Issue link: http://winesandvines.uberflip.com/i/70670
WINEMA k ING spectrum technical prowess, political savvy and people skills of David Block. I was additionally pleased to learn that he could take in stride a little postmodern criticism from my corner re- garding his presentation on the sensory effects of filtration. Together with Roger Boulton, arguably enology's top chemi- cal engineer, and sensory ace Hildegarde Heymann, Block con- ducted a red wine filtration study to look into the widespread anecdotal claim (I count myself among these rumor mongers) that filtration can affect wine's flavor and aroma, and strip the wine of its true character. Block expressed skepticism based on the fact that the pore size of a wine filtration is many orders of magnitude larger than the functional size of any volatile com- pound, so it can't be that the filters are blocking passage of volatiles. Other possibilities he envisioned include adsorption by filter media and oxygen pick-up during pumping. To prevent the later, a 2009 Oakville Merlot test wine was pushed with N2 gas through two types of 0.45 μ filters (PVDF and PES compo- sition.) An empty filter housing was used as a control. Wines were collected throughout the run and compared against an unprocessed control, rated for a spectrum of sensory character- istics at zero, one, two, four, six, 12 and 24 weeks. While significant variations in analytical anthocyanins and tan- nins were seen, they did not occur in a pattern one would as- sociate with adsorption. What the study did detect, and perhaps for the first time, was evidence of bottle shock (i.e. a temporary non-significant diminution in cherry aroma was noted that also occurred in the empty housing wine). In a subsequent study, a significant temporary increase in astringency was observed that abated after 16 weeks. Filtration of a 2010 Muscat/Chardonnay blend "did not have an effect." My points to Dr. Block are these: 1. Failing to show an effect is not the same thing as demonstrat- ing that there isn't one. In statistical parlance, the "significance" of an effect is measured by the probability of Type I Error (α): declar- ing an effect that isn't really there. This is quite a different thing from the probability of Type II Error (β), the chance of missing a real effect. The more cautious we are in preventing Type I Error, the greater the chance of Type II Error becomes. The only way out is to increase n, the number of sensory trials. In the case, for ex- ample, that a real difference exists could be detected in a duo-trio test 10% of the time (a 45:55 split), if we set our α (the chance of false alarm) at 10% (very liberal by conventional standards), then with n=40 trials, with normally distributed data it turns out that our chance of missing an effect is a whopping 90%. If we want both α and β to be 10%, our n=1,000 sensory trials. Since this is Tools Made to Take Out Big Weeds Big Savings on Weed Badger Save $500 when purchased with a new machine package Mower head Save $250 when purchased with a new machine package Cyclone Spade "Standard" Magnum Coil Teeth on Rotor When herbicide fails, or when weeds are too tough for a touch up machine, get search and rescue tools that work! Call factory-direct for pricing and availability. 800-437-3392 • www.weedbadger.com 56 Wines & Vines JULY 2012 David Block heaped praise upon the new Jess S. Jackson Sustainable far too expensive, we generally can make no claim to the quality of our evidence regarding missing an effect. In all fairness, the study was indeed sensitive enough to show a bottle shock effect. 2. Much more to the postmodern point, Block's skepticism has not taken into account the possibility that a filter may disrupt tan- nin colloidal structure, which many of us believe to offer a home for volatiles. In my own experience, ultrafiltration of red wines using a 200,000 molecular weight porosity (about a fifth the pore size of a 0.45 micron filter) shows dramatic loss of color in the filtrate, which is higher in veggie, oaky and microbial aromas. Sue Ebeler was the first person to tell me, a decade ago, of differences in pyrazine volatility due to macromolecular structure, and any- one conversant with micro-oxygenation has seen this effect when tannins are refined. The disruption of this structure by filtration, similar to the collapse of the emulsion when one attempts to ster- ile filter milk, seems worth considering. If there is anything to this effect, it is more to be expected in reds than whites. I encourage Dr. Block and colleagues to contin- ue these investigations, and applaud the industry and enthusiasm they are bringing to the daunting task of compiling sufficient data to solidify a position based on the facts. In future experiments, it would also be wise not to crossflow filter the test wine prior to the trials. Living proof Of all the work currently undertaken at Davis, from the postmod- ern perspective I am perhaps most excited by the groundbreaking microbiology research of David Mills. A key tenet of postmodern thinking is microbial balance.5 In the same way Integrated Pest Management (IPM) has largely replaced broad spectrum chemical eradication practices in vineyards, per- haps fostering microbial competition is a viable alternative to dra- conian microbial control measures in the cellar. In an Oenodev experiment a decade ago, I saw flash-pasteurized reds inoculated with Brettanomyces soar past the standard panic point of 500 colony-forming units/ml all the way to an unheard-of 107 CFU/ ml, leading me to suspect that there might be a substantial unchar- acterized microbial ecology in wines that suppresses Brett—the vinous equivalent of dark matter. Perhaps Brett is essentially a hospital disease, an opportunistic pathogen caused by sanitation. Until recently, our picture of wine microbiology, both in species range and population counts, has been limited by our ability to culture organisms on petri dishes. Mills' talk pro- vided a window on the new technology of DNA/RNA-based