| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
2Nationaal Herbarium Nederland, Universiteit Leiden branch, P.O. Box 9514, 2300 RA Leiden, The Netherlands; 3Nationaal Herbarium Nederland, Universiteit Utrecht branch, Heidelberglaan 2, 3584 CS Utrecht, The Netherlands; 4Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3DS, United Kingdom
Received for publication June 24, 2003. Accepted for publication November 7, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Annonaceae • Bayesian inference • maximum parsimony • Miliusa • Miliuseae • rbcL • trnL intron • trnL-trnF intergenic spacer
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; APG II, 2003
). Sister to Annonaceae are Eupomatiaceae (Qiu et al., 2000
), which were previously included as a subfamily within Annonaceae and have always been considered closely related based on morphological characters (e.g., karyology; Morawetz, 1988
).
Although the position of Annonaceae among the flowering plants and their monophyly are not disputed (e.g., Fries, 1959
; Keßler, 1993
; Doyle et al., 2000
), the relationships of the genera within the family are not well understood. Morphological characters useful for the delimitation of genera and species have overlap at higher taxonomic levels (e.g., tribal level). More or less formal classifications based on intuition or phenetic analysis of morphological characters (Hutchinson, 1923
, 1964
; Sinclair, 1955
; Fries, 1959
; Walker, 1971
; Van Heusden, 1992
; Van Setten and Koek-Noorman, 1992
; Keßler, 1993
; Koek-Noorman et al., 1997
) do not accurately predict relationships between genera in Annonaceae.
DNA sequence data provide an alternative source of characters suitable for building a phylogenetic classification of the family, but few molecular studies have been performed on Annonaceae. Van Zuilen (1996)
and Meade (2000)
used molecular data (RAPDs, RFLPs, the trnL-trnF intergenic spacer, and ITS sequences) to establish relationships within and between a small number of selected genera in Annonaceae. The only comprehensive molecular study on the family level was carried out by Bygrave (2000
; also partly published in Doyle et al., 2000
) on 130 Annonaceae taxa collected worldwide. Using only rbcL gene sequence data, he was able to resolve many suprageneric relationships, but other relationships among and within genera remained unresolved.
This study focuses on the Asian genus Miliusa A. DC., which comprises approximately 35 species in the lowland forests from Southeast Asia to Australia and New Guinea. Some species are confined to drier, monsoon-affected areas and are deciduous. The genus is generally characterized by strongly pubescent receptacles, outer petals similar to the sepals, inner petals predominantly with a saccate base, and stamens with a connective not extending over the theca (the so-called miliusoid stamen type, Fig. 1a). During the revision of the Miliusa species of the Flora Malesiana region and Australia (Mols and Keßler, 2003b
), it became evident that not all Miliusa species had inner petals with a saccate base (i.e., M. amplexicaulis Ridl. and M. parviflora Ridl.). Additionally, M. amplexicaulis differs from the other Miliusa species in having a thick arc of glandular tissue at the base instead of running along the midrib. These morphological differences might warrant the exclusion of these species and related taxa from Miliusa.
|
). In total, these genera comprise approximately 130 species, which is 6% and 13% of Annonaceae species worldwide and in Asia, respectively. Their diagnostic characters are valvate sepals and petals, a low number of stamens and carpels, and the miliusoid stamen type (Fig. 1a, with the exception of Mezzettia; Hooker and Thomson, 1855
; Keßler, 1993
). The more general stamen type in Asian Annonaceae is the uvarioid type (Fig. 1b). The genera within the tribe are distinguished by the size of the outer petals (equal to the inner petals or to the sepals), fusion of the sepals (connate at the base or free), shape of the inner petals (mitre-shaped or ovate and with or without a saccate base and glandular ring), number of stamens (less than 10 or more than 20), number of ovules (1–2 or more), size of the fruits (up to 1 cm or larger) and presence of a stipe (Mols and Keßler, 2003a
). In previous publications the name Saccopetaleae has been used for this tribe, but due to nomenclatorial rules and decisions the name Miliuseae should be used in stead (Mols and Keßler, 2003b
). The relationships among the genera in Miliuseae have not previously been examined in a phylogenetic context. The main objectives of this study were to investigate: (1) whether Miliuseae are monophyletic; (2) whether Miliusa is monophyletic and which are its sister groups; and (3) which morphological characters are phylogenetically informative in Miliuseae.
To answer these questions, the gene rbcL, the intron of trnL, and the intergenic spacer between trnL and trnF, all belonging to the plastid genome, were chosen as phylogenetic markers. The rbcL gene encodes for the large subunit of ribulose 1,5- bisphosphate carboxylase/oxygenase and has been widely used for phylogenetic reconstruction at the family level or above in angiosperms (e.g., Chase et al., 1993
; Ingrouille et al., 2002
). The trnL (UAA) intron (a group I intron) and the trnL-trnF intergenic spacers are noncoding regions (Gielly and Taberlet, 1994
, 1996
). These markers have been used in recent studies in Annonaceae (Bygrave, 2000
; L. W. Chatrou et al., unpublished data), enabling us to take advantage of the large database of sequences already available.
|
MATERIAL AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
in which major clades were found within Annonaceae. All members of Miliuseae were found in his unsupported group 2, referred to as the malmeoid, piptostigmoid, miliusoid (MPM) clade after Doyle and Le Thomas (1994
, 1996
). This MPM clade in the study by Bygrave contained Asian, African, and American taxa. To determine the position of Miliusa in Miliuseae and to determine whether the tribe is monophyletic, 114 taxa were analyzed (see Appendix Supplemental Data accompanying online version of this article). These taxa include seven species of Miliusa (sampling based on availability of material suitable for DNA extractions, but covering the morphological and geographical variation within the genus) and four of the five other genera of Miliuseae (no suitable material of Phoenicanthus was available). In addition, more than 60 representatives of other Asian, African, and American taxa were included as molecular (Bygrave, 2000
; L. W. Chatrou et al., unpublished data) and morphological data suggested a close relationship of these genera with Miliuseae. Finally, three Melodorum accessions were added as representatives of another, more distantly related, large clade found in the family cladogram (Bygrave, 2000
; L. W. Chatrou et al., unpublished data).
