Adaptation strategies of smallholder farmers to climate variability and change in Konso, Ethiopia

Abstract

This study investigates the adaptation strategies employed by smallholder farmers in Konso, Ethiopia, to cope with climate change and variability. A survey of 355 smallholder households revealed that farmers utilize various indigenous techniques, including terracing, agroforestry, intercropping, and traditional irrigation practices, to adapt to climate change. The adoption of these strategies is influenced by factors such as education level, landholding size, income level, access to climate information, credit, and extension services. Despite these efforts, smallholder farmers in Konso continue to face challenges in ensuring food security due to recurrent droughts and unpredictable rainfall patterns. To enhance the sustainability of indigenous farming systems in Konso, this study recommends providing regular weather information, accessible credit services, crop insurance options, and capacity building through extension services, with a focus on inclusive approaches that cater to uneducated farmers.

Introduction

Climate change presents significant challenges to agriculture in the Sub-Saharan African region¹. Smallholder farmers are adopting various adaptation strategies to mitigate the impacts of climate change. Rain-fed agricultural areas, which account for approximately 95% of the region's food production, are particularly vulnerable to droughts and unpredictable rainfall patterns². Therefore, it is crucial to transform the African agricultural sector, which is largely dominated by smallholder farmers, to better withstand the challenges posed by climate change. Agriculture plays a pivotal role in Ethiopia's economy, contributing 42% of the country's GDP and providing employment for 80% of the population². Despite its importance, the agricultural sector in Ethiopia is heavily reliant on rainfed practices and is predominantly composed of smallholder farmers, who account for more than 90% of the country’s annual production³.

The livelihoods of millions of smallholder farmers in Ethiopia are significantly impacted by frequent droughts and unpredictable rainfall patterns⁴. Since the 1980s, Ethiopia has experienced approximately 16 major national drought events. More recently, in 2015 and 2017, approximately 10 million and 5 million people, respectively, were affected by severe drought conditions⁵. Furthermore, soil erosion has led to the degradation of 50% of Ethiopia’s highlands and has caused an approximate two percent reduction in annual land productivity⁶. As a result, famine and food shortages have become recurring challenges over the past few decades⁷. The unpredictability of rainfall and the increasing frequency of drought events are projected to persist in the future⁸. Therefore, enhancing the implementation of adaptation strategies is crucial for addressing both current and future climate change-related challenges. This approach will help mitigate the adverse effects on the livelihoods of smallholder farmers and improve their resilience to climate change and variability.

An increasing body of literature emphasizes the crucial role of adaptation strategies in mitigating climate-related challenges and enhancing the livelihoods of smallholder farmers. Numerous studies conducted in Ethiopia have identified a positive correlation between the adoption of adaptation strategies and improved crop productivity^{4,9,10,11,12,13}. For instance, in northwest Ethiopia, Adego et al.⁴ observed a 2.79 quintal/year increase in maize productivity attributed to the implementation of adaptation strategies. Similarly, Asrat and Simane¹⁴ reported that smallholder farmers employing adaptation strategies achieved a 24.1% greater yield than nonusers in the Dabus watershed. Furthermore, several studies have indicated that smallholder farmers who adopt adaptation strategies are less vulnerable to climate change impacts than are those who do not^9,10,11,15. In Ethiopia, adaptation strategies are increasingly being integrated into watershed management approaches². These strategies, which include intercropping, agroforestry, crop rotation, and the use of diverse crop varieties, not only enhance crop productivity and resilience to climate change but also contribute to reducing greenhouse gas emissions^3,16,17.

Despite the potential benefits of adaptation strategies in enhancing resilience, increasing productivity, and contributing to mitigation efforts, the adoption of these strategies by Ethiopian smallholder farmers remains low to moderate³. The adoption of adaptation strategies by smallholders is influenced by a combination of institutional, socioeconomic, and biophysical factors^{8,9,10,13,18,19}. Institutional factors, such as access to credit facilities, extension services, and climate information, play a pivotal role in facilitating the adoption of adaptation strategies. Extension agents contribute to the adoption process by disseminating knowledge and raising awareness about effective adaptation practices¹⁰. Additionally, access to financial capital is essential for alleviating the financial constraints faced by smallholder farmers, thereby promoting the adoption of adaptation strategies²⁰. Studies conducted by Asfaw et al.²¹ and Nhemachena et al.²² have highlighted the positive association between access to credit services and the adoption of adaptation strategies, as credit facilities can help mitigate cash constraints. Conversely, research conducted by Marie et al.¹⁸ and Sertse et al.¹⁹ identified the lack of access to credit services and financial resources as significant barriers hindering smallholders' willingness to adopt climate change adaptation strategies in Ethiopia. Furthermore, access to accurate climate and technological information is crucial for reducing climate-related risks and informing effective adaptation strategies^23,24. Recent studies have indicated that limited information regarding climate variability and inadequate knowledge about suitable adaptation options are major obstacles impeding smallholders' preparedness to adopt adaptation strategies^18,19,25.

