Introduction

Mountain regions are susceptible to global climatic change and their ecosystems predicted to be severely affected in the future (Bezák & Halada, 2010). Mountain ecosystems are key components in the supply of ecosystem goods and services to local communities (Grȇt-Regamey et al., 2013). These communities are therefore vulnerable to the effects of climate change due to dependence on mountain systems, poverty and limited access to resources (Gioli et al., 2019; Gupta et al., 2019). Mountain forests also ultimately experience anthropogenic pressures linked to provisioning services exacerbating their vulnerability to climate change (Egan & Price, 2017). Though mountains have been identified as hotspots for climate change, little is known about the progressive impacts this change will have on different ecosystems and subsequently the communities dependent on them.

Research has alluded to the occurrence of climate change in some mountain regions and the ensuing impacts (Maghsood et al., 2019; Poudel & Duex, 2017; Said et al., 2019). Though these impacts and their magnitudes may differ due to biophysical and socio-ecological differences, there are commonalities that have been identified. The phenomenon of climate change largely includes increased occurrence of extreme weather events and natural hazards, devastating local communities (IPCC, 2014; Spinoni et al., 2020). Droughts, fires, floods and storms are the most common climate hazards resulting from higher temperatures and increased rainfall variability (Ahmed et al., 2019; Alemayehu & Bewket, 2017). Droughts cause severe water stress which reduces water available for domestic use and agriculture causing substantial economic losses (Bryant, 2022; Haslinger et al., 2019). Floods and storms resulting from erratic high intensity rainfall lead to damage to property, structures and the outbreak of waterborne diseases (Alemayehu & Bewket, 2017). The increasing occurrence of extreme weather events globally is undoubtedly a signal of the onset of climate change impacts (IPCC, 2014). Studies on the impacts of climate variability in mountain regions have been mostly focused on the Himalayan Mountain ranges in Asia (Xu et al., 2009; Briner et al., 2012; Negi et al., 2012; Akbar, 2017; Tariyal, 2017). The Himalayas have been experiencing higher temperatures, which have increased the rate of glaciers melting and reduced agricultural productivity (Tariyal, 2017). Other mountain regions such as the South American Andes have experienced landslides, rockslides, avalanches and floods due to increasing occurrence of intense storms (Bedoya-Soto et al., 2018; Poveda, et al., 2020). In Africa, most mountain studies have focused on species responses to climate change and its impact on coffee production on Mt Kilimanjaro in Tanzania (Zech et al., 2011; Dulle et al., 2016; Molina-Venegas et al., 2020; Wagner et al., 2021).

Research indicates that southern Africa is warming up at a faster rate in comparison to the average global rates, with studies identifying a rapid increase in the minimum annual temperatures in South Africa (Mosase & Ahiablame, 2018). Rising temperatures indicate that the impacts of climate change in southern Africa are likely to be more acute in the future, yet the region remains one of the poorest and highly reliant on natural resources (Ziervogel, 2018). The prevalence of floods in the region has increased over the last few decades resulting in a greater number of studies on flood analyses and risk assessment (Busayo et al., 2022; Gründemann et al., 2018; Li et al., 2016; Mazibuko et al., 2021).

The identification of localized impacts of climate change is essential in the development and implementation of local adaptation strategies (Reimann et al., 2021; Vogel et al., 2020). South African mountain lowlands are home to rural and semi-rural households who are largely low-income earners and directly dependent on mountain ecosystems (Ngwenya et al., 2019). However, there are only a few studies on the impacts of climate change on South African mountains (Mukwada & Manatsa, 2018; Mukwada, 2022). Therefore, the objectives of this study are to examine the rainfall and temperature trends of Maluti-a-Phofung Municipality over the last 120 years and to assess the impact of changing climates on low-income households living along the Drakensberg Mountains.

Materials and methods

Study area

The Maluti-a-Phofung Municipality is located in the Thabo Mofutsanyane District in the Free State Province of South Africa. Covering an area of approximately 4 421 km2, the municipality is divided into 35 wards (Maluti-a-Phofung Municipality, 2017). Phuthaditjhaba is the administrative urban centre of Qwaqwa situated in the south-western part of the municipality and surrounded by rural villages (Fig. 1).

