Notes

Outline

The New Phylogeny of the Lilioid Monocotyledons Alan W. Meerow USDA-ARS-SHRS National Germplasm Repository 13601 Old Cutler Road Miami, FL 33158 USA

Phylogenetic analysis (cladistics) Based upon principles formally enumerated by Hennig (1966). Defines any inclusive group of organisms (a clade), regardless of taxonomic rank, by the presence of one or more shared, derived character states (synapomorphies). Such a group is described as being monophyletic. To accept a taxonomic grouping based on shared primitive character states (plesiomorphies) results in polyphyletic (taxonomic groups with multiple evolutionary origins) or paraphyletic groups (groups from which one or more members of common descent are excluded). Misinterpretations of homology (parallel evolution or convergence) also results in polyphyletic groups.

Slide 3

Phylogenetic analysis (cladistics) Parsimony: the shortest possible phylogenetic tree (cladogram), i.e., requires the least number of steps (character state changes), is the most accurate. Computer programs used for cladistic analysis attempt to find the shortest possible (i.e., the most parsimonious) phylogenetic tree produced by a particular character state matrix.

Phylogenetic analysis (cladistics) The larger the number of informative (versus neutral or ambiguous) characters in the matrix, the smaller the number of equally parsimonious trees. Initial polarization of characters states is most frequently accomplished by outgroup comparison.  A designated outgroup(s) is/are generally included in the matrix. Various weighting schemes or other assumptions about character evolution can be applied to some or all of the data. Several confidence tests of a particular phylogenetic resolution are employed by systematists, the most widely used being the bootstrap analysis (Felsenstein 1985, 1988; Hillis and Bull 1993; Sanderson 1989). A high bootstrap value for a particular clade is a sign of robustness; a low value means that the clade is not well supported.

"Cladistics + molecular approaches to..." Cladistics + molecular approaches to phylogeny reconstruction have provided quantum leaps in taxonomic science over the past 20 years. Some of the most rapid and radical changes in our understanding of flowering plant phylogeny have been concentrated among the moncotyledons. Illustrative of the speed at which new information is being generated, there have been three major publications on the evolutionary biology and classification of the monocotyledons since 1995 (Rudall et al. 1995; Kubitzki 1998; Wilson and Morrison 2000). Significant changes in our understanding of monocot phylogeny have occurred between the release of each.

"Huber (1969)" Huber (1969) radically challenged concepts of familial and ordinal limits of the monocotyledons. Emphasized less conspicuous characters, particularly embryological characters, over gross floral or vegetative morphology Highlighted the heterogeneity present in many traditional monocot families, especially Liliaceae Juss. Refined and placed into phylogenetic context by Dahlgren and coworkers (Dahlgren and Clifford 1982; Dahlgren and Rasmussen 1983; Dahlgren, Clifford, and Yeo 1985). In Dahlgren et al.’s (1985) synthesis, superorder Liliiflorae encompasses 5 orders: Dioscoreales, Asparagales, Melanthiales, Burmanniales and Liliales.

"Dahlgren et al." Dahlgren et al. (1985): the families of monocots rich in geophytes are classified into two orders, Asparagales and Liliales, that have evolved many traits in parallel. Dahlgren et al. (1985) listed 16 characters that differentiated Liliales and Asparagales, but most do not occur in all taxa and several at least are plesiomorphic states.

"Cladistics + molecular approaches to..." Cladistics + molecular approaches to phylogeny reconstruction have provided quantum leaps in taxonomic science over the past 20 years. Some of the most rapid and radical changes in our understanding of flowering plant phylogeny have been concentrated among the moncotyledons. Illustrative of the speed at which new information is being generated, there have been three major publications on the evolutionary biology and classification of the monocotyledons since 1995 (Rudall et al. 1995; Kubitzki 1998; Wilson and Morrison 2000). Significant changes in our understanding of monocot phylogeny have occurred between the release of each.

"Huber (1969)" Huber (1969) radically challenged concepts of familial and ordinal limits of the monocotyledons. Emphasized less conspicuous characters, particularly embryological characters, over gross floral or vegetative morphology Highlighted the heterogeneity present in many traditional monocot families, especially Liliaceae Juss. Refined and placed into phylogenetic context by Dahlgren and coworkers (Dahlgren and Clifford 1982; Dahlgren and Rasmussen 1983; Dahlgren, Clifford, and Yeo 1985). In Dahlgren et al.’s (1985) synthesis, superorder Liliiflorae encompasses 5 orders: Dioscoreales, Asparagales, Melanthiales, Burmanniales and Liliales.

