WATER AND WASTEWATER MANAGEMENT FOR SUSTAINABLE VITICULTURE AND OENOLOGY IN SOUTH PORTUGAL – A REVIEW

.


INTRODUCTION
Water use in viticulture and oenology demands improved quantification to support present and future adaptation strategies of the wine industry to climate change, while increasing water use efficiency and minimizing environmental burdens (Chiusano et al., 2015;Costa et al., 2016;Martins et al., 2018).The challenges posed to the wine industry extend to wastewater (WW) generation and management in both viticulture and oenology phases (Castex et al., 2015;Peth et al., 2017).This is particularly important for Mediterranean regions, characterized by hot dry summers and mild winters, and increasingly exposed to extreme climate events (Fraga et al., 2018;Lopes et al., 2018).
Portugal is a typical Mediterranean wine producer, with around 191,000 ha of vineyards and a wine production of about 6.5 MhL in 2017 (IVV, 2018).It is the 11 th largest producer worldwide and the 9 th largest amongst global wine exporters (OIV, 2018).
The wine-producing region of Alentejo (South Portugal) is one of 14 Portuguese wine regions.It has a total vineyard area of 23,879 ha and accounts for approximately 12.5% of the country's total wine grape production area (IVV, 2018).Most of the Alentejo region has a Csa climate, according to the Köppen-Geiger updated classification (Peel et al., 2007), characterized by warm and dry summers (IPMA, 2017).Precipitation is very low or absent in summer and air vapour pressure deficit (VPD air ) can be extremely high (up to 8 kPa), generating high crop evapotranspiration losses (COTR-ATEVA, 2009; Barroso et al., 2017).Bioclimatic indices such as the Dryness Index (DI) (Riou et al., 1994), which help to assess regional suitability for wine production and predict climatic impact on viticulture, show a tendency for increasingly severe drought conditions in Alentejo (Figure 1) (Fraga et al., 2018;Lopes et al., 2018).This is in line with the fact that the region is experiencing more frequent heat waves, which poses increasing risks to the sector (Lopes et al., 2018;Silvestre et al., 2018).In this context, irrigation emerged as a major tool to overcome the constraints posed by adverse climate conditions and support risk management, e.g.due to heat waves (Silvestre et al., 2018).Consequently, the total irrigated vineyard area in Alentejo is nowadays largely above 10,000 ha (more than 50% of total surface) (Costa et al., 2019).This was a fast increment, if we consider that in 2002 there were only 4,600 ha with irrigated vineyards (CVRA, 2002), and the trend should remain, supported by the Alqueva dam (EDIA, 2018).(Riou et al., 1994), for (a) Portalegre, (b) Évora and (c) Beja (Alentejo region, South Portugal) using datasets from local weather stations.DI = Wo + P -Tv -Es, where Wo is the soil water reserve at April 1 (assumed as 200 mm); P is the accumulated precipitation (mm); Tv is the vineyard's potential transpiration (mm); Es is the soil water loss by evaporation (mm).The DI estimates soil water availability taking into account vine's transpiration, soil evaporation and precipitation between April 1 and September 30 (in the Northern Hemisphere).The box-plots are bounded on top by the 3 rd quartile (Q3), on the bottom by the 1 st quartile (Q1) and divided by the median (Q2).The average value of the series is represented by (x) and (o) indicates the DI values.The top whisker is defined as Q3+1.5xIQR and the lower whisker as Q1-1.5xIQR,where IQR=Q3-Q1.
Water accounting in irrigated agriculture is often characterized by poor metrics and scarce data reporting (Corbo et al., 2014;Santiago-Brown et al., 2015;Liu et al., 2017;Pfister et al., 2017;EU Commission, 2017).In addition, indicators to assess the sustainability of the wine industry vary between regions and countries, making it difficult to compare "companies/farms" performance (Santini et al., 2013; b) Corbo et al., 2014;Oliveira et al., 2019).In Portugal, recent efforts in the wine sector to improve its environmental sustainability are reported in several studies (Neto et al., 2013;Quinteiro et al., 2014;Engel et al., 2015;CVRA, 2016), but information remains scarce.In addition, auditing for environmental performance demands more precise water use monitoring in farms and wineries, as well as data on wastewater production and/or quality (EPA, 2004).In parallel, more homogeneous standards/metrics for sustainability, independently of the terroir are on demand (Costa et al., 2016;Oliveira et al., 2019).
Therefore, the major aims of this review are to point out the risks of increasingly dry conditions for irrigated viticulture and oenology in South Portugal, while emphasizing the role(s) of improved water metrics for more efficient water use and WW management in the vineyard and in the winery.

