| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Development and Morphogenesis |
Department of Cell Biology and Molecular Genetics, University of Maryland, College Park, Maryland 20742 USA
Received for publication October 10, 2002. Accepted for publication March 25, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
90° during maximal phase) are often positioned in spiral patterns, whereas primordia arising on the subtending axis (as defined by displacement angles of 
90° during maximal phase) are arranged in whorled patterns. A secondary geometrical relationship derived from the literature shows an inverse correlation between the primordial size : available space ratio and the magnitude of the Fibonacci numbers in spiral phyllotaxis or the number of leaves per whorl in whorled phyllotaxis. The data available for terrestrial plants suggest that their phyllotactic patterning may also be specified by these same geometrical relationships. Major exceptions to these correlations are attributable to persistent embryonic patterning, leaflike structures arising from stipules, congenital splitting of young primordia, and/or non-uniform elongating of internodes. The geometrical analysis described in this paper provides the morphological context for interpreting the phenotypes of phyllotaxis mutants and for constructing realistic models of the underlying mechanisms responsible for generating phyllotactic patterns.
Key Words: aquatic plants • geometrical model • phyllotaxis • shoot apical meristem
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Arber, 1920
; Sculthorpe, 1967
). Recent phylogenetic analyses of angiosperm evolution have identified the fossil aquatic plant family Archaefructaceae as the sister group to all angiosperms (Sun et al., 2002
) and several living aquatic plants, including Ceratophyllum, Cabomba and Potamogeton, as having arisen from early divergent angiosperm lineages (Qiu et al., 2001
). Moreover, terrestrial angiosperms have re-invaded aquatic environments a minimum of 205 times (Cook, 1970
, 1999
). Aquatic plants adopt an exceptional range of habits including submerged, floating, and emergent shoots with or without roots anchored into the substrate. Thus, aquatic angiosperms manifest the ability of the angiosperm lineage to evolve developmental innovations in response to ecological selection pressures. The underlying assumption of this paper is that aquatic angiosperms are also expressing the complete range of different phyllotaxes (leaf arrangements). Therefore, we anticipate that aquatic plants may be especially useful for elucidating the basic geometrical relationships specifying phyllotactic patterning.
The initial phyllotactic pattern is established with the positioning of leaf primordia on the surface of the shoot apical meristem (Lyndon, 1998
). This pattern may persist to form the observed arrangement of mature leaves, or it may be altered to develop a new pattern as the consequence of differential growth of various subapical regions. Three general patterns are routinely observed on mature shoots: alternate (one leaf inserted at the node), opposite (two leaves), and whorled (three or more leaves). Alternate phyllotaxis, which is the most common pattern observed on the shoot tips of terrestrial plants, is routinely described in terms of the spirals of leaf primordia radiating from the meristem center (Erickson, 1983
; Jean, 1992
, 1994
). If leaves are assigned a number in the order of their origin, then the intervals in leaf numbers between successive leaves on these spirals can typically be related to the enigmatic Fibonacci number series. However, the underlying causes of phyllotactic patterning remain elusive despite research into plausible chemical gradients (Thornley, 1975
; Veen and Lindenmayer, 1977
; Meinhardt, 1982
), biomechanical stresses (Green et al., 1996
; Green, 1999
) and molecular regulation (Leyser and Furner, 1992
; Clark et al., 1997
; Itoh et al., 2000
) operating in the shoot apical meristem.
In this paper, a geometric approach was used to characterize the phyllotactic patterns in aquatic plants having Fibonacci spirals and non-Fibonacci whorls because previous biomechanical and biochemical models for phyllotactic patterning were largely divorced from the fundamental geometry of the meristems responsible for these patterns. Using a polar-coordinate model of shoot apical meristems, we have discovered striking geometric relationships between the positions of new primordia and the resulting phyllotaxis. These new insights may help to separate the underlying mechanism of leaf positioning from the more specialized process of primordial initiation in Fibonacci spirals. Moreover, these geometric relationships may provide the structural context for the expression of molecular mechanisms involved in leaf initiation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Watson and Dallwitz, 2000
). The advanced mode was chosen to screen 583 angiosperm families. Settings were changed from the default settings to those recommended for data retrieval. The ecological type was set to hydrophyte to identify an initial list of angiosperm families with aquatic members and a second initial list of angiosperm families lacking aquatic members. The character of leaves-insertion was then used to screen for the presence of alternate, opposite, and whorled species in the families of the initial hydrophyte, initial nonhydrophyte, and total lists.
Various taxonomic sources (Cook, 1970
; Gleason and Cronquist, 1991
; Heywood, 1993
; Zomlefer, 1994
) were used to determine whether each hydrophyte family in the initial list had sufficient aquatic genera to warrant being considered a genuine hydrophyte family. Large families composed of more than 100 genera that were almost exclusively composed of terrestrial genera were eliminated from the initial hydrophyte list. Smaller families with aquatic species in 10% or less of their genera were also excluded from the final hydrophyte list. Lastly, families composed of a single genus with less than 10% aquatic species were also excluded. The leaf insertion data for the final hydrophyte list were manually adjusted to reflect the exclusion of families from the initial list based on the criteria outlined. Similarly, those same families were manually added to the initial nonhydrophyte list for inclusion in the final nonhydrophyte list.
