| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
Department of Biology, University of North Carolina, Chapel Hill, North Carolina 27599-3280 USA
Received for publication February 2, 2001. Accepted for publication May 17, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: classification • matK • molecular phylogeny • rbcL • Theaceae • Theoideae
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
circumscription of Theaceae D. Don (nom. cons.) includes 
40 genera and 600 species in four subfamilies: Ternstroemioideae (including Sladenia Kurz), Theoideae, Bonnetioideae, and Asteropeioideae. Dahlgren (1983)
also includes Pelliciera rhizophorae Triana & Planchon and Tetramerista Miquel in the family, but excludes Bonnetioideae. Goldberg (1986)
, Thorne (1992)
, and Takhtajan (1997)
have the most restrictive circumscriptions based on morphology, limiting the family to the two major subfamilies Ternstroemioideae and Theoideae (plus Sladenioideae: Sladenia in Takhtajan). Theaceous plants can be distinguished from related families (Lawrence, 1951
) by the spiral arrangement of the perianth parts, the several series of numerous stamens, and the presence of involucral bracts that often grade into sepals, yet all of these features can be found in other families of flowering plants such as Actinidiaceae, Pentaphylacaceae, Symplocaceae, and Tetrameristaceae, suggesting that these character states are likely to be pleisiomorphic.
Ribulose-1,5-bisphosphate carboxylase/oxygenase (rbcL) DNA sequence data analyses published by Morton and colleagues (Morton et al., 1996, 1997
; Morton, Karol, and Chase, 1997
) divide the family into two distantly placed lineages, with Asteropeioideae (as Asteropeiaceae) near Physenaceae and Caryophyllales, while Ternstroemioideae and Theoideae are part of a larger Ericales clade (APG, 1998
). The studies by Morton and colleagues also found subfamilies Theoideae and Ternstroemioideae to each be monophyletic but not sister to each other, suggesting paraphyly for the family. More powerful analyses of two-gene (Savolainen et al., 2000
) and three-gene (Soltis et al., 2000
) data sets fail to unite the two subfamilies into a monophyletic lineage although the strict consensus tree was unresolved. These findings support the recognition of two distinct families, Theaceae and Ternstroemiaceae, as demonstrated in the multigene publications and as suggested by the APG (1998)
, but it should be accepted with caution since only two or three taxa were sampled for each family sensu stricto (s.s.). Theaceae will follow a narrow definition (= Theoideae of Cronquist, 1981
) for the remainder of the text.
Theaceae s.s. includes 7–21 genera depending on the classification system employed (see Table 1 for representative classification systems and list of possible recognized genera). Members are most diverse in tropical and subtropical Asia, but representatives also occur in warm-temperate Asia and North America and in the American tropics. Members of the family are characterized by the presence of generally large, showy, solitary flowers borne in terminal leaf axils. The suite of characters used to identify Theaceae are not unique to the family; similar features can be found in members of Actinidiaceae, Symplocaceae, and Tetrameristaceae. Tsou's preliminary palynological studies (1997, 1998) identify the production of a specialized pseudopollen in all of the Theaceae genera examined. Tsou (1997)
describes this feature as "the only autapomorphy so far determined" for this family.
|
; Airy-Shaw, 1936
) emphasize floral characters such as the size, number, and degree of fusion of the bracteoles, sepals, and petals in the erection of tribes and genera. These characters appear to vary considerably and often continuously from species to species within a genus and even within species. Sealy (1958)
and Keng (1962)
placed more emphasis on fruit dehiscence and gross morphology, as well as seed characters, such as the presence or absence of a wing and the amount of endosperm present. Generic circumscriptions are as problematic as tribal taxonomy in Theaceae. Monotypic genera such as Apterosperma Chang, Dankia Gagnepain, and Stereocarpus (Pierre) H. Hallier are often segregated out of larger genera, resulting in several probable paraphyletic groupings.
A number of taxa were considered as potential closest relatives of Theaceae since the results of recent molecular data analyses (Morton et al., 1996, 1997
; Soltis et al., 2000
) do not provide significant statistical support for the selection of one specific taxon as an appropriate outgroup. The taxa selected as outgroups for this study were based on a combination of previous morphological and anatomical work and an analysis of rbcL sequences available from GenBank (NCBI [National Center for Biotechnology Information]) and include members of Clethraceae, Cyrillaceae, Symplocaceae, and Ternstroemiaceae. Trees were rooted using Hydrangea L. as the ultimate outgroup.
