Content uploaded by Robert K Jansen
Author content
All content in this area was uploaded by Robert K Jansen on Mar 11, 2016
Content may be subject to copyright.
74
TAXON 56 (1)
•
February 2007: 74–88Barber & al. • Hybridization in Macaronesian Sideritis
INTRODUCTION
Congruent patterns shared between independent
phylogenies are generally considered to represent reli-
able estimates of relationships (Hillis, 1987, 1995; Mi-
yamoto & Fitch, 1995). Many researchers support com-
bining datasets in the absence of strong incongruence
(Bull & al., 1993; Chippindale & Wiens, 1994; Brower
& al., 1996; Mason-Gamer & Kellogg, 1996; Johnson &
Soltis, 1998). It has been demonstrated that analyses of
combined datasets can improve phylogenetic reconstruc-
tion and may even recover clades that do not appear in
any of the independent analyses (Olmstead & Sweere,
1994; Whitten & al., 2000; DeBry, 2003; Gatesy & al.,
2003). In cases where incongruence is localized to par-
ticular taxa, or to specific areas of a tree, pruning of the
conflicting taxa or clades may permit the datasets to be
combined for analysis (De Queiroz & al., 1995; Kellogg
& al., 1996). In some instances, however, incongruence
between two datasets may be so well-supported as to
suggest different evolutionary histories for the molecular
markers involved. Incongruence may be assessed ini-
tially by simple inspection of competing topologies and
attendant support measures for conflicting placements of
taxa. A number of statistical tests are also available that
yield a more precise estimate of incongruence between
competing topologies at the individual character level
(Templeton, 1983) and at the global level (Farris & al.,
1994, 1995; Shimodaira & Hasegawa, 1999), although
the accuracy of some of these tests has been increasingly
Hybridization in Macaronesian Sideritis (Lamiaceae): evidence from
incongruence of multiple independent nuclear and chloroplast sequence
datasets
Janet C. Barber1, Courtney C. Finch1, Javier Francisco-Ortega2, Arnoldo Santos-Guerra3 &
Robert K. Jansen4
1 Department of Biology, 3507 Laclede Avenue, Saint Louis University, St. Louis, Missouri 63103, U.S.A.
barberjc@slu.edu (author for correspondence)
2 Department of Biological Sciences, Florida International University, University Park, Miami, Florida
33199, and The Center for Tropical Plant Conservation, Fairchild Tropical Botanic Garden, 10901 Old
Cutler Road, Coral Gables, Florida 33156, U.S.A.
3 Jardín de Aclimatación de La Orotava, Calle Retama Num. 2, E-38400 Puerto de La Cruz, Tenerife,
Canary Islands, Spain
4 Section of Integrative Biology and Institute of Cellular and Molecular Biology, University of Texas,
Austin, Texas 78712, U.S.A.
With 23 suffrutescent to woody perennial species distributed in diverse ecological zones, Sideritis subgenus
Marrubiastrum constitutes one of the largest plant radiations in the Macaronesian archipelagos. In an earlier
study, we investigated the evolution of Sideritis in Macaronesia using chloroplast restriction site (R FLP) data,
but inferences were limited by the lack of a nuclear marker. A second study used sequence data to determine
the continental origin of the Macaronesian group, but that study included only seven island taxa in a much
larger sampling of continental taxa. For the present study, we generated new datasets from sequences of the
nuclear ribosomal DNA internal transcribed spacers (ITS) and two plastid regions (trnL intron, trnT-trnL
intergenic spacer) in order to reconstruct relationships among all extant island taxa using both nuclear and
chloroplast data. Relationships based upon plastid data suggest that there may be a geographical component
to cpDNA variation. Individual phylogenies reconstructed from the nuclear and chloroplast sequence data
were incongruent, and differing placements of taxa were well supported in each of the two datasets. This
incongruence has enabled us to identify several instances of potential cytoplasmic introgression, suggesting
that hybridization may have been important in the evolution of Sideritis in Macaronesia. Because the ITS
phylogeny concurs with current taxonomic circumscriptions in the Macaronesian subgenus, we reassess
patterns of diversification based upon the nuclear tree.
KEYWORDS: hybridization, Lamiaceae, Macaronesia, oceanic islands, phylogenetic incongruence,
Sideritis
75
Barber & al. • Hybridization in Macaronesian SideritisTA XO N 56 (1)
•
February 2007: 74–88
questioned in recent studies (Dolphin & al., 2000; Yoder
& al., 2001; Barker & Lutzoni, 2002; Darlu & Lecointre,
2002; Dowton & Austin, 2002; Grant & Kluge, 2003).
Strong incongruence between competing topologies
may preclude a combined analysis, but does not neces-
sarily invalidate either hypothesis. Carefully examined,
incongruence between phylogenies based on indepen-
dent molecular datasets can illuminate evolutionary pro-
cesses within a group of organisms (Doyle, 1992; Wendel
& Doyle, 1998). In particular, disagreement often occurs
between trees based on different genes for a number of
plant groups that include putative hybrids (e.g., Kellogg
& al., 1996; Mason-Gamer & Kellogg, 1996; Rieseberg
& al., 1996; Sang & al., 1997; Seelanan & al., 1997; Fer-
guson & Jansen, 2002). Introgressive hybridization is
widely considered to be a common occurrence and an of-
ten important evolutionary force among vascular plants
(Stebbins, 1950; Grant, 1981; Rieseberg, 1995; Ellstrand
& al., 1996; Arnold, 1997; Levin, 2000; Rieseberg & al.,
2003). However, despite the generally accepted impor-
tance of hybridization in plant evolution, there is no con-
sensus with respect to its frequency or its ultimate role in
the diversification of oceanic island plants. Some authors
(e.g., Crawford & al., 1987; Sanders & al., 1987; Lowrey,
1995) consider hybridization a rare phenomenon in insu-
lar settings, whereas others (Gillett, 1972; Raven, 1972;
Carr, 1995; Smith & al., 1996) describe a more frequent
incidence.
Sideritis (L.) is one of the three largest genera in the
primarily Old World Lamiaceae subfamily Lamioideae
(Cantino, 1992). The genus contains approximately 150
species of annual and perennial taxa with a predomi-
nantly circum-Mediterranean distribution. Two sub-
genera are recognized. Most species (ca. 125) belong to
the continental subgenus Sideritis which is divided into
three sections: a polyphyletic section comprising all an-
nual taxa (Bentham, 1848) and two monophyletic peren-
nial sections (Bentham, 1834; Obón de Castro & Rivera
Nuñez, 1994). Nine diploid chromosome numbers are
known for the continental taxa, ranging from 2n = 16 to
2n = 34. The smaller subgenus Marrubiastrum (Pérez
de Paz & Negrín Sosa, 1992) comprises the 23 species
endemic to the Macaronesian archipelagoes of Madeira
and the Canary Islands (Fig. 1). These islands are volca-
nic in origin and have never been connected to Africa or
Europe (Carracedo, 1994; Carracedo & al., 2002). They
have wide ranges of both geological age (0.8 to 21 million
years) and distance from continental source areas (100–
620 km). The influence of moisture-laden northeasterly
trade winds combined with altitudes ranging from sea
level to more than 3,600 meters has produced a remark-
able diversity of ecological habitats. Macaronesian Sid-
eritis are found in all described ecological zones in the
two archipelagos. Populations are generally small with
highly restricted distributions, and all but four species
are single-island endemics (Pérez de Paz & Negrín Sosa,
1992). Compared to continental species, morphologi-
cal diversity of the insular taxa is extensive, including a
much greater development of woodiness. In contrast to
a general absence of chromosomal diversity in island
plant genera (Stuessy & Crawford, 1998), Macaronesian
Sideritis are particularly intriguing for the presence of
a dysploid series encompassing 11 diploid chromosome
numbers among the 24 species (vs. 9 for the 125 conti-
nental species).
Two earlier studies based on restriction fragment
length polymorphisms (RFLPs) of the chloroplast ge-
nome and DNA sequence variation (chloroplast and nu-
clear) supported a single introduction of Sideritis to the
islands (Barber & al., 2000, 2002). In the latter study, the
closest continental relative to the insular subgenus was a
Moroccan annual species. This was interpreted as sup-
port for a western Mediterranean origin for the Macaro-
nesian group, and for Carlquist’s (1974) hypothesis of in-
sular evolution of woodiness from herbaceous ancestors;
broader sampling of continental taxa, both annual and
perennial, is needed to verify these hypotheses. In the
RFLP study, anomalous placement of infraspecific taxa
of several species, as well as the polyphyly of three acces-
Fig. 1. Maps showing (above) location of the Macarone-
sian archipelagos of Madeira and the Canary Islands rela-
tive to Morocco and the Iberian peninsula; and (below) the
seven islands of the Canarian archipelago. Macaronesian
Sideritis are distributed on all islands of these two archi-
pelagos.
76
TAXON 56 (1)
•
February 2007: 74–88Barber & al. • Hybridization in Macaronesian Sideritis
sions of S. canariensis from different islands, suggested
that hybridization may have played a role in the evolution
of Sideritis in Macaronesia (Barber & al., 2000).
The study reported here was undertaken to recon-
struct an independent phylogeny of the island subgenus
using a nuclear marker for comparison with the earlier
RFLP phylogeny (Barber & al., 2000). We wished to test
our interpretations of evolution in Macaronesian Sideri-
tis, particularly the suggestion of hybridization. A second
objective was to generate complete parallel datasets for
all taxa in both earlier studies; therefore, we sequenced
two additional chloroplast regions for 23 island taxa
not included in the 2002 study. Finally, we reanalyzed
the RFLP dataset (Barber & al., 2000) applying a more
stringent level of bootstrap support for clade acceptance.
This allowed us to compare hypotheses of relationships
among the Macaronesian taxa as reconstructed from
both cpDNA sequence and RFLP data.
MATERIALS AND METHODS
DNA extraction, PCR amplification and se-
quencing. — Sampling for this study included 55 spe-
cies of Sideritis plus two outgroups; voucher details
are presented in the Appendix. All Macaronesian taxa
were included, as well as three annual species and 22
taxa representing the two continental perennial sections.
Phylogenetic information remains sparse about generic
relationships among the ca. 50 genera of subfamily La-
mioideae, making selection of outgroups difficult. Bal-
lota and Stachys were used based on an earlier study
(Barber & al., 2002) that strongly supported monophyly
of both the genus Sideritis and the insular subgenus Mar-
rubiastrum. Lindqvist & Albert (2002) included several
representatives of Sideritis in their study of the Hawaiian
endemic mint genera. A portion of their study based on
5S-NTS sequence data supported monophyly of Sideri-
tis, although their cpDNA data (rbcL; trnL intron) weakly
supported its inclusion in a polyphyletic Stachys.
