FIGURE 1 - uploaded by Akitoshi Iwamoto
Content may be subject to copyright.
Ceratophyllum demersum L. (A) Shoot in aquatic environment. (B) Vegetative buds covered with prophylls, in the axil of a leaf. (C) Staminate fl ower consisting of numerous stamens and surrounding bracts. (D) Pistillate fl ower consisting of only one pistil with long stigma and surrounding bracts. b, bract; l, leaf; p, prophylls; st, stamen; sti, stigma. Scale bars = 1 cm (A); 500 μm (B–D). Photographs are reprinted from Iwamoto (2012) with permission.
Source publication
Premise of the study:
The phylogenetic position of Ceratophyllum is still controversial in recent molecular analyses of angiosperms, with various suggestions of a sister group relation to all other angiosperms, eudicots, monocots, eudicots + monocots, and magnoliids. Therefore, the morphological characters of Ceratophyllum are important for resolv...
Contexts in source publication
Context 1
... are unisexual and appear not to be subtended by leaves. Staminate fl owers are composed of nu- merous stamens that lack fi laments ( Fig. 1C ), and pistillate fl owers that bear only a single pistil with a long appendage on top of the ovary ( Fig. 1D ; Iwamoto et al., 2003 ). Both staminate and pistillate fl owers are surrounded by a whorl of foliaceous organs, identifi ed either as tepals ( Sehgal and Ram, 1981 ;Endress, 1994 ) or bracts ( Jones, 1931 ;Aboy, 1936 ;Les, 1986Les, , 1993Iwamoto et al., 2003 ). ...
Context 2
... mucilaginous substances that might interfere with observations. develop into lateral branches. Flowers are unisexual and appear not to be subtended by leaves. Staminate fl owers are composed of nu- merous stamens that lack fi laments ( Fig. 1C ), and pistillate fl owers that bear only a single pistil with a long appendage on top of the ovary ( Fig. 1D ; Iwamoto et al., 2003 ). Both staminate and pistillate fl owers are surrounded by a whorl of foliaceous organs, identifi ed either as tepals ( Sehgal and Ram, 1981 ;Endress, 1994 ) or bracts ( Jones, 1931 ;Aboy, 1936 ;Les, 1986Les, , 1993Iwamoto et al., 2003 ). Although six species are recognized within Ceratophyllum , they show few ...
Context 3
... e leaf primordia were initiated in a whorl (white rectangle; Fig. 10A ). At the earliest stage, the primordia were integrated and appeared as an annular bulge, making it diffi cult to distinguish the borders between the ...
Context 4
... -Th e observations of shoot tips lead to the possible view of C. demersum phyllotaxy as originally decussate ( Fig. 10B ). We observed one regular vegetative bud at each node, and, infrequently, another vegetative bud initiated at the opposite side of the node. ...
Context 5
... this study, we propose a hypothesis explaining the shift to the complicated phyllotaxis of Ceratophyllum . Two leaf primordia initi- ate originally at one node at opposite sides and bear two stipule-like appendages each ( Fig. 10C ), which is the same to the phyllotactic pattern of prophylls of Ceratophyllum ( Fig. 4E ). Later, the append- ages become multiplied ( Fig. 10D ), as we can observe in the stipule- like appendages of the prophylls on vegetative shoots ( Fig. 5D ), similar to the leaf development of Limnophila indica (Plantagina- ceae), although the interpetiolar organs are regarded as a part of compound leaves ( Rutishauser, 1999 ). ...
Context 6
... leaf primordia initi- ate originally at one node at opposite sides and bear two stipule-like appendages each ( Fig. 10C ), which is the same to the phyllotactic pattern of prophylls of Ceratophyllum ( Fig. 4E ). Later, the append- ages become multiplied ( Fig. 10D ), as we can observe in the stipule- like appendages of the prophylls on vegetative shoots ( Fig. 5D ), similar to the leaf development of Limnophila indica (Plantagina- ceae), although the interpetiolar organs are regarded as a part of compound leaves ( Rutishauser, 1999 ). Finally, leaf primordia and stipule-like appendage primordia become integrated and initiate at the earliest stage as an annular bulge in Ceratophyllum ( Fig. 10A ), which cannot be separated into 8-10 parts. ...
