The experiment was conducted in a commercial vineyard in northeastern Spain (Aldeanueva de Ebro, La Rioja) over three consecutive years (2020, 2021 and 2022). The vineyard design was 2.6 m x 1.2 m and 3205 vines per hectare of Tempranillo cultivar spur-pruned in a bilateral Royat Cordon system and grafted on R-110 rootstock. The vineyard had previously been managed under Spanish regulation for integrated pest management (IPM), with the application of herbicide to avoid weed presence and of drip water irrigation.

As Mairata et al. (2024) have explained, all soil management treatments were established in February 2019 and reapplied annually between March and April. These treatments were applied to a strip that was up to 25 cm wide on each side of the vine row ( i.e., 50 cm in total), and comprised three organic mulch substrates: grapevine pruning debris (GPD) from previous years, straw mulch (STR) composed of wheat ( Triticum sp) and sourced from the Government of La Rioja (Logroño, Spain), and spent mushroom compost (SMC) derived from mushroom ( Agaricus bisporus) and containing a rich composition of animal manure and urea provided by “Sustratos de La Rioja SL.” Two conventional soil management practices were examined: under-row tillage (TILL) and herbicide application (HERB). The herbicide, only applied in the HERB treatment, comprised Terafit (25 %, Flazasulfuron) and Atila (36 %, Glyphosate) at a rate of 100 litres per hectare. The experimental design followed a randomised complete block divided into experimental plots, with each soil management treatment consisting of three plots containing 40–50 vines each.

The regional laboratory of the Government of La Rioja analysed the composition of the three organic mulches (Table 1) and the soil samples (Table 2). SMC mulch had an ash content of 48.1 %, higher than that of the STR (5.7 %) and GPD (3.7 %) mulches. The organic matter content of the STR and GPD mulches was approximately 95 %, whereas that of SMC was 51.9 %. However, SMC exhibited higher nitrogen concentration (2.38 %) than STR (0.77 %) and GPD (0.88 %). Moreover, the C/N ratio of SMC mulch was lower (12.6) than that of STR (71.3) and GPD (63.6). The concentrations of other essential elements, such as phosphorus (P), magnesium (Mg) and calcium (Ca), were up to three times higher in SMC than those measured in GPD and STR.

Table 1. Physical and chemical properties of the organic mulches based on spent mushroom compost (SMC), straw (STR) and grapewine pruning debris (GPD) were analysed by the regional laboratory of the Government of La Rioja (Logroño, Spain).

ppm: parts per million

The study region is characterised by a Mediterranean climate with warm and dry summers (Table S1). In addition, soil temperature (°C) and volumetric soil water content (%) were recorded in each plot (n = 3) every 30 min at three soil depths (5, 15 and 25 cm) with Sentek equipment (Sentek Pty Ltd., Stepney, Australia) and “Drill & Drop” probes ( Pou et al., 2021). The equipment was installed in October 2020, and soil data were recorded in 2021 and 2022. Before the establishment of the soil treatments in December 2018, soil texture and nutritional composition were analysed (Table 2) in ten shallow soil samples (0-30 cm) distributed throughout the vineyard. Each soil sample was a mixture of ten independent drill samples. Furthermore, soil composition was examined at the end of the experimental study (2022) in all plots following the same methodology applied throughout the experiment. The soil samples were dried at 40 °C for one week and sent to the regional laboratory of the Government of La Rioja for analysis of soil texture (sand, silt and clay), pH, electrical conductivity (1:5; soil/water using a conductivity metre), organic matter, macro-nutrients (NPK), oligo-nutrients (Mg and Ca), micro-nutrients (Fe, Mn, Zn and Cu), oligo-nutrients (Mg, Ca, SO4) and Na as described in Mairata et al. (2023).

Table 2. Texture and general properties of the soil composition of the field (Aldeanueva de Ebro, Spain) at the beginning (2018) and at the end of the field experiment in each soil management treatment (2022).

ppm: parts per million

Grapevine and soil water status, leaf gas exchange and whole-plant hydraulic conductance were analysed in the same six plants (n = 6) per soil treatment at four key stages: flowering, fruit set, veraison and ripening. Fruit set analyses in 2020 were not conducted. Midday and predawn leaf water potentials (Ψ _leaf and Ψ _soil, respectively) were measured in one leaf per plant (n = 6) using a Scholander pressure chamber (Soil Moisture Equipment Corp., Santa Barbara, CA, USA). On the same day, net photosynthesis rate (A _N), stomatal conductance (g _s) and transpiration rate (E) were measured in one sun-exposed leaf per plant (n = 6) using a portable open gas exchange system (Li- 6400; Li-Cor Inc., Lincoln, NE, USA) at a CO ₂ concentration of 400 mmol CO ₂/mol air. Leaf gas exchange measurements were conducted on sunny days from 10:00 to 12:00. Intrinsic water use efficiency (WUE _i) was calculated as the ratio of A _N to g _s ( Mairata et al., 2022; Pou et al., 2022). Whole-plant hydraulic conductance (K _plant) was calculated using the method described by Nardini et al. (2000), which is based on Ohm’s law analogy for continuous the soil-plant-atmosphere flow: K _plant = E _max/(Ψ _soil – Ψ _leaf), where E _max is the maximum daily transpiration rate.

