Larry Hufford

Phylogeny of Saxifragales (Angiosperms, Eudicots): Analysis of a Rapid, Ancient Radiation

Rapid, ancient radiations pose one of the most difficult challenges for phylogenetic estimation. We used DNA sequence data of 9,006 aligned base pairs from five genes (chloroplast atpB, matK, rbcL, and 18S and 26S nrDNA) to elucidate relationships among major lineages of Saxifragales (angiosperms, eudicots). These relationships were poorly supported in previous studies, apparently because the lineages originated in rapid succession. Using an array of methods that explicitly incorporate assumptions about evolutionary process (weighted maximum parsimony, maximum likelihood, LogDet/paralinear transformed distances), we show that the initial diversification of Saxifragales was indeed rapid. We suggest that the poor resolution of our best phylogenetic estimate is not due to violations of assumptions or to combining data partitions having conflicting histories or processes. We show that estimated branch lengths during the initial diversification are exceedingly short, and we estimate that acquiring sufficient sequence data to resolve these relationships would require an extraordinary effort (approximately 10(7) bp), assuming a linear increase in branch support with branch length. However, our simulation of much larger data sets containing a distribution of phylogenetic signal similar to that of the five sampled gene sequences suggests a limit to achievable branch support. Using statistical tests of differences in the likelihoods of topologies, we evaluated whether the initial radiation of Saxifragales involved the simultaneous origin of major lineages. Our results are consistent with predictions that resolving the branching order of rapid, ancient radiations requires sampling characters that evolved rapidly at the time of the radiation but have since experienced a slower evolutionary rate.

Rosidae and their relationships to other nonmagnoliid dicotyledons : a phylogenetic analysis using morphological and chemical data

Parsimony analysis was used to develop phylogenetic hypotheses for Rosidae and other nonmagnoliid dicotyledons, especially Asteridae. Rosidae were placed among «lower» Hamamelidae as the sister group of Platanus and Hamamelidaceae. «Higher» Hamamelidae (Fagales, juglandales, and Casuarinaceae), Dilleniidae, and Asteridae nest within a paraphyletic Rosidae. With some expansion, the traditional Asteridae are monophyletic. For example, the problematic Columellia was placed among Asteridae as the sister group of Caprifoliaceae. Asteridae were placed as the sister group of Ericales among rosids circumscribed as Corniflorae in recent classifications of Dahlgren (...)

Elucidating deep-level phylogenetic relationships in Saxifragaceae using sequences for six chloroplastic and nuclear DNA regions

To elucidate relationships at deep levels within Saxifragaceae we analyzed phylogenetically a data set of sequences for six DNA regions, four representing the chloroplast genome (rbcL, matK, trnL-trnF, psbA-trnH) and two from the nuclear genome (ITS and expansion segments of the 26S rDNA). A total of 6676 bp was aligned per taxon, 4559 bp and 1878 bp from the chloroplast and nuclear genomes, respectively. Chloroplast and nuclear trees agreed closely, prompting analysis of a combined, six-gene data set. Application of both parsimony and maximum likelihood methods yielded similar topologies. The use of different ITS alignments and the exclusion of hard-to-align ITS regions had little impact on either the final nuclear-based topology, or the shortest trees from the analysis of six genes. The affinities of two monotypic genera (Saxifragella and Saxifragodes) endemic to Tierra del Fuego were elucidated. Saxifragella is an early branching member of the North Temperate genus Saxifraga s. str.; Saxifragodes is sister to Cascadia, a genus endemic to Oregon and Washington. Long-distance dispersal from east Asia or western North America to South America may have played an important role in forming these and other similar disjunctions in the family. A number of well-supported clades are present, including Saxifraga s. str., Micranthes, Saxifragopsis/Astilbe, Chrysosplenium/Peltoboykinia, and the Boykinia and Heuchera groups. The use of additional characters has provided greatly increased resolution and internal support at deep levels. Saxifragaceae comprise two major lineages: Saxifraga s. str. (including Saxifragella) and all other genera of the family (the heucheroids). This major split is accompanied by general biogeographical and morpihological differences. Whereas Saxifraga s. str. is largely arctic to alpine in occurrence, the heucheroid clade is largely temperate in distribution, Saxifraga s. str. has a relatively uniform floral morphology (generally actinomorphic; 5 sepals, 5 petals, 10 stamens, 2 carpels), whereas the heucheroid clade encompasses actinomorphic and zygomorphic forms, as well as variation in the number of sepals, petals, stamens, and carpels. Deep-level relationships within both Saxifraga s. str. and the heucheroid clade are well resolved and supported. A phylogenetic classification of the family is provided.

