| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
What's this? |
Systematics and Phytogeography |
2Department of Phanerogamic Botany, Swedish Museum of Natural History, P.O. Box 50007, 104 05 Stockholm, Sweden; 3Laboratoire de Botanique, Centre IRD de Nouméa, B.P. A5 Nouméa Cedex, New Caledonia
Received for publication April 28, 2004. Accepted for publication December 6, 2004.
ABSTRACT
The phylogeny of a representative group of genera and species from the Sapotaceae tribe Chrysophylleae, mainly from Australia and New Caledonia, was studied by jackknife analyses of sequences of nuclear ribosomal DNA. The phylogeny conflicts with current opinions on generic delimitation in Sapotaceae. Pouteria and Niemeyera, as presently circumscribed, are both shown to be nonmonophyletic. In contrast, all species currently assigned to these and other segregate genera confined to Australia, New Caledonia, or neighboring islands, form a supported clade. Earlier classifications in which more genera are recognized may better reflect relationships among New Caledonian taxa. Hence, there is need for a revision of generic boundaries in Chrysophylleae, and particularly within the Pouteria complex, including Leptostylis, Niemeyera, Pichonia, Pouteria pro parte (the main part of section Oligotheca), and Pycnandra. Section Oligotheca have been recognized as the separate genus Planchonella, a monophyletic group that needs to be resurrected. Three clades with strong support in our jackknife analysis have one Australian species that is sister to a relatively large group of New Caledonian endemics, suggesting multiple dispersal events between this small and isolated tropical island and Australia. The phylogeny also suggests an interesting case of a relatively recent and rapid radiation of several lineages of Sapotaceae within New Caledonia.
Key Words: biogeography • conservation • New Caledonia • Planchonella • Pouteria • rDNA • Sapotaceae • systematics
Sapotaceae are a family of trees and shrubs with 53 genera and more than 1200 species (Govaerts et al., 2001
). They occur worldwide, but mostly in tropical and subtropical regions. The systematics of this family is complicated largely because of extensive homoplasy of morphological characters (e.g., numerous cases of evolutionary parallelisms). The most recent classification of the family is that of Pennington (1991)
, who recognized five tribes: Chrysophylleae, Isonandreae, Omphalocarpeae, Mimusopeae, and Sideroxyleae. Pennington adopted a wide generic concept and reduced many small genera into larger ones. In New Caledonia, 78 endemic species were recognized and placed in 14 genera by Aubréville (1967)
, but five genera were retained by Pennington (1991)
, who transferred many species to either Niemeyera or Pouteria.
The first phylogenetic study of Sapotaceae, based on a data set of molecular sequences from the cpDNA gene ndhF, found three main evolutionary lineages in the family (Anderberg and Swenson, 2003
). Sarcosperma is sister to the rest of the family, which forms two large clades. Taxa belonging to the tribes Chrysophylleae and Omphalocarpeae form one of these clades, and the other is formed by three tribes: Isonandreae, Mimusopeae, and Sideroxyleae. As a sequel, Swenson and Anderberg (in press) analyzed a combination of molecular and morphological data using a similar sample of taxa. This study confirmed the basal diversification between the two groups of tribes and provided further information on the phylogenetic relationships within the two clades. Within Chrysophylleae, the results of the study support not only the monophyly of Delpydora, Leptostylis, Micropholis, and Pradosia, but also the polyphyly for the two largest genera, Chrysophyllum and Pouteria. They further suggested that Pouteria section Oligotheca, earlier known as the genus Planchonella, should be resurrected. A weak phylogenetic signal also indicated that all sampled species from Australasia formed a clade. However, insufficient variability in the gene sequences and a high level of homoplasy in morphological characters resulted in an overall low resolution of the phylogeny within Chrysophylleae, particularly regarding the relationships among Pouteria species. To clarify species relationships and the evolutionary history of this group, we explored here the phylogenetic signal in a more variable genetic marker.
The remarkable richness of species and high levels of endemism in many tropical islands offer promising models for evolutionary studies. Mechanisms of evolution can be studied in detail using island communities because of their geographic isolation and highly diverse climatic and ecological conditions (Eibl et al., 2001
; Bramley et al., 2004
). It is not surprising, therefore, that tropical islands attract a high level of attention from specialists concerned with a broad range of problems, from systematics and evolution to conservation.
The islands of New Caledonia in the southwest Pacific represent an ideal context to study the evolutionary origin of local diversity. There are c. 3250 native species of seed plants in an area of about 17 000 km2, which has a remarkably high level of endemism; that is, 76% at the species level and about 14% at the genus (Lowry, 1998
; Morat et al., 2001
). Sapotaceae endemism in New Caledonia, at close to 100%, is not an exception (Jaffré et al., 2001
).
