Content uploaded by John C. Semple
Author content
All content in this area was uploaded by John C. Semple on Aug 26, 2015
Content may be subject to copyright.
Chapter 4
A review of chromosome numbers
in Asteraceae with hypotheses on
chromosomal base number evolution
John C. Semple and Kuniaki Watanabe
INTRODUCTION
In spite of the great variability in the chromosome
numbers … a pattern can be seen when the evidence
is fully reviewed. — obinson et al. 1981, p. 8
Asteraceae are the largest family of owering plants and
have long been of cytological interest. The rst chromo-
some counts for members of the family were published
more than a century ago ( Juel 1900; Land 1900; Merrell
1900). The total number of chromosome number reports
has increased dramatically with major e orts to determine
chromosome numbers of large numbers of composites
being conducted in the 1960 –1980 period (e.g., aven et
al. 1960; Anderson et al. 1974). Prior to DNA sequence-
based phylogenetic analyses, hypotheses on chromosomal
base numbers in Asteraceae were hampered by a lack of
understanding of which genera were basal within tribes
and which tribes were basal within the family. For exam-
ple, Cronquist (1981) reported that Asteraceae had a range
of base numbers from x = 2 to x = 19+ and suggested
that perhaps x = 9 was ancestral. Earlier, Solbrig (1977)
had also concluded x = 9 was the ancestral base number
of the family based on an analysis of habit and frequency
of chromosome numbers. Bremer (1994) merely noted
that chromosome number data were conveniently sum-
marized in Solbrig (1977) and subsequent indices. In
more recent years following the introduction of molecu-
lar techniques for analyzing phylogenies through DNA
restriction fragment length polymorphisms and base pair
sequence analyses, authors have compared molecular
results with chromosomal basal number data in order
to reach conclusions on ancestral base numbers within
groups of genera and among tribes (e.g., Baldwin et al.
2002; Ito et al. 2000; Chapter 37). Accessing data in all
of the tens of thousands of publications reporting chro-
mosome numbers in Asteraceae has not been convenient
until very recently, when much of the information was
put online in Watanabe’s (2008) Index to Chromosome
Numbers in Asteraceae (http://www.lib.kobe-u.ac.jp/prod
ucts/asteraceae/index.html). This paper presents analyses
of chromosome numbers in the online database in light
of recent understanding of the phylogeny of Asteraceae
(e.g., Funk et al. 2005). The rst objective of the study
was to compile a summary database of every genus in
the family. The second objective was to determine the
chromosomal base number for every genus in the fam-
ily for which data were available. The third objective
was to plot chromosome counts and basal chromosome
numbers of every taxon onto the supertree (= metatree)
phylogeny (Funk et al. 2005). The fourth object was to
formulate hypotheses on patterns of chromosomal base
number evolution in the family having “fully reviewed”
the evidence.
Semple and Watanabe62
MATERIALS AND METHODS
Two datasets were used as primary sources of informa-
tion on chromosome numbers. The most critical of these
were the data available online at http://www.lib.kobe-u
.ac.jp/products/asteraceae/index.html, Watanabe’s Index
to Chro mo some Numbers of Asteraceae (2008). The database
has been updated multiples times as data from addi-
tional publications are added to the matrix. Our analysis
is based on entries in the database as of March 2007.
This included records on more than 38,000 chromosome
number reports at the time of our analysis listed by author
and taxon; and 41,000 as of September 2007. Each search
of counts in a genus included a summary of the number
of reports, but not a calculated total of the actual numbers
of counts included in the records. When a publication
reported multiple counts for a single taxon, these were
not listed separately. However, the information could be
tallied from data presented in the search results. Also,
searches for some genera include the names of taxa for
which no chromosome counts have been reported in the
literature. Thus, the number of “records” listed at the
top of a search report needed to be recalculated to yield
the actual number of individual chromosome number
reports for a genus included in the summary data matrix
created for this paper. The Watanabe dataset included
data extracted from 4521 publications at the time of our
analysis, and 4800 as of September 2007. Details on these
can be accessed from the search reports generated by the
web site and are not listed here. Searches of the database
for this chapter were facilitated by working o -line di-
rectly from the Microsoft Excel™ data le (*.xls) created
by Watanabe. Generic nomenclature in the Watanabe
database follows Bremer (1994) with post March 2007
changes to re ect treatments of taxa in Flora of North
America (Flora of North America Editorial Committee
2006).
The second data source for chromosome counts was
the rst author’s research database les on chromosome
number reports for asters, goldenrods and miscellaneous
other genera of Astereae (primarily taxa of the subtribe
Chrysopsidinae Nesom and miscellaneous other North
American Astereae). The Semple datasets collectively
included information on 10,835 individual counts at
the time of our analysis and are based on individual
voucher data. These were compiled for research on cyto-
geographic and taxonomic studies. Nomenclature in the
Semple databases follows that of generic treatments of
the Asteraceae in Flora of North America (vols. 19–21) with
a data eld indicating the name under which the count
was originally published. These datasets are not avail-
able online. Each of the data les (*.ask) was created in
askSam™ v.5.1.2.367 (Seaside Software Inc. dba askSam
Systems, Perry, Florida).
A database summarizing information by genus on
chromosome number data was constructed using ask-
Sam working from the Watanabe and Semple datasets
plus information in generic treatments in Flora of North
America (vols. 19–21). Eighteen data elds were included
for each genus document. These are listed in Table 4.1
and included data on nomenclature and numbers of spe-
cies, number of chromosome number reports (actual or
estimated total number of counts), a list of chromosome
numbers reported in the literature, ancestral and derived
base numbers determined in this study, other cytological
data (ploidy levels, aneuploidy, dysploidy and supernu-
merary chromosomes), geographic distribution informa-
tion, and a eld for miscellaneous observations, e.g., al-
ternative sources of data, etc.
The completed generic summary database was searched
to generate reports on a number of di erent cytological
and taxonomic questions. Lists of genera with cytologi-
cal data were generated for each tribe, subtribe or clade.
Separate lists of all genera sorted alphabetically, by chro-
mosomal base number, and by numbers of reports were
also generated. eports were saved as *.ask les, which
can be exported as *.txt, *.rtf, *.html and several other
le formats.
RESULTS AND DISCUSSION
Data on 1587 genera of Asteraceae and 15 genera of
Calyceraceae and Goodeniaceae were included in the ge-
neric summary database. The results of analyses of num-
bers of counts and reports by genus are summarized by
tribe/clade in Table 4.2 listed in the order of branching of
clades on the supertree phylogeny of Funk et al. (2005).
