psilate sculpturing (Figures 2. o, p). This spe- cies is closely related to C. discolor, but differs from it by subglabrous upper leaf surface (very sparsely pilose – strigose in C. persicus), red vein and petiole (green in C. persicus). According to exine sculpturing pattern, two main types (striate) and non-striate (psilate) were recognized in the Cotonoster. Most of the specimens belong to types striate. Table 4. Distribution and coding of main petal characteristics in studied species. Taxon Number of conical projections in 50 µm2 Distinct or not distinct Boundaries between cells Closely or not closely conical projections Folding or not folding of top of conical projections Oriented or not oriented of conical projections C. subgen. Cotoneaster C. melanocarpus 12 _ + + _ C. subgen.Chaenopetalum C. multiflorus 9 + _ _ + C. suavis 9 _ + + _ C. hissaricus 13 + _ + + C. morulus 12 + _ + + C. tytthocarpus 19 _ + _ _ C. luristanicus 16 _ + + _ C. turcomanicus 11 _ + + + C. nummularioides 12 + _ _ + C. kotschyi 17 _ + _ _ C. assadii 17 + _ + _ C. nummularius 25 _ + _ _ C. ovatus 8 + _ + _ C. esfandiarii 18 _ + + + C. discolor 17 + + _ + C. persicus 13 + _ _ _ 139Pollen and floral micromorphological Studies of the genus Cotoneaster Medik. (Rosaceae) and its systematic importance Figure 1. SEM micrographs of pollen grains of C. integerimus (a-b), C. melanocarpus (c-d), C. turcomanicus (e-f), C. morulus (g-j), C. ovatus (k-l), C. assadii (m-n), C. luristanicus (o-p), C. nummularioides (q-r). 140 Nemat Allah Raei Niaki, Farideh Attar, Mansour Mirtadzadini, Kazem Mahdigholi Type (I): striate This type is recognized by distribution of lira throughout pollen surface. This type is subdivided according to the pattern of perforation between lira. Subtype I Striate pollen, which has clear fingerprint-like ridg- es with few small perforations and with long intervals of ridges. This subtype is observed in these species: C. nummularius, C. esfandiarii, C. hissaricus Pojark., C. tyt- thocarpus and C. multiflorus (Figures 2 f, h, j, l, n). The first two species surely belong to the subgenus Chaenopetalum, section Chaenopetalum, series Racemi- flori (classification according Fryer and Hylmo, 2009). They have semi-dense inflorecensc and red fruit (except C. esfandiarii) and the two latter species are members of the Hissarici series that they have lax inflorecens and black fruit. This type also is observed in the last species; C. multiflorus from subgenus Chaenopetalum, section Multiflori, series Multiflori with open inflorescence and lower surface of leaves scarcely hairs but pollen of this species has very large perforation similar to that is seen in subtype IV. Subtype type II (A-B) This subtype differs from subtype I by having prom- inent perforations between ridges. This type is subdivid- ed according to interval of ridges; subtype II-A and type subtype II- B with short and long intervals, respectively. Subtype II-A is observed in C. integerrimus Medik., C. melanocarpus and C. turcomanicus (Figures 1. b, d, f). Subtype II-B is seen in C. morulus ( Figure 1. j) Subtype III This subtype differs from subtype II, subtype III has short ridges (0.15 to 0.30 µm) and can be seen in C. assadii Khat., C. ovatus and C. luristanicus (Figures 1. l, n, p). The first two species are members of the series Racemif lori, the latter species is the member of the Hissarici series. Subtype IV This subtype is diagnosed by having very large per- forations. Subtype V This subtype is recognized by having obscure ridges due to very moderate slope of ridge. This subtype is seen in C. discolor, C. kotschyi and C. nummularioides Pojark. (Figures 2. b, d; Figure 1. r). First species belongs to the series Racemiflori and the latter two species are in Hissa- rici series. Type (II): Psilate This type is diagnosed by having no ridge on the pollen surface. This type is seen in C. persicus (Figures 2. o-r). This species is a member of the series Racemiflori and lacks any perforation on the surface. Petal morphology The micromorphological characters of petals of 16 species belonging to two subgenera of the Cotoneaster were studied. Also, according to previous studies on pet- als of other genera of Rosaceae, the ornamentations of the adaxial surface and the lower surface of the petals are described. Adaxial surface of petals: On the adaxial surface of all petals conical (finger- like or tubercle) shape projection are observed. C. melanocarpus: The epidermal cells of the petal surface are loosely packed with distinct outline. This species exhibits