Grapevine Phenology, Vegetative and Reproductive Characteristics of Vitis vinifera L. cv Chardonnay in the Cape South Coast Region in South Africa

Abstract

Climate change necessitates the exploration of new, cooler viticultural regions globally. Chardonnay is an early ripening variety which is subjected to temperature extremes. This study aimed to investigate the response of Chardonnay in cool climatic regions in the Cape South Coast region of South Africa over two growing seasons in 2021–2022 and 2022–2023 in three commercial vineyards. An evaluation of the climatic, vegetative and reproductive characteristics was performed. Seasonal variation was the biggest driver of the Growing Degree Days (GDD) at the sites. Overall, the 2021–2022 season was warmer than the 2022–2023 season, but the microclimatic conditions were impacted by the cultivation practices which were applied. The canopy density and total leaf surface varied between the different sites (p < 0.01) and by season × site (p < 0.05). Site and the site × season interaction were the main drivers of the environmental conditions and cultivation practices. Canopy characteristics impacted the sugar accumulation rate over the two seasons. Grape berry transpiration was impacted by the environmental conditions at the sites. Chemical composition varied with soil depth. From the results of our study, although Chardonnay is suitable for cultivation in the Cape South region, site-specific conditions impact fruit development and the quality at harvest.

1. Introduction

Globally, Chardonnay is one of the most economically important grape cultivars with over 210,000 hectares planted [1]. Overall, it is a very versatile cultivar predominantly used to produce champagne (France), bottled fermented sparkling wines in other wine-producing countries and regions (Italy—Franciacorta; South Africa—Cap Classique), and high-quality table wine. Most Chardonnay plantings are found in France (51%) and the United States of America (USA) (31%), with smaller proportions in other wine-producing countries [2]. Chardonnay is the cornerstone of Champagne and Burgundy and is used to make a wide variety of champagne and table wine styles, ranging from lean and crisp styles with judicious oak use to creamy, buttery and oak-aged versions. Within the South African context, Chardonnay is the fourth most-planted white grape cultivar (13%), covering 6685 hectares of the total plantings [3]. These Chardonnay plantings are predominantly found in the Robertson, Stellenbosch and Paarl wine regions. More recently, there has been a shift in the plantings from traditional regions to cool-climate areas of South Africa such as the Walker Bay and Overberg districts and the Hemel-en Aarde Valley and Elgin wards. This shift is mainly driven by global warming, which leads to the classic Chardonnay regions being unable to provide sufficient chilling requirements for dormancy release, consequently causing delayed or uneven budbreak [4]. Cooler conditions are more favorable for slow ripening, which facilitates the making of terroir specific wines.

Despite the versatility of Chardonnay, its cultivation in emerging cool-climate regions poses numerous challenges. Firstly, Chardonnay is an earlier ripening cultivar which is very sensitive to events like spring frost and rain, which can strongly influence its phenological stages [5]. Secondly, the ideal growing season temperature (GST) for Chardonnay varies between 14 and 16 °C while, for later ripening cultivars such as Cabernet Sauvignon, the GST requirement varies between 16.8 and 20.2 °C [6]. Gladstones [7] has characterized the mean temperature in the warmest month for the Elgin and Walker Bay wine regions as 19.7 and 20.7 °C respectively. Overall, the Cape South Coast wine region is characterized by cooling wines because of its proximity to the ocean. Temperatures have a major impact on grapevine phenology, sugar accumulation and disruptions in synchronicity between fruit development and metabolites, which consequently affect wine styles [8,9]. Moreover, short-term extreme temperature fluctuations can cause disruptions to important metabolic processes such as photosynthesis, respiration and transpiration [10]. Thirdly, extreme weather conditions (e.g., hailstorms, black and white frost, and droughts) can result in complete crop losses [11,12]. Furthermore, the Walker Bay and Elgin regions typically have 750 mm and 800–1200 mm of rain annually. Overall, the soil in the Elgin and Walker Bay regions are characterized as Bokkeveld shale and Table Mountain sandstone, which is well drained with high clay content. The Walker Bay region has a maritime climate, while Elgin is known as the coolest region primarily due to its elevation at 200–500 m above sea level and the presence of cloud cover because of wind. Summer Southerly breezes off the cold Atlantic and the black South Easter create cloud cover in the region which filters light and results in slower ripening.

Topography is of equally important consideration, as the slope aspect and elevation of a vineyard can determine the frequency and risks of extreme weather events and determine the success of grapevine growth and the quality of the fruit produced [13,14]. Lastly, appropriate viticulture practices are also needed to adapt to cool-climate regions. The timing and extent of canopy practices can also alter grapevine phenology, ripening and yield [14,15]. It is unclear whether Chardonnay, when grown in emerging cool-climate regions, experiences these challenges. In fact, site suitability and characterization studies have been performed extensively for numerous cultivars in South African wine regions [16,17,18,19,20,21,22]. Despite all these studies, none of them has focused on Chardonnay in cool climatic regions. The objective of this study was to investigate the phenological development, vegetative growth and reproductive performance of Vitis vinifera L. cv. Chardonnay cultivated in the cool-climate Cape South Coast region of South Africa. By examining seasonal variations across two growing seasons and assessing site-specific environmental and viticultural factors—including rootstock–scion interactions, canopy structure, berry composition and soil properties—this study aims to characterize how Chardonnay responds to the emerging cool-climate viticultural zones. This study aimed to evaluate the viticultural characteristics of Chardonnay vineyards in the Cape South Coast region of South Africa with typical cool climates.

