Novelties in the embryology of Parodia (Cactaceae) and its potential application to the genus taxonomy

Abstract

Background and Aims

The problem related to the circumscription of Parodia originates in the classification of Acanthocephala, Eriocephala and Wigginsia as subgenera of Notocactus sensu lato, and continues after the merger of Notocactus sensu lato with Parodia sensu stricto, which originated Parodia sensu lato. Embryological data were not considered in the decisions leading to the circumscription of Parodia sensu lato, but recent studies have demonstrated its potential applicability in discussions involving the genus. Therefore, this study aimed to describe embryological data related to ovule development and the processes of megasporogenesis and megagametogenesis, and to apply them to existing circumscriptions.

Methods

Floral buds at different stages of development and pre-anthesis flowers from six Parodia species were analysed. For sample processing, standard techniques used in anatomical studies were employed. The material was analysed and photographed under polarized light, epifluorescence microscopy and light microscopy.

Key Results

The analyses revealed the absence of a developmental pattern during ovule formation in Parodia, with species exhibiting a triad or tetrad of megaspores, apomixis, mitotic division of archespores, and different mechanisms of cell selection during ovule development. The species also exhibited an atypical pattern of callose deposition during megasporogenesis, which could be considered exclusive to Cactaceae. Furthermore, one critically endangered species showed ovule malformation.

Conclusions

The observation of exclusive features in these species allowed the application of the data to one of the less inclusive circumscriptions, in which Acanthocephala, Brasiliparodia, Eriocephala, Notocactus, Parodia and Wigginsia are considered distinct but related genera. In addition, the embryological data also demonstrated potential use as a tool in understanding the reasons that may cause the population decline of critically endangered species.

INTRODUCTION

Parodia is a genus of cacti native to South America belonging to the tribe Notocacteae, subfamily Cactoideae (Cactaceae). Within this genus, there are 61 species found in the flat regions of the pampa biome (southern Brazil, Uruguay, north-eastern Argentina, and eastern Paraguay) and along the eastern slopes of the Andes (north-western Argentina and eastern Bolivia) (Machado et al., 2008; IUCN, 2023). The morphological diversity of the genus is evident in its vegetative characteristics, with globular to short cylindrical stems that have ribbed or non-ribbed tubercles bearing areoles. These areoles contain spines of various colours, shapes and textures, which vary among species. The floral morphology also exhibits a wide range of diversity, with flowers that can range from green to red in colour, varying in shape from funnel-shaped to bell-shaped, and with areoles that may bear bristles, spines and/or woolly trichomes (Buxbaum, 1967; Carneiro et al., 2016). A concerning fact for this genus is that more than half of its species are classified under one of the higher threat categories of the International Union for Conservation of Nature (IUCN) Red List of Threatened Species, where 15% of Parodia species are listed as critically endangered (CR), 23% as endangered (EN) and 21% as vulnerable (VU).

The issue of nominal problems has been present in the history of Parodia since its creation. In fact, the name of the genus emerged as a replacement for the illegitimate name Hickenia (Spegazzini, 1923). This name replacement not only corrected the nominal error, but also led to the creation of a list of synonyms which, over time, has grown significantly (Anceschi and Magli, 2018). According to Anceschi and Magli (2018), Parodia includes 18 genera under its synonymy, of which 10 are considered illegitimate or invalid names. The remaining synonyms consist of legitimate genera that, for various reasons, ended up being synonymized under Parodia. Among these, some have been more questioned than others, as in the case of Acanthocephala, Brasiliparodia, Eriocephala, Notocactus and Wigginsia.

The debate concerning these genera originated in 1967 with Franz Buxbaum, an Austrian researcher who classified Acanthocephala, Eriocephala and Wigginsia as subgenera of Notocactus, creating Notocactus sensu lato (s.l.). Buxbaum believed that the differences between the three genera and the type species, which he incorrectly considered to be Echinocactus scopa, were merely gradual and less significant than those observed in the subgenus Neonotocactus, which, despite considerable differences, was not separated from Notocactus (Buxbaum, 1967). Over a decade later, German scientist Friedrich Ritter reignited the discussion by meticulously identifying distinct characteristics in the spines, flowers, fruits and seeds. According to Ritter, these differences were significant enough to re-establish the genera Acanthocephala, Eriocephala, Notocactus and Wigginsia, and even create a new one called Brasiliparodia (Ritter, 1979). However, the characteristics identified by Ritter were not enough to keep the genera separate. Less than a decade after its publication, the International Organization for Succulent Plant Study (IOS) decided to merge Notocactus s.l. and Parodia. This fusion gave rise to Parodia s.l. and was justified by the lack of practical differences that could distinguish the two genera (Hunt and Taylor, 1986).

Although the circumscription of Parodia continues to be a topic of discussion today, few studies have sought evidence that could strengthen or weaken the union of Acanthocephala, Brasiliparodia, Eriocephala, Notocactus and Wigginsia in Parodia s.l. (Gerloff and Neduchal, 2004; Van Vliet, 2009, 2014; Gentz et al., 2023). Among these studies and throughout the history of the genus, only Gentz et al. (2023) have utilized embryological data in an attempt to contribute to resolving the problematic circumscription of Parodia, revealing the potential taxonomic use of such data through the observation of characteristics that allowed the distinction between Parodia sensu stricto (s.s.) and Notocactus s.l. While this marks a significant advancement in embryological studies for the genus, the available data are limited to the anatomy of the ovule in the final stage of development. Furthermore, the stages leading up to the formation of this structure remain largely unknown, indicating a relatively unexplored and fragmented field of study within the family Cactaceae. Therefore, given the dearth of embryological research on Parodia and the potential usefulness of such data in resolving taxonomic issues at various hierarchical levels (Cronquist, 1988; Tobe, 1989; Stuessy, 1990), our goal was to describe the development of the ovule and the processes of megasporogenesis and megagametogenesis in this genus. Through this, we aim to expand our understanding of the embryology of Parodia and, in turn, enrich the embryology of Cactaceae. Additionally, we hope to identify new embryological features that can be applied to existing discussions regarding its circumscription.

