Subscriber login Close [x]
remember me
You are not logged in.

The Australian Wine Research Institute refutes Paul White's 'Scientifically Speaking' feature in Harpers' Closures supplement in December 2005

Published:  18 January, 2007

Professor Sakkie Pretorius says:
In this letter I do not intend to enter into a detailed critical review of the article. Rather, this response addresses passages from the article that are factually incorrect, or which use incomplete AWRI data which, used in that way, are misleading, or which misrepresent the AWRI's position on various issues. Additionally, some areas that we regard as scientifically unsound for an article that purports to be a review of the science of wine closures are also addressed.

In general, the tone and content of the Scientifically Speaking article appears to be very negative and critical with regard to the AWRI. We find this curious, and suggest that it may stem from a lack of awareness on the part of the author of our integrated and demonstrably successful work in helping winemakers to avoid the potential problem of what we call 'post-bottling reduction'.

Post-bottling reduction relates to the appearance, or an increase in the intensity of 'reductive' aromas in wine at any time after bottling, and is a function of the composition of the wine at bottling. The AWRI began to develop and to actively disseminate its strategies for the avoidance of this potential problem through our various extension and industry development activities, from mid 2001.

We do not believe that any other organisation or individual can claim to have been as proactive as the AWRI in raising awareness of this issue, and the success of the strategies in Australia suggests that the science behind the hypotheses on which the strategies are based is sound. Our extension and industry development activities are complementary to our ongoing research projects which seek to elucidate greater understanding of the mechanisms and of the compounds responsible for post-bottling reduction, and it is apparent that nobody can claim to have a full understanding of the science involved.

We note here that there does not appear to be any marked difference between our position and that of Mr Alan Limmer regarding the currently accepted science behind what he calls SLO (sulfur-like odour), and as stated in Limmer (2005) "the chemistry is reasonably well known and readily available if you know where to look".

However, in his criticism of the AWRI the author of the Scientifically Speaking article appears to rely heavily on Mr Limmer's papers on this subject, and attributes a number of quotations that are critical of the AWRI to Mr Limmer.

Much of the discussion in the Scientifically Speaking article appears to be based on the false premise that a there is a high incidence of post-bottling reduction in wines sealed with screwcaps.

With Australian wines, where the AWRI has particular expertise, this is demonstrably not the case. By any measure of technology adoption the rate and extent of the uptake of screwcaps in Australia and New Zealand has been extraordinary, and we suggest that such uptake would not have occurred if there were systemic problems with the new technology.

It is surely obvious to all in the wine business that the overwhelming majority of wines closed with screwcaps do not show post-bottling reduction, and therefore we further suggest that claims to the contrary such as those contained in the Scientifically Speaking article or in Mr Limmer's papers, should be treated with skepticism.

Mr Limmer claims to have been the first to raise the issue of SLO or post-bottling reduction in wines sealed with screwcaps "long before any reports of its occurrence", in his Stonecroft Wines newsletter of February 2002. This statement is incorrect, but the author of the Scientifically Speaking article seems to have accepted this mistaken belief at face value.

This is particularly concerning given that, between June and October 2001, the AWRI published three substantial papers containing results from the original "closure trial" which discuss and in two cases present data on reduction in the wine sealed with screwcaps and other closures. Those papers were published in three leading wine industry publications which are widely distributed in Australia and New Zealand.

We therefore believe it is important that the inaccuracies and unfounded criticisms of the AWRI contained in the Scientifically Speaking article are corrected, and that readers of Harpers are made aware of the complementary roles of the AWRI and how they have been harnessed as an integrated strategy to avoid post-bottling reduction with demonstrable success, particularly in Australian wine.

We do not believe that the "Scientifically Speaking" title of the article is consistent with its content and approach, and the article does not represent a proper review of the science of wine bottle closures.

The Fore-closure introduction to the supplement refers to the 2004 New Zealand Screwcap Wine Seal Initiative International Screwcap Symposium.

