bs_bs_banner Botanical Journal of the Linnean Society, 2013, 172, 555–571. With 3 figures Comparison of pollen grain morphological features of selected species of the genus Crataegus (Rosaceae) and their spontaneous hybrids DOROTA WROŃSKA-PILAREK1*, JAN BOCIANOWSKI2 and ANDRZEJ M. JAGODZIŃSKI3,4 1 Department of Forestry Natural Foundations, Poznań University of Life Sciences, Wojska Polskiego 71 d, 60-625 Poznań, Poland 2 Department of Mathematical and Statistical Methods, Poznań University of Life Sciences, Wojska Polskiego 28, 60-637 Poznań, Poland 3 Institute of Dendrology, Polish Academy of Sciences, Parkowa 5, 62-035 Kórnik, Poland 4 Department of Game Management and Forest Protection, Poznań University of Life Sciences, Wojska Polskiego 71 c, 60-625 Poznań, Poland Received 26 March 2012; revised 31 August 2012; accepted for publication 15 January 2013 The aim of our study was to verify, on the basis of statistical analyses of nine quantitative morphological features of pollen grains, the hypothesis that pollen grains of three parental species of Crataegus (C. laevigata, C. monogyna, C. rhipidophylla) differed from the pollen of three spontaneous hybrids of these species (C. ¥ macrocarpa, C. ¥ media, C. ¥ subsphaericea). Contrast analysis revealed that a majority of the pollen features of hybrid species were characterized by significantly higher mean values than those of parental species. Analysis of pollen shape classes indicated that the parental species clearly differed from each other in contrast with hybrids, which were characterized by a similar proportion of pollen in individual pollen shape classes. Statistical analyses showed that the pollen grains of two parental species, C. laevigata and C. monogyna, were most similar to one another. Pollen grains of typical C. rhipidophylla were similar to the pollen of hybrids and the mean values of almost all studied pollen features [P, E, Exp, Exp/P, Le, d, d/E (PAI)] of C. rhipidophylla var. rhipidophylla were intermediate between those of C. monogyna and C. rhipidophylla var. lindmanii. This corroborates Zieliński’s conjecture that C. rhipidophylla is probably an old, conserved hybrid between C. monogyna and C. calycina (= C. lindmanii = C. rhipidophylla var. lindmanii). According to the analysis of canonical variables, C. ¥ macrocarpa and C. ¥ media pollen grains were most similar. C. ¥ subsphaericea and C. rhipidophylla var. rhipidophylla and C. rhipidophylla var. lindmanii formed another pair (group), and C. monogyna and C. laevigata constituted separate ‘single species groups’. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571. ADDITIONAL KEYWORDS: interspecific variability – pollen morphology – pollen variability – statistical analysis. INTRODUCTION The genus Crataegus L. (hawthorns) comprises approximately 150–1200 species (depending on the species concept employed), distributed mainly in the Northern Hemisphere. Among these are 50–100 Old World Crataegus spp. occurring in Europe, northern Africa and western Asia (Christensen, 1992). *Corresponding author. E-mail: pilarekd@up.poznan.pl Crataegus is a well-defined genus, traditionally included in the tribe Crataegeae, subfamily Maloideae, of Rosaceae (Phipps, 1983, 1988; Robertson et al., 1991; Christensen, 1992; Christensen & Janjic, 2006). According to Campbell et al. (2007), Crataegus is included in the subtribe Pyrinae (formerly subfamily Maloideae). Christensen & Zieliński (2008) included Crataegus in the tribe Pyreae. The intrageneric classification of maloid genera has been studied by Phipps (1983), Campbell & Dickinson (1990) and Phipps et al. (1991). © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 555 556 D. WROŃSKA-PILAREK ET AL. The taxonomy of Crataegus is problematic owing to biological and historical factors (Dönmez, 2008). As in other large maloid genera, frequent interspecific hybridization, extensive polyploidy and apomixes all have the potential to blur the boundaries between species (Byatt, Ferguson & Murray, 1977; Dönmez, 2004; Talent & Dickinson, 2005; Dickinson, Lo & Talent, 2007; Ennos et al., 2012). Christensen (1992) also emphasized that Crataegus has a complicated taxonomic history, because numerous species have been described by several authors and the circumscriptions of the species have varied widely. In his studies, Christensen (1992, 1997) presented a synthesis of hawthorns in Europe and proposed a decrease in the numbers of recognized taxa and a new nomenclature. Within the boundaries of Crataegus, Phipps et al. (1990) and Phipps, O’Kennon & Lance (2003) recognized 40 series in about 15 readily distinguishable sections, although there was considerable variation in each of these groups. Christensen (1992) distinguished four series and three subseries in section Crataegus. During the course of our study, the pollen morphology of three Crataegus spp. [C. laevigata (Poir.) DC., C. monogyna Jacq., C. rhipidophylla Gand.] and three spontaneous hybrids of these species (C. ¥ macrocarpa Hegetschw., C. ¥ media Bechst., C. ¥ subsphaericea Gand.) were investigated (Table 1). According to Christensen’s (1992) classification of the genus, the species analysed belong to the large section Crataegus, series Crataegus, subseries Erianthae (Pojarkova) Christensen (C. laevigata) and subseries Crataegus (all remaining taxa; Table 1). It is evident from the palynological studies of Crataegus carried out so far that their pollen grains are very variable. Pollen grains of the described species are in radially symmetrical monads. Two pollen types are most frequent (tricolpate and tricolporate), but trisyncolporate, tetracolporate, pericolporate and inaperturate pollen types also occur. The length of the polar axis (P) varies greatly, ranging from 15 to 65 mm; similarly, the equatorial diameter (E) can range from 15 to 47 mm. Pollen grain sizes are mostly medium, rarely small or large. The pollen shape (P/E ratio) varies from suboblate, through spheroidal to prolate. Many different exine sculpture types can be found in Crataegus, including striate-perforate, regulateperforate, rugulate-striate-perforate and microreticulate (Reitsma, 1966; Byatt, 1976; Kuprianowa & Alyoshina, 1978; Eide, 1981; Fedoronchuk & Savitskii, 1985; Dickinson & Phipps, 1986; Xin, Zhang & Wang, 1986; Hebda, Chinnappa & Smith, 1988; González Romano & Candau, 1989; Hebda & Chinnappa, 1990; Moore, Webb & Collinson, 1991; Christensen, 1992; Hebda & Chinnappa, 1994; Zhou, Wie & Wu, 2000; Dönmez, 2008). Palynological studies on intraspecific hybrids have generally focused on artificial hybrids and have rarely been concerned with spontaneous hybrids found in nature. Such investigations have been based on the analysis of selected, single quantitative or qualitative features of pollen grains (usually comprising pollen size and exine sculpture) of individual interspecific hybrids. The results obtained are not consistent. Some authors have maintained that the pollen sizes of interspecific, artificial hybrids are significantly larger than those of their parents and that this is associated with ploidy (Nair, Katiar & Srivastava, 1977; Srivastava, Pal & Nair, 1977; Rizaeva, Akhmedova & Abdullaev, 1985; Hossain, Inden & Asahira, 1990; Van der Walt & Littlejohn, 1996; Franssen et al., 2001; Rhee et al., 2005), whereas others have claimed that hybrids (usually polyploids) can have pollen sizes similar or smaller than those of their parents (Olsson, 1974; Srivastava, 1978; Rao, Kumar & Nair, 1979; Ohashi, Hoshi & Iketani, 1991; Chaturvedi, Ram & Pal, 1993; Delaporte, Conran & Sedgley, 2001; Franssen et al., 2001; Jiang et al., 2001; Lu et al., 2002; Dönmez, 2008; Karlsdóttir et al., 2008). The objective of our study was to verify, based on the statistical analyses of nine quantitative features of 900 pollen grains, whether and to what degree pollen grains of the three parental species studied (C. laevigata, C. monogyna, C. rhipidophylla) differ from the pollen grains of three spontaneous hybrids (C. ¥ macrocarpa, C. ¥ media, C. ¥ subsphaericea). MATERIAL AND METHODS The taxonomic classification of Crataegus and the studied taxa was adopted from Christensen (1992), with the exception of C. ¥ kyrtostyla Fingerh., considered to be a hybrid between C. monogyna and C. rhipidophylla. As explained by Christensen & Zieliński (2008), in the past, C. monogyna ¥ rhipidophylla has often been referred to as C. ¥ kyrtostyla Fingerh. (1829) [see, for example, Franco (1968) and Christensen (1992)], but recently an original specimen of C. kyrtostyla was located at Bonn and, although the specimen is in a fairly poor condition, it is evident that Lippert’s (1978) contention that C. kyrtostyla is synonymous with C. monogyna is correct (Christensen, 1997). As a consequence, C. subsphaericea Gand. (1871) is now the oldest legitimate name for the hybrid C. monogyna ¥ rhipidophylla (Christensen, 1997). Therefore, following Christensen & Zieliński (2008), we adopted the name C. ¥ subsphaericea for this taxon. The palynological terminology follows Punt et al. (2007) and Hesse et al. (2009). The study was conducted on three parental species of Crataegus and © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 No. Series Subseries Species/taxon Locality Position Collector, herbarium 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Erianthae Erianthae Erianthae Erianthae Erianthae Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus laevigata (Poir.) DC. Crataegus laevigata (Poir.) DC. Crataegus laevigata (Poir.) DC. Crataegus laevigata (Poir.) DC. Crataegus laevigata (Poir.) DC. Crataegus monogyna Jacq. Crataegus monogyna Jacq. Crataegus monogyna Jacq. Crataegus monogyna Jacq. Crataegus monogyna Jacq. Crataegus rhipidophylla Gand. Crataegus rhipidophylla Gand. Crataegus rhipidophylla Gand. Crataegus rhipidophylla Gand. Crataegus rhipidophylla Gand. Crataegus ¥ macrocarpa Hegetschw. Crataegus ¥ macrocarpa Hegetschw. Crataegus ¥ macrocarpa Hegetschw. 51°55′6.239″N, 17°2′7.548″E 50°17′49.344″N, 16°39′7.848″E 51°18′26.424″N, 17°3′31.14″E 51°28′18.911″N, 16°54′30.383″E 50°24′2.304″N, 18°3′49.463″E 49°28′51.744″N, 22°42′3.671″E 52°14′53.951″N, 17°5′17.951″E 53°9′44.136″N, 22°48′13.355″E 49°32′48.983″N, 21°50′58.092″E 52°52′34.103″N, 18°41′36.707″E 49°28′52.968″N, 20°1′53.508″E 49°44′30.551″N, 19°35′43.403″E 53°18′0.36″N, 17°26′51.144″E 50°17′45.312″N, 16°52′25.86″E 50°17′49.344″N, 16°39′7.848″E 52°16′53.328″N, 17°1′34.067″E 50°26′47.471″N, 16°15′54.431″E 53°29′26.303″N, 16°19′7.536″E Kaczmarek, KOR Kosiński, KOR Gostyńska, KOR Gostyńska, KOR Browicz, KOR Jakuszewski, KOR Gostyńska, KOR Niedziałek, KOR Gostyńska, KOR Gostyńska, KOR Browicz, Zieliński, KOR Browicz, Zieliński, KOR Boratyński, KOR Kosiński, KOR Kosiński, KOR Gostyńska, KOR Browicz, Gostyńska, KOR Gostyńska, KOR 19 20 21 22 23 24 25 26 27 28 29 30 Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus Crataegus ¥ macrocarpa Hegetschw. Crataegus ¥ macrocarpa Hegetschw. Crataegus ¥ media Bechst. Crataegus ¥ media Bechst. Crataegus ¥ media Bechst. Crataegus ¥ media Bechst. Crataegus ¥ media Bechst. Crataegus ¥ subsphaericea Gand. Crataegus ¥ subsphaericea Gand. Crataegus ¥ subsphaericea Gand. Crataegus ¥ subsphaericea Gand. Crataegus ¥ subsphaericea Gand. Ostrowo, Wielkopolskie Province Bystrzyca Kłodzka, Dolnośla˛skie Province Kocie Góry, Dolnośla˛skie Province Żmigród, Dolnośla˛skie Province Mechnica, Opolskie Province Krościenko, Podkarpackie Province Kórnik, Wielkopolskie Province Leśniki, Podlaskie Province Rymanowa Zdrój, Podkarpackie Province Aleksandrów, Kujawsko-Pomorskie Province Nowy Targ, Małopolskie Province Sucha Beskidzka, Małopolskie Province Czarmuń, Kujawsko-Pomorskie Province Stronie Śla˛skie, Dolnośla˛skie Province Bystrzyca Kłodzka, Dolnośla˛skie Province Borówiec, Wielkopolskie Province Kudowa, Dolnośla˛skie Province Kamienna Góra, Zachodniopomorskie Province Dubiecko, Podkarpackie Province Życzanów, Małopolskie Province Pre˛gowo, Pomorskie Province Kulin, Kujawsko-Pomorskie Province Ciechocinek, Kujawsko-Pomorskie Province Milicz, Dolnośla˛skie Province Chapsko, Kujawsko-Pomorskie Province Sobieszyn, Lubelskie Province Grzegorzowice, Świe˛tokrzyskie Province Smogorzów, Opolskie Province Kulin, Kujawsko-Pomorskie Province Wrotkowo, Warmińsko-Mazurskie Province 49°49′32.808″N, 22°23′24.467″E 49°30′2.988″N, 20°40′37.991″E 54°15′17.855″N, 18°28′46.019″E 52°40′12.215″N, 19°7′35.471″E 52°52′51.671″N, 18°47′38.292″E 51°31′51.815″N, 17°16′32.412″E 52°37′37.559″N, 17°56′4.848″E 51°35′36.815″N, 22°9′44.424″E 50°52′19.991″N, 21°9′6.011″E 51°9′23.615″N, 17°45′33.911″E 52°40′12.215″N, 19°7′35.471″E 54°14′43.223″N, 22°20′49.379″E Browicz, Gostyńska, KOR Gostyńska, KOR Gostyńska, KOR Gostyńska, KOR Bojarczuk, KOR Gostyńska, KOR Gostyńska, KOR Boratyński, KOR Gostyńska, KOR Gostyńska, KOR Gostyńska, KOR Naklicki, KOR POLLEN FEATURES OF SELECTED CRATAEGUS TAXA © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 Table 1. List of localities of Crataegus taxa studied. KOR – Herbarium of the Institute of Dendrology, Polish Academy of Sciences, Kórnik, Poland 557 558 D. WROŃSKA-PILAREK ET AL. three spontaneous hybrids of these species. The list of the taxa analysed is shown in Table 1. Pollen samples were collected in the Herbarium of the Institute of Dendrology of the Polish Academy of Sciences in Kórnik (52°14′12″N, 17°05′55″E) – KOR (Poland). Several randomly selected flowers were collected from each individual shrub. The plant material was collected by specialists, experienced dendrologists from the Institute of Dendrology of the Polish Academy of Sciences in Kórnik. We have selected herbarium specimens representing different taxa, and the specimens were re-reviewed by Professor Jerzy Zieliński, a leading dendrologist specializing in Rosaceae, including hawthorns (Zieliński, 1977, 1982; Christensen & Zieliński, 2008). Each shrub was represented by 30 pollen grains. Plant material was derived from 30 natural sites located in Poland, so that each taxon had five localities (Table 1). All samples were acetolysed according to the method described by Wrońska-Pilarek (1998, 2011) and Wrońska-Pilarek & Jagodziński (2011). The acetolysing mixture was made up of nine parts of acetic acid anhydride and one part of concentrated sulphuric acid, and the process of acetolysis lasted 2.5 min. Observations were carried out with both a light microscope (Biolar 2308, Nikon HFX-DX) and a scanning electron microscope (SEM) (Hitachi S-3000N). A pollen sample consisted of 30 pollen grains from each shrub. In total, 900 pollen grains were measured. Only mature, correctly formed pollen grains were studied. However, malformed pollen grains were also noticed in the samples, and their percentage was determined by examining 400 pollen grains (four randomly selected samples of 100 pollen grains) for each taxon. Nine quantitative features were analysed, i.e. the length of the polar axis (P), equatorial diameter (E), length of the ectocolpi (Le), thickness of the exine along the polar axis (Exp), distance between the apices of two colpi (d) and four ratios: P/E, Le/P, Exp/P and d/E. Two qualitative features were also analysed: pollen outline and shape. The normality of P, E, P/E, Exp, Exp/P, Le, Le/P, d and d/E data distributions was tested using Shapiro– Wilk’s normality test (Shapiro & Wilk, 1965). A oneway analysis of variance (ANOVA) was carried out to determine the effects of species on P, E, P/E, Exp, Exp/P, Le, Le/P, d and d/E. The minimal and maximal values of the characteristics and the arithmetic means and coefficients of variation were calculated. When critical differences were noted, multiple comparisons were carried out using least-significant difference (LSD) for each trait; based on this, homogeneous groups (not significantly different from each other) were determined for the traits analysed. The relationships between P, E, P/E, Exp, Exp/P, Le, Le/P, d and d/E were estimated using Pearson’s correlation coefficients (Sokal & Rohlf, 1995). The Mahalanobis distance (Mahalanobis, 1936) between the species, determined using P, E, P/E, Exp, Exp/P, Le, Le/P, d and d/E, can be treated as the phenotypic distance between the species. Graphical distribution of the species and hybrids on a plane, including all the features jointly, was possible through the application of canonical variable analysis. All data analysis was performed using the statistical package GenStat v. 10.1 (GenStat, 2007). RESULTS GENERAL POLLEN MORPHOLOGICAL DESCRIPTION A description of the pollen grain morphology of Crataegus spp. studied is given below and illustrated with SEM photographs (Figs 1–15). Pollen grains occur as monads. The pollen grains of Crataegus spp. were tri-zonocolporate (Figs 1–8). All analysed pollen grains were of medium size (25.1–50 mm; Table 2) according to Erdtman’s (1952) pollen size classification. The average length of the polar axis (P) for the three parental Crataegus spp. was 36.50 mm (range, 28–46 mm) and for the three hybrids 39.07 mm (range, 30–50 mm). Parental species usually had smaller pollen than hybrids. The exception was C. rhipidophylla, the pollen grains of which were relatively large and equalled some of the hybrid pollen grains. Among parental species, the smallest P features were found for the pollen of C. laevigata (28 mm) and the largest for C. rhipidophylla (46 mm). However, in hybrids, the smallest P was found in C. ¥ macrocarpa (30 mm) and the largest in C. ¥ subsphaericea (50 mm; Table 2). The mean length of the equatorial diameter (E) for the parental Crataegus spp. was 33.20 mm (range, 24–46 mm) and for the three hybrids 36.04 mm (range, 26–46 mm). In parental species, the shortest equatorial diameter (E) occurred in the pollen of C. monogyna (24.0 mm), and the longest was in C. rhipidophylla (46 mm). In hybrids, the shortest equatorial diameter (E) was found in C. ¥ macrocarpa (26 mm) and the longest in C. ¥ subsphaericea (46 mm; Table 2). In all taxa studied, the outline in polar view was mostly elliptic or circular, more rarely triangular with obtuse apices, whereas, in equatorial view, the outline was mostly elliptic, rarely circular (Figs 1–5). The mean P/E ratio for parental Crataegus spp. was 1.11, ranging from 0.81 in C. rhipidophylla to 1.67 in C. monogyna, and for hybrids was 1.09, ranging from © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 POLLEN FEATURES OF SELECTED CRATAEGUS TAXA Figures 1–15. See caption on next page. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 559 560 D. WROŃSKA-PILAREK ET AL. Figures 1–15. Polar and equatorial views, apertures and exine sculpture in scanning electron microscopy. Figures 1–3. Crataegus laevigata, C. rhipidophylla, C. monogyna: pollen grains in equatorial views, colpi and exine sculpture. Figures 4, 5. Crataegus rhipidophylla, C. ¥ media: pollen in polar views, three colpi and exine sculpture. Figure 6. Crataegus laevigata: two pollen grains in equatorial and polar views, colpi and exine sculpture. Figure 7. Crataegus ¥ macrocarpa: long colpus and exine sculpture. Figure 8. Crataegus ¥ macrocarpa: colporus and exine sculpture. Figure 9. Crataegus ¥ media: seven pollen grains in polar and equatorial views, colpori and exine sculpture. Figures 10–15. Crataegus laevigata, C. monogyna, C. rhipidophylla, C. ¥ macrocarpa, C. ¥ media, C. ¥ subsphaericea: striate exine sculpture with striae, grooves and perforations. 0.86 in C. ¥ subsphaericea to 1.77 in C. ¥ macrocarpa (Table 2). Pollen grains of the parental Crataegus spp. differed distinctly with respect to their pollen shape (Table 3). Crataegus monogyna had the largest number of elongated pollen grains from such shape classes as prolate-spheroidal (50.7%) to subprolate (33.3%). The proportion of the remaining classes was low, ranging from 2.0 to 7.3%, and oblate pollen grains were not found in this species. Pollen grains of C. laevigata were primarily spheroidal (49.3%) and oblate-spheroidal (20.7%). In addition, this species was distinguished by having a relatively high proportion of oblate pollen (12.7%), a moderate frequency of prolate-spheroidal pollen (16%) and a low proportion of pollen grains from the subprolate class (1.3%). No prolate pollen grains occurred in C. laevigata. Crataegus rhipidophylla was characterized by the most uniform distribution of pollen in individual shape classes. Its pollen grains belonged mainly to the following three shape classes: spheroidal (30%), prolate-spheroidal (26%) and subprolate (24.7%). The remaining classes were less frequent (1.3–10.7%; Table 3). Pollen grains of the three hybrids, in contrast with parental species, were characterized by similar shapes. In all taxa, prolate-spheroidal pollen grains were most numerous (44.7–48.7%) and grains from the spheroidal and subprolate classes were also frequent, ranging from 20.7 to 28% and 19.3 to 26%, respectively. Oblate-spheroidal and prolate pollen grains occurred rarely (5.3–6.7%) and very rarely (1.3–2%), respectively. No oblate pollen grains were found in any of the analysed species (Table 3). In hybrids, prolate-spheroidal pollen grains were most numerous (46.4%), pollen grains from subprolate (22.9%) and spheroidal classes were frequent and similar, whereas oblate-spheroidal (5.8%) and prolate (1.6%) pollen grains were the least numerous. No oblate pollen grains were found (Table 3). The mean exine thicknesses were 0.69 mm (0.2– 2.0 mm) (parental species) and 0.82 mm (0.4–2.0 mm) (hybrids; Table 2). The exine was thinnest in C. monogyna (0.2 mm) and thickest in C. laevigata and C. rhipidophylla (2.0 mm). In hybrids, the exine thickness was slightly less variable, ranging from 0.4 mm in C. ¥ media and C. ¥ subsphaericea to 2.0 mm in C. ¥ macrocarpa. The relative thickness of the exine (Exp/P ratio) for parental species and hybrids was similar, and averaged 0.019 (0.009–0.040) and 0.021 (0.008–0.063; Table 2), respectively. In all taxa studied, the exine sculpture was striate with or without perforations. Striae were long and more or less parallel to the colpus (Figs 1, 7, 12) or short or medium length and interwoven, forming bends (Figs 2, 6, 10). Circular or elliptic perforations with different, usually small, diameters were found at the bottom of the grooves (Figs 9, 11, 15). Pollen grains usually possessed three apertures (colpori; Figs 4, 5). Colpi were arranged meridionally, regularly, more or less evenly spaced, fusiform in outline and were long: mean lengths of 