2. Materials and Methods

2.1. Vineyard Characteristics and Experimental Layout

The study was conducted in the 2021–2022 and 2022–2023 growing seasons (December–February) in three commercial Vitis vinifera L. cv. Chardonnay vineyards in the Cape South Coast region of South Africa. The three vineyard sites (Site A, Site B and Site C) were situated in the Walker Bay and Elgin districts (Figure 1;

Table 1. Characteristics of sites used in this study.

Site	Scion Clone	Rootstock	Year Planted	Planting Density	Row Orientation
A	CY 95 I	Richter 110	2005	3333	N–S
B	CY 277 D	Richter 110	2009	6667	NE–SW
C	CY 548	101-14 Mgt	2004	2667	N–S

). Three mini-plots comprising three rows (four to six vines between poles) were identified using Normalized Difference Vegetation Index (NDVI) imaging. To ensure that the respective mini-plots were homogenous, the NDVI images were confirmed with respect to the ground truth for each site. Therefore, three mini-plots were selected per site as biological replicates, consisting of 18 vines each, for a total of 54 vines per site.

2.2. Climatic Evaluation

The microclimate within the canopy was determined at each site using Tinytag loggers (Gemini Data Loggers Ltd.; West Sussex, UK). Tinytags were installed within the canopy at one mini-plot per site. The temperature was measured every 15 min from the beginning of December until the end of March (96 daily measurements) in both seasons. The temperature thresholds (≤20 °C; 20–25 °C; 25–30 °C; 30–35 °C; >35 °C) were calculated from December to March. The Growing Degree Days (GDD) was determined according to Equation (1).

$$GDD={\displaystyle {\sum }_{t=0}^n\frac{(T_{\max }+T_{\min })}{2}}-T_b$$

(1)

where T_max and T_min are the daily maximum and minimum temperature, respectively, and T_b is the base temperature for grapevine growth, which was considered to be 10 °C as per [23].

2.3. Soil Chemical Analysis

Soil analysis was conducted on 1 kg soil samples collected from soil pits at three different depths (0–30 cm, 30–60 cm and 60–90 cm) in October 2022. The soil pits were dug 1.5 m deep, 2 m long and 0.6 m wide, using an excavator. The soil was described and characterized by VinPro viticultural services, Paarl, South Africa, and the soil samples were analyzed at independent laboratories for chemical and physical analyses.

The analysis of the micro- (Cu, Mn and Zn) and macro-element (Ca, K, Mg and Na) contents, carbon (C) and the pH of each sample was conducted by Elsenburg Analytical Laboratories (Western Cape Department of Agriculture, Stellenbosch, South Africa). The analysis of total nitrogen (N) was conducted by Bemlab (Pty) Ltd. (Strand, Western Cape, South Africa), and electrical conductivity (EC) was determined by SGS (Pty) Ltd., Cape Town, South Africa. The soil water content was calculated for each sample following the drying method described by Kok et al. [24].

2.4. Phenological and Vegetative Measurements

The phenological characteristics at the respective sites were evaluated using the Eichorn and Lorenz system on a weekly basis [25]. Each of the three mini-plots at each of the sites were evaluated for phenological characterization. Daily macro-climate data (temperature) were obtained from three automatic weather stations situated across the Walker Bay and Elgin wine districts, in closest proximity to each site, using the iLeaf platform (https://ileafweather.com/LogIn.aspx, accessed on 21 July 2025).

Three vines—one in each panel per mini-plot—were selected for weekly canopy length, width and height measurements over the course of each season, from bunch closure (E-L stage 32) until harvest ripeness (E-L stage 38) [25]. Canopy length was measured from canopy end-to-end, width from the morning to afternoon side of the canopy, and height from the cordon to the shoot tip [26].

2.5. Grape Carpological Parameters

In each mini-plot, five bunches were selected on both the morning and afternoon side of the vines for weekly destructive measurements, and a single berry bunch was selected in the middle panel of each mini-plot for weekly non-destructive measurements. Destructive measurements were conducted (berry weight, length, and width) weekly from bunch closure (E-L stage 32) until harvest ripeness (E-L stage 38). Berries for destructive measurements were sampled from the top, middle and bottom of the tagged bunches (30 berries per mini-plot), and 20 berries were sampled randomly from the top, middle and bottom of surrounding untagged bunches. These berry samples (150 berries per site) were processed immediately after sampling and their individual weights, lengths and widths were determined.

Non-destructive measurements of berry length and width were conducted weekly on a selected top, middle and bottom berry of each cluster tagged for non-destructive measurements.

2.6. Classical Berry Quality Parameters

A total of 150 berries per site (50 berries × 3 mini-plots) were sampled weekly. The berries were crushed using a Logik stick blender, Gauteng, South Africa (200 W). The juice was sieved after crushing and Total Soluble Solids (TSS) were measured with an ATAGO pocket refractometer (Atago Co., Ltd., Tokyo, Japan). Sugar loading was determined using the method described by Deloire [27].

2.7. Transpiration Rate

Berry lengths and widths were determined at inception, and the berry transpiration rate was measured hourly over three hours in the laboratory for three randomly selected berries from each mini-plot. This was done on a weekly basis from bunch closure until harvest, based on the method of Lescourret et al. [28]. The pedicel of each berry was dipped in paraffin wax before recording its fresh weight every hour for a total of three hours. The temperature of the laboratory environment was also tracked using a digital thermo-hygrometer (TFA Dostmann GmgH & Co.KG; Wertheim, Germany). The transpiration rate of each berry was obtained by calculating the slope of fresh weight loss as a function of time, and the transpiration rate per unit area was obtained by adding the fruit surface area into the calculation. Fruit transpiration was calculated using the models described by Ben-Yehoshua [29] and Fishman & Génard [30].

$$T_f={\displaystyle \frac{(P_f-P_a)\times A_f\times \rho \times M_w}{R\times T}}=\Delta W$$

