Smallholder cattle farmers’ perceptions, adoption potential, and utilization of trees and shrubs as livestock feeds in the Eastern Democratic Republic of the Congo

ABSTRACT

The sustainability of the livestock sector is negatively affected by insufficient quality and quantity of feeds. The effects of population growth and climate change further exacerbate the availability of feeds. Therefore, developing feasible low-cost strategies to improve year-round feed availability for enhanced livestock production is necessary. The use of trees and shrubs fodder has been recommended as a potential strategy to address feed scarcity. This study evaluated the perceptions, adoption potential, and utilization of trees and shrubs as livestock feeds in the Eastern Democratic Republic of the Congo (DRC) by employing a survey of 805 randomly sampled cattle farmers. The findings revealed that farmers utilized about 62 trees and shrubs for animal fodder. The farmers employed experience-based perceptions to evaluate the suitability of trees and shrubs for fodder based on animal and tree-related criteria. A probit model analysis results indicated that the utilization of trees and shrubs as livestock feeds were influenced by several household demographic factors, livestock farms, farming system factors, and institutional characteristics. The findings provide relevant insights into informing policy formulation and implementation that promote tree and shrub fodder utilization for improved livestock production under smallholder production systems.

KEYWORDS:

• Milk production
• smallholder cattle farmers
• probit model
• protein supplements
• trees and shrubs

1. Introduction

Population growth, urbanization, shifts in dietary preferences, and increasing consumer demand for livestock products are re-defining livestock production systems globally (Food and Agriculture Organization, 2020). In 2019, the United Nations estimated that by 2050 the world population would increase from 7.7 to 9.7 billion people, where Sub-Saharan Africa (SSA) would account for almost double growth (United Nations, 2020). Consequently, the demand for food, specifically animal protein in SSA, is expected to double by 2050 (Maina et al., 2020). This population growth raises a concern about the capability of the SSA countries to meet the dietary needs of its people, yet livestock production in this region remains sub-optimal (Food and Agriculture Organization, 2020). Urbanization is also rapidly increasing, and it is expected that 75% of the population will live in urban areas by the year 2050, making SSA the fastest urbanizing region (United Nations, 2020). The urban populace adopts new eating behaviours characterized by the consumption of higher amounts of animal protein and eating a higher proportion of their food away from home (Maina et al., 2020; Steinfeld et al., 2006). This anticipated high demand for livestock-derived foods represents market opportunities for smallholder farmers to increase milk and meat production, thus contributing to food security and poverty alleviation. However, the increase in livestock production in SSA countries at the national level is crucial to reducing the need for high-cost imports and saving currency, which might be diverted to productive investments (Ndambi et al., 2007).

In SSA, livestock is a predominant resource to rural smallholder households and is the main source of income, food and nutritional needs, employment, insurance, a store of wealth and social safety nets, and manure for soil fertilization (Food and Agriculture Organization, 2020; The World Bank, 2021). In the Democratic Republic of the Congo (DRC), cattle farming accounts for 11 per cent of the national potential of 40 million head of livestock, contributing 9.2 per cent of the country’s agricultural gross domestic product (GDP) (Mugumaarhahama et al., 2021; National Investment Promotion Agency, DRC, 2022). Despite this contribution and favourable environmental conditions, the current livestock output remains far below the country’s demand. This low production is associated with many complex and inter-related factors such as feed scarcity, the poor genetic potential of local breeds for functional traits, climate, widespread diseases, undeveloped marketing systems, infrastructure, and war (Bisimwa et al., 2018; Cox, 2011; Maass et al., 2012; Mugumaarhahama et al., 2021; Mutwedu et al., 2022; Paul et al., 2016). Nonetheless, feed scarcity in terms of quality and quantity, as well as seasonal fluctuation, was identified as the most important constraint to livestock development and productivity in DRC (Bacigale et al., 2014; Katunga et al., 2014). In the Eastern DRC particularly, this situation is intensified by the farming systems, which are predominantly transhumant and agro-pastoralists (Mugumaarhahama et al., 2021).

Similar to the rest of SSA, the fodder production systems in DRC are predominantly rain-fed, and smallholder farmers have limited capacities to practise fodder conservation technologies (such as hay and silage) for use during the lean periods and appropriate feeding regimes (Bacigale et al., 2014). The rain-fed fodder in the tropics, particularly pastures and crop residues, has low digestibility and is often deficient in essential nutrients, notably proteins required for increased feed intake and rumen microbial fermentation for improved performance of the ruminants (Abdulrazak et al., 2000; Maina et al., 2020; Mudzengi et al., 2020; Notenbaert et al., 2017). Alternative conventional supplements such as concentrates and agro-by products are not readily available and, when available, are costly for the majority of smallholder farmers who are resource-poor (Maass et al., 2012; Mutwedu et al., 2022). Furthermore, increased population growth and the shrinking of pastoral land due to cropping activities and building development have exacerbated the feed scarcity situation (Ayenew, 2019; Maina et al., 2020; Mugumaarhahama et al., 2021). Therefore, inadequate nutrition of cattle often culminates in substantial economic losses to the farmers due to low rates of weight gain or milk production, poor body condition, reduced productive and reproductive performance, and susceptibility to pests and disease infestations (Henry et al., 2018; Michalk et al., 2019). In response to this challenge, there is a need to provide alternative cheap-to-produce and locally available feed resources to supplement the feeds produced on smallholder farms. In this regard, both the national and transnational research organizations and development discourse have, over the past decades, advocated for the utilization of trees and shrubs to enhance the nutritional resource base and as a sustainable intensification of livestock strategy in the SSA countries (Ayenew, 2019; Food and Agriculture Organization, 2020; Franzel et al., 2014; Wambugu et al., 2011).

Trees and shrubs are instrumental in delivering the triple-win strategy of enhancing livestock productivity, mitigating harsh climate in pastoral and agro-pastoral areas, and improving food security and rural livelihoods (Henry et al., 2018; Notenbaert et al., 2017). This strategy is attributed to its multiple roles in contributing to the welfare of the cattle farming communities (by providing fuelwood, construction materials, shade, and veterinary medicine); and in mitigating climate change (through biological nitrogen fixation, carbon sequestration, and reduction of greenhouse gas emissions) (Ayenew, 2019; Brandt et al., 2020; Dubeux et al., 2017; Paul et al., 2020b; Pello et al., 2021). The trees and shrubs are highly valued because of their high volume of biomass production, nutritional value, and adaptation to poor soil and harsh climatic conditions (Barnes et al., 2021; Barwani et al., 2022; Maina et al., 2020; Pello et al., 2021). Nutritionally, experimental animal feed trials have shown that tender shoots, leaves, pods, twigs, and fruits of trees and shrubs are rich in proteins, vitamins, and minerals compared to pastures and crop residues (Ayenew, 2019; Juma et al., 2006; Osuga et al., 2008). Integrating trees and shrubs into animal diets improves palatability, feed intake, digestibility, and lactation performance (Ayenew, 2019; Derero & Kitaw, 2018; Osuga et al., 2011). The importance of leguminous trees and shrubs underscores their attribute of drought resistance, which increases the survival rates and the productivity of the animals when pastures depreciate in quantity and quality (Jamala et al., 2013; Kassie et al., 2018).

