OPCE-STR.vp Acta Bot. Croat. 69 (1), 71–82, 2010 CODEN: ABCRA 25 ISSN 0365–0588 Multilocus genomic associations among selected taxa of genus Potentilla (Rosaceae) in Poland using RAPD analysis JEREMI KO£ODZIEJEK1*, TOMASZ CIEŒLIKOWSKI2, TOMASZ SAKOWICZ3 1 University of £ódŸ, Department of Geobotany and Plant Ecology, University of £ódŸ, Banacha 12/16, 90-237 £ódŸ, Poland 2 Centre for Medical Biology Polish Academy of Sciences, Lodowa 106, 93-232 £ódŸ, Poland 3 University of £ódŸ, Department of Cytogenetics and Plant Molecular Biology, Banacha 12/16, 90-237 £ódŸ, Poland Nine taxa of Potentilla species from Poland representing P. sect. Terminales (Dõll.) Gren et Godr. and P. sect. Aureae (Wolf) Juz. were analyzed via a series of random amplified polymorphic DNA (RAPD) analyses to test (1) the hypothesis that the six species repre- senting P. subsect. Collinae Juz. of the P. Terminales sect., i.e. P. collina Wibel, P. thyrsiflora Zimmeter, P. wimannania Günther et Schummel, P. leucopolitana P. J. Müll., P. ´gabarae Kolodziejek, P. koernickei Zimmeter, are genetically differentiated enough to be considered as separate taxa, and (2) the position of populations of P. thyrsiflora and P. collina with respect to the Terminales sect. (P. argentea L) and the Aureae sect. (P. taber- naemontani Ascherson and P. incana P. Gaertner, B. Meyer et Scherb.). RAPD-based ge- netic similarity values using the UPGMA method and corresponding dendrogram exhib- ited incomplete accordance between RAPD and morphological variations. According to 'overgenomic' associations based on a series of genomic loci selected at random, P. thyrsiflora and P. collina are closely related and similarly related to the species: P. argentea, P. tabernaemontani and P. incana hypothesized as being 'parental'. Within Terminales sect., our dendrogram shows P. argentea to be relatively isolated from the other members of the section analysed, P. thyrsiflora and P. collina. Key words: average similarity, multilocus diversity, Potentilla, RAPD, Rosaceae, taxonomy Abbreviations: RAPD – random amplified polymorphic DNA; UPGMA – unweighted pairs group method using mathematical averages Introduction The genus Potentilla L. comprises approximately 490 species distributed mainly in the northern hemisphere (ERIKSSON et al. 1998, SOJÁK 2005). Within Potentilla seven subgen- ACTA BOT. CROAT. 69 (1), 2010 71 * Corresponding author, e-mail: kolo@biol.uni.lodz.pl U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees era have been distinguished, i.e. Schistophyllidium Juz., Micropogon Bge., Fragariastrum Ser., Closterostyles Torr. et Gr., Hypargyrium Fourr., Dynamidium Fourr. and Argentina (Lam.) Jepson (JUZEPCZUK 1941, SOJÁK 2005). This study included seven species of the P. sect. Terminales (Dõll.) Gren. et Godr. (P. argentea and six species of P. subsect. Collinae) and two species of the P. sect. Aureae (Wolf) Juz. (P. tabernaemontani Ascherson and P. incana P. Gaertner, B. Meyer et Scherb.) within P. subgen. Hypargyrium. In Europe, a total of 25 taxa have been described in the P. subsect. Collinae, most of them being not very well defined (as separate species, subspecies or varietas). This diver- gence in taxonomic viewpoints is accompanied by problems in distinguishing them and mistakes in the use of synonyms. Species identification strictly on the basis of morphologi- cal features is however difficult, because all species are closely related and morphologi- cally similar. The high degree of polymorphism is partly caused by apomictic reproduction (pseudogamy) in conjunction with polyploidy and hybridization, as first pointed out by MÜNTZING (l928, 1931). In Poland, there are 12 taxa of P. subsect. Collinae, i.e. P. collina Wibel, P. thyrsiflora Zimmeter, P. silesiaca R. Uechtr., P. wimannania Günther et Schum- mel, P. leucopolitana P. J. Müll., P. microdons Schur, P. schultzii F. W. Schultz, P. koernickei Zimmeter, P. leucopolitanoides B³ocki, P. scholziana Callier, P. tynieckii B³ocki and P. ´gabarae Ko³odziejek (P. leucopolitana P. J. Müll. ´ P. incana P. Gertner, B. Meyer et Scherb) (KO£ODZIEJEK unpublished). Since it is widely accepted (ASCHERSON