Plant evo-devo of the gynoecium of Angiosperms (flowering plants)

Plant Evo-Devo of the Gynoecium

Rolf Sattler

Latest version of September 20, 2023.

Keywords: Plant evo-devo, plant morpho evo-devo, gynoecium, carpel, gynoecial appendage, spatial shifting, heterotopy, topographic approach, fossil angiosperms and pre-angiosperms, flower concept


Definition of the Carpel
Do all Gynoecia have Carpels
Redefining the Carpel
A Topographic Approach to the Gynoecium
Plant Evo-Devo, especially Morpho Evo-Devo of the Gynoecium
Gynoecia of Fossil Angiosperms and Pre-Angiosperms
Processes in Plant Evo-Devo (Plant Morpho Evo-Devo) of the Gynoecium
Plant Evo-Devo (Plant Morpho Evo-Devo) of the Gynoecium and the Concept of the Flower

Note: Versions published before August 8th did not include the topographic approach.


To a great extent, plant evo-devo of the gynoecium still implies the classical notion of the carpel as an appendage that bears and encloses ovule(s), which means that a carpel is a closed megasporophyll. However, developmental investigations have demonstrated that although many taxa of angiosperms have classical carpels, in a considerable number of other taxa of angiosperms the placenta or a single ovule arises in the centre of the gynoecium or at its periphery in the axil of gynoecial primordia. To decide whether these gynoecia are carpellate or not, one would have to know how far the gynoecial primordia extend into the centre of the gyoecium. But since it is difficult or impossible to delimit growth centres such as gynoecial primordia, the carpel is redefined as an appendage that encloses ovule(s), which clears the way to a topographic approach, which recognizes different positions of the placentae and ovules. This approach has far-reaching consequences: in plant evo-devo, it underlines the fundamental importance of the developmental and evolutionary process of spatial shifting that leads to heterotopy of the placenta and ovule(s), and, as a consequence, it changes the search for the ancestor(s) of angiosperms. In addition to heterotopy, heterochrony, and several processes leading to heteromorphy also need to be considered in plant evo-devo of the gynoecium and the search for the ancestor(s) of angiosperms.


Plant Evo-Devo investigates the evolution of plant development. It integrates evolutionary theory, morphology, and molecular genetics. MorphoEvoDevo emphasizes the morphological aspects of evo-devo (Wanninger 2015). Hence, plant morpho evo-devo emphasizes the morphological aspects of plant evo-devo (
Petrone-Mendoza et al. 2023). In this article, the focus will be on plant morpho evo-devo. Thus, the focus will be on the morphological evolutionary changes in the development of the gynoecium. According to mainstream thinking, the gynoecium consists of carpels. However, the carpel concept seems to be only of limited usefulness. Therefore, I propose a topographic approach to the gynoecium that overcomes problems and pseudo-problems in the application of the carpel concept.

Definition of the Carpel

According to mainstream thinking (classical morphology), a carpel is an appendage that bears and encloses ovule(s). In other words, it is a closed megasporophyll (e.g., Endress 2019; Kaplan 2022). But in an extensive review of the literature, Brückner (2000) distinguished ten different interpretations of the carpel or gynoecial unit based on different interpretations of the gynoecium, its appendages, and its evolutionary and phylogenetic origin. These divergent interpretations have led to countless arguments and controversies. Lorch (1963), in a historical overview of the carpel concept, concluded that "even on its own, classical morphology upon advancing toward smaller and less obvious targets would have found itself confronted with the breakdown of the concept of the carpel.” I think, however, that plant evo-devo has the potential to provide a framework based on empirical studies of the development of gynoecia. From this perspective, as I shall outline below,
I propose a topographic approach to the gynoecium in which processes such as spatial shifting (heterotopy) of the placenta or a single ovule and the notion of the gynoecial appendage are considered more basic than the classical carpel concept, although the latter will retain limited usefulness.

Do all Gynoecia have Carpels?