New sequences were generated for ITS, the trnL
intron and trnT-trnL intergenic spacer regions for 23
Macaronesian Sideritis taxa using the same DNA extrac-
tions as in the RFLP study of Barber & al. (2000). PCR
amplification for ITS and chloroplast regions followed
the methods used in Barber & al. (2002). Amplification
of ITS was straightforward and, as in the 2002 study,
sequences exhibited no heterogeneity; therefore, we did
not perform cloning for any taxa. All new sequences
were assembled and edited in Sequencher™ (version 4.2;
Gene Codes Corporation, Inc., Ann Arbor, Michigan,
U.S.A.) , and incorporated into the aligned data matrix
from Barber & al. (2002). Genbank accession numbers
are listed with the voucher information (Appendix).
Phylogenetic analyses. — Phylogenetic analyses
were performed under parsimony, maximum likelihood,
and Bayesian inference methods. For the parsimony
analyses, we coded indels as characters, following the
simple indel coding method of Simmons & Ochoter-
ena (2000). Parsimony analyses were conducted with
PAUP* v. 4.0b10 (Swofford, 2002) excluding uninfor-
mative characters and using a heuristic search strategy
with TBR branch swapping and MULTREES option.
For each data partition, 1,000 random addition replicates
were performed. Support for groups was evaluated by
estimating 10,000 bootstrap pseudoreplicates (Felsen-
stein, 1985) using a simple addition sequence and TBR,
but saving only one tree per replicate. This “fast” method
that has been shown in simulation studies (DeBry &
Olmstead, 2000; Mort & al., 2000) to produce estimates
virtually identical to saving multiple trees per replicate.
Maximum likelihood estimates were also conducted in
PAUP* with identical search parameters, including all
characters except indels. The program Modeltest (version
3.06; Posada & Crandall, 1998) was used to identify the
best-fit substitution model for each data partition. Clade
support was estimated using the same procedure as in
parsimony but with bootstrap pseudoreplicates limited to
1,000. Bayesian inference used MrBayes (version 3.0b4;
Ronquist & Huelsenbeck, 2003). Each analysis imple-
mented four simultaneous chains and ran for 1 × 106 gen-
erations. Tree space was sampled every 100th generation
for a sample total of 10,000 trees. Likelihood scores were
plotted against generation to determine when the search
reached stationarity. Trees found in the interval prior to
this point (the “burn-in” period) were discarded before
computing a consensus of the remaining trees. Three in-
dependent runs were completed for each data partition
to ensure that analyses were converging on the optimal
tree set.
We reanalyzed the RFLP data using the 35-taxon
dataset from Barber & al. (2000). Parsimony analyses
performed 1,000 random addition replicates, excluding
uninformative characters and using a heuristic search
strategy with TBR branch swapping and MULTREES
option. Support for groups was evaluated by estimating
10,000 bootstrap pseudoreplicates using a simple addi-
tion sequence and TBR, saving only one tree per repli-
cate. Unlike the earlier study which retained all clades
with ≥ 50% bootstrap support, we applied a more strin-
gent measure and retained only those clades appearing in
≥ 70% of the pseudoreplicates.
For direct comparison with the RFLP dataset, the
cpDNA sequence dataset was pruned to include the same
taxa and reanalyzed under a parsimony criterion. Only
two taxa had differing placements in the RFLP and cp-
DNA sequence phylogenies. One of these, S. candicans,
is the sole Madeiran species. When the RFLP study was
77
Barber & al. • Hybridization in Macaronesian SideritisTA XO N 56 (1)
•
February 2007: 74–88
performed, the only DNA available was extracted from
a plant grown from seeds provided by a botanical gar-
den, whereas all later sequence data were generated from
DNA extractions of plant material directly collected in
Madeira. Placement of sequences both of these acces-
sions (seed-grown and field-collected) in the cpDNA phy-
logeny makes it appear likely that the accession used for
the RFLP study was misidentified as the field-collected
accession. Because it is not possible to redo the RFLP
analysis, this taxon was omitted from further analyses
of the cpDNA data. The second taxon whose placement
differed is S. soluta subsp. gueimaris. Misidentification
was ruled out in this case because both RFLP and se-
quence data for this taxon were generated from the same
DNA extraction, and resequencing it yielded an identical
sequence. Alternative placement of this taxon in the cp-
DNA sequence phylogeny was due to a single character
with only weak bootstrap support (64%). We therefore
combined the RFLP and cpDNA sequence data and ana-
lyzed them under parsimony.
Statistical tests for incongruence. — For a global
assessment of incongruence between the 57-taxon ITS
and cpDNA phylogenies, we used the incong ruence length
difference (ILD) test of Farris & al. (1994, 1995) which
is implemented in PAU P* as the partition homogeneity
test (Swofford, 2002). To test character incongruence, we
conducted Templeton tests (Templeton, 1983) using the
parsimony strict consensus trees for both 57-taxon da-
tasets. A final comparison of competing topologies was
performed via a one-tailed Shimodaira-Hasegawa (SH)
test (Shimodaira & Hasegawa, 1999) as implemented in
PAUP* (“RELL” method; 1,000 bootstrap replicates).
For the SH test, we compared likelihood scores for the
best tree from each Bayesian search vs. 50 trees with the
best likelihood scores from the rival data partition.
RESULTS
Phylogenetic analyses. — The length of the ITS re-
gion (including the 5.8S gene) in Sideritis ranged from
574 to 642 bp. The aligned dataset contained 690 char-
acters, of which 117 (16.9%) were phylogenetically in-
formative (Table 1). Sequence divergence ranged from
0–10.0% within Sideritis, and from 10.0–16.8% between
Sideritis and the outgroups. Scoring indels as charac-
ters resulted in an additional 37 informative characters.
Parsimony analysis found 931 minimum length trees of
308 steps (including autapomorphies); one of the mini-
mum length trees is shown in Figure 2. The CI was 0.66
and RI was 0.92, with both indices calculated excluding
autapomorphies. Maximum likelihood analysis pro-
duced a single tree (not shown) that was topologically
congruent with the parsimony strict consensus tree. The
Bayesian analysis recovered the same clades resolved in
the parsimony strict consensus and likelihood analyses.
Relationships resolved showed no conflict in the trees
recovered under the three optimality criteria. Support
values from the ITS likelihood bootstrap were generally
equal to or lower than those from the parsimony boot-
strap. A total of 17 and 18 nodes, respectively, had sup-
port ≥ 70 in the MP and ML bootstrap analyses. Under
Bayesian inference, 21 nodes had posterior probabilities
(p.p.) ≥ 95 (Table 1).
The length of the trnL intron in Sideritis is ca. 540
bp; the trnT-trnL intergeni c sp acer is ca. 700 bp i n le ngth.
These sequences were combined into a single dataset for
phylogenetic analysis. Lengths of the concatenated se-
quences ranged from 1,093–1,191 bp; the aligned data-
set contained 1,226 characters, 50 (4.1%) of which were
phylogenetically informative (see Table 1). Sequence
divergence ranged from 0–2.9% within Sideritis, and
from 5.0–6.1% between Sideritis and the outgroup taxa.
Scoring indels as characters resulted in an additional 13
informative characters. Parsimony analysis found 16,000
minimum length trees of 103 steps (including autapomor-
phies). The CI (excluding uninformative characters) was
0.84 and RI was 0.93. Maximum likelihood analysis of
the cpDNA sequence data found two equally likely trees
that differed from each other only in the recovery of one
additional node that lacked bootstrap support; both ML
trees were topologically consistent with the strict con-
sensus of parsimony trees. As with ITS, support values
from the cpDNA sequence ML bootstrap were similar to
Table 1. Comparison of the 57-taxon datasets and resulting trees for analyses of Macaronesian Sideritis. Parsimony tree
lengths include autapomorphies; consistency (CI) and retention (RI) indices are calculated excluding uninformative char-
acters. Tree support values compare node support for the three optimality criteria under which datasets were analyzed.
Tree suppor t values
No. of No. of. MP ML Bayesian
informative nodes No. of No. of nodes with
characters resolved nodes with nodes with posterior
(sequence/ No. of Tree in strict bootstrap bootstrap probability
Dataset indel) trees length CI RI consensus ≥ 70 ≥ 70 ≥ 95
ITS 117/37 931 308 0.66 0.92 33 17 18 21
cpDNA combined 50/13 16,000 103 0.84 0.93 15 8 8 15
78
TAXON 56 (1)
•
February 2007: 74–88Barber & al. • Hybridization in Macaronesian Sideritis
or lower than those from the MP bootstrap. Eight nodes
in both MP and ML bootstrap analyses had bootstrap
support (BS) ≥ 70 (Table 1). Bayesian analysis of the
cpDNA sequence data partition reached stationarity at
ca. 30,000 generations; thus, 300 trees were discarded as
burn-in before computing a consensus of the remaining
9,700 trees. Fifteen nodes in the Bayesian analysis of the
cpDNA sequence data had p.p. ≥ 95. Because the main
focus of this paper is on Macaronesian Sideritis, we com-
bined cpDNA sequence data with the earlier RFLP data
for analysis (see following), and have not shown the trees
for the full 57-taxon chloroplast sequence dataset (they
are available from the first author on request).
The reanalysis of the RFLP dataset (excluding
S. candicans) recovered 36 minimum length trees of 269
steps. The parsimony search of the 34-taxon cpDNA se-
quence dataset using the same five continental taxa for
outgroups as the RFLP dataset yielded a single tree of
45 steps. Because the continental taxa used to root the
tree are phylogenetically distant, branch length leading
to the Macaronesian group increased from 3 steps in
the 57-taxon tree to 62 steps in the 34-taxon tree; boot-
strap support increased from 95% to 100%. Despite the
great increase in branch length, relationships among the
Macaronesian taxa in this second analysis were identical
to those of the Macaronesian clade in the 57-taxon cp-
DNA sequence trees. Thus, relationships do not appear
to be affected by long-branch attraction. Figure 3 shows
one of 292 minimum length trees of 319 steps (CI = 0.79,
RI = 0.91) from parsimony analysis of the combined
DNA data (RFLP, trnT-trnL IGS/trnL intron sequences,
and indels). Relationships are the same as in Barber &
al. (2000), with recovery of one well-supported clade
that corresponds to clade 2 and a number of small clades
among the remaining taxa that were also present in the
earlier study.
Incongruence tests. — All tests indicated sig-
nificant conflict between the 57-taxon cpDNA and ITS
topologies. In the Templeton test for the ITS dataset, 56
characters required more changes and two required fewer
to fit onto the cpDNA vs. its own strict consensus tree
(p < 0.0001). Thirteen characters in the cpDNA dataset
required more changes to fit onto the ITS strict consen-
sus than on its own (p = 0.001); no characters required
fewer steps. As a global assessment, the ILD test indi-
cated non-homogeneity for the ITS and cpDNA datasets
(p < 0.01). Similarly, homogeneity was rejected in the SH
tests (p < 0.05) of ITS vs. cpDNA, and vice-versa. Sig-
nificant rejection of homogeneity by all tests, combined
with several clear instances of strong incongruence be-
tween the ITS and cpDNA phylogenies, argued against
combining the data for further analysis.