and GRAEBNER 1904, WOLF 1908, ASKER and FRÕST 1970, KURTTO et al. 2004) that the species of P. subsect. Collinae has been derived from natural hybridization between different Potentilla species, we attempted to use RAPD analysis to deduce the putative parents of this species. In morphology, the taxa of P. subsect. Collinae shows traits both from the Terminales (P. argentea) and the Aurea sec- tions (P. tabernaemontani and P. incana). The characteristic features of the P. subsect. Collinae are primarily style conical shape with a few papillae at base, slightly clavate at apex; leaflets 5–7 oblong-obovate or oblanceolate, variously toothed, sparsely pubescent to white-tomentose or sericeous beneath with mixture of four types of hairs, i.e. crispate, straight, curved and imperfectely stellate. In contrast, P. argentea has a subcylindrical style, a dense crispate indumentum on the lower side of leaves and leaves with inrolled margins. The P. subsect. Collinae differs from the P. tabernaemontani and P. incana by its mixed (straight, curved, flexuous and incomplete stellate) hairs on the upper and lower side of the leaves. P. tabernaemontani have only straight hairs. From P. incana it differs by the lack of the stellate hairs. Apart from by the different indumentum, the members of the Collinae subsect. can easily be separated from species of the Aureae sect. (P. tabernaemon- tani and P. incana) by their style shape. In species of the Aureae sect., the styles towards the apex are widened. The members of the Collinae subsect. occurs in isolated sites in the plains of Poland and have slightly different habitats. The species prefer open habitats, mostly on dry sandy soils and calcareous rendzinas. Potentilla argentea and P. incana can be found growing abun- dantly on dry mineral soils over the whole territory, while P. tabernaemontani occurs lo- cally on dry soils in West, Central and South Poland. However, in a few localities, for in- stance on the Wy¿yna Czêstochowska upland the seven taxa, i.e. P. collina, P. thyrsiflora, P. wimannania, P. leucopolitana, P. tabernaemontani, P. incana and P. argentea occurred side by side on gravelly or sandy, not or only slightly calcareous ground (KO£ODZIEJEK 2004). 72 ACTA BOT. CROAT. 69 (1), 2010 KO£ODZIEJEK J., CIEŒLIKOWSKI T., SAKOWICZ T. U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees In the present paper we explore the value of a detailed, molecular (RAPDs) approach for unravelling complex variation at low taxonomic level, in order to use this approach to (1) test the hypotheses whether the morphologically defined species of the Collinae subsect. represent genetically distinct units. (2) secondly, we consider the position of P. thyrsiflora and P. collina with respect to the Terminales (P. argentea) and Aureae sections (P. tabernaemontani and P. incana), in order to obtain a better taxonomic classification of Potentilla. Material and methods Plant material and genomic DNA isolation In this study, we investigated six species of P. subsect. Collinae, i.e. P. collina, P. thyrsiflora, P. wimannania, P. leucopolitana, P. koernickei, P. ´gabarae and a group repre- senting the Terminales and Aureae sections, i.e. P. argentea, P. tabernaemontani and P. incana considered (ASCHERSON and GRAEBNER 1904, WOLF 1908, ASKER and FRÕST 1970) as the potential parental species for the group. The geographic localities and information on the population samples are given in table 1. Some of the plants investigated were preserved in the form of herbarium specimens and deposited in the Department of Geobotany and ACTA BOT. CROAT. 