Background

In Sub-Saharan Africa, the urban population is increasing at an astonishing rate (Saghir and Santoro 2018). For example, in Uganda, the urban population has grown by about 4.5 million between 2002 (2.9 million) and 2014 (7.4 m) (UBOS 2016) accelerating the amount of waste generation (Aryampa et al. 2019) and the demand for food (Sabiiti and Katongole 2016, 2014). Effective management of municipal solid waste (MSW), especially organic waste is a challenge to many urban areas in Uganda (Komakech 2014a, b). Most solid wastes are burnt, dumped indiscriminately in landfills leading to pollution of the air, soil and water bodies, posing health and livelihood risks to many urban dwellers (Komakech et al. 2014b). Treatment of the organic fraction of the MSW through composting is one feasible and cost-effective method of reducing waste volumes while releasing vital nutrients for the soils of developing countries (Komakech 2014a; Danso et al. 2017; Nigussie et al. 2015; Tweib et al. 2011). According to Tweib et al. (Tweib et al. 2011), composting can be carried out with little capital and operating costs. For example, a recent economic estimation by Romero et al. (Romero et al. 2013) in Spain showed that the production cost of raw compost leachate as fertilizer could be as low as 0.67€ per litre compared to 4.61 € per litre for commercial fertilizer.

There is potential for organic waste and MSW compost utilization and management through urban agriculture in Uganda (Komakech 2014a; Nsimbe et al. 2018). Treatment of MSW is important to reduce the overuse of chemical fertilizers in agriculture since they cause severe pollution of water resources (Diacono and Montemurro 2011; Ojo et al. 2014; Savci 2012; Udeigwe et al. 2015) and a decrease in the amount of soil organic matter (Massah and Azadegan 2016; Roba 2018; Wu et al. 2020). Moreover, results from the annual agricultural survey carried out in 2018 (UBOS 2018) indicate that 77% of the households in Uganda use organic fertilizers (e.g., livestock manure, crop residues and ash) to improve agricultural production. The use of the derived organic compost represents an appropriate win–win waste management strategy and sustainable soil fertility management practice (Kumar et al. 2020; Rogger et al. 2011), especially for farmers in urban areas (Komakech 2014a; Menyuka et al. 2018).

With the aims to address the mounting solid waste management problems in Ugandan municipalities, reduce greenhouse gas emissions from open dumping landfills and use the generated compost as a safe alternative fertilizer for crop production, the Uganda National Environment Management Authority (NEMA) with support from the World Bank and the government of Netherlands initiated a municipal solid waste composting project in 2005 under the Clean Development Mechanism (CDM) in nine municipalities including Mukono, Jinja, Mbale, Soroti, Lira, Mbarara, Kasese, Kabale, and Fort Portal in the first phase of the project (Lederer et al. 2015; Okot-Okumu and Nyenje 2011). As part of the cost-recovery plan, each municipality was expected to generate revenue through the sale of the recyclables such as plastics to urban residents and composts to its farmers. To sustain the functionality of these compost plants, there is a need to understand whether or not urban farmers utilize these MSW compost as a soil conditioner in agricultural production and the factors that influence its utilization. Studies on the use and determinants of use of agricultural technologies (Melesse 2018; Mwangi and Kariuki 2015) and soil fertility improvement (Abebe and Debebe 2019; Ali et al. 2018; Babasola et al. 2018; Nazziwa-Nviiri et al. 2017) have been done in Uganda, Ghana, Nigeria and Ethiopia. These studies have shown that socio-demographic factors (Mwangi and Kariuki 2015; Abebe and Debebe 2019; Babasola et al. 2018) influenced the use of organic fertilizers. In Uganda, previous studies have mostly involved the use of homemade compost mainly from manure and crop residues, kitchen refuse (Nsimbe et al. 2018), chemical fertilizers (Nazziwa-Nviiri et al. 2017) and fecal sludge (Danso et al. 2017). However, to our knowledge, there is limited evidence about the utilization of MSW compost as a soil conditioner in crop production by urban farmers where CDM plants have been established and associated factors in Uganda. Such information is important in guiding investment decisions by farmers and businessmen in MSW compost reuse businesses through the conversion of waste to organic fertilizers, and urban planning. This study aimed to assess the current utilization of the MSW compost from CDM plants in agricultural production and its associated factors among urban farmers in Mbarara and Fort Portal municipalities of Uganda. Urban areas in this study have been defined as those areas under the municipality administration according to the Uganda National Urban Policy (MoLHUD 2017).

