American Journal of Botany 87(1): 96–107. 2000.
PHYLOGENETIC STRUCTURE IN THE GRASS FAMILY
(POACEAE): EVIDENCE FROM THE NUCLEAR GENE
PHYTOCHROME B1
SARAH MATHEWS,2,3 ROCKY C. TSAI,2,4
AND
ELIZABETH A. KELLOGG5
2
Department of Organismic and Evolutionary Biology, Harvard University Herbaria, 22 Divinity Avenue,
Cambridge, Massachusetts 02138 USA; and
5 Department of Biology, University of Missouri-St. Louis, 8001 Natural Bridge Road, St. Louis, Missouri 63121 USA
Phylogenetic analyses of partial phytochrome B (PHYB) nuclear DNA sequences provide unambiguous resolution of
evolutionary relationships within Poaceae. Analysis of PHYB nucleotides from 51 taxa representing seven traditionally
recognized subfamilies clearly distinguishes three early-diverging herbaceous ‘‘bambusoid’’ lineages. First and most basal
are Anomochloa and Streptochaeta, second is Pharus, and third is Puelia. The remaining grasses occur in two principal,
highly supported clades. The first comprises bambusoid, oryzoid, and pooid genera (the BOP clade); the second comprises
panicoid, arundinoid, chloridoid, and centothecoid genera (the PACC clade). The PHYB phylogeny is the first nuclear gene
tree to address comprehensively phylogenetic relationships among grasses. It corroborates several inferences made from
chloroplast gene trees, including the PACC clade, and the basal position of the herbaceous bamboos Anomochloa, Streptochaeta, and Pharus. However, the clear resolution of the sister group relationship among bambusoids, oryzoids, and pooids
in the PHYB tree is novel; the relationship is only weakly supported in ndhF trees and is nonexistent in rbcL and plastid
restriction site trees. Nuclear PHYB data support Anomochlooideae, Pharoideae, Pooideae sensu lato, Oryzoideae, Panicoideae, and Chloridoideae, and concur in the polyphyly of both Arundinoideae and Bambusoideae.
Key words:
gene family; grasses; nuclear gene; phylogeny; phytochrome B; Poaceae.
The grasses (Poaceae) have received abundant scientific attention, both phylogenetic and otherwise. The historical prominence of the grasses as an object of botanical
research reflects their almost ubiquitous biogeographical
presence and their pervasive economic importance since
the very beginnings of human civilization. Approximately one-third of the world’s dry land is covered by some
species of Poaceae, and the majority of the world’s human population relies heavily, if not predominantly, on
cereal grasses such as rice, corn, or wheat for its daily
sustenance.
Most taxonomic treatments of the Poaceae recognize
six or seven major subfamilies of grasses, Bambusoideae,
Oryzoideae, Pooideae, Panicoideae, Arundinoideae,
Chloridoideae, and Centothecoideae, which are further
subdivided into 40 tribes consisting of 600–750 genera
(Campbell, 1985; Dahlgren, Clifford, and Yeo, 1985;
Clayton and Renvoize, 1986; Renvoize and Clayton,
1992; Watson and Dallwitz, 1992). The scientific effort
to distinguish properly, and ultimately to classify into natural groups, the 10 000 species of the grass family began
with systematic attempts to group species according to
the morphology of characteristic grass structures such as
the spikelet (Brown, 1810, 1814). Eventually, this classical morphological approach to grass systematics was
supplemented and complemented by data consisting of
novel characters derived from detailed studies of anatomy
(Prat, 1936; Reeder, 1957; Row and Reeder, 1957;
Brown, 1958; Tateoka, Inoue, and Kawano, 1959), physiology (Hattersley and Watson, 1992), cytology (Avdulov, 1931), and most recently, DNA sequences.
Recent phylogenetic studies of the grass family have
led to a family tree that provides a consistent resolution
of genealogical relationships and that can ultimately
serve as the basis for a taxonomic system that accurately
reflects patterns of evolution within the family. Most conclusions regarding grass phylogeny, however, are based
on data from the chloroplast genome (rbcL sequence
data: Doebley et al., 1990; Barker, Linder, and Harley,
1995; Duvall and Morton, 1996; Barker, 1997; ndhF sequence data: Clark, Zhang, and Wendel, 1995; Catalán,
Kellogg, and Olmstead, 1997; chloroplast restriction site
variation: Davis and Soreng, 1993; Soreng and Davis,
1998; rpoC2 sequence data: Cummings, King, and Kellogg, 1994; rps4 sequence data: Nadot, Bajon, and Lejeune, 1994). A comparably comprehensive nuclear data
set from the grasses is lacking.
Only four published studies incorporate data from nuclear genes. The pioneering study of Hamby and Zimmer
(1988) included nuclear ribosomal (nr) RNA sequences
from nine species representing four subfamilies. Their
nrRNA phylogeny indicated early the promise of molecular data, but it included too few taxa for a test of plastid
gene trees. Hsaio and colleagues sequenced the internal
transcribed spacers (ITS) of nrDNA, focusing on Pooideae (Hsaio et al., 1995) and on Arundinoideae (Hsaio
et al., 1998). They found support for generic relation-
1 Manuscript received 2 June 1998; revision accepted 4 May 1999.
The authors thank Matt Lavin, Lynn Clark, Jerry Davis, Weiping
Zhang, Barbara Ertter, Roberta Mason-Gamer, Ben Zaitchik and Elizabeth Farnsworth for DNA and plant material, Elisa Freeman for technical assistance, David Swofford for the use of PAUP*4.0, Lynn Clark
and an anonymous reviewer for comments on the manuscript, and acknowledge financial support from NSF grants BIR-9508467 to SM and
DEB-9419748 to EAK.
3 Author for correspondence (email: [email protected]).
4 Current address: 5505 HLS Holmes Mail Ctr., Cambridge, Massachusetts 02138 USA.
96
January 2000]
MATHEWS
ET AL.—PHYB PHYLOGENY IN GRASSES
ships, but support for relationships among tribes was
minimal. Moreover, sequences from arundinoids were
difficult to align (Hsaio et al., 1998), suggesting that ITS
may have limited utility in broader studies in grasses.
Mathews and Sharrock (1996) sampled short fragments
(;300 bp) from three phytochrome loci and inferred a
phylogeny from the combined data. The phytochrome
phylogeny included a taxonomic sample more comparable to chloroplast gene trees. It was strikingly well resolved and corroborated certain results from analyses of
chloroplast data. Perhaps the most notable result was the
resolution of a well-supported BOP clade consisting of
bambusoid, oryzoid, and pooid genera, a clade that was
only poorly supported by ndhF data (Clark, Zhang, and
Wendel, 1995) and that was not resolved by the other
plastid data sets (Doebley et al., 1990; Davis and Soreng,
1993; Cummings, King, and Kellogg, 1994; Nadot, Bajon, and Lejeune, 1994; Barker, Linder, and Harley, 1995;
Duvall and Morton, 1996; Soreng and Davis, 1998). Despite the apparent promise of the phytochrome study, it
included none of the taxa relevant to early evolution in
the family, and only a sparse taxonomic sample of its
derived clades. Additionally, the completeness of the data
set depended on sampling each of three different phytochrome loci from each taxon.
In this paper, we describe results from our study of a
single phytochrome paralog, phytochrome B (PHYB), to
generate a nuclear data set comprehensive enough to test
plastid phylogenies. We sampled partial PHYB sequences
(1.2 kb of exon I) from seven major subfamilies of Poaceae, including taxa that potentially bear on early and
secondary diversification within the family. We also included several taxa characterized by equivocal placement
in chloroplast phylogenies.
