Potassium fertigation and organic fertilisation in ‘Syrah’ grape in Northeastern Brazil: yield, must characteristics and phenolic compounds

The effect of both potassium and organic fertilisation on grape yield and must composition originating from wine grapevines cultivated in the São Francisco river basin, in the semi-arid region of Northeastern Brazil, is not well known. Since soils from that region usually have low organic matter content, we hypothesise that soil application of organic fertiliser and potassium application through drip irrigation enhances the availability of this nutrient in the soil, and it influences grape yield and quality. Hence, five doses of potassium (0, 20, 40, 80 and 160 kg ha -1 ) applied through a drip irrigation system and two doses of goat manure as organic fertiliser applied into the soil (0 and 7.5 m 3 ha -1 ) after pruning were evaluated concerning yield and berry quality of grapevine ‘Syrah’ over four growing seasons (April 2013 to July 2015) in Petrolina, state of Pernambuco, Brazil. The experimental design was randomised blocks with 5 replications. We evaluated yield (t ha -1 ), weight of 100 berries (g), must volume (mL), total soluble solids (°Brix), pH and total titratable acidity (‰), K, Mg and Ca contents and phenolic compounds (total polyphenols, anthocyanins and flavonoids). The increasing application of potassium at the different doses of organic fertiliser did not alter the crop yield, except in the second season. Nonetheless, K concentration in grape must was influenced in the four growing seasons, with higher values at fitted K 2 O doses by polynomial regression of 83.62 kg ha -1 (first season), 101.25 kg ha -1 (second season), 120.00 kg ha -1 (third season) and 77.99 kg ha -1 (0 m 3 ha -1 organic fertiliser) and 96.07 kg ha -1 (7.5 m 3 ha -1 organic fertiliser) in the fourth season. The addition of increasing doses of potassium interfered with the wine grape quality, which is an important factor to be considered in relation to its acidity. Organic fertilisation did not affect the most quality characteristics of the grapes.


INTRODUCTION
Potassium (K) is the most abundant cation in the grape berry and it plays a role in berry sugar accumulation, berry maturation, berry water relations, cellular growth, disease resistance, abiotic stress tolerance, mitigating senescence and in the flow of xylem and phloem vessels (Mpelasoka et al., 2003;Brunetto et al., 2015;Haddad and Kamangar, 2015;Keller et al., 2015;Rogiers et al., 2017).In addition, K significantly influences the pH and total titratable acidity of grape must, but it has little effect on the concentration of phenolic compounds, such as anthocyanins and flavonoids (Delgado et al., 2004;Morris et al., 1987).
Grape growing develops worldwide, even within semi-arid areas such as the Lower-middle region of the São Francisco River basin in the Brazilian Northeast.Particularly for this region, it was demonstrated that grape yield is minimally impacted by K fertilisation (Silva et al., 2023).However, high levels of exchangeable potassium (K + ) in the soil can alter grape components, including total soluble solids and pH of the grape must (Ciotta et al., 2016;van Leeuwen et al., 2018) as well as to lead a magnesium deficiency in the grapevine (Ramos and Martínez de Toda, 2019).
Organic matter improves several soil attributes, but there are different sources of it for fertilisation purposes.In this context, due to the goat farming commonly performed in the semi-arid region of Brazil, goat manure is an alternative to be utilised as organic fertiliser (Rocha et al., 2015;Silva et al., 2016).This region is particularly deficient in soil organic matter due to its fast decomposition, induced by high temperatures (Assis et al., 2010) and little input of organic material (Queiroz et al., 2018).The interaction between organic matter and potassium fertiliser has not been studied for the purpose of wine grape production in that region.
The use of 30 m³ ha -1 of goat manure increased the soil phosphorus content and the cation exchange capacity and also increased the yield of grapevines 'Syrah' grown under this semi-arid condition.Concurrently, a decrease in the total titratable acidity of grape must was observed (Rocha et al., 2015).Increments in Ca, Mg, Mn and NO 3 -concentrations in the soil were also observed (Silva et al., 2016) with goat manure fertilisation and nitrogen fertigation in 'Syrah' grape.
The effects of K and manure fertilisation on the yield of wine grapes grown in the Brazilian semi-arid and its influence on grape must composition are not well known.Since soils from that region usually have low organic matter content, as reported herein, we hypothesise that soil application of organic fertiliser and potassium application through drip irrigation enhances the availability of this nutrient in the soil, and it influences grape yield and quality.Thus, this study aimed to evaluate whether different levels of K and organic fertilisation affect the yield and must composition of wine grapes 'Syrah' cultivated in a semi-arid region.

