Live from the Bird Cams Lab: Our First Q&A Session

By Miyoko Chu, June 19, 2018

On June 15, 17 participants joined our Bird Cams staff for our very first live Q&A session online. We connected through a link to a Zoom webinar, and because it was our first time using this tool, our team was both excited and nervous to see how it would turn out. At ten minutes to noon, we felt relieved to see our first participants coming on board and typing comments into the live chat. Soon we were well on our way for a full hour of lively discussion.

First, we introduced ourselves—Bird Cams leader Charles Eldermire, communications specialist Ben Walters, graduate student Rachael Mady, and director of Communications Miyoko Chu—and participants tuning in from across the country, from California, Texas, Florida, and Georgia, and North Carolina to Ohio, Wisconsin, Massachusetts, New York, and more. Participants tested our online polling tool by sharing how long they’ve been watching the cams—ranging from less than a year to more than four years.

To bring everyone up to date, we then shared an overview of the project and recent activities, now summarized in our recent blog posts, “Welcome to the Bird Cams Lab,” “Wonder Board Brainstorm: Your Questions, Summarized,” and “Seven Cam-Testable Questions You Identified in Sorting Activity.” Then we put up another poll so participants could choose a topic to delve into for the remainder of the hour. The majority picked vocalizations, but there was also a lot of interest in nestling and feeding behaviors.

To shape the discussion, we explained that we use four criteria to help decide if we can successfully investigate a question: whether it is measurable, specific, feasible in a reasonable amount of time, and interesting (both for yielding new scientific discoveries and for being personally motivating for participants to pursue).

We then thought about the example question, “Do the hawks use different kinds of calls in different situations at the nest?” As the Bird Cams staff provided commentary, participants brainstormed different kinds of vocal behaviors they’ve noticed when watching the cams and some considerations when collecting data:

LisaB : The chicks whistle when they see a parent flying toward the nest
David : Frequency and intensity of vocalization; Pre-fledge v post-fledge
Kymberlee: Frequency of calls/vocalization as correlated with certain behaviors. This could be assessed with longitudinal data as well.
SLESCH: consider whether a hawk is in distress, hungry, another predator is near, time for relief from brooding, etc?
Valerie C: There seem to be 4-5 basic vocalizations… chirps, warning calls, whistles etc. Correlate & log to behavior
FloridaGirl: agreed – prey delivered
K Williams: Vocalizations between chicks.
jean: Seems like we would need sound spectral analysis to actually make this measurable?
Kymberlee: Statistically significant -those vocals with behavior so there are objective indicators.
Sara: Could we grab a clip of a vocalization and make the audio available for everyone to hear – then we could all be hearing the same vocalization and then relate an action
Suzanne: that is a great idea sara
David C: Would we restrict “eligible” calls only to those we can see and hear vs only hearing them.
Valerie C: Maybe concentrate on BR since she would be in all the current and archived footage
Kymberlee: Interrater reliability can be assessed between viewer ratings
Sara: that’s a good idea about concentrating on BR
LisaB: The RTH screeing can be heard on cam but it never occurs at the nest
David: Most birds vocalize when anxious, excited or seeking a mate.
David: RTH’s seem to increase calls as fledging approaches.
treegal : dichotomous key between vocalization and actions?
K Williams: There is also vocalization between the mating pair when they hand off prey or trade manning the nest.
Kymberlee: Just looking at BR would be interesting, but then we may not get (perhaps) the differences between maternal and paternal influences/calls
Sara: yes, it would be able to be specific if we can create that ‘training archive’
Valerie C: Let’s identify the basic vocalizations and then focus on a specific call/behavior
David: We should use “Vocalization” as a major topic and build a set of questions under it. Use this as a pilot to create model.

Is the question measurable? Our staff explained that in order to delve into footage from the past season, we will break up the recorded video into 10-second clips which will be presented on the citizen-science platform for participants to help classify. The chat above shows how participants were able to generate a clearer idea of the kinds of vocal behaviors we could look for, and some ways to document them, including the possibility of using sound spectrograms to compare sounds visually. Overall, we did feel that investigating the topic was feasible, even if there might be some challenges such as associating certain vocalizations with factors that might not always be visible on the cam screen.

Is the question specific? Although the question is broad, participants noted the value of creating a catalog, both to understand the range of vocalizations for further study and to create a reference set that could be used when tagging footage.

Is the question feasible in a reasonable amount of time? We’re not sure how long it will take to tag a full season of data; we suppose that depends on how many volunteers will help us out! We could release a batch of footage to see how long it takes people to process. If it’s a very long time, we could consider analyzing footage from just part of the season.

Is the question interesting? Since this question won our poll for the day, we know that studying vocalizations would be motivating for at least some participants. Based on Rachael’s initial perusal of the scientific literature, documenting more details of vocalizations at the nest seems scientifically valuable as well.

In our last poll of the day, most participants indicated that today’s live session was very useful. Our staff felt the same way, and really appreciated the thoughts that everyone shared. We were also relieved that the webinar format seemed to work without major glitches, and we’re looking forward to improving a few things on our end for future sessions.

Now we’d love to continue the conversation in a new forum where we can go in-depth on each of the seven testable questions. Please join us as we lead up to the final vote where we’ll pick a question so we can start collecting data! Visit the new Question Design forum.

Suggestions or ideas about the live Q&A session? Please leave your comments below.