Phylogeny and Subfamilial Classification of the Grasses (Poaceae) Author(s): Grass Phylogeny Working Group, Nigel P. Barker, Lynn G. Clark, Jerrold I. Davis, Melvin R. Duvall, Gerald F. Guala, Catherine Hsiao, Elizabeth A. Kellogg, H. Peter Linder Source: Annals of the Missouri Botanical Garden, Vol. 88, No. 3 (Summer, 2001), pp. 373-457 Published by: Missouri Botanical Garden Press Stable URL: http://www.jstor.org/stable/3298585 Accessed: 06/10/2008 11:05 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=mobot. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org. Missouri Botanical Garden Press is collaborating with JSTOR to digitize, preserve and extend access to Annals of the Missouri Botanical Garden. http://www.jstor.org Volume 88 Annals Number 3 of the 2001 Missouri Botanical Garden PHYLOGENY AND Grass Phylogeny Working Group2,3 SUBFAMILIAL CLASSIFICATION OF THE GRASSES (POACEAE)' ABSTRACT A large collaborative effort has yiel(led a comprehensive study of the phylogeny and a new suhfanilial classification of the grass family (Poaceae/Graminieae)T. he stu(ly was (con(luc(tedon an integratedla ndlr epresentatives et of 62 grasses (0.6% of the species and ca. 8% of the genera) plus four outgroup taxa using six molecular sequence (lata sets ({ndhFl, rbcL, rpoC2, phyB, ITS2, and (;BSSI or waxy), chloroplast restriction site (lata, and(m orphologicali data. A parsimony analysis using 2143 informativec haracters (the comblineda nalysis) resulted in a single most parsimonioust ree of 8752 steps with an RI of 0.556 and bootstrap support of > 90% for more than half of the internal no(les. Significant relationships that appear consistently in all analyses of all (lata sets and are strongly supported by the combined analysis include the following: Joinvilleaceae are sister to a monophyletic Poaceae; the earliest (liverging lineages of the Poaceae are Anomochlooideae, Pharoideae, and Puelioideae, respectively; and(a ll remaining grasses form a clade. Multiple monophyletic clades were recovere(, including Bambusoideae s. str., Ehrhartoideae, Pooideae s.l., Aristidoi- deae, l)anthonioideae, Chloridoideaes . str., Chloridoideae s.l., Panicoideae, Parianeae, Olyreae s. str., Oryzeae, Stipeae, Meliceae, Lygeum + Nardus, and Molinia + Phragmites. 'The PACCAI) Clade is monophyletic, containing Aristidoi- deae, Danthonioideae, Arundinoideae s. str., Chloridoideae s.l., Centothecoideae, Panicoideae, Eriachne, Micraira, and Gynerium.B ased on the phylogeny, a classification of 11 previously published subfamilies (Anomochlooideae, Pharo- ideae, Puelioideae, Bambusoideae, Ehrhartoideae, Pooideae, Aristidoideae, Arundinoideae, Chloridoideae, Centothe- coideae, and Panicoideae) and 1 new subfamily (Danthonioideae) is proposed. Several changes in the circumscription of traditionally recognized subfamilies are included. Previous phylogenetic work and classifications are reviewed in relation to this classification and circumscription, and major characteristics of each subfamily are discussed and de- scribed. The matrix, trees, and updated data matrix are available at (http://www.virtualherbarium.org/grass/gpwg/ default.htm). Key words: cereals, classification, DNA sequence data, evolution, grass, phylogeny, Poaceae. I Workp resented here was supportedi n part by NSF grants DEB-9806584 and DEB-9806877 to LGC, DEB-9727000 to JID, DEB-9419748 and DEB-9815392 to EAK, and BIR-9508467 to SYM. Miwa Kojima prepared the line illustra- tions of leaf anatomy and spikelets. We thank T. Cope, J. Everett, S. W. L. Jacobs, S. Phillips, S. A. Renvoize, and P. F. Stevens for helpful comments on the manuscript. 2 This paper is to be cited as authored by the Grass Phylogeny WorkingG roup, or GPWG. The group includes the following members, listed here in alphabetical order; there is no senior author. Nigel P. Barker, Departmento f Botany, Rhodes University, P.O. Box 94, Grahamstown,6 140, South Africa; Lynn G. Clark, Department of Botany, Iowa State University, Ames, Iowa 50011-1020, U.S.A.; Jerrold I. Davis, L. H. Bailey Hortorium, Cornell University, 462 Mann Library, Ithaca, New York 14853, U.S.A.; Melvin R. Duvall, Department of Biological Sciences, Northern Illinois University, DeKalb, Illinois 60115-2861, U.S.A.; Gerald F. Guala, Fairchild Tropical Garden, 11935 Old Cutler Road, Miami, Florida 33156, U.S.A.; Catherine Hsiao, 6005 CrossmontC ourt, San Jose, California 95120, U.S.A.; Elizabeth A. Kellogg, Department of Biology, University of Missouri-St. Louis, 8001 Natural Bridge Road, St. Louis, Missouri 63121, U.S.A.; H. Peter Linder, Institut fur Systematische Botanik, Zollikerstrasse 107, CH-8008, Zurich, Switzerland; ANN. MISSOURBI OT. GARD.8 8: 373-457. 2001. 374 Annals of the Missouri Botanical Garden The economic and ecological significance of the and are somewhat more distantly related to Flagel- grasses (Poaceae) has led to widespread interest in lariaceae (Dahlgren et al., 1985; Campbell & Kel- their evolution and classification. The cereals, sug- logg, 1987; Linder, 1987; Doyle et al., 1992; Kel- arcane, bamboos, and forage and weedy grasses are logg & Linder, 1995; Briggs et al., 2000); this group of pre-eminent importance in human economy. forms the graminoid clade, a subset of the order Grasses, which occur in virtually every terrestrial Poales (APG, 1998). A sister-group relationship be- habitat, cover as much as one-fifth of the Earth's tween Poaceae and Joinvilleaceae consistently has land surface (Shantz, 1954). Long recognized as a been supported (Campbell & Kellogg, 1987; Doyle "natural" group, the grass family includes approx- et al., 1992; Clark et al., 1995; Soreng & Davis, imately 10,000 species in over 700 genera (Dahl- 1998), although rbcL sequence data suggest that gren et al., 1985; Tzvelev, 1989; Watson & Dall- Joinvilleaceae + Ecdeiocoleaceae is the sister witz, 1992; Renvoize & Clayton, 1992). Efforts to clade to Poaceae (Briggs et al., 2000). produce a comprehensive, formal taxonomic struc- The grass family was recognized as distinctive ture of the family began over 200 years ago, while and coherent long before the term monophyly was serious study of grass evolution began late in the ever applied. The fruit (caryopsis) is unique to the 19th century. family, with the outer integument developmentally The Grass Phylogeny Working Group (GPWG) fused to the inner wall of the ovary. The embryo is was established in 1996 to (1) combine a series of lateral and, unlike most monocot embryos, is highly existing data sets to produce a comprehensive phy- differentiated, with clear shoot and root meristems, logeny for the grass family; (2) focus taxon sampling leaves and vascular system. The pollen, as is typ- in the development of existing and future data sets; ical for the whole order Poales, has only one ap- and (3) reevaluate the subfamilial classification of erture, but in grasses the pollen wall lacks scrobi- the grass family based on the results of the phylo- culi. In all but the earliest-diverging lineage, the genetic analyses. We combined and performed cla- grass spikelet consists of a set of distichous bracts, distic analyses on eight data sets (one structural, the basal two empty (glumes) with a series of one four plastome, and three nuclear) from 62 grasses to 4 many bracts (lemmas) above, each lemma sub- and outgroup taxa. The phylogenetic results and a contracted floral axis on which is borne a revised subfamilial classification of the fam- tending grass a two or three reduced ily are in this presumed prophyll (palea), presented paper. perianth parts (lodicules), the androecium, and the REVIEW OF GItASS PHYLOGENYA ND gynoecium (see discussion under Spikelet). Although subdivision of the grasses into CLASSIFICATI)N groups we today call tribes began in the 18th century (see Historically, the Poaceae were thought to be re- reviews in Calder6n & Soderstrom, 1980; Gould & lated to Cyperaceae (Engler, 1892; Cronquist, Shaw, 1983; Pohl, 1987), works by Brown (1810, 1981) based on floral reduction and chemical char- 1814) represent the earliest attempt to define acters, but evidence accumulated during the past groups of tribes, or what we now call subfamilies. 15 years unequivocally shows that the similarities Brown (1814) divided the grasses into the tribe are convergent. Phylogenetic studies based on mor- Paniceae (roughly equivalent to the modern Pani- phological and molecular characters show that the coideae) and the tribe Poaceae (roughly equivalent grasses are most closely related to Joinvilleaceae, to the Festucoideae of Hitchcock & Chase, 1950) Restionaceae, Anarthriaceae, and Ecdeiocoleaceae, based on spikelet compression, articulation, and floret number. Brown is credited with describing Roberta of Sci- grass spikelets in detail and recognizing them as J. Mason-Gamer,D epartment Biological of Idaho branched structures, as well as ences, University Idaho, Moscow, 83844, U.S.A.; noting the tendency Sarah Y. Mathews, Division of Biological Sciences, Uni- for the panicoids to grow in warm climates and the versity of Missouri-Columbia,2 26 Tucker Hall, Columbia, pooids in cooler climates (Gould & Shaw, 1983; Missouri 56211, U.S.A.; MarkP . Simmons, The Ohio State Pohl, 1987). Brown's division of the into two University Herbarium, Ohio State University, 1315 Kin- family near Ohio Robert So- major groups was formalized Bentham Road, Columbus, 43212, U.S.A.; J. by (1878), reng, Department of Botany, Natural History Museum, was retained by Bentham and Hooker (1883) and Smithsonian Institution, Washington, D.C. 20560-0166, Hackel (1887), and persisted well into the 20th U.S.A.; Russell E. Spangler, Departmento f Ecology, Evo- Hitchcock, 1935; Hitchcock & lution, and Behavior, University of century (e.g., Minnesota, 1987 Upper Buford Circle, St. Paul, Minnesota 55108, U.S.A. Chase, 1950). 3Author for correspondence: Elizabeth A. Kellogg, Several classifications for the grasses based on tkellogg@umsl.edu. spikelet and inflorescence morphology were pro- Volume 88, Number 3 Grass Phylogeny Working Group 375 2001 Phylogeny and Classification of Poaceae posed in the 19th century (see reviews in Calder6n of Diarrhena, Nardus, and Lygeum, the remaining & Soderstrom, 1980; Gould & Shaw, 1983; Camp- seven clusters corresponded to the subfamilies Fes- bell, 1985; Pohl, 1987), with usually nine or ten tucoideae, Oryzoideae, Arundinoideae, Centothe- tribes recognized. Some tribes, for example Pani- coideae, Panicoideae, Eragrostoideae, and Bam- ceae, Andropogoneae, and Bambuseae, contain busoideae. Watson et al. (1985) used the DELTA largely the same genera now as nearly 200 years system to conduct comprehensive phenetic analy- ago. Others, such as the various "pooid"t ribes, in- ses of the family, and their character list continues cluded disparate elements and are now seen as ar- to be developed. Watson and Dallwitz (1992) ini- tificial. tially recognized five subfamilies and subsequently Whether explicit or not, a different perspective updated their classification to include seven (Wat- on the evolution of grasses and relationships within son & Dallwitz, 1999; http://biodiversity.uno.edu/ the family began to emerge by the end of the 19th delta/grass/www/class.htm); these are Stipoideae, century. Workerss uch as Celakovsky (1889), Goe- Pooideae, Bambusoideae, Centothecoideae, Arun- bel (1895), and Schuster (1910) carefully analyzed dinoideae, Chloridoideae, and Panicoideae. Sub- spikelet structure and proposed that Streptochaeta, sequent phenetic analyses of immunological data or something very much like it, was representative (Esen & Hilu, 1989) and plastid DNA reassociation of the most primitive grasses. With the development (Hilu & Johnson, 1991) were limited in sampling of leaf anatomical (Duval-Jouve, 1875; Prat, 1932), but in each case produced four major groups. embryological (van Tieghem, 1897), and cytological Only within the past 15 years have cladistic (Avdulov, 1931) data, a profound reassessment of methods been applied to questions of grass phylog- evolutionary relationships among grasses began. eny and evolution. The first attempt to produce an Additional data on embryo anatomy (Reeder, 1957, explicit hypothesis of relationships was the mor- 1961, 1962), starch grains (Tateoka, 1962), lodi- phological phylogeny of Kellogg and Campbell cules (Jirisek & Jozifova, 1968; Guedes & Dupuy, (1987), who analyzed 33 characters scored for vir- 1976), and leaf anatomy (Brown, 1958; Metcalfe, tually all grass genera. The pooids (including Sti- 1960) accumulated and were also incorporatedi nto peae), Panicoideae, Chloridoideae, and Bambuso- evolutionary and classification schemes. Several ideae were consistently monophyletic in their classification systems were published in the 20th analyses, but Arundinoideae were polyphyletic, and century (e.g., Roshevits, 1937, 1946; Tateoka, the pooid clade formed the basal lineage in the 1957; Prat, 1960; Stebbins & Crampton, 1961; family. Bambusoideae s.l. (including herbaceous Jacques-Felix, 1962; Caro, 1982; Clayton & Ren- tribes such as Anomochloeae, Phareae, Strepto- voize, 1986; Tzvelev, 1989; Renvoize & Clayton, chaeteae, and Streptogyneae) were interpreted as 1992; Watson & Dallwitz, 1992); major ones that monophyletic based on the presence of arm and are global in scope are compared in Table 1. The fusoid cells; several tribes often included in the number of subfamilies recognized ranged from 2 tra(litional Bambusoideae were placed in other (Tzvelev, 1989) to 13 (Caro, 1982). All but the Wat- clades (e.g., Brachyelytreae, Diarrheneae, and son and Dallwitz (1992) classification, which is Phaenospermatidae in the pooid clade). avowedly phenetic, were based on presumed evo- Hamby and Zimmer (1988) and Doebley et al. lutionary relationships. The major change was the (1990) published the first molecular phylogenies for subdivision of the old Festucoideae (or Pooideae) the family, based respectively on ribosomal RNA into several subfamilies; Panicoideae were retained and plastid gene rbcL (ribulose 1,5-bisphosphate almost without modification. Other differences carboxylase/oxygenase, large subunit) sequence among the major classification systems primarily data. Relatively few taxa were sampled in both relate to the treatment of Arundinoideae and Bam- studies, but both supported the core Pooideae as busoideae. Clayton and Renvoize (1986) in partic- well as the group that came to be known as the ular published a number of diagrams depicting re- PACC clade (Davis & Soreng, 1993), containing lationships based on their synthesis of knowledge subfamilies Panicoideae, Arundinoideae, Centothe- at that time. These diagrams have served as a start- coideae, and Chloridoideae. ing point for much subsequent work. The first extensive application of molecular data Phenetic analyses of the grass family generally to grass phylogeny was undertaken by Davis and found groups consistent with the five or six subfam- Soreng (1993), using plastid DNA restriction site ilies commonly recognized by the mid 1980s. Hilu variation for 31 taxa representing the six subfami- and Wright( 1982), in a cluster analysis of morpho- lies of Clayton and Renvoize (1986). This study logical and anatomical data, found eight major marked the beginning of wider sampling in the tra- groups with strong support. Except for the cluster ditional Bambusoideae (= Bambusoideae s.l.), long C) Table 1. Comparisono f the major 20th century classification systems of the Poaceae. 0) Roshevits (1946) Tateoka (1957) Prat (1960) Caro (1982) Clavton & Renvoize Tzvelev (1989) Watson & Dallwitz Grass Phylogeny (1986) (1992) WorkingG roup (present paper) Bambusoideae Pharoideae Bambusoideae Bambusoideae Bambusoideae Bambusoideae Bambusoideae Bambusoideae Bambuseae Bambuseae Bambuseae s.l. Bambuseae Bambuseae Bambuseae Bambuseae Bambuseae Arundinarieae 77 Shibataeeae 77 Dendrocalameae Melocanneae Oxytenanthereae Oryzoideae Olyroideae Olyreae Olyreae Olyreae Olyreae Olyreae Olyreae Olyreae Olyreae Olyreae Olyreae Parianeae Parianeae Pariana ? Parianeae Parianeae ,, In Olyreae ? 9 In ,, Olyreae Buergersiochloeae Anomochlooideae Anomochlooideae Anomochloeae Anomochloeae Incertae Sedis Anomochloeae Anomochloeae Anomochloeae Anomochloeae Anomochloeae Streptochaetoideae > Streptochaeteae Streptochaeteae Streptochaeta Streptochaeteae Streptochaeteae Streptochaeteae Streptochaeteae Streptochaeteae P) Pharoideae 0 c[Io1 Phareae In Olyreae 9 Phareae Leptaspideae Phareae Phareae P) (D _. Puelioideae In Bambuseae 9 In Bambuseae In Bambuseae Atractocarpeae Puelieae Puelieae G) In Bambuseae 9 In Bambuseae In Bambuseae In Bambuseae Guaduelleae Guaduelleae Q- (D :3 Pooideae Ehrhartoideae See Arundoideae Incertae Sedis 9 Streptogyneae Streptogyneae Streptogyneae Incertae Sedis Oryzoideae Pooideae Oryzeae Oryzeae Oryzeae Oryzeae Oryzeae Oryzeae Oryzeae Oryzeae ? 9 Diarrheneae Diarrheneae Diarrheneae See Pooideae See Arundoideae 9 Brachyelytreae Brachyelytreae Brachyelytreae See Pooideae rN o Table 1. Continued. o... CD Phenospermeae See Arundoideae Phaenospermatae Phaenospermatae Phaenospermatae See Pooideae o Phyllorachieae 9 Phyllorachideae Phyllorachideae Phyllorachideae Phyllorachideae z Ehrhartoideae In Oryzeae See Arundoideae Incertae Sedis Ehrharteae Ehrharteae Ehrharteae Ehrharteae Ehrharteae (D Centhostecoideae Centothecoideae Centothecoideae co Centotheceae See Arundoideae Incertae Sedis Centhosteceae Centotheceae Centosteceae Centotheceae Centotheceae Micraireae Thysanolaeneae Pooideae Festucoideae Festucoideae Pooideae Pooideae Pooideae Hordeae Triticeae Hordeae Triticeae Triticeae Triticeae Triticeae Brachypodieae Brachypodieae Brachypodieae Brachypodieae Bromeae Bromeae Bromeae Bromeae Bromeae Festuceae Festuceae Festuceae Poeae Poeae Poeae Poeae Aveneae Agrosteae Aveneae Aveneae Aveneae Agrostideae Agrostideae Phalarideae Phalarideae In Phleeae , Centotheceae I3 0 :) Streptogyneae Ehrharteae 0 o Phaenospermeae C . -h Garnotieae CD Arundinelleae Arundinelleae 0 3 Panicoideae Eragrostoideae Chloridoideae Eragrostoideae Chloridoideae Chloridoideae Chloridoideae G) Q- Thysanolaeneae D In Eragrostideae In Cynodonteae Triodieae In Eragrostideae Pappophoreae Pappophoreae Pappophoreae Pappophoreae Pappophoreae Pappophoreae Orcuttieae Orcuttieae Orcuttieae Chlorideae Chlorideae Chlorideae Cynodonteae Cynodonteae Chlorideae Cynodonteae ?? ,, Zoysieae Lappagineae Zoysieae Spartineae Eragrosteae In Chlorideae Eragrosteae Eragrostideae Eragrostideae o r o 0 Table 1. Continued. 3 CD Sporoboleae In Chlorideae Sporoboleae 00 Leptureae Leptureae various isolated Centropodia,M erx- genera muellera rangei 0cDC0T CD Panicoideae Panicoideae Panicoideae Panicoideae Panicoideae Panicoideae CIO Steyermarkochloeae Steyermarkochloeae See Arundinoideae Steyermarkochloeae In Eragrosteae 9 9 Eriachneae In Arundineae See Arundinoideae Incertae Sedis ? 9 Hubbardieae 9 In Isachneae Hubbardieae In Festuceae Isachneae Isachneae Isachneae Isachneae Isachneae Isachneae Paniceae Paniceae Paniceae Paniceae Paniceae Paniceae Paniceae Boivinelleae Bovinelleae Boivinelleae ,, ,, Melinideae Melinideae Melinideae ,, ,, In Paniceae 9 ,? Anthephoreae In Paniceae Lecomtelleae 9 In ,, Zoysieae Trachyeae 9 In ,, Zoysieae 9 Neurachneae In Melinidae 9 Arthropogoneae In Paniceae Arundinelleae See Arundoideae See Phragmitiformes Arundinelleae Arundinelleae Arundinelleae Arundinelleae Andropogoneae Andropogoneae Andropogoneae Andropogoneae Andropogoneae Andropogoneae I,, Andropogoneae ,< ) o 0 Maydeae Maydeae ,, Maydeae Maydeae various isolated CD t genera including V<,< Phaenosperma Incertae Sedis Incertae Sedis 0) C Anomochloa Eriachneae o1) 0 Centotheca . Micraireae (Q Ehrharta Streptogyneae Lygeum, Nardus O Cyperochloa J0 Micraira Gynerium oo 0C ) Streptogyna CD C) Thysanolaena and CD other genera CA Co 380 Annals of the Missouri Botanical Garden presumed to include the most ancestral elements of as the monophyly of the core Bambusoideae, in ad- the grass family. Davis and Soreng's (1993) results dition to supporting the PACC and pooid clades. supported an expanded pooid clade, the PACC The topology recovered by Liang and Hilu (1996) (now PACCAD) clade, and suggested that the tra- from analysis of matK (maturase K) sequence data ditional Bambusoideae were not monophyletic. was similar to the rbcL topologies, with the PACC Nadot et al. (1994) analyzed sequences of the and pooid clades sister to each other, and Oryza plastid gene rps4 (ribosomal plastid small subunit, sister to that clade and a woody bamboo sister to protein 4) for 26 genera of grasses. Their sampling the whole family. By this time, reassessment of sub- was heavily weighted toward the pooid grasses, but familial classification was necessary; Clark and they did include three genera of woody bamboos Judziewicz (1996) resurrected Anomochlooideae and Zizania and Oryza of the ehrhartoids (ory- and Pharoideae to accommodate the basal lineages zoids). They recovered both a monophyletic pooid of the family, which could not be retained in a clade (including Stipa) and the PACC clade. The monophyletic Bambusoideae. bambusoid/oryzoid taxa were paraphyletic and Soreng and Davis (1998) combined a structural formed a polychotomy with the PACC clade. Cum- data set (including morphological, anatomical, mings et al. (1994), using sequence data from the chromosomal, and biochemical characters as well plastid rpoC2 (RNA polymerase II, [" subunit) as structural features of the chloroplast genome) gene, sampled only 13 genera, but did derive a and an expanded chloroplast restriction site data monophyletic PACC clade and a monophyletic set to analyze phylogenetic relationships within the pooid clade. The rbcL sequence analysis of Barker grass family. They confirmed the basal positions of et al. (1995) focused on the subfamily Arundino- Anomochlooideae and Pharoideae, monophyly of an ideae. Both the PACCa nd pooid clades were shown expanded Pooideae, monophyly of Panicoideae, to be monophyletic, although the traditional Arun- Centothecoideae, and Chloridoideae, and polyphyly dinoideae appeared as polyphyletic. Bambuso- of the traditional Arundinoideae. The core Bam- ideae, represente(db y a woody bamboo (Bambusa) busoideae, supported as monophyletic in other and Ehrhartoideae (Zizania and Oryza), were par- analyses along with the ehrhartoid grasses, ap- aphyletic to the rest of the family. peared as a set of clades paraphyletic to the [Bra- Clark et al. (1995) were the first to include a chyleytrum + (Pooideae + PACC)] clade. Soreng broad sample of bambusoid and ehrhartoid taxa. and Davis (1998) also identified structural syna- Using ndhF (NADH dehydrogenase, subunit F) se- pomorphies for major clades, including, for exam- quence data, they confirmed the polyphyly of the ple, loss of the epiblast and gain of an elongated traditional Bambusoideae and demonstrated that mesocotyl internode in the PACC clade. Anomochloa and Streptochaeta, two broad-leaved Barker et al. (1999) used sequences of the grass- Neotropical forest genera, formed the earliest di- specific insert in the chloroplast gene rpoC2 (here- verging branch of the family, with Pharus, another after referred to only as rpoC2) to study relation- broad-leaved tropical forest genus, constituting the ships among a broad sample of "arundinoid" taxa. next most basal branch. Their results also con- They were the first to include molecular data on firmed strong support for monophyly of the PACC such traditionally arundinoid genera as Centropo- clade, an expanded pooid clade (including Stipeae, dia, Merxmuellera, Notochloe, Tribolium, Mona- Phaenospermatideae, Brachyelytreae, and Diarrhe- chather, Pentaschistis, Prionanthium, Cortaderia, neae), a derived, monophyletic core bambusoid and Spartochloa. Because Arundinoideae were clade (Olyreae + Bambuseae), and the polyphyly known to be polyphyletic, previous classifications of the traditional Arundinoideae. They also recov- were not helpful in placing these genera. rpoC2 se- ered a weakly supported clade including the core quences of Anomochloa and Streptochaeta could not bambusoids, ehrhartoids, and pooids, which they be aligned with those of other grasses, so their basal named the BOP clade (here updated to the BEP position could not be tested. Relationships among Clade based on nomenclatural priority of Ehrhar- the Bambusoideae, Pooideae s.l., and the PACC toideae over Oryzoideae). They concluded that clade varied depending on the analytical method many features previously used to define the tradi- and inclusion of phylogenetically informative in- tional Bambusoideae, including the presence of sertion/deletion characters. Consistent with previ- arm and fusoid cells and pseudopetiolate leaf ous studies, they identified Panicoideae and Chlor- blades among others, were probable synapomor- idoideae as monophyletic. They showed clearly that phies for the family. a large clade corresponding to Danthonioideae is The rbcL study of Duvall and Morton( 1996) con- monophyletic and that this corresponds at least in firmed the basal placement of Anomochloa, as well part to the clade with haustorial synergids identi- Volume 88, Number 3 Grass Phylogeny Working Group 381 2001 Phylogeny and Classification of Poaceae fled by Verboom et al. (1994). They also showed sistent with certain broad phylogenetic groups (Kel- that the genus Merxmuellera is polyphyletic, with logg, 1998). Unique combinations of linkage groups one species, M. rangei, most closely related to Cen- are synapomorphic for subfamilies Pooideae (Moore tropodia and the chloridoids. et al., 1995; Gale & Devos, 1998), Panicoideae Hilu et al. (1999) sequenced the chloroplast (Moore et al., 1995; Gale & Devos, 1998; Wilson gene matK for 62 species of Poaceae and produced et al., 1999), and Ehrhartoideae (Kennard et al., a tree that was quite similar to those found in pre- 1999). In addition, unique linkages support mono- vious studies. Streptochaeta and Anomochloa were phyly of Triticeae (Devos et al., 1993) and Andro- the earliest diverging lineages, although paraphy- pogoneae (Wilson et al., 1999). letic instead of monophyletic. The matK data sup- Phylogenetic analyses of individual molecular ported a PACC clade and a clade including Pooi- data sets within the last decade have converged on deae and Bambusoideae. Oryzoideae ( a set of well-supported relationships within Po- Ehrhartoideae) was sister to the PACC clade, rather aceae. Changes in the circumscriptions of subfam- than the pooid/bambusoid clade, but this was not ilies, and in the number of subfamilies recognized, strongly supported. clearly are necessary. The GPWG analyses pre- Zhang (2000) used the intron in the chloroplast sented here provide robust support for the major gene rpll6 (ribosomal protein 16) to construct a clades within the grass family, and provide the ba- phylogeny of the grasses and confirmed (again) the sis for the first family-wide subfamilial classifica- early divergence of Anomochloa and Streptochaeta, tion based on an explicit phylogenetic hypothesis. although his data did not support the monophyly of the pair. The next branch was Pharus. The rpll6 MATERIALSA NI) METHODS data supported a PACC clade and a BEP Clade; O()RGANIZATIONO F THlE CPWC Puelia olyriformis was sister to the BEP Clade, with modest bootstrap support. The Grass Phylogeny Working Group was formed Analyses of nuclear sequence data have provid- explicitly to combine available (lata on the phylog- ed results complementary to those obtained for mo- eny of the grass family an(l to use these (lata to lecular plastid data sets. Mathews and Sharrock )rop)ose a new classification. Most contributors had (1996) and Mathews et al. (2000) sequenced loci already published papers on grass phylogeny an(d in the phytochrome gene famlily antl resolved a to- were invite(l to contri)ute their (lata, both pul)- pology similar to that (lerived from IIndtlhs equence lished and unpublished. Each contributor retaine(l data, although the phytochrome data provided sig- control over his or her data an(l was free to publish nificantly stronger support for the B3EP Clade than at any time, but the group agreed that the entire did the ndhF data. Additionally, the basal positions data set woul(l be published as a single paper. Most of Anomochloa, Pharus, and Puelia (as noted in of the collaboration has been conducted via e-mail, Zhang, 1996) were confirmed. Hsiao et al. (1999) an(d the entire group has never met in a single inferred phylogenetic relationships within the place. This may serve as a model for future collab- grasses based on sequences of the internal tran- orations in plant systematics. scribed spacer (ITS) region of nuclear ribosomal A list of taxa was drawn up in 1995 by LGC, DNA. As in the other studies, Streptochaeta and JID, and EAK to improve parallel sampling for all Pharus were resolved as the basal lineages, and data sets (Table 2; Appendix I). This list was cho- monophyly of the PACC clade and monophyly of sen to include as many of the major lineages in the the Pooideae were strongly supported. Unlike pre- family as possible, based on our knowledge from vious studies, however, some of Hsiao et al.'s (1999) previous studies. Although sampling of taxa is still analyses found that the traditional Arundinoideae not perfectly parallel, many sequences were gen- were monophyletic. erated for this particular set of taxa. DNA was ex- Combined analysis of sequence data from two changed as necessary among members of the group. chloroplast genes (ndhF, rbcL) and one nuclear The list was expanded slightly based on results ac- gene (phytochrome B) provided strong support for quired during the study. the placement of Puelia + Guaduella as the next All sequences available at the end of 1997 were most basal lineage of the family after Anomoch- assembled by EAK into a single large data set in looideae and Pharoideae (Clark et al., 2000). These NEXUS format. The data set was then distributed results necessitated the description of a new sub- via e-mail to all participants, who had the oppor- family, the Puelioideae. tunity to comment on it. A "final" version of the Mapping studies of the nuclear genome are in data set was then distributed. Based on the results their infancy, but genome rearrangements are con- of the first round of analyses (GPWG, 2000), the 382 Annalso f the MissourBi otanicaGl arden Table 2. Summary of genes and taxa included in the combined analysis. Taxa are listed approximately in the order in which they appear in Figure 1. cp rs = chloroplast restriction sites; GBS = GBSSI; struc. = structural data; * = composite taxon, represented by sequences from several genera; # = composite taxon, represented by sequences of different species within the same genus (as in Appendix I). For details of species, authorities, origina1 publications, and GenBank accession numbers, see Appendix I. Merxmuellera m. = Merxmuellera macowanii, Merxmuellera r. Merxmuellerar angei. Genus cp rs ndhF phyB rbcL rpoC GBS ITS struc. Flagellaria x x x x x Elegia# x x x Baloskion x x x x Joinvillea# x x x x x x x Anomochloa x x x x x x Streptochaeta# x x x x x Pharus# x x x x x x Guaduella x x x Puelia x x x x Eremitis x x x x x Pariana x x x x Lithachne# x x x x x x Olyra# x x x x x Buergersiochloa x x x Pseudosasa* x x x x x x Chusquea# x x x x x x x Streptogynca x x x Ehrharta# x x x x x x Oryza x x x x x x x x Leersia# x x x x x Phaenosperma x x Brachyelytrum x x x x Lygeum x x x x x x x Nardus x x x x x x Anisopogon x x x x x x Ampelodesmos x x x x Stipa# x x x x x x Nassella# x x x x X Piptatherum# x x X X Brachypodium# x x x x X Melica# x x x X X X Glyceria# x x x X X X Diarrhena# x x x X X Avena* x x x x x X X Bromus# x x x x x X X Triticum* x x x x X X X Aristida# x x x x x X X Stipagrostis x x x X X Amphipogon# x x x x X X Arundo x x x x X X Molinia* x x x x x X X Phragmites x x x x x X X Merxmuelleram . x x x X X X Karroochloa x x x X X X Danthonia# x x x x x X X Austrodanthonia x x x X X X Merxmuellerar . x x x X X X Centropodia x x x X X X Eragrostis# x x x x x X X Uniola x x x Zoysia# x x x Volume 88, Number 3 Grass Phylogeny Working Group 383 2001 Phylogeny and Classification of Poaceae Table 2. Continued. Genus cp rs ndhF phyB rbcL rpoC GBS ITS struc. Distichlis X X X Pappophorum* X X X X Spartina# X X X X X Sporobolus# X X X X X Eriachne# X X X Micraira# X X X X Thysanolaena X X X X X X Gynerium X X X X X Chasmanthium# X X X X X X X Zeugites X X Danthoniopsis# X X X X Panicum# X X X X X X Pennisetum# X X X X X X X X Miscanthus* X X X X X X X X Zea X X X X X X X taxon list was expanded to include several more groups. This choice was based on previous work danthonioid taxa, and the matrix was recompiled (summarized in Kellogg & Linder, 1995) indicating by JID. Although all participants in the GPWG that these represent the closest relatives of the were invited to undertake data analyses and com- grasses. The results of Briggs et al. (2000) suggest ment on the final version of matrices, abstracts, and that the Ecdeiocoleaceae should be included in text, this was not required. Thus the analyses and outgroup comparisons for the grasses in the future. text of this paper reflect largely the work of LGC, Within the grass family, 62 exemplar species were JID, and EAK, with input from several members of chosen to represent the commonly recognized sub- the group. The GPWG website was created and is families Anomochlooideae, Arundinoideae, Bam- maintained by GFG. busoideae, Centothecoideae, Chloridoideae, Ehr- hartoideae (=Oryzoideae), Panicoideae, Pha- TAXA roideae, and Pooideae, as well as species from sev- The taxa used in this analysis include four gen- eral genera whose placement was uncertain (Am- era representing the families Flagellariaceae, Res- phipogon, Anisopogon, Brachyelytrum, Buerger- tionaceae (two genera), and Joinvilleaceae as out- siochloa, Danthonia, Danthoniopsis, Eriachne, Table 3. Tree statistics for subsets of the (lata matrix. The percent missing data is the number of cells that are missing for the particular block when included in the total data matrix, and thus is equivalent to the number of missing taxa times the total number of informativec haracters plus missing data within sequences for scored taxa. Chloroplast r.s. = Chloroplastr estriction sites. Total # # Inf. % Missing Data set # Taxa characters characters data Length CI RI # Trees Morphological 66 53 50 16.0 227 0.300 0.690 38,000+ Chloroplastr .s. 45 364 293 42.2 939 0.312 0.569 7 ndhF 65 2210 680 4.7 2894 0.379 0.582 16 phytochromeB 40 1182 417 45.5 1997 0.369 0.522 1 rbcL 37 1344 213 44.8 651 0.448 0.660 1 rpoC2 34 777 150 49.9 374 0.503 0.648 33 GBSSI 19 773 213 71.2 720 0.479 0.504 1 ITS 47 322 127 28.8 745 0.349 0.541 24 cp sequence data 66 4331 1043 25.2 3952 0.399 0.597 8 All cp data 66 4695 1336 26.5 4903 0.381 0.589 3 Nuclear 57 2277 757 51.9 3513 0.382 0.512 8 All molecular 66 6972 2093 33.6 8488 0.378 0.554 6 Combined data 66 7025 2143 33.2 8752 0.375 0.557 1 384 Annals of the Missouri Botanical Garden Table 4. List of structuralc haracters and states. The first number in parentheses indicates the number of the same character in Soreng and Davis (1998), and the second number indicates the number of the same character in Kellogg and Campbell (1987); a "-" indicates that the character was not used in one or both of those analyses. Culm 1 (1;-). Perennating woody culms: 0 = absent; 1 = present. 