| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
What's this? |
Systematics and Phytogeography |
South China Botanical Garden, Chinese Academy of Sciences, Guangzhou, 510650, China; Department of Botany, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20013-7012 USA; Rancho Santa Ana Botanic Garden, 1500 North College Avenue, Claremont, California 91711-3157 USA; College of Life Sciences, South China Agricultural University, Guangzhou, 510642, China
Received for publication May 25, 2006. Accepted for publication May 25, 2007.
ABSTRACT
Within the Poaceae, inflorescence diversification and its bearing on phylogeny and evolution are exceedingly complex. We used phylogenetic information of the "finger millet clade," a group of grasses with digitate inflorescences, to study the inflorescence diversification. This clade appears monophyletic in the morphological and molecular phylogenetic analyses. Three well-supported clades are shown in our cpDNA-derived phylogeny, with clades I and III consisting of species of Chloris and Microchloa, respectively, and clade II including species of Cynodon, Dactyloctenium, and Eleusine. Variation appears at different times throughout development. Changes involving primordium number and arrangement occur very early, changes involving duration of primordium activity occur much later. Characters derived from the comparison of developmental sequences were optimized onto the most parsimonious tree. The developmental characters were congruent with the molecular phylogeny. Two developmental characters may not be homologous in the Chloris subclade and the Cynodon subclade.
Key Words: Chloridoideae • development • finger millet clade • inflorescence • phylogeny • Poaceae
Thirteen inflorescence forms found in the Chloridoideae (Poaceae) can be categorized into four basic inflorescence types: paniculate, digitate, racemose, and spicate (Liu et al., 2005a
). These types can be characteristic of certain suprageneric groups, but analysis of the phylogenetic distribution of inflorescence types in their cladogram suggests that different forms have arisen independently many times (Liu et al., 2005b
). Homoplasy of inflorescence types in the Chloridoideae phylogeny obscures the mechanism by which inflorescence diversification has arisen.
Mature inflorescences are seen only at a single stage of development, and this may not be sufficient to understand their morphological diversification and relationships (Doust and Kellogg, 2002
; Stuessy, 2003
). Therefore, it may be preferable to look for homologies during the development of an inflorescence rather than in the mature form.
We have chosen to concentrate on the so-called finger millet clade (subtribe Chloridineae), which has 17 genera and approximately 150 species, sensu clade C1 in Hilu and Alice (2001)
. The finger millet clade has been found to be monophyletic in morphological (Van den Borre and Watson, 1997
; Liu et al., 2005b
) and molecular (Hilu and Alice, 2001
) studies. It includes genera whose inflorescences appear digitate. Homology of inflorescence structures was not recognized by previous workers, who placed members of the clade in different tribes (Clayton and Renvoize, 1986
). Chloris, Cynodon, Dactyloctenium, Eleusine, and Microchloa are the larger genera within the finger millet clade. By investigating these genera, we hope to discern developmental changes that have lead to the diversity seen in mature inflorescences (Figs. 1–6).
|
), or as several long paraclades circling a truncated main axis (Cámara-Hernández and Rua, 1991
; Liu et al., 2005a
). Observing inflorescence development will allow many questions to be properly formulated.
To define developmental homologies and to evaluate evolutionary diversification, we must trace developmental characters based on an explicit phylogenetic hypothesis. Accurate assessments of developmental and phylogenetic homology will impact estimates of the inflorescence diversification as well as our perception of the evolutionary processes through which they have arisen (Hufford, 2003
; Kellogg, 2004
). The present study aims to (1) explore the precise developmental sequences and clarify the nature of the digitate branching system, and (2) assess the homology of developmental characters in the finger millet clade through integration of molecular phylogenetic and developmental analyses.
MATERIALS AND METHODS
Taxon sampling and sequencing
Nine ingroup and two outgroup taxa were sampled for the phylogenetic reconstruction (Appendix 1). The outgroup taxa were chosen from the x = 10 Eragrostis clade (Hilu and Alice, 2001
; Roodt and Spies, 2003
). Ingroup taxa were chosen to show all of the different inflorescence forms in the finger millet clade. All samples were freshly collected in the field and were identified using keys in Keng (1959)
and Chen (1990).