Apical meristem measurements
Using an Olympus SZH10 research stereoscope at magnifications up to 140x, 4–14 healthy shoots per species were hand-dissected (Table 1). The meristems were kept hydrated while enough leaves were carefully removed in order to expose the youngest primordium. The hydrated apices were transferred in depression slides to a Wild (Wild Heerbrugg Ltd., Heerbrugg, Switzerland) compound microscope where the apical domes and youngest primordia were drawn in sagittal view with a camera lucida at 200x. The image from a slide micrometer was also drawn to determine the magnification of the meristem images. When appropriate, enlarged copies of published median longitudinal sections of shoot apical meristems and supplemental textual descriptions were used to augment the live materials (Table 1).
|
). Using trigonometry, these height and width measurements were then converted to polar coordinates (Table 2). The radius (r) of the shoot apex corresponded to the radius of the defined circle, and the position of the youngest leaf primordium was represented by the displacement angle (), which was delimited by the radius connecting the highest point on the dome summit and the radius drawn to the insertion point of the youngest primordium. The angular displacement () was converted into a linear displacement (x) along the circle by multiplying the arc defined by theta and the circle circumference, or
 | (1) |
 | (2) |
= 0°) equated to an x/r of zero, and a of 90° would always yield an x/r of 
/2 or 1.57.
|
|
values of 90°, indicating that the circle was inscribed on their apical domes. Their x/r values were calculated as the sum of /2 and y/r, the latter factor representing the additional distance from the circle to the youngest primordium, normalized for the radius. All measurements were incorporated into a Microsoft Excel 97 spreadsheet for performing routine calculations of all variables, for correcting for magnification if needed, and for graphing the relationships among the variables.
All meristem drawings representing different species with alternate, opposite, and whorled phyllotaxes were analyzed to generate model values. For most meristems, the minimal phase occurred just after the initiation of the youngest primordium and the maximal phase occurred just prior to the initiation of the next primordium (for discussion, see Esau, 1965
). Meristem measurements were plotted as minimal phase (lowest or x/r value) and maximal phase (highest or x/r value) for each species.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
The whorled, alternate, and opposite groups generated from the database of flowering plant families were also examined for patterns that would prove informative as to the factors potentially involved in different phyllotaxes. Interestingly, the monocots had very few species with opposite phyllotaxis, but these species were almost exclusively classified as hydrophytes. By contrast, the large number of dicot opposites were sorted into both hydrophyte and nonhydrophyte categories (data not shown).
Geometric analysis of aquatic plant meristems
Ideally, morphological models should be constructed so that the model parameters can be directly related to fundamental structural properties. The traditional rectilinear coordinates of height and width measurements were converted to polar coordinates because height and width, independent of each other, cannot be related to a specific meristem property, such as size, shape, or primordium position. In the model presented here, the displacement angle (), or its normalized linear equivalent (x/r), represents an independent measure of primordial position (Fig. 1). The meristem can be divided into two distinct regions consisting of the apical dome shaped like a hemisphere or hemispherical sector and the subtending axis shaped like a regular cylinder or truncated cone. The dividing line between these two regions corresponds to a value of 90°, also expressed as 1.57 in linear terms. Hence, lower or x/r values correspond to a primordial position close to the dome summit, and higher or x/r values correspond to a position away from the summit on the subtending axis beneath the dome. The radius is inversely related to the curvature of the apical dome: a broad meristem with slight curvature is characterized by a high r value, whereas a narrow meristem with pronounced curvature is characterized by a low r value.
Given the cyclical changes in meristem size during primordium initiation, Figs. 2A, 3A, and 4A present minimal and maximal phase data for each species. Particular attention was given to and corresponding x/r values associated with the maximal phase, which represented the shoot apex just prior to the emergence of the new primordium. The switch between primordial initiation on the apical dome to the subtending axis was indicated by a solid line on each graph at = 90° (or x/r = 1.57). To reduce the labels needed to identify the points on the graphs, data for associated minimal and maximal phases were connected with a dotted line in Figs. 2A–4A.
|
|
|
Alternate phyllotaxis
The geometrical characteristics of the shoot apical meristems of two lycophytes (Isoetes engelmannii and I. riparia), two monocotyledons (Oryza sativa and Potamogeton perfoliatus), and two dicotyledons (Proserpinaca palustris and Ranunculus flabellaris) that display alternate phyllotaxis of mature leaves were graphed in Fig. 2. For example, the shoot apex of Potamogeton perfoliatus was characterized by a radius of 50 µm and a displacement angle of 32° at the minimal phase and by a radius of 37 µm and an angle of 79° at the maximal phase (Fig. 2A). In general, the meristems of all species except O. sativa remained grouped together. These meristems had radii ranging between 36 µm and 74 µm and displacement angles between 32° and 79° at their maximal phases, signifying that the youngest primordia on these meristems were initiated on the apical dome itself. The meristem of O. sativa, with a slightly smaller radius of 29 µm, had displacement angles equal to or greater than 90° in both phases and corresponding x/r values of 1.57 at minimal phase and 2.24 at maximal phase. Thus, the youngest primordium on this meristem developed on the axis subtending the apical dome. Interestingly, unlike other primordia in this category, this primordium rapidly encircles the apical meristem soon after its inception (Sharman, 1945
).