Context 7
... the append- ages become multiplied ( Fig. 10D ), as we can observe in the stipule- like appendages of the prophylls on vegetative shoots ( Fig. 5D ), similar to the leaf development of Limnophila indica (Plantagina- ceae), although the interpetiolar organs are regarded as a part of compound leaves ( Rutishauser, 1999 ). Finally, leaf primordia and stipule-like appendage primordia become integrated and initiate at the earliest stage as an annular bulge in Ceratophyllum ( Fig. 10A ), which cannot be separated into 8-10 parts. Th e annular bulge may correspond to the bases of two large leaf primordia completely sur- rounding the node ( Fig. 10B ). ...
Context 8
... leaf primordia and stipule-like appendage primordia become integrated and initiate at the earliest stage as an annular bulge in Ceratophyllum ( Fig. 10A ), which cannot be separated into 8-10 parts. Th e annular bulge may correspond to the bases of two large leaf primordia completely sur- rounding the node ( Fig. 10B ). Th is leaf structure and development bears some similarities to that of Galium , where an annular bulge is formed fi rst, then two opposite leaves develop, followed by a num- ber of interpetiolar leaf-like stipules ( Goebel, 1905 ;Majumdar and Pradyot Kumar, 1958 ;Rutishauser, 1999 ), even though two opposite leaves and stipules initiate simultaneously on the annular bulge and cannot be distinguished. ...
Context 9
... our view, the previously described 8-10-merous leaf whorls in Ceratophyllum could be interpreted as consisting of two leaves with additional lat- eral members that are called here stipule-like appendages. In the leaves of Ceratophyllum , leaf appendages have been so completely integrated that we cannot distinguish between the initial two leaves and their lateral appendages, even at the earliest stage ( Fig. 10A ), and each subunit of leaves receives its own vascular trace from the stele ( Schneider and Carlquist, 1996 ). ...
Context 10
... suggest that in Ceratophyllum , the axillary meristem of one leaf integrating two larger leaves develops as a vegetative bud (V 1 , Fig. 10B ), while that of the other leaf is typically repressed ( V 2 , Fig. 10B ) . Loiseau and Grangeon (1963) observed that two oppo- site vegetative buds developed at the same node and one of them was much smaller than the other in Ceratophyllum demersum. Th is case represents an intermediate situation between only one vegetative bud ( Fig. ...
Context 11
... suggest that in Ceratophyllum , the axillary meristem of one leaf integrating two larger leaves develops as a vegetative bud (V 1 , Fig. 10B ), while that of the other leaf is typically repressed ( V 2 , Fig. 10B ) . Loiseau and Grangeon (1963) observed that two oppo- site vegetative buds developed at the same node and one of them was much smaller than the other in Ceratophyllum demersum. Th is case represents an intermediate situation between only one vegetative bud ( Fig. 7A, B ) and two equal vegetative buds ( Fig. 8A, B ) at one node. Th e ...
Similar publications
- [...]
Elucidating the origin of flowers has been a challenge in botany for a long time. One of the central questions surrounding the origin of flowers is how to interpret the carpel, especially the relationship between the phyllome part (carpel wall) and the ovule. Recently, consensus favors the carpel originating from the fusion of an ovule-bearing part...
Citations
... There are two ancient angiosperm lineages-Nymphaeales and Ceratophyllalesthat are conspicuous in the occurrence of deviations from axillary branching, especially with respect to flower arrangement [8,[11][12][13][14][15][16][17][18][19][20][21][22][23][24][25]. Both lineages are aquatic and not sister to each other. ...
... All species of Ceratophyllum are specialized aquatics with underwater pollination [25,[39][40][41][42][43][44][45][46][47]. They lack any traces of roots. ...
... Under this interpretation, leafy appendages of each node belong to two strongly dissected opposite leaves. Further refinement of this interpretation is provided by Iwamoto et al. [25] who interpreted all but central leafy appendage of the opposite leaves as multiplied stipules (stipule-like appendages). On the other hand, Raynal-Roques [22] developed a theory that all appendages of each whorl belong to one and the same dissected leaf. ...