69 (1), 2010 73 MULTILOCUS GENOMIC ASSOCIATIONS OF POTENTILLA Tab. 1. Collection data from Potentilla material sampled from cytogenetic analyses. The chromosome counts after ILNICKI and KOLODZIEJEK (2008). Population code Taxon 2n Sample collection site A P. collina 28,35 Silesia prov., K³obuck 50°57’N/18°59’E B P. collina 28,35 Silesia prov., Rzêdkowice 50°35’N/19°27’E C P. collina 28,35 Silesia prov., Mirów 50°37’N/19°28’E D P. thyrsiflora 21,28,35 Silesia prov., Cisowa near Pilica 50°26’N/19°42’E E P. thyrsiflora – Silesia prov., Ostra Góra near Siewierz 50°27’N/19°09’E F P. wimannania 35 Silesia prov., Mirów 50°37’N/19°28’E G P. leucopolitana 35 Pomerania prov., Cha³upy 54°43’N/18°23’E H P. ´gabarae – Silesia prov., Jaroszów village near ¯arki 50°39’N/19°21’E I P. koernickei 28,35,42 Pomerania prov., Czarna Woda near Czersk 53°51’N/18°08’E J P. argentea 35 Silesia prov., Cisowa near Pilica 50°26’N/19°42’E K P. argentea 35 £ódŸ prov., Bolimów Landscape Park 52°05’N/20°10’E L P. tabernaemontani – Ma³opolska prov., Szaflary 49°25’N/19°33’E M P. tabernaemontani – Silesia prov., Jaroszów village near ¯arki 50°39’N/19°21’E N P. incana 28 Ojców Nationale Park 50°14’N/19°50’E O P. incana 28 Sl¹sk prov., Kusiêta village nar Olsztyn 50°45’N/19°16’E U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees Plant Ecology of the University of Lodz. Others were transplanted into the experimental garden for further studies. The selected species of Collinae are the most common taxa of the subsection occurring in Poland. Two of them, i.e. P. collina and P. thyrsiflora are en- countered in relatively the greatest number of localities and their populations are the rich- est, consisting of more than fifty individuals. The relative abundance and incidence of the accessions observed for these two species suggests more frequent genetic material ex- change between them and the species of the Terminales and Aureae sections. The studied plant material was collected during field trips in Poland. For RAPD analy- sis we used five plants representing each population. Each plant selected was separated from another by at least 0.5 m. All of them were dug in May – July. Leaves were dried in fine-grained silica gel. The samples of the plant genomic DNA were extracted from approximately 0.25 g of fresh and young leaves from field-collected plants. The frozen material, prior to the cell-lysis, was disrupted by grinding in liquid nitrogen, then digested by RNase-A in a lysis buffer at 65°C. The extraction itself, performed using DNeasy Plant Mini Kit (Qiagen), was followed by the spectrophotometrical quantification at 260 nm. The UV-spectrophoto- meter Hitachi U-2000, Japan was used for determination of DNA purity and concentration. Similar to the other experimental work (e.g. ROMÁN et al. 2003, SAKOWICZ and CIEŒLIKOWSKI 2006), ours followed the template-mixing strategy where equal amounts of working solution DNAs from each group of individuals of the same species were pooled as the 'species-template DNA' prior to the PCR reaction. DNA was extracted from single plants and each population was represented by the bulk of a variable number of individuals, depending on the availability of samples after collection. That was to increase the relative contribution of the common target sequences and consequently to highlight the spe- cies-specific features (amplicons) at the step of amplification, before starting numerical analysis. RAPD amplification The random polymorphic DNA amplification was performed according to the subtle modified (temperature profile) procedure of WILLIAMS at al. (1990). The reaction mixture included 10 mM Tris-HCl pH 8.3, 50 mM KCl, 2 mM MgCl2 (Finnzyme, Finland), 0.2 mM dNTP (Promega, USA). The aliquot of 25 mL contained: the reaction mixture, 30 ng of primer (SIGMA-ARK, Germany), 5 ng of genomic DNA and 2 units of termostable poly- merase Dynazymeä(Finnzyme, Finland) and Milli-Q water (Millipore, Austria). The tem- perature profile was as follows: initial denaturation at 95 °C for 1 min., followed by 35 cy- cles (denaturation at 94 °C for 1 min., annealing for 1 min. and extension at 72 °C for 2 min.) with a final extension on the step at 72 °C for 10 min. The annealing temperature was primer-specific and was each time 5°C below its melting temperature supplied by the man- ufacturer – SIGMA-ARK (Tabs. 2, 3). The amplification was performed in thin-wall vials (MJ Research, USA) with a thermocycler Uno II (Biometra, Germany). The amplification products were separated against molecular-weights marker (100 bp DNA ladder, MBI Fermentas) in 1.5% agarose gel and TAE buffer (40 mM TRIS-acetate, 1 mM EDTA, pH 7.8) at 80 V in MGU 602T electrophoresis unit (CBS Scientific, USA). The agarose gel stained with ethidium bromide was visualized under ultra-violet light and documented us- ing computer image system (Vilber Lourmat, France, Agarose, TAE, EtBr from SERVA – 74 ACTA BOT. CROAT. 