Integrating Local Knowledge and Community Practices for Flood Resilience in the Volta Basin

Abstract

Flooding, exacerbated by climate change, urbanization, and poor land-use practices, is a growing challenge for rural households in the Volta Basin. This study examines the effectiveness of flood management practices in improving household resilience in Benin and Togo. Using a mixed-methods approach, including focus group discussions, individual interviews, and structural equation modeling, we analyze three categories of flood management practices: Endogenous Knowledge-Based Practices (EKPs), Community Engagement-Based Practices (CEPs), and Agricultural Technology-Based Practices (ATPs). The results show significant contributions from CEPs to resilience and highlight the role of social cohesion and collective action. EKPs also have a positive impact, reflecting the importance of local knowledge, especially in Benin. However, the adoption of ATPs varies, with greater effectiveness observed in Togo than in Benin. Factors such as age, gender, education, and access to advisory services influence the acceptability and effectiveness of these practices. The findings highlight the need for tailored, integrative interventions that combine traditional knowledge and community participation with modern technologies to strengthen resilience in flood-prone communities. This study provides actionable insights for policymakers and development practitioners who aim to improve disaster risk reduction and climate resilience strategies in the Volta Basin.

1. Introduction

Flooding has emerged as a serious and increasingly serious natural hazard in the Volta Basin, which includes regions in Ghana, Benin, and Togo. This phenomenon is primarily caused by the confluence of climate change, rapid urbanization, and inadequate land-use practices. The Volta Basin’s vulnerability to extreme weather events, such as heavy rainfall and coastal storms, poses a significant threat to both rural and urban households, often leading to devastating consequences, such as loss of life, damage to infrastructure, displacement, and long-term economic disruption [1,2]. The Intergovernmental Panel on Climate Change (IPCC) has highlighted that extreme weather events are becoming more frequent and intense, largely due to climate change, exacerbating existing vulnerabilities in regions such as the Volta Basin [2]. Furthermore, urbanization is associated with increased flood risk due to unplanned expansion into flood-prone areas, exacerbating challenges for communities in coping with flood hazards [3].Effective flood management practices are essential for improving household resilience, which is defined as the ability of individuals, families, and communities to withstand, adapt to, and recover from shocks such as floods [4,5,6]. Resilience is a multidimensional concept that includes economic, social, and environmental dimensions and enables households to reduce their vulnerability to floods and recover more quickly after disasters [5,6]. Innovative and inclusive flood management strategies, such as early warning systems, improved drainage infrastructure, sustainable land-use practices, and digital flood forecasting tools have been identified as critical components for strengthening household resilience [7]. These practices not only help mitigate the immediate impacts of floods but also strengthen long-term adaptive capacity by minimizing economic losses, sustaining livelihoods, and protecting essential assets [5,7]. However, the implementation and effectiveness of these flood management practices vary significantly across the Volta Basin, influenced by differences in socioeconomic status, education, access to extension services, and financial resources [8,9]. Consequently, the impact of flood management practices on household resilience is neither uniform nor consistent. While certain households may benefit from timely interventions and a supportive policy environment, others remain vulnerable due to barriers such as limited access to information, inadequate infrastructure and cultural resistance to change [9]. This disparity highlights the need for targeted approaches that take into account the unique vulnerabilities and capacities of various social groups in the Volta Basin [8]. This paper examines how three types of flood management practices—Endogenous Knowledge-Based Practices (EKPs), Community Engagement-Based Practices (CEPs), and Agricultural Technology-Based Practices (ATPs)—influence household resilience in the Volta Basin. Particular attention is given to how these practices strengthen or constrain the adaptive capacity of different social groups in Benin and Togo. It also examines the influence of factors such as gender, education, age and access to extension services on the successful adoption of flood management practices, and their subsequent impact on resilience outcomes are examined [6,9]. Several recent studies have examined community flood resilience, particularly in contexts of socio-economic vulnerability (e.g., [4,5,6,7,8,9]). However, many of these works have highlighted persistent implementation challenges, limited community uptake of formal strategies, and the exclusion of traditional knowledge systems. These gaps justify our evidence-based and participatory approach in the Volta Basin. This study aims to contribute to the growing literature on climate resilience and disaster risk reduction and provide actionable insights for policymakers and development practitioners committed to improving flood resilience in communities across the Volta Basin. The results of this study are critical not only for understanding the effective implementation of flood management practices in the Volta Basin, but also for designing interventions that promote equitable access to these technologies. It highlights the importance of integrative, community-focused approaches to disaster risk management that address the varied vulnerabilities and capacities of different households across the region. Integrating local knowledge and practices into flood management strategies is critical to promote resilience and ensure interventions are tailored to the specific needs of communities facing the challenges of flooding. While this study is primarily exploratory and inductive, it operates under the assumption that each of the three categories of flood management practices positively contributes to household resilience, though in potentially different ways depending on social and geographic context. The study is unique because it uses a Structural Equation Modeling (SEM) framework to incorporate local knowledge and identify flood management strategies from the bottom up, based on the focus group discussions and interviews. Furthermore, the cross-country comparison between Benin and Togo highlights the varying pathways through which households build resilience to flooding in the Volta Basin.

