Preliminary molecular phylogenetic studies in Pachyanthus (Miconieae, Melastomataceae).
Pachyanthus A. Rich. (Melastomataceae: Miconieae) is traditionally circumscribed as a group of 20 species of Caribbean shrubs, mostly local Cuban endemics (Borhidi, 1985; Judd & Skean, 1991; Liogier, 1957). However, the exact number of species is unknown because generic delimitations within Miconieae, a neotropical group of over 2,000 species, are mostly unclear and in many instances arbitrary and non-monophyletic (Wurdack, 1973; Judd, 1986, 1989: Judd & Skean, 1991; Michelangeli et al., 2004). The genus was described by A. Richard in 1845 based on a specimen collected on Isla de Pinos (Isla de la Juventud), Cuba, which was named Packyanthus cubensis (Richard, 1845).
Richard (1845) suggested that Pachyanthus was near Tetrazygia L.C. Rich. ex DC., but differed in having flowers with five sepals and petals and a trilocular ovary. The genus was placed next to Miconia Ruiz & Pavon by Grisebach (1866), who distinguished it on the basis of its truncate to fove to six denticulate calyx lobes, rigid, dry berry, and larger flowers (in almost all the species). Liogier (1957) defined Pachyanthus as a genus characterized by terminal, few-flowered inflorescences, with large, usually five- or six-merous flowers having a tomentose-furfuraceous, cylindrical or flared hypanthium/calyx, with each calyx lobe having a dorsal appendage. On the other hand, Judd and Skean (1991), suggested that Pachyanthus may be a monophyletic group because its species show moderate to large six-merous flowers (secondarily five-merous), more or less robust fruits with a cylindrical-campanulate hypanthium that is coriaceous and persistent (often torn into segments similar to sepals), and inflorescences with flattened axes usually bearing only one to seven flowers. Judd and Skean (1991) also pointed out that stellate hairs, sometimes reduced, are present in the group.
Borhidi (1985) recognized 24 species of Pachyanthus on Cuba and Hispaniola and one problematic species from Colombia. Judd and Skean (1991) excluded the Colombian species, Pachyanthus corymbiferus (Naudin) Cogn., accepting the classification of Naudin (1851) in placing this species in the monotypic genus Chalybea Naudin. Additionally, they transferred to Pachyanthus the Central American species, Miconia lundelliana, as earlier suggested by Wurdack (1988).
Some species of Pachyanthus have been considered phenetically similar to members of the Caribbean genus, Calycogonium DC. (Judd, 1986; Urban, 1926). Judd (1986) also noted similarities with some Antillean species of Ossaea, pointing out that the relationships among these taxa are in need of investigation. Finally, we note that some species that have been placed in Pachyanthus in the past are now considered to belong in Tetrazygia, e.g. Tetrazygia urbaniana and Tetrazygia tuerckheimii (Cogn.) Ekman ex Urb. (Judd & Skean, 1991; Liogier, 2000).
It is evident that there has never been a consistent generic delimitation of the genus Pachyanthus. Neither is there agreement among systematists that have studied this genus as to which morphological characters should be used to distinguish it, e.g., inflorescence position, number of flowers per inflorescence, floral merosity and size, and/or fruit form.
The objective of the current study is to produce a preliminary molecular phylogeny of Pachyanthus and related genera in order to test the monophyly of Pachyanthus. We also aim to develop a hypothesis regarding the relationships of Pachyanthus to other Antillean genera of Miconieae such as Calycogonium, Tetrazygia, and certain species of Miconia. In this study, we also discuss the potential synapomorphies of Pachyanthus s.s. (and clades containing species of Pachyanthus s.l.) based on ongoing unpublished morphological studies in this genus by the first author.
Materials and Methods
Taxon Sampling and Outgroup Selection
In order to study phylogenetic relationships of Pachyanthus and related Antillean species, 39 taxa were sampled; 38 of these belong to Miconieae and one to Merianieae (Appendix 1). The ingroup taxa include 17 species currently placed in Pachyanthus of the 24 recognized by Borhidi (1985). These species represent most of the morphological variation seen in the genus and the remaining species were not available because they are known from only a few specimens. Pachyanthus corymbiferous (Naudin) Cogniaux, a species from the Andes of Colombia, was not included in this analysis because it has been already shown that it is more closely related to Huilaea Gleason in the tribe Blakeeae, and should be considered as Chalybea corymbiferous Naudin, as originally described (Judd & Skean, 1991; Wurdack, 1988; Morales & Gonzalez, 2005). Additionally, 21 species of Calycogonium, Miconia, and Tetrazygia were included in the analysis in order to test the monophyly of the genus. These species are mostly Antillean Miconieae mainly Cuban endemics, and they have morphological similarities to, or are potentially the closest relatives of Pachyanthus (see Michelangeli et al. this volume). Miconia laevigata, and widespread species, distantly related to the Antillean taxa mentioned above was included as part of the outgroup. Finally, we used Graffenrieda latifolia of the tribe Merianieae, which is considered sister to Miconieae (Clausing & Renner, 2001; Michelangeli et al., 2004), to root the resulting trees. Voucher specimens of all taxa utilized in the DNA analyses are deposited at HAJB, FLAS, or JBSD (Appendix 1).