Ceratophyllum is an ancient and phylogenetically isolated angiosperm lineage. Comparisons between Ceratophyllum and other angiosperms are hampered by uncertainty in inferring organ homologies in this genus of specialized aquatics. Interpretation of shoot morphology is especially problematic in Ceratophyllum. Each node has several leaf-like appendages interpreted as verticillate leaves, modified parts of one and the same leaf or parts of two leaves under decussate phyllotaxis. Vegetative branches are axillary, but reproductive units (interpreted as flowers or inflorescences) are commonly viewed as developing from collateral accessory buds. We studied shoot development in Ceratophyllum submersum, C. tanaiticum, and C. demersum using scanning electron microscopy to clarify shoot morphology and branching patterns. Our data support the idea that the phyllotaxis is essentially decussate with appendages of stipular origin resembling leaf blades. We conclude that a leaf axil of Ceratophyllum possesses a complex of two serial buds, the lower one producing a vege-tative branch and the upper one developing a reproductive unit. The reproductive unit is congenitally displaced to the subsequent node, a phenomenon known as concaulescence. Either member of the serial bud complex may be absent. There is a theory based on a synthesis of molecular and morphological data that Chloranthaceae are the closest extant relatives of Ceratophyllum. Serial buds and concaulescence are known in Hedyosmum (Chloranthaceae). Our new interpretation facilitates morphological comparisons between Hedyosmum and Ceratophyllum.
... Consistent with the aquatic habit, the vegetative body of Ceratophyllum lacks stomata or roots at any stage of development; it also lacks xylem conducting elements. The characteristic "leaves" are dichotomously divided 1-4 times with linear denticulate segments; they appear whorled (verticillate), but could plausibly be interpreted as decussate, with several leaf-like organs at each node (Schaeppi, 1935;Iwamoto et al., 2015). The fruits are one-seeded, often with characteristic spine-like appendages. ...
... Even though the pistillate flowers of Ceratophyllum (not considering the involucre) are rather simple structures, their morphological interpretation is far from straightforward. The main problem is a long filiform distal extension to the gynoecium ( Figure 14B), sometimes called the stigma, which is normally attached on the same side as the ovule (Troll, 1933;Shamrov, 1983;Endress, 1994;Igersheim and Endress, 1998;Endress, 2001;Iwamoto et al., 2003;Endress and Doyle, 2015). This extension can be interpreted as median-adaxial in its attachment relative to the main axis on which the flowers develop (but note that flowersubtending bracts are lacking in Ceratophyllum unless all leafy segments in each node are interpreted as parts of the same dissected leaf with a group of collateral axillary buds: Raynal-Roques, 1981;Rutishauser and Sattler, 1987;Rutishauser, 1999;Iwamoto et al., 2015). ...
... The main problem is a long filiform distal extension to the gynoecium ( Figure 14B), sometimes called the stigma, which is normally attached on the same side as the ovule (Troll, 1933;Shamrov, 1983;Endress, 1994;Igersheim and Endress, 1998;Endress, 2001;Iwamoto et al., 2003;Endress and Doyle, 2015). This extension can be interpreted as median-adaxial in its attachment relative to the main axis on which the flowers develop (but note that flowersubtending bracts are lacking in Ceratophyllum unless all leafy segments in each node are interpreted as parts of the same dissected leaf with a group of collateral axillary buds: Raynal-Roques, 1981;Rutishauser and Sattler, 1987;Rutishauser, 1999;Iwamoto et al., 2015). Most commonly, distal carpel extensions are dorsal in basal angiosperms (i.e., angiosperms other than monocots and eudicots). ...
Molecular phylogenetic analyses have revealed a superclade of mesangiosperms with five extant lineages: monocots, eudicots, magnoliids, Ceratophyllum and Chloranthaceae. Both Ceratophyllum and Chloranthaceae are ancient lineages with a long fossil record; their precise placement within mesangiosperms is uncertain. Morphological studies have suggested that they form a clade together with some Cretaceous fossils, including Canrightia, Montsechia and Pseudoasterophyllites. Apart from Canrightia, members of this clade share unilocular gynoecia commonly interpreted as monomerous with ascidiate carpels. Alternatively, the gynoecium of Ceratophyllum has also been interpreted as syncarpous with a single fertile carpel (pseudomonomerous). We investigate patterns of morphological, anatomical and developmental variation in gynoecia of three Ceratophyllum species to explore the controversial interpretation of its gynoecium as either monomerous or pseudomonomerous. We use an angiosperm-wide morphological data set and contrasting tree topologies to estimate the ancestral gynoecium type in both Ceratophyllum and mesangiosperms. Gynoecia of all three Ceratophyllum species possess a small (sometimes vestigial) glandular appendage on the abaxial side and an occasionally bifurcating apex. The ovary is usually unilocular with two procambium strands, but sometimes bilocular and/or with three strands in C. demersum. None of the possible phylogenetic placements strongly suggest apocarpy in the stem lineage of Ceratophyllum. Rescoring Ceratophyllum as having two united carpels affects broader-scale reconstructions of the ancestral gynoecium in mesangiosperms. Our interpretation of the glandular appendage as a tepal or staminode homologue makes the Ceratophyllum ovary inferior, thus resembling (semi)inferior ovaries of most Chloranthaceae and potentially related fossils Canrightia and Zlatkocarpus. The entire structure of the flower of Ceratophyllum suggests strong reduction following a long and complex evolutionary history. The widely accepted notion that apocarpy is ancestral in mesangiosperms (and angiosperms) lacks robust support, regardless of which modes of carpel fusion are considered. Our study highlights the crucial importance of incorporating fossils into large-scale analyses to understand character evolution.
