A Systematic Review on Drivers of Water-Use Behaviour among Agricultural Water Users

Abstract

Sustainable Development Goal 6 (SDG 6) is closely linked to the sustainable management of water resources and sanitation worldwide. SDG 6.4, in particular, aims to significantly improve water-use efficiency across all sectors by 2030. It is important to acknowledge the significant role that behavioural aspects of water users in an agricultural context play in contributing to water-use efficiency. This systematic review aims to provide an up-to-date synthesis of the current knowledge of water-use behaviours in agriculture to stay on track in achieving SDG 6. This systematic literature review investigates the factors influencing water-use behaviour among agricultural water users globally. Using the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) method, we retrieved a total of 867 records, of which 47 matched the eligibility criteria. The 47 relevant studies were primarily conducted in the United States and China with key themes including sustainable agricultural practices, technology adoption for productivity, climate change adaptation, and modelling and uncertainty in water conservation. Additionally, the review identified six distinct water-use behaviours investigated from 2020 to 2024, which were each driven by its unique set of factors. Overall, the findings from the systematic review indicate that there has been a geographical gap in research efforts over the past five years, and it is important for water-use behaviour-related research to be extended to other countries that are water-stressed. Furthermore, the researchers recommend that future studies should develop comprehensive behavioural models and adopt holistic approaches to better understand and promote sustainable water-use behaviours in agriculture. These efforts are vital for achieving sustainable water management and broader SDGs.

1. Introduction

On average, agriculture accounts for 70% of global freshwater withdrawals [1]. The per capita distribution of freshwater resources has changed consistently with population growth. Over the 30 years leading up to 2017, food production increased by 100%, and it is estimated that by 2050, a further 60% increase in food production will be necessary to meet the growing global population’s food requirements [1,2]. This anticipated higher demand for food in the future is expected to directly affect agricultural water usage, with agricultural production already under threat due to pressures on land and water systems [1].

Against the backdrop of rising food demand, the FAO [2] emphasises the critical role of increasing land and water productivity in achieving food security, sustainable agriculture, and the objectives outlined in the 2030 Agenda for Sustainable Development Goals (SDGs). Among the SDGs, SDG 6 is closely linked to the sustainable management of water resources and sanitation worldwide. SDG 6.4, in particular, aims to significantly improve water-use efficiency across all sectors by 2030. Furthermore, it aims to secure long-term withdrawals and freshwater supplies in order to alleviate water shortages and significantly reduce the number of people affected by water scarcity [3].

Water-use efficiency (WUE) in agriculture has become a critical priority due to the increasing constraints on water resources [4]. This is evident in the research that is directed towards understanding and improving agricultural WUE. For instance, studies have explored water-use efficiency in the context of climate change [5,6,7], technological innovation [8,9,10], and agricultural practices [11,12,13,14,15]. Additionally, researchers have studied the dynamics of the environment and their implications on WUE in agriculture [16,17,18]. Moreover, while studying WUE in light of the aforementioned aspects is important, it is also crucial to acknowledge the significant role that the behavioural aspects of irrigators play in contributing to water-use efficiency. For example, Callejas Moncaleano et al. [19] mentioned that one reason behind inefficient water use is human behaviour.

However, understanding irrigators’ water-use behaviour is important even outside of a WUE context. Several studies have emphasised the necessity of tailored policies that account for psychological and behavioural factors (e.g., [20,21,22,23,24]). In addition, understanding water-use-related behaviour is also crucial for water resource management [25]; scaling targeted solutions in agricultural water management challenges [26]; predicting responses to environmental changes accurately [27]; improving environmental programs’ effectiveness and efficiency [22]; designing effective strategies to promote water management technique [28]; and facilitating more effective intervention studies for behavioural change [29]. The aforementioned research demonstrates the overall importance of understanding behaviour for sustainable water management.

In addressing the complex interplay between human behaviour and sustainable water management, numerous studies have been undertaken employing various theories and models analysing human environmental behaviours. Callejas Moncaleano et al. [19] provide a comprehensive overview of these models, including the norm activation model (NAM); the new environmental paradigm (NEP); the theory of planned behaviour (TPB); the theory of values; the values, beliefs and norms theory (VBN); the theory of environmentally significant behaviour; and the risk, attitude, norms, abilities and self-regulation (RANAS). Although these conventional behavioural models have been extensively used, they have also been augmented to increase their predictive efficacy [30,31]. For example, studies have augmented the NAM [32,33,34,35], TPB [36,37,38,39,40,41], and the VBN [13,42,43]. The extension of these standard models indicates a growing recognition of the need to refine these models to better capture the complexities of behaviour in specific contexts.

The augmentation of conventional behavioural models is based on either considering potential specific factors or using empirical evidence to guide the augmentation process. In this regard, a literature review on the factors influencing water-use behaviour in agriculture could be a valuable starting point. According to the authors’ best knowledge, two reviews [19,44] have been conducted related to the water-use behaviour in agricultural contexts. Callejas Moncaleano et al. [19] reviewed the contextual and behavioural factors influencing water-use efficiency (WUE) and proposed a conceptual framework to link these factors to address the knowledge gaps on human water-use behaviour in light of WUE. The study by Meempatta et al. [44] reviewed the decision-making process of irrigators and the impact of their decisions on resource use, societal welfare, and environmental sustainability. Despite the valuable insights offered by the reviews of Callejas Moncaleano et al. [19] and Meempatta et al. [44], an up-to-date review is crucial to stay on track in achieving SDG 6, particularly target 6.4 (increase water-use efficiency across all sectors) and 6.5 (implement integrated water resources management at all levels). An up-to-date review is especially important, given the rapid advancements in agricultural technologies [45,46], climate change conditions [47,48], and changing social and economic conditions [49,50]. This paper addresses the existing knowledge gap by providing an up-to-date review of the factors influencing water-use behaviour in diverse agricultural contexts. Our review aims to provide an update on the current literature on water-use behaviour in agricultural contexts to ensure that interventions and strategies align with the latest body of knowledge.

2. Materials and Methods

This study will follow the framework developed by Koutsos et al. [51] for systematic reviews that set out a roadmap of guidelines for conducting an effective systematic review for agricultural research. This framework extends the basic steps provided by the PRISMA method. The systematic review framework proposed by Koutsos et al. [51] involves six steps: scoping, planning, identification, screening, eligibility/assessment, and presentation, as outlined in Figure 1.

Figure 1. A systematic review framework for agricultural research. Source: Adopted from Koutsos et al. [51].