Height and width measurements were also taken for the same group of alternate species and graphed in Fig. 2B. A min-max line was drawn to indicate the growth dynamics during the plastochron. The shoot apical meristem of Potamogeton perfoliatus, our earlier example, had a height and width of 8 µm and 53 µm, respectively, during the minimal phase and 30 µm and 73 µm, respectively, at its maximal phase. The dimensions of the minimal phases of alternate meristems ranged from 3 µm to 32 µm in height and from 37 µm to 92 µm in width, and their maximal phases ranged from 7 µm to 40 µm in height and from 47 µm to 126 µm in width. The meristems of Oryza sativa exhibited an exceptional height of 48 µm at maximal phase. As indicated by the positive slope of the min-max lines, the meristems of all species except O. sativa increased in both height and width during the plastochron. In addition, the majority of the slopes appeared relatively flat, signifying that their plastochronic dynamics emphasized lateral as opposed to axial growth. The min-max line for Potamogeton perfoliatus had a slope of 1.13, which indicated that this meristem experiences roughly equal growth in both directions. The negative slope for the meristematic growth of O. sativa revealed that this meristem underwent axial growth at the expense of lateral growth.
Because Ranunculus flabellaris and Proserpinaca palustris meristems can be found above or below the water's surface, geometrical measurements were taken from both emergent and submersed meristems to evaluate the effect of environmental conditions on meristem geometry. In both species, the maximal phase of the emergent meristem had a smaller r value, which corresponded to a less protuberant shape. The displacement angles of maximal phase of the emergent and submersed P. palustris meristems were almost identical. A greater difference was observed in R. flabellaris where the emergent displacement was greater than that in the submersed form (Fig. 2A). In both species, the overall geometry of these meristems fell within the typical range for alternate plants. The displacement angles did not exceed 90°, indicating that the primordia were initiated on the meristem dome. Although the emergent meristems of both heterophyllic species had steeper min-max lines than the corresponding submersed meristems, these slopes remained below a value of one (Fig. 2B). Thus, the plastochron dynamics of these meristems were not significantly affected by their position relative to the water's surface.
Opposite phyllotaxis
The dimensions of radius (r), angular displacement (), and its linear counterpart (x/r) were collected from 10 species including three monocotyledons (Najas guadalupensis, another Najas sp., and Zannachellia palustris) and seven dicotyledons (Callitriche heterophylla, Cabomba sp., Dianthera americana, Ludwigia pervensus, Bacopa monnieria, Bacopa carolinia, and Rotalia indica) that display opposite phyllotaxis of mature leaves (Fig. 3A). Typical of the shoot apical meristems of most dicotyledons, D. americana was characterized by an r value of 82 µm and a value of 25° during its minimal phase. At the maximal phase of the plastochron, its r value decreased to 52 µm and value increased to 59°. The meristems of all dicotyledons except Callitriche heterophylla had radii in the range of 40–77 µm and in the range of 39°–77° during the maximal phase. Thus, these meristems initiated new primordia on their apical domes. The four remaining species, the dicotyledon Callitriche heterophylla (including both the emergent [CHe] and submersed [CHs] specimens), and the monocotyledons N. guadalupensis (NG), Najas sp. (NS), and Z. palustris had apical meristems whose values were 90° and x/r values of 1.57 (CHe), 1.96 (CHs), 2.47 (NG), and 2.92 (NS) at their maximal phases. Therefore, these meristems initiated new primordia on the subtending axis below the apical dome.
The min-max lines, representing the plastochronic changes in meristems for opposite plants, are seen in Fig. 3B. Again, using Dianthera americana as a typical opposite plant, the minimal phase of its meristem was 8 µm in height and 69 µm in width, and the maximal phase was 26 µm in height and 89 µm in width. As a group, the apical meristems of dicot opposite species had ranges of 2–50 µm in height and 42–72 µm in width during the minimal phases and 17–32 µm in height and 48–89 µm in width during the maximal phases. The monocotyledons Najas sp. (NS), N. guadalupensis (NG), and Z. palustris had the greatest heights within this group, but they exhibited average widths. Conversely, the meristems of the dicotyledon Cabomba sp. had widths outside those seen for other opposites, 108 µm at minimum and 97 µm at maximum, but maintained heights typical of other meristems of plants developing opposite phyllotaxis. The meristems of all dicotyledons excluding Cabomba sp. had min-max lines with positive slopes. Four meristems displayed min-max lines with slopes less than one, indicating that their plastochronic dynamics emphasized lateral and not axial growth. Two meristems had slopes greater than one, representing a greater emphasis on axial growth. The remaining species, including all three monocotyledons plus the dicotyledon Cabomba sp., produced min-max lines with negative slopes, which illustrated a pronounced emphasis on axial growth during their plastochrons.
Included among the opposite plants was one heterophyllic species, Callitriche heterophylla. A comparison of the emergent vs. submersed meristems revealed that during the maximal phase, the emergent meristem had a slightly larger radius, 24 µm vs. 19 µm, and a smaller linear displacement (x/r), 1.57 vs. 1.96 (Fig. 3A). However, based on the displacement values of both these meristems, primordia were initiated on the subtending axis beneath the apical dome. Even smaller differences were seen in the height and width measurements of submersed vs. emergent apices (Fig. 3B). The steep slopes of both min-max lines reflected the predominance of axial growth during their plastochrons.
