Smallholders’ Water Management Decisions in the Face of Water Scarcity from a Socio-Cognitive Perspective, Case Study of Viticulture in Mendoza

Abstract

Grape producers in the Province of Mendoza (Argentina) are extremely vulnerable to the current water crisis, especially smallholders who have very limited resources to adapt. The discourse on adaptation options is mainly technocratic with a focus on modern irrigation systems not accessible to the majority of grape producers. Thus, this research aims at shedding light and providing information for the design of inclusive adaptation strategies by identifying, with a socio-cognitive model, feasible adaptation options according to grape producers’ perceptions and the related implementation barriers. Grape producers’ water scarcity and adaptation appraisal were explored through qualitative interviews in the Northern Oasis (Mendoza) to better understand how producers’ intentions are shaped through perceptual and socio-cognitive processes. To do so, a socio-cognitive model on grape producers’ adaptation to water scarcity (GPAWS) was developed based on two similar models. The analysis reveals that, as overall grape producers share a similar concern with the risk of water scarcity, their different adaptive behaviours tend to be mostly derived from their differences in adaptation appraisal. Moreover, producers’ adaptation intentions are mainly reactive and limited to answer short term, immediate risks. Most of the grape producers perceive feasibility and plan the implementation of reasonable efficiency measures. However, multiple barriers consequently limit the implementation of such adaptation options perceived as feasible by the producers. The results of this research can support government actors, agriculture research institutes, but also the cooperatives of producers seeking to encourage farmers’ adaptation, by identifying which adaptation options could be implemented according to the type of producers and their adaptation appraisal, but also why certain feasible measures are not being implemented.

1. Introduction

“In all, 2.3 billion people already live in water-stressed countries, with over 733 million of them—or about 10 percent of the global population—living in countries with high and critical water stress. An even bigger number of people living in agricultural areas—3.2 billion—are threatened by the risks that water scarcity poses for global food security and sustainable development. These are often the most vulnerable: smallholders and family farmers who see their harvests severely damaged by drought.”

[1] (paragraph 6)

Multiple regions of the world are increasingly facing serious threats of water scarcity and projections indicate that the world will face a water deficit of 40% by 2030 [2,3]. As water scarcity results from the mismatch between water availability and societal demand [4], the combined effects of increasing intensity and frequency of drought due to climate change and increasing societal demand, as well as mismanagement are additionally expected to exacerbate pressure on water resources, thus representing a risk to suffer major impacts especially for smallholders in the global south [2,5,6,7,8]. For these producers, water scarcity represents a challenge to be addressed alongside structurally difficult socio-economic conditions of their production context (e.g., little access to inputs, technical support, etc.) and a fluctuating economy of the countries in which they thrive. Policies and initiatives have been designed and are being implemented to help smallholders improve their water management decisions at the farm level.

Indeed, in an effort to support the design of such initiatives, research on smallholders’ decision-making regarding water management in relation to water scarcity has shed light on the multiple factors that influence their adoption of practices to mitigate or prevent the impacts on their production. For example, Rockstrom and colleagues [9] analysis of Eastern and Southern African smallholders’ water management decisions in a rainfed agricultural water-scarcity and drought-prone production context showed that the experience of dry spells influences their risk-aversion strategies, also leading to counter-beneficial practices (e.g., nutrient mining and low investments). In their study of smallholders’ adaptive behaviours to water scarcity in Kenya, Wens et al. [10] showed that a variety of social, cultural and cognitive conditions influence their decisions to adopt different practices and thus the possible effect of adaptation programmes and policies.

In this paper, we contribute to this strand of literature by focusing on social and cognitive dimensions of smallholders’ decision-making to address water scarcity in the context of grape smallholder producers in a Latin American country such as Argentina. This represents a clear example in which smallholders face structural agricultural production challenges due to limited access to resources, water scarcity and a recurrent national economic crisis which, while pushing much of its population below the poverty line and the inflation rate to above 40% values [11], affects their ability to ensure capacity to cope with more frequent and intense droughts.

More specifically, we focus on the province of Mendoza, given its importance for world-renowned wine production, making this sector the first source of economic income for the province, but also the first consumer of water for irrigation [12,13]. This region is suffering recurrent water scarcity, mainly affecting smallholders’ livelihoods with worrying consequences in increasing the share of the population living below the poverty line in the country [12,13]. This water scarcity is being triggered by the recurrent El Niño Southern Oscillation (ENSO) event [13,14], and as shrinking water flows from the Andes’ glaciers from which they largely depend [13,15]. As a result of this challenging context and the ongoing economic crisis, many of the smallholders in the area have started to abandon part of their vineyards due to their limited capacity to adapt [16].

