American Journal of Botany 89(6): 929–939. 2002.
TACKLING
SPECIOSE GENERA: SPECIES COMPOSITION
AND PHYLOGENETIC POSITION OF
SENECIO
SECT.
JACOBAEA (ASTERACEAE) BASED PLASTID AND NRDNA SEQUENCES1
PIETER B. PELSER,2 BARBARA GRAVENDEEL, RUUD VAN DER MEIJDEN
ON
AND
Nationaal Herbarium Nederland, Universiteit Leiden, P.O. Box 9514, 2300 RA Leiden, The Netherlands The molecular phylogeny of Senecio sect. Jacobaea (Asteraceae; Senecioneae) was studied to clarify species composition and interspecific relationships of Senecio sect. Jacobaea. This information is necessary for studies seeking explanations of the evolutionary success of Senecio, in terms of high species numbers and the evolution of chemical defense mechanisms. Parsimony analyses with 60 species of the tribe Senecioneae, representing 23 genera and 11 sections of Senecio, based on DNA sequence data of the plastid genome (the trnT-L intergenic spacer, the trnL intron, and two parts of the trnK intron, flanking both sides of the matK gene) and nuclear genome (ITS1, 5.8S, and ITS2 gene and spacers) show that sect. Jacobaea is a strongly supported monophyletic group. Fifteen species have been identified as members of section Jacobaea, including three species that have been consistently ascribed to this section in taxonomic literature and 12 species that were either placed in other sections of Senecio or not exclusively ascribed to sect. Jacobaea. This section was traditionally circumscribed as a group of European, biennial, or perennial herbs with pinnately incised leaves, but the results of this study show that one annual species, a species from northeastern Asia, and a species growing in the Himalayas are members of sect. Jacobaea as well. Furthermore, not all species in the section have pinnately incised leaves. The genera Emilia, Packera, and Pseudogynoxys form the sister clade of sect. Jacobaea, but this relationship lacks strong bootstrap support and thus remains provisional. Key words:
Asteraceae; nr ITS; sect. Jacobaea; Senecio; Senecioninae; systematics; trnK; trnL-F.
Attempts to reconstruct the phylogeny of speciose genera are often complicated by severe difficulties in obtaining a representative taxon sampling. In order to arrive at a stable classification, ideally, all species of a speciose genus should be included. But restrictions in time and material make this an impossible task to complete in short-term research projects. Another option is to subdivide the genus into monophyletic groups and to study these separately. This might minimize the effects of poor taxon sampling on the analysis (Starr and Ford, 2001). Monographs with subdivisions into subgenera or sections could be a good starting point to select such groups (Reznicek, 1990). Bohs and Olmstead (1997), however, mention four problems with this strategy. Many large genera have no modern worldwide monographs and traditional taxonomic subdivisions often prove to be highly unnatural. Furthermore, many species in speciose genera are often not ascribed to subgeneric or sectional groups. Finally, even if the group under study is monophyletic, it is a very difficult task to select suitable outgroup species among the many other species in the
genus, because phylogenetic relationships to other groups are generally unknown. Crins (1990), Oliver and Oliver (1991), Linne von Berg et al. (1996), and Miller and Bayer (2001) reached similar conclusions for Carex, Erica, Allium, and Acacia, respectively. Senecio L. (Asteraceae; Senecioneae) is one of about 50 plant genera comprising over 500 species (Mabberley, 1997). Senecio has a worldwide distribution and many of its species are notorious for their poisonous pyrrolizidine alkaloids (PAs) that are responsible for killing more grazing livestock than all other poisonous plants together (Jeffrey, 1978). The large size of the genus (between 1000 and 3000 species; Jeffrey et al., 1977; Nordenstam, 1978; Bremer, 1994; Vincent, 1996) has troubled attempts to make an infrageneric classification of Senecio, and therefore, the evolutionary history of this genus is still poorly known. A second problem is the remarkable amount of morphological variation in habit, leaf shape and texture, indument and flower color, both between and within species of Senecio (Barkley, 1978). Moreover, interspecific hybridization is presumably widespread in this genus and other genera in the Senecionineae (Barkley, 1978, 1988; Bain and Jansen, 1995; Comes and Abbott, 1999). These problems have prevented attempts to make a modern worldwide monograph of Senecio and/or of its infrageneric groups. Furthermore, the majority of the approximately 150 sections of Senecio are not distinguished from one another by unique character states or syndromes of character states. This has resulted in quite a number of vague section circumscriptions and compositions that are extremely variable and not mutually exclusive. Sectional divisions in Senecio should therefore be regarded as informal groups of species that share a resemblance in their gross morphology rather than being putative monophyletic groups. Jeffrey tried to overcome these problems with the in-
Manuscript received 30 October 2001; revision accepted 3 January 2002. The authors thank J. F. Bain, R. J. Bayer, D. J. Crawford, J. W. Kadereit, E. B. Knox, J. L. Panero, and J. Francisco-Ortega for sending DNA samples of various Senecioneae species; G. Hol, M. Macel, A. Mahoney, P. C. M. van Steijn, G. A. van Uffelen, K. Vrieling, E. Wiebe, and various botanical gardens for providing seeds or tissue samples of many species used in the analyses; the director and curator of the Herbarium of Higher Plants and Mosses Collections of the Bulgarian Academy of Sciences for the loan of herbarium specimens of S. pancicii; T. M. Barkley, T. F. Stuessy, P. L. D. Vincent, and many others for advice; M. C. M. Eurlings, B.-J. van Heuven, H. W. Nell, J. W. Pelser, K. A. D. Pelser, and W. Star for technical assistance; and J. D. Jellema, A. P. T. M. Vogel, and other employees of the Hortus Botanicus Leiden for excellent care of the Senecioneae plants grown in their gardens and greenhouses. 2 Author for reprint requests (e-mail:
[email protected]). 1
929
930
AMERICAN JOURNAL
frageneric classification of Senecio by selecting ‘‘modal species around which other species might reliably be clustered. . . ’’ (Jeffrey et al., 1977, p. 48). In this approach, groups of presumably related species were made on the basis of macro- and micromorphological characters and ‘‘some fifteen years’ working knowledge of the group’’ (Jeffrey et al., 1977, p. 48). These groups are supposedly suitable operational taxonomic units to use in following phylogenetic analyses (Jeffrey, 1992). Such analyses were, however, never published and it thus remains to be tested if Jeffrey’s classification reflects the evolutionary history of Senecio. Although the phylogeny of the entire genus Senecio has never been studied, there are several studies of the phylogeny of selected groups within the tribe Senecioneae. These studies, based on both morphological and molecular data, show that Senecio is paraphyletic or even polyphyletic (Knox and Palmer, 1995; Kadereit and Jeffrey, 1996). Because classifications should be based on groups of evolutionary historical reality to be of universal value for biological research in general (Sanders and Judd, 2000), monophyletic satellite genera such as Kleinia, Packera, and Dendrosenecio are nowadays usually split off from Senecio sensu lato (s.l.) in the process to arrive at a monophyletic delimitation of Senecio sensu stricto (s.s.) (Jeffrey et al., 1977; Jeffrey, 1979; Bremer, 1994). Until now, however, the generic classification in the Senecioneae is largely unresolved. Because of its enormous size, studies on the evolutionary history of Senecio s.l. should be confined to infrageneric groups of this genus of which the phylogenetic history can be reconstructed within a relatively short period of time. The assemblage consisting of Senecio jacobaea and closely related species is a suitable candidate for such studies. Senecio jacobaea is a widespread herb, sometimes a noxious weed, with its main distribution