2.1. Step 1: Scoping

The initial scoping phase is crucial, as it lays the foundation for subsequent stages in the systematic review process. Establishing a robust review protocol involves crafting a detailed research question and study design. We employed the Population, Intervention, Comparison, and Outcome (PICO) framework to structure a clearly defined research question. Table 1 outlines each component of the PICO framework formulated for this systematic review. Our research question is based on the specified PICO elements shown in Table 1: “What are the main factors influencing water-use behaviour among agricultural irrigators globally, taking into account psychological factors, decision-making processes, technological interventions, policy frameworks, and socioeconomic conditions?”. Regarding the study design, our systematic review will adhere to the comprehensive protocol proposed by Koutsos et al. [51], which entails six steps, as depicted in Figure 1.

Table 1. PICO elements for a systematic review.

PICO Component	Description
Population	Irrigators involved in water-use practices across various regions and contexts globally.
Intervention	Factors influencing water-use behaviour in agriculture, including but not limited to psychological factors, decision-making processes, technological interventions, policy frameworks, and socioeconomic conditions.
Comparison	Comparisons between different regions, countries, farming systems, and water management approaches
Outcome	Understanding the determinants and patterns of water-use behaviour in agriculture.

Note: Source: Authors’ compilation.

Furthermore, the authors involved with this systematic review independently screened and selected studies for inclusion to mitigate bias in study selection and interpretation. The web-based tool Rayyan was utilised to streamline the screening process. Rayyan is widely recognised as an effective tool for speeding up the laborious task of study selection within systematic reviews, fostering collaboration among researchers and boosting productivity [52]. The reviewing process for each reviewer was conducted in a blinded manner, ensuring that they could only view their own decisions and not those of the other reviewers involved.

Figure 2 provides an overview of the systematic integration of Rayyan in our review process. Firstly, search results from relevant databases were uploaded into Rayyan. Subsequently, Rayyan automatically identified duplicates, which the main author carefully managed to ensure the retention of the appropriate version. Following this, independent screening by reviewers took place with each article being categorised as ‘Include’, ‘Maybe’, or ‘Exclude’. Any discrepancies in categorisation and articles marked as ‘Maybe’ were thoroughly discussed among the authors to reach a consensus. Once the screening process was completed, subsequent steps in our systematic review could proceed.

Figure 2. Steps for Rayyan integration in systematic review. Source: Authors’ compilation.

A pilot review was conducted to familiarise the authors with the systematic review process and the use of the digital tool Rayyan. The pilot review also helped to refine the search strategy by testing search terms, databases, and inclusion/exclusion criteria and identifying challenges that needed to be resolved before conducting the final review. The pilot review was performed based on the keywords derived from the PICO framework in Table 1. The tailored keywords for each PICO element are presented in Table 2.

Table 2. Specific keywords tailored to each PICO element.

PICO Component	Tailored Keywords
Population	irrigators; farmers; producers; growers; water users; irrigation practitioners; agricultural practitioners; crop growers
Intervention	water use behaviour; water management practices; psychological drivers; irrigation techniques; decision-making; behavioural drivers
Comparison	regional; countries; farming systems; irrigation schemes; water management approaches; farming system variations; international perspectives
Outcome	determinants of water use behaviour; patterns of water use behaviour; water management practices; agricultural water usage patterns; behavioural insights in irrigation

Note: Source: Authors’ compilation.

2.2. Step 2: Planning

This study employed a systematic approach to the literature review, utilising Scopus, Web of Science, and EBSCOHost databases (specifically Academic Search Ultimate, Africa-Wide Information, CAB Abstracts with Full Text, E-Journals, GreenFILE, and MEDLINE). Scopus and Web of Science are well known for their coverage of various disciplines [53], whereas EBSCOHost databases provide a wide range of academic resources, including full-text articles, abstracts, and specialised information relevant to the study’s aims. Table 3 outlines the specific Boolean operators used for each database to ensure relevant search results.

Table 3. Databases and associated Boolean operators.

Database	Boolean Operators
Web of Science	(farm* or crop or crops or agricultur* or horticultur* or cropland*) and (“water use” or “water usage” or “water management” or “irrigation technique*” or “irrigation management*”) and (behav* or pattern* or practice* or insight* or adapt*) and title (manag* or use or usage or technique*) and (water* or irrigat*) and (behav* or pattern* or practice* or insight* or adapt*)
Scopus	(farm* or crop or crops or agricultur* or horticultur* or cropland*) and (“water use” or “water usage” or “water management” or “irrigation technique*” or “irrigation management*”) W/5 (behav* or pattern* or practice* or insight* or adapt*) and ti (manag* or use or usage or technique*) and (water* or irrigat*) and (behav* or pattern* or practice* or insight* or adapt*)
EBSCOHost (Academic Search Ultimate, Africa-Wide Information, CAB Abstracts with Full Text, E-Journals, GreenFILE, MEDLINE)	(farm* or crop or crops or agricultur* or horticultur* or cropland*) and (“water use” or “water usage” or “water management” or “irrigation technique*” or “irrigation management*”) N5 (behav* or pattern* or practice* or insight* or adapt*)) and ti (manag* or use or usage or technique*) and (water* or irrigat*) and (behav* or pattern* or practice* or insight* or adapt*)

Note: Source: Authors’ compilation derived from the keywords identified using the PICO framework. Note: The “*” symbol is a wildcard used to include all variations of a word (e.g., “farm” captures “farm”, “farms”, “farmer”, “farming”, etc.).

An additional search was conducted using the advanced search function in Google Scholar. Table 4 outlines the approach followed for this advanced search, detailing the specific search fields and terms used. The results obtained from Google Scholar were exported in ENW file format following the search. These search results were then uploaded to Rayyan for further screening and analysis.

Table 4. Google Scholar additional search protocol.

Google Scholar Advanced Search Field	Description
With all of the words	irrigators farmers water use irrigation
With the exact phrase	behave *
With at least one of the words	producers growers water users irrigation practitioners agricultural practitioners crop growers water management practices psychological drivers irrigation techniques behavioural drivers
Where my words occur	anywhere in the article
Return articles dated between	2020–2024

Note: Source: Authors’ compilation derived from the keywords derived from the PICO framework. Note: The “*” symbol is a wildcard used to include all variations of a word (e.g., “behav” captures “behavior”, “behaviour”, “behaviors”, “behaviours”, etc.).

2.3. Eligibility Criteria

For the selection of articles, the following inclusion criteria were set: (1) studies published within the last five years (2020 to 2024) to ensure the relevance of the research; (2) all studies published in peer-reviewed English language journals; (3) studies that investigate various interventions or practices related to water use in agriculture; (4) studies involving farmers, agricultural workers, or any individuals directly involved in agricultural activities related to water use will be considered; and (5) peer-reviewed papers, reports, and relevant publications that meet the criteria for eligibility will be included.