Outgroup choice
Outgroups were species of Anaxagorea based on the placement of this genus as closest sister to all other Annonaceae using molecular (L. W. Chatrou et al., unpublished data; Doyle et al., 2000
), biogeographical (Maas and Westra, 1984
, 1985
), and morphological (Doyle and Le Thomas, 1994
, 1996
) characters.
DNA extraction, PCR amplification, and sequencing
Total genomic DNA was extracted using either a modified cetyl trimethyl ammonium bromide (CTAB) method (Doyle and Doyle, 1987
) or the DNeasy Plant Mini Kit (QIAGEN, Leusden, Netherlands). The CTAB extraction was done by adding 1300 µL of CTAB (65°C) and 13 µL 2-mercaptoethanol to 50 mg of dried leaf material and homogenizing this with a micropestle in a preheated mortar. A total of 1 mL of this solution was heated to 65°C in a waterbath. After 20 min, 1 mL of chloroform : isoamylalcohol (24 : 1) was added. The sample was left to shake for 90 min and then centrifuged at 13 000 rpm for 10 min. The supernatant (300 µL) was cleaned and precipitated using the Wizard PCR Preps DNA Purification System or Wizard DNA Clean-Up System (Promega, Leiden, Netherlands). All DNA was extracted from silica-gel- dried leaves collected in the field or in botanical gardens.
The rbcL gene was amplified using the following two primer combinations: 1F/724R and 636F/1460R (Fay et al., 1997
). The trnL intron and trnL-trnF intergenic spacer were amplified using the primer combinations c/d and e/f, respectively (Taberlet et al., 1991
). The thermal cycling protocol comprised 28 cycles, each with 1 min denaturation at 94°C, 1 min annealing at 50°C, an extension of 2 min at 72°C, concluding with an additional extension of 10 min at 72°C. All PCR products were cleaned using a QIAquick PCR Purification Kit (QIAGEN). The samples were sequenced with the PCR primers and electrophoresed on an ABI 377 automatic sequencer (Applied Biosystems, Nieuwekerk a/d IJssel, Netherlands).
Phylogenetic analysis
Sequences of trnL-trnF were aligned automatically with the Clustal method in Megalign 4.03 (DNAstar, 1999, Madison, Wisconsin, USA) and subsequently adjusted by hand (according to the recommendation of Kelchner, 2000
), resulting in the removal of 43 characters (mainly mononucleotide repeats) due to ambiguous alignment. All DNA sequences were deposited in Genbank (see Appendix in Supplemental Data accompanying the online version of this article). Aligned data matrices have been submitted to TreeBASE (study accession number: S1034; matrix accession number: M1757) and can be obtained upon request from the corresponding author.
Phylogenetic analyses were performed with PAUP* version 4.0b10 (Swofford, 2002
) using maximum parsimony (MP) with the heuristic search option (10 times simple addition). The options tree bisection-reconnection (TBR), accelerated transformation optimization (ACCTRAN), and save all minimal trees (MULTREES) were invoked. Because a large number of trees were found and the memory limit quickly reached, an additional heuristic search using 1000 replicates with 10 trees saved per replicate was performed to investigate whether shorter trees could be found; when the latter was the case, one of the shorter trees was used as a starting tree for a new heuristic search. Character states were specified as unordered and equally weighted (Fitch, 1971
). Indels were coded as single characters (absence or presence). Insertions/deletions were treated as missing data.
The MP analyses were performed with three different data sets: the rbcL data, the trnL-trnF data, and both data sets combined. The separate trnL-trnF units have been combined in one data set in stead of separate ones because the different units form a connected noncoding region. This makes it possible to compare the noncoding vs. coding region. Bootstrap analyses were performed to evaluate internal support (Felsenstein, 1985
). Bootstrap percentages (BP) were calculated using 2000 replicates, saving one tree per replicate (Hedges, 1992
). Because uninformative characters can lower the BP (Harshman, 1994
), only the informative characters were included. Bremer support (Bremer, 1998; Donoghue et al., 1992
) was calculated using AutoDecay 4.0.2 (Eriksson, 1998
) with 100 addition sequence replicates per run, for the combined analysis only. The separate data sets were tested for congruence with the partition homogeneity test (PHT; Farris et al., 1995
; 1000 replicates, 10 trees saved per replicate). However, as this type of test is said to be unreliable (Yoder et al., 2001
; Reeves et al., 2001
), the separate bootstrap trees were compared for "hard" incongruence (i.e., BP 
85%; Seelanan et al., 1997
; Wiens, 1998
) as well. Transition/transversion (ts/tv) ratios were calculated on one of the most parsimonious trees (arbitrarily chosen).
