I've noticed the same thing when using higher efficiency extraction methods. Kavain must be slightly more water soluble than the other kavalactones, thus if you increase the total amount of kavalactones extracted, the proportion of less water soluble chemicals will go up. Thus the chemotype will appear to change. I wonder if this is the reason why Deleted User gets different chemotypes than what some others report. If he is using organic solvents to extract the kavalactones prior to assaying them, then you would expect the chemotype to apeear with less kavain than if water was used as the solvent. It has been reported that flavokawain proportions also go up with higher extraction efficiencies as well.
"Extract chemical composition was strongly influenced by extraction solvent. Metabolic fingerprints from aqueous and ethanolic extracts plotted in principal component space formed two distinct groups driven by extraction solvent where the use of either 100% water or 95% ethanol was responsible for 71.1% of the variation among all samples explained by PC1 (
Figure S2). The detected ion
m/z, retention time pairs that contribute the most to the loadings for PC1 were 315.1132
m/z, 9.1823 min and 285.1021
m/z, 9.4699 min, which correspond to the masses and retention times of FLKA and FLKB, respectively. Compound quantification showed that extracts prepared with 95% ethanol resulted in higher yields and greater consistency among replicates, compared with extracts prepared with 100% water. This result is similar to previous studies that found water produced kava extracts with decreased compound concentrations compared to extracts prepared with ethanol. Specifically, K, DHK, M, and DHM concentrations were 1.5–5x higher in samples extracted with 95% ethanol than in those extracted with 100% water. The concentrations of FLKA & FLKB were up to fifty times higher in samples extracted by 95% ethanol than in those extracted with 100% water although a significant number of the water extracts contained concentrations of FLKA or FLKB that were below detectable limits (
Figure 1). Extracts prepared with 95% ethanol consistently contained greater quantities of FLKA and FLKB than corresponding water extracts, and were highly variable across kava products (ranging from undetectable concentrations up to 14.7 ppm"
Measuring the Chemical and Cytotoxic Variability of Commercially Available Kava (Piper methysticum G. Forster)