| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Development and Morphogenesis |
Department of Botany, University of Toronto, 25 Willcocks Street, Toronto, Ontario, Canada M5S 3B2
Received for publication February 26, 2002. Accepted for publication August 27, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: allometry • constraint • enantiostyly • floral development • inflorescence architecture
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Guerrant, 1989
; Wake, 1991
; Diggle, 1992
; Richards and Barrett, 1992
). However, constraints in the structure, composition, and dynamics of developmental systems can place limits on character evolution (Maynard Smith et al., 1985
). For example, a lack of heritable variation in left-right asymmetries can constrain the evolution of form. In Drosophila, selection experiments on body asymmetries, including eye facet number (Maynard Smith and Sondhi, 1960
), eye size (Coyne, 1987
), wing folding behavior (Purnell and Thompson, 1973
), and thoracic bristle number (Tuinstra, de Jong, and Scharloo, 1990
), have shown that while selection increased the degree of asymmetry, it had no effect on the direction of the asymmetry. These results demonstrate that fluctuating asymmetries may have a heritable component, but that the direction of asymmetry is not heritable, thus limiting the evolution of asymmetric morphologies.
Floral asymmetries often involve radial or dorsiventral asymmetries in perianth structures or sexual organs (for a discussion of the terminology of floral symmetry see Giurfa, Dafni, and Neal [1999]
). The patterns of zygomorphy that occur in flowers such as Asterids are likely constrained by early developmental patterns such as petal number, the overall orientation of the bud relative to the stem, and the orientation of dorsiventral differentiation (Ree and Donoghue, 1999
). Because of these developmental constraints, only a limited number of forms of asymmetry in the Asterids are possible.
Enantiostylous flowers exhibit another form of asymmetry, a medial-lateral asymmetry (Fig. 1). In enantiostyly, the style of a flower is deflected either to the left (left-styled) or to the right (right-styled) of the floral axis and is therefore a floral polymorphism. Enantiostyly has evolved in at least ten angiosperm families in both the monocotyledons and dicotyledons (Jesson, 2002
). Despite the multiple origins of enantiostyly in flowering plants, several similarities in floral morphology occur among unrelated enantiostylous species. For example, enantiostyly is often associated with heteranthery, the specialization of anthers into brightly colored feeding anthers and a cryptically colored pollinating anther (Graham and Barrett, 1995
). The pollinating anther is deflected in the opposite direction from the style and likely plays a role in precise pollen transfer. The majority of enantiostylous species exhibit monomorphic enantiostyly (at least 25 genera from ten families; Jesson, 2002
), in which left- and right-styled flowers occur on the same individual. Dimorphic enantiostyly, in which individuals are genetically determined to be entirely left- or right-styled (Jesson and Barrett, 2002a
), is only reported in five species from three monocotyledon families.
|
; Jesson and Barrett, 2002b
). Observations of the stigma and the pollinating anther contacting opposite sides of a pollinator's body have led to the suggestion that enantiostyly functions to effect proficient pollination between flowers of opposite style orientation (Dulberger, 1981
). However, if a plant has left- and right-styled flowers open simultaneously (as in monomorphic enantiostyly), a pollinator visiting the two opposite forms on the same individual can cause geitonogamous self-pollination. Experimental evidence has shown that selfing rates are significantly lower in plants manipulated to be entirely left- or entirely right-styled than those possessing both floral forms (Barrett, Baker, and Jesson, 2000
; Jesson and Barrett, 2002b
). Thus, it is unclear why a consistent direction of stylar deflection, such as that found in dimorphic enantiostyly, has not evolved more often, given the functional advantages that this polymorphism provides by reducing the costs associated with selfing (Jesson and Barrett, 2002b
).
It has been suggested that selection on the direction of stylar deflection on flowers of an individual may be constrained developmentally or structurally, perhaps by a lack of positional information within the bud (Barrett, Jesson, and Baker, 2000
). We were therefore interested in examining how developmental processes may constrain the evolution of enantiostyly, particularly dimorphic enantiostyly. Specifically, we were interested in addressing four questions related to the developmental biology of enantiostyly: (1) For species with monomorphic enantiostyly, what is the arrangement of left- and right-styled flowers on inflorescences? (2) In species where the direction of stylar bending is fixed (i.e., dimorphic enantiostyly), when are the differences between floral forms first visible and how does this compare to species with monomorphic enantiostyly? (3) How early in floral development does the style deflect from the central axis in different enantiostylous species? (4) What are the developmental patterns of the gynoecium (particularly the style) and androecium (the pollinating and feeding anthers) for enantiostylous species occurring in unrelated families?
