Desktop version

Home arrow Environment

  • Increase font
  • Decrease font


<<   CONTENTS   >>

MULTIMODAL FLORAL COMMUNICATION

No floral trait exists in a vacuum. Floral color accompanies other elements of display (pattern, shape, texture, orientation and architecture), chemical traits (scent, surface chemistry, nectar, pollen, resins and oils) and timing (diel rhythms of flower opening, scent emission and reward presentation, blooming phenology), all of which impact pollinator attraction and fidelity. Despite the fact that most floral traits and their associated sensory physiology in animal pollinators have been studied in isolation, most flowers advertise themselves in complex, multimodal sensory channels (Raguso 2004, 2008b). The focus of this chapter is to understand when and how floral scent works in concert with other floral traits. The multimodal signaling concept structure is outlined in three classes of nonmutually exclusive hypotheses (Table 16.1): (1) Content-related hypotheses (what information is furnished by flowers?), (2) Efficacy-related hypotheses (how do floral signals function against variable biotic and abiotic backgrounds?), and (3) Inter-signal interaction hypotheses (how do cognitive aspects of signal perception impact signal function?). This framework has been utilized in recent reviews of concerted changes in floral color and scent (Raguso and Weiss 2015) and holistic evaluations of floral phenotypic integration (Junker and Parachnowitsch 2015). These hypotheses provide conceptual tools for interpreting the complex interactions between scent and other floral traits revealed through manipulative bioassays.

Floral Manipulation - A Brief Prehistory

Nearly a century ago, Clements and Long (1923) outlined ingenious methods for manipulating the color, orientation and symmetry of living flowers in situ, providing a wealth of tools for subsequent studies (Figure 16.1a). Such methods and their modern counterparts include resupination (rotating floral orientation (Fulton and Hodges 1999), painting flowers to alter their colors (Melendez-Ackerman et al. 1997) and floral “mutilation” (removing petals, cutting nectar spurs or unzipping hoods to reveal other floral organs w'ithin (Pellegrino et al. 2017). Daumer (1958) used a similar approach, inverting the ray florets of Helianthus rigidus (Asteraceae) and other species with UV-absorbing centers, to which bees responded by probing for nectar at the flower’s periphery rather than its center (Figure 16.lb—d).

Clements and Long (1923) decoupled the visual and olfactory displays of hawkmoth-pollinated flowers (Oenothera cespitosa; Onagraceae), using colored crepe paper to conceal these fragrant flowers, whereas Knoll (1926a) sandwiched scented flowers of Lonicera implexa (Caprifoliaceae) between glass, compelling hawkmoths to choose between visual targets and a displaced scent plume. Knoll (1926b) extended this approach in his study of Arum nigrum (Araceae), a brood site- deceptive plant that traps its pollinators within a kettle-like inflorescence using fecal scent and heat. Knoll experimentally decoupled scent and heat by adding a scented (dissected) spadix to heated or unheated glass model spathes to construct floral “chimeras”. In the century since Knoll’s work, the “model spathe” approach has been adopted by botanists world-wide to decouple visual, olfactory and thermal floral stimuli from Arum family plants in behavioral assays with their pollinators (Patt et al. 1995; Miyake and Yafuso 2003).

Finally, the simplest (but not always straightforward) way to experimentally modify floral scent is to add single compounds or reconstituted blends to living flowers, typically using emitter devices or septa that control emission rates (Dobson et al. 2005). Manning (1956) pioneered this approach by adding essential oils to blooming and nonblooming Cynoglossum officinale (Boraginaceae) plants, to test their impact on bumblebee attention and foraging decisions. The novel odorants had no effect on approaches but reduced landings by half when flowers were present, either masking or disrupting the odor-color stimulus learned previously by the bees. Subsequent studies have established

Early efforts at experimental floral manipulation,

FIGURE 16.1 Early efforts at experimental floral manipulation, (a) Floral “mutilation” or targeted dissection of Aconitum and Delphinium flower parts (Ranunculaceae) by Clements and Long (1923), experimentally modifying floral symmetry, display size and reward accessibility in contrast to unmodified controls (left; drawings by Edith Clements). (From Clements, F.E. and F.L. Long. Experimental Pollination: an Outline of the Ecology of Flowers and Insects. Carnegie Institution of Washington, Washington, DC, 1923.) (b-d) Studies of ultraviolet (UV) absorbance patterns by Daumer (1958), showing differences in human visible vs. UV wavelengths for Helianthus rigida (b), a schematic drawing of the experimental inversion of ray florets to relocate the UV absorbance target (c), and an apparatus used to couple reconstructed floral displays with sugar rewards in bee behavioral choice assays (d). (From Daumer, K. Z. Vergl. Physiol., 41, 49-100, 1958. Image in panel (a) is reprinted with permission from the Carnegie Institute of Science. Images in panels b-d are reprinted with permission from Springer-Nature, license 4605651447443.)

dose-dependent impacts of floral scent on the behavior of pollinators and natural enemies (Galen et al. 2011), highlighting the importance of measuring and matching natural emissions.

These foundational studies encompass three categories of experimental manipulation: (1) adding scent to living flowers (floral augmentation), (2) decoupling or subtracting specific cues from living flowers (floral deconstruction), and (3) reconstituting the essence of floral display (floral reconstruction) (Raguso 2006). Below, I review progress made using each of these three approaches, emphasizing more recent work (along with the specific hypotheses that were tested in each case) and highlighting emerging tools, ideas and technologies whenever possible.

 
<<   CONTENTS   >>

Related topics