"Dahlgren et al." Dahlgren et al. (1985): the families of monocots rich in geophytes are classified into two orders, Asparagales and Liliales, that have evolved many traits in parallel. Dahlgren et al. (1985) listed 16 characters that differentiated Liliales and Asparagales, but most do not occur in all taxa and several at least are plesiomorphic states.

"The two important and consistent..." The two important and consistent characters that separate the two orders are the presence of phytomelan in the seed coat of Asparagales (Huber 1969), and the universal absence of septal nectaries in Liliales (Rudall et al. 2000). The boundaries between the two orders are difficult to define on morphological grounds alone, though multiple gene sequences support these two orders as monophyletic groups (Chase et al. 2000).

"Phylogenetic analyses of the monocotyledons..." Phylogenetic analyses of the monocotyledons (morphological and gene sequence data) have supported this classification with some amendment (Duvall et al. 1993; Stevenson and Loconte 1995; Chase et al. 1995a, b, 2000). Melanthiales is no longer recognized as distinct from Liliales [Angiosperm Phylogeny Group (AGP) 1998]. Burmanniaceae is placed within Dioscoreales (Caddick et al. 2000). Iridaceae Juss. and Orchidaceae Juss. have been transferred from Liliales to Asparagales, primarily on the basis of DNA sequence data (Chase et al. 1995a, 2000).

"The most recent analysis of..." The most recent analysis of molecular data (Chase et al. 2000) across all of the monocotyledons utilized a combined matrix of three genes: plastid rbcL, plastid atpB and nuclear 18S ribosomal DNA.

"The exact relationships among the..." The exact relationships among the lilioid orders Asparagales, Dioscoreales, Liliales and Pandanales are not yet well-resolved.

"The results of these and..." The results of these and other analyses has resulted in a formal reclassification of the flowering plants along a strict criterion of monophyly, published by the Angiosperm Phylogeny Group (APG 1998).

Liliales Dahlgren et al. (1985) recognized ten families in Liliales:

Liliales Plastid DNA sequences have since resulted in Iridaceae (including Geosidridaceae) and Orchidaceae (including Apostasiaceae and Cypripediaceae) being transferred to Asparagales (Chase et al. 1995a). The most current classification (AGP, 1998) recognizes the following nine families: Alstroemeriaceae, Campynemataceae Dumort, Colchicaceae, Liliaceae, Luzuriagaceae Kunth, Melanthiaceae Batsch, Philesiaceae Dumort, Ripogonaceae, and Smilacaceae Vent.

Liliales Cladistic analyses of combined plastid genes rbcL and trnL-F resolves four main lineages within the Liliales (Rudall et al. 2000): 1) Liliaceae (including Calachortaceae and some former members of Uvulariaceae), Philesiaceae, and Smilacaceae 2) Campynemataceae 3) the colchicoid lilies (Colchicaceae including Petermannia F. Muell.and Uvularia L.), Alstromeriaceae and Luzuriaga R. & P. 4) Melanthiaceae (including Trilliaceae Lindl.). The relationships between these lineages are not well resolved.

Liliales Rudall et al. (2000) suggest combining Philesiaceae and Ripogonaceae with Smilacaceae.

Liliaceae Liliaceae sensu stricto appear to consist of two main groups (Rudall et al., 2000). The larger clade based on plastid sequences is made up of three subclades.

Liliaceae 1) a Clintonia Raf.-Gagea Salisb. clade.

Liliaceae 2) the core Liliaceae (Lilium L., Fritillaria L., Nomocharis Franch., Cardiocrinum Endl.)

Liliaceae 3) a Tulipa L.-Erythronium L. group.

Liliaceae The smaller main clade represents part of what Dahlgren et al. (1985) treated as Uvulariaceae (Tricyrtis Wall.and allies).

Liliaceae The affinities of Calochortus remain controversial. Rudall et al. 2000 sister to Liliaceae in the rbcL/trnL-F trees. embedded between the two main clades of the family in the combined analyses.

Liliaceae Patterson et al. (1998), using the more rapidly evolving chloroplast gene ndhF, resolved Calochortus as sister to Tricyrtis. Tamura (1998a) recognized Calochortaceae, isolating Calochortus in the monogeneric bulbous tribe Calochorteae Melchior.  The remaining four rhizomatous genera (including Tricyrtis) were placed in the tribe Tricyrtideae K. Krause.

Colchicaceae This cormous and rhizomatous family includes the horticultural genera Gloriosa L., Sandersonia Hook., Littonia Hook., and Colchicum L. and 5-7 other genera, including Uvularia, the only North American genus of the family.