Given the benefits of the trees and shrubs, efforts are being made to increase the uptake of trees and shrubs fodder among the resource-poor smallholder farmers as cattle feeds. Organizations such as International Centre for Tropical Agriculture (CIAT), World Agroforestry Centre, and International Livestock Research Institute (ILRI) have been researching and disseminating the trees and shrubs fodder technologies among the smallholder cattle keepers in the tropics. However, their utilization by smallholder farmers for cattle feeding has proven unsatisfactory (Franzel et al., 2014; Wambugu et al., 2011). Roothaert and Franzel (2001) studied the farmers’ preferences for trees and shrubs species in Kenya. The authors found that the farmers in Kenya have integrated about 160 species of trees and shrubs within the farming systems across the farm boundaries, within the food crops, grazing land, and at the homestead. The preference of the trees and shrubs among the smallholder farmers was based on palatability, ability to satisfy hunger, and contributions to animals’ overall health. Paul et al. (2020a) focused on farmers’ adoption of improved forages in dairy production systems in Tanzania. The authors reported that the recognized ecological benefits, labour costs, and positive community attitude influenced the adoption potential of the farmers. Despite the insights, the studies focusing on the utilization of trees and shrubs as cattle feeds, notably in the DRC context, remain a significant literature gap. The few studies conducted in DRC focus on challenges facing the utilization of trees and shrubs among smallholder farmers (Barnes et al., 2021; Katunga et al., 2014; Muoni et al., 2019). This study, therefore, aimed to assess the socio-economic determinants of the utilization of trees and shrubs as livestock feeds by smallholder cattle farmers in the Eastern DRC. The findings of the study would provide insights for selecting and scaling up the most valuable trees and shrubs technologies for improved cattle nutrition and feeding to enhance productivity.

2. Materials and methods

2.1. Study area, sampling, and data collection

The study was conducted in the territories of Kalemie and Kabare in the Eastern DRC (Figure 1). The territory of Kalemie is located in the province of Tanganyika, between latitudes 6° 00’ S and 6° 50’S and longitudes 26° 40’ E and 27° 30 E, with an average altitude of 880 metres above sea level. The average annual rainfall is 999 mm. The average annual minimum and maximum temperatures are 18°C and 32°C, respectively, with a mean annual temperature of 27°C. It covers 30,512 km² with estimated population of 707,021 (Development Indicator Analysis Unit, 2020). The territory of Kabare is located in the province of South-Kivu between latitudes 2° 30’ S and 2° 65’ S and longitudes 28° 30’ E and 28° 8’ E, at an average altitude of 2225 metres above sea level. The average annual rainfall is 1897 mm. The average annual minimum and maximum temperatures are 12.6°C and 24.4°C, respectively, with a mean annual temperature of 18°C. It covers 1,960 km² with estimated population of 868,616 (Development Indicator Analysis Unit, 2020).

Figure 1. Map of the study areas, Eastern DRC. Map made using shape files from the common geographical reference (http://www.rgc.cd).

The criterion for enrolling smallholder cattle farmers into the study was the possession of at least one head of cattle. A multistage sampling technique was used to select territories, groups, and smallholder cattle farmers. In the first stage, the purposive choice for Kalemie and Kabare territories for this study was informed by their prominence in livestock farming and proximity to research centres, both local (e.g. National Institute for Agricultural Study and Research, INERA) and transnational (e.g. International Institute of Tropical Agriculture, IITA) that promote uptake of trees and legumes among the smallholders. In the second stage, 17 groups were selected from two territories based on the intensity of cattle farming, road access, and absence of conflict or violence. In Kalemie territory, five out of seventeen groups that fulfilled the criteria were selected, including Moni, Kasanga-Mtoa, Miketo, Mugonda, and Tumbwe-fief. While, in Kabare territory, twelve out of seventeen groups were selected, including Bugobe, Bugorhe, Bushwira, Bushumba, Cirunga, Irhambi-Katana, Kagabi, Luhihi, Miti, Mudaka, Mudusa, and Mumosho. Lists of all smallholder cattle farmers in each territory, which were compiled by the territorial livestock officers, served as a sampling frame for this study. A simple random sampling technique was used to select the survey respondents. Overall, 805 smallholder cattle households were randomly selected for interviews.

A cross-sectional design was used to collect data from smallholder cattle farmers through a household survey in May and June 2021 using a well-structured and pre-tested questionnaire. The questionnaire was designed by conducting an exhaustive literature review to extract all the variables related to the utilization of trees and shrubs as livestock feeds. Then the variables were transformed into re-testable questions (Kothari, 2004). Pilot tests were conducted twice on the same ten respondents to evaluate the questionnaire for relevance, design, clarity, interpretation, completeness, and time taken per interview. To ensure the internal consistency of the questionnaire, the responses were subjected to a reliability test using the Cronbach test (Kothari, 2004). The scores below 0.7 were identified and modified.

The interviews were conducted in Swahili or French languages, and the results were translated into English. The survey captured cattle farmer-related variables, including socio-economic characteristics, livestock feeds, feeding management, trees and shrubs, and specific information regarding farmers’ knowledge, production, importance, perceptions, preferences, and the utilization of cattle feeds prior to the interviews. All interviewers were either research technicians or university students doing internships at INERA in the territories of Kalemie and Kabare. In each territory, the same interviewers participated at all sites. Questions were administered to the person responsible for cattle farming, usually the household head, within an average of 45 min. Data were recorded on tablet devices by the interviewers. Each interviewer was able to administer eight questionnaires per day.

Cattle farmers were asked to mention all trees and shrubs consumed by cattle, goats, and sheep. The trees and shrubs were identified by their local names and botanical names, which were cross-checked in botanical books using photos from the field (Latham et al., 2021a, 2021b), and unrecognized species were identified at the Natural Science Research Centre (CRSN-Lwiro), located 45.3 km North of Bukavu City in South-Kivu Province, DRC. The survey also captured the edible parts consumed by the animals, the period of the year they provide such fodder to the animals, the niche of production, and other roles played by trees and shrubs. The number of animals owned by farmers was converted into a tropical livestock unit (TLU) using the conversion factors of 0.7 for cattle and 0.1 for sheep and goats (Maass et al., 2012).

