Tuesday, August 5, 2008 - 1:30 PM

OOS 11-1: Testing the hypothesis of mushroom mimicry for Dracula felix orchids

Bitty A. Roy1, Bryn T.M. Dentinger2, Tobias Policha1, Tommy Jenkinson3, and Rocio Manobanda4. (1) University of Oregon, (2) Royal Ontario Museum and University of Toronto, (3) San Francisco State University, (4) National Herbarium, Quito, Ecuador

Background/Question/Methods Orchids in the genus Dracula appear to mimic mushrooms, both in terms of visual and fragrance cues.  However, there is no published quantitative data on visitation rates or evidence of pollinia removal in the wild, and no one has attempted to identify putative models.  In short, for this to be mimicry, the resemblance must be adaptive.  We tested the mimicry hypothesis for D. felix by observing visitors and by doing experimental manipulations in a cloud forest on the western side of Ecuador.

Results/Conclusions Small fungus visiting flies were the most common visitors to the flowers and we observed pollinia on these flies.  The flies treated the flowers the same way they did mushrooms: they guarded them, and mated within the flowers.  Many species of mushrooms that were similar in shape and size to the flowers commonly co-occurred within centimeters of these orchids.  We performed a mushroom augmentation experiment to determine whether visitation was increased in the presence of mushrooms.  There was a significant treatment effect for fungus fly visitation (F5.292,14.3, P=0.0190); flies visited the treatments with mixtures of mushrooms and flowers more often than they did either mushrooms or flowers alone.  Furthermore, individual flies flew back and forth between mushrooms and flowers.  We performed a second experiment to determine how important fragrance and visual cues were for enabling the flies to find the flowers. The three treatments were: exposed flowers, flowers covered with a muslin bag that hid the flowers but allowed fragrance to escape, or the muslin bags alone (controls).  There was a significant treatment effect, F = 7.192,33, P=0.0479; significantly more fungus fly visited flowers than the bag controls, but there was not a significant difference in visitation between the unbagged and bagged flowers, suggesting that fragrance cues are critical.  Our data support the fungal mimicry hypothesis.  Pollination of plants such as these orchids that are rare and isolated and that flower when it is always raining, is by no means guaranteed.  Fungal mimicry appears to be a highly successful strategy for these orchids; they are extremely attractive to fungus visiting flies, are visited at high rates, the flies can fly long distances carrying pollinia, and we have some evidence of specificity.