We investigated the development of enantiostyly by comparing inflorescence architecture and floral organ development among a range of enantiostylous species. We chose to take a broad, comparative approach by examining multiple lineages in which enantiostyly occurs, rather than focusing on a smaller species sample. To examine whether the direction of style deflection in monomorphic enantiostylous plants was predictable, we compared the arrangement of left- and right-styled flowers within inflorescences of three monocotyledon species (Heteranthera mexicana and Monochoria korsakowii [Pontederiaceae] and Cyanella lutea [Tecophilaeaceae]) and one dicotyledon species (Solanum rostratum [Solanaceae]). Comparative evidence indicates that enantiostyly has separate evolutionary origins in each of these families (Kohn et al., 1996
; Soltis, Soltis, and Chase, 1999
; summarized in Fig. 2). To investigate whether asymmetry in sex organs occurs early in development, we compared related dimorphic and monomorphic enantiostylous species of Heteranthera in the Pontederiaceae (Heteranthera multiflora and Heteranthera limosa, respectively) as well as in two unrelated species in which the direction of stylar deflection within the flower was predictable (dimorphic enantiostylous Wachendorfia paniculata [Haemodoraceae] and monomorphic enantiostylous S. rostratum [Solanaceae]). To examine the role of heterochrony in the timing of floral organ differentiation, particularly in the feeding and pollinating anthers, we compared growth allometries of sex organs in nine enantiostylous taxa representing five distinct lineages (Jesson, 2002
).
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
For all species, we noted the organization of the organs of the mature flower (bract, sepal, and petal, or tepal, stamen, and gynoecium), as well as the structural features of stamen dimorphism, the position of the pollinating and feeding anthers, the position of stamen insertion, and any patterns of perianth fusion.
Inflorescence patterning
To determine whether patterning of left- and right-styled flowers within an inflorescence was predictable with respect to inflorescence architecture, we recorded the direction of style deflection and flower position in glasshouse populations of three enantiostylous species: M. korsakowii, H. mexicana (Pontederiaceae), and S. rostratum (Solanaceae), and two field populations of C. lutea (Tecophilaeaceae). See http://ajbsupp.botany.org/v90/ for details of collection localities and number of plants measured.
Inflorescences of S. rostratum are unbranched scorpioid cymes, whereas flowers of C. lutea, M. korsakowii, and H. mexicana are all borne on indeterminate racemes. In all species, we recorded the direction of style deflection and flower position for the first five flowers produced on an inflorescence. For glasshouse populations, we noted the direction of the style for the first five flowers in anthesis, the date of flowering, and the nodal position of the flower on the plant (1 = bottom flower, which is the first to flower). For C. lutea, we sampled two natural populations and recorded the direction of deflection of all newly opened flowers on a plant, their nodal position, and the number of flowers previously opened.
We used logistic regression (see Neter, Wasserman, and Kutner [1990]
) to examine differences in left- and right-styled flower patterning on an inflorescence. We tested all predictors and interactions and removed terms via stepwise deletion. Tests of significance were obtained by comparing the changes in the full main effects model to a model with the main effects removed or interactions added (equivalent to type III sums of squares). All analyses were conducted using S-plus 4 (Mathsoft, 1997
).
For glasshouse populations of M. korsakowii, H. mexicana, and S. rostratum, the model response was the direction of stylar bending (either left or right) on flowers at nodal positions 2–5. The predictors in the model were nodal position (2–5), date of flowering, and direction of the first emerging flower on an inflorescence (nodal position 1). In field populations of C. lutea, the flower located at the second nodal position was most frequently open, and hence, its direction of deflection was known. To examine if the direction of deflection of a flower could be predicted by the direction of other flowers on the inflorescence, we considered the flower at the second node as a predictor variable. Other predictor variables also included the nodal position of a flower (i.e., the total number of previous flowers produced). The direction of flowers at nodal positions 1, 3, 4, and 5 (if known) were considered response variables.
