当前位置：文档库 › 2005Further disintegration of the Scrophulariaceae

2005Further disintegration of the Scrophulariaceae

INTRODUCTION

Lamiales sensu APG II (The Angiosperm Phylogeny Group, 2003) constitute a major clade of flowering plants, with approximately 17,800 species (Judd & al., 2002) and 23 families, representing 12.3% of eudicot diversity (Stevens, 2001, onwards). They belong to a major asterid clade, the Lamiids (Bremer & al., 2002). The monophyly of Lamiales is comparatively uncontro-versial and well supported by molecular (e.g., Backlund & al., 1998; Oxelman & al., 1999; Albach & al., 2001; Bremer & al., 2002) and phytochemical data (Jensen, 1992; Scogin, 1992). Morphologically, they typically are characterized by opposite leaves, sympetalous zygomor-phic flowers, oligosaccharides, frequent production of 6-oxygenated flavones, embryos of Onagrad type and cap-sular fruits. Although the majority of the many well-known families in Lamiales are well supported (e.g., Olmstead & al., 2001), the relationships among them are obscure, despite several recent molecular studies (e.g., Olmstead & Reeves, 1995; Backlund & al., 1998; Oxelman & al., 1999; Olmstead & al., 2001).

The concept “Scrophulariae” occurs in Durande, Notions Elém. Bot.: 265. 1782, but the name of the fam-ily is conserved as Scrophulariaceae Jussieu. The most influential classifications for the 19th century concept of Scrophulariaceae were those of Bentham (1846), who in addition recognized Selaginaceae Choisy, Wettstein (1895), who recognized Lentibulariaceae, Plantagin-aceae, and Orobanchaceae as separate families, and Hallier (1903), who made the broadest circumscription of the family.

Even if doubts sometimes were expressed regarding the naturalness of Scrophulariaceae (e.g., Thieret, 1967; Barringer, 1984), it was not until the study by Olmstead & Reeves (1995) based on plastid ndhF and rbcL gene sequences that the concept of Scrophulariaceae needed a revolutionary revision in order to fit within a phyloge-netic framework. Olmstead & Reeves (1995) discovered two clearly separated clades consisting of scrophularia-ceous taxa. One clade (“scroph I”) included Buddleja L., Selago L., Verbascum L. and the type genus Scrophularia L., whereas the other (“scroph II”) includ-ed Antirrhinum L., Digitalis L., Veronica L., Plantago L., Hippuris L., and Callitriche L. In their study the enig-matic woody Paulownia Siebold & Zucc., which was usually classified in Scrophulariaceae but transferred to Bignoniaceae Juss. by Takthajan (1980), and Schlegelia Miq., which was originally placed in Bignoniaceae but argued by Armstrong (1985) to fit better within

411

Further disintegration of Scrophulariaceae

Bengt Oxelman1, Per Kornhall1, Richard G. Olmstead2& Birgitta Bremer3

1Department of Systematic Botany, Evolutionary Biology Centre, Uppsala University, Norbyv?gen 18D SE-752 36 Uppsala, Sweden. bengt.oxelman@ebc.uu.se. (author for correspondence), per.kornhall@ebc.uu.se 2Department of Biology, Box 355325, University of Washington, Seattle, Washington 98195, U.S.A. olm-stead@https://www.wendangku.net/doc/cb7761183.html,

3The Bergius Foundation at the Royal Swedish Academy of Sciences, Box 50017 SE-104 05 Stockholm, Sweden. birgitta.bremer@bergianska.se

A phylogenetic study of plastid DNA sequences (ndhF, trnL/F, and rps16) in Lamiales is presented. In partic-ular, the inclusiveness of Scrophulariaceae sensu APG II is elaborated. Scrophulariaceae in this sense are main-ly a southern hemisphere group, which includes Hemimerideae (including Alonsoa, with a few South American species), Myoporeae, the Central American Leucophylleae (including Capraria), Androya, Aptosimeae, Buddlejeae, Teedieae (including Oftia, Dermatobotrys, and Freylinia), Manuleeae, and chiefly Northern tem-perate Scrophularieae (including V erbascum and Oreosolen). Camptoloma and Phygelius group with Buddlejeae and Teedieae, but without being well resolved to any of these two groups. Antherothamnus is strongly supported as sister taxon to Scrophularieae. African Stilbaceae are shown to include Bowkerieae and Charadrophila. There is moderate support for a clade of putative Asian origin and including Phrymaceae, Paulownia, Rehmannia, Mazus, Lancea, and chiefly parasitic Orobanchaceae, to which Brandisia is shown to belong. A novel, strongly supported, clade of taxa earlier assigned to Scrophulariaceae was found. The clade includes Stemodiopsis, Torenia, Micranthemum and probably Picria and has unclear relationships to the rest of Lamiales. This clade possibly represents the tribe Lindernieae, diagnosed by geniculate anterior filaments, usually with a basal swelling.

KEYWORDS:Lamiales, ndhF, phylogeny, rps16, Scrophulariaceae, trnL/F.

Scrophulariaceae, were both left in uncertain positions within Lamiales.

Several subsequent molecular phylogenetic studies focusing either on other taxa of Lamiales or more inclu-sive groups (e.g., Soltis & al., 1998; Olmstead & al., 2000, 2001; Albach & al., 2001; Bremer & al., 2002) have confirmed the general pattern revealed by Olmstead & Reeves (1995).

The parasitic plants that have variously been placed in Scrophulariaceae or Orobanchaceae Vent. have been shown to constitute an additional monophyletic group (dePamphilis & al., 1997; Wolfe & dePamphilis, 1998; Young & al., 1999). Olmstead & al. (2001) merged the rbcL/ndhF dataset and the data from the plastid rps2 gene that had been useful for the recognition of Orobanchaceae. They also added some other taxa, and were thereby able to identify an additional distinct scro-phulariaceous clade, Calceolariaceae. In addition, Halleria L. was shown to group with Stilbaceae Kunth, and Mimulus L. did not belong to any of these five clades of ex-Scrophulariaceae taxa. Beardsley & Olmstead (2002) identified Mimulus as a member of another clade, Phrymaceae. Globularia L. has been demonstrated to belong to the Veronicaceae clade sensu Oxelman & al. (1999), Olmstead & al. (2001), Kornhall & al. (2001), and Bremer & al. (2002).

Olmstead & al. (2001) recognized five distinct phy-logenetic lineages composed mainly of taxa previously assigned to Scrophulariaceae. Nevertheless, several of their included genera (e.g., Schlegelia, Paulownia, Mimulus) did not group with any of these five lineages. This fact, together with the restricted sampling (39 of ca. 280 genera in Scrophulariaceae in a traditional sense, (see Watson & Dallwitz, 1992 onwards), calls for more extensive sampling of genera previously assigned to Scrophulariaceae.