New Caledonia is a part of the former East Gondwana continent, and its geological history suggests a long-standing isolation of 85–65 million years (i.e., the Late Cretaceous) from the main continental landmasses of Australia and/or New Zealand, associated with the opening of the Tasman Sea and New Caledonia–Taranaki Basin (Knox, 1982
; Veevers and Li, 1991
). Some authors have suggested that the island could have experienced a substantial or total submergence during the Early Eocene (Lillie and Brothers, 1970
; Aitchison et al., 1995
). This does not explain, however, the presence of several ancient lineages of gymnosperms and angiosperms, most of them yet to be studied in a biogeographic context, but Nothofagus (Nothofagaceae) certainly indicate a long-standing presence in New Caledonia (Swenson et al., 2001
). Nevertheless, the outstanding level of endemism on the islands might be a result of both historical vicariance and long-distance dispersal, with Australia and New Guinea as its main sources (Lowry, 1998
). Unfortunately, the difficulties in obtaining palynological data from New Caledonia pre-Quaternary sediments (Jarzen, 1980
) do not allow for a detailed reconstruction of the vegetation history of the territory.
The evolution of biodiversity in tropical forests was long considered to have resulted from a gradual accumulation of species through time with relatively low extinction rates, according to the "museum" model of Stebbins (1974)
. This mode of evolution could be expected in areas with long-term stable environments in these ecological systems. Some recent phylogenetic studies have questioned this view, suggesting an alternative model of rapid speciation in large species-rich genera (Richardson et al., 2001
; Goldblatt et al., 2002
). Both models seem appropriate to explain parts of the evolution of Polyscias (Araliaceae) in New Caledonia (Eibl et al., 2001
). Results of this study suggest that sympatric/allopatric distribution of closely related species, differences in elevation and edaphic conditions, and a recent breakdown in reproductive barriers leading to hybridization, may all have played a role in producing the observed phylogenetic pattern within the genus. A large group of endemic plant species such as Sapotaceae, if included in the framework of studies on New Caledonian biota, has a strong potential to provide valuable information about specific evolutionary patterns on isolated islands.
In accordance with the outlined problems, the main purposes of the present study are to (1) identify an additional molecular marker; (2) explore the phylogenetic relationships among the apparently closely related genera of Chrysophylleae, with emphasis on Australian and New Caledonian taxa; (3) test the monophyly of Niemeyera and Pouteria; (4) test the hypothesis of a single origin of Sapotaceae in New Caledonia against that of multiple colonization events; and (5) provide a more robust phylogenetic framework for the evolution of the New Caledonian taxa and new insights into the patterns of evolution of morphological characters within Australasian Sapotaceae.
MATERIALS AND METHODS
Nomenclature
For tribal classification, we follow Pennington (1991)
and use the names from the latest checklist of the family by Govaerts et al. (2001)
, unless otherwise stated.
Taxon sampling
Sampling of Sapotaceae for this study was designed to provide adequate representation of the New Caledonian endemics and taxa from adjacent areas, including Australia, the Pacific Islands, and the Neotropics. Nearly half of the Chrysophylleae recognized from New Caledonia, 32 of 78 species (Aubréville, 1967
), as well as nine still undescribed taxa, were included. Six species of Pouteria from Australia, one from South America, and one from Africa were sampled to represent the range of the genus. South American Pradosia (three species) have been confused with Pouteria and Niemeyera and were therefore also included, along with one species each of Englerophytum and Magodendron from Africa and New Guinea, respectively. The latter have been thought to be among the closest relatives of Pouteria (Anderberg and Swenson, 2003
; Swenson and Anderberg, in press). Our choice of outgroup, two species of Xantolis, was also based on their results. This genus is sister to all other Chrysophylleae (including Omphalocarpeae). For a list of included taxa, voucher specimens, and accessions see the Appendix in Supplemental Data accompanying the online version of this article.
DNA extraction
DNA was extracted from leaves taken from herbarium specimens or from leaves collected from living plants and dried in silica gel. About 20 mg of leaf tissue was ground in one or two 40-s cycles in a Mini-Bead Beater (BioSpec Products, Bartlesville, Oklahoma, USA) in buffer using the DNAeasy plant DNA extraction kit (Qiagen, Valencia, California, USA). After grinding, extraction proceeded according to the manufacturer's instructions.