DNA Isolation and Sequencing
Genomic DNA was isolated from material dried in silica gel or from herbarium material, using the method of Doyle and Doyle (1987), but modified to use 5 [micro]L of proteinase K in the CTAB solution instead of [beta]-mercaptoethanol. The internal transcribed spacer region of the nuclear ribosomal DNA (nrITS) and the trnS-G spacer of the chloroplast were amplified and sequenced. Polymerase chain reactions (PCR) were performed on an Eppendorf Mastercycler EP Gradient S thermocycler in ~25 [micro]L reactions using Eppendorf products: 1-3 [micro]L of template, 2.5 [micro]L of 10x buffer (including 15 mM Mg [(OAc).sub.2] premixed), 2.5 [micro]L of 25 mM Mg[(OAc).sub.2], 5 [micro]L of TaqMaster PCR Enhancer, 0.5 [micro]L of 10 [micro]M primers, 0.5 [micro]L of 10 mM dNTPs, and 1 unit of Taq (polymerase), with water comprising the remaining volume. The PCR parameters for ITS were 99[degrees]C, 10 min; 94[degrees]C hold for polymerase addition; 33x (94[degrees]C, 45 s; 65[degrees]C, 1 min; 72[degrees]C, 1 min); 72[degrees]C, 3 min with primers 17SE (ACGAATTCATGGTCCGGTGAA GTGTTCG) and 26SE (TAGAATTCCCCGGTTCGCTCGCCGTTAC) from Sun et al. (1994). The PCR parameters for trnS-G ([trnS.sup.GCU][-trnG.sup.UUG]) were 94[degrees]C, 3 min; 30x (94[degrees]C, 45 s; 58[degrees]C, 45 s; 72[degrees]C, 1 min); 72[degrees]C, 3 min with primers [trnS.sup.GCU] (AGATAGGGATTCGAACCCTCGGT) and 3'[trnG.sup.UUG] (GTAGCGGG AATC GAACCCGCATC) from Shaw et al. (2005).
The PCR product was cleaned with Microclean[TM] following manufacturer's protocols, eluted with 50 [micro]L of 10 mM Tris-Cl (pH 8.5), and stored at 4[degrees]C. Cycle sequencing was performed with the same amplification primers using Big Dye dideoxy terminator and Better Buffer[TM]. The cycle sequencing parameters were 96[degrees]C, 10 s; 25 x (96[degrees]C, 10 s; 50[degrees]C, 5 s; 60[degrees]C, 4 min).The products were run in an automated sequencer, Perkin Elmer Applied Biosystems, Inc. (ABI)3130, of the Interdisciplinary Center for Biotechnology Research at the University of Florida. All sequence data were edited using Sequencher[TM]. Voucher information and GenBank accession numbers for the taxa included in this study are summarized in Appendix 1.
Sequences Alignment and Phylogenetic Analysis
DNA sequences were initially aligned with ClustalX (Thompson et al., 1997), and then the gap edges adjusted manually using Se-Al v2.0a11 (Rambout, 1996). Once the alignment was finished, both ends of the matrices were trimmed to avoid sequence artifacts. During the phylogenetic analyses gaps were coded as missing values, and all base characters were weighted equally. Three different phylogenetic analyses were conducted as follows: the nrITS (39 taxa) and the plastid trnS-G (34 taxa) data were each analyzed separately, and then a final combined analysis was conducted for the 34 taxa for which we had data from both regions. Parsimony analyses of the individual and combined datasets where performed in Nona (Goloboff, 1993), running it from Winclada (Nixon, 1999-2002). One thousand random addition sequence replicates holding 20 trees per replicate were conducted. Each starting tree was first exhaustively swapped with SPR then TBR. After the random addition sequences were complete, trees were swapped to 500,000 trees using TBR. Winclada (Nixon, 1999-2002) was also used for tree viewing and character optimization.
Bootstrap support values were calculated from 1,000 replicates. Each replicate was performed with 20 random taxon-entry sequences and swapped using the parameters described above. Branches were collapsed if supported ambiguously. The consensus tree was retained from each replicate, and all 1,000 consensus trees used to calculate the support value.
For the nrITS region 798 nucleotide sites were unambiguously aligned and included in the analysis. Of these sites, 62 were potentially informative. The trnS-G spacer matrix was 854 bp long and also had 62 potentially informative sites. The analysis of the nrITS data set by itself resulted in 15 most parsimonious trees (MPTs) of 119 steps (CI=0.59; RI=0.84). The trnS-G spacer data matrix yielded 54 MPTs of 95 steps (CI=0.72; RI=90). Analysis of the combined matrix of all 39 taxa resulted in 22 MPTs of 234 steps (CI=0.59; RI=0.83). Lastly, analysis of the reduced combined matrix resulted in 4 MPTs of 220 steps (CI=0.61; RI=0.83).
Analysis of nrITS Sequences Data
The strict consensus tree, obtained from the analysis of nrITS sequence data, clearly demonstrates that Pachyanthus is not monophyletic (Fig. 1). Pachyanthus monocephalus and Pachyanthus moaensis are placed in a clade along with Calycogonium rosmarinifolium and Miconia cerasiflora (Fig. 1, Clade A), while the remaining species of Pachyanthus are placed within the larger sister clade, which also includes species of Miconia, Tetrazygia, and Calycogonium.
