Molecular phylogeny of Nassauvia (Asteraceae, Mutisieae) based on nrDNA ITS sequences

The phylogeny of the genus Nassauvia and closely related genera was reconstructed using sequences from the internal transcribed spacer regions (ITS) of nuclear ribosomal DNA. The genus Triptilion is nested within Nassauvia, making the latter genus paraphyletic. Neither of the two subgenera Nassauvia and Strongyloma is resolved as monophyletic, and none of the sections of subgenus Nassauvia is recovered as monophyletic. The evolution of the compound secondary inflorescences has been complex in Nassauvia, with the highly aggregated forms representing the original condition in the genus. However, the ancestral condition is equivocal in several clades, and there are alternative reconstructions for the gains–losses of the variously aggregated conditions. There has been at least one gain of solitary capitula in Nassauvia. The evolution of flavonoid chemistry has been complex in Nassauvia, and flavonoids are of limited phylogenetic-taxonomic utility in the genus. Gains–losses of flavonols occur only on terminals whereas changes in flavones and C-glycosyl flavones occur at various levels in the tree. Gains–losses of methylation of flavones and flavonols occur only on terminals.


Introduction
Anyone studying the flora of southern South America can hardly overlook the morphologically striking genus Nassauvia Comm.ex Jussieu (Asteraceae, Mutisieae, Nassauviinae) with its unusual clusters of flowering heads.Nassauvia consists of 38 species (Cabrera 1982;Arroyo and Marticorena 1988) distributed in southern South America, especially along the Andes, in Patagonia, and on the Malvinas (Falkland Islands).It is the type genus of subtribe Nassauviinae, which contains about 24 genera (Cabrera 1971), and it has been previously suggested as being closely related to Calopappus, Moscharia, Polyachyrus, and Triptilion (Crisci 1974;Bremer 1993).Although studies on the morphology (Lessing 1832;de Candolle 1838;Bentham 1873;Cabrera 1982), phenetics (Crisci 1974), and morphological cladistics (Freire et al. 1993) of Nassauvia have already been published, comprehensive molecular phylogenetic insights are lacking.Katinas et al. (2008a, b) provide a broad overview of relationships within Nassauviinae using nuclear and plastid sequences, but given the objective of their study, only four species of Nassauvia were included.
Previous attempts have been made to estimate phylogenetic relationships within Nassauvia from morphology.On the basis of intuitive assessment of morphological characters and states, Cabrera (1982), in his revision of the genus, described evolutionary relationships among infrageneric taxa.He differentiated the primitive section Panargyrum from the more derived sections Mastigophorus, Nassauvia, and Caloptilium.He also suggested that subgenus Strongyloma was sister to subgenus Nassauvia.Hunziker et al. (1991), focusing on morphology and chromosome numbers (x = 11 in Nassauvia), suggested that the genus evolved from Calopappus, and that section Mastigophorus might be the most primitive.Triptilion was suggested as having evolved from section Panargyrum.
The most comprehensive morphological phylogenetic analysis of Nassauvia was conducted by Freire et al. (1993), who sampled all species of Nassauvia and Triptilion, Caloppapus, Moscharia, and Polyachyrus for 35 morphological characters.The more distant Cephalopappus of subtribe Nassauviinae was used as outgroup.The results supported in large measure the classification of Cabrera (1982), except that section Panargyrum was recovered as paraphyletic.Furthermore, Calopappus was nested within Nassauvia rendering the latter genus paraphyletic.Subgenus Strongyloma was resolved as the sister group of subgenus Nassauvia.
Because of the striking variation of the types of synflorescence in Nassauvia, ranging from single heads to large clusters, some previous discussions have been directed to understanding the evolution of these types.Stuessy et al. (1988;unpublished work) suggested that single capitula, for example in the closely related genus Calopappus, may have evolved the small heads typical of section Mastigophorus followed by smaller aggregations of individual heads as in sections Panargyrum and Strongyloma, culminating in the dense aggregations of sections Caloptilium and Nassauvia.
A putative environmental correlation was observed in the morphological trend from more open secondary inflorescences (synflorescences) to tighter and more conspicuous aggregations in going from the lower and warmer regions to the higher and colder Andes.Plotting synflorescence types on the morphologically-based cladogram of Freire et al. (1993) showed tendencies for more dense synflorescence aggregations to occur in more highly derived clades, but with several notable exceptions (e.g., N. juniperus, N. pentachaeta and N. aculeata; Fig. 6 in Freire et al. 1993).Tortosa et al. (2004) and Katinas et al. (2008a, b) argued that the compact synflorescence is the ancestral condition in Nassauvia and that more lax inflorescences are the derived condition.Katinas et al. (2008a, b) suggested that the general trend has been from complex secondary head type in more arid conditions to loss of complexity associated with the ecological transition to more mesic conditions.
Flavonoid compounds have been shown to be of taxonomic utility at different levels throughout the Asteraceae (Bohm and Stuessy 2001), but there are relatively few studies of tribe Mutisieae (Bohm and Stuessy 2001, Chap. 14).A survey of the distribution of flavonoid compounds in the leaves of species of Nassauvia was carried out to ascertain whether specific classes of compounds characterize species, sections, or subgenera, and to trace the distribution of compounds on the molecular phylogeny to interpret evolution of the compounds within the genus.
Therefore, the purposes of this paper are: 1 to estimate a molecular phylogeny of Nassauvia on the basis of sequences of nuclear ribosomal ITS, including a test of the monophyly of the genus; 2 to compare the new phylogenetic inferences with morphological hypotheses; 3 to analyze the implications of all data for infrageneric and intergeneric relationships; 4 to comment on the evolution of synflorescences within the genus; and 5 to interpret the pattern of evolution of leaf flavonoid chemistry.

