The traits of angiosperm flowers are strikingly diverse, where they vary in terms of color, scent, size, and the type and amount of reward they offer to pollinating animals (Conner, 1997). The evolution of floral traits has always been a focus of evolutionary biologists, but the evolutionary mechanism is still unclear. As early as 1970, Stebbins (1970) proposed the “most effective pollinator principle” where a flower’s traits will be molded by those pollinators that are most effective and frequent in the local area (Mayfield et al. 2001; Huang, 2014). This view was supported by studies on species and phylogenetic levels (Whittall and Hodge, 2007; Pauw et al. 2009; Paudel et al. 2016; Johnson & Raguso, 2016). For instance, the emergence of many specialized pollination systems in plants is related to cooperation between plants and pollinators, and selection for their traits, where the different traits and foraging preferences of pollinators mediate the formation of different flower traits (Hattori et al. 2015). Classically, plant assemblages pollinated by the same pollinator are under similar selective pressures, so they should exhibit convergence in terms of their floral characteristics (Johnson and Steiner, 2000; Johnson and Raguso, 2016). By contrast, ecotypic divergence will occur when plants are pollinated by behaviorally or morphologically different insects in different populations (Fenster et al. 2004; Sun et al. 2014; Peter and Johnson, 2014). Therefore, flower traits reflect adaptations to specific pollinator groups (Fenster et al. 2004; Faegri and van der Pijl, 1979; Pyke, 2016), thereby leading to the emergence of the conceptual framework of pollination syndromes. The evolution of a structure should be consistent with the function that it performs, and thus, function is the main driving force related to structural evolution. Therefore, in order to understand the special structures and morphology of flowers, adaptive pollination (or improving reproductive fitness) has been the main focus of researchers, such as changes in flower color and shape (Irwin & Strauss, 2005; Veiga et al. 2016), heterostyly (Gilmartin, 2015; Yuan et al. 2017), the staminal lever mechanism (Claßen-Bockhoff et al. 2004), and a flexible style for pollination (Li et al. 2001), which are plant adaptations to various environmental pressures to satisfy the function of reproduction. Due to the identification of these features, researchers have increasingly focused on the diversity and sophistication of plant morphological structures under various selection pressures (biotic and abiotic), and the function of plant staminode are among these structures.
Staminodes are stamens that have lost the main function of producing pollen and they often perform important secondary flower functions (Jennifer and Harder, 2000). Studies of the phylogenetic distribution of staminodes indicate that staminodes usually appeared during evolutionary reduction of the androecium as a result of long-term natural selection in the evolutionary process. Previous studies by botanists and evolutionary biologists focused on morphological descriptions of staminodes, and deeper research into functional staminodes has only begun to conducted in recently (Botnaru and Schenk, 2015, Hou et al. 2022a; b). Staminodes form a physical barrier structure in some genuses, such as Penstemon, Verticordia and Darwinia, and it is generally considered that the ecological function of these staminodes is to protect the internal structures of flowers from insects (Rodríguez-Riaño, 2015). However, recent studies have found that these staminodes may have functions in insect screening, where plants present biomechanical barriers in the form of staminodes to limit visitors from reaching flower rewards (or increase the cost of rewards) in order to screen for effective pollinators (Córdoba and Cocucci, 2011; Hou et al. 2022a, b). The screening mechanism is an important component of the screening game between plants and pollinators, and it serves as a useful theoretical framework for understanding the maintenance and drivers of animal and plant coevolution. In addition, studying screening mechanisms provides a theoretical basis for understanding insect-mediated plant flower morphology evolution as well as the specialized and generalized pollination mechanisms involving plants and pollinators. However, only a few studies have investigated insect screening mechanisms and that is mainly descriptive reports (Córdoba and Cocucci, 2011; Hou et al. 2022a), which are not sufficient for developing a theoretical understanding of insect screening mechanisms.
Many plant flowers have movable parts that must be actively handled by insects to let sexual organs be contacted (Córdoba and Cocucci, 2011). Similarly, Delphinium caeruleum flowers have a complex structure, where two blue staminodes form a “double door” structure which cover anthers and stigmas. The “double door” staminodes can be opened under a certain external force to expose the male and female stamen structures hidden under the staminodes. The pollinator must open the “double door” structure (staminodes) to achieve rewards, and complete pollination by contacting with the stamens and pistils below the staminodes, and thus, this structure may form an effective screen (biomechanical screening) that selects for insects with greater strength. Córdoba and Cocucci (2011) designated the mechanical strength required to open a forcible floral mechanism as operative strength. In addition, D. caeruleum has a spur formed by the extension of petals, with nectar at the end of the spur, which may also form a typical “length” screening structure where only insects with a long proboscis might access the nectar in the spur (Newman et al. 2014). Therefore, we hypothesized that successful pollination of D. caeruleum might be achieved through a complex mechanism mediated by a combination of length screening and biomechanical screening.
Based on the hypotheses described above, we examined the ecological functions of the staminodes and nectar spur in D. caeruleum to determine whether this composite mechanism performs the function of screening for effective pollinators. We conducted our study in three population of D. caeruleum in northwest China, i.e., in Hezuo, Haibei, and Tianzhu. We investigated the ecological functions of these special flower morphological traits (a compound coevolution mechanism combined with length coevolution), and then explored the evolutionary mechanism associated with the flower morphology and structure. In particular, we plan to solve following questions. (1) What are the differences in the flower traits and pollinators among different populations? (2) Do the delicate structures of the staminodes and nectar spur play roles in screening efficient pollinators for D. caeruleum? (3) If this is the case, will the different combinations of insects in different populations lead to differences in the operative strength of staminodes and the spur length (i.e., different local groups might differ in terms of biomechanical and length matching)?