Relative growth rates
To investigate the development of flower structure and the relative growth stages of sexual organs, we dissected 15–70 buds per species (see http://ajbsupp.botany.org/v90/) ranging from 1 to 20 mm in length using a Stemi Sv8 (Zeiss, Oberkochen, Germany) dissecting microscope. To examine the initiation of stylar deflection and the relative growth rates of the floral organs, we measured bud length, style length, and filament length of one pollinating stamen and one feeding stamen, the length of one pollinating and feeding anther, and the distance from the stigma to the floral axis. Measurements were made with an ocular micrometer (to the nearest 0.1 mm) or with digital calipers.
To investigate differences in allometric growth rates, we regressed the length of the style and the stamen (filament length + anther length) against the length of the floral bud. All data were log transformed prior to analysis. We used linear least squares regression to calculate the relative growth rates of the different organs. We used analysis of covariance (Sokal and Rohlf, 1995
) to compare differences in the slopes and intercepts of the length of pollinating and feeding stamens for all species.
The initiation of left-right asymmetries
To investigate asymmetries during floral development, we used scanning electron microscopy to examine buds of two monomorphic (H. limosa and S. rostratum) and two dimorphic enantiostylous species (H. multiflora and W. paniculata). We compared these species because in H. multiflora, W. paniculata, and S. rostratum (see RESULTS), we were able to predict the direction of style deflection of the flowers (either left- or right-styled). To examine differences in dimorphic and monomorphic enantiostyly, bud development in H. multiflora was compared to H. limosa. Inflorescences and flower buds were fixed in FAA and vacuum infiltrated overnight (138 kPa). Specimens were dehydrated in a graded ethanol series, CO2 critical point dried using an Autosamdri 814 Critical Point Dryer (Tousimis Research, Rockville, Maryland, USA), gold coated using a Cressington 108 Sputter Coater (Cressington Scientific Instruments, Cranberry Township, Pennsylvannia, USA), and viewed in a Hitachi scanning electron microscope (Model S-2500, Hitachi, Tokyo, Japan) at 10 kV. We viewed 3–12 buds per floral stage across at least three floral stages for each species to compare qualitative trends. The buds viewed had been measured previously in the allometric study.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
Dilatris corymbosa has two dark-colored pollinating anthers and a single, orange feeding anther that is always adjacent to the tepals furthest from the stem. Wachendorfia paniculata does not have differentiated feeding and pollinating stamens. All three stamens have anthers of equal size and color; however, one stamen has a longer filament and is always in a position reciprocal to the style. This stamen is positioned nearest to the stem of the inflorescence. There is also no division of labor in the anthers of P. lanuginosum, because flowers only have one stamen. Solanum rostratum, like Heteranthera spp. and C. lutea, has dimorphic stamens. In S. rostratum, the single pollinating anther is dark yellow or brown and is fused to the corolla in a ventral position in the flower. The four feeding anthers are bright yellow and are located in a cone in the center of the flower.
Inflorescence patterning
Among the four monomorphic enantiostylous species investigated there were striking differences in the extent to which stylar orientation could be predicted by flower position within inflorescences. This variation ranged from almost complete predictability to an apparently random arrangement of left- and right-styled flowers within an inflorescence. The patterning of left- and right-styled flowers on an inflorescence in S. rostratum is predictable (Fig. 4; Table 1) as the direction of the style is always deflected towards the stem. Shoot growth patterns of S. rostratum are monochasial and determinate. Inflorescences are quickly overtopped by one lateral shoot from the node below. The position of this lateral developing shoot alternates between distal axils on the left and right sides of the inflorescence. Flowers that originate from a left axil lateral shoot are right-styled, while flowers originating from a right axil lateral shoot are left-styled. Thus, the direction of deflection alternates along the inflorescence cyme and along the vegetative stem of the plant. Observations of stylar deflection for the first five developing flowers on an inflorescence demonstrated this pattern of alternation between left- and right-styled flowers. However, the first flower produced was either left- or right-styled. This led to a highly significant interaction between the direction of the focal flower and the nodal position of the flower (Table 1). Because of the monochasial pattern of inflorescence architecture, the ratio of flowers on an individual was 1 : 1. Of the 30 plants examined, the proportion of right-styled flowers on a plant was 0.53 (1 SE = 0.05; N = 150 flowers).
|
|
In C. lutea, the direction of stylar deflection qualitatively appeared to alternate along successive floral nodes (Fig. 4). However, the interaction between the direction of the focal flower and the nodal position of the flower accounted for 0.02 deviance units out of a total of 142.31, thus explaining only 0.001% of the total variance (Table 1). Hence, in this species there is only a limited amount of patterning of left- and right-styled flowers within an inflorescence. In C. lutea, there was no effect of population or plant on the proportion of left- and right-styled flowers within an inflorescence. The mean proportion of right-styled flowers on a plant was 0.57 at Brandweg (1 SE = 0.05, N = 320 flowers) and 0.41 at MacGregor (1 SE = 0.05, N = 195 flowers).
