Soil health in the vineyards of North-Western Italy

1 Introduction

Soil health is the “ability of the soil to sustain the productivity, diversity, and environmental services of terrestrial ecosystems” (1). While this definition encompasses all soil ecosystem services, the term is usually applied to the capacity that soils have to support plant growth and food production. In this context, soil health serves as a synthetic indicator of soil biological, physical, and chemical fertility. By integrating these aspects, it provides a comprehensive overview of both the interrelationships among properties and their connection with soil functions. To integrate multiple soil properties into a single synthetic metric, soil health indexes (SHIs) have been developed. They range from the simple ranking of a specific soil property with respect to the other soils of the area, to more complex procedures that involve machine learning and AI techniques (e.g., 2–4). The most well-known example of the first approach is the Cornell Assessment of Soil Health (CASH) (5), which has been applied to a wide range of agricultural soils, mainly cultivated with annual crops.

In the CASH model, soil properties are grouped according to the effect they are expected to have on plant growth and crop yield. In the past years, great emphasis has been given to the chemical indicators of soil fertility, but the CASH model also appropriately describes the soil health status by taking into account physical and biological indicators. For all of them, three model types are assumed: “more-is-better” (which applies for example to organic matter), “less-is-better” (commonly used for contaminants), and “optimum value” model (e.g., often used for pH). The use of these models should be adapted based on the pedological conditions of the area. For example, if all soils lie in the same range of acidic pH, the “more-is-better” model may be more efficient than the “optimum” one.

Recently, the applicability of SHIs to viticulture and other permanent crops has been questioned (6). In terms of soil fertility, viticulture represents a specific case indeed. High-quality wines are often produced in environments where moderate stress optimizes the synthesis of secondary metabolites in grapes. This contributes to the desirable flavor profile (7). However, the stress must not be severe enough to impair plant function or reduce yield (8). This concept is embodied in the definition of terroir, where climate, soil, management and cultural conditions converge, allowing the establishment of a complex and specific equilibrium (9). Thus, the standard SHIs calculation based on the “more-is-better” model is likely poorly applicable: quality, and not merely quantity, is the objective of viticulture in the terroir (10). In terroir-driven viticultural systems, soil conditions that would be classified as suboptimal or limiting within conventional SHI frameworks may, in fact, represent optimal functional states for high-quality wine production. This conceptual mismatch highlights the need to reinterpret soil health in viticulture as a context-dependent property, where the balance between limiting and favorable conditions, rather than their maximization, defines soil functionality. Furthermore, appropriate SHIs are essential to improve soil health in the terroir as vine growers would not be willing to use soil-friendly management practices if the benefits for crop production are not clearly demonstrated (11).

At the regional and administrative levels, the terroir is transposed into the wine-producing appellation areas. In Italy, the appellation areas (DOC and DOCG, the latter corresponding to highest quality) represent a combination of climatic and topographic conditions (elevation, aspect), administrative boundaries, and compelling vineyard management techniques. Wines originating in specific areas are expected to have superior quality with respect to those produced from the same grapevine cultivar in other areas. However, in Italy there is no explicit link between appellation areas and soil types, despite the designation of the appellation areas takes into account several geographical and environmental characteristics. Few studies have examined this relationship. In Tuscany, the Chianti administrative DOC and DOCG areas were found to both encompass a wide range of soil types (12). Different soil types may, however, share similar physical and/or chemical characteristics; therefore, the soil characteristics of the DOCG areas might represent a first approximation of the optimal conditions for high-quality wine production and may serve in the evaluation of soil health in high-quality viticulture.

We hypothesized that soil characteristics and overall soil health differ between DOCG and DOC areas because, as stated above, wines of the highest quality likely originate from more limiting soil health conditions, such as e.g. low amounts of available nutrients, relatively high pH, and a relatively low SOM (e.g. 13). When these conditions remain within tolerable thresholds, they can improve berry composition and quality, while when excessive they can compromise the production. For example, severe water stress can reduce yield and affect ripening (14); alkaline pH associated with high active lime can induce Fe chlorosis (15); severe nutrient deficiencies can result in poor yield and grape quality (16).