Whorled phyllotaxis
The normalized linear and angular displacements were plotted against dome curvature for eight whorled species including two monocotyledons, Anacharis densa (= Elodea densa) and Hydrilla sp., and six dicotyledons, Ceratophyllum densa, Hippuris vulgaris, Myriophyllum heterophyllum, M. spicatum, M. aquaticum, and Rotalia indica (Fig. 4A). Anacharis densa, commonly used in text as an example of whorled phyllotaxis, had a meristematic radius of 42 µm during its minimal phase and 44 µm during its maximal phase, which indicated that the curvature of its apical dome remains constant over the plastochron. In both phases, angular displacement () was greater than 90° as evidenced by a minimal normalized linear displacement of 2.5 and a maximal linear displacement of 3.9. The eight whorled species sampled had a wide range of meristematic radii with maximal phases between 36 µm and 68 µm, and normalized linear displacements between 1.8 and 4.1. These data show that, without exception, all whorled species in this study initiated primordia on the axis subtending the apical dome.
During the plastochron, Anacharis densa had a height of 80 µm and width of 100 µm at minimal phase and a height of 148 µm and a width of 136 µm at maximal (Fig. 4B). Overall, the apical meristems of whorled species ranged from 19 µm to 94 µm in height and 47 µm to 168 µm for width during the minimal phase. Every meristem in this group increased in height and width during the plastochron with maximal phases that ranged from 64 µm to 148 µm for height and 76 µm and 168 µm for width. As a result, almost all min-max lines had positive slopes; therefore, these species emphasized axial growth. Exceptions included Myriophyllum heterophyllum (submersed) and M. aquaticum meristems, which had slopes close to one, corresponding to a balance between axial vs. lateral growth.
Whorled submersed and emergent meristems were observed in the cases of Myriophyllum heterophyllum and Hippuris vulgaris. Submersed meristems of Hippuris vulgaris had greater displacement angles and heights than emergent meristems, while all Myriophyllum heterophyllum meristems tended to have overlapping ranges. Both species were similar to other whorled plants in that primordia were initiated on the subtending axis of the apical dome. In addition, both species experienced plastochron growth dynamics similar to the other whorled plants.
Overall comparisons
The radii of the apical domes of aquatic plants fell within an overlapping range from 36 µm to 77 µm at the maximal phase. Thus, apical curvature had no apparent influence on the phyllotactic pattern chosen. By contrast, maximal phase displacement angles equal to or greater than 90° were exclusive to whorled species. Therefore, a large displacement from the dome, which resulted in the initiation of primordia on the subtending axis of the dome, was a necessary characteristic of whorled phyllotaxis. With the exception of one alternate (Oryza sativa) and several opposite species (Callitriche heterophylla, Najas guadalupensis, Najas sp., and Zannachellia palustris), all meristems with either alternate or opposite phyllotaxis exhibited smaller displacement angles associated with primordial initiation on the apical dome. Similarly, the growth dynamics in the meristems of whorled plants and those exceptional plants listed before had pronounced axial growth at the expense of lateral growth, which differed from the greater emphasis on lateral growth in all other alternate and many opposite plants.
Heterophyllic species were present in all three categories of phyllotactic patterns. In general, the submersed meristems of heterophyllic species were characterized by greater heights and greater normalized linear displacements at the maximal phase than comparable emergent meristems with the result that the submersed meristems extended farther above the existing leaf primordia. However, no differences in the phyllotactic patterns originating from submersed vs. emergent meristems were observed in the heterophyllic species examined in this study.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
) that are directly related to meristem curvature and primordial position, respectively. The resulting data in Figs. 2A, 3A, and 4A establish that the primordia arranged in alternate and opposite phyllotaxis tend to have low values at maximal phase, i.e., they arise high on the apical dome, while primordia arranged in whorled phyllotaxis tend to have values >90° at maximal phase, i.e., they arise off the apical dome on the subtending axis.
We propose that these observations disclose several fundamental geometrical relationships that represent ground states for phyllotactic patterning. Typically, the apical dome bears its primordia in alternate arrangements that are usually characterized as two spirals radiating across the dome. On the other hand, the subtending axis bears its primordia in whorled arrangements that consist of successive rings of several primordia wrapped around the axis. Therefore, it seems reasonable to conclude that these observations reveal the primary geometrical principle relating apical geometry to primordial phyllotaxis: (A) primordia arising on the apical dome tend to assume a spiral arrangement, and (B) primordia arising on the subtending axis tend to assume a whorled arrangement (Fig. 5). In the latter case, the primordia of whorled species can initially arise on the axis as a unified ring, which will then differentiate into separate primordia (Equisetum and Ceratophyllum: Rutishauser and Sattler, 1987
), or they can arise as discrete primordia positioned in an whorled arrangement (Anacharis: Clowes, 1961
).