2.2. Conceptual framework

This study underpins the theory of household utility maximization (Singh et al., 1985) to develop a conceptual framework that explores households’ motivations and constraints to the utilization of trees and shrubs as livestock feeds, as presented in Figure 2. In the framework, the household context variables include demographics, institutional, livestock farm and farming, livestock product marketing, livestock feed production, feeding factors, tree and shrub production, perceptions, knowledge, and preferences. The household context interacts with the perceived costs and benefits of trees and shrubs alongside the influence of intervening variables (including the legal and policy framework and intervention of agricultural research organizations). It is hypothesized that households will utilize trees and shrubs as livestock feeds if the benefits derived from utilization outweighs its associated costs. Household utilization of trees and shrubs as livestock feeds is expected to improve livestock productivity (with respect to reproductive and productive performance), sustainability, and household welfare (in regard to incomes and food security).

Figure 2. Conceptual framework of the study.

2.3. Analytical framework and variables

The smallholder farmer’s choices are based on their perceived utility, that is, if the expected net benefits of trees and shrubs $({\mathrm{B}}^{\mathrm{AF}})$are greater than the expected payoffs of the alternative feed sources $({\mathrm{C}}^{\mathrm{AF}})$, as shown in Equation (1). (1) $({\mathrm{B}}^{\mathrm{TS}}-{\mathrm{C}}^{\mathrm{TS}})\ge ({\mathrm{B}}^{\mathrm{AF}}-{\mathrm{C}}^{\mathrm{AF}})$(1) where $({\mathrm{B}}^{\mathrm{TS}}-{\mathrm{C}}^{\mathrm{TS}})$ are the benefits of trees and shrubs $({\mathrm{B}}^{\mathrm{TS}})$ including low cost of production, drought resistance, animal satiation, and improved milk production and costs or setbacks of trees and shrubs $({\mathrm{C}}^{\mathrm{TS}})$ such as toxicity and thorniness. While $({\mathrm{B}}^{\mathrm{AF}}-{\mathrm{C}}^{\mathrm{AF}})$ denotes the benefits of alternative feeds $({\mathrm{B}}^{\mathrm{AF}})$ and their associated costs $({\mathrm{C}}^{\mathrm{AF}})$ (Franzel et al., 2014; Kiptot & Franzel, 2012). The comparative Equation (1) underpins the theory of utility maximization, which assumes that when an economic agent (individual or household) faces a choice among two or more alternatives, the agent will select an alternative that yields the highest utility (Singh et al., 1985). If the difference in the expected net benefits for the two livestock feeding regimes is denoted by π, as shown in Equation (2), then a household will utilize tree and shrub forage if π>0 and vice versa (Singh et al., 1985). (2) $({\mathrm{B}}^{\mathrm{TS}}-{\mathrm{C}}^{\mathrm{TS}})-({\mathrm{B}}^{\mathrm{AF}}-{\mathrm{C}}^{\mathrm{AF}})>\pi$(2) where π is the latent variable representing the differences in benefits and costs associated with the utilization of tree and shrub forages and alternative feeds. The specification presented in Equation (2) implies that determinants that will raise π will enhance the utilization of trees and shrubs fodder and vice versa. The empirical modelling of a household choice of tree and shrub fodder can be denoted by a binary latent variable (Y) such that: (3) ${\mathrm{Y}}_{\mathrm{i}}=\{\begin{array}{c}1if{\pi }_{\mathrm{i}}={\pi }_{\mathrm{i}}+\epsilon =({\mathrm{B}}^{\mathrm{TS}}-{\mathrm{C}}^{\mathrm{TS}})-({\mathrm{B}}^{\mathrm{AF}}-{\mathrm{C}}^{\mathrm{AF}})\ge 0\\ 0,otherwise\end{array}$(3) where ϵ denotes the standard error term, Equation (3) can be estimated using probit regression analysis to determine the factors influencing the propensity to utilize trees and shrubs as livestock feeds. The magnitude π is presumed to be highly heterogeneous depending on various contextual specific household factors, including gender, age, differential information, knowledge asymmetries, cattle production systems and capacities, farming experiences, incomes, and liquidity constraints (Poole, 2017). The empirical specification for Equation (3) is expressed in Equation (4). (4) ${\mathrm{Y}}_{\mathrm{i}}=\varphi \left({\beta }_0\sum _{\mathrm{n}}^{\mathrm{m}}{\beta }_{\mathrm{j}}{\mathrm{X}}_{\mathrm{ni}}\right)+\epsilon$(4)

Where $\varphi$ is the cumulative standard normal distribution function, X is a vector of factors hypothesized to affect the choice of tree and shrub fodder as livestock feeds, including attributes such as age, gender, and education level, as shown in Table 1.

Table 1. Descriptions and descriptive statistics of the sampled householdsis.