2 (2;-). Culm internodes: 0 = solid; 1 = hollow. Leaf 3 (3;-). Leaf sheath margins: 0 = free; 1 = fused for at least 1/4 of length. 4 (4;-). Adaxial ligule type: 0 = membrane (with or without fringe of hairs); 1 = fringe of hairs only. 5 (5;-). Abaxial (contra-) ligule: 0 = absent; 1 = present. 6 (-;-). Leaf blade: 0 = absent; 1 = present. 7 (6;-). Pseudopetiole: 0 = absent; 1 = present. Spikelet 8 (-;-). Floret with a structure identifiable as a palea, this recognized as present when a flower arises on a contracted axis above an enshrouding prophyll (or something like it), in the axil of a lemma: 0 = absent; 1 = present. 9 (-;-). Spikelet pairs: 0 = absent; 1 = present. 10 (7;-). Pedicel of spikelet: 0 = absent; 1 = present. 11 (8; 3, 4). Proximal female-sterile florets in female-fertile spikelets: 0 = absent; 1 = present. 12 (9; 6). Number of female-fertile florets per female-fertile spikelet: 0 = two or more; 1 = one. 13 (10;-). Awn or mucro on fertile or sterile lemma: 0 = absent; 1 = present. 14 (-;-). Number of awns: 1 1 awn; 2 = 13-23 awns (unique to Pappophoruzm3); 3 awns. 15 (11;-). Awn attachment:0 = terminal / subterminal; 1 = from a sinus; 2 = dorsal. 16 (12; 1). Disarticulation above glumes: 0 = absent; 1 = present. 17 (13;-). Germinationf lap in lemma: 0 = absent; 1 = present. Flower 18 (14;-). Lodicules: 0 = absent; 1 = present. 19 (15; 7). Lodicule number: 2 = two; 3 = three. 20 (16;-). Fusion of anteriorp air of lodicules: 0 = free; 1 = fused. 21 (17; 8). Distally membranousp ortion of lodicule: 0 = absent; 1 = present. 22 (18; 9). Lodicule vascularization:0 = very faint to absent; 1 = prominent. 23 (19; 1(). Inner whorl, posterior stamen: 0 = absent; 1 = present. 24 (19; 10). Inner whorl, anterior stamen pair: 0 = absent; 1 = present. 25 (19; 10). Outer whorl, anterior stamen: 0 = absent; 1 = present. 26 (19; 10). Outer whorl, posterior stamen pair: 0 = absent; 1 = present. 27 (-;-). Anthers tetrasporangiate,d ithecal = 0; anthers bisporangiate, monothecal = 1. 28 (20;-). Styles fused at least at base: 0 = absent; 1 = present. 29 (21;-). Number of stigmas: 1 = one; 2 = two; 3 = three; 4 = four. 30 (22;-). Highest order of stigmatic branching present: 1 = simple (unbranched,o r with branches comiposedo f single elongate papillate receptive cells, or with very short branches composed of a few papillate receptive cells, but in the latter case the stigmas linear in outline); 2 = primary (branches well developed, composed of series of dispersed papillate receptive cells, with secondary branches absent or minimally developed, stigmas lanceolate or [broader)3; = secondary (secondary to tertiaryb ranches well developed, branches composed of series of dispersed papillate receptive cells). 31 (23;-). Number of locules and ovules per pistil (all three families have one ovule per locule): 1 = oile; 2 = two; 3 = three. Embryogeny 32 (-;-). Haustorial synergids: 0 = absent; 1 = present. Fruit and Embryo 33 (24; 11). Hilum: 0 = long-linear, > 1/3 length of grain; 1 = nonlinear, < 1/3 length of grain, elliptical or broader to punctiform. 34 (25;-). Embryo position and structure:0 = embedded, simple; 1 = lateral, grass-type. 35 (26; 15). Embryo epiblast: 0 = absent; 1 = present. 36 (27; 16). Embryo scutellar tail: 0 = absent; 1 = present. 37 (28; 17). Embryo mesocotyl internode: 0 = negligible; 1 = elongate. 38 (29; 18). Embryonic leaf margins: 0 = meeting; 1 = overlapping. 39 (30;-). Endosperm lipid: 0 = absent; 1 = present. 40 (31;-). Endosperm starch grain syndromes: 0 = Triticum-type( simple grains only, dimorphic in size, round or lenticular, free); 1 = Festuca-type (highly compound grains present, with or without simple grains also present); 2 = Andropogon-type( simple and compound grains both present, the latter consisting of few granules); 3 = Panicum-type (simple grains only, uniform in size, small to medium, angular or sometimes smooth walled, densely packed); 4 = Brachyelytrum-type( simple only, large). Seedling 41 (32; 20). Lamina of first seedling leaf: 0 = absent; 1 = present. Volume 88, Number 3 Grass Phylogeny Working Group 385 2001 Phylogeny and Classification of Poaceae Table 4. Continued. Vegetative Anatomy 42 (-;-). Differentiation of leaf epidermal cells into long and short (cork) cells: 0 = absent (i.e., cells ? undiffer- entiated); 1 = present (Campbell & Kellogg, 1987). 43 (34; 21). Multicellular microhairs:0 = absent; 1 = present. 44 (35; 22). Occurrence in multicellular microhairs of a broad, short terminal cell, often with a longer basal cell, the walls of the terminal and basal cells similar in thickness: 0 = absent; 1 = present. 45 (36; 31). Arm cells: 0 = absent; 1 = present. 46 (37;-). Fusoid cells: 0 = absent; 1 = present. Chromosomes 47 (-;-). Base chromosome number is same as state number except that 0 = 10; 1 = 11; 2 = 12; 3 = 13; 4 = 18; 5 = 19. Biochemistry 48 (38; 30). Carbon fixation pathway:0 = C,; 1 = C4 NADP-ME classical-type; 2 = C4 NADP-ME Aristida-type;3 = C4 NAD-ME;4 = C4 NADP-ME Arundinelleae-type; 5 = C4 NADP-ME Eriachne-type. 49 (39; 30). Carbon fixation PCK: 0 = absent; 1 = present. Deletion in Phytochrome B 50 (-;-). 3-bp DNA deletion in phytochrome B: 0 = 3-bp DNA present (i.e., non-deleted state); 1 = DNA absent (i.e., deleted state; the deleted codon is at position 402 in the alignment of Mathews et al., 1995). Chloroplast Genome Structure 51 (40;-). 6.4 kb inversion in the large single-copy region of the chloroplast genome, relative to the gene arrangement in Nicotiana: 0 = absent; 1 = present. 52 (41;-). trnT inversion in the large single-copy region of the chloroplast genome, relative to the gene arrangement in Nicotiana: 0 = absent; 1 = present. 53 (42;-). 15 bp insertion in ndhF at position 101951 of the chloroplast genome of Oryza sativa: 0 = absent; 1 present. Lygeum, Micralira, Nard(is, Pari(t(a, Phaenosper- here, should break up each conglomerate taxon into ma, Puelia, Streptogyna, 7Thysanolaelna). real species (i.e., exempflar taxa). For 31 of the terminal taxa in the matrix, all 'he numb)er of taxa was (dictated by the numb)ers molecular data were taken from a single species; of available sequences in the largest of the original for an additional 27, data were from two or more data sets (rdh F andt chloroplast restriction sites). species of the same genus (noted l)y # in Table 2; Recent work on large phylogenies suggests that Appendix I). In eight cases, however, data from sev- phylogenetic accuracy is improve(d by a very dense eral genera were combined to create a "conglom- sample of taxa (e.g., Hillis, 1996, 1998; Graybeal, erate" taxon (asterisks in Table 2). For example, 1998). Producing a large data set with perfectly although one listed representative of the Andropo- parallel sampling, however, would have required ei- goneae is Miscanthus, there is no rbcL sequence ther a centralized effort in a single lab, or a formal, available for that genus. There is, however, a se- coordinated, and separately funded effort among quence for Sorghum. Thus the Sorghum sequence multiple labs, rather than the decentralized ap- for rbcL was combined with the Miscanthus se- proach used here. quences for ndhF, creating a fictive taxon, an ap- proach used previously by Kellogg and Linder CHARACTE RS (1995). This assumes that both genera are part of a monophyletic higher-level group (in this case, Andropogoneae, which are The data matrix included 7025 characters as- certainly monophyletic; Spangler et al., 1999). The results of such combi- sembled from the following sources: nations are in that as- 1. NADH subunit F potentially misleading, they dehydrogenase, (ndhF)- sume certain combinations of characters that may Clark et al. (1995, 2000); Davis et al. (this not ever actually occur in a single plant. We feel paper); Spangler et al. (1999). that the number of characters involved, however, is 2. Ribulose 1,5-bisphosphate carboxylase/oxygen- small, and the addition of phylogenetically infor- ase, large subunit (rbcL)-Barker et al. (1995); mative characters by including the line of data out- Barker (1997); Doebley et al. (1990); Duvall and weighs the risk of misleading results. Any subse- Morton (1996). quent studies, particularly those for which there are 3. RNA polymerase II, [3" subunit (rpoC2)-Cum- more than two representatives of taxa combined mings et al. (1994); Barker et al. (1999). 386 Annals of the Missouri Botanical Garden 4. Chloroplast restriction sites-Davis and Soreng different ways the two programs count resolutions (1993); Soreng and Davis (1998). of polytomies. 5. Phytochrome B (phyB)-Mathews and Sharrock PAUP* analyses used 10 random addition se- (1996); Mathews et al. (2000). quences, MULPARS on, TBR branch swapping, 6. Internal transcribed spacer of the nuclear ribo- and MAXTREES set to automatically increase by somal RNA (ITS)-Hsiao et al. (1998, 1999). 100. Bootstrapa nalyses (bts) used the full heuristic 7. Granule bound starch synthase I (GBSSI, or option, 500 or 1000 replicates. Bremer support( ab- waxy)-Mason-Gamer et al. (1998). breviated here as brs; Bremer, 1988; Kallersji et 8. Morphology-Soreng and Davis (1998, and ad- al., 1992; also called decay index, cf. Donoghue et ditional members of the GPWG, this paper). al., 1992) was also calculated. For tree lengths up Information on numbers of characters and taxa to 11 steps longer than the shortest tree (up to 8763 for each matrix is in Table 3, and the structural steps), all trees were saved and the strict consensus character list is in Table 4. The morphological computed. Because of memory limitations the (structural) matrix is in Appendix II. The first four method of negative constraints (Baum et al., 1994) data sets represent the chloroplast genome, and the was used to compute higher Bremer supportv alues. next three the nuclear genome. Five data sets, The search for optimal trees was found to be quite ndhF, rbcL, rpoC2, phyB, and GBSSI, are all protein inefficient with this method and often led to inflated coding sequences; introns of GBSSI were not in- support values. To minimize this problem, each cluded in the alignments. The full data matrix in- search was done with 10,000 random addition se- cluded 66 taxa and 7025 characters, for a total of quences. Even so, the search frequently found trees 463,650 cells. The amount of missing data for the shorter than the negative constraint tree, indicating total data set is 33.2% and varies among genes and that the previous searches had missed some trees. taxa (Table 3). The full data matrix can be obtained Computing Bremer support thus took almost two from LGC, JID, EAK, or HPL, or from the GPWG weeks of computer time on a G3. For the tree pre- website, or at Tree BASE (http://herbaria.harvard. sented here, we arbitrarily chose a cut-off of 34 edu/treebase/index.html). steps, so brs values above that are simply reported Alignments were provided by the contributors as "> 34." except for the ITS data, for which the alignment To assess robustness of the results to choice of was constructed by EAK, beginning with an initial markers, each data set was analyzed by itself. The alignment in ClustalW (Thompson et al., 1994) and morphological data set was omitted from one anal- then continuing by eye. It became apparent that ysis, and the chloroplast data were analyzed sepa- ITS1 could not be aligned reliably across the fam- rately, as were the nuclear data. For analyses of ily, so it was omitted from the data set. Later after individual data sets, PAUP* was set to perform extensive data exploration, several regions of am- heuristic searches using maximump arsimony,g aps biguous alignment were also omitted from ITS2. were coded as missing data, multistate taxa were Gaps were treated as missing data. A few indels, coded as uncertain, and starting trees were ob- identified as phylogenetically informative in anal- tained by ten random addition sequences, holding yses of individual data sets, were coded as binary one tree at each step; branch swapping used tree characters and included in the structural data ma- bisection and regrafting (TBR), steepest descent trix (Appendix II). was not in effect, and MULPARS was in effect. Bootstrap analyses of individual data sets were DATA ANALYSIS done to facilitate comparisons with combined anal- yses. All bootstrapso f individual data sets included Data were analyzed by parsimony algorithms, as 500 bootstrap replicates; MAXTREES was set to implemented in PAUP*4.0 d64 (Swofford, 1998) on 500 to minimize times for searches. a Power Macintosh G3, and Nona (Goloboff, 1993) Analyses conducted with Nona ver. 1.6 (Golo- on an Intel-chip-based workstation running Win- boff, 1993) used the default settings amb- (clades dows NT. Data sets were analyzed individually by resolved only if they have unambiguous support) JID, LGC, EAK, and HPL to be sure that e-mail and poly= (polytomies allowed). Tree searches in- transmittal of such a large file had not introduced volved 1000 Wagner tree initiations using random any errors (for which we suggest the term "network- taxon entry sequences, followed by tree bisection induced homoplasy"). Numbers of informative reconnection (tbr) swapping with up to 20 most- characters and tree lengths were the same for the parsimonious trees retained in each search (hold/ two programs, although in some cases the number 20, mult*1000); shortest trees retained from the of equally parsimonious trees differed because of subsearches were then tbr-swapped to completion, Volume 88, Number 3 Grass Phylogeny Working Group 387 2001 Phylogeny and Classification of Poaceae with up to 10,000 trees held in memory (holO000, different from the model specified for sequence max*). Structural character autapomorphies of ter- data. Calculation of base frequencies and transi- minals, and synapomorphies of clades, were deter- tion/transversionr atios would be meaningless. Sev- mined by optimizing the morphological data on eral neighbor-joining analyses were done with mor- most-parsimonious trees obtained by the various phological and restriction site data omitted, but this analyses, using Winclada ver. 0.9.99m6.1 (Nixon, also required omitting several taxa for which dis- 2000). Strict-consensus bootstrap frequencies for tances were then undefined because of missing just the total evidence analysis (see Soreng & Da- data. By the time data sets and taxa were omitted, vis, 1998) were computed with Clados ver. 1.9.95 the results were difficult to compare to those of par- (Nixon, 1993) running Nona (Goloboff, 1993) as a simony algorithms. Several maximum likelihood daughter process for the tree searches, using a copy analyses were also undertaken on sequence data of the data set from which uninformative characters alone. These did not reach completion even after had been removed (with the "mop" function of Win- three to five days of analysis time. As with the clada). One thousand bootstrap replicates were con- neighbor-joining analyses, missing data and differ- ducted, using the same ambiguity and polytomy ent models of evolution for the different genes made settings as in the basic analyses. Each replicate the results of questionable validity. consisted of 10 random taxon entry sequences fol- lowed by tbr swapping with up to 10 trees retained RESULTS from each subsearch (ho/10, mult* 10), and with further tbr swapping then conducted on the result- Consensus trees for analyses of the individual ing trees from the 10 subsearches, with 101 trees data sets are presented in Appendix III-A to H and held (ho 101, max*). tree statistics in Tables 3 and 5. Note that the taxa Uninformative characters were excluded for all included are generally selected from more compre- analyses, so all tree statistics reported in this paper hensive analyses that have been published else- (consistency index [CI] and retention index [RI]) where, as described in Materials and Methods. reflect only potentially phylogenetically informative Many of the trees differed in topology, but in no characters. case was a strongly supported group in one tree Chloroplast and nuclear trees were compared us- contradicted by a strongly supported group in an- ing the incongruence length difference test (random other tree. We interpreted this as lack of significant partition test of Farris et al., 1994), as implemented conflict. Nonetheless, the ILD test indicated signif- in PAUP*. They were also compared using simple icant differences between the nuclear and chloro- inspection, as recommended by Mason-Gamer and plast data sets, between nuclear protein-coding and Kellogg (1996). To compare tree topologies, con- chloroplast, between ndhF and phyB, and between straint trees were constructed as necessary in ndhF and rbcL. These differences persisted in most MacClade (Maddison & Maddison, 1993); these cases even when taxa with conflicting placements were then loaded, the constraint enforced, and a were removed. In the only exception to this obser- heuristic search undertaken using the same param- vation, ndhF and phyB were not significantly dif- eters as in unconstrained searches. ferent if the PACCAD Clade was reduced to Pani- The combined data were constrained to fit to- coideae, Chloridoideae, and the clade of Molinia pologies suggested in previous studies by loading a plus Phragmites. This provides weak evidence that constraint tree in PAUP* and then searching for differences in resolution of the PACCAD Clade the most parsimonious tree compatible with that (Panicoideae, Arundinoideae s. str., Chloridoideae constraint tree. Constrained and unconstrained s.l., Centothecoideae, Aristidoideae, Danthonioi- trees were compared using the Wilcoxon signed deae) are partly responsible for the significant dif- ranks test (WSR) as suggested by Templeton (1983) ferences. Differences between ndhF and rbcL, how- and implemented by Mason-Gamer and Kellogg ever, are puzzling because both are part of the same (1996). Significance values were determined using linkage group. Because of the ambiguity of the re- a two-tailed test. sults, we did not attempt to do all possible pairwise The entire data set could not be analyzed with comparisons of trees. Despite the differences in the neighbor-joining or maximum likelihood algo- data sets, we chose to combine the data in a single rithms. The inclusion of morphological and restric- analysis. Different histories for the various genes tion site data with sequences made it nonsensical remain a formal possibility. However, in other in- to specify a single model of evolution. While a mod- vestigations we have seen that the ILD test may el could in principle be hypothesized for morpho- return significant differences if there is extensive logical or restriction site data, it would have to be missing data (as we have in some data sets here) CAD o Table 5. Bootstrap support values for subsets of the total data matrix. Numbers of nodes at particular support values are given as fractions of the total number of nodes, and also as decimals. Poly = polyphyletic; para = paraphyletic. Anom. = Anomochlooideae; Phar. = Pharoideae; Puel. = Puelioideae. *Panicoideae helre include Danthoniopsis;i f it is excluded then support values are much higher. Molec. Chloro- Total data data plast Nuclear Structural ndhF cprs rbcL rpoC2 phyB GBSSI ITS # Nodes 100 27/64 25/64 22/64 2/55 0 23/63 2/43 8/35 2/32 4/38 3/17 1/45 (0.42) (0.39) (0.34) (0.04) (0.36) (0.05) (0.23) (0.06) (0.11) (0.18) (0.02) # Nodes 90-99 9/64 14/64 14/64 12/55 1/64 11/63 8/43 7/35 6/32 13/38 4/17 5/45 (0.14) (0.22) (0.22) (0.22) (0.02) (0.17) (0.19) (0.20) (0.19) (0.34) (0.24) (0.11) # Nodes 70-89 11/64 8/64 7/64 5/55 4/64 4/63 9/43 6/35 4/32 10/38 1/17 4/45 (0.17) (0.13) (0.11) (0.09) (0.06) (0.06) (0.21) (0.17) (0.12) (0.26) (0.06) (0.09) Fraction nodes > 70 0.73 0.73 0.67 0.35 0.08 0.59 0.45 0.60 0.37 0.71 0.48 0.22 Poaceae 100 100 100 97 Para 100 98 99 Not tested 96 Not tested Not tested Spikelet Clade 100 99 98 87 Para 94 81 98 Not tested 54 Not tested 83 Bistigmatic Clade 100 100 100 97 <50 100 <50 Not tested Not tested 79 100 98 0on 3 BEP + PACCAD Clade 100 100 99 81 Poly 100 Not tested Para Not tested 75 Not tested Not tested c- (0 Bambusoideae 97 98 97 Para Poly 100 Para Para Para 92 Poly 75 M* o Ehrhartoideae 100 100 100 92 Para 98 72 100 51 99 Not tested 54 Pooideae 100 100 93 95 Para 88 <50 70 88 94 77 19 - . :3 BEP 71 90 62 50 Para 53 Para Para 56 89 Poly Para 0 Aristidoideae 100 100 100 84 84 100 Not tested 98 67 Not tested Not tested 76 Chloridoideae 86 83 86 51 Para <50 98 54 63 93 95 Poly G) Panicoideae* 65 Poly Para 63 Para Poly Para 95 93 99 61 Poly QO Danthonioideae 98 97 98 Poly Para 99 Not tested 83 70 Not tested CD Poly Poly PACC 100 100 99 77 <50 100 95 73 73 90 Poly <50 Arundinoideae s. str. 77 78 <50 Poly Para <50 Not resolved 23 Poly 52 Not tested Poly Centothecoideae Para Poly Para Para Poly Para Not tested 64 Para 98 Not tested Para Volume 88, Number 3 Grass Phylogeny Working Group 389 2001 Phylogeny and Classification of Poaceae or if a single terminal taxon differs in its placement ported at bts 86 (brs 8). A clade corresponding to (Z. Magombo, pers. comm.). Because of the lack of Arundinoideae s. str.-Arundo, Amphipogon, Mol- obvious points of conflict between the data sets, and inia, and Phragmites-receives modest support because of the clear congruence at the deep nodes from this analysis (bts 77, brs 6), but the sister with which we are concerned, we interpret the sig- relationship of Molinia and Phragmites is well sup- nificant ILD tests as misleading. ported (bts 100; brs 16). The other major clade (the Analyses of the complete data set were faster BEP Clade) is less well supported (bts 71; brs 8) than analyses of many of the individual data sets, and includes Bambusoideae s. str., Ehrhartoideae as has been found in studies of other large data (= Oryzoideae), and Pooideae. Bambusoideae are sets (Soltis et al., 1998). For example, a heuristic monophyletic (bts 97; brs 15), as is the clade in- search of the complete data set in PAUP* on a cluding the herbaceous bamboos (bts 100; brs 18). Macintosh G3 with 10 random addition sequences Likewise Ehrhartoideae are monophyletic (bts 100; took 19.6 seconds. brs 24), as are Oryzeae (bts 100; brs > 34). Pooi- With all data combined, there were 2143 parsi- deae include Brachyelytrum (bts 100; brs 15), and mony informative characters, which produced a sin- most nodes within the pooid clade are strongly sup- gle tree of 8752 steps, consistency index (CI) of ported. 0.375, and retention index (RI) of 0.557 (Figs. 1 Despite the strong phylogenetic pattern shown by and 2). Bootstrap analyses (1000 replicates) indi- the combined analysis, placement of some taxa re- cated that 27 branches were supported in 100% of mains ambiguous. The major uncertainty remains the bootstrap replicates, 9 branches in 90-99%, the monophyly of the BEP Clade. As noted earlier and 11 branches in 70-89% (Table 5). Put another (GPWG, 2000), it is almost equally parsimonious way, of 64 internal nodes, slightly more than half to place Pooideae as sister to the PACCAD Clade, (36) have bootstrap values over 90% and a clear and this makes evolution of particular morpholog- majority (47) have values over 70%. Bootstrap val- ical characters more parsimonious. The Pooideae ues were virtually identical whether done using plus PACCAD group appears in analyses of rbcL strict consensus bootstrap in Nona (Goloboff, 1993) (Appendix III-C), chloroplast restriction sites (Ap- or the frequency-within-replicates bootstrap in pendix III-A), morphology (Appendix III-H), and PAUP*4.0 (Swofford, 1998); for comparison with ITS (Appendix III-F), whereas the BEP Cla(le is individual analyses, we report the values from retrieve(l by analyses of nidhF (Appendix III-B), PAUP*4.0. rpoC2 (Appendix Ill-I)), and phyB (Appendix III- The analysis of the combined data confirms many E). GBSSI (Appendix III-G) forms a novel topology, results of previous studies and clarifies some rela- in which neither the PACCAD nor the BEP Cla(les tionships that were previously ambiguous. The two is monophyletic. An analysis combining rbcL, chlo- species of Restionaceae form a clade. Joinvillea is roplast restriction sites, ITS, and morphology re- sister to a monophyletic Poaceae. The three earliest trieves, not surprisingly, a clade that links the Pooi- diverging lineages are the Anomochlooideae, Phar- deae with the PACCAD Clade. Bootstrap analysis, oideae, and Puelioideae, in that order, together ac- however, finds that the Pooideae + PACCAD clade counting for 30 species of grasses. The vast major- occurs in only 23% of the replicates, although it ity of extant grasses fall into two distinct lineages. appears in 40% if Streptogyna is considered part One of these is the PACC clade (Davis & Soreng, of the clade. The BEP Clade was not found in any 1993), here called the PACCAD Clade (Panicoi- of the bootstrap partitions. deae, Arundinoideae s. str., Chloridoideae s.l., Cen- Constraining the entire data set to place Pooi- tothecoideae, Aristidoideae, Danthonioideae) to re- deae sister to the PACCAD Clade resulted in a sin- flect the inclusion of two additional subfamilies gle tree eight steps longer than the most parsimo- within the clade. Within this clade, Panicoideae s. nious tree. The net change of eight steps, however, str. (excluding Danthoniopsis) are monophyletic (bts was produced by changes of one or two steps in 94; brs 10), as are the core Paniceae sampled here 107 characters from throughout the data set. A Wil- (bts 100; brs 25) and Andropogoneae (bts 100; brs coxon signed rank test (Templeton, 1983; Mason- 32). Other strongly supported groups in the PAC- Gamer & Kellogg, 1996) resulted in a test statistic CAD Clade correspond to Aristidoideae (bts 100; of 2654; for n = 107, this corresponds to p < 0.406 brs 25) and Danthonioideae (bts 98; brs 15). The (two-tailed test). This means that we cannot rule out traditional Chloridoideae are supported at bts 99 the possibility that Pooideae are indeed sister to (brs 16), and the clade including the chloridoids the PACCAD Clade. This is true even if the mor- plus Centropodia glauca and Merxmuellera rangei phological characters are excluded (z = 1.146; P (Chloridoideae s.l.) is also reasonably well sup- < 0.254). 390 Annals of the Missouri Botanical Garden 118 Flagellaria 115 1316 36 I Elegia 1113 13 aloskion 94 Joinvillea 106 134 140 Volume 88, Number 3 Grass Phylogeny Working Group 391 2001 Phylogeny and Classification of Poaceae Flagellaria 100 I Elegia >34 Baloskion Joinvillea 100 Anomochloade d00 1 5 StreptochaetaJ 100 Anomochlooideae Pharus -Pharoideae >34 100 Guaduela Puelioiddeeaaee 32 -- - Puelia 1100 100 Eremitis >34 73 4 Pariana 100 4 Lithachne 97 1 1 180 6 4 36 314 Olyra Bambusoideae Buergersiochloa 10 0 3 6 1 5 628 I Pseudosasa 127 2 Chusquea 140 | Streptogyna - Incertae sedis 2 l oO{ Ehrharta 7 Ehrhartoideae 24 >34r Oryza >34 Leersia 53 Phaenosperma 28 1 Anisopogon I-00 Ampelodesmos 100 701 8 299 77 Stipa 12 60 Nassella 3 >34 2 Piptatherum 1 93 Brachypodium 882 96 9 no~o9 ^4i Avena Bromus Pooideae 7 10 >34 Triticum 88 88 Diarrhena 100 Melica 100 >34 E Glyceria 15 100 Lygeum >34 Nardus 100 Brachyelytrumn Arstida Aristidoideae 61 25 Stipagrostis A tidodeae ~988a Q I Merxmuellera m. 1558 100 Karroochloa 1 >001 Austrodanthonia Dant honioideae 100 41 27 Danthonia 20 79 -- Amphipogon 3 7J 7L' Arundo LQ AMro linia Arundinoideae 76 2 32 9 16 _P hragmites 3 9 8 Merxmuellerar . 3 11 Centropodia 27 86 100 - 7 Eragrostis 3 8 4 9 0 CUhnlioorlaU(Pno ld 99 4 Pappophorum Chloridoideae 100 Zoysia 1 85 7 196 Spartina 25 4 7 ' Sporobolus 2 Distichlis Eriachne - Incertae sedis 65 - 68 Thysanolaena 3 L- Zeugites Centothecoideae 8 7 3 10 Chasmanthiumn_ ?8 8 Gynerium Incertae sedis 20 868 68 3 65 100 Danthoniopsis Panicum Panicoideae 2 194 25 Pennisetum 10 O|--00 Miscanthus 32 -Zea Micraira - ncertae sedis Figure 2. Same tree as Figure 1, showing percent of bootstrap replicates above lines and Bremer support below. Brackets indicate the revised classification for the Poaceae. Figure 1. Single most parsimonious tree for the grasses and relatives, based on eight sets of data. Length = 8752 steps, CI = 0.375, RI = 0.557. Numbers above branches are numbers of unambiguous changes. Branches are drawn proportionalt o length. 392 Annals of the Missouri Botanical Garden The combined analysis places Streptogyna as firmly resolved by the combined data set, although sister to Ehrhartoideae, but this result is not strong- both are clearly members of the PACCAD Clade. ly supported (bts 40; brs 2). This partly reflects This almost certainly reflects missing data. In ad- missing data, in that only ndhF and phyB sequenc- dition to morphological data, Eriachne is repre- es are available for Streptogyna, in addition to mor- sented only by rbcL and ITS sequences, and Mi- phological data. ndhF places S. americana sister to craira by ndhF, rpoC2, and ITS. They are both Ehrhartoideae, whereas phyB places it as sister to isolated taxa, and in individual analyses fall at the the entire BEP Clade, and morphological data fail base of other well-supported clades. The position to resolve its position. of Micraira as sister to the entire PACCAD Clade The combined data place the woody bamboos, appears only in the combined analysis, and like- Pseudosasa and Chusquea, in a clade (bts 68; brs wise the position of Eriachne as sister to the Arun- 2), as would be expected from previous studies dinoideae s. str. + Chloridoideae + Aristidoideae (Zhang, 2000; Zhang & Clark, 2000). The pair ap- + Danthonioideae clade is both novel and poorly pears monophyletic in chloroplast restriction site, supported. Bootstrapa nalysis of the combined data morphological, and phyB trees. The two are para- set placed Eriachne and Micraira as sisters in 51% phyletic, however, in trees using rbcL and ndhF, of the 1000 replicates, a position not supported by although this result is not well supported in these the most parsimonious tree. trees. "Pseudosasa" is a composite taxon, made up Aristidoideae and Danthonioideae are both of data from several different genera, and this may clearly monophyletic, and each is strongly sup- also affect its placement in the combined tree. ported by both bootstrap and decay analyses (bts Phaenosperma and Anisopogon are clearly mem- 100 and 98; brs 25 and 15, respectively). In the bers of the expanded pooid clade, where they are combined tree they appear as sister taxa. The ar- placed by all data sets, either singly or in combi- istidoid/danthonioid clade is not stronlglys upport- nation. Their position within the clade, however, ed, however (bts 61; brs 8), and is reflected only in remains uncertain. They are sister taxa when all the phyB tree. rbcL places Aristidoideae sister to data are combined, but this result is not strongly Chloridoideae, whereas ndhF and cllloroplast re- supported (bts 53; brs 1); together they are sister striction sites put Aristidoideae sister to the rest of to the Stipeae, also a poorly supported result (bts the PACCAD Clade, and ITS places the subfamily 28; brs 1). In ndhF, chloroplast restriction site, and sister to Amphipogon + Chloridoideae. rpoC2 in- phyB trees, Anisopogon is placed on a branch that dicates that Aristidoideae is derived from within diverges after the Lygeum + Nardus clade, but be- Arundinoideae. ndhF places Danthonioideae sister fore the rest of the Pooideae (i.e., Stipeae, Meli- to Panicoideae + Centothecoideae, whereas chlo- ceae, Diarrhena, Brachypodium, Aveneae, and roplast restriction sites do not resolve the position Poeae). ITS places it sister to Aveneae/Poeae, and of Danthonia. GBSSI retrieves a novel arrangement rpoC2 places it sister to Stipa. In no case is the in which Danthonioideae are polyphyletic, but this placement strongly supported. The position of result is not strongly supported and is likely af- Phaenosperma is based only on ndhF and morpho- fected by skewed taxon sampling in the GBSSI data logical data, and the latter are largely uninformative set. rpoC2 suggests that Danthonioideae are sister about its position. to Amphipogon. Meliceae are monophyletic in all gene trees and The relationships of Zeugites, Thysanolaena, in the combined tree. Their position, however, Chasmanthium, Danthoniopsis, and (ynerium to varies among the individual gene trees. The com- each other and to the Panicoideae are not resolved bined tree provides good evidence that Meliceae by this analysis. The entire group is well supported diverged after Lygeum + Nardus (bts 82; brs 7), as monophyletic (bts 87; brs 8), but other relation- but evidence is weak that it was the next diverging ships are less clear. Only morphologicala nd ndhF branch (bts 29; brs 1). Other possible placements data are available for Zeugites, so its placement include sister to Stipeae (ndhF, phyB), sister to may be affected by missing data. Diarrhena + Brachypodium + Aveneae/Poeae (cp The morphological data have little effect on the restriction sites), sister to a clade of Brachypodium analysis. When they are omitted, 6 trees are found + Brachyelytrum + (Lygeum + Nardus) (ITS), or in two islands (length 8488, CI = 0.378, RI = paraphyletic at the base of the Pooideae (GBSSI). 