| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
(American Journal of Botany. 2008;95:1015-1029.) doi: 10.3732/ajb.0800085 © 2008 Botanical Society of America, Inc. |
What's this? |
Systematics and Phytogeography |
2 Department of Plant Biology, Southern Illinois University Carbondale, Carbondale, Illinois 62901-6509 USA
Received for publication 4 March 2008. Accepted for publication 23 May 2008.
ABSTRACT
Loranthaceae (73 genera and ca. 900 species) comprise mostly aerial hemiparasitic plants. Three monotypic genera considered relicts are root parasites. The family is diverse in tropical areas, but representatives are also found in temperate habitats. Previous classifications were based on floral and inflorescence morphology, karyological information, and biogeography. The family has been divided into three tribes: Nuytsiae, Elytrantheae (subtribes Elytranthinae and Gaiadendrinae), and Lorantheae (subtribes Loranthinae and Psittacanthinae). Nuytsiae and Elytrantheae are characterized by a base chromosome number of x = 12, whereas subtribes Loranthinae (x = 9) and Psittacanthinae (x = 8) numbers are derived via aneuploid reduction. To elucidate the phylogeny of the family, we analyzed sequences from five genes (nuclear small and large subunit rDNA and the chloroplast genes rbcL, matK, and trnL-F) representing most genera using parsimony, likelihood, and Bayesian inference. The three root parasites, Nuytsia, Atkinsonia, and Gaiadendron, are supported as successive sister taxa to the remaining genera, resulting in a monophyletic group of aerial parasites. Three major clades are resolved each corresponding to a subtribe. However, two South American genera (Tristerix and Notanthera) and the New Zealand genus Tupeia, which were previously classified in subtribe Elytranthinae, are weakly supported as part of a clade representing the South American subtribe Psittacanthinae.
Key Words: matK • parasitic plants • phylogeny • rbcL • ribosomal DNA • Santalales • trnL-F intergenic spacer
Among the 12 clades of parasitic plants, the sandalwood order (Santalales) encompasses the widest array of habits including nonparasites, root parasites, and several variants of aerial (stem) parasites (Malécot and Nickrent, 2008
; Der and Nickrent, 2008). Aerial parasitism has arisen five times in different lineages within the order (Nickrent, 2002; Vidal-Russell and Nickrent, 2008), each of which includes plants commonly referred to as "mistletoes." A mistletoe is a hemiparastic shrub that attaches to a host stem and is also a member of Loranthaceae, Misodendraceae, "Santalaceae," or Viscaceae (thus, the term describes a habit among some members of santalalean clades). The success of this life form is apparent in that nearly all of the major species-level radiations have taken place in mistletoes, such as the genera Phoradendron and Viscum (Viscaceae) and Amyema and Psittacanthus (Loranthaceae). The genus Thesium ("Santalaceae") is the sole exception with over 300 mainly Old World species of root hemiparasites. With 73 genera and over 900 species, Loranthaceae (loranths) have the greatest number of mistletoes in Santalales. Although a few species occur in temperate Europe, Asia, Australia, New Zealand, and South America, its greatest diversity is in the Old and New World tropics, particularly in seasonally dry habits of Africa and Australia. The last classification that treated the family worldwide was by Engler (1897) and generic concepts among loranths have changed dramatically since that time. We present here the first comprehensive multigene phylogeny (sampling 60 of the 73 genera) for this important santalalean family.
Characteristics of Loranthaceae
Most loranths have dichlamydous flowers (i.e., with two perianth whorls) where the highly reduced calyx is referred to as the calyculus. When present it is found as a small lobe or rim at the top of the inferior ovary. The corolla can be choripetalous (petals free) or gamopetalous (petals fused), and petal number varies from four to nine. There are as many stamens as petals, and the degree of fusion to the corolla differs among the genera. The anthers are two- or four-locular, the majority being basifixed. Pollen grains have a characteristic triangular to trilobate shape. The ovary is inferior and uni- to plurilocular, and it does not have true ovules. A mamelon arises from the base of the ovary where the embryo sacs are formed. Because these mistletoes do not have ovules, technically they also do not have true seeds, although this term is used for the functional unit. Fruits are pseudoberries that contain one viscin-coated seed. Chromosome numbers in the family are conservative and normally characterize groups of related genera. The base chromosome number for the family is x = 12, but other numbers are observed such as x = 8, 9, and 10 (Barlow and Wiens, 1971). Most members of Loranthaceae are stem-parasitic plants, but three monotypic genera are root parasites: Nuytsia floribunda from Western Australia, Atkinsonia ligustrina from eastern Australia and Gaiadendron punctatum from Central and South America.
Compared to other angiosperms, Loranthaceae can be seen as highly specialized (owing to its parasitic habit). However, previous authors suggested plesiomorphic and apomorphic features for the family that will be addressed in the context of the phylogeny reported here. Plesiomorphies include terrestrial root parasitism, the presence of a haustorial system with epicortical roots (Hamilton and Barlow, 1963; Kuijt, 1969), and inflorescences composed of aggregations of cymose inflorescence units or triads (Barlow, 1966; Barlow and Wiens, 1973). Floral plesiomorphic conditions include an ovary that is partially chambered into loculi at the base with ovular lobes (Maheshwari et al., 1957); small, open, pale, actinomorphic, choripetalous, six-merous, entomophilous, hermaphroditic flowers; and anthers that are dorsifixed and versatile (Barlow and Wiens, 1973).
Life history of Loranthaceae
All species in Loranthaceae depend on a biotic agent for pollination. This dependence is evidenced by their floral morphology distinguished by two main syndromes: entomophily and ornithophily. Flowers in the first group are usually small (2–10 mm), white or greenish, and choripetalous, whereas in the last group, the flowers are typically large (30–160 mm), gamopetalous, and brightly colored. Small flowers are found in the New Zealand genus Tupeia, Cecarria from New Guinea, Barathranthus from Asia, and in many of the x = 8 taxa from the New World tropics. In the New World group, which includes genera such as Struthanthus, Cladocolea, Dendropemon, Phthirusa, and Oryctanthus, these floral reductions have made alpha taxonomy difficult; thus their relationships at the specific and generic level are often not clear.
Bird-pollinated flowers in Loranthaceae may be entirely red, or they may display banding patterns of contrasting colors, such as orange, yellow, green and even black. In many species, pollen is released in a burst and is deposited on the pollinator. Examples of bird-pollinated flowers can be seen in Helixanthera, Taxillus, Macrosolen, and Dendrophthoe.
