| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
2Division of Botany and Zoology, Australian National University, Canberra ACT 0200, Australia 3CSIRO, Plant Industry, Centre for Plant Biodiversity Research, Australian National Herbarium, GPO Box 1600, Canberra ACT 2601, Australia
Received for publication October 20, 2000. Accepted for publication February 13, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: external transcribed spacer • trnK intron • matK, psbA-trnH spacer • Gastrolobium • Fabaceae • molecular phylogenetics; • Mirbelieae
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
60 species. This genus is endemic to the southwest of Western Australia, except two species that are widespread throughout northern Australia (G. brevipes and G. grandiflorum; Fig. 1). Furthermore, it is one of the largest legume genera in the southwest of Western Australia, where it forms a major component of the understory in many areas, such as sandplains with their accompanying vegetation ("Kwongan"), which is usually heath or mallee (shrubby eucalypt woodland).
|
). Members of Gastrolobium accumulate sodium monofluoroacetate (e.g., Aplin, 1971
; Twigg et al., 1996), which makes them highly toxic, and severe stock losses have occurred in the past due to fluoroacetate poisoning, which led to an eradication program, particularly in the wheat-belt region of southwestern Western Australia. As a consequence, many species are now rare or threatened with extinction, making Gastrolobium both ecologically and economically important.
Taxonomic history of Gastrolobium
Throughout its taxonomic history, the circumscriptions of Gastrolobium and its allied genera, particularly Oxylobium Jackson, have changed considerably. As a result, species have been transferred from one genus to another on several occasions. A major component of the problem of the circumscription of Gastrolobium is due to the fact that morphological data have, to date, failed to fully resolve the relationships within the tribes Mirbelieae and Bossiaeeae (Crisp and Weston, 1987, 1995
), especially the Gastrolobium/Oxylobium generic group.
Gastrolobium was described by Brown (1811)
as a monotypic genus, diagnosed by a stipitate ovary with two ovules, which distinguished it from Oxylobium (below), though Brown (1811)
did not mention this fact explicitly. Species were added to Gastrolobium over time by various authors, and Bentham (1864)
provided the first revision of this group in Flora Australiensis. Again, it was primarily ovule number that separated Gastrolobium from Oxylobium, with Gastrolobium having two ovules and Oxylobium four or more ovules (Bentham, 1864
). Both genera contained species that produced fluoroacetate, and Oxylobium contained species from both eastern and western Australia (Fig. 1).
Kuntze (1891)
subsumed Oxylobium and Gastrolobium into the earlier genus Callistachys Vent. However, Oxylobium was later conserved against Callistachys. Nemcia was described by Domin (1923)
, including 12 species characterized by 4–6 ovules, trifid bracts, and condensed racemose inflorescences. This work was largely ignored, and the concepts of Gastrolobium and Oxylobium remained as they had been since Bentham (1864)
.
Gardner and Bennetts (1956)
provided a revision of the toxic plants of Western Australia, which included a number of species of Gastrolobium and Oxylobium. However, this was not a complete revision of the group, because it did not include the nontoxic species of either genus. Furthermore, these toxic species were interleaved in this artificial key, the authors apparently being unable to distinguish easily between the two genera, and again the concept of Bentham (1864)
was used as the division between Gastrolobium and Oxylobium, relying on ovule number as the main character to separate the two genera.
Crisp and Weston (1987)
published the first major review of Gastrolobium since Bentham (1864)
. They presented a phylogeny of the tribe Mirbelieae based on morphology and reinstated and expanded both Nemcia and Podolobium F.Muell., the latter of which is an eastern Australian genus closely aligned with Oxylobium. Gastrolobium fell into the ‘Callistachys' group, which consisted of Brachysema R.Br., Callistachys, Jansonia Kipp., Gastrolobium, Nemcia, Podolobium, and Oxylobium lineare. The analysis of Crisp and Weston (1987)
, however, was done at a higher level to resolve tribal relationships within the Mirbelieae, using either genera or subgeneric groups as terminals.
Crisp and Weston (1987)
changed the circumscription of Gastrolobium to include all toxic species of Gastrolobium and Oxylobium, so that for the first time, species with more than two ovules were included within Gastrolobium. This left only one species of Oxylobium occurring in Western Australia (O. lineare), which required further work to determine its generic affinities. Their reduced concept of Oxylobium comprises five species endemic to eastern Australia, mostly along the central and southern coastal plain and the adjacent escarpment of the Great Dividing Range (Fig. 1). The nontoxic species of Gastrolobium and Oxylobium were removed into Nemcia (but see Twigg et al., 1996).
Nemcia, as defined by Crisp and Weston (1987)
, contained species with axillary racemes often reduced to one or two flowers (though some had condensed, terminal racemes with many flowers), and included the nontoxic species transferred from Gastrolobium and Oxylobium (see Aplin, 1971
), thereby using secondary metabolites as an aid in the resolution of this taxonomically difficult group.
Genera such as Brachysema, Jansonia, and Leptosema Benth. were distinguished by floral characteristics that have been interpreted by later authors as indicative of bird pollination (e.g., Keighery, 1982
). These characters include red petals, a reduced standard petal and enlarged keel petals, and copious nectar. Gastrolobium and Oxylobium are primarily bee pollinated, except G. grandiflorum, which has large, red flowers, but lacks the "bird–flower" modifications of genera such as Brachysema, like a reduced standard petal. However, most of the assumptions of bee- or bird-pollination are largely inference based on floral structure, which often came from empirical data, such as sightings of birds visiting flowers (e.g., Keighery, 1980, 1982, 1984
).
The evolution of bird-pollination in some Australian legumes was discussed by Crisp (1994)
, where a phylogeny of Brachysema, Jansonia, Nemcia, and Oxylobium lineare was presented, but did not include Gastrolobium. Crisp (1994)
also tested the monophyly of these genera with a species-level phylogeny using morphology. Even though this analysis did not include Gastrolobium, Nemcia was shown to be paraphyletic, while Brachysema was demonstrated to be monophyletic.