0.554). The topology of the strict consensus (Ap- The ambiguity cannot be ascribed to missing data, pendix III-K) is similar to that of the combined tree although additional sampling among early-diverg- except for the position of Zeugites, which is sister ing Pooideae might be warranted. to Danthoniopsis, and Gynerium,w hich is sister to The positions of Eriachne and Micraira are not Panicoideae. In the consensus of the six trees, the Volume 88, Number 3 Grass Phylogeny Working Group 393 2001 Phylogeny and Classification of Poaceae relationship of Pseudosasa and Chusquea is unre- Differences appear only in the placement of the solved, as is the relationship of Phaenosperma and Meliceae, Eriachne, Micraira, Zeugites, and Dan- Anisopogon, and the position of Meliceae in the thoniopsis, all poorly supported areas of the trees. pooid clade. These are areas that were poorly sup- ported even in the combined tree, and thus already DISCUSSION known to be ambiguous. The most notable differ- WELL-SUPPORTED CLADES ence is the increased support for the BEP Clade, Some relationships appear consistently in all which is supported at a bootstrap value of 90%. analyses of all data sets and are strongly supported The number of nodes with support greater than by the combined analysis. Among these are the fol- 90% (Table 5) is somewhat greater without the lowing (in order from the bottom of the tree): structural data, but the overall consistency index is 1. Joinvilleaceae are sister to Poaceae. not changed appreciably (Table 3). 2. Poaceae are monophyletic. The results for the entire data set largely reflect 3. The earliest diverging lineage of Poaceae is the results for the chloroplast data alone. The chlo- Anomochlooideae (even if Anomochloa and Strep- roplast data contribute 1336 potentially phyloge- tochaeta prove to be two separate lineages, they netically informative characters, or about 62% of would still be the two earliest-diverging lineages in the total. Analysis of these data alone produces 2 the family). trees (length = 4903, CI = 0.381, RI = 0.589) 4. The next diverging lineage is Pharoideae. that differ only in the relative positions of Phae- 5. The next diverging lineage is Puelioideae. nosperma and Anisopogon (Appendix III-I). The 6. All remaining grasses form a clade, which ap- numbers of strongly supported nodes are about the pears to have diversified well after the origin of the same as for the entire data set (Table 5), and dif- family. ferences between the chloroplast tree and the entire 7. Bambusoideae s. str., Ehrhartoideae, Pooi- data set are all in poorly supported areas of the tree deae s.l., Aristidoideae, Danthonioideae, Arundi- (see below). noideae s. str., Chloridoideae s. str., Chloridoideae The chloroplast tree is only slightly affected by s.l., and Panic oideae are all monophyletic. mixing sequence data with restriction site data. If 8. Bambuseae, Parianeae, (lyreae s. str., Ory- the restriction site data are exclulded so that the zeae, Stipeae, Meliceae, and LygeiLm + Nardus, data set consists only of ndh F, rbcI , and rpoC2 and Molinia + Phragmnites are all monophyletic. data, the tree is virtually identical to the chloroplast 9. The PACCAI) Clade- now including Pani- tree except that Pseiidosasa plus Chusquea, an(l coideae, Arundinoideae s. str., Chloridoideae s.l., Phaenosperma plus Anisopogon form monophyletic Centothecoideae, Aristidoideae, Danthonioideae, pairs rather than being paraphyletic (not shown). Eriachne, Micraira, and Gynerium-is monophylet- Piptaltherum and Nassella are paraphyletic rather ic. than sisters, and the Meliceae are sister to the core As noted in the introduction, all of these rela- Pooideae rather than to the Stipeae. tionships have been supported by previous studies Analysis of only the three nuclear genes (phyB, and none is unique to the combined analysis. Pre- GBSSI, and ITS) required elimination of nine taxa vious studies, however, were limited because they for which nuclear data were not available. The were based on a single gene, a modest number of analysis thus included 57 taxa and 757 characters morphological characters, and/or a restricted sam- and found eight trees (length = 3513, CI = 0.382, ple of taxa. Because of the strong support for the RI = 0.512) on one island (Appendix III-J). The relationships found in the present study, we propose nuclear trees were not as well supported as the a revised subfamilial classification (see Taxonomic chloroplast tree or the tree for the entire data set, Treatment). The revisions primarily reflect changes which presumably reflects extensive missing data in circumscriptions of the Bambusoideae and Arun- for GBSSI, and a generally smaller number of in- dinoideae and involve only a small fraction of the formative characters. Only two nodes were sup- species in the family. Over three quarters of the ported in 100% of the 1000 bootstrap replicates, species are included in the subfamilies Pooideae, and 11 had values between 90 and 99%. Chloridoideae, and Panicoideae, the circumscrip- Analysis of chloroplast data plus the data from tions of which are changed only slightly by the re- the two protein-coding nuclear genes (that is, ex- visions. cluding morphological and ITS data) has little ef- fect on either topology or for the tree, MOLECULAR CHARACTERS support per- haps because omitting morphology and ITS only Virtually all of the phylogenetic signal in this eliminates 178 characters, or about 8% of the total. analysis comes from the molecular data (Appendix 394 Annals of the Missouri Botanical Garden III-K), as expected. The combinable component was difficult and confirmed our suspicion that it consensus (Bremer, 1990) of the molecular trees may not be useful at this level of divergence. ITS1 (Kellogg, 1998) is remarkably well resolved; all and parts of ITS2 had to be eliminated because of nodes found in this consensus are strongly sup- difficulty in assessment of sequence similarity. The ported in the combined analysis presented here. rpoC2 sequences used here code for repeated ami- When the molecular data are analyzed alone, all no acid motifs inserted into the protein. The inser- strongly supported nodes from the combined anal- tion appears only in the grasses and thus consti- ysis are recovered, and support for the BEP Clade tutes a synapomorphyf or the family (Cummingse t is increased. al., 1994), although we did not code it as such in Previous theoretical (Graybeal, 1998) and em- this analysis. The repeats are similar but not iden- pirical (Soltis et al., 1998) studies have indicated tical to each other in sequence, making alignment that large numbers of characters may be necessary problematical, a point discussed at length by Bark- to resolve phylogenetic patterns, a conclusion only er (1995) and Barker et al. (1999). Efforts to im- partially supported by this study. The molecular prove alignments necessarily reduce apparent ho- data alone and the entire data matrix included moplasy; this may result in the high CI mentioned 2093 and 2143 phylogenetically informative char- above. Although phylogenetic results from these acters, respectively (Table 3). These data sets found molecules are similar to those from the other genes, the largest percentage of nodes with bootstrap val- by themselves the two sets of sequences do not per- ues above 70% (0.73 in both cases) but did not mit confident assessments of relationships among have the highest consistency or retention indices. subfamilies. The highest CI was produced by the rpoC2 data Trees from ndhF and phyB, individlually,a re par- alone (150 informative characters), although this ticularly well resolved and well su,pported.T heir could be in part an artifact of alignment (see be- congruence in early-divergingb ranches contributes low), and the highest RI by the rbcL data alone (213 to the strength of the overall topology of the com- informative characters). The fraction of nodes with bined data. In particular,p hyB provi(les consider- bootstrap values over 70% was almost as high for able support for the BEP Clade, a to,,ology that is the phytochrome B data alone (with 417 informative only weakly supported by ndhF, an(l not at all by characters) as for all data combined. We conclude several other data sets. The two data sets do appear that, while large numbers of informative characters to conflict in relationships among members of the may provide increased reliability, small numbers PACCAD Clade, and this may be an area for future are not necessarily misleading or inaccurate sam- investigations. ples of the whole. As noted in Methods, sequences for a given ge- Other studies have shown that the number of nus were in some cases taken from (lifferent, con- taxa included may affect phylogenetic accuracy generic species. This procedure assumes that the (e.g., Hillis, 1996, 1998; Graybeal, 1998), although genus is monophyletic, an assumption that is almost this is not necessarily the case (Poe & Swofford, certainly correct in many cases (e.g., Joinvillea, 1999). Certainly future studies should include more Streptochaeta),a nd perhaps not as likely in others. taxa than just the set of exemplars used here. How- For example, the three species of Stipa sampled ever, the results numbered 1 to 9 above have been here have been placed in the genera Achnatherum, found in analyses of virtually every individual data Stipa, and Jarava (Barkworth & Fverett, 1987; set, as well as in the combined tree, and we would Barkworth, 1993; Jacobs & Everett, 1997), which be surprised if they were overturned by inclusion are distinct and possibly not a monophyletic group of more taxa. within Stipeae (Jacobs et al., 2000). While this Most of the molecular data come from chloroplast problem is not likely to compromise our conclu- genes, so it is not surprising that the tree from the sions regarding subfamily relationships, it means chloroplast alone closely matches the tree for the that relationships among species of the Stipeae (or entire data set. The ndhF data set is missing the other tribes or genera where composite terminal least data (Table 3) and has the most informative taxa were used) cannot be addressed by this anal- characters of the molecular data sets, presumably ysis. because it is the longest molecule. Our results con- firm the utility of this molecule for resolving rela- STRUCTURAL CHARACTERS tionships among grass genera (Clark et al., 1995; Giussani et al., in press). The structural characters, comprising the mor- Alignment is a particular problem for the rpoC2 phological data set (Table 4), were optimized on the and ITS data used here. Alignment of the ITS data phylogeny (Fig. 3). Our results suggest that some Volume 88, Number 3 Grass Phylogeny Working Group 395 2001 Phylogeny and Classification of Poaceae 3 3147 Flagellaria 3053 Joinvillea 31 2 242938 2 28424345464751 A4 n hAnomochloa 354547 11111141 0-Streptochaeta 2 9 43 1 1 1112 'Puelia 1EP 221 020000 2 122135374050 1o-AooCoPACCAD 0000131 Figure 3 (pp. 395-397). Same tree as Figure 1, with structuralc haracters mapped on using ACCTRANo ptimization. Charactern umber is above the branch, and the state to which the character changes is below. Filled circles represent unique occurrences of character states; open circles represent homoplasies. of the morphological characters may be useful for or not), 11 (presence or absence of proximal re- delimiting groups within tribes or subfamilies, but duced flowers), 12 (number of flowers per spikelet), are too variable to be useful in delimiting subfam- 13 (presence or absence of awns), 15 (attachment ilies; these include 2 (culms hollow or not), 3 (leaf of awns), 16 (disarticulation above or below sheath margins fused or free), 10 (pedicel present glumes), 20 (lodicule fusion), 28 (style fusion), 29 396 Annals of the Missouri Botanical Garden 5 1215253045 1 0 1 0 Streptogyna 35 /r />-~Ehrharta 111923242846 0 17 1 2 1 1 0 0 41 Oryza OO 23242530 194546 Leersia pOO- o0003 13 1 1 12232429 1 5 40 Pseudosasa 0113 2 1147 13 1 -0 Chusquea I 0 r-Buergersiochloa \_ 17 gLithachne 13 1 A (-_ 0-- "13 Olyra 0 I 2529 1630 Eremitis 0 1 01Y 2324 01 Pariana 2 4047 0- Brachyelytrum 2 1344 10182943 ....0-0 / 01 -. Lygeum j28364347 /OOO-O-( 7123047 0000 0011 13 Nardus 1 1 3 0000 1 I 31647 3 13202130 -Melica \ 122238 00 0/ 1 1 o - Glyceria 13 Diarrhena L, r~-B rachypodium 404,7 153940 07 21 Avena 3 35 Bromus Triticum 0 7 131640 14153638 Phaenosperma 2019 47 1 Anisopogon 2 12 12 3 o Ampelodesmos 25 _ Piptatherum 1940 Stipa 2 0 -Nassella Figure 3. Continued. Volume 88, Number3 Grass Phylogeny WorkingG roup 397 2001 Phylogeny and Classificationo f Poaceae 182547 o0o-- Micraira 1126 xO-- Chasmanthium 2835 1 0 1 5 2845 21 1 712223740 1 Thysanolaena 11 0 01 2 33 Zeugites 1 2 212545 2 122135374050 0000 00- ,OOO0-- Gynerium 0000131 4 133340 {?? Danthoniopsis 2 134047 1121748 9 1747 Miscanthus 1111 010 2830 16 Zea 48 28 ? 15 ^ Panicum 0 ~ Pennisetum 48 [*-Eria rchne 12174748 Aristida 48 --12 4\t~~ 1421 Stipagrostis 4 1340 31 Merxmuellera macowanii 1532 47 21 4Danthonia ( 2 1 47 111 1~\-30K3 arroochloa 1 Austrodanthonia 47 0 33 o olinia (2V 112845 2 1 1 1 Phragmites ~ 1217222844 / Amphipogon 1 A0 3 Arundo 33 4348 13- I 1547 Merxmuellera rangei 1 6 Centropodia 13 35748 1- 14 Pappophorum 103 20 3 Eragrostis 112835 1344 Uniola 1J'J1 10 0 1 4 40 Distichlis . / 0 0,3,4 13 _ o--Zoysia 1012161849 / 1 2 28 01001 o-Spartina 101647 111 9 Sporobolus 9 Figure 3. Continued. 398 Annals of the Missouri Botanical Garden (number of stigmas), 30 (highest order of stigmatic flower and spikelet affects characters 8, 18, and 19. branching), 39 (lipid in the endosperm), 40 (en- C4 photosynthesis (chars. 48 and 49) is known to dosperm starch grain syndrome), and 47 (base be a set of characters that do not co-vary; its origin chromosome number). is still poorly understood. These characters are dis- For other structural characters, our assessment cussed in more detail below. of character states is inadequate for use in phylo- genetic analysis. For example, the woodiness of SPIKELETS AND FLOWERS bamboo culms (char. 1) is due to numerous, closely spaced vascular bundles around the periphery of Spikelet. The flowers of most grasses are ar- the culm, each bundle with massive sclerenchyma ranged in bracteate units known as spikelets. These fiber caps on both sides, combined with heavily usually consist of two bract-like appendages sclerified ground tissue, but isolated fiber strands (glumes) at the base of a central axis (rachilla) on may also occur (Soderstrom,1 981; Liese, 1998). It which are borne one or more florets, all in a disti- is not clear if the hardness of culms in Arundo, chous pattern. According to Clayton (1965: viii), Thysanolaena, and Gyneriuma nd a number of other the "grass spikelet never ceases to fascinate, for the "woody" taxa of the PACCAD Clade (Watson & simplicity of its theme is matched by the elegance Dallwitz, 1992) is derived in the same way; these of its variations." The apparent simplicity of the taxa need more careful anatomical study. Similarly, grass spikelet notwithstanding, its origin and ho- the fringed membranous ligule of some Pooideae mologies, as well as those of the grass flower, have (char. 4) may not be the same as the ciliate mem- been much debated (see reviews in Clifford, 1987, brane found in many members of the PACCAD and Soreng & Davis, 1998). Clade and requires developmental investigation. The phylogeny highlights the difficulty of dis- Spikelet pairing (char. 9) appears in various pani- cussing the evolution of the grass spikelet. The coid grasses and also in Pharus, but the pattern of standard spikelet (or some modificationo f it) is pre- development of the pairs in Pharus is unknown. sent in all members of the Pooideae and the PAC- Membranousl odicules (char. 21) are scored as be- CAD Clade (Figs. 4 and 5). 'The pattern of bracts ing the same wherever they occur, but development and flowers, however, is variable among Pharo- of these structures has never been compared. Haus- ideae, Puelioideae, Bambusoideae, and Ehrharto- torial synergids are apparently uniquely derived ideae (Fig. 4). Pharoideae bear single flowers, each within the Danthonioideae, but many taxa have not with a lemma, a palea, and a pair of glumes, but been investigated for this character. Embryo char- no rachilla extension (Fig. 4B; Judziewicz, 1987; acters (35, 36, 37, 38) are often phylogenetically Soderstrom et al., 1987). Puelioideae have multi- informative, but lack of observation of critical taxa flowered spikelets, in which each flower has a lem- makes their use difficult in some cases. Starch ma and palea, and the whole unit has a pair of grains (char. 40) are classified according to their glumes (Fig. 4A). Proximal incomplete florets oc- apparent structure when viewed with the light mi- cur, but distal reduction is seen only in Guaduella. croscope. With much recent work done on the bio- In Bambusoideae, the unisexual, one-flowered chemistry and molecular genetics of starch granule spikelets of Olyreae are standard, but the bracteate, formation (e.g., Whistler et al., 1984; Frazier et al., rebranching spikelets (pseudospikelets) of many 1997), this character could and should be recir- Bambuseae are difficult to interpret (Fig. 4F; Jud- cumscribed, although then much scoring will need ziewicz et al., 1999). The multiflowereds pikelet of to be redone. Arm cells and fusoid cells (chars. 45 Streptogyna presents no difficulties of interpreta- and 46) are now seen to be ancestral in the family, tion, but in the Ehrhartoideae,e xtra proximal, ster- yet their development and ultrastructurea re poorly ile bracts (usually called sterile lemmas) are com- studied, and their physiological function is un- mon (Fig. 4G), and extreme reduction of glumes is known. Comparisonso f chromosome base numbers known (Oryzeae). It is not clear whether the prox- (char. 47) will almost certainly become more pre- imal sterile bracts of Ehrhartoideae are truly ho- cise because of recent studies of nuclear genome mologous to the proximali ncomplete floretst hat oc- arrangement( e.g., Gale & Devos, 1998). cur in Puelioideae and some bambusoids, or if they Finally, some of the structural characters are are phylogenetically (and possibly developmentally genuine morphological puzzles, ones for which and genetically) distinct. The uncertain position of strict comparison is difficult. These include the flo- Streptogyna makes the homology assessment even ral bracts (glumes, lemmas, paleas, lodicules), more ambiguous. which occur only in the grasses, and may not even Most characters of the spikelet and the floret occur in Anomochlooideae. Homology of the grass (chars. 9-12 and 14-17) are treated as inapplicable Volume 88, Number 3 Grass Phylogeny Working Group 399 2001 Phylogeny and Classification of Poaceae 1 mm 1 cm E F I 1 mm 1 C B Figure 4. Spikelets and spikelet equivalents of early-divergingl ineages and the BEP clade. -A. Puelia schuman- niana, Puelioideae (Letouzey1 2930, US). -B. Pharus mezii, Pharoideae (Hinton 16059, US; redrawnf rom Judziewicz, 1987). -C. Streptochaetas picata, Anomochlooideae (Bailey & Bailey 723, US; redrawnf rom Judziewicz & Soderstrom, 1989, from originals by A. Tangerini at US). -D. Anomochloa marantoidea, Anomochlooideae (Calderon 2046, US; redrawn from Judziewicz & Soderstrom, 1989, from originals by A. Tangerini at US). -E. Stipa comata, Pooideae (Pearson s.n., ISC). -F. Guadua chacoensis, Bambusoideae (Nee 35467, ISC). -G. Ehrharta bulbosa, Ehrhartoideae (Barker1 119, ISC). -H. Festuca idahoensis, Pooideae (Pohl 15642, ISC). 400 Annals of the Missouri Botanical Garden 1mmi I1 -0-? ,?i E 1 mm F G 2mm 1 mm immi C Figure 5. Spikelets of the PACCAD clade. -A. Arundo donax, Arundinoideae (Bradley & Sears 3558, ISC). -B. Dichanthelium oligosanthes, Panicoideae (Lelong 2063, ISC). -C. Aristida arizonica, Aristidoideae (Griffiths7 373, ISC). -D. Tridensf lavus, Chloridoideae (Thorne 18302, ISC). -E. Andropogon gerardii, Panicoideae (Clark s.n., teaching collection, ISC). -F. Danthonia californica, Danthonioideae (Pohl 9459, ISC). -G. Centotheca lappacea, Centothecoideae (Liang 66250, ISC). -H. Chloris cucullata, Chloridoideae (Malacara & Gutierrez3 3, ISC). Volume 88, Number 3 Grass Phylogeny Working Group 401 2001 Phylogeny and Classification of Poaceae when grass-type spikelets and florets are absent, as the structure still could be derived as a prophyll if in the non-grass outgroups (see char. 8). Although the axis rotated through 180? (Clifford, 1987). The pseudospikelets occur in some genera of Bambu- gemmiparousb racts of many Bambuseae are essen- seae, the two in the present study are regarded as tially glumes with a bud in the axil. If the bud having true spikelets and florets, and thus scorable develops it becomes a second- or higher order for features of these structures. pseudospikelet (Judziewicz et al., 1999). Glumes Neither Anomochloa nor Streptochaeta (Fig. 4C may be highly reduced or lost, as in Oryzeae. A and D) has structures clearly homologous with glume-like prophyll at the base of the pseudospike- glumes, lemmas, or paleas, and thus neither can be let is observed in many Bambuseae (Fig. 4F), al- described as having grass-type spikelets or florets. though the axis bearing the prophyll is not elon- We therefore follow Clark and Judziewicz (1996) in gated (McClure, 1966). using the term "spikelet equivalent" to refer to the Numbers The rachilla or not flowering units of the inflorescences in the Anom- offlorets. may may extend the most distal and reduced ochlooideae to emphasize this lack of beyond floret, recognizable or modified florets be below or above homology. Characters of the spikelet and the floret may present (or both below and above) the fertile ones. For this are scored as inapplicable or ambiguous in these two for char. 13 analysis, we have assumed that the grass flower is genera except (see Appendix IV). terminal to the axis on which it is borne. In the The grass spikelet may have originated either be- Anomochlooideae, there is no identifiable fore or after the palea divergence of Anomochlooideae. If and thus the flower to be terminal to before, then the spikelet was appears truly extensively modified the main sympodial axes of the in inflorescence in An- the long history of Anomochlooideae. The origin as discussed in and Davis certainly must have occurred before omochloa, Soreng (1998); divergence of the same is true under Soderstrom's inter- Pharoideae. We refer to the clade of all (1981) grasses ex- of the of cept Anomochlooideae as the Clade. pretation spikelet equivalent Spikelet Within the Bracts outside of the Streptochaeta. Spikelet Clade, however, spikelets subtend inflores- the flowers are borne on lateral as indi- cence axes and often have a blade. Short to elon- branches, cated by the presence of a prophyll (i.e., the gate prophylls are palea) usually present on the branches in the a(laxial on the branch. subtended by these bracts. Such bracts occur proximal, position )pri- This is clear in those taxa with multiflow- marily in Bambuseae and pattern Andropogoneae,b ut they ered or with one floret a are not necessarily homologous between the two spikelets spikelets ancd rachilla extension. There are a number of taxa with groups (see Renvoize & Clayton, 1992). Well-de- a single floret and no rachilla extension, including veloped sul)tending bracts are usually absent in the in which a is other memlers of the Pharoideae, well-developed palea Spikelet Clade, although found in the floret. It is to there may be a ridge or scar which simple enough imagine presumablyr ep- the reduction of the branch to the where resents the subtending bract at the base of the in- apex point no evidence of a rachilla extension can be ob- florescence branch. served, but, as Soreng and Davis (1998) noted, the Glumes. We have assumed here that glumes are presence of a single-flowered floret (or equivalent) homologous across the Spikelet Clade. Glumes are appears to be plesiomorphic for the family. This typically defined as the two sterile bracts at the implies that either the rachilla extension and ad- base of the spikelet, but additional sterile bracts ditional fertile florets evolved subsequently, or that (usually called sterile lemmas or sterile florets even multiflowered spikelets evolved before the diver- if there is no evidence of any, even vestigial, floral gence of the Pharoideae, and reduction to a single axis) may occur between the glumes and the flower- fertile floret occurred in that lineage (Soreng & Da- bearing lemmas (e.g., Ehrhartoideae, vis, Chusquea, 1998). Clearly, single-flowered spikelets many Centothecoideae). In general, the first evolved a number of times in various in (lower) lineages glume is abaxial and the second the BEP + PACCAD Clade. (upper) glume is adaxial (E. A. Kellogg, pers. obs.; Clifford, 1987), Lemma. Each floral axis is subtended by a lem- but the first glume may be adaxial in position, as ma (Fig. 6A and B), a structure that appears to be in a number of Paniceae (Clifford, 1987). Grassl universally present across the Spikelet Clade. The (1956) and Stapleton (1997) argued that the first lemma apparently is formed wholly by the spikelet glume is actually a prophyll for Andropogoneaea nd meristem, and thus is a bract on the rachilla (Clif- Bambuseae, respectively, and that the prophyll was ford, 1987). Lemma morphology is extremely vari- displaced upward to assume the position and func- able, but the number of nerves is consistently odd, tion of a glume. Even if the first glume is abaxial, varying from 1 to 15 (Clifford, 1987). Some taxa 402 Annals of the Missouri Botanical Garden ]E Ii L I! N e P e K: 0 I st' I \\ i II i i | LJ r I i: I! C i ;.f v : E E E E E lo . r CD 0 ijj A Figure 6. Grass flowers, fruits, and embryos. -A. Floral diagram of a grass with three lodicules and six stamens. -B. Floral diagram of a grass with two lodicules and three stamens. -C. Flower of Yushania (Bambuseae, original by D. Friedrick). -D. Lodicules of Pooideae (Poa, redrawn from Jirasek, 1968). -E. Lodicules of Chloridoideae (Muhlenbergia,r edrawnf rom Soderstrom, 1967). -F. Lodicules of Panicoideae (Setaria, redrawnf rom Jirasek, 1968). -G. Lodicules of Bambusoideae (Chusquea), showing the anterior pair (lower two) and the posterior one (upper). H. Generalized ovule section (Danthonioideae) showing haustorial synergids (stippled) (redrawnf rom Verboome t al., 1994). -I. Longitudinal section of a panicoid embryo showing presence of a scutellar tail (st) and the elongated mesocotyl internode (mi). J. Longitudinal section of a pooid embryo showing presence of an epiblast (ep). -K. Cross section of a panicoid embryo apex showing overlapping embryonic leaf margins. -L. Cross section of a pooid embryo apex showing embryonic leaf margins that meet. M, N. Caryopsis of Eustachys (Chloridoideae). -M. Hilum side, showing a punctiform hilum (h). -N. Embryo side, showing the large embryo (e). O, P. Caryopsis of Chusquea( Bam- busoideae) (redrawnf rom McClure, 1973). -0. Hilum side, showing a linear hilum (h). -P. Embryo side, showing the small embryo (e). c-coleoptile; co-coleorhiza; e-embryo; ep-epiblast; es-embryo sac; h-hilum; i-inner integument; 1-lemma; lo-lodicule; mi-mesocotyl internode; o-outer integment; p-palea; pl-placenta; r-rach- illa; s-scutellum; st-scutellar tail; t-vascular trace to placenta; w-ovary wall. Volume 88, Number 3 Grass Phylogeny Working Group 403 2001 Phylogeny and Classification of Poaceae have more than two glumes at the base of the spike- cules continue to be controversial. The grass flower let (see Glumes), but the sterile lemma of the pan- has been interpreted variously as dichlamydeous icoid spikelet does appear to represent an evolu- (with two perianth whorls, the palea representing tionary reduction from a fertile floret. two fused sepals and the lodicules the petals), mon- one the Palea. The 6A and is ochlamydeous (with only perianth whorl, palea (Fig. B) widely in- lodicules, and the to a terpreted as a prophyll (Linder, 1987; present, palea homologous Stapleton, no the lod- 1997; Clayton, 1990; Soreng & Davis, 1998; Jud- prophyll), achlamydeous (with perianth, icules modified bracts or or ziewicz et al., 1999), and prophylls do occur in the representing stipules), the flower a inflorescences of many Bambuseae and pseudanthial (with representing highly Andropo- reduced branch and the lodicules derived goneae (Judziewicz et al., 1999; E. A. Kellogg, un- system, from leaves or The mon- published obs.) relative to subtending bracts in branches) (Clifford, 1987). pre- of the flower is cisely the same way a prophyll is related to a ochlamydeous interpretation grass the most widely accepted. Recent molecular genet- subtending leaf. Although the "palea as prophyll" ic studies for a explanation may be more parsimonious, the deri- provide support petaloid homology of the lodicules Schmidt & vation of the palea from the fusion of two (Irish, 1998; Ambrose, sepals Ambrose et thus the ach- also has been 1998; al., 2000), rejecting supported (Schuster, 1910; Stebbins, and Ambrose 1974; Irish, 1998; Schmidt & Ambrose, lamydeous pseudanthial hypotheses. 1998). Nu- et al. (2000) also fiez provided genetic evidence that (1968) and Clifford (1987) have suggested that the and also the lemma have char- the odd-nerved, often 1-keeled palea possibly paleas in the Ory- acteristics in common with an outer zeae perianth whorl, represent a separate origin of a palea-like thus that the structure, but this interpretation is inconsistent suggesting dichlamydeous interpreta- tion be revived. and Pozzi with the phylogeny. There is no reason to might Bossinger (1990) suppose et al. described mutations in in which paleas have been replaced with something different (2000) barley the lemma is converted to a leaf. This in may suggest Oryzeae. Single-flowered spikelets lacking a pa- that the lemma is more leaf-like than lea sepal-like. (e.g., Alopecurus)p robably represent loss of the that these two rachilla extension and the Kellogg (2000a) suggested interpre- palea on the floral axis. tations are not and that the Vegetative branching in the grasses and other mutually exclusive, complex structure of the monocots grass spikelet may reflect typically involves a leaf with a bud or branch in simultaneous of both leaf and floral de- its axil, although bud displacement is expression If observed in some palms and the latter grasses (Dahlgren et velopmental "programs." interpretation al., 1985; Serebryakova,1 971; Fisher & is correct, then it Dransfield, may be meaningless to discuss whether the lemma is a leaf or a 1977). The first appendage of the branching axis is "really" sepal. an adaxial two-keeled bract (= prophyll). This pro- Androecium. All six stamens (arranged in two phyll encloses the bud, but often persists once the alternating whorls of three stamens each, Fig. 6A; branch develops. The origin of the vegetative pro- see Clifford, 1987) are plesiomorphically present phyll is not clear, but whether it is a single struc- within the study group and at the point of origin of ture or the result of fusion of two bracts, ultimately the grass family, but the entire outer whorl is lost it is foliar in nature (Stebbins, 1974). The relation- in the Restionaceae. Within the all six ship between the family, subtending leaf and the grass prophyll stamens (both whorls) are maintained in the three is generally constant, and therefore a prophyll subfamilies for the loss of marks the presence of a lateral branch. If the earliest-diverging (except veg- the inner anterior pair in Anomochloa, as noted etative branching pattern is reiterated in the inflo- The outer whorl is maintained rescence, then the above). throughout palea ought to be homologous to most of the grass family, except for the various au- a prophyll, and prophylls should occur in fully losses and noted bracteate inflorescences as the adaxial first tapomorphic polymorphisms ap- above. Loss of the entire inner whorl pendage of branches in the axils of (e.g., Fig. 6B) subtending is interpreted as a synapomorphy of the BEP + bracts. PACCAD Clade, but in one subclade within this Flower. Grass flowers are made up of a gynoe- large group (the bambusoid/ehrhartoid alliance) cium, androecium, and two or three flap-like struc- there are three or more independent reversions to tures (lodicules) that force the floret open at ma- presence of this whorl, and possibly a secondary turity (Fig. 6A-G). The literature on the anatomy loss in Leersia. An alternative interpretation of the and development of the grass flower should prob- outer stamen whorl, in the context of the present ably be reinterpreted in the light of the present reconstruction of phylogenetic relationships, and phylogeny. The origin and homology of the lodi- involving no secondary origins, would have the in- 404 Annals of the Missouri Botanical Garden ner whorl lost independently in Pooideae and the are additional differences in expression of the en- PACCAD Clade, as well as in a series of small zymes involved in photosynthesis (Sinha & Kellogg, lineages within the bambusoid/ehrhartoid alliance. 