A large proportion of these smallholders depend on traditional surface irrigation systems that are resource-intensive and are characterized by significant water loss. Initiatives designed to improve water use efficiency in these farms have differentiated producers as viable or non-viable, depending on their technical capacity to implement what was framed as modern technical irrigation solutions [16]. This dominant institutional response has overlooked the social, economic and related decision-making context of producers with a resulting low level of adoption of these irrigation techniques and continuous water mismanagement and loss. As suggested by the above-mentioned research focusing on social and cognitive perspectives, this technical-based approach might have overlooked opportunities to design better water management strategies and related incentives based on a better understanding of the possible role of people’s beliefs about risks and mitigation solutions, which can be important drivers of private adaptation decisions [17,18]. With the aim to contribute to fill this knowledge gap, this research aims at shedding light and providing information for the design of inclusive adaptation strategies based on the insights provided by a socio-cognitive understanding of barriers and opportunities perceived by producers that might influence their water scarcity risk and adaptation appraisal and finally their water management decisions.

In the following sections, we first provide a conceptual approach to consider the socio-cognitive factors that influence smallholders’ decision-making processes to address and cope with water scarcity risks. In the methodological section, we introduce the study area background information regarding climatic risks and socio-institutional context of smallholders’ vineyard production and adaptation interventions. Then, we present the results and summarize the key findings. At last, we discuss and interpret the major findings and briefly conclude.

2. Socio-Cognitive Framework to Analyse Farmers’ Adaptation Decisions

In order to promote smallholders’ adaptation to adverse climate risks, authors have increasingly called for a better understanding of the factors that influence farmers decision-making, with special attention to both their perception of the climate risks as well as their appraisal of what are effective adaptation measures [19]. For example, previous studies found that smallholders with limited resources tend to have short planning horizons and struggle in adapting a long-term vision [20,21]. Even when perceiving climate risks, some smallholders might still decide to not proactively adapt but only react to shocks through short-term coping measures [22,23]. Longer-term adaptation measures might force smallholders (who typically have a small margin of risk-taking capacity) to make a trade-off between the potential (and uncertain) long-term benefits of adaptation measures and the required investments which might include intensive labour and/or financial resources needed for their establishment (in the short-term) [21,24].

Most of the empirical research has focused on resource constraints (financial, technical and institutional) as the major determinant of adaptation [17]. However, an increasing part of the scientific community started to consider the importance of psychological factors in understanding adaptation processes [17,19,25,26]. People’s beliefs about risks, chances and adaptation options have often been neglected, even though they were found to drive much of the process of private adaptation decision-making [17,18]. In response to this evidence, Grothmann and Patt [17] developed the Model of Private Proactive Adaptation to Climate Change (MPPACC) based on the Protection Motivation Theory (PMT) [27]. This model analyses both the socio-economic variables (external social context) and the cognitive factors (internal psychological factors), which together help explain private adaptive behaviours [17,18,19].

This socio-cognitive approach differentiates between two key perceptual processes [17], namely: risk appraisal and adaptation appraisal. The former has been more largely considered in research on risk perception and refers to the individual assessment of the probability of threats and the related damage potential to things he/she values, under the condition of no change in his/her behaviour. The second refers to an individual perceptual assessment of his/her ability to avert being harmed by the threat, along with the costs of taking such action. Adaptation appraisal has been increasingly acknowledged by the research community focusing on smallholders’ adaptation, as the perceived adaptive capacity is crucial (i.e., alongside the objective resource endowment needed to adapt) to transform the individual’s motivation into adaptation action. In their socio-cognitive model, Grothmann and Patt [17] suggest that adaptation appraisal can only come after risk appraisal, as a specific threshold of threat need to be reached and perceived by the adapter and that avoidant responses (e.g., negating risks, wishful thinking, or fatalism) might not prevent the economic or physical damage, but only the negative emotional consequences of the perceived risks of these impacts. In this perspective, a person would do nothing to adapt (even if he/she has a high-risk perception) if his/her perceived capacity to adapt is low. In this line, these authors suggest the need to not only communicate risks but also potential effective and accessible adaptation options.

Building on MPPACC [17] and Mitter et al. [18], we propose a Grape Producers’ Adaptation to Water Scarcity (GPAWS) socio-cognitive model (see Figure 1) allowing us to take into account specific contextual factors and the risks and production challenges faced by grapes smallholders. This model acknowledges that alongside producers’ characteristics and contextual factors, their “risk experiences” (inspired by the MPPACC) with water scarcity influence their risk appraisal. Moreover, the GPAWS includes the influence of the social discourse regarding water scarcity as it is expected to shape grape producers’ perceptions of water scarcity risk. The component “other risks” is included in the model, as grape producers in the regions are facing numerous risks from all kind of natures, such as grape losses due to hail. Thus, there is a need to understand how important the risk of water scarcity is for producers in comparison with the other risks and difficulties they might be facing, as it can influence their water scarcity risk appraisal. The “technocratic discourse” component refers to the dominant institutional response previously mentioned and is expected to negatively influence smallholders’ perceived adaptive capacity when they cannot access modern irrigation systems.