Low-copy nuclear genes remain underutilized in phylogenetic studies, despite the need for nuclear tests of
plastid phylogenies (Doyle, 1992, 1997). The nuclear sequences commonly sampled in phylogenetic studies are
from the high-copy ribosomal loci. Low-copy nuclear
genes often are avoided as a source of phylogenetic characters because identification of strictly orthologous sequences may be confounded by the presence of multiple
related loci in a genome. This concern is valid; paralogy
may be mistaken for orthology if related sequences within a genome are evolving in concert (Zimmer et al., 1980;
Sanderson and Doyle, 1992) or if we fail to sample all
members of a gene family, even if they are evolving independently (but see below). Nevertheless, multigene
families have provided phylogenetically useful characters. Examples include genes for alcohol dehydrogenase
(e.g., Gaut and Clegg, 1991; Sang, Donoghue, and
Zhang, 1997; Small et al., 1998), phosphoglucose isomerase (Gottlieb and Ford, 1996), phytochromes (Mathews
and Sharrock, 1996; Mathews, Tsai, and Kellogg, 1997;
Lavin et al., 1998; Mathews and Donoghue, 1999), granule-bound starch synthase (Mason-Gamer, Weil, and Kellogg, 1998; R. Evans, University of Maine, personal communication), arginine decarboxylase (Galloway, Malmberg, and Price, 1998), and vicilin (Whitlock and Baum,
1999).
A simple way to sample putatively orthologous loci in
the phytochrome gene family is to use locus-specific amplification primers. Three of the four major phytochrome
97
gene lineages occurring in angiosperms predate angiosperm origins (Mathews and Sharrock, 1997). They are
divergent in function and in sequence (Sharrock and
Quail, 1989; Clack, Mathews, and Sharrock, 1994; Quail,
1994) and can be distinguished by amplification primers
(Mathews, Tsai, and Kellogg, 1997; Mathews, unpublished data). Thus, they can be sampled as single-copy
sequences in taxonomic lineages in which there is no
evidence of recent gene duplication events. This may be
the case in most grasses, where Southern blot analyses
and surveys using the polymerase chain reaction (PCR)
have provided evidence that PHYA, PHYB, and PHYC
occur singly in genomes of grasses (Mathews and Sharrock, 1996). The known exception is the genome of
maize, in which two PHYB have been mapped; however,
a single PHYB has been mapped and sequenced in sorghum (Childs et al., 1997).
MATERIALS AND METHODS
Sampling strategy—Fifty-one species of the grass family, representing 29 tribes and seven traditionally recognized subfamilies, were included in the phytochrome B (PHYB) data set. The sampling distribution of taxa is given in Table 1, which is arranged according to the
classification of Clayton and Renvoize (1986). The present data set
comprises five short PHYB sequences analyzed in Mathews and Sharrock (1996) and 46 newly generated sequences.
DNA cloning and sequencing—A 1.2-kilobase (kb) region of exon
I of the PHYB gene (Fig. 1a) encompassing the phytochrome chromophore attachment site was amplified from total DNA of 40 newly sampled species of Poaceae (sources of DNA given in Table 1). The five
PHYB sequences previously analyzed in Mathews and Sharrock (1996)
each comprised ;300 base pairs (bp). All newly sampled sequences
were amplified using a stepdown polymerase chain reaction (PCR) protocol (Hecker and Roux, 1996) beginning at 708 or 738C. PHYB-specific
primer pairs that were used in the amplification of the novel sequences
are listed in Fig. 1b. We used a degenerate primer pair to obtain clones
from Aristida, Danthonia, Lolium, Puelia, Streptogyna, and Thysanolaena. The forward degenerate primer sequence is 5-ACIGGITAYGAYMGIGTIATG-3; the reverse degenerate sequence is 5-GTYTCDATSARYCKAACCATYTC-3. It amplifies a comparable-sized target, but the forward member of this pair is 153 nucleotides downstream
of the forward member of the PHYB-specific pair; these nucleotide sites
are thus coded as missing in the final data set.
Putative PHYB PCR products were excised after electrophoresis on
a 1.2% agarose preparation gel and extracted using the QIAquikTM gel
extraction protocol (QIAGEN Inc., Valencia, California, USA). DNA
inserts were ligated to pGEMt-T or pGEMt-T Easy vectors (Promega,
Madison, Wisconsin, USA) during incubation overnight at 48C. XL1Blue Epicurian Colit subcloning-grade competent cells (Stratagene, La
Jolla, California, USA) were transformed with ligation products and
incubated overnight at 378C. White colonies were cultured overnight in
terrific broth, followed by isolation of plasmid DNA via alkaline lysis/
polyethylene glycol minipreparation or QIAprepy Spin Miniprep Kits
(QIAGEN Inc., Valencia, California, USA). Two plasmid DNAs per
taxon were screened by restriction enzyme digestion using PstI and
SphI, or EcoRI. ABIPRISMy DyeDeoxy terminator cycle sequencing
of positive clones was performed in standard extension/termination mixes (Perkin-Elmer, Norwalk, Connecticut, USA) using the universal and
internal sequence-specific primer pairs shown in Fig. 1b.
Phylogenetic analysis—A data matrix of 51 taxa by 1182 nucleotides
was obtained by manually adding accumulating sequences to a core
alignment from the Multiple Alignment Construction and Analysis
98
AMERICAN JOURNAL
OF
BOTANY
[Vol. 87
TABLE 1. Taxa included in the phytochrome B (PHYB) data set. The subfamilial and tribal divisions follow the classification system of Clayton
and Renvoize (1986), except for Anomochlooideae and Pharoideae (Clark and Judziewicz, 1996). BZ 5 Ben Zaitchik, EF 5 Elizabeth
Farnsworth, EAK 5 Elizabeth A. Kellogg, HPL 5 H. Peter Linder, JD 5 Jerrold Davis, LC 5 Lynn Clark, PP 5 Paul Peterson, RG 5 Roberta
Mason-Gamer, RS 5 Robert Soreng, SD 5 Soejatmi Dransfield, SM 5 Sarah Mathews, WZ 5 Weiping Zhang, XL 5 Ximena Londoño, ML
5 Matt Lavin, BBG 5 Berlin Botanic Garden, PI 5 U.S.D.A. Plant Introduction Station (Pullman, Washington). Taxa with asterisks are
represented by sequences (306 bp) previously analyzed in Mathews and Sharrock (1996).
SUBFAMILY/Tribe/Species
ANOMOCHLOOIDEAE
Anomochloeae
Anomochloa marantoidea Brongn.
Streptochaeteae
Streptochaeta angustifolia Soderstr.
PHAROIDEAE
Phareae
Pharus lappulaceus Aubl.
BAMBUSOIDEAE
Parianeae
Pariana radiciflora Aubl.
Eremitis Doell sp. nov.
Streptogyneae
Streptogyna americana C. E. Hubbard
Puelieae
Puelia ciliata Franchet
Bambuseae
Bambusa (Hackel) Ekman sp.*
Chusquea oxylepis Kunth
Pseudosasa japonica (Sieb. & Zucc. ex Steud) Makino ex Nakai
Olyreae
Olyra latifolia L.
Buergersiochloa bambusoidea Pilger
Lithachne pauciflora Beauv.
Ehrharteae
Ehrharta erecta Lam.
Diarrheneae
Diarrhena obovata (Gleason) Brandenburg
Oryzeae
Oryza sativa L.
Zizania aquatica L.
POOIDEAE
Bromeae
Bromus inermis Leyss.*
Triticeae
Triticum aestivum L.
Elymus L. sp.
Peridictyon sanctum O. Seberg, S. Frederiksen, & C. Baden
Brachypodium pinnatum Beauv.
Poeae
Poa pratensis L.*
Lolium perenne L.
Sesleria insularis Sommier
Aveneae
Ammophila arenaria (L.) Link
Phalaris arundinacea L.
Stipeae
Nassella viridula (Trin.) Barkworth
Meliceae
Glyceria grandis S. Watson
Melica cupanii Guss.
Lygeae
Lygeum sparteum L.
