| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Structure and Development |
New York Botanical Garden, Bronx, New York 10458 USA
Received for publication February 28, 2003. Accepted for publication July 1, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: androecium • floral development • loss and suppression • symmetry • Vochysiaceae
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Stafleu, 1952
). A second west African genus, Korupodendron (Litt and Cheek, 2002
), was discovered after this work was completed and is not included here.
|
|
|
; Stafleu, 1948
, 1952
, 1953
, 1954
; Keay and Stafleu, 1953
; Kawasaki, 1998
) indicate that the position of the fertile stamen varies by genus and is correlated with petal number (Fig. 2). According to these workers, genera with more than one petal (Vochysia, Salvertia, and Erismadelphus) have the stamen in front of a petal, whereas those with only one petal (Qualea, Ruizterania, Callisthene, and Erisma) have the stamen in front of the fifth sepal. Staminodes, according to Stafleu (1952)
, are always antepetalous.
Stafleu's (1952)
discussion of the androecium in Vochysiaceae implies that the variability in the position of the fertile stamen is the result of reduction from two complete whorls of stamens to a single stamen. In the case of antepetalous stamens, an inner whorl stamen was retained, whereas in the case of antesepalous stamens, an outer whorl stamen was retained. This explanation is unsatisfactory for two reasons. First, there is little supporting evidence. According to Stafleu, species with an antesepalous fertile stamen and antepetalous staminodes (species of Qualea, Ruizterania, and Erisma) are examples in which stamens from two whorls are present in one flower. However, the position of staminodes relative to perianth parts is often difficult to determine based on gross morphological examination, and even the identity of these organs as staminodes and/or reduced petals may be obscure. Moreover, Myrtaceae (Myrtales), the proposed closest relatives of the family (e.g., Chase et al., 1993
; Conti et al., 1996
; Savolainen et al., 2000
; Soltis et al., 2000
), are not useful as a point of reference for understanding floral organization, as they have actinomorphic flowers generally with a proliferation of stamens.
The second and more significant concern that arises from Stafleu's discussion (Stafleu, 1952
) is the implication that the single fertile stamen is not homologous across the family. Both stamen positions are found in each of the two tribes (Fig. 2), thus if both tribes are monophyletic Stafleu's explanation would require four different reduction events (Fig. 3). Defined as "one fertile stamen," this character would almost certainly be a synapomorphy for Vochysiaceae in any phylogenetic analysis. This simplistic character definition would, however, obscure the fact that in different genera the stamen was derived independently. Under these circumstances, the single fertile stamen could not be considered one character shared by all the genera.
Alternative hypotheses to explain this character are (1) there is a developmental/genetic switch that toggles between antesepalous and antepetalous and (2) all primordia are initiated in the same position, but in one-petaled taxa the stamen is displaced during development. We do not, at the moment, have the molecular tools to examine the first hypothesis. The second hypothesis is based on the conjecture that the primordium of the fertile stamen originates between the gynoecium and the petal. Space-packing requirements might then result in the displacement of the stamen to an antesepalous position (Fig. 4). The second hypothesis is the subject of this study.
|
described it as antepetalous, as in other multi-petaled taxa, whereas Kopka and Weberling (1984)
reported it as antesepalous. Second, although the stamen of Erisma has consistently been described as antesepalous (Warming, 1875
; Stafleu, 1954
; Kopka and Weberling, 1984
; Kawasaki, 1998
), our own examination of mature buds and open flowers indicated otherwise.
This study was designed to determine where stamen primordia are initiated and how many are formed in one-petaled and multi-petaled taxa. This information is used (1) to clarify the position of the fertile stamen, particularly in Erisma and Erismadelphus, and (2) to determine the number of reduction events necessary to explain the distribution of fertile stamen positions in the family and thus to address the issue of floral homologies. Floral structure in the genera of Vochysiaceae is described in Litt and Stevenson (2003)
.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). With the exception of those of Erismadelphus, all were fixed in the field in formalin : propionic acid : 50% ethanol (FPA, 1 : 1 : 18 v/v) and stored in 70% ethanol. Buds of Erismadelphus were obtained from herbarium material, rehydrated in 10% Aerosol OT, and stored in FPA. For Qualea, Callisthene, Erisma, and Erismadelphus and for organogenesis in Vochysia, all species for which appropriate developmental stages were available were used. For serial sections of Vochysia, species were chosen from the available material to represent as many of Stafleu's three taxonomic sections and eight subsections as possible (Stafleu, 1948
)(see Litt and Stevenson, 2003
for details).