MATERIALS AND METHODS

Plant materials

Six species were selected for the study, each representing one of the former genera currently synonymized with Parodia. These include: Parodia erinacea (Wigginsia), Parodia haselbergii (Acanthocephala), Parodia leninghausii (Eriocephala), Parodia microsperma (Parodia s.s.), Parodia ottonis (Notocactus s.s.) and Parodia rechensis (Brasiliparodia). The specimens of the collected species were deposited in the ICN Herbarium of the Federal University of Rio Grande do Sul (UFRGS) under the numbers 202995, 202993, 202984, 202996, 202989 and 202982, corresponding to the same order as the species listed above.

To conduct the proposed analyses, dozens of floral buds at various stages of development were collected, with a minimum of three pre-anthesis flowers for each species. At least three different individuals were sampled for each species. For each stage of the megasporogenesis and megagametogenesis processes, an average of 10–30 ovules per ovary were analysed. The plant materials were collected from the collection of the Porto Alegre Botanical Garden in Porto Alegre, Rio Grande do Sul (RS), Brazil; Cactário Horst in Imigrante (RS); the private property of Filipe Bernardi in Caxias do Sul (RS); and the private collection of Jones Caldas da Silva in Viamão (RS).

Anatomical analyses under light and epifluorescence microscopy

Upon collection, the material was fixed in a solution of 1 % glutaraldehyde and 4 % formaldehyde in 0.1 m phosphate buffer pH 7.2 (McDowell and Trump, 1976). The fixed material was then washed in 0.1 m phosphate buffer pH 7.2, dehydrated gradually and sequentially in ethanol, and embedded in plastic resin (hydroxyethyl methacrylate) (Johansen, 1940; Gabriel, 1982; Gerrits and Smid, 1983). After embedding, 4-μm sections were obtained using a microtome (RM2265, Leica, Nussloch, Germany) and stained with 0.1 % Toluidine Blue O in 0.1 m sodium phosphate buffer pH 4.4 (Feder and O’Brien, 1968). To visualize callose during megasporogenesis, sections were stained with Aniline Blue (Smith and McCully, 1978). Starch grains in mature gametophytes were identified using polarized light.

Photomicrographs in bright-field and epifluorescence were captured using a light microscope (DMR HC, Leica, Wetzlar, Germany) equipped with a digital imaging system (AxioCam, Zeiss, Germany) using the free image capture software Carl Zeiss Zen Lite 2012. A 340–380 nm excitation filter was used for capturing epifluorescence photomicrographs. Figures were created using Adobe Photoshop CC 2017 software.

Terminologies

The terms archesporium and archespore were used in accordance with the definitions of Pozner (2001). The ovule type classification was based on the definitions of Bocquet and Bersier (1960).

RESULTS

A summary of the distinct traits observed in the studied species is provided in Table 1. Specific details regarding the ovule development and megasporogenesis of the six studied species can be observed in Figs 1–6. Figure 7 illustrates the process of megagametogenesis, using P. erinacea as an example. Additionally, Fig. 8 displays the morphology of the ovule and mature female gametophyte for the other analysed species.

Table 1.

Distinct traits observed among the species of Parodia.

Trait	P. erinacea	P. haselbergii	P. leninghausii	P. microsperma	P. ottonis	P. rechensis
Archespore number and PCD extension	1 cell	1 cell	2 cells	1 or 2 cells	2 cells	1 cell
Mitotic division of archespores	Absent	Absent	Absent	Present	Absent	Absent
Meiosis extension	1 cell	1 cell	1 or 2 cells	1 or 2 cells	1 or 2 cells	1 cell
Meiosis gradient	Absent	Absent	Absent	Present*	Present*	Absent
Apomixis	Present	Absent	Absent	Absent	Absent	Absent
Number of megaspores	1	3	3	3	3	4
Cell selection mechanism	Pre-meiotic	Pre-meiotic	Pre-meiotic and post-meiotic	Pre-meiotic and meiotic	Pre-meiotic and post-meiotic	Pre-meiotic
Antipodals	Persistent**	Persistent**	Ephemeral	Persistent**	Persistent**	Ephemeral
Nucellar projection	Absent	Absent	Absent	Present	Absent	Absent

PCD, pre-meiotic cellular differentiation.

^*Present in ovules where the meiosis extends to two cells.

^**In the fully developed ovule.

Fig. 1.