1996). For example, all Chloridoideae except for Alternatively, if a PACCAD + Pooideae lineage is the C3 Eragrostis walteri and Merxmuellerar angei considered, loss of the inner stamen whorl could be form aspartate, use the NAD-malic enzyme, and interpreted as a synapomorphy of that clade; the have double bundle sheaths, with the inner one of presence of this whorl in some genera of Ehrhar- thick-walled cells. (Centropodiah as not been bio- toideae and Bambusoideae might then be inter- chemically typed, but is anatomically the same as preted as retention of a plesiomorphy, while the ab- other NAD-ME taxa.) Stipagrostis,i n Aristidoideae, sence in others would be interpreted as having is similar to the chloridoids in having two bundle arisen independently of the loss in the PACCAD + sheaths, the outer of which appears to be carbon- Pooideae. reducing (Sinha & Kellogg, 1996). Aristida itself forms malate, using the NADP-malic enzyme; un- BIOCHEMISTRY like other NADP taxa, it has two bundle sheaths. The ultrastructureo f the outer sheath is similar to The phylogeny suggests that the C4 photosyn- that of NADP species, but the inner sheath is more thetic pathway has evolved multiple times within like an NAD or PCK plant; the extent to which the PACCAD Clade. In the C4 pathway, the Calvin- these sheaths are developed varies throughoutt he Benson cycle, and hence Rubisco, is relegated to genus (Brown, 1977; Carolin et al., 1973). Eriachne the bundle sheath cells surrounding the veins (Hat- also uses NADP-ME and has a double bundle tersley & Watson, 1992; Sinha & Kellogg, 1996). sheath, but unlike Aristida the inner sheath has Phosphoenolpyruvate carboxylase then catalyzes thick-walled cells. The C4 Panicoideae also vary CO2 reduction in the mesophyll to produce the four- biochemically and histologically, although large carbon compound oxaloacetate. This compound is groups are uniform. For example, the entire tribe then reduced to malate or aspartate and transported Andropogoneae (ca. 100 genera and 1000 species) to the bundle sheath, where the newly fixed CO2 is uses NADP-ME and has a single bundle sheath. immediately removed and taken up by Rubisco. Both the phylogeny and structural/biochemical This keeps CO2 concentration high at the active site data indicate that the C4 pathwayi s not homologous of Rubisco, preventing competition by O2. Consis- wherever it occurs. The close relationship of the C4 tent with the constant flow of materials between the lineages, however, suggests that there were a set of bundle sheath and mesophyll, C4 species (Fig. 7C) changes at the base of the entire PACCAD Clade have closer vein spacing than C3 species (Fig. 7B). that made the pathway easier to evolve. If this were Other anatomical manifestations of C4 photosynthe- true, then those changes, whatever they were, sis include enlarged bundle sheath cells (Kranz would be homologous, even though the final man- anatomy), closely packed chlorenchyma cells, and ifestations of the pathway are not. in some, radiate chlorenchyma (Fig. 7C). Despite the foregoing generalizations, C4 photo- IMPLICATIONSF OR MORPHOLOGICALE VOLUTION synthesis actually represents a suite of characters, The phylogeny provides a unique and rather than a powerful single genetic and phylogenetic tool for description of evolutionaryp attern (Kellogg, change. Only down-regulation of Rubisco in the 2000a). Major clades and evolutionary transitions mesophyll, up-regulation of PEP carboxylase, and are summarized in 8. Additional detail can closer vein spacing are common to all C4 lin- Figure grass be found at The four-carbon that http://www.virtualherbarium.org/grass/ eages. compound transports the carbon to the bundle sheath be malate or gpwg/default.htm. may aspartate, the decarboxylating enzyme may be a Sister relationship of Poaceae and Joinville- malic enzyme using NAD as a co-factor (NAD-ME), aceae. The presence of long and short cells in the or using NADP (NADP-ME). If the former, addi- leaf epidermis (char. 42), with at least some of the tional decarboxylation activity may be provided by short cells containing silica bodies, unambiguously PEP carboxykinase, so PCK activity is dependent supports the sister relationship of Joinvilleaceae upon the presence of NAD-ME (Kanai & Edwards, and Poaceae. This arrangement is apparently 1999). Some C4 grasses have only one bundle unique among angiosperms (Campbell & Kellogg, sheath, whereas others have two. In those with two 1987; Kellogg & Linder, 1995). The presence of bundle sheaths the inner sheath may be made up the 6.4 kb inversion in the chloroplast genome of thick-walled cells, forming a conventional mes- (char. 51) is also an unambiguous synapomorphy tome sheath, or it may be parenchymatous. There supportingt his sister relationship, although it is not Volume 88, Number 3 Grass Phylogeny Working Group 405 2001 Phylogeny and Classification of Poaceae A.k. A Figure 7. Leaf anatomy.- A. Dinochloa maclelandii (Soderstrom2 607). -B. Poa sp. (Carolina Biological Supply Co.). -C. Bouteloua sp. (Carolina Biological Supply Co.). ac-arm cell; bc-bulliform cell; ch-chlorenchyma; fc- fusoid cell; is-intercellular space; ms-mestome sheath; p-phloem; ps-outer parenchyma sheath; rch-radiate chlorenchyma;s g-sclerenchyma girder; st-stomatal apparatus;x -xylem. known whether Ecdeiocoleaceae possess this in- gion of the chloroplast genome (char. 52), unam- version (R. J. Soreng, pers. comm.). biguously support the monophyly of Poaceae. Two other features unique to and characteristic of Po- Monophyly of Poaceae. Two characters in this aceae but not included in this analysis are the cary- analysis, the highly differentiated grass embryo opsis and the presence of intraexinous channels in (Fig. 61 and J) and its lateral position (char. 34) the pollen wall (Linder & Ferguson, 1985; Camp- and the trnT inversion in the large single-copy re- bell & Kellogg, 1987; Kellogg & Linder, 1995). 0O) Panicoideae ? * Centothecoideae ? * Aristidoideae ? Danthonioideae * Arundinoideae * Chloridoideae ? Eriachneae ? Micraireae Pooideae * 3 o -X Ehrhartoideae * .o o O r Streptogyneae D o' Bambusoideae * c) Puelioideae * D Pharoideae * Anomochlooideae Joinvilleaceae Other Poales ? * Volume 88, Number 3 Grass Phylogeny Working Group 407 2001 Phylogeny and Classification of Poaceae The caryopsis is a single-seeded, usually dry, in- es except Anomochlooideae) is then likely the re- dehiscent fruit with the pericarp fused to the seed sult of a cumulative series of changes that occurred coat in the hilar region and otherwise closely ad- during the early history of the family. Clayton nate (Sendulsky et al., 1987). The caryopsis devel- (1990), however, pointed out that there are well- ops from a unilocular ovary containing a single developed bracts in the spikelets of Restionaceae, ovule (Fig. 6H). Within Poaceae, the basic cary- and what could be interpreted as subtending bracts opsis has been modified to the fleshy (baccoid) or are present, although not necessarily well devel- the achene-like (nucoid) caryopses of some woody oped, in Joinvilleaceae (Dahlgren et al., 1985). bamboos (Sendulsky et al., 1987) or the follicoid or Clayton (1990) also noted that if the palea in grass- cistoid caryopses of some chloridoids in which the es is interpreted as a prophyll, then there is no seed is free from the pericarp or separates from it homolog for it among other Poales; however, pro- when moistened. phylls are occasionally found at the base of spike- The highly differentiated grass embryo and its lets in Restionaceae (H. P. Linder, pers. obs.). lateral position at the base of the caryopsis (char. The results of this analysis and prior studies 34; Fig. 61 and J) are synapomorphiesf or the family (Clark et al., 1995; Soreng & Davis, 1998) support (Campbell & Kellogg, 1987; Kellogg & Linder, the following as plesiomorphic within the grass 1995). In the grasses, the embryo has leaves, vas- family: an herbaceous, perennial, rhizomatoush ab- cular tissue, and clearly localized shoot and root it; pseudopetiolate and relatively broad leaf blades meristems before the fruit is dispersed, and thus looks much more like a than the bearing multicellular microhairs and anatomically seedling embryos with commissural veins, fusoid cells of relatives (Fig. 7A), and non-grass (Fig. 61 and J; Reeder, 1957; et Constant features of the alternating long and short cells on the epidermis Sendulsky al., 1987). with at least some of the short cells silica grass embryo are the scutellum, coleoptile, and co- including leaves with an adaxial and leorrhiza. A scutellar cleft bodies; ligule open (Fig. 6I) may or may not sheaths; a highly bracteate inflorescence; one-flow- separate the scutellum from the coleorrhiza,a nd the ered or six stamens epiblast, an extra flap of tissue the scu- spikelets spikelet equivalents; opposite in two whorls with dithecal an- tellum, may be present (Fig. 6J) or absent (Fig. 61). tetrasporangiate, Whether or not the leaf meet or thers; monoporate pollen with intraexinous chan- embryonic margins overlap varies throughout the 6K and nels in the wall; a uniloculate, uniovulate family (Fig. gynoe- cium with three stigmas and one order of L). stigmatic in the inflorescence occurred between branching; a basic caryopsis with a linear Changes hilum; a the time that Joinvillea divergetda nd the time of highly differentiated, laterally positioned embryo a and a divergence of Anomochlooideae. Relative to most with scutellum, coleoptile, coleorhiza, neg- of their poalean sister families (excluding Restion- ligible mesocotyl internode; Festuca-type starch aceae, Ecdeiocoleaceae, and Centrolepidaceae),t he grains in the endosperm; and the C3 photosynthetic grasses, including Anomochlooideae, have well-de- pathway.T he grass spikelet and lodicules may have veloped bracts (what are normally called subtend- evolved before the divergence of the Anomochlooi- ing bracts, prophylls, glumes, lemmas, and paleas) deae and the rest of the family, and were lost in the subtending and enclosing contracted inflorescence Anomochlooideae (cf. Soreng & Davis, 1998), but branches and flowers (Figs. 4 and 5). This appears these features are not plesiomorphic under the pre- to be a derived character marking the origin of the sent optimization. Bisexual flowers are probably family, although this was not used explicitly as a also plesiomorphic, but unisexuality evolved early characteri n this analysis. Under this interpretation, and in a number of different lineages within the the characteristic grass spikelet (found in all grass- family. Figure 8. Summaryp hylogeny of the grasses indicating significant morphological,e cological, and molecular (cpDNA = chloroplast DNA) events in the evolution of the family. Infrequentl osses, parallel gains, and reversals are not shown for these characters. The 12 subfamilies recognized by the GPWG appear in boldface. Poales sensu APG include Cyperaceae. Markedt axa: (star) At least some included species have unisexual flowers/florets;( ?) At least some included species have a C4 carbon fixation pathway, Kranz anatomy, or both. Dark circles indicate nodes strongly supported by all data combined (bootstrap> 99; Bremer support > 16). Subfamilies with common names: Aristidoideae (wiregrasses, etc.), Arundinoideae (reeds, etc.), Bambusoideae (bamboos), Chloridoideae (lovegrasses, tef, etc.), Danthonioideae (oat- grasses, pampas grass, etc.), Ehrhartoideae( rice, wild-rice, etc.), Panicoideae (maize, panic grasses, millets, sorghum, sugar cane, etc.), and Pooideae (barley, brome grasses, oats, rye, wheat, etc.). 408 Annals of the Missouri Botanical Garden Anomochlooideae. The basal divergence be- let equivalent of Anomochloa is indeed homologous tween Anomochlooideae and the rest of the Poaceae to a standard grass lemma, a palea is lacking and (the Spikelet Clade) is well supported based on mo- the flower is terminal, whereas in a true grass floret lecular evidence. Monophyly of Anomochlooideae, the flower is borne on a lateral axis as indicated by however, is supported morphologically only by the the presence of the palea if it is interpreted as a unreversed presence of the adaxial ligule as a prophyll. Some authors interpret the distal three fringe of hairs, a character that appears elsewhere bracts of the spikelet equivalent of Streptochaeta as in the family. The pulvinus at the summit of the lodicules and the next proximal two bracts as a pseudopetiole is a possible synapomorphy for this bifid palea (e.g., Clayton, 1990), but there is no clade but requires further study to determine sim- compelling evidence for this. The spikelet equiva- ilarities with the structure in other grasses. Molec- lent of Streptochaeta might represent a condensed ular support for this clade may be due at least in branching system (Soderstrom, 1981), but it is not part to long-branch attraction (see Unresolved a pseudospikelet as found in the Bambuseae. In questions). any case, lack of a palea in Streptochaeta also im- Anomochlooideae have been recognized as a plies that the flower is terminal. separate family (Nakai, 1943), a point of view that The is consistent with the We Spikelet Clade (Pharoideae + [Puelioideae completely phylogeny. + {BEP + This have chosen here to retain Anomochlooideae within PACCAD]]). clade, which in- cludes all of the grasses except for Anomochlooi- Poaceae because of the strong synapomorphies deae, is defined the of linking them by unambiguous presence (notably the caryopsis and the highly true grass spikelets, florets (char. 8), and lodicules differentiated embryo, see Fig. 8). Retention of An- (char. 18). The condition of the omochlooideae in Poaceae is also plesiomorphic taxonomically in line with all studies of the spikelet is clearly the presence of a pedicel (char. conservative, previous two and consistent with the efforts of the APG 10), glumes, and a well-developed lemma and family, palea in the floret. The to limit monotypic or small families et al., single-flowered spikelet may (Chase be synapomorphic for the family above the point of 2000a, b). divergence of Anomochlooideae are variable with to Anomochlooideae, with a transfor- respect mation to multiflowered leaf and the spikelets above Pharoideae embryonic margins epiblast. Embry- and then numerous reversals, but the first true onic leaf margins meet and the epiblast is present spikelets in grasses may have been multiflowered in Anomochloa, whereas the embryonic leaf mar- (see discussion under and the is absent in Spikelet). The plesiomorphic gins overlap epiblast Strepto- condition for lodicules is chaeta & Both clearly three (char. 19), (Judziewicz Soderstrom, 1989). gen- unfused and with a era have an scutellar cleft. The (char. 20), distally membranous inconspicuous portion (char. 21). Presence of unisexual flowers coleoptile is usually represented as a more or less may be conical the leaves and synapomorphic for this clade, with numer- "cap" protecting embryonic ous reversals to bisexual shoot but the florets, but it may be more apex, coleoptile margins are entirely that arose times. A free and overlapping in Streptochaeta, whereas in likely unisexuality multiple base chromosome number of x = 12 (char. 47) was Anomochloa the margins at the base of the cole- established before the of Pharoideae. optile are fused but free toward the divergence apex, as is also seen in Pharoideae (Reeder, 1953; Judziewicz & Pharoideae. Monophyly of this clade is strong- Soderstrom, 1989). ly supported by the presence of resupinate leaf The inflorescences of both Anomochloa and blades, oblique lateral veins in the leaf blades, and Streptochaeta are bracteate, but the lack of clear uncinate hairs wholly or partially covering the fe- homology of these bracts with those of the standard male lemmas (Judziewicz, 1987). The female lem- grass spikelet has been noted. The spikelet equiv- mas exhibit a laminar anatomical structure similar alents in both Anomochloa and Streptochaeta are to that of Anomochlooideae, but the transversely one-flowered and bisexual. The upper bract in An- elongated cell layer is subjacent to the adaxial epi- omochloa exhibits a laminar anatomical structure, dermis. In Pharoideae, Anomochlooideae, and with the transversely elongated cell layer subjacent some Bambusoideae (Ghopal & Ram, 1985), the to the abaxial epidermis of the bract (Judziewicz & coleoptile margins are free for at least a portion of Soderstrom, 1989). This laminar structure is not their length, but the distribution of this feature in found in bracts of Streptochaeta. (Similar but not the rest of the BEP + PACCAD Clade is not well identical laminar anatomy characterizes female documented. Pharoideae embryos have an epiblast lemmas in Pharoideae.) As Soreng and Davis and a small scutellar cleft, and embryonic leaf mar- (1998) pointed out, if the upper bract of the spike- gins meet (Judziewicz, 1987). The Pharoideae uni- Volume 88, Number 3 Grass Phylogeny Working Group 409 2001 Phylogeny and Classification of Poaceae formly exhibit one-flowered, unisexual, paired phological synapomorphies support its monophyly: (char. 9) spikelets (Fig. 4B). When present in the loss of the pseudopetiole (char. 7), reduction to two male spikelets, the three lodicules are small, which lodicules (char. 19), loss of the inner whorl of sta- may be plesiomorphic or may represent a reduc- mens (chars. 23 and 24), and loss of arm and fusoid tion. Disarticulation is variable, in that the whole cells (chars. 45 and 46). The pseudopetiole is re- inflorescence may disarticulate as in Scrotochloa, gained in Bambusoideae, as well as in a few mem- or the whole inflorescence or branches usually dis- bers of the PACCAD Clade. Arm and fusoid cells articulate in Pharus, or female spikelets disarticu- are also regained in Bambusoideae. The inner late above the glumes in Leptaspis, and perhaps whorl of stamens is interpreted as having been re- also in the other genera (Soderstrome t al., 1987). gained three or four times within the bambusoid/ These, along with the uncinate hairs, appear to be ehrhartoid clade. Within the BEP + PACCAD adaptations to epizoochorous dispersal. Multicel- Clade, the lamina on the first seedling leaf is lost lular microhairs (char. 43) are lost in Pharoideae. only in Bambusoideae and Oryzeae. Unisexual flo- rets have evolved in most of this clade, The Bistigmatic Clade (Puelioideae + + lineages [BEP e.g., Olyreae (Bambusoideae), Zizania (Ehrharto- PACCAD]). This clade is marked by three mor- ideae), Lamarckia (Pooideae), several genera of phological synapomorphies: transformation from Chloridoideae and Centothecoideae, and very com- three to two stigmas (char. 29), transformationf rom monly in Panicoideae. Most two lineages include taxa one to orders of stigmatic branching (char. 30), with one floret per spikelet and taxa with multiple and presence of the 15 bp ndhF insertion (char. florets per spikelet. The presence or absence of an 53). Multiple florets per spikelet (char. 12; Fig. 5) epiblast is variable, as is the presence or absence may have arisen within Puelioideae as shown in of the scutellar cleft, in the PACCAD Figure 3, or in although the common ancestor of the Pue- Clade the scutellar cleft is + generally present (Reed- lioideae (BEP + PACCAD)c lade. Regardless of er, 1957). which scenario is correct, reversals to one floret The large number of reversals hypothesized in occurred numerous times in the BEP + PACCAD this part of the tree raises a number of intriguing Clade. Disarticulation above the glumes (char. 16) questions evolution. We do clearly is established of regarding morphological before the divergence Pue- not know anything about the development or un- lioideae. derlying genetics of the characters, so we are forced Puelioideae. The forest habitat and broad, into the agnostic assumptions that gains and losses pseudopetiolate leaf blades of the Anomochlooi- are equally likely, and that pseudopetioles, arm deae and Pharoideae are retained in this subfamily, cells, fusoid cells, epiblasts, and unisexual flowers but no unique morphological synapomorphies for are all developmentally and genetically the same Puelioideae have been identified. The culms ap- wherever they occur. The changes that we interpret parently do not produce aerial branches, nor basal as reversals could actually represent retained prim- tillering as in Anomochlooideae and Pharoideae. itive characters if loss of these characters is more The presence of proximal female-sterile florets in likely than their regain. Equally possible, the the spikelet (char. 11) is an unreversed synapo- changes interpreted as reversals could represent morphy for Puelioideae in this analysis, but is also the origin of novel characters that look superficially a synapomorphy for Paniceae and autapomorphic similar to ancient ones. We have some evidence for for multiple other taxa on the tree. In Puelioideae, the latter (see below) in that arm cells in the bam- the pattern of sexuality in the spikelets is somewhat busoids are actually morphologically different from more complex than in many other subfamilies, but those in Anomochlooideae and Pharoideae (Zhang within a spikelet, at least some proximal florets are & Clark, 2000). Finally, the character optimizations male. In Guaduella, the 1 to 3 proximal florets are reflect the hypothesis that the BEP Clade is mono- male, and additional florets are bisexual with the phyletic. If, as we outline below, the Pooideae are distalmost one or few reduced, but in Puelia the actually sister to the PACCAD Clade-a hypothesis proximal 3 to 6 florets are male or neuter, with the that is neither favored nor excluded by the data- single apical floret unisexual and female. Multicel- then the pattern of morphological evolution is dif- lular microhairs (char. 43) are lost in Puelia, and ferent. a reversion to three stigmas (char. 29) occurs in some species of Puelia. The BEP Clade (= BOP clade of Clark et al., 1995). This clade is supported by molecular se- The BEP + PACCAD Clade. This clade in- quence data, particularly from ndhF, rpoC2, and cludes the vast majority of grass species. Six mor- phyB (see Results), but other data sets support a 410 Annals of the Missouri Botanical Garden Pooideae + PACCAD clade (Soreng & Davis, Pooideae. Monophyly of the pooid clade is 1998). In this analysis, constraining Pooideae + strongly supported by molecular data including PACCAD as monophyletic was only slightly less cpDNA restriction site data (Soreng et al., 1990; parsimonious than BEP + PACCAD (see Results). Davis & Soreng, 1993; Nadot et al., 1994; Soreng In addition, no morphological synapomorphies sup- & Davis, 1998, 2000). Parallel-sided subsidiary porting the BEP Clade have been identified. Loss cells, lack of microhairs, nonvascularizedl odicules of the lemma awn is optimized to this node, but (Fig. 6D), and the presence of an epiblast and lack awns are regained in many taxa within the BEP of a scutellar cleft in the embryo (Fig. 6J) are char- Clade. The lack of sequence data for Streptogyna acteristic of a majority of the subfamily but do not contributes to the uncertainty of its position within constitute unequivocal synapomorphies. In this the BEP Clade and may also affect assessment of analysis, the loss of stylar fusion (char. 28) is an the monophyly of the clade. Streptogyna appears as unreversed synapomorphyf or the Pooideae. Loss of sister to Ehrhartoideae (Fig. 1), but in other anal- the scutellar tail (char. 36) is widespread in the yses of these data it appears as sister to the rest of clade, but polymorphismsp revent its unambiguous the BEP Clade. optimization. An unreversed transformationt o faint or absent vascularization of the lodicules Bambusoideae. Monophyly of the true bamboos (char. 22) occurs within the Pooideae after the divergence of (i.e., olyroid + woody bamboos) is supported by Brachyelytrum.A transformationt o the molecular data in this and other embryonic analyses (Clark et leaf & margins meeting (as opposed to overlapping; al., 1995; Zhang, 1996; Zhang Clark, 2000). of the Fig. 6L) also occurs after the divergence of Bra- Morphologically, secondary gain pseudope- but is reversed in tiole and loss of the lamina of chyelytrum Phaenosperma (or (char. 7) secondary Phaenosperma + Multicellular micro- the first seedling leaf Anisopogon). (char. 41) are synapomor- hairs or absence of arm (char. 43) are known only in Lygeum + Nar- phies. Although only presence dus; although this character is scored for the cells was scored in this analysis, Zhang and Clark only abaxial leaf it found that the of surface, appears that Pooideae, at (2000) presence strongly asym- least above this divergence, are the arm cells is a only group of metrically invaginated (Fig. 7A) po- grasses to lose completely the ability to make mul- tential synapomorphy for this clade. Fusoid cells ticellular microhairs anywhere on the are characteristic of the Bambusoideae plant (except (Fig. 7A), possibly the lodicules). Chromosomale volution in but it is not known whether their presence repre- Pooideae is complex (see char. but x = 12 is sents retention of the 47), plesiomorphic condition or apparently plesiomorphic in the BEP so reversal after loss of fusoid cells at the base of the Clade, numbers such as x = 10 and x = 11 in the earlier- BEP Clade. Bambuseae are here supported by the diverging lineages of Pooideae may well be derived presence of perennating woody culms (char. 1), ab- from this condition. The presence of x = 12 in axial ligules (char. 5), and Panicum-type starch and some A of the inner Phaenosperma, Ampelodesmos, Stipeae grains (char. 40). secondary gain be a x = 7 is a stamen whorl 23 and occurred at least may retention; clearly synapomor- (chars. 24) phy of the core Pooideae Bra- once but (here represented by possibly several times. Olyreae have a base chromosome number of x = chypodium,A vena, Bromus, and Triticum)T. wol od- synapomorphic icules (char. 19) are found at the base of 11 (char. 47), but the tribe is also characterized Pooideae, by but a reversal to three occurs in Stipeae (in which unisexual spikelets. another transformation,t o two, occurs in Nassella); Ehrhartoideae. This lineage is strongly sup- this is undoubtedly an oversimplificationo f the pat- ported by molecular data, and is characterized by tern in the Stipeae in which lodicule numberv aries the presence of one female-fertile floret per spike- considerably (Vickery et al., 1986). Loss of the dis- let, often with one or two proximal female-sterile tally membranousp ortion of the lodicule (char. 21) florets (char. 11). This character is coded as am- is a synapomorphy for Meliceae. The earliest-di- biguous in Oryza and Leersia, but if the vestigial verging lineages of the pooid clade have one floret structures at the base of the spikelets in these gen- per spikelet (char. 12) (although a rachilla exten- era are interpreted as highly reduced glumes, then sion is present in Brachyelytrum),m ultiple florets the presence of proximal female-sterile florets is an appear in Meliceae, single florets characterize the unambiguous synapomorphy. Two lodicules (char. (Phaenosperma + Anisopogon) + Stipeae clade, 19) are found in this clade; in addition, the inner multiple florets are found at the base of the core whorl of stamens (chars. 23 and 24) is regained, pooids, and many taxa within the core pooids have styles are not fused (char. 28), and fusoid cells one floret per spikelet. Multiple independent ori- (char. 46) are lost. gins of multiple florets per spikelet can be hypoth- Volume 88, Number 3 Grass Phylogeny Working Group 411 2001 Phylogeny and Classification of Poaceae esized, but subsequent reduction to one floret per lished for the BEP + PACCAD Clade, but no trans- spikelet has clearly occurred in several groups. Pat- formations to any other number occur in the PAC- terns of divergence within this clade are complex CAD Clade. and still are being evaluated, so some inferences The positions of Micraira and Eriachne in the regarding character evolution are likely to change. phylogeny are not well resolved, presumably due to a lack of sequence data for both genera (see Results The PACCAD Clade. Over half the species of and also Unresolved Questions). The two species of the grass family are included in this clade. Even Eriachne are from quite distinct parts of the genus, as early as the 1930s (Avdulov, 1931; Prat, 1932, based on the informal classification of Lazarides 1936; Roshevits, 1937), taxa of this clade have (1995). This undoubtedly affects the interpretation been grouped together. Hilu and Wright (1982) of character state transformations within the entire were the first to retrieve this clade in a formal anal- clade. ysis, and subsequently support for the monophyly Early in the evolution of the PACCAD Clade, of the clade is found in all molecular analyses to some lineages developed the capacity for C4 pho- date with sufficient sampling (Hilu & Esen, 1988; tosynthesis, apparently as an adaptation to high Hilu & Johnson, 1991; Davis & Soreng, 1993; Na- light/high temperature conditions and perhaps also dot et al., 1994; Barker et al., 1995; Clark et al., to falling levels of atmospheric CO2 (Sage & Mon- 1995; Duvall & Morton,1 996; Liang & Hilu, 1996; son, 1999). Most members of the Panicoideae, all Mathews & Sharrock, 1996; Soreng & Davis, 1998; but two Chloridoideae, the Aristidoideae (except for Hsiao et al., 1999; Mathewse t al., 2000) except for Sartidia), and the Eriachneae are C4. The poor res- Cummings et al. (1994), in which an oryzoid clade olution of the phylogeny at the base of the PACCAD nested within the PACC clade. Davis and Soreng Clade makes it impossible to determine precisely (1993) named this the PACC clade based on the how many origins of C4 photosynthesis there were, four subfamilies that were then recognized as com- but certainly there were at least two, and possibly prising the clade, but we here modify the name to more. The data are also consistent with a polymor- reflect the recognition of two additional subfamilies, phism at the base of the PACCAD Clade. the Aristidoideae and the Danthonioideae. The PACCAD Clade is robustly supported based The Panicoideae + Centothecoideae Clade. on molecular data and additionally is supported by This clade was recovered in virtually all suban- the presence of an elongated mesocotyl internode alyses, and had reasonable support (bts 85, brs 8) (char. 37) and the loss of the epiblast (char. 35; in the combined analysis. The presence of non-lin- Fig. 61). The latter character reverses in the clade, ear hila (char. 33; Fig. 6M) is a potential synapo- so that secondary gain of the epiblast is an apparent morphy for this clade. Although support for the synapomorphy for Centhothecoideae. Two charac- monophyly of Panicoideae (excluding Gynerium ters (chars. 