2. Methodology

2.1. Sample and Research Context

This study focuses on the Volta Basin, specifically the district of Boukoumbé in Benin, and the districts of Oti Sud and Dakpen in Togo (Figure 1). These locations share similarities in their vulnerability to flooding but also have distinct characteristics shaped by their geographical, socio-economic, and environmental contexts. In this region, natural disasters, particularly floods and droughts, potentially affect approximately 30,000 people annually in the basin, cause an estimated USD 25 million in economic losses in built-up areas each year, and impact around 48,000 hectares of agricultural land and 35,000 hectares of pasture [10]. Under projected climatic and socio-economic scenarios, the vulnerability of people, assets, services, and ecosystems to flood and drought risks in the basin is expected to increase [10].

Located in the Atacora department of northern Benin, Boukoumbe is known for its hilly terrain and proximity to the Atacora Mountains. The district’s high elevation creates diverse microclimates, but deforestation and soil erosion in the highlands increase surface runoff during heavy rainfall and increase the risk of flooding in low-lying areas. The district’s economy is predominantly agrarian, with an emphasis on subsistence agriculture, which also includes crops such as millet, sorghum, and yam. Additionally, the area is culturally significant as the center of the Betamariba people, known for their unique mud tower houses called Takienta. However, reliance on traditional agricultural practices contributes to unsustainable land-use, reducing the country’s resilience to flooding. Boukoumbe faces challenges due to limited infrastructure development. Many communities lack access to adequate drainage systems, making them more vulnerable to seasonal flooding. In addition, poor road conditions make it difficult to provide timely disaster relief and assistance during extreme events.

Oti Sud is located in northern Togo in the basin of the Oti River, an important tributary of the Volta River. Its flat terrain and proximity to major watercourses make it particularly vulnerable to seasonal flooding. Heavy rainfall often leads to river flooding and affects settlements and agricultural areas along the flood plain. The district’s economy is heavily dependent on agriculture, with crops such as corn, rice, and peanuts being the main source of food for most households. Fishing along the Oti River is also important. However, floods impact these livelihoods by damaging crops, displacing households, and depleting the resources needed for reconstruction. Oti Sud faces significant infrastructure deficits, including inadequate drainage systems and limited access to early warning systems. Many riverside communities are particularly at risk of displacement during floods, and recovery efforts are often hampered by poor connectivity and resource constraints.

Both Boukoumbe (Benin) and Oti Sud (Togo) face similar challenges, including environmental degradation, limited adaptability, and vulnerable populations. Local authorities and NGOs in both countries have initiated community-based disaster risk reduction efforts, but funding and coordination gaps limit the impact of these efforts. While traditional practices remain central to flood risk management, they are poorly integrated into modern approaches, reducing overall resilience. By focusing on these districts, this study provides a regional understanding of how flood management practices shape flood risk and resilience in the Volta Basin.

2.2. Research Design

Based on our conceptual framework, we formulated the following general hypothesis to guide the quantitative analysis: the adoption of Endogenous Knowledge-Based Practices (EKPs), Community Engagement-Based Practices (CEPs), and Agricultural Technology-Based Practices (ATPs) positively influences household resilience to floods in the Volta Basin. This relationship may vary depending on the household’s sociodemographic characteristics and national context (Benin or Togo). To verify this hypothesis, we conducted this research in two phases: identifying the flood management practices used in the Volta Basin and assessing their impact on household resilience.

2.2.1. Identification of Flood Management Practices Used in the Volta Basin

The first phase was to identify the flood management practices used by the population. We used focus group discussions and unstructured individual interviews with key informants to collect data. These included local institutions and users of the Volta Basin in each study area, including professional agricultural organizations, local government representatives, elders, farmers, pastoralists, natural resource co-management actors, and technicians of structures involved in flood management, such as the Red Cross, Météo-Bénin, Direction Générale de l’Eau (DGEau), etc.

To facilitate consideration of gender, a focus group was organized with each stakeholder category: women, youth, and men. On average, fifteen participants were involved in each focus group discussion. The unstructured individual interviews in Benin included 68 households represented by household heads, 30% of whom were women. The aim of the unstructured individual interviews was to confirm the various flood management practices identified during the focus groups. To validate these practices, a local workshop was held in each country to introduce the practices to the community. This workshop contributed to refining the identified practices and group them into three categories: Endogenous Knowledge-Based Practices (EKPs), Community Engagement-Based Practices (CEPs), and Agricultural Technology-Based Practices (ATPs).

2.2.2. Assessment of the Impact of Flood Management Practices on Household Resilience

The aim of this phase was to assess which flood management practices households use and how these practices contribute to their resilience. Each type of flood management practice was used as an item to which the respondent indicated agreement on a 5-point Likert scale, ranging from ‘strongly disagree’ to ‘strongly agree’ (

Table 1. Latent constructs and their observed indicators.