In H. mexicana, we found no significant predictors for the direction of stylar deflection (Fig. 4; logistic regression results not shown). As with the other species, the proportion of left- and right-styled flowers on a plant was near equal. The proportion of right-styled flowers on a plant in H. mexicana was 0.52 (1 SE = 0.02; N = 250 flowers). Five plants of H. mexicana were observed to have a single flower with a straight style.
Allometric growth patterns
The pattern and timing of floral development varies considerably between species. In W. paniculata and D. corymbosa, bending was first observed at a bud length of 5 and 7 mm, respectively (approximately 25% and 40% of the mature length, respectively; Fig. 5). In P. lanuginosum, this occurred at a length of approximately 9 mm (45% of mature flower length; Fig. 5). In contrast, no evidence of stylar bending was found prior to anthesis in C. lutea, H. limosa, H. rotundifolia, H. multiflora, M. australasica, and S. rostratum.
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Comparisons of the development of enantiostyly
Enantiostylous species from diverse evolutionary origins exhibit obvious differences in the structure of their mature flowers. For example, in some species there is a morphological division of anthers into feeding and pollinating functions (e.g., S. rostratum), while in others there are no differences in anther function or structure (M. australasica) or there is only one stamen (P. lanuginosum). Analysis of the growth allometries of these species suggest that the heterochronic processes of initiation and termination of stamen growth (rather than differences in growth rate) account for the structural variation of stamens, although this was not apparent in scanning electron micrographs of W. paniculata and S. rostratum (results not shown). It is possible that nonlinear growth early in development accounts for this difference, or that perhaps small, undetected changes become augmented later in the developmental process.
There are also striking differences in the role of inflorescence architecture on the arrangement of left- and right-styled flowers in enantiostylous species. In S. rostratum, the direction of a flower is almost completely predictable once the direction of the style of the first flower is determined, whereas in H. mexicana, the position of left- and right-styled flowers appear to be randomly determined. In C. lutea and M. korsakowii, inflorescence architecture and environmental factors can account for some of the variation in the position of left- and right-styled flowers, as the direction of stylar deflection is not completely predictable. This suggests that the cues to stylar bending may be determined differently among species of independent evolutionary origins.
Broad similarities also exist among enantiostylous species in the structure and development of their flowers. In all species examined enantiostyly is initiated relatively late in floral development. In W. paniculata, D. corymbosa, and P. lanuginosum, bending of the style was observed in the floral bud when buds were approximately 25–45% of full size. In all other species studied, the style was not deflected in the bud, suggesting that stylar bending occurs just prior to or simultaneous with the opening of the bud at anthesis. The expression of enantiostyly late in floral development has been noted in other enantiostylous taxa. In Monochoria hastata (syn. M. hastaefolia; Pontederiaceae) and Cassia didymobotrya (Leguminosae), the bending of the style takes place 6–12 h prior to anthesis (Iyengar, 1923
; Dulberger, 1981
, respectively). The bending of the pollinating stamen also occurs during this time in M. hastata and M. vaginalis (Iyengar, 1923
). It is possible that developmental constraints, such as limited space within a bud, lead to the consistent late appearance of enantiostyly during development. Thus, enantiostyly may only be expressed when the tight packaging that occurs in early bud development is released during floral expansion.
A left- or right-deflection in position of the pollinating stamen was not seen in the bud in any species examined, and thus, must occur very late in development. Despite the reciprocal positioning of the style and a pollinating anther at anthesis, the initiation of androcial and gynoecial structures occur at different developmental stages. In all species examined, the stamens were initiated and differentiated into anther and filament well before gynoecial differentiation into the ovary, style, and stigma. Thus, the reciprocal positioning of the stigma and anther that occurs in many unrelated taxa is likely to be due to convergent selection, rather than through allometric processes in early organ development. This convergent selection on stigma and anther height has been documented in other species. For example, Hufford (1988a
, b)
examined diversification of organ lengths in the genus Eucnide. He found that while stigma and anther heights were always similar in mature organs, the growth rates and timing of organ initiation were very diverse within this clade. This indicates that similarities in the mature structures of flowers can result from quite distinct developmental processes.
