| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
Institute of Systematic Botany, University of Zurich, Zollikerstr. 107, CH-8008 Zurich, Switzerland
Received for publication May 30, 2002. Accepted for publication September 24, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Alzateaceae • chloroplast DNA • floral development • molecular phylogeny • Oliniaceae • Penaeaceae • perianth homology • Rhynchocalycaceae
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Recent molecular studies recognize 14 families within Myrtales: Alzateaceae, Combretaceae, Crypteroniaceae, Heteropyxidaceae, Lythraceae, Melastomataceae, Memecylaceae, Myrtaceae, Oliniaceae, Onagraceae, Penaeaceae, Psiloxylaceae, Rhynchocalycaceae, and Vochysiaceae (e.g., Conti, Litt, and Sytsma, 1996
; APG, 1998
; Savolainen et al., 2000
). Several studies, both nonmolecular (Johnson and Briggs, 1984
) and molecular (Conti et al., 1997
; Clausing and Renner, 2001
), investigated the relationships among myrtalean families. Phylogenetic studies of rbcL sequences (Conti, Litt, and Sytsma, 1996
; Conti et al., 1997
) confirmed a clade, earlier defined from nonmolecular data (Johnson and Briggs, 1984
), that comprises four myrtalean taxa with a Western Gondwanan distribution: the South African Penaeaceae and Rhynchocalycaceae, the South and East African Oliniaceae, and the Central and South American Alzateaceae. This Western Gondwanan clade is strongly supported as sister to a fifth taxon, the Southeast Asian Crypteroniaceae (Clausing and Renner, 2001
; Conti et al., 2002
). These five taxa are characterized by flowers with an ephemeral anther endothecium (Tobe and Raven, 1983a
, b
, 1984a
, b
, c
, d
, 1987a
, b
; Johnson and Briggs, 1984
). The present study focuses on the four Western Gondwanan taxa: Penaeaceae, Oliniaceae, Rhynchocalycaceae, and Alzateaceae.
The Penaeaceae are a clearly circumscribed taxon that comprises 23 species in seven genera endemic to the southern and southwestern parts of the Cape Province of South Africa. They are shrubs and undershrubs confined to the "fynbos" vegetation typical of the Cape Floristic Region. In the past, Penaeaceae have either been placed in Myrtales or close to Thymelaeaceae (e.g., Gilg, 1894b
; Supprian, 1894
), the latter sometimes also included in Myrtales (e.g., Cronquist, 1981
). More recent authors agree on the placement of Penaeaceae in Myrtales and the exclusion of Thymelaeaceae from this order (e.g., Dahlgren and Thorne, 1984
; Johnson and Briggs, 1984
; Thorne, 1992
; Takhtajan, 1997
). The family is relatively well described through studies on pollen morphology (Patel, Skvarla, and Raven, 1984
), vegetative anatomy (Carlquist and Debuhr, 1977
; van Vliet and Baas, 1984
; Dickie and Gasson, 1999
), embryology (Tobe and Raven, 1984c
), and a series of mainly taxonomical papers by Dahlgren (Dahlgren, 1967a
, b
, c
, 1968
, 1971
) and Dahlgren and van Wyk (1988)
. The flowers of Penaeaceae have only a single whorl of perianth organs that alternate with the stamens. The perianth has variously been interpreted as being apetalous (e.g., Cronquist, 1981
; Dahlgren and Thorne, 1984
; Johnson and Briggs, 1984
; Dahlgren and van Wyk, 1988
), i.e., the flowers as being obhaplostemonous (number of stamens equals number of sepals and stamens are in alternisepalous position) or as consisting of petals only (Rao and Dahlgren, 1968
), i.e., the flowers as being haplostemonous (number of stamens equals number or petals and stamens are in alternipetalous position). Other authors did not specify the nature of the perianth (e.g., Dahlgren, 1967a
, b
, c
, 1968
, 1971
; Rourke, 1995
).
The Oliniaceae comprise about eight species confined to the single genus Olinia. They are trees and shrubs found in the montane and coastal forests of eastern and southern Africa, and also on the island of St. Helena, where the genus was most probably introduced from South Africa (Cufodontis, 1960
). Most authors of traditional classifications associated Oliniaceae with myrtalean families (Cronquist, 1981
; Dahlgren and Thorne, 1984
; Thorne, 1992
; Takhtajan, 1997
). A sound taxonomic treatment of the family is not yet available (Sebola and Balkwill, 1999
). The homology of the three perianth whorls of Oliniaceae has been much debated. From the outside of the flower inwards, the following three alternating whorls have been identified: (1) a whorl of truncate, minutely tooth-like structures, (2) a whorl of conspicuous, petal-like organs, and (3) a whorl of small, scale-like organs. The stamens are positioned in the same radii as these latter scale-like organs. Two hypotheses have been proposed for the homology of these three whorls: the first interprets the flowers as being haplostemonous, i.e., the outermost organs are interpreted as a reduced calyx, the middle whorl as a corolla, and the innermost organs as either scales of unspecified nature (e.g., Verdcourt, 1975
), as staminodes (e.g., Cronquist, 1981
), or as stipules of the petals (e.g., Rao and Dahlgren, 1969
; Sebola and Balkwill, 1999
). The second hypothesis describes the flowers as being obhaplostemonous, i.e., the organs of the outermost whorl are interpreted as "teeth" of unspecified nature, the middle whorl as calyx, and the innermost whorl as corolla (e.g., Gilg, 1894a
; Dahlgren and van Wyk, 1988
).
The Rhynchocalycaceae comprise the single species Rhynchocalyx lawsonioides, a rare, evergreen tree endemic to the Eastern Cape and KwaZulu-Natal in South Africa, that grows mainly on forest margins, often near watercourses. The genus was first described in Lythraceae (Oliver, 1894
), but later was included in an expanded Crypteroniaceae (van Beusekom-Osinga and van Beusekom, 1975
). However, the distinctive embryology of Rhynchocalyx prompted its elevation to family rank (Johnson and Briggs, 1984
; Tobe and Raven, 1984d
). The flowers of R. lawsonioides are clearly obhaplostemonous, i.e., the stamens are in alternisepalous position.
The Alzateaceae comprise the single species Alzatea verticillata, represented by trees restricted to submontane tropical forests of Bolivia, Peru, Panama, and Costa Rica. Alzatea verticillata was earlier referred to Lythraceae (e.g., Lourteig, 1965
) and was later included in the broadly circumscribed Crypteroniaceae (van Beusekom-Osinga and van Beusekom, 1975
). Based on wood anatomical (van Vliet and Baas, 1984
) and embryological features (Tobe and Raven, 1984a
), the genus was then raised to family rank (Graham, 1984
). The flowers of A. verticillata are generally described as apetalous (e.g., Graham, 1984
; Johnson and Briggs, 1984
; but see van Beusekom-Osinga and van Beusekom, 1975
).
To date, no molecular phylogenetic study based on an intensive taxon sampling of this Western Gondwanan clade of Myrtales has been published and neither has the floral morphology been comparatively investigated in a phylogenetic context. In addition, intrafamilial relationships in both Penaeaceae and Oliniaceae have never been addressed using molecular data. Therefore, several unresolved phylogenetic and morphological questions remain to be answered. For example: What are the detailed phylogenetic relationships among and within Penaeaceae, Oliniaceae, Rhynchocalycaceae, and Alzateaceae? Are the traditionally recognized genera of Penaeaceae monophyletic? What are the perianth organs of Oliniaceae and Penaeaceae homologous to? Here we present the results of phylogenetic analyses performed on DNA sequences from six chloroplast markers for most species of this Western Gondwanan clade of Myrtales. In addition, floral development and floral structure in Penaeaceae, Oliniaceae, and Rhynchocalycaceae are described and compared. We analyze the results of our comparative morphological study in the context of the molecular phylogeny to address questions of perianth organ homology in Penaeaceae and Oliniaceae. Further, we attempt to elucidate the evolution of a number of floral features in Penaeaceae, some of which relate to pollination syndromes and others that are important for intrafamilial classification. The resulting paper provides an example of how the integration of molecular phylogenetic and comparative developmental/morphological approaches can contribute to the problem of organ homology assessment.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
, for a total of 30 sampled taxa. As outgroup taxon we used Crypteronia paniculata from the Crypteroniaceae, which unequivocally have been shown to be sister to the clade comprising the four ingroup families (Conti, Baum, and Sytsma, 1999; Clausing and Renner, 2001
; Conti et al., 2002
). Taxon names, voucher information, and GenBank accession numbers are archived at http://ajbsupp.botany.org/v90. We extracted total genomic DNA from silica dried leaf material. Leaf tissue was homogenized using glass beads and a Retsch MM 2000 shaker (Retsch, Haan, Germany). For the subsequent DNA extraction, we applied the DNeasy Plant Mini Kit (Qiagen, Basel, Switzerland) to all species except Brachysiphon rupestris, Endonema lateriflora, and Stylapterus fruticulosus, which were extracted with the sodium dodecyl sulphate (SDS) extraction protocol as described by Eichenberger, Gugerli, and Schneller (2000)
.
