Molecular systematics, character evolution, and pollen morphology of Cistus and Halimium (Cistaceae)

Céline Pélissier; Thierry Otto

The lack of a comparative approach makes it impossible to determine the main factors influencing colonization and evolution in plants. Here we conducted the first comparative study of a characteristic Mediterranean lineage (white-flowered Cistus) taking advantage of its well-known phylogenetic relationships. A two-scale approach was applied to address the hypothesis of higher levels of isolation in mountain than in lowland species. First, a time-calibrated phylogeny using plastid sequences of Cistaceae suggested that the origin of Cistus species postdated both the refilling of the Mediterranean Sea (5.59–5.33 Ma) and the onset of the Mediterranean climate (3.2 Ma). Two hundred and sixty-three additional, plastid sequences from 111 populations showed different numbers of haplotypes in C. laurifolius (7), C. monspeliensis (2) and C. salviifolius (7). Although haplotype sharing among disjunct populations was observed in all species, phylogeo-graphic analyses revealed haplotype lineages exclusive to Europe or Africa only in the mountain species (C. laurifolius). Isolation by either geographical distance or sea barriers was not significantly supported for the lowland species (C. monspeliensis; C. ladanifer from a previous study). The same is true for the less habitat-specific species of the lineage (C. salviifolius). Comparative phylogeography of the Cistus species leads us to interpret a general pattern of active colonization surpassing Mediterranean barriers. In contrast, ecological conditions (precipitation, temperature, soils) appear to have determined the distribution of the Cistus species of Mediterranean mountains. This study further provides molecular evidence for multiple colonization patterns in the course of successful adaptation of Cistus species to Mediterranean habitats.

Flow cytometry, using propidium iodide and 4',6-diamidano-2-phenylindole staining, was used to estimate the nuclear DNA content (2C) and the proportion of A-T base pairs in 16 species of the Mediterranean genus Cistus. Genome sizes were shown to be constant within species, since no significant intraspecific variation in 2C DNA content was detected. At the genus level, up to about 1.5-fold differences in absolute DNA amounts were observed, ranging from 3.92 pg in C. crispus to 5.88 pg in C. monspeliensis. The (AT) : (GC) ratio was close to 1, and was similar for all species examined, ranging from 47.87% A-T content in C clusii, to 50.67% in C. populifolius. Pink-flowered species (subgenus Cistus) had lower DNA amounts than white-flowered species (subgenera Leucocistus and Halimioides). However, the distribution of DNA amounts in Cistus appeared to be continuous and did not permit a clear separation of infra-generic ranks in the genus.

The tribe Naucleeae has recently been recircumscribed on the basis of both morphological and molecular [rbcL, trnT-F, internal transcribed spacer (ITS)] evidence, and has been found to be the sister group of the tribe Hymenodictyeae Razafim. & B. Bremer. In order to find pollen morphological support for this new classification, the pollen and orbicules of 65 species, representing 23 Naucleeae