Nardeae
Nardus stricta L.
CENTOTHECOIDEAE
Centotheceae
Chasmanthium latifolium Link
Voucher
GenBank sequence
accession numbera
LC 1299
GBAN-AF137291
LC 1304
GBAN-AF137328
LC 1329
GBAN-AF137321
LC & WZ 1344
LC & WZ 1343
GBAN-AF137317
GBAN-AF137304
Johnston 433
GBAN-AF137329
Kobayashi et al. 1541
GBAN-AF137324
EAK V6
LC 1069
WZ 8400708
GBAN-U61188
GBAN-AF137298
GBAN-AF137323
XL & LC 911
SD 1382
LC 1297
GBAN-AF137315
GBAN-AF137295
GBAN-AF137307
EAK V44
GBAN-AF137302
LC & WZ 1216
GBAN-AF137301
BZ s.n.
GBAN-X57563
GBAN-AF137333
ML s.n.
GBAN-U61193
RG s.n.
EAK s.n.
EAK H5575
PI-440176
GBAN-AF137331
GBAN-AF137303
GBAN-AF137319
GBAN-AF137294
ML s.n.
SM 402
PI-253719
GBAN-U61214
GBAN-AF137308
GBAN-AF137325
RS 3389
RS 3427
GBAN-AF137289
GBAN-AF137320
ML s.n.
GBAN-AF137314
JD & RS s.n.
PI-383702
GBAN-AF137305
GBAN-AF137310
RS 3698
GBAN-AF137309
BBG: Royl and Schiers s.n. 1988
GBAN-AF137313
EAK V13
GBAN-AF137297
January 2000]
MATHEWS
ET AL.—PHYB PHYLOGENY IN GRASSES
99
TABLE 1. Continued.
SUBFAMILY/Tribe/Species
ARUNDINOIDEAE
Arundineae
Danthonia spicata Roem. and Schult.
Phragmites australis Trin. ex Steud.
Hakonechloa macra Honda
Molinia caerulea (L.) Moench
Anisopogon avenaceus R. Br.
Aristideae
Aristida longiseta Steud.
Thysanolaeneae
Thysanolaena maxima (Roxb.) Kuntze
CHLORIDOIDEAE
Eragrostideae
Eragrostis cilianensis (All.) Mosher*
Cynodonteae
Bouteloua gracilis (Willd. ex H.B.K.) Lag.*
Sporoboleae
Sporobolus giganteus Nash
PANICOIDEAE
Andropogoneae
Sorghum halepense Pers.
Zea mays L.
Miscanthus sinensis Anderss.
Capillipedium parviflorum Stapf
Arundinelleae
Arundinella hirta (Thunberg) C. Tanaka
Danthoniopsis dinteri C. E. Hubbard
Paniceae
Panicum capillare L.
Pennisetum alopecuroides (L.) Spreng
GenBank sequence
accession numbera
Voucher
EAK V10
No voucher
EAK V21
RS 3305
HPL 5590
GBAN-AF137299
GBAN-AF137322
GBAN-AF137306
GBAN-AF137312
GBAN-AF137290
SM s.n.
GBAN-AF137292
EF s.n.
GBAN-AF137330
ML s.n.
GBAN-U61200
ML s.n.
GBAN-U61191
PP 10008
GBAN-AF137327
PI-255738
ML s.n.
Arnold Arboretum 301-80C
PI-301782
GBAN-AF137326
GBAN-AF137332
GBAN-AF137311
GBAN-AF137296
PI-246756
PI-207548
GBAN-AF137293
GBAN-AF137300
ML s.n.
EAK s.n.
GBAN-AF137316
GBAN-AF137318
a The prefix GBAN- has been added to link the online version of American Journal of Botany to GenBank but is not part of the actual GenBank
accession number.
Workbench (MACAW) software system (Schuler, Altschul, and Lipman,
1991). Phylogenetic analyses were performed using parsimony on
PAUP*4.0 (Swofford, 1998). Characters were equally weighted with
respect to codon position in all cladistic analyses, and GAPMODE was
set to MISSING for the single-codon gap found in all panicoid, arundinoid, chloridoid, and centothecoid taxa. Most parsimonious trees were
retained from a heuristic search employing 500 random addition replicates (hold 5 1000) and the tree-bisection-reconnection (TBR) branch
swapping algorithm, with STEEPEST, USENONMIN, ALLSWAP, and
MULPARS options invoked. Four alternative topologies were compared
with the minimal-length trees by performing additional heuristic searches under the same settings, but with constraints loaded. We used constraint analyses to evaluate conflict with phylogenetic patterns emerging
from other molecular data sets rather than to evaluate the many conflicts
apparent between most molecular data sets and traditional classifications. To compare levels of support for the monophyly of constituent
clades, decay indices (Bremer, 1988; Donoghue et al., 1992) were determined using AutoDecay 4.0 (Eriksson, 1998) and the same heuristic
search settings employed to obtain minimal-length trees. In addition, a
bootstrap consensus tree from 1000 replicates was generated using an
heuristic search strategy employing simple taxon addition and TBR,
retaining groups with frequency greater than 50% (Felsenstein, 1985).
We used the likelihood ratio test to investigate the assumption that
evolution of phytochrome B sequences in grasses is clocklike (Felsenstein, 1993; Huelsenbeck and Rannala, 1997). In order to facilitate computation of log likelihoods, we trimmed the data set so that no column
in the data matrix included sites coded as missing. A matrix of 26 taxa
and 1141 nucleotide sites resulted. Two maximum likelihood searches
of these data using PAUP*4.0 (Swofford, 1998) invoked the evolutionary model of Hasegawa, Kishino, and Yano (1985), but rate heterogeneity across sites was allowed to vary under a discrete approximation
to a gamma distribution with four rate categories, and the shape parameter, gamma, was estimated by maximum likelihood. The two searches
varied only with respect to whether or not branch lengths were constrained to fit a model of clocklike evolution. Starting trees for NNI
branch swapping were obtained from ten random addition replicates.
The log likelihoods from the two searches were compared to determine
whether the likelihood of the search without a clock had been significantly increased by allowing branch lengths to vary at a unique rate
(Felsenstein, 1993). Additionally, the proportions of Jukes-Cantor corrected nonsynonymous differences among taxa were estimated using
MEGA (Kumar, Tamura, and Nei, 1993), and pairwise comparisons of
evolutionary rates relative to PHYB from Anomochloa marantoidea
were estimated by the method of Wu and Li (1985).
RESULTS
We observed no sequence variation among multiple
PHYB from single taxa, consistent with our expectation
that PHYB occurs singly in the genomes of most grasses.
The final data matrix for parsimony analyses, comprising
1182 bp of the PHYB gene, contained 610 variable sites
and 430 informative characters, with 66, 33, and 331
phylogenetically informative sites at the first, second, and
third codon positions, respectively. Seven percent of the
cells in the matrix were coded as missing. Nongrass sister
taxa, Flagellaria and Joinvillea (Campbell and Kellogg,
1987; Doyle et al., 1992), were included as outgroups in
our initial analyses, despite their fragmentary nature
(;300 bp) relative to the sequences from grasses (1182
bp). In all trees, Anomochloa and a second grass, Strep-
100
AMERICAN JOURNAL
Fig. 1. (a) Generalized structure of the PHYB gene (Quail, 1994).
(b) Schematic of the PHYB insert and primer sequences. Arrows indicate direction of synthesis from primers. Degeneracy of each sequence
is indicated parenthetically. I 5 inosine.
tochaeta, are the earliest diverging grasses (not shown),
corroborating the results from analyses of ndhF (Clark,
Zhang, and Wendel, 1995) and rbcL (Duvall and Morton,
1996). We did not include Flagellaria (GenBank
U61203) and Joinvillea (GenBank U61205) in subsequent analyses in order to minimize the amount of missing data; instead, Anomochloa marantoidea was the outgroup.