Phylogenetic Relationships in Sarraceniaceae Based on rbcL and ITS Sequences

Sarraceniaceae, a New World family of carnivorous plants, consist of Sarracenia, Heliamphora, and the monotypic Darlingtonia. Despite extensive interest in these plants, few explicit phylogenetic hypotheses are available for the family. We investigate phylogenetic relationships in the family using sequences of the internal transcribed spacer regions of nuclear ribosomal DNA (ITS) and of the chloroplast gene rbcL. One representative of each genus was sequenced for rbcL, whereas representatives of all species in the family, except for H. heterodoxa and H. ionasii, were sequenced for ITS. Both the ITSand rbcL-based phylogenies demonstrate that Sarracenia and Heliamphora form a monophyletic group that is the sister group of Darlingtonia. Sarracenia and Heliamphora are each well supported monophyletic genera. In our restricted sampling of Heliamphora, H. nutans is the sister of H. minor and H. tatei. Our results provide little resolution of relationships among species of Sarracenia, but identify one clade that consists of all species except S. alata. Within this clade are two smaller clades comprising:-1) S. purpurea and S. leucophylla,-and-2) S. flava, S. minor, and S. psittacina. Sarraceniaceae have received considerable attention from biologists because of carnivory throughout the family. All members of Sarraceniaceae are found in nutrient-poor habitats and like other carnivorous plants use animals as a supplemental source of nitrogen that is taken up in the form of ammonia (Christensen 1977; Bradshaw and Creelman 1984; Jaffe et al. 1992). Sarraceniaceae, Cephalotaceae, and Nepenthaceae use pitcher-like (ascidiate) leaves as passive traps for the collection of animals. As has been suggested previously (Thanikaimoni and Vasanthy 1972; DeBuhr 1975) and shown in recent phylogenetic analyses (Albert et al. 1992; Chase et al. 1993), the ascidiate leaves shared by these three families are evolutionary parallel-

A phylogenetic analysis of Cunoniaceae

We present a cladistic analysis of genera of Cunoniaceae based on 44 structural characters using Wagner parsimony. A synthetic outgroup based on the attributes of Hamamelidaceae, Rosaceae and Fagaceae was used to polarize character states and root resulting cladograms. In all of the resulting cladograms, a clade consisting of Gillbeea, Brunellia, Spiraeanthemum, and Aistopetalum formed the sister group of the rest of the Cunoniaceae. The continued recognition of Brunellia in a monogeneric family separate from Cunoniaceae was not supported by our results. The hypothesis that apocarpy was secondarily derived within the Cunoniaceae in Brunellia and Spiraeanthemum was supported. The sister group of the above set of taxa shared the derived possession of interpetiolar stipules (although they originated in parallel in the Spiraeanthemum-Aistopetalum group). The group defined by interpetiolar stipules, consisting of the bulk of the Cunoniaceae, includes Eucryphia and Bauera, which have been placed often in monotypic families. Eucryphia is the sister group of a clade defined by bicarpellate gynoecia, which are derived within Cunoniaceae. Bauera is a member of a clade of subshrubs that includes Acrophyllum and Anodopetalum. The southern hemisphere family Cunoniaceae is centered largely in eastern Australia, New Guinea, and New Caledonia with a few genera in southern Africa, Madagascar, and the neotropics. Cunoniaceae have been discussed as including the most primitive members of Rosidae. As such, the family was allied traditionally with Rosaceae and Saxifragaceae s.l. (Baillon 1871; Engler 1928a; Hutchinson 1959; Schulze-Menz 1964; Wettstein 1935). Kalkman (1988), for example, considered Cunoniaceae to be the most plausible sister group of Rosaceae. Ingle and Dadswell (1956) noted the resemblance of wood features among Cunoniaceae, Dilleniaceae and Theaceae. Takhtajan (1969) linked Rosidae to Dilleniales through Cunoniaceae. Recently, similarities have been re-emphasized between Cunoniaceae and certain basal Hamamelidae, particularly Hamamelidaceae (Dickison 1989). Hence, the Cunoniaceae appear positioned at a critical transitional point in the evolution of higher dicotyledonous groups (see also Ehrendorfer 1977; Ehrendorfer et al. 1984) and, therefore, assume considerable phylogenetic importance. Despite the key position of Cunoniaceae for understanding the radiation of higher dicotyledonous groups, numerous questions remain about patterns of character evolution and relationships within the family. This cladistic analysis of the genera of Cunoniaceae addresses these questions. Nineteen to 26 genera generally are circumscribed in the family. They are woody plants with decussate leaves and most have interpetiolar stipules. The decussate leaf arrangement has been hypothesized (Dickison and Rutishauser 1990; Kalkman 1988) to support the monophyly of Cunoniaceae. The genera have diverse inflorescence, flower and fruit forms which have created problems for understanding relationships within the family. A significant problem is the relationship of two problematic genera, Bauera and Eucryphia, to Cunoniaceae. Although often allied with Cunoniaceae (Bausch 1938; Dickison 1978; Hils 1989; Huber 1963; Jay 1968; Patel 1990; Planchon 1854), each is generally placed in a monotypic family. Both Bauera and Eucryphia have a decussate leaf arrangement. Eucryphia has interpetiolar stipules. Until recently, Bauera had been considered exstipulate; however, it has been shown to have paired, free stipules with the opposite leaves at each node (Dickison and Rutishauser 1990; Hils 1989). Dickison and Rutishauser (1990) hypothesized that both Bauera and Eucryphia are nested within Cunoniaceae as