According to mainstream thinking (classical morphology), one of the defining characteristics of angiosperms (flowering plants) is the possession of classical carpels (closed megasporophyhlls). Hence, all gynoecia in angiosperms are considered carpellate. However, based on developmental studies, this claim has been contradicted or has at least become questionable if we accept the classical definition of the carpel according to which a carpel bears and encloses ovule(s). Of course, many taxa of angiosperms have carpels according to this definition, but for a considerable number of other taxa, it is difficult or impossible to apply the carpel concept (see, for example, Sattler and Lacroix 1988). Developmental investigations indicate that in these cases the centre of the gynoecium is transformed into a single ovule or a placenta that bears the ovules (see, for example, Macdonald and Sattler 1973, Sattler and Lacroix 1988), or an ovule is formed in the axil of gynoecial primordia (Pauzé and Sattler 1979, Zhang et al. 2019). Nonetheless, classical morphologists such as Endress (2019) and Kaplan (2022) insist that all gynoecia are carpellate. Endress (2019) claimed that in gynoecia with a terminal ovule or placenta (where the ovule or placenta arises in the centre of the gynoecium) the "young carpels … are 'rooted' within the remaining floral apex." He does not explain what exactly this means and he does not provide evidence for such rootedness, but it implies that the ovule or placenta does not arise from the floral apex but from carpellary tissue. Although a carpel primordium, like any other primordium or growth centre, cannot be sharply delimited at its basis, extending it over the floral apex, which is another growth centre, lacks an empirical developmental basis. It may imply a congenital fusion of the extension with underlying tissue and the floral apex (see Sattler 1974). But congenital fusion, by definition, is in principle unobservable. Hence, the admission of this concept removes the empirical basis of morphology. Furthermore, if a central terminal ovule is claimed to belong to a carpel, the question would be to which one when more than one are present. For example, in
Myrica gale, the gynoecium has two gynoecial appendages and one terminal ovule (Macdonald and Sattler 1973). If these two gynoecial appendages are interpreted as carpels, then the ovule should belong to one of them, shouldn’t it? But to which one? No known observation, including vascularization, indicates that the ovule is more closely connected to one of the two carpels. Thus, it is left hanging in between.
From the perspective of plant evo-devo, the challenge is to discover the observable developmental processes that are subsumed and often hidden by the concept of congenital fusion, which often does not imply any kind of fusion. Sattler (1978) distinguished the following processes: zonal growth, heterotopy, meristem extension, and interprimordial growth (meristem fusion). Sokoloff et al. (2018, p. 18) equate congenital fusion with zonal growth, which, however, is only one of the processes subsumed under the concept of congenital fusion. The de novo formation of such zonal growth during evolution does not imply any kind of fusion. Heterotopy also does not involve any kind of fusion, neither does meristem extension. However, interprimodial growth due to meristem fusion is a process that can be observed during ontogeny. Thus, among all the processes subsumed under the concept of congenital fusion, interprimordial growth seems to be the only one that constitutes a process of fusion. However, further clarification of developmental processes subsumed under the term congenital fusion appears desirable. In any case, for plant evo-devo the basis must be the elucidation of developmental processes that change during evolution and that are at least in principle observable. If this condition is not fulfilled, plant evo-devo and morphology lose their empirical basis.
One might avoid the problem of congenital fusion by assuming that as a result of the formation of gynoecial primordia the floral apex is used up and thus disappears so that the base of the gyoecial primordia extends into the centre of the gynoecium. However, in taxa such as
Basella rubra (Sattler and Lacroix 1988), the centre of the gynoecium retains the organization of a typical floral apex, which then is gradually transformed into an ovule. Hence, the gynoecium has been considered acarpellate. However, this conclusion can be avoided through a definition of the carpel.

