Although many angiosperms are serviced by flying pollinators, reports of wings as pollen vectors are rare. Flame azalea (Rhododendron calendulaceum) is visited by diverse insects, yet previous observations suggested that only butterfly wings may transfer pollen to stigmas. We used an experimental approach to determine whether butterfly wings are the primary vehicle of pollination in flame azalea. Over two seasons of observations, only butterflies (Papilio glaucus and Speyeria cybele) contacted both anthers and stigmas, yet because of differences in wing-flapping behavior, P. glaucus transferred pollen most efficiently. In contrast, bee species specialized either on pollen or nectar but did not contact both anthers and stigmas. A field experiment revealed that flowers excluding butterflies experienced almost complete fruit failure, whereas fruit set in open flowers did not differ from those that were hand pollinated. Additionally, butterflies had 56-fold more azalea pollen on their wings than bodies, while azalea stigmas bore both pollen and wing scales. These results suggest that plants with many visitors contacting reproductive organs may still specialize on a single guild of visitors for pollination and that wing-borne pollen transfer is a key mode of flame azalea pollination.
Butterfly pollen loads
Pollen carried on wings and bodies of butterflies collected in the field near R. calendulaceum plants in June of 2011 and 2014 (data in separate tabs). Individual pollen grains counted for 2011 butterflies and half of 2014 butterflies before switching to categorical ranges for 2014 butterflies (data for both ranges and counts shown here, in separate tabs).
Butterfly pollen load data.xlsx
2011 wings v body data
Data from observations of _Papilio glaucus_ butterflies foraging on _Rhododendron calendulaceum_ flowers in the field. Showing contact between butterfly wings and an anther & stigma versus contact between butterfly's body and an anther & stigma. Observed contact = 1, absence of contact = 0. Note that these data should not be used to draw any conclusions about which visits had BOTH wings and bodies contact anthers and stigmas. Data collected June 2011.
Flower visitor observations
Observations of flower visitors on _Rhododendron calendulaceum_ and flower parts contacted; taken June 2014.
Papilio glaucus v. Speyeria cybele on flame azalea
Field observations of contact between a butterfly and anthers versus stigmas during visits to _Rhododendron calendulaceum_ flowers by the butterflies _Papilio glaucus_ and _Speyeria cybele_. Contact with flower part during a flower visit = 1, absence of contact with a flower part = 0.
Pollen & wing scale load on stigmas from naturally visited flowers
After observing a visit by a butterfly in the field, stigmas from R. calendulaceum flowers were clipped and attached to a microscope slide in the field. Data show presence/absence (=Y/N) of pollen and butterfly wing scales on stigmas as scored after examining stigmas under the microscope.
Pollen & Scale load on stigmas from naturally visited flowers.xlsx
Pollinator exclusion experiment data
Data from experimental exclusion of flower visitors to _Rhododendron calendulaceum_. Treatments applied 6/12/14 - 6/13/14 and fruit set scored 8/4/14-8/7/14. Treatments defined as follows: A = flowers open to all visitors; B = mesh + cage (all visitors excluded); C = cage (only large visitors, e.g. big butterflies, excluded); D = maximum fruit set (unopened flwrs covered with ziplock until open, then pollinated by hand with outcrossed pollen).
Stigma load from single butterfly visits on emasculated flowers
_Rhododendron calendulaceum_ twigs bearing budding inflorescences were cut in the field and kept in water in the lab. Flowers were emasculated upon anthesis, and emasculated inflorescences were attached to the end of a pole and presented to foraging _Papilio glaucus_ butterflies near flowering _R. calendulaceum_ plants in the field. Stigmas of emasculated flowers were removed and attached to a microscope slide in the field following a visit, and scored for the presence/absence (=Y/N) of pollen grains and butterfly wing scales upon inspection under the microscope.