Included in Table 4.2 by tribe/clade are the numbers of
genera included in this study compared with the number
of genera reported for each tribe/clade in Bremer (1994),
the percent of genera with at least one chromosome num-
ber report, an estimate of the number of species, and an
estimate of the number of count reports.
An est imated 58,320 chromosome nu mber reports were
summarized; 58,124 of these reports were for Asteraceae.
The actual number of chromosome number count deter-
minations made on individuals is not known because this
information was not always included in a publication. In
some cases, a report was based on a chromosome count
from a single individual of a taxon. In other cases, several
hundred to more than a thousand counts were reported
for a single taxon in one publication (e.g., Semple 1989).
Papers reporting very large numbers of counts ( > 100)
for a taxon were usually cytogeographic studies. At least
one chromosome count has been reported for a taxon in
978 genera of Asteraceae (61.6%); no data were available
for 611 genera in the family. Thirteen genera were found
Chapter 4: A review of chromosome numbers in Asteraceae 63
to have more than 1000 chromosome number reports
in total ( Table 4.3). The majority of genera have ten or
fewer chromosome number reports (598 genera; 37.7%);
203 genera (12.8%) have been sampled only once.
The number of species of composites is estimated to
be 22,472 (= total of all estimated numbers of species by
genus) in 1587 genera. Cronquist (1981) estimated there
to be about 1000 genera and 20,000 species in Asteraceae.
Bremer (1994) recognized 1535 genera and raised the
number of species to around 23,000. Bremer (1994) was
the primary database on most genera included in our
analysis, and thus it is not surprising that our numbers of
genera and species are similar to those in Bremer (1994).
The larger number of genera of Astereae reported here is
the direct result of numerous DNA sequence studies pub-
lished in recent years and incorporated into the generic
limits followed in Flora of North America.
Some other errors were also undoubtedly introduced
into the summary of genera in the database due to nomen-
clatural problems and data entry errors. The International
Plant Names Index (http://www.ipni.org/index.html) was
frequently consulted in the creation of the summary data
matrix in order to reduce the number of errors due to
nomenclature. Checking synonymy sometimes revealed
a case of double counting of a species and including its
cytological data under two genera. The case with asters
is informative and indicative of possible sources of error
at the time our analysis was rst completed and involves
the two authors of this paper. In the Watanabe database
in March 2007, counts for asters were generally listed
in the genus Aster L. under which the majority of the
counts were originally reported. The online index con-
tained 1753 records for Aster s.l. and one record for the
North America aster genus Symphyotrichum Nees. The
Semple database had 4578 reports for Symphyotrichum and
only about 100 reports for the Eurasian genus Aster s.str.
(clearly re ecting a geographic bias in data entry to date).
However, when the numbers of reports were tallied for
the Eurasian species included in the Watanabe database,
it contained 2128 reports for 27 Eurasia species of Aster
s.str. The number of reports for North American species
of Symphyotrichum was larger in the Semple database than
the Watanabe database because the former included more
than 600 unpublished counts to be reported in un nished
cytogeographic studies. The conclusion to draw from the
asters case is simple: anyone searching a database on chro-
mosome numbers must pay attention to the generic con-
cepts followed in entering the data. Many of these kinds
of potential errors were sorted out using the synonymy
in generic treatments in Flora of North America. We are
pleased to note that changes to the nomenclature of asters
in the Watanabe database post March 2007 were made
so that reports of counts for North American species
of asters are listed when searching Doellingeria, Eurybia,
Table 4.1. Data fi elds included in the summary database on genera.
PHYL [ Number for phylogenetic ordering of clades/tribes in reports 0–37
TRIBE [
CLADE [ Any major but informal subtribal groupings
SUBTRIBE [
GENUS [
AUTH [ Authority(-ies) of generic name
SPP [ Number of sp eci es
REPS [ Number of published reports in on-line Index Chromo Asteraceae
X= [ Base number (not always obvious)
X2= [ Derived base numbers (not always obvious)
2n= [ All sporophytic numbers, meiotic and mitotic
POLY [ Yes/no poly pl oid y present
PLEVELS [ 2x, 4x, 6x, etc.
DYSP [ Yes /no dysploidy present; base number shift up or down
ANEU [ Aneuploid numbers reported (interpretation of Index data)
SUPERS [ Yes /no supern um er ari es (fragm ent s, B’s, etc.)
LOC [ General information on distribution; continent, country; state or province for North American taxa
OBS [ Notes on cytology, classifi cation, problems to check; some synonyms
64 Semple and Watanabe
Table 4.2 . Summary of numbers of genera with and without chromosome data by tribe /clade.
No. Tribe
Included
in this
study
No. of
genera in
Bremer
(1994)
% of total
genera
includeda
Genera
with
counts
Genera
without
counts
% genera
included
with counts
Estimated
number
of species
Estimated
number of
count reports
Basal Grade
1 Barnadesieae 9 9 100% 6 3 67% 92 28
2Stifftia clade 3 2 150% 1 2 33% 18 1
3 Mutisieae 55 58 95% 27 28 49% 685 238
4 Gochnatieae 3 3 100% 2 1 67% 77 2
5Hecastocleis clade 1 1 100% 1 1 100% 1 1
Carduoideae
6 Dicomeae 7 7 100% 3 4 43% 103 5
7 Oldenburgieae 1 1 100% 1 0 100% 4 3
8 Tarchonantheae 2 2 100% 2 0 100% 17 3
9 Cardueae 83 83 100% 53 30 64% 2,557 4,093
10 Pertyeae 4 4 100% 2 2 50% 69 58
11 Gymnarrheneae 1 1 100% 1 1 100% 1 2
Cichorioideae
12 Gundelieae 2 2 100% 2 0 100% 3 9
13 Cichorieae 100 98 102% 80 20 80% 1,850 11,635
14 Arctotideae 17 17 100% 8 9 47% 209 66
15 Liabeae 14 14 100% 12 2 86% 159 88
16 Vernonieaeb105 98 107% 42 63 40% 897 1001
17 unassigned 3 3 100% 1 2 33% 30 3
Asteroideae
18 Senecioneae 120 120 100% 65 55 54% 3,196 2,784
19 Calenduleae 8 8 100% 6 2 75% 112 194
20 Gnaphalieae 181 162 112% 95 86 52% 2,014 1,419
21 Astereae 215 170 126% 140 75 65% 2,638 20,052
22 Anthemideae 110 109 101% 69 41 63% 1,732 4,598
23 Inuleae 67 67 100% 35 32 52% 716 729
24 Athroismeae 3 3 100% 2 1 67% 27 2
Helenieae–Helianthoid clade
25 Helenieae 13 13 100% 12 1 92% 117 441
26 Coreopsideae 24 20 120% 16 8 67% 420 980
27 Neurolaeneae 1 2 50% 1 0 100% 13 11
28 Tageteae 33 32 103% 23 10 70% 265 598
29 Chaenactideae 3 3 100% 3 0 100% 20 101
30 Bahieae 18 18 100% 17 1 94% 73 240
31 Polymnieae 2 2 100% 2 0 100% 9 67
32 Heliantheae 132 108 122% 95 37 72% 1,350 3,010
Chapter 4: A review of chromosome numbers in Asteraceae 65
Symphyotrichum etc., rather than collectively under Aster
s.l. The asters case demonstrates the advantage of on-
line databases that can be updated and modi ed often,
which is not the situation with printed databases or static
online databases. The e ort needed to keep a database
such as Watanabe’s Index up-to-date is large and time-
consuming. For genera in other parts of the world, we
have less con dence in decisions made while creating the
summary of genera database. For nomenclature-related
problems in genera within the same branch of the super-
tree of Funk et al. (2005), errors in assigning species and
their chromosome counts to the correct genus have little
or no signi cance to the tribal and family level conclu-
sions presented below.