Methods

Study area

The study was conducted in Fort Portal and Mbarara Municipalities of western Uganda. These municipalities were purposively selected because both have the highest rates of organic fertilizer use (39.6%) as reported by UBOS (UBOS 2018), have CDM project compost plants, with each plant capable of composting between 2.3–3.5% of the wastes collected per day (NEMA 2017). The mean annual rainfall and mean annual temperature in Mbarara is 832 mm and 21 °C, respectively, and soils are mostly Luvisols and acidic clay loam (Kaizzi 2014; Wortmann and Eledu 1999). Fort Portal, receives a mean annual rainfall of 1310 mm, with a mean temperature of 20 °C (Wortmann and Eledu 1999), and the soils are characterized as Lixic Ferralsols (Wortmann and Eledu 1999; Okello-Oleng et al. 2021).

Sampling design and sample size

A cross-sectional survey was conducted between February and April 2019 to collect primary data from farming household heads or their spouses. All the divisions and villages in the study municipalities were included in the sampling frame. From each village, a list of farmers, generated with the help of local council chairpersons was used as a sampling frame. We determined the sample size for the number of households to participate in the survey using the Krejcie and Morgan (1970) table, covering both MSW compost users and non-users. Individual urban farmers (farmers above 18 years) from each village were randomly selected and interviewed using a semi-structured questionnaire (Nigussie et al. 2015). The questionnaire was pretested with 20 farmers from Fort Portal Municipality and appropriate modifications were made. The questions asked included the socio-demographic characteristics of the household heads such as age in years, gender, educational level, marital status, land size, land ownership, land tenure, farming experience, access to extension services, access to infrastructure, access to credit facilities, engagement in non-farm activities, annual household income from farming, farmer experience with compost, number of household members and access to other soil conditioners (e.g., livestock manure, crop residues and ash). During the interview process, individual farmers were asked whether or not they utilized MSW compost in their farms. Furthermore, farmers were asked whether or not they were willing to pay for MSW compost and cost per Kg they were willing to pay for such compost. Field observations and discussions with key informant respondents (17 from Fort Portal and 32 from Mbarara) were used to supplement the household interviews.

Study variables

The dependent variable in the study was the level of utilisation of the MSW compost by farmers (1 if utilizing MSW compost, 0 otherwise, Table 1). The independent variables were chosen based on previous studies (Abebe and Debebe 2019; Babasola et al. 2018; Akpan et al. 2012; Huang and Karimanzira 2018; Zhou et al. 2010) and they included socio-demographic factors (e.g., gender, age, marital status, level of education, number of people in the household, farming experience, access to extension services, access to good infrastructure, access to credit facilities, access to markets, membership of farmer association, engagement in non-farm activities, land ownership, farm size, the land tenure system, level of income, access to other soil conditioners (e.g., livestock manure, crop residues and ash), experience in the use of compost, willingness to take part in compost making and cost per Kg of compost the farmer was able to pay.

Table 1 The description of the variables used in the study and their expected signs
Full size table

Data analysis

Primary data was entered into Excel and imported into IBM SPSS version 25 and STATA version 14. Descriptive statistic using percentages was used to express the level of utilization of MSW compost. A probit regression model was used to analyse factors that influence respondents’ utilization of MSW compost. Multicollinearity among the explanatory variables used in the model was tested using correlation matrices (Nigmatullin 2008). We determined the statistical significance at a p-value ≤ 0.05. In the probit model, the categorical (usually dichotomous) dependent variable was modelled as a linear (or log-linear) function of a combination of explanatory variables (Noreen 1988). The probit model assumes that while we only observe the values of 0 and 1 for the variable Y, there is a latent, unobserved continuous variable Y* that determines the value of Y (Sebopetji and Belete 2009). The probit model was preferred over the logit model because it includes believable error term distribution as well as realistic probabilities (Nagler 1994). The Probit model is specified as follows:

Let us suppose Yi is a binary response variable with only two possible outcomes (1 if the farmer is utilizing MSW compost and 0 otherwise).

