Among environmental constraints, water scarcity is probably the most important threat all over the world (Jury and Vaux, 2005; Kijne, 2006; Berger et al., 2010), compromising agricultural production at several latitudes (FAO, 2018). Rapidly, the possibilities to generate and to manage new water sources for agriculture will be limited, and the instability of water resources will not only be detrimental to crop productivity, but will also generate substantial socioeconomic impacts (Postel, 2000; Polade et al., 2017). Considering the competition for water use among agriculture, human consumption, and industrial sectors, crop water demands could double by 2050, whereas the availability of freshwater is predicted to drop by 50%, owing to global climate change (CC; Gupta et al., 2020). Viticulture production is not an exception, since future projections of CC-driven changes or “climate crisis” suggest a lack of water to maintain current levels of production in all regions of the world, which will particularly impact Mediterranean ecosystems (Hannah et al., 2013). Furthermore, suitable zones for grapevine production based on temperature may be greatly affected in the Mediterranean regions (Santillán et al., 2019). Viticulture adaptation to CC in these regions (as most crops do) will require integrated strategies and major adaptive levers to cope with water availability and grapevine productivity and an increase in evapotranspiration that encompass different levels of organization: the crop (cultivar and rootstock), the cropping system (management techniques used), and the farming system, including farmers (del Pozo et al., 2019; Naulleau et al., 2021).

A recent systematic study identified current knowledge to evaluate adaptation strategies in the main vineyards worldwide (Naulleau et al., 2021), whose findings were as follows: (1) evaluation of a combination of adaptation strategies provides better solutions for adapting to CC; (2) multi-scale studies allow local constraints and opportunities to be considered; and (3) only a small number of studies have developed multi-scale and multi-lever approaches to quantify feasibility and effectiveness of adaptation (Naulleau et al., 2021). For instance, ecophysiological studies have contributed to maximize the use and productivity of water, i.e., irrigation technification, regulated deficit irrigation (Hsiao et al., 2007), soil mulch, and optimization of the orchard density and architecture (Ripoche et al., 2010). However, in mid-term, it will be necessary to incorporate increasing productivity, fruit quality and disease tolerance, criteria of water productivity [WP; g (MS)/mm (H₂O) transpired], and water stress tolerance in cultivars and rootstock breeding (Warschefsky et al., 2016; Simonneau et al., 2017; Gupta et al., 2020). Thus, a water shortage scenario will demand to assess interaction effects of cultivar x rootstock x environment and their impacts in fruit attributes associated with its quality (Ibacache et al., 2016; Cochetel et al., 2017; van Leeuwen et al., 2019; Villalobos-González et al., 2019).

Under water deficit conditions, perennial plants display a continuum of mechanisms for dealing with low water availability between two edges: 1) drought evasion, which is found in species bearing high stomatal sensitivity (near-isohydric), or 2) drought tolerance, found in species bearing low stomatal sensitivity response to the ambient (near-anisohydric) and functional and morphological traits toward adaptation such as osmo-regulation (Schultz, 2003; Blum, 2011; Gupta et al., 2020). Stomatal closure is linked to systemic signaling from the roots rather than the shoot, as evidence that root physiological status plays a key role in controlling the shoot behavior (Lawlor, 2002). Traditionally, the plant hormone abscisic acid (ABA) plays an essential function as a phytochemical signal involved in the shoot-root communication, because of drop in soil water potential (Zhang et al., 2006). It is also considered the most prominent player in drought stress, directly affecting stomatal conductance (gs) at the guard cell level (Gupta et al., 2020). The physiological and molecular mechanisms driving the ABA effects are yet to be summarized (Gambetta et al., 2020).

