Effect size in multiple groups comparison

LEfSe can be used to analyze multiple groups comparison. I was wondering how to calculate the effect size (LDA score) in the case with multiple groups.
In lefse.py line 219, the effect size always is estimated from the first and second classes.

effect.size ← abs(mean(LD[sub_d[,“class”]==“‘+p[0]+’”]) - mean(LD[sub_d[,“class”]==“‘+p[1]+’”]))

If the LDA score was calculated only using first and second classes, how to interpret the log LDA scores plotted in the class with the highest mean.