Among the species sampled, M. laevigata (section Miconia) is isolated, while the representatives of Miconia sect. Chaenopleura constitute a strongly supported clade (Fig. 1, Clade B). The remaining species form a clade (without bootstrap [BS] support), and within this group, a distinctive subclade (Fig. 1, Clade C) has only species of Pachyanthus (BS=67%). Other species of Pachyanthus are scattered within the large clade that constitutes the sister group of the Pachyanthus-s. str. clade, i.e., in Clades D, E, F, and H (Fig. 1). We note that Pachyanthus discolor, Pachyanthus mayarensis, and Pachyanthus neglectus are evidently related (Fig. 1, Clade F), as are Pachyanthus blancheanus and Pachyanthus pedicellatus (Fig. 1, Clade H).
Tetrazygia crotonifolia and T. urbaniana are sister taxa (BS=67%; Fig. 1, Clade G), and are related to several other species of Tetrazygia (as well as Calycogonium, Miconia) and P. blancheanus and P. pedicellatus (Fig. 1). Four species of Tetrazygia (Fig. 1, Clade I) form a monophyletic group with very high support (BS=99%).
[FIGURE 1 OMITTED]
Analysis of trnS-G Chloroplast Sequences Data
The trees resulting from the chloroplast gene trnS-G sequence analysis are similar to the results of the ITS analysis in supporting a non-monophyletic Pachvanthus (even though five species are missing from this analysis, i.e., P. blancheanus, P monocephalus, P. neglectus, Tetrazygia discolor, and T. urbaniana). A strict consensus of the cladograms resulting from the analysis is presented in Fig. 2. Like the ITS analysis, nearly half of the species of Pachyanthus constitute a single clade (Fig. 2, F'), and the membership of this Pachyanthus-s. str. clade is identical to that of the ITS-based analysis (Figs. 1, 2).
The first node of the consensus tree shows a polytomy. Miconia laevigata is isolated, and P. moaensis is placed with M. cerasiflora and C. rosmarinifolium (Fig. 2, Clade A'), in agreement with the results of the ITS analysis. Again, the species belonging to Miconia section Chaenopleura form a monophyletic group (Fig. 2, Clade B'). The remaining species in the analysis form a large and well-supported clade (BS=97%). Relationships are problematic within this large group but five subclades are evident (Fig. 2, Clades C'-G'). Species of Pachyanthus are found within three of these clades (i.e., Fig. 2, Clades C', F', and G'). P. pedicellatus and Pachyanthus reticulatus (Fig. 2, Clade C') are resolved as sister taxa with strong support (BS=98%), but these two species do not form part of the same clade in the nrITS results (Fig. 1). Clade F' has eight species of Pachyanthus (the same as in the nrITS analysis), and as in the nrITS analysis P. cubensis is sister to the remaining species. However, in the trnS-G analysis Miconia delicatula is resolved as the sister species of the Pachyanthus clade (Fig. 2, Clade F'). Three species of Pachyanthus are placed in Clade G' (Fig. 2), along with species of Calycogonium and Tetrazygia. Relationships within this clade are similar to those within Clade D + E + F in the ITS analysis.
[FIGURE 2 OMITTED]
As in the ITS analysis, Tetrazygia bicolor, Tetrazygia barbata, Tetrazygia coriacea, and Tetrazygia lanceolata form a clade (Fig. 2, E') with at least moderate support (BS=85%). But, Tetrazygia as currently circumscribed is not monophyletic because other species group elsewhere on the cladogram (i.e., Fig. 2, Clades, D' and G').
Combined Data Analysis of nrITS and trnS-G Sequences
The trees resulting from the combined data analyses (see strict consensus cladogram, Fig. 3), not surprisingly, also strongly indicate that Pachyanthus, Miconia, Calycogonium, and Tetrazygia are not monophyletic. As in the trees resulting from the analyses of trnS-G and ITS sequences, these trees strongly support (BS=98%) the monophyly only of a Pachyanthus sensu stricto (which is represented here by P. cubensis, Pachyanthus tetramerus, Pachyanthus lundellianus, Pachyanthus wrightii, Pachyanthus angustifolius, Pachyanthus poiretii, Pachyanthus clementis, and Pachyanthus mantuensis; see Fig. 3, Clade E"). Within this monophyletic group, P. cubensis is well supported as sister to the remaining species; (Fig. 3). The placement of M. delicatula as the sister taxon to the Pachyanthus s. str. group receives some support (BS = 77%), and this relationship was also evident in the trnS-G analysis. The remaining species of Pachyanthus are placed elsewhere: P, moaensis is in a clade with M. cerasiflora and C. rosmarinifolium (Fig. 3, Clade A"), which is sister to the remaining taxa; P. pedicellatus, and P. reticulatus are placed alone in the larger basal polytomy, while Pachyanthus monopleurus, P. discolor, and P mayarensis are interdigitated with various species of Tetrazygia and Calycogonium in Clade C" (BS=52%; Fig. 3).
[FIGURE 3 OMITTED]
As in the previous trees the three species representing Miconia section Chaenopleura form a well-supported clade (BS=99%; Fig. 3, Clade B"). A clade with the same species of Tetrazygia that come together in all analyses (i.e., T. barbata, T. lanceolata, T. bicolor and T. coriacea; Fig. 3, clade D") is strongly supported (BS= 100%) and is placed as sister to a Pachyanthus s. str. + M. delicatula clade (with a relatively low BS of 56%). We note that Tetrazygia decorticans, Tetrazygia aurea, Tetrazygia elegans, and T. crotonifolia are placed within Clade C" (Fig. 3), but their relationship to one another is in need of further study.