Polymerase chain reaction and DNA sequencing
Six chloroplast regions, the rps16 intron, rpl16 intron, the trnS-G intergenic spacer, the atpB-rbcL intergenic spacer, the psbA-trnH intergenic spacer, and part of the matK exon, were amplified for all taxa. Primers, corresponding annealing temperatures, and literature sources for the primers are presented in Table 1. All polymerase chain reactions (PCR) were performed on a Biometra TGradient thermocycler (Biometra, Göttingen, Germany), applying a thermal cycling program consisting of 34 cycles of 0.5 min at 95°C, 1 min at the primer-specific annealing temperatures given in Table 1, and 1.7 min at 72°C, followed by a terminal extension of 10 min at 72°C. In order to successfully detect amplified DNA target regions and possible contamination, the PCR products were examined on agarose gels with a negative control. Successfully amplified PCR products were purified with the QIAquick PCR Purification Kit (Qiagen).
|
) was used to edit and assemble complementary strands. Base positions were individually checked for agreement between the complementary strands. We aligned the sequences visually in Sequencher 3.1.1. Variable positions were double-checked to ensure that base calls were consistent at informative positions. To ensure that unexpected results, such as the paraphyly of Brachysiphon, were not the product of any mistake during laboratory work, we re-extracted and resequenced all five species of Brachysiphon. For Brachysiphon fucatus and Stylapterus micranthus, second accessions were extracted, sequenced, and compared to the first accessions.
Indel coding and inversions
We coded insertion or deletion events (indels, listed in Table 2) as additional, binary characters applying the "simple indel coding method" described by Simmons and Ochoterena (2000)
. Only potentially informative indels bordered by unambiguously aligned nucleotide stretches were coded. We omitted single-nucleotide indels resulting from the differing lengths of strings of the same nucleotide (e.g., four A's vs. five A's), as this type of indel may be the product of experimental error (Downie et al., 1998
; McDade and Moody, 1999
) or evolutionary lability (Small et al., 1998
).
|
).
Phylogenetic analyses
Cladistic analyses were performed using the Maximum Parsimony (MP) optimization of the software package PAUP* (Phylogenetic Analyses Using Parsimony) 4.0b8 (Swofford, 2000
). We included only informative nucleotide and indel characters in the analyses, and all characters and character state transitions were weighted equally (Fitch, 1971
). The individual data matrices for each of the six DNA regions plus a combined data matrix of the six regions were analyzed by employing an heuristic search strategy using 1000 replicates with random taxon-addition, tree-bisection-reconnection (TBR) branch swapping, and keeping all most-parsimonious trees. Relative support for different clades was estimated with the bootstrap (Felsenstein, 1985
) option implemented in PAUP* employing a full heuristic search with 1000 replicates, simple addition sequence, and TBR branch swapping. To assess congruence among pairwise combinations of the six data sets we carried out 1000 replicates of the partition homogeneity test (Farris et al., 1995
) as implemented in PAUP*, employing a full heuristic search with simple taxon-addition sequence, tree-bisection-reconnection (TBR) branch swapping, and saving all most-parsimonious trees. In addition, a homogeneity test with the same settings was conducted to assess congruence between the indel characters and the substitution characters.
Morphological/anatomical/histological methods
Out of the currently eight recognized species of Oliniaceae we studied floral development and morphology of two species (Olinia emarginata and O. ventosa). For Penaeaceae we collected floral material for all genera except for the monotypic genus Glischrocolla. We studied floral development and morphology of several species including Brachysiphon acutus, Endonema retzioides, Penaea dahlgrenii, Saltera sarcocolla, and Stylapterus fruticulosus. Both in Penaeaceae and Oliniaceae, respectively, our studies revealed that floral development and basic floral organization are very similar within each family. In the results section of the present study floral development and morphology of one representative of Penaeaceae (Endonema retzioides) and Oliniaceae (Olinia emarginata) are described. However, the conclusions on organ homology drawn from the study of these two representative species are valid for the entire families. Rhynchocalycaceae is represented by its only member Rhynchocalyx lawsonioides. For scanning electron microscopy (SEM) and microtome sectioning, living material was fixed in formalin-acetic acid-alcohol (FAA) and subsequently stored in 70% ethanol. For SEM investigations, the material was dehydrated in ethanol and acetone, critical point dried, and sputter coated with gold. For microtome sectioning, specimens were embedded in 2-hydroxyethyl methacrylate (Kulzer's Technovit 7100; Heraeus Kulzer, Wehrheim, Germany), cut using a Microm HM rotary microtome 355 (Microm, Walldorf, Germany) at 5 µm, and stained with toluidine blue and ruthenium red (for detailed description of embedding and staining method, see Igersheim [1993]
and Igersheim and Cichocki [1996]
). Permanent slides are deposited at the Institute of Systematic Botany of the University of Zurich (Z).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
|
The monophyly of Olinia and Penaeaceae is well supported by 100% BS and 19 and 14 indel characters, respectively. Within Olinia, two well-supported clades were found. One consists of Olinia emarginata and O. vanguerioides (BS = 81%) and the other comprises the three remaining species (BS = 97%; Fig. 1). Each of the clades is supported by two indel characters (C8, D4, and C7, E9, respectively; Table 2).
Within Penaeaceae, several well-supported clades are defined. The strict consensus tree (Fig. 1) supports the sister-group relationship of Endonema with the rest of the family (BS = 87%). The two species of the genus Endonema form a clade with 100% BS. In turn, the clade of Glischrocolla formosa and Brachysiphon rupestris (BS = 91%) is sister to the clade formed by Penaea, Stylapterus, Brachysiphon pro parte (p.p.), Saltera, and Sonderothamnus (BS = 93%), a relationship further supported by a deletion (E5; Table 2) of 226 bp in the intergenic spacer between psbA and trnH. This topology contradicts the strict consensus of the rpl16 intron trees, supporting (BS = 57%) a sister-group relationship between a clade with Sonderothamnus, Saltera, and Brachysiphon mundii and the rest of Penaeaceae. In the strict consensus tree of the combined analysis, Sonderothamnus, Saltera, and Brachysiphon mundii form a strongly supported clade (BS = 100% and an insertion in the rpl16 intron), which, in turn, forms part of a strong (BS = 95%) trichotomy with Stylapterus micranthus and Brachysiphon acutus and a third clade comprising the remaining species of Penaeaceae. This latter clade receives 99% BS and shares an 8-bp deletion in the rpl16 intron (B4; Table 2). Other well-supported groups are confined to a clade comprising all Penaea species except Penaea dahlgrenii (BS = 98% in the combined analysis). The members of this Penaea clade also share a 16-bp indel in the trnS-G spacer (C11; Table 2). Within the Penaea clade, four of the five subspecies of Penaea cneorum, together with Penaea mucronata, form a solid clade (BS = 97%) that is part of a trichotomy with the fifth subspecies of Penaea cneorum and Penaea acutifolia.
Floral development and structure
Floral development and structure are described with emphasis on characters that help to clarify floral organization and organ homology among Rhynchocalycaceae, Oliniaceae, and Penaeaceae.
