| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
2Department of Plant Biology and Center for Systematic Biology, Southern Illinois University, Carbondale, Illinois 62901-6509 USA; 3Institute of Systematic Botany, University of Zurich, Zollikerstrasse 107, 8008 Zurich, Switzerland; 4Department of Biology, Morrill Science Center, University of Massachusetts, Amherst, Massachusetts 01003-5810 USA; 5Department of Botany, University of Florida, Gainesville, Florida 32611-8526 USA; 6Department of Biology, Washington State University, Pullman, Washington 99164-4236 USA
Received for publication March 12, 2002. Accepted for publication June 4, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: atp1 • atpB • Hydnora • Lactoridaceae • matR • Prosopanche • rbcL • ribosomal DNA
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Soltis et al., 2000
), basal members of this clade (Parkinson, Adams, and Palmer, 1999
; Qiu et al., 1999
, 2000
), or the eudicots (Hoot, Magallón, and Crane, 1999
). These data have been instrumental in reshaping current thought about higher-level relationships and indeed have provided the impetus to construct a revised classification of the flowering plants (APG, 1998
). In the APG classification, 27 families were listed as "position uncertain," and among these six were holoparasite families whose evolutionary relationships to other angiosperms have long been controversial: Balanophoraceae sensu lato (s.l.), Rafflesiaceae s.l., and Hydnoraceae. The phylogenetic position of the latter family within angiosperms has been the source of much disagreement. Cronquist (1988)
placed Hydnoraceae in Rafflesiales, an order thought to be related to Santalales of subclass Rosidae. In the system of Takhtajan (1997)
, the family was placed in its own order, allied with Rafflesiales in Rafflesianae of subclass Magnoliidae. The latter placement was more in accord with traditional systems that placed these orders with Aristolochiales. Molecular phylogenetic analyses of angiosperms have consistently documented the existence of a eudicot clade with tricolpate pollen and a grade composed of taxa with mainly monosulcate pollen (Chase et al., 1993
; Soltis et al., 2000
). The presence of monosulcate pollen in Hydnora represented an inconsistency in those classifications that placed Hydnoraceae among the eudicots (Cronquist, 1981
) as opposed to the magnoliids (Thorne, 1992
; Takhtajan, 1997
). In this paper we use the results of our molecular phylogenetic analyses to address this long-standing dilemma and explore the reasons for conflict among previous classifications.
Hydnoraceae contains only two genera: Hydnora, with approximately five species from Africa, the Arabian Peninsula, and Madagascar (Musselman and Visser, 1989
), and Prosopanche, with two species from South and Central America (Cocucci, 1965
). Hydnoraceae are quite distinctive; indeed Hydnora has been called the "strangest plant in the world" (Musselman and Visser, 1986
). This epithet is deserved given the highly modified vegetative and floral morphology of these plants. Hydnoraceae are the only angiosperms known that lack leaves (or modifications such as scales). Two types of roots exist in Hydnora: horizontal rhizome-like "pilot roots" that are hexagonal in cross section and vermiform outgrowths from the ridges of the pilot roots called "haustorial roots," whose function is to attach to the host (Fig. 1).
|
Vegetative and floral features of Prosopanche are unmistakenly similar to Hydnora; however, some differences exist. The androecium is fused into a dome-like structure, and staminodes are present as the second (innermost) whorl (Cocucci, 1975
). Only the upper flower parts (those above the ovary) emerge above the soil where they are then visited by a number of pollinators including small nitidulid beetles that are attracted by a pineapple-like smell. These beetles mate within the protogynous flower, which has a temperature elevated over ambient. The fruit begins development underground and with further development splits irregularly and circumscissally, exposing the pleasant-smelling endocarp. The fruit and numerous seeds are likely dispersed by nocturnal mammals (Cocucci and Cocucci, 1996
).
The biogeographic pattern of Hydnoraceae strongly suggests a Gondwanan distribution, although it might also be explained by long-distance dispersal. Given the holoparasitic nature of the plants and their host requirements, vicariance is a more likely explanation. Prosopanche americana frequently parasitizes legumes (principally Prosopis), whereas P. bonacinae occurs on host plants representing at least nine other families. Hydnora species occur mainly on hosts in Euphorbiaceae (H. africana and H. triceps) and Acacia (H. esculenta and H. johannis).