To ensure that the papers included in this systematic review are relevant, the following exclusion criteria were established: (1) book chapters were omitted to keep the focus on peer review articles; (2) papers that were work-in-progress or labelled as “short papers” were excluded; (3) to ensure that this systematic review does not duplicate existing secondary literature captured in other review types (such as scoping reviews, critical reviews, narrative reviews, opinion pieces, and systematic reviews), we specifically excluded these review types. Our systematic review focuses exclusively on primary studies, avoiding the duplication of research already covered by other systematic reviews; (4) studies not focused on water-use behaviour in agricultural contexts were also excluded; and (5) theses and dissertations are excluded from consideration to maintain consistency with the focus on peer-reviewed articles.

2.4. Step 3: Identification/Search

The predefined search queries were executed, and the results were exported. For the databases listed in Table 3, the search results were exported in RIS file format, and for Google Scholar (Table 4), the search results were exported in EWN file format. Both RIS and EWN file formats are compatible with Rayyan. The results were then uploaded and briefly scanned in Rayyan to confirm their relevance before proceeding with in-depth screening in the next step. During this initial screening, it was ensured that the search results accurately reflected the intended scope of the review. No additional changes were made to the search terms, as the formulated strategy was comprehensive and aligned with the keywords derived from the PICO framework.

The reporting of this systematic review was guided by the standards of the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) Statement (Figure 3). We retrieved 867 records, of which 709 were obtained from electronic databases (WOS, Scopus, and EBSCOHost), and 158 additional records were obtained from Google Scholar. After removing 361 duplicates, 506 records remained for screening of their titles and abstracts.

Figure 3. PRISMA flow chart to illustrate the article search and the inclusion process. Source: Authors’ compilation based on Page et al. [54].

2.5. Step 4: Screening Articles

At this stage, each reviewer categorised the 506 records in Rayyan as either “include”, “exclude”, or “maybe” based on their titles and abstracts (see Figure 3). Out of the 506 records, 381 were marked as “exclude”, while 71 were classified as “include”. Additionally, 31 records were tagged as “maybe”, and 23 records had conflicting classifications among the authors. The authors engaged in discussions to reach an agreement for the 31 “maybe” records and the 23 conflicting records. Eventually, 7 out of the 31 “maybe” articles and 5 out of the 23 conflicting records were included in the final analysis. Therefore, a total of 83 records remained to be assessed for eligibility.

2.6. Step 5: Eligibility/Assessment

The 83 articles that remained after the title and abstract screening were chosen for full-text analysis. Initially, attempts were made to retrieve the full texts of all articles. However, full texts for 3 out of the 83 articles could not be obtained, because access to these articles required payment that was not within our financial resources. Therefore, a total of 80 full-text records were evaluated against the predefined eligibility criteria to determine whether they aligned with the objectives of this study. A total of 33 articles were not deemed suitable and were removed based on the criteria, which resulted in 47 articles being considered eligible.

Based on the methodology described by van Dinter et al. [55], a quality assessment of the 47 articles was carried out to ensure that this systematic review maintained an acceptable standard. Table 5 displays the detailed criteria used to evaluate the quality of all 47 studies. The colour-coded system used in Appendix A, Table A1, shows how well each article met each criterion. Studies below a score of 4 out of 8 were removed, while those with a higher score were kept, ensuring high-quality input for our systematic review. Nevertheless, no studies scoring below four were identified, meaning the total number of articles to be included in our systematic review remained at 47.

Table 5. Quality checklist.

Quality Criteria	Question
Q1	Are the aims of the study clearly stated?
Q2	Are the scope, context and experimental design clearly defined?
Q3	Are the variables in the study likely to be valid and reliable?
Q4	Is the research process documented adequately?
Q5	Are all the study questions answered?
Q6	Are the negative findings presented?
Q7	Are the main findings regarding creditability, validity, and reliability clearly stated?
Q8	Do the conclusions relate to the purpose of the study? Are they reliable?

Note: Source: Adopted from Dinter et al. [55].

3. Results

This section provides a detailed analysis of the findings from our systematic review, offering insights into their interpretation and the broader implications drawn from the literature.

3.1. Publication Trends

In this review, the 47 articles selected for inclusion are summarised in Appendix A, Table A2. Table A2 provides an overview of the authors of each study, the year of publication, the country where the study was conducted, the type of water-use behaviour examined, and a summary of the key findings of each study. The year of publication for the 47 studies is illustrated in Figure 4. The publication output varied from 2020 to 2024, with the highest number of publications observed in 2021 and 2022, each of these years totalling 15 publications. Notably, the low number of publications in 2024 can be attributed to the fact that the searches for this systematic review were conducted in March 2024 and that publications in 2024 would likely increase after this systematic review.

Figure 4. Number of papers published in each year over the five years (2020–2024). Source: Authors’ compilation.

3.2. Geographic Distribution of Studies

The treemap shown in Figure 5 visually summarises the distribution of study counts across different countries within the final 47 articles identified. In Figure 5, the size of each rectangle in the treemap corresponds to the number of studies conducted in that specific country. The United States and China are the top countries in terms of research numbers on water-use behaviour in agriculture: the United States conducted eight studies, while China conducted seven studies between 2020 and 2024.

Figure 5. Treemap of study distribution across countries (2020–2024). Source: Authors’ compilation.

Furthermore, we employed keyword co-occurrence analysis to examine the thematic focus of water-use behaviour studies conducted in the U.S. (8 studies) and China (7 studies). The analysis was conducted using Nvivo 14 software, and density visualisation plots were derived from the results. Separate density plots were created for the eight studies conducted in the U.S. (Figure 6) and the seven studies in China (Figure 7). These density plots highlight the predominant themes emerging from each country’s research on water-use behaviour in agricultural contexts, enabling a comparative analysis between the U.S. and China.

Figure 6. Bibliometric network analysis of agricultural water-use behaviour studies conducted in the U.S. Source: Authors’ compilation. Note: Nodes represent keywords extracted from the literature. Larger and brighter nodes indicate keywords that appear more frequently or are central to the network. Connections (edges) between nodes represent co-occurrences of keywords within the selected studies. The colour scheme in the density visualization plot ranges from bright red to blue. Key nodes are represented in bright red, indicating higher density or significance, while the surrounding areas are shaded in blue, showing lower density or significance.