in western Eurasia. This species has been introduced into North and South America, South Africa, Australia, and New Zealand (Harper and Wood, 1957; Bain, 1991). Senecio jacobaea is an important pest in many countries and is well known for its toxic PAs (Bain, 1991). Therefore, information about the phylogeny of the group of S. jacobaea and closely related species can provide important insights into the evolution of PAs in this assemblage. Senecio jacobaea is the type species of sect. Jacobaea (Mill.) Dumort. This section is usually circumscribed as a section of perennial or rarely biennial herbs, with a subglabrous to floccose indument. The leaves of these plants are more or less pinnately incised. Supplementary bracts are usually more than one-quarter as long as the involucre of the flower heads. The achenes are subcylindrical and hairy or glabrous (Chater and Walters, 1976). Unfortunately, section descriptions of sect. Jacobaea by different authors (e.g., Chater and Walters, 1976; Jeffrey, 1992; Shishkin, 1995) are well applicable to many species from other sections of Senecio s.s. This resulted in many different views of the species composition of sect. Jacobaea (see Table 1 for an overview of the main taxonomic subdivisions). Three species are generally regarded to form the core of sect. Jacobaea: S. aquaticus, S. erucifolius, and S. jacobaea (Table 1). These species have always been ascribed to sect. Jacobaea in taxonomic literature and have never been placed in other sections. Additionally, numerous other species have been more or less frequently associated with sect. Jacobaea. Many of these species are likely to be synonyms of one of the former species or prove to be only distantly related to S. ja-
OF
BOTANY
[Vol. 89
cobaea. The species composition of sect. Jacobaea as is used by different authors overlaps with those of many other mainly European sections of Senecio s.s., but shows most overlap with sect. Senecio s.l. (including sect. Annui, Obaejaca, Obaejacoideae, Vernales, and Vulgares). This is because species of sect. Senecio s.l. show striking morphological resemblances to species of sect. Jacobaea: plants of both sections are European herbs with similarly shaped pinnately divided leaves. In fact, the only distinguishing character between both sections that is mentioned in literature is their life history strategy. Plants from sect. Jacobaea are presumed to be perennial, while most representatives of sect. Senecio s.l. are annuals. This lack of distinguishing characters between both sections has led Alexander (1979) to include the species of sect. Jacobaea in sect. Senecio s.l. Recent studies using plastid restriction fragment length polymorphism (RFLP) data (Knox and Palmer, 1995; Kadereit and Jeffrey, 1996), however, suggest a remarkably different phylogenetic position of sect. Jacobaea than would be expected from morphological data. In these RFLP studies, sect. Jacobaea appears to be only distantly related to other mainly European sections of Senecio s.s. Relatively close allies seem to be the succulent sect. Rowleyani and the genera Delairea, Gynura, Kleinia, Packera, and Solanecio. Unfortunately, these relationships are only weakly supported and the result of studies with a very poor taxon sampling with respect to members of sect. Jacobaea, due to their aim to reconstruct the phylogeny of other taxonomic groups in the tribe Senecioneae (Dendrosenecio and Senecio sect. Senecio s.l., respectively). The main aims of this study were to use phylogenetic analyses of DNA sequence data to investigate (1) the species composition and (2) the interspecific relationships of sect. Jacobaea and (3) to study the phylogenetic position of this section in the tribe Senecioneae. We used a broad taxon sampling strategy and DNA sequence data from the trnT-L intergenic spacer, the trnL intron, two parts of the trnK intron, flanking both sides of the matK gene (plastid genome), and the ITS1– 5.8S-ITS2 region (nuclear genome). The trnT-L and ITS regions were selected because they have proven their usefulness in phylogenetic studies at the species and genus levels in Senecioneae (e.g., Bain and Golden, 2000; Alvarez Fernandez et al., 2001). The trnK region was chosen because this region has been used for phylogenetic reconstructions at a variety of taxonomic levels in angiosperms, among which is Asteraceae (Denda et al., 1999). MATERIALS AND METHODS Taxon selection—A total of 60 taxa were selected to represent the three core species of sect. Jacobaea and species from other sections of Senecio s.s. and genera in the tribe Senecioneae. The selection includes species from 23 genera representing all three subtribes of Senecioneae (Blennospermatinae, Tussilagininae, and Senecioninae, sensu Bremer, 1994). Blennosperma nanum was used as outgroup. This species belongs to the subtribe Blennospermatinae, which is assumed to be a sister group to the other two subtribes (Bremer, 1994; Barkley, 1999; but see Alvarez Fernandez et al., 2001). Sources of plant material—Plant material was obtained from various sources. Most plants were grown from seeds in the Hortus Botanicus Leiden. These seeds were acquired from seedbanks of botanical gardens and private collections. Some plants were collected in the field. Tissue samples of these plants were preserved on silicagel until DNA extraction. Other accessions included DNA or tissue samples kindly contributed by coworkers in Senecioneae and herbarium specimens from L and SOM (acronyms following Holm-
June 2002]
PELSER
ET AL.—MOLECULAR PHYLOGENY OF
gren, Holmgren, and Barnett [1990]). Voucher specimens are archived on the American Journal of Botany Supplementary Data Site (http://ajbsupp.botany. org/v89/pelser.xls).
DNA extraction, PCR amplification, and sequencing—Total genomic DNA was extracted from tissue samples according to the method of Doyle and Doyle (1987) or using DNeasy Plant Mini Kit (QIAGEN, Leusden, The Netherlands). In total, four plastid regions (the trnT-L intergenic spacer, the trnL intron, and two regions of the trnK intron) and three nuclear regions (ITS1, 5.8S, and ITS2) were sequenced. Polymerase chain reaction (PCR) and sequencing amplification of the trnTL intergenic spacer and trnL intron were performed with primers a, b, c, and d, designed by Taberlet et al. (1991). In total 36 cycles of PCR amplification were carried out under the following conditions: denaturation for 45 s at 948C, annealing for 45 s at 488C, and extension for 2 min at 728C. The trnT-L intergenic spacer was only sequenced for 37 taxa due to PCR amplification problems. The 59 and 39 trnK intron (flanking sequences of matK) were partially sequenced using primers 39F (59-TGCGGCTAGGATCTTTTACACA-39), 546R (59-TTTTTCAACCCAATCGCTCTTT-39), 1023F (59-GATTTGGGCCGATTTCTC-39), and 1559R (59-GCACACGGCTTTCCCTCTG-39), designed by the authors. The conditions were as follows: denaturation for 1 min at 948C, annealing for 30 s at 528C, and extension for 1 min at 728C. Both regions were PCR amplified in 29 cycles. Internal transcribed spacer (ITS) regions 1 and 2 and the 5.8S gene were PCR amplified and sequenced with primers ITS4 and ITS5 (White et al., 1990). The PCR amplification conditions were: 36 cycles, denaturation for 1 min at 948C, annealing for 1 min at 528C, and extension for 1 min at 728C. The sequenced samples were run on an ABI 377 automated sequencer (Applied Biosystems, Nieuwerkerk a/d IJssel, The Netherlands) using standard dye-terminator chemistry following the manufacturer’s protocol. All DNA sequences were submitted to GenBank (see http://ajbsupp.botany.org/v89/ pelser.xls for accession numbers) and Treebase.
DNA sequence alignment—DNA sequences were aligned using the ‘‘Clustal’’ option in Megalign 4.03 (DNAstar, 1999, Madison, Wisconsin, USA) and by subsequent manual adjustment where necessary. The aligned data sets are available from the first author. Sequence limits of the trnL intron, ITS1, 5.8S, and ITS2 were determined with known sequences of trnL, 18S, 5.8S, and 25S (Shinozaki et al., 1986; Baldwin, 1992; Kim and Jansen, 1994). Pairwise sequence divergence values were calculated with PAUP* version 4.0b8 (Swofford, 2001).