In grapevines, the hydraulic and biochemical modes of stomatal regulation are interdependent, making a strict division between them extremely challenging both theoretically and experimentally (Medrano et al., 2003; Ollat et al., 2016; Buckley, 2019; Gambetta et al., 2020). Moreover, the relative contribution of these mechanisms is still unknown and likely dependent on genotype and environment (Lovisolo et al., 2016; Coupel-Ledru et al., 2017; Hochberg et al., 2018; Dayer et al., 2020). It has also been demonstrated that leaf hydraulic conductance (K_leaf) was downregulated by exogenous ABA, with strong variations depending on the genotype (Coupel-Ledru et al., 2017). Interestingly, variation between isohydric and anisohydric genotypes correlated with K_leaf sensitivity to ABA, with K_leaf being unresponsive to exogenous ABA in the most anisohydric genotypes (Coupel-Ledru et al., 2017). Recent work suggests that all genotypes regulate stomatal conductance to protect against more severe damage, such as petiole or leaf cavitation and leaf shedding (Dayer et al., 2020). However, it is not clear to what extent differences in regulation of vine water use between cultivars result from innate genotypic differences or environmental factors (Hochberg et al., 2018). Likewise, grapevines appear to almost always operate within a “safe” margin of water potentials in which stem cavitation is extremely rare (Charrier et al., 2018). Still, the exact vine mortality thresholds are still unknown. Thus, numerous gaps remain in understanding of what really configures a drought-adapted grapevine cultivar, making it difficult to robustly address future climate challenges (Gambetta et al., 2020). Also, unraveling mechanisms to explain how the regulation of aerial (cultivar) drought tolerance could be enhanced by a root-driven feedback regulation, where the role of rootstock might be crucial to determine such responses and sustain productivity, pointing toward an integrative approach that digs into the complexity of cultivar x rootstock x environment interaction (Franck et al., 2020).

Most of the worldwide vineyards are grafted on commercial rootstocks which are hybrids of mostly three species, namely, Vitis berlandieri, V. riparia, and V. rupestris, that were developed before 1930 from American Vitis species to control phylloxera damage (Serra et al., 2014; Berdeja et al., 2015). Rootstocks also provide support for cultivation under challenging soil conditions, including the presence of nematodes and insects, high salinity or active lime, and drought (Meggio et al., 2014; Serra et al., 2014; Walker et al., 2014). Limited long-term information on rootstock effects over yield and its components are available. Nevertheless, it is established that the response is primarily associated with the vigor level conferred to the scion by the rootstock (Dry and Loveys, 1998). This influences bud fruitfulness and vine productivity (Satisha et al., 2010; Ibacache et al., 2016). To sustain grapevine productivity and quality in CC warming, an increase for irrigation water will occur, generating big freshwater demands, considering the low adoption of measurements of RDI or PRD. Therefore, agricultural adaptation efforts that anticipate these multiple possible effects in Mediterranean agroecosystems are needed (Hannah et al., 2013; Wolkovich et al., 2018). By assessing the current effects of CC on V. vinifera, it is key to understand the plasticity associated with the ability to uptake water from soil in the continuous root/vine/environment (Santillán et al., 2019; van Leeuwen et al., 2019).

Although scion-rootstock interactions in drought tolerance have been studied (Serra et al., 2014; Tomás et al., 2014; Bianchi et al., 2020), the diversity of rootstocks adapted to dryer conditions is limited. Recent studies in rootstock effects were evaluated, and differences in fruit yield, pruning weight, budburst, fruit set, bunch weight, berry weight, berry diameter, and rachis weight between nine rootstocks in semiarid conditions were determined along with their effects on nutrient uptake (Ibacache et al., 2016, 2020). Therefore, grapevine rootstocks will undoubtedly play a fundamental role in the adaptation to future CC, especially to water shortage (Serra et al., 2014; Ollat et al., 2016; Delrot et al., 2020) and to improve WP, but mechanisms driving these processes are still elusive. In a grapevine meta-analysis contrasting stomatal conductance in response to water availability, rootstock genotype explained the greatest contribution to variability (19.1%) followed by the scion genotype (16.2%) (Lavoie-Lamoureux et al., 2017). Moreover, the effect of soil water-holding properties was analyzed and showed a scion-dependent effect which was dominant over rootstock effect in predicting gs values. Overall results suggest that a continuum exists in the range of stomatal sensitivities to water stress in V. vinifera, rather than an isohydric—anisohydric dichotomy, which is further enriched by diversity of scion-rootstock combinations and interactions with soils and intensity of water deficits (Levin et al., 2020).

In Chile, naturalized rootstocks were collected from arid regions of Northern Chile (Milla-Tapia et al., 2013) and studied in response to water deficit. Some selected genotypes induced significantly higher tolerance for morpho-physiological traits irrespective of scion and seasons associated with higher root growth at early stages (Franck et al., 2020). Further transcriptomic analysis was performed, determining that major differences in transcriptional behavior occurred at root level, suggesting scion-driven transcriptional regulation in response to water deficit (Franck et al., 2020). Despite the importance of grapevine phenology, studies on the effect of rootstock on the development of phenological stages are scarce in literature and have been carried out in few varieties and under different edaphoclimatic conditions (Loureiro et al., 2016; van Leeuwen and Destrac-Irvine, 2017; van Leeuwen et al., 2019).

