| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Structure and Development |
2Department of Ecology and Evolutionary Biology, 75 North Eagleville Road, Unit 3043, University of Connecticut, Storrs, Connecticut 06269 USA 3Department of Biology, Indiana University, Bloomington, Indiana 47405 USA
Received for publication October 27, 2000. Accepted for publication February 8, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Berberidaceae • clonal growth • heteroblasty • leaf development • mayapple • ontogenetic contingency • Podophyllum peltatum • preformation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
), in the size or shape of the axis, i.e., the phenomenon of Erstarkungswachstum (strengthening growth) (Troll and Rauh, 1950
), and in rooting ability, etc. (Allsopp, 1965, 1967
). In perennials, variation among metamers that arises within growing seasons, i.e., seasonal heteroblasty (Jones, 1999
), may be superimposed on patterns of whole-plant variation.
The study of heteroblasty at the whole-plant level provides critical insights into (1) functional differences among vegetative metamers (Kearsley and Whitham, 1997
; Diggle, 1999
; Karban and Thaler, 1999
), (2) developmental links between heteroblasty and phase change (Kerstetter and Poethig, 1998
; Jones, 1999
), and (3) the molecular genetic regulation of leaf shape (Goliber et al., 1999
). Seasonal shoot heteroblasty has received less recent attention, despite the sometimes dramatic variation in leaf form, e.g., variation between bud scales and foliage leaves. The functional consequences of seasonal heteroblasty may be readily apparent, but less obvious differences also have been found among foliage leaves produced at different times within the growing season (Winn, 1999
).
The degree to which the developmental basis of whole-plant heteroblasty parallels that of seasonal heteroblasty has yet to be thoroughly studied in single species. A presumed link between the two modes of heteroblasty was first suggested by Goebel (1880)
, who proposed that a process of "arrested" foliage leaf development could account for both the morphologically simpler forms of seedling leaves and the simple nature of bud scales. Detailed developmental studies of heteroblasty at the level of the whole plant (Franck, 1976
; Mueller, 1982
) and seasonally within shoots (e.g., Foster, 1929
; Kaplan, 1973
) have failed to provide support for a fundamental tenet of Goebel's hypothesis of arrest, i.e., that primordia are developmentally identical until growth is "arrested" at different points in the different leaves. Nonetheless, Goebel's hypothesis was key in its recognition of how differences in timing of developmental processes can result in divergent forms deriving from similar points of inception.
Interest in the developmental relationship between foliage leaves and bud scales in temperate plants is longstanding (see Foster, 1928
). Eichler (1861)
proposed that following the formation of the initial leaf primordium, or Primordialblatt, the developing primordium becomes divided into two regions. The lower leaf zone, or Unterblatt, comprises the broader, basal region of the leaf that is confluent with the stem. The upper leaf zone, or Oberblatt, is the freely projecting distal region that gives rise to the lamina and ultimately the petiole (original not seen; as cited in D. R. Kaplan, University of California, Berkeley, unpublished manuscript; Foster, 1928
). Morphological comparisons of mature leaves along the shoot reveal how different regions may become elaborated among leaves with different final forms (e.g., Troll, 1937
), but developmental investigations of the timing of onset and differentiation of different leaf zones are necessary to show how that elaboration is achieved and the extent to which common modes of development are shared.
An additional facet of seasonal heteroblasty is the common occurrence of preformation. Preformation may be complete, wherein all organs that expand in a growth season are initiated in the preceding year or earlier, or preformation may be partial, wherein some organs that expand during a given growing season are initiated in the previous year or years (e.g., preformed leaves) and some are initiated in the current year (e.g., neoformed leaves). Preformation has profound ecological consequences, as it may limit a plant's ability to adjust its growth to current environmental conditions, but it also allows relatively rapid expansion at the beginning of the next growth season (Diggle, 1997
; Geber, Watson, and de Kroon, 1997
; Watson, Hay, and Newton, 1997
). Likewise, preformation may present a phylogenetically determined developmental constraint (see also Geber, Watson, and de Kroon, 1997
)—at present, an area in need of investigation. Many species that exhibit strong seasonal heteroblasty also have some degree of preformation, but an absolute dependence of one condition on the other is not essential, e.g., some species exhibit preformation but not marked leaf heteroblasty (e.g., Aydelotte and Diggle, 1997
; Diggle, 1997
; Meloche and Diggle, in press). Nevertheless, in species that exhibit both, the relationship between heteroblastic leaf development and the seasonal schedule of leaf initiation and development remains unexplored. For example, does one constrain the other?
The herbaceous perennial mayapple, Podophyllum peltatum L. (Berberidaceae), presents an interesting example of an herbaceous plant with strong seasonal shoot heteroblasty and preformation. First, mayapple has an unusual aboveground morphology. Each spring one of two types of structures emerges: the vegetative "shoot" consists only of a single, manifestly peltate leaf with an orbicular blade atop an elongate petiole (the shoot meristem remains below ground; see Fig. 1I); the reproductive "shoot" is truly a shoot bearing two latently peltate leaves and a terminal flower (Fig. 1M). From year to year, an individual rhizome system can alternate between producing vegetative and sexual aerial structures (Foerste, 1884
). Second, these structures are borne on rhizomes that produce scale leaves prior to initiating the aerial structure (Holm, 1899
), so seasonal shoot growth is heteroblastic. Third, these aerial structures are present at the end of the preceding growing season and, thus, are preformed sometime during that season, or earlier (Bastin, 1894
). Finally, in mayapple, as in many plants exhibiting preformation, the developmental "decision" to produce a single vegetative leaf on a determinate vegetative shoot, vs. a reproductive shoot, has profound consequences for the ecology of individual rhizome systems. That is, the type of aerial shoot produced determines the reproductive status (and associated carbon loss) of the rhizome for the entire growing season, as well as the plant's likelihood of branching (Geber, Watson, and de Kroon, 1997
) and its potential to acquire carbon because sexual shoots senesce later and have greater leaf areas (Benner and Watson, 1989
; Geber, de Kroon, and Watson, 1997
; Watson and Lu, 1999
).