Rhynchocalyx lawsonioides (Rhynchocalycaceae)
Flowers are bisexual; penta- to heptamerous, except for the two-merous gynoecium; obhaplostemonous; and the ovary is superior (Figs. 3, 4). Sepals are inserted on the rim of a short hypanthium (Figs. 3, 4). They are broad-based and triangular, and their aestivation is valvate (Figs. 3, 12, 19). Sepals are 8–11 cell layers thick, and each sepal is served by three vascular bundles (Fig. 18). Petals alternate with the sepals and are inserted on the adaxial side of the hypanthium rim (Figs. 4, 5). Their development is retarded relative to the sepals in young buds (Figs. 4, 5). In early developmental stages, petals are incurved, club-shaped, with a narrow base and a slightly thickened apex (Fig. 5). In older stages, petals have a linear claw and an expanded, circular blade with an irregularly shaped outline (Fig. 13). Petal blades are 3–5 cell layers thick. Each petal is served by a single vascular bundle (Fig. 19). Stamens are produced opposite, i.e., just below the petals on the adaxial side of the hypanthium rim (Figs. 4, 5). In bud, stamens are wrapped by the petal blades (Fig. 13), which at anthesis spread to expose the stamens (not shown). Stamens are strongly incurved in bud, with their pollen sacs directed towards the hypanthium (Figs. 5, 19). The apex of the ovary is tapering into a short style with a terminal, papillate stigma (Fig. 13). Scattered cells with oxalate druses are present in all floral organs except for the petals. In the calyx they are restricted to the basal part of the sepals and are concentrated in the adaxial parenchymatic tissue (Fig. 20).
|
|
|
Endonema retzioides (Penaeaceae)
Flowers are bisexual and tetramerous, and the ovary is superior. The perianth consists of a single whorl of organs, which are inserted on the rim of a long, tubular hypanthium (Fig. 16). Perianth organs have a broad base and are triangular in shape in young stages (Fig. 9). Aestivation is apert in early developmental stages (Fig. 9) and valvate in later stages (Fig. 25). The perianth organs become exceedingly thick and fleshy during floral development (Fig. 24), and at anthesis they are numerous cell layers thick. Each perianth lobe is supplied by three main vascular bundles (Fig. 24). In early developmental stages, floral buds show strikingly large, triangular, "empty" areas between the adjacent flanks of two neighboring perianth lobes and the corresponding stamen primordium, which is located just below on the adaxial side of the hypanthium rim (Figs. 10, 17). Stamens are strongly incurved in bud with their pollen sacs directed towards the hypanthium (Figs. 11, 17, 25). At anthesis, the stamens are upright and exserted from the floral tube (Fig. 16). The carpels alternate with the stamens (Fig. 11). Already at relatively young developmental stages, the commissural regions (i.e., the regions where neighboring carpels are fused) extend beyond the carpel tips, giving the false impression of the carpels being positioned in the same sectors as the stamens (Fig. 17). At anthesis, the style ends in a quadrangular, capitate stigma, which consists of four commissural lobes. Scattered cells with oxalate druses are present in all floral organs except for the uppermost part of the style. In the perianth organs they are restricted to the basal part and are concentrated in the adaxial parenchymatic tissue (Fig. 26).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Conti et al., 2002
), with Alzateaceae sister to the three African taxa and Rhynchocalycaceae sister to the clade of Penaeaceae and Oliniaceae. This African clade was also strongly supported by the analyses of Clausing and Renner (2001)
and Conti et al. (1997)
. The strong molecular support for the clade formed by the three African families, excluding the New World Alzateaceae, contrasts with results based on nonmolecular characters (Johnson and Briggs, 1984
), which support sister relationships between Rhynchocalycaceae and Alzateaceae and between Oliniaceae and Penaeaceae, respectively. The data matrix by Johnson and Briggs (1984)
did not contain any exclusive synapomorphies for the three African families, but it included two derived features shared by Rhynchocalycaceae and Alzateaceae, i.e., libriform fibers (walls with simple pits) in the wood and a reduced carpel number. However, libriform fibers are also present in Oliniaceae (van Vliet and Baas, 1984
), and reduced carpel number (relative to the number of perianth organs and stamens) is also present in some members of Olinia (Rao and Dahlgren, 1969
). The occurrence of a glandular swelling or mucro (areola) at the leaf tip in Rhynchocalycaceae, Oliniaceae, and some species of Penaeaceae (Saltera, Sonderothamnus) has been proposed as a possible synapomorphy for the three African families (Dahlgren and van Wyk, 1988
).
Our cpDNA results corroborate the sister-group relationship between Oliniaceae and Penaeaceae that was already supported in previous molecular (Conti et al., 1997
, 2002
; Clausing and Renner, 2001
) and nonmolecular studies (Johnson and Briggs, 1984
), but only when a 15-bp deletion in the rpl16 intron (B1; Table 2) is added to the combined data set (BS = 58%; see Fig. 2). Earlier molecular studies also found only low or moderate bootstrap support for the Oliniaceae/Penaeaceae clade (Conti et al., 1997
, BS < 50%; Clausing and Renner, 2001
, BS = 68%/74%). The only synapomorphy supporting the clade with Penaeaceae and Oliniaceae in the analysis by Johnson and Briggs (1984)
was what they called "reduction of stamen filaments." However, filaments, although small in Oliniaceae, are well developed in both families, and in many species of Penaeaceae they are even massive and long (e.g., in Endonema, Glischrocolla, Saltera; see Fig. 16 for Endonema retzioides; see also Dahlgren and van Wyk [1988]
). Patel, Skvarla, and Raven (1984)
noted distinct similarities in pollen structure of Oliniaceae and Penaeaceae, including a psilate surface and a particular exine structure with an exceedingly thick foot layer and tectum, and a thin columella layer with an infratectal granular layer.
The three African families are each very distinctive, both morphologically and molecularly (BS = 100% for both Oliniaceae and Penaeaceae; see Fig. 1), but their detailed relationships are weakly resolved. The long branches subtending the crown groups of Oliniaceae and Penaeaceae may, in part, explain the apparent contradiction between distinctness and resolution (see Fig. 27, phylogram). A possible explanation for the observed lack of resolution is that the stem lineages of Oliniaceae, and especially Penaeaceae, represent groups of relatively ancient origin that accumulated many apomorphies during their independent evolutionary trajectories. The radiation that gave rise to the ancestors of Rhynchocalycaceae, Oliniaceae, and Penaeaceae has probably been obscured by the changes that have later accumulated over a long period of time, both at the molecular and nonmolecular level, thus making it difficult to establish phylogenetic relationships among the three families. After a relatively long period of independent evolution, Oliniaceae and Penaeaceae appear to have radiated rather rapidly, a process that left its signature in the short branches of the crown groups of these two families. A similar evolutionary scenario has been proposed for the radiation of the South African Cape genus Phylica (Rhamnaceae; Richardson et al., 2001
). In the future, we plan to use molecular dating approaches to further investigate the implications of the observed pattern for the rate of molecular evolution and time of radiation of the groups studied here.
|
Phylogenetic relationships within Oliniaceae
The short branches (Fig. 27) and polytomies (Fig. 1) within Oliniaceae mirror the confusing taxonomy of this family (Sebola and Balkwill, 1999
), where a variable number of species, from as few as one (e.g., Sonder, 1862
) to seven or eight have been recognized (e.g., Dahlgren and van Wyk, 1988
; Sebola and Balkwill, 1999
). Geographically, most species are restricted to South Africa, while one species (Olinia vanguerioides) occurs in Zimbabwe and Mozambique and one, O. rochetiana, in northeastern South Africa, extending north into tropical East Africa. Verdcourt (1978)
states that O. rochetiana is closely related to O. vanguerioides, differing in little but flower size. Among the five species sampled, our results strongly support a clade with O. emarginata and O. vanguerioides (BS = 81%) and a clade including O. capensis, O. ventosa, and O. radiata (BS = 97%). In addition, each clade is supported by two indels (Fig. 1, Table 2). Geographically, the latter three species are restricted to coastal forests in the western and/or southern parts of South Africa, whereas O. emarginata extends from the southeast to the northeast of South Africa and O. vanguerioides occurs even farther north in Zimbabwe and Mozambique (van Wyk and van Wyk, 1997
; Sebola and Balkwill, 1999
). Morphologically, species delimitation is problematic within Olinia, and a careful monographic treatment and morphological/anatomical study of the genus are greatly needed. For this reason it is currently difficult to determine morphological synapomorphies that are congruent with the clades retrieved in our molecular study. Stomata types may represent a potentially useful synapomorphy, as they are reported to be mostly anomocytic on the leaves of O. emarginata and O. vanguerioides and mostly paracytic in O. radiata and O. ventosa (Mújica and Cutler, 1974
).
