Exploring the impact of drought on farmers’ skills gap and training needs in South Africa

Abstract

The study investigated the impact of drought on farmers’ preparedness and mitigation practices in South Africa's agricultural sector and identified their skills gaps in drought management and associated training needs. Using a multi-stage purposive sampling method, 192 farmers (120 subsistence and 72 small-scale commercial) were interviewed across dry agroecological zones in the Free State, Eastern Cape, Northern Cape, and Western Cape provinces of South Africa. The results indicate that the primary socioeconomic impacts were reduced household food security (reported by 80% of participants) and increased poverty (≥60%). Environmentally, reduced water availability (>80%) and grazing land deterioration (>64%) were the most cited impacts. Key coping strategies included supplementary feeding (71%) and livestock sales (55%). The results of the ordinal regression analysis indicated that age and gender were associated with a lower capacity to cope with drought, whereas education, income source, and access to extension services enhanced farmers’ coping ability (P ≤ 0.05). Additionally, age, education, access to extension services, and agricultural training negatively influenced farmers’ need for drought management training (P ≤ .05). Notably, a widespread lack of skills in drought monitoring, early warning systems, and water management was identified across farming types. The study concluded that targeted government interventions should prioritize adequately resourcing extension services and systematically building farmers’ capacities to manage the escalating frequency and severity of droughts.

Introduction

Drought is a complex and far-reaching natural hazard with implications for both society and the natural environment, contributing to reduced agricultural output, social tensions, and environmental degradation (Bahta and Nyaki, 2025). Globally, drought severity has increased between 1981 and 2025 (Gebrechorkos et al., 2025). Since 1900, these events have become more frequent, resulting in annual economic losses estimated between 3% and 8%. Approximately 37% of the global land surface has been affected by water-related challenges since 1980 [Organisation for Economic Co-operation and Development (OECD), 2025], with drought alone causing estimated annual losses of nearly US$307 billion worldwide (Thomas et al., 2024). In 2022 alone, drought disasters caused approximately US$202 billion in damages and affected 185 million people globally (Wollburg et al., 2024). In Africa, drought-induced economic losses have been estimated at US$2.4 billion, with Southern Africa contributing approximately US$354 million to this total (Bahta, 2021).

The agricultural sector is particularly vulnerable to drought due to its heavy reliance on water availability, which places its long-term sustainability at risk (Nhamo et al., 2025). In South Africa, the sector supports around 930,000 jobs and plays a critical role in ensuring food and nutrition security, as well as fostering rural development (Statistics South Africa, 2025). As the 30th driest country in the world, South Africa has experienced recurrent droughts throughout the 20th century (Pili and Ncube, 2022). The 2015–2016 drought, among the most severe in recent decades, caused an 8.4% decline in agricultural production and a 15% reduction in the national livestock herd (Slayi et al., 2023). With economic losses estimated at US$250 million and affecting 2.7 million households, the event deepened the sector's vulnerability (Zhou et al., 2022). Projected population growth to 79.2 million by 2050, coupled with intensifying climate variability, is expected to place substantial strain on South Africa's agricultural systems, necessitating sustainable drought management to protect food and nutrition security (Strydom and Struweg, 2016; World Health Organization, 2024).

The national government provides strategic oversight of drought management in South Africa's agricultural sector, with execution delegated to provincial and local municipalities (Makaya et al., 2020). Government departments such as the Department of Agriculture (DoA), Department of Water and Sanitation, and the Department of Cooperative Governance and Traditional Affairs (DCGTA) work in collaborations with the National Disaster Management Centre (NDMC) and the South African Joint Committee in drought and disaster management (Liebenberg, 2015; Makaya et al., 2020; Vogel et al., 2016). Within this framework, the NDMC is instrumental in early warning system planning, stakeholder coordination, and research, while DoA leads the dissemination of information, education, and program implementation within farming communities (Liebenberg, 2015).

Extensive research has documented the impact of drought on South Africa's agricultural sector (Bahta and Nyaki, 2025; Mbuqwa et al., 2024; Mzimela and Moyo, 2024; Mzimela et al., 2025), highlighting the critical role of agricultural extension officers and disaster management personnel in building sectoral resilience. However, most of these studies focus on specific farmer groups, either subsistence, small-scale, or commercial, with few investigating the broader agricultural sector's skills gap and training needs in drought preparedness and mitigation. Furthermore, discussions on the impacts of drought have been mostly centered on agricultural losses and food availability, yet the role of extension and farmer training in reducing vulnerability remains underexplored. The current study, therefore, assessed the impacts of drought on farmers’ preparedness and mitigation practices in South Africa's agricultural sector and identified existing skills gaps and associated training needs in drought management.

Materials and methods

Ethical approval

The study received ethical approval from the Stellenbosch University Social, Behavioural and Education Research Committee (REC-2021–19116) in accordance with the South African National Health Act 61 of 2003. All procedures adhered to the regulations of the Research Ethics Committee, including obtaining written informed consent from all participating farmers. To ensure confidentiality, all data were anonymized and the identities of research participants were protected. Gatekeeper permission for accessing farmers was granted by the DoA. Furthermore, as data collection occurred during the COVID-19 pandemic, all protocols outlined in the South African Disaster Management Act No. 27 of 2002 were strictly followed.

Study sites and sampling procedures

Table 1 summarizes the environmental conditions and sample sizes across the surveyed areas. Provinces, districts, and local municipalities were selected using a multi-stage purposive sampling approach based on aridity and the historical recurrence of drought. Aridity was determined using a simple rainfall-based index, with “semi-arid” areas receiving between 250 and 500 mm of annual rainfall and “arid” areas receiving less than 250 mm (Halimani et al., 2021; Intergovernmental Panel on Climate Change, 2007). In this context, the Free State and Eastern Cape provinces were classified as “semi-arid,” whereas the Northern Cape and Western Cape were classified as “arid.”

Table 1.