Figure 7. Bibliometric network analysis of agricultural water-use behaviour studies conducted in China. Source: Authors’ compilation. Note: Nodes represent keywords extracted from the literature. Larger and brighter nodes indicate keywords that appear more frequently or are central to the network. Connections (edges) between nodes represent co-occurrences of keywords within the selected studies. The colour scheme in the density visualization plot ranges from bright red to blue. Key nodes are represented in bright red, indicating higher density or significance, while the surrounding areas are shaded in blue, showing lower density or significance.

The density visualisation plot (Figure 6) for the eight studies conducted in the U.S. reveals how research efforts have focused on understanding and enhancing water-use behaviour in agricultural contexts. Key themes such as “water conservation”, “soil moisture sensors”, “irrigation management”, and “agriculture” indicate that these topics have been prominent in the studies analysed. Additionally, there is increasing attention towards “climate change beliefs”, “adaptation”, and “irrigation”, underscoring a growing recognition of climate impacts and the necessity for adaptive strategies in agricultural water management. The visualisation where “water conservation” is the largest and bright red, close to nodes like “soil moisture sensors”, “agriculture”, and “irrigation”, suggests strong connections and frequent co-occurrences among these concepts in eight U.S. studies. Furthermore, the connection between “climate change beliefs” and “adaptation” suggests that U.S. farmers’ perceptions and beliefs about climate change are an important factor shaping their adaptive behaviours and water-use practices.

The keyword co-occurrence analysis of the seven studies on water-use behaviour in China, shown in Figure 7, reveals six significant keywords. The analysis highlights “management” as the largest and brightest node, closely linked to “scarcity”, indicating a strong research focus on managing water resources in response to scarcity issues in China. This underscores significant efforts in developing and implementing management strategies to address water scarcity challenges. The proximity of nodes for “determinants” and “land” suggests a strong relationship between factors influencing water use and land-related issues. Furthermore, “smallholder farmers” is closely located to “land”, emphasising the importance of land use practices by smallholder farmers in research. This clustering shows recent research emphasis on understanding the various determinants and land management practices, particularly involving smallholder farmers, in studies conducted in China.

The density visualisation plots provide a detailed overview of the main themes and concepts explored in research on water-use behaviour in the U.S. and China. To complement these insights, Appendix A, Table A3 summarises the key topics addressed in each study.

Table A3 shows that studies on water-use behaviour in Chinese agriculture cover key topics such as government policies and social security measures affecting water-management practices; the implementation and impact of water pricing policies; farmers’ decision-making processes and risk preferences; adoption of water-saving technologies; the role of social networks and economic incentives in water management strategies; and awareness of environmental consequences related to agricultural water use. These topics reflect the diverse factors influencing water-use behaviours and strategies in Chinese agriculture.

Furthermore, Table A3 reveals that studies on water-use behaviour in U.S. agriculture explored key topics such as irrigation practices and scheduling, the adoption of water conservation technologies, multilevel modelling of water-use behaviour, and policy implications of water conservation efforts. Overall, Table A3 highlights the different regional priorities and challenges faced by farmers in the U.S. and China.

3.3. Key Theme and Topics

Next, a word frequency analysis was conducted using Nvivo 14 software to identify key themes and focus areas discussed across the 47 articles included in this systematic review. Figure 8 presents the word frequency cloud derived from the full-text body of the 47 articles and places greater prominence on words that occurred most frequently across the 47 articles. The words “climate”, “irrigation”, “water”, “change”, “farmers”, and “adoption” were found to be the most frequently occurring words expressed in red, indicating that these words occurred with the highest frequency.

Figure 8. Word frequency cloud derived from the full 48 documents included in this systematic review. Source: Authors’ compilation. Note: Words in the centre are the largest and appear in red, indicating the highest frequency. These central words are surrounded by slightly smaller words in bold black, representing moderate frequency. The bold black words are further surrounded by smaller words in grey, indicating lower frequency.

In addition to the insights obtained from Figure 8, a word frequency table (Table 6) was constructed to provide a deeper understanding of the most prominent words identified across the 47 articles in this review. Table 6 displays the top 30 words based on their frequency in the 47 articles, giving an overview of their occurrence. From Table 6, the top three words by frequency are “water”, “irrigation”, and “farmers”, with high word counts of 7674, 4270, and 2783, respectively. The word “water” has a weighted percentage of 2.01%, meaning it makes up 2.01% of all words across the 47 articles. Furthermore, the ranking of words in the top ten of occurrence, such as “farmers”, “adoption”, and “management”, highlights the vital role of agricultural practitioners in adopting water-saving methods and efficiently managing water resources. In addition, the inclusion of words such as “climate” and “change” suggests a growing concern regarding the impacts of climate change on water resources and agricultural practices.

Table 6. Word frequency.

Word Ranking	Word	Length	Count	Weighted Percentage	Word Ranking	Word	Length	Count	Weighted Percentage
1	water	5	7674	2.01	16	crop	4	1001	0.26
2	irrigation	10	4270	1.12	17	risk	4	888	0.23
3	farmers	7	2783	0.73	18	data	4′	873	0.23
4	climate	7	1817	0.47	19	information	11	847	0.22
5	change	6	1679	0.44	20	level	5	845	0.22
6	adoption	8	1659	0.43	21	factors	7	828	0.22
7	management	10	1589	0.42	22	research	8	823	0.22
8	agricultural	12	1558	0.41	23	saving	6	817	0.21
9	farm	4	1344	0.35	24	land	4	816	0.21
10	farmers’	8	1310	0.34	25	soil	4	790	0.21
11	adaption	10	1276	0.33	26	used	4	789	0.21
12	technology	10	1216	0.32	27	area	4	758	0.20
13	model	5	1176	0.31	28	variables	9	749	0.20
14	study	5	1138	0.30	29	agriculture	11	744	0.19
15	practices	9	1042	0.27	30	policy	6	740	0.19

Note: the weighted percentage in a word frequency query assigns a portion of the word’s frequency to each group to ensure the overall total does not exceed 100%. Source: Authors’ compilation.

3.4. Keyword Co-Occurrence Network

The word frequency cloud (Figure 8) and Table (Table 6) highlight several key themes covered across the literature: water management and irrigation; farmers and their perceptions; climate change and its impacts; and agricultural practices, management, and technology adoption.

