I'm trying to run a MANOVA on shoreline habitat data. The data was collected in an odd way. For each reach of shoreline, data was collected for each habitat variable (shade, veg and IWM). Each variable was described as having a certain percentage of a particular category of the variable. For example, (not from my data) Shade (the habitat variable) at Reach 1 was comprised of 50% "No Shade (one of the categories)", 25% "6%-25% shade", and 25% "Greater than 75% shade." Therefore each of the three variables are described by amounts of each of the categories and the percentages always equal 100 when added together. So my question is, how do I run a MANOVA on this to determine if the reaches are statistically different in R? Thanks in advance for any advice! Reach No Shade 1-5% Shade 6-25% Shade 26-75%Shade >75% Shade <25% Veg 26-50% Veg 51-75%Veg >75% Veg No IWM 1-10% IMW 11-50%IWM >50% IWM MS01 0.5609 0.2631 0.0103 0.1658 0 0 0.0294 0 0.9706 0.6693 0.2281 0.1026 0 MS02 0.2916 0.5672 0.0623 0.0789 0 0.0222 0 0 0.9778 0.1312 0.4787 0.1011 0.289 MS03 0.477 0.0411 0.0609 0.184 0.2371 0.0411 0 0 0.9589 0.477 0 0.225 0.298 MS04 0.7158 0.2037 0 0.0805 0 0.2569 0 0 0.7431 0.4589 0.2037 0.3374 0 MS05 0.4938 0.5062 0 0 0 0 0 0 1 0.558 0.442 0 0 [[alternative HTML version deleted]]