John, Don made a very good suggestion in his reply: you can calculate the mean number of days for a species to reach a particular phenological stage, thereby making "day" the response variable. My point is that it sometimes makes sense to treat ordinal data the same way you treat continuous data (although a lot of the time it does not) and I would not rule it out without at least considering it. The trick is to determine the "spacing" between categories. Is a class 2 one unit greater than a class 1 and a class 4 three units greater? If you take this approach, you need to make your rationale clear when you're reporting your results.
Cheers, Peter