Scrophulariaceae sensu APG II (2003) constitutes approximately what Olmstead & Reeves (1995) identi-fied as the “scroph I” clade, that is, Buddlejaceae K. Wilh. (see Oxelman & al., 1999 for phylogenetic cir-cumscription), Manuleeae Benth. (incl. Selagineae, see Kornhall & al., 2001), the two large genera Scrophularia and Verbascum, plus the tribe Hemimerideae Benth. and Myoporaceae R. Br., including Androya Perrier (Oxel-man & al., 1999; Bremer & al., 2002). The assignment of Hemimerideae to this clade has received only weak sup-port in previous studies. Several taxa that have been sug-gested to have affinities with Myoporaceae (e.g., Capraria L., Anticharis Endl. and Peliostomum E. Mey.) have not been studied by molecular methods. The inclu-siveness of Manuleeae is not entirely clear, and perhaps most importantly, a more extensive sampling of taxa that have been attributed to Scrophulariaceae is needed before a revised classification of Scrophulariaceae in accordance with phylogenetic relationships can be pre-sented. Fischer (2004) has presented a tentative classifi-cation of all genera belonging to Scrophulariaceae sensu lato (i.e., approximately in the sense of Hallier, 1903) into tribes and informal higher groups (“families”).

The aims of this study are to identify the inclusive-ness of the “scroph I” clade, sensu Olmstead & Reeves (1995), using plastid DNA sequence data and to infer relationships of representatives of major groups of taxa previously assigned to Scrophulariaceae, for which plas-tid DNA sequences have not been obtained before.

MATERIAL AND METHODS

Sampling. — Most previous molecular studies have used sequences of rbcL and ndhF, but many other plastid loci have been introduced as well, including rps2 (dePamphilis & al., 1997), trnL/F intron/spacer region (Freeman & Scogin, 1999; McDade & Moody, 1999), rps16(Wallander & Albert, 2000; Bremer & al., 2002), matK(Bremer & al., 2002), trnT/F spacer region, and the trnV spacer (Bremer & al., 2002). We use the ndhF, trnL/F, and rps16DNA sequence regions, partly because a substantial number of bulk Lamiales taxa have already been sequenced for these, and partly because previous studies have indicated that these regions are particularly informative in Lamiales. Initially, we included all Lamiales taxa with sequence information from either of the trnL/F, rps16, or ndhF regions deposited in EMBL/ Genbank as per May 2002. However, because of uneven representation, we used an exemplar approach for groups whose monophyly is not in question. In addition, we tried to obtain material for DNA extraction from repre-sentatives of genera that have been suggested to be relat-ed to Scrophulariaceae sensu APG II. We also included representatives from other parts of Scrophulariaceae sensu Hallier and from Stilbaceae. We tried to find addi-tional representatives of every taxon, and sequence at least one of the three regions to verify the taxonomic identity of the sequences. Information on sampled taxa, and EMBL/Genbank accession numbers can be found in the Appendix.

Sequencing. —Most of the sequencing was per-formed either at the Evolutionary Biology Centre, Uppsala University, Sweden, or at the Department of Botany, University of Washington, Seattle, U.S.A. Total DNA was usually extracted from dried plant material using some variant of a standard CTAB/Chloroform extraction protocol, often followed by DNA purification using the Qiaquick PCR purification kit (Qiagen). Polymerase chain reactions (PCR) of the targeting regions were performed using combinations of primers published in Taberlet & al. (1991), Oxelman & al. (1997,

412

1999), Kornhall & al. (2001), and Popp & Oxelman (2001). For trnL/F two new primers were constructed: trnL BOC: GGCGRAATYGGTAGACGCTACG, and trnL BOF-R: CCAGATTTGAACTGGTGACACGAG. PCR products were purified using the Qiaquick PCR purification kit (Qiagen). In some instances, when the products were present as faint bands only, we used nest-ed primers in a second round of PCR in order to obtain enough products for sequencing. Automated sequencing was performed either on an ABI377 (Applied Biosystems), or on a MegaBACE 1000 capillary machine (Amersham Pharmacia Biotech) using manu-facturer’s protocols.

Assembly and alignment. —Sequences were assembled using Sequencher v. 3.1.1 (GeneCodes Corporation). The resulting sequences were aligned using the guidelines in Oxelman & al. (1997), and the “simple gap coding” described by Simmons and Ochotorrena (2000).

Phylogenetics. — Separate phylogenetic analyses were performed for each region initially. Taxa for which only one region has been sequenced are not presented in any of the results, unless the relationships found have not previously been reported in the literature. We made three combined analyses: one where all taxa were included, regardless of completeness (A1), one where at least two of the three regions were available for each terminal taxon (A2), and one with taxa where all three regions were available (A3). For all analyses, we used maximum parsimony as the optimality criterion, and the program PAUP* ver. 4.0b10 (Swofford, 1999) to find the most parsimonious trees using the TBR search algorithm and 50 random sequence addition replicates, saving a maxi-mum of ten trees in each. For each analysis, the strict consensus tree was calculated from these trees. Bootstrapping was performed with 500 replicates, TBR search, and three random additions per replicate. No more than 10 trees were saved per random addition repli-cate.

A Bayesian phylogenetic analysis was performed on dataset A2 using MrBayes v. 3.0b4 (Huelsenbeck & Ronquist, 2001) on computers running Linux. The appropriateness of different models was evaluated using the program Modeltest v. 3.06 (Posada and Crandall, 1998). We ran the program for 1,000,000 generations, four parallel chains, and with every 100th tree saved.

In order to detect possible incongruence resulting from analyses of the three regions, a partitioned Bremer support analysis (PBS) was conducted using TreeRot v. 2 (Sorensen, 1999). The trees are rooted with sequences from Oleaceae Hoffsgg. & Link, which has been shown to be sister to the rest of Lamiales in previous chloroplast DNA phylogenies (Savolainen & al., 2000; Bremer & al., 2001, 2002; Olmstead & al., 2001).

RESULTS

The number of sequences, previously unreported sequences generated by us, aligned positions, parsimony informative positions and indels, consistency index (CI), and retention index (RI), for each region and the com-bined analyses are summarized in Table 1. The strict con-sensus tree from 129 most parsimonious trees found in the A2 parsimony analysis (at least two sequenced regions available) is shown with bootstrap percentages above branches in Figs. 1–6. Details of the summary tree in Fig. 1 are shown in Figs. 2–6, and referred to in the discussion of each group. The strict consensus tree from the A3 analysis is shown in Fig. 7 with bootstrap and PBS values. Table 2 shows the bootstrap support for comparable nodes in the three analyses. In most cases there is higher bootstrap support in the A2 and A3 analy-ses than in the A1 analysis, and the differences are often considerable. The A2 and A3 analyses had similar sup-port levels for comparable groups.

The Bayesian analysis was performed under a gener-al time reversible model with a proportion of invariant sites and a gamma distribution, as selected by Modeltest. The first 100,000 generations were discarded as burn-in. The frequencies of different nodes are indicated below branches on the A2 trees in Fig. 1–6. Generally, the par-simony and the Bayesian analyses are highly congruent, usually with considerably higher frequencies in the Bayesian analysis. Cases where the parsimony and Bayesian analyses differ include the position of

413

Calceolaria , which in the parsimony analysis is sister group to the core of Lamiales (excluding Tetrachondraceae and Oleaceae, Fig. 1). The bootstrap support for this is less than 50%. In the Bayesian analy-sis, Calceolaria ends up as sister group to Gesneriaceae with a posterior probability of 1. Colpias and Alonsoa form a monophyletic group with .56 posterior probabili-ty, whereas they form a poorly (51% bootstrap frequen-cy) supported grade in the parsimony analysis (Fig. 3).The positions of Rehmannia and Paulownia are both poorly resolved within the ABLLMOPPV clade by the parsimony bootstrap analysis, but form consecutive sis-ter groups to Orobanchaceae with 100 % posterior prob-abilities in the Bayesian analysis.