Ballota
Stachys
S. romana
S. montana
S. syriaca
S. athoa
S. perfoliata
S. clandestina
S. scardica
S. euboae
S. taurica
S. hyssopifolia
S. hirsuta
S. glacialis
S. chamaedryfolia
S. javalambrensis
S. endr. emporitana
S. sericea
S. incana
S. dianica
S. glauca
S. marmorinensis
S. antiatlantica
S. algarviensis
S. tragoriganum
S. murgetana
S. cossoniana
21
10
15
15
1
1
3
13
12
2
1
2
3
1
2
2
3
1
1
2
16
10
Annuals (widespread)
Eastern
Mediterranean
perennials
Western
Mediterranean
perennials
Annual (Morocco)
Macaronesian perennials
*
*
*
*
*
*
*
*
+
+
*
S. discolor
S. dasygnaphala
S. sventenii
S. marmorea
S. lots
y
i
S. pumila
S. pumila
S. candicans
S. dendrochahorra
S. canariensis
S. canariensis
S. canariensis
S. cretica spicata
S. cretica cretica
S. eriocephala
S.
g
omerae
g
omerae
S.
g
omerae perezii
S. nutans
S. macrostach
y
s
1
2
1
1
1
1
S. ferrensis
S. barbellata
1
S. c
y
stosiphon
S. infernalis
S. soluta soluta
S. soluta
g
ueimaris
S. oro. oroteneriffae
S. oro. ara
y
ae
1
21
31
S. kuegleriana
S. nervosa
S. brevicaulis
2
2
2
4
32
1
6
*
*
*
*
**
*
*
+
+
AB
Fig. 2. One of 931 minimum length trees of 308 steps from parsimony analysis of ITS data including indels coded as char-
acters for 57 taxa of Sideritis. Branch lengths from the parsimony trees are shown above nodes. Suppor t values are indi-
cated below: an asterisk (*) indicates nodes supported by parsimony (MP) and likelihood (ML) bootstrap values ≥ 70% and
Bayesian posterior probabilities (p.p.) ≥ 0.95; a plus sign (+) designates nodes supported by MP/ML bootstrap values < 70%
but ≥ 50% and p.p. ≥ 0.95. Dashed lines identify branches that collapse in the strict consensus tree. A, continental taxa of
Sideritis showing monophyly of the two perennial sections and polyphyly of the annual taxa; B, perennial taxa comprising
the Macaronesian endemic subgenus Marrubiastrum.
79
Barber & al. • Hybridization in Macaronesian SideritisTA XO N 56 (1)
•
February 2007: 74–88
DISCUSSION
Clade support and incongruence. — Several
clades in the ITS and cpDNA sequence phylogenies
had low bootstrap support but high Bayesian posterior
probabilities (Table 1). Some authors (e.g., Suzuki & al.,
2002; Simmons & al., 2004) have suggested that Bayes-
ian posterior probabilities inflate clade credibility, and
that bootstrap values are more conservative. Others
contend that Bayesian inference is a less biased estima-
tor of phylogenetic accuracy than either parsimony or
likelihood bootstrapping (Wilcox & al., 2002; Alfaro &
al., 2003). It is perhaps worth noting that many of these
studies are based upon computer simulations rather than
empirical datasets. Still, caution is warranted when
evaluating support values. Whether they differ or agree
in apparent confidence, bootstrap percentages and pos-
terior probabilities are computed in very different ways
and are not directly comparable. Furthermore, some
studies have noted conflicting hypotheses with high
posterior probabilities, as well as a sensitivity of Bayes-
ian inference to model parameters (Buckley & al., 2002;
Douady & al., 2003). The question of whether support is
underestimated by bootstrap values or inflated by pos-
23
91
10
62
2
1
2
12
1
1
3
1
1
1
1
3
3
8
3
7
2
1
7
3
1
2
Clade 2
100
100
100
100
S. discolor (C)
S. dasygnaphala (C)
S. soluta gueimaris (T)
S. canariensis (T)
S. oro. arayae (T)
S. oro. oroteneriffae (T)
S. dendrochahorra (T)
S. sventenii (C)
S. pumila (F)
S. pumila (L)
S. marmorea (G)
S. gomerae perezii (G)
S. macrostachys (T)
S. nutans (G)
S. cretica spicata (G)
S. lotsyi (G)
S. gomerae gomerae (G)
S. barbellata (P)
S. eriocephala (T)
S. brevicaulis (T)
S. cystosiphon (T)
S. infernalis (T)
S. kuegleriana (T)
S. nervosa (T)
S. canariensis (P)
S. cretica cretica (T)
S. ferrensis (H)
S. canariensis (H)
S. soluta soluta (T)
S. hyssopifolia
S. syriaca
S. scardica
S. montana
S. romana
100
73
70
95
100
96
91
95
79
100
2
70
1
2
3
4
Fig. 3. One of 292 minimum length trees of 319 steps from parsimony analysis of combined chloroplast data (RFLP, se-
quence data and coded indels). Branch lengths are shown above nodes and bootstrap support ≥ 70% below. Dashed lines
identify branches that collapse in the strict consensus tree. The shaded bar identifies continental species used as out-
groups (same as those of Barber & al., 2000). Clade 2 of Barber & al. (2000) is indicated by the bold arrow; all other Macaro-
nesian taxa comprised clade 1 in the earlier analysis. Letter in parentheses after taxon name indicates island distribution:
C = Gran Canaria; G = La Gomera; F = Fuerteventura; H = El Hierro; L = Lanzarote; P = La Palma; T = Tenerife; M = Madeira.
Four numbered clades demonstrate potential geographic structuring of cpDNA variation (see text for discussion).
80
TAXON 56 (1)
•
February 2007: 74–88Barber & al. • Hybridization in Macaronesian Sideritis
terior probabilities (or both) will continue to be debated.
Nonetheless, in the present study we are confident in the
relationships resolved because analyses under all three
optimality criteria recover identical clades in each of the
two data partitions.
Although relationships within datasets are consis-
tent under different optimality criteria, there is strongly
supported conflict between phylogenies reconstructed
from the two data partitions. All tests of congruence
were significant for conflict between the cpDNA and
ITS topologies. A number of studies over the last decade
have concluded that some of these statistical tests lack
utility for assessing incongruence. The Templeton test
(Templeton, 1983) and the ILD test (Farris & al., 1994,
1995) have both been criticized as being too conservative
(Lutzoni & Vilgalys, 1995; Cunningham, 1997), and the
latter has been suggested to constitute such a poor mea-
sure of combinability and congruence that its use should
be discontinued (Dolphin & al., 2000; Yoder & al., 2001;
Barker & Lutzoni, 2002; Darlu & Lecointre, 2002; Dow-
ton & Austin, 2002; Grant & Kluge, 2003). Here we find
that the results of these two tests agree with the SH test
in determining that the datasets are significantly differ-
ent. Although combined analysis is thus contraindicated,
these differences shed some light on the evolution of Sid-
eritis in Macaronesia (see discussion following).
Relationships based upon chloroplast data. —
Parsimony analysis of the combined cpDNA data con-
firm results of the earlier RFLP study (Barber & al.,
2000). Relationships among the Macaronesian taxa were
not altered by adding cpDNA sequence and indel data,
although bootstrap support increased for several clades.
Only a few relationships in the cpDNA phylogeny agree
with those based on ITS data. In both cases, S. soluta
subsp. gueimaris is in a clade with the two subspecific
taxa of S. oroteneriffae. Barber & al. (2000) noted the
sympatry and morphological similarity of these taxa.
While the three Gran Canarian taxa—S. dasygnaphala,
S. discolor, S. sventenii—are clearly closely related, they
form a basal polytomy in the ITS phylogeny but are part
of a derived clade in the cpDNA trees. Overall, however,
relationships based on the two data partitions are more
dissimilar than similar.
Infraspecific taxa of three species. — Sideri-
tis cretica, S. gomerae, S. soluta—are not sister in the
chloroplast phylogeny, and multiple accessions of S.
canariensis from Tenerife, La Palma and El Hierro are
polyphyletic. Furthermore, the El Hierro accession of
S. canariensis is strongly supported as sister to S. fer-
rensis (the only other species on that island), which may
potentially be a result of chloroplast capture. A num-
ber of the relationships reconstructed from chloroplast
data suggest that a geographical component to cpDNA
variation may be present. Similar correlations have
been found in several other studies (Soltis & al., 1992;
Mason-Gamer & al., 1995; Wolf & al., 1997). Among
Macaronesian Sideritis, taxa frequently cluster by island,
and often have overlapping or adjacent distributions
within islands. Four such groups are identified by num-
ber in Figure 3. For example, all taxa in group 1 are on
Tenerife and the distribution of each overlaps with that of
at least one of the other taxa in the group. Furthermore,
in the revision of subg. Marrubiastrum (Pérez de Paz &
Negrín Sosa, 1992), one of six newly described hybrids
(Sideritis × bornmuellerii ) is named as a hybrid between
S. canariensis and S. oroteneriffae subsp. oroteneriffae.
Similarly, group 2 taxa are all Gomeran in distribution
and each has small contact zones with one of the other
two taxa in the group. Again, a new hybrid between S.
gomerae subsp. gomerae and S. cretica subsp. spicata
was described in the 1992 revision. Although no hybrids
are known between taxa in groups 3 and 4, all are re-
stricted to localized areas on Tenerife and narrow con-
tact zones between constituent taxa are present.
Evidence for hybridization. — Differing placements
of several taxa in the two phylogenies offer support for
our earlier hypotheses of hybridization within Macaro-
nesian Sideritis (Barber & al., 2000). Table 2 compares
mean and total character difference between these puta-
tive hybrids and their associated taxa in the cpDNA and
ITS data partitions. In the cpDNA phylogeny (Fig. 3),
S. cretica subsp. spicata from La Gomera is placed in a
well-supported clade (group 2) with two other Gomeran
species, whereas S. cretica subsp. cretica from Tenerife
is in a clade that contains mostly other Tenerifan species.
Mean character difference (MCD) and total character
difference (TCD) between the two subspecies (6.4% and
16, respectively) for cpDNA data is much higher than
between S. cretica subsp. spicata and the two taxa with
which it clusters in the cpDNA tree (MCD 0.4–1.6%;
TCD of 1 and 4 characters, respectively). In the nuclear
ITS tree (Fig. 4), the infraspecific taxa of S. cretica are
sister to each other in the MP strict consensus, although
support for this relationship is low. Mean and total char-
acter difference between the subspecies in ITS is 2.8%
and 4 characters. Similarly, chloroplast data place the
two subspecies of S. gomerae (MCD 9.8%; TCD 24) in
different clades with high support, whereas in the nu-
clear ITS tree they form a well-supported clade that also
includes S. nutans. Mean character difference between
these three taxa in ITS ranges from 0–0.7%, with only
a single character difference. These are the only Maca-
ronesian taxa exhibiting a rosette shrub growth form.
Together, they comprise section Empedocleopsis, sup-
ported by the nuclear data as the sole monophyletic taxo-
nomic section of subgenus Marrubiastrum.
