Determinants of Small-Scale Farmers’ Participation in Social Capital Networks to Enhance Adoption of Climate Change Adaptation Strategies in OR Tambo District, South Africa

Abstract

Globally, climate change remains one of the most pressing challenges, and it is also an obstacle to the fundamental achievement of the Sustainable Development Goals. The purpose of the study was to examine the determinants of small-scale farmers’ participation in social capital networks to enhance adoption of climate change adaptation strategies. Multistage and purposive sampling were used to carry out the study. A cross-sectional research design was used to carry out the study and structured questionnaires were used as a data collection tool. The data collected were analyzed using descriptive statistics, the Binary Logit model, and the Ordered Logit model. The findings of the study show that gender, household size, age, marital status, education, and employment status affect a farmer’s participation in social capital networks. The results also show that household size, employment status, and income level affect the extent of participation in social networks. As the paper is the first to look at the determinants of the participation of farmers in social capital networks in Eastern Cape, the results are of paramount importance to policy formulators in order to formulate policies that will encourage farmers to join localized farmer-based social capital networks to adopt climate change adaption measures.

1. Introduction

Globally, climate change remains one of the most pressing issues and it is also an obstacle to the fundamental attainment of the Sustainable Development Goals. According to Ortiz-Bobea et al. [1], climate variability over the last 60 years has led to a 21% reduction in agricultural yields globally. Climate variation and the prevalence of extreme weather events are anticipated to upsurge within the twenty-first century, having a negative impact on the growth of agricultural systems, especially in countries in the Global South [2]. Climate variation has adverse effects on agricultural production in Africa because Africa relies on rain-fed agricultural systems. As is the case with many African nations, South Africa is often recognized as being extremely susceptible to climate variability. In the last twenty years, South Africa has turned out to be increasingly susceptible to climate variability events such as cyclones, temperature extremes, and unpredictable rainfalls [3]. In the face of climate change calamities, small-scale farmers are the most affected as they are often recognized as farmers with less resources to cope with climate change [4].

Most small-scale farmers fail to adapt to climate change not because they do not want to but because they lack resources. To adapt well, economic, environmental, and social resources are all critical. Farmers with limited resources are often able to acquire management practices and adaptation measures through the formation of networks that make them able to share risks [5]. The formation of networks produces social capital, which is referred to as a shared value in the form of customs, thoughts, sentiments, trust, societal organizations, shared relations, and institutions, which enable cooperation for common benefits [6]. Poor service delivery has been habitual in countries in the Global South and impoverished societies deal with climate change-related events such as floods by collaboration and engaging with each other in order to enable service provision. Having access to a social network that may support one in times of need may lessen risk aversion and allow one to try innovative technology [6].

Social capital networks assist small-scale farmers to pull their limited resources together, thus enabling them to maximize adoption of adaptation approaches to climate variation [6]. Social capital networks have a vital role in the adaptation process of small-scale farmers because external assistance is often insufficient or late [7]. A social capital network gives farmers the chance to trade goods, knowledge, and labor. The ties between farmers, which also act as a means for knowledge sharing, can result in a coordinated effort to adapt to the environment. Because of the potential for collaborative knowledge acquisition and access to technologies tailored to local conditions, belonging to a social organization also encourages the adoption of ways to adapt to climate change [6]. Social capital networks serve as channels for money transfers that could ease the farmer’s credit restrictions. They also serve as informational channels for innovative technology. Social networks can enable people to work together to solve a problem using joint action.

In recent years, researchers have become increasingly interested in the impacts of changes in climate on agriculture production and how to overcome the negative impacts of climate variation on agriculture, food security, and poverty. A small number of research studies have looked at how household social capital network engagement affects the use of climate variability coping techniques [6,8,9,10]. Ogunleye et al. [6] found a positive influence of social capital networks on the adoption of climate change adaptation strategies in Nigeria, whereas Kehinde et al. [8] found that membership in cooperatives positively influenced adoption of EU improved pesticides.

There is a paucity of studies on factors influencing participation in social capital networks among farmers. The influence of social networks on the adoption of climate change adaptation strategies has been studied in Nigeria [6,8], Vietnam [11], South Africa [12], and Indonesia [13]. To our knowledge, no study has looked at the determinants of small-scale farmers’ participation in social capital networks in the Eastern Cape Province. Therefore, this study aims to fill a knowledge gap caused by the inadequate use of social capital networks in adaptation regulations, which is also made worse by a lack of sufficient data documentation. This study tries to bridge this knowledge gap by looking at the determinants of small-scale farmer’s participation in social capital networks to enhance the adoption of climate change adaptation strategies.

The remainder of the paper will unfold as follows: Section 2 describes the methodology. Section 3 presents the results of the study and discussion. Lastly, Section 4 covers the conclusion and recommendations.

