Flight-calling in Wood-Warblers (Parulidae): Does Species-Specific Behavior Drive Evolution of Calls?

Methods

Summary

To investigate these hypotheses, we will conduct fieldwork at two study sites along the diurnal migration of warblers in Michigan’s Keweenaw Peninsula. At both the west end of Manitou Island and Bete Grise we will conduct stationary counts of all westbound warblers flying over these locations. Preliminary data have shown that migrant flights are very concentrated at these two locations making it possible to tally the majority 3of individuals from a single location{Weidner 1992}. We will use observer-tallied counts to assess the magnitude of migration at both locales. To more directly analyze the behavior and composition of warbler flocks we will use audio recording of flight calls during morning flights at both study sites. We will record flight calls by placing SMM-A2 cabled acoustic microphones at each study site, each of which will be attached to a remote, passive SongMeter SM4 recorder at a sampling rate of 24 khz (Wildlife Acoustics) {Smith et al. 2014}. Recording will be programmed to start at sunrise and continue to 1300 hrs, which preliminary data have shown to be a time when cessation of diurnal flights occur. To our knowledge this is the first study to propose to directly investigate the role of flight calls in flock composition using both aural and visual evidence simultaneously. Many authors have proposed functions for flight calling behavior, but traditional studies using audio recordings of nocturnal migrations provide an indirect line of evidence to answer these questions. Audio recording will be collected from a 30-day window capturing peak migration periods in both spring and fall  2017-2019. Recordings will be downloaded and all species identities and individuals will be assigned using the SPCC algorithm to assign similarilty values {Keen 2016ba}to flight calls in RavenPro software. To test hypothesis 1, we will synchronize audio recordings with video recordings taken with a Canon XF300 camera for the same daily diurnal migration period during a subset capturing the largest diurnal flights. From this, we will extract the number of individuals in each flock of migrants, the number of conspecific individuals in each flock, and the approximate distances and positions of each calling bird using video frame analysis. We will use linear mixed models to explore the influence of flock position, number of adjacent conspecific individuals, and distance to flock individuals on flight call emission and calling rate. This will allow us to address whether flight calling behavior varies with distance and position within migrant flocks and whether such calls function primarily in interspecific or intraspecific communication. From these data we will also look directly at species-specific biases in calling behavior and reliance on flight calls for flock unity. We will contrast data from the same time period at Bete Grise and Manitou Island study sites to determine if flight calling behavior varies with geographic conditions and is more prevalent in high risk conditions, such as to maintain flock unity before water crossings.

                  To test hypothesis three we will use the current wood-warbler phylogeny {Lovette et al. 2010} to  look for phylogenetic signal in flight calling behavior. We will calculate independent contrasts for flight call frequency and sound attenuation for each species using CAIC {Freckleton et al. 2006} to see if significant evolutionary patterns exist in flight calling behavior. We will also map indices for species-specific differences in flock organization onto the phylogeny, which will be compared to flight call spectral characteristics using least squares regression.