Axes of asymmetry are present soon after organ initiation in H. limosa, H. multiflora, S. rostratum, and W. paniculata. However, the axis of asymmetry lies in the dorsiventral plane, rather than in the medial-lateral plane. In H. multiflora, H. limosa, and S. rostratum, the pollinating anther is consistently located on the ventral side of the flower, while in W. paniculata, it is located on the dorsal side. The early initiation of dorsiventral asymmetries is common in many zygomorphic species, and in many species, zygomorphy is expressed at floral initiation (Endress, 1999
). In other species, dorsiventral asymmetries can be expressed later in development due to differences in heterochrony or the late development of structures such as glands or spurs (for review see Tucker [1999]
).
Despite the early initiation of dorsal-ventral asymmetry, we found no consistent differences in left-right axes early in development. In buds of H. multiflora that were greater than 90% of mature length, the stigma rotated so that the tilt of the stigma was away from the direction of stylar bending. This was the earliest indication that we noticed for the direction of stylar bending. Thus, despite the genetic determination of dimorphic enantiostyly in H. multiflora (Jesson and Barrett, 2002a
), we found no differences between the style morphs until very late in development. Tucker (1999)
also noted that stylar bending in several enantiostylous species in the Leguminosae (e.g., C. didymobotrya and C. fasciculata) occurred late in development. This is not necessarily the case in other enantiostylous species. Iyengar (1923)
found that both M. hastata and M. vaginalis exhibited one or two unequal spurs on the filaments of all stamens, with the largest spur always occurring on the side of deflection of the pollinating stamen. These spurs were unequal before bending of the stamen and were hence an early indicator of the direction of stylar bending. We did not observe any spurs in M. australasica, although this species does not possess a pollinating anther. Thus, it would be interesting to contrast floral development among species of Monochoria as this may reveal other differences in the timing of initiation of asymmetries.
Constraints on the evolution of monomorphic and dimorphic enantiostyly
Dimorphic enantiostyly appears to have evolved from monomorphic enantiostyly only two or three times (Barrett, Jesson, and Baker, 2000
). Within the Pontederiaceae, we found no significant differences in development between dimorphic enantiostylous H. multiflora and monomorphic enantiostylous H. limosa. While the type of enantiostyly differs at the plant level in these two species, it remains very similar at the developmental level and in the mature flower structure. In H. multiflora, it appears that fixing the direction of stylar deflection does not require separate developmental processes. In this way, enantiostyly may have similarities to the evolution of dioecy from monoecy. In monoecious species, plants produce both male and female flowers, whereas plants with dioecious species produce flowers of only one sex. Renner and Ricklefs (1995)
have argued that one of the reasons that dioecy commonly evolves from monoecy is because the developmental machinery is already present in the ancestral state. However, the infrequent occurrence of dimorphic enantiostyly (compared to dioecy) suggests the genetic determination of left from right (or vice versa) required for a transition from monomorphic to dimorphic enantiostyly is much more difficult than the origin of dioecy from monoecy. Alternatively, selection for dimorphic enantiostyly may not be as strong as for dioecy (but see Jesson and Barrett, 2002b
, c
).
Clues to the constraints in the positional information determining left from right may be found by examining the patterning of left- and right-styled flowers on an inflorescence. In S. rostratum, the direction of style bending is almost completely predictable (Fig. 4). Consecutive flowers on a plant alternate between being left- or right-styled, with the style always turned towards the axis of the raceme (termed pendulum asymmetry; see Charlton, 1998
). This pattern was first noticed in S. rostratum by Todd (1882)
and Knuth (1906)
and also occurs in enantiostylous species of Cassia and Chamaecrista (Tucker, 1999
). While it is unknown how the direction of deflection of the first flower is initiated, the consistent patterning of flowers suggests that the position of the flower in relation to the axis of the inflorescence plays a critical role in determining stylar direction. The consistent deflection of a style towards the axis of the cyme implies that an axis of symmetry that is already present in the shoot plays a role in determining left from right. Thus, in S. rostratum, it is possible that monomorphic enantiostyly did not evolve through the development of a new axis of symmetry, but rather developed from positional cues that occurred from an already established axis.