The combined data set without the binary indels was also analyzed using Bayesian inference (BA) with MrBayes version 3.0b (Huelsenbeck and Ronquist, 2001
and references). This analysis was performed to test the robustness of the evolutionary signal in the data set, because it is based on different optimality criteria and parameter settings used in phylogeny reconstruction in comparison with MP. For the BA, the data set was split into four partitions (rbcL, noncoding part downstream of rbcL, trnL intron, trnL-trnF intergenic spacer). A DNA substitution model was assigned to each partition using MrModeltest version 1.1b (a simplified version of Posada and Crandall's [1998]
Modeltest version 3.06; Nylander, 2002
). For those partitions for which a model was set using a gamma (
)-distributed rate variation across sites, the 
-value was set to unlinked. Furthermore, a rate multiplier was enforced to constrain the average rate across the partitions to one. The Markov Chain Monte Carlo analyses (MCMC; Geyer, 1991
) were run for 3 000 000 generations with four simultaneous MCMC chains to calculate the posterior probabilities (PP). Prior probabilities for all trees were equal. A random tree was used as the starting point for the analysis, and one tree per 10 generations was saved. The burn-in values were determined empirically from the likelihood values. Finally, 50% majority consensus trees were calculated together with the branch lengths and approximations of the PP for the observed bipartitions. Because MrBayes only accepts one outgroup taxon, we chose Anaxagorea luzonensis for this purpose. The analysis was repeated four times to assure sufficient mixing by confirming conversion to the same PP and topology.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
and Olmstead et al. (1992)
. The final alignment had a total length of 1433 bp, of which 1394 bp belonged to the rbcL gene and 39 bp to the noncoding region. This noncoding region was quite variable and contained seven indels, ranging in size from 4 to 11 bp, four of which were phylogenetically informative. The Phaeanthus splendens and Mitrephora celebica accessions were omitted from this analysis, because the rbcL region could not be amplified for these species. Of the 1440 characters used (bp + indels), 262 were variable, 160 of which (11%) were potentially phylogenetically informative. Pairwise sequence divergence varied between 0% (in nine comparisons between taxa, e.g., Miliusa horsfieldii and M. mollis) and 5.1% (Melodorum cf. fruticosum II vs. Anaxagorea javanica). The ts/tv ratio was 1.5. The third codon position contributed more steps than the first and second positions and also displayed a higher consistency index (CI) (Table 1).
|
|
|
Phylogenetic analysis yielded >10 000 trees of length 577 (CI = 0.78, RI = 0.86; Table 2), with none of the individual trees fully resolved. In comparison with rbcL, the strict consensus obtained from this analysis showed more resolution and resulted in more clades supported by high BP. The following clades were monophyletic (Fig. 3): Miliusa (A; BP 96%), Orophea (B; BP 66%), Alphonsea (C; BP 85%), several Polyalthia clades (F1, F2, F4 and S; BP 93%, 99%, 62%, and 84%), Popowia (P; BP 96%), Phaeanthus (Q; BP 77%), Pseuduvaria including Petalolophus and Craibella (N; BP 99%), Marsypopetalum including Polyalthia p.p. (L; BP 94%), Neo-uvaria (G; BP 100%), Sapranthus with Tridimeris (K; BP 72%), Meiogyne including Fitzalania (H; BP 50%), Mitrephora (O; BP <50%), and Melodorum (T; BP 100%). Compared with the rbcL results, five additional clades were supported: U (most Asian taxa; BP 82%), a clade combining U + Monocarpia (R; BP 95%), a clade V (combining U + R + SAm + S + Afr + Cremastosperma microcarpum + Malmea dielsiana + Onychopetalum periquino + Pseudoxandra lucida; BP 98%), a clade V + T (BP 83%), and a clade V + T + E (BP 100%). Compared with the rbcL results, clades A, B, and S included additional species.
|
85%) BP. The combined data set comprised a total of 2520 characters, including 92 indels. Of these characters, 653 were variable, 338 of which (13%) were potentially phylogenetically informative. Phylogenetic analysis yielded >10 000 MPTs of length 1184 (CI = 0.63 and RI = 0.78; Table 2).
Bootstrap analysis of this combined data set provided more supported clades than either individual data set (Fig. 4). The following genera were recovered: Miliusa (A1 + A2 + A3; BP 97%), Orophea (B1 + B2; BP 86%), Alphonsea (C; BP 86%), several Polyalthia clades (F1–F4 [including Enicosanthum] and S; BP 98%, 64%, 82%, 71%, and 90%), Mitrephora (O; BP 63%), Popowia (P; BP 100%), Phaeanthus (Q; BP 82%), Pseuduvaria including Petalolophus and Craibella (N; BP 100%), Marsypopetalum including Polyalthia p.p. (L; BP 98%); Neo-uvaria (G; BP 100%), Meiogyne including Fitzalania (H; BP < 50%); Central American clade including Stelechocarpus burahol (K; BP < 50%); and Melodorum (T; BP 100%). Compared with the rbcL and trnL-trnF results, several clades (A1–A3, B1, B2, C, N) were well resolved at the species level.
|
) with a gamma-distributed rate and an invariable proportion of sites (GTR + + I) to be the least-rejected model. The noncoding region downstream of the rbcL gene was assumed to evolve under the Jukes-Cantor model (Jukes and Cantor, 1969
) with a gamma-distributed rate only (JC + ). In both trnL-trnF partitions, the DNA sequence evolution was best explained using the Hasegawa-Kishino- Yano model (Hasegawa et al., 1985
) with a gamma-distributed rate (HKY + ).