Phylogenetic relationships within Penaeaceae
Phylogenetic relationships among species of Penaeaceae have not been investigated from a molecular standpoint or with a rigorous analytical approach. Delimitation and relationships of genera within Penaeaceae have been controversial in the past. Morphologically, the family is rather uniform and differences pertain mainly to the relative size and shape of the vegetative and reproductive organs, whereas distinct, qualitative differences are rare. Already in 1894, Supprian noted that due to the close relationships among all Penaeaceae, and the occurrence of taxa with transitional morphological features, generic delimitation was difficult. These difficulties are also reflected in earlier taxonomic treatments of the family, for example de Candolle (1857)
recognized six genera, which Baillon (1877)
reduced to only three and Gilg (1894b)
extended again to five. The currently accepted generic circumscriptions and systematic relationships among the genera are mainly based on taxonomic revisions by Dahlgren (Dahlgren, 1967a
, b
, c
, 1968
, 1971
). He recognized seven genera and 21 species, although he also found it challenging to clearly circumscribe genera, especially due to the occurrence of transitional species, for example, Stylapterus ericifolius between Stylapterus and Brachysiphon and Brachysiphon mundii between Brachysiphon and Saltera/Sonderothamnus (Dahlgren, 1967a
). Since Dahlgren's work, two new species have been described, Penaea dahlgrenii (Rourke and McDonald, 1989
) and Brachysiphon microphyllus (Rourke, 1995
). Rourke (1995)
stated that these new species are not easily assigned to any particular genus, further emphasizing the lack of clear-cut boundaries between the genera. A detailed comparative study of leaf anatomy of Penaeaceae (Dickie and Gasson, 1999
) concluded that, due to variability and inconsistency, leaf anatomical characters are not useful for systematic purposes within the family, a conclusion that corroborates Dahlgren's view (Dahlgren, 1968
). Furthermore, leaf anatomical characters, when mapped on our molecular tree, revealed no congruence with molecular clades (not shown). Similarly, wood anatomical features track more closely habitat type (e.g., xeromorphic vs. mesomorphic adaptations) than taxonomic groupings (Carlquist and Debuhr, 1977
).
The difficult interpretation of generic circumscription is also reflected in the short branches found within the Penaeaceae crown group (Fig. 27). Despite the relatively low number of informative characters within Penaeaceae, our analysis of the combined data sets strongly supports several clades (marked I–VI in Fig. 1) and suggest that some of Dahlgren's genera, specifically Brachysiphon and Stylapterus (Dahlgren, 1967a
), are not monophyletic (see Fig. 1).
Clade I (BS = 100%, Fig. 1; branch length [BL] = 17, Fig. 27) comprises the two species of the genus Endonema and is sister to the rest of the Penaeaceae. This result strongly confirms the proposed monophyly of the genus (Dahlgren, 1967c
). Endonema lateriflora and E. retzioides share large, tubular, and brightly colored flowers typical of a bird-pollination syndrome (Dahlgren, 1967c
). Both species have strongly incurved stamens in bud, a feature regarded as a synapomorphy for Endonema (Dahlgren, 1967c
), but this feature is also present in Brachysiphon acutus and B. mundii (Fig. 28; J. Schönenberger, personal observation). Strongly incurved stamens also characterize Oliniaceae, Rhynchocalycaceae, and Alzateaceae; hence they appear to represent the ancestral condition in Penaeaceae. Weberling (1988)
identified a potential synapomorphy for Endonema, i.e., the specialized inflorescence with two or more flowers situated in the leaf axil on a young branch that continues vegetative growth, while inflorescences are terminal in the rest of Penaeaceae. Both species are restricted to a relatively small area in the Riversonderend Mountains in the Cape Province and occur in similar habitats on south-facing slopes, often near a watercourse (Dahlgren, 1967c
).
|
). Glischrocolla differs from Endonema mainly in having upright stamens in bud, five- vs. three-colporate pollen, and terminal vs. lateral inflorescences. Brachysiphon sensu Dahlgren (1968)
is a taxonomically difficult genus that is not easily circumscribed. Dahlgren (1968)
stated that B. acutus and B. fucatus are in some respects similar to members of the genus Stylapterus, while B. rupestris and B. mundii in some features resemble Saltera and Sonderothamnus. Using Dahlgren's criteria, B. microphyllus, which was recently described (Rourke, 1995
), could be placed either in Brachysiphon or Sonderothamnus and only the occurrence of bracts with entire margins favored its assignment to the former. The occurrence of two erect and two pendant ovules in the middle part (axially) of each locule, a consistent feature of Endonema and Glischrocolla (Dahlgren, 1967b
, c
), appears occasionally in B. rupestris (Phillips, 1926
, 1951
) and might provide a morphological link to G. formosa. Both G. formosa and B. rupestris grow on ledges or in crevices on sandstone rock precipices, the former at rather high (1200–1500 m) and the latter at lower altitudes (30–400 m). At present, the two species are only known from a few populations and are narrowly distributed in close, but not overlapping, areas of the Western Cape Province. Clades III, IV, and V form a strongly supported monophyletic group (BS = 93%, BL = 4, 226-bp deletion in the psbA-trnH spacer). Further, all members of this group have only erect ovules (two per locule, except for Saltera sarcocolla, which has four) inserted basally in the locule (Fig. 28). Relationships among the three clades are not resolved.
Clade III (BS = 100%, BL = 10, insertion in the rpl16 intron) comprises Sonderothamnus (two species), Saltera (monospecific), and Brachysiphon mundii. Dahlgren (1968)
listed a number of similarities between Sonderothamnus and Saltera, including specialized leaf tips (with areolae), which are always present in Sonderothamnus and at least in some individuals of Saltera. Denticulate-fimbriate bract and bracteole margins represent another feature shared by both genera. From a morphological point of view, B. mundii does not share any obvious synapomorphies with Saltera and Sonderothamnus, but we have been unable to detect any traits that clearly distinguish it from these genera and would thus contradict the close relationship among the taxa of clade III. Before Dahlgren's treatment of Penaeaceae (Dahlgren, 1968
), the two species of Sonderothamnus were included in a broadly circumscribed Brachysiphon (e.g., Gilg, 1894b
; Phillips, 1951
; Barker, 1963
). Dahlgren knew B. mundii only from a single specimen in post-floral stage (the type specimen) and he remarked that the collection was rather incomplete. Therefore, it is possible that a more detailed study of anthetic flowers may reveal characters linking B. mundii with Saltera and Sonderothamnus. The two species of Sonderothamnus and B. mundii have narrow distributions, while Saltera sarcocolla is relatively widespread in the southwestern part of the Cape Province. The distribution of Saltera overlaps with that of Sonderothamnus, while B. mundii has a more eastern distribution and is the only species within Penaeaceae that is restricted to limestone rocks (Dahlgren, 1968
; Goldblatt and Manning, 2000
).
Clade IV (BS = 95%, BL = 4) comprises Stylapterus micranthus and Brachysiphon acutus, a pair of species that had never been suggested to be closely related. To rule out the possibility that this unexpected result may be the product of experimental error, we double-checked our sequencing results (as described in the Materials and Methods section) and confirmed our species identifications (see http://ajbsupp.botany.org/v90/). As the floral structure of these two species does not suggest a sister relationship, it is necessary to propose an alternative evolutionary process that may explain this unexpected pattern. In this regard, it is important to note that our phylogenetic reconstruction is derived exclusively from chloroplast DNA sequences; hence, it represents a maternal phylogeny. Therefore, the possibility cannot be ruled out that the relationship between B. acutus and S. micranthus may represent a further example of hybridization followed by chloroplast capture (see, e.g., Soltis and Kuzoff, 1995
; Wolfe and Elisens, 1995
; Sang, Crawford, and Stuessy, 1997
; Conti et al., 1999).
Clade V (BS = 99%, BL = 7; deletion in the rpl16 intron, B4, Table 2) comprises the remaining species of the genera Brachysiphon and Stylapterus and all species of Penaea. Most species in the clade have white or yellow flowers, simply valvate aestivation of the perianth lobes, relatively short pollen sacs, and stigmatic areas that are restricted to the angles between commissural stigma lobes. Exceptions for either one of the abovementioned characters are found in B. microphyllus and B. fucatus (see Fig. 28). Clade V is formed by a polytomy with four branches, of which only one supports a strongly monophyletic group (clade VI; BS = 98, BL = 6; insertion C11, Table 2).