Amplification and sequencing
We performed PCR reactions with 10 µM of the primers 18SF (GAA CCT TAT CGT TTA GAG GAA GG) and 26RN (CCG CCA GAT TTT CAC GCT GGG C) designed by C. Rydin (Stockholm University, Sweden). The reactions were carried out in 25 µL of reaction mixture using prepared beads from Pharmacia Biotech (Uppsala, Sweden). The thermal cycling profile was generally that suggested by the manufacturer: 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, 50°C for 30 s, 72°C for 90 s, and an additional cycle at 72°C for 8 min. The amplified fragments were controlled for their quality by electrophoresis in 0.8% agarose gel and purified by application of QIAquick PCR purification kit from Qiagen (VWR International AB, Stockholm, Sweden). We used the Big Dye Terminator Sequencing kit (PE Applied Biosystems, Warrington, Cheshire, UK) and primers 18SF, 26SR, 5.8SC (TGCGTTCAAAGACTCGAT), and 5.8SN (ATCGAGTCTTTGAACGCA) for sequencing reactions. The last group of primers was designed by Youngbae Suh (Seoul National University, Korea) and was used in a few cases to ensure that entire fragments of both internal transcribed spacers (ITS1 and ITS2) and the 5.8S region of 18S–26S nuclear rDNA were sequenced. The annealing sites of these primers are located within the conserved 5.8S region. The resulting fragments were separated and analyzed on an ABI377 Prism Automated DNA Sequencer (Perkin Elmer, Applied Biosystems, Warrington, Cheshire, UK).
Assembly and alignment
We assembled sequences of amplified DNA fragments and cross-checked them against sequence chromatograms with Staden software (Staden et al., 1998
). The resulting sequences were aligned with the ClustalX program (Thompson et al., 1997
) included in BioEdit software (Hall, 2004
). The final alignment was corrected manually to minimize the number of the total mutational changes. The 60 sequences of rDNA from our study were submitted to GenBank (accession numbers AY552102– AY552161). The aligned data matrix and information used in analyses can be obtained from the first or second author upon request.
The complete amplified sequences, comprising parts of 18S and 26S rDNA, and the complete 5.8S rDNA, ITS1, and ITS2, were subjected to cladistic analyses. Two relatively short continuous fragments (bp positions in the final alignment: 291–301 and 341–352) could not be aligned unambiguously and were therefore excluded, but we also carried out additional analyses on the complete set of characters to test the effects of the short excluded fragments. Gaps were treated as missing data within the alignment, but they also were coded as additional characters following the simple method of Simmons and Ochoterena (2000)
. These authors argue that gaps should be treated as homologous binary characters if they are of the same position, length, and sequence (insertions), but as separate characters if they differ in length, sequence, or position. To explore the effect of additional gap characters, we carried out analyses both with and without gaps.
Phylogenetic analysis
Jackknife analyses (Farris et al., 1996
) were performed, as implemented in PAUP* version 4.0b10 (Swofford, 2002
), on sequence data, both including and excluding gaps coded as characters. We used the following settings: 1000 jackknife replicates with a single random addition sequence, TBR branch swapping, collapse of branches if minimum length is zero, and steepest descent not in effect. To avoid overflow when gaps coded as characters were deleted, the settings were changed to 10 000 jackknife replicates, each with 10 random additions, and MULTREES not in effect. Jackknifing investigates the structure in a matrix without permutation, but excludes an assigned fraction of characters, here set to 35%. Groups with support frequencies below 50% are not recognized.
Estimation of the oldest differentiation
To estimate pairwise genetic distances among taxa from our sample, we used several models of substitutions for combined sequences of ITS1 and ITS2, as implemented in PAUP, from the simplest uncorrected p value to the most complicated general time-reversible (GTR) model. Mean pairwise genetic distances were estimated within strongly supported monophyletic groups, where an Australian species was sister to a New Caledonian clade. The estimates were between an Australian species and a New Caledonian group and were used to calculate the oldest Australian/New Caledonian disjunctions. We consider the highest values of differentiation and the slowest published rates of substitutions in ITS, 1.72 x 10–9 per site per yr (Richardson et al., 2001
), as reasonably conservative estimates of the oldest ages.
We refrained from carrying out more sophisticated analyses of genetic differentiation because applying a molecular clock in this study would be problematic for several reasons. First, we lack reliable calibration points because of the paucity of fossil records of Sapotaceae, especially in New Caledonia. Second, significant mutation rate heterogeneity in rDNA, putatively associated with edaphic conditions in New Caledonia, has been demonstrated in Metrosideros, a genus whose members have similar life histories (Wright et al., 2004
). We believe that the issues of the mutation rate heterogeneity and applicability of a molecular clock to study trans-Coral Sea disjunctions involving the New Caledonian flora require a special study, which we will pursue elsewhere.