Description of the experimental area
The study was carried out in an experimental field of Brazilian Agricultural Research Corporation (Embrapa), in Petrolina, state of Pernambuco, Brazil (9° 8' 8.9" S latitude, 40° 18' 33.6" W longitude, 373 m altitude).We evaluated the grapevine 'Syrah' (Vitis vinifera L.) grafted onto Paulsen 1103 rootstock and planted in a 0.35 ha vineyard in April 2009, in a 1.0 × 3.0 m spacing between plants and rows, respectively.Plants were supported by a trellis system and grown with lateral branches with six spurs per plant and three buds per spur.The grapevines were irrigated by a drip irrigation system, with emitters spaced at 0.5 m and 2.5 L h -1 flow rate.
The gross irrigation depth (I, mm) supplied the water demand corresponding to the crop evapotranspiration (ETc, mm) (Table 1).ETc was calculated as a product of the reference evapotranspiration (ETo, mm) estimated by the Penman-Monteith method (Allen et al., 2021) and the crop coefficient (kc) for each phenological stage of the 'Syrah' grapevines grown in this experimental area (Bassoi et al., 2007).The soil of the vineyard was categorised as a Typic Plinthustalf (Soil Survey Staff, 2014) and presented a loamy sand texture, with 81, 13 and 6 g kg -1 of sand, silt and clay, respectively.Chemical and physical analyses were performed as described by scientific laboratory methods (Teixeira et al., 2017).
The study lasted for three years, comprising four consecutive growing seasons.Throughout all, weather data were recorded by an automatic station located 60 m away from the experimental field (Figure 1).
They were arranged into a randomised complete block design with five replications, assuming a split-plot design.The K 2 O and OF doses were assigned within the subplots and whole plots, respectively.The experimental unit (EU) comprised 16 plants per plot, and 8 plants per plot were considered for analysis.
The chemical analysis of the OF (Table 3) was performed according to scientific methodology (Silva, 2009).It was applied atop the soil prior to the spur pruning carried out on April 29, 2023, October 4, 2013, May 30, 2014and March 9, 2015.These dates marked the beginning of the first, second, third and fourth growing seasons, respectively.
Potassium fertilisation was performed for 10 weeks, starting one week after the spur pruning, via fertigation, using an injector pump with a flow rate of 360 L h -1 .The potassium fertilisers used were potassium chloride (K 2 O = 60 %), potassium nitrate (K 2 O = 45 %) and potassium sulfate (K 2 O = 50 %), and they were applied separately in the proportion of 50 %, 40 % and 10 % of the total K 2 O dose applied, respectively.These fertilisers were used to achieve the desired amount of potassium to be applied by fertigation.
The accompanying ions were balanced using complementary fertilisation.