In addition, further insights were gained by constructing a keyword co-occurrence network (Figure 9) to identify key themes, clusters, and connections within our included studies. The keyword co-occurrence network was generated in Vosviewer, where 383 keywords were extracted, among which 33 keywords meet the threshold of three, meaning that the minimum occurrence of a keyword is three. Each node in Figure 9 represents a keyword with the size of the node indicating the frequency of occurrence of that keyword. Larger nodes signify keywords that are more significant within the corpus. The distance between nodes reflects the relatedness of keywords based on their co-occurrence, where nodes closer to each other indicate that keywords appear together more frequently in the same articles. The thickness of the lines between the nodes indicates the strength of the co-occurrence relationship. Furthermore, keywords are grouped into clusters that are thematically connected. Additionally, the colour of each node denotes the cluster to which the keyword belongs, and the words within each cluster tend to co-occur more frequently with each other than with words in other clusters.

Figure 9. Keyword co-occurrence network. Source: Authors’ compilation. In the keyword co-occurrence network, the different colours represent different clusters. Each cluster indicates a group of keywords that frequently occur together, highlighting related themes or topics within the dataset.

A total of four clusters occurs in Figure 9, each with distinct colours. In the first cluster (red), keywords such as “adaptation”, “conservation”, “innovation”, “management”, “sustainability”, and “water management” are all related to sustainable agricultural practices and resource management. In the second cluster (green), keywords such as “adoption”, “agriculture”, “irrigation”, “productivity”, and “technologies” are all associated with the adoption and use of technologies in agriculture to improve productivity, particularly in water management. Conversely, phrases like “climate change”, “adaptation strategies”, “impacts”, “risk”, and “water scarcity” are included in the third cluster (blue), suggesting a thematic emphasis on understanding and addressing the impacts of climate change on water availability and agricultural practices. Lastly, the fourth cluster (yellow) highlights keywords such as “farmers”, “model”, “technology”, “uncertainty”, and “water conservation”, suggesting a thematic focus on modelling approaches and uncertainties in agricultural water management.

Viewing the interlinkages between clusters in Figure 9 reveals that “adaptation” (cluster one, red) is closely linked to “irrigation” and “agriculture” (cluster two, green), which implies that efficient irrigation practices are an important component of adaptive strategies in agriculture. In addition, “adaptation” (cluster one, red) is also linked to “climate change” and “impacts” (cluster three, blue), which highlights the need for adaptive strategies to mitigate the adverse effects of climate change and the related impacts of climate change.

3.5. Types of Water-Use Behaviours Investigated

Furthermore, a detailed breakdown of the various specific water-use behaviours investigated across the 47 studies was provided. From the 47 articles, a total of six distinct categories of water-use behaviour have been identified. The following categories of behaviours were used: (1) adaptation to climate change; (2) adoption of water-saving technologies; (3) adoption of water conservation techniques/practices; (4) complying with water-saving policies; (5) water-use behaviour in light of institutional performance; and (6) seasonal irrigation water usage patterns. A sunburst chart (Figure 10) was created to visually represent the distribution of studies across the six water-use behaviour categories. It is important to note from Figure 10 that some studies address multiple water-use behaviours. Figure 10 reveals that the adoption of water conservation techniques/practices (behaviour 3) emerges as the dominant focus among the identified water-use behaviours, with a notable 24 studies dedicated to this specific behaviour. Following closely, the adoption of water-saving technologies (behaviour 2) emerges as another prominent behaviour investigated with 17 studies dedicated to studying the adoption of water-saving technologies. However, the remaining five water-use behaviours identified (behaviours 1, 4, 5, and 6) received considerably less attention. Specifically, six studies focused on adaptation to climate change (behaviour 1), while four studies examined compliance with water-saving policies (behaviour 4). Additionally, only one study examined water-use behaviour in relation to institutional performance (behaviour 5), and similarly, only one study focused on seasonal irrigation water usage patterns (behaviour 6).

Figure 10. Distribution of studies across water-use behaviour categories. Source: Authors’ compilation. Note that the number in the middle layer of the sunburst chart represents a specific water-use behaviour: (1) adaptation to climate change; (2) adoption of water-saving technologies; (3) adoption of water conservation techniques/practices; (4) complying with water-saving policies; (5) water-use behaviour in light of institutional performance; and (6) seasonal irrigation water usage patterns. Note: Alotaibi and Kassem [20]; Ataei et al. [21]; Bourceret et al. [22]; Seidl et al. [24]; Yuan et al. [25]; Dessalegn et al. [26]; Sabzevar et al. [27]; Sabbagh and Gutierrez [28]; Wang et al. [41]; Woldeselassie et al. [56]; Michalak [57]; Usman et al. [58]; Davenport et al. [59]; Wheeler et al. [60]; Abebe et al. [61]; Wang et al. [62]; Mi et al. [63]; Berthold et al. [64]; Xiuling et al. [65]; Fan and McCann [66]; Bailey et al. [67]; Cavazza et al. [68]; Roberts et al. [69]; Sajid et al. [70]; Reints et al. [71]; Danso et al. [72]; Cui et al. [73]; Mekonnen et al. [74]; Muenratch and Nguyen [75]; Alauddin et al. [76]; Hoang-Thi et al. [77]; Ortiz et al. [78]; Momenpour et al. [79]; Johnson et al. [80]; Irmak et al. [81]; Bogdan and Kulshreshtha [82]; Salas-Zapata et al. [83]; Rafii [84]; Velasco-Muñoz et al. [85]; Fischer and Sanderson [86]; Asthana [87]; Ricart and Rico-Amorós [88]; Ren and Yang [89]; Mahdavi [90]; Aldaya et al. [91]; Leroy [92]; Gao et al. [93].

Next, the six water-use behaviours in Figure 10 were further analysed by exploring the driving factors behind each behaviour as identified in the relevant studies. Considering the diverse agricultural contexts reflected across the 47 studies, a deeper understanding of the variables that have been found to influence each water-use behaviour will provide valuable insights into the broad spectrum of determinants driving water-use behaviours in diverse agricultural landscapes. It is clear from Figure 11 that the variables found to influence climate change adaptation in agricultural water-use-related contexts are multifaceted. Farmers’ ability to implement adaptive measures has been found to be influenced by economic and financial factors, institutional and information access, farm characteristics and resources, sociodemographic factors, and perceptions and experience. The economic and financial factors identified include cost and availability of resources [56], financial resources [57], debt levels [60], debt-to-equity ratios [60], and access to credit [58,61]. Institutional and information access factors encompass support from government institutions [57], access to information [57], information on weather forecasting [60], information on climatic and natural hazards [58], and agricultural extension services [58,61]. Farm characteristics and resources include crop characteristics and requirements [56,60], irrigation timing and frequency [56], supplemental irrigation [56], crop type [57], farm size [57,58,60,61], tenancy status [58], irrigated area [60], and soil quality [58]. Sociodemographic factors are age, education, and household size [58]. Lastly, perception and experience factors involve future drought exposure belief [59], farm sensitivity appraisal [59], drought risk perception [59], climate change perceptions [59,61], and previous adaptation experience [61].

