Case Study 2: The Roquefort Caves and the Geography of a Mold

There is a village in southern France called Roquefort-sur-Soulzon. It sits at the edge of the Causse du Larzac, a limestone plateau in the Aveyron department, about 90 kilometers northwest of Montpellier. The village has roughly 600 residents. It has had cheese for at least 2,000 years — a Roman document from the first century CE refers to a cheese being made on the plateau, and there are local stories (whose dates are flexible) of a shepherd accidentally leaving his lunch in a cave and returning weeks later to find the bread covered in blue-green mold and the cheese transformed.

The cheese is Roquefort. It is a sheep's-milk blue cheese, ripened with Penicillium roqueforti — a mold named for the village. Roquefort is one of the world's most legally protected foods. To be called Roquefort, the cheese must be made from raw milk of the Lacaune breed of sheep (or two specific other breeds), curdled and salted in a particular way, and aged in the natural caves of Mont Combalou under the village. Cheese aged anywhere else, even using identical milk and identical molds, is not legally Roquefort.

This case study is about why those caves matter, what is happening inside the cheese, and why a single appellation has held for over a millennium.

What the caves are

The caves of Combalou were formed by a geological collapse, perhaps two million years ago, in which a section of the limestone plateau slumped into itself. The collapse created a network of fissures and chambers running through the mountain. Some of those fissures connect to the outside, and one specific feature — the fleurines, vertical air shafts — connects the deep caves to the surface above.

The fleurines are the engine of Roquefort. As the outside air heats and cools across day and night, and across seasons, air masses rise and fall through the fleurines. The air entering the caves passes through the surrounding limestone, which has its own moisture content and microclimate. The result, inside the aging caves, is:

  • Temperature: stable around 10–12°C (50–54°F) year-round.
  • Humidity: roughly 90–95%.
  • Air movement: gentle but constant — never stagnant.

These conditions are extraordinarily consistent. They are also, as it happens, exactly what Penicillium roqueforti needs to thrive, what the cheese's surface microflora needs to remain in balance, and what the slow proteolytic and lipolytic chemistry of aging requires to proceed at the right rate.

The cheese can be aged elsewhere. The chemistry will proceed. But the rate, the balance, and the specific volatile-aroma profile will not be the same. Roquefort made anywhere else is a similar cheese, not the same cheese.

The mold's biology

Penicillium roqueforti is a filamentous fungus — a mold — that grows by extending hyphae (thread-like cells) through whatever it is colonizing. It produces spores that are blue-green to gray-green, depending on the substrate and the strain.

The species has at least four major genetic lineages, distinguishable by DNA analysis. The lineages found in Roquefort cheese are not the same as the lineages found in silage or in spoiled bread; centuries of culture have selected for cheese-specific strains. The genetic differences correlate with metabolic differences: cheese-strain P. roqueforti tolerates higher salt, grows better in low oxygen, and produces a different mix of secondary metabolites than wild P. roqueforti does.

What the mold does inside a wheel of Roquefort:

  1. Lipolysis. P. roqueforti produces several lipases — fat-cleaving enzymes — that break down milk triglycerides into free fatty acids. The fatty acids that accumulate include butyric acid (4-carbon — sharp, slightly cheesy), caproic acid (6-carbon — goaty), caprylic acid (8-carbon — sharp), capric acid (10-carbon — sharp, slightly waxy), and others. Free fatty acids are far more aromatic than triglyceride-bound fatty acids; the lipolysis is what gives blue cheeses their pungency.
  2. Methyl ketone production. A fraction of the released fatty acids are partially oxidized by the mold into methyl ketones. The most important are 2-pentanone (5-carbon), 2-heptanone (7-carbon), and 2-nonanone (9-carbon). Methyl ketones are intensely aromatic — fruity, slightly sharp, characteristically blue-cheese-like. They are responsible for much of the signature blue-cheese aroma.
  3. Proteolysis. P. roqueforti also produces proteases that degrade casein into peptides and free amino acids. The proteolysis softens the cheese's texture (a young Roquefort is firm and crumbly; an aged one becomes soft and almost spreadable) and contributes to flavor through free glutamate, free aspartate, and other umami amino acids.
  4. Spore production and spread. The visible blue veins of Roquefort are P. roqueforti sporulating. The veins follow the air channels through the cheese — channels created by mechanical piercing of the wheels with stainless steel skewers during early aging. The mold's hyphae grow inward from the punctures along the air gradient.