Accordingly, our objectives were: i) to assess the soil health status of vineyard soils in the Piemonte region (NW-Italy) based on existing soil legacies; and ii) to evaluate whether soil health indicators can discriminate terroirs for the three most commonly cultivated red grape varieties (Barbera, Dolcetto, and Nebbiolo) across DOCG and DOC areas. This second objective extends conventional soil health frameworks by identifying how a functional equilibrium between stress and fertility supports high-quality wine production.

2 Materials and methods

2.1 Data selection and analytical methods

All data were obtained from publicly available databases maintained by the Regione Piemonte (NW Italy) administration geoportal and databases (Supplementary Table 1). The soil data source was the “Atlante delle analisi del suolo”, that includes all soil samples analyzed by the regional labs. The database contains approx. 27,500 records, which were then filtered by land use (vineyard) and sampling depth (0–30 cm). The total amount of samples used in this study was 2,812, i.e., all vineyard topsoils in the database. In addition to land use, the actual grapevine variety was available in 2,044 cases. The samples were then linked to a soil type unit in the Piemonte Soil Map 1:50,000 (USDA Soil Taxonomy), to a geo-lithological unit, elevation, slope and aspect (derived from a digital terrain model with 10 m resolution), and to the presence of one or more DOC or DOCG areas. There are 19 DOCG and 41 DOC wines in Piemonte (Supplementary Table 2) The DOCG wines produced from red grapevine varieties are the most common (14), and they mainly derive from Barbera, Dolcetto, Nebbiolo grapevines; we therefore focused on these varieties.

The records related to Barbera, Dolcetto and Nebbiolo were 1,343 of which 823 cases are situated in areas of DOCG appellations. Therefore, 519 sites were either DOC or areas where more than one appellation is allowed: two DOCs or a DOCG and a DOC. The localization of the samples is reported in Figure 1.

The analytical data available in the database were: sand, silt, clay (pipette method), pH (water), total CaCO₃ (volumetric), active CaCO₃ (KMnO₄-extractable), soil organic matter (SOM, dry combustion), total nitrogen (N, dry combustion), C/N ratio, cation exchange capacity (CEC, BaCl₂ pH 8.1) and exchangeable Ca, Mg and K, available phosphorus (Pav, Olsen method), available Fe, Mn, Zn and Cu (DTPA extraction) (Supplementary Table 3). All analyses were performed according to the official Italian methods of soil analysis (17). In addition to these analytical data, soil bulk density (BD) and soil hydrologic characteristics were calculated from pedotransfer functions. Bulk density was estimated using the pedotransfer function developed by Jeffrey (18), easily applicable in hilly areas (19). The available water capacity (AWC) was estimated from the water contents at field capacity (FC) and that at the permanent wilting point (PWP), both estimated according to Toth et al. (20). The use of pedotransfer functions to calculate missing measured physical properties may lead to erroneous estimates in case of stony soils (21) and does not take into account the tridimensional arrangement of soil solids into aggregates. Despite these disadvantages, pedotransfer functions are often used, also in the calculation of soil health indexes (e.g. 22).