3. Methods

3.1. Study Area

The province of Mendoza is located in the Andean central-west region of Argentina (see Figure 2) and is characterised by a semi-arid climate. This territory can be described as a desert oasis in which grape production happens in intensive artificial irrigation systems mainly within the Mendoza river catchment (Northern Oasis) [13,28,29]. The water flows of this catchment are dominated by snow-glacier dynamics [8] and are the most vulnerable, with a water scarcity index of around 98.17% [30]. Generally, a total of 1500 km² of the agricultural production area is under irrigation from both surface water and groundwater in the Northern Oasis [29]. Viticulture represents 46% of the agricultural land area (43.992 hectares) in this oasis being largely (i.e., 73%) composed of vineyards that are less than 10 hectares and that mainly (i.e., for 73% of plots) use surface irrigation (e.g., flood irrigation, furrow irrigation) and, in a smaller proportion (i.e., 23%), drip irrigation [31].

As the region’s water flows depend on Andes’ glaciers melting dynamics, water availability is being affected by the observed substantial loss of glaciers and snow drought conditions due to climate change [5,13,15,33,34]. These trends are expected to persist in the next decades with projected decreases of 3% to 30% in mean annual precipitations by the end of this century [35] and of a mean annual river discharge of approximatively −16% (Mendoza River) by the end of the century [33].

Around 70% of the hectares cultivated in the northern part of the Mendoza river basin rely on groundwater when surface water is insufficient, leading to aquifer overexploitation [13,33]. These supply-side problems, together with the low efficiency (30%) of the irrigation water distribution systems [33] and farm-level water mismanagement add to the pressures on Argentinian viticulture water resources posed by climate events as heatwaves, hail and frost impacting both yields and the wine grape quality [36]. The Northern Oasis is divided in six irrigation zones (see Figure 3), which can be differentiated by their location along the Mendoza River catchment. Upstream irrigation zones (1 to 3) tend to have higher water flow level compared to downstream irrigation zones (4 to 6). Therefore, farmers located in downstream irrigations zones tend to be more impacted by the water crisis than farmers located upstream.

3.2. Water Scarcity Management Context

Based on the 1884 Water Act, the province’s governance of water management adopts a supply-based distribution approach in which water volumes are allocated based on the irrigator’s farm size. Under this approach, the General Department of Irrigation (DGI) responds to water scarcity crisis emergencies by reducing the volume of water for each type of licence in equal proportions, without taking into account farming contexts such as crop type and growth cycle or soil types. This water allocation system has proven ineffective in supporting smallholders’ ability to address water scarcity [32]. Capacity to respond to water scarcity is also challenged by the lack of regulation and legislation around water use and distribution as well as the lack of data and transparency regarding annual water use by farming premises [13,16,37]. Institutional responses within this context have adopted a technocratic approach to water crises focusing on improving water-use efficiency through modern irrigation systems that tend to overestimate smallholders’ capacity to access resources needed to implement these measures [16].

More concretely, past institutional efforts devised two types of measures to support adaptation options for grape producers’ water management in Mendoza, namely: artisanal practices, and infrastructural/technological options [28,32,36]. Since the majority (70%) of grape producers in the area have limited access to the resources needed to implement the latter (i.e., high costs and knowledge requirements, DGI), institutional efforts led by the National Institute for Agricultural Technology (INTA) focused on accessible and feasible adaptation options, proposing the concept of Measures of Reasonable Efficiency (MRE) [38]. This approach suggests to minimize requirements of structural investments and rather emphasize the optimization of existing irrigation systems (see example in Figure 4).

3.3. Data Recompilation and Analysis

Building on the approach of Mitter et al. [18] and Grothmann and Patt [17], this research adopts a qualitative approach combining interviews and experiential case encounter [39] to capture smallholders’ perceptions and views regarding water management decisions in response to water scarcity. Through interviews, we captured smallholders’ subjective views on their farm water management and social context to deal with water scarcity events and risks [18]. Experiential case encounter was a complementary method to enhance our practical understanding of the case study as we accompanied producers in their irrigation activities (e.g., opening sluice gates) and agronomists in their field visits, thus getting a personal experience with the case study while observing and taking notes of their comments on water management. The following Figure 5 provides an overview of the research approach, data collection and analysis steps.

Structured and semi-structured interviews and experiential case encounters were all conducted in Spanish and data were registered in a logbook. We conducted 28 structured on-farm interviews with producers and ten semi-structured interviews with agronomists and relevant institutional actors. We identified people to interview either through direct contacting them to ensure sample variability or indirectly based on suggestions by key informants or via snowball sampling (i.e., interviewees’ suggestion of a new contact) [18,40]. The majority of the respondents were contacted through the contacts provided by two cooperatives of producers (indirect, gatekeeper): Asociación de Viñateros de Mendoza (AVM); Asociación de Cooperativas Vitivinícolas Argentinas (ACOVI).

