| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
What's this? |
Systematics |
2Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3DS, UK; 3Section of Integrative Biology and Plant Resources Center, University of Texas, Austin, Texas 78712 USA; 4Ohio State University Herbarium, Department of Evolution, Ecology and Organismal Biology, Columbus, Ohio 43212 USA; 5The Lewis B. and Dorothy Cullman Program for Molecular Systematics Studies, The New York Botanical Garden, Bronx, New York 10458-5126 USA
Received for publication April 21, 2004. Accepted for publication January 12, 2005.
ABSTRACT
Phylogenetic relationships within the epidendroid orchids with emphasis on tribes Epidendreae and Arethuseae were assessed with parsimony and model-based analyses of individual and combined DNA sequence data from ITS nuclear ribosomal DNA and plastid trnL intron, the trnL-F spacer, matK (gene and spacers), and rbcL regions. Despite the absence of boostrap support for some of the relationships, a well-resolved and supported consensus was found, for which most clades were present in more than one individual analysis. Most clades of this consensus attained high posterior probabilities with a Bayesian approach. Circumscription of Arethuseae and Epidendreae are different from most orchid systems based on morphology, but they correspond to a combination of patterns from several less comprehensive orchid phylogenetic analyses previously published. A new circumscription of Epidendreae includes only Neotropical subtribes (Bletiinae, Chysiinae, Laeliinae, Ponerinae, and Pleurothallidinae), whereas Arethuseae include Coelogyninae (all Old World) and Arethusinae (pantropical). Many previously included genera will need to be moved to other tribes. Taxa previously assigned to be Old World Epidendreae are related to different groups of Old World orchids, and this study can serve as a guide for sampling strategies in future studies to resolve troublesome epidendroid orchid clades.
Key Words: Arethuseae • classification • Epidendreae • Epidendroideae • molecular phylogenetics • Orchidaceae
In recent years, at least one morphological (Freudenstein and Rasmussen, 1999
) and several molecular phylogenetic studies of Orchidaceae (Cameron et al., 1999
, rbcL; Freudenstein and Chase, 2001
, nad1b-c intron; Cameron and Chase, 2000
, 18S; Cameron, 2004
, psaB; Freudenstein et al., 2004
, rbcL and matK) have been published. Although with extensive sampling of lineages representative of the whole family, these studies were mostly based on single DNA regions, whereas several studies combining two or three DNA regions have been published for some tribes, subtribes, or subfamilies (e.g., Whitten et al., 2000
, Cymbidieae; Pridgeon et al., 2001
, Pleurothallidinae; Gravendeel et al., 2001
, Coelogyne and related genera; Kores et al., 2000
, Orchidoideae; Salazar et al., 2003
, Cranichideae). At the same time that the combined DNA region approach provided insights into many specific questions, their limited sampling did not allow a broader picture of orchid phylogenetic patterns. Conversely, in broad studies with one or two genes (Cameron et al., 1999
; Freudenstein et al., 2004
), the levels of variation permitted conclusions about subfamily delimitation and relationships but contributed little to the understanding at the tribal and subtribal levels.
Epidendreae were described by Kunth (1815)
and included at the time most tropical epiphytic orchids, as opposed to Orchideae, which contained terrestrial, mostly temperate genera (see Table 1 for an overview of classification systems for Epidendreae and Arethuseae). Subsequent systems such as Bentham and Hooker (1883)
and Schlechter (1926)
, kept this circumscription that included a large number of genera. This broad circumscription was maintained until the system of Dressler and Dodson (1960
, not presented). More recent systems such as Brieger et al. (1970–1984
, not presented), Dressler (1981
, 1993
; Table 1) and Szlachetko (1995
; Table 1), circumscribed subfamily Epidendroideae as equivalent to tribe Epidendreae in older systems and established tribe Epidendreae as a more narrowly circumscribed taxon, although with broad differences in the subtribes included. The position and circumscription of Arethuseae was also variable. Some systems restricted this tribe to members with soft pollinia and frequently included them in other subfamilies, such as Vanilloideae, and placed the genera with firm pollinia in several tribes of Epidendroideae. Later systems such as Dressler (1981
, 1993
) included a mixture of taxa with soft and hard pollinia, although Szlachetko (1995)
followed the older idea, placing taxa with soft pollinia in Arethuseae (Vanilloideae) and the ones with hard pollinia in Bletieae (Epidendroideae).
|
and Dressler (1981)
, or cladograms presenting putative ideas of phylogenetic relationships and character evolution (Dressler, 1993
; Szlachetko, 1995
).
Freudenstein and Rasmussen (1999)
presented the first morphological cladistic analysis of the whole of Orchidaceae, in which members of Epidendreae sensu Dressler (1993)
were placed in several different clades in the most parsimonious trees (but not in the strict consensus). Sirhookera (Adrorhizinae) and Jensoa (Calypsoeae) were successive sisters to a clade containing Dactylostalix (Calypsoeae), Pleurothallis (Pleurothallidinae), Isochilus (Laeliinae), Glomera (Glomerinae), Agrostophyllum, Ceratostylis, and Appendicula (Podochileae), Arpophyllum (Arpophyllinae), Meiracyllium (Meiracylliinae), and Epidendrum, Cattleya, and Schomburgkia (Laeliinae). The morphological analysis did not provide a clear distinction between Epidendreae and members of Calypsoeae and Podochileae.
Using rbcL sequence data, Chase et al. (1994)
provided the first attempt at a phylogenetic analysis of Epidendroideae, but Epidendreae were not represented (only 34 terminals were included). Neyland and Urbastch (1996)
provided an analysis based on another plastid gene, ndhF, but limited sampling (36 taxa) and low levels of sequence divergence at the tribal and subtribal level severely restricted reaching many conclusions. Some patterns appearing in the ndhF trees were not present in Cameron et al. (1999)
, e.g., the clade that resembled Epidendreae composed of the New World subtribes Arpophyllinae, Laeliinae, and Chysiinae. Bletiinae and Ponerinae were not sampled in Neyland and Urbastch (1996)
, but the clade sister to Epidendreae was composed of Coelia, Tipularia (Calypsoeae), and Dendrobium.