2. Methodology

2.1. Study Area

The study was carried out in Port St. John’s Local Municipality (Figure 1). Port St Johns Local Municipality (PSJ) is located in the OR Tambo District of the Eastern Cape on the seaside of the Indian Ocean in the former Transkei. It has one town and has approximately one hundred and thirty rural areas. It occupies 11% of the district’s surface area, making it the smallest municipality of the 5 local municipalities in the OR Tambo [14]. The Department of Social Development has made numerous efforts to improve things for the municipality’s poor population, but the issue still exists. The majority of the population depend on social assistance, have a high prevalence of illiteracy, and have limited access to utilities like water and electricity (Stats SA, 2011). According to the Köppen–Geiger climate classification (https://en.climate-data.org/location/146968/) (accessed on 25 November 2023), the climate in PSJ is described as a moderately humid subtropical coastal climate. The region’s summertime temperatures range from an average maximum of 25 degrees Celsius to an average minimum of 20 degrees Celsius. Port St. Johns receives between 1100 and 1400 mm of rainfall each year.

2.2. Research Design

The study used a quantitative approach to identify determinants of small-scale farmer’s participation in social capital networks to enhance the adoption of climate change adaptation strategies. A survey research design was used for this study; specifically, a cross-sectional research design was used to carry out the study. The survey was conducted through the use of questionnaires to gather demographic, institutional, and other information related to the study.

2.3. Conceptual Framework

Different farmer’s characteristics influence their decision on whether to adopt climate-related adaptation strategies or not (Figure 2). Garcia-Yi [15] states that a farmer’s decision-making process is affected by the five capitals from the sustainable livelihood approach. These capitals are human, social, physical, financial, and natural. Lugandu [16] claims that whether farmers choose to implement climate change adaptation methods depends on how they perceive those strategies in comparison to other agricultural technology. Farmers may use modern technologies for a number of reasons. Some farmers may act rationally; therefore, the knowledge they have access to, their socioeconomic condition, and their agricultural enterprises may have an impact on how they perceive things [17]. Given sufficient knowledge about a new technology and its potential advantages, farmers adopt climate change adaptation strategies. Access to a social capital network which may provide assistance in times of need lowers fear of risk and may encourage people to try new technology. The availability of help, whether an act of kindness or in the form of unofficial credit, supplements the resources available to households, potentially enhancing their financial independence when making production-related decisions. Figure 2 summarizes how this paper conceptualizes the influence of social capital network on the adoption of climate change adaptation strategies.

2.4. Sampling Technique and Sampling Procedures

The study used a multistage sampling approach. The first stage was purposively selecting Port St Johns Local Municipality due to its climatic aspects. With the help of the extension officers who acted as gate keepers in this research, wards that were severely affected by the floods were then purposively selected from Port St Johns Local Municipality. Out of twenty wards in Port St Johns local Municipality, seven wards (ward 4, 5, 6, 8, 9, 10, and 11) were selected from the study area. A list of all villages in each of the selected wards was obtained from the respective extension officers. Convenient sampling was used to select twelve villages that were severely affected by the floods from the selected wards. This was performed with the help of the extension officers. Two villages were selected from Ward 4, three villages from Ward 5, one village from Ward 6, one village from Ward 8, one village from Ward 9, four villages from Ward 10, and two villages from Ward 1. According to Stats SA [18], there are 18,190 farming households in Port St Johns Local Municipality. The Yamane [19] formula was used to calculate the sample size, as follows:

$$n=\frac{N}{1+N{\left(e\right)}^2},$$

(1)

where n is the sample size, N is the population size, and e is the sampling error

$$n=\frac{\mathrm{18,190}}{1+\mathrm{18,190}\left({0.05}^2\right)}$$

(2)

The sample size for the study was 391 small-scale farmers. These farmers were conveniently selected from the twelve villages. Data were collected from 238 small-scale farmers only because the area was affected by floods before data collection, and this made some of the areas inaccessible.

2.5. Sources and Methods of Data Collection

This study utilized primary data collected through face-to-face interviews of small-scale farmers using a structured questionnaire. The target respondents for the study were small-scale farmers from Port St Johns Local Municipality. The selection of villages and farmers was conducted purposively with the help of the agricultural extension officers who often interact with the small-scale farmers. Primary data focused on issues like access to extension officers, distance to the nearest output market, access to weather information, and access to agricultural credit. Primary data also looked at socioeconomic factors and demographic aspects. Information about climate change adaptation strategies adopted by small-scale farmers and social capital networks was also obtained from these selected farmers using questionnaires. To avoid misinterpretation of the questions since the sample was large, the researcher administered the questionnaire. This was achieved with the help of field enumerators who were trained before the commencement of the data collection phase.

2.6. Data Analysis

Comparisons of various elements under investigation were conducted using Stata and SPSS. To attain the research objectives, different procedures were adopted. Firstly, descriptive statistics were used to identify the different types of social capital networks and climate change adaptation strategies used by farmers to cope with climate change. Socioeconomic and demographic characteristics of small-scale farmers were also described using descriptive statistics. Descriptive statistics were used to describe the data, including the mean, standard deviation, and coefficient of variation.