It is evident in all the analyses (nrITS; trnS-G; combined nrITS and trnS-G) that Pachyanthus as currently circumscribed is not monophyletic, as has been shown for most of the major genera of the Miconieae by Michelangeli et al. (2004). Although sampled less densely, Calycogonium and Tetrazygia are also shown to be polyphyletic in the current investigation. These results are not surprising because the genera of Miconieae have long been considered to be poorly characterized (Gleason, 1932), difficult to discern, and quite arbitrary (Cogniaux, 1891; Macbride, 1941; Wurdack, 1972). This fact is easily seen in the complex keys to the genera of Miconieae in various tropical floras, e.g., Liogier (1957), Adams (1972), Liogier (2000), and Wurdack (1973, 1980).
However, it is possible to define a well supported (BS=98%, Fig. 3) monophyletic group, here called Pachyanthus s. str., that includes several species of Pachyanthus endemic to Cuba (P. angustifolius, P. clementis, P. mantuensis, P. poiretii, P. tetramerus, and P. wrightii), a species from Central America, P. lundellianus, and the type of the genus, P. cubensis, also endemic to Cuba. These species form a clade in all analyses (see Figs. 1, 2, and 3). Based on an ongoing taxonomic revision of Pachyanthus (Becquer, in prep.) we are confident that one additional Cuban species, P. longifolius Jennings, also belongs to this well supported and morphologically distinctive clade (see discussion of morphological features below).
Pachyanthus s. str is characterized by conspicuous flowers (2-4 cm in diameter) that are (four-) six-merous and have thick, ovate to obovate, conspicuously unguiculate petals. The petals are generally white (but are pink in P. cubensis). The margins of the petals are membranaceous and plaited near the base. The petals are thus distinctive, and can be used to readily distinguish members of this clade. Based on on-going morphological phylogenetic studies (Becquer, in prep.), it is probable that most of these features are synapomorphic for this clade. It is of interest that petal morphology provides the most diagnostic features for the Pachyanthus s. str. clade because petal characters (in conjunction with other floral characters) are useful in differentiating (and possibly synapomorphic for the members of) other genera and/or species complexes in the Miconieae (Judd, 1986, 1989; Judd & Skean, 1991; Michelangeli, 2000). Nonetheless, the presence of clawed petals in Pachyanthus has not been reported in recent literature, despite that Grisebach (1866) mentioned it as a distinctive feature of P cubensis. Even though clawed petals occur in a few other groups of Melastomataceae, e.g., some species of Henriettea DC., the combination of this character with the other floral features (as enumerated above) can be used to distinguish species of Pachyanthus s. str. from the remaining genera of Miconieae (Fig. 4a).
The stamens, 10-12 (rarely eight), are isomorphic and deflexed to one side of flower. Anthers are ovate to oblong, straight, yellow, without connective appendages or, if these are present, then they are inconspicuous; dorsal glands are also absent. The anthers dehisce by small apical pores. If we take into account the great variability of anther shape in certain groups of Melastomataceae (Michelangeli et al., 2004), the anthers of Pachyyanthus s. str. are quite simple. This was noted by Judd and Skean (1991) who suggested a basal divergence of Pachyanthus relative to the origin of most species of Miconia.
The calyx is persistent and coriaceous, and in anthesis the calyx lobes are almost totally free, rotate to reflex, thus, a calyx tube as seen in most species of Miconieae is not evident. Judd and Skean (1991) pointed out that in some species of Pachyanthus the calyx lobes tear into sepal-like segments at anthesis. During this study we have seen that in those species the calyx appears fused in bud, but upon close inspection it is evident that there are valvate demarcation lines prior to anthesis, and the "lobes" are the product of their separation. This feature may be a synapomorphy for all of Pachyanthus s. str. except for the basal species P. cubensis (Becquer, 2008). External calyx teeth are usually conspicuous, again except for P. cubensis.
The species of Pachyanthus s. str. have axile placentation, as is characteristic of nearly all Miconieae (Renner, 1993; Judd et al., 2002; Clausing et al., 2000). There are numerous seeds per ovary, i.e., ca. 60 to 500, which are small (i.e., 0.8-1 mm long) in most species, but the seeds of P. cubensis are 1.5-1.6 mm. The seeds are obpyramidal-angular with a smooth testa (Fig. 4d).
The morphological differences between P. cubensis and the other species of Pachyanthus s. str. can be understood phylogenetically, since P. cubensis is sister to the rest of the species of Pachyanthus s. str., as discussed above. The position of P. cubensis may also explain why, even though it is possible to see putative hybrids in the field between some of the sympatric species, e.g., P. poiretii and P. angustifolius, P. cubensis is not known to form hybrids, even when co-occurring with other species (personal observation of first author). Relationships within Pachyanthus s. str. are investigated in more detail in a morphology-based phylogenetic analysis, which will be published with the first author's taxonomic treatment of this group (Becquer, in prep.).
The close relationship of M. delicatula to Pachyanthus is surprising given that there are no obvious morphological characters that unite the former species to the latter group, besides the fact that both share the same type of habitat in western Cuba. Perhaps with increasing taxon sampling within the large Caribbean clade (see Michelangeli et al. this volume) this relationship, as well as the relationship of M. delicatula to other species will be clarified.
