Evolution of Soil Over the Past 1 Million Years Determines Terroir in Wine

WARNING: This post is a lot longer than usual, so if you’re short on time, I’d recommend skipping to the conclusions and coming back to the rest another time 😉

As wine consumers, we hear the word “terroir” thrown around a lot when talking about the flavors, aromas, and other characteristics of particular wines. In a very basic sense, terroir is the expression of a physical place through the flavors and quality of a wine. There are a lot of factors that influence the quality and characteristics of a wine which in concert constitute this concept of terroir, including the climate of the vineyard where the grapes were grown, the soil, the geography, as well as the characteristics of the grape variety itself.

Many studies have shown that the geography component of terroir is particularly important, as the shape of the ground, the climate in the area, the availability of water, the strength and frequency of the winds, and the placement and frequency of rocks in the soil all interact with the natural characteristics of the grape itself to create a unique wine profile. It is assumed that certain types of soil are important in the expression of terroir; however, the research that has been done has only shown weak associations between the two. This association between soil type and terroir appears to be weak due to the fact that the soil of

any given vineyard has likely experienced significantly different transformations and successions related to weathering, erosion, mixing with organic particles, and human intervention. While two vineyards may appear to have the same type of soil, the processes of how that soil came to be throughout the millennia may be drastically different, thus further complicating macroscopic and microscopic soil-environment interactions that ultimately contribute to the complexity of terroir.

Thinking of it another way: say two people did a math problem and came up with the answer 15. Person 1 was doing a simple math problem involving the addition of 12 and 3, while Person 2 was doing a much more complicated problem involving derivatives, differential equations, and other confusing calculus methods. The final answer is the same, but the way they got to the answers is different. One is simple, while the other is much more complicated. Taking the whole process into consideration, you end up with two completely different math problems even though the answers were the same. Applying this concept to terroir, you might have two vineyards with similar soil types, however the changes and experiences in geographical areas experienced could be very different, thus resulting in wines that are completely different from one another, perhaps with one expressing a more complicated flavor profile and the other expressing a more delicate or simple profile. This analogy isn’t perfect, but it makes sense in my head ;). Soil is a “metastable complex system” which while the majority of the soil can be characterized as one type (i.e. clay, etc), the complexities that make up the remainder of the soil are unique to each site, resulting in the unique concept of terroir in the wine made from grapes growing in each of those soils.

The goal of the study presented today aimed to link the differences between soil characteristics of different vineyards with the changes experienced by that soil throughout the Quaternary period (i.e. the last 1 million years of the earth’s history), be it through natural means or through human intervention, and to link these characteristics and changes with the concept of terroir.

Methods

The area of focus in this study was in Montepulciano in South Eastern Tuscany region in Central Italy. Throughout the last 1 million years, the area has undergone many geographical changes, including the overlapping of layers by plate tectonics, marine depositions, lagoon sedimentation, and fluvial cutting. In other words, under the topsoil, different areas within the Montepulciano region experienced a variety of geographical and subterranean changes that resulted in unique soil layer compositions depending upon where you look. The paper presented today goes into great detail of exactly what happened with the geography at specific periods during the last 1 million years, but I will leave out the details for now for the sake of relative brevity of this post.

The climate of the area is considered to be sub-oceanic Mediterranean with some variability depending upon the geography of the specific area of focus. The top of the Chianti-Cetona ridge experiences an annual average temperature of less than 12 oC, while the valley floor experiences as annual average temperature of 14 oC. The summer months tend to be driest, while it can be very rainy in the spring and fall. Many of the vineyards are located on sloped land of 300-450 m a.s.l.

Vineyards included in the study were selected based on vegetation make-up, plant homogeneity, age, and proximity to meteorological stations (for climatic data). Planting density, training systems, canopy and bud characteristics, and grape production levels were all relatively similar between the vineyards selected for study. For each vineyard, the following measurements were recorded: grape yield per vine, number of clusters, average cluster weight and berry weight, sugar content, accumulation rate and titratable acidity of grapes. Once mature, 50kg of grapes were harvested from each vineyard and processed via small-lot winemaking techniques.

In terms of the soils, the following were measured: hydrological characteristics (drainage, runoff, permeability, and water table depth), bulk density, moisture content, aggregate stability, total porosity, and bulk density. In terms of climatic

data, the following were measured: daily temperature, daily rainfall, biweekly soil temperature, and moisture content. Data including rainfall, air and soil temperatures at 50cm, Winkler’s Index, and number of days the soil was dry in the moisture control section were also available for analysis.