Crisp, Gilmore, and Van Wyk (in press) provide a molecular phylogeny of the genistoid legume tribes, though only two species of the ‘Callistachys' group are used in this tribal phylogeny, so nothing can be deduced about the relationships within this group. A sound, well-resolved phylogeny of Gastrolobium and its close relatives is therefore required in order to resolve the taxonomic dilemmas surrounding this group and bring stability to these genera.
Molecular phylogeny of Gastrolobium and related genera
As morphological data alone have been insufficient in resolving the relationships of the Gastrolobium group, molecular data were the obvious choice to try to find a robust phylogeny on which to base future taxonomic classifications. This study has used the psbA-trnH intergenic spacer, the trnK 5' intron (both from chloroplast DNA), and the 3' end of the external transcribed spacer (ETS, from nuclear ribosomal DNA) in an attempt to resolve the relationships of Gastrolobium and its close relatives.
The psbA-trnH intergenic spacer region lies in the inverted repeat region of the chloroplast genome, near the boundary with the large single-copy region, adjacent to the trnK gene (Sugiura, 1992
). The psbA chloroplast gene belongs to the Photosystem II (PSII) protein complex and codes for the PSII D1-protein; the trnHHis (GUG) gene belongs to the transfer RNA gene system and transfers for the amino acid histidine. Phylogenetic studies reported this spacer to be of more use at higher taxonomic levels, particularly intergeneric levels (for example, Aldrich et al., 1988
; Sang, Crawford, and Stuessy, 1997
; Asmussen and Liston, 1998
; Kim et al., 1999
), though Kim et al. (1999)
did find it somewhat useful at the infrageneric level.
The trnK intron, which includes the matK coding region, has been used to reconstruct phylogenies in a number of different families, such as the Apiaceae (Plunkett, Soltis, and Soltis, 1996, 1997
), Cornaceae (Xiang, Soltis, and Soltis, 1998
), Cupressaceae (Gadek et al., 2000
), Fabaceae (Hu et al., 2000
), Juglandaceae (Stanford, Harden, and Parks, 2000
), Nymphaeaceae (Les et al., 1999
), Orchidaceae (Jarrell and Clegg, 1995
), Pinaceae (Wang et al., 1999
), Poaceae (Liang and Hilu, 1996
; Hilu and Liang, 1997
; Hilu and Alice, 1999
), Polemoniaceae (Steele and Vilgalys, 1994
; Johnson and Soltis, 1995
; Johnson et al., 1996
), and Saxifragaceae (Johnson and Soltis, 1994, 1995
). Hilu and Liang (1997)
evaluate the rate, patterns, and types of nucleotide substitutions in the matK gene, functional constraints, and phylogenetic utility of the gene, using data from a number of different plant families, and report that the 5' end of the trnK intron is larger and contains more informative characters than the 3' end. Accordingly, the 5' section of the trnK intron was selected for use in this study.
The external transcribed spacer has recently been shown to be much larger and contain more phylogenetically informative characters than the internal transcribed spacer (ITS), providing a large number of characters for use in phylogenetic analyses (Baldwin and Markos, 1998
; Bena et al., 1998
). To date, the ETS has been used mainly to study phylogeny within the Asteraceae (Baldwin and Markos, 1998
; Linder et al., 2000
), and the Fabaceae (Bena et al., 1998
). However, sequences have been generated from the ETS region in a number of families, sometimes covering the entire intergenic spacer (IGS) of ribosomal DNA, including the Asteraceae (Baldwin and Markos, 1998
; Linder et al., 2000
), Brassicaceae (Rathgeber and Capesius, 1990
), Cucurbitaceae (King et al., 1993
), Fabaceae (Rogers and Bendich, 1987a, b
; Schiebel et al., 1989
; Bena et al., 1998
), and Solanaceae (Schmidt-Puchta, Gunther, and Sänger, 1989
; Borisjuk et al., 1994
; Volkov et al., 1996
).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Table 1 shows the 94 taxa used in this analysis including their authorities, along with GenBank accession numbers for the sequences obtained. Vouchers of all specimens used in the analysis are deposited at the Australian National Herbarium (CANB), and the collector name and number for each accession are provided in Table 1, along with a brief locality description.
|
, following the work of Crisp and Weston (1995)
, and sampled throughout the 5-nucleate embryo sac clade. Isotropis cuneifolia was used to root the tree, as this genus occurs at the base of the 5-nucleate embryo sac group. Other outgroups used were Jacksonia horrida, Latrobea hirtella, Mirbelia depressa, M. dilatata, Phyllota phylicoides, Pultenaea dentata, and P. reticulata. Outgroup genera that appeared closely related to Gastrolobium were sampled more extensively, including Callistachys (1/1 species), Oxylobium (5/6 species), and Podolobium (6/6 species).
Ingroup sampling
A pilot study suggested that Brachysema, Jansonia, and Nemcia were nesting within Gastrolobium, so these genera were sampled more extensively than originally planned. The study included 9/10 species of Brachysema, 1/1 species of Jansonia, and 16/39 species of Nemcia, and one undescribed species (all undescribed taxa are marked in Table 1 with ‘ms’). Species of Nemcia were added to the sample as it became clear that this genus is polyphyletic, and were chosen to represent the diversity of this group. Within Gastrolobium s.s. (sensu stricto), 48/60 species were sampled, including 13 undescribed species. Of the 12 species of Gastrolobium not sampled, six were unavailable recent discoveries. For the other six, fresh or CTAB-preserved material was unavailable and herbarium material of these failed to amplify. It was felt that the final sample size was sufficient to test the monophyly (or nonmonophyly) of each group and to resolve relationships within them.
DNA isolation, amplification, and sequencing
Total DNA was isolated as outlined in Bayer, Hufford, and Soltis (1996)
. Methods outlined in Gilmore, Weston, and Thompson (1993)
were used to isolate DNA from herbarium tissue and to purify recalcitrant DNAs. When these methods failed, DNAs were run through a QIAquickTM PCR Purification Kit (QIAGEN, Hilden, Germany).