|
; Meijer, 1974
), no complete developmental study of the entire seasonal growth cycle exists. However, DeMaggio and Wilson (1986)
have provided a detailed study of floral organogenesis and anatomy. Here we report the development of the annual rhizome segment from its inception through initiation and determination of all organs prior to dormancy. There are a number of reasons for our interest in the development of these metamers. First, the analysis provides information about the developmental basis of leaf heteroblasty in mayapple. Where do developmental pathways of different leaf homologs diverge and converge? Second, the analysis will clarify the relationship between the sequence of leaf shapes produced along the shoot and shoot type determination. That is, it can address the question of whether the type, number, or sequence of leaf homologs produced prior to initiation of the emergent structure depends on shoot status (sexual or vegetative). As a corollary, we can determine when sexual shoots are morphologically distinguishable from vegetative shoots. Third, we are able to examine the relationship between heteroblasty and preformation. While preformation severely constrains the morphology of the emergent aerial structure in the year of emergence, preformation does not necessarily limit developmental plasticity up to that point. In other words, can the sequence of heteroblastic leaves be used to indicate points where shoot (i.e., rhizome) development is plastic, prior to its final dormancy before emergence?
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Aerial structures (N = 365) were randomly assigned to harvest dates as they emerged in late March. We removed all flowers on sexual shoots on 25 April to increase the potential number of sexual renewal shoots in our collections (see Geber, de Kroon, and Watson, 1997
). Approximately 12 aerial structures and their attached rhizome segments (previous, current, and renewal) were harvested every 5–7 d until mid-May and every 3–5 d from mid-May until the end of July. If the designated rhizome system was damaged, a substitute of the same shoot type, i.e., vegetative or sexual, was chosen haphazardly from nearby plants. On branched rhizomes, if the two renewal shoots differed in size, the largest shoot in the forward-growing position was analyzed.
We determined the number and position of all leaves initiated on two years of rhizome segments: current rhizome segments and renewal rhizome segments (see below). In current rhizome segments, the positions of all leaves were determined by counting leaf scars. In renewal rhizome segments, the number of leaves on the elongating rhizome segment were counted and added to the number of leaves observed in histological sections of shoot apices to yield a total number of leaves initiated by the shoot at the time of collection. The shoot apices of the renewal rhizome segments were fixed in formalin, acetic acid, and alcohol (FAA; Berlyn and Miksche, 1976
). Apices were embedded in Paraplast, sectioned at 10 µm and stained with safranin and fast green (Berlyn and Miksche, 1976
). Apices viewed with scanning electron microscopy were dehydrated through a graded ethanol series, critical point dried, sputter coated with gold, and viewed in a Zeiss DSM 982 Gemini scanning electron microscope (SEM) (Thornwood, New York, USA) at 20 kV.
Measurements of meristems and leaf primordia were made using an ocular micrometer on a Zeiss Axioskop (Thornwood, New York, USA). Minimum and maximum meristem phase diameters were measured from median longitudinal sections of shoot apices that had just initiated leaf 8 (a representative rhizome scale leaf), leaves 10 and 11 (bud scale leaves), and leaves 12 and 13 for leaves at these positions. Heights of leaf primordia were measured from median sections; the height was measured as the distance between the central, upper surface of the leaf primordium, and its base, determined at the center of its diagonal insertion on the stem.
All descriptive statistics were calculated using SAS (1990)
. Means and standard deviations of positions along the rhizome segment of renewal buds and foliage leaves were calculated using PROC TTEST, and Wilcoxon rank sum scores were calculated using PROC NPAR1WAY.
Line drawings of mature morphology (Fig. 1) were done "by eye." Camera lucida drawings were made from hand sections stained with phloroglucinol using a Wild M10 (Leica Microsystems, Bannockburn, Illinois, USA) dissecting microscope. Photomicrographs were taken with an Olympus OM4T (Melville, New York, USA) mounted on the Axioskop.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
), the series of physically attached, annual rhizome segments. At this site, individual rhizome systems consisted of an average of seven annual segments; as renewal shoots were produced distally, the oldest annual rhizome segment decayed proximally (Landa et al., 1992
). Most commonly, rhizomes were linear; renewal rhizome segments tended to arise in a forward-growing direction. In the less common bifurcating or "branched" rhizome systems (e.g., Fig. 3), two or rarely three renewal rhizome segments were formed at the distal tip of a distal segment. Roots formed on the ventral surface, but only distally, in the region of growth curvature (Fig. 3; Bastin, 1894)
. Roots did not senesce annually, but instead persisted until the decay of the rhizome segment on which they were borne.