RESULTS
The jackknife analysis of the partial sequence of rDNA included 60 taxa and 906 characters, of which 806 were from aligned sequences (insertions and ambiguously aligned fragments excluded) and 100 were from the matrix of coded gaps. Of the total number of used characters, 489 were constant, 166 autapomorphic, and 251 potentially parsimony informative. The tree topology of the jackknife analysis, in which gaps are coded as characters, does not conflict with that from the jackknife analysis in which only sequence data were included. However, inclusion of information from gaps resulted in a much better resolution and increased support values for some nodes. Therefore, we will discuss the results of the combined analysis of sequence and gap data. The arrows in Fig. 1 indicate which branches receive support above the 50% level when gap-coded characters are included.
|
in his treatment of the family, e.g., Chrysophyllum, Leptostylis, Pichonia, and Pradosia. On the other hand, the results also show that the circumscription of genera such as Niemeyera, Pouteria, and Pycnandra sensu Pennington (1991)
, and of Iteiluma, Ochrothallus, Planchonella, Pycnandra, and Pyriluma, as circumscribed by Aubréville (1967)
, do not correspond to monophyletic groups. The problems of both systems are highlighted by a strongly supported group (100% jackknife support) of five taxa (clade B in Fig. 1), which includes four species belonging to two genera that were accepted by Pennington (Pichonia and Pouteria), and one taxon (Planchonella daenikeri) of doubtful generic affiliation. The same five species were placed in four genera by Aubréville (1963
, 1964
, 1967)
, i.e., Pichonia, Planchonella, Rhamnoluma, and Sersalisia.
Biogeographical patterns
In contrast to earlier systematic treatments of Sapotaceae, in which taxa such as Chrysophyllum and Pouteria often are present in two or more geographic areas, our results are mostly congruent with geography. All species from Australia and New Caledonia appear in a monophyletic group, albeit with a rather weak jackknife support (72%, clade A in Fig. 1). This clade is sister to Magodendron mennyae, the only species in this sample endemic to New Guinea. All species from the other areas, such as Africa, continental Asia, and South America/Caribbean, are more basal in relation to this group.
Within the Australian/New Caledonian clade (clade A), one pattern is repeated three times: an Australian species is sister to a clade of several New Caledonian species (clades B, Cb, and D). The only two exceptions to this pattern are in clade Ca, where Pouteria arnhemica from Australia is placed with two New Caledonian species (Pouteria sphaerocarpa and Pyriluma dothioense) in a trichotomy, albeit with weak jackknife support; and another clade of two endemics from New Caledonia, Pouteria baueri and Pouteria rubicunda, both formally included in the genus Beccariella by Aubréville (1967)
.
Pairwise divergence between Australian and New Caledonian species
The strongest value of the mean pairwise genetic differentiation between an Australian species and a New Caledonian clade was obtained by application of the GTR model, which takes into consideration unequal substitution rates between nucleotides. The values varied from 5.4% for clade B to 9.0% for clade D (Table 1). Differences in the mean estimates based on various models of molecular evolution were relatively small. For example, the weakest values were provided by the uncorrected p value, which corresponds to the simplest model and varied between 5.2% and 8.0% for clades B and D, respectively. We observed the strongest GTR model-based individual differentiation between an Australian and a New Caledonian species belonging to clade D (of 36 total pairwise comparisons) between Niemeyera whitei and an undescribed species of Leptostylis (10.8%).
|
Systematics of Chrysophylleae
Our jackknife analysis found no support for the monophyly of Pouteria, nor does it suggest that Chrysophyllum, Englerophytum, or Pradosia are closely related to Australasian Sapotaceae (Fig. 1). A clade corresponding to Pradosia was recovered with maximum jackknife support, but was found in a basal polytomy among other genera from lack of synapomorphies in the variable rDNA sequence data. Similarly, the two species of Pouteria from Africa and South America group together with a jackknife value of only 52%. In the broad analysis of Sapotaceae (Swenson and Anderberg, in press), Pouteria proved to be polyphyletic and the two species from Africa and South America sampled here belong to different evolutionary lineages conforming to the two geographic areas. In the present study, with limited sampling from outside Australia and New Caledonia, a weak phylogenetic signal pulls them together, but they are phylogenetically distant from Pouteria section Oligotheca of Australasia. Hence, the broad treatment of Pouteria, suggested by Baehni (1942)
and partly adopted by Pennington (1991)
and Vink (2002)
, is not supported by the most recent cladistic analyses based on molecular data from the cpDNA gene ndhF (Anderberg and Swenson, 2003
), nor by combined molecular and morphological data (Swenson and Anderberg, in press), nor by the rDNA molecular data used here.