Fig. 1
figure 1

Location of Maluti-a-Phofung Municipality, in the Free State province of South Africa

Full size image

The climate of Maluti-a-Phofung Municipality is characterised by hot summers and cold winters with Phuthaditjhaba averaging 24.4o C in the summer months (November – March) and -0.5o C in the winter (May – August) (Molefe et al., 2013; Taylor, 2023). Mean annual rainfall in the municipality is around 653 mm (Molefe et al., 2013). The main economic activities in the municipality are agricultural activities such as cattle ranching, dairy and crop production (Maluti-a-Phofung Local Municipality, 2017).

The municipality has a population of 335 784 with the greater proportion of the people living in Qwaqwa where the population was estimated to be about 280 000 in 2011 (Stats SA, 2011; Denoon-Stevens & Mocwagae, 2019). Due to the high unemployment rate (42%) most people in the municipality are self-employed and depend on subsistence farming to survive (Stats SA, 2011). The municipality has one of the highest poverty levels both in the Free State Province and nationally (Buschke & Seaman, 2014; Hlalele, 2019). According to Statistics South Africa (2011), the municipality has 60% of its inhabitants earning less than South African Rand (ZAR)1 650/ month.

The Maloti-Drakensberg region supplies South Africa with 25% of its water (Mukwada et al., 2018) and mountains are largely made up of grasslands with sparse trees and woody shrubs. Maluti-a-Phofung Local Municipality has been plagued by droughts since 2015 which has led to massive water shortages, negatively impacting farmers and households (SA Government, 2020). The municipality also regularly experiences service delivery protests by local residents, due to water shortages and electricity cuts. With 68% of households having no immediate access to portable water in the municipality, the situation is worse in Qwaqwa where 88.8% of households experience water access challenges (Mocwagae & Mphambukeli, 2023). These households ultimately have to collect water from alternative sources such as municipality tankers and rivers (Taylor, 2023). Residents ranked climate change, negligence and corruption as the leading causes of the water crisis (Mocwagae & Mphambukeli, 2023).

Most households both in the rural and urban parts of Qwaqwa are connected to the main electricity grid and piped water networks though supply is erratic. Other socio-economic characteristics of Qwaqwa reveal general underdevelopment and the prevalence of communicable diseases such as HIV and tuberculosis (Cronjé & Barker, 2006; Victor, 2009).

Climate data

Climatic data were obtained from Climate Explorer for Maluti-a-Phofung Municipality to study the climatic trends. Climate Explorer (www.climexp.knmi.nl) is a World Meteorological Organization web-based application for climate research managed by the Royal Netherlands Meteorological Institute (KNMI). The data consisted of monthly rainfall and maximum temperature data spanning 120 years, that is, from 1900 to 2020. The analysis of such data would be critical to identify if significant changes in climate have occurred over the last century, thus a regression analysis was used. Regression analysis is commonly used to understand the relationship between variables using regression models. Average monthly data was downloaded and the annual average calculated from 1900 to 2020 and plotted on a graph using Microsoft Excel. The trend line or regression line was also plotted with the equation and the coefficient correlation. The significance of the coefficient correlation was tested by finding the p-value.

To assess if the temperature and rainfall trends of the local municipality are changing, hypotheses were developed. Firstly, to test if the average maximum annual temperature has been increasing and secondly to test if the average annual rainfall of Maluti-a-Phofung Local Municipality has been decreasing over the past 120 years. These hypotheses were informed by predicted Southern African climate data trends of rising temperatures and declining rainfall levels.

Mountain household surveys

To assess the impacts of climate change on the mountain community, a survey was conducted by selecting households from 12 villages in Qwaqwa. The total number of households sampled was 383 and this sample size was calculated using the Cochrane’s equation. The questionnaires were proportionally distributed amongst all the villages based on population size. Households were sampled using systematic sampling where every tenth household was selected in each village. Where household members were not available or refused to participate, the next household would be selected. A household questionnaire was used to provide socio-economic context of the local community, understand the level of dependence on mountain ecosystems, perceptions on climate change and the hazards experienced. High levels of poverty have been associated with heavy dependence on natural ecosystems and results in a two-fold vulnerability to climate change (Nerfa et al., 2020) because of low adaptive capacity and loss of natural safety-nets.