Clade VI comprises all Penaea species except for P. dahlgrenii. The latter species was described by Rourke and McDonald (1989)
only after Dahlgren's (1971)
monograph of Penaea. In the protologue of P. dahlgrenii, Rourke and McDonald (1989)
remarked that the species is so unlike any other Penaea that its placement in the genus might be regarded as questionable. Our results lend some support to this view and confirm the monophyly of Penaea sensu Dahlgren (1971)
. The sister relationship of P. dahlgrenii with Stylapterus ericoides is weakly supported by molecular data (BS < 50%, BL = 2) and seems unlikely from a biogeographic standpoint, as both species are narrow endemics separated by rather long distances (ca. 150 km; Dahlgren, 1967a
; Rourke and McDonald, 1989
). In addition, because all species of Penaeaceae have ant-dispersed seeds with elaiosomes, dispersal over long distances seems rather unlikely (Bond and Slingsby, 1983
). Considering the weak support for this clade, it is possible that additional DNA sequence data may change the relationships of Penaea dahlgrenii and Stylapterus ericoides. Within clade VI (Penaea sensu Dahlgren) the five subspecies of P. cneorum (Dahlgren, 1971
) do not form a monophyletic group, but are paraphyletic with P. mucronata (Fig. 1). Both P. mucronata and P. cneorum are relatively widespread and variable in their phenotypic characters (Dahlgren, 1971
). To resolve phylogenetic relationships and species delimitation of these two species more informative characters and a much denser sampling at the population level are necessary. Another weakly supported clade within clade V comprises Stylapterus ericifolius, S. fruticulosus, and Brachysiphon fucatus. This clade seems unlikely from a morphological perspective, as B. fucatus differs conspicuously from the two Stylapterus species in flower color, perianth aestivation, and stigmatic characteristics. Conversely, some morphological characters, including the smaller and more rounded commissural stigma lobes on the style apex, long papillate instead of ciliate dehiscence margins of anthers, and a different ovary shape, suggest that S. ericifolius (and S. dubius, not included here) may be transitional in their morphology between Stylapterus and Brachysiphon (Dahlgren, 1967a
). Finally, the unresolved relationships of Brachysiphon microphyllus within clade V reflect the difficulty of assigning this morphologically distinctive species to any genus within Penaeaceae (Rourke, 1995
).
Perianth homology
The question of perianth homology in Penaeaceae, and especially in Oliniaceae, has been extensively debated (e.g., Fernandes and Fernandes, 1962
; Rao and Dahlgren, 1968
; Verdcourt, 1978
; Sebola and Balkwill, 1999
). Different interpretations of homology were based on evidence from floral anatomy and morphology (e.g., Rao and Dahlgren, 1969
) and floral organization and architecture (e.g., Dahlgren and van Wyk, 1988
), but were never integrated in an explicit phylogenetic framework. Here, we interpret the results of our analyses of floral development and structure in the context of the molecular phylogeny.
In Myrtales, the perianth generally consists of calyx and corolla (e.g., Dahlgren and Thorne, 1984
; Johnson and Briggs, 1984
). Among the closest relatives of Oliniaceae and Penaeaceae, the perianth either is typically biseriate, as in most Crypteroniaceae and in Rhynchocalycaceae, or the petals are absent, as in Crypteroniaceae p.p. (in Crypteronia, e.g., van Beusekom-Osinga and van Beusekom, 1975
) and in Alzateaceae (e.g., Graham, 1984
; but see van Beusekom-Osinga and van Beusekom, 1975
; they describe Alzateaceae as having rudimentary petals). As our study and earlier molecular phylogenetic analyses (Clausing and Renner, 2001
; Conti et al., 2002
) identified Rhynchocalycaceae as sister to the clade formed by Oliniaceae and Penaeaceae, we decided to compare perianth development and structure of these three families.
Although anthetic flowers of the three families appear to be rather different, the study of early developmental stages revealed clear similarities. In all three families young floral buds have a whorl of broadly attached organs (i.e., sepals of Rhynchocalycaceae, middle whorl Oliniaceae, perianth whorl of Penaeaceae) with valvate aestivation that enclose and protect the other floral organs (Figs. 3, 6, 9). Also anatomically and histologically these broadly attached perianth lobes of the three families share several features, i.e., they are relatively thick, they are served by three main vascular bundles, and they contain cells with oxalate druses restricted to the adaxial parenchymatic tissue in their basal part (Figs. 18–26). These shared features indicate that the middle whorl of perianth organs of Oliniaceae and the perianth organs of Penaeaceae are homologous to the sepals of Rhynchocalycaceae. The combination of traits that characterizes these organs, including their large size and protective function in early development, valvate aestivation, broad attachment, and three main vascular bundles, is typical for sepals (Endress, 1994
). In Rhynchocalycaceae and Oliniaceae a whorl of narrowly attached organs (i.e., petals of Rhynchocalycaceae, inner whorl of Oliniaceae) alternating with the sepals is initiated on the inner side of the hypanthium rim. These narrowly attached organs emerge in a more or less transversal position from the hypanthium rim and later bend downwards over the stamens (Figs. 4, 5, 7, 8). In both families these organs are delayed in their development as compared to the sepals and, although they are involved in pollinator attraction (Dahlgren and van Wyk, 1988
), they are relatively small at anthesis. At the beginning of anthesis they bend back and outwards and expose the stamens (see also Dahlgren and van Wyk, 1988
). Also anatomically and morphologically these innermost perianth organs of Rhynchocalycaceae and Oliniaceae are similar, as they are thin and served by a single vascular bundle (Figs. 18–26). The shared features indicate the homology of the petals of Rhynchocalycaceae and the inner perianth organs of Oliniaceae. The combination of delayed development, relatively thin organs, narrow attachment, supply by a single vascular bundle, and attractiveness to pollinators is characteristic for petals (Endress, 1994
).
In Penaeaceae, a conspicuous "empty" area between adjacent perianth lobes and the stamen positioned between them marks floral buds in early developmental stages (Figs. 10, 11). Such an "empty" area is unusual in a developing flower, where organs are normally densely packed to reduce damage from desiccation, frost, heat, or insects (Endress, 1994
). Exceptions to such dense packing are mainly present in groups where organs are aborted during floral development. Examples are found in the Lamiales, where the adaxial stamen is mostly reduced (Endress, 1999
), or in certain groups of Fabaceae, where petals and/or stamens are initiated early during floral ontogeny but are later aborted at different stages of development (Tucker, 1988
, 2001
). The "empty" areas in the floral buds of Penaeaceae are probably the remnants of a whorl of petals. Most likely, petals are still initiated on the meristematic flower apex of Penaeaceae, but their development is suppressed immediately after initiation, producing the conspicuous "empty" areas, which lack any distinct primordia. Initiation necessarily precedes appearance, and organ primordia may be present without being detectable as three-dimensional structures (Endress, 1992
, 1999
). Based on their vascular anatomical studies of Glischrocolla formosa, Rao and Dahlgren (1968)
interpreted the perianth lobes of Penaeaceae as petals, and the flowers, therefore, as being haplostemonous, but in later publications, Dahlgren and collaborators (Dahlgren and Thorne, 1984
; Dahlgren and van Wyk, 1988
) described the flowers of Penaeaceae as obhaplostemonous. Our own morphological developmental studies support the interpretation of perianth organs of Penaeaceae as sepals; hence, they also support the obhaplostemonous nature of the flowers.
An interesting feature that characterizes both Oliniaceae and Penaeaceae is that the perianth organs display a mixture of ecological functions. The sepals of both families are responsible for the protection of young floral organs. In Oliniaceae the sepals lose that function in later developmental stages; protection of stamens and carpels is then provided by the petals, which close the entrance to the floral tube (Fig. 15). At anthesis, the sepals of Oliniaceae are colored, relatively large and lingulate in shape, and share the function of pollinator attraction with the smaller, but also colored petals (see also Dahlgren and van Wyk, 1988
). In Penaeaceae the sepals provide protection to the other floral organs until anthesis. Pollinator attraction is shared between the colored sepals and the colored hypanthium (see also Dahlgren and Thorne, 1984
).