Redefining the Carpel

Instead of concluding that there are carpellate and acarpellate gynoecia, an alternative would be a redefinition of the carpel concept in such a way that most of the acarpelllate or questionable gynoecia would become carpellate. According to one redefinition, the carpel is a gynoecial appendage that encloses ovule(s) but does not necessarily bear them (Sattler and Lacroix 1988; Greyson 1994; Leins and Erbar 2010, in their glossary only). With regard to Caryophyllales, Ronse De Craene (2021) pointed out a "progressive detachment of ovules from the carpellary tissue."
It seems what matters most is that the placenta or ovule(s) are enclosed and whether they are borne on the gynoecial appendage or the floral apex is not of great importance for the survival of the species. Therefore, instead of claiming that all Angiosperms have carpels that bear ovules, I conclude that most angiosperms have gynoecial appendages that alone or in conjunction with underlying tissues, enclose ovules. The enclosure of ovules then becomes the defining characteristic for most angiosperms.
Although this redefinition of the carpel concept is broader than the classical carpel concept and supersedes problems and pseudo-problems of the latter, it still has a weakness, because in any particular case someone might ask whether the ovule(s) are borne on the carpel or not, which means that it may imply an either/or logic: Are the ovule(s) borne on the carpel or on the floral apex. But how do we answer this question? We would have to delimit the carpel primordium from the floral apex. It is, however, difficult or even impossible to delimit growth centres such as the carpel primordium and the floral apex. And therefore it may be difficult or impossible to decide whether ovule(s) are borne on the carpel or the floral apex. To overcome this problem, which may be a pseudo-problem,
we proposed to redefine a carpel simply "as an appendage which ENCLOSES ovule(s)" (Sattler and Perlin 1982, p. 181). This redefinition does not require difficult or impossible delimitations and therefore I consider it the most appropriate redefinition. When I refer to redefinition, this one is meant.
Although this redefinition of the carpel broadens the scope of the term, it may lead to confusion unless we always specify whether the term is used in the traditional sense as a closed megasporophyll or as redefined, whereas using the term gynoecial appendage and gynoecial primordium does not require further specification and therefore appears preferable. In my book Organogenesis of Flowers (Sattler 1973), I consistently used the terms gynoecial appendage and gynoecial primordium. Using these terms does not imply whether an ovule or placenta is borne on them.

A Topographic Approach to the Gynoecium

Like the redefinition of the carpel as an appendage that encloses ovule(s), the topographic approach does not require difficult or impossible delimitation of the gynoecial primordia. In addition, the topographic approach provides a topography of placentae and ovule(s) that specifies the position of the placentae and ovule(s). We can distinguish the following positions: basal (=central or terminal), axillary, and appendicular; and for the latter, submarginal, ventral, dorsal, and laminar; and, for syncarpous gynoecia, free central, axile, parietal, and superficial. Intermediate positions can also be recognized such as, for example, a near basal ovular position in taxa such as Hordeum vulgare (Sattler 1973). As a result, a continuum of positions can be envisaged.
This topographic terminology has been used by taxonomists for a long time and seems to work well. Since it does not rely on a delimitation of growth centres such as the floral apex and gynoecial primordia, it transcends the classical carpel concept. It can be seen as a liberation from the limitations of the classical carpel concept that has cast a long shadow on gynoecial morphology since the publication of Goethe's
Metamorphosis of Plants (1890) (that involved the carpel concept but not the term) and the introduction of the term by F. Dunal in 1817 (see Lorch 1963, p. 271). However, for many morphologists this approach may entail an evasion of interpretation. They would consider the proposed topographic approach only descriptive. But some philosophers, including Nietzsche, have pointed out that all descriptions are interpretations because all descriptions involve concepts that interpret them. Thus, the proposed topographic approach can be seen as a topographic interpretation of the gynoecium, an interpretation that avoids the phyllome versus caulome categorizations. Furthermore, it avoids the pseudo-problem to find a boundary, such as the boundary of the gynoecial priordium, that does not exist in nature.
One great advantage of the topographic approach is that it allows adherents and defenders of opposite views to meet: those who defend a phyllomic (carpellate) interpretation of the gynoecium can meet with those who hold a cauline view because they can agree that in free central and basal placentation the placenta or ovule arises in the centre of the gynoecium. Thus, one can let go of the commitment to morphological categories of phyllome and caulome, leaf and stem. One can see that understanding processes such as spatial and temporal shifting is more basic than structural categories. One can see the value of process morphology that is beyond categories (Sattler 1992).
The topographic approach can redirect research toward more productive avenues since it can liberate us from common futile debates about boundaries that do not exist in nature such as the boundaries of gynoecial primordia. Thus, questionable distinctions between carpellate and acarpellate gynoecia can be let go of. When the placentation is clearly appendicular, the classical carpel concept is still applicable and useful. And the redefinition of the carpel concept is more comprehensive than the classical concept, if it does not imply a delimitation of gynoecial primordia and therefore avoids the distinction of appendicular and cauline placentation, phyllospory and stachyospory (Lam 1950).
(As an autobiographical note, I might mention that my morphological thinking about the gynoecium evolved from an acceptance of the classical carpel concept in my doctoral thesis (Sattler 1962), to a distinction of carpellate and acarpellate gynoecia (Sattler 1974), to a redefinition of the carpel (Sattler and Perlin 1982), and finally, in this article, to a topographic approach. I think that the topographic approach is the most comprehensive one. Nonetheless, the classical carpel concept retains limited usefulness, especially when applied within its range of applicability. The redefinition of the carpel is compatible with the topographic approach if it does not require a delimitation of gynoecial primordia).
The topographical approach and its terminology is useful in plant evo-devo in as much as it is based on observable developmental events and their change during evolution. If concepts such as that of the carpel are used, they must be based on developmental events that are at least in principle observable and do not rely on questionable or impossible delimitations of growth centres such as the floral apex and gynoecial primordia.

