Comparison of analytical methods for the determination of condensed tannins in grape skin

Background and Aims: Condensed tannins extracted from grapes are an important organoleptic component of wine. Assessing tannin level in grapes requires an effective analytical method. This study compared the efficacy of three analytical methods. Methods and Results: Condensed tannin levels were measured in 38 grape skin samples by protein and methylcellulose precipitation and by high-performance liquid chromatography (HPLC)-phloroglucinolysis. Total phenolics were determined by absorbance at 280 nm. Tannin levels by protein and methylcellulose precipitation were of similar magnitude, while levels determined by HPLC-phloroglucinolysis were generally higher. No correlation was found between the level of phenolics and tannins by any of the methods employed. Neither was there a correlation between tannin levels determined by HPLC-phloroglucinolysis and by precipitation, although previously a strong correlation had been reported between protein precipitation and HPLC-phloroglucinolysis in wine. Conclusions: The ratio of tannin to non-tannin phenolics is highly variable between cultivars. Each analytical method measures a different fraction of the tannin extracted from grape skin. Significance of the Study: The disparity between analytical methods requires that method selection reflect the decision-making that will utilise the information produced. For example, protein precipitation of tannins by bovine serum albumin (BSA) has been correlated with perceived astringency; consequently this may be the best indicator of potential mouthfeel.