|
: Fig. 3.8; Jean, 1994
; Lyndon, 1998
), and (B) the smaller the primordial size relative to the overall annulus (or ring) size, the greater the number of primordia arising in the resulting whorl (Fig. 5). In the latter case, the aquatic Hippuris vulgaris has an elongated meristem with several primordial rings forming on the subtending axis below the apical dome; an increase in primordial number in each whorl is directly correlated with an increase in annulus diameter (McCully and Dale, 1961
). Several whorled terrestrial species, including Equisetum arvense (Bierhorst, 1959
), Ephedra altissima (Gifford, 1943
), and Casuarina distyla (Williams and Metcalf, 1985
), have this same correlation (Fig. 6).
|
This interpretation is strongly supported by morphological evidence from the apical meristems of grass mutants. Our geometrical model predicts that a change in the site of primordial initiation from the axis to the dome should result in the corresponding switch from whorled to spiral phyllotaxis. In particular, the rice sho mutants are characterized by flatter meristems, young primordia on the apical dome, and nonclasping leaves. It is entirely consistent with our model that sho mutants have an irregular spiral phyllotaxis with horizontal divergence angles approaching the 138° angle characteristic of Fibonacci-based patterns (Itoh et al., 2000
). Furthermore, it follows from the geometrical model that a decrease in the ratio of primordial size relative to annulus diameter should result in more than one primordium being initiated in the annulus. In wild-type maize, the primordium encircles the meristem to such an extent that the margins of the primordium overlap each other. By contrast, the maize mutant abphyl exhibits a profound increase in meristem size without a compensatory increase in primordial size. As a result, the axis can accommodate the simultaneous initiation of two primordia in an opposite arrangement (Greyson et al., 1978
: cf. Fig. 1 vs. Fig. 4). In conclusion, these mutants confirm the fundamental validity of applying the "whorl of one" concept to grass phyllotaxis.
The opposite monocots also offer an intriguing opportunity to evaluate our model. In principle, opposite phyllotaxis can be viewed as a special case of whorled phyllotaxis consisting of two primordia that arise at the same level. Because monocot embryos develop a single cotyledon, all monocots must initially express alternate phyllotaxis by default. Thus, opposite monocots have subsequently switched from alternate to whorled phyllotaxis. Interestingly, the limited number of monocot families that are reported as typically having opposite phyllotaxis is almost exclusively composed of aquatic species (Dallwitz et al., 2000
; Watson and Dallwitz, 2000
). It is noteworthy that all aquatic opposite monocots in Fig. 3 initiate their primordia on elongated, narrow apices, with displacement angles greater than 90°, thus corresponding to the same location for primordial initiation in all aquatic whorled species (Fig. 4A). Thus, it follows as a reasonable conclusion that these monocots switched to a whorled phyllotaxis of two primordia once the leaves arose off the apical dome in an annulus large enough to fit two primordia (Fig. 5).
In marked contrast to the restricted occurrence of opposite phyllotaxis in monocot families, this phyllotactic pattern is widespread among dicot families. Almost all dicots produce two cotyledons that arise in an opposite arrangement near the apex of the globular embryo (Gifford and Foster, 1989
; Kaplan and Cooke, 1997
). That these opposite patterns are initiated on the apical domes of dicot embryos implies that these embryos are not constrained by the same geometrical relationships observed on other whorled species. Thus, dicot embryos are apparently able to override the basic geometrical relationships to produce the equivalent of an opposite phyllotaxis on their apical dome. This opposite arrangement of dicot cotyledons is converted to postembryonic arrangements of different spiral phyllotaxes in up to 85% of dicot seedlings (Meicenheimer, 1998
), which is consistent with the new leaves being initiated on the apical dome. However, the other 15% of dicot species remain locked into the embryonic pattern of opposite cotyledons that becomes their postembryonic pattern of leaf phyllotaxis. Despite their primordia having displacement angles <90°, most opposite aquatic dicots measured in Fig. 3 have this mode of persistent opposite phyllotaxis. By contrast, the apical meristems of Callitriche heterophylla produce new primordia with displacement angles around 90°, which may be indicative of a genuine whorled phyllotaxis composed of two small primordia arising on the subtending axis. Lastly, several dicot families develop monocotyledonous embryos (for references, see Gifford and Foster, 1989
), and the phyllotactic arrangements of these plants, as would be expected from geometric principles, are invariably spiral.
In summary, the analysis of meristem geometry of aquatic plants depicted in Figs. 2–4 leads to the realization that their apices produce new primordia in two fundamental patterns: (1) spiral patterns, which are produced by new primordia arising at different levels on the apical dome, and (2) whorled patterns, which result from one or more primordia arising in a complete ring around the subtending axis. These geometrical relationships are validated by their ability to account for the unexpected axial positions of new primordia in the instances of the alternate phyllotaxis of rice plants and the opposite phyllotaxis of aquatic monocots. However, these relationships are frequently overridden by unknown biological mechanisms in the case of most dicots that have persistent opposite phyllotaxis.
Evolution of whorled phyllotaxis in aquatic plants
The aquatic species examined in this paper have almost exclusively generated whorled phyllotaxis via the formation of lateral primordia on an elongated meristem. Thus, the repeated evolution of whorled phyllotaxis in aquatic plants (see introduction) appears to depend in part on the potential of elongated meristems to evolve in aquatic environments. Two observations hint at potential selection pressures operating on meristem shape itself. Firstly, the height of emergent meristems of all heterophyllic species, excluding Ranunculus flabellaris, was significantly reduced relative to submersed meristems (Figs. 2–4B). Conversely, xeric plants like cacti often possess flat or sunken meristems that are embedded underneath the surrounding subapical structures (Boke, 1951
, 1954
, 1955
). Thus, aquatic environments seem to be more permissive toward exposed meristems in contrast to terrestrial environments where the desiccating effects of even intermittent exposure appear to select for more protected meristems.
