| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
2Department of Biology, Wake Forest University, Winston-Salem, North Carolina 27109-7325 USA; 3New York Botanical Garden, Bronx, New York 10458-5126 USA
Received for publication May 8, 2001. Accepted for publication August 28, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
33 currently recognized) of Vaccinium. Results indicated that, in general, traditional generic circumscriptions were not corroborated, but several well-supported clades were found: an Andean clade (including at least some members of 18 of the 23 Neotropical genera sampled), a Meso-American/Caribbean clade, an East Malesian clade (including the Old World taxa Paphia and Dimorphanthera), an Agapetes clade comprised of some Asian Vaccinium and Agapetes, and a Bracteata-Oarianthe clade (Vaccinium spp.). This study is the first to address phylogenetic relationships among members of Vaccinieae on a worldwide basis and the results indicate that the large genus Vaccinium is not monophyletic, but likely represents a grade group out of which several more specialized clades evolved.
Key Words: blueberries • Ericaceae • matK • nrITS • tropics • Vaccinieae • Vaccinium
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
and http://www.nybg.org/bsci/res/lut2/ for information on Neotropical blueberries). Currently the Neotropical Vaccinieae have 30 recognized genera, many more than the four or five genera recognized in the Old World tropics. Vaccinium is the only genus in the tribe that occurs in both the temperate and tropical zones of the Old and New World. Calyx, corolla, and especially stamen morphology are often emphasized in the delimitation of genera within Vaccinieae (Sleumer, 1941
; Luteyn, 1991
) although characters of the fruit, seed, and vegetative parts are also considered important. However, many characters traditionally used to delimit genera fail to adequately distinguish among them (Stevens, 1972
) and the limits of most genera are poorly understood (Stevens, 1985
). This study addresses the phylogenetic relationships within Vaccinieae with particular attention to the identification of major clades within the group. It also addresses the monophyly of Vaccinium because this taxon is critical to understanding the evolutionary relationships in tropical Vaccinieae. Sequence data from the nuclear internal transcribed spacer region (nrITS) and the chloroplast matK gene are used to investigate the phylogenetic relationships within the tribe. Previous studies (Kron and King, 1996
; Kron, 1997
; Kron et al., 1999
) have shown matK and nrITS to be useful for phylogenetic analysis in Ericaceae. This investigation is intended to lay the groundwork for future, more detailed studies within the group. |
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
33 sections of the genus (as recognized by Sleumer [1941
] with modifications by Airy Shaw [1948
], Stevens [1969, 1972, 1974
], Vander Kloet [1988
], and Vander Kloet and Dickinson [1992
]). Taxa were selected to represent the range of geographical (with the exception of the African/Madagascaran taxa, which were unavailable for this study) and morphological diversity within Vaccinieae and include temperate and tropical taxa from both the Old and New World. Ninety-three species (excluding outgroup taxa) were analyzed for the individual and combined analyses of matK and nrITS (Table 1). Leaf material for the Neotropical species was obtained in the field by J. L. Luteyn. Other material was obtained from the living collections of S. P. Vander Kloet (Acadia University, Nova Scotia, Canada) and G. Argent (Royal Botanical Garden Edinburgh, Scotland, UK).
|
) using the Epicentre MasterPure Plant Leaf DNA Purification kit (Epicentre Technologies, Madison, Wisconsin, USA) or the modified cetyltrimethylammonium bromide method of Doyle and Doyle (1987)
. The nrITS and matK regions were amplified from the same accession for each species using standard polymerase chain reaction (PCR) protocol. Amplified products were cleaned via the Qiagen Qiaquick PCR Cleanup protocol (Qiagen, Santa Clarita, California, USA) and sequenced at the Wake Forest University School of Medicine Automated DNA Sequencing Facility, Winston-Salem, North Carolina, USA. Sequences were downloaded and edited in Sequencher 3.1 (GeneCodes, Ann Arbor, Michigan, USA). The edited sequences were aligned manually.