In Table 4.2, a wide range in the percent of genera
for which at least one chromosome count has been re-
ported among tribes /clades is presented. At least one
chromosome number has been reported for all 36 pri-
mary clades in the family. All genera have been sampled
in a number of the smaller tribes, e.g., Gundelieae and
Polymnieae. For tribes with more than ten genera, the
range of those sampled was 40% –100%. For the seven
tribes with more than 100 genera, the average number
of genera sampled was 60.75%. Vernonieae were the least
well sampled with chromosome counts reported for only
40% of the genera using data in the Watanabe Index that
was updated late in this study with the assistance of Dr.
Harold obinson. However, prior to assigning counts
originally published under the generic name Vernonia
to the many genera that have been segregated from it,
only 23% of the genera in Vernonieae had at least one
chromosome number reported. The average number of
genera sampled for the six other large tribes was 64.2%,
which is slightly more than for the entire family; the six
tribes Senecioneae, Gnaphalieae, Astereae, Anthemideae,
Heliantheae and Eupatorieae include about 57% of the
genera in the family.
A very large range in chromosome numbers and chro-
mosomal base numbers occurs in Asteraceae. More than
180 di erent mitotic counts have been reported: 2n = 4,
4+1–3, 5, 6, 6+1–2Bs, 7, 8, 8+1–6Bs, 9, 10, 10+1–2B, 11,
12, 12+1, 12+1–4Bs, 13, 14, 14+1–2, 15, 16, 17, 18, 18+1,
Table 4.3. Thirteen genera with more than 1000
chromosome count reports.
No. of
reports Genus Tribe
4578 Symphyotrichum Astereae
4549 Solidago Astereae
4017 Taraxacum Cichorieae
2129 Aster Astereae
2128 Crepis Cichorieae
1905 Eupatorium Eupatorieae
1884 Brachyscome Astereae
1709 Hieracium Cichorieae
1605 Senecio Senecioneae
1600 Xanthisma Astereae
1489 Centaurea Cardueae
1400 Artemisia Anthemideae
1158 Erigeron Astereae
Table 4.2 . Summary of numbers of genera with and without chromosome data by tribe /clade.
No. Tribe
Included
in this
study
No. of
genera in
Bremer
(1994)
% of total
genera
includeda
Genera
with
counts
Genera
without
counts
% genera
included
with counts
Estimated
number
of species
Estimated
number of
count reports
33 Millerieae 34 38 89% 25 9 74% 358 737
34 Madieae 36 36 100% 36 0 100% 200 1,445
35 Perityleae 5 5 100% 4 1 80% 76 177
36 Eupatorieae 168 170 99% 89 79 53% 2,350 3,316
Incertae sedis
Galeana 1 1 100% 0 1 0% 3 0
Villanova clade 2 2 100% 1 1 50% 10 2
Welwitschiella 1 1 100% 0 1 0% 1 0
Totals 1,587 1,493 978 611 61.6% 22,472 58,136
a The total number of genera is based on Bremer (1994) or the tribal description in Flora of North America (2006).
b Number of genera counted and percentages based on data provided by Dr. Harold Robinson to update Watanabe database.
Continued.
Semple and Watanabe66
18+1–4B, 18+2, 19, 20, 20+1–5, 20+1–6B, 21, 22, 22+1–
3, 24, 24+1, 24+1B, 24+5–9, 25, 26, 27, 28, 29, 30, 30+2,
30+2B, 31, 32, 33, 34, 34+1frag, 35, 36, 36+1, 36+1–2,
37, 38, 39, 40, 40–45, 40–47, 40+2Bs, 42, 42–44, 44, 45,
45–50, 46, 47+3, 48, 48+1, 48+3Bs, 50, 50–52, 51, 51–52,
52, 53, 54, 54+1–5supers, 55, 56, 56–58, 57, 58, 58–59,
60, 61, 62, 63, 64, 65, 66, 67, 68, 69–72, 70, 72, 72–74,
76, 78, 80, 81, 84, 85, 86, 87, 88, 90, 92, 95, 96, 98, 100,
108, 110, 112, 114, 120, 122, 130, 138, 140, 146, 154, 160,
160+, 171, 176, 180, 184, 198, ca. 228, ca. 288, ca. 324,
ca. 432. The most frequent number in the database was
2n = 18 because it is the most frequent number reported
in Astereae, which has the largest number of counts re-
ported, and in several other larger tribes. Two species in
the tribe Astereae have the very low sporophytic number
of 2n = 4, Brachyscome dichromosomatica C.. Carter and
Xanthisma gracile (Nutt.) D.. Morgan & .L. Hartman.
The highest number reported is 2n = ca. 432 (48x ; x = 9)
for Olearia albida Hook. f. (Beuzenberg and Hair 1984),
also in the tribe Astereae.