To understand the effects and expected impacts of CC warming due to water deficit on grapevine productivity and fruit maturity, we conducted a multi-rootstock approach using two contrasting cultivars in regard to hydric behavior grafted on selected naturalized rootstocks to assess the array of response as study model for understanding adaptive responses that might confer better drought adaptation to specific clone/rootstock combinations on vegetative and fruit expression in hyper arid environment.

This study was designed to determine to what extent the plant water relations are modified by using naturalized rootstocks and whether this fit primary productivity and adaptation to harsh conditions typical of a hyper-arid zone that is expected to occur in Mediterranean regions due to CC (Morales-Castilla et al., 2019; IPCC, 2021). Considering the observed decline of precipitation over central Chile, which has been greatly accentuated by an uninterrupted sequence of dry years from 2010 to the present, with annual rainfall deficits ranging between 25 and 45% (Garreaud et al., 2020), the 50% deficit irrigation was a feasible projected decrease. This ongoing, multiyear dry spell has been referred to as the Central Chile Mega Drought (MD) due to its unprecedented longevity and large spatial extent in the historical record (CR2, 2015). Nevertheless, the phenological, physiological, and productive responses necessarily respond to the climatic conditions that are provisional between the growing seasons. Within the productive variables, grapevine phenology is fundamental for the planning of agricultural practice within the fields. For example, irrigation, fertilization, and the application of phytosanitary and hormonal products are programmed based on the date of occurrence of phenological states such as budburst, flowering, fruit set, and veraison. Indeed, the main factor that affects grapevine phenology is related to temperature (Parker et al., 2011). In this sense, the “season” factor, related to different climatic conditions (Table 1), was the most significant for the different phenological stages (Supplementary Table 1). The growing degree days accumulation (between budburst and harvest) did not vary between seasons (1,300–1,400 heat units), being like those reported in other wine-growing zones of Chile for the cv CS (Verdugo-Vásquez et al., 2016). Regarding the effect of the use of rootstocks and irrigation on the grapevine phenology, there is little information in literature (Keller et al., 2012; Sabbatini and Howell, 2013). It has been reported that modifications in the date of occurrence due to the use of rootstocks are related to indirect scion response, such as canopy density, as influenced by the rootstock's direct impact on scion vigor (Sabbatini and Howell, 2013). In this study, there were specific effects at the beginning of the season (budburst) for the cv CS (differences <5 days, Season S2), but the differences observed at the beginning of the season were not maintained throughout the season without significant differences for flowering, veraison, and harvest. For cv Sy, the differences due to the use of rootstocks were at the end of the season (harvest), but were also not consistent between seasons. It was observed that there is no consistency in the results between cultivars and seasons, which is why long-term studies are required to determine whether the use of rootstocks allows advancing or delaying phenological stages stably and the mechanisms by which differences are generated. For the “irrigation” factor, there were more consistent results between cultivars and seasons, wherein the decrease in irrigation (T1) advanced harvest date (at the same level of total soluble solids) for both cultivars. This advance was related to the accumulation of sugars, rather than differences in the beginning of the ripening period (veraison) since there were no differences due to the irrigation factor for budburst, flowering and veraison. At the within-field scale, it was determined that soil conditions (slope and total soil water availability) can affect grapevine phenology at the beginning of the season (budburst), and is associated with differences in the initial soil moisture (Li et al., 2016; Verdugo-Vásquez et al., 2021b). However, in this study, despite the differences in the water applied since the beginning of the season and the absence of rain in winter, there were no significant differences due to the irrigation factor for budburst.

Regarding fruit maturity, the “rootstock” factor only modified some specific dates and was not consistent between seasons. Similar results were observed by Keller et al. (2012), as they reported that scion effects and differences due to yearly climate variation far outweighed any differences due to rootstock for fruit maturity. On the other hand, the “irrigation” factor was more consistent in the results, showing that the decrease in irrigation increases the accumulation of sugars in the berries and decreases the titratable acidity, being the earliest harvest, as mentioned above. These results coincide with those reported in literature (Acevedo-Opazo et al., 2010; Cabral et al., 2021; Pérez-Álvarez et al., 2021), according to the time and intensity of water stress (Romero et al., 2022). The differences observed in maturity were more related to the general balance of grapevine (Ravaz index) than to berry weight in this study (Table 5).

