Tannin content in red wines is positively correlated with astringency perception and wine grade; however, tannin quantification is one of the main challenges. In this study, tannin content was quantified using three analytical methods in commercial red wines from Vitis vinifera and interspecific cold-hardy hybrids including Marquette, Frontenac, and Petite pearl cultivars. Protein (PP) and methylcellulose precipitation (MCP) methods were compared to a HPLC-DAD method, which is based on the interaction between tannins and a hydrophobic surface (RPC). Frontenac wines were the poorest in tannins and Cabernet sauvignon wines were the richest regardless of the method used. In cold-hardy red wines, the tannin content was higher in Marquette with high alcohol content, which suggested that the tannins were extracted from seeds rather than skins. The high limit of quantification of the PP method and the presence of anthocyanin di-glucosides in cold-hardy wines were parameters suggesting that protein and methylcellulose precipitation methods were neither suitable nor reliable for the quantification of tannins in cold-hardy red wines. The tannin content quantified by RPC was positively correlated to tannin quantified by MCP, suggesting that the RPC method would be relevant for the quantification of tannins in red wines.
Keywords: HPLC-PLRP; interspecific hybrid grapes; methylcellulose precipitation; proanthocyanidins; protein precipitation.