Plant Evo-Devo, especially Morpho Evo-Devo of the Gynoecium

In plant morpho evo-devo it is important that we study first the morphological development of gynoecia, which must provide an empirical basis. Then we can ask how different gynoecia may have been transformed into one another. And we find that it is through a number of processes. One of them is spatial shifting, which implies heterotopy, especially heterotopy of the placenta or ovule(s). As a result of heterotopy, the placenta or ovule(s) may arise on the gynoecial appendages, which leads to classical carpels (closed megasporophylls), or in the axil of the gynoecial appendages, or in the centre of the gynoecium. From a topographic perspective, we would simply say that the placenta or ovule(s) may be appendicular, axillary, or basal, and in syncarpous gynynoecia, free central, axile, parietal, or superficial. Furthermore, a continuum of these positions may be envisaged. A single axillary ovule occurs in Illicium lanceolatum (Zhang et al. 2019), Illicium henryi (Wang et al.) and Ochna atropurpurea (Pauzé and Sattler 1979). An ovule-bearing branch is formed in the axil of gynoecial appendages of atypical gynoecia of Michelia figo (Zhang et al. 2017). In Myrica gale (Macdonald and Sattler 1973), Basella rubra (Sattler and Lacroix 1988), and other taxa the centre of the gynoecium is transformed into an ovule, which means the ovule is basal. Sattler and Lacroix (1988, p. 926) listed many taxa in which the placenta is free central or the ovule is basal (for a more complete list contact Prof. Christian Lacroix, Biology Dept., University of Prince Edward Island, Charlottetown, P. E. I.).