Phylogenetic analyses—Phylogenetic analyses were performed with PAUP*, using the maximum parsimony algorithm with the option ‘‘Heuristic search.’’ Regions of the alignment in which homology could not be assessed reliably were excluded from the analyses (Table 2). Transitions and transversions were equally weighted. The data matrix for the phylogenetic analyses was composed of the DNA sequence alignment in which insertion/deletion events were treated as missing data, and an additional matrix in which gaps were binary coded for their presence or absence. Initial trees in the heuristic search were built ten times with a random addition sequence of the taxa. The options tree bisection-reconnection (TBR) branch-swapping, ACCTRAN optimization, and MULPARS were invoked. Bootstrap support (BS) for monophyletic groups (Felsenstein, 1985) was calculated with 5000 bootstrap replicates, using the same settings as for the general heuristic search analyses, but holding only ten trees per replicate. Decay values (Bremer, 1988; Donoghue et al., 1992) were calculated with AutoDecay 4.0.2 (Eriksson, 1998) and PAUP* using 100 addition sequence replicates for each run. All seven DNA regions were analyzed separately and combined. Partition homogeneity tests (1000 replications, saving ten trees per replicate; Farris et al., 1995), as is implemented in PAUP*, were used to test for incongruence between the phylogenetic signal of individual data sets.
SENECIO
SECT.
JACOBAEA
931
RESULTS Sequence variation—The amplified trnL intron regions are between 421 base pairs (bp) (S. ambiguus, S. chrysanthemoides, S. cineraria, S. jacobaea, S. pancicii, and S. subalpinus) and 451 bp (Emilia coccinea) in length. The length of ITS1 varies between 211 bp (S. oxyriifolius) and 269 bp (Blennosperma nanum). The 5.8S region has a constant length of 164 bp for all species examined. ITS2 has a length of 213 (Blennosperma nanum)–225 bp (S. viscosus) for the species sequenced in this study. The trnT-L intergenic spacer and the flanking trnK intron regions of matK were only partially amplified. Due to amplification problems, we could not sequence all DNA regions for some species (Table 2). Pairwise sequence divergence values for the combined plastid and nuclear data are between 0.1% (between S. alpinus and S. subalpinus) and 14.0% (between Blennosperma nanum and S. lineatus). See Table 2 for character statistics. Phylogenetic analyses—Partition homogeneity analyses showed that all sequenced plastid regions (the trnT-L intergenic spacer, the trnL intron, and the two parts of the trnK intron) are congruent (P 5 0.38), as is also true for the nuclear partitions (ITS1, 5.8S, and ITS2; P 5 0.25). There is, however, significant incongruence between the plastid data set and the nuclear dataset (P 5 0.0001). Parsimony analysis of the combined plastid regions yielded more than 200 000 most parsimonious trees (MPTs) with a length of 519 steps, a consistency index (CI) of 0.78 (0.63 excluding uninformative characters) and a retention index (RI) of 0.82. Analysis of the combined nuclear data resulted in 640 MPTs with a length of 1591 steps, a CI of 0.46 (0.42 excluding uninformative characters) and an RI of 0.68 (Table 3; cladograms not shown). Visual inspection of the bootstrap consensus trees of the plastid data set and the nuclear data set (Fig. 1) showed that most topological differences between both consensus trees are very weakly supported. In fact, only one incongruent taxon placement is supported with a BS over 70%. This concerns the placement of either the clade S. cannabifolius–S. paludosus or the clade S. abrotanifolius–S. adonidifolius–S. incanus–S. minutus as the most basal group in sect. Jacobaea (as delineated in Fig. 2). Other events of incongruence are supported with lower BS and are therefore considered to be insignificant (Hillis and Bull, 1993). A combined analysis of the plastid data set and the nuclear data set (Yoder, Irwin, and Payseur, 2001) resulted in a total of 18 MPTs. These trees require 2168 character state changes and have a CI of 0.53 (0.44 excluding uninformative characters) and an RI of 0.69 (Table 3). A strict consensus tree accompanied by BS values, decay indices, and branch lengths is given in Fig. 2. This cladogram shows that the clade S. abrotanifolius–S. adonidifolius–S. incanus–S. minutus is the most basal group in sect. Jacobaea. DISCUSSION Phylogenetic utility of the DNA regions sequenced—The seven sequenced DNA regions showed significant differences in sequence variation between the species and in the number of potentially phylogenetic informative positions (Table 2). The four sequenced plastid regions (the trnT-L intergenic spacer, the trnL intron, and the two parts of the trnK intron) were much more conserved (between 12.2 and 19.9% variable po-
932 TABLE 1.
AMERICAN JOURNAL
OF
BOTANY
[Vol. 89
Overview of several alternative sectional classifications of a selection of Senecio species used in our analyses. Reichenbach (1831–1832)
Dumortier (1827)
Sect. Jacobaea (Mill.) Dumort.
Sect. Cinerarioideae Rouy Sect. Cordati Rchb.f.
S. S. S. S. S. S. S.
adonidifolius aquaticus doria erucifolius jacobaea paludosus squalidus — —
Sect. Doria (Fabr.) Rchb. s.l.
—
Sect. Delphinifolius Rchb.f. Sect. Ecalyculati DC Sect. Extremiorientalis Schischk. Sect. Incani DC
— —
Ser. Indici DC Sect. Scaposi O. Hoffm. Sect. Senecio s.l.
Sect. incert.
S. S. S. S. S.
—
S. S. S. S. S. S.
— — S. S. S. S.
— — — — S. sylvaticusr S. viscosusr S. vulgaris
abrotanifolius aquaticus erucifolius incanus jacobaea
De Candolle (1838)
alpinus doria nemorensis paludosus — — — —
S. S. S. S.
— — squalidus sylvaticus viscosus vulgaris —
abrotanifolius adonidifoliusf aquaticus cannabifolius erucifolius jacobaea — —
S. dorian S. nemorensisn S. paludosusn
Reichenbach fil. (1853)
S. S. S. S.
abrotanifolius aquaticus erucifolius jacobaea
S. S. S. S. S.
— alpinus subalpinus doria nemorensis paludosus
— S. alpinus S. S. S. S. S. S. S. S.
— ambiguus cineraria incanus chrysanthemoides — minutuss sylvaticusr viscosusr vulgarisr —
— — — S. incanus
S. S. S. S.
— — squalidusu sylvaticusv viscosusv vulgarisv —
Boissier (1875)
S. S. S. S. S. S. S.
abrotanifoliusd,t ambiguusi aquaticusa,i cinerariab,i erucifoliusi jacobaeai subalpinusi — —
S. nemorensisi S. paludosusj
Hoffmann (1890–1894)
S. aquaticusa S. erucifolius S. jacobaea
— — S. nemorensisk S. paludosusk
— —
— —
— —
— S. cinerariab
— — S. sylvaticusr S. viscosusr S. vulgarisr
— — S. sylvaticust S. vulgarist
—
—
Notes in superscript following species names: included by author as a: S. erraticus; b: S. bicolor; c: S. carniolicus; d: S. carpathicus; e: S. cordatus; f: S. artemisiaefolius; g: S. fuchsii; h: subspecies of S. jacobaea. Included by author in i: sect. Jacobaea Boiss.; j: sect. Crociserides DC.; k: sect. Crociseris (Rchb.) Hallier & Wohlf.; l: sect. Oliganthi Boiss.; m: sect. Pseudo-Oliganthi Sofieva; n: sect. Sarracenici DC.; o: sect. Umbrosae Rouy; p: sect. Incanae O. Hoffm.; q: sect. Carniolici Schischk.; r: sect. Obaejaca (Cass.) Dumort.; s: sect. Obaejacoideae DC.; t: sect. Annui O. Hoffm.; u: sect. Vernales Rchb.f.; v: sect. Vulgares Rchb.f.