Socioeconomic factors significantly influence the adoption of adaptation strategies among smallholder farmers. Previous research has indicated that higher educational attainment is positively correlated with smallholders' decisions to adopt climate change adaptation measures, suggesting that educated farmers are more aware of the risks associated with climate variability^14,15. Apart from educational attainment, empirical studies have shown that households with a greater number of economically active members are more likely to adopt labor-intensive agricultural technologies and engage in adaptation strategies^9,11,21,26. Furthermore, age plays a role in the adoption of adaptation strategies, as older farmers tend to have more experience in farming and a better understanding of past climatic conditions, increasing their inclination to adopt adaptive measures^9,15. Additionally, landholding size has been identified as a significant determinant influencing the adoption of adaptation strategies, with larger landholdings being associated with a greater likelihood of adopting adaptive practices^14,15. In Ethiopia, smallholder farmers often engage in both on-farm and off-farm economic activities. Involvement in off-farm activities has been shown to positively influence households' decisions to adopt adaptation strategies⁹. According to Sani et al.¹¹, having access to off-farm income opportunities improves smallholders' financial capabilities, enabling them to invest in essential farm inputs such as seeds and fertilizers.

Adaptation strategies and climate risk management interventions should be tailored to address localized vulnerabilities and impacts on smallholder farmers. However, previous studies analyzing the adoption of adaptation strategies in Ethiopia have predominantly relied on national-level surveys^2,3,16. The aggregated findings from these studies offer limited insights into the adoption of adaptation strategies at the household level in specific geographic locations. Furthermore, most existing studies on smallholders' adoption of adaptation strategies have been conducted in the highlands of Ethiopia, where climate shocks are generally less severe than in semiarid regions^8,9,10,11,13. The Konso region presents a unique opportunity for studying the impact of adaptation strategy adoption on agriculture, given its position in the semiarid region and its culturally integrated conservation practices. Despite the long-standing use of sophisticated soil and water conservation strategies by Konso smallholder farmers, which have been sustained for centuries, their vulnerability to the adverse effects of climate change is increasing due to recurrent droughts, unpredictable rainfall, and rising temperatures. However, there is a paucity of research focusing on adaptation strategies in Konso. Therefore, this study aimed to investigate the adaptation strategies employed by smallholder households in Konso. Specifically, this paper aims to (i) assess the adaptation strategies adopted by smallholders across various agroecologies, (ii) identify the determinants influencing the adoption of these strategies, and (iii) explore the impacts of adaptation strategies on crop productivity.

Materials and methods

Study area

The Konso area is an arid expanse characterized by stone-walled terraces located in the Rift Valley section of Ethiopia²⁷. This cultural landscape stands as a testament to human ingenuity and resilience, showcasing the remarkable engineering feats achieved by communities striving to thrive in a challenging environment. The origins of these intricate "cultural landscape sculptures" can be traced back more than 400 years, reflecting centuries of continuous cultural strategies devised to adapt to and mitigate the harsh environmental conditions for human habitation. Recognizing its cultural significance and unique heritage, UNESCO designated the Konso Cultural Landscape as a World Cultural Heritage Site in 2011. The landscape spans an area between 5° 12′–5° 40′ N and 37° 01′–37° 43′ E (Fig. 1), covering a total area of 2273.38 km².

Figure 1

Source: The map was generated based on CSA data of 2021 (https://data.humdata.org/dataset/ethiopia-cod-ab?) using ArcGIS Version 10.5.

Location map of the study area.

Topographically, the Konso region is characterized by rugged terrain with varying elevations, predominantly featuring hilly and mountainous landscapes interspersed with ravines and valleys. According to the Ethiopian agroecological zone (AEZ) classifications, approximately 80% of the Konso land falls within the semiarid (Kolla) agroecological zone (500–1500 m), while the remaining 20% lies within the subhumid (Woynadega) agroecological zone (1500–2300 m). The primary rainy season in Konso is known as spring (belg)²⁸. Konso has an average annual temperature of 23 °C and receives an average annual rainfall of 801.96 mm. According to the CSA²⁹ census report, the Konso region is home to a total population of 277,964 people residing in 55,673 households. The community's livelihood predominantly relies on livestock and crop production, with sorghum serving as a staple food crop.

Sampling and data sources

The study employed a multistage sampling technique to gather the data. Initially, I adhered to traditional agroecological zoning and identified two distinct zones: the Kolla and Woynadega AEZs. In the second stage, the lowest administrative units (kebeles) were randomly selected from each agroecological zone. The sample size was then determined proportionately based on the number of households within each zone. Subsequently, I utilized a systematic random sampling method to select 355 households from lists provided by the Kebele administration. The household questionnaire surveys were administered to the selected household heads to gather information on socioeconomic characteristics, the adoption of adaptation strategies, and the challenges hindering the adoption of these strategies. To validate the household survey findings and gain insights into the current state of adaptation strategies, soil and water conservation challenges, and proposed solutions, key informant interviews were conducted with development agents from the kebele (lower administrative division), as well as district and zone heads of agriculture offices. Additionally, data from focus group discussions (FGDs) and direct field observations were employed to supplement the information obtained from the household surveys. A total of 6 FGDs were conducted, each comprising 6–10 participants. The FGD participants included elderly men and women from the community, youth aged between 18 and 30 years, and local leaders. The first author and his assistants acted as organizers, moderators, and note-takers for each FGD session.