Grape yield and qualitative and nutritional aspects of must
Harvests were carried out at 117 (first and second seasons), 118 (third season) and 121 (fourth season) days after pruning (dap), respectively, on August 23, 2013, January 29, 2014, September 26, 2014and July 8, 2015.The grapes were weighted on a scale (0.01 g readability) to calculate the yield (t ha -1 ) from the average weight of grapes per plant.The weight of 100 grape berries was determined on a scale with a 0.01 g accuracy; subsequently, the grapes were crushed to obtain the must and determine its volume (mL), total soluble solids (TSS) with portable digital refractometer (°Brix), pH with a digital benchtop pH meter and total titratable acidity (TTA) by titration with 0.1 N NaOH, using bromothymol blue as an indicator (Ciotta et al., 2016).
Chemical analyses were performed to determine the mineral content (K, Ca and Mg) of the grape must.This analysis was conducted using the dry method in a muffle furnace at 500 °C, with K determined by flame photometry and Ca and Mg by atomic absorption spectrophotometry (Rizzon and Miele, 2012).

Phenolic compounds analysis
To determine the total extractable polyphenols (TEP), 1.0 g of peel and 5.5 g of pulp were weighed, then 50 % methanol and 70 % acetone were added (25 mL each), followed by centrifugation (15,000 rpm) for 15 minutes after each addition of reagent, and readings were performed with the extract.Folin-Ciocalteau and gallic acid were used as standard in the UV/VIS spectrophotometer at 280 nm wavelength.TABLE 3. Chemical composition of goat manure used as organic fertiliser.
pH EC dS m -1 The calibration curve was used to determine the quantity of total polyphenols according to the absorbance reading obtained.Anthocyanins and flavonoids were determined by weighing 0.5 g of grape peel and adding a solution (25 mL) with ethanol and 1.5 N HCl; after 24-hour rest, the samples were filtered using hydrophilic cotton.Readings were performed with the extract in UV/VIS spectrophotometer at 535 and 374 nm wavelength.

Statistical analysis of results
The analysis of variance (ANOVA) was performed to account for differences between treatments, setting the significance level at 5 % (p ≤ 0.05).Shapiro-Wilk's and Bartlett's tests were adopted to check the assumptions of normality of residuals and homoscedasticity of variances, respectively.The regression analysis was performed if a significant main effect of interaction (K 2 O × OF) was found (p ≤ 0.05) or a single main effect of K 2 O.Then, the relationship between K 2 O doses and the variable of interest was characterised by linear or quadratic functions according to the significance of the regression coefficients.For a significant single main effect of OF, we performed the Student's t-test or the Welsh's t-test (if the groups do not have equal variances) to compare the means of the variable of interest.Additionally, the Principal Component Analysis (PCA) was applied to the technological and phenolic ripening and the mineral concentration in grape must obtained in each growing season.PCA was performed using PAleontological STatistics-PAST software version 4.02 (Natural History Museum/University of Oslo).

Grape yield and grape must quality
Potassium and organic fertilisation did not influence grape yield in three growing seasons, which was equal to 4.0 t ha -1 (0 m 3 ha -1 OF) and 4.2 t ha -1 (7.5 m 3 ha -1 OF) in the first season, 8.6 t ha -1 (0 m 3 ha -1 OF) and 8.2 t ha -1 (7.5 m 3 ha -1 OF) in the third season, and 2.6 t ha -1 (0 m 3 ha -1 OF) and 2.4 t ha -1 (7.5 m 3 ha -1 OF) in the fourth season.
K 2 O application significantly interfered with must pH in the first season at the different OF doses used and in the third season for the single effect of K 2 O doses.For the first season, the data were described by a quadratic regression model, with maximum fitted values of 80.00 and 114.28 kg ha -1 of K 2 O for OF doses of 0 and 7.5 m³ ha -1 , respectively.For the third season, the fitted regression model was linear, with a 0.0006 decrease in pH for each unit increase in K 2 O doses (Table 4).
For TTA (Table 5), the influence (p ≤ 0.01) of K 2 O doses was observed in all growing seasons, demonstrating the strong relationship of this variable with the increase in K 2 O availability in the soil.Interaction between OF and K 2 O was observed only in the first season.During the first growing season, in the presence of organic fertilisation (7.5 m 3 ha - 1 ), there was a linear effect of potassium fertigation on total titratable acidity (TTA), with an increase of 0.009 % per unit increase in the K 2 O doses applied, with no fit for the OF dose of 0 m 3 ha -1 .In the second growing season, there was a quadratic polynomial effect corresponding to the single effect of potassium fertigation on TTA.The fitted K 2 O dose of 103.3 kg ha -1 led to the highest value of TTA, 6.45 ‰.
In the third growing season, there was a linear effect of the K 2 O doses on TTA, with an increase of 0.00449 % per unit increment in the K 2 O dose applied.In the fourth growing season, there was a quadratic polynomial fit with the highest value of TTA, 7.06‰, at the fitted K 2 O dose of 84.5 kg ha -1 .