The remaining species of Pachyanthus belong to other clades, as assessed in this preliminary analysis of Antillean species. When our results are compared with those of Michelangeli et al. (2004), it is seen that most of the species included in our analyses (i.e., clades C", D" and E" of Fig. 3) constitute a clade, which corresponds to clade G (in Fig. 2 of Michelangeli et al., 2004) and is almost exclusively restricted to the Caribbean region. More complete sampling of Antillean species of Miconieae is necessary before the phylogenetic placement of these species of Pachyanthus can be clarified, but it is clear that P. pedicellatus, P. reticulams, P. monopleurus, P discolor, P. mayarensis (and related taxa not included in these analyses i.e.P, oleifolius Griseb.; Becquer, 2008) are more closely related to various species of Calycogonium and Tetrazvgia than they are to species of the Pachyanthus s. str. clade.
[FIGURE 4 OMITTED]
The most isolated species is P. moaensis, which consistently groups with C. rosmarinifolium and M. cerasiflora (Fig. 3, clade A"), a group which likely also includes P. monocephalus (see Fig. 1, clade A). These species have several potential synapomorphies, e.g., small flowers (<1 cm), obovate to spatulate petals (Fig. 4c), and strongly fused internal calyx lobes. They have basal-axile placentation, with the placental tissue reduced (except in P. monocephalus), while most Miconieae have axile placentation with expanded and/or somewhat divided placentae. The fruits have only a few seeds (usually 15 to 20), and these are usually bigger than those of the Pachyanthus s. str. clade, i.e., 1.7-1.9 mm long, but 2.86-3 mm in M. cerasiflora (Fig. 4g). More study of this group is needed, especially a detailed survey of variation in placental form and position.
We stress that many groupings are poorly supported in our cladograms (i.e., BS values below 50%), the sampling of Antillean species needs to be increased, and the species considered here need to be addressed within the context of variation found in the Miconieae as a whole (with numerous continental representatives). The topology of our trees could easily change after the inclusion of additional species of Calycogonium, Tetrazvgia, or Miconia and other genes in view of the potential for conflict among data sets. Nonetheless, we discuss a few other intriguing groupings in the following paragraphs. These preliminary hypotheses will be tested through our ongoing phylogenetic studies of the tribe.
The small clade comprising P. mavarensis, P discolor, and P. neglectus (see nrITS analysis, Fig. l, and combined analysis, Fig. 3, subclade within clade C") even though it does not receive strong bootstrap support is supported by some morphological characters. These species share symmetric obovate petals, and fused internal calyx lobes at anthesis; their anthers are linear-lanceolate, narrowing toward the apex, usually curved, each with a conspicuous dorsal-apical pore. Also, these species have large seeds (i.e., 1.5-1.8 mm long) with a conspicuously papillose testa (Fig. 4e). These seeds characteristics, along with petal shape, the degree of fusion of the internal calyx lobes, and different anthers features, clearly separate these species of Pachyanthus from the Pachyanthus s.str, clade.
On the other hand, P. monopleurus groups with T. elegans in all analyses (Figs. 1,2, and 3). Pachyanthus monopleurus has some distinctive features such as the presence of peltate scales, absence of acrodromous venation and globose and cucullate external calyx lobes. Based on some of these characters, Urban (1924) considered that this species was not closely related to other species of Pachyanthus.
Pachyanthus reticulatus and P. pedicellatus form a clade in the trnS-G trees and in the combined analysis, but P. pedicellatus groups with P. blancheanus (a species from Haiti, Hispaniola) in the nrITS trees. Pachyanthus pedicellatus and P. blancheanus have an uncommon feature: a completely inferior ovary, but with the peripheral portion of the locules extending laterally up into the hypanthium, thus extending beyond the ovary apex, i.e., the point at which the ovary joins with the style. In other words, the hypanthium extends above the apex of the ovary with extensions of the locules creating an internal, hollow gap within the abaxial and adaxial surfaces of the hypanthium. Structurally, this could also be seen as the central portion of the ovary being deeply sunken with respect to the lateral portion. This character needs further study because the same condition is found in T. decorticans (Becquer-Granados, 2007), and other species, such as T. discolor, Calycogonium heterophyllum and C. glabratum have a fully inferior ovary (but without the peripheral ovary extensions). The rest of the species of Pachyanthus included in the analysis possess a partially inferior ovary, so that there is a central free portion of the ovary inside the laterally divergent hypanthium, which is fused with the ovary in the lower portion. Also they have ovate-lanceolate to linear-lanceolate anthers, i.e., the anthers are narrowed toward the apex, each with a conspicuous dorsal-apical pore. P. reticulatus has similar anthers, but they usually are curved, and are purple, whereas the anthers of R pedicellatus and R blancheanus are yellow (Fig. 4f).
Pachyanthus blancheanus was treated as a relative of P. cubensis by Urban (1927). Borhidi (1983) considered this taxon as a subspecies of P. cubensis, but it is very clear (both from morphology and nucleotide sequence data) that P. blancheanus is not a member of the Pachyanthus s. str. clade, and it, thus, cannot be a close relative of R cubensis. It is best maintained as a distinct species.
As mentioned above, Calycogonium is not resolved as monophyletic. Tetrazygia, an Antillean group of ca. 25 species (Howard, 1989), also is not monophyletic in any of our analyses. It is interesting that in the nrlTS analysis T. discolor, the type species of this genus, does not come together with any other species of Tetrazygia. Judd and Skean (1991) and Michelangeli et al. (2004) pointed out that this genus, as currently defined, is not monophyletic. The sampling of species within both of these genera needs to be improved before we can discuss their relationships. It is clear, however, that some species of Pachyanthus probably are closely related to particular species of Calycogonium and/or Tetrazygia. It appears that current generic delimitations among Pachyanthus (as broadly circumscribed), Calycogonium, and Tetrazvgia do not accurately reflect evolutionary relationships. The only strongly supported clades (which can also be easily diagnosed on the basis of morphological characters) are Pachyanthus s.s. (Fig. 3, clade D") and the basal-axile placentation clade (Fig. 3, clade A").