Maximum parsimony analysis of the 51 PHYB sequences from grasses yielded nine equally parsimonious
trees (one shown in Fig. 2) in two islands that differ with
respect to (1) placement of Molinia as sister either to
Phragmites 1 Hakonechloa or to the chloridoids, and (2)
placement of Chasmanthium and Thysanolaena as sister
either to the panicoids or Phragmites 1 Hakonechloa.
Minimal-length trees are 2433 steps and have consistency
and retention indices of 0.35 (uninformative characters
excluded) and 0.58, respectively. The strict consensus
tree, with tribal and subfamilial divisions designated according to Clayton and Renvoize (1986), is given in Fig.
3. The herbaceous ‘‘bambusoid’’ Streptochaeta occurs in
a basal polytomy with Anomochloa (the two are sister
taxa in 60% of the bootstrap replicates when Flagellaria
and Joinvillea are included; not shown). The monophyly
of the rest of the family is supported by 37 base substitutions and bootstrap and decay (d) values of 100% and
OF
BOTANY
[Vol. 87
18, respectively. The next diverging taxa also are herbaceous ‘‘bambusoids,’’ first Pharus and then Puelia. The
early divergence of Pharus is highly supported: 38 base
substitutions and bootstrap and decay values of 100% and
16, respectively, support the monophyly of the grasses
exclusive of Pharus. Similarly, the divergence of the rest
of the grasses from Puelia is supported by 22 base substitutions and bootstrap and decay values of 93% and 6,
respectively. The nonbasal grasses are divided into two
major clades, one that includes all sampled panicoid,
arundinoid, chloridoid, and centothecoid species and one
that includes the rest of the sampled bambusoid, and all
the sampled oryzoid and pooid species.
The bambusoid–oryzoid–pooid clade (corresponding
with the BOP clade of Clark, Zhang, and Wendel, 1995)
is strongly supported as monophyletic in the PHYB phylogeny by 17 base substitutions, a bootstrap value of
96%, and a decay value of 8. Moreover, the internal structure of this clade is completely resolved and 57% of its
clades have bootstrap support .90%. For example, the
‘‘core’’ Bambusoideae (Bambuseae, Olyreae, and Parianeae), the Oryzoideae, and an expanded Pooideae that
includes the ex-arundinoid Anisopogon and the ex-bambusoid Diarrhena, occur in 98, 99 and 96% of bootstrap
replicates, respectively. There is modest support (15 base
substitutions, bootstrap 5 70%, d 5 2) for the sistergroup relationship of Pooideae with Oryzoideae.
Within Pooideae, nine of the 15 internal branches have
bootstrap values of .90%, and six have decay values of
. 6. A clade of Lygeum 1 Nardus occurs as sister to the
remainder of the subfamily in 83% of the bootstrap replicates (d 5 2).
Within Oryzoideae, Oryzeae (monophyletic in 100%
of the bootstrap replicates) and Ehrharta are sister taxa.
Within Bambusoideae, Parianeae occurs in 100% of the
bootstrap replicates and is sister to Olyreae in 83%, but
the Bambuseae are poorly supported (bootstrap ,70%, d
5 2). Placement of the herbaceous bambusoid Streptogyna is equivocal. In all nine of the shortest trees, it is
sister to the rest of the BOP clade, but there is little support for this arrangement (bootstrap ,70%, d 5 2). Constraint analysis reveals that Streptogyna is placed with the
rest of the Bambusoideae of the BOP clade in trees that
are just two steps longer than the minimal-length trees
from unconstrained analysis. In contrast, constraining
Streptogyna to fit the topology of the ndhF tree (Clark,
Zhang, and Wendel, 1995), which places it with Oryzoideae, is more costly, adding five steps to minimal-length
trees.
The clade consisting of the panicoid, arundinoid, chloridoid, and centothecoid grasses (corresponding with the
PACC clade of Davis and Soreng, 1993) is strongly resolved as monophyletic in the PHYB phylogeny, supported by 34 base substitutions, a bootstrap value of
100%, and a decay value of 7. However, in comparison
with the BOP clade, the internal structure of this clade is
less resolved. Two traditionally recognized subfamilies,
Panicoideae and Chloridoideae, are well supported with
100 and 94% bootstrap support, respectively. Monophyly
of the Centothecoideae remains undetermined because
just Chasmanthium was sampled. Chasmanthium and the
arundinoid Thysanolaena are sister taxa in 94% of bootstrap replicates. The arundinoids are polyphyletic, as has
January 2000]
MATHEWS
ET AL.—PHYB PHYLOGENY IN GRASSES
been noted in nearly all other phylogenetic studies. Aristida and Danthonia are sister taxa (bootstrap 5 91%, d
5 5), and are sister to the rest of the PACC clade in 81%
of the bootstrap replicates (d 5 3); Phragmites and Hakonechloa are sister taxa in 99% of the bootstrap replicates
(d 5 8), but their relationship with another arundinoid,
Molinia, is not retained in the consensus. Among subfamilies, Panicoideae is the best sampled, and displays the
most internal structure. Andropogoneae 1 Arundinella
and Paniceae are monophyletic in 98 and 100% of the
bootstrap replicates, respectively, and are sister taxa in
95% of the bootstrap replicates. Danthoniopsis (Arundinelleae) is sister to the rest of the Panicoideae in 95% of
the bootstrap replicates. Constraining Danthonia to a
clade with chloridoid taxa, a topology observed in both
the rbcL (Barker, Linder, and Harley, 1995) and early
chloroplast DNA restriction site (Davis and Soreng,
1993) trees resulted in three equally parsimonious trees
requiring 2447 steps, 14 steps longer than the minimallength trees. Constraining Danthonia and Aristida into a
trichotomy with Chloridoideae consistent with Barker’s
(1997) rbcL tree requires ten extra steps.
The clock and nonclock searches of the pruned PHYB
data set using maximum likelihood resulted in single and
identical trees; their topology is identical with a pruned
parsimony consensus tree. Comparison of the log likelihoods of the two trees suggests that the log likelihood is
significantly increased when branch lengths are allowed
to vary at unique rates (22 log L 5 89.52, df 5 24, P
, 0.005). Thus, evolution of exon I of PHYB in grasses
apparently is not clocklike. The Wu and Li (1985) test,
however, revealed just a single significant rate difference
between the PHYB sequences of Panicum and Phragmites (data not shown).
DISCUSSION
We found that nucleotide data from the single phytochrome paralog, PHYB, provides strong evidence of phylogenetic structure within the grass family. PHYB data
from 51 species resolves a basal grade of herbaceous
‘‘bambusoids’’ and strongly supports the monophyly of
two principal nonbasal clades of the grass family: the
first, a panicoid–arundinoid–chloridoid–centothecoid
group corresponding with the strongly supported PACC
clade of Davis and Soreng (1993); and second, a bambusoid–oryzoid–pooid group corresponding with the
weakly supported BOP clade of Clark, Zhang, and Wendel, 1995. In PHYB trees that included Flagellaria and
Joinvillea, neotropical Anomochloa and Streptochaeta are
the earliest diverging grasses, corroborating the pattern
revealed by data from ndhF (Clark, Zhang, and Wendel,
1995), from GBSSI (granule-bound starch synthase I;
Mason-Gamer, Weil, and Kellogg, 1998), and from combined chloroplast structural and restriction site characters
(Soreng and Davis, 1998), and supporting the placement
of Anomochloa as sister to the rest of the grasses by rbcL
data (Duvall and Morton, 1996). Similarly, the position
of Phareae as the next diverging branch, another feature
of the ndhF, GBSSI, and plastid restriction site trees, is
strongly corroborated. A third ‘‘bambusoid’’ element, the
West African Puelia, is here resolved as the next diverging branch and sister to the BOP and PACC clades.