All three regions were amplified by the polymerase chain reaction (PCR) using Taq DNA polymerase and the following conditions. The PCR samples were heated to 94°C for 3 min prior to the addition of DNA polymerase to denature unwanted proteases and nucleases. The double-stranded PCR products were produced via 30 cycles of denaturation (94°C for 1 min), primer annealing (48°C for 1 min), and extension (72°C for 1 min). A 7-min final extension cycle at 72°C followed the 30th cycle to ensure the completion of all novel strands. See Table 2 for all primer sequences and references. Double-stranded PCR products were cleaned using QIAquickTM PCR Purification Kits (QIAGEN, Hilden, Germany) prior to sequencing.
|
trnK 5' intron sequences
The PCR reaction mixture consisted of 35 µL of sterile water, 5 µL of 10x reaction buffer, 3 µL of 25 mmol/L magnesium chloride solution, 2 µL of 40 mmol dNTP solution in equimolar ratio, 10 pmol of each primer (3914f and 1110R), 5–25 ng of template DNA, and 0.25 µL of Taq polymerase in a total volume of 50 µL. (The 1110r primer was designed by R. Bayer for use in the Asteraceae, which worked well in the legumes in this study. The 1110r primer lies 1110bp from the 3' end of the trnK gene in the tobacco chloroplast genome and within the matK coding region.) Some taxa required the use of four primers (3914f and Gast12; Gast11 and 1110r) to amplify this region, particularly when herbarium material was used. Gast11 and Gast12 were designed in a conserved part of the trnK intron and provide overlapping sequences (Table 2).
External transcribed spacer sequences
Specific primers were developed by initially using a long-range PCR amplification of the entire intergenic spacer region between the 18S and 26S subunits of rDNA, using the universal primers of Baldwin and Markos (1998)
. The 18S-IGS primer was then used to sequence the 3' end of the ETS region. The 5' end of this region yielded a conservative site suitable for the design of another primer, Gast1 (Table 2), which allowed the amplification of 350 bp of sequence. The PCR reaction mixture consisted of 70 µL of sterile water, 10 µL of 10x reaction buffer, 6 µL of 25 mmol/L magnesium chloride solution, 4 µL of 40 mmol dNTP solution in equimolar ratio, 20 pmol of each primer (18SIGS and Gast1), 10–50 ng of template DNA, and 0.5 µL of Taq polymerase in a total volume of 100 µL.
Sequencing of PCR products
The double-stranded PCR products were used as templates in cycle sequencing reactions, which employed the same primers that were used for PCR amplification to sequence both strands. The double-stranded PCR products were sequenced using the dideoxy chain termination method (Sanger, Nicklen, and Coulson, 1977
) with the use of the Big Dye Terminator RR Kit® (Perkin Elmer Applied Biosystems, Norwalk, Connecticut, USA) at CSIRO, Plant Industry. An annealing temperature of 60°C was used for both primers. The cycle sequencing protocol followed manufacturer's instructions.
Sequences were assembled using SequencherTM 3.0 (Gene Codes Corporation, Ann Arbor, Michigan, USA), then manually adjusted following the principles of noncoding sequence alignment using secondary structure (Kelchner and Clark, 1997). Indels were placed where they minimized the number of inferred length mutations, unless clear evidence was seen for nonhomologous length mutation events. Unambiguous indels have been coded as additional characters (Golenburg et al., 1993
).
Sequence data analysis
Sequence data were analyzed using parsimony as implemented in PAUP 4.0b3a (Swofford, 1997
) on a Macintosh G3 computer. The data matrix contained 75 ingroup taxa, taken from the ‘Callistachys' group, and 19 outgroup taxa. Phylogenetic reconstruction was performed on unweighted characters by heuristic searches with simple addition of taxa. An island search was employed to search for further most parsimonious trees, with a random addition sequence of 100 replicates using a heuristic search (Maddison, 1991
). The three sets of sequences were analyzed individually and together. A partition homogeneity test was conducted to test the compatibility of the three data sets.
The robustness of clades was tested using two methods: bootstrapping (Felsenstein, 1985
) and decay analysis (Bremer, 1988
). One thousand replicates were used for the bootstrap. The decay analysis was facilitated by the program AutoDecay (Eriksson, 1998
). The decay values were then extracted using AutoDecay and visualized using the tree-drawing package, TreeView (Page, 1996
).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
In the combined matrix, sequence divergence varies from 0.435% between Nemcia hookeri and N. obovata to 21.451% between Pultenaea reticulata and Isotropis cuneifolia; in the psbA-trnH spacer it varies from 0% between Podolobium alpestre and P. procumbens to 21.364% between Gastrolobium parviflorum and Isotropis cuneifolia; in the trnK 5' intron it ranges from 0% between Gastrolobium heterophyllum and G. nutans to 20.2% between Isotropis cuneifolia and Nemcia sp. nov. A; and in the ETS it ranges from 0% between several sets of taxa (N. leakeana/N. luteifolia/N. rubra; N. coriacea/N. hookeri/N. obovata/N. plicata; Gastrolobium stowardii/G. carinatum ms; G. revolutum/G. tetragonophyllum/G. parviflorum; G. floribundum/G. propinquum; and G. appressum/G. oxylobioides) to 28.444% between Callistachys lanceolata and Isotropis cuneifolia.
The number of unambiguous indels in each sequence varies considerably, with numerous indels present in the psbA-trnH spacer (ranging in size from 2 to 241 bp) to very few in the ETS (all of which were autapomorphic and, therefore, phylogenetically uninformative). Only the numbers of unambiguously coded indels are given in Table 3, which range in size from 2 to 241 bp in the psbA-trnH spacer, 5 to 20 bp in the trnK 5' intron, and none in the ETS.
Phylogenetic reconstruction
A heuristic search of all potentially phylogenetically informative nucleotide characters from the total combined data matrix, including indels, revealed 360 trees of 2327 steps, with confidence index (CI) = 0.404, retention index (RI) = 0.631. A 50% majority-rule tree also shows the decay and bootstrap values calculated for each clade (Fig. 2). Only five branches in the majority-rule tree collapse in the strict consensus. These are shown with dashed lines (Fig. 2). A phylogram shows the number of synapomorphies supporting each branch (Fig. 3) in one of the equally most parsimonious trees. The partition homogeneity test indicated the data sets are not significantly different (P = 0.08) and can therefore be combined into one analysis.
|
|
Clade A (decay [D] = 3, synapomorphies [SYN] = 16)
The ‘Oxylobium’ group contains Oxylobium (excluding O. lineare), Mirbelia, Callistachys, and three species of Podolobium (P. alpestre, P. procumbens, and P. scandens). Oxylobium and Podolobium both occur in eastern Australia (Fig. 1), Mirbelia occurs in both eastern and western Australia, and Callistachys is endemic to the southwest of Western Australia.
