Relationship of Hypochaeris salzmanniana (Asteraceae, Lactuceae), an endangered species of the Iberian Peninsula, to H. radicata and H. glabra and biogeographical implications

Botanical Journal of the Linnean Society, 2004, 146, 79–95. With 8 figures © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 79 Blackwell Science, LtdOxford, UKBOJBotanical Journal of the Linnean Society0024-4074 The Linnean Society of London, 2004? 2004 1461 7995 Original Article RELATIONSHIPS BETWEEN THREE SPECIES OF HYPOCHAERIS K. TREMETSBERGER ET AL . *Corresponding author. E-mail: [email protected] Relationship of Hypochaeris salzmanniana (Asteraceae, Lactuceae), an endangered species of the Iberian Peninsula, to H. radicata and H. glabra and biogeographical implications KARIN TREMETSBERGER1*, SALVADOR TALAVERA2, TOD F. STUESSY1, MARIA ÁNGELES ORTIZ2, HANNA WEISS-SCHNEEWEISS1 and GERTRUD KADLEC1 1Department of Higher Plant Systematics and Evolution, Institute of Botany, University of Vienna, Rennweg 14, A-1030 Vienna, Austria 2Departamento de Biología Vegetal y Ecología (Botánica), Facultad de Biología, Universidad de Sevilla, Avenida Reina Mercedes, E-41080 Seville, Spain Received September 2003; accepted for publication March 2004 Hypochaeris salzmanniana DC. (Asteraceae, Lactuceae) is an endangered species on the Iberian Peninsula, known from only eight coastal populations. Most authors have treated it as a variety, subspecies or simply as a synonym of H. glabra L. On the basis of morphological and cytological characters, Talavera recently separated H. salzmanniana (2n = 8) from H. glabra (2n = 10). Material of H. salzmanniana, H. glabra and H. radicata was collected from Spain, Italy, Sicily and Tunisia in order to assess taxonomic status and population relationships. Amplified Fragment Length Polymorphism (AFLP) analysis revealed three well-differentiated species. A close relationship between H. salzmanniana and H. radicata is also confirmed by AFLP analysis and chromosome number (2n = 8), morphology, and rDNA localization (FISH, fluorescence in situ hybridization). Hypochaeris salzmanniana and H. radicata share three fixed diagnostic AFLP fragments out of 348 fragments scored. The population structure of H. salzmanniana reveals distinct groups in southern Spain that are separated geographically. High differentiation among a western (Conil to Zahara), an intermediate (Punta Paloma and Los Algarbes) and an eastern (Algeciras and La Línea) group may reflect ancient separation. Population sizes and genetic compatibility differ greatly among populations and can be used to explain levels of within-population genetic diversity, together with recent documented loss of habitats resulting from tourist developments. Population structures of H. radicata and H. glabra show a similar geographical patterning: strongly differentiated populations from the Betic Cordillera and from the Iberian Massif, which are separated at present by the Guadalquivir river. Geological events at the end of the Tertiary (Tortonian–Messinian Miocene) might help explain patterns of differentiation in these three species of sect. Hypochaeris.© 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95. ADDITIONAL KEYWORDS: AFLP – conservation – genetic variation – Spain. INTRODUCTION Hypochaeris salzmanniana DC. was described as a distinct species by A. P. de Candolle (1838: 91). Subsequent authors working on the flora of the western Mediterranean have included H. salzmanniana with H. glabra L. as a variety (Amo, 1872: 493; Willkomm, 1893: 112), subspecies (Jahandiez & Maire, 1934: 831; Emberger & Maire, 1941: 1164), or simply as a mere synonym of H. glabra (De Filipps, 1976). Only recently, Talavera (1980, 1987) resurrected H. salzmanniana out of H. glabra not only as a good taxon based on morphological and karyological characters, but also for the first time as a species related to H. radicata L. rather than H. glabra. Förther & Podlech (2003) recently characterized populations from south-west Spain and north-west Morocco (coastal areas) as ssp. salzmanniana, being differentiated by conspicuously inflated peduncles Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022 80 K. TREMETSBERGER ET AL. © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 from ssp. maroccana Förther & Podlech in mountains of Morocco and Algeria to 1700 m a.s.l. [= H. arachnoidea sensu Oberprieler, 2002, and Oberprieler & Vogt, 2002)]. Morphological, cytological, compatibility and ecological differences exist among H. salzmanniana, H. radicata and H. glabra. Hypochaeris salzmanniana has hirsute, rarely glabrous, peduncles, strongly inflated under the capitula during anthesis, and ovate outer involucral bracts with wide scarious margins and setose hairs at the end of the central nerve, and with ligules conspicuously exserted from the involucre (Talavera, 1987). Hypochaeris glabra has glabrous, cylindrical peduncles, lanceolate and glabrous involucral bracts, rarely with a few setose hairs, and with ligules barely exserted from the involucre (Talavera, 1987). Field studies in 2002 by Talavera and Stuessy et al. have provided new morphological data to suggest recognition of H. salzmanniana as a good species. In comparison to its presumptive closest relative, H. radicata, the leaves of the former are more undulating at the margin, the leaf rosettes are usually very full, the involucral bracts often have long dark hairs, and the branches are ascending (Fig. 