Plant Syst Evol (2011) 295:23–54 DOI 10.1007/s00606-011-0458-7 ORIGINAL ARTICLE Molecular systematics, character evolution, and pollen morphology of Cistus and Halimium (Cistaceae) Laure Civeyrel • Julie Leclercq • Jean-Pierre Demoly Yannick Agnan • Nicolas Quèbre • Céline Pélissier • Thierry Otto • Received: 23 September 2010 / Accepted: 13 April 2011 / Published online: 12 June 2011 Ó Springer-Verlag 2011 Abstract Pollen analysis and parsimony-based phylogenetic analyses of the genera Cistus and Halimium, two Mediterranean shrubs typical of Mediterranean vegetation, were undertaken, on the basis of cpDNA sequence data from the trnL-trnF, and trnS-trnG regions, to evaluate limits between the genera. Neither of the two genera examined formed a monophyletic group. Several monophyletic clades were recognized for the ingroup. (1) The ‘‘white and whitish pink Cistus’’, where most of the Cistus sections were present, with very diverse pollen ornamentations ranging from striato-reticulate to largely reticulate, sometimes with supratectal elements; (2) The ‘‘purple pink Cistus’’ clade grouping all the species with purple pink flowers belonging to the Macrostylia and Cistus sections, with rugulate or microreticulate pollen. Within this clade, the pink-flowered endemic Canarian species formed a monophyletic group, but with weak support. (3) Three Halimium clades were recovered, each with 100% bootstrap support; all Halimium species had striato-reticulate L. Civeyrel (&) Y. Agnan C. Pélissier T. Otto Laboratoire Ecologie Fonctionnelle et Environnement, Université de Toulouse, UPS, 31062 Toulouse Cedex 9, France e-mail: laureciv@cict.fr J. Leclercq UMR DAP, CIRAD, TA A-96/03, Avenue Agropolis, 34398 Montpellier Cedex 5, France J.-P. Demoly APBF, 15 bis rue de Marignan, 75008 Paris, France N. Quèbre ETH Zentrum, IBZ, Universitätstrasse 16, CHN-G35.1, 8092 Zurich, Switzerland pollen. Two Halimium clades were characterized by yellow flowers, and the other by white flowers. Keywords TrnL-F TrnS-G Pollen Exine Cistaceae Cistus Halimium Introduction Specialists on the Cistaceae usually acknowledge eight genera for this family (Arrington and Kubitzki 2003; Dansereau 1939; Guzmán and Vargas 2009; Janchen 1925): Cistus, Crocanthemum, Fumana, Halimium, Helianthemum, Hudsonia, Lechea and Tuberaria (Xolantha). Two of these, Lechea and Hudsonia, occur in North America, and Crocanthemum is present in both North America and South America. The other genera are found in the northern part of the Old World. The two shrubby genera of the family, Cistus (22 species) and Halimium (nine to 14 species), are restricted to the Mediterranean basin and are mainly found in open vegetation (matorral) (Dansereau 1939). They are both sun lovers and are large to small shrubs, reaching up to 3 m in height (Cistus ladanifer L.). Cistus and Halimium are recognized as distinct by most botanists, although they share some characters, for example chromosome number (2n = 18), which is different from all the other genera of the family (Helianthemum 2n = 20–24; Tuberaria 2n = 14; Fumana 2n = 32; Xolantha 2n = 36; Crocanthemum 2n = 20 in Demoly and Montserrat 1993). They hybridize in the wild and in cultivation, making genus delimitation more tedious. Intergeneric hybrids have been described: 9 Halimiocistus revolii (Coste) Dans., with two nothosubspecies, 9 H. sahucii (Coste and Soulié) Janch., and 9 H. humilis Demoly, and a fourth has been reported (Cistus 9 heterogenus Bornet, n.n.) (Demoly 1998). Some 123 24 intrageneric hybrids of Cistus are more difficult to obtain artificially than the intergeneric hybrids (Demoly 1996). Classification of these two genera has mainly been based on morphological characters. Cistus has pink or white flowers with five locules in each ovary (except for Cistus ladanifer which has 6–12). Halimium has yellow or white flowers with three locules in each ovary. Nevertheless, white flowers and four locules in each ovary are found in both genera, as shown by Demoly (1998). Moreover, the number of sepals can be variable in both genera. Pinkflowered Cistus species, which were considered by Dansereau (1939) as ‘‘basal’’, generally have five large sepals (with the exception of Canarian species). Other species in Cistus or Halimium have either three large sepals or 3 ? 2 sepals (three large and two small). All these distinctive characters sometimes overlap and transitions between the two genera are almost continuous. Guzmán and Vargas (2005) proposed a phylogenetic hypothesis for 20 species of Cistaceae based on plastid and nuclear DNA sequence data. More recently, they conducted a total evidence analysis combining nuclear and plastid DNA sequences to assess the adaptive radiation of Mediterranean Cistus (Guzmán et al. 2009). In a third paper, Guzmán and Vargas (2009) reconstructed the first phylogeny comprising a representative sample of all known Cistaceae genera using sequences of plastid DNA. Guzmán and Vargas (2009), Nandi (1998), and Arrington and Kubitzki (2003) followed Ukraintseva’s Cistaceae pollen classification (1993), in which Tuberaria, Halimium, and Cistus share the ‘‘same pollen type’’. Palynological studies of Cistaceae published by Ukraintseva in 1993 are mainly based on light microscopy from her previous work in Russian (Ukraintseva 1991), and on previous work by Kultina (in Russian; cited in Ukraintseva 1993). Because it involves work dealing with the whole family, her simplification, grouping Tuberaria, Halimium, and Cistus under the same pollen type, is understandable. One of her objectives was to provide tools for paleoecological or paleofloristic reconstructions of the past by paleoflorists (Ukraintseva 1993), but her classification is oversimplistic when dealing with Cistaceae species. Pollen of the Cistaceae attracted the attention of botanists and paleoclimatologists long before Ukraintseva, as they are clearly typical of Mediterranean vegetation. Several palynologists have studied the pollen morphology of Cistus or of Halimium (Heydacker 1963; Jean and Pons 1962, 1963; Jiménez-Albarrán 1984; Marquez et al. 1996; Palacios-Chavez et al. 1999; Reille 1990; Saens de Rivas 1979; Ukraintseva 1993), but the number of species studied varied depending on the authors, from two (Heydacker 1963) to 17 (Saens de Rivas 1979). Strong contradictions have been encountered in the literature for pollen size and exine thickness (mainly between Saens de Rivas 1979; Jean and Pons 1963 and Ukraintseva 1991, 1993) and there 123 L. Civeyrel et al. is no complete study of these two genera. In our study, we examined all the available species of Cistus and Halimium in order to evaluate palynological characters (Table 1). Non-coding chloroplast regions are another source of information for evaluating relationships between closelyrelated taxa. They display higher levels of variation than coding regions and have been used extensively for lower taxonomic studies (Borsch and Quandt 2009; Clegg and Zurawski 1992; Downie and Palmer 1992; Mort et al. 2007; Shaw et al. 2005; 2007). Mapping characters on to a molecular phylogeny provides insights into patterns of character evolution independently of the characters themselves. The trnL-trnF region (hereafter trnL-F) is one of the most frequently used molecular markers in phylogenetic reconstruction (Borsch and Quandt 2009; Quandt et al. 2004; Shaw et al. 2005; Taberlet et al. 1991). It consists of two transfer RNA genes, trnLUAA trnFGAA, separated by non-coding regions: the spacer trnL-trnF, and the trnL intron, the latter being the only group I intron in the plastid genome of land plants (Borsch and Quandt 2009). The trnL intron is generally more conserved than the trnL-F spacer and, for that reason, it is assumed to structure the topology, whereas the relatively high variation of the trnL-F spacer is assumed to resolve inter-specific relationships (Borsch and Quandt 2009; Lahaye et al. 2007). We also used a second region to increase the robustness of the molecular phylogeny by including the trnSGCUtrnGUCC intergenic region (hereafter trnS-G), for which universal primers have been designed. This region is rapidly evolving and has been used to assess genetic variation within populations of Corythophora (Hamilton 1999). Gaskin and Schaal (2003) showed that in Tamarix the trnSG spacer can be five times more variable than the trnL-trnF spacer and thus may provide more variable characters than even ITS. Unfortunately, this high level of variation can render some sequences unalignable between distant genera within the same family (Olson 2002; Shaw et al. 2005) and they are sometimes difficult to amplify (Shaw et al. 2005). The objectives of this study were to assess phylogenetics relationships for Cistus and Halimium. Morphological and palynological characters were used to establish limits between the two genera and to provide new information on the evolutionary history of these genera in the Mediterranean region, and especially on Mediterranean islands where speciation of this group has occurred. Materials and methods Material Samples for DNA, morphological, or pollen character studies were collected in the field or taken from herbarium Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) 25 Table 1 Plant vouchers for DNA and pollen studies Taxon Origin Pollen voucher Pollen no. DNA voucher trnS-G trnL trnL-F Cistus albidus L. France Civeyrel no. 1192 332 Civeyrel no. 1191 This study This study This study Cistus albidus L. f. subalbus (Dun.) Dans. Cultivated Demoly no. 69 363 Cistus asper Demoly & Mesa Canary Is H Demoly no. 1931 1931 Demoly no. 1934 This study This study This study Cistus chinamadensis ssp. chinamadensis Bañares & Romero Canary Is T Demoly no. 4 18/05/1995 cult CBN Brest 391 Demoly no. 1211 This study This study This study Cistus chinamadensis ssp. gomerae Bañares & Romero Canary Is G Demoly no. 6 23/05/1995 cult CBN Porq. 390 Demoly no. 907 This study This study This study Cistus chinamadensis ssp. ombriosus Demoly & Marrero Canary Is H Demoly no. 1924 1924 Demoly no. 1922 This study This study This study Cistus clusii Dunal Morocco Faure sn 30/05/1931 (K) 330 Civeyrel no. 1447 This study This study This study Cistus clusii ssp. multiflorus Demoly Spain Charpin sn 9/06/1971 (G) 365 Cistus clusii ssp. multiflorus Demoly Ballearic Is Mall D’en Rabassa no. 10 (G) 366 Cistus creticus L. Greece Demoly no. 2584 This study This study This study Civeyrel no. 336 This study This study This study Civeyrel no. 1460 This study This study This study Demoly no. 2218 This study This study This study Civeyrel no. 1456 This study This study This study Demoly no. 2582 This study This study This study Cistus creticus L. Morocco Demoly no. 10 359 Cistus creticus L. (= C. villosus L.) Yugoslavia Kew no. 51843 (K) 288 Cistus creticus L. grp corsicus France Corsica Collet sn April 1899 (K) 289 Cistus creticus L. grp tauricus (= C. villosus L.) Greece Mattfeld sn 2/07/1926 (K) 280 Cistus creticus L. grp tauricus (= C. villosus L.) Crimea Davis no. 33102 (K) 287 Cistus crispus L. France Civeyrel sn 16/5/1993 336 Cistus crispus L. France Gauthier sn 20/05/1894 (MPU) 326 Cistus crispus L. Spain Sennen no. 235 (MPU) 327 Cistus heterophyllus Desf. Morocco Cistus heterophyllus Desf. Algeria Bove sn April/1839 (K) 314 Cistus heterophyllus Desf. ssp. carthaginensis (Pau) Crespo & Mateo Spain Demoly no. 8 14/5/1990 281 Cistus horrens Demoly Canary Is G C Demoly no. 2218 Cistus horrens Demoly Canary Is G C Lowe no. 38 bis (K) Cistus inflatus Pourr. ex Demoly Spain Cistus inflatus Pourr. ex Demoly France G. Hibon no. 435 (P) 273 Cistus inflatus Pourr. ex Demoly Spain Bourgeau sn 07/6/1863 (K) 320 Cistus inflatus Pourr. ex Demoly Cistus ladanifer L. ssp. ladanifer Spain Spain Durieu no. 398 (K) 321 Cistus ladanifer L. ssp. ladanifer France Demoly no. 125 333 Cistus ladanifer L. ssp. ladanifer f. albiflorus (Dun.) Dans. France Demoly no. 126 334 2218 301 323 123 26 L. Civeyrel et al. Table 1 continued Taxon Origin Pollen voucher Pollen no. Cistus ladanifer L. ssp. mauritianus Pau & Sennen (= Cistus ladanifer L. var. petiolatus Maire) Morocco Cistus ladanifer L. ssp. mauritianus Pau & Sennen (= Cistus ladanifer L. var. petiolatus Maire) Spain Demoly no. 128 282 Cistus ladanifer L. ssp. mauritianus Pau & Sennen (= Cistus ladanifer L. var. petiolatus Maire) Algeria Faure sn 16/04/32 (K) 307 Cistus ladanifer L. var. sulcatus Demoly (= C. palhinhae Ingram) Portugal Cistus ladanifer L. var. sulcatus Demoly (= C. palhinhae Ingram) Cultivated Ingram sn 02/06/1949 (K) 305 Cistus laurifolius L. France Demoly no. 112 361 Cistus laurifolius L. Turkey Civeyrel no. 1182 313 Cistus laurifolius L. ssp. atlanticus (Pitard) Sennen & Mauricio Morocco Crookhank no. 38 (K) 308 Cistus libanotis (auct. non L.) var. sedjera (Pomel) Dans. Algeria Letourneux sn 30 3 1862 (P) 274 Cistus libanotis L. Spain Bourgeau sn 01/04/1850 (K) 315 Cistus libanotis L. f. major n.n. Cultivated Demoly no. 136 357 Cistus monspeliensis L. France Demoly no. 89 331 Cistus monspeliensis L. Canary Is P Sprague & Hutchinson no. 196 (K) 303 Cistus monspeliensis L. France Oléron Is Demoly sn 07/05/1994 375 Cistus munbyi aff. Morocco Cistus munbyi Pom. Algeria Faure sn 10/05/1931 (K) Cistus ochreatus Chr. Sm. Canary Is G C Demoly no. 2248 Cistus ochreatus Chr. Sm. Canary Is G C Cistus ochreatus Chr. Sm. Canary Is G C Cistus osbeckiifolius ssp. tomentosus Bañares & Demoly Canary Is T Cistus osbeckiifolius ssp. tomentosus Bañares & Demoly Canary Is T Demoly no. 2494 Cistus osbeckiifolius Webb. ex Christ. Canary Is T Cistus osbeckiifolius Webb. ex Christ. DNA voucher trnS-G trnL trnL-F Civeyrel no. 1462 Demoly no. 2583 This study This study This study Demoly no. 1788 This study This study This study Demoly no. 638 This study This study This study Demoly no. 124 This study This study This study Civeyrel no. 1464 This study This study This study Civeyrel no. 331 This study This study This study Demoly no. 2073 This study This study This study This study This study This study This study 309 2248 Demoly no. 1709 Demoly no. 1717 Austin no. M3 25/7/1960 (K) This study 311 Demoly no. 1957 2494 Demoly no. 2480 This study Demoly sn 28/04/1993 341 Demoly no. 1210 This study This study This study Canary Is C Bramwell no. 2101 (K) 310 Cistus palmensis Bañares & Demoly Canary Is P Demoly no. 1989 Demoly no. 1908 This study This study This study Cistus parviflorus Lam. Crete Civeyrel no. 1446 This study This study This study Cistus parviflorus Lam. Greece 