Because of the close affinities between these parasitic plants and their dispersal vectors and pollinators, birds are thought to have contributed to the diversification of loranthaceous mistletoes (Feehan, 1985; Restrepo et al., 2002). Loranthaceae, like nearly all mistletoes, are highly dependent on their dispersers for reaching a suitable host. Different assemblages of birds have been described as dispersers of Loranthaceae in Africa, Australia, Asia, and South America (Docters van Leeuwen, 1954; Davidar, 1983; Liddy, 1983; Watson, 2001; Restrepo et al., 2002); however, a unique interaction has been observed in the temperate forest of South America where an arboreal marsupial disperses the seeds of Tristerix (Amico and Aizen, 2000). The association of the plants with their dispersers is very close, and it has been proposed that in some cases these species have coevolved (Reid, 1991).
All aerial Loranthaceae have viscous seeds, an essential adaptation that permits attachment to the host branch. Soon thereafter, germination begins and the radicular end of the embryo forms an attachment disc (the holdfast). Actual penetration of the host branch follows, then an internal haustorial system (endophyte) forms where the parasite establishes a connection to the host xylem. In most loranths the shoot system develops from the epicotyl, whereas in Tristerix aphyllus the seedling axis degenerates and shoot growth is adventitious from the internal endophyte.
Taxonomy and classification
Delimitation of genera within the family has long presented taxonomic difficulties. Tieghem (1894) accepted 118 genera of Loranthaceae, yet just three years later Engler (1897) demoted all of these to various sections within one genus, Loranthus. Subsequent work by Danser (1929, 1933) provided the basic framework for today’s classification. He improved Engler’s system by recognizing a number of new genera and by reinstating many of those proposed by Tieghem. More recent work has been focused at the continental scale, for example Balle (1954) and Polhill and Wiens (1998) in Africa; Barlow (1966, 1974) in Asia, Australia, and New Zealand; and Kuijt (1988, 2003 and reference therein; Feuer and Kuijt, 1979) in the New World. Danser (1933) divided the family into three tribes: Nuytsieae, which includes only Nuytsia, characterized by a unilocular ovary and spreading cotyledons; Elytrantheae with plurilocular ovaries and with cotyledons that spread during germination; and Lorantheae, with unilocular ovaries and the cotyledons hidden in the endosperm during germination. Tribe Elytrantheae was subdivided into two subtribes: Elytranthinae, which includes stem parasites with baccate fruits and immobile anthers, and Gaiadendrinae, which are root parasites with a drupaceous fruit and dorsifixed versatile anthers. Tribe Lorantheae was subdivided into two subtribes: Loranthinae and Psittacanthinae, both of which lack endosperm.
Chromosome counts for most of the loranth genera were made by Barlow and Wiens (1971). Based on these cytological data, Barlow and Wiens (1971, 1973) reclassified some genera and split the assemblage of Phrygilanthus into separate genera following previous concepts. Danser’s tribes and subtribes were retained, some South American genera were transferred from subtribe Loranthinae to Psittacanthinae and an unnamed tribe was added that included Ileostylus and Muellerina. Loranthinae are characterized by a base chromosome number x = 9 and Psittacanthinae by x = 8. The other two tribes (Elytrantheae and Nuytsieae) have x = 12.
Past molecular work has shown that Loranthaceae are monophyletic and sister to a clade composed of Schoepfiaceae and Misodendraceae (Nickrent and Duff, 1996; Nickrent et al., 1998; Nickrent and Malécot, 2001; Vidal-Russell and Nickrent, 2008). Cabrera (2002) constructed a molecular phylogeny of 43 genera of Loranthaceae that used the chloroplast gene matK. He found that the western Australian root parasite Nuytsia was sister to all other genera, followed by the eastern Australian endemic Atkinsonia. All the New World genera, except Gaiadendron and Tristerix, formed a well-supported clade, but taxon sampling from South America was limited. He recovered a clade that included several genera in subtribe Loranthinae. Within it, taxa were grouped with moderate support according to geographical distributions, thus distinguishing an Australasian and two African clades.
A more recent publication reported results of a molecular analysis for 47 genera in Loranthaceae based on nuclear ITS and chloroplast trnL-F (Wilson and Calvin, 2006). As seen in Cabrera (2002), Nuytsia was resolved as sister to all other genera in Loranthaceae, but most of the other nodes were not supported. Those clades that were supported (i.e., with bootstrap values >80) had already been recovered by Cabrera (2002).
No modern classification (either traditional or molecular) has included all genera of the family. The current study expands taxon sampling and uses more sequence information, thus providing the basis to establish such a classification. The evolutionary relationships between genera will be addressed, and the resulting phylogeny will be used to examine character evolution in the family with particular emphasis focused on floral features.
MATERIALS AND METHODS
Taxon sampling
Sixty of the 73 genera of Loranthaceae were sampled, which represents the full distribution of the family worldwide (Table 1). Although sequence variation in the chloroplast gene rbcL was generally conservative among genera, it was included to gain resolution at deeper levels in the phylogeny; sampling thus included only exemplars of major clades. Missing taxa are mainly from southeast Asia, some of which are very rare, and others are thought to be extinct. DNA extraction was attempted but without success from herbarium specimens of some of these rare taxa. Misodendraceae (two species) and Schoepfiaceae (four species) were used as outgroups. GenBank numbers for all species are shown in Table 1; 202 sequences are newly reported for this study. Alignments and phylogenetic trees are available through TreeBASE (http://treebase.org) study number S2095.
|
). Polymerase chain reactions (PCR), purification, and sequencing were performed as described in Vidal-Russell and Nickrent (2007). The chloroplast gene matK was amplified with 78f (Vidal-Russell and Nickrent, 2007) and 163f (5' AGG TTA CTA ATT GTG AAA CG 3') as forward primers and 1564r (5' ATG ATT RAC TAG ATC GTT GA 3') as the reverse primer. For the other chloroplast regions, published primers were used (trnL-F: Taberlet et al. [1991]; rbcL: Vidal-Russell and Nickrent [2008]). The nuclear small subunit ribosomal DNA (SSU rDNA) was amplified with primers reported in Vidal-Russell and Nickrent (2008) and approximately 2000 bp of the large subunit ribosomal DNA (LSU rDNA) were amplified in two pieces with the primers 27f (Vidal-Russell and Nickrent, 2008) and S5f (5' CGT GCA AAT CGT TCG TCT 3') as forward and S6r (5' CGC CAG TTC TGC TTA CCA 3') and 2134r (5' GGA CCA TCG CAA TGC TTT GT 3') as reverse.