The following were noted by the authors as “invariant characteristics”: elevation, slope, aspect, radiation, soil particle size, rock fragments, stoniness of the surface, drainage, runoff, root depth, bulk density, field capacity, wilting point, available water capacity, and aggregate stability.

Vineyard/viticulture results were compared with the invariant characteristics data as well as the climate data. Major geological events were also reconstructed using specialized geographic software.

Results

Note: there are an incredible amount of details in the results section of this paper—for sake of length/time, I’ll just highlight the major findings, though if you have specific questions about certain things, just ask and I’ll dig deeper into the paper to see if I can find an answer for you.

• Annual average rainfall amounts and annual average temperatures were lower over the long term.
• The lengths and intensities of summer droughts were due to variable rainfall events, and not an increase in temperature (temperatures were more regular compared with rainfall events).
• The following soil properties varied very little among the different sites: topography, radiation, root depth, stoniness, bulk density, organic carbon, pH, and cation exchange capacity.
• The following soil properties varied significantly among the different sites: soil texture, available water capacity, lime, and electrical conductivity.
• Functional soil characteristics were significantly different between soil types.
• Based on hydrological data, the sites were able to be separated into these distinct groups:
o San Gimignano, Strada, and Valiano soils. These soils were without a water deficit and saturation during the entire growing season.
o San Quirico, Quercia, Poggio Golo, and Valiano aquic soils. These soils possessed qualities of near saturation during early spring, and a moderate water deficit in the summer.
o Cusona soils. These soils were found to have a strong water deficit in the summer months.
o Monte soils. These soils were found to have a strong water deficit in the summer and near saturation conditions during the early spring.
• Air and soil temperatures were higher in the vineyards at low and average elevations.
• Winkler’s Index was significantly lower at the highest elevations.
• Annual rainfall did not vary significantly at different sites, as well as the annual number of dry days per season.
• Grape production was significantly associated with soil type.
o The highest grape yield was observed for Valiano soils, and the lowest grape yields were observed for Cusona and Monte soils.
• Grape weight was significantly associated with soil type.
o The largest grapes were observed in San Gimignano and Valiano soils.
o Since the surface:volume ratio is increased with larger berries, the ratio of polyphenols:juice is decreased, thus wine quality is also decreased.
o The smallest grapes were observed in Monte, San Qurico, Quercia, Cusano, and Valiano aquic soils (thus higher wine quality).
• Grapes from vineyards on Cusano soils had the highest average sugar levels, while grapes from vineyards on San Gimignano and Poggio Golo soils had the lowest average sugar levels.
o The same trend was noted for total acidity.
• Wines made from grapes grown in Valiano aquic, Cusona, and Monte soils showed good structure, typicity, and strong berry/cherry flavors. Sensory profile stability, however, was found to be very low in these wines.
• Wines made from grapes grown in Quercia, Poggio Golo, and San Quirico soils showed good structure, typicity, and mid-levels of berry/cherry flavors. Sensory profile stability was very strong in these wines.
• Wines made from grapes grown in Strada soils showed medium structure, medium astringency, and low levels of berry/cherry flavors. Sensory profile stability was strong in these wines.
• Wines made from grapes grown in San Gimignano and Valiano soils showed poor structure, high astringency, and low typicity.
• Sugar content, sugar accumulation rate, grape yield, cluster weight, 100 berry weight, and total acidity were significantly associated with just two invariant environmental characteristics: available water content and root depth.
o These associations explained 62% of the variation in the model using PCA analysis.
• Combining the results of yearly variability, invariant environmental characteristics, variable soil and climate characteristics, the authors were able to group different soil types into different “viticultural suitabilities”:
o Group 1: Cusona, Monte, and Valiano aquic soils. Generally, these soils were noted for “severe permanent limitations to crop cultivation” based on various factors including water table levels, water stress in general, and poor root development. Wine made from grapes grown in these soils had good structure and decent balance, but the overall quality was highly variable from year to year.
o Group 2: San Quirico, Poggio Golo, and Quercia soils. Generally, these soils showed “intermediate growing conditions”, though they were noted for having an excess of water in the spring and a deficit of water during the summer months. Wines made from grapes grown in these soils had good structure, typicity, and stability, as well as strong cherry and berry flavors.
o Group 3: San Gimignano, Strada, and Valiano soils. Generally, these soils were “excessively fertile”, with a lot of water and free oxygen available to the plants throughout the entire growing season. This resulted in basically no stress to the vines, and basically big fat watered-down grapes. Wines made from grapes grown in these soils were generally the lowest quality, with poor structure, typicity, and balance.
• In terms of landscape development with the past 1 million years, the authors go into great detail about each of the eras and how the geography had changed. Again, for space limitations, I won’t go into those details, but if you have specific questions, just ask!
• The evolution of soil formation in all of the soil types examined in this study was that it likely formed after the Neolithic and Early Bronze Age, as many artifacts from the Etrusians and Romans were quite often found in the soil profile.
o San Gimignano and Strada soils formed on marine sands rich with marine salts, organic oxides, and iron oxides. These soils also possessed good porosity. Combining these qualities plus climate characteristics, the soil could be easily broken down resulting in soil with good rooting capabilities.
o San Quirico and Quercia soils formed on marine clay and silts, which resulted in a much more solid/impermeable medium and ultimately poor rooting capabilities. These soils evolved into “blocky” sections of soil, due to a combination of poor porosity and drying/wetting cycles throughout the growing season.
o Valiano and Valiano aquic soils formed on Pleistocene fluvial-lacustrine clay. These soils are low in salt content, since the soil was formed by continental rocks and not oceanic rocks. These soils evolved to have good drainage as a result of “angular blocky aggregates” (think larger lumps of solid soil separated by pockets of air that allowed for good drainage and rooting capabilities).
o Human influence continued to change these soils further, which added another level of complexity to the mix since people were doing different things to the soil at different locations.

