Factors influencing farming households’ climate change adaptation strategies in Central Vietnam

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 1 / 22 OPEN ACCESS Citation: Nhuong BH, Truong DD, Huan LH, Lan BTH, Hang ND, Tam DD (2025) Factors influencing farming households' climate change adaptation strategies in Central Vietnam. PLoS One 20(7): e0328058. https:// doi.org/10.1371/journal.pone.0328058 Editor: Sandra Ricart, Politecnico de Milano, SPAIN Received: June 21, 2024 Accepted: June 25, 2025 Published: July 10, 2025 Peer Review History: PLOS recognizes the benefits of transparency in the peer review process; therefore, we enable the publication of all of the content of peer review and author responses alongside final, published articles. The editorial history of this article is available here: https://doi.org/10.1371/journal. pone.0328058 Copyright: © 2025 Nhuong et al . This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, RESEARCH ARTICLE Factors influencing farming households’ climate change adaptation strategies in Central Vietnam Bui Huy Nhuong1, Dinh Duc Truong 2*, Le Huy Huan2, Bui Thi Hoang Lan2, Nguyen Dieu Hang2, Duong Duc Tam2 1 University Board of Management, National Economics University (NEU), Vietnam, 2 Faculty of Environmental, Climate Change and Urban Studies, NEU, Vietnam * truongdd@neu.edu.vn Abstract This study investigates the determinants of household-level adaptation strategies to climate variability and saltwater intrusion in the coastal regions of Central Vietnam.

Using a Multinomial Logistic Regression (MNL) model, the analysis is based on a cross-sectional survey of 356 farming households, focusing on four primary adapta- tion measures: vegetable production, shrimp farming, adoption of salt-tolerant rice varieties, and the lotus-fish farming model. The results reveal that socio-economic, demographic, environmental, and institutional factors significantly influence the adoption of specific adaptation strategies. Key determinants include gender, edu- cation, age, farming experience, household income, land characteristics, access to information, credit services, membership in civic organizations, and participation in training programs. Male-headed households and those with greater access to climate information and social networks were more likely to adopt diverse and complex adap- tation strategies. In contrast, households with limited land resources, lower incomes, or lacking institutional support were less adaptive. The findings highlight the hetero- geneity of adaptive behaviors and the need for tailored interventions. From a policy perspective, enhancing institutional capacity—especially through targeted training, increased access to subsidized credit, and support for community-based organiza- tions—can significantly strengthen farmers’ adaptive capacities. Moreover, the study contributes to filling key research gaps in the Southeast Asian context by integrating socio-economic and environmental variables into a unified analytical framework. These insights are critical for designing inclusive and effective climate adaptation pol-

icies aimed at safeguarding rural livelihoods and promoting sustainable agricultural development in climate-vulnerable regions.

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 2 / 22 1. Introduction Climate change and the increasing frequency of extreme weather events have had profound effects on agricultural systems, particularly in countries which geographical and climatic conditions render them highly sensitive to environmental variability [1–3]. These unpredictable climatic shifts not only undermine the quality and productivity of agricultural outputs but also compel farmers to invest more heavily in adaptation measures, ultimately reducing profit margins and exacerbating economic losses [4–7]. Alarmingly, approximately 52% of global agricultural land is currently classified as moderately to severely degraded [5]. The compounded effects of climate change and extreme weather can trigger a cascade of socio-economic instabilities, contrib- uting to poverty and development challenges across household, community, and national scales [6–10]. For farmers, a key approach to mitigating these adverse impacts is to adjust agricultural systems in ways that enhance resilience to changes in temperature, precipitation patterns, and climate-induced hazards [11,12]. Broadly defined, climate change adaptation refers to adjustments in natural or human systems in response to actual or anticipated climatic stimuli or their effects, with the aim of moderating harm or exploiting potential benefits [13,14]. These behavioral adjustments can significantly reduce agricultural vulnerability and exposure to climatic risks [15,16]. Adaptation,

however, is inherently complex, encompassing multi-dimensional and multi-scalar processes [17–19]. It may occur in various forms—autonomous or planned, and proactive or reactive—and can also be classified by temporal duration, spatial scope, and strategic approach [20,21]. Farmers’ selection of adaptation strategies is typically influenced by a combination of demographic, socio-economic, institutional, and agro- ecological factors. Nonetheless, empirical evidence suggests that the effect of these determinants varies depending on the specific livelihood systems and agricultural contexts in which farmers operate [22,23]. Consequently, there is a pressing need to identify the key factors shaping adaptation choices and the capacity of farmers to mitigate climate change impacts on their environments, livelihoods, and associated socio-economic and institutional structures, thereby informing effective policy instru- ments and management frameworks [24–26]. The Asian region, including Vietnam, is among the most climate-vulnerable areas globally, characterized by rapidly shifting climatic conditions that pose substantial challenges for communities attempting to respond and adapt [27,28]. Vietnam, in particular, is identified as one of the five countries most severely affected by cli- mate change, owing to its long coastline, high population density, and concentra- tion of socio-economic activities in low-lying coastal areas [27,29,30]. According to Department of Climage Change (2020), a 100 cm rise in sea level could inundate