DNA was extracted with CTAB described in Doyle and Doyle (1987)
. Two chloroplast markers including trnL intron and rps16 intron were selected because they are highly informative noncoding regions (Shaw and Small, 2004
). These cpDNA regions were amplified using primers of Taberlet et al. (1991)
for the trnL intron and of Oxelman et al. (1997)
for the rps16 intron. DNA amplifications were performed in 50 µL reactions with approximately 10–50 ng of total DNA, 10 mmol Tris buffer/L (100 mmol Tris-HCl/L pH 8.8, 250 mmol KCl/L), 0.2 µmol dNTPs/L, 10 µmol of each primer/L, 5 U of Taq polymerase, and 25 µmol MgCl2/L. The polymerase chain reactions (PCR) was run on a Peltier Thermal Cycler DNA engine DYAD (MJ Research, Watertown, Massachusetts, USA) starting at 94°C for 3 min; followed by 33 cycles of 94°C for 1 min, 52°C for 1 min, and 72°C for 1 min; and ending with a final extension of 72°C for 7 min. To be cleaned for sequencing, the PCR products were then run on a 1% low-melting agarose gel, and the bands excised and gelased, using the GELase (Epicentre Technologies, Madison, Wisconsin, USA) method. Dideoxy cycle sequencing was conducted using BigDye version 3 reagents (Applied Biosystems, Foster City, California, USA) with an ABI model 377 automated sequencer (Applied Biosystems). Cycling conditions included an initial denaturing step at 94°C for 5 min, followed by 30 cycles of 96°C for 10 s, 50°C for 5 s, and 60°C for 4 min. Contigs were assembled using Sequencher 3.1.1 (Gene Codes, Ann Arbor, Michigan, USA). Sequences (Appendix 1) have been deposited in GenBank. They were then aligned by ClustalX version 1.83.1 (PC version) (Thompson et al., 1997
) and manually adjusted using the program MacClade 4.0 (Maddsion and Maddsion, 2000
).
Phylogenetic analyses
Phylogenetic analyses were performed with PAUP* 4.0b10 (Swofford, 2002
) using maximum parsimony (MP) (Farris et al., 1970
; Fitch, 1971
) and maximum likelihood (ML) (Felsenstein, 1981
) methods. The heuristic searches for MP analysis were conducted with 100 random taxon addition replicates, tree-bisection-reconnection (TBR) branch swapping, and collapse of zero-length branches. Character state changes were equally weighted in the analysis. Gaps were treated either as missing data or as new characters. The support values of clades revealed in the maximum parsimony tree(s) (MPTs) were examined using 1000 bootstrap replicates with the heuristic search options. Chi-square tests of PAUP* were run to reveal nucleotide bias among taxa. Congruence between the two chloroplast data sets was assessed with the incongruence length difference (ILD) test (Farris et al., 1995
). The ILD test has been criticized as a method for assessing character incongruence (Yoder et al., 2001
; Barker and Lutzoni, 2002
). However, Hipp et al. (2004)
has recently shown that the ILD test can serve as a conservative initial test of data partition congruence. The appropriate model for DNA substitution for the maximum likelihood analysis was determined using Modeltest version 3.06 (Posada and Crandall, 1998
). Once the best-fit model was determined, maximum likelihood searches were performed on the data set (Felsenstein, 1981
).
The best-fit analysis of the combined cpDNA data sets including trnL intron and rps16 intron (including gap characters) was performed with WinClada version 1.00.08 (Nixon, 2002
) running the program NONA (Goloboff, 1993
) as a daughter process. In the NONA analysis, three sequential parsimony ratchet runs were chosen, and each replicate included 200 iterations with 20 trees held in memory. Support for cladistic relationships was assessed using bootstrap analysis (Felsenstein, 1985
) with 1000 replications that included 10 searches and 10 starting trees in each replication.
Developmental analysis
Approximately 40 inflorescences in various developmental stages for all taxa (Appendix 1) were collected and preserved in FAA (1 part formalin, 1 part glacial acetic acid, 18 parts 70% ethanol). After dissection with a Nikon SM-Z645 stereomicroscope (Nikon, Tokyo, Japan), the samples were fixed overnight in 2% (v/v) glutaraldehyde in phosphate buffer at 4°C. The samples were then rinsed three times for 5 min each, and then kept 2 h in 1% osmium tetroxide (OsO4, pH = 6.8) at room temperature. Samples were dehydrated in a series of ethanol solutions (70%, 90%, and 100% three times), critical point dried in an Eiko DX-1 critical-point dryer (Eiko Engineering, Hitachi, Japan), and sputter-coated with gold in a JFC-1100 sputter coater (JEOL, Tokyo, Japan). Inflorescences were examined and photographed with a Jeol JSM-6360V scanning electron microscope (SEM) (JEOL, Tokyo, Japan) at 10 kv. Use of continuous quantitative characters in phylogenetic analysis has been criticized by some authors (e.g., Stevens, 1991
) but defended by others based on simulation studies and other empirical evidence (Poe and Wiens, 2000
). Because of the lack of consensus on this topic and because our aim was to define the homology of continuous quantitative characters, we decided to include continuous quantitative characters in our analyses.