123 Atchley sn 1932 (K) 1989 316 Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) 27 Table 1 continued Taxon Origin Pollen voucher Pollen no. Cistus parviflorus Lam. Greece Atchley no. 677 (K) 317 Cistus parviflorus Lam. Greece Gandoger no. 7696 (K) 322 Cistus parviflorus Lam. Turkey Balansa no. 87 (MPU) 360 Cistus parviflorus Lam. (= C. asperrimus) Cistus populifolius L. ssp. major (Dun.) Heywood Cyprius Tracey no. 36 (K) 312 Cistus populifolius L. ssp. major (Dun.) Heywood Morocco Trettewy no. 448 (K) 306 Cistus populifolius L. ssp. major (Dun.) Heywood Cistus populifolius L. ssp. populifolius Spain Ellman & Hubbard no. 611 (K) 318 Cultivated Demoly no. 111 338 Cistus populifolius L. ssp. populifolius France Cistus populifolius L. ssp. populifolius (= C. narbonensis Rouy & Foucaud) Portugal Bourgeau no. 1778 (P) 272 Cistus populifolius L. ssp. populifolius (= C. narbonensis Rouy & Foucaud) France Schultz no. 19265 1884 (P) 276 Cistus pouzolzii Delile France Delille sn June 1837 (MPU) 350 Spain Cistus pouzolzii Delile Morocco Maire sn (MPU) 351 Cistus salviifolius L. France Civeyrel sn 16/05/93 335 Cistus salviifolius L. France Corsica Demoly sn 3/05/1994 383 Cistus sintenisii de Lit. Albania Demoly no. 84 356 Cistus sintenisii de Lit. Albania Civeyrel no. 1463 325 Cistus symphytifolius Lam. Canary Is T Cistus symphytifolius Lam. Canary Is T Cistus symphytifolius Lam. (Pico de Cabras) Canary Is T Cistus symphytifolius Lam. (Punta Gorda) Canary Is P Cistus symphytifolius Lam. (Punta Gorda) Cistus symphytifolius Lam. var. canus Demoly Canary Is P De La Perraudière sn 27/05/ 1855 (K) Demoly no. 2462 Demoly no. 1851 DNA voucher trnS-G trnL trnL-F Civeyrel no. 1454 This study This study This study Civeyrel no. 1459 This study This study This study Civeyrel no. 1461 This study This study This study Civeyrel no. 335 This study This study This study Demoly no. 84 This study This study This study Civeyrel no. 1453 This study This study This study Demoly no. 2463 This study This study This study Demoly no. 1832 This study This study This study This study This study This study This study 329 2462 1851 Demoly no. 1845 Canary Is P Demoly no. 882 Cistus symphytifolius Lam. var. canus Demoly Canary Is P Demoly no. 2444 Cistus symphytifolius Lam. var. canus Demoly Cistus symphytifolius Lam. var. villosus Demoly Canary Is P Cistus symphytifolius Lam. var. villosus Demoly Canary Is T Demoly no. 2023 2023 Canary Is T Demoly no. 1952 Demoly no. 2319 This study 2319 Demoly no. 1954 This study 123 28 L. Civeyrel et al. Table 1 continued Taxon Origin Fumana ericoides Pau subsp. montana (Pomel) Güemes & Muñoz France Halimium antiatlanticum Maire & Wilczek Morocco Maire sn (MPU) 271 Halimium antiatlanticum Maire & Wilczek Morocco Maire no. 173 (MPU) 342 Halimium antiatlanticum Maire & Wilczek Morocco Maire sn 08/04/1935 (MPU) 343 Halimium atlanticum Humb. & Maire Morocco Halimium atlanticum Humb. & Maire Morocco De Wilde & al. no. 2951 (P) Halimium atriplicifolium (Lam.) Spach Spain Brummitt et al. no. 5943 (K) Halimium calycinum (L.) K.Koch (= H. commutatum Pau) Portugal Halimium calycinum (L.) K.Koch (= H. commutatum Pau) Spain Belmonte et al. sn 16/04/1982 (K) 299 Halimium halimifolium (L.) Willk. Portugal Sauvage no. 157 (MPU) 349 Halimium halimifolium f. coronatum Sennen Morocco Halimium halimifolium f. lasiocalycinum (Boiss. & Reut.) Gross. Morocco Maire sn 12/06/1928 (MPU) 346 Halimium halimifolium f. lasiocalycinum (Boiss. & Reut.) Gross. Morocco Maire sn 25/04/1924 (MPU) 347 Halimium halimifolium ssp. halimifolium (L.) Willk Morocco Maire sn 06/04/1921 (MPU) 344 Morocco Demoly no. 1045 Morocco Maire sn 21/06/1930 (MPU) Halimium halimifolium multiflorum auct. non Halimium halimifolium multiflorum auct. non ssp. Salzm. ssp. Salzm. Pollen voucher Pollen no. DNA voucher trnS-G trnL trnL-F Civeyrel sn 27/06/ 2005 This study This study This study Civeyrel no. 1587 This study This study This study 275 Civeyrel no. 1595 This study This study This study 290 Civeyrel no. 1450 This study This study This study Civeyrel no. 1451 This study This study This study Civeyrel no. 1457 This study This study This study Demoly no. 1045 This study This study This study 1045 345 Halimium lasianthum (Lam.) Spach var. formosum (Curt.) Gross. Portugal Civeyrel no. 1465 This study This study This study Halimium lasianthum ssp. alyssoides (Lam.) Greuter & Burdet (= H. alyssoides (Lam.) Koch) France Demoly no. 2080 This study This study This study Halimium lasianthum ssp. alyssoides (Lam.) Greuter & Burdet Portugal Vertcourt no. 4396 (2) (K) 291 Halimium lasianthum ssp. lasianthum (Lam.) Spach Spain Breckle no. 1008 (K) 295 Halimium lasianthum ssp. lasianthum (Lam.) Spach Spain Lindberg sn 10/04/1926 (K) 296 Halimium lasianthum ssp. lasianthum (Lam.) Spach Spain Demoly sn 25/05/1995 394 Halimium lasiocalycinum ssp. rhiphaeum (Pau & Font Quer) Maire Morocco Dr. Font Quer no. 406 (G) 388 123 Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) 29 Table 1 continued Taxon Origin Halimium lasiocalycinum ssp. rhiphaeum var. elatum Pau & Font Quer Morocco Halimium lasiocalycinum ssp. rhiphaeum var. elatum Pau & Font Quer Morocco Halimium ocymoides (Lam.) Willk. Halimium ocymoides (Lam.) Willk. f. elongatum (Vahl.) Gross. Spain Pollen voucher Pollen no. Maire sn 24/06/1926 (MPU) trnS-G trnL trnL-F Civeyrel no. 1583 This study This study This study Civeyrel no. 1445 This study This study This study Demoly no. 641 This study This study This study Demoly no. 1356 This study This study This study Civeyrel no. 1449 This study This study This study Demoly no. 2571 This study This study This study 348 Spain Bourgeau no. 240 (K) 292 Halimium ocymoides (Lam.) Willk. f. sampsucifolium (Cav.) Gross. Portugal Atchley no. 127 (K) 293 Halimium umbellatum (L.) Spach France Urgel no. 51906 (K) DNA voucher Halimium umbellatum (L.) Spach France Halimium umbellatum (L.) Spach (= H. verticillatum (Brot.) Sennen) Portugal 298 Halimium umbellatum (L.) Spach f. syriacum (Boiss.) Willk. Lebanon Gombault no. 4274 (P) 278 Halimium umbellatum (L.) Spach ssp. viscosum (Willk.) Bolos & Vigo Algeria Bourgeau sn (P) 277 Halimium umbellatum (L.) Spach ssp. viscosum (Willk.) Bolos & Vigo (= H. viscosum (Willk.) P. Silva) Halimium umbellatum (L.) Spach ssp. viscosum (Willk.) Bolos & Vigo (= H. viscosum (Willk.) P. Silva) France France Demoly no. 164 392 Halimium voldii Kit Tan, Perdetzoglou & Raus Greece Demoly no. 2571 2571 Helianthemum caput-felis Boiss. Morocco Civeyrel no. 1455 This study This study This study Helianthemum oelandicum (L.) DC. subsp. incanum (Willk.) López (= H. canum (L.) Hornem.) France Civeyrel no. 1192 This study This study This study Helianthemum syriacum (Jacq.) Dum.-Cours. Spain Civeyrel no. 1466 This study This study This study Muntingia calabura L. Porto-Rico Chase 346 (K) This study This study This study Tuberaria guttata (L.) Raf. France Civeyrel no. 1194 This study This study This study Abbreviations for the Canary Islands: G, La Gomera; G C, Gran Canaria; H, El Hierro; P, La Palma; T, Tenerife specimens (from G, K, MPU and P herbaria). When possible, we sampled recognized sub-species or varieties, or strongly isolated populations. The list of samples is given in Table 1 with the reference voucher and geographical origin. Samples used for pollen or DNA studies, voucher specimens, and sources are listed in Table 1. All the 106 sequences of Cistaceae were produced for this study. For trnL-F, all available species or subspecies of Cistus and Halimium and representatives of the other European genera 123 30 of Cistaceae were included and constituted the ingroup, with Muntingia calabura L. as the outgroup. The genus Muntingia appeared as one of the sister taxa of Cistaceae (Alverson et al. 1998, fig. 3, p. 880). It was also included in the outgroups in Guzmán and Vargas (2009). It shares some characters with the Cistaceae, for example crumpled petals in buds (Bayer et al. 1998). Muntingia was unalignable for trnS-G, so we used our trnL-F analyses and the results of Guzmán and Vargas (2009) to select Fumana ericoides Pau subsp. montana (Pomel) Güemes and Muñoz as the outgroup to root the trees for the combined analysis. Methods Palynology Pollen grains were removed from anthers and acetolysed according to Erdtman’s (1960) standard acetolysis method. For electron microscopy, pollen grains were air dried on SEM stubs from 100% ethanol and coated with platinum by use of an SPD 050 Balzers sputter-coater. They were examined with an Hitachi S-2400 scanning electron microscope (SEM) at the Royal Botanic Gardens, Kew. Exine ornamentation and structure were described from SEM images at the same magnification (15,0009) on the same region between two apertures on the equator. For light microscopy, pollen grains were mounted in glycerol jelly sealed with paraffin and observed with an optical microscope. The type and number of apertures were examined by light and electron microscopy. We considered three palynological characters for this study: pollen shape, Fig. 1 SEM micrographs of exine surface ornamentation types, all at identical magnification (originally 915,000). a Cistus laurifolius 313, reticulate exine b C. heterophyllus subsp. carthaginensis 281, microreticulate exine. c C. albidus 332, rugulate exine. d Halimium lasianthum ssp. lasianthum 296, striatoreticulate exine. e H. calycinum 299, striato-reticulate exine. f C. populifolius var. populifolius 306, largely reticulate exine. g C. monspeliensis 303, micro-reticulate with smooth supratectal ornamentation. h C. inflatus 273, micro-reticulate with smooth supratectal ornamentation. i C. salviifolius 335, microechinate ornamentation. All scale bars = 2 lm 123 L. Civeyrel et al. exine thickness, and exine sculpture. Character states were defined according to reference works in pollen terminology (Punt et al. 1994, 2007; Hesse et al. 2009; http://www. paldat.org/ 2010) and completed by personal observations. Pollen shape terminology was defined by Erdtman (1943, 1952), and shape classes were based on the relationship P/E between the polar axis (P) and the equatorial diameter (E); they were determined on 25 pollen grains for each sample (when possible). Exine thickness was also measured on 25 pollen grains for each sample. Thickness states are quantitative characters and it was sometimes difficult to define limits in a continuous dataset. We considered the presence of consequent gaps within the numerical order of data and reduced it to four states\2.5 lm, between 2.5 and 2.8 lm, between 2.8 and 3.15 lm and over 3.15 lm. Exine thickness was somehow related more to exine ornamentation than to pollen size. In the literature, interpretation of the exine sculpture of Cistaceae varies from one author to another, maybe because of misinterpretation, so for this study we redefined the terms we used. Exine is reticulate when muri form a network-like pattern with lumina wider than 1 lm (Fig. 1a), and is microreticulate if the lumina and the muri are less than 1 lm, and equivalent in size (Fig. 1b) (Punt et al. 2007; Hesse et al. 2009). There has been some confusion in the literature of Cistaceae between microreticulate and rugulate ornamentation. For example Cistus albidus (Fig. 1c) is described as rugulate by Jean and Pons (1963), and microreticulate by Reille (1990). Rugulate exine, defined by Iversen and Troels Smith (1950), Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) describes a type of ‘‘ornamentation of elongated muri more than 1 lm long intermediate between reticulate and striate’’. We would add here, to complete that definition, that the lumina should be much smaller on average than the muri (Fig. 1c). Striate ornamentation of exine has been described for some of our taxa (Saens de Rivas 1979), but again with some confusion. For example, pollen of Halimium umbellatum (Table 3; Fig. 5i) has been described as striato-reticulate by Márquez et al. (1996), reticulate by Jiménez-Albarrán (1984) and striate by Saens de Rivas (1979). The definition of striate ornamentation given by Iversen and Troels-Smith (1950) in Punt et al. (2007) describes an exine with ‘‘elongated and generally parallel elements separated by grooves’’. This is close to the definition given for the term rugulate, and to distinguish between the two ornamentations we complete this definition by: ‘‘elements (lumina and muri) should be more than 1 lm wide on average’’. We have already given a definition of reticulate, and things would have remained simpler without the term striato-reticulate, which somewhat confuses any interpretation. Erdtman (1952) describes a striato-reticulate exine as ‘‘a pattern in which parallel or subparallel muri are cross-linked to form a reticulum in the grooves’’. In the case of Halimium umbellatum, the exine is obviously striato-reticulate as interpreted by Márquez et al. (1996) (Fig. 5i). Nevertheless, within the reticulate type of exine, muri and lumina can be approximately of the same size, but when the lumina is much wider than the muri (1.59 wider), we used the term largely reticulate (Fig. 1a, f) as suggested by Reille (1990). For some Cistus species, the exine has been described as retipilate (Saens de Rivas 1979). This word is used to describe a pattern of exine ornamentation when the reticulum is formed by a row of pila instead of muri (Erdtman 1952). Hesse et al. (2009) pointed out that there is no example of a reticulum formed by rows of pila instead of muri. Earlier observations were based on light microscopy (Hesse et al. 2009). Moreover, the four species examined in our study and described as being retipilate were basically reticulate, so we described the type of reticulation according to the size of the muri and lumina completed by the type of supratectal ornamentation. This gave us two more states, micro-reticulate with smooth supratectal ornamentation (Fig. 1g, h), or microechinate ornamentation (Fig. 1i). Molecular techniques DNA was extracted using fresh leaf tissues or dried material conserved in silica gel using a DNeasy Plant Mini Kit (Qiagen, Courtaboeuf, France) according to the manufacturer’s instructions. Double-stranded products of 31 CpDNA were amplified from total DNA using the universal primers designed by Hamilton (1999, trnS-G) or those of Taberlet et al. (trnL-F, 1991), and the PCR procedures given by the two authors (Hamilton 1999; Taberlet et al. 1991). PCR was performed in a 50 lL reaction mixture which contained 0.4 lM of each forward and reverse