Phylogenetic analysis
Outgroup taxa were chosen based on previous analyses of Ericaceae that included taxon sampling throughout the family (Kron et al., 1999
; Kron, Judd, and Crayn, 1999
). These analyses, and a more restricted analysis of the wintergreen group and its relatives (Powell and Kron, 2001)
, indicate that Leucothoe fontanesiana and the Andromeda polifolia + Zenobia pulverulenta clade are likely the closest relatives to Vaccineae. The monophyly of Vaccinieae is well supported by both molecular (Kron et al., 1999
; Kron, Judd, and Crayn, 1999
) and morphological data (Kron, Judd, and Crayn, 1999
and unpublished data). The most noticeable synapomorphy for the clade is the presence of an inferior ovary.
Due to the large size of the individual and combined data sets, two approaches to phylogenetic analysis were taken. For each data matrix PAUP* 4.0 (Swofford, 1999
) and the Parsimony Ratchet (Nixon, 1999a, b
; available via WINCLADA at www.cladistics.com) were used to construct most parsimonious trees. The PAUP analyses were run under the following parameters: heuristic search, uninformative characters removed, 100 random replicates, MULPARS on, gaps as missing data. Parsimony ratchet analyses were run with uninformative characters deactivated, ten passes of 200 iterations each, 10% of informative characters perturbed, one tree held at each iteration, random constraint level = 10, amb = poly-, gaps as missing data. Bootstrap analysis in PAUP used the "fast bootstap" option with 100 random replicates.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Analysis of the nrITS data resulted in trees 587 steps long (consistency index [CI] = 0.42, retention index [RI] = 0.75) in both the PAUP and parsimony ratchet (ratchet) methods (Figs. 1 and 2). However, the PAUP analysis ran out of memory after >27 h at replication #2 with 46 499 trees saved. Each pass of the ratchet analysis completed within 10 min and a total of 154 trees were found after ten passes (<2 h). The resulting strict consensus tree (Figs. 1 and 2) indicated that more islands of trees had been encountered in the ratchet analysis than in the PAUP analysis. Six clades that were present in the strict consensus of all most parsimonious trees found in the PAUP analysis were not found (i.e., were collapsed) in the ratchet analysis (Figs. 1 and 2). These clades lacked bootstrap support above 50% in the PAUP analysis and were generally confined to the deeper nodes of the tree (Fig. 1). In the ratchet analysis, two clades were present in the strict consensus of 154 trees that were not present (collapsed) in the PAUP analysis. Both clades were confined to the tips of the branches (Fig. 2). A similar pattern of results was obtained in the analysis of the matK data (parameters for analysis were the same as that for the nrITS analyses). The PAUP analysis ran out of memory after >27 h, at replication #1 with 114 100 trees saved. The ratchet found 223 most parsimonious trees in <3 h (ten passes at 15 min each). Both methods resulted in trees 1026 steps long (CI = 0.48, RI = 0.64). The strict consensus (Figs. 3 and 4) of the trees found in the PAUP analysis and the ratchet analysis differed only in the presence of a branch within the "Andean" clade in the PAUP analysis (bootstrap <50%) that is collapsed in the ratchet analysis (Fig. 4). Combined analysis of matK and ITS data for 93 taxa used the same parameters as for the individual analyses in both PAUP and the ratchet (Figs. 5 and 6). The ratchet analysis found 303 most parsimonious trees (Length [L] = 1673, CI = 0.44, RI = 0.67) whereas the PAUP analysis found 13 200 trees one step longer (L = 1674, CI = 0.44, RI = 0.67). Run time for ten passes of the ratchet analysis was <3 h compared to >8 h in PAUP. Comparison of the strict consensus from each analysis shows that the difference between the two results is in the relationships among species of Demosthenesia within the Andean clade (Fig. 6).
|
|
|
|
|
|
The genus Vaccinium is not monophyletic in this analysis. Six of the seven clades in the basal polytomy are comprised mostly or entirely of species of Vaccinium. In two of these clades (Agapetes and Orthaea/Notopora clades) other taxa are nested among species of Vaccinium, while V. poasanum is found within the Meso-American/Caribbean clade (see below). Three of the seven clades consist of the sister pair V. padifolium and V. cylindraceum, a three-taxon clade that includes V. consanguineum, V. meridionale, plus V. floribundum, and lastly, V. tenellum, which is unresolved with respect to the six other clades in the polytomy.