WATER USE AND WASTEWATER PRODUCTION IN VINEYARDS AND WINERIES
A strategic approach towards more sustainable water use in dry regions' wine production must involve the combined assessment of water needs and WW production/management in both vineyards and wineries.This is highly relevant for Mediterranean regions, e.g.South Portugal, where values of 88 to 264 L of water can be spent per litter of wine produced (Engel et al., 2015).More recent literature points to even higher values of annual water footprint, reaching 360 L per 0.75 L bottle (450 L per litter of wine), as referred by Saraiva et al. (2019) for South Portugal wines.The Portuguese wine industry is based on small to medium size vineyards and small wineries, with 98% of them having production volumes below 2000 hL (IVV, 2016).Therefore, a "Farm-Winery" approach, in which vineyards and wineries are considered in a single stream grape system, will represent the majority of Portuguese viticulture and oenology stakeholders and can be used to support the analysis of water use and management in the Portuguese vine and wine sector (Oliveira et al., 2019).

Water use and wastewater production issues in the vineyard
In dry areas, viticulture accounts for the highest fraction of water used in wine production, reaching values comprised between 70 to almost 90% of total water use in some cases (Ene et al., 2013;Correia, 2015;Saraiva et al., 2019).Despite grapevine being considered well-adapted to dry conditions, irrigated cultivation uses a considerable water volume per season under Mediterranean conditions (EDIA, 2018; Table I).
Water use in the vineyard depends on several factors: 1) Plant -canopy leaf area and canopy gas exchanges, stem and root morphology, plant hydraulics, phenology, genotype; 2) Soil -depth, texture, organic matter content, water availability, soil temperature (T s ); 3) Atmospheric conditions -rainfall (R), air vapor pressure deficit (VPD air ), air temperature (T air ), wind speed (W s ); 4) Agronomic choices and practices -soil preparation, plant density, rootstock/variety combination, row orientation, floor management, training system, canopy management (Figure 2).
In the Alentejo wine region, irrigation needs are expected to vary between 2,500-3,000 m 3 /ha per year for an average production of 7.5 to 10 t/ha (EDIA, 2018).Nevertheless, these indicative values are 30-50% higher than the ones observed under deficit irrigation, which is already being implemented in commercial vineyards in the region for some years (see Table I).
More sustainable grapevine irrigation should supply enough water, at the right moment, to guarantee a profitable yield and winemakers' desired berry composition without compromising vine longevity.Deficit irrigation strategies should be based on a precise monitoring of atmospheric conditions, soil water content and plant water status.In addition, Alentejo has a wide annual climate variability (see Figure 1 and Table I), with increasingly drier winters and warmer temperatures in recent years, which pose additional risks to the wine sector in the region.Even though the degree of corporatization has been increasing in the region, and with it the adoption of better management strategies (INE, 2013;2016), there is still a large variation among farms regarding irrigation efficiency, with timing and volume of irrigation being decided mostly through experience and observation, instead of atmosphere/crop/soil monitoring and precise water metrics (COTR-ATEVA, 2009; Levidow et al., 2014).Audits to water use in farms and wineries are also not a widely adopted practice (Radke et al., 2015).This suggests that a wider network of simple monitoring sensors, more technical support and broader implementation of sustainability programs for the sector, will be a step forward to optimize water management at company and region level.