Secondly, an Ordered Logit Model was used to examine the factors determining the extent of participation in social capital networks. In this study, the extent of participation means the number of social capital networks that a participant is a member of. The Ordered Logit Model is expressed as Nick and Campbell [20]:

Z*i = β₀ + β₁X₁ + β₂X₂ + β₃X₃ + β₄X₄ + β₅X₅ + β₆X₆ + β₇X₇ + β₈X₈ + e_i,

(3)

where Z*i is the extent of participation in social capital networks (SCNs). Five social capital networks were found in Port St Johns Local Municipality and they were ordered according to: 2 social capital networks = low participation, 3 social capital networks = medium, and 4+ social capital networks = high. e_i is the error term, β₁–β₈ are coefficients of the explanatory variables to be estimated, and X_i are independent variables, where X₁ = age, X₂ = gender, X₃ = household size, X₄ = marital status, X₅ = educational attainment, X₆ = employment status, X₇ = income, X₈ = land ownership.

Thirdly, the Binary Logistic Model was employed to examine the factors influencing the participation of small-scale farmers in social capital networks for the improved adoption of climate change adaptation methods. This model was chosen due to the binary nature of the dependent variable, where it takes a value of 1 if a farmer belongs to at least one social capital network and 0 if not. The Binary Logit Model is represented as per Sha [21] and Williams [22]:

$$ln\left(\frac{p}{1-p}\right)={\beta }_0+{\beta }_1{X}_1+{\beta }_2{X}_2+{\beta }_3{X}_3+{\beta }_4{X}_4+{\beta }_5{X}_5+{\beta }_6{X}_6+{\beta }_7{X}_7+{\beta }_8{X}_8+e_i$$

(4)

where p is a binary dependent variable which takes the value of 1 if the farmer is a participant in a social capital network and 0 otherwise, ${\beta }_0$ is a constant, ${\beta }_1$–${\beta }_8$ are coefficients of the explanatory variables to be estimated, and X_i are independent variables, where X₁ = age, X₂ = gender, X₃ = household size, X₄ = marital status, X₅ = educational attainment, X₆ = employment status, X₇ = income, X₈ = land ownership, ‘e’ is the error term.

Lastly, the Binary Logit Model was used to investigate the influence of social capital networks on the selection of climate change adaptation strategies. The dichotomous dependent variable was coded 1 if a farmer picked the gth adaptation method in response to perceived climate change and 0 otherwise. Greene [23] asserts that the Binary Logit Model has the following structure:

$$In\left(\frac{P_i}{1-P_i}\right)={\pi }_o+{\pi }_1{N}_{1i}+e_i$$

(5)

where P_i is the probability that a farmer will adopt the gth adaptation strategy, π_o is the constant, π_i are parameters of the explanatory variables to be estimated, and N_i is the independent variable.

Therefore, the model can be expressed as

$$In\left(\frac{P_i}{1-P_i}\right)={\pi }_o+{\pi }_1 \mathrm{social} \mathrm{capital} \mathrm{network}$$

(6)

Choice of Variables Used in the Empirical Analysis and Justification of Inclusion

The explanatory variables input in the model specification were included based on empirical evidence from the literature [24]. The variables included in the model are shown in

Table 1. Description of variables used in the regression models.

Variable	Description and Unit of Measurement	Sign
Participation in social capital network	Is the household head a member of social capital network? Yes = 1; No = 0	+/−
Age	Age of household head in years	+/−
Gender	Gender of household head: Male = 1; Female = 0	+/−
Household size	The number of persons permanently living within a household	+/−
Marital status	Marital status of household head: Married = 1 Unmarried = 0	+/−
Educational attainment	Highest education qualification attained by household head	+/−
Income	Level of monthly income	+/−
Land ownership	Land ownership status: communal owner = 1; leased = 0	+/−

. Age affects a person’s tendency to underestimate the future and propensity to join social capital. Depending on the kind of organization, age may have different effects on involvement. The positive correlation between age and trust may lead to an increased probability of engaging in social capital networks [25]. The gender of a household can lead to disparities in preferences and impediments to the joining of social capital networks due to different roles and limits. In contrast to men, women in rural Africa may have more time constraints due to communal gender conventions, which might limit their participation in social networks. Households headed by women might not be able to join organizations that demand membership fees or other donations [26]. Acquiring knowledge and building trust are related to education. Education also boosts a person’s confidence to speak up in front of others. Households with greater levels of education might be more interested in joining organizations since they can make use of their positive externalities more readily [27]. This study assumes that education has a positive relationship with participation in social capital networks. Household size is expected to have a positive relationship with membership in social capital networks because a large family size means more labor endowment in a household so they will be able to devote some of their time to social capital activities such as attending meetings. Ref. [28] found a positive relationship between farmers’ participation in cooperatives social networks and household size.

Income is usually considered in evaluating the degree of financial stability in agricultural households. A household’s income also affects their willingness to join social capital networks. Compared to households that have full time jobs and earning a high income, unemployed or households with low levels of income may join social capital networks due to the easy access of informal credits in these networks [27]. More affluent farmers are less risk-averse [29] so they might join social capital network because they can afford to purchase farm inputs on their own without assistance from other farmers in a social network. The relationship between social capital and land ownership has not received adequate research. However, ref. [30] found positive relationship between social capital and land ownership.