Latent/Endogenous Variables	Countries	Indicators *
Endogenous Knowledge-Based Practices (EKPs)	Benin	▪ Observing the direction of the cattle’s eyes toward the sky (eastward) for a few weeks ▪ Appearance of migratory birds, locusts, and certain reptile species ▪ Early loss of foliage, early or late flowering of certain plant species in the dry season.
	Togo	▪ Observing the movements of magnan ants from watercourses or wetlands to higher altitudes ▪ The migration of wild birds, such as wild ducks, to dwellings, accompanied by their cries, or the frequent appearance of the gendarme weaver
Community Engagement-Based Practices (CEPs)	Benin	▪ Setting up a community flood and drought risk management committee ▪ Setting up a WhatsApp group to exchange information on disaster risks ▪ Setting up an environmental club in secondary schools ▪ Rationalization of food reserves by the community to cope with natural disasters
	Togo	▪ Farmers’ associations for mutual support and assistance in the event of a disaster; ▪ Mechanisms for disseminating warnings through village chiefs and village development committees ▪ Your active participation in community simulation exercises
Agricultural Technology-Based Practices (ATPs)	Benin	▪ Adoption or development of land fertility management practices ▪ Adoption of agroecological practices (intercropping, ecosystem conservation) ▪ Adoption of innovative water management practices
	Togo	▪ Adoption or development of agroecological practices ▪ Adoption of extra-early varieties cultivated just after floods ▪ The capture of surface runoff water on mountain slopes
Household resilience (adapted from [6])	All the countries	▪ I can replace my house quickly when it is affected by floods ▪ I am confident that my house will not be submerged by the highest floods ▪ I am confident that my house will not collapse or be swept away by the highest floods ▪ I am confident that my household has enough food to eat during the flood season ▪ I am confident that my household will not need to borrow food or money from informal sources during the flood season ▪ I am confident that my household can find a safe place to evacuate to if there is an extreme flood event in the future ▪ I am confident that children and elderly people are safe during extreme floods ▪ I am confident that the health of my family members will not be negatively affected by floods ▪ I want to learn new farming practices to cope with floods ▪ I have used new farming practices to cope with floods

* Each indicator was measured using a five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree).

). While the same categories of practices were identified in both countries, there were some differences in the indicator/item for their implementation.

We adopted the household resilience for flood index, as presented in Table 1 [6]. This index includes three properties of household resilience to floods: households’ confidence in securing food, income, health, and evacuation during floods and recovery after floods; households’ confidence in securing their homes not being affected by a large flood event; and households’ interests in learning and practicing new flood-based farming practices [6].

The assessment of the impact of flood management practices on resilience involved 312 households represented by the head of household in the Volta Basin, i.e., 156 per country. The sample size was determined based on recommended thresholds for structural equation modeling, which suggest a minimum of 10 observations per estimated parameter [11]. This number was also adjusted to ensure adequate representation across gender, age groups, education levels, and access to advisory services in both countries. The respondent profile, including gender (male/female), age, formal education (none, primary, secondary and post-secondary/university), and access to flood management advisory services (yes/no), was also collected and used as exogenous or independent variables in the evaluation model. In the first sub-section of the results, we provide a description of the household profile involved in the study.

2.3. Data Analysis

We assessed the relationships among the household profile, flood management practices, and household resilience to floods using Structural Equation Modeling (SEM) in Jamovi 2.3.28. The SEM approach was chosen for its ability to test complex causal relationships between latent constructs, which are not directly observable but measured through multiple indicators [11]. SEM integrates factor analysis and multiple regression, allowing for simultaneous analysis of measurement and structural models [11,12]. This fits well with our conceptual framework, where EKPs, CEPs, and ATPs are modeled as latent constructs influencing household resilience. All latent constructs—EKPs, CEPs, ATPs, and household resilience—are specified as endogenous variables in the model, given that they are influenced by exogenous observed variables, such as gender, age, education, and access to advisory services.

The estimation method used is Diagonally Weighted Least Squares (DWLS), as this method effectively deals with the problem of non-normal data and accepts ordinal variables [13]. DWLS is particularly well-suited for ordinal data, such as Likert-scale responses, as it does not rely on the assumption of multivariate normality required by traditional Maximum Likelihood Estimation (MLE). This makes it especially appropriate for social science research involving non-normally distributed or categorical data [13]. The DWLS method provides more accurate parameter estimates and standard errors for ordinal and non-normal data, and it is robust for small to moderate sample sizes [13,14]. Moreover, DWLS improves the reliability of model fit indices in the presence of skewed distributions or categorical variables [15].

However, DWLS also presents some limitations. It can be computationally intensive for large models with many variables and is generally less efficient than MLE when applied to continuous, normally distributed data [14,15]. Despite these limitations, DWLS was considered the most appropriate estimation method for this study, given the ordinal nature of the measurement items and the cross-cultural context, in which the assumption of multivariate normality is often violated.

The method was implemented in Jamovi software version 2.3.28 using the SEMLj module. First, we estimated an overall SEM at the Volta Basin scale. Then, we used country as a multigroup factor to understand the effect of context on the structural relationships between variables.

Before SEM estimation, we assessed the reliability of all latent variables using the Cronbach’s α and ensured that their value was high than 0.75, indicating strong internal reliability [16]. Then, we estimated the robustness of the estimated model using the Comparative Fit Index (CFI), the Goodness-of-Fit Index (GFI), the Adjusted Goodness-of-Fit Index (Adj. GFI), the Standardized Root Mean Residual (SRMR), and the Root Mean Square Error of Approximation (RMSEA). These fit indices were selected because they provide a complementary and robust evaluation of model fit, particularly in the context of ordinal data and moderate sample sizes. The Comparative Fit Index (CFI) compares hypothesized models to a baseline model in which all variables are assumed to be uncorrelated. CFI is less sensitive to sample size than other indices, and values close to 0.95 or higher are considered indicators of excellent model fit [11,15]. The Goodness-of-Fit Index (GFI) and its adjusted version (AGFI) measure the proportion of variance accounted for by the estimated population covariance. GFI and AGFI values above 0.90 generally suggest a good fit, although they can be sensitive to model complexity and sample size [15,17]. The Root Mean Square Error of Approximation (RMSEA) evaluates the discrepancy between the observed and estimated model per degree of freedom. RMSEA values below 0.08 indicate reasonable fit, and values below 0.05 are indicative of close fit. However, RMSEA tends to be inflated in small samples [13,17]. The Standardized Root Mean Square Residual (SRMR) measures the standardized difference between observed and predicted correlations. It is useful for assessing absolute model fit, with values below 0.08 indicating good fit [14]. As recommended by [11,17], using multiple fit indices lowers the risk of depending solely on one indicator, which could be impacted by sample size, model complexity, or data characteristics, and enables a more nuanced interpretation of model performance.