The homology of the outer whorl of the Oliniaceae perianth, formed by small tooth-like structures, has long been debated. In a study of floral anatomy of Olinia, Rao and Dahlgren (1969)
suggested that the "teeth" correspond to reduced sepals and the middle whorl to petals, hence the flowers were interpreted as haplostemonous. The innermost whorl was interpreted as being formed by the fusion of pairs of adjacent petal stipules. This view was later adopted by Sebola and Balkwill (1999)
. However, no structural evidence was found to support this hypothesis, and the main argument was based on organ arrangement, i.e., on the opposite position of stamens and organs of the inner whorl (Rao and Dahlgren, 1969
). Our investigations show that the tooth-like structures on the hypanthium rim of Oliniaceae are clearly different from the perianth organs of both Rhynchocalycaceae and Penaeaceae. In addition, the "teeth" arise only late during floral development (Fig. 6), a pattern that would be highly unusual for a reduced floral organ whorl. If the "teeth" were interpreted as reduced sepals, one would expect them to be initiated in the normal sequence, i.e., as the first organs on the developing floral apex, which are then aborted during later stages of development. This temporal sequence of floral organ initiation and subsequent abortion has been documented for all categories of floral organs and for a broad range of taxa, for example, for the reduced calyces of some Acanthaceae (Schönenberger and Endress, 1998
). However, if the tooth-like structures of the Oliniaceae perianth do not represent the calyx, to what do they correspond? Mayr (1969)
summarizes possible interpretations of comparable structures (the so-called epicalyx) in Lythraceae, another family of Myrtales. The epicalyx in Lythraceae has been interpreted as consisting of either simple outgrowths in the region of the congenitally united sepals (see also Dahlgren and Thorne, 1984
), as a separate whorl of bracteole-derived structures, or as fused calyx stipules. The often bifid structure of the "teeth" in Oliniaceae (e.g., Fig. 14) fits best with the interpretation of united stipules. However, the final answer to the question of the nature of the teeth-like structures remains elusive.
Myrtalean flowers generally have a double perianth (e.g., Cronquist, 1981
; Takhtajan, 1997
) and the ancestral condition of the androecium appears to be diplostemony, i.e., the flowers have twice as many stamens as petals or sepals and the outer whorl of stamens is in episepalous position (Ronse Decraene and Smets, 1991
). Flowers of Crypteroniaceae have a double perianth (Dactylocladus, Axinandra) or are apetalous (Crypteronia) and are diplostemonous (Axinandra) or obhaplostemonous (Crypteronia, Dactylocladus) (van Beusekom-Osinga and van Beusekom, 1975
; Pereira, 1996
). As in Rhynchocalyx, the petals in the flowers of Dactylocladus and Axinandra are enveloping the stamens in bud (Pereira, 1996
). Flowers of Alzatea are obhaplostemonous (e.g., Lourteig, 1965
; Graham, 1984
) and mostly described as apetalous (e.g., Dahlgren and Thorne, 1984
; Graham, 1984
), whereas van Beusekom-Osinga and van Beusekom (1975)
described the flowers of Alzatea as having rudimentary petals. Thus, there are two obvious evolutionary trends in the flowers of Crypteroniaceae, Alzateaceae, and Rhynchocalycaceae, i.e., (1) the reduction of petals and (2) the reduction or loss of the episepalous whorl of stamens, which leads to obhaplostemonous flowers. Not only do our developmental and structural studies suggest that the flowers of Oliniaceae and Penaeaceae should also be interpreted as obhaplostemonous, but our phylogenetic study also supports this interpretation, as it is most parsimonious to assume a common basic organization of the flowers in the entire clade formed by the four Western Gondwanan families plus Crypteroniaceae. In this phylogenetic framework (Fig. 1), obhaplostemony (i.e., stamens alternating with the sepals, represented by the middle perianth whorl in Oliniaceae and by the only perianth whorl in Penaeaceae) constitutes a clear synapomorphy for the clade. Conversely, if the floral "teeth" of Oliniaceae were interpreted as sepals and the organs of the middle whorl as petals, and/or the perianth organs of Penaeaceae as petals, the flowers would be haplostemonous, a less parsimonious interpretation of character state distribution.
Conclusions
Our cpDNA study confirms earlier results in supporting Alzateaceae as sister to the African families and Rhynchocalycaceae as sister to the clade formed by Oliniaceae and Penaeaceae. Further, our study provides a molecular phylogenetic hypothesis of the detailed relationships within Oliniaceae and Penaeaceae. Our results resolve Endonema as sister to all other Penaeaceae and strongly support several groups within the family, but contradict previous generic circumscriptions of Penaeaceae by suggesting that Brachysiphon and Stylapterus sensu Dahlgren (1967a
, 1968)
are not monophyletic. Definite taxonomic adjustments within the family will have to await results based on the analysis of additional molecular, preferably nuclear markers and a detailed study of floral morphology.
In addition, our molecular data indicate a relatively long, independent evolution of the stem lineages and a relatively recent origin of the crown groups of Penaeaceae and Oliniaceae, respectively. This phylogenetic pattern raises important questions pertaining to, for example, whether the radiations of the two crown groups occurred simultaneously and also whether these radiations coincided with climatic changes, as suggested for other South African taxa (Richardson et al., 2001
). The answers to questions on the tempo of evolution in these African families should be addressed with molecular dating methods.
Our developmental and structural analyses of the flowers of Rhynchocalycaceae, Oliniaceae, and Penaeaceae demonstrate that the flowers of all three families have the same basic organization, i.e., they are obhaplostemonous. The calyx of Rhynchocalycaceae is homologous to the middle perianth whorl in Oliniaceae and to the single perianth whorl of Penaeaceae. The corolla in Rhynchocalycaceae is homologous to the inner whorl of perianth organs in Oliniaceae. The flowers of Penaeaceae are apetalous or, more precisely, the petals are highly reduced. Calyx and corolla homology of the three families is supported by congruent developmental patterns, morphological/anatomical similarities, and phylogenetic reconstruction.