Gynoecia of Fossil Angiosperms and Pre-Angiosperms

The recognition of various positions of the placenta and ovule(s) is not only important for an understanding of gynoecial morphology in extant angiosperms but is also highly relevant for the search for fossil angiosperms and pre-angiosperms. As long as we are blinded by the assumption that all angiosperms have classical carpels (closed megasporophylls), we will look for fossils whose gynoecia resemble classical carpels at least in some ways, maybe megasprophylls that are not yet completely closed. However, bearing in mind the diversity of placentation and ovular position, we are not locked into this view; we can also envisage other possibilities. Thus
Combina gen. nov., discovered in the middle Triassic by Santos and Wang (2022), can be considered a precursor of angiosperms. Combina has an axillary ovule that is almost fully enveloped by a bract. This condition resembles the one reported in taxa such as Illicium as noted above. If this condition is ancestral to the angiospermous gynoecium, then classical carpels (closed megasporophylls) would have evolved through shifting (heterotopy) of the ovule unto the margin of the gynoecial appendage and an increase in its number as it can be seen, for example, in a pea pod. Having many ovules instead of only one axillary one could be seen as advantageous.
Santos and Wang (2022, Fig. 4) proposed an evolutionary trend leading from the fossil Drepanolepis, whose ovule is in the axil of a bract that does not enclose it, to Combina where the bract almost completely encloses the ovule, and then to taxa like Illicium where the ovule is fully enclosed. If this hypothesis can be confirmed, the question remains whether this was the only way how the gynoecium of angiosperms originated or whether additional ways existed as proposed by some authors (see, for example, Croizat 1960, 1962, 1964; Heads 1984). In other words, are angiosperms monophyletic or polyphyletic? In any case, appendicular placentation can no longer be used as a criterion for the monophyletic origin of angiosperms since not all angiosperms have such placentation. However, according to the redefinition of the carpel and the topographical approach, most angiosperms have their placenta or ovule(s) enclosed. This enclosure might have evolved in different ways. Archaefructus, a fossil angiosperm from the early Cretaceous (Sun et al. 2002) has carpels (closed megasporophylls), but they are unusual because the ovules are inserted along the midrib of the carpel, that is, they have an adaxial position (Wang 2018). Their number varies from one to twelve (ibid.). If there is only one ovule, where exactly is it positioned? If it is positioned at or near the axil of the carpel, it would be similar to the position of the axillary ovule of Combina and Illicium. Then, starting with the axillary ovule of Combina, the carpel of Archaefructus could have evolved through an amplification of the number of ovules along the midrib of the carpel. Subsequently, spatial shifting of ovule formation from the midrib towards the margins could have produced the conduplicate angiospermous carpel such as the pea pod. Furthermore, spatial shifting of axillary ovule formation into a ventral median position of an ascidiate carpel could have led to gynoecia of primitive angiosperms such as that of Amborella. Further shifting into the centre of the gynoecium could have produced the basal ovule in gynoecia such as that of Myrica. However, the basal ovule in Myrica and other taxa could also be considered primitive. Nubilora, a Triassic fossil, whose ovules are directly borne on the floral axis, may support this conjecture (Wang 2018, Santos and Wang 2022). The Middle Jurassic Qingganninginfructus formosa had a single basal bitegmic ovule (Han et al. 2023). Hence in the Middle Jurassic we find already basal ovules.
Although the origin and evolution of angiosperms and the gynoecium remains hypothetical, we can conclude that heterotopy of ovules occurred in both extant and fossil angiosperms. And this heterotopy includes appendicular, axillary, and basal (central) positions,
the latter two which require a topographic approach or at least a redefinition of the carpel, if one wants to retain this term; but, as I suggested, the term gynoecial appendage would be preferable.

Processes in Plant Evo-Devo (Plant Morpho Evo-Devo) of the Gynoecium

Since plant evo-devo (plant morpho evo-devo) investigates developmental changes in evolution, it leads to the processes underlying these developmental changes. As pointed out above, one important process is spatial shifting (heterotopy). Within extant Angiosperms, positional shifts may have occurred in various directions (Sattler and Lacroix 1988). Once the placenta or ovule(s) have been enclosed, such shifts would not endanger the survival of such species.
Besides spatial shifting (heterotopy), another process is temporal shifting (heterochrony). An extreme example of heterochrony occurs in
Balanophora elongata: the embryo sac develops already within an elongate floral apex (Zimmermann 1959, p. 531). Hence, no gynoecial appendages and ovules are formed.
In addition to heterotopy and heterochrony, Zimmermann (1959) referred to heteromorphy. The following processes (leading to heteromorphy) played an important role in plant morpho evo-devo of the gynoecium (for details see Sattler 1974, Greyson 1994, Leins and Erbar 2010, Ronse De Craaene 2018, 2021):

  1. 1.Differentiation, such as the differentiation of the gynoecial appendages into ovary, style, and stigma.
  2. 2. Varying proportions, such as the varying proportions of ovary, style, stigma, and other components (e.g., Ronse De Craene 2021).
  3. 3. Zonal growth and interprimordial growth (see Sattler 1978).
  4. 4. Postgenital fusions between various organs. In contrast to congenital fusion, which is in principle unobservable, postgenital fusion is observable.
  5. 5. Reduction or amplification in size and number, as, for example, the size of organs and the number of ovules.
  6. 6. Transference of function (Corner 1958). For example, in Stylidium adnatum the function of the style has been transferred to the androecial tube (Sattler 1973, 1974).
Takhtajan (1972), like Zimmermann (1959), also a forerunner of plant morpho evo-devo, distinguished the processes of deviation, prolongation and abbreviation. Heterotopy and the above processes (# 1-6) would be examples of deviation, whereas prolongation and abbreviation would be examples of heterochrony.
Considering all relevant processes provides a more complete picture of plant evo-devo (plant morpho evo-devo) of the gynoecium.