The Scientifically Speaking article makes many inaccurate or incomplete references, and appears to overlook the content of a presentation made by the AWRI's Peter Godden at that symposium, and of the resulting published paper (Godden et al. 2005), in discussions which are critical of the AWRI's position on various issues. This is inconsistent with an article that purports to be a scientific review, and the effect is that our position on various matters is misrepresented. It is particularly concerning given that the author of the Scientifically Speaking article formed part of the audience at the Screwcap Symposium.

"The closure simply reacted with 'dirty' wine that was the result of incompetent winemaking" [Page 26, paragraph 3].

The AWRI has never blamed winemakers for post-bottling reduction, and the implication that the AWRI regards winemakers who have experienced this problem as "incompetent" is erroneous and is rejected by the AWRI and its staff, and I am sure it would also be rejected by the many winemakers in Australia and other parts of the world that have approached and worked proactively with the AWRI, in order to gain greater understanding to avoid or to overcome this potential problem.

Our position, which we believe is undeniable, remains that the propensity of a wine to develop 'reductive' aromas post-bottling is a function of the wine, and that post-bottling reduction is not the 'fault' of the closure but may be exacerbated by the closure if the wine has a propensity for such aromas to develop. A corollary is that the propensity of a wine to develop such aromas can be affected by the winemaking process, but that is not to blame the winemaker for post-bottling reduction.

The role of the AWRI is to work cooperatively with winemakers, and in the case of the incidence of post-bottling reduction of wines closed with low oxygen permeation closures the success of that work is demonstrable.

"For the past year I've been treating all screwcapped wines on my tasting bench to 'before and after' comparisons like this", [with reference to copper fining wines in tasting glasses using copper coins] "try adding a few pennies to a glass and see for yourself". [Page 27, column 2, paragraph 4].

In his Screwcap Symposium presentation, Peter Godden discussed data from one of our AWRI Advanced Wine Assessment Courses which indicates a higher incidence of reduction in wines sealed with cork compared to wines sealed with screwcaps. Two subsequent courses have provided similar data.

One problem in ascertaining the true incidence of post-bottling reduction is distinguishing this phenomenon from similar or identical aromas that may already have been present in wine at bottling. Notwithstanding, these data indicate that the approach of the author in treating only screwcapped wines with copper is unsound, and would appear to be markedly unscientific.

Additionally, it is possible to obtain what might be considered very high, and certainly quite variable concentrations of copper in wine by adding a single copper coin to different glasses, let alone "a few" coins, and the consequent degree of fining would be uncontrolled and variable.

Copper reacts with many wine components including many flavour and aroma compounds, and thus any wine subjected to this treatment would be expected to exhibit different sensory properties "before and after".

"One of the most spurious claims is that corks have a thousandfold variability rate [of oxygen ingress]based on misinterpretation of a MOCON study using dry corks as a model" [with reference to the Australian Closure Fund paper], [Page 28, column 1, paragraph 2].

Without entering into a discussion of the Australian Closure Fund paper here, Peter Godden's Screwcap Symposium presentation and Godden et al. 2005 each include data on the oxygen permeation of twelve Reference 2 corks from the AWRI's original closure trial which show a 1227-fold range. (Note that this is not the "MOCON study using dry corks as a model" referred to above.)

The author of the Scientifically Speaking article has overlooked the existence of these data in his discussion of the variability in oxygen permeability of cork, which appears to be a serious omission in an article that purports to include scientific discussion of the oxygen permeation of cork.

The author goes on to state that "a recent peer-reviewed Bordeaux study by Lopez [sic]clearly demonstrates that corks do not have anywhere near thousandfold variability" [Page 28, column 1, paragraph 3]. However, it is not possible to draw such a conclusion from the Lopez paper, the purpose of which is to report on the development of a method to measure oxygen permeation non-destructively in-bottle, and not to examine the variability of permeation that exists in cork.