Of the 30 000 trees obtained from BA the burn-in was estimated, resulting in more than 7500 trees being discarded. The remaining trees (Table 3) yielded a consensus tree (not shown here, PP given on strict consensus of the MP analysis: Fig. 4) with a topology similar to the bootstrap consensus tree of the combined analysis. Three clades with BP 70% did not receive high PP: Q (PP 69%), U (PP < 50%), and an internal clade of O (PP < 50%). Seven additional clades obtained strong support (PP 95%) in BA only: placement of Platymitra (D) as sister to Alphonsea (C; PP 100%), Meiogyne including Fitzalania (H; PP 100%), Mitrephora (O [and an internal clade]; PP 100%, 99%), and several (internal) Polyalthia clades (F2 and F4; PP 100%, 98%, 96%).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
), most of the changes within rbcL were in the third codon position. These changes were also phylogenetically informative as the less conservative third codon positions did not show more homoplasy than the first and second positions (Table 1). Therefore, the third codon positions were not excluded from the analyses as this resulted in a loss, and not gain, of phylogenetic information (data not shown). The combined trnL-trnF and rbcL data set led to more resolution and higher bootstrap values (Table 2). Because only a limited number of potentially phylogenetically informative characters were present to infer the relationships among the 114 taxa, analyses with a smaller number of taxa (mainly focusing on the Miliuseae) were performed. These analyses did not result in a better resolved cladogram and/or higher bootstrap percentages (data not shown).
Polyphyly of Miliuseae
The five genera formerly assigned to Miliuseae do not form a clade, as Mezzettia (E) is found outside clade U near the outgroup. This genus was only tentatively placed in Miliuseae (Keßler, 1993
) based on the reduced number of carpels and stamens, which also occur in Orophea and Phoenicanthus (Van der Heijden and Keßler, 1990
). The genus lacks the typical miliusoid stamen type considered to be the most important diagnostic character of the tribe. Instead, the stamens in this genus are peculiar in having introrse thecae with a single sporangium (Fig. 1c). Other characters also suggest an isolated position of Mezzettia in Annonaceae. Its seeds have three integuments, which are only found in a few other taxa in the family (Christmann, 1986
, 1989
). Furthermore, the satellite chromosomes have a small satellite (Okada and Ueda, 1984
), whereas in all other Annonaceae they are large. Additionally, Okada and Ueda (1984)
observed 2n = 14 for Mezzettia parviflora, which is only found in nine annonaceous genera worldwide (Morawetz, 1986
; Morawetz and Le Thomas, 1988
). Three of these nine taxa belong to a group referred to as the ambavioids (Doyle and Le Thomas, 1994
, 1996
), which are characterized by the presence of three integuments in the seeds, chromosome number x = 7, monosulcate pollen, and irregular endosperm ruminations. The last two characters are also shared by Anaxagorea, which is considered the sister group of the ambavioids. Molecular and morphological characters indicate that Mezzettia is part of the ambavioid group, which is placed as sister to the rest of Annonaceae (except Anaxagorea).
The genera Alphonsea, Orophea, and Platymitra can neither be confirmed nor rejected as members of Miliuseae. Few of the generic relationships are supported by high bootstrap percentages. Until more resolution is found, clade U (Fig. 4) is considered to be the clade containing the genus Miliusa and its closest relatives. Because we do not want as yet to give this large clade formal taxonomic status, it will be referred to as the miliusoid clade. The non-Miliuseae genera in this clade were previously ascribed to several other tribes or groups (Keßler, 1993
). The miliusoid clade consists of 24 genera (Fig. 4), but further research is needed to elucidate the relationships among these genera.
Monophyly of Miliusa
For Miliusa (A), seven of the approximately 35 species have been included. These species end up in three clades (A1–A3, Fig. 4). Clade A1 consists of five species, which differ only slightly in number of ovules, shape and size of the leaves, and apices of the inner petals. Miliusa velutina differs from the other included species in its deciduous habit, shape and size of the leaves, sepals and petals, fruits, and number of stamens and carpels. It much more resembles the species from clade A2, M. horsfieldii and M. lineata, especially in the size and shape of the fruits. However, the placement of M. velutina as the sister to the rest of A1 and not in A2 corresponds with original sections recognized in Miliusa. All species in clade A1 have one or two ovules (section Miliusa), whereas the species in A2 have more than two ovules (section Saccopetalum). Additionally, M. velutina is one of the few species in the genus [together with M. parviflora and M. andamanica (King) Finet & Gagnep.] with non-saccate inner petals. This indicates that the species with non-saccate inner petals should be included in the genus. Miliusa horsfieldii and M. lineata (A2) had identical sequences (only differing in the number of base pairs we were able to obtain). This supports the view that the morphologically almost identical M. lineata and several other species should be synonymized with M. horsfieldii (Mols and Keßler, 2003b
).
The third clade A3 contains two specimens of Miliusa mollis var. mollis and forms an unresolved branch with the other Miliusa clades. This species is a member of a group of six taxa (M. amplexicaulis, M. fusca Pierre, M. glandulifera C. E. C. Fischer, M. glochidioides Hand.-Mazz., M. mollis var. mollis, and M. mollis Pierre var. sparsior Craib), which deviate from the other Miliusa species in shape of the inner petals (broadly ovate to almost triangular, with a basal glandular tissue ring, and without a saccate base) and connate sepals (J. B. Mols, personal observation). This study shows that these species should be included in Miliusa.
The most likely sister group of Miliusa based on both MP and BI analysis is clade F1 containing three Polyalthia species. However, bootstrap support is low and no apparent morphological similarities between clade F1 and Miliusa are known.
Monophyly of other genera
Although most intergeneric relationships were not clarified, we found that most genera are monophyletic and supported by high bootstrap percentages as far as can be concluded on the basis of a limited sampling for some of the genera. Only three genera appear to be not monophyletic. For two genera, Stelechocarpus and Desmopsis (found in clades J and K), the polyphyly does not receive strong support.