3. Results

3.1. Characteristics of Respondents

Table 2. Descriptives statistics of respondents.

Household Characteristics	Modalities	Benin		Togo		Study Area
		N	%	N	%	N	%
Sex	Female	41	26	40	26	81	26
	Male	115	74	116	74	231	74
	Total	156	100	156	100	312	100
Formal education	None	91	58	68	44	159	51
	Primary	27	17	66	42	93	30
	Secondary	34	22	18	12	52	17
	University	4	3	4	3	8	3
	Total	156	100	156	100	312	100
Advisory services	No	57	37	91	58	148	47
	Yes	99	63	65	42	164	53
	Total	156	100	156	100	312	100
Age	Young	57	37	57	37	114	37
	Adult	66	42	70	45	136	44
	Old	33	21	29	19	62	20
	Total	156	100	156	100	312	100

presents the characteristics of the study participants. Male respondents dominated in both countries, constituting 74% of the study population. This distribution was uniform across both Benin and Togo. Female respondents represented 26% of the sample, indicating potential gender disparities in representation. A significant proportion of respondents in both countries lacked formal education (51% overall), with Benin showing a higher percentage (58%) than Togo (44%). Primary education was more prevalent in Togo (42%) compared to Benin (17%). Respondents in Benin were more likely to have received advisory services (63%) compared to those in Togo (42%). Overall, 53% of respondents had access to advisory services. The age distribution was fairly similar between the two countries, with adults (44%) being the majority group, followed by young (37%) and old (20%).

The data highlight gender disparity and limited access to formal education, particularly among women. The higher percentage of respondents who received advisory services in Benin suggest better outreach or access mechanisms compared to Togo.

3.2. Measurement Model

The measurement model is an important part of Structural Equation Modeling (SEM), which examines the relationship between latent variables and their measures. In this research, we used Cronbach’s α to evaluate the reliability of the measurement model (

Table 3. Assessment of measurement reliability.

Country	Endogenous Variables	Mean	SD	Cronbach’s α
Benin	EKPs	3.26	1	0.92
	CEPs	3.33	0.94	0.88
	ATPs	4.14	0.51	0.86
	Resilience	2.93	0.58	0.89
Togo	EKPs	2.88	1.26	0.91
	CEPs	2.98	1.15	0.94
	ATPs	2.96	1.21	0.89
	Resilience	2.87	0.85	0.90

). The Cronbach values are higher than the 0.8 threshold, indicating a high level of reliability for each variable (Table 3).

A good fit was demonstrated across various indices.

Table 4. Fit indices.

Model	CFI	GFI	Adj. GFI	SRMR	RMSEA
Volta Basin scale (Overall SEM)	0.915	0.999	0.993	0.048	0.08
Multigroup factor analysis	0.906	0.998	0.982	0.05	0.09

presents the fit indices of the two models (Volta Basin scale and multigroup factor analysis). The CFI, GFI, and Adj.GFI values are close to 1. This indicates that the user and baseline models demonstrate an excellent fit. Both the RMSEA and SRMR values meet the recommended thresholds. The R-squared values elucidate the variance in the dependent variables accounted for by the independent variables in the model.

3.3. Structural Relationships

This section illustrates the results of the structural equation model that shows how the three categories of flood management practices contribute to household resilience in the Volta Basin. The model also illustrates how household characteristics determine the flood management practices adopted. Figure 2 provides a visual summary of the significant structural relationships identified in the SEM. The arrows represent causal paths, and their thickness indicates the strength of the standardized coefficients. It highlights that Community Engagement-Based Practices (CEPs) have the strongest positive impact on household resilience across the Volta Basin, highlighting the centrality of collective action in enhancing adaptive capacity. Agricultural Technology-Based Practices (ATPs) also demonstrate a substantial positive influence, particularly in Togo, while Endogenous Knowledge-Based Practices (EKPs) exhibit a moderate yet significant effect in both countries.

Figure 2 further illustrates how socio-demographic variables, such as education, gender, age, and access to advisory services, affect the likelihood of adopting each practice. For example, access to advisory services strongly influences the adoption of CEPs and ATPs, while education negatively correlates with EKPs. These nuanced relationships highlight the need to adapt interventions to specific household profiles. The overall results of the SEM are presented in

Table 5. Parameter estimates for the overall model.