As discussed in Nickrent et al. (1998)
, holoparasitic plants present unique problems for those interested in inferring their phylogeny. As typically accompanies the holoparasitic habit, Hydnoraceae have highly modified or reduced vegetative and floral features. These modifications and losses prevent comparison to characters present in more conventional (photosynthetic) plants, thus impeding phylogenetic analysis based on morphology. A similar phenomenon may also occur at the molecular level where gene losses and rate accelerations have been documented for all three subcellular genomes in some holoparasites (Nickrent and Starr, 1994
; Nickrent et al., 1998
). When such divergent sequences are analyzed along with less divergent ones, long-branch artifacts (Felsenstein, 1978
) can occur, thus casting doubt upon the inferred phylogeny. Despite such difficulties, substitution rates among such holoparasites are heterogeneous, thus those groups with relatively lower rates are less likely to result in artifactual relationships. Relative rates tests involving Prosopanche were not significantly different (less than two standard errors) than comparisons using nonparasitic plants (Nickrent and Starr, 1994
). Preliminary analyses using nuclear small-subunit (18S) rDNA sequences for over 200 angiosperms placed Hydnora and Prosopanche near Aristolochiaceae (Nickrent and Duff, 1996
; Nickrent et al., 1998
), thus supporting traditional classifications. Because this analysis did not go to completion, and because only a single gene was used, we wished to confirm this result with additional (independent) data. We reasoned that if the results obtained from genes derived from separate subcellular compartments were congruent, it would be more likely that the actual organismal phylogeny is being detected.
We generated sequence data from nuclear (SSU and LSU rDNA) and mitochondrial (atp1 and matR) genes for Hydnora and Prosopanche. Two chloroplast genes (rbcL and atpB) were also included from photosynthetic angiosperms but were not included for Hydnoraceae because these genes appear to be absent in these plants. This was inferred from negative results of numerous polymerase chain reaction (PCR)-based experiments (D. L. Nickrent, personal observation). Three different data sets were constructed with differing taxon density and gene sampling. From these data sets, our major objective was to determine the position of Hydnoraceae within the global angiosperm phylogeny.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
, 2000
) and the three-gene analysis of Soltis et al. (2000)
. Large subunit rDNA sequences on nonparasites are reported in Zanis et al. (2002)
. Previously unpublished sequences include Hydnora africana large-subunit rDNA (GBAN-AF503353 and GBAN-AF503354), atp1 (GBAN-AF503356), matR (GBAN-AF503358), and Prosopanche americana large-subunit rDNA (GBAN-AF503355), atp1 (GBAN-AF503357) and matR (GBAN-AF503359). (The prefix GBAN- has been added to link the online version of the American Journal of Botany to GenBank but is not part of the actual accession number.) Placeholders were used in five instances such that no missing data were incorporated into the matrix other than rbcL and atpB for Hydnora and Prosopanche. The generic placeholders were Menispermum + Cissampelos (Menispermaceae), Brassica + Arabidopsis (Brassicaceae), Pisum + Vicia (Fabaceae), Placospermum + Persoonia (Proteaceae), and Oryza + Triticum (Poaceae). For the 77 taxon data set, 77 genera in 48 familes were included. Accession information including GenBank numbers have been archived on the American Journal of Botany Supplementary Data website, http://www.ajbsupp.botany.org/v89/.
Molecular and analytical methods
The protocols used here for extracting genomic DNA, PCR amplification, cloning, and sequencing have been reported (Qiu et al., 1993
; Nickrent, 1994
; Kuzoff et al., 1998
; Soltis et al., 2000
). Both manual and automated DNA sequencing methods were used. All alignments were conducted by eye. For chloroplast protein-coding genes (rbcL and atpB), alignment was straightforward given the lack of length variation, hence all positions were included. Alignment of the mitochondrial protein-coding genes (atp1 and matR) was guided by the use of translated sequences. For atp1, 19 sites in two regions and for matR 278 sites in 19 regions were removed owing to ambiguous alignment using the exclude characters command in PAUP* (Swofford, 2001
). In most cases the removed regions represented sequence that was autapomorphic for the outgroup taxa. For SSU rDNA, two regions (positions 190–199 and 1340–1342) were removed and for LSU rDNA 13 regions totalling 156 positions were removed. These alignments are available on the American Journal of Botany Supplementary Data website.
To address the question of phylogenetic position of Hydnoraceae within angiosperms, nuclear SSU rDNA sequences were obtained for Hydnora and Prosopanche and analyzed using three data sets differing in taxon density and genes. The three-gene data set included nuclear SSU rDNA (1792 sites), plastid rbcL (1403 sites), and atpB (1440 sites) for 461 taxa totaling 4635 sites. Hydnora and Prosopanche lack the plastid genes rbcL and atpB, hence these sites were coded as missing in these two taxa. The intention here is to stabilize the overall tree topology by adding these plastid genes. The position of Hydnoraceae will then be determined by the data that are present (nuclear rDNA). As compared with the next two, this data set has the broadest sampling within angiosperms, particularly eudicots.