Producers interviewed were located mainly in the downstream areas of the Northern Oasis, where water scarcity has hit the most (5th irrigation zone). Although we focused more on smallholders (i.e., less than 20 hectares) largely depending on surface irrigation, we also interviewed a few medium and large producers using drip irrigation to understand how the production’s size and irrigation system influence cognitive factors and adaptation. Respondents were only producing grapes (i.e., not involved in wine production) and were the ones responsible for water management decision-making. Agronomists and relevant institutional actors were selected based on their expertise and knowledge of the smallholders’ production context. The majority of the agronomists interviewed work at the Instituto Nacional de Tecnología Agropecuaria (INTA), while a minority work for the Programa de Apoyo a Pequeños Productores Vitivinícolas en Argentina (PROVIAR).

The interview guide for grape producers included open-ended and closed questions focusing on the socio-cognitive model (GPAWS). A pilot was conducted with extensionists to adjust the interview questions and structure. This proved to be a critical step and helped us focus the interview guide on the components of the GPAWS framework rather than on the interactions between the components to avoid cognitive burden on producers. For the semi-structured interviews with agronomists and key institutional actors, the interview guide was adapted to each respondent’s expertise and helped us to gain a deeper understanding on the external socio-institutional context of the GPAWS (adaptation barriers, objective adaptive capacity, institutional support, etc.).

Following the Grounded Theory inductive methodological approach [41], we transcribed interviews and proceeded to coding using ATLAS.ti 9 [42], ensuring codes were based on the GPAWS components perceived self-efficacy, perceived adaptation costs, perceived risk impacts, and adaptation intentions. We established interpretative rules to identify data regarding key GPAWS concepts as revealed by responses to multiple questions. Our interpretative rules helped us to inductively determine the extent of producers’ Perceived Adaptive Capacity (PAC) and Risk Appraisal (RA). For instance, the PAC level for each producer was assessed based on producers’ answers to five questions; questions 1 to 3 lead to yes or no answers, while questions 4 and 5 lead to answers defined as high/medium/low efficacy and self-efficacy. According to our interpretative rule, in order to have a high PAC, producers’ responses had to consist of a positive response to questions 1 and 3, and of a response at least “high” and “medium” to questions 4 and 5. The data were put into Excel to facilitate the comparison and identification of trends.

In order to comply with ethical standards in interviewing, we asked for informed consent, and communicated and ensured anonymity to the respondents, using data synthetically coded to identify respondents. In the following results section, small producers are referred to as SP, medium producers as MP, large producers as LP, agronomists as AG, and institutional actors as IN.

4. Results

In

Table 1. Location, irrigation system and socio-economic characteristics of grape producers interviewed and their farms (n = 28).

Sample Description		Location		Irrigation System
		Downstream	Upstream	Drip	Furrow	Flood
Producers type	Small	18	5	3	18	2
	Medium	3	0	0	3	0
	Large	2	0	1	1	0
Age groups	Under 40 y/o	11	1	0	11	1
	Over 40 y/o	12	4	4	11	1
Education levels	Primary	6	1	0	7	0
	Secondary	6	0	0	5	1
	Tertiary	11	4	4	10	1
Relation with the vineyards	Owner	20	4	4	19	1
	Employee	3	1	0	3	1
Member of cooperative	Yes	15	1	1	14	1
	No	8	4	3	8	1
Irrigation water rights by source	Superficial	15	2	1	15	1
	Ground and superficial	8	3	3	7	1

we summarize the main characteristics of the 28 producers interviewed. The average farm size of the sample is 21 hectares, but that of the majority (i.e., smallholders) is 6 hectares. All respondents sell low-quality grapes to wine producing cooperatives mostly targeting domestic markets. Producers were interviewed in the six different irrigation zones of the North Oasis with the majority located in the 5th irrigation zone of Lavalle (downstream). The respondents’ average age was 48 years old, which is relatively young compared to the average age of the producers in that region (68 years old). Approximately half of the respondents are members of a cooperative of producers, an important factor to consider, as these producers tend to receive advice from the cooperative’s agronomist regarding irrigation practices. Around two thirds of the respondents mentioned receiving assistance and/or services from institutions such as: FECOVITA (Federation of Argentinian Vitivinicultural Cooperatives); COVIAR (“Corporación Vitivinícola Argentina”); INTA (Instituto Nacional de Tecnología Agropecuaria). The majority of respondents with groundwater rights could not make use of it due to their non-functioning wells.

In terms of the future water scarcity risk appraisal, a large proportion of small- and medium-size producers tend to perceive medium to high probability of impacts from current and future events. Interestingly, the only two large producers interviewed appear to minimize those risks. Producers’ location in the watershed does not appear to be related to the extent of risks, differently from the age characteristics of the respondents. Indeed, producers over 40 years appear to have a lower risk appraisal than producers under 40 years old (

Table 2. Smallholders’ water scarcity risk appraisal in relation to their characteristics. Results represent the proportion (%) of the respondents.