More than 170 di erent meiotic counts have been re-
ported: 2n = 2II, 3II, 3II+1– 8B, 4II, 4II+2sup, 4II+1–3IBs,
4II+1–2 IIBs, 5II, 5II+1, 5II+1–4Bs, 5II+10 I, 6II, 6II+1–2 Bs,
7II+2I, 8II, 8II+1I, 8II+1–2B, 9II, 9II+1–2I, 9II+2Bs, 9II–12II,
10II, 10II+1, 10 II+1–2Bs, 10 II+10I, 11II, 11II+1I, 11II+8I, 12II,
12II+3Bs, 13II, 13–14II, 13II+1I, 9II+9I, 14II, 15II, 15II–16II,
15II–17II, 15II+1B, 15II+1f rag, 16II, 16II+1I, 16II+1I, 16II–
18II, 17II, 17II–24II, 17II+1–4fr a g s, 17II+1I, 17II+1II, 17II+5I,
17II+6B, 18II, 18II–20II, 18II–27II, 18II–20II, 18II+1f rag,
18II+1I, 19II, 19II+1f rag, 19II+1I, 19II+2–3Bs, 19II+2–3frag,
19II+4–7B, 20II, 20II+1I, 20II+1f rag, 21II, 21II+1I, 22II,
23II, 24II, 24II–27II, 24II–30II+8–20I, 25II, 25II+1–6f ra g,
25II+Bs, 26II, 26II+1–3Bs, 27II, 27II+6I, 27–28II, 27II–30II,
28II, 28II+2I, 28II–29II, 29II, 29II+1I, 30II, 30II+2I, 32II+1II,
32II–34II, 33II, 33II–34II, 34II, 34II–36II, 34II+2I, 36II, 36II–
38II, 38II, 39II, 40II, 41II, 42II–44II, 43II, 44II, 44–45II, 45II,
46II, 47–48II, 47II+3I, 48II, 50II, 50II+1–8supers, 51II, 52II,
54II, 59II, 59–60II, 60II, 64II, 68II, 68II–69II, 70II, 72II,
80II, 86II, 89–96II, 100II, 102II–108II, ca. 108II, ca. 110II,
ca. 131II. The range in meiotic counts is the same as for
mitotic counts with the exception of only reaching about
30x to 32x. Meiotic irregularities and precocious divi-
sions of some bivalents account for many of the reported
numbers. These make determining with certainty the
chromosome number of polyploid individuals more dif-
cult. In our experience, interpreting meiosis is generally
more di cult than counting mitotic chromosomes, and
this is particularly true because such a large number of
composites are of polyploid origin. However, growing
live plants for root tip squashes to obtain mitotic counts is
often not possible.
Two categories of chromosomal base numbers occur
in Asteraceae. First, x numbers include the ancestral base
numbers (plesiomorphies) and the base numbers derived
from these via dysploidy. Dysploidy is the change in
the chromosomal base number through a rearrangement
of chromatin and loss or gain of a centromere without
necessarily changing the amount of chromatin in the
karyotype. In Asteraceae, dysploidy decreases are com-
mon to very common in some clades, while increases
appear to be rare or very rare depending upon how the
higher x numbers are interpreted. Base numbers of x = 2,
3, 4, 5, 6, 7, 8, 9, 10, and 11 occur in the family. Some
of these numbers may be the result of a dysploid increase,
others are undoubtedly the result of a series of dysploid
decreases. There are many, many cases of dysploid series
from higher to lower base numbers in Asteraceae, and
these are found in nearly all of the 36 main branches of
the phylogeny. Dysploidy is unknown in a few of the
branches due to a lack of data (no counts or very few
counts). Even some of the smaller branches with few taxa
have some dysploidy. Dysploidy occurs in 102 genera
with x base numbers and in 112 genera with derived x
base numbers. In total, dysploidy occurs in 214 genera,
21.9% of the 978 genera with counts reported.
Numerous secondarily derived base numbers (x) are
also common in the family. These evolved in several
di erent ways. Derived base numbers can result from
allo polyploid combinations of x numbers. For example,
the x = 9 base number in Chrysopsis (Astereae) is derived
from hybridizing x = 4 and x = 5 parental taxa and sub-
sequent chromosome number doubling and diploidization
(Semple and Chinnappa 1980). Alternatively, derived base
numbers can result from autopolyploidy and subsequent
diploidization of the karyotype resulting in a x that is a
multiple of the ancestral x number of the clade. Nearly the
entire Olearia II clade in Astereae appears to be based on
a diploidized 12x ploidy level (Cross et al. 2002; Chapter
37). Dysploid decreases also occur in clades with derived
x. The following derived base numbers occur in the fam-
ily: x = 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4,
and 3. Dysploidy has also occurred in polyploids of these
derived numbers resulting in much larger x numbers and
dysploid numbers derived from them.
Polyploidy is common in Asteraceae and occurs in
most major clades. In total, polyploidy occurs in 570 gen-
era, 58.3% of the 978 genera with counts reported; this
includes all genera of the major Helenioid – Helianthoid
clade. Polyploidy occurs in 247 genera without x base
numbers, 25.3% of the 978 genera with counts re-
ported. Polyploidy is common in the most basal branch
of Asteraceae, subfam. Barnadesioideae, in which only
Schlechtendalia is known to occur at a presumed diploid
level with a dysploid derived base number. The follow-
ing ploidy levels occur in Asteraceae: 2x, 3x, 4x, 5x, 6x,
7x, 8x, 9x, 10x, 12x, 14x, 15x, 16x, 18x, 20x, 22x, 24x,
32x, 36x, and 48x. Frequencies of ploidy levels are sum-
marized in Table 4.4. Diploids are most frequent and
Chapter 4: A review of chromosome numbers in Asteraceae 67
Table 4.4. Frequencies of ploidy levels in Compositae.
Ploidy level
Number of genera % of 978 genera with counts
Genus TribeOnly level
With other
levels Only level
With other
levels
Base numbers (x)
2x270 440 27.6% 45.0%
3x0 34 0.0% 3.5%
4x25 193 2.6% 19.7%
5x0 16 0.0% 1.6%
6x20 93 2.0% 9.3%
7x0 6 0.0% 0.6%
8x1 41 0.1% 4.2% Paragynoxys Senecioneae
9x0 5 0.0% 0.5%
10x3 25 0.3% 2.6%
12x1 24 0.1% 2.5% Pachystegia Astereae
14x0 5 0.0% 0.5%
12x + 14x1 4 0.1% 0.4% Soliva Anthemideae
15x0 1 0.0% 0.1% Werneria Senecioneae
16x0 3 0.0% 0.3% Antennaria Gnaphalieae
Raoulia Gnaphalieae
Werneria Senecioneae
18x0 2 0.0% 0.2% Tetradymia Senecioneae
Antennaria Gnaphalieae
20x0 2 0.0% 0.2% Antennaria Gnaphalieae
Werneria Senecioneae
22x0 1 0.0% 0.1% Olearia II Astereae
32x0 1 0.0% 0.1% Olearia II Astereae
36x0 1 0.0% 0.1% Olearia II Astereae
48x0 1 0.0% 0.1% Olearia II Astereae
Derived base numbers (x²)
2x289 300 29.6% 30.7%
3x0 27 0.0% 2.8% Amauriopsis Bahieae (apomict)
4x8 129 0.8% 13.2%
5x0 10 0.0% 1.0%
6x1 41 0.1% 4.2% Erechtites Senecioneae
7x0 2 0.0% 0.2%
8x0 21 0.0% 2.1%
9x0 0 0.0% 0.0%
10x0 7 0.1% 0.7%
12x0 9 0.0% 0.9%
16x0 2 0.0% 0.1% Chromolaena Eupatorieae
Leptinella Anthemideae
20x0 1 0.0% 0.1% Leptinella Anthemideae
24x0 1 0.0% 0.1% Leptinella Anthemideae
Semple and Watanabe68
were the only ploidy level occurring in 270 genera with x
base numbers and in 289 genera with diploidized x base
numbers, 27.6% and 29.6% of the 978 genera with counts
reported, respectively. Higher ploidy levels occur with
decreasing frequency as the ploidy level increases; 7.4%
(x) and 2.9% (x) of the 978 genera with counts reported
include ploidy levels of 8x to 10x; 3.9% (x), and 1.3% (x)
of the 978 genera with counts reported include ploidy
levels of 12x and higher levels.