Table I
Climate conditions and irrigation water volumes applied under sustained deficit irrigation (SDI) conditions implemented in a field trial at a commercial vineyard in the Alentejo winegrowing region (Reguengos) along three consecutive years (Costa et al., 2019).Average rainfall (mm); Reference Evapotranspiration (ETo); Tair -air temperature; Irrigation volume: water volume applied during the irrigation period; SDI management was based on vine leaf water potential threshold of ~0.4 to 0.5 MPa Clima e volumes de água de rega aplicados ao longo de três anos consecutivos em condições de rega deficitária sustentada (SDI) implementada num ensaio conduzido numa vinha comercial da região do Alentejo (Reguengos) (Costa et al., 2019).Precipitação média (mm); Evapotranspiração de referência (ET o ); T air -temperatura do ar; Volumes de rega: dotação aplicada durante o período de rega.A gestão da rega deficitária foi baseada em limiares na medição do potencial hídrico foliar de base (limiares entre ~0.4 a).) 0.5 MPa.A much less described item in conventional viticulture is the volume of WW generated in the vineyard (WW vineyard ), mainly related to spraying of plant protection products (PPP) and/or improper disposal/management of PPP residues (Doruchowski et al., 2014;Otto et al., 2015).The WW vineyard may represent a minor percentage of the overall volume of water used in irrigated vineyards, but its impact on soil and groundwater contamination can be significant.For example, Rochard and Codis (2004) indicate that for a volume of 108 to 1400 L of water used per spraying treatment, a total amount of 41 L and 354 L of WW can be generated, respectively.In turn, the volume of clean water used to wash sprayers has been shown to vary between 95 L and 190 L per day of spraying application (IFV, 2010).Such range of variation can be due to technical and operational factors, such as spraying frequency, sprayers' dimension, operators' experience, or cleaning equipment specificities (e.g.high pressure cleaning saves about 50% of water relatively to conventional sprayers), but also to the characteristics of the terroir (IFV, 2010).The WW volumes derived from washing protective clothing and empty pesticide containers should be accounted as well, but data is often unavailable.

Year
The wine industry is strongly committed to reduce the amount of PPP in viticulture (EIP-AGRI, 2019), namely by adoption of Integrated Crop Management (ICM) and biological control measures and/or novel dosing methods based on dynamic variation of canopy leaf area, phenology and climate conditions (e.g. based on modelling software and decision support systems) (Kuflik et al., 2009;Gil et al., 2011Gil et al., , 2014;;Pérez-Expósito et al., 2017).For example, it is now possible to optimize the applied volume of PPP as function of canopy area/volume, resulting in a more precise site-specific application, with considerable water savings (Gil et al., 2014;Campos et al., 2019).Nevertheless, this occurs mostly in large-scale operations and companies.

Water use and wastewater production issues in the winery
Water use in the winery is mainly related to cleaning operations -equipment, tanks, vats, barrels, presses, de-stemmers, reception hoods, as well as taps, floors, walls and pipes (Pirra and Bianchi, 2007;Andreattola et al., 2009;Oliveira et al., 2019).Water consumption depends on type and size of the winery, the type of wine produced and the adopted winemaking technology (Andreattola et al., 2009;GWRDC, 2011).
The maximum water consumption in the winery occurs during vintage and in the 1 st racking period (Duarte et al., 2004;Oliveira and Duarte, 2016;Oliveira et al., 2019).The WW flows are proportional to the vintage duration (GWRDC, 2011).Duarte et al. (2004) and Oliveira and Duarte (2016) proposed a simpler approach in which two or more activities are combined.They consider two periods: vintage and 1 st racking (Period I) -characterized by high peak flows and high pollution loads -and the remaining activities e.g.bottling, are encompassed in Period IIcharacterized by reduced water flows and medium/low pollution loads.The lack of distinction between these two periods may explain the large variation reported in literature for the volumes of water used in wineries (L water/L wine produced) (Table II).
Similarly to vineyards, WW production in wineries remains not fully characterized.Data collected in Alentejo shows that 60-86% of the total WW is produced in Period I (Oliveira and Duarte, 2015), which is similar to Italian wineries, where 78% of the WW is generated during vintage and 1 st racking (Lofrano et al., 2009).Likewise, WW characteristics need a better assessment, as they depend on the working period and the type of winemaking technologies used.
The winery WW contains grape pulp, skins and seeds (Devesa-Rey et al., 2011;Oliveira and Duarte, 2016).This complex matrix of different flows comprises suspended solids, readily biodegradable compounds (fructose, glucose and ethanol) and other difficult-to-remove compounds (e.g.surfactants and polyphenols), which may be problematic when directly discharged in treatment system facilities (Oliveira et al., 2009;Mosse et al., 2011;Okada et al., 2013).The organic matter present in WW generally ranges from 1 to 25 g of COD L -1 and depends on the working period (Petruccioli et al., 2002;Lofrano et al., 2009).Higher organic loads are normally originated in Period I, but high levels of COD can also be recorded in Period II (50 g COD L -1 ) probably due to inadequate lees removal (Oliveira et al., 2009).In the same way, the biodegradability ratio (BOD 5 /COD) of WW depends on the working periods, being higher for the WW generated in Period I (Fernández et al., 2007;Oliveira et al., 2009).
Keeping the complete record of WW flow streams helps winery managers to estimate the required storage and treatment infrastructures, because it depends on the size of the winery, the grape variety and/or the harvest period (Oliveira et al., 2009;Mosse et al., 2011;Welz et al., 2016).Other critical issues for efficient WW management in the wineries are the wide range of pH values in WW, as well as the high dissolved solids content and the presence of indigenous microorganisms, e.g.bacteria and yeast (Eusébio et al., 2005;Mosse et al., 2011).

