For those
variables that showed significant interactions between salinity and species, separate one-way
ANOVAs were performed on data taken from low and medium salinity treatments (since
interactions may confuse the interpretation of the effects of the main factors).