Polar (longitudinal) axis diameter (Dp) and equatorial (transverse) axis diameter (Dt) of primordium were calculated using SEM photos, and interpretation of Dp/Dt corresponding to different primordium shapes follows Erdtman (1952)
. Twenty-three characters were selected on the basis of intergeneric variations (Table 1). They consisted of 17 binary and six multistate characters. All multistate characters were treated as unordered. Six quantitative characters were coded following the step matrix gap weighting (SMGW) method of Wiens (2001)
. The SMGW method assigned to each quantitative character a unique mean value, and the changes between these states are specified in a step matrix (Appendix S1 in Supplemental Data accompanying the online version of this article).
|
) was used to optimize (Fitch, 1971
) inflorescence characters on the most parsimonious tree inferred from the combined molecular data set. RESULTS
Phylogeny analysis
Chi-square tests of homogeneity of base frequencies across taxa were run in PAUP* for the nonambiguous sites for (1) all 11-taxon trnL intron data (
2 = 4.7645, df = 30, P = 1), rps16 intron data (2 = 12.3213, df = 30, P = 0.99), or combined trnL intron and rps16 intron data (2 = 38.0047, df = 30, P = 0.15), and for (2) ingroup 9-taxon concatenated trnL intron and rps16 intron, using the nonambiguous sites (2 = 27.6696, df = 24, P = 0.27). None of the tests revealed nucleotide bias among taxa. The trnL intron alignment comprised 591 positions, of which 315 were parsimony informative characters. The rps16 intron alignment comprised 853 positions, of which 451 were parsimony informative characters. Because these two cpDNA data sets were found to be congruent using the ILD test (P = 1.00), we combined them. The combined data set of 11 taxa comprised 1444 positions, 766 of which were parsimony informative.
With gaps scored as missing data, the MP analysis of the combined data set generated one MPT with a total length of 990 steps, a consistency index (CI) excluding uninformative characters of 0.96, a retention index (RI) of 0.98, and a rescaled consistency index (RC) of 0.94 (Fig. 7). Microchloa, Cynodon, Eleusine, Dactyloctenium, and Chloris formed a monophyletic group in one most parsimonious tree. Within this finger millet clade, three major clades were recovered, with clades I and III consisting of species of Chloris and Microchloa, respectively, clade II including species of Cynodon, Eleusine, and Dactyloctenium. Bootstrap support for clades I, II, and III was 99%, 100%, and 99%, respectively. Clade III is revealed to be sister to the clades I and II. Even when the gaps were treated as new characters, the topology of the tree did not change. Modeltest suggested that the GTR + I + G was the best model (–ln likehood score = 5392.73). The base frequencies were A = 0.33, C = 0.17, G = 0.16, T = 0.35, and the substitution rate matrix was as follows: A/C = 0.14, A/G = 0.25, A/T = 0.10, C/G = 0.14, C/T = 0.22, G/T = 0.15, the Ti/Tv = 0.81. The gamma shape parameter was 1.85. The maximum likelihood tree (not shown) had an almost identical topology to the MPT (Fig. 7) except that Chloris barbata and clade II formed a monophyletic group, sister to C. virgata. Owing to the lack of resolution of monophylesis for Chloris in the ML analysis, the MP tree was chosen to map the developmental characters on our cladogram.
|
|
|
|
|
|
Spikelet differentiation on the 2° axis of Chloris is basipetal (Fig. 15) and starts with the inception of two alternate concave glume primordia. The first formed primordium develops into a lower glume and the second one into an upper glume (Fig. 16). The glume arrangement along the 2° axis is similar in all examined species where the lower glume of spikelet is central, in contrast to lateral upper glume of spikelets on the abaxial surface of the 2° axis. In C. barbata, four lemma primordia on each spikelet are initiated in turn, while in C. virgata only two lemma primordia are initiated. The first floret of every spikelet differentiates and matures first (Figs. 20, 22, 24), while the remainder of the floret primordia are undifferentiated. When the spikelet axis elongates in C. barbata, the lemma primordium of the first floret arises (Fig. 16). Subsequently, the first floret primordium initiates (Fig. 17). The lemma of the second floret initiates alternately (Fig. 18). Simultaneously, the first floret primordium differentiates. The palea primordium initiates first (Fig. 19). Three stamen primordia arise after the formation of the palea (Fig. 20). Two of them are initiated on the lateral flanks of the primordium and one, abaxially (Fig. 21). Following the initiation of lemma of the third floret (Fig. 22), the gynoecial primordium of the first floret initiates and develops a gynoecial ridge surrounding the ovule primordium (Fig. 23). The lemma of the fourth floret initiates (Fig. 24). After that, the two lodicules of the first floret initiate in a whorl outside the stamen primordia (Fig. 25), and at the same time, the stamen primordia of the first floret expand to form thecae. Later, the filaments of each stamen gradually elongate, the style branches, and stigma develop. In C. barbata, the fourth floret primordium is reduced to a lemma vestigium without an awn. Therefore, there are three protruding awns on every spikelet at the end of anthesis (Figs. 1, 24). In C. virgata, only the second floret develops, with the lemma forming an awn. Therefore, there are two protruding awns on every spikelet at the end of anthesis (Figs. 2, 26). Finally, the 2° axis of both species ends in a blind extension (Fig. 27).
