| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
2Departamento de Genética, Instituto de Biociências, Universidade Federal do Rio Grande do Sul, Caixa Postal 15053, 91501-970 Porto Alegre, RS, Brazil; 3Departamento de Botânica, Universidade Federal do Paraná, Caixa Postal 19031, 81531-970 Curitiba, PR, Brazil; 4Centro de Biologia Genômica e Molecular, Faculdade de Biociências, Pontifícia Universidade Católica do Rio Grande do Sul, Ipiranga 6681, 90610-001 Porto Alegre, RS, Brazil; 5Departamento de Botânica, Instituto de Biociências, Universidade Federal do Rio Grande do Sul, Av. Bento Gonçalves, 9500, Prédio 43433, 90509-900 Porto Alegre, RS, Brazil
Received for publication September 12, 2002. Accepted for publication March 28, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
1000 bp), and the rps4 plastid gene (570 bp). Three major clades were highly supported, independent of the marker and phylogenetic method used; one included the subgenera Distephana, Dysosmia, Dysosmioides, Passiflora, and Tacsonioides, a second, the subgenera Adopogyne, Decaloba, Murucuja, and Pseudomurucuja, and a third, the subgenus Astrophea. We call these the Passiflora, Decaloba, and Astrophea clades, respectively. The position of subgenus Deidamioides is undefined. The monophyly of Passiflora could not be statistically corroborated, and the relationships among the major clades and of these clades with the related genera remain unresolved. Our results indicate that a reevaluation of the monophyly of Passiflora and its infrageneric classification is necessary.
Key Words: ITS • Passiflora • Passifloraceae • phylogenetic analysis • rps4 • trnL-trnF
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
, Passifloraceae are divided into two tribes: Paropsieae (with six genera, all Old World species in Africa and Madagascar) and Passifloreae (with 14 genera, five of them present in the New World and nine in the Old World). Judd et al. (1999)
suggested that the monophyly of Passifloraceae is supported most strongly by the presence of a corona in the flowers. The Paraopsieae, which contains shrubs and trees that lack tendrils, probably represent a paraphyletic basal complex within the family. Passifloreae, in contrast, are clearly monophyletic, as evidenced by their vine habit, axillary tendrils, and specialized flowers. The genus Passiflora, which includes the largest number of species described thus far (about 400 for the New World and 20 for Asia; Cervi, 1997
), is included among the Passifloreae. The taxonomy of Passiflora is based on several floral and vegetative structures, leading to a complex taxonomic subdivision in subgenera, sections, and series. As a matter of fact, Killip (1938)
and MacDougal (1994)
asserted that among angiosperms no other group presents such diversity in leaf form. Its flowers also vary widely in size and color, with the corona and perianth diversely oriented and developed, all of which may have arisen as a result of coevolutionary relationships with insect pollinators (MacDougal, 1994
).
According to Killip (1938)
, the genus should be subdivided into 22 subgenera and the Decaloba and Astrophea subgenera into 13 sections. Escobar (1989)
described one additional subgenus. Killip's (1938)
classification was entirely based on morphological characters, especially those of the floral structure. Only two other studies deal with the infrageneric relationships in Passiflora (Sánchez et al., 1999
; de Melo et al., 2001
), but none employed modern phylogenetic methods, DNA sequence data, or suggested any infrageneric grouping.
In the present investigation we analyzed the relationships of 61 species of Passiflora, formally classified in 11 subgenera, and representatives of four other genera, using two noncoding DNA segments, the nuclear ribosomal internal transcribed spacers (ITS), and the plastid trnL-trnF spacers. These regions have been widely used in plant phylogenetics because of their high rate of nucleotide substitutions (Taberlet et al., 1991
; Baldwin, 1992
; Baldwin et al., 1995
; Rauscher, 2002
) and their power in elucidating infrageneric relationships (Gielly and Taberlet, 1996
; Bakker et al., 1998
; Molvray et al., 1999
; Nishikawa et al., 1999
; Hardig et al., 2000
). The rps4 gene (encoding protein 4 of the small plastid ribosomal subunit) was also studied in a smaller number of species to broadly access the major phylogenetic relationships within Passiflora. This gene was chosen mainly because of its slower evolutionary rate when compared with the other two regions studied here and its previously successful use in the phylogenetic reconstruction of Poaceae (Nadot et al., 1994
) and Iridaceae (Souza-Chies et al., 1997
). The data were evaluated using several phylogenetic reconstruction methods. Specific objectives were (a) to examine the nature of the variation in these two differently positioned spacer regions and in a coding region, assessing their phylogenetic utility in Passiflora; (b) to investigate the monophyly of the genus; and (c) to clarify the infrageneric phylogenetic relationships.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
DNA was extracted from fresh leaves (collected directly from nature or derived from seeds grown in the laboratory) or from herbarium material, using the cetyltrimethyl ammonium bromide (CTAB) method (Doyle and Doyle, 1987
). Amplification was done by polymerase chain reaction (PCR), using the primers described by Desfeux and Lejeune (1996)
for ITS and the primers e and f of Taberlet et al. (1991)
for trnL-trnF. The rps4 gene was amplified with primers rps5 and trnS described by Souza-Chies et al. (1997)
. The amplified material was purified with shrimp alkaline phosphatase and exonuclease I (Amersham Biosciences, Piscataway, New Jersey, USA), and the two strands were directly sequenced, using the internal primers ITS3 and ITS3-reverse (Desfeux and Lejeune, 1996
) for ITS. Sequencing was done using the BigDye Terminator Sequencing Kit (PE Applied Biosystems, Foster City, Michigan, USA) on an ABI Prism 310 (PE Applied Biosystems). Sequences were deposited with GENBANK (see Appendix 1 in the Supplementary Data accompanying the online version of this article).