DISCUSSION

The general structure of the tree in Figure 1 is in accordance with previously published chloroplast DNA phylogenies of Lamiales (e.g., Oxelman & al., 1999;Olmstead & al., 2001; Bremer & al., 2002). With Olea-ceae designated as the outgroup, Tetrachondraceae appears as sister to the rest. In agreement with Olmstead & al. (2001) and Bremer & al. (2002), there is emerging support that the bulk of “core” Lamiales does not include Calceolariaceae, Gesneriaceae Rich. & Juss., Sanango Bunting & Duke, and Peltanthera Benth. The relation-ships among these are, however, contradictory. The PBS analysis (Fig. 7) gave conflicting numbers for two nodes (Calceolaria L. as sister group to the rest of Lamiales including Gesneriaceae, and monophyly of Sanango/

Streptocarpus Lindley), and in both cases the rps16data are in conflict with trnL/F and ndhF . We have rechecked the identities of these sequences as well as the alignment,and we can find no obvious explanation to this pattern.There is support both from ndhF (bootstrap) and trnL/F (bootstrap) for a sister-group relationship between Sanango and Gesneriaceae. The support for Peltanthera as sister taxon to Sanango and Gesneriaceae is weak

414

Fig. 1. Simplification of the strict consensus tree from the 129 most parsimonious trees found in the A2 analysis with bootstrap values, when >50%, given above nodes. Bayesian posteriors under a GTR + gamma model are given below branches. Calceolariaceae is represented by a single sequence in this analysis. References to detailed subtrees are given inside parentheses. ABLLMOPPV = Acanthaceae, Bignoniaceae, Lamiaceae, Lentibulariaceae, Martyniaceae,Orobanchaceae, Pedaliaceae, Phrymaceae, and Verbenaceae.

Scrophulariaceae (Fig. 3)Stilbaceae (Fig. 4)Lindernieae (Fig. 5)ABLLMOPPV (Fig. 6)Plantaginaceae (Fig. 2)Gesneriaceae s.l.Calceolariaceae

Tetrachondraceae Oleaceae

9998527978100

8988781001001.

1..62

1.1.

1.1.1.1.-1. 1.

from our data, only trnL/F contributes with positive sup-port in the three-gene dataset (Fig. 7). However, both of these relationships are supported by rbcL (Oxelman &al., 1999). Bremer & al. (2002) report strong support for these relationships, and Jensen (2000) report biochemi-cal evidence that may indicate a relationship between Peltanthera and Gesneriaceae.

In the following discussion, we will present and dis-cuss more detailed results from particular groups of the “core” Lamiales. We use the family classifications of Olmstead & al. (2001) and APG II and the tribal rank to classify the genera studied taxonomically. We also dis-cuss some still unsequenced scrophulariaceous taxa and their putative relationships.

Plantaginaceae (Veronicaceae sensu Olmstead & al.2001). This group (Fig. 2) was originally recognized by Olmstead & Reeves (1995) as “Scroph II” and was cor-roborated with more taxa in Olmstead & al. (2001). The group is supported in the cpDNA tree (parsimony boot-strap: A2 - 79%; A3 - 91%), and there is also support for a sister-group relationship to the rest of the core Lamiales. Within Plantaginaceae, there are several well-supported subgroups. Cheloneae Benth., comprised sole-ly of New World taxa, have been identified by several previous molecular studies, including restriction sites (Wolfe & al., 1997, 2002), the trnL intron (Freeman &Scogin, 1999), ndhF , rbcL , and rps2(Olmstead & al.,2001), and matK and nrDNA ITS sequences (Wolfe &al., 2002). In our A2 tree, as well as in Wolfe & al.,(2002), and some of the trees obtained by Olmstead & al.(2001), Russelieae (Russelia Jacq. and Tetranema Benth.) forms the sister group to Cheloneae, although in none of the trees the relationship receives strong support.

Our study and Olmstead & al. (2001) also concur that Cheloneae and Russelieae form a monophyletic group with Antirrhineae Dumort., Digitalis , Globularia ,Poskea Vatke, Campylanthus Roth., Hippuris ,Callitriche , Plantago , Hemiphragma Wallich., and Veronica . This clade consists of several well-delimited groups, of which several often have been recognized at the family level (Globulariaceae DC., Hippuridaceae Vest., Callitrichaceae Bercht. & J. Presl. and Plantagina-ceae sensu stricto). Ghebrehiwet & al. (2000) examined Antirrhineae cpDNA relationships in detail, and the sis-ter-group relationship to the rest of the above-mentioned genera is well established.

Globularia and Poskea form a strongly supported group, which is corroborated by morphological data (Barringer, 1993). Campylanthus is resolved as sister to these, a result which needs corroboration from other data. The proximity of Campylanthus to Digitalideae Dumort. has been suggested (see Hjertsson, 2003) and has also been substantiated in a phytochemical study by R?nsted & Jensen (2002) and a brief molecular study by

415

Fig. 2. Phylogram of one of the most parsimonious trees from the A2 analyses with taxa from Plantaginaceae. Bootstrap values, when >50%, are indicated above branches. Bayesian posteriors under a GTR + gamma model are given below branches. * denotes sequences combined from different vouchers. Species epithets and occasionally numbers within parentheses are given to differentiate between DNA samples within genera, when there are more than one (see Appendix).

G l o b u l a r i e a e

Hjertsson (1997).

Digitalis appears paraphyletic in relation to Isoplexis Loudon according to ndhF data (results not shown).According to ITS data (Albach, 2001), Erinus L. is also part of Digitalideae.

Digitalideae and Globularieae form strongly sup-ported monophyletic groups irrespective of whether divergent Callitrichaceae and Plantaginaceae s.s. are included. Hemiphragma appears to belong to the Plantagineae/Veroniceae clade rather than to Digitalideae (Olmstead & al., 2001).

The least understood group in the former

416

Fig. 3. Phylogram of one of the most parsimonious trees from the A2 analysis with taxa from Scrophulariaceae sensu

APG II. Bootstrap values, when >50%, are given above nodes (or on the left when this was not possible). Bayesian posteriors under a GTR + gamma model are given below branches. * indicates sequences combined from different vou-chers. Species epithets and occasionally numbers within parentheses are given to differentiate between DNA samples within genera, when there are more than one (see Appendix).