Pachvanthus, as traditionally defined, is not a monophyletic group. As shown here, Pachyanthus s. str. forms a clade, which is characterized by the petals ovate to obovate, and conspicuously unguiculate; calyx lobes separating into sepal-like segments at anthesis; anthers ovate to oblong, straight, yellow, usually without connective appendages; anthers dehiscing by small apical pores; and seeds numerous, small (1-2 mm), obpyramidal, with a smooth testa. The remaining species traditionally included in Pachyanthus do not have any of these features. The other species of Pachyanthus do not form a single clade, but instead belong within several other, poorly-supported clades. Additional detailed molecular and morphological studies are needed in the Miconieae to clarify phylogenetic relationships among these taxa, and to ascertain whether any of these clades should be treated as new genera or recircumscribed as part of currently recognized genera in the Caribbean region. While preliminary hypotheses are presented in this paper, especially relating to species traditionally placed in Pachyanthus, generic reclassification must await more detailed analyses to sort out the lineages within what have traditionally been classified as Calycogonium, Miconia, and Tetrazvgia, which are also shown not to be monophyletic.
Appendix 1 Taxa used for Molecular Phylogenetic Study Taxa Country GenBank GenBank # nrITS # trnsS-G Pachranthus angustilblius Cuba EU055870 EF549719 Griseb. P blancheanus Urb. & Ekman Haiti EU055871 P. clementis P. Wilson Cuba EU055872 EF549720 P. cubensis A. Rich. Cuba EU055873 EF549721 P. discolor Norlind Cuba EU055874 EF549722 P. lundellianus (L.O. Williams) Cuba EU055875 EF549723 Judd & Skean P. mantuensis Britton Cuba EU055876 EF549724 & P. Wilson P. mayarensis Urb. Cuba EU055877 EF549725 P. moaensis Borhidi Cuba EF418918 EF549726 P. monocephalus (Urb.) Borhidi P. monopleurus Urb. Cuba EU055878 P. neglectus Borhidi Cuba EU055880 P. pedicellattus Urb. Cuba EU055881 EF549728 P. poiretii Griseb. Cuba EU055882 EF549729 P. reticulatus Britton Cuba EU055883 EF549730 & P. Wilson P. tetramerus Urb. & Ekman Cuba EU055884 EF549731 P. wrightii Griseb. Cuba EU055885 EF549732 Calycogonium glabratzrm DC. Cuba EU055645 EF549708 C. grisebachii Triana Cuba EU055646 EF549709 C. cf. heterophyllum Naudin Cuba EU055647 EF549710 C. revolutum Alain Cuba EU055650 EF549711 C. rosmarinifolium Griseb. Cuba EU055651 EF549712 Miconia cerasiflora Urb. Cuba EU055743 EF549713 M. cubensis (Wright ex Griseb.) Cuba EU055762 EF549714 Wright in Sauv. M. delicatula A. Rich. Cuba EU055764 EF549715 M. laevigata (L.) DC. Dominican EU069392 EF549716 Republic M. pteroclacia Urb. Cuba EU055820 EF549717 M. turquinensis Urb. & Ekman Cuba EU055856 EF549718 Tetrazygia aurea R.A. Howard Cuba EU055889 EF549733 & W.R. Briggs T. barbata Borhidi Cuba EU055890 EF549734 T. bicolor Cogn. Cuba AY460543 EF549735 T. coriacea Urb. Cuba EU055891 EF549736 T. crotoniflia (Desr.) DC. Dominican AY460544 EF549737 Republic T. discolor (L.) DC. Dominica EF418920 T. elegans Urb. Cuba EU055893 EF549738 T. lanceolata Urb. Cuba EU055894 EF549739 T. urbaniana (Cogn. ex Urb.) Dominican EF418921 Croizat ex Moscoso Republic T. decorticans Becquer Cuba EU069393 EF549740 Groffenrieda latifolia Dominica AY460450 EF549707 (Naudin) Triana Taxa Voucher (herbarium) Pachranthus angustilblius HFC 81592 (HAJB) Griseb. P blancheanus Urb. & Ekman T. Clase 4190 (JBSD) P. clementis P. Wilson HFC 82465 (HAJB, FLAS) P. cubensis A. Rich. HFC 82418 (HAJB, FLAS) P. discolor Norlind HFC 82243 (HAJB, FLAS) P. lundellianus (L.O. Williams) J. Richard Abbott 19636 Judd & Skean (HAJB, FLAS) P. mantuensis Britton HFC 82419 (HAJB, FLAS) & P. Wilson P. mayarensis Urb. HFC 82496 (HAJB, FLAS) P. moaensis Borhidi HFC 82504 (HAJB, FLAS) P. monocephalus (Urb.) Borhidi P. monopleurus Urb. HFC 83731 (HAJB) P. neglectus Borhidi HFC 83737 (HAJB) P. pedicellattus Urb. HFC 81126 (HAJB, FLAS) P. poiretii Griseb. HFC 81675 (HAJB) P. reticulatus Britton HFC 82484 (HAJB) & P. Wilson P. tetramerus Urb. & Ekman HFC 82425 (HAJB, FLAS) P. wrightii Griseb. HFC 82269 (HAJB, FLAS) Calycogonium glabratzrm DC. HFC 82248 (HAJB, FLAS) C. grisebachii Triana HFC 