|
; here we use "shoot" to refer to the entire annual rhizome segment and its aerial structure. Rhizome segments were elongate, plagiotropic axes that reorient to orthotropic growth in the distal region (Fig. 3). Longitudinal shoot symmetry was mesotonic (Troll, 1937
), wherein a few proximal short internodes with bilateral transverse symmetry were followed by longer internodes with radial transverse symmetry and ultimately by short internodes with dorsiventral transverse symmetry in the region of upturn. By the end of the growing season, each current rhizome segment bore a series of rhizome scales, bud scales, and the aerial vegetative foliage leaf or a flowering shoot in the distal region of upturn (schematized in Fig. 2). The aerial structure senesced in mid- to late summer (Watson and Lu, 1999
), leaving behind a scar that revealed whether the aerial structure was vegetative or reproductive (these regions of upturn have been referred to as "nodes" in the literature).
|
.) The lamina was supported by an elongate, centrally inserted petiole that extended to lengths >30 cm above ground in many colonies. Radial transverse symmetry of the unifacial petiole was accompanied by an endoscopic orientation of xylem in all vascular bundles (Figs. 1J–L; see also Ellis and Fell, 1963
). At its juncture with the lamina, the petiole retained its radial symmetry and overall orientation of vascular bundles, although individual vascular bundles shifted in position in association with the developing lamina (Fig. 1J). At the point of its transition into the leaf base, the xylem in all vascular bundles remained oriented inward as the leaf base became dorsiventally symmetrical where it encircled the upturned shoot apex. The shoot apex initiated one to a few additional leaves before becoming dormant; this bud remained dormant unless the forward growing portion of the rhizome was damaged.
Foliage leaves on sexual shoots
If the annual rhizome segment was sexual, the plagiotropic shoot reoriented to orthotropic growth and underwent considerable internodal elongation between the last bud scale and the first foliage leaf; this internode elevated both foliage leaves and the single terminal flower above ground (Figs. 1M, 2). Within this elongate internode, the xylem poles of vascular bundles were oriented endoscopically in a complex concentric pattern (Fig. 1Q), in contrast to the plagiotropic portion of the rhizome segment where the vascular bundles were arranged concentrically (Fig. 1R; Bastin, 1894)
. Each lobed lamina could range from irregularly orbicular to reniform in outline, with petioles inserted at the base of the lamina (Fig. 6). In transection, laminas of sexual foliage leaves are indistinguishable from those of vegetative foliage leaves. Petioles were relatively short, ranging from 8 to 15 cm in length. Petioles were dorsiventral in transectional symmetry at their point of attachment to the lamina (Fig. 1N), reflecting the latently peltate insertion of the lamina. The unifacial petiole morphology, indicated by the radial symmetry of the petiole at midlength (Fig. 1O), was accompanied by an endoscopic orientation of vascular bundles that extended into the apical region of the petiole (Fig. 1N). The leaf base of sexual foliage leaves did not completely encircle the shoot axis. Instead, its rounded dorsiventral symmetry reflected the small diameter of the floral axis and the nearly simultaneous insertion of the leaf base of the second foliage leaf (Fig. 1P).
Bud scales
Two to several vaginate bud scales protected the vegetative foliage leaf or sexual shoot as it overwintered and during emergence above ground the following spring. When the renewal shoot became dormant in the fall, bud scales were generally 2–3 cm long, forming an imbricate sheath of scales that surrounded the aerial structure (Figs. 1F, 4). Along their length, bud scales exhibited dorsiventral transverse symmetry (Fig. 1G, H); vascular bundles were arranged in a single plane, resulting in a concentric arrangement in transection (Fig. 1H). Vascular bundles converged at the apex as the scales tapered to a short, nearly unifacial sector distally (Fig. 1G), often bearing tiny lobes of a vestigial lamina; this lamina is more apparent at early stages of scale development, prior to expansion (e.g., Fig. 37). In the following spring, bud scales often doubled in length as they protected the expanding aerial structure, becoming membranous prior to withering (Fig. 3).
|
Seasonal growth of the annual rhizome segment
Axillary buds that gave rise to renewal shoots could be distinguished from other axillary buds in the region of rhizome upturn by their larger size and position in the forward direction of the rhizome (Fig. 3, RS). These renewal buds were actually initiated during the year before their expansion into renewal rhizome segments (e.g., Fig. 4, RS), thus 2 yr prior to functioning as a current rhizome segment bearing an expanded aerial structure.
Two years prior to the "current" year, the renewal rhizome bud was initiated as a small mound of densely staining meristematic cells on the axis of the renewal shoot, positionally opposed to the adaxial surface of the subtending leaf. Surrounding this pocket of cells was an arcuate shell zone, characterized by a pattern of radially arranged, anticlinally dividing cells. Subsequent development of the axillary bud was characterized by an increase in height of the dome of the bud, a restriction of meristematic activity into a cluster of cells opposite the subtending leaf, near the juncture of the leaf and the rhizome axis, and a clearly delimited morphogenetic zone (Fig. 7). Shortly after initiation, the first leaf, or single prophyll, is initiated in an adorsed position.
In mid-March of the current year, at the time the aerial structures borne on the current rhizome segments were beginning to emerge above ground, renewal rhizome buds were <0.5 cm in length (e.g., DeMaggio and Wilson, 1986
: Fig. 1). By early April, these buds bore six to seven scale leaves, but the shoot axis had not yet begun to elongate (Figs. 8 and 9A, B). Elongation of the plagiotropic rhizome segments occurred though expansion of the internodes separating rhizome scale leaves following their initiation. During rhizome elongation and subsequent development, scale leaves no longer formed a tight-fitting helmet protecting the shoot apex, but instead became stretched into an appressed sheath surrounding the rhizome axis (Figs. 1A, 3). The number and position of scale leaves could be determined from the scars remaining after the sheath senesced. Elongation of the entire rhizome segment could be quite variable. In this colony, current rhizome segment length ranged from 0.7 (i.e., very little elongation between rhizome scales) to 14.1 cm in vegetative shoots and from 1.5 to 16.2 cm in reproductive shoots.