|
Spikelet differentiation on 2° axes of C. dactylon and C. arcuatus is amphipetal (Fig. 35) and starts with the formation of two alternate concave glume primordia (Fig. 36) on the abaxial surface of 2° axis. In addition, the 2° axis ends in an undifferentiated spikelet primordium (Fig. 37). The initiation of florets along the spikelet axis is acropetal (Figs. 38–41). While the spikelet axis elongates, the lemma primordium of the first floret arises (Fig. 38). Almost simultaneously, the first floret primordium initiates (Fig. 39), and the lemma of second floret initiates alternately (Fig. 40). The second floret primordium then initiates (Fig. 41), and the remaining part of spikelet meristem develops into a degenerative floret. The palea primordium of the first floret initiates first (Fig. 42), and simultaneously three stamen primordia of first floret arise. Sequentially, the gynoecial primordium of the first floret initiates from the remaining part of floret primordium (Fig. 43) and develops a gynoecial ridge surrounding the ovule primordium. At the same time, stamen primordia of the first floret expand to form thecae. After that, the two lodicules of the first floret initiate in a whorl outside the thecae (Fig. 44). Later, stamen filaments gradually elongate, and style branches and stigma develop. By anthesis, the floral organs of every spikelet are completely enveloped by glumes (Fig. 3). Finally, the 2° axis ends in a degenerative spikelet (Fig. 45).
|
Spikelet differentiation on the 2° axis of E. indica is basipetal (Fig. 53) and starts with the formation of two keeled, concave glume primordia (Fig. 54) on the abaxial surface of 2° axis. In addition, the 2° axis ends in a differentiated spikelet primordium (Fig. 55). The initiation of florets along the spikelet axis of E. indica and E. coracana is acropetal (Figs. 56–58, 60). While the spikelet axis elongates, the lemma primordium of first floret arises, and almost simultaneously, the first floret primordium initiates (Fig. 56). The lemma of the second floret initiates alternately, and the second floret primordium then initiates (Fig. 56). The lemma of third floret initiates alternately, and the third floret primordium initiates (Fig. 57). Then the first floret primordium differentiates and the palea primordium initiates. Three stamen primordia arise after the formation of palea following the initiation of lemma of the fourth floret (Fig. 58). Then, the gynoecial primordium of first floret initiates from the remaining part of the floret primordium (Fig. 59). Accompanying the formation of the fourth floret primordium (Fig. 60), a gynoecial ridge forms surrounding the ovule primordium (Fig. 61). The lemma of fifth floret initiates (Fig. 62), and the two lodicules of the first floret subsequently initiate in a whorl outside the stamen primordia (Fig. 63); at the same time, stamen primordia of the first floret expand to form thecae. Later, the filaments of the stamens gradually elongate, and style branches develop. By anthesis, the floral organs of each spikelet are completely enveloped by the glumes. Finally, the 2° axis ends in a perfect spikelet in both species of Eleusine (Fig. 4).
|
Spikelet differentiation on the 2° axis of Dactyloctenium aegyptium is amphipetal (Fig. 71) and starts with the formation of concave glume primordia (Fig. 72) on the abaxial surface of 2° axis. In addition, the 2° axis ends in a microtip (Fig. 73). The initiation of florets along the spikelet axis of D. aegyptium is acropetal (Figs. 74–76), and while the spikelet axis elongates, the lemma primordium of the first floret arises. Almost simultaneously, the first floret primordium initiates (Fig. 74), and the lemma of the second floret initiates alternately; the second floret primordium then initiates (Fig. 75). The lemma of the third floret initiates alternately (Fig. 76), and then the first floret primordium differentiates. The palea primordium initiates first, and three stamen primordia arise after the formation of palea (Fig. 77). Following the initiation of the third floret primordium (Fig. 78), the gynoecial primordium of the first floret differentiates. Accompanying the differentiation of second floret, the gynoecial ridge develops and surrounds the ovule primordium (Fig. 79). Subsequently, the two lodicules of the first floret initiate in a whorl outside the stamen primordia (Fig. 80), and at the same time, stamen primordia of first floret expand to form thecae. Later, stamen filaments elongate when the two lodicules (Fig. 81) and style branches develop. By anthesis, the floral organs of each spikelet are completely enveloped by glumes, and finally, the 2° axis ends in a microtip (Fig. 5).