Clade B (D = 27, SYN = 45)
The Podolobium ilicifolium group contains three species of Podolobium, P. aciculiferum, P. aestivum, and P. ilicifolium. These species all have prickly leaves and recurved calyces, and occur on the east coast and associated escarpment of Australia (Fig. 1).
Clade C (D = 19, SYN = 30)
The ‘Gastrolobium’ group contains all species of Gastrolobium, as well as the genera Brachysema, Jansonia, and Nemcia, plus Oxylobium lineare, a doubtful species of Oxylobium that Crisp and Weston (1995)
made clear belongs in another genus and the only one occurring in Western Australia.
Clade D (D = 1, SYN = 11)
This clade contains a number of species, including some that form smaller clades, including the Gastrolobium spinosum group (D = 13, SYN = 23; G. spinosum, G. triangulare, and G. trilobum); the G. bilobum/G. parviflorum group (D = 4, SYN = 14; G. parviflorum, G. ‘revolutum’ ms, G. ‘stenocarpum’ ms, G. tetragonophyllum, G. bilobum, G. congestum ms, G. grandiflorum, and G. tergiversum ms). There are also a number of species that form a grade at the base of Clade D and occur only on or on the margins around granite outcrops, including G. acrocaroli ms, G. callistachys, G. graniticum, G. involutum ms, G. semiteres ms, and G. stenophyllum.
Clade E (D = 1, SYN = 10)
The "tomentose-leaved" group comprises G. densifolium, G. rotundifolium, G. tomentosum, and G. villosum.
Clade F (D = 9, SYN = 12)
The "sandplain" group. This clade contains a number of species of Gastrolobium that occur throughout the sandplains of middle and northern southwest Western Australia, and includes G. crassifolium, G. floribundum, G. diablophyllum, G. glaucum, G. hians ms, G. laytonii, G. microcarpum, G. polystachyum, G. propinquum, and G. pycnostachyum.
Clade G (D = 20, SYN = 34)
Three morphologically disparate species make up this clade, G. heterophyllum, G. nutans, and G. pusillum.
Clade H (autapomorphies = 20)
This clade contains a single species only, G. brownii, situated directly between clade G and clade I.
Clade I (D = 3, SYN = 12)
This group contains a number of species of Nemcia that are intermediate in morphology between Gastrolobium and Nemcia (N. hookeri, N. obovata, N. plicata, and N. spathulata), plus Gastrolobium bennettsianum, G. stowardii, and G. carinatum ms (aff. bennettsianum). These species have shortly racemose inflorescences, generally in the axils of the leaves.
Clade J (D = 1, SYN = 7)
Brachysema latifolium (the type species of Brachysema), Nemcia pulchella, and Gastrolobium truncatum are contained in this clade. There is also a group of Gastrolobium s.s. species (D = 9, SYN = 13), including G. appressum, G. calycinum, G. hamulosum, G. oxylobioides, G. reflexum ms, G. rigidum, G. spectabile, and G. tenue ms, that share glaucous leaves with strongly reticulate venation and an intramarginal vein.
Clade K (D = 2, SYN = 8)
This clade includes a number of Nemcia species, N. alternifolia ms, N. emarginata, N. reticulata, and N. retusa, as well as Oxylobium lineare. These species all have strongly tomentose calyces, which may be two-tone in color, and generally have inflorescences reduced to a few flowers in the leaf axils.
Clade L (D = 3, SYN = 8)
This group contains all bird-pollinated species within the greater Gastrolobium group except two (Brachysema latifolium and Gastrolobium grandiflorum), as well as three bee-pollinated species. This includes all but one species of Brachysema (B. bracteolosum, B. celsianum, B. melanopetalum, B. minor, B. modestum, B. praemorsum, B. sericeum, and B. subcordatum), Jansonia formosa, and the red-flowered Nemcia group, N. leakeana (the type species of Nemcia), N. luteifolia, N. rubra, and N. vestita. The bee-pollinated species of Nemcia included within this clade are N. coriacea, N. crenulata, and N. pyramidalis.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Crisp, 1994
), molecular data have clarified relationships. It is clear from this analysis that Gastrolobium s.s. is paraphyletic, with strong support for the inclusion of Brachysema, Jansonia, Nemcia, and O. lineare within Gastrolobium. Nemcia itself is polyphyletic within this clade.
Past classifications have circumscribed various genera primarily using floral characters that appear related to pollination syndrome (e.g., bird pollination vs. bee pollination), inflorescence structure, and fluoroacetate content (see Crisp and Weston, 1987
). Thoughts on the homology of such characters need to be reconsidered, especially in light of this analysis, which shows these to be homoplastic. More care needs to be taken when choosing morphological characters for phylogenetic reconstruction. Many characters in the past were more relevant to phenetic analyses rather than cladistic analyses, yet were used in cladistic analysis (e.g., Crisp and Weston, 1987, 1995
). This is not to say that morphology does not provide important phylogenetic information in the Gastrolobium group, simply that it does not provide enough resolution exclusive of other data. Indeed, this analysis gives some support for a large, mostly bird-pollinated lineage within Gastrolobium (clade L), showing that the floral characters that appear related to bird pollination are indeed phylogenetically informative. However, morphological characters were unable to satisfactorily work out the broader relationships of this lineage, which the molecular data more satisfactorily resolve.
The red-flowered species (Clade L, including Brachysema, Jansonia, and some species of Nemcia) form a well-supported clade (D = 3, SYN = 8), although three other species of Nemcia which have short, dense, many-flowered racemes with large, orange flowers are also within this clade (N. coriacea, N. crenulata, and N. pyramidalis). This conflicts with previous morphological work (Crisp, 1994
), which showed that bird pollination in the genera Brachysema, Jansonia, and Nemcia was due to convergence, and had arisen twice. This study shows that there is one main lineage within the greater Gastrolobium that appears to be bird pollinated, with only two other species (Brachysema latifolium and Gastrolobium grandiflorum) occurring singly outside this group, with bird pollination originating three times within Gastrolobium.
