Table 2. Distributions for each observed variable and duration conditioned on states.

Observed Variable	Searching	Fishing	Cruising
	generalized Pareto	generalized extreme value	Gaussian mixture
	uniform	wrapped Cauchy	Laplace-Gaussian mixture
	Kumaraswamy	uniform	loglogistic
	Beta	uniform	loglogistic
	Laplace	Gaussian mixture	Student’s t
	Laplace	Gumbel	Student’s t
Duration	generalized extreme value	lognormal	generalized extreme value

Notes: When Beta and Kumaraswamy distributions are used, data is transformed to scale from 0 to 1.