Climatic conditions and number of respondents in the surveyed areas of the Free State and Cape provinces, South Africa.

						Climatic conditions
Province	District	Municipality	Land area(km2)	GPS coordinates	Ecological zone	Annual rainfall (mm)	Mean annual temperature (OC)	Main agricultural activities	Farmers
Free State	Lejweleputswa	Tswelopele	6 524	28.0223°S, 25.9468°E	Semi-arid	400–600	15–16	Crop production—maize; sunflower	27
	Xhariep	Letsemeng	9 828	29.3246°S, 25.0940°E	Semi-arid	350–500	16–17	Livestock production—sheep; goat; ostrich	18
Northern Cape	Namakwa	Karoo-Hoogland	27 985	31.8438°S, 20.9030°E	Arid	100–200	17–18	Livestock production - sheep	20
	Pixley Ka Seme	Emthanjeni	11 388	30.6551°S, 24.0088°E	Arid	190–260	13–14	Livestock production—sheep; goat; game	27
Western Cape	West Coast	Matzikama	5 372	31.6125°S, 18.5226°E	Arid	30–260	17–18	Horticulture production and aquaculture—vegetables; raisins; grapes	9
	Central Karoo	Beaufort West	21 917	32.3529°S, 22.5841°E	Arid	150–235	17–18	Livestock production—sheep; goat	27
Eastern Cape	Chris Hani	Engcobo	11 046	31.6764°S, 28.0093°E	Semi-arid	300–400	11–14	Subsistence livestock and crop production—cattle; sheep; goat; maize	32
	O.R Tambo	Nyandeni	2 474	31.5697°S, 29.1391°E	Semi-arid	470–550	17–20	Subsistence livestock and crop production—cattle; sheep; goat; maize	32

Sources: https://en.climate-data.org/; Halimani et al. (2021).

Participants were selected from a list of farmers provided by the local DoA extension support services in each targeted community. The list had 622 farmers in the Free State, Eastern Cape, Northern Cape, and Western Cape provinces. Farmers were then randomly selected from this list; however, only those who were willing to participate in the study were interviewed. A total of 192 farmers were interviewed face-to-face between February and May 2022. This comprised 45 in the Free State, 64 in the Eastern Cape, 47 in the Northern Cape, and 36 in the Western Cape. The number of participants was determined using Slovin's formula as follows:

$$n={\displaystyle \frac{N}{1+N{(e)}^2}={\displaystyle \frac{622}{1+622{(0.06)}^2}=192.02\approx 192}}$$

where N = the total record number of farmers in the four provinces (622), n = sample size required, which is 192; e = the acceptable sampling error of 0.06. However, the authors acknowledge that the sampling method may be biased and not a true representation of farmers in the provinces.

Data collection

Trained enumerators administered a pre-tested, semi-structured questionnaire in the farmers' preferred local language (English, IsiXhosa, Sesotho, or Afrikaans) to collect data on socioeconomic profiles, farm characteristics, farmers’ perceptions of drought impacts, their knowledge of drought adaptation skills, and their drought management training needs. A prototype of the questionnaire was pre-tested on nine farmers in the West Coast district and one farmer in the Stellenbosch municipality, Western Cape province. This pre-test assessed whether the questionnaire was correctly completed and whether respondents genuinely understood the questions, and were able and willing to provide the requested information. Based on the findings, the questionnaire was refined by rephrasing and reorganizing certain questions to eliminate ambiguity, sensitivity, and difficulty, while ensuring clarity, logical flow, and appropriateness of instructions. The final version, comprising 67 questions, was administered, with each interview lasting an average of 45 minutes. Participation was voluntary, and respondents could withdraw at any time before submission without providing a reason. All responses were anonymized, and data were kept confidential and aggregated for analysis. In addition to the interviews, farm visits were conducted to observe farmers’ drought adaptation strategies first-hand.

Statistical analyses

All the survey data were analyzed using SAS 9.4 (SAS Institute Inc. Cary, NC, USA). Descriptive statistics generated using PROC FREQ were used to analyze farmers’ socioeconomic characteristics and drought management skills. The general linear model procedure (PROC GLM) was used to analyze the effect of farm type (i.e., subsistence versus small-scale commercial) on livestock flock/ herd sizes. The farmers’ drought management skills needs were ranked using the Kruskal–Wallis test (NPAR1WAY procedure). An ordinal regression model (PROC LOGISTIC) was used to determine the factors influencing a farmer's ability to cope with the impacts of drought and the need for drought management training (i.e., water management, drought monitoring, and early warning systems). The logit model fitted independent variables such as age, gender, education, farmer type, ecological zone, farm size, farming experience, the source of income, access to extension services, and agricultural training (Table 2). The forward selection model option embedded in PROC LOGISTIC (SAS Institute Inc. Cary, NC, USA) was used to select independent variables that would be included in the final model. The model used for farmers’ ability to cope with the impacts of drought or farmers' need for drought management training was as follows:

$$\log \left({\displaystyle \frac{P}{1-P}}\right)={\beta }_0+{\beta }_1{X}_1+{\beta }_2{X}_2+\cdots +{\beta }_n\beta X_n+\epsilon$$

Table 2.

Description of independent variables used to create a logistic regression model for factors influencing the ability of a farmer to cope with the impacts of drought and choice of key drought management training needs in the surveyed areas of the Free State and Cape provinces.