We also point out that any implication in the Scientifically Speaking article that the variability of oxygen permeation that may exist in twelve 'first grade' corks is representative of cork in general is unsound. We also note that the authors of the paper rightly, following scientific protocol, declare that they received funding from a cork company for their study, and that the same company supplied all of the cork-based closures examined. However, these factors are not noted in the Scientifically Speaking article.

The author goes on to attribute a quotation to Alan Limmer stating that "The natural corks had about a twofold variation between grades, and close to twofold variation within grades" [Page 28, column 1, paragraph 3]. There seems to be some confusion here as the Lopez paper only reports on the testing of one grade of natural cork.

But the author attributes a further quotation to Mr Limmer, stating that the results of that study "are more in accordance with AWRI trial resultsthan the MOCON thousandfold variation" [Page 28, column 1, paragraph 3]. The AWRI data from the closure trial in fact showed a 1227-fold variation across the reference 2 cork sample.

The author goes on to attribute the following quotation to Mr Limmer, "Remember that the MOCON data for screwcap shows a fourfold variation" [Page 28, column 1, paragraph 4].

This passage appears to relate to our data-set for the oxygen permeability of screwcaps and other closures, which shows the 1227-fold range for the reference 2 corks, i.e. the same data-set that the author failed to mention in his earlier discussion. That he should now rely on this data-set in this passage is, in our opinion, a particularly selective and unscientific use of our data.

Further, the author apparently then relates the data from that data-set concerning the permeability of screwcaps to data for the permeation of corks from the Lopez study, while at the same time ignoring the data which shows 1227-fold variability in the permeation of the reference 2 corks. We regard this as scientifically unsound and misleading.

Some of our most serious concerns with the Scientifically Speaking article relate to the discussion of Tyson Stelzer's book, Taming the Screw9. The article states that the book "seriously exposes this group [including Peter Godden] as underqualified to comment on fundamentals of wine chemistry" [Page 28, column 2, paragraph 1].

The implicit assertion in the Scientifically Speaking article that the AWRI reviewed or was responsible for the final text of Taming the Screw is incorrect. Members of AWRI staff regularly review text for technical accuracy on behalf of writers and journalists, in the interests of scientifically accurate information on wine and wine science entering the public domain.

With regard to Taming the Screw, on 23 February 2005 Peter Godden received a copy of the 'second draft' manuscript of this publication by email, and was asked by Mr Stelzer to "cast his eye" over portions of the manuscript, and to "check that I have represented AWRI research accurately".

Consequently, Peter Godden and to a lesser extent other members of AWRI staff reviewed portions of that manuscript and the resulting comments were verbally relayed to Mr Stelzer by Peter Godden. The acknowledgement of Peter's input on page 291 of the book accurately reflects the manner in which the comments were relayed to Mr Stelzer.

The AWRI was not given or asked to review any further versions of the manuscript. It is apparent when comparing the comments made by AWRI staff on the version of the manuscript received on 23 February 2005 with the final version, that all of the errors that we are aware of in the final version did exist in the earlier version, and that they were all noted by AWRI staff members and were all relayed to Mr Stelzer. Therefore, the assertion that the nature of Peter Godden's involvement with Taming the Screw "seriously exposes [him] as underqualified to comment on fundamentals of wine chemistry" is utterly rejected.

"Whereas in previous publications the AWRI appears to have been at a loss to explain the high incidence of reduced characters in screwcap wines, it now appears to be heading in the direction of Limmer's previously published explanations. Unfortunately, the authors still seem not to have cracked the underlying chemistry yet, continuing to espouse the old red herring that reduced faultiness is a winemaking fault, not a closure issue" [Page 28, column 2, paragraph 3].

We have particularly serious concerns regarding this passage. Firstly, as previously discussed, the premise that there is high incidence of reduced characters in screwcapped wines is incorrect. Thus, the notion that the AWRI has been at a loss to explain anything is based on a false premise, and is therefore rejected.

