Deceptive Ceropegia sandersonii uses an arabinogalactan for trapping its fly pollinators

Abstract

Introduction

Around 130 million years ago, angiosperms started to emerge (Crane et al., 1995; Magallón et al., 2015), and since then, surfaces that reduce or modulate insect attachment have evolved. These surfaces often protect the plants from herbivores, while deceptive trap flowers and carnivorous plants use such surfaces to trap insects (Poppinga et al., 2010; Bröderbauer et al., 2012) for pollination purposes and to use them as a food source, respectively.

Plants reduce the ability of insects to adhere to their surfaces through a variety of mechanisms, such as surface sculpturing, contamination and/or aquaplaning. Anti-adhesion via surface texture is achieved by convex, dome-like, papillae-like or tabular-shaped cells that result in roughness (Poppinga et al., 2010). Such an arrangement of cells on the plant surface prevents claw interlock, by not providing the adequate edges or ridges that insect claws can successfully lock on to (Juniper et al., 1989; Vogel & Martens, 2000). Furthermore, these surfaces also act to greatly reduce the overall contact area for insect footpad adhesion thus reducing possible adhesion forces (Gorb & Gorb, 2002; Gorb et al., 2005; Poppinga et al., 2010). An anisotropic arrangement of lunate cells or small trichomes can also achieve the same effect (Juniper et al., 1989; Gaume et al., 2002; Poppinga et al., 2010; Gorb & Gorb, 2011; Bauer et al., 2012). While all these surface types reduce the contact area on a microscopic scale, the formation of three-dimensional epicuticular waxes reduces the contact area on a nanoscopic scale (Gorb, 2001; Gorb & Gorb, 2017). This is achieved by radial ridges (see Bohn & Federle, 2004) on plant surfaces by superimposed wax crystals or cuticular folds (Bohn & Federle, 2004; Prüm et al., 2012; Surapaneni et al., 2021) among others. Anti-adhesive properties via contamination are achieved by interfering with the adhesive properties of the insect footpads (Knoll, 1914; Poppinga et al., 2010). Here epicuticular wax layers display either filamentous or tubular crystals (Federle et al., 1997; Gaume et al., 2004; Borodich et al., 2010) or platelets (Juniper et al., 1989; Gaume et al., 2004; Gorb et al., 2005; Purtov et al., 2013), both of which may easily break off the plant surface and attach to the insect's adhesive footpads, thus contaminating the foot surface. Another strategy to reduce insect attachment is to soak up the adhesive liquids that insects produce on their footpads through a highly porous wax layer present on the plant epidermis (Gorb et al., 2017; Gorb & Gorb, 2017). Footpad adhesion can also be disturbed via films of water or fatty oils on the plant surface (Knoll, 1922; Bohn & Federle, 2004). These can either reduce adhesion directly or can contaminate the feet of the insect. Some plants may even combine more than one of the above described features to boost their anti-adhesive properties. Among such plants are the deceptive kettle trap flowers Ceropegia stapeliiformis Haw. (Vogel, 1965) and Ceropegia sandersonii Decne. ex Hook. f. (Vogel, 1961; Heiduk et al., 2020) that have corolla lobe epiderms that are equipped with convex outer walls topped by a papilla secreting a single small liquid droplet on its tips. When an insect footpad contacts a droplet, it immediately attaches to the pad making the footpad nonfunctional (Vogel, 1965). The corolla of Ceropegia (Apocynaceae) trap flowers constitute a kettle at its base (ostiolum), followed by the tube and free corolla tips which are, however, often brought together at the tips again as in C. sandersonii, where the fused tips form a parachute-like structure that ‘caps’ the corolla (Vogel, 1965; Fig. 1a,b). In this species, kleptoparasitic fly pollinators are attracted to the flowers by their scents mimicking prey of the flies (honey bees) (Heiduk et al., 2016). The scents are produced by the flowers in osmophores found in specific regions on the underside of the corolla cap (Fig. 1a,b). In-between the osmophores are the gliding zones (Fig. 1b). Centrally, the purple-black ‘uvula’ (Fig. 1a) protrudes into a cone-like structure, representing the centrally fused ends of the five gliding zones of the cap (cf. Heiduk et al., 2020).

Fig. 1

Open in figure viewerPowerPoint

A flower of Ceropegia sandersonii and a fly pollinator. (a) Functional units of the flower. (b) The underside of the corolla cap consists of osmophores (O) and gliding zones (G) (see Heiduk et al., 2020). (c) The gliding zone with conical-shaped epidermal cells, each with a bristle-like central protuberance and a liquid droplet on its tip. (d) Cryo-SEM image showing the hexagonal cells of the gliding zone. (e) A Desmometopa fly pollinator with solidified droplets on the tip of its front leg, and a pollinarium attached to the head (P, pollinium). (f) Tarsus of Desmometopa with adhesive pad and claws (scanning electron microscopy). (g) A tarsus of Desmometopa contaminated with solidified droplets.

Pollinators land on the flower, enter the corolla through one of five openings and, when moving to a gliding zone, slip and slide down the funnel to finally get temporarily trapped in the kettle, where downward-pointed hairs prevent them from escaping (Vogel, 1961). After pollination, the hair-like trapping mechanism becomes nonfunctional, allowing the flies to escape and potentially pollinate another flower (Vogel, 1961).

This complex pollination mechanism has already been studied from some key perspectives, with most recent research focusing on the development of the flowers (Heiduk et al., 2020) and on the chemical nature of the scents produced by the osmophores and their importance in attracting and fooling the pollinators (Heiduk et al., 2016). Many questions remain about the chemical and physical aspects of the droplets in the gliding zone. It seems that the droplets somehow contaminate the pollinators feet (Vogel, 1965). In order to better understand the molecular and physical mechanisms that control the function of the droplets, we studied in detail how the liquid droplets found in the gliding zones of C. sandersonii flowers behave when getting in contact with the footpads of fly pollinators. Due to the metastable nature of these droplets, which made a detailed physical characterization extremely difficult, we focused mainly on characterizing the chemical composition of the droplets, which may assist further in understanding function. Droplets were collected to analyze their chemical composition. Our data revealed that the droplets consist of a water-soluble arabinogalactan polymer that solidifies on the feet of pollinators, interfering with the adhesive properties of the insect footpads by contaminating them.