In other species, the direction of deflection is not as predictable. While flowers of C. lutea on a given node are more likely to deflect to the opposite direction than the previous older flower, this pattern is not consistent. Therefore, factors other than pendulum asymmetry clearly play a role. We found no consistent predictors of stylar direction within inflorescences in H. mexicana. While stylar direction seems to be completely random in this species, it is likely that other factors not examined in this study (such as hormonal or environmental cues) may determine style direction. Further work is needed to determine whether the patterning of left and right styles within an inflorescence is the result of random accidents of development.
In species such as S. rostratum in which stylar deflection is predictable, left- and right-styled flowers are determined by the position of the flower relative to the axis of the inflorescence. Fixing the direction of deflection (as in dimorphic enantiostyly) would necessitate changing the required positional cue to one that is consistently left or right regardless of flower position. While aborting 50% of the flowers on a plant would produce entirely left- and entirely right-styled individuals in this species, the fitness consequences of this are likely to be severe.
Coen and workers (Carpenter and Coen, 1990
; Luo et al., 1996
) have shown that positional information in zygomorphic flowers is conveyed by two genes: CYCLOIDEA and DICHOTOMA. These genes establish a distinction between dorsal, lateral, and ventral organs, as well as dorsiventral asymmetry within individual organs. However, enantiostyly also requires the establishment of a medial-lateral axis (or the utilization of an existing apical-basal axis; see Fig. 1). Monomorphic enantiostyly may result from differential growth rates along apical-basal or medial-lateral axes. Dimorphic enantiostyly, however, requires differential growth rates consistently on one side. Therefore, additional genes or signaling pathways must be required to establish left from right as well as medial from lateral and dorsal from ventral. If this is true, then the evolution of dimorphic enantiostyly may be constrained by the absence of gene products that perceive and transduce appropriate positional information or by a lack of cues to provide such positional information.
Comparisons with the development of other stylar polymorphisms
Three other stylar polymorphisms are well documented in flowering plants: distyly, tristyly, and stigma-height dimorphism (Barrett, Jesson, and Baker, 2000
). While there have been no studies of the development of stigma-height dimorphism, the organogenesis of distylous and tristylous flowers have been investigated in some detail (Richards and Barrett, 1984
, 1987
, 1992
; Richards and Koptur, 1993
; Faivre, 2000
). As in enantiostyly, heterostylous species from different evolutionary origins show many ontogenetic differences. In tristylous species, the dimorphism of stamen height found within a morph is evident at initiation. The different organ heights can occur in a radial arrangement in two different stamen series (as in the Lythraceae and the Oxalidaceae) or in a dorsiventral arrangement (as in the Pontederiaceae) so that stamens in the same series can have different heights (Richards and Barrett, 1992
). In the enantiostylous species studied, the different organ heights also occur in a dorsiventral arrangement. Richards and Barrett (1992)
found no evidence that differences in stamen height within a flower were due to changes in growth rate, a pattern also found in this study. We found no evidence in enantiostylous species of differences in the relative growth of the feeding and the pollinating anther in all species examined. Instead, our allometric studies showed evidence for differences in stamen development from initiation, although this was not seen in SEM studies.
An examination of the development of stylar polymorphisms in an evolutionary context is important as it can be used to test evolutionary hypotheses. Faivre (2000)
compared the ontogeny of species with putative ancestral floral morphologies to heterostylous flowers. She found that transitions from an approach herkogamous morphology (in which the stigma is positioned above the anthers) to a heterostylous one was possible, as predicted in theoretical models by Lloyd and Webb (1992)
. However, the change in stylar growth patterns between a species with approach herkogamy and a closely related heterostylous species led Faivre (2000)
to suggest that a further developmental step was probably required for the evolution of heterostyly to occur. In monomorphic enantiostylous H. limosa and dimorphic enantiostylous H. multiflora, we found few differences in organ initiation and development. This gives some support to the hypothesis that dimorphic enantiostyly evolved from monomorphic enantiostyly, rather than from a separate evolutionary transition in which differences in development would more likely occur. It would be useful to compare development in other closely related pairs of monomorphic and dimorphic enantiostylous species to see if this pattern of similarity remains consistent.