Morphology can be selected to change rapidly in response to change in pollination syndrome when a plant moves towards bird pollination (see review in Crisp, 1994
), so convergence among a number of different lineages is quite possible. In fact, only one or two genes may be responsible for flower color and shape (Gottleib, 1984; Coen, 1991
; Coen and Meyerowitz, 1991
), such that minimal genetic change may dramatically alter floral morphology. In the Brachysema and red-flowered Nemcia clade (clade L), a variety of floral shapes and colors are found. For example, the red-flowered Nemcia species do not have a reduced standard petal or enlarged keel petals, whereas the Brachysema species do. The colors range from white (e.g., B. modestum) through green (B. bracteolosum), red (e.g., B. subcordatum, Jansonia formosa, and Nemcia rubra) to black or very dark purple (B. melanopetalum). It is possible that these species shared a common ancestor that evolved towards bird pollination and then underwent an adaptive radiation, expanding into many shapes and colors. This may have been facilitated by the release of developmental constraints on the ancestral, yellow and red, bee-pollinated flowers in the rest of the Gastrolobium clade. This may also be true for other genera within the Mirbelieae and Bossieeae, as red flowers with elongated keels are found in species of Bossiaea Vent., Chorizema Labill., Daviesia Sm., Gompholobium Sm., Leptosema, Mirbelia Sm., and Sphaerolobium Sm.
Fluoroacetate is found in a number of clades within the Gastrolobium clade (clades D–J), but not in clades K or L (see, Aplin, 1971
; Twigg et al., 1996). It is possible that production of fluoroacetate is the plesiomorphic condition in this group (acquired in the ancestor of clade C), which was then lost from some lineages, most notably in the common ancestor of the red-flowered group (clade L) and a group of yellow-flowered Nemcia species (clade K). Toxin strength does not otherwise appear to decline in derived clades in the tree (except in the species of Nemcia intermediate with Gastrolobium in clade I, where trace levels have been recorded). Usually, fluoroacetate is either present or absent in these groups, implying that a mutation in the fluoroacetate metabolic pathway to interrupt production could have occurred, which could have led to a drastic reduction in fluoroacetate production, as found in N. spathulata by Twigg et al. (1996), or even a complete absence of fluoroacetate in some of the more derived clades.
Ovule number has been shown to be homoplastic throughout the tribe Mirbelieae (Crisp and Weston, 1987, 1995
). This study has shown this character to be equally homoplastic throughout Gastrolobium and related groups, and no support can be found for its use in past classifications to distinguish among various genera in this group, such as Gastrolobium, Nemcia, and Oxylobium.
Characteristics of the major clades
Most major clades (Figs. 2, 3) show consistency in morphology and ecology among their included species as described in detail below. Groups contained within clade D, which consists entirely of species from Gastrolobium s.s. and is sister to the rest of Gastrolobium s.l. (sensu lato; clade C), include the G. parviflorum complex and members of the G. bilobum group (the type of the genus), which all share condensed, many-flowered racemes and have cuneate, emarginate leaves, plus the strongly supported G. spinosum clade (D = 13, SYN = 23), which all have spinose leaves and short, few-flowered racemes. Clade D also contains a number of species occurring solely on granite outcrops and their immediate margins, though these do not form a clade (G. acrocaroli, G. callistachys, G. graniticum, G. involutum, G. semiteres, and G. stenophyllum). In fact, the only species occurring in the same habitat that is not within clade D is G. spectabile, which occurs within clade J. These granite-inhabiting species all share a similar inflorescence and floral structure (long, open racemes with long internodes, relatively large flowers, and strongly recurved calyx lobes) in addition to habitat.
Clade E contains the tomentose-leaved Gastrolobium species. These four species all share details of the inflorescence structure (strongly hairy, with short floral internodes and with large, lanceolate bracts that persist longer than in most species of Gastrolobium, that are caducous), and all except G. densifolium have leaves that are tomentose on the abaxial surface. These species are the sister group to the "sandplain" group (clade F), though this is weakly supported (D = 1, SYN = 10). These sandplain species are open, spreading shrubs that have tough, often glaucous leaves and long, open racemes, and generally have widespread distributions, occurring throughout the sandplains of southwestern Western Australia.
There is strong support for clade G (D = 20, SYN = 31), though the composition of this group is somewhat puzzling. Two of the three species, G. heterophyllum and G. nutans ms, share similar leaves and inflorescences, but the placement of G. pusillum there is interesting, although this species does not strongly resemble any other species of Gastrolobium s.l. It could be that this small group of species are well differentiated, with a number of morphological autapomorphies making them appear quite different.
The position of Gastrolobium brownii is interesting, as it was one of three species out of 22 transferred to Nemcia by Crisp and Weston (1987)
, along with G. pusillum and G. truncatum, only to be transferred back to Gastrolobium by Crisp and Weston (1995)
because of uncertainties in relationships based on morphology. Like most of Gastrolobium sens. str., G. brownii is well known to accumulate fluoroacetate. Gastrolobium brownii and G. truncatum both have inflorescences similar to many of those in the Nemcia group (short, few-flowered axillary racemes), and both sit with or near these species in the phylogeny presented here. Gastrolobium brownii is closely related to a group of species intermediate in morphology between Gastrolobium s.s. and Nemcia (clade I), which includes N. hookeri, N. obovata, N. spathulata, and N. plicata. These species of Nemcia have more in common with Gastrolobium s.s., such as short racemes, recurved calyx lobes, stipitate ovaries, and ovoid fruits, than with the circumscription of Nemcia provided by Crisp and Weston (1987)
. Gastrolobium truncatum is sister to Nemcia pulchella (with which it shares details of inflorescence structure such as short racemes and petal coloration) and also appears related to Brachysema latifolium (which has racemes of red, bird-pollinated flowers). Thus, the molecular data agree with the morphology in placing this group of species intermediate between Gastrolobium s.s. and Nemcia.