Water use indicators for the vineyard and the winery
Water metrics in wine production must rely on robust indicators and methods that assist growers and wine managers in understanding water cycle streams in their farms and wineries, respectively (Christ and Burrit, 2013;Peth et al., 2017).Medrano et al. (2015) consider different indicators to evaluate water use efficiency, ranging from leaf eco-physiological level (e.g.Net photosynthesis/Transpiration = Instantaneous Water Use Efficiency) to crop level (Yield/Water use = Crop water use efficiency).In a more practical approach, Skewes (1998) proposed several basic indicators to evaluate water productivity in the vineyard.Skewes (1998) includes the ratio between grape yield and the volume of irrigation water used (t/m 3 ), but also considers the ratio between profit and the volume of water used (profit (€)/water volume (m 3 )).In addition, and to incorporate the potential environmental impact of the WW vineyard , we may consider as well the volume of WW derived from PPP application (WW PPP ), which would be translated into another potential indicator of water usage/productivity: volume of WW PPP produced per kg of harvested grapes.However, this indicator can have a large variation, due to the inter-annual variability of weather conditions (see Table I) and related pest pressure, as it occurs under Mediterranean conditions.
In the winery, the volume of WW generated per litter of wine produced (WW winery ) is a widely adopted metric (Table II).However, calculation of a global indicator for small and small to medium farmwineries must take into account the type of grape that is processed (e.g.white vs. red), the labour periods (Periods I and II) and the implementation of best available technics (BAT) (see Table III).This allows stakeholders to evaluate WW winery output along different labour periods, for different grape types, production specificities (e.g.kg grapes production/year), vinification rate and annual water consumption in oenological processes, assuming that all water consumed is discharged as wastewater.This approach would support decision making on winery technologies to reach treated wastewater quality requirements.

Water use indicators: limitations and possible applications to Alentejo wine production
The environmental impact of water use by agriculture is usually amplified in dry climates.Wine production in dry areas (e.g.Alentejo -Portugal) can promote water abstraction and/or pollution of surface and/or groundwater due to WW PPP runoff in the field and/or by WW winery mismanagement.Therefore, a set of indicators is needed to support the assessment of environmental impacts of the wine industry, with focus on water and wastewater management.Indicators focused on water stress/water scarcity can support water use monitoring and management at regional and local levels, contributing to assess the impact of agriculture on water availability, consumption and pollution (OECD, 2011;EU Commission, 2012;Moore et al., 2015;Liu et al., 2017;Xu and Wu, 2017).Among these indicators we can consider the Relative Water Stress Index (RWI) (Vörösmarty et al., 2000), also defined as the Water Availability Index (WAI) (EU Commission, 2012), which provides a measure of water demand pressures from domestic, industrial and agriculture sectors in relation to local and upstream water supplies.In turn, the Water Stress Index (WSI; Pfister et al. 2009), is a variation of the RWI that incorporates climate variability, which can be relevant for regions such as Alentejo, characterized by a wide inter-annual climate variation (see Figure 1 and Table I).

