Polyphenols and Variety
admin | January 18, 2011I was recently asked to speak at the Australian Olive Association National Conference on what makes good show oils ‘tick’. The Australian EVOO show system is unique in that most shows ‘measure’ all of the exhibited oils in some way to either ensure eligibility or to allow the oils to be objectively allocated to classes based on their style. As a result of this, a significant body of data on the chemistry of Australian olive oils has been collected. Specifically the free fatty acidity (which is an indicator of the quality of the olives and how quickly and carefully the olives were turned into oil), and polyphenol level (which is a good indicator of how bitter and peppery the oil is –aka style) have been collected by a number of shows over 5 or more years. As Chairman of these shows I get sent much of these data by the organisers at the conclusion of the show in a format that allowed me to get them together fairly easily. So while puting the seminar data together, along the way I collated data of the polyphenol levels of varietal oils. 2348 of them!
I selected 3 long running EVOO shows in Australia. These were 1) Royal Perth – a show that whilst being dominated by exhibits from Western Australia is well supported by boutique producers and has a wide variety of cultivars represented (typically 150+ entries). 2) Royal Canberra – much smaller, but is almost exclusively supported by small growers from cooler regions (80 odd entries) and 3) The Australian National – the biggest show with the most diverse range of exhibitors – micro volume through to millions of litres, but most importantly the oils exhibited here are from all states of Australia and therefore represent all climates – from cold to hot and even sub-tropical! (more than 200 entries per year).
I collated the data from the 2005-2010 Royal Perth and Australian National, and from the 2006-2010 Royal Canberra EVOO shows. These represent over 2,300 EVOO’s (ok – a few were surely exhibited in more than one of these shows, but it doesn’t happen that often – my guess is that <5% would fall into this category). So it’s a pretty good representation of what is/was being made.
Based on the information given by the exhibitor regarding whether the oil was varietal or a blend, I calculated the median polyphenol level by variety. To be included in the list, the oil must have been 100% from of a single variety and that variety must have appeared more than 30 times in these shows. Under this criterion Coratina just snuck in, as did Arbequina, but all of the others were very well represented – being major varieties in Australia (or in the case of WA Mission, Western Australia).
Here are the median (middle value) of polyphenols by variety. Note that in Australia (following the lead from the scientific community involved in measuring polyphenols), the results are given in “caffeic acid equivalents”. I won’t go into what that means specifically, but think of it as a unit of measure like a centimeter. If you see someone else quote a polyphenol in gallic acid equivalents or tyrosol equivalents then its like the same thing being measured in inches or pre-roman Calligulas or something. In short, oils measured in the same units can be compared, but you can’t compare an oil measured in tyrosol equivalents (which seems common in Italy), with another compared in gallic acid or caffeic acid equivalents.
Anyway – here are the results……
A few surprises. Manzanillo – wow. Considered as a table olive it does make oils that pack a punch. The much maligned Barnea – pretty well only grown in Australia and Israel showed that it isn’t meek and mild (not that it is a surprise to those who grow it, but perhaps to those who know nothing about it). Picual – I would have thought that it would be higher than low to mid range. Having said that, it is typically grown in Australia in warm to hot climates where it is irrigated – Conditions not conducive to high polyphenol production. The Italian mob – Corregiolla, Frantoio and WA Mission are closely grouped.
Coratina is lower than expected but it is a relatively new variety and as such is not strongly represented sample wise amongst the other varieties. If I were to repeat this in 5 years, I’d be surprised if Coratina didn’t move amongst the top 2 varieties in polyphenols.
Hi Richard-
I am a long time reader, but first time commenting. Thank you for going though all the trouble to make the nice graph. It’s really interesting to see how much the level of polyphenol varies (didn’t expect Manzanillo to have such a high number).
It would be interesting to do similar polyphenol studies for olives grown in California…wonder how will it will correlate due to the difference in geographic and climate.
Hi Selina
I’m happy that you have found the blog useful. Yes it would be interesting to see some polyphenol levels on oils from other parts of the world including California. Unfortunately the traditional analysis is rather expensive (around $US80) as the lab test requires a lot of manual handling. It can be done by NIR cutting the cost by 3/4 but that still requires that hundreds of oils be analysed in the traditional way before a good calibration can be made. So it might be a while before we see a lot of data on polyphenols.
It’s also worth noting that relating polyphenols to climate is harder than you might think as the level depends a lot on the ripeness of the fruit at picking, the type of processing and to a lesser extent on malaxation conditions. They also drop quite quickly with age. So for these reasons if you were doing a survey (like in the blogpost) you need a hell of a lot of examples to get an ‘average’ picture of varietal differences.
Richard G.
Hi Richard and Selina,
First: Richard, I too would like to thank you very much for these data; I’d also like to see the SD, or a ballpark eyeballing if the math is too much hassle even for a statistician 😉 . And, indeed, I’d love to see you put these data to full scrutiny, analyzing all the data you get (must include FFA, PV, polys — what else?) by varietal, geographic origin, and any other way you can, and publishing it formally. (You’re a victim of the high expectations you generate by example, sir 😉 ).
Selina: Rod Mailer did an excellent survey of the variation of Australian olive oil characteristics by geographic area, which at the least does track average temperatures: amongst other things, it confirms the show findings here reported by Richard on varietals x polys, and also that there are clear trends for cooler areas to produce higher oleic acid and poly levels. They also used these raw data to generate a proper peer-reviewed report, which broke this all down more elegantly; I can send you a PDF if you’d like and will post me a bot-proof email address. Also, Richard did a good blog post about this report with additional analysis, but for the life of me I can’t find it … Richard …?
Hi Michael
The data is pretty secondary in that the different shows (being independent) ask different questions of exhibitors on the entry form. For example, some only ask a business address (for billing purposes) which may or not be the grove address (many olive grove owners in Australia live and work in the cities), while others ask both. The only thing that has been consistently requested are details about what variety(ies) the oil was made from. The acidity and polyphenol levels were determined after a sample of every oil was send to a single lab for analysis by the organisers.
So cross tabulation with other variables is not possible.
Richard.