The cheesemaker's choreography

A wheel of Roquefort takes between three months and a year to age. The choreography is, abbreviated:

  • Day 1. Sheep's milk (raw, unpasteurized — Roquefort is one of the protected raw-milk cheeses) is warmed, inoculated with both standard lactic-acid bacteria starters and a measured spore preparation of P. roqueforti. Rennet is added; the curd sets.
  • Days 1–3. Curd is cut, drained, salted, and packed into perforated cylindrical molds. Drainage continues; the wheels firm up.
  • Days 4–7. Wheels are unmolded and dry-salted on the surface, turned daily.
  • Day 7 or so. Wheels are piqué — pierced with long, thin steel needles, traditionally about 30 to 60 punctures per wheel. The punctures create air channels through which oxygen can reach the P. roqueforti spores already present in the cheese's interior.
  • Weeks 2–4. Wheels are moved into the Combalou caves. P. roqueforti begins to sporulate visibly along the air channels. Blue veins appear.
  • Weeks 4–12 (and sometimes longer). Cheese is aged in the caves. The caves' relative humidity, temperature, and air flow are not actively controlled — they are what they are because of the mountain's geology and the fleurines.
  • Wrapping. At a point that the affineurs (the specialists who manage aging) judge correct, wheels are wrapped in foil and held at lower temperature to halt further mold growth and stabilize the cheese.

What scaling teaches us

The cheese-industry has tried, repeatedly, over the past century, to produce Roquefort-like cheese in industrial caves elsewhere — climate-controlled aging rooms in commercial dairies in Wisconsin, in Australia, in Japan. The cheeses produced are good. They are not Roquefort, even when made with starter cultures and P. roqueforti spores derived from the original caves.

The reasons are several:

  • The microbial environment of a 2,000-year-old cave is irreproducible in a year-old climate-controlled room. The wild flora of the Combalou caves — bacteria on the walls, yeasts in the air, other molds in the rock — contribute to the cheeses aged there. A new aging room has only the organisms that arrive with the cheese itself.
  • The fleurines' particular humidity and air-flow rhythm is not exactly reproducible. Mechanical air handling produces relatively constant conditions; the fleurines produce a rhythm of slow oscillations that may matter to the cheese's surface microflora.
  • The milk is from sheep grazed on the Causse du Larzac plateau — a specific high-pasture environment with specific plants, specific minerals, specific seasonal patterns. Sheep raised on different pastures produce different milk.
  • The starter cultures are local. Twenty-plus generations of milk-handling in the village have shaped the lactic-acid-bacteria population in equipment, hands, and air.

This is partly the terroir argument applied to cheese — and unlike some terroir claims for wine that have not held up well to blind tasting, the Roquefort terroir argument is at least partly defensible scientifically. The cave's microflora is real. The geological air flow is real. Sheep's milk from this pasture differs measurably from sheep's milk from other pastures.

It is not magic. It is a stable assemblage of organisms, conditions, and milk that has been refined over a thousand-plus years of practice in one specific place.

The honest counter-point

It would be a mistake, however, to use the Roquefort case as a story about French cheese being intrinsically superior to other blue cheeses. Stilton (England), Gorgonzola (Italy), Cabrales (Asturias, northern Spain), Bavaria's Bayerischer Blauschimmelkäse, Denmark's Danablu, and many regional blue cheeses across Europe and beyond each have their own combinations of milk, mold, and aging environment. Each is the product of its own multi-century evolution.

And — perhaps more importantly — the P. roqueforti in Cabrales caves in northern Spain is not the P. roqueforti of Roquefort. Cabrales is aged in natural caves in the Picos de Europa mountains; the local mold strains evolved separately. The cheeses are cousins, not children.

The story of P. roqueforti across Europe is essentially a parallel-evolution story: the same species, growing wild on grain stores and on cheese throughout temperate Europe, encountered milk in many places and entered many cheese-making traditions independently. Each tradition selected its own strain. Each cave evolved its own air. Each milk had its own pasture. The result is dozens of regional blue cheeses, each in some sense "the local P. roqueforti on the local sheep's or cow's or goat's milk in the local conditions."

Analyze this

For your classroom or self-study:

  1. The European Union grants Roquefort Protected Designation of Origin (PDO) status, which means cheese made anywhere outside the protected area, even with identical methods, cannot be called "Roquefort." Is this defensible scientifically (because the cheese genuinely depends on the place), legally (because tradition deserves protection), or commercially (because European producers have lobbied for it)? Make a case for one or more of these and identify what the others might say.
  2. P. roqueforti tolerates low oxygen. P. camemberti (the white-rind Brie/Camembert mold) does not — it grows on the surface. How does this single physiological difference produce two visually and texturally distinct cheese categories from molds that are otherwise close cousins?
  3. Industrial blue cheese is often inoculated with concentrated spore preparations and aged in controlled-atmosphere rooms. Does it taste like Roquefort? (Try a taste test with a study group: a Roquefort, an industrial blue, and one other regional blue such as Gorgonzola.) What can you detect in each?
  4. Penicillium species also include P. chrysogenum (the original penicillin source) and various human pathogens. How does the cheese industry ensure that the P. roqueforti and P. camemberti strains used for cheesemaking are safe? (Hint: research the concept of "QPS" status — Qualified Presumption of Safety — and the genome-level differences between cheesemaking and toxin-producing Penicillium strains.)
  5. If you wanted to learn the chemistry of blue cheese without flying to France, what are three things you could do at home with reasonable confidence of teaching yourself?