(2)

T_f: Transpiration rate of fruit (g h⁻¹)

P_f: Steady-state gas partial pressure in the fruit intercellular spaces (bar)

P_a: Steady-state gas partial pressure in the ambient atmosphere (bar)

A_f: fruit surface area (cm²)

ρ: permeation coefficient of fruit surface to water vapor

R: gas constant (83 cm³ bar mol⁻¹ K⁻¹)

T: Temperature on the absolute scale (K)

M_w: molecular mass of water (18 g mol⁻¹)

∆W: fruit weight loss (g h⁻¹)

$$RH={\displaystyle \frac{P}{P*}}$$

(3)

P: Steady-state gas partial pressure (bar)

P*: Saturation vapor pressure (bar)

$$P*=0.008048 \times \exp [0.0547 \times (\mathrm{T} - 273.15)]$$

(4)

Combining Equations (2)–(4), we can calculate the skin water conductance (ROU):

$$\rho ={\displaystyle \frac{\Delta W\times R\times T}{A_f\times M_w\times P^{*}\times (H_f-H_a)}}$$

2.8. Reproductive Growth

Grapes were harvested at commercial maturity for each producer, ranging between 19–23 Brix, at the vineyard sites. Grapevines within each mini-plot were harvested individually, and their yield and bunch numbers were recorded.

2.9. Plant Balance Assessment

Spur and cane number per vine were counted in each mini-plot at each site. One cane per vine was selected and its node number and length were determined. The vines were pruned (except for Cazenave pruning at Site A) during the dormancy months of July to August each year (2021–2023). The canes were weighed using a Salter-Brecknell ElectroSamson digital hand-held scale (Avery Weigh-Tronix, LLC, Fairmont, WV, USA). The Ravaz Index [31], an indicator of vine balance, was determined by dividing the total yield of a vine (kg) by the total pruning weight of the vine (kg); see Equation (5). Cane weight below 10 g was indicative of low vigor, 20–40 g medium vigor and higher than 60 g per vine indicated high vigor [31].

Ravaz index = Yield/Pruning Weight

(5)

2.10. Statistical Analysis

To assess the effects of Season, Site and their interaction on all measured parameters, a linear mixed-effects model was applied using custom scripts developed in Matlab R2023a (MathWorks, Natick, MA, USA). Statistical significance was investigated at p-value ≤ 0.05, ≤0.01 and ≤0.001 (*, ** and ***, respectively). Days after Anthesis was incorporated as a continuous covariate to account for temporal variations in the dependent variables. Given the differing ranges of Days after Anthesis across seasons, which could introduce bias due to non-overlapping covariate values, the analysis was restricted to the overlapping range common to all seasons to ensure comparability. Model assumptions were evaluated from residuals using the Shapiro–Wilk test (normality) and Levene’s test (homogeneity). Post-hoc comparisons among factor levels were performed using the Tukey–Kramer HSD procedure to control the family-wise error rate.

Principal component analysis (PCA) was also performed to determine the relationships between the main variables analyzed and the site and season factors. Exploring the score and loading plots provides a better understanding of the sources of variability in the data, and can reveal clusters and trends in the data. The variables used to conduct the PCA were the phenological and vegetative measurements, grape carpological parameters, classical berry quality parameters, transpiration rate measurements, reproductive growth measurements and plant balance assessment parameters. To perform this analysis, MATLAB R2023a and PLSToolbox version 9.0 for MATLAB (Eigenvector Inc., Manson, WA, USA) were used.

3. Results

3.1. Climatic and Soil Characterization

The 2021–2022 season was warmer when compared to the 2022–2023 season. Most of the temperatures for the respective sites were ≤20 °C or ranged between 20 and 25 °C. Seasonal (December–March) within-canopy mean temperatures were 20.7, 21.1 and 20.9 °C for Sites A, B and C in 2021–2022, and 20.2, 20.8 and 20.6 °C in 2022–2023. Inter-annual means were ~0.3–0.5 °C cooler in 2022–2023 across all sites, with Site B consistently the warmest and Site A the coolest (

Table 2. Monthly temperatures and number of hours at temperature thresholds for Chardonnay sites.

Season	Site	Month	Mean Temperature per Site (°C)	Mean Temperature per Month (°C)	Maximum Temperature (°C)	Number of Hours
						≤20 °C	20–25 °C	25–30 °C	30–35 °C	>35 °C
2021–2022	A	December	20.7	19.0	19.32	15.76	4.95	2.33	0.71	0.24
		January		21.8	22.05	11.64	5.42	3.90	2.10	0.90
		February		21.5	21.77	12.61	3.86	4.96	2.14	0.43
		March		20.0	20.27	14.77	4.26	3.26	1.58	0.13
2021–2022	B	December	21.1	19.0	19.15	16.14	5.05	2.14	0.67	0.00
		January		21.9	22.13	10.81	6.55	4.26	1.65	0.74
		February		21.8	22.04	12.00	5.18	4.00	2.25	0.57
		March		21.0	21.32	13.5.2	4.52	3.39	2.00	0.58
2021–2022	C	December	20.9	18.8	19.03	15.48	5.10	2.71	0.71	0.00
		January		22.4	22.69	10.81	4.87	4.87	2.61	0.84
		February		21.8	22.06	12.39	3.68	4.04	3.39	0.5
		March		20.2	20.44	14.77	3.84	3.00	2.10	0.29
2022–2023	A	December	20.2	20.0	20.28	13.39	5.39	3.97	1.16	0.1
		January		21.3	21.61	12.27	4.6	5.13	1.47	0.53
		February		20.8	21.09	12.79	5.61	3.71	1.32	0.57
		March		18.7	18.92	16.16	4.94	2.23	0.52	0.16
2022–2023	B	December	20.8	20.5	20.75	12.35	6.45	4.00	1.16	0.03
		January		22.0	22.29	11.13	5.4	4.4	2.40	0.67
		February		21.2	21.44	12.43	6.11	3.64	1.32	0.5
		March		19.4	19.74	15.1	5.26	2.65	0.61	0.39
2022–2023	C	December	20.6	19.8	20.07	14.29	4.45	3.35	1.87	0.03
		January		21.5	21.75	12.83	3.63	3.43	3.63	0.47
		February		21.0	21.24	13.82	4.46	2.57	2.39	0.75
		March		18.9	19.19	13.82	4.46	2.57	2.39	0.75