Within the rest of clade J, a group of Gastrolobium s.s. species form a strongly supported group (D = 9 and SYN = 13). These species have similar morphology, including glaucous leaves with strongly reticulate venation and an intramarginal vein, inflorescences with long floral internodes, very pubescent calyces, and deep orange standard petals.
Clade K contains a group of Nemcia species, plus Oxylobium lineare. Nemcia alternifolia and N. reticulata are sister species sharing standard petals that are almost entirely maroon on the back, an identical inflorescence type (solitary or paired flowers in the axils) and strongly tomentose calyces. Similarly tomentose calyces are shared with N. emarginata and N. retusa, however the latter two species have cuneate and emarginate leaves, two-toned hairs on the calyces (silver at the base, and golden brown at the top), and inflorescences clustered at the branchlet terminus with numerous flowers. Oxylobium lineare has similar leaves to those of Nemcia reticulata, but has a long raceme with many flowers that have uniformly colored hairs on the calyces, and may be a reversion to the typical, long Gastrolobium-type raceme as seen in the more basal Gastrolobium groups. Many species of Nemcia, and a few in Gastrolobium (such as G. heterophyllum), have short axillary shoots with a short, terminal raceme, so many developmental changes may not be required to further reduce this to a solitary flower.
The red-flowered group (clade L) includes Brachysema, Jansonia, and species of Nemcia with red flowers, and some Nemcia species with orange flowers in condensed, dense racemes. There is some support for a sister relationship between Brachysema celsianum and Jansonia formosa, though this is not found in all trees, and they share riverine habitats and straggly habits. The other Brachysema species have similar floral architecture. Three out of four red-flowered Nemcia species group together strongly and consistently (D = 5, SYN = 7), and all four group together in some of the most parsimonious trees. The three presumably bee-pollinated Nemcia species share a condensed terminal raceme with many, large flowers that are strongly orange in color and have large, crenulate leaves.
Phylogenetic utility of the various loci
Together, the three sequence regions used in this analysis provide a robust phylogeny. In a data set this large (94 taxa), numerous characters are required to obtain much resolution, and any DNA region alone is unlikely to yield a sufficient number of informative characters. For example, trnK 5' intron, which has the greatest number of informative characters (Table 2), only has 2.5 informative characters, on average, per taxon. In contrast, the combined analysis has 587 informative characters (including coded indels), or 6.3 characters, on average, per taxon.
Some regions appear more phylogenetically useful than others, however. The trees produced from only the trnK 5' intron and ETS data sets (not presented) more closely resemble the tree from the combined analysis than the psbA-trnH spacer data set. This may be due to the large number of indels (particularly deletions that can be large) present in the psbA-trnH spacer, most of which are not phylogenetically informative. These large deletions leave a many potential characters unavailable for many taxa. The analyses performed with and without coded indels produced trees with identical topologies, but support for individual clades increased slightly.
Different combinations of data sets also appear more phylogenetically informative than others, with the combined chloroplast data set most closely resembling the tree from the combined data. It also has the largest number of informative characters. The combined trnK/ETS data set is next most similar, with the topology of the psbA-trnH/ETS data set the most different from the combined analysis. There is no strong disagreement between nuclear and chloroplast sequences, however, with all topologies being quite similar, and the result of the partition homogeneity test showed no significant difference between the data sets, justifying the combination of the data. We believe that the best phylogeny is one with all data present, because the more characters added, the more robust the phylogeny becomes, though caution is recommended, and we believe that conflict between data sets must be taken into consideration.
Implications for taxonomy
This study has important implications for the taxonomy of the Gastrolobium group. The analysis provides strong support for the inclusion of Brachysema, Jansonia, and Nemcia within Gastrolobium s.s., so the continued recognition of these genera as currently circumscribed is untenable. There are two options for the taxonomy of this group. One would be to lump all species into Gastrolobium, and the other would be to split the Gastrolobium clade into different genera. The internal support of many branches within Gastrolobium (Figs. 2, 3) is quite low (decay = 1 for many of them), with many of these branches forming a ladder. In light of this low support and the overall shape of the tree, even though some individual groups have very strong support, the further splitting of Gastrolobium would not be the optimal solution. In addition, some of the genera would be difficult to distinguish morphologically, such as Brachysema and Jansonia, and Gastrolobium and Nemcia, due to the similar combinations of characters present in these genera. The lumping of Brachysema, Jansonia, Nemcia, and Oxylobium lineare appears to have the strongest support from this analysis.
Conclusions
Molecular data have an important role to play in estimating phylogenies in the Mirbelieae/Bossiaeeae. This study has shown that a taxonomically difficult group can be resolved using molecular data where morphology has achieved only partial success and is an important step forward in the systematics of Australian legumes. The circumscription of Gastrolobium may need to be expanded to include Brachysema, Jansonia, Nemcia, and Oxylobium lineare, pending further analysis using additional DNA markers. A future paper will present a systematic revision of the greater Gastrolobium, with descriptions of all species. Further investigation is required to test the possible sister relationship of the Podolobium ilicifolium group to Gastrolobium and of Gastrolobium to other genera within the tribe.