Appropriate material of Ruizterania and Salvertia was not available. A reevaluation of generic limits in Vochysiaceae is underway at this point, and preliminary analyses of DNA sequence data and morphological features indicate that Ruizterania may be nested within Qualea, from which it was segregated by Marcano-Berti (1969)
. The data also indicate that the monotypic Salvertia may not be distinct from Vochysia. More important, these genera do not differ from Qualea and Vochysia, respectively, in any characters that affect the patterns investigated in this study. Thus we assume that our observations of floral development and vasculature in Qualea and Vochysia also apply to Ruizterania and Salvertia respectively.
Material for scanning electron microscopy (SEM) work, undertaken to investigate organ initiation, was dissected and dehydrated in an ethanol/acetone series, critical-point dried, and coated with gold/palladium. Images were produced with a JEOL 5410LV at 5 kV.
Material for serial sectioning to examine vasculature was dehydrated in an ethanol/toluene series and embedded in paraffin. Relatively mature buds were used to ensure as much development of the vasculature as possible. Specimens that were difficult to section were exposed at the tip in their paraffin blocks and soaked for at least 1 wk in 2 : 5 glycerin : ethanol (70%) with acetic acid added to 10% of the total volume. Microtome sections were made at 10–15 µm. Those that were particularly rich in tannins were treated for up to 2 h with Stockwell's bleach (Schmid, 1977
). All were stained with Johansen's safranin in 50% ethanol and counterstained with either chlorazol black E or fast green followed by orange G.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
. In all Vochysiaceae the floral apex is concave, therefore floral organs in general develop underneath (equivalent to "in front of" or "adaxial to") the previously initiated organs. Descriptions of floral vasculature are based on serial sections, presented in acropetal sequence.
Vochysia
All species of Vochysia that were examined (Litt and Stevenson, 2003
) had three petals and two staminodes, the most common configuration in this genus (Stafleu, 1948
). Suitable material was not available of species with one or no petals. Little variability was found among the species for Vochysia used in this study.
Organ initiation (V. tucanorum, V. pumila, V. elliptica)(Figs. 5–8)
Three petal primordia are initiated on the flanks of the developing floral cup. The first two primordia appear between the third and fifth and the first and third sepals. These are followed by the third petal primordium, which arises between the second and fifth sepals (Fig. 5). A stamen primordium is then initiated directly below (adaxial to) each of these (Figs. 6, 7). The order of initiation could not be determined, but the central one, which becomes the fertile stamen, elongates rapidly, becoming noticeably larger than the three petals and two staminodes. The six organs (three petals and three stamens) are squeezed into the area defined by the floral cup, and although their bases are not affected, the distal portions shift to fill the space (Fig. 8). At this stage neither the style nor the filament has elongated. The style shows some growth prior to anthesis, but substantial elongation of both structures occurs at anthesis.
|
|
|
Whereas five petal primordia are always initiated, this is not the case with stamen primordia. In some buds, five may be initiated, but it is more common to see three, directly in front of (under or adaxial to) the petal primordia (between sepals three and five, one and three, and two and five). The stamen primordium that is directly in front of the full-sized petal develops into the fertile stamen. The stamen primordia of E. bracteosum appear to develop as downward elongations of the petal primordia rather than as separate discrete mounds (Figs. 17, 18). In E. floribundum, the primordia of the staminodes forms in this way, but not the primordium of the fertile stamen. This first appears as a discrete mound below the primordium of the major petal.