Phylogenetic analysis
Sequences were aligned manually in the program Se-Al version 2.0a11 (Rambaut, 2004). With protein-coding genes, the translated amino acid sequences were used to aid alignment. Phylogenetic tree inferences for each gene individually and for the concatenated data set were obtained through maximum parsimony (MP), maximum likelihood (ML) and Bayesian inference (BI). In MP analyses, heuristic searches were performed with 1000 random addition sequences holding 10 trees for each iteration and by using the tree-bisection-reconnection (TBR) algorithm for branch swapping. A maximum limit of 100000 trees was imposed. A parsimony ratchet method (Nixon, 1999) was performed with 20 independent searches of 200 iterations and randomly weighting 25 characters per iteration for SSU rDNA using the programs PAUPRat version 1 (Sikes and Lewis, 2001) and PAUP* version 4.0b10 (Swofford, 2003).
Nodal support for MP was obtained through maximum parsimony bootstrap resampling (MPBS) run for 100 replications. For each bootstrap pseudoreplicate, heuristic searches were performed with 100 random addition sequences holding 10 trees for each iteration, with the TBR branch-swapping algorithm. When trees longer than the most parsimonious tree were found, only 10 were saved per replicate. For matK, rbcL, LSU rDNA, and the concatenated nuclear data set, analyses could not proceed because of the high number of best trees found during a particular replicate. In these cases, a time limit of 60 min per additional replicate was imposed.
For ML, the model of molecular evolution appropriate for each individual gene partition was selected using the program Modeltest version 3.6 (Posada and Crandall, 1998) using the hierarchical likelihood ratio test and the Akaike information criterion (Posada and Buckley, 2004). Frequently both methods selected the same model, but when they differed, the model with fewer parameters was used. Model selection can be found in Table 2. Heuristic tree searches were conducted starting with a neighbor joining tree, then by performing TBR branch swapping. Nodal support was obtained through bootstrap resampling (MLBS), with 100 replications using the program GARLI version 0.951 (Zwickl, 2006).
|
; Ronquist and Huelsenbeck, 2003) with models chosen for each partition by the program MrModeltest version 2.2 (Nylander, 2004) (Table 2). Two independent analyses were run with four chains each. The Markov chain Monte Carlo was set to run for five million generations, saving trees and parameters every 100th generation. The run was set to stop if topological convergence was reached between the two runs, which was determined by the presence of a standard deviation in split frequencies that was lower than 0.01 (discarding 25% as burn-in). Upon run completion, inspection of the likelihood scores vs. generation plots showed that these scores had always reached stationary before the first 25% of the samples; thus discarding this fraction as burn-in was conservative. Model parameters were estimated as part of the analysis. Uniform prior probabilities were assigned to all parameters except the state frequencies for which a Dirichlet prior distribution was assigned. When more than one partition was analyzed (i.e., the concatenated data set), parameter estimations were unlinked allowing partitions to evolve at different rates.
Morphological analysis
Flower characteristics associated with pollination syndromes (i.e., corolla color, flower symmetry, petal fusion) were coded for all genera. This matrix was used for parsimony character reconstruction on the consensus tree of the concatenated data set using the program Mesquite version 2.01 (Maddison and Maddison, 2007).
RESULTS
Statistics relating to the results of the various analyses of the separate and concatenated partitions are shown in Table 2. Although the level of resolution between the different gene partitions differed, there were no conflicting topologies that received high support. Among the five individual gene partitions, the lowest percentage of informative characters was from SSU rDNA followed by LSU rDNA, rbcL, trnL-F, and finally matK. Although the percentage of informative sites for the nuclear gene partition was less than half that of the chloroplast partition, the number of shortest trees recovered with MP was similar (384 and 336, respectively). Because the SSU rDNA partition gave very little resolution of relationships among genera, it is not shown. The combined SSU and LSU rDNA (nuclear) tree was less resolved than the LSU rDNA partition alone. For this reason, only the LSU rDNA partition tree is discussed next.
Nuclear LSU ribosomal gene partition
Parsimony analysis of the nuclear LSU partition did not recover Nuytsia as sister to the remaining genera but as one of four clades that arise from a polytomy along the spine of the tree (Fig. 1). The second clade of the polytomy is composed of Alepis and Peraxilla (clade A), and the third clade contains seven genera traditionally classified as subtribe Elytranthinae (clade B). The fourth clade (C) includes the remaining loranth genera but receives only moderate support and is not recovered using the chloroplast genes. Clade C is composed of three clades that arise from a polytomy, the first of which is Atkinsonia, the second Gaiadendron plus clade D, and the third the remaining loranths. Clade D contains two x = 12 genera (Tupeia [New Zealand] and Notanthera [South America]), Desmaria (South America) with x = 16, and the x = 8 clade (E) that includes seven genera. Although Tristerix and Ligaria do not occur with clade D and E taxa (as they do with the chloroplast genes), they resolve with little support in the next most derived position on the tree. In an unresolved position in clade F is clade G containing Loranthus and Cecarria. Clade I includes six Australasian and Indomalayan genera that have traditionally been considered part of tribe Lorantheae. The remaining loranths (clade J) include Asian taxa such as Scurrula, Taxillus, and Helixanthera as well as the remaining loranths that are mainly African.
|
|
|
|
This study represents the first multigene molecular phylogeny of Loranthaceae with robust sampling at the generic level. In comparison to similar intergeneric analyses within other families of Santalales, resolution of relationships among loranths proved to be particularly difficult, mainly owing to less phylogenetic signal in genes typically used at this level. For example, both rbcL and SSU rDNA sequences were useful for resolving relationships among genera in "Santalaceae" and Viscaceae (Der and Nickrent, 2008). This higher resolution is partly attributable to the general increase in substitution rates seen in more evolutionarily derived groups (i.e., Viscaceae). It appears that within loranths morphological and karyotypic character evolution has proceeded such that clear differences can be seen between many genera, but this differentiation is not marked by comparable changes in the nuclear and chloroplast genomes. A similar discordance was found for Argyroxiphium (Asteraceae) of Hawaii, which differs markedly in morphology from its tarweed ancestors in California but lacks concomitant genetic differentiation (Baldwin, 1997). The weak support for the relationships of the main aerial parasite clades (A+B, E, and F) and some monotypic genera (Desmaria, Notanthera, Tupeia) can probably be attributed to rapid radiations that occurred when this habit evolved in the Oligocene (Vidal-Russell and Nickrent, 2008).