1A). Hypochaeris glabra and H. salzmanniana are annuals, whereas H. radicata is perennial. Hypochaeris glabra has 2n = 10 chromosomes (Adame & Talavera, 1980) whereas H. salzmanniana has 2n = 8 (Talavera, 1981) as does H. radicata (Adame & Talavera, 1980). Hypochaeris radicata is normally self-incompatible, H. glabra self-compatible, and H. salzmanniana shows variation in its compatibility system with plants being self-compatible or self-incompatible (M. Á. Ortiz & S. Talavera, unpubl. data). Hypochaeris glabra and H. radicata are widespread in the Mediterranean region and are worldwide weeds. Hypochaeris glabra seldom occurs in the same localities as H. salzmanniana (e.g. in Barbate and Punta Paloma, see Table 1), and H. salzmanniana does not occur with H. radicata. Hypochaeris salzmanniana is restricted to coastal areas on both sides of the Strait of Gibraltar. The species is reported from Morocco between Tanger and Rabat (Oberprieler, 2002; Förther & Podlech, 2003). In Spain, only eight populations are known along the coast of Cádiz, from Conil to La Línea (Fig. 2; Table 1). Six of these survive among beach developments: the western populations (Conil to Zahara, pops 1–4) and the population of Algeciras (pop. 7), which all live in sand dunes near the coast, and the threatened population of La Línea (pop. 8), which lives at the base of the Rock of Gibraltar. The only populations undisturbed by the tourism industry (Los Algarbes and Punta Paloma, pops 5 and 6) reside in the ecotone constituted by Quaternary dunes with flysch materials on the rocky spur of the Sierra de San Bartolomé (1 km inland). The reported pattern and size of populations of H. salzmanniana raises questions regarding their origins and biogeographical relationships. Because the populations occur in a linear fashion, it would be of interest to test whether or not genetic relationships among these eight populations conform to this linear Figure 1. Hypochaeris salzmanniana (A; Talavera, Stuessy et al. 5) and a typical recent dune habitat along the southern Spanish coast (B; Conil de la Frontera: Playa El Palmar). A B Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022 RELATIONSHIPS BETWEEN THREE SPECIES OF HYPOCHAERIS 81 © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 model. That is, did the populations originate in one location and then spread serially along the coast, or is a more complex explanation required? An attractive hypothesis is that those populations growing on fossil dunes may represent relict populations, perhaps now isolated genetically from the others. In order to assess relationships among H. salzmanniana and its relatives, H. glabra and H. radicata, and to determine infraspecific genetic variation within the three species, we used the DNA fingerprinting technique Amplified Fragment Length Polymorphism (AFLP) (Vos et al., 1995). AFLPs have already been shown efficacious at the population as well as interspecific level for revealing genetic relationships in Hypochaeris and other plant groups (e.g. Pfosser et al., 2002; Schönswetter et al., 2003; Stuessy et al., 2003; Tremetsberger et al., 2003a). Karyotypic analysis of H. salzmanniana, H. radicata and H. glabra is also provided as a comparison with the AFLP data. The purposes of this paper therefore are Table 1. Localities, collectors and number of individuals of Hypochaeris salzmanniana, H. radicata and H. glabra sampled for the AFLP study Taxa and localities Collectors and number No. individuals analysed H. salzmanniana DC. 1Spain, Cádiz, Conil de la Frontera: Playa El Palmar Talavera, Stuessy et al. 5-1 to 5-4 38 2Spain, Cádiz, Los Caños de Meca Talavera, Stuessy et al. 24 5 3Spain, Cádiz, Barbate de Franco Talavera, Stuessy et al. 14 5 4Spain, Cádiz, Zahara de los Atunes Talavera & Ortiz 1/03 5 5Spain, Cádiz, Punta Paloma: Los Algarbes Talavera & Ortiz 2/03 5 6Spain, Cádiz, Punta Paloma Talavera, Stuessy et al. 32; Talavera & Ortiz 3/03 19 7Spain, Cádiz, Algeciras: Palmones Talavera, Stuessy et al. 33 5 8Spain, Cádiz, La Línea de la Concepción Talavera, Stuessy et al. 35 5 H. radicata L. 9Brazil, São Paulo, São Paulo Talavera et al. BRA 30 1 10 Colombia, Bogotá, Bogotá Stuessy 31 2 11 Germany, Nordrhein-Westfalen, Bochum Stuessy s.n. 1 12 Italy, Fòggia, Gargano Tremetsberger s.n. 10 13 Italy, Règgio di Calàbria, Aspromonte Tremetsberger s.n. 6 14 Italy, Règgio di Calàbria, Barriteri Tremetsberger s.n. 2 15 Italy, Règgio di Calàbria, Palmi Tremetsberger s.n. 7 16 Italy, Sicily, Nebrodi F. & L. Ehrendorfer 5FE 4 17 Italy, Sicily, Palermo Royal Botanical Garden Madrid 646283 (seeds) 10 18 Mexico, México, Valle de Toluca Guevara s.n. 1 19 Spain, Burgos, Quintanar de la Sierra Ortiz s.n. 1 20 Spain, Cádiz, near Vejer de la Frontera Talavera, Stuessy et al. 25; Talavera et al. s.n. 6 21 Spain, Huelva, Aracena Talavera et al. s.n. 5 22 Spain, Huelva, National Park Doñana Talavera et al. s.n. 8 23 Spain, Huelva, Santa Ana la Real Talavera, Stuessy et al. 48 5 24 Spain, Huelva, Valverde del Camino Talavera, Stuessy et al. 46 5 25 Spain, Málaga, Gaucín Talavera, Stuessy et al. 40 3 26 South Africa, Western Cape, near Ceres F. & L. Ehrendorfer 83 1 27 Tunisia, Jendouba, near Aïn Draham F. & L. Ehrendorfer s.n. 9 H. glabra L. 28 Chile, Región VIII, Ñuble Baeza 3924 3 29 Spain, Cádiz, Barbate de Franco Talavera, Stuessy et al. 15 2 30 Spain, Cádiz, Punta Paloma Talavera, Stuessy et al. 31 2 31 Spain, Canary Islands, Gran Canaria Stuessy s.n. 2 32 Spain, Huelva, Hinojos