1 Introduction

"Adaptation for survival" and "survival of the fittest" are expressions found in everyday conversations about the continued existence strategies for humans to hold on to life. This is not only applicable to humans but all living organisms; the more reason adaptation is a concept that holds the same meaning across multiple disciplines. In socio-environmental studies, adaptation is defined as a process or outcome of a process that warrants minimising the hazard or its associated risks, or achieving the advantages associated with climate variability and climate change [1]. It is also defined as a process whereby strategies for moderating, coping with and taking advantage of the negative effects of climate-related occurrences are enhanced, developed, and implemented [2]. Adaption therefore emanates in the advent of threats and damages, and in quest of the management strategies needed by any living organism to build and foster resilience for survival. Most importantly, there are concerns for human population due to their vulnerability to various disasters with respect to the health challenges and livelihood disruptions being faced. A vast body of literature [3,4,5,6,7,8,9,10] have addressed these concerns considering a wide range of disasters across the world. While some focused on natural disasters, others are on man-made disasters with relevant and necessary information on planning, adaptation and recovery measures required to deal with the disasters.

Drawing from history, a number of disasters are public health-related and being characterised as endemics, epidemics or pandemics which have severely affected cities and the livelihoods of people. In 2019, there was the emergence of COVID-19 with severe health effects on people across the world to an extent that it was declared a pandemic. Like other pandemics in the past, compliance with public health and social measures is key in preventing the spread and control of the pandemic. Hence, measures such as lockdown, isolation and physical distancing are put in place around the world in order to curb the spread of the COVID-19 to save lives [11,12,13]. However, such restrictions on movement have serious negative implications for economic growth and social development including death, job losses, reduction of people’s income, and disruption in social interactions and congregations [14,15,16,17,18,19,20,21,22,23,24,25].

Community resilience is one way that the literature [26,27,28,29,30] addresses the challenges and consequences of the pandemic. For instance, a recent study [29] analysed literature on cities and the COVID-19 pandemic, and highlighted six thematic areas as follows: (1) air quality changes; (2) impacts of weather and meteorological factors on the infection risk; (3) monitoring the transmission of the virus in urban water systems; (4) green recovery in cities; (5) contributions of smart city solutions; and (6) Built environment factors and transportation. These thematic areas provide valuable insights into the various aspects of urban planning, design, and management in the context of the pandemic with a focus on adaptation, resilience, and sustainable livelihood. More so, with an emphasis on community-level reactions and resources, another study [30] offered insights into the various critical institutional, social, economic, infrastructure, and demographic factors contributing to resilience of communities to the COVID-19 pandemic. These factors cover a broad spectrum and are critical to a community's capacity to contain, mitigate, and recover from the effects of the pandemic.

Community-level assessment is, however, a summative measure that is limited in its capacity to understand the different factors contributing to resilience at the basic unit of analysis. Researching COVID-19 pandemic resilience at the household level is preferable because households are the basic units where individuals make daily decisions regarding health, resources, and compliance with preventive measures. Focusing on households allows for a nuanced analysis of the pandemic impacts, adaptation and resilience which can better inform targeted interventions at community level. Further, the COVID-19 pandemic affected households of different social status with different level of impacts. It should be noted that different households, prior to the pandemic, had different ways or a particular behaviour towards addressing certain situations affecting their livelihoods. With the added burdens due to the impacts of the COVID-19 pandemic, the question of how households were resilient, that is, holding on to life during the lockdown and carrying on afterwards, is of great concern.

The negative impacts of COVID-19 caused people of different works of life to change their behaviour, strategies and methods of managing their daily activities to cope with the pandemic. This assertion is established by a number of studies [31,32,33,34,35,36,37,38,39]. Some of these studies focused on people as employees in organisations to determine how they cope with COVID-19 pandemic, as such, addressing their adaptation for business. Some were likewise household-based and have focused on adaptation of families to COVID-19 pandemic at their homes. From the studies that have considered households, little is known about how the different categories of households based on their socioeconomic characteristics adapt to COVID-19 pandemic. Hence, this study assesses households’ adaptation to COVID-19 pandemic across different residential densities in Abeokuta, Nigeria. Although, density in terms of concentration of population in different areas has been established to impact the spread of COVID-19 by influencing the ease of person-to-person transmission [40,41,42,43,44]. The possibility of intimate contact between people is higher in crowded settings, like cities, which facilitates the ability to spread the virus. High-density housing, crowded areas, and public transit can all lead to higher transmission rates. In these kinds of situations, it becomes imperative to put into practice efficient public health measures, such as social distancing and wearing of mask to slow down the transmission of the virus. However, from the perspective of this study, density is used to reflect social status across spatial strata. This is because different residential densities do reflect the socioeconomic characteristics of households in urban areas with functional land use zoning system, therefore providing a socio-spatial context for this study.