Using many more rbcL DNA sequences, Cameron et al. (1999)
performed the first broadly sampled molecular study of Orchidaceae. The levels of variation were enough to build a clear picture of the subfamilial relationships, but at the tribal and subtribal levels there was low resolution and little bootstrap support. The epidendroid orchids were divided into a "lower" grade and a set of "higher" clades. Members of Epidendreae and Arethuseae were placed in several different clades, although without high bootstrap support. Many of these relationships seemed artifactual and differed from those in Neyland and Urbatsch (1996)
. However, they should not be taken as clear phylogenetic hypotheses because the levels of variation and bootstrap support were low. Another phylogenetic study was based on mitochondrial DNA data of an intron within nad1 (Freudenstein et al., 2000
; Freudenstein and Chase, 2001
), which also showed a clade composed of Laeliinae, Pleurothallidinae, Bletia, and Calypsoeae, all of which would qualify as Epidendreae. However, as in rbcL, variation was low and consequently, only seven clades had jackknife support above 70%. Many groups were oddly placed relative to previous studies. Combining rbcL and matK (Freudenstein et al., 2004
) changed substantially the estimates of relationship within Epidendroideae obtained with rbcL alone, indicating that problems in the rbcL data set were due to sampling errors (i.e., too few variable sites). Both genes, however, were still not sufficient to resolve many of the polytomies within the subfamily.
It is difficult to discuss the placement of Epidendreae and related tribes without comparing them to Arethuseae, especially Bletiinae. Dressler (1993)
suggested that the ancestor of advanced Epidendroideae was part of this group. Arethuseae include taxa with a mosaic of gradual changes from soft, mealy pollinia (which other authors have considered members of Neottioideae) to those that are firmer and well defined. Dressler's (1993)
system used a restricted concept of Arethuseae, with only two subtribes, Arethusinae and Bletiinae. Although Goldman et al. (2001)
showed that Arethusinae if narrowly defined are monophyletic, Bletiinae were indicated to be a completely artificial assemblage of genera (this pattern also emerged in Cameron et al., 1999
). On the other hand, the rbcL and matK data used in Goldman et al. (2001)
were probably not variable enough, and there was not sufficient sampling of taxa in Epidendreae and Coelogyneae to evaluate whether any members of Arethuseae belong in these groups. For that reason, no formal taxonomic changes were presented in that work.
In this study, we aimed to delimit Epidendreae and Arethuseae by extensively sampling the putative component subtribes (based on the topologies of Neyland and Urbatsch, 1996
; Cameron et al., 1999
; van den Berg et al., 2000
; Freudenstein and Chase, 2001
; Goldman et al., 2001
; Freudenstein et al., 2004
) as well as genera throughout Epidendroideae. We used three lower epidendroids (Cephalanthera, Epipactis, and Listera) as outgroups, and all major clades of advanced epidendroids were represented in the matrix in an effort to determine whether any genera currently considered to be Epidendreae in fact have relationships elsewhere or vice versa. We were particularly interested to determine if Old World and New World subtribes/ genera that some authors (e.g., Dressler, 1993
) have ascribed to Epidendreae were intermixed.
MATERIALS AND METHODS
Plant materials
Plant material and voucher information for this analysis are given in Appendix 1 (see Supplemental Data accompanying online version of this article). The outgroups, Epipactis helleborine (L.) Crantz, Cephalanthera damasonium (Miller) Druce and Listera smallii Wiegand, were chosen from among the members of "lower" Epidendroideae observed in previous studies (Cameron et al., 1999
; Freudenstein and Chase, 2001
; Goldman et al., 2001
; Freudenstein et al., 2004
). Representatives of all other main clades of Epidendroideae were included. Within Epidendreae, sampling aimed for representation of all Old World and New World subtribes listed in Dressler (1993)
, with a larger sampling of subtribes richer in genera, such as Laeliinae and Pleurothallidinae. Selected subgroups within Laeliinae were also sampled because we had previous knowledge of genera that were placed outside the subtribe, such as Ponera, Isochilus, Helleriella (van den Berg et al., 2000
), Dilomilis (Cameron et al., 1999
; Freudenstein et al., 2000
; van den Berg et al., 2000
; Pridgeon et al., 2001
), and Basiphyllaea (Goldman et al., 2001
). In Epidendreae II (Old World taxa), we were unable to obtain material of Adrorhizon or Sirhookera; therefore, Adrorhizinae were not represented. Sobralia and Elleanthus were not included in the analysis because they have been shown to be distantly related to Epidendreae (Neyland and Urbatsch, 1996
; Cameron et al., 1999
; Goldman et al., 2001
).
DNA extraction, amplification, and sequencing
DNA was extracted mostly from fresh leaves, flowers, and silica-gel-dried leaves and flowers, using in most cases a modified version of the 2x CTAB procedure of Doyle and Doyle (1987)
. For samples that presented difficulties due to polysaccharides (e.g., Polystachya), DNA was extracted using the Nucleon Phytopure kit (Amersham Plc., Little Chalfont, UK). DNAs were purified either by cesium chloride/ethidium bromide gradients or QIAQuick silica columns (Qiagen, Ltd., Crawley, West Sussex, UK; following the manufacturer's protocols for PCR products) and sometimes by a combination of both methods. For amplification and sequencing of ITS, we used the primers 17SE and 26SE of Sun et al. (1994)
, and a PCR program consisting of 28 cycles of 94°C denaturation for 1 min, 50°C annealing for 1 min, and 72°C extension for 3 min, with 72°C for 7 min of final extension. To help reduce the effects of secondary structure on amplification and sequencing, 4% dimethyl sulfoxide was added to both the PCR and cycle sequencing mixes. For the trnL intron and trnL-F spacer (hereafter trnL-F), we used the four universal primers (c, d, e, and f) of Taberlet et al. (1991)
and a PCR program consisting of 28–30 cycles of 94°C denaturation for 1 min, 50°C annealing for 30 s, and 72°C of extension for 1 min. Most species were amplified and sequenced with primers c and f, but difficult samples had to be amplified in two, nonoverlapping halves (with primer pairs c, d and e, f) with the consequent insertion of missing characters (?) in the region corresponding to the primers d/e, which are direct complements. The matK region was amplified as a single piece, using the primers-19F (Molvray et al., 2000
) and trnK2R (Johnson and Soltis, 1994
). PCR conditions were a hot start with 2 min of initial denaturation at 94°C, followed by 28–30 cycles of 94°C denaturation, 52°C annealing for 45 s, and 72°C for an initial time of 2 min 30 s with auto-extension of 8 s per cycle. rbcL was amplified with the primers 1F and 1360R (Kores et al., 1997
) and the internal primers 636F and 724R (Muasya et al., 1998
). Purification of PCR products was performed with QIAquick (Qiagen, Ltd.) and Concert (Gibco BRL, Ltd., now Invitrogen, Paisley, UK) silica columns following the manufacturers' protocols. For ITS only, an extra wash was added with 35% guanidinium chloride solution to help remove primer dimers. PCR products were sequenced in both directions, using the Big Dye Terminator kit on an ABI 377 automated sequencer following manufacturer's protocols (Applied Biosystems, Inc., Warrington, Cheshire, UK). The following primers were used for sequencing in addition to the ones used for amplification: matK 163F, matK 458F, matK 556R, matK 877F (Molvray et al., 2000
), matK 731F, matK 881R (Pridgeon et al., 2001
), matK 1155F, and matK 1592R (Goldman et al., 2001
).