69 (1), 2010 KO£ODZIEJEK J., CIEŒLIKOWSKI T., SAKOWICZ T. U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees Germany). The gel stained with ethidium bromide was visualized under ultra-violet light and documented using computer image system (Vilber Lourmat, France). Numerical analysis of DNA fingerprinting electrophoretic patterns The computer assisted analysis was carried out with the aid of BioNumerics pro- gramme (Applied Maths, Kortijk, Belgium), and the unweighted pairs group method using mathematical averages – UPGMA was used for the data cluster analysis. Numbers at branches are bootstrap values (%) generated after 1000 replications. According to specific- ity of the RAPD method, the electrophoretic patterns were compared on the basis of the whole densitometric curve shape and the Pearson’s product-moment correlation coeffi- cient has been applied. The final dendrograms describing averaged phenetic similarities were calculated according to UPGMA algorithm and euclidean distances (Statistica 5.1, Statsoft). The bootstrap method (FELSENSTEIN 1985) employed to evaluate the reliability of tree topology was evaluated after 1000 samples. The calculations were performed with the MEGA 4.0. software (TAMURA et al. 2007). Results RAPD-based genetic diversity among the six species of P. subsect. Collinae RAPDs generated a total of 110 bands using 13 decamer primers (an average of 8.5 bands per assay) of which 39.1% were polymorphic. The size of amplification products ranged between 300–800 bp (Tab. 2). The dendrograms formed (not shown) as a result of 13 experiments were diverse in their shape and global homology level. For seven out of the thirteen RAPD tests, P. leucopolitana and P. koernickei were rather similar and their electrophoretic patterns ACTA BOT. CROAT. 69 (1), 2010 75 MULTILOCUS GENOMIC ASSOCIATIONS OF POTENTILLA Tab. 2. Names, percentage of polymorphic bands and melting temperatures (Tm) of different primers used to RAPD analysis for six species of P. subsect. Collinae. Code Sequence 5’ to 3’ Percentage of polymorphic bands Tm (°C) J-01 TGGGTCCCTC 42.9 34 J-02 CGGCGGACTA 40.0 38 J-03 GGCGGATAGC 50.0 34 J-04 GAGTCAGCAG 36.4 32 J-05 GTCAGGGCA 37.5 32 J-06 GTCAGGGCAA 33.3 32 J-07 TCACGTCCAC 40.0 32 J-08 CAGGGGTGGA 41.7 34 J-09 CAGGGGTGGA 40.0 34 J-10 GGGCCGTTTA 28.6 32 J-11 GCCCTCGGAT 44.4 32 J-12 GTGCGCGACC 37.5 29 J-13 TGAGGGTCCC 36.4 34 39.1 (average) U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees formed common subcluster. These two species were weakly associated with P. ´gabarae. In 8 experiments, three populations of P. collina were clustered altogether, while in 3 RAPD tests they were clustered together with P. wimannania. In an analysis of the total RAPD data set, three very distinct clusters of RAPD pheno- types (Fig. 1) appeared. The first distinct minor cluster consisted of P. leucopolitana and P. koernickei. These species showed a similarity coefficient of 0.43 with the cluster support- ing by a bootstrap value of 82%. The middle cluster included P. thyrsiflora (population D). The third cluster was further divided into two, somewhat less distinct, subclusters of which the left included population of P. collina located in Mirów (population C), P. wimannania, P. ´gabarae, and while the right subcluster was formed by two populations (A, C) of P. collina and P. thyrsiflora (population E). In the present study, significant divergence was found between two P. thyrsiflora popu- lations, and the genetic distance values between populations was 0.48. Populations from different sites did not cluster together. One RAPD phenotype of P. thyrsiflora observed in plants from the Cisowa population clustered at a high level (0.49) with the P. leucopolitana and P. koernickei cluster, and the other phenotype of P. thyrsiflora observed in the Ostra Góra population from clustered well within the P. collina subcluster. The associations de- termined