Data analysis

To assess the statistical significance of differences between groups, I utilized a chi-square test. Furthermore, the study employed employed a multinomial logit model (MNL) to investigate the factors influencing the adoption of adaptation strategies. A concise specification of the model is provided below:

Multinomial logit model specification

In this study, the study employed employed the MNL model to analyze the determinants influencing the adoption of adaptation strategies. This model is particularly suitable for this study data analysis, as it accommodates situations where individuals choose one option from more than two categories based on the option that offers the highest utility³⁰. The MNL model allows for the examination of decisions across multiple categories by determining the choice probabilities associated with each category³¹. To define the MNL model, let Y be a random variable taking on values from 1 to J, where J is a positive integer and let X represent a set of explanatory variables⁸. In this context, Y serves as the dependent variable, representing the adoption of specific adaptation strategies from a set of potential measures, while X denotes the explanatory variables influencing the selection of these adaptation strategies. The associated probabilities for each category are denoted as P1, P2, …, Pj, with the constraint that their sum equals one (P1 + P2 + … + Pj = 1). This relationship elucidates how variations in X influence the response probabilities, P(y = j/x), where j ranges from 1 to J. Once the probabilities for j = 2 to J are determined, the probability P(y = j/x) can be calculated, ensuring that the total probabilities sum to one.

$$P\left(y=\text{1/x}\right)=1-\left({P}_{2}+{P}_{3}+\text{...}{P}_{j}\right)$$

(1)

In this study, the MNL model was estimated by normalizing one category, referred to as the "base category". Of the eight adaptation strategies examined in the present study, the base category was defined as "no adoption of adaptation strategies". The theoretical premise underlying the model posits that, in all instances, the estimated coefficient should be compared to the base group³². For an outcome variable with j categories, the generalized form of the probabilities can be expressed as follows:

$${\text{Pr}}\left({y}_{i}=\text{j/x}\right)=\frac{{\text{exp}}\left({x}{\prime}{\beta }_{{j}{\prime}}\right)}{1+{\sum }_{k=1}^{j}{\text{exp}}\left({x}{\prime}{\beta }_{{j}{\prime}}\right)}\text{, j}=\text{1, 2...J for j}>{1}$$

(2)

While the parameter estimates from the MNL model indicate the direction of the independent variables' effect on the dependent variable, they do not quantify the magnitude of the change. To address this, differentiating Eq. (2) with respect to the explanatory variables allows us to derive the marginal effects of these variables, which are expressed as follows:

$$\frac{\partial {p}_{j}}{\partial {\text{xk}}}={p}_{j}\left({\beta }_{\text{jk}}={\sum }_{k=0}^{j}{p}_{j}{\beta }_{\text{jk}}\right)$$

(3)

Multicollinearity can significantly impact the parameter estimates of the logit model. Therefore, it is crucial to test for the variance inflation factor (VIF) to ascertain the presence of multicollinearity among continuous explanatory variables. The VIF is calculated using the following formula:

$${\text{VIF}}=\frac{1}{1-{\text{Ri}}^{2}}$$

(4)

where Ri² is the squared multiple correlation coefficient between X_i and the other independent variables.

Ethics approval

All methods in this study were carried out in accordance with the ethical standards, relevant guidelines, and regulations of Hawassa University. The researchers obtained ethical approval from the College of Social Sciences and Humanities Ethical Review Committee at Hawassa University.

Consnt to participate

Before collecting data, informed verbal consent was obtained from the respondents of the questionnaire, Focus Group Discussion (FGD) participants, and Key Informants (KIs).

Results and discussion

Smalholder farmer adoption of adaptation strategies

Smallholder farmers in Konso have been utilizing a range of indigenous adaptation strategies, including terracing, agroforestry, organic manure application, diverse crop varieties, soil bunds, intercropping, reduced tillage, and irrigation, to mitigate the effects of climate change and variability.

Agroecology-based adaptation to climate variability and change

The adoption of adaptation strategies by smallholder farmers significantly differed between the subhumid (Woynadega) and semiarid (Kolla) agroecologies (p < 0.01). As illustrated in Table 1, adoption rates ranged from 15 to 59% among smallholders in Woynadega and from 24.3 to 96.6% in Kolla. The higher adoption rate of Kolla agroecology suggested that farmers perceive more significant impacts from climate variability and change than do those in Woynadega. This heightened perception likely stems from Kolla's inherent susceptibility to climate variability and the community's adaptive experiences. Konso smallholders are renowned for their innovative terracing systems, a primary adaptation strategy to combat the effects of climate variability and change. Descriptive analysis revealed that 82.8% of the smallholders in Ko Lla had adopted stone-walled terraces, whereas only 30.7% of those in Woynadega had adopted such terraces. This disparity may be attributed to the greater moisture stress in Kolla than in Woynadega. Faced with unreliable rainfall and recurrent droughts, Konso farmers have ingeniously transformed their challenging terrain into a landscape adorned with indigenous stone-walled terraces and channels, effectively conserving rainwater and safeguarding against soil erosion (Fig. 2b).

