The Gambia is a small country occupying the two banks of The Gambia River which are North and South of the river ([12]). Considering the 2013 Population and Housing Census, The Gambia has a population of 1,882,450 with a density of 176 persons per km² ([13]). This research covered the whole country hence the focus of the assessment will be based on the impact on households and the coping strategies of communities along the River Gambia. The whole five regions; West Coast Region (WCR), North Bank Region (NBR), Lower River Region (LRR), Central River Region (CRR), and Upper River Region (URR)) and the two administrative regions (Kanifing (KMC) and Banjul (BJL) were covered in the assessment. In the FGD, the five regions were also covered leaving the other two administrative regions (KMC and BJL). The reason is that, the study is designed to know the local coping strategies of the people living in the rural and remote communities along the River Gambia. The selection of this research area is mainly motivated by settlements along the River Gambia which are much closer to the river.

As a means of better understanding the flood impacts and household coping strategies of local communities along the River Gambia, this research employed the convergent parallel mixed method of case study designed ([24]) to be able to access: the impacts of floods on household along the River Gambia. Figure 1, the convergent parallel mixed methods design helped in the simultaneous collection and analysis of both quantitative and qualitative data that provided a comprehensive understanding of the research problem. The results from each method were then compared and integrated during the interpretation phase to corroborate findings and draw holistic conclusions. The local coping strategies adopted by local households to mitigate, prepare, and respond to flooding. Our analysis used a mix of qualitative and quantitative data collected from both primary and secondary sources. This study was nationwide and adopted a mixed-methods design ([3]). The quantitative and qualitative data were gathered in the same phase of the research process during the data collection, analyzed independently, and interpreted together to produce results ([9]).

Figure 1. A framework of the study design showing the convergent parallel mixed method study approach.

We employed one of the purposeful samplings; the Criterion Sampling Method for FGD participants. 10 FGDs were conducted with 7 participants in each group. This was done in five regions of the Gambia, namely the West Coast Region, North Bark Region, Lower River Region, Central River Region, and Upper River Region ([13]). In each region, two FDGs were conducted, one for females and the other for males. The FDG was convinced that interviewing community members in a group discussion would be very helpful ([31]).

The participants were selected purposively in consultation with the “Alikalos” (village heads). The following criteria were used during the selection of participants: 1) 30 years of age and above, 2) living in the community for more than 3 years, 3) Having experience of flood events, 4) currently a permanent resident in the area. Questions were open-ended and were well structured to capture the flood’s local coping strategies and the impacts of the flood. The main questions asked were:

What are the local coping strategies community members use before and during floods?How did you acquire this knowledge you’ve mentioned? How effective is your local strategy toward the management of floods in your community? What is well-known to be destroyed during floods in your community? How serious is the damage in your community?

The design of this research was a community-based cross-sectional research approach ([43]). The impact assessment approach made use of socioeconomic, biophysical, and demographic variables. The research was conducted by the use of questionnaires for each selected household. The questionnaires were pretested for reliability and validity before the actual data collection was done ([21]; [16]). Household heads were interviewed with the help of a semi-structured interviewer-administered questionnaire by simple random sampling in each selected community ([1]). The data collection was determined by the factors, of both women and men, age 30 and above, household head, or a representative participation ([26]). Any person who has stayed in the research area for a minimum of 3 years was excluded from the research because he/she may have not experienced the last year’s rainy season. A sample of 384 was generated (n = pq (z²e)²), and it was further multiplied by 10% for sampling error to obtain 422 respondents. The final 422 respondents were divided into five regions according to the GBoS 2013 population census on the number of households for each Local Government Area. In each region, 3 communities along the River Gambia were randomly selected.

Qualitative data analysis: The interviews were Focus Group Discussions (FGD) and recorded using an audio recorder. Voice recording was done during discussions and field notes were kept for proper record writing. Afterward, the voice recordings were transcribed. The transcribed data were vigorously proofread to check the content of the writing to the voice records before the analysis. Later on, interview transcriptions and field notes from the observation process were read thoroughly to understand the issue from a holistic point of view. Direct quotes were used to reflect participants’ views more strikingly and to depict the phenomenon more clearly by unfolding it for the readers.

