Univariate data analysis in Primer?

I have abundance data for a marine invertebrate from 50 sampling units at 8 sites over the course of several years. I would like to compare this data between years to see if density manipulation work made any significant changes in its abundance.

My data is not normal and can't be made normal with any transformations that I have tried. I have performed a simple Kruskall-Wallis test between the sampling dates and did not find a difference but wondered if there was another way to test for a difference.

I understand that Primer does not make any assumptions about data but have always considered it useful only for community data- Can anyone tell me if an ANOSIM routine be appropriate in this case where I only have one variable? I have run the routine but wasn't sure if my similarity matrix had any meaning when it was only based on data from 1 variable....
Any insight would be appreciated. Cheers.