Floral biology of Ziziphus mauritiana (Rhamnaceae)
- First Online:
- Cite this article as:
- Tel-Zur, N. & Schneider, B. Sex Plant Reprod (2009) 22: 73. doi:10.1007/s00497-009-0093-4
- 241 Views
Floral development of the synchronous dichogamous species Ziziphus mauritiana, as followed by light and scanning electron microscopy (SEM), was divided into 11 stages using a series of landmark events. Main cellular events happen synchronously in the female and the male structures, such as meiosis in micro- and macrosporocyte cells, tetrad microspore formation and appearance of the functional megaspore cell, and onset of embryo sac differentiation coinciding with mitosis in the microspores. The last stage was characterized by anthesis and continued development of the flower, beginning with anther dehiscence (male phase) and proceeding to the female phase, which was characterized by style elongation. Flowers exhibit synchronous protandrous dichogamy; anthesis takes place in the morning (group A, e.g., clone Q-29) and afternoon (group B, e.g., clone B5/4). Stigma receptivity started after the male phase and occurred synchronously and complementarily with pollen dispersal in the two clones. Pollen viability and production were similar in the two clones, but the pollen diameter of Q-29 was significantly larger than that of B5/4. This study provides the basis for understanding the biological mechanisms regulating floral development, thus expanding the prospects for Z. mauritiana breeding programs and for further molecular and genetic studies of this species.
KeywordsFloral organogenesisProtandrousRhamnaceaeSynchronous dichogamy
The genus Ziziphus (Rhamnaceae) comprises about 170 species native to the tropics and subtropics (Islam and Simmons 2006). Ziziphus mauritiana (Lamk.), known as ber, desert apple or Indian plum, is an evergreen thorny shrub or a small tree up to 15 m height, with many drooping branches. Under severe environmental conditions, it is a compact shrub.
Z. mauritiana is an economically important fruit crop, cultivated on marginal lands on a commercial scale, especially in India, where it is endemic (Arndt et al. 2001). Z. mauritiana is tolerant to extreme environmental conditions, including drought, high temperatures and saline water (Chrovatia et al. 1993; Mizrahi and Nerd 1996; Clifford et al. 1998). Cultivars introduced to Israel from India and planted in an experimental plot in the Southern Dead Sea Basin produced high yields (up to 80 kg/tree per year), despite the extreme temperatures prevailing there and saline water irrigation (Mizrahi and Nerd 1996). Thus, Z. mauritiana has a suitable genetic makeup for selection and breeding as a sustainable crop for the arid and semiarid regions of the world, which is an increasingly crucial challenge in the face of global climate change and ever-diminishing freshwater supplies (Pareek 2001).
Ziziphus has hypanthium-type flowers (a cup or tube bearing the floral parts above the base of the ovary of the flower) with five membranous hood-like petals. The ovary has two chambers, each with a single ovule, and is broadly attached at the base. Each fruit can bear two viable embryos. The five stamens are epipetalous, each being surrounded by a petal. The pistil is central, terminating in two stigmatic lobes (Galil and Zeroni 1967). Z. mauritiana flower buds emerge from an axillary position in a cyme inflorescence. The flowers open only for 1 day, with flowers at the top of the plant being the first to open. Each cyme has 12–14 buds (Vashishtha and Pareek 1979). Fruits are drupes, similar to small apples, with a crisp white flesh. Fruits may be oval or round, 2.5–6.3 cm long (Pareek 1997), depending on the cultivar. They are sweet and rich in antioxidant compounds and vitamins (Jawanda and Bal 1978; Machuweti et al. 2005).