The carbon isotope ratio (δ ¹³C) was determined from 50 mixed berries per plot collected at harvest. Three grape samples, one per plot, and three instrumental replicates of each one were analysed. The berry samples were oven-dried without seeds ( Mairata et al., 2022) and powdered in an ultra-centrifugal mill (ZM1, Retsch, Haan, Germany). Aliquots of 2 ± 0.1 mg of grape powder were combusted in an elemental analyser (Thermo Flash EA 1112 Series, Bremen, Germany), separated by chromatography and directly injected into a continuous-flow isotope ratio mass spectrometer [Thermo Finnigan Delta Plus, Bremen, Germany] ( Bchir et al., 2016). Carbon isotope discrimination was expressed as δ ¹³C = [(Rs-Rb)/Rb] x 1000 (Farquhar et al., 1984), where Rs and Rb are the ¹³C/ ¹²C ratio of grape sample and PDB (PeeDee Belemnite), respectively.

At harvest, the yield mass, number of clusters and cluster weights of 36 plants per plot uniformly distributed along the row were recorded. At pruning, the annual shoot mass per vine, the number of shoots per vine and the average shoot weight were determined in six representative plants per experimental plot of each soil management treatment. Moreover, the Ravaz index was calculated as the ratio between the yield and the pruning mass ( Pou et al., 2022). These variables were recorded annually from 2020 to 2022 except for pruning data in 2021, which was unavailable due to prior pruning carried out by the vineyard owner.

Vine vigour was assessed by canopy reflectance using Crop-Circle ACS-430 [Holland Scientific, Inc., Lincoln, NE, USA] ( Pou et al., 2022; Taskos et al., 2015). Measurements were carried out one metre from the canopy and in one go: ten measurements per second, amounting to 35 seconds for the entire plot. The ACS-430 sensor recorded the reflectance of an active light sensor at 630 nm (red), 730 nm (red-edge) and 780 nm (NIR). From the reflectance values, NDVI (normalised difference vegetation index) and NDRE (normalised difference red edge) were calculated according to Taskos et al. (2015):

NDVI = (NIR – red)/(NIR + red)

NDRE = (NIR – red edge)/(NIR + red edge)

At flowering and veraison, 30 healthy leaves (one per plant) from fruit-bearing shoots were sampled from each plot to determine total leaf blade nitrogen content (organic and inorganic). At flowering, leaves from one side of the first bunch were collected, and at veraison, leaves from the opposite side were collected following the methodology outlined by Romero et al. (2013). The sampling was carried out on alternate sides of the trellis. The collected leaf blades were washed three times, rinsed with distilled water and oven-dried at 70 °C for 48 hours. Subsequently, they were ground using an ultra-centrifugal mill (ZM1, Retsch, Haan, Germany) and sieved through a mesh (< 0.5 mm). Two hundred milligrams of ground sample was analysed following the Dumas method ( Etheridge et al., 1998) and using a Leco CNS analyser (St. Joseph, MI, USA). The concentrations reported are equivalent to grams of nitrogen per 100 grams of dry matter (g N/100 g dry matter).

From 2020 to 2022, grape and must quality variables were assessed via random berry sampling. In the three replicates of the five soil management treatments, 500 random berries were sampled within the entire row to analyse physical, chemical and oenological variables. Total soluble solids (TSS) were analysed using a refractometer, and pH, total acidity (TA), malic acid and potassium were determined using EEC methods. Tartaric acid in the berry juice was measured using the Rebelein method. All the methodologies for these measurements, including yeast assimilable nitrogen (YAN), have been described by Portu et al. (2017). Following the methods described by Pou et al. (2022), two hundred berries were weighed and extracted twice using 50 mL of 1 % HCl in a mixer. The mixture was heated to 40 °C while being stirred. An additional 1 % HCl was introduced, followed by further heating and stirring to 60 °C. The mixture was subsequently cooled to 10 °C using an ice bath and filtered through a cloth, and the extract volume was recorded. Following dilution, total phenolics were measured as the total polyphenol index (TPI) using spectrophotometric absorbance at 280 nm and determined by a Helios Omega spectrometer (Thermo Fisher Scientific). Total anthocyanin was measured by decolorising with sulphur dioxide.