|
Spikelet differentiation on the 2° axis of Microchloa is basipetal (Fig. 89) and starts with the formation of two concave glume primordia (Fig. 90) on the abaxial surface of 2° axis. In addition, the 2° axis ends in a microtip (Fig. 91), while the lemma primordium of first floret arises (Fig. 91). Almost simultaneously, the first floret primordium initiates. The palea primordium initiates first, and three stamen primordia (E = 20 µm) arise sequentially (Fig. 92). The gynoecial primordium of the first floret initiates and develops a gynoecial ridge that surrounds the ovule primordium (Fig. 93). Subsequently, two lodicules of the first floret initiate in a whorl outside the stamen primordia (Fig. 94). At the same time, the stamen primordia of the first floret expand to form thecae. The style branches develop later. By anthesis, the distal spikelets are incompletely developed and enveloped by glumes (Fig. 95), and finally, the 2° axis ends in a microtip.
|
Distribution of developmental characters on a molecular tree
Three characters were uninformative when mapped onto the cladgram (Fig. 96, Table 1). Of the remaining characters that were informative, two appear as homoplastic synapomorphies (characters 2 [Dt of 1°-axis primordium] and 3 [100·Dp/Dt of 1°-axis primordium]), and six characters appear as nonhomoplastic synapomorphies (characters 7 [arrangement of 2°-axis primordia], 9 [length of 2°-axis at the moment of 3°-axis initiation], 10 [end of 2° axis], 11 [peduncle length], 14 [P of distal spikelet primordium], and 16 [initiation succession of spikelets on 2° axis]).
|
Phylogeny
Ingroup taxa were chosen to cover the major morphological groups in the finger millet clade (Hilu and Alice, 2001
; Liu et al., 2005a
). Although these results need to be confirmed with data from more taxa, it is interesting to explore the implications of the typology discovered. With two Eragrostis species used as outgroups (Hilu and Alice, 2001
), Microchloa, Cynodon, Eleusine, Dactyloctenium, and Chloris formed a monophyletic group in the most parsimonious tree. Within this finger millet clade, three major clades were recovered, with clades I and III consisting of species of Chloris and Microchloa, respectively, and clade II including species of Cynodon, Eleusine, and Dactyloctenium. Clade III was found to be sister to clade I/II.
Our results suggest that Eleusine and Dactyloctenium should be included in the subtribe Choridinae. The monophyletic group composed by clade I/II has high bootstrap support (100%) and contains two well-supported clades. Clade I consists of plants with one fertile floret per spikelet, and clade II consists of plants with two to many fertile florets per spikelet. Therefore, Clayton and Renvoize (1986)
included Chloris, Cynodon, and Microchloa in the Chloridinae and Eleusine and Dactyloctenium in the Eleusininae. Peterson et al. (2001)
included Eleusine and Dactyloctenium in the Eleusininae. Hilu and Alice (2001)
grouped Eleusine and Dactyloctenium in a weakly supported clade with phylogenetic affinities with other genera in the Chloridinae. The relationship between clades I and II in this study reflects morphological similarities in developmental characters such as two to nine 2° axes surrounding the 1° axis (character 6), semiverticillate or verticillate arrangement of the 2°-axis primordia (character 7), laterally compressed spikelets (character 17), and 2–5 florets per spikelet (character 19). Because the two subtribes Eleusininae and Chloridinae do not seem to be natural taxonomic units, especially for those taxa undergoing rapid speciation, these groups should be reclassified after more species have been sampled to gain a better understanding of their evolutionary relationships (Peterson et al., 1997
; Jacobs and Everett, 2000
).
Clade III, which comprises two species of Microchloa, has strong bootstrap support (99%). Our analysis confirms the sister relationship between Microchloa and Chloris allies because these genera form a strongly supported finger clade (99%). The distinct developmental characters of clade III are the single 2°-axis primordium (character 6) and the unilateral arrangement of the 2° axis (character 7). Because Microchloa is one of the early derived lineages, we hypothesize that digitate inflorescences with a single 2° axis arose relatively early in the evolutionary history of finger millet clade.
Combined analysis
The overall picture of development is one of early similarity followed by later diversification. Developmental characters were optimized onto one parsimonious cladogram and were mostly congruent with the molecular phylogeny. The combined analysis had an identical topology with the molecular tree (Fig. 7). Clade I/II as a whole and clade II are not defined by any developmental characters, but clade III is defined by the basipetal succession of spikelet initiation on the 2° axis (character 16). Dactyloctenium (clade II) is defined by two characters: semiverticillate arrangement of the 2°-axis primordia (character 7) and a microtip in the end of the 2° axis (character 10).
The equatorial axis diameter of the 1°-axis primordium (character 2) and 100·Dp/Dt of 1°-axis primordium (character 3) originated twice, once within Cynodon (clade II) and once within Chloris (clade I). The parallel derivation of homoplastic synapomorphies suggests that these character states (characters 2 and 3) in Cynodon may not be homologous with the structures in Chloris.