Talavera, Stuessy et al. 43 2 33 Spain, Huelva, National Park Doñana Talavera et al. s.n. 2 34 Spain, Huelva, Valverde del Camino Talavera, Stuessy et al. 45 2 Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022 82 K. TREMETSBERGER ET AL. © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 to: (i) test the specific status of H. salzmanniana; (ii) assess its relationships to H. radicata and H. glabra; (iii) determine infraspecific genetic structures within the three species; and (iv) interpret the observed patterns in the context of biogeographical history and conservation perspectives. MATERIAL AND METHODS SAMPLING Leaves were sampled in silica gel from individuals in all known populations of Hypochaeris salzmanniana from the Iberian Peninsula (Fig. 2, Table 1). Five individuals were analysed in Los Caños, Barbate, Zahara, Los Algarbes, Algeciras, and La Línea (pops 2, 3, 4, 5, 7, and 8). Nineteen individuals were analysed in Punta Paloma (pop. 6). This population exhibited very low genetic variation after analysis of ten individuals sampled in 2002 (Talavera, Stuessy et al. 32; inds a to j). Nine additional individuals were analysed after the population had been resampled in 2003 (Talavera & Ortiz 3/03; inds k to s). In the large population Playa El Palmar (pop. 1), 38 individuals were analysed from four linear transects from sea to road across the dune, each approximately 1 km apart, so that genetic variation could be related to distance from the ocean. Along these transects each second square-metre quadrant was numbered sequentially from sea to road and one to two individuals, if present, were sampled in quadrants next to the ocean (low numbers: 1–3 in transects 1, 3 and 4; 4–9 in transect 2) and in quadrants far from the ocean (high numbers: 11–14 in transects 1, 2 and 4; 17–19 in transect 3). A total of 87 H. salzmanniana plants was investigated. Populations of H. radicata and H. glabra were similarly sampled (Table 1), with emphasis on populations of the former taxon from throughout its native range in the Mediterranean region (Spain, Italy, Sicily and Tunisia). Furthermore, a few samples were collected from introduced accessions throughout the world. Fewer populations of H. glabra were compared because morphological and karyological data indicated a more distant relationship to H. salzmanniana, despite the earlier taxonomic opinions of some authors that they were close relatives (see Introduction). These samples are from Spain, where H. glabra is native, the Canary Islands and Chile, where it is introduced and invasive. A total of 87 plants was investigated in H. radicata and 15 plants in H. glabra. Vouchers of all populations sampled are on deposit at SEV and WU. KARYOTYPES The karyotypes of H. salzmanniana [Spain, Cádiz, Playa El Palmar (pop. 1); see Table 1], H. radicata (Chile, Région VIII, Lota, Stuessy et al. 15477; WU), and H. glabra [Spain, Huelva, Valverde del Camino (pop. 34); see Table 1] were determined. Surface-ster- ilized seeds were germinated on wet filter paper on Petri dishes. Two days after germination, seedlings were pretreated with 0.1% colchicine for 2 h at room temperature and 2 h at 4∞C, fixed in 3:1 ethanol/acetic acid for 12 h at room temperature, and stored at -20∞C until use. Feulgen staining with Schiff’s reagent was carried out according to standard protocol (Weiss et al., 2003). From each species, a preparation with at least 15 well-spread chromosome plates was chosen for analysis. Chromosomes for fluorescence in situ hybridization (FISH) were prepared by enzymatic digestion/squashing as described by Weiss-Schneeweiss et al. (2003). Figure 2. Distribution of Hypochaeris salzmanniana in southern Spain, showing all known populations (all sampled; map by ARCVIEW GIS v.3.2; Environmental Systems Research Institute, Inc.). 6°00' 36°10' 36°00' 36°10' 36°00' 5°50' 5°40' 5°30' 5°20' 6°00' 5°50' 5°40' 5°30' 5°20' 01020 Kilometres Punta Paloma (pop. 6 ) Los Algarbes (pop. 5 ) Algeciras (pop. 7 ) La Línea (pop. 8 ) Zahara (pop. 4 ) Los Canos (pop. 2 ) Barbate (pop. 3 ) Conil (pop. 1) ~ Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022 RELATIONSHIPS BETWEEN THREE SPECIES OF HYPOCHAERIS 83 © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 The quality of spreads was checked by phase-contrast microscopy and only those preparations with adequate numbers of well-spread metaphases (10–15) were selected for FISH. Slides were frozen at -80∞C, and after cover slip removal, preparations were stored at -20∞C. FISH was carried out according to the method of Weiss-Schneeweiss et al. (2003). Probes used for FISH were: 18S-25S rDNA from Arabidopsis thaliana in plasmid pSK+, labelled with digoxygenin (Roche), and 5S rDNA from Beta vulgaris in plasmid pBx1–2, labelled with biotin (Roche). Both probes were labelled by nick translation according to the manufacturer’s instruction (Roche). Hybridization was carried out overnight. Stringent washes were performed as in Weiss-Schneeweiss et al. (2003). For digoxygenin/ biotin detection antidig-FITC solution (Roche)/extravidin Cy3 (Sigma) was applied to the slides at 37∞C for 1 h. The preparations were mounted in antifade buffer Vectashield (Vector Laboratories, UK) containing DAPI counterstain (2 mg/mL) and stored at 4∞C. Analyses of preparations were made with a ZEISS Axioscope epifluorescence microscope. Images were acquired with a CCD camera (Zeiss), and files were processed using ADOBE PHOTOSHOP software