The five-gene data set included nuclear SSU rDNA (1737 sites), plastid rbcL (1399 sites), atpB (1498 sites), atp1 (1285 sites), and matR (2307 sites) for 77 taxa totaling 8226 sites. The strategy here is to include more sequence data to further stabilize angiosperm relationships. Taxon density is lower than the three-gene data set; however, good representation of magnoliids and eudicots was achieved. This data set contained both nuclear and mitochondrial sequences for Hydnoraceae, thus more than one gene was influencing the position of the family in the analysis. The inclusion of mitochondrial data also allowed tests to be made of separate process partitions to determine whether the same signal is being received from gene sequences derived from distinct subcellular compartments.
The six-gene data set included nuclear SSU (18S) rDNA (1661 sites), LSU (26S) rDNA (3469 sites), plastid rbcL (1398 sites), atpB (1497 sites), atp1 (1284 sites), and matR (2189 sites) for 38 taxa totaling 11 528 sites. Taxon density is lower in this data set because many LSU rDNA sequences were not available (particularly the eudicots). This data set was used to examine the effect on tree topology with increased sequence data. It also allowed tests to be made to determine whether results from the two nuclear partitions are congruent.
All analyses of the above three data sets were conducted using PAUP* 4.0 (Swofford, 2001
). Maximum parsimony with heuristic searches, tree bisection-reconnection (TBR) branch swapping, and bootstrap analyses (100 replications) was used. For the 461-taxon three-gene analysis, five searches, each with 100 initial replications of NNI saving five trees per replication, were conducted. The shortest trees were then used as starting trees for subsequent searches with TBR. These TBR searches were allowed to run for 5–7 d. Previous congruence analyses as well as the similar tree topologies obtained for individual genes (reviewed in Soltis et al., 2000
) provide justification for combining the six genes used in this study into a single data set.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Three-gene analysis
Each of the five searches of the 461 taxon matrix resulted in more than 4000 trees, with tree lengths ranging from 37 026 to 37 051 (trees not presented here). Four searches placed Hydnoraceae within Aristolochiaceae s.l., and one placed it as sister to Saururaceae/Piperaceae.
Five-gene analysis
Heuristic searches of the five-gene data set recovered 24 shortest trees of length 10 951. These 24 trees occurred in one island found in 693 of 1000 random taxon addition replications. The single other island of six trees one step longer was hit 307 times. The strict consensus of these trees is nearly identical in topology to the bootstrap (BS) consensus, differing only in the position of Chloranthales (sister to monocots in the strict consensus, part of a polytomy in the BS tree). As shown by the BS consensus tree (Fig. 2), 54 of the 66 resolved nodes (five polytomies exist) received BS of 90% or greater. Forty-four percent and 40% of the informative characters derive from the chloroplast and the mitochondrial gene partitions, respectively (Table 1). The consistency index (CI) values are consistently higher for mitochondrial genes than for the nuclear or chloroplast genes. This trend (higher mitochondrial partition) can also be seen in the number of resolved nodes with BS values of 90% or greater. The first branching taxon is Amborella followed by Nymphaeales and a clade composed of Illiciales and Austrobaileya. Four clades comprising the eumagnoliids emerge next from a polytomy: (1) Chloranthales, (2) monocots plus Ceratophyllum, (3) eudicots, and (4) a clade composed of Laurales, Magnoliales, Winterales, and Piperales. Bootstrap support for the latter clade was 81% and was 100% for the eudicots. Piperales and Winterales were resolved as sister (84% BS) and within Piperales, two clades occur: Piperaceae plus Saururaceae and Aristolochiaceae s.l. The latter family includes genera traditionally placed in Aristolochiaceae (Aristolochia, Asarum, and Saruma) as well as Lactoris (Lactoridaceae) and Hydnora plus Prosopanche (Hydnoraceae). Relationships within Aristolochiaceae s.l. are not fully resolved in this analysis. Hydnora is strongly supported as sister to Prosopanche, as is Saruma with Asarum; however, relationships within Aristolochiaceae s.l. are poorly resolved.
|
|
). In contrast, distance analyses (using any of the substitution models available in PAUP* with neighbor-joining or unweighted pair group methods with arithmetic averaging) do not provide strong support for the inclusion of Hydnoraceae within Aristolochiaceae. Here, Hydnoraceae most frequently associate with Fabaceae and Brassicaceae, both of which exhibit long branches.