The household questionnaire was semi-structured and mostly included closed-ended questions. The questionnaire was divided into three sections. The first section consisted of the demographic questions such as age, education and employment status, while the second section was composed of questions on natural resource use, climate hazards and their impacts on the households. The third section primarily focused on ecosystem governance policies and adaptation. Consent was obtained from all the respondents prior to the interviews, which were conducted in the local Sesotho language by trained research assistants. These interviews allowed for greater interaction with the respondents thus increasing their understanding of the main concepts of the study.

Household survey questionnaire analysis

Data from the household questionnaire were mainly quantitative in nature therefore, a combination of descriptive analysis and hypothesis testing was conducted. The results of the household survey were analysed using Spatial Analysis for Social Sciences (SPSS) for descriptive statistics. The association of independent (age, employment status, level of education, monthly income, number of years lived in the area) and dependent variables (for example, access and harvesting of mountain resources, impacts of drought and extreme weather) was tested using Chi-square test for independence.

Results and discussion

Temperature and rainfall trends

The average maximum annual temperatures and rainfall trends of Maluti-a-Phofung Municipality for over 120 years were analysed using a regression analysis. The regression analysis shows that the average maximum annual temperatures have been steadily increasing since 1900. The highest average maximum temperature was recorded in 2019 (Fig. 2).

Fig. 2
figure 2

Average maximum temperature trend in Maluti-a-Phofung Municipality between 1900 and 2020

Full size image

The results show increases in temperature are highly significant as depicted by the p value < 0.0001. Therefore, the hypothesis that the temperature in Maluti-a-Phofung Municipality is increasing is accepted.

Average annual rainfall has been increasing gradually over the past 120 years (Fig. 3). The regression analysis of the average annual rainfall for the municipality found that rainfall recorded has been increasing at a highly significant rate (R2 = 0.1244, P < 0.0001), therefore, the hypothesis that the amount of rainfall has been decreasing is rejected. The study by Mukwada and Manatsa (2018) similarly found that both precipitation and temperature have been rising in the Namahadi Catchment Area in the Drakensberg Mountains since 1960, though only the temperature increase was significant. In addition, the rainfall results are dissimilar to the findings by Taylor et al. (2016) who found no significant change in the annual rainfall over a 30 to 40 year period in the Drakensberg. Furthermore, results from other parts of the continent show contradictory findings where precipitation had been declining since 1980 though not significantly in the Bamenda Highlands of Cameroon (Innocent et al., 2016). Also, a study on glacial recession in the Rwenzori Mountains in East Africa found that temperatures were increasing by 0.5 °C for four subsequent decades since 1960 (Taylor et al., 2006). However, future projections of East African mountains (Taita Hills and Kilimanjaro) indicate higher rainfall as variability of seasons increases (Nsengiyumva, 2019). The recurrence and severity of droughts has been increasing along the Drakensberg Mountains of the Free State Province and Qwaqwa has been affected by these since the severe drought event in 2015 (Mukwada, 2022; Orimoloye et al., 2022). This area therefore experiences extreme weather events consisting of periodic droughts and high intensity rainfall which indicate increasing rainfall variability. Climate change is anticipated to have widespread impacts whose magnitude is location specific and dependent on the socio-economic characteristics of local communities (Hallegatte & Rozenberg, 2017).

Fig. 3
figure 3

Average annual rainfall trends in Maluti-a-Phofung Municipality over 120 years (1900 – 2020)

Full size image

Sample household characteristics

A survey was conducted on 383 households located the villages of Qwaqwa. The study respondents constituted of mostly females (66%) and 34% males. The respondents were household heads primarily aged 60 years and older (31%) and between 30 to 45 years (26%). Only 1% of respondents were less than 18 years old and these consisted of child headed households.

Village household survey results showed that most of the respondents had attained some secondary education (44%) while the lowest number of respondents (13%) had no formal education. Those who had attained a tertiary educational level in the form of a diploma, degree or skills training constituted only 19% of the respondents (Fig. 4b).