Sideritis canariensis is one of only four Macarone-
sian taxa that occur on multiple islands. Populations are
81
Barber & al. • Hybridization in Macaronesian SideritisTA XO N 56 (1)
•
February 2007: 74–88
S. marmorea (G)
S. lotsyi (G)
S. ferrensis (H)
S. barbellata (P)
S. pumila (L)
S. pumila (F)
S. dendrochahorra (T)
S. candicans (M)
S. canariensis (P)
S. canariensis (H)
S. canariensis (T)
S. cre spicata (G)
S. cre cretica (T)
S. gom gomerae (G)
S. gom perezii (G)
S. nutans (G)
S. eriocephala (T)
S. macrostachys (T)
S. kuegleriana (T)
S. nervosa (T)
S. brevicaulis (T)
S. cystosiphon (T)
S. infernalis (T)
S. sol soluta (T)
S. sol gueimaris (T)
S. oro arayae (T)
S. oro oroteneriffae (T)
S. discolor (C)
S. dasygnaphala (C)
S. sventenii (C)
A
B
C
D
44
40
44
44
44
44
44
44
42
42
36
36
36
40
40
38
38 - 42
44
44
44
34
36
44
44
44, 46
34
36
36
36
36
A
B
C
D
LF
SLS
PF
LF
LF
LF
LF
PF, NLS
SLS
SLS
PF
SLS
PF
PF
NLS
NLS
CD, NLS
NLS, SLS
CD
NLS, LF
NLS
NLS
NLS
NLS
NLS, SLS
NLS
SLS
SLS
SLS
HD
Fig. 4. Strict consensus tree (ITS data) showing only Macaronesian species of Sideritis. Letter in parentheses after taxon
name indicates island distribution (abbreviations as in Fig. 2). Diploid chromosome number is shown to the left and eco-
logical zone distribution to the right of taxon name. Open circles denote ecological zones influenced by humid tradewinds;
solid circles indicate arid zones. Hatched circles identify taxa found in both humid and arid zones. Bolded black branches
denote clades supported by Bayesian posterior probabilities (p.p.) ≥ 0.95 and bootstrap percentages ≥ 70 in both parsi-
mony and likelihood analyses. Cross-hatched branches designate clades with Bayesian p.p. ≥ 0.95, but parsimony and/or
likelihood bootstrap values < 70 but ≥ 50%. Clades A and B demonstrate interisland colonization between identical eco-
logical zones; clades C and D show limited intraisland radiation. Ecological zone abbreviations are as follows: CD, coastal
desert; SLS, southern lowland scrub; NLS, northern lowland scrub; LF, laurel forest; PF, pine forest; HD, high altitude
desert.
Table 2. Comparison of mean (uncorrected p-distance) and total character difference between putative hybrid taxa in the
cpDNA and nuclear ITS data partitions. Letter after taxon name identifies island distribution (abbreviations as in Fig. 2).
cpDNA ITS
Mean char. Total char. Mean char. Total char.
Reference taxon difference difference difference difference
S. canariensis (H) S. ferrensis (H) 0.016 4 0.056 10
S. canariensis (P) 0.044 11 0.007 1
S. canariensis (T) 0.139 35 0.042 6
S. cretica subsp. spicata (G) S. lotsyi (G) 0.004 1 0.049 7
S. gomerae subsp. gomerae (G) 0.016 4 0.084 12
S. cretica subsp. cretica (G) 0.064 16 0.028 4
S. gomerae subsp. gomerae (G) S. gomerae subsp. perezii (G) 0.098 24 0.007 1
S. nutans (G) 0.045 11 0.000 0
S. soluta subsp. gueimaris (T) S. oroteneriffae var. arayae (T) 0.020 4 0.014 2
S. oroteneriffae var. oroteneriffae (T) 0.016 4 0.000 0
S. soluta subsp. soluta (T) 0.108 27 0.035 5
82
TAXON 56 (1)
•
February 2007: 74–88Barber & al. • Hybridization in Macaronesian Sideritis
confined to the laurel forest zone on the islands of Tener-
ife, El Hierro and La Palma. Accessions of S. canariensis
from each of these islands are polyphyletic in the cpDNA
phylogeny (Fig. 3), differing from e ach other by 4.4–13.9%
(MCD) and 11–35 characters. In the ITS tree, they form
a well-supported clade (designated clade A, Fig. 4) with
a mean and total character difference of 0.7–4.2% and
1–6 characters. Two of the three populations (El Hierro
and Tenerife) group with other taxa from the same island
in the cpDNA trees (Fig. 3), again suggesting a poten-
tial link between geographical distribution and cpDNA
variation. In addition, Pérez de Paz & Negrín Sosa (1992)
identified S. canariensis as a parental taxon in three of
six hybrids newly described in their revision, implying
that crossing barriers in this species may be low. Prior
to the advent of molecular phylogenetics, hybridization
was generally considered a relatively rare contributor to
the evolution of insular floras (Humphries, 1979; Gan-
ders & Nagata, 1984). However, early molecular studies
of Macaronesian plant groups (e.g., Argyranthemum,
Francisco-Ortega & al., 1996; woody Sonchus, Kim &
al., 1996), found that hybridization had in fact played a
role in diversification of the island lineages. More recent
molecular phylogenies (e.g., Mort & al., 2002; Percy &
Cronk, 2002; Ackerfield & Wen, 2003) continue to chal-
lenge the earlier assumption. Sideritis provides addi-
tional evidence of the potential role of hybridization in
the evolution of the Macaronesian flora.
In contrast to the foregoing discussion, relationships
of the infraspecific taxa of S. soluta remain unclear. This
taxon is diffusely distributed in pine forests in deep can-
yons on the south side of Tenerife, with a distinct sub-
species gueimaris comprising a single isolated popula-
tion in the canyon of Güimar at the northeastern limit of
the species’ distribution. Distribution of S. soluta subsp.
gueimaris overlaps with that of S. oroteneriffae, with
which it is easily confused (Negrín Sosa & Pérez de Paz,
1988; Pérez de Paz & Negrín Sosa, 1992). Nuclear ITS
data place S. soluta subsp. gueimaris in a clade with the
two subspecies of S. oroteneriffae (MCD 0–1.3%; TCD
0–2), although support for this resolution is low. Chlo-
roplast RFLP data resolve these three taxa with several
other Tenerifan species and the three Gran Canarian taxa
(clade 1; Fig. 3). Mean and total character difference be-
tween the two subspecies of S. soluta is 10.8% and 27 in
the cpDNA dataset, and 3.5% and 5 in ITS.
Although hybridization does not explain fully the
high degree of dissimilarity between phylogenetic hy-
potheses, it is strongly implicated as one of the major
causes of incongruence between the data partitions.
Geographical structure is apparent in cpDNA varia-
tion, suggesting hybridization and introgression of the
chloroplast genome among species on the same island.
Conversely, ITS variation has a much higher level of
concordance with morphology. ITS sequences for pu-
tative hybrids exhibited no heterogeneity which could
mean that sufficient time has passed to permit complete
homogenization of divergent parental ITS sequences, or
alternatively, that parental taxa were not divergent when
the original hybridization event occurred. Data gathered
for this study do not permit us to distinguish between
these hypotheses, nor to determine whether hybridization
occurred early or late in the insular radiation of Sideritis.
However, hybrids among island taxa have only recently
been formally described (Pérez de Paz & Negrín Sosa,
1992) and it may be that hybridization in the group is due
to relatively recent anthropogenic disturbances that have
resulted in the breakdown of ecological barriers.
Patterns of diversification. — Many studies of
Macaronesian plant groups have described a pattern of
diversification consistent with interisland colonization
between the same or similar ecological zones (Kim &
al., 1996; Mes & ’T Hart, 1996; Francisco-Ortega & al.,
1996, 2001, 2002; Allan & al., 2004; Trusty & al., 2005).
In contrast, our earlier study based on RFLP data sug-
gested that the pattern for Sideritis was primarily one of
within-island radiation (Barber & al., 2000). However,
the ITS phylogeny presents more of a mixed pattern. At
least two clades (Fig. 4, clades C and D) do in fact show
limited evidence of intraisland radiation, but two other
clades (A and B) clearly show interisland colonization
between the same ecological zones. This mixed pattern
has also been found for the insular taxa of the woody
Sonchus alliance (Lee & al., 2005). Thus, diversifica-
tion of Sideritis in the islands conforms to patterns seen
in other studies of Macaronesian plant groups. There is
strong support for the sister relationship of S. dendrocha-
horra and S. candicans, suggesting that the Madeiran ar-
chipelago was colonized from Tenerife. Although colo-
nization from the Canaries to Madeira requires dispersal
against existing wind and ocean currents, this pattern
has been demonstrated for a number of Macaronesian
plant groups in addition to Sideritis (Böhle & al., 1996;
Kim & al., 1996; Panero & al., 1999; Francisco-Ortega
& al., 2002; Mort & al., 2002; Carine & al., 2004; Trusty
& al., 2005; Lee & al., 2005).
Few studies of the Macaronesian flora have explic-
itly addressed the initial site of introduction to the is-
lands. Many lack sufficient resolution to determine the
ancestral island and in some cases where one is iden-
tified, robust support is lacking. Fairfield & al. (2004)
posited initial colonization of Aichryson Webb & Ber-
thelot to Fuerteventura, one of the easternmost Canary
Islands situated ca. 100 km from the African continent.
An investigation of Olea europaea L. (Hess & al., 2000)
also suggested introduction to Fuerteventura, followed
by stepping-stone dispersal across the Canaries from
east to west. Francisco-Ortega & al. (2002) identified the
83
Barber & al. • Hybridization in Macaronesian SideritisTA XO N 56 (1)
•
February 2007: 74–88
Canary Islands as the ancestral archipelago for Crambe
L., but were unable to distinguish whether Tenerife or
Fuerteventura was the ancestral island. A phylogenetic
analysis of Descurainia Webb & Berthel. suggests that
Tenerife is ancestral for the Macaronesian taxa of that
genus (Goodson & al., 2006). In contrast to these stud-
ies, Lee & al. (2005) determined that the Macaronesian
woody Sonchus alliance most likely originated on Gran
Canaria. Phylogenetic reconstructions for two coleop-
teran lineages also show this pattern. Gran Canaria is
identified as the ancestral island in a population level
assessment of Brachyderes Schönherr (Emerson & al.,
2000a). Although there have been multiple introduc-
tions of Calathus Bonelli beetles to Macaronesia, Gran
Canaria appears ancestral for at least one island clade
(Emerson & al., 2000b). In the present study, the early
divergence in the ITS phylogeny of the three Gran Ca-
narian Sideritis species (Fig. 4) suggests that initial in-
troduction into Macaronesia was also to that island. We
cannot, of course, rule out the possibility that the origi-
nal colonizer is no longer extant.
Chromosome evolution. — The most intriguing
aspect of Macaronesian Sideritis remains the question
of chromosomal evolution. Given the reported paucity
of chromosomal diversity (Stuessy & Crawford, 1998)
among oceanic island plants, why is this group such a
spectacular exception? Chromosomal diversity in insu-
lar Sideritis reflects that of the continental species, but
the rate of diversification appears to be greatly acceler-
ated: 11 diploid numbers for 23 Macaronesian species vs.
