Grapevine responses to site-specific spatiotemporal factors in a Mediterranean climate

Highlights

• Long-terms spatiotemporal dynamics of grapevine attributes were analyzed.

• Effects of factors associated with water availability on vines were quantified.

• A negative trend in spatial dependency was found for all grapevine attributes.

• Precipitation affected vine performance in wet years, mediated by terrain.

• Differential irrigation had a stronger impact following a relatively dry winter.

Abstract

Water availability in vineyards varies in space and time. The spatiotemporal variability of water availability to vines is affected by terrain, meteorology, and irrigation. This study aimed to analyze the response of vegetative and reproductive attributes of vines to spatiotemporal variability in water availability. We quantified the spatial autocorrelation of terrain covariates and grapevine attributes and determined the relative influence (RI) of the covariates on these attributes. In each of four growing seasons (2017–2020), in a Vitis vinifera cv. “Sauvignon Blanc” vineyard, five vegetative and reproductive attributes were collected for 240 vines. Terrain covariates included elevation, slope, aspect, topographic wetness index, and categorical landforms; and meteorological covariates consisted of annual rainfall and chilling hours. A Moran’s I statistic was computed for each spatial covariate and each grapevine attribute during each season to define temporal changes in spatial variability. A multivariate analysis using gradient boosted regression trees algorithm was applied to extract the RI of the covariates. The results showed high spatial autocorrelation of the terrain covariates and a strong negative shift in spatial dependency of the grapevine attributes throughout the experiment. Yield and number of clusters per vine were highly affected by seasonal precipitation (RI of 46.15% and 42.59%), while changes in inter-seasonal cluster weight and pruning weight were highly subjected to irrigation amounts (RI of 23% and 26.66%), with complementary terrain influences. The number of canes per vine was mainly affected by terrain characteristics. Long-term changes in grapevine attributes depended on meteorological shifts, while higher precipitation amounts were associated with weaker responses of vines to irrigation strategies. The spatial patterns of terrain affected water distribution in the vineyard and controlled the spatial dynamics of grapevine attributes. Knowledge regarding the space-time trends of water availability effects on grapevine attributes may assist decision making of irrigation practices in vineyards.

Introduction

Soil water availability is a major factor in grapevine growth and productivity (Medrano et al., 2003). Water distribution in vineyards often varies in space and time (Acevedo-Opazo et al., 2008), depending on various factors. Spatial variability of water availability in the field may be governed by the spatial patterns of terrain attributes such as elevation, slope, and aspect (Ohana-Levi et al., 2019, Willwerth and Reynolds, 2020), while seasonal (temporal) dynamics in vineyard water regime depend mainly on meteorological conditions such as precipitation amounts (Wilson et al., 2020) and temperature levels (de Rességuier et al., 2020), as well as the irrigation provided to the grapevines (Ohana-Levi et al., 2020a) alongside other agrotechnical practices.

The spatial distribution of water availability is affected by multiple factors and impacts the spatial variability of the grapevines vegetative and reproductive attributes. Soil moisture varies from dry conditions in upper slopes to wet conditions in lower slopes, where upslope hydrological process affect downslope water availability (Brolsma and Bierkens, 2007). This process was found to affect the spatial patterns of vegetation performance, which rely on the distribution of soil water availability across the field. In an experiment in an “Aglianico” vineyard in Southern Italy, Basile et al. (2020) found differences in physiological and vegetative responses of grapevines to slope gradient; for example, downslope vines had lower water stress than upslope vines. Similarly, Santesteban et al. (2011a) found differences in grapevine water stress in different slope aspects in a “Tempranillo” vineyard in Southern Navarre, Spain. However, flat terrain seems to be less favorable to grapevine performance, and Irimia and Patriche (2013) reported the optimal slope range to be 5–15% (i.e., 2.85°–8.5°). Similar slope ranges were reported in a study conducted in Illinois by Kaan Kurtural et al. (2008). Ohana-Levi et al. (2020a) conducted a time-series analysis of daily images of evapotranspiration (ET) derived from remote sensing retrievals during four growing seasons in a “Pinot noir” vineyard in California. They found that soil type and elevation were the main contributors to ET spatial variability. Furthermore, topographic wetness index (TWI), a calculation of the ratio between the absolute area and slope, was found to be associated with soil moisture in a “Sangiovese” vineyard, Central Italy (Costantini et al., 2009) and with water stress and canopy cover in a “Cabernet Sauvignon” vineyard in Israel (Bahat et al., 2021). The response of grapevine attributes to water availability is commonly spatially dependent and relies on the interrelations among multiple factors such as soil characteristics, terrain, and irrigation management. These relationships among field properties are inconsistent and depend on natural and controllable factors such as the climatic region, vine variety, and rootstock (Smith and Whigham, 1999).