2.7. Multicollinearity Check

The multicollinearity of the variables may be a problem in regression models. The VIF (Variance Inflation Factor) was used for testing this, and a VIF under 5 indicates that there is no issue of multicollinearity in the regression model [31].

3. Results and Discussion

3.1. Descriptive Statistics

Table 2. Distribution of socio-economic characteristics of participants (n = 200).

	All Participants	Social Capital Network (SCN) Participants	Non-Social Capital Network Participants
Categorical Variables	Percentage (%)	Percentage (%)	Percentage (%)
Gender
Male	30.5	27.73	34.57
Female	69.5	72.27	65.43
Age
18–35	27	20.17	37.04
36–55	31.5	30.25	33.33
56+	41.5	49.58	29.63
Marital status
Unmarried	52.5	43.70	65.43
Married	47.5	56.30	34.57
Education level
No schooling	15.5	16.81	13.58
Primary	27	29.41	23.46
Secondary	47	43.70	51.85
Tertiary	10.5	10.08	11.11
Employment status
Unemployed	82.50	78.99	87.65
Formally employed	6.50	5.88	7.41
Self-employed	11	15.13	4.94
Income level
R0–2000	83	81.51	85.19
R2001–4000	9.50	9.24	9.88
>R4000	7.50	9.24	4.94
Continuous Variable
Household Size	Mean	Standard Deviation	Minimum	Maximum
	6.8	3.5	1	22

shows the social-economic characteristics of the farming households in the study area. The gender distribution of the participants shows that females were more represented than males as both social capital network members and non-social capital networks participants. Females accounted for 72.27% of social capital network members and 65.43 of non-social capital networks respondents, whereas males accounted for 27.73% and 34.57%, respectively. The findings further reveal that, even when combined (social capital participants and non-social capital networks farmers), female-headed farming households were more than double the male-headed farming households (69.5% were females, whereas 30.5% were males). These findings are consistent with Flato et al. [32]’s findings, which demonstrate that numerous small-scale farmers’ families are headed by women because most men move to bigger cities to better their welfare and be able support their families. The findings from this research further align with those of Simelane [33], Abegunde et al. [29], and Masuku and Manyatsi [34]. These researchers similarly observed that communal farming in South Africa is predominantly undertaken by women in rural regions. Moreover, the results resonate with the South Africa Household Survey [35], which reported that half of the households in the Eastern Cape Province are headed by females.

The most represented age group among all the respondents was that aged 56 and above. The results show that less youths engage in farming because they relocate to larger cities in pursuit of better opportunities. According to Srinivasan and White [36], due to the devaluation of farming and rural living, fewer young people are interested in farming. They have no desire to pursue farming or a rural life.

For social capital network participants, the most frequent age group was that aged 56 and above, which constituted 49.58%. The least observed age group for social capital network members was that aged between 18 and 35 (n = 20.17%). For non-social capital networks= participants, the most observed age group was youth (18–35), who constituted 37.04%. The least observed age group for non-social capital network participants was elderly people who are aged 56 and above (n = 29.63%). The statistics above show that less youths participate in social capital networks. This might be due to obstacles that they face when they want to venture into farming. These obstacles include limited access to information, innovation, and credit facilities in rural areas. When it comes to marital status, most respondents were not married. Non-married respondents constituted 52.5% of the study respondents and married constituted 47.5%.

In terms of education, the findings of the study indicate that most smallholder farmers had secondary-level education (47%). More than twenty percent (27%) of the farmers attained primary-level education and 15.5% indicated that they did not go to school. A significant number of the farmers did not go to school (15.5%). The substantial number of farmers without formal education can be attributed to the after effects experienced in the area during the apartheid era. During the apartheid era, the people in the former Transkei had less access to education [37].

The least observed level of education was those with tertiary-level education; they constituted 10.5% of the respondents. A few respondents (10.5%) indicated having tertiary-level education; this is because educated people tend to move to bigger cities in search of employment and to better their livelihood [38]. The buttress results found by Stas SA [18] indicate that 23.5% of people in Port St Johns did not go to school and 3.9% have tertiary-level education. The results also indicated that the minimum number of household size is one, whereas the maximum is twenty two.

On average, farming households had seven members in their families. The study’s findings echo those of Taruvinga et al. [39], who found in their study that the average household size stood at seven members, and this was an advantage in terms of free and readily available labor force in the form of farm work. In terms of employment status, 82.50% of the farmers were unemployed, whereas 6.50% were formally employed and 11% were self-employed. The high level of unemployment among the farmers is due to high unemployment in South Africa, which is currently at 32.9% [35].

With regard to income level, most farmers (83%) indicated that their monthly income ranges from R0 to 2000. The observed sources of income were child support grants (46%), old age grant (28.6%), agriculture income (13.6%), salary (5.6%), remittances (4.4%), and lastly old age pension (2.6%). This low level of income among the households is supported by the high unemployment rate found in the area of 82.5% (descriptive statistics), which makes the farmers highly dependent on social grants, which were the main source of income among farmers.