The only genus with high support for polyphyly is Polyalthia, which is one of the largest genera among Asian Annonaceae, consisting of approximately 150 species. Generic delimitation has been based on sepal and petal aestivation, shape, and size. Several authors (e.g., Rogstad and Le Thomas, 1989
; Schatz and Le Thomas, 1990
; Van Setten and Koek-Noorman, 1992
; Doyle and Le Thomas, 1994
; Johnson and Murray, 1999
) have suggested that Polyalthia should be split into several smaller genera. This has not been done because no suitable characters were found to easily distinguish "natural groups." In total, 28 species of Polyalthia were included in this study, among which 26 were from Asia, one from Madagascar (P. pendula), and one from Africa (P. stuhlmannii). Additionally, one species of the African genus Greenwayodendron was included, which previously has been considered a separate section within Polyalthia (e.g., Fries, 1959
). In total, seven clades (F1–F4, L, S, and Afr) were found, containing species assigned to Polyalthia (Fig. 4).
Greenwayodendron is situated as sister to the African genus Piptostigma (Afr) and seems not closely related to the other clades containing species of Polyalthia, warranting its status as a separate genus. The second Polyalthia clade (S) contains three species that are part of what Rogstad (1989)
referred to as the Polyalthia hypoleuca complex, which share several unique morphological characters (concerning the bark, leaves, and seeds) not found elsewhere within Polyalthia. Our results indicate that this complex of species should be transferred to a new genus. The placement of this complex outside clade T among the South-American genera (SAm1 + 2) is supported by pollen morphology, because all members of clades S, Sam 1, and Sam 2 have monosulcate, boat-shaped pollen (Rogstad and Le Thomas, 1989
).
The remaining five clades of Polyalthia species are all part of clade U. Clade F1 is the weakly supported sister group of Miliusa in the combined analyses. It consists of three species. The African species P. stuhlmannii and P. pendula are similar in their sepals, stamens, number of ovules, fruits, and seeds. A difference is found in the pollen as P. pendula is described as inaperturate (Schatz and Le Thomas, 1990
) and P. stuhlmannii as monosulcate (Le Thomas, 1988
; Doyle and Le Thomas, 1996
). The Asian P. cerasoides differs mainly from these two species in having smaller flowers that are always solitary, smaller globose fruits, and disulculate pollen (Le Thomas, 1988
).
Clade F2, containing the type species P. subcordata, should be considered as Polyalthia sensu stricto (s.s.). The genus Haplostichanthus, which is nested within this clade, consists of only six species found in the Philippines, Moluccas, New Guinea, and Australia. In her revision of the genus, Van Heusden (1994b)
indicated a close relationship with some Polyalthia species based on leaf and flower characters.
Clade F3 is closely related (though with low branch support) to clade F2. It contains only three species, but two of these, Polyalthia debilis and P. suberosa, are quite similar in their habit, petals, and fruits. The close relationship between clades F2 and F3 is supported by morphology because all the species have two or more ovules, which is a key character for Polyalthia section Polyalthia.
In clade F4, besides Polyalthia species, several species of Enicosanthum are present. The latter includes a large number of species formerly assigned to Polyalthia. Enicosanthum and Polyalthia are similar in general appearance and fruit and seed characters. The main reason of segregating the genus is the presence of imbricate sepals and petals in Enicosanthum and valvate sepals and petals in Polyalthia. The results presented here indicate that Polyalthia species of clade F4 should be included in Enicosanthum and that when dealing with the classification of Polyalthia more emphasis should be placed on fruit rather than floral characters. Clade F4 can be distinguished from clades F2 and F3 by the presence of a single ovule (Polyalthia section Monoon).
Clade L consists of P. littoralis and Marsypopetalum. According to Van Heusden (1992)
P. crassa R. N. Parker (and related species, i.e., P. littoralis and P. modesta (Pierre) Finet & Gagnep.; P. J. A. Keßler, personal observation) might be erroneously placed in Polyalthia because it clearly resembles Marsypopetalum due to similarity of the leaves, perianth, stamens, and fruits. According to all analyses, Polyalthia littoralis (and relatives) should be transferred to Marsypopetalum.
According to our results, Pseuduvaria (N) also contains the monotypic genus Petalolophus endemic to New Guinea. Petalolophus megalopus is unique in having large winged inner petals. Pseuduvaria flowers are generally unisexual. In contrast, Petalolophus has bisexual flowers and, larger inner petals. However, Pseuduvaria novaguineensis J. Sinclair has long peduncles, short pedicels with an articulation in the middle, bisexual flowers, dissimilar outer and inner stamens, and smooth, globose monocarps just as Petalolophus megalolopus (Su, 2002
), which supports the molecular placement of Petalolophus in Pseuduvaria.
Monotypic Fitzalania, endemic to Australia, is placed within Meiogyne (H). Based on size and shape of inner and outer petals and position of the lateral ovules, the genus was placed in an informal group together with Orophea, Platymitra, Popowia, Pseuduvaria, and Petalolophus (Van Heusden, 1992
). In contrast, endosperm ruminations and number of seeds places Fitzalania together with, among others, Haplostichanthus, Desmopsis, Polyaulax, Oncodostigma, Guamia, and Ancana (Van Setten and Koek-Noorman, 1992
). The last four genera are now included in Meiogyne (Van Heusden, 1994a
). Van Heusden indicated that it was not warranted to include Fitzalania in Meiogyne based on the floral differences. Our results, however, show Fitzalania to be nested within Meiogyne, although support for this is low, suggesting that seed characters may be more informative in this case.
