Post-Harvest Losses Along the Main Value-Added Chains and Strategies for Reduction in the Soybean Sector in Benin ^†

Abstract

The aim of this study was to provide information on post-harvest losses (PHLs) along the main value-added chains in the soybean industry, and on farmers’ strategies for reducing these losses in Benin. Data were collected using a subjective method (questionnaires) from 152 direct actors in the soybean value chain and an objective method (technological monitoring using technological monitoring sheets) from 27 processing units using a two-stage sampling technique in eighteen of the country’s municipalities. The data were analyzed using descriptive statistics. The results revealed that post-harvest losses in soybean value chains ranged from 0.03% to 7.98%. The highest loss percentages were 7.98% (production), 10.0% (processing), and 4.27% (marketing), and were obtained during the subjective measurements. The link most affected by PHL was the production link. These losses were mainly observed during threshing, winnowing, sorting, and harvesting operations. The main causes of these losses were poor mechanization and a lack of skills for post-harvest operations. The main strategies developed by players to cope with this situation remained traditional (timely harvesting, use of appropriate equipment, and good storage practices). However, additional efforts in terms of infrastructure, adapted subsidies/credits, and recruitment of specialists are needed to mechanize the operations most sensitive to losses and to reinforce the capacities of agricultural actors in order to considerably reduce post-harvest losses.

1. Introduction

Population growth, which is more prevalent in developing countries, is driving up food demand. Paradoxically, more than half the food produced is wasted or lost in post-harvest operations [1]. One of the crucial strategies to enhance food security and to promote sustainable development mainly in rural areas is to mitigate such wastes and losses. Indeed, reducing post-harvest losses (PHLs) in this way would make a lasting contribution to fighting hunger, raising incomes, and improving food security in developing countries [2]. However, the FAO and the World Bank estimated that between 10% and 20% of the volume of grain produced in sub-Saharan Africa is lost every year, for an amount of USD 4 billion corresponding to the minimum annual needs of around 48 million people [3]. In Benin, preliminary surveys suggest a considerable soybean PHL due to inadequate storage and processing infrastructure, though precise figures remain undocumented. This underscores the need for urgent and well-targeted strategies that can boost post-harvest efficiencies.

In Benin, soybean is a highly important commercial crop. It accounted for around 62% of the national legume production [4,5]. Production rose from 57,000 tons in 2009 to 253,954 tons in 2020, for a total soybean export value of USD 66.6 million in 2020 [6]. This production makes Benin the second largest soybean producer in West Africa after Nigeria [7]. Soybeans rank fourth among Benin’s export crops, after cotton, pineapple, and cashew nuts [8]. Soybean has emerged as a pivotal sector in enhancing rural livelihoods and stimulating economic activities across Benin. Moreover, this legume plays an important role in reducing malnutrition, as it is a source of vegetable protein for human and animal consumption. In rural areas, given its richness in magnesium, vitamins, and mineral salts, soy is a substitute for meat, fish, and eggs [9]. Due to its nutritional diversity and economic importance, soybean is a core component of the nutritional and socio-economic framework of rural communities.

Given soybean’s place in Benin’s socio-economic sphere, finding a solution to reduce or even prevent PHL is imperative. Beyond farmers, PHL in soybeans jeopardizes the livelihoods of traders, processors, and rural households who are dependent on affordable protein, amplifying economic vulnerabilities across Benin’s agrarian economy. However, the available literature on this crop focuses more on soil fertility improving properties [10,11,12], improved varieties developed [13,14], variety adoption factors [15,16], and its contribution to improving farmers’ livelihoods [17,18]. This specific focus reveals a significant research gap concerning the critical post-harvest stage of the soybean value chain. Thus, there is a lack of knowledge regarding the aspects related to post-harvest losses of soybeans in Benin. Soybeans have been the subject of very limited research on post-harvest losses despite the critical importance of generating information to identify sensitive stages requiring intervention. Such information is also useful in political decision making and designing effective strategies to reduce both quantitative and qualitative losses. Addressing this gap is then imperative for the formulation of evidence-based policies that can transform the efficiency of soybean processing and distribution networks in Benin. The present article fills this gap in the literature on PHL in the soybean sector. Its aim was to provide insights into the percentages and critical loss stages of post-harvest losses (PHLs) along the main soybean value chains, as well as farmers’ strategies for reducing these losses in Benin. This study aligns with emerging frameworks on sustainable food systems, viewing PHL reduction as a nexus of economic, nutritional, and social outcomes in rural developing contexts.