Clade B (Fig. 1) gained maximum support (100% jackknife value) and corresponds to the small genus Pichonia, formally circumscribing only one species growing in New Caledonia (Aubréville, 1967
). Pennington (1991)
expanded the genus to six when he transferred all New Caledonian taxa of Rhamnoluma to Pichonia. This action is supported by our analysis, but our results also indicate that not all members of this clade, i.e., Pichonia, are yet identified. We can here add the Australian species Pouteria sericea, and the so far improperly classified species Planchonella daenikeri from New Caledonia (nomenclatural action will be taken elsewhere). Pennington (1991)
suggested a close relationship between Pichonia and Pouteria, or possibly a section thereof. It is evident from this study, as well as from that of Swenson and Anderberg (in press), that reducing Pichonia to Pouteria would be an inappropriate solution, an act violating the primary principle of monophyly, because the type species of Pouteria is confined to South America and belongs to another part of the phylogeny. Pichonia has several morphological synapomorphies to explore in a more expanded circumscription, including staminodes fixed at the same level as the stamens and one-seeded fruits whose seeds lack an endosperm. This combination of characters is rare in Chrysophylleae and lends strong support for recognition of Pichonia as a distinct genus. A phylogenetic analysis of additional species of Australasian Sapotaceae with similar combinations of characters, however, is needed to confirm whether the full composition of the group has been identified.
Clade C (Fig. 1) is strongly supported (97%) and corresponds to Pouteria section Oligotheca. These species were earlier recognized by some authors (including Aubréville, 1967
) as Planchonella, but were reduced to Pouteria by Pennington (1991
, p. 192) who argued that he "was not able to find any characters of vegetative, floral, fruit or seed morphology to justify the separation of these species [i.e., New World Pouteria] as a unit from the Asian ones." It is probably true that clear-cut morphological synapomorphies may not exist to separate the American group of Pouteria from the species previously recognized as Planchonella, which are distributed in Australia and the southwest Pacific. However, our results suggest that the presence/absence of endosperm and exserted/included radicle may be useful characters to differentiate between these genera. All taxa of section Oligotheca included here, as far as known, have an endosperm and an exserted radicle, whereas most Pouteria species lack an endosperm and have an included radicle (Pennington, 1991
). The presence of endosperm in 10 of about 170 Pouteria species from tropical America could easily be a result of parallel evolution. On the other hand, the absence of endosperm in the seeds of Australian and New Caledonian Sapotaceae is possibly a homoplastic synapomorphy for the clades B and D. However, the distribution of this character is poorly understood. First, fruits of several species such as Planchonella daenikeri have never been observed. Second, the distinction is not often clear between the fully non-endospermous condition, in which seeds have an embryo with plano-convex cotyledons such as in Pouteria dominigensis, and the typical endospermous seed with foliaceous cotyledons (e.g., Pouteria australis). A series of intermediate conditions has been found in several species such as Pouteria annamensis, P. macrantha, and P. malaccensis, all from tropical Asia and representatives of section Oligotheca (Pennington, 1991
). There is probably a persistent parallelism and homoplasy in evolution of this and many other characters within Sapotaceae (Swenson and Anderberg, in press).
Clade D (Fig. 1) is another strongly supported clade. Here, our cladistic analyses of rDNA sequences found no support for the broad circumscription of Niemeyera suggested by Pennington (1991)
, but rather suggest close relationships among the New Caledonian taxa that have been placed in Leptostylis, Niemeyera, and Pycnandra. Leptostylis is a well-diagnosed genus possessing several morphological synapomorphies (Aubréville, 1967
; Pennington, 1991
), and its monophyly received strong jackknife support. However, our results indicate that the other two genera, as circumscribed by Pennington (1991)
, are not monophyletic. Members of Pycnandra appear in two different clades. One of these (clade E), represented by Pycnandra comptonii, P. controversa, P. neocaledonica, P. griseosepala, and two undescribed species, has strong jackknife support (95%), and its members are morphologically similar to the type species of the genus, P. benthamii. This group may thus be taken to represent Pycnandra sensu stricto. The second group comprises P. carinocostata, two specimens of P. fastuosa, and an additional undescribed taxon. This group is more closely related to the two Niemeyera species included in the study, N. francei and N. sessilifolia, than to Pycnandra sensu stricto, and could thus merit recognition as a separate genus. The morphology of these plants has not yet been studied in detail, so potential synapomorphies cannot at present be identified. Moreover, because the two clades comprising Pycnandra species are part of a large polytomy within clade D (Fig. 1), it is not possible to assess whether they represent sister groups. Pennington (1991)
, following some previously suggested ideas, considered dividing Pycnandra into two groups. His argumentation was, however, mainly based on seed characters (smooth vs. ruminate seed coat and wide vs. narrow seed scar) that are not known for most Pycnandra species. It is therefore impossible to evaluate whether the two groups of species identified in our analysis are congruent with Pennington's rationale for dividing the genus in two.