The soil health index (SHI) was evaluated following the Comprehensive Assessment of Soil Health (CASH) in two steps (5). For the physical indicators of soil health, we used soil AWC and BD; SOM and C/N ratio were used as biological indicators; pH, available P, exchangeable K, Mg, and micronutrients were used as chemical indicators. Magnesium and micronutrients (Fe, Mn, Zn) were combined into a single indicator, as indicated in the CASH model despite Mg is not a micronutrient, computed by averaging the four scores (5). Each soil sample was scored using the approach “more-is-better” or the “optimum” range. The “optimum” model was used for: BD (optimum at 1.3 g cm^-3), pH (optimum at 6.5), and C/N (optimum at 10), according to recommendations provided by Celi et al. (23). The other variables were divided by the maximum recorded value in the database (“more-is-better”). Consequently, each indicator was standardized in a 0–1 range, where records with a value of 1 signify the optimal conditions attainable within the area. A global SHI was obtained by averaging all singles SHIs without weighting them, as suggested by Andrews et al. (24), because they found few differences using this simple approach with respect to other more complex methods. The minimum number of valid cases was 339 for extractable Fe. For all soil records, at least 4 indicators of soil health representing the three indicator types (chemical, physical, and biological) of the CASH model were available. To check that no bias on indicators was introduced because of the high number of missing cases for micronutrients, some additional statistical analyses were performed as shown in Supplementary Figure 1. No trend was visible in the average value of soil chemical health index with increasing data completeness.

2.2 Statistical analyses

Significant differences between groups (e.g., DOCG vs DOC vs Both, or Barbera-Dolcetto-Nebbiolo vs Others) were evaluated using one-way ANOVA (p = 0.05), with Tukey HSD post-hoc test. Before performing ANOVA, all variables were tested for homoscedasticity, and those with non-equal variances were log10 transformed. Chi-squared was calculated to infer the distribution of frequencies of the different soil types. Pearson correlation coefficients were also calculated for selected variables. The data are presented either as boxplots or population distributions. All statistical analyses were performed using SPSS, and the plots were generated with R version 4.2.3.

3 Results

3.1 Characteristics of vineyards and vineyard soils

The variability in soil type was large, as vineyards have been planted both on poorly developed Entisols and on Pleistocene Alfisols. The highest proportion of vineyards was on Entisols of the hills of central Piemonte (in a famous wine-producing area called Langhe and Monferrato) and in pedemontain areas (n=1,730), followed by the slightly more developed Inceptisols of hilly and pedemontain areas (n=545) and the Alfisols (n=491). Forty-four Vertisols and two Mollisols were also present. Elevation ranged from 92 to 1397 m a.s.l. and slope from 0 to 42% (Table 1).

Table 1. Site, physical and chemical soil (0–30 cm) characteristics of Piemonte vineyard soils.

BD, bulk density; FC, field capacity; PWP, permanent wilting point; CaCO3T, total carbonates as calcite equivalent; CaCO3A, active carbonates; SOM, soil organic matter; CEC; cation exchange capacity, Caex, Mgex, Kex, exchangeable cations; BS, base saturation; Pav, available P; Feav, Mnav, Znav, Cuav, DTPA-extractable elements.

As expected, soil physical and chemical characteristics varied across the whole region (Table 1). All particle size classes ranged from very low contents to more than 60% (clay) or 80-90% (sand and silt); the average bulk density was 1.39 kg dm^-3, and the available water contents were low (13% v/v on average), despite an average field capacity of 29%. Soil variability was less visible in some chemical properties. The average pH was 7.7 as most soils contained carbonates and approximately 90% of the samples had SOM contents below 50 g kg^-1. The total N contents showed a good correlation with SOM (r= 0.934, p<0.001, n=2094). The average CEC was 17 cmol₍₊₎ kg^-1, more related with clay contents (standardized regression coefficient 0.671, p<0.001) than SOM (standardized regression coefficient 0.240, p<0.001). The base saturation was, on average, above 90%, in agreement with pH and carbonate content. More than half of the samples had less than 10 mg kg^-1 available P, which is considered as a threshold for P-poor agricultural soils (25). The extractable microelement contents (Zn, Cu, Mn, and Fe) also indicate paucity in most soils, with the exception of Cu contents (10% of cases above 100 mg kg^-1, and four cases above 300 mg kg^-1).

