Hi, We have a list of relative use of each observed plant species (in percentage) in each season by a primate species, and I'm try to apply the Kruscal-Wallis test to test whether the seasonal pattern is statistically significant. In each season there are about 30-40 plant species (about half) were not observed to be used (basically 0%). My question is, do I keep those ZEROs in the test or simply ignore them and use only those percentages with a positive number?
Thanks a lot in advance. WL