Supernumerary chromosomes of various kinds have
been reported in 143 genera of Asteraceae. Of these, 45
genera have been reported to have B chromosomes. It
was not determined by us whether or not these reports
were for true B chromosomes ( Jones and ees 1982)
such as found and well studied in Xanthisma gracile and
X. texanum DC. or were merely supernumerary chromo-
somes mislabeled as B chromosomes. Some reports may
even have been errors in reporting the distal portion of
the satellite chromosome as a supernumerary when the
satellite was well separated from the proximal portion of
the chromosome. For example, the large distal portion of
the satellite chromosomes in Eurybia and Symphyotrichum
(both Astereae) could easily be mistaken for separate small
supernumerary chromosomes ( J.C. Semple, pers. obs.) or
separate autosomal chromosomes ( Watanabe et al. 2007).
The frequency of aneuploidy was also analyzed. True
aneuploidy is the gain or loss of single chromosomes
without changing the base chromosome number. Due
to the large number of ambiguous chromosome counts
published as “circa” reports or as errors in reports based
on sectioning techniques, it was unclear if ranges in
numbers about a base number or multiple of the base
number in polyploids were indications of aneuploidy oc-
curring in a taxon or if these ranges were counting er-
rors. Therefore, no reliable frequency of aneuploidy can
be reported here.
Chromosomal base number evolution in Asteraceae
Speculating on base chromosome numbers o ers,
perhaps, the nest of all vehicles for intellectual
auto-stimulation. — e-mail from J.L. Strother
to J.C. Semple, 16 June 2006
Ancestral base numbers for each of the 36 main branches
of the supertree phylogeny (Funk et al. 2005) were deter-
mined, as were the base numbers for Goodeniaceae and
Calyceraceae. The latter two families have a base number
of x = 9 with lower base numbers of x = 8 and x = 7 de-
rived by downward dysploidy. In the rbcL DNA phylog-
eny of Asterales (Gusta son et al. 1996), the basal grade
in Goodeniaceae included Anthotium . Br., Dampiera
. Br., Lechenaultia . Br. and Brunonia Smith. All four
genera have base numbers of x = 9 (Peacock 1963). In
Goodeniaceae, Goodenia and Coopernookia with x = 8 or 7
were in a derived position in the family (Gusta son et al.
1996). A phylogeny of Asteraceae with ancestral chromo-
somal base numbers superimposed is shown in Fig. 4.1.
An ancestral base number of x = 9 is hypothesized for
Barnadesieae with x = 8 being derived by downward dys-
ploidy. The genera Arnaldoa, Chuquiraga and Dasyphyllum
have polyploid chromosome numbers with x = 27 based
on counts in Watanabe’s online index and Watanabe et al.
(2007). Doniophyton has reported numbers of 2n = 24II,
48, and 25II suggesting base numbers of x = 25 and
x = 24 derived from an ancestral x = 27 via downward
dysploidy. The hexaploid ploidy level would have reduced
the rate of evolution allowing these genera to retain ple-
siomorphic traits for the family. Chromosome numbers
reported for Barnadesia (2n = 12II, 14II, 25II, 50–52, 52, 54,
62, ca. 50II) suggest more karyotype evolution has taken
place in the genus than other related genera or some of
the counts are inaccurate. Schlechtendalia has a base num-
ber of x = 8, which is likely derived by dysploid decrease.
However, it is di cult to infer dysploid reduction from
2n = 54 to 2n = 18 or 16 at a bound. Thus it is possible
that the x = 8 base number for Schlechtendalia has been
derived from the ancestral x = 9 by dysploid reduction
and 2n = 54 for Barnadesia and Dasyphyllum is a hexaploid
state based on the original base chromosome number,
x = 9. Stuessy et al. (1996) considered Schlechtendalia to
be primitive within Barnadesioideae, but this is not sup-
ported by the derived position of the genus on the super-
tree (Funk et al. 2005); the phylogram in Stuessy et al.
(1996) shows little similarity to the generic arrangement
in the supertree. A basal position for Schlechtendalia based
on new molecular sequence data, however, is an alterna-
tive that still cannot be refuted (see Chapter 13).
The chromosome number/habit situation in Barna-
desieae looks similar to primitive angiosperm families
with the high base chromosome numbers and with the
woody habits in the woodland or forests (= the closed
plant community). These high chromosome numbers,
the woody habits (tree, shrub or liana) and their habitats
in the closed plant communities are linked very closely.
In contrast, the herbaceous members, Acicarpha spathulata
(Calyceraceae), Schlechtendalia luzulaefolia and Hecastocleis
shockleyi (Asteraceae) have the lower chromosome num-
ber of 2n = 16 and their habitats are open plant commu-
nities such as the maritime coastal sand-dune (Acicarpha
spathulata) and arid semi-desert (Hecastocleis shockleyi ).
They have very specialized morphology such as succulent
(Acicarpha spathulata) or spiny (Hecastocleis shockleyi ) le aves,
and seeds embedded within the receptacle (Acicarpha
spathulata). In sister families of Asteraceae, members of
Goodeniaceae and Calyceraceae are herbs and have the
low chromosome base number x = 9 and occur in open
plant communities. In more basal Asteraceae, the n = 8
Chapter 4: A review of chromosome numbers in Asteraceae 69
for Schlechtendalia luzulaefolia and Hecastocleis shockleyi is
a derived number. These specialized habitat taxa have a
more restricted recombination system and more imme-
diate tness instead of genetic exibility. Such a genetic
system appeared to be causally connected with the de-
pendence on ample seed production as the only means of
propagation in short lived plants and with rapid popula-
tion establishment in labile and brie y available habitats
(Grant 1958; Stebbins 1958; Ehrendorfer 1970).