Independent variable	Description of an independent variable
Age (youth vs adults)	Age of the farmer [0 = Young (younger than 40), 1 = Adult, older than 40]. Research shows that older farmers cope better with drought using experience and knowledge (Nyam et al., 2021). Experienced farmers rely on indigenous knowledge (Shikwambana and Malaza, 2022). Thus, older farmers likely demonstrate greater drought coping ability, preferring skill refinement training, while younger farmers seek new technologies
Gender (male vs female)	Gender of the farmer (0 = Male, 1 = Female). Females now have better farming experience and network access (Antwi-Agyei and Stringer, 2021), potentially increasing their influence on drought coping Gender may affect training participation, as women often encounter barriers.
Educational level (no formal education—primary vs secondary—tertiary)	Educational level of the farmer (0 = No formal education to primary education, 1 = Secondary and tertiary education). Lower literacy among farmers increases vulnerability to drought impacts through links to marginalization and poverty (Muroyiwa et al., 2022). Farmers without secondary or tertiary education often misjudge seasonal rainfall (Muthelo et al., 2019). Higher education enhances adaptation, receptiveness to new information, and capacity to implement effective coping strategies.
Type of farmer (subsistence versus small-scale commercial)	Type of farmer (farmer typology) (0 = Subsistence farmer, 1 = Small-scale commercial farmer). Commercial farmers’ greater resources and technologies enable better drought measures than subsistence farmers, who face food insecurity (Muroyiwa et al., 2022). Thus, commercial farming positively influences drought coping.
Ecological zone (arid area vs semi-arid area)	Ecological zone (0 = Semi-arid area, 1 = Arid area). Farmers in arid areas have long periods of drought-associated water scarcity (Muthelo et al., 2019), giving them greater experience that enhances knowledge and coping strategies compared to those in less drought-prone areas (Muroyiwa et al., 2022) Hence, farming in arid areas strongly predicts drought-coping ability.
Farm size (≤100 ha vs >100 ha)	Farm size of the farm (0 = Less than 100 ha, 1 = More than 100 ha). Farm size correlates with economic viability and technology adoption (Mapiye, 2022). Larger farms likely cope better with drought and may need specialized training in areas like crop management and mechanization.
Farming experience (≤10 years vs >10 years)	Farming experience of the farmer (0 = Less than 10 years, 1 = More than 10 years). Farming experience provides practical knowledge and an understanding of adopting mitigation strategies for recurrent droughts (Muroyiwa et al., 2022). Therefore, over 10 years of experience is expected to determine farmers’ drought coping ability.
Source of income (farming vs other incomes)	Source of income (0 = Farming as a source of income, 1 = Other income sources). Access to alternative income sources helps farmers buffer against climatic risks like drought (Mapiye, 2022), so this access is expected to greatly influence their coping ability.
Access to extension services (no vs yes)	Access to Extension services of the farmer (0 = Do not have access to agricultural extension services, 1 = Have access to agricultural extension services). Access to extension services strongly predicts farmers’ technology adoption and coping strategies (Muroyiwa et al., 2022). These services provide crucial information and training, which directly enhances a farmer's capacity to manage drought effectively.
Agricultural training (no vs yes)	Agricultural training (0 = Do not have some agricultural training, 1 = Have some agricultural training). Agricultural training enhances farming skills, increasing the likelihood that farmers will adopt technical measures to cope with drought (Muthelo et al., 2019). Consequently, trained farmers are better equipped to implement effective drought strategies.
Land tenure (no vs yes)	Land tenure (0 = Do not own land, 1 = Have land ownership). Secure land tenure boosts training investment and technology adoption.
Information and communication technologies (no vs yes)	Farmer access to a phone (0 = Do not have access to a phone, 1 = Have access to a phone). Access to mobile phones and the internet enables training and information access.
Availability of labor (no vs yes)	Labor availability (0 = Do not have available labor, 1 = Have available labor). Larger households can more easily participate in training due to shared tasks.

where

P = probability of a farmer's ability to cope with the impacts of drought/ farmer's need for drought management training;

β₀ = intercept;

β₁, β₂, …, β _n  = coefficients of independent variables;

χ₁, χ₂, …, χ _n  = independent variables.

ε = random residual error

For the current farmers’ ability to cope with drought, the empirical specifications of the determinants underlying the binomial logit were as follows:

$$\begin{array}{rl}& Y_1(\mathrm{Ability}\mathrm{to}\mathrm{copy}\mathrm{with}\mathrm{the}\mathrm{impact}\mathrm{of}\mathrm{drought})\\ =& {\beta }_0+{\beta }_1(\mathrm{Age})+{\beta }_2(\mathrm{Gender})+{\beta }_3(Education)\\ & +{\beta }_4(\mathrm{Type}\mathrm{of}\mathrm{farmer})+{\beta }_5(\mathrm{Ecological}\mathrm{zone})\\ & +{\beta }_6(\mathrm{Farmsize})+{\beta }_7(\mathrm{Source}\mathrm{of}\mathrm{income})\\ & +{\beta }_8(\mathrm{Access}\mathrm{to}\mathrm{extension}\mathrm{services})\\ & +{\beta }_{10}(\mathrm{Agricultural}\mathrm{training})\end{array}$$

where ${\beta }_0$ = Intercept and ${\beta }_{1\dots .}{\beta }_{10}$ coefficients.

For farmers’ need for drought management training, the empirical specifications of the determinants underlying the binomial logit were as follows:

$$\begin{array}{rl}& Y_1(\mathrm{Farmer}{\mathrm{s}}^{\mathrm{\prime }}\mathrm{need}\mathrm{for}\mathrm{drought}\mathrm{management}\mathrm{training})\\ & ={\beta }_0+{\beta }_1(\mathrm{Age})+{\beta }_2(\mathrm{Gender})+{\beta }_3(\mathrm{Education})\\ & +{\beta }_4(\mathrm{Type}\mathrm{of}\mathrm{farmer})+{\beta }_5(\mathrm{Ecological}\mathrm{zone})\\ & +{\beta }_6(\mathrm{Farmsize})+{\beta }_7(\mathrm{Source}\mathrm{of}\mathrm{income})\\ & +{\beta }_8(\mathrm{Farming}\mathrm{experience})\\ & +{\beta }_9(\mathrm{Access}\mathrm{to}\mathrm{extension}\mathrm{service})\\ & +{\beta }_{10}(\mathrm{Agricultural}\mathrm{training})+{\beta }_{11}(\mathrm{Access}\mathrm{to}\mathrm{phone})\\ & +{\beta }_{12}(\mathrm{Labor}\mathrm{availability})\\ & +{\beta }_{13}(\mathrm{Land}\mathrm{tenure})+{\beta }_{14}(\mathrm{Farming}\mathrm{full}\mathrm{or}\mathrm{parttime})\end{array}$$

where ${\beta }_0$ = Intercept and ${\beta }_1\cdots {\beta }_{14}$ coefficients.Marginal effects were developed to explain the effects of independent variables in terms of probabilities as follows:

$$x_1,{\displaystyle \frac{1}{N}\sum _{i=1}^N{\displaystyle \frac{\beta .p_{i.}(1-p_i)}{100}.}}$$