Variable	Descriptions and measurement	Pooled sample Mean/proportion	Users (1) n = 511	Non-users (0) (n = 294)	Difference (1)-(0)
Location	Location of household head (0 = Kabare;1 = Kalemie)	0.50	0.55	0.41	−3.85***
Gender	Gender of household head (0 = female;1 = male)	0.83	0.85	0.80	−1.94*
Age	Age of year of household head (HH (years)	44.76	43.81	46.40	2.25**
Marital status	Marital status of the HH (1 = married; 0 = otherwise)	0.89	0.90	0.88	−0.94
Education level	Education level of the HH (1 = primary 2 = secondary 3 = university	1.18	1.16	1.20	0.66
Household size	Household members above 18 years old (Number)	7.63	7.70	7.51	−0.72
Dependant size	Number of household members below 18 years	3.86	3.98	3.65	−1.68
Household income source	1 = crop-livestock 0 = livestock only	0.26	0.29	0.22	−2.01*
Monthly income	Household monthly income (USD)	65.18	68.88	58.74	−2.51**
Group membership	Membership in a social or economic group (1 = Yes)	0.40	0.42	0.36	−1.69
Land size	Household land size holding (hectares)	1.34	1.33	1.35	0.14
Land tenure	Tenure of landholding (0 = none 1 = customary 2 = purchased)	0.92	0.89	0.99	2.03
Livestock farming experience	Years of livestock farming	13.23	13.12	13.44	0.34
Cattle farming purpose	Purpose of cattle farming (1 = income source 0 = otherwise)	0.71	0.74	0.65	−2.68***
Cattle holding	Number of cattle (Tropical livestock units, TLU)	11.41	13.11	8.44	−4.14***
Improved cattle holding	Number of improved cattle (TLU)	0.10	0.12	0.05	−2.50**
Goats TLU	Number of goats (TLU)	0.22	0.20	0.25	2.12**
Sheep TLU	Number of sheep (TLU)	0.05	0.05	0.05	0.33
Meat production	Livestock reared for meat production purposes (1 = yes)	0.04	0.03	0.06	2.16**
Milk production	Livestock reared for milk production purposes (1 = yes)	0.70	0.79	0.54	−7.66***
Milk production quantity	Milk production per day (litres)	4.31	4.90	3.30	−3.44***
Milk sales	Milk sales per day (Litres)	2.90	3.42	1.99	−3.80***
Milk price	Price of milk per litre (USD)	0.31	0.36	0.23	−3.83***
Cattle feeding system	Cattle feeding system (0 = Zero grazing; 1 = Grazing)	0.84	0.84	0.84	−0.18
Grass production	Household produces grass (1 = Yes)	0.37	0.35	0.39	1.31
Grass production area	Grass production area (hectares)	0.04	0.05	0.03	−1.19
Legume production	Household produces legumes (1 = Yes)	0.17	0.23	0.08	−5.33***
Legume production area	Legume production area (hectares)	0.03	0.04	0.01	−2.47**
Crop residue usage	Household uses crop residues (1 = yes)	0.53	0.52	0.53	0.26
Feed conservation	Household conserves feeds (1 = yes)	0.12	0.12	0.12	−0.01
Tree and shrub usage	Household utilizes trees and shrubs as cattle feed (1 = yes)	0.63	–	–	–
Tree and shrub production	Household produces trees and shrubs	0.44	0.69	0	−25.37***

***p < 0.01, **p < 0.05, *p < 0.1; TLU = Tropical livestock units; USD = United States dollars

2.4. Data analysis

After data collection, all tablet devices were synchronized with the Kobotoolbox platform server through an internet connection to upload all filled forms to the server (KoBoToolbox, 2020). An excel spreadsheet with data was then downloaded from the Kobotoolbox platform server, verified for input errors, and coded. The data was checked and cleaned to ensure consistency for any missing variables and outliers. Computation of inferential statistics, including the t-tests and statistical descriptions, was used to explore the correctness of the data and any underlying patterns in the dataset. Data analysis was executed with STATA software, version 16 (StataCorp, 2019). The independent variables were tested for multicollinearity using variance inflation factors (VIF). Unless otherwise indicated, the significance level was set at p < 0.05. The probit regression analysis was used to assess the likelihood of independent variables determining the utilization of trees and shrubs as livestock feeds. The average marginal effects were estimated to determine the contribution of the independent variables to the probability of utilization of trees and shrubs.

3. Results and discussion

3.1. Socio-economic characteristics of the smallholder cattle farmers in the study areas

The descriptive statistics for the smallholder cattle farmers are presented in Table 1. Most of the sampled household heads were males (83%) and married (89%), having a mean of 8 household adult equivalent members and 4 dependents. The majority of the household heads were of middle age (45 years), having attained primary school education. This concurs with the findings of Mutwedu et al. (2022), who reported that most farmers involved in livestock were married men aged between 41 to 60 years with a primary school education level. The low level of education in the study areas may be explained by the less access to the education in rural areas due to poverty and the long period of political instability, which is consistent with other studies on the livelihood strategies of smallholder farmers in the Eastern DRC (Cox, 2011; Mugumaarhahama et al., 2021). About 26% of the households depended on mixed crop and livestock production for their livelihoods, earning per month a mean income of United States Dollars (USD) 65.18. An estimated 40% of the households had membership in either social or economic groups on social networks.

The majority of the households (92%) had a customary tenure form of landholding with a mean land size of 1.3 hectares. The observation of the landholding size is consistent with the rising trends of sub-divisions of land in SSA (Food and Agriculture Organization, 2021). The Food and Agriculture Organization (2021) report indicates that the trend will negatively affect farmers’ production of livestock feeds. The results showed that the farmers from this study had a mean livestock farming experience of 13 years. This experience would have implications for utilizing trees and shrubs in cattle feeding. The findings showed that most of the farmers had higher tropical livestock units of cattle (11.41 units), followed by goats (0.22 units) and sheep (0.05 units). However, cases of improved cattle were very low (0.10 units). The majority of the farmers reared cattle for income generation (71%), mostly for milk production (70%), and less for meat production purposes (4%). The farmers produced an average of 4 litres of milk per day, allocating 3 litres for sale at USD 0.31 per litre. Apportioning 75% of the milk produced to sales is consistent with the farmers’ economic assignment on cattle production. In terms of the feeding systems, the majority of the farmers (84%) practised free range grazing, with about 37% of the households producing grass and about 17% producing legumes. Farmers allocated slightly more land to grass production (0.04 hectares) than legumes (0.03 hectares). An observation on the utilization of crop residues is consistent with the predominating trend of the smallholders’ dependence (in this case, 53%) on crop residues, including maize, beans, and sugarcane, in cattle feeding. These results are consistent with the observation of Paul et al. (2020b) who reported that in Northern Tanzania farmers kept large local cattle herd sizes, few improved breeds, daytime grazing, widespread crop residue feeding and low productivity.

Very few farmers (12%) practised feed conservation innovations such as hay and silage. The observation was consistent with that of Notenbaert et al. (2017), who attributed the low uptake of conservational technologies to a lack of technical know-how and access to or unaffordability of the required ingredients. The high dependence on crop residues and low uptake of conservation strategies highlight the need to supplement locally produced feeds, especially during lean periods. Over 63% of the farmers utilized trees and shrubs for feeding their cattle, with about 44% of households cultivating them within their farms.

The mean differences following an independent t-test among the tree and shrub users and non-users are summarized in Table 1. There were significantly more farmers in Kalemie (55%) utilizing trees and shrubs as cattle feeds than in Kabare (41%). The majority of the tree and shrub user households were male-headed and significantly younger than the non-users. Although not significantly different, the tree and shrub users were more educated and had a higher number of household adult equivalent members and dependants than the non-users. A significantly higher number of tree and shrub users (29%) reported mixed crop and livestock farming as their source of income than the non-users (22%) and consequently earned significantly higher monthly income (USD 68.88) than the non-users (USD 58.74). Although not significant, more tree and shrub users (42%) reported having membership in a social or economic group than the non-users (36%). There were no significant differences between the tree and shrub users and non-users regarding the size of land owned and livestock experience. There were significant differences (p < 0.01) between the tree and shrub users and non-users regarding the cattle owned, where the users owned more cattle (13.11 units) than non-users (8.44 units). Similarly, users of trees and shrubs own significantly more improved cattle breeds (0.12 units) than non-users (0.05 units).