with RAPD markers among P. subsect. Collinae showed genetic similarity coeffi- cients ranging from 0.22 (population C of P. collina and P. wimannania) to 0.55 (P. leucopolitana, P. koernickei and P. thyrsiflora) revealing medium levels of genetic varia- tion among the species studied. RAPD-based genetic diversity among P. collina, P. thyrsiflora, and the putative parents Using 19 decamer primers, a total of 168 RAPD marker loci were scored (an average of 8.8 bands per primer) and 43.4% were polymorphic over all accessions (Tab. 3). For most of the RAPD tests the electrophoretic patterns of P. collina and P. thyrsiflora formed common cluster (15/19 RAPDs) or were weakly connected in a stair shape manner (4/19 RAPDs), while the species representing Aureae sect. remained separated (not 76 ACTA BOT. CROAT. 69 (1), 2010 KO£ODZIEJEK J., CIEŒLIKOWSKI T., SAKOWICZ T. 0,45 0,40 0,35 0,30 0,25 0,20 0,15 0,10 4 (G) 5 (I) 2 (D) 1 (C) 3 (F) 6 (J) 1(A) 1(C) 2 (E) A v e ra g e d g e n o m ic s im il a ri ty Fig. 1. Taxonomic similarity among the Potentilla subsect. Collinae Juz. species – the final dendro- gram. The averaged similarity between nine populations found on the basis of thirteen RAPD analyses. U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees shown). The internal homology level of P. collina and P. thyrsiflora ranged between 75% and 30% (in stair-shaped dendrograms). The separation of P. collina and P. thyrsiflora was distinct in two tests. These two species were occasionally included into the Aureae sect. In 11 experiments, P. tabernaemontani and P. incana of the Aureae sect. were gathered to- gether, while P. argentea clustered out of this section, or was weakly associated with it. Cluster analysis, presented in the form of a final dendrogram grouped all species into two groups (Fig. 2). The first group is represented by two species, P. collina and P. thyrsiflora with the high genetic similarity of 0.88. The second, mixed cluster is formed by three species, i.e. P. argentea, P. tabernaemontani and P. incana. (supported by a 64% BS). Out of these species, P. tabernaemontani and P. incana are joined as a distinct minor subcluster with a low genetic similarity (0.35), supported by a bootstrap value of 89%. Interestingly, the analysis of species similarity showed that RAPD markers placed P. argentea (Terminales sect.) closer to P. tabernaemontani and P. incana (Aureae sect.) than to P. collina and P. thyrsiflora, both from the Terminales section (Fig. 2). The association of P. argentea and the subcluster containing P. tabernaemontani and P. incana was supported by a moderate bootstrap (BS) value of 64%. P. tabernaemontani and P. incana (classified in the same section) showed a similarity coefficient of 0.35 with the cluster supported by a bootstrap value of 81%. ACTA BOT. CROAT. 69 (1), 2010 77 MULTILOCUS GENOMIC ASSOCIATIONS OF POTENTILLA Tab. 3. Names, percentage of polymorphic bands and melting temperatures (Tm) of different primers used to RAPD analysis for Potentilla collina and P. thyrsiflora, and the putative parents, i.e. P. argentea, P. tabernaemontani and P. incana. Code Sequence 5’ to 3’ Percentage of polymorphic bands Tm (°C) S-01 GTCAGGGCAA 37.5 32 S-02 GGGCTCGTGA 54.5 34 S-03 GAGTCAGCAG 44.4 32 S-04 TGGGGGTCCC 33.3 36 S-05 GGCGGATAGC 50.0 34 S-06 CAGGGGTGGA 28.6 34 S-07 CCTGGGCCAC 44.5 36 S-08 CCCGCCTCCC 50.0 38 S-09 GAGCACTAGC 45.5 34 S-10 GAGCACGGGA 44.4 34 S-11 ATCTGCGAGC 50.0 32 S-12 TCCGATGCTG 42.9 32 S-13 ACCGTCGGCA 37.5 34 S-14 GCTTGCACCG 45.5 35 S-15 CTACCGTGGC 58.3 36 S-16 AGGGGCGGTC 33.3 36 S-17 GTCCACACGG 42.6 32 S-18 GTGTGAGAGA 37.5 31 S-19 CCAGTGCATG 44.4 31 43.4 (average) U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees Discussion In the investigation presented here, 39.1% of the RAPD fragments were polymorphic. Compared to other plant species, the members of Collinae subsect., therefore, showed a medium level of genetic variability. In a study of the widespread Tanacetum vulgare, 85% polymorphic bands were observed (), in the rare Vicia pisiformis 7.2% () and in apomictic species of Rosa sect. Caninae only 3% (). These differences are not