The miliusoid clade also contains a clade K, consisting of endemic Central American genera. That this clade is placed among Asian taxa suggests multiple invasions of Central America, at least once by species from more widespread Neotropical genera such as Annona and Guatteria and by endemic Central American taxa derived from Asian ancestry. This Asian and Central American link for Annonaceae was also suggested by Schatz (1987)
. He refers to Gentry (1982a
, b
) and several others who describe remnants of a small tropical Laurasian floristic element in Central America.
Monophyly of subgenera
Besides the monophyly of the genera of the Miliusoid clade, several well-supported infrageneric relationships were found. Orophea consists of approximately 50 species, and Keßler (1988)
recognized two subgenera. Subgenus Orophea contains approximately 30 species characterized by the presence of an indument on the young shoots and carpels, six ovules per carpel, and cylindrical fruits and seeds. Subgenus Sphaerocarpon Keßler (including the genus Mezzettiopsis Ridl.; Leonardia and Keßler, 2001
) consists of 22 species characterized by the absence of indument on the young shoots and carpels, two ovules per carpel, and globose fruits and seeds. In this study, two subclades can be recognized, clade B1 consisting of O. brandisii, O. celebica, O. enterocarpa, O. enneandra, and O. myriantha, and clade B2 containing O. creaghii, O. kerrii, O. polycarpa, and O. cf. malayana (Fig. 4). Clade B1 corresponds with subgenus Orophea and clade B2 with subgenus Sphaerocarpon. Leonardia and Keßler (2001)
performed a phylogenetic analysis of the latter subgenus based on morphological characters and showed that subgenus Sphaerocarpon is monophyletic, which is supported by our molecular analysis.
The original members of Miliuseae were almost all found in a clade consisting of mainly Asian, Central-American, and some African taxa (miliusoid clade: U), which were previously ascribed to other tribes within Annonaceae. Miliuseae cannot be considered monophyletic as Mezzettia previously belonging to this tribe, is positioned outside the miliusoid clade, and several new taxa need to be included. The intergeneric relationships in the miliusoid clade remained poorly resolved, but the genera included, except Polyalthia, appear to be monophyletic. The genus Miliusa was found to be monophyletic and includes the nonsaccate species. A tentative sister group of Miliusa seems to be a group of three Polyalthia species, which most likely represents a much larger number of Polyalthia species, from East Africa, Madagascar (Schatz and Le Thomas, 1990
), and mainland Asia. In Miliusa, Orophea, and Polyalthia, the molecular clades found are congruent with sectional divisions. It can be concluded that molecular data support the previous generic and intrageneric delimitation based on morphology. Pollen, fruit, and seed characters, and number of ovules per carpel especially appear to be phylogenetically informative and promising for the elucidation of the intergeneric relationships of the miliusoid clade.
|
FOOTNOTES |
|---|
5 E-mail: mols{at}nhn.leidenuniv.nl
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Bremer K. 1988 The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42: 795-803[CrossRef][ISI]
Bygrave P. C. 2000 Molecular systematics of Annonaceae Juss. Ph.D. dissertation, University of Reading, Reading, UK
Chase M. W. et al 1993 Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80: 528-580[CrossRef][ISI]
Christmann M. 1986 Beiträge zur Histologie der Annonaceen-Samen. Botanische Jahrbücher für Systematik, Pflanzengeschichte und Pflanzengeographie 106: 379-390
Christmann M. 1989 Die tritegmischen Annonaceen-Samen. Botanische Jahrbücher für Systematik, Pflanzengeschichte und Pflanzengeographie 110: 433-439
Donoghue M. J. R. G. Olmstead J. F. Smith J. D. Palmer 1992 Phylogenetic relationships of Dipsacales based on rbcL sequences. Annals of the Missouri Botanical Garden 79: 333-345
Doyle J. J. Doyle 1987 A rapid isolation procedure for small quantities of fresh leaf tissue. Phytochemistry Bulletin Botanical Society of America 19: 11-15
Doyle J. A. P. C. Bygrave A. Le Thomas 2000 Implications of molecular data for pollen evolution in Annonaceae. In M. M. Harley, C. M. Morton, and S. Blackmore [eds.], Pollen and spores: morphology and biology, 259–284. Royal Botanical Gardens, Kew, UK
Doyle J. A. A. Le Thomas 1994 Cladistic analysis and pollen evolution in Annonaceae. Acta Botanica Gallica 141: 149-170[ISI]
Doyle J. A. A. Le Thomas 1996 Phylogenetic analysis and character evolution in Annonaceae. Bulletin du Muséum National d'Histoire Naturelle, 4e série, Sect. B, Adansonia 18: 279-334
Eriksson T. 1998 AutoDecay vers. 4.0.2. Department of Botany, Stockholm University, Stockholm, Sweden
Farris J. S. M. Källersjo A. G. Kluge C. Bult 1995 Testing significance of incongruence. Cladistics 10: 315-319[CrossRef][ISI]
Fay M. F. S. M. Swenson M. W. Chase 1997 Taxonomic affinities of Medusagyne oppositifolia (Medusagynaceae). Kew Bulletin 52: 111-120[CrossRef]
Felsenstein J. 1985 Confidence limits on phylogenetics: an approach using the bootstrap. Evolution 39: 283-291
Fitch W. M. 1971 Toward defining the course of evolution: minimal change for a specific tree topology. Systematic Zoology 20: 406-416[CrossRef][ISI]
Fries R. E. 1959 Annonaceae. In A. Engler and K. Prantl [eds.], Die natürlichen Pflanzenfamilien, 2nd ed., 17a, IIB, 1–171. Duncker and Humblot, Berlin, Germany
Gentry A. H. 1982a Neotropical floristic diversity: phytogeographical connections between Central and South America, Pleistocyne climatic fluctuations, or an accident of the Andean orogeny?. Annals of the Missouri Botanical Garden 69: 557-593[CrossRef][ISI]
Gentry A. H. 1982b Phytogeographic patterns as evidence for a Chocó refuge. In T. Prance [ed.], Biological diversification in the tropics, 112– 136. Columbia University Press, New York, New York, USA
Geyer C. J. 1991 Markov chain Monte Carlo maximum likelihood. In M. Keramidas [ed.], Computing science and statistics: proceedings of the 23rd symposium on the interface, 156–163. Interface Foundation of North America, Fairfax, Virginia, USA
Gielly L. P. Taberlet 1994 The use of chloroplast DNA to resolve plant phylogenies: noncoding versus rbcL sequences. Molecular Biology and Evolution 11: 769-777[Abstract]
Gielly L. P. Taberlet 1996 A phylogeny of the European gentians inferred from chloroplast trnL (UAA) intron sequences. Botanical Journal of the Linnean Society 120: 57-75[CrossRef]
Harshman J. 1994 The effect of irrelevant characters on bootstrap values. Systematic Biology 43: 419-424[CrossRef]
Hasegawa M. H. Kishino T. Yno 1985 Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 21: 160-174
Hedges S. B. 1992 The number of replications needed for accurate estimation of the bootstrap P value in phylogentic studies. Molecular Biology and Evolution 9: 366-369[ISI][Medline]
Holmgren P. K. N. K. Holmgren L. C. Barnett 1990 Index herbariorum, part I: the herbaria of the world, 8th ed. New York Botanical Garden, Bronx, New York, USA.