The tree and shrub users produced significantly more milk (4.90 litres per day) than the non-users (3.30 litres per day). Correspondingly, the users sold more milk (3.42 litres per day) at a higher price of USD 0.36 per litre compared to the non-users (1.99 litres per day) who sold a litre at USD 0.23. Milk production and higher sales would incentivise the farmers to utilize trees and shrubs to enhance productivity and income. Results further showed that there were no significant differences between the household production of grasses and the grass production area. However, more tree and shrub non-users used grass in cattle feeding than users. The tree and shrub users allocated significantly more land to grass production (0.05 hectares) than the non-users (0.03 hectares). More tree and shrub users (23%) produce legumes than non-users (0.08%). There were no significant differences in the utilization of crop residues and feed conservation. The tree and shrub users produced their fodder by cultivating the trees and shrubs on-farm (69%), while the non-users barely produced any fodder on the farm. This attitude would be attributed to the transhumance nature of the cattle producers hereof, who lack land tenure for tree and shrub cultivation. Overall, there were significant differences between the tree and shrub users and non-users on numerous variables. Therefore, it qualified for the econometric estimation of the dataset to determine the factors influencing the utilization of trees and shrubs as cattle feeds.

3.2. Farmer production of trees and shrubs

Table 2 presents the results of the attributes considered in the production of trees and shrubs among smallholder cattle farmers. The farmers cultivated the trees and shrubs along the farm boundaries (12.5%), homestead compound (4.6%), within food crops (2.1%), within grazing land (0.1%), and off-farm (0.5%). There were more farmers in Kabare who preferred cultivating the trees and shrubs along the farm boundaries (24.2%), compared to Kalemie (0.8%). In contrast to the farmers in Kalemie, the farmers in Kabare planted trees and shrubs within food crops (3.2%) and grazing land (0.2%).

Table 2. Farmer production of trees and shrubs.

Utilization	Pooled sample (%)	Kabare (n = 405) (%)	Kalemie (n = 400) (%)
Parts utilised as livestock feeds
Leaves only	41.4	27.9	55.0
Leaves and twigs	18.3	25.4	11.0
Leaves and pods	0.7	1.2	0.3
Fruits and pods	0.2	0.5	0
Season of tree and shrub usage
Through the year	38.0	33.6	42.5
Dry season	23.9	21.2	26.5
Rainy season	1.6	2.2	1.0
Production of trees and shrubs
Farm boundary (live fences or windbreaks)	12.5	24.2	0.8
Homestead compound (farm woodlands)	4.6	9.1	1.0
Within food crops (alley cropping)	2.1	3.2	0
Within grazing land (production hedges)	0.1	0.2	0
Off-farm	0.5	0.7	0.3
Other usages of trees and shrubs
Soil fertilization	33.0	29.4	36.8
Soil erosion prevention	32.4	32.1	32.8
Fuel wood	38.0	23.0	53.3
Wind breaks	26.0	21.5	30.5
Construction materials	30.8	21.0	40.8
Live fence	18.5	17.0	20.0
Human medicinal purposes	18.4	6.9	30.0
Veterinary medicine	28.2	17.3	39.3
Environmental protection	20.7	22.2	19.3

The trees and shrubs served multipurpose functions to the cattle farmers, including fuelwood (38%), soil fertilization (33%), provision of construction materials (30.8%), veterinary medicine (28.2%), live fences (18.5%), and human medicine purposes (18.4%). The results showed that the farmers in Kalemie utilized trees and shrubs more for a number of welfare purposes than in the Kabare location. In all seasons, the farmers utilized trees and shrubs throughout the year, with farmers in Kalemie utilizing more (42.5%) than their Kabare counterparts (33.6%).

The trees and shrubs were also utilized more during the dry season than the rainy season. This attitude is probably due to the abundance of other feed resources during the rainy season. Most farmers utilized leaves only in feeding their cattle, followed by the combinations of leaves and twigs, leaves and pods, and fruits and pods. This finding is consistent with Roothaert and Franzel (2001), who observed that the preferred livestock feeding parts of trees and shrubs were twigs plus leaves.

3.3. Farmer’s most preferred tree and shrub species for cattle feeding

Farmers in the Kabare and Kalemie territories utilized a total of 62 trees and shrubs species to feed their cattle. However, farmers in Kalemie territory utilized more species (36) compared to their Kabare counterparts (26 species). Table 3 presents the most preferred tree and shrub species for cattle feeding. Notably, Vernonia amygdalina (Del.) was the most preferred species by the farmers in Kalemie (56.25%) than in Kabare (5.90%). The tree is locally known as ‘Mubirizi’ and belongs to the Asteraceae family. It is one of the multipurpose trees that can rapidly grow, regenerate and produce much biomass for forage (Kedir & Feki, 2021). Evidence opines that supplementation of Vernonia amygdalina in livestock feeding improves animal feed intake, nutrient digestibility, and body weight gain (Kedir & Feki, 2021; Mengistu et al., 2020). The Persea americana (Mill.), locally known as the ‘Avoca’ tree, belongs to the family of Lauraceae, was the second most preferred tree notably by the farmers of Kabare (28.10%) compared to those in Kalemie (1.25%). The tree is a tropical fruit that thrives under harsh climatic conditions and is cultivated majorly for the production of fruits. Supplementing P. americana forage in livestock feeding has been shown to improve meat quality (Leontopoulos et al., 2021). Calliandra calothyrsus (Meisn.) was also utilized in Kabare (16.30%) and Kalemie (1.25%) by cattle farmers. The farmers cultivate C. calothyrsus for fuelwood, shade, and ornamental purposes. According to Cook et al. (2005), C. calothyrsus increases the dry matter feed intake when fed to animals.

Table 3. Tree and shrub species preferred as cattle feeds.

