A snapshot of SO 2 concentations in Austalian ines Sulfu dioxide (SO 2 ) is a pesevative idely used in foods and beveages. In inemaking, SO 2 is a vey common additive, due to its antioxidant and antimicobial functions. Even in ines hee a conscious decision is made not to add SO 2, small amounts ill be pesent because it is natually poduced by yeast. Maximum quantities of total SO 2 ae egulated in Austalian ines and in many of Austalia s expot makets. Undestanding the levels of SO 2 found in Austalian ine is impotant fo monitoing egulatoy compliance and also gives an indication of cuent tends in inemaking pactices, especially hen compaed ith histoic data. To gain an undestanding of the cuent use of SO 2 in Austalian ines, data ee analysed fo moe than 3, ines fom the 216 and 217 vintages hich had been submitted to Wine Austalia fo expot appoval. As pat of the expot pocess each poduce povides the ph, fee SO 2 and total SO 2 levels (among othe analytes) of the ine to be expoted. These data, along ith the majo gape vaiety in each ine, ee povided to the AWRI in an anonymised fom by Wine Austalia fo analysis. In total, data fom 3,167 ines (2,119 eds and 1,48 hites) ee evieed ith 2,278 fom the 216 vintage and 889 fom the 217 vintage. When selecting the data, only vaieties that had moe than 24 enties in the dataset ee chosen. In the case of blends, the vaiety that made up moe than 5% of the blend as used to categoise the ine. Wines hee no single vaiety constituted moe than 5% ee not included in the dataset. The statistical analysis of the data is summaised in Table 1. Table 1. Statistical summay of SO 2 data fo 3,167 Austalian ines (216 and 217 vintages) submitted fo expot appoval. All SO 2 esults ae in mg/l. All ines mean SD n min 1 st Qu median 3 d Qu max Red White 216 217 fso 2 3.6 9.8 3167 27 32 37 126 tso 2 93.7 34.8 3167 72 92 117 233 fso 2 3.5 9.5 2119 27 32 36 89 tso 2 83.8 3.2 2119 66 82 11 219 fso 2 31. 1.4 148 26 32 37 126 tso 2 113.9 34.6 148 98 116 134 233 fso 2 3.4 9.6 2278 27 32 36 126 tso 2 94. 34.2 2278 73 92 115 233 fso 2 31.1 1.3 889 27 32 37 61 tso 2 93. 36.2 889 7 93 118 221 Octobe 218 Technical Revie No. 236 11
Fee SO 2 When SO 2 is measued in ine, it is usually measued in to foms: fee SO 2 and total SO 2. The fee potion is esponsible fo the antioxidant and antimicobial popeties in ine. Figue 1 shos the distibution of fee SO 2 esults, boken don into ed and hite ines. It is inteesting to note that thee ae no appaent diffeences in the distibution of SO 2 concentations beteen ed and hite ines. Due to inteactions ith colou components in ed ine the efficacy of fee SO 2 vaies beteen ed and hite ines, paticulaly in the amount needed to achieve equivalent antimicobial activity. This diffeence does not appea to be eflected in Austalian inemaking pactices, based on this dataset. It is possible this esult is due to a geate concen about management of oxidation in packaged ine athe than micobial spoilage. Anothe inteesting obsevation fom the distibution in Figue 1 is the small gouping of ines at the exteme left of the histogam fo both eds and hites. This gouping of lo fee SO 2 ines (<1 mg/l) most pobably epesents the inceased occuence of no pesevative added ines in the maketplace. It can be seen that this is still a vey small pat of the oveall population, ith the vast majoity of ines having fee SO 2 levels above 2 mg/l and moe than 95% having levels beteen 2 and 4 mg/l. Less than 1% of ines had fee SO 2 levels ove 5 mg/l, ith one ed ine (89 mg/l) and one hite ine (126 mg/l) being exteme outlies. This similaity beteen ed and hite ines is continued acoss vintages, as can be seen in Figue 2, hich shos no significant diffeences beteen vintages and equivalent distibutions fo ed and hite ines. These esults ae consistent ith the obsevations fom 1986 to 214 fo hite ines (Godden et al. 215) that shoed a geneal incease in fee SO 2 esults until 212 and then a elatively constant value aound 31 mg/l of fee SO 2. The distibution of 4 3 colou Mean StDev N 3.45 9.51 2119 3.99 1.39 148 Fequency 2 1. 17. 5 35. 52. 5 7. 87. 5 15. 122. 5 Fee SO 2 mg/l Figue 1. Distibution of fee SO 2 values (mg/l) fo ed and hite ines 12 Technical Revie No. 236 Octobe 218