sitions [VP]) than the ITS1 and ITS2 sequences (64 and 63.1% VP, respectively). The 5.8S gene shows a similar VP percentage as the plastid regions (18.9%). These findings correlate with the differences in the number of potentially phylogenetic informative positions (PIP) between the plastid regions (between 5.5 and 7.6% PIP), ITS1 and ITS2 (51.7 and 47.5% PIP, respectively), and 5.8S (9.8% PIP). On average, the plastid data set contains 6.6% PIP, whereas in the nuclear data set, on average 54.5% of the positions are potentially informative. These differences are reflected in the results of the parsimony analyses of the plastid and nuclear data sets: the nuclear data set yields better resolved cladograms than the plastid data set (Figs. 1 and 2). The characters of the plastid data set, however, show less homoplasy (CI of 0.78 and RI of 0.82) than those of the nuclear data set (CI of 0.46 and RI of 0.68). Our findings of the percentage of VP and PIP and the CI and RI values are in accordance with the results of other studies in Asteraceae using nr ITS and/or trnL-F sequences (e.g., Baldwin, 1992; Kim and Jansen, 1994; Panero et al., 1999; Bain and Golden, 2000; Francisco-Ortega et al., 2001). Alvarez Fernandez et al. (2001) found similar differences in the number of PIP and the extent of homoplasy between nr ITS and trnL-F sequences in their study of Doronicum (Senecioneae). Species composition of sect. Jacobaea—A total of 14 taxa included in our study prove to be closely related to S. jacobaea. Two of these species, together with S. jacobaea, were already generally considered to form the core of sect. Jaco-
baea (Table 1). These two species are S. aquaticus and S. erucifolius, which have a Eurasian distribution. The other 12 taxa were previously ascribed to other sections of Senecio s.s. or have not been consistently placed in sect. Jacobaea. These taxa include S. alpinus, S. pancicii, and S. subalpinus, three morphologically similar taxa growing in central and east European mountainous regions. All three species have been ascribed to sect. Jacobaea, but S. alpinus and S. subalpinus have also been placed in other sections (Table 1). Senecio abrotanifolius and S. adonidifolius are two European species that are frequently ascribed to sect. Jacobaea (Table 1). These species prove to be closely related to S. jacobaea as well. Senecio minutus is the only annual species in this group of close relatives of S. jacobaea. This species grows in south and central Spain and was previously placed in sect. Senecio s.l. (De Candolle, 1838) and sect. Delphinifolius (Chater and Walters, 1976) (Table 1). Senecio minutus shares the finely pinnatisect leaf shape of S. abrotanifolius and S. adonidifolius. This character is also present in S. delphinifolius, a species that was ascribed by Muschler (1909) and Konechnaya (1981) to sect. Jacobaea and that could not be included in our analyses. Senecio cannabifolius grows in northeastern Asia and was placed by De Candolle (1838), Kitamura (1942), and Koyama (1969) in sect. Jacobaea (Table 1). Senecio paludosus has been listed in sect. Jacobaea (Dumortier, 1827; Kitamura, 1942; Koyama, 1969), but is more commonly regarded as a member of sect. Doria s.l. (including. sect. Crociserides, Crociseris, Oliganthi, Pseudo-Oliganthi, Sarracenici, and Umbrosae; Table 1), because of its entire
June 2002] TABLE 1.
PELSER
SENECIO
SECT.
JACOBAEA
933
Extended.
Rouy (1903)
S. S. S. S. S. S.
ET AL.—MOLECULAR PHYLOGENY OF
adonidifolius aquaticush cineraria erucifolius incanus jacobaea
S. alpinus — S. doriao S. nemorensisg,o S. paludosusj
Chater and Walters (1976)
Koyama (1969)
S. S. S. S. S. S. S. S. S. S. S. S.
abrotanifolius adonidifolius alpinus aquaticus cannabifolius chrysanthemoides doria erucifolius jacobaea nemorensis paludosus subalpinus
S. S. S. S. S. S. S. S. S. S.
abrotanifolius adonidifolius alpinuse aquaticus carpetanus erucifolius jacobaea pancicii squalidus subalpinus
— — —
— — S. doria S. nemorensis S. paludosusk
— — — —
— — — —
— — S. sylvaticusr S. viscosusr S. vulgarisr
— — —
S. minutus — — S. ambiguusp S. cinerariab,p S. incanusp — — S. sylvaticus S. viscosus S. vulgaris
—
—
—
Konechnaya (1981)
S. S. S. S. S. S. S.
Jeffrey (1992)
abrotanifolius aquaticus cineraria erucifolius incanusc jacobaea subalpinus
— — S. doria S. nemorensism S. paludosusj
S. S. S. S. S. S.
ambiguus cineraria erucifolius incanus jacobaea nemorensis
— — S. doriak S. paludosusk
— — — —
— — — —
— — S. sylvaticus S. viscosus S. vulgaris
— — S. sylvaticus S. viscosus S. vulgaris
—
leaves (Boissier, 1875; Hoffmann, 1890–1894; Chater and Walters, 1976; Jeffrey, 1992; Shishkin, 1995). This species has a Eurasian distribution. Senecio chrysanthemoides is a species that grows in the Himalayas. De Candolle (1838) placed this taxon in his series Indici because of its geographic distribution. Hooker (1881) and Koyama (1969) placed S. chrysanthemoides in sect. Jacobaea (Table 1). Finally, three species formerly ascribed to sect. Incani are
S. abrotanifolius S. adonidifolius
Shishkin (1995) a
S. aquaticus S. erucifolius S. jacobaea
S. S. S.
— — doriaj nemorensism paludosusj subalpinusj — — cannabifolius cineraria incanusc,q
S. S. S. S.
— abrotanifoliusd sylvaticusr viscosusr vulgarisr
S. S. S. S.
—
Present study (2001)
S. S. S. S. S. S. S. S. S. S. S. S. S. S. S.
abrotanifolius adonidifolius alpinus ambiguus aquaticus cannabifolius chrysanthemoides cineraria erucifolius incanus jacobaea minutus paludosus pancicii subalpinus — — S. carpetanus S. doria S. nemorensis — — — —
S. S. S. S.
— — squalidus sylvaticus viscosus vulgaris —
grouped together with core members of sect. Jacobaea. Senecio incanus from the central and east European mountain areas is the type species of sect. Incani. Senecio incanus was already ascribed by Reichenbach (1831–1832), Rouy (1903), and Jeffrey (1992) to sect. Jacobaea (Table 1). Konechnaya (1981) placed S. carniolicus in sect. Jacobaea. This taxon is closely related to S. incanus and sometimes regarded to be a subspecies of S. incanus (e.g., Chater and Walters, 1976). Senecio cineraria is usually ascribed to sect. Incani as well (Ki-
TABLE 2. Number of sequenced species of Senecioneae, alignment length, number of excluded positions due to alignment problems, transition/ transversion rates calculated on the strict consensus tree of the combined plastid and nuclear data sets (Fig. 2), number of variable positions and potentially phylogenetic informative positions and number of potentially phylogenetic informative gaps.