Consider also a vector of independent variables xi which is assumed to influence Yi.

Then the probit model takes the form:

$$\Pr \, \left( {Y_{i} = 1| \, x_{i} } \right) \, = F\left( {\beta ^{\prime}x_{i} } \right) \, = \, \Phi \left( {\beta ^{\prime}x_{i} } \right)$$
(1)

where Pr denotes the probability that an individual uses compost or not, Yi is the binary choice variable representing use and Φ is the cumulative distribution function of the standard normal distribution. β is a vector of unknown parameters.

It is assumed that the latent variable Yi* can be specified as follows:

$$Y_{i} * = \beta_{0} + \sum\limits_{n = 1}^{N} {\beta_{n} x_{ni} + u_{i} }$$
(2)

And that:

Yi = 1 if Y* > 0.

Yi = 0 otherwise.


where xi represents a vector of explanatory variables, ui is a random disturbance term, N is the total sample size, and β is a vector of unknown parameters to be estimated by the method of maximum likelihood.

Model specification;

$$\begin{aligned} {\text{Yi}} = & \beta 0 + \beta {\text{1X1}} + \beta {\text{2X2}} + \beta {\text{3X3}} + \beta {\text{4X4}} + \beta {\text{5X5}} + \beta {\text{6X6}} + \beta {\text{7X7}} + \beta {\text{8X8}} + \beta {\text{9X9}} + \beta {1}0{\text{X1}}0 + \beta {\text{11X11}} \\ & + \beta {\text{12X12}} + \beta {\text{13X13}} + \beta {\text{14X14}} + \beta {\text{15X15}} + \beta {\text{16X16}} + \beta {\text{17X17}} + \beta {\text{18X18}} + \beta {\text{19X19}} + \, \beta {2}0{\text{X2}}0 + {\text{ui}} \\ \end{aligned}$$
(3)

where,

Yi = is the probability of utilization of MSW compost (1 if the farmer is utilizing MSW compost, 0 otherwise); X1, gender of household head (male = 1, otherwise 0); X2, age of the household head (year); X3, marital status (1 if couple, 0 otherwise); X4, education of household head; X5, household size; X6, access to extension services; X7, access to infrastructure; X8, access to credit facilities; X9, access to market for produce; X10, membership of farmer association; X11, annual income from farming in Uganda shillings; X12, engagement in non-farm activities; X13, farming experience; X14, land ownership (1 if land is available, 0 otherwise); X15, total size of farmland in acres; X16, household land tenure system; X17, household access to other soil conditioners (1 if household has access, 0 otherwise); X18, experience with compost fertilizer use in years (1 if the household applied compost, 0 otherwise); X19, cost at which farmer is willing to pay for use of MSW compost as a soil conditioner (1 if willing to contribute money, otherwise 0); X20, cost per kilogram of compost; β0 is the regression coefficient and ei is random error term. The model estimates marginal effects of an explanatory variable on the expected value of the dependent variable and the coefficients are more informative and useful for policy decision making (Anang 2016).

The marginal effects are estimated by differentiating Eq. (1) with respect to xi according to Greene (Greene 2000).

$$\frac{\partial y}{{\partial x}} = \phi \left( {\beta^{\prime}x_{i} } \right)\beta_{i}$$
(4)

where ϕ represents the probability density function of the standard normal distribution.