TABLE III
Global and dedicated indicators for wastewater flows in the winery (WW winery ) (Adapted from Duarte et al., 2004;Oliveira e Duarte, 2016 andOliveira et al., 2019) Indicadores globais e dedicados para fluxos de água residual tratada em adega (Adaptado de Duarte et al., 2004;Oliveira e Duarte, 2016e Oliveira et al., 2019).The use of regional and global indicators could be optimized if combined with other methodologies such as Water Footprint (WFP), or Life Cycle Assessment (LCA), for a more robust analysis of the environmental impact of agricultural production (Brown et al., 2011;Liu et al., 2017).Indeed, the WFP accounts for both direct and indirect use of water by a consumer or producer and it works as a spatial-temporal indicator of freshwater use in agrofood products, including wine (Hoekstra et al., 2011;Ene et al., 2013;Lamastra et al., 2014;Bonamente et al., 2015;Saraiva et al., 2019).However, WFP methodology is not consensual (Perry, 2014) and inconsistencies in underlying used water databases pose concern among researchers (Vanham and Bidoglio, 2013) Although the major component is still green water, the distribution of the WFP components was in this case 82% for green water, 15% for grey water and 3% for blue water.Moreover, as WFP values for food products vary with geographic location and climate, these factors must be considered when analysing the wine sector and related wine WFP.Indeed, inter and intra-annual variation of WFP has been observed for several crops including cereals, vegetables and apples (Zhuo et al., 2016).This is an important drawback, as the WFP is only representative for the reported year, while the decision-making process requires long-term and more robust serial historic data, which is not always available (Liu et al., 2017).Under typical Mediterranean climate conditions, WFP values are thus expected to vary with the highly variable rainfall (green water) and irrigation needs (blue water) (e.g.30-100% from year to year).This is in line with Vázquez-Rowe et al. (2012), who emphasized the impact of the harvest year when reporting the environmental impact of wine production and the need for including a timeline analysis in the wine sector.To minimize such a variation, the WFP estimation has been standardized by the ISO norm 14046 (ISO, 2014;Lovarelli et al., 2016).The LCA is another standardized methodology to assess environmental burdens of a product from its production to its disposal or recycling (cradle-tograve approach) according to the ISO standard 14040 series, (ISO, 2014) and it has been used to characterize the wine supply chain (Neto et al., 2013;Quinteiro et al., 2014).Neto et al. (2013) identified four stages of wine production with relevant environmental impact: viticulture, wine production (winemaking to storage), wine distribution and bottle production.Nevertheless, the LCA approach can fail to truly represent environmental impacts related to water and WW management (Comandaru et al., 2012) and an improved temporal resolution (e.g. more frequent assessments in time) is needed to attain correct conclusions based on LCA of crop production (Pfister and Bayer, 2014).This must be tested for grape and wine production.

STRATEGIES TO IMPROVE WATER USE AND WASTEWATER MANAGEMENT IN THE VINEYARD AND WINERY
In order to minimize WW production, several strategies can be adopted to save water in the vineyard and in the winery.We further describe different approaches in the context of Mediterranean wine production.

Strategies for the vineyard
Water saving and water protection strategies in the vineyard may include short and long-term approaches.On the short-term, more efficient irrigation practices (e.g.deficit irrigation) and more precise irrigation scheduling can minimize the negative impacts of climate change on viticulture in South Mediterranean countries and help to save water (Chaves et al., 2010;Iglesias and Garrote, 2015;Medrano et al., 2015; UN-WATER, 2015) (Table I).Literature reports differences in water use efficiency by different grapevine genotypes, which should be considered for irrigation management purposes (Costa et al., 2016).The use of locally adjusted crop coefficients (Kc adj ) in deficit irrigation, derived not only as function of soil and climatic conditions and agronomic practices (Allen et al., 1998;Myburgh, 2016), but also as function of genotype, phenological stage, differences between canopy size, row orientation, training system, vine spacing and different levels of soil water or salinity stress or specific floor management practices (e.g.surface mulching or the use of cover crops), will enable an accurate determination of the adjusted (or actual) crop evapotranspiration (ETc adj ) (see Allen andPereira, 2009 andPereira et al., 2015 for a detailed general review).Also, the use of the "dual" Kc approach, that splits the Kc into the algebraic sum of a basal crop coefficient (Kcb) and a soil evaporation coefficient (Ke), and allows to separately account the contribution of soil evaporation (E) and grapevine (and cover crops, when present) transpiration (T) to the actual ETc (Allen et al., 2005;Farahani et al., 2007;Fandiño et al., 2012;Cancela et al., 2015;Ferreira et al., 2017), should be envisaged in order to better manage irrigation with water savings and decreased leaching risks.In fact, soil management practices (e.g.mulching) can minimize soil evaporation and help to control excessive soil temperatures (Dalmago, 2004) and promote soil fertility and water infiltration (Keller, 2015;Medrano et al., 2015).Moreover, the use of low competition species and adequate management of cover crops, together with installation of water basins to retain winter's rainwater, can promote water infiltration and increase water availability during summer period.Flowmeters in different sectors of the vineyard can help to quantify water use and improve water inputs' inventory.Breeding and selection of varieties, clones and rootstocks with higher resistance to water and heat stress can also contribute to water savings (Gonçalves and Martins, 2012;Keller, 2015;Carvalho et al., 2019), though this is a typical medium to long-term strategy.
Meanwhile, another very important component for sustainable water use in wine production is the precise application of PPP, as well as the management of related WW PPP .Several techniques can reduce spray drift pollution from pesticide spraying in agricultural systems, as explained by Otto et al. (2015).A 38% reduction in drift pollution (e.g. using low-drift nozzles), up to a maximum of 98% reduction could be achieved if hedgerows co-occur alongside fields (EU Commission, 2016).The use of low or anti-drift equipment and techniques are thus recommended to be included in environmental regulatory programmes on a regional scale (OECD, 2014;Otto et al., 2015).In addition, more precise knowledge on vineyard canopy architecture (height, width, volume, density and exposed leaf area) can help to optimize PPP and fertilizer treatments (Rosell and Sanz, 2012;Gil et al., 2014).The wider use of wastewater collection infrastructures in the vineyard or evaporation of water from remnants in dehydration systems (Doruchowski et al., 2014) will also help to minimize WW pollution.