Region
TrnL-F intergenic spacer TrnL intron TrnK intron (59 of matK) TrnK intron (39 of matK) Total plastid data ITS1 5.8S ITS2 Total nuclear data Total plastid and nuclear data
No. sequenced species
Alignment length (bp)
Exclude no. positions
Transition/ transversion rate
37 55 60 57 60 56 56 56 56 60
754 549 566 641 2510 354 164 296 814 3324
122 0 18 129 269 4 0 1 5 274
0.45 1.05 0.58 0.67 0.69 1.28 14 1.66 5.65 1.24
No. variable positions
126/632 79/549 67/548 70/512 342/2241 224/350 31/164 186/295 441/809 783/3050
(19.9%) (14.4%) (12.2%) (13.7%) (15/3%) (64%) (18.9%) (63.1%) (54.5%) (25.7%)
No. informative positions
48/632 30/549 35/548 35/512 148/2241 181/350 16/164 140/295 337/809 485/3050
(7.6%) (5.5%) (6.4%) (6.8%) (6.6%) (51.7%) (9.8%) (47.5%) (41.7%) (15.9%)
No. informative gaps
Total no. informative characters
6 7 3 4 20 7 0 9 16 36
54 37 38 39 168 188 16 149 353 521
934
AMERICAN JOURNAL
OF
BOTANY
[Vol. 89
TABLE 3. Values and statistics from parsimony analyses of the plastid data set, nuclear data set, and the combined (total analysis) data set for species of Senecioneae.
Combined plastid dataa
No. most parsimonious trees Length of most parsimonious trees Consistency index (excluding uninformative characters) Retention index No. clades with bootstrap support .70% No. clades with bootstrap support 50–69% a b
.200 000 519 0.78 (0.63) 0.82 12 9
Combined nuclear datab
640 1591 0.46 (0.42) 0.68 19 9
Combined plastid and nuclear data (total analysis)
18 2168 0.53 (0.44) 0.69 23 11
trnT-L intergenic spacer, trnL intron, and two parts of trnK intron. ITS1, 5.8S, and ITS2.
tamura, 1942; Chater and Walters, 1976; Shishkin, 1995), but Boissier (1875), Rouy (1903), Konechnaya (1981), and Jeffrey (1992) suggested that this species is better placed in sect. Jacobaea (Table 1). Senecio ambiguus is morphologically very similar to S. cineraria and was ascribed by Jeffrey (1992) to sect. Jacobaea (Table 1). Both species have a Mediterranean distribution. Senecio cineraria is also locally naturalized elsewhere. Our results show that sect. Incani is not monophyletic. This section is characterized by the tomentose, grey indument of the leaves, stems, and capitula, a character that proves to be homoplasious when plotted on the molecular phylogeny presented here. We propose here to include the 12 species mentioned above in sect. Jacobaea. These species are closely related to the core species of sect. Jacobaea and form together a strongly supported monophyletic group (99% BS). Unfortunately, due to a lack of plant material of some species and the large number of species, not all European taxa of Senecio could be included in our analyses. Because of this, and the fact that macromorphological synapomorphies could not be found for sect. Jacobaea in the present study, additional members of this section might be identified in future studies. Phylogeny of sect. Jacobaea—The typical element of sect. Jacobaea is formed by eight very closely related taxa: S. alpinus, S. ambiguus, S. aquaticus, S. chrysanthemoides, S. cineraria, S. jacobaea, S. pancicii, and S. subalpinus. Phylogenetic relationships between these taxa could not be resolved satisfactorily, due to a high degree of genetic uniformity in both the plastid and nuclear DNA regions studied (average pairwise sequence divergence between 0.1 and 0.8%). Senecio erucifolius is placed basal to this clade of closely related species. There are two alternative hypotheses about the most basal clade of sect. Jacobaea. The analysis of the plastid data set (trnT-L intergenic spacer, the trnL intron, and two parts of the trnK intron) suggests that the clade composed of S. cannabifolius and S. paludosus is the sister group of all other species in sect. Jacobaea. The nuclear ITS data, however, indicate that the clade formed by S. abrotanifolius, S. adonidifolius, S. incanus, and S. minutus is the most basal clade of the section. Both competing hypotheses are supported with moderate to high BS (79 and 95%, respectively). Separate parsimony analysis of the plastid data set yielded more than 200 000 MPTs and could not be swapped to completion. A heuristic search through the nuclear data set resulted in 640 MPTs. A combined data set of plastid and nuclear data supports the basal placement of the S. abrotanifolius–S. adonidifolius–S. incanus–S.
minutus clade with a BS of 91%. This combined analysis results in fewer MPTs (only 18), which are more resolved, and the resulting strict consensus tree also has a better resolution, when compared with analyses of the individual data sets. Furthermore, such a combined analysis gives a higher overall BS (Table 3; Figs. 1–2). We are therefore in favor of combining both data sets, despite the results of the partition homogeneity test (Soltis et al., 1998; Yoder, Irwin, and Payseur, 2001). Detailed morphological studies have not been performed in the present study and macromorphological synapomorphies for sect. Jacobaea have not been identified. Nonetheless, species resembling each other in general habit are grouped together (e.g., S. bicolor and S. ambiguus). More elaborate studies need to be performed to get a better view of the systematic value of morphological characters in sect. Jacobaea. All species in sect. Jacobaea, except for S. cannabifolius and S. chrysanthemoides, have their main distribution in Europe. Phylogeographic patterns in this group are, however, not distinguishable. Phylogenetic position of sect. Jacobaea within the subtribe Senecioninae—The results of the present study are not conclusive about the sister group relationships of sect. Jacobaea. The genera Emilia, Packera, and Pseudogynoxys form the sister clade of sect. Jacobaea in the strict consensus cladogram of the combined plastid and nuclear data (Fig. 2), but this relationship is only weakly supported (,50% BS). Knox and Palmer (1995), using plastid RFLP data, identified a sister group relationship between sect. Jacobaea and a succulent group of Senecioninae (Senecio sect. Rowleyani and the genus Delairea). This last group formed together with species of Gynura, Kleinia, Packera, and Solanecio a paraphyletic group in which sect. Jacobaea is nested. Our data, however, support the monophyly of the clade composed of Gynura, Kleinia, Solanecio (termed the ‘‘Gynuroid complex’’ by Jeffrey [1986]), and the succulent Senecio s.s. sections Delairea, Peltati, and Rowleyani. The supposed close relationship between this clade and sect. Jacobaea cannot be confirmed here. Kadereit and Jeffrey (1996) also included Emilia, Kleinia, and Pseudogynoxys in their plastid RFLP analysis. In their study, Kleinia groups together with members of sect. Senecio s.l. and sect. Doria s.l. Emilia and Pseudogynoxys are placed basal to this clade and the clade formed by members of sect. Jacobaea. Their analysis is, however, not accompanied by branch support values, so that we have no indication of the statistical support for these findings. In the most recent phylogenetic survey of the genus Packera, Bain and Golden (2000) found that their accession of S.
June 2002]
PELSER
ET AL.—MOLECULAR PHYLOGENY OF
SENECIO
SECT.
JACOBAEA
935
Fig. 1. Bootstrap consensus cladograms for Senecioneae of the plastid data set (the trnT-L intergenic spacer, the trnL intron, and 59 and 39 trnK introns) and the nuclear data set (ITS1, 5.8S, and ITS2). Five thousand bootstrap replicates, holding ten trees per replicate. Bootstrap values are indicated above the branches. (A) Bootstrap consensus cladogram of the plastid data set. (B) Bootstrap consensus cladogram of the nuclear data set. Accessions of Lordhowea insularis, S. anethifolius, S. incanus, and S. medley-woodii are missing because of amplification problems. S. 5 Senecio.