L i m o s e l l e a e

Scrophulariaceae is probably Gratioleae. The circum-scription of Gratioleae by Bentham has been discussed by Wettstein (1895) and Thieret (1967) among others.Olmstead & al. (2001) reported monophyly of Amphianthus Torrey, Bacopa Aublet, and Gratiola L.,and a weakly supported relationship with Angelonia Bonpl. We present data that assign Scoparia L., Stemodia L., and Mecardonia Ruís & Pavón to Gratioleae. Many genera remain to be tested for their relationships with this group, with many members occurring in Neotropical areas. This is even more evident for Angelonieae. The close relationship between Angelonia and Basistemon Turcz. previously suggested by Barringer (1993) is cor-roborated, and these in turn form a strongly supported group together with the South American genera Monttea and Melosperma , consecutively followed by a less strongly supported relationship with the Neotropical Ourisia Comm. Plantaginaceae deserve more study, in particular with respect to Gratioleae/Angelonieae.Recently, Plantaginaceae have been studied in more detail by Albach & al. (2005).

Scrophulariaceae sensu APG II (Fig. 3). —The concept of Scrophulariaceae has changed consider-ably with the greater understanding gained via molecular phylogenetics. Olmstead & al. (2001) and APG II (2003)delimit the family to a clade consisting of Scrophularieae Dumort., Manuleeae Benth., Buddlejeae Bartl. (Buddle-jaceae sensu Oxelman & al. 1999), Myoporeae Rchb.,Leucophylleae Miers, Aptosimeae Benth. & Hook. f.,and Hemimerideae Benth. In this study, we confirm the monophyly of this assemblage (89% bootstrap support in our A2 analysis; 99% in our A3 analysis; Fig. 7). We add one to several taxa to each of the tribes relative to previ-

ous molecular studies. In addition, tribe Teedieae Benth.is identified as a member of this clade. In the following text, each of the tribes is discussed in more detail.

Scrophularieae. —The close relationship between Scrophularia and Verbascum has been suggest-ed by previous molecular studies, as well as similarities in seed and embryo characteristics (Thieret, 1967), and leaf anatomy (Lersten & Curtis, 1997). The close relation between Oreosolen Hook. f. and Scrophularia (Fig. 3) is here reported for the first time, but is not surprising from a morphological point of view. Both the Himalayan Oreosolen and the closely related Nathaliella B. Fedtsch.from Central Asia agrees well with Scrophularia in flo-ral morphology and general leaf architecture. The rela-tionship also makes sense biogeographically, since Scrophularia , Oreosolen , and Verbascum all have main-ly Northern Hemisphere distributions, as opposed to most other Scrophulariaceae. The strong support for a sister-group relationship between the South African Antherothamnus N. E. Br. and Scrophularia/Oreosolen/Verbascum is not evident morphologically.The position of Antherothamnus is robust to method of analysis, i.e.,neighbor-joining, maximum likelihood, and Bayesian inference with various evolutionary models (results not shown), and is supported by all three chloroplast genes (Fig. 7). Similar to Scrophularia , Antherothamnus has a fully developed staminode, a character that is not found in Manuleeae. Antherothamnus now appears to be a link between southern African ancestors and Eurasian Scrophularieae. [ N.B. The Scrophularia macrantha trnL sequence on EMBL/Genbank is probably a misidentifi-cation, because it groups with Collinsieae with strong support.]

Limoselleae. —Manuleeae encompass Selagineae Wettstein (Selaginaceae Choisy) since the latter are nest-ed within Manuleeae (Kornhall & al., 2001). However,the unexpected finding that Limosella L. also is nested within this group (Fig. 3, Kornhall & Bremer, 2004) sug-gests that the correct name for this tribe is Limoselleae Dumort., for priority reasons (Kornhall & Bremer, 2004).Limosella includes small plants that, as the vernacular English name, mudworts, indicates, grow in wet areas.Their choice of habitat and the small seeds might have facilitated a global distribution of the taxon by migrating birds (Darwin, 1872). The inclusion of Selagineae in Limoselleae on molecular evidence is also strongly cor-roborated by morphological characters. The “Selagi-neae” gynoecium, with one ovule per locule has appar-ently emanated several times within Limoselleae.Maintaining a separate Selagineae and Cronquist’s (1981) view that Selagineae belonged in the Globularia-ceae results from accepting the uni-ovulate ovary as a cardinal character. In spite of several authors pointing to the close connection between the two tribes based on

417

Fig. 4. Phylogram of one of the most parsimonious trees from the A2 analysis with representatives from Stilbaceae. Bootstrap values, when >50%, are indicated above the nodes. Bayesian posteriors under a GTR +gamma model are given below branches. Species epithets and occasionally numbers within parentheses are given to differentiate between DNA samples within genera, when there are more than one (see Appendix).

Anastrabe

Bowkeria verticillata Ixianthes retzioides (2)Campylostachys Stilbe

Retzia capensis (1)

Euthystachys Nuxia

Charadrophila

Halleria lucida (1)

50 changes

9294

10092100689799B o w k e r i e a e

S t i l b a c e a e s . s .

1..59 1. 1..75

.92

other characters (Junell, 1961; Argue, 1993; Hilliard,1994), the two tribes were not united before Kornhall &al. (2001). Long branches characterize the basal parts of Limoselleae, and the relations between Jamesbrittenia ,Limosella and Lyperia are not totally clear. Kornhall &Bremer (2004) elaborate in more detail on the interpreta-tion of morphological features of Limosella in relation to the chloroplast phylogeny found here.

Buddlejeae. —The results presented here corrob-orate the circumscription of Buddlejeae by Oxelman &al. (1999). Wallick & al. (2000) have, with more exten-sive sampling of Buddleja and trnL intron sequences,further strengthened the suggestion that Buddleja might be paraphyletic.

T eedieae. —There is strong support for a mono-phyletic group, Teedieae, consisting of the genera Oftia Adans., Teedia Rudolphi, Dermatobotrys Bolus, and Freylinia Colla. This group in turn forms a strongly sup-ported monophyletic group consisting also of the like-wise woody Buddlejeae, Phygelius Mey., and Camptoloma Benth. Wallick & al. (2000) reported a sis-ter-group relationship between Buddlejeae and Teedieae,but they included only Teedia and Oftia in their study, so their results are in agreement with ours. Phygelius and, in particular, Camptoloma are poorly resolved with respect to Teedieae and Buddlejeae. Barringer (1993) segregated the African woody taxa Teedieae, Freylinieae, and Bowkerieae from Cheloneae where Bentham originally placed them. Freylinieae sensu Barringer included also Phygelius , Antherothamnus , and Manuleopsis Thell., a relationship that is not supported by the present study.Barringer did not consider the epiphytic Dermatobotrys in this context. Morphologically, Buddlejeae, Teedieae,and Phygelius all have anthers with separate thecae,whereas the anther thecae of Camptoloma are confluent (as in Manuleeae and Scrophularieae). It is difficult to determine the polarity of this character. Possibly, the leafy inflorescences typical for Oftia , Teedia ,Dermatobotrys , and Freylinia could be interpreted as a synapomorphy for this group, as the bracts in Buddlejeae

and Camptoloma are small, and in Phygelius gradually turning leaf-like. We conclude that Teedia , Oftia ,Dermatobotrys , and Freylinia are best accommodated in Teedieae, whereas at the present it cannot be determined which relationship Phygelius and Camptoloma have to Teedieae and Buddlejeae. Teedieae, Buddlejeae,Camptoloma , Phygelius , Manuleeae, and Scrophularieae form a very strongly supported monophyletic group based on chloroplast DNA sequences.