Plant Evo-Devo (Plant Morpho Evo-Devo) and the Concept of the Flower

Plant Evo-Devo (Plant Morpho Evo-Devo) of the gynoecium is also relevant to the concept of the flower. According to the predominant classical morphology, "a flower is a reproductive short shoot bearing microsporophylls (stamens)… and megasporophylls (carpels) as its appendages or leaf homologues" (Kaplan 2022, p. 1069, edited by Peter Endress). In view of the available developmental data, this definition is no longer generally valid. The appendages of the androecium range from leaf-like to stem-like and even short shoot-like structures (Rutishauser and Sattler 1985, Sattler 1988). And it is questionable whether all gynoecia consist of closed megasporophylls. How then can we define the flower?
According to Claßen-Bockhoff ( 2016), the flower is the "sporangia bearing tip of the shoot," in which "stamens and carpels are sporangiophores and as such 'de novo' structures not necessarily homologous with vegetative leaves." Whereas one can see the androecium as consisting of sporangiophores that may be more or less leaf-like, stem-like, short shoot-like, or don't fit any of the classical categories, the gynoecium is more complex consisting of more or less leaf-like appendages (gynoecial appendages) that enclose the sporangiophores, which in the simplest case consist only of one ovule and in more elaborate cases of a placenta with ovules. As pointed out above, the position of the placentae and ovules is variable. In gynoecia with classical carpels the sporangiophores are formed on the carpels, whereas in other gynoecia, for example, in Myrica, they are basal, or as, for example, in Illicium, they are axillary. Alternatively, one could consider the gynoecial sporangiophore as a dual structure consisting of a more less foliaceous appendage and a sporangia-bearing structure.
According to process morphology, stamens and carpels, gynoecial appendages, placentae and ovules are seen as process combinations that need not be fitted into the classical categories of stem and leaf (Sattler 1988, 1992, Sattler and Rutishauser 2023).


In Plant Evo-Devo (Plant Morpho Evo-Devo), the emphasis should be on observable developmental changes and the processes that may lead to these changes during development and evolution. Spatial shifting (leading to heterotopy) is a basic process that led to gynoecia with ovules formed in different positions: on the gynoecial appendage or in its axil or in the centre of the gynoecium. In the latter case, ovules are not “as if glued on the surface of the gynoecium or the floral apex,” as Endress (2019) misrepresented our investigations; but the floral apex, retaining at first its typical organization, is gradually transformed into a single ovule or a placenta (see, for example, Sattler and Lacroix 1988, Figs. 12-15). From the topographical perspective, one would not even refer to the floral apex, but simply to a basal ovule or a free central placenta. This view has an empirical basis; it does not require difficult to impossible delimitations of gynoecial primordia.
In addition to spatial shifting (heterotopy), temporal shifting (heterochrony), differentiation, varying proportions, interprimordial and zonal growth, reduction, amplification, and transference of function are processes that played an important role in plant evo-devo (plant morpho evo-devo) of the gynoecium.
As a result of a plant evo-devo (plant morpho evo-devo) perspective based on developmental studies of gynoecia, angiosperms cannot be defined by the possession of classical carpels (closed megasporophylls), although many taxa have such carpels. As the name angiosperms indicates, they are defined by having the ovules enclosed, enclosed by gynoecial appendages and in many cases by additional underlying tissue formed by zonal growth, as, for example, in inferior ovaries. Within the enclosure and protected by it, placentae and ovules could change their position without any detrimental effect on the survival of the plants.
In this article on plant evo-devo of the gynoecium, the focus has been on plant morpho evo-devo. This focus will have to be enlarged to include data from developmental molecular genetics. Based on developmental genetics, Mathews and Kramer (2012) concluded "that the carpel is a complex organ consisting of a foliaceous appendage and the placenta." When this is recognized that the carpel consists of two components, the foliaceous appendage (which I called the gynoecial appendage) and the placenta or ovule(s), then we can also recognize that the relative position of the two components may change and that this change may eclipse the classical carpel when the placenta or ovule(s) is free central, basal, or axillary.
Recognizing that in extant angiosperms ovules are not always appendicular is also relevant for the search of fossil angiosperms and pre-angiosperms. It opens up different plant evo-devo vistas for the evolution of the gynoecium. Thus, the Triassic
Combina can be considered a possible ancestor of angiosperms. It has bracts that bear an ovule in their axil and enclose it almost completely. Spatial shifting of axillary ovule formation into a ventral median position of an ascidiate carpel could have led to gynoecia of primitive angiosperms such as that of Amborella. If we assume a monophyletic origin of angiosperms, further shifting of ovule formation into the centre of the gynoecium could have produced gynoecia such as that of Myrica in which the ovule is basal. However, there is evidence of basal ovules already in the Triassic and Jurassic. Spatial shifting of ovule formation into the adaxial midrib region of the enclosing appendage and an increase in the number of ovules could have produced the carpels of the early Cretaceous fossil Archaefructus. And spatial shifting of the adaxial ovules toward the margins of the gynoecial appendage could have led to the conduplicate carpel (folded megasporophyll) of extant angiosperms such as that of Magnoliaceae.


