The study was conducted in the major tef-producing areas of the eastern part of the Amhara Region, namely Kobo, Habru, and Kalu districts of the North and South Wollo Zones, as Fig. 1 indicated below. The experimental sites were purposively selected based on their high potential for tef production. Kobo District is located at 12°08'60.00"N latitude and 39°37'59.99"E longitude, with an altitude of approximately 1468 meters above sea level. Habru District lies at 11°32'58.00"N latitude and 39°48'58.89"E longitude, with an altitude of about 1825 meters above sea level, while Kalu District is situated at 11°00'00.00"N latitude and 39°49'59.99"E longitude, with an altitude of 1492 meters above sea level. The main livelihood of the local community is based on a mixed farming system, where smallholder farmers engage in both crop cultivation and livestock rearing. The study areas experience a bimodal rainfall pattern characterized by two distinct rainy seasons. The main rainy season, locally known as Kiremt, occurs from June to September and provides the primary source of moisture for crop production. The short rainy season, called Belg, extends from February to May and supports the cultivation of short-cycle crops and early land preparation for the main cropping season. Although total annual rainfall varies from year to year, it generally provides limited moisture for reliable crop production. Despite these challenges, the area remains an important agricultural zone where research and development efforts focus on the generation, adaptation, and innovation scaling of improved technologies, particularly high-yielding, early maturing, and disease-resistant tef varieties to enhance crop productivity and improve the livelihoods of smallholder farmers.

For this study, experimental sites and host farmers were selected in collaboration with the district and kebele offices of agriculture. Host farmers were chosen based on their willingness to provide land for the experiment and the suitability of their soil for conducting the trials. Before implementation, training was provided to agricultural experts and development agents on the specific field activities to be carried out. The experiments were conducted at Farmer Training Centers (FTCs) and on farmers’ fields. The number of FTCs used in each district was determined based on their potential and functionality. Variation in plot sizes across the sites resulted from land scarcity and the fragmented nature of farmland in the study areas.

The evaluation of improved tef variety, along with recommended production packages, was conducted using the Farmer Research and Extension Group (FREG) approach. Agronomic practices were implemented according to standard recommendations. NPS fertilizer was applied at a rate of 121 kg/ha at planting, while Urea fertilizer was applied at a total rate of 100 kg/ha in two splits 50% at planting and the remaining 50% at the vegetative growth stage. Other agronomic practices, such as weeding and crop protection, were carried out by farmers under the close supervision of development agents and researchers.

The experimental treatments included three management practices: improved Mena tef variety with improved management, local tef varieties with improved management, and local tef varieties under traditional (farmers’ own) management. Finally, field days and field visits were organized at vegetative and maturity stages, during which host farmers and neighboring farmers participated in evaluating the performance of the tested technologies.

The experiment was done in Kobo, Habru, and Kalu districts on farmers’ fields and farmers’ training center (FTC), as shown in Table 1 below.

During the result demonstration phase, Farmers’ Research and Extension Groups (FREGs) were established to evaluate the improved tef production practices. These small groups, composed of farmers and extension workers, played a key role in assessing the performance of the improved tef variety at different growth stages. The FREG approach is a participatory research method in which researchers, extension workers, and groups of farmers collaboratively conduct research on selected topics based on the farmers’ needs and interests directly in their fields (Edalkachew Z., 2008). The tef varieties were planted on farmers’ fields, and nearly all members of the farmers’ groups, along with researchers and extension workers, actively participated in the presentation, observation, and evaluation of the varieties during both the vegetative and maturity stages.

Data were collected through focus group discussions, individual interviews using a simple checklist, and field observations. The biological data collected comprised grain yield, straw yield, plant height, and panicle length. Economic data encompassed market prices of grain and straw, as well as input costs such as seed, fertilizer, labor, and chemicals. Social data includes farmers’ perceptions and preferences, along with the opinions of other stakeholders, which were assessed using a simple preference ranking method.

The collected data were analyzed using analysis of variance (ANOVA) through Gen Stat software to evaluate yield and yield-related traits. Duncan’s Multiple Range Test (DMRT), as described by Steel and Torrie (1980), was employed to compare the means among the different tef varieties. For profitability assessment, a partial budget analysis was conducted following the guidelines of CIMMYT (1988) to determine the economic advantage of the improved technologies over conventional farming practices. Furthermore, a pairwise ranking technique was used to analyze farmers’ perceptions and preferences towards improved and local tef varieties associated with different production packages.

The ANOVA results presented in Table 2 show the effects of different treatments on yield and yield-related traits (plant height, panicle length, biomass yield, and grain yield) of tef across Kobo, Habru, and Kalu. The treatments compared were the Improved (Mena) variety with improved management, the Local variety with improved management, and the Local with farmer management. At Kobo, Mena variety with improved management recorded the highest mean values for plant height (119.0 cm), panicle length (44.0 cm), biomass yield (5.33 t ha^–1), and grain yield (0.88 t ha^–1), with significant differences observed only for plant height. At Habru, Mena variety with improved management has a significant difference in plant height and panicle length. At Kalu, Mena maintained its advantage in grain yield (1.36 t ha^–1) and panicle length (45.9 cm), although differences among treatments were not statistically significant. The coefficients of variation (CV) were relatively high in Kalu, particularly for grain yield, indicating notable variability among treatments and across locations due to pest or disease damage.