Fig. 4
figure 4

a Respondent age distribution; b Level of education attained; c Employment status and b Income per capita distribution

Full size image

Figure 4c shows that 57% of respondents in the survey were unemployed with the least number of respondents being formally employed (12%). Additionally, the majority (56.8%) of households earned less than ZAR 2000 or less per month while 5.6% of households had per capita income of ZAR 10 000 or more (Fig. 4d).

Households’ dependence on mountain natural resource

The socio-ecological interactions between households and mountain ecosystem goods and services are an indicator of the type and magnitude of impact that climate change will have on resource dependent households.

Most households perceived their access to mountain products as limited. Only 40% of households noted that they had access to mountain natural resources while 60% perceived they did not. However, 43% of the households reported that they were directly involved in harvesting mountain resources. Households harvested an array of mountain resources which included firewood, animal dung, water and wild edible plants or herbs.

Firewood (75%) was the most harvested mountain resource followed by animal dung (15%) and wild edible plants and herbs (6%) (Fig. 5). Building poles and water were the least harvested resources as they were harvested by only 1% of the households respectively.

Fig. 5
figure 5

Household harvesting of mountain resources in Qwaqwa

Full size image

Households that used electricity as a primary source of energy were 84% while 7% used firewood (Fig. 6).

Fig. 6
figure 6

Household primary energy sources

Full size image

Most households used electricity for their primary energy needs, that is, for heating, cooking and lighting. Only 7% of households used firewood as a primary energy source despite it being the most harvested mountain resource. This trend may indicate that firewood is mostly used as an alternative or secondary source of energy These results vary from other areas with similar socio-economic characteristics in South Africa where despite households having access to electricity, they still used firewood as the primary source of energy particularly for cooking and heating (Madubansi & Shackleton, 2007; Masekela & Semenya, 2021; Uhunamure et al., 2017). The major reason for this phenomenon was the cost of electricity and the limited access to free basic electricity (Masekela & Semenya, 2021). Studies from other developing countries also show dissimilar patterns of high firewood and non-timber forest product use (Sharma et al., 2009). Higher mountain resource use trends in other areas also emanate from the higher availability of natural resources proximate to local households.

Perceptions of climate change evidence

Climate change perceptions are primarily based on awareness, insight and individual experiences. Households who perceived climate change is occurring in Qwaqwa highlighted certain environmental trends that they believed reflect evidence of climate change (Fig. 7).

Fig. 7
figure 7

Household evidence of climate change

Full size image

Changing seasons (24%) was the most common evidence of climate change highlighted by households in all the villages. Households also identified reduced rainfall (21%), high temperatures (16%), erratic weather conditions (9%) and drought (9%). The household perceptions reflect that there is an awareness that the climate in Qwaqwa is changing due to changes in the occurrence of other specific phenomena, such as, changes in plant growth and low productivity of crops (Fig. 7). The ability to link changes in the climate to environmental changes indicates a higher level of understanding of climate change as a real phenomenon. Though this understanding was apparent most households could not identify the causes or source of these changes in the climate (Manandhar et al., 2015). Similarly, only 1% of households in Qwaqwa identified extreme weather as evidence of climate change (Fig. 7) yet 50.9% of households (Fig. 8) reported that they were impacted by extreme weather. These percentages show that households do not necessarily associate extreme weather with climate change. However, household perceptions aligned correctly with the climatic data in terms of rising temperatures in the area and were divergent in terms of the amount of rainfall the area was receiving. Rainfall trends (Fig. 3) show a significant increase in rainfall in the municipality while by 21% households perceive a reduction in rainfall amount. These results differ from the study by Manandhar et al. (2015) which found that the local community’s perception of the rainfall trends correctly aligned with climatic observations. The reason for the perception of lower rainfall trends in Qwaqwa may be due to rainfall trends becoming more erratic and the occurrence of mid-season droughts.

Fig. 8
figure 8

Prevalence of natural hazard occurrence

Full size image

Impacts of natural hazards on households

Households in Qwaqwa experience a wide range of natural hazards. The results in Fig. 8 show that the most prevalent natural hazards cited by household respondents were extreme weather events (50.9%) followed by drought (25.5%) and floods (5.2%).