Agapetes and Bracteata-Oarianthe clades
The Agapetes clade comprises several species of temperate and tropical Asian Vaccinium and species of Agapetes (Fig. 5). This clade is strongly supported and comprises two sister clades (A and B). Clade A contains Vaccinium lancifolium as sister to V. filiforme (both of section Rigiolepis) and this clade (moderately supported) is sister to V. caudatifolium (section Galeopetalum). Clade B, well supported, contains Agapetes serpens as the sister group of A. buxifolia, and this clade is sister to V. gaultheriifolium (section Galeopetalum). Agapetes hosseana is the sister group of A. serpens + A. buxifolia + V. gaultheriifolium, and this clade is sister to V. nummularia (section Conchophyllum). These relationships indicate that neither Agapetes subgenus Agapetes nor Vaccinium section Galeopetalum are monophyletic. The Bracteata-Oarianthe clade, comprised of species of Vaccinium from New Guinea and Borneo, is related to the Agapetes clade, but with <50% bootstrap support. Within the Bracteata-Oarianthe clade the sister group relationship of V. cercidifolium to V. leptospermoides + V. summifaucis + V. finisterrae + V. horizontale is strongly supported. These species represent Vaccinium sections Bracteata (V. cercidifolium, V. horizontale, V. summifaucis) and Oarianthe (V. finisterrae, V. leptospermoides). In this analysis neither section Bracteata nor section Oarianthe are monophyletic.
Myrtillus clade
This clade received <50% support in the bootstrap analysis (Fig. 5) and contains species of Vaccinium from sections Hemimyrtillus, Macropelma, and Myrtillus. Costera endertii is also in this clade, although its relationship is unresolved with respect to the Vaccinium species. This analysis suggests that V. sect. Myrtillus is paraphyletic and that section Macropelma is derived from within it. Relationships within the Myrtillus clade will be addressed in more detail in a future study.
Orthaea/Notopora and Vaccinium clades
Orthaea is not monophyletic in this analysis, as O. fimbriata is placed sister to the Cavendishia + Thibaudia jahnii clade within the Andean clade (see below). However, the monophyly of a group containing the remaining species of Orthaea sampled (O. apophysata and O. venamensis) is very strongly supported (Fig. 5). These two species of Orthaea are sister to Notopora with very strong bootstrap support (100%). Successively, Vaccinium crenatum and Gaylussacia + Vaccinium crassifolium are sister to Notopora + Orthaea, although there is no bootstrap support for the latter relationship. However, in the individual analysis of the matK data, V. crenatum and Orthaea + Notopora are successively sister to the East Malesian clade. However, there is no support for this arrangement in the nrITS analysis. This difference is the only occurrence in this study of a strongly supported clade in one analysis (matK) that is not supported in the other (nrITS).
In the combined analysis (Fig. 5) Vaccinium macrocarpon and V. vitis-idaea are sister species (100% bootstrap support) and this clade is sister to the Orthaea/Notopora clade + Gaylussacia + V. crassifolium but this relationship has little support.
East Malesian clade
The close relationship of Paphia meiniana and P. stenantha to Dimorphanthera is strongly supported in this analysis (Fig. 6), and neither Paphia nor Dimorphanthera are monophyletic. Dimorphanthera keysseri is sister to Paphia stenantha and this clade is sister to Paphia meiniana. The remaining species of Dimorphanthera sampled form a strongly supported clade. Relationships within this clade are resolved but weakly supported, with D. megacalyx sister to D. womersleyi and this clade sister to D. amoena. Dimorphanthera dekockii is sister to D. amoena + D. womersleyi + D. megacalyx, and this clade is sister to D. kempteriana.