esults fo hite ines as also simila, ith an aveage standad deviation fo the fee SO 2 fo the last thee yeas of that study being 9.7 mg/l compaed to 1.4 mg/l in this dataset. Fo ed ines the mean value found in this study (3.5 mg/l) as highe than the last thee yeas in the Godden et al. (215) study (27.1, 28.9 and 28 mg/l) but not by a lage amount, suggesting that the geneal tend of inceasing fee SO 2 levels in ed ines may also have peaked o plateaued. The standad deviations fo ecent values in the last thee yeas of the Godden et al. (215) study ee also vey close to those found in this dataset. Figue 3 shos the fee SO 2 concentations boken don by gape vaiety. Fo the majo vaieties elatively little vaiation is evident. Once again this pobably eflects inemake decision-making being focused on post-packaging stability as ell as a ecognition that at 8 7 6 Fee SO₂ mg/l 5 4 3 32 32 32 33 2 1 colou Label Vintage 216 217 Figue 2. Box plot of fee SO 2 levels (mg/l) fo 216 and 217 vintage ed and hite ines, ith outlies excluded Fee SO₂ mg/l 8 7 6 5 4 3 2 1 R-Cabenet Sauvignon R-Duif R-Genache R-Melot R-Montepulciano R-Nebbiolo R-Neo d'avola R-Petit Vedot R-Pinot Noi R-Ruby Cabenet R-Sangiovese R-Shiaz R-Tempanillo Vaiety Figue 3. Boxplot of the fee SO 2 values boken don by gape vaiety W-Chadonnay W-Fiano W-Pinot Gigio W-Pinot Gis W-Riesling W-Sauvignon Blanc W-Semillon W-Vementino W-Viognie Octobe 218 Technical Revie No. 236 13
these levels fee SO 2 has little o no impact on ine s sensoy popeties. The mean fee SO 2 concentations fo Pinot Noi and Genache (23.7 and 27.3 mg/l) ae significantly loe (based on a Tukey s means compaison test) than the majo ed vaieties Cabenet Sauvignon (32.9 mg/l), Shiaz (31.9 mg/l) and Melot (31.9 mg/l) hich ee not statistically diffeent fom each othe. In the hite ines no vaiety as found to be statistically diffeent fom all the othe hite vaieties. Total SO 2 Total SO 2 is a measue of all the foms of sulfu dioxide, bound and unbound, found in ine and as such is not as closely linked to the technical function of sulfu dioxide as fee SO 2. It is an impotant measue, hoeve, because in Austalia and most of Austalia s expot makets the maximum alloed concentation of sulfu dioxide in ine is egulated as total SO 2. Fo this eason, it is impotant to have a pictue of the levels found in Austalian ines. Table 1 pesents a summay of the statistical esults fom the suvey fo total SO 2 and Figue 4 shos the distibution of total SO 2 values fo ed and hite ines. As is appaent fom the histogam thee is a significant diffeence in the mean total SO 2 values fo ed (83.8 mg/l) and hite (113.9 mg/l) ine. The spead of esults, hoeve, is simila fo hites and eds (standad deviations of 3.2 and 34.6 mg/l espectively). The highe level of total SO 2 in hite ines is typical and is an atefact of hite ine poduction hee highe levels of acetaldehyde (hich is a stong binde of SO 2 ) lead to a geate accumulation of bound SO 2 ith consequently highe total SO 2 levels. The bound potion of the total SO 2 has elatively little impact on the antioxidant and antimicobial function of SO 2 in ine. 2 15 colou Mean StDev N 83.76 3.24 2119 113.9 34.59 148 Fequency 1 5 3 6 9 12 15 18 21 Total SO 2 mg/l Figue 4. Distibution of total SO 2 values (mg/l) fo ed and hite ines 14 Technical Revie No. 236 Octobe 218