irregular folds and rugose tuberculate pattern. The surface of each cell exhibits striate to rugose pattern. The ruga and striae are condensed and forming ruminate pattern on the tubercle of folds (Figure 4. a). C. multiflorus: The petal surface cells are distinct and loosely packed with distinct cell walls. The margin of cells is smooth. The central part of the cells is raised into small regular finger-like projections. A tubercle is formed in the middle of the finger-like projection with ruminate patterns. The surface of each cell exhibiting striate to rugose pattern (Figure 4. b). C. suavis: Adaxial surface has loosely packed cells with prominent cell boundaries and more or less thick folds, forming tubercle in the middle of the folds. The surface as a whole is striate to rugose but at the tubercle becomes ruminate (Figure 4. c). C. hissaricus: Petal surface of this species exhib- its closely packed epidermal cells. The cell surfaces are raised into broad finger-like projections or tubercles. The surface as a whole shows striate pattern which is paral- 141Pollen and floral micromorphological Studies of the genus Cotoneaster Medik. (Rosaceae) and its systematic importance Figure 2. SEM micrographs of pollen grains of C. kotschyi (a-b), C. discolor (c-d), C. multiflorus (e-f), C. tytthocarpus (g-h), C. hissaricus (i-j), C. nummularius (k-l), C. esfandiarii (m-n), C. persicus (o-p), multiflorus (q), C. discolor (r). 142 Nemat Allah Raei Niaki, Farideh Attar, Mansour Mirtadzadini, Kazem Mahdigholi Figure 3. LM micrographs of pollen grains of C. integerimus (a), C. melanocarpus (b), C. persicus (c), C. discolor (d), C. assadii (e), C. num- mularius (f), C. esfandiarii (g), C. ovatus (h), C. hissaricus (i), C. turcomanicus (j), C. morulus (k), C. tytthocarpus (l), C. luristanicus (m), C. kotschyi (n),C. nummularioides (o), C. multiflorus (p), fertil and steril pollen grain (q). 143Pollen and floral micromorphological Studies of the genus Cotoneaster Medik. (Rosaceae) and its systematic importance lel all over the surface except at the tubercles which are intermingled with together in these pearts (Figure 4. d). C. morulus: The epidermal cells of the petal surface were loosely packed. Central part of each cell is raised into a fold surrounded by thick flat boundaries with distinct outline and rugose-tuberculate surface pattern which is condensed in the central fold or tubercle giving ruminate appearance. Cell margin is flat with smooth patterns (Figure 4. e). C. tytthocarpus: The petal surface of this species exhibits closely packed epidermal cells. The cell surface is raised into broad finger-like projections and more or less thick folds, forming tubercle in the middle of the cell. The surface as a whole is rugose but at the tubercle becomes ruminate (Figure 4. f). C. luristanicus: The epidermal cells of the petal sur- face are closely packed without distinct outline, showing rugose tuberculate pattern. The cell surface is raised into irregular projection giving appearance of simple folds or V-shaped folds. Ruga are observed all over the surface running parallel to each other or intermingling at the tubercle (Figure 4. g). C. turcomanicus: The epidermal cells of petal sur- face are closely packed with rugose-tuberculate surface pattern and distinct outline. The cell surfaces are raised into big regular projections giving appearance of folds (Figure 4. h). C. nummularioides: The petal surface exhibits rugose-ruminate pattern. The epidermal cells are dis- tinct and loosely packed with thin walls. The elevated radial walls also show smooth pattern. The central part of the cells is raised into small finger-like projection (Figure 4. i). C. kotschyi: Petal surface of this species exhibits closely packed without distinct outline. The cell surface is raised into broad finger-like projections and more or less thick folds, forming tubercles in the mid of the cell. The surface as a whole shows rugose to striate pattern which is parallel all over the surface except at the tubercle where these are intermingled with together (Figure 4. j). C. assadii: The epidermal cells of the petal surface are loosely packed with distinct outline, showing rugose tuberculate pattern. The surface between radial walls of each cell exhibit striate pattern, the central part of the cells is raised into small irregular projections giving appearance of folds. The striae are condensed and form- ing ruminate pattern on the folds (Figure 4. k). C. nummularius: Petal surface is composed of close- ly packed cells. Surface of