The phenological stages were analysed independently to preserve information through the vine vegetative cycle. Exploratory data analysis was conducted to assess data normality through Q-Q plots. For all variables, an initial analysis of variance (ANOVA) was carried out using a generalised linear model (GLM) to detect differences among independent variables (soil management treatment, year, and phenological stage), as well as their interactions. Subsequently, a linear mixed model (LMM) was employed for variables without interaction between treatment and year, treating the experimental year as a random variable. A linear model (LM) was conducted when interactions were observed within each subgroup. When differences among the various soil management practices were identified, a post-hoc Tukey analysis was conducted to enable multiple comparisons between soil management treatments. Detailed statistical results are provided in Tables S2 to S5. Data analysis was performed using RStudio software (version 4.3.1), and graphs were generated by GraphPad Prism (version 8.0.1). Any statistical significance was accepted with a p < 0.05.

Climatic data for the critical grapevine growth period (May to September) is summarised in Table S1. The year 2022 is characterised by warmer and drier conditions, with accumulated precipitation of 89.6 mm and growing degree-days (GDD) of 1924.4 °C. By contrast, in 2020 and 2021, higher precipitation levels of 117.4 mm and 142 mm were recorded, respectively, with corresponding GDD values of 1684 °C and 1587 °C.

Table 2 provides an overview of the physical and chemical soil properties before the application of the soil treatments and at the end of the experimental assay (2022). The texture analysis revealed a soil composition of 50.1 % sand, 32.7 % silt and 17.3 % clay, indicating that it is loamy to sandy-loam soil with a significant amount of sand particles. The soil management treatment did not alter the soil structure. The initial soil pH was alkaline (8.2), with an electrical conductivity (CE) of 0.16 dS/m. The C/N ratio of 9.4 indicates a balanced relationship conducive to microbial activity and organic matter decomposition despite the low percentage (1.1 %). The nutrient analysis revealed a relatively low nitrogen (N) content (0.84 ppm; parts per million), moderate phosphorus (P) levels (up to 51 ppm) and a relatively high concentration of potassium (K) concentration (254.3 ppm). Sodium (Na) content was relatively low (10.8 ppm), and Magnesium (Mg) and calcium (Ca) were abundant (reaching 275.1 ppm and 21400.8 ppm, respectively). Iron (Fe) concentration was moderate (44.7 ppm), and manganese (Mn) was 53.1 ppm. Finally, Zinc (Zn) and copper (Cu) concentrations were moderate (3.7 and 7.6, respectively). After four years of the experiment, soil properties of the SMC treatment exhibited disparities relative to the other soil treatments. Specifically, the SMC treatment decreased the soil pH to 8.11, whereas the other treatments increased the pH to 8.46. Moreover, SMC increased the soil’s electrical conductivity to 0.52 dS/m and enhanced levels of organic matter and nutrients such as N, K and Zn.

Figure 1. Volumetric soil water content (VWR) at 5 cm (A), 15 cm (B) and 25 cm (C) of the soil profile and accumulated daily precipitation (mm) of the soil management treatments from the end of May (DOY 140) to the end of September (DOY 260) in 2021 and 2022 (Aldeanueva de Ebro, Spain). The mean of the soil probes (n = 3) and the standard deviation are shown. The vertical black lines represent the phenological stages at flowering (F), fruit set (S), veraison (V) and ripening (R).

Throughout 2021 and 2022, soil water content (Figure 1) and soil temperature (Figure S1) were consistently monitored at three depths (5, 15 and 25 cm) from mid-May to late September (indicated by day of the year, DOY), aligning with the main grapevine phenological stages: flowering (F), fruit set (S), veraison (V) and berry ripening (R). Daily precipitation (Figure 1) and air temperature (Figure S1) are also represented in the figures. Soil water content (Figure 1) remained relatively stable across both years. Specifically, water content at 25 cm depth (Figure 1C) exhibited fewer fluctuations and minimised differences between treatments compared to shallower layers. In particular, the 5 cm depth (Figure S1) displayed significant oscillations throughout the growing cycle.