3.2. Distribution of Farming Households by Farming Data and Institutional Factors

The results in

Table 3. Distribution of farming households by farming data and institutional factors.

Categorical Variables	Frequency (n = 200)	Percentage (%)
Land ownership
Communal	195	97.50%
Leased	5	2.50%
Farming type
Crop production only	103	51.50%
Animal production only	4	2%
Both	93	46.50%
Access to extension services
No	97	48.5%
Yes	103	51.5%
Access to formal agricultural credit
No	179	89.5%
Yes	21	10.5%
Access to weather information
No	79	39.5%
Yes	121	60.5%
Climate change awareness
No	84	42%
Yes	116	58%
Distance to the nearest output market (km)
0–10	109	54.5%
11–20	30	15%
>20	61	30.5%
Mean	17.11
Standard deviation	19.36
Minimum	0
Maximum	85
Farmland size
0.1–2	116	58%
2.1–4	64	32%
>4	20	10%

Source: Field survey, 2023.

show the distribution of farming households by farming data and institutional factors. The results from Table 3 show that most of the land is communally owned (97.5%). In this study, 2.5% of the smallholder farmers indicated that their land was leased. In terms of farming type, most farmers were into crop farming only (51.5%) and 46.5% of the farmers were into both crop and livestock farming. Most of the small scale farmers’ field sizes ranged between 0.1 and 2 ha (58%), while 32% of the farmers had farmland size between 2.1 and 4 ha and 10% had farming land that is greater than four hectares.

The results from Table 3 indicate that most farmers (51.5%) had access to extension services in the study area. The results are consistent with Loki et al. [40], who found that most farmers have access to extension services in the Eastern Cape. More than a third (48.5%) of the farmers did not have access to extension services in the study area. Extension services are crucial informational resources for coping with and adapting to climate change and they also play a vital role in the formation of social capital networks [41]. The results from Table 3 indicate that most small-scale farmers (89.5%) do not have access to formal credit. Only 10.5% of the respondents had access to formal credit from banks. Most farmers do not have access to formal credit because banks and other financial organizations only issue credit to farmers who can provide the collateral needed to repay the loan. In the study area, most of the land is communally owned and this makes it challenging for banks to grant loans to the majority of farmers who lack any kind of security. Access to credit remains a challenge among rural farmers in the Eastern Cape, and this was also found by Mdiya et al. [42].

More than half of the farmers from the study area had access to weather information; they constituted 60.5% of the respondents, whereas 39.5% had no access to weather information. Of those that had access to weather information, the sources included television, radio, and cell phones. From the results in Table 3, 58% of the farmers were aware of climate change, whereas 42% were not. Farmers with knowledge of climate variability, what is causing changes in climate, and climate variability impacts tend to employ adaptation strategies and coping techniques to counteract its negative impacts [43]. With regard to the distance to the nearest output markets, most farmers indicated they travel 0–10 km in order to sell their farm produce. They constituted 54.5% of respondents. Most of the farmers in the study area sell their output locally and some farmers grow for home consumption as they rely on family labor. More than 10% (15%) of the farmers travel 11–20 km in order to sell their produce to the nearest town, which is Port St Johns, and 30.5% travel more than twenty kilometers in order to sell their products. Most farmers sell their output locally because access to markets is challenging due to bad roads that were worsened by the floods, few possibilities for travel, and inadequate means of product transportation.

3.3. Adaptation Strategies Adopted Due to Use of Social Capital Networks

Table 4. Adopted adaptation strategies in the study area.

Adaptation Strategy	Percentage (%)
Changing planting dates	74.19
Use of new improved varieties	33.33
Use of organic manure	51.61
Mixed farming	7.53
Crop diversification	18.28

shows the different adaptation strategies adopted by the farmers in the study area. The climate change adaptation measures adopted in the study area include changing planting dates, use of new and improved variety, use of organic manure, mixed farming, and crop diversification. The adopted adaptation methods are typically used in conjunction with other techniques. The most used adaptation strategy is changing planting dates (74.19%), followed by use of organic manure (51.61%).

3.4. Empirical Results

3.4.1. Factors Influencing Participation in Social Capital Networks

Table 5. Determinants of participation in social capital networks—Binary Logistic Regression.