primer and 0.025 U ll-1 Taq-polymerase in 10 mM Tris–HCl pH 9, 50 mM KCl, 2.5 mM MgCl2, 0.1% Triton X100, and 0.2 mg ml-1 BSA. PCR amplifications were also carried out using 1–4 ll total DNA and the Master mix TAQ PCR Qiagen (Qiagen). PCR reactions were sent for purification and sequencing to Genome Express (Meylan, France). Sequences were edited, corrected and aligned using Sequencher 4.2.2 software (Gene Code Corporation, Ann Arbor, Michigan, USA). Consensus sequences were manually aligned in a matrix under Paup 4.0b10 software for Macintosh (Swofford 2002). Analyses For trnL-F, 54 sequences were analysed separately, and 52 for trnS-G. All but an average of the first 36 bases at the 50 end and 52 at the 30 end of trnL-F, and the first 40 bases at the 50 end and 54 at the 30 end of the of trnS-G were sequenced. Alignment was straightforward for trnL-F; the length of the individual sequences varied from 618 bp (incomplete sequence for Cistus symphytifolius PicoDC) to 736 bp for trnL-F, and from 250 bp (incomplete sequence for Halimium umbellatum) to 866 bp for trnS-G. The total length of the aligned matrix is given in Table 2. Indels were found for trnL-F and trnS-G, and were quite variable in length (Table 2), often consisting of a repetition of short sequences of base pairs next to the indel itself (SSR: simple sequence repeat), or a short tandem repeat (STRs) or microsatellites (Borsch and Quandt 2009). Most of these indels can be phylogenetically informative. None of the indels were coded in the analysis, but inserted regions were retained and coded as missing. The total, matrix length, and number of informative characters are given in Table 2. Maximum parsimony analyses of the sub-matrix were conducted separately for trnL-F and trnS-G. Two combined analyses were then performed. All analyses were implemented with PAUP 4.0b10 (Swofford 2002) with the options: unit weight, heuristic search, 1,000 replicates of random taxon-additions, and TBR swapping with ACCTRAN optimization to save the shortest trees. Then, keeping all the shortest trees in memory, and with MulTrees on, an extensive search was conducted to find all the most parsimonious trees with a number limit of 20,000–50,000 trees. Successive approximation weighting (hereafter SW; Farris 1969), with characters reweighted according to their rescaled consistency index (RC) based 123 32 L. Civeyrel et al. Table 2 Summary statistics for plastid DNA data analyses of phylogenetic relationships in Cistus and Halimium Maximum parsimony analyses trnL-trnL-F trnS-G trnL-trnL-F ? trnS-G Combined molecular Number of taxa x alignment length 54 9 833 52 9 1141 52 9 1922 Number of indels 28 36 54 (18 ? 36) Indels length 1–26 1–202 1–202 Phylogentically informative characters (%) 102 (12.25%) 140 (12.27%) 226 (11.76%) UW Number of shortest trees [20,000 [20,000 [50,000 UW Tree length 331 491 725 UW Consistency indexa UW Retention index 0.855 (0.7377) 0.8797 0.8676 (0.7336) 0.8862 0.8524 (0.7284) 0.8863 UW Rescaled consistency ind. 0.7521 0.7689 0.7555 SW Number of shortest trees [20,000 [20,000 [50,000 SW Tree length 256,697 390,418 567,752 SW Consistency indexa 0.9544 (0.8923) 0.9665 (0.9087) 0.9602 (0.9046) SW Retention index 0.9609 0.9684 0.9668 SW Rescaled consistency index 0.9171 0.9359 0.9283 UW, unit weight analyses; SW, successive weighting analysis a Without uninformative characters on the best fit of characters on any of the trees, was conducted on the two combined analyses. The reasons for using SW were explained in a previous paper (Civeyrel et al. 1998). Re-weighting rounds were repeated until the tree length did not change in two consecutive iterations. The base weight of 1,000 applied in SW was removed for tree presentation. Confidence in specific clades of the resulting topology was estimated by bootstrap analysis. The settings used were: 1,000 replicates, keeping bootstrap frequencies from 50 to 100%, random addition of taxa, sampling characters with equal probability but applying weights (from SW), and TBR swapping (Tree Bisection and Reconnection) with 5 replications, but only keeping the optimum tree from each replicate, even if not optimum over all replicates. All illustrated trees used ACCTRAN optimization. We categorized bootstrap supports according to Kress et al. (2002) and considered strong for support [85%, moderate 70–85%, and weak 50–70%. Results Molecular results DNA sequence summary Table 2 provides a summary of statistics for the two molecular datasets. The aligned trnS-G data matrix contained more parsimony-informative characters (12.86% for 123 1141 bp) than the trnL-F region (12.25% for 833 bp) (Table 2). A parsimony analysis of the trnL-F dataset was conducted for 54 taxa. The removal of Muntingia calabura L. for the combined molecular analysis reduced the numbers of indels from 28 to 18, and the length of the aligned matrix from 833 to 780 bp. We removed a second taxon, Halimium lasianthum ssp. alyssoides (Lam.) Greuter & Burdet, that was too poor for trnS-G. Parsimony analysis of the trnL-F matrix (54 taxa) The parsimony analysis of the trnL-F matrix produced [20,000 trees with Muntingia calabura L. as the outgroup. The topology for the unit-weighted analysis (UW) was not different from the successive weighted analysis (SW), but the bootstrap supports were stronger. Only the former will be discussed. Flower colour defines groups quite well so we mapped this character on the strict consensus tree to help visualise groups to be discussed later (Fig. 2a). From the base to the top of the strict consensus tree, Fumana ericoides Pau subsp. montana (Pomel) Güemes and Muñoz Fig. 2 a Complete trnL-trnL-F dataset, strict consensus tree of the c 20,000 equally parsimonious trees from the successive weighting analysis. Petal colours are mapped on the tree to help visualise colour clades. b trnS-G dataset, strict consensus tree of the 20,000 equally parsimonious trees from the successive weighting analysis. Geographical origin is mapped on the branches. Numbers above branches are successive weighted bootstrap values. Branches not present in the strict consensus tree from the unit weight analysis are indicated with an arrow Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) 33 123 34 L. Civeyrel et al. The parsimony analysis of the trnS-G matrix produced 20,000 trees (tree limit). We used the result of trnL-F to select Fumana ericoides subsp. montana as the outgroup. The distribution of taxa was mapped on the tree according to the hot spot or floristic entities given in the distribution map (Fig. 3). The topology for the SW strict consensus tree was more resolved than that of UW, with three branches not present in the UW analysis and higher BS. Only the strict consensus tree from the SW is discussed (Fig. 2b). From the base to the top of the strict consensus tree we found the strongly supported clade of the three species of Helianthemum in polytomy with Fumana. Tuberaria guttata (L.) Raf. as the sister group of the Cistus and Halimium clades. For the ingroup, we only summarized the main differences between the trnL-F and trnS-G analyses. We found none of the Cistus present at the base of the tree from the trnL-F analyses, but instead we found the Halimium clades H 1 and H 3 and H 2, followed by a large group of Cistus with white or whitish pink flowers. This clade was also present in the UW analysis, but with the small clade grouping Cistus aff. munbyi and Cistus clusii in polytomy with the rest of the Cistus, and not apart as in the SW tree. Subspecies of C. ladanifer L., subspecies of C. laurifolius and a clade comprising Cistus sintenisii de Lit. and Cistus parviflorus Lam. had high BS. Two clades, the two subspecies of C. populifolius and a three-species clade comprising Cistus monspeliensis L., Cistus salviifolius L., and Cistus pouzolzii Delile, which were not present in the trnL-F analysis, had weak or no support at all (Fig. 2b). The purple pink Cistus clade was even better supported with a BS of 100%. Cistus crispus L. was again found at the base and sister to two strongly supported Fig. 3 Distribution map of the genera Cistus and Halimium. For each floristic entity the number of species, subspecies and varieties is given, followed by the number of species in brackets. Mediterranean hot spots are underlined (from Médail and Quezel 1997). The distribution of species encompasses more or less the limits of the Mediterranean climate but goes further north formed a polytomy with Muntingia, followed by the three species of Helianthemum and Tuberaria guttata (L.) Raf. as the sister group of the Cistus and Halimium clades. The Helianthemum clade and the clade formed by Tuberaria with Cistus and Halimium were strongly supported by 94% bootstrap support (hereafter BS) and each of them by 100% BS. Cistus sintenisii de Lit. was sister to the rest of the ingroup, followed by a clade formed by Cistus aff. munbyi and Cistus clusii Dunal, supported by 100% BS. From the above we found a large polytomy including 2 separate clades grouping some of the Halimium species, H 1, with white flowers and H 2, with yellow flowers, all supported by very high BS. The following were an unresolved group of white or whitish pink (Cistus parviflorus) species and infraspecific taxa of Cistus, in which only subspecies of C. ladanifer L. were grouped together (Fig. 2a). At the top of the tree we found a polytomy grouping a third clade of Halimium, H 3 with yellow flowers, the two subspecies of C. laurifolius L. with white flowers, and the remaining Cistus, all with purple pink flowers. There was strong support for the ‘‘purple pink Cistus clade’’ with a BS of 94%. Cistus crispus L. was the sister taxon to a large Cistus polytomy with only one clade comprising Cistus albidus L. and Cistus creticus L., which was strongly supported, and three quite weakly supported clades mixing Canarian taxa and Cistus heterophyllus. Parsimony analysis of the trnS-G matrix (52 taxa) 123 Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) groups, Cistus albidus L., Cistus creticus L., and Cistus heterophyllus Desf. on one side, and a complex polytomy, with 14 Canarian taxa, on the other side (Fig. 2b). Combined parsimony analysis of the trnS-G and trnL-F matrix (52 taxa) The combined parsimony analysis of the trnS-G and trnL-F matrix produced more than 50,000 trees. Fumana ericoides subsp. montana was again used as the outgroup and to root the tree. The topology for the SW strict consensus tree was more resolved than that of the UW, with 6 branches not present in the UW analysis and higher BS, but in the UW analysis the H 2 clade was not in polytomy with the ‘‘white 35 and whitish pink Cistus’’. However, because there was no BS for this branch in the UW, this did not really change the interpretation of the analysis much. Only the strict consensus tree from the SW is discussed (Fig. 4). At the base of the tree, the three species of Helianthemum were in polytomy with Fumana, and Tuberaria guttata (L.) Raf. was again the sister group of the Cistus and Halimium clades. The topology of the tree was a mixture of the two separate analyses but with stronger support on average. Each marker contributed to better resolution of the phylogeny. The three Halimium clades, H 1, H 2 and H 3, all had 100% BS, followed by the same large group of white and whitish pink Cistus. This former clade was very close to what was found in the trnS-G, but with the difference Fig. 4 Combined parsimony analysis of the trnS-G and trnLF matrix. Strict consensus tree of the 50,000 equally parsimonious trees from the successive weighting analysis. Exine ornamentation is mapped on the tree to help visualise pollen clades. Numbers above branches are successive weighted bootstrap values. Branches not present in the strict consensus tree from the unit weight analysis are indicated with arrows 123 36 that Cistus salviifolius was now sister to the C. ladanifer clade instead of being with Cistus monspeliensis L. and Cistus pouzolzii. This former clade was without any support. The ‘‘purple pink Cistus clade’’ was supported with a BS of 100%, again with Cistus crispus L. at the base and sister to two strongly supported groups found already in the trnS-G analysis. The 14 Canarian taxa clade found previously was better resolved, but still with some taxa in polytomy, and the only weakly supported clade was the group formed by Cistus ochreatus Chr. Sm., Cistus palmensis Bañares & Demoly and Cistus symphytifolius Lam. (Fig. 4). We mapped exine ornamentation on this strict consensus tree according to our observations (Fig. 4). Palynological results Pollen grains in Cistus and Halimium are stenopalynous single grains, tricolporate comprising three long ectoapertural colpi and three equatorial endoapertural pori. Their forms vary from oblate spheroidal to prolate, but 90% of the samples examined were spheroidal to sub-prolate. The exine was simplicolumellate with most samples with a microreticulate and striato-reticulate exine. Exine thickness varied from 2 to 5 lm (Table 3). Six groups of taxa were distinguished for the combined analysis, hereafter the outgroup, the three Halimium clades H 1, H 2 and H 3, the ‘‘white and whitish pink Cistus clade’’ (hereafter WWPC) and the ‘‘purple pink Cistus clade’’ (hereafter PPC), in which the Canarian and the nonCanarian taxa could be distinguished. We discuss: 1 the pollen data taxa for each of these groups to establish whether there is a common trend in their palynological morphology; and 2 the discrepancies between our observations and the literature or in the literature itself, when applicable. Pollen morphology of outgroups We used the palynological characters given by the literature, with the exception of the surface of the exine, because of the discrepancies already found for the ingroup. We reinterpreted the exine surface from the original SEM images given by the authors when available. All species were straightforward, except Fumana ericoides, where two very different types of exine have been proposed. For Jean and Pons (1963) and Heydacker (1963), the exine is reticulate but it is retipilate for Saenz de Rivas (1979). In the absence of more data, the exine was coded for this species as equivocal. For other taxa, there is some confusion in the descriptions given by Jean and Pons (1963), the exine being described as 123 L. Civeyrel et al. striate, but often authors talk about a deeper reticulum. For Saenz de Rivas (1979), the state striato-reticulate does