The occurrence of wildfires (0.8%) and mudflow (0.3%) were identified by respondents as the least common hazards (Fig. 8).

Extreme weather, specifically storms and hail, were the most prevalent climatic hazards experienced by the households followed by drought. These results are similar to a mountain study by Gurung et al. (2021) which reported that respondents perceived an increase in hailstorms and winter droughts. The study also found that it was not only the frequency of the drought but the severity as well. However, the study by Manandhar et al. (2015) found that over 70% of local households identified droughts and floods as hazards they experience.

The prevalence of hazards can potentially impact the status and livelihood of a household negatively because of limited access to natural resources. Since most households identified extreme weather as the most prevalent hazard, it was also found to have the highest overall impact on households (Fig. 9).

Fig. 9
figure 9

Impacts of natural hazards on households

Full size image

These results mirror those by Gurung et al. (2021) where hailstorms and strong winds were aspects of extreme weather that were identified as having the highest increase in impact while wildfires had the least impact. The occurrence of extreme weather events have been similarly observed in the Himalayas where cloudburst triggered flash floods accounting for loss of life and 70% of property damage (Sati, 2011). In Qwaqwa, drought (25.5%) was perceived as having a greater impact than floods (12%) which is similar to what was found in the study by Manandhar et al. (2015) which established that drought had the most impact on the local people, followed by floods and landslides. Inversely, findings by Gurung et al. (2021) who found that the impacts of floods were higher and had increased more than impact of drought.

Damage to property was identified as the highest impact of the extreme weather and all the other hazards while destruction of plants was highest under the drought in this study. Drought impact on livelihoods (12%) was lower in Qwaqwa (Fig. 9) which is contrary to Manandhar et al. (2015) where more than 70% of households perceived that droughts and floods negatively impacted their livelihoods. The difference in perception is likely due to fewer Qwaqwa households engaging in climate dependent crop farming as a source of livelihood. Most households have small gardens largely used to supplement diets. The impact of the identified climatic hazards on the local households extended to the socio-ecological interactions between households and mountain ecosystems, that is access harvesting of natural resources.

Socio-economic characteristics and perceptions of climate change impacts

The association between independent variables, gender, education, employment status, income, number of years lived in the village and dependent variables, that is, different aspects of mountain resource access, use and the impacts of climate hazards were assessed using the Chi-square test of independence. The Chi-square test is ideally used to analyse categorical data after cross-tabulation and the overall results depicted in Table 1.

Table 1 Chi-square analysis test of independence p-values between dependent and independent variables
Full size table

Gender

The association between gender and access, harvesting, natural hazard occurrence and the impacts of climate change was analysed using Chi-square test of independence. The results show no significant association between gender and the dependent variables (Table 1).

Age

The relationship between age (older vs younger) and individual dependent variables was analysed using Chi-square test of independence. This analysis revealed that there was a significant association between age and perception on natural hazard occurrence, X2(6, N = 382) = 14.25, p = 0.0196. Older respondents (> 45 years) perceived more hazards occurred in Qwaqwa than younger respondents (< 45 years). One reason for this is that older respondents are likely to have experienced more hazards over the course of the years and thus were more aware of the changes in the frequency and intensity of hazards.

Education

The analysis of the relationship between the level of education of the respondent and individual dependent variables using Chi-square test of independence. The test revealed that there was a marginal significant relation between education and harvesting of mountain resources, X2 (3, N = 379) = 7.73, p = 0.053. More household heads who had attained some level of education (49%) reported that they did not harvest mountain resources possibly because of greater access to alternative sources of income through employment.

Employment status

The household respondent’s employment status and various dependent variables were analysed using Chi-square test. The results of the test show that the association between employment status and the perception that climate hazards had an impact on mountain resources availability was significant, X2(3, N = 295) = 11.82, p = 0.008. Unemployed respondents (54%) affirmed that hazards negatively impact availability of natural resources followed by retired (19%), self-employed (14%) and formally employed (12%).