Andean clade
A large clade comprised of Neotropical taxa that include Cavendishia, Demosthenesia, Diogenesia, Macleania, Psammisia, Satyria, Thibaudia, and several other named genera (Fig. 6) is strongly supported in the combined analysis. This clade is also found in each of the individual analyses with strong support (Figs. 2 and 4). For the most part, the taxa within this clade have much of their diversity concentrated in the montane forests of the northern Andes, so this group is referred to as the "Andean Clade." The Andean clade is strongly supported in both individual analyses, but resolution within it is generally poor. However, resolution was improved in the combined analysis with two clades (A and B) resolved as sister at the base of the Andean clade, although neither has bootstrap support.
Within Clade A, all of the sampled species of Satyria, except S. boliviana, form a monophyletic group. (Satyria boliviana is placed in Clade B with Thibaudia densiflora.) Satyria is sister to a clade (although with <50% bootstrap support) containing Orthaea fimbriata, Thibaudia jahnii, and three species of Cavendishia. In this analysis, Cavendishia is paraphyletic. The close relationship between Satyria and the Cavendishia/Orthaea fimbriata/Thibaudia jahnii clade is also found in the individual analysis of the matK data (Fig. 4), but the two clades are unresolved in the nrITS analysis (Fig. 2). Other relationships indicated in Clade A include a clade comprised of Demosthenesia, Diogenesia, and Themistoclesia. Representatives of these taxa form a weakly supported clade (53%), and of these, only Themistoclesia is supported as monophyletic in this analysis. The two species of Anthopterus sampled form a clade that is sister to Thibaudia floribunda. This clade is the sister group of the Demosthenesia/Diogenesia/Themistoclesia clade, and together they are sister to the Satyria/Cavendishia/Orthaea fimbriata/Thibaudia jahnii clade.
Clade B has little resolution within it, but a few groups are worth noting. Of the twelve genera represented in this portion of the tree, only Macleania is monophyletic if M. megabracteolata is excluded (see below). Sphyrospermum boekii, S. campanulatum, Disterigma pallidum, and D. pernettyoides form a moderately supported clade that is weakly supported as sister to a clade that contains Ceratostema megabracteatum, Sphyrospermum buxifolium, and S. microphyllum. However, more intensive sampling and more data are required before any conclusions can be made about detailed relationships among these and other Neotropical Vaccinieae.
Meso-American/Caribbean Clade
This is a well-supported clade (Fig. 6) comprised of species that are generally found in Central America and the Caribbean. Within this clade, the relationships among species are weakly supported, but it is clear that Macleania megabracteolata has been misclassified, as the remaining Macleania sampled are in the Andean clade. Due to this evidence and morphological similarities with Gonocalyx, this species is now recognized (Luteyn, 2001a)
as Gonocalyx megabracteolatum. The Meso-American/Caribbean clade is sister to the Andean clade in the combined analysis with a bootstrap value of 53%. The same relationship is recovered in the individual analysis of the matK data (Fig. 4; bootstrap <50%), but in the nrITS analysis, the Meso-American/Caribbean clade is sister to a clade containing the East Malesian Paphia and Dimorphanthera (bootstrap <50%). In the combined analysis, the East Malesian clade is sister to the Andean + Meso-American/Caribbean clade, although without bootstrap support.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
) was much faster than the heuristic search used in PAUP* 4.0 (Swofford, 1999
) in the large data matrices analyzed in this study. Although alternate strategies for exploring more islands of parsimonious trees have been used for PAUP, these methods are still slow relative to the speed of the parsimony ratchet approach. However, we had difficulty converting a tree file produced in WINCLADA to a file that could be opened in other graphics software. This was a distinct disadvantage to the WINCLADA/NONA software; it was much easier to save a tree to another format in PAUP. However, for phylogenetic analysis, the ratchet approach was relatively easy to use, and because of its speed and superior ability to explore more islands of most parsimonious trees, it is to be preferred for the analysis of large data sets (see also Soltis et al., 2000)
such as those used in this study. Differences in topologies of the trees produced by the two methods are worth noting. The ratchet explores more islands of most parsimonious trees than the PAUP method (Nixon, 1999a
). This is evident in the decreased resolution seen in the deeper parts of the tree in the ratchet strict consensus compared to the PAUP strict consensus (e.g., nrITS results, Figs. 1 and 2). The differences at the tips of the trees in the PAUP analysis vs. the ratchet (e.g., relationships of Demosthenesia in Fig. 6) are likely due to the fact that the ratchet does not attempt to swap to completion on each island as PAUP does.