Six-gene analysis
Maximum parsimony heuristic searches of the six-gene data set resulted in four most-parsimonious trees of length 11 640. These four trees occurred in one island found in 850 of 1000 random taxon addition replications. Four other islands (seven trees) of longer trees were hit a total of 150 times. The strict consensus of the four shortest trees is nearly identical to the BS consensus tree that is shown in Fig. 3 (the main difference between them being the position of Chloranthaceae). Although taxon sampling is more limited compared with the 77-taxon data set, representatives of all the orders are present, and the topology of this tree is similar to the consensus of five-gene analysis. Of the 33 resolved nodes present on this tree, 21 received BS support of 90% or greater. The nodes that have lower BS support are clustered mainly within the eudicots and along the "spine" of the tree that represents the eumagnoliid clades. It is of interest that these nodes receive low BS support when any of the separate partitions are analyzed separately (trees not shown). Of the total 2389 informative characters, 43% derive from the nuclear rDNA partition, but nearly 5000 base pairs (bp) of sequence was collected to obtain these characters (Table 1). As a percentage of the partition length, the chloroplast genes have more informative characters than the other two partitions (25.5%). As with the 77-taxon data set, analysis of the mitochondrial partition results in more resolved nodes with BS values of 90% or greater (14 of 35). Also agreeing with the 77-taxon analysis, Piperales are strongly monophyletic (100% BS) and are sister to Winterales (but here with lower BS support). Within Piperales, Hydnoraceae still emerge from a paraphyletic Aristolochiaceae s.l. that also includes Lactoris. Support for this clade is high (94% BS), an increase apparently attributable to the inclusion of LSU rDNA data.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
. In that study, two clades of Aristolochiaceae sensu stricto (s.s.) were resolved at 100% BS: Aristolochia + Thottea and Asarum + Saruma clade (Hydnora or Prosopanche were not sampled). In 1 of the 18 shortest trees, Lactoris was shown as sister to the former clade, but with low BS support (58%). In a morphological cladistic study of basal angiosperms (Doyle and Endress, 2000
), the following characters were identified as unambiguously changing on the node giving rise to Aristolochioideae (Aristolochia), Asaroideae (Saruma, Asarum), and Lactoris: inflorescence a solitary flower (or occasionally with one or two additional flowers) and dry, indehiscent fruits. Aside from the cymose inflorescence, González and Rudall (2001)
pointed out the general lack of morphological features linking Lactoris to Aristolochiaceae. The features cited by Endress (1994)
supporting a close relationship between the families were considered symplesiomorphies for the entire order Piperales by these authors. The molecular data presented here, specifically the six-gene analysis, provide further indication that Aristolochiaceae s.l. are monophyletic and that Hydnora, Prosopanche, and Lactoris are components of this expanded family.
Affinities of Hydnoraceae: historical review and perspectives
First described as a fungus nearly 230 yr ago, Hydnora was discovered by Thunberg (1775)
, a student of Linnaeus. The family name (Hydnorinae) of Agardh (1821)
was conserved in the Montreal Code. The relationships between Hydnoraceae and Aristolochiaceae, particularly Thottea, were discussed by Meyer (1833)
. During the same approximate time period (1818), Rafflesia was discovered in Sumatra (Brown, 1822
), but an explicit relationship between Hydnoraceae and Rafflesiaceae was not suggested until 1844 (Brown, 1844
). During the 1860s, Prosopanche was first discovered in South America (De Bary, 1868
), and later additional species of Hydnora were described from Africa (Beccari, 1871
; Decaisne, 1873
). The association of Hydnoraceae with Rafflesiaceae likely stems mainly from the fact that both are parasites with flesh-colored flowers, although the number of shared morphological features is certainly limited. The association between Hydnoraceae, Rafflesiaceae (or Cytinaceae), and Aristolochiaceae continued through the latter part of the 19th century, and some authors, such as Baillon (1886)
, actually classified Hydnoraceae as a tribe of Aristolochiaceae. The treatment of Hydnoraceae in Pflanzenfamilien by Solms-Laubach (1894
, p. 285) considered the evidence favoring the two competing concepts as illustrated by the following translation:
What was previously said in the case of Rafflesiaceae is also true here. Most authors characterize Hydnora as a branch of Rafflesiaceae. Contradicting this is the construction of the androecium and of the fruit. In addition, there is the completely divergent structure of the seed, the presence of a perisperm, as well as the rough composition of the cellulose walls in the ‘nutrient tissue’ (Nährgewebe), as well as many other significantly divergent aspects. Based on the entire construction of the flower, one could definitely make a case for directly classifying it as Aristolochiaceae, even though this classification does not explain away the differences in the construction of the seed.
By the turn of the century and into the 1920s, additional species of Hydnora (Jumelle and Perrier de la Bâthie, 1912
) and Prosopanche (Chodat, 1915
) were described and anatomical studies were published for the ovules and seeds (Tieghem, 1897
; Dastur, 1921
). In a later version of Pflanzenfamilien, Harms (1935
, p. 288) wrote the following about familial relationships:
The family was usually attached to the Tribe Rafflesiaceae (for example of R. Brown) or Cytinaceae. However, the construction of the androecium and gynoecium contradicts this. Additional substantial differentiating features are the different construction of the seed, the presence of a perisperm and the deposit of the reserve materials in the cellulose walls (Solms Laubach). E. Meyer had thoroughly discussed the relationships to Aristolochiaceae (especially Thottea); and according to Solms-Laubach as well, the characteristics of the seed construction speak against such a classification, that Baillon, for example (among others), had posited.
