Boesenbergia longiflora (Zingiberaceae) and descriptions of five related new taxa

Yüklə 321,14 Kb.
Pdf görüntüsü
ölçüsü321,14 Kb.
1   2   3   4   5   6   7   8   9   ...   14

Fig. 4. One of 37 best, Maximum Likelihood phylograms (-ln=2218.749) for Boesenbergia 

longiflora and closely related taxa based on analysis of nuclear ribosomal ITS data for all 

samples and clones. Support values (Bootstrap/Posterior Probability) for critical branches are 

shown above branches. Bold typeface highlights strongly supported positions of the different 

Boesenbergia longiflora copies generated by cloning. Asterisks indicate samples that should 

also be cloned to clarify a few polymorphisms. Underlined samples are placed in conflicting 

positions on the chloroplast phylogeny. Note: lower branches were pruned to fit page and are 

not to scale.


Gard. Bull. Singapore 65(1) 2013

Fig. 5. Single best Maximum Likelihood phylogram (-ln=3953.067) for 

B.  longiflora and 

closely related taxa based on analysis of chloroplast trnK intron data. Support values (Bootstrap/

Posterior Probability) for critical branches are shown above branches. Numbers in bold typeface 

indicate strong support. Underlined samples are placed in conflicting positions on the nuclear 

ribosomal phylogeny. Note: lower branches were pruned to fit page and are not to scale.


Boesenbergia longiflora and related taxa

The identification of at least two disparate ITS copies for B. longiflora requires 

further discussion. Only one sample of 

B. longiflora was available for inclusion in 

this study. This sample, M11P48

 was from cultivated plants grown by the first author, 

but originally collected in Burma (Kress 03-7305, US). Multiple clones representing 


 B. longiflora ITS type were distributed among members of the B. kingii clade 

(ITS copy “A”) while the other 

B. longiflora copy fell sister to the B. kerrii clade 

(ITS copy “B”) as is shown in Fig. 4. Although a number of tests have been proposed 

to distinguish between the processes of hybridisation and incomplete lineage sorting 

(Joly et al. 2009) all have limitations. We were restricted here by the single sample for 

B. longiflora, but hopefully, more samples will be available for future analyses.

The  majority  of  the  polymorphic  samples  were  identified  as  B. kingii based 

on morphology. Six of the twelve samples required cloning to generate clean ITS 

sequences. These samples fall into two categories. The first group included samples 

with multiple ITS sequence types, but those sequences were all quite similar with only 

a few different bases or slight length variations. Analyses of the ITS data partition 

placed all of these sequences in the B. kingii sensu lato clade as described above 

(B. hamiltonii + B. kingii + B. maxwellii

), and with moderate to high support (75% 

BS, 0.98 PP) but with poor resolution within the clade (Fig. 4). Further, most of the 

B. kingii

 samples clustered with moderate support (70% BS, 1.00 PP). Among the 

ingroup members, B. kingii is the most broadly distributed species and is the most 

genetically diverse taxon. 

Chloroplast trnK intron (including the protein coding 

matK gene)

A total of 36 potentially parsimony-informative characters (15 for ingroup only) were 

used to generate almost 100,000 maximum parsimony trees. However, maximum 

likelihood analyses resulted in a single best tree (Fig. 5). The likelihood tree had a 

topology identical to several of the MP trees. As with the results of the ITS analyses, 

both  MP  and  ML  analyses  identified  two  clades  within  the  ingroup,  one  including 

samples of B. collinsii + B. kerrii + 

B. longiflora + B. maxwellii (65% BS; 0.85 PP) 

and a clade comprising Bhamiltonii and B. kingii

 (63% BS; 0.83 PP). Three taxa were 

resolved and supported as monophyletic: B. collinsii

 (65% BS, 0.94 PP) B. kerrii (63% 

BS, 0.94 PP) and B. maxwellii

 (99% BS, 1.00 PP). The relationship of B. maxwellii 

to B. hamiltonii and B. kingii differed here relative to the ITS results, but only with 

weak (65% BS; 0.85 PP) support. Sequences for two samples were not available for 

this data set, B. hamiltonii M3212 and B. maxwellii MP1450. The two samples with 

names underlined in Fig. 4, B. kingii M11P77 and M11P78, were identical and clearly 

supported as members of the B. kingii group in the phylogeny produced by analyses of 

the nuclear ITS dataset. Results of the chloroplast data analyses (Fig. 5) clearly (88% 