Melanthiaceae Problematic taxonomic history (Zomlefer 1997). Appears to be well defined morphologically by extrorse anthers and three styles (though these characters occur elsewhere in Liliales). Tamura (1998c) did not include Trilliaceae (Zomlefer 1996; Tamura 1998d) in his treatment. Trilliaceae resolves as embedded within Melanthiaceae in many molecular analyses (Rudall et al. 2000).

Alstroemeriaceae New World endemic family generally resolves as an isolated lineage most closely related to the genus Luzuriaga and a monophyletic Colchicaceae (Chase et al. 1995a, 2000; Rudall et al. 2000).

Alstroemeriaceae Bayer (1998) recognized five genera: Alstroemeria L., Bomarea Mirb., Leontochir Phil., Schickendantzia Pax, and Taltalia Ehr. Bayer. The latter two genera are segregates from Alstroemeria and are not supported by cladistic analyses of chloroplast DNA variation (Aagesen and Santo 1998).  In Aageson and Santo’s (1998) analysis, Bomarea and Leontochir are sister genera, and the Andean species of Alstroemeria are embedded within the Brazilian species of the genus.

Asparagales Thirty-one families were included in Asparagales by Dahlgren et al. (1985). Analyses of rbcL sequence data (Chase et al., 1995a) resulted in the transfer of  Orchidaceae and Iridaceae from Lililes (Dahlgren et al, 1985) to Asparagales. Several families treated by Dahlgren et al. (1985) within Asparagales have been moved to Liliales. The Angiosperm Phylogeny Group recognizes 29 families in the order (AGP, 1998).

Asparagales Asparagales consistently forms two groups (Rudall et al., 1997). The “lower” asparagoids a predominant simultaneous microsporogensis. frequently inferior ovaries. The “higher” asparagoids uniformly successive microsporogenesis. frequent occurrence of superior ovaries.

Asparagales Relationships between the families within each group have unfortunately presented problems (Chase et al., 1995a), and macromorphological synapomorphies for many of the families are not apparent (Fay et al., 2000). Fay et al. (2000) presented analyses of four plastid sequence data sets that produced trees largely congruent with the rbcL topology of Chase et al. (1995a), but with increased bootstrap support for many of relationships resolved among the families (Fig. 3).

Asparagales Within the higher asparagoids, Amaryllidaceae and Alliaceae form a sister relationship with Agapanthaceae sister to both.  Analysis of plastid sequences alone place Alliaceae as sister to an Agapanthaceae/ Amaryllidaceae clade (Fay and Chase, 1996; Meerow et al., 1999). Themidaceae (the former tribe Brodiaeeae of Alliaceae) is allied with Hyacinthaceae and Aphyllanthaceae.

Asparagales Agavaceae is included in a clade with Anthericaceae and several smaller families (including Hostaceae).

Slide 37

Asparagales - Agavaceae Bogler and Simpson (1995); Bogler (1995) and other cladistic studies resolve two main clades. Superior-ovaried (Yucca and Hesperaloe) Inferior-ovaried (Agave, Furcraea and Beschorneria and herbaceous genera). Nolinaceae separate family.

Asparagales A broadly circumscribed Convallariaceae is united with Asparagaceae and Laxmanniaceae.

Slide 40

Asparagales Themidaceae Fay and Chase (1996), on the basis of a phylogenetic analysis of rbcL sequence data, removed Agapanthus from Alliaceae, and resurrected the family Themidaceae for the western North American and Mexican genera of Alliaceae (tribe Brodiaeeae).

Asparagales Hyacinthaceae This family has been recognized as a natural group within Liliaceae sensu lato on the basis of anatomical (Fuchsig 1910) and embryological (Schnarf 1929; Wunderlich 1937; Buchner 1948) characters. Speta (1998) recognizes about 67 genera and 900 species in the family, subdivided into five subfamilies of which four are well-supported by molecular data (Chase et al. 1995; Fay and Chase 1996). Molecular systematic work in progress (J. Manning pers. comm.) favors a more conservative circumscription of genera than that introduced by Speta (1998). Important horticultural genera include Eucomis L’ Hér., Hyacinthus L., Lachenalia Jacq. F. ex Murray, Muscari Mill., Ornithogalum L., Scilla L., and Veltheimia Gled.

Asparagales Hyacinthaceae

Asparagales Alliaceae In Fay and Chase’s (1996) rbcL trees, Alliaceae forms two subclades: An American/South African group (Tulbaghia is the only endemic African genus of the family) and An Allium/Milula clade.