Data analysis
Sequences were aligned using the ClustalX 1.81 program (Thompson et al., 1994
, 2001
) and manually refined. Phylogenetic analyses were performed in PAUP*, version 4.0b10 (Swofford, 1998
). All the analyses described later were done with each spacer considered as a different data set (ITS or trnL-trnF) and also with both markers combined in a single alignment. ITS1 and ITS2 were considered together. Because of technical problems, we could not obtain usable sequences for some species for some markers, so that only 55 species were studied for ITS and 60 for trnL-trnF. For rps4, we sequenced material from 32 species, in order to obtain representatives of the major clades of Passiflora, which were compared with three of the four outgroups.
Equally weighted parsimony (maximum parsimony [MP]) analyses were performed by a heuristic search with tree bisection-reconnection (TBR) branch swapping, the MULPARS option, and 1000 random-addition replicates. Bootstrap statistical support (Felsenstein, 1985
) was carried out with 1000 replications of heuristic search and simple taxon addition, with the all trees saved option. Topological congruence between the trees obtained by the different data sets was evaluated with Templeton's test (1983)
, and heterogeneity among markers was assessed using the partition homogeneity test or incongruence length difference test (IDL) developed by Farris et al. (1994)
with 1000 iterations. The g1 statistic (Hillis, 1991
) of skewed tree-length distribution was calculated from 10 000 random trees, to measure the phylogenetic information content of the two spacer regions independently and for the combined data.
The appropriate model of nucleotide substitution for maximum-likelihood (ML) analysis was determined using the MODELTEST 3.06 program (Posada and Crandall, 1998
). This procedure implements a hierarchical likelihood-ratio test to determine the model that best fits the data. For ML tree estimation, heuristic searches with as-is, TBR branch swapping, and MULPARS option, were conducted with PAUP*. Because obtaining support estimates for the ML trees' branches by bootstrap analysis is computationally too intensive with a data set of this size, we conducted a bootstrap analysis using the neighbor-joining method under a maximum-likelihood model, with parameters settings estimated by the MODELTEST, as described by Xiang et al. (2002)
. Topology comparisons between the trees obtained with the two spacers were done with the KH (Kishino and Hasegawa, 1989
) and SH (Shimodaira and Hasegawa, 1999
) tests.
The new MetaPIGA 1.02 software was used for phylogenetic inference under the ML criterion using a metapopulation genetic algorithm (Lemmon and Milinkovitch, 2002
). This method was developed to supply better and faster estimates of ML trees with large data sets. The program was run using the following relevant settings: HKY85 (Hasegawa et al., 1985
) nucleotide substitution model, four populations, and 10 replications. Lemmon and Milinkovitch (2002)
suggest that four population metaPIGA runs seem to give the best compromise between computing speed and accuracy of topology inference. All other settings were used as suggested in the program manual (Lemmon and Milinkovitch, 2002
). The 50% majority-rule consensus tree was constructed in PAUP*.
Bayesian analyses of the data were performed using MrBayes 2.01 (Huelsenbeck and Ronquist, 2001
) to generate a posterior probability distribution using Markov chain Monte Carlo (MCMC) methods. No a priori assumptions about the topology of the tree were made and all searches were provided with a uniform prior. The models of DNA substitution used were those estimated by MODELTEST and used in the standard ML analyses. The MCMC processes were set so that four chains were run simultaneously for 500 000 generations, with trees being sampled every 100 generations for a total of 5000 trees. Burn-in, or the time for each parameter to reach a stationary state, was determined when visual inspection indicated that the log-likelihood values achieved an asymptote over a large number of generations. To calculate the posterior probability of each bipartition, a 50% majority-rule consensus tree was constructed from the remaining trees using PAUP*.
For the distance analyses, trees were constructed using the neighbor-joining method (NJ; Saitou and Nei, 1987
) using proportional (p), Kimura two parameter, and logDet (Steel, 1994
; Lockhart et al., 1994
) distances. LogDet or paralinear distances were calculated to test the possible influence of the nucleotide composition difference in the phylogeny (Nei and Kumar, 2000
). Reliability of the trees was tested using 1000 bootstrap replications (Hedges, 1992
). Topology comparisons between the ITS and trnL-trnF data sets were performed using the KH and SH tests. Relative rate evaluations, based on the two-cluster test of Takezaki et al. (1995)
, were performed using the PHYLTEST program (Kumar, 1996
). Average nucleotide diversities and their standard errors within each subgenus were calculated by the SendBS program (Takezaki, 2001
) with the p distance option and 2000 bootstrap replications.