astonishing, since the level of genetic variability strongly depends on the plant’s life history traits (). The species of the subsection are perennial, facultatively apomictic plants with a narrow ecological am- plitude and these biological characteristics all contribute to create and maintain the ob- served medium level of genetic variability. Within Terminales sect., our dendrogram shows P. argentea to be relatively isolated from the other members of the section analysed, P. collina and P. thyrsiflora (Fig. 2). Previ- ous morphological studies have also found the same differentiation between P. argentea and other members of the section (LEHT 1997). In earlier studies (KO£ODZIEJEK 2007, KO£ODZIEJEK and GABARA 2007, 2008), based on morphological analyses in P. subsect. Collinae material, i.e. P. collina, P. thyrsiflora, P. wimannania, P. leucopolitana, P. ´gabarae and P. koernickei, was identified at the species level. However, these conclusions do not fit our observations based on RAPD markers. The clustering obtained with morphological characters after research based largely upon her- barium material (KO£ODZIEJEK data not shown) was not altogether compatible with the dendrogram based on RAPD marker. The absence of a relationship between the morpho- logical and genetic similarities was also found for wild populations of other plants (GREENE et al. 2004, STEINER and SANTOS 2001). Several reasons may account for the discordance between the morphological traits and RAPD marker. First, molecular markers represent a sample of the plant genome, and even so, are used to make an inference concerning the whole genome. Second, morphological variation is strongly associated with environmental 78 ACTA BOT. CROAT. 69 (1), 2010 KO£ODZIEJEK J., CIEŒLIKOWSKI T., SAKOWICZ T. 1,00 0,95 0,90 0,85 0,80 0,75 0,70 0,65 0,60 0,55 0,50 1 (A+B+C) 2 (D+E) 3 (L+M) 4 (N+O) 5 (J+K) A v e ra g e d g e n o m ic s im il a ri ty Fig. 2. Taxonomic similarity among Potentilla collina, P. thyrsiflora, and the putative parents, i.e. P. argentea, P. thyrsiflora and P. incana – the final dendrogram. The averaged similarity among five species found on the basis of nineteen RAPD analyses. U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees variation; the morphological similarities observed may be due to different combinations of alleles producing similar phenotypes that might result in morphological similarities or dif- ferences that are not proportional to the underlying genetic differences. Third, the P. subsect. Collinae is a very variable species complex in some respects appearing intermedi- ate between P. argentea, P. tabernaemontani, P. incana and regarded by some authors as derived from hybridisation between them, the occurrence of polyploids and ability to re- produce both sexually and apomictically (MÜNTZING 1928, ASKER and FRÕST 1970, HOLM and GHATNEKAR 1996, GREGOR at al. 2002). Apparently, taxa of the Collinae subsect. occa- sionally hybridise with each other and with their parents (KURTTO et al. 2004). The similarity between P. leucopolitana and P. koernickei has been a matter of debate. These two species have been traditionally considered to be closely related, and P. koer- nickei have even been treated as forms within P. leucopolitana by some authors (e.g. WOLF 1908). In contrast, other authors have classified both taxa as separate species (ZIMMETER 1887). However, recent morfological studies (data not shown), clearly support the separa- tion of P. leucopolitana from P. koernickei. Our results also support these findings since the dendrogram separated both species. The RAPD variation in four other species of P. subsect. Collinae, i.e. P. praecox, P. alsatica, P. leucopolitana and P. alpicola and the putative parents, i.e. P. argentea, P. incana and P. tabernaemomtani has been investigated earlier GREGOR et al. 2002). Accord- ing to their research, P. argentea is probably a parental species of P. praecox, P. alsatica, P. leucopolitana and P. alpicola, and P. tabernaemontan might be the parent species of P. lindackeri. Based on chemotaxonomical studies, ASKER and FRÕST (1970) showed a close relationship between taxa