21 and 50) are and possible Danthoniopsis) was synapomor- strong (see Results), rela- for the PACCADC lade, but because of a lack tionships of the centothecoid phies taxa, Gynerium, and of data or lack of a structurei n Micraira, Danthoniopsis to the Panicoideae and to each other placement of these transitions is ambiguous. The lack of lod- remain unresolved, but the placement of Gynerium icules in Micraira as sister to traditional Panicoideae is a novel result. prevents unambiguous place- ment of the loss of the distally membranousp ortion Panicoideae. The presence of proximal female- of the lodicule (char. 21), and Micraira remains un- sterile florets (char. 11) and the transformation to sampled for the presence or absence of the 3 bp the classical NADP-ME C4 subtype (char. 48) are deletion in phytochrome B (char. 50). Solid culm unambiguous synapomorphies for Danthoniopsis + internodes (char. 2) are shown here as synapo- Panicoideae. Some reversions to the C: pathway oc- morphic, although hollow ones reappear in other cur within the Paniceae among unsampled taxa, members of the clade. Non-linear hila (char. 33; and at least one secondary transformation to the Fig. 6M) are widespread in the PACCADC lade, but NAD-ME C4 subtype occurs in Panicum. This the point of origin is ambiguous. The Panicum-type clade is also supported by the presence of one fe- starch grain syndrome (char. 40) may be a syna- male-fertile floret (char. 12) as a reversal and the pomorphy for the PACCAD Clade, with a reversal gain of a germination flap (char. 17), but the place- to the Festuca-type in the clade containing Eri- ment of this latter transformation is ambiguous. The achne, Aristidoideae, Danthonioideae, Arundino- loss of disarticulation above the glumes (char. 16) ideae, and Chloridoideae (the Ligule of Hairs is a synapomorphy for Panicoideae excluding Dan- Clade, as defined below), but other optimizations thoniopsis. The presence of paired spikelets (char. are possible. Two lodicules (char. 19) are estab- 9) is a synapomorphy of Andropogoneae in this 412 Annals of the Missouri Botanical Garden analysis, but paired spikelets do occur within Pan- relationship of these two clades is relatively mod- iceae (e.g., Brachiaria, Digitaria, Paspalum). est. The gain of the NAD-ME C4 subtype (char. 48) is a Centothecoideae. of this possible synapomorphy for the entire clade, Monophyly subfamily as constituted is not in however, and if so it would then revert to C3 in M. currently strongly supported rangei. The gain of chloridoid-type microhairs this analysis. The secondary gain of an epiblast is a is un- (char. 44) is a synapomorphy for the traditional (char. 35) possible synapomorphy( but known for as is fusion of the Chloridoideae, although the character does occur Thysanolaena), styles elsewhere in the PACCAD Clade, and many genera (char. 28). of chloridoids also include species with panicoid- The Ligule of Hairs Clade (Eriachne + [[Aristi- type microhairs (Jacobs, 1987; Van den Borre, doideae + Danthonioideae] + [Arundinoideae + 1994; Van den Borre & Watson, 1994). The peri- Chloridoideae}]). The adaxial ligule as a fringe of carp is often free or loose, but this feature is not hairs (char. 4), awned lemmas (char. 13), and com- uniform and is also found in non-chloridoid grass- pound starch grains (char. 40) are synapomorphies es. of this clade, but characters 4 and 13 reverse mul- tiple times, and character 40 once, within this UNRESOLVED QUESTIONS clade, as well as elsewhere on the tree (Fig. 3). The of this clade is a novel Anomochlooideae. Anomochloa recovery finding, but further Monophyly of and are each distinctive but investigation is warranted, given the lack of se- Streptochaeta genera data for Eriachne. The transformationt o appear to have little in common. We have not quence yet embryonic leaf found a derived character margins meeting (char. 38; Fig. 6L) uniquely morphological is an unreversed of the four subfam- that unites them as members of a clade. Al- synapomorphy single ilies above Eriachne. though the analyses presented here indicate that the two form a monophyletic group, analyses of sin- The Aristidoideae + Danthonioideae Clade. gle data sets sometimes show them to be paraphy- Although each of these subfamilies is well sup- letic or unresolved (Mathews et al., 2000; Hilu et ported as monophyletic, their sister relationship is al., 1999; Zhang, 2000). Because both genera oc- another novel result, and one that is only moder- cupy long branches in gene trees, they may form a ately supported. Nonetheless, the presence of a ba- clade only because of long-branch attraction (Fel- sic pattern of three awns (char. 14; Fig. 5C and F) senstein, 1978). Molecular studies of other species is an unreversed synapomorphyf or this clade. Re- of Streptochaeta would help break up the long appearance of the distal membranousp ortion of the branch to S. angustifolia and might affect the lodicules (char. 21) also may be a synapomorphy, monophyly of the clade. Resolution of Anomochloa although this reverses within the Danthonioideae. and Streptochaeta as two separate basal lineages obviously would affect interpretations of character Aristidoideae. Gain of a germinationf lap (char. evolution within the and transformationt o a base chromosomen um- family. 17) ber of x = 11 (char. 47) are unambiguous syna- Position of Streptogyna. As noted in Results, pomorphies for the clade. the position of Streptogyna is ambiguous, appar- ently caused by lack of data. There are two Danthonioideae. The species presence of haustorial in the is as an unreversed genus, one in the New Worldt ropics and the synergids (char. 32) interpreted other in Africa. Neither has been collected fre- synapomorphy,b ut wider sampling within the clade is needed. quently, and we do not know of any plants in cul- tivation. Morphologically,t he genus would fit com- Arundinoideae. No unambiguous morphologi- fortably within the Bambusoideae, but molecular cal support for the monophylyo f this subfamily was data suggest that it is an early-diverging member found, although a reversal to hollow culms (char. 2) of the BEP Clade or the Ehrhartoideae.T he char- occurs in this clade. acters it shares with Bambusoideae are thus pre- ancestral, not indicative of Chloridoideae. Chloridoideae, including Cen- sumably relationship. Accurate of is for tropodia and Merxmuellera are placement Streptogyna necessary rangei, supported based on molecular no clearcut mor- interpretation of character evolution in the early- data, although members of Bambusoideae, Ehrharto- phological synapormorphies have been identified. diverging ideae, and Pooideae. Monophylyo f Centropodia+ M. rangei is well sup- ported as is that of the traditional Chloridoideae Early-diverging Pooideae. The combined anal- (i.e., Chloridoideae s. str.), but supportf or the sister ysis confirmst he position of Brachyelytruma s sister Volume 88, Number 3 Grass Phylogeny Working Group 413 2001 Phylogeny and Classification of Poaceae to the rest of the Pooideae, and Lygeum + Nardus supported as monophyletic by our data. The list of as the next-diverging lineage; both these results are genera included in and excluded from each sub- well supported. The next diverging lineages include family, however, is based on a rather limited sample Phaenosperma, Anisopogon, Stipeae, Ampelodesmos, of species and genera, combined with inferences Meliceae, and Diarrheneae, but the order of diver- from classical morphological studies. In particular, gence is not resolved by any data collected to date. the exact circumscriptions of Danthonioideae, In the case of Phaenosperma, Anisopogon, Diar- Arundinoideae, and Centothecoideae are not pre- rhena, and Ampelodesmos, the problem may be as- cisely determined by this study. A comprehensive cribed to insufficient sequence data in this analysis. effort by multiple systematists is needed to improve For our sample of Stipeae and Meliceae, however, understanding of the many poorly known species appreciable sequence data are available, yet the and genera. relative positions of the two lineages remain un- Centothecoideae. Of the here clear. If the phylogenetic problem is indeed soluble groups recognized as subfamilies, Centothecoideae are the only one with molecular data, the sample of genera and spe- not as the com- cies in each tribe have to be increased sub- strongly supported monophyletic by may bined We have retained the subfamilial stantially. A combined analysis. analysis of cpDNA restric- name and tion sites and expanded the circumscription to include morphology (Soreng & Davis, 2000) a member of the Arundi- represents the broadest taxon Thysanolaena, formerly sample for Pooideae noideae. As with the remainder of the PACCAD among studies to date. The order of divergence of a clear of the limits of the centothe- these lineages affects of the evolution Clade, picture interpretation coid clade on much more of such characters as depends data, particu- parallel-sided subsidiary larly on the remaining centothecoid genera, but a cells, loss of microhairs, and trends in reduction of is under G. chromosome number The latter study way (J. Sanchez-Ken, pers. (Kellogg, 1998). may correlate with a marked increase in comm.). genome size (Bennetzen & Kellogg, 1997) and may suggest Circumscriptiono f tribes. This paper does not possible mechanisms of genome evolution. address tribal circumscription. This will require far more extensive sampling, particularly in Pooideae, The PACCAD Clade. Relationships among the Panicoideae, Chloridoideae, and Bambusoideae, major lineages in the PACCAD Clade are not re- which constitute the four largest subfamilies. solved by this or any other phylogenetic analysis to Choice of outgroups for such studies is now clear, date. In the combined analysis, the branches at the however. base of the clade are short, marked by relatively few mutations each 41, and 16 steps; Biogeography. Present-day distributions do not (11, Fig. 1). This suggests that the PACCAD radiation indicate much about where the may have grasses originated. occurred relatively rapidly. If this is true, then re- Restionaceae are clearly a Gondwanang roup, with lationships may remain difficult to resolve with cer- representatives in Africa and Australia. Joinville- tainty. The clade also contains a number of taxa of aceae, however, are insular, occurring on Borneo, uncertain placement, many of which have received New Caledonia, and Pacific Islands. The basal lin- little or no attention in of the are found in the re- phylogenetic studies. The eages grasses tropical tribe which includes the Australian gions of South America, Africa, and Asia; the An- Eriachneae, Eriachne and is omochlooideae are restricted to South and Central genera Pheidochloa, represented here an rbcL America & Soderstrom, the only by sequence of Eriachne triodioi- (Judziewicz 1989), des and an ITS sequence of E. triseta. The Pharoideae are genus pantropical (Soderstrom et al., and the Puelioideae are restricted to Micraira, the only member of the Australian tribe 1987), tropical is an ndhF se- Africa Micraireae, (Soderstrom & Ellis, 1987; Clark et al., represented only by 2000). Due to the absence of an early fossil record, quence of M. lazaridis and by an ITS sequence of it is not clear how M. Such as this distribution was established, subulifolia. genera Cyperochloa, Stey- whether across the and At- ermarkochloa, the were not in- by long-distance dispersal Crinipes group lantic and Indian Oceans, or whether across a cluded in con- this combined analysis. An rbcL se- tinuous Gondwanan equatorial forest. Either way, quence of Cyperochloa places it with the the continent of cannot be determined centothecoids, whereas a origination sequence of the crinipoid with current data. genus Styppeiochloa places it sister to Arundineae s. str. (Barker, 1997; Linder et al., 1997). Timing and causes of diversification. The ear- The subfamilies recognized within the PACCAD liest unequivocal grass fossils are pollen grains Clade are, except for Centothecoideae, strongly from the Paleocene of South America and Africa, 414 Annals of the Missouri Botanical Garden deposited approximately 60 to 55 million years consistent with available data, that the grasses (my) ago (Jacobs et al., 1999), although some grains achieved their Gondwanan distribution by dispersal of Monoporitesf rom the upper Maastrichtian (Cre- (Soreng & Davis, 1998), as has been suggested for taceous) may also represent remains of grasses other taxa with an apparent Gondwanan distribu- (Linder, 1987). The earliest known grass macrofos- tion (e.g., Adansonia, Baum et al., 1998; Atheros- sil appears in an early Eocene formation (ca. 55 permataceae, Renner et al., 2000). mya) in North America (Crepet & Feldman, 1991). The combination of fossil data and molecular Based on the fossil record therefore, the family is clock evidence suggests that the major diversifica- at least 55 my and possibly as much as 70 my old. tion of the grasses occurred between 15 and 25 Establishment of all major lineages had occurred mya, long after the origin of the family at 55 to 70 by the mid-Miocene (Jacobs et al., 1999), which is mya. This is consistent with the observed branch about the time that grass-dominatede cosystems ap- lengths on the phylogeny in Figure 1. There may peared. have been many more representatives of the An- Attempts to date particular nodes on the clado- omochlooideae, Pharoideae, and Puelioideae (or gram using molecular clocks are confounded by even additional lineages) that are now extinct, but non-clocklike behavior of several of the genes grasses are generally rare in the fossil record until (Gaut et al., 1996, 1997; Mathewse t al., 2000; Kel- the Miocene (Jacobs et al., 1999). The simplest ex- logg & Russo, unpublished obs.). Using sequences planation is that the family diversified long after its of GBSSI, which has been shown to exhibit clock- origin. The novel characters that arose after the di- like mutation, Gaut and Doebley (1997) placed the vergence of Joinvillea-the caryopsis, differentiat- divergence of maize and Pennisetum at 25 mya, ed embryo, reduction in perianth-therefore did whereas Kellogg and Russo (unpublished) place the not lead immediately or directly to the current dom- divergence of Danthoniopsisd interi from the rest of inance of the family. Other characteristics acquired the panicoids at ca. 16 mya. The two dates conflict later in the evolution of the family may have been with each other, but do suggest that the PACCAD more important in its diversification and ecological Clade originated in the early Miocene or late Oli- success. Possibilities include such characters as gocene. formation of intercalary meristems or the acquisi- All C4 lineages are included in the PACCAD tion of mechanisms for drought tolerance. We do Clade, so paleontological evidence for C4 photosyn- not know the phylogenetic distribution of interca- thesis can establish a minimum age for the common lary meristems, however, and it is possible that in- ancestor of the clade. The earliest known C4 grass tercalary meristems of the leaves evolved after such macrofossil is dated at 12.5 mya (Nambudiri et al., meristems in the stems. This character needs to be 1978), and the earliest isotopic evidence for C4 is investigated further. Acquisition of drought and ca. 15 mya (Kingston et al., 1994; Latorre et al., heat tolerance would also be worth investigating, 1997). This suggests that the origin of the PACCAD but would require a precise definition of what is Clade occurred no later than 15 mya and possibly meant by each term. The cellular components of as early as 25 mya. such physiological responses are being identified Both fossil data and molecular clock estimates and could perhaps be studied across a range of seem at odds with the apparent Gondwanan distri- taxa. bution of many grass taxa (see for example Simon & Jacobs, 1990). The Gondwanan distribution of CONCLUSIONS such derived groups as the subfamily Danthonioi- deae might suggest that the PACCAD Clade origi- We present here a resolved and strongly sup- nated sometime before the breakup of Gondwana, ported phylogeny of the grass family. It can be used which would then place the origin of the family long to understand the diversification of morphology, before the earliest known fossils were deposited. genes, and genomes, to interpret comparative stud- This cannot be ruled out, of course, because it is ies of cereal crops and forage grasses, and to de- an assumption based on negative evidence. If, how- velop hypotheses of adaptation to past and future ever, we assume that groups within the PACCAD environments. Some phylogenetic questions remain Clade originated before the breakup of Gondwana unresolved, and these affect inferences about such (a process hard to date precisely but perhaps 100- important characters as C4 photosynthesis. None- 70 mya), then we would have to assume that the theless, this phylogeny is one of the most compre- family originated more than 200-140 mya, before hensive and robust available for any family of the time of the first appearance of angiosperms in plants, making the grasses an excellent clade for the fossil record. It seems more likely, and more studies of evolutionary pattern and process. Volume 88, Number 3 Grass Phylogeny Working Group 415 2001 Phylogeny and Classification of Poaceae TAXONOMICT REATMENT Thysanolaeneae placed in the Centothecoideae and Gynerium as Incertae Sedis. Centropodia and Merx- Twelve subfamilies are recognized formally in muellera rangei are placed in Chloridoideae. Pooi- this classification system (Table 1). A description deae have grown by inclusion of Brachyelytreae, is provided for each subfamily, and where appro- Lygeeae, Nardeae, Phaenospermatideae, Diarrhe- priate, synonymy is indicated. To permit easy com- neae, Stipeae, and Ampelodesmeae, all formerly parison with previous work, we have listed for each classified within either Bambusoideae or Arundi- subfamily which of the tribes recognized by Clayton noideae by some authors; note, however, that Clay- and Renvoize (1986) are to be included. In some ton and Renvoize (1986) placed Lygeeae, Nardeae, cases (e.g., Pharoideae or Danthonioideae), the new and Stipeae in Pooideae in agreement with the clas- circumscription of subfamilies makes tribal recog- sification proposed here. A detailed comparison of nition largely unnecessary. For example, the sub- the GPWG classification with the major grass clas- family Pharoideae includes three genera in a single sification systems of the 20th century is presented tribe; the tribe is effectively redundant and serves in Table 1. no useful function in the subfamilial classification. Primary sources for suprageneric names were the Nonetheless we list the names for comparison. STAR Database (http://matrix.nal.usda.gov:8080/ Our sample of taxa was explicitly designed to star/supragenericname.html), the Catalog of New explore relationships among major clades that can World Grasses (http://mobot.mobot.org/W3T/ be recognized at the subfamilial level, but it is not Search/nwgc.html), and Clayton and Renvoize dense enough to evaluate tribal limits. We have in (1986). Diagnoses of the subfamilies were extracted many cases combined molecular data from several from various sources including Clayton and Ren- species to represent a genus (as is also commonly voize (1986) and Watson and Dallwitz (1992). done for morphological analyses), and in a few cas- Tribes in Chloridoideae and Panicoideae (except for es have combined data from several genera that the exclusion of Eriachneae) follow the treatment represent a putatively monophyletic group. Such of Clayton and Renvoize (1986); tribes listed for combinations assume, rather than test, monophyly. the other subfamilies generally are treated accord- We therefore refrain from formal discussion of tribal ing to more recent studies and/or consultation with limits, which cannot be addressed by our (lata; specialists in those groups. these limits will have to be re-evaluated by future studies. Three tribes and two genera are placed as Poaceae (R. Br.) Barnh., Bull. Torrey Bot. Club Incertae Sedis at the end of the classification, al- 22: 7. 1895. (Nom. alt. Gramineae Juss., Gen. though the genera may be provisionally placed as P1.: 28. 1789.) noted below. A the fol- This classification reflects our monophyletic family, recognizable by attempt to use the as the basis for subfamilies lowing synapomorphic morphological characters: phylogeny recognizing Inflorescence bracteate. Perianth reduced or while remaining nomenclaturally conservative. Ex- highly Pollen for scrobiculi, but with intraex- cept Centothecoideae, all subfamilies lacking. lacking recog- inous channels. Seed coat fused to inner nized are well as ovary wall supported monophyletic in our at While we could create an unranked clas- maturity, forming a caryopsis. Embryo highly dif- analyses. ferentiated with obvious shoot and root mer- sification for the grasses using our leaves, phylogeny, we and lateral in feel that the practical interests of the istems, position. potential us- ers of this classification currently are best served I. Anomochlooideae Pilg. ex Potztal, in Willd- by retaining the Linnaean hierarchy. Nonetheless enowia 1: 772. 1957. TYPE: Anomochloa we have applied informal names to several of the Brongn. Figure 4C and D. well-supported clades (see above). The most significant changes in our proposed Syn.: Streptochaetoideae (Nakai) Butzin, Neue Unters. Blute Gram.: 148. 1965. subfamily classification are the breakup of the tra- ditional Bambusoideae and Arundinoideae and the Plants perennial, rhizomatous, herbaceous, of expansion of Pooideae. The diversity encompassed shaded tropical forest understories. Culms hollow by the traditional Bambusoideae (or Bambusoideae or solid. Leaves with phyllotaxis either distichous s.l.) is now recognized as Anomochlooideae, Phar- or spiral; abaxial ligule absent; adaxial ligule a oideae, Puelioideae, Bambusoideae s. str., and Ehr- short fringe of cilia or absent, not membranous; hartoideae. Elements of the traditional Arundino- blades usually relatively broad, venation parallel, ideae are now recognized as Aristidoideae, with pseudopetioles short to very long, these with Danthonioideae, and Arundinoideae s. str., with dark, turgid swellings (pulvini) at both ends (An- 416 Annals of the Missouri Botanical Garden omochloa) or only at the summit (Streptochaeta); II. Pharoideae (Stapf) L. G. Clark & Judz., Taxon sheaths non-auriculate. Inflorescences spicate, with 45: 643. 1996. TYPE: Pharus P. Browne. Fig- complicated branching patterns, bracts outside of ure 4B. the spikelet equivalents present, large and with a blade or small and bladeless, or absent. Ultimate Syn.: Leptaspidoideae (Tzvelev) C. 0. Morales, Sendtnera 5: 244. 1998. Nom. structures of the inflorescence superfl. (spikelet equiva- lents) of uncertain homology with typical grass Plants perennial, rhizomatous, monoecious, her- spikelets but one-flowered and bisexual; bracts baceous, of shaded tropical to warm temperate for- within the spikelet equivalents with phyllotaxis dis- est understories. Culms hollow or solid. Leaves dis- tichous or spiral, lacking uncinate macrohairs, tichous; abaxial ligule absent; adaxial ligule a sometimes awned but if so, the awns single; lodi- fringed membrane; blades resupinate, relatively cules absent, or, in Anomochloa, their position oc- broad, with pseudopetioles prominent and twisted, cupied by a ring of short brownish cilia borne on with lateral nerves diverging obliquely from mid- a low membranous ring; stamens 4 or 6; ovary gla- nerve and running straight to margins; sheaths non- brous, apical appendage absent, haustorial syner- auriculate. Inflorescences paniculate, the main axis gids presumed absent, style 1, stigma(s) 1 or 3. and branches disarticulating or not, covered with Caryopsis with the hilum linear, shallow and incon- uncinate macrohairs, bracts outside of the spikelets spicuous; endosperm hard, containing compound absent. Spikelets unisexual, one-flowered, mostly in starch grains; embryo large, epiblast present or not, male-female pairs on short branchlets, or some fe- scutellar cleft present but shallow, mesocotyl inter- male spikelets solitary. Female spikelets large, node absent, embryonic leaf margins overlapping short-stalked; glumes 2, shorter than the floret; or not. Basic chromosome numbers: x = 11 or 18 lemma tubular or inflated, covered wholly or in part (note: Clark & Judziewicz, 1996, erroneously cited by uncinate macrohairs, awnless; palea well devel- these as 12 or 18). oped; lodicules absent; ovary glabrous, apical ap- pendage absent, haustorial synergids ptresumed ab- Foliar anatomy. Mesophyll nonradiate, an ad- sent, style 1, stigmas 3. Caryopsis with the hilum axial palisade layer absent, with fusoid cells very linear, extending the full length; end(osperm hard, large and well developed, arm cells only weakly without lipid; embryo small, epiblast present, scu- developed; Kranz anatomy absent; midrib complex; tellar cleft present but shallow, mesocotyl internode adaxial bulliform cells present. absent, embryonic leaf margins overlapping. Male Foliar Stomata with low spikelets small, short- to long-stalked, membra- micromorphology. nous; glumes 2, shorter than the floret; lodicules 3 dome-shaped and triangular subsidiary cells; bi- or if cellular microhairs the 0, present then minute, elliptic, glabrous, and very large (0.075-0.15 mm), cell one and a half times as nerveless; stamens 6. Basic chromosome number: pointed apical usually x= 12. long as the basally constricted basal cell; papillae absent. Foliar anatomy. Mesophyll nonradiate, an ad- axial palisade layer absent, fusoid cells and Photosyntheticp athway. Presumed large C3. well developed, arm cells weakly to moderately well developed; Kranz anatomy absent; midrib complex; INCLUDED TRIBES: inflated adaxial interstomatal cells present, bulli- form cells Anomochloeae C. E. in Fam. poorly developed or absent. Hubb., Hutchinson, Fl. P1. 2: 219. 1934. TYPE: Anomochloa Foliar micromorphology. Stomata with parallel- Brong. sided to dome-shaped subsidiary cells; bicellular Streptochaeteae C. E. Hubb., in Hutchinson, Fam. microhairs and papillae absent. Fl. PI. 2: 205. 1934. TYPE: Streptochaeta Presumed Schrad. ex Nees. Photosynthetic pathway. C3. Notes. There is no unique morphological syn- INCLUDED TRIBE (NOW IDENTICAL TO SUBFAMILY apomorphyf or this subfamily, but both tribes lack AND THUS REDUNDANT): lodicules and they apparently also lack grass-type Phareae in As noted above Stapf, Thiselton-Dyer, Fl. Cap. 7: 319. spikelets. (Unresolved Questions), 1898. TYPE: Pharus P. Browne. this lineage may not be monophyletic, in which case two subfamilies would need to be recognized. Notes. In his original description of the tribe, The subfamily includes 4 species. Stapf specifically included Olyra (based on its uni- Volume 88, Number 3 Grass Phylogeny Working Group 417 2001 Phylogeny and Classification of Poaceae sexual spikelets), but did not explicitly list Pharus axial palisade layer absent, fusoid cells well de- or Leptaspis, although his choice of the name Phar- veloped, arm cells only weakly developed; Kranz eae implicitly recognized the membership of Pha- anatomy absent; midrib complex or less commonly rus in the tribe and automatically placed Pharus as simple; adaxial bulliform cells present. its type, according to Article 10.6 of the Code Foliar Stomata with dome- (Greuter et al., 2000). As long as Olyra was re- micromorphology. to cells; microhairs tained in the same tribe as shaped triangular subsidiary Pharus, Phareae was a absent or multicellular, uniseriate micro- superfluous name for the Olyreae. When Pharus (Puelia) hairs present (Guaduella);p apillae present or more and Leptaspis are segregated into their own tribe, absent. and Olyra is excluded, then Phareae becomes the commonly valid, correct name for the tribe. Clark and Jud- Photosynthetic pathway. Presumed C3. ziewicz (1996) based the name of the subfamily on this tribal name. Tzvelev (1989) argued that the INCLUDED TRIBES: name Phareae was illegitimate because the type of Guaduelleae Soderstr. & R. P. Ellis, in Soderstrom the previously described tribe Olyreae was includ- et al. (editors), Grass Syst. Evol.: 238. 1987. ed in it, and provided the name Leptaspideae for TYPE: Guaduella Franch. this tribe. Morales (1998) agreed with Tzvelev and Puelieae Soderstr. & R. P. Ellis, in Soderstrom et rejected the name Pharoideae for this subfamily, al. (editors), Grass Syst. Evol.: 238. 1987. according to Article 52.1 of the Code (Greuter et TYPE: Puelia Franch. al., 2000), replacing it with Leptaspidoideae. Under Article 52.3, however, "A name that was nomen- Notes. This subfamily, which comprises ap- claturally superfluous when published is not ille- proximately 14 species, is poorly known, and mor- gitimate ... if it is based on the stem of a legitimate phological, anatomical, cytological, and ecological generic name." We therefore accept the name Phar- studies are needed. oideae for this subfamily, as Pharus is a legitimate generic name. The subfamily includes 12 IV. Bambusoideae Luerss., Grundz. Bot., ed. 5: species. 451. 1893. TYPE: Bambusa Schreb. Figures III. Puelioideae L. G. Clark, M. Kobay., S. Ma- 4F, 6C, G, O, P, 7A. thews, Spangler & E. A. Kellogg, Syst. Bot. 25: Syn.: ()lyroideae Pilger, Nat. Pfl.-Fam. ed. 2, 14(1: 168. 181-187. 2000. TYPE: Puelia Franch. Figure 1956. 4A. Parianoideae (Nakai) Butzin, Neue Unters. lIiite (ram.: 148. 1965. Plants perennial, rhizomatous, herbaceous, of Plants shaded rainforest understories. Culms hollow. perennial (rarely annual), rhizomatous, herbaceous or of and for- Leaves distichous; abaxial woody, temperate tropical ligule absent (Guaduel- montane la) or present (Puelia); adaxial ligule a ests, tropical high grasslands, riverbanks, fringed and sometimes savannas. Culms hollow or solid. membrane; blades relatively broad, pseudopetiola- Leaves abaxial absent te, venation parallel; sheaths non-auriculate. Inflo- distichous; ligule (Olyreae) or present (Bambuseae);a daxial ligule membranous rescences racemose or paniculate, bracts outside of or the sometimes with two chartaceous, fringed or unfringed;b lades usually spikelets present. Spikelets and several the 1 to 3 flo- relatively broad, pseudopetiolate, venation parallel; glumes florets, proximal sheaths often auriculate. Inflorescences ra- rets male, the next several florets spicate, female-fertile, cemose or with distal paniculate, completing development of incomplete florets (Guaduella), or the all spikelets in one period of growtha nd subtending proximal 3 to 6 florets male or neuter with the sin- bracts and distal floret female prophylls usually absent, or pseudo- gle (Puelia), disarticulating with basal bracts above the glumes and between the florets spikelets bud-bearing producing (Guad- two or more orders of spikelets with different or not phas- uella) (Puelia); lemmas lacking uncinate ma- es of and well sometimes maturity subtending bracts and prophylls crohairs, awnless; palea developed, lodicules stamens usually present. Spikelets (or spikelets proper of tubular; 3, membranous, ciliate; the pseudospikelets) bisexual (Bambuseae) or uni- 6; ovary glabrous or hairy, an apical appendage sexual of 2 or several present or not, haustorial synergids presumed ab- (Olyreae), consisting 0, 1, glumes, 1 to many florets; lemma lacking uncinate sent, styles 2 or 3, the bases close, stigmas 2 or 3. macrohairs, if awned, the awns single; palea well Caryopsis with a long-linear hilum; embryo small. Basic chromosome number: x = 12. developed; lodicules usually 3 (rarely 0 to 6 or many), membranous, vascularized, often ciliate; Foliar anatomy. Mesophyll nonradiate, an ad- stamens usually 2, 3, or 6 (10 to 40 in Pariana, 6 418 Annals of the Missouri Botanical Garden to 120 in Ochlandra); ovary glabrous or hairy, ittate (Phyllorachideae), somewhat broad to usually sometimes with an apical appendage, haustorial narrow, sometimes pseudopetiolate, venation par- synergids absent, styles 2 or 3, sometimes very allel; sheaths sometimes bearing auricles. Inflores- short but close, stigmas 2 or 3. Caryopsis with hi- cences paniculate or racemose, bracts outside of lum linear (or rarely punctate), extending its full the spikelets rarely present (Humbertochloa). length (or rarely less than full length); endosperm Spikelets bisexual or unisexual, with glumes 2 (ab- hard, without lipid, containing compound starch sent in some Oryzeae), sterile florets 0 to 2, and grains; embryo small, epiblast present, scutellar female-fertile floret 1, disarticulating above the cleft present, mesocotyl internode absent, embry- glumes or infrequently primary branches disartic- onic leaf margins overlapping. Basic chromosome ulating as units; lemma lacking uncinate macro- numbers: x = 7, 9, 10, 11, and 12. hairs, if awned, the awn single; palea well devel- oped; lodicules 2, membranous or Foliar anatomy. Mesophyll nonradiate, an ad- rarely fleshy, vascularized; stamens 3 or 6 axial absent, fusoid cells and heavily usually (some- palisade layer large times or well 1, 2, 4); ovary glabrous, apical appendage developed, arm cells usually well developed haustorial and Kranz absent, absent; synergids absent, styles 2, free, strongly invaginated; anatomy fused or for their full midrib or adaxial bulliform cells basally length (Zizaniopsis), complex simple; close, stigmas 2. Caryopsis with the hilum long- present. linear; endosperm hard, without lipid, containing Foliar micromorphology. Stomata with dome- compound starch grains (rarely simple); embryo shaped, triangular, or parallel-sided subsidiary small, epiblast usually present (absent in Ehrhar- cells; bicellular microhairs present, panicoid-type; ta), scutellar cleft usually present (absent in Leersia papillae common and abundant. and Potamophila), mesocotyl internode absent (pre- sent but short in C3. Microlaena), embryonic leaf usu- Photosynthetic pathway. ally with overlapping margins (meeting in Pota- mophila). Basic chromosome numbers: x = 12 INCIUDED TRI:LS: (10 in Microlaena; 15 in Zizania). Bambuseae Dumort., Anal. Fam. Pi.: 63. 