Quantitative Data Analysis: In the quantitative stage, participants’ answers obtained from the data collection tools (questionnaire) have been processed using the STATA 11. Both descriptive and inferential analyses were performed on the data collected during the survey. Microsoft Excel Office version 2016 software was used to manage the raw data and STATA 11 statistical analysis tools were used to analyze and do data interpretation. The obtained data were presented using descriptive statistics (percentages, frequencies, and pie charts). The relationship between respondents’ demographic variables (eg. region, gender, age, ethnic group, religion, marital status, education, occupation, household size, and the number of years stayed in the community) was examined using Pearson’s chi-square and Binary Logistic Regression. During the Binary Logistic Regression, we considered demographic variables (eg. region, gender, age, ethnic group, religion, marital status, education, occupation, household size, number of years stayed in the community) of respondents as the independent variable (coping ways or strategy) as the dependent variable.

A total of 422 households responded to this research. Among the respondents, 294 (69.7%) were males, while 128 (30.3%) were females. More than one-third 183 (43.8%) of the respondents are not educated and 88 (20.9%) are farmers. More than half 218 (51.1%) of the respondents have lived in their communities for more than 20 years. Regarding household size, 250 (59.2%) have a household size of 6 or above (see Table 1).

Household experience of flood events shows that 416 (98.6%) had experienced a flood in their community. 298 (70.6%) stated that their houses had experienced floods where water entered their buildings (Table 2).

The result shows that the local people have less information about the impending floods, out of the respondents 9 (2.1%) said that they have information about the impending floods. 373 (88.4%) of the respondents said that they did not know any evacuation centre during flood disasters (Table 3).

Table 1. Demography of household Heads of communities along the River Gambia.

Table 2. Household experience of flood events.

Table 3. Household Knowledge/Preparedness on the effects of flood.

The result of household knowledge or preparedness on the effects of the flood (Table 2) shows that the majority of the local people have knowledge of flood events but are not fully informed and prepared for early warning. This is in agreement with the research carried out by [30], which reveals that the local people do not have an adequate response to an early warning system to reduce the effect of flood disasters. In the same vein, a similar survey was carried out in Pakistan where the result suggested that there was an ineffective early flood warning system that could be a result of many factors ([34]). In addition, the result of this research has shown that the majority of the local people could not identify locally, a designated location or place for an evacuation centre during a flood disaster contrary to what was found in the research conducted in Indonesia ([31]; [36]).

Table4 shows Pearson Chi-square results for the impacts of the flood on households along the River Gambia. In the results, region, and gender were highly significantly associated with death impact as a result of flood. The region, gender, ethnic group, and occupation were also highly significantly associated with the impact of the flood on injury to household members. Region, gender, age, ethnic group, occupation, settlement type, household size, and number of years stayed in the community were all highly significantly associated with the impact of water entering houses.

Table 4. Pearson chi-square result for the impact of the flood on household.

Pearson chi-square result for households’ preparedness for flood risk indicated that region, age, ethnic group, marital status, educational level, occupation, settlement type, household size, and years in the community were all statistically significantly associated with the household preparedness toward flood events. This is similar to the research carried out by ([10]) where their Pearson chi-square shows age and household size have an influence on the preparedness of local people for the flood. The chi-square analysis test (Table 4) shows that region, gender, ethnic group, and occupation were the only demographic variables that have shown a statistically significant association with the question about death as an impact of flood in their communities. Furthermore, region, gender, age, ethnic group, occupation, settlement type, household size, and number of years stayed in the community were all highly significantly associated with the impact of water entering the house and its consequences. This suggested that local people know and are aware of the impact of floods in their communities. A similar research carried out by [32] indicated that the local people (80%) are aware of the flood impacts or consequences.

Table5 shows Pearson chi-square results of households’ preparedness to flood communities along the River Gambia. Region, age, ethnic group, religion, education level, settlement type, household size, and years in the community were all significantly associated with household access to electricity. Region, age, ethnic group, religion, educational level, occupation, settlement type, household size, and years in the community were all significantly associated with household members having health or asset insurance in case of flood disaster. Region, age, ethnic group, marital status, educational level, occupation, settlement type, household size, and years in the community were all highly significantly associated with household preparedness for flood events.

Table 5. Pearson chi-square result for households’ preparedness to flood risk.

This study also tried to find out whether the local people had been interacting with flood risk maps but the majority of the respondents said that they did not have flood maps. A similar research was carried out in Ethiopia and it was established that flood risk maps were of great help in local coping strategies in flood management ([11]).