The results of the different inference phylogenetic methods within each data set were quite similar, so only selected results are presented here. (All other trees are presented in the Supplementary Data accompanying the online version of this article.) All trees were rooted with Paropsia because it is representative of Paropsieae and was shown to be more basal than Adenia by Chase et al. (2002)
.
A possible substitution saturation in the ITS region was investigated by plotting the ITS pairwise sequence p distances against the more conserved trnL-trnF pairwise distances. The minimum sizes of the flowers (using the descriptions of Killip [1938
], MacDougal [1994
], and Cervi [1997
]) of different clades (see Results) were compared using the Wilcoxon-Mann-Whitney test to determine if their averages were significantly different.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
2 = 34.8; 1 df; P < 0.0001). The pairwise p distances, including the outgroups, ranged from 1.9% to 38%, with an average value of 20%. The trnL-trnF intergenic region had a total aligned length of 416 bp (individual sequences ranged from 239 to 313 bp), 61 being parsimony-informative. The pairwise p distances, including the outgroups, ranged from 1.8% to 10%, with an average value of 4.4%. The GC content is low in this region (36%), and no marked differences were found among the subgenera. The alignment of the rps4 gene is 576 bp in length (unaligned sequences ranged from 542 to 560 bp), 58 being parsimony-informative. The GC content is very low in this region (27%), and again, no marked differences were found among the subgenera. The pairwise p distances, including the outgroups, ranged from 0.2% to 8.1%, with an average value of 3.5%.
Phylogenetic analysis
The selected model for the maximum-likelihood analysis using MODELTEST was TrN+I+G (–lnL = 8227.1035) for ITS, HKY + G (–lnL = 2131.0359) for trnL-trnF, and TVM + G (–lnL = 2164.3538) for rps4 (see Posada and Crandall [1998
] for details of the models). The gamma shape parameters estimated for the three DNA regions were also different (ITS, 1.839; trnL-trnF, 0.5640; rps4, 0.7806). For the combined data of the spacer regions, the model selected was TrNef+I+G (–lnL = 10 381.6836) with a gamma shape parameter of 1.078. Figures 1–3 present the phylogenetic trees obtained using the standard ML method on the ITS, trnL-trnF, and combined data sets, respectively. In the great majority of the trees we can observe three major clades within Passiflora, which are generally supported by high bootstrap values: one composed of all studied species of subgenera Distephana, Dysosmia, Dysosmioides, Passiflora, and Tacsonioides; another comprising all species of sugenera Adopogyne, Decaloba, Murucuja, and Pseudomurucuja; and a third composed of the four species of subgenus Astrophea. We call these the Passiflora, Decaloba, and Astrophea clades, respectively.
|
|
comparing the Bayesian posterior probabilities support values with the nonparametric bootstrap values and with the results of Lemmon and Milinkovitch (2002)
on the metaGA approach. They concluded that the higher support values for the Bayesian and metaGA analyses are appropriate and provide much closer estimates of phylogenetic accuracy than the estimates given by corresponding bootstrap proportions (but see Suzuki et al. [2002
] for a rebuttal).
|
). For the combined trees, the g1 statistic is –1.058. The maximum-parsimony trees estimated using these parameters were very similar to the respective ML trees (see Supplementary Data accompanying the online version of this article).
|
There are some differences among the trees on the positioning of a few Passiflora species within the major clades that are worth noting. Passiflora lancetillensis and P. microstipula, from subgenus Deidamioides, form a sister group to the Decaloba clade in the trnL-trnF tree (Fig. 2) with a high divergence and therefore could possibly be considered a fourth main clade. But in the ITS tree (Fig. 1), P. lancetillensis is placed within the Decaloba clade. Passiflora rubra is also placed as a highly divergent sister to the remaining species of the Decaloba clade in the trnL-trnF trees (Fig. 2) but not in the ITS trees. Also noteworthy is the position of P. foetida in the ITS tree (Fig. 1), as sister to all other taxa in the Passiflora clade, a placement not seen in the trnL-trnF trees (Fig. 2).
By contrast with the high support among markers and phylogenetic methods for the major Passiflora clades, the relationships among the outgroups and clades received very low support (bootstrap and posterior probabilities values) in each tree and with many disagreements among them. For example, in the ITS maximum-likelihood tree (Fig. 1), the outgroups are not phylogenetically adjacent, because Tetrastilys ovalis and Adenia keramanthus are positioned as sister to the Passiflora clade (with a support value of 97%), while Mitostemma brevifilis is placed as a sister to the Decaloba + Astrophea clades (with support of <50%). In the trnL-trnF tree (Fig. 2), T. ovalis was included within the Passiflora clade as sister to P. speciosa, but here the other three outgroup species did group together, although with medium (73%) bootstrap support. The combined ML tree (Fig. 3) presents a topology that has components found in both single marker trees, with T. ovalis positioned as sister of the Passiflora clade (bootstrap support = 100%), Adenia keramanthus positioned among the three clades, and Mitostemma brevifilis appearing as the most basal species, but both with support values below 50%.