Figure 11. The factors influencing adaptation to climate change in agricultural water-use contexts. Source: Authors’ compilation. Note: Woldeselassie et al. [56]; Michalak [57]; Usman et al. [58]; Davenport et al. [59]; Wheeler et al. [60]; Abebe et al. [61].

The results in Figure 12 indicate that the implementation of water-saving techniques (water-use behaviour 2) is impacted by a range of factors such as demographic and socioeconomic variables, awareness and perceptions, institutional and policy influences, environmental and farm characteristics, psychological and behavioural elements, information access, and interaction and collaboration. Demographic and socioeconomic variables include education level [62,64,65,67,70], off-farm work participation [62], household income level [67], age [28,64,65,71], family size [73], and health [68]. Awareness and perceptions encompass environmental awareness [41], water scarcity perception [27], perceived ease of use of technology [41], knowledge about water management [27], awareness towards water pollution and disasters [20,73], and awareness of water management interventions [70]. Institutional and policy factors comprise access to extension services [62,69], subsidies [63,70,72], cooperative membership [63], and supportive policies [68]. Environmental and farm characteristics involve crop types [20,66,67], irrigation practices [67], planting scale [63,73], regional climates [71], income from crops [71], farm size [65,66,70], irrigation complexity [71], climate variability [66] and seasonal precipitation patterns [67]. Psychological and behavioural elements include subjective norms [41], attitudes [27,41], perceived control [41], self-efficacy and risk-taking [27], effort, and performance expectancy [28,69]. Informational variables include general access to information [27,66] and access to reliable data [64]. Interaction and collaboration include neighbours’ decisions [63], coordination and cooperation among relevant actors [68], friendships among irrigators [74], spatial proximity between farms [74], and shared knowledge among farmers [74].

Figure 12. Factors influencing the adoption of water-saving technologies in agricultural water-use contexts. Source: Authors’ compilation. Note: Alotaibi and Kassem [20]; Sabzevar et al. [27]; Sabbagh and Gutierrez [28]; Wang et al. [41]; Wang et al. [62]; Mi et al. [63]; Berthold et al. [64]; Xiuling et al. [65]; Fan and McCann [66]; Bailey et al. [67]; Cavazza et al. [68]; Roberts et al. [69]; Sajid et al. [70]; Reints et al. [71]; Danso et al. [72]; Cui et al. [73]; Mekonnen et al. [74].

The adoption of water conservation techniques/practices (Figure 13) in agricultural contexts is influenced by a wide range of variables that can be categorised into several key areas. Demographic and socioeconomic factors play a significant role, including variables such as age and education level [67,70,71,76,82,83], household income [77,80], household assets [77], land ownership [76,82], and gender [26,80]. Awareness and perception variables were also found to be significant with influences including environmental knowledge [27,75,84], awareness of extraction impacts and environmental consequences [20,41,75,85,87], perception of climate change and water scarcity [27,77,80], and perceived water availability [80]. The third category of variables focuses on technological, institutional and policy factors, which encompass groundwater tariffs [75], training programs [41,70], institutional factors [26,79,80], and supportive regulations and policies [26,81]. Informational variables were also significant, including access to prior weather information [76], general information [67], and access to information [27]. The next category identified focuses on environmental and farm characteristics, including topography and soil type [69,76,79], irrigated land area, elevation, rainfall [77,81], climatic conditions [80], farm size and location [70,78,82,83,88], water quality and context [26,86,88], and cultivated crop variety [20]. Another crucial category identified is the psychological and behavioural factors, which entail variables such as subjective norms [21,41,75], perceived behavioural control [41,75], attitudes [21,27,41,84], self-efficacy [84], personal norms and intentions [21], and environmental values [86]. Lastly, the adoption of water conservation techniques/practices was found to be influenced by interaction and collaboration factors, including social networks [85], social norms [21], social capital [27], and other social factors [79]. These categories collectively highlight the complex interplay of factors that drive the adoption of water-saving technologies in agriculture.

Figure 13. Factors influencing the adoption of water-saving conservation techniques/practices in agricultural water-use contexts. Source: Authors’ compilation. Note: Alotaibi and Kassem [20]; Ataei et al. [21]; Dessalegn et al. [26]; Sabzevar et al. [27]; Wang et al. [41]; Bailey et al. [67]; Roberts et al. [69]; Sajid et al. [70]; Reints et al. [71]; Muenratch and Nguyen [75]; Alauddin et al. [76]; Hoang-Thi et al. [77]; Ortiz et al. [78]; Momenpour et al. [79]; Johnson et al. [80]; Irmak et al. [81]; Bogdan and Kulshreshtha [82]; Salas-Zapata et al. [83]; Rafii [84]; Velasco-Muñoz et al. [85]; Fischer and Sanderson [86]; Asthana [87]; Ricart and Rico-Amorós [88].

Figure 14 summarises the factors influencing compliance with water-saving policies in agricultural water-use contexts. Farm characteristics, including farm productivity [24,89], farmland value and excess water holdings [24], farm area under irrigation [24], soil type and agronomic factors [91], and mobility of resource units [89] have been found to play an essential role in farmers’ decisions to comply with water-saving policies. In addition, economic factors were also found to be important, such as farm financial metrics [24,25,91], labour considerations [91], water pricing, and market conditions [91]. Environmental factors are also crucial determinants, encompassing climatic variables and precipitation levels [25,91], regional variations in water resources [89], and territorial features [91]. Furthermore, psychological and behavioural aspects were also identified as important, entailing factors such as attitudes [22,90], subjective norms [22,90], perceived behavioural control [22,90], adventurous perceptions and tolerance for profit variation [25]. Moreover, demographic and socioeconomic dimensions, including gender and education levels [24], were found to be significant. Perceptions related to climate change risks were also found to act as a significant driver [24].

Figure 14. Factors influencing compliance with water-saving policies in agricultural water-use contexts. Source: Authors’ compilation. Note: Bourceret et al. [22]; Seidl et al. [24]; Yuan et al. [25]; Ren and Yang [89]; Mahdavi [90]; Aldaya et al. [91].