Table6 shows the binary logistic result of how the local people responded to their flood local coping strategies. The result reveals that KMC (CL 9.250 - 3.739 and p = .000), NBR (CL −1.038 - 4.824 and p = .002) and CRR (CL 3.535 - .141 and p = .034), male gender (CL −.303 - 3.014 and p = .016), household head youth age 20 - 29 (CL 11.669 - 1.553 and p = .010), Mandinka ethnic group (CL 3.557 - .782 and p = .002), informal education (CL −.945 - 4.270 and p = .002), rural settlement type (CL .042 - 4.377 and p = .055), household size 5 below (CL 4.224 - .416 and p = .017), 6 - 10 (CL 4.545 - 1.519 and p = .000) and 11 - 20 (CL 6.120 - 2.764 and p = .000), and number of years stayed in the community 6 - 10 (CL 7.909 - .089 and p = .045 ) and 16 - 20 (CL 2.283 - .123 and p = .029) were all significantly associated with the household heads’ knowledge on their local coping strategies.

Table 6. The binary logistic result on how local communities’ members viewed their coping strategies was sufficient.

Knowledge of flood among the local communities during FGD

Participants show a local knowledge of flood management. All participants pointed out that both river and flash floods occurred mainly during the rainy season (June to October). They pointed out that whilst flooding can be induced by heavy rainfall, it frequently happens where there is less preparation and socioeconomic strength to curb the effect of flood.

“…yes rainfall is the main cause of flood in this community but our way of settlement and financial strength are not able to handle the effect of the flood as you can see…” (Participant 2 in Dasilameh—Male group).

It was said by the participants that water spread rapidly in their villages due to the intensity and duration of rainfall. The longer the rainfall, the faster the flood water and the more destruction it makes on farmlands and buildings. Based on this, women in LRR were able to identify a particular place in the village square, a high land which they viewed as a temporary evacuation centre during severe floods. Moreover, the men group in CRR-North showed a very high level of awareness of vulnerability and exposure to certain locations in their villages, they’ve reiterated that households located in swampy (lowland), mud houses, and are not raised higher above the ground are the ones mostly affected during flood events.

It was unanimously agreed in all the discussions that the greatly and frequently affected asset is their farmlands. However, roads are the second most destroyed as mentioned by the majority. Buildings came third hence some participants said that their houses were not seriously affected because they built them on high land instead of lowland.

The impact of the flood in The Gambia is greatly felt as much research has shown in another part of the world where floods in some cases claimed lives. [28] and [15] in their research noticed that the impact did not only stop at destroying roads and buildings but also on vulnerable lives whose resilience is very low. Run-offs, coverage areas, and impacts of the flood are on the increase as stated by the participants in FGD. Most of the informants have a view that the main cause of flood could be anthropogenic activities like blocking the drainage system, building houses or fences on the run-off, building flimsy house foundations, etc. Similar causes were also identified in research carried out in Nigeria ([2]). In addition, this research also reveals that floods are capable of destroying people’s assets like farmlands, buildings, roads, livestock, etc. This is also mentioned by [2], [5], and [5] in their research on the effect/impact of floods on the people.

Based on the FGD, participants had shown that flood has a great impact on their livelihood. Out of the total participant who said “yes” flood has damaged their assets, Figure 2 shows that farmlands (32%) and roads (31%) are the two most affected assets out of the four assets discussed. Out of the 70 participants in the five regions during the FGD, 69 participants agreed that their farmlands were destroyed by flood, 63 out of 70 agreed that their buildings were destroyed, 66 out of 70 also said that the roads were destroyed, 16 out of 70 said that their livestock were destroyed by flood water. “…We are always seeking help both from the government and philanthropists to prevent our farms from flooding because we are financially constrained.” (FGD Male group NBR). Local people even though they have some local strategies for coping with floods, all admitted that their strategies are not sufficient thus they needed help to be able to properly manage the adverse effects of floods in their communities. “…yes we have our local means of managing flood but we are very often overwhelmed by the strength the flood water enters our farmlands and houses, that is why we need help to be able to do our farming well, especially during the rainy season” (FGD, Female group LRR).

Figure 2. A bar chart showing the response of participants of FGD in the five regions on assets damaged.

This research revealed that households have their local coping strategies that are always employed. Local strategies like; sandbags to block or reduce the speed of flood water, household reinforced buildings or raised elevations of the house to obstruct the entry of flood waters into their houses, and constructing contour bonds and dikes to prevent flood were all identified during the research which was in consonant with the research carried out by [41] and [19]. Based on the result (Table 7), the local community members had much more concentration on the control of flood water and their assets than on the control of flood water sources ([18]).

Table 7. A table showing the local people’s flood coping strategies during FGD.