1829. Foliar anatomy. Mesophyll nonradiate, an ad- TYPE: Bamuttsa Schreb. axial Kunth ex Fl. 1: 172. palisade layer usually absent, fusoid cells ab- Olyreae Spenn., Friburg. sent or sometimes and 1825. TYPE: L. present (Zizania Zizaniop- Olyra (Including Buerger- sis), arm cells absent or siochloeae Blake, Blumea, present; Kranz anatomy Suppl. 3: 62. 1946; midrib or adaxial bulliform Parianeae C. E. Hubbard, in Hutch., Fam. Fl. absent; simple complex; cells P1. 2: 219. present. 1934.) Foliar micromorphology. Stomata with dome- Notes. The current circumscription of this sub- shaped or triangular subsidiary cells; bicellular mi- family is much narrower than the traditional view. crohairs In their recent and Clark present, panicoid-type; papillae often pre- analysis, Zhang (2000) sent in Oryzeae, otherwise absent. recovered two robustly supported clades, the oly- roid bamboos and the woody bamboos, which they Photosynthetic pathway. C3. recognized as tribes Olyreae and Bambuseae, re- spectively. Following Zhang and Clark (2000), INCLUDED TRIBES: Buergersiochloeae and Parianeae are included in Ehrharteae Nevski, Trudy Bot. Inst. Akad. Nauk Olyreae. This subfamily includes approximately SSSR 4: 227. 1937. TYPE: Ehrharta Thunb. 1200 species. Oryzeae Dumort., Observ. Gramin. Belg.: 83. 1824. TYPE: Oryza L. V. Ehrhartoideae Link, Hort. Berol. 1: 233. Phyllorachideae C. E. Hubb., in Hook. Ic. P1. 34: 1827. TYPE: Ehrharta Thunb. Figure 4G. t. 3386, p. 5. 1939. TYPE: Phyllorachis Tri- Syn.: Oryzoideae Kunth ex Beilschm., Flora 16(2): 52, men. 109. 1833. Notes. Although we did not sample Phyllorach- Plants annual or perennial (rhizomatous or sto- ideae, we place it here based on morphological sim- loniferous), herbaceous to suffrutescent, of forests, ilarity. Nonetheless, any future studies of this clade open hillsides, or aquatic habitats. Culms hollow or should include this tribe to test its relationship to solid. Leaves distichous; abaxial ligule absent; ad- Ehrharteae and Oryzeae. Under the present circum- axial ligule a fringed or unfringed membrane, or a scription, this subfamily includes approximately fringe of hairs; blades rarely basally cordate or sag- 120 species. Volume 88, Number 3 Grass Phylogeny Working Group 419 2001 Phylogeny and Classification of Poaceae VI. Pooideae Benth., Fl. Hongk. 407. 1861. or circular and less than 1/3 the length of the fruit; TYPE: Poa L. Figures 4E, H, 6D, J, L, 7B. endosperm hard or sometimes soft or liquid (some Poeae), with or without lipids (some Poeae), con- Syn.: Avenoideae Link, Hort. Berol. 1: 108. 1827. Festucoideae Link, Hort. Berol. 1: 137. 1827. taining compound starch grains, or simple starch Glycerioideae Link, Hort. Berol. 1: 160. 1827. grains (Brachyelytreae, Bromeae, Triticeae, some Echinarioideae Link, Hort. Berol. 1: 197. 1827. Stipeae); embryo small, epiblast present (rarely ab- Cynosuroideae Link, Hort. Berol. 1: 198. 1827. sent), scutellar cleft absent (rarely present, but not AnthoxanthoideaeL ink, Hort. Berol. 1: 232, 271. 1827. deeply incised), mesocotyl internode absent (rarely Agrostidoideae Kunth ex Beilschm., Flora (Beib.) 16(2): 52, 104. 1833. short, Brachyelytrum), embryonic leaf margins Stipoideae Burmeist., Handb. Naturgesch. 199. 1837. meeting (infrequently margins overlapping). Basic Hordeoideae Burmeist., Handb. Naturgesch.2 02. 1837. chromosome numbers: x = 7 (Bromeae, Triticeae, Phalaroideae Burmeist., Handb. Naturgesch. 208. Poeae 1837. generally, few Brachypodieae), 2, 4, 5, 6, 8, Secaloideae Fl. France 14: 13 in a few Poeae and Rouy, 2, 298. 1913. 9, 10, 11, 12, represented the other tribes, generally medium or large. Plants annual or perennial (rhizomatous, stolon- iferous, or neither), herbaceous, of cool Foliar an ad- temperate anatomy. Mesophyll nonradiate, and boreal regions, extending across the tropics in axial palisade layer absent, fusoid cells absent, arm the cells Kranz midrib high mountains. Culms hollow (rarely solid). absent; anatomy absent; simple; Leaves distichous; abaxial ligule absent; adaxial adaxial bulliform cells present. ligule scarious or membranous, the margin not or Foliar micromorphology. Stomata with parallel- infrequently short ciliate fringed (rarely long cili- sided cells; bicellular microhairs absent ate, Anisopogon); blades somewhat broad to subsidiary usually (rarely present, Lygeum, where chloridoid, Nardus, narrow, rarely pseudopetiolate (Phaenosperma), ve- where unicellular microhairs absent nation parallel; sheaths sometimes auriculate. In- panicoid), few ab- florescences (rarely present, Stipeae); papillae usually spicate, racemose, or paniculate, sent, when more than one bracts outside of the present rarely per long spikelets absent or rarely pre- cell. sent (e.g., Sesleria, Echinaria, Ammochloa). Spike- lets bisexual, infrequently unisexual or mixed, usu- Photosynthetic pathway. C3. ally with two glumes (rarely without glumes, Lygeum, or the first absent, Hainardia, Lolium, INCI,LUDI)E! TRIEKS: Nardus, except on terminal spikelets), 1 to many female-fertile florets with apical or infrequently Ampelodesmeae (Conert) Tutin, Bot. J. Linn. Soc. basal reductions, compressed laterally, infrequently 76: 369. 1978. TYPE: Ampelodesmos Link. not or dorsally compressed, disarticulating above Brachyelytreae Ohwi, Bot. Mag. Tokyo 55: 361. the glumes (infrequently below the glumes, some 1941. TYPE: Brachyelytrum P. Beauv. Poeae, or at the nodes of the inflorescence, various Brachypodieae (Hack.) Hayek, Oesterr. Bot. Z. genera); lemma lacking uncinate macrohairs, if 74(10): 253. 1925. TYPE: Brachypodium P. awned, the awn single; palea usually present and Beauv. well developed, but variable and sometimes very Bromeae Dumort., Observ. Gramin. Belg.: 83. reduced or absent; lodicules 2 (rarely 3, Anisopo- 1824. TYPE: Bromus L. gon, Ampelodesmeae, many Stipeae and few Poeae; Brylkinieae Tateoka, Canad. J. Bot. 38: 962. 1960. fused, Meliceae; rarely absent, Lygeum, Nardus, TYPE: Brylkinia F. Schmidt. and few Poeae), usually lanceolate, broadly mem- Diarrheneae (Ohwi) C. S. Campb., J. Arnold Arbor. branous apically (fleshy, truncate, Meliceae), often 66: 188. 1985. TYPE: Diarrhena P. Beauv. lobed (Triticeae, Poeae), obscurely few-nerved, or Lygeeae J. Presl, Wsobecny Rostl. 2: 1708, 1753. infrequently + distinctly few-nerved, not or con- 1846. TYPE: Lygeum Loefl. ex L. spicuously ciliate on the margins; stamens usually Meliceae Link ex Endl., Fl. Poson.: 116. 1830. [as 3 (infrequently 1 or 2); ovary glabrous or pubes- "Melicaceae"] TYPE: Melica L. cent, rarely with an apical appendage (Bromus, Nardeae W. D. J. Koch, Syn. Fl. Germ. Helv.: 830. Diarrhena) or rostellum (e.g., Brachyelytrum, Ros- 1837. TYPE: Nardus L. traria), haustorial synergids absent, styles usually Phaenospermatideae Renvoize & Clayton, Kew 2, close, stigmas 2 (rarely 1, Lygeum, Nardus, and Bull. 40: 478. 1985. TYPE: Phaenosperma a few others, or 3, scattered genera). Caryopsis with Munro ex Benth. the hilum linear and up to as long as the fruit, or Poeae R. Br., Voy. Terra Austral. 2: 582. 1814. subbasal and punctiform, linear, ellipsoidal, ovate, TYPE: Poa L. (Including Aveneae Dumort., 420 Annals of the Missouri Botanical Garden Observ. Gramin. Belg.: 82. 1824; Agrostideae fusoid cells absent, arm cells absent; Kranz anat- Dumort., Observ. Gramin. Belg.: 83. 1824.) omy absent (Sartidia) or present (Stipagrostis, Ar- Stipeae Dumort., Observ. Gramin. Belg.: 83. 1824 istida), when present with one (Stipagrostis) or two [as "Stipaceae"]. TYPE: Stipa L. (Aristida) parenchyma sheaths, although both not Triticeae Dumort., Observ. Gramin. Belg.: 82, 84, equally well developed throughout the genus; mid- 91. 1824. TYPE: Triticum L. rib simple; adaxial bulliform cells present. Notes. Relationships among some of the major Foliar micromorphology. Stomata dome-shaped lineages of the core Pooideae clade remain unre- or triangular; bicellular microhairs present, pani- solved, and conflicts between molecular data and coid-type; papillae absent. morphologically based tribal classifications exist C3 C4 (e.g., Poeae vs. see Photosynthetic pathway. (Sartidia); (Ar- Aveneae; Soreng & Davis, istida, NADP-ME; not 2000). This is one of several reasons that we do not Stipagrostis, biochemically but NAD-ME; & offer a formal classification of tribes at this typed, anatomically Hattersley point. Watson, 1992). Relationships among the earlier diverging lineages of the whole pooid clade are only weakly supported, INCLUDED TRIBE IDENTICAL TO THE and also require further investigation. The tribal (NOW SUBFAMILY AND THUS classification presented here is almost certain to REDUNDANT): change as additional data accumulate, and thus Aristideae C. E. Hubbard, in Bor, Grasses Burma, should be taken only as an indication of the taxa Ceylon, India & Pakistan: 685. 1960. TYPE: included within the subfamily. The subfamily in- Aristida L. cludes approximately 3300 species. Notes. The presence of a basal column of the awn is a potential morphological synapomorphy for VII. Aristidoideae Caro, Dominguezia 4: 16. this clade. Sartidia diverges from Stipagrostis and 1982. TYPE: Aristida L. Figure 5C. Aristida in other respects, and should be sampled Plants annual or herba- in future analyses. The subfamily includes perennial, caespitose, approx- 350 ceous, xerophytic or less commonly mesophytic, of imately species. temperate, subtropical and tropical zones, often in VIII. Arundinoideae Handb. Natur- open habitats. Culms solid or hollow. Leaves dis- Burmeist., tichous; abaxial ligule absent or present as a line gesch.: 204. 1837. TYPE: Arundo L. Figure of hairs; adaxial ligule a fringed membrane or a 5A. fringe of hairs; blades relatively narrow, without Syn.: PhragmitoideaeP arodi ex Caro, Dominguezia4 : 13. pseudopetioles, venation parallel; sheaths non-au- 1982. riculate. Inflorescences paniculate, bracts outside of the spikelets absent. Spikelets with bisexual flo- Plants perennial (rarely annual), rhizomatous, female-fertile floret and no rach- stoloniferous, or caespitose, herbaceous to some- rets, glumes 2, 1, illa or what woody, of temperate and tropical areas, me- extension, cylindrical laterally compressed, above the lemma with three sophytic or xerophytic, the reeds found in disarticulating glumes; marshy the awns habitats. Culms hollow or less solid. awns, separate from each other, or fused commonly below into a twisted Leaves column; short, less than distichous; abaxial ligule absent or palea rarely half the lemma lodicules or present as a line of hairs length; present rarely (Hakonechloa); adaxial absent, when ligule a fringed or unfringed membrane or a present 2, free, membranous, gla- fringe brous, heavily vascularized; stamens 1 to of hairs; blades 3; relatively broad to narrow, without ovary haustorial pseudopetioles, venation parallel; sheaths glabrous, apical appendage absent, syn- usually 2. non-auriculate. Inflorescences ergids absent, styles 2, free, close, stigmas Cary- usually paniculate, with the hilum short or endo- rarely spicate or racemose, bracts outside of the opsis long-linear; without spikelets absent. with bisexual hard, florets, sperm lipid, containing compound Spikelets starch 2, a sterile lemma sometimes fe- grains; embryo small (Sartidia) or large (Ar- glumes present, male-fertile florets 1 to reduction istida, Stipagrostis), epiblast absent, scutellar cleft several, apical or absent dis- present (Sartidia), mesocotyl internode usually present, usually laterally compressed, above the lemma un- elongated, embryonic leaf margins meeting. Basic articulating glumes; lacking chromosome numbers: x = 12. cinate 11, macrohairs, if awned, awn usually single, sometimes awns three, but then lacking a basal col- Foliar anatomy. Mesophyll radiate or nonra- umn; palea usually well developed; lodicules 2, diate (Sartidia), an adaxial palisade layer absent, free (rarely joined at the base), fleshy, glabrous or Volume 88, Number 3 Grass Phylogeny Working Group 421 2001 Phylogeny and Classification of Poaceae infrequently ciliate, not or scarcely vascularized to monophyletic arundinoid clade, although Linder et heavily vascularized; stamens (1 to)3; ovary gla- al. (1997) linked Arundo, Phragmites, and Molinia brous, apical appendage absent, haustorial syner- by the presence of hollow culm internodes, a punc- gids absent, styles 2, usually free, close, stigmas 2. tiform hilum, and convex adaxial rib sides in the Caryopsis with the hilum short or long-linear (Mol- leaf blade. This subfamily clearly requires further inia); endosperm hard, without lipid, containing study. The subfamily includes 33 to 38 species, compound starch grains; embryo large or small counting the crinipoids. (Amphipogon),e piblast absent, scutellar cleft pre- sent, mesocotyl internode elongated, embryonic leaf IX. Danthonioideae Barker & H. P. Linder, sub- margins meeting or overlapping (Hakonechloa).B a- fam. nov. TYPE: Danthonia DC. Fl. Franc. 3: sic chromosome numbers: x = 6, 9, 12. 32. 1805. Figure 5F. Foliar anatomy. Mesophyll nonradiateo r rarely Haec subfamiliaa b aliis subfamiliisP oacearumsy ner- radiate (Arundo,A mphipogon), without an adaxial gidis haustorialibusli,g ula ciliata, embryonem esocotyle- palisade layer, without fusoid cells, arm cells ab- done praedito,s picula pluriflorav el si uni- vel biflora nunc rhachillai n extensionem sent or Kranz ab- desinente,s tylorum b asi- present (Phragmites); anatomy bus plerumqued istantibuas tquea natomia" Kranze"t mi- sent; midrib simple; adaxial bulliformc ells present. cropilisc hloridoideics arentibusb ene distincta. Distinctf romt he others ubfamilieso f the grasses the Foliar micromorphology. Stomata with low by haustoriasl ynergids,a nd by the conjunctiono f a ciliate dome-shaped or triangular subsidiary cells; bicel- ligule, the presenceo f an embryom esocotyl,a several- lular microhairs present or less commonly absent, flowereds pikelet,w hich,i f 1- or 2-floweredh, as a rachilla when present of panicoid-type except in Amphipo- extensionu, suallyd istinctlys eparateds tyle bases, t he ab- sence of Kranza natomya, nd the absence of chloridoid gon, which has unique microhair morphology;p a- microhairs. pillae absent except in Amphipogon. Plants perennial (caespitose, rhizomatouso r sto- Photosyntheticp athway. C,. loniferous) or less commonly annual, herbaceous or rarely suffrutescent, of mesic to xeric open habitats INCLUDEI) TRIBE (NOW II)FNTICA, T() SU3BFAMIL,Y in grasslands, heathlands, and open woodlands. AND THUS REDUNI)ANT): Culms solid or very rarely hollow. Leaves disti- Arundineae Dumort., Obs. Gram. Belg.: 82. 1824. chous; abaxial ligule usually absent (sometimes TYPE: Arundo L. present in Cortaderia, Karroochloa, and Pentas- chistis); adaxial ligule a fringe of hairs or a fringed Notes. The traditional Arundinoideaew ere well membrane; blades relatively narrow, without a known as a dustbin group (e.g., Clayton & Renvo- pseudopetiole, venation parallel; sheaths not auric- ize, 1986; Kellogg & Campbell, 1987). A number ulate except in Pentameris thuarii. Inflorescences of studies indicated that this subfamily as tradi- paniculate or less commonly racemose or spicate, tionally circumscribed was polyphyletic (e.g., Bark- bracts outside of the spikelets absent (but the sub- er et al., 1995; Clark et al., 1995), although some tending leaf + spatheate and disarticulating with support for a monophyletic Arundinoideae (includ- the inflorescence in Triboliump usillum). Spikelets ing Arundinoideae s. str., Danthonioideae, Aristi- bisexual (but sometimes without bisexual florets in doideae, Micraira, and Eriachne) was found by Cortaderia) or unisexual (Cortaderia, Lamprothyr- Hsiao et al. (1999). The results of the combined sus), glumes 2 and usually equal, female-fertile flo- analysis presented here suggest that a monophyletic rets 1 to 6(to 20), with apical reduction and a rach- core arundinoid group does exist, even though in- illa extension usually present, laterally compressed, dividual data sets do not strongly supportt he group. disarticulating above the glumes and between the The exact generic membership of the subfamily re- florets, less commonly below the glumes; lemma mains to be determined; however, we include the lacking uncinate macrohairs, awn single and from following genera: Amphipogon, Arundo, Dregeo- a sinus; palea well developed, sometimes relatively chloa, Hakonechloa, Molinia (and Moliniopsis if short; lodicules 2, free (rarely joined), fleshy or recognized), and Phragmites. We provisionally rarely with an apical membranousf lap, glabrous or place the crinipoid group (Crinipes, Dichaetaria, ciliate, often with microhairs, sometimes heavily Elytrophorus, Leptagrostis, Nematopoa, Piptophyl- vascularized; stamens 3; ovary glabrous or rarely lum, Styppeiochloa, and Zenkeria) here as well, with apical hairs (Pentameris), apical appendage based on molecular evidence from Linder et al. absent, haustorial synergids present, only weakly (1997) and Barker (1997). No morphological syn- developed in a few taxa, styles 2, the bases usually apomorphies have been identified to support the widely separated, stigmas 2. Caryopsis with the hi- 422 Annals of the Missouri Botanical Garden lum short or long-linear; endosperm hard, contain- perate woodlands and tropical forests. Culms solid ing compound starch grains (simple in Prionan- or hollow. Leaves distichous; abaxial ligule absent thium); embryo large or small, epiblast absent, or present as a line of hairs (Calderonella, Thysan- scutellar cleft present, mesocotyl internode elon- olaena); adaxial ligule membranous or ciliate, or gated, embryonic leaf margins meeting (overlapping membranousw ith ciliate margins; blades relatively in Danthonia decumbens).B asic chromosome num- broad to narrow, often pseudopetiolate, venation bers: x = 6, 7, 9. parallel; sheaths sometimes auriculate. Inflores- cences racemose or bracts outside of Foliar anatomy. Mesophyll nonradiate, an ad- paniculate, the spikelets absent. bisexual or unisex- axial palisade layer absent, fusoid cells absent, arm Spikelets to cells ual, absent; Kranz absent; midrib (1 to)2- many-flowered w ith reduction either anatomy simple, above or below the fertile florets, often compressed usually with one bundle, an arc of bundles in Cor- laterally; lemma lacking uncinate if taderia; adaxial bulliform cells present or absent. macrohairs, awned, the awn single; palea usually well devel- Foliar micromorphology. Stomata with dome- oped, sometimes relatively short; lodicules 2 or ab- shaped or parallel-sided subsidiary cells (rarely sent, + cuneate, many-nervedo r less commonlyn ot high dome-shaped or slightly triangular);b icellular or scarcely vascularized; stamens (1 to)2 or 3; ovary microhairs present, panicoid-type, sometimes ab- glabrous, apical appendage absent, haustorial syn- sent; papillae usually absent but often present in ergids presumed absent, styles 2, free or fused, Chionochloa and Merxmuellera. close, stigmas 2. Caryopsis with the hilum basal, C3. punctiform;e ndosperm hard, without lipid, contain- Photosyntheticp athway. ing simple or compound starch grains; embryo small or large, the epiblast present, scutellar cleft INCLUDED TRIBE (NOW IDENTICAL TO SUBFAMILY AND THUS present, mesocotyl internode present, embryonic REDUNDANT): leaf margins overlapping. Basic chromosome num- Danthonieae Zotov, New Zealand J. Bot. 1 (1): 86. ber: x = 12 (x = 11 or 12? in Thysanolaena). 1963. (Including CortaderieaeZ otov,N ew Zea- land J. Bot. 1 (1): 83. 1963.) TYPE: Danthonia Foliar anatomy. Mesophyll nonradiate, often DC. with an adaxial palisade layer, fusoid-like cells fre- quently present as extensions of the outer paren- Notes. The presence of haustorial synergids in chyma bundle sheath, arm cells absent; Kranza nat- the ovule and distant styles support the monophyly omy absent; midrib simple; adaxial bulliform cells of this clade (Verboome t al., 1994). Bilobed pro- large. phylls also may be a synapomorphy,b ut this feature has not been investigated sufficiently in the rest of Foliar micromorphology. Stomata with dome- the The results of this indicate robust shaped and/or family. study triangulars ubsidiary cells; bicellular molecular for the of this clade microhairs support monophyly present, panicoid-type; papillae absent. (excluding Centropodiaa nd Merxmuellerar angei), Photosyntheticp athway. C3. but its placement within the larger PACCAD Clade is equivocal. Pending further studies of the diver- INCLUDED TRIBES: sity of the danthonioid grasses, we recognize only one tribe, which includes the following genera (sen- Centotheceae Ridl., Mat. Fl. Malay Pen. 3: 122. su Barker et al., 2000): Austrodanthonia, Chaeto- 1907. TYPE: CentothecaP . Beauv. bromus, Chionochloa, Cortaderia, Danthonia, Joy- Thysanolaeneae C. E. Hubb., in Hutch., Fam. Fl. cea, Karroochloa, Lamprothyrsus, Merxmuellera P1. 2: 222. 1934. TYPE: Thysanolaena Nees. (minus M. rangei), Notochloe, Notodanthonia, Pen- Notes. for the of this sub- tameris, Pentaschistis, Plinthanthesis, Prionan- Support monophyly family as recognized here is moderate, and no mor- thium, Pseudopentameris,R ytidosperma,S chismus, have been identified. and Tribolium. The subfamily includes phological synapomorphies approxi- The sister relationship between the centothecoid mately 250 species. and panicoid clades, however, is relatively robust. The of and X. Centothecoideae Soderstr. "Centostecoi- positions Gynerium Danthoniopsis are [as unstable. A majority of the Centotheceae are char- deae"], Taxon 30: 615. 1981. TYPE: Cento- acterized by unusual leaf anatomy, including the theca Desv. Figure 5G. presence of palisade mesophyll and laterally ex- Plants annual or perennial (rhizomatous or sto- tended bundle sheath cells. Additional study of this loniferous), herbaceous or reedlike, of warm tem- clade is under way (J. G. Sanchez-Ken, pers. Volume 88, Number 3 Grass Phylogeny Working Group 423 2001 Phylogeny and Classification of Poaceae comm.). The subfamily includes approximately 45 lar or dome-shaped subsidiary cells; bicellular mi- species. crohairs present, panicoid-type, rarely absent; pa- pillae absent or present (mostly in the XI. Panicoideae Link, Hort. Berol. 1: 202. 1827. Andropogoneae). TYPE: Panicum L. Figures 5B, E, 6F, I, K. C3, C4 NAD-ME Syn.: Andropogonoideae Burmeist., Handb. Photosynthetic pathway. (PCK, Naturgesch.: 201. 1837. and NADP-ME), and some C3/C4 intermediates. Rottboellioideae Burmeist., Handb. Naturgesch.: 202. 1837. INCLUDED TRIBES: Saccharoideae (Rchb.) Horan., Char. Ess. Fam.: 34. 1847. AndropogoneaeD umort. [as "Andropogineae"],O b- serv. Gramin. Belg.: 84. 1824. TYPE: Andro- Plants annual or perennial (rhizomatous, stolon- pogon L. iferous, caespitose or decumbent), primarily her- Arundinelleae Stapf, Fl. Cap. 7: 314. 1898. TYPE: baceous, of the tropics and subtropics, but also di- Arundinella Raddi. verse in the temperate zone. Culms solid or less Hubbardieae C. E. Hubb., in Bor, Grasses India commonly hollow. Leaves distichous; abaxial ligule Burma Ceylon Pakistan: 685. 1960. TYPE: usually absent, occasionally present as a line of Hubbardia Bor. hairs; adaxial ligule a fringed or unfringed mem- Isachneae Benth., J. Linn. Soc. Bot. 19: 30. 1881. brane, or a fringe of hairs, or sometimes absent; TYPE: Isachne R. Br. blades relatively broad to narrow, sometimes pseu- Paniceae R. Br., Voy. Terra Austr. 2: 582. 1814. dopetiolate, venation parallel; sheaths usually non- TYPE: Panicum L. auriculate. Inflorescences panicles, racemes, or Steyermarkochloeae Davidse & R. P. Ellis, Ann. spikes, or complex combinations of these, bracts Missouri Bot. Gard. 71: 994. 1984. TYPE: outside of the spikelets present (Andropogoneae) or SteyermarkochloaD avidse & R. P. Ellis. absent (Paniceae). Spikelets bisexual or unisexual Notes. While for the (if the latter plants dioecious or monoecious), fre- support panicoid/centothe- is quently paired in combinations with long and short coid clade high, relationships within the clade remain unclear. No robust for the Pani- pedicels, usually with glumes 2, sterile lemma 1, phylogeny or coideae is and female-fertile floret 1, yet available, work is in dorsally compressed although pro- less not et al., in Duvall et al., in commonly compressed or laterally com- gress (Giussani press; results indicate that the Pani- pressed, disarticulating below the glumes (above press). Preliminary the glumes in Arundinelleae) or the inflorescence ceae as currently circumscribed may not be mono- axes breaking apart; lemma and that the Panicum is lacking uncinate ma- phyletic, large genus poly- et G6mez-Martinez& crohairs, if awned, the awn single; palea well de- phyletic (Zuloaga al., 2000; + veloped (Paniceae) or reduced to absent Culham, Arundinella (Andro- 2000). Andropogoneae ap- pogoneae); lodicules 2 or sometimes pear to be absent, monophyletic (Spangler et al., 1999); stamens other genera of the Arundinelleae are to be cuneate, free, fleshy, usually glabrous; 3; likely distributed the ovary usually glabrous, apical appendage absent, among Andropogoneae, Paniceae, haustorial and even the Centothecoideae synergids absent, styles 2, free or fused, perhaps (Kellogg, This close, 2000b). subfamily includes stigmas 2 (rarely 1 or 3). Caryopsis with the approximately hilum 3270 usually short; endosperm hard, without lipid, species. containing simple or less commonly compound XII. Chloridoideae Kunth ex Flora starch Beilschm., grains; embryo usually large, epiblast absent or 16(2): 52, 105. 1833. TYPE: Chloris Sw. Fig- rarely present, scutellar cleft present, mesocotyl ures 7C. internode elongated, embryonic leaf margins over- 5D, H, 6E, M, N, Syn.: PappophoroideaBe urmeist.,H andb. Naturgesch. lapping or rarely meeting. Basic chromosome num- 205. 1837. bers: x = 5, (7), 9, 10, (12), (14). EragrostoideaPei lger,N at. Pfl.-Fame. d. 2, 14d: 167. 1956. Foliar anatomy. Mesophyll radiate or nonra- an adaxial fusoid cells Plants annual or diate, palisade layer absent, perennial (rhizomatous, s tolon- absent in and arm iferous, caespitose or decumbent), herbaceous except Homolepis Streptostachys, cells usually absent; Kranz anatomy present or ab- (rarely woody), of dry climates, especially in the sent; midrib simple or and rarely complex; adaxial bul- tropics subtropics, also found in the temperate liform cells zone. Culms solid or hollow. Leaves distichous; ab- present. axial ligule usually absent, rarely present as a line Foliar micromorphology. Stomata with triangu- of hairs; adaxial ligule a fringed or less commonly 424 Annals of the Missouri Botanical Garden unfringed membrane; blades relatively narrow, INCERTAE SEDIS: without pseudopetioles, venation parallel; sheaths non-auriculate. Inflorescences CentropodiaR eichenb., Merxmuellerar angei usually paniculate, (Pilg.) Conert paniculate with spicate branches, racemose, or spi- cate, bracts outside of the spikelets absent. Spike- Notes. Reduction in the number of veins in the lets bisexual or sometimes unisexual (if so the lemma is a general trend within the subfamily but plants dioecious or monoecious), with glumes 2, is clearly not a synapomorphy.E xcept for the C3 rarely a sterile lemma, and female-fertile florets 1 Eragrostis walteri and Merxmuellera rangei, the to many, apical reduction usually present, usually Chloridoideae are uniformlyC 4w ith both the NAD- laterally compressed, sometimes dorsally com- ME and PCK subtypes. The current tribal classifi- pressed, usually disarticulating above the glumes cation for this subfamily conflicts with molecular (below in a few Eragrostis species); lemma lacking data and is likely to be modified (Hilu et al., 1999). uncinate macrohairs, if awned, the awns single or This subfamily includes approximately 1400 spe- if multiple, lacking a basal column; palea well de- cies. veloped; lodicules 2 or absent, fleshy, glabrous;s ta- mens 1 to 3; ovary glabrous, apical appendage ab- XIII. Incertae Sedis sent, haustorial synergids absent, styles 2, free, Eriachneae (Ohwi) Eck-Borsb., Blumea 26: 128. close, stigmas 2. Caryopsis with the pericarp often 1980. free or loose; hilum short; endosperm hard, without Micraireae Pilger, Nat. Pfl.-Fam. Ed. 2, 14d: 167. lipid, containing simple or compound starch grains; 1956. embryo large or rarely small, epiblast present or Streptogyneae C. Calder6n & Soderstr., Smithsoni- rarely absent, scutellar cleft present, mesocotyl in- an Contr. Bot. 44: 18. 1980. ternode elongated, embryonic leaf margins meeting Cyperochloa Lazarides & L. Watson, Brunonia 9: or rarely overlapping. Basic chromosome numbers: 216. 1987. x =(7), (8), 9, 10. GyneriumW illd. ex P. Beauv., Ess. Agrostogr.1 38, 153, t. 24. 1812. Foliar anatomy. Mesophyll usually radiate, Notes. These five taxa are left Incertae Sedis without an adaxial palisade layer, fusoid cells ab- because the data here do not arm cells Kranz presented firmly sup- sent, absent; anatomy present; port their inclusion in any of the 12 subfamilies. midrib simple; adaxial bulliform cells present. This approach has also been taken by the APG Foliar Stomata with dome- (1998) for taxa of uncertain micromorphology. placement. Some pos- bicellular mi- sible placements of the five taxa above will or cells; require shaped triangulars ubsidiary publication of new names, and we feel stronglythat crohairs present, usually chloridoid-type; papillae nomenclatural changes should not be made until absent or present. appreciable data support the conclusion. That said, C3 walteri, recent data Photosyntheticp athway. (Eragrostis unpublished (J. G. Sanchez-Ken, pers. Merxmuellera rangei), otherwise C4 (PCK, NAD- comm.) suggest that Gyneriumc an be placed as its but as NADP-ME in own tribe in Panicoideae, and the tribal name ME, may reported Pappophorum, be available the time this is & Watson, 1992; the latter be by paper published by Hattersley may an (Sanchez-Ken & Clark, 2001). It is likely that Cy- error). perochloa will be placed in Centothecoideae, but this is based on its morphological similarities to INCLUDED TRIBES: Spartochloa and not on any data on Cyperochloa itself. Streptogyneae will probably fall within Ehr- Cynodonteae Dumort., Observ. Gramin. Belg.: 83. hartoideae, but limitations of our data and lack of 1824. TYPE: Cynodon Rich. support in our trees make us cautious about placing Eragrostideae Stapf, Fl. Cap. 7: 316. 1898. TYPE: it there unequivocally; there may be an argument Eragrostis Wolf. for recognition of the tribe as its own subfamily. Leptureae Dumort., Observ. Gramin. Belg.: 83. The name Micrairoideae has been published (Pil- 1824. TYPE: LepturusR . Br. ger, 1956). Our data are too limited and the place- Orcuttieae Reeder, Madrofio1 8: 20. 1965. TYPE: ment of the group too uncertain to add it as a thir- Orcuttia Vasey. teenth subfamily, although flora writers may choose Pappophoreae Kunth, Rev. Gramin. 1: 82. 1829. to do so. Our data on Eriachne are weak, and show TYPE: PappophorumS chreb. only that the genus does not fall within the Pani- Volume 88, Number 3 Grass Phylogeny Working Group 425 2001 Phylogeny and Classification of Poaceae coideae, where it has been placed traditionally. Its a one-way ticket to genomic obesity? P1. Cell 9: 1509- placement near the base of the PACCAD Clade is 1514. Bentham, G. 1878. Flora Australiensis 7: 449-670. based on a single-stranded rbcL sequence from one & J. D. Hooker. 1883. Gramineae. Pp. 1074- species, and an ITS sequence from a second. The 1215 in Genera Plantarum,v ol. 3, pt. 2. L. Reeve, Lon- two species represent two sections of the genus, one don. of which has actually been recognized as its own Borre, A. Van den. 1994. Taxonomyo f the Chloridoideae genus. We therefore feel that Incertae Sedis best (Poaceae), with Special Reference to the Genus Era- grostis. Unpublished Ph.D. Dissertation, Australian Na- reflects what we know of the position of the tribe- tional University, Canberra. its position is uncertain. & L. Watson. 1994. The infragenericc lassification of Eragrostis (Poaceae). Taxon 43: 383-422. LiteratureC ited Bossinger, G. 1990. Klassifizierung von Entwicklungs- mutanten der Gerste anhand einer Interpretation des Ambrose, B. A., D. R. Lerner,P . Ciceri, C. M. Padilla, M. Pflanzenaufbausd er Poaceae aus Phytomeren. Inaugu- F. Yanofsky & R. J. Schmidt. 2000. Molecular and ge- ral-Dissertation, Rheinischen Friedrich-Wilhelms- netic analyses of the silky] gene reveal conservation in Universitat zu Bonn. floral organ specification between eudicots and mono- Bremer, K. 1988. The limits of amino acid sequence data cots. Molec. Cell 5: 569-579. in angiosperm phylogenetic reconstruction. Evolution APG (Angiosperm Phylogeny Group). 1998. An ordinal 42: 795-803. classification for the families of flowering plants. Ann. 1990. Combinable component consensus. Cladis- Missouri Bot. Gard. 85: 531-553. tics 6: 369-372. Avdulov, N. P. 1931. Kario-sistematicheskoyei ssledova- Briggs, B. G., A. D. Marchant,S . Gilmore & C. L. Porter. niye semeystva zlakov. Trudy Prikl. Bot. Prilozheniye 2000. A molecular phylogeny of Restionaceae and al- 44: 1-352. [Karyosystematics tudies in the grass family. lies. Pp. 661-671 in K. L. Wilson & D. A. Morrison Supplement 44 to The Bull. Appl. Bot. Genet. P1.