All trees of the rps4 data set, despite its much smaller size, corroborate the existence of three major Passiflora clades, but there is disagreement on the relationship among the clades and their relationship with the outgroups when comparing the different methods. For example, in the ML tree the higher-order relationships are statistically unresolved (Fig. 5), but in the NJ and Bayesian analyses (see Supplementary Data) there is moderate support for the non-monophyly of Passiflora. Therefore, the monophyly of the genus Passiflora was not statistically supported by any data set or phylogenetic method, and in several trees there was even statistical support for a non-monophyletic Passiflora.
|
|
and Reeves et al. (2001)
in other organisms. At present, it is not clear whether heterogeneous sets of data should or should not be combined to obtain an overall phylogenetic picture. On the other hand, for ML and NJ trees, the KH and SH tests did not show significant differences between the spacers.
One interesting difference that can be observed visually is the higher divergence among the species of the Decaloba clade (implied by the usually longer branch lengths) as compared to those of the Passiflora clade, especially marked for the trnL-trnF region (Fig. 2). Also noteworthy is the usually longer branch leading to the Decaloba clade, as contrasted to that leading to the Passiflora clade. Nucleotide diversity (
) for the ITS region is significantly higher in the Decaloba clade (0.102 ± 0.013) as compared to the Passiflora clade (0.043 ± 0.078). The relative rate Z statistic (Z = 3.549) calculated with the PHYLTEST package rejects rate constancy (P < 0.05) between the two. A similar result was found for the trnL-trnF intergenic region (Z = 2.399, P < 0.05) and for the rps4 gene (Z = 2.976, P < 0.05), suggesting that these different evolutionary rates are not locus specific but characteristic of the clades. To test if these differences were not due to the different number of species sampled in them, new analyses were performed consisting of 10 random subsamples of species with an identical number of species in both clades. The results corroborated the significant differences between the two clades. Because of its much smaller sample size, we did not test the Astrophea clade for these divergence differences.
Anyone acquainted with Passiflora could intuitively perceive that the different subgenera could also have dissimilar flower sizes. Indeed, a Wilcoxon-Mann-Whitney test showed that the flowers of the Passiflora clade have a significantly larger (P < 0.001) minimum size than the flowers of the Decaloba and Astrophea clades (median values: 19 mm vs. 6.5 mm, respectively).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
By contrast, all phylogenetic results presented here suggest the existence of three well-defined clades in Passiflora. The largest comprise all representatives of Killip's (1938)
subgenera Distephana, Dysosmia, Dysosmioides, Passiflora, and Tacsonioides. Because the subgenus Passiflora is by far the most diverse of all, we named this set as the Passiflora clade. A second major clade comprises all species of Killip's (1938)
and MacDougal's (1994)
subgenus Decaloba, as well P. tulae of subgenus Murucuja, P. cuprea of subgenus Pseudomurucuja, and P. multiflora of subgenus Adopogyne. Using the same reasoning stated above, we called this the Decaloba clade. The third clade proposed here comprises the four species of the subgenus Astrophea (P. haematostigma, P. mansoi, P. citrifolia, and P. macrophylla), which has very high support values in most of the trees presented here.
The subgenus Deidamioides (represented here by P. lancetillensis and P. microstipula) presents ambiguous positions in the trees derived from the different markers. In the more conserved trnL-trnF spacer and rps4 gene, this subgenus is very divergent from the other subgenera, and it was placed in some trnL-trnF trees (e.g., the MP tree available in the supplementary data) as a very basal independent clade, suggesting that it could be considered as a fourth major clade of Passiflora. However, in the ITS trees (e.g., Fig. 1), P. lancetillensis is placed within the Decaloba clade, with a high bootstrap value. These ambiguities, and the low number of species sampled, determine that the status of this subgenus should remain undefined until further data are analyzed.
Though we could not find any published evidence that explicitly corroborates the existence of these three clades in Passiflora, two papers seem to agree with our results. De Melo et al. (2001)
described the basic chromosome number of several subgenera of Passiflora and some related genera. From the interpretation of their tree (their Fig. 6) in a phylogenetic context, a grouping of subgenera was suggested that is consistent with the three clades described here. On the other hand, Downie et al. (1996)
studied the rpoC1 chloroplast intron in several families and found that it was absent in four species of the subgenus Decaloba, but present in six species of the Passiflora clade (three subgenera). It should also be remembered (see Results) that the five subgenera that form the Passiflora clade all have statistically significant larger and more conspicuous flowers than those of the Decaloba and Astrophea clades.