approximately 40,000 km²—equivalent to 12.1% of the nation’s current land area— potentially displacing up to 17.1 million people [27]. This rise would devastate the Red River and Mekong Deltas, two of the most critical agricultural and economic zones in the country [30]. Over the past decade, the economic damage attributable to climate change has been estimated at 2.5% to 3.0% of GDP annually, and in the absence of effective adaptation policies, projections indicate this figure may rise to and reproduction in any medium, provided the original author and source are credited. Data availability statement: All data are in the manuscript and/or supporting information files. Funding: This study is funded by the National Economics University, Vietnam. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist.

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 3 / 22 10% of GDP per year by 2050 [26,31]. The agricultural sector, which remains central to Vietnam’s economy and food security, is particularly vulnerable. Climate change exerts adverse effects on agricultural production, threatening the live- lihoods of farming communities, national food security, and the value of agricultural exports [32,33]. As one of the world’s top five rice-exporting nations, Vietnam’s agricultural supply chains are increasingly exposed to climate-induced risks [26,34]. Despite various initiatives, climate adaptation efforts in Vietnam continue to face multiple constraints, including high exposure to hazards, heightened sensitivity, and limited adaptive capacity, especially in resource-scarce rural areas [35–37]. Globally, a substantial body of literature has explored the determinants of climate change adaptation strategies among communities and agricultural households in both developed and developing economies [38–40]. These studies typically identify behavioral, psychological, and socio-economic characteristics as key influencing factors. Theoretical frameworks underlying such research often derive from behavioral models, technology acceptance theories, and collective action paradigms [41–43]. Despite the substantial progress in the global literature on climate change adaptation among farming households, important research gaps remain—particularly in the context of transitional economies such as Vietnam, where institutional

reforms, environmental vulnerability, and livelihood diversification intersect in complex ways. First, while a significant body of empirical research has examined household-level adaptation in regions such as Sub-Saharan Africa and South Asia, there is a noticeable dearth of studies focusing on Southeast Asia, especially in Vietnam. Within this region, rural house- holds residing in coastal and salinity-affected zones face unique climate risks, including saltwater intrusion, erratic rainfall, and seasonal droughts. Yet, household-level analyses of how these communities perceive, respond to, and cope with such hazards remain limited. As a result, current adaptation models may lack external validity when applied to South- east Asian coastal contexts, where socio-cultural and agro-ecological dynamics differ substantially from those studied in African or South Asian cases. Second, the majority of existing research tends to analyze adaptation using either a single strategy (e.g., crop diversification or irrigation) or aggregate adaptation indices, without unpacking the distinct types of strategies households may pursue simultaneously. Such an approach overlooks the inherent heterogeneity, sequencing, and trade-offs embedded in farmers’ adaptation decisions. In highly variable environments, households often adopt a combination of strategies based on resource constraints, perceived risks, and institutional access. Therefore, there is a critical need to disaggregate adaptation behavior across specific climate risks—such as drought versus saltwater intru-

sion—and to recognize that different strategies may be complementary, substitutive, or even conflicting. Third, although socio-economic, environmental, and institutional drivers of adaptation have been studied individually, few studies have integrated these dimensions within a coherent, multi-level analytical framework. This lack of integration has resulted in fragmented explanations that fail to capture the complex interplay of structural conditions, environmental pressures, and enabling institutions. Consequently, policy responses derived from such partial understandings may be less effective in addressing the real constraints faced by farmers on the ground. Fourth, behavioral and psychological determinants—such as risk perception, time preferences, trust in institutions, and intra-household dynamics—are frequently overlooked or only superficially addressed in adaptation studies. Yet these factors critically shape how individuals interpret climate risks, eval- uate adaptation options, and commit to action. In particular, farmers’ subjective expectations about future climate trends, their past experiences with crop failure, or the influence of gendered power relations within households can significantly influence whether and how adaptation occurs. The absence of these elements from most quantitative analyses leaves an important explanatory gap in understanding why some households adapt while others do not, even when faced with similar climatic and economic conditions [44,45].