Results obtained in this study are in general agreement with the earlier classification by Danser (1933), which was later modified by Barlow and Wiens (1971, 1973). However, the molecular phylogenetic trees presented here resulted in rearrangement of some genera, suggesting a need to change the circumscription of some tribes. Each clade will be discussed, including how it correlates with previous classifications and noting morphological features characteristic of that clade.
Nuytsia floribunda has always been recognized as different from other loranths and was classified in its own tribe by Danser (1933), a relationship supported by molecular data that placed it as sister to all other Loranthaceae (Figs. 2 and 3). This monospecific genus is a small tree of western Australia with a winged, wind-dispersed achene—a fruit type unique in the family. Moreover, Nuytsia nonspecifically parasitizes the roots of nearby plants using one of the most unusual haustorial types in Santalales. The haustorium contains sclerenchymatous prongs that act like scissors that transversely sever the host root (Beyer et al., 1989; Calladine et al., 2000). Some of its characteristics, like being phanerocotylar (cotyledons that spread during germination) and its base chromosome number of x = 12, resemble tribe Elytranthinae, while others, such as its unilocular ovary, resemble tribe Lorantheae.
The other root parasites, Atkinsonia from eastern Australia and Gaiadendron from Central and South America, are the only members of subtribe Gaiadendrinae in Tribe Elytrantheae (Danser, 1933). Atkinsonia is a small shrub, whereas Gaiadendron can be a shrub or an aerial parasite (Kuijt, 1963). Our data indicate that these two genera diverged early in the evolutionary history of the family after Nuytsia. The three root parasites do not form a clade, but they are successive sister taxa to the remaining members of the family. This topology suggests that aerial parasitism arose once in Loranthaceae, not four times as proposed by Wilson and Calvin (2006).
Tribe Elytrantheae is characterized by a plurilocular ovary and by fusion of the nucellus with the middle of the ovary, thus producing an enlarged mamelon. All genera are phanerocotylar and have a base chromosome number x = 12. This tribe has traditionally included two subtribes: Gaiadendrinae (just described) and Elytranthinae, the latter with 15 extant genera (Table 3), four of which were not sampled in this study. If the subtribe is circumscribed without Notanthera, Tristerix, and Tupeia (i.e., our clades A and B), it is strongly supported as monophyletic. The two New Zealand genera, Alepis and Peraxilla (clade A), long considered to be closely related (Barlow, 1966), are sister to the other clade B taxa. Within clade B, the sister relationship of Amylotheca and Loxanthera is well supported. These two genera, as well as Decaisnina have racemes of flowers in triads, whereas Lepidaria and Macrosolen have monads in spikes, racemes, and capitula. All genera in clades A and B have epicortical roots except Lysiana, which produces localized infections. The inflorescence of this genus is reduced to a two-flowered, axillary umbel. Another specialized feature of Lysiana is the high degree of fusion of the nucellus with the ovary wall where it is not possible to distinguish ovules from the mamelon (Cocucci and Venturelli, 1982). On the other hand, Lysiana and Peraxilla have an ovary with four loculi (Bhatnagar and Johri, 1983), which is considered a plesiomorphy. In other genera of this clade, the loculi of the ovary disappear (e.g., Macrosolen), or they are retained only in the basal portion of the ovary (e.g., Amylotheca).
|
; Bhatnagar and Johri, 1983). Previous classifications subdivided the tribe into two subtribes, Psittacanthinae and Loranthinae, with Psittacanthinae composed of all South American taxa plus Ligaria and Desmaria (Table 3). In this study, genera with x = 8 are recovered as monophyletic (clade E). Other South American genera (Tristerix, Notanthera, Ligaria, and Desmaria) together with the New Zealand genus Tupeia, are present in clade D, but the relationships are not well resolved (Fig. 3). Among all loranth clades, clade D has the greatest heterogeneity in base chromosome number: Tupeia, Tristerix, and Notanthera with x = 12, Ligaria with x = 10, Desmaria with x = 16, and the remaining South American genera with x = 8. Resolving the relationships and biogeography of this clade is crucial to an understanding of the history of the entire family.
Tupeia differs from the other New Zealand genera, Alepis and Peraxilla, in clade A, by lacking epicortical roots and by having a raceme of triads. However, the three genera share the x = 12 base chromosome number. In Tupeia the infection type is localized, and the endophyte grows within the host cortex, a type also found in Tristerix among clade D taxa (Kuijt, 1982; Calvin and Wilson, 2006).
Kuijt (1985) placed the following small-flowered (2–10 mm) genera of South and Central America in an assemblage separate from the other New World genera: Cladocolea, Dendropemon, Ixocactus, Panamanthus, Phthirusa, Oryctanthus, Oryctina, and Struthanthus. Oryctina, Ixocactus, and Panamanthus were not sampled in this study, but the remaining five genera of clade E appear in two well-supported clades that arise from a trichotomy, the third member of which is a clade containing the large-flowered (>30 mm) genera Aetanthus and Psittacanthus. Struthanthus is highly supported as sister to Cladocolea, a result that is not surprising given their morphological similarity. Kuijt (1981) placed them together in the Cladocolea-Struthanthus complex, differentiating these taxa by inflorescence features: Cladocolea with racemes of single flowers and Struthanthus with racemes of triads. The other small-flowered genera, Dendropemon, Oryctanthus, and Phthirusa, form a clade, which is in agreement with relationships proposed by Kuijt (1991). The first two genera are characterized by indeterminate racemes or spikes of bracteolate monads, while Phthirusa has an indeterminate raceme or spike of bracteolate triads (Kuijt, 1981).
The large-flowered genera Aetanthus and Psittacanthus form a well-supported clade in agreement with their similar inflorescence and floral morphology. Psittacanthus is a large genus with 119 species (Kuijt, 2008, in press) with various inflorescence types, all probably derived from a raceme of triads with pedicellate flowers (Kuijt, 1981). These three New World clades (i.e., the two small-flowered and the large-flowered clades) are highly supported as sister to Tripodanthus, which has medium sized (15–35 mm) white flowers arranged in a determinate raceme of triads. Given this topology, some hypotheses about floral evolution can be made. Among the clade D genera, several have reduced flowers that are likely insect-pollinated (Tupeia, Desmaria, Notanthera, Tripodanthus, and Notanthera). Thus, no matter how the Tristerix plus Ligaria clade resolves relative to this polytomy, it appears that traits related to insect pollination are plesiomorphic whereas tubular, bird-pollinated flowers evolved secondarily twice, i.e., in the ancestors of the Tristerix-Ligaria and Psittacanthus-Aetanthus clades.