Ovule development and apomeiosis process in P. erinacea. (A) Ovule primordium formation and subdermal initial cell differentiation. (B) Ovule in early developmental stage, with an archesporium composed of a single archespore. (C) Ovule with a megaspore mother cell (arrow: third layer in the outer integument). (D) Elongated megaspore mother cell with small vacuoles. Compare the structural development of the ovule with that of the ovule in C (arrow: third layer in the outer integument). (E–G) Details of changes in the megaspore mother cell. (E) Formation of small vacuoles, distinguishable as white dots at the cell poles. (F) Elongation of the megaspore mother cell and filling of small vacuoles with polysaccharides. (G) Fusion of small vacuoles resulting in the formation of two single vacuoles, one at each pole of the cell, filled with polysaccharides. ar, archesporium/archespore; cl, central layer; dl, dermal layer; fu, funiculus; ii, inner integument; mmc, megaspore mother cell; nu, nucellus; oi, outer integument; si, subdermal initial cell; sl, subdermal layer.

Fig. 2.

Ovule development and megasporogenesis in P. haselbergii. (A) Ovule primordium formation and subdermal initial cell differentiation (arrows: secondary ovule primordia). (B) Ovule in early developmental stage, with the archesporium composed of a single archespore. (C) Ovule in megaspore mother cell stage (arrow: third layer in the outer integument). (D) Ovule in megaspore dyad stage (arrow: third layer in the inner integument). The chalazal megaspore in anaphase indicates the transition to the megaspore triad stage. (E) Ovule in megaspore triad stage (arrow: third layer in the inner integument). (F–J) Details of the megasporogenesis process. (F) Megaspore mother cell. (G) Megaspore dyad transitioning to the triad stage. (H) Megaspore triad. (I) First degenerated megaspore (arrow). (J) Two degenerated megaspores (arrows) and the establishment of the chalazal megaspore as the functional megaspore, which is undergoing vacuolization. ar, archesporium/archespore; cl, central layer; dl, dermal layer; dy, megaspore dyad; fu, funiculus; ii, inner integument; mmc, megaspore mother cell; nu, nucellus; oi, outer integument; si, subdermal initial cell; sl, subdermal layer; tr, megaspore triad.

Fig. 3.

Ovule development and megasporogenesis in P. leninghausii. (A) Ovule primordium formation and subdermal initial cell differentiation (arrows: secondary ovule primordia). (B) Ovule in early developmental stage, with the archesporium composed of two archespores. (C) Ovule in megaspore mother cell stage (arrow: third layer in the outer integument). (D) Ovule in megaspore dyad stage. (E–H) Details of the megasporogenesis process. (E) Megaspore dyad. (F) Megaspore triad. (G) First degenerated megaspore (arrow). (H) Two degenerated megaspores (arrows) and establishment of the chalazal megaspore as the functional megaspore. (I) Ovule in vacuolated functional megaspore stage (* detail of functional megaspore). ar, archesporium/archespore; cl, central layer; dl, dermal layer; dy, megaspore dyad; fm, functional megaspore; fu, funiculus; ii, inner integument; mmc, megaspore mother cell; nu, nucellus; oi, outer integument; si, subdermal initial cell; sl, subdermal layer.

Fig. 4.

Ovule development and megasporogenesis in P. microsperma. (A) Ovule primordium formation and subdermal initial cell differentiation (arrows: secondary ovule primordia). (B) Ovule in early developmental stage, with the archesporium composed of two archespores, one of which is finishing a mitotic division and giving rise to two new cells (arrows). (C) Ovule with two megaspore mother cells. (D) Ovule in megaspore dyad stage, with the chalazal megaspore in anaphase indicating the transition to triad (arrows: megaspores of the dyad). Another ovule (*), developing below the mentioned ovule, displaces the micropylar region of the first one. (E, F) Projection of the dyad of megaspores beyond the limits of the micropyle, along with the nucellus. The chalazal megaspore in metaphase indicates the transition to triad. (G) Megaspore triad. (H) Ovule in functional megaspore stage, with structure in advanced development. (I) Detail of the vacuolating functional megaspore (arrows: degenerated megaspores). ar, archesporium/archespore; cl, central layer; dl, dermal layer; dy, megaspore dyad; fm, functional megaspore; fu, funiculus; ii, inner integument; mmc, megaspore mother cell; nu, nucellus; oi, outer integument; si, subdermal initial cell; sl, subdermal layer; tr, megaspore triad.

Fig. 5.

Ovule development and megasporogenesis in P. ottonis. (A) Ovule in early developmental stage, with the archesporium composed of two archespores. (B) Ovule in megaspore mother cell stage (arrow: third layer in the inner integument). (C) Ovule in megaspore dyad stage (* detail) (arrow: third layer in the inner integument). (D) Ovule in megaspore triad stage, with micropylar megaspore in metaphase (** detail) (arrow: third layer in the outer integument). (E) Ovule in functional megaspore stage. (F) Megaspore triad and megaspore mother cell in the same ovule (arrows: megaspores of the triad). (G) Megaspore triad and immature binucleate gametophyte in the same ovule (arrows: megaspores of the triad; arrowheads: nuclei of immature gametophyte). ar, archesporium/archespore; dy, megaspore dyad; fm, functional megaspore; fu, funiculus; ig, immature gametophyte; ii, inner integument; mmc, megaspore mother cell; nu, nucellus; oi, outer integument; tr, megaspore triad.

Fig. 6.