The soils of the three selected red grapevine varieties (Barbera, Dolcetto, and Nebbiolo) were at a slightly lower average elevation than the other vineyard soils (260 vs 290 m a.s.l., p< 0.001) but had similar slope and aspect (Table 2). They showed a higher proportion of Entisols than expected (69%, Chi-squared p<0.001), and consequently, the other soil orders were less frequent. They had less sand (24 vs 29%) and slightly higher amounts of silt and clay (p<0.001). Consequently, the hydrological properties also differed, with higher amounts of water at the permanent wilting point (p<0.001) and a lower AWC. Despite the lower overall variability in chemical properties of the vineyard soils of the region, those cultivated with Barbera, Dolcetto or Nebbiolo showed a higher pH, and higher amounts of active carbonates (p<0.001). Differences were also visible in the amounts of available elements: P, Mn, Zn, and Cu were always less available (p<0.001) in the soils of the three grapevine varieties than in the other vineyards, but exchangeable Mg and K were more abundant (p<0.001).

Table 2. Differences among vineyard soils from the most common red grape varieties (Barbera, Dolcetto, Nebbiolo: BDN), and all other vineyard soils (0–30 cm) in Piemonte.

The soils where Barbera, Nebbiolo, and Dolcetto vines grow differed, depending on the appellation (Figure 2). In most cases, the soils of areas where more than one appellation is allowed showed intermediate values between DOCG and DOCs. The DOCG areas were at lower elevation (Figure 2A) and were clay and silt-richer (Figures 2B, C). Consequently, they showed higher FC and PWP than the DOC areas (Figures 2F, G). DOCG and DOC areas significantly differed in all the considered chemical characteristics with the exception of total N (Figure 2K). The DOCG soils were significantly more basic (Figure 2H), with more available Ca, Mg, and K (Figures 2N–P) than the other appellation areas. In DOCG areas, the CEC was higher as well (Figure 2M), although they had lower organic matter content (Figure 2J). They were also characterized by a lower C/N ratio (Figure 2l), and lower available P content (Figure 2Q). Most of the general chemical properties (i.e., clay, sand, PWP, CEC, Mgex, Kex, Pav) were shared between DOCG areas and those where both DOCG and DOC appellations are allowed, as shown in Figure 2 (results of the post-hoc Tukey test). Conversely, DOC areas, as well as those with more than one appellation, were always richer in extractable Fe, Mn, and Zn than the DOCG areas (Figures 2R–T), and the same situation was visible for extractable Cu (Figure 2U). In the DOCG areas, Entisols and Inceptisols were more frequently observed than in the DOC areas (Chi-squared p<0.001, n=1342).

3.2 Soil health indexes

On average, vineyard soils across Piemonte exhibited a physical soil health index of 0.62 (Figure 3A), primarily driven by a favorable bulk density rather than AWC (Supplementary Table 4). In fact, more than 50% of the soils had a BD score above 0.84, while the median was 0.43 for AWC (Supplementary Table 4). These two indicators were poorly correlated with each other, although significantly (r=-0.269, p<0.001, n=1,554). The biological health index was lower than the physical one, with an average of 0.49 (Figure 3B). The transformation of organic matter, as indicated by the C/N ratio, was adequate, with 50% of the soils showing a score above 0.61 (Supplementary Table 4). The chemical index of soil health showed a left-skewed distribution (Figure 3C), with an average of 0.23. This was due to the low scores of pH, exchangeable K, microelements (particularly for Zn and Mn) (Supplementary Table 4). The global SHI (Figure 3D) mediates among the three health indicators, with a frequency distribution less skewed and more symmetrical; the mean and the median were both at 0.43.

The soils of the three selected red-berry varieties differed from the others, showing lower physical and chemical indexes, as well as a lower global SHI (Figures 3E, G, H). The biological health indicator instead was similar in all vineyard soils (Figure 3F).