Ileostylus and Muellerina (clade H) are sister to subtribe Loranthinae. These two genera have been recognized as a separate (but unnamed) tribe by Barlow and Wiens (1971), mainly because they have a base chromosome number of x = 11, unique for the family. Danser (1933), who included both genera in Loranthinae, indicated that Ileostylus was the "most primitive" in that subtribe. Ileostylus and Muellerina both are characterized by epicortical roots arising from the base of the plant and indeterminate inflorescences with racemes of triads. The four species of Muellerina are endemic to eastern Australia, whereas the single species, Ileostylus micranthus, is endemic to New Zealand. Clade H taxa represents the second of two apparently independent aneuploid reductions derived from x = 12 ancestors (the other being clade E). The third aneuploid reduction (to x = 9) is seen in all members of subtribe Loranthinae (Barlow and Wiens, 1971). Clade G containing Cecarria and Loranthus is highly supported; however, its position on the tree is not well resolved as it alternates between being sister to Ileostylus and Muellerina or sister to the remaining Loranthinae. Barlow and Wiens (1973) proposed that Cecarria represents a relict near the stem of the Old World line, a position not in disagreement with the phylogeny reported here. Cecarria has localized haustoria, but at least in some species of Loranthus, epicortical roots from the base of the plant can be formed. With the exception of these two genera, subtribe Loranthinae (clades I plus J) is highly supported as monophyletic.
The Amyema complex (Amyema, Dactyliophora, Benthamina, Diplatia, Sogerianthe) is well supported (clade I) and includes genera characterized by large inflorescences and flowers in triads; however, this general type is frequently reduced in various taxa (Danser, 1933; Kuijt, 1981). Most genera have epicortical roots but some Amyema species produce localized infections. Diplatia is the only genus in this clade that has endophytic growth (cortical strands). The two species of Amyema (A. glauca and A. queenslandica) sampled in this study did not form a clade, thus suggesting that the entire genus of 92 species is not monophyletic, a result in agreement with Calladine and Waycott (1999). Additional molecular work is required to establish affinities among all its components.
Within clade J, a clade with Dendrophthoe and Tolypanthus (from Indomalaya and Asia) are sister to another large clade (18 genera) found in Africa and Madagascar. These two genera have epicortical roots, as does Helixanthera coccinea. As a genus, Helixanthera is not monophyletic as evidenced by H. cylindrica being sister to the remaining taxa in clade J. This genus, with ca. 50 species, is widespread from Asia to Africa. As with Amyema, these results suggest that further work should be done to test the monophyly of the genus.
Only weakly supported as monophyletic, the African/Malagasy genera are characterized by their highly specialized bird-pollination mechanisms. Several genera have zygomorphic flowers with vents in the corolla tube (used by birds to open the flower) and coiled, explosive filaments. Their inflorescences are indeterminate with monads, and the flowers are five-merous with the petals fused into a tube. Most genera belong to one of two well supported clades, the exceptions being Plicosepalus, Socratina, Taxillus, and Vanwykia, which were described as "primitive" among African genera by Polhill and Wiens (1998) because they retained more pleisomorphic characters and because their pollination mechanism is less specialized. Except for Socratina, whose haustorial connection is unknown, the other three genera have epicortical roots that emerge only from the base of the plant; other African genera have single haustorial attachments. Taxillus is a genus with ca. 30 species distributed mainly in Asia with only one species reaching Africa. The accession sampled in this study was from Malaysia; thus no statements can be made about relationships with the African species. In this study Taxillus was sister to Scurrula, which agrees with their highly similar floral morphology and their mutual possession of epicortical roots. Vanwykia was described as closely related to the taxilloid genera (Taxillus, Socratina, Septulina, and Bakerella), but in this study it is strongly supported as sister to Socratina. The latter genus is endemic to Madagascar and Vanwykia is found in eastern and southeastern Africa; therefore a dispersal event to Madagascar from a common ancestor with Socratina is implied. Oedina, Agelanthus, and Oncocalyx belong to one of the African clades that are able to form cortical stands within the host, and the last two can form secondary shoots from this endophytic system. The other genera in this clade are characterized by localized haustorial connections. In the other African clade, only Moquiniella forms cortical strands and apparently lacks the capacity to form secondary shoots. The remaining five genera in this clade have single haustorial attachments.
Biogeographic implications
The molecular phylogeny presented herein has obvious implications for the biogeographic history of Loranthaceae, particularly when viewed in relation to the chronogram reported in Vidal-Russell and Nickrent (2008). Space does not permit discussion here; however, formal biogeographic analyses have been conducted (Vidal-Russell, 2007), and this topic will be fully explored in a future publication. The following represent general observations and interpretations, particularly those that relate to the roles played by pollinating and fruit-dispersing birds during the evolutionary history of Loranthaceae. Of the three basalmost loranth lineages, Atkinsonia is exclusively insect pollinated, and Nuytsia and Gaiadendron are visited by both insects and birds. All three of these genera have flowers whose morphology matches the reconstruction of the ancestral entomophilous type (i.e., open, choripetalous, and yellow). We argue here that these early (Cretaceous) loranths were pollinated exclusively by insects and that their floral morphology was shaped by this selectional environment. Present-day visitation by birds is likely a secondary event where opportunistic birds obtain nectar from these flowers. In the Old World, Loranthaceae are primarily pollinated and dispersed by oscine birds (order Passeriformes, suborder Passeri). These birds, in the families Dicaeidae and Nectariniidae, open the flowers with their bills by pinching the apex or by "unzipping" the flower along a corolla slit (Davidar, 1983; Feehan, 1985). The close association with pollinating birds appears to have driven selection for various floral traits. Some African genera such as Erianthemum, Actinanthella, and Oedina are highly specialized for bird-pollination. Their flowers change color at maturation, and their tubular corollas split along the petal junctions to form window-like fenestrae. This fenestration results from tension generated by differential growth of the stamens, which are fused to the petals below the fenestrae but free above (Kirkup, 1998). The pollinating sunbirds insert their beaks through the fenestrae, thus triggering rapid flower opening, inward coiling of the filaments, and deposition of pollen on the bird’s head. Simultaneously, the petals recurve and the style moves forward. Pollination and dispersal of Loranthaceae in Australia and New Zealand is performed by honeyeaters (Melphagidae), one of the first lineages to diverge within the oscines during the Eocene (Barker et al., 2004). Peraxilla from New Zealand has developed an explosive mechanism that is activated by the pollinating bird (Ladley et al., 1997).
