Legume integration in smallholder farming systems for food security and resilience to climate change

PLOS One | https://doi.org/10.1371/journal.pone.0327727 August 13, 2025 4 / 20 goats, sheep, donkeys, mule, etc. were raised to provide food and income for HHs and used to ploughing land and their manure as natural fertilizer. Farmland holdings in the region are small, thus farmers use numerous cropping methods, including intercropping and crop rotation, to maximize output per unit area and lower the risk of crop failure due to climate change. Intercropping has become a widespread practice in the research area [26], and farmers make additional revenue while alleviating the food scarcity period. Cereals like maize and sorghum are often intercropped with common beans, faba beans, and field pea [24]. About 57 to 100% of the maize production area is intercropped in early and late belg (small rains) seasons with beans [25]. Field peas and faba beans are commonly grown in rotation with barley, wheat, and potatoes in the highlands. For the current study, we purposely selected potential legume producing districts from Konso, Gamo, Gofa, and South Omo Zones based on the information obtained from regional agriculture office. Household (HH) data The HH survey was conducted during 2022–2023. Information was collected on legume and other crop production and production constraints at the HH level from zones (the administrative structure next to the region) and seventeen districts within the zones. From each zones, two to three districts were purposely chosen, taking into account the area size and

population number of the respective districts. Furthermore, the selection process incorporated variations in agro-ecology and the potential for legume production, achieved through consultations with local administrative officials and agricultural expertise from the zones. From each district, three kebeles (the smallest administrative structure) were selected and added to 51 kebeles in two farming seasons and three agro-ecologies (highland, midland and lowland) (S1 Fig). Sampling techniques and sample size determination The study population or farms in the research region were chosen using stratified random sampling. The zones were con- sidered as strata. After determining the required sample size, the proportions used to select the samples from the study area. For determining the sample size for the strata, we adapted the following formula from [26]: n = [∑ ( Ni2∗pi∗ (1 – pi) /wi ) ]/[ (N2∗e2/Z2_ {α/2}) + ∑ (Ni∗p∗ (1 – p)) ] Where n is the sample size needed, N is the total population size, L is the total number of strata (Zone), Z alpha over two is the value of the upper tail probability of the standard normal, distribution Z = 1.96 for α = 0.05 significance level, Pi is the subpopulation proportion of farmers who have knowledge about legume-rhizobium inoculant technology for stratum i and i = 1, 2,…, L, e is the level of precision/marginal error/, Ni is the size of stratum (Zone) i and wi is the estimated proportion of Ni to N Considering N1 = N. Gamo ≤ 35518 N2 = N. Gofa ≤ 35436 N3 = N_S. Omo ≤ 18745 N4 = N. Konso ≤ 26519 N5 = N.

Basketo ≤ 1634 N6 = N. Amharo ≤ 6056 e ≤ 0.035 marginal error. It varies from 1% to 5%, p = 0.50, since no research has been done on this issue and there was no information about p. The sample size n is the equation for n = 779. District agri- cultural experts selected kebeles for a study based on potential and agro-ecology. Data collection Structured questionnaires were designed to collect data on socio-demographic and economic, agricultural, and service- related factors. It also included HH crop production priority (considering legumes, cereals, root crops, vegetables, fruits, and spices), owned livestock, production challenges, and their land holding sizes. The questionnaires were developed in English and then translated into Amharic. Agricultural expertise from districts and agricultural extension workers from each kebele were trained and used for data collection. We used Open Data Kit (https://opendatakit.org), a digital data collec- tion tool and uploaded it to a secured data repository (Kobo Toolbox, https://kobotoolbox.org). In areas where the data