Is climate change research on marine mammals scientific?
Although it is beyond dispute that marine mammals respond to the physical effects upon habitat suitability, it is usually not clear what the relationship between a particular effect and the response from the marine mammal will be. Where data from time-series are analysed, as in the case of Forcada et al., (2005), they are used to test post-hoc for relationships between climate and biological variables. There is a tendency in these circumstances to test for all possible relationships using a range of physical and biological variables. Such post-hoc testing is fraught with pitfalls because invariably the final apparently statistically significant relationships are not down-weighted in their significance by all the other non-significant relationships that were investigated alongside those that proved to be statistically significant. Of course, there may be a priori reasons for accepting that a particular relationship is true, but the approach to examining time-series rarely provides an an