101
The basal branches of the grass family tree—Members of the subfamily Bambusoideae traditionally have
been the focus of speculations concerning the most primitive members of the grass family (e.g., Celakovský,
1889; Arber, 1934; Soderstrom, 1981a). They share apparently plesiomorphic features, such as spikelet-like
branches (‘‘pseudospikelets’’) and indeterminate inflorescences. The current taxonomic circumscription of the
Bambusoideae s.l. (e.g., Clayton and Renvoize, 1986) includes the Bambuseae (mostly woody bamboos, but including Puelia), as well as the herbaceous ‘‘bambusoid’’
tribes Streptochaeteae, Anomochloeae, Phareae, Streptogyneae, Parianeae, and Olyreae. Molecular data suggest,
however, that Bambusoideae s.l. are polyphyletic. For example, data from ndhF (Clark, Zhang, and Wendel,
1995), PHYB (this study), plastid restriction sites (Soreng
and Davis, 1998), and GBSSI (Mason-Gamer, Weil, and
Kellogg, 1998) establish that the herbaceous ‘‘bambusoids,’’ Anomochloa, Streptochaeta, and Pharus, are
members of the earliest diverging lineages in the grass
family, while Parianeae, Olyreae, Streptogyneae, and
Bambuseae are nested within the BOP clade. Anomochloa 1 Streptochaeta and the pharoid lineage have been
elevated to subfamilial status, in Anomochlooideae and
Pharoideae, respectively (Clark and Judziewicz, 1996).
The strong resolution in the PHYB phylogeny of another herbaceous ‘‘bambusoid,’’ Puelia, as the next diverging branch suggests that it also is better placed outside the core Bambusoideae. This West African genus is
distinguished by several morphological features, including external ligules, dimorphic florets, and extension of
the rachilla. Soderstrom contended that Puelia was ancient (Soderstrom, 1981a), and placed it outside a core
Bambusoideae (Soderstrom, 1981a; Soderstrom and Ellis,
1987). Additional molecular data suggest that Puelia and
Guaduella, a second West African taxon also segregated
from the core Bambusoideae by Soderstrom and Ellis
(1987), form a third basal lineage warranting subfamilial
rank (GPWG, 1998a, b).
Together these findings support the view of Clark,
Zhang, and Wendel (1995) that certain presumed synapomorphies of the Bambusoideae s.l., such as the presence of a short mesocotyl internode during embryonic
development and the presence of arm cells and fusoid
cells, are better considered as synapomorphies of the
whole family that are lost from later lineages. Thus, these
characters of the embryo and leaf can be added to the
long list of morphological characters that distinguish the
grasses (Campbell and Kellogg, 1987).
Support for the BOP clade—The PHYB phylogeny is
unique in its highly supported resolution of a BOP clade.
The BOP clade is only weakly supported by the ndhF
data (6-bp mutations, d 5 1 in Clark, Zhang, and, Wendel, 1995) and does not occur in other chloroplast gene
trees. In rbcL and recent plastid restriction site trees,
Pooideae are sister to the PACC clade (Barker, 1997;
Barker, Linder, and Harley, 1995; Duvall and Morton,
1996; Soreng and Davis, 1998). Clark and colleagues
(1995) hypothesized that the weak support for the BOP
group in their ndhF data may have resulted from rapid
divergence among its constituent lineages. The absence
of this strongly resolved element of the PHYB phylogeny
102
AMERICAN JOURNAL
OF
BOTANY
[Vol. 87
January 2000]
MATHEWS
ET AL.—PHYB PHYLOGENY IN GRASSES
from chloroplast trees, and from trees inferred from combined plastid data sets (L. Clark, J. Davis, and E. Kellogg,
unpublished data), is puzzling. It is possible that molecular changes in PHYB coincided with morphological radiation of this clade, while changes in the chloroplast
genome did not. Alternatively, ambiguity in the plastid
data sets with regard to relationships among the BOP
subfamilies might result from patterns of plastid lineage
sorting that conflict with the nuclear and/or species phylogeny, or from reticulation.
Phylogenetic structure within the BOP clade—Oryzoid genera are placed within Bambusoideae s.l. in many
classifications (e.g., Clayton and Renvoize, 1986; Soderstrom and Ellis, 1987; Kellogg and Watson, 1993.) In
the most comprehensive analysis of morphological data,
the oryzoid tribes Oryzeae and Ehrharteae are nested
within a monophyletic Bambusoideae s.l. (Kellogg and
Watson, 1993), while evidence of their close relationship
emerges from analyses of both morphological data (Hilu
and Wright, 1982; Kellogg and Campbell, 1987; Kellogg
and Watson, 1993) and plastid data (Cummings, King,
and Kellogg, 1994; Clark, Zhang, and Wendel, 1995; Soreng and Davis, 1998). In the ndhF consensus tree, Oryzeae 1 Ehrharteae, together with Streptogyna, are one of
the trichotomous lineages of the BOP clade (Clark,
Zhang, and Wendel, 1995). In contrast, the PHYB phylogeny places the monophyletic oryzoid group in a moderately supported sister-group relationship with Pooideae
(70% bootstrap support, d 5 2), while the ‘‘core’’ Bambusoideae (Bambuseae, Olyreae, and Parianeae) are resolved as a basal lineage in the BOP clade. Possible morphological synapomorphies of a pooid-oryzoid alliance
include lodicule number (2), first seedling internodes that
elongate, presence of adventitious roots at the coleoptilar
node, a narrow first seedling blade, and absence of the
embryonic scutellar cleft.
The placement of Streptogyna remains problematic.
Based on the amphiatlantic tropical distribution of extant
species of Streptogyna, Soderstrom (1981a) included it
among taxa he considered archaic. This is not inconsistent with its placement in the PHYB phylogeny, where it
is sister to the rest of the BOP clade. However, as noted
above, the ndhF phylogeny places Streptogyna as sister
to the oryzoids (Clark, Zhang, and Wendel, 1995). Constraining the PHYB phylogeny to this arrangement adds
five steps to the length of minimal-length trees, while
constraining it to occur with the core Bambusoideae adds
just two steps.
Our support for a monophyletic pooid group is consistent with a large body of existing morphological and molecular evidence. Kellogg and Campbell (1987), Cummings, King, and Kellogg (1994), and Soreng and Davis
(1998) supported a (Nardus 1 Lygeum) group basalmost
in a larger pooid clade, a topology that is corroborated
by the PHYB phylogeny. The relative order of divergence
of the pooid tribes Meliceae and Stipeae has not been
103
resolved unambiguously in any phylogenetic analysis
thus far. The PHYB data set unites Meliceae with Stipeae
(represented by Nassella), but this topology is not well
supported and contradicts other findings (e.g., Hsiao et
al., 1995; Catalán, Kellogg, and Olmstead, 1997; Soreng
and Davis, 1998); thus, it should be confirmed with additional data from Stipeae. Chloroplast data sets from
ndhF and from restriction site 1 morphological characters resolve Diarrhena within the Pooideae (Clark,
Zhang, and Wendel, 1995; Catalán, Kellogg, and Olmstead, 1997; Soreng and Davis, 1998), justifying the contention of Kellogg and Campbell (1987) that this ‘‘bambusoid’’ grass belonged in Pooideae. Similarly, the arundinoid genus Anisopogon is firmly placed in Pooideae by
PHYB data, a pattern also noted in analysis of ndhF (J.
Davis, Cornell University, personal communication).
Thus, the PHYB data concur with other data in supporting
a revised circumscription of the Pooideae that is more
inclusive than current taxonomic treatments of the subfamily.