The remaining two water-use behaviours (behaviours 5 and 6) are illustrated together in Figure 15, since only one study has been found for each of these behaviours. In light of water-use behaviour 5 (institutional performance), the categories of influencing variables include farm characteristics (location within the irrigation scheme/area, access to groundwater), interaction and collaboration (social capital, group size of irrigators in specific regions), technological and institutional (irrigation technique used, organisation and association membership) and awareness and perception (perceptions of fairness in water distribution, perceptions of water scarcity) [92]. For seasonal irrigation water usage patterns (behaviour 6), the influencing variables include farm characteristics (soil properties, crop type), technological factors (number of irrigation days in the season, early-season irrigation intensity, duration of the irrigation period), and environmental factors (climate predictors) [93].

Figure 15. Influential factors in agricultural water-use behaviour in light of institutional performance (water-use behaviour 5) and seasonal irrigation water usage patterns (water-use behaviour 6). Source: Authors’ compilation. Note: Leroy [92]; Gao et al. [93].

4. Discussion

In this systematic review, we explored the different water-use behaviours in agricultural settings as reported by the 47 studies included in this review. Our analysis highlighted several key themes and trends in the studies reviewed, providing valuable insights into current practices and areas needing further investigation. Firstly, the geographic distribution of the 47 studies revealed a concentration of research efforts in a few key countries over the past five years. The United States leads with eight studies, which is followed closely by China with seven studies. As highlighted in the introduction, it is clear that understanding behavioural aspects is crucial for achieving water-use efficiency, sustainable agriculture, and broader water resource management goals. However, it is concerning that research efforts directed at understanding water-use behaviour in agricultural contexts over the past five years are concentrated in only a few countries. This geographic distribution of studies is concerning, especially in light of water stress affecting various regions globally, as reported by the Food and Agricultural Organization [94]. The FAO [94] applied the SDG indicator 6.4.2 (this indicator measures the level of water stress and is defined as the ratio of total fresh water withdrawn by all major sectors (agricultural, industrial and municipal) to total renewable freshwater resources, after considering environmental flow requirements) as a measurement of water stress and illustrated that the highest levels of water stress are found in East Asia, western Asia, central Asia, southern Asia, and northern and southern Africa. By not adequately studying water-use behaviour in these countries, we risk hindering progress towards sustainable water management and ultimately lowering the likelihood of achieving SDG 6 before 2030.

Moreover, our results from the word frequency and the keyword co-occurrence network analysis revealed several key themes. The keyword co-occurrence network analysis identified four main clusters, each representing a distinct theme. The first cluster’s theme was identified as “Sustainable Agricultural Practices”, emphasising the importance of long-term resource availability and minimising environmental impacts through efficient water management in agriculture. This finding is sensible given the extensive attention that sustainable agricultural practices in general have received in the literature [95,96,97,98,99], particularly focusing on sustainable water-management practices [100,101,102,103]. Foguesatto et al. [95] highlighted the various potential benefits of adopting sustainable regional and global agricultural practices. However, Foguesatto et al. [95] further mentioned that despite these potential benefits, the adoption rate of these practices remains low in many countries. Considering the points mentioned above might explain why our systematic review revealed sustainable agricultural practices as a prominent theme, since a better understanding of water-use behaviour in the context of sustainable agricultural practices could potentially increase the adoption rate of these practices.

The theme of the second cluster centres around technology adoption for productivity, highlighting the role that technological advancements play in enhancing agricultural productivity, particularly in the context of water management and water-use efficiency. Jararweh et al. [104] argued that given the restricted farming space, water shortages, climate change, and continually changing environmental circumstances, new and novel smart agricultural technologies must be developed. However, according to Levidow et al. [105], while innovative irrigation practices can enhance water efficiency, provide economic advantages, and reduce environmental burdens, farmers do not fully realise these benefits due to various challenges. Barriers include the cost of adopting new technology [104,106], lack of adequate knowledge and data [63,104], and risk of reduced yield [63]. Policies and educational programs are vital to support farmers in overcoming these challenges [107]. Ultimately, integrating advanced technologies into agriculture is crucial to achieving sustainable water management and is seen as an important way to reduce water scarcity [108]. This importance is underscored by our results, which show that 17 of the 47 papers addressed the adoption of water-saving technology as a water-use behaviour, demonstrating an increasing need to understand and promote water-saving technologies in agricultural contexts.

Furthermore, the results of the keyword co-occurrence showed that the theme of the third cluster centres around climate change, adaptation strategies, and water scarcity, particularly concerning agricultural practices. The emergence of climate change adaptation and water scarcity as common themes in our systematic review is sensible, given that agriculture is widely recognised as one of the most vulnerable sectors to climate change [109,110,111,112]. According to Mehrazar et al. [113], given the vulnerability of the agriculture sector to climate change, studying the potential impacts of climate change is crucial to ensure food and water security in the future. Moreover, Hanjra and Qureshi [114] highlighted that with increasing population and income growth, there will be a corresponding increase in demand for food and water. Hanjra and Qureshi [114] further mention that irrigation will be the first sector to lose water as water competition by non-agricultural sectors will intensify, which will have implications for food security, hunger, poverty and ecosystem health and services. There is a growing need to develop effective adaptation measures to safeguard agricultural productivity against the adverse effects of climate change and water scarcity. Strategies such as climate-resilient crop varieties [115,116,117], improved irrigation techniques [117,118,119], and sustainable water-management practices [112,120] emerge as critical components in mitigating the adverse effects of climate change on water availability and agricultural systems. However, addressing climate change’s adverse effects, Srivastav et al. [121] emphasised the importance of following a holistic approach. This aligns with our systematic review’s findings, which revealed the multifaceted nature of farmers’ adaptation to climate change in water-use-related contexts. Future studies on farmers’ adaptive behaviour in water-use-related contexts should consider a holistic approach in modelling their adaptative behaviours.

The fourth cluster identified in our keyword co-occurrence analysis highlighted the theme of modelling and uncertainty in water conservation. The significance of employing models to understand behaviour cannot be overstated. Through behavioural models, researchers gain a better understanding of the complex interplay of factors influencing water usage in agriculture. These models not only serve an academic purpose but are also important to inform policymakers and stakeholders to develop effective water-saving behaviour change strategies [122], efficient resource allocation [123], and effective policy design [124]. It is, therefore, crucial to measure behaviour accurately. According to Reimer et al. [125], it is crucial to consider contextual factors when measuring behaviour, as the variation in these factors can make it difficult to predict and measure farmers’ adoption/behaviour. However, our systematic review revealed six water-use behaviours and the various variables influencing these behaviours over the past five years (2020 to 2024). Therefore, our systematic review can be a valid starting point for future studies when modelling behaviour. Additionally, uncertainty emerged as a significant theme in the fourth cluster, underscoring its pivotal role in water management decisions. Dessai and Hulme [126] emphasised that uncertainty is inherently part of decision making in environmental management, influencing the sensitivity of management decisions to uncertainties in environmental predictions. For example, Sun et al. [127] showed that projections of future climate change are plagued with uncertainties, making it difficult for decision-makers to plan adaptation measures. Our keyword co-occurrence analysis results align with the points raised by Dessai and Hulme [126] and Sun et al. [127], as the word “uncertainty” is linked to both “climate change” and “management”, suggesting that uncertainty is an issue affecting various aspects of water management in light of climate change. Therefore, when modelling behaviour, it is crucial to account for the influence of uncertain environmental factors on farmers’ water-use behaviour.