Dep	Pred	Volta Basin		Benin		Togo
		Estimate	p	Estimate	p	Estimate	p
EKPs	Sex1 (Male)	−0.01	0.971	0.02	0.93	−0.06	0.783
EKPs	Education1 (Primary)	−0.26	0.144	−0.57	0.023	0.04	0.871
EKPs	Education2 (Secondary)	−0.10	0.647	−0.15	0.616	−0.38	0.353
EKPs	Education3 (University)	−0.98	0.036	−0.48	0.527	−1.36	0.022
EKPs	Advisory1 (Yes)	0.48	<0.001	0.38	0.019	0.24	0.368
EKPs	Age	0.02	<0.001	0.01	0.023	0.03	0.001
CEPs	Sex1 (Male)	−0.39	0.002	−0.40	0.021	−0.41	0.026
CEPs	Education1 (Primary)	0.13	0.373	−0.01	0.961	0.18	0.384
CEPs	Education2 (Secondary)	0.38	0.03	0.38	0.109	0.20	0.528
CEPs	Education3 (University)	0.16	0.654	0.18	0.687	−0.08	0.875
CEPs	Advisory1 (Yes)	0.82	<0.001	1.36	<0.001	0.24	0.25
CEPs	Age	−3.81 × 10−4	0.929	0.00	0.568	0.00	0.531
ATPs	Sex1 (Male)	−0.25	0.087	0.15	0.132	−0.43	0.047
ATPs	Education1 (Primary)	−0.26	0.131	−0.21	0.071	0.12	0.596
ATPs	Education2 (Secondary)	0.18	0.373	−0.07	0.701	−0.07	0.839
ATPs	Education3 (University)	−0.58	0.184	0.32	0.184	−0.69	0.153
ATPs	Advisory1 (Yes)	0.66	<0.001	−0.08	0.315	0.45	0.062
ATPs	Age	0.01	0.1	0.00	0.755	0.02	0.039
Resilience	EKPs	1.28	0.001	1.38	0.011	1.16	0.042
Resilience	CEPs	3.73	<0.001	2.42	<0.001	5.20	<0.001
Resilience	ATPs	1.44	0.002	−1.50	0.194	3.34	<0.001

, showing the structural relationship between the predictors and dependents variables, respectively, in Volta Basin, Benin, and Togo. We provide a thorough explanation of the structural relationships between the factors under evaluation in the following subsections.

3.3.1. Factors Affecting Flood Management Practices in the Volta Basin

At the Volta Basin scale, including Benin and Togo, age, formal education, access to flood risk management advice, and gender determine the types of management practices developed by households.

Age shows a positive influence on Endogenous Knowledge-Based Practices (EKPs) (0.020, p < 0.001), suggesting that older individuals develop more EKPs than young. The positive influence of age on EKPs is the same in both countries. Over time, older people have accumulated a wealth of endogenous knowledge about flood risk management. This helps explain why age is positively associated with EKPs in the Volta Basin. In Togo, age is positively associated with Agricultural Technology-Based Practices (0.020, p = 0.039), indicating that older populations in Togo may engage more in ATPs compared to younger.

Formal education affects flood management practices developed by the household, particularly EKPs and CEPs. Higher education negatively affects EKPs (−0.98, p = 0.036) at the level of the Basin, suggesting that those with advanced education rely less on traditional or local knowledge in developing their flood risk management. Similar results occurred in all countries; people with a primary school education (−0.57, p = 0.023) and a university degree (−1.36, p = 0.022) in Benin and Togo, respectively, rely less on EKPs compared to people without formal education.

Formal education, mainly secondary education, positively impacts CEPs (0.38, p = 0.03), suggesting that moderate formal education encourages community-oriented practices. People with formal education are receptive to collective actions, which explains the positive link between formal education and CEPs for flood management in the Volta Basin.

Advisory services on flood management emerged as a critical driver across all practices. In the Volta Basin, advisory services show significant positive impacts across EKPs (0.48, p < 0.001), CEPs (0.82, p < 0.001), and ATPs (0.66, p < 0.001). These results indicate that access to advisory services encourages widespread adoption of all practices. In the context of our study, advisory services are provided by NGOs (for example, NGO Alpha-Omega Environnement (AOE) in Benin). The positive influence of advisory services on flood management practices, mainly on EKPs (0.38, p = 0.019) and CEPs (1.36, p < 0.001), is perceptible in Benin, but is not significant in Togo, while the effects are also positive. Our findings show that the Volta Basin demonstrates the most robust response to advisory services, while Benin relies heavily on them for CEPs.

In the Volta Basin, there is a negative association between gender and CEPs (−0.39, p = 0.002), indicating that male-headed households are less likely to rely on CEPs. Similar negative effects of gender were observed on CEPs in both countries and on ATPs in Togo. This result is noteworthy and indicates that women have confidence in community initiatives and agricultural technologies to reduce the risks of flooding. This highlights gender-based disparities in participation in community and technical practices for flood management in the Volta Basin.

In summary, advisory services are the strongest drivers across all practices of flood management in the Volta Basin. In Benin, advisory services are crucial for EKPs and CEPs, but ATP engagement is limited. Higher education is strongly negatively associated with EKPs. In Togo, gender disparities are pronounced, with male-headed households less involved in CEPs and ATPs, while advisory services have weaker effects compared to other regions, especially for CEPs.