	Part of Farmer Group	Part of NGO Project	Part of Cooperative	Religious Group	Family Group
Gender	−1.16 ** (0.60) [0.31]	−0.75 (0.59) [0.47]	0.33 (0.54) [1.39]	−0.63 (0.74) [0.53]	1.50 ** (0.71) [4.48]
Age	−0.13 (0.35) [0.88]	−0.15 (0.38) [0.86]	−0.25 (0.39) [0.78]	0.90 * (0.49) [2.45]	0.55 (0.57) [1.74]
Marital status	−0.88 * (0.51) [0.41]	−0.24 (0.52) [0.78]	−0.97 (0.60) [0.38]	−0.40 (0.64) [0.67]	1.11 (0.88) [3.02]
Level of education	−0.11 (0.30) [0.89]	0.17 (0.31) [1.19]	−0.69 ** (0.34) [0.50]	0.52 (0.43) [1.69]	0.40 (0.54) [1.49]
Household size	0.01 (0.06) [1.01]	−0.01 (0.06) [0.99]	−0.03 (0.07) [0.97]	−0.12 * (0.07) [0.89]	−0.22 *** (0.09) [0.80]
Employment status	−0.79 *** (0.30) [0.45]	0.23 (0.44) [1.26]	−0.19 (0.40) [0.83]	0.37 (0.56) [1.45]	0.19 (0.70) [1.21]
Income level	−0.02 (0.38) [0.98]	−0.28 (0.48) [0.76]	−0.22 (0.43) [0.80]	−0.44 (0.51) [0.65]	−0.11 (0.70) [0.90]
Land Ownership	−1.10 (0.88) [0.33]	18.24 (15,393.12) [83,315,486.99]	0.33 (1.11) [1.39]	−1.30 (1.05) [0.27]	−1.62 (1.17) [0.20]
Constant	5.80 *** (1.65) [330.02]	−15.36 (15,393.12) [0.00]	4.53 *** (1.78) [92.95]	2.71 (1.90) [14.96]	2.92 (2.01) [18.63]
Model Summary
Sig	0.00	0.00	0.00	0.00	0.00
Exp(B)	5.82	7.30	8.55	10.24	13.69
Nagelkerke	0.16	0.06	0.11	0.11	0.21

Source: Field survey, 2023. Sig at * 10%, ** 5% and *** 1%. ( ) Standard error. [ ] Exp(B).

shows the factors affecting the utilization of social networks by smallholder farmers in the study area. Multicollinearity was checked first before running the model. The relationship between the independent variables was checked using VIFs and tolerance. All the VIFs were less than five and the tolerance was above 0.20, meaning that the independent variables were not correlated to each other. The Binary Logistic Regression model was used to determine factors influencing participation in different social capital networks. All the models were significant at the 1% level (Table 5). However, low Nagelkerke values between 0.06 and 0.21 were observed, showing that there were other variables that affect participation in social capital networks that were not included in the model. Table 5 shows that gender was significant in being part of a farmer group and family group (5% level) and household size in relation to religious beliefs (10% level) and family groups (1% level). Age was also significant in engaging with religious groups (10% level), marital status for farmer groups (10% level), educational levels for cooperatives (10% level), and employment status for farmer groups (1% level).

Table 5 shows that gender was significant with a negative coefficient for being part of farmer groups. This means that females were less likely to be part of farmer groups, while men were likely to be part of family groups. The negative relationship between a female’s participation in a farmer group can be associated with females’ involvement in decision-making processes. Men normally join farmer groups because they are the head of households. These study results are similar to those of Ingutia and Sumelius [44], whose findings indicate that a female’s involvement in farmers’ groups is adversely affected by social-cultural issues that restrict females being involved in making decisions. Gender was significant, with a positive coefficient for being part of a family group. This means that females were more likely to be part of family groups. This might be because most of the farming households in the study area are female headed.

Household size was significant and negatively related to being part of a religious groups. As household size increased, the lower the likelihood that smallholder farmers will be part of religious groups. The relationship between family size and being part of a religious network can be due to the fact that strong familial ties tend to pull away one’s focus and effort from joining other networks. A large household size indicates potential labor endowment of the household, so bigger families can conduct farming activities without the help of religious groups. A large family size means that families can deal with the aftermath of floods such as building fences destroyed by floods and clearing the paths to their fields closed by fallen trees. The availability of family labor can influence farmers to adopt labor-intensive adaptation strategies such as applying organic manure to the fields. The results are contrary to those of Zakari et al. [45], who found that an increase in household size causes an increase in a household’s participation in social capital networks.

Age was found to positively affect farmers’ participation in religious networks. An increase in age resulted in an increased likelihood of being part of religious social networks. The results can be associated with more involvement of people in religious groups as they grow. Young people usually replace religion with sports and music. The participation of older farmers in religious social networks can influence their attitude towards the adoption of climate change because religious interactions also affect how people perceive climate change and how they deal with the effects of climate change events. This finding was consistent with that of Adong et al. [46], who found that the elderly were 0.9% more likely to engage in social capital networks than younger individuals. Different results were observed by Ghana by Katungi et al. [24] who found a negative relationship between age and membership in social networks.

Marital status was significant and negatively related to being part of a farmer group. This means that married smallholder farmers were less likely to be part of farmer groups. Married households tend not to join farmer groups because according to Zhao et al. [47], married households usually have a higher level of income stability than unmarried ones, and hence they might not join social capital networks. Married farmers may not choose to join farmer groups because they can make informed decisions when it comes to adopting climate change adaptation strategies without influence from a group of farmers. This is because according to Atube et al. [48], marital status is positively related to decision making among households. Banful, Nkonya, and Oboh [49], Umar, Musa, and Kamsang [50], and Kaliba et al. [51] found that married households were more likely to adopt climate change adaption strategies because they tend to have other distinct networks such as non-governmental organization projects compared to unmarried farmers who depend on other farmers as their source of farming information.