Table 1 Smallholders’ adoption of adaptation strategies across agroecologies.

Figure 2

(a) Moringa stenopetala (b) terraced landscape (c) microbasins (d) fruit trees.

Locally, the terrace walls are referred to as 'kawwata', while the terraces themselves are known as 'xeeranta'. These terracing practices enable farmers to cultivate even steep slopes while preventing erosion and retaining soil moisture. The terraces, often reaching heights of up to 6 m, vary in width based on the slope's gradient. The construction of these terraces typically begins with digging trenches for the foundation, with wall foundations embedded approximately 10–25 cm into the ground for stability. The walls are meticulously crafted by interlocking large and small stones, with gaps filled using soil as a binding agent. The terrace walls typically rise 20–40 cm above the retained field level to prevent rapid runoff that could damage the walls³³. Farmers in both agroecologies recognize that terracing mitigates soil erosion and optimizes rainwater infiltration and retention, thereby maintaining soil moisture. Focus group discussions (FGDs) and insights from key informants in both areas affirm that terracing is crucial for preserving soil moisture and reducing the risk of crop failure due to moisture stress and drought. In a similar vein, GebreMichael³⁴ noted in Konso that terraces serve multiple purposes, including slope modification, water harvesting, removal of stones from fields, provision of space for drought-resistant fodder cultivation, and functioning as a farm plot fence and field boundary.

In addition to terracing, Konso farmers integrate other water harvesting and control structures crafted from local stone and available materials. To optimize rainwater capture and infiltration, microbasins, locally known as 'Kolba' or 'Korayita', were constructed on the terraces (Fig. 2c). These microbasins collect rainwater, allowing it to infiltrate to the desired depth. Table 1 reveals that half (50.2%) of the smallholders in Kolla had adopted traditional irrigation, whereas 43.2% of those in Woynadega had adopted traditional irrigation. Small ridges and terrace walls are strategically built to divert rainwater from roads, pathways, and streams to fields. Farmers utilize small stone canals and mud banks to channel runoff from flooding and minor springs, thereby preserving soil moisture. Even the most minor springs of water are captured and redirected to the fields²⁸. Terrace walls typically feature spillways approximately 30–35 cm wide positioned slightly above the field level to facilitate downhill water flow from one terrace to another. These spillways, known as 'tehota,' are drainage outlets resembling windows made of flat stones placed atop terrace walls. They serve to prevent stormwater damage and enhance irrigation on lower terraces. Another essential water control structure employed by Konso farmers is the flood protection wall, locally termed the 'toomota.' Constructed from massive boulders along stream banks, these flood protection walls shield surrounding fields from potential flooding.

In the realm of agroforestry, a significantly greater proportion of smallholders in Kolla (96.6%) than in Woynadega (45.5%) have adopted this practice, as indicated in Table 1. In Konso, traditional agroforestry practices predominantly involve cultivating multipurpose trees on terraced fields, which play a pivotal role in adapting to climate variability and change. Smallholders from both Woynadega and Kolla agroecologies recognize the multifaceted benefits of agroforestry. These authors highlighted its role in providing food, diversifying income streams, reducing soil erosion, enriching soil organic matter, and serving as a valuable source of livestock fodder, local construction materials, and fuelwood. Consequently, the incorporation of trees into farming systems is experiencing a surge. Some farmers are even transitioning croplands into dedicated tree plots, driven by the myriad of ecological benefits and income diversification opportunities that agroforestry offers. GebreMichael³⁵ observed that certain smallholders in Konso have shifted from cultivating crops to establishing multipurpose tree plots due to a combination of three key factors: (1) the need to adapt to recurrent droughts; (2) the labor-saving nature of tree cultivation, which allows household labor to engage in supplementary activities such as daily wage labor in urban areas, handicrafts, and small-scale trading in villages; and (3) the lucrative demand and pricing of timber in local markets. The literature corroborates these findings, suggesting that agroforestry not only diversifies income sources but also acts as a buffer against economic shocks while delivering crucial ecological services^2,16,36.

The landscape of the study area is characterized by the presence of scattered native tree species across farmlands. Among these, the multipurpose trees Moringa stenopetala (locally known as 'shelaqta') and Terminalia brownii ('oypatta') stand out due to their significant socioeconomic importance. These three species were commonly observed across farms in both Woynadega and Kolla agroecosystems (Fig. 2a). Moringa stenopetala is predominantly cultivated around residential compounds within villages to ensure easy access, while also being dispersed across farmlands. Typically, each household compound in the village hosts 10–25 Moringa stenopetala trees. The dense presence of these trees within villages imbues the landscape with vibrant green hues, even within the semiarid terrains of Konso (Fig. 3a). This tree holds special significance because its fresh leaves are commonly consumed as cooked vegetables and feature prominently in daily meals. Moreover, the leaves are a sought-after commodity in local markets, serving purposes ranging from medicinal use to water purification³⁴. Its remarkable drought tolerance ensures year-round vegetable availability. However, periodic pest infestations ('setayta') during the wet season pose a significant challenge to the provision of sustained service.