The research questions for this study are: (i) what are the socioeconomic characteristics of households in different residential densities in Abeokuta, Nigeria? (ii) What is the impact of COVID-19 pandemic on households in different residential densities? (iii) What are the adaptive strategies used by households during COVID-19 in different residential densities? And (iv) how do the impacts of COVID-19 pandemic on households and their adaptation strategies vary across different residential densities? The results of this study would be useful to better understand the variation of COVID-19 impact across different social classes in the society and how resilience can be built. The reason being that the society consists of not a single social class while all kinds of shocks and stresses like the pandemic and its consequences influence each class differently depending on its capacity and available resources. Furthermore, the emergence of COVID -19 and its declaration as a pandemic is not the first public health disaster and might not be the last. This is because various forces such as climate change and human encroachment into natural wildlife habitats may increase the frequency of pandemics in the future, hence, better knowledge of the underlying patterns and dynamics of pandemics, their effects on cities, and necessary preparations, responses, and adaptation measures are needed [45]. This study therefore provides individuals, stakeholders, and the general populace with information not only on how the current pandemic has affected and still affecting all and sundry but also shedding more light on probable solutions for adaptation at household level during the pandemic and its imminent new waves.

2 Materials and methods

This study was conducted in Abeokuta using the quantitative research design. Households in Abeokuta constituted the study population that were surveyed across low, medium and high residential planning densities in the city.

2.1 The study area

Abeokuta is the Capital City of Ogun State in Southwest Nigeria and is geographically located on latitudes 7° 10' N to 7° 15' N and longitudes 3° 17'E to 3°26' E (Fig. 1). It lies on the eastern bank of the Ogun River, and is 77 km (48 miles) north of Lagos by rail and 130 km (81 miles) by waterway [46]. Abeokuta was founded in 1830, near rock outcrops, therefore surrounded by the Olumo Rock and several caves and shrines. The region that the city lies has a tropical climate with distinct dry and wet seasons, which last about 130 days; the average annual precipitation and temperature are about 1,270 mm and 28 °C respectively, and the estimated average annual potential evaporation is 1,100 mm [47, 48]. The natural drainages for the city are Ogun River and Oyan River, and its water supply [49]. Abeokuta and its surrounding areas had an estimated population in 2021 is 544,000 which is an increase of 2.06% from 2020 (https://www.macrotrends.net/cities/21975/abeokuta/population). This study was conducted in Abeokuta because the capital city experienced stringent social and economic restrictions associated with the total lockdown during the pandemic, especially due to its proximity to Lagos State with the highest COVID-19 cases, in which the first cases was discovered in Nigeria.

Fig. 1
figure 1

Map of Abeokuta within the contexts of Ogun State and Nigeria

Full size image

2.2 The research design

This study utilised a quantitative research design based on a cross-sectional survey for the phases of data collection, analysis and interpretation. The study was conducted between November, 2020 and August, 2021. Within this period, we embarked on a preliminary observation of the physical, social, economic and environmental conditions of the study area. This aided the identification of localities and their delineation into residential density zones patterned around their socioeconomic attributes.

2.3 Study population and sampling

The households in Abeokuta constituted the study population as they dwell in different zones delineated based on residential planning densities of low, medium and high. While our initial plan was to conduct a household survey, we were unable to access household data. As a result, we employed an alternative method, which is commonly used in urban planning literature [50,51,52,53,54], involving the use of city and building population statistics to select a representative sample of households. In the determination of sample size, we utilised the projected population of Abeokuta for 2020 which was 533, 000 from the base population of 434,000 in the 2006 census. Using the recommended population proportion of 50%, a minimum sample size of 384 respondents was calculated to be sufficient for a confidence level of 95% that the true population is within ± 5% of the sample size. In adjusting for missing questionnaires and other unforeseen circumstances, 10% of the computed sample size was added, thus making a total of 422 respondents determined as sample size for this study. A two-stage sampling technique was employed in this study. The first stage involved the use of stratified random sampling method. The study area comprising different localities was stratified into homogenous group based on residential densities (high, medium and low). There is a total of 45 existing localities from which 14, 17 and 14 were identified as high, medium and low density area respectively. Afterwards, there was a selection of localities (high -5, medium – 4 and low – 3) from the identified density areas using simple random sampling technique. Furthermore, there was a selection of streets (high – 15, medium – 11 and low – 8) from the identified density areas using simple random sampling technique. The second stage involved the selection of buildings from the selected streets using systematic sampling technique. After selecting the first building using simple random sampling, one out of every five buildings were selected on each street which amounted to a total number of 422 buildings. A household from each building was surveyed which resulted into the selection of 422 respondents from the three residential density areas (high – 180, medium – 137 and low – 105) as sample size for this study.