). The respective sites varied in terms of the accumulated GDD, ranging between Category I and Category II with accumulated heat below 1370 °C days and between 1370 and 1650 °C days (Figure A1), consistent with cool to very cool conditions. The canopy temperatures varied between the seasons and the sites. Overall, higher canopy temperatures (18.8–22.4 °C) were recorded in the first season when compared to the second season (18.7–22.0 °C) (Table 2).

The phenology of the 2021–2022 and 2022–2023 seasons varied drastically. In 2021–2022, cool, windy and rainy weather conditions prevailed in the initial stages of the season, followed by intense heat in January [32]. The 2022–2023 season was characterized as a drier season [33]. Overall, it is clear the E-L stages in the 2022–2023 season were earlier than in the previous season (

Table A1. Chardonnay sampling dates with corresponding Days After Anthesis (DAA) and E-L stages for both seasons.

Site	2021–2022 Season			2022–2023 Season
	Sampling Dates	Days After Anthesis	E-L Stage	Sampling Dates	Days After Anthesis	E-L Stage
Site A	29 December 2021	51	32	11 January 2023	72	34
	5 January 2022	58	33	18 January 2023	79	35
	26 January 2022	79	35	25 January 2023	86	36
	2 February 2022	86		1 February 2023	93
	10 February 2022	94	36	8 February 2023	100	37
	17 February 2022	101	37
	24 February 2022	108
	2 March 2022	114	38
Site B	29 December 2021	75	32	11 January 2023	76	35
	5 January 2022	82	33	18 January 2023	83
	26 January 2022	103	35	25 January 2023	90	36
	2 February 2022	110		1 February 2023	97	37
	10 February 2022	118	36	8 February 2023	104
	17 February 2022	125	37
Site C	29 December 2021	48	32	11 January 2023	72	35
	5 January 2022	55	33	18 January 2023	79
	26 January 2022	76	35	25 January 2023	86	36
	2 February 2022	83	36	1 February 2023	93	37
	10 February 2022	91		8 February 2023	100
	17 February 2022	98	37
	24 February 2022	105	38

). Significant differences were observed in the temperatures on a macro-climatic level. The weather station at Site B was consistently higher when compared to the other two weather stations (Table 2).

The soil chemical analysis did not differ significantly at the respective sites (

Table 3. Soil chemical analysis for the three sites in the Walker Bay and Elgin Wine districts.

	pH	EC (mS/m)	Total Nitrogen (%)	Calcium (cmol/kg)	Magnesium (cmol/kg)	Sodium (mg/kg)	Potassium (mg/kg)	Copper (mg/kg)	Zinc (mg/kg)	Manganese (mg/kg)	Total Carbon (%)	Water %
Site
Site A	6.75 ± 1.12	9.26 ± 3.02	0.10 ± 0.04	4.79 ± 3.98	1.05 ± 0.75	64.33 ± 12.06	17.33 ± 6.66	1.25 ± 0.66	0.56 ± 0.36	2.53 ± 3.50	1.44 ± 1.04	2.85 ± 0.45
Site B	6.05 ± 0.58	8.50 ± 1.91	0.09 ± 0.07	2.59 ± 2.57	1.52 ± 0.16	50.00 ± 3.00	60.67 ± 60.96	1.28 ± 1.05	0.87 ± 0.81	0.41 ± 0.35	2.05 ± 2.15	3.01 ± 0.76
Site C	5.96 ± 0.75	13.68 ± 4.04	0.10 ± 0.04	3.63 ± 2.72	1.32 ± 0.35	70.67 ± 14.57	52.67 ± 52.37	1.79 ± 0.64	1.05 ± 0.63	2.33 ± 3.31	0.93 ± 0.86	3.05 ± 0.68
Depth
0–30	7.08 ± 0.65 a	10.14 ± 1.92	0.15 ± 0.02 a	6.89 ± 1.81 a	1.61 ± 0.30	54.67 ± 8.02	89.67 ± 56.72	1.56 ± 1.01	0.97 ± 0.73	4.51 ± 3.21	2.88 ± 1.43 a	2.82 ± 0.86
30–60	6.16 ± 0.64 ab	9.99 ± 6.24	0.08 ± 0.04 ab	3.00 ± 1.48 b	1.07 ± 0.48	68.00 ± 15.59	22.67 ± 7.57	1.35 ± 0.40	0.68 ± 0.17	0.54 ± 0.24	1.21 ± 0.50 ab	2.85 ± 0.19
>60	5.51 ± 0.08 b	11.31 ± 2.89	0.06 ± 0.01 b	0.85 ± 0.14 b	1.22 ± 0.57	62.33 ± 16.17	18.33 ± 5.03	1.40 ± 1.00	0.83 ± 0.85	0.23 ± 0.07	0.32 ± 0.21 b	3.24 ± 0.60
Significance
Site	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns
Depth	*	ns	*	**	ns	ns	ns	ns	ns	ns	*	ns

a,b: Means with a common letter between rows are not significantly different at 5% probability. *, and ** are indicative of significance at a p-value of 0.05, 0.01 and 0.001, respectively; ns indicates non-significance. Different letters indicate significant differences at p ≤ 0.05.