			Main Characteristics of the Respondents
			Production Size			Age		Location
GPAWS Components	Type of Answer	All Producers (n = 28)	Small Producers (n = 23)	Medium Producers (n = 3)	Large Producers (n = 2)	Producers under 40 Years Old (n = 12)	Producers over 40 Years Old (n = 16)	Downstream (n = 23)	Upstream (n = 5)
Water scarcity risk appraisal	HIGH	47	48	67	0	75	25	48	40
	MEDIUM	21	22	33	0	17	25	26	0
	LOW	32	30	0	100	8	50	26	60
Perceived WS risks for 2022	HIGH	54	52	67	50	58	50	52	60
	MEDIUM	18	17	33	0	17	19	17	20
	LOW	28	31	0	50	25	31	31	20
Perceived WS risks for 2032	HIGH	64	65	100	0	75	56	74	20
	MEDIUM	22	26	0	0	17	25	9	80
	LOW	14	9	0	100	8	19	17	0
Perceived WS risks impact	HIGH	46	52	33	0	67	31	48	40
	LOW	54	48	67	100	33	69	52	60
WS perceived as the major risk	YES	50	48	67	50	75	31	52	40
	NO	50	52	33	50	25	69	48	60
WS risk experience	HIGH	93	91	100	100	100	87	91	100
	LOW	7	9	0	0	0	13	9	0

), while the older generation of producers considers water scarcity risk in the long term highly uncertain, but also part of a natural cycles as suggested by one respondent in this group stating “I do think we cannot predict nature, I believe it’s a cycle, and water level will come back as before” (SP10).

However, in spite of this awareness of the natural water scarcity cycles, most of the respondents have expressed major concerns regarding the current and future water scarcity crisis in the region. This is reflected in actual limitations being faced, which forced some to reduce their irrigated area stating that “the waterflow has decreased a lot in these 10 years, I used to irrigate an area of ten hectares, but nowadays I can only irrigate four hectares, as water is insufficient” (SP6). Almost all producers (excluding SP9, SP10) experienced episodes of water stress in their vineyards, and the majority described suffering each year from water scarcity. It should be noted that none of the respondents saw opportunities in water scarcity, while most actually indicated that alongside water scarcity they face also other climate risks, as suggested by the following quotes:

“I lost my harvest due to hail in 2018 and 2019, this is why my priority is to buy protection against hail; irrigation will come after.”

(SP18)

“Yesterday, I lost 80% of my production in two minutes, due to hail. I went outside and I could not do anything besides watching a year of work going away with hailstones as large as a tennis ball.”

(SP10)

In addition to these climate risks, producers are facing non-climatic risks which can affect their ability to cope with climate risks, for example affecting their production rentability and economic stability, as this quote suggests in relation to the economic risks:

“The grape prices are in pesos and it does not follow the current inflation. For the last three years, the price of the wine has not increased, but we have an inflation of 50%, meaning that everything costs more. For example, in one year, the price of agricultural inputs increased by 100%, but we received the same income as before. This is the main problem.”

(SP13)

This economic situation in which their production has to thrive is also made more challenging by the difficulty in finding labour forces in numbers and quality, which has been an issue for multiple years, not only during the harvest, but nowadays, during the whole year.

It appears that production size (possibly as a proxy of broader access to assets) plays a role in producers’ appraisal of their adaptative capacity, as all medium and large holder producers perceive high PAC compared to a more distributed perception among smallholders. Producers’ age appears to also influence perceived adaptive capacity, as producers under 40 years old tend to have a higher PAC than producers over 40 years old. The age seems to especially influence producers’ perception of the adaptation cost (“worth it”), but also their perceived ability to implement certain adaptation options (“self-efficacy”) (

Table 3. Perceived Adaptive Capacity in relation to smallholders’ characteristics. Results represent the proportion (%) of the respondents.

			Main Characteristics of the Respondents
			Production Size			Age		Location
GPAWS Components	Type of Answer	All Producers (n = 28)	Small Producers (n = 23)	Medium Producers (n = 3)	Large Producers (n = 2)	Producers under 40 Years Old (n = 12)	Producers over 40 Years Old (n = 16)	Downstream (n = 23)	Upstream (n = 5)
Perceived Adaptive Capacity	HIGH	33	20	100	100	36	31	26	57
	LIMITED	37	44	0	0	50	25	39	29
	LOW	30	36	0	0	14	44	34	14
Perceived efficacy	HIGH	60	52	100	100	71	63	57	71
	MEDIUM	30	36	0	0	29	25	30	29
	LOW	20	12	0	0	0	12	13	0
Perceived self-efficacy	HIGH	60	60	66	50	50	81	57	71
	LOW	33	36	34	0	43	19	35	29
Perceived adaptation cost	YES	60	52	100	100	71	50	65	43
	NO	30	36	0	0	7	50	34	14

). Moreover, the production size tends to influence producers’ PAC, as small producers have a lower PAC than large producers. The type of irrigation system also matters, as producers with drip irrigation tend to have a lower risk appraisal than producers with surface irrigation, as they usually experience no water stress thanks to the efficiency of their irrigation system. There appears to be no specific pattern between the PAC and the producers’ location.