The tribe Vaccinieae is one of the largest clades within Ericaceae and is comparable in terms of numbers of species to Rhododendron L. and Erica L. (sensu lato; Oliver, 2000)
. It is interesting to note that variation in floral morphology is arguably comparable between Vaccineae and Erica, with possibly somewhat less variation seen in Rhododendron. Thus the rank at which these clades are recognized does not necessarily provide accurate information about their morphological diversity. In contrast to the Erica and Rhododendron clades, Vaccinieae have been divided into 30–35 genera most of which grow in the Neotropics. Although sampling of Vaccinieae in this study is approximately one-tenth of the described species, the taxa chosen were carefully selected and represent 28 of the 30–35 genera recognized. This study is the first to address phylogenetic relationships within Vaccineae from a global perspective and the combined matK and nrITS analysis has recovered some well-supported clades that can be used as the basis for future studies.
It is clear that Vaccinium is polyphyletic. It fragments throughout the tree in both the individual (Figs. 1–4) and combined (Figs. 5 and 6) analyses. The results of this study support a previous analysis (Kron et al., 1999
) that indicated Vaccinium is probably polyphyletic. The clades of Vaccinium that are supported in this study do not indicate any strong pattern of temperate vs. tropical taxa, nor do they consistently group by continent. More intensive sampling is necessary before any biogeographic conclusions can be made. It may be significant that none of the sampled species of Vaccinium are placed in the Andean clade. Nevertheless, species of Vaccinium do occur in the northern Andes (e.g., V. crenatum). Only one species, V. poasanum, is found in the Meso-American/Caribbean clade. Other Neotropical Vaccinium sampled include V. floribundum, V. consanguineum, and V. meridionale, all of which form a clade that is unresolved with respect to the remainder of the tribe.
Other New World members of Vaccinieae included in the analysis but not part of the large Meso-American/Caribbean + Andean clade include the Orthaea/Notopora clade, the sister pair of Gaylussacia and V. crassifolium, and V. tenellum. Of these, the position of Gaylussacia as sister to V. crassifolium invites further investigation. Gaylussacia possesses a ten-locular ovary (Palser, 1961
) which is distinguished from the pseudo-ten-locular condition found in many other members of Vaccinieae (Stevens, 1985
; Vander Kloet and Dickinson, 1992
; referred to as pseudo-ten-locular because the inner wall of the fruit pushes into the original five locules). In addition, the fruit in Gaylussacia is a drupe with ten stones. These characteristics have been used to mark Gaylussacia as a genus distinct from Vaccinium. Gaylussacia also has a disjunct distribution with some species in the southeastern USA, but most of the species in Brazil (three species are scattered in the Andes). More intensive sampling of Gaylussacia and representatives of pseudo-ten-locular Vaccinium may help to resolve the relationship of Gaylussacia to other temperate Vaccinium.