9 diploid numbers for the 125 species of the continental
subgenus. Barber & al. (2000) hypothesized that the ap-
pare nt elev at ed ra te of ch ro mosom al ch an ge in t he in su la r
subgenus could be due to the dynamics of reproduction in
small, isolated populations, including bottlenecks caused
by new colonizations of islands or ecological zones. We
also noted that inbreeding (if present) could contribute
to chromosomal instability that could in turn be exacer-
bated by self-compatibility. Breeding system in Sideritis
remains to be investigated, but one study of S. discolor,
an endangered Macaronesian endemic, found that 85%
of genetic variation occurred within populations (Batista
& al., 2004) rather than between them, consistent with
the expected distribution of variation for outbreeding
species (Hamrick, 1990).
Diploid chromosome numbers mapped onto the
RFLP phylogeny (Barber & al., 2000) suggested a bi-
modal pattern of change, implying that both ascending
and descending aneuploidy had occurred within the
Macaronesian subgenus. No clear pattern, however, is
evident in the combined cpDNA analysis (Fig. 3). The
ITS phylogeny corresponds more closely with taxonomic
boundaries, but is insufficiently resolved for a definitive
interpretation. Nevertheless, mapping diploid chromo-
some numbers onto the ITS tree (Fig. 4) reveals a pattern
that is consistent with aneuploid increase in the subge-
nus; i.e., basal branches have 2n = 36, whereas higher
numbers occur in the most derived clades. A similar
trend in which increasing chromosome numbers are cor-
related with phylogenetically derived clades has been
documented in slipper orchids (Cox & al., 1997, 1998).
Our present knowledge of Macaronesian Sideritis
precludes a rigorous assessment of chromosomal evo-
lution, but points the way for future work. The mecha-
nisms of chromosomal evolution in Sideritis have yet to
be elucidated. Marrero (1992) suggested that the varia-
tion present in the Macaronesian taxa could be due to
changes in chromosome structure caused by centric fis-
sion or centric fusion (i.e., Robertsonian changes). Al-
though such changes are relatively rare in plants (see
review by Jones, 1998), with most reports from gymno-
sperms or monocots, documented instances in eudicots
are known (Kyhos, 1965; De Azkue & Martínez, 1988;
Pijnacker & Schotsman, 1988; Gibby & al., 1996; Cerbah
& al., 1998). Karyotype data for Macaronesian Sideritis
are scant (Marrero, 1992), but the few that are available
show that species with lower diploid numbers have more
symmetrical karyotypes than those with higher diploid
numbers, consistent with a hypothesis of Robertsonian
change.
Alternative possibilities exist that could explain the
striking chromosomal diversity in Sideritis. It may be
that chromosome evolution in Sideritis is simply highly
labile, as it is in some other genera, e.g., Nicotiana
(Wheeler, 1945; Chase & al., 2003), Crocus (Brighton,
1976, 1978). A decrease in chromosome number could
result from large translocations followed by the loss of
the centromere, as has been demonstrated in genera such
as Chaenactis (Kyhos, 1965), Haplopappus (Jackson,
1962), and Coreopsis (Smith, 1974; Crawford & Smith,
1982). Polyploidization could allow the loss of one or
more chromosomes without negative effects. Polyploidy
has come to be viewed as an important evolutionary
force (see review by Otto & Whitton, 2000), and es-
timates of its frequency in plants range from 30–80%
(Masterson, 1994). Polyploidy has not been documented
in Sideritis, but neither has it been ruled out. That the
Macaronesian taxa all have chromosome numbers at
the high end of the range for the genus suggests that the
possibility of polyploidy deserves further inquiry. Fi-
nally, an increasing number of studies (Kalendar & al.,
2000; Jiang & al., 2004; Raskina & al., 2004a, b; Lai &
al., 2005) have shown that repetitive elements such as
telomeric sequences, ribosomal loci, and mobile DNA
sequences (e.g., transposons) may vary extensively in
sequence motif and copy number between even closely
related species, and that this variation may contribute to
chromosomal change and, ultimately, genome evolution.
84
TAXON 56 (1)
•
February 2007: 74–88Barber & al. • Hybridization in Macaronesian Sideritis
CONCLUSIONS
Although numerous molecular phylogenies of Maca-
ronesian plant groups have been published in the last de-
cade, only a small proportion of them have been based on
data from more than a single molecular marker. Our own
initial study of Sideritis was based solely on data from
the maternally inherited chloroplast genome, and many
of our inferences with respect to evolutionary patterns in
the Macaronesian subgenus have been reassessed in light
of the current study. Here we reconstructed relationships
using DNA sequences of nuclear ITS and a combined
cpDNA dataset comprising RFLP and sequence data. The
resulting phylogenies offer strong support for our earlier
hypotheses that hybridization has played a role in evolu-
tion of the island taxa. However, our interpretations of the
patterns of diversification and chromosomal change have
been modified. Although within-island radiation appeared
to be the predominant pattern in the earlier cpDNA phy-
logeny, we now f ind that intra-island colonization between
the same ecological zones is equally important. The pat-
tern of change in chromosome number interpreted on the
ITS phylogeny is consistent with aneuploid increase, but
resolution remains insufficient for a definitive interpreta-
tion of chromosomal evolution. However, the application
of modern cytogenetic techniques to previously intractable
questions of mechanistic change, combined with improved
phylogenetic resolution which we are pursuing with new
markers, hold exciting promise for solving the puzzle of
chromosomal evolution in Macaronesian Sideritis.
ACKNOWLEDGMENTS
The authors thank A. Marrero, E. Barreno, S. Sá-Fontinha,
David Draper Munt and Concepción Obón de Castro for assistance
with field collections in the Canary Islands, Madeira, Portugal and
Spain. We are grateful to the botanic gardens that donated seeds
for some of the continental taxa used in our study, and to the Plant
Resources Center at the University of Texas and the Instituto de
Investigacione Agrarias de Canarias for the use of their facilities.
We are indebted to Richard Olmstead and a second reviewer for
constructive criticism of an earlier version of the manuscript. This
work was supported in part by a Beaumont Faculty Development
grant to JCB from Saint Louis University and by the National Sci-
ence Foundation DEB-9801594 to RKJ and JCB.
LITERATURE CITED
Ackerfield, J. & Wen, J. 2003. Evolution of Hedera (the ivy
genus, Araliaceae): insights from chloroplast DNA data.
Int. J. Pl. Sci. 16 4: 593– 602.
Alfaro, M.E., Zoller, S. & Lutzoni, F. 2003. Bayes or boot-
strap? A simulation study comparing the performance of
Bayesian Markov chain Monte Carlo sampling and boot-
strapping in assessing phylogenetic confidence. Molec.
Biol. Evol. 20: 255–266.
Allan, G., Francisco-Ortega, J., Santos-Guerra, A.,
Boerner, E. & Zimmer, E.A. 2004. Molecular phyloge-
netic evidence for the geographic origin and classifica-
tion of Canary Island Lotus (Fabaceae: Loteae). Molec.
Phylog. Evol. 32: 123–138.
Arnold, M.L. 1997. Natural Hybridization and Evolution.
Oxford University Press, New York.
Barber, J.C., Francisco-Or tega, J., Santos- Guerra, A., Mar-
rero, A. & Jansen, R.K. 2000. Evolution of endemic Sid-
eritis (Lamiaceae) in Macaronesia: insights from a chloro-
plast DNA restriction site analysis. Syst. Bot. 25: 633–647.
Barber, J.C., Francisco-Ortega, J., Santos-Guerra, A.,
Turner, K.G. & Jansen, R.K. 2002. Origin of Macaro-
nesian Sideritis L. (Lamiaceae) inferred from nuclear and
chloroplast sequence datasets. Molec. Phylog. Evol. 23:
293–306.
Barker, F.K. & Lutzoni, F.M. 2002. The utility of the incon-
gruence length difference test. Syst. Biol. 51: 625–637.
Batista, F., Bouza, N., González-Pérez, M.A., Caujapé-
Castells, J. & Sosa, P.A. 2004. Genetic variation within
and between populations of two endangered endemic spe-
cies of the laurel forest from the Canary Islands, Myrica
rivas-martinezii (Myricaceae) and Sideritis discolor (La-
miaceae). Austral. J. Bot. 52: 471–480.
Bentham, G. 1834. Labiatarum Genera et Species. J. Ridge-
way, London.
Bentham, G. 1848. Sideritis L. Pp. 212–226 in: De Candolle,
A. (ed.), Prodromus Systematis Naturalis Regni Vegetabi-
lis. Victoris Masson, Paris.
Böhle, U.-R., Hilger, H.H. & Martin, W.F. 1996. Island
colonization and evolution of the insular woody habit in
Echium L. (Boraginaceae). Proc. Natl. Acad. Sci. U.S.A.
93: 1174 0 –11745.
Brighton, C.A. 1976. Cytological problems in the genus Cro-
cus (Iridaceae). I. Crocus vernus aggregate. Kew Bull. 31:
33– 46.
Brighton, C.A. 1978. Cytological problems in the genus Cro-
cus (Iridaceae). II. Crocus cancellatus aggregate. Kew
Bull. 32: 33–44.
Brower, A.V.Z., DeSalle, R. & Vogler, A. 1996. Gene trees,
species trees, and systematics: a cladistic perspective.
Annual Rev. Ecol. Syst. 27: 423–450.
Buckley, T.R., Arensburger, P., Simon, C. & Chambers, G.K.
2002. Combined data, Bayesian phylogenetics, and the ori-
gin of the New Zealand cicada genera. Syst. Biol. 51: 4–18.
Bull, J. J., Huelsenbeck, J.P., Cunningham, C.W., Swofford,
D.L. & Waddell, P.J. 1993. Partitioning and combining
data in phylogenetic analysis. Syst. Biol. 42: 384–397.
Cantino, P.D. 1992. Toward a phylogenetic classification of
the Labiatae. Pp. 27–37 in: Harley, R.M. & Reynolds, T.
(ed s.), Advances in Labiate Science. Royal Botanic Gar-
dens Press, Kew.
Carine , M.A, Russell , S.J., Santos-Guerra , A. & Francisco-
Ortega, J. 2004. Relationships of the Macaronesian and
Mediterranean f loras: molecular evidence for multiple
colonizations into Macaronesia and back-colonization of
the continent in Convolvulus (Convolvulaceae). Amer. J.
Bot. 91: 1070–1085.
85
Barber & al. • Hybridization in Macaronesian SideritisTA XO N 56 (1)
•
February 2007: 74–88
Carlquist, S. 1974. Island Biology. Columbia University
Press, New York.
Carr, G.D. 1995. A fully fertile intergeneric hybrid derivative
from Argyroxiphium sandwicense ssp. macrocephalum
× Dubautia menziesii (Asteraceae) and its relevance to
plant evolution in the Hawaiian Islands. Amer. J. Bot. 82:
1574–1581.
Carracedo, J.C. 1994. The Canary Islands: an example of
structural control on the growth of large oceanic-island
volcanoes. J. Volcanol. Geoth. Res. 60: 225–241.
Carracedo, J.C., Pérez, F.J., Ancochea, E., Meco, J.,
Hernán, F., Cubas, C.R., Casillas, R., Rodriguez, E.