At maturity, Erisma flowers always possess one fully developed petal and one fertile stamen; however, the number of accessory organs varies. The number of lateral stamen primordia that are initiated seems to vary, as can the number of petal primordia that abort. As a result, the identity of a structure at maturity can only be determined with certainty when both a staminode and a rudimentary petal (or aborted vestige) are present at one position (Fig. 19). These pairs occur most often directly adjacent to the petal/fertile stamen pair (between sepals one and three and sepals two and five)(Fig. 2). If there is only one accessory organ, as occurs frequently on either side of sepal four, it cannot be determined to which whorl it belongs. Nonetheless, it may be reasonable to assume that these are most often rudimentary petals, because petals are always initiated in these positions, whereas stamens are much less frequently so.
The petal may grow relatively large in the bud, wrapping around the stamen and gynoecium and becoming highly wrinkled and folded. The filament and style elongate to varying degrees at anthesis with considerable twisting in one or both. This twisting obscures the position of attachment of the stamen.
Vasculature (E. bracteosum, E. uncinatum, E. floribundum, E. japura)(Figs. 27–29)
Vasculature to the petal and fertile stamen is similar to that seen in Vochysia. Although some weakly developed vascular tissue was observed in the rudimentary petals and staminodes of some specimens, this vasculature was isolated and not connected with the vascular bundles that supply the calyx (Figs. 27–29). It is interesting to note that procambial differentiation is most likely bidirectional because organ development is arrested with the procambium and vasculature remaining in the early stages of differentiation.
|
Organ initiation
Immature material was not available for studies of early development, but dissections of mature buds of Erismadelphus exsul show that the stamen is in front of a petal, as described by Keay and Stafleu (1953)
. Staminodes are antepetalous as well.
Vasculature (Erismadelphus exsul)
Vasculature is similar to that seen in Vochysia and Erisma. Each staminode is supplied by a branch of the bundle that also supplies the petal directly behind it, as in Vochysia.
Qualea
The species of Qualea examined in this study vary in the number of staminodes and rudimentary petals they possess. As in Erisma, the specific identity of structures is often difficult to determine from mature material. The observed configuration of these structures in each species is noted in the descriptions of development and vasculature. Preliminary floral dissections showed that at maturity the fertile stamen is not directly in front of the petal (Fig. 36), as it is in Vochysia, Erisma, and Erismadelphus, and that the position of these two organs relative to each other varies slightly among species.
|
1. Qualea parviflora, Q. grandiflora, Q. multiflora (Figs. 20–24)—The flowers of these species, which are probably closely related (Stafleu, 1953
; Litt, 1999
), have no staminodes or reduced petals. Stages in petal initiation were observed only in Q. grandiflora, in which a petal primordium is initiated between sepals one and three and a second between sepals three and five (Fig. 20). The primordium between sepals three and five continues to develop, whereas the other disappears as the floral apex expands. Occasional flowers of several Qualea species have two petals, but the position of the second petal has not been documented.
Three stamen primordia are initiated (Fig. 21) in all three species, although the third primordium, which is much smaller, was not observed in all specimens. The position of the lateral primordia is difficult to determine conclusively because the sepals overlap extensively, the floral apex is irregular in shape, and there is only one petal primordium to use as a landmark (see legend for Fig. 21). The lateral primordia abort immediately and generally are not present at subsequent stages of development. Occasionally, one will persist as a protrusion on the floral cup near the base of the ovary (Fig. 23).
The primordium that will develop into the fertile stamen arises in an ambiguous position (Figs. 21, 22). It is partly under (adaxial to) the petal primordium but not directly so, as in Vochysia and Erisma (Figs. 7, 17). On the other hand, it is not directly in front of the fifth sepal, as reported in published floral descriptions (e.g., Warming, 1875
; Stafleu, 1953
). As the organs develop, there is some shifting of position within the area enclosed by the sepals; it appears, however, to be the petal that is displaced. The edge of the petal behind the enlarging fertile stamen is relatively displaced backwards as the floral cup expands; the petal appears to pivot on its edge in front of the third sepal. The petal is tangential to the ovary or nearly so, facing the stamen broadside, and its position in the bud is asymmetrical (Fig. 24).