In general, producers perceive their adaptive capacity as insufficient to face the water crisis, which in some cases (SP1, SP9, SP19, SP27) is associated with a passive attitude and conviction of an inability to change and to be able to achieve a different outcome. Indeed, almost all producers mentioned their inability to face the current water crisis on their own and needing more involvement of government agencies (i.e., reliance on external support). The extent of adaptation appraisal and, inversely, reliance on external support had a direct relation with the intentions to adapt. Limited PAC appears also to be associated to producers’ willingness to take credit and is rather related to their willingness to implement MRE given their low investment requirement and dependence on external inputs, with respect to more technically-advanced irrigation alternatives (e.g., drip irrigation mainly implemented by respondents with high PAC). As observed, all the producers with non-protective reactions had a low adaptation appraisal, which tends to confirm the assumption that adaptation appraisal is negatively correlated with avoidance. The following quote from a producer with a low PAC illustrates how his/her fatalism but also his/her perception of the adaptation benefits can lead to avoidance:

“I was thinking of waterproofing the main irrigation canal, but I’m tired, I will not do anything. If it were 20 years ago, I would have done it. Now I am too old and it will not bring anything to me. It is already too late. If you look at how much I get paid for one kilo of grapes, you would understand.”

(SP9)

This example also demonstrates that the lack of rentability does not only impact the producers’ objective capacity, but also their PAC and willingness to adapt. Even though the quoted producer knew that waterproofing the canals was feasible (high self-efficacy) and will be efficient to save water and beneficial for the production (high efficacy), he/she did not implement it (SP9). This observation shows that a producer perceiving a certain adaptation option as very efficient and accessible, the adaptation cost might still be perceived as too high (cost in term of physical effort, and time) thus hampering its adoption.

In terms of measures applicability, most of the respondents are already using the following MRE: furrow irrigation, land levelling, amass water flow, and weed control. Moreover, the majority of the respondents have the intentions to implement more MRE such as sluice-gate. Other essential measures of reasonable efficiency, such as waterproofing internal irrigation canals, are less considered, even though this measure can considerably reduce producers’ vulnerability to the current water crisis. This is illustrated by an agronomist in the following quotation:

“My colleagues and I found that waterproofing the internal irrigation canals can improve by 50% the producers’ irrigation efficiency, which means half of the water entering the vineyards could be saved. This is the kind of affordable and efficient adaptation option that is under-implemented.”

(AG4)

Therefore, multiple techniques of reasonable efficiency are available and accessible, but still many producers among the respondents do not use them. Agronomists recognized that making general conclusions on which irrigation techniques should be implemented or not, but also their corresponding efficiency, is difficult, as each vineyard has very different conditions (e.g., soil type) (AG1, AG4). This could mean that even though producers might perceive they have the capacity to implement MRE (high self-efficacy), they might not do so as the effectiveness (low perceived efficacy) of certain MRE (e.g., weed control) is uncertain and need to be assessed depending on the vineyards (e.g., soil management). In addition to the MRE, very few producers plan to implement other options such as constructing a private water reservoir. The producers’ low self-efficacy combined with the high perceived adaptation cost (money, space) could explain why only 11% of the respondents have the intention to build an individual water reservoir. It should also be noted that a minority of producers, who mostly had low PAC, plan to sell or abandon their vineyard.

At last, producers’ adaptation appraisal appears in this study to be more relevant than producers’ water scarcity appraisal, to understand their different adaptative behaviours, but also the producers’ appraisal regarding the applicability of the different solutions. Indeed, the fact that almost all the respondents experienced water scarcity and mentioned overall being preoccupied by this risk makes the identification of trends between these variables difficult, due to the low variability in water scarcity risk perception among respondents. However, it is not because the risk appraisal was less relevant in this study that it is a less important component of the GPAWS. Risk appraisal is essential, as producers will usually not adapt to a risk they do not perceive. Thus, as overall the respondents were concerned by the risk of water scarcity, the answer to analyse their different adaptive behaviours tended to be mostly derived from their differences in adaptation appraisal.