from P. subsect. Collinae and P. argentea and P. tabernaemon- tani. According to theim, the taxa of P. subsect. Collinae could have emerged from crosses between P. argentea of the Terminales sect. on one side, and P. tabernaemontani or P. incana of the Aureae sect. on the other side. Even other taxa from the Aureae sect., such as P. tommasiniana F. W. Schultz and P. gaudinii Gremli (= P. pusilla Host), can participate in the emergence of new species belonging to P. collina (GUSTAFSSON 1947). The molecular data thus are consistent with the supposed intermediacy in morphological features, which has been cited to support the hybrid origin of the taxa of P. subsect. Collinae (ASCHERSON and GRAEBNER 1904, WOLF 1908, BALL et al. 1968). As has been noted to be the case for many hybrids and hybrid derivatives (RIESEBERG 1995), the morphology of the taxa of P. subsect. Collinae is not strictly intermediate between its putative parental species, but rather consists of a mixture of qualititative characters that match one or the other parental species as well as intermediacy in quantitative characters. Conclusion Morfological differences and traditional taxonomy practice seem to justify recognition of the six taxa P. collina, P. thyrsiflora, P. wimannania, P. leucopolitana, P. ´gabarae and P. koernickei as a separate taxa at species level. However, the grouping of nine populations representing the six species within Collinae subsect. in the RAPD trees, as partitioned in traditional taxonomic treatments, was not altogether compatible with the well defined spe- cies. The authors state (ASKER and FRÕST 1970, GREGOR et al. 2002). that molecular data agree with the intermediacy of species of Collinae subsect.between P. argentea (Termi- ACTA BOT. CROAT. 69 (1), 2010 79 MULTILOCUS GENOMIC ASSOCIATIONS OF POTENTILLA U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees nales sect.), P. tabernaemontani and P. incana (Aureae sect.), but our own results do not agree with this. Further analyses are needed to determine the correct infrageneric taxonomic treatment of the Collinae subsect. as it is outside the Terminales sect. in our and LEHT (1997) analyses. To obtain a deeper insight into the position of this species, it will be necessary to in- crease both the number of Potentilla populations and the number of specimens analysed. Acknowledgements Part of this work was supported by grant No. 2 P04F 040 26 from the Ministry of Sci- ence and Information Society Technologies (KBN). References ASCHERSON, P., GRAEBNER, P., 1904: Potentilla L. In: ASCHERSON, P., GRAEBNER, P. (eds), Synopsis der Mitteleuropäischen Flora 6, 664–872. Wilhelm Engelmann, Leipzig. ASKER, S., FRÕST, S., 1970: The »Potentilla collina problem« – a chemotaxonomic ap- proach. Hereditas 66, 49–70. BALL, P. W., PAW£OWSKI, B., WALTERS, S. M., 1968: Potentilla L. In: TUTIN, T. G., HEY- WOOD, V. H., BORGES, N. A., MOORE, D. M., VALENTINE, D. H., WALTERS, S. M., WEBB, D. A. (eds), Flora Europaea 2, 36–47. Cambridge University Press, Cambridge. BLACK-SAMUELSSON, S., ERIKSSON, G., GUSTAFSSON, L., GUSTAFSSON, P., 1997: RAPD and morphological analysis of the rare plant species Vicia pisiformis (Fabaceae). Biologi- cal Journal of the Linnean Society 61, 325–343. ERIKSSON, T, DONOGHUE, M. J., HIBBS, M. S., 1998: Phylogenetic analysis of Potentilla us- ing DNA sequences of nuclear ribosomal internal transcribed spacers (ITS), and impli- cations for the classification of Rosoideae (Rosaceae). Plant Systematics and Evolution 211, 155–179. FELSENSTEIN, J., 1985: Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, 783–791. GREENE, S.L., GRITSENKO, M., VANDEMARK, G., 2004: Relating morphologic and RAPD variation to collection site environment in wild populations of red clover (Trifolium pratense L.). Genetic Resources and Crop Evolution 51, 643–663. GREGOR, T., ROLLIK, J., WEISING, K., 2002: RAPD-Untersuchungen und Chromosomen- zählungen in der Potentilla-collina-Gruppe (Rosaceae). Berichte der Bayerische Bota- nische Gesellschaft 72, 159–167. GUSTAFSSON, A., 1947: Apomixis in higher plants, 2. The causal aspect of apomixis. Lunds Universitets Arsskrift 43, 71–178. HOLM, S., GHATNEKAR, L., 1996: Sexuality and no apomixis found in crossing experiments with diploid Potentilla argentea. Hereditas 125, 77–82. ILNICKI, T., KO£ODZIEJEK, J., 2008: Chromosome numbers of Potentilla subsect. Collinae (Rosaceae) from Poland. Caryologia 61, 170–175. JUZEPCZUK, S., 1941: Potentilla L. In: SHISHKIN, B. K., JUZEPCZUK, S. V. (eds), Flora URSS 10, 78–223 (In Russian). Academy of Science URSS, Moskow. 