From the analysis of the context and gaps in previous studies, this study proposes three main research questions (RQ). RQ1. What are the key socio-economic, institutional, and environmental factors that shape the heterogeneous adoption of climate change adaptation strategies among farming households in coastal and salinity-prone areas of transitional economies like Vietnam?

overlooked behavioral and psychological aspects of adaptation, including gender dynamics, experience, and perception of risk, which are critical in shaping real-world decision-making under uncertainty. Fifth, by disaggregating adaptation responses to specific stressors such as saltwater intrusion and drought, the research provides context-sensitive insights that can inform localized adaptation planning. Finally, the study contributes practical value by generating evidence-based policy recommendations tailored to different household profiles, thereby supporting the development of inclusive and scal- able adaptation strategies aligned with national climate resilience goals. 2. Materials and methods 2.1 Study site This study was conducted in Quang Binh province in the Central region of Vietnam, specifically in Tuyen Hoa district (17.7882° N, 106.2051° E) in the upper reaches of the Gianh River, which has been severely affected by drought and saltwater intrusion over the past 10 years. Gianh River is originated from the 2,017 m high Co Pi mountainside area in the Truong Son range, flowing through the districts of Minh Hoa, Tuyen Hoa, Bo Trach, Quang Trach, and Ba Don commune, then going to the East Sea at Cua Gianh. The Gianh River upstream is located in the northwest of Tuyen Hoa district, from the mountains on the Lao Bao peninsula. Mountainous terrain and steep hills, with small streams, waterfalls, and lakes, characterize this area. In the Tuyen Hoa district, the Gianh River upstream flows through the production areas of three

communes Chau Hoa, Phong Hoa, and Duc Hoa [46,47] Tuyen Hoa is always affected by northern air masses in winter (Southeast wind) and hot, dry West winds in summer (Laos wind). The average annual temperature ranges from 24°C - 25°C, the average annual rainfall is about 2,100– 2,300 mm and is divided into two distinct seasons: the rainy season and the low-rainy season. The rainy season is from September to March of the following year, the average temperature is from 22°C - 23°C, the lowest temperature is in January (about 10°C – 14°C) [48]. Rainfall is concentrated in September, October and November, accounting for 70% − 80% of the total annual rainfall reaching 2,000–2,300 mm, due to the influence of the northeast monsoon combined with disturbances that cause heavy rain such as storms and high pressure. Tropical lows, tropical convergence zones, leading to flooding in the delta. The rainy season is from April to August, the average temperature is about 26°C −27°C. The hot- test month can reach 39°C (June, July and August), the average rainfall only accounts for 20% − 30% of the total annual rainfall [35,47]. The economy of Tuyen Hoa district and the three above communes is mainly based on agriculture. In particular, water from the river has played an essential role in developing agriculture and enriching the locality. This is also where

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 5 / 22 people rely on the Gianh River for daily life and agricultural production. The river water is also used for irrigation, aqua- culture, and providing domestic water for the community. According to Mai and Truong (2022), with current changes in weather and environment, the Gianh River upstream area is facing challenges regarding water resources. The pro- longed heatwave and increasing seawater intrusion have also caused unwanted impacts on the lives and production of the communities. Reality shows that heatway and saltwater intrusion incidents have increased significantly in recent years compared to the period 2000–2010 and usually occur from April to July every year. These two types of disasters often come together and affect agricultural livelihoods. Firstly, it affects aquaculture: increases the cost of aquaculture activities (if salinity increases, farmers must add sugar to the water or pump more fresh water to neutralize and reduce pond salinity to < 25%). If the phenomenon of saltwater intrusion occurs for a long time, farmers cannot pump water into the pond, causing the shrimp to get sick (because there is not enough oxygen to breathe), so aquaculture output can decrease and cause total loss. Second, it affects farming activities: making the soil saline and acidic, rice plants are unable to grow, productivity decreases by 60–70%, people cannot cultivate and convert to shrimp farming ponds,

brackish water crab or rice fish have higher economic efficiency, most of the land area is in low-lying areas. In addition, drought causes lack of water to irrigate the rice growing area, mainly in the Winter-Spring crop. Planted rice having not enough irrigation water will might reduce productivity. If the drought lasts about 1 month, it can cause crop failure (due to dead rice or flat rice grains) [35,46–48]. 2.2 Empirical model development Adaptation has been widely acknowledged as a critical pillar in the policy response to climate change, particularly in the context of agricultural resilience and rural livelihoods [23,49]. A substantial body of literature has examined the ways in which various regions—especially in developing countries—respond to climate-induced challenges, as well as the key factors shaping these adaptive behaviors [50–52]. Existing studies have demonstrated that farmers’ adaptation strategies are highly context-specific and influenced by a range of factors, including but not limited to farm size, land type, agro- climatic conditions, and broader ecological, institutional, socio-economic, cultural, and political contexts [41,42,50,52]. More recent analyses have identified a complex interplay of determinants that extend beyond structural factors to include institutional accessibility, political frameworks, social networks, cultural norms, behavioral and cognitive dimensions, and gender dynamics [53]. Despite the growing interest in climate change adaptation within the Global South, including parts

of Asia and Africa, there remains a notable research gap concerning the specific barriers and enabling conditions for adopting adaptive strategies in Vietnam, particularly at the household level [44,45,54]. In response to this gap, the present study investigates the factors influencing farming households’ decisions to adopt climate variability adaptation strategies in Vietnam. The analysis encompasses a range of determinants across three broad categories: (i) socio-economic variables such as farm size, income levels, civic organization membership, and farming experience; (ii) demographic characteristics including household size and age of the household head; and (iii) institutional factors such as access to credit, markets, technical training, and climate-related information. These factors are conceptualized in the analytical framework summa- rized in Fig 1, which illustrates the multidimensional nature of adaptation decision-making among rural households. To identify essential variables in farmers’ choice of adaptation strategies, we performed a multinomial logit analysis model (MNL) as previous studies [55–58]. The model allowed determining the relationship between the probability of choosing an adaptation option and a set of explanatory variables, then coming up with the determination of probabilities of different selected adaptive options [59]. In MNL model, let Y be a random dependent variable with values {1 to J} represents the adaptation strategies chosen