Results

Socio-demographic characteristics of respondents and utilization of MSW compost

The study revealed that among the study participants in Fort Portal, the majority were females (56.5%), were aged above 36 years (56.2%), were married (71.5%), had no experience in composting (73.7%), had access to extension services (67.6%), had access to other soil conditioners (60.1%), had access to infrastructure (good roads and storage, 94.2%), had access to credit facilities (74.2%) and markets (93.6%), were engaged in non-farm activities (78.8%), owned land (90.3%), and had less than two acres of land (57.6%). Nevertheless, only 35.5% were members of farmers’ associations. In Mbarara, the majority were males (51.8%), married (65.7%), had secondary level education (34.4%), had a household size of 3–6 persons (45.5%), had access to a soil conditioner (62.7%), had access to extension services (60.4%), had access to good infrastructure (87.5%), had access to credit facilities (70.8%) and markets (83.6%), had prior experience in composting (44.0%), earned an annual income of less than 200,000 Uganda shillings (56 US dollars) from farming (46.8%), were involved in non-farm activities (66.6%) and owned land (77.2%) of less than two acres (33.1%). Only 30.6% were members of the farmer associations and had a farming experience of more than 10 years (24.6%) (Table 2). Also, in this study, only 1.7% of the farmers in Fort Portal municipality utilized MSW compost compared to 2.2% of the farmers in Mbarara municipality.

Table 2 Socio-demographic characteristics of the study participants and utilization of MSW compost in Fort Portal and Mbarara municipalities, western Uganda, 2018 (n = 359 for Mbarara and 361 for Fort Portal)
Full size table

Factors associated with utilization of MSW compost

The probit regression results of factors influencing farmers’ utilization of MSW compost in Fort Portal and Mbarara municipalities are presented in Table 3. The likelihood ratio chi-square (LR Chi2) for the model was statistically significant at 1% level of significance, which revealed the model had good explanatory power. The validity of the Probit model in estimating utilization of MSW compost is consistent with a related study from Ghana (Danso et al. 2017). Results from the Probit regression model showed that five out of the 20 variables included in the model were positively significant in influencing the utilization of MSW compost in Mbarara, whereas four variables were positively significant in influencing farmers’ decision to utilize MSW compost in Fort Portal (Table 3). In Fort Portal, the factors that were positively significant in influencing farmers’ decision to use MSW compost were the level of income from farming activities, access to other soil conditioner, the experience a farmer had in using a soil conditioner. The cost of compost was negatively significant in influencing farmers’ decision to use MSW compost. However, in Mbarara, the level of income from farming activities, access to other soil conditioners and the experience a farmer had in using a soil conditioner significantly and positively influenced farmers’ decision to use MSW compost. Membership of a farmer association and the cost per Kg of compost were the factors that significantly and negatively influenced the decision to use MSW compost (Table 3).

Table 3 Probit regression results of factors influencing farmers’ decision to use MSW compost in Fort Portal and Mbarara municipalities, Uganda
Full size table

Discussion

The level of utilization of CDM compost found in our study (1.7% for Fort Portal Municipality and 2.2% for Mbarara Municipality) were much lower than what was reported (25%) for municipal solid waste compost in urban and peri-urban areas of Accra, Ghana (Danso et al. 2006). The generally low utilization of the compost in urban areas might be explained by constraints associated with urban compost use (e.g., poor quality of compost including broken bottles and polythene, price and transportation costs, high water requirements of plants following compost application and absence of markets for organically produced crops), negative perceptions about compost as reported by Danso et al. (2006) among urban and peri-urban farmers in Accra, Ghana, and a lack of awareness about the benefits of compost for soil fertility (Dandeniya and Caucci 2020). From a practical point of view, these findings as well as our study emphasize the need to strengthen extension programmes and training on compost use, provision of long-term compost price subsidy program through operational tax waivers and provision of transport to resource-constrained smallholder farmers especially in areas with weak commercial fertiliser distribution networks as already being implemented in Western Kenya (Makau et al. 2016) and improving quality through compost segregation in urban areas of Uganda.