In most cases, temporal variation in developmental characters would not have been detected by investigating mature morphology. For example, D. aegyptium (Fig. 5) differs from C. virgata (Fig. 2) in having a lax digitate inflorescence (Figs. 2, 5). Initiation of the 2°-axis promordia occured early in both genera, but further spread of the axes did not occur in C. virgata, resulting in a contracted inflorescence at maturity. In contrast, the spread of 2° axes occurred late in the development of D. aegyptium, leading to a distinctly spreading inflorescence.
In addition to traditional morphological studies, developmental analysis can also provide novel explanations for interpreting differences among genera. In the monophyletic finger millet clade, the end of the 2° axis has been used as a diagnostic taxonomic character (Chen, 1990
). In species of Microchloa, the 2° axis terminates with a microtip, while in others, such as species of Cynodon, the 2° axis terminates with a spikelet. Early in their development, however, all species have a similar terminal morphology of the 2° axis. Differences appear at maturity because spikelets are initiated basipetally along the 2° axis in species of Microchloa with the microtip remaining undifferentiated in its initial developmental stage, while in Cynodon spikelets are initiated amphipetally along the 2° axis, with a spikelet primordium at the end of the 2° axis late in development. Therefore, the "character" of the end of the 2° axis is composed of another developmental character, i.e., initiation of a spikelet on a 2° axis. In contrast, in species of Cynodon, the end of the 2° axis is determined later in development and thus is truly nonhomologous to the end of 2° axis in species of Microchloa.
The role of developmental variation in inflorescence formation in the finger millet clade can be diagrammed using the simple model in Fig. 97. Changes involving primordium differentiation that result from activity of the apical meristem for all orders of axes occur very early. Changes involving axis elongation, including the arrest of the 1° axis, the elongation of 2° axes, and the formation of 3° axes (i.e., spikelets), occur much later. For different grass groups, axis elongation plays a critical role in the external appearance of the inflorescence. For example, the arrest of the 1° axis in the finger millet clade determines that the main axis terminates in a remnant and forms a digitate inflorescence, while in the bristle clade (Doust and Kellogg, 2002
), the 1° axis continually develops to maturity and forms a paniculate inflorescence. Therefore, the remarkable discrimination of development of the same character suggests that the various developmental characters may be under different genetic control (Kellogg, 2004
).
|
; Liu et al., 2005a
). Studies of gene expression involved in axis elongation are needed to elucidate inflorescence diversification in the finger millet clade. The external appearance of maize inflorescence is regulated by three RAMOSA genes, i.e., RA1, RA2, and RA3 (Vollbrecht et al., 2005
; Bortiri et al., 2006
; Satoh-Nagasawa et al., 2006
). The RA2 is critical for shaping the initial steps of inflorescence formation in the grass family because RA2 controls the expression of other genes and regulates the activity of the 2°-axis primordia. The RA2 also controls the expression pattern and is conserved in other grasses including rice, barley, and sorghum (Bortiri et al., 2006
). Given this current knowledge, any of the three genes could be involved in the activity of the 2°-axis primordia in the finger millet clade. Accordingly, analysis of the RA2 gene expression and the interplay between three RAMOSA genes in the finger millet clade may elucidate the molecular mechanism underlying the branching pattern of digitate inflorescences. APPENDIX 1.
Voucher information and GenBank accession numbers for sequences for the trnL and rps16 introns for the taxa used in this study. Voucher specimens are deposited in the following herbaria: IBSC = Herbarium of South China Botanical Garden, the Chinese Academy of Sciences; QL = Qing LIU
Taxon—GenBank accessions: trnL, rps16; Source; Voucher specimen.
Chloris barbata Sw.—AY787812, DQ242043; Hainan, China; QL 05026, IBSC. C.virgata Sw.—AY787813, DQ242043; Yunnan, China; QL 02022, IBSC. Cynodon dactylon (L.) Pers.—AY787808, DQ242038; Guangdong, China; QL 04035, IBSC. C. arcuatus J. Presl et C. Presl—AY787807, DQ242039; Hainan, China; QL 02016, IBSC. Eleusine coracana (L.) Gaertn.—AJ489464a, DQ242040; Anhui, China; QL 03020, IBSC. E. indica (L.) Gaertn.—AY787806, DQ242041; Yunnan, China; QL 02028, IBSC. Dactyloctenium aegyptium (L.) P.Beauv.—AY787809, DQ242042; Hainan, China; QL 05074, IBSC. Microchloa kunthii Desv.—AY787814, DQ242045; Yunnan, China; QL 02060, IBSC. M. indica (L. f.) P. Beauv.—AY787815, DQ242046; Hainan, China; QL 02012, IBSC. Eragrostis pilosa (L.) P. Beauv.—AY787810, AY136859a; Guangdong, China; QL 02091, IBSC. E. cylindrica (Roxb.) Nees—AY787811, DQ242047; Yunnan, China; QL 02060, IBSC.