In mature buds, the petal is highly wrinkled and folded, and in some species is large enough to wrap around the stamen and gynoecium twice. As in Vochysia and Erisma, substantial elongation of the filament and style does not occur until anthesis. At that time, there may be considerable twisting of the style and of the filament, particularly at its base.
2. Qualea mori-boomii (Figs. 25–26)—The flowers of this species have one pair each of rudimentary petals and staminodes (Fig. 36). Three petal primordia are initiated in sequence between sepals one and three, three and five, and two and five respectively (Fig. 25). The middle primordium (between sepals three and five), which becomes the full-sized petal, enlarges more rapidly than the others. Three stamen primordia are initiated lower on the sides of the floral cup (Fig. 26). The two lateral primordia, which develop into staminodes, form directly under the lateral petal primordia (between sepals one and three and sepals two and five). The middle primordium, which develops into the fertile stamen, arises nearly directly under (in front of) the middle petal primordium and is only very slightly offset if at all. However, the angle of emergence of the two organs differs; a line drawn to bisect the petal primordium will not bisect the stamen (Fig. 26), and the petal already appears slightly askew relative to the center of the flower.
Vasculature
Vasculature to the petal and the fertile stamen is somewhat variable in the six species examined. In all cases, and as in the other genera, the vascular tissue in the pedicel forms a ring surrounding the pith. Medullary bundles were not seen in any of the species examined. As in the other genera described, branches from the vascular cylinder enter the floral cup, and the remaining vascular tissue of the cylinder supplies the gynoecium. In all cases, the bundle supplying the petal also branches to supply small bundles that become lateral traces in sepal three on one side and sometimes sepal five on the other side. Only where noted below was the stamen bundle observed to branch before entering the stamen. The species differ in details of the origin of the stamen trace.
1. Qualea parviflora (Figs. 30–33)—No staminodes or reduced petals were observed in this species. As the vascular cylinder separates into bundles, one segment divides laterally into two bundles that will supply the stamen and the petal (Fig. 30). The bundle supplying the stamen appears to extend slightly laterally through the floral cup to enter the stamen (Figs. 31, 32). The petal trace is also displaced slightly in the floral cup, in the direction opposite the displacement of the stamen bundle (Figs. 31, 32).
|
for Q. rosea, but only one pair of structures of indeterminate identity was observed. Marcano-Berti (1989)
does not mention staminodes or rudimentary petals in his description of Q. mori-boomii; one pair of each was found in this study. As the vascular cylinder separates into discrete bundles, one large bundle, which will supply both the stamen and the petal, can be identified. Higher in the floral cup, two bundles branch off from the adaxial side of this bundle (Fig. 34). These anastomose and the resulting bundle develops laterally in the floral cup for a short distance and enters the stamen (Fig. 35). The remaining large bundle becomes the petal midrib (Figs. 35, 36). Staminodes/rudimentary petals were not visible in all specimens sectioned. When present, they did not appear to have vascular tissue (Fig. 36). However, their points of attachment to the floral cup alternate with the median vascular bundles of the sepals, supporting the hypothesis of a position alternate with the sepals.
3. Qualea lineata—Both staminodes and rudimentary petals were observed in this species, although Stafleu (1953)
reported only staminodes. The vasculature of this species is similar to that observed in the preceding two species, except that the large bundle that will supply both the petal and the fertile stamen is composed of numerous extremely small bundles. Several of these bundles separate from the adaxial side of this group and anastomose into one bundle that enters the stamen. Vascular tissue was observed in neither the staminodes nor the reduced petals.
4. Qualea dichotoma (Figs. 37–39)—No staminodes or rudimentary petals were observed in this species. As the vascular cylinder separates in the floral cup, one segment divides laterally into three bundles (Fig. 37). The two flanking bundles anastomose, and the resulting bundle extends laterally to enter the stamen (Fig. 38). The remaining large bundle, which supplies the petal, moves in the floral cup to become oriented at an angle to the ovary (Fig. 39) as in the other species.
|
Vasculature (C. major, C. fasciculata)
The pattern of vasculature to the petal and stamen in these two species is essentially identical to that seen in Q. mori-boomii and Q. rosea.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
) of reduction from two whorls. The other component of the hypothesis, that the fertile stamen is displaced in single-petaled species during floral development, is not supported by the data. In fact, in species of Qualea and most likely in Callisthene, the petal appears to be displaced during development (Figs. 23, 24).