Materials and methods
Taxon sampling To test the monophyly of Nassauvia, we constructed a data matrix of 54 total species, including 23 species (24 accessions) of Nassauvia, 29 species from 8 other genera of Mutiseae, and Schlechtendalia luzulifolia (Barnadesiodeae) as an outgroup (Table 1).We included in this sample several genera (i.e., Calopappus, Polyachyrus and Triptilion) considered closely related to Nassauvia on the basis of analyses of morphological and molecular characters (Crisci 1974;Bremer 1993;Katinas et al. 2008a, b).

Molecular methods
Amplification and DNA sequencing Total DNA was extracted from herbarium specimens in accordance with either the CTAB procedure (Doyle and Doyle 1990) or by use of the DNeasy Plant Mini Kit from Qiagen (Valencia, CA, USA).The nrDNA ITS regions were amplified by use of primers 4 and 5 from White et al. (1990).The doublestranded PCR products were produced via 35 cycles of 95°C for 1 min, 48°C for 1 min, and 72°C for 1 min, with a 10-min final extension cycle at 72°C.PCR products were purified using the QIAquick Gel Extraction Kit (Qiagen) and sequenced using the same primers in 10 ll reactions using 2 ll BigDye, 1 ll primer (20 pmolar), and template DNA and purified water to reach the reaction volume.Cycle sequencing used 25 cycles of 96°C for 10 s, 50°C for 5 s, and 60°C for 4 min.Sequencing products were visualized on an ABI Prism 377 automated DNA sequencer.
Phylogeny reconstruction Sequences were edited and assembled using AutoAssembler (PE-ABI, Foster City, CA, USA).Alignment of the large ITS data set for Mutiseae was easily accomplished by eye using Se-Al (Rambaut 1996).Data were exported as a nexus file and subsequently analyzed using parsimony (MP) in PAUP* (Swofford 2000) and maximum likelihood (ML) using Garli (available at http://www.bio.utexas.edu/faculty/antisense/garli/Garli.html).For MP analyses, characters were equally weighted and gaps were treated as missing data and not included in subsequent analyses.Initial MP searches used NNI branch swapping with 1,000 random addition replicates saving 5 trees per replicate.This generated a pool of starting trees that were subsequently used for more rigorous searches using TBR branch swapping.Relative support for the clades recovered was assessed by use of 1,000 jackknife replicates with 37% character deletion, enforcing the emulate ''jac'' option, TBR branch swapping, and two random additions per jackknife replicate and holding two trees per replicate.Before ML analyses ModelTest (Posada and Crandall 1998) was used to select the appropriate molecular model employing the AIC criterion.Heuristic ML analyses were conducted using best-fit model with Garli (Zwickl 2006).Relative support for the ML topology was assessed by use of 500 bootstrap replicates (Felsenstein 1985) as implemented in Garli.
Flavonoid chemistry Vegetative material was collected from natural populations (Table 2), and standard paper (2D) and thin-layer chromatographic methods were used to isolate and purify compounds (Mabry et al. 1970;Markham 1982).Purified compounds were identified by ultraviolet spectroscopy, by use of standard diagnostic reagents (Mabry et al. 1970;Markham 1982).Flavonoid glycosides were hydrolyzed according to the method of Wilkens and Bohm (1976), and sugars were identified by use of the circular chromatography method of Becker et al. (1977).Glycosides resistant to acid hydrolysis were assumed to be C-glycosyl flavonoids (sugars attached by carbon-carbon rather than oxygen bonds).Chromatographic and spectral data were also useful in identifying these carbon glycosides.Lack of sufficient amounts of plant material precluded exhaustive analyses of all compounds.Some identifications of flavonoid glycosides were based on cochromatography on thin-layer plates with several solvent systems.In rare instances when there was insufficient yield for spectral analyses, the class of compound was inferred from R F values (relative mobilities) in several solvent systems using paper and thin-layer chromatography.Color change of compounds on chromatograms when fumed with ammonia over ultraviolet light was also used to infer the structure of flavonoids (Mabry et al. 1970;Markham 1982).
Character evolution Patterns of flavonoid evolution and inflorescence architecture were explored using MacClade (Maddison and Maddison 2005) and the ML topology.Five categories of flavonoids were variously present within Calopappus acerosus, species of Nassauvia, and Triptilion spinosum: flavonols, flavones, C-glycosyl flavones, methylated C-glycosyl flavones, and methylated flavonols (Table 2).These were each considered a separate character with two states (present or absent), and were traced on to the ML estimate of phylogeny.The coding of inflorescence type followed that of Freire et al. (1993), with three character states (synflorescence, pseudocephalium, and solitary capitula) present among the ingroup taxa.