Data analysis
Electropherograms were edited and assembled using Sequencher 3.0 and 3.1 (Genecodes Inc., Ann Arbor, Michigan, USA), and the resulting sequences were aligned manually following the guidelines in Kelchner (2000)
. A manually coded binary gap matrix with all non-autapomorphic, unambiguous indels in the trnL-F and matK gene data sets was included. Gaps in the ITS data set were considered missing data due to the less obvious alignment of this region among reasonably distant taxa. The rbcL matrix used mostly sequences from Cameron et al. (1999)
but also many sequences produced for this paper. Analyses were performed using PAUP* (Phylogenetic analysis using parsimony (*and other methods), version 4 (Swofford, 1998
) with Fitch parsimony (equal weights, unordered characters; Fitch, 1971
) as the optimality criterion. Initially, we performed four separate searches, grouping regions roughly by category: (1) ITS, (2) plastid noncoding (trnL-F region and trnK introns bordering matK), (3) matK, and (4) rbcL. The last two were analyzed separately to allow a comparison between our rbcL results and those of Cameron et al. (1999)
. A fifth analysis included the data from all DNA regions. Each search consisted of 1000 random taxa-addition replicates of subtree-pruning-regrafting (SPR) with the number of trees limited to 10 per replicate to prevent extensive swapping on islands with many trees. The resulting trees were then used as starting trees for TBR (tree-bisection and reconnection) swapping (because this algorithm is sometimes able to reach shorter trees than SPR) with an upper limit of 10 000 trees. Internal support was evaluated using 1000 replicates of bootstrapping (Felsenstein, 1985
), with simple taxon-addition and the TBR algorithm, saving 10 trees per replicate. Bremer support (decay index) was calculated for the most parsimonious tree presented using Autodecay 5.0 (Eriksson, 2001
) and PAUP*. A model-based analysis was performed with Bayesian inference (Larget and Simon, 1999
; Lewis, 2001
), using MrBayes 3.01 (Ronquist and Huelsenbeck, 2003
). The model used was GTR + Inv + gamma, as indicated by nested likelihood-ratio tests using Modeltest 3.06 (Posada and Crandall, 1998
). MrBayes was run with four random-initiated chains for 200 000 generations, sampling trees every 10 generations. Trees were checked for stability, which happened at around 49 000 chains; therefore, the first 50 000 trees were discarded as the burn-in. The remaining trees (15 000) were used to assess topology and posterior probabilities in a majority-rule consensus. All sequences used in this paper have been submitted to GenBank (See Appendix for accession numbers. The data matrices are available upon request from CvdB and MWC.)
Combinability of the separate matrices was assessed by looking for incongruent groups with high bootstrap support (Whitten et al., 2000
; Reeves et al., 2001
). Differently placed groups with low support were taken to be the result of sampling error (too few data to get a clear answer). Furthermore, truly incongruent patterns should result in lower bootstrap support and decreased resolution for a given clade in the combined analysis relative to that in one or more of the separate analyses. In this study, no cases of such strongly supported incongruence were observed.
RESULTS
General features of the DNA regions used are presented in Table 2. The most variable data set was ITS (63.5% potentially informative sites), which also had the lowest consistency (CI) and retention (RI) indices (CI = 0.30 and RI = 0.47; ITS characters optimized on the combined tree). The trnL-F region, matK gene, and spacers bordering matK had similar variation (around 28%), with a CI between 0.51 (matK gene) and 0.61 (matK bordering spacers) and higher RIs than ITS (trnL-F region, 0.55; matK gene, 0.53; matK bordering spacers 0.61). The least variable data set was rbcL with only 11.1% potentially informative sites, CI = 0.48, and RI = 0.58. ITS sequences of Basiphyllaea corallicola and Thelasis carinata were unavailable because we were unable to amplify the ITS regions for these samples. The ITS sequence obtained for Angraecum magdalenae was excluded because two sequences were obtained that were highly divergent and more or less unalignable (this genus often produces paralogous sequences; Carlsward et al., 2003
).
|
, excluding Coelia. This clade was also present in the individual ITS analysis. The inclusion of Chysis in this clade did not obtain bootstrap support >50%, whereas Coelia was sister (with BP <50%) to Goveniinae (Fig. 1). Within Epidendreae, three main clades in addition to Chysis are present: Laeliinae (including Arpophyllum and Meiracyllium; BP 100), Pleurothallidinae (BP 98), Bletiinae sensu stricto (BP 100), and Ponerinae (BP 100), with the last two sister to each other (BP 77). These same clades were recovered also in the individual ITS, matK, and noncoding plastid analyses, although their relationships were unresolved. The clade sister to Epidendreae (Fig. 2; BP < 50) includes "the vandoid orchids," namely, Cymbidieae, Vandeae, Polystachyinae, Agrostophyllum, and Earina. Collectively sister to Epidendreae plus "vandoid orchids," there is a clade with two subgroups: Goveniinae (BP 50) plus Coelia (mostly NW; BP < 50) and a cluster of Old World groups (Podochileae, BP 53; Phaiinae, BP < 50; and Collabiinae, BP < 50). Sister to all clades previously mentioned is a group that includes Arethusinae (BP < 50) plus Coelogyninae (BP 74). Finally, just above the root of the tree Dendrobieae (BP 69) and Malaxideae (BP 100) were consecutively sister to all clades mentioned earlier (but with BP < 50). All these larger group relationships were unresolved in the individual analyses.
|
|
|
Molecular evolution
The levels of variation differed noticeably among the different DNA regions. As expected, the two ITS spacer regions, ITS1 and ITS2, have the greatest number of variable sites with the fastest rate of change (Table 2). These two regions have twice the number of variable sites as the coding 5.8S in relation to the total number of sites. The CI and RI of these regions are clearly lower than 5.8S and the plastid regions because the number of changes per site in ITS is three to four times higher than for the other DNA regions in this study. Increased taxon sampling might potentially increase the RI by recovering phylogenetic structure from more homoplasious characters. In all plastid regions except rbcL (less variable), there are fewer differences in the levels of variation among data sets. Values of CI and RI in coding and noncoding regions were comparable, suggesting little influence of alignment on these measures. For a noncoding region, the trnL-F region is not particularly variable (for a detailed discussion, see Bakker et al., 2000
), whereas the matK region has a large number of variable sites. Although Kores et al. (2000)
suggested that it might be a pseudogene because of the presence of indels not part of triplets and of internal stop codons, we found no internal stops and all indels were in triplets. First and second positions of matK were similarly variable, and there was an excess of third position substitutions, although by only 1.5 times (as opposed to around 4–5 times in rbcL and atpB; Savolainen et al., 2000
). The transition/transversion ratio (ts : tv) in matK was around 1.0, whereas in rbcL, it is 1.70, similar to other studies with plastid-coding genes (e.g., Savolainen et al., 2000
; rbcL = 1.65, atpB = 2.09). Despite these differences in the ts : tv, we can not conclude from these parameters that in these plants matK is a pseudogene.