the cell is raised into finger- like to folded projections or tubercles. The surface as a whole is rugose but at the tubercle becomes dense rumi- nate and cell boundaries are not clear (Figure 4. l). C. ovatus: Petal surface of this species shows the cell boundaries prominently, the surface exhibits finger-like projections with rugose pattern (Figure 4. m). C. esfandiarii: The petal surface of this species exhibits closely packed epidermal cells. The cell surface is raised into broad finger-like projections or tubercles, sometimes flattened into folds. The surface as a whole shows striate pattern except the top of the projections or tubercles where the striate show parallel and ruminate pattern (Figure 4. n). C. discolor: The petal surface of this species exhibit closely packed epidermal cells. The cell surface is raised into broad finger-like projections or tubercles, some- times flattened into folds. The surface as a whole shows striate pattern except the top of the projections or tuber- cles where the striate are ruminate pattern (Figure 4.o). C. persicus: Petal surface of this species is loosely packed with traceable cell boundaries. The epidermal cell appears to be polygonal with raised folds. The cen- tral part of the cells is raised into small semi-regular projection giving appearance of folds. Sometimes the ruga are condensed forming ruminate pattern on the folds (Figure 4. p). Abaxial surface of petal: On the abaxial surface of petal two basic types of ornamentation are seen: 1) The striate surface as a whole, parallel and does not show the cell boundaries prominently (This form is observed in: C. melanocarpus, C. multiflorus, C. luristan- icus, C. kotschyi, C. ovatus, C. discolor) (Figures 5. A, b, g, j, m, i). 2) The hive-shape with four to seven-sided houses in this form the boundaries between cells are clear. This form can be seen in the rest of the species studied (Fig- ures 5.c, d, e, f, h, k, l , n, p). These decorations were probably immature deco- rations. Because in some species such as C. esfandiarii, there was an intermediate of these two forms, and in species such as C. ovatus and C. discolor, in different individuals, there was one of the two forms. Infrageneric variation Both clustering and PCA analyses of the Cotoneaster species studied produced similar groupings and there- fore only PCA analyses tree characters are presented here (Figures. 6 and 7). The result based on pollen morphological: In this plot (Figures, 6), it can be seen that the two Hissarici 144 Nemat Allah Raei Niaki, Farideh Attar, Mansour Mirtadzadini, Kazem Mahdigholi Figure 4. a-o. Micromorphological micrographs of ornamentation of adaxial surface in the studied species of Cotoneaster. C. melanocarpus (a), C. multiflorus (b), C. suavis (c), C. hissaricus (d), C. morulus (e), C. tytthocarpus (f), C. luristanicus (g), C. turcomanicus (h), C. nummu- larioides (i), C. kotschyi (j), C. assadii (k), C. nummularius (l), C. ovatus (m), C. esfandiari (n), C. discolor (o), C. persicus (p). 145Pollen and floral micromorphological Studies of the genus Cotoneaster Medik. (Rosaceae) and its systematic importance Figure 5. a-o. Micromorphological micrographs of ornamentation of abaxial surface in the studied species of Cotoneaster. C. melanocarpus (a), C. multiflorus (b), C. suavis (c), C. hissaricus (d), C. morulus (e), C. tytthocarpus (f), C. luristanicus (g), C. turcomanicus (h), C. nummu- larioides (i), C. kotschyi (j), C. assadii (k), C. nummularius (l), C. ovatus (m), C. esfandiari (n), C. discolor (i), C. persicus (p). 146 Nemat Allah Raei Niaki, Farideh Attar, Mansour Mirtadzadini, Kazem Mahdigholi and Racemifl ori sections, which have the most species of this genus, are completely separated from each other and it can be said that using of pollen traits is probably eff ective in separating the sections and using these traits for placing a species in a particular section is probably helpful. Th e Cotoneaster subgenus members have con- siderable distance each other. C. suavis from the Aitch- isonioides section and C. multifl orus from Multifl ori sub- Figure 6. PCA plot of Cotoneaster species based on pollen morphological characters. Figure 7. PCA plot of Cotoneaster species based on fl oral morphological characters. 