The major clades of Loasaceae: phylogenetic analysis using the plastid matK and trnL-trnF regions

Phylogenetic analyses of Loasaceae that apply DNA sequence data from the plastid trnL-trnF region and matK gene in both maximum-parsimony and maximum-likelihood searches are presented. The results place subfamily Loasoideae as the sister of a subfamily Gronovioideae-Mentzelia clade. Schismocarpus is the sister of the Loasoideae-Gronovioideae-Mentzelia clade. The Schismocarpus-Loasoideae-Gronovioideae-Mentzelia clade is the sister of Eucnide. Several clades in Loasoideae receive strong support, providing insights on generic circumscription problems. Within Mentzelia, several major clades receive strong support, which clarifies relationships among previously circumscribed sections. Prior taxonomic and phylogenetic hypotheses are modeled using topology constraints in parsimony and likelihood analyses; tree lengths and likelihoods, respectively, are compared from constrained and unconstrained analyses to evaluate the relative support for various hypotheses. We use the Shimodaira-Hasegawa (SH) test to establish the significance of the differences between constrained and unconstrained topologies. The SH test rejects topologies based on hypotheses for (1) the placement of gronovioids as the sister of the rest of Loasaceae, (2) the monophyly of subfamily Mentzelioideae as well as Gronovioideae and Loasoideae, (3) the monophyly of Loasa sensu lato as circumscribed by Urban and Gilg, and (4) the monophyly of Mentzelia torreyi and Mentzelia sect. Bartonia.

A phylogenetic analysis of Hydrangeaceae based on sequences of the plastid gene matK and their combination with rbcL and morphological data

DNA sequences of the plastid gene matK were used alone and in combination with rbcL and morphological data in analyses of phylogenetic relationships in Hydrangeaceae. A suggested relationship of Hydrostachys to Hydrangeaceae was examined, but the maximum parsimony analyses of the matK data and the data set that combines matK and rbcL place Hydrostachys outside of Hydrangeaceae. The DNA sequence data sets both alone and in combination produced congruent results. A Jamesia + Fendlera clade (= subfamily Jamesioideae) was the sister of the rest of the family (= subfamily Hydrangeoideae). Two tribes, Philadelpheae and Hydrangeeae, are recognized in Hydrangeoideae. Philadelpheae included three primary clades: (1) Philadelphus + Carpenteria, (2) Deutzia + Kirengeshoma, and (3) Fendlerella + Whipplea. Relationships in Hydrangeeae remain poorly resolved. A Cardiandra + Deinanthe clade was placed robustly as the sister of the rest of the Hydrangeeae only in analyses in which matK sequences were combined with other data. Broussaisia, Decumaria, Dichroa, Pileostegia, Platycrater, and Schizophragma were nested among species of Hydrangea. Relationships in this Hydrangea clade (the most inclusive monophyletic group that included species of Hydrangea) were poorly resolved, except for the monophyly of (1) Decumaria, Pileostegia, and Schizophragma; (2) Broussaisia, Dichroa, Hydrangea hirta, and Hydrangea macrophylla; (3) Platycrater, Hydrangea involucrata, and Hydrangea aspera; and (4) Hydrangea anomala and Hydrangea section Cornidia. Conflict between the morphological and matK data weaken support for the monophyly of both Philadelpheae and the Hydrangea clade; however, the addition of the morphological data strengthens support for Hydrangeoideae, Hydrangeeae, and the sister group relationship of the Hydrangea clade and Cardiandra + Deinanthe.

