I didn’t know that most GBIF data is at sea. But it’s on my todo list for the competition and I’ll let you know if it’s useful!
I anticipate something like using a GBIF data to create a species distribution model. If you interpret that model as habitat suitability, then the habitat suitability nearby the colonies might be a useful predictor. But who knows.
In case others read this thread:
3.5k locations where chinstrap penguins have been found.
The data are biased, messy etc. etc. but might be interesting.