The relative amount of light harvesting is a second potential selection pressure operating on phyllotactic pattern. In computer simulations, light harvesting is maximized by whorled phyllotaxis provided that the leaves within each whorl do not overlap and the internodal distance equals or exceeds leaf length (Niklas, 1998
). Thus, it could be argued that whorled phyllotaxis may be favored in both submerged aquatic and shaded terrestrial environments where light penetration is often limiting (for compensating factors, see Niklas, 1998
). Because most terrestrial plants lack the elongated meristems characteristic of aquatic plants with whorled phyllotaxis, then the question becomes, How do certain terrestrial plants develop equivalent types of this phyllotactic pattern?
Exceptional examples of whorled phyllotaxis in terrestrial plants
It is critical to confirm that the fundamental geometrical relationships underlying phyllotactic arrangements described for aquatic plants are applicable to all plants. Our initial survey of the meristem geometry of terrestrial plants (data not shown) indicated that they do generally follow the fundamental geometric relationships illustrated in Fig. 5. In particular, a few terrestrial plants, such as Equisetum arvense, Ephedra altissima, and Casuarina distyla, produce elongated meristems with multiple primordia arising in each whorl on the subtending axis (Fig. 6A–C). However, the less protuberant, compressed meristems of most other terrestrial plants appear to be geometrically restricted from generating whorled arrangements of young primordia. Then it comes as no surprise that terrestrial plants tend to favor spiral arrangements of their primordia, with the exception of the 15% of dicots locked in an embryonic opposite pattern, as described earlier.
Terrestrial plants have, nevertheless, adopted several different developmental mechanisms for altering the primordial phyllotaxis into either whorl-mimics or pseudowhorls. One example of a nongeometrical mechanism for creating whorl-mimics involves the modification of stipules into leaflike organs. For example, the dicot Galium rubioides (Fig. 6E) produces a whorl of leaflike structures that consists of two leaves and 4–6 stipules that are transformed into leaflike organs (Takeda, 1916
; Troll, 1943
). Although Rutishauser (1984)
argues that these modified stipules are genuine foliage leaves, the weight of morphological evidence, such as the absence of associated buds, discloses their stipular nature. Thus, we interpret the whorl-mimicking phyllotaxis of this genus as originating from the interaction of two nongeometric factors: the basic pattern of a locked opposite arrangement resulting from the persistent embryonic condition, which is subsequently modified through enhanced stipular growth.
A second nongeometrical mechanism appears to involve the separation of a single primordium into two primordia. For example, the dicot Nerium oleander exhibits an opposite (i.e., decussate) phyllotaxis during early vegetative growth. Because N. oleander develops a flat meristem (Williams et al., 1982
), this postembryonic pattern must also be attributed to the persistence of embryonic patterning. However, the higher nodes of this plant have an intriguing transitional state between the original decussate and the later tricussate phyllotaxes (Fig. 6F; Williams et al., 1982
). This transitional state has one large primordium (L8, upper right) opposite two adjacent small primordia (L8, lower left), which is reasonably interpreted as representing the other large primordium being congenitally split into two primordia. In higher whorls (L9), the transitional state is obscured as the smaller primordia assume the size of the large primordium and are repositioned at 120° divergence angles around the apex. Another example of this type of whorl-mimic is Abelia, which has transitions from trimerous to tetramerous whorls (Douady and Couder, 1996c
: Fig. 13). Thus, these whorl-mimics are also generated by persistent embryonic patterning and altered primordial growth.
The third common mechanism for creating a whorled appearance in terrestrial plants does not involve the alteration of primordial organs. Instead, the subapical elongation of certain adjacent internodes is altered so that primordia that were initiated at different levels on the meristem are now closely positioned on the mature shoot, thereby creating pseudowhorls (for many examples, see Schoute 1921
, 1925
, 1936
, 1938
). In Peperomia verticillata, pseudowhorls result from the limited elongation of certain internodes in both spiraled and whorled species (Kwiatkowska, 1999
). Those pseudowhorls derived from spiral patterns are more likely to experience a fluctuating leaf number in contrast to those resulting from the combining of two whorled nodes, which are always double the number found in the original whorl. Lastly, while almost all examples of the pseudowhorled arrangement are found on terrestrial plants, each vertical shoot of the marine plant Halophila engelmannii bears a single pseudowhorl that is derived from a compressed distichous phyllotaxis (data not shown).
It is clear from this brief survey that phyllotactic patterning in many aquatic plants illustrates the potential for a morphogenetic process to occupy what might be called the biophysical ground state. By contrast, this same process in certain terrestrial plants illustrates the potential for some plants to evolve specific biological mechanisms that override the ground state to achieve another developmental outcome.