ITS sequence data
The aligned nrDNA matrix for 53 members of Mutiseae, including 23 species from Nassauvia, comprised 708 characters.Maximum parsimony (MP) and maximum likelihood (ML) analyses recover highly similar topologies, thus only the ML topology is discussed here (Fig. 1).Our analyses provide strong support (95% ML bootstrap) for a Together, these subclades form a weakly supported clade (61% ML bootstrap) that is sister to the remainder of Nassauvia.The four accessions from the three members of section Panargyrum (i.e., N. looseri, N. darwinii, and N. aculeata) presently sampled are all placed within a single, weakly supported clade (59% ML bootstrap); however, these taxa are not resolved as monophyletic (Fig. 1).Our analyses place two members of section Nassauvia (i.e., N. chubutensis and N. pulcherrima) within this same clade.
Weakly recovered as sister to this clade is a well supported clade (97% ML bootstrap) comprising the remaining 11 taxa from section Nassauvia sampled for this study.First branching in this clade is one of two accessions of N. revoluta; the second accession of this species is placed as sister to N. pinnigera (80% ML bootstrap, but\50% MP jackknife), with N. argentea resolved as sister to this clade with less than 50% ML bootstrap (Fig. 1).Although this clade is only weakly supported, the moderate support for the sister relationship between N. pinnigera and N. revoluta-1 to the exclusion of N. revoluta-2 suggests that this taxon is not monophyletic.The remaining seven species from section Nassauvia (i.e., N. argyrophylla-N.dusenii) are resolved in a weakly supported clade (\50% ML bootstrap) that is sister to a moderately supported (75% ML bootstrap) subclade comprising the remaining individual sampled from subgenus Strongyloma (i.e., N. axillaris) and two taxa from section Mastigophorus (N.pygmaea and N. juniperina).

Synflorescence evolution
The molecular phylogeny suggests that the evolution of inflorescences has been complex.The original condition within Nassauvia is the highly aggregated pseudocephalium; the first split within the clade includes a smaller clade consisting of subgenus Strongyloma, N. lagascae (section Caloptilium of subgenus Nassauvia and Triptilion) in which the ancestral condition is equivocal (Fig. 2).There could have been two losses of the pseudocephalium one in the ancestor of subgenus Strongyloma and one in Triptilion; alternatively, there could have been one loss of the pseudocephalium followed by one gain in N. lagascae (Fig. 2).The other major, and much larger, clade consists of all members of section Nassauvia together with sections Mastigophorus and Panargyrum (Fig. 2).Section Mastigophorus includes species that often, but not always, have solitary capitula.Although N. axillaris is placed in section Mastigophorus by analyses of ITS data, it groups with other representatives of subgenus Strongyloma in a plastid phylogeny (Maraner and Stuessy unpubl.).Indeed, the preponderance of evidence, including inflorescence type, places N. axillaris in subgenus Strongyloma (Freire et al. 1993).Two species of section Mastigophorus (i.e., N. juniperina and N. pygmaea) were included in the current analysis and, as discussed earlier, the latter species is unique in this section by having a pseudocephalium.Although additional sampling within the section is needed to determine the evolution of inflorescence types, it is clear that solitary capitula have been gained at least once whereas the pseudocephalium has been retained in N. pygmaea (Fig. 2).Within the clade that includes section Parnargyrum, together with N. chubutensis, and N. pulcherrima from section Nassauvia, there has been either one loss of the pseudocephalium with a subsequent gain in N. darwinii, or two losses of the pseudocephalium (Fig. 2).