In this study, we sampled extensively in Theaceae and Ternstroemiaceae using two variable chloroplast DNA regions (rbcL and matK + portions of the flanking spacer regions) to confirm the monophyly of Theaceae, assess the circumscription of tribes, and evaluate the utility of various classification systems. This research also addresses several important taxonomic questions including the relationship between Franklinia and Gordonia Ellis, the monophyly of Hartia Dunn. and Stewartia L., and the monophyly of Gordonia sensu lato (s.l.).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Molecular methods
Total genomic DNAs were extracted from 1.5–3.0 g of fresh leaf tissue or 0.5–1.0 g of silica-dried plant material of a single individual using a minor modification of Doyle and Doyle (1987)
. The aqueous phase was extracted with 24 parts chloroform : 1 part isoamyl alcohol. When fresh or silica dried material was unavailable, a 1-cm diameter disk of leaf tissue from herbarium specimens was used. DNA was resuspended in TE buffer (10 mmol/L tris-HCl, 1 mmol/L EDTA, pH 8.0) following isopropyl alcohol precipitation at –20°C for approximately 24 h (2 wk for herbarium material). DNA was purified by reprecipitation using 1/10 volume 3 mol/L sodium acetate (pH 4.8) and 70% ethanol, spooling on a glass hook, rinsing in wash buffer (76% ethanol, 10 mmol/L ammonium acetate), air-drying, and resuspending in TE buffer.
Amplification of the rbcL gene utilized a number of forward and reverse primers 20–30 bases in length, producing an 1382 base pair (bp) product using Promega Taq DNA polymerase (Promega, Madison, Wisconsin, USA) according to the manufacturer's directions with an annealing temperature of 48–50°C. A list of all primers utilized in the amplification and sequencing are available from the authors. Many of the rbcL primers are modified from published Zurawski primers (Taylor and Swann, 1994
). Amplification of the matK gene and flanking spacer regions (= matK throughout the remainder of the paper) used primers anchored in the 3' and 5' trnK gene under conditions similar to the rbcL gene. The amplification products were 2600 bp in length but only the 2118 bp at the 3' end were sequenced (characters 1–284 corresponding to the aligned 5' intergenic spacer region between 5' trnK and matK, characters 285–1738 corresponding to the aligned matK gene, and 1738–2918 corresponding to the aligned 3' intergenic spacer between matK and 3' trnK). Herbarium material generally required amplification in fragments 400–600 bp in length.
Amplified rbcL products were purified using glassmilk and a chaotropic sodium iodide solution (a modification of the Vogelstein and Gillespie [1979
] protocol) and were sequenced directly using a 32P alpha labeled dATP in the dideoxynucleotide method of Sanger, Nicklen, and Coulson (1977)
with Sequenase version 2.0 (United States Biochemical, Cleveland, Ohio, USA). We sequenced matK products using automated sequencing methodology of the ABI Prism Terminator Cycle Sequencing Ready Reaction Kit (original dyes with AmpliTaq DNA Polymerase, Perkin Elmer, Foster City, California, USA) at half reaction volumes. Products were cleaned in Sephadex G-50 (fine) Centri-Sep spin columns (Princeton Separations, Adelphia, New Jersey, USA). Samples were dried under vacuum and run on an ABI 377 (Applied Biosystems, Foster City, California, USA) autosequencer at the Iowa State University DNA Sequencing Facility (Ames, Iowa, USA). Raw sequences were assembled and edited using Sequencher (Gene Codes Corporation, Ann Arbor, Michigan, USA), and manually aligned in Se-Al version 1.0 (Rambaut, 1996
). Inferred insertion/deletion (indel) events in the matK data set were coded as presence/absence characters at the end of the matK matrix (characters 2919–2945).
Computational methods
All analyses were conducted in PAUP* version 4.0 (beta test version, Swofford, 1998
) and run to completion unless otherwise noted. Fitch (1971)
equally weighted parsimony analyses were conducted with 500 random sequence addition replicates, tree bisection-reconnection (TBR) branch swapping, saving all shortest trees. Multiple random sequence additions were chosen to minimize the likelihood of being trapped on any particular tree island (Maddison, 1991
). Robustness of clades was evaluated by decay analyses (Bremer, 1988
; Donoghue et al., 1992
) and bootstrap analysis (Felsenstein, 1985
). Constraint trees for decay analyses were created using Autodecay version 4.0 (Eriksson, 1998
) and then run in PAUP* with 100 random addition replicates. Bootstrap analysis used 100 random addition replicates with TBR branch swapping, saving a maximum of 1000 trees for 100 bootstrap replicates. Several different analyses were conducted for the matK matrix to evaluate the effect of missing data for some taxa and the effect of indel coding as individual presence/absence characters. Data for taxa common to both data sets were subjected to partition homogeneity tests as implemented in PAUP* with 10 000 replicates, no branch swapping. Results of this test and a visual examination of the tree topology and branch support for individual data analyses were used to determine whether combined analyses were appropriate. Combined data analyses were conducted as described for the individual data sets above.