Ovule development and megasporogenesis in P. rechensis. (A) Ovule in early developmental stage, with the archesporium composed of a single archespore. (B) Ovule in megaspore mother cell stage. (C) Ovule in megaspore dyad stage, with two megaspores in metaphase indicating transition to tetrad. (D) Megaspore triad, with micropylar megaspore in metaphase (arrow). (E) Ovule in tetrad megaspore stage. (F) Functional megaspore undergoing vacuolation (arrows: degenerated megaspores). (G–K) Detailed process of megasporogenesis, with each detail accompanied by an image obtained after aniline blue test under epifluorescence. (G) Megaspore mother cell. (H) Megaspore dyad. (I) Megaspore triad with initial deposition of callose inside micropylar and intermediate megaspores (arrowheads). (J) Megaspore tetrad with callose deposition inside all three megaspores (arrowheads) slightly bigger than in the triad phase. (K) Functional megaspore (arrow). Presence of callose inside degenerated megaspores (arrowhead). ar, archesporium/archespore; dy, megaspore dyad; fu, funiculus; ii, inner integument; mmc, megaspore mother cell; nu, nucellus; oi, outer integument; te, megaspore tetrad; tr, megaspore triad.

Fig. 7.

Megagametogenesis in P. erinacea. (A) Binucleate gametophyte, after the first mitosis of megagametogenesis (arrows: degenerated nucellar cells). (B) Second mitosis with both nuclei in metaphase (arrow: degenerated nucellar cells). (C) Tetranucleate gametophyte (arrowheads: nuclei). (D–F) Octonucleate gametophyte (arrows: degenerated nucellar cells). (G) Fully developed ovule. (H) Detail of the mature gametophyte. (I) Starch grains present in the mature gametophyte, observed through the use of polarized light. an, antipodal cell; eg, egg cell; fu, funiculus; ii, inner integument; mg, mature gametophyte; n1–n8, nuclei; nu, nucellus; oi, outer integument; pn, polar nuclei; sy, synergid.

Fig. 8.

Fully developed ovule and gametophyte in Parodia. (A) P. haselbergii. (B) P. leninghausii. (C) P. microsperma. (D) P. ottonis. (E) P. rechensis. (F) Detailed view of the gametophyte of P. rechensis, allowing observation of two antipodal cells in the process of degeneration. (G–K) Starch grains present in the mature gametophyte, detected using polarized light. (G) P. haselbergii. (H) P. leninghausii. (I) P. microsperma. (J) P. ottonis. (K) P. rechensis. an, antipodal cell; eg, egg cell; sn, secondary nucleus; st, starch grains; sy, synergid.

Ovule initiation

The ovary of the six species is multi-carpellary, unilocular, and has parietal placentation. The placenta is divided into three meristematic zones: the first zone, also known as the dermal layer; the second zone, known as the subdermal layer; and the third zone, or central layer, which characterizes the Parodia ovule as trizonal. The initiation of the ovule primordium occurs in the third zone of the placental meristem, accompanied by the subdermal and dermal layers in its development (Figs 1A, 2A, 3A and 4A). After a brief development, the ovule primordium divides, giving rise to secondary primordia (Figs 2A, 3A and 4A), which develop into clustered ovules with branched funiculi.

At the apex of the subdermal layer in the ovule primordium, a cell stands out from the rest due to its larger size and more prominent nucleus and nucleolus (Figs 1A, 2A, 3A and 4A). This cell, known as the subdermal initial cell, undergoes periclinal division to give rise to the parietal cell and the archesporium, which characterizes the Parodia ovule as crassinucellate. The archesporium may contain one or two archespores (Table 1; Figs 1B, 2B, 3B, 4B, 5A and 6A). The differentiation process of this cell causes it to stand out from the surrounding nucellar cells, with a larger size and more prominent nucleus and nucleolus. In one species, specifically P. microsperma, the mitotic division of one of the archespores was observed (Fig. 4B).

Upon the initiation of archesporium differentiation, the inner and outer integuments originate from the dermal layer, with the former beginning before the latter (Fig. 3B). The development of the integuments marks the archesporium stage (Figs 1B, 2B, 3B, 4B, 5A and 6A). Both integuments are predominantly composed of two layers of cells, becoming pluristratified in the apical region as the ovule progresses in its development (Figs 1C, 2E, 3I, 4H, 5D and 6E, for example). In some cases, in the region of origin of the integuments a third layer can be observed, which may vary in degree of development (Figs 1C, D, 2C–E, 3C and 5B–D). In all species, the ovules are classified as campylotropous due to the curvature observed in the nucellus and the region of origin of the integuments.

Megasporogenesis

From the division of the subdermal initial cell, three types of megasporangium development can occur in Parodia: the first one, where the pre-meiotic cellular differentiation (PCD), which is characterized by the increase in cell volume, dense cytoplasm and larger nucleus and nucleolus, is restricted to only one cell, or archespore, and this cell differentiates into a megaspore mother cell [observed in P. erinacea (Fig. 1B–D), P. haselbergii ( Fig. 2B, C, F) and P. rechensis (Fig. 6A, B, G)]; the second, where the PCD extends to one or two archespores, with both potentially developing into a megaspore mother cell in the latter case [observed in P. microsperma (Fig. 4B, C)]; and the third, where the PCD extends to two archespores, and one or both of them can differentiate into a megaspore mother cell [observed in P. leninghausii (Fig. 3B, C) and P. ottonis (Fig. 5A, B, F, G)]. Note that, in the megaspore mother cell stage, the integuments are already in an advanced stage of development, in contrast to what is observed in the archesporium stage.