Although there are some consistent species groupings within the Passiflora and Decaloba clades, only a few of them have high levels of statistical support and are consistent among markers and phylogenetic methods, such as the P. actinia + P. elegans + P. sidaefolia and the P. quadrangularis + P. alata subgroups. Most of these subgroups, however, are not consistent with Killip's (1938)
subgenera or sections.
One question that can be considered is the taxon sampling of our study and its bearing on the reliability of the major clades found. We are aware that a larger sample of species could result in more reliable trees. However, we suggest that our results and phylogenetic conclusions are presently adequate. Thus, our sampling represents about 15% and 49% of Passiflora's species and subgenera, respectively. These sample sizes are by no means unusual or small in phylogenetic studies of large and diversified taxa. For example, Miller et al. (2002)
investigated the monophyly of Ipomoea (Convolvulaceae) using just 35 (5.8%) of about 600 species of this genus; Bortiri et al. (2002)
studied just 22 (11%) of over 200 species of Prunus; and Cuénoud et al. (2002)
studied just one species of each genus for the phylogenetic analyses of Caryophyllales. Perhaps the clearest example of the utility of a limited taxonomic sampling is the most recent work that tries to infer ordinal and familial classification for angiosperms (such as that of the Angiosperm Phylogeny Group, 1998
) with a relatively limited sample of the angiosperm species. Rosenberg and Kumar (2001)
also suggested that incomplete sampling might not be a problem for phylogenetic inferences (but see Zwickl and Hillis [2002
] for a different position). Our results with 11 (49%) of the 23 putative subgenera suggest that at least the subgenera studied here should be grouped in three major clades: Passiflora, Decaloba, and Astrophea.
Sequence variation
The GC content in the ITS region varied from 47% to 67% in the different species of Passiflora, a variability that has also been observed in other groups of plants (e.g., Nishikawa et al., 1999
; Gaut et al., 2000
). Interestingly, the GC content differs significantly between the Decaloba and non-Decaloba clades, and although this may be a bias factor for most phylogenetic methods, the use of the logDet distance (Lockhart et al., 1994
; Steel, 1994
) should correct for it. The use of this distance with a subsample of our data yielded the same main clades presented here, suggesting that the ITS phylogeny was not biased by the heterogeneity in nucleotide content (see Supplementary Data).
Heterogeneity in the rates of DNA change such as that observed in the Decaloba/non-Decaloba clades has also been documented at several taxonomic levels in plants (Bousquet et al., 1992
; Gaut et al., 1993
; Suh et al., 1993
; Gielly and Taberlet, 1996
; Eyre-Walker and Gaut, 1997
; Aïnouche and Bayer, 1999
; Bortiri et al., 2002
). Many factors supposedly influence the rate of molecular evolution, such as generation time, selection, DNA replication or repair, and metabolic rates (Wu and Li, 1985
; Britten, 1986
; Bousquet et al., 1992
; Li, 1993
; Martin and Palumbi, 1993
; Eyre-Walker and Gaut, 1997
; Gaut et al., 1997
). Unfortunately, little is known about such factors in Passiflora, although Benson at al. (1975)
asserted that members of the subgenus Decaloba have shorter generation times than species of the subgenus Passiflora, a factor that may have accelerated its evolutionary rate. Another possible explanation for the very short branches within the Passiflora clade is a rapid radiation of the species of this clade, an explanatory factor that was attributed to similar results described in other groups of angiosperms (e.g., Bortiri et al., 2002
).
Comparison among markers
The discrepancies between the trees obtained with ITS and trnL-trnF are mainly due to the position of a few species (e.g., P. rubra and P. foetida) and the outgroup relationships. Some of these discrepancies could be due to the different forms of inheritance of nuclear and chloroplast markers. The majority of the angiosperms have maternal chloroplast inheritance, but Turnera ulmifolia, a closely related member of the Malpighiales, has mixed, paternally biased inheritance (Shore and Triassi, 1998
). (Note that Chase et al. [2002
] placed the Turneraceae within the Passifloraceae.) Courriveau and Coleman (1988)
found chloroplast paternal inheritance in P. edulis; and similar results were obtained by A. P. Lorenz et al. (Federal University of Rio Grande do Sul, unpublished data) in P. elegans and P. actinia. In accordance with this reasoning, the trees based on the two chloroplast markers, trnL-trnF and rps4, generally had fewer discrepancies among themselves than when they were compared with trees based on ITS sequences.
Another factor that may have contributed to these discrepancies, mainly at the higher level relationships, is the observed saturation in the ITS region, when contrasted to the trnL-trnF data (Fig. 5). The phylogenetic utility of the ITS spacers within the Passifloraceae seems to be restricted to the interspecific comparisons, and they should be used with caution at the intergeneric level.