... The effect of pressure of the bract can be expressed on either side of the floral primordium along a median gradient, favoring a polarity in flower sequence initiation and development, and eventually also organ loss. Ceratophyllum presents a good example where the effect of the pressure is more pronounced on the adaxial side of the flower primordium ( Figure 7) [31,32]. Here, the bract subtending a staminate flower overtopping the floral bud presses against the adaxial side and stamens may fail to develop as organs can only initiate unidirectionally from the abaxial to the adaxial side ( Figure 7D-F). ...
... Scale bar in A = 5 mm, in B = 500 µm, in C-F = 50 µm. ((A,B,E,F) reproduced from Iwamoto et al., 2015; (C,D) from Iwamoto et al., 2003)[31,32]. ...
Mechanical forces acting within the plant body that can mold flower shape throughout develop-ment received little attention. The palette of action of these forces ranges from mechanical pres-sures on organ primordia at the microscopic level up to the twisting of a peduncle that promotes resupination of a flower at the macroscopic level. Here, we argue that without these forces acting during the ontogenetic process, the actual flower phenotype would not be achieved as it is. In this review, we concentrate on mechanical forces that occur at the microscopic level and determine the fate of the flower shape by the physical constraints on meristems at an early stage of development. We thus highlight the generative role of mechanical forces over the floral phenotype and underline our general view of flower development as the sum of interactions of known physiological and genetic processes, together with physical aspects and mechanical events that are entangled to-wards the shaping of the mature flower.
... : (Iwamoto et al. 2003(Iwamoto et al. , 2015(Iwamoto et al. , 2012(Iwamoto et al. , 2017) (Iwamoto et al. 2003) , (Iwamoto et al. 2015) ...
... : (Iwamoto et al. 2003(Iwamoto et al. , 2015(Iwamoto et al. , 2012(Iwamoto et al. , 2017) (Iwamoto et al. 2003) , (Iwamoto et al. 2015) ...
The external mechanical forces on a floral meristem play a key role in floral development. A novel experimental system was developed to apply mechanical forces on the floral meristem of Arabidopsis thaliana, causing various morphological changes in floral development. A micromanipulator with a silicon device (= micro device), which is shaped to fit the contour of the abaxial side of the young primordium on a floral meristem, is used to apply contact pressure on the floral primordia. Mechanical forces were successfully applied on the floral primordia using this experimental system, and morphological changes were induced in floral development. One of the important issues in this experimental system is measuring the mechanical forces imposed on the floral primordia, and I tried to apply “Pollination Simulation” for the measurement. “Pollination Simulation” is a device which equips a double-bending load cell as its microforce sensor, and it was originally developed to measure the forces needed to release the lever mechanism in the genus Salvia exerted by pollinators. In this novel experimental system, the mechanical forces on the floral meristem were measured in several nano-Newton order by the device. The measurements depend on the shapes of the micro devices and floral primordia. These results demonstrate that the device “Pollination Simulation” is applicable to measure the mechanical forces on the floral meristem, inducing the morphological changes in floral development in the novel experimental system.
... The retention or recovery of the fifth staminode and its heterochronic shift should be linked with the spatial constriction within the flower bud. Iwamoto et al. (2003Iwamoto et al. ( , 2015 suggested that the initiation and development of floral organs could be suppressed by mechanical pressures on the floral meristem. In fact, the flower buds at early stages in Zingiberaceae including Globbeae observed in this study are tightly packed insides the sepals (e.g., Figures 4D,E, 7D,E), and there could be a strong mechanical pressure between the different elements of the abaxial common primordium, including the labellum and other floral organs, which might cause the suppression of initiation of the abaxial outer androecial primordium and its heterochronic shift (Ronse De Craene, 2018). ...