Relationships among genera within the Neotropical Andean clade indicate that major generic realignments will be necessary. Of the 13 genera that had two or more representatives in the analysis, only Anthopterus, Macleania, and Themistoclesia are supported as monophyletic (each genus represented by two or three species). Some genera are widely fragmented between the Andean and other well-supported clades. Disterigma trimerum is placed in the Meso-American/Caribbean clade, while the other two species of Disterigma sampled (D. pallidum, D. pernettyoides) are placed in the Andean clade with three species of Sphyrospermum (Fig. 6; Clade B). Orthaea is fragmented between the Andean clade and the Orthaea/Notopora clade. Generic fragmentation within the Andean clade includes Thibaudia, which is polyphyletic. Six species of this genus were sampled and none of them form a clade: four species are found in Clade B and the remaining two species are in different clades within Clade A. Although taxon sampling may have a role in the fragmenting of genera seen in this study, studies of members of Macleania and Psammisia that used more data and were more intensively sampled (Kron, Powell, and Luteyn, 2000
; 30 species of Macleania and Psammisia total) indicated that neither genus is monophyletic (this takes the transfer of M. megabracteolata into account). These results would indicate that many of the floral features used in traditional classifications of Vaccinieae are likely subject to convergence or parallelism resulting from strong selection pressure from pollinators. In addition, this group is relatively poorly known, with many species known only from type specimens. Thus generic circumscriptions may have suffered from incremental additions of poorly understood species to already existing taxa.
Despite the general pattern of generic fragmentation indicated by this analysis, some newly identified clades are well supported. These include the Andean clade, Meso-American/Caribbean clade, East Malesian clade, and the Orthaea/Notopora clade. In addition, two clades (Agapetes and the Bracteata-Oarianthe clade) comprised of currently recognized species of Agapetes and Vaccinium are each strongly supported.
The Andean clade is the largest of the groups found in this analysis but does not include all currently recognized Neotropical members of Vaccineae. Other Neotropical members occur in the Orthaea/Notopora clade, the Meso-American/Caribbean clade, and in a clade containing V. floribundum + V. consanguineum + V. meridionale. The Andean clade has never been formally recognized, and future studies will investigate morphological characters that may be synapomorphies for this group.
Within the Andean clade, Satyria is monophyletic, providing S. boliviana is excluded from it. This Neotropical genus has not been monographed, but the clade containing six of the seven species sampled is strongly supported in other analyses with different taxonomic sampling. Relationships of Satyria have been debated in the literature. Stevens (1974)
suggested a close relationship between Dimorphanthera and Satyria based on anther morphology and petiole anatomy. However, a close relationship between these taxa is not supported in this study. Maguire, Steyermark, and Luteyn (1978)
considered Satyria to be an isolated taxon, not closely related to other Neotropical Vaccinieae. MacBride (1944)
suggested that Satyria was more like Thibaudia than Cavendishia, but intermediate between these two genera. Smith (1932)
placed Satyria in his Cavendishia group (including Cavendishia and Orthaea) based on the possession of unequal stamens. He distinguished between Saytria and the other two genera because Satyria has equal filaments and unequal anthers, as opposed to unequal stamens in Cavendishia and Orthaea, which have the filaments and/or anthers unequal. The analysis of Kron et al. (1999)
indicated a sister relationshp between Satyria and Cavendishia, but only one species of each genus was sampled in that analysis. The results presented here also suggest a close relationship between Satyria (excluding S. boliviana) and a clade containing three species of Cavendishia, one species of Orthaea, and one species of Thibaudia. Luteyn (1976, 1983, 1987)
has also suggested a close relationship of Orthaea to Cavendishia. This study does not support a monophyletic Orthaea, but does indicate that some of the species currently placed in Orthaea may be closely related to some species of Cavendishia.