Figure 3

(a) Moringa stenopetala within a traditional village (b) agroforestry site (c) riverbank farmland ('yeela') (d) sorghum intercropped with other crops.

Terminalia brownii is another prevalent tree species found either cultivated or growing naturally on farmlands. It serves multiple purposes, including livestock fodder, local construction material, and fuelwood. Additionally, various shrubs and tree fodder species, such as Acacia asak, Balanites aegyptiaca, Berchemia discolor, Cajanus cajan, Cordia africana, Ehretia cymosa, Ficus, and Rhus natalensis, are commonly utilized. Furthermore, fruit trees such as mango (Fig. 2d), avocado, lemon, papaya, and banana plants are cultivated, particularly on irrigated farms. These fruits are primarily consumed by households but can also be sold for additional income. Additionally, stimulant crops such as coffee, heat, and tobacco are cultivated on select farms within the Woynadega agroecology.

In terms of intercropping, the findings reveal a statistically significant difference in adoption of these two agroecologies (χ² = 27.13, p < 0.01). Approximately 40% of the smallholder farmers in Woynadega practice intercropping, whereas only 24.3% in Kolla engage in this agricultural practice. The greater adoption rate of intercropping in Woynadega could be attributed to land scarcity and its greater suitability for diverse crops compared to Kolla. To mitigate the risk of crop failure due to drought, Konso farmers employ intercropping by integrating trees with food crops such as cereals, pulses, and root crops, complemented by various management strategies. Certain crops, such as sorghum, exhibit drought tolerance, while others, such as Haricot beans, have shorter maturation periods. To protect ripening sorghum plants from wind damage, farmers often tie their heads together (Fig. 3d)²⁸. Intercropping sorghum with other crops not only minimizes disease and pest risks but also facilitates nutrient and moisture uptake from different soil layers. In instances where sorghum faces pest infestations, farmers can rely on pigeon peas and haricot beans, and vice versa. Cassava, a tuberous crop, serves as a resilient backup among root crops. In the case of crop failures associated with sorghum and other crops, tubers, which were grown during periods of ample rainfall in previous years, remain preserved underground and can be harvested during periods of food scarcity. When not needed for household consumption, cassava can be sold as a cash crop. Consistent with earlier research, the findings underscore the protective benefits of intercropping against crop damage caused by diseases, pests, and droughts^3,16.

In terms of adopting new crop varieties, Table 1 reveals a notable disparity between the two agroecologies. A majority of smallholders in Kolla (74.2%) have adopted improved crop varieties, whereas only a small fraction in Woynadega (14.8%) have done so. The introduction of these new crop varieties in Konso aimed to bolster resilience against the impacts of climate variability and change. Farmers appreciate the early maturation of these new varieties; however, they also voice concerns about their reduced resistance to pests and susceptibility to bird predation. Key informants and participants in FGDs highlighted that, compared to traditional crop varieties cultivated three decades ago, contemporary crop varieties, such as sorghum, are more appealing to birds. Consequently, farmers expend additional labor to prevent birds from damaging these crops.

Terraced fields cultivated with a mix of various crop species are enriched with manure sourced from livestock, including cattle, goats, and sheep, to sustain soil fertility. Table 1 highlights the significant difference in the adoption of manure application between the two agroecologies (χ² = 12.1, p < 0.01). In the study region, livestock are either grazed on pastures in Kolla or tethered in uncultivated areas adjacent to the village periphery. To safeguard crops, tethering and stall feeding are prevalent practices that serve dual purposes: generating income and procuring manure, thereby aiding in adaptation to the impacts of climate variability and change. Livestock are primarily fed sorghum and maize stalks or leaves from trees grown on farms, partly for this specific purpose. Additionally, failed or thinned crops are repurposed as livestock feed or sold as fodder. While the primary objective behind livestock rearing is income generation and resilience against droughts and crop failures, the resulting manure is instrumental in preserving soil fertility. Despite its recognized benefits, manure application is not uniformly practiced across all farm plots. Plots located closer to homesteads tend to receive more manure than those situated farther from homesteads. In contrast, plots located farthest from homesteads often remain untreated. Notably, manure is seldom applied to riverbank farmlands ('yeela') (Fig. 3c). These lands are believed to naturally benefit from organic matter and fertile soil carried by irrigation canals from other areas. Generally, the propensity to apply manure diminishes as the distance between the plot and the homestead increases.

Crop productivity

In addition to mitigating the impacts of climate variability and change, indigenous Konso smallholders employ adaptation strategies to bolster crop productivity. Smallholders across both agroecological zones (AEZs) emphasized that terraces play a pivotal role in reducing soil erosion and preserving soil moisture, thereby enhancing crop yields (Fig. 3c). Consistent with these findings, prior studies conducted in Konso have also highlighted the positive correlation between terracing and increased crop yields^28,37,38. Similarly, Di Falco et al.³⁹ observed a beneficial impact of terracing on crop productivity in the north-central highlands of Ethiopia. Apart from terracing, smallholders employ irrigation techniques to maintain soil moisture levels and boost crop productivity. Whenever feasible, the terraced fields are irrigated. Smallholders attest that irrigation not only elevates yields but also facilitates year-round cropping. Nonetheless, the availability of perennial streams suitable for irrigation remains limited. Within the terraced landscapes of Konso, indigenous agroforestry practices aimed at augmenting crop productivity are prevalent (Fig. 3b). Traditionally, Konso smallholders intermingle indigenous trees with food crops to capitalize on their multifaceted benefits³⁸. As a result, Konso's terraced farms typically feature a diverse array of multipurpose trees capable of providing food and enhancing crop yields.