2.4 Data collection

Data were collected through administration of questionnaire on households affected by the COVID-19 pandemic across different residential densities in Abeokuta, Ogun State. The data collection was between April and June, 2021. The field surveys were embarked on using questionnaire as the research instrument. The questionnaire addressed issues on socioeconomic characteristics of households, the impact of COVID-19 pandemic on them, and their adaptation strategies to the pandemic. The socioeconomic characteristics include gender, age, occupation and income while the questions on impact and adaptation strategies were developed as agreement scales using 3-point Likert responses (1 –Disagree; 2– Just Agree; and 3 –Agree). These are the COVID-19 Impact Agreement Scale (CIAS) and COVID-19 Adaptation Agreement Scale (CAAS). The questionnaire was validated through expert judgment as the questions were vetted by senior researchers. A total of 422 (high – 180, medium – 137 and low – 105) copies of the questionnaire were administered during the survey while 403 (high – 173, medium – 132 and low – 98) were retrieved from respondents.

2.5 Ethical considerations

In the course of data collection, personal and sensitive information as well as those for re-identification of individuals were not obtained with the questionnaire. Each questionnaire contained in the consent form and the purpose of the research and other guidelines were stated in it. Each participant was asked to carefully read and sign the form to indicate their understanding of the research and interest to participate in the research. Participants were assured of non-disclosure of research data but only for academic and research purposes.

2.6 Data analysis and interpretation

During data collation, 12 questionnaires were excluded from analysis due to the nature of incompleteness and unreliability of the information provided. Therefore, a total of 391 (high—167, medium – 128 and low – 96) copies of the questionnaire were considered during data analysis. Descriptive and inferential statistics were employed at the data analysis phase using the Statistical Package for Social Sciences (SPSS 20) and Microsoft Excel (MS-Excel 13). The descriptive statistics comprised the cross-tabulation analysis which was performed on socioeconomic, flood impact and adaptation attributes across the residential zones. The inferential statistics involved the testing of the following hypotheses:

  • Ho1: There is no statistically significant difference in impact of COVID-19 pandemic on households across residential zones in Abeokuta.

  • Ho2: There is no statistically significant difference in households’ adaptive strategies during COVID-19 pandemic across residential zones in Abeokuta.

The data collected on COVID-19 impacts and adaptation were computed as composite scores and means through the Likert responses that were weighted according to their designated values (1-Disagree, 2—Just Agree, 3—Agree). This conversion is necessary in transforming ordinal data to ratio data for composite measurement and can be adaptable for parametric test for a normally distributed data [55,56,57]. The composite mean outcomes of the transformation were later used for the hypothesis testing. The Analysis of Variance (ANOVA) alongside its complementary tests which are the Levene and Welch’s statistics were used for the hypothesis testing.

The results of the descriptive statistics for this study are presented in percentages and for better insights and understanding into the discussion, the interpretations are provided. The results for socioeconomic characteristics are interpreted as: minority (< 50.0%), average (50.0%), majority (> 50). Those of impact and adaptive strategies are interpreted as follows: very low (≤ 20.0%), low (≥ 20.1% to ≤ 40.0%), medium = Just Agree (≥ 40.1% to ≤ 60.0%), high (≥ 60.1% to ≤ 80.0%), and very high (≥ 80.1% to ≤ 100.0%). For the inferential statistics, the data used for analysis are on 3-point Likert responses and the weight allotted to each response to standardise it from ordinal to ratio scale has the interpretation of 1 = Disagree (≤ 1.49), 2 = Just Agree (≥ 2.50 to ≤ 3.49), and 3 = Agree (≥ 3.50 to ≤ 4.49). Furthermore, the ANOVA results are interpreted as a measure of equality of means. The Levene results are interpreted as a measure of equality of variances, while the Welch results are interpreted when there is unequal variances.