The situation in Niemeyera is less ambiguous. In our results, the nine species of Niemeyera sensu Pennington belong to no fewer than five different clades, and the genus is thus clearly polyphyletic as presently circumscribed. Three of the species included in our sample represent type species of the segregate genera Corbassona, Ochrothallus, and Sebertia, genera earlier accepted by Aubréville (1967
; cf. Fig. 1). Each of the clades containing species of Niemeyera has strong jackknife support (94–100%) and can possibly be recognized using diagnostic morphological characters. One interesting feature, for instance, is that Niemeyera acuminata, which corresponds to Aubré ville's genus Sebertia, has a very characteristic bluish sap when growing on ultramaphic soils.
Ochrothallus sensu Aubréville (1967)
is endemic to New Caledonia and comprises nine species, three of which are included in our sample. Aubréville based this genus on the presence of anisomerous flowers with 5 sepals and 7–10 corolla lobes. The type of Ochrothallus currently referred to as N. sessilifolia forms a well-supported clade with N. francei, which could be retained with a more restricted generic concept, with the exclusion of at least one species, O. gordoniifolius. This species attaches with weak support at the base of two Niemeyera species, a clade with problematic affinity, because the type of the genus is the Australian species N. prunifera (F. Muell.) F. Muell., a taxon most probably related to N. whitei (Fig. 1). Thus, until additional species of Ochrothallus are sampled, we cannot evaluate the composition nor identify possible synapomorphies or diagnostic features for this genus.
Corbassona sensu Aubréville (1967)
is a small genus endemic to New Caledonia of only two species (both sampled here). Aubréville distinguished this genus from Niemeyera by the form and size of the seed scar, and it further differs by its growth habit (shrubs or small trees vs. large trees). The two species form a strongly supported group, but like with many other clades identified here, they are part of a sub-basal polytomy among the New Caledonian species within clade D.
Our results strongly suggest that a core group of New Caledonian Sapotaceae is monophyletic and circumscribes taxa currently placed in Leptostylis, Niemeyera, and Pycnandra (clade D). Aubréville (1967)
recognized species from the very same clade in at least seven genera, that is, Corbassona, Leptostylis, Niemeyera, Ochrothallus, Pycnandra, Sebertia, and Trouettea. We can conclude that this group needs future, detailed studies on a broader sample of all segregate genera recognized by Aubréville (1967)
to unravel the generic limits and explore relationships within this large clade of New Caledonian Sapotaceae.
Biogeography of Chrysophylleae
In the absence of adequate fossil records, phylogenetic studies of biota can shed some light on its history. Phylogenetic reconstructions within plant groups from New Caledonia and the Australia/Coral Sea region lend support for two main biogeographic patterns, which appear to have shaped the unique composition of the New Caledonian flora. Nothofagus (Manos, 1997
; Swenson et al., 2001
) and Arillastrum (Ladiges et al., 2003
) probably represent examples of vicariance diversification in accordance with geological hypotheses of the breakup of Gondwana, whereas Polyscias section Tieghemopanax (Eibl et al., 2001
) and Mearnsia (Wright et al., 2004
) indicate a more recent colonization (or a series of such events) by means of long-distance dispersal from Australia and New Zealand, respectively.
Tree topology from our jackknife analysis suggests monophyly for the group comprising all genera of Chrysophylleae distributed in Australia and on southwest Pacific islands (clade A). Moreover, in accordance with the general trends in New Caledonian flora (Lowry, 1998
), our results indicate strong affinities between Sapotaceae in Australia and New Caledonia. Because of the lack of fossil data and reliable estimates of mutation rates for rDNA, we cannot precisely determine the age of differentiations between taxa from these areas. Application of previously suggested rates of substitutions in rDNA to mean pairwise values of genetic differentiation may, however, provide some indication. The slowest of the published rates (recently reviewed by Richardson et al., 2001
) may provide a conservative estimate of differentiation age between sister lineages from different landmasses. Using this rate, we calculate that the oldest differentiation between sister Australian and New Caledonian taxa might have taken place approximately 32.4 mya. This value is not much older than the Oligocene-Miocene transition and indicates that most of the disjunctions would be associated with the Neogene, long after New Caledonia became isolated at the end of the Cretacous, some 65 mya (McLoughlin, 2001
). If a vicariance event, such as New Caledonia drifting from Australia, is instead invoked to explain the oldest disjunction (clade D), then the rDNA substitution rate must have been about five times slower than that estimated for other families, or about two times slower than the slowest rate so far reported (Richardson et al., 2001
). If the estimated maximum age just suggested for the New Caledonian species pairs in clades B, Cb, or D is correct, it follows that the differentiation within this archipelago has taken place in situ and was preceded by a dispersal event. In fact, such a pattern between Australia and New Caledonia is well supported and repeated at least three times in our phylogenetic reconstruction. Multiple dispersal events between the two areas are therefore suggested.