147Pollen and floral micromorphological Studies of the genus Cotoneaster Medik. (Rosaceae) and its systematic importance section are also placed far from each other, which, due to the their few representatives in Iran, this separation cannot be interpreted as a meaningfull seperation. But, as we can see, the dispersal of the species in this chart indicates that pollen traits alone are not suitable for the separation, and that some species that are macromor- phologically similar to each other, such as C. discolor and C. persicus are placed far from each other. As shown in Figure 7, the first component variance is 77.07 and the second component variance is 20.26. The CCCP and FFCP traits have a significant positive correlation with the first component and the other three traits show a negative correlation with this component. Additionally, there is a significant positive correlation between the quantitative NoCP trait with the first com- ponent, and the remaining four qualitative traits show small positive and negative correlation with the second component. Finally, the PCA analysis showed that petal traits in cotoneaster, as expected, are not separating traits, and the Hissarici and Racemiflori series species, which are the most common species in Iran, were overlapping in the terms of the separating petal traits. Subgenus Cotoneaster and some other series (Multiflori, Aitchiso- nioides), although they are separated, but because they have few representatives in Iran, it can be said that this separation is probably not meaningful and it can be relied only when more individuals of these subgenus, sections and series are studied. DISCUSSION Species delimitation and taxonomic consideration by pollen character Many researchers have proven that taxonomic charac- ters are of great interest for the correct identification of dif- ferent plant groups (Ullah et al., 2018a; Ullah et al., 2018b). The genus Cotoneaster like other tree and shrubby genera of Rosaceae such as Amygdalus L., Pyrus L., Cra- taegus L., Rosa L. is a morphologically difficult genus. Occurrence of hybridization which is a result of specific structure of flower, leads to appearance of individuals with intermediate characters. According to some stud- ies on the family (Hebda and Chinnappa 1990) and also some genera such as Amygdalus (Vafadar et al. 2010), Pyrus (Xu and Yao 1990, Zamani et al. 2010), Rubus (Wronska- Pilarek et al. 2006), striate sculpturing is the predominant ornamentation in the family. An important feature in Cotoneaster different from other genera such as Pyrus, Rubus, Amygdalus and Rosa (Xu and Yao 1990; Wronska-Pilarek et al. 2006; Vafadar et al. 2010; Fatemi et al. 2012) is the presence of both tri- and tetracolporate pollen in the same specimen which is generally related to different levels of ploidy (Borsch and Wilde 2000). On the basis of a comprehensive study on pollen morphology of the family Rosaceae in Canada (Hebda and Chinnappa 1990) it has been stated that variation in sculpturing is a diagnostic tool by which taxa can be identified, usually at the generic and often at the specific level. According to sculpturing, two main types striate (ridges and valleys) and non-striate (mainly psilate and verrucate) were recognized in the family (Hebda and Chinnappa, 1990). Moore (1991) has emphasized that pollen morphology in taxa of Rosaceae is very variable, even among the populations of the same species. Also, the grain size is the least reliable feature that is related to the comparatively frequent occurrence of hybridization in this family. This problem is remarkable in this study in the case of shape and sculpturing, even in different specimens of the same species. The importance of pollen morphological characters and their fitness for the most actual subgeneric taxonomic grouping are discussed in the following. Subgen. Cotoneaster. ser. Cotoneaster In this research, two species of subgenus Cotoneas- ter were studied. As shown in Table 2 and 3, the pollen characters in these two species are very similar to each other. This confirms the previous results that said that subgenus Cotoneaster is monophyletic (Li et al. 2014). In addition, because the pollination of this subgenus is highly dependent on a particular group of bees, the similarity of pollen grains in this subgenus can be evo- lutionary. Subgen. Chaenopetalum. ser. Hissarici Similar ornamentation pattern in C. nummularioides and C. kotschyi and dissimilarity from others are in line with other similarities between these species (including subcoriaceous, small (15 × 13 mm) and ovate or broadly elliptic leaves, compact inflorescence and number of flow- er (2 -5) per inflorescence, black and small fruits, navel open and also the same geographical distribution in Iran). C. hissaricus and C. tytthocarpus are very similar to each other in