|
FOOTNOTES |
|---|
2 Current address: Department of Botany, University of Wisconsin, Madison, Wisconsin 53706 USA
3 Author for reprint requests (jschonenberg{at}wisc.edu
)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Baillon H. 1877 Histoire des plantes—XLVIII Pénaeacées. Hachette, Paris, France
Barker W. F. 1963 A new species of Penaeaceae. Journal of South African Botany 29: 167-169
Baum D. A. R. L. Small J. F. Wendel 1998 Biogeography and floral evolution of baobabs (Adansonia, Bombacaceae) as inferred from multiple data sets. Systematic Biology 47: 181-207[CrossRef][Web of Science][Medline]
Bond W. J. P. Slingsby 1983 Seed dispersal by ants in shrub lands of the Cape Province and its evolutionary implications. South African Journal of Science 79: 231-233[Web of Science]
Carlquist S. L. Debuhr 1977 Wood anatomy of Penaeaceae (Myrtales): comparative, phylogenetic, and ecological implications. Botanical Journal of the Linnean Society 75: 211-227[CrossRef]
Clausing G. S. S. Renner 2001 Molecular phylogenetics of Melastomataceae and Memecylaceae: implications for character evolution. American Journal of Botany 88: 486-498
Conti E. D. Baum K. Sytsma 1999a Phylogeny of Crypteroniaceae and related families: implications for morphology and biogeography. XVI International Botanical Congress, St. Louis—Abstract volume: 250 (Abstract)
Conti E. T. Eriksson J. Schönenberger K. J. Sytsma D. A. Baum 2002 Early Tertiary out-of-India dispersal of Crypteroniaceae: evidence from phylogeny and molecular dating. Evolution 56: 1931-1942[CrossRef][Web of Science][Medline]
Conti E. A. Litt K. J. Sytsma 1996 Circumscription of Myrtales and their relationships to other rosids: evidence from rbcL sequence data. American Journal of Botany 83: 221-233[CrossRef][Web of Science]
Conti E. A. Litt P. G. Wilson S. A. Graham B. G. Briggs L. A. S. Johnson K. J. Sytsma 1997 Interfamilial relationships in Myrtales: molecular phylogeny and patterns of morphological evolution. Systematic Botany 22: 629-647[CrossRef][Web of Science]
Conti E. D. E. Soltis T. M. Hardig J. Schneider 1999b Phylogenetic relationships of the Silver Saxifrages (Saxifraga, Sect. Ligulatae Haworth): implications for the evolution of substrate specificity, life history, and biogeography. Molecular Phylogenetics and Evolution 13: 536-555[CrossRef][Web of Science][Medline]
Cronquist A. 1981 An integrated system of classification of flowering plants. Columbia University Press, New York, New York, USA
Cufodontis G. 1960 Die Identifizierung von Tephea Delile und andere die Oliniaceae betreffende Feststellungen. Österreichische Botanische Zeitschrift 107: 106-112[CrossRef]
Dahlgren R. 1967a Studies in Penaeaceae. Part I. Systematics and gross morphology of the genus Stylapterus A. Juss. Opera Botanica 15: 3-40
Dahlgren R. 1967b Studies on Penaeaceae III. The genus Glischrocolla. Botaniska Notiser 120: 57-68
Dahlgren R. 1967c Studies on Penaeaceae IV. The genus Endonema. Botaniska Notiser 120: 69-83[Web of Science]
Dahlgren R. 1968 Studies on Penaeaceae. Part II. The genera Brachysiphon, Sonderothamnus and Saltera. Opera Botanica 18: 5-72
Dahlgren R. 1971 Studies on Penaeaceae. VI. The genus Penaea L. Opera Botanica 29: 5-58
Dahlgren R. R. F. Thorne 1984 The order Myrtales: circumscription, variation, and relationships. Annals of the Missouri Botanical Garden 71: 633-699[CrossRef][Web of Science]
Dahlgren R. A. E. van Wyk 1988 Structures and relationships of families endemic to or centered in Southern Africa. Monographs in Systematic Botany from the Missouri Botanical Garden 25: 1-94
de Candolle A. 1857 Penaeaceae. In A. de Candolle [ed.], Prodromus systematis naturalis regni vegetabilis, vol. 14, 483–491. Masson, Paris, France
Dickie J. B. P. E. Gasson 1999 Comparative leaf anatomy of the Penaeaceae and its ecological implications. Botanical Journal of the Linnean Society 131: 327-351[CrossRef]
Downie S. R. S. Ramanath D. S. Katz-Downie E. Llanas 1998 Molecular systematics of Apiaceae subfamily Apioideae: phylogenetic analyses of nuclear ribosomal DNA internal transcribed spacer and plastid rpoC1 intron sequences. American Journal of Botany 85: 563-591[Abstract]
Eichenberger K. F. Gugerli J. J. Schneller 2000 Morphological and molecular diversity of Swiss common bean cultivars (Phaseolus vulgaris L., Fabaceae) and their origin. Botanica Helvetica 110: 61-77
Endress P. K. 1992 Evolution and floral diversity: the phylogenetic surroundings of Arabidopsis and Antirrhinum. International Journal of Plant Sciences 153: S106-S122[CrossRef]
Endress P. K. 1994 Diversity and evolutionary biology of tropical flowers. Cambridge University Press, Cambridge, UK
Endress P. K. 1999 Symmetry in flowers: diversity and evolution. International Journal of Plant Sciences 160: 3-23[CrossRef]
Farris J. S. M. Källersjö A. G. Kluge C. Bult 1995 Testing significance of incongruence. Cladistics 10: 315-319[CrossRef][Web of Science]
Felsenstein J. 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791[CrossRef][Web of Science]
Fernandes A. R. Fernandes 1962 O género Olinia Thunb. em Angola. Memórias da Junta de Investigações do Ultramar 38: 9-20
Fitch W. M. 1971 Toward defining the course of evolution: minimal change for a specific tree topology. Systematic Zoology 20: 406-416[CrossRef][Web of Science]
Gene Codes. 1998 Sequencher, version 3.1.1. Gene Codes, Ann Arbor, Michigan, USA
Gilg E. 1894a Oliniaceae. In A. Engler and K. Prantl [eds.], Die natürlichen Pflanzenfamilien, vol. III, Abt. 6, 213–216. W. Engelmann, Leipzig, Germany
Gilg E. 1894b Penaeaceae. In A. Engler and K. Prantl [eds.], Die natürlichen Pflanzenfamilien, vol. III, Abt. 6, 208–213. W. Engelmann, Leipzig, Germany
Goldblatt P. J. Manning 2000 Cape plants: a conspectus of the Cape Flora of South Africa. Strelitzia 9: 1-743
Graham S. A. 1984 Alzateaceae, a new family of Myrtales in the American Tropics. Annals of the Missouri Botanical Garden 71: 757-779[CrossRef][Web of Science]
Hamilton M. B. 1999 Four primer pairs for the amplification of chloroplast intergenic regions with intraspecific variation. Molecular Ecology 8: 513-525[CrossRef][Medline]
Igersheim A. 1993 The character states of the Caribbean monotypic endemic Strumpfia (Rubiaceae). Nordic Journal of Botany 13: 545-559[CrossRef][Web of Science]
Igersheim A. O. Cichocki 1996 A simple method for microtome sectioning of prehistoric charcoal specimens, embedded in 2-hydroxyethyl methacrylate (HEMA). Review of Palaeobotany and Palynology 92: 389-393[CrossRef]
Johnson L. A. S. B. G. Briggs 1984 Myrtales and Myrtaceae—a phylogenetic analysis. Annals of the Missouri Botanical Garden 71: 700-756[CrossRef][Web of Science]
Kelchner S. A. 2000 The evolution of non-coding chloroplast DNA and its application in plant systematics. Annals of the Missouri Botanical Garden 87: 482-498[CrossRef][Web of Science]
Lourteig A. 1965 On the systematic position of Alzatea verticillata R. & P. Annals of the Missouri Botanical Garden 52: 371-378[CrossRef]
Manen J.-F. A. Natali F. Ehrendorfer 1994 Phylogeny of Rubiaceae-Rubieae inferred from the sequence of a cpDNA intergenic region. Plant Systematics and Evolution 190: 195-211[CrossRef][Web of Science]
Mayr B. 1969 Ontogenetische Studien an Myrtales-Blüten. Botanische Jahrbücher 89: 210-271
McDade L. M. L. Moody 1999 Phylogenetic relationships among Acanthaceae: evidence from non-coding trnL-trnF chloroplast DNA sequences. American Journal of Botany 86: 70-80
Mújica M. B. D. F. Cutler 1974 Taxonomic implications of anatomical studies on the Oliniaceae. Kew Bulletin 29: 93-123[CrossRef]
Oliver D. 1894 Rhynchocalyx lawsonioides Oliv. In D. Oliver [ed.], Hooker's Icones Plantarum, vol. 24, 2348. Dulau, London, UK
Oxelman B. M. Lidén D. Berglund 1997 Chloroplast rps16 intron phylogeny of the tribe Silenae (Caryophyllaceae). Plant Systematics and Evolution 206: 393-410[CrossRef][Web of Science]
Patel V. C. J. J. Skvarla P. H. Raven 1984 Pollen characters in relation to the delimitation of Myrtales. Annals of the Missouri Botanical Garden 71: 858-969[CrossRef][Web of Science]
Pereira J. T. 1996 Crypteroniaceae. In E. Soepadmo, K. M. Wong, and L. G. Saw [eds.], Tree flora of Sabah and Sarawak, vol. 2, 135–149. Forest Research Institute Malaysia, Sabah Forestry Department, and Sarawak Forestry Deparment, Kuala Lumpur, Malaysia
