Research Interests
I am interested in behavioral ecology, specifically birdsong and aggressive communication.  My current dissertation research with swamp sparrows centers on one aspect of birdsong, vocal performance, and how this is used as a signal by males in intra- and intersexual interactions. 
Vocal performance is the ability to produce physically demanding songs, and likely functions in both courtship and territory defense.  Vocal performance is affected by morphological constraints which cause trade-offs between vocal components that cannot be maximized simultaneously.  In Emberizid sparrows, a trade-off has been demonstrated between trill rate (the rate of repetition of syllables) and frequency bandwidth (the range of frequencies encompassed by a syllable). This tradeoff between bandwidth and trill rate is thought to be a consequence of vocal tract morphology and mechanics.  The trade-off is characterized by a negative upper bound regression line relating trill rate to bandwidth, and vocal performance can be measured by a song’s deviation from the upper bound regression.


(left) A swamp sparrow, Melospiza georgiana

(right) Acoustic space for trilled songs (modified from Podos and Nowicki, 2004, in Nature’s Music).  Low performance songs are those with a large deviation from the upper bound for the population (orange), while high performance songs have a small deviation (yellow).

(above) Example of swamp sparrow song. Frequency bandwidth is the range of frequencies in a song (blue arrow). Trill rate is the rate at which syllables (black box) are repeated.
Dissertation Overview
I work with a population of swamp sparrows in Conneaut Marsh in western Pennsylvania, near the University of Pittsburgh’s Pymatuning Laboratory of Ecology.  There I collaborate with other birdsong researchers from the Nowicki Lab at Duke University.

My proposed dissertation research is comprised of the following four studies on vocal performance.

(1) Male modulation of vocal performance
If vocal performance is a signal of quality in aggressive interactions, then males should be under selection to maximize their vocal performance in aggressive contexts.  I used playback to simulate territorial intrusions and recorded response to test two mechanisms by which performance could be maximized:  males might choose out of the two to four song types in their individual repertoires the one with the highest vocal performance, or they might increase the performance level of whatever song they chose to sing.  Males were found to do the latter, by increasing both trill rate and frequency bandwidth. This result is the first to indicate flexible modulation of vocal performances in a signaling interaction and is consistent with this signal feature having a role in aggressive communication.

(2) Aggression and vocal performance
I am using acoustic analysis techniques to measure recordings from two previous studies on aggressive signaling in swamp sparrows.  I am interested in whether the vocal performance of males is related to their level of aggression. 

(3) Discrimination of vocal performance
Males have been shown to modulate their performance, but are these changes in performance meaningful?  Are other males able to discriminate the difference between a singer’s low and high performance versions of a song?  Are males able to discriminate between a low performance individual and a high performance individual?  I will address these questions in two playback experiments.

(4) Vocal performance in territorial defense
I hypothesize that vocal performance plays an important role in territory defense. I plan to simulate interactions between territory owners and intruders with manipulated levels of vocal performance to test a prediction my hypothesis, that simulated owners with higher vocal performance will incur fewer intrusions than those with lower vocal performance.