Ziziphus flowers exhibit synchronous protandrous dichogamy, i.e., flowers of an individual plant mature in synchrony and anther dehiscence precedes stigma receptivity with little or no overlap between the sexual stages (Renner 2001). Such temporal separation of sexual functions has been referred to as a “temporal dioecism” (Cruden 1988). Genotypes of Ziziphus species are divided into two groups according to the timing of anthesis—morning or afternoon—designated groups A and B, respectively. In Z. mauritiana, anther dehiscence begins shortly after anthesis and terminates within 2 h (Desai et al. 1986) to 4 h (Josan et al. 1980). Flowering phases of the two types overlap, thus, utilizing complementary sexual morphs in orchard design are vital for successful cross-pollination and optimum yields. Z. mauritiana is reported to be self-incompatible (Godara 1980). The flowering period is prolonged; in Israel, the flowering season begins early July and continues till the end of October. Late pruning may delay the flowering season (unpublished data). Flowers are visited by different species of insects, including wasps, flies, butterflies, and bees (Mishra et al. 2004).
Galil and Zeroni (1967) found that two processes take place rapidly in the pistil of Z. spina-christi following anthesis: elongation of the style and stigma development. Stigmas are considered receptive when they can support germination of compatible pollen grains. Based on the observations of a sticky, shiny secretion from the stigma and on hand pollination tests, Z. mauritiana is reported to be receptive on the day of anthesis (Vashishtha and Pareek 1979; Desai et al. 1986). Yet, detailed studies of style elongation and stigma development and the relationship of these two processes to stigma receptivity in Ziziphus are lacking.
Similarly, almost no information is available on the genetic and molecular factors controlling synchronous dichogamy. A detailed morphological description of flower development would therefore provide a vital foundation for future studies that would ultimately give a better understanding of the mechanisms that regulate synchronous dichogamy. To this end, we undertook a systematic anatomical analysis of flower development based on histological sections and scanning electron microscopy (SEM) techniques. We followed the course of events through the period of flower initiation, development, and differentiation and divided the developmental process into 11 specific stages. In parallel, we studied pollen production, diameter, and viability and followed the temporal sequence of stigma receptivity after anthesis.
Materials and methods
Study site and plant material
This study was conducted in an experimental plot at the Sede-Boqer Campus of Ben-Gurion University of the Negev, located in the Negev Highlands, Israel (30°52″N, 34°46″E, 430 m above sea level). The Negev Highlands are characterized by cold and mostly sunny winters with mean daily maximum/minimum temperatures of 14.9/3.8°C, and by hot, dry summers with mean daily maximum/minimum temperatures of 32/17°C. Average annual rainfall is 80 mm with considerable deviation from year to year. Two Z. mauritiana clones, Q-29 (morning anthesis, group A) and B5/4 (afternoon anthesis, group B), previously selected in trial plantings in Israel, were used in the present work (voucher specimens are kept at the herbarium of the Hebrew University of Jerusalem, Israel). These clones were grafted onto Z. mauritiana seedlings and planted among other Ziziphus species in the experimental plot in native loess soil. Chemical fertilization (Poly-Feed DRIP 23:7:23 + 2MgO with micronutrients, Haifa Chemicals Ltd.) was supplied through a drip system. The plants were irrigated weekly with 56 l during the hot season and 16 l during the cold wet season (November–March). The grafted trees were 1.5–2.5 years old at the time of the study.
Clone Q-29 was chosen among several genotypes planted at the experimental plot in Sede Boqer, all with similar flower morphology, due to its profuse flower production and prolonged flowering season. Shoot tips, floral buds, and flowers from Q-29 were sampled for histological and SEM analysis at various stages of development. For histological analysis, tissues were fixed with FAA (formalin:acetic acid:alcohol, 1:1:3), and then placed on ice in a vacuum desiccator to facilitate the penetration of the fixative into the plant tissue. Hard plant tissues, characteristic of desert plants, hindered embedding and resulted in collapsed tissues, thus, we experimented with several embedding and cutting methods including cryostat sectioning using the Leica CM 3050 S with Tissue-Tek® O.C.T. The optimum method was fixation, dehydration with an ethanol series, and embedding in Paraplast Plus® (McCormick™ Scientific) according to Jackson (1991). Blocks were sectioned at 8–10 μm (Leica RM2235 rotation microtome) and stained with Safranin-Fast Green (Rusin 1999). Several microtome blade models made by various manufacturers were tried for sectioning and the best results were obtained with Patho Cutter-II 350 80 mm microtome blades (ERMA Inc., Tokyo, Japan). The sections were studied and photographed with an Axioimagera1 LED (Zeiss) microscope and an Axiocam HRC camera (Zeiss).
