Which Correlation is correct?

#1
Hi there, I'm having a bit of a problem with the last bit of analysis for my thesis.

I would like to compare wave exposure of sites along the U.S. west coast against sea urchin spine repair rings.

My hypothesis is that a site with high wave exposure causes lots of spine breakage. Whenever a spine is broken, a new ring forms, and they look like tree rings.

So basically, my theory is higher wave variation means more spine rings.

I have 11 sites, each with its own set of repair cycle data (75 points) and its own set of water level stdev's (~75).

Is there anyway to compare these? I am using JMP and my advisor suggested I just take the average of each data set per site and use Kendall's tau ranking multivariate analysis. Is there a test like KT that doesn't use ranks, but instead uses raw data? Should I use a different method of correlation? I would like to keep the sites separate so that I can rank them after the analysis has been completed.
 
#2
Another way to think of it is this: If I took a box/whisker plot of wave exp per site and I wanted to compare it to a box/whisker plot of spine repair rings per site. Does that make sense? I think the problem with KT is that it assumes values are paired across rows.