3.3.2. Impacts of Endogenous Knowledge-Based Practices (EKPs) on the Resilience of Households

EKPs value traditional and local knowledge in dealing with floods and reflect a deep-rooted cultural and ecological understanding. At the Volta Basin scale, EKPs have a moderate positive impact on household resilience (1.28, p = 0.001). This suggests that using local knowledge contributes significantly to preparation and coping strategies. In the Volta Basin, EKPs are often rooted in collective memory and long-standing practices adapted to the region’s unique ecological conditions. In Benin, EKPs have a similarly positive but slightly stronger effect on resilience (1.38, p = 0.011). This reflects the strong role of traditional knowledge in Benin’s rural communities, where access to modern resources may be limited, making indigenous flood management strategies essential. In Togo, the effects of EKPs are positive but relatively weaker (1.16, p = 0.042) compared to the Volta Basin and Benin. This could indicate increasing reliance on modern technologies or gaps in the transfer of endogenous knowledge.

Overall, EKPs remain a cornerstone of resilience across regions, but their strength varies, likely due to differences in cultural adherence to traditional practices and exposure to modern interventions.

3.3.3. Impacts of Community-Based Practices (CEPs) on the Resilience of Households

CEPs emphasize collective action and social structures to improve flood preparedness and recovery. At the Volta Basin scale, CEPs have the strongest impact on household resilience (3.733, p < 0.001), suggesting that community cohesion and participation are critical in coping with floods. Shared responsibility enables effective communication, pooling of resources, and mutual help in times of crisis. The same results are recorded for both countries. CEPs emerged to be the most effective practice in all regions of Volta Basin, highlighting the importance of promoting strong social networks and collective decision-making in flood-prone areas.

3.3.4. Impacts of Agricultural Technology-Based Practices (ATPs) on the Resilience of Households

ATPs focus on modern agricultural innovations, including flood-resistant seeds, land and water management practices, etc. to mitigate the impact of floods on livelihoods.

At the Volta Basin scale, ATPs have a significant positive impact on household resilience (1.44, p = 0.002), reflecting the role of adaptive technologies in protecting agricultural productivity. Since agriculture is an important livelihood in the Volta Basin, access to flood-resistant crops, irrigation systems, and other technologies can significantly improve household resilience. Unfortunately, the impact of ATPs on household resilience in Benin is not significant (−1.50, p = 0.194), indicating limited adoption or effectiveness of agricultural technologies. This may be due to financial constraints, lack of technical support, or lack of awareness among rural households. In Togo, ATPs have a substantial positive effect on household resilience (3.34, p < 0.001). This indicates a strong integration of agricultural innovations in flood management, which could be attributed to better dissemination of the technology or targeted interventions in Togo.

4. Discussion

4.1. Factors Affecting Flood Management Practices in the Volta Basin

This study explored the key factors influencing flood management practices in the Volta Basin, with particular emphasis on Endogenous Knowledge-Based Practices (EKPs), Community Engagement-Based Practices (CEPs), and Agricultural Technology-Based Practices (ATPs). The results show that demographic variables, such as age, formal education, extension services, and gender, play a crucial role in household-level decision-making on flood management. These findings inform the design of context-specific and integrative flood risk management strategies in Benin, Togo, and the wider Volta Basin region, as they highlight the need for tailored approaches that take into account local realities and knowledge systems.

4.1.1. Age and Endogenous Knowledge-Based Practices (EKPs)

The relationship between age and EKPs shows a positive correlation, suggesting that older people are more inclined to use endogenous knowledge when dealing with flood risks. This trend is particularly evident in both Benin and Togo, where traditional knowledge built through experience, is highly valued. In Togo, this age-related trend extends to ATPs, suggesting that older people may also be more likely to adopt agricultural technologies into their agricultural practices as complementary flood risk mitigation strategies. Such findings emphasize the importance of integrating traditional knowledge into formal flood management frameworks, thereby promoting intergenerational knowledge transfer that can improve community resilience [18,19].

Furthermore, the preservation of traditional ecological knowledge is essential for effective flood risk management. Traditional ecological knowledge can significantly influence flood risk management strategies and provide insights that are often overlooked in modern approaches [18]. This integration not only respects the cultural heritage of communities but also increases the effectiveness of flood management practices by leveraging local expertise and adaptation strategies refined over generations.

4.1.2. Formal Education and Flood Management Practices

The impact of formal education on flood management practices paints a nuanced picture. At the basin scale, higher levels of formal education are associated with lower reliance on EKPs, as those with advanced education tend to favor scientific and technical knowledge over traditional practices. This trend is consistent in both Benin and Togo, where formal education shows significant negative correlations with EKPs. This discrepancy between formal education and indigenous knowledge highlights the need for educational initiatives that promote the integration of traditional practices into formal flood management strategies [19,20].

Conversely, formal education, through secondary education, appears to promote greater community engagement, as individuals with intermediate educational backgrounds are more likely to participate in collective actions related to flood management. This positive relationship between secondary education and CEPs suggests that educational programs should not only promote community participation but also aim to harmonize traditional and formal knowledge systems. Such an approach can improve community-oriented flood management practices and ultimately lead to more resilient communities [19].

4.1.3. Advisory Services as Key Drivers of Flood Management Practices

Advisory services have emerged as a critical driver influencing EKPs, CEPs, and ATPs across the Volta Basin. These services, often provided by non-governmental organizations, play a vital role in disseminating knowledge and promoting the adoption of effective flood management practices among communities. In Benin, the influence of advisory services on EKPs and CEPs is particularly pronounced, indicating their essential role in supporting both endogenous and community-based practices. However, in Togo, while advisory services continue to have a positive impact, their influence on CEPs is less significant, highlighting regional disparities in service delivery [19,20].