82261 (HAJB, FLAS) C. cf. heterophyllum Naudin HFC 81151 (HAJB, FLAS) C. revolutum Alain HFC 82510 (HAJB, FLAS) C. rosmarinifolium Griseb. HFC 82482 (HAJB, FLAS) Miconia cerasiflora Urb. HFC 82487 (HAJB, FLAS) M. cubensis (Wright ex Griseb.) HFC 82455 (HAJB, FLAS) Wright in Sauv. M. delicatula A. Rich. HFC 82421 (HAJB, FLAS) M. laevigata (L.) DC. Penneys 1317 (FLAS) M. pteroclacia Urb. HFC 81639 (HAJB) M. turquinensis Urb. & Ekman HFC 81648 (HAJB, FLAS) Tetrazygia aurea R.A. Howard HFC 82461 (HAJB, FLAS) & W.R. Briggs T. barbata Borhidi HFC 82254 (HAJB, FLAS) T. bicolor Cogn. HFC 82286 (HAJB, FLAS) T. coriacea Urb. HFC 82423 (HAJB, FLAS) T. crotoniflia (Desr.) DC. Williams & Whitten s.n. (FLAS) T. discolor (L.) DC. Penneys 1287 (FLAS) T. elegans Urb. HFC 81655 (HAJB) T. lanceolata Urb. HFC 82431 (HAJB, FLAS) T. urbaniana (Cogn. ex Urb.) Skean 4137 (ALBION Croizat ex Moscoso College) T. decorticans Becquer HFC 82464 (HAJB, HAC, FLAS, B) Groffenrieda latifolia Penneys 1303 (FLAS) (Naudin) Triana Countries from which each specimen was collected are listed. Vouchers are listed by collector or serial number (e.g., HFC: Herbarium Florae Cubensis), and collector number; herbaria where vouchers are deposited are indicated in parentheses (acronyms from Index Herbariorum, Holmgren & Holmgren, 1998)
Acknowledgements We thank the staff of the herbarium of the Florida Museum of Natural History, University of Florida, especially W. Mark Whitten, Norris Williams, and Kent Perkins. Thanks also to Karen L. Kelley and Lynda Schneider from the Electron Microscopy Core Laboratory at the University of Florida. We also thank Javier Francisco Ortega and Eugenio Santiago for organizing the Symposium "Evolucion en las Antillas: la Perspectiva Molecular" at the IX Latin American Botanical Congress in Santo Domingo, Dominican Republic in June 2006. where this research was originally presented. A grant from the International Association of Plant Taxonomy (IAPT) supported, in part, the research of the first author. The National Science Foundation funded portions of this research through grants WJ (DEB0515636) and FAM (DEB-0515665).
Adams. C. D. 1972. Flowering plants of Jamaica. Robert MacLehose & Co., University Press, Glasgow.
Becquer-Granados E. R. 2007. Tetrazygia decorticans (Miconieae, Melastomataceae), a new species from Cuba. Wildenowia 37:313-317.
Borhidi, A. 1983. New names and new species in the flora of Cuba and Antilles. III. Acta Bot. Hung. 29: 181-215.
--. 1985. Phytogeographic survey of Cuba. 1. The phytogeographic characteristics and evolution of the flora of Cuba. Acta Bot. Hung. 31 : 3-34.
Clausing, G., K. Meyer & S. S. Rennet. 2000. Correlations among fruit traits and evolution of different fruit within the Melastomataceae. Bot. J. Linn. Soc. 133: 303-326.
Clausing, G., Renner S.S. 2001. Molecular phylogenetics of Melastomataceae and memccylaceae: Implications for character evolution. Amer. J. Bot. 88: 486-498.
Cogniaux, A. 1891. Melastomataceae. In: A. & C. de Candolle, eds. Monographiae phanerogamarum 7. Masson, Paris.
Doyle, J. J. & J. L. Doyle. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem. Bull. 19: 11-15.
Gleason, H. A. 1932. A synopsis of the Melastomataceae of British Guiana. Brittonia 1:127-184
Goloboff, P. 1993. Nona. Software and Documentation by the Author, Tucuman, Argentina.
Grisebach, A. H. R. 1866. Catalogus plantarum cubensium. Wihelm Engelmann, Leipzig.
Holmgren, P. K. & N. H. Holmgren. 1998 Index Herbariorum. New York Botanical Garden. http:// sciweb.nybg.org/science2/IndexHerbariorum.asp. (continuously updated).
Howard, R. A. 1989. Flora of the Lesser Antilles. Vol. 5. Dicotyledoneae. Part. 2. Arnold Arboretum, Harvard University, Jamaica Plain, Massachusetts.
Judd, W. S. 1986. Taxonomic studies in the Miconieae (Melastomataceae). I. Variation in inflorescence position. Brittonia 38: 150-161.
--. 1989. Taxonomic studies in the Miconieae (Melastomataceae). III. Cladistic analysis of axillary-flowered taxa. Ann. Missouri Bot. Gard. 76: 476-495.
--& J. D. Skean. 1991. Taxonomic studies in the Miconieae (Melastomataceae). IV. Generic realignments among terminal-flowered taxa. Bull. Fl. Mus. Nat. Hist., Biol. Sci. 36:25-84.