The Probit regression model showed how the significant factors would predict the future use of the MSW compost by farmers. The findings indicated that a unit increase in a farmer joining a farmer association in Mbarara leads to a 7.5% decrease in the probability of farmers’ decisions to use MSW compost, contrary to the expected benefits of group membership, where farmers are empowered to achieve higher yields. This is contrary to a study carried out in Nepal (Kumar et al. 2020) where membership to cooperative associations positively influenced farmers’ decision to use improved agricultural technologies and extension services. The study also showed that a unit increase in the level of income from farming increases the chances of a farmer using MSW compost by 3.7% and 3.6% in Fort Portal and Mbarara, respectively. A farmer earning from his produce is motivated to invest in productivity improvement technologies (e.g., purchase of inputs), resulting in improved utilization of compost (Singh et al. 2016). Furthermore, a unit increase in accessibility to other soil conditioners lead to a 19.4% and 19.1% increase in the likelihood of farmers using MSW compost in Fort Portal and Mbarara, respectively. This means that farmers who have access to other types of soil enhancers have a high likelihood of using MSW compost. However, a previous study conducted in Mbarara district in 2013 shows that farmers are likely to abandon the use of fertilizers because of ease of accessibility to livestock dung (Kasirye 2013). This study also showed that an increase in the price of the MSW compost would lead to a 16.4% reduction in level of farmer’s decision to use MSW compost in both Fort Portal and Mbarara municipalities. This is because most farmers already had a fixed amount of money (fee set by the municipal council at the launch of the compost plants) they were willing to pay for the MSW compost. Therefore, any increase in the amount beyond the municipal price may lead to a decrease in its use. This finding is similar to studies of Blessing et al. (2010) in Imo State, Nigeria which found that the price of fertilizer was a significant determinant of a farmer’s decision to adopt or not to adopt the use of fertilizer. Finally, this study also revealed that the experience a farmer had in using any soil conditioner had a positive significant relationship with farmers’ decision to use MSW compost. This finding is consistent with studies conducted in Brazil (Morello et al. 2018) where the level of experience with fertilizers had a positive and significant effect on fertilizer adoption. An increase by a year of experience in the use of fertilizers by a farmer leads to a 4.3% increase in the level of farmer decision to use MSW compost in both Fort Portal and Mbarara municipalities. Farming experience increases the likelihood of farmers using the compost as experienced farmers have much knowledge and also information about the success and failure of using fertilizers than farmers with less experience.

Conclusion and policy implications

This study explored the level of utilization and the factors affecting the farmers’ decision to use MSW compost from the CDM plants in two urban areas of Uganda. Results showed that the level of utilization of MSW compost in both municipalities is generally low. Results obtained from the probit model showed that annual farm income, access to a soil conditioner, experience with the use of fertilizers positively influenced farmers’ decision to use MSW compost while the price of MSW compost and membership in a farmers’ group negatively influenced farmers’ decision to use MSW compost. The negative influence of group membership on the use of MSW compost in Mbarara, should be a concern to policymakers in a country that currently invests in agricultural extension through groups such as that under the Operation Wealth Creation Scheme. Overall, the findings of this study suggest the need for government to create an enabling environment by investing in policies aimed at improving accessibility to soil conditioners, creating opportunities that maximize income from households, interventions focused on the farmer-to-farmer experience sharing, improved dissemination of information through extension programmes and other innovative communication approaches harnessing the benefits of digital tools, and long-term subsidy schemes for the price of compost through operational tax waivers and provision of transport for farmers in urban areas of Uganda. Finally, to guarantee quality and to improve adoption of compost generated at the CDM plants in Uganda, there is a need for research on supply chain analysis and CDM plant operators to invest in research on the quality of MSW compost generated at their facility, conduct cost–benefit analysis of compost application and therefore set a price commensurate with quality, and develop guidelines and rates of application of the MSW compost. Furthermore, a market research using the new product design (Urban and Hauser 1993) should be conducted focusing on farmer perception, preference and adaptive conjoint measurement (tradeoff) analysis. This analysis will help the CDM plant operators to tailor their products and services to different customer segments in quality (price setting, etc.) and therefore address the issue of famer perception about fertilisers.