with only those functions that applied equally to all pixels in the image. For rDNA localization, a minimum of 30 well-spread metaphases and prometaphases was analysed for each species. AFLPS Total DNA was extracted from dry leaf material according to a modified CTAB-protocol and qualitychecked on 1% TAE-agarose gels as in Tremetsberger et al. (2003b). The AFLP procedure followed established protocols (Vos et al., 1995; PE Applied Biosystems, 1996) with modifications. Restriction-ligation reactions were carried out as in Tremetsberger et al. (2003b). The reaction mix contained 1.1 mL 10 ¥ T4 DNA ligase buffer (Promega), 1.1 mL 0.5 M NaCl, 0.55 mL 1 mg/mL BSA (New England Biolabs), 1 mL 50 mM MseI-adaptor, 1 mL 5 mM EcoRI-adaptor, 1 U MseI (New England Biolabs), 5 U EcoRI (Promega), 1 U T4 DNA ligase (Promega), and c. 0.5 mg template DNA, made up with water to a final volume of 11 mL. Ligated DNA fragments were diluted and preselective and selective amplifications performed as in Tremetsberger et al. (2003b). The reaction mix for the preselective amplification contained 1.14 mL 10 ¥ RedTaq polymerase buffer (Sigma), 0.2 U RedTaq polymerase (Sigma), 0.22 mL 10 mM dNTPs (Fermentas), 0.58 mL preselective primers (4.8 pmol/mL each of MseI and EcoRI preselective primers), and 2 mL diluted product of restriction/ligation, made up with water to a final volume of 10 mL. As selective primers with three and four selective nucleotides were chosen (see below), separate PCRs with one and two selective nucleotides in the MseI primer (MseI-C and -CT; EcoRI-A) were performed. The preselective PCR products were checked on a 1.5% TBE-agarose gel and diluted ten-fold with TE0.1 buffer. An initial screening of selective primers using 39 primer combinations was performed on five individuals from three populations of H. radicata and H. salzmanniana. A second screening used 15 primer combinations with clear bands on 13 individuals from eight populations of H. radicata and H. salzmanniana. From this, we have chosen six selective primer combinations with clear bands evenly distributed over the AFLP profiles for an application to all individuals investigated. These are: MseI-CTCG/ EcoRI-ATC(Fam), MseI-CAC/EcoRI-ACG(Hex), MseI- CTA/EcoRI-ACC(Ned), MseI-CTG/EcoRI-ACA(Fam), MseI-CTC/EcoRI-AGG(Hex), and MseI-CTGA/EcoRI- AAC(Ned). The reaction mix for the selective amplification contained 1 mL 10 ¥ RedTaq polymerase buffer, 0.2 U RedTaq polymerase, 0.22 mL 10 mM dNTPs, 0.54 mL of each selective primer (MseI-primer: 5 pmol/ mL; EcoRI-primer: 1 pmol/mL), and 2 mL diluted product of the preselective amplification, made up to a final volume of 10 mL with water. The fluorescence-labelled selective amplification products were run on a 5% denaturing polyacrylamide gel on an automated sequencer (ABI 377, Perkin Elmer). Before running, 0.8 mL NED- and HEX-labelled, and 0.4 mL FAM- labelled selective amplification products were mixed with 1.2 mL loading dye [containing 64.8 mL deionized formamide, 25.2 mL loading buffer, and 10 mL Gene- Scan-500 (ROX) size standard], and denatured at 95∞C for 2 min. Raw data were scored and exported as a presence/absence matrix as in Tremetsberger et al. (2003b) using ABI Prism GENESCAN Analysis Software v.2.1 (PE Applied Biosystems) and GENOGRAPHER (v.1.1.0; Montana State University, 1998, see http://hordeum.msu.montana.edu/genographer/). DATA ANALYSIS To test correspondence of data gathered by each of six primer combinations, Jaccard (1908) similarity matrices on all 189 individuals analysed were constructed for each primer combination separately using R PACKAGE (v.4.0; Casgrain & Legendre, 2000). The similarity matrices were used to compute pairwise correlations among them by a Mantel test (standardized Mantel statistic r; R PACKAGE v.4.0; Casgrain & Legendre, 2000). One-tailed test statistic probabilities were obtained through 999 permutations (R PACKAGE v.4.0; Casgrain & Legendre, 2000). The Jaccard (1908) distance matrix of all primer combinations combined (R PACKAGE; Casgrain & Legendre, 2000) was subsequently imported into PAUP* (v.4.0b8; Swofford, 1998) and used to construct Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022 84 K. TREMETSBERGER ET AL. © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 a UPGMA dendrogram for all individuals analysed, and for individuals of each species separately. A cophenetic correlation coefficient, which measures the goodness-of-fit of a model to the data, was calculated with NTSYSpc (v.2.2h; Applied Biostatistics Inc.). Support for each node was tested by 1000 bootstrap replicates with PAUP* (v.4.0b8; Swofford, 1998) using the UPGMA method in conjunction with Nei & Li’s (1979) genetic distances on the original presence/absence matrix, since PAUP* does not feature the Jaccard index. The Nei & Li (1979) distance and neighbourjoining algorithms were also applied to the data and resulted in very similar dendrograms (not shown). For each species, the total number of fragments, the number of fixed fragments (i.e. those occurring in all individuals analysed of a species), the diagnostic fragments (i.e. those occurring in only one species), and the fixed diagnostic fragments (i.e. those occurring in all individuals analysed of only one species) were assessed. A Mantel test was used to correlate geographical with genetic distances in H. salzmanniana. A Jaccard distance matrix among all pairwise combinations of individuals was computed using the