Furthermore, the findings from our systematic review revealed the multifaceted nature of the variables driving farmers’ water-use behaviours in diverse agricultural contexts. Our systematic review identified six water-use behaviours, each with its unique drivers and barriers. However, several overarching themes emerge despite the differences among the six water-use behaviours identified. Firstly, it is clear that a wide range of factors influence farmers’ decisions and actions regarding their water usage. Various factors such as economic considerations, demographic variables, socioeconomic variables, farmers’ awareness and perceptions, institutional and policy influences, technical factors, farm and environmental characteristics, psychological and behavioural elements, informational factors and interaction and collaboration among relevant stakeholders are all crucial components. The complex interplay of these various variables in influencing agricultural water-use behaviours highlights the need for comprehensive approaches to water management policies and interventions. The importance of following a comprehensive approach to water management is evident, as highlighted by past initiatives such as the Comprehensive Assessment of Water Management in Agriculture (CA) [128].

Secondly, our results underscore the complexity of water-use behaviours in agricultural settings. It is crucial to acknowledge the limitations of existing behavioural models in capturing this complexity. Academically, the recognition of these complexities in human decision making is evident, with various studies augmenting traditional behavioural models by integrating additional constructs [36,37,39,41]. Our findings provide a significant starting point for future studies aiming to measure water-use behaviour in agricultural contexts. Subsequent research can build upon our findings to determine essential constructs for a more comprehensive understanding.

Lastly, considering that the 47 articles included in our systematic review are from diverse regions globally, it is crucial to recognise that the unique contextual factors of each agricultural setting significantly influence water-use behaviours. Farm characteristics, such as farm size, location and type of crop produced, irrigation complexity, and scale of plantings, interact directly with environmental factors such as the climate, soil quality, and seasonal precipitation patterns. Recognising the contextual interplay is essential for resource efficiency [129], reducing water waste and pollution [130], promoting water sustainability research [131], and designing tailored interventions [132].

5. Conclusions

This systematic review aimed to explore and synthesise the various water-use behaviours in agricultural contexts as reported in the literature from the past five years (2020 to 2024). Our findings provided critical insights, identifying key factors influencing various water-use behaviours in agricultural contexts as well as thematic trends emerging from the 47 studies included in our review. Our review draws several important conclusions about water-use behaviour in agricultural contexts. Firstly, the geographic distribution of the studies showed a significant concentration of research in a few countries over the past five years, notably the United States and China. The concentration of studies in certain countries highlights a geographical gap in research efforts. Addressing this gap is crucial for global progress towards sustainable water management and achieving SDG 6 by 2030. Secondly, our review identified six primary water-use behaviours studied in diverse agricultural settings, which are influenced by a wide range of factors ranging from farm characteristics, economic considerations, and institutional and environmental conditions to psychological and behavioural aspects. The complex interplay of variables found to influence these water-use behaviours highlights the importance of a comprehensive and context-specific approach to promoting sustainable water-management practices in agriculture. Lastly, the themes that emerged from our systematic review highlighted the critical areas of sustainable agricultural practices, technology adoption for productivity, climate change adaptation and water scarcity, and the importance of modelling and addressing uncertainty in water management. These themes underscore the significant progress made in recent years but also highlight persistent challenges.

The implications of our findings extend to policy, emphasising the need for policies that account for the various factors influencing farmers’ water-use behaviours. This includes economic incentives, educational programs, and support for technology adoption to enhance water-use efficiency. Furthermore, given the significant influence of regional and environmental factors, water management policies should be tailored to specific agricultural contexts considering local climatic conditions, soil types, and water availability to increase the effectiveness. This includes promoting sustainable agricultural practices to optimise water-use efficiency and minimise environmental impacts. Additionally, facilitating the adoption of innovative water-saving technologies can significantly enhance agricultural productivity while conserving precious water resources. Investing in climate adaptation strategies, such as resilient crop varieties and improved irrigation techniques, is crucial for building resilience to changing climatic conditions and mitigating the impacts of water scarcity. Furthermore, policymakers should prioritise the development of accurate behavioural models to inform decision making and tailor interventions effectively.

In addition, we recommend that governments enhance funding dedicated to teaching farmers about efficient water-use practices and fostering sustainable water-use behaviours. Such efforts are crucial for ensuring sustainable water management and securing water resources for future generations.

Acknowledging certain limitations, the present study followed a systematic procedure to identify keywords and screening sources. However, this process has inherent limitations due to the subjectivity of the authors in selecting and screening studies. Additionally, the reliance on available databases might have excluded relevant studies not indexed in those sources. Language bias is another limitation, as the review included only studies published in English. Moreover, due to financial constraints, we were unable to access three full-text articles, potentially introducing bias by not fully representing all relevant literature.

In conclusion, we recommend that future research focuses on modelling water-use behaviours. Although we identified six water-use behaviours in our systematic review, we will illustrate the application of our findings using one specific behaviour: adaptation to climate change in agricultural water-use contexts (water-use behaviour 1, Figure 11). Our systematic review highlighted that factors such as economic and financial considerations, institutional and information access, farm characteristics and resources, sociodemographic factors, and perceptions and experiences influence farmers’ adaptation to climate change.

We have augmented the theory of planned behaviour (TPB) with these categories to show how these influencing variables can be integrated into standard behavioural models such as the TPB. The TPB was chosen for illustration purposes due to its widespread use in understanding human behaviour, which is supported by extensive empirical evidence across various fields in the literature. Figure 16 depicts the augmented adaptation to climate change TPB model. By integrating insights from our systematic review, researchers can develop more tailored models to better understand and predict water-use behaviour in diverse agricultural contexts. While Figure 16 focuses on the variables influencing climate change adaptation in water-related contexts identified in this systematic review, it is still important for future studies to consider the contextual setting of their specific research.

Figure 16. Augmented theory of planned behaviour (TPB) model for adaptation to climate change in agricultural water-use contexts. Source: Authors’ compilation.