The marginal effects measured the effect of predictor variables on the probability of a farmer's ability to cope with the impacts of drought and a farmer's need for drought management training for a unit change in the independent variable. Probability (P) values ≤ .05 were deemed significant.

Results

Farmers’ socioeconomic profile

The socioeconomic profiles of participating farmers revealed notable distinctions between subsistence and small-scale commercial groups. The majority of respondents (62%) were male, comprising 65% of subsistence farmers and 58% of small-scale commercial farmers in the surveyed areas. Age distribution showed that nearly half of subsistence farmers were over 60 years old, compared to one-third of their commercial counterparts. Educational attainment was relatively high, with most subsistence (55%) and small-scale commercial (72%) farmers having completed secondary or post-secondary education. Regarding agricultural training, over 60% of small-scale commercial farmers and nearly half of subsistence farmers had received formal instruction. Land tenure arrangements differed markedly; almost 30% of small-scale commercial farmers owned private land, whereas 48% of subsistence farmers relied on communal land. Farm size also varied substantially, with most subsistence farmers (70%) cultivating less than 50 hectares, while half of the small-scale commercial farmers managed operations exceeding 100 hectares. Approximately 90% of all respondents were full-time farmers, with over 70% possessing more than a decade of farming experience. Labor practices differed, as 55% of small-scale commercial farmers employed paid workers compared to 35% of subsistence farmers. Livestock sales served as the primary income source for 65% of respondents, followed by field crop sales. Notably, one-third of subsistence farmers relied on social grants and pensions as their main income streams.

Farm agricultural activities

Of the small-scale commercial farmers surveyed, half were located in arid areas, with over 75% engaged in livestock production, compared to two-fifths of subsistence farmers in the same region. In semi-arid areas, mixed farming was practiced by 70% of small-scale commercial farmers and 80% of subsistence farmers. Maize was the dominant crop, grown by 30% of small-scale commercial farmers and 45% of subsistence farmers, followed by vegetables (20% and 25%, respectively). Only 2% of small-scale commercial farmers cultivated horticultural crops. Among subsistence farmers, most production (64%) was for home consumption, while 32% was sold and 4% used as livestock feed. In contrast, 67% of small-scale commercial farmers grew crops primarily for sale, with nearly a quarter used as feed.

Livestock production varied between groups, with small-scale commercial farmers and those in arid areas owning significantly larger flocks of sheep and goats. Across both groups, the main reasons for keeping livestock were meat production (80%) and live animal sales (93%). About half of the small-scale commercial farmers raised sheep and goats for fiber. Subsistence farmers additionally kept livestock for ceremonies (40%), social status (15%), and milk (14%). Water sources for crop production were primarily dams and boreholes, used by 11% of small-scale commercial farmers and 2% of subsistence farmers. In arid and semi-arid areas, 7% and 9% of respondents, respectively, relied on dam water for irrigation. For livestock, boreholes were the main drinking water source for 20% of small-scale commercial farmers and 30% of subsistence farmers.

The distribution of livestock breeds varied considerably between farming systems and agro-ecological zones. Among subsistence farmers in arid and semi-arid areas, cattle production was dominated by non-descript crossbreeds, reported by 66% of respondents, followed by Nguni (25%), Bonsmara (6%), and Brahman (3%). In contrast, small-scale commercial farmers in the same regions predominantly raised Bonsmara cattle (58%), with other breeds including non-descript crossbreeds (25%), Afrikaner (10%), Nguni (6%), and Limousine (1%). Sheep breed preferences also reflected environmental adaptation. In arid zones, the hardy Dorper breed was most common, kept by 50% of farmers, while in semi-arid areas, Merino sheep were more prevalent (58% of respondents). Other sheep breeds kept included crossbreeds and Meatmaster, though in smaller proportions. Goat production displayed distinct patterns across regions. In arid areas, non-descript crossbreeds made up the largest share (49%), followed by Boer goats (35%), Angora (12%), and smaller numbers of Saanen (2%) and Alpine (1%). In semi-arid areas, non-descript crossbreeds were also the most common (59%), with Boer goats comprising 28% and Angora 10% of the goat herds.

3.3 Farmers’ perceptions of the impact of drought

Across all farmer types and agroecological zones, the concept of drought was universally understood. Farmers defined drought primarily as a lack of rainfall persisting over a longer period of time (>2 dekads), often accompanied by extremely high temperatures. The data confirms that drought is a pervasive challenge, affecting over 95% of all respondents. Notably, every farmer interviewed had been impacted by the severe, multi-year drought that persisted from 2015 to 2018. Despite its severity, half of the subsistence and small-scale commercial farmers failed to anticipate its onset, leaving a third of them completely unprepared for its arrival. Droughts’ major social impacts were profound and widely reported. Key impacts included diminished household food security (cited by 80% of farmers), increased poverty (60%), and loss of on-farm or off-farm employment (24%). While both groups suffered, subsistence farmers generally reported being more severely affected than their small-scale commercial counterparts. The sole exception was the reported increase in poverty levels, which was slightly higher among small-scale commercial farmers. The majority of subsistence farmers (85%) and small-scale commercial farmers (80%) reported farm income losses. Small-scale commercial farmers reported increased livestock mortality (72%) and incidences of pest and diseases (60%) compared to subsistence farmers (65 and 56%, respectively). Regarding environmental impacts, subsistence farmers reported greater reductions in water availability (87%), grazing land deterioration (70%), vegetation loss (61%), and rising temperatures (60%) than small-scale commercial farmers (74%, 56%, 45%, and 48%, respectively). Table 3 shows the socioeconomic factors that influence farmers’ ability to cope with the impacts of drought. Ordinal regression marginal effects results show that farmers’ ability to cope with drought was influenced by age, gender, education, the source of income, and access to agriculture extension services (P ≤ .05). A decrease in young farmers and females, and an increase in level of education, farming income, and access to agricultural extension services increased the likelihood of coping with the impacts of drought (P ≤ .05).