Myoporeae. —With small variations, Myopora-ceae have generally been circumscribed as a mostly Southern Hemispheric taxon including Myoporum Sol.,Eremophila R. Br., and Bontia L. Oftia , here considered to belong to Teedieae (see above), has sometimes been considered to belong here (e.g., Wettstein, 1895), but Dahlgren & Rao (1971) rejected this based on morpho-logical evidence. The monophyly of Myoporum ,Eremophila , and Bontia is here strongly supported, a result consistent with the palynological data presented by Niezgoda & Tomb (1975) and anatomical data presented by Carlquist & Hoekman (1986).

Leucophylleae have been suggested to be related to Myoporeae (palynological data: Niezgoda & Tomb,1975; anatomical data: Carlquist & Hoekman, 1986;molecular data: Oxelman & al., 1999; Olmstead & al.,2001). The close relationship between Leucophyllum Bonpl. and Capraria L., both with Latin American dis-tribution, reported by Niezgoda & Tomb (1975) based on palynological evidence, and by Lersten & Curtis (2001)based on anatomical evidence is here corroborated. [N.B.The Leucophyllum minus trnL sequence deposited on EMBL/Genbank (AF034878) does not form a mono-phyletic group with our Leucophyllum frutescens sequence. Instead the trnL sequence labeled Pedicularis procera (AF034877) groups with L. frutescens . The identities of the L. minus and P . procera sequences need confirmation.]

Karrfalt & Tomb (1983) and Lersten & Beaman (1998) claimed homology between the oil cavities found in several Leucophyllum species and those found in the

418

Fig. 5. Phylogram of one of the most parsimonious trees from the A2 analysis with taxa from Lindernieae. Bootstrap values, when >50%, indicated above nodes. Bayesian posteriors under a GTR + gamma model are given below bran-ches. Species epithets and occasionally numbers within parentheses are given to differentiate between DNA samples within genera, when there are more than one (see Appendix).

Picris

Torenia polygonioides

Torenia baillonii (2)

50 changes

Micranthemum glomeratum Stemodiopsis buchananii

10086

100

981.

three Myoporeae genera, a hypothesis that is consistent with the results presented here. However, Lersten &Beaman (1998) found no cavities in L. revolutum Rzed or in Eremogeton grandiflorus Standl. & L. O. Williams,which is generally considered to be closely related to Leucophyllum . More detailed molecular phylogenetic studies on Leucophyllum , Eremogeton , and the little known Faxonanthus Greenm. from Guatemala, are war-ranted.

There is a strongly supported sister-group relation between Myoporeae and Leucophylleae and between that group and the Madagascar genus Androya (100% in both the A2 and the A3 analyses).

Aptosimeae. — The circumscription of Aptosimeae, characterized by alternate leaves and a dilated corolla tube, is one of the very few that seems to have been stable over the history of Scrophulariaceae classification. Bentham included Aptosimum , Pelio-stomum (sometimes included in Aptosimum ), and Anticharis , and has been followed by Wettstein among others. Monophyly of Aptosimeae is strongly supported by our chloroplast DNA sequences, and a sister group relation with Androya , Leucophylleae, and Myoporeae is moderately supported. All have 3-colpate, diporate pollen (Erdtman 1952; Niezgoda & Tomb, 1975; Punt,1980). Thus, these features may be synapomorphies for this group (“Myoporaceae”).

Hemimerideae. —The chiefly South African tribe Hemimerideae (including Alonsoa ) is supported by our data (70% bootstrap) and conforms well to the cir-cumscription reviewed in Steiner (1996), except that the chloroplast DNA sequence data support an inclusion of Colpias E. Mey. Steiner & Whitehead (1996) argued that Colpias rather should belong Bowkerieae, based on a basic chromosome number of x = 20, a tubular corolla,and a staminode corresponding to a fifth fertile stamen,characters that are shared with Bowkerieae. The amphi-Atlantic Alonsoa groups weakly with Diclis and Hemimeris . All have a basic chromosome number of x =7 (Steiner, 1996). The South American species of Alonsoa form a monophyletic group based on ndhF sequences (results not shown). Their shared ancestry is further supported by having 2n = 56 instead of 2n = 28 in the South African taxa. Thus, the interpretation of the absence of oil secretion in South American species of Alonsoa as a synapomorphy (Steiner, 1996) is reinforced.Whether the origin of Alonsoa predates the breakup of Gondwanaland needs to be further investigated. Nemesia

419

Fig. 6. Phylogram of one of the most parsimonious trees from the A2 analyses with representatives for the families Acanthaceae, Bignoniaceae, Lamiaceae, Lentibulariaceae, Martyniaceae, Orobanchaceae, Pedaliaceae, Phrymaceae and Verbenaceae. Bootstrap values, when >50%, are indicated above nodes. Bayesian posteriors under a GTR + gamma model are given below branches. * denotes sequences combined from different vouchers. Species epithets and occa-sionally numbers within parentheses are given to differentiate between DNA samples within genera, when there are more than one (see Appendix).

Barleria

Chlamydacanthus Crossandra *

Thunbergia alata

Pinguicula Elytraria *Catalpa Kigelia Tecoma stans (2)Martynia

Proboscidea Stachytarpheta dichotoma (2)Schlegelia parviflora 2

Berendtia Hemichaena Phryma Mimulus

Brandisia Cyclocheilon Lindenbergia philippinensis Lancea Mazus stachydifolius Rehmannia *

Paulownia

Callicarpa dichotoma Lamium

Sesamum 50 changes

100869152

6899100100

10010097100100

581005952Pedaliaceae Lamiaceae

O r o b a n c h a c e a e

Phrymaceae Verbenaceae

Martyniaceae

Bignoniaceae Lentibulariaceae

A c a n t h a c e a e

-1. 1.

-1..92.82

Vent. is here shown to be closely related with Diascia .Steiner (1996) showed that these taxa share the basic chromosome number x = 9, but emphasized differences

in the folding of lateral corolla lobes in bud and cited unpublished chloroplast rps2gene sequences as evidence for Nemesia belonging outside Hemimerideae. Our data

Fig. 7. Strict consensus tree from the A3 analysis. Parsimony bootstrap values, when >50%, are indicated above nodes.Below nodes the values from the PBS analysis are indicated in the following order: ndhF , trnL/F and rps16. *denotes sequences combined from different vouchers. Species epithets and occasionally numbers within parentheses are given to differentiate between DNA samples within genera, when there are more than one (see Appendix).

Alonsoa unilabiata Diclis reptans

Hemimeris sabulosa

Colpias Diascia

Nemesia strumosa Androya

Bontia Myoporum mauritianum Eremophila Capraria Leucophyllum frutescens *Aptosimum sp.