Phylogenetic structure within the PACC clade—Recent chloroplast gene trees and the PHYB phylogeny concur in the strong resolution of a monophyletic PACC
group consisting of panicoid, arundinoid, chloridoid, and
centothecoid grasses (e.g., Barker, Linder, and Harley,
1995; Clark, Zhang, and Wendel, 1995; Duvall and Morton, 1996; Barker, 1997; Soreng and Davis, 1998). Within
this group, the ‘‘Arundinoideae’’ have long been considered a miscellaneous assemblage of genera, lacking even
a single character that would unite its disparate taxa or a
subset thereof. Moreover, most comprehensive phylogenetic analyses support the view that the subfamily is paraphyletic or polyphyletic (e.g., Kellogg and Campbell,
1987; Barker, Linder, and Harley, 1995; Clark, Zhang,
and Wendel, 1995; Mathews and Sharrock, 1996; Barker,
1997; Soreng and Davis, 1998; but see Hsiao et al.,
1998). In the PHYB phylogeny, three reedy Arundineae
[Molinia, (Hakonechloa, Phragmites)] are a clade in six
of the nine shortest trees, perhaps representing an arundinoid core. The fourth arundinoid, Danthonia, is resolved as the sister to Aristida, (of the Aristideae of Clayton and Renvoize, 1986), and the two are sister to the
rest of the PACC clade. Danthonia is our single sample
from a sizable group of arundinoid genera (the Danthonieae), all of which apparently share the putative synapomorphy of haustorial synergids in the embryo sac (Verboom, Linder, and Barker, 1994). In the rbcL tree of
Barker (1997), separate lineages of reedy arundinoids,
danthonioids, and aristidoids are resolved. However, in
contrast to the PHYB phylogeny, the latter two are in a
clade with Chloridoideae. Constraining the PHYB phylogeny to create a trichotomy of Aristida, Danthonia, and
Chloridoideae is costly, requiring ten extra steps.
The monophyly of Chloridoideae is supported in analysis of anatomical and physiological characters (including, among other features, the predominance of the C4
←
Fig. 2. A phylogeny of PHYB DNA sequences from 51 grasses (1182 nucleotides, 430 informative characters). Heuristic parsimony analysis (500
random addition replicates with TBR swapping in PAUP*4.0; Swofford, 1998) yielded nine most parsimonious trees of 2433 steps; CI 5 0.35
(autapomorphies excluded); RI 5 0.58. Numbers above branches are inferred branch lengths under ACCTRAN optimization.
104
AMERICAN JOURNAL
OF
BOTANY
[Vol. 87
January 2000]
MATHEWS
ET AL.—PHYB PHYLOGENY IN GRASSES
photosynthetic pathway). Recent molecular analyses concur in the monophyly of this subfamily as circumscribed
in current taxonomic treatments, as do the PHYB data,
which resolve a well-supported chloridoid clade, despite
the fragmentary nature of the sequences from Bouteloua
and Eragrostis.
The monophyly of Panicoideae is supported by morphological features such as paired spikelets with a single
female fertile floret (Kellogg and Campbell, 1987). Subsequent phylogenetic analyses have supported the monophyly of the panicoid tribe Andropogoneae and its sistergroup relationship to the Paniceae (e.g., Davis and Soreng, 1993; Barker, Linder, and Harley, 1995; Duvall and
Morton, 1996; Mathews and Sharrock, 1996), while the
Arundinelleae are polyphyletic (e.g., Barker, Linder, and
Harley, 1995; Spangler et al., 1999). Analysis of PHYB
data strongly corroborates the polyphyly of Arundinelleae
in placing Danthoniopsis as sister to the rest of the panicoids, and Arundinella as sister to the Andropogoneae.
The true affinities of centothecoid genera have
emerged just recently, having been recognized as a tribe
within Bambusoideae (e.g., Watson and Dallwitz, 1992),
as a tribe with arundinoid affinities (e.g., Kellogg and
Campbell, 1987; Soderstrom and Ellis, 1987), or as a
separate subfamily (Soderstrom, 1981b; Clayton and
Renvoize, 1986). Chloroplast restriction site data were
the first to place the centothecoid Chasmanthium firmly
in the PACC clade (Davis and Soreng, 1993), a position
confirmed by both rbcL and ndhF data (Barker, Linder,
and Harley, 1995; Clark, Zhang, and Wendel, 1995). The
plastid data further revealed the relationship of centothecoids with the arundinoid Thysanolaena. Data from
PHYB strongly support the relationship of Chasmanthium
with Thysanolaena, and six of the nine shortest trees are
consistent with the plastid data in placing them as sister
to the Panicoideae.
The utility of PHYB sequences—Our analyses illustrate the power and utility of sampling a single paralog
of the phytochrome multigene family for resolving relationships within Poaceae. Phylogenetic structure within
Poaceae is well defined by the PHYB data. The basal
branches and principle clades are highly supported, 69%
of all clades in the phylogeny are supported by bootstrap
values of .90%, and structure within the BOP clade is
fully resolved and well supported. Thus, the nuclear
PHYB phylogeny is a robust test of the plastid phylogenies, which mostly are corroborated. Phytochrome sequences may be especially useful for resolving relationships within Poaceae because of their relatively rapid rate
of nucleotide substitution; estimates suggest that they
have evolved faster than the plastid sequences from
which most molecular phylogenies of Poaceae have been
inferred (Mathews, Lavin, and Sharrock, 1995; Olmstead
105
and Reeves, 1995). Although we presented evidence that
evolution of the PHYB sequences in grasses is not clocklike, we cannot attribute any conflict observed among the
plastid and PHYB phylogenies to this phenomenon. For
example, we wondered whether the well-supported resolution of the BOP clade by the PHYB data might result
from more rapid nucleotide evolution in this clade relative to evolution in the PACC clade, which also is resolved by the plastid data sets. This is not the case, however; relative rate tests revealed that sequences from BOP
taxa are evolving more slowly relative to Anomochloa
than sequences from PACC taxa, although the Wu and
Li (1985) test does not suggest that the differences are
significant (data not shown). Additional points of conflict
are the placement of Streptogyna, Danthonia, and Aristida; we did not analyze these conflicts in further detail
because they are being investigated in combined analyses
of eight data sets from the grasses, of which the PHYB
data set is one (GPWG, 1998a, b).
LITERATURE CITED
ARBER, A. 1934. The Gramineae: a study of cereal, bamboo, and grass.
University Press, Cambridge, UK.
AVDULOV, N. P. 1931. Kario-sistematicheskoye issledovaniye semeystva zlakov. Prilozheniye 44 k Trudy po prikladnoy botanike, genetike I selektsii, 428 pages. Leningrad. [Karyosystematic studies in
the grass family. Supplement 44 to the Bulletin of Applied Botany,
Genetics and Plant Breeding. Leningrad. Russian text, pages 1–
352; German summary, pages 353–425; index, pages 426–428. English translation (mislabeled as supplement 43) published for the
Smithsonian Institution and the National Science Foundation,
Washington, D.C., by the Indian National Scientific Documentation
Centre, New Delhi, 1975. TT 70–53085.]
BARKER, N. P. 1997. The relationships of Amphipogon, Elytrophorus,
and Cyperochloa (Poaceae) as suggested by rbcL sequence data.
Telopea 7: 205–213.
———, H. P. LINDER, AND E. H. HARLEY. 1995. Polyphyly of the
Arundinoideae (Poaceae): evidence from rbcL sequence data. Systematic Botany 20: 423–435.
BREMER, K. 1988. The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42: 795–803.
BROWN, R. 1810. Prodromus Florae Novae Hollandiae et Insulae VanDiemen, v. 1, viii 1 145–592 pages. J. Johnson, London, UK.
———. 1814. General remarks, geographical and systematical, on the
botany of Terra Australis. In M. Flinders, A voyage to Terra Australis, vol. 2, 553–613. London, UK.
BROWN, W. V. 1958. Leaf anatomy in grass systematics. Botanical Gazette 119: 170–178.
CAMPBELL, C. S. 1985. The subfamilies and tribes of Gramineae (Poaceae) in the southeastern United States. Journal of the Arnold
Arboretum 66: 123–199.