The Meso-American/Caribbean clade contains taxa that are seemingly dissimilar morphologically. However, this clade is supported in both individual analyses and combined analyses (Figs. 2, 4, 6). More comprehensive sampling of Central American and Caribbean taxa and the analysis of morphological features may help to characterize this group of species. The relationship of the Meso-American/Caribbean clade to the Andean clade is weakly supported in both the combined results (Fig. 6) and the matK analysis, but in the nrITS analysis the Meso-American/Caribbean clade was sister to the East Malesian clade (Fig. 2). However, neither of these relationships were strongly supported. It is interesting to note, however, that of the predominantly Old World tropical clades found in this analysis, the East Malesian clade is never closely related to the Agapetes or the Bracteata-Oarianthe clades in any of the individual or combined analyses.
The East Malesian clade is strongly supported as monophyletic, and this result is in agreement with the work of Stevens (1972, 1974)
, who proposed that the Malesian species of Agapetes (= Paphia) were likely more closely related to Dimorphanthera than to other Agapetes. Stevens (1972)
also proposed that the Asian continental Agapetes were more closely related to the Asian Vaccinium. In this study, a strongly supported Agapetes clade contains the Asian members of Agapetes plus several Asian Vaccinium species. Interestingly, the Bracteata-Oarianthe clade (weakly supported as sister to the Agapetes clade) contains several species of Vaccinium from New Guinea. So far as known, members of this combined group all have pseudo-ten-locular ovaries (occurring in a few other Vaccinieae) and superficial phellogen (Stevens, 1972
). Unfortunately, the enigmatic Agapetes scortechinii (King and Gamble) Sleumer could not be included in this study. The other New Guinean taxa in this analysis (Paphia stenantha, Dimorphanthera spp.) fall in the East Malesian clade, but this clade does not include any species of Vaccinium. These results indicate that either Vaccinium will need to be enlarged to include all species currently recognized in the tribe, or be broken up, with some groups of species formerly considered Vaccinium given different names. However, these changes await more complete morphological and molecular data.
This study has identified several well-supported clades within Vaccinieae that generally do not correspond to current taxonomically recognized groups. This is especially true of the species of Vaccinium sampled in the analysis, but is also the case in genera such as Agapetes, Dimorphanthera, Disterigma, Orthaea, and Thibaudia. The results presented here suggest that Vaccinium as currently recognized represents a grade group out of which several clades of Vaccinieae have evolved. In four of these clades many of the members possess larger flowers (i.e., >1 cm long) than those usually found in the currently recognized members of Vaccinium. Taxa from the Old World tropics do not form a single clade, nor do the Neotropical taxa. This suggests that the diversity of species/genera in Vaccinieae seen in the tropics may be due to the independant diversification of several clades rather than a result of a single tropical origin. This study provides little resolution at the base of the tree, and therefore basic biogeographic patterns in Vaccinieae still remain unclear. Future work addressing these issues will require more sampling within Vaccinium and more data.
|
|
FOOTNOTES |
|---|
4 Author for reprint requests (kronka{at}wfu.edu
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Chase M. W. H. G. Hills 1991 Silica gel: an ideal material for field preservation of leaf samples for DNA studies. Taxon 40: 215-220[CrossRef][Web of Science]
Doyle J. J. Doyle 1987 A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11-15
Kron K. A. 1997 Phylogenetic relationships of Rhododendroideae. American Journal of Botany 84: 973-980[Abstract]