Brückner, C. 2000. Clarification of carpel number in Papaverales, Capparales, and Berberidaceae. Botanical Review 66 (2): 155-307.

Claßen-Bockhoff, R. 2016. The shoot concept of the flower: Still up to date? Flora 221: 46-53.

Corner, E. J. H. 1958. Transference of function. Journal of the Linnean Society, Botany 56: 33–40.

Croizat, L. 1960.
Principia Botanica. Caracas: Leon Croizat.

Croizat, L. 1962.
Space, Time, Form: The Biological Synthesis. Caracas: Leon Croizat.

Croizat, L. 1964. Thoughts on high systematics, phylogeny and floral morphogeny, with a note on the origin of the Angiospermae. Candollea 19:17-96.

Endress, P. K. 2019. The morphological relationship between carpels and ovules in angiosperms: pitfalls of morphological interpretation. Botanical Journal of the Linnean Society 189: 201-227.

Goethe, J. W. von. 1790. Versuch die Metamorphose der Pflanzen zu erklären. Gotha: C. W. Ettinger (translated by Arber, A. 1946. Goethe's Botany. Chronica Botanica 10: 63-126.

Greyson, R.I. 1994. The Development of Flowers.
New York/Oxford: Oxford University Press.

Han, L. et al. 2023. New fossil evidence suggests that Angiosperms flourished in the Middle Jurassic. Life 13(3), 819;

Heads, M. 1984. Principia Botanica: Croizat's contribution to botany. Tuatara 27: 26-45.

Kaplan, D. R. (edited by C. D. Specht) 2022.
Kaplan's Principles of Plant Morphology. CRC Press.

Lam, H. J. 1950. Stachyospory and phyllospory as factors in the natural system of the cormophyta. Svensk Botanisk Tidskrift 44: 517-534.

Leins, P. and Erbar, C. 2010.
Flower and Fruit. Stuttgart: Schweizerbart Science Publishers.

Lorch, J. 1963. The carpel - a case-history of an idea and a term. Centaurus 8: 269-291.

Macdonald, A. D. and Sattler, R. 1973. Floral development of
Myrica gale and the controversy over floral concepts. Canadian Journal of Botany 51:1965-1976.

Mathews, S. and Kramer, E. 2012. The evolution of reproductive structures in seed plants: A re-examination based on insights from developmental genetics. New Phytologist 194: 910-923.

Pauzé, F. et Sattler, R. 1979. La placentation axillaire chez
Ochna atropurpurea DC. Canadian Journal of Botany 57: 100-107.

Petrone-Mendoza, E., Vergara-Silva, F., Olson, M. E. 2023. Plant morpho evo-devo. Trends in Plant Science July 07, 2023.