Fertile stamen position
Erisma
In contradiction to published descriptions (Warming, 1875
; Stafleu, 1954
; Kopka and Weberling, 1984
; Kawasaki, 1998
), the fertile stamen of Erisma is antepetalous. Dissections of mature buds before elongation of the filament show that the stamen lies directly in front of the petal (Fig. 19). At anthesis, the filament twists at the base. Superficially, the stamen may appear to be attached adjacent to the petal. However, if the stamen is removed, the scar marking the point of attachment is seen directly in front of the petal. Further evidence comes from observations of floral development in Erisma bracteosum, in which the petal and stamen primordia appear as one radially elongated primordium forming the petal from the top portion and the fertile stamen from the bottom portion (Fig. 17). Moreover, the pattern of vasculature to the stamen and petal is identical to that seen in the unequivocally antepetalous Vochysia (Figs. 10–12, 27–28).
Qualea and Callisthene
As in Erisma, the twist at the base of the filament of most species of Qualea and Callisthene can lead an observer to misjudge the position of the stamen and conclude that it is antesepalous (e.g., Warming, 1875
; Stafleu, 1952
, 1953
). In fact the base of the stamen is not centered in front of the fifth sepal. Rather, it lies towards the edge of that sepal. The position at which the primordium is initiated also indicates that the stamen is not antesepalous (Figs. 21, 22, 26). Furthermore, the pattern of vasculature is fundamentally similar to that seen in Erisma, Vochysia, and Erismadelphus, in that the vascular supply to the stamen is continuous with the midrib of the petal (Figs. 30, 34, 37). Thus, the stamen can be considered offset from antepetalous rather than antesepalous. In the antesepalous condition, the vasculature of the stamen should be continuous with the median bundle of the associated (fifth) sepal.
Stafleu (1953)
said that staminodes in Vochysiaceae are always in front of petals, even in cases in which the fertile stamen is not. This study supports his observations. Callisthene has no staminodes, but in Qualea the erratic position is confined to the fertile stamen.
The species examined in this study differ in the exact position of the stamen. In the two species of Callisthene examined, as well as in Q. mori-boomii, Q. rosea, and Q. lineata, the stamen is little offset from antepetalous, whereas in Q. parviflora, Q. multiflora, and Q. grandiflora the offset is more substantial. This difference can be seen at the time of stamen initiation. In Q. grandiflora, Q. parviflora, and Q. multiflora, the stamen primordium is clearly not directly in front of the petal (Figs. 21, 22); however, the primordium of Q. mori-boomii is only slightly offset (Fig. 26). The greater degree of separation between the petal and stamen in Q. parviflora is also reflected in the vasculature. In this species the stamen and petal bundles separate laterally from a common supply, rather than radially into an inner and an outer bundle as in most of the other species (compare Figs. 30, 34, 38); they also separate farther down in the floral cup. However, in all cases anatomical data support the interpretation of the single fertile stamen as fundamentally antepetalous.
Phylogenetic implications
The results of this study have clarified and simplified the distribution of the stamen position character in the family; Vochysia, Salvertia, Erisma, and Erismadelphus have the antepetalous position and only Qualea, its segregate Ruizterania, and Callisthene differ in having an offset stamen. Preliminary analyses of DNA sequences and morphology (Litt, 1999
) strongly support these last three genera as a monophyletic group (hereinafter the "QRC" clade). Thus, the stamen position character can be explained by one reduction at the base of the family and one change of position within the family. Because only one reduction event is required, the single fertile stamen is homologous across Vochysiaceae, and this character is a synapomorphy for the family. The offset position is a synapomorphy of the QRC clade (Fig. 40).
|
; and A. Litt, personal observation). Other unifying characters, such as the capsular fruit, are likely to be plesiomorphic, but further analysis is needed. The studies reported here have uncovered additional differences between the putatively monophyletic QRC group and the rest of the family: the former are the only genera lacking medullary bundles and are of course the only genera with an offset fertile stamen and displaced petal. Thus the monophyly of the tribe Vochysieae is called into question.