We observed the floral development of Hemiorchis burmanica and two species of Gagnepainia, which make up the sister group of Globba in Globbeae, as well as a selected number of species of Globba. Our observations revealed that in Gagnepainia, a "fifth staminode," develops as part of the labellum, while it is generally lost in other Zingiberaceae. The fifth staminode, however, was not initiated in the labellum of Hemiorchis. A fifth staminode was also observed in Globba geoffrayi but was not found in all other species of Globba investigated. While the staminode becomes an integral part of the labellum in Gagnepainia, it remains small and almost aborts at later stages of development in G. geoffrayi. These results indicate that the development of a fifth staminode could be a recurring plesiomorphic trait for Globbeae, only retained in G. geoffrayi. The retention of the fifth staminode and its heterochronic shift may be linked with the mechanical constriction within the flower bud. The results may also support the interpretation of an atavistic expression of a once lost staminode.
... Ceratophyllum position within the phylogeny of angiosperms has been controversial for some time; the first phylogenetical analysis placed C. demersum in a sister group of angiosperms. The Angiosperm Phylogeny Group (APG IV) and more recent studies using more data sets but with low support values, have placed Ceratophyllum close to the eudicots (Iwamoto et al. 2015;Chase et al. 2016). ...
Intraspecific variability plays a pivotal role in short and long term responses of species to environmental fluctuations. This variability, expressed through different traits of individuals, can potentially influence species sensitivity to chemical contamination. This intraspecific variability is currently not taken into account in ecotoxicological risk assessment, whereas it can mislead its results. To examine this hypothesis, the importance of intraspecific variability in the response to copper (Cu) was quantified in controlled conditions for three aquatic macrophyte species, Lemna minor, Myriophyllum spicatum and Ceratophyllum demersum. Variations among genotypes of each of these 3 species were compared to interspecific variability. Results have highlighted a significant genotypic variability, whose importance depends on the species considered. Indeed, L. minor demonstrated a low variability, contrarily to M. spicatum whose variability in growth inhibition by Cu was higher than interspecific differences. In order to specify the extent and the mechanisms of genotypic variability in M. spicatum, other experiments involving measurements of life-history traits have been conducted on 7 genotypes exposed to Cu. Results showed that some genotypes were up to eightfold more sensitive to Cu than others (at concentrations ranging between 0.15 and 0.5 mg/L). These differences in sensitivity were partly explained by the traits measured, but physiological or transcriptomic endpoints may explain more precisely the source of these differences in sensitivity. Finally, 3 experiments with fluctuations in nutrient concentrations, light intensity and Cu pre-exposure have demonstrated that phenotypic plasticity plays an important role in L. minor sensitivity to Cu. Indeed, the weakening of individuals, as a result of unfavorable environmental conditions, can lead to a two-fold increase in sensitivity to Cu.[...]
... The regulation of the contortion of the androecium is highly efficient in the case of Melanophylla, as in c. 100 examined flowers of each species studied here, the direction of contortion, when present in the androecium, was precisely fixed relative to corolla handedness. The most obvious causal explanation of stamen contortion is that the contortion of the corolla is already well manifested by the time of widening of the anthers, so that the stamens grow in an asymmetric environment and mechanical forces (Iwamoto, Izumidate & Ronse De Craene, 2015;Bull-Hereñu et al., 2017; created by petals govern asymmetry of stamen development. A similar hypothesis was proposed by Schoute (1935) to explain the influence of positional information from spirally arranged sepals to a contort corolla in species with unfixed direction of petal contortion. ...
Transitions in corolla symmetry are an important aspect of angiosperm floral evolution. Contort petal aestivation is common in several groups of eudicots. In rosids, the direction of overlap between adjacent petals (handedness) of the contort corolla is often labile among flowers in a single inflorescence, but in asterids, handedness is usually stable at a supraspecific level. Taxa with contort corolla are unevenly distributed among asterids, and detailed developmental data are often lacking. We provide the first developmental study of flowers in Melanophylla (Torricelliaceae, Apiales), the only known campanulid with a contort corolla, and demonstrate that the corolla handedness is labile within a single inflorescence. Labile handedness distinguishes Melanophylla from members of lamiids with a contort corolla where handedness is stable. In Melanophylla, the handedness is determined by the arrangement of bracteoles. Petals are asymmetric from early developmental stages. The androecium is also usually contort, with handedness always opposite to that of the corolla. Anthers have broad, flat connectives, which is unusual in asterids. The gynoecium is pseudomonomerous, with the fertile carpel in a left or right-transversal position, depending on the handedness of the corolla and androecium. Symmetry patterns of all floral whorls, including the pseudomonomerous gynoecium, are strongly correlated in Melanophylla, in contrast with the unstable carpel orientation in monomerous gynoecia of Apiales studied so far. The tricarpellate pseudomonomerous gynoecia of Melanophylla and other early divergent Apiales resemble those of Dipsacales.