Asparagales Alliaceae Meerow et al.’s (1999) combined plastid sequence analysis supported this resolution of Alliaceae, though in trees resulting from the trnL-F matrix alone, Tulbaghia is sister to the rest of the family. Fay and Chase (1996) proposed three subfamilies, Allioideae (Allium and Miulla), Tulbaghioideae (Tulbaghia) and Gilliesioideae (for all endemic American genera, e.g., Leucrocoryne, Iphieon). In Rahn’s (1998) treatment, 13 genera are recognized.

Asparagales Amaryllidaceae Amaryllidaceae is one of the few families of Asparagles well-defined by other than molecular characters, namely umbellate cymes, inferior ovaries, and unique alkaloid chemistry (Meerow and Snijman, 1998). The four most recent infrafamilial classifications of Amaryllidaceae are those of Traub (1963), Dahlgren et al. (1985), Müller-Doblies and Müller-Doblies (1996) and Meerow and Snijman (1998).

Asparagales Amaryllidaceae Traub's scheme included Alliaceae, Hemerocallidaceae and Ixioliriaceae as subfamilies, following Hutchinson (1934, 1959) in part.  Within his subfamily Amarylloideae, he erected two informal taxa, "infrafamilies" Amarylloidinae and Pancratioidinae, both of which were polyphyletic (Meerow, 1995). Dahlgren et al. (1985) dispensed with any subfamilial classification above the level of tribe, recognizing eight, and treated as Amaryllidaceae only those genera in Traub's Amarylloideae.  Stenomesseae Traub and Eustephieae (Pax) Hutch. were combined. Meerow (1995) resurrected Eustephieae from Stenomesseae and suggested that two new tribes might need to be recognized, Calostemmateae D. & M-D. and Hymenocallideae (D. & U. M-D.) Meerow.

Asparagales Amaryllidaceae Müller-Doblies and Müller-Doblies (1996) recognized ten tribes (among them Calostemmateae) and nineteen subtribes, many of them monogeneric. Meerow and Snijman (1998) recognized 13 tribes, with two subtribes only in one of them.  A discussion of character evolution within the family can be found in Meerow (1995) and Meerow et al. (1999).

Asparagales Amaryllidaceae Fay and Chase (1996), on the basis of a phylogenetic analysis of rbcL sequence data, recircumscribed Amaryllidaceae to include Agapanthus, previously included in Alliaceae, as subfamily Agapanthoideae.  Meerow et al. (1999) opted to recognize a monotypic Agapanthaceae, which has been adopted by the Angiosperm Phylogeny Group (AGP, 1998).

Asparagales Amaryllidaceae Tribe Amaryllideae J. St.-Hil., entirely southern African with the exception of pantropical Crinum, was sister to the rest of Amaryllidaceae with very high bootstrap support. The tribe is also well marked morphologically: cartilaginous leaf fibers, sclerenchymatous scapes, unique bisulculate pollen with spinulose exines, unitegmic ovules, and bulbiform seeds.

Amaryllideae

Asparagales Amaryllidaceae The remaining two African tribes of the family, Haemantheae (Pax) Hutch. (including Gethyllideae Dumort) and Cyrtantheae Salisb., were well supported, but their position relative to the Australasian Calostemmateae and a large clade comprising the Eurasian and American genera, is not clear. Unexpected sister relationship of the American genera and Eurasian clade.

Cyrtantheae

Haemantheae

Slide 55

Still elusive…. The relationships of the Eurasian genera have not yet been fully resolved.

Asparagales Amaryllidaceae Meerow et al. (2000a, b), American Amaryllidaceae, internal transcribed spacer (ITS) of nuclear ribosomal DNA. Two major subclades. “Hippeastroid” clade diploid (n = 11) primarily extra-Andean element comprising the genera treated as the tribe Hippeastreae in most recent classifications (Dahlgren et al. 1985, Müller-Doblies and Müller-Doblies 1996; Meerow and Snijman 1998).

Asparagales Amaryllidaceae

Asparagales Amaryllidaceae The second subclade constitutes the tetraploid-derived (n = 23), Andean-centered tribes. Characterized by 3 consistent deletions, two in the ITS1 and one in the ITS2 regions. A petiolate-leafed clade containing elements of both Eucharideae (pax) Hutch. and Stenomesseae was resolved with a bootstrap = 93%. Several genera within the hippeastroid subclade resolve as polyphyletic (Rhodophiala Presl., Zephyranthes Herb.) Possible reticulate evolution (i.e., early hybridization)?