Within the vineyards of Barbera, Dolcetto, and Nebbiolo, differences in health indexes were visible depending on whether the area belongs to a DOCG or DOC appellation, or both (Figures 4A–D). The differences were mainly related to the chemical and physical health indexes (Figure 4C), significantly lower at DOCG sites (p<0.001). Differences were, however, also visible in soil health among the three varieties, with the Dolcetto vineyards showing statistically homogeneous health indexes among areas (Figures 4I–L). The Barbera vineyards where a double appellation is allowed always behave as the DOCGs (Figures 4E–H). The physical (Figure 4M), biological (Figure 4N) and chemical indexes (Figure 4O) of soil health of the Nebbiolo DOCG vineyard soils were instead always lower than both DOC and double appellation areas. However, the global SHI (Figure 4P) did not differ from that of the areas where both DOC and DOCGs are allowed (Figure 4P).

4 Discussion

4.1 Vineyard soils and appellation areas

The analysis of the soil database of the Piemonte clearly demonstrated that the soils of the three principal red grapevine varieties (Barbera, Dolcetto, and Nebbiolo) were systematically finer (i.e., richer in clay and silt). Also, within both Barbera and Nebbiolo vineyards, the DOCG areas were the finest. These differences in texture, an inherent soil property, are at the basis of the definition of a terroir (e.g., 10). Texture in fact affects soil hydrological parameters and plant water status (26), that determine grape quality. The higher water reservoir guaranteed by a finer texture has been associated with lower phenolic contents and less intense wine color and shade in commercial Spanish vineyards (27), with larger berries and lower anthocyanin concentration in Turkish (28) and Sicilian vineyards (29). Also, California growers prefer lower clay contents to induce some physiological stress (6). At the same time, Van Leeuwen et al. (14) reported a higher grape quality in clayey soils because of the moderate water deficit induced early in the season. Also Costantini et al. (30) found that clayey soils represent the optimal conditions for Sangiovese vines, because a certain amount of water is maintained at depth during the summer dry period. Indeed clay-rich soils have a greater water holding capacity, but the water tends to be less available to plants as the roots extract water more easily in the presence of a coarse texture (31). A fine soil texture is therefore not necessarily associated to an excess of vine vigor and, in the Piemonte vineyard soils, the PWP was on average more than 50% of the FC, sharply decreasing water availability to plants. Moreover, as soil hydraulic characteristics were calculated from pedotransfer functions derived by particle size distribution, the impact of soil structure on water fluxes was not considered in our analysis. The real FC is likely to be lower than the estimated one because of the interaggregate space, while the PWP estimate is probably closer to the actual one in clay-rich soils. These uncertainties could therefore lead to lower-than-predicated AWC in many Piemonte vineyard soils. As the level of soil development was lower in red-berry vineyards (i.e., more Entisols than expected), the higher clay proportion was linked to the soil parent material, with a larger proportion of marls, claystones and siltstones, and not to the presence of clay-rich B horizons. Because of the stabilizing effect of clay minerals on organic compounds (32), a positive correlation between clay content and soil organic matter was expected, and often found in other studies on large datasets (e.g., 33). Instead, clay was significantly anticorrelated with SOM (r= -0.204, p<0.001). The SOM contents were low in all vineyard soils of the region, suggesting that vineyard management does not favor the accumulation of organic matter. Rather, the transformations of organic residues seem to lead to organic carbon losses, as indicated by the C/N ratio of 8.8-8.9. The low ratio, coupled with the low SOM content, might explain the limited effect of organic matter on the CEC (more related to clay contents). Also, the DOCG areas had a significantly lower C/N ratio than the other red-berry vineyard soils (i.e., 8.6, Figure 2L). Based on these considerations, we can therefore hypothesize that the integrated production guidelines of the Piemonte Region, which require to limit tillage operations, do not favor the stabilization of soil organic matter. During spring and summer after mowing the grass cover (or after herbicide application), residues are left on the soil surface, and are not incorporated into the soil to avoid disturbances. This may possibly lead to losses of carbon as CO₂ and prevent the stabilization of organic compounds in the soil despite the favorable intrinsic soil properties (34).