not exist for Cistaceae; he describes either reticulate or striate exine with nothing in between. However, his images show that subparallel muri are crossed-linked to form a reticulum in the grooves. For all outgroups except Fumana we coded the surface of the exine as striatoreticulate. Pollen morphology of Halimium The authors who have examined the exine ornamentation of Halimium have described it as striate (Saenz de Rivas 1979), reticulate (Jean and Pons 1962, 1963; Jiménez-Albarrán 1984; Marquez et al. 1996), or striato-reticulate (Heydacker 1963; Jean and Pons 1962, 1963), but are usually coherent within the group. The exine thickness and grain shapes were quite variable, sometimes within the same species. The first Halimium clade H 3 (Fig. 4) was at the base of the tree. It comprised two species of Halimium with a striato-reticulate exine (Figs. 1e, 5a–c). There was a difference in the thickness of the exine and the shape of the pollen grains. Halimium atlanticum pollen grains were thicker and with a more spheroidal shape than those of H. calycinum, but their exine ornamentation was very much the same. The second Halimium clade H 2 (Fig. 4) comprised six taxa of Halimium in the combined analysis and seven in the trnL-F analysis (H. lasianthum ssp. alyssoides belonged to this group for trnL-F). All had a striato-reticulate exine with the exception of H. lasiocalycinum ssp. rhiphaeum, whose exine tended to be intermediate between rugulate and striato-reticulate (Fig. 5h), but with wider muri than the typical rugulate found in PPC (Fig. 1c). The third Halimium clade H 1 comprised four taxa of Halimium in the combined analysis (Fig. 4), three belonging to the same species, H. umbellatum (Fig. 5i), and the fourth, H. voldii (Fig. 5k–m), which has been described recently from Greece (Greuter and Raus 2000), could be regarded as a subspecies of H. umbellatum to which it shows similarities (http://www.cistuspage.org. uk/Halimium%20voldii.htm 2010). All these four Halimium taxa have white flowers. We also present a plate showing the pollen of Halimium antiatlanticum (Fig. 5n–p), but we did not find the plant in Morocco, and did not have a sequence to discuss for that taxon. However, on the basis of the samples examined in the Paris (P) and Montpellier (MPU) herbaria, we considered this taxon as a synonym of Halimium halimifolium f. lasiocalycinum (Boiss. & Reut.) Gross. The exine was striato-reticulate (Fig. 5n, p) but with a tendency to be rugulate (Fig. 5p). No. Taxon Ref. P E P/E Shape Exine thickness Exine surface Cistus albidus L. Jean and Pons 1963 48–55 40–46 1.20 Sub prolate 2 lm Rugulate Cistus albidus L. Reille 1990 2–3 lm Microreticulate Cistus albidus L. Saenz de Rivas 1979 53 31 1.71 Prolate 1.4 lm Rugulate 332 Cistus albidus L. This study 47.26 lm ± 2.09 44.09 lm ± 2.15 1.07 ± 0.05 Prolate spheroidal 2.2 lm ± 0.36 Rugulate 1931 Cistus asper Demoly & Mesa This study 59.71 lm ± 3.90 58.83 lm ± 2.90 1.01 ± 0.03 Prolate spheroidal 2.19 lm ± 0.13 Microreticulate 391 Cistus chinamadensis ssp. chinamadensis Bañares & Romero This study 63.6 lm ± 2 61.5 lm ± 2 1.03 ± 0.03 Prolate spheroidal 2.16 lm ± 0.10 Microreticulate 390 Cistus chinamadensis ssp. gomerae Bañares & Romero This study 63.60 lm ± 0.76 60 lm ± 1.12 1.06 ± 0.03 Prolate spheroidal 2.19 lm ± 0.11 Microreticulate 1924 Cistus chinamadensis ssp. ombriosus Demoly & Marrero This study 58.95 lm ± 3.38 58.61 lm ± 2.83 1.01 ± 0.04 Prolate spheroidal 2.17 lm ± 0.17 Microreticulate Cistus clusii Dunal Saenz de Rivas 1979 53 45 1.18 Sub prolate 2.8 lm Striate 330 Cistus clusii Dunal This study 2–2.5 lm Striato-reticulate 365 Cistus clusii ssp. multiflorus Demoly This study 44.50 lm ± 3.17 35.54 lm ± 3.06 1.26 ± 0.10 Sub prolate 2.43 lm ± 0.41 Striato-reticulate Cistus clusii ssp. multiflorus Demoly This study 2–2.5 lm Striato-reticulate Cistus creticus L. This study 36.06 lm ± 1.53 32.54 lm ± 1.30 1.11 ± 0.05 Prolate spheroidal 2.02 lm ± 0.27 Rugulate Cistus creticus L. (= C. incanus L.) Jean and Pons 1963 39–47 32–40 1.19 Sub prolate 3 lm Reticulate 288 Cistus creticus L. (= C. villosus L.) This study 2–2.25 lm Rugulate 289 Cistus creticus L. grp corsicus This study 2.26 lm ± 0.15 Microreticulate 287 Cistus creticus L. grp tauricus (= C. villosus L.) This study 2 lm Microreticulate 280 Cistus creticus L. grp tauricus (= C. villosus L.) This study Cistus crispus L. Jean and Pons 1963 40–41 30–32 1.31 Sub prolate 2 lm Rugulate Cistus crispus L. Saenz de Rivas 1979 42 37 1.14 Prolate spheroidal 1.4 lm Rugulate 336 Cistus crispus L. This study 40.86 lm ± 1.73 39.54 lm ± 1.12 1.03 ± 0.04 Prolate spheroidal 2.41 lm ± 0.32 Rugulate 326 Cistus crispus L. This study 327 Cistus crispus L. This study Cistus heterophyllus Desf. Saenz de Rivas 1979 57.33 lm ± 4.22 45.80 lm ± 3.63 1.03 ± 0.07 Prolate spheroidal Microreticulate Rugulate Rugulate 50 46 1.09 Prolate spheroidal 1.4 lm Rugulate to microreticulate 37 123 366 359 Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) Table 3 Selected palynological characters for Cistaceae taxa studied with combined SEM and LO observations 38 123 Table 3 continued No. Taxon Ref. P E P/E Shape Exine thickness Exine surface 314 Cistus heterophyllus Desf. This study 281 Cistus heterophyllus Desf. ssp. carthaginensis (Pau) Crespo & Mateo This study 76.27 lm ± 9.16 63.07 lm ± 6.72 1.21 ± 0.07 Sub prolate 2.20 lm Microreticulate 2218 Cistus horrens Demoly This study 64.15 lm ± 2.96 57.38 lm ± 2.53 1.12 ± 0.03 Prolate spheroidal 2.33 lm ± 0.14 Microreticulate Microreticulate Cistus horrens Demoly This study 56.36 lm ± 4.18 48.12 lm ± 3.66 1.17 ± 0.09 Sub prolate 2.96 lm ± 0.20 Microreticulate Cistus inflatus Pourr. ex Demoly This study 45.20 lm ± 6.88 41.20 lm ± 5.31 1.09 ± 0.06 Prolate spheroidal 2.45 lm ± 0.37 Microreticulate with smooth supratectal elements 273 Cistus inflatus Pourr. ex Demoly This study 2.5 lm Microreticulate with smooth supratectal elements 321 Cistus inflatus Pourr. ex Demoly This study Microreticulate with smooth supratectal elements 323 Cistus inflatus Pourr. ex Demoly This study Microreticulate with smooth supratectal elements Cistus inflatus Pourr. ex Demoly (= C. hirsutus Lam.) Jean and Pons 1963 48–50 45–48 1.05 Prolate spheroidal 4 lm Reticulate Cistus inflatus Pourr. ex Demoly (= C. psilosepalus Sweet) Màrquez et al.1996 38.25–46.15 35.25–44.18 1.06 Prolate spheroidal 2.5–4 lm Reticulate Cistus inflatus Pourr. ex Demoly (= C. psilosepalus Sweet) Saenz de Rivas 1979 48 46 1.04 Prolate spheroidal 4.2 lm Retipilate Cistus ladanifer L. Jean and Pons 1963 50–61 50–59 1.02 Prolate spheroidal 4 lm Reticulate Cistus ladanifer L. Màrquez et al.1996 31.29–39.47 28.4–38.21 1.06 Prolate spheroidal 2.8–3.2 lm Reticulate Cistus ladanifer L. Saenz de Rivas 1979 51 46 1.11 Prolate spheroidal 4.2 lm Reticulate 333 Cistus ladanifer L. This study 3–4 lm Large reticulate 334 Cistus ladanifer L. ssp. ladanifer f. albiflorus (Dun.) Dans. This study 43.80 lm ± 2.26 41.52 lm ± 2 1.06 ± 0.04 Prolate spheroidal 3.21 lm ± 0.35 Large reticulate 307 Cistus ladanifer L. ssp. mauritianus Pau & Sennen This study 3–4 lm Large reticulate 282 Cistus ladanifer L. ssp. mauritianus Pau & Sennen This study 3–4 lm Large reticulate 305 Cistus ladanifer L. ssp. sulcatus Demoly (= C. palhinhae Ingram) This study 3–4 lm Large reticulate Cistus ladanifer L. ssp. sulcatus Demoly (= C. palhinhae Ingram) Saenz de Rivas 1979 4.2 lm Reticulate 62.3 lm ± 4.5 50 58.1 lm ± 4.1 46 1.07 ± 0.05 1.09 Prolate spheroidal Prolate spheroidal L. Civeyrel et al. 301 320 No. Taxon Ref. P E P/E Shape Exine thickness Exine surface Cistus laurifolius L. Jean and Pons 1963 48–50 47–49 1.02 Prolate spheroidal 4–5 lm Reticulate Cistus laurifolius L. Saenz de Rivas 1979 52 46 1.13 Prolate spheroidal 4.2 lm Reticulate 313 Cistus laurifolius L. This study 2–3 lm Large reticulate 308 Cistus laurifolius L. ssp. atlanticus (Pitard) Sennen & Mauricio This study 48.94 lm ? 9.82 40.04 lm ? 6.41 1.22 ? 0.10 Sub prolate 4.23 lm ± 0.55 Large reticulate Cistus libanotis L. Saenz de Rivas 1979 48 47 1.02 Prolate spheroidal 2.8 lm Striate 315 Cistus libanotis L. This study 45.9 lm ± 6.6 37.4 lm ± 5.6 1.23 ± 0.12 Sub prolate 3.74 lm ± 0.31 Striato-reticulate 357 Cistus libanotis L. f. major n.n. This study 36.40 lm ± 2.20 33.66 lm ± 2.06 1.08 ± 0.04 Prolate spheroidal 3.22 lm ± 0.22 Striato-reticulate Cistus monspeliensis L. Jean and Pons 1963 45–50 43–50 1.02 Prolate spheroidal 3–4 lm Reticulate Cistus monspeliensis L. Reille 1990 4 lm Reticulate Cistus monspeliensis L. Saenz de Rivas 1979 55 51 1.08 Prolate spheroidal 4.2 lm Retipilate 331 Cistus monspeliensis L. This study 37.04 lm ? 1.94 36.94 lm ? 2.02 1 ± 0.03 Spheroidal 2.71 lm ± 0.31 Microreticulate with smooth supratectal elements 303 Cistus monspeliensis L. This study 47.30 lm ? 4.27 47.05 lm ? 1.64 1.01 ? 0.09 Prolate spheroidal 3.76 lm ± 0.30 Microreticulate with smooth supratectal elements 375 Cistus monspeliensis L. This study Microreticulate with smooth supratectal elements Cistus munbyi Pom. This study 42.02 lm ? 2.31 37.14 lm ? 2.49 1 ? 0.05 Spheroidal 2.21 lm ± 0.10 Striato-reticulate Cistus ochreatus Chr. Sm. This study 63.45 lm ? 3.46 57.20 lm ? 1.91 1.11 ? 0.05 Prolate spheroidal 2.36 lm ± 0.33 Microreticulate 311 Cistus ochreatus Chr. Sm. This study 2–3 lm Microreticulate 2494 Cistus osbeckiifolius ssp. tomentosus Bañares & Demoly This study 61.60 lm ? 2.3 56.06 lm ? 1.8 1.10 ? 0.03 Prolate spheroidal 2.27 lm ± 0.16 Microreticulate 341 Cistus osbeckiifolius Webb. ex Christ. This study 47.18 lm ? 2.07 42.76 lm ? 1.76 1.10 ? 0.04 Prolate spheroidal 2.28 lm ± 0.41 Microreticulate 2–3 lm Microreticulate 64.06 lm ? 2.51 59.66 lm ? 3.07 1.08 ? 0.04 Prolate spheroidal 2.24 lm ± 0.12 Microreticulate 310 Cistus osbeckiifolius Webb. ex Christ. This study 1989 Cistus palmensis Bañares & Demoly This study 316 Cistus parviflorus Lam. This study 76.44 lm ? 4.66 66.15 lm ? 5.44 1.16 ? 0.05 Sub prolate 5.17 lm ± 0.53 Large reticulate 317 Cistus parviflorus Lam. This study 48.69 lm ? 3.22 41.19 lm ? 4.30 1.19 ? 0.08 Sub prolate 3.16 lm ± 0.31 Large reticulate 360 Cistus parviflorus Lam. This study 3 lm Striato-reticulate 39 123 309 2248 Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) Table 3 continued 40 123 Table 3 continued No. Taxon Ref. P E P/E Shape Exine thickness Exine surface 322 Cistus parviflorus Lam. This study 3–4 lm Striato-reticulate 312 Cistus parviflorus Lam. (= C. asperrimus) This study 3 lm Large reticulate Cistus populifolius L. Reille 1990 4–5 lm Large reticulate Cistus populifolius L. Saenz de Rivas 1979 57 53 1.08 Prolate spheroidal 4.2 lm Reticulate Cistus populifolius L. This study 48.24 lm ? 2.69 47.20 lm ? 2.75 1.02 ? 0.02 Prolate spheroidal 3 lm ± 0.13 Large reticulate Cistus populifolius L. ssp. populifolius Jean and Pons 1963 40–50 40–46 1.05 Prolate spheroidal 4–6 lm Reticulate Cistus populifolius L. ssp. populifolius Màrquez et al.1996 37.25–44.90 36.28–44.96 1.01 Prolate spheroidal 2.88–4.44 lm Reticulate 272 Cistus populifolius L. ssp. populifolius (= C. narbonensis Rouy & Foucaud) This study 55.68 lm ? 3.27 52.13 lm ? 3.42 1.07 ? 0.05 Prolate spheroidal 3.34 lm ± 1.12 Large reticulate 276 Cistus populifolius L. ssp. populifolius (= C. narbonensis Rouy & Foucaud) This study 3–4 lm Large reticulate 318 Cistus populifolius L. ssp. major (Dun.) Heywood This study 3–4 lm Large reticulate 306 Cistus populifolius L. ssp. major (Dun.) Heywood This study 76.56 lm ? 8.52 71.24 lm ? 6.73 1.07 ? 0.04 Prolate spheroidal 4.78 lm ± 0.19 Large reticulate Cistus pouzolzii Delile Jean and Pons 1963 52–59 45–56 1.12 Prolate spheroidal 2–3 lm Rugulate 351 Cistus pouzolzii Delile This study 47,68 lm ± 3,91 42,58 lm ± 3,90 1.12 ? 0.04 Prolate spheroidal 2.94 lm ± 0.33 Striato-reticulate 350 Cistus pouzolzii Delile This study Cistus salviifolius L. Jean and Pons 1963 48–51 44–50 1.05 Prolate spheroidal 5 lm Cistus salviifolius L. Màrquez et al.1996 45.29–49.22 44.95–48.74 1.01 Prolate spheroidal 5 lm Reticulate Cistus salviifolius L. Reille 1990 5 lm Reticulate Cistus salviifolius L. Saenz de Rivas 1979 58 51 1.14 Prolate spheroidal 4.2 lm Retipilate 383 Cistus salviifolius L. This study 44.06 lm ? 2.51 42.35 lm ? 2.43 1.04 ? 0.05 Prolate spheroidal 3 lm ± 0.07 Microechinate 335 Cistus salviifolius L. This study 356 Cistus sintenisii de Lit. This study 38.94 lm ? 2.38 38.93 lm ? 2.14 1 ? 0.04 Spheroidal 2.76 lm ± 0.38 Striato-reticulate 325 Cistus sintenisii de Lit. This study 329 Cistus symphytifolius Lam. This study 59.9 lm ? 1.66 56.1 lm ? 1.85 1.06 ? 0.02 Prolate spheroidal 2.30 lm ± 0.16 Microreticulate 2462 Cistus symphytifolius Lam. (Pico de Cabras) This study 59.80 lm ? 2.86 55.18 lm ? 2.45 1.08 ? 0.04 Prolate spheroidal 2.29 lm ± 0.17 Microreticulate 338 Striato-reticulate Microechinate L. Civeyrel et al. Striato-reticulate No. Taxon Ref. P E P/E Shape Exine thickness Exine surface 1851 Cistus symphytifolius Lam. (Punta Gorda) This study 58.65 lm ? 2.85 56.72 lm ? 3.01 1.04 ? 0.04 Prolate spheroidal 2.26 lm ± 0.14 Microreticulate 2023 Cistus symphytifolius Lam. var. canus Demoly This study 64.42 lm ? 4.28 57.64 lm ? 4.12 1.12 ? 0.04 Prolate spheroidal 2.19 lm ± 0.51 Microreticulate 2319 Cistus symphytifolius Lam. var. villosus Demoly This study 67.23 lm ? 2.93 61.73 lm ? 2.51 1.09 ? 0.03 Prolate spheroidal 2.45 lm ± 0.38 Microreticulate 343 Halimium antiatlanticum Maire & Wilczek This study 44.40 lm ? 3.54 40.30 lm ? 2.38 1.10 ? 0.06 Prolate spheroidal 2.38 lm ± 0.26 Striato-reticulate 45.46 lm ? 2.19 39.75 lm ? 4.03 1.15 ? 0.09 Sub prolate 2.42 lm ± 0.27 Striato-reticulate 342 Halimium antiatlanticum Maire & Wilczek This study 271 Halimium antiatlanticum Maire & Wilczek This study 2–3 lm Striato-reticulate 275 Halimium atlanticum Humb. & Maire This study 2–3 lm Striato-reticulate 1587 Halimium atlanticum Humb. & Maire This study 54.34 lm ? 2.55 50.82 lm ? 1.62 1.07 ? 0.04 Prolate spheroidal 3.30 lm ± 0.27 Striato-reticulate 1595 Halimium atlanticum Humb. & Maire This study 54.69 lm ? 3.75 50.07 lm ? 3.18 1.09 ? 0.04 Prolate spheroidal 3.40 lm ± 0.24 Striato-reticulate Halimium atriplicifolium (Lam.) Spach JiménezAlbarrán 1984 57.29–55.83 53.56–52.19 1.06 Prolate spheroidal Halimium atriplicifolium (Lam.) Spach Saenz de Rivas 1979 60 53 1.13 Prolate spheroidal 2.8 lm 2.92 lm ± 0.24 290 299 Reticulate Striate Halimium atriplicifolium (Lam.) Spach This study 53.80 lm ? 3.83 45.52 lm ? 4.60 1.19 ? 