Hooker J. D. T. Thomson 1855 Flora Indica I: 86–153. W. Pamplin, London, UK
Huelsenbeck J. P. F. Ronquist 2001 MrBayes: Bayesian inference of phylogenetic trees. Bioinformatics 17: 754-755
Hutchinson J. 1923 Contributions towards a phylogenetic classification of flowering plants 2. The genera of Annonaceae. Kew Bulletin 7: 241-261
Hutchinson J. 1964 The genera of flowering plants. Dicotyledons 1. Clarendon Press, Oxford, UK
Ingrouille M. J. M. W. Chase M. F. Fay D. Bowman M. van der Bank A. de Bruijn 2002 Systematics of Vitaceae from the viewpoint of plastid rbcL DNA sequence data. Botanical Journal of the Linnean Society 138: 421-432[CrossRef]
Johnson D. M. N. A. Murray 1999 Four new species of Polyalthia (Annonaceae) from Borneo and their relationship to Polyalthia insignis. Contributions from the University of Michigan Herbarium 22: 95-104
Jukes T. H. C. R. Cantor 1969 Evolution of protein molecules. In H. M. Munro [ed.], Mammalian protein metabolism. Academic Press, New York, New York, USA
Källersjö M. J. S. Farris M. W. Chase B. Bremer M. F. Fay C. J. Humphries G. Petterson O. Seberg K. Bremer 1998 Simultaneous parsimony jacknife analysis of 2538 rbcL DNA sequences reveals support for major clades of green plants, land plants, seed plants and flowering plants. Plant Systematics and Evolution 213: 259-287[CrossRef][ISI]
Kelchner S. A. 2000 The evolution of non-coding chloroplast DNA and its application on plant systematics. Annals of the Missouri Botanical Garden 87: 482-498[CrossRef][ISI]
Keßler P. J. A. 1988 Revision der Gattung Orophea Blume (Annonaceae). Blumea 33: 1-80
Keßler P. J. A. 1993 Annonaceae. In K. Kubitzki, J. G. Rohwer, and V. Bittrich [eds.], The families and genera of vascular plants, vol. 2, 93– 129. Springer-Verlag,Berlin, Germany
Koek-Noorman J. A. K. Van Setten C. M. Van Zuilen 1997 Studies in Annonaceae XXVI. Flower and fruit morphology in Annonaceae: their contribution to patterns in cluster analysis. Botanische Jahrbücher für Systematik, Pflanzengeschichte und Pflanzengeographie 119: 213-230
Leonardia A. A. P. P. J. A. Keßler 2001 Additions to Orophea subgenus Sphaerocarpon (Annonaceae): revision and transfer of Mezzettiopsis. Blumea 46: 141-163[ISI]
Le Thomas A. 1988 Variation de la region apérturale dans le pollen des Annonacées. Taxon 37: 644-656[CrossRef][ISI]
Maas P. J. M. L. Y. Th. Westra 1984 Studies in Annonaceae II. A monograph of the genus Anaxagorea A. St. Hil. Part 1. Botanische Jahrbücher für Systematik, Pflanzengeschichte und Pflanzengeographie 105: 73-134
Maas P. J. M. L. Y. Th. Westra 1985 Studies in Annonaceae II. A monograph of the genus Anaxagorea A. St. Hil. Part 2. Botanische Jahrbücher für Systematik, Pflanzengeschichte und Pflanzengeographie 105: 145-204
Meade C. V. 2000 A systematic revision of the Uvaria L. group (Annonaceae) in Continental Asia. Ph.D. dissertation, University of Dublin, Trinity College, Dublin, Ireland
Mols J. B. P. J. A. Keßler 2003a Studies in the Miliuseae. Review of the taxonomic history of a polyphyletic "tribe.". Telopea 10: 113-124
Mols J. B. P. J. A. Keßler 2003b The genus Miliusa (Annonaceae) in the Austro-Malesian area. Blumea 48: 421-462.[ISI]
Morawetz W. 1986 Systematics and karyoevolution in Magnoliidae: Tetrameranthus as compared with other Annonaceae genera of the same chromosome number. Plant Systematics and Evolution 154: 147-173[CrossRef][ISI]
Morawetz W. 1988 Karyosystematics and evolution of Australian Annonaceae as compared with Eupomatiaceae, Himantandraceae, and Austrobaileyaceae. Plant Systematics and Evolution 159: 49-79[CrossRef][ISI]
Morawetz W. A. Le Thomas 1988 Karyology and systematics of the genus Ambavia and other Annonaceae from Madagascar. Plant Systematics and Evolution 158: 155-160[ISI]
Nylander J. 2002 MrModeltest version 1.1b. Department of Systematic Zoology, EBC, Uppsala, Sweden
Okada H. K. Ueda 1984 Cytotaxonomical studies on Asian Annonaceae. Plant Systematics and Evolution 144: 165-177[CrossRef][ISI]