The terraced farmlands of Konso, delineated by stone walls, showcase a rich mosaic of intercropped crops and trees. Konso's smallholders are adept at intensive agriculture, primarily cultivating cereals such as maize and sorghum, which are interspersed with a plethora of other crops. These include pulses such as chickpea, cowpea, pigeon pea, haricot bean, common bean, hyacinth bean, horse bean, mung beans, and lentils; oilseeds such as sunflower, linseed, and castor; and tubers such as cassava, taro, yam, sweet potato, and potato. On average, each plot accommodates a diverse assortment of 10–20 crop species encompassing cereals, pulses, oilseeds, tubers, and vegetables. Interestingly, crop diversity tends to be more pronounced in Woynadega, with 14–20 species commonly observed, particularly in proximity to homesteads. In contrast, Kolla exhibited slightly lower diversity, typically ranging between 10 and 13 species. GebreMichael³⁴ observed that Konso's agricultural practices involve the simultaneous and sequential sowing of a minimum of 7 different crops through a double-cropping system across two rainy seasons. This substantial crop diversity per plot signifies the promising potential for sustainable yields. Echoing this sentiment, Watson²⁸ portrayed Konso's fields as a 'riot of different crops', with each farmer striving to cultivate staple foods complemented by a selection of other useful or delectable additions.

In addition to the traditional early crops cultivated in Konso, smallholders have embraced newly introduced, improved crop varieties that exhibit rapid maturation under dry, rainy conditions, aiming to optimize crop productivity. This practice coincides with findings of Tadesse³⁸, who emphasized the significance of enhancing agricultural productivity in dryland agroecologies through the adoption of improved crop varieties. Such adoption not only enables these regions to better mitigate natural risks but also fosters sustainable increases in crop yields. Supporting this viewpoint, research by CIAT and USAID³ underscored the importance of adopting improved crop varieties to boost yields and mitigate yield losses attributable to pests and diseases. Overall, Konso's smallholders demonstrate remarkable skill in integrating a diverse array of cereals, pulses, root crops, and tree species on terraced farmlands enriched by manure application. Their adeptness lies in striking a delicate balance among these crops. This balance ensures adequate subsistence food production during challenging years while simultaneously maximizing cash crop yields during favorable years.

Konso's indigenous adaptation strategies serve multiple purposes, not only by enhancing crop productivity and facilitating adaptation to climate variability and change but also by mitigating greenhouse gas emissions. For example, terraces that retain soil moisture foster the growth of diverse crops and tree species, which effectively capture and store carbon. Furthermore, the practice of intercropping, coupled with the adoption of improved crop varieties, diminishes the reliance on chemical inputs and pesticides, thereby reducing their atmospheric emissions. This is corroborated by findings from CIAT and USAID³, which underscored the benefits of adopting improved crop varieties in minimizing chemical input and pesticide use. A study by Abegaz et al.¹⁷ conducted across Ethiopia's highlands further emphasized the role of improved crop varieties in soil organic carbon sequestration. Additionally, the application of manure as fertilizer not only enriches the soil but also curtails methane emissions and minimizes the use and emission of nitrous oxide from inorganic fertilizers. Previous research has similarly highlighted the methane-reducing benefits of manure-based fertilization^2,36. While Konso's indigenous adaptation strategies demonstrably reduce emissions, quantifying the extent of this reduction was not within the purview of this study and warrants further, more detailed investigation. In essence, these indigenous adaptation strategies in Konso are intricately interlinked and mutually reinforcing, serving a multitude of purposes. The synergistic intercropping of diverse crop and tree species, combined with various indigenous farming practices, is anticipated to play a pivotal role in ensuring food and fodder security in the Konso region. From an agroecological perspective, the prevalence of diverse indigenous farming strategies is more pronounced in Woynadega than in Kolla owing to its higher level of intensification, adaptability to a range of tree and crop varieties, and reduced susceptibility to the impacts of climate variability and change.

Determinants of climate change adaptation strategy adoption

In the examination of the determinants influencing the adoption of adaptation strategies, I identified a significant positive correlation between access to credit services and the implementation of various adaptation measures, such as terracing, agroforestry, manure application, adoption of crop varieties, and irrigation. Specifically, access to credit services was found to boost the adoption rates of terracing, agroforestry, and irrigation by 7.6%, 10.7%, and 11.1%, respectively, as illustrated in Table 2. Consistent with this findings, previous studies have also highlighted that access to credit facilities can alleviate the cash constraints faced by smallholders, thereby fostering the adoption of adaptation strategies^20,21,22.