2. Materials and Methods

The general methodology for this study was inspired by the FAO [2] recommendations, emphasizing the combination of subjective and objective approaches to measure post-harvest losses (PHLs). A two-stage sampling method was used: purposive sampling (to select municipalities) and random sampling (to select producers, processors, and traders). The choice of municipalities was based on their soybean production levels within the Agricultural Development Poles (PDAs).

2.1. Study Area

The study was conducted in the main soybean-producing municipalities in Benin, distributed across six of the country’s seven Agricultural Development Poles—PDAs (Figure 1). Eighteen municipalities were selected based on their significant roles in various soybean value chains (

Table 1. Sampling procedure and sample size.

PDA	Municipalities	Number of Respondents
		Producers	Processors	Traders
PDA2 « Alibori-Sud, Borgou-Nord et 2KP »	Bembéréké, Kalalé, Banikoara, Sinendé, and Kandi	38	5	5
PDA3 «Atacora-Ouest»	Cobly and Natitingou	40	5	4
PDA4 «Borgou-Sud, Donga et Collines»	Nikki, Parakou, Savè, and Dassa-Zoumé	10	3	2
PDA5 « Zou et Couffo»	Bohicon and Zogbodomey		9
PDA6 «Plateau»	Pobè		5	5
PDA7 «Ouémé, Atlantique et Mono»	Cotonou, Porto-Novo, Dangbo, and Adjarra		9	12
Total	18	88	36	28

). Within each municipality, villages and/or urban neighborhoods were randomly selected for collecting subjective data, while objective measurements were purposively conducted.

2.2. Sampling

Simple random sampling was used to select a sample size of 152 respondents for this study, calculated using Yamane’s formula [19] (Equation (1)). A total of 88 producers, 36 processors, and 28 traders were surveyed for subjective assessments. This sampling method was chosen due to the accessibility, collaboration, and spatial distribution of the actors across the PDAs. The sampling process involved two steps:

• Based on the list of soybean-producing municipalities obtained from the Department of Agricultural Statistics, those with high production levels were selected (as shown in Table 1). The research unit consists of producers, processors, and traders from the selected municipalities.
• The 152 respondents were distributed as outlined in Table 1.

Among the 152 respondents, five soybean processors per value chain unit were selected for technological monitoring to collect objective data, excluding the soybean grain value chain since the practices are nearly identical [20]. The municipalities of Dassa-Zoumè, Zogbodomey, Bohicon, Porto-Novo, and Adjarra were chosen for objective measurements within the processing segment.

$$n_o={\displaystyle \frac{N}{1+N\left(e^2\right)}}$$

(1)

where

• $n_o$: Minimum sample size for each actor category;
• $N$: Population size for each actor category;
• $e^2$: Margin of error, fixed at 5%.

2.3. Selection of the Value Chains

According to the FAO [2], the evaluation of post-harvest losses (PHLs) along value chains must be preceded by the characterization and comprehensive analysis of the chains, which allows the identification of the processes most likely to result in post-harvest losses. The inventory of all the soybean value chains (SVCs) led to the identification of nine SVCs: (i) soybean grain SVC, (ii) soybean cheese SVC, (iii) soybean flour SVC, (iv) soybean milk SVC, (v) soybean mustard (afitin) SVC, (vi) soybean oil SVC, (vii) soybean cake SVC, (viii) soybean biscuit SVC, (ix) soybean meal SVC [21,22]. For the selection of SVCs in this study, two criteria were used: the priorities defined in the National Soybean Development Program (PNDF) and the significance of the SVCs within the soybean sector in terms of their contribution to local and even national wealth creation. Consequently, four SVCs were selected for this study: soybean grain SVC, soybean cheese SVC, soybean flour SVC, and soybean milk SVC.