R PACKAGE (v.4.0; Casgrain & Legendre, 2000) and compared to geographical distances between individuals (in km; distances within populations set to 0). The standardized Mantel statistic r is the linear correlation between genetic and geographical distances and was tested for significance with 999 permutations of rows and columns in one (the genetic or geographical) distance matrix to obtain the distribution under the null hypothesis of no correlation. The percentages of different AFLP phenotypes and of polymorphic fragments were assessed as measures of within-population genetic diversity for each population of H. salzmanniana (ARLEQUIN v.2.0; Schneider, Roessli & Excoffier, 2000). The Shannon diversity index was calculated from five randomly chosen individuals in a population as HSh = –S[pi*ln(pi)], where pi is the relative frequency of the ith fragment in a population, and correlated with population size (estimated number of individuals) and self-incompatibility (percentage of self-incompatible plants; Ortiz & Talavera, unpubl. data) using a Pearson correlation (1- tailed significance) in SPSS (v.10; SPSS Inc.). The effects of geographical vicinity within a transect and proximity or distance to the ocean have been tested as factors influencing the partitioning of genetic variation in the large population Playa El Palmar (pop. 1). ARLEQUIN (v.2.0; Schneider et al., 2000) was used to carry out an Analysis of Molecular Variance (AMOVA; Excoffier, Smouse & Quattro, 1992) with 1023 permutations among three hierarchical levels: (i) among transects, (ii) among quadrants next to the ocean and far from the ocean in each transect, and (iii) among individuals within each of the two groups of nearby quadrants (next to the ocean and far from the ocean) in each transect. Furthermore, the differentiation between the four main groups of H. salzmanniana, determined by UPGMA analysis, was quantified through pairwise FST values and tested for significance by 1023 permutations of haplotypes (distance method: pairwise difference; ARLEQUIN v.2.0; Schneider et al., 2000). RESULTS A total of 348 unambiguously scoreable AFLP fragments was obtained from analysis of six primer combinations on 189 individuals of Hypochaeris glabra, H. radicata and H. salzmanniana. The primer combinations MseI-CAC/EcoRI-ACG(Hex) yielded 71 fragments, MseI-CTCG/EcoRI-ATC(Fam) 68, MseI-CTA/ EcoRI-ACC(Ned) 62, MseI-CTG/EcoRI-ACA(Fam) 57, MseI-CTC/EcoRI-AGG(Hex) 47, and MseI-CTGA/ EcoRI-AAC(Ned) 43. A correlation test performed on each pairwise combination of six Jaccard similarity matrices obtained from analysis of each primer combination separately yielded generally high values (Table 2). The lowest value for Mantel’s r is 0.859 among MseI-CTG/EcoRI-ACA(Fam) and MseI-CTGA/ EcoRI-AAC(Ned), and the highest is 0.931 among MseI-CAC/EcoRI-ACG(Hex) and MseI-CTA/EcoRI- ACC(Ned). All values are significant (one-tailed significance = 0.001) after 999 permutations. DIFFERENTIATION OF H. GLABRA, H. RADICATA AND H. SALZMANNIANA Comparisons among populations of H. glabra, H. radicata and H. salzmanniana using AFLP show a clear distinction among the three species (Fig. 3; cophenetic correlation = 0.934). Each species consists of a highly supported (100% BS) cluster of populations. Hypochaeris radicata and H. salzmanniana are more similar to one another (level of divergence 0.327; Fig. 3) than to H. glabra, which branches off at a level of divergence of 0.351. However, a strong relationship between H. radicata and H. salzmanniana is not highly supported (51% BS). The total numbers of AFLP fragments are 142 in H. salzmanniana and 269 in H. radicata (accessions from Spain, Italy, Tunisia, and from worldwide introduced accessions). For a direct comparison it seems more appropriate to consider the same geographical area in both species. The Spanish populations of H. radicata alone (a total of 33 individuals) have 210 fragments, which is still considerably more than in H. salzmanniana. Hypochaeris glabra has 124 fragments, although this value might be underestimated, because only 15 individuals were investigated. The Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022 RELATIONSHIPS BETWEEN THREE SPECIES OF HYPOCHAERIS 85 © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 number of fixed fragments is 30 in Hypochaeris radicata (41 when only Spanish populations are considered), 63 in H. salzmanniana, and 78 in H. glabra. Diagnostic fragments are important for species determination, especially when they are fixed. The number of nonfixed diagnostic fragments is 133 in Hypochaeris radicata, 26 in H. salzmanniana, and 23 in H. glabra (Fig. 4A). The number of fixed diagnostic fragments is one in Hypochaeris radicata, nine in H. salzmanniana, and 17 in H. glabra (Fig. 4B). Shared diagnostic fragments (i.e. those occurring in two of the three species) are important for assessing evolutionary relationships among species. Hypochaeris radicata and H. salzmanniana share 52 nonfixed diagnostic fragments; H. radicata and H. glabra 30; and H. salzmanniana and H. glabra four (Fig. 4A). When shared diagnostic fragments are fixed in all individuals of the two species, the closer relationship of these two species becomes emphasized. Hypochaeris radicata and H. salzmanniana share three fixed diagnostic fragments; H. radicata and H. glabra two; and H. salzmanniana and H. glabra none (Fig. 4B). Cytological comparisons between H. radicata and H. salzmanniana (both 2n = 