For SEM analysis, the plant material was fixed overnight with 3.7% formaldehyde, 50% ethanol, and 5% acetic acid and then dehydrated with an increasing ethanol gradient (up to 100%). The fixed tissues were critical-point-dried with liquid CO2, mounted on aluminum stubs with double-sided tape, coated with 11 nm of gold by using an automated sputter coater (Polaron E6700, high-vacuum evaporator), and then examined with a SEM (JEOL JSM-5610LV).
Flowers were collected at regular periods following anthesis. Whole flowers from the two cultivars were dissected under a stereomicroscope (SV6 FL, Zeiss) and stigma-style tissues were removed into 20 mM phosphate buffer pH 4.5, containing 0.1 M guaiacol and 0.1 M H2O2, until an orange-brown color was observed (approximately 1–2 min). At least ten stigmas from each clone were tested at each developmental stage. The tissues were studied and photographed with a stereomicroscope (SV6 FL, Zeiss) and an Axiocam HRC camera (Zeiss).
To determine the number of pollen grains per anther, ten flowers from each clone were harvested before stamen dehiscence, and the five anthers from each flower (50 anthers) were gently removed and vortexed in a tube containing 300 μl of 0.5 M sucrose. The pollen suspension was injected into a hemocytometer and the total number of grains within the 25-square counting area was determined. The number of pollen grains per anther was calculated from the following formula according to the hemacytometer procedure (Sigma, Product information Z359629): (mean number of pollen grains per square × 104 × 0.3)/50.
The viability of the pollen grains of the two clones was assayed using fluorescein diacetate (FDA) at a final concentration of 2 μg/ml (Heslop-Harrison and Heslop-Harrison 1970). Pollen grains from undehisced anthers were collected at the beginning of anthesis, brought to the lab and dispersed on a microscope slide in a drop of FDA stain. The slide was incubated for 4–5 min at room temperature and examined with an Axioimagera1 LED (Zeiss) fluorescent microscope, with a blue excitation (450–490 nm) filter set. Brightly fluorescing grains were scored as viable, while weakly fluorescing grains or those not exhibiting fluorescence were scored as non-viable. At least 500 grains were examined per flower. In addition, the equatorial diameter of 100 viable pollen grains was measured using the AxioVision AC Rel. 4.5 program (Zeiss).
Differences between the two clones in pollen performance—viability, diameter, and grains per anther—were analyzed using Student’s t test with confidence limits set at α ≤ 0.05.
Morphological indications and respective lengths of floral buds in Ziziphus mauritiana, clone Q-29, at various stages of development
Morphological indications and developmental activity
Bud length (mm)
Transition to reproductive stage and meristem initiation
Initiation of sepal primordium
Initiation of petal and stamen primordia
Initiation of carpel primordium
Stamen differentiation into anther and filaments
Meiosis: microsporocyte at early-prophase I
Microsporocyte at mid-prophase I. Macrosporocyte initiation.
Microsporocyte at late-prophase I. Meiosis in ovules
Microsporocyte at metaphase I
Tetrad formation of microspores and functional megaspore cell and three degenerating megaspores
Microspores development. Free young microspore with small vacuoles
Ovule showing large megaspore cell, and inner and outer integuments visible
Microspores with a large vacuole and a visible nucleus, the ovule with both inner and outer integument fully differentiated, micropyle visible, embryo sac formed
Before anthesis: nearly mature pollen and enlarged ovule
Anthesis and post-anthesis: 3 and 24 h after anthesis
Stage 1: floral initiation
Stage 2: initiation of sepal primordium
The floral meristem became broadened and undulated on the upper surface, these protuberances will become the sepal primordia (Fig. 1b). The floral meristem developed into a circular structure with sepal primordia (Fig. 1c). The five-sepal primordia arose in a spiral arrangement at the border of the floral primordium (Fig. 1d). Stage 2 continued with the elongation of sepals; the floral apex is nearly flat with a slight central depression (Fig. 1e).