Further, we reject any notion that we have changed our position on the issue of post-bottling reduction. Indeed, the suggestion is fallacious in that the passage is referring to Godden et al. (2005) which is the published version of Peter Godden's November 2004 Screwcap Symposium presentation, and thus pre-dates the December 2004 publication of the first of Mr Limmer's papers referred to in the Scientifically Speaking article.

We also point out that Godden et al. (2005) represents two conference presentations which did not seek or purport to explain the "underlying chemistry" of the subject. Indeed, the AWRI has never published a paper which seeks or purports to explain the "underlying chemistry" of post-bottling reduction.

We also reject the implication that the AWRI is at fault for not having "cracked the underlying chemistry", and would equally reject any assertion by others that the chemistry or compounds responsible for reduced characters in bottled wine have been fully elucidated.

"One can't help but sense a bias towards screwcap in this report" [Godden et al. 2005], [Page 29, column 1, paragraph 2].

We strongly refute the allegation of bias in Godden et al. (2005). The content of that paper is consistent with Peter Godden's Screwcap Symposium and Enoforum Congress presentations, (Piacenza, Italy March 2005), and comparative data for all the closures tested at the 63-month time point of our closure trial is presented, and is objectively discussed.

To conclude, we believe that Harpers has made a very useful contribution to the debate of a number of wine science subjects and we hope that once the issues raised by this letter have been appropriately dealt with, that the AWRI and Harpers can continue to work in the constructive manner which has characterised our relationship in the past.

Professor Sakkie Pretorius

Managing Director


1 Godden, P.; Francis, L.; Field, J.; Gishen, M.; Coulter, A.; Valente, P.; Hj, P.; Robinson, E. (2001) Wine bottle closures: physical characteristics and effect on composition and sensory properties of a Semillon wine, 1. Performance up to 20 months post-bottling. Australian Journal of Grape and Wine Research 7: 64-105.

2 Limmer, A. (2005) The chemistry of post-bottling sulfides in wine. Chemistry in New Zealand (Sept): 2-5.

3 Limmer, A. (2005) Do corks breathe? Or the origin of SLO. Australian and New Zealand Grapegrower and Winemaker (497a): 89-98.

4 Limmer, A, (2005) Do corks breathe? New Zealand WineGrower (Autumn): 38-41.

5 Godden P.; Lattey, K.; Francis, L.; Gishen, M.; Cowey, G.; Holdstock, M.; Robinson, E.; Waters, E.; Skouroumounis, G.; Sefton, M.; Capone, D.; Kwiatkowski, M.; Field, J.; Coulter, A.; D'Costa, N.; Bramley, B. Towards offering wine to the consumer in optimal condition - the wine, the closures and other packaging variables: a review of AWRI research examining the changes that occur in wine after bottling. Australian and New Zealand Wine Industry Journal 20 (5): 20-30.

6 Godden, P.W.; Francis, I. L.; Field, J.; Gishen, M.; Coulter, A.D.; Valente, P.; Hj, P.B.; Robinson, E.M.C. (2001) Results of an AWRI trial investigating the technical performance of various types of wine closure. III. Wine sensory properties up to 20 months post-bottling. Australian Grapegrower and Winemaker (453): 103-110.

7 Godden, P.W.; Francis, I. L.; Field, J.; Gishen, M.; Coulter, A. D.; Valente, P.; Hj, P.B Robinson, E. M. C. (2001) Wine bottle closures: sensory properties of a Semillon wine. Performance up to 20 months post-bottling. Australian and New Zealand Wine Industry Journal 16: 93-112.

8 Lopes P.; Saucier, C.; Glories, Y. (2005) Nondestructive colorimetric method to determine the oxygen diffusion rate through closures used in winemaking. Journal of Agricultural and Food Chemistry 53: 6967 - 6973.

9 Stelzer, T. (2005) Taming the Screw: a manual for winemaking with screw caps. Brisbane, Qld: Wine Press: 305p.

10 Godden, P.W.; Gishen, M. (2005) Trends in the composition of Australian wine. Australian and New Zealand Wine Industry Journal 20(5): 21-46.