by farming households (1 to J) and let Xi be the set of variables described as socio-economic characteristics, institutional factors, and climatic properties. To show how changes in the elements are determined by summing the probabilities of P (Yi) = j Xi . The probability of a dependent variable belonging to the nth category in a multinomial logistic regression model is expressed as below [59,60].

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 6 / 22 πj = exp(∑K k=1 βjkxk) 1 + ∑J–1 j=1 βjkxk with j = 1, 2, . . . . J – 1 (1) In models with dependent variables of more than two types, it is necessary to identify a base type for comparison or analy- sis. The base category (J) can be chosen as 0 for the dependent variable and consisting of categories 0, 1, 2, 3, and 4. The model was linearized by taking the natural logarithm of these odds ratios to obtain logistic models. If J is chosen as the base type, then the probability of the dependent variable being in the base type is determined as given in the equation below. πj = P(Y = J) = 1 1 + ∑J–1 j=1 exp[(∑K K=1 βjkxk)] with j = 1, 2, . . . . J – 1 (2) The overall relationship between an individual independent variable and the dependent variables were tested by testing the coefficient – 2likelihood based on summing the probabilities associated with the predicted and actual outcomes. –2LogLikelihood = ∑N n=i ([Yi lnP (Yi)) + (1 – Y i) (3) Fig 1. Proposed study model of impact factors. Source: Study design (2024). https://doi.org/10.1371/journal.pone.0328058.g001

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 7 / 22 The marginal effect of exogenous factors is given by: δj = Pj Xk = Pj [ βjk – ∑J–1 J=1 Pjβjk ] = Pj(βj – β) (4) The marginal effect measures the expected change in the probability of a particular choice being made for a unit change in the explanatory variable. The signs of the marginal effects and the corresponding coefficients could be different because the coefficients depended on the signs and magnitudes of all the other coefficients [61–64]. Table 1 described the variables expected to be included in the model to analyse factors affecting the choice of livelihood strategies to adapt to climate change of farming households. The adoption of these strategies was hypothesized to be influenced by demo- graphic and socio-economic characteristics, such as gender, age of the household head, household size, education level, farm size, annual income, and number of livestock and breeds owned; institutional factors, such as credit services, agri- cultural extension services, social networks, access to information and awareness of climate change. 2.3 Data collection and analysis This study employed both primary and secondary data sources to ensure comprehensive and contextually grounded analysis of the factors influencing climate change adaptation strategies among farming households. Primary data were collected through a combination of qualitative and quantitative methods. Specifically, focus group discussions and in-depth

interviews were conducted with local officials, commune leaders, and villagers to explore community-level perceptions and challenges associated with climate change adaptation. These methods were instrumental in contextualizing the research, refining the questionnaire design, and capturing localized knowledge and lived experiences that quantitative data alone could not fully reveal. The qualitative insights also informed the selection of variables for subsequent modeling and anal- ysis. In addition to field-generated data, secondary data were collected from several authoritative sources to complement Table 1. Description of explanatory variables in the model. No. Explanatory variables Code Expected signs Measurement 1 Gender of respondent GEN +/- 1: male 0: female 2 Household size SIZE +/- Continuous 3 Education level of respondent EDU +/- Continuous 4 Age of respondent AGE +/- Continuous 5 Household monthly income INC +/- Continuous 6 Farming experience EXP +/- Continuous 7 Percentage of soil salinity SAL +/- Continuous 8 Location of farmland POS +/- Continuous 9 Farmland area ARE +/- Continuous 10 Access information from the state INF + 1: yes; 0: no 11 Membership of the local associations MEM + 1: yes; 0: no 12 Access credit from policy banks CRE + 1: yes; 0: no 13 Participate in training courses TRA + 1: yes; 0: no Source: Study design (2024). https://doi.org/10.1371/journal.pone.0328058.t001