Stage 3: initiation of petal and stamen primordia
Differentiation of petal and stamen primordia and a central concavity (arrow) was clearly distinguishable (Fig. 1f).
Stage 4: initiation of carpel primordium
Stage 5: stamen differentiation into anther and filament
Stage 6: ovule initiation
Stage 7: meiosis
Stage 8: tetrad formation
Anthers and ovules continued to develop. At the end of meiosis, each microsporocyte formed a tetrad of four haploid microspores isolated by a massive callose wall around the tetrad and between each monad (Fig. 4e), thereby completing the microsporogenesis phase. At the end of the meiotic division, when the macrosporogenesis phase was complete, a large functional megaspore and three degenerating megaspores were evident (Fig. 4f).
Stage 9: development of microspores and ovule, and embryo sac formation
Stage 10: before anthesis
Stage 11: anthesis and post-anthesis
We observed cessation of normal style elongation in 40–50% of the flowers during the flowering season. Styles that failed to elongate did not significantly change in length from anthesis, but peroxidase activity was clearly evident as spots on the apical part of the stigma 5 h after anthesis (Fig. 9c). Styles that did not elongate 24 h after anthesis retained peroxidase activity.
Peroxidase activity was also tested for clone B5/4, in which anthesis takes place in the afternoon, at 2:30 p.m. under local conditions. Peak of peroxidase activity occurred the following day, between 9:30 a.m. and 2:30 p.m., thus, stigma receptivity in this clone occurred synchronously with pollen dispersal in Q-29 (group A), facilitating pollen transfer and pollination between the flowers of the two different groups.
Pollen performance in clones Q-29 and B5/4
Pollen grains per anther
1073 ± 130a
1170 ± 279a
Pollen viability (%)
50.3 ± 3.6a
55.7 ± 6.1a
Pollen diameter (μm)
27.3 ± 0.4a
23.1 ± 0.3b
The morphological characteristics of flower development in Z. mauritiana, from transition and reproductive stages to the mature flower, were followed by applying histological and SEM techniques. We demonstrated parallel occurrence of the main cellular events in the female (carpel) and the male (stamens) structures, i.e., meiosis in both micro- and macrosporocyte cells occurred at the same stage, tetrad microspore formation occurred simultaneously with the appearance of the functional megaspore cell and the three degenerating megaspores and the onset of embryo sac differentiation coincided with mitosis in the microspores.
Petals of Colubrina species (Rhamnaceae) were reported to arise from a common stamen–petal primordium (Sattler 1973). Similar petal development was described for Z. mauritiana, i.e., each petal/stamen pair seems to arise by tangential splitting of an individual primordium (Medan and Hilger 1992). The link between petals and stamens is quite strong in Rhamnaceae, as both organs tend to be fused basally (Medan and Aagesen 1995), a phenomenon which was corroborated in our work by the observed simultaneous movement of both organs following anthesis.
Our staging system provides important information of the relative timing of morphological events that accompany Z. mauritiana flower development and may facilitate future investigation of the relationships between genes and morphological traits.
Knowledge of the timing and duration of stigma receptivity is vital for developing breeding programs for crops (Stone et al. 1995). Stigma receptivity or the ability of the stigma to support pollen germination was monitored from anthesis up to tissue necrosis. The reproductive strategy of Ziziphus species is synchronous dichogamous protandry; thus, pollen is shed before the stigma is receptive. This mechanism eliminates or significantly reduces the incidence of self- or cross-pollination between individuals of the same group. We observed and described three different stigmatic stages—immature, mature, and degenerated—in the synchronous maturation of the stigma of the two groups. Peroxidase activity increased as the stigma “matured”, reaching a peak when the stigma was most receptive to pollen (Dupuis and Dumas 1990; Dafni and Motte Manues 1998; McInnis et al. 2006). In Z. mauritiana, peroxidase activity was clearly visible after the end of the male phase and peaked 5–24 h following anthesis. A process of cell degeneration started in the stigmatic tissue shortly thereafter and the tissues finally became necrotic, showing a brown color. Previous work on other species has demonstrated different relationships between the two processes. For example, in Actinidia deliciosa, cessation of stigma receptivity occurs with degeneration of the papillae and loss of cellular integrity (Gonzalez et al. 1995), while Petunia flowers remain receptive despite necrotic stigmata (Herrero and Dickinson 1979). A relationship between stigma degeneration and stigma receptivity was not revealed in this study. Further work is thus needed to determine whether cell degeneration results in a cessation of stigma receptivity in Z. mauritiana.