2.4. Data Collection and Analysis Methods and Tools

Four data collection techniques were employed. First, a literature review grid was used to conduct a document review. Second, an exploratory survey was carried out using a semi-structured interview guide with around twenty respondents randomly selected by SVC outside the main sample. The key points of the interviews and the information gathered focused on the main value chains and the stakeholders involved in each sector, the stages and operations prone to post-harvest losses (PHLs) at each SVC link, storage and preservation practices, the causes and/or factors contributing to post-harvest losses, the estimated values of these losses, and strategies for reducing them. The results from this phase were used to finalize the questionnaires and objective measurement sheets for the actual data collection process. Thirdly, a quantitative survey was conducted using a structured questionnaire to collect subjective data from soybean producers, processors, and traders. Finally, a technological monitoring sheet was used during participatory observations (technological monitoring) within processing units to collect objective measurements. During these technological monitoring sessions conducted in March 2022, quantitative losses were recorded at each processing operation. This combination of subjective and objective approaches is one of the FAO’s recommendations for measuring post-harvest losses (PHLs). For each SVC, three repetitions were carried out during the technological monitoring in different municipalities. The measurement equipment, consisting of a 50 kg capacity scale (Zhongshan Camry Electronic Co., Ltd., Guangdong, China) with 1 g precision (Figure 2), was used to measure the initial and final quantities of the product at each operation. At the end of each transformation process step, the quantities of products left behind or discarded by the processor were collected and weighed using the 1 g precision scale (Yongkang Nengzhi Industry & Trade Co, Yongkang, China). These collected or discarded quantities then represent the objective losses recorded at that operation.

The collected data were subjected to statistical analysis using IBM SPSS Statistics v.25.0 software. This analysis involved calculating frequencies, means, and standard deviations.

2.5. Determination of Post-Harvest Loss (PHL) Indicators

The loss indicators were calculated following the FAO’s recommendations for the different stages and selected value chains [23]. Thus, both quantitative (or weight-based) losses and relative losses (loss rate) were determined. Quantitative losses, expressed in kilograms, were collected from agricultural operations (subjective measurements) and obtained through objective (or physical) measurements.

Relative losses (loss rate) are expressed as percentages and measure the intensity of losses occurring during various harvesting and post-harvest operations. They are obtained by dividing the quantity lost at each operation by the quantity handled. For instance, the drying loss rate is calculated by dividing the quantity of grains lost during drying by the quantity of grains dried (

Table 2. T Formulas for determining PHLs by level.

Levels	Formulas		Variables
Operation	$TPO={\displaystyle \frac{QPO}{QMO}}\times 100$	(2)	TPO: Rate of loss during operation QPO: Quantity lost during operation QMO: Quantity handled during operation
Value chain	$TPCVA=\sum _{i=1}^{n}TPO$	(3)	TPCVA: Loss rate per value chain n: Number of value chain transactions TPO: Loss rate per transaction

Source: [22].

). The use of the quantities handled (in this case, the quantities subjected to drying) as the denominator ensures that the loss rates are between 0 and 100 percent (Formula (2)).

The determination of the most sensitive stages to losses was performed. The value chains most sensitive to losses were evaluated using Formula (3). For each value chain, losses were determined by summing the losses recorded at each individual operation.

3. Results

The socio-economic profiles of the respondents, the key values and loss hotspots across the value chain stages of the soybean sector, and the strategies employed by stakeholders to mitigate post-harvest losses constitute the main findings of this study.

3.1. Socio-Economic and Demographic Characteristics of Respondents

The socio-economic and demographic characteristics are presented by the value chain stage or stakeholder category (producers, processors, and traders).

3.1.1. Soybean Producers

Soybean production was predominantly a male-dominated activity (87.1%). Most soybean producers were married (92.1%) and lacked any form of literacy (76.1%) (

Table 3. Socio-economic and demographic characteristics of soybean producers.

Variables		PDA2	PDA3	PDA4	Total
Gender	Female	15.79	12.50	0	12.50
	Male	84.21	87.50	100.00	87.50
Age (mean)		43.66 ± 12.61	36.48 ± 9.21	49.4 ± 14.62	41.05 ± 12.18
Marital status (%)	86.84	95.00	100.00	92.05
	2.63	5.00	0	3.41
	10.53	0.00	0	4.55
Average contribution of soybean to farm income (%)		36.20	55.70	48.00	46.40
Literacy level (%)	None	71.05	85.00	60.00	76.14
	Ability to read	0	2.50	0.00	1.14
	Ability to write	0	2.50	10.00	2.27
	Ability to read and write	28.95	10.00	30.00	20.45
Average number of dependents		12.82 ± 9.26	5.95 ± 3.27	14.8 ± 6.37	9.92 ± 7.69
Average number of farm workers in the household		4.32 ± 3.04	2.78 ± 1.98	4.3 ± 2.58	3.61 ± 2.64
Years of experience in soybean production		9.24 ± 8.64	4.8 ± 2.91	9.5 ± 3.86	7.25 ± 6.50
Membership of cooperative or producer group (%)		10.53	0.00	0.00	4.55