& Ahijado, A. 2002. Cenozoic volcanism II: the Canary
Islands. Pp. 439–472 in: Gibbons, W. & Moreno, T. (eds.),
The Geology of Spain. Geological Society, London.
Cerbah, M., Souza-Chies, T., Jubier, M.F., Lejeune, B. &
Siljak-Yakovlev, S. 1998. Molecular phylogeny of the
genus Hypochaeris using internal transcribed spacers
of nuclear rDNA: inference for chromosomal evolution.
Molec. Biol. Evol. 15: 345–354.
Chase, M.W., Knapp, S., Cox, A.V., Clarkson, J.J., Butsko,
Y., Joseph, J., Savolainen, V. & Parokonny, A.S. 2003.
Molecular systematics, GISH and the origin of hybrid
taxa ain Nicotiana (Solanaceae). Ann. Bot. 92: 107–127.
Chippindale, P.T. & Wiens, J.J. 1994. Weighti ng, pa rtition-
ing, and combining characters in phylogenetic analysis.
Syst. Biol. 43: 278–287.
Cox, A.V., Pridgeon, A.M., Albert, V.A. & Chase, M.W.
1997. Phylogenetics of the slipper orchids (Cypripedioi-
deae: Orchidaceae): nuclear rDNA ITS sequences. Pl.
Syst. Evol. 208: 197–223.
Cox, A.V., Abdelnour, G.J., Bennett, M.D. & Leitch, I.J.
1998. Genome size and karyotype evolution in the slipper
orchids (Cypripedioideae: Orchidaceae). Amer. J. Bot. 85:
681–687.
Crawford, D.J. & Smith, E.B. 1982. Allozyme variation in
Coreopsis nucensoides and C. nuecensis (Composit ae).
a progenitor-derivative species pair. Evolution 36: 379–
386.
Crawford, D.J., Whitkus, R. & Stuessy, T.F. 1987. Plant evo-
lution and speciation on oceanic islands. Pp. 183–199 in:
Urbanska, K.M. (ed.), Differentiation Patterns in Higher
Plants. Academic Press, New York.
Cunningham, C.W. 1997. Can three incongruence tests pre-
dict when data should be combined? Molec. Biol. Evol.
14: 733–740.
Darlu, P. & Lecointre, G. 2002. When does the incongru-
ence length difference test fail? Molec. Biol. Evol. 19:
432– 437.
De Azkue, D. & Martínez, A. 1988. DNA content and chro-
mosome evolution in the shrubby Oxalis. Genome 30:
52–57.
DeBry, R.W. 2003. Identifying conf licting signal in a multi-
gene analysis reveals a highly resolved tree: the phylog-
eny of Rodentia (Mammalia). Syst. Biol. 52: 604–617.
DeBry, R.W. & Olmstead, R.G. 2000. A simulation study of
reduced tree-search effort in bootstrap resampling analy-
sis. Syst. Biol. 49: 171–179.
De Queiroz, A., Donoghue, M.J. & Kim, J. 1995. Sepa-
rate versus combined analysis of phylogenetic evidence.
Annual Rev. Ecol. Syst. 26: 657–681.
Dolphin, K., Belshaw, R., Orme, C.D.L. & Quicke, D.L.J.
2000. Noise and incongruence: interpreting results of the
incongruence length difference test. Molec. Phylog. Evol.
17: 401–406.
Douady, C.J., Delsuc, F., Boucher, Y., Doolittle, W.F. &
Douzery, E.J.P. 2003. Comparison of Bayesian and max-
imum likelihood bootstrap measures of phylogenetic reli-
ability. Molec. Biol. Evol. 20: 248–254.
Dowton, M. & Austin, A.D. 2002. Increased congruence
does not necessarily indicate increased phylogenetic ac-
curacy—the behavior of the incongruence length differ-
ence test in mixed-model analyses. Syst. Biol. 51: 19–31.
Doyle, J.J. 1992. Gene trees and species trees: molecular
systematics as one-character taxonomy. Syst. Bot. 17:
144 –163.
Ellstrand, N.C., Whitkus, R. & Rieseberg, L.H. 1996. Dis-
tribution of spontaneous plant hybrids. Proc. Natl. Acad.
Sci. U.S.A. 93: 5090–5093.
Emerson, B.C., Oromí, P. & Hewitt, G.M. 2000a. Colo-
nization and diversification of the species Brachyderes
rugatus (Coleoptera) on the Canary Islands: evidence
from mitochondrial DNA COII gene sequences. Evolu-
tion 54: 911–923.
Emerson, B.C., Oromí, P. & Hewitt, G.M. 2000b. Interpret-
ing colonization of the Calathus (Coleoptera: Carabidae)
on the Cana ry Islands and Ma deira throug h the application
of the parametric bootstrap. Evolution 54: 2081–2090.
Fairfield, K.N., Mort, M.E. & Santos-Guerra, A. 2004.
Phylogenetics and evolution of the Macaronesian mem-
bers of the genus Aichryson (Crassulaceae) inferred from
nuclear and chloroplast data. Pl. Syst. Evol. 248: 71–83.
Farris, J.S., Kallersjö, M., Kluge, A.G. & Bult, C. 1994.
Testing significance of incongruence. Cladistics 10: 315–
319.
Farris, J.S., Kallersjö, M., Kluge, A.G. & Bult, C. 1995.
Constructing a significance test for incongruence. Syst.
Biol. 44: 570–572.
Felsenstein , J. 1985. Confidence limits on phylogenies: an ap-
proach using the bootstrap. Evolution 39: 783–791.
Ferguson, C.J. & Jansen, R.K. 2002. A chloroplast DNA
phylogeny of eastern Phlox (Polemoniaceae): implica-
tions of congruence and incongruence with the ITS phy-
logeny. Amer. J. Bot. 89: 1324–1335.
Francisco-Ortega, J., Barber, J.C., Santos-Guerra, A.,
Febles-Hernández, R. & Jansen, R.K. 2001. Origin
and evolution of the endemic genera of Gonosperminae
(Asteraceae: Anthemideae) from the Canary Islands:
evidence from nucleotide sequences of the internal tran-
scribed spacers of the nuclear ribosomal DNA. Amer. J.
Bot. 88: 161–169.
Francisco-Ortega, J., Fuertes-Aguilar, J., Kim, S.-C.,
Santos-Guerra, A., Crawford, D.J. & Jansen, R.K.
2002. Phylogeny of the Macaronesian endemic Crambe
section Dendrocrambe (Brassicaceae) based on internal
transcribed spacer sequences of nuclear ribosomal DNA.
Amer. J. Bot. 89: 1984–1990.
Francisco-Ortega, J., Jansen, R.K. & Santos-Guerra, A.
1996. Chloroplast DNA evidence of colonization, adap-
tive radiation, and hybridization in the evolution of the
Macaronesian flora. Proc. Natl. Acad. Sci. U.S.A. 93:
4085–4090.
Ganders, F.R. & Nagata, K.M. 1984. The role of hybridiza-
tion in the evolution of Bidens on the Hawaiian Islands.
86
TAXON 56 (1)
•
February 2007: 74–88Barber & al. • Hybridization in Macaronesian Sideritis
Pp. 179–194 in: Grant, W.F. (ed.), Plant Biosystematics.
Academic Press, New York.
Gatesy, J., Amato, G., Nordell, M., Desalle, R. & Hayashi,
C. 2003. Combined support for wholesale taxic atavism
in gaviline crocodylians. Syst. Biol. 52: 403–422.
Gibby, M., Hinnah, S., Marais, E.M. & Albers, F. 1996. Cy-
tological variation and evolution within Pelargonium sec-
tion Hoarea (Geraniaceae). Pl. Syst. Evol. 203: 111–142.
Gillett, G.W. 1972. The role of hybridization in the evolution
of the Hawaiian f lora. Pp. 141–159 in: Valentine, D.H.
(ed .), Taxonomy, Phytogeography and Evolution. Aca-
demic Press, New York.
Goodson, B.E., Santos-Guerra, A. & Jansen, R.K. 2006.
Molecular systematics of Descurainia (Brassicaceae) in
the Canary Islands: biogeographic and taxonomic impli-
cations. Taxo n 55: 671–682.
Grant, T. & Kluge, A. 2003. Data exploration in phylogenetic
inference: scientific, heuristic or neither. Cladistics 19:
379– 418.
Grant, V. 1981. Plant Speciation. Columbia University Press,
New York.
Hamrick, J.L. 1990. Isozymes and the analysis of genetic
structure in plant populations. Pp. 87–105 in: Soltis, D.E.
& Soltis, P.S. (eds.), Isozymes in Plant Biology. Chapman
& Hall, London.
Hess, J., Kadereit, J.W. & Vargas, P. 2000. The colonization
history of Olea europaea L. in Macaronesia based on in-
ternal transcribed spacer 1 (ITS-1) sequences, randomly
amplified polymorphic DNAs (RAPD), and intersimple
sequence repeats (ISSR). Molec. Ecol. 9: 857–868.
Hil lis, D.M. 1987. Molecula r versus morphological a pproaches
to systematics. Annual Rev. Ecol. Syst 18: 23–42.
Hillis, D.M. 1995. Approaches for assessing phylogenetic ac-
curacy. Syst. Biol. 44: 3–16.
Humphries, C.J. 1979. Endemism and evolution in Macaro-
nesia. Pp. 171–199 in: Bramwell, D. (ed.), Plants and Is-
lands. Academic Press, New York.
Jackson, R.C. 1962. Interspecific hybridization in Haplo-
pappus and its bearing on chromosome evolution in the
Blepharodon section. Amer. J. Bot. 49: 119–132.
Jiang, N., Bao, Z., Zhang, X., Eddy, S.R. & Wessler, S.R.
2004. Pack-MULE transposable elements mediate gene
evolution in plants. Nature 431: 569–573.
Johnson, L.A. & Soltis, D.E. 1998. Assessing congruence:
empirical examples from molecular data. Pp. 297–348 in:
Soltis, D.E., Soltis, P.S. & Doyle, J.J. (eds.), Molecular
Systematics of Plants II: DNA Sequencing. Kluwer Aca-
demic Publishers, Boston.
Jones, K. 1998. Robertsonian fusion and centric fission in
karyotype evolution of higher plants. Bot. Rev. 64: 273–
289.
Kalendar, R., Tanskanen, J., Immonen, S., Nevo, E. &
Schulman, A.H. 2000. Genome evolution of wild bar-
ley (Hordeum spontaneum) by BARE-1 retrotransposon
dynamics in response to sharp microclimatic divergence.
Proc. Natl. Acad. Sci. U.S.A. 97: 6603–6607.
Kellogg, E.A., Appels, R. & Mason-Gamer, R.J. 1996.
When genes tell different stories: the diploid genera of
Triticeae (Grami neae). Syst. Bot. 21: 321–347.
Kim, S.- C., Crawford, D.J., Francisco-Or tega, J. & Santos-
Guerra, A. 1996. A common origin for woody Sonchus
and five related genera in the Macaronesian islands: mo-
lecular evidence for extensive radiation. Proc. Natl. Acad.
Sci. U.S.A. 93: 7743–7748.
