Can we increase human safety by understanding the language of bears?

Methods

Summary

The research area is about 2 square miles of sedge meadows and tidal flats.  Within this area between 5 and 30 bears may be seen feeding and interacting.  Daily methodology includes observing bears and anticipating interactions, strategically positioning the team before an interaction occurs in order to record each action and reaction as the interaction unfolds.  As recording commences, the video will be given a unique number that includes the exact date and time of each recording.  Cameras are also 'time-synced' to ensure simultaneous playback during analysis.  Other variables that may influence the interaction are also recorded via other methods.  These include: use of a weather device for wind speed and direction, barometric pressure, and temperature, visibility factors such as fog, tall grass, availability of foods, demographic details of each bear (age, sex, etc).

Challenges

The greatest challenge of this project is time and project design.  This research has never been performed, therefore the design is a constantly evolving process of some trial and error guided by behavioral research methods and experience with bears.  Because data collection methods are in the field, based on naturally occurring interactions with no controls or manipulation, the time investment to obtain enough data to make conclusions is significant.  As such it requires a commitment of years. 

Pre Analysis Plan

During post-season analysis videos will be reviewed multiple times in succession, each time focusing on the presence or absence of a particular body posture or vocalization in one subject bear (i.e. Bear 1) as well as the same for the reaction of Bear 2.  An ethogram will be developed that categorizes and describes each posture or vocalization, when and how it is used, and what reaction it prompts.  Analysis of vocalizations will include use of 'Ravenlite' software to visualize the pitch and frequency.  

Without placing meaning on actions or reactions, we will attempt to uncover patterns in the usage of each posture, vocalization, or combination.  Ultimately this will allow us to draw conclusions on their usage.