After completing the differentiation, the megaspore mother cell of P. haselbergii, P. leninghausii, P. microsperma, P. ottonis and P. rechensis initiates meiosis I, resulting in the formation of a dyad of megaspores (Figs 2D, G, 3D, E, 4D, 5C and 6C, H). In P. microsperma, we observed a projection of the dyad along with the nucellus to extend beyond the boundaries of the micropyle (Table 1; Fig. 4E, F). Both megaspores of the dyad initiate meiosis II, but only the chalazal megaspore completes it, resulting in the formation of a linear triad of megaspores in most of the mentioned species (Figs 2E, H, 3F, 4G and 5D, F, G), except in P. rechensis (Table 1). In this species, meiosis II occurs in both the chalazal and micropylar megaspores, leading to the formation of a linear tetrad of megaspores (Fig. 6E, J). Meiosis II can be synchronous or asynchronous, with the latter allowing the visualization of linear triads of megaspores (Fig. 6D, I). Despite this difference, in this species, as well as in P. haselbergii, P. leninghausii, P. microsperma and P. ottonis, the female gametophyte is monosporic in origin, with the chalazal megaspore starting the vacuolation process, resulting in the formation of two vacuoles, one at each pole of the cell, while the other megaspores undergo degeneration (Figs 2I, J and 3G, H). As a result, the chalazal megaspore becomes the functional megaspore from which the female gametophyte will develop (Figs 2J, 3I, 4H, I, 5E and 6F).

This particular development pattern described does not apply to P. erinacea, the only apomictic species among the six studied (Fig. 1). In this species, the archespore directly differentiates into a megaspore mother cell (Fig. 1B–D), which does not initiate meiotic division and, as result, does not give rise to megaspores. This results in a longitudinal elongation of the cell, with the formation of small vacuoles that later accumulate polysaccharides and merge, forming two larger vacuoles, one at each pole of the megaspore mother cell (Fig. 1C–G).

In all studied species, callose deposition was not observed on the walls of the megaspore mother cell, dyad, triad or tetrad. Callose was only observed, although in low quantities, inside the megaspores that undergo degeneration, and subsequently in the degenerated megaspores (Fig. 6G–K). The accumulation of this substance begins in the triad phase, and is barely noticeable due to its low concentration (Fig. 6I). As the megaspores develop, the concentration gradually increases (Fig. 6J). A significant concentration is only observed when both megaspores, or three in the case of P. rechensis, are degenerated, and the chalazal megaspore is established as the functional megaspore (Fig. 6K).

Megagametogenesis

The process of megagametogenesis follows a similar pattern in all six species analysed; therefore, we used P. erinacea as an illustrative model (Fig. 7). Starting from the functional megaspore in most species and the megaspore mother cell in P. erinacea, three successive mitotic divisions without cytokinesis give rise to two (Fig. 7A), four (Fig. 7C) and eight nuclei (Fig. 7D–F), which are haploid in P. haselbergii, P. leninghausii, P. microsperma, P. ottonis and P. rechensis, and diploid in P. erinacea. Once formed, the nuclei occupy the pole region of the gametophyte, evenly distributed and separated by a central vacuole containing polysaccharides (Fig. 7D–F). The development of the gametophyte is accompanied by the degeneration of surrounding nucellar cells (Fig. 7A–D, F).

After cytokinesis and cellular differentiation, the gametophyte of the Polygonum type is composed of seven cells and eight nuclei: three antipodes located at the chalazal pole; two synergid cells and one oosphere, which together form the oospheric apparatus at the micropylar pole; and a central cell, which may have either two polar nuclei (Fig. 7G, H) or a single secondary nucleus formed after the fusion of polar nuclei (Fig. 8A–F). In P. leninghausii and P. rechensis, the antipodal cells are ephemeral and begin to degenerate before floral anthesis (Table 1; Fig. 8B, F). All species exhibit mature gametophytes containing starch grains in their interior (Figs 7I and 8G–K).

Ovule malformation in P. rechensis

In P. rechensis, the structure of the ovule may be partially or completely affected during its formation and development (Fig. 9). Approximately half of the examined ovaries showed some deformities, such as compromise of the ovary wall resulting in loose cells and abnormal intercellular spaces (Fig. 9A), malformation of the coverings with the ovule’s outer integument consisting of only one layer of cells and subsequent malformation of the micropyle, and degeneration of the funiculus and chalaza cells (Fig. 9B–D).

Fig. 9.

Ovule malformation in different developmental stages in P. rechensis ovules. (A) Structural impairment of the ovary and placenta resulting in ovule abortion (arrow: collapsed placenta; arrowhead: intercellular spaces due to loosened cells). (B) Beginning of the degeneration process in a young ovule. In this stage, the outer integument does not develop the second layer of cells (arrow) and the funiculus presents abnormal intercellular spaces (arrowheads). (C) Ovule in triad megaspore stage. Both integuments fail to develop the second layer of cells, with the outer integument showing cellular degeneration near the micropyle (arrow). Abnormal intercellular spaces in the funiculus (*). (D) Ovule in advanced development stage, with malformation of the integuments, cellular degeneration of the outer integument near the micropyle (arrow), and degeneration of cells in the funiculus and chalaza region (arrowheads). ar, archesporium/archespore; fu, funiculus; ig, immature gametophyte; ii, inner integument; mi, micropyle; oi, outer integument; ov, ovule; tr, megaspore triad.