The egulatoy limit fo total SO 2 in Austalia fo dy ines is 25 mg/l. All ines in the dataset fell belo this figue, hich is to be expected as they had all been submitted to the egulatoy authoity fo expot licence. The vast majoity of ines ee significantly belo this figue (99% of ed ines belo 16 mg/l and 99% of hite ines belo 19 mg/l) and ould easily meet the equiements of Austalia s majo expot makets. As ith fee SO 2, the diffeences in total SO 2 beteen vintage 216 and 217 ines ee small and poved to be statistically insignificant, as shon in Figue 5. When compaed to the histoical data fom Godden et al. (215), the hite ines continued the donads tend obseved fom 212 214 (median values 129, 123 and 12 mg/l) ith medians of 117 and 114.6 mg/l espectively. Inteestingly the compaison fo ed ine total SO 2 is not so staightfoad. In the Godden et al. (215) pape, total SO 2 median values had declined significantly fom 211 to 214 (84 to 67 mg/l) hile the data in this study shoed total SO 2 medians of 83 and 78 mg/l fo 216 and 217 espectively. It as noted by the authos of the ealie study, hoeve, that hile the tend can be consideed intenally consistent, the esults fo total SO 2 may have been skeed as the time each sample had been in bottle hen analysed as unknon and natual degadation of some samples may have loeed the oveall values. The data in this study came fom ines hich ee about to be expoted so it can be assumed that the data as less affected by natual degadation of sulfu dioxide duing bottle ageing. Since hite ines geneally spend much less time in bottle, the diffeence beteen the to studies is likely to be much less fo hite ines, giving moe consistent compaisons fo total SO 2. 25 2 Total SO₂ mg/l 15 1 83 117 78 114.6 5 colou Label Vintage 216 217 Figue 5. Box plot of total SO 2 (mg/l) levels fo 216 and 217 vintage ed and hite ines, ith outlies excluded Octobe 218 Technical Revie No. 236 15
25 Total SO₂ mg/l 2 15 1 5 R-Cabenet Sauvignon R-Duif R-Genache R-Melot R-Montepulciano R-Nebbiolo R-Neo d'avola R-Petit Vedot R-Pinot Noi R-Ruby Cabenet R-Sangiovese R-Shiaz R-Tempanillo Vaiety Figue 6. Boxplot of the total SO 2 values boken don by gape vaiety W-Chadonnay W-Fiano W-Pinot Gigio W-Pinot Gis W-Riesling W-Sauvignon Blanc W-Semillon W-Vementino W-Viognie The beakdon by vaiety of ed ine total SO 2 (Figue 6) shos that ith the exception of Pinot Noi, the diffeences in means fo the ed vaieties ee not statistically significant. Pinot Noi (ith the loest mean value of 7.9 mg/l) as statistically diffeent fom the othe majo ed vaieties. This may eflect concens of inemakes about potential bleaching of colou by sulfu dioxide in this vaiety leading to moe consevative additions. In hite ines thee as no statistically significant diffeence in means among the thee majo vaieties of Chadonnay, Semillon and Sauvignon Blanc, hich epesented moe than 65% of the hite ines in the dataset. Conclusion This dataset of vintage 216 and 217 ines suggests that ecent tends identified in the use of SO 2 in Austalian ine have continued. All ines examined in study this easily comply ith the egulatoy limits in Austalia s expot makets. Acknoledgements The data fo this study ee kindly supplied by Wine Austalia ith special thanks to Steve Guy, Rachel Tiggs and Belinda van Eyssen. This ok as suppoted by Austalia s gapegoes and inemakes though thei investment body, Wine Austalia, ith matching funds fom the Austalian Govenment. The AWRI is a membe of the Wine Innovation Cluste in Adelaide, South Austalia. Refeence Godden, P., Wilkes, E., Johnson, D. 215. Tends in the composition of Austalian ine 1984 214. Aust. J. Gape Wine Res. 21(S1): 741 753. Eic Wilkes, Goup Manage Commecial Sevices, eic.ilkes@ai.com.au 16 Technical Revie No. 236 Octobe 218