Future considerations
We recommend, and it is in our plans, that future work should concentrate on certain critical areas, such as increased taxon sampling, mainly of the subgenera not yet studied and of the larger subgenera, such as Decaloba, that were underrepresented here. Furthermore, considering the lack of adequate resolution found in our trees at the higher order relationships, it is necessary to increase the amount of sequence data to clarify issues such as the monophyly of Passiflora and its relationship with its major clades and with the other genera of the Passifloraceae. Also of interest is an understanding of the causes underlying the different evolutionary dynamics found here among the major Passiflora clades.
|
FOOTNOTES |
|---|
6 Author for correspondence (e-mail: loreta.freitas{at}ufrgs.br
)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Angiosperm Phylogeny Group. 1998 An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85: 531-553[CrossRef][ISI]
Bakker F. T. D. Hellbrügge A. Culham M. Gibby 1998 Phylogenetic relationships within Pelargonium sect. Peristera (Geraniaceae) inferred from nrDNA and cpDNA sequence comparisons. Plant Systematics and Evolution 211: 273-287[CrossRef][ISI]
Baldwin B. G. 1992 Phylogenetic utility of the internal transcribed spacers of nuclear ribosomal DNA in plants: an example from the Compositae. Molecular Phylogenetics and Evolution 1: 3-16[CrossRef][Medline]
Baldwin B. G. M. J. Sanderson M. J. Porter M. F. Wojciechowski C. S. Campbell M. J. Donoghue 1995 The ITS region of nuclear ribosomal DNA: a valuable source of evidence on angiosperm phylogeny. Annals of the Missouri Botanical Garden 82: 247-277[CrossRef][ISI]
Benson W. W. K. S. Brown L. E. Gilbert 1975 Coevolution of plants and herbivores: passion flower butterflies. Evolution 29: 659-680[CrossRef][ISI]
Bortiri E. S.-H. Oh F.-Y. Gao D. Potter 2002 The phylogenetic utility of nucleotide sequences of sorbitol 6-phosphate dehydrogenase in Prunus (Rosaceae). American Journal of Botany 89: 1697-1708
Bousquet J. S. A. H. Strauss A. H. Doerksen R. A. Price 1992 Extensive variation in evolutionary rate of rbcL gene sequences among seed plants. Proceedings of National Academy of Sciences USA 89: 7844-7848
Britten R. J. 1986 Rates of DNA sequence evolution differ between taxonomic groups. Science 231: 1393-1398
Cervi A. C. 1997 Passifloraceae do Brasil. Estudo do gênero Passiflora L., subgênero Passiflora. Fontqueria 45: 1-92
Chase M. W. S. Zmarzty M. D. Lledó K. J. Wurdack S. M. Swensen M. F. Fay 2002 When in doubt, put it in Flacourtiaceae: a molecular phylogenetic analysis based on plastid rbcL DNA sequences. Kew Bulletin 57: 141-181[CrossRef]
Courriveau J. L. A. W. Coleman 1988 Rapid screening method to detect potential biparental inheritance of plastid DNA and results for over 200 angiosperm species. American Journal of Botany 75: 1443-1458[CrossRef][ISI]
Cuénoud P. V. Savolainen L. W. M. Powell R. J. Grayer M. W. Chase 2002 Molecular phylogenetics of Caryophyllales based on nuclear 18S rDNA and plastid rbcL, atpB, and matK DNA sequences. American Journal of Botany 89: 132-144
de Melo N. F. A. C. Cervi M. Guerra 2001 Karyology and cytotaxonomy of the genus Passiflora L. (Passifloraceae). Plant Systematics and Evolution 226: 69-84[CrossRef][ISI]
Desfeux C. B. Lejeune 1996 Systematics of Euromediterranean Silene (Caryophyllaceae): evidence from a phylogenetic analysis using ITS sequence. Comptes Rendus, Academie des Sciences, Paris 319: 351-358
Downie S. R. E. Llanas D. S. Katz-Downie 1996 Multiple independent losses of the rpoC1 intron in angiosperm chloroplast DNA's. Systematic Botany 21: 135-151
Doyle J. J. J. L. Doyle 1987 A rapid DNA leaf material. Phytochemical Bulletin 19: 11-15
Escobar L. K. 1988 Passifloraceae. In P. Pinto and G. Lozano [eds.], Flora de Colombia, vol. 10, 1–138. Universidad Nacional de Colombia, Bogotá, Colombia
Escobar L. K. 1989 A new subgenus and five new species in Passiflora (Passifloraceae) from South America. Annals of the Missouri Botanical Garden 76: 877-885.[CrossRef][ISI]
Eyre-Walker A. B. S. Gaut 1997 Correlated rates of synonymous site evolution across plant genomes. Molecular Biology and Evolution 14: 455-460[Abstract]
Farris J. S. M. Källersjö A. G. Kluge C. Bult 1994 Testing significance of congruence. Cladistics 10: 315-319[CrossRef][ISI]