Differences in soil water content were observed over both years (Table S6). STR and SMC mulching treatments resulted in higher moisture levels. Moreover, the SMC treatment also enhanced water retention capacity during precipitation events. At a depth of 25 cm, notable changes in water content were observed whenever precipitation exceeded 10 mm. Nonetheless, on two occasions (DOY 212 in 2021 and DOY 175 in 2022), a substantial increase in soil water content was observed due to drip irrigation inputs, which mitigated differences between the water content levels of the analysed soil treatments. Temperature data (Figure S1) followed a similar trend to water content. Soil temperatures across the soil profile were generally consistent, though more variation was noted among sensors at 5 cm depth. Organic mulches, especially STR, effectively reduced soil temperatures by up to 5 °C compared to conventional HERB and TILL treatments during the year's hottest days.

Significant interactions were observed in soil water potential (Ψ _soil), leaf water potential (Ψ _leaf), plant conductivity (K _plant), photosynthesis rate (A _N), stomatal conductance (g _s) and intrinsic water use efficiency (WUE _i) between phenological stages, years and soil management treatments. For this reason, differences in soil management treatments were analysed independently for each phenological stage (flowering, fruit set, veraison, and ripening) in 2020 (Table 3), 2021 (Table 4) and 2022 (Table 5).

Table 3. Mean values of midday (Ψ _leaf) and pre-dawn (Ψ _soil) water potentials, whole plant hydraulic conductance and leaf gas exchange measures at flowering, fruit set, veraison and ripening (cv. Tempranillo) 2020. For each dependent variable, different letters indicate statistically significant differences. Statistical differences were accepted when p < 0.05.

Ψ _leaf = leaf water potential; Ψ _soil = leaf water potential; K _plant = plant hydraulic conductivity; A _N = net photosynthesis rate; g _s = stomatal conductance; WUE _i = intrinsic water use efficiency.

In general, Ψ _leaf and Ψ _soil showed higher values in the STR mulch treatment. By contrast, the SMC treatment exhibited more variability, with lower soil water potentials at flowering and fruit set in 2021 and at veraison and ripening in 2022. However, both STR and SMC achieved higher soil water potentials at veraison and ripening in 2021. The GPD, HERB and TILL treatments showed similar soil and leaf water potentials. Soil water potential values remained relatively stable between -0.3 MPa and -0.5 MPa across most phenological stages, except at veraison and ripening in 2022, when they decreased to -0.7 MPa. This trend was also observed in leaf water potentials with higher variability, generally oscillating between -1 MPa and -1.6 MPa.

Table 4. Mean values of midday (Ψ _leaf) and pre-dawn (Ψ _soil) water potentials, whole plant hydraulic conductance and leaf gas exchange measures at flowering, fruit set, veraison and ripening (cv. Tempranillo) in 2021. For each dependent variable, different letters indicate statistically significant differences. Statistical differences were accepted when p < 0.05.

Ψ _leaf = leaf water potential; Ψ _soil = leaf water potential; K _plant = plant hydraulic conductivity; A _N = net photosynthesis rate; g _s = stomatal conductance; WUE _i = intrinsic water use efficiency.

Plant hydraulic conductivity (K _plant), calculated from transpiration and soil/leaf water potentials, showed scattered differences between soil management treatments without a clear behavioural trend. Leaf gas exchange values (A _N, g _s and WUE _i) differed depending on the phenological stage and year. In 2020, differences were only observed during ripening. Organic mulch treatments (SMC, STR and GPD) exhibited higher photosynthesis rates than conventional management practices (HERB and TILL). GPD showed the highest stomatal conductance values (0.21 mol H ₂O/m ²⋅s), while HERB showed the lowest (0.12 mol H ₂O/m ²⋅s). Additionally, SMC showed higher WUE _i values than GPD. In 2021, differences between all the phenological stages were recorded. During flowering and fruit set, SMC exhibited lower photosynthesis and stomatal conductance values but higher WUEi than the other treatments. However, at veraison and ripening, SMC and STR (only at veraison) showed higher photosynthesis and stomatal conductance values, thus reducing the WUE _i ratio. In 2022, STR showed higher stomatal conductance ratios at flowering, while SMC showed lower ratios, inversely affecting the WUE _i ratio. During the fruit set, SMC and HERB showed higher stomatal conductance ratios. SMC achieved a higher WUE _i ratio due to a higher photosynthesis ratio. At version, STR had higher photosynthesis and stomatal conductance values than other treatments, without differences in WUE _i. Carbon isotopic discrimination (δ ¹³C) data from Table S7 revealed no differences between soil management treatments in any of the analysed years.

Table 5. Mean values of midday (Ψ _leaf) and pre-dawn (Ψ _soil) water potentials, whole plant hydraulic conductance and leaf gas exchange measures at flowering, fruit set, veraison and ripening (cv. Tempranillo) in 2022. For each dependent variable, different letters indicate statistically significant differences. Statistical differences were accepted when p < 0.05.