This disparity in the effectiveness of advisory services highlights the need for targeted interventions to enhance service delivery in regions where it is lacking. By ensuring equitable access to advisory services, communities can be better equipped to adopt effective flood management practices, thereby improving their overall resilience to flood risks. Furthermore, the integration of local knowledge with advisory services can create a more robust framework for flood risk management, as suggested by the work of Hoang et al., who emphasize the importance of combining traditional and modern approaches in flood risk management [21].

4.1.4. Gender Disparities in Flood Management

Gender differences significantly impact flood management practices, with male-headed households generally exhibiting lower levels of trust in CEPs and ATPs, indicating women’s reliance on collective action and agricultural technologies to reduce flood risk. Promoting gender equality in flood management is not only a matter of social justice but also a strategic imperative for strengthening community resilience. Studies indicate that diverse perspectives contribute to more effective decision-making and problem-solving in the context of disaster risk management [19,20]. Encouraging an integrative approach that values the contributions of all community members, flood management strategies can be more comprehensive and effective, ultimately leading to better outcomes for vulnerable populations.

Gender differences in flood management practices reveal a complex environment in which, despite numerous challenges, women demonstrate strong confidence in community initiatives and agricultural technologies as effective means of mitigating flood risks. Research shows that women play a critical role in community engagement and decision-making processes related to flood management, often using their unique perspectives and experiences to strengthen resilience within their communities [22,23]. This empowerment is crucial, as it not only promotes women’s sense of agency but also contributes to the overall effectiveness of flood risk management strategies.

Women in flood-prone areas often strongly believe that community initiatives can effectively address the challenges of flooding. Their involvement in local government and community-based organizations allows them to advocate for sustainable practices and technologies that can reduce flood risk. For example, studies have shown that women actively participate in the development and implementation of flood preparedness plans and use their knowledge of local conditions and social networks to strengthen community resilience [22]. This commitment is critical, as it ensures that flood management strategies are not only inclusive but also tailored to the specific needs and vulnerabilities of the community.

Additionally, women’s trust in agricultural technologies as a means to combat flooding is increasingly recognized. Many women farmers are adopting innovative agricultural practices and technologies that improve their adaptability to flood events. For example, the use of drought-resistant crop varieties and improved irrigation techniques have been shown to empower women in agriculture and enable them to maintain productivity even in adverse weather conditions [24]. This adoption of technology not only mitigates the immediate impacts of flooding, but also contributes to the long-term food security and economic stability of their families and communities.

The psychological aspect of women’s trust in these initiatives cannot be overlooked. Research suggests that increased awareness and understanding of flood risks leads to greater support for mitigation efforts among women [25]. This informed risk perception empowers women to take proactive actions in their communities, fostering a culture of preparedness and resilience. By participating in training programs and workshops focused on flood management and agricultural technologies, women are better equipped to make informed decisions that improve the resilience of their households and communities to floods [22,23].

However, it is important to recognize that while women demonstrate confidence and skills in these areas, structural barriers persist and may limit their full participation. Societal norms and gender roles often limit women’s access to resources, information, and decision-making platforms [23]. Addressing these barriers through targeted interventions, such as capacity-building programs and inclusive policy frameworks, is critical for maximizing women’s potential in flood management. Empowering women and amplifying their voices can lead to more inclusive and effective community flood risk management strategies.

In conclusion, the factors affecting flood management practices in the Volta Basin are complex and interrelated. Age, formal education, extension services, and gender all play a crucial role in household-level decision-making regarding flood risks. By understanding these dynamics, stakeholders can develop more effective, inclusive, and context-specific flood risk management strategies that build on local knowledge and promote community engagement. This holistic approach is essential for building resilience in the face of increasing flood risks exacerbated by climate change and socioeconomic developments [19,20,26].

4.2. Impacts of Flood Management Practices on Household Resilience in the Volta Basin

The results of this study highlight the important role that various flood management practices play in improving household resilience in the Volta Basin. In particular, the impacts of Endogenous Knowledge-Based Practices (EKPs), Community Engagement-Based Practices (CEPs), and Agricultural Technology-Based Practices (ATPs) illustrate how these strategies contribute to the adaptive capacity of households exposed to flood risks. By examining these practices, we can better understand their impact on community resilience and the overall effectiveness of flood management strategies in the region.

4.2.1. Endogenous Knowledge-Based Practices (EKPs)

The moderate positive impact of EKPs on household resilience (1.28, p = 0.001) highlights the importance of traditional and local knowledge in flood management. This practice reflects a deep-rooted cultural and ecological understanding that has been passed down through generations. In the context of the Volta Basin, EKPs are often rooted in collective memory and long-standing practices adapted to the unique ecological conditions of the region. In Benin, the slightly stronger effect of EKPs on resilience (1.38, p = 0.011) suggests that traditional knowledge is particularly crucial in rural communities where access to modern resources may be limited. This reliance on indigenous flood management strategies is critical for improving household resilience, as it enables communities to effectively prepare for and manage flood events [27,28].