80 ACTA BOT. CROAT. 69 (1), 2010 KO£ODZIEJEK J., CIEŒLIKOWSKI T., SAKOWICZ T. U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees KESKITALO, M., LINDÉN, A., VALKONEN, J. P. T., 1998: Genetic and morphological diversity of Finnish tansy (Tanacetum vulgare L., Asteraceae). Theoretical and Applied Genetics 96, 1141–1150. KO£ODZIEJEK, J., 2004: Distribution of the Potentilla species in the northern part of the Landscape Park »Orle Gniazda« (Silesia-Cracow upland) (in Polish). Fragmenta Flo- ristica et Geobotanica Polonica 11, 263–270. KO£ODZIEJEK, J., 2007: Morphological analysis of carpel styles of Polish members of the Potentilla collina group (Rosaceae). Genus (Supplement) 14, 35–39. KO£ODZIEJEK, J., GABARA, B., 2007: Characteristics of achenes in Potentilla collina group (Rosaceae). Acta Societatis Botanicorum Poloniae 76, 35–42. KO£ODZIEJEK, J., GABARA, B., 2008: Palynological study of Polish taxa of Potentilla subsect. Collinae (Rosaceae). Acta Botanica Croatica 67, 139–146. KURTTO, A., LAMPINEN, R., JUNIKKA, L., (eds) 2004: Atlas Florae Europaeae. Distribution of vascular plants in Europe. Rosaceae (Spiraea to Fragaria, excl. Rubus) 13, 195–203. The Committee for Mapping the Flora of Europe and Societas Biologica Fennica Vanamo, Helsinki. LEHT, M., 1997: The genus Potentilla L. in Estonia, Latvia and Lithuania: distribution, morphology and taxonomy. Dissertationes Biologicae Universitatis Tartuensis 27, 1–186. MÜNTZING, A., 1928: Pseudogamie in der Gattung Potentilla. Hereditas 11, 267–283. MÜNTZING, A., 1931: Note on the cytology of some apomictic Potentilla-species. Hereditas 15, 166–178. NYBOM, H., BARTISH, I. V., 2000: Effects of life history traits and sampling strategies on ge- netic diversity estimates obtained with RAPD markers in plants. Perspectives in Plant Ecology, Evolution and Systematics 3, 93–114. ROMÁN, B., ALFARO, C., TORRES, A. M., MORENO, M. T., [ATOVI], Z., PUJADAS, A., RUBIALES, D., 2003: Genetic relationships among Orobanche species as revealed by RAPD analysis. Annals of Botany 91, 637–642. RIESEBERG, L. H., 1995: The role of hybridization in evolution: old wine in new skins. American Journal of Botany 82, 944–953. SAKOWICZ, T., CIEŒLIKOWSKI, T., 2006: Phylogenetic analyses within three sects of the ge- nus Vicia. Cellular Molecular Biology Letters. 11, 594–615. SOJÁK, J., 2005: Potentilla L. s.l. (Rosaceae) in Flora Europae Orientalis (Notes on Potentilla, 18). Candollea 60, 59–78. STEINER, J. J., LOS SANTOS, G. G., 2001: Adaptive ecology of Lotus corniculatus L. geno- types: I plant morphology and RAPD maker characterizations. Crop Science 41, 552–563. TAMURA, K., DUDLEY, J., NEI, M., KUMAR, S., 2007: MEGA4: Molecular Evolutionary Ge- netics Analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24, 1596–1599. WERLEMARK, G., NYBOM, H., 2001: Skewed distribution of morphological character scores and molecular markers in three interspecific crosses in Rosa sect. Caninae. Hereditas 134, 1–13. ACTA BOT. CROAT. 69 (1), 2010 81 MULTILOCUS GENOMIC ASSOCIATIONS OF POTENTILLA U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees WILLIAMS, J. G. K, KUBELIK, A. R., LIVAK, K. J., RAFALSKI, J. A., TINGEY, S. V., 1990: DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Research 18, 6531–6535. WOLF, T., 1908: Monographie der Gattung Potentilla L. Bibliotheca Botanica 71, 1–714. ZIMMETER, A., 1887: Schlüssel zur Bestinmung der deutschen, Österreichischungarischen und schweizer Arten der Gattung Potentilla. Botaniker-Kalender, 66–83. 82 ACTA BOT. CROAT. 69 (1), 2010 KO£ODZIEJEK J., CIEŒLIKOWSKI T., SAKOWICZ T. U:\ACTA BOTANICA\Acta-Botan 1-10\Kolodziejek2.vp 9. travanj 2010 12:43:21 Color profile: Disabled Composite 150 lpi at 45 degrees