In order to summarize the results presented in the above sections narrative, Figure 6 highlights the major factors that, based on our interpretative rules applied within the GPAWS framework, influence smallholders’ adaptation intentions to address water scarcity in their farms. To start with, risk appraisal appears to be influenced not only by recent experiences and the characteristics of their production context, but also by the complex interactions between water scarcity risk and the need to also address other type of climatic (e.g., hail, heat waves, etc.) and, especially non-climatic factors (e.g., economic crises and labour scarcity) which can indirectly (e.g., increasing their fatalist avoidance attitude) or directly, together with adaptation appraisal, influence their adaptation intentions.

Overall, we found at least six major dimensions that emerge from our GPAWS-based interpretation of respondents’ perspectives to address irrigation water scarcity and improve adoption of farm water management adaptation measures. First, personal characteristics related to experience and knowledge regarding farm water management might influence their ability to identify water losses in their vineyards, irrigation practices and thus their perceived need to actually look for ways to reduce water losses (IN2). For example, an agronomist noticed that, even though producers might know how to irrigate their vineyards, they often do not question the way they irrigate:

“During an assembly, a few months ago, producers participating were asked to say if they thought they irrigated well. Almost all of them answered yes. If a producer cannot identify where he has water losses that could be avoided, he will not change anything.”

(IN2)

Interviews with agronomists suggest that this might be related to the producers’ experience as, according to them, older generations of producers might have better practical knowledge in irrigation practices compared to the younger generation of producers (AG2; AG4; SP21; MP26). Even for the cases in which younger producers have learned from their older peers, interviews revealed that irrigation experiences and techniques learned from parents and grandparents (e.g., furrow irrigation inherited by the family tradition; SP21) might not be fit to address the currently changed causes and dynamics of water scarcity occurrence (IN2; AG1; SP21; SP15). This suggests that these irrigation techniques inherited by young producers might be outdated and unfit, and actually be causing major water losses.

Second, social-cultural barriers emerging from the social discourse on water scarcity (i.e., in the contextual social dimension of the GPAWS) might also influence producers’ perceived causes of water scarcity risk and relatedly associated responsibility to address it. An insightful example of this is some producers’ discourse on the responsibilities associated to the water scarcity crisis which reveals a defensive attitude. In this respect, an agronomist stated:

“I think that one of the important challenges to producers’ adaptation is their defensive attitude in the face of the water crisis. Many producers put the guilt on the water authorities (DGI), without realizing that they can also make a change in multiple ways. Many solutions are accessible to these producers.”

(AG4)

Third, the economic situation in Argentina combined with the lack of rentability in viticulture can objectively limit the adaptive capacity and represent a major barrier to producers’ adaptation decisions influencing their perceived financial capacity to adopt alternative options (e.g., drip irrigation, repairing wells or constructing a new reservoir). So, even MRE options requiring low investments might be perceived as a barrier by some producers, as they might need to take a credit to access the required materials (e.g., PVC for waterproofing the irrigation canal). Even when multiple lines of credits are officially offered at acceptable interest rates, some smallholders interviewees indicated major bureaucratic barriers and documentations they are not able to address (SP12; SP21; SP18; SP28; PP18).

Fourth, some structural institutional barriers related to the water supply system constrain producers’ capacity to adapt (IN1; IN2; IN3; AG3; AG4). Indeed, producers are informed only a few days in advance about the exact date and time they will receive their irrigation turns, leaving little time to implement the vineyards preparation measures needed to improve water use efficiency. For example, interviewees indicated that some non-material MRE require time to be implemented before actual irrigation water is distributed in the field. Additional structural barriers are associated to the limited resources (e.g., time, staff, transport, etc.) that the water authority (DGI) and producers cooperative have to identify and support the implementation of vineyards-tailored MRE options for all producers and thus improve the irrigation methods at scale (IN2; IN3; AG4).

Fifth, objective limitations in accessing resources and their limited margins for experimentation also influence the producers’ adaptation appraisal possibly determining their more reactive than preventive attitude towards water scarcity risks. Indeed, almost all the respondents mentioned having reduced the area cultivated in response to their experiencing of irrigation water scarcity, as suggested by the exemplary quote of one producer:

“We decided to implement drip irrigation this month as we lost too much production last year with the lack of water. We know, doing such transition in the summer is dangerous, as it causes stress to the plants. But we have no choice, we had to do the transition, if we want a decent harvest.”

(SP15)

This same producer added that they waited until they experienced major impacts before adopting response measures (SP15). In face of high uncertainty regarding water scarcity occurrences and their limited margin for experimentation, some producers adopt a short-term perspective, as suggested by one producer’s point of view regarding his/her adaptation plan for the future:

“I live day by day. I wake up and see what can be done, and tomorrow is another day. Decisions are taken day by day, as everything is so uncertain, you do not know what can happen the day after in Argentina, it is unpredictable, we have no visibility for the future.”