The soils of the three selected red-berry varieties had low amounts of available P and micronutrients, in line with the high pH and the presence of carbonates, which lead to the precipitation of Ca-phosphates and of insoluble micronutrient forms. When Ca concentration is above 1 x 10^–4 M, clustering of Ca²⁺ ions starts at P sorption sites with the final formation of Ca-phosphates (35). The threshold concentration for these reactions is well below that corresponding to the equilibrium with CaCO₃ (36), and P deficiencies are therefore common in calcareous vineyard soils. In the soils of Piemonte, this will contribute to maintain an acceptable stress level and decrease vines vigor despite the clayey texture, ultimately affecting wine quality. In fact, a high carbonate concentration in the soil has been related to higher sugar contents in grapes and to good wine structure (30). The soils of the Barbera, Dolcetto, and Nebbiolo vineyards had higher contents of both total carbonates and active lime, and the DOCG areas were the most calcareous. Similarly, the availability of micronutrients is decreased by a number of precipitation products that impact Zn and Mn in particular (37). Although lithology-inherited soil properties are the most plausible reasons for the differences in micronutrients among vineyard soils, we cannot exclude some effects of management practices, particularly on Cu. The maximum Cu level in DOCG areas was 60 mg kg^-1, and in the same areas the average was 18 mg kg^-1. These values correspond to one third and to less than one half of the maximum and average of the DOC areas. In viticultural areas, the concentration of Cu is often correlated to the age of the vineyards, with the oldest ones showing higher concentrations because of past management practices (e.g., 38). In the DOCG areas, vineyards are not younger than in the other areas, therefore we can hypothesize a different management, with lower inputs of Cu. However, no data on vineyard management were collected for this study, nor are they available in the database; further investigations are thus needed to understand the reasons for the differences in Cu contents.

4.2 Health indexes in vineyard soils

Soil health is often lower in vineyards than in other agricultural soils, and the need for appropriate health indexes was recently emphasized by Gonzalez-Maldonado et al. (6). The health of the Piemonte vineyard soils cannot be compared to other agricultural soils of the same region because of lack of reference indexes, and few data are available about the SHI of other vineyards around the world. In some French vineyards, before the adoption of conservative agricultural practices, a soil health index that took into account soil structure, carbon transformations and nutrient cycling was on average 0.3-0.4 on a 0–1 scale (39), comparable therefore to what we found.

The main characteristics that decrease soil health in Piemonte vineyards are related to the chemical indicators. Namely, the low amounts of macronutrients and micronutrients, and the basic soil pH. In the DOCG areas, the chemical health indicator was particularly low. The basic pH and the consequences it has on nutrient availability are common features of all poorly developed soils on calcareous sedimentary rocks such as those of the Piemonte Tertiary Basin (40), and represent therefore an intrinsic soil property. However, while the management of other common fruit crops in those areas, fertilization helps overcoming nutrient shortage, low inputs of fertilizers are recommended in vineyards to avoid an excess of production (41). Furthermore, in DOCG areas, the parsimony in the use of fertilizers may be a management choice to increase wine quality. Since we calculated the SHI taking as reference population the Piemonte vineyard soils, we cannot therefore compare vineyards with other crops. However, the asymmetrical frequency distribution shown in Table 1, and the low average index show that only a few vineyards in the region have ample nutrient contents or favorable pH. The low chemical health indicator seems therefore a characteristic of the regional vineyard soils, and does not represent a soil degradation stage as far as viticulture is concerned.

The biological health indicator instead did not show any difference between DOCG areas and the other vineyards (Figure 4) nor between the Barbera, Dolcetto and Nebbiolo vineyards and the other ones (Figure 3F), in agreement with the lack of differences in organic matter contents and C/N ratio. Still, improving biological soil health should be possible in areas where soils are similarly clay-rich and the C-saturation of the clay fraction is not yet reached. In fact, the OC to clay ratio of the Piemonte vineyard soils is well below the line reported by Georgiou et al. (42) for the mineral-associated organic fraction, suggesting therefore that a more appropriate soil management could effectively improve soil health (not to the point of favoring undesired vine vigor).