DISCUSSION

Embryological aspects of Parodia

There is limited knowledge on the initiation process of the ovule in Cactaceae. Most studies focus on later stages of development or mature structures (Engleman, 1960; Flores and Engleman, 1976; Hernández-García and García-Villanueva, 1991; Fuentes-Pérez et al., 2009; De Almeida et al., 2010; Cisneros et al., 2011; Jiménez-Durán et al., 2014; Camacho-Velázquez et al., 2018; Gentz et al., 2023; González et al., 2023). In the genus Parodia, the ovules in all six species exhibit three distinct layers, with cell divisions primarily occurring in the central layer in a periclinal manner, while the subdermal and dermal layers follow suit in an anticlinal direction. These subdermal and dermal layers play a crucial role in ovule development, with the former being the site of differentiation for the archesporial initial cell, which will eventually divide to form the parietal cell and the archesporium, and the latter responsible for the formation of both internal and external integuments. A similar ovule development pattern has been observed in various flowering plant families (Bouman, 1984). Among the Cactaceae family, only Pachycereus militaris has been studied regarding the origin of its integuments, which were also determined to be dermal in nature (Núñez-Mariel et al., 2001). However, the origin of ovule primordia in this family has not yet been investigated. In Cactaceae, only a small fraction of studies involving ovules have focused on early stages, with most beginning their analyses from the archesporial stage (Engleman, 1960; Núñez-Mariel et al., 2001; Strittmatter et al., 2002; Hernández-Cruz et al., 2019), a developmental phase during which some information about ovule initiation has already been lost.

The presence of an archesporium with one or more archespores in Cactaceae has already been reported (Engleman, 1960; Johri et al., 1992; Núñez-Mariel et al., 2001; Strittmatter et al., 2002). In P. microsperma, one of the species with a bicellular archesporium, we observed that the archespores have the ability to undergo mitotic division (Fig. 4B), a rare event in megasporangia (Pozner, 2001). In the present study, the term ‘archesporium’ is used according to the definition of Pozner (2001), which defines ‘archesporium’ as a cell or group of cells that undergoes PCD, and gives rise, either wholly or partially, to the meiotic tissue. The cell, or cells, that make up the archesporium are referred to as an ‘archespore’. An archespore can be distinguished from the surrounding cells by its larger cell volume, dense cytoplasm, and larger nucleus and nucleolus, which indicate PCD and the potential formation of meiotic tissue or a megaspore mother cell. However, the occurrence of PCD does not always guarantee the formation of such tissue, as it can occur in both a pre-mitotic and pre-meiotic interphase. It is only during the G2 phase of meiosis I that a cell commits to the meiotic cycle and is considered a megaspore mother cell, which can be identified from prophase I onwards (Pozner, 2001).

Two phenomena are related to PCD: the extension phenomenon and the gradient phenomenon. The extension phenomenon is directly linked to the number of archespores expressing PCD, while the gradient phenomenon relates to the synchrony, or lack thereof, of PCD in developing archespores. Both phenomena can also be applied to meiosis and follow the same logic. However, the gradient phenomenon in both PCD expression and meiosis only occurs if the archesporium consists of more than one archespore. Therefore, in an archesporium where PCD extends to more than one archespore it is possible for some of them to exhibit more pronounced differentiation than others, demonstrating asynchrony due to the presence of a PCD gradient. Alternatively, all cells may begin and continue PCD synchronously, thanks to the absence of a gradient. The extension and gradient of PCD and meiosis are of great importance in ovule development (Pozner, 2001). According to Pozner (2001), these phenomena act as mechanisms of cellular selection, serving as (1) pre-meiotic selection mechanisms, (2) meiotic selection mechanisms, and (3) post-meiotic selection mechanisms.

In Parodia, the three selection mechanisms can be found. In the genus, the archesporium has been shown to be a complex structure, with PCD extending to one or two cells in the studied species. These species do not exhibit a PCD gradient and may or may not present a meiosis gradient (Table 1). Out of the six studied species, three exhibit only the pre-meiotic selection mechanism, which are P. erinacea, P. haselbergii and P. rechensis. In these species, PCD acts as a selection mechanism by extending to a single cell from the beginning, preventing the development of other archespores (Figs 1B, 2B and 6A); there is no meiosis gradient. In P. microsperma the selection mechanism can also be pre-meiotic. However, in some cases the meiotic selection mechanism is observed. In this case, two archespores may occur and both can develop and differentiate into megaspore mother cells (Fig. 4C), but only one will complete meiosis (Fig. 4C–H); a meiosis gradient was observed (Fig. 4C). On the other hand, in P. leninghausii and P. ottonis, more than one megaspore mother cell may develop, as both species exhibit post-meiotic selection mechanisms. Meiosis can be completed in more than one cell in both species, but in P. leninghausii the phenomenon of meiotic gradient does not occur, and the development of two gametophytes in the same ovule was not observed. In contrast, in P. ottonis, both the meiotic gradient phenomenon and the development of more than one gametophyte in the same ovule can be observed (Fig. 5F, G). In both species, the pre-meiotic selection mechanism can also occur, which prevents the two archespores from always developing into megaspore mother cells.