Despite variation in internode lengths between rhizome scales, the number of scale leaves initiated prior to the initiation of bud scales in the region of upturn was relatively constant, ranging in this colony from seven to nine among most rhizome segments in most years. Leaf initiation occurred at a relatively constant rate through the transition from rhizome to bud scale leaves, as shown by continuous increase in leaf number with time from leaf positions 8 through 12 (Fig. 9A). Based on the mean number of scale leaves initiated on shoots collected from mid-April to early June, the length of the plastochon between successive leaf initiations was 
10 d. During this period, rhizome segments were elongating rapidly as internodes began to expand between previously initiated leaves (Fig. 9B).
|
In current rhizome segments, the renewal axillary bud was located in the axils of leaves 7–12 in vegetative shoots, but occurred most frequently (51%) in the axil of the 10th leaf (mean = 9.95) (Fig. 10A). Renewal buds also occurred most frequently (48%) in the axil of the 10th leaf in sexual shoots, but were also found frequently (36%) in the axil of the 11th leaf (mean = 10.38). Both the mode and the mean position of renewal buds in the current year were significantly different between shoot types: renewal buds occurred at one leaf position later in sexual shoots (Wilcoxon rank-sum test of modal differences: Z = 3.91; P < 0.0001).
|
Following bud scales, the apical meristem could initiate a leaf primordium that developed into a vegetative foliage leaf, followed by additional primordia that quickly became restricted in their growth as the meristem became dormant. Alternatively, the shoot apex could initiate two leaves in short succession and then undergo a transition to a floral meristem. In order to assess whether an initiating primordium at a given position was likely to become a vegetative foliage leaf, or a bud scale leaf on a yet-to-be-determined shoot, we again used positions of vegetative foliage leaves vs. the first leaf of sexual shoots borne on current shoots to estimate the most likely positions of these leaves in renewal rhizome segments. In the current rhizome segment, vegetative foliage leaves occurred over a range of positions from 10 to 15, with 39.5% occurring modally at position 13, and 32% occurring at position 12 (Fig. 10B). In contrast, the first leaf of sexual shoots occurred as early as position 12 and as late as position 16, with 39.4% modally at position 14 (mean = 13.9). Thus, both the means and the modal positions of the vegetative foliage leaves were significantly earlier than the first leaf on sexual shoots (Wilcoxon rank-sum tests: Z = 8.26, P < 0.0001).
Similar patterns were observed in determined renewal rhizome segments that were collected later in the growing season. In all renewal rhizome segments (except one), vegetative foliage leaves occurred at positions 11–15, with 39% at a mode of position 12 (mean = 12.65) (Fig. 10D). The first foliage leaves on sexual shoots occurred as early as position 12 and as late as position 16, with 47% at a mode of position 14 (mean = 13.97). Again the difference in the position of the foliage leaf differed significantly between shoot types (Wilcoxon rank-sum test of difference in mode: Z = 5.07, P < 0.0001; t test of differences in means: t = 6.11; P < 0.0001, df = 83).
Development of the vegetative foliage leaf
We begin with a description of development of vegetative foliage leaves because these leaves represent the fullest expression of all regions of the leaf and thus provide the framework for interpreting the comparative development of scale leaves and sexual foliage leaves. Because leaves at positions 12 and 13 had the highest probability of differentiating into vegetative foliage leaves, the following description of vegetative foliage leaf development concentrates on leaves at those positions.
The phyllotactic arrangement of leaves in positions 12–13 and higher was examined to determine whether a consistent shift in leaf divergence angle preceded vegetative foliage leaves, perhaps indicating that the meristem altered its rhythm of leaf initiation between production of bud scale leaves and vegetative foliage leaves. However, such a shift was not detected in renewal rhizome segment apices that produced a vegetative foliage leaf (recognizable from serial transverse sections). Leaf divergence angles between the earlier bud scale leaf and the vegetative foliage leaf were variable, ranging from 180° to 137° (e.g., Fig. 11).
|
Superimposed on tunica-corpus organization was clear cytohistological zonation, with a prominent initial zone of enlarged, rectangular vacuolate cells with weakly staining cytoplasm and chromatin (Fig. 12). The initial zone was roughly hemispherical in shape, with the surface at the apical summit spanning 7–9 cells of T1 and T2 and extending centrally to a depth of about six cells, thus encompassing a few layers of cortical initials as well. The morphogenetic zone was characterized by smaller, more densely staining cells with chromatic nuclei. This zone flanked the initial zone, where distally it included randomly oriented corpus cells, and more proximally it included radial files of the rib and peripheral meristems.