Kron K. A. R. Fuller D. M. Crayn P. A. Gadek C. J. Quinn 1999 Phylogenetic relationships of epacrids and vaccinioids (Ericaceae s.l.) based on matK sequence data. Plant Systematics and Evolution 218: 55-65[CrossRef][Web of Science]
Kron K. A. W. S. Judd D. M. Crayn 1999 Phylogenetic analyses of Andromedeae (Ericaceae subfam. Vaccinioideae). American Journal of Botany 86: 1290-1300
Kron K. A. J. M. King 1996 Cladistic relationships of Kalmia, Leiophyllum and Loiseleuria (Phyllodoceae, Ericaceae) based on rbcL and nrITS data. Systematic Botany 21: 17-29[CrossRef][Web of Science]
Kron K. A. E. A. Powell J. L. Luteyn 2000 Molecular systematics of Macleania and Psammisia and their relationship to other members of Vaccinieae. American Journal of Botany 87: (Supplement) 137
Luteyn J. L. 1976 A revision of the Mexican-Central American species of Cavendishia (Vacciniaceae). Memoirs of the New York Botanical Garden 28: 1-138
Luteyn J. L. 1983 Ericaceae—Part 1. Cavendishia.In J. Luteyn [ed.], Flora Neotropica monograph, 1–290. The New York Botanical Garden, Bronx, New York, USA
Luteyn J. L. 1987 Orthaea (Ericaceae-Vaccinieae): new species and redefinition of the genus. Nordic Journal of Botany 7: 31-37
Luteyn J. L. 1991 Key to the subfamilies and genera of neotropical Ericaceae. Nordic Journal of Botany 11: 623-627
Luteyn J. L. 2001a Two new species and two new combinations in Mesoamerican Ericaceae. Brittonia 53, in press
Luteyn J. L. 2001b Biogeography of the neotropical Vaccinieae (Ericaceae): patterns of diversity, adaptation, and endemism. Botanical Review In press
MacBride J. F. 1944 Vaccinium and relatives in the Andes of Peru. University of Wyoming Publications 11: 37-46
Maguire B. J. A. Steyermark J. L. Luteyn 1978 The botany of the Guayana Highland, Ericaceae. Memoirs of the New York Botanical Garden 29: 139-203
Nixon K. C. 1999a The parsimony ratchet, a new method for rapid parsimony analysis. Cladistics 15: 407-414[CrossRef][Web of Science]
Nixon K. C. 1999b WINCLADA (beta), version 0.9.9. Published by the author, Ithaca, NY. [Available at http://www.cladistics.com]
Oliver E. G. H. 2000 Systematics of Ericeae (Ericaceae-Ericoideae): species with indehiscent and partially dehiscent fruits. Contributions from the Bolus Herbarium 19: 1-483
Palser B. F. 1961 Studies of floral morphology in the Ericales. V. Organography and vascular anatomy in several United States species of the Vaccinaceae. Botanical Gazette 123: 79-111[CrossRef]
Powell E. A. K. A. Kron 2001 Phylogenetics of the wintergreen group (Gaultheria, Pernettya, Diplycosia; Ericaceae). Systematic Botany 26: 808–817 [Web of Science]
Sleumer H. 1941 Vaccinioideen-Studien. Botanische Jahrbücher 71: 375-510
Smith A. C. 1932 The American species of Thibaudieae. Contributions from the U.S. National Herbarium 28: 311-547
Soltis D. E. et al 2000 Angiosperm phylogeny inferred from 18s rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnean Society 133: 381-461[CrossRef]
Stevens P. F. 1969 Taxonomic studies in the Ericaceae. Ph.D. dissertation, University of Edinburgh, Edinburgh, Scotland, UK
Stevens P. F. 1972 Notes on the infrageneric classification of Agapetes with four new taxa from New Guinea. Notes from the Royal Botanic Garden Edinburgh 32: 13-28
Stevens P. F. 1974 Circumscription and relationships of Dimorphanthera (Ericaceae) with notes on some Papuasian species. Contribution Herbariorum Australiense 8: 1-34
Stevens P. F. 1985 Notes on Vaccinium and Agapetes (Ericaceae) in Southeast Asia. Journal of the Arnold Arboretum 66: 471-490[Web of Science]
Swofford D. L. 1999 PAUP* 4.0: phylogenetic analysis using parsimony (*and other methods). Sinauer, Sunderland, Massachusetts, USA
Vander Kloet S. P. 1988 The genus Vaccinium in North America. Research Branch Agriculture Canada Publication 1828
Vander Kloet S. P. T. A. Dickinson 1992 The taxonomy of Vaccinium section Hemimyrtillus. Botanical Magazine, Tokyo 105: 601-614[CrossRef][Web of Science]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