Income per capita

Household monthly income was analysed for association with resource access, harvesting, natural hazard occurrence, perceptions on impacts of drought and extreme weather. The Chi-square test of independence showed that income per capita and household perception on impacts of extreme weather had a significant relationship, X2 (12, N = 218) = 22.6, p = 0.031). Low-income households reported more extreme weather impacts than higher income households. Households earning less that ZAR 2000 highlighted 60% of the impacts while those who earned between ZAR 2001 – 5000 identified 26% of the impacts and those earning more than ZAR 10 000 reporting only 6% of extreme weather impacts.

Number of years lived

The households living in Qwaqwa are exposed to the climate hazards occurring in the area. Chi-square test of independence used to analyse the relationship between the perception of natural hazard occurrence and number of years that the household has lived in Qwaqwa found a highly significant relationship, X2 (15, N = 382) = 91, p = 6.4e−12. Households that had lived the longest, that is, more than 10 years (72%) reported the greater number of hazards experienced than those that have lived in Qwaqwa for 10 years or less.

Household resource use patterns

Qwaqwa households are generally less directly reliant on mountain ecosystems to meet their primary needs. The reason for lower levels of dependence on the mountain ecosystems is likely due to a greater dependence on man-made systems to provide mainly energy and water.

Household mountain resource use patterns in Qwaqwa are also lower in comparison to other semi-rural areas in South Africa where low-income households actively choose to use firewood more as it is considered cheaper than electricity and is widely available from local forests (Ngwenya et al., 2019; Williams & Shackleton, 2002). In such areas, the importance and value of fuelwood is perceived as higher such that it is used as a source of livelihood when it is harvested for sale (Matsika et al., 2013). The Maluti-a-Phofung Municipality currently owes the national power utility, Eskom approximately ZAR 7.2 billion (Business Live, 2023) which is partly a result of corruption in the municipality and limited revenue collection particularly in the informal parts of Phuthaditjhaba (Mocwagae, 2020; Mocwagae & Mphambukeli, 2023). Consequently, the municipality has limited capacity to supply electricity consistently to all the villages. The current national energy crisis in South Africa (Akinbami et al., 2021) has therefore worsened the energy supply problems already in existence in Qwaqwa over the last few years.

Most households in Qwaqwa resort to using firewood during times of extended power cuts to meet their energy requirements. This situation has enhanced households’ reliance on natural energy sources such as firewood as it is the more affordable and readily available option in comparison to other sources of energy such as gas. Though natural systems are irrefutably important as safety-nets, they generally tend to be undervalued with little or no planning and management of their exploitation. Climate change is expected to contribute to biodiversity loss and coupled with overexploitation of natural resources, poses a danger in the destruction of the natural safety-nets for households through the disappearance of forests. Mountain ecosystem goods also act as a safety-net for households in times of stress from the loss of a job or the death of a bread winner, thus increasing their importance for lower income households (Adhikari et al., 2018; Negi et al., 2012). The negative impacts of climate change also eventually reduce the adaptive capacity of local communities through the elimination of these safety-nets.

Impacts of climate change on poor households

Socio-economic factors such as income determine the level of dependence on mountain resources and adaptive capacity of households (Chauhan et al., 2021), hence climate change impacts low-income households the most. The employment status of the respondents had a significant relationship with the perception that climate hazards impact resource availability. The results show that the majority of those with this perception were unemployed (54%), indicating that natural resources are more important to the unemployed possibly to supplement diets, energy and as a source of livelihood.

Furthermore, households in Qwaqwa indicated that extreme weather was the most common natural hazard experienced with impacts including destruction of property and crops. The lowest income group earning ZAR 2000 or less (60%) reported that they were impacted by extreme weather events, that is, storms or hail more than the higher income groups. Poorer households are less likely to invest in high quality building materials to build or reinforce their houses to withstand extreme weather (Ullah et al., 2021) and hence become more prone to experiencing damage to property which was highlighted as the most prevalent impact. Adaptation generally requires significant financial input which the poor do not have access to (Alemayehu & Bewket, 2017). Rural communities in developing countries reflect higher levels of vulnerability to climate hazards due to low adaptive capacity (Mainali & Pricope, 2017; Simane et al., 2016; Tessema & Simane, 2019). Qwaqwa households can therefore be categorised as having a low adaptive capacity since income is directly linked with the ability to adapt and prepare for future hazards.