Felsenstein J. 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791[CrossRef][ISI]
Gaut B. S. L. G. Clark J. F. Wendel S. V. Muse 1997 Comparisons of the molecular evolutionary process at rbcL and ndhF in the grass family (Poaceae). Molecular Biology and Evolution 14: 769-777[Abstract]
Gaut B. S. S. V. Muse M. T. Clegg 1993 Relative rate of nucleotide substitution in the chloroplast genome. Molecular Phylogenetics and Evolution 2: 89-96[CrossRef][Medline]
Gaut B. S. L. P. Tredway C. Kubik R. L. Gaut W. Meyer 2000 Phylogenetic relationships and genetic diversity among members of the Festuca-Lolium complex (Poaceae) based on ITS sequence data. Plant Systematics and Evolution 224: 33-53[CrossRef][ISI]
Gielly L. P. Taberlet 1996 A phylogeny of the European gentians inferred from chloroplast trnL (UAA) intron sequences. Botanical Journal of Linnean Society 120: 57-75[CrossRef]
Hardig T. M. P. S. Soltis D. E. Soltis 2000 Diversification of the North American shrub genus Ceanothus (Rhamnaceae): conflicting phylogenies from nuclear risbosomal DNA and chloroplast DNA. American Journal of Botany 87: 108-123
Hasegawa M. H. Kishino T. Yano 1985 Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 22: 160-174[CrossRef][ISI][Medline]
Hedges S. B. 1992 The number of replications needed for accurate estimation of the bootstrap P value in phylogenetic studies. Molecular Biology and Evolution 9: 366-369[ISI][Medline]
Hillis D. M. 1991 Discriminating between phylogenetic signal and random noise in DNA sequences. In M. M. Myamoto and J. Craft [eds.], Phylogenetic analysis of DNA sequences, 278–294. Oxford University Press, New York, New York, USA
Huelsenbeck J. P. F. Ronquist 2001 MrBayes: Bayesian inference of phylogeny. Department of Biology, University of Rochester, Rochester, New York, USA
Judd W. S. C. S. Campbell E. A. Kellogg P. F. Stevens 1999 Plant systematics: a phylogenetic approach. Sinauer Associates, Sunderland, Massachusetts, USA
Killip E. P. 1938 The American species of Passifloraceae. Field Museum of Natural History, Botanical Series 19: 1-613
Kishino H. M. Hasegawa 1989 Evaluation of the maximum likelihood estimate of the evolutionary tree topology from DNA sequence data, and the branching order in Hominoidea. Journal of Molecular Evolution 29: 170-179[CrossRef][ISI][Medline]
Kumar S. 1996 PHYLTEST: phylogeny hypothesis testing software. Pennsylvania State University, University Park, Pennsylvania, USA
Lemmon A. M. C. Milinkovitch 2002 The metapopulation genetic algorithm: an efficient solution for problem of large phylogeny estimation. Proceedings of National Academy of Sciences, USA 99: 10516-10521
Li W. H. 1993 So, what about the molecular clock hypothesis?. Current Opinion in Genetics and Development 3: 896-901[CrossRef][Medline]
Lockhart P. J. M. A. Steel M. D. Hendy D. Penny 1994 Recovering evolutionary trees under a more realistic model of sequence evolution. Molecular Biology and Evolution 11: 605-612[ISI]
MacDougal J. M. 1994 Revision of Passiflora subgenus Decaloba section Pseudodysosmia (Passifloraceae). Systematic Botany Monographs 41: 1-146
Martin A. P. S. R. Palumbi 1993 Body size, metabolic rate, generation time, and the molecular clock. Proceedings of National Academy of Sciences, USA 90: 4087-4091
Miller R. E. T. R. Buckley P. S. Manos 2002 An examination of monophyly of morning glory taxa using Bayesian phylogenetic inference. Systematic Biology 51: 740-753[CrossRef][ISI][Medline]
Molvray M. P. J. Kores M. W. Chase 1999 Phylogenetic relationships within Korthalsella (Viscaceae) based on nuclear ITS and plastid trnL-F sequence data. American Journal of Botany 86: 249-260
Nadot S. R. Bajon B. Lejeune 1994 The chloroplast gene rps4 as a tool for the study of Poaceae phylogeny. Plant Systematics and Evolution 191: 27-38[CrossRef][ISI]
Nei M. S. Kumar 2000 Molecular evolution and phylogenetics. Oxford University Press, Oxford, UK
Nishikawa T. K. Okazaki T. Uchino K. Arakawa T. Nagamine 1999 A molecular phylogeny of Lilium in the internal transcribed spacer region of nuclear ribosomal DNA. Journal of Molecular Evolution 49: 238-249[CrossRef][ISI][Medline]
Posada D. K. A. Crandall 1998 MODELTEST: testing the model of DNA substitution. Bioinformatics 14: 817-818
Rauscher J. T. 2002 Molecular phylogenetics of the Espletia complex (Asteraceae): evidence from nrDNA ITS sequences on the closest relatives of Andean adaptive radiation. American Journal of Botany 98: 1074-1084