Peliostomum Antherothamnus Scrophularia californica Verbascum arcturus Manuleopsis Selago thunbergii Limosella *Buddleja davidii Emorya suaveolens (2)Nicodemia Camptoloma rotundifolium Oftia Teedia Dermatobotrys *Phygelius Anastrabe

Ixianthes retzioides (2)Campylostachys Stilbe Retzia capensis (1)Euthystachys

Nuxia Charadrophila Halleria lucida (1)

Barleria Chlamydacanthus Thunbergia alata

Berendtia

Hemichaena Phryma Mimulus

Brandisia Lindenbergia Lancea Mazus stachydifolius

Rehmannia *

Paulownia

Catalpa Lamium Proboscidea Schlegelia Stachytarpheta dichotoma (2)Lantana *Micranthemum Torenia (2)Stemodiopsis Sesamum Angelonia pubescens Basistemon Melosperma Monttea Ourisia Gratiola Stemodia Scoparia Antirrhinum Campylanthus Globularia cordifolia (1)Poskea Isoplexis Chelone obliqua (1)Russelia Tetranema Peltanthera Sanango Streptocarpus *Calceolaria Ligustrum

Polypremum

Tetrachondra patagonica 12,25,10

10053

2,-1,0

10017,2,0

700,1,2760,0,14,2,410010014,4,2

100100

12,7,1010,14,718,4,6

100651,2,1921005,0,0

9,5,17393

6,-1,-25,0,1

984,2,1952,0,1

100

10,8,3

1,0,1920,1,1923,-1,1

4,10,7100702,0,01,1,11,3,5

0,0,4

3,0,21,1,115,7,61,0,03,0,34,5,2

25,12,015,1,819,7,714,4,611,3,426,3,6

39,9,202,0,-12,1,0

5,0,-1

4,6,81,0,029,11,17

4,5,4

1,0,0

-2,2,13,0,2

14,4,11

-1,1,34,1,62,0,06,4,-1

21,20,9

5,2,-2

11,6,913,7,07,4,24,3,41,1,018,1,10

4,2,-14,0,38,0,-1

6,6,-5

0,1,06,0,-3

11,7,11

44,6,22

100821009989100941006698100100

1001001001006310010069

1001009959100

8010066

9310075100100

99995282

100548895S c r o p h u l a r i a c e e a e

S t i l b a c e a e

P l a n t a g i n a c e a e

Lindernieae Acanthaceae Pedaliaceae Phrymaceae Orobanchaceae Bignoniaceae Lamiaceae Martyniaceae 420

reject this, and instead favor the similarities in the androecium pointed out by Hilliard & Burtt (1984) as being homologous for Nemesia and Diascia.

Stilbaceae. —The circumscription of this family continues to expand (Fig. 4). Bremer & al. (1994) pre-sented molecular data that supported the close relation-ship between Retzia Thunb. and Stilbe Berg. proposed by Goldblatt & Keating (1976) on morphological grounds. Oxelman & al. (1999) showed that Nuxia Comm., previ-ously classified in Buddlejaceae, also belong here. These findings are supported by phytochemical data (Damtoft & al., 1984; Frederiksen & al., 1999). Olmstead & al. (2001) added Halleria, and in this study we show that the chloroplast DNA phylogeny supports putting the tribe Bowkerieae (Bowkeria, Anastrabe, and Ixianthes; Barringer, 1993) and Charadrophila Marloth here. Charadrophila capensis is a rare plant found in shade on permanent wet cliffs in the Cape Province. Its position has been disputed ever since Marloth’s original descrip-tion (Marloth, 1898). He placed it in Scrophulariaceae, but Engler, in a note to the original description placed it in Gesneriaceae, which it resembles superficially. Weber (1989) followed Marloth and placed it in Scrophulari-aceae arguing for a close relationship with Alonsoa. Its position in Stilbaceae was unexpected, and detailed stud-ies are needed in order to establish homology hypotheses in this expanded circumscription of Stilbaceae.

Lindernieae. —A novel strongly supported clade of taxa earlier assigned to Scrophulariaceae was found that includes Stemodiopsis Engl., Torenia L., Micranthe-mum Desf., and probably Picria Lour. (= Conobea) and has unclear relationships to the rest of Lamiales (Fig. 4). This clade probably represents the tribe Lindernieae, diagnosed by geniculate anterior filaments, usually with a basal swelling. If this character proves to be synapo-morphous, then for example also Lindernia All., Craterostigma Hochst., Crepidorhopolon E. Fischer, Hartliella E. Fischer, and Artanema D. Don. may belong here. Neither Stemodiopsis, Picria, or Micranthemum possess the characteristic basal swelling of the base of the filament, but like Stemodiopsis, they have geniculate or curved anterior filaments/staminodes. Lindernieae have mostly been classified in Gratioleae (Fischer, 1997), but have a pretty distinct morphology (e.g., alve-olate endosperm of a certain type; Fischer, 1992).

Orobanchaceae, Phrymaceae and relatives.—There is moderate support for a clade including Phrymaceae, Paulownia, Rehmannia Libosch., Mazus Lour., Lancea Hook. f. & Thomson, and chiefly parasitic Orobanchaceae, to which Brandisia Hook. f. & Thomson, previously not known to be parasitic, is shown to belong (Fig. 6). The inclusion of Lancea and Mazus in Phrymaceae, as advocated by Beardsley & Olmstead (2002), is not supported here.

ACKNOWLEDGEMENTS

Reija Dufva, Per Erixon, Nahid Heidari, Jan-Eric Mattsson, and Pat Reeves are gratefully acknowledged for invaluable work in the lab. Numerous persons helped us by providing plant materi-al, we are especially grateful to the help from Dirk Albach, Fanny Astholm, Mandy Balkwill, Mark W. Chase, S?ren R. Jensen, Pieter Maas, Veronica Mayer, Heidi Meudt, Mats Thulin, Andrea Wolfe, the staff at Uppsala Botanical Garden, and the curators of the GB, S, and UPS herbaria. K?re Bremer and Mats Thulin are gratefully acknowledged for comments on earlier versions of the manuscript. The study was financed by grants from the Swedish Research Council to BO and BB, respectively.

LITERATURE CITED

Albach, D. C.2001. Paraphyly of Veronica(Veronicaceae;

Scrophulariaceae): evidence from the internal transcribed spacer (ITS) sequences of nuclear ribosomal DNA.J. Pl.

Res.114: 9–18.

Albach, D. C., Meudt, H., & Oxelman, B. 2005. Piecing together the “new” Plantaginaceae. Amer. J. Bot.92: 297–315.

Albach, D. C., Soltis, P. S., Soltis, D. S. & Olmstead, R. G.

2001. Phylogenetic analysis of Asterids based on sequences of four genes. Ann. Missouri Bot. Gard.88: 163–212.

Argue, C. L.1993. Pollen morphology in the Selagineae, Manuleae (Scrophulariaceae), and selected Globulari-aceae, and its taxonomic significance. Amer. J. Bot. 80: 723–733.

Armstrong, J. E.1985. The delimitation of Bignoniaceae and Scrophulariaceae based on floral anatomy, and the place-ment of problem genera. Amer. J. Bot.72: 755–766. Backlund, M., Oxelman, B. & Bremer, B.1998.

Phylogenetic relationships within the Gentianales based on ndhF and rbcL sequences, with a particular reference to the Loganiaceae. Amer. J. Bot.87: 1029–1043. Barringer, K. A.1984. Seed morphology and the classification of the Scrophulariaceae. Amer. J. Bot.71 (5, pt. 2): 156.

[Abstr.]