Leaf primordia at positions 12 and 13 were initiated as thickened lateral lips on the flanks of the apical meristem (Fig. 12). Growth of primordia at positions 12 and 13 was initially apical, dominated by vertical extension. The procambium differentiated acropetally toward the densely cytoplasmic distal surface of the primordium. By a height of 150 µm, the upper leaf zone began to thicken adaxially. Transverse sections through slightly older leaves (250 µm in height) show that this thickening resulted from the formation of a cross zone of meristematic cells (Fig. 17, CZ) that later rounded out the proximal petiolar region.
|
250 µm in height, the lateral flanks of the lower leaf zone extended more than halfway around the circumference of the meristem (Fig. 20) and the primordium was dominated by vertical extension of a rounded but slightly asymmetrical upper leaf zone (Figs. 20, 21). This asymmetry of the upper leaf zone was due to the apically oriented growth of the lamina and the concomitant adaxial extension of the proximal region of the incipient petiole over the shoot apical meristem. Primordia of this height exhibited an acropetal gradient of longitudinal differentiation, such that distal zone of the leaf was more or less uniformly meristematic (Fig. 21).
|
|
600 and 700 µm, most primordia were characterized by densely staining cells in the lamina and subjacent, radially thickened petiole (Fig. 28). However, among rhizome segments, there was variation in the degree of staining density of developing leaves at this stage.
Leaves that were clearly determined to be vegetative foliage leaves could be distinguished by the time they reached 700–800 µm in height. Morphologically, vegetative foliage leaves exhibited further elaboration of the lamina, with pronounced, downward growth of the leaf lobes (Fig. 34). At its juncture with the axis, the leaf base was relatively narrow; the radial diameter of the lamina equaled or exceeded the diameter of leaf base. Longitudinal sections through vegetative foliage leaves at this stage show elongation of petioles and considerable meristematic activity in the distal region of the petiole. This activity eventually became restricted to a band of meristematic cells just beneath the lamina (Fig. 37). In addition, the leaflets were elongate with densely cytoplasmic tips, indicating gradual centrifugal differentiation of the lamina.
Growth of subsequently initiated primordia was inhibited such that these primordia remained relatively small and buried in the base of the petiole of the foliage leaf as it increased in size (Fig. 37). Subsequent growth of the vegetative foliage leaf was characterized by increasing elongation of the petiole and concurrent downward expansion and elaboration of the leaflets.
Development of bud scales
Because leaf primordia at positions 10 and 11 had the highest probability of developing into bud scale leaves, primordia at these positions were the focus of the developmental investigations. The phyllotactic arrangement of bud scale leaves was alternate, with varying divergence angles below 140° between successive leaves (Fig. 13). Within an individual renewal rhizome segment, angles of leaf divergence among bud scales varied considerably and resulted in an irregular helical pattern by which these leaves came to fully enclose the terminal shoot apex.
During initiation of bud scale leaves, the shoot apical meristem was slightly broader in diameter than in vegetative foliage leaves, ranging from 210 µm in the minimal phase to 285 µm in the maximal phase. Two tunica layers characterized by anticlinal cell divisions were apparent, and the corpus incorporated five or six subtending layers of cells (Fig. 14). Cytohistological zonation overlay tunica-corpus organization, as described earlier for vegetative foliage leaves, although the initial zone was less deep.
Because bud scale leaf initiation occurred on the flanks of the apical dome, it appeared confluent with the band of smaller, densely cytoplasmic cells of the morphogenetic zone (Fig. 14). Young bud scale primordia showed less pronounced apical growth than leaves produced at later positions, but still exhibited some meristematic activity of the apical summit of the primordium. At a height of 160 µm, the primordium was a dorsiventral, thickened peg showing adaxial thickening in the medial plane (Fig. 14). Procambium extended acropetally into the distal region of primordium, terminating in a zone of nearly uniform, cytoplasmically dense cells. A small cluster of lightly staining cells flanked by more densely staining cells (indicated by an arrow in Fig. 14) revealed the inception of differentiation of the central zone of the incipient lamina.
As the primordium neared 250 µm in height, the distal, uppermost surface tapered to a rounded apex (Figs. 22, 23). In longitudinal section, the distal region of the leaf was characterized by the presence of densely staining cells, except in the central upper surface of the primordium (Fig. 23, arrow). The central region of the upper surface, which corresponded to the cluster of more vacuolate cells in the younger primordium, was often apparent as a small depression in the center of the distal surface. The apical zone of densely staining cells marked the initial expression of the distal surface of the lamina. Procambium differentiated acropetally toward the apical point, delimiting the differentiating ab- and adaxial surfaces of the leaf. Transverse sections through a slightly smaller (200 µm) bud scale primordium revealed a meristematic cross zone giving rise to adaxial files of cells in the median plane (Fig. 18).
As the leaf primordium increased in height beyond 500 µm, the lamina became distinguished from the subjacent regions of the leaf by a distinct constriction or sinus (Fig. 29), similar to, although not initially as pronounced in, leaves at later positions (compare Figs. 26 and 29). In leaves at positions 10 and 11, the leaf base is broader in diameter than the periphery of the incipient lamina, such that the entire leaf is more conical than cylindrical (compare Figs. 34 and 35). The peripheral lips (margins) of the lamina remained moderately cytoplasmic, while a region of vacuolate cells, variable in size among bud scale leaves, reflected centrifugal differentiation of lamina relative to the margins (Fig. 30, arrow). The extent of morphological expression of leaflets in a rudimentary lamina was more variable among bud scales and reflected differences in the relative time of differentiation of cells in the margins. Subsequent development was characterized by increased cell expansion and vacuolation throughout the leaf (Figs. 31, 38), whereas leaves at later positions, but at the same stage of development, were more cytoplasmic in the distal regions (compare Figs. 28 and 31). Eventually, the lamina ceased development and became a shriveled apical cap as subsequent development of the bud scale was dominated by expansion of the lower leaf zone.