Petiolate-leafed Andean subclade

The new tribe Clinantheae

Asparagales Amaryllidaceae In both subclades, there is a small tribe that is sister to the rest of the subclade, the Eustephieae in the Andean group, and the Griffineae Rav. in the hippeastroid clade.  These two small tribes may represent either ancestral or merely very isolated elements of their respective clades.

Asparagales Convallariaceae Sensu Dahlgren et al. (1985): rhizomatous perennial herbs with a primarily northern hemisphere distribution, particularly abundant in eastern Asia. The seeds of the berry fruits lack phytomelan. Three tribes are recognized: Polygonateae Benth., Ophiopogoneae Endl., and Convallarieae (Conran and Tamura 1998). Some workers recognize a fourth, Aspidistreae (Dahlgren et al. 1985).

Asparagales Convallariaceae The rbcL analysis of Chase et al. (1995a) suggested that the Ophiopogoneae should be allied with Ruscaceae Spreng. ex Hutch. and Asparagaceae Juss. Rudall et al.’s (1997) analysis of rbcL sequences indicated that Convallariaceae were polyphyletic, and intergrade with Nolinaceae Nakai, Dracaenaceae Salisb. and Ruscaceae, all woody families of woody plants. Yamashita and Tamura (2000) used the plastid gene tmK (inclusive of matK), along with rbcL to investigate the same problem, and were not able to resolve a monophyletic Convallariaceae. However, the tribes Polgonateae and Ophiopogoneae were resolved as monophyletic, and the Convallarieae and Aspidistreae formed a clade, results contrary to Rudall et al.’s (1997) conclusions. The Angiosperm Phylogeny Group (AGP 1998) included Nolinaceae, Dracaenaceae and Ruscaceae within Convallariaceae.

Asparagales Asphodelaceae Well-defined by 4 synapomorphies (Smith and Van Wyk, 1998): parenchymatous inner bundle-sheath cells (except Bubinella and Kniphofia). simultaneous microsporogenesis. aril or stropiole developed from an annular invagination at the distal part of funcile (Stenar, 1928; Schnarf, 1929). chrysophanol in the roots (Van Wyk et al., 1995b, c).

Asparagales Asphodelaceae

Asparagales Hemerocallidaceae The modern consensus on this morphologically diverse family (Clifford et al. 1998) unites the daylilies (Hemerocallis L.) with New Zealand flax (Phormium J. R. Forst. & G. Forst.) and 11 other genera, including Dianella Lam. Ex Juss. Previously treated as a monogeneric family (Hemerocallis; Dahlgren et al. 1985). The evidence for this unsuspected alliance is from rbcL sequence analyses (Chase et al., 1995a), as well as palynological (Kosenko, 1994) and serological evidence (Chupov, 1987).

Asparagales Iridaceae Classified in Liliales near Colchicaceae by Dahlgren et al. (1985) on the basis of extrorse anthers, non-phytomelanous seeds, mottled tepals, perigonal nectaries and nuclear endosperm development. Perigonal nectaries are now known to represent an independent, derived state in Iridaceae, as are mottled tepals, and septal nectaries are the ancestral state for the family (Goldblatt 1998). The more ancestral Iridaceae are characterized by helobial endosperm formation. Despite lack of clear cut morphological links to Asparagales, multiple gene sequence analyses place Iridaceae well within this order (Chase et al. 1995a, 2000; Fay et al. 2000).

Asparagales Iridaceae  The family occupies an isolated position among the well-resolved clades of the lower asparagoids, and probably represents a relatively ancient divergence from the rest of the order (Goldblatt 1998). Three subfamilies are recognized by Goldblatt (1998).

Asparagales Iridaceae Isophysidoideae (1 monotypic genus, Isophysis, in Tasmania). Nivenioideae (7 genera, Australia, South Africa and Madagascar, three with woody aerial stems).

Asparagales Iridaceae Iridoideae 27 genera, cosmopolitan.

Asparagales Iridaceae Ixioideae (27 genera, mostly African).

Lingering Questions Within the Asparagales: a precise understanding of the relationships among the basal, “lower” families. Within Liliales, the relationships among the component families appear more resolute, but uncertainty remains concerning the exact affinities of Calochortus the relationships of Alstroemeriaceae and the accurate alignment of the genera formerly treated as Uvulariaceae. Finally, the exact relationships among the lilioid orders Asparagales, Dioscoreales, Liliales and Pandanales are not yet well-resolved.

Acknowledgments