32.45 mm (24.00–46.00 mm) (parental species) and 35.82 mm (24.00–48.00 mm) (hybrids) (Figs 1–3; Table 2). On average, the length of the colpi constituted 89% of the polar axis length in parental species and 92% in hybrids. On average, species had slightly shorter colpi than hybrids. Their width was variable and usually greatest in the equatorial region. The polar area index (PAI) averaged 0.20 (0.05– 0.40) in parental species and 0.15 (0.05–0.35) in hybrids. The lowest mean values of this index in parental species were recorded in C. rhipidophylla (0.16) and the highest in C. monogyna (0.24); in hybrids, mean values of PAI were less variable and smaller, ranging from 0.15 in C. ¥ macrocarpa to 0.16 in C. ¥ subsphaericea (Table 2). The percentage of deformed pollen grains in the samples (400 grains per taxon) was variable and ranged from 20 to 75% (Fig. 16). The results are surprising because, on average, the share of deformed pollen grains from parental species was higher (48%) than in the hybrids (30%). The highest frequency of deformed pollen was found in samples of C. rhipidophylla (75%) and the lowest in C. ¥ macrocarpa (20%). In parental species, the lowest percentage of deformed pollen grains occurred in C. monogyna (25%), an intermediate percentage in C. laevigata (45%) and the highest percentage in C. rhipidophylla (75%). In hybrids, the percentages of deformed pollen grains were 20% in C. ¥ macrocarpa, 25% in C. ¥ subsphaericea and 45% in C. ¥ media. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 POLLEN FEATURES OF SELECTED CRATAEGUS TAXA Figures 1–15. Continued © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 561 562 D. WROŃSKA-PILAREK ET AL. Figures 1–15. Continued © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 P (mm) Taxon C. laevigata C. monogyna C. rhipidophylla C. ¥ macrocarpa C. ¥ media C. ¥ subsphaericea LSD0.001 ANOVA P > F Min–Max 28–40 30–44 32–46 30–46 32–46 30–50 E (mm) Mean CV (%) 33.73d 37.09c 38.67b 38.17bc 38.69b 40.35a 1.19 < 0.001 7.16 6.48 7.48 10.07 6.8 10.27 Exp (mm) Taxon C. laevigata C. monogyna C. rhipidophylla C. ¥ macrocarpa C. ¥ media C. ¥ subsphaericea LSD0.001 ANOVA P > F Min–Max 0.4–2 0.2–1.6 0.4–2 0.6–2 0.4–2 0.4–2 Min–Max 26–38 24–40 26–46 26–44 28–44 30–46 P/E Mean CV (%) Min–Max Mean CV (%) 31.35c 32.09c 36.15ab 35.55b 35.41b 37.17a 1.36 < 0.001 7.34 9.37 13.26 11.14 8.55 10.13 0.875–1.385 0.938–1.667 0.810–1.462 0.9–1.769 0.889–1.5 0.857–1.412 1.081b 1.166a 1.086b 1.080b 1.099b 1.089b 0.044 < 0.001 9.07 11.46 13.33 10.03 9.54 8.8 Mean CV (%) 30.43d 31.95c 34.96b 34.85b 35.84ab 36.77a 1.42 < 0.001 8.31 10.68 10.6 12.28 10.33 11.8 Exp/P Le (mm) Mean CV (%) Min–Max Mean CV (%) 0.716c 0.492d 0.857ab 0.888a 0.833abc 0.739bc 0.137 < 0.001 53.69 51.3 47.77 40.46 45.22 47.04 0.01–0.059 0.005–0.04 0.009–0.059 0.013–0.063 0.009–0.063 0.008–0.044 0.021ab 0.013c 0.022ab 0.024a 0.022ab 0.018b 0.004 < 0.001 55.02 51.53 46.8 42.78 47.67 47.45 Le/P d (mm) Min–Max Mean C. laevigata C. monogyna C. rhipidophylla C. ¥ macrocarpa C. ¥ media C. ¥ subsphaericea LSD0.001 ANOVA P > F 0.8125–1 0.6–1 0.75–1 0.6842–1 0.75–1 0.56–1 0.902a 0.862b 0.904a 0.912a 0.925a 0.912a 0.023 < 0.001 CV (%) 4.48 9.39 6.36 5.77 6.48 6.05 Min–Max 2.0–12 4.0–12 2.0–12 2.0–12 2.0–10 2.0–10 24–38 24–40 28–46 26–44 24–44 28–48 d/E (PAI) Mean CV (%) Min–Max Mean CV (%) 6.367b 7.647a 5.707bc 5.280c 5.287c 5.787c 0.725 < 0.001 29.81 23.87 34.9 39.51 33.33 31.69 0.063–0.353 0.105–0.4 0.045–0.333 0.05–0.3529 0.05–0.2632 0.05–0.2941 0.203b 0.239a 0.161c 0.148c 0.150c 0.157c 0.021 < 0.001 28.21 23.95 37.77 37.16 33.49 32.77 563 Taxon Min–Max POLLEN FEATURES OF SELECTED CRATAEGUS TAXA © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 Table 2. Range (minimum–maximum, Min–Max), mean values and coefficient of variation (CV) of pollen features studied. One-way analyses of variance (ANOVAs) were performed separately for each of the traits. Same letters indicate a lack of statistically significant difference between the analysed taxa according to the least-significant difference (LSD) test (P < 0.001) 564 D. WROŃSKA-PILAREK ET AL. Table 3. Percentage of pollen grains distributed among shape classes (P/E ratio) according to Erdtman’s (1952) classification: oblate-spheroidal (0.89–0.99); spheroidal (1.00); prolate-spheroidal (1.01–1.14); subprolate (1.15–1.33); prolate (1.34–2.00); perprolate (> 2.01) Pollen shape class Taxon Oblate Oblate-spheroidal Spheroidal Prolate-spheroidal Subprolate Prolate C. laevigata C. monogyna C. rhipidophylla C. ¥ macrocarpa C. ¥ media C. ¥ subsphaericea 12.7 20.7 2.0 10.7 6.7 5.3 5.3 49.3 6.7 30.0 28.0 20.7 21.3 16.0 50.7 26.0 44.7 46.0 48.7 1.3 33.3 24.7 19.3 26.0 23.3 7.3 7.3 1.3 2.0 1.3 1.3 C. xsubsphaericea C. xmedia C. xmacrocarpa C. rhipidophylla C. laevigata C. monogyna 0 10 20 30 40 50 60 70 80 Deformed pollen grains (%) Figure 16. Percentage of deformed pollen grains. INTERSPECIFIC VARIABILITY Statistically significant differences were found among the taxa examined for all nine pollen grain features analysed (P < 0.001; Table 2). The results of ANOVAs are summarized in Table 2. Most of the observed quantitative features were significantly correlated (Table 4). However, feature d was not correlated with P, E, P/E, Exp or Exp/P, Exp was not correlated with P, and Le/P was not correlated with Exp or Exp/P. Fourteen of 28 significantly correlated pairs of features were characterized by positive correlation coefficients (Table 4). The analysis of phenotypic distance (Mahalanobis distance) between the species revealed that pollen grains of C. laevigata and C. monogyna were most similar with respect to all the features examined treated jointly (Mahalanobis distance 1.68), whereas they differed significantly from the pollen grains of C. rhipidophylla and from those of the hybrids (Mahalanobis distances 2.07 and 2.04, respectively) (Table 5). The three hybrids differed slightly from one another and from C. rhipidophylla, and differed significantly from the pollen of C. laevigata and C. monogyna (Table 4). The diagram of the first two canonical variables (Fig. 17) was used to divide the taxa examined into four groups. The first group comprised C. ¥ macrocarpa and C. ¥ media. The second included C. ¥ subsphaericea and C. rhipidophylla. The other two groups included just one species, C. monogyna or C. laevigata (Fig. 17). The first two canonical variables accounted for 91.92% of the total multivariate variability among species and hybrids. In the contrast analysis between parental species and their hybrids (Table 6), for the majority of the features analysed, hybrids were characterized by a significantly higher mean of the given feature in comparison with the mean of the parental form (i.e. negative contrast value). In the case of P, E and Le features, all hybrids were characterized by a significantly higher mean in comparison with the parental forms (Table 6). For Exp, Exp/P and Le/P, two of the three hybrids exhibited a significantly higher mean of these features in comparison with the parental forms. In the case of Exp and Exp/P, these included C. ¥ macrocarpa (C. laevigata ¥ C. rhipidophylla) and C. ¥ media (C. monogyna ¥ C. laevigata) and, for Le/P, C. ¥ subsphaericea (C. monogyna ¥ C. rhipidophylla) and C. ¥ media. Considerably fewer hybrid/ feature combinations were found to have a lower mean of the given feature in comparison with the mean of its parental forms. This happened in all hybrids in the case of d and d/E and in C. ¥ subsphaericea and C. ¥ media for the P/E feature. Hybrids did not differ significantly from the mean of the parental species in only the following four cases: in C. ¥ macrocarpa for P/E and Le/P features and in C. ¥ subsphaericea for Exp and Exp/P features (Table 6). © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 POLLEN FEATURES OF SELECTED CRATAEGUS TAXA 565 Table 4. Correlation matrix for the observed features Feature P E P/E Exp Exp/P Le Le/P d d/E P E P/E Exp Exp/P Le Le/P d d/E 1 0.575*** 0.284*** 0.054 -0.146*** 0.833*** 0.090** -0.020 -0.191*** 1 -0.609*** 0.257*** 0.139*** 0.534*** 0.144*** -0.029 -0.343*** 1 -0.232*** -0.287*** 0.170*** -0.089** 0.008 0.216*** 1 0.975*** 0.079* 0.062 -0.059 -0.138*** 1 -0.091** 0.038 -0.05 -0.094** 1 0.624*** -0.165*** -0.314*** 1 -0.277*** -0.304*** 1 0.941*** 1 *P < 0.05, **P < 0.01, ***P < 0.001. Table 5. Phenotypic distance between the taxa studied, calculated as the Mahalanobis distance based on P, E, P/E, Exp, Exp/P, Le, Le/P, d and d/E Taxon C. laevigata C. monogyna C. rhipidophylla C. ¥ macrocarpa C. ¥ media C. ¥ subsphaericea C. laevigata C. monogyna C. rhipidophylla C. ¥ macrocarpa C. ¥ media C. ¥ subsphaericea 0 1.68 2.07* 1.91 1.96 2.38* 0 2.04* 2.29* 2.26* 2.24* 0 0.77 0.85 0.96 0 0.36 0.92 0 0.81 0 *P < 0.05. DISCUSSION Statistical analyses of the morphological variability of the pollen grains of the taxa of Rosaceae have rarely been undertaken. Extensive research on the range of pollen grain variation of several species of Rosa L., at the inter- and intraspecific levels, and on flower variation, at the individual and population levels, were conducted by Wrońska-Pilarek & Jagodziński (2009, 2012). There are a few other published studies in which coefficients of variation of selected pollen grain morphological traits have been analysed for the taxa of Rubus L. and Spiraea L. (Naruhashi & Takano, 1980; Polyakova & Gataulina, 2008; Wrońska-Pilarek, Jagodziński & Maliński, 2012). In the palynological literature, it is difficult to find information associated with comparative studies on species (parentals) and their spontaneous hybrids. Researchers have focused on pollen grain morphological features of selected species (parentals) and their artificial hybrids. Usually, only a few basic quantitative features of pollen grains (pollen size, P/E ratio, exine sculpture elements, less frequently the number of apertures) or qualitative features (exine sculpture type, operculum shape) have been analysed. The results of the investigations carried out so far are equivocal. Some researchers maintain that taxa (hybrids) of higher chromosome number (frequently polyploids) are habitually characterized by larger pollen size than diploid species (parentals); others are of the opposite opinion. It is no wonder that controversy exists because, as explained by Grant (1971), polyploidy is correlated with many factors (climate, latitude, elevation, type of habitat, hybridity, life form, breeding system, cell and chromosome size and structure, sex chromosome mechanism or genotype). Rizaeva et al. (1985) found that, in nine species of Gossypium L. (Malvaceae) and in various hybrids involving these species, pollen size was associated with ploidy. Tetraploid, pentaploid and hexaploid hybrids exceeded their parentals in pollen diameter (E), which, in the F1 diploid hybrids, was intermediate between that of the parental species. The greatest variation in pollen size was found in the pentaploid hybrid (Gossypium barbadense L. ‘S6037’ ¥ [G. harknessii Brandegee ¥ G. trilobum (DC.) Kearney] ¥ ‘S6037’). The same conclusions were also reached by Sørensen (1989) in his comparative studies of the pollen morphology of Pachyrhizus erosus (L.) Urb., P. tuberosus Spreng. and P. ahipa © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 566 D. WROŃSKA-PILAREK ET AL. Figure 17. Distribution of six Crataegus taxa as indicated by first two canonical variables. Table 6. Results of contrast analysis between parental species and their hybrids Contrast Feature C. ¥ media C. monogyna ¥ C. laevigata C. ¥ macrocarpa C. laevigata ¥ C. rhipidophylla C. ¥ subsphaericea C. monogyna ¥ C. rhipidophylla P E P/E Exp Exp/P Le Le/P d d/E -3.28*** -3.69*** 0.024* -0.229*** -0.00435*** -4.65*** -0.043*** 1.72*** 0.071*** -1.97*** -1.80*** 0.003 -0.101** -0.00170 -2.16*** -0.009 0.76*** 0.034*** -2.47*** -3.05*** 0.036** -0.064 -0.00067 -3.32*** -0.028*** 0.89*** 0.043*** *P < 0.05, **P < 0.01, ***P < 0.001. Parodi and their hybrids (Fabaceae). The thesis was further corroborated by Hossain et al. (1990), who maintained that pollen grains of amphidiploids, interspecific hybrids between Brassica oleracea L. var. capitata L. ‘Yoshin’ and B. rapa L. [as B. campestris L.] var. pekinensis (Lour.) Kitam. ‘Kaga’ and B. rapa var. chinensis (L.) Hanelt ‘Honsaiai’ (Brassicaceae), were significantly longer and wider than those of their diploid parentals, presumably because of the phenotypic expression of the hybrid genomes and ploidy effects. Van der Walt & Littlejohn (1996) recorded significant differences in pollen size of Protea L. (Proteaceae) between clones, interspecific hybrids and species. Rhee et al. (2005) reported that, in interspecific hybrid Lilium (L.) (Liliaceae) after in vitro chromosome doubling, in hybrid FA96-18, the pollen length was 67.5 mm in haploids and 107.3 mm in diploids. The researchers quoted below reached different conclusions. Olsson (1974) investigated pollen grain features from parental plants of Linaria vulgaris Mill. and L. repens Mill. (Plantaginaceae) and controlled crosses of F1 and F2 generations. In this case, a high percentage of grains with a reduced equatorial diameter (E) was typical of the F1 individuals. The median values of equatorial diameter were smaller © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 POLLEN FEATURES OF SELECTED CRATAEGUS TAXA than the smallest values of the equatorial diameter for the parental species by 3–4 mm, i.e. by c. 20%. Similar results were found for the length of the polar axis (P) in parental species and hybrids. According to Srivastava (1978), pollen grains of three cultivars of Cajanus cajan (L.) Huth and their hybrids (Fabaceae) are characterized by similar length values of the polar axis (P). The morphological analysis indicated dominance of the male parent in the hybrid pollen grains. Investigations conducted by Delaporte et al. (2001) on interspecific hybrids of Eucalyptus L’Hér (Myrtaceae), which focused on numerous morphological traits, including operculum length and width, failed to provide unambiguous results. The operculum lengths of hybrids from the F1 generation and from the openpollinated second generation were similar and, on average, were 10.9 and 10.1 mm, whereas, in the four Eucalyptus spp. analysed, they were slightly higher and ranged from 13.1 to 13.3 mm. The operculum width is a feature which is much more variable than the length, and thus measurement results were more variable. The operculum width was found to be similar in hybrids from the F1 generation and from the open-pollinated second generation (10.9 and 10.3 mm), whereas, in individual parental species, the measurements of this feature differed widely (2.1– 14.5 mm). Franssen et al. (2001) examined pollen morphological variation among species of Amaranthus L. (Amaranthaceae) and interspecific hybrids from the point of view of differences between the monoecious and dioecious Amaranthus spp. According to these researchers, pollen