Conclusions

Though I left out a lot of the specific details of the study in this summary post, it is clear that there is some sort of connection between the soil and general environment and the aromatic and structural characteristics as well as the overall quality of the wine made from grapes grown in a particular environment. Throughout the last one million years, the slow evolution of the soils resulted in soils that while they may possess similar qualities and characteristics (topography, radiation, root depth, stoniness, bulk density, organic carbon, pH, and cation exchange capacity), there were some characteristics that appeared to change the soil significantly enough that each soil expressed its own unique terroir through the finished wine created from grapes grown in that soil (soil texture, available water capacity, lime, and electrical conductivity).

In particular, the authors noted that human intervention had a huge impact on the differences between soils, and things such as heavy machinery, bulldozing, and other human-made technologies implement to prepare the soil for planting likely made a significant impact on the overall expression of terroir in the finished wines.

The quality of wines produced from each geographical area is not surprising when compared with the structural and environmental characteristics of the soil. Those environments with a lot of water and poor drainage resulted in poorer quality wine, while those environments under more water and

physiological stress resulted in the highest quality wines in this study. Knowing how the different soils developed and what wine is like when created from grapes grown in those soils may provide important information to vineyard managers and winemakers in terms of predicting how a particular vineyard will perform as well as how different soils should be handled or treated in order to have a soil worthy of planting a vineyard upon.

While this study was full of interesting information on how geological and environmental changes over the past 1 million years have shaped the soil of vineyards today, I found myself wanting to know just one more thing: Can you determine the exact compounds or soil characteristics that allow you to identify terroir? What I mean is this: we see from the results of this study that different soils undergoing different changes over evolutionary time produce different types of wine. We also know from the results of this study that only a small number of soil characteristics were actually significantly different between these different soils, while many other of the characteristics were statistically similar. How do these soil differences correlate to chemical differences found in the finished wines? Can you correlate one or more chemical compounds in the finished wine with the soil characteristics that differ between sites? Would you be able to use this information to predict where any given wine was made (or where the grapes were grown, anyway). Finally, can you correlate any one or more historical events to the specific change in the soil that caused the specific change in the wine? OK, maybe this isn’t THAT important, but it’d be cool for pure nerd’s sake!

The follow up experiment I would like to see is this: let’s take the same vineyard sites as this study used (or at least a group of several vineyard sites that produce the same wine in a slightly unique manner). Next, plant experimental vines of the same grape (same clone) at all sites. Then, after a few seasons when you have grapes mature enough for winemaking, harvest the grapes and made wine the exact same way for each site. At the same time, measure a ton of soil and environmental characteristics, as well as historical soil manipulation data. Analyze the soil and finished wine chemistry (as well as the grape chemistry prior to winemaking) and perform correlation tests and Principle Components Analysis (or some other analysis like PCA) to see if any one particular soil characteristic can be tied to uniqueness that is the terroir expressed at each individual vineyard site. Of course, this is a very rough design here and I’m sure there are some problems with it, but you get the idea!

I’d love to hear what you all think of this very long study! Would you have liked to have seen something else done during this study? What would you add/change? Please feel free to leave your comments!

Source: Costantini, E.A.C., Bucelli, P., and Priori, S. 2012. Quaternary landscape history determines soil functional characters of terroir. Quaternary International 265: 63-73.