General considerations about the mechanistic basis of phyllotactic patterns
The observations reported in this paper have significant implications for future efforts to construct mechanistic models for phyllotactic patterning. In the past, it has been universally assumed that the meristem is realistically modeled as a transverse polar projection (for recent examples, see Jean and Barabé, 1998
). In particular, Green (1999)
represents the shoot apex as a planar surface bearing concentric, leaf-generating annuli surrounding the apex center. Other workers make explicit statements that meristem geometry need not be considered as an important parameter in their physical models and computer simulations of phyllotactic patterning (e.g., Douady and Couder, 1996a
, b
). Although these models are generally successful to the extent that they can generate either whorled (e.g., Green et al., 1996
; Green, 1999
) or spiral patterns (e.g., Mitchison, 1977
; Douady and Couder, 1996a
, b
, c
), they have been unable to account for the transitions between these patterns without contrived explanations about optimal packing. Our results argue very strongly that the site of primordial initiation specifies the subsequent phyllotactic pattern, which means that pattern transitions can readily be accomplished by subtle alterations in the dynamics of meristematic growth vs. primordial initiation.
Our observations indicate that the three types of mature phyllotaxis, alternate, opposite, or whorled, are each generated by several different developmental processes. For instance, whorled phyllotaxis can arise from various processes, including whorled patterns of primordial initiation, altered dynamics of primordial growth, or non-uniform internodal elongation. If the geometrical relationships presented here prove to represent a true ground state that holds for all angiosperms, then future modeling efforts should focus on the simple geometrical relationships specifying phyllotactic pattern as opposed to the more complicated situations in which uncharacterized biological processes override those relationships.
It remains plausible that separate physicochemical mechanisms may be operating to generate either spiral phyllotaxis on the dome or whorled phyllotaxis on the axis. However, the evidence available suggests that a single mechanism interprets meristem geometry in order to produce the appropriate phyllotaxis for each potential site of primordial initiation. Given that many angiosperm lineages have repeatedly re-invaded the aquatic environments, it is striking that no aquatic plants with elongated meristems have retained spiral phyllotaxis, which might be expected in the unlikely circumstance that each whorled aquatic had to evolve its own mechanism de novo for generating that phyllotaxis. We are currently evaluating reaction-diffusion systems and mechanical-buckling models to determine if they can generate the main phyllotactic patterns in the appropriate locations on shoot apices.
|
FOOTNOTES |
|---|
2 Author for reprint requests (telephone: 301-405-1628; fax: 301-314-9081; e-mail: wjkelly{at}wam.umd.edu
)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Bierhorst D. W. 1959 Symmetry in Equisetum. American Journal of Botany 46: 170-179[CrossRef][ISI]
Boke N. H. 1951 Histogenesis of the vegetative shoot in Echinocereus. American Journal of Botany 38: 23-38[CrossRef][ISI]
Boke N. H. 1954 Organogenesis of the vegetative shoot in Pereskia. American Journal of Botany 41: 619-637[CrossRef][ISI]
Boke N. H. 1955 Development of the vegetative shoot in Rhysalis cassytha. American Journal of Botany 42: 1-10
Bostrack J. J. W. F. Millington 1962 On the determination of leaf form in an aquatic heterophyllous species of Ranunculus. Bulletin of the Torrey Botanical Club 89: 1-20
Clark S. E. R. W. Williams E. M. Meyerowitz 1997 The CLAVATA1 gene encodes a putative receptor kinase that controls shoot and floral meristem size in Arabidopsis. Cell 89: 575-585[CrossRef][ISI][Medline]
Clowes F. A. L. 1961 Apical meristems. Blackwell Scientific, London, UK.
Cook C. D. K. 1970 Water plants of the world. Dr. W. Junk, The Hague, Netherlands
Cook C. D. K. 1999 The number and kinds of embryo-bearing plants which have become aquatic: a survey. Perspectives in Plant Ecology, Evolution and Systematics 2: 79-102
Dallwitz M. J. T. A. Paine E. J. Zurcher 2000 User's guide to Intkey: a program for interactive identification and information retrieval, 1st ed. http://biodiversity.uno.edu/delta
Deschamp P. A. T. J. Cooke 1985 Leaf dimorphism in the aquatic angiosperm Callitriche heterophylla. American Journal of Botany 72: 1377-1387[CrossRef][ISI]
Dodoens R. 1578 A Niewe Herball, or Historie of Plantes (English translation by Henry Lyte Esquyer). Gerard Dewes, London, UK
Douady S. Y. Couder 1996a Phyllotaxis as a dynamical self organizing process. Part I: The spiral modes resulting from time-periodic iterations. Journal of Theoretical Biology 178: 255-274[CrossRef]