having similar morphological (size and shape of leaves, color and size of fruit, villose and depressed calyx, open navel) and pollen characters (P/E, type of sculptur- ing, shape of pollen) which distinguish these taxa from 148 Nemat Allah Raei Niaki, Farideh Attar, Mansour Mirtadzadini, Kazem Mahdigholi other species of the series. C. hissaricus comes from Tajik- istan and Afghanistan and C. tytthocarpus occurring in Tajikistan. Also, both of species are tetraploid (Fryer & Hylmo 2009). Morphological, micromorphological char- acters of pollen, origin center and chromosome num- ber suggest they could be regarded as related species, although their area of distribution in Iran is not the same (first species distribution is in NW Iran and latter is in NE Iran). The pollen morphology in C. morulus is very heterogenus, the exine having variable sculpturing type, and does not provide much useful information for the interspecific delimitation within the series Hissarici. Subgen. Chaenopetalum. ser. Racemiflori Widely distributed C. ovatus (with ovate leaves and red fruit) and C. assadii. (with obovate leaves and red – orange fruit) are considered closely related by the similar ornamentation, polar axis length, pore number in area unit, inter ridge width and apocolpium length. With respect to their similar macromorphological char- acters (size of leaves and fruits, number of flowers in per infloresens, habit of plant) and distribution area, the overlapping pollen morphologies of C. ovatus and C. assadii provide support for the same origin of these species. According to Khatamsaz (1993), C. esfandiarii is placed in the Cotoneaster subgenus (by erect petals and 2-3 style) but Fryer and Hylmo placed it in the Chaeno- petalum subgenus, Racemiflori section (by spread petals and 2- 3 style). In this survey, based on the exine sculp- turing, C. esfandiarii resembles members of subgenus Chaenopetalum more to species of subgenus Cotoneaster, but the judgment in this case requires further studies. C. persicus and C. discolor are much alike in their pollen ornamentation. C. discolor pollen (with obscure ridge) differs from that of C. persicus (psilate sculpture) usually by having of a few number perforations in area unit. Many of these species are relatively specific in their habitat requirements on the dry slopes (e.g. C. persicus, C. prunoisus, C. kotschyi) or wet regions (C. assadii) and may prove to be important habitat indicators. Also, the presence of pollen grains of these species in the depths of a region can be partly informed of the climate of that area in a particular geological period. In conclusion, our findings revealed the palynologi- cal characteristics (e.g., perforation number, size and exine sculpturing) of the genus Cotoneaster. The similar- ity of exine structure and ornamentation, as well as the similarity of the various parameters analyzed at inter- specific level makes it hard to establish taxonomical boundaries and clearly shows the affinity of species as far as morphological characteristics are concerned. Species delimitation and taxonomic consideration by micromorphological petal Taxonomic perspective Shaheen et al. (2016) analyzed the shape of petal epi- dermal cells and their wall patterns within Rosaceae and concluded that family had a high degree of petal micro- morphological variation, but we found only little dif- ferences among Cotoneaster species. Our result showed that there was not a significant variation at interspecific level in the 16 studied species. Unlike other genera of the Rosaceae, e.g. Rubus, Crataegus (Christensen 1992; Christensen and Hansen 1998; Sharifnia & Behzadi Shakib 2012; Hamzeh’ee et al. 2014), Sibbaldia (Tahir and Rajupt 2010) and Rosa (Sharma et al. 2005), Coto- neaster species petals decorating the microscopic level, did not show significant variation (exception number of conical projections). The petal epidermal features among species were fairly similar to each other. The shape of petal cells in the all of species was conical to finger- shape projection on the adaxial side. Conical cells may increase petal brightness and therefore increase pollina- tor visitation rates (Glover and Martin 1998; Comba et al. 2000; Dyer et al. 2006; Ojeda et al. 2009). The micro- morphological properties of petal surfaces showed some variations. Number of conical projections is an impor- tant diagnostic character. The abaxial epidermis surface of these petals had a uniform pattern and cells with dif- ferent sizes joined together in a fixed pattern. Asexual seed