0.08 Sub prolate Halimium calycinum (L.) K.Koch (= H. commutatum Pau) JiménezAlbarrán 1984 54.85–57.99 47.15–48.97 1.17 Sub prolate Striato-reticulate Halimium calycinum (L.) K.Koch (= H. commutatum Pau) Saenz de Rivas 1979 66 50 1.32 Sub prolate 2.8 lm Striate Striato-reticulate This study 59.16 lm ? 3.51 45.22 lm ? 2.52 1.31 ? 0.08 Sub prolate 2.68 lm ± 0.32 Striato-reticulate Jean & Pons 1963 50–58 40–48 1.23 Sub prolate 4 lm Striato-reticulate Halimium halimifolium (L.) Willk. JiménezAlbarrán 1984 49.98–49.42 44.1–42.92 1.14 Prolate spheroidal Halimium halimifolium (L.) Willk. Saenz de Rivas 1979 43 42 1.02 Prolate spheroidal 2.8 lm Striate 349 Halimium halimifolium (L.) Willk. This study 48.70 lm ? 3.06 43.94 lm ? 2.46 1.11 ? 0.07 Prolate spheroidal 3.46 lm ± 0.73 Striato-reticulate 347 Halimium halimifolium f. lasiocalycinum (Boiss. & Reut.) Gross. This study 45.90 lm ? 2.24 40.25 lm ? 2.38 1.14 ? 0. 06 Prolate spheroidal 346 Halimium halimifolium f. lasiocalycinum (Boiss. & Reut.) Gross. This study Reticulate Striato-reticulate Striato-reticulate 41 123 Halimium calycinum (L.) K.Koch (= H. commutatum Pau) Halimium halimifolium (L.) Willk. Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) Table 3 continued 42 123 Table 3 continued No. Taxon Ref. P E P/E Shape Exine thickness Exine surface 345 Halimium halimifolium ssp. multiflorum auct. non Salzm. This study 49.78 lm ? 3.88 39.30 lm ? 3.84 1.27 ? 0.13 Sub prolate 1.94 lm ± 0.35 Striato-reticulate 1045 Halimium halimifolium ssp. multiflorum auct. non Salzm. This study 54.21 lm ? 2.83 48.09 lm ? 3.00 1.13 ? 0.05 Prolate spheroidal 2.50–2.80 lm Striato-reticulate 344 291 Halimium halimifolium ssp. halimifolium (L.) Willk. This study 44.4 lm ? 2.33 37.22 lm ? 2.50 1.20 ? 0.08 Sub prolate 2.09 lm ± 0.22 Striato-reticulate Halimium lasianthum ssp. alyssoides (Lam.) Greuter & Burdet Jean & Pons 1963 45–50 39–49 1.08 Prolate spheroidal 3 lm Reticulate Halimium lasianthum ssp. alyssoides (Lam.) Greuter & Burdet JiménezAlbarrán 1984 49.39–51.94 44.44–47.74 1.11 Prolate spheroidal Halimium lasianthum ssp. alyssoides (Lam.) Greuter & Burdet Màrquez et al.1996 44.55–49.79 43.36–51.05 1.00 Spheroidal 2.3–3.4 lm Reticulate Halimium lasianthum ssp. alyssoides (Lam.) Greuter & Burdet Saenz de Rivas 1979 50 44 1.14 Prolate spheroidal 2.8 lm Striate Halimium lasianthum ssp. alyssoides (Lam.) Greuter & Burdet This study 47.83 lm ? 2.99 40.17 lm ? 2.99 1.19 ? 0.07 Sub prolate 2.13 lm ± 0.21 Striato-reticulate 53.52 lm ? 3.74 45.56 lm ? 2.75 1.18 ? 0.06 Sub prolate Reticulate 295 Halimium lasianthum ssp. lasianthum (Lam.) Spach This study 2.93 lm ± 0.32 Striato-reticulate 296 Halimium lasianthum ssp. lasianthum (Lam.) Spach This study 2–3 lm Striato-reticulate 394 Halimium lasianthum ssp. lasianthum (Lam.) Spach This study 2–3 lm Striato-reticulate Halimium lasianthum ssp. alyssoides (Lam.) Spach (= H. alyssoides (Lam.) Koch) JiménezAlbarrán 1984 52.58–53.73 42.24–50.3 1.08 Prolate spheroidal Halimium lasiocalycinum ssp. rhiphaeum (Pau & Font Quer) Maire This study 40.36 lm ? 1.25 36.02 lm ? 2.01 1.12 ? 0.05 Prolate spheroidal Halimium ocymoides (Lam.) Willk. JiménezAlbarrán 1984 49.92–51.55 42.39–43.04 1.19 Sub prolate Halimium ocymoides (Lam.) Willk. Màrquez et al.1996 40.36–45.83 38.47–45.88 1.02 Prolate spheroidal 1.9–3.3 lm Reticulate Halimium ocymoides (Lam.) Willk. Saenz de Rivas 1979 52 43 1.21 Sub prolate 2.8 lm Striate 292 Halimium ocymoides (Lam.) Willk. f. elongatum (Vahl.) Gross. This study 2–3 lm Striato-reticulate 293 Halimium ocymoides (Lam.) Willk. f. sampsucifolium (Cav.) Gross. This study 49.08 lm ? 3.95 42.06 lm ? 3.62 1.17 ? 0.06 Sub prolate 2.16 lm ± 0.31 Striato-reticulate Halimium umbellatum (L.) Spach Jean and Pons 1963 52–60 45–52 1.15 Sub prolate 3 lm Reticulate Halimium umbellatum (L.) Spach JiménezAlbarrán 1984 50.6–51.27 46.96–48.08 1.07 Prolate spheroidal 348 Reticulate 2.06 lm ± 0.17 Striato-reticulate Reticulate L. Civeyrel et al. Reticulate No. 298 Taxon Ref. P E P/E Shape Exine thickness Exine surface Halimium umbellatum (L.) Spach Màrquez et al.1996 43.16–51.35 41.51–48.37 1.05 Prolate spheroidal 3–4 lm Striato-reticulate 2.92 lm ± 0.28 Striato-reticulate Halimium umbellatum (L.) Spach This study 55.68 lm ? 5.80 46.64 lm ? 3.91 1.20 ? 0.10 Sub prolate Halimium umbellatum (L.) Spach ssp. viscosum (Willk.) Bolos & Vigo JiménezAlbarrán 1984 52–53.82 46.73–49.28 1.07 Prolate spheroidal Halimium umbellatum (L.) Spach ssp. viscosum (Willk.) Bolos & Vigo Saenz de Rivas 1979 66 54 1.22 Sub prolate 2.8 lm Striate 392 Halimium umbellatum (L.) Spach ssp. viscosum (Willk.) Bolos & Vigo This study 45.00 lm ? 3.38 38.70 lm ? 2.57 1.16 ? 0.06 Sub prolate 2.88 lm ± 0.39 Striato-reticulate 277 Halimium umbellatum (L.) Spach ssp. viscosum (Willk.) Bolos & Vigo This study 2–3 lm Striato-reticulate 2571 Halimium voldii Kit Tan, Perdetzoglou & Raus This study 55.09 lm ? 3.08 56.32 lm ? 2.42 0.98 ? 0.03 Oblate spheroidal 3.30 lm Striato-reticulate Fumana ericoides Pau subsp. montana (Pomel) Güemes & Muñoz Saenz de Rivas 1979 68 68 1.00 Spheroidal 4.2 lm Retipilate Fumana ericoides Pau subsp. montana (Pomel) Güemes & Muñoz Jean and Pons 1963 52–58 59–65 0.89 Oblate spheroidal Helianthemum caput-felis Boiss. Saenz de Rivas 1979 43 32 1.34 Prolate 2.8 lm Striate 42–50 30–35 1.41 Prolate 2 lm Striate Reticulate OUTGROUP Reticulate Helianthemum caput-felis Boiss. This study Helianthemum oelandicum (L.) DC. Jean and Pons 1963 Striato-reticulate Helianthemum oelandicum (L.) DC. subsp. incanum (Willk.) López This study Helianthemum oelandicum (L.) DC. subsp. incanum (Willk.) López (= H. canum (L.) Hornem.) Saenz de Rivas 1979 38 32 1.18 Sub prolate 2.8 lm Striate Helianthemum oelandicum (L.) DC. subsp. incanum (Willk.) López (= H. canum (L.) Hornem.) Màrquez et al.1996 42.88–57.08 33.52–41.81 1.33 Sub prolate 1.22–2.39 lm Striato-reticulate Striato-reticulate This study Saenz de Rivas 1979 66 48 1.37 Prolate 2.8 lm Striato-reticulate Striate Helianthemum syriacum (Jacq.) Dum.-Cours. (= H. lavandulaefolium DC.) Jean and Pons 1963 50–65 45–50 1.21 Sub prolate 3 lm Striate Tuberaria guttata (L.) Raf. Saenz de Rivas 1979 52 38 1.37 Prolate 2.8 lm Striate 43 123 Helianthemum syriacum (Jacq.) Dum.-Cours. Helianthemum syriacum (Jacq.) Dum.-Cours. (= H. lavandulaefolium DC.) Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) Table 3 continued 1.07 40–42 43–45 This study Jean and Pons 1963 Tuberaria guttata (L.) Raf. Tuberaria guttata (L.) Raf. 123 Pollen grains examined for this study were compared with previous studies (Ref.). For each sample are given, when known, the polar axis (P) and equatorial diameter (E), the P/E ratio, the shape of the pollen from P/E, and observations on exine thickness and surface. When pollen samples were examined by SEM, only an approximation of the thickness is given, on the basis of one measurement only. For outgroups, no SEM was carried out and interpretation is given only when our opinion differs from the literature. We have included some synonymy when the name found on the herbarium sample was different from the commonly accepted name and also in order to test pollen variation within the species Striato-reticulate Prolate spheroidal 3.5 lm Exine thickness Shape P/E E P Ref. Taxon No. Table 3 continued Striate L. Civeyrel et al. Exine surface 44 Pollen morphology of Cistus The WWPC (Fig. 4) exhibited the highest diversity of exine ornamentation for all the clades discussed in this study, with some unique patterns not found elsewhere. The large shrub-like species Cistus ladanifer, C laurifolius, and C. populifolius had the unique, largely reticulate, exine ornamentation (Fig. 6a–h) shared by some samples of C. parviflorus, a much smaller species (Fig. 6I). The subspecies of Cistus ladanifer were examined for this study. C. ladanifer ssp. mauritianus (Fig. 6a–c), occurring in North Africa and southern Spain, showed no significant palynological difference from C. ladanifer subsp. ladanifer or from C. ladanifer subsp. sulcatus, which is regarded by some authors as a different species, C. palhinhae Ingram. To avoid any confusion about the pollen morphology of C. ladanifer subsp. sulcatus (C. palhinhae) we examined specimens collected by Ingram himself (pollen sample 305). The pollen characters of Cistus laurifolius ssp. atlanticus (Fig. 6d–f) were morphologically comparable with those of the subspecies C. laurifolius ssp. laurifolius. The same observation was made for subspecies of C. populifolius. Although authors have agreed on the ornamentation for all species but C. parviflorus, we followed Reille (1990) for characterization of this type of sculpture and defined it as largely reticulate. The shape of most of the taxa observed was prolate spheroidal with minor variations. The exine was very thick, usually 3–4 lm. Cistus parviflorus was not examined in previous studies and had some interesting variation. Five samples of this species were studied, three had a very large reticulum (Fig. 6i–l), but samples 360 and 322 had a striato-reticulate exine (Fig. 6j, m–o). It could have been interesting to compare their DNA with those of other species but the two samples came from very old herbarium specimens (Table 1) and we only had one DNA accession. Four species of Cistus had a striatoreticulate exine but were not grouped together. Two species had not been studied before within this group, Cistus munbyi (Fig. 7a–b) and C. sintenisii (Fig. 7c–d). Cistus pouzolzii has been described as rugulate by Jean and Pons (1963). Our sample from North Africa (Fig. 7e) had a slightly more serrated exine that could be regarded by some authors as relaxed rugulate compared with the French sample (Fig. 7f), but we definitely regarded it as striato-reticulate exine. Three species had supratectal ornamentation—Cistus salviifolius with microechinate elements (Fig. 1i) and C. monspeliensis and C. inflatus with smooth elements (Fig. 1g–h). For these species there was no difference in interpretation, we only used a different term to define their exine. Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) 45 Fig. 5 SEM micrographs of pollen grains. a–c Halimium altlanticum 1595, a pollen grain in polar view, b equatorial view, c striatoreticulate exine. d–e H. halimifolium ssp. multiflorum 1045, d pollen grain in polar view, e striato-reticulate exine. f–h H. lasiocalycinum ssp. rhiphaeum 348, f equatorial view, g aperture, h striato-reticulate exine. i H. umbellatum ssp. viscosum 392, striato-reticulate exine. j H. atriplicifolium 290 striato-reticulate exine. k–m H. voldii 2571, k pollen grain in equatorial view, l polar view, m striato-reticulate exine. n–o H. antiatlanticum 271, p H. antiatlanticum 343, n, p striato-reticulate exine, o pollen grain in polar view and equatorial view At the base of the ‘‘purple pink Cistus clade’’ (hereafter PPC), we found Cistus crispus and a clade formed by three species C. albidus, C. creticus, and C. heterophyllus with a rugulate to microreticulate exine. C. crispus (Fig. 7g) and C. albidus L. (Figs. 1c, 7q) have a typically rugulate exine, except for Reille (microreticulate, 1990), with lumina 123 46 L. Civeyrel et al. Fig. 6 SEM micrographs of pollen grains. a–c Cistus ladanifer var petiolatus 282, a prolate spheroidal pollen grain in polar view, b in equatorial view, c largely reticulate exine. d–f C. laurifolius ssp. atlanticus 308. d prolate spheroidal pollen grain in polar view, e equatorial view, f largely reticulate exine. g–h C. populifolius ssp major. g 306 largely reticulate exine, h 318 pollen grain in polar view. i–l C. parviflorus 316, i largely reticulate exine, k pollen grain in equatorial view, l polar view. m–o C. parviflorus 360. m striatoreticulate exine. o pollen grain in equatorial view. j, n C. parviflorus 322, j pollen grains in equatorial and polar view, n striato-reticulate exine under 0.5 lm. We examined two taxa of Cistus heterophyllus, a North African sample (Fig. 7h) and the rare C. heterophyllus subsp. carthaginensis (Figs. 1b, 7i–j) from Spain, both have a microreticulate exine. We examined five samples belonging to C. creticus, from Corsica (Fig. 7n–o) to others in Crimea and Morocco, the exine was rugulate in 123 Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) 47 Fig. 7 SEM micrographs of pollen grains. a–b Cistus munbyi 309, a spheroidal pollen grain in equatorial view, b striato-reticulate exine. c–d C. sintenisii 325, c spheroidal pollen grain in polar view, d striato-reticulate exine. e–f C. pouzolzii striato-reticulate exine. e 351 from Morocco, f 350 from France. g C. crispus 336 pollen grain in polar view. h C. heterophyllus 314 pollen grain in equatorial view. i–j C. heterophyllus subsp. carthaginensis 281. i pollen grain in equatorial view, j polar view. k C. heterophyllus 314 microreticulate exine. l C. creticus 359 prolate spheroidal pollen grain in equatorial view. m C. creticus 288 rugulate exine. n–o C. creticus var. corsicus 289, n prolate spheroidal pollen grain in equatorial view, o microreticulate exine. p C. creticus var. tauricus 280 microreticulate exine. q C. albidus 332 prolate spheroidal pollen grain in equatorial view r–s C. asper 1931 r prolate spheroidal pollen grain in polar view. s microreticulate exine three cases (Fig. 7m) and microreticulate in the other two (Fig. 7o–p). There was no geographical pattern for exine ornamentation. Obviously, there is a continuum of state for this character. Discrepancies between authors show that they had the same problem of continuum between the two states (Table 3). The shapes for these four species were not 123 48 L. Civeyrel et al. very consistent and varied from prolate to prolate spheroidal with an exine thickness below 2.5 lm. The Canarian taxa that formed a monophyletic group were also characterised by a micro-reticulate exine sometimes bordering on rugulate (Figs. 7, 8, 9). None of these had been studied before. The exine thickness was usually less than 2.5 lm (Fig. 8d), except for Cistus horrens (No. 301, Table 3). The shape was prolate spheroidal, except for the same sample Cistus horrens (No. 301, Table 3). Fig. 8 SEM micrographs of pollen grains. a–b Cistus chinamadensis subsp. chinamadensis 391. a prolate spheroidal pollen grain in equatorial view. b microreticulate exine. c–e C. chinamadensis subsp. gomerae 390. c pollen grains. d microreticulate exine. e exine thickness. f–g C. chinamadensis subsp. ombriosus 1922. f microreticulate exine. g pollen grain in equatorial view. h–i C. horrens 301. h subprolate pollen grain in equatorial view. i microreticulate