• Fitzsimmons, Lauren P.
  • Foote, Jennifer R.
  • Ratcliffe, Laurene M.
  • Mennill, Daniel J.


Our understanding of animal communication is expanding from a dyadic framework of one signaler and one receiver to a broader communication network model, yet empirical studies of communication networks are scarce. To investigate whether territorial males eavesdrop on interactions occurring outside of their territory boundaries and to quantify the neighborhood-level effects of song contests, we simulated diurnal dyadic countersinging exchanges in the undefended spaces between established territories of black-capped chickadees (Poecile atricapillus). In each of 10 neighborhoods, we used stereo playback to simulate interactions between 2 unknown rivals. We simulated 2 types of song contests that differed only in the relative timing and patterning of the songs of the contestants; aggressive treatments contained frequency matching and song overlapping, whereas submissive treatments contained neither matching nor overlapping. We used a 16-microphone acoustic location system to record males in the neighborhood surrounding the playback apparatus. Territorial chickadees responded more intensely to the aggressive treatments than the submissive treatments. Neighborhood song output (number of songs produced by all individuals in the recording area) was twice as high after aggressive playback than after submissive playback. Males with territories bordering the playback apparatus had higher song output than males who were more than one territory removed from the playback apparatus. We did not find an influence of male dominance rank on playback responses. Our results reveal that territorial male chickadees eavesdrop on and respond to interactions occurring outside of their territory boundaries.


Field methods

We conducted a 2-treatment playback experiment in each of 10 black-capped chickadee neighborhoods at the Queen's University Biological Station near Kingston, Ontario, Canada (44°34′N, 76°19′W). Playbacks were conducted in 2005 and 2006 between 30 April and 15 May, and 810 and 0945 h. At this time of year, females are fertile and male–male countersinging interactions are common (Mennill and Otter 2007). We banded adult chickadees in the winter of each year with aluminum Canadian Wildlife Services bands as well as unique combinations of colored leg bands (N = 149 individuals in 2005 and 236 individuals in 2006). We determined the winter dominance ranks of males by observing pairwise interactions at feeding stations (N = 2811 interactions in 2005 and 8423 interactions in 2006). A bird was scored as dominant if it supplanted or chased an opponent, resisted a supplanting attack by an opponent, elicited a submissive posture in an opponent, or fed while an opponent waited to approach a feeder (for details, see Ratcliffe et al. 2007). We classified “high-ranking males” as the top-ranking male in flocks with 2 or 3 males or the top 2 males in flocks with 4 or 5 males. We classified “low-ranking males” as the bottom-ranking male in flocks with 2 or 3 males or the bottom 2 males in flocks with 4 or 5 males. We classified “mid-ranking males” only in flocks with 3 or 5 males.

Our ALS consisted of an array of 16 omnidirectional microphones connected to a central computer by 2200 m of microphone cable. The microphones were housed in rain guards made of PVC tubing mounted on top of 3-m wooden poles. Microphone poles were elevated off the ground and attached to trees with bungee cords. Input from all microphones was digitized using a multichannel data acquisition card (National Instruments DAQ-6260) and stored as 16-channel digital sound files using Chickadee V1.9 recording software (J. Burt, Seattle, WA). This design was an extension of the 8-microphone system used by Mennill et al. (2006). Each 16-channel microphone array recorded an area of approximately 160 000 m2 and encompassed the territories of 7–10 male chickadees. Recorded neighborhoods consisted of birds familiar with one another from the previous winter, from either the same winter flock or the adjacent winter flocks.

In April of each year, as winter flocks began to break up and breeding pairs began to defend territories, we visited each pair every 2–4 days. We mapped breeding territories according to the method of Bibby et al. (1992), recording the movements and territorial interactions of each pair on a detailed map by using landscape features and grid flags as landmarks. We considered a pair's territory to be the maximum extent of space exclusively occupied by the pair after the period of winter flock breakup but before the female's fertile period.

Playback design

Our playback apparatus consisted of 2 loudspeakers (Sony SRS-77G) mounted on 1.8-m poles and separated from each other by 24 m. We positioned the 2 loudspeakers in undefended spaces between the established territories of resident males in order to simulate a countersinging interaction between birds attempting to insert themselves in an undefended space.

As part of a larger study of avian communication networks, we recorded each neighborhood of chickadees for 2 or 3 days and then moved the ALS into a different neighborhood. Playback sessions were conducted on the final day of recording in each neighborhood so that playback sessions were never run on successive days. We recorded each neighborhood for a control period prior to broadcasting the 2 playback treatments. Each neighborhood received one treatment where we broadcast a highly “aggressive” interaction between 2 simulated rivals and another treatment where we attempted to broadcast a more “submissive” interaction between 2 simulated rivals (Figure 1). Song overlapping and frequency matching are directed signals of aggression in naturally occurring chickadee song contests (Mennill and Otter 2007; Fitzsimmons et al. 2008). In our aggressive treatments, the songs of one simulated male consistently overlapped the songs of the other and the simulated males were frequency matched within 50 Hz. In the submissive treatments, by contrast, the songs of the 2 simulated males were broadcast at different frequencies (frequency difference: 496 ± 4 Hz) and their songs did not overlap in time.