Ronse De Craene, L. P. 2018. Understanding the role of floral development in the evolution of angiosperm flowers: clarifications from a historical and physico-dynamic perspective. Journal of Plant Research 131: 367-393.
DOI 10.1007/s10265-018-1021-1

Ronse De Craene, L P. 2021. Gynoecium structure and development in core Caryophyllales: a matter of proportions. Botanical Journal of the Linnean Society 195: 437-466.

Rutishauser, R. and Sattler, R. 1985. Complementarity and heuristic value of contrasting models in structural botany. I. General considerations. Botanische Jahrbücher für Systematik 107: 415-455.

Santos, A. A. and Wang, X. 2022. Pre-carpels from the Middle Triassic of Spain. Plants 11 (21): 2833.

Sattler, R. 1962. Zur frühen Infloreszenz- und Blütenentwicklung der Primulales sensu lato mit besonderer Berücksichtigung der Stamen-Petalum-Entwicklung. Botanische Jahrbücher 81: 358-396.

Sattler, R. 1973.
Organogenesis of Flowers. A Photographic Text-Atlas. University of Toronto Press

Sattler, R. 1974. A new approach to gynoecial morphology. Phytomorphology 24: 22-34.

Sattler, R. 1978. “Fusion” and “continuity” in floral morphology. Notes of the Royal Botanic Garden, Edinburgh 36: 397-405.

Sattler, R. 1988. A dynamic multidimensional approach to floral development. In Leins, P., Tucker, S. C., and Endress, P.
Aspects of Floral Development. Berlin/Stuttgart: J. Cramer, pp. 1-6.

Sattler, R. 1992. Process morphology: structural dynamics in development and evolution. Canadian Journal of Botany 70: 708-714.

Sattler, R. and Perlin, L. 1982. Floral development of
Bougainvillea spectabilis Willd., Boerhaavia diffusa L. and Mirabilis jalapa L. (Nyctaginaceae). Botanical Journal of the Linnean Society 84: 161-182.

Sattler, R. and Lacroix, C. 1988. Development and evolution of basal cauline placentation in
Basella rubra. American Journal of Botany 75: 918-927.

Sattler, R. and Rutishauser, R. 2023. Fundamentals of plant morphology and plant evo-devo (evolutionary developmental biology). Plants 12 (1), 118;

Sokoloff, D. D. et al. 2018. Types of organ fusion in angiosperm flowers (with examples from Chloranthaceae, Araliaceae and monocots). Biologia Serbica 40 (1): 16-46.

Sun, G. et al. 2002. Archaefructaceae, a new basal angiosperm family. Science 296: 899-904.

Takhtajan, A. 1972. Patterns of ontogenetic alterations of higher plants. Phytomorphology 22: 164-170.

Wang, X. 2018.
The Dawn of Angiosperms. Uncovering the Origin of Flowering Plants. 2nd edition. Springer.

Wang, Z., Duan, X., Li, J. and Zhang, X. The carpel. Morphological development of
Illicium henryi and its vascular bundles structure. In prep.

Wanninger, A. 2015. Morphology is dead - long live morphology! Integrating MorphoEvoDevo into molecular EvoDevo and phylogenomics. Frontiers in Ecology and Evolution 3, Article 54.

Zhang, X, Liu, W., Wang, X. 2017. How the ovules get enclosed magnoliaceous carpels. PLoS One 12(4): e0174955.

Zhang, X. Zhang, Z., Zhao, Z. 2019. Floral ontogeny of Illicium lanceolatum (Schisandraceae) and its implication of carpel homology. Phytotaxa 416: 200-210.

Zimmermann, W. 1959. Die Phylogenie der Pflanzen. 2nd edition. Jena: Gustav Fischer Verlag.


I am grateful to Regine
Claßen-Bockhoff, Louis Ronse De Craene and Rolf Rutishauser for their valuable comments and suggestions.

Latest update on September 20, 2023.

See also
Morphological development (organogenesis) of flowers

Plant evo-devo and morphological research by Rolf Sattler and collaborators


RapidWeaver Icon

Made in RapidWeaver