2x = 8) also reveal their similarity and a clear distinction from H. glabra (2n = 2x = 10) (Fig. 5). Karyotypically, as analysed by Feulgen-staining and FISH with 5S and 18S-25S rDNA, the former two species are indistinguishable. Their karyotypes have similar length and are symmetrical with metacentric and submetacentric chromosomes (Fig. 5C, E). Analyses of rDNA (5S and 18S–25S rRNA genes) have shown the same number and localization in karyotypes of H. radicata and H. salzmanniana (Fig. 5D, F). The karyotype of H. glabra is also symmetrical, but differs in chromosome number (2n = 10 vs. 2n = 8; Fig. 5A). The number of 5S and 18S-25S rDNA loci is the same as in the other two species, but their localization is different (Fig. 5B, D, F). In H. glabra, 5S rDNA is localized distally on the short arm of chromosome 4 (Fig. 5B), whereas in the other two species it is localized in the proximal part of the long arm of chromosome 2 (Fig. 5D, F). The 18S-25S rDNA locus occurs distally on the short arm of chromosome 2 in H. glabra (Fig. 5B), and on the short arm of chromosome 3 in the other two species (Fig. 5D, F). INFRASPECIFIC VARIATION WITHIN H. SALZMANNIANA The AFLP data also reveal similarities and distances among individuals of H. salzmanniana in southern Spain (Fig. 6). Four main clusters were found. The first cluster includes the geographically neighbouring populations Conil, Los Caños and Barbate (pops 1–3; with the exception of individual 2c clustering apart). The individual plants from the three locations are completely intermingled in the UPGMA dendrogram, suggesting a large panmictic group. The second cluster consists of population Zahara (pop. 4) and is most similar to the first cluster (pops 1–3). The third cluster comprises the geographically closely neighbouring populations Los Algarbes and Punta Paloma (pops 5 and 6). In fact, two collections in Punta Paloma (pop. 6) in two subsequent years (2002: Talavera, Stuessy et al. 32, inds a to j; 2003: Talavera & Ortiz 3/03, inds k to s) are separated by pop. 5 with the exception of ind. 6s collected in 2003 clustering with the 2002 collection. The fourth cluster comprises populations Algeciras and La Línea (pops 7 and 8), which grow in geographical proximity on the eastern side of the Strait of Gibraltar. The highest differentiation is between pops (5, 6) and (4) (FST = 0.662) and pops (5,6) and (7,8) (FST = 0.644; Table 3). The lowest differentiation is between pops (1–3) and (4) (FST = 0.318) and Table 2. Mantel’s r showing pairwise correlation among six Jaccard similarity matrices obtained from each of six primer combinations used (a total of 348 fragments and 189 individuals). One-tailed significance of all values is 0.001 after 999 permutations (R PACKAGE; Casgrain & Legendre, 2000) MseI-CTCG/ EcoRI-ATC(Fam) MseI-CAC/ EcoRI-ACG(Hex) MseI-CTA/ EcoRI-ACC(Ned) MseI-CTG/ EcoRI-ACA(Fam) MseI-CTC/ EcoRI-AGG(Hex) MseI-CAC/ EcoRI-ACG(Hex) 0.888 MseI-CTA/ EcoRI-ACC(Ned) 0.889 0.931 MseI-CTG/ EcoRI-ACA(Fam) 0.880 0.871 0.890 MseI-CTC/ EcoRI-AGG(Hex) 0.881 0.886 0.897 0.868 MseI-CTGA/ EcoRI-AAC(Ned) 0.874 0.902 0.927 0.859 0.891 Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022 86 K. TREMETSBERGER ET AL. © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 Figure 3. UPGMA dendrogram of Jaccard distances (based on 348 AFLP fragments) showing general relationships among individuals of Hypochaeris salzmanniana, H. radicata, and H. glabra. Cophenetic correlation coefficient 0.934. Bootstrap values for the three species (1000 replicates) above branches; Jaccard distances below branches. 0.000.050.100.150.200.250.300.35 H. glabraH. radicataH. salzmanniana 100 100 100 0.234 0.101 0.024 0.207 Jaccard distance Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022 RELATIONSHIPS BETWEEN THREE SPECIES OF HYPOCHAERIS 87 © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 between pops (1–3) and (7, 8) (FST = 0.364). The Mantel test revealed an overall positive, albeit low, correlation between geographical distances in km and genetic distances measured by the Jaccard index among all individuals investigated by Mantel’s r = 0.625 (one-tailed significance = 0.001 after 999 permutations; R PACKAGE, Casgrain & Legendre, 2000). Levels of within-population genetic variation vary among populations (Table 4). The large population Conil (pop. 1; c. 2 000 000 plants) has the largest variation (16.7% polymorphic fragments, Shannon diversity 9.52), and the 38 individuals analysed in this population harbour 125 of 142 fragments entirely found in H. salzmanniana. The population Los Algarbes (pop. 5) has the lowest variation (2.0% polymorphic fragments, Shannon diversity 2.31), and the five individuals analysed in this population harbour 96 fragments. Individuals 5b and 5d share the same AFLP phenotype. Identical AFLP phenotypes are also found in Punta Paloma (pop. 6), in which three (b–d) and four (f, h–j) individuals of the 2002 collection Figure 4. Pie diagrams of nonfixed (A) and fixed (B) diagnostic AFLP fragments (i.e. fragments occurring in only one or in only two species of Hypochaeris salzmanniana, H. glabra, and/or H. radicata). A B H. radicata and H. glabra H. radicata and H. glabra H. glabra H. glabra H. glabra and H. salzmanniana (4) H. salzmanniana H. salzmanniana H. salzmanniana and H. radicata H. salzmanniana and H. radicata H. radicata H. radicata 12 3 9 17 52 133 26 23 30 Table 3. Pairwise FSTs between four main groups of Hypochaeris