grains of the dioecious species usually had a greater number of apertures on the visible surface. However, pollen diameters (E) did not differ between the monoecious and dioecious plants. Pollen of the hybrids was similar in size to that of the maternal parent, but the number of apertures was intermediate between those of the parental species. This indicates that pollen characteristics may be controlled by the female, contrary to the claim of Srivastava (1978), and that hybrids may be more prevalent than originally thought. Jiang et al. (2001) ascertained that the pollen sizes of F1 hybrids did not exceed those of parental taxa (Gastrodia elata Blume forma glauca S. Chow and G. elata forma elata; Orchidaceae). These results indicated that pollen grains from the F1 hybrids expressed general or transitional morphology of the pollen grains of the parental taxa. Lu et al. (2002) observed the shape, size, germinal aperture and exine sculpture of the pollen grains of a new hybrid Eriobotrya japonica (Thunb.) Lindl. ‘Zaozhong 6’ and parents ‘Moriowase’ and ‘Jiefangzhong’ (Rosaceae). The analysed pollen features of the three cultivars differed in terms of size and exine sculpture, with those of ‘Zaozhong 6’ being transitional [e.g. the pollen shapes (P/E ratio) of ‘Morio- 567 wase’, ‘Jiefangzhong’ and the hybrid (‘Zaozhong 6’) were 1.346, 1.268 and 1.193, respectively, and the average lengths of the polar axis (P) were 20.71, 24.33 and 23.83 mm, respectively]. Dönmez (2008), examining the pollen morphology of Turkish Crataegus taxa, did not observe a significant increase in pollen size in polyploid species. The only article dealing with natural (spontaneous) hybrids was published by Karlsdóttir et al. (2008), who examined the pollen size and shape of 22 natural, triploid Betula L. hybrids and parental species (Betulaceae). The mean diameter of the pollen grains from the triploid hybrids was not statistically significantly different from that of B. nana L. pollen, but was significantly smaller than the mean value of B. pubescens Ehrh. pollen. The results obtained in our investigations were also ambiguous. A strong argument in favour of the thesis that the spontaneous hybrids studied were characterized by greater pollen grain diameter in comparison with parental species results from contrast analysis (Table 5). For most of the analysed pollen features, hybrids were characterized by significantly higher values than the mean of the parental species. Moreover, differences between parental species and hybrids were also confirmed by the shape analysis of pollen grains of the taxa examined (Table 3). Pollen grains of the parental species clearly differed from one another with respect to this feature, in contrast with hybrids, which were characterized by similar proportions of pollen grains in individual shape classes, but it should be emphasized that, in C. rhipidophylla, the proportion of pollen grains in individual pollen shape classes was similar to that in hybrids. The remaining statistical analyses examining the relationships among individual taxa showed that the pollen grains of two parental species, C. laevigata and C. monogyna, were most similar to each other and differed significantly, not only with respect to the pollen grains of the three spontaneous hybrids, but also with regard to the pollen of the third parental species, C. rhipidophylla (Table 5). However, based on the results of the canonical variable analysis, C. ¥ macrocarpa (C. laevigata ¥ C. rhipidophylla) and C. ¥ media (C. monogyna ¥ C. laevigata) were the most similar pollen grains. A second pair was formed by C. ¥ subsphaericea (C. monogyna ¥ C. rhipidophylla) and C. rhipidophylla, whereas C. monogyna and C. laevigata formed separate, ‘single species groups’ (Fig. 17). The effectiveness of the multi-feature analysis of the Crataegus spp. and hybrids was corroborated by the fact that the first two canonical variables accounted for 91.92% of the overall variability. The usefulness of the method of canonical variables is further corroborated by its © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 568 D. WROŃSKA-PILAREK ET AL. widespread application (Rybiński et al., 2009, 2011; Seidler-Łożykowska & Bocianowski, 2012). Pollen grains of C. rhipidophylla are larger than those of C. laevigata and C. monogyna, and are more similar to the pollen of the hybrids examined. Although, in classical taxonomy, C. rhipidophylla is considered as a ‘good species’, in other words it is characterized by distinct features which distinguish it from the remaining Crataegus spp., Zieliński (1977), based on morphology (features of leaves, stipules, fruit and sepals), classified C. curvisepala Lindm. (= C. rhipidophylla) as a species occupying an intermediate position between C. monogyna and C. calycina Peterm. (= C. lindmanii Hrabětová). Zieliński (1977) maintained that it is probably an old, conserved hybrid between C. monogyna and C. calycina [= C. rhipidophylla var. lindmanii (Hrabětová) P.A.Schmidt]. This thesis was also supported by the analysis of Christensen (1992), which, based on morphological features, found that 30% of Crataegus spp. were hybrid in origin. Therefore, we decided to determine whether the pollen morphology supports Zieliński’s (1977) theory. Even though our study is predominantly based at the species level, we compared the pollen morphology of C. rhipidophylla at the varietal level. We collected, in addition, three samples of C. rhipidophylla var. lindmanii, each of 30 pollen grains (Góra Zelejowa, Świe˛tokrzyskie Province, 50°49′5.988″N, 20°27′33.012″E; Ojcowski Park Narodowy, Małopolskie Province, 50°12′42.44″N, 19°49′49.33″E; Dopikujdowa, Małopolskie Province, 49°31′6.959″N, 19°47′18.599″E), and compared these for all the palynological features obtained for C. rhipidophylla var. rhipidophylla. The samples were also re-reviewed by Professor Jerzy Zieliński. There were statistically significant differences (P < 0.0001) between C. rhipidophylla var. rhipidophylla and C. rhipidophylla var. lindmanii in P (mean values: 38.67 vs. 42.02 mm, respectively), E (36.15 vs. 39.67 mm, respectively), Exp (0.857 vs. 1.066 mm, respectively), Exp/P (0.024 vs. 0.053, respectively) and Le (34.96 vs. 37.51, respectively). For the remaining pollen morphological features, there were no statistically significant differences between the varieties studied. This results corroborate Zieliński’s (1977) conjecture, because the mean values of almost all the pollen features studied [P, E, Exp, Exp/P, Le, d, d/E (PAI)] for C. rhipidophylla var. rhipidophylla are intermediate between those of C. monogyna (the lowest values of the pollen grains studied) and C. rhipidophylla var. lindmanii (= C. calycina = C. lindmanii) (the highest values of the pollen grains studied) (Table 2). Moreover, Christensen’s (1992, 1997) classification confirmed the complex character of this taxon because, according to that researcher, the present C. rhipidophylla embraces two species distinguished earlier, namely C. curvisepala and C. lindmanii (= C. calycina). Considerable variation of C. rhipidophylla was also confirmed by Christensen (1992), who divided this species into two varieties: C. rhipidophylla var. rhipidophylla, with fruits crowned by spreading or reflexed sepals, and C. rhipidophylla var. lindmanii (Hrabětová-Uhrová) Christensen, with fruits crowned by erect or suberect sepals. According to this study, the characteristics of the pollen