Random Ramblings of a Weary Winemaker: Microbes: The Good, The Bad, and The Ugly

by Peter Wellington

As I started to think about how I would organize this blog I came to the realization that there really should be a fourth category – the “maybe good, maybe bad, maybe innocuous”, with a large aside on the pros and cons of inoculating with pure yeast strains. 

The list of species that can grow in grape juice is quite small, and even fewer can grow in wine.  The low pH and high sugar and/or alcohol levels make for a very hostile environment.  No pathogenic or toxin producing microbes can grow or even survive in wine, which led Louis Pasteur to declare wine the most hygienic of all beverages.  Only two species, Saccharomyces cerevisiae (wine yeast) and Oenococcus oeni (malolactic bacteria), are widely accepted as “good” wine microbes.  Even malolactic bacteria are considered spoilage organisms if malolactic fermentation (MLF) isn't desired...

Saccharomyces cerevisiae is widely distributed in nature, and thousands of different strains have been identified.  It is responsible for most leavened bread and virtually all beer, wine and other alcoholic beverage fermentations.  Whether you add one of scores of available pure yeast strains or just let nature take its course, it is wine yeast that are responsible for alcoholic fermentation.  Commercially available wine yeast strains are not bioengineered “Frankenyeasts”, they are merely yeasts that have been isolated from successful fermentations.  Often they are named for their origin (e.g. Bordeaux Red, Montrachet).

Quite a few artisanal wineries and even some large producers do not use commercial yeast strains (or at least not intentionally).  Some confusion has been created by the terminology used to describe this practice.  We read or hear terms like wild yeast, native yeast, vineyard yeast, indigenous yeast, etc., with pure yeast strains referred to as commercial or industrial yeast.  “Wild yeast” refers to non-Saccharomyces genera, and really shouldn't be used to describe wine yeast of unknown origin.  Although wine yeasts have been found on grapes in vineyards, they don't seem to be abundant enough to be consistently responsible for wine fermentations; apparently the main source of incidental wine yeast inoculation is the winery itself.  Brand new facilities have experienced difficulty getting uninoculated grapes to ferment.  A few years ago, someone (Clark Smith, perhaps) published a brief discussion of how best to describe “natural” fermentations, and came to the conclusion that “not intentionally inoculated” was most accurate.

Oenococcus oeni (the bacterium formerly known as Leuconostoc oenos) is the preferred species for performing malolactic fermentation, a process that converts malic acid to lactic acid.  This reduces acidity (lactic acid is a weaker acid than malic), and, more importantly, stabilizes the wine against unwanted ml fermentation in bottle.  ML fermentation in bottle is undesirable because it also produces  CO2 and aromas reminiscent of cheese or sauerkraut. 

Wild yeasts such as Hansenula, Kloeckera, and non-cerevisiae species of Saccharomyces fall into the maybe good, maybe bad category.  In grape must that is not inoculated with a wine yeast culture (and particularly in unsulfited must), wild yeasts often initiate fermentation.  Almost all these species have low alcohol tolerance and are soon overwhelmed by wine yeast, which finish the fermentation.  All yeasts produce “secondary” metabolites other than alcohol and CO2.  Proponents of “natural” fermentation contend that the secondary metabolites produced by wild yeasts are important aspects of wine quality and regional character.  Others believe that the potential benefits of natural fermentation are minor, disappear during aging, and are outweighed by the risks.  The biggest risks of encouraging wild yeast are increased volatile acidity (VA – vinegar), stuck fermentations and unpredictability.  A supporter of the use of pure yeast strains once asked, “You wouldn't make wine from randomly harvested grape varieties, so why would you use randomly selected yeast?” A few years ago one of the yeast companies introduced a couple of mixed wild yeast / wine yeast cultures for wine; they didn't catch on.  There is debate as to how much of the observed differences in natural fermentations are attributable to wild yeast versus different population dynamics of the wine yeast.  Some winemakers try to replicate natural fermentations by “micro-inoculating” with pure strain wine yeast.

In addition to wild yeast, a number of mold species can grow on unfermented grapes.  These are  generally only a problem if the incoming grapes have rot and are not treated with sulfites. 