BS, 1.00 PP) placed these two samples in a clade with B. hamiltonii

Combined analyses

The combined data set required a number of samples to be excluded. Because the ITS 

data required cloning for some samples, results of the ITS MP analyses were used to 

select the clone with the shortest branch for each sample (as suggested by Beilstein et 


Gard. Bull. Singapore 65(1) 2013

al. 2008) for subsequent use in the combined analyses. This was done only when all 

clones for a particular sample either formed a monophyletic lineage or clustered within 

a moderately or well-supported lineage (BS≥75%). In the instance of B. kingii samples 

M11P77 and M11P78

, there was positional conflict between the nuclear and plastid 

datasets. Those samples were excluded from the combined analysis. As mentioned, 

there was only a single sample of 

B. longiflora available for this study. The nuclear 

data set included two distinctly different copies, A and B. Rather than totally exclude 

this critical taxon, we selected the nuclear copy B that best matched the chloroplast 

phylogeny, while recognising we do not know which better represents the evolutionary 

history of the species.

Maximum parsimony analysis produced 468 shortest trees. Maximum likelihood 

analyses produced three equally likely trees, one of which is shown in Fig. 6. Both 

MP and ML analyses identified two clades within the in-group most similar to those 

of the ITS data analyses, a B. kingii clade including samples of B. hamiltonii + B. 

kingii + B. maxwellii and a 

B.  longiflora sensu lato clade including samples of B. 

collinsii + B. kerrii 

B. longiflora. The B. kingii sensu lato clade received 62% BS and 

0.78 PP. Unlike the ITS data analyses results, the combined data analyses produced 

highly resolved topologies within this clade. Significantly, B. maxwellii is resolved as 

monophyletic (83% BS; 1.00 PP). Boesenbergia hamiltonii and Bkingii samples are 

grouped together with 51% BS and 0.95 PP. Boesenbergia hamiltonii was also resolved 

as monophyletic, but with weak support (<50% BS, 0.86 PP). The B. longiflora sensu 

lato clade was strongly supported (95% BS; 1.00 PP) with significant internal structure 

and support. Boesenbergia collinsii

 was supported as monophyletic (100% BS; 1.00 

PP) as was B. kerrii

 (93% BS, 1.00 PP). The single sample of B. longiflora was sister 

to B. kerrii 

(64% BS, 0.64 PP), but this placement should be viewed with caution.

Multiple accessions for each taxon (except 

B. longiflora) were included to assess 

genetic variation within and among the newly identified taxa. The amount of sequence 

variation within each taxonomic clade was generally less than the amount of variation 

between clades as is shown by the phylograms in Figs. 4

-6. Despite the higher level of 

sequence diversity within B. kingii, it is supported as monophyletic. The results of both 

maximum parsimony and maximum likelihood analyses of DNA data presented above 

consistently identified five mostly strongly supported clades within the B. longiflora 

complex: B. collinsiiB. hamiltonii, B. kerriiB. kingii and B. maxwellii. The data also 

support two sister relationships, one between B. collinsii and B. kerrii, and one between 

B. hamiltonii and B. kingii. The position of 

B. longiflora and B. maxwellii are less clear. 

The relationship of 

B. longiflora to these taxa is uncertain due to polymorphism in the 

nuclear (ITS) data set, but it is clearly closely related to them. Boesenbergia maxwellii 

jumps between the two clades depending upon data set examined. Finally, additional 

data are needed to test between various hypotheses (incomplete concerted evolution 

due to genetic drift post polyploidisation, hybridisation, or incomplete lineage sorting) 

to explain the high level of genetic variation detected in the nrITS of B. kingii. Overall, 

although there are a number of questions remaining, the molecular data allowed for the 

delineation of six taxa.


Boesenbergia longiflora and related taxa

Yüklə 321,14 Kb.

Dostları ilə paylaş:
1   2   3   4   5   6   7   8   9   ...   14

Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur © 2024
rəhbərliyinə müraciət

    Ana səhifə