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 8 / 22 and validate the primary findings. These included socio-economic and climatic data Quang Binh Statistics Department, and the Department of Climate Change under the Ministry of Natural Resources and Environment (MONRE). These sources provided critical background information on agricultural practices, historical climate variability, and household demographics in the study area. The sample size for the household survey was determined using a simplified formula recommended by Greene (2000) [65] expressed as: n = N 1 + N ∗ e2 (5) where n represents the required sample size, N is the total population, and e is the allowable margin of error, set at 5% in this study. Based on official statistics from the Quang Binh Statistics Office (2021) [48], the combined population of the three selected communes—Chau Hoa, Phong Hoa, and Duc Hoa— were approximately 14,521 individuals. With an average household size of 4.5 persons per household in rural areas [46], the estimated total number of households was 3,226. Applying the above formula yielded a required minimum sample size of 356 households to ensure statistical reli- ability and representativeness. The three communes were purposefully selected due to their high concentration of farming households and significant exposure to climate risks [46,48]. Given the relatively even distribution of households across the selected sites, a proportional random sampling approach was adopted. Accordingly, 119 households were surveyed

in Phong Hoa, 119 in Duc Hoa, and 118 in Chau Hoa. Lists of households were obtained from communal authorities, from which respondents were selected using simple random sampling techniques to minimize selection bias. To ensure ethical compliance and participant engagement, surveys were conducted in the evenings when household members were more likely to be present. Prior to the interviews, each participant was provided with a clear explanation of the study’s objectives and procedures. Informed consent was obtained in writing through a consent form, in which partici- pants voluntarily indicated their willingness to participate and signed their names. All selected households agreed to take part in the study. The official survey was conducted over a two-month period, from November to December 2023. Upon completion of data collection, the survey responses were screened for completeness and consistency. Cleaned data were then coded and entered into the Statistical Package for the Social Sciences (SPSS) version 23.0 for quanti- tative analysis. The initial phase of analysis included descriptive statistics, frequency distributions, and cross-tabulations to summarize household characteristics and assess general attitudes and perceptions regarding climate change and its effects on agricultural livelihoods. ANOVA tests were also conducted to examine differences in awareness and adaptive behavior across socio-economic subgroups. To identify the key determinants of adaptation strategy choices, the study

employed the MNL model. This model was appropriate given the categorical nature of the dependent variable—house- holds’ choice of specific adaptation strategies—and allowed for the estimation of the likelihood that a household selects a particular strategy given a set of independent socio-economic, demographic, and institutional variables. The MNL results provided empirical evidence on the significance and direction of each factor’s influence on adaptive decision-making in the context of climate-related stressors. 3. Results 3.1 Socio-economic characteristics of the sample Table 2 presents the socio-economic characteristics of the sampled farming households. The gender distribution of house- hold heads indicated that 59.78% were male and 40.22% were female, suggesting that male-headed households consti- tuted the majority of the sample. The average household size was 5.23 members. Larger households were observed to utilize a broader range of livelihood resources, derive greater income from environmental activities, and exhibit a higher likelihood of adopting climate change adaptation strategies compared to smaller households. The age of the household

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 9 / 22 head was employed as a proxy for farming experience. It was found that younger household heads generally possessed less familiarity with the socio-ecological context and agricultural practices than their older counterparts. Age categories were established based on standard psychomotor developmental criteria, classifying respondents into three groups: young, adult, and elderly. The results showed that the majority of respondents fell into the adult category, with an aver- age age of 45.24 years. This demographic profile is advantageous for the study, as middle-aged individuals are likely to have accumulated substantial experiential knowledge about local climate variability, its impacts, and relevant adaptation practices. Education was found to play a significant role in enhancing adaptive capacity. Educated respondents demonstrated a higher propensity to adopt adaptation measures, likely due to improved access to information, technology, and diversified income sources. Survey results revealed that 14.89% of respondents were illiterate, while 85.11% were literate. This illiter- acy rate is relatively high compared to national averages, consistent with patterns observed in other rural areas of Quang Binh province. The high proportion of illiterate farmers may hinder their capacity to respond effectively to climate change due to limitations in accessing adaptation-related knowledge and technologies. Farm size, defined as the total area of

agricultural land owned by a household, was measured in hectares. The data indicated that land ownership ranged from 0.1 to more than 3 hectares, with an average of 1.65 hectares per household. Livestock ownership was used as a proxy for household wealth, given that livestock remains a critical indicator of economic status in rural areas. The average livestock holding across sampled households was equivalent to 1,363 kilograms in terms of stockholding weight. Monthly household income was also assessed as a key economic indicator. The mean monthly income of surveyed households was 16,890,560 VND (approximately 700 USD). These figures reflect the relatively low-income status of the sampled population and underscore the economic vulnerability of farming households to climate-related shocks. Institutional factors Table 2. Socio-economic description of the sample. Variables Categories Frequency (N = 356) Percentage Means Gender Female 172 40.22 – Male 184 59.78 Age Young 99 27.81 45.24 Adulthood 231 64.89 Old 26 7.30 Household size (person) 1–4 153 42.98 5.23 5–8 149 41.85 9–12 54 15.17 Education level Illiterate 53 14.89 – Primary 109 30.62 Secondary 116 32.58 High school 48 13.48 College and above 30 8.43 Landholding (hectare) Less than 3 251 70.51 1.65 More than 3 105 29.49 Livestock holding (kg) 0-1,000 331 92.98 1,363 More than 1,000 25 7.02 Monthly income Below 10 million VND 179 50.28 16,890,560 11−30 million VND 156 43.82 More than 30 million VND 21 5.90 Source: Study results (2024).

