Testing for resource use and selection by marine birds: a comment.
The method recently advocated by Haney and Solow (1992: J. Field Ornithol. 63:43- 52) for the analysis of resource use and selection does not recognize the underlying structure of the data, so is inappropriate. In particular, it ignores the unit-sum constraint, requires independence between observations, and does not allow for extra-binomial variability. Compositional analysis, which is tailored to the data structure, and provides power and flexibility in its application, is recommended instead.