The overarching question that I am interested in is: what is the genetic basis of adaptation and speciation?

Co-advised by Prof. Toby Bradshaw, an evolutionary geneticist, and Prof. Tom Danniel, an insect neurobiologist, I study pollinator-driven floral adaptation in my PhD. Flowering plants have a bewildering diversity in floral features, such as shapes, colors, scents, and symmetry etc. It is suggested that plant-pollinator interaction is key to understanding the origination of this diversity. There are two important questions to ask: a, what genetic changes modify the design of plant to cope with selection pressure? b, what kind of selection pressure does pollinator poses to plant?

Project 1: Genetic mapping of flower feature variations in monkeyflower

I use a common wildflower Mimulus (monkeyflower) as a model system to understand the genetic basis for pollinator-driven adaptation. By using both forward genetics (QTL mapping, mutagenesis screen) and reverse genetics approach(comparative genomics, candidate gene method and stable transformation for functional verification), I try to map the genetic variations responsible for the floral difference (e.g. color and scent) among a group of closely related Mimulus species (shown below).


The phylogeny of Erythranthe section in Mimulus, adapted from Beardsley et al., 2003

Mimulus mutants

Mimulus lewisii flower has two anthocyanin patterns: pink region on the petal lobes and small red dots on the nectar guide. We found that the plant also has two sets of transcription factor genes to control the two patterns respectively. Mutagenesis and transgenesis can cleanly diminish or enhance one color patten, while not influencing the other pattern. Upper left: pelan, has dimished pink region; Upper right: negan, has diminished red dots; Lower left, roi, has enhanced pink region; Lower right, red tongue, has enhanced red dots

Project 2: Exploration of fitness landscape of flower shape variations

I employ a novel combination of 3D-printing, electronic sensing, and computer vision to quantify the fitness effects of flower morphology variations. I measure the fitness parameter for equation-generated artificial flowers under the visit of real pollinators (hawkmoth and hummingbird). The goal is to construct a fitness landscape over flower morphological space to understand 1, how the flower morphology influence plant’s and pollinator’s fitness, respectively and 2, how different groups of pollinators drive flower morphology divergence.

This is the top view and side view of some 3D printed artificial flowers, only varying corolla curvature.

If you are interested, please watch the following videos which documented how animal pollinators visit 3D printed flowers.

Video 1. A hawkmoth Manduca sexta visit a 3D-printed flower. A machine vision program identifies the position of the moth and marks it with a white dot. Once the moth enters the 12.5cm proximity of the center flower, the electric sensors will start to record the nectar level change and acceleration change. The nectar will be replenished automatically by a micro-injector refiller once the moth leave the circle (Note the change of nectar information on the upper left corner).

Video 2. High speed video of a hawkmoth’s visit. The video was played at a speed 2.5 times slower than real time. The right side panel shows the nectar level and the accelerometer signal (the small part on top of the flower, which functions as a proxy for real flowers’ reproductive parts).


Video 3. This video shows a hummingbird’s visit to 3D-printed flower equipped with electronic sensors.