https://doi.org/10.1371/journal.pone.0328058.t002

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 10 / 22 were also examined, particularly in terms of access to agricultural extension services and weather-related information. The findings indicated that 78.21% of households had access to agricultural extension services, while 81.34% reported receiving information on weather and climate conditions. These institutional supports are considered crucial for enhancing farmers’ capacity to anticipate and respond to climatic variability and extreme events. 3.2 Livelihood vulnerability to climate change and adaptation strategies by households The present study investigated the level of awareness and the types of adaptation strategies adopted by farming house- holds in response to disasters induced by climate change in the local context of Quang Binh province. A key focus was placed on understanding how communities perceive climatic changes and how this perception informs their coping and adaptive behaviors. The findings reveal that a substantial majority—approximately 85.23% of surveyed households— demonstrated an awareness of the ongoing manifestations of climate change and the increasing frequency of extreme weather events occurring in their locality. This high rate of awareness suggests that climate change and its impacts have become observable and tangible realities for the local population. Households reported that the most commonly experienced climate-related hazards in the past five years include floods,

tropical storms, saltwater intrusion, and drought. These hazards have not only occurred with increasing frequency but also with greater intensity and duration, posing heightened threats to agricultural livelihoods. When asked to compare the fre- quency of such disasters to previous years (with 2018 serving as a reference point), 51.9% of respondents perceived that droughts had become more frequent, followed by 42.2% who observed an increase in flooding, and 31.3% who reported a rise in saltwater intrusion. Interestingly, although only 36.6% of respondents noted an increase in storm frequency, 63.4% believed that the frequency of storms had decreased. Nonetheless, the majority emphasized that the intensity of individ- ual storms has significantly escalated. For example, the storm Wutip, which struck in October 2018, was cited by multiple respondents as particularly destructive, underscoring the need to assess not only the occurrence but also the severity of climatic hazards (Table 3). Respondents further identified the temporal patterns of specific climate-induced disasters. Flooding is typically observed from August to October and may extend into November, often associated with the typhoon season. In contrast, droughts and saltwater intrusion are most prevalent between April and July, with the peak intensity occurring between April and June. During this period, the lack of rainfall combined with high temperatures and salinity intrusion severely undermines rice production—the primary livelihood activity in the area. Interviewees consistently reported that saltwater

intrusion, when combined with drought, causes substantial losses in rice yields, particularly in low-lying paddy fields with poor irrigation infrastructure. These conditions significantly disrupt planting calendars, reduce agricultural productivity, and increase the risk of long-term livelihood insecurity. The seasonal concurrence of storm-induced flooding and drought- related phenomena places significant pressure on the already limited adaptive capacity of local communities. In the months of August through November, storms often result in flash floods, damaged irrigation systems, and waterlogging of crops. Conversely, during the dry season from April to July, extended dry spells and salinity intrusion compromise water Table 3. Perception of types of accidents occur every month of the year. Unit: % Month 1 2 3 4 5 6 7 8 9 10 11 12 Salinization 2.1 3.2 2.3 15.8 20.2 25.5 10.3 5.1 8.1 7.4 8.3 2.1 Drought 0 0 0.8 26.2 66.1 88.3 54.4 6.3 0 0 0 0 Flooding 0 0 0 0 0 0 2.12 43.1 93.2 72.6 14.1 0.8 Storm 0 0 0 0 0 0 0 20.5 86.4 80.7 56.5 1.2 Source: Study results (2024). https://doi.org/10.1371/journal.pone.0328058.t003

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 11 / 22 availability for irrigation and domestic use, further compounding vulnerability. The spatial and temporal overlap of multiple stressors necessitates context-specific adaptation strategies tailored to distinct hazard profiles across different times of the year. Importantly, the subjective observations of the surveyed population regarding changing climate patterns and disaster risks align closely with national assessments and projections. The observed trends in increased drought severity, pro- longed flood seasons, and rising storm intensity are consistent with the findings outlined in Scenario of climate change, sea level rise for Vietnam [27]. This convergence between community-level perceptions and scientific projections lends further credibility to the argument that climate change has transitioned from a theoretical concern to a lived experience among rural households in Vietnam. The insights gained from this study emphasize the importance of localized climate knowledge and the role of community-based observations in informing adaptation planning. They also highlight the necessity of enhancing farmers’ adaptive capacity through timely dissemination of climate information, disaster preparedness education, and investment in resilient agricultural infrastructure. The findings provide a foundation for future interventions aimed at improving resilience and fostering sustainable rural development in the face of accelerating climatic threats.