Another question that arises here is the combinability of the four DNA data sets. Although several tests of combinability are described in the literature, they have been shown to be of limited value in detecting incongruence (Reeves et al., 2001
; Yoder et al., 2001
). Instead, we used a careful inspection of all supported groups in individual analyses, as suggested by Wiens (1998)
and Reeves et al. (2001)
. No well-supported incongruent patterns were found; incongruence between DNA data sets in orchids has been found thus far only within genera at the species level (such as Calopogon, Goldman et al., 2004
; Cattleya, van den Berg, 2000
; Cymbidium, van den Berg et al., 2002
). Undoubtedly, the data set with greatest contribution in resolving clades was the ITS, which can be explained by its greater number of variable sites and their higher rate of substitution. The topologies found in this study did not obtain high levels of bootstrap support; however, coincidence of patterns among the different DNA regions may serve as additional evidence for the patterns found.
Support in phylogenetic analyses of Orchidaceae
Although several recent studies (Cameron et al., 1999
; Cameron and Chase, 2000
; Freudenstein and Chase, 2001
; Freudenstein et al., 2003) have obtained good resolution of relationships among orchid subfamilies, a well-resolved tree within Epidendroideae remained elusive. Although the combination of two plastid data sets used by Freudenstein et al. (2004)
presented many more well-supported groups in relation to the rbcL data set of Cameron et al. (1999)
, Epidendroideae still contained a large polytomy with many small groups. In this paper, including a more rapidly evolving region (ITS) and also a rather variable plastid region (trnL-F intron and intergenic spacer; Bakker et al., 2000
) seems to have improved resolution, although parsimony analyses assessed with resampling methods still yielded many poorly supported clades, a result also confirmed by low Bremer support (decay) for several clades. The Bayesian estimate produced a well-resolved tree with high posteriors. It is clear that boostrap percentages and posteriors are not equivalent measures. However, our example seems to illustrate how the Bayesian approach appears to be more efficient in dealing with data sets with few substitutions. Alfaro et al. (2003)
found in simulation studies that Markov chain Monte Carlo posteriors are generally less-biased estimators of phylogenetic support than nonparametric bootstrap under the optimality criteria of either maximum parsimony or maximum likelihood. Considering that no well-supported group in our parsimony analysis was contradicted in the Bayesian analysis, we believe that the latter indicates many more patterns with some level of support. However, caution should be used in interpreting Bayesian probabilities to avoid giving them the same status as resampling methods of support, because they have been shown to be over-estimates of confidence (Huelsenbeck et al., 2002
; Suzuki et al., 2002
).
Comparison with previous orchid phylogenetic analyses
As an overall comparison, our combined analysis was in many of the clades identified at odds with the large rbcL study of Cameron et al. (1999)
and is more similar to the ndhF topologies of Neyland and Urbatsch (1996)
, nad1b-c intron (Freudenstein et al., 2000
; Freudenstein and Chase, 2001
) and combined rbcL and matK (Freudenstein et al., 2004
). Although this is surprising considering that rbcL had a much more extensive sampling of taxa, it could be easily explained by the fact that ndhF, nad1 intron, and matK combined included more variable positions than rbcL. Epidendreae composed of Pleurothallidinae, Laeliinae (including Meiracyllium), and Chysiinae and probably a relationship of this clade to Coelia and Calypsoeae were already clear in the ndhF maximum likelihood and parsimony tree from the nad1 intron: in the rbcL + matK study, these clades were unresolved. Comparing these studies with ours shows clearly that the level of variation of rbcL was only satisfactory to resolve the subfamilies of orchids and relationships in groups that are supposed to have diverged early (e.g., Vanilloideae and lower epidendroids). In groups such as Epidendroideae, the number of variable positions was low, and many unlikely placements were undoubtedly due to character sampling error. A good example was the position of Polystachya, which was embedded in Laeliinae and which in turn was sister to Vandeae; likewise the position of Pleurothallidinae with Coelia, Acanthephippium, and Calanthe seemed unlikely on morphological grounds. It is clear from the comparisons in our study that the most efficient strategy to get more resolved topologies within Epidendroideae is to collect additional noncoding or more variable plastid regions rather than genes like rbcL or atpB. To improve the results from analysis of ITS would require an even more extensive sampling of taxa across Epidendroideae, due to the large number of changes per site and improvement of alignments made possible by the inclusion of intercalating taxa.
Delimitation of tribe Epidendreae
Table 3 summarizes the circumscriptions found in this study in relation to the systems of Dressler (1981
, 1993
). From our results, Epidendreae appear to be composed exclusively of New World subtribes: Laeliinae, Pleurothallidinae, Ponerinae, Bletiinae, and Chysinae. This is the same set of subtribes of Epidendreae sensu Szlachetko (1995)
, although in his system Arpophyllum was placed in another tribe. Subtribal circumscriptions also differ broadly from this author. If Epidendreae were to include the Old World tribes included there by Dressler, then the tribe would have to include as well the vandoid orchids (Vandeae and Cymbidieae), which would expand them beyond a reasonable level. Coeliinae (sole genus Coelia) were not included in Epidendreae in either the ITS alone and combined analyses; Coelia was placed with Calypsoeae (sister to Goveniinae). However, there was no support greater than BP 50 for this relationship, and the Bayesian analysis supported Coelia as sister to Epidendreae. So far, the positions of Calypso and Govenia/Aplectrum have been inconsistent, and it is not clear whether Coelia belongs with them or with Epidendreae. It is also possible that Calypsoeae + Coelia could also be a clade sister to Epidendreae, a pattern that appears in a combined analysis of the plastid data (results not shown), however, with BP under 50%.
|
and Pridgeon et al. (2001)
. Both represent small clades with few species sister to larger, species-rich subtribes. Nevertheless, both Arpophyllum and Dilomilis/ Neocogniauxia are branch-length divergent relative to the larger groups, and they deviate from the morphological patterns that characterize their sister clades. For example, Dilomilis/ Neocogniauxia lack some of the synapomorphies that mark Pleurothallidinae sensu Dressler (1993
; e.g., articulation between ovary and pedicel). In the case of Arpophyllum, plant habit is similar to some Laeliinae (e.g., Laelia, Cattleya), but also to Octomeria (Pleurothallidinae). Its flower morphology is unique. Baker (1972)
found that the foliar anatomy of Arpophyllum is different from Laeliinae, and thus this similar external morphology could be interpreted as a parallelism.