jacobaea is sister to Packera and Pericallis. Although this relationship is well supported (87% BS), taxon sampling of Old World Senecio s.l. species was confined to S. jacobaea and Pericallis cruenta in their study. According to the results of our analyses, Pericallis groups together with Cineraria, Dendrosenecio, S. lineatus, and S. scandens, and this clade is placed much more basal in the Senecioninae. These findings are, however, only weakly supported (,50% BS). Pericallis and Packera share a ‘‘helianthoid’’ pollen type (Skvarla et al.,
1977; Bain and Walker, 1995) in which the exine is filled with small holes or internal foramina. This character is more typical for species from the subtribe Tussilagininae, as representatives of the Senecioninae typically have ‘‘senecioid’’ pollen without these holes (Bain, Tyson, and Bray, 1997; Bain and Golden, 2000). Determination of the pollen type of species of sect. Jacobaea might provide additional information about the phylogenetic position of this section. A close relationship between sect. Jacobaea and Packera is also suggested by the delimi-
936
AMERICAN JOURNAL
OF
BOTANY
[Vol. 89
Fig. 2. Strict consensus cladogram for Senecioneae of 18 MPTs of the combined plastid and nuclear data. Bootstrap values and decay indices are indicated above the branches. Minimum assigned branch lengths are given below the branches. Length of the MPTs 5 2168, CI 5 0.53 (including uninformative characters; 0.44 excluding them), RI 5 0.69.1, species not ascribed to sections in literature;2, sensu Jeffrey (1993);3, sensu Chater and Walters (1979). Distribution data: Af: Africa, Am: America, As: Asia, Au: Australia, EA: Eurasia. S. 5 Senecio.
tation of sect. Jacobaea by Chater and Walters (1976). They included Packera cymbalaria under its synonym S. resedifolius in sect. Jacobaea. Phylogenetic analyses based on nuclear ITS sequence data (Bain and Jansen, 1995; Bain and Golden, 2000), however, show that Packera cymbalaria is nested within Packera. In contrast to the overall macromorphological resemblance between species of sect. Jacobaea and sect. Senecio s.l. (Alexander, 1979), molecular data presented here and by Knox and Palmer (1995) and Kadereit and Jeffrey (1996) clearly indicate that both sections are only distantly related. Our molecular data show that sect. Senecio s.l., represented in this paper by S. squalidus, S. sylvaticus, S. viscosus, and S. vulgaris, is more closely related to representatives of sections
from Australia, North America, and Africa. Senecio squalidus and S. nebrodensis (the latter one not included in our analyses) are two examples of species that were previously included in sect. Jacobaea, but now prove to be nested within sect. Senecio s.l. based on molecular data. Senecio squalidus was included in sect. Jacobaea by Dumortier (1827) and Chater and Walters (1976). Chater and Walters (1976) also placed several taxa in sect. Jacobaea that are supposed to be closely related to S. squalidus (S. aethnensis, S. cambrensis, S. fruticulosus, and S. siculus; Ashton and Abbott, 1992; Harris and Ingram, 1992a, b; Abbott et al., 2000). Hybrids between S. jacobaea and S. squalidus have been recorded (Harper and Wood, 1957), but Benoit, Crisp, and Jones (1975) concluded that these observations are erroneous. The results of the present
June 2002]
PELSER
ET AL.—MOLECULAR PHYLOGENY OF
study and those of, e.g., Kadereit and Jeffrey (1996) and Comes and Abbott (1999) indicate that S. squalidus and its close allies are better placed in sect. Senecio s.l. Senecio nebrodensis from the mountains of Spain is another species that was erroneously ascribed to sect. Jacobaea by De Candolle (1838), Boissier (1875), Hoffmann (1890–1894), Muschler (1909), and Chater and Walters (1976). This species is, however, closely related to S. viscosus (Emig and Kadereit, 1993; Kadereit et al., 1995; Purps and Kadereit, 1998; Comes and Abbott, 1999) and firmly nested within sect. Senecio s.l. (Kadereit and Jeffrey, 1996; Comes and Abbott, 1999). Also, many species of the Eurasian sect. Doria s.l., which are generally characterized by their undivided leaves and the low number of ligules, have been associated with species of sect. Jacobaea (Table 1). It is not clear to us why species of sect. Doria s.l. have been regarded to be close allies of S. jacobaea, since both groups are very different in their overall morphology. Results of our analysis and those of Knox and Palmer (1995) and Kadereit and Jeffrey (1996) confirm that the sections are not closely related. Phylogenetic surveys of the Senecioneae (Bremer, 1994; Knox and Palmer, 1995; Kadereit and Jeffrey, 1996; Panero et al., 1999; Bain and Golden, 2000, and this study) clearly show the problems of tackling speciose genera. The enormous size of the Senecioneae makes it very difficult to reach representative sampling sizes for studies into the evolutionary history of this group. Additionally, and partly because of this, the generic classification in the Senecioneae is still largely unresolved. It is obvious that Senecio s.s., as currently circumscribed, is a polyphyletic assemblage (Bremer, 1994; Knox and Palmer, 1995; Kadereit and Jeffrey, 1996; Vincent, 1996). Because only monophyletic groups of species have evolutionary historical reality (Sanders and Judd, 2000), a monophyletic delimitation of Senecio is of utmost importance. There are two ways to arrive at a monophyletic Senecio. One way is to transfer monophyletic groups of species from Senecio s.l. to other genera, to reinstate genera, or to raise new genera. This has been the prevailing strategy in the last three decades (e.g., Robinson and Brettell, 1973a, b, c, d, 1974; Robinson, 1974; Nordenstam, 1978; Jeffrey and Chen, 1984; Jeffrey, 1986, 1992). Another option is to abolish genera nested within Senecio s.l. and to transfer their species to Senecio. In either way, many nomenclatural changes are necessary, but these should only be performed when the phylogeny of the Senecioneae is sufficiently known. The present study is again another indication that a classification of the Senecioneae is only possible after a well-supported framework of the phylogeny of this group is reconstructed (Knox and Palmer, 1995). The results of the present study are a significant contribution to the knowledge about the evolutionary history of sect. Jacobaea and Senecio in general. (1) First, it has become clear what species are members of sect. Jacobaea, apart from the three core species that have been consistently ascribed to this section. (2) In addition to this, phylogenetic relationships within sect. Jacobaea have been clarified, although evolutionary relationships between S. jacobaea and seven of its closest relatives are not well resolved. (3) Finally, we have gained more knowledge about the phylogenetic position of sect. Jacobaea in the tribe Senecioneae. In future studies we will examine the currently unresolved evolutionary relationships among S. jacobaea and its closest relatives and the taxonomic status of several taxa in this group in more detail, using faster evolving molecular markers (e.g.,
SENECIO
SECT.