Barringer, K. A.1993. Five new tribes in the Scrophulariaceae. Novon3: 15–17.

Beardsley, P. M. & Olmstead, R. G.2002. Redefining Phrymaceae: the placement of Mimulus, tribe Mimuleae, and Phryma. Amer. J. Bot. 89: 1093–1102.

Bentham, G.1846. Ordo CXLII Scrophulariaceae. In: de Candolle, A. de (ed.), Prodromus Systematis Naturalis Regni Vegetabilis, vol. 10. Victoris Masson, Paris. Bremer, K., Backlund, A., Sennblad, B., Swenson, U., Andreasen, K., Hjertsson, M., Lundberg, J., Backlund, M. & Bremer, B.2001. A phylogenetic analysis of 100+ genera and 50+ families of Euasterids based on morpho-logical and molecular data with notes on possible higher level morphological synapomorphies. Pl. Syst. Evol. 229: 137–169.

Bremer, B., Bremer, K., Heidari, N., Erixon, P., Olmstead, R. G., Anderberg, A. A., K?llersj?, M. &

421

Barkhordarian, E.2002. Phylogenetics of Asterids based on 3 coding and 3 non–coding chloroplast DNA markers and the utility of non–coding DNA at higher taxonomic levels. Molec. Phylogen. Evol. 24: 274–301.

Bremer, B., Olmstead, R. G., Struwe, L. & Sweere, J. A.

1994. rbcL sequences support exclusion of Retzia, Desfontainia, and Nicodemia from the Gentianales. Pl.

Syst. Evol.190: 213–230.

Carlquist, S. & Hoekman, D. A.1986. Wood anatomy of the Myoporaceae: ecological and systematic considerations.

Aliso11: 317–334.

Cronquist, A.1981. An Integrated System of Classification of Flowering Plants. Columbia Univ. Press, New York. Dahlgren, R. & Rao, V. S.1971. Genus Oftia Adans. and its systematic position. Bot. Not.124: 451–472. Damtoft, S., Franzyk, H., Jensen, S. R. & Nielsen, B. J.

1984. Iridioids and verbascoids in Retzia. Phytochemistry 34: 239–243.

Darwin, C.1872. The Origin of Species by Means of Natural Selection.John Murray, London.

dePamphilis, C. W., Young, N. D. & Wolfe, A. D.1997.

Evolution of plastid gene rps2in a lineage of hemipara-sitic and holoparasitic plants: many losses of photosynthe-sis and complex patterns of rate variation. Proc. Natl.

Acad. Sci. U.S.A. 94: 7367–7372.

Erdtman, G.(ed.). 1952. Angiosperms. Pollen Morphology and Plant Taxonomy. Almqvist & Wiksell, Stockholm. Fischer, E.1992. Systematik der afrikanischen Lindernieae (Scrophulariaceae). Trop. Subtrop. Pflanzenwelt81: 1–365.

Fischer, E.1997. Revision of the genus Stemodiopsis Engl.

(Scrophulariaceae-Gratioloideae). Bot. Jahrb. Syst.119: 305–326.

Fischer, E.2004. Scrophulariaceae. Pp 333–432 in: Kadereit, J. W. (ed.), The Families and Genera of Flowering Plants, vol. VII. Flowering Plants. Dicotyledons: Lamiales (except Acanthaceae including Avicenniaceae). Springer, Berlin.

Frederiksen, L. B., Damtoft, S. & Jensen, S. R.1999.

Biosynthesis of iridoids lacking C-10 and the chemotaxo-nomic implications of their distribution. Phytochemistry 52: 1409–1420.

Freeman, C. E. & Scogin, R.1999. Potential utility of chloro-plast trnL(UAA) gene intron sequences for inferring phy-logeny in Scrophulariaceae. Aliso18: 141–159. Ghebrehiwet, M., Bremer, B. & Thulin, M.2000. Phylogeny of the tribe Antirrhineae (Scrophulariaceae) based on mor-phological and ndhF sequence data. Pl. Syst. Evol.220: 223–239.

Goldblatt, P. & Keating, R. C.1976. Chromosome cytology, pollen structure, and relationship of Retzia capensis. Ann.

Missouri Bot. Gard.63: 321–325.

Hallier, H.1903. Ueber die Abgrenzung und Verwandtschaft der einzelnen Sippen. Bull. Herb. Boissier2: 181–207. Hilliard, O.1994. The Manuleae: a Tribe of Scrophulariaceae.

Edinburgh Univ. Press, Edinburgh.

Hilliard, O. & Burtt, B. L.1984. A revision of Diascia section Racemosae. J. S. African Bot.50: 269–340. Hjertsson, M.1997. Systematics of Lindenbergia and Campylanthus(Scrophulariaceae). Comprehens. Summ.

Uppsala Diss. Fac. Sci. Technol.331: 46.

Hjertsson, M.2003. Revision of the disjunct genus

Campylanthus(Scrophulariaceae). Edinburgh J. Bot.60: 131–174.

Huelsenbeck, J. P. & Ronquist, F.2001. MRBAYES, Bayesian inference of phylogeny. Bioinformatics17: 754–755.

Jensen, S. R.1992. Systematic implications of the distribution of iridoids and other chemical–compounds in the Loganiaceae and other families of the Asteridae. Ann.

Missouri Bot. Gard.79: 284–302.

Jensen, S. R.2000. Chemical relationships of Polypremum procumbens, Tetrachondra hamiltonii and Peltanthera floribunda. Biochem. Syst. Ecol.28: 45–51.

Judd, W. S., Campbell, C. S., Kellogg, E. A., Stevens, P. F. & Donoghue, M. J.2002. Plant Systematics: A Phylogenetic Approach, ed. 2. Sinauer Associates, Sunderland, Massachusetts.

Junell, S. 1961. Ovarian morphology and taxonomical position of Selagineae.Svensk Bot. Tidskr. 55: 168–192. Karrfalt, E. E. & Tomb, A. S. 1983. Air spaces, secretory cav-ities, and the relationship between Leucophylleae (Scrophulariaceae) and Myoporaceae. Syst. Bot.8: 29–32. Kornhall, P., Heidari, N. & Bremer, B.2001. Selagineae and Manuleeae, two tribes or one? Phylogenetic studies in the Scrophulariaceae. Pl. Syst. Evol.228: 199–218. Kornhall, P. & Bremer, B.2004. New circumscription of the tribe Limoselleae (Scrophulariaceae) that includes the taxa of the tribe Manuleeae. Bot. J. Linn. Soc.146: 453–467. Lersten, N. R. & Beaman, J. M.1998. First report of oil cav-ities in Scrophulariaceae and reinvestigation of air spaces in leaves of Leucophyllum frutescens. Amer. J. Bot. 85: 1646–1649.

Lersten, N. R. & Curtis, J. D.1997. Anatomy and distribution of foliar idioblasts in Scrophularia and Verbascum (Scrophulariaceae). Amer. J. Bot. 84: 1638–1645. Lersten, N. R. & Curtis, J. D.2001. Idioblasts and other unusual internal foliar secretory structures in Scrophulariaceae. Pl. Syst. Evol. 227: 63–73. Marloth, R.1898. Charadrophila Marloth nov. gen. Bot.