3 Results

The results are on socioeconomic characteristics of households, impact of COVID-19 pandemic on households across residential zones, and households’ adaptive strategies during COVID-19 pandemic across residential zones. The results are visualised and presented in bar charts. The hypothesis testing results are on COVID-19 impact on households and their adaptation strategies across the different residential densities.

3.1 Socioeconomic characteristics of the respondents

Every household has its peculiar socioeconomic attributes that are defined by gender, age and occupation, amongst others. The attributes were considered due to their importance in identifying the status of respondents and how such status is defined by the areas they live. The results of socioeconomic characteristics of the respondents are hereunder presented (Figs. 2–7). Figure 2 shows the results of gender distribution. In the high residential density zone, male respondents comprised 54.7% and female respondents constituted 45.3%..In the medium residential density zone, male respondents accounted for 52.3% and female respondents consisted of 47.7%. In the low residential density zone, male respondents accounted for 53.8% and female respondents constituted 46.2%.

Fig. 2
figure 2

Gender of respondents across residential density areas

Full size image

The distribution of ages of the respondents (Fig. 3) in the high residential density zone revealed that the young adults constituted 46.7%, old adults accounted for 29.3%, the youths constituted 13.3%, and the elderly respondents consisted of 10.7%. In the medium residential density zone, the young adults constituted 52.8%, old adults comprised 32.1%, the youths and elderly also accounted for 7.5%, apiece. In the low residential density zone, having young adults constituted 43.6%, old adults accounted for 30.8%, youths also accounted for 20.5% and elderly respondents constituted 5.1%.

Fig. 3
figure 3

Age of respondents across residential density areas

Full size image

Distributions across the residential zones are shown for marital status of respondents (Fig. 4). In the high residential density zone, the married accounted for 54.7%, single accounted for 25.3%, separated also accounted for 13.3% and divorced respondents constituted 6.7%. In the medium residential density zone, the married comprised 75.5%, single accounted for 17.0%, the separated and divorced also consisted of 5.7% and 1.9%, respectively. In the low residential density zone, having married accounted for 71.8%, single accounted for 20.5%, separated also constituted 5.1%, and divorced respondents comprised 2.6%.

Fig. 4
figure 4

Marital status of respondents across residential density areas

Full size image

The respondents also have their distributions regarding educational status presented across the residential density zones (Fig. 5). In the high residential density zone, respondents with primary education comprised 55.8%, tertiary education accounted for 20.3%, secondary education also consisted of 13.2%, and respondents with no formal education accounted for 10.7%. In the medium residential density zone, those with primary education accounted for 22.8%, and secondary education constituted 32.6%. Those with tertiary education and no formal education accounted for 34.2%and 10.4%, respectively. In the low residential density zone, those having primary education accounted for 29.0%, secondary education consisted of 37.3%, tertiary education accounted for 21.6% and no formal education respondents constituted 12.1%.

Fig. 5
figure 5

Educational status of respondents across residential density areas

Full size image

Occupation of respondents across the residential zones is shown in Fig. 6. In the high residential density zone, the civil servants accounted for 38.5%, private employers constituted 28.2%, business persons accounted for 25.6%, and professional practitioners consisted of 7.7%. In the medium residential density zone, civil servants accounted for 24.5%, private employers constituted 26.4%. The business persons and professional practitioners comprised 30.2% and 18.9%, respectively. In the low residential density zone, civil servants consisted of 21.3%, businesses persons accounted for 28.0%, private employers accounted for 18.7% and professional practitioners accounted for 5.1%.

Fig. 6
figure 6

Occupation of respondents across residential density areas

Full size image

The difference in income distributions across the residential zones is presented in Fig. 7. In the high residential density zone, the low income earners accounted for 56.3%, lower middle income earners accounted for 24.7%, upper middle income earners accounted for 12.3%, and high income earners respondents comprised 6.7%. In the medium residential density zone, the low income earners accounted for 15.5%, lower middle income earners accounted for 24.7%, as well as the upper middle and high income earners also consisted of 32.6% and 25.1%, respectively. In the low residential density zone, low income earners accounted for 9.5%, lower middle income earners accounted for 11.6%, upper middle income earners also accounted for 27.3%, and high income earners consisted of 51.6%.