In the New World, pollination functions are performed by both oscines and hummingbirds (order Apodiformes), whereas fruit dispersal is by suboscines (order Passeriformes, suborder Tyranni). Many hummingbird species visit loranth flowers (genera such as Aetanthus, Ligaria, Psittacanthus, and Tristerix) as do some species of Diglossa, nectar-robbing birds that are responsible for pollinating these flowers in the Andes (Graves, 1982; Amico et al., 2007). In addition, a large number of New World loranths are insect pollinated, particularly those in the x = 8 small-flowered clade.
Aerial parasitism in Loranthaceae is estimated to have evolved on Gondwana during the Oligocene, ca. 28 mya (Vidal-Russell and Nickrent, 2008) when Australia, Antarctica, and South America were still connected. The interaction between Melphagidae and Loranthaceae likely began at this time, fueling diversification and the subsequent dispersal of these aerial parasites to New Zealand. Two migrational waves, one from New Zealand and one from Australia, resulted in the spread and diversification of these loranth lineages throughout Australasia, Indomalaya, and eventually Africa. Migration of Asian genera in subtribe Loranthinae (e.g., Amyema, Dendrophthoe, and Benthamina) into Australia was probably effected by Dicaeum (presently the only species of mistletoe bird found in Australia), which is thought to have migrated into Australia in the Pliocene (Reid, 1988).
At present, the picture of biotic interactions among New World loranths and their pollinators is not clear, mainly owing to the paucity of fossils for hummingbirds, the oldest of which (30 mya) is from Europe (Mayr, 2004). Two groups within clade D (Fig. 3) are pollinated by birds: (1) Tristerix and Ligaria and (2) Psittacanthus and Aetanthus. Given the topology of the molecular tree, it is likely that these plants arrived at similar floral morphologies via independent interactions with these birds and that such interactions became established more recently than the situation with honeyeaters and loranths in Australia.
Upon arriving in Africa, the ancestral loranth (x = 9) underwent a massive adaptive radiation that generated over 20 genera of subtribe Loranthinae. This radiation can be linked to two sources: the development of the savannah habit during middle Miocene (Jacobs, 2004) and interaction with pollinating birds during the Oligocene (Barker et al., 2004; Beresford et al., 2005). Although mistletoes are found in tropical rainforests, they are particularly suited to more open habitats where full sun permits the active transpiration necessary to maintain more negative water potentials than their hosts. Floral features such as zygomorphy, bright and contrasting corolla colors, and explosive pollen dehiscence mechanisms evolved in response to interaction with sunbirds (Nectariniidae), which are responsible for pollinating many loranths in Africa. Sunbirds are sister to flowerpeckers (Ericson et al., 2003), and both families belong to a group of oscines that diversified and dispersed from Australia around 45 mya (Barker et al., 2004). These two bird families likely played an important role in the diversification and dispersal of Loranthaceae through Asia and into Africa during the Tertiary.
FOOTNOTES
1 The authors thank the numerous collectors listed in Table 1 who helped obtain specimens for this project. Thanks to MO for access to specimens, G. Amico for his useful comments, and S. Sipes for generously allowing use of her automated DNA sequencer. The authors appreciate the useful comments provided by two anonymous reviewers that improved this manuscript. Financial support was provided by a fellowship from the Fulbright Commission Argentina (R.V.-R.), the SIUC Graduate School and grants from the National Science Foundation (D.L.N.). 
3 Author for correspondence (e-mail: vidalr{at}crub.uncoma.edu.ar); present address: Laboratorio Ecotono, CRUB, Universidad Nacional del Comahue, Quintral 1250 (8400) Bariloche, Rio Negro Argentina
LITERATURE CITED
Amico, G., AND M. A. Aizen. 2000. Mistletoe seed dispersal by a marsupial. Nature 408: 929–930.
Amico, G. C., R. Vidal-Russell, AND D. L. Nickrent. 2007. Phylogenetic relationships and ecological speciation in the mistletoe Tristerix (Loranthaceae): The influence of pollinators, dispersers, and hosts. American Journal of Botany 94: 558–567.
Baldwin, B. G. 1997. Adaptive radiation of the Hawaiian silversword alliance: Congruence and conflict of phylogenetic evidence from molecular and non-molecular investigations. InT. J. Givnish, and K. J. Sytsma [eds.], Molecular evolution and adaptive radiation, 103–116. Cambridge University Press, Cambridge, UK.
Balle, S. 1954. Sur quelques Loranthoidées d’Afrique. Bulletin des Séances de l’Académie Royale des Sciences Colon Bruxelles 25: 1619–1635.
Barker, F. K., A. Cibois, P. Schikler, J. Feinstein, AND J. Cracraft. 2004. Phylogeny and diversification of the largest avian radiation. Proceedings of the National Academy of Sciences, USA 101: 11040–11045.
Barlow, B. A. 1966. A revision of the Loranthaceae of Australia and New Zealand. Australian Journal of Botany 14: 421–499.[CrossRef][Web of Science]
Barlow, B. A. 1974. A revision of the Loranthaceae of New Guinea and the south-western Pacific. Australian Journal of Botany 22: 531–621.[CrossRef][Web of Science]
Barlow, B. A., AND D. Wiens. 1971. The cytogeography of the loranthaceous mistletoes. Taxon 20: 291–312.[CrossRef]
Barlow, B. A., AND D. Wiens. 1973. The classification of the generic segregates of Phrygilanthus (= Notanthera) of the Loranthaceae. Brittonia 25: 26–39.[CrossRef][Web of Science]
Beresford, P., F. K. Barker, P. G. Ryan, AND T. M. Crowe. 2005. African endemics span the tree of songbirds (Passeri): Molecular systematics of several evolutionary ‘enigmas.’ Proceedings of the Royal Society of London, B, Biological Sciences 272: 849–858.