An x = 9 ancestral base number is hypothesized for
the next three branches on the supertree (Fig. 4.1).
The chromosome counts reported for the Sti tia clade
are 2n = 54 (Gibbs and Ingram 1982; Watanabe et al.
2007). This also is hypothesized to be a paleopolyploid
with x = 27 derived from an x = 9 ancestor. Additional
counts for this clade are needed. Genera in Mutisieae
include chromosome counts indicating that both poly-
ploidy and dysploidy have occurred multiple times. Base
numbers of x or x = 8, 9, 10, 11, 12, 14, 15, 22, 23, 24,
25, 26, 27, 36 are indicated by the many counts for the
tribe. We hypothesize that multiple downward dysploid
events from polyploids based on x = 9 account for all,
or nearly all, of the base numbers listed. These are the
result of long dysploid series from polyploids of x = 27.
There was a reduction in chromosome number from
x = 27 (e.g., Acourtia) to x = 14 and 11 (Chaetanthera)
with a change in growth form from shrub to herb habit
within Mutisieae. In Gochnatieae, four counts have been
reported; 2n = 54 for Cyclolepis and 2n = 54, 2n = ca.
23II and 2n = 44 for Gochnatia. A paleop oly ploid ba se
of x = 27 is hypothesized, again being derived from an
x = 9 ancestor. Therefore, the basal grade of tribes native
to South America all are hypothesized to have an ances-
tral chromosomal base number of x = 9. Alternatively,
the Sti tia clade, Mutisieae and Gochnatieae could have
an ancestral base number of x = 27, with all other num-
bers in the three clades derived from this presumably
diploidized hexaploid number. Such a possibility would
then necessitate a long, and undocumented, dysploidy se-
ries from x = 27 to x = 11, 10, 9 and 8 in the next series
of tribes/clades on the supertree. We remind the reader
of Strother’s comment on base numbers at this point in
the discussion. If Barnadesieae were also hypothesized to
be x = 27 and this is assumed to be basal for Asteraceae,
Fig. 4.1. Chromosomal base number evolution in Asteraceae. Hypothesized base numbers are superimposed on the summary
tree of the supertree (= metatree) phylogeny presented by Funk et al. (2005).
e
aecaine
dooG
e
aecarecy
laC
eaei
sedanra
B
itSf aitf edalc
eaes
ituM
ea
eitanhco
G
eaei
rohciC
ea
eretsA
ea
editcan
eahC
eae
ditotcrA
eaedimehtnA
ea
eihaB
eaebaiL
eaelunI
eaein
myloP
Vea
einonre
eae
msiorhtA
.rts
.s eaeh
tnaileH
muibmyr
oC
.rt
s.s
eaeineleH
e
aeirelliM
ea
enoicen
eS
eae
dispoero
C
eaeidaM
e
aeludnel
aC
eae
nealorue
N
e
aelytireP
e
aeilahpa
nG
Teaet
eg
a
ea
eirotapu
E
sie
lcotsace
H
ea
emociD
aigrubn
edlO
Teaeht
nanohcr
a
e
aeudra
C
ea
eytreP
a
nehrran
myG
e
aeilednu
G
Carduoideae Cichorioideae s.str. Asteroideae
South America
Sub-Saharan Africa
Northern Africa & Mediterranean
Asia
Eurasia & Europe
North America & Mexico
Australia & New Zealand
Widespread or unresolved
x= 9
x= 10
x2= 19
917 19 17
8
8
999999
9
?
7
10 9
11
10
10 10 10
10
10
10
[10]
10
9
14
13
10
10
10 10
19
19
19
18
11 15
19
19
19
9
27 9
27
9
27
9
27
Semple and Watanabe70
then all chromosome numbers in the family would be x
numbers. This does not seem likely at this time.
The next branch on the supertree includes just
Hecastocleis with one report of 2n = 16. A base number
of x = 8 is indicated for this North American genus. We
hypothesize that it is derived by downward dysploidy
from x = 9. Additional chromosome counts are needed
to test this hypothesis.
Funk et al. (2005) noted that next nine major branches
on the supertree were likely African in origin. The red
lines of the South American grade were replaced by blue,
lavender and green lines on their phylogeny. Shifts from
x = 9 to x = 10 and 11 are hypothesized to have occurred
accompanying the shift in geography. Ancestral base
numbers of x = 10 or 11 (Dicomeae), x = 10 (Cardueae,
Gymnarrhena, Vernonieae, Senecioneae, Calenduleae,
Gna ph al ieae, Anthemideae, Inuleae [ including Pluche-
eae ] and Athroismeae) and x = 9 (Oldenburgia, Tarchon-
antheae, Gundelieae, Cichorieae, Arctotideae, Liabeae,
and Astereae) are hypothesized based on the known
chromosome numbers of basal members of these clades.
Therefore, the base number of Carduoideae is x = 10.
The base number of Cichorioideae could be either x = 10
or x = 9. We hypothesize that it was ancestrally x = 10.
The four core Asteroideae tribes also are likely to have
been ancestrally x = 10. In Calenduleae, Nordenstam
(1994) concluded a base number of x = 10 appeared
likely, and we agree that this is most parsimonious with
x = 8 and x = 7 derived by downward dysploidy. In
Anthemideae and Gnaphalieae, decreases from x = 10
early in their histories to x = 9 and 7, respectively, are
hypothesized. Watanabe et al. (1999) noted the di culty
in determining the ancestral base number in Gnaphalieae
due to a lack of chromosome counts for African taxa.
The few counts available for members of the subtribe
elhaniinae suggest base numbers of x = 9, 8, and 7,
but the majority of genera have not yet been sampled
cytologically even once. Counts with x = 10 have been
reported in a few genera of Anthemideae. Only Astereae
shifted to x = 9 via downward dysploidy before diversi-
fying. However, the two most basal genera in Astereae,
Denekia and Printzia, are unknown cytologically. Should
either of these be found to have x = 10 as a base number,
then Astereae also would be ancestrally x = 10.