Fig. 7
figure 7

Income of respondents across residential density areas

Full size image

3.2 Impact of the COVID-19 pandemic on respondents

The impact of COVID-19 pandemic on respondents are shown across the residential zones (Fig. 8). The impact is of two categories which are social and economic. The social impact include variables such as unable to join social and religious gatherings, impairment of social relationships inside their homes and inability to enjoy outdoor activities in their neighborhood. In the high residential density zone (Fig. 8a), 90.8% stated they were unable to attend social and religious gatherings. Respondents accounting for 73.3% were unable to enjoy outdoor activities in their neighborhood. Similarly, 73.3% claimed that the pandemic affected social relationships inside their homes. Respondents that accounted for 49.3% stated an increase in drug abuse while an increase in domestic violence was accounted for by 45.4% respondents. Lastly, 14.7% respondents experienced the loss of a family member or friend. In the medium residential density zone (Fig. 8b), 86.8% respondents were unable to enjoy outdoor activities in their neighborhood. Similarly, 79.0% claimed that the pandemic had affected social relationships inside their homes, and 77.0% stated they were unable to attend social and religious gatherings. Reports on an increase in domestic violence were made by 65.8% respondents while 51.0% reported an increase in drug abuse and 9.5% experienced the loss of a family member or friend. In the low density zone (Fig. 8c), 79.5% respondents could not engage in outdoor recreation within their neighbourhood, 77.4% apiece were unable to attend social and religious functions, and stated that social relation within their households was disrupted by the pandemic. Those that reported cases of increase in drug abuse were 64.1%, while those who experienced increased domestic violence were 48.0%. Respondents that lost their family member or friend during the pandemic were 14.6%.

Fig. 8
figure 8figure 8

a Impact of the COVID-19 Pandemic on Respondents in High Density Area. b Impact of the COVID-19 Pandemic on Respondents in Medium Density Area. c Impact of the COVID-19 Pandemic on Respondents in Low Density Area

Full size image

The economic impact of COVID-19 pandemic was also assessed using six variables which are: loss of major occupation, no finance to access healthcare services, cannot pay utility bills, loss of other income sources, increased hunger, and loss of shelter. In the high residential density zone, 87.0% respondents agreed on issues regarding inability to foot their utility bills, 85.2% reported inability to access healthcare services due to no funds, and 83.4% experienced the loss of their main occupation. Similarly, 74.6% respondents reported cases of increased hunger, and 61.4% stated they lost other income sources. Those who became homeless due to loss of shelter were over 51.3%. In the medium residential density zone, the respondents who could not pay utility bills accounted for 81.0% and those that could not pay to access healthcare services were 73.6%. Respondents constituting 66.1% reported about loss of other income sources, 64.7% experienced loss of major occupation, and 60.4% experienced increased hunger. Those who lost their shelter accounted for 39.7% of respondents. In the low density zone, 20.6% experienced loss of other income sources, and 18.0% reported increased hunger. In addition, 10.0% respondents experienced loss of major occupation, 8.7% stated they could not afford utility bills, 7.6% had no finance to access healthcare services, and 6.4% experienced loss of shelter.

3.3 Adaptive strategies employed by respondents during COVID pandemic

The adaptive strategies employed by respondents during COVID-19 pandemic are shown across the residential zones (Table 1). These strategies also fall into two categories which are social and economic. The social adaptive strategy contained six variables, namely: engaged in spiritual exercises(such as praying, spiritual bathing and enchantment), engaged in online communication and social interactions(calling and chatting on WhatApps, Facebook, Twitter, Instagram, TikTok, etc.), received help from family and friend, avoided listening to COVID-19 news and updates, applied COVID-19 personal safety measures(use of personal protective equipment(PPE) like nose, mask, hand gloves, face-shield, overall, etc.), and applied COVID-19 distance safety measures (social distancing). In the high density zone (Fig. 9a), 87.7% respondents avoided listening to COVID-19 news and updates, 81.0% engaged in online communication and social interactions, and 80.5% applied COVID-19 distance safety measures. More so, 74.7% engaged in spiritual exercises, and 73.3% received help from family and friend. Those accounting for 68.0% applied COVID-19 personal safety measures. In the medium density zone (Fig. 9b), 76.2% respondents applied COVID-19 personal safety measures, 74.6% applied COVID-19 distance safety measures, 73.2% received help from family and friend, and 70.8% engaged in spiritual exercises. Additionally, 64.7% avoided listening to COVID-19 news and updates and 57.9% engaged in online communication and social interactions. In the low density zone (Fig. 9c), 92.0% applied COVID-19 personal safety measures. Further, 87.2% respondents applied COVID-19 distance safety measures, 84.8% engaged in online communication and social interactions, and 83.4% avoided listening to COVID-19 news and updates. Those that engaged in spiritual exercises accounted for 70.2% while 6.40% received help from family and friend.