Species	Local names	Family names	Kalemie (%)	Kabare (%)
Vernonia amygdalina Del.	Mubirizi	Asteraceae	56.25	5.90
Erythrina abyssinica Lam.	Cigoho	Fabaceae	4.75	0.50
Leucaena leucocephala (Lam.) De Wit	Leucaena	Fabaceae	3.75	5.70
Mangifera indica L.	Hembe	Anacardiaceae	3.50	1.00
Tephrosia vogelii Hook. f	Bubaka	Fabaceae	3.00	1.50
Tithonia diversifolia (Hemsl.) A. Gray	Tithonia	Asteraceae	2.25	13.10
Albizia gummifera C. A. Sm.	Mushebeya	Fabaceae	1.75	0.70
Persea americana Mill.	Avoca	Lauraceae	1.25	28.10
Calliandra calothyrsus Meisn.	Calliandra	Fabaceae	1.25	16.30
Ficus glumosa Delile	Mutudu	Moraceae	0.50	8.10

The Tithonia diversifolia is among the shrubs preferred and utilized by the farmers in Kabare (13.10%) than in Kalemie (3.25%) for cattle feeding. The farmers interviewed opined that T. diversifolia has several advantages, including high production, fast growth and regrowth after cutting, and tolerance to acidic soils. Research shows that T. diversifolia has high crude protein content and minerals as well as low fibre content, and its inclusion in livestock diets would boost nutrient intake and nutrient digestibility (Mauricio, 2017; Osuga et al., 2012; Sirait & Simanihuruk, 2021).

Ficus glumosa (Delile) was mostly preferred by the farmers in Kabare (8.10%) than in Kalemie (0.50%). Farmers in Kalemie preferred Erythrina abyssinica and Mangifera indica (4.75% and 3.5%) compared to farmers in Kabare (0.50 and 1%), respectively. Similarly, farmers in Kalemie territory preferred Tephrosia vogelii, and Albizia gummifera to farmers in Kabare. Farmers in Kabare preferred Leucaena leucocephala compared to farmers in Kalemie. Barwani et al. (2022) recently evaluated the chemical composition and in vitro gas production profile of V. amygdalina, E. abyssinica, C. calothyrsus, L. leucocephala, Sesbania sesban, M. indica, T. diversifolia, and F. glumosa from the Eastern DRC. The authors reported that these trees and shrubs had higher crude protein content and in vitro digestibility, suggesting their suitability as protein supplements in livestock diets, particularly during the dry season. Subsequently, Barwani et al. (2023) reported a daily increase in milk production of 3% and 15% of crossbred cows fed on a basal diet of Guatemala grass (Tripsacum andersonii) and supplemented daily with 2.83 kg of DM of dried leaves of L. leucocephala and Manihot esculenta, respectively.

3.4. Farmer’s evaluation of trees and shrubs for cattle feeding

Farmers employed either animal-related criteria such as palatability, ability to improve animal satiation, growth or productivity effects or tree-related criteria, including compatibility with cultivated crops, drought resistance, improving soil fertility and toxicity, as shown in Table 4. For the livestock-related criteria, the most mentioned attribute was the ability of the tree and shrub fodder to satisfy the hunger of the animal. The finding on animal satiation is consistent with the findings of Roothaert and Franzel (2001) in a study carried out in Kenya. The authors indicated that the farmers ranked animal satiation as the commonly used criteria to assess the quality of fodder. However, more farmers in Kalemie (63.5%) utilized the animal satiation criteria than in Kabare (35.3%).

Table 4. Self-reported criteria farmers use to evaluate the tree and shrub forages as cattle feeds.

Criteria	Pooled sample (n = 805) (%)	Kabare (n = 405) (%)	Kalemie (n = 400) (%)
Livestock-related criteria (positive attributes)
Animal satiation	49.3	35.3	63.5
Palatability	47.3	47.3	54.8
Improves animal health	34.4	27.7	41.3
Increased milk production (cows)	16.4	10.4	22.5
Improved growth	20.5	16.0	25.0
Improved meat production	10.9	1.2	20.8
Livestock-related criteria (Negative attributes/side effects)
Cough	0.1	0.2	0
Diarrhoea	2.0	1.5	2.5
Loss of appetite	0.2	0	0.5
Bloat	0.7	1.5	0
Death	1.0	0.7	1.3
Tree-related criteria (positive attributes)
Compatibility with cultivated crops	20.0	13.3	26.8
Drought resistance	50.8	40.7	61.0
Improves soil fertility	26.5	23.0	30.0
Tree-related criteria (negative attributes)
Toxicity	5.8	4.7	7.0
Thorniness	1.1	0.2	2.0
Toxicity and thorniness	0.4	0	0.8

Farmers in Kalemie employed the qualities of palatability and ability of the fodder to improve animal health and production (in regards to milk, meat, and growth) compared to farmers in Kabare territory. The negative-related criteria, such as the ability of the trees to impose adverse health effects, were less observed than the positive livestock-related attributes. The tree and shrub fodder mostly caused diarrhoeas (2.0%), death (1.0%), bloat (0.7%), loss of appetite (0.2%) and cough (0.1%).

Cattle farmers from Kalemie territory reported using the tree criteria thereof compared to Kabare. Based on positive tree-related criteria, drought resistance (50.8%) was the most employed criteria, followed by the ability to improve soil fertility (26.5%) and compatibility with other crops. The species such as Leucaena leucocephala, Calliandra calothyrsus and Tephrosia vogelii were reported by farmers as potential plants that improves soil fertility. Jose and Dollinger (2019) reported that leguminous trees improve soil health by drawing nitrogen from the air and transfers it to the soil through their roots and fallen leaves, thereby replenishing depleted soils with nutrient-rich organic matter. The authors further reported that when such trees are planted by farmers in association with improved grasses, they increase carrying capacity and enhance the productivity of grazing cattle.

Farmers reported negative tree-related criteria as toxicity (5.8%), thorniness (1.1%), and both toxicity and thorniness (0.4%). However, the farmers interviewed reported that the toxic substances contained in the tree and shrub fodder could provide beneficial anthelminthic effects to livestock.

Table 5 presents the rankings of the most preferred trees and shrubs fodder qualities. Farmers ranked the tree based on its ability to satisfy hunger, with the highest-ranking being V. amygdalina at 74.3%, while the lowest was T. diversifolia at 45.5%. The farmer’s assessment shows that V. amygdalina was highly ranked on palatability (84.3%) compared to other highly preferred trees and shrub fodder. C. calothyrsus improved milk productivity (20.8%), T. diversifolia improved animal growth (45.5%), and V. amygdalina (81.4%) was ranked highest in improving meat production and animal health. Similarly, V. amygdalina was ranked as the most drought-resistant, while C. calothyrsus was hailed as a soil fertilizing tree and the most compatible tree with cultivated crops.

Table 5. Farmers’ evaluation of the quality of the most preferred tree and shrub species.