Flavonoid components and evolution
A total of 32 flavonoid compounds was detected in Calopappus, Nassauvia, and Triptilion (Table 2).The classes of compounds included simple flavonols and their oxygen (O) glycosides, flavones, and their O glycosides, and C-glycosyl flavones in which the sugars are attached to the flavonoid nuclei by carbon rather than oxygen bonds.Two other classes of flavonoids were flavonols and C-glycosyl flavones in which there is methylation of a hydroxyl group.Two of the components were not identified to class because of the small amount of plant material and small concentrations of compounds in the plants.Four compounds were classified only to class, with two flavones and two flavonols inferred from chromatographic properties.
Tracing the flavonoid classes on to the ML ITS tree suggests a complex pattern of secondary chemistry evolution, with each class showing several gains-losses.Optimizing flavonols on to the ML topology indicates that the ancestral condition for the Nassauvia clade is presence of flavonols.Within Nassauvia, parsimony reconstructs either four separate losses of flavanols or two losses followed by two independent gains of flavanols (Fig. 3a).
The patterns of evolution are complex for flavones and their presence is inferred as ancestral (Fig. 3b).Employing equivocal cycling in MacClade indicates that there are ten possible reconstructions of this feature.The most common pattern recovered involves three losses of flavones, with one each noted for N. lagascae and N. fuegiana (Fig. 1 and clade A, Fig. 3b) and a loss for the clade comprising N. argentea to N. pulcherrima, with eight independent gains within this large clade (clade F, Fig. 2b).
The absence of C-glycosyl flavones is interpreted as ancestral in the Nassauvia clade.The parsimony reconstruction reveals unequivocally that the presence of C-glycosyl flavones has arisen five times within this clade.There are two independent gains in each of the first two branching lineages and a single gain with four apparent reversals noted in the large clade comprising N. revoluta-2 to N. argentea (clade A, Fig. 3c).The evolution of methylated C-glycosyl flavones is less complicated.Our analyses indicate that this class of secondary compounds has arisen three times within the Nassauvia clade (Fig. 3d).The absence of methylated flavonols is constructed as ancestral, and there have been either five gains or four gains and with a subsequent loss in the N. chubutensis-N.darwinii subclade (clade B, Fig. 3e).

Relationships among genera and the major clades of Nassauvia
The placement of the genera Calopappus and Triptillion in the ITS phylogeny may be considered in the context of previous taxonomic-phylogenetic treatments.The former genus is monotypic, and has been treated as either an independent genus (Crisci 1974;Cabrera 1982;Katinas et al. 2008a, b) or as a member of Nassauvia (Bremer 1993).The primary features distinguishing Calopappus from Nassauvia are the larger capitula and differences in pollen, whereby the former has an ectosexine that is thicker than the endosexine, and with both these layers columellate and divided by a zig-zag internal tectum (Crisci 1974a;Tellerı ´a et al. 2003).The morphological cladistic analysis of Freire et al. (1993) showed Calopappus nested within Nassauvia.The results of this study are in agreement with Crisci and Freire (1986), who viewed Calopappus as sister to a Nassauvia-Triptilion group, and the molecular phylogenetic study of Simpson et al. (2009)  Fig. 2 Maximum likelihood topology inferred from nrDNA ITS data with inferred character state changes in inflorescence.Black shows pseudocephalia, blue shows synflorescence, and red shows solitary capitula consists of 12 species of annual or perennial herbs, has been treated by most workers as a distinct genus (Cabrera 1982;Bremer 1993;Katinas et al. 2008a, b).The genus is morphologically similar to Nassauvia, with the nature of the pappus being the only consistent distinguishing character (Cabrera 1982;Bremer 1993).In a cladistic analysis of morphological characters (Freire et al. 1993), Triptilion was resolved as sister to a clade that includes Calopappus and Nassauvia.The ITS phylogeny places Triptilion within Nassauvia (Fig. 1) and is in agreement with the results of Simpson et al. (2009) that place Triptilion within Nassauvia.Cabrera (1982) considered Triptilion as most closely related to basal elements within Nassauvia section Panargyrum, but the molecular phylogeny places it in a subclade with N. lagascae (section Caloptilium) and that subclade is sister to species of subgenus Strongyloma in a weakly supported clade (61% ML bootstrap; Fig. 1).
There is general support for the sister group relationship between the two subgenera of Nassauvia (Cabrera 1982;Freire et al. 1993).Cabrera (1982) considered section Caloptilium (N.lagascae) to be ''derived'' and it is embedded in section Panargyrum on the basis of morphological analyses (Freire et al. 1993).The molecular phylogeny suggests different placement of this monotypic section, with sections Caloptilium, Triptilion and subgenus Strongyloma allied in one clade (Fig. 1).The second major clade includes two subclades, one comprising primarily section Panargyrum and the other including members of sections Mastigophorus and Nassauvia.Section Mastigophorus seems to be derived from within section Nassauvia (Fig. 1).By contrast, the morphological phylogeny of Freire et al. (1993) places section Mastigophorus in a large clade that is sister to section Nassauvia rather than nested within it.