The generic composition of each subtribe has to be redefined in some cases, especially for Laeliinae and Bletiinae. Both ITS and the combined analyses gave a clear separation for all subtribes within Epidendreae. Laeliinae should include all the genera listed in Dressler (1993)
, except Isochilus, Ponera, and Helleriella (which should be included in Ponerinae), Dilomilis and Neocogniauxia (Pleurothallidinae), and Basiphyllaea (Bletiinae). Ponerinae are a resurrected subtribe with a new circumscription presented here. Although the name has been used in many systems (Schlechter, 1926
; Brieger et al., 1970–1984
; Szlachetko, 1995
), it always included genera with a column foot, which otherwise had been placed in Laeliinae. This study indicates that Ponerinae are a small subtribe including only Ponera, Isochilus, and Helleriella. All other genera previously included in Ponerinae (i. e., the Scaphyglottis alliance and Jacquiniella) are part of Laeliinae. Scaphyglottis punctulata, which was once transferred to Helleriella (Garay and Sweet, 1974
), should be retained in Scaphyglottis. Bletiinae are part of Epidendreae rather than Dressler's (1993)
Arethuseae and here comprise a small subtribe with only three New World genera: Bletia, Hexalectris, and Basiphyllaea. All other genera placed in Bletiinae in Dressler (1993)
and other previous systems belong to different clades in Epidendroideae, mainly in Coelogyninae and Collabieae (Goldman et al., 2001
). Ponerinae and Bletiinae are sister clades. They both have a column foot but differ in that Ponerinae have a reed-stem habit, whereas Bletiinae have corms. Meiracylliinae, placed outside Laeliinae in the morphological analysis of Freudenstein and Rasmussen (1999)
and the classification of Brieger et al. (1970–1984
; Podochileae) and Dressler (1971
, 1981
, 1993
), is deeply embedded in Laeliinae. Its unusual column structure is autapomorphic.
All members of the group defined in Dressler (1993)
as Epidendreae II (Old World: Polystachyinae, Glomerinae, and probably Adrorhizinae) are related to other tribes of mainly Old World distribution. These will be discussed together with the remainder of Epidendroideae next.
There are almost no exclusive characters to define Epidendreae from a morphological standpoint. The tribe is excessively variable, especially within subtribes Pleurothallidinae and Laeliinae. Furthermore, many of the character states found within these subtribes are found in other Old World genera in several different subtribes. These are probably remarkable vegetative and floral convergences that explain the broadly different circumscriptions used for Epidendreae. Most characters previously used, such as presence and absence of a column foot, reed-stem habit, and pollinarium shapes appear to be too variable, with limited value even in defining the subtribes within Epidendreae. It may be nearly impossible to establish one synapomorphy for the tribe, although some trends occur. The Epidendrum seed type appears in Coeliinae and Laeliinae (Dressler, 1993
), whereas the Pleurothallis seed type occurs only in Pleurothallidinae. The Pleurothallis type, however, differs from the Epidendrum type mostly in having small warts that are easily soluble (Rauh et al., 1975
); the testa cells of the two types are otherwise similar. Ideally, it would be worthwhile breaking these complex "types" into their component characters, but this would require a great deal more work. A further complication is that the data on seed types is largely unpublished and is only summarized in Dressler (1993)
. Although these seed types appear to be exclusive to Epidendreae, the Elleanthus seed type that occurs in Arpophyllum, Meiracyllium, and also in Chysiinae, occurs in other tribes of Epidendroideae as well. Velamen types also seem to be characteristic in Arpophyllinae, Laeliinae, and Pleurothallidinae, whereas the Calanthe type (considered plesiomorphic in Orchidaceae; Dressler, 1993
) occurs in some Bletiinae. The distribution of seed and velamen type variation could be an indication that anatomical characters may ultimately provide an opportunity to discover synapomorphies in groups, such as Epidendreae, for which macromorphological characters are otherwise too homoplastic. However, most previous anatomical studies did not look at the key taxa that are now indicated by the new phylogenetic hypotheses based on DNA data, and new studies are necessary to include these new taxa and refine the scoring of character states.
Delimitation of Arethuseae
From our results, this tribe also has a distinct circumscription in relation to both systems of Dressler (1981
, 1993
); it should now include only Arethusinae (as in Dressler, 1993
) and Coelogyninae (previously part of tribe Coelogyneae in Dressler, 1993
). Calopogon and Arundina are part of Arethusinae instead of Bletiinae, whereas Bletilla (formerly Bletiinae), Dilochia (formerly Arundininae, subtribe unplaced) and Glomera (formerly Glomerinae, Epidendreae) are included in Coelogyninae. Dressler (1993)
had already suggested that his tribe Coelogyneae was related to Arethuseae, having in common some morphological characters such as an emergent, clam shell-shaped stigma. In the circumscription presented here, Arethuseae are a chiefly Old World tribe, except for the North American genera Calopogon and Arethusa, the latter with a disjunct relationship to Japanese Eleorchis. In relation to Dressler's (1981
, 1993
) delimitation, Arethuseae no longer include Bletiinae, the main subtribe in terms of number of genera (the type genus Bletia and a few of the genera placed there are part of Epidendreae in our analyses). Most of the Old World genera placed in Bletiinae by Dressler (1993)
that are not part of Coelogyninae are here placed with Collabiinae and Phaiinae (collectively Collabieae), and these two subtribes do not appear to be related to Arethuseae (they do not fall together in the parsimony tree and are relatively distant in the Bayesian results, although this low level of support does not refute them having a relationship). The inclusion of Glomera in Coelogyninae, which first occured in the combined matK and rbcL analysis of Goldman et al. (2001)
, is a circumscription that occurred in Bentham's (1881)
system, but no one has used this subsequently. Because our original Glomera sample (the same used by Cameron et al., 1999
) was not identified to species, we sequenced another member of the genus (G. pulchra) to confirm its placement. Dressler (1993)
had placed Glomera and many other Old World genera, such as Earina and Agrostophyllum, in Glomerinae (Epidendreae). In our results, Glomerinae have to be considered a nomenclatural synonym of Coelogyninae, whereas Agrostophyllum and Earina, if considered a distinct subtribe, have to be called Agrostophyllinae (a subtribe proposed by Szlachetko, 1995
); they appear to belong to the "vandoid clade." The overall results we found agree with subgroups found by Goldman et al. (2001)
. However, with a larger sampling of Epidendroideae in this study, the polyphyletic nature of Arethuseae in Dressler's (1981
, 1993
) systems can be fully assessed, and the genera previously placed in Arethuseae can now be assigned more confidently to other tribes and subtribes. Reinterpretation of morphological characters previously assigned to Arethuseae and Bletiinae in the literature is troublesome because most studies are focused in taxa that belong now in other subtribes. As previously mentioned for Epidendreae, many anatomical patterns will need to be reassessed afresh, considering that members previously placed in Bletiinae are now in three different clades (Bletiinae s. s., Collabiinae and Coelogyninae).