Beyer, C., W. Forstreuter, AND H. C. Weber. 1989. Anatomical studies of haustorium ontogeny and remarkable mode of penetration of the haustorium in Nuytsia floribunda (Labill.) R. Br. Botanica Acta 102: 229–235.[Web of Science]
Bhatnagar, S. P., AND B. M. Johri. 1983. Embryology of Loranthaceae. In D. M. Calder, and P. Bernhardt [eds.], The biology of mistletoes, 47–66. Academic Press, New York, New York, USA.
Cabrera, J. F.2002. Molecular phylogeny and historical biogeography of Loranthaceae (Santalales) inferred from matKsequence data. M.S. thesis, Department of Plant Biology, Southern Illinois University, Carbondale, Illinois, USA.
Calladine, A., J. S. Pate, AND K. W. Dixon. 2000. Haustorial development and growth benefit to seedlings of the root hemiparasitic tree Nuytsia floribunda (Labill.) R. Br. in association with various hosts. Annals of Botany 85: 733–740.
Calladine, A., AND M. Waycott. 1999. Evolution in the Australian loranthaceous mistletoes. In Proceedings of XVI International Botanical Congress, 1999, St. Louis, Missouri, USA, 409 (abstract no. 3558).
Calvin, C. L., AND A. C. Wilson. 2006. Comparative morphology of epicortical roots in Old and New World Loranthaceae with reference to root types, origin, patterns of longitudinal extension, and potential for clonal growth. Flora 201: 51–64.
Cocucci, A. E., AND M. Venturelli. 1982. El ovulo y el gineceo en Loranthaceae. Boletín de la Sociedad Argentina de Botánica 21: 131–141.
Danser, B. H. 1929. On the taxonomy and the nomenclature of the Loranthaceae of Asia and Australia. Bulletin du Jardin Botanique Buitenzorg 10: 291–374.
Danser, B. H. 1933. A new system for the genera of Loranthaceae-Loranthoideae, with a nomenclator for the Old World species of this subfamily. Verhandelingen der Koninklijke Akademie van Wetenschappen te Amsterdam Afdeeling Natuurkunde Sect. 2: 1–128.
Davidar, P. 1983. Similarity between flowers and fruits in some flowerpecker pollinated mistletoes. Biotropica 15: 32–37.[CrossRef][Web of Science]
Der, J. P., AND D. L. Nickrent. 2008. A molecular phylogeny of Santalaceae (Santalales). Systematic Botany 33: 107–116.[CrossRef][Web of Science]
Docters van Leeuwen, W. M. 1954. On the biology of some Javanese Loranthaceae and the role birds play in their life-history. Beufortia. Misc. Publ. 4: 105–207.
Engler, A. 1897. Loranthaceae. In A. Engler, and K. Prantl [eds.], Die Natürlichen Planzenfamilien, 124–140.
Ericson, P. G. P., M. Irestedt, AND U. S. Johansson. 2003. Evolution, biogeography, and patterns of diversification in passerine birds. Journal of Avian Biology 34: 3–15.[CrossRef][Web of Science]
Feehan, J. 1985. Explosive flower opening in ornithophily: A study of pollination mechanisms in some central African Loranthaceae. Botanical Journal of the Linnean Society 90: 129–144.[CrossRef][Web of Science]
Feuer, S., AND J. Kuijt. 1979. Fine structure of mistletoe pollen. II. Pollen evolution in the genus Psittacanthus Mart. Botaniska Notiser 132: 295–309.
Graves, G. R. 1982. Pollination of a Tristerix mistletoe (Loranthaceae) by Diglossa (Aves, Thraupidae). Biotropica 14: 316–317.[CrossRef][Web of Science]
Hamilton, S. G., AND B. A. Barlow. 1963. Studies in Australian Loranthaceae. II. Attachment structures and their interrelationships. Proceedings of the Linnean Society of New South Wales 88: 74–90.
Huelsenbeck, J., AND F. Ronquist. 2001. MrBayes. Bayesian inference of phylogeny. Bioinformatics (Oxford, England) 17: 754–755.[CrossRef]
Jacobs, B. F. 2004. Palaeobotanical studies from tropical Africa: Relevance to the evolution of forest, woodland and savannah biomes. Philosophical Transactions of the Royal Society of London, B. Biological Sciences 359: 1573–1583.[CrossRef][Medline]
Kirkup, D. 1998. Pollination mechanisms in African Loranthaceae. In R. Polhill, and D. Wiens [eds.], Mistletoes of Africa, 37–60. Royal Botanic Gardens, Kew, UK.
Kuijt, J. 1963. On the ecology and parasitism of the Costa Rican tree mistletoe, Gaiadendron punctatum (Ruiz and Pavón) G. Don. Canadian Journal of Botany 41: 927–938.[CrossRef]
Kuijt, J. 1969. The biology of parasitic flowering plants. University of California Press, Berkeley, California, USA.
Kuijt, J. 1981. Inflorescence morphology of the Loranthaceae—An evolutionary synthesis. Blumea 27: 1–73.
Kuijt, J. 1982. Epicortical roots and vegetative reproduction in Loranthaceae (s.s.) of the New World. Beitraege zur Biologie der Pflanzen 56: 307–316.
Kuijt, J. 1985. Morphology, biology and systematic relationships of Desmaria (Loranthaceae). Plant Systematics and Evolution 151: 121–130.[CrossRef][Web of Science]
Kuijt, J. 1988. Revision of Tristerix (Loranthaceae). Systematic Botany Monograph No. 19 American Society of Plant Taxonomists.
Kuijt, J. 1991. Panamanthus, a new monotypic genus of neotropical Loranthaceae. Annals of the Missouri Botanical Garden 78: 172–176.[CrossRef][Web of Science]
Kuijt, J. 2003. Two new South American species of Struthanthus (Loranthaceae) posing a challenge to circumscription of neotropical genera. Botanical Journal of the Linnean Society 142: 469–474.[CrossRef][Web of Science]
Kuijt, J. In press.Monograph of Psittacanthus (Loranthaceae). Systematic Botany Monographs.
Ladley, J. J., D. Kelly, AND A. W. Robertson. 1997. Explosive flowering, nectar production, breeding systems, and pollinators of New Zealand mistletoes (Loranthaceae). New Zealand Journal of Botany 35: 345–360.[Web of Science]
Liddy, J. 1983. Dispersal of Australian mistletoes: The cowiebank study. In D. M. Calder, and P. Bernhardt [eds.], 101–116, Academic Press, New York, New York, USA.