It is curious that while paraphrasing Solms-Laubach, Harms focused upon the evidence against a relationship between Hydnoraceae and Aristolochiaceae, even though the original statement (above) appears to have considered this as a viable possibility.
A series of detailed studies of Prosopanche was published by Cocucci (1965
, 1975
, 1976
) that treated the taxonomy, floral morphology, and anatomy of the genus. A tendency toward fusion of the stamens and gynoecium as well as embryological features (such as the presence of unitegmic ovules) prompted Coccuci and Cocucci (1996)
to propose a relationship between Hydnoraceae and Mitrastemonaceae (Rafflesiales). This scheme showed Hydnora and then Prosopanche being derived from Mitrastemon, which was itself derived from Annonaceae. In parallel, Aristolochia (via Thottea) and Rhizanthes (via Pilostyles) were derived from an anonaceous ancestor. Although Aristolochiaceae were peripherally involved, the family was not proposed to be a close relative of Hydnoraceae.
Cronquist (1981)
placed Hydnoraceae in Rafflesiales and then allied this order with Santalales (Rosidae). Although he acknowledged that the family had traditionally been associated with Aristolochiaceae (based on perianth features), he considered the groups distinct and stated, "In my opinion, the Rafflesiales are singularly misplaced in the Aristolochiales." The presence of monosulcate pollen in Hydnora, however, presented a complication because nearly every other angiosperm with this condition was classified in Magnoliidae. Cronquist escaped this dilemma by proposing that the monosulcate pollen in Hydnora represented a reversion to a more primitive type, paralleling other simplification trends seen when plants adopt the parasitic habit. Molecular phylogenetic analyses using nuclear and mitochondrial genes do not suggest a close relationship between Hydnoraceae and Rafflesiaceae; indeed, the most recent data suggest that Rafflesiaceae are a clade in the eudicots (Nickrent, 2002
; A. Blarer et al., unpublished data).
The tendency to consider Hydnoraceae closely related to Rafflesiaceae also likely influenced the circumscription and scoring of character states in previous morphological cladistic analyses. Although both Aristolochiaceae and Hydnoraceae were included in the study by Dahlgren and Bremer (1985)
, the two did not form a clade. Instead, the latter was sister to Rafflesiaceae on a clade supported by seven characters. For one character (plants without chlorophyll), the states scored for the two holoparasites were not homologous and in others the range of character states present in the families was not adequately captured by the scoring. For example, pollen type for the two families was given as "monosulcate, inaperturate or ulcerate" despite the presence of colporate or porate pollen in Cytinus (Takhtajan et al., 1985
) and two- to three-porate pollen in Prosopanche (see Table 2). The presence of monosulcate pollen in Hydnora compelled Takhtajan (1997)
to follow a more traditional classification and place Hydnoraceae among the magnoliids. With reference to Hydnorales, he states, "Together with the next order, Rafflesiales are related to the Aristolochiaceae, especially to the Asaroideae, and have probably originated directly from their immediate ancestors." This placement is in agreement with molecular data with respect to Hydnoraceae, but again it has not escaped the pitfall of associating this family with Rafflesiaceae.
|
; Nickrent et al., 1998
) but bootstrap support is higher, particularly with the six-gene analysis. Here, Aristolochiaceae s.l. includes the shrub Lactoris (Lactoridaceae), Aristolochiaceae s.s., and Hydnoraceae. Given these results, the two clades will here be discussed Piperaceae + Saururaceae and Aristolochiaceae s.l. Despite the use of over 11 kb of sequence data derived from genes representing all three subcellular genomes, relationships within Aristolochiaceae s.l. are still unresolved. Thus, additional (molecular) data will be required to provide greater resolution of these relationships. Three possible tree topologies could emerge that would allow the three component families to remain monophyletic: (Hydnoraceae (Aristolochiaceae, Lactoridaceae)), (Aristolochiaceae (Hydnoraceae, Lactoridaceae)), or (Lactoridaceae (Hydnoraceae, Aristolochiaceae)). Given that it is unlikely that the monophyly of Hydnoraceae will be disrupted by any additional data, and given that Lactoris is monospecific, the other possible topologies that could result from further analysis would all involve paraphyly of Aristolochiaceae. Without additional data, it is premature to propose a reclassification of Aristolochiaceae s.l. based on phylogenetic principles.