Galil and Zeroni (1967) described the failure of the style to elongate in Z. spina-christi, accompanied by an undeveloped stigma. In practice, those flowers perform as male flowers, increasing the availability of pollen, then abscising. Our observations were consistent with those of Galil and Zeroni (1967), i.e., the process of style elongation in both clones stopped in 40–50% of flowers. However, in Z. mauritiana, peroxidase activity was observed in the stigmas of those styles, and the flowers did not drop off. Further experiments, such as hand pollination of flowers with undeveloped stigmas are needed in Z. mauritiana to determine whether such flowers contribute only as pollen donors or whether they can be pollinated and hence bear fruits.
Pollen stainability with the FDA experiment revealed low pollen viability in both clones. Low pollen viability may be associated with chromosome variations: for example, the meiotic irregularities in the tetraploid vine cactus Selenicereus megalanthus (Lichtenzveig et al. 2000) or with extreme environmental conditions, as reported for Mangifera indica L. (Issarakraisila and Considine 1994). Pollen viability in Z. mauritiana varies among the different cultivars: it may be as high as 91%, as was reported for the cultivars Chauhara, Sanur-1, and Shamber (Desai et al. 1986) or as low as 10%, as was observed for the octaploid cultivar Illaichi (Khoshoo and Singh 1963). Further research should include cytological work and plant development studies under a range of environmental conditions in order to identify the factor(s) that affect pollen viability in Z. mauritiana.
One of the factors affecting fertilization is pollen limitation, which seems to be quite a common phenomenon in many plant species. Burd (1994) showed that 62% of the 258 species of angiosperms reviewed were pollen limited at some time or place. The number of pollen grains per anther varies widely among the different plant species. An average of 993 pollen grains per anther were observed in the insect-pollinated Euphorbia boetica (Narbona et al. 2005), a number similar to that in Z. mauritiana. Since Ziziphus flowers contain a two-chambered ovary, each with a single ovule, the number of viable total pollen grains per ovule is about 1,350–1,600. Under our local conditions, a ratio of 1,350–1,600 viable pollen grains per ovule may be a limitation; however, if the undeveloped flowers in which the process of style elongation was halted are virtually male flowers, then pollen availability may not be a limiting factor at the whole plant level. Thus, further work to elucidate the contribution of flowers with undeveloped stigmas—as pollen donors or as normal flowers—as well as testing pollen germination in vivo will contribute valuable information toward answering the still open question regarding a possible pollen limitation.
Pollen grains in Z. mauritiana are tricolporate, as was previously reported in other Rhamnaceae species such as Ziziphus lotus L., Z. spina-christi, Rhamnus alaternus, and R. utilis (Lobreau-Callen 1976; Nasri-Ayachi and Nabli 1995). Similar to previous studies reporting that pollen grains ranged in diameter from 27 to 30 μm (Desai et al. 1986), our study found an average pollen diameter of 23.1 and 27.3 μm for B5/4 and Q-29, respectively. The importance of pollen size lies in the fact that it can reflect the storage capacity for particular nutrients that may affect pollen tube growth (Roulston et al. 2000). Further work is needed to elucidate whether genotypes with large pollen grains have a competitive advantage over those with smaller grains under extreme environmental conditions. Should this hypothesis prove correct, then the pollen diameter trait may be used for selecting superior cultivars in breeding programs.
The authors thank Mr. J. Mouyal and Ms. R. Jeger (Ben-Gurion University of the Negev) for their skillful technical assistance.