). The chemical composition was rather altered by the depth at which the samples were taken. The pH decreased with depth, suggesting a more acidic medium with an effective soil depth greater than 60 cm (Table 3). The significantly higher (p ≤ 0.01) calcium in the soil points to the fertilization practices followed by the producers. The availability of calcium also became lower with decreasing soil depth, which is also a function of the availability of the nutrients due to pH.

3.2. Vegetative Measurements

The canopy density (width) and total leaf surface area differed significantly (p ≤ 0.01) between seasons (

Table 4. Vegetative characteristics of Chardonnay in the 2021–2022 and 2022–2023 growing seasons in the Cape South Coast.

		Canopy Length (cm)	Canopy Width (cm)	Total Leaf Surface Area (cm3)
Season
2021–2022		124 ± 22	71 ± 16	65.7 ± 20.4
2022–2023		96 ± 11	30 ± 6	33.9 ± 7.0
Site
Site A		109 ± 17	49 ± 21 b	51.6 ± 15.9 b
Site B		110 ± 17	52 ± 21 b	39.0 ± 12.9 c
Site C		114 ± 30	56 ± 29 a	62.9 ± 29.2 a
Season × Site
2021–2022	Site A	116 ± 20	64 ± 15	63.6 ± 11.5 b
	Site B	121 ± 13	68 ± 13	48.8 ± 7.3 c
	Site C	138 ± 24	83 ± 17	88.5 ± 18.1 a
2022–2023	Site A	100 ± 9	30 ± 3	37.7 ± 3.3 d
	Site B	97 ± 11	32 ± 2	27.2 ± 6.4 e
	Site C	90 ± 10	30 ± 3	37.3 ± 5.6 d
Significance
Season		ns	ns	ns
Site		ns	**	**
Season × Site		ns	ns	*

a–e: Means with a common letter between rows are not significantly different at 5% probability. * and ** are indicative of significance at p-values of 0.05, 0.01 and 0.001, respectively; ns indicates non-significance.

). The wetter growing season (2021–2022) contributed to the higher vigor observed.

In our study, Chardonnay clones CY 95, CY 247 and CY 548 were used at the respective sites. Clones CY 95 and CY 277 were characterized as having above average production and vigor at Site A and Site B correspondingly. Clone CY 548 at Site C was characterized as having average to lower production. Site B had the smallest canopy compared to the other two sites in 2022–2023, which can be ascribed to the fertilization strategy used by the producer. The latter phenomenon could be due to: (i) the proximity of the ocean in this study and the consequent exposure to high wind levels, and (ii) the planting density at the sites. In our study, we observed increasing canopy height with decreasing planting density.

3.3. Grape Carpological Parameters

Destructive and Non-Destructive Berry Measurements

Berry measurements were conducted on intact berries in the field as well as destructively. There was a significant difference in the fresh weight of berries harvested destructively (p ≤ 0.01) between the seasons. The berry fresh weight was higher in the 2022–2023 season than in 2021–2022 (

Table 5. Destructive and non-destructive grape berry measurements of Chardonnay over two growing seasons.

		Non-Destructive		Destructive
		Berry Length (mm)	Berry Width (mm)	Fresh Weight (g)	Berry Length (mm)	Berry Width (mm)
Season
2021–2022		12.3 ± 2.1	12.2 ± 2.2	1.20 ± 0.40 b	12.1 ± 1.3 a	11.7 ± 1.3 a
2022–2023		13.1 ± 1.2	12.6 ± 1.2	1.26 ± 0.38 a	11.6 ± 2.0 b	11.4 ± 1.9 b
Site
Site A		12.5 ± 1.5	11.8 ± 2.0	1.18 ± 0.33 b	11.4 ± 1.9 c	11.1 ± 1.9 c
Site B		12.1 ± 2.0	12.0 ± 1.3	1.18 ± 0.39 b	11.9 ± 1.3 b	11.7 ± 1.3 b
Site C		13.7 ± 1.6	13.4 ± 1.7	1.34 ± 0.42 a	12.3 ± 1.6 a	11.9 ± 1.6 a
Season × Site
2021–2022	Site A	11.5 ± 2.1 c	11.5 ± 2.3	1.21 ± 0.34 b	12.1 ± 1.2 bc	11.7 ± 1.2 b
	Site B	11.9 ± 1.5 c	11.7 ± 1.4	1.17 ± 0.43 bc	12.0 ± 1.4 c	11.7 ± 1.4 b
	Site C	14.0 ± 1.9 a	13.5 ± 2.3	1.22 ± 0.42 b	12.2 ± 1.3 ab	11.8 ± 1.3 b
2022–2023	Site A	12.8 ± 1.4 b	12.0 ± 1.4	1.14 ± 0.33 c	10.6 ± 2.3 d	10.4 ± 2.2 c
	Site B	13.2 ± 1.1 ab	12.4 ± 1.0	1.20 ± 0.34 b	11.9 ± 1.2 c	11.6 ± 1.2 b
	Site C	13.4 ± 1.1 ab	13.2 ± 1.0	1.46 ± 0.38 a	12.3 ± 1.8 a	12.0 ± 1.8 a
Significance
Season		ns	ns	**	***	***
Site		ns	ns	***	***	***
Season × Site		*	ns	***	***	***

a–c: Means with a common letter between rows are not significantly different at 5% probability. *, ** and *** are indicative of significance at p-values of 0.05, 0.01 and 0.001, respectively; ns indicates non-significance.