As noted, evidence from sister-group analysis is not helpful in understanding the evolutionary history of the androecium of Vochysiaceae. Based on the phylogenetic position of Vochysiaceae within the Myrtales, the condition of having only one fertile stamen is an autapomorphy. Nonetheless it is interesting to note that obhaplostemony, the condition of having a single whorl of antepetalous stamens, is found scattered throughout Myrtales (Dahlgren and Thorne, 1984
; Ronse De Craene and Smets, 1995
). This study has shown that the Vochysiaceae can be considered fundamentally obhaplostemenous, with the androecium of the QRC group being a derived variant of that condition. Among families in which the first or only whorl of stamens is antepetalous, there are several, including Myrtaceae and Onagraceae of Myrtales, in which common petal/stamen primordia have been observed (Ronse De Craene et al., 1993
). In this study such primordia were observed in the two species of Erisma in which organogenesis was examined. However, in a broad examination of this feature, Ronse De Craene et al. (1993)
concluded that there is little evidence that these common primordia are of systematic value.
Loss vs. suppression
Tucker (1984
, 1988
, 1997
) and Basile and Basile (1993)
have addressed the significance of reduction of floral parts. Tucker (1988)
drew a distinction between loss and suppression; loss can be seen as the endpoint on a continuum of degrees of suppression, but the effects of loss and suppression on floral development can be different (Tucker, 1988
). When organs are suppressed, primordia are initiated, but at some stage they cease development. At maturity the flower may show little or no sign that the primordia ever existed; nonetheless, primordia can serve as placeholders in the development of remaining floral organs so that the spatial arrangement of the remaining organs is maintained. In contrast, when organs are truly lost, no primordia are initiated. Without a complete whorl of primordia to determine spacing, organs in that whorl or the next may arise in unexpected and sometimes unpredictable positions (Tucker, 1988
; Ronse De Craene and Smets, 1995
).
Tucker (1984
, 1997
) suggested that features of early floral development (organogeny) are more likely to be constant across broader taxonomic categories than features of middle or late development (organ growth and tissue differentiation, respectively). The latter are likely to vary among genera or species. Thus, because true loss of organs is manifest early in floral development, one might expect it to be stable across genera or an entire family. Suppression, on the other hand, can occur at a wide variety of stages during development and is likely to be variable even within genera.
Within Vochysiaceae, reduction in floral parts is a constant feature, but the numbers of those parts varies among and within genera and in some cases even within species or individuals. According to Tucker's hypothesis, this variability would implicate suppression as the important factor in organ reduction. Alternatively, loss and suppression may be viewed as part of a developmental continuum, differing only in degree. In this view, suppression may occur at many stages of development. When it occurs early enough, it is perceived as loss.
Androecium
In the androecium of all three genera examined, two stamens are completely lost, and two are suppressed to a greater or lesser extent. However, descriptions of monotypic Salvertia and some species of Vochysia indicate that they may possess up to four staminodes (Stafleu, 1948
), in which case, no stamens have been completely lost. In some Erisma flowers four stamen primordia also may be initiated, but this was not conclusively seen in any specimens. In Vochysia, Erisma, and some species of Qualea, two of the three primordia that are initiated become staminodes (the third becomes the fertile stamen). In other species of Qualea, suppression of the two lateral stamen primordia occurs immediately, and there are no staminodes.
Corolla
Salvertia and Erismadelphus have a complete corolla of five petals; all other genera have fewer. All species of Vochysia examined in this study have three petals, and in all species three petal primordia were initiated. In Vochysia, therefore, two petals are lost (possibly more in species with one or no petals). Qualea, Callisthene, Ruizterania, and Erisma have one petal. In Erisma five primordia were always initiated (Figs. 16–18), and thus only suppression is involved in corolla reduction.