--, W. S., C. S. Campbell, E. A. Kellogg, P. E Stevens, & M. J. Donoghue. 2002. Plant Systematics: A Phylogenetic Approach, ed. 2. Sinauer Associates, Sunderland, Massachusetts.
Liogier, A. H. I= Hno. Alain]. 1957. Melastomataceae. In Flora de Cuba, vol. 4, by Hno. Leon and Hno. Alain. Contr Ocas Mus Nat Colegio "De La Salle" 10: 1-556.
Liogier, A. H. 2000. Melastomataceae. In Flora de la Espanola. IX. Jardin Botanico Nacional Dr. Rafael Ma. Moscoso, Editora Corripio, C. por A., Santo Dopmingo, Dominican Republic.
Macbride, J. F. 1941. Flora of Peru (Melastomataceae). Fieldiana, Bot. 13:249 521.
Michelangeli, 17. A. 2000. A cladistic analysis of the genus Tococa (Melastomataceae) based on morphological data. Syst. Bot. 25: 211-234.
--, D. S. Penneys, J. Giza, D. Soltis, M. H. Hils & J. D. Skean, Jr. 2004. A preliminary phylogeny of the tribe Miconieae (Melastomataceae) based on nrITS sequence data and its implications on inflorescence position. Taxon 53: 279-290.
Morales, M. E. & F. Gonzalez. 2005. Redescubrimiento de Chalyhea Nandin e implicaciones en la delimitacion generica de Huilaea Wurdack (Melastomataceae). Rev. Acad. Colomb. Cienc. 29 (111):171-178
Naudin, C. 1851. Melastomacearum. Ann. Sci. Nat. Bot. et Bio. Veg. ser. 3, 16: 83-246.
Nixon, K. C. 1999-2002. WinClada. Soflware and Documentation by the author, Ithaca, NY
Rambout, A. 1996. Se-Al: Sequence Alignment editor, v2.0a11. University of Oxford, Oxford, UK. http://evolve.zoo.ox.ac.uk/.
Rennet, S. S. 1993. Phylogeny and classification of Melastomataceae and Memecylaceae. Nord. J. Bot. 13: 519-540.
Richard, A. 1845. Fanerogamia o plantas vasculares. In Sagra, R. de la, Historia fisica, politica y natural de la Isla de Cuba. Parte 2. Historia Natural, Vol: 10. Paris.
Shaw, J., E. B. Liekey, J. T. Beck, S. B. Farmer, W. Liu, J. Miller, K. C. Siripun, C. T. Winder, E. E. Schilling & R. L. Small. 2005. The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis. Amer. J. Bot. 92:142-166.
Sun, Y., D. Z. Skinner, G. H. Liang & S. H. Hulbert. 1994. Phylogenetic analysis of Sorghum and related taxa using internal transcribed spacer of nuclear ribosomal DNA. Theor. Appl. Genet. 89: 26-32.
Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F., & Higgins, D.G. 1997. The CLUSTAL X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25: 4876-4882.
Urban, I. 1924. Sertum antillanum, xix. Repert. Spec. Nov. Regni Veg. 19: 298-308.
--. 1926. Sertum antillanum, xxvi. Repert. Spec. Nov. Regni Veg. 22:222-241.
--. 1927. Plantae haitienses novae vel rariores IV. a cl. E. L. Ekman 1924-26 lectae. Ark. Bot. 21A(5)
Wurdack, J. J. 1972. Certamen Melastomataceis XVIII. Phytologia 22: 399-418.
--. 1973. Melastomataceae. In T. Lasser (ed.). Flora de Venezuela. No. 8. Instituto Botanico, Ministerio de Agricultura y Cria, Caracas.
--. 1980. Melastomataceae. In G. Harling & B. Sparre, eds. Flora of Ecuador. Ni. 13. Univ. Goteborg & Riksmuseum, Stockholm.
--. 1988. Certamen Melastomataceis XXXVIII. Phytologia 64:293-301.
Published online: 15 April 2008
Eldis R. Becquer-Granados (1,4). Kurt M. Neubig (2) * Walter S. Judd (2) * Fabian A. Michelangeli (3) * J. Richard Abbott (2) * Darin S. Penneys (2)
(1) Jardin Botanico Nacional, Universidad de La Habana, Havana, Cuba
(2) Department of Botany, University of Florida, Gainesville, FL 32611-8526, USA; e-mail; firstname.lastname@example.org, email@example.com, firstname.lastname@example.org
(3) The New York Botanical Garden, 200th St. & Southern Blvd., Bronx, NY 10458-5126, USA; e-mail: email@example.com
(4) Author for Correspondence; e-mail: firstname.lastname@example.org
|Printer friendly Cite/link Email Feedback|
|Author:||Becquer-Granados, Eldis R.; Neubig, Kurt M.; Judd, Walter S.; Michelangeli, Fabian A.; Abbott, J. Ri|
|Publication:||The Botanical Review|
|Date:||May 1, 2008|
|Previous Article:||Floristic richness and affinities in the West Indies.|
|Next Article:||Multiple events of dispersal and radiation of the tribe Miconieae (Melastomataceae) in the Caribbean.|
|Plant evolution and biodiversity in the Caribbean Islands--perspectives from molecular markers.|
|Multiple events of dispersal and radiation of the tribe Miconieae (Melastomataceae) in the Caribbean.|