Preliminary analyses consistently identified three lineages within Theaceae. To better assess the relationship between these lineages, three constraints trees were constructed by rearranging a simple neighbor-joining tree to represent all three possible sister relationships of the three clades. The possible topologies are: tree topology 1 ((Thee + Stew) Gord); tree topology 2 (Thee (Gord + Stew)); and tree topology 3 ((Thee + Gord) Stew). Likelihood scores for three possible tree topologies were calculated under 16 different models of evolution (Jukes-Cantor [Jukes and Cantor, 1969
], Kimura two-parameter [Kimura, 1980
], Hasegawa-Kishino-Yano [Hasegawa, Kishino, and Yano, 1985
], and General Time Reversible [Lanave et al., 1984
; Tavaré, 1986
; Rodríguez et al., 1990
], each with or without gamma rate estimation [G] and proportion of invariant sites estimation [I]) to determine whether any particular topology was significantly better than the other two (as described by Felsenstein [1988]
assuming one degree of freedom). Use of the likelihood ratio test in this instance is based on information from tribal classification based on morphology and the current data set as evaluated under the parsimony criterion. Although this application violates the condition that the different tree topologies be specified a priori (Swofford et al., 1996
; Goldman, Anderson, and Rodrigo, 2000
), it can give an estimation of relative support for any particular topology over the others.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
The rbcL analysis produced 2576 trees of length (L) 627 steps (retention index [RI] = 0.6307, consistency index [CI] = 0.4211, rescaled consistency index [RC] = 0.2656 excluding uninformative characters) in one tree island (Maddison, 1991
). Figure 1 depicts the strict consensus of the equally parsimonious trees with bootstrap values indicated. Decay indices parallel bootstrap values with highest decay values on branches with higher bootstrap support as shown on all relevant figures. A number of clades identified correspond to groups recognized by Airy-Shaw (1936)
including a Gordoniinae clade (bootstrap 85%, decay = 3 steps), a Stewartiinae clade (bootstrap 87%, decay = 3 steps), and a Camellieae clade (bootstrap 64%, decay = 1 step). The strict consensus tree fails to unite the two families Theaceae and Ternstroemiaceae as a monophyletic lineage. Although there is no statistical support (bootstrap <50%, decay = 1 step), Styrax is consistently placed sister to representatives of Ternstroemiaceae.
|
Initial analyses were conducted on the entire matrix, treating gaps as missing data. The analysis for 50 taxa (some incomplete sequences) produced 13 626 equally most parsimonious trees. Two Theaceae sequences were not completed due to technical difficulties: Apterosperma oblata (missing 75%) and Adinandra millettei (missing 25%). These two taxa were excluded from analyses with 48 taxa (results not shown). Five outgroup taxa (Rhododendron hippophaeoides, Fouquieria splendens, Actinidia chinensis, Sarracenia purpurea, and Hydrangea quercifolia) lacked 23–43% of their sequence data, primarily for the flanking spacer regions, and were excluded (along with the two Theaceae sequences listed above) in analyses for 43 taxa (results not shown). Placement of taxa based on incomplete sequences is tentative. The inclusion of taxa for which some sequence data was missing did not alter the topology of the tree (except within Camellia s.l.), indicating that missing data was not causing tree instability in the larger analyses. Figure 2 provides a strict consensus tree for the most extensive analysis (50 taxa, indels deleted but coded as independent presence/absence characters) with bootstrap values indicated. Taxa for which incomplete sequences were used are indicated with dashed branches. All equally parsimonious trees (738 trees of L = 506) had a CI of 0.7174 and an RI of 0.9112 (excluding uninformative characters). The RC was 0.6537. Bootstrap evaluation of the gene sequence matrix subcomponents (matK sequence vs. sequence minus indels plus indel coding) supported identical branches with similar levels of support (data not shown).
|
Combined analyses
Partition homogeneity test (for the 39 taxa common to both data sets) showed no significant heterogeneity between the two data sets (P = 0.8785). The combined data analyses (64 taxa with much missing data, matK indels excluded but coded independently) produced 13 449 equally most parsimonious trees in three tree islands (Maddison, 1991
). The tree islands differ in the sister group relationship of the three major clades (Stewartia + Hartia = Stewartieae, Franklinia + Gordonia + Schima = Gordonieae, all others = Theeae). Another major difference between these tree islands are the sister group relationships relative to Theaceae. In tree islands 1 and 2, Symplocos are sister to Theaceae, while in tree island 3 Clethra is the sister taxon. One of the equally parsimonious trees has been redrawn in Fig. 3. The strict consensus of the equally most parsimonious trees was highly resolved and is shown in Fig. 4. The combined analyses provided the same general topology as the individual matrix analyses, but with higher bootstrap and decay values. Other studies of combined data analyses have shown a significant decrease in overall search time, better resolution than the separate analysis trees, and fewer equally most parsimonious trees (Soltis et al., 1998
). The results presented here show similar trends.