Table 1 Residential density areas in Abeokuta
Full size table
Fig. 9
figure 9figure 9

a Adaptive Strategies Employed by Respondents during COVID-19 Pandemic in High Density Area. b Adaptive Strategies Employed by Respondents during COVID-19 Pandemic in Medium Density Area. c Adaptive Strategies Employed by Respondents during COVID-19 Pandemic in Low Density Area

Full size image

The economic adaptive strategy likewise contained seven variables which are: lived off savings, received online loans from digital money lenders, made incomes through online jobs, received loans from commercial banks, reduced household’s consumption, sold properties and assets, and received government palliatives. In the high density zone, 89.3% respondents made incomes through online jobs and reduced household’s consumption and 84.3% reduced their household’s consumption. Moreover, 78.9% received online loans from digital money lenders, 70.8% lived off savings, and 70.6% sold properties and assets. Those that received government palliatives are 61.3% and those that received loans from commercial banks are 54.0%. In the medium density zone, 64.0% respondents lived off savings, 62.5% received loans from commercial banks, and 62.3% sold properties and assets. Those that received government palliatives are 59.3%, 58.4% received online loans from digital money lenders, 58.0% reduced their household’s consumption, and (57.1%) made incomes through online jobs. In the low density zone, 46.2% respondents received loans from commercial banks. Furthermore, 9.6% received online loans from digital money lenders, 8.3% received government palliatives, 6.2% sold properties and assets, and 5.2% lived off savings. The respondents with 5.0% are those who made incomes through online jobs, and 4.3% are those who reduced their household’s consumption.

3.4 Research Hypothesis Testing

The two hypotheses formulated for this study were tested using the Analysis of Variance (ANOVA) with complementary test as the Levene and Welch’s statistics. The level of significance for all the statistics used is 0.05 (∝ = 0.05). The results are presented in (Tables 2, 3, 4 and 5). To determine the variation in impact of COVID-19 pandemic across the high, medium and low residential zones, the Levene’s test was first conducted to check if there is equality of variance across the groups. The results [LT (2, 388) = 59.068, p = 0.000] indicated that the variances are unequal (Table 3). The ANOVA results [F (2, 388) = 30.406, p = 0.000] then showed that the impact of COVID-19 differed significantly across the residential zones (Table 4). For reason of considerable differences in variances due to unequal number of respondents across residential zones (unequal group sizes), the Welch’s test was conducted. The results [FW (2, 223.810) = 25.088, p = 0.000] of the Welch’s test also indicated a statistically significant difference in COVID-19 impact across the residential zones (Table 5). Households were most impacted in the high residential zone (M = 2.38, SD = 0.819), followed by the medium residential zone (M = 2.06, SD = 0.498) and less impacted in the low residential zone (M = 1.67, SD = 0.763) (Table 2). This implies the rejection of the null hypothesis that there is no statistically significant difference in impact of COVID-19 pandemic on households across residential zones in Abeokuta. Therefore, an alternative hypothesis that a statistically significant difference exist in impact of COVID-19 pandemic on households across residential zones in Abeokuta is accepted.

Table 2 Descriptive statistic summary
Full size table
Table 3 Test of homogeneity of variances
Full size table
Table 4 ANOVA Summary
Full size table
Table 5 Robust tests of equality of means
Full size table

Further, the results [LT (2, 388) = 44.255, p = 0.000] of Levene’s test also indicated that unequal variances for households’ adaptation to COVID-19 pandemic across the high, medium and low residential zones (Table 3). The ANOVA results [F (2, 388) = 39.539, p = 0.000] showed that the COVID-19 adaptive strategies employed by households varied significantly across the high, medium and low residential zones (Table 4). Similarly, due to unequal number of respondents across residential zones (unequal group sizes) leading to considerable differences in variances, the Welch’s test was conducted. The results [FW (2, 242.413) = 41.497, p = 0.000] of the Welch’s test also indicated a statistically significant difference in COVID-19 adaptive strategies across the residential zones (Table 5). Households adapted most in the low residential zone (M = 2.54, SD = 0.679), followed by the medium residential zone (M = 1.97, SD = 0.639) and adapted less in the high residential zone (M = 1.66, SD = 0.910) (Table 2). This implies the rejection of the null hypothesis that there is no statistically significant difference in households’ adaptation strategies during COVID-19 pandemic across residential zones in Abeokuta. Hence, the acceptance of an alternative hypothesis that a statistically significant difference exist in households’ adaptive strategies during COVID-19 pandemic across residential zones in Abeokuta.