- (editors), Monocots: Systematics and Evolution. CSIRO Breed., Leningrad. Russian text, pp. 1-352; German Press, Sydney. summary, pp. 353-425; index pp. 426-428. English Brown, R. 1810. Prodromus florae Novae Hollandiae et translation( mislabeled as supplement 43) published by insulae Van-Diemen, vol. 1. J. Johnson, London. the Smithsonian Institution and the National Scientific 1814. General remarks, geographical and syste- Documentation Centre, New Delhi. 1975. TT 70- matical, on the botany of Terra Australis, Appendix 3. 53085.] Pp. 580-583 in M. Flinders (editor), A Voyage to Terra Barker, N. P. 1995. A Molecular Phylogeny of the Suh- Australis; Undertaken for the Purpose of Ph.D. of Completing family Arundinoideae. Thesis, University Cape the Discovery of that Vast Country, and Prosecuted in Town. the Years 1801, 1802, and 1802. W. Bulmer, l,ondon. 1997. The relationships of Amphipogon,E lytro- Brown, W. V. 1958. Leaf and as rbc, anatomy in grass systematics. phorus, Cyperochloa( lPoaceae) suggested by Bot. Gaz. data. 7: 205-213. (Crawfordsville) 1 19: 170-178. sequence Telopea 1977. The Kranz , H. P. Linder & E. H. 1995. of syndrome and its subtypes in Harley. Polyphyly Mem. Arundinoideae Evidence fromr bcl, sequence grass systematics. 'rorrey B ot. Club 23: 1-91. (Poaceae): Calderon, C. E. & T. R. Soderstrom. 1980. The of data. Syst. Bot. 20: 423-435. genera Bambusoideae , & . 1999. of the (Poaceae) of the American continent: Sequences grass- in Key and comments. Smithsonian Contr. Bot. 44: 1-27. specific insert the chloroplast rpoC2 gene elucidate generic relationships of the Arundinoideae (Poaceae). Campbell, C. S. 1985. The subfamilies and tribes of Gra- Bot. 23: 327-350. mineae (Poaceae) in the southeastern United States. J. Syst. C. M. Morton & H. P. Linder. 2000. The Dan- Arnold Arbor. 66: 123-199. , thonieae: Generic and & E. A. Kellogg. 1987. Sister composition relationships. Pp. group relationships 221-229 in S. W. L. Jacobs & J. E. Everett of the Poaceae. Pp. 217-224 in T. R. K. (editors), Soderstrom, Grasses: and Evolution. CSIRO W. Hilu, C. S. Systematics Publishing, Campbell & M. E. Barkworth( editors), Victoria. Grass Systematics and Evolution. Smithsonian Institu- Collingwood, Barkworth,M . E. 1993. North American tion Stipeae: Taxo- Press, Washington, D .C. nomic notes and other comments. 74: 1-25. , P. E. Garwood& L. P. 1986. Bambusoid Phytologia Specht. & J. Everett. 1987. Evolution in the Stipeae: affinities of the North American temperate genus Bra- Identificationa nd relationships of its monophyletict axa. chyelytrum (Gramineae). Bull. Torrey Bot. Club 113: Pp. 251-264 in T. R. Soderstrom, K. W. Hilu, C. S. 135-141. Campbell & M. E. Barkworth( editors), Grass System- Caro, J. A. 1982. Sinopsis taxonomica de las gramineas atics and Evolution. Smithsonian Institution Press, argentinas. Dominguezia 4: 1-51. Washington,D .C. Carolin, R. C., S. W. L. Jacobs & M. Vesk. 1973. The Baum, D. A., K. J. Sytsma & P. C. Hoch. 1994. A phy- structure of the cells of the mesophyll and parenchy- logenetic analysis of Epilobium (Onagraceae)b ased on matous bundle sheath of the Gramineae. Bot. J. Linn. nuclear ribosomal DNA sequences. Syst. Bot. 19: 363- Soc, London 66: 269-273. 388. Celakovsky, L. 1889. Uber den Archenbau der Brasilian- ,R. L. Small & J. F. Wendel. 1998. Biogeography ischen GrasgattungS treptochaetaS chrader.S itzungsber. and floral evolution of baobabs (Adansonia, Bombaca- Konigl. Bohm. Ges. Wiss. Prag, Math.-Naturwiss.C 1.3 : ceae) as inferred from multiple data sets. Syst. Biol. 47: 14-42, table 2. [Unpublished English translation by 181-207. Mrs. G. Saad, 1974, for the Smithsonian Institution and Bennetzen, J. L. & E. A. Kellogg. 1997. Do plants have the National Science Foundation, Washington, D.C.; in 426 Annals of the Missouri Botanical Garden the files of the McClure Bamboo Library, Department Slipped-strand mispairing in a plastid gene: rpoC2 in of Botany, Smithsonian Institution, Washington,D .C.] grasses (Poaceae). Molec. Biol. Evol. 11: 1-8. Chase, M. W., M. F. Fay & V. Savolainen. 2000a. Higher- Dahlgren, R. M. T., H. T. Clifford & P. F. Yeo. 1985. The level classification in the angiosperms: New insights Families of the Monocotyledons. Springer-Verlag,B er- from the perspective of DNA sequence data. Taxon4 9: lin, Heidelberg, New York. 685-704. Davis, J. I. & R. J. Soreng. 1993. Phylogenetic structure , D. E. Soltis, P. S. Soltis, P. J. Rudall, M. F. Fay, in the grass family (Poaceae) as inferred from chloro- W. J. Hahn, S. Sullivan, J. Joseph, M. Molvray, P. J. plast DNA restriction site variation. Amer. J. Bot. 80: Kores, T. J. Givnish, K. J. Sytsma & J. C. Pires. 2000b. 1444-1454. Higher-level systematics of the monocotyledons:A n as- Dehesh, K., J. Tepperman, A. H. Christensen & P. H. sessment of current knowledge and a new classification. Quail. 1991. phyB is evolutionarilyc onserved and con- Pp. 3-16 in K. L. Wilson & D. A. Morrison (editors), stitutively expressed in rice seedling shoots. Molec. Monocots: Systematics and Evolution. CSIRO Press, Gen. Genet. 225: 305-313. Sydney. Devos, K. M., M. D. Atkinson, C. N. Chinoy, H. A. Fran- , R. G. Olmstead, D. Morgan, D. H. Les, cis, R. L. Harcourt, R. M. D. Koebner, C. J. Liu, P. B. D. Mishler, M. R. Duvall, R. A. Price, H. G. Hills, Masoj'c, D. X. Xie & M. D. Gale. 1993. Chromosomal Y.-L. Qiu, K. A. Kron, J. H. Rettig, E. Conti, J. D. rearrangements in the rye genome relative to that of Palmer, J. R. Manhart,K . J. Sytsma, H. J. Michaels, W. wheat. Theor. Appl. Genet. 85: 673-680. J. Kress, K. G. Karol, W. D. Clark, M. Hedren, B. S. Doebley, J., M. Durbin, E. M. Golenberg, M. T. Clegg & Gaut, R. K. Jansen, K.-J. Kim, C. F. Wimpee, J. F. D. P. Ma. 1990. Evolutionarya nalysis of the large sub- Smith, G. R. Furnier, S. H. Strauss, Q.-Y. Xiang, G. M. unit of carboxylase (rbcL) nucleotide sequence data Plunkett, P. S. Soltis, S. M. Swensen, S. E. Williams, P. among the grasses (Poaceae). Evolution 44: 1097-1108. A. Gadek, C. J. Quinn, L. E. Eguiarte, E. Golenberg, Donoghue, M. J., R. G. Olmstead, J. F. Smith & J. D. G. H. Learn, Jr., S. W. Graham, S. C. H. Barrett, S. Palmer. 1992. Phylogenetic relationships of Dipsacales Dayanandan & V. A. Albert. 1993. Phylogenetics of based on rbcL sequences. Ann. Missouri Bot. Gard. 79: seed plants: An analysis of nucleotide sequences from 249-265. the plastid gene rbcL. Ann. Missouri Bot. Gard. 80: Doyle, J. J., J. I. Davis, R. J. Soreng, D. Garvin & M. J. 528-580. Anderson. 1992. Chloroplast DNA inversions and the Chatterton, N. J., C. Hsiao, K. H. Asay, K. B. Jensen & origin of the grass family (Poaceae). Proc. Natl. Acad. R. R. Wang. 1992. Nucleotide sequence of the internal Sci. U.S.A. 89: 7722-7726. transcribed spacer region rDNA in barley, Hordeumv ul- Duval-Jouve, M. J. 1875. Histotaxie des feuilles des Gra- gare L. (Gramineae).P 1. Molec. Biol. 20: 165-166. minees. Ann. Sci. Nat. Bot. 8: 227-346. Chen, Z., S. Muthukrishnan,G . H. Liang, K. F. Schertz Duvall, M. R. & B. R. Morton. 1996. Molecular phylo- & G. E. Hart. 1993. A chloroplast DNA deletion locat- genetics of Poaceae: An expanded analysis of rbcL se- ed in RNA polymerase gene rpoC2 in CMS lines of quence data. Molec. Phylogenet. Evol. 5: 353-358. sorghum. Molec. Gen. Genet. 236: 251-259. , M. T. Clegg, M. W. Chase, W. D. Clark, W. J. Clark, L. G. 1991. The function of fusoid cells in bamboo: Kress, H. G. Hills, L. E. Eguiarte, J. F. Smith, B. S. An hypothesis. Amer. J. Bot. 78 (suppl.): 22. [Abstract.] Gaut, E. A. Zimmer & G. H. Learn, Jr. 1993. Phylo- & E. J. Judziewicz. 1996. The grass subfamilies genetic hypotheses for the monocotyledonsc onstructed Anomochlooideae and Pharoideae (Poaceae). Taxon4 5: from rbcL sequence data. Ann. Missouri Bot. Gard. 80: 641-645. 607-619. W. Zhang & J. F. Wendel. 1995. A phylogeny of , J. D. Noll & A. H. Minn. In press. Phylogenetics the grass family (Poaceae) based on ndhF sequence of Paniceae (Poaceae). Amer. J. Bot. data. Syst. Bot. 20: 436-460. Ellis, R. P. 1981. Leaf anatomy of the South African Dan- , M. Kobayashi, S. Mathews, R. E. Spangler & E. thonieae (Poaceae). V. Merxmuellera macowanii, M. A. Kellogg. 2000. The Puelioideae, a new subfamily of davyi and M. aureocephala. Bothalia 13: 493-500. Poaceae. Syst. Bot. 25: 181-187. Engler, A. 1892. Syllabus der Vorlesungeni iber specielle Clayton, W. D. 1965. Introduction to Arber's "The Gra- und Medicinisch-pharmaceutisch Botanik. Gebrtider mineae." Pp. i-xxxii in The Gramineae by A. Arber, Borntrager,B erlin. 1965 reprint. J. Cramer,N ew York. Esen, A. & K. W. Hilu. 1989. Immunological affinities .1990. The spikelet. Pp. 32-51 in G. P. Chapman among subfamilies of the Poaceae. Amer. J. Bot. 76: (editor), Reproductive Versatility in the Grasses. Cam- 196-203. bridge Univ. Press, Cambridge,U .K. Farris, J. S., M. Kallersjo, A. G. Kluge & C. Bult. 1994. & S. A. Renvoize. 1986. Genera Graminum, Testing significance of incongruence. Cladistics 10: Grasses of the World. Her Majesty's Stationery Office, 315-319. London. Felsenstein, J. 1978. Cases in which parsimonya nd com- Clifford, H. T. 1987. Spikelet and floral morphology.P p. patibility methods will be positively misleading. Syst. 21-30 in T. R. Soderstrom,K . W. Hilu, C. S. Campbell Zool. 27: 401-410. & M. E. Barkworth (editors), Grass Systematics and Fisher, J. B. & J. Dransfield. 1977. Comparativem or- Evolution. Smithsonian Institution Press, Washington, phology and development of inflorescence adnation in D.C. rattan palms. Bot. J. Linn. Soc. 75: 119-140. Crepet, W. L. & G. D. Feldman. 1991. The earliest re- Frazier, P. J., P. Richmond & A. M. Donald (editors). mains of grasses in the fossil record. Amer. J. Bot. 78: 1997. Starch: Structure and Functionality. Royal Soci- 1010-1014. ety of Chemistry,I nformationS ervices, Cambridge,U .K. Cronquist,A . 1981. An IntegratedS ystem of Classification Gale, M. D. & K. M. Devos. 1998. Comparativeg enetics of Flowering Plants. Columbia Univ. Press, New York. in the grasses. Proc. Natl. Acad. Sci. U.S.A. 95: 1971- Cummings, M. P., L. M. King & E. A. Kellogg. 1994. 1974. Volume 88, Number 3 Grass Phylogeny Working Group 427 2001 Phylogeny and Classification of Poaceae Gaut, B. S. & J. F. Doebley. 1997. DNA sequence evi- Hilu, K. W. & A. Esen. 1988. Prolamin size diversity in dence for the segmental allotetraploid origin of maize. the Poaceae. Biochem. Syst. Ecol. 16: 457-465. Proc. Natl. Acad. Sci. U.S.A. 94: 6809-6814. & J. L. Johnson. 1991. Chloroplast DNA reasso- ,S. V. Muse, W. Clark & M. T. Clegg. 1992. Rel- ciation and grass phylogeny. P1. Syst. Evol. 176: 21- ative rates of nucleotide substitution at the rbcL locus 33. in monocotyledonous plants. J. Molec. Evol. 35: 292- & K. Wright. 1982. Systematics of Gramineae:A 303. cluster analysis study. Taxon 31: 9-36. , B. R. Morton, B. McCaig & M. T. Clegg. 1996. ,L. A. Alice & H. Liang. 1999. Phylogeny of Po- Substitution rate comparisons between grasses and aceae inferred from matK sequences. Ann. Missouri palms: Synonymousr ate differences at the nuclear gene Bot. Gard. 86: 835-851. Adh parallel rate differences at the plastid gene rbcL. Hiratsuka, J., H. Shimada, R. Whittier, T. Ishibashi, M. Proc. Natl. Acad. Sci. U.S.A. 93: 10274-10279. Sakamoto, M. Mori, C. Kondo, Y. Honji, C. R. Sun, B. , L. G. Clark, J. F. Wendel & S. V. Muse. 1997. Y. Meng, Y. Q. Li, A. Kanno, Y. Nishizawa, A. Hirai, Comparisons of the molecular evolutionary process at K. Shinozaki & M. Sugiura. 1989. The complete se- rbcL and ndhF in the grass family (Poaceae). Molec. quence of the rice (Oryza sativa) chloroplast genome: Biol. Evol. 14: 769-777. Intermolecular recombination between distinct tRNA Ghopal, B. H. & H. Y. Mohan Ram. 1985. Systematic genes accounts for a major plastid DNA inversion dur- significance of maturee mbryoo f bamboo. P1. Syst. Evol. ing the evolution of the cereals. Molec. Gen. Genet. 148: 239-246. 217: 185-194. Giussani, L. M., J. H. Cota-Sanchez, F. 0. Zuloaga & E. Hitchcock, A. 1935. Manual of Grasses of the United A. Kellogg. In press. A molecular phylogeny of the States. U.S.D.A. Misc. Publ. 200. grass subfamily Panicoideae (Poaceae) shows multiple & A. Chase. 1950. Manual of Grasses of the Unit- origins of C4 photosynthesis. Amer. J. Bot. ed States, 2nd ed. U.S.D.A. Misc. Publ. 200. Goebel, K. 1895. Ein Beitrag zur Morphologied er Graser. Hsiao, C., S. W. L. Jacobs, N. P. Barker& N. J. Chatterton. Flora 81: 17-19. 1998. A molecular phylogeny of the subfamily Arun- Goloboff, P. 1993. Nona, version 1.16. Buenos Aires: Dis- dinoideae (Poaceae) based on sequences of rDNA. Aus- tributed by the author. tral. Syst. Bot. 11: 41-52. Gomez-Martinez,R . & A. Culham. 2000. Phylogeny of the , N. J. Chatterton& K. H. Asay. 1999. A subfamily Panicoideae with emphasis on the tribe Pan- molecular phylogeny of the grass family (Poaceae) iceae: Evidence from the chloroplast trnL-F cpDNA re- based on the sequences of nuclear ribosomal DNA gion. Pp. 136-140 in S. W. I. Jacobs & J. E. Everett (ITS). Austral. Syst. Bot. 11: 667-688. (editors), Grasses: Systematics and Evolution. CSIRO Igloi, G. L., A. Meinke, I. Dory & H. Kossel. 1990. Nu- Publishing, Collingwood, Victoria. cleotide sequence of the maize chloroplast rpo B/C1/C2 Gould, F. W. & R. B. Shaw. 1983. Grass Systematics, 2nd operon: Comparison between the (lerived protein pri- ed. Texas A&M Univ. Press, College Station. mary structures fiom various organisms with respect to GPWG (Grass Phylogeny WorkingG roup). 2000. A phy- functional domains. Molec. Gen. Genet. 221: 379-394. logeny of the grass family (Poaceae), as inferred from Irish, E. E. 1998. Grass spikelets: A thorny problem. eight character sets. Pp. 3-7 in S. W . . Jacobs & J. E. BioEssays 20: 789-793. Everett (editors), Grasses: Systematics and Evolution. Jacobs, B. F., J. D. Kingston & L. I. Jacobs. 1999. The CSIRO Publishing, Collingwood, Victoria. origin of grass-dominated ecosystems. Ann. Missouri Grassl, C. 0. 1956. The morphologyo f the grass spikelet Bot. Gard. 86: 590-643. with special reference to Saccharum. Proc. Int. Soc. Jacobs, S. W. L. 1987. Systematics of the chloridoid grass- Sugar Technologists 1: 764-780. es. Pp. 277-286 in T. R. Soderstrom,K . W. Hilu, C. S. Graybeal, A. 1998. Is it better to add taxa or characters Campbell & M. E. Barkworth( editors), Grass System- to a difficult phylogenetic problem? Syst. Biol. 47: 9- atics and Evolution. Smithsonian Institution Press, 17. Washington,D .C. Greuter, W., J. McNeill, F. R. Barrie, H. M. Burdet, V. & J. Everett. 1997. Jarava plumosa (Gramineae), Demoulin, T. S. Filgueiras, D. H. Nicolson, P. C. Silva, a new combination for the species formerly known as J. E. Skog, P. Trehane, N. J. Turland & D. L. Hawk- Stipa papposa. Telopea 7: 301-302. sworth (editors). 2000. InternationalC ode of Botanical , M. E. Barkworth& C. Hsiao. 2000. Re- Nomenclature (Saint Louis Code). Regnum Veg. 138. lationships within the stipoid grasses (Gramineae).P p. Gu6eds, M. & P. Dupuy. 1976. Comparativem orphology 75-82 in S. W. L. Jacobs & J. E. Everett (editors), of lodicules in grasses. Bot. J. Linn. Soc. 73: 317-331. Grasses: Systematics and Evolution. CSIRO, Mel- Hackel, E. 1887. Gramineae. Pp. 1-97 in A. Engler & K. bourne. Prantl (editors), Die natiirlichen Pflanzenfamilien. En- Jacques-F6lix, H. 1962. Les graminees (Poaceae) gelmann, Leipzig. d'Afrique tropicale, 1: G6n6ralit6s, classification, de- Hamby, R. K. & E. A. Zimmer. 1988. Ribosomal RNA scription de genres. Bull. Agron. 246: 1-4. sequences for inferring phylogeny within the grass fam- Jirasek, V. 1968. Morphologic der Schuppchen (Lodicu- ily (Poaceae). P1. Syst. Evol. 160: 29-37. lae) von Grasern und ihre Terminologie, ein weiterer Hattersley, P. W. & L. Watson. 1992. Diversification of Beitrag zur Kenntnis des Baues der Lodiculae. Acta photosynthesis. Pp. 38-116 in G. P. Chapman (editor), Univ. Carol., Biol. 1968: 321-344. Grass Domestication and Evolution. Cambridge Univ. & M. Jozifova. 1968. Morphology of lodicules, Press, Cambridge, U.K. their variability and importance in the taxonomy of the Hillis, D. M. 1996. Inferringc omplex phylogenies. Nature Poaceae family. Bol. Soc. Argent. Bot. 12: 324-349. 383: 130. Johnston, C. R. & L. Watson. 1976. Microhairs: A uni- . 1998. Taxonomic sampling, phylogenetic accu- versal characteristic of non-festucoid grass genera? racy, and investigator bias. Syst. Biol. 47: 3-8. Phytomorphology2 6: 297-301. 428 Annals of the Missouri Botanical Garden Judziewicz, E. J. 1987. Taxonomya nd Morphologyo f the 1992b. The structure and evolution of the female Tribe Phareae (Poaceae: Bambusoideae). Unpublished flower of the African Restionaceae. Bot. J. Linn. Soc. Ph.D. Dissertation, University of Wisconsin, Madison. 109: 401-425. & T. R. Soderstrom. 1989. Morphological, ana- & G. Davidse. 1997. The systematics of Tribolium tomical, and taxonomic studies in Anomochloa and Desv. (Danthonieae, Poaceae). Bot. Jarhb. Syst. 119: Streptochaeta (Poaceae: Bambusoideae). Smithsonian 445-507. Contr. Bot. 68. & I. K. Ferguson. 1985. On the pollen morphol- , L. G. Clark, X. Londono & M. Stern. 1999. ogy and phylogeny of the Restionales and Poales. Grana American Bamboos. Smithsonian Institution Press, 24: 65-76. Washington,D .C. & G. A. Verboom. 1996. Generic limits in the Kallersjo, M., S. J. Farris, A. G. Kluge & C. Bult. 1992. Rytidosperma( Danthonieae, Poaceae) complex. Telopea Skewness and permutation.C ladistics 8: 275-287. 6: 597-627. Kanai, R. & G. E. Edwards. 1999. The biochemistry of & N. P. Barker. 1997. Phylogeny and C4 photosynthesis. Pp. 49-87 in R. F. Sage & R. K. evolution in the Crinipes group of grasses (Arundino- Monson (editors), C4 Plant Biology. Academic Press, ideae: Poaceae). Kew Bull. 52: 91-110. San Diego. Lou, H., B. Zhang & D.-F. Qi. 1989. M13 cloning and Katayama, H. & Y. Ogihara. 1996. Phylogenetic affinities nucleotide sequence of the sorghum chloroplast gene of the grasses to other monocots as revealed by molec- for the large subunit of ribulose-l,5-bisphosphate car- ular analysis of chloroplast DNA. Curr.G enet. 29: 572- boxylase. Acta Biochim. Biophys. Sin. 21: 179-188. 581. Maddison, W. P. & D. R. Maddison. 1993. MacClade,v er- Kellogg, E. A. 1998. Relationships of cereal crops and sion 3. Analysis of Phylogeny and CharacterE volution. other grasses. Proc. Natl. Acad. Sci. U.S.A. 95: 2005- Sinauer, Sunderland, Massachusetts. 2010. Mason-Gamer,R . J. & E. A. Kellogg. 1996. Testing for 2000a. The grasses: A case study in macroevo- phylogenetic conflict among molecular data sets in the lution. Ann. Rev. Ecol. Syst. 31: 217-238. Triticeae. Syst. Biol. 45: 524-545. 2000b. Molecular and morphologicale volution in , C. F. Weil & E. A. Kellogg. 1998. Granule-bound Andropogoneae. Pp. 149-158 in S. W. L. Jacobs & J. starch synthase: Structure, function, andtp hylogenetic E. Everett (editors), Grasses: Systematics and Evolution. utility. Molec. Biol. Evol. 15: 1658-1673. CSIRO, Melbourne. Mathews, S. & R. A. Sharrock. 1996. The phytochrome & C. S. Campbell. 1987. Phylogenetic analyses of gene family in grasses (Poaceae): A phylogeny and ev- the Gramineae. Pp. 310-322 in T. R. Soderstrom, K. idence that grasses have a subset of the loci found in W. Hilu, C. S. Campbell & M. E. Barkworth( editors), dicot angiosperms. Molec. Biol. Evol. 13: 1141-1150. Grass Systematics and Evolution. Smithsonian Institu- , M. Lavin & R. A. Sharrock. 1995. Evolution of tion Press, Washington,D .C. the phytochrome gene family and its utility for phylo- & H. P. Linder. 1995. Phylogeny of Poales. Pp. genetic analyses of angiosperms. Ann. Missouri Bot. 511-542 in P. J. Rudall, P. J. Cribb, D. F. Cutler & C. Gard. 82: 296-321. J. Humphries (editors), Monocotyledons: Systematics ,R. C. Tsai & E. A. Kellogg. 2000. Phylogenetic and Evolution. Royal Botanic Gardens, Kew, England. structure in the grass family (Poaceae): Evidence from Kennard, W, R. Phillips, R. Porter & A. Grombacher. the nuclear gene phytochromeB . Amer. J. Bot. 87: 96- 1999. A comparativem ap of wild rice (Zizania palustris 107. L. 2n = 2x = 30). Theor. Appl. Genet. 99: 793-799. McClure, F. A. 1966. The Bamboos:A Fresh Perspective. Kingston, J. D., B. D. Marino & A. Hill. 1994. Isotopic HarvardU niv. Press, Cambridge,M assachusetts. evidence for Neogene hominid paleoenvironments in . 1973. Genera of bamboos native to the New the Kenya Rift Valley. Science 264: 955-959. World (Gramineae:B ambusoideae). Smithsonian Contr. Klijsgen, R. B., A. Gierl, Z. S. Schwarz-Sommer& H. Bot. 9: 1-148. Saedler. 1986. Molecular analysis of the waxy locus of Metcalfe, C. R. 1960. Anatomy of the Monocotyledons.I . Zea mays. Molec. Gen. Genet. 203: 237-244. Gramineae. Clarendon Press, Oxford. Latorre, C., J. Quade & W C. McIntosh. 1997. The ex- Moore, G., K. M. Devos, Z. Wang & M. D. Gale. 1995. pansion of C4 grasses and global change in the late Grasses, line up and form a circle. Curr.B iol. 5: 737- Miocene: Stable isotope evidence from the Americas. 739. Earth Planet. Sci. Lett. 146: 83-96. Morales, C. 0. 1998. Nomenklatorische Probleme in Be- Lazarides, M. 1995. The genus Eriachne (Eriachneae, Po- zug auf die Namen Phareae und Leptaspideae (Grami- aceae). Austral. Syst. Bot. 8: 355-452. neae). Sendtnera 5: 243. LeRoux, L. G. & E. A. Kellogg. 1999. Floral development Nadot, S., R. Bajon & B. Lejeune. 1994. The chloroplast and the formation of unisexual spikelets in the Andro- gene rps4 as a tool for the study of Poaceae phylogeny. pogoneae (Poaceae). Amer. J. Bot. 86: 354-366. P1. Syst. Evol. 191: 27-38. Liang, H. & K. W. Hilu. 1996. Application of the matK Nakai, T. 1943. Anomochloaceae. In Ordines, Familiae, gene sequences to grass systematics. Canad. J. Bot. 74: Tribi, Genera, a Prof. Nakai, Edita 222. 125-134. Nambudiri, E. M. V., W. D. Tidwell, B. N. Smith & N. P. Liese, W. 1998. The Anatomy of Bamboo Culms. Tech- Hebbert. 1978. A C4 plant from the Pliocene. Nature nical Report 18. InternationalN etworkf or Bamboo and 276: 816-817. Rattan, Beijing. Nishizawa, Y. & A. Hirai. 1987. Nucleotide sequence and Linder, H. P. 1987. The evolutionaryh istory of the Poales/ expression of the gene for the large subunit of rice ri- Restionales-A hypothesis. Kew Bull. 42: 297-318. bulose 1, 5-bisphosphate carboxylase. Jap. J. Genet. 62: . 1992a. The gynoecia of Australian Restionaceae: 389-395. Morphology,a natomya nd systematic implications. Aus- Nixon, K. C. 1993. Clados, version 1.4.98 (computers oft- tral. Syst. Bot. 5: 227-245. ware and manual). Distributed by the author. Volume 88, Number 3 Grass Phylogeny Working Group 429 2001 Phylogeny and Classification of Poaceae . 2000. Winclada, version 0.9.99m6.1 (computer Sage, R. F. & R. K. Monson. 1999. C4 Plant Biology. Ac- software and manual). Distributed by the author. ademic Press, San Diego. Nfiez, 0. 1968. El problemad e la palea de Oryza L. Bol. Sanchez-Ken, J. G. & L. G. Clark. 2001. Gynerieae, a Soc. Argent. Bot. 12: 57-97. new Neotropical tribe of grasses (Poaceae). Novon 11: Pilger, R. 1956. Gramineae II, Unterfam. Micrairoideae. 350-352. P. 167 in A. Engler & K. Prantl, Die naturlichen Pflan- Schmidt, R. J. & B. A. Ambrose. 1998. The blooming of zenfamilien (ed. 2) Band 14d. Duncker & Humblot, grass flower development. Curr.O pinion P1. Biol. 1: 60- Berlin. 67. Poe, S. & D. L. Swofford.1 999. Taxon sampling revisited. Schuster, J. 1910. tber die Morphologie der Grasbliite. Nature 398: 299-300. Flora 100: 213-266, plates 2-5. Pohl, R. W. 1987. Man and the grasses: A history. Pp. Seberg, O. & I. Linde-Laursen. 1996. Eremium, a new 355-358 in T. R. Soderstrom,K . W. Hilu, C. S. Camp- genus of the Triticeae (Poaceae) from Argentina. Syst. bell & M. E. Barkworth( editors), Grass Systematics and Bot. 21: 3-15. Evolution. Smithsonian Institution Press, Washington, Sendulsky, T., T. S. Filgueiras & A. G. Burman. 1987. D.C. Fruits, embryos, and seedlings. Pp. 31-36 in T. R. Sod- Pozzi, C., P. Faccioli, V. Terzi, A. M. Stanca, S. Cerioli, erstrom, K. W. Hilu, C. S. Campbell & M. E. Barkworth P. Castiglioni, R. Fink, R. Capone, K. J. Miiller, G. (editors), Grass Systematics and Evolution. Smithsonian Bossinger, W. Rohde & F. Salamini. 2000. Genetics of Institution Press, Washington,D .C. mutations affecting the development of a barley floral Serebryakova,T . I. 1971. Morfogenezp obegov i evoluiut- bract. Genetics 154: 1335-1346. siia zhiznennykh form zlakov [Shoot Morphogenesisa nd Prat, H. 1932. L'epiderme des gramin6es: etude anato- Evolution of Life Forms in Grasses]. Izdatel'stvo" Nau- mique et systematique. Ann. Sci. Nat. Bot. 14: 117- ka," Moskva. 324. Shantz, H. L. 1954. The place of grasslands in the earth's 1936. La syst6matique des Graminees. Ann. Sci. cover of vegetation. Ecology 35: 143-145. Nat. Bot., ser. 10, 18: 165-258. Simon, B. K. & S. W. L. Jacobs. 1990. Gondwanang rasses 1960. Vers une classification naturelle des gra- in the Australian flora. Austrobaileya 3: 239-260. minees. Bull. Soc. Bot. France 107: 32-79. Sinha, N. R. & E. A. Kellogg. 1996. Parallelism and di- Reeder, J. R. 1953. The embryo of Streptochaetaa nd its versity in multiple origins of C4 photosynthesis in the bearing on the homologyo f the coleoptile. Amer. J. Bot. grass family. Amer. J. Bot. 83: 1458-1470. 49: 77-80. Soderstrom,T R. 1967. Taxonomics tudy of subgenus Po- 1957. The embryo in grass systematics. Amer.J . dosemum and section Epicampeso f Muhlenbergia( Gra- Bot. 44: 756-769. mineae). Contr. U.S. Natl. Herb. 34: 75-189 + 14 pl. . 1961. The grass embryo in systematics. Pp. 91- 1981. The grass subfamily Centostecoideae. Tax- 96 in Recent Advances in Botany, Vol. 1. Univ. Toronto on 30: 6144-16. Press, Toronto. & R. . llis. 1987. The position of bamboo gen- 1962. The Bambusoid embryo: A reappraisal. era and allies in a system of grass classification. Pp. Amer. J. Bot. 49: 639-641. 225-238 in rr.R . Soderstrom, K. W. Hilu, C. S. Camp- Renner, S. S., D. B. Foreman & I). Murray.2 ()()().T iming bell & M. E. Barkworth( editors), Grass Systematics and transantarctic disjunctions in the Atherospermataceae Evolution. Smithsonian Institution Press, Washington, (Iaurales): Evidence from coding and noncoding chlo- D.C. roplast sequences. Syst. Biol. 49: 579-591. & E. J. Judziewicz. 1987. The Phareae Renvoize, S. A. & W. D. Clayton. 1992. Classificationa nd and Streptogyneae of Sri I,anka: A morphological-ana- evolution of the grasses. Pp. 3-37 in (. P. Chapman tomical study. Smithsonian Contr. Bot. 65. (editor), Grass Evolution and Domestication.C ambridge Soltis, D. E., P. S. Soltis, M. E. Mort, M. W. Chase, V. Univ. Press, Cambridge,U .K. Savolainen, S. B. Hoot & C. M. Morton. 1998. Inferring Rohde, W., D. Becker & F. Salamini. 1988. Structural complex phylogenies using parsimony:A n empirical ap- analysis of the waxy locus from Hordeumv ulgare. Nucl. proach using three large DNA data sets for angio- Acids Res. 16: 7185-7186. sperms. Syst. Biol. 47: 32-42. Rosengurtt, B., A. Laguardia& B. R. Arrillaga de Maffei. Soreng, R. J. & J. I. Davis. 1998. Phylogenetics and char- 1972. El caracter lipido del endosperma central en es- acter evolution in the grass family (Poaceae): Simulta- pecies de Gramineas. Bol. Fac. Agron. Univ. Montevi- neous analysis of morphological and chloroplast DNA deo 124: 1-43. restriction site character sets. Bot. Rev. 64: 1-85. Roshevits, R. Yu. 1937. Zlaki: Vvedenie v Izuchenie Kor- & . 2000. Phylogenetic structure in Po- movykh Zlakov. [Grasses: An introduction to the study aceae subfamily Pooideae as inferred from molecular of fodder and cereal grasses.] Sel'khozgiz, Gosudar- and morphological characters: Misclassification vs. re- stvennoe Izdatel'stvo Kolkhoznoi I Sovkhoznoi Litera- ticulation. Pp. 61-74 in S. W. L. Jacobs & J. E. Everett tury. Agricultural Publishing House, Moscow, Lenin- (editors), Grasses: Systematics and Evolution. CSIRO grad. [English translationp ublished for the Smithsonian Publishing, Collingwood, Victoria. Institution and the National Science Foundation,W ash- & J. J. Doyle. 1990. A phylogenetic ington, D.C., by the Indian National Scientific Docu- analysis of chloroplast DNA restriction site variation in mentation Centre, New Delhi. 1980. TT 72-51033.] Poaceae subfam. Pooideae. P1. Syst. Evol. 172: 83-97. . 1946. Sistema zlakov v svyazi s ikh evolyutsiei. Spangler, R., B. Zaitchik, E. Russo & E. A. Kellogg. KamarovB otanical Institute, U.S.S.R. Academy of Sci- 1999. Andropogoneae evolution and generic limits in ences, Leningrad. [Portugueset ranslationb y T. Sendul- Sorghum (Poaceae) using ndhF sequences. Syst. Bot. sky. 1969. Evolucao e sistematica das gramineas. Bol. 24: 267-281. Inst. Bot. (Sao Paulo) 5: 1-20.] Stapleton, C. M. A. 1997. The morphologyo f woody bam- 430 Annals of the Missouri Botanical Garden boos. Pp. 251-267 in G. P. Chapman( editor), The Bam- Vickery, J. W., S. W. L. Jacobs & J. Everett. 1986. Taxo- boos, Academic Press, London. nomic studies in Stipa (Poaceae) in Australia. Telopea Stebbins, G. L. 1974. Flowering Plants: Evolution above 3: 1-132. the Species Level. Belknap Press of Harvard Univ. Wang, Z. Y., F. Q. Zheng, J. P. Gao, X. Q. Wang, M. Wu, Press, Cambridge, Massachusetts. J. L. Zhang & M. M. Hong. 1994. Identificationo f two & B. Crampton.1 961. A suggested revision of the transposon-like elements in rice Wx gene. Sci. in China grass genera of temperate North America. Pp. 133-145 Ser. B-Chem. Life. Sci. & Earth Sci. 37: 437-447. in Recent Advances in Botany, Vol. 1. Univ. Toronto Watson, L. & M. J. Dallwitz. 1992. The Grass Genera of Press, Toronto. the World. CAB International,W allingford,U .K. Sugiura, M. 1989. Oryzas ativa chloroplast DNA 134,525 & . 1999. Grass Genera of the World:D e- bp. Nagoya University, Center for Gene Research, Na- scriptions, Illustrations, Identification, and Information goya, Japan. Retrieval; Including Synonyms, Morphology,A natomy, Swofford, D. 1998. PAUP*. Phylogenetic Analysis Using Physiology, Phytochemistry, Cytology, Classification, Parsimony (*and other methods). Version 4. Sinauer, Pathogens, World and Local Distribution, and Refer- Sunderland, Massachusetts. ences. http://biodiversityunoedu/delta/ Version: 18th Takaiwa, F., K. Oono, Y. Iida & M. Sugiura. 1985. The August 1999. complete nucleotide sequence of a rice 25S rRNA gene. ,H. T. Clifford & M. J. Dallwitz. 1985. The clas- Gene 37: 255-259. sification of the Poaceae: Subfamilies and supertribes. Tateoka, T 1957. Miscellaneous papers on the phylogeny Austral. J. Bot. 33: 433-484. of the Poaceae (10). Proposition of a new phylogenetic Whistler, R. L., J. N. BeMiller & E. F. Paschall (editors). system of Poaceae. J. Jap. Bot. 32: 275-287. 1984. Starch: Chemistry and Technology. Academic 1962. Starch grains of endosperm in grass sys- Press, Orlando. tematics. Bot. Mag. (Tokyo)7 5: 336-343. Wilson, W. A., S. E. Harrington,W . L. Woodman,M . Lee, , S. Inoue & S. Kawano. 1959. Notes on some M. E. Sorrells & S. R. McCouch. 1999. Inferences on grasses. IX. Systematic significance of bicellular micro- the genome structure of progenitor maize through com- hairs of leaf epidermis. Bot. Gaz. 121: 80-91. parative analysis of rice, maize and the domesticated Templeton, A. R. 1983. Phylogenetic inference from re- panicoids. Genetics 153: 453-473. striction endonuclease cleavage site maps with partic- Zhang, W. 1996. Phylogeny and Classificationo f the Bam- ular reference to the evolution of humans and the apes. boos (Poaceae: Bambusoideae). Ph.D. Thesis, Iowa Evolution 37: 221-244. State University, Ames, Iowa. Terrell, E. E. 1971. Survey of occurrences of liquid or soft . 2000. Phylogeny of the grass family (Poaceae) endosperm in grass genera. Bull. TorreyB ot. Club 98: from rp/16 intron sequence data. Molec. Phylogenet. 264-268. Evol. 15: 135-146. Thompson, J. D., D. G. Higgins & T. J. Gibson. 1994. & L. G. Clark. 2000. Phylogeny and classification CLUSTALW : Improving the sensitivity of progressive of the Bambusoideae (Poaceae). Pp. 35-42 in S. W. L. multiple sequence alignment through sequence weight- Jacobs & J. E. Everett (editors), Grasses: Systematics ing, position-specific gap penalties and weight matrix and Evolution. CSIRO Publishing, Collingwood,V icto- choice. Nucl. Acids Res. 22: 4673-4680. ria. Tieghem, Ph. van. 1897. Morphologied e l'embryon et de Zuloaga, F. O., O. Morrone& L. Giussani. 2000. A cla- la plantule chez les Gramin6es et les Cyperacees. Ann. distic analysis of the tribe Paniceae: A preliminarya p- Sci. Nat. Bot. 3: 259-309. proach. Pp. 123-135 in S. W. L. Jacobs & J. E. Everett Tzvelev, N. N. 1989. The system of grasses (Poaceae) and (editors), Grasses: Systematics and Evolution. CSIRO their evolution. Bot. Rev. 55: 141-203. Publishing, Collingwood,V ictoria. Verboom,G . A., H. P. Linder & N. P. Barker. 1994. Haus- Zurawski, G., M. T. Clegg & A. H. D. Brown. 1984. The torial synergids: An importantc haracter in the system- nature of nucleotide sequence divergence between bar- atics of danthonioid grasses (Arundinoideae:P oaceae). ley and maize chloroplast DNA. Genetics 106: 735- Amer. J. Bot. 81: 1601-1610. 