Burning question:

There's so much monitoring data already- camera trap archives, acoustic recordings, GPS tracks - but almost all of it was collected to answer presence/absence or population questions. What I'm after is the layer above that: behavior. Not "is the species there," but "what are they doing, and for how long."

I'm working toward behavioral time-budget analysis as an early signal of ecosystem stress, and I keep hitting the same wall: how do you actually pull structured behavioral data out of collections that were never designed to capture it?

What I'm trying to figure out:

• What's genuinely been automated here, vs. still manual annotation?
• Is anyone extracting behavior (not just detection) from camera trap or bioacoustic archives at any kind of scale?
• Are there datasets already behaviorally coded, or is everyone building that layer from scratch?

Pointers to people, papers, tools - all welcome. Curious what's worked and what's been a dead end.

Thanks!
Maggie