Developmental flexibility of leaves at positions 11 and higher
Because a young primordium at position 12 or 13 could differentiate into a vegetative foliage leaf if the shoot were to become vegetative, or into a bud scale leaf if the shoot remained undetermined, there was no way to know for certain which type of leaf would have resulted from the primordia described. Therefore, it is possible that some of the micrographs described earlier of vegetative foliage leaves are actually pictures of bud scales. We have examined nearly 200 nonflowering apices from several different colonies with developing leaves at comparable positions and have found no observable differences in shape, position, or internal structure between bud scale leaves at positions 12 and 13 and vegetative foliage leaves until these primordia reached heights of 700 µm. Near a height of 700 µm, vegetative foliage leaves began to exhibit downward growth of the lamina and an elongation and continued increase in diameter of the petiole (Fig. 34). In longitudinal section, leaves had densely cytoplasmic cells at tips of lobes and in the distalmost region of the petiole (Fig. 37). In contrast, leaves that would develop into bud scale leaves at this position continued on the same course of development as those at position 10: cell vacuolation proceeded acropetally within the lower leaf zone and centrifugally within the lamina as the leaf increased in size (e.g., Fig. 38). Centrifugal vacuolation of the lamina led to cessation of its growth. There was little elongation of the petiole, and the diameter of the proximal leaf base increased in size relative the distal region.
The height of the leaf at which these differences became apparent and the degree to which these differences were expressed were somewhat variable among developing leaves at this position, resulting in a large number of leaves with intermediate characteristics between 700 and 1200 µm. Thus, it was not always possible to ascertain the final fate of leaves at these positions until they were somewhat larger or unless the meristem had become floral (described below). For example, the primordium shown in Fig. 40 was relatively narrow at its base with a broad lamina, but the downward growth of the leaflets was not as pronounced in other primordia of similar height (e.g., Fig. 34). On the other hand, some primordia exhibited some elaboration of the lamina but a relatively broad leaf base (Fig. 41). Even bud scale leaves at this position borne on determined sexual shoots, i.e., leaves in which the final fate was known to be that of a bud scale leaf, exhibited some elaboration of the leaflets of the lamina (Fig. 42).
|
Development of rhizome scale leaves
As mentioned earlier, the apical meristem of a renewal rhizome segment originates in the previous growing season and initiates 6–7 rhizome scales prior to subsequent growth the following spring. During April and early May, rhizome scales were initiated alternately at divergence angles of between 170° and 180°, in a distichous phyllotactic pattern (Fig. 15). The shoot apical meristem was broader and less domed than later in the growing season (Fig. 16). Its minimal diameter prior to the initial protuberance of a new leaf primordium was 230 µm, and its maximal diameter was 340 µm.
Rhizome scale leaves arose as dorsiventral lips low on lateral flank of the apical meristem, at the proximal edge of the morphogenetic zone. Even shortly after initiation, by heights of 150 µm, these primordia appeared less densely cytoplasmic than later leaves, presumably due to lower levels of meristematic activity (Fig. 16, leaf 8). Acropetal differentiation was apparent in both longitudinal and transverse section as vacuolation along the abaxial surface, proximally contiguous with the vacuolating subjacent cortical cells (Figs. 16, 19). Procambium extended into the primordium acropetally as slightly elongate and more densely staining cells in longitudinal section (Fig. 16). A small cluster of four to five epidermal cells on the uppermost surface of the primordium was already more vaculoate than neighboring cells at this stage of development (Fig. 16, arrow).
By a height of 250 µm, the primordium of a scale leaf is distinctly hood-shaped, with a rounded mound on the distal surface and the initial stages of an encircling leaf base (Fig. 24). In contrast to the vegetative foliage leaf and the bud scale, the proximal extension of the adaxial surface is greater while the distal extension is less (compare Fig. 25 with Figs. 21 and 23). This proximal extension results from the activity of a cross zone of meristematic activity (Fig. 19). Concomitantly, densely cytoplasmic cells were restricted to proximal and distal pockets (Fig. 25, asterisks), separated by intervening cells already differentiating (Fig. 25, arrow). This pocket of differentiating cells corresponded to the small cluster of differentiating cells in the same position initially observed in the epidermis of younger primordia (compare Figs. 16 and 25, arrows). Procambium differentiated acropetally into the distal region of the primordium, separating the ab- and adaxial regions of the leaf (Fig. 25). Vacuolation in the abaxial region of the leaf extended acropetally from subjacent cortical cells and was more pronounced than that in the adaxial side.
By a height of 400 µm, the primordium had increased its expansion over and around the shoot apical meristem, with the proximal edge nearly contacting the next youngest primordium (Fig. 32). The rounded distal mound corresponding to the developing lamina in later leaves was less distinct at the same height in rhizome scales. Longitudinal sections revealed the continued expansion of a zone of vaculoating cells in the center of the distal leaf surface (Fig. 33, arrow). In some rhizome scales, continued meristematic activity at the perimeter of the lamina, particularly on the distal edge of the primordium, resulted in the formation of a slightly protuberant margin (Fig. 33).
By a height of nearly 700 µm, developing rhizome scales showed pronounced differentiation of both ab- and adaxial leaf zones in conjunction with acropetally differentiating procambium (Fig. 39). In more mature rhizome scales, the lamina was sometimes manifest as a small, lobed apical cap (e.g., Fig. 36). The expression of lamina lobing varied among rhizome segments.