|
|
|
lnL = 3.51923) data. The tree topology resulting in the best likelihood score was consistent regardless of the model of evolution used with rbcL equivocal, matK spacer favoring tree topology 2 (Thee (Gord + Stew)) , and the other three partitions (matK coding, allmatK, combined matK and rbcL) favoring tree topology 3 ((Thee + Gord) Stew). |
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Monophyly of Theaceae s.l
The most parsimonious trees, as summarized in the strict consensus trees (Figs. 1–5), do not support a monophyletic Theaceae s.l. The data provide strong support for the monophyly of Theaceae s.s. and moderate support for the monophyly of Ternstroemiaceae. The recognition of two distinct, nonsister lineages is also supported by at least one additional chloroplast intergenic spacer DNA data set (Prince, 2000
).
The circumscription of family Theaceae has been modified greatly over the past 100 yr with the primary trend toward a systematic pruning of taxa (subfamily Bonnetioideae, Tetramerista, Pelliciera, etc.), thus the suggestion of a [still] paraphyletic Theaceae is not unreasonable nor unexpected. Indeed, when the family name Theaceae was first published, Ternstroemia and its close relatives were placed in a separate family, Ternstroemiaceae (Mirbel, 1813
). The data analyzed here provide support for the recognition of two distinct families, a narrowly defined Theaceae (= Theoideae of Cronquist) and Ternstroemiaceae (= Ternstroemioideae of Cronquist).
Tribal and subtribal classification
The phylogenies provided in Figs. 3 and 5 are working hypotheses of relationships within Theaceae. Several hypotheses of relationships have been proposed for genera of the family (Table 1). Most authors recognize at least two tribes or subtribes, but the specific composition varies considerably from one system to another. The most striking differences between classification systems involve the placement of Stewartia (including Hartia) as either an independent tribe/subtribe, or as a subtribe within Gordonieae; and the circumscription of Gordonia and the subsequent placement of segregate genera Laplacea and Polyspora.
The strict consensus trees (Figs. 4 and 5) retained three major lineages within Theaceae, namely subtribes Gordoniinae and Stewartiinae and tribe Camellieae (= Theeae) as circumscribed by Airy-Shaw (1936)
. These lineages have high (
90%) bootstrap support in the combined rbcL and matK analyses (Fig. 5). The data clearly do not support the tribal and subtribal classifications of Sealy (1958)
, Keng (1962)
, Melchior (1964)
, Ye (1990)
, or Takhtajan (1997)
since they all include Polyspora and Laplacea within Gordonia. In addition, the placement of Schima and Franklinia in the same subtribe as Stewartia by Sealy (1958)
and Melchior (1964)
is not supported. Melchior postulated a close relationship between Camellia s.l., Pyrenaria s.l., Laplacea, and Polyspora, by placing all in the same tribe Theeae, a relationship supported by the available data. Our data support the recognition of three lineages at the tribal level: Theeae, Gordonieae, and Stewartieae.
Resolution of relationships within each of these major lineages varies, but is least clear in the tribe Theeae. The low level of resolution in the Theeae lineage could be explained by either a shift in the substitution rate, by a relatively rapid radiation of the group, or by higher levels of extinction in the other tribes. One of the authors is investigating relationships within this tribe using more rapidly evolving regions of the genome.
Generic circumscription
Camellia s.l. may be divided into as many as nine genera based on recent synonymy: Camellia, Camelliastrum, Dankia, Glyptocarpa, Parapiquetia, Piquetia, Stereocarpus, Theopsis, and Yunnanea. Characters used to diagnose these segregate genera are variable and (probably) continuously distributed, such as the number and location of flowers in the inflorescence (axillary vs. terminal, solitary vs. multiple), the number of ovules per locule, number of locules per ovary, versatility of the anther, etc. All contemporary scientists working in this family recognize one large, broadly defined genus Camellia with approximately 135–300 species (Sealy, 1958
; Chang and Bartholomew, 1984
).