FOOTNOTES
101 aArchived GenBank sequences 
1 The authors thank Z.-R. Song and S.-F. Zhang for help with field collection; X.-L. Wang, J.-Y. Yao, and S.-L. Zhou for assistance in the molecular laboratory; X.-Y. Hu and Z.-W. Guo for their kind help with the SEM operation; and E. A. Kellogg, A. N. Doust, and G. H. Rua for valuable comments. This study was financially supported by the Smithsonian Institution Short-Term Visitor Awards to Q.L., South China Botanical Garden Director's Fund to Q.L. (2005–1140), Project of Bureau of Life Sciences and Biotechnology (2002–1091) of the Chinese Academy of Sciences to N.X.Z. (2002–1091), and partly supported by Operation Fund of Molecular Systematics Laboratory of SCBG (0790003002). The authors also thank two anonymous reviewers for their constructive comments that greatly improved the manuscript.
6 Author for correspondence (liuqing{at}scib.ac.cn
)
LITERATURE CITED
Barker F. K. Lutzoni F. M.. 2002. The utility of the incongruence length difference test. Systematic Biology 51: 625-637.[CrossRef][Web of Science][Medline]
Bortiri E. Chuck G. Vollbrecht E. Rocheford T. Martienssen R. Hake S.. 2006. RAMOSA2 encodes a LATERAL ORGAN BOUNDARY domain protein that determines the fate of stem cells in branch meristems of maize. Plant Cell 18: 574-585.
Cámara-Hernández J. Rua G. H.. 1991. The synflorescence of Poaceae. Beiträge zur Biologie der Pflanzen 66: 297-311.
Chen S. L.. 1990. Gramineae (4)—subfamily Eragrostoideae Pilger. In S. L. Chen [ed.], Flora reipublicae popularis sinicae, vol. 10 1-131 Science Press, Beijing, China.
Clayton W. D. Renvoize S. A.. 1986. Genera graminum. Kew Bulletin Additional Series XIII Her Majesty's Stationery Office, London, UK 389.
Doust A. N. Kellogg E. A.. 2002. Inflorescence diversification in the panicoid "bristle grass" clade (Paniceae, Poaceae): evidence from molecular phylogenies and developmental morphology. American Journal of Botany 89: 1203-1222.
Doyle J. J. Doyle J. L.. 1987. A rapid DNA isolation procedure for small quantities of fresh tissues. Phytochemical Bulletin 19: 11-15.
Erdtman G.. 1952. Pollen morphology and plant taxonomy—angiosperms Almqvist and Wilksell, Stockholm, Sweden. 466.
Farris J. S. Källersjö M. Kluge A. G. Bult C.. 1995. Testing significance of incongruence. Cladistics 10: 315-319.[CrossRef][Web of Science]
Farris J. S. Kluge A. G. Eckardt M. J.. 1970. A numerical approach to phylogenetic systematics. Systematic Zoology 19: 172-191.
Felsenstein J.. 1981. Evolutionary trees from DNA sequences: a maximum likelihood approach. Journal of Molecular Evolution 17: 368-376.[CrossRef][Web of Science][Medline]
Felsenstein J.. 1985. Confidence limits of phylogenies: an approach using the bootstrap. Evolution 39: 783-791.[CrossRef][Web of Science]
Fitch W.. 1971. Toward defining the course of evolution: minimum change for a specific tree typology. Systematic Zoology 20: 406-416.[Abstract]
Goloboff P. A.. 1993. NONA version 2.0 Computer program published by the author, Tucumán, Argentina (available at http://www.cladistics.com).
Hilu K. W. Alice L. A.. 2001. A phylogeny of Chloridoideae (Poaceae) based on matK sequences. Systematic Botany 26: 386-405.[Web of Science]
Hipp A. L. Haii J. C. Sytsma K. J.. 2004. Congruence versus phylogenetic accuracy: revisiting the incongruence length difference test. Systematic Biology 53: 81-89.
Hufford L.. 2003. Homology and developmental transformation: models for the origins of the staminodes of Loasaceae subfamily Loasoideae. International Journal of Plant Sciences 164: S409-S439.[CrossRef][Web of Science]
Jacobs S. W. L. Everett J.. 2000. Grasses: systematics and evolution CSIRO, Collingwood, Victoria, Australia.
Kellogg E. A.. 2004. Evolution of developmental traits. Current Opinion in Plant Biology 7: 92-98.[CrossRef][Web of Science][Medline]
Keng Y. L.. 1959. Flora illustralis plantarum primarum sinicarum Gramineae Science Press, Beijing, China.