Other relationships within Epidendroideae
The relationships of the remaining Epidendroideae were more difficult to infer with our data due to limited taxonomic sampling. Many patterns are just beginning to emerge because of the greater numbers of data in this study compared with previous work. The clade that we have called the "vandoid orchids" appeared in the ITS, plastid noncoding, parsimony combined, and Bayesian combined analyses and is of special interest because it includes most members of subfamily Vandoideae of Dressler (1981)
, which he abandoned in his subsequent (1993) classification. Our data indicate that it is a natural group and should include Cymbidieae, Vandeae (which include Polystachyinae), and Agrostophyllinae. Preliminary results also indicate that Adrorhizinae and Bromheadiinae fall with the vandoid clade (S. Fernando and M. Chase, unpublished data). This pattern clarifies the weakly supported results of Cameron et al. (1999)
, which placed Polystachya near Laeliinae, probably as consequence of sampling error. Polystachyinae being sister to Vandeae and then this clade sister to Cymbidieae is in complete agreement with the diagram presented in Dressler (1981
, p. 155). All taxa in this vandoid clade have complex pollinaria with well-developed stipes, caudicles, and viscidia. Their columns are also similarly complex with arms and wings of various forms. Although this is a natural group, the levels of differentiation and embedded position in Epidendroideae demonstrate clearly that treating them as a separate subfamily, Vandoideae (Dressler, 1981
, non-1993), is inappropriate considering that Epidendroideae would be rendered paraphyletic. Although Agrostophyllinae were themselves a well-supported group (Earina and Agrostophyllum, part of Glomerinae in Dressler, 1993
), they were placed without support as sister to Vandeae plus Cymbidieae in both combined and ITS analyses, whereas Polystachyinae were sister to Agrostophyllinae in the noncoding plastid analysis. Despite the lack of bootstrap support above 50% in the combined analyses for the whole vandoid clade, it attained PP 100 in the Bayesian analysis, and these coincidences in individual and combined analyses point to it being a natural group. Morphology also supports this relationship, with elaters being a common feature in Agrostophylum, Earina, and Polystachya (Hallé, 1986
). Some genera formerly placed in Glomerinae such as Aglossorhyncha, Glossorhyncha, Ischnocentrum, and Sepalosiphon have yet to be sequenced, and their placements need to be clarified.
Another clade that appeared in most analyses is Podochileae, including Podochilinae and Thelasinae (BP 53, PP 100). If Eriinae are to be recognized as a subtribe separate from Podochilinae, greater sampling is required; they are clearly related (BP 93). Ridleyella (Ridleyellinae) was placed sister to this clade in the combined analysis but was sister of Thelasiinae in the plastid noncoding, matK, and rbcL analyses. We were unable to amplify ITS from Ridleyella. Ridleyellinae was created initially by Brieger et al. (1970–1984)
for Ridleyella, considered to be a member of Thelasiinae by Dressler (1981)
and again placed in a separate subtribe in Dressler's (1993)
system; although in the text he mentioned, it probably was part of Thelasiinae. The clade in our parsimony combined analysis that is sister to Podochileae (BP < 50) includes Acanthephippium, Collabiinae (Collabium, Nephelaphyllum, and Ancistrochilus) and Phaiinae (Phaius and Calanthe, which may be congeneric); collectively these could be treated as Collabieae. The Bayesian tree included in this group Malaxideae (PP 100) as sister to this expanded version of Collabieae. Further studies are necessary to assess relationships between Collabieae and Podochileae, decide whether Collabieae includes one or two subtribes, and determine where Acanthephippium and Malaxideae belong.
Finally, Dendrobieae, Malaxidae, and Nervilia were successively sister to all Epidendroideae in the parsimony combined analysis, whereas Dendrobiinae were sister to Vandeae in the Bayesian analysis (PP 55), but they will not be discussed due to the low level of sampling used here. As a general conclusion of this study, Epidendreae and Arethuseae can now be considered to be clearly delimited from remaining Epidendroideae. We also obtained a glimpse of the other relationships within Epidendroideae, which were not possible with previous analyses of single DNA regions. Clearly, the increased number of variable sites of plastid spacers and the more rapid rate of change in the variable positions in ITS increased significantly both the resolution and boostrap support for many groups. Increased taxon sampling and more DNA regions with more variable positions will be necessary to achieve a well-supported estimate of relationships among the different clades of Epidendroideae.
FOOTNOTES
1 The authors thank Paul Kores, Mia Molvray, W. Mark Whitten, and Norris Williams for access to sequence data. This research was supported by Conselho Nacional de Pesquisas, CNPq, Brazil, The American Orchid Society, and the trustees of the Royal Botanic Gardens, Kew. 