Table 3.

Socioeconomic factors influencing the ability of farmers to cope with the impacts of drought in the surveyed areas of the Free State and Cape provinces, South Africa.

Variable	Margin	Standard error	t value	P>|t|	[95% conf. interval]
Age	−0.182	3.22	2.20	.028	−0.234	−0.130
Gender	−0.134	2.45	2.12	.034	−0.172	−0.095
Education	0.441	1.36	−11.62	<.0001	0.324	0.558
Type of farmer	−0.009	1.34	0.25	.802	−0.011	−0.006
Ecological zone	−0.105	2.82	1.45	.147	−0.135	−0.075
Farm size	0.011	1.71	−0.24	.807	0.008	0.014
Experience in farming	−0.017	1.01	0.64	.523	−0.021	−0.012
Source of income	0.092	1.70	2.11	.035	0.119	0.066
Access to extension services	0.077	1.51	−1.98	.048	0.055	0.099
Agricultural training	−0.027	1.10	0.95	.343	−0.035	−0.019

Table 4 shows the socioeconomic factors that influence the farmer's need for drought management training. Ordinal regression marginal effects showed that age, education, access to extension services, and agricultural training negatively influenced farmers’ need for drought management training (P ≤ .05). A decrease in the number of young farmers (<40 years), level of education, access to extension services, and agricultural training increased the farmers’ need for drought management training.

Table 4.

Factors influencing farmers’ choice of key drought management training needs in the surveyed areas of the Free State and Cape provinces, South Africa.

Variable	Margin	Standard error	t value	P>|t|	[95% conf. interval]
Age	−4.031	0.69	100.02	<.0001	−4.828	−3.235
Gender	−0.008	0.87	0.16	.875	−0.010	−0.006
Education	−0.924	0.69	22.94	<.0001	−1.107	−0.742
Type of farmer	−0.013	0.88	0.24	.807	−0.015	−0.010
Ecological zone	−0.091	0.83	1.86	.063	−0.108	−0.073
Farm size	−0.020	1.12	0.31	.760	−0.024	−0.016
Experience in farming	0.068	0.79	−1.47	.141	0.055	0.082
Source of Income	0.049	0.79	−1.06	.288	0.040	0.059
Access to extension services	−0.181	0.90	3.44	.001	−0.216	−0.145
Agricultural training	−0.234	1.21	3.33	.001	−0.281	−0.188
Access to phone	0.003	0.81	−0.06	.954	0.002	0.003
Availability of labor	−0.005	0.704	0.13	.896	−0.006	−0.004
Land tenure	−0.076	0.84	1.55	.120	−0.091	−0.061
Full/ part-time farming	0.084	1.05	−1.37	.171	0.067	0.100

Farmers’ agricultural drought management skills

Table 5 presents the agricultural drought management skills reported by farmers. Across all farmer types and agroecological zones, the majority of farmers possessed drought management skills, including supplementary feeding (71% of the respondents), livestock sales (55%), use of fodder banks (44%), and appropriate stocking rates (43%) and rotational grazing (∼40%). They also possessed soil, water, crop, and business management skills (<30% of the respondents).

Table 5.

Percentage of farmers with agricultural drought management skills in the surveyed areas of the Free State and Cape provinces, South Africa.