). Additionally, berries from Site C consistently exhibited the highest fresh weights and dimensions, especially during the second season, suggesting favorable microclimatic or soil conditions at this location, or more vigorous vine performance (as also supported by the vegetative growth data). Berry length and width showed a similar trend and differed significantly (p ≤ 0.001) between the seasons. However, the variation between destructive and non-destructive values in some parameters points to a need for caution when interpreting field-based measurements alone, as they may be influenced by positional or sampling biases. Irrespective of the type of berry weight determinants, similar results were obtained.

3.4. Grape Berry Primary Metabolites

Total Soluble Solids (TSS) and sugar accumulation followed a similar trend across all Chardonnay sites in the respective seasons (Figure 2a,b). In the 2021–2022 season, Chardonnay sugar content was highest at harvest at Site A (161 mg/berry), followed closely by Site C (156 mg/berry) and Site B (153 mg/berry) (Figure 2c). In the 2022–2023 season, the corresponding order was Site C (240 mg/berry), Site B (199 mg/berry) and Site A (181 mg/berry) (Figure 2d).

3.5. Transpiration Rate

The transpiration rate per unit area and berry skin water conductivity (rou) differed significantly (p ≤ 0.05) between the sites in the later stages of berry development (

Table 6. Chardonnay grape berry evapotranspiration over two growing seasons (2021–2022 and 2022–2023).

		Transpiration Rate (g/day)	Transpiration Rate per Unit Area (g/(Day cm2))	Skin Water Conductivity (ROU)
Season
2021–2022		0.055 ± 0.023	0.010 ± 0.005	37.3 ± 19.6
2022–2023		0.070 ± 0.048	0.014 ± 0.011	43.6 ± 31.7
Site
Site A		0.060 ± 0.046	0.012 ± 0.010 ab	39.6 ± 31.1 b
Site B		0.075 ± 0.040	0.016 ± 0.010 a	52.1 ± 27.4 a
Site C		0.049 ± 0.024	0.009 ± 0.004 b	30.8 ± 13.5 c
Season × Site
2021–2022	Site A	0.049 ± 0.017	0.010 ± 0.004	34.7 ± 17.3
	Site B	0.058 ± 0.029	0.012 ± 0.007	44.2 ± 27.0
	Site C	0.048 ± 0.022	0.009 ± 0.003	33.5 ± 9.1
2022–2023	Site A	0.071 ± 0.062	0.015 ± 0.013	44.6 ± 40.0
	Site B	0.093 ± 0.041	0.019 ± 0.010	60.1 ± 25.7
	Site C	0.049 ± 0.025	0.009 ± 0.005	28.5 ± 16.1
Significance
Season		ns	ns	ns
Site		ns	*	*
Season × Site		ns	ns	ns

a–c: Means with a common letter between rows are not significantly different at 5% probability. * are indicative of significance at p-values of 0.05, 0.01 and 0.001, respectively; ns indicates non-significance.

).

3.6. Grapevine Balance

Grapevine yield did not differ significantly between the seasons (

Table 7. Grapevine balance comparison for Chardonnay over two growing seasons and three sites in the Cape South Coast.

		Yield (kg)	Bunches per Vine	Shoot Length (cm)	Vine Cane Weight (kg)	Ravaz Index
			Season
2021–2022		1.86 ± 1.47	22.2 ± 16.6 a	92 ± 21 a	0.45 ± 0.35	5.13 ± 5.00
2022–2023		1.88 ± 1.47	18.2 ± 9.7 b	59 ± 21 b	0.30 ± 0.24	6.04 ± 3.34
			Site
Site A		1.73 ± 0.88 b	16.9 ± 7.2 b	63 ± 22 c	0.29 ± 0.13 b	6.53 ± 3.48 a
Site B		0.87 ± 0.59 c	11.4 ± 6.1 c	78 ± 24 b	0.22 ± 0.10 c	5.41 ± 7.00 ab
Site C		2.88 ± 1.95 a	33.1 ± 7.8 a	88 ± 31 a	0.75 ± 0.40 a	4.00 ± 2.14 b
			Season × Site
2021–2022	Site A	1.86 ± 0.97 b	19.0 ± 8.0 bc	81 ± 14	0.37 ± 0.13 c	5.70 ± 3.59 ab
	Site B	0.87 ± 0.59 c	11.4 ± 6.1 d	92 ± 16	0.24 ± 0.11 d	5.41 ± 7.00 ab
	Site C	3.48 ± 1.65 a	44.3 ± 16.5 a	107 ± 28	0.92 ± 0.40 a	3.83 ± 1.56 b
2022–2023	Site A	1.60 ± 0.77 b	14.7 ± 5.7 cd	45 ± 10	0.22 ± 0.06 d	7.35 ± 3.20 a
	Site B	N.A.	N.A.	65 ± 23	0.21 ± 0.10 d	N.A.
	Site C	2.28 ± 2.07 b	23.2 ± 11.9 b	69 ± 20	0.58 ± 0.33 b	4.17 ± 2.60 b
			Significance
Season		ns	*	***	ns	ns
Site		***	***	***	***	***
Season × Site		*	***	ns	***	**

a–d: Means with a common letter between rows are not significantly different at 5% probability. *, ** and *** are indicative of significance at p-values of 0.05, 0.01 and 0.001, respectively; ns indicates non-significance. Different letters indicate significant differences at p ≤ 0.05.

). However, a significant difference (p ≤ 0.01) was observed at each of the sites (Table 7). Grapes were not harvested at Site B in 2022–2023 due to high fungal pressure, which led to the below average yield in the second season. The season x site interaction varied significantly (p ≤ 0.05). Bunch number per vine varied significantly between the seasons (p ≤ 0.05). Site C had the highest bunch numbers when compared to the other two sites (Table 7). Higher vigor stimulated higher yields and a greater bunch number per vine.