Jahrb. Syst.26: 358–359.

McDade, L. A. & Moody, M. L.1999. Phylogenetic relation-ships among Acanthaceae: evidence from noncoding trnL-trnF chloroplast DNA sequences. Amer. J. Bot.86: 70–80. Niezgoda, C. J. & Tomb, A. S.1975. Systematic palynology of tribe Leucophylleae (Scrophulariaceae) and selected Myoporaceae. Pollen & Spores17: 495–516. Olmstead, R. G., dePamphilis, C. W., Wolfe, A. D., Young, N. D., Elisens, W. J. & Reeves, P. A.2001. Disintegration of the Scrophulariaceae. Amer. J. Bot.88: 348–361. Olmstead, R. G., Jansen, R. K., Kim, K.-J. & Wagstaff, S. J.

2000. The phylogeny of the Asteridae s.l. based on chloro-plast ndhF sequences. Molec. Phylogen. Evol. 16: 96–112. Olmstead, R. G. & Reeves, P. A.1995. Evidence for the poly-phyly of the Scrophulariaceae based on chloroplast rbcL and ndhF sequences. Ann. Missouri Bot. Gard.82: 176–193.

Oxelman, B., Backlund, M. & Bremer, B.1999.

Relationships of the Buddlejaceae s.l. investigated using parsimony jackknife and branch support analysis of chloroplast ndhF and rbcL sequence data. Syst. Bot.24: 164–182.

Oxelman, B., Lidén, M. & Berglund, D.1997. Chloroplast rps16intron phylogeny of the tribe Silenae

422

(Caryophyllaceae). Pl. Syst. Evol. 206: 393–410. Popp, M. & Oxelman, B.2001. Inferring the history of the polyploid Silene aegaea(Caryophyllaceae) using plastid and homoeologous nuclear DNA sequences.Molec.

Phylogen. Evol.20: 474–481.

Posada, D. & Crandall, K. A.1998. MODELTEST, testing the model of DNA substitution. Bioinformatics14: 817–818.

Punt, W.1980. Pollen morphology. Pp. 162–191 in: Leeuwen-berg, A. J. M. (ed.), Engler and Prantl’s Die natürlichen Pflanzenfamilien, Angiospermae: Ordnung Gentianales, Fam. Loganiaceae, vol. 28b(1). Duncker and Humblot, Berlin.

R?nsted, N. & Jensen, S. R.2002. Iridoid glucosides and caf-feoyl phenylethanoid glycosides from Campylanthus sal-saloides and Campylanthus glaber. Biochem. Syst. Ecol.

30: 1091–1095.

Savolainen, V., Fay, M. F., Albach, D. C., Backlund, A., Bank, M. van der, Cameron, K. M., Johnson, S. A., Lledó, M. D., Pintaud, J.-C., Powell, M., Sheahan, M.

C., Soltis,

E., Weston, P., Whitten, W. M., Wurdack,

K. J. & Chase, M. W.2000. Phylogeny of the Eudicots: a nearly complete familial analysis based on rbcL gene sequences. Kew Bull.55: 257–309.

Scogin, R.1992. The distribution of acetoside among angiosperms. Biochem. Syst. Ecol.20: 477–480. Simmons, M. P. & Ochoterena, H.2000. Gaps as characters in sequence-based phylogenetic analyses. Syst. Biol.49: 369–381.

Soltis, D. E., Soltis, P. S., Mort, M. E., Chase, M. W., Savolainen, V., Hoot, S. B. & Morton, C. M.1998.

Inferring complex phylogenies using parsimony: an empirical approach using three large DNA datasets for angiosperms. Syst. Biol.47: 32–42.

Sorensen, M. D.1999. TreeRot.Boston Univ., Boston. Steiner, K. E.1996. Chromosome numbers and relationships in tribe Hemimerideae (Scrophulariaceae). Syst. Bot.21: 63–76.

Steiner, K. E. & Whitehead, V. B.1996. The consequences of specialization for pollination in a rare South African shrub, Ixianthes retzioides(Scrophulariaceae). Pl. Syst. Evol.

201: 131–138.

Stevens, P. F.2001. Angiosperm Phylogeny Website.

https://www.wendangku.net/doc/cb7761183.html,/MOBOT/Research/APweb/ welcome.html

Swofford, D. L.1999. P AUP, Phylogenetic Analysis Using Parsimony (*and Other Methods).Sinauer Associates, Sunderland, Massachusetts.

Taberlet, P., Gielly, L., Pautou, G. & Bouvet, J.1991.

Universal primers for amplification of three non–coding regions of chloroplast DNA. Pl. Molec. Biol.17: 1105–1109.

Takhtajan, A. L.1980. Outline of the classification of flower-ing plants (Magnoliophyta). Bot. Rev.46: 225–359.

The Angiosperm Phylogeny Group.2003. An update of the angiosperm phylogeny group classification for the orders and families of flowering plants: APG II. Bot. J. Linn. Soc.

141: 399–436.

Thieret, J. W.1967. Supraspecific classification in the Scrophulariaceae: a review. Sida3: 87–106. Wallander, E. & Albert, V. A. 2000. Phylogeny and classifi-cation of Oleaceae based on rps16and trnL-F sequence

data.Amer. J. Bot.87: 1827–1841.

Wallick, K. P., Elisens, W. J., Molvray, M. & Kores, P. J.

2000. A phylogenetic study of Buddlejaceae s.s. inferred from the trnL-F region of cpDNA. Amer. J. Bot.87 (6, suppl.): 183. [Abstr.]

Watson, L. & Dallwitz, M. J.1992. The Families of Flowering Plants: Descriptions, Illustrations, Identification, and Information Retrieval.https://www.wendangku.net/doc/cb7761183.html,/ delta/angio/

Weber, A.1989. Family position and conjectural affinities of Charadrophila capensis Marloth. Bot. Jahrb. Syst.111: 87–119.

Wettstein, R. von.1895. Scrophulariaceae. Pp. 39–107 in: Engler, A. & Prantl, K. (eds.), Die natürlichen Pflanzenfamilien, vol. 4(3b). Wilhelm Engelmann, Leipzig.

Wolfe, A. D., Elisens, W., Watson, L. & dePamphilis, C.W.

1997. Using restriction–site variation of PCR-amplified cpDNA genes for phylogenetic analysis of tribe Cheloneae (Scrophulariaceae). Amer. J. Bot.84: 555–564.

Wolfe, A. D., Datwyler, S. L. & Randle, C. 2002. A phyloge-netic and biogeographic analysis of the Cheloneae (Scrophulariaceae) based on ITS and matK sequence data.

Syst. Bot.27: 138–148.

Wolfe, A. D. & dePamphilis, C. W.1998. The effect of relaxed functional constraints on the photosynthetic gene rbcL in photosynthetic and nonphotosynthetic parasitic plants. Molec. Biol. Evol.15: 1243–1258.

Young, N. D., Steiner, K. E. & dePamphilis, C. W.1999. The evolution of parasitism in Scrophulariaceae/ Orobanchac-eae: plastid gene sequences refute an evolutionary transi-tion series. Ann. Missouri Bot. Gard.86: 876–893.

423

424

425