Despite variations in PCD extent, meiotic gradient and types of cell selection mechanisms, the majority of Parodia species exhibit the formation of a megaspore triad after meiosis (Table 1; Figs 2E, H, 3F, 4G and 5D). In these species, the chalazal cell of the megaspore dyad initiates meiosis II before the micropylar cell (Figs 2G and 4F); the former completes meiosis II and produces two megaspores, while the latter only begins but does not finish it (Fig. 5D). This demonstrates that the formation of the megaspore triad in Parodia is due to the suppression of the meiosis II in the micropylar megaspore of the dyad, rather than a lack of cytokinesis as suggested in other Cactaceae genera (Engleman, 1960; Núñez-Mariel et al., 2001). According to Bouman (1984), this suppression of meiosis in the micropylar cell of the dyad is considered a reduction trend additional to the reduction of megaspore numbers. However, in P. rechensis this suppression does not occur, resulting in a tetrad of megaspores after the two phases of meiosis (Fig. 6G–J).

Meanwhile the suppression of meiosis may occur only in the micropylar megaspore of the dyad or not at all in the other species; in P. erinacea the meiotic suppression is complete (Fig. 1), making it an apomictic species, meaning it reproduces through apomixis. As defined by Nogler (1984), apomixis is ‘asexual reproduction through seeds’, resulting in a progeny genetically identical to the mother plant. The embryo, which will become the future sporophyte, can develop within a female gametophyte (gametophytic apomixis – apospory or diplospory) or from a somatic cell of the ovule (sporophytic apomixis) (Koltunow, 1993; Cornaro et al., 2023). In diplosporic gametophytic apomixis, a generative cell, such as the megaspore mother cell, produces a female gametophyte after three successive mitotic divisions. Meiosis in the generative cell can be initiated and later inhibited (meiotic diplospory) or not be initiated at all (mitotic diplospory) (Nogler, 1984; Koltunow, 1993). Based on our observations, P. erinacea reproduces through mitotic diplosporic gametophytic apomixis. This is the first report of apomixis in the genus Parodia and the sixth report in the Cactaceae family (Hörandl, 2024). Further studies could reveal whether, as in most species with diplosporic apomixis, the male gametophyte does not contribute to the genetics of the embryo (Koltunow, 1993) or if it does contribute to the formation of the endosperm.

Another novelty for Parodia and, in this case, a unique report for the Cactaceae family, is the absence of callose deposition in the cell walls of cells involved in the megasporogenesis process. Callose is a specialized polysaccharide that, in certain situations, can be part of the plant cell wall composition (Chen and Kim, 2009; Ünal et al., 2013). During megasporogenesis, for example, the identification of callose in specific regions of the cell wall of the megaspore mother cell is common, as well as being visible during the formation of the cell plate that forms between the megaspores. Its presence in these cases is often attributed to cellular isolation, so that the cells related to megasporogenesis can undergo independent and distinct development from surrounding cells (Rodkiewicz, 1970; Bouman, 1984; Ünal et al., 2013). However, in Parodia, during the entire process of megasporogenesis, callose is not observed as a constituent of the cell wall (Fig. 6G–H). Instead, what is observed is the simple presence of this substance inside the megaspore triad or tetrads (Fig. 6I, J), reaching its highest concentration when the megaspores degenerate (Fig. 6K). Thus, although further studies are necessary, it can be inferred that, in Parodia, callose is not related to cellular isolation. Considering that the megaspore that will become functional does not have any traces of this substance at any point during megasporogenesis, the presence of callose in this genus seems to be more related to the degeneration of non-functional megaspores. Additional studies must be carried out to confirm this hypothesis.

The use of embryological characters in the taxonomic context of Parodia

One of the motivations for studying the ovule initiation and processes of megasporogenesis and megagametogenesis in Parodia was to identify embryological characteristics that could be applied in the taxonomic context of the genus. The megagametogenesis phase of ovule development has shown the most consistency, making it less relevant for the proposed taxonomic approach. This is because the female gametophyte, which follows the Polygonum type and is formed by seven cells after three consecutive mitoses, is not only found in the six species studied, but also in the majority of angiosperms, making it a basic type of development (Willemse and Van Went, 1984). Therefore, it can be concluded that the most significant embryological characteristics for the discussion surrounding the circumscription of Parodia are observed earlier and throughout the process of megasporogenesis. However, even during these moments, as well as during megagametogenesis, the species also share certain characteristics. The presence of morphological similarities between species of the former genera, now considered synonyms of Parodia, was previously mentioned by Buxbaum (1967). In this case, the author argued for the distinction between Notocactus s.l. and Parodia s.s., and attributed the convergence of certain characteristics as evidence of their common origin. Furthermore, many of the shared characteristics among the studied species are also commonly described in other angiosperms (Bouman, 1984; Willemse and Van Went, 1984; Johri et al., 1992).

As a result of our study, certain characteristics have demonstrated their relevance in discussing the circumscription of Parodia. These include (1) the mitotic division of archespores and (2) the presence of nucellar projection in P. microsperma, (3) the cellular selection mechanism, (4) the apomixis in P. erinacea, and (5) the presence of a megaspore tetrad in P. rechensis. In Parodia species, the cellular selection mechanism can be pre-meiotic, meiotic, or post-meiotic. The meiotic selection mechanism was only observed in P. microsperma, a species that also exhibited other unique features, such as the ability for mitotic division of its archespores and nucellar projection during megasporogenesis. Similar analysis by Gentz et al. (2023) had previously shown that Parodia s.s., which includes P. microsperma, stands out from other formerly grouped genera currently included in Parodia s.l., due to the nucellar projection during the mature gametophyte phase and the shift from endostomic to exoendostomic micropyle. Although the cellular selection mechanism may be shared among the other species, each one presents a unique quality that sets it apart from the others. For species with only the pre-meiotic cellular selection mechanism, the distinguishing features include (1) P. erinacea with gametophytic apomixis of the mitotic diplospory type, (2) P. rechensis with the formation of a megaspore tetrad, and (3) P. haselbergii, whose development most closely resembles the commonly described pattern for other Cactaceae genera (e.g. Engleman, 1960; Núñez-Mariel et al., 2001; Jiménez-Durán et al., 2014). As for species with a post-meiotic cellular selection mechanism, the singularity is attributed to the presence of a gradient during meiosis and the development of two female gametophytes in P. ottonis, while P. leninghausii lacks this gradient and only develops one female gametophyte.