production or apomixis, which is often associated with hybridization and polyploidy (Marshall & Brown 1981; Nogler 1984), has been reported in five Maloid genera e.g. Amelanchier Medik. (Campbell et al. 1985) and Cotoneaster (Hjelmqvist 1962). Such plants will therefore produce some com- pletely maternal progeny through apomixes (Stebbins 1950). Consequently, apomixis genes can be much older than the clones they are currently contained in (Van Dijk 2003). Apomixis also has been reported frequent- ly in Cotoneaster (Rothleutner et al. 2016). Since some of the maternal traits can be preserved for a long time through apomicies (Stebbins 1950), one of the reasons for not changing the ornamentation of the adaxial sur- face of the petals and their similarity to each other in different species is apomixis. The interesting thing is that micromorphological traits of petal in two species C. hissaricus and C. morulus, very similar to each oth- er. These two species have similar macromorphological characters (shape and size of leaf, size and color of flow- er, shape and color of fruit) and regional distribution in Iran (Azerbaijan province). 149Pollen and floral micromorphological Studies of the genus Cotoneaster Medik. (Rosaceae) and its systematic importance Evolutionary perspective Pollination is done by bees in Cotoneaster mainly by the short-tongued bumble bees (Bombus terrestris and bombus lucorum) and honey bees (Apis mellifera) which visited species in both subgenera of the genus, concen- trating on the subgenus Cotoneaster during early sum- mer and on Chaenopetalum after mid-June. The sec- tion Cotoneaster is recommended as particularly valu- able for bee forage. Plants of the subgenus Cotoneaster were visited more by these bees in May and early June, a critical period when other forage may be scarce. The common carder bee (Bombus pascuorum) and the early bumble bee (Bombus pratorum) almost exclusively vis- ited plants in the subgenus Cotoneaster throughout the season (Corbet et al. 1992). These findings and simi- lar studies suggest that pollination of Cotoneaster and bee nutrition strongly linked together (Toth et al. 2011). Also, the Cotoneaster petals are white (especially in the section Chaenopetalum), for this reason the petal cell ornamentation on the adaxial surface is very important in attracting bees. The periclinal wall pattern of petal cells in all species studied is conical. Different species of Cotoneaster have the same pollinators and therefore there is not much difference between the adaxial sur- face ornamentation of the petals. Previous research has shown that flowers and their pollinators in many plants evolve together and has suggested that the rise of bees coincided with the largest flowering plant clade, the eud- icots (Cappellari et al. 2013). Probably, pollination by certain species of bees is the only way to reproduce sex- ually in Cotoneaster and for this reason, the various spe- cies of this genus have evolved with each other in terms of petals, along with the particular species of this bee. Ecological perspective It seems that petal traits are stable in different spe- cies of this genus and do not change under the influ- ence of the ecological conditions. Because all species that have been collected from different climates of Iran have almost the same ornamentation in their petals. As can be seen, in species with long and open inflo- rescences that have large flowers, the number of coni- cal projections per unit area is lower and the boundary between the cells is quite distinct (C. melanocarpus, C. multiflorus, C. suavis, C. ovatus). This form of inflores- cence and flower is found in species that have large, thin, and crusty leaves. On the other hand, these leaf traits are seen in mesophytic species. Thus, the high density of papillae on the adaxial surface of the petals can be a reason to deal with the dryness of the air. So meso- phytic species do not have a high density of papillae. It is important to note that species with same traits, may also be found in semi-arid regions (C. suavis, C. ova- tus). It can be concluded that petal traits maybe have been evolved once at the time of splitting species of this genus and in subsequent periods they have not been changed under the climate differences (similar results are obtained for leaves in the species of this genus). ACKNOWLEDGEMENTS We are grateful Narges Fatemi for her assistance in acetolysing of samples, Behnam Rahmani for his kind assistance by SEM. REFERENCE Akcin EO. 2009. Micromorphology and anatomical