6 Author for correspondence (e-mail: vandenberg{at}uefs.br
). Current address: Laboratório de Sistemática Molecular, Departamento de Ciências Biológicas, Universidade Estadual de Feira de Santana, BR116 Km3, 44031-460, Feira de Santana, Bahia, Brazil
7 Current address: Harvard University Herbaria, 22 Divinity Avenue, Cambridge, Massachusetts 02138 USA
LITERATURE CITED
Alfaro M. E. S. Zoller F. Lutzoni 2003 Bayes or bootstrap? A simulation study comparing the performance of Bayesian Markov chain Monte Carlo sampling and bootstrapping in assessing phylogenetic confidence. Molecular Biology and Evolution 20: 255-266
Baker R. K. 1972 Foliar anatomy of the Laeliinae (Orchidaceae). Ph.D. dissertation, Washington University, St. Louis, Missouri, USA
Bakker F. T. A. Culham R. Gomez-Martinez J. Carvalho J. Compton R. Dawtrey M. Gibby 2000 Patterns of nucleotide substitution in angiosperm cpDNA trnL(UAA)-trnF(GAA) regions. Molecular Biology and Evolution 17: 1146-1155
Bentham G. 1881 Notes on Orchideae. The Journal of the Linnean Society 18: 281-360
Bentham G. J. D. Hooker 1883 Genera plantarum (Orchidaceae), vol. 3, part 2. L. Reeve & Co., London, UK
Brieger F. G. R. Maatsch K. Senghas 1970–1984 Die Orchideen. Paul Parey, Berlin, Germany
Cameron K. M. 2004 Utility of plastid psaB gene sequences for investigating intrafamilial relationships within Orchidaceae. Molecular Phylogenetics and Evolution 31: 1157-1180[CrossRef][Web of Science][Medline]
Cameron K. M. M. W. Chase 2000 Nuclear 18S rDNA sequences of Orchidaceae confirm the subfamilial status and circumscription of Vanilloideae. In K. L. Wilson and D. A. Morrison [eds.], Monocots: systematics and evolution, 457–464. CSIRO Publishing, Melbourne, Australia
Cameron K. M. M. W. Chase W. M. Whitten P. J. Kores D. C. Jarrell V. A. Albert T. Yukawa H. G. Hills D. H. Goldman 1999 A phylogenetic analysis of the Orchidaceae: evidence from rbcL nucleotide sequences. American Journal of Botany 86: 208-224
Carlsward B. S. W. M. Whitten N. H. Williams 2003 Molecular phylogenetics of Neotropical leafless Angraecinae (Orchidaceae): reevaluation of generic concepts. International Journal of Plant Sciences 164: 43-51[CrossRef][Web of Science]
Chase M. W. K. G. Cameron H. G. Hills D. Jarrell 1994 Molecular systematics of the Orchidaceae and other lilioid monocots. In A. M. Pridgeon [ed.], Proceedings of the 14th World Orchid Conference, 61–73. HMSO, Glasgow UK
Doyle J. J. J. L. Doyle 1987 A rapid DNA isolation method for small quantities of fresh tissues. Phytochemical Bulletin of the Botanical Society of America 19: 11-15
Dressler R. L. 1960 The relationships of Meiracyllium (Orchidaceae). Brittonia 12: 222-225[CrossRef]
Dressler R. L. 1971 Nomenclatural notes on the Orchidaceae, V. Phytologia 21: 440-443
Dressler R. L. 1979 The subfamilies of Orchidaceae. Selbyana 5: 197-206
Dressler R. L. 1981 The orchids: natural history and classification. Harvard University Press, Cambridge, Massachusetts, USA
Dressler R. L. 1993 Phylogeny and classification of the orchid family. Dioscorides Press, Portland, Oregon, USA
Dressler R. L. C. H. Dodson 1960 Classification and phylogeny in the Orchidaceae. Annals of the Missouri Botanical Garden 47: 25-68[CrossRef]
Eriksson T. 2001 Autodecay, version 5.0. Software program distributed by the author. Bergius Foundation, Royal Swedish Academy of Sciences, Stockholm, Sweden
Felsenstein J. 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791[CrossRef][Web of Science]
Fitch W. M. 1971 Towards defining the course of evolution: minimum change for a specific tree topology. Systematic Zoology 20: 406-416[CrossRef][Web of Science]
Freudenstein J. V. M. W. Chase 2001 Analysis of mitochondrial nad1b-c intron sequences in Orchidaceae: utility of length-change characters. Systematic Botany 26: 643-657[Web of Science]
Freudenstein J. V. F. N. Rasmussen 1999 What does morphology tell us about orchid relationships? A cladistic analysis. American Journal of Botany 86: 225-248
Freudenstein J. V. D. M. Senyo M. W. Chase 2000 Mitochondrial DNA and relationships in the Orchidaceae. In K. L. Wilson and D. A. Morrison [eds.], Monocots: systematics and evolution, 421–429. CSIRO Publishing, Melbourne, Australia
Freudenstein J. V. C. van den Berg D. H. Goldman P. J. Kores M. Molvray M. W. Chase 2004 An expanded plastid DNA phylogenetic analysis of Orchidaceae and analysis of jackknife clade support strategy. American Jounral of Botany 91: 149-157[CrossRef]
Garay L. A. H. R. Sweet 1974 Flora of the Lesser Antilles: Orchidaceae. Harvard University Press, Jamaica Plain, Massachusetts, USA
Goldman D. H. J. V. Freudenstein P. J. Kores M. Molvray D. C. Jarrell W. M. Whitten K. M. Cameron R. K. Jansen M. W. Chase 2001 Phylogenetics of Arethuseae (Orchidaceae) based on plastid matK and rbcL sequences. Systematic Botany 26: 670-695[Web of Science]
Goldman D. H. R. K. Jansen C. Van den Berg I. J. leitch M. F. Fay M. W. Chase 2004 Molecular and cytological examination of Calopogon (Orchidaceae, Epidendroideae): Circumscription, phylogeny, polyploidy and potential hybrid speciation. American Journal of Botany 91: 707-723
Gravendeel B. M. W. Chase E. F. de Vogel M. C. Roos T. H. M. Mes K. Bachmann 2001 Molecular phylogeny of Coelogyne (Epidendroideae; Orchidaceae) based on plastid RFLPs, matK, and nuclear ribosomal ITS sequences: evidence for polyphyly. American Journal of Botany 88: 1915-1927
Hallé N. 1986 Les èlatéres des Sarcanthinae et additions aux Orchidaceae de la Nouvelle-Caledonie. Bulletin du Muséum National d'Histoire Naturelle, Paris, 4e, series 8, section B,. Adansonia 3: 215-239
Huelsenbeck J. P. B. Larget R. E. Miller F. Ronquist 2002 Potential applications and pitfalls of Bayesian inference in phylogeny. Systematic Biology 51: 673-688[CrossRef][Web of Science][Medline]
Johnson L. A. D. E. Soltis 1994 matK DNA sequences and phylogenetic reconstruction in Saxifragaceae s.s. Systematic Botany 19: 143-156
Kores P. J. K. M. Cameron M. Molvray M. W. Chase 1997 The phylogenetic relationships of Orchidoideae and Spiranthoideae (Orchidaceae) as inferred from rbcL plastid sequences. Lindleyana 12: 1-11
Kelchner S. A. 2000 The evolution of noncoding chloroplast DNA and its application in plant systematics. Annals of Missouri Botanical Gardens 87: 482-498[CrossRef]