JACOBAEA
937
amplified fragment length polymorphisms [AFLPs]). In addition to this we will use the results of our phylogenetic analyses to gain more knowledge about the evolution of the diversity of PA types found by M. Macel (Institute of Evolutionary and Ecological Sciences, The Netherlands, unpublished data) in sect. Jacobaea. Furthermore, we will search for morphological synapomorphies for sect. Jacobaea as circumscribed here, to find additional support for this section. These characters might also contribute to a better understanding of the phylogenetic position of sect. Jacobaea and the phylogenetic value of morphological characters that are frequently used in classifications of Senecio s.l. LITERATURE CITED ABBOTT, R. J., J. K. JAMES, J. A. IRWIN, AND H. P. COMES. 2000. Hybrid origin of the Oxford Ragwort, Senecio squalidus L. Watsonia 23: 123– 138. ALEXANDER, J. C. M. 1979. The Mediterranean species of Senecio sections Senecio and Delphinifolius. Notes from the Royal Botanic Garden Edinburgh 37: 387–428. ALVAREZ FERNANDEZ, I., J. FUERTES AGUILAR, J. L. PANERO, AND G. NIETO FELINER. 2001. A phylogenetic analysis of Doronicum (Asteraceae, Senecioneae) based on morphological, nuclear ribosomal (ITS) and plastid (trnL-F) evidence. Molecular Phylogenetics and Evolution 20: 41–64. ASHTON, P. A., AND R. J. ABBOTT. 1992. Multiple origins and genetic diversity in the newly arisen allopolyploid species, Senecio cambrensis Rosser (Compositae). Heredity 68: 25–32. BAIN, J. F. 1991. The biology of Canadian weeds. Senecio jacobaea L. Canadian Journal of Plant Science 71: 127–140. BAIN, J. F., AND J. L. GOLDEN. 2000. A phylogeny of Packera (Senecioneae; Asteraceae) based on internal transcribed spacer region sequence data and a broad sampling of outgroups. Molecular Phylogenetics and Evolution 16(3): 331–338. BAIN, J. F., AND R. K. JANSEN. 1995. A phylogenetic analysis of the aureoid Senecio (Asteraceae) complex based on ITS sequence data. Plant Systematics and Evolution 195: 209–219. BAIN, J. F., B. S. TYSON, AND D. F. BRAY. 1997. Variation in pollen wall ultrastructure in New World Senecioneae (Asteraceae), with special reference to Packera. Canadian Journal of Botany 75: 730–735. BAIN, J. F., AND J. WALKER. 1995. A comparison of the pollen wall ultrastructure of aureoid and non-aureoid Senecio species (Asteraceae) in North America. Plant Systematics and Evolution 195: 199–207. BALDWIN, B. G. 1992. Phylogenetic utility of the internal transcribed spacers of nuclear ribosomal DNA in plants: an example from the Compositae. Molecular Phylogenetics and Evolution 1: 3–16. BARKLEY, T. M. 1978. Senecio. North American Flora series II, 10, 50–139. New York Botanical Garden, Bronx, New York, USA. BARKLEY, T. M. 1988. Variation among the Aureoid Senecios of North America: a geohistorical interpretation. Botanical Review 54: 82–106. BARKLEY, T. M. 1999. The segregates of Senecio, s.l., and Cacalia, s.l., in the flora of North America North of Mexico. Sida 18: 661–672. BENOIT, P. M., P. C. CRISP, AND B. M. G. JONES. 1975. Senecio L. In C. A. Stace [ed.], Hybridization and the flora of the British Isles, 404–410. Academic Press, London, UK. BOHS, L., AND R. G. OLMSTEAD. 1997. Phylogenetic relationships in Solanum (Solanaceae) based on ndhF sequences. Systematic Botany 22: 5–17. BOISSIER, P. E. 1875. Flora orientalis 3. H. Georg, Basel and Geneva, Switzerland. BREMER, K. 1988. The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42: 795–803. BREMER, K. 1994. Asteraceae: cladistics and classification. Timber Press, Portland, Oregon, USA. CHATER, A. O., AND S. M. WALTERS. 1976. Senecio L. In T. G. Tutin, V. H. Heywood, N. A. Burges, D. M. Moore, D. H. Valentine, S. M. Walters, and D. A. Webb [eds.], Flora Europaea 4, 191–205. Cambridge University Press, Cambridge, UK. COMES, H. P., AND R. J. ABBOTT. 1999. Reticulate evolution in the Mediterranean species complex of Senecio sect. Senecio: uniting phylogenetic and population-level approaches. In P. M. Hollingsworth, R. M. Bateman,
938
AMERICAN JOURNAL
and R. J. Gornall [eds.], Molecular systematics and plant evolution, 171– 198. Taylor and Francis, London, UK. CRINS, W. J. 1990. Phylogenetic considerations below the sectional level in Carex. Canadian Journal Botany 68: 1433–1440. DE CANDOLLE, A. P. 1838. Prodromus 6. Treuttel and Wu¨rtz, Paris, France. DENDA, T., K. WATANABE, K. KOSUGE, T. YAHARA, AND M. ITO. 1999. Molecular phylogeny of Brachycome (Asteraceae). Plant Systematics and Evolution 217: 299–311. DONOGHUE, M. J., R. G. OLMSTEAD, J. F. SMITH, AND J. D. PALMER. 1992. Phylogenetic relationships of Dipsacales based on rbcL sequences. Annals of the Missouri Botanical Garden 79: 333–345. DOYLE, J. J., AND J. L. DOYLE. 1987. A rapid isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11–15. DUMORTIER, B. C. J. 1827. Florula Belgica. Casterman, Tournay, Belgium. EMIG, W., AND J. W. KADEREIT. 1993. The comparative biology of the closely related Senecio nebrodensis and S. viscosus, a narrow endemic and a widespread ruderal. Nordic Journal of Botany 13: 369–375. ERIKSSON, T. 1998. AutoDecay ver. 4.0.2 (program distributed by the author). Department of Botany, Stockholm University, Stockholm, Sweden. FARRIS, J. S., M. KA¨LLERSJO¨, A. KLUGE, AND C. BULT. 1995. Testing significance of incongruence. Cladistics 10: 315–319. FRANCISCO-ORTEGA, J., S.-J. PARK, A. SANTOS-GUERRA, A. BENABID, AND R. K. JANSEN. 2001. Origin and evolution of the endemic Macaronesian Inuleae (Asteraceae): evidence from the internal transcribed spacers of the nuclear ribosomal DNA. Biological Journal of the Linnean Soceity 72: 77–97. FELSENSTEIN, J. 1985. Confidence limits on phylogenies: an approach using bootstrap. Evolution 39: 783–791. HARPER, J. L., AND W. A. WOOD. 1957. Biological flora of the British Isles. Senecio jacobaea. Journal of Ecology 45: 617–637. HARRIS, S. A., AND R. INGRAM. 1992a. Molecular systematics of the genus Senecio L. I. Hybridization in a British polyploid complex. Heredity 69: 1–10. HARRIS, S. A., AND R. INGRAM. 1992b. Molecular systematics of the genus Senecio L. II. The origin of S. vulgaris. Heredity 69: 112–121. HILLIS, D. M., AND J. J. BULL. 1993. An empirical test of bootstrapping as a method for assessing confidence in phylogenetic analysis. Systematic Biology 42: 182–192. HOFFMANN, O. 1890–1894. Compositae. In H. G. A. Engler and K. A. E. Prantl [eds.], Natu¨rliche Pflanzenfamilien I, 4 (5), 87–387. Engelmann, Leipzig, Germany. HOLMGREN, P. K., N. H. HOLMGREN, AND L. C. BARNETT. 1990. Index herbariorum part I: the herbaria of the world, edition 8. New York Botanical Garden, Bronx, New York, USA. HOOKER, J. D. 1881. The Flora of British India 3. Reeve, London, UK. JEFFREY, C. 1978. Compositae. In V. H. Heywood [ed.], Flowering plants of the world, 263–268. Mayflower Books, New York, New York, USA. JEFFREY, C. 1979. Generic and sectional limits in Senecio (Compositae), II: evaluation of some recent studies. Kew Bulletin 34: 49–58. JEFFREY, C. 1986. Notes on Compositae, IV: the Senecioneae in east tropical Africa. Kew Bulletin 41: 873–943. JEFFREY, C. 1992. Notes on Compositae, VI: the tribe Senecioneae (Compositae) in the Mascarene Islands with an annotated world check-list of the genera of the tribe, notes on Compositae VI. Kew Bulletin 47: 49– 109. JEFFREY, C., AND Y.-L. CHEN. 1984. Taxonomic studies on the tribe Senecioneae (Compositae) of Eastern Asia. Kew Bulletin 39: 205–446. JEFFREY, C., P. HALLIDAY, M. WILMOT-DEAR, AND S. W. JONES. 1977. Generic and sectional limits in Senecio (Compositae), I: Progress report. Kew Bulletin 32: 47–67. KADEREIT, J. W., H. P. COMES, D. J. CURNOW, J. A. IRWIN, AND R. J. ABBOTT. 1995. Plastid DNA and isozyme analysis of the progenitor-derivative species relationship between Senecio nebrodensis and S. viscosus (Asteraceae). American Journal of Botany 82: 1179–1185. KADEREIT, J. W., AND C. JEFFREY. 1996. A preliminary analysis of cpDNA variation in the tribe Senecioneae (Compositae). In D. J. N. Hind and H. J. Beentje [eds.], Compositae: Systematics. Proceedings of the International Compositae Conference, Kew, 1994, vol. 1, 349–360. Royal Botanic Gardens, Kew, UK. KIM, K.-J., AND R. K. JANSEN. 1994. Comparisons of phylogenetic hypotheses among different data sets in dwarf dandelions (Krigia, Asteraceae): additional information from internal transcribed spacer sequences of nuclear ribosomal DNA. Plant Systematics and Evolution 190: 157–185.