In Togo, EKPs still have a positive impact on resilience (1.16, p = 0.042), but their effect is relatively weaker than in Benin and the entire Volta Basin. This may indicate an increasing dependence on modern technologies or gaps in the transfer of traditional knowledge. The varying strength of EKPs across the region highlights the need for targeted interventions that promote the preservation and integration of traditional practices into formal flood management systems. By valuing and incorporating local knowledge, stakeholders can increase the effectiveness of flood management strategies and improve household resilience [27]. These findings are consistent with findings by [18] in Uganda and [19] in Nepal, who noted that local knowledge plays a vital role in early detection and low-cost adaptation strategies among flood-prone communities.

4.2.2. Community Engagement-Based Practices (CEPs)

The results show that CEPs have the strongest impact on household resilience across the Volta Basin (3.733, p < 0.001). This highlights the crucial role of community cohesion and participation in flood management. The ability to share responsibilities, pool resources, and support each other in times of crisis strengthens the adaptability of households. The strong positive correlation between CEPs and resilience suggests that promoting social networks and collective decision-making is essential for effective flood management [29]. In both Benin and Togo, the effectiveness of CEPs highlights the importance of promoting strong social structures that facilitate communication and collaboration among community members. Involving local communities in flood preparedness and recovery efforts not only strengthens community bonds but also increases overall household resilience. This collective approach to flood management is particularly relevant in the context of climate change, where the frequency and intensity of floods is expected to increase. By prioritizing community engagement, stakeholders can develop more robust flood management strategies that enable households to respond effectively to flood risks [8,29,30,31].

Our findings align with the work of [20], who emphasized the importance of social networks and participatory structures in enhancing community-level flood resilience in the United States. Similar conclusions were drawn in Malawi, where strong community cohesion was found to be a key predictor of effective flood preparedness and response [30]. This underscores the relevance of CEP approaches across diverse geographic contexts.

4.2.3. Agricultural Technology-Based Practices (ATPs)

ATPs also play an important role in improving household resilience, with a positive impact of 1.44 (p = 0.002) on the Volta Basin scale. These practices focus on modern agricultural innovations, such as flood-resistant seeds and improved land and water management techniques, which are critical for mitigating the impact of floods on livelihoods in the context of climate change [32]. Since agriculture is a primary source of income for many households in the Volta Basin, access to adaptive technologies can significantly improve resilience [33,34].

However, the results suggest a disparity in the effectiveness of ATPs between the two countries. In Benin, the impact of ATPs on household resilience is not significant (−1.50, p = 0.194), suggesting limited adoption or effectiveness of agricultural technologies. This may be due to financial constraints, lack of technical support or inadequate awareness of available technologies among rural households [32]. In contrast, Togo shows a significant positive effect of ATPs on resilience (3.34, p < 0.001), indicating greater integration of agricultural innovations into flood management. This difference may be due to better technology diffusion or targeted interventions in Togo, highlighting the importance of removing barriers to technology adoption in Benin [33,34].

The strong effect of ATPs on resilience in Togo is consistent with findings by [21] in the Mekong Delta and [32] in Northern Benin, which showed that access to adaptive technologies, including early-maturing seeds and water management innovations, significantly reduced flood vulnerability in agricultural communities. However, the weaker impact in Benin also echoes results from [33], who highlight structural barriers, such as uneven extension services and financial access, which hinder technology adoption in vulnerable rural areas.

Overall, while all three categories of flood management practices—EKPs, CEPs, and ATPs—contribute to household resilience, their relative influence varies by context and mechanism of action. CEPs consistently demonstrate the strongest effect on resilience, underscoring the critical role of social capital, collective action, and shared learning in enabling adaptive responses to floods. EKPs also play a substantial role, particularly in areas where access to formal services is limited, by drawing on locally embedded ecological knowledge. The impact of ATPs appears more context-dependent, with higher effectiveness in regions benefiting from structured extension services and technology dissemination. These findings highlight the need for integrated approaches that leverage the complementarities between traditional knowledge, community engagement, and technological innovation. Promoting synergies among these practices can enhance the adaptive capacities of rural households and ensure more robust and inclusive flood resilience strategies across diverse socio-ecological settings in the Volta Basin.

5. Conclusions

This study highlights the strategic importance of integrating Endogenous Knowledge Practices (EKPs), Community-Based Engagement Practices (CEPs), and Agricultural Technology-Based Practices (ATPs) to enhance the flood resilience of households in the Volta Basin. Drawing on evidence from Benin and Togo, the findings underscore that locally rooted knowledge and community mechanisms are essential pillars of effective climate adaptation strategies.

The strong influence of CEPs on resilience suggests that policies should actively support community organization, information-sharing networks, and grassroots environmental initiatives. EKPs, grounded in generations of observation and experience, also play a vital role and should be institutionalized and integrated into early warning systems, risk communication, and local planning processes. Meanwhile, the differentiated effects of ATPs across contexts call for targeted and locally adapted promotion of climate-smart agricultural technologies that align with local capacities and socio-cultural conditions.

For policymakers and development partners, this study suggests that resilience-building efforts will be more effective when strategies are participatory, inclusive, and co-designed with the communities they aim to serve. National adaptation plans and disaster risk reduction frameworks should, therefore, institutionalize local knowledge and community practices alongside technical innovations. Such integrative approaches will enhance not only the relevance but also the sustainability and ownership of resilience interventions. We acknowledge that some measurement uncertainty may arise due to the differences in item phrasing and interpretation between the two countries. To improve cross-context comparability, future research should focus on harmonizing indicators.