salzmanniana (determined through UPGMA analysis, Fig. 6). Significance of each value is 0.000 after 1023 permutations (ARLEQUIN v.2.0; Schneider et al., 2000) Pops 1–3 Pop. 4 Pops 5, 6 Pops 4 0.318 Pops 5, 6 0.418 0.662 Pops 7, 8 0.364 0.510 0.644 Table 4. Population size (estimated no. of individuals in population) and genetic diversity in populations of Hypochaeris salzmanniana based on 142 AFLP fragments (ARLEQUIN v.2.0; Schneider et al., 2000). None of the populations has fixed private fragments Population size % different AFLP phenotypes % polymorphic fragments Shannon diversity No. of private fragments Conil (pop. 1) 2 000 000 100 16.7 9.52 5 Los Caños (pop. 2) 500 100 9.8 9.90 1 Barbate (pop. 3) 250 100 7.8 7.82 – Zahara (pop. 4) 200 100 4.0 4.15 2 Los Algarbes (pop. 5) 20 000 80 2.0 2.31 – Punta Paloma (pop. 6) 50 47 4.3 3.44 3 Algeciras (pop. 7) 600 100 6.3 6.97 2 La Línea (pop. 8) 250 100 8.9 9.00 2 Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022 94 K. TREMETSBERGER ET AL. © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 the Ministerio de Ciencia y Tecnología (PB96-1352 and REN2002-04634-C05-03 to S. Talavera and REN2002-04354-C02-02 to M. Arista) and from Junta de Andalucía (group RNM-204). REFERENCES Adame P, Talavera S. 1980. Números cromosómicos para la flora española: 130–131. Lagascalia 9: 256–257. Amo M. 1872. Flora fanerogámica de la Península Ibérica 4. Granada. Asensi A, Díez-Garretas B. 1987. Andalucía Occidental. In: Peinado-Lorca M, Rivas-Martínez S, eds. La vegetación de España. Alcalá de Henares: Universidad de Alcalá, 199– 230. Blanca G. 1993. Origen de la Flora Andaluza. In: Valdés B, ed. Introducción a la flora Andaluza. Sevilla: Junta de Andalucía, 19–35. Bocquet G, Widler B, Kiefer H. 1978. The Messinian Model – a new outlook for the floristics and systematics of the Mediterranean area. Candollea 33: 269–287. Braga JC, Aguirre J. 2001. Coralline algal assemblages in upper Neogene reef and temperate carbonates in Southern Spain. Palaeogeography, Palaeoclimatology, Palaeoecology 175: 27–41. Busack SD. 1986. Biochemical and morphological differentiation in Spanish and Moroccan populations of Discoglossus and the description of a new species from Southern Spain (Amphibia, Anura, Discoglossidae). Annals of Carnegie Museum 55: 41–61. de Candolle AP. 1838. Prodromus Systematis Naturalis Regni Vegetabilis 7. Paris. Casgrain P, Legendre P. 2000. The R package for multivariate and spatial analysis, v.4.0. (development release 2). Québec: University of Montréal. De Filipps RA. 1976. 158. Hypochoeris L. In: Tutin TG, Heywood VH, Burges NA, Moore DM, Valentine DH, Walters SM, Webb DA, Chater AO, De Filipps RA, Richardson IBK, eds. Flora Europaea, Vol. 4: Plantaginaceae to Compositae (and Rubiaceae). Cambridge: Cambridge University Press, 308–310. Duggen S, Hoernle K, Van den Bogaard P, Rüpke L, Morgan JP. 2003. Deep roots of the Messinian salinity crisis. Nature 422: 602–606. Emberger L, Maire R. 1941. Catalogue des plantes du Maroc 4. Algiers: Imprimerie Minerva. Estabrook GF. 2001. Vicariance or dispersal: The use of natural historical data to test competing hypotheses of disjunction on the Tyrrhenian coast. Journal of Biogeography 28: 95–103. Excoffier L, Smouse P, Quattro J. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restriction site data. Genetics 131: 479–491. Fernández-Palacios A, Fernández-Palacios C, Gil BJ. 1988. Guias de naturalistas de la Provincia de Cádiz. I. El Litoral. Cádiz: Diputación de Cádiz. Förther H, Podlech D. 2003. Die Gattung Hypochaeris L. sect. Hypochaeris (Compositae) im westlichen Afrika. Sendtnera 8: 35–43. Galán de Mera A. 1995. Contribución a las floras de la provincia de Cádiz y de la Península Tingitana. Lagascalia 18: 55–62. García-París M, Jockusch EL. 1999. A mitochondrial DNA perspective on the evolution of Iberian Discoglossus (Amphibia: Anura). Journal of Zoology 248: 209–218. García-París M, Alcobendas M, Alberch P. 1998. Influence of the Guadalquivir River Basin on mitochondrial DNA evolution of Salamandra salamandra (Caudata: Salamandridae) from Southern Spain. Copeia 1998: 173–176. Guerra-Merchán A, Serrano F. 1993. Tectosedimentary setting and chronostratigraphy of the Neogene reefs in the Almanzora corridor (Betic Cordillera, Spain). Geobios 26: 57–67. Gutiérrez JM, Martin A, Domínguez S, Moral JP. 1991. Introducción a la geología de la Provincia de Cádiz. Cádiz: Universidad de Cádiz. Jaccard P. 1908. Nouvelles recherches sur la distribution florale. Bulletin de la Société Vaudoise de Sciences Naturales 44: 223–270. Jahandiez E, Maire R. 1934. Catalogue des plantes du Maroc, 3. Algiers: Imprimerie Minerva. de Jong H. 1998. In search of historical biogeographic patterns in the Mediterranean terrestrial fauna. Biological Journal of the Linnean Society 65: 99–164. Krijgsman W, Hilgen FJ, Raffi I, Sierro FJ, Wilson DS. 1999. Chronology, causes, and progression of the Messinian salinity crisis. Nature 400: 652–655. Lonergan L, Johnson C. 1998. Reconstructing orogenic exhumation histories using synorogenic detrital zircons and apatites: an example from the Betic Cordillera, SE Spain. Basin Research 10: 353–364. Lonergan L, White N. 1997. Origin of the Betic-Rif mountain belt. Tectonics 16: 504–522. Meulenkamp JE, Sissingh W, Ilyina LB, Kovac M, Barrier E. 2000. Late Tortonian (8.4–7.2 Ma). In: Dercourt J, Gaetani M, Vrielynck B, Barrier E, Biju-Duval B, Brunet MF, Cadet JP, Crasquin S, Sandulescu M, eds. Atlas Peri-Tethys, Palaeogeographical maps. Paris: CCGM/CGMW. Myers N, Mittermeier RA, Mittermeier CG, Da Fonseca GAB, Kent J. 2000. Biodiversity hotspots for conservation priorities. Nature 403: 853–858. Nei M, Li W-H. 1979. Mathematical model for studying genetic variation in terms of restriction endonucleases. Proceedings of the National Academy of Sciences, USA 76: 5269– 5273. Oberprieler C. 2002. Hypochaeris. In: Valdés B, Rejdali M, Achhal El Kadmiri A, Jury SL, Montserrat JM, eds. Catalogue des plantes vasculaires du Nordiska du Maroc, incluant des clés d’identification, 2. Madrid: Consejo Superior de Investigaciones Científicas, 686–689. Oberprieler C, Vogt R. 2002. Hypochaeris arachnoidea Poir., a hitherto neglected species in NW Africa. Willdenowia 32: 231–236. Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022 RELATIONSHIPS BETWEEN THREE SPECIES OF HYPOCHAERIS 95 © 2004 The Linnean Society of London, Botanical Journal of the Linnean Society, 2004, 146, 79–95 Orszag-Sperber F, Butterlin J, Clermonte J, Colchen M, Guiraud R, Poisson A, Ricou LE. 1993. Tortonian (11.5–6 Ma). In: Dercourt J, Ricou LE, Vrielynck B, eds. Atlas Tethys palaeoenvironmental maps. Paris: Gauthier Villars. PE Applied Biosystems. 1996. AFLPTM plant mapping protocol. Foster City, CA: PE Applied Biosystems. Pendón JG. 1978. Sedimentación turbidítica en las unidades del Campo de Gibraltar. Granada: Universidad de Granada. Pfosser MF, Guzy-Wróbelska H, Sun B-Y, Stuessy TF, Sugawara T, Fujii N. 2002. The origin of species of Acer (Sapindaceae) endemic to Ullung Island, Korea. Systematic Botany 27: 351–367. Riding R, Braga JC, Martín JM, Sánchez-Almazo IM. 1998. Mediterranean Messinian Salinity Crisis: constraints from a coeval marginal basin, Sorbas, southeastern Spain. Marine Geology 146: 1–20. Samuel R, Stuessy T, Tremetsberger K, Baeza C, Siljak- Yakovlev S. 2003. Phylogenetic relationships among species of Hypochaeris (Asteraceae, Lactuceae) based on ITS, plastid trnL intron, trnL-F spacer, and matK sequences. American Journal of Botany 90: 496–507. Sánchez-Almazo IM, Spiro B, Braga JC, Martín JM. 2001. Constraints of stable isotope signatures on the depositional palaeoenvironments of upper Miocene reef and temperate carbonates in the Sorbas Basin, SE Spain. Palaeogeography, Palaeoclimatology, Palaeoecology 175: 153–172. Sánchez-Gómez M, Balanyá JC, García-Dueñas V, Azañón JM. 2002. Intracrustal tectonic evolution of large lithosphere mantle slabs in the western end of the Mediterranean orogen (Gibraltar arc). Journal of the Virtual Explorer 8: 23– 34. Schneider S, Roessli D, Excoffier L. 2000. Arlequin, Version 2.000: a software for population genetics data analysis. Geneva: University of Geneva. Schönswetter P, Tribsch A, Schneeweiss GM, Niklfeld H. 2003. Disjunctions in relict alpine plants: phylogeography of Androsace brevis and A. wulfeniana (Primulaceae). Botanical Journal of the Linnean Society 141: 437–446. Seidenkrantz M-S, Kouwenhoven TJ, Jorissen FJ, Shackleton NJ, Van der Zwaan GJ. 2000. Benthic foraminifera as indicators of changing Mediterranean-Atlantic water exchange in the late Miocene. Marine Geology 163: 387–407. Soria JM, Fernández J, Viseras C. 1999. Late Miocene stratigraphy and paleogeographic evolution of the intramontane Guadix Basin (Central Betic Cordillera, Spain): implications for an Atlantic-Mediterranean connection. Palaeogeography, Palaeoclimatology, Palaeoecology 151: 255–266. Stuessy TF, Tremetsberger K, Müllner AN, Jankowicz J, Guo Y-P, Baeza CM, Samuel RM. 2003. The melding of systematics and biogeography through investigations at the populational level: examples from the genus Hypochaeris (Asteraceae). Basic and Applied Ecology 4: 287–296. Swofford DL. 1998. PAUP*: phylogenetic analysis using parsimony (*and other methods), v.4.0b8. Sunderland, MA: Sinauer Associates. Talavera S. 1980. Nota taxonómica sobre el género Hypochoeris L. en Andalucía Occidental. Lagascalia 9: 232–234. Talavera S. 1981. Números chromosómicos para la flora española: 182–256. Lagascalia 10: 225–256. Talavera S. 1987. Hypochaeris. In: Valdés B, Talavera S, Fernández-Galiano E, eds. Flora vascular de Andalucía Occidental, 3. Barcelona: Ketres Editoria, 102–105. Tremetsberger K, Stuessy TF, Guo Y-P, Baeza CM, Weiss H, Samuel RM. 2003a. Amplified Fragment Length Polymorphism (AFLP) variation within and among populations of Hypochaeris acaulis (Asteraceae) of Andean southern South America. Taxon 52: 237–245. Tremetsberger K, Stuessy TF, Samuel RM, Baeza CM, Fay MF. 2003b. Genetics of colonization in Hypochaeris tenuifolia (Asteraceae, Lactuceae) on Volcán Lonquimay, Chile. Molecular Ecology 12: 2649–2659. Vos P, Hogers R, Bleeker M, Reijans M, Van de Lee T, Hornes M, Frijters A, Pot J, Peleman J, Kuiper M, Zabeau M. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Research 23: 4407–4414. Weiss H, Stuessy TF, Grau J, Baeza CM. 2003. Chromosome reports from South American Hypochaeris (Asteraceae). Annals of the Missouri Botanical Garden 90: 56–63. Weiss-Schneeweiss H, Stuessy TF, Siljak-Yakovlev S, Baeza CM, Parker J. 2003. Karyotype evolution in South American species of Hypochaeris (Asteraceae, Lactuceae). Plant Systematics and Evolution 241: 171–184. Willkomm M. 1893. Supplementum Prodromi Florae Hispanicae. Stuttgart. Wright S. 1943. Isolation by distance. Genetics 28: 114–138. Downloaded from https://academic.oup.com/botlinnean/article/146/1/79/2420314 by Universidad de Sevilla user on 19 May 2022