Previous studies have also experimented with significantly different seasons, which predominate over the physiological parameters, such as Ψ_stem and gas exchange, that present dynamic fluctuations itself (Bascuñán-Godoy et al., 2017; Buesa et al., 2017; Romero et al., 2018; Levin et al., 2020). However, there is a high degree of co-regulation in the plant to cope with water deficits through their stomata. Since stomatal closure and regulation of leaf gas exchanges with the atmosphere, it is a key process in response to moderate water deficits both in the soil and in the atmosphere, therefore integrating internal signaling and environmental cues and complex genetic control and thus providing multiple layers of regulation to balance water loss and CO₂ assimilation in dry environments (Chaves et al., 2016; Levin et al., 2020). In a physiological framework, several studies have already reported the different degrees of responses between cultivars, of which CS and Sy display contrast hydric strategy (Hochberg et al., 2012; Franck et al., 2020; Levin et al., 2020). In this fashion, the meta-analysis showed the extent of reductions in physiological parameters measured at T1 in contrast to T0. Thus, when integrating the effect magnitude of water deficit in CS self-grafted plants, it was determined that the decline of g_s was greater than that of the A_n. On the contrary, in Sy self-grafted plants, the decline magnitude was greater in A_n than in g_s (Figure 1). On the other hand, the rootstocks alleviated the percentual reductions of A_n caused by T1, which agreed with what was reported by Franck et al. (2020), who observed that the naturalized rootstocks grown in pots and under field conditions displayed a better performance.

During water deficit, photosynthesis is limited by both stomatal closure and impairment of the photosynthetic machinery (i.e., metabolic factors) in order to prevent dehydration, during which root system is the major interface between the plant and water availability of a drying soil (Gambetta et al., 2020). Thus, the rootstocks can contribute to the scion water loss through a combination of hydraulic and hormonal root-to-shoot signaling (Lovisolo et al., 2010). In this sense, the relationship of A_n vs. g_s observed in two contrasting cultivars suggested that different rootstock genotypes (140 Ru, R32, R70) changed the A_n sensitivity to g_s variations given the coefficients r obtained (Figure 2). Moreover, in the case of CS, the g_s values (mol H₂O m² s⁻¹) in which the A_n reaches the “Plateau” increased with the use of rootstocks (0.17 SG <0.19 R32 <0.22 R70 <0.26 140 Ru). Meanwhile, Sy decreased with R70 (0.23 mol H₂O m² s⁻¹), remained with 140 Ru (0.26 mol H₂O m² s⁻¹), and increased with R32 (0.29 mol H₂O m² s⁻¹), suggesting a modulation effect. Considering A_n and g_s behavior, and from an efficiency perspective (WUE_i), rootstock may also modulate the variability in response to stomatal closure (correlation r), where calculated slopes demonstrated that 140 Ru and R32 were the most conservative for CS and Sy, respectively, in relation to WUE_i—g_s through regressions lines, enabling to compare the slopes between genotypes to highlight environmental and genetic differences (Tortosa et al., 2019).

It has been described that the regulation of stomatal closure is mediated by hydraulic, chemical, physical, and even electrical signals (Beis and Patakas, 2010). Among these signals, an integrated modeling approach suggested that both hydraulic and chemical signals are likely important for the rootstock-specific stomatal regulation. In addition, the coupled chemical-hydraulic factors most precisely describe the stomatal conductance underlying gas exchange of grafted grapevines, since factors controlling ABA biosynthesis (either in leaves or roots, or root system architecture) caused differences in the hydraulic conductance between the rhizosphere and the soil–root interface (Peccoux et al., 2018). In this sense, this study provides results that reinforces the previous conclusions obtained in which the hydraulic variability between grape cultivars is in turn strongly influenced by the ambient VPD (Villalobos-González et al., 2019; Gambetta et al., 2020). In the face of duration and intensity of water stress given by SIΨ, CS was sensitive to the accumulated effects of deficit irrigation at the end of the growing seasons. A naturalized rootstock R32 showed the least stress, even surpassing 140 Ru, which has been shown to be drought tolerant (Romero et al., 2018). Meanwhile, the SIΨ of Sy was indifferent to deficit irrigation in much of the experimental period. The SIΨ values agrees with what was reported by Zúñiga et al. (2018). However, the stress levels in this study were higher than those previously reported and are explained by the conditions of greater aridity in the experimental site as a natural laboratory for CC adaptation studies.