———, AND E. A. KELLOGG. 1987. Sister group relationships of the
Poaceae. In T. R. Soderstrom, K. W. Hilu, C. S. Campbell, and M.
E. Barkworth [eds.], Grass systematics and evolution, 217–224.
Smithsonian Institution, Washington, DC, USA.
CATALÁN, P., E. A. KELLOGG, AND R. G. OLMSTEAD. 1997. Phylogeny
of Poaceae subfamily Pooideae based on chloroplast ndhF gene
sequences. Molecular Phylogenetics and Evolution 8: 150–166.
←
Fig. 3. Strict consensus of the nine most parsimonious trees from parsimony analysis (500 random addition replicates with TBR swapping in
PAUP*4.0; Swofford, 1998) of PHYB DNA sequences from 51 grasses (1182 nucleotides, 430 informative characters). All trees are 2433 steps; CI
5 0.35 (autapomorphies excluded); RI 5 0.58. Tribal designation (labels at far right) follows Clayton and Renvoize (1986). BA 5 Bambusoideae,
OR 5 Oryzoideae, PO 5 Pooideae, CH 5 Chloridoideae, AR 5 Arundinoideae, PA 5 Panicoideae, CE 5 Centothecoideae. Numbers of above
branches are bootstrap percentages $70% from 1000 replicates. Numbers below branches are decay values determined using AutoDecay 4.0
(Eriksson, 1998).
106
AMERICAN JOURNAL
CELAKOVSKÝ, L. 1889. Über den ärchenbau der Brasilianischen Grasgatung Streptochaeta Schrader. Sitzungsberichte der Königlichen
Böhmischen Gesellschaftder Wissenschaften. Mathematischnaturwissenschaftliche Classe 3: 14–42, Table 2. [Unpublished translation by Mrs. Geti Saad, 1974, for the Smithsonian Institution and
the National Science Foundation, Washington, D. C.; in the files of
the McClure Bamboo Library, Department of Botany, Smithsonian
Institution, Washington, DC.]
CHILDS, K. L., F. R. MILLER, M.-M. CORDONNIER-PRATT, L. H. PRATT,
P. W. MORGAN, AND J. E. MULLET. 1997. The sorghum photoperiod
sensitivity gene, Ma3 encodes a phytochrome B. Plant Physiology
113: 611–619.
CLACK, T., S. MATHEWS, AND R. A. SHARROCK. 1994. The phytochrome
apoprotein family in Arabidopsis is encoded by five genes: the
sequences and expression of PHYD and PHYE. Plant Molecular
Biology 25: 413–427.
CLARK, L. G., AND E. J. JUDZIEWICZ. 1996. The grass subfamilies Anomochlooideae and Pharoideae. Taxon 45: 641–645.
———, W. ZHANG, AND J. F. WENDEL. 1995. A phylogeny of the grass
family based on ndhF sequence data. Systematic Botany 20: 436–
460.
CLAYTON, W. D., AND S. A. RENVOIZE. 1986. Genera graminum. Her
Majesty’s Stationery Office, London, UK.
CUMMINGS, M. P., L. M. KING, AND E. A. KELLOGG. 1994. Slippedstrand mispairing in a plastid gene: rpoC2 in grasses (Poaceae).
Molecular Biology and Evolution 11: 1–8.
DAHLGREN, R. M. T., H. T. CLIFFORD, AND P. F. YEO. 1985. The families
of the monocotyledons. Springer, Berlin, Germany.
DAVIS, J. I., AND R. J. SORENG. 1993. Phylogenetic structure in the
grass family (Poaceae) as inferred from chloroplast DNA restriction
site variation. American Journal of Botany 80: 1444–1454.
DOEBLEY, J., M. DURBIN, E. M. GOLENBERG, M. T. CLEGG, AND D. P.
MA. 1990. Evolutionary analysis of the large subunit of carboxylase (rbcL) nucleotide sequence data among the grasses (Poaceae).
Evolution 44: 1097–1108.
DONOGHUE, M. J., R. G. OLMSTEAD, J. F. SMITH, AND J. D. PALMER.
1992. Phylogenetic relationships of Dipsacales based on rbcL sequences. Annals of the Missouri Botanical Garden 79: 333–345.
DOYLE, J. J. 1992. Gene trees and species trees: molecular systematics
as one-character taxonomy. Systematic Botany 17: 144–163.
———. 1997. Trees within trees: genes and species, molecules and
morphology. Systematic Biology 46: 537–553.
———, J. I. DAVIS, R. J. SORENG, D. GARVIN, AND M. ANDERSON. 1992.
Chloroplast DNA inversions and the origin of the grass family (Poaceae). Proceedings of the National Academy of Sciences, USA 89:
7722–7726.
DUVALL, M. R., AND B. R. MORTON. 1996. Molecular phylogenetics of
Poaceae: an expanded analysis of rbcL sequence data. Molecular
Phylogenetics and Evolution 5: 352–358.
ERIKSSON, T. 1998. AutoDecay ver. 4.0 (program distributed by the
author). Department of Botany, Stockholm University, Stockholm,
Sweden.
FELSENSTEIN, J. 1985. Confidence limits on phylogenies: An approach
using the bootstrap. Evolution 39: 783–791.
———. 1993. PHYLIP (Phylogeny Inference Package), version 3.5c.
Department of Genetics, University of Washington, Seattle, Washington, USA.
GALLOWAY, G. L., R. L. MALMBERG, AND R. A. PRICE. 1998. Phylogenetic utility of the nuclear gene arginine decarboxylase. Molecular Biology and Evolution 15: 1312–1320.
GAUT, B. S., AND M. T. CLEGG. 1991. Molecular evolution of alcohol
dehydrogenase 1 in members of the grass family. Proceedings of
the National Academy of Sciences, USA 90: 2060–2064.
GOTTLIEB, L. D., AND V. S. FORD. 1996. Phylogenetic relationships
among the sections of Clarkia (Onagraceae) inferred from the nucleotide sequences of PgiC. Systematic Botany 21: 45–62.
GPWG (GRASS PHYLOGENY WORKING GROUP). 1998a. Proposal for a
subfamilial reclassification of the Poaceae. In Monocots II, Abstracts, 15. Royal Botanic Gardens Sydney, Australia.
———. 1998b. A phylogeny of the grass family (Poaceae), as inferred
from eight character sets. In Monocots II, Abstracts, 18. Royal
Botanic Gardens, Sydney, Australia.
HAMBY, R. K., AND E. A. ZIMMER. 1988. Ribosomal RNA sequences
OF
BOTANY
[Vol. 87
for inferring phylogeny within the grass family (Poaceae). Plant
Systematics and Evolution 160: 29–37.
HASEGAWA, M. H., H. KISHINO, AND Y. YANO. 1985. Dating of the
human-ape splitting by a molecular clock of mitochondrial DNA.
Journal of Molecular Evolution 21: 160–174.
HATTERSLEY, P. W., AND L. WATSON. 1992. Diversification of photosynthesis. In G. P. Chapman [ed.], Grass evolution and domestication. Cambridge University Press, Cambridge, UK.
HECKER, K. H., AND K. H. ROUX. 1996. High and low annealing temperatures increase both specificity and yield in touchdown and stepdown PCR. Biotechniques 20: 478–485.
HILU, K. W., AND K. WRIGHT. 1982. Systematics of Gramineae: a cluster analysis study. Taxon 31: 9–36.
HSIAO, C., N. J. CHATTERTON, K. H. ASAY, AND K. B. JENSEN. 1995.
Molecular phylogeny of the Pooideae (Poaceae) based on nuclear
rDNA (ITS) sequences. Theoretical and Applied Genetics 90: 389–
398.
———, S. W. L. JACOBS, N. P. BARKER, AND N. J. CHATTERTON. 1998.