grains of the Crataegus taxa studied are considered to be auxiliary features in the taxonomy, because they do not allow the diagnosis of individual species and only allow the distinction of groups of species (Fig. 17). Similar results were obtained by Reitsma (1966), Byatt (1976), Byatt et al. (1977), Eide (1981), Hebda et al. (1988) and Hebda & Chinnappa (1994). These results support Christensen’s (1992) hypothesis that the interpretation and practical use of pollen structure in Crataegus and other genera of Maloideae is severely limited. According to Muniyamma & Phipps (1979), in Crataegus, the occurrence of malformed pollen grains is related to extensive hybridization. Formation of the deformed pollen grains may be dependent on disturbed meiosis, noted in some of the sexual hybrids of Crataegus (Byatt et al., 1977). In the taxa studied, the deformation of pollen grains includes mainly the fracture of the aperture area and the loss of turgor pressure and change in pollen shape (such pollen grains are flat, not spherical or ellipsoidal) (Figs 5, 6, 9). This is mainly associated with acetolysis, in which pollen grains are exposed to concentrated chemicals and high temperature. Nevertheless, this process is necessary to obtain comparable measurement results of pollen grain morphology with published data from other palynologists, who, in most cases, measured acetolysed pollen grains (Reitsma, 1966; Byatt, 1976; Eide, 1981; Hebda et al., 1988; González Romano & Candau, 1989; Hebda & Chinnappa, 1990, 1994; Moore et al., 1991; Christensen, 1992; Zhou et al., 2000; Dönmez, 2008). Fedoronchuk & Savitskii (1985) found that a high proportion of deformed and sterile pollen grains was characteristic for species of probable hybrid origin and also of wide-ranging species, especially C. monogyna (80–85% abnormal grains). Our studies have not fully confirmed this hypothesis. A similar result was found only for C. rhipidophylla var. rhipidophylla samples (75% deformed grains); for the remaining taxa, the proportion of abnormal pollen grains was much lower and ranged from 20 to 45% (Fig. 16). Completely different results from those published by Fedoronchuk & Savitskii (1985) were obtained for C. monogyna, in which we found only 25% malformed pollen. The results are surprising and show that, on average, the percentage of deformed pollen grains in parental species is higher (46%) than © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 POLLEN FEATURES OF SELECTED CRATAEGUS TAXA in hybrids (30%) (Fig. 16). However, this interpretation is not straightforward. In the case of parental species, the high mean value of malformed pollen grains is affected by C. rhipidophylla var. rhipidophylla samples. Taking into account C. monogyna, C. laevigata, C. rhipidophylla var. lindmanii and the three hybrids, the share of deformed pollen grains is similar (20–45%). On this basis, one might lean towards the idea that the parental and hybrid taxa examined do not differ significantly in terms of the percentage of deformed pollen grains. ACKNOWLEDGEMENTS We kindly thank Professor Jerzy Zieliński (Institute of Dendrology, Polish Academy of Sciences, Kórnik, Poland) for the identification of plant material and Dr Lee E. Frelich (University of Minnesota, St Paul, MN, USA) for linguistic support and valuable comments on an early draft of the manuscript. We would like to thank the reviewers for their detailed and valuable comments on the manuscript. REFERENCES Byatt JI. 1976. Pollen morphology of some European species of Crataegus L. and of Mespilus germanica L. (Rosaceae). Pollen et Spores 18: 335–349. Byatt JI, Ferguson IK, Murray BG. 1977. Intergeneric hybrids between Crataegus L. and Mespilus L.: a fresh look at an old problem. Botanical Journal of the Linnean Society 74: 329–343. Campbell CS, Dickinson TA. 1990. Apomixis, pattern of morphological variation, and species concepts in subfam. Maloideae (Rosaceae). Systematic Botany 15: 124–135. Campbell CS, Evans RC, Morgan DR, Dickinson TA, Arsenault MP. 2007. Phylogeny of subtribe Pyrinae (formerly the Maloideae, Rosaceae): limited resolution of a complex evolutionary history. Plant Systematics and Evolution 266: 119–145. Chaturvedi M, Ram T, Pal M. 1993. Pollen morphology in Chorisia species and their hybrid. Phytomorphology 43: 25–28. Christensen KI. 1992. Revision of Crataegus sect. Crataegus and nothosect. Crataeguineae (Rosaceae–Maloideae) in the old world. Ann Arbor, MI: The American Society of Plant Taxonomists. Christensen KI. 1997. Typification of Crataegus kyrtostyla Fingerh. In: Wisskirchen R, ed. Notulae and Floram Germanicam I. Feddes Repertorium 108: 1–109. Christensen KI, Janjic N. 2006. Taxonomic notes on European taxa of Crataegus (Rosaceae). Nordic Journal of Botany 24: 143–147. Christensen KI, Zieliński J. 2008. Notes on the genus Crataegus (Rosaceae–Pyreae) in southern Europe, the Crimea and western Asia. Nordic Journal of Botany 26: 344–360. 569 Delaporte K, Conran J, Sedgley M. 2001. Morphological analysis to identify the pollen parental of an ornamental interspecific hybrid Eucalyptus. Scientia Horticulturae 89: 55–72. Dickinson TA, Lo E, Talent N. 2007. Polyploidy, reproductive biology, and Rosaceae: understanding evolution and making classifications. Plant Systematics and Evolution 266: 59–78. Dickinson TA, Phipps JB. 1986. Studies in Crataegus (Rosaceae, Maloideae), XIV. The breeding system of Crataegus crus–galli sensu lato in Ontario. American Journal of Botany 73: 116–130. Dönmez AA. 2004. The genus Crataegus L. (Rosaceae) with special reference to hybridization and biodiversity in Turkey. Turkish Journal of Botany 28: 29–37. Dönmez EO. 2008. Pollen morphology in Turkish Crataegus (Rosaceae). Botanica Helvetica 118: 59–70. Eide F. 1981. Key for northwest European Rosaceae pollen. Grana 20: 101–118. Ennos RA, Whitlock R, Fay MF, Jones B, Neaves LE, Payne R, Taylor I, De Vere N, Hollingsworth PM. 2012. Process-Based Species Action Plans: an approach to conserve contemporary evolutionary processes that sustain diversity in taxonomically complex groups. Botanical Journal of the Linnean Society 168: 194–203. Erdtman G. 1952. Pollen morphology and plant taxonomy. Angiosperms. An introduction to palynology 1. Stockholm: Almquist & Wiksell. Fedoronchuk NM, Savitskii VD. 1985. Palynomorphological study of the Ukrainian species of the genus Crataegus (Rosaceae). Botanicheskii Zhurnal 70: 1190–1196 [in Russian]. Franco J. 1968. Crataegus calycina Peterm. and C. curvisepala Lindman. In: Heywood VH, ed. Flora Europaea Notulae Systematicae ad Floram Europaeam spectantes 7. Feddes Repertorium 79: 1–68. doi:10.1002/fedr. 19680790102. Franssen AS, Skinner DZ, Al–Khatib K, Horak MJ. 2001. Pollen morphological differences in Amaranthus species and interspecific hybrids. Weed Science 49: 732–737. GenStat. 2007. GenStat release 10 reference manual. Rothamsted, Hertfordshire: Lawes Agricultural Trust. González Romano ML, Candau PA. 1989. Contribution to palynological studies in the Rosaceae. Acta Botanica Malacitana 14: 105–116. Grant V. 1971. Plant speciation. New York, London: Columbia University Press. Hebda RJ, Chinnappa CC. 1990. Studies on pollen morphology of Rosaceae in Canada. Review of Palaeobotany and Palynology 64: 103–108. Hebda RJ, Chinnappa CC. 1994. Studies on pollen morphology of Rosaceae. Acta Botanica Gallica 141: 183–193. Hebda RJ, Chinnappa CC, Smith BM. 1988. Pollen morphology of the Rosaceae of western Canada I. Agrimonia to Crataegus. Grana 27: 93–113. Hesse M, Halbritter H, Zetter R, Weber M, Buchner R, Frosch–Radivo A, Ulrich S. 2009. Pollen terminology. An illustrated handbook. Vienna: Springer. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 570 D. WROŃSKA-PILAREK ET AL. Hossain MM, Inden H, Asahira T. 1990. Pollen morphology of interspecific hybrids of Brassica oleracea and B. campestris. HortScience 25: 109–111. Jiang L, Wan SY, Wang SB, Yu CJ. 2001. Scanning electron microscopy (SEM) observation of pollen morphology in Gastrodia elata Bl. f. glauca S. Chow and G. elata Bl. f. elata and their hybrids. Journal of Huazhong Agricultural University 20: 89–91. Karlsdóttir L, Hallsdóttir M, Thórsson AT, Anamthawat–Jónsson K. 2008. Characteristics of pollen from natural triploid Betula hybrids. Grana 47: 52–59. Kuprianowa LA, Alyoshina LA. 1978. Pollen dicotyledonarum florae partis Europeae URSS Lamiaceae– Zygophyllaceae. Leningrad: Nauka. Lippert W. 1978. Zur Gliederung und Verbreitung der Gattung Crataegus in Bayern. Die Veröffentlichungen der Bayerischen Botanischen Gesellschaft 49: 165–198. Lu XM, Chen JY, Zhang LM, Zheng SQ, Yu D, Liao RY. 2002. Observation and comparison on pollen morphology of a new hybrid loquat variety ‘Zaozhong 6’ and its parentals. Acta Horticulturae Sinica 29: 271–273. Mahalanobis PC. 1936. On the generalized distance in statistics. Proceedings of the National Institute of Science of India 12: 49–55. Moore PD, Webb JA, Collinson ME. 1991. Pollen analysis. London: Blackwell Scientific Publications. Muniyamma M, Phipps JB. 1979. Cytological proof of apomixis in Crataegus (Rosaceae). American Journal of Botany 66: 149–155. Nair PKK, Katiar K, Srivastava GS. 1977. Pollen morphology of an Erythrina hybrid and their parentals. Current Science 46: 824–825. Naruhashi N, Takano H. 1980. Size variation of pollen grains in some Rubus species. Journal of Phytogeography and Taxonomy 28: 27–32. Ohashi H, Hoshi H, Iketani H. 1991. Taxonomy and pollen morphology of hybrids between Sorbus and Micromeles in the genus Sorbus (Rosaceae subfamily Maloideae). Journal of Japanese Botany 66: 110–124. Olsson U. 1974. A biometric study of the pollen morphology of Linaria vulgaris (L.) Miller and L. repens (L.) Miller (Scrophulariaceae) and their hybrid progeny in F1 and F2 generations. Grana 14: 92–99. Phipps JB. 1983. Crataegus – a nomenclator for sectional and serial names. Taxon 32: 598–604. Phipps JB. 1988. Crataegus (Maloideae, Rosaceae) of the southeastern United States, I. Introduction and series Aestivales. Journal of the Arnold Arboretum 69: 401–431. Phipps JB, O’Kennon RJ, Lance RW. 2003. Hawthorns and medlars. Cambridge: Royal Horticultural Society. Phipps JB, Robertson KR, Rohrer JR, Smith PG. 1991. Origin and evolution of subfam. Maloideae (Rosaceae). Systematic Botany 16: 303–332. Phipps JB, Robertson KR, Smith PG, Rohrer JR. 1990. A checklist of the subfamily Maloideae (Rosaceae). Canadian Journal of Botany 68: 2209–2269. Polyakova TA, Gataulina GN. 2008. Morphology and variability of pollen of the genus Spiraea L. (Rosaceae) in Siberia and the Far East. Contemporary Problems of Ecology 1: 420–424. Punt W, Hoen PP, Blackmore S, Nilsson S, Le Thomas A. 2007. Glossary of pollen and spore terminology. Review of Palaeobotany and Palynology 1431: 1–81. Rao PM, Kumar CR, Nair PKK. 1979. Pollen morphology of Cotyledon grandiflora and Echeveria sp. and their hybrid. Current Science 48: 909–910. Reitsma TJ. 1966. Pollen morphology of some European Rosaceae. Acta Botanica Neerlandica 15: 290–379. Rhee HK, Cho HR, Kim KJ, Kim KS. 2005. Comparison of pollen morphology in interspecific hybrid lilies after in vitro chromosome doubling. Acta Horticulturae 673: 639– 643. Rizaeva SM, Akhmedova MZ, Abdullaev AA. 1985. Pollen viability and pollen grain morphology in interspecific cotton hybrids differing in origin and ploidy. Sel’skokhozyaistvennaya Biologiya 9: 63–66 [in Russian]. Robertson KR, Phipps JB, Rohrer JR, Smith PG. 1991. A synopsis of genera in Maloideae (Rosaceae). Systematic Botany 16: 376–394. Rybiński W, Szot B, Bocianowski J, Rusinek R. 2011. Geometric properties of grasspea seeds and their mechanical loads. International Agrophysics 25: 271–280. Rybiński W, Szot B, Rusinek R, Bocianowski J. 2009. Estimation of geometric and mechanical properties of seeds of Polish cultivars and lines representing selected species of pulse crops. International Agrophysics 23: 257–267. Seidler-Łożykowska K, Bocianowski J. 2012. Evaluation of variability of morphological traits of selected caraway (Carum carvi L.) genotypes. Industrial Crops and Products 35: 140–145. Shapiro SS, Wilk MB. 1965. An analysis of variance test for normality (complete samples). Biometrika 52: 591–611. Sokal RR, Rohlf FJ. 1995. Biometry: the principles and practice of statistics in biological research. New York: Freeman WH and Co. Sørensen M. 1989. Pollen morphology of species and interspecific hybrids in Pachyrhizus Rich. ex DC. (Fabaceae: Phaseoleae). Review of Palaeobotany and Palynology 61: 319–339. Srivastava V. 1978. Pollen morphology of Cajanus cajan (L.) Willd. (Leguminosae) cultivars and their hybrids. Grana 17: 107–109. Srivastava V, Pal M, Nair PKK. 1977. A study of the pollen grains of Amaranthus spinosus Linné and A. dubius Mart ex Thellung and their hybrids. Review of Palaeobotany and Palynology 23: 287–291. Talent N, Dickinson TA. 2005. Polyploidy in Crataegus and Mespilus (Rosaceae Maloideae): evolutionary inferences from flow cytometry of nuclear DNA amounts. Canadian Journal of Botany 83: 1268–1304. Van der Walt ID, Littlejohn GM. 1996. Pollen morphology, male hybrid fertility and pollen tube pathways in Protea. South African Journal of Botany 62: 236–246. Wrońska-Pilarek D. 1998. Pollen morphology of the Polish species of the genus Ribes L. Acta Societatis Botanicorum Poloniae 67: 275–285. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571 POLLEN FEATURES OF SELECTED CRATAEGUS TAXA Wrońska-Pilarek D. 2011. Pollen morphology of Polish native species of the Rosa genus (Rosaceae) and its relation to systematics. Acta Societatis Botanicorum Poloniae 80: 221–232. Wrońska-Pilarek D, Jagodziński AM. 2009. Pollen morphological variability of Polish native species of Rosa L. (Rosaceae). Dendrobiology 62: 71–82. Wrońska-Pilarek D, Jagodziński AM. 2011. Systematic importance of pollen morphological features of selected species from the genus Rosa (Rosaceae). Plant Systematics and Evolution 295: 55–72. Wrońska-Pilarek D, Jagodziński AM. 2012. Intra- and inter-individual variability of selected quantitative features of pollen grain morphology based on the example of Rosa canina L. (Rosaceae). Dendrobiology 67: 25–39. 571 Wrońska-Pilarek D, Jagodziński AM, Maliński T. 2012. Morphological studies of pollen grains of the Polish endemic species of the genus Rubus (Rosaceae). Biologia 67: 87– 96. Xin X, Zhang YM, Wang J. 1986. Observations on the pollen morphology of major Crataegus species in China. Scientia Agricultura Sinica 3: 94–95 [in Chinese]. Zhou LH, Wie ZX, Wu ZY. 2000. Pollen morphology of Maloideae of China (Rosaceae). Acta Botanica Yunnanica 22: 47–52 [in Chinese]. Zieliński J. 1977. Crataegus curvisepala Lind. and C. microphylla C. Koch in Bulgaria. Arboretum Kórnickie 22: 29–38. Zieliński J. 1982. Crataegus heldreichii Boiss. in Bulgaria. Fragmenta Floristica et Geobotanica Polonica 28: 625–627. © 2013 The Linnean Society of London, Botanical Journal of the Linnean Society, 2013, 172, 555–571