Liu Q. Zhao N. X. Hao G.. 2005a. Inflorescence structures and evolution in subfamily Chloridoideae (Gramineae). Plant Systematics and Evolution 251: 183-198.[CrossRef][Web of Science]
Liu Q. Zhao N. X. Hao G.. 2005b. Phylogenetic relationships of the Chloridoideae (Gramineae): a cladistic analysis. Journal of Tropical and Subtropical Botany 13: 432-442.
Maddsion D. R. Maddsion W. P.. 2000. MacClade, version 4.0 Sinauer, Sunderland, Massachusetts, USA.
Niklas K. J.. 1987. Pollen capture and wind-induced movement of compact and diffuse grass panicles: implications for pollination efficiency. American Journal of Botany 74: 74-89.[CrossRef][Web of Science]
Nixon K. C.. 2002. WinClada, version 1.00.08 Computer program published by the author, Ithaca, New York, USA (available at http://www.cladistics.com).
Oxelman B. Lidén M. Berglund D.. 1997. Chloroplast rps16 intron phylogeny of the tribe Sileneae (Caryophyllaceae). Plant Systematics and Evolution 206: 393-410.[CrossRef][Web of Science]
Peterson P. M. Soreng R. J. Davidse G. Filgueiras T. S. Zuloaga F. O. Judziewicz E. J.. 2001. Catalogue of New World grasses (Poaceae). II. Subfamily Chloridoideae. Contributions from the United States National Herbarium 41: 1-255.
Peterson P. M. Webster R. D. Valdes-Reyna J.. 1997. Genera of New World Eragrostideae (Poaceae: Chloridoideae). Smithsonian Contributions to Botany 87: 1-50.
Poe S. Wiens J. J.. 2000. Characters selection and the methodology of morphological phylogenetics. In J. J. Wiens [ed.], Phylogenetic analysis of morphological data 20-36 Smithsonian Institution Press, Washington, D.C., USA.
Posada D. Crandall K. A.. 1998. MODELTEST: testing the model of DNA substitution. Bioinformatics 14: 817-818.
Roodt R. Spies J. J.. 2003. Chromosome studies in the grass subfamily Chloridoideae. I. Basic chromosome numbers. Taxon 53: 557-566.
Satoh-Nagasawa N. Nagasawa N. Malcomber S. Sakai H. Jackson D.. 2006. A trehalosa metabolic enzyme controls inflorescence architecture in maize. Nature 441: 227-230.[CrossRef][Medline]
Shaw J. Small R. L.. 2004. Addressing the "hardest puzzle in American pomology": phylogeny of Prunus sect. Prunocerasus (Rosaceae) based on seven noncoding chloroplast DNA regions. American Journal of Botany 91: 985-996.
Stevens P. F.. 1991. Character states, morphological variation, and phylogenetic analysis: a review. Systematic Botany 16: 553-583.[CrossRef][Web of Science]
Stuessy T. F.. 2003. Morphological data in plant systematics. In T. F. Stuessy, V. Mayer, and E. Hörandl [eds.], Deep morphology: toward a renaissance of morphology in plant systematics 299-315 Gantner, Koeltz, Königstein, Germany.
Swofford D. L.. 2002. PAUP*: phylogenetic analysis using parsimony (*and other methods), version 4.0b10 Sinauer, Sunderland, Massachusetts, USA.
Taberlet P. Gielly L. Pautou G. Bouvet J.. 1991. Universal primers for amplification of three non-coding regions of choroplast DNA. Plant Molecular Biology 17: 1105-1109.[CrossRef][Web of Science][Medline]
Thompson J. D. Gibson T. J. Plewniak F. Eanmougin F. Higgins D. G.. 1997. The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 25: 4876-4882.
Van Den Borre A. Watson L.. 1997. On the classification of the Chloridoideae. Australian Systematic Botany 10: 491-531.[CrossRef][Web of Science]
Vegetti A. C.. 1986. Contribución al conocimiento de las inflorescencias en Chloris y Cynodon (Poaceae). Kurtziana 18: 109-120.
Vollbrecht E. Springer P. S. Goh L. Buckler E. S. Martienssen R.. 2005. Architecture of floral branch systems in maize and related grasses. Nature 436: 1119-1126.[CrossRef][Medline]
Wiens J. J.. 2001. Character analysis in morphological phylogenetics: problem and solutions. Systematic Biology 50: 689-699.
Yoder A. D. Irwin J. A. Payseur B. A.. 2001. Failure of the ILD to determine data combinability for slow Loris phylogeny. Systematic Biology 50: 408-424.[Web of Science][Medline]
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Reinheimer, F. O. Zuloaga, A. C. Vegetti, and R. Pozner Diversification of inflorescence development in the PCK clade (Poaceae: Panicoideae: Paniceae) Am. J. Botany, March 1, 2009; 96(3): 549 - 564. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