Species of Qualea show both loss and suppression of petals. In Q. mori-boomii three primordia are initiated; thus, two petals are lost as in Vochysia. Development of the two lateral primordia is limited, and they form two threadlike rudiments (Fig. 36). In Q. grandiflora, only two primordia were observed (Fig. 20), thus an additional petal has been completely lost. One of the primordia disappears immediately as the floral apex expands; the other completes development. Thus although all Qualea species have one full-size petal, the two species for which clear observations could be made differ in the number initiated and therefore the number that can be considered lost. The timing of suppression of those that do not complete development also varies.
Thus of the three genera examined in this study, Vochysia has loss of two petals, Erisma has suppression of four to varying degrees, and Qualea has both loss and suppression. Because the number of fully developed petals varies from five to none within the family, differences in the pattern of loss and suppression might be expected among the genera. However, the pattern is not consistent: both Qualea and Erisma have one petal at maturity, but Erisma has five initiated, whereas Qualea has three (perhaps two in some cases). Furthermore, whereas Vochysia has three petals to the one of Qualea, both have three initiated (at least in some Qualea species).
Thus Vochysiaceae show both loss and suppression in both the corolla and the androecium, with less variability in the androecium. The variability among Qualea species in the number of petals initiated is not in complete accord with Tucker's (1984
, 1988
) hypothesis that events early in development are more likely to be constant within genera. However, floral development is flexible enough to provide examples contrary to any hypothesis, as also discussed by Endress (1999)
. Furthermore, a third petal primordium may exist so briefly in Q. grandiflora that it was not observed in this study. In the androecium of Qualea and in the corolla and androecium of the other genera examined, no variability was observed in the number of primordia initiated, as predicted by Tucker (1984
, 1988
). There is, however, considerable variation among the genera. This variation is less difficult to explain if loss and suppression are viewed as being parts of a developmental continuum; processes can move slightly forwards or backwards on the continuum without difficulty and without greatly perturbing the overall developmental process or mature floral structure.
Tucker (1988)
, Endress (1990)
, and Ronse De Craene and Smets (1995)
have noted that when loss occurs in the corolla whorl, the position of the stamens may become irregular. The complete loss in Qualea of two petals (in some cases possibly three) might then be implicated in the erratic position of the fertile stamen. However, staminodes of Qualea are antepetalous, and the stamen and staminodes of Vochysia are likewise unaffected by the loss of corolla parts. Furthermore, although the position of the fertile stamen in Qualea is unusual, it is reasonably predictable, being always more or less offset from directly in front of the single petal. Thus there are no observations that lead to an obvious and consistent developmental explanation for the offset position of the fertile stamen in Qualea. It does, however, provide a distinctive synapomorphy for the clade consisting of Qualea, Ruizterania, and Callisthene, in contrast to the strictly antepetalous position seen in the other genera of both tribes.
|
FOOTNOTES |
|---|
2 Current address: Yale University, Department of Molecular, Cellular, and Developmental Biology, P.O. Box 208104, New Haven, Connecticut 06520 USA (amy.litt@yale.edu)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Boesewinkel F. D. M. Venturelli 1987 Ovule and seed structure in Vochysiaceae. Botanische Jahrbücher für Systematik 108: 547-566
Chase M. W. et al 1993 Phylogenetics of seed plants: an analysis of nucleotide sequences from the plastic gene rbcL. Annals of the Missouri Botanical Garden 80: 528-580[CrossRef][ISI]