Thus, upon applying the data obtained from this current study to the circumscriptions related to Parodia [i.e. Buxbaum (1967), Ritter (1979) and Hunt and Taylor (1986)], we can observe a greater affinity with Ritter’s (1979) proposed circumscription. In Kakteen in Südamerika, Ritter argues for the restoration of the genera classified as subgenera of Notocactus s.l. by Buxbaum (1967), re-establishing the genera Acanthocephala, Eriocephala, Notocactus and Wigginsia, and creating the new genus Brasiliparodia. This action is justified by the identification of various distinct morphological characteristics among the genera, particularly in relation to the vegetative body, flowers, fruits and seeds; not coincidentally, Ritter is considered a taxonomic splitter. Furthermore, Ritter also supported the idea of a relationship between these genera and the genus Parodia, but never their union in a single genus. The data presented in this study demonstrate a significant variation during ovule formation, making it difficult to establish a pattern for the genus Parodia s.l. It is important to note that these differences are being observed in an embryological process that, typically, is conserved and follows a pattern in well-established genera (Bouman, 1984; Johri et al., 1992). Additionally, this is not the first study involving embryological data in which the results align with Ritter’s (1979) circumscription; in Gentz et al. (2023), the characteristics observed in fully developed ovules could further support the author’s idea. Thus, we emphasize the need to integrate morphological and embryological data with molecular data in this case, in order to better understand the relationships and delimitations of Parodia and its aggregates.

Implications of ovule malformation in P. rechensis

The occurrence of defects and morphological changes in P. rechensis ovules is an important issue due to the current conservation status of the species and difficulties in ex situ conservation efforts (Larocca and Machado, 2013). This species is only known to exist in two populations in Caxias do Sul, Rio Grande do Sul, Brazil, and is currently classified as critically endangered (CR) on the IUCN Red List (Larocca and Machado, 2013), the Red Book of Brazilian Flora (Martinelli and Moraes, 2013) and the List of Threatened Plant Species in Rio Grande do Sul (SEMA, 2014). The longer-known population, which includes the species’ type specimen, currently has a maximum of only ten individuals. However, in 2021 a second population was discovered, providing hope for the future of the species (IMA, 2021). The main contributors to the species’ decline include illegal collecting and habitat alteration (CNCFlora, 2012; Larocca and Machado, 2013; IMA, 2021). The latter is primarily caused by the presence of invasive exotic species such as Pinus, which suffocates and shades the specimens due to leaf litter accumulation and growth habits, and wild boars, which trample the plants as they move through the areas where the species is found (N. F. Fagundes and A. D. Nilson, pers. comm. Porto Alegre Botanical Garden, Brazil). Additionally, our data suggest a potential new threat to the decline of the species: ovule malformation, the precursor to seed development.

It is widely understood that the future sporophyte develops from the embryo inside a seed. However, in order for a new sporophyte to be able to grow, the seed must be viable. To produce a viable seed, the events that occur within the ovule, such as those during megasporogenesis and megagametogenesis, must be initiated and completed without any absence or interruption of the common embryological stages of the species. If this does not happen, the seed’s development can be disrupted and it may become non-viable (Koltunow, 1993). Changes in the structure of the ovule can also result in a non-viable seed, as it can hinder the pollen tube from reaching the female gametophyte in the case of alterations in the micropylar region, or affect the formation of the seed’s protective coats in the case of alterations in the ovule’s integuments, which serve to protect the developing embryo.

During the study of the processes of megasporogenesis and megagametogenesis, abnormalities were observed during the formation and development of P. rechensis ovules. In almost half of the analysed ovaries, the ovules followed a normal development, able to form an archesporium, triads/tetrads and gametophytes. However, at unspecified times and for unknown reasons, the structure of the ovule began to degenerate, affecting even the products of megasporogenesis and megagametogenesis that were already formed. It is important to note that such abnormalities are occurring in a species that is facing a strong population decline. Nothing similar was observed in any of the other five species studied in the present research, or in the 15 species of Parodia studied in Gentz et al. (2023). Therefore, we believe that the malformation of ovules may also be a contributing factor to the observed decline in the species. Our hypothesis is linked to the fact that abnormal ovules do not develop, and as a consequence no viable seeds are formed in these cases, which can lead to a decrease in the population. Conducting genetic studies and focusing on the count of viable and non-viable seeds is necessary to gain a better understanding of the reasons for the malformation of ovules in P. rechensis and how it may be affecting the species’ populations.

FUNDING

This work was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES), Finance Code 001, which awarded a grant to the first author, in part by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq; grant number 303840/2019-6) as a research grant to the second author, and in part by Plant Anatomy Laboratory (LAVeg) own resources.

CONFLICT OF INTEREST

The authors have no competing interests to declare that are relevant to the content of this article.