A molecular phylogeny of the subfamily Arundinoideae (Poaceae)
based on sequences of rDNA. Australian Systematic Botany 11:
41–52.
HUELSENBECK, J. P., AND B. RANNALA. 1997. Phylogenetic methods
come of age: testing hypotheses in an evolutionary context. Science
276: 227–232.
KELLOGG, E. A., AND C. S. CAMPBELL. 1987. Sister group relationships
of the Poaceae. In T. R. Soderstrom, K. W. Hilu, C. S. Campbell,
and M. E. Barkworth [eds.], Grass systematics and evolution, 217–
224. Smithsonian Institution, Washington, DC, USA.
———, AND L. WATSON. 1993. Phylogenetic studies of a large data
set. I. Bambusoideae, Andropogonodae, and Pooideae (Gramineae).
Botanical Review 59: 273–343.
KUMAR, S., K. TAMURA, AND M. NEI. 1993. MEGA: Molecular Evolutionary Genetics Analysis, version 1.01. Pennsylvania State University, University Park, Pennsylvania, USA.
LAVIN, M., E. ESHBAUGH, J.-M. HU, S. MATHEWS, AND R. A. SHARROCK.
1998. Monophyletic subgroups of the tribe Millettieae (Leguminosae) as revealed by phytochrome nucleotide sequence data.
American Journal of Botany 85: 412–433.
MASON-GAMER, R. J., C. F. WEIL, AND E. A. KELLOGG. 1998. Granulebound starch synthase: structure, function, and phylogenetic utility.
Molecular Biology and Evolution 15: 1658–1673.
MATHEWS, S., AND M. S. DONOGHUE. 1999. The root of angiosperm
phylogeny inferred from duplicate phytochrome genes. Science
286: 947–950.
———, M. LAVIN, AND R. A. SHARROCK. 1995. Evolution of the phytochrome gene family and its utility for phylogenetic analysis of
angiosperms. Annals of the Missouri Botanical Garden 82: 296–
321.
———, AND R. A. SHARROCK. 1996. The phytochrome gene family in
grasses (Poaceae): a phylogeny and evidence that grasses have a
subset of the loci found in dicot angiosperms. Molecular Biology
and Evolution 13: 1141–1150.
———, AND ———. 1997. Phytochrome gene diversity. Plant, Cell,
and Environment 20th Anniversary Special Issue: Photomorphogenesis 20: 666–671.
———, R. C. TSAI, AND E. A. KELLOGG. 1997. The advantage of sampling single phytochrome loci versus multiple loci for inference of
phylogeny: An example from the grasses. American Journal of Botany 84 (supplement): 628 (Abstract).
NADOT, S., R. BAJON, AND B. LEJEUNE. 1994. The chloroplast gene rps4
as a tool for the study of Poaceae phylogeny. Plant Systematics
and Evolution 191: 27–38.
OLMSTEAD, R. G., AND P. A. REEVES. 1995. Evidence for the polyphyly
of the Scrophulariaceae based on chloroplast rbcL and ndhF sequences. Annals of the Missouri Botanical Garden 82: 176–193.
PRAT, H. 1936. La systematique des graminées. Annals des Sciences
Naturelles, Botanique series 10, 14: 117–324.
QUAIL, P. H. 1994. Phytochrome genes and their expression. In R. E.
Kendrick and G. H. M. Kronenberg [eds.], Photomorphogenesis in
plants, 71–104. Kluwer Academic Publishers, The Netherlands.
REEDER, J. R. 1957. The embryo in grass systematics. American Journal of Botany 44: 756–769.
RENVOIZE, S. A., AND W. D. CLAYTON. 1992. Classification and evo-
January 2000]
MATHEWS
ET AL.—PHYB PHYLOGENY IN GRASSES
lution of grasses. In G. P. Chapman [ed.], Grass evolution and domestication, 3–37. Cambridge University Press, Cambridge, UK.
ROW, H. C., AND J. R. REEDER. 1957. Root-hair development as evidence of relationships among genera in the Gramineae. American
Journal of Botany 44: 596–601.
SANDERSON, M. J., AND J. J. DOYLE. 1992. Reconstruction of organismal
and gene phylogenies from data on multigene families: concerted
evolution, homoplasy, and confidence. Systematic Biology 41: 4–
17.
SANG, T., M. J. DONOGHUE, AND D. ZHANG. 1997. Evolution of alcohol
dehydrogenase genes in peonies (Paeonia): phylogenetic relationships of putative nonhybrid species. Molecular Biology and Evolution 14: 994–1007.
SCHULER, G. D., S. F. ALTSCHUL, AND D. J. LIPMAN. 1991. A workbench
for multiple alignment construction and analysis. Proteins: Structure, Function, and Genetics 9: 180–190.
SHARROCK, R. A., AND P. H. QUAIL. 1989. Novel phytochrome sequences in Arabidopsis thaliana: structure, evolution, and differential
expression of a plant regulatory photoreceptor family. Genes and
Development 3: 1745–1757.
SMALL, R. L., J. A. RYBURN, R. C. CRONN, T. SEELANAN, AND J. F.
WENDEL. 1998. The tortoise and the hare: choosing between noncoding plastome and nuclear Adh sequences for phylogeny reconstruction in a recently diverged plant group. American Journal of
Botany 85: 1301–1315.
SODERSTROM, T. R. 1981a. Some evolutionary trends in the Bambusoideae (Poaceae). Annals of the Missouri Botanical Garden 68:
15–47.
———. 1981b. The grass subfamily Centostecoideae. Taxon 30: 614–
616.
———, AND R. ELLIS. 1987. The position of the bamboo genera and
allies in a system of grass classification. In T. R. Soderstrom, K.
107
W. Hilu, C. S. Campbell, and M. E. Barkworth [eds.], Grass systematics and evolution, 225–238. Smithsonian Institution, Washington, DC, USA.
SORENG, R. J., AND J. I. DAVIS. 1998. Phylogenetics and character evolution in the grass family (Poaceae): simultaneous analysis of morphological and chloroplast DNA restriction site characters. Botanical Review 64: 1–85.
SPANGLER, R., B. ZAITCHIK, E. RUSSO, AND E. A. KELLOGG. 1999. Andropogoneae evolution and generic limits in Sorghum (Poaceae)
using ndhF sequences. Systematic Botany 24: 267–281.
SWOFFORD, D. L. 1998. PAUP*4.0: Phylogenetic analysis using parsimony, beta test version 4d64. Sinauer, Sunderland, Massachusetts,
USA.
TATEOKA, T., S. INOUE, AND S. KAWANO. 1959. Notes on some grasses.
IX. Systematic significance of bicellular microhairs of leaf epidermis. Botanical Gazette 121: 80–91.
VERBOOM, G. A., H. P. LINDER, AND N. P. BARKER. 1994. Haustorial
synergids: an important character in the systematics of danthonioid
grasses (Arundinoideae: Poaceae). American Journal of Botany 81:
1601–1610.
WATSON, L., AND M. J. DALLWITZ. 1992. The grass genera of the world.
CAB International, Wallingford, Oxon, UK.
WHITLOCK, B. A., AND D. A. BAUM. 1999. Phylogenetic relationships
of Theobroma and Herrania (Sterculiaceae) based on sequences of
the nuclear gene vicilin. Systematic Botany 24: 128–138.
WU, C.-I., AND W.-H. LI. 1985. Evidence for higher rates of nucleotide
substitution in rodents than in man. Proceedings of the National
Academy of Sciences, USA 82: 1741–1745.
ZIMMER, E. A., S. L. MARTIN, S. M. BEVERLY, Y. W. KAN, AND A. C.
WILSON. 1980. Rapid duplication and loss of genes coding for the
alpha-chains of hemoglobin. Proceedings of the National Academy
of Sciences, USA 77: 2158–2162.