I work with environmental measurements that represent ratios of things (i.e., flow: volume/time, concentrations: mass/volume) and not absolute values (e.g., weights, mass, length, number of individuals).
I understand that Compositional Data are those data that represent parts of a whole ---- like proportions and percentages that add up to a constant value like 1 or 100%, respectively ---- and that compositional data require special treatment when performing certain analysis like correlation (e.g., log-ratio transformation, CoDA).
Do measurements like 3 mg/L or 10 mg/kg (which would be equivalent to 3 and 10 ppm, respectively) represent compositional data? If so, should I be using CoDA for analyzing those data?