Belize, Central America

Postdoctoral fellow / Laboratory Manager

Ph.D. Biology (Ecology, Evolution and Systematics Division) - Saint Louis University, St. Louis, 2003

M.S. (R) Biology (Cell and Molecular Division) - Saint Louis University, St. Louis, 1997

B.S. Biology & B.A. Chemistry - Saint Louis University, St. Louis, 1995

About Me:

I came to the Kellogg lab in September 2003, a few months before completing a Ph.D. program at Saint Louis University, St. Louis, Missouri. As the Laboratory Manager, I am responsible for training new graduate students in molecular systematics techniques and analyses, including DNA extraction, PCR, sequencing and phylogenetic analyses using parsimony and maximum likelihood. I am also in charge of ordering supplies for the laboratory and maintaining the accounts current.

Research Interests:

My research combines the disciplines of ecology, evolution and systematics. My main interest is the evolution of reproductive systems and what role these may play in adaptive radiations. The passionflowers (genus Passiflora) of the Caribbean have great potential for answering this general question. In my dissertation work I showed that among hummingbird-pollinated species of the genus extent of incompatilbility varies, as does the effective gender of individual flowers. Is the plant/pollinator interaction responsible for these differences or are there other factors at play (ecological, genetic constraints,etc.)? Apart from hummingbird-pollinated lineages, does this pattern also occur in other lineages, for example those that are bat-or wasp-pollinated? I can address these questions by continuing to use the Passiflora system. I will need a) further observations on pollinators, b) more extensive experimental work doing controlled pollinations, and c) a better resolved phylogeny on which to map reproductive characters for all the Caribbean passionflowers. With the combination of phylogenetic and experimental data, I will also be able to answer questions as to the relative plasticity of important pollination characters such as corona type, especially in threatened species.

Currently, I am also working on phylogenetic problems addressing morphological diversification in the economically important grass family.

Previous Research:

My dissertation research focused on the evolution of reproductive systems in passionflowers, genus Passiflora (Passifloraceae). The genus is comprised of approximately 520 species, mostly neotropical vines. The two most conspicuous structures found in every passionflower are the androgynophore, an elongated stalk on which all the reproductive parts of the flower (including the ovary!) sit, and the corona, which is usually composed of brightly coloured filaments arranged in one or more rows. In the 16th century monastic scholars saw aspects of the passion of Christ reflected in these and other floral parts. Hence the name “passionflowers”. The androgynophore was interpreted as representing the column of flagellation and the corona of filaments, the crown of thorns.

Variation in the colour, size, shape and fusion patterns of these floral parts and also the floral tube (hypanthium) are indicative of the different pollination systems found in the genus. Passionflowers are usually pollinated by insects (e. g. bees and wasps) and hummingbirds. Bee-pollinated passionflowers are usually pan-shaped, emit sweet odours and have brightly coloured coronas that can serve as insect guides (see P. rubra above on right and this image). Hummingbird passionflowers are usually tubular, scentless and have reduced coronas (see P. murucuja below on left and this image). Bat pollination has been reported for only four species that bloom at night and release unpleasant odours.

I studied the pollination, compatibility systems, and phylogenetic relationships of Caribbean passionflowers for my dissertation. Most of this reproductive biology work focused on a group of 11 Greater Antillean species believed to be closely related. Each species is usually endemic to a single island. Except for the green-flowered P. penduliflora (see below on left), these species have pink-red, tubular, scentless flowers that open during the day, characters typical of hummingbird-pollination. They can be divided into two groups: those with a reduced, divided uniseriate corona (image) and those in which the filaments of the corona have fused to form a tube. This corona tube in Greater Antillean species is unique among New World representatives of the genus (see image on left).

I conducted fieldwork in Jamaica and Puerto Rico to study in detail the reproductive ecology of species that were representative of the different floral types in the Greater Antillean group. Pollinators were observed and mist-netted in the field and pollen was removed from them. Experimental hand-pollinations were done for later analysis in the lab using fluorescence microscopy techniques (see image on right). My results indicated that all of the tubular pink-red flowers are hummingbird-pollinated (image). However, those with a divided uniseriate corona had a wider suite of visitors including different size classes of hummingbirds and passerines, while those with a fused tubular corona were only pollinated by large hummingbirds. Other hummingbirds appeared to be excluded by the tubular corona. Plants with divided corona flowers were found to be bisexual or polygamous and self-incompatible. Those with fused corona flowers were self-compatible and non-autogamous. The musty smelling green-flowered representative of the group was found to be nocturnal, self-incompatible and was pollinated by the Greater Antillean long-tongued bat, Monophyllus redmani

In addition to reproductive ecology studies I traveled throughout the Greater Antilles collecting leaf material for DNA extraction. I obtained sequence data for the internal transcribed spacer and an unidentified single-copy gene to create a phylogenetic hypothesis of relationships in the Greater Antillean group. I then used this phylogeny to interpret the evolution of pollination systems and corona type in the Greater antillean group. The results indicated that hummingbird-pollination in the Greater Antillean lineage evolved once from a Mesoamerican insect-pollinated ancestor while bat-pollination evolved independently from hummingbird-pollination. The lineage was divided into two sister clades. The first was comprised of a monophyletic uniseriate- divided-corona clade. The other was a paraphyletic fused-tubular-corona clade in which a divided-uniseriate-corona evolved independently. The uniseriate-divided-corona in the Greater Antillean lineage and its Mesoamerican ancestor were derived from a bi-seriate-divided-corona.

Selected Publications:

Kay, E. and P. Bernhardt. Hummingbird-pollination and self-compatibility in Passiflora tulae (Passifloraceae). (Submitted to Plant Systematics and Evolution).

Kay, E. 2003. Floral Evolutionary Ecology of Passiflora spp. (Passifloraceae): subgenera Murucuia, Pseudomurucuja and Astephia. Ph.D. dissertation, Saint Louis University.

Kay, E. and P. Bernhardt. 2002. Evolution of Vertebrate Pollination in Passiflora spp. of the Greater Antilles. Abstracts of the Tropical Biology Meeting, Panama City, Panama.

Kay, Elma E. 2001. Observations on the pollination of Passiflora penduliflora. Biotropica 33(4): 709-713.

Contact me if you have questions or comments.

All images and text copyright Elma Kay