Table 2 Parameter estimates of the marginal effects of the multinomial logit models for the adoption of adaptation strategies.

In contrast to conventional beliefs, this study revealed a negative association between livestock ownership and the adoption of terracing. This could be attributed to the fact that livestock rearing competes for labor resources that might otherwise be engaged in terrace construction. Additionally, the limited availability of grazing land could be a deterrent. Moreover, postharvest grazing on Kolla farmlands could compromise soil and water conservation structures. However, livestock ownership had a positive influence on the adoption of manure application, agroforestry, and new crop varieties. The favorable relationship between livestock ownership and manure application can be attributed to smallholders rearing livestock both for income generation and as a source of natural fertilizer. As previously discussed, the prevalent cut-and-carry stall-feeding system in the study area ensures that livestock serve as a consistent source of manure. Consequently, the easy collection and application of manure to fields to maintain soil fertility make livestock ownership a significant factor, enhancing the likelihood of manure adoption by 15.3% (p < 0.05). The positive correlation between livestock ownership and agroforestry status suggested that the presence of various tree species on farmlands serves as a valuable fodder resource for livestock. Furthermore, the association between livestock ownership and the adoption of new crop varieties indicates that livestock can be leveraged as a cash source. This, in turn, empowers smallholders to invest in purchasing high-quality seeds for improved crop varieties. The intricate relationships between livestock ownership, manure application, new crop varieties, and agroforestry underscore the complementary nature of these systems. For instance, livestock contribute to manure production and act as a financial resource for purchasing seeds of new crop varieties, while trees within agroforestry systems offer fodder for livestock. Consistent with this findings, prior research has also identified a positive link between livestock ownership and the adoption of climate change adaptation strategies^11,15.

Contrary to our expectations, this study identified an inverse relationship between access to relief and the adoption of certain adaptation strategies, such as soil bunds, intercropping, reduced tillage, and irrigation. This unexpected finding may be attributed to smallholders' growing frustration resulting from persistent crop failures due to recurring droughts. This prolonged exposure to adverse climatic conditions might have cultivated a sense of dependence on relief assistance among them. Additionally, the socioeconomic background of smallholders, including their educational level, might shape their perception and understanding of the purpose and benefits of relief aid. Key informants corroborated this perspective by highlighting that some recipients of food aid did not witness tangible improvements in their livelihoods. Instead, they appeared to anticipate and rely on continuous relief from either governmental bodies or nongovernmental organizations (NGOs). In contrast, this study findings revealed a significant positive correlation between access to relief and the adoption of certain adaptation strategies, namely, terracing, manure application, and the use of new crop varieties. This positive relationship between relief access and these specific adaptation strategies might stem from their complementary nature. For instance, cash-based relief could empower smallholders to invest in acquiring new crop varieties, purchasing high-quality seeds, procuring livestock, or even hiring labor for terrace construction. Consequently, access to relief significantly bolsters the likelihood of adopting terracing, manure application, or new crop varieties by 32.4%, 13.6%, and 32.5%, respectively (p < 0.01). This finding aligns with the observations of Bryan et al.⁴⁰, who noted that smallholders receiving more food aid exhibited heightened confidence in their ability to combat the challenges posed by climate change.

In terms of agroecological contexts, the MNL model highlighted that farming within the Woynadega agroecological zone significantly and positively influences the adoption of terracing, new crop varieties, and reduced tillage (as shown in Table 2). Specifically, residing in the Woynadega zone increases the likelihood of implementing terracing, adopting new crop varieties, and using reduced tillage by 41%, 30.2%, and 31.2%, respectively (p < 0.01).

The heightened propensity for terracing adoption in Woynadega could be attributed to its rugged terrain, which offers an ample supply of materials essential for terrace construction, especially stones. Conversely, the analysis of marginal effects indicated that being situated in Woynadega reduces the likelihood of utilizing soil bunds and irrigation by 1% and 8.4%, respectively (as depicted in Table 2). This finding may stem from the fact that smallholders in the Konso Woynadega region predominantly rely on terracing methods and face limited access to irrigation resources, given the scarcity of perennial streams in the area. Additionally, the reduced use of soil bunds in Woynadega could be due to the prevalence of stone terraces, which makes the need for soil bunds less pronounced owing to the limited availability of stones in the Kolla region.

Regarding the adoption of adaptation strategies based on gender, the analysis indicates that being part of a female-headed household reduces the probability of implementing terracing, agroforestry, manure application, soil bunds, intercropping, and reduced tillage by 1.4%, 27.2%, 0.1%, 5.9%, and 21.9%, respectively (as illustrated in Table 2). This trend may be attributed to the prevalent gender roles in rural Konso and broader Ethiopia, where women typically shoulder a greater share of domestic responsibilities than men. Additionally, women often face greater constraints in accessing information and benefiting from socioeconomic opportunities than their male counterparts. Consistent with these findings, prior research has also demonstrated that households led by men tend to be more inclined toward adopting new agricultural technologies than are those headed by women^9,41.