The inclusion of Hydnoraceae in Aristolochiaceae s.l. adds yet another dimension to an already morphologically diverse order. Parasitism in Hydnoraceae represents one out of approximately ten independent evolutionary events that lead to this nutrition mode in flowering plants. The possible tree topologies discussed above gain additional interest when framed around the question of the origin of parasitism in Aristolochiales.
It is likely that overall flower morphology (flesh-colored tubular flowers) suggested to early workers an association between Hydnoraceae and Aristolochiaceae. There are a number of morphological features that are potential synapomorphies between these two families (Table 2). Characters that are compatible with both families (but not representing the entire range) are: entomophily; bisexual, epigynous, three-merous, monochlamydous, synsepalous, flowers; anther dehiscence extrorse and longitudinal with a secretory tapetum; pollen monosulcate; placentation parietal; endosperm development cellular; and embryo minute and undifferentiated. Ovary position varies in Aristolochiaceae, and it has been suggested that the epigynous condition in Asarum can be reversed (Kelly, 1997
). For pollen features, both monosulcate and inaperturate types occur in Aristlochiaceae and Lactoris, and pollen morphology is equally diverse in Hydnoraceae (Table 2). Moreover, during the course of evolution of holoparasitic angiosperms, morphological structures are often lost, reduced, or convergent. For these reasons, it is prudent to not engage in excessive speculation as to the homology of the morphological features that remain (e.g., epigynous perianth insertion, monochlamydous perianths, fusion of androecium and gynoecium, etc.) until more detailed analyses of the genetic basis for such characters are conducted.
Conclusions
Hydnora and Prosopanche comprise a clade that is resolved as a component of Aristolochiaceae s.l. (including Lactoridaceae). This clade is sister to another composed of Saururaceae and Piperaceae, together comprising Piperales. Separate analyses of nuclear and mitochondrial gene partitions result in generally congruent topologies for the resulting shortest parsimony trees. Combinations of gene sequences gave higher bootstrap support for the Aristolochiaceae s.l. plus Hydnoraceae. Although several morphological similarities can be found between Aristolochiaceae and Hydnoraceae, determining whether such shared character states are homologous is difficult owing to high levels of morphological variation within the families. This study highlights the use of molecular data in placing morphologically derived and therefore phylogenetically problematic plant families as has been done with other groups such as Podostemaceae (Soltis et al., 1999
), Hydrostachyaceae (Albach et al., 2001
), Aphloiaceae, and Ixerbaceae (Soltis et al., 2000
).
|
FOOTNOTES |
|---|
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Albach D. D. Soltis M. Chase P. Soltis 2001 Phylogenetic placement of the enigmatic angiosperm Hydrostachys. Taxon 50: 781-805[CrossRef][ISI]
APG. 1998 An ordinal classification for the families of flowering plants. Annals of the Missouri Botanical Garden 85: 531-553[CrossRef][ISI]
Baillon H. E. 1886 Hydnoraceae trib. Aristolochiacearum. Librairie Hachette, Paris, France
Beccari O. 1871 Descizione die due nuove specie d'Hydnora l'Abissinia. Nuovo Giornale Botanico Italiano 3: 6-7
Brown R. 1822 An account of a new genus of plants, named Rafflesia. Transactions of the Linnaean Society London 13: 201-234
Brown R. 1844 Description of the female flower and fruit of Rafflesia arnoldi, with remarks on its affinities; and an illustration of the structure of Hydnora africana. Transactions of the Linnaean Society of London 19: 221-238
Chase M. W. et al 1993 Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastid gene rbcL. Annals of the Missouri Botanical Garden 80: 528-580[CrossRef][ISI]
Chodat R. 1915 Les espèces du genre Prosopanche. Bulletin de la Société Botanique de Genève 7: 65-66
Cocucci A. E. 1965 Estudios en el género Prosopanche (Hydnoraceae) I. Revisión taxonómica. Kurtziana 2: 53-74
Cocucci A. E. 1975 Estudios en el género Prosopanche (Hydnoraceae) II. Organización de la flor. Kurtziana 8: 7-15
Cocucci A. E. 1976 Estudios en el género Prosopanche (Hydnoraceae) III. Embriología. Kurtziana 9: 19-39