Education was also a significant factor that affects participation in cooperatives. Education shows a negative correlation with participation in cooperatives. The results indicate that an increase in education will decrease the use of cooperative social networks. The negative correlation between education and joining cooperatives could be associated with the reason behind joining cooperatives in the area. Farmers in the area indicated that they join cooperatives because they help them with inputs and information. Educated people can obtain information from other sources such as the internet so they might not join cooperatives in PSJ. The results are opposite to those of Dendup and Aditto [52], who found that education increases a farmer’s participation in cooperative social networks because these networks benefit the farmers by giving them access to farm equipment and inputs. Mostly, education is usually correlated with employment and income, so this means that educated farmers may not join cooperatives because they are able to buy their own inputs without any help from cooperatives. Educated farmers also understand what climate change is and know the importance of adopting strategies meant to mitigate any arising challenges. There is an increased likelihood that an educated farmer can adopt climate change adaptation strategies even if they are not part of cooperatives.

Employment status was significant for being part of a farmer group. Employment status shows a negative correlation with participation in a farmer group. This indicates that employed farmers are less likely to join farmer groups. This could be because most small-scale farmers in rural areas join farmer groups for financial and input support, so when one is employed, they can afford to buy inputs and farming equipment on their own without any support from a social network. Also, employed farmers have less desire to join farmer groups because they do not have time as they spend most of their time at work. The results concur with the findings from Adong et al. [46] in Uganda.

3.4.2. Factors Affecting Extent of Participation in Social Capital Networks

The extent of utilizing social networks by smallholder farmers is shown in

Table 6. Determinants on the number of social capital networks—Ordinal Logistic Regression.

Variable	β	Std. Error	Sig
Household size	0.10	0.05	0.03 *
Gender	−0.46	0.34	0.18
Age	−0.66	0.50	0.18
Marital status	−0.58	0.36	0.10
Education	−0.89	0.72	0.22
Employment status	−1.51	0.79	0.06 *
Income	−1.69	0.93	0.07 *
Land ownership	−1.16	2.28	0.61
Summary Statistics
−2 Log Likelihood	278.48
Chi-square	34.25	Sig	0.00
Nagelkerke	0.20

* 5% significance level.

The extent of participation was the dependent variable (two social capital networks = low participation, three social capital networks = medium and four+ social capital networks = high). In this study, the extent of participation means the number of social capital networks that a farmer is a member of. It was analyzed using the Ordered Logit Model. Multicollinearity was checked first before running the model using VIFs and tolerance. All the VIFs were less than five and the tolerance was above 0.20, meaning that the independent variables were not dependent on each other.

The Ordered Logit Model was significant at the 1% level (Table 6), with a Nagelkerke of 0.20. Variables such as household size, employment status, and income were significant at the 5% level. Participation in various social capital networks means farmers will obtain information from various sources and this can influence their adoption of climate change adaption strategies.

As shown in

Table 7. Influence of social capital networks on selection of climate change adaptation techniques—Binary Logistic Regression.

	Crop Diversification	Improved Crop Variety	Mixed Farming	Use of Organic Manure	Crop Rotation
Social capital network membership	0.14 (0.44) [0.002] *	0.36 (0.53) [0.000] *	0.09 (0.36) [0.011] **	0.50 (0.60) [0.000] *	0.29 (0.05) [0.000] *
Model summary
Sig	0.002 *	0.000 *	0.011 **	0.000 *	0.000 *
Exp(B)	0.30	0.37	0.25	0.41	0.37
Nagelkerke	0.05	0.18	0.03	0.26	0.13

() standard error, [] p-value, ** 5% significance level, * 1% significance level.

, gender, age, marital status, and education level were statistically insignificant and household size, employment status, and income were statistically significant.

Household size is significant and has positive coefficient as a determinant on the number of social capital networks that a farmer is a member of. An increase in household size correlates with the utilization of various social networks. A plausible reason for this trend might be that households with more members have a surplus of labor. This allows them to distribute tasks efficiently, freeing up time for engagement in diverse social networks. These findings resonate with those of Etim et al. [53], who identified a positive relationship between household size and active participation in multiple social capital networks. Tatlonghari et al. [54] also found a positive relation between large family sizes and utilization of different social capital networks.

Employment status was found to be a significant factor that also affects the number of social capital networks that a farmer is a member of. Employment status is negatively related with the utilization of social networks. This means there is reduced utilization of social networks when smallholder farmers are formally employed. Unemployed farmers are likely to join numerous social capital networks. The results are in consonance with priori expectations. The descriptive results of the study show that most farmers in the area are unemployed and according to ref [47], unemployed farmers usually join social networks for access to credit and input distribution. The negative relationship between employed farmers and participation in different social capital networks can also be associated with a lack of time among employed farmers. Employed farmers are usually busy at work so they will not have time to commit to many social capital networks.