Nutrient and phenolic compounds content in the grape must
For K concentration in grape must (Table 7), significant effects were observed in all growing seasons.In the first season, there was a polynomial fit for the K 2 O doses, and the highest K concentration (3384 mg L -1 ) was observed at the fitted K 2 O dose of 83.62 kg ha -1 .In the second season, there was a polynomial fit, with the lowest value of K (1953.2mg L -1 ) observed at the fitted K 2 O dose of 101.25 kg ha -1 .In the third growing season, only the K 2 O dose factor influenced the concentration of K in the must, and the fit was linear with an increase of 5.66 mg L -1 for each unit increase of K 2 O.
In the fourth season, there was an interaction between the studied factors (K 2 O doses × OF doses) with the polynomial fit (p ≤ 0.05) for the concentration of K in the grape must.
For the OF dose of 0 m 3 ha -1 , there was a decrease in the concentration of K as a function of the K 2 O doses, with the lowest value (1038 mg L -1 ) reaching the fitted K 2 O dose of 77.99 kg ha -1 .For the OF dose of 7.5 m 3 ha -1 , there was an increasing polynomial fit as a function of the K 2 O doses up to the adjusted value of 96.07 kg ha -1 , with an accumulation of 3295 mg L -1 of K in the must.
The concentration of Ca in the must was not influenced (p > 0.05) by the treatments in any of the growing seasons, showing mean values of 27.02 (first season), 17.32 (second season), 37.66 (third season) and 124.12 mg L -1 (fourth growing seasons).
The concentration of Mg in the must (Table 8) was influenced by the interaction of treatments (K 2 O and OF doses) in the first growing season and by the single factor K 2 O doses in the fourth growing season.For the first season, the fit was quadratic, with the highest values (39.37 and 42.01 mg L -1 at OF doses of 0 and 7.5 m 3 ha -1 , respectively) observed at fitted K 2 O doses of 57.17 kg ha -1 (0 m 3 ha -1 ) and 72.85 kg ha -1 (7.5 m 3 ha -1 ).In the fourth growing season, Mg concentration was influenced by the K 2 O dose factor, with the quadratic polynomial fit and the highest value (30.61 mg L -1 ) at the fitted K 2 O dose of 120.00 kg ha -1 .
K 2 O and OF application did not significantly affect the phenolic compounds, except in the first growing season for anthocyanins (p ≤ 0.05), assuming the interaction between K 2 O and OF doses and in the third growing season for flavonoids in response to K 2 O doses (p ≤ 0.01) as a single factor.For the concentration of anthocyanins (Table 9), for  the OF dose of 7.5 m³ ha -1 , a quadratic fit was observed with maximum values at the K 2 O dose of 82.61 kg ha -1 .For the concentration of flavonoids, the fit was linear, with a decrease of 0.203 mg 100 g -1 for each unit increase in K 2 O doses.