stud- ies on petals of 11 Turkish Onosma L. (Boraginceae) Taxa. Bangl. J. Bot. 16: 157-164. Arianmanesh R. I. Mehregan M. Assadi T. Nejadsattari 2016. Comparative Morphology of the genus Tama- rix (Tamaricaceae) in Iran. ILNS. 60: 1-12. Barthlott W. 1981. Epidermal and seed surface characters of plants: Systematic applicability and some evolu- tionary aspects. Nord. J. Bot. 1: 345-355. Bartish IV, Hymlö B, Nybom H. 2001. RAPD analysis of interspecific relationships in presumably apomictic Cotoneaster species. Euphytica 120: 273-280. Borsch T, Wilde V. 2000. Pollen variability within species, populations, and individuals, with particular refer- ence to Nelumbo In: Harley M. M, Moron C. M. & Blackmore S. (ed.), Pollen and Spores: Morphology and Biology, p. 285-299. Kew: Royal Botanic Gar- dens. Campbell CS, R. C. Evans., D. R. Morgan., T. A. Dickin- son & M. P. Arsenault. 2007. Phylogeny of subtribe Pyrinae (formerly the Maloideae, Rosaceae): limited resolution of a complex evolutionary history. Plant. Syst. Evol. 266: 119-145. Campbell CS, C. W. Greene., B. F. Neubauer & J. M. Hig- gins. 1985. Apomixis in Amelanchier laevis, shadbush (Rosaceae, Maloideae). Am. J. Bot. 72: 1397-1403. Cappellari, S. C., H. Schaefer., C. C. Davis. 2013. Evolu- tion: Pollen or Pollinators — Which Came First?. Curr. Biol. 23: R316- R318. Christensen KI. 1992. Revision of Crataegus sect. Cra- taegus and nothosect. Crataeguineae (Rosaceae- Maloideae) in the Old World. Syst. Bot. Monogr. 35: 1-199. 150 Nemat Allah Raei Niaki, Farideh Attar, Mansour Mirtadzadini, Kazem Mahdigholi Christensen KI, H. V. Hansen. 1998. SEM studies of epi- dermal patterns of petals in Angiosperms. Council for Nordic Publications in Botany. 135: 1-91. Comba L, S. Corbet., H. Hunt., S. Outram., J. S. Parker & B. J. Glover. 2000. The roles of genes influencing the corolla in pollination of Antirrhinum majus. Plant. Cell. Environ. 23: 639-647. Corbet SA. & A. Westgarth-Smith. 1992. Cotoneaster for bumblebees and honey bees. J. Apic. Res. 31: 9-14. Dickore WB. & G. Kasperek. 2010. Species of Cotoneaster (Rosaceae, Maloideae) indigenous to, naturalizing or commonly cultivated in Central Europe. WILLDE- NOWIA. 40:13–45. Dyer, G.A., H. M. Whitney., S. E. J. Arnold., B. J. Glover & L. Chittka. 2006. Behavioural ecology: bees associ- ate warmth with floral colour. Nature. 442: 525. Erdtman G. 1952. Pollen Morphology and Plant Taxon- omy: Angiosperms, An introduction to Palynology. Stockholm: Almqvist & Wiksell/Waltham, p.553. – MA: Chron. Bot. Hafner. Fatemi N, Attar F, Assareh MH, Hamzehee B. 2012. Pol- len morphology of the genus Rosa L. (Rosaceae) in Iran. Iranian. J. Bot. 14 (2): 156-164. Fryer J, B. Hylmö. 2009. Cotoneasters: A Comprehensive Guide to Shrubs for Flowers, Fruit, and Foliage. Tim- ber Press. Portland, Oregon. 344 pp. Glover, B.J. & C. Martin. 1998. The role of petal cell shape and pigmentation in pollination success in Antirrhi- num majus. Heredity. 80: 778-784. doi:10.1046/j.1365- 2540.1998.00345.x Harley MM. 1992. The potential value of pollen morphol- ogy as an additional taxonomic character in subtribe Ociminae (Ocimeae: Nepetoideae: Labiatae), pp 125- 138. – In: Harley, R.M., Reynolds, T. (Eds.), Advances in Labiatae Science, Royal Botanic Gardens. Kew, Richmond, Surrey, UK. Hebda RJ, Chinnappa CC. 1990. Studies on pollen mor- phology of Rosaceae in Canada. – Rev Palaeobot Palynol. 64: 103-108. Hsieh TH, Huang TC. 1997. Note on the Flora of Taiwan (28)- The Genus Cotoneaster Medik. (Rosaceae). Tai- wania. 42(1): 43-52. Hamzeh’ee, B., F. Attar., M. H. Assareh., A. A. Maas- soumi., SH. Kazempour Osaloo & K. I. Christensen. 2014. Taxonomic notes on Crataegus, ser. Crataegus, subser. Erianthae (Rosaceae), new species and record, using morphology and micromorphological evi- dence. Nord. J. Bot. 32: 26-37 Hjelmqvist H. 1962. The embryo sac development of some Cotoneaster species. C. Bloms publications. Hamer O., Harper D.A. & Ryan P.D. 2012. PAST: Pale- ontological Statistics software package for education and data analysis. Palaeonto. Electro. 4: 9. Kay, Q. O. N., H. S. Daoud & C. H. Stirton. 1981. Pig- ment distribution, light reflection and cell structure in petals. Bot. J. Linean. 83: 57-83. Kumar A, Panigrahi G. 1995. Cotoneaster, In Revisionary studies of the Family Rosaceae in India, Dehra Dun, India. 