(SP5)

Sixth, even if adaptation measures are adopted at the farm scale, the effectiveness in addressing irrigation water scarcity might be hampered by water mismanagement at the landscape scale. One interviewee revealed, for example, that the presence of urban waste in the canals represents a common physical barrier affecting irrigation zones especially downstream, as canals pass through urban areas. This appears to lead to major water losses which, according to the respondent, amount up to 20% of the water in the distribution system (IN3).

5. Discussion

Through the use of GPAWS, we could analyse which socio-economic and environmental contextual factors are influencing producers perceptions regarding risks and adaptation alternatives and their decisions to adopt possible solutions [17,18]. This study contributes to analyse grape producers’ adaptation decision-making and can be useful to innovate the way viticulture’s adaptation to water scarcity has been traditionally addressed (i.e., with limited embeddedness in producers’ heuristics and production context) [28,32,36].

Our results confirm the findings of other studies [17,18,19,25,43], as risk appraisal appears to be positively related to adaptation intentions, but also that alongside climate risks, other non-climatic risks influence producers’ water scarcity risk and adaptation options’ appraisal [32]. In addition, perceived adaptation costs appear to have a major influence on producers’ adaptation intentions alongside perception of the extent to which adaptation options are effective and affordable [19,25].

In line with findings from other studies, only a minority of small grape producers intend to implement investment-intensive long-term strategies (e.g., drip irrigation) (e.g., in Tanzania [22]). Two socio-cognitive barriers seem to be related to this apparently short-sighted perspective on water management investments on the side of the producers. On one side, these smallholders face a cognitive barrier while facing multiple risks that might even concretize their impacts at the same time (as de facto during the interviews), such as those related to climate extreme events and economic crisis. This implies that, in making decisions on whether to invest in water management improvements, they face deep uncertainties regarding not only the intensity and recurrence of extremes, but also about whether they are able to recover their investment costs as, for example, the national currency might significantly devaluate so that their access to inputs markets becomes very limited. On the other side, interviews to these producers also revealed that there are social barriers related to scarce institutional support.

Our findings confirm the importance and similarity of the institutional-support gaps also found in many other smallholders’ production contexts, such as the lack of access to credits (e.g., in Ghana [44]) and support for inputs needed to invest in water use efficiency measures [45], combined with the multiple risks and uncertainty context [32] and limited own resources to buffer investment risks (e.g., in Ethiopia [46]). These contextual limitations, similar to what others found in smallholders’ production situations (e.g., in Ethiopia [47]; in Ghana [44]; [21]), hamper the adoption of improvements, even when technically-feasible and relatively low-cost improvements in water use efficiency are available. As noted by other studies, there might still be multiple barriers as, for example, showed by a systematic assessment of five case studies on farmers’ adaptation behaviours [48] illustrating that, in spite of the technical feasibility of alternatives, farmers perceive and have to weigh in their decision multiple risks (e.g., price volatility) in their responses to short-term immediate risks.

In addition, addressing water scarcity risks requires interventions to not only focus on the farm scale alternatives to improve water use efficiency at a reasonable cost. Indeed, our interviews reveal that it is important to also take into account current activities of and water mismanagement by other water demanding sectors (e.g., urban waste polluting irrigation canals) that can affect agricultural production and scarcity of water resources at the quality standard required. Then, the design of alternatives to address water scarcity might need to consider and be informed by a more integrated water management (IWM) approach that takes into account water dynamics within the boarder landscape scale (e.g., solutions that acknowledge the urban-agricultural-water nexus). The need to promote this landscape-informed IWM approach to water scarcity is also in line with the findings of a large literature review by Van den Brandeler et al. [49], who called attention to the fact that urban and agricultural uses often compete for water quantity and quality, while the design and implementation of urban integrated water management pollution is rarely informed by water management solutions in rural (agricultural production) contexts.

6. Conclusions

In the face of climatic and non-climatic risks, small grape producers’ survival is under threat and many have started to abandon part of their vines. Analysing grape producers’ perceptions is essential to develop inclusive and feasible adaptation strategies to the water crisis in Mendoza. The results show that grape producers are very concerned by the risk of water scarcity, as almost all of them already experienced water stress in their vineyards for numerous years. As grape producers shared a similar concern with the risk of water scarcity, their different adaptive behaviours tended to be mostly derived from their differences in adaptation appraisal, but also contextual factors. With limited resources, highly diverse and multiple risks, most of the producers are restrained to implement coping measures (short-term) that do not prevent future monetary or physical damage, while only a limited share of farmers chose resource-intensive and long-term strategies. The measures of reasonable efficiency are under-implemented even though, they appear to be the most accessible and adapted options to reduce smallholders’ vulnerability to the water crisis. The findings of this study can support government actors, agriculture research institutes, but also the cooperatives of producers and their agronomists whose objective is to encourage grape producers’ adaptation. Indeed, the results help to identify which adaptation options could be implemented according to the type of producers and their adaptation appraisal, but also why certain feasible measures are not being implemented, and contribute to adjust the adaptation strategies while considering the adaptation barriers.