Organic fertiliser effect
We also observed the single main effect of organic fertilisation on certain 'Syrah' wine grapes traits at harvest.Significant differences were found between the means of yield (p ≤ 0.01) and must volume (p ≤ 0.01) at the second growing season.Therefore, the Student's t-test showed that the goat manure fertilisation promoted a decrease in yield (Figure 2A) but an increase in must volume (Figure 2B).The flavonoid content was only affected by the organic fertilisation in the third growing season (p ≤ 0.01).According to the Student's t-test, the mean flavonoid content in grape must of unfertilised plants was higher than in fertilised plants (Figure 2C).The mineral content in grape must was also influenced by the organic fertilisation alone and we noticed this single main effect on Ca concentration in the second (p ≤ 0.01) and third growing season (p ≤ 0.05) and on Mg concentration in the fourth growing season (p ≤ 0.01).Thus, the Student's t-test showed that the mean Ca concentration was higher in grape must of unfertilised plants than in fertilised plants (Figures 2D and 2E).Nevertheless, Welsh's t-test indicated that the organic fertilisation resulted in an improvement in Mg mean concentration (Figure 2F).

Principal Component Analysis (PCA)
Based on the results of the PCA, the first two principal components (PC1 and PC2) were considered in each of the growing seasons (Figure 3), which accounted for more than 55 % of the explanation of the total variance.For the first growing season (Figure 3A), 55.14 % of the data variance was explained by the first two components, of which 34.67 % was explained by PC1 and 20.47 % was explained by PC2.
The variables K, flavonoids, anthocyanins, TSS, TTA, pH showed a high contribution in the variation of PC1, while the nutrients Ca and Mg (highly) and the total polyphenols (TEP) contributed to the variation of PC2, leading to the formation of two groups: the first formed by the K 2 O doses of 20 kg ha -1 (7.5 m³ ha -1 ), 40 kg ha -1 (0 and 7.5 m³ ha -1 ) and 80 kg ha -1 (7.5 m³ ha -1 ) with the greatest influence on the variables pH, TSS, K, flavonoids and anthocyanins; and the second by the K 2 O doses of 0 and 80 kg ha -1 , with greater influence of 0 m³ ha -1 on the variables Ca, TTA and TEP.In addition, the nutrients Ca and Mg were negatively correlated with treatments with the highest applied dose of K 2 O (160 kg ha -1 ), regardless of the organic fertiliser; that is, when grape plants were fertilised with high levels of K 2 O, there was a reduction of these nutrients (Ca and Mg) in the berry.
In the second season (Figure 3B), 61.89 % of the total variance of the data was explained by the first two components together, 34.84 % by PC1 and 27.05 % by PC2.The variables TEP, pH, TSS, Ca and Mg had a greater contribution to PC1, while anthocyanins, flavonoids, TTA and K had a greater contribution to PC2.An important highlight was observed in this season regarding the OF factor, differing from that observed in the univariate analysis, where distinct groups of data without (0 m 3 ha -1 ) and with (7.5 m 3 ha -1 ) organic fertiliser were identified, regardless of the levels of K 2 O, separated mainly by the axis of PC2.In addition, treatments with higher levels of K 2 O (40 and 80 kg ha -1 of K 2 O applied) and with OF were associated with higher TTA values.The addition of OF and K 2 O in general, especially at the highest doses, increased the pH of the must.
In the third season (Figure 3C), the first two principal components explained 63.79 % of the total variance of the data, 39.33 % for PC1 and 24.46 % for PC2.The variables TEP, anthocyanins, TTA, K and Mg had a greater contribution to PC1, while the variables pH, soluble solids and Ca had a higher contribution to PC2.As in the second season, but in a more evident way, there was separation of groups in relation to the use of OF; the presence of OF promoted an increase in pH and reduction of flavonoids, TSS and Ca in the must at most doses studied.The addition of 160 kg of K 2 O ha -1 increased TTA and the concentrations of K and Mg but led to musts with lower concentrations of anthocyanins and TEP.These variables were found at higher concentrations in grape musts from treatments with lower doses of K 2 O and in which OF was not used.
For the fourth season (Figure 3D), the total variance of the data, explained by the first two components, was 58.8 %, with PC1 and PC2 explaining 33.42 % and 25.38 %, respectively.The application of 80 and 160 K 2 O ha -1 together with organic fertiliser increased K concentrations in the must and, on the other hand, reduced the levels of TEP, TTA and pH, which were negatively correlated with the concentration of K in the must.