Our phylogeny does not, however, identify a hypothetical center of origin of Australasian Chrysophylleae. To solve this problem and better understand the biogeographic mechanism in the region, our sample must be expanded to include additional species from Australia and neighboring areas. For example, species of Pouteria sect. Pierrisideroxylon, with representatives in New Guinea and the southwest Pacific islands (Vink 2002
), will allow us to include more taxa and areas of the region. Morphological characters such as presence of copious endosperm, thin foliaceous cotyledons, and exserted radicle, suggest that sect. Pierrisideroxylon is distinct from South American Pouteria and is probably more closely related to the large group of Australian/southwest Pacific taxa, which also have this combination of character states (Pennington, 1991
). Many of the species belonging to this section were also included in Planchonella, together with other Australian and New Caledonian taxa in earlier systematic treatments of Sapotaceae (cf. Govaerts et al., 2001
). If the assumption of close relationship between sect. Pierrisideroxylon and Australasian representatives of sect. Oligotheca proves to be correct, then the trans–Coral Sea disjunction would be the main biogeographical pattern within Australasian Chrysophylleae.
FOOTNOTES
1 The authors thank two anonymous reviewers for critically reading the manuscript. The Swedish Research Council is acknowledged for grants provided to Ulf Swenson for phylogenetic and biodiversity studies of Sapotaceae, and to Arne A. Anderberg for phylogenetic studies of Ericales. Tanguy Jaffré, Gordon McPherson, and Pete Lowry are thanked for fresh leaf material provided from the field. The authors are grateful to Mari Källersjö, Pia Eldenäs, and Bodil Cronholm for various help and discussions in the lab. 
4 Author for correspondence (E-mail: ulf.swenson{at}nrm.se
)
LITERATURE CITED
Aitchison J. C. G. L. Clarke S. Meffre D. Cluzel 1995 Eocene arc-continent collision in New Caledonia and implications for regional Southwest Pacific tectonic evolution. Geology 23: 161-164
Anderberg A. A. U. Swenson 2003 Evolutionary lineages in Sa\[chpotaceae (Ericales): a cladistic analysis based on ndhF sequence data. International Journal of Plant Sciences 164: 763-773[CrossRef]
Aubréville A. 1963 Notes sur les Poutériées Océaniennes (Sapotacées). Adansonia 3: 327-335
Aubréville A. 1964 Les Sapotacées: taxonomie et phytogéographie. Adansonia Mémoires 1: 1-157
Aubréville A. 1967 Flore de la Nouvelle Calédonie et Dépendances. 1. Sapotaceae, Paris, France
Baehni C. 1942 Mémoires sur les Sapotacées. 2. Le genre Pouteria. Candollea 9: 147-476
Bramley G. L. C. R. T. Pennington R. Zakaria S. S. Tjitrosoedirdjo Q. C. B. Cronk 2004 Assembly of tropical plant diversity on a local scale: Cyrtandra (Gesneriaceae) on Mount Kerinci, Sumatra. Biological Journal of the Linnean Society 81: 49-62[CrossRef]
Eibl J. M. G. M. Plunkett P. P. Lowry II 2001 Evolution of Polyscias sect. Tieghemopanax (Araliaceae) based on nuclear and chloroplast DNA sequence data. Adansonia, 3ème séries 3 23: 23-48
Farris J. S. V. A. Albert M. Källersjö D. Lipscomb A. G. Kluge 1996 Parsimony jackknifing outperforms neighbor-joining. Cladistics 12: 99-124[CrossRef][Web of Science]
Goldblatt P. V. Savolainen O. Porteous I. Sostaric M. Powell G. Reeves J. C. Manning T. G. Barraclough M. W. Chase 2002 Radiation in the Cape flora and the phylogeny of peacock irises Moraea (Iridaceae) based on four plastid DNA regions. Molecular Phylogenetics and Evolution 25: 341-360[CrossRef][Web of Science][Medline]
Govaerts R. D. G. Frodin T. D. Pennington 2001 World checklist and bibliography of Sapotaceae. Royal Botanic Gardens, Kew, UK
Hall T. 2004 BioEdit version 7.0.0. Distributed by the author, website: www.mbio.ncsu.edu/BioEdit/bioedit.html
Jaffré T. P. Morat J.-M. Veillon F. Rigault G. Dagostini 2001 Composition and characterisation of the native flora of New Caledonia. IRD, Nouméa: IRD, Documents Scientifiques et Techniques. II 4. Institut de Recherche pour le Développement, Nouméa, New Caledonia
Jarzen D. M. 1980 Héritage perdu, New Caledonia September/October 1979. Canadian Association of Palynologists Newsletter 3: 4-5
