I am studying a rare population of a herbaceous plant that has a pronounced clumped distribution pattern in their habitat (sloping woodland or semi-open vegetation). I read that an adaptive cluster sampling would be appropriate as a sampling design for a field survey to determine the microhabitat variables affecting its distribution and abundance. I am wondering whether the data requirements and assumptions in adopting ACS would be in congruence with analyses such as comparison of means and multivariate analyses employing ordination (such a non-metric multidimensional scaling or canonical correspondence analysis).