DISCUSSION
Excess potassium caused significant differences in the pH of Cabernet-Sauvignon grapes (Morris et al., 1987), as observed in this study in the first and third growing seasons (Table 4), while no effect of K on the pH of Cabernet-Sauvignon grape must was observed (Poni et al., 2003) as reported herein for second and fourth seasons (Table 4).As 50 % of the total K absorbed is accumulated in the berries, the chemical composition of the must, by excess K, can alter other components such as total titratable acidity (Delgado et al., 2004;Morris et al., 1987;van Leeuwen et al., 2018), without influencing crop yield (Silva et al., 2023) mainly in soils with high concentrations of K.The increase in soil potassium caused an increase of 0.047 in must TTA in 'Cynthiana' wine grapes due to the increase in total soluble solids (Morris et al., 1987).The same behaviour has been observed (Karimi, 2017), i.e., an increase in TTA values due to the addition of K 2 O.Such information shows that the increase in TTA caused by the increment of K 2 O in the soil in the present experiment (Table 5) results from the addition of potassium, being observed in the four growing seasons.TSS showed no differences related to potassium fertilisation (Morris et al., 1987) as observed in the first, second, third (without OF) and fourth seasons, while the only response to potassium doses was together OF in the third season (Table 6).
The increase of K concentration in grapes with the addition of K 2 O in the soil (Table 7) is corroborated by other studies with 'Cabernet-Sauvignon' (Poni et al., 2003) and for 'Zweigelt' (Griesser et al., 2017), where the berries contain one of the highest concentrations of K among the parts of the plant.Another study (Howell and Conradie, 2013) reported a higher concentration of K in the composition of the must (2600 mg L -1 ), with frequent applications of K 2 O via fertigation (91.9 kg ha -1 ).More than 50 % of the total K absorbed by plants is accumulated in the berries, and its functions are enzymatic activation and synthesis of reactions, contributing to fruit maturation (Brunetto et al., 2015).Positive correlations between the soil extractable K and the foliar K, the K in the must and the K in the wines were reported by Martínez-Vidaurre et al. (2023).
Although Ca is an essential element for the structure of grape berries, being responsible for the rigidity of their cell wall (Brunetto et al., 2015), no response of grape Ca concentration was found for both treatments.On the other hand, there was a response in Mg concentration in grapes and the variation of K concentration in the plants in the first, second and fourth seasons (Table 8).Higher Mg values in the juice of grapes produced with a K application of 91.9 kg ha -1 were reported (Howell and Conradie, 2013) as well as the variation of K + in the plants altered the concentration of Mg in the grapes (Griesser et al., 2017).The increase in Mg concentration improves chlorophyll synthesis in the plant, but the increase in concentration in grape berries is possibly related to their maturation and the high absorption of K in this process (Brunetto et al., 2015).
Differences in the values of anthocyanins and flavonoids were found only, respectively, in the first and third seasons (Table 9).The values of anthocyanin and phenolic compounds observed in this experiment were higher than in other studies (e.g., Bassoi et al., 2007) with different irrigation managements for the same variety and under the same edaphoclimatic conditions.Possibly, potassium fertilisation has a greater influence on the phenolic compounds evaluated than water deficit does, thus increasing their concentration in the grape berries.
No differences were observed in relation to yield and must volume in three growing seasons, but observed an evolution of different elements in the soil as organic matter, phosphorus content and cation exchange complex (Rocha et al., 2015).Increasing Ca, Mg, Mn and NO 3-in the same region were reported in a previous study carried out over three growing seasons from April 2010 to September 2011 with goat manure and nitrogen fertigation (Silva et al., 2016).OF applied in this experiment was probably not sufficient to change yield, must volume, flavonoids, calcium and magnesium content (Figure 2).Especially with respect to the maintenance of organic matter content in the soil, soil cultivation has caused significant losses in the irrigated Brazilian semi-arid area, being more expressive in annual cropping.It is highly recommended that cropping systems consider more effective cultivation control strategies, such as crop rotation, favouring the addition of plant residues to the soil for effective soil organic matter recycling (Assis et al., 2010).The amount of organic matter in the soil of Brazilian semi-arid zones is usually low, and the use of management systems that aim their accumulation and maintenance at appropriate levels will contribute to the increase in the CEC, as well as to improve their chemical and physical properties (Queiroz et al., 2018).
In general, the Ca content was negatively correlated with the values of pH; on the other hand, it was positively correlated with the levels of anthocyanins and flavonoids in most seasons.These effects may be related to the addition of K 2 O in the soil, which is a determinant for increasing the values of most parameters evaluated (Walker and Blackmore, 2012).The relationships between the variables evaluated herein have been observed in several other studies involving potassium fertilisation (Ciotta et al., 2016;Walker and Blackmore, 2012) and water deficit (Vila Nova et al., 2019) in grapes, despite the differences in climate and soil fertility of the different sites.
Harvests were performed in August 2013, January and September 2014 and July 2015.Possibly, seasonality influenced the variability of the studied variables, which was observed in the correlation between Ca and TEP and the concentration of K in the must.A study demonstrated the effect of seasonality in the São Francisco Valley region, where phenolic components of wine, such as flavonoids, ranged from 1150.0 to 2142.2 mg L -1 between 2014 and 2016 for 'Syrah' grapes (Oliveira et al., 2019).In another study carried out in the same region (Oliveira et al., 2020), the intra-annual climate variability influenced the composition of grapes and wines as grape harvesting occurs throughout the year.