Shoot lengths were significantly (p ≤ 0.001) higher in the 2021–2022 season (Table 7). Subsequently, the Ravaz index was higher in this season (Table 7). The same parameters were significantly lower in the 2022–2023 season. Site C had significantly higher shoot lengths, pruning mass and Ravaz index. This vegetative robustness translated into larger berries (Table 5) at Site C, with significantly greater berry length, width and fresh weight in both destructive and non-destructive assessments, especially during the 2022–2023 season. The higher berry size and mass may reflect better source–sink capacity, potentially driven by canopy size and photosynthetic area. Additionally, due to severe fungal pressure, grapes at Site B were not harvested in the 2022–2023 season, resulting in lower yield and an artificially low Ravaz Index. No significant interaction was recorded between site and season in terms of shoot lengths, but Site C had consistently the highest shoot lengths. Shoot mass and the Ravaz index differed significantly (p ≤ 0.001) with the season x site interaction.

3.7. Principal Component Analysis (PCA)

To explore the relationships between viticultural traits across seasons and sites, a Principal Component Analysis (PCA) was performed using vegetative measurements, classical parameters, transpiration rates and grapevine balance data. The first two principal components (PC1 and PC2) explained 71.06% of the total variance (Figure 3), with PC1 accounting for 48.93% and PC2 for 22.13%.

PC1 was associated with positive loadings of variables linked to vegetative vigor and structural development, such as shoot length, total canopy surface area, canopy height and width, vine shoot weight and yield, together with negative loadings such as Ravaz index and transpiration rate, relating this axis to the overall vine performance. PC2 was associated with fruit and ripening traits, including berry fresh weight, berry dimensions, pH and Brix with positive loadings and titratable acidity with negative loading. This axis thus differentiated samples associated with berry development and ripening. Site-wise, the Site C samples were consistently associated with positive scores along PC1, driven by high vegetative growth and larger berry dimensions. In contrast, samples from Sites B and A were mainly associated negatively along PC1, reflecting low vigor and reduced yield. The PC2 axis differentiates both vintages due to the inter-season variability observed in the berry development and ripening characteristics.

4. Discussion

Seasonal variation was the biggest driver of variation in the GDD at all sites (Figure A1). The recorded canopy temperatures were primarily driven by soil type, cover crops, row orientation and the proximity of the ocean. Overall, the first season had higher GDD when compared to the second season.

Cultivation practices such as row orientation and canopy management contribute to the micro-climatic conditions [17,22,34,35]. All of the sites were established on a southern facing slope which is cooler and with reduced light interception. Sites A and C had N-S orientated rows, which likely received higher Photosynthetically Active Radiation (PAR) in the bunch zone. Site B had a NE-SW row orientation, which received lower PAR in the bunch zone as the sun was continuously perpendicular above the rows. Orselli [33] has reported similar findings for Pinot noir in the Cape South Coast region.

Blancquaert et al. [36] reported the link between increased temperature and light in the canopies of Cabernet Sauvignon in the Stellenbosch wine region. However, the quantification of light in the canopy is outside the scope of this study.

The pH of the soil was impacted by the sampling depth, which led to the macro and micro-elements becoming less available for plant use. Nitrogen is a key macro-element supporting plant physiology responses and vegetative growth, and is thus a critical macro-nutrient for grapevines. The presence of total nitrogen can also be due to the use of cover crops at the respective sites. Carbon % also decreased, which can be ascribed to the use of cover crops. Soil chemical characteristics have been reported by other authors to impact vine growth and the qualitative aspects of wine [13,37,38]. Soil nitrogen concentration influenced vegetative vigor and yield, with soil nitrogen supplementation through nitrogen-based fertilizer use being a common viticultural practice [39].

Jones [5] reported that the ideal average growing season temperature for Chardonnay ranges between 14 and 17.5 °C. Greer & Weedon [40] reported that the rate of ripening of Chardonnay berries was heightened at 25 °C in Australia. The 2022–2023 growing season was characterized by drier growing conditions, which resulted in lower vigor [32,33]. The observed variations in vegetative growth can also likely be explained by the rootstock × scion interaction and the cultivation practices employed by the producers in this study.

Anderson et al. [41] evaluated thirteen French and Californian Chardonnay clones over four growing seasons to produce sparkling wine in California. The French clones were consistent in their viticultural characteristics, while the American clones varied in growth components. From the latter, different clones led to variations in grapevine performance.

Tandonnet et al. [42] and Marguerit et al. [43] reported that rootstocks induced vegetative vigor which influenced the size of the canopy. The lower vigor seen at Site A and Site B was consistent with reports by Woolridge et al. [44] on the low vigor imparted by rootstock Richter 110, when grafted on Chardonnay. Rootstock 101-14 Mgt induced higher vigor at Site C. Similar observations were reported by Sampaio [45], where Chardonnay grafted to 101-14 Mgt showed increased vigor. At the site level, canopy height varied significantly (p ≤ 0.01), while canopy density did not (p = 0.30) (Table 2).

Pienaar [46] reported a smaller leaf area and higher lateral shoot growth, while Merlot noir grapevines which were sheltered from excessive wind had longer shoots. Overall, the study area is characterized by high wind speeds during critical phenological stages such as flowering, due to its proximity to the ocean. A shortcoming of this study is the lack of access to wind data from weather stations. Our findings correspond to those of researchers [47] who found that interplant competition results in increased competition among vines for resources, subsequently reducing shoot and root growth.