Across locations, grand mean of grain yields for the Mena variety under improved management, the local variety with improved management, and the local variety under farmer practices were 0.96 t ha^–1, 0.98 t ha^–1, and 1.31 t ha^–1 at Kobo, Habru, and Kalu, respectively. Although no significant differences were observed in grain and biomass yields, it is important to note that all yield and yield-related parameters exhibited numerical differences among treatments. Moreover, significant variations in plant height and panicle length were observed at Kobo and Habru, respectively. These findings are consistent with previous research showing that improved variety and management combinations can influence vegetative growth traits even when yield gains are not statistically significant (Belissa and Bacha 2023; Jerjero et al. 2025).

In general, the key finding of the Mena variety with improved management has an insignificant yield difference over the local variety under improved and farmer management systems across all locations. The improvement was most evident in grain and biomass yields, demonstrating the potential of integrating improved varieties with better management practices to enhance tef productivity. However, the variability observed across sites, as reflected by the CV values, suggests that environmental conditions and pest or disease damage play a crucial role in determining yield performance.

The partial budget analysis demonstrates clear economic differences among the tested treatments when evaluated using the marginal rate of return (MRR). The MRR is an essential decision-making tool in farm-level economic analysis because it measures the additional return obtained for each additional unit of cost incurred when shifting from a lower-cost to a higher-cost treatment. In practical terms, an MRR above 100%, meaning a return of at least 1 ETB for every 1 ETB invested, is generally considered acceptable for smallholder farmers, as it compensates for production risks and limited capital resources.

The results presented in Table 3 show that the improved Mena tef variety, when combined with recommended management practices, generated higher relative net benefits than the local varieties (Bursa and Magna) managed under farmers’ practices in the Habru and Kalu districts. This indicates that investing in improved seed and management is economically advantageous in these areas. In contrast, in the Kobo district, the improved Mena variety with the full management package did not provide additional net benefit compared with the Boset variety under farmers’ practice. This outcome may be attributed to the early maturity and high yield potential of the Boset variety under moisture-deficit conditions, which can reduce production risks and labor requirements while maintaining profitability. Similar findings were reported by Assefa et al. (2021), who noted that Boset variety performed better in terms of yield and net benefit under moisture-stressed environments.

The MRR results further support these observations. In Habru and Kalu districts, the improved Mena variety with recommended management produced positive marginal returns of 33.46 Birr and 1.34 Birr, respectively. This means that each additional Birr invested in the improved package generated 33.46 Birr in Habru and 1.34 Birr in Kalu after covering additional input costs. According to CIMMYT (1998), an MRR greater than 100% indicates strong economic feasibility, while an MRR below 50% suggests limited attractiveness. Therefore, the investment was profitable with a minimum rate of return in Habru and Kalu. However, improved Mena variety with improved management showed no additional net benefit and was economically dominated by the Boset variety; consequently, the MRR was not calculated.

Generally, the improved Mena variety with recommended management practices proved economically limited viable in Habru and Kalu districts, but not in Kobo district, where the Boset variety under farmer management was more advantageous. These findings highlight the importance of selecting crop varieties and management strategies that match specific agro-ecological conditions. Site-specific recommendations are crucial to ensure that farmers adopt options that maximize net benefit while providing sufficiently high marginal returns to justify additional investment under risk-prone smallholder farming systems.

Farmers’ participation in the variety selection process plays a paramount role in identifying preferred traits for promoting the best tef varieties. Discussions were conducted with farmers to assist them in identifying selection criteria. The identified criteria in Kobo for selecting the best tef varieties encompass plant height, panicle length, earliness, productivity, seed color, tiller number, disease resistance, straw palatability, and marketability, as shown in Table 4 below.

In the Kobo district, a total of 22 farmers participated in the selection process. The results of pairwise ranking and final preference value analysis indicate that the improved Mena tef variety was the most preferred. This preference was based on several factors, including panicle length, seed color, disease resistance, marketability, earliness, plant height, tiller number, and straw palatability for animal feed. The local Boset variety with improved management was ranked second in preference, while the local Boset variety under farmer practices was ranked last, as shown in Table 5.

In the Habru district, the farmers’ preferred traits for evaluating tef varieties were seed color, productivity, disease tolerance, seed size, marketability, tiller number, earliness, and plant height, as indicated in Table 6 above. To quantify farmers’ preferences, a sum score was assigned to each variety based on the level of agreement among farmers, using a five-point Likert scale where 1 = very poor, 2 = poor, 3 = medium, 4 = good, and 5 = very good. Then the criteria rank was multiplied by the rank of each variety in this process 29 farmers actively participated. Such participatory approaches are widely recognized for their importance in variety selection, as they integrate farmers’ knowledge, market demands, and agro-ecological suitability, ultimately increasing the likelihood of adoption of improved tef varieties (Ahmed et al. 2023; Sudding et al. 2025).