|
|
FOOTNOTES |
|---|
4 Author for reprint requests, current address: Department of Biology, Virginia Commonwealth University, 816 Park Ave, P.O. Box 842012, Richmond, Virginia, 23284-2012 USA (email: the.lizard{at}prontomail.com
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Aplin T. E. H. 1971 Poison plants of Western Australia: the toxic species of Gastrolobium and Oxylobium. Western Australian Department of Agriculture Bulletin 3772: 1-66
Asmussen C. B. A. Liston 1998 Chloroplast DNA characters, phylogeny, and classification of Lathyrus (Fabaceae). American Journal of Botany 85: 387-401[Abstract]
Baldwin B. G. S. Markos 1998 Phylogenetic utility of the external transcribed spacer (ETS) of 18S–26S nrDNA: congruence of ETS and ITS trees of Calycadenia (Compositae). Molecular Phylogenetics and Evolution 10: 449-463[CrossRef][Web of Science][Medline]
Bayer R. J. L. Hufford D. E. Soltis 1996 Phylogenetic relationships in Sarraceniaceae based on rbcL and ITS sequences. Systematic Botany 21: 121-134
Bena G. M.-F. Jubier I. Oliviera B. Lejeune 1998 Ribosomal external and internal transcribed spacers: combined use in the phylogenetic analysis of Medicago (Leguminosae). Journal of Molecular Evolution 46: 299-306[CrossRef][Web of Science][Medline]
Bentham G. 1864 Flora Australiensis, vol. 2. Reeve and Co., London, UK
Borisjuk N. L. Borisjuk G. Petjuch V. Hemleben 1994 Comparison of nuclear ribosomal RNA genes among Solanum species and other Solanaceae. Genome 37: 271-279[Medline]
Bremer K. 1988 The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42: 795-803[CrossRef][Web of Science]
Brown R. 1811 Gastrolobium. In W. T. Aiton [ed.], Hortus Kewensis, 2nd ed., part iii, 16. Longman, London, UK
Coen E. S. 1991 The role of homeotic genes in flower development and evolution. Annual Review of Ecology and Systematics 42: 241-279
———, andE. M. Meyerowitz 1991 The war of the whorls: genetic interactions controlling flower development. Nature 353: 31-37[CrossRef][Medline]
Crisp M. D. 1994 Evolution of bird-pollination in some Australian legumes (Fabaceae). In P. Eggleton and R. Vane-Wright [eds.], Phylogenetics and ecology, 281–309. Academic Press, London, UK
———,S. R. Gilmore B.-E. Van Wyk 2000 Molecular phylogeny of the genistoid tribes of papilionoid legumes. In P. S. Herendeen and A. Bruneau [eds.], Advances in legume systematics, part 9, 249–276. Royal Botanic Gardens, Kew, UK
———, and P. H. Weston 1987 Cladistics and legume systematics, with an analysis of the Bossiaeeae, Brongniartieae and Mirbelieae. In C. H. Stirton [ed.], Advances in legume systematics, part 3, 65–130. Royal Botanic Gardens, Kew, UK
———, and ———. 1995 Mirbelieae. In M. D. Crisp, and J. J. Doyle [eds.], Advances in legume systematics, part 7, Phylogeny, 245–282. Royal Botanic Gardens, Kew, UK
Domin K. 1923 Nemcia, a new genus of the Leguminosae. Preslia 2: 26-31
Eriksson T. 1998 AutoDecay version 4.0 (program distributed by the author). Bergius Foundation, Royal Swedish Academy of Sciences, Stockholm. Available on the web from http://www.bergianska.se/personal/TorstenE/
Felsenstein J. 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791[CrossRef][Web of Science]
Gadek P. A. D. L. Alpers M. H. Heslewood C. J. Quinn 2000 Relationships within Cupressaceae sensu lato: a combined morphological and molecular approach. American Journal of Botany 87: 1044-1057
Gardner C. A. H. W. Bennetts 1956 The toxic plants of Western Australia. West Australian Newspapers Ltd., Periodicals Division, Perth, Australia
Gilmore S. P. H. Weston J. A. Thompson 1993 A simple, rapid, inexpensive, and widely applicable technique for purifying plant DNA. Australian Systematic Botany 6: 139-148
Golenburg E. M. M. T. Clegg M. L. Durbin J. Doebley D. P. Ma 1993 Evolution of a noncoding region of the chloroplast genome. Molecular Phylogenetics and Evolution 2: 52-64[CrossRef][Medline]
Gottlieb L. D. 1984 Genetics and morphological evolution in plants. American Naturalist 123: 681-709[CrossRef][Web of Science]
Hilu K. W. L. A. Alice 1999 Evolutionary implications of matK indels in Poaceae. American Journal of Botany 86: 1735-1741
———, and H. Liang 1997 The matK gene: sequence variation and application in plant systematics. American Journal of Botany 84: 830-839[Abstract]
Hu J.-M. M. Lavin M. F. Wojciechowski M. J. Sanderson 2000 Phylogenetic systematics of the tribe Millettieae (Leguminosae) based on chloroplast trnK/matK sequences and its implications for evolutionary patterns in Papilionoideae. American Journal of Botany 87: 418-430
Jarrell D. C. M. T. Clegg 1995 Systematic implications of the chloroplast-encoded matK gene on the tribe Vandeae (Orchidaceae). American Journal of Botany 82: 137
Johnson L. A. J. L. Schultz D. E. Soltis P. S. Soltis 1996 Monophyly and generic relationships of Polemoniaceae based on matK sequences. American Journal of Botany 83: 1207-1224[CrossRef][Web of Science]
———, andD. E. Soltis 1994 matK DNA sequences and phylogenetic reconstruction in Saxifragaceae s. str. Systematic Botany 19: 143-156
———, and ———. 1995 Phylogenetic inference in Saxifragaceae sensu stricto and Gilia (Polemoniaceae) using matK sequences. Annals of the Missouri Botanic Gardens 82: 149-175[CrossRef][Web of Science]
Keighery G. J. 1980 Bird pollination in south Western Australia: a checklist. Plant Systematics and Evolution 135: 171-176[CrossRef][Web of Science]
———. 1982 Bird-pollinated plants in Western Australia. In J. A. Armstrong, J. M. Powell, and A. J. Richards [eds.], Pollination and evolution, 77–89. Royal Botanic Gardens, Sydney, Australia
———. 1984 Pollination of Jansonia formosa Kipp. ex Lindl. (Papilionaceae). Western Australian Naturalist 16: 21
Kelchner S. A. L. G. Clark 1997 Molecular evolution and phylogenetic utility of the chloroplast rpl16 intron in Chusquea and the Bambusoideae (Poaceae). Molecular Phylogenetics and Evolution 8: 385-397[CrossRef][Web of Science][Medline]