Drought management skill
	Type of farmer				Overall (n = 192)
	Subsistence (n = 72)	Small-scale commercial (n = 120)	Arid (n = 83)	Semi-arid (n = 109)
Crop management
Planting of drought-resistant varieties of crops	10.4	9.9	10.8	27.5	20.3
Introduction of cultivars with shorter growing seasons	6.8	3.7	7.2	12.8	10.4
Changing planting and harvesting times	12.5	4.7	2.4	28.4	17.2
Crop rotation	13.1	6.3	7.2	71.6	19.3
Crop diversification	15.1	6.3	4.8	33.9	21.4
Introduction of legume cover crops	8.3	4.7	4.8	19.3	13.0
Reducing the area planted	16.2	5.2	6.0	33.0	21.4
Water management
Irrigation technologies and water scheduling	7.3	5.2	13.3	11.9	12.5
Water conservation techniques	20.8	11.5	27.1	35.8	32.3
Rainwater harvesting	13.5	14.6	21.7	33.0	28.1
Drilling boreholes	9.9	8.3	26.5	11.9	18.2
Protection of open water sources	15.1	5.7	19.3	22.0	20.8
Recycling of wastewater	6.8	10.4	9.64	22.9	17.2
Construction of storage tanks	19.2	9.4	27.7	29.4	28.7
Building dams for water storage	9.4	4.7	14.5	13.8	14.1
Soil management
Zero/ minimum tillage	7.8	2.1	3.6	14.7	9.9
Mulching	14.6	6.8	4.8	33.9	21.4
Composting	15.1	7.8	3.6	37.6	22.9
Increased soil cover crop	14.6	6.8	6.0	33.0	21.4
Intercropping	15.6	4.2	2.4	33.0	19.8
Construction of contours and terraces	6.77	1.6	1.2	13.76	8.3
Leaving fields fallow	6.3	2.6	2.4	13.8	8.9
Integrated pest management	6.3	3.7	3.6	14.7	9.9
Construction spreader channels to prevent soil erosion	4.2	0.5	2.4	6.4	4.7
Rangeland management
Use of appropriate stocking rates	28.7	14.1	62.7	27.5	42.7
Creation of paddocks	13.5	7.3	24.1	18.4	20.8
Rotational grazing	20.3	18.8	42.2	36.1	39.1
Night grazing	7.29	2.1	9.6	9.2	9.4
Rangeland assessment	21.9	12.5	45.8	25.7	34.4
Removal of invasive plant species	12.0	6.3	22.9	14.7	18.2
Irrigation of pastures and rangelands	6.3	1.6	10.8	5.5	7.8
Reseeding pastures and rangelands with improved grasses	9.9	4.2	22.9	7.8	14.1
Livestock management
Keeping adapted breeds	25.0	12.5	55.4	23.9	37.5
Use of improved reproduction methods	9.9	7.8	28.9	9.2	17.7
Use of livestock genetic improvement techniques	8.9	5.7	19.3	11.0	14.6
Supplementary feeding	43.8	27.6	77.1	67.0	71.4
Use of fodder banks	27.6	16.7	50.6	39.5	44.3
Selling of livestock	35.4	19.3	73.5	40.4	54.7
Culling less reproductive animals	16.2	9.4	42.2	12.8	25.5
Relocation of livestock to areas with better pastures	8.9	8.9	31.3	7.3	17.7
Water restriction and deprivation techniques	7.8	3.1	18.1	5.5	10.9
Use of water stress alleviators in drinking water	9.4	5.7	26.5	6.4	15.1
Vaccination against diseases and parasites	23.4	13.0	55.4	22.0	36.5
Business management
Water accounting and auditing	5.7	4.2	13.3	7.3	9.9
Creation of an alternative low input system	5.7	4.2	13.3	7.3	9.9
Saved money for emergencies	14.6	12.5	37.4	19.3	27.1
Risk assessment and management	6.8	6.3	20.5	7.3	13.0
Diversifying the farm	7.3	8.9	22.9	11.0	16.2
Insurance	7.3	4.2	15.7	8.3	11.5
Credit and borrowing from financial institutions	7.3	5.7	20.5	7.3	13.0

Farmers’ drought management training needs

Table 6 presents the proportion of farmers who had access to drought-related information, aid, and training. Among subsistence farmers, about 25% had access to drought information, compared to 15% of small-scale commercial farmers. Extension officers were the source of half of this information. The majority of subsistence farmers (74%) relied on indigenous knowledge to cope with the effects of agricultural drought. In terms of social networks, 76% of subsistence farmers and 88% of small-scale commercial farmers belonged to farmer groups, where knowledge and experiences were regularly exchanged. Regarding aid, approximately 30% of subsistence farmers and 15% of small-scale commercial farmers had previously received relief from government or farmer organizations. This assistance came in the form of cash handouts (15%), vouchers (55%), or livestock feed (28%). When it came to training, one-quarter of small-scale commercial farmers and 15% of subsistence farmers had participated in agricultural training over the past decade. However, over 84% of all respondents (across both farmer types) expressed a need for training in agricultural drought management. Regardless of farmer category or agroecological zone, most farmers (75%) preferred workshops or seminars as the mode of training delivery.

Table 6.

Percentage of farmers with access to drought information, aid, and training in the surveyed areas of the Free State and Cape provinces, South Africa.

Variable	Type of farmer		Overall (n = 192)
	Subsistence (n = 120)	Small-scale commercial (n = 72)
Access to drought information
Yes	39.2	41.7	40.1
No	60.8	58.3	59.9
Government extension services	0.8	-	0.5
Farmers’ organizations	4.2	4.2	4.2
Community leaders	5.0	13.9	8.3
Indigenous knowledge	19.2	19.4	19.3
Print/ Visual media	23.3	29.2	25.5
Possess indigenous knowledge
Yes	74.2	33.8	59.0
No	25.8	66.2	41.0
Received drought relief/aid
Yes	46.7	40.3	44.3
No	53.3	59.7	55.7
Source of relief/aid
Government extension services	53.3	0.1	33.4
Farmers’ organizations	1.7	59.7	0.2
Community leaders	0.8	4.2	0.2
Neighbors/Family	44.2	35.0	66.2
Kind of assistance
Advisory	55.8	69.0	59.7
Farming tools	4.2	-	0.0
Medication and vaccines	1.7	-	0.0
Funds	1.7	16.9	0.1
Feed	36.7	14.1	40.2
Access to farmers’ networks
Yes	75.8	87.5	80.2
No	24.2	12.5	19.8
Received agriculture drought-related training in the last 10 years
Yes	15.0	25.0	18.8
No	85.0	75.0	81.8
Need drought management training
Yes	90.8	84.7	88.5
No	9.2	15.3	11.5
Best way to be trained
Workshops/Seminars	75.8	79.2	77.1
Print/Visual media	24.2	20.5	22.9

Table 7 outlines the specific drought management skills farmers identified as necessary. The most critical need, cited by both farmer groups and across all ecological zones, was training in drought monitoring and early warning systems (P ≤ .05). Water management skills were ranked second in importance, also with statistically significant agreement across farmer types and zones (P ≤ .05).

Table 7.

Drought management skills needed by farmers in the surveyed areas of the Free State and Cape provinces, South Africa.

Drought management skill	Rank (mean rank) a
	Subsistence	Small-scale commercial	Significance	Arid	Semi-arid	Significance
Drought monitoring and early warning systems	1 (5.4)	1 (5.3)	*	1 (4.8)	1 (5.7)	*
Vulnerability and risk assessment	3 (12.7)	5 (9.1)	*	6 (10.8)	5 (11.3)	*
Crop production	5 (13.7)	7 (11.4)	*	5 (10.2)	6 (14.2)	*
Water management	2 (9.3)	2 (8.2)	*	2 (8.4)	2 (9.3)	*
Soil management	8 (22.3)	8 (13.2)	*	8 (15.5)	8 (16.7)	*
Rangeland management	4 (13.0)	3 (8.2)	*	4 (10.1)	4 (10.6)	*
Livestock management	7 (14.5)	6 (9.2)	*	3 (9.2)	7 (14.9)	*
Business management	6 (14.1)	4 (8.6)	*	7 (11.2)	3 (9.7)	*

a

The lower the rank, the greater the importance of the skill, *Significance at P ≤ .05.