The transition to flowering
Because there are clear differences from inception in development of the sexual foliage leaves, we were able to distinguish between vegetative (bud scale or foliage) and sexual foliage leaf development. The earliest stages of transition to floral apices were characterized morphologically by a more domed apex and a leaf primordium that, shortly after initiation, appeared adaxially thickened even to the point of an upward orientation (Fig. 45). A deep sinus separated this leaf from the apical meristem, which was also increasing in height (Figs. 45, 47). Histologically, the most notable change in the meristem associated with the transition to flowering was a gradual increase in cytoplasmic density of the previous initial zone, leading ultimately to a loss of cytohistological zonation and the formation of three clear tunica layers (see also, DeMaggio and Wilson, 1986
), characteristic of a mantle-core cell arrangement of an early floral meristem (Esau, 1965
). Elongation of the floral axis occurred beneath the level of insertion of leaf primordia, such that both the floral meristem and the two foliage leaves became elevated above the point of insertion of the previous bud scale leaf (Fig. 47). Subsequent stages of floral development of the apex are described in DeMaggio and Wilson (1986)
.
Foliage leaves on sexual shoots
By a height of 160 µm, the primordium of the first foliage leaf on sexual shoots was distinguishable in longitudinal section from previous primordia by its pronounced median thickness and the sinus mentioned previously that separates this primordium from the apical meristem (Fig. 45). Because these changes in the apex were occurring at the same time that the internode below the first foliage leaf was elongating, there was a relatively low position of insertion of the preceding bud scale. Initiation of the second leaf of the floral axis was first apparent as a broad zone on the opposite flank of the shoot apical meristem (Fig. 47). Both differences in position of initiation of the foliage leaves and changes in the meristem of the floral axis indicated that the apex was destined to become floral prior to any significant development of the second foliage leaf (Figs. 45, 47). While the leaf primordia were initiated in close succession, the first was slightly larger. This size difference persists through subsequent growth and expansion, such that one of the two foliage leaves is commonly larger at maturity (e.g., Fig. 6).
Several features of primordial development distinguished the foliage leaves on sexual shoots by a height of 325 µm: vacuolation along the adaxial region of the leaf, a precocious and pronounced elongation of the petiole, a marked zone of densely staining cells on the distal outer region of the primordium that eventually gave rise to a highly asymmetric, localized region of leaflet initiation, and a leaf base restricted in the degree of lateral extension at the point of leaf insertion (Fig. 46). The procambium of these leaves extended directly into the distal region of meristematic cells (Fig. 46). By a height of 475 µm, the lamina extended asymmetrically away from the shoot apex (Fig. 48). The downward growth from the leaflet tips and the vacuolation of the summit of the primordium were similar to that occurring during development of the lamina of vegetative foliage leaves.
Subsequent development of the floral foliage leaf primordia occurred via the continued downward extension of the leaflets and a comparative restriction in further elongation of the petiole until emergence of the shoot from the soil (Figs. 49, 50). Longitudinal sections through older foliage leaves showed meristematic activity in the leaflets and a meristematic zone at the distal region of the petiole, also similar to that in vegetative foliage leaves.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
]), and (3) the heteroblastic series arises from differential growth of the two leaf zones (Eichler, 1861
; Foster, 1928
; Troll, 1939
; Denne, 1960
; Kaplan, 1973
). We have shown that several aspects of development are common to each of the three leaf homologs produced during vegetative growth, i.e., the rhizome scale leaf, the bud scale leaf, and the vegetative foliage leaf. All of these primordia exhibit early establishment of an encircling and overarching leaf base, as well as some expression of a lamina initiated from the apical region of the primordium. In all primordia, differentiation of the primordium is initially acropetal, and differentiation of the lamina or laminar region is centrifugal. Despite these similarities, primordia of rhizome and early bud scales can be distinguished from each other and from primordia at later positions (12 and 13) at inception or shortly after by differences in position of initiation on the meristem, shape of the young primordium, patterns of differentiation, and subsequent differences in the relative rate and extent of development of the two major leaf zones.
Rhizome scales
Relative to vegetative foliage leaves, rhizome scales rapidly overarch the shoot apical meristem due to precocious expression of the leaf base by early formation of a cross zone and extension of the adaxial surface of the leaf. Delimitation of the procambium occurs in association with a comparatively early differentiation of ab- and adaxial surfaces of the leaf base and reflects the earlier onset of differentiation and maturation in these scales. Initial elaboration of the lamina is delayed, such that it is first apparent as a distal mound on a hood-like primordium. Once the lamina is initiated, vacuolation and maturation proceed rapidly, prior to significant development of leaflets (compare Fig. 39 with Figs. 37 and 38). Thus rhizome scale leaves are characterized by an early expression and differentiation of the leaf base relative to bud scales and vegetative foliage leaves, coupled with a delayed initiation and then very rapid differentiation of the lamina.
Bud scale leaves
In comparison to rhizome scales, bud scales exhibit more pronounced apical growth and a longer period of meristematic activity. The lamina is delimited earlier and grows for a longer period than in rhizome scales, relative to the overall height of the primordium, before centrifugal differentiation causes a cessation of growth of its lobed margins. In many respects, both the shapes of the developing bud scale primordia and patterns of meristematic activity and subsequent differentiation are intermediate between those of the rhizome scale leaf and the foliage leaf.