). The respondents had, on average, seven years of experience in soybean production and were generally not affiliated with any soybean producer organization (4.6%). Additionally, soybean production contributed slightly less than half (46.4%) to the agricultural income of producers. This contribution was slightly higher for producers in PDA3 (55.7%). Across all the surveyed PDA zones, about four out of ten household members were engaged in agricultural activities.

3.1.2. Soybean Processors

Soybean processing was primarily undertaken by women (94.4%) (

Table 4. Socio-economic and demographic characteristics of soybean processors.

Variables		PDA2	PDA3	PDA4	PDA5	PDA6	PDA7	Total
Gender	Female	100	100	66.67	88.89	100	100	94.44
	Male	0.00	0.00	33.33	11.11	0.00	0.00	5.56
Age (mean)		41 ± 12.18	32.8 ± 4.96	53 ± 9.64	43.67 ± 8.37	39.8 ± 7.43	46.60 ± 11.33	42.08 ± 10.21
Marital status (%)	Married
	Single
	Widowed
Average contribution of soybean processing to agricultural income (%)		60.0	68.0	36.66	72.22	74	74.44	67.77
Literacy level (%)	None	19.23	15.38	0.00	23.08	15.38	26.92	72.22
	Ability to read	0.00	0.00	33.33	33.33	0.00	33.33	8.33
	Ability to read and write	0.00	14.29	28.57	28.57	14.29	14.29	19.44
Years of experience in soybean processing		8.80 ± 4.54	5.88 ± 3.80	11.67 ± 4.16	9.22 ± 2.11	9.80 ± 5.45	16.0 ± 15.64	10.67 ± 8.70
Membership of processors organizations (%)		0.00	0.00	100.00	22.22	0.00	0.00	13.89
Main final product sold (%)	Soy cheese	80.00	100	33.33	77.78	80.00	44.44	69.44
	Soy milk	0.00	0.00	33.33	11.11	0.00	0.00	5.56
	Soy porridge	0.00	0.00	0.00	0.00	20.00	55.56	16.67
	Soy flour	20.00	0.00	33.33	11.11	0.00	0.00	8.33

). Both male and female processors were adults, with an average age of 42 years, and approximately two-thirds were illiterate. The respondents had an average of 11 years of experience in soybean processing, which contributed about 68% to their agricultural income. The main final products obtained from processing were soybean cheese (69.4%) and soybean porridge (16.7%).

3.1.3. Soybean Traders

The majority of the traders were illiterate (82.1%), and three-quarters (75.0%) were women (

Table 5. Socio-economic and demographic characteristics of soybean traders.

Variables		PDA2	PDA3	PDA4	PDA6	PDA7	Total
Gender	Female	60.00	75.00	50.00	100.00	75.00	75.00
	Male	40.00	25	50.00	0.00	25.00	25.00
Age (mean)		41.82 ± 2.19	27.50 ± 2.08	54.50 ± 7.77	37.80 ± 2.17	46.08 ± 13.98	41.82 ± 12.19
Average contribution of soybean to agricultural income (%)		53.42	47.50	50.00	64.00	54.10	53.20
Literacy level (%)	None	80.00	100.00	100.00	60.00	83.33	82.14
	Ability to read	20.00	0.00	0.00	0.00	0.00	3.57
	Ability to read and write	0.00	0.00	0.00	40.00	16.67	14.29
Years of experience in soybean marketing		13.50 ± 9.21	5.5 ± 2.65	15 ± 14.14	11.4 ± 4.83	17.00 ± 8.40	13.57 ± 9.11
Membership of cooperatives or organizations (%)		0.00	0.00	50.00	0.00	0.00	3.57

). Their average experience in soybean trading was 14 years across all the PDA zones, although it was relatively lower in PDA3 (5.5 years). On average, soybean trading contributed 50% to the agricultural income of the respondents, and membership in trader organizations was notably low (3.6%).