The physical indicator of soil health follows more closely a normal distribution and has an average value above 0.6, similar in DOCG areas and other vineyards. The AWC is relatively high because of the soil texture, and the bulk density values indicate only slightly compacted soils (1.4 g cm^-3 on average in DOCG areas). The similarities in soil physical indicators among the red-berry appellation areas are therefore more closely related to the inherent soil properties than to soil management.

Interestingly, differences were visible among the appellation areas of the three red-berry varieties (Figure 4). Barbera DOC-only areas showed higher SHI than the areas where DOCG wines are produced or where multiple appellations are allowed, because of better physical and chemical conditions. The DOCG-only Nebbiolo vineyards were instead systematically different from the other areas of the same grape variety, with lower global, chemical and biological indicators, but higher physical indexes. Although no differentiation between DOC and DOCG areas was made by (43), these authors found a high specificity of Nebbiolo vineyards in terms of climate, which was drier than other areas (i.e., 575 mm year^-1). This confirms the hypothesis that the fine texture may favor a moderate vine stress, while helping to avoid severe damage to vegetation.

4.3 Limitations of soil health indexes calculation in vineyards

This study was carried out with the data available in the regional databases, and many physical, hydrological, and biological properties were lacking. This has certainly impacted some of our results. As hydrology parameters were lacking (as in most soil databases), the physical soil health indicators obtained through pedotransfer functions tend to reflect particle size distribution and do not include the effect of soil aggregation on pore size distribution, which cannot be assessed on disturbed soil samples. The improvement of soil structure thanks to organic matter has been known for a long time, as well as its effects on water infiltration and storage (44). More recently, several alternative management techniques have been proven to be effective in decreasing compaction and increasing porosity and water storage, such as biochar (45), zeolites (46) or organic amendments (47). However, only when measured data on physical properties are available the effect of a tailored management technique might be appreciated; in all other cases, the physical indexes derived from pedotransfer functions are dependent on inherent properties, and will only partially reflect any change in management that improves soil health.

Also, the lack of differences among grapevine varieties and appellation areas in biological soil health can be related to the selection of properties included in the indicator, constrained by data availability. Indicators of soil health should in fact, be sensitive to changes and, among the biological ones, the labile carbon and the protein contents (5), the microbial biomass and N forms (6), or other microbial populations or parameters (48) have been used. Unfortunately, these data are rarely available in soil databases and, therefore, even when analyzed in a soil sample, they cannot be compared to a reference population.

A more general methodological issue is related to the use of the “more is better” approach in high-quality viticulture. In this work, we clearly demonstrated that the soil characteristics associated with the red grape varieties, particularly within the DOCG areas, differ from those of the broader vineyard soils of the region. These findings indicate that optimal viticultural performance is not achieved through the maximization of individual soil properties, but rather through a balanced equilibrium between limiting and favorable conditions. This equilibrium is clearly linked to the terroir concept and, in this context, soil health should be interpreted in a functional, crop-specific manner, where moderate constraints contribute positively to grape composition. Consequently, conventional soil health assessment frameworks based on the search for high soil fertility require adaptation to better capture the complexity of terroir-driven systems. Future research should therefore focus on refining soil health indicators to account for crop-specific responses, local soil variability, and management interactions, thereby improving their applicability in high-quality viticultural environments. The substitution of the “more is better” and “less is better” approaches with the “optimum value” approach could be a first step to better link soil health with crop needs in terroir viticulture, but the determination of the optimum values may be difficult, as shown by the differences we found between DOC and DOCG areas of the same grapevine variety. On the other hand, even if optimum values were available, our results indicate that appellation areas are characterized by an equilibrium between favorable and negative soil conditions, in which intrinsic properties and management-affected ones do not simply add to each other but interact.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsoil.2026.1844190/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: José A. González-Pérez, Spanish National Research Council (CSIC), Spain

Reviewed by: Domenico Cicchella, University of Sannio, Italy

Miguel Ángel Olego, University of León, Spain