FIGURE 1 .
FIGURE 1. Precipitation (P) and air temperature (maximum: T max ; mean: T mean ; and minimum: T min ) data recorded during the first and second growing seasons (A) and third and fourth growing seasons (B) of 'Syrah' grape in Petrolina.

FIGURE 2 .FIGURE 3 .
FIGURE 2. Mean and standard deviation bars of wine grape traits for each organic fertiliser rate.* and ** indicate a difference at the 0.05 and 0.01 significance level according to Student's t test or Welsh's t test.

TABLE 1 .
Reference Table 2 presents the results of the soil chemical analysis performed in January 2013.

TABLE 2 .
Chemical analysis of the Typic Plinthustalf in the 0-0.20 m and 0.20-0.40m layers.

TABLE 4
. pH in 'Syrah' grapes as function of different doses of potassium (K 2 O) and organic fertiliser (OF).ns=not significant * significant at 5 % probability of error.Alexsandro Oliveira daSilva et al.

TABLE 5 .
Total titratable acidity (TTA) in 'Syrah' grape as function of different doses of potassium (K 2 O) and organic fertiliser (OF).

TABLE 6 .
Total soluble solids in must (TSS), volume of must (V) and mass of 100 berries (M) of 'Syrah' grape as function of different doses of potassium (K 2 O) and organic fertiliser (OF).

TABLE 7 .
Potassium concentration in 'Syrah' grape as function of different doses of potassium (K 2 O) and organic fertiliser (OF).

TABLE 8 .
Magnesium concentration in 'Syrah' grape as function of different doses of potassium (K 2 O) and organic fertiliser (OF).
ns = not significant * significant at 5 % probability of error.Alexsandro Oliveira da Silva et al.

TABLE 9 .
Concentration of anthocyanins and flavonoids in 'Syrah' grape as function of different doses of potassium (K 2 O) and organic fertiliser (OF).