Criteria	Species
	Persea americana	Vernonia amygdalina	Calliandra calothyrsus	Tithonia diversifolia
Livestock-related criteria
Animal satiation	61.5	74.3	45.8	45.5
Palatability	66.7	84.3	66.7	72.3
Improves animal health	43.6	60.0	33.3	27.3
Increased milk production	15.4	12.9	20.8	9.1
Improved growth	23.1	20.0	20.8	45.5
Improved meat production	0	81.4	8.3	0
Tree-related criteria
Compatibility with cultivated crops	15.4	17.1	25.0	9.1
Drought resistance	69.2	88.6	58.3	63.6
Improves soil fertility	10.3	34.3	62.5	27.3

The farmers’ assessments on the ranking of the scores most preferred trees and shrubs are shown in Table 6. The scores were rated on the scale of 1 to 5, that is, 1 = bad, 2 = poor, 3 = fair, 4 = good and 5 = excellent. C. calothyrsus ranked highly in terms of growth; the second-rated tree was V. amygdalina, while F. glumosa had the lowest rating. C. calothyrsus had the highest ranking in regard to regrowth and drought resistance. Conversely, F. glumosa had the lowest rating on drought tolerance and the ability for regrowth. V. amygdalina was rated highly on compatibility when intercropped, while C. calothyrsus and T. diversifolia had zero ratings. V. amygdalina and C. calothyrsus showed the highest rating on palatability, followed by T. diversifolia and F. glumosa.

Table 6. Farmers scoring the quality of the most preferred fodder tree and shrub species on selected criteria.

Species	Growth	Regrowth	Drought resistance	Compatibility with crops	Palatability
Persea americana	2.56 (1.23)	2.95 (1.12)	3.33(1.06)	1.85(1.13)	3.51(1.23)
Vernonia amygdalina	0.39 (1.11)	0.40 (1.15)	0.37 (1.11)	0.34 (1.00)	0.49 (1.39)
Calliandra calothyrsus	0.46 (1.14)	0.45 (1.12)	0.46 (1.44)	0	0.49 (1.21)
Tithonia diversifolia	0.11 (0.59)	0.09 (0.52)	0.11 (0.62)	0	0.10 (0.56)
Ficus glumosa	0.05 (0.39)	0.05 (0.40)	0.05 (0.42)	0.04 (0.34)	0.06 (0.47)

The figures in parentheses are the robust standard errors

3.5. Determinants of utilization of trees and shrubs as cattle fodder

The study employed a probit model to estimate the factors that influence the decision of the farmers to utilize trees and shrubs as cattle feeds. The empirical model results and marginal effects are presented in Table 7. The results indicate gender differences in utilizing the trees and shrubs technologies, with male farmers being more likely to use the technologies than females. In Eastern DRC, Mugumaarhahama et al. (2021) attributed this observation to less engagement of female farmers in livestock farming, especially cattle farming. Further, Mutwedu et al. (2022) reported that this low commitment could be attributed to cultural beliefs and practices considering livestock farming is restricted to men as families look up to men for solutions for tackling whole family needs. This finding indicated the need to encourage women to adopt this sustainable agricultural practice through extension workers. Similar findings on the uptake of agroforestry were reported in Kenya and Malawi, and this was attributed to socio-economic inequalities and barriers which women face, including access to extension services, production, and market information (Maindi et al., 2020; Thangata et al., 2002). Nyberg et al. (2021) reported that female farmers in SSA are known to have less access to extension services, which may explain why male farmers in this study were likely to use more trees and shrubs fodder than female farmers. However, women farmers should therefore be specifically targeted based on their needs in the study area to increase the adoption of tree and shrub fodder. These findings are in line with the observations of Ayantunde et al. (2020) who reported that access to extension services increases farmers’ likelihood of adopting agricultural technologies. The findings underline the importance of access to information in promoting the adoption of sustainable agricultural practices. It has been reported that the weak agricultural extension services in the DRC fail to deliver knowledge and technology to rural areas due to lack of coordination, lack of a unified and clear policy and mandate, lack of funding, ageing and low skills of agents, and lack of mobility and interactions of agents with key actors (Ragasa et al., 2016). Strengthening the agricultural extension services is therefore critical for sustainable agricultural practices in the DRC. Effective extension systems are essential for the dissemination of proven agricultural technologies, off-farm testing and evaluation of the technologies, building the capacity of smallholder farmers, and linking farmers to appropriate development agencies and research institutions (Ayantunde et al., 2020).

Table 7. Probit model estimates on determinants of utilization of trees and shrubs as cattle feeds.

	Coefficient	SE	AME	p-value
Gender	0.336**	0.140	0.106	0.017
Age	−0.002	0.005	−0.001	0.649
Marital status	−0.144	0.173	−0.045	0.404
Education level	−0.173***	0.066	−0.055	0.008
Household size adult equivalent	0.006	0.025	0.002	0.798
Dependant size	0.045	0.032	0.014	0.153
Household income source	0.077	0.137	0.024	0.576
Monthly income	0.001	0.001	0.000	0.494
Group membership	0.394***	0.122	0.124	0.001
Land size	−0.028	0.028	0.124	0.317
Livestock farming experience	0.001	0.005	−0.009	0.838
Cattle holding TLU	0.009*	0.005	0.003	0.087
Improved cattle holding	0.324**	0.158	0.102	0.041
Goats TLU	−0.299*	0.156	−0.094	0.056
Sheep TLU	−0.257	0.307	−0.081	0.402
Cattle farming purpose	0.088	0.110	0.028	0.427
Meat production	−0.161	0.260	0.168	0.537
Milk production	0.535***	0.142	−0.051	0.000
Milk production quantity	−0.064**	0.030	−0.020	0.031
Milk sales	0.076**	0.038	0.024	0.043
Milk price	0.225*	0.134	0.071	0.095
Grass production	−0.052	0.154	−0.016	0.737
Grass production area	0.216	0.302	0.068	0.474
Legume production	0.814***	0.173	0.256	0.000
Legume production area	0.156	0.414	0.049	0.706
Crop residue usage	0.674***	0.190	0.212	0.000
Feed conservation	0.021	0.169	0.007	0.901
Location	0.869***	0.235	0.274	0.000
Constant	−1.279***	0.347		0.000
Goodness of fit measures
Prob > chi^2	0.000
Pseudo R^2	0.1546
LR chi^2 (28)	163.40
Log-likelihood	−446.669
Number of observations	805

***p < 0.01, **p < 0.05, *p < 0.1

SE = standard error

AME = Average marginal effects

More findings from this study showed that the level of formal education of cattle farmers was not a significant determinant in utilizing tree and shrub innovations. This finding could be explained by the tendency of the more educated people to prefer off-farm employment compared to agriculture. This result is consistent with Jera and Ajayi (2008), who opined that less-educated farmers gain experience through their involvement in agricultural work, and hence they could be well informed on trees and shrubs fodder technologies.