A total of 29 farmers participated in the selection process. The final result of preference analysis reveals that local Magna with improved management prioritized first, Mena with improved practices ranked second, and local Magna variety with farmer practices ranked last (Table 7 below).

In the Kalu district, the identified and weighted selection criteria for choosing the best tef varieties were disease resistance, marketability, seed color, tillering capacity, panicle length, straw palatability, plant height, and earliness. These criteria were ranked based on their importance, as shown in Table 8 below.

The final acceptability ranking at the Kalu district showed that Mena variety with improved management was the most preferred variety, followed by Bursa with improved management and Bursa with local management (farmer practices). Mena tef variety top ranking reflects its superior performance in disease resistance, seed color, and size of grain for market and consumption purposes, number of tillers, and panicle traits, along with relative yield performance and compatibility in the study area. The results highlight that improved management practices significantly enhance varietal performance, with improved varieties responding best to these conditions.

In general, the results of the pairwise ranking indicate that farmers showed a strong preference for the improved Mena tef variety combined with improved production practices compared to the local varieties (Boset in Kobo and Bursa in Kalu) under both improved management and traditional farmer practices (Tables 5, 9). The improved Mena variety with improved production practices was highly and positively perceived across several important agronomic and market-related traits.

Farmers prioritized key selection criteria such as seed color, panicle length, disease resistance, earliness, grain productivity, plant height, tiller number, seed size, straw palatability, and marketability. These traits are directly linked to both household food security and income generation. For instance, disease resistance and earliness reduce production risks and vulnerability to erratic rainfall, while higher productivity and tiller count contribute to increased grain yield. Similarly, desirable seed color and larger seed size enhance market preference and price, and good straw palatability increases the value of crop residues for livestock feed.

Based on these criteria, the improved Mena tef variety with improved production practices was ranked first and identified as the most preferred option in both Kobo and Kalu districts. This indicates that farmers recognized the combined advantage of genetically improved varieties and better crop management practices in enhancing overall performance. The second-ranked option was the local varieties (Boset in Kobo and Bursa in Kalu) when grown under improved management practices, suggesting that improved agronomic practices alone can significantly enhance the performance of local varieties. Lastly, local varieties grown under farmers’ traditional practices were ranked lowest, reflecting comparatively lower performance in the prioritized traits.

Therefore, the findings suggest that farmers’ preferences are strongly influenced by productivity, resilience, and market-related attributes. The superior ranking of the improved Mena variety highlights the importance of integrating improved varieties with recommended production practices to achieve better agronomic and economic outcomes in Kobo and Kalu districts.

The fundamental planning tasks were demonstration activity were centered on the principle of “learning by doing,” emphasizing active participation from researchers, extension staff, and farmers. The primary objective was to optimize tef yield while engaging stakeholders in selecting varieties and management practices that align with their preferences, thereby facilitating long-term technology adoption.

Field days were organized at both the vegetative and maturity stages of the crop, during which farmers evaluated each tef variety and its production package based on criteria they established according to their interests and needs. Farmers provided valuable feedback, highlighting that the improved Mena tef variety possesses superior seed size and color suitable for both market and consumption, resistance to yellow rust disease, palatable straw for animal feed, and the ability to maintain green morphology even at the maturity stage. However, they expressed dissatisfaction with its loose panicle structure, which did not fully meet their preferences.

In comparison, the local Magna variety in Habru district, locally known as “dalga tef,” was appreciated for its semi-loose panicle, large seed size, high grain yield, shattering resistance, and green stalk retention. In Kalu district, the local Bursa variety was praised for its white seed color, large seed size, tall plant height with compact panicles, and high productivity, though it was criticized for susceptibility to yellow rust (locally called “mich”) and non-palatable straw. In Kobo, Boset variety was widely adopted due to its early maturity and high productivity, making it suitable for triple cropping systems during the Meher (June–September), irrigation (December–February), and Belg (March–May) seasons. Nonetheless, farmers reported issues with grain shattering during unexpected rainfall at harvest.

Figs 2, 3, 4, 5 below illustrate the field day, field visit, and overall performance of the demonstrated tef varieties across the study locations. The improved variety, Mena, was evaluated at all sites, while the local checks, Magna in Habru, Bursa in Kalu, and Boset in Kobo, served as local checks.

These observations underscore the importance of incorporating farmers’ experiential knowledge and preferences into varietal selection. While improved varieties such as Mena demonstrate superior agronomic traits, local varieties continue to hold value due to their adaptation to specific agro-ecological conditions and cropping patterns, highlighting the need for location-specific recommendations (Ahmed et al. 2023; Sudding et al. 2025). Farmers’ feedback provides essential insights for breeding programs and extension strategies to ensure both productivity and adoption success.