Relationships among sections of Nassauvia
The most recent monograph of Nassauvia (Cabrera 1982) recognized five sections within the genus.Included in this study is N. lagascae of the monotypic section Caloptilium, and two members from section Mastigophorus, four members of section Panargyrum, and 13 accessions (12 species) of section Nassauvia.The results of our MP and ML analyses indicate that the sections of the genus, as currently defined, are not monophyletic.There is strong support (100% ML and 100% MP bootstrap) for a subclade comprising three of the four members of section Strongyloma (Fig. 1), but strongly excluded from this subclade is N. axillaris.Similarly, the four members of section Panagyrum are all placed within a single clade, but nested within this subclade is N. chubutensis with N. pulcherrima supported as sister to this clade with moderate to strong support (59% ML bootstrap, but 99% MP bootstrap).Most taxa sampled from section Nassauvia are placed within a large, well supported clade (97% ML and 100% MP bootstrap); however, nested within this and rendering these taxa paraphyletic is a moderately supported clade (75% ML and 61% MP bootstrap) including two members of the putatively non-monophyletic section Mastigophorus and N. axillaris (section Strongyloma).

Evolution of inflorescences, geographic distribution, and ecology
One of the characteristic features of Nassauvia and related genera is the reduction of the number of florets within capitula and the secondary aggregation of capitula, with the largest and most dense aggregations forming pseudocephalia.Secondary inflorescences (synflorescences) are rare but widely distributed in Asteraceae (Kunze 1969;Crisci 1974).Because they are rather peculiar morphologically and rare, synflorescences have been viewed as evolutionarily advanced (Stuessy 1978).However, within Nassauvia and closely related genera such as Calopappus and Triptilion there have been various phylogenetic interpretations of the different levels of aggregations, which vary from solitary capitula to highly condensed pseudocephalia.The pseudocephalia of sections Nassauvia and Caloptilium have generally been viewed as derived.However, in the morphological cladistic analysis of Freire et al. (1993), all other inflorescence types were shown as having originated from the synflorescence type found in Triptilion.Freire et al. (1993) thus viewed the solitary capitula of section Mastigophorus as derived.In contrast, Stuessy (1978) suggested that the solitary capitula of section Mastigophorus represent the ancestral condition in Nassauvia, with all other inflorescence types evolving from solitary heads.The results of Tortosa et al. (2004) and Katinas et al. (2008a, b) indicate that the more complex synflorescence type is ancestral in Nassauvia, in agreement with the hypothesis of Freire.
Inflorescence type may be examined relative to the geographic distribution and ecology of the elements in Nassauvia within the framework of the general hypothesis that the highly aggregated capitula would be at an advantage at higher altitudes where pollinators are at a premium (Bingham 1998).Presumably, the larger aggregations would be more efficient in attracting pollinators and pollination would be more efficient (Stuessy unpubl.).The members of section Nassauvia occur primarily in the Andes, and are characterized by the ancestral condition of highly aggregated pseudocephalia.This suggests that Nassauvia originated in the mountains, and there were subsequent migrations to other areas.An ancient divergence within Nassauvia produced the ancestor of subgenus Strongyloma, which subsequently migrated to lower elevations in Patagonia, with the evolution of synflorescences consisting of a dichasial inflorescence on short stalks originating from the axils of the upper leaves.