Extreme weather events also have a profound impact on rural communities dependent on natural ecosystems (Kieslinger et al., 2019) by altering ecosystem services supply to communities.

Despite increasing rainfall trends in Qwaqwa, households cited drought as the second most common natural hazard they experienced. The last major drought was recorded in 2015 though water shortages are a regular occurrence. In Qwaqwa this phenomenon may be explained by increased rainfall variability and poor resource governance. The interaction of climate change and anthropogenic practices on mountains further complicates the dynamics of impacts experienced (Said et al., 2019) and hence the adaptation process. Qwaqwa households indicated that the atmospheric changes in the area led to low productivity and declining biodiversity of indigenous plants. Some studies have shown that climate change ultimately impacts socio-economic systems such as agriculture, forestry and tourism, however other impacts of the phenomenon have not been well-articulated (Knowles & Scott, 2021; Gentle & Thwaites, 2016; Dhanai et al., 2014). The uncertainty associated with the impacts of warmer temperatures in specific areas ultimately affects the management of these areas and their human practices (Rasmussen et al., 2019). High temperatures and dry conditions are likely to reduce forest productivity and the availability of natural resources, hence eliminating the natural safety-net function provided by the forests. Mountain areas with high levels of poverty, ecosystem dependence, and subsistence agriculture are therefore more vulnerable to climate change (Pandey et al., 2016).

Poor governance and climate change

Local government has been identified as being a key partner in climate change adaptation (Grumbine & Xu, 2021). In a bid to align to international treaties and agreements on climate change and adaptation, the South African government has developed policies and frameworks gleaned from these treaties for the promotion of adaptation initiatives (Amusan & Olutola, 2016; DEA, 2011; Seo, 2017). The top-down approach adopted is used to drive these initiatives to the lower levels of governance and ultimately the local communities. Governing institutions also determine community adaptive capacity since they hold key decision-making power, financial resources, develop legislative environment, and vital networks with other institutions that can enhance adaptive capacity (Montana et al., 2016).

However, poor governance limits household access to basic water and energy forcing them to seek available and cheaper alternatives. During these times the direct dependence on natural systems escalates. Maluti-a-Phofung Local Municipality was placed under administration in 2018 due to issues of financial maladministration (The Local Government Handbook, 2021). Since the local government is struggling with basic service delivery issues it is unlikely to have the financial capacity and expertise to provide support to Qwaqwa households in coping with the impacts and adapting to climate change.

Conclusions and recommendations

Temperature and rainfall trends have been significantly rising in Maluti-a-Phofung Municipality since 1900. Qwaqwa households have been experiencing the harsh impacts of climate change through the occurrence of hazards. These households identified extreme weather and drought as having the greatest impact through damage to property and destruction of plants. Unemployed respondents perceived that climate hazards had an impact on mountain resource availability, indicating a greater dependence on mountain resources by this group. Low-income households were also found to be more negatively impacted by extreme weather events thus indicating low-adaptive capacity. Respondents were able to identify changing seasons, reduced rainfall and higher temperatures as evidence of climate change in Qwaqwa.

Qwaqwa mostly had low-income households with high levels of unemployment. Only 40% of households perceived they had access to mountain resources while 43% indicated that they harvest mountain resources. Firewood was the most harvested mountain resource though mainly used as a secondary energy source during power cuts. The results of this study show that mountain resources are increasingly becoming important in Qwaqwa due to the energy and water crisis. However, mountain ecosystems are also experiencing threats from climate change which will undermine their capacity to act as safety-nets. Households will have to develop suitable low-cost adaptive strategies with the support of external actors such as the government, private organisations and agencies. Identifying the impacts of climate change in the mountain communities of Maluti-a-Phofung Local Municipality is vital in the development of appropriate adaptation strategies, increasing adaptation capacity and the identification of areas of intervention by the local government. However, these tasks may be more difficult to achieve in Qwaqwa due to the prevailing service delivery and governing challenges.