The development of enantiostyly in species of diverse evolutionary origins is characterized by similar developmental processes. The position of sex organs in mature flowers is determined by bending of the style and pollinating stamen and these events occur late in development. Furthermore, variations in stamen size are not due to differences in growth rates. This suggests that the diverse ways in which enantiostyly can be expressed in flowering plants may be constrained by either developmental or physical structures. Despite this, the patterning of left- and right-styled flowers within inflorescences varies markedly between families. It is possible that such differences may play a role in limiting the evolution of dimorphic enantiostyly to only three monocotyledon families.
|
FOOTNOTES |
|---|
2 Author for reprint requests, current address: School of Biological Sciences, Victoria University of Wellington, P.O. Box 600, Wellington, New Zealand (linley.jesson{at}vuw.ac.nz
; FAX: 64 4 463 5331)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Barrett S. C. H. L. K. Jesson A. M. Baker 2000 The evolution and function of stylar polymorphisms in flowering plants. Annals of Botany 85: (Supplement A):253-265
Carpenter R. E. S. Coen 1990 Floral homeotic mutations produced by transposon-mutagenesis in Antirrhinum majus. Genes and Development 4: 1483-1493
Charlton W. A. 1998 Pendulum symmetry. In R. V. Jean and D. Barabe [eds.], Symmetry in plants. World Scientific, Singapore
Coyne J. 1987 Lack of response to selection for directional asymmetry in Drosophila melanogaster. Journal of Heredity 78: 119.
Diggle P. K. 1992 Development and the evolution of plant reproductive characters. In R. Wyatt [ed.], Ecology and evolution of plant reproduction. Chapman and Hall, New York, New York, USA
Dulberger R. 1981 The floral biology of Cassia didymobotrya and C. auriculata (Caesalpinaceae). American Journal of Botany 68: 1350-1360[CrossRef][ISI]
Endress P. 1999 Symmetry in flowers: diversity and evolution. International Journal of Plant Science 160: S3-S23[CrossRef][ISI][Medline]
Faivre A. E. 2000 Ontogenetic differences in heterostylous plants and implications for development from a herkogamous ancestor. Evolution 54: 847-858[ISI][Medline]
Giurfa M. A. Dafni P. R. Neal 1999 Floral symmetry and its role in plant–pollinator systems. International Journal of Plant Sciences 160: S41-S50[CrossRef][ISI][Medline]
Graham S. W. S. C. H. Barrett 1995 Phylogenetic systematics of Pontederiales: implications for breeding system evolution. In P. J. Rudall, P. J. Cribb, D. G. Cutler, and C. J. Humphries [eds.], Monocotyledons: systematics and evolution, 415–451. Royal Botanic Gardens, Kew, UK
Guerrant E. O. 1989 Early maturity, small flowers and autogamy: a developmental connection?. In J. H. Bock and Y. B. Linhart [eds.], The evolutionary ecology of plants, 61–84. Westview Press, Boulder, Colorado, USA
Hufford L . 1988a The evolution of floral morphological diversity in Eucnide (Loasaceae): the implications of modes and timing of ontogenetic changes on phylogenetic distribution. In P. Leins, S. C. Tucker, and P. K. Endress [eds.], Aspects of floral development, 103–119. Schweizerbartsche Verlagbuchhandlung, Stuttgart, Germany
Hufford L. 1988b Potential roles of scaling and post-anthesis developmental changes in the evolution of floral forms in Eucnide (Loasaceae). Nordic Journal of Botany 8: 147-157
Iyengar M. O. 1923 On the biology of the flowers of Monochoria. Journal of the Indian Botanical Society 3: 170-177
Jesson L. K. 2002 The evolution and functional significance of enantiostyly in flowering plants. Ph.D. dissertation, University of Toronto, Toronto, Ontario, Canada
Jesson L. K. S. C. H. Barrett 2002a The genetics of mirror-image flowers. Proceedings of the Royal Society of London, Series B 269: 1835-1839[Medline]