Ψ _leaf: leaf water potential; Ψ _soil: leaf water potential; K _plant: plant hydraulic conductivity; A _N: net photosynthesis rate; g _s: stomatal conductance; WUE _i: intrinsic water use efficiency.

Table 6 summarises three years of data for the production and vegetative growth per plant, including yield weight (YW), average cluster weight (CW), number of clusters per vine (CV), pruning weight (PW), average shoot weight (ASW) and number of main shoots (NMS). Additionally, it includes the Ravaz Index (RI), which represents the ratio of harvest weight to pruning weight. Two-way ANOVA analysis highlights differences between soil management treatments ( p-value < 0.05). The SMC treatment exhibited higher YW and PW values than the other soil treatments, except for GPD in PW. This difference was reflected in increased CW and ASW. However, no differences were detected between the soil management treatments for CV, NMS and RI.

Table 6. Mean values (2020-2022) of harvest and pruning data (cv. Tempranillo) of the five soil managements analysed. Significant differences between soil management treatments were represented when p < 0.05.

YW = yield weight (kg/vine); CW = cluster weight (kg); CV = clusters/vine; PW = pruning weight (kg/vine); ASW = average shoot weight (g); NMS = number of main shoots/vine; RI = Ravaz Index (yield/pruning mas).

Table 7 shows the NDVI (normalised difference vegetation index), NDRE (normalised difference red edge), and leaf nitrogen content (N) at flowering and veraison over the three analysed years. No interactions between soil management treatment and years were observed in the two phenological stages analysed. At flowering, no differences in NDVI were detected. However, NDVI and NDRE rates increased at veraison compared to flowering, while nitrogen concentration decreased. For the flowering stage, the SMC treatment obtained higher NDRE values than STR and higher nitrogen amounts than STR and GPD. More pronounced differences between the soil treatments were observed during veraison, with variations in all the variables: specifically, the SMC treatment exhibited higher NDVI, NDRE and leaf nitrogen values than STR, GPD, and HERB.

Table 7. Mean values (2020-2022) of normalised difference vegetation index (NDVI), normalised difference red edge (NDRE), and leaf blade nitrogen content (N, g N/100 g dry matter) of the five soil management treatments at flowering and veraison (cv. Tempranillo). Significant differences between soil management treatments were represented when p < 0.05.

Phenology	Treatment	NDVI	NDRE	N
Flowering	SMC	0.757	0.252 a	3.63 a
	STR	0.731	0.218 b	3.02 b
	GPD	0.741	0.232 ab	3.11 b
	HERB	0.733	0.225 ab	3.20 ab
	TILL	0.738	0.235 ab	3.30 ab
Veraison	SMC	0.820 a	0.350 a	2.42 a
	STR	0.783 b	0.278 b	1.89 b
	GPD	0.789 b	0.292 b	2.00 b
	HERB	0.790 b	0.294 b	2.08 b
	TILL	0.799 ab	0.314 ab	2.15 ab

Grape yield and must quality parameters were evaluated over the three years, as shown in Table 8. The SMC soil treatment recorded the highest pH values (3.65), followed by the GPD treatment (3.59). The STR, HERB and TILL treatments showed the lowest pH values, ranging from 3.51 to 3.49. Soil management treatments affected malic acid content, with SMC having the highest concentration (1.8 g/L), and HERB (1.16 g/L) and TILL (1.22 g/L) the lowest amounts. Potassium levels in the must differed depending on the soil treatment, with SMC again showing the highest amounts (1442.9 mg/L), and HERB (1168.5 mg/L) and TILL the lowest (1166 mg/L). Must yeast assimilable nitrogen (YAN) levels also differed among treatments, with SMC showing the highest concentration (210.3 mg/L), followed by the TILL treatment (174.4 mg/L), and the STR treatment with the lowest value (113.1 mg/L). The STR and GPD treatments showed the highest berry mass (226.4 g/100 berries and 221.2 g/100 berries, respectively), and HERB and TILL the lowest (202.8 g/100 berries and 202.7 g/100 berries, respectively). Anthocyanin content was lower with the SMC treatment (1.23 mg/L) than with the TILL treatment (1.46 mg/L). No significant differences were found between soil management treatments in total acidity (TA), tartaric acid (TAR) and total polyphenol index (TPI).

This three-year study aimed to evaluate the impact of alternative soil management strategies based on organic mulches on vine development, plant physiology and grape composition. The discussion is intricate due to the multiple variables studied, including soil water content and temperature, yield and pruning mass, leaf nutritional content, leaf physiological variables, soil and plant water potentials, isotopic carbon discrimination in grapes, plant reflectance parameters and the assessment of grape and must quality.