Kim S.-C. D. J. Crawford R. K. Jansen S.-G. Arnoldo 1999 The use of a non-coding region of chloroplast DNA in phylogenetic studies of the subtribe Sonchinae (Asteraceae: Lactuceae). Plant Systematics and Evolution 215: 85-99[CrossRef][Web of Science]
King K. R. A. Torres U. Zentgraf V. Hemleben 1993 Molecular evolution of the intergenic spacer in the nuclear ribosomal RNA genes of Cucurbitaceae. Journal of Molecular Evolution 36: 144-152[CrossRef][Web of Science][Medline]
Kuntze O. 1891 Revisio generum plantarum vascularium omnium atque cellularium multarum secundum leges nomeclaturae internationales: cum enumeratione plantarum exoticarum initinere mundi collectarum: mit Erläuterungen. Part 1. A. Felix, Leipzig, Germany
Les D. H. E. L. Schneider D. J. Padgett P. S. Soltis D. E. Soltis M. Zanis 1999 Phylogeny, classification and floral evolution of water lilies (Nymphaeaceae; Nymphaeales): a synthesis of non-molecular, rbcL, matK, and 18S rDNA data. Systematic Botany 24: 28-46[CrossRef][Web of Science]
Liang H. K. W. Hilu 1996 Application of the matK gene sequences to grass systematics. Canadian Journal of Botany 74: 125-134[CrossRef]
Linder C. R. L. R. Goertzen B. V. Heuvel J. Francisco-Ortega R. K. Jansen 2000 The complete external transcribed spacer of 18S–26S rDNA: amplification and phylogenetic utility at low taxonomic levels in Asteraceae and closely allied families. Molecular Phylogenetics and Evolution 14: 285-303[CrossRef][Web of Science][Medline]
Maddison D. R. 1991 The discovery and importance of multiple islands of most parsimonious trees. Systematic Zoology 40: 315-328[CrossRef]
Page R. D. M. 1996 TREEVIEW: an application to display phylogenetic trees on personal computers. Computer Applications in the Biosciences 12: 357-358
Plunkett G. M. D. E. Soltis P. S. Soltis 1996 Evolutionary patterns in Apiaceae: inferences based on matK sequence data. Systematic Botany 21: 477-495[CrossRef][Web of Science]
Plunkett G. M. D. E. Soltis P. S. Soltis 1997 Clarification of the relationship between Apiaceae and Araliaceae based on matK and rbcL sequence data. American Journal of Botany 84: 565-580[Abstract]
Rathgeber J. I. Capesius 1990 Nucleotide sequence of the intergenic spacer and the 18S ribosomal RNA gene from mustard (Sinapsis alba). Nucleic Acids Research 18: 1288
Rogers S. O. A. J. Bendich 1987a Heritability and variability in ribosomal RNA genes of Vicia faba. Genetics 117: 285-295
———, and ———. 1987b Ribosomal DNA genes in plants: variability in copy number and in the intergenic spacer. Plant Molecular Biology 9: 509-520[CrossRef][Web of Science]
Rogstad S. H. 1992 Saturated NaCl-CTAB solution as a means of field preservation of leaves for DNA analyses. Taxon 41: 701-708[CrossRef][Web of Science]
Sang T. D. J. Crawford T. F. Stuessy 1997 Chloroplast DNA phylogeny, reticulate evolution, and biogeography of Paeonia (Paeoniaceae). American Journal of Botany 84: 1120-1136[Abstract]
Sanger F. S. Nicklen A. R. Coulson 1977 DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences, USA 74: 5463-5467
Schiebel K. G. Waldburg J. Gerstner V. Hemleben 1989 Termination of transcription of ribosomal RNA genes of mung bean occurs within a 175 bp repetitive element of the spacer region. Molecular and General Genetics 218: 302-307
Schmidt-Puchta W. I. Gunther H. L. Sänger 1989 Nucleotide sequence of the intergenic spacer (IGS) of the tomato ribosomal DNA. Plant Molecular Biology 13: 251-253[CrossRef][Web of Science][Medline]
Stanford A. M. R. Harden C. R. Parks 2000 Phylogeny and biogeography of Juglans (Juglandaceae) based on matK and ITS sequence data. American Journal of Botany 87: 872-882
Steele K. P. R. Vilgalys 1994 Phylogenetic analysis of Polemoniaceae using nucleotide sequences of the plastid gene matK. Systematic Botany 19: 126-142
Sugiura M. 1992 The chloroplast genome. Plant Molecular Biology 19: 149-168[CrossRef][Web of Science][Medline]
Swofford D. 1997 PAUP: phylogenetic analysis using parsimony, pre-release version 4.0b3a. Laboratory of Molecular Systematics, Smithsonian Institution, Washington, D.C., and Sinauer, Sunderland, Massachusetts, USA
Twigg L. E. D. R. King L. H. Bowen G. R. Wright C. T. Easton 1996 Fluoroacetate content of some species of the toxic Australian genus, Gastrolobium, and its environmental persistance. Natural Toxins 4: 122-127[Medline]
Volkov R. S. Kostishin F. Ehrendorfer D. Schweizer 1996 Molecular organization and evolution of the external transcribed rDNA spacer region in two diploid relatives of Nicotiana tabacum (Solonaceae). Plant Systematics and Evolution 201: 117-129[CrossRef][Web of Science]
Wang X.-R. Y. Tsumura H. Yoshimaru K. Nagasaka A. E. Szmidt 1999 Phylogenetic relationships of Eurasian pines (Pinus, Pinaceae) based on chloroplast rbcL, matK, rpl20-rps18 spacer, and trnV intron sequences. American Journal of Botany 86: 1742-1753
Xiang Q.-Y. D. E. Soltis P. S. Soltis 1998 Phylogenetic relationships of Cornaceae and close relatives inferred from matK and rbcL sequences. American Journal of Botany 85: 285-297[Abstract]
This article has been cited by other articles:
|
|
 |
|
  W. J. Kress, K. J. Wurdack, E. A. Zimmer, L. A. Weigt, and D. H. Janzen Use of DNA barcodes to identify flowering plants PNAS, June 7, 2005; 102(23): 8369 - 8374. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Shaw, E. B. Lickey, J. T. Beck, S. B. Farmer, W. Liu, J. Miller, K. C. Siripun, C. T. Winder, E. E. Schilling, and R. L. Small The tortoise and the hare II: relative utility of 21 noncoding chloroplast DNA sequences for phylogenetic analysis Am. J. Botany, January 1, 2005; 92(1): 142 - 166. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Andreasen and B. G. Baldwin Reexamination of relationships, habital evolution, and phylogeography of checker mallows (Sidalcea; Malvaceae) based on molecular phylogenetic data Am. J. Botany, March 1, 2003; 90(3): 436 - 444. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