Jesson L. K. S. C. H. Barrett 2002b Solving the puzzle of mirror-image flowers. Nature 417: 707.[CrossRef][Medline]
Jesson L. K. S. C. H. Barrett 2002c Enantiostyly in Wachendorfia (Haemodoraceae): the influence of reproductive systems on the maintenance of the polymorphism. American Journal of Botany 89: 253-263
Knuth P. 1906 Handbook of flower pollination. Clarendon Press, Oxford, UK
Kohn J. R. S. W. Graham B. Morton J. J. Doyle S. C. H. Barrett 1996 Reconstruction of the evolution of reproductive characters in Pontederiaceae using phylogenetic evidence from chloroplast DNA restriction-site variation. Evolution 50: 1454-1469[CrossRef][ISI]
Lloyd D. G. C. J. Webb 1992 The selection of heterostyly. In S. C. H. Barrett [ed.], Evolution and function of heterostyly, 179–208. Springer-Verlag, Berlin, Germany
Lord E. M. 1981 Cleistogamy: a tool for the study of floral morphogenesis, function and evolution. Botanical Review 47: 421-449[ISI]
Luo D. R. Carpenter C. Vincent L. Copsey J. E. Coen 1996 Origin of floral asymmetry in Antirrhinum. Nature 383: 794-799[CrossRef][Medline]
Mathsoft. 1997 S-Plus. Data Analysis Products Division. Mathsoft, Seattle, Washington, USA
Maynard Smith J. R. Burian S. Kauffman P. Alberch J. Campbell B. Goodwin R. Lande D. Raup L. Wolpert 1985 Developmental constraints and evolution. Quarterly Review of Biology 60: 265-287[CrossRef]
Maynard Smith J. K. C. Sondhi 1960 The genetics of a pattern. Genetics 45: 1034-1050
Neter J. W. Wasserman M. H. Kutner 1990 Applied linear statistical models. Richard D. Irwin, Homewood, New Jersey, USA
Purnell D. J. J. N. J. Thompson 1973 Selection for asymmetrical bias in a behavioral character of Drosophila melanogaster. Heredity 31: 401-405[ISI][Medline]
Ree R. H. M. J. Donoghue 1999 Inferring rates of change in flower symmetry in asterid angiosperms. Systematic Biology 48: 633-641[CrossRef][ISI]
Renner S. S. R. E. Ricklefs 1995 Dioecy and its correlates in the flowering plants. American Journal of Botany 82: 596-606[CrossRef][ISI]
Richards J. H. S. C. H. Barrett 1984 The developmental basis of tristyly in Eichhornia paniculata (Pontederiaceae). American Journal of Botany 71: 1347-1363[CrossRef][ISI]
Richards J. H. S. C. H. Barrett 1987 Development of tristyly in Pontederia cordata (Pontederiaceae) I. Mature floral structure and patterns of relative growth of reproductive organs. American Journal of Botany 74: 1831-1841[CrossRef][ISI]
Richards J. H. S. C. H. Barrett 1992 Development of heterostyly. In S. C. H. Barrett [ed.], Evolution and function of heterostyly, 85–128. Springer-Verlag, Berlin, Germany
Richards J. H. S. Koptur 1993 Floral variation in Guettarda scabra (Rubiaceae). American Journal of Botany 80: 31-40[CrossRef][ISI]
Sokal R. R. F. J. Rohlf 1995 Biometry. Freeman, New York, New York, USA
Soltis P. S. D. E. Soltis M. W. Chase 1999 Angiosperm phylogeny inferred from multiple genes as a tool for comparative biology. Nature 402: 402-404
Todd J. E. 1882 On the flowers of Solanum rostratum and Cassia chamaecrista. American Naturalist 16: 281-287[CrossRef]
Tucker S. C. 1999 Evolutionary lability of symmetry in early floral development. International Journal of Plant Sciences 160: S25-S39[CrossRef][ISI][Medline]
Tuinstra E. J. G. de Jong W. Scharloo 1990 Lack of response to family selection for directional asymmetry in Drosophila melanogaster: left and right are not distinguished in development. Proceedings of the Royal Society of London, Series B 241: 146-152[Medline]
Wake D. B. 1991 Homoplasy: the result of natural selection, or evidence of design limitation. American Naturalist 138: 543-567[CrossRef][ISI]
This article has been cited by other articles:
|
|
 |
|
  B. Marazzi and P. K. Endress Patterns and development of floral asymmetry in Senna (Leguminosae, Cassiinae) Am. J. Botany, January 1, 2008; 95(1): 22 - 40. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||