To respond effectively to the multifaceted impacts of climate change, a notable proportion of surveyed farming households—approximately 42.13%—reported having adopted at least one or more adaptation strategies tailored to their specific livelihood conditions. The descriptive statistical analysis revealed a diversified array of coping mecha- nisms utilized by farmers in Tuyen Hoa district, reflecting both agricultural innovation and livelihood diversification. Among the identified adaptation practices, shrimp farming emerged as the most prevalent strategy, adopted by 76.69% of households. This was followed by lotus-fish integrated farming systems (64.33%), and the cultivation of vegetables (55.9%) as alternative or supplementary sources of income and food security. Furthermore, 66.85% of respondents reported switching to new salt-tolerant rice varieties—a crucial adaptation in the context of increasing soil salinity driven by saltwater intrusion. In terms of technological and infrastructural adaptation, approximately 23.88% of households had installed small-scale irrigation systems, which, although limited in coverage, are essential in areas with erratic rainfall and unreliable surface water availability. Adjustments in agricultural calendars also constituted a common adaptation approach, adopted by 49.72% of households. These adjustments were made to synchronize agricultural activities with seasonal weather variations. Specifically, aquaculture activities were typically initiated between February and

May (lunar calendar), the winter-spring rice crop was planted from late December to April, and the summer-autumn crop was sown from mid-May to mid-August annually. Additionally, 34.83% of respondents indicated participation in non-agricultural employment as a strategy to diversify income sources and reduce dependence on climate-sensitive sectors (Table 4). Table 4. Adaptation strategies of farming households against saltwater intrusion. Adaptation strategies n % of respondents Shrimp farming 273 76.69 Lotus and fish farming 229 64.33 Vegetable production 199 55.90 Small-scale irrigation system 85 23.88 Salt-tolerant rice varieties 238 66.85 Agricultural calendar changing 177 49.72 Non-agricultural work participation 124 34.83 Source: Study results (2024). https://doi.org/10.1371/journal.pone.0328058.t004

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 12 / 22 3.3 Challenges for climate change adaptation To better understand the constraints to adaptation implementation, farmers were asked to identify and rank key challenges that hinder the effective adoption of these strategies. This evaluation was conducted using a 5-point Likert scale ranging from 1 (“completely disagree”) to 5 (“strongly agree”). The mean scores of each statement were calculated and ranked accordingly to reflect the perceived severity of the challenges. The results of the ranking exercise highlighted “lack of water for irrigation” as the most critical constraint. This issue was consistently emphasized by respondents as a major barrier to sustaining agricultural production, particularly in the context of coastal Vietnam, where freshwater scarcity is exacerbated by climate-induced phenomena such as prolonged droughts, saltwater intrusion, and river salinization. Given the reliance of farming systems on water sources—rivers, rainfall, canals, and groundwater—any disruption to water availability significantly undermines productivity and food security. Although governmental organizations (GOs), non-governmental organizations (NGOs), and local civic groups have engaged in efforts to address water management issues, the challenges remain pervasive and unresolved in many areas. The second highest-ranked constraint was “lack of cultivable land.” With ongoing demographic expansion and rural-

urban transformation, increasing amounts of farmland are being converted to non-agricultural uses, such as housing development, industrial infrastructure, and public facilities. This trend of agricultural land loss poses a serious threat to cli- mate resilience, as it reduces the physical space available for adaptation-oriented diversification and sustainable farming practices. The third most frequently cited constraint was “lack of capital.” Financial limitations were reported by 48% of the respondents as a significant obstacle preventing the adoption of adaptation strategies. Limited access to credit and low savings rates reduce farmers’ ability to invest in climate-resilient technologies, improved seeds, irrigation systems, and other necessary inputs. In this context, access to microfinance, rural credit services, and public subsidies may play a criti- cal role in facilitating wider adaptation adoption. Other identified barriers included “lack of information” and “lack of technology,” which were noted by approximately 32% and 30% of the respondents, respectively. These constraints highlight the role of institutional support and information dissemination in climate adaptation processes. Without timely access to reliable climate forecasts, agricultural extension services, or demonstrations of new technologies, farmers are likely to rely on traditional practices that may no longer be sufficient under changing environmental conditions. In summary, the findings emphasize that adaptation to climate change among smallholder farmers in Central Vietnam

is constrained by a combination of physical, financial, and informational barriers. Addressing these barriers—particularly those related to irrigation water, land availability, capital resources, access to climate information, and agricultural technol- ogies—is essential for enhancing the adaptive capacity of rural communities. Doing so will contribute not only to the resil- ience of farming systems but also to the broader goals of sustainable agricultural development and rural poverty reduction in Vietnam (Fig 2). 3.4 Factors influencing households’ choice of climate change adaptation strategies To investigate the determinants influencing the adoption of climate change adaptation strategies among farming house- holds in the study area, a MNL model was employed. This model is particularly suitable for analyzing discrete choice data where the dependent variable represents the selection among multiple, unordered alternatives. In this study, the depen- dent variable was defined as the household’s choice of one of four main adaptation strategies: (1) vegetable production, (2) shrimp farming, (3) adoption of salt-tolerant rice varieties, and (4) lotus-fish integrated farming. Each of these strate- gies reflects a distinct approach to mitigating the adverse effects of climate variability and environmental stressors specific to the local agro-ecological context. For robustness of analysis and binary classification purposes, these four adaptation strategies were pooled, and an

index-based transformation was applied. Specifically, households that adopted adaptation strategies at a level above the

