Thursday, October 20, 2005
Sunday, October 16, 2005
Herbivory - hypotheses...

Our herbivory results are above. I think we have made it sufficiently clear that the work has some limitations - limited number of samples, crude estimates (although I bet that applied to a sufficient number of samples it would yield meaningful results), insuffient area etc. However, since we were not, I am guessing, willing to spend the next three years on this (the length of a typical science project in this field) it was designed to give us a back of the envelope calculation of a phenomenon. Very often this is the first step in the development of a longterm study.
The results permit a certain amound of exploratory data analysis. Additionally, it allows us to generate a series of hypotheses. For instance on the basis of these results we might want to test the hypotheses that:
there is significant species to species variation in herbivory in the urban forest
species that are closely related (the two maples, for example) are likely to have similar defences against herbivory.
Since these numbers seem low, perhaps herbivory in semi-natural forest is greater (where there is a greater variety of potential herbivores
....
Can you generate additional hypotheses?
The point of these hypotheses is that (assuming a sufficient interest and sufficient resources) one would develop the procedures, design the sampling protocol, and exectute an experiment to see if these hypotheses stand up to scrutiny. That is the first step of most research is an attempt to quantify a pattern - then one evaluates how to test hypotheses relating to the pattern. Ultimately scientists like understand the mechanisms that generate the hypotheses... often the most difficult step. The execution of such a project may be a challenge - the pleasure tends to be in the ideas and in the interpretation of results. Often the simplest of results can revolutionize the world (the discovery of the structure of DNA was inventive chemistry - but the implications were revolutionizing!)
L