Phillips E. P. 1926 The genera of South African flowering plants. Botanical Survey of South Africa 10: 1-702
Phillips E. P. 1951 The genera of South African flowering plants. Botanical Survey Memoir 25: 1-923
Rao V. S. R. Dahlgren 1968 Studies on Penaeaceae V. The vascular anatomy of the flower of Glischrocolla formosa. Botaniska Notiser 121: 259-268[Web of Science]
Rao V. S. R. Dahlgren 1969 The floral anatomy and relationships of Oliniaceae. Botaniska notiser 122: 160-171[Web of Science]
Richardson J. E. et al 2001 Rapid and recent origin of species richness in the Cape flora of South Africa. Nature 412: 181-183[CrossRef][Medline]
Ronse Decraene L.-P. E. Smets 1991 The impact of receptacular growth on polyandry in the Myrtales. Botanical Journal of the Linnean Society 105: 257-269[CrossRef]
Rourke J. P. 1995 A new species of Brachysiphon (Penaeaceae) from the Southern Cape, South Africa. Nordic Journal of Botany 15: 63-66[CrossRef][Web of Science]
Rourke J. P. D. J. McDonald 1989 A new species of Penaea (Penaeaceae), from the Langeberg range, southern Cape. South African Journal of Botany 55: 400-404[Web of Science]
Sang T. D. J. Crawford T. F. Stuessy 1997 Chloroplast DNA phylogeny, reticulate evolution, and biogeography of Paeonia (Paeoniaceae). American Journal of Botany 84: 1120-1136[Abstract]
Savolainen V. et al 2000 Phylogeny of the eudicots: a nearly complete familial analysis based on rbcL gene sequences. Kew Bulletin 55: 257-309[CrossRef]
Schönenberger J. P. K. Endress 1998 Structure and development of the flowers in Mendoncia, Pseudocalyx, and Thunbergia (Acanthaceae) and their systematic implications. International Journal of Plant Sciences 159: 446-465[CrossRef]
Sebola R. J. K. Balkwill 1999 Resurrection of two previously confused species, Olinia capensis (Jacq.) Klotsch and O. micrantha Decne. (Oliniaceae). South African Journal of Botany 65: 97-103
Simmons M. P. H. Ochoterena 2000 Gaps as characters in sequence-based phylogenetic analyses. Systematic Biology 49: 369-381[CrossRef][Web of Science][Medline]
Small R. L. J. A. Ryburn R. C. Cronn T. Seelanan J. F. Wendel 1998 The tortoise and the hare: choosing between noncoding plastone and nuclear adh sequences for phylogeny reconstruction in a recently diverged plant group. American Journal of Botany 85: 1301-1315
Soltis D. E. R. K. Kuzoff 1995 Discordance between nuclear and chloroplast phylogenies in the Heuchera group (Saxifragaceae). Evolution 49: 727-742[CrossRef][Web of Science]
Sonder W. 1862 Olinieae. In W. H. Harvey and O. W. Sonder [eds.], Flora capensis, vol. 2, 519–520. L. Reeve, London, UK
Supprian K. 1894 Beiträge zur Kenntnis der Thymelaeaceae und Penaeaceae. Botanische Jahrbücher für Systematik, Pflanzengeschichte und Pflanzengeographie 18: 306-353
Swofford D. L. 2000 PAUP*: phylogenetic analysis using parsimony (*and other methods). Sinauer, Sunderland, Massachusetts, USA
Takhtajan A. 1997 Diversity and classification of flowering plants. Columbia University Press, New York, New York, USA
Thorne R. F. 1992 An updated phylogenetic classification of the flowering plants. Aliso 13: 365-389
Tobe H. P. H. Raven 1983a An embryological analysis of Myrtales: its definition and characteristics. Annals of the Missouri Botanical Garden 70: 71-94[CrossRef][Web of Science]
Tobe H. P. H. Raven 1983b The embryology of Axinandra zeylanica (Crypteroniaceae) and the relationships of the genus. Botanical Gazette 144: 426-432[CrossRef][Web of Science]
Tobe H. P. H. Raven 1984a The embryology and relationships of Alzatea Ruiz & Pav. (Alzateaceae, Myrtales). Annals of the Missouri Botanical Garden 71: 844-852[CrossRef][Web of Science]
Tobe H. P. H. Raven 1984b The embryology and relationships of Oliniaceae. Plant Systematics and Evolution 146: 105-116[CrossRef][Web of Science]
Tobe H. P. H. Raven 1984c The embryology and relationships of Penaeaceae (Myrtales). Plant Systematics and Evolution 146: 181-195[CrossRef][Web of Science]
Tobe H. P. H. Raven 1984d The embryology and relationships of Rhynchocalyx Oliv. (Rhynchocalycaceae). Annals of the Missouri Botanical Garden 71: 836-843[CrossRef][Web of Science]
Tobe H. P. H. Raven 1987a The embryology and relationships of Crypteronia (Crypteroniaceae). Botanical Gazette 148: 96-102[CrossRef]
Tobe H. P. H. Raven 1987b The embryology and relationships of Dactylocladus (Crypteroniaceae) and a discussion of the family. Botanical Gazette 148: 103-111[CrossRef]
Tucker S. C. 1988 Loss versus suppression of floral organs. In P. Leins, S. C. Tucker, and P. K. Endress [eds.], Aspects of floral development, 69–82. Gebrüder Bornträger, Berlin, Germany
Tucker S. C. 2001 The ontogenetic basis for missing petals in Crudia (Leguminosae: Caesalpinioideae: Detarieae). International Journal of Plant Sciences 162: 83-89[CrossRef][Web of Science]
van Beusekom-Osinga R. J. C. F. van Beusekom 1975 Delimitation and subdivision of the Crypteroniaceae. Blumea 22: 255-266
van Vliet G. J. C. M. P. Baas 1984 Wood anatomy and classification of the Myrtales. Annals of the Missouri Botanical Garden 71: 783-800[CrossRef][Web of Science]
van Wyk B. P. van Wyk 1997 Field guide to trees of Southern Africa. Struik Publishers, Cape Town, South Africa
Verdcourt B. 1975 Oliniaceae. In R. M. Polhill [ed.], Flora of Tropical East Africa, vol. 2, 1–4. Crown Agents for Oversea Governments and Administrations, London, UK
Verdcourt B. 1978 Oliniaceae. In E. Launert [ed.], Flora Zambesica, vol. 4, 323–327. Flora Zambesica Managing Committee, London, UK
Weberling F. 1988 The architecture of inflorescences in the Myrtales. Annals of the Missouri Botanical Garden 75: 226-310[CrossRef][Web of Science]
Wolfe A. D. W. J. Elisens 1995 Evidence of chloroplast capture and pollen-mediated gene flow in Penstemon sect. Peltanthera (Scrophulariaceae). Systematic Botany 20: 395-412[CrossRef][Web of Science]
This article has been cited by other articles:
|
|
 |
|
  L. P. Ronse De Craene Are Petals Sterile Stamens or Bracts? The Origin and Evolution of Petals in the Core Eudicots Ann. Bot., September 1, 2007; 100(3): 621 - 630. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Rutschmann, T. Eriksson, K. A. Salim, and E. Conti Assessing Calibration Uncertainty in Molecular Dating: The Assignment of Fossils to Alternative Calibration Points Syst Biol, August 1, 2007; 56(4): 591 - 608. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Fior, P. O. Karis, G. Casazza, L. Minuto, and F. Sala Molecular phylogeny of the Caryophyllaceae (Caryophyllales) inferred from chloroplast matK and nuclear rDNA ITS sequences Am. J. Botany, March 1, 2006; 93(3): 399 - 411. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Marazzi, P. K Endress, L. P. d. Queiroz, and E. Conti Phylogenetic relationships within Senna (Leguminosae, Cassiinae) based on three chloroplast DNA regions: patterns in the evolution of floral symmetry and extrafloral nectaries Am. J. Botany, February 1, 2006; 93(2): 288 - 303. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Shaw, E. B. Lickey, J. T. Beck, S. B. Farmer, W. Liu, J. Miller, K. C. Siripun, C. T. Winder, E. E. Schilling, and R. L. Small The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis Am. J. Botany, January 1, 2005; 92(1): 142 - 166. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Baum, S. DeWitt Smith, A. Yen, W. S. Alverson, R. Nyffeler, B. A. Whitlock, and R. L. Oldham Phylogenetic relationships of Malvatheca (Bombacoideae and Malvoideae; Malvaceae sensu lato) as inferred from plastid DNA sequences Am. J. Botany, November 1, 2004; 91(11): 1863 - 1871. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. O. Simoes, M. E. Endress, T. van der Niet, L. S. Kinoshita, and E. Conti Tribal and intergeneric relationships of Mesechiteae (Apocynoideae, Apocynaceae): evidence from three noncoding plastid DNA regions and morphology Am. J. Botany, September 1, 2004; 91(9): 1409 - 1418. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Shaw and R. L. Small Addressing the "hardest puzzle in American pomology:" Phylogeny of Prunus sect. Prunocerasus (Rosaceae) based on seven noncoding chloroplast DNA regions Am. J. Botany, June 1, 2004; 91(6): 985 - 996. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