Douady S. Y. Couder 1996b Phyllotaxis as a dynamical self organizing process. Part II: The spontaneous formation of periodicity and the coexistence of spiral and whorled patterns. Journal of Theoretical Biology 178: 275-294[CrossRef]
Douady S. Y. Couder 1996c Phyllotaxis as a dynamical self organizing process. Part III: The simulation of the transient regimes of ontogeny. Journal of Theoretical Biology 178: 295-312[CrossRef]
England W. H. R. J. Tolbert 1964 A seasonal study of the vegetative shoot apex of Myriophyllum heterophyllum. American Journal of Botany 51: 349-353[CrossRef][ISI]
Erickson R. O. 1983 The geometry of phyllotaxis. In J. E. Dale and F. L. Milthrope [eds.], The growth and functioning of leaves, 53–88. University Press, Cambridge, UK
Esau K. 1965 Plant anatomy, 2nd ed. John Wiley & Sons, New York, New York, USA
Gifford E. M. 1943 The structure and development of the shoot apex of Ephedra altissima. Bulletin of the Torrey Botanical Club 70: 15-25[CrossRef]
Gifford E. M. 1954 The shoot apex in angiosperms. Botanical Review 20: 477-529
Gifford E. M. A. S. Foster 1989 Morphology and evolution of vascular plants, 3rd ed. W. H. Freeman, New York, New York, USA
Gleason H. A. A. Cronquist 1991 Manual of vascular plants of northeastern United States and adjacent Canada, 2nd ed. New York Botanical Garden, Bronx, New York, USA
Green P. B. 1999 Expression of pattern in plants: combining molecular and calculus-based biophysical paradigms. American Journal of Botany 86: 1059-1076
Green P. B. C. S. Steele S. C. Rennich 1996 Phyllotactic patterns; a bio-physical mechanism for their origin. Annals of Botany 77: 515-527
Greyson R. I. D. B. Walden J. A. Hume 1978 The ABPHYL syndrome in Zea mays. II. Patterns of leaf initiation and the shape of the shoot meristem. Canadian Journal of Botany 56: 1545-1550[ISI]
Heywood V. H. 1993 Flowering plants of the world, updated edition. Oxford University Press, New York, New York, USA
Itoh J. A. Hasegawa H. Kitano Y. Nagato 1998 A recessive heterochronic mutation, plastochron1, shortens the plastochron and elongates the vegetative phase in rice. Plant Cell 10: 1511-1521
Itoh J. H. Kitano M. Matsuoka Y. Nagato 2000 SHOOT ORGANIZATION genes regulate shoot apical meristem organization and the pattern of leaf primordium initiation in rice. Plant Cell 12: 2161-2174
Jean R. V. 1992 On the origins of spiral symmetry in plants. In I. Hargittai [ed.], Spiral symmetry, 323–351. World Scientific, Singapore, China
Jean R. V. 1994 Phyllotaxis, a systematic study in plant morphogenesis. Cambridge University Press, New York, New York, USA
Jean R. V. D. Barabé 1998 Symmetry in plants. World Scientific, River Edge, New Jersey, USA
Kaplan D. R. 2001 The science of plant morphology: definition, history, and role in modern biology. American Journal of Botany 88: 1711-1741
Kaplan D. R. T. J. Cooke 1997 Fundamental concepts in the embryogenesis of dicotyledons: a morphological interpretation of embryo mutants. Plant Cell 9: 1903-1919[CrossRef][ISI][Medline]
Koller D. H. R. Highkin O. H. Caso 1960 Effects of gibberellic acid on stem apices of vernalizable grasses. American Journal of Botany 47: 518-524[CrossRef][ISI]
Kwiatkowska D. 1999 Formation of pseudowhorls in Peperomia verticillata (L.) A. Dietr. shoots exhibiting various phyllotactic patterns. Annals of Botany 83: 675-685
Leyser H. M. O. I. J. Furner 1992 Characterization of three shoot apical meristem mutants of Arabidopsis thaliana. Development 116: 397-403[ISI]
Lyndon R. F. 1998 The shoot apical meristem: its growth and development. Cambridge University Press, Cambridge, UK
Matsuo T. K. Hoshikawa 1993 Science of the rice plant, vol. 1. Morphology. Food & Agriculture Policy Research Center, Tokyo, Japan
McCully M. E. H. M. Dale 1961 Variations in leaf number in Hippuris. A study of whorled phyllotaxis. Canadian Journal of Botany 39: 611-625
Meicenheimer R. D. 1998 Decussate to spiral transitions in phyllotaxis. In R. V. Jean and D. Barabe [eds.], Symmetry in plants, 25–143. World Scientific, River Edge, New Jersey, USA
Meinhardt H. 1982 Models of biological pattern formation. Academic Press, London, UK
Mitchison G. H. 1977 Phyllotaxis and the Fibonacci series. Science 196: 270-275
Nagasawa N. M. Miyoshi H. Kitano H. Satoh Y. Nagato 1996 Mutations associated with floral organ number in rice. Planta 198: 627-633[ISI]
Niklas K. J. 1998 Light harvesting "fitness landscapes" for vertical shoots with different phyllotactic patterns. In R. V. Jean and D. Barabe [eds.], Symmetry in plants, 759–773. World Scientific, River Edge, New Jersey, USA
O'Brien T. P. M. E. McCully 1969 Plant structure and development, a pictorial and physiological approach. MacMillan, New York, New York, USA
Paolillo D. J., Jr. E. M. Gifford Jr 1961 Plastochronic changes and the concept of the apical initials in Ephedra altissima. American Journal of Botany 48: 8-16[CrossRef][ISI]
Qui Y. J. Lee B. A. Whitlock F. Bernasconi-Quadroni O. Dombrovska 2001 Was the ANITA rooting of the angiosperm phylogeny affected by long-branch attraction?. Molecular Biology and Evolution 18: 1745-1753