Reeves G. M. W. Chase P. Goldblatt P. Rudall M. F. Fay A. V. Cox B. Lejeune T. Souza-Chies 2001 Molecular systematics of Iridaceae: evidence from four plastid regions. American Journal of Botany 88: 2074-2087
Rosenberg M. S. S. Kumar 2001 Incomplete taxon sampling is not a problem for phylogenetic inference. Proceedings of National Academy of Sciences, USA 98: 10751-10756
Saitou N. M. Nei 1987 The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 9: 945-967
Sánchez I. F. Angel M. Grumm M. C. Duque M. Lobo J. Tohme W. Roca 1999 Variability of chloroplast DNA in the genus Passiflora L. Euphytica 106: 15-26[CrossRef][ISI]
Shimodaira H. M. Hasegawa 1999 Multiple comparisons of log-likelihoods with applications to phylogenetic inference. Molecular Biology and Evolution 16: 1114-1116[ISI]
Shore J. S. M. Triassi 1998 Paternally biased cpDNA inheritance in Turnera ulnifolia (Turneraceae). American Journal of Botany 85: 328-332[Abstract]
Souza-Chies T. T. G. Bittar S. Nadot L. Carter E. Besin B. Lejeune 1997 Phylogenetic analysis of Iridaceae with parsimony and distance methods using the plastid gene rps4. Plant Systematics and Evolution 204: 109-123[CrossRef][ISI]
Steel M. 1994 Recovering a tree from the Markov leaf colouration it generates under a Markov model. Applied Mathematics Letters 7: 19-23
Suh Y. L. B. Thien H. E. Reeve E. A. Zimmer 1993 Molecular evolution and phylogenetic implications of internal transcribed spacer sequences of ribosomal DNA in Winteraceae. American Journal of Botany 80: 1042-1055[CrossRef][ISI]
Suzuzi Y. G. V. Glazco M. Nei 2002 Overcredibility of molecular phylogenies obtained by Bayesian phylogenetics. Proceedings of National Academy Sciences, USA 99: 16138-16143[CrossRef]
Swofford D. L. 1998 PAUP*: phylogenetic analysis using parsimony (*and other methods). Version 4. Sinauer Associates, 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]
Takezaki N. 2001 SendBS Program. Available at website: www.cib.nig.ac.jp/dda/ntakezak.html#sendbs
Takezaki N. A. Rzhetsky M. Nei 1995 Phylogenetic test of the molecular clock and linearized tree. Molecular Phylogenetics and Evolution 12: 823-833
Templeton A. R. 1983 Phylogenetic inference from restriction endonuclease cleavage site maps with particular reference to the evolution of humans and the apes. Evolution 24: 218-227
Thompson J. D. D. G. Higgins T. J. Gibson 1994 ClustalW: improving the sensivity of progressive multiple sequence weighting, positions-specific gap penalties and weight matrix choice. Nucleic Acids Research 22: 4673-4680
Thompson J. D. D. G. Higgins T. J. Gibson 2001 ClustalX. Program available at ftp://ftp-igbmc.u-strasbg.fr/pub/clustalx/
Wilcox T. P. D. J. Zwickl T. A. Heath D. M. Hillis 2002 Phylogenetic relationships of dwarf boas and comparison of Bayesian and bootstrap measures of phylogenetic support. Molecular Phylogenetics and Evolution 25: 361-371[CrossRef][ISI][Medline]
Wu C. I. W. H. Li 1985 Evidence for higher rates of nucleotide substitution in rodents than in man. Proceedings of National Academy of Sciences, USA 82: 1741-1745
Xiang Q.-Y. M. L. Moody D. E. Soltis C. Fan P. S. Soltis 2002 Relationships within Cornales and circumscription of Cornaceae—matK and rbcL sequence data and effects of outgroups and long branches. Molecular Phylogenetics and Evolution 24: 35-57[CrossRef][ISI][Medline]
Yoder A. D. J. A. Irwin B. A. Payseur 2001 Failure of the ILD test to determine data combinability for slow loris phylogeny. Systematic Biology 50: 408-424[ISI][Medline]
Zwickl D. J. D. M. Hillis 2002 Increased taxon sampling greatly reduces phylogenetic error. Systematic Biology 51: 588-598[CrossRef][ISI][Medline]
This article has been cited by other articles:
|
|
 |
|
  A. K. Hansen, L. K. Escobar, L. E. Gilbert, and R. K. Jansen Paternal, maternal, and biparental inheritance of the chloroplast genome in Passiflora (Passifloraceae): implications for phylogenetic studies Am. J. Botany, January 1, 2007; 94(1): 42 - 46. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P. LORENZ-LEMKE, V. C. MUSCHNER, S. L. BONATTO, A. C. CERVI, F. M. SALZANO, and L. B. FREITAS Phylogeographic Inferences Concerning Evolution of Brazilian Passiflora actinia and P. elegans (Passifloraceae) Based on ITS (nrDNA) Variation Ann. Bot., April 1, 2005; 95(5): 799 - 806. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||