Nowadays, water use (i.e., the water consumed) and WUE (i.e., the efficiency of this consumed water to assimilate carbon, produce biomass, or fruit yield) are crucial parameters, especially in areas with increasing water scarcity that requires adaptation to CC, such as south Europe, West Asia, Western Australia, Chile, North Africa, and parts of South Africa (FAO, 2018; van Leeuwen et al., 2019; Naulleau et al., 2021). Despite the importance of rootstocks for the total water productivity and WUE of the crop, the variability of WUE in rootstocks has largely been underexplored, in addition to obtaining contradictory results (reviewed by Medrano et al., 2018). Moreover, the main role of rootstocks in plant water economy leads to the consideration of the genetic variability of WUE as a complementary target for current research (Medrano et al., 2018). In this regard, differences based upon the interaction among cultivar scion: commercial rootstock support the necessity of exploring differences both in cultivar hydric behavior and the interaction with the rootstocks, leading to change in crop performances depending on the cultivar (Tortosa et al., 2019; Franck et al., 2020; Romero et al., 2022). Thus, a tolerant rootstock, such as 140 Ru, was superior in the hydric conservative CS vines, whereas R32 overcame the 140 Ru performance in Sy vines, which were measured in traits such as water productivity, trunk circumference, and yields and pruning weights (Figure 6). Latter results may indicate enhanced abilities for water uptake and assimilation since near-anisohydric cultivars bears a “risky strategy” of water use. Indeed, 140 Ru exhibited better adaptive behavior as CS vines maintained slight unchanged A_n, g_s, and WUE_i levels under both well-watered and persistent water stress conditions. This might be associated to greater root water-uptake capacity and whole-plant hydraulic conductance, translating to high productivity and vigor (Romero and García-García, 2020; Romero et al., 2022). With respect to normal rainfall scenarios which could represent the T0 irrigation treatment (100% ETa applied), it was observed that rootstocks behavior differed in each cultivar under these conditions. For cv CS, 140 Ru was the one that presented the best behavior regarding the productive variables analyzed (N° bunches per plant, pruning weight, trunk circumference, among others), while R32 was the one that presented the best behavior with respect to the productive variables analyzed for cv Sy (yield, trunk circumference, water productivity, among others). These results highlight the need to carry out studies that consider different rootstock-scion combinations under different edaphoclimatic conditions.

In concert, CS displayed a conservative behavior in terms of lower g_s and high rates of A_n-enhanced WUE_i (Zamorano et al., 2021). Conversely, cv Sy showed increased WUE_i due to reduced g_s near-anisohydric behavior. This was characterized by less prone to rapid stomatal closure under water stress maintaining higher stomatal aperture and exhibiting substantial reductions in Ψ_stem. Thus, adjusting locally with lower stomata sensitivity to drought-induced ABA and being more dependent on hydraulic signals such as leaf water status (Coupel-Ledru et al., 2017). In this regard, the integral water stress observed in this study highlighted this condition since cv Sy was the one that accumulated more stress. This cultivar is efficient in terms of water use but shows limited heat dissipation due to its reduced g_s, which can favor the occurrence of leaf sunburn under severe heat stress and drought (Costa et al., 2012). Other authors have demonstrated the mechanisms by which rootstocks modify the mentioned productive variables and found that the grafted plant modifies the absorption of light, increasing the assimilation of carbon compounds and therefore increasing the yield (Corso et al., 2016; Bascuñán-Godoy et al., 2017).