exine. j–k C. ochreatus 311. j pollen grain in polar view. k microreticulate exine. l–m C. osbeckiifolius subsp. osbeckiifolius 341. l pollen grain in equatorial view. m microreticulate exine. n–o C. osbeckiifolius subsp. tomentosus 2494 n pollen grain in polar view. o microreticulate exine. p–s C. palmensis 1989. p pollen grain in equatorial view. q pollen grain in polar view. r aperture. s microreticulate exine 123 Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) 49 Fig. 9 SEM micrographs of Cistus symphytifolius pollen grains. a–b C. symphytifolius var. canus 2023. a prolate spheroidal pollen grain in equatorial view, b microreticulate exine. c–d C. symphytifolius (Pico de Cabra) 2462, c prolate spheroidal pollen grain in equatorial view, d microreticulate exine. e–f C. symphytifolius (Punta Gorda) 1851, e prolate spheroidal pollen grain in equatorial view, f microreticulate exine. g–h C. symphytifolius var. villosus 2319, g microreticulate exine, h prolate spheroidal pollen grain Discussion Systematics Molecular analyses The white and whitish pink Cistus clade had the weakest support, and had some interesting particularities. It includes some subspecies or varieties on both sides of the Gibraltar strait and is very heterogenous with regard to sections, no fewer than six sections have been defined within this group. We examined the species with varieties or subspecies distributed on both sides of the Gibraltar strait. Cistus ladanifer subsp. mauritianus Pau and Sennen, Cistus laurifolius L. ssp. atlanticus (Pitard) Sennen and Mauricio, Cistus populifolius var. major Dun. are the Moroccan counterpart subspecies or varieties of the European species. Within each species, the sequences were identical for all taxa except for Cistus populifolius var. major Dun. which had two mutations not present on Cistus populifolius L. var. populifolius. There were two small clades with good BS, Cistus parviflorus Lam. and Cistus sintenisii de Lit., which are two species from the east of the Mediterranean, and Cistus clusii Dunal with Cistus aff. Munbyi. The six generic sections were Ledonella Dun., Stephanocarpus (Spach.) Gren. and Godr., Stephanocarpoidea Rouy. and Foucaud, Ledonia Dun., Ladanium (Spach.) Gren. and Godr., and Halimioides Wilk.. The section Ledonella was monospecific with C. parviflorus only. This species was grouped by Dansereau (1939) with the other pink-flowered species (subgenus Erythrocistus), but this is not supported by our analyses and cannot be maintained. The section Stephanocarpus comprised two species, C. monspeliensis and C. sintenisii, but this grouping was not supported either. These two sections seemed to be quite If we compare the plastid region analysed, the phylogeny obtained with trnS-G was better resolved than that with only trnL-F. Some clades were present in both analyses with quite good support: the three Halimium clades, H 1, H 2, and H 3, and the PPC clade. The rest of the Cistus formed a clade with trnS-G only, not with trnL-F. Both plastid regions contributed to the better resolution of the strict consensus tree of the combined analysis. The Halimium species did not group together but instead formed three monophyletic groups, followed by a group of Cistus, mainly with white flowers, and the purple pink Cistus clade (Fig. 3). Petal colour, which was mapped on one of the single gene analyses (Fig. 2a), defined three monophyletic groups for the Halimium species, two yellow-flowered and one white-flowered. For Cistus, the two monophyletic groups were also quite well defined by their colour, the purple pink flowers being found grouped in one clade, and the only other pinkish species was very pale in comparison, and was grouped with the white species of Cistus. There were several recognised monophyletic clades for the ingroup and all but one had high BS, the ‘‘white and whitish pink Cistus’’ group being the weakest. Neither of the two genera examined formed a monophyletic group. The two clades of Cistus never grouped together and the situation was even worse for the Halimium species, where the very long length of the branches prevented any grouping. 123 50 artificial. Moreover, C. parviflorus and C. sintenisii formed a well supported group in the combined analysis (Fig. 3). These two species are eastern Mediterranean species and fertile hybrids between the two are easily produced. The two species share, with C. monspeliensis also, a stephanocapsule. C. pouzolzii belonged to the monospecific section Stephanocarpoidea and was grouped with C. monspeliensis but with no support. A long style characterizes C. pouzolzii, and this character is found in the purple species also, but, because no hybrids have been recorded between these two groups, they are probably genetically distant. It seems for the moment that this is simple convergence. The section Ledonia, which comprises three species, C. salviifolius, C. inflatus, and C. populifolius, was paraphyletic in our study. Two of the tallest species of Cistaceae, C. ladanifer and C. laurifolius, belong to the section Ladanium, sharing a reduced number of sepals (3). Again this section was not monophyletic, because of the position of C. salviifolius at the base of the clade formed by the subspecies of C. ladanifer. The character common to C. salviifolius and the other two species of the section Ladanium is the paucity of the numbers of flowers per inflorescence; in C. ladanifer they are even completely unifloral. However, Dansereau (1941) noted that two forms of C. salviifolius can be found in the wild, some trifloral (pure C. salviifolius) and some with many more flowers per inflorescence (probably introgressed with C. monspeliensis). The last section of this group of Cistus is Halimioides with the remaining three species C. libanotis, C. clusii, and C. munbyi. C. libanotis was not grouped with any other species and the other two taxa, C. clusii and C. aff. munbyi, had an intermediate position between the Halimium and the rest of the Cistaceae. Within the purple pink Cistus clade, it was easy to distinguish two groups belonging to two Cistus sections. First we found the four species of the section Cistus which were paraphyletic (C. crispus, C. creticus, C. albidus, and C. heterophyllus), and second a monophyletic group of Canarian taxa which all belonged to the section Macrostylia Willk. The four species of the subgenus Cistus and those of the subgenus Macrostylia do not hybridize freely with any other section of Cistus. When artificial hybridization is carried out, the progeny is sterile and very weak, and does not survive for long (Demoly 1996). Comparison of Cistus albidus and C. creticus Within the section Cistus, the distribution of Cistus albidus and Cistus creticus is rather interesting. These species overlap very little around the Mediterranean. Cistus creticus occurs mainly in the eastern part of the Mediterranean basin whereas C. albidus is found only in the western part. 123 L. Civeyrel et al. However, the two species can be found in sympatry in some parts of Italy and Morocco. There is some evidence that the distribution of C. creticus might not have always been the same in the past and has been different from what we find now. When the work for Flora iberica (Demoly and Montserrat 1993) was started, some very old herbarium specimens of Cistus creticus from Portugal and Catalonia were discovered. Most of those stations have now disappeared but one relictual population of C. creticus has since been found in the Albacete province and is now under protection. For some authors C. creticus comprises three subspecies, Cistus creticus subsp. eriocephalus, C. creticus subsp. corsicus, and C. creticus subsp. creticus (Greuter et al. 1984) with some geographical structuring, but the areas of these taxa are not separated (Falchi et al. 2009; Paolini et al. 2009). It can be wondered whether C. creticus and C. albidus might be two sub-units of a large ancestral species that has been fragmented into two geographical entities with a limited outcrossing barrier between them, or whether they are two ancient species that are now differentiated enough to be able to remain distinct, even when they are in sympatry. Their sequences were very similar, with few differences, and it would be useful to see how they behave where they grow in sympatry. Ecologically, these two species share some interesting similarities. They are the only taxa of the purple pink clade able to grow in chalky soil. On the whole, most Cistus species are found on acidic rather than basic soil. Some ecological differences can be found between the two species, C. albidus being less dynamic in acidic soil and it is also more sensitive to cold. In Morocco it is found only up to 1,100 m whereas C. creticus can reach 2,100 m (Raynaud 1992). In Spain it has been collected up to 1,300 m (Grant et al. 2006) and in Europe its altitudinal limit is 1,400 m (Tutin et al. 1968). When these two species are in sympatry, some fertile hybrids can be found (C. 9 canescens Sweet), but they do not take the place of their parents, nor do they seem to backcross preferentially with one of the parents, producing a large hybrid population with a wide spectrum of intermediate characters (Dansereau 1940). Despite these hybridization events, which are a frequent phenomenon within Cistaceae, these two taxa are considered as two distinct species and the hybrids do not take over because they are probably less adapted to their environment than the two parents. In Corsica and Sardinia the differentiation of C. creticus is correlated with geology. The distribution of haplotypes separates them into two groups, one found on granite and the other on schist. The accumulation of mutations between these two groups indicates that this species has been there for a long time (Falchi et al. 2009). Heterogeneity of Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) habitat also seems to be involved in the genetic diversity of C. albidus (Grant et al. 2006). The Canarian taxa The purple pink-flowered Cistus reach their maximum endemicity in the Canary Islands. Only four purple pinkflowered taxa are found outside this archipelago. The Canary Islands form a volcanic archipelago, with seven main islands, which have been dated from 20 to 21 million years (Fuerteventura) to less than 1 MA (El Hierro) (Emerson 2003; Garcia-Talavera 1999). Cistus species are only present on the five western islands. Fuerteventura and Lanzarote, which are among the older Islands and also the closest to the continent, do not have any Cistus species on them. They also lack forest habitats, and are of low elevation. In contrast, Gran Canaria, which has both forest and high elevation, and is also the oldest, ranging in age from 14 to 16 MA (Emerson 2003; Hoernle et al. 1991), is the first where Cistus species are encountered. Four species occur on Gran Canaria, C. ochreatus, the ubiquitous C. monspeliensis, which is also found on all five islands (Barquı́n-Diez and Voggenreiter 1988), a new recently described species C. grancanariae Marrero-Rodr., Almeida & C. Rı́os (Marrero-Rodrı́guez et al. 2008), belonging to the subgenus Leucocistus (not included in our sampling), and C. horrens which was the sister species to the other Canarian taxa on the combined molecular tree. The plant Fig. 10 The combined molecular tree, restricted to Canarian taxa, is mapped on the Canary Islands. The branches of the tree end at the place from which plants of group A or group B were collected. 51 called C. symphytifolius var. leucophyllus (Spach), which is restricted to Gran Canaria, has been given species status and is now C. ochreatus. Previous studies showed, on the basis of allozymes, that it was already distinct from other C. symphytifolius varieties (Batista et al. 2001). Of the Canarian taxa, C. horrens seems to be the species ‘‘best adapted’’ to xeric conditions (Demoly 2004). Two polytomies, A and B, were found (Fig. 10). The taxa present in polytomy B were mainly from the south of the three islands on which they occur (La Gomera, Tenerife, and La Palma) with the exception of C. symphytifolius Puntagorda which occurs on the northeastern part of La Palma. The members of the polytomy A (Fig. 10) were mainly from the north, east, or west of three islands (El Hierro, Tenerife, and Gran Canaria). Cistus symphytifolius is a widespread species (Batista et al. 2001). We sampled five C. symphytifolius taxa showing some variability and they did not group together, even on the same island, Tenerife, which is not surprising, because previous studies showed that little gene flow occurred between populations (Batista et al. 2001). Cistus chinamadensis comprises three subspecies, each constituted by a single population on a single island, and each of which is critically endangered (Bañares et al. 2008; Moreno 2008). They did not group together; the same pattern was found by Batista et al. (2001) with the two subspecies known at that time. The two subspecies of C. osbeckiifolius occurring on Tenerife were also in different clades. Because we only used chloroplast data further investigations are needed. Numbers in brackets refer to geological ages estimated in millions of years (Emerson 2003) 123 52 L. Civeyrel et al. The three Halimium clades Morocco and Cistuses In the combined analysis, clade H 1 was between the PPC clade and the WWPC clade, but this position was not supported at all by bootstrap. Two species belonged to this clade: Halimium umbellatum and the recently described species H. voldii from Greece (Tan and Iatrou 2001); both are members of the section Halimium of that genus. It is the only group of Halimium with white flowers and it shares some morphological characters with Cistus of the section Halimioides: white flowers, linear leaves with an inwardly rolled margin, and an umbellate inflorescence (also present in C. laurifolius and C. libanotis). Clades H 2 and H 3 were characterised by only yellow flowers and two sections of the genus Halimium, Chrysorhodion and the monospecific section Commutati (with only H. calycinum). Because of the position of Halimium atlanticum grouped with H. calycinum, the section Chrysorhodion was not monophyletic and either H. atlanticum should be transferred to the section Commutati or only a single section should be kept. In H 2, two wellsupported groups can be morphologically distinguished. The two species H. atriplicifolium and H. lasianthum have homogeneous leaves whereas the rest of the group