Most of the bad guys in winemaking are bacteria.  Lactic acid bacteria other than Oenococcus are the most frequent cause of problems during fermentation.  They can multiply faster than wine yeast, gobbling up nutrients the yeast need and releasing acetic acid, which is toxic to yeast.  This can result in a stuck fermentation and elevated VA.  One particularly virulent species that emerged in the 1990's was named Lactobacillus kunkei in “honor” of the esteemed UC Davis wine microbiologist, Ralph Kunkee.  Sulfite addition at crush is usually sufficient to suppress lactic acid bacteria, but some wineries use lysozyme, a natural antibiotic found in tears and saliva and isolated commercially from egg whites.

Acetobacter sp. (vinegar bacteria) can also produce large amounts of VA quickly.  They need oxygen, so they can be controlled by protecting wine from air.  During active fermentation CO2 displaces air from fermenters; after fermentation, tanks and barrels must be kept “topped up” or blanketed with an inert gas such as CO2 or argon.

There are some incidental microbes that don't grow in wine but have caused severe headaches for wineries.  Foremost amongst these are the molds that produce trichloroanisoles and tribromoanisoles.  These are the compounds responsible for cork taint and, even worse, “cellar taint”.  They can grow in wooden tanks and barrels, wine hoses, drains, and even posts and beams.  Use of chlorine bleach sanitizers, once common in the wine industry, has become a big no-no, as these can provide substrate for the formation of TCAs. 

Finally, the truly ugly, many winemakers' greatest fear: Brettanomyces (“Brett”).  My comments here will be limited; if you want more in depth information, a Google search will yield thousands of references.  Brett is hard to detect, can produce compounds that smell and taste awful, and can grow even in bottled dry wine.  Historically, the aromas imparted by Brett were accepted as part of the regional character of Bordeaux and other areas.  The spoilage occurred slowly, without signs of microbial activity such as cloudiness and effervescence and was assumed to be an aspect of age related bouquet.

Brett has been isolated from the skins of grapes and other fruits, but used barrels are generally implicated as the presumed major source of introduction to wineries.  It is difficult to track Brett because it grows slowly, is inhibited by SO2 and is not easy to culture.  Wines that test negative for Brett right before bottling can go through full blown Brett spoilage in bottle.  While Brett can use alcohol as a carbon source and can grow anaerobically (without oxygen), it grows more rapidly with some oxygen and/or reducing sugars.  Residual sugar in wine increases the risk of Brett growth, as does the use of new oak with its contribution of sugars that are not fermentable by Saccharomyces.  New oak may be a major factor in the high historic incidence of Brett in Bordeaux.

Common descriptors of Brett spoilage in red wine include “horsey”,  “mousy”, “wet dog” and “bandaid” for aromas and “metallic” for taste.  The main compounds responsible are 4-ethylphenol (4EP), 4-ethylguaiacol (4EG) and isovaleric acid (IVA).  There is a school of thought that small amounts of these compounds can make a wine more complex and interesting, but virtually nobody cares for wines dominated by these aromas.  White wines don't have the substrate for 4EP and 4EG production, but can develop high IVA levels and smell like vomit or worse.

Brett is sensitive to SO2, which inhibits its growth but does not kill it.  It can usually be controlled in wineries through stringent sanitation and maintenance of sufficient SO2 levels in aging wine.  SO2 levels decline after bottling, eventually to the point where any dormant Brett cells can begin to grow.  Many wines are sterile filtered as insurance against Brett.  Otherwise, dry wines (less than 0.3% RS) that have completed MLF (less than 0.01% malic acid) could be bottled unfiltered with little fear of anything ever growing.  An alternative to sterile filtration is pre-bottling treatment with Velcorin™ (dimethyl dicarbonate), which will kill Brett and any other living thing in wine and breaks down very quickly.  Velcorin loses effectiveness if the wine is anything less than brilliantly clear prior to treatment,.  You need a permit, training, and special equipment to use it, and many of us are philosophically opposed to the concept as well.  One other factor can be a very effective inhibitor of Brett: alcohol.  As far as I know, Brett has not been found in wines of 15% alcohol or more.  In our case, if we have a wine above 15%, we feel comfortable bottling unfiltered, but we don't aim for 15% in most of our wines.