4 Discussion

The impact of the COVID-19 pandemic is observable in several urban settlements around the world. In this study, in determining the nature and level of impact, analysing the characteristics of the different households is very important in understanding the specific socioeconomic groups being affected and how to address the needs of each group. The findings on socioeconomic characteristics of participants dwelling in high, medium, and low residential density zones of Abeokuta, Nigeria show interesting insights. First and foremost, all zones showed a relatively balanced gender distribution, with the high-density zone exhibiting a majority of males. In addition, the educational status indicated a shift towards the majority having primary education in the high-density zone and minority having secondary education in the low-density zone. Occupational diversity was observed across all zones, with civil servants, private employers, and business persons being the majority. Lastly, the income distribution showed the majority of high-income earners in the low-density zone, while low-income earners were the majority in the high-density zone. Overall, these findings on socioeconomic characteristics showed that the distribution of respondents descriptively varied across the high, medium, and low residential density zones, therefore providing a socio-spatial context for the study.

The impact of the COVID-19 pandemic on selected households was further assessed using their social and economic aspects of life. Across the residential density zones, respondents experienced notable disruptions in social interactions and outdoor activities. However, those in the high-density zone indicated a very high impact in terms of being unable to attend social and religious gatherings and experiencing domestic violence. In comparison, respondents in low-density zone showed a relatively low impact on their social aspects of life. Economically, the respondents in high-density zone faced significant challenges, such as being unable to pay utility bills and access healthcare services, as well as experiencing a loss of main occupation. On the other hand, those in the low-density zone demonstrated greater resilience, with a lower proportion facing economic hardships. In terms of adaptive strategies, respondents across all zones turned to spiritual exercises, online communication, and social interactions to cope with the pandemic. Notably, those in the low-density zone exhibited a very high adoption of COVID-19 personal safety measures, while the high-density zone relied on online jobs and reduced household consumption as significant adaptive strategies. This demonstrates the resourcefulness and adaptability of households in different residential zones during the pandemic. Lastly, the inferential statistics results provide further evidence of significant variance in the impact of COVID-19 pandemic on households, on the one hand, and their adaptation strategies, on the other hand, across residential zones in Abeokuta. As a result, the level of impact in high, medium, and low residential density zones is high, medium, and low, respectively. Whereas, the level of adaptation in high, medium, and low residential density zones is low, medium, and high.

The findings on socioeconomic characteristics in this study support an array of literature [58,59,60,61] that has empirically found that low, medium, and high density areas are inhabited by the high-, medium-, and low-income earners. Thus validating the reciprocity of urban density and income structure in Abeokuta. More so, the results on the impact of pandemics on households align with previous research [14, 17, 20] which found that the COVID-19 pandemic disrupted social interactions and outdoor activities, leading to increased reliance on online communication. For instance, “the lives of the general public have been greatly disrupted and restricted in terms of their mobility, social interactions, and the performance of their daily activities. Many individuals have been suddenly “forced” to study or work from home and may experience changes in physical and social health” [17]. Furthermore, economic challenges were prevalent, affecting various income groups differently. For instance, it is stated that the lockdown affected people by putting setback on economic activities which in turn negatively affected the income of households, therefore making the poor becoming poorer because they have no backup to easily fall on [14]. A number of studies [52, 62,63,64] have also stated that any disruption of high impact with low adaptation on the side of victims suggests high vulnerability or low resilience, and vice-versa. This implies that as far as the vulnerability of households is high, their resilience is low. Hence, the findings of this study showcase issues of low resilience of households to COVID-19 pandemic in the high density zone, medium resilience in the medium density zone, and high resilience in the high density zone. Thus, linking high vulnerability and low resilience with social and economic poverty.

Considering the extensive body of literature examined in their work [29], it is crucial to note that the findings of this are related in two ways. First, both studies highlight the differential impacts of the pandemic on different population groups. As it showcased the disproportionate impacts on minorities and urban poor, emphasising the need for regenerative urban design measures [29], this study similarly explains the effects of the pandemic on households in different residential density zones, emphasizing significant interruptions to social contacts and financial difficulties in high-density zones. Second, both studies emphasised the importance of adaptive strategies to deal with the pandemic's effects. Their work [29] also emphasised the need for regenerative urban design measures to promote a just, resilient, and sustainable recovery from the pandemic while this study highlights the use of social connections and online communication as adaptive techniques, amongst others. Findings from another study [30] also relate to this study by highlighting the importance of considering the social and economic dimensions of life in assessing the impacts of pandemics on communities. It recognised that urban and rural communities may have different resilience priorities and that resilience-building strategies need to be tailored to address the specific challenges and strengths of each community [30]. This study similarly suggests that the effects of the pandemic vary across different residential density zones, with people in high-density zones experiencing more significant impacts on their social and economic well-being. The study highlights that households in different zones adopted different adaptive strategies, depending on their unique circumstances. However, this study is more specific to urban setting with no insights provided regarding the rural setting.

