I have an NMDS analysis in which I used to visualize distances between macroinvertebrate community compositions within four treatments referring to acid mine drainage condition. The plot looks like this, with all treatments overlapping one another, but the area each treatment occupies getting smaller:
What I want to do is quantify the area (and therefore the variance) of each treatment. In other words, I want to measure the area of the polygons around each treatment. I have two questions regarding this:
1) Is it viable to measure the area of the polygon, given the non-metric nature of NMDS?
2) Is there a way to measure area in R, or should I use ImageJ?
I realize that this is probably a crude/rudimentary way to quantify variance of an NMDS, so I welcome any alternative methods as well.