Based on the preponderance of clearly downward dys-
ploid events in Asteraceae, it seems likely that upward
dysploid events are much more di cult to successfully
complete. An increase in base number could result from
trisomic aneuploidy of a single chromosome homologue
that does not produce a lethal increase in gene product
from the three copies of each gene. Aneuploidy would
readily provide the new centromere needed for the in-
crease in base number, and chromosome rearrangements
and loss or suppression of critical genes could result in
a stabilized new higher base number. Such an evolu-
tionary process involves more di cult steps than simply
rearranging existing chromatin on fewer centromeres to
achieve a dysploid decrease. This di erence in likeli-
hood would account for the rarity of dysploid increases
in the family and the relative commonness of dysploid
decreases. Therefore, we have hypothesized very few an-
cestral dysploid increases in favor of many long dysploid
series with gaps in base numbers from high to low due to
extinctions or lack of discovery.
The two other tribes in this middle portion of the
supertree are hypothesized to have derived base numbers.
Pertyeae have chromosome numbers indicating possibly
derived base numbers of x = 14, and 13. We hypothesize
that these are not derived by serial upward dysploidy
from base number of x = 10 or 9 or 8, but rather they
are derived by a series of downward dysploid events
from a polyploid ancestor with n = 20, 18 or 16. The
same series of events is a documented pattern in the
Helenioid – Helianthoid clade and also appears to have
occurred in Vernonieae with x = 17 being derived from
x = 10, 9 and 7 ancestors. This appears to be the “easier”
evolutionary process than multiple dysploid increases to
reach x = 14. Corymbium forms the other mid tree clade
with a derived base number, but in this case x = 8 is in-
dicated by the single count of 2n = 16. Two downward
dysploid events from an x = 10 ancestor are hypothesized
in this branch of the supertree.
Numerous and sometimes well documented downward
dysploid series have occurred in Cichorieae, Astereae and
Gnaphalieae. Some of these cases are classical studies in
cytotaxonomy and need not be discussed further here,
e.g., Crepis and Brachyscome. Polyploidy is also frequent
in these tribes resulting in them being some of the more
intensively studied tribes cytologically over many years.
Funk et al. (2005) noted a second major geographic
shift in the location of composite evolution indicated on
their supertree diagram by a shift from blue, lavender and
green lines to yellow lines for North American origins.
This is the large terminal Helenioid – Helianthoid clade
of the phylogeny. Baldwin et al. (2002) discussed this
portion of the tree in detail noting the high derived an-
cestral base numbers for all the tribes in the clade; they
hypothesized that x = 18 was ancestral with multiple
upward dysploidy events to yield x = 19. Decades earlier,
Smith (1975) and later obinson et al. (1981) hypothesized
x = 17–19 as basal for Heliantheae s.l., with obinson et
al. presenting arguments suggesting x = 19 being derived
via aneuploidy from 2n = 4x = 20. During this same pre-
DNA sequence time period, base numbers of x = 8 or 9
(Stuessy 1977) and x = 8 –12 (Solbrig et al. 1972) were
suggested for Heliantheae s.l. We hypothesize that x = 19
is ancestral for the entire clade (Fig. 4.1) because we be-
lieve multiple dysploid increases are much less likely than
Chapter 4: A review of chromosome numbers in Asteraceae 71
Literature cited
Anderson, L.C., Kyhos, D.W., Mosquin, T., Powell, A.M.
& Raven, P.H. 1974. Chromosome numbers in Compositae.
IX. Haplopappus and other Astereae. American Journal of Botany
61: 665 – 671.
Baldwin, B.G., Wessa, B.L. & Panero, J.L. 2002. Nuclear
rDNA evidence for major lineages of helenioid Heliantheae
(Compositae). Systematic Botany 27: 161–198.
Beuzenberg, E.J. & Hair, J.B. 1984. Contributions to a chro-
mosome atlas of the New Zealand ora – 27: Compositae.
New Zealand Journal of Botany 22: 353–356.
Bremer, K. 1994. Asteraceae: Cladistics & Classi cation. Timber
Press, Portland.
Cronquist, A. 1981. An Integrated System of Classi cation of
Flowering Plants. Columbia University Press, New York.
Cross, E.W., Quinn, C.J. & Wagsta , S.J. 2002. Molecular
evidence for the polyphyly of Olearia (Astereae: Asteraceae).
Plant Systematics and Evolution 235: 99–120.
Ehrendorfer, F. 1970. Evolutionary patterns and strategies in
seed plants. Taxon 19: 185–195.
Flora of North America Editorial Committee (eds.). 2006.
Flora of North America, vols. 19–21, Asteraceae, parts 1–3. Ox-
ford University Press, New York.
Funk, V.A., Bayer, R.J., Keeley, S., Chan, R., Watson, L.,
Gemein holzer, B., Schilling, E., Panero, J.L., Baldwin,
B.G., Garcia-Jacas, N., Susanna, A. & Jansen, R.K.
2005. Ever ywhere but Antarctica: using a supertree to un-
derstand the diversity and distribution of the Compositae.
Biologiske Skrifter 55: 343–374.
Gibbs, P. E. & Ingram, R. 1982. Chromosome numbers of
some Brazilian owering plants. Notes from the Royal Botanic
Garden, Edinburgh 40: 399–407.
Grant, V. 1958. The regulation of recombination in plants. Cold
Spring Harbor Symposia on Quantitative Biology 23: 337–363.
Gustafsson, M.H.G., Backlund, A. & Bremer, B. 1996.
Phyl ogeny of the Asterales sensu lato based on rbcL sequences
with pa rticular reference to the Goodeniaceae. Plant Systematics
and Evolution 199: 217–242.
Ito, M., Yahara, T., King, R.M., Watanabe, K., Oshita, S.,
Yokoyama, J. & Crawford, D.J. 2000. Molecular phylog-
eny of Eupatorieae (Asteraceae) estimated from cpDNA FLP
and its implication for the polyploid origin hypothesis of the
tribe. Journal of Plant Research 113: 91–96.
Jones, R.N. & Rees, H. 1982. B Chromosomes. Academic Press,
London.
Juel, H.O. 1900. Vergleichende Untersuchungen über typi-
sche und parthenogenetische Fortp anzung bei der Gattung
Antennaria. Kongliga Svenska Vetenskapsakademiens Handlingar
33: 1–59.
Land, W.J. 1900. Double fertilization in Compositae. Botanical
Gazette 30: 252–260.
numerous dysploid decreases. In our database, x = 19
taxa occur in Helenieae, Tageteae, Bahieae, Heliantheae,
Millerieae, Madieae, and Perityleae. Dysploid derived
numbers of x = 18, 17, 15, and 11 appear to be an-
cestral in Coreopsideae, Eupatorieae, Polymnieae, and
Neurolaeneae, respectively. The ancestral base number
in Chaenactideae may be x = 9, with x = 17 being sec-
ondarily derived, but this is not certain (Mooring 1965;
Baldwin et al. 2002).