The comparison of the above findings with previous studies highlights the universal nature of the COVID-19 pandemic and its impact while also underscoring the influence of local socioeconomic contexts on households' coping and adaptation mechanisms. Meanwhile, the novelty of this study is that it contributes to the current understanding of how the COVID-19 pandemic affected households and how they adapted across residential density zones in Abeokuta, Nigeria. The findings underscore the importance of recognising socio-spatial differences in formulating effective responses and support mechanisms during health disasters.

5 Conclusion and implications

This study concludes that the social and economic impact of COVID-19 pandemic on households differ across residential zones in Abeokuta, as well as the adaptive strategies they employed. It is worth mentioning that significant variations in COVID-19 impact across residential zones can lead to resource disparities. As the world contends with future uncertainties, this study provides valuable insights into how individuals and families can hold on to life in the face of shocks and stressors. This entails fostering resilience and promoting equitable outcomes for the diverse urban households. The findings have both practical and theoretical implications. On the practical implications, it is deduced that (i) the high residential zone with higher pandemic impact might experience overwhelming strain on resources such as housing and healthcare systems, leading to resource shortages and reduced quality of life and healthcare, while other zones might have more manageable situations. (ii) The variation in impact could strain social cohesion, as households facing severe consequences may experience heightened stress and tension, while those less affected might struggle to empathize, potentially impacting collective efforts to combat the pandemic. (iii) The differences in impact can exacerbate economic inequality. The high residential zone that is heavily affected by the pandemic may experience more severe economic downturns due to lockdowns and business closures, affecting employment and income, and lack swift recovery. In contrast, other zones with milder impact may recover more swiftly. (iv) Lastly, the different residential zones could face distinct challenges in implementing effective adaptation strategies due to the difference and social and economic capacities, capital or assets. Against this background, the identification of specific socioeconomic characteristics and adaptive strategies provides valuable information for policymakers and stakeholders to design targeted interventions and support measures for households in the different residential density zones. It is recommended that post COVID-19 palliatives and mitigating measures in high residential zone are to be of priority concern and directed towards the poor. Prioritising the physical development of high residential zone in terms of housing and infrastructure for social and economic improvement is a good strategy for better adaptation and fostering resilience. For the theoretical implication, the study contributes to the existing body of knowledge on the impact of pandemics on households, establishing a socio-spatial context and highlighting the need for socioeconomic context-specific responses.

This study is without its limitations. First, the research was conducted in Abeokuta, Nigeria, which is a secondary city in the southwestern region of Nigeria. This reflects that selecting participants from this city alone had not fully represented households in this region, and the findings are not generalisable for the region. Additionally, the findings have also not reflected the experiences of households in other regions of Nigeria, similarly limited in generalisability. Nevertheless, it still provided information on how a part of the country experienced and responded to the pandemic and could be a template for conducting regional and national-based research. Secondly, the study is based on household perceptions, hence, the data relied on self-reported responses that are very subjective. Nonetheless, the subjectiveness is good to understand personal experiences for a bottom-up policy formulation. More so, the research methods and data were empirically validated to minimise response biases. Thirdly, the study utilised a cross-sectional design that limits the establishment of causality between variables. However, the design allowed for analysing multiple variables across different socioeconomic groups within the short period of COVID-19, thus allowing the researchers to get a grasp of the impact and adaptive strategies during the pandemic.

The limitations of this study give room for further studies. It is recommended that future studies could focus on regional analysis of COVID-19 pandemic impact and adaptation considering the relationships between socioeconomic and environmental variables. More so, the assessment of factors influencing adaptive strategies across different residential densities and regions may offer nuanced approaches to enhancing community resilience. Additionally, a mixed-method design could be adopted to gain the advantage of quantitative and qualitative research methods for better outcomes and insights. Finally, longitudinal studies could be engaged to analyse the long-term effects of the pandemic on households' socioeconomic status and well-being to provide valuable insights into post-pandemic recovery.