In terms of the adoption of adaptation strategies relative to age, the findings suggest that with each unit increase in the age of the household head, there is a significant 0.8% decrease in the likelihood of adopting new crop varieties. This trend may be attributed to older smallholders potentially having limited access to updated information on improved crop varieties, stemming from restricted access to contemporary information sources. Conversely, as the age of the household head increases by one unit, the probability of adopting soil bunds and reduced tillage increases by 0.6% and 0.5%, respectively (as depicted in Table 2). This increase could be attributed to the extensive experience of older smallholders with these traditional strategies, which have been employed in the region for an extended period.

When assessing the association between educational levels and the adoption of adaptation strategies, the data presented in Table 2 reveal both positive and negative correlations. Specifically, the educational status of a household head was found to have a significant positive impact on the adoption of intercropping, increasing the likelihood of education by 33.5% (p < 0.01). This positive association can be attributed to the notion that education serves as a crucial facilitator for smallholders to embrace innovative technologies and refine existing practices. Educated household heads are often better equipped to access, comprehend, and apply information, thereby fostering increased adoption of adaptation strategies^14,15.

When examining the relationship between educational attainment and the adoption of adaptation strategies, the data presented in Table 2 reveal both positive and negative correlations. Specifically, the educational level of a household head was found to have a significant positive impact on the adoption of intercropping, increasing the likelihood of education by 33.5% (p < 0.01). This positive association can be attributed to the notion that education serves as a crucial facilitator for smallholders to embrace innovative technologies and refine existing practices. Educated household heads are often better equipped to access, comprehend, and apply information, thereby fostering increased adoption of adaptation strategies^14,15.

When assessing the association between household size and the adoption of adaptation practices, this study findings indicate a positive correlation between the two for terracing, agroforestry, manure application, soil bunds, reduced tillage, and irrigation. This positive linkage likely stems from the fact that household members predominantly serve as the primary labor force for agricultural activities. Thus, larger households possess a greater pool of labor, enhancing the feasibility and implementation of labor-intensive adaptation techniques. Moreover, this study results suggest that households with more economically active members are more inclined to adopt labor-intensive agricultural technologies. This observation aligns with prior research that has identified a significant positive association between household size and the utilization of climate change adaptation strategies^11,25.

Regarding the influence of landholding on the adoption of adaptation practices, this study results indicate a significant positive relationship between larger landholdings and the adoption rates of agroforestry, manure application, crop varieties, reduced tillage, and irrigation, with increases of 6.2%, 6.9%, 5%, 5.5%, and 7.2%, respectively. This observation is consistent with prior research that underscores the positive impact of larger landholdings on the adoption of climate change adaptation strategies^14,15. Larger land areas enable smallholders to diversify their crops, thereby mitigating the risks associated with climate change¹⁰. Furthermore, the proximity of farmlands to market centers was found to influence the adoption of adaptation strategies. Specifically, the likelihood of implementing terracing, agroforestry, crop varieties, and soil bunds decreased significantly with increasing distance from the market. Proximity to markets provides smallholders with advantages in terms of easier access to information and exchange opportunities. Conversely, greater distances to market centers result in higher transaction costs for both input acquisition and product sales, thereby diminishing the relative benefits of adopting advanced agricultural technologies^11,15.

Conclusion

The study revealed that smallholders employ a variety of indigenous adaptation strategies in response to the challenges posed by climate variability and change. These strategies include terracing, agroforestry, manure application, adoption of improved crop varieties, soil bunds, intercropping, reduced tillage, and irrigation. Notably, the combined use of terracing with integrated crop-tree-livestock systems has proven particularly effective in mitigating the risks associated with climate variability and change. Despite the adoption of these diverse adaptation strategies, the agricultural productivity and food security of smallholders remain vulnerable due to persistent challenges such as recurrent droughts, erratic rainfall patterns, and increasing temperatures over prolonged periods. Additionally, the study underscores the significant influence of socioeconomic and institutional factors, such as access to credit facilities, climate information, and extension services. These factors either facilitate or hinder the adoption of adaptation strategies within the study area.

This study underscores the autonomous development of indigenous adaptation strategies in Konso by generations of smallholders. The region's rugged topography, susceptibility to erosion, recurrent droughts, and unpredictable rainfall patterns have compelled them to devise intricate indigenous adaptation mechanisms over centuries. While these autonomous strategies have historically enabled them to cope with climate variability and change, the intensifying impacts of climate change have stretched their already limited resources and adaptation capacities, exacerbated by the increasing frequency of droughts and erratic rainfall patterns. Recently, the adoption of these indigenous adaptation strategies has been hampered by a complex web of challenges. These include limited irrigation potential, financial constraints, labor shortages, sporadic extension and training services, and land scarcity. Given these challenges, there is an evident need for external interventions to bolster local livelihoods and ensure the sustainability of the traditional farming system in Konso. This necessitates the provision of regular weather forecasts, accessible and affordable credit facilities, and crop insurance to support the local community. Additionally, policy-driven initiatives are essential for introducing drought-resistant and early-maturing crop varieties. Diversifying livelihood options through opportunities in nonfarm or off-farm employment sectors is also crucial for reducing the vulnerability of smallholders to the impacts of climate variability and change^42,43,44,45.