Later scale leaves and vegetative foliage leaves
Primordia that arise at positions 12 and 13 commonly show more prolonged apical growth and meristematic activity than scale leaves initiated at earlier positions. The lamina, when it arises, occupies a greater proportion of the distal region of the primordium (compare Fig. 21 with Figs. 23 and 25). At the same time, centrifugal spread of differentiation of the central apical surface occurs more slowly as the primordium increases in height (compare Fig. 28 with Figs. 31 and 34), resulting in a diameter that nearly equals the diameter of the leaf base when primordia are between 400 and 700 µm in height.
Until developing leaves at positions 12 and 13 reached a height of 700 and 800 µm, vegetative foliage leaves cannot be distinguished from bud scales (compare Figs. 26, 27, and 28 with known bud scales in Figs. 29, 30, and 31). Once this size is achieved, a suite of features associated with continued growth of the lamina and elaboration of the petiole distinguish the two leaf types. Vegetative foliage leaves exhibit continued downward elongation of the leaflets and corresponding meristematic activity in these leaflets, increases in height of the petiole, the presence of a dense band of meristematic activity in the junction of the petiole and lamina, and a corresponding reduction in growth of subsequent leaf primordia. In contrast, subsequent development of bud scale leaves at this position is characterized by vacuolation and differentiation of the lamina, little expansion of the petiole, and both vertical and horizontal expansion of the leaf base due to increases in cell size (compare Figs. 37 and 38).
Foliage leaves on sexual shoots
In contrast to the common elements of development shared by leaf homologs produced by the vegetative (i.e., nonreproductive) meristem, development of the foliage leaves on sexual shoots is markedly divergent. Primordia on sexual shoots differ in their position of insertion low on the flank of the apical meristem and pronounced apical growth giving rise to a sharp crevice that separates the leaf primordium from the initiating floral bud. Subsequent development of these leaves involves elaboration of the distal region of the leaf, precocious expansion of the petiole, and a more dorsivental initiation of the leaflets relative to vegetative leaves (compare Figs. 48 and 27). Pronounced elaboration of the lamina is accompanied by very little expression of an encircling leaf base; petiole elongation soon ceases prior to bud dormancy in the fall. Each of these post-initiation features of primordial development are similar to, and likely homologous with, precocious expression of the upper leaf zone (lamina and petiole), coupled with delayed expression of the lower leaf zone (leaf base), relative to vegetative leaf homologues. In this regard, sexual shoot foliage leaves share a mode of development similar to that described for laminar bud scales or floral bracts in other species (Troll, 1939
; D. R. Kaplan, University of California, Berkeley, unpublished manuscript), despite their larger size.
Leaf heteroblasty and shoot type determination in mayapple
Mayapple presents an intriguing system for examining the interaction between development, ecology, and life history evolution. A key developmental event, shoot type determination, is discrete and has immediate and future demographic consequences (Geber, de Kroon, and Watson, 1997
) that translate into long-term evolutionary potential for fitness differences among genotypes if they vary in the internal and external cues that affect the developmental decision (Geber, Watson, and de Kroon, 1997
). To begin to investigate these interactions and to fully understand internal and external cues that affect the shoot type determination, we must first understand exactly when shoot type determination occurs during seasonal growth of the renewal rhizome segment. Differences in the sequence or positions of leaf homologs in vegetative vs. reproductive shoots could provide a predictive morphological tool for detecting whether determination occurred prior to the formation of an actual floral meristem or a vegetative foliage leaf.
We found no differences in either the sequence or the nature of development of leaf homologs between the two different shoot types in this study. In all rhizome segments, meristems ultimately giving rise to new renewal rhizome segments are initiated on renewal rhizome segments of the previous year; the first rhizome scale leaves are initiated by these new meristems shortly thereafter. By the time the aerial structures of the now-current rhizome segments expand in the following spring, new renewal buds have initiated 6–7 rhizome scale leaves. During early spring of the following year, meristems of the renewal rhizomes continue to initiate additional rhizome scale leaves that are added prior to internode expansion (compare Fig. 1A and B). All rhizome scale leaves exhibit similar development. Expansion of the renewal rhizome does not begin in earnest until around the time of anthesis, often during late April (Fig. 1B). In early May, during the period of very rapid rhizome elongation (see also Geber, de Kroon, and Watson, 1997
), the shoot apex begins to initiate primordia that ultimately differentiate as bud scale leaves.
Despite the lack of difference in the sequence or type of leaf homologs produced prior to shoot type determination, it is possible that other morphological or physiological traits might indicate that shoots had become determined prior to the appearance of the respective primordia. Destructive sampling precluded us from knowing the eventual fate of the meristems we studied, but had we known these fates, we might have detected differences in sizes of rhizome segments that predicted ultimate meristem fate. It is well documented in mayapple that mature sexual rhizome segments are, on average, larger than vegetative rhizome segments (Sohn and Policansky, 1977
; Benner and Watson, 1989
; de Kroon, Whigham, and Watson, 1991
; Landa et al., 1992
; Geber, de Kroon, and Watson, 1997
). This difference occurred in the colony we examined as well, although in this colony, as in others (referenced above), there was considerable overlap in the ranges of lengths of rhizome segments of the two shoot types. By measuring rhizome elongation nondestructively and repeatedly on the same rhizome segments throughout the season and determining shoot type at the end of the season Geber, de Kroon, and Watson (1997)
found that renewal rhizome segments that ultimately be