Discussion

As anticipated, all participants, irrespective of farmer type or agroecological zone, were familiar with the term “drought”. This universal awareness likely stems from the pervasive and well-documented impacts of drought across South Africa's agricultural landscape. Consistent with this familiarity, the majority of participants were able to articulate specific drought effects, corroborating findings by Agri and SA (2020) who noted that most farmers are yet to recover from the impacts of the 2015 drought. However, a notable and unexpected finding was the general lack of anticipation and preparedness for drought onset among farmers. Given the recurrent nature of drought in South Africa, it was reasonable to expect that farmers would have developed and institutionalized coping mechanisms in response to historical occurrences (Katiyatiya et al., 2022; Mdoda et al., 2024). This disconnect between experience and proactive adaptation warrants further investigation into the barriers preventing the internalization of past events into future preparedness.

The disproportionate social challenges reported by subsistence farmers relative to their small-scale commercial counterparts reinforce established narratives regarding the heightened vulnerability of resource-constrained farming communities. These findings align with Ebhuoma et al. (2020), who emphasized the compounded fragility of subsistence-oriented households in the face of climatic shocks. This vulnerability is also illustrated by Lottering et al. (2021), who documented elevated levels of food insecurity, poverty, and malnutrition among subsistence farmers in uMsinga, KwaZulu-Natal, during drought periods. The cascading effects of prolonged droughts extend beyond immediate agricultural loss to encompass demographic shifts. Vetter et al. (2020) observed that persistent drought over the past decade has triggered significant rural-to-urban migration, particularly among young men seeking alternative livelihoods. This trend is further compounded by structural issues such as land access constraints, which, according to Moreda (2023), propel youth toward urban centers in search of viable income opportunities. For small-scale commercial farmers, the financial architecture of their operations introduces additional layers of risk. Mathinya et al. (2022) noted that reliance on credit and loans to sustain productivity renders this group particularly susceptible to debt default during drought years, exacerbating poverty cycles. Income losses (articulated by participants as primary drought impacts) were frequently linked to high livestock mortality and increased pest and disease incidences. These were linked to high livestock losses and the emergence of pests and diseases, which could be potentially attributed to the presence of larger herds and a lack of preparedness among small-scale commercial farmers (Vetter et al., 2020).

The greater climate-related environmental impacts experienced by subsistence farmers relative to small-scale farmers reflect inequitable access to resources, which limits investment in water infrastructure, supplementary feed, and rangeland management. These constraints weaken adaptive capacity and increase exposure to climate-related drought. Overall, the key environmental impacts of drought reported by farmers concurred with existing literature. Lottering et al. (2021) found that 74% of respondents in uMsinga identified water scarcity, rising temperatures, and forest degradation as critical drought-induced challenges. Similarly, Vetter et al. (2020) reported that during the 2016 drought in KwaZulu-Natal, a high percentage of animal deaths were reported in areas with overgrazing and soil erosion. Drought-driven environmental stressors reduce crop, forage, and livestock production, compromising livestock-system resilience, extension-service effectiveness, and farming-household livelihood security (Bahta and Myeki, 2022; Letsoalo et al., 2023). In terms of drought management skills, farmers demonstrated a range of drought management strategies consistent with those reported by Mare et al. (2018) such as the early sale of livestock and the cultivation of fodder crops such as lucerne and maize to serve as fodder banks.

The marginal effects derived from ordinal regression analysis offer critical insights into the socio-demographic determinants of drought coping capacity. Increased probabilities of effective coping were associated with younger farmers, male gender, higher educational attainment, greater farming income, and access to agricultural extension services. These findings are consistent with Lottering et al. (2021), who identified age, gender, education, and income as significant predictors of mitigation adoption. Education, in particular, emerges as a pivotal enabler; Hassan (2020) argued that formal schooling enhances farmers’ receptivity to novel technologies and information, thereby strengthening adaptive capacity. Post-primary education is especially instrumental in facilitating the acquisition and processing of information from extension services and other types of media (Mthombeni et al., 2021). Gender dynamics also play a role; Bahta (2021) reported that male farmers’ participation in social networks such as farmers’ associations enhances resource access and resilience to the impacts of drought. Moreover, the intersection of education and age yields synergistic benefits. Khowa (2021) noted that there is a positive association between education level and age, as educated young farmers easily obtain, comprehend, and use agricultural drought coping strategies. Younger farmers generally exhibit greater adaptability to technological and methodological innovations compared to older farmers (Vetter et al., 2020). Agricultural extension services remain indispensable conduits for disseminating information on productive technologies, credit access, and drought early warning systems (Mulungu and Tschopp, 2025).

Regarding expressed drought management training needs, the negative marginal effect for age suggests that younger farmers were more inclined to identify a need for drought management training compared to older farmers. This finding aligns with Bakawa et al. (2025), Chisadza et al. (2025), and Hassan and Nhemachena (2008), who characterized younger farmers as more innovative and receptive to adopting climate-resilient practices. Conversely, older farmers may rely more heavily on traditional knowledge systems or perceive themselves as adequately equipped, thereby reducing their need for formal training (Hassan and Nhemachena, 2008). Education similarly exhibited a negative marginal effect, indicating that farmers with higher levels of formal education were less likely to express a demand for drought management training (Jagadeesh et al., 2025; Loki and Mdoda, 2025). This may suggest that educated farmers possess greater access to information or alternative adaptation strategies through non-formal channels (Naicker et al., 2025). In contrast, farmers with lower levels of education may depend more heavily on structured training interventions to enhance adaptive capacity (Makamane et al., 2025). The influence of extension access and prior agricultural training further reinforces this pattern; farmers already receiving regular technical support or having participated in capacity-building initiatives perceive themselves as better prepared, thereby expressing reduced demand for additional training (Hassan and Nhemachena, 2008; Makamane et al., 2025). Similar patterns have been reported by Loki and Mdoda (2023), who highlighted the role of prior exposure to extension and capacity building activities in reducing training gaps.