have two kinds of leaves depending on whether the shoot is sterile or fertile. Sterile stems have petiolated and trinerved leaves. On shoots bearing flowers, H. halimifolius H. lasiocalycinum, and H. halimifolium ssp. multiflorum have sessile and uninerved leaves, and H. ocymoides has petiolated and one-nerved leaves. H. lasiocalycinum is often considered by botanists as a distinct species but there is only one base pair of divergence with H. halimifolius and nothing with H. halimifolium ssp. multiflorum, Both H. lasiocalycinum and H. halimifolium ssp. multiflorum should be regarded as varieties or subspecies of H. halimifolius as already suggested by some authors (Ball 1877; Grosser 1903). Halimium atlanticum in group H 3, has always been regarded as a close relative of H. lasianthum, but this was not confirmed by our analyses. There are 24 mutations between H. atlanticum and H. calycinum. These two species have a very divergent ecology, H. atlanticum is found above 1,500 m in the Rif in northern Morocco whereas the second is a species found on coastal sand dunes in Portugal, Spain, and the north of Morocco. Their distributions never overlap. They share three morphological characters: revolute leaf margins, pauciflorous inflorescences, and small size, but many other characters distinguish them. We included two samples of H. atlanticum, H. atl. 1587 and H. atl. 1595 collected on two different mountains. They were hardly distinguishable but they differed by three mutations, which is a lot more than between some species, for example C. osbeckiifolius or C. chinamadensis. Morocco is the place where the most species are found for both Cistus (12) and Halimium (7) (Fennane et al. 1999). This country has a complex geography with numerous mountains that originated from paleo-islands within the Tethys (Dercourt et al. 1992), lately connected when the Mediterranean sea was formed. Because of its diverse environments and complex palaeohistory, Morocco is a species-rich country with a high percentage of endemic taxa. The common ancestor of the Canarian species probably originated from Morocco, first because it is the nearest continental land to the Canary Islands and, second, because all the closest three relatives (Cistus heterophyllus, C. albidus, and C. creticus) of the Canarian section Macrostylia still occur in mainland Morocco. 123 Acknowledgments We are most grateful to curators or keepers of the herbaria G, K, MPU, and P for permitting the examination of specimens and removal of pollen samples. We should like to thank Peter Biggins from CIRAD for revising the English, and also all colleagues of the palynology unit at the R. B. G. Kew for their assistance, the Conservatoire Botanique National de Brest and Robert Page for their help, and Olivier Filippi for cultivating our plants collected on field trips, especially from Morocco and the Canary Islands. We also wish to thank Elise Van Campo, Laurent Chibret, and Renaud Lahaye for their support. References Alverson WS, Karol KG, Baum DA, Chase MW, Swensen SM, Mccourt R, Sytsma KJ (1998) Circumscription of the Malvales and relationships to other Rosidae: evidence from rbcL sequence data. Am J Bot 85:876–887 Arrington JM, Kubitzki K (2003) Cistaceae. In: Kubitzki K, Bayer C (eds) The families and genera of vascular plants. Malvales, Capparales and non-betalain Caryophyllales, vol 5. Springer, Berlin Ball J (1877) Spicilegium Florae Marocannae. J Linn Soc Bot 16:343–348 Bañares Á, Blanca G, Güemes J, Moreno JC, Ortiz S (eds) (2008) Atlas y Libro Rojo de la Flora Vascular Amenazada de España. Adenda 2008. Dirección General de Medio Natural y Polı́tica Forestal (Ministerio de Medio Ambiente, y Medio Rural y Marino-Sociedad Española de Biologı́a de la Conservación de Plantas). Madrid Barquı́n-Diez E, Voggenreiter V (1988) Prodromus del atlas fitocorológico de las Canarias occidentales. Parte I: Flora autóctona y especies de interés general Batista F, Bañares A, Caujapé-Castells J, Carqué E, Marrero-Gómez M, Sosa PA (2001) Allozyme diversity in three endemic species of Cistus (Cistaceae) from the Canary Islands: intraspecific and interspecific comparisons and implications for genetic conservation. Am J Bot 88:1582–1592 Bayer C, Chase MW, Fay MF (1998) Muntingiaceae, a new family of dicotyledons with malvalean affinities. Taxon 47:37–42 Borsch T, Quandt D (2009) Mutational dynamics and phylogenetic utility of noncoding chloroplast DNA. Plant Syst Evol 282:169–199 Molecular systematics, character evolution and pollen morphology of Cistus and Halimium (Cistaceae) Civeyrel L, Le Thomas A, Ferguson IK, Chase MW (1998) Critical reexamination of palynological characters used to delimit Asclepiadaceae in comparison to the molecular phylogeny obtained from plastid matK sequences. Mol Phylogenet Evol 9:517–527 Clegg MT, Zurawski G (1992) Chloroplast DNA and the study of plant phylogeny: present status and future prospects. In: Soltis DE, Soltis PS, Doyle JJ (eds) Molecular systematics of plants. Chapman and Hall, pp 1–13 Dansereau P (1939) Monographie du genre Cistus. Boissiera 4:1–90 Dansereau P (1940) Études sur les hybrides de Cistes. Ann Épiphyt 6 (Ser. 2):7–26 Dansereau P (1941) Etudes sur les hybrides de cistes—IV. Corrélation des caractères du C. salviifolius L: VI. Introgression dans la section Ladanium. Anim Biodivers Conserv 38:59–67 Demoly JP (1996) Les hybrides binaires rares du genre Cistus L. (Cistaceae). Anales Jard Bot Madrid 54:241–254 Demoly JP (1998) Notes taxonomiques sur le nothogenre 9 Halimiocistus Janch. (Cistaceae). J Bot Soc Bot France 6:31–38 (9 Halimiocistus humilis Demoly) Demoly JP (2004) Description d une espèce confondue dans le passé avec Cistus symphytifolius Lam. et, parfois, avec C. ocreatus Chr. Sm. Acta Bot Gallica 151:231–232 Demoly JP, Montserrat P (1993) Cistus. In: Castroviejo S et al (eds) Flora Iberica, vol 3, pp 319–337. Consejo de Investigaciones Cientifı́cas, Madrid Dercourt J, Bassoullet JP, Baud A, Butterlin J, Camoin G, Cavelier C, Cecca F, Enay R, Fourcade E, Guiraud R, Lorenz C, Marcoux J, Masse JP, Orszag F, Philip J (1992) Paeloenvironmental atlas of the Tethys from Permian to recent. In: 29th International geological congress, Kyoto, Japan Downie SR, Palmer JD (1992) Use of chloroplast DNA rearrangements in reconstructing plant phylogeny. In: Soltis DE, Soltis PS, Doyle JJ (eds) Molecular systematics of plants. Chapman and Hall, pp 14–35 Emerson BC (2003) Genes, geology and biodiversity: faunal and floral diversity on the island of Gran Canaria. Anim Biodivers Conserv 26:707–716 Erdtman G (1943) An introduction to pollen analysis. Waltham, Massachusetts Erdtman G (1952) Pollen morphology and plant taxonomy, angiosperms. Almqvist and Wiksell, Stockholm Erdtman G (1960) The acetolysis method. A revised description. Svensk Bot Tidskr 54:561–564 Falchi A, Paolini J, Desjobert JM, Melis A, Costa J, Varesi L (2009) Phylogeography of Cistus creticus L. on Corsica and Sardinia inferred by the trnL-F and rpl32-trnL sequences of cpDNA. Mol Phylogenet Evol 52:538–543 Farris JS (1969) A successive approximations approach to character weighting. Syst Zool 18:374–385 Fennane M, Ibn Tattou M, Mathez J, Ouyahya A, El Oualidi J (eds) (1999) Flore Pratique du Maroc, vol 1. Trav Inst Sci 36:I–XV Garcı́a-Talavera F (1999) La Macaronesia. Consideraciones geológicas, biogegráficas y paleoecológicas. In: Fernández-Palacios JM, Bacallado JJ, Belmonte JA (eds) Ecologı́a y cultura en Canarias, 39–64. Museo de la Ciencia, Cabildo Insular de Tenerife, Santa Cruz de Tenerife, Canary Islands, Spain Gaskin JF, Schaal BA (2003) Molecular phylogenetic investigation of US invasive Tamarix. Syst Bot 28:86–95 Grant OM, McNeilly T, Incoll LD (2006) Genetic diversity of Cistus albidus in south-east Spain does not relate to mesoclimate. Funct Pl Biol 33:247–255 Greuter W, Raus T (2000) Med-Checklist Notulae, 19. Willdenowia 30:232 Greuter W, Burdet HM, Long G (1984) Med Checklist. A critical inventory of vascular plants of the circum-Mediterranean 53 countries. Conservatoire et Jardin Botanique de la Ville de Genève, Genève Grosser W (1903) Cistaceae. In: Engler A (éd) Das Pflanzenreieh, Regni vegetabilis conspectus. vol 14, p 161 Guzmán B, Vargas P (2005) Systematics, character evolution, and biogeography of Cistus L. (Cistaceae) based on ITS, trnL-trnF, and matK sequences. Mol Phylogenet Evol 37:644–660 Guzmán B, Vargas P (2009) Historical biogeography and character evolution of Cistaceae (Malvales) based on analysis of plastid rbcL and trnL-trnF sequences. Org Divers Evol 9:83–99 Guzmán B, Lledó MD, Vargas P (2009) Adaptive radiation in mediterranean Cistus (Cistaceae). PLoS ONE 4(7):e6362. doi: 10.1371/journal.pone.0006362 Hamilton MB (1999) Four primer pairs for the amplification of chloroplast intergenic regions with intraspecific variation. Mol Ecol 8:521–523 Hesse M, Halbritter H, Zetter R, Weber M, Buchner R, FroschRadivo A, Ulrich S (2009) Pollen terminology. An illustrated handbook, VI. Springer, Wien Heydacker F (1963) Les types polliniques de la famille des Cistaceae. Pollen Spores 5:41–50 Hoernle K, Tilton G, Schminke HU (1991) Sr–Nd–Pb isotopic evolution of Gran Canaria: evidence for shallow enriched mantle beneath the Canary Islands. Earth Planet Sci Lett 106:44–63 Iversen J, Troels Smith J (1950) Pollenmorfologiske definitioner og typer. Danmarks Geol Unders 3:1–54 Janchen E (1925) Cistaceae. In: Engler A, Prantl K (eds) Die Naturlichen Pflanzenfamilien, 2nd edn., Bd. 21:289–313. W Engelmann, Leipzig Jean MT, Pons A (1962) Une clef de détermination palynologique pour les Cistacées de la flore de France. Naturalia MonspFasc 14:87–92 Jean MT, Pons A (1963) Contibution à l’étude palynologique des Cistacées de la flore de France. Ann Sci Nat Bot Paris 4:159–204 Jiménez-Albarrán MJ (1984) Estudio palinologico del Genero Halimium (Dun.) Spach (Cistaceae). Anal Asoc Pal Lengua Esp 1:21–29 Kress WJ, Prince LM, Williams KJ (2002) The phylogeny and a new classification of the gingers (Zingiberaceae): evidence from molecular data. Am J Bot 89:1682–1696 Lahaye R, Klackenberg J, Källersjö M, Van Campo E, Civeyrel L (2007) Phylogenetic relationships within Secamonoideae (Apocynaceae s.l.) based on four chloroplast sequences. Ann Mo Bot Gard 94:378–396 Márquez J, Diaz-Losada E, Rodrı́guez-Gracia V, Suàrez-Cervera M (1996) Cistaceae. In: Atlas de Polen de Galicia, pp 113–127, Ourense Marrero-Rodrı́guez Á, Almeida R, Rı́os C (2008) Cistus grancanariae sp. nov. (Cistaceae) una nueva especie para Gran Canaria (Islas Canarias). Bot Macaronésica 27:73–88 Médail F, Quézel P (1997) Hot-spots analysis for conservation of plant biodiversity in the mediterranean basin. Ann Mo Bot Gard 84:112–127 Moreno JC (coord) (2008) Lista Roja 2008 de la flora vascular española. Dirección General de Medio Natural y Polı́tica Forestal (Ministerio de Medio Ambiente, y Medio Rural y Marino, y Sociedad Española de Biologı́a de la Conservación de Plantas), Madrid Mort ME, Archibald JK, Randle CP, Levsen ND, O’Leary TR, Topalov K, Wiegand CM, Crawford DJ (2007) Inferring phylogeny at low taxonomic levels: utility of rapidly evolving cpDNA and nuclear ITS loci. Am J Bot 94:173–183 Nandi OI (1998) Floral development and systematics of Cistaceae. Plant Syst Evol 212:107–134 Olson ME (2002) Intergeneric relationships within the CaricaceaeMoringaceae clade (Brassicales) and potential morphological 123 54 synapomorphies of the clade and its families. Int J Plant Sci 163:51–65 Palacios-Chávez R, Quiroz-Garcı́a DL, Arreguı́n-Sánchez MDLL (1999) Morfologı́a de los granos de polen de la familia Cistaceae de México. Acta Botanica Mexicana 49:1–14 Paolini J, Falchi A, Quilichini Y, Desjobert JM, Cian MC, Varesi L, Costa J (2009) Morphological, chemical and genetic differentiation of two subspecies of Cistus creticus L. (C. creticus subsp. eriocephalus and C. creticus subsp. corsicus). Phytochemistry 70:1146–1160 Punt W, Blackmore S, Nilsson S, Le Thomas A (1994) Glossary of pollen and spore terminology. LPP Foundation (Laboratory of Paleobotany and Palynology), Utrecht Punt W, Hoen PP, Blackmore S, Nilsson S, Le Thomas A (2007) Glossary of pollen and spore terminology. Rev Palaeobot Palynology 143:1–81 Quandt D, Müller K, Stech M, Hilu KW, Frey W, Frahm JP, Borsch T (2004) Molecular evolution of the chloroplast trnL-F region in land plants. Monogr Syst Bot Mo Bot Gard 98:13–37 Raynaud C (1992) Cistaceae. In: Fennane M, Mathez J (eds) Elements pour la Flore pratique du Maroc. Naturalia Monsp, sér. Bot. 56:154–158, 171–220 Reille M (1990) Leçons de Palynologie et d’Analyse pollinique. Editions du CNRS, Paris 123 L. Civeyrel et al. Saenz De Rivas C (1979) Pollen morphology of Spanish Cistaceae. Grana 18:91–98 Shaw J, Lickey EB, Beck JT, Farmer SB, Liu WS, Miller J, Siripin KC, Winder CT, Schilling EE, Small RL (2005) The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. Am J Bot 92:142–166 Shaw J, Lickey EB, Schilling EE, Small RL (2007) Comparison of whole chloroplast genome sequences to choose noncoding regions for phylogenetic studies in angiosperms: the tortoise and the hare III. Am J Bot 94:275–288 Swofford DL (2002) PAUP: phylogenetic analysis using parsimony. Version 4.0b10 for Macintosh. Sinauer Associates, Sunderland Taberlet P, Gielly L, Pautou G, Bouvet J (1991) Universal primers for amplification of three non-coding regions of the chloroplast DNA. Pl Mol Biol 17:1105–1109 Tan K, Iatrou G (2001) Endemic plants of Greece—the Peloponnese (Halimium voldii). Gads Forlag, Copenhagen, pp 215–217 Tutin TG, Heywood VH, Burges NA, Moore DM, Valentine DH, Walters SM, Webb DA (1968) Flora Europaea, vol 2. Rosaceae to Umbelliferae. Cambridge University Press, Cambridge Ukraintseva VV (1991) The palynological study of the genus Cistus (Cistaceae). Bot J (St. Peterburg). 76:969–985 Ukraintseva VV (1993) Pollen morphology of the family Cistaceae in relation to its taxonomy. Grana 2:33–36

RELATED PAPERS

RELATED TOPICS

Log In

Molecular systematics, character evolution, and pollen morphology of Cistus and Halimium (Cistaceae)

Molecular systematics, character evolution, and pollen morphology of Cistus and Halimium (Cistaceae)

Molecular systematics, character evolution, and pollen morphology of Cistus and Halimium (Cistaceae)

Molecular systematics, character evolution, and pollen morphology of Cistus and Halimium (Cistaceae)

Molecular systematics, character evolution, and pollen morphology of Cistus and Halimium (Cistaceae)

Related Papers

RELATED PAPERS

RELATED TOPICS