Morphological and numerical variation patterns of floral organs in two species of Eranthis

Highlights

• Floral structures are variable in Eranthis.

• The stamen number is the main contributor to structural variation.

• Morphological variation in Eranthis mainly occurred in petals.

• Both sepal and petal numbers fit the modified error function.

Abstract

There is wide diversity in flower structure and pattern in nature. The number and morphological variation of flower organs in Eranthis of the family Ranunculaceae is high, but the pattern of flower structure variation in this genus is unclear. In this study, the morphological and numerical variation in organs of Eranthis stellata Maxim. and Eranthis lobulata W. T. Wang were observed with stereomicroscopy and scanning electron microscopy, and the patterns of variation were summarized statistically. The results showed the following: (1) The structural patterns of the two species of Eranthis were not stable. There were 232 flower structural patterns among 300 individuals of E. stellata and 55 among 100 individuals of E. lobulata. The number of floral organs in E. lobulata was significantly less than that in E. stellata. (2) The morphological variation in flower organs in Eranthis mainly occurred in petals, especially in the upper lip, pseudonectaries and lower lip, e.g., the number of lobes, and pseudonectaries and inconsistency in the upper lip edge. (3) The coefficient of variation (CV) of flower organs was greater than 20%, with some values close to 40%, which indicates that the number of floral organs of the two species differ substantially, and that the number of floral organs was extremely unstable. The range in stamen number was the largest while that in sepal number was the smallest, and this fluctuation might be caused by the spiral phyllotaxis of floral organs.

Introduction

Flowers are peculiar reproductive structures found in angiosperms, and their phenotypes are one of the essential characteristics used in plant classification. Variation in the number of floral organs widely exists within populations or even individual species (Herrera, 2009; Ronse De Craene, 2016), and the phenomenon of indefinite organs is of particular systematic significance. In the basal group of angiosperms, the number of flower organs is unstable, and stamens and carpels are generally numerous; for example, in Amborellaceae, 9-11 tepals and 12-21 stamens occur in male flowers, while female flowers typically have 7-8 tepals, five carpels, and 1-2 sterile stamens (Endress and Igersheim, 2000). In Nymphaeaceae, (4-) 5-12 sepals are generally present, petals and stamens are numerous, and there can be up to 35 carpels (Wiersema and Hellquist, 1997). In addition, some groups, such as Sarcandra (Chloranthaceae), have a flower consisting of only one stamen and one carpel (Wu, 1999). In monocots, most groups are trimerous, and the number of organs is consistent and stable, while in Alismataceae, the number of carpels varies from 9 to 27 (Salisbury, 1926). In contrast to what is observed in the basal taxa, the number of floral organs in high-level taxa on the evolutionary branch of core eudicots tends to be stable (pentamerous) and small (mostly have two to five carpels in a whorl) (Endress, 2011), while for several core eudicots that have a large number of organs, such as Cactaceae, the number of stamens is usually more than 50 and can even be 1500 in some species (Parfitt and Gibson, 1997). However, the high numbers are the result of a secondary increase, in contrast to the situation on Ranunculaceae and other lower eudicots (Ronse De Craene and Smets, 1992; Endress, 2014). The variation in the number of floral organs is determined by multiple factors, such as genetic or epigenetic regulation, environmental changes, and randomness in gene expression (Thattai and van Oudenaarden, 2004).

Ranunculales, as the basal group of eudicots, is a vital transition group and sister to all other eudicots(Chase et al., 2016). High merism (such aspolyandry) and spiral phyllotaxis may be beneficial to variations in organ number, which indicates that members of Ranunculales are more plastic than core eudicots and monocots (Damerval and Becker, 2017). Ranunculales is composed of seven families, including Ranunculaceae, Papaveraceae, Berberidaceae, Lardizabalaceae, Menispermaceae, Circaeasteraceae, and Eupteleaceae (the latter without a perianth). For most Ranunculales, petals can secrete nectar as a reward to pollinators, and the petals are also called nectar leaves. Ranunculaceae is one of the core groups of Ranunculales, with 59 genera and 2500 species (Wang et al., 2009). The morphology and structure of the flowers in this family are diverse; the perianth differentiates into a calyx and corolla (e.g., Aquilegia L.) or has only tepals (e.g., Anemone), the flowers are radially (e.g., Ranunculus L.) or bilaterally symmetrical (e.g., Delphinium L.), and the number of perianth organs is low and stable (e.g., Dichocarpum W. T. Wang et Hsiao.) or large but unstable (e.g.,Laccopetalum giganteum (Weddell) Ulbrich) (Tamura, 1993; Endress and Matthews, 2006; Carrive et al., 2020). Previous studies of Ranunculaceae have focused on phylogeny, floral morphogenesis, nectary structure, pollination biology and molecular mechanisms of organ development, especially in Aquilegia and Nigella L., which are model plants for studies of organ development (Ren et al., 2011; Zhao et al., 2011; Sharma and Kramer, 2013; Ballerini et al., 2019; Espinosa et al., 2020). However, research on the variation in the morphology and floral organ number in this family is limited, except in the case of Anemone rivularis Buch.-Ham. ex DC. var. flore-minore Maxim., and results have suggested that differences in the number of floral organs are accompanied by variation in the morphology of the floral organs (Chang et al., 2005). Bachmann and Chambers (1978) proposed that numerical variation was caused by the accidental fusion or division of floral primordia and that the variation curve obeyed a Poisson distribution. Studies on flower organ developmental genes showed that mutations in the class C gene AGAMOUS (AG) could lead to an increase in the number of floral organs (Bowman et al., 1992). AP3-3 is an essential gene controlling the presence or absence of petals, and the deletion of AP3-3 causes the formation of calyx or stamen in the petal positions (Sharma et al., 2011; Zhang et al., 2013). Statistical analysis and mathematical modelling of the flower organ recognition gene expression domain suggested that the change in perianth number in 18 species of 5 genera of Ranunculaceaemay be due to two developmental processes (random transfer of the homeotic gene expression boundary and semi-concentric organ arrangement) (Kitazawa and Fujimoto, 2014).

Eranthis Salisb. belongs to Cimicifugeae of Ranunculaceae, with eight to ten species occurring from Europe to Asia (Erst et al., 2020), three of which are mainly distributed in north-eastern China and Sichuan. According to the records of the Flora of China (Li and Tamura, 2001), the floral organs in Eranthis vary widely: 5-8 petaloid sepals; 5-8 bilobed or slightly concave petals, sometimes with pseudonectaries; 10 or many stamens and 4-9 carpels. We are interested in examining the patterns of floral organ number variation in Eranthis as well as whether morphological variation exists in the floral organs. It remains unclear whether the variation pattern is consistent within the genus.

In this study, we selected two related geographically differentiated species (Park et al., 2019), namely, Eranthis stellata Maxim. (which is distributed in north-eastern China) and Eranthis lobulata W. T. Wang (which is distributed in Sichuan). Here, we aim to: (1) analyse the number of floral organs in E. stellata and E. lobulata; (2) observe the morphological variation in flower organs; and (3) evaluate the floral organ number variation pattern within Eranthis.