The observation that subsistence farmers possessed greater indigenous knowledge of drought than small-scale commercial farmers is consistent with Mathinya et al. (2022), who attributed this disparity to differential resource endowments. Resource constraints compel subsistence farmers to rely on traditional practices, which, while accessible and affordable, may simultaneously limit the uptake of modern technologies. In uMsinga, KwaZulu-Natal, for instance, indigenous water conservation strategies such as rainwater harvesting and well utilization have been adopted with notable success due to their low cost and local availability (Lottering et al., 2021). Some of these indigenous techniques are being adopted because they are less expensive, easily accessible, and widely available locally, which suits the resource-limited subsistence farmers.

The finding that few farmers had access to drought information, with extension agencies providing half of the information, was not expected. This suggests gaps in information dissemination, highlighting the importance of improved communication needed between extension services and farmers (Makamane et al., 2024). Extension services are mandated to facilitate knowledge transfer on drought, climate variability, and adaptive decision-making, yet their reach appears constrained (Raidimi and Kabiti, 2019). Encouragingly, a high proportion of subsistence farmers demonstrated indigenous coping knowledge, consistent with previous studies (Chisadza et al., 2015; Mfitumukiza et al., 2020). Additionally, substantial membership in farmer networks among both subsistence and small-scale commercial groups indicates the presence of robust peer-to-peer learning platforms. These networks serve as effective channels for disseminating drought-related information and sharing experiential knowledge (Rakgwale and Oguttu, 2020).

Some subsistence and small-scale commercial farmers had previously received government and farmer-organization aid/relief in line with previous studies in the Eastern Cape and Free State provinces, where farmers received different forms of assistance during drought (Ngaka, 2012). However, the turnaround time for the relief has been a constraint to agricultural production (Katiyatiya et al., 2022). The variations in training possessed by the farmers could be attributed to possible skills shortage among extension officers and drought management personnel. Weak linkages between research institutions and extension services, together with infrequent training opportunities, likely impede knowledge and skills development among farmers (Khumalo et al., 2025; Koch and Terblanché, 2013; Raidimi and Kabiti, 2019).

Farmers prioritized drought monitoring and early warning systems as the most critical skills, followed by water management, a preference likely shaped by the prohibitive costs associated with advanced technologies. A study of local early warning systems for drought in Limpopo by Andersson et al. (2020) found that farmers possessed inadequate early warning tools and were unaware of South African Weather Service forecasts, relying instead on indigenous forecast indicators to direct their farming activities. In the same study, farmers were challenged by resource limitations, which affected their ability to respond and, to some extent, adaptive mindsets to the impacts of drought. Advanced drought monitoring technologies, such as remote sensing, remain financially out of reach for the majority of farmers (Khapayi and Celliers, 2016).

The low proportion of subsistence and small-scale commercial farmers who have received agricultural training over the past decade, relative to the high number expressing a need for drought management training, underscores the importance of integrating agricultural extension services with digital climate information platforms and early warning systems. However, early warning information alone, without context-specific advisory support, is often insufficient to inform farm-level decision-making and may therefore have limited practical utility (Mulungu and Tschopp, 2025). Extension officers identified in this study as the primary source of approximately half of the drought-related information accessed by farmers are strategically positioned to bridge this gap by translating climate information into actionable management practices. In addition, the high participation of farmers in farmer groups highlights the potential of peer-to-peer learning networks as complementary dissemination pathways, facilitating locally relevant knowledge exchange and enhancing the uptake of drought preparedness strategies.

Skills related to crop, livestock, and soil management were ranked lower by participants, likely reflecting a continued reliance on indigenous knowledge systems. These systems offer locally adapted, cost-effective, and straightforward solutions that align closely with farmers’ immediate environmental and socioeconomic contexts (Lottering et al., 2021; Yeleliere et al., 2022). Farmers with strong indigenous knowledge and water management competencies are better positioned to autonomously adapt to variable rainfall patterns and seasonal shifts by modifying planting dates, selecting appropriate crop varieties, and diversifying crop combinations (Nhemachena et al., 2020 ).

Conclusion

Drought resilience among South African farmers remains constrained by more than uneven adoption of adaptive practices. The evidence points to persistent competency, information, and institutional deficits that limit farmers’ capacity to anticipate, prepare for, and respond to drought. Across production systems and agroecological zones, farmers identified competency gaps in drought monitoring, early-warning interpretation, water conservation, farm business planning, and crop and rangeland management. These deficiencies are compounded by limited formal training, variable access to credible climate information, and weak engagement with extension services. Strengthening resilience, therefore, requires a policy shift from fragmented technical support to integrated, demand-driven extension systems that translate climate information into timely, actionable farm decisions. Extension officers should be resourced and mandated to act as intermediaries between seasonal forecasts, drought alerts, and locally adapted advisory services. Policy should prioritize farmers with limited education, younger producers, and groups historically underserved by extension, whose restricted access to information and support may deepen vulnerability. Training investments should focus on practical use of early warning systems, drought monitoring, and water management, delivered through workshops, seminars, demonstration sites, and farmer learning networks. Embedding South African Weather Service platforms within provincial extension frameworks could improve the credibility, timeliness, and usability of advisories. Future research should evaluate integrated early warning-extension models and assess context-specific communication tools capable of reaching remote and drought-prone farming communities with timely, actionable guidance.