Kyhos, D.W. 1965. The independent aneuploid origin of two
species of Chaenactis (Compositae) from a common an-
cestor. Evolution 19: 26–43.
Lai, J., Li, Y., Messing, J. & Dooner, H.K. 2005. Gene move-
ment by Helitron transposons contributes to the haplo-
type variability of maize. Proc. Natl. Acad. Sci. U.S.A.
102: 9068–9073.
Lee, C., Kim, S.-C., Lundy, K. & Santos-Guerra, A. 2005.
Chloroplast DNA phylogeny of the woody Sonchus
alliance (Asteraceae: Sonchinae) in the Macaronesian
Islands. Amer. J. Bot. 92: 2072–2085.
Levin, D.A. 2000. The Origin, Expansion, and Demise of
Plant Species. Oxford University Press, Oxford.
Lindqvist, C. & Albert, V.A. 2002. Origin of the Hawaiian
endemic mints within North American Stachys (Lamia-
ceae). Amer. J. Bot. 89: 1709–1724.
Lowrey, T.K. 1995. Phylogeny, adaptive radiation, and
biogeography of Hawaiian Tetra molopium (Asteraceae,
Astereae). Pp. 195–220 in: Wagner, W.L. & Funk, V.A.
(ed s.), Hawaiian Biogeography: Evolution on a Hot Spot
Archipelago. Smithsonian Institution Press, Washington.
Lutzoni, F. & Vilgalys, R. 1995. Integration of morphological
and molecular data sets in estimating fungal phylogenies.
Canad. J. Bot. 73: S649–S659.
Marrero, A. 1992. Chromosomal evolutionary trends in the
genus Sideritis subgenus Marrubiastrum. Pp. 247–256 in:
Harley, R.M. & Reynolds, T. (eds.), Advances in Labiate
Science. Royal Botanic Gardens Press, Kew.
Mason-Gamer, R.J., Holsinger, K.E. & Jansen, R.K. 1995.
Chloroplast DNA haplotype variation within and among
populations of Coreopsis grandiflora (Asteraceae).
Molec. Biol. Evol. 12: 371–381.
Mason-Gamer, R.J. & Kellogg, E.A. 1996. Testing for phy-
logenetic conflict among molecular data sets in the tribe
Triticeae (Grami neae). Syst. Biol. 45: 524–545.
Masterson, J. 1994. Stomatal size in fossil plants: evidence
for polyploidy in majority of angiosperms. Science 264:
421–423.
Mes, T.H.M. & ’T Hart, H. 1996. The evolution of growth
forms in the Macaronesian genus Aeonium (Crassulaceae)
inferred from chloroplast DNA RFLPs and morphology.
Molec. Ecol. 5: 351–363.
Miyamoto, M.M. & Fitch, W.M. 1995. Testing species phy-
logenies and phylogenetic methods with congruence.
Syst. Biol. 44: 64–76.
Mort, M.E., Soltis, D.E., Soltis, P.S., Francisco-Ortega, J. &
Santos-Guerra, A. 2002. Phylogenetics and evolution of
the Macaronesian clade of Crassulaceae inferred from nu-
clear and chloroplast sequence data. Syst. Bot. 27: 271–288.
Mort, M.E., Soltis, P.S., Soltis, D.E., & Mabry, M.L. 2000.
Comparison of three methods for estimating internal sup-
port on phylogenetic trees. Syst. Biol. 49: 160–171.
Negrín Sosa, L. & Pérez de Paz, P.L. 1988. Consideracio-
nes acera del género Sideritis L. (Lamiaceae) en Tenerife
(Islas Canarias). Lagascalia 15: 295–318.
Obón de Castro, C. & Rivera-Nuñez, D. 1994. A Taxonomic
Revision of the Section Sideritis (Genus Sideritis) (Labia-
tae). J. Cramer, Berlin.
Olmstead, R.G. & Sweere, J.A. 1994. Combining data in
phylogenetic systematics: an empirical approach using
87
Barber & al. • Hybridization in Macaronesian SideritisTA XO N 56 (1)
•
February 2007: 74–88
three molecular data sets in the Solanaceae. Syst. Biol.
43: 467–481.
Otto, S.P. & Whitton, J. 2000. Polyploid incidence and evo-
lution. Annual Rev. Genet. 34: 401–437.
Panero, J., Francisco-Ortega, J., Jansen, R.K. & Santos-
Guerra, A. 1999. Molecular evidence for multiple origins
of woodiness and a New World biogeographic connection
of the Macaronesian Island endemic Pericallis (Astera-
ceae: Senecioneae). Proc. Natl. Acad. Sci. U.S.A. 96:
13886 –13891.
Percy, D.M. & Cronk, Q.C. 2002. Different fates of island
brooms: contrasting evolution in Adenocarpus, Genista,
and Te line (Genisteae, Fabaceae) in the Canary Islands
and Madeira. Amer. J. Bot. 89: 854–864.
Pérez de Paz, P. & Negrín Sosa, L. 1992. Revisión Tax-
onómica de Sideritis L. Subgénero Marrubiastrum (Mo-
ench) Mend.-Heuer (Endemismo Macaronésico). J. Cra-
mer, Berlin.
Pijnacker, L.P. & Schotsman, H.D. 1988. Nuclear DNA
amounts in European Callitriche species (Callitricha-
ceae). Acta Bot. Neerl. 37: 129–135.
Posada, D. & Crandall, K.A . 1998. MODELTEST: testing the
model of DNA substitution. Bioinformatics 14: 817–818.
Raskina, O., Belyayev, A. & Nevo, E. 2004a. Quantum spe-
ciation in Aegilops: molecular cytogenetic evidence from
rDNA cluster variability in natural populations. Proc.
Natl. Acad. Sci. U.S.A. 101: 14818–14823.
Raskina, O., Belyayev, A. & Nevo, E. 2004b. Activity of the
En/Spm-like transposons in meiosis as a base for chromo-
some repatterning in a small, isolated, peripheral popula-
tion of Aegilops speltoides Tausch. Chromosome Res. 12:
153–161.
Raven, P.H. 1972. Evolution and endemism in the New Zea-
land species of Epilobium. Pp. 259–274 in: Valentine,
D.H. (ed.), Taxonomy, Phytogeography and Evolution.
Academic Press, New York.
Rieseberg, L.H. 1995. The role of hybridization in evolution:
old wine in new skins. Amer. J. Bot. 82: 944–953.
Rieseberg, L.H., Raymond, O., Rosenthal, D.M., Lai, Z.,
Livingstone, K., Nakazato, T., Durphy, T.L., Schwarz-
bach, A.E., Donovan, L.A. & Lexer, C. 2003. Major
ecological transitions in wild sunflowers facilitated by
hybridization. Science 301: 1211–1216.
Rieseberg, L.H., Whitton, J. & Linder, C.R. 1996. Molecu-
lar marker incongruence in plant hybrid zones and phylo-
genetic trees. Acta Bot. Neerl. 45: 243–262.
Ronquist, F. & Huelsenbeck, J.P. 2003. MrBayes 3: Bayes-
ian phylogenetic inference under mixed models. Bioinfor-
matics 19: 1572–1574.
Sanders, R.W., Stuessy, T.F., Marticorena, C. & Silva,
O.M. 1987. Phytogeography and evolution of Dendros-
eris and Robinsonia, tree–Compositae of the Juan Fer-
nandez Islands. Opera Bot. 92: 195–215.
Sang, T., Crawford, D.J. & Stuessy, T.F. 1997. Chloroplast
DNA phylogeny, reticulate evolution, and biogeography
of Paeonia (Paeoniaceae). Amer. J. Bot. 84: 1120–1136.
Seelanan, T., Schnabel, A. & Wendel, J.F. 1997. Congruence
and consensus in the cotton tribe (Malvaceae). Syst. Bot.
22: 259–290.
Shimodaira, H. & Hasegawa, M. 1999. Multiple compari-
sons of log-likelihoods with applications to phylogenetic
inference. Molec. Biol. Evol. 16: 1114–1116.
Simmons, M.P. & Ochoterena, H. 2000. Gaps as characters
in sequence-based phylogenetic analyses. Syst. Biol. 49:
369 –381.
Simmons, M.P., Pickett, K.M. & Miya, M. 2004. How
meaningful are Bayesian support values? Molec. Biol.
Evol. 21: 188–199.
Smith, E.B. 1974. Coreopsis nuecensis (Compositae) and a
related new species from southern Texas. Brittonia 26:
161–171.
Smith, J.F., Burke, C.C. & Wagner, W.L. 1996. Interspe-
cific hybridization in natural populations of Cyrtandra
(Gesneriaceae) on the Hawaiian Islands: evidence from
RAPD markers. Pl. Syst. Evol. 200: 61–77.
Soltis, D.E., Soltis, P.S. & Milligan, B.G. 1992. Intraspecific
chloroplast DNA variation: systematic and phylogenetic
implications. Pp. 117–150 in: Soltis, P.S., Soltis, D.E. &
Doyle, J.J. (eds.), Molecular Systematics of Plants. Chap-
man & Hall, New York.
Stebbins, G.L. 1950. Variation and Evolution in Plants.
Columbia University Press, New York.
Stuessy, T.F. & Crawford, D.J. 1998. Chromosomal stasis
during speciation in angiosperms of oceanic islands. Pp.
307–324 in: Stuessy, T.F. & Ono, M. (eds.), Evolution and
Speciation of Island Plants. Cambridge University Press,
Cambridge.
Suzuki, Y., Glazko, G.V. & Nei, M. 2002. Overcredibility of
molecular phylogenies obtained by Bayesian phylogenet-
ics. Proc. Natl. Acad. Sci. U.S.A. 99: 16138–16143.
Swofford, D.L. 2002. PAUP*: Phylogenetic Analysis Using
Parsimony, vers. 4.0b10. Sinauer Associates, Sunderland.
Tem plet on, A .R. 1983. Phylogenetic inference from restric-
tion endonuclease cleavage site maps with particular ref-
erence to the evolution of humans and the apes. Evolution
37: 221–244.
Trusty, J.L., Olmstead, R.G., Santos-Guerra, A., Sá-
Fontinha, S. & Francisco-Ortega, J. 2005. Molecular
phylogenetics of the Ma caronesian- endemic genus Bystro-
pogon (Lamiaceae): palaeo-islands, ecological shifts and
interisland colonizations. Molec. Ecol. 14: 1177–1189.
Wendel, J.F. & Doyle, J.J. 1998. Phylogenetic incongruence:
window into genome history and molecular evolution. Pp.
265 –296 in: Soltis, D.E., Soltis, P.S. & Doyle, J.J. (eds.),
Molecular Systematics of Plants II: DNA Sequencing.
Kluwer Academic Publishers, Boston.
Wheeler, H.-M. 1945. A contribution to the cytology of the
Australian-South Pacific species of Nicotiana. Proc.
Natl. Acad. Sci. U.S.A. 31: 177–185.
Whitten, W.M., Williams, N.H. & Chase, M.W. 2000. Sub-
tribal and generic relationships of Maxillarieae (Orchida-
ceae) with emphasis on Stanhopeinae: combined molecu-
lar evidence. Amer. J. Bot. 87: 1842–1856.