Sampling for this study was limited to representatives of Camellia s.s. and Glyptocarpa camellioides only. The strict consensus tree for the 50 taxon matK analysis (indels included, results not shown) clearly unites these taxa, but the exclusion of indel data and subsequent coding as presence/absence characters results in loss of resolution for this particular clade (Fig. 2). The combined rbcL and matK analyses group four of the six representatives, but only with weak bootstrap support (Figs. 4 and 5). Independent studies are underway to address the circumscription of Camellia s.l. The data presented here do not conflict with the recognition of a broadly circumscribed genus Camellia.
The generic circumscription of Pyrenaria s.l. (Pyrenaria, Parapyrenaria, Sinopyrenaria, and Tutcheria) could not be addressed as only representatives of the genus Tutcheria provided reliable sequence data. Morphological characters used to diagnose these different genera include flower location (terminal vs. axillary), the number and texture of the sepals and petals, a dehiscent fruit vs. indehiscent fruit, and the number of seeds per locule. Yang (1998)
used cytological, morphological, and molecular evidence to evaluate relationships within Pyrenaria s.l. His findings support the recognition of a large, broadly circumscribed genus in which Sinopyrenaria, Tutcheria, and Parapyrenaria are included.
Stewartia s.l. was represented by seven taxa, two (evergreen) Hartia representatives, and five (deciduous) Stewartia representatives. Hartia sequences are clearly allied to Stewartia monodelpha, resulting in a paraphyletic Stewartia s.s. The first author is collecting data for a nuclear intergenic spacer region to confirm relationships in Stewartieae.
The monophyly of the genus Gordonia s.l. was not supported by our sequence data. All analyses placed Gordonia lasianthus (and Gordonia brandegeei in the matK and some combined analyses) as a member of Gordonieae, and the remaining species (Polyspora and Laplacea) as members of Theeae. The data also supported the recognition of two distinct lineages, an Old World Polyspora lineage and a New World Laplacea lineage. Additional sampling is required to determine whether all Old World Laplacea species would fall into the Polyspora clade.
The inclusion of Gordonia lasianthus, Polyspora, and Laplacea into a broadly defined genus Gordonia (Keng, 1980
) was based on the overall similarity of the capsular fruit and the apically winged seeds. In all cases, the fruit is a loculicidally dehiscent capsule, usually of five (or more) carpels. Mature fruits have a persistent columella and often retain some of the bracts and/or sepals. The seeds bear an apical wing, a feature not found elsewhere in the subfamily (although other wing arrangements are found). It is possible that the capsule similarities and the presence of the apical wing are not homologous structures in these three groups. Developmental studies on the seeds of the sister genus Schima and representatives of Polyspora by Tsou (1997, 1998)
confirm a different pattern of wing development for these two genera.
The circumscription of Schima s.l. is tentative given the large amount of missing data for Apterosperma. The major morphological differences between the two genera appear to be the number of staminal whorls, two in Apterosperma vs. three to five in Schima, and the wingless nature of the seed in Apterosperma. Our data places Apterosperma in Theeae with moderate to strong bootstrap support (Figs. 2 and 4). This finding does not conflict with developmental data of Tsou (1998)
and has been confirmed by rpl16 intron sequence data (Prince, 2000
). One brief comment should be made regarding species concepts in the genus Schima. Bloembergen (1952)
recognized a single "complex-polymorphous" species (Schima wallichii) with several geographically distinct components. The results presented here show longer internal branch lengths and higher bootstrap support for those branches than for other clades in which species concepts are less controversial. These data, along with morphological data from common garden experiments by one of the authors (C. R. Parks, unpublished data) provide support for the recognition of several distinct species of Schima.
Franklinia is a monotypic genus formerly known from the southeastern coastal plain of the United States. Franklinia alatamaha Marshall has been extirpated from the single known wild population site in Georgia, possibly due to overcollection in the late 1800s, yet remains in cultivation (primarily) in arboreta. Earlier literature synonymized Franklinia alatamaha with Gordonia lasiathus or recognized it as a distinct species of Gordonia (Gordonia pubescens Cavanilles or Gordonia franklini L'Héritier). There are a number of significant morphological differences between these two species including fruit dehiscence and shape (loculicidal and septicidal dehiscence of a globose fruit in Franklinia, loculicidal dehiscence only of an acuminate columnar fruit in Gordonia), seed wing morphology (perimeter in Franklinia vs. apical in Gordonia), and peduncle length (subsessile in Franklinia, peduncle several centimeters long in Gordonia). Molecular data clearly distinguishes Franklinia from Gordonia and places it as a closer relative of the Asian genus Schima with which it shares a similarly shaped fruit and seed.