Kores P. J. P. H. Weston M. Molvray M. W. Chase 2000 Phylogenetic relationships within the Diurideae (Orchidaceae): inferences from plastid matK DNA sequences. In K. L. Wilson and D. A. Morrison [eds.], Monocots: systematics and evolution, 449–456. CSIRO Publishing, Melbourne, Australia
Kunth K. S. 1815 Nova genera et species plantarum. Librairie Graeco-Latino-Germanicae, Paris, France
Larget B. D. Simon 1999 Markov chain Monte Carlo algorithms for the Bayesian analysis of phylogenetic trees. Molecular Biology and Evolution 16: 750-759[Web of Science]
Lewis P. O. 2001 Phylogenetic systematics turns over a new leaf. Trends in Ecology and Evolution 16: 30-37
Mansfeld R. 1937 Über das System der Orchidaceae-Monandrae. Notizblatt des Botanischen Gartens und Museums zu Berlin-Dahlem 13: 666-676[CrossRef]
Molvray M. P. J. Kores M. W. Chase 2000 Polyphyly of mycoheterotrophic orchids and functional influences on floral and molecular characters. In K. L. Wilson and D. A. Morrison [eds.], Monocots: systematics and evolution, 441–448. CSIRO Publishing, Melbourne, Australia
Muasya A. M. D. A. Simpson A. Culham M. W. Chase 1998 An assessment of suprageneric phylogeny in Cyperaceae using rbcL sequences. Plant Systematics and Evolution 211: 257-271[CrossRef][Web of Science]
Neyland R. L. E. Urbatsch 1996 Phylogeny of subfamily Epidendroideae (Orchidaceae) inferred from ndhF chloroplast gene sequences. American Journal of Botany 83: 1195-1206[CrossRef][Web of Science]
Posada D. K. Crandall 1998 Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817-818
Pridgeon A. M. R. Solano M. W. Chase 2001 Phylogenetic relationships in Pleurothallidinae (Orchidaceae): combined evidence from nuclear and plastid DNA sequences. American Journal of Botany 88: 2286-2308
Rauh W. W. Barthlott N. Ehler 1975 Morphologie und Funktion der Testa staubförmiger Flugsamen. Botanisches Jahrbücher fur Systematik, Pfanzengeschichte und Pflanzengeographie 96: 353-374
Reeves G. M. W. Chase P. Goldblatt T. de Chies B. Lejeune M. F. Fay A. V. Cox P. J. Rudall 2001 A phylogenetic analysis of Iridaceae based on four plastid sequences regions: trnL intron, trnL-F spacer, rps4 and rbcL. American Journal of Botany 88: 2074-2087
Ronquist F. J. P. Huelsenbeck 2003 MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572-1574
Salazar G. A. M. W. Chase M. A. Soto Arenas M. Ingrouille 2003 Phylogenetics of Cranichidae with emphasis on Spiranthinae (Orchidaceae, Orchidoideae): evidence from plastid and nuclear DNA sequences. American Journal of Botany 90: 777-795
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][Web of Science][Medline]
Schlechter R. 1926 Das System der Orchidaceae. Notizblatt des Botanischen Gartens und Museums zu Berlin-Dahlem 9: 563-591[CrossRef]
Sun Y. D. Z. Skinner G. H. Liang S. H. Hulbert 1994 Phylogenetic analysis of Sorghum and related taxa using internal transcribed spacers of nuclear ribosomal DNA. Theoretical and Applied Genetics 89: 26-32[Web of Science]
Suzuki Y. G. V. Glazko M. Nei 2002 Overcredibility of molecular phylogenies obtained by Bayesian phylogenetics. Proceedings of the National Academy of Sciences, USA 99: 16138-16143
Swofford D. L. 1998 PAUP*: Phylogenetic analysis using parsimony (*and other methods), version 4. Sinauer Associates, Sunderland, Massachusetts, USA
Szlachetko D. L. 1995 Systema Orchidalium. Fragmenta Floristica et Geobotanica (Supplement) 3 1-152
Taberlet P. L. Gielly G. Pautou J. Bouvet 1991 Universal primers for amplification of three noncoding regions of chloroplast DNA. Plant Molecular Biology 17: 1105-1109[CrossRef][Web of Science][Medline]
van den Berg C. 2000 Molecular phylogenetics of Epidendreae with emphasis on subtribe Laeliinae (Orchidaceae). Ph.D. dissertation, University of Reading, Reading, UK
van den Berg C. A. Ryan P. J. Cribb M. W. Chase 2002 Molecular phylogenetic analysis of Cymbidium Sw. (Orchidaceae: Maxillarieae): sequence data from internal transcribed spacers (ITS) of nuclear ribosomal DNA and plastid matK. Lindleyana 17: 102-111
van den Berg C. W. E. Higgins R. L. Dressler W. M. Whitten M. A. Soto Arenas A. Culham M. W. Chase 2000 A phylogenetic analysis of Laeliinae (Orchidaceae) based on sequence data from internal transcribed spacers (ITS) of nuclear ribosomal DNA. Lindleyana 15: 96-114
Whitten W. M. N. H. Williams M. W. Chase 2000 Subtribal and generic relationship of Maxillarieae (Orchidaceae) with emphasis on Stanhopeinae: combined molecular evidence. American Journal of Botany 87: 1842-1856
Wiens J. J. 1998 Combining data sets with different phylogenetic histories. Systematic Biology 47: 568-581[CrossRef][Web of Science][Medline]
Yoder A. D. J. A. Irwin B. A. Payseur 2001 Failure of the ILD to determine data combinability for slow loris phylogeny. Systematic Biology 50: 408-424[Web of Science][Medline]
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  C. van den Berg, W. E. Higgins, R. L. Dressler, W. M. Whitten, M. A. Soto-Arenas, and M. W. Chase A phylogenetic study of Laeliinae (Orchidaceae) based on combined nuclear and plastid DNA sequences Ann. Bot., August 1, 2009; 104(3): 417 - 430. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. G. Conran, J. M. Bannister, and D. E. Lee Earliest orchid macrofossils: Early Miocene Dendrobium and Earina (Orchidaceae: Epidendroideae) from New Zealand Am. J. Botany, February 1, 2009; 96(2): 466 - 474. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. L. Davies and M. Stpiczynska Labellar Micromorphology of Two Euglossine-pollinated Orchid Genera; Scuticaria Lindl. and Dichaea Lindl. Ann. Bot., November 1, 2008; 102(5): 805 - 824. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Figueroa, G. A. Salazar, H. A. Zavaleta, and E. M. Engleman Root Character Evolution and Systematics in Cranichidinae, Prescottiinae and Spiranthinae (Orchidaceae, Cranichideae) Ann. Bot., March 1, 2008; 101(4): 509 - 520. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. M. Whitten, M. A. Blanco, N. H. Williams, S. Koehler, G. Carnevali, R. B. Singer, L. Endara, and K. M. Neubig Molecular phylogenetics of Maxillaria and related genera (Orchidaceae: Cymbidieae) based on combined molecular data sets Am. J. Botany, November 1, 2007; 94(11): 1860 - 1889. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. S. Carlsward, W. M. Whitten, N. H. Williams, and B. Bytebier Molecular phylogenetics of Vandeae (Orchidaceae) and the evolution of leaflessness Am. J. Botany, May 1, 2006; 93(5): 770 - 786. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