Cocucci A. E. A. A. Cocucci 1996 Prosopanche (Hydnoraceae): somatic and reproductive structures, biology, systematics, phylogeny and potentialities as a parasitic weed. In M. T. Moreno, J. I. Cubero, D. Berner, D. Joel, L. J. Musselman, and C. Parker [eds.], Advances in parasitic plant research, 179–193. Junta de Andalucia, Dirección General de Investigación Agraria, Cordoba, Spain
Cronquist A. 1981 An integrated system of classification of flowering plants. Columbia University Press, New York, New York, USA
Cronquist A. 1988 The evolution and classification of flowering plants. New York Botanical Gardens, Bronx, New York, USA
Dahlgren R. K. Bremer 1985 Major clades of angiosperms. Cladistics 1: 349-368
Dastur R. H. 1921 Notes on the development of the ovule, embryosac, and embryo of Hydnora africana Thunb. Transactions of the Royal Society of South Africa 10: 27-31
De Bary A. 1868 Prosopanche burmeisteri, eine neue Hydnoree aus Süd-Amerika. Abhandlungen der Naturforschende Gesellschaft zu Halle 10: 243-272
Decaisne M. J. 1873 Note sur trois espèces d'Hydnora. Bulletin de Societie Botanique Francois 20: 75-77
Doyle J. A. P. K. Endress 2000 Morphological phylogenetic analysis of basal angiosperms: comparison and combination with molecular data. International Journal of Plant Sciences 161: (Supplement 6) S121-S153[CrossRef]
Endress P. K. 1994 Floral structure and evolution of primitive angiosperms: recent advances. Plant Systematics and Evolution 192: 79-97[CrossRef][ISI]
Felsenstein J. 1978 Cases in which parsimony or compatibility will be positively misleading. Systematic Zoology 27: 401-410[CrossRef][ISI]
González F. P. Rudall 2001 The questionable affinities of Lactoris: evidence from branching pattern, inflorescence morphology, and stipule development. American Journal of Botany 88: 2143-2150
Harms H. 1935 Rafflesiaceae, Hydnoraceae, and Balanophoraceae. In A. Engler and H. Harms [eds.], Die Natürlichen Planzenfamilien, 243–281, 282–295, 296–339. Duncker & Humblot, Leipzig, Germany
Hoot S. B. S. Magallón P. R. Crane 1999 Phylogeny of basal eudicots based on three molecular data sets: atpB, rbcL, and 18S nuclear ribosomal DNA sequences. Annals of the Missouri Botanical Garden 86: 1-32
Jumelle H. H. Perrier de la Bâthie 1912 Quelques phanérogames parasites de Madagascar. Revue Générale de Botanique 24: 321-328
Kelly L. M. 1997 A cladistic analysis of Asarum (Aristolochiaceae) and implications for the evolution of herkogamy. American Journal of Botany 84: 1752-1765[Abstract]
Kuzoff R. K. J. A. Sweere D. E. Soltis P. S. Soltis E. A. Zimmer 1998 The phylogenetic potential of entire 26S rDNA sequences in plants. Molecular Biology and Evolution 15: 251-263[Abstract]
Meyer E. H. F. 1833 De Hydnora. Nova Acta Physico-medica Academiae Caesareae Leopoldino Carolinae Naturae Curiosorum 16: 765.
Musselman L. J. J. H. Visser 1986 The strangest plant in the world!. Veld and Flora 71: 109-111
Musselman L. J. J. H. Visser 1989 Taxonomy and natural history of Hydnora (Hydnoraceae). Aliso 12: 317-326
Nickrent D. L. 1994 From field to film: rapid sequencing methods for field collected plant species. Biotechniques 16: 470-475[ISI][Medline]
Nickrent D. L. 2002 Orígenes filogenéticos de las plantas parásitas. In J. A. López-Sáez, P. Catalán, and L. Sáez [eds.], Plantas Parásitas de la Península Ibérica e Islas Baleares, 29–56. Mundi-Prensa Libros, S. A., Madrid, Spain
Nickrent D. L. 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 Andalucia, Dirección General de Investigación Agraria, Cordoba, Spain
Nickrent D. L. R. J. Duff A. E. Colwell A. D. Wolfe N. D. Young K. E. Steiner 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 Academic, Boston, Massachusetts, USA
Nickrent D. L. E. M. Starr 1994 High rates of nucleotide substitution in nuclear small-subunit (18S) rDNA from holoparasitic flowering plants. Journal of Molecular Evolution 39: 62-70[ISI][Medline]
Parkinson C. L. K. L. Adams J. D. Palmer 1999 Multigene analyses identify the three earliest lineages of extant flowering plants. Current Biology 9: 1485-1488[CrossRef][ISI][Medline]
Qiu Y.-L. M. W. Chase D. H. Les C. R. Parks 1993 Molecular phylogenetics of the Magnoliidae: cladistic analysis of nucleotide sequences of the plastid gene rbcL. Annals Missouri Botanical Garden 80: 587-606[CrossRef][ISI]