There are several ways that x = 19 could have evolved,
and we do not favor one possibility over another due to a
lack of conclusive evidence. In each scenario, downward
dysploidy would have been involved, but the timing dif-
fers. Athroismeae are sister to the entire clade, and this
has a base number of x = 10. The x = 19 base number
could be an allopolyploid derived from x = 10 and x = 9
parental taxa. The dysploid event would have occurred
before the polyploid event. Alternatively, a tetraploid
taxon with n = 20 and x = 10 could have undergone
subsequent downward dysploidy and diploidization to
yield x = 19. There is evidence that both scenarios have
occurred in other tribes, although the base numbers in-
volved are di erent. obinson et al. (1981) noted a ten-
dency for higher base numbers to stabilize at numbers
less than double the original lower base number. This
observation lends support to x = 19 evolving with poly-
ploidy rst (2n = 4x = 20) and subsequent disploidy and
karyotype diploidization.
Within the Helenioid–Helianthoid clade, short to long
downwa rd dysploid series occur in many di erent genera.
In most of these, gaps occur in the series, either due to
a lack of data or extinction of intermediate base number
taxa. Many taxa remain to be sampled cytologically and
some of the gaps may be lled in with new data. Other
series will likely remain incomplete.
The evolution of base number s in Eupatorieae is stri king
because x = 17 is likely both ancestral and derived (Ito et
al. 2000; this study did not include any x = 18 taxa). Like
most other tribes in the Helenioid–Helianthoid clade,
a derived high ancestral base number is plesiomorphic,
in this case x = 19 –18. Many subtribes within the tribe
have a base number of x = 10 derived by downward dys-
ploid series. In some subtribes, secondarily derived base
numbers of x = 20, 19, 18, 17, 16 and 14 occur as a result
of polyploidy on x = 10 and subsequent downward dys-
ploidy for a second time in the history of those phylads.
Of note is the unusual x = 25 base number in one of two
groups of Neomirandea (Watanabe et al. 1995). We choose
not to speculate on the origin of this remote, high chro-
mosome number pending further data. Without DNA
sequence phylogenies, it is unlikely that such redundant
patterns of karyotype evolution would have been fully
resolved. The combination of morphological, cytological
and molecular studies together reveal the details in the
history of Eupatorieae and the family as a whole.
Semple and Watanabe72
Merrell, W.D. 1900. A contribution to the life history of
Silphium. Botanical Gazette 29: 99–133.
Mooring, J.S. 1965. Chromosome studies in Chaenactis and
Chamaechaenactis (Compositae, Helenieae). Brittonia 17: 17–25.
Nordenstam, B. 1994. Tribe Calenduleae. Pp. 365–376 in:
Bremer, K., Asteraceae: Cladistics & Classi cation. Timber Press,
Portland.
Peacock, W.J. 1963. Chromosome numbers and cytoevolution
in the Goodeniaceae. Proceedings of the Linnean Society of New
South Wales 88: 8–27.
Raven, P., Solbrig, O., Kyhos, D. & Snow, R. 1960. Chro-
mo some numbers in Compositae. I. Astereae. American Journal
of Botany 47: 124–132.
Robinson, H., Powell, A.M., King, R.M. & Weedin, J.F.
1981. Chromosome num bers in Compositae, X II: Helianthe ae.
Smithsonian Contributions to Botany 52: 1–28.
Semple, J.C. 1989. Geographical distribution of B chromo-
somes of Xanthisma texanum (Compositae: Astereae). II. Local
variation within and between populations and frequency vari-
ations through time. American Journal of Botany 76: 769–776.
Semple, J.C. & Chinnappa, C.C. 1980. Phylogenetic im-
plications of meiosis in wild and cultivated interspeci c hy-
brids in Chrysopsis (Compositae-Astereae): C. godfreyi (n = 5),
C. gossypina × ssp. cruiseana (n = 9) and C. godfreyi (n = 5) ×
linearifolia (n = 5). Canadian Journal of Botany 58: 172–181.
Smith, E.B. 1975. Chromosome numbers of North American
Coreopsis with phyletic interpretations. Botanical Gazette 136:
78–86.
Solbrig, O.T. 1977. Chromosomal cytology and evolution in
the family Compositae. Pp. 267–281 in: Heywood, V.H.,
Harborne, J.B. & Turner, B.L. (eds.), The Biology and Chemistry
of the Compositae, vol. 1. Academic Press, London.
Solbrig, O.T., Kyhos, D.W., Powell, M. & Raven, P.H. 1972 .
Chromosome numbers in Compositae. VIII. Heliantheae.
American Journal of Botany 59: 869–878.
Stebbins, G.L. 1958. Longevity, habitat, and release of genetic
variability in the higher plants. Cold Spring Harbor Symposia on
Quantitative Biology 23: 365–378.
Stuessy, T.F. 1977. Heliantheae—systematic review. Pp. 621–
671 in: Heywood, V.H., Harborne, J.B. & Turner, B.L. (eds.),
The Biology and Chemistry of the Compositae, vol. 2. Academic
Press, London.
Stuessy, T.F., Sang, T. & DeVore, M.L. 1996. Phylogeny
and biogeography of the subfamily Barnadesioideae with im-
plications for early evolution of the Compositae. Pp. 463–
490 in: Hind, D.J.N. & Beentje, H.J. (eds.), Proceedings of
the International Compositae Conference, Kew, 1994, vol. 1,
Compositae: Systematics. oyal Botanic Gardens, Kew.
Watanabe, K. 2008. Index to Chromosome Numbers in Asteraceae.
http://www.lib.kobe-u.ac.jp/products/asteraceae/index.html.
Watanabe, K., King, R.M., Yahara, T., Ito, M., Yokoyama,
J., Suzuki, T. & Crawford, D.J. 1995. Chromosomal cy-
tology and evolution in Eupatorieae (Asteraceae). Annals of the
Missouri Botanical Garden 82: 581–592.
Watanabe, K., Short, P.S., Denda, T., Konishi, N., Ito, M.
& Kosuge, K. 1999. Chromosome numbers and karyotypes
in the Australian Gnaphalieae and Plucheeae (Asteraceae).
Australian Systematic Botany 12: 781–802.
Watanabe, K., Yahara, T., Hashimoto, G., Nagatani, Y.,
Soejima, A., Kawahara, T. & Nahazawa, M. 2007.
Chro mo some numbers and karyotypes in Asteraceae. Annals
of the Missouri Botanical Garden 94: 643–654.