Conti E. A. Litt K. J. Sytsma 1996 Circumscription of the Myrtales and their relationships to other rosids: evidence from rbcL sequence data. American Journal of Botany 83: 221-233[CrossRef][ISI]
Dahlgren R. R. F. Thorne 1984 The order Myrtales: circumscription, variation, and relationships. Annals of the Missouri Botanical Garden 71: 633-699[CrossRef][ISI]
Dumortier B. C. 1829 Analyse des familles des plantes avec l'indication des principaux genres qui s'y rattachent. Vochysiaceae. J. Casterman, Tournay, France
Endress P. K. 1990 Patterns of floral construction in ontogeny and phylogeny. Biological Journal of the Linnean Society 39: 153-175[CrossRef]
Endress P. K. 1999 Symmetry in flowers: diversity and evolution. International journal of Plant Sciences 160: (Suppl.) S3-S23[CrossRef][ISI][Medline]
Kawasaki M. L. 1998 Systematics of Erisma. Memoirs of the New York Botanical Garden 81: 1-40
Keay R. W. J. F. A. Stafleu 1953 Erismadelphus. Mededelingen van het Botanisch Museum en Herbarium van de Rijksuniversiteit te Utrecht 114: 594-599
Kopka S. F. Weberling 1984 Zur Morphologie und Morphogenese der Blüte von Vochysia acuminata Bong. subsp. laurifolia (Warm.) Stafleu (Vochysiaceae). Beiträge zur Biologie der Pflanzen 59: 273-302
Litt A. 1999 Floral morphology and phylogeny of Vochysiaceae. Ph.D. dissertation, City University of New York, New York, New York, USA
Litt A. M. Cheek 2002 Korupodendron songweanum, a new genus and species or Vochysiaceae from West-Central Africa. Brittonia 54: 13-17
Litt A. D. W. Stevenson 2003 Floral development and morphology of Vochysiaceae I. The structure of the gynoecium. American Journal of Botany 90: 1533–1547
Lleras E. 1978 Trigoniaceae. Flora Neotropica Monograph 19: 1-73
Marcano-Berti L. 1969 Un nuevo genero de las Vochysiaceae. Pittieria 2: 3-28
Marcano-Berti L. 1989 Vochysiaceae: novidades y correciones. Pittieria 18: 5-14
Ronse De Craene L. P. D. Clinckemaillie E. Smets 1993 Stamen-petal complexes in Magnoliatae. Bulletin du Jardin Botanique National de Belgique 62: 97-112[CrossRef]
Ronse De Craene L. P. E. F. Smets 1995 The distribution and systematic relevance of the androecial character oligomery. Botanical Journal of the Linnean Society 118: 193-247[CrossRef]
Savolainen V. M. W. Chase S. B. Hoot C. M. Morton D. E. Soltis C. Bayer M. F. Fay A. Y. de Bruijn S. Sullivan Y.-L. Qiu 2000 Phylogenetics of flowering plants based on combined analysis of plastid atpB and rbcL gene sequences. Systematic Botany 49: 306-362
Schmid R. 1977 Stockwell's bleach, an effective remover of tannin from plant tissues. Botanische Jahrbücher für Systematik 98: 278-287
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]
Stafleu F. A. 1948 A monograph of the Vochysiaceae. I. Salvertia and Vochysia. Recueil des Travaux Botaniques Néerlandais 41: 398-540
Stafleu F. A. 1952 A monograph of the Vochysiaceae. II. Callisthene. Acta Botanica Neerlandica 1: 222-242
Stafleu F. A. 1953 A monograph of the Vochysiaceae. III. Qualea. Acta Botanica Neerlandica 2: 144-217[ISI]
Stafleu F. A. 1954 A monograph of the Vochysiaceae. IV. Erisma. Acta Botanica Neerlandica 3: 459-480
Tucker S. C. 1984 Origin of symmetry in flowers. In R. A. White and W. C. Dickinson [eds.], Contemporary problems in plant anatomy, 351–394. Academic Press, New York, New York, USA
Tucker S. C. 1988 Loss versus suppression of floral organs. In P. Leins, S. C. Tucker, and P. K. Endress [eds.], Aspects of floral development, 69–82. J. Cramer, Berlin, Germany
Tucker S. C. 1997 Floral evolution, development, and convergence: the hierarchical-significance hypothesis. International Journal of Plant Sciences 158: (Suppl.) S143-S161[CrossRef]
Warming E. 1875 Vochysiaceae. In C. F. P. Martius [ed.], Flora brasiliensis, vol. 13, 17–116. R. Oldenburg,Munich, Germany
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP |
= QRC clade. Abbreviations: circum. = circumferential; incl. phloem = included phloem; indehisc. = indehiscent; seed wing comp. = seed wing composition; unilat. = unilateral