OF
BOTANY
[Vol. 89
KITAMURA, S. 1942. Compositae Japonicae 3. Memoirs of the College of Science Kyoto Imperial University, series B, Biology 16: 155–292. KNOX, E. B., AND J. D. PALMER. 1995. The origin of Dendrosenecio within the Senecioneae (Asteraceae) based on plastid DNA evidence. American Journal of Botany 82: 1567–1573. KONECHNAYA, G. Y. 1981. The carpological and anatomical characters of the species of the genus Senecio L. (Asteraceae) with reference to their taxonomy. Botanicheskii Zhurnal S.S.S.R. 66: 834–842. KOYAMA, H. 1969. Taxonomic studies on the tribe Senecioneae of East Asia. II. Enumeration of the species of eastern Asia. Memoirs of the Faculty of Science, Kyoto University, series Biology 2(2): 137–183. LINNE VON BERG, G., A. SAMOYLOV, M. KLAAS, AND P. HANELT. 1996. Plastid DNA restriction analysis and the infrageneric grouping of Allium (Alliaceae). Plant Systematics and Evolution 200: 253–261. MABBERLEY, D. J. 1997. The plant-book: a portable dictionary of the vascular plants. Cambridge University Press, New York, New York, USA. MILLER, J. T., AND R. J. BAYER. 2001. Molecular phylogenetics of Acacia (Fabaceae: Mimosoideae) based on the plastid matK coding sequence and flanking trnK intron spacer regions. American Journal of Botany 88: 697–705. MUSCHLER, R. 1909. Systematische und pflantzengeographische Gliederung der afrikanischen Senecio-Arten. Botanische Jahrbu¨cher 43: 1–74. NORDENSTAM, B. 1978. Taxonomic studies in the tribe Senecioneae (Compositae). Opera Botanica 44: 1–84. OLIVER, E. G. H., AND I. M. OLIVER. 1991. Studies in the Ericoideae (Ericaceae). VII. New species in Erica, section Pseuderemia, from southern Africa. Bothalia 21: 137–142. PANERO, J. L., J. FRANCISCO-ORTEGA, R. K. JANSEN, AND A. SANTOS-GUERRA. 1999. Molecular evidence for multiple origins of woodiness and a New World biogeographic connection of the Macaronesian island endemic Pericallis (Asteraceae; Senecioneae). Proceedings of the National Academy of Science, USA 96: 13 886–13 891. PURPS, D. M. L., AND J. W. KADEREIT. 1998. RAPD evidence for a sister group relationship of the presumed progenitor-derivative species pair Senecio nebrodensis and S. viscosus (Asteraceae). Plant Systematics and Evolution 211: 57–70. REICHENBACH, H. G. L. 1831–1832. Flora germanica excursiora. Cnobloch, Leipzig, Germany. REICHENBACH, H. G. 1853. Senecio. In H. G. L. Reichenbach and H. G. Reichenbach [eds.], Icones florae germanicae et helveticae, 16: 35–46. Hofmeister, Leipzig, Germany. REZNICEK, A. A. 1990. Evolution in sedges (Carex, Cyperaceae). Canadian Journal of Botany 68: 1409–1432. ROBINSON, H. 1974. Studies in the Senecioneae (Asteraceae), VI: the genus Arnoglossum. Phytologia 28: 294–295. ROBINSON, H., AND R. D. BRETTELL. 1973a. Studies in the Senecioneae (Asteraceae). I. A new genus, Pittocaulon. Phytologia 26: 451–453. ROBINSON, H., AND R. D. BRETTELL 1973b. Studies in the Senecioneae (Asteraceae). II. A new genus, Nelsonianthus. Phytologia 27: 53–54. ROBINSON, H., AND R. D. BRETTELL. 1973c. Studies in the Senecioneae (Asteraceae). III. The genus Psacalium. Phytologia 27: 254–264. ROBINSON, H., AND R. D. BRETTELL. 1973d. Studies in the Senecioneae (Asteraceae). IV. The genera Mesadenia, Syneilesis, Miricacalia, Koyamacalia and Sinacalia. Phytologia 27: 265–276. ROUY, C. C. 1903. Flore de France 8. Paris, France. SANDERS, R. W., AND W. S. JUDD. 2000. Incorporating phylogenetic results into floristic treatments. Sida 18: 97–112. SHINOZAKI, K., ET AL. 1986. The complete nucleotide sequence of tobacco plastid genome: its gene organization and expression. EMBO Journal 5: 2043–2049. SHISHKIN, B. K. 1995. Senecio. In B. K. Shishkin and E. G. Bobrov [eds.], Flora of the U.S.S.R. 26, 801–908. Bishen Singh Mahendra Pal Singh & Koeltz Scientific Books, Dehra Dun, India. SKVARLA, J. J., B. L. TURNER, V. C. PATEL, AND A. S. TOMB. 1977. Pollen morphology in the Compositae and in morphologically related families. In V. H. Heywood, J. B. Harborne, and B. L. Turner [eds.], The biology and chemistry of the Compositae 1, 141–265. Academic Press, London, UK. SOLTIS, D. E., P. S. SOLTIS, M. E. MORT, M. W. CHASE, V. SAVOLAINEN, S. B. HOOT, AND C. M. MORTON. 1998. Inferring complex phylogenies using parsimony: an empirical approach using three large DNA data sets for angiosperms. Systematic Biology 47: 32–42. STARR, J. R., AND B. A. FORD. 2001. The taxonomic and phylogenetic utility
June 2002]
PELSER
ET AL.—MOLECULAR PHYLOGENY OF
of vegetative anatomy and fruit epidermal silica bodies in Carex section Phyllostachys (Cyperaceae). Canadian Journal of Botany 79: 362–379. SWOFFORD, D. L. 2001. PAUP*: phylogenetic analysis using parsimony (* and other methods). Version 4. Sinauer, Sunderland, Massachusetts, USA. TABERLET, P., L. GIELLY, G. PAUTOU, AND J. BOUVET. 1991. Universal primers for amplification of three non-coding regions of plastid DNA. Plant Molecular Biology 17: 1105–1109. VINCENT, P. L. D. 1996. Progress on clarifying the generic concept of Senecio based on an extensive world-wide sample of taxa. In D. J. N. Hind and H. J. Beentje [eds.], Compositae: Systematics. Proceedings of the Inter-
SENECIO
SECT.
JACOBAEA
939
national Compositae Conference, Kew, 1994, vol. 1, 597–611. Royal Botanic Gardens, Kew, UK. WHITE, T. J., T. BRUNS, S. LEE, AND J. TAYLOR. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In M. Innis, D. Gelfand, J. Sninsky, and T. J. White [eds.], PCR protocols: a guide to methods and applications, 315–332. Academic Press, San Diego, California, USA. YODER A. D., J. A. IRWIN, AND B. A. PAYSEUR. 2001. Failure of the ILD to determine data combinability for slow loris phylogeny. Systematic Biology 50: 408–424.