Berry development is very dependent on the prevailing environmental conditions during flowering and berry set, which is followed by three stages of berry development [48]. Stage I is characterized by a period of rapid cell division and enlargement (Stage I), followed by a lag phase (Stage II) and fruit ripening (Stage III). Berry development in the 2022–2023 season was characterized by the occurrence of rainfall during Stage I of berry development, which contributed to the larger berries observed. Cortell et al. [49] and Keller et al. [50] reported similar findings in Chardonnay and Pinot noir. Overall, Chardonnay is very susceptible to millerandage, with this phenomenon resulting in poor fruit set at Site B. As previously mentioned, this site is also prone to higher wind speeds, which likely impacted flowering and therefore negatively impacted the berry weights in both seasons (Table 3). Additionally, the observed between-season increase in berry fresh weight may be linked to the timing of rainfall and water availability during early berry development, which can favor cell expansion and sugar accumulation.

The prevailing weather conditions were a strong driver of berry fresh mass measurements during Stage I of berry development. The sugar content in berries is linked to the assimilation of sucrose through photosynthesis, which is hydrolyzed to glucose and fructose. Sites A and C were characterized with a higher sugar accumulation when compared to Site B in the 2021–2022 season (Figure 2c). This phenomenon is likely due to the higher leaf area at Site C in this season. In the 2022–2023 season, Sites B and C had higher sugar accumulation compared to Site A. Our result is consistent with that of [51], who reported that grapevine leaves are source tissues which can synthesize and export solutes.

The high wind exposure throughout berry development led to higher evapotranspiration [46]. The skin water conductivity is determined by the structure and chemistry of the epicuticular wax layer, which may change with age or be disrupted by various factors (e.g., temperature, sun exposure and pathogens) [51]. The epicuticular wax layer is predominantly genetically controlled [51,52,53,54], which likely contributed to the differences seen in transpiration rates and skin water conductivity. Accordingly, the higher transpiration rates observed at Site B are plausibly linked to clonal differences affecting epidermal anatomy and wax deposition at that site.

Therefore, seasonal and site-specific factors that influenced vegetative vigor between the seasons and sites likely also played a role in the determination of yield and bunch number per vine. Another likely contributing factor is the pruning system applied by the producer. Steyn et al. [55] reported that vine vigor influences the production capacity of a vine, which affects grape production. Grapevine performance is not hindered and the crop load will remain balanced if a good vine balance is maintained [55]. The pruning method (cazenave) used at Site A contributed to an increase in yield and bunch number per vine at this site in both seasons, as reported by [56] for Chardonnay in Southern Brazil.

An overriding site and season effect was obvious amongst the sites. The observed variation is likely due to the differences in vegetative vigor seen between seasons and sites, which was induced by the rootstock. A link between cane weights and vigor was also reported by [57] for Chasselas and Merlot. However, the results were skewed as the producer pre-pruned the vineyards prior to conducting the measurement. Vine balance varied between sites and with the site x season interaction; this can be ascribed to the environmental conditions, rootstock and vineyard management practices applied at the respective sites.

The observed trends by site in the PCA highlight how vine performance and berry development are shaped by the combined effects of environment, soil and management practices. Site C was associated with higher vigor and larger berries, while Site B—especially in 2022–2023—presented low vegetative growth and poor yield due to fungal pressure. Site A showed intermediate responses, with variation across the seasons. The PCA results thus provide support for the univariate trends described previously, reinforcing the need for site-adapted viticultural strategies in emerging cool-climate regions, where seasonal variation and microclimatic differences can significantly impact the grapevine physiology and fruit quality.

5. Conclusions

This work represents the first detailed multi-season characterization of Vitis vinifera L. cv. Chardonnay in the emerging cool-climate Cape South Coast region of South Africa, integrating phenological, vegetative, reproductive, physiological and soil parameters. The novelty of this study lies in its site- and season-specific assessment of Chardonnay’s performance in a region where limited empirical data exist, despite the increasing shift of plantings toward such cooler zones in response to climate change.

The findings demonstrate that seasonal variation was the dominant driver of differences in growing degree days, canopy microclimate, vegetative vigor and berry development. Site-specific factors, including the rootstock–scion combination, row orientation, planting density and management practices, modulated these responses, resulting in distinct vine growth and yield patterns across locations. The wetter 2021–2022 season favored vegetative growth and yield, whereas the drier 2022–2023 season reduced vigor but, in some cases, increased berry size through favorable water conditions during early development. The PCA confirmed that sites could be clearly differentiated based on vigor, berry size and vine balance indicators, with Site C consistently outperforming others and Site B showing reduced performance, partly due to fungal pressure.

A critical reflection on these outcomes underscores both the adaptability and sensitivity of Chardonnay in this environment. While the cultivar proved suitable for Cape South Coast production, its yield stability and fruit composition are vulnerable to inter-annual climatic variability, disease pressure and site-specific constraints. Soil chemical profiles were broadly similar among the sites, suggesting that aboveground performance differences are more strongly driven by climate, vine physiology and management than by nutrient limitations under current practices.

While this study offers valuable insights, certain aspects provide scope for further exploration. Conducting the work within commercial vineyard settings ensured its real-world relevance, although it also meant that some variables (e.g., pruning systems, fertilization regimes) reflected standard grower practices rather than experimental controls.

Future research should expand this work across more seasons and incorporate controlled trials to isolate the effects of rootstock–scion combinations, planting densities and canopy management on vine physiology and fruit quality. Integrating detailed light and environmental measurements, plant water status monitoring and berry metabolomics would further elucidate the mechanisms linking environmental and management factors to fruit composition. Such research is expected to strengthen the viticultural decision-making framework needed to optimize Chardonnay production in South Africa’s cool-climate regions under evolving climatic conditions.