Monitoring plan for winery wastewaters
Water use in the winery represents the smallest part of water inputs for wine production, considering production of grapes under irrigated conditions.However, WW winery production can have a major environmental impact (Christ and Burritt, 2013) (Oliveira et al., 2009;Mosse et al., 2011;Da Ros et al., 2016).Development and use of effective and cheaper alternatives for WW winery treatment is crucial to small/medium wineries to accomplish legal requirements for recycling or disposal (GWRDC, 2011).Cost-effective options for WW winery recycling/disposal must rely on higher energy efficiency and low maintenance costs (Mosse et al., 2011;Oliveira and Duarte, 2015;Kyzas et al., 2016).

Winery wastewater recycling/disposal
In order to reduce costs with WW winery recycling, there are some strategies to be implemented.For example, if the ultimate aim is to reuse treated WW winery for irrigation purposes, domestic WW should not be combined with WW winery .This will prevent contamination by bacteria, viruses and parasites, which require further treatments and a disinfection step.Moreover, the guidelines available for WW winery reuse, in Old World Viticulture countries, include mainly microbiological parameters (Brissaud, 2008;Oliveira and Duarte, 2016).
Regarding WW winery , inorganic parameters are of particular concern; the sodium adsorption rate (SAR) is recommended to be below 6 mmol 1/2 L -1/2 , to prevent adverse soil structural changes (Laurenson et al., 2010).To reduce SAR, some strategies should be applied in the winery (Kumar and Christen, 2009;Mosse et al., 2011), namely the reuse of washing water.The use of alternative cleaning agents based on potassium hydroxide and magnesium hydroxide, although more expensive, may be an option.The ozone treatment as a disinfection procedure will allow the decrease of the conductivity and COD, thus contributing to attain the compliance of the legal limits for uses in crop irrigation (Lucas et al., 2009;Cullen and Norton, 2012).The reduction of COD can also be achieved by screening out solids larger than 0.5-1.0mm with basket screens and by reducing the contact period between solids and WW, minimizing mass transfer.The replacement of citric acid by phosphoric acid is also an advantage (Brissaud, 2008).According to the World Health Organization recommendations for WW reuse (WHO, 2006), the amount of organic matter to be applied via irrigation should not exceed 500 mg L -1 , expressed as BOD5, to avoid changes in soil properties (EPA, 2004).The same source states that application of an urban WW containing a BOD5 between 110-400 mg L -1 may be beneficial for crops (WHO, 2006).However, countries with more restrictive legislation only allow the use of WW with 20-30 mg L -1 of organic matter expressed as BOD5 (Tsagarakis et al., 2004;Brissaud, 2008).Recent Portuguese legislation sets maximum values of 10-40 mg L -1 , expressed as BOD5, for wastewater reuse in crop irrigation, according to quality classes (DL 119/2019).Also, the physical-chemical analysis could be insufficient to evaluate the potential ecological risk, since it does not allow the assessment of possible combined effects of different contaminants mixed together, as well as their bioavailability.Therefore, bioassays are recommended for ecological risk assessment of WW relative to soil or other matrices, whenever WW should be used as an organic amendment (Oliveira et al., 2009;Mosse et al., 2011).