The temporal distribution of water availability is affected by meteorological attributes, such as precipitation and temperature, and controllable attributes, such as irrigation amounts. Precipitation has immediate influence on soil moisture, while deeper soils are more affected by low intensity rainfall events. Dry conditions lead to sharp decrease in soil moisture with low spatial variability. Occurrences of long dry periods are known to have harmful effects on yield in terms of quantity and quality (Ramos and Mulligan, 2005, Ramos and Martínez-Casasnovas, 2006). In an experiment in a California vineyard, Wilson et al. (2020) found that the main component affecting water availability early in the season was precipitation during the wet season, which had an impact on leaf area development. Furthermore, larger amounts of seasonal precipitation, leading to more stored water in the soil, were related to an extended period of canopy growth. They also showed that deficit in soil moisture was higher and dynamic during April and May, when canopy develops, and then stabilized at constant levels of water lost from the soil. Temporal effects in the seasonal scale may also be attributed to other factors such as chilling hours. A chilling requirement must be fulfilled for buds to synchronously burst and is believed to be between 0 and 7.2 °C, while grapevines typically require between 50 and 400 chilling hours during winter to satisfy endodormancy. Insufficient chilling can lead to delayed and desynchronized budding and slowing of crop development (Londo and Johnson, 2014, Fraga et al., 2019). Soil water content is also known to affect budbreak and further occurrence of phenological stages (Ramos and Martínez-Casasnovas, 2006, Van Leeuwen et al., 2009). Therefore, modeling the interactions of temporally varying components is required to fully understand between-season dynamics of grapevine vegetative and reproductive attributes.

Space-time distribution of grapevine growth and yield components can be managed during the growing season by irrigation (Bellvert et al., 2020). Irrigation practices are intended to control the water stress of the grapevines through management of soil water availability, and consequently the vegetative growth, physiological conditions, yield components, and wine quality (Santesteban et al., 2011b, Munitz et al., 2016). However, irrigation management is highly dependent on the amount, timing and intensity of precipitation. Wilson et al. (2020) reported that changes in irrigation amounts have an immediate effect on leaf area growth, and that lower levels of soil moisture due to lower amounts of precipitation are more likely to generate a stronger response of the plant to the irrigation regime. Nolz et al. (2016) showed the interactive effects of precipitation and irrigation levels on soil water content according to precipitation amounts, soil depth, and dryness of the soil prior to infiltration of precipitation/irrigation water.

Various studies illustrated the effects of the spatial factors in the field on vegetative, physiological, and reproductive attributes (Acevedo-Opazo et al., 2008, Ohana-Levi et al., 2019, Yu et al., 2021). Many works were conducted to quantify the impact of precipitation amounts, timing, and intensity on water availability in vineyards and consequently on grapevine performance (Ramos and Martínez-Casasnovas, 2009, Camps and Ramos, 2012, Ramos, 2017, Agam et al., 2019). However, to the best of our knowledge no studies focused on grapevine response to multivariate effects of spatial field attributes and temporal meteorological and controlled factors. The innovation of this present research stems from two main notions. The first focuses on generating an approach for applying an integrative space-time analysis of the response of grapevine attributes to spatially and temporally varying field properties. The second notion is to provide an overview on the interactions of factors affecting water availability, including direct water sources such as precipitation and irrigation, and indirect contributors, such as terrain properties that affect water distribution in the field. The suggested approach was established using a multivariate dataset collected during four years from a “Sauvignon Blanc” vineyard, enabling a long-term perception of processes and dynamics over space.

The main objective was to model the response of grapevine vegetative and reproductive attributes to water availability in space and time. The specific objectives were to (1) define, measure, and select the spatial and temporal variables to include in the model; (2) quantify the spatial autocorrelation of the spatial variables and grapevine vegetative and reproductive attributes; and (3) determine the relative influence (RI) of the different covariates on grapevine vegetative and reproductive attributes.