General conclusions
The results of analyses of molecular data provide the basis for a tentative revised classification of Theaceae. The family as circumscribed by Cronquist (1981)
and Takhtajan (1997)
is likely paraphyletic. The results support the recognition of Ternstroemiaceae and Theaceae (sensu Mirbel, 1813
; APG, 1998
). The results support portions of Airy-Shaw's (1936)
and Tsou's (1998)
classifications but with some major differences as well. The overall topology argues for some shifts in rank of subtribes. With their classifications in mind, we propose a tentative classification for Theaceae of three tribes: Theeae Szyszylowicz including Camellia Linnaeus, Pyrenaria Blume, Polyspora Sweet ex. G. Don, Laplacea Kunth, and Apterosperma H. T. Chang; Gordonieae De Candolle with Gordonia Ellis, Franklinia W. Bartram ex H. Marshall, and Schima Reinwardt ex Blume; and Stewartieae Choisy with Stewartia Linnaeus (Table 2).
|
and Tsou (1998)
. Based on morphological and cytological data Keng considered Gordonieae the least specialized, while Tsou's cladistic analyses of ten morphological and developmental characters identified Camellieae (= Theeae) as the basalmost lineage. Stewartieae includes a single genus, Stewartia with (relatively) small seeds, copious endosperm, and a capsule that lacks a central columella. These plants are primarily Old World temperate trees and shrubs, with only two representatives in the eastern United States. While there is some support for the basal position of Stewartieae under parsimony criteria, there is no statistical significance for alternative tribal arrangements in Theaceae under likelihood criteria. At this time there is insufficient evidence to conclude Stewartieae are the basalmost lineage in the family, hence the recognition of three clades (tribes) of equal rank. The Gordonieae and Theeae clades include plants that have (relatively) larger seeds, reduced endosperm, and capsules with some distinctive features (columellae, locular wall slits) when compared with Stewartieae. Gordonieae includes Gordonia s.s. (plus some Laplacea), Franklinia, and Schima. Two of the genera (Gordonia and Franklinia) appear to be restricted to the New World warm temperate and subtropical regions, with the third genus (Schima) restricted to the Old World warm temperate to tropical regions.
The tribe Theeae encompasses ten times the species diversity (400+ species) as the other two tribes (30 species each) and includes Apterosperma, Camellia s.l., Pyrenaria s.l., Polyspora, and Laplacea pro parte. Theeae representatives are often shrubs or small trees, although some species of Polyspora are valuable timber trees in Malesia and Indonesia. Representatives are distributed in both the Old and New World but are especially diverse in Southeast Asia and Indonesia. Polyspora and Laplacea each include 15–20 species, but are far less diverse than Pyrenaria, with 35–40 species, and Camellia, with 300 species.
There are a number of reasons why Theeae might have significantly higher species diversity than the other two tribes. The more tropical distribution of Theeae may have allowed this tribe to survive historic climatic events of the Tertiary better than the more temperate Stewartieae, but that does not explain the lower diversity of Gordonieae. Theeae also have a high incidence of polyploidy, a condition not found in either of the other two tribes. Polyploidy has long been suggested as an important factor in the evolution and diversification of plants (e.g., Ehrlich, Holm, and Parnell, 1974
; Dobzhansky et al., 1977; Judd et al., 1999
; Wendel, 2000
). Theeae have the most diverse fruits and seeds of the family, ranging from dry capsules with flat, winged seeds (Polyspora and Laplacea) to somewhat fleshy walled indehiscent or dehiscent capsules with large angular seeds (Pyrenaria s.l. and Camellia). The added diversity of fleshy fruits and seeds may significantly improve dispersability for this tribe.
The nucleotide sequences analyzed in this study are not variable enough to confidently address relationships within Theeae. The authors are participating in a larger collaboration to collect additional information from a nuclear gene intron that will improve resolution in the largest tribe of Theaceae. The data do confirm the recognition of Theaceae and Ternstroemiaceae as separate families in the Order Ericales of the Angiosperm Phylogeny Group (APG, 1998
). The data also support the recognition of Franklinia as distinct although closely related to Gordonia. Finally, the data provide overwhelming evidence of the polyphyly of Gordonia s.l., supporting the recognition of three distinct genera: Gordonia, Laplacea, and Polyspora.