Integrative strategies for a more sustainable wine production chain
The wine industry needs improved water metrics and integrative management strategies related to water fluxes in the vineyard and in the winery.To deal with water scarcity in the Mediterranean region, rain harvesting can be considered, furthering the water supply to the vineyard (Stec and Zelenáková, 2019).
In addition, a conceptual overview of the existing water and WW flows along the wine production chain (vineyard and winery phases) should be envisaged, considering the "Farm-Winery" scenario, which is typical in regions such as Alentejo (Figure 3).Two distinct WW treatment approaches are proposed: the first relates to vineyard's WW (WW PPP ) generated by the activities (e.g.crop protection) taking place in the vineyard, whereas the second is related to WW winery (Figure 3).Consequently, we consider two WW treatment plants in this analytical approach: an agrochemical dedicated WW treatment plant (AWWTP) and an industrial WW treatment plant (IWWTP) (Figure 3).Optimal water management in the winery must identify flows and loads as part of the working Periods I and II, as well as be a function of the type of grape processed (red vs. white grapes).This allows for segregation of flows and selection of the most biodegradable in order to improve WW winery quality (Level 1 quality), which can be recycled within the winery, namely to wash the floors.In parallel, low quality water flows (e.g. with lower biodegradability and high polyphenol content) (Level 2 quality) are suggested to be used after being treated, to irrigate landscape crops surrounding the wineries or the vineyard, or as a small additional supply to vineyard's irrigation.Rain harvesting is another option to consider for reducing water consumption in wineries.It's worth noting that such decisions do require a previous cost-benefit analysis in order to ensure the economic viability and potential new projects.
More integrative approaches to optimize the implementation of climate change adaptation measures in viticulture and to optimize water use in the wine production chain should consider closer interaction and information exchange between all stakeholders (growers, winery managers, technical advisors, local and regional institutions, as well as government entities).A more efficient water management will be achieved when water metrics accommodate data from vineyards, up to the whole region and related governance.

CONCLUSIONS AND FUTURE PROSPECTS
Water must be protected and efficiently used, especially in dry areas.This paper provides a resumed overview and discussion over the water use and WW management in viticulture and oenology in South Portugal.This review helps to clarify and emphasize existing limitations concerning water issues (e.g.efficiency, metrics) in wine production in dry regions and suggests possible approaches to optimize water and WW management in vineyards and wineries.
The typical inter-annual climate variability of Mediterranean climates and more extreme climate events offer increasing challenges when estimating water needs and water use in irrigated vineyards.Indeed, water consumption can be highly variable in the field (due to climate conditions), but also in the winery.In this case, the size of the winery, the type of wine produced and the winemaking periods (I and II) are important variation factors.
Simple water and WW indicators to be used in vineyards and wineries were highlighted and should contribute to: 1) optimize water and WW management in irrigated vineyards in dry areas such as Alentejo (Portugal); 2) optimize water and WW management in the winery; 3) support improved water management in accordance to the specificities of the wine production chain, and ultimately 4) minimize the potential environmental impact of the wine industry.Future studies should test the potential combination of water stress/water scarcity indicators with other analysis tools (e.g.WFP and LCA) to support the wine industry in assessing its environmental impact at farm and regional levels.Indicators such as Carbon Footprint or Ecological Footprint should complement studies focused on water use and WW production in the wine sector.Sustainability and certification programs must consider the water component in both the vineyard and winery, especially in dry areas.These initiatives improve stakeholders' perception of sustainability issues and sustainability certification, which require robust water metrics and data reporting.Still, analysis of medium to long-term impacts of WW on crop and soil needs more detailed studies, particularly for wine production under dry and warm climates.

Figure 1 .
Figure1.Ten-years interval variability of the Dryness Index (DI, mm)(Riou et al., 1994), for (a) Portalegre, (b) Évora and (c) Beja (Alentejo region, South Portugal) using datasets from local weather stations.DI = Wo + P -Tv -Es, where Wo is the soil water reserve at April 1 (assumed as 200 mm); P is the accumulated precipitation (mm); Tv is the vineyard's potential transpiration (mm); Es is the soil water loss by evaporation (mm).The DI estimates soil water availability taking into account vine's transpiration, soil evaporation and precipitation between April 1 and September 30 (in the Northern Hemisphere).The box-plots are bounded on top by the 3 rd quartile (Q3), on the bottom by the 1 st quartile (Q1) and divided by the median (Q2).The average value of the series is represented by (x) and (o) indicates the DI values.The top whisker is defined as Q3+1.5xIQR and the lower whisker as Q1-1.5xIQR,where IQR=Q3-Q1.