749. Volume 88, Number3 Grass Phylogeny WorkingG roup 431 2001 Phylogeny and Classificationo f Poaceae Appendix I. Taxa included. For each data set, species name, voucher, and reference are listed, as well as GenBank accession numbers for gene sequences. EAK = Elizabeth Kellogg; HPL = Peter Linder; JID = Jerrold Davis; LGC = Lynn Clark; NPB = Nigel Barker;P MP = Paul Peterson; RJS = Robert Soreng; SJ = Surrey Jacobs; WZ = Weiping Zhang; XL = Ximena Londono; BBG = Berlin Botanic Garden; BHC = L. H. Bailey HortoriumC onservatory;F TG = Fairchild Tropical Garden; NTBG = National Tropical Botanical Garden (Hawaii); PI = USDA Plant Introduction Station (Pullman, Washington)a s source of seed. Genus Species Voucher Reference GenBank # ndhF Flagellaria indica L. LGC & WZ 1305 (ISC) Clark et al. (1995) U22007 Elegia stipularis Mast. Eldenas 2 (BOL) This paper AF251443 Baloskion tetraphyllum( Labill.) B. Kew-6565-1977 (BH) This paper AF251444 G. Briggs & L. A. S. Johnson Joinvillea ascendens Gaudich. ex NTBG-800379 (living) Clark et al. (1995) U21973 Brongn. & Gris. Anomochloa marantoidea Brongn. LGC 1299 (ISC) Clark et al. (1995) U21991 Streptochaeta angustifolia Soderstr. LGC 1304 (ISC) Clark et al. (1995) U21982 Pharus latifolius L. LGC 1302 (ISC) Clark et al. (1995) U21992 Guaduella marantifolia Franch. Kobayashi et al. 1539 Clark et al. (2000) AF164777 (ISC) Puelia ciliata Franch. Kobayashi et al. 1541 Clark et al. (2000) AF164779 (ISC) Eremitis sp. nov. LGC & WZ 1343 (ISC) Zhang & Clark (2000) AF182353 Pariana radicifora Sagot ex Doll LGC & WZ 1344 (ISC) Zhang & Clark (2000) AF182354 Lithachne hurmilis Soderstr. LGC 1298 (ISC) Clark et al. (1995) U21977 Olyra latifolia L,. XL & LGC 911 (1SC) Clark et al. 1995 U21971 Buergersiochloa bambusoidesP ilg. Iransfield 1382 (K) Zhang & Clark (2000) AF 182341 Arundinaria gigantea (Walter)M uhl. WZ 8400703 (ISC) Clark et al. (1995) U21846 Chusqlue latifolia 1. (;. Clark LGC & XL 4/7 (ISC) Clark et al. (1995) U21989 Streptogyna americana C. fE. Hubb). lohl & Davidse 12310 Clark el al. (1995) U21965 (ISC) Ehrharta calycina Sin. NPB s.n. (BOL) Clark et al. (1995) U21995 Oryza sativa L. Sugiura (1989) Clark et al. (1995) X159()1 Leersia virginica Willd. LGC 1316 (ISC) Clark et al. (1995) U21974 Phaenosperma globosum Munroe x LGC 1292 (ISC) Clark et al. (1995) U22005 Benth. Brachyelytrum erectum (Schreb.) P. LGC 1330 (ISC) Clark et al. (1995) U22004 Beauv. Lygeum spartum L. RJS 3698 (BH) This paper AF251445 Nardus stricta L. BBG: Royl & Schiers This paper AF251446 s.n. 1988 (B) Anisopogon avenaceus R. Br. HPL 5590 (BOL) This paper AF251447 Ampelodesmos mauritanica (Poir.) T. BBG: Royl & Schiers This paper AF251448 Durand & Schinz s.n. 1988 (B) Stipa barbata Desf. PI-229468 (BH) This paper AF251449 Nassella viridula (Trin.)B arkworth PI-387938 (BH) This paper AF251450 Oryzopsis( =Piptath- racemosa (Sm.) Ricker ex LGC & WZ 1288 (ISC) Clark et al. (1995) U21924 erum) Hitchc. Brachypodium distachyon (L.) P. Beauv. PI-422452 (BH) This paper AF251451 Melica altisssima L. PI-325418 (BH) This paper AF251452 Glyceria striata (Lam.) Hitchc. JID & RJS s.n. (BH) This paper AF251453 Diarrhena obovata (Gleason) Bran- LGC & WZ 1216 (ISC) Clark et al. (1995) U21998 denburg Avena sativa L. material from R. Wise Clark et al. (1995) U22000 (ISU) Bromus inermis Leyss. PI-314071 (BH) This paper AF251454 432 Annals of the MissouriB otanicalG arden Appendix I. Continued. Genus Species Voucher Reference GenBank # Hordeum vulgare L. material from R. Wise Clark et al. (1995) U22003 (ISU) Aristida purpurea Nutt. var. longi- Gabel 2700 (ISC) Clark et al. (1995) U21966 seta (Steud.) Vasey ex Rothr. Stipagrostis zeyheri (Nees) DeWinter NPB 1133 (BOL) This paper AF251455 Amphipogon strictus R. Br. HPL 5634 (BOL) This paper AF251456 Arundo donax L. LGCs .n. (ISC) Clark et al. (1995) U21997 Molinia caerulea (L.) Moench LGC 1165 (ISC) Clark et al. (1995) U21994 Phragmites australis (Cav.) Trin. ex LGC 1294 (ISC) Clark et al. (1995) U21996 Steud. Merxmuellera macowanii (Stapf) Conert NPB 1008 (BOL) This paper AF251457 Karroochloa purpurea (L.f.) Conert & HPL 5360 (BOL) This paper AF251458 Tuirpe Danthonia californica Bolander PI-232247 (BH) This paper AF251459 Austrodanthonia laevis (Vickery) H. P. HPL 5633 (BOL) This paper AF251460 Linder Merxmuellera rangei (Pilg.) Conert NPB 960 (GRA) This paper AF251461 Centropodia glauca (Nees) Copt NPB 967 (BOL) This paper AF251462 Eragrostis curvula (Schrad.) Nees LGC 1303 (ISC) Clark et al. (1995) U21988 Uniola paniculata L. JID s.n. (BH) This paper AF251463 Zoysia matrella (L.) Merr. LGC 1174 (ISC) Clark et al. (1995) U21975 Distichlis spicata (L.) E. Green Allred s.n. (BH) This paper AF251464 subsp. stricta (Torr.)R . F. Thorne Pappophorum bicolor E. Fourn. Pohl 12464 (ISC) This paper AF352581 Spartina pectinata Link LaDuke s.n. (BH) This paper AF251465 Sporobolus indicus (L.) R. Br. LGC 1293 (ISC) Clark et al. (1995) U21983 Micraira lazaridis L. G. Clark, LGC 1157 (ISC) Clark et al. (1995) U21972 Wendel & Craven Thysanolaena maxima (Roxb.) Kuntze FTG (living) Clark et al. (1995) U21984 Gynerium sagittatum (Aubl.) P. LGC & P Asimbaya This paper AF251466 Beauv. 1472 (ISC) Chasmanthium laxum (L.) H. O. Yates D. Lewis s.n. (ISC) Clark et al. (1995) U27296 Zeugites pittieri Hack. LGC 1171 (ISC) Clark et al. (1995) U21987 Danthoniopsis petiolata (J. B. Phipps) LGC 1173 (ISC) Clark et al. (1995) U22008 Clayton Panicum virgatum L. LGC 1164 (ISC) Clark et al. (1995) U21986 Pennisetum alopecuroides( L.) Spreng. RJS s.n. (BH) This paper AF251467 Miscanthus japonicus Andersson Arnold Arboretum3 01 Spangler et al. (1999) AF117417 80c (living) Zea mays L. cv. 'B73' Materialf rom M. Lee Clark et al. (1995) U21985 (ISU) rbcL Flagellaria indica Chase 206 (NCU) Chase et al. (1993) L12678 Elegia capensis (Burm. f.) Schel- Chase 209 (NCU) Duvall & Morton( 1996) L12675 pe Baloskion tetraphyllum No voucher Katiyama & Ogihara D38296 (1996) Joinvillea plicata (Hook. f.) Newell Thren8 4 (NO) Duvall & Morton( 1996) L01471 & B. C. Stone Anomochloa marantoidea LGC 1299 (ISC) Duvall & Morton( 1996) AF021875 Guaduella marantifolia Kobayashi et al. 1539 Clark et al. (2000) AF164778 (ISC) Volume 88, Number3 Grass Phylogeny WorkingG roup 433 2001 Phylogeny and Classificationo f Poaceae Appendix I. Continued. Genus Species Voucher Reference GenBank # Puelia ciliata Kobayashi et al. 1541 Clark et al. (2000) AF164780 (ISC) Lithachne humilis LGCs .n. (ISC) Duvall & Morton( 1996) U13231 Bambusa multiplex (Lour.) Sanders 62-616 (UCR) Duvall & Morton( 1996) M91626 Raeusch. ex Schult. & Schult. f. Chusquea circinata Soderstr.& C. Quail Botanic Garden Duvall & Morton( 1996) U13227 Calderon (living) Oryza sativa No voucher Nishizawa & Hirai D00207 (1987) Leersia oryzoides( L.) Sw. LGCs .n. (ISC) Duvall & Morton( 1996) U13228 Stipa dregeana Steud. var. dre- McDowell s.n. (BOL) Barker et al. (1995) geana Avena sativa No voucher Duvall et al. (1993) L15300 Bromus inermis Leyss. No voucher Seberg & Linde-Laursen Z49836 (1996) Hordeum vulgare No voucher Zurawski et al. (1984) X00630 Aristida congesta Roem. & NPB 1130 (BOL) Barker et al. (1995) U31359 Schult. Stipagrostis zeyheri NPB 1133 (BOL) Barker et al. (1995) U31378 Amphipogon strictus HPL 5634 (BOL) Barker (1997) U88403 Arundo donax NPB 1131 (BOL) Barker et al. (1995) U31360 Moliniopsis japonica (Hack.) Hayata Kobayashi 1253 Barker et al. (1995) U31439 Phragmites australis NPB 1132 (BOL) Barker et al. (1995) U29900 Merxmuellera macowani NPB 1008 (BOL) Barker et al. (1995) U31438 Karroochloa purpurea HPL 5360 (BOL) Barker et al. (1995) U31437 Danthonia spicata (I.) P. Beauv. ex EAK V1 (GH) Barker et al. (1995) U31102 Roem. & Schult. Centropodia glauca HPL 5410 (BOL) Barker et al. (1995) U31100 Eragrostis capensis (Thunl.) Trin. NPB 1 135 (BOL) Barker et al. (1995) U31104 Enneapogon scaber Lehm. NPB 1023 (BOL) Barker et al. (1995) U31103 Eriachne triodioides Domin EAK s.n. (GH) This paper AF352580 Thysanolaena maxima Kew 1979-3225 Warr Barker et al. (1995) U31380 (living) Gynerium sagittatum Kew 1991-1276 Kall Barker et al. (1995) U31105 (living) Chasmanthium latifolium (Michx.) H. O. Snow 5944 Barker et al. (1995) U31101 Yates Pennisetum glaucum (L.) R. Br. No voucher Doebley et al. (1990) L14623 Sorghum bicolor (L.) Moench No voucher Lou et al. (1989) 1515164A Zea mays No voucher Gaut et al. (1992) Z11973 rpoC2 Joinvillea plicata No voucher Barker et al. (1999) AF001864 Olyra latifolia HPL 5742 (BOL) Barker et al. (1999) U90825 Bambusa vulgaris Schrad. ex J. C. Durban Botanic Garden Barker et al. (1999) U90824 Wendl. (living) Ehrharta dura Nees ex Trin. NPB 1118 (BOL) Barker et al. (1999) AF064761 Oryza sativa No voucher Hiratsuka et al. (1989) X15901 Lygeum spartum Kew (living) Cummings et al. (1994) L25381 Nardus stricta Kew (living) Cummings et al. (1994) L25382 Anisopogon avenaceus HPL 5590 (BOL) Barker et al. (1999) U92263 Stipa dregeana McDowell s.n. (BOL) Barker et al. (1999) U90826 Briza maxima L. EAK s.n. (GH) Cummings et al. (1994) L25376 Bromus tectorum L. EAK s.n. (GH) Cummings et al. (1994) L25377 434 Annals of the Missouri Botanical Garden Appendix I. Continued. Genus Species Voucher Reference GenBank # Aristida congesta Roem. & NPB 1130 (BOL) Barker et al. (1999) U90827 Schult. subsp. barbi- collis (Trin. & Rupr.) DeWinter Stipagrostis zeyheri subsp. zeyheri NPB 1133 (BOL) Barker et al. (1999) U90828 Amphipogon strictus HPL 5634 (BOL) Barker et al. (1999) U94335 Arundo donax NPB 1131 (BOL) Barker et al. (1999) U92264 Moliniopsis japonica Kobayashi 1253 Barker et al. (1999) U95081 Phragmites australis NPB 1132 (BOL) Barker et al. (1999) U95130 Merxmuellera macowanii NPB 1008 (BOL) Barker et al. (1999) U95076 Karroochloa purpurea HPL 5360 (BOL) Barker et al. (1999) U94824 Danthonia spicata EAK V1O( GH) Barker et al. (1999) U93362 Austrodanthonia laevis HPL 5633 (BOL) Barker et al. (1999) U96313 Merxmuellera rangei NPB 960 (GRA) Barker et al. (1999) U95077 Centropodia glauca HPL 5410 (BOL) Barker et al. (1999) U92265 Eragrostis capensis NPB 1135 (BOL) Barker et al. (1999) U96317 Enneapogon scaber NPB 1023 (BOL) Barker et al. (1999) U96319 Spartina alterniflora EAK s.n. (GH) Cummings et al. (1994) L25386 Micraira lazaridis LGC 1157 (ISC) Barker et al. (1999) U96318 Thysanolaena maxima Kew, 1979-3225 Warr Barker et al. (1999) U96315 (living) Gynerium sagittatum Kew, 1991-1276 Kall Barker et al. (1999) U94392 (living) Chasmanthiuml latifolium Snow 5944 Barker et al. (1999) U94334 Panicum maximum Jacq. NPB 1125 (BOL) Barker et al. (1999) AF000021 Pennisetum sp. No voucher Cummingse t al. (1994) L25383 Sorghum bicolor No voucher Chen et al. (1993) Z14983 Zea mays No voucher Igloi et al. (1990) X86563 PhytochromeB Flagellaria indica RJS 77 394 (BH) Mathews & Sharrock U61203 (1996) Joinvillea ascendens Moore 10438 (NY) Mathews & Sharrock U61205 (1996) Anomochloa marantoidea LGC 1299 (ISC) Mathews et al. (2000) AF137291 Streptochaeta angustifolia LGC 1304 (ISC) Mathews et al. (2000) AF137328 Pharus lappulaceus Aubl. LGC 1,329 (ISC) Mathews et al. (2000) AF137321 Puelia ciliata Kobayashi et al. 1541 Mathews et al. (2000) AF137324 (ISC) Eremitis sp. nov. LGC & WZ 1343 (ISC) Mathews et al. (2000) AF137304 Pariana radiciflora LGC & WZ 1344 (ISC) Mathews et al. (2000) AF137317 Lithachne pauciflora (Sw.) P. Beauv. LGC 1297 (ISC) Mathews et al. (2000) AF137307 Olyra latifolia XL & LGC 911 (ISC) Mathews et al. (2000) AF137315 Buergersiochlo( bambusoides Dransfield 1382 (K) Mathews et al. (2000) AF137295 Pseudosasa japonica (Sieb. & Zucc. EAK V6 (A) Mathews et al. (2000) AF137323 ex Steud.) Makino ex Nakai Chusquea oxylepis (Hack.) Ekman LGC 1069 (ISC) Mathews et al. (2000) AF137298 Streptogyna americana Johnston 433 Mathews et al. (2000) AF137329 Ehrharta erecta Lam. EAK V44 (GH) Mathews et al. (2000) AF137302 Oryza sativa no voucher Dehesh et al. (1991) X57563 Lygeum spartum RJS 3698 (BH) Mathews et al. (2000) AF137309 Nardus stricta BBG: Royl & Schiers s.n. Mathews et al. (2000) AF137313 Anisopogon avenaceus HPL 5590 (BOL) Mathews et al. (2000) AF137290 Nassella viridula Lavin s.n. (MONT) Mathews et al. (2000) AF137314 Brachypodium pinnatum (L.) P. Beauv. PI-440176 (GH) Mathews et al. (2000) AF137294 Melica cupanii Guss. PI-383702 (BH) Mathews et al. (2000) AF137310 Volume 88, Number3 Grass Phylogeny WorkingG roup 435 2001 Phylogeny and Classificationo f Poaceae Appendix I. Continued. Genus Species Voucher Reference GenBank # Glyceria grandis S. Watson JID & RJS s.n. (BH) Mathews et al. (2000) AF137305 Diarrhena obovata LGC & WZ 1216 (ISC) Mathews et al. (2000) AF137301 Phalaris arundinacea L. RJS 3427 (BH) Mathews et al. (2000) AF137320 Bromus inermis Lavin s.n. (MONT) Mathews et al. (2000) U61193 Triticum aestivum L. Mason-Gamers .n. (GH) Mathews et al. (2000) AF137331 Aristida purpurea subsp. longiseta Lavin s.n. (MONT) Mathews et al. (2000) AF137292 Molinia caerulea RJS 3305 (BH) Mathews et al. (2000) AF137312 Phragmites australis Keller s.n. (GH) Mathews et al. (2000) AF137322 Danthonia spicata EAK V1O( GH) Mathews et al. (2000) AF137299 Eragrostis cilianensis (All.) Vignolo Lavin s.n. (MONT) Mathews et al. (2000) U61200 ex Janch. Sporobolus giganteus Nash PMP 10008 (US) Mathews et al. (2000) AF137327 Thysanolaena maxima Farnsworths .n. (GH) Mathews et al. (2000) AF137330 Chasmanthium latifolium EAK V13 (A) Mathews et al. (2000) AF137297 Danthoniopsis dinteri (Pilg.) C. E. PI-207548 (GH) Mathews et al. (2000) AF137300 Hubb. Panicum capillare L. Lavin s.n. (MONT) Mathews et al. (2000) AF137316 Pennisetum alopecuroides EAK s.n. (A) Mathews et al. (2000) AF137318 Miscanthus japonicus Arnold Arboretum3 01- Mathews et al. (2000) AF137311 80C (living) Zea mays Lavin s.n. (MONT) Mathews et al. (2000) AF137332 Chloroplastr estriction site polymorphisms Flagellaria indica BHC-77394 Soreng & Davis (1998) Baloskion tetraphyllum Kew-6565-1977 (BH) Soreng & Davis (1998) Joinvillea ascendens NTBG-800379 Davis & Soreng (1993) (H. Moore 10438) Anomochloa marantoid(ea LGC 1299 (ISC) Soreng & D)avis( 1998) Streptochaeta sodiroanaL Hack. PMP 9525 (US) Soreng & Davis (1998) Pharus latifolius IHC from USZ I)avis & Soreng (1993) Eremitis sp. USNHG-1 53, Soderstrom Soreng & Davis (1998) 2182 (US) or USNHG- 286 (US) Lithachne humilis BHC from U. S. National Davis & Soreng (1993) Zoological Gardens Olyra latifolia PMP 7311 (US) Soreng & Davis (1998) Pseudosasa japonica BHC-71467 Davis & Soreng (1993) Chusquea aff. subulata L. G. Clark PMP 9499 (US) Soreng & Davis (1998) Ehrharta calycina PI-208983 (BH) Soreng & Davis (1998) Oryza sativa no voucher Hiratsuka et al. (1989) Leersia virginica RJS 3399 (BH) Davis & Soreng (1993) Brachyelytrum erectum RJS 3427 (BH) Davis & Soreng (1993) Lygeum spartum RJS 3698 (BH) Soreng & Davis (1998) Nardus stricta BBG: seed from Royl & Davis & Soreng (1993) Schiers s.n. 1988, Hempel s.n. 1987 (B) Anisopogon avenaceus HPL 5590 (BOL) Soreng & Davis (1998) Ampelodesmos mauritanica BBG: Royl & Schiers s.n. Soreng & Davis (1998) 1988 (B) Stipa barbata PI-229468 (BH) Davis & Soreng (1993) Nassella viridula PI-387938 (BH) Soreng & Davis (1998) Piptatherum miliaceum (L.) Coss. PI-284145 (BH) Davis & Soreng (1993) Brachypodium pinnatum PI-440170 (BH) Davis & Soreng (1993) Melica altissima PI-325418 (BH) Davis & Soreng (1993) Glyceria striata JID & RJS s.n. (BH) Davis & Soreng (1993) Diarrhena obovata Seed from Tiedye 5186 Davis & Soreng (1993) (DAO) 436 Annals of the MissouriB otanicalG arden Appendix I. Continued. Genus Species Voucher Reference GenBank # Avena barbata Pott ex Link No voucher Soreng & Davis (1998) Bromus inermis RJS 3428 (BH), PI- Davis & Soreng (1993) 314071 (BH) Triticum aestivum L. cv. 'Susqu- RJS s.n. (BH) Soreng & Davis (1998) ehanna' Aristida purpurea Allred s.n. (BH) Soreng & Davis (1998) Amphipogon strictus HPL 5634 (BOL) Soreng & Davis (1998) Arundo donax FTG-83-130 (BH) Soreng & Davis (1998) Molinia caerulea RJS 3305 (BH) Soreng & Davis (1998) Phragmites australis RJS 3884 (BH) Davis & Soreng (1993) Danthonia californica PI-232247 (BH) Davis & Soreng (1993) Eragrostis curvula PI-365034 (BH) Davis & Soreng (1993) Uniola paniculata JID s.n. (BH) Soreng & Davis (1998) Zoysia sp. JID s.n. (BH) Soreng & Davis (1998) Distichlis spicata subsp. stricta Allred s.n. (BH) Soreng & Davis (1998) Spartina pectinata LaDuke s.n. (BH) Soreng & Davis (1998) Sporobolus giganteus PMP 10008 (US) Soreng & Davis (1998) Chasmanthium latifolium Cornell University gar- Davis & Soreng (1993) dens (living) Panicum virgatum USDA 421520 (BH) Soreng & Davis (1998) Pennisetum alopecuroides RJS s.n. (BH) Davis & Soreng (1993) Miscanthus sinensis Andersson var. RJS s.n. (BH) Davis & Soreng (1993) gracillimus Hitchc. ITS Joinvillea plicata Wilson 7126 Hsiao et al. (1999) AF019784 Streptochaeta sodiroana PMP & Annable 9525 Hsiao et al. (1999) AF019785 (US) Pharus latifolius PMP & Annable 6944 Hsiao et al. (1999) AF019786 (US) Lithachne humilis Utah State University Hsiao et al. (1999) AF019787 s.n. (living) Chusquea latifolia LGC & XL 417 (ISC) Hsiao et al. (1999) AF019788 Microlaena stipoides (Labill.) R. Br. Kew 1973-15875 (living) Hsiao et al. (1999) AF019791 Oryza sativa No voucher Takaiwae t al. (1985) Leersia hexandra Sw. Jacobs 7782 Hsiao et al. (1999) AF019793 Brachyelytrum erectum IntermountainH erbarium Hsiao et al. (1999) AF019794 1669 Lygeum spartum Catalan 1593 Hsiao et al. (1999) AF019797 Nardus stricta IntermountainH erbarium Hsiao et al. (1999) AF019796 203443 Anisopogon avenaceus Dalby 94/01 Hsiao et al. (1999) AF019800 Ampelodesmos mauritanica Kew 150-90.00982 (liv- Hsiao et al. (1999) AF019799 ing) Stipa ichu (Ruiz & Pavon) Renvoize& Flores 5301 Hsiao et al. (1999) AF019803 Kunth (K) Nassella leucotricha (Trin. & Houck s.n. Hsiao et al. (1999) L36520 Rupr.) R. W. Pohl Piptatherum songaricum (Trin. & Hsiao 199 Hsiao et al. (1999) AF019802 Rupr.) Roshev. ex Ni- kitina Brachypodium mexicanum (Roem. & University of Leicester Hsiao et al. (1999) AF019805 Schult.) Link Botanic Gardens 347 Melica californica Scribn. Curto 719 Hsiao et al. (1999) L36518 Glyceria striata Curto 826 Hsiao et al. (1999) L36516 Diarrhena americana P. Beauv. IntermountainH erbarium Hsiao et al. (1999) AF019798 218465 Volume 88, Number3 Grass Phylogeny WorkingG roup 437 2001 Phylogeny and Classificationo f Poaceae Appendix I. Continued. Genus Species Voucher Reference GenBank # Avena longiglumis Durieu Fritz, CN Hsiao et al. (1999) Z11758 Bromus mnermis Hsiao 103 Hsiao et al., (1994) L11579 Hordeum vulgare Hsiao 200 Chattertone t al. (1992) Z11759 Aristida purpurea IntermountainH erbarium Hsiao et al., 1999 AF019807 209381 Stipagrostis zeyheri subsp. zeyheri NPB 1133 Hsiao et al. (1999) AF019845 Amphipogon caricinus F. Muell. Macfarlane 2155 Hsiao et al. (1998) AF019849 Arundo donax Hsiao 196, Evans s.n. Hsiao et at. (1999) AF019809 Molintia caerulea Kewl973 10386 Hsiao et al. (1999) AF109857 Phragmites australis Chatterton s. a. Hsiao et at. (1999) AF019810 Merxmuellera macowant'i Kew 142-83.01715 Hsiao et al. (1998) AF019863 Karroochloa purpurea HPL 5360 Hsiao et al. (1998) AF019874 Danthonia califoricn Curto 974 Hsiao et at. (1999) AF019813 Rytidosperma pumilum (Kirk) H. P. HPL 5747 Hsiao et at. (1998) AF019878 Linder Merxmuellera rangei NPB 960 (GRA) Hsiao et al. (1998) AF019862 Centropodia glauca NPB 967 Hsiao et al. (1998) AF019861 Eragrostis dielsii Pulg. cx Diels & Jacobs 7195 Hsiao et al. (1999) AF019834 Pritz. Spartina gracilis Trin. IntermountainH erbarium Hsiao et al. (1999) AF019844 194828 Sporobolus airoides (Torr.) Torr. Curto s. n. Hsiao et al. (1999) AF019842 Eriachne triseta Nees cx Steud. Jacobs 7184 Hsiao et al. (1999) AF019818 Micratir(L subulifolia F. Mucll. Clarkson 1 0300 Hsiao et al. (1999) AF019859 Thysanolaena maxima Kew1979-3225 Hsiao et al. (1999) AF019854 Gynerium sagtttaturn Kewl991-1276Kall Hsiao et al. (1999) AF O1 9853 Chasmanthiutm latij6lium ltiterl)iountainH erbarium Hsia(ae t al. (1999) A 0O19 815 216008 PatnicuLm )isalcranttlr rThunth. Hsitao 160, 1)-19486 Hsiao et al. (1999) AF0l 9829 Pennisetum setaceum (Forssk.) Chiov. Cutrto 976 Hsiao et at. (1999) AF O1 9 833 Miscantlh us smiersuis 1va(ns s.n. Hsiao ct al. (1999) AV O1 9822 Zea mnaysL . sub)sp. mexicaLnaL Hsiao 197 Hsiao et al. (1999) AF O1 981 7 (Schrad.) Ittis GBSSI Aniomochloai marantoiden LGC 1299 (ISC) Mason-Gamere t al. AF079290 (1998) laippulaceus LGC 1.329 (ISC) Mason-Gamere t al. AF079298 (1998) Eremitis sp. nov. LGC & WZ 1343 (ISC) Mason-Gamere t al. AF079295 (1998) Pariana radicflora LGC & WZ 1344 (ISC) Mason-Gamere t al. AF079297 (1998) Chusquea exasperata L. G. Clark LGC et al. 1003 (ISC) Mason-Gamere t al. AF079293 (1998) Oryza sativa No voucher Wang et at. (1994) X65183 Lygeum spartum RJS 3698 Mason-Gamere t al. AF079289 (1998) Melica cupanii PI-383702 (A) Mason-Gamere t al. AF079296 (1998) Glycerin grandis JID & RJS s.n. Mason-Gamere t al. AF079291 (1998) Hordeum vulgare No voucher Rohde et al. (1988) X07932 Merxmuellera macowanal NPB 1008 (BOL) This paper AF353520 Karroochloa purpurea HPL 5360 (BOL) This paper AF353519 Austrodanthonia laevis HPL 5633 (BOL) This paper AF353517 438 Annals of the Missouri Botanical Garden Appendix I. Continued. Genus Species Voucher Reference GenBank # Merxmuellera rangei NPB 960 (GRA) This paper AF353521 Centropodia glauca NPB 967 (BOL) This paper AF353518 Danthoniopsis dinteri PI-207548 (A) Mason-Gamere t al. AF079251 (1998) Pennisetum alopecuroides Park Seed 3650 (A) Mason-Gamere t al. AF079288 (1998) Sorghum bicolor PI-156549 (A) Mason-Gamere t al. AF079258 (1998) Zea mays No voucher Klosgen et al. (1986) X03935 Appendix II. Matrix of structural characters, as assembled for analysis in NONA (Goloboff, 1993). Taxa in the matrix appear in groupings according to what was known about the phylogeny at the time the taxon samlplingl ist was prepared. Thus, the four outgroupsc ome first, followed by the early-divergingt axa, then bambusoids, rices, pooids, etc. Abbreviations of taxon names and associated underlines are required by the program. Charactersa nd c haracters tates are described in Table 4, and are optimized on the cladogram in Figure 3. Codes used for polymorphisns (presence of two or more states) and subset ambiguities (when one or more states are not present, but the observed attributec annot be assigned to any of the states not eliminated) are as follows: ? = unobserved; - = inapplicable; \ = intermediate/ uncertain homology/unassignablet o defined states; A = [01]; C = [03]; D = [12]; E = [13]; F = [231; H = [034]; J = [234]; K = [01341; L = [14]; N = [29]; Q = [07]; R = [012]; S = [57]; T = [127]. Charactern umbers 00000000011111111112222222222333333333344444444445555 Taxon 789016789012346789012345678901 234567890123456789023 Flagellaria \01-01AO- _-- _- -0\o \1111003130-0----??00?-0050- -\000 Elegia 000--0-0- ----- o-0\ ,1100103310-0----??1-0---?0- -???1 Baloskion 070--0-0- ,1100102D2?-0----011-0---TO- -?0?0 Joinvillea 01000110- - ----0\ \1111OA3130-0----???110\140- -1100 Anomochloa 0001011\- \0---\0\ \10110\111?011100?101101140- -0110 Streptochaeta 01010110-\\\\---\0\\\\111101311?010101?10?100110- -0110 Pharus 0000011111010--A0130??111101311?01110101110-1120- -0110 Guaduella 0100011101100--1013011111100221?01???????110A1?O- -???1 Puelia 0100111101110--1013011111101F21?01???????10-1120- -0??1 Eremitis 0100011101010--001301100010\111?011101?1011011\0- -01?1 Pariana 010001110A010--0\1301111110A211?011101??011011DO- -0??1 Lithachne 01000111A1010--11130110011012F1?0111010?01101110- -0111 Olyra 0100011101010--11130110011012F1?0111010101101110- -01?1 Buergersiochloa 010001110101110101301100A1012?1?01???????11011?0- -0??1 Pseudosasa 110011110100A1010130111111013210011101?301101120- -01?1 Chusquea 100011110111A--1013011001101221?0111010?01101100- -01?1 Streptogyna 010011A101001111013011000101F11?0 11101? 00120-0?? Ehrharta Oryza 010OOO01A101\lA10\11201111110A221?0110010101101020- -0110 Leersia 0100010101\lA10\0120A1000100231?011001010110A020- -?1?0 Phaenosperma OA00011101010--0013010011002D?01110104 0020- -???1 Brachyelytrum 0000010101011101012011001100221?011A01?4110-0010- -?111 Lygeum 000001010\000--1?0----00110\121?0110000111110000- -0111 Nardus 0100010100011101?0----00110\111?0110000111100030- -01?? Anisopogon 0100010101011311013010001100F21001?????1?10-00-0- -0111 Ampelodesmos 00000101010011A1013010001100221?0110000?110-0020- -?1?1 Stipa 01000101010111A101FO10001100J2100110000Kll0-0ORO- -?1?1 Nassella 01000101010111A1012010001100221?0110000Cl10-00\0- -01?1 Piptatherum 010001010101A101013010000100221?01????01110-0020- -?1?1 Brachypodium_d 0100010101001101?1201?001100221?01AO0000110-0OSO- -01?1 Melica_a 01100101010AO--0012100001100231?011000011AO-0090- -01?1 Volume 88, Number3 Grass Phylogeny WorkingG roup 439 2001 Phylogeny and Classificationo f Poaceae Appendix II. Continued. Charactern umbers 00000000011111111112222222222333333333344444444445555 Taxon 123456789012 345678901 23456789013458901234567890123 Glyceria-s 0110010101000--1012A00000100231?011000011AO-0000-01? Diarrhena 01000101010AO--101201000A100221?011AOA01110-0000-01? Avena 0100010101001E21012010001100221?011AO\1llAO-0070-01?1 Bromus 011001010100A1R1012010001100221?010000001AO-0070-0171 Triticum 010001010000A101012010001100221?01100\001AO-0070-01?1 Aristida 00OA010101011E011120110011002210010110011A100012-11?1 Stipagrostis OA010101010113011A2011001100221?010110?D111000130???1 Amphipogon 01010101010113A1112000001101221?110110?1?11100-0-?1?1 Arundo 110001010100AEA1012001001100221?110110?1?A100020-?111 Molinia 010101010100A101?12A01001100221?0101100111100090-1111 Phragmites \10101010110A1010120010011012210\101100111101020-1111 M_merxmuellera-m_ 0001010101001E11012011001100221101???????11000?0-???1 Karroochloa 0101010101001E110120010011002211110110???1100060-???1 Danthonia 010101010100131101200100110022110101100D11100040-11?1 Austrodanthonia 01010101010013110120010011002211110110???1100020-???1 R-merxmuellera-r_ 0?01010101001E11012001001100221??????????10-0060-???1 Centropodia 0001010101001111012001001100221011?????L?11000630???1 Eragrostis 0\01010101000--1012001001100221?11111001111A000301111 Uniola 0\01010101100--1012001001101221?110110??11110003\?1?1 Zoysia 000A1OOA0001110000 ----0A1OA221?1111100D111100031?1?1 Distichlis 0000010101000--1012001001100221?1111100H?11100030?1?1 Pappophorum 0?010101010A120101200\001100221?\111100??111000E-???l Spartina 0101010100010--000----001101221?1111100111A100Q31?1?1 Sporobolus 00OAO10101010--O1A200100AA002D1?1111100E11110093Al1?1 Eriachne OA01A101010AA1A101200100A100221?010111??111000?5-???? Micraira 00OAO1A101AAO--100----000101221?0101???Hl1100000-???l Thysanolaena 1100111101010--101200000A100221?11?????????11010DO-1111 Gynerium 100AA1\101000--101201\000100221?117??????11010\0-???1 Chasmanthiunml 0100010101100--1012001001001221?1111110311100020-1111 Zeugites 010001110101A1OA01200\001101221?111101?1?11000\0-???1 I)anthoniopsis OAO1A1A1AllllE11112001001100221?01?????4111000NL-1??i Panicum 0100010101110--01120A100110022101101110E111000\3011?1 Pennisetum 010A01010111A1001A200100110A221?1101110311100091-11?1 Miscanthus 0000010111111110012001001100221?110111?211100051-11?1 Zea 010001011A10O--0?12001001101211?\101110311100001-11?1 440 Annals of the MissouriB otanicalG arden Appendix III. Consensus trees for individual data sets and combinations of data sets. Numbers above branches indicate percent of 500 bootstrap replicates, except for K (all molecular data), for which 1000 replicates were done. Tree statistics are listed in Table 3. The GPWG classification is overlain on each tree for comparison with Figure 2. -A. Chloroplast restriction sites; strict consensus of seven trees. -B. ndhF; strict consensus of 16 trees. -C. rbcL; single most parsimonioust ree. -D. rpoC2;s trict consensus of 33 trees. -E. PhytochromeB ; single most parsimonious tree. -F. ITS; strict consensus of 24 trees. -G. GBSSI, single most parsimonious tree. -H. Structural data; strict consensus of 38,000 trees. -I. Chloroplast data; strict consensus of two trees. -J. Nuclear data; strict consensus of eight trees. -K. All molecular data; strict consensus of six trees. Flagellaria Baloskion Joinvillea Anomochloa A m h 91 CI Streptochaet7 Anomochlo Pharus Phar. Eremitis 100 67 Lithachne Bambus. Olyra 98 9 8 Brachyelytrum "-~~~~9 9 Lygeum Nardus Anisopogon 7 0 Ampelodesmos -8 9 6 _ Stipa 88 51 Nassella Po. 62 81 Piptatherum - I Brachypodium Avena 59 68 ~ Bromus 1 3 100 Triticum Diarrhena X Melica Glyceria Aristida - Aristid. Amphipogon Arundo MrMoolilniniaia Arundin. Phragmites Danthonia - Danthoni. 95 80 Eragrostis Uniola 9.8 84 Zoysia Chlorid. 97 98t Spartina Sporobolus Distichlis 58j Chasmanthium Centothec. ' 74 ,- Pennisetum Miscanthus Panic. Panicum Pseudosasa Bambus. Chusquea 72 I Ehrharta Oryza Ehrhart. Leersia Appendix III-Figure A, cp restriction sites Volume 88, Number 3 Grass Phylogeny Working Group 441 2001 Phylogeny and Classification of Poaceae Flagellaria 1 00 Elegia Baloskion Joinvillea 96 Anomochloa Anomochlo Streptochaeta Anmcl . Pharus Phar. 100 10 1Guaduella ] Pue I. .1 ~0~~0n ~1 00 Eremitis 1 0 0 8 1 Pariana 100- 7i: -- Lithachne 63 -I Olyra Bambus. Buergersiochloa 9 4 1 00 Pseudosasa Chusquea 6 9 Streptogyna - Incertae sedis 98 I Ehrharta Oryza Ehrhart. 10 0- Leersia Phaenosperma 1 00 9 5 Anisopogon 6 99 Ampelodesmos 98 SPtiipptaa therum Nassella 1 00 Melica 72 Glyceria 9 6 Brachypodium ~96- Po. Avena o88 67 00 Bromus Triticum Diarrhena 1 10 00 1d00u L I u l-- ~ y geum ^~~ Nardus 1 Brachyelytrun- 100 Aristida I ] Arist id. Stipagrostis I• AMroulnindioa Arundin. 1 00- Phragmites 1 00 Merxmuelleram . 1I9 9 Karroochloa Dant honi. ' Austrodanthonia 00 Danthonia 53 - Danthoniopsis Pan i c. 67 ZeuggitieTsh ysanolaenCa entothec. 69 Gynerium - Incertae 100 sedis Panicum 74 70 Pennisetum Panic. 00-- Miscanthus Zea Chasmanthium- Centothec. 1 00 I Merxmuellerar . Centropodia 99 Eragrostis Uniola 0 0 1 000 Zoysia Chlorid. 9 7 Q_- -l62j-- Spartina - Sporobolus Distichlis Pappophorum Micraira Incertae sedis Appendix III-Figure B, ndhF 442 Annals of the Missouri Botanical Garden - Anomochlo. ] Pueli. Po. Arist id. Chlorid. Incertae sedis Arundin. 6 4 Thysanolaena 87 Chasmanthium J Centothec. Gynerium - Incertae sedis 95 1 Pennisetum 1 00[-- Miscanthus Panic. Zea Lithachne Pseudosasa Bambus. Chusquea 100 Oryza Ehrhart. Leersia Appendix III-Figure C, rbcL Volume 88, Number 3 Grass Phylogeny Working Group 443 2001 Phylogeny and Classification of Poaceae Joinvillea 51 Ehrharta I EtOryza ] Ehrhart. 61 Oryza Olyra B Pseudosasa B aambubus.s . 56 99 Lygeum Nardus 88 ~~~~~88 I Anisopogon 63 Stipa Po. 8 2 1 Avena I Bromus 67 Aristida Stipagrostis 100 Molinia Phragmites Arundin. Arundo Amphipogon Merxmuelleram . 70 g97 Karroochloa 99 I Austrodanthonia Dant honi. | 73 -,~~~~~~~~73 Danthonia 51 I Merxmuellerar . 63 I Centropodia 1000 Eragrostis Chlorid. 9' 9-I Pappophorum ~I ~ Spartina Micraira -Incertae sedis Thysanolaena Cent ot hec. Chasmanthium Gynerium -Incertae sedis 80 I Panicum 93 I Pennisetum Paann ic. 98 I Miscanthus I Zea Appendix III-Figure D, rpoC2 444 Annals of the Missouri Botanical Garden Flagellaria Joinvillea Anomochloa Streptochaeta Anomochlo. Pharus Phar. Puelia - Pueli. Eremitis Pariana Buergersiochloa Lithachne Bambus. Olyra Pseudosasa Chusquea Ehrharta EOryza jEhrhart. Lygeum Nardus Anisopogon Nassella Melica Po. Glyceria Brachypodium Diarrhena Avena Bromus Triticum Streptogyna Incertae sedis Aristida A rist id. Danthonia - Danthoni. Molinia Phragmites J Arundin. Eragrostis Chlorid. Sporobolus i Thysanolaena - Cent ot hec. Chasmanthium Danthoniopsis Panicum Pennisetum Panic. Miscanthus Zea Appendix III-Figure E, phy B Volume 88, Number 3 Grass Phylogeny Working Group 445 2001 Phylogeny and Classification of Poaceae Joinvillea Streptochaeta - Anomochlo. Pharus - Phar. Z5- Lithachne -7 Bambus. Chusquea _ Ehrharta R6Q9|- Oryza Ehrhart. L-- eersia 83 ,83 _-4-Brachyelytrum -- 90 Lygeum Nardus Brachypodium 67 Melica Glyceria Anisopogon 99 Stipa o g9~ Q ~8 1- Nassella Piptatherum 6 4 1 Avena 19 L9 0 Bromus Triticum Ampelodesmos Diarrhena 76 isti I--- Aristida. I__| ~idStipagrostis 62 1 Amphipogon - Arundin. 1-00 Spartina Chlorid. Sporobolus Chorid. 1 8 _ 93 Karroochloa *--- Danthoni. Austrodanthonia Dant honi. Arundo Arundin. Merxmuellera m. - Dant honi. Merxmuellera r. - Ch o r i d. Danthonia Dant honi. Molinia Arundin Phragmites Arundn. Centropodia Chlorid. Eragrostis ~J 79 Pac Panicum __=__ IPennisetum Pa Miscanthus Thysanolaena Chasmanthium Centothec. Zea Panic. Gynerium Eriachne Incertae sedis Micraira Appendix III-Figure F, ITS 446 Annals of the Missouri Botanical Garden Anomochloa - Anomochlo. Pharus - Phar. 100 | Eremitis Pariana Barmbus. Oryza Ehrhart. 98 j Lygeum Triticum Po. 64 7776 4 7 Glyceria Melica Danthoniopsis 61 Pennisetum 63 Panic. 93 | Miscanthus Zea 67 100 Karroochloa Danthoni. Austrodanthonia 96 9 5 Merxmuellerar . 5 CChhloorriidd.. 1 00 Centropodia Merxmuelleram . - Danthoni. Chusquea Bambus. Appendix III-Figure G, GBSSI Volume 88, Number3 Grass Phylogeny WorkingG roup 447 2001 Phylogeny and Classificationo f Poaceae Flagellaria 97 = Elegia Baloskion Joinvillea Streptochaeta _ Anomochlo. Anomochloa Pharus - Phar. Pariana Eremitis Bambus. Chusquea Puelia Pue I -Guaduella Puel_ - -Pseudosasa 80 LI -Lithachne BBaamm bbu uss Olyra . Buergersiochloa Zeugites - Centothec. Streptogyna - Incertae sedis Oryza LI Leersia Ehrhart. Ehrharta Phaenosperma Brachyelytrum Lygeum Nardus Diarrhena 86 Melica Glycerias Ampelodesmos S_t. ip. a. . Po. er u Plitarnerum Anisopogon Nassella Brachvpodium Triticumo Triticum Q A A Ps:',&#io