Income was significant and it negatively affected participation in social capital networks. Utilization of social networks is low for smallholder farmers with high monthly income. An increase in income resulted in a decrease in participation in social capital networks. This might be because an increase in income makes a farmer to be able to cover their agricultural expenses, so they do not see the need to join social capital networks which usually help small-scale farmers with informal credit and inputs. High income is also associated with employment, so farmers who are employed can have challenges in balancing work and many social networks. This infers that such farmers prefer to join few social capital networks that they are able to commit to all, especially in terms of time. The results are congruent to those of Kehinde et al. [8], who found that an increase in the income of households decreases the level of participation in social networks because farmers that earn substantial amounts of income are less likely to need external funds.

3.4.3. Influence of Social Capital Networks on Selection of Climate Change Adaptation Techniques

To delve into the influence of social capital networks on the adoption of specific climate change adaptation techniques, the Binary Logit Model was employed (Table 7). The choice of each mitigation strategy was binary. The study identified adaptation strategies such as crop diversification, improved varieties, mixed farming, utilizing organic manure, and changing planting dates. All the models were significant at the 1% and 5% levels and had Nagelkerke values ranging between 0.05 and 0.26, indicating a better fit of the model.

Table 7 shows that membership of a social network positively influences the adoption of crop diversification, organic manure, changing planting dates (1% level), and mixed farming (5% level). The result shows that membership of a social capital network increases the likelihood of implementing crop diversification, mixed farming, organic manure, and changing planting dates. Participation in social capital networks was significant and had a positive coefficient with all adaptation strategies. This indicates that when a smallholder farmer is integrated into a social capital network, there is a heightened likelihood of adopting measures to counteract the impacts of climate change. These findings reinforce the preceding anticipation, suggesting that social capital networks play a pivotal role in promoting the adoption of climate change adaptation strategies. An explanation for these outcomes may be that social capital networks make farmers connect and build trust, which in turn brings about a positive mindset and a change of attitudes in farmers. Another reason for these results may be associated with a reduction in risk aversion when a farmer participates in a social network. The correlation between participation in a social capital network and adoption of climate change adaptation strategies is because social capital networks help farmers share information and farm inputs. Being a member of a social capital network is essential in raising awareness and the dissemination of educative information [55]. Farmers may talk about their difficulties with their peers in a social capital network, where they can receive advice on how to deal with certain issues which could be challenging. Farmers may have easier availability of knowledge and resources if they belong to a group focused on agriculture and may then implement best practices that other farmers may have adopted concerning climate change. These observations align with those of Ogunleye et al. [6] in Nigeria, who identified a positive and notable link between membership in social capital networks and the uptake of climate change adaptation strategies. In China, Ren et al. [56] and in Kenya, Birir [57] also found a positive influence of social capital networks on adoption of climate change adaptation strategies.

4. Conclusions and Recommendations

The purpose of this study was to assess factors that determine participation in social capital networks in order to enhance adoption of climate change in Port St Johns Local Municipality. To achieve this objective, 238 farmers were selected from Port St Johns Local Municipality and data were analyzed descriptively and empirically using the Binary Logit and Ordered Logit Model. The descriptive results of the study revealed that females dominated in the study area, and they participate in social capital networks more than males. The higher participation of females in social capital networks than males could be explained by less access to resources among females. There is minimal participation of youth in agriculture and also in social capital networks. The empirical results found that gender, household size, age, marital status, education, employment status, and income affect a farmer’s participation in social capital networks. When it comes to the extent of participation in social capital networks, household size, employment status, and income were found to influence the number of social capital networks that a farmer is a member of. Social capital networks were found to positively influence the uptake of climate change strategies.

Following the results revealed in this study, there are farmers that do not participate in any social capital networks. Since social capital networks have a positive influence on the adoption of climate change adaptation strategies, the government should educate farmers about the importance of social capital networks and the benefits of acting jointly in the face of climate change events. These findings can also be utilized to create policies that will encourage farmers to join localized farmer-based social capital networks to adopt climate change adaptation measures. Policy interventions should encourage resource-poor farmers to participate in social capital networks because these networks will foster group interactions, improve exchange of information, and improve climate change adaptation. This study found low levels of youth participation in farming and also in social capital networks. This might be due to lack of information; therefore, extension officers should educate farmers about the importance of social networks. The government should also come up with policies that will motivate youth to participate in agriculture. The positive influence of social capital networks on the adoption of climate change adaptation strategies calls for extension services in the study area to create and promote youth social capital networks or to create farming-related programs that will encourage youth to join. Given the significant role of females in social capital networks, this study proposes more gender-specific policies or programs that address the unique challenges and barriers faced by women in agriculture. Tailoring interventions to address gender disparities and enhance women’s access to resources could be impactful.

This study did not focus on other social capital network characteristics such as membership fee and meeting attendance in the area. This study did not investigate factors that influence farmers not to join social capital networks. It is suggested that further research could be conducted to understand why some farmers do not join social capital networks. This paper also identified low levels of youth participation in both farming and social capital networks. Other researchers can delve deeper into the reasons behind this trend and determine specific, targeted strategies to engage younger demographics. These could involve integrating modern technology, social media platforms, and youth-centric approaches to farming and climate change adaptation.