Olmstead R. G. H. J. Michaels K. M. Scott J. D. Palmer 1992 Monophyly of the Asteridae and identification of their major lineages inferred from DNA sequences of rbcL. Annals of the Missouri Botanical Gardens 79: 249-265
Posada D. K. A. Crandall 1998 Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817-818
Qiu Y. L. J. Lee F. Bernasconi-Quadroni D. E. Soltis P. S. Soltis M. Zanis E. A. Zimmer Z. Chen V. Savolainen M. W. Chase 2000 Phylogeny of basal angiosperms: analysis of five genes from three genomes. International Journal of Plant Sciences 161: Supplement 6 3-27[CrossRef]
Reeves G. M. W. Chase P. Goldblatt P. Rudall M. F. Fay A. T. Cox B. Lejeune T. Souza-Chies 2001 Molecular systematics of Iridaceae: evidence from four plastid DNA regions. American Journal of Botany 88: 2074-2087
Rogstad S. H. 1989 The biosystematics and evolution of the Polyalthia hypoleuca complex (Annonaceae) of Malesia, I. Systematic treatment. Journal of the Arnold Arboretum 70: 153-246[ISI]
Rogstad S. H. A. Le Thomas 1989 Pollen characters of the Polyalthia hypoleuca complex (Annonaceae): their significance in establishing monophyly and candidate outgroups. Bulletin du Muséum National d'Histoire Naturelle, 4e série, Sect. B., Adansonia 3: 257-278
Schatz G. E. 1987 Systematic and ecological studies of Central American Annonaceae. Ph.D. dissertation, University of Wisconsin, Madison, Wisconsin, USA
Schatz G. E. A. Le Thomas 1990 The genus Polyalthia Blume (Annonaceae) in Madagascar. Bulletin du Muséum National d'Histoire Naturelle, 4e série, Sect. B, Adansonia 2: 113-130
Seelanan T. A. Schnabel J. W. Wendel 1997 Congruence and consensus in the cotton tribe (Malvaceae). Systematic Botany 22: 259-290[CrossRef][ISI]
Sinclair J. 1955 A revision of the Malayan Annonaceae. Gardens' Bulletin Straits Settlements 14: 149-516
Soltis D. E. et al 2000 Angiosperm phylogeny inferred from 18S rRNA, rbcL and atpB sequences. Botanical Journal of the Linnean Society 133: 381-461[CrossRef]
Su Y. C. F. 2002 Systematics and phylogeny of Pseuduvaria (Annonaceae). Ph.D. dissertation, University of Hong Kong, Hong Kong, China
Swofford D. L. 2002 PAUP*: phylogenetic analysis using parsimony (* and other methods). Vers. 4. Sinauer, Sunderland, Massachusetts, USA
Taberlet P. L. Gielly G. Pautou J. Bouvet 1991 Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105-1109[CrossRef][ISI][Medline]
Tavare S. 1986 Some probabilistic and statistical problems on the analysis of DNA sequences. Lectures on Mathematics in the Life Sciences 17: 57-86
Van der Heijden E. P. J. A. Keßler 1990 Studies on the tribe Saccopetaleae (Annonaceae) III. Revision of the genus Mezzettia Beccari. Blumea 35: 217-228[ISI]
Van Heusden E. C. H. 1992 Flowers of Annonaceae: morphology, classification and evolution. Blumea Supplement 7: 1-218
Van Heusden E. C. H. 1994a Revision of Meiogyne (Annonaceae). Blumea 38: 487-511[ISI]
Van Heusden E. C. H. 1994b Revision of Haplostichanthus (Annonaceae). Blumea 39: 215-234[ISI]
Van Setten A. K. J. Koek-Noorman 1992 Fruits and seeds of Annonaceae: morphology and its significance for identification. Bibliotheca Botanica 142
Van Zuilen C. M. 1996 Patterns and affinities in the Duguetia alliance (Annonaceae): molecular and morphological studies. Ph.D. dissertation, Utrecht University, Utrecht, Netherlands
Walker J. W. 1971 Pollen morphology, phytogeography and phylogeny of the Annonaceae. Contributions of the Gray Herbarium of Harvard University 202: 1-130
Wiens J. J. 1998 Combining data sets with different phylogenetic histories. Systematic Biology 47: 568-581[CrossRef][ISI][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[ISI][Medline]
This article has been cited by other articles:
|
|
 |
|
  M. D. Pirie, M. P. B. Vargas, M. Botermans, F. T. Bakker, and L. W. Chatrou Ancient paralogy in the cpDNA trnL-F region in Annonaceae: implications for plant molecular systematics Am. J. Botany, June 1, 2007; 94(6): 1003 - 1016. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||