Phylogeny of Amorpheae (Fabaceae: Papilionoideae)

The legume tribe Amorpheae comprises eight genera and 240 species with variable floral form. In this study, we inferred a phylogeny for Amorpheae using DNA sequence data from the plastid trnK intron, including matK, and the nuclear ribosomal ITS1, 5.8S, and ITS2. Our data resulted in a well-resolved phylogeny in which the tribe is divided into the daleoids (Dalea, Marina, and Psorothamnus), characterized by generally papilionaceous corollas, and the amorphoids (Amorpha, Apoplanesia, Errazurizia, Eysenhardtia, and Parryella), characterized by non-papilionaceous flowers. We found evidence for the paraphyly of Psorothamnus and for the monophyly of Dalea once D. filiciformis is transferred to monophyletic Marina. Errazurizia rotundata is more closely related to Amorpha than to the other errazurizias, and Eysenhardtia is supported to be monophyletic. The monotypic Parryella and Apoplanesia are placed within the amorphoids. Among Papilionoideae, trnK/matK sequence data provide strong evidence for the monophyly of Amorpheae and place Amorpheae as sister to the recently discovered dalbergioid clade.

Phylogenetic Relationships within Lithophragma (Saxifragaceae): Hybridization, Allopolyploidy, and Ovary Diversification

We explore phylogenetic relationships in Lithophragma through parsimony and maximum like- lihood estimation analyses of internal transcribed spacer sequences of 18S-26S ribosomal DNA. Results based on internal transcribed spacer sequences are compared with those from previous studies based on chloroplast DNA restriction site, morphological, and flavonoid variation. Phylogenetic analysis of internal transcribed spacer sequences produces a highly resolved topology with six main clades. These results suggest that three previously described species of Lithophragma are not monophyletic. Based on this topology and previous findings, two species not recognized in the most recent monograph, L. thompsonii and tetraploid L. bolanderi, are hypothesized to have arisen through hybridization and allopolyploidy, respectively. Flowers of Lithophrag- ma exhibit a diverse array of ovary positions ranging from what has been described as superior to deeply inferior. Analysis of ovary position in light of our phylogenetic results reveals a complex pattern of diversi- fication in Lithophragma. This pattern is explored through character mapping and correlation analyses and is found to be inconsistent with either an active or a passive trend toward greater inferiority. A weak relation- ship between variation in ovary position and the topological position of each taxon suggests homoplastic

THE ROLES OF ONTOGENETIC EVOLUTION IN THE ORIGINS OF FLORAL HOMOPLASIES

Homoplasy, manifest as convergence, parallelism, or reversal, is an important aspect of floral diversity. Few studies have focused on the evolution of floral homoplasies, and the role of ontogenetic evolution in their origins has been largely ignored. Four clades of angiosperms, the genera Besseya (Scrophulariaceae) and Eucnide (Loasaceae), the family Hydrangeaceae, and the order Piperales, were used to examine the roles of ontogenetic evolution in creating floral homoplasies. The possible modes of ontogenetic evolution were found to contribute unequally to the origin of homoplasies. Juvenilizing terminal deletions were common in the origin of homoplastic states but only slightly more so than novel substitutions and repatternings. Few homoplastic states arose via additions. Comparative studies on the origins of homoplasies are currently limited by the absence of developmental data for groups on which cladistic analyses have been conducted, by multiple, equally parsimonious cladograms, and by the low number of homoplastic states that are often available in clades.