Knox G. J. 1982 Taranaki basin, structural style and tectonic setting. New Zealand Journal of Geology and Geophysics 25: 125-140[Web of Science]
Ladiges P. Y. F. Udovicic G. Nelson 2003 Australian biogeographical connections and the phylogeny of large genera in the plant family Myrtaceae. Journal of Biogeography 30: 989-998[Web of Science]
Lillie A. R. R. N. Brothers 1970 The geology of New Caledonia. New Zealand Journal of Geology and Geophysics 13: 145-183
Lowry II P. P. 1998 Diversity, endemism, and extinction in the flora of New Caledonia: a review. In C.-I. Peng and P. P. Lowry II [eds.], Proceedings of the International Symposium on Rare, Threatened, and Endangered Floras of Asia and the Pacific Rim, 181–206. Academia Sinica, Taipei, Taiwan
Manos P. S. 1997 Systematics of Nothofagus (Nothofagaceae) based on rDNA spacer sequences (ITS): taxonomic congruence with morphology and plastid sequences. American Journal of Botany 84: 1137-1155[Abstract]
McLoughlin S. 2001 The breakup history of Gondwana and its impact on pre-Cenozoic floristic provincialism. Australian Journal of Botany 49: 271-300[CrossRef]
Morat P. T. Jaffré J. M. Veillon 2001 The flora of New Caledonia's calcareous substrates. Adansonia, 3ème séries 3 23: 109-127
Pennington T. D. 1991 The genera of Sapotaceae. Royal Botanic Gardens, Kew, UK
Richardson J. F. R. T. Pennington T. D. Pennington P. M. Hollingsworth 2001 Rapid diversification of a species-rich genus of neotropical rainforest trees. Science 293: 2242-2245
Simmons M. P. H. Ochoterena 2000 Gaps as characters in sequence-based phylogenetic analyses. Systematic Biology 49: 369-381[CrossRef][Web of Science][Medline]
Staden R. K. Beal J. K. Bonfield 1998 The Staden package. In S. Misener and S. A. Krawets [eds], Computer methods in molecular biology 132: bioinformatics methods and protocols, 115–130. Humana Press Inc., Totowa, New Jersey, USA
Stebbins G. L. 1974 Flowering plants, evolution above the species level. Harvard University Press, Cambridge, Massachusetts, USA
Swenson U. A. Backlund S. McLoughlin R. S. Hill 2001 Nothofagus biogeography revisited with special emphasis on the enigmatic distribution of subgenus Brassospora in New Caledonia. Cladistics 17: 28-47[CrossRef][Web of Science]
Swenson U. A. A. Anderberg In press Phylogeny, character evolution, and classification of Sapotaceae (Ericales). Cladistics
Swofford D. L. 2002 PAUP*: phylogenetic analysis using parsimony (* and other methods), version 4. Sinauer, Sunderland, Massachusetts, USA
Thompson J. D. T. J. Gibson F. Plewniak F. Jeanmougin D. G. Higgins 1997 The ClustalX Windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 24: 4876-4882
Veevers J. J. Z. X. Li 1991 Review of seafloor spreading around Australia. II. Marine magnetic anomaly modeling. Australian Journal of Earth Sciences 38: 391-408[CrossRef][Web of Science]
Vink W. 2002 Some Malesian species of Pouteria (Sapotaceae). Blumea 47: 95-147[Web of Science]
Wright S. D. R. D. Gray R. C. Gardner 2004 Energy and the rate evolution: inferences from plant rDNA substitution rates in the Western Pacific. Evolution 57: 2893-2898[Web of Science]
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  J. W. Clayton, P. S. Soltis, and D. E. Soltis Recent Long-Distance Dispersal Overshadows Ancient Biogeographical Patterns in a Pantropical Angiosperm Family (Simaroubaceae, Sapindales) Syst Biol, August 14, 2009; (2009) syp041v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Grandcolas, J. Murienne, T. Robillard, L. Desutter-Grandcolas, H. Jourdan, E. Guilbert, and L. Deharveng New Caledonia: a very old Darwinian island? Phil Trans R Soc B, October 27, 2008; 363(1508): 3309 - 3317. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Goldberg, S. A Trewick, and A. M Paterson Evolution of New Zealand's terrestrial fauna: a review of molecular evidence Phil Trans R Soc B, October 27, 2008; 363(1508): 3319 - 3334. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. E. Smedmark and A. A. Anderberg Boreotropical migration explains hybridization between geographically distant lineages in the pantropical clade Sideroxyleae (Sapotaceae) Am. J. Botany, September 1, 2007; 94(9): 1491 - 1505. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