|
FOOTNOTES |
|---|
2 Author for reprint requests, current address: Rancho Santa Ana Botanic Garden, 1500 North College Ave., Claremont, CA 91711-3157 USA (linda.prince{at}cgu.edu
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Albert V. A. S. E. Williams M. W. Chase 1992 Carnivorous plants: phylogeny and structural evolution. Science 257: 1491-1495
Anderberg A. A. B. Ståhl M. Källersjö 1998 Phylogenetic relationships in the Primulales inferred from rbcL sequence data. Plant Systematics and Evolution 211: 93-102[CrossRef][ISI]
APG [Angiosperm Phylogeny Group]. 1998 An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85: 531-553[CrossRef][ISI]
Bloembergen S. 1952 A critical study in the complex-polymorphous genus Schima (Theaceae). Reinwardtia 2: 133-183
Bremer K. 1988 The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42: 795-803[CrossRef][ISI]
Chang H.-T. B. Bartholomew 1984 Camellias. Timber Press, Portland, Oregon, USA
Chase M. W. et al 1993 Phylogenies of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80: 528-580[CrossRef][ISI]
Cronquist A. 1981 An integrated system of classification of flowering plants. Columbia University Press, New York, New York, USA
Dahlgren R. M. T. 1983 General aspects of angiosperm evolution and macrosystematics. Nordic Journal of Botany 3: 119-149
Dobzhansky T. F. J. Ayala G. L. Stebbins J. W. Valentine 1977 Patterns of speciation in evolution. W. H. Freeman, San Francisco, USA
Donoghue M. J. R. G. Olmstead J. F. Smith J. D. Palmer 1992 Phylogenetic relationships of the Dipsacales based on rbcL sequences. Annals of the Missouri Botanical Garden 79: 333-345
Doyle J. J. J. L. Doyle 1987 A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11-15
Ehrlich P. R. R. W. Holm D. R. Parnell 1974 The process of evolution, 2nd ed. McGraw-Hill, New York, New York, USA
Eriksson T. 1998 AutoDecay, version 4.0 (program distributed by the author). Department of Botany, Stockholm University, Stockholm, Sweden
Felsenstein J. 1985 Confidence intervals on phylogenies: an approach using the bootstrap. Evolution 39: 783-791[CrossRef][ISI]
Felsenstein J. 1988 Phylogenies from molecular sequences: inference and reliability. Annual Review of Genetics 22: 521-565[CrossRef][ISI][Medline]
Fitch W. M. 1971 Toward defining the course of evolution: minimum change for a specific tree topology. Systematic Zoology 20: 406-416[CrossRef][ISI]
Goldberg A. 1986 Classification, evolution, and phylogeny of the families of dicotyledons. Smithsonian Institution Press, Washington, D.C., USA
Goldman N. J. P. Anderson A. G. Rodrigo 2000 Likelihood-based tests of topologies in phylogenetics. Systematic Biology 49: 652-670[CrossRef][ISI][Medline]
Hasegawa M. H. Kishino T. Yano 1985 Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 21: 160-174
Judd W. S. C. S. Campbell E. A. Kellogg P. F. Stevens 1999 Plant systematics: a phylogenetic approach. Sinauer, Sunderland, Massachusetts, USA
Jukes T. H. C. R. Cantor 1969 Evolution of protein molecules. In H. N. Munro [ed.], Mammalian protein metabolism, 21–132. Academic Press, New York, New York, USA
Keng H. 1962 Comparative morphological studies in Theaceae. University of California Publications in Botany 33: 269-384
Keng H. 1980 On the unification of Laplacea and Gordonia (Theaceae). Gardens' Bulletin (Singapore) 33: 303-311
Kimura M. 1980 A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16: 111-120[CrossRef][ISI][Medline]
Kron K. A. M. W. Chase 1993 Systematics of the Ericaceae, Empetraceae, Epacridaceae and related taxa based upon rbcL sequence data. Annals of the Missouri Botanical Garden 80: 735-741[CrossRef][ISI]
Lanave C. G. Preparata C. Saccone G. Serio 1984 A new method for calculating evolutionary substitution rates. Journal of Molecular Evolution 20: 86-93[CrossRef][ISI][Medline]
Lawrence G. H. M. 1951 Taxonomy of vascular plants. Macmillan, New York, New York, USA
Maddison D. R. 1991 The discovery and importance of multiple islands of most-parsimonious trees. Systematic Zoology 40: 315-328[CrossRef]
Melchior H. 1925 Theaceae. In A. Engler and E. Prantl [eds.], Die naturlichen Pflanzenfamilien, 2nd ed., vol. 21, 109–154. Wilhelm Engelmann, Leipzig, Germany
Melchior H. 1964 Theaceae. In A. Engler [ed.], Syllabus der Pflanzenfamilien, Band II, 166–168. Gebrüder Borntraeger, Berlin, Germany