Evolution of flavonoid compounds
The evolution of flavonoid classes in the Nassauvia-Triptilion group (hereafter referred to as Nassauvia unless specifically considering Triptilion) has been rather complex, and as such flavonoids are not useful for delimiting groups within the genus.The presence of flavonols is the ancestral condition; four of the species of Nassauvia sampled for this study have lost these compounds.These taxa are placed in two separate subclades, each with a single taxon that has flavonols (Fig. 3a).This pattern can be explained by four separate losses; two losses with a single reversal in each subclade; or some combination of those scenarios.A comparison of the taxonomic distribution of flavonol loss with gain of methylated flavonols (Fig. 3a, e) shows that in several instances the loss of flavonols occurs in the same taxa showing gain of methylated flavonols.This pattern could be explained because the flavonols serve as a substrate for methylation and thus the methylated derivatives are accumulated at the expense of sequestering simple flavonols (Bohm and Stuessy 2001, Chap. 5).
Interpreting the evolution of flavones is more complex.MacClade provides 10 equally parsimonious reconstructions of these compounds.The presence of flavones is the ancestral condition within Nassauvia.The most common pattern recovered involves three losses of flavones.Two losses are inferred within the N. lagascae-N.fuegiana subclade (clade B;Fig 3b).A third loss can be inferred for the remaining ingroup taxa (clade A; Fig. 3b), with eight reversals to flavone presence; however, other combinations of gains and losses within this subclade are possible.
The parsimony reconstruction of the evolution of C-glycosyl flavones indicates that these compounds were absent in the ancestor of Nassauvia.Five separate gains of C-glycosyl flavones are inferred; two gains are inferred within the first two branching subclades of Nassauvia.A fifth gain of C-glycosyl flavones is ancestral for the N. argentea-N.revoluta subclade (clade A; Fig. 3c), with four independent losses in this same subclade.Methylated C-glycosyl flavones occur in a small subset of the taxa where C-glycosyl flavones are found (Fig. 3c, d).We infer three independent gains of methylated C-glycosyl flavones (Fig. 3d).Because methylation is a late step in the biosynthetic pathway to C-glycosyl flavones (Bohm and Stuessy 2001, Chap. 5), it follows that methylated C-glycosyl flavones would occur only in those taxa that synthesize simple C-glycosyl flavones.
The absence of methylated flavonols is constructed as ancestral in Nassauvia (Fig. 3e).There have been either three or four gains of methylated flavones; three unequivocal gains are inferred for N. fuegiana, N. dentata, and the subclade including N. juniperina and N. axillaris (clade A; Fig. 3e).The ancestral state for the N. chubutensis-N darwinii subclade is equivocal (clade B; Fig. 3e).Thus it is equally parsimonious that this represents two separate gains, or a single gain with a subsequent reversal.
The evolution of flavonoid classes during the diversification of Nassauvia has been complex.The gains or losses have occurred throughout the clade, in some cases only in terminals and in others at deeper branches in the phylogeny.However, looking more closely it is evident that gains and losses of flavonols (including methylated derivatives) occur only in terminal taxa whereas changes in flavones and C-glycosyl flavones occur both early in the diversification of the clade and in numerous terminals (Fig. 3a, b, c).The gain of methylation occurs only in terminals (Fig. 3d, e).These collective patterns of gains and losses, although interesting and comprehensible within a biosystematic context, make the use of flavonoids per se of very limited taxonomic-phylogenetic utility in Nassauvia.

Fig. 3
Fig. 3 MacClade reconstructions of changes in classes of flavonoids on the ML nrDNA ITS tree for the following classes of secondary compounds: a flavonols (red loss); b flavones (red loss); c C-glycosyl flavones (red gain); d methylated C-glycosyl flavones (red gain); and e methylated flavonols (red gain).Subclades discussed in text are indicated by blue brackets; for e, *N.dentata; # N. fuegiana

Table 1
Collections of Nassauvia and related Mutisieae genera analyzed in this study

Table 2
Distribution of flavonoid compounds in Calopappus, Nassauvia, and Triptilion that also resolved Calopappus as sister to a Nassauvia-Triptilion clade.In contrast to Calopappus, the genus Triptilion, which