Qiu Y. Y. Lee F. Bernasconi-Quadroni D. E. Soltis P. S. Soltis M. Zanis E. A. Zimmer Z. Chen V. Savolainen M. W. Chase 1999 The earliest angiosperms: evidence from mitochondrial, plastid and nuclear genomes. Nature (London) 402: 404-409
Qiu Y.-L. J. Lee F. Bernasconi-Quadroni D. E. Soltis P. S. Soltis M. Zanis E. A. Zimmer Z. Chen V. Savolainen M. W. Chase 2000 Phylogeny of basal angiosperms: analyses of five genes from three genomes. International Journal of Plant Sciences 161: S3-S27[CrossRef][ISI]
Savolainen V. M. W. Chase Y.-L. Qiu 2000 Phylogenetics of flowering plants based on combined analysis of plastid atpB and rbcL gene sequences. Systematic Biology 49: 306-362[CrossRef][ISI][Medline]
Solms-Laubach H. 1894 Hydnoraceae. In A. Engler and K. Prantl [eds.], Die Natürlichen Planzenfamilien, Part III, 282–285. Wilhelm Engelmann, Leipzig, Germany
Soltis D. E. M. E. Mort P. S. Soltis C. Hibsch-Jetter E. A. Zimmer D. Morgan 1999 Phylogenetic relationships of the enigmatic angiosperm family Podostemaceae inferred from 18S rDNA and rbcL sequence data. Molecular Phylogenetics and Evolution 11: 261-272[CrossRef][ISI][Medline]
Soltis D. E. et al 2000 Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnaean Society 133: 381-461[CrossRef]
Swofford D. L. 2001 PAUP*: phylogenetic analysis using parsimony (* and other methods), version 4.0b9. Sinauer, Sunderland, Massachusetts, USA
Takhtajan A. 1997 Diversity and classification of flowering plants. Columbia University Press, New York, New York, USA
Takhtajan A. L. N. Meyer R. Kosenko V. N. Kosenko 1985 Pollen morphology and classification in Rafflesiaceae s. l. Botanichnyi Zhurnal 70: 153-162
Thorne R. F. 1992 An updated phylogenetic classification of the flowering plants. Aliso 13: 365-389
Thunberg C. P. 1775 Beskrifning paa en ganska besynnerlig och obekant svamp, Hydnora africana. Konglia Vetenskaps Akademiens Handlingar 36: 69-75
Tieghem P. V. 1897 Sur la structure de l'ovule et de la graine chez les Hydnoracees. Journal de Botanique 11: 233-238
Zanis M. D. Soltis P. Soltis S. Mathews M. Donoghue 2002 The root of angiosperms revisited. Proceedings of the National Academy of Sciences, USA 99: 6848-6853
This article has been cited by other articles:
|
|
 |
|
  P. S. Soltis, S. F. Brockington, M.-J. Yoo, A. Piedrahita, M. Latvis, M. J. Moore, A. S. Chanderbali, and D. E. Soltis Floral variation and floral genetics in basal angiosperms Am. J. Botany, January 1, 2009; 96(1): 110 - 128. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Arias and J. H. Williams Embryology of Manekia naranjoana (Piperaceae) and the origin of tetrasporic, 16-nucleate female gametophytes in Piperales Am. J. Botany, March 1, 2008; 95(3): 272 - 285. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Y. Smith and R. A. Stockey Establishing a fossil record for the perianthless Piperales: Saururus tuckerae sp. nov. (Saururaceae) from the Middle Eocene Princeton Chert Am. J. Botany, October 1, 2007; 94(10): 1642 - 1657. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. U. Tennakoon, J. F. Bolin, L. J. Musselman, and E. Maass Structural attributes of the hypogeous holoparasite Hydnora triceps Drege & Meyer (Hydnoraceae) Am. J. Botany, September 1, 2007; 94(9): 1439 - 1449. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. S. Soltis and D. E. Soltis The origin and diversification of angiosperms Am. J. Botany, October 1, 2004; 91(10): 1614 - 1626. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Williams and W. E. Friedman The four-celled female gametophyte of Illicium (Illiciaceae; Austrobaileyales): implications for understanding the origin and early evolution of monocots, eumagnoliids,and eudicots Am. J. Botany, March 1, 2004; 91(3): 332 - 351. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. C. Davis and M. W. Chase Elatinaceae are sister to Malpighiaceae; Peridiscaceae belong to Saxifragales Am. J. Botany, February 1, 2004; 91(2): 262 - 273. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. J. Barkman, S.-H. Lim, K. M. Salleh, and J. Nais From the Cover: Mitochondrial DNA sequences reveal the photosynthetic relatives of Rafflesia, the world's largest flower PNAS, January 20, 2004; 101(3): 787 - 792. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Fan and Q.-Y. Xiang Phylogenetic analyses of Cornales based on 26S rRNA and combined 26S rDNA-MATK-RBCL sequence data Am. J. Botany, September 1, 2003; 90(9): 1357 - 1372. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