Maddison, W. P., AND D. R. Maddison. 2007.Mesquite: A modular system for evolutionary analysis, version 2.01. Website http://mesquiteproject.org.
Maheshwari, P., B. M. Johri, AND S. N. Dixit. 1957. The floral morphology and embryology of the Loranthoideae (Loranthaceae). Journal of Madras University B 27: 121–136.
Malécot, V., AND D. L. Nickrent. 2008. Molecular phylogenetic relationships of Olacaceae and related Santalales. Systematic Botany 33: 97–106.[CrossRef][Web of Science]
Mayr, G. 2004. Old world fossil record of modern-type hummingbirds. Science 304: 861–864.
Nickrent, D. L. 1994. From field to film: Rapid sequencing methods for field collected plant species. BioTechniques 16: 470–475.[Web of Science][Medline]
Nickrent, D. L. 2002. Mistletoe phylogenetics: Current relationships gained from analysis of DNA sequences. In P. Angwin [ed.], Proceedings of the 48th Western International Forest Disease Work Conference, Kona, Hawaii, USA, 48–57. USDA Forest Service, Redding, California, USA.
Nickrent, D. L., AND R. J. Duff. 1996. Molecular studies of parasitic plants using ribosomal RNA. In M. T. Moreno, J. I. Cubero, D. Berner, D. Joel, L. J. Musselman, and C. Parker [eds.], Advances in parasitic plant research, 28–52. Junta de Andalucía, Dirección General de Investigación Agraria, Córdoba, Spain.
Nickrent, D. L., R. J. Duff, A. E. Colwell, A. D. Wolfe, N. D. Young, K. E. Steiner, AND C. W. dePamphilis. 1998. Molecular phylogenetic and evolutionary studies of parasitic plants. In D. E. Soltis, P. S. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants II, DNA Sequencing, 211–241. Kluwer, Boston, Massachusetts, USA.
Nickrent, D. L., AND V. Malécot. 2001. A molecular phylogeny of Santalales. In A. Fer, P. Thalouarn, D. M. Joel, L. J. Musselman, C. Parker, and J. A. C. Verkleij [eds.], Proceedings of the 7th International Parasitic Weed Symposium, 69–74. Faculté des Sciences, Université de Nantes, Nantes, France.
Nixon, K. C. 1999. The parsimony ratchet, a new method for rapid parsimony analysis. Cladistics 15: 407–414.[CrossRef][Web of Science]
Nylander, J. A. 2004. MrModelTest2. Computer program distributed by the author. Evolutionary Biology Centre, Uppsala University, Uppsala.
Polhill, R., AND D. Wiens. 1998. Mistletoes of Africa. Royal Botanic Gardens Kew, Richmond-Surrey, UK.
Posada, D., AND T. R. Buckley. 2004. Model selection and model averaging in phylogenetics: Advantages of Akaike information criterion and Bayesian approaches over likelihood ratio tests. Systematic Biology 53: 793–808.
Posada, D., AND K. A. Crandall. 1998. MODELTEST: Testing the model of DNA substitution. Bioinformatics (Oxford, England) 14: 817–818.[CrossRef]
Rambaut, A. 2004. Se-Al Sequence Alignment Editor. Department of Zoology, University of Oxford, Oxford, UK.
Reid, N. 1988. The mistletoebird and Australian mistletoe: Coevolution or coincidence? Emu 87: 130–131.[Web of Science]
Reid, N. 1991. Coevolution of mistletoes and frugivorous birds. Australian Journal of Ecology 16: 457–469.[CrossRef][Web of Science]
Restrepo, C., S. Sargent, D. Levey, AND D. Watson. 2002. The role of vertebrates in the diversification of New World mistletoes. In D. Levey, W. Silva, and M. Galetti [eds.], Seed dispersal and frugivory: Ecology, evolution and conservation, 83–98. CABI Publishing, New York, New York, USA.
Ronquist, F., AND J. Huelsenbeck. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics (Oxford, England) 19: 1572–1574.[CrossRef]
Sikes, D. S., AND P. O. Lewis. 2001. PAUPRat: PAUP* implementation of the parsimony ratchet, Beta software, version 1. Distributed by the authors, Department of Ecology and Evolutionary Biology, University of Connecticut, Storrs. Website http://users.iab.uaf.edu/
derek_sikes/software2.htm.
Swofford, D. L. 2003. PAUP*: Phylogenetic analysis using parsimony (*and other methods). Sinauer, Sunderland, Massachusetts, USA.
Taberlet, P., L. Gielly, G. Pautou, AND J. Bouvet. 1991. Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105–1109.[CrossRef][Web of Science][Medline]
Tieghem, P. 1894. Sur le groupement des esèces en genres dans les Loranthacées a calice dialysépale et anthéres basifixes. Bulletin de la Societe Botanique de France 41: 497–511.
Vidal-Russell, R. 2007. The first aerial parasites in the sandalwood order (Santalales): Molecular phylogenetic and biogeographic investigations. Ph.D. dissertation, Southern Illinois University, Carbondale, Illinois, USA.
Vidal-Russell, R., AND D. L. Nickrent. 2007. A molecular phylogeny of the feathery mistletoe Misodendrum. Systematic Botany 32: 560–568.[CrossRef][Web of Science]
Vidal-Russell, R., AND D. L. Nickrent. 2008. The first mistletoes: Origins of aerial parasitism in Santalales. Molecular Phylogenetics and Evolution 47: 523–537.[CrossRef][Web of Science][Medline]
Watson, D. M. 2001. Mistletoe—A keystone resource in forests and woodlands worldwide. Annual Review of Ecology and Systematics 32: 219–249.[CrossRef][Web of Science]
Wilson, C. A., AND C. L. Calvin. 2006. An origin of aerial branch parasitism in the mistletoe family, Loranthaceae. American Journal of Botany 93: 787–796.
Zwickl, D. J. 2006. Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion. Website www.bio.utexas.edu/faculty/antisense/garli/Garli.html, University of Texas, Austin, Texas, USA [accessed 21 January 2007].
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  G. C. Amico and D. L. Nickrent Population structure and phylogeography of the mistletoes Tristerix corymbosus and T. aphyllus (Loranthaceae) using chloroplast DNA sequence variation Am. J. Botany, August 1, 2009; 96(8): 1571 - 1580. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