3: 1-292 Khatamsaz M. 1985. Studies on the genus Cotoneaster Medik. in lran, new records and key. Iran J Bot. 3: 55-62. Khatamsaz M. 1992. Rosaceae. In: Assadi et al. (eds.) Flora of Iran no. 6. Research Institute of Forests and Rangelands, Tehran. Koehne E. 1893. Deutsche Dendrologie. Stuttgart. Li, F., Q. Fan., Q. Li., S. Chen., W. Guo., D. Cui & W. Liao. 2014. Molecular phylogeny of Cotoneaster (Rosaceae) inferred from nuclear ITS and multiple chloroplast sequences. Plant. Syst. Evol. 300:1533– 1546 DOI 10.1007/s00606-014-0980-5 Lu, L.D. & A. R. Brach. 2003. Cotoneaster. In: Wu, Z. Y. and Raven, P. H. (eds.) Flora of China. Science Press, Beijing, China and Missouri Botanical Garden Press, St. Louis. Vol. 9, pp. 85-108. Mabberley DJ. 2008. Mabberley’s Plant-Book. A port- able dictionary of plants, their classification and uses. Third Edition. Cambridge University Press, Cam- bridge, U.K. 1021 pp. Marshall, D. R., A. H. D. Brown. 1981. The evolution of apomixis. Heredity. 47: 1-15. Moore PD, Webb JA, Collinson ME. 1991. Pollen Analy- sis. 2nd ed. Blackwell Scientific Publications, London. Nogler GA. 1984. Gametophytic apomixis. In: Johri, B.M. (ed.), Embryology of angiosperms. Springer-Verlag, Berlin. pp. 475-518. Ojeda, I., J. Francisco-Ortega & Q. Cronk. 2009. Evolu- tion of petal epidermal micromorphology in Legumi- nosae and its use as a marker of petal identity. Ann. Bot. 104: 1099–1110, 2009 Omer, S. A., M. Tahir., M. Rajput & S. S. Tahir. 2017. Micromorphological studies on petals of Spiraea L. species (Rosaceae) From Pakistan. Pak. J. Bot. 49 (1): 283-287. Perveen A, Qaiser M. 2014. Pollen flora of Pakistan–XXI. Rosaceae. - Pak. J. Bot. 46(3): 1027-1037. Podani J. 2000. Introduction to the Exploration of Multi- variate Data [English translation]. Leide, Netherlands: Backhuyes Riedl H. 1969. Cotoneaster. In: Rechinger K. H. (ed.) Flo- ra Iranica. Graz, Austria: Akademische Druck- und Verlagsanstalt. Vol. 66 Radford AE, Dickison WC, Massey JR, Bell CR. 1974. - Vascular Plant Systematics. Harper & Row Publish- ers, New York. 151Pollen and floral micromorphological Studies of the genus Cotoneaster Medik. (Rosaceae) and its systematic importance Rothleutner, J. J., M. W. Friddle & R. N. Contreras. 2016. Ploidy levels, relative genome sizes, and base pair composition in Cotoneaster. J. Am. Soc. Hortic. Sci. 141: 457–466. Shaheen, Sh., S. Younas., N. Harun., M. Jalal., S. Ramzan., M. Jaffer & S. S. Javad Sarwar. 2016. Petal anatomy: can it be a taxonomic tool? JAEBS. 6 (4S) 218-226. Sharifnia, F. & S. Behzadi Shakib. 2012. Epidermal petal patterns of 13 Iranian Rubus L. (Rosaceae) species. Ann. Biol. Res. 3: 2734-2740. Sharma G, M. L. Sharma., S. K. Sharma & H. P. S. Kang. 2005. Single-step surface replication of flower petals for SEM. Microsc. Anal. 19: 21-23. Stebbins GL. 1950. Variation and evolution in plants. New York, 643 p. Sulborska, A., E. Weryszko-Chmielewska & M. Chwil. 2012. Micromorphology of Rosa rugosa Thunb. petal epidermis secreting fragrant substances. Acta Agro- botanica. 65: 21–28. Tahir, S.S., M. Tahir & M. Rajput. 2010. SEM studies of petal structure of corolla of the species Sibbaldia L. (Rosaceae). Pak. J. Bot. 42: 1443-1449. Toth EN, R. Filep & A. Farkas. 2011. Nectary structure of Cotoneaster roseus. Acta. Biol. Szeged. 55: 243-246. Ueda Y, Tomita H. 1989. Morphometric analysis of pol- len exine patterns in Roses. J. Japan. Soc. Hort. Sci. 58: 211-220. Ullah F, Zaman W, Papini A, Zafar M, Shah SN, Ahmad M, Saqib S, Gul S, Sohail A. 2018a. Using multiple microscopic techniques for the comparative system- atic of Spergula fallax and Spergula arvensis (Caryo- phyllaceae). Microsc Res Tech. 1–9. Ullah F, Papini A, Shah SN, Zaman W, Sohail A, Iqbal M. 2018b. Seed micromorphology and its taxonomic evidence in subfamily Alsinoideae (Caryophyllaceae). Microsc Res Tech. 1–10. Van Dijk PJ. 2003. Ecological and evolutionary oppor- tunities of apomixis: insights from Taraxacum and Chondrilla. Philosophical Transactions of the Royal Society B: Biological Sciences. 358: 1113-1121. Vafadar M, Attar F, Maroofi H, Mirtadzadini M. 2010. Pollen morphology of Amygdalus L. (Rosaceae) in Iran. Acta Soc. Bot. Pol. 79 (1): 63-71 Wronska-Pilarek D, Malinski T, Lira J. 2006. - Pollen morphology of Polish species of genus Rubus – part I. Rubus gracilis. Dendrobiology. 56: 69-77. Xu F, Yao YX. 1990. Observation on pollen morphology and exine ultrastructure of Pyrus in China. Chinese. J. Bot. 3: 33-41. Yü TT, Lu LT. 1974. Cotoneaster. In: Yü, T.T. (ed.), FRPS 36: 107-178. (in Chinese) Yu TT. 1963. Taxa nova Rosacearum sinicarum. JSE. 8: 14. Zamani A, Attar F, Maroofi H. 2010. Pollen morphology of the genus Pyrus (Rosaceae) in Iran. - Acta Biolog- ica (Szeged). 54(1):51-56. http://www.sci.u-szeged.hu/ ABS