The findings on membership showed that belonging to a social or agricultural group increased the likelihood of utilizing trees and shrubs. Similar findings were observed by Wambugu et al. (2011) and Pello et al. (2021), who argue that group membership facilitates flow, exchange, and sharing of information and experiences, incentivising the uptake of technologies. The households that owned higher tropical livestock units of indigenous cattle and their improved breeds were more likely to utilize tree and shrub fodder. This result could be attributed to the increased demand for fodder to sustain the productivity of cattle (Kassie et al., 2018; Maina et al., 2020). However, goat ownership did not significantly affect the utilization of trees and shrubs as did cattle. The unexpected negative effect could be explained by the assignment of less importance to goats compared to cattle. Farmers who perceived that tree and shrub fodder could increase milk production were more likely to utilize the technology. This attitude implies that the farmers’ perceived benefits of using trees and shrubs as fodder for livestock were high, consequently influencing the adoption decision.

The farmers’ high-value assignment would explain the finding on the positive attributes in evaluating the quality of the trees and shrubs. This finding is consistent with Meijer et al. (2015), who observed that farmers’ prior knowledge of the benefits of technology creates a positive attitude towards the technology and increases the likelihood of adoption. Similarly, Barnes et al. (2021) observed that farmers adopted legume cultivation technology because it provided fodder to animals during periods of drought besides addressing other constraints, namely poor soil quality and limited access to soil chemical fertilizers. Therefore, trees and shrubs would be considered an attractive option for smallholder farmers in the coming decades.

Analysis of data obtained in the current study showed that an increase in one litre of milk per day is likely to decrease the probability of utilizing trees and shrubs by 2%. This finding suggests that special attention should be paid to trees and shrubs fed to dairy cows. In Eastern DRC, the availability of concentrate feed remains limited and expensive when available (Maass et al., 2012; Mutwedu et al., 2022). High feed prices drive up production costs, leading to high milk prices and making it inaccessible to poor rural households. The likelihood of farmers adopting trees and shrubs increased with the milk sales and milk price. The availability of a milk market in Eastern DRC should incentivise cattle farmers to increase milk production to meet the high market demand for dairy products through the use of trees and shrubs as a source of high quality nutrients able to stimulate high milk production at a low cost thus lowering the milk price and increasing household access to the milk. The use of tree and shrub fodder in addition to natural pastures is a cheaper alternative for smallholder cattle farmers and consequently improves the growth and lactation performance of animals (Maasdorp et al., 1999; Osuga et al., 2011). Jera and Ajayi (2008) found that the adoption of tree-based fodder technology provided a market for smallholder dairy cows in Zimbabwe.

The results from this study highlight to the effect of the agro-ecological zone in adopting this technology. Data analysis showed that farmers in Kalemie were more likely to use trees and shrubs than farmers in Kabare. In addition, the location influenced the preference of specific trees and shrubs by cattle farmers. For instance, out of the 62 species listed, 36 were found in Kalemie, while 26 were from Kabare. The location further affected cattle feeding systems as it varied with the environment. For example, in Kalemie, cattle farmers kept more livestock units and depended on free-range pastures or transhumance that allowed them to get enough trees and shrubs fodder directly for their livestock.

Additionally, it was observed that in Kalemie, farmers relied on customary land tenure, i.e. they have the right to own land temporarily. This system does not motivate them to develop sustainable agricultural practices such as planting trees and shrubs. Therefore, this regime does not promote good cohabitation between cattle keepers and crop farmers. Promoting the adoption of trees and shrubs in livestock production would have to be considered alongside revising the land tenure system in Kalemie.

Land scarcity and population pressure limit cattle farmers from accessing grazing pastures; therefore, using trees and shrubs is an alternative. This was evidenced by the findings from the Kabare area, which had a higher adoption level compared to Kalemie. Similar results were found in the South-Kivu province of the DRC by Bacigale et al. (2014) and Muoni et al. (2019). The authors reported that increasing land pressure and degradation have changed the context by reducing grazing land, subdividing farms, and increasing conflicts between cattle keepers and farmers, leading cattle farmers to rely on zero-grazing systems, tethering systems, collecting green forages from fields, and roadside systems for livestock feeding. Therefore, smallholder farmers require other sources for livestock feeds, and trees and shrubs are a viable option. Besides higher adoption of trees and shrubs technologies in Kabare, farmers also used more exotic or introduced species (58%) than farmers in Kalemie (12%). Mugumaarhahama et al. (2021) observed that in the South-Kivu province of the DRC, farmers used the little space they had for growing trees and shrubs species and food crops because of increased demographic growth that has led to land scarcity. This finding shows the importance of taking into account the particularity of each agro-ecological zone in promoting the use of trees and shrubs to increase its dissemination and adoption among smallholder cattle farmers.

4. Conclusions and policy recommendations

This study provides the factors for adopting trees and shrubs as livestock feeds in Eastern DRC. The findings show that the farmers utilized 62 trees and shrub species as fodder. Farmers cultivate trees and shrubs across the boundaries, within food crops, and in off-farm niches. Besides being cattle fodder, the trees and shrubs are important in providing fuelwood, construction materials, medicinal benefits, and soil fertilization. The farmers listed regrowth as one of the preference factors in the use of trees and shrubs fodder; however, further research is needed focusing on its regenerative capacity to ensure a win-win situation for both livestock and the environment. Policymakers need to provide both male and female farmers with additional incentives, such as training on the benefits of trees and shrubs, to scale up their use and promote it. They can also promote an appropriate extension strategy that motivates smallholder farmers to include trees and shrubs in the cattle feeding plan. The findings also show that cattle farmers in Kalemie are more likely to utilize the trees and shrubs because they are transhumant and do not have the right to land. Policymakers should promote the right to access land and enable cattle farmers to develop sustainable agricultural practices and avoid land conflict in the coming decades. The farmers evaluate the trees and shrubs, employing animal-related and tree-related criteria. Despite the farmer’s knowledge, it is crucial to determine the nutritional quality and feeding impact of trees and shrubs used by smallholder cattle farmers to appropriately recommend their utilization as livestock feeds. This study reveals that gender, group membership, ownership of cattle, ownership of improved cattle, milk production perception, milk sales, milk price, legume production perception, crop residue usage and farmer location positively and significantly influenced the likelihood of adopting trees and shrubs.

Disclosure statement

No potential conflict of interest was reported by the author(s).