... Unlike modern Ceratophyllum fruits, which have very short pedicels below the bracts (e.g., see fig. 2 of Csiky et al., 2010 and fig. 1D of Iwamoto et al., 2015), the fruits of Donlesia probably have long gynophores and very short pedicels. ...
A new species of Donlesia (Ceratophyllaceae) from the Early Cretaceous Cheyenne Sandstone of Kansas, USA is reported. The fruits of Donlesia cheyennensis sp. nov. are achenes with four lateral spines, a stylar spine, and a long pedicel. The four lateral spines are arranged on two perpendicular planes. The leaves associated with D. cheyennensis fruits dichotomize three to four times. The fruit morphology of D. cheyennensis differs from Donlesia dakotensis, from the younger Dakota Formation, other fossil and extant members of Ceratophyllaceae in its tetra-radial symmetry, the presence of four lateral spines, and its long pedicel. The presence of Donlesia in the Cheyenne Sandstone extends the fossil record of Ceratophyllaceae to the earliest late Albian and this indicates that the Ceratophyllaceae clade probably diverged from its sister clade at least 105 million years ago in the Early Cretaceous.
... These genomes will be of great significance for comparative work that seeks to elucidate the evolutionary developmental origin of angiosperm wood. Finally, mention should be made of a bizarre rootless aquatic dicot angiosperm, Ceratophyllum (Iwamoto et al. 2015), which lacks xylem, even primary xylem. As the xylogenesis pathway has been deleted in this plant, it represents a "natural knockout" experiment, which might one day be attractive to researchers. ...
Angiosperm trees now rival the largest conifers in height and many species reach over 80 m high. The large tree life form, with extensive secondary xylem, originated with the progymnosperms and gymnosperms in the Devonian and Carboniferous. However evidence suggests that the ancestor of extant angiosperms was not a tree but either a herb or understory shrub. Angiosperm fossil woods are rare in the early Cretaceous but become common in the mid-Cretaceous. The “reinvention” of wood in the Cretaceous produced a novel xylary morphospace that has since been extensively explored by subsequent evolution. Today, large timber trees are absent in the early diverging lineages of the angiosperms, and conventional wood has been lost in the monocots. There are a few timber trees in the magnoliid clade. Most timber trees are in the rosid clade, particularly the fabids (e.g. Fabaceae) but also in the Malvids (e.g. Meliaceae). Timber trees are less common in the strongly herbaceous asterid clade but some important timbers are also found in this lineage such as teak, Tectona grandis (Lamiaceae). Genomic resources for angiosperm trees are developing rapidly and this, coupled with the huge variation in woody habit, make angiosperm trees a highly promising comparative system for understanding wood evolution at the molecular level.
Stipules are generally regarded as the outgrowths of the leaf base in angiosperms. Other interpretations see stipules as independent organs comparable to leaves. Stipules have been recognized as an important trait for plant taxonomy and identification, and there has been great progress in the understanding of their morphology, development, origin, function, and gene regulation over time. Therefore, this review will briefly summarize past research and aims to clarify the occurrence, location, and morphology of stipules in the families recognized by APG IV and reconstruct their ancestral states. Additionally, the developmental morphology of different types of stipule is presented through scanning electron microcopy observations and a survey of the existing literature. The difference between stipules and ligules is discussed in relation to the occurrence of postgenitally or congenitally fused ‘continuous’ stipules. A distinction is made between ‘true’ stipules and ‘pseudostipules’. The origin of stipules at the base of the leaf is explained from different perspectives. About one-third of the families are reported to have stipules, mostly concentrated in the Rosid clade with the highest level of diversity. On the basis of the ancestral state reconstructions, stipules may be absent in the ancestors of angiosperms, but are present in the ancestor of Rosids, with a pair appearing on both sides of the petiole base. The transition between paired and annular stipules is discussed, the latter arising postgenitally or congenitally. Several hypotheses are discussed to explain the abundance of stipules in the Rosid clade and their limited presence in the Asterid clade.