Recently, a signaling communication peptide has been identified in response to drought, where CLE25 peptide is produced in the roots and moves systemically through the plant vasculature to leaves to drive ABA production by activating biosynthetic enzyme NCED3 (Takahashi et al., 2018). This burst of ABA synthesis leads to stomatal closure and improved water balance, thereby promoting drought survival. This insight into small-peptide signaling in Arabidopsis may help to unravel conserved mechanisms in crops for root-to-shoot mobilization of stress signals (Gupta et al., 2020). Hence, the interaction with the rootstock and the ability of this to later explore and uptake water is fundamental for efficiently facing water deficit. Many traits and mechanisms are involved in the response of a rootstock by scion combination to water demand/water availability ratio. Hence, determining the optimal combination may enhance this adaptative processes. Rootstocks can differ by their capacity to extract water from the soil, which is primary linked to root biomass and to the hydraulic conductivity of the roots (Gomès et al., 2021). Lately, adaptation of viticulture requires a proper exploration of an optimal interaction of cultivar and rootstock, particularly when exploring new geographical areas, new training systems, new management practices, or new varieties, both for rootstocks and scions (Gomès et al., 2021). There are additional studies of water productivity in vineyards, but knowledge of the effect of irrigation reductions and its combined effect with grafted vines using rootstocks on secondary metabolism is still growing (Cáceres-Mella et al., 2018).

Hence, adaptations to CC require modifications in genotypes and viticulture techniques that can influence both phenology expression and grape ripening since rootstocks are able to trigger transcriptional changes on berry secondary metabolism which is relevant for berry composition and sensory properties (Berdeja et al., 2015). Moreover, studies for rootstocks conferring higher drought tolerance to the scion, driving carbon flux toward both accumulation of phenolic compounds, and alteration of anthocyanin profile, thereby altering grape quality at harvest, might be a target (Zombardo et al., 2020). In this regard, rootstocks that have novel drought tolerance mechanisms (i.e M4 rootstock) have shown greater synthesis of phenolic compounds such as stilbenes and flavonoids with enhanced capacity to scavenge and regulate the reactive oxygen species (ROS) levels that are generated under stress conditions and cause oxidative damage (Corso et al., 2015). In the naturalized R32 rootstock tissues grafted to cv Sy, the upregulation of genes of Phenylpropanoid metabolic process and pigment accumulation was determined in response to water deficit (Franck et al., 2020), enhancing plant survival in the presence of abiotic stress. M-rootstocks have also displayed adaptive traits, such as reducing the stomatal conductance and stem water potential while maintaining high photosynthetic activity with high Water Use Efficiency in water-limiting conditions (Bianchi et al., 2020). The capacity of M4 to satisfy the water demand of the scion under limited water availability has shown a delayed stomatal closure, allowing higher photosynthetic activity, which is also related to a reduced activation of ABA signaling both in the root and the leaf level (Prinsi et al., 2021). Therefore, the use of drought tolerant genotypes (scion, clones, and rootstocks) represents an environmentally friendly and cost-effective tool for adaptation to a changing climate (van Leeuwen et al., 2019).

Rootstock did not modify the main phenological stages, while irrigation treatment allowed modifying the harvest date. Moreover, harvest date and acidity were modified by deficit irrigation treatment and rootstocks did not modulate phenological stages. Adaptation of grapevines to expected lower water availability might be improved by using suitable tolerant rootstocks. In addition, maturity index can be modified through irrigation management.

The regulation and behavior of several physiological parameters related to the plant water status were contrasting among the cultivars studied. In turn, this behavior varied depending on how stressful the environmental conditions were between growing seasons. Under water deficit conditions, when photosynthesis was mainly limited by stomatal conductance, rootstocks showed the ability to adjust the sensitivity by which photosynthesis was restricted.

The dynamic responses and grapevine adaptation to water deficit were highly dependent on the cultivar hydric strategy (near-isohydric or near-anisohydric) and the interaction with the rootstock. Hence, the vine fitness performance will be determined by new environmental demands that will be imposed by climatic challenges, and such growth and developmental responses to drought, higher temperatures, or combined abiotic stresses will rely on the proper combination of both cultivar and rootstock.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AZ-S conceived and designed the experiments. EV-S conduced the physiological analyses and laboratory determinations with the guidance of AZ-S, NV-V, and ID. AZ-S wrote the manuscript with the help of EV-S, NV-V, and ID. EV-S, NV-V, and AZ-S prepared all the figures. All authors contributed to the discussion of ideas, revised, and approved the final manuscript.

This work was supported by Consejo Nacional de Ciencia y Tecnología CONICYT–Fondecyt Regular (Grant No. 1140039 2014/INIA) to AZ-S, and by Agencia Nacional de Investigación y Desarrollo ANID—Postdoctoral Fondecyt (Grant No. 3180252 2018/INIA) to NV-V.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

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