ResearcH Interests

Testing the Sensory Drive Hypothesis in the Context of Social Communication:

Although widely-discussed, sensory drive has received little empirical support, particularly in terrestrial systems. I am testing the predictions of the sensory drive hypothesis using West Indian Anolis lizards as a model system.

At the intraspecific level, I am testing the predictions of the sensory drive hypothesis using allopatric populations of the lizard A. cristatellus from two distinct environments, mesic and xeric. I found that populations from mesic and xeric conditions occupy two distinct habitats with respect to light intensity and spectral quality. There is no evidence for divergence between populations in visual system. However, populations have diverged in dewlap color in a way that appears adaptive with respect to the light environment: dewlap detectability decreases significantly when a dewlap from one type of habitat is seen under the spectral conditions of the other. These results strongly suggest that selection for signal efficacy has favored divergence in dewlap designs as populations adapt to different light environments, a finding that supports the sensory drive hypothesis. Furthermore, these findings bolster the possibility that reproductive isolation might evolve as a by-product of natural selection favoring adaptations of the dewlap design to local habitat light conditions. 

Detailed phylogenies are currently available for the major clades of West Indian anoles, making them an ideal model to test the prediction that signal design and spectral sensitivity will converge in species that independently occupy similar habitat light environments. Preliminary analysis from 13 species of anoles suggests that dewlap’s design has evolved due to selection for efficacy of communication. For example, species living in open habitat (i.e., high light conditions) exhibit dewlap designs that are easily detectable under high light intensity, whereas species from close canopy forest (i.e., low light conditions) exhibit dewlap designs that are easily detectable under low light levels.

Species Recognition and Signal Design

Theory predicts that signals used for species recognition should be highly reliable and detectable. Exactly what makes a signal highly “reliable” is unclear. Therefore, which signal designs are favored for the process of species recognition is an area of much debate. My work with anoles showed that signal detectability is achieved through the use of brightness and chromatic contrast.  Preliminary analysis of dewlap designs indicate that sister taxa tend to differ more in dewlap brightness than dewlap chroma. This result is somewhat surprising, because brightness is often considered an unreliable component of the signal. The next questions in this line of research are the following: When do animal rely on brightness versus chroma in assessing visual information? Is one more reliable than the other, or are both components complementary?

Signal Design, Habitat Conditions, and Predation Pressure

The current consensus among behavioral and evolutionary ecologist is that predation pressure and habitat conditions can impact the design of visual displays and that signals can rapidly change in response to both of these selective forces. However, empirical evidence for rapid evolution of signal displays in response to habitat conditions and/or predation pressure in vertebrates is limited to a handful of studies, all of them from aquatic systems. In collaboration with J. B. Losos and T. Schoener, last May I began a long-term experiment to elucidate the interactions between habitat conditions (e.g., vegetation profile and background noise) and predation pressure on the evolution of conspicuous visual displays by conducting a replicated field experiment in the wild, over a period of four years, using A. sagrei as a model system. Our experimental manipulations are designed to allow us to collect data across several generations during the duration of this experiment, thus providing the possibility to track changes in signal design over "evolutionary time."