Wilcox, T.P., Zwickl, D.J., Heath, T.J. & Hillis, D.M. 2002.
Phylogenetic relationships of the dwarf boas and a com-
parison of Bayesian and bootstrap measures of phyloge-
netic support. Molec. Phylog. Evol. 25: 361–371.
Wolf, P.G., Murray, R.A. & Sipes, S.D. 1997. Species-in-
dependent geographical structuring of chloroplast DNA
haplotypes in a montane herb Ipomopsis (Polemonia-
ceae). Molec. Ecol. 6: 283–291.
Yoder, A.D., Irwin, J.A. & Payseur, B.A. 2001. Failure of
the ILD to determine data combinability for slow loris
phylogeny. Syst. Biol. 50: 408–424.
88
TAXON 56 (1)
•
February 2007: 74–88Barber & al. • Hybridization in Macaronesian Sideritis
Appendix. Collection locations, voucher numbers, and Genbank accession numbers for Sideritis trnL intron, trnT-trnL
intergenic spacer, and ITS sequences. An asterisk identifies Genbank accessions reported in a previous study (Barber
& al., 2002). Vouchers are deposited in the following herbaria: Jardín Botánico Canario Viera y Clavijo (LPA), Las Palmas
de Gran Canaria, Canary Islands, Spain; Instituto Canario de Investigaciones Agrarias (ORT), Puerto de la Cruz, Tenerife,
Canary Islands, Spain; University of Texas (TEX), Austin, Texas, U.S.A. Abbreviations: CI, Canary Islands.
Taxon: collecti on location; voucher number (herbar ium); Genbank acc ession numbers: trnL intron; trnT-trnL intergenic spacer; ITS.
Outgroup species
Ballota hispanica Benth.: Spain, Castellón, Azuebar; ASG s.n. (ORT); AF335680*, AF335719*, AF335641*.
Stachys hirta L.: Madeira, Pta. São Lourenço; JB 233 (TEX); AF335682*, AF335721*, AF335643*.
Ingroup species
Sideritis L. subgen. Marrubiastrum (Moench) Mend.-Heuer: S. barbellata Mend.-Heuer emend. P. Pérez & L. Negrín: CI-La
Palma, Aldea de Charco; JB 262 (TEX); DQ900796, DQ900773, DQ900750. S. brevicaulis Mend.-Heuer emend. P. Pérez & L. Negrín:
CI-Tenerife, above Masca; JB 226 (TEX); DQ900797, DQ900774, DQ900751. S. canariensis L.: CI-El Hierro, Talbano; JB 257 (TEX);
AF335644*, AF335683*, AF335605. S. canariensis L.: CI-La Palma, Br eña Alta, 4 km S of tu nnel; JB 228 (TEX); DQ900798, DQ90 0775,
DQ900752. S. canariensis L.: CI-Tenerife, Carretera de Maquinal; JB 227 (TEX); DQ900799, DQ900776, DQ900753. S. candicans Ai-
ton var. candicans: Madeira, Encumeada; JB 230 (TEX); AF335645*, AF335684*, AF335606*. S. cretica L. subsp. cretica: CI- Tenerife,
Bco. de Talavera; JB 218 (TEX); DQ900800, DQ900777, DQ900754. S. cretica L. subsp. spicata (Pitard) L. Negrín & P. Pérez: CI-La
Gomera, carretera San Sebastian & Hermigua; JB 213 (TEX); DQ900801, DQ900778, DQ900755. S. cystosiphon Svent.: CI-Tenerife,
Tam ai mo; JB 222 (TEX); DQ900802, DQ900779, DQ900756. S. dasygnaphala (Webb & Berth.) Clos emend. Svent.: CI-Gran Canaria,
Pico de las Nieves; ASG 18691 (LPA); DQ900803, DQ900780, DQ900757. S. dendrochahorra Bolle: CI-Tenerife, between Bailadero
& Taganana; JB 224 (TEX); DQ900804, DQ900781, DQ900758. S. discolor Bolle: CI-Gran Canaria, Bco. Los Tilos; ASG 8424 (ORT);
DQ900805, DQ900782, DQ900759. S. eriocephala Marr. ex Negr. & Pérez: CI-Tenerife, Cañadas del Teide; JB 266 (TEX); AF335646*,
AF335685*, AF335607*. S. ferrensis P. Pérez & L. Negrín: CI-El Hierro, Mocanal; JB 204 (TEX); DQ900806, DQ900783, DQ900760.
S. gomerae Bolle subsp. gomerae: CI-La Gomera, San Sebastian; JB 256 (TEX); AF335647*, AF335686*, AF335608*. S. gomerae
Bolle subsp. perezii L. Negrín: CI-La Gomera, road to Benchijigua; JB 200 (TEX); DQ900807, DQ900784, DQ900761. S. infernalis
Bolle emend. Svent.: CI-Tenerife; Bco. del Infierno; JB 225 (TEX); DQ900808, DQ900785, DQ900762. S. kuegleriana Bornm.: CI-
Tenerife, Bco. de Talavera; ASG 16767 (ORT); DQ900809, DQ900786, DQ900763. S. lotsyi (Pitard) Bornm.: CI-La Gomera, carr. Ga-
rajonay 2 km E of National Park; JB 206 (TEX); DQ900810, DQ900787, DQ900764. S. macrostachys Poir.: CI-Tenerife, Anaga; JB 254
(TEX); AF335648*, AF335687*, AF335609*. S. marmorea Bolle: CI-La Gomera, Riscos de Puntallana; ASG 32073 (ORT); DQ900811,
DQ900788, DQ900765. S. nervosa (Christ) Lindinger: CI-Tenerife, Pta. de Teno; JB 214 (TEX); DQ900812, DQ900789, DQ900766. S.
nutans Svent.: CI-La Gomera, El Santo; JB 201 (TEX); DQ900813, DQ900790, DQ900767. S. oroteneriffae L. Negrín & P. Pérez var.
arayae L. Negrín & P. Pérez: CI-Tenerife, pine forest above Candelaria; ASG 33.953 (ORT); DQ900814, DQ900791, DQ900768. S. oro-
teneriffae L. Negrín & P. Pérez var. oroteneriffae: CI-Tenerife, Montaña de Ayosa; JB 207 (TEX); DQ900815, DQ900792, DQ900769.
S. pumila (Christ) Mend.-Heuer: CI-Fuerteventu ra, Pico de la Zarza, ca. 700 m; ASG 22.226 (ORT); DQ900816, DQ900793, DQ900770.
S. pumila (Christ) Mend.-Heuer: CI-Lanzarote, Peñas del Chache; JB 278 (TEX); AF335649*, AF335688*, AF335610*. S. soluta Clos
subsp. gueimaris L. Negrín & P. Pérez: CI-Tenerife, Bco. de Badajoz; JB 221 (TEX); DQ900817, DQ900794, DQ900771. S. soluta Clos
subsp. soluta: CI-Tenerife, Topo de la Grieta; JB 217 (TEX); AF335650*, AF335689*, AF335611*. S. sventenii (Kunkel) Mend.-Heuer:
CI-Gran Canaria, Ayagaures; JB 195 (TEX); DQ900818, DQ900795, DQ900772.
Sideritis L. subgen. Sideritis sect. Hesiodia (Moench) Bentham: S. cossoniana Ball: Morocco, Antiatlas; ASG s.n. (ORT);
AF335652*, AF335691*, AF335613*. S. montana L.: seed source unidentified; native to Mediterranean region, Balkan Peninsula; JB
212 (TEX); AF335651*, AF335690*, AF35612*. S. romana L.: Italy, San Bartolomeo al Mare; JB 209 (TEX); AF335653*, AF335692*,
AF335614*.
Sideritis L. subg. Sideritis sect. Empedoclea (Rafin.) Bentham: S. athoa Papanic. & Kokkini: Greece, Mt. Athos; JB 292 (TEX);
AF335654*, AF335693*, AF335615*. S. clandestina (Bory & Chaub.) Hayek: Greece, Tayhete, Peloponnese; JB 286 (TEX); AF335655*,
AF335694*, AF335616*. S. euboaea Heldr.: Greece, Euboae; JB 290 (TEX); AF335656*, AF335695*, AF335617*. S. perfoliata L.: seed
source unidentified; native to Greece, Turkey, Cyprus & Syria; JB 294 (TEX); AF335657*, AF335696*, AF335618*. S. scardica Griseb.:
Greece, Thessalien, Nomos Larisis; JB 211 (TEX); AF335658*, AF335697*, AF335619*. S. syriaca L.: Greece, Mt. Dirfys; JB 210
(TEX); AF335659*, AF335698*, AF335620*. S. taurica Stephan ex. Willd.: seed source unidentified; native to Turkey; JB 289 (TEX);
AF335661*, AF335700*, AF335622*.
Sideritis L. subgen. Sideritis sect. Sideritis: S. algarviensis Rivera & Obón: Portugal, Cabo S. Vicente & Sagres; JB 288 (TEX);
AF335662*, AF335701*, AF335623*. S. antiatlantica (Maire) Rejdali: Morocco, Antiatlas Mountains; ASG s.n. (ORT); AF335663*,
AF335702*, AF335624*. S. chamaedryfolia Cav.: Spain, Alicante, Vellena Peña Rubia; JB 275 (TEX); AF335664*, AF335703*,
AF335625*. S. dianica Rivera , Obón, de la Torre & Ba rber: Spain, Alic ante, Pego, Alt de Pascual; JB 280 (TEX); AF335665*, AF335704*,
AF335626*. S. endressii Willk. subsp. emporitana (Cadev) Rivera & Obón: Portugal, Cabo Espichel; JB 282 (TEX); AF335666*,
AF335705*, AF335627*. S. glacialis Boiss.: seed source unidentified; native to Spain; JB 272 (TEX); AF335668*, AF335707*,
AF335629*. S. glauca Cav.: Spain, Alicante, Sierra de Orihuela; JB 248 (TEX); AF335669*, AF335708*, AF335630*. S. hirsuta L.:
Spain, Teruel, Puerto de Ragudo; JB 287 (TEX); AF335670*, AF335709*, AF335631*. S. hyssopifolia L.: Spain, Picos de Asturien;
JB 202 (TEX); AF335672*, AF335711*, AF335633*. S. incana L.: Spain, Valencia, Enguera; JB 246 (TEX); AF335673*, AF335712*,
AF335634*. S. marmorinensis Obón & Rivera: Spain, Murcia, Atamaria; JB 250 (TEX); AF335674*, AF335713*, AF335635*. S. mur-
getana Obón & Rivera: Spain, Murcia, Altorreal; JB 249 (TEX); AF335675*, AF335714*, AF335636*. S. pungens Benth. subsp. java-
lambrensis Reverchon: Spain, Teruel, Pico Javalambre; JB 244 (TEX); AF335676*, AF335715*, AF335637*. S. sericea Pers.: Spain,
Valencia, Enguera; JB 293 (TEX); AF335677*, AF335716*, AF335638*. S. tragoriganum Lagasca: Spain, Castellón, Soneja & Azuebar;
JB 241 (TEX); AF335678*, AF335717*, AF335639*.