Mirbel M. 1813 Notes pour servir à l'histoire naturelle de la famille des Orangers de M. A.-L. de Jussieu. Nouveau Bulletin des Sciences par la Société Philomatique 75: 377-383
Morgan D. R. D. E. Soltis 1993 Phylogenetic relationships among members of Saxifragaceae sensu lato based on rbcL sequence data. Annals of the Missouri Botanical Garden 80: 631-660[CrossRef][ISI]
Morton C. M. M. W. Chase K. A. Kron S. M. Swensen 1996 A molecular evaluation of the monophyly of the order Ebenales based upon rbcL sequence data. Systematic Botany 21: 567-586[CrossRef][ISI]
Morton C. M. K. G. Karol M. W. Chase 1997 Taxonomic affinities of Physena (Physenaceae) and Asteropeia (Theaceae). Botanical Review 63: 231-239
Morton C. M. S. A. Mori G. T. Prance K. G. Karol M. W. Chase 1997 Phylogenetic relationships of Lecythidaceae: a cladistic analysis using rbcL sequence and morphological data. American Journal of Botany 84: 530-540[Abstract]
Olmstead R. G. H. J. Michaels K. M. Scott J. D. Palmer 1992 Monophyly of the Asteridae and idenfication of their major lineages inferred from DNA sequences of rbcL. Annals of the Missouri Botanical Garden 79: 249-265
Prince L. M. 2000 Additional evidence of monophyly, paraphyly, and polyphyly in genera of Theaceae. American Journal of Botany 87: 178 (Abstract)
Rambaut A. 1996 Se-Al, Sequence alignment editor, version 1.0 alpha 1. Department of Zoology, Oxford University, Oxford, UK
RodrÍguez F. J. L. Oliver A. MarÍn J. R. Medina 1990 The general stochastic model of nucleotide substitution. Journal of Theoretical Biology 142: 485-501[ISI][Medline]
Sanger F. S. Nicklen A. R. Coulson 1977 DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences, USA 74: 5463-5467
Savolainen V. M. W. Chase S. B. Hoot C. M. Morton D. E. Soltis C. Bayer M. F. Fay A. Y. de Bruijn S. Sullivan Y.-L. Qiu 2000 Phylogenetics of flowering plants based on combined analysis of plastid atpB and rbcL gene sequences. Systematic Biology 49: 306-362[CrossRef][ISI][Medline]
Sealy J. R. 1958 A revision of the genus Camellia. Royal Horticultural Society, London, UK
Soltis D. E. P. S. Soltis M. E. Mort M. W. Chase V. Savolainen S. B. Hoot C. M. Morton 1998 Inferring complex phylogenies using parsimony: an empirical approach using three large DNA data sets for angiosperms. Systematic Biology 47: 32-42[CrossRef][ISI][Medline]
Soltis D. E. et al 2000 Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnean Society 133: 381-461[CrossRef]
Swofford D. L. 1998 PAUP*: phylogenetic analysis using parsimony (* and other methods), version 4.0. Sinauer, Sunderland, Massachusetts, USA
Swofford D. L. G. J. Olsen P. J. Waddell D. M. Hillis 1996 Chapter 11: phylogenetic inference. In D. M. Hillis, C. Moritz, and B. K. Mable [eds.], Molecular systematics, 2nd ed., 407–514. Sinauer, Sunderland, Massachusetts, USA
Takhtajan A. 1997 Diversity and classification of flowering plants. Columbia University Press, New York, New York, USA
Tavaré S. 1986 Some probabilistic and statistical problems on the analysis of DNA sequences. Lectures on Mathematics in the Life Sciences 17: 57-86
Taylor J. W. E. Swann 1994 DNA from herbarium specimens. In B. Herrmann and S. Hummel [eds.], Ancient DNA, 166–181. Springer-Verlag, New York, New York, USA
Thorne R. F. 1992 Classification and geography of the flowering plants. Botanical Review 58: 225-348
Tsou C.-H. 1997 Embryology of the Theaceae–anther and ovule development of Camellia, Franklinia, and Schima. American Journal of Botany 81: 369-381
Tsou C.-H. 1998 Early floral development of Camellioideae (Theaceae). American Journal of Botany 85: 1531-1547
Vogelstein B. D. Gillespie 1979 Preparative and analytical purification of DNA from agarose. Proceedings of the National Academy of Sciences, USA 76: 615-619
Wendel J. F. 2000 Genome evolution in polyploids. Plant Molecular Biology 42: 225-249[CrossRef][ISI][Medline]
Xiang Q.-Y. D. E. Soltis P. S. Soltis 1998 Phylogenetic relationships of Cornaceae and close relatives inferred from matK and rbcL sequences. American Journal of Botany 85: 285-297[Abstract]
Yang S.-X. 1998 Systematics, diversification, and geographical distribution of Pyrenaria sensu lato (Theaceae). Ph.D. dissertation, Kunming Institute of Botany, Yunnan, China
Ye C.-X. 1990 A discussion on relationships among the genera in Theoideae (Theaceae). Acta Scientarum Naturalium Universitatis Sunyatseni 29: 74-81
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