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 13 / 22 sample mean were coded as ‘1’ (i.e., indicating active application of adaptive strategies), while those adopting strategies at or below the average level were coded as ‘0’ (i.e., no or limited adaptation). This dichotomization facilitated the identifi- cation of significant factors distinguishing adaptive versus non-adaptive households. Prior to estimating the MNL model, the potential multicollinearity among independent variables was assessed using Variance Inflation Factor (VIF) [65]. All VIF values were found to be below the threshold value of 2.0, indicating an absence of serious multicollinearity concerns. This confirms that the explanatory variables included in the model were sufficiently independent of one another, ensuring the reliability of coefficient estimates. Model adequacy and overall goodness-of-fit were assessed through the Likelihood Ratio (LR) Chi-square test and Pearson Chi-square statistics. The LR Chi-square value was 81.032, statistically significant at the 1% level (p < 0.01), suggesting that the explanatory vari- ables jointly contribute to the variation in adaptation strategy choices and that the model significantly improves upon the null model with no predictors. Furthermore, the Pearson Chi-square statistic for goodness-of-fit was 182.342, also highly significant (p < 0.01), providing strong evidence that the model has a good fit to the observed data and demonstrates ade- quate explanatory power.

The empirical results from the MNL model indicate that the selection of climate change adaptation strategies among farming households is not random but is systematically influenced by a range of factors. These include household-level socio-economic characteristics (such as education, income, household size, and farming experience), institutional factors (such as access to extension services, information on climate change, and participation in civic organizations), organiza- tional capacity (such as access to credit and irrigation infrastructure), and environmental characteristics (including land- holding size, frequency of extreme weather events, and water availability). Each of these variables exerts a differentiated impact on the probability of selecting a specific adaptive strategy, reflecting the heterogeneity in household capacities and climate exposures. These findings underscore the importance of designing context-specific adaptation policies that are sensitive to the underlying socio-economic and environmental drivers influencing household behavior. In particular, strengthening institutional support systems, enhancing access to agricultural services and information, and addressing resource-related constraints (e.g., water, land, and capital) are crucial for promoting the adoption of effective and sustain- able adaptation practices at the grassroots level (Table 5). Specifically, the findings in Table 6 indicated that gender (GEN) significantly influenced adaptation choices. Male-

headed households were more likely to adopt shrimp farming (+31.6%), lotus-fish farming (+41.2%), and salt-tolerant rice Fig 2. Challenges for climate change adaptation of farmers. Source: Study results (2024). https://doi.org/10.1371/journal.pone.0328058.g002

PLOS One | https://doi.org/10.1371/journal.pone.0328058 July 10, 2025 14 / 22 Table 5. Results of MNL analysis. Vegetable production Shrimp farming Salt-tolerant rice varieties Lotus and fish farming VIF Variables B Sig. Exp(B) B Sig. Exp(B) B Sig. Exp(B) B Sig. Exp(B) Intercept 0.234 0.912 0.059 0.932 9.783 0.00 4.534 0.032 GEN −0.354 0.567 0.745 1.321 0.042** 2.987 1.732 0.003*** 6.324 0.989 0.312** 3.313 1.83 SIZE −0.301 0.432 0.789 −0.112 0.789 0.981 1.564 0.000*** 4.324 0.645 0.045** 3.432 1.64 EDU 1.623 0.000*** 4.021 −0.708 0.071** 0.645 0.608 0.067** 1.785 −0.211 0.645 0.678 1.46 AGE −0.178 0.345 0.843 −0.213 0.654 0.708 0.432 0.078** 0.822 −0.032 0.756 0.867 1.75 INC 0.702 0.065** 2.019 0.205 0.322 1.453 0.004 0.998 1.323 0.322 0.876 1.423 1.58 EXP 0.342 0.424 1.432 −0.599 0.323 0.607 0.199 0.798 1.543 −0.543 0.321 0.876 1.44 SAL −0.601 0.587 0.643 −0.132 0.876 0.895 2.567 0.061* 10.342 1.432 0.312 2.764 1.39 POS −0.089 0.832 0.912 0.673 0.069** 2.868 −0.345 0.436 0.654 0.312 0.543 1.312 1.28 ARE −0.243 0.656 0.802 0.883 0.102** 2.124 1.812 0.000*** 5.756 1.312 0.032** 2.977 1.74 INF −0.061 0.912 0.899 0.064 0.899 1.069 0.032 0.998 1.434 −0.433 0.534 0.677 1.49 MEM −0.312 0.632 0.657 −0.073 0.876 0.911 0.045 0.997 1.543 −0.534 0.123 0.656 1.63 CRE −0.189 0.872 0.799 1.321 0.065** 3.567 0.732 0.245 1.432 0.766 0.032** 2.543 1.81