3.2. Values and Critical Stages of PHL

At the three stages of the soybean value chain, post-harvest losses ranged from 0.03% to 7.98%. The production and marketing stages recorded the highest (7.98%) and lowest (0.03%) loss rates, respectively.

3.2.1. Soybean Production Stage

The losses at the production stage of the soybean value chain are estimated at 7.9% (Figure 3). The highest loss rates occurred during threshing, winnowing, and harvesting (2.0%, 1.7%, and 1.4%, respectively). The lowest loss rates were recorded during sorting, drying, transportation from field to home, storage/packaging, and transportation to market/distribution (0.8%, 0.8%, 0.7%, 0.4%, and 0.1%, respectively).

3.2.2. Soybean Processing Stage

• Soybean cheese value chain

The subjective and objective post-harvest loss (PHL) rates for the soybean cheese value chain are illustrated, with significant losses observed during winnowing (3% for subjective losses and 0.5% for objective losses) and sorting (3.9% for subjective losses and 1.2% for objective losses) (Figure 4).

• Soya milk value chain

The subjective and objective post-harvest loss (PHL) rates for the soya milk value chain show that for the subjective measures, losses were observed during grain transport, sorting/winnowing, soaking, filtration/milk extraction, and sterilization, with rates of 2%, 2.8%, 1%, 2.2%, and 4%, respectively. For the objective measures, losses were primarily seen during sorting/winnowing and sterilization, with rates of 1.4%, 1.1%, and 2.4%, respectively (Figure 5).

• Soybean flour value chain

The subjective and objective post-harvest loss (PHL) rates for the soybean flour value chain reveal that subjectively, the stages of sorting/winnowing, roasting, dehulling, and milling were the most prone to losses, with rates of 1.7%, 1%, 1%, and 1.8%, respectively. Objectively, the highest losses occurred during winnowing/sorting (0.7%) and milling (0.4%) (Figure 6).

3.2.3. Soybean Trading Stage

Soybean traders were grouped into retailers, semi-wholesalers, and wholesalers. Figure 7 illustrates the PHL rates during the marketing stage. The semi-wholesalers recorded the highest average loss rate (0.9%) compared to the retailers (0.2%) and wholesalers (0.3%). The semi-wholesalers experienced the highest losses during sorting (6.5%), whereas the wholesalers and retailers recorded lower losses, not exceeding 1%. The wholesalers experienced their highest losses during procurement (0.9%), while the retailers were the least loss-prone, with 0.5% recorded during sorting.

3.3. Strategies for Reducing PHL at Critical Stages

3.3.1. Soybean Production Stage

The majority of the soybean producers (86.4%) identified timely harvesting as the best strategy for reducing PHL (Figure 8). Other strategies included using proper bags (preferably jute) and ensuring that storage areas were free from direct contact with walls, roofs, or other structures to prevent temperature fluctuations. Proper winnowing, timely threshing, and using appropriate tarpaulins were also cited as effective measures.

3.3.2. Soybean Processing Stage

Using appropriate equipment (66.7%), mechanizing operations, and completing tasks quickly (45.5%) were the most used strategies by the processors (Figure 9).

3.3.3. Soybean Trading Stage

To reduce PHL in this sector (Figure 10), the traders made more use of good storage practices (39.3%). This included ventilation if the stocks were to be used as seed. To reduce losses even further, good collection was necessary (28.6%), with the aim of sorting good grain from bad. It was essential to use bags that were in good condition (25.0%) in order to reduce transport losses. On-time delivery (21.4%) reduced losses when the customer was satisfied on time. It was also very important, according to some retailers, to protect the bags with tarpaulins during transport (21.4%) to prevent them from deteriorating during the journey.

4. Discussion

Reducing post-harvest losses (PHLs) is an effective method for ensuring food security. It also enables farmers in disadvantaged regions to fully exploit the potential of their production.

4.1. Values and Critical Points for PHL

The study revealed that the operations most responsible for post-harvest losses in soybean production, regardless of the chain segment (production, processing, and marketing), include threshing, winnowing, sorting, and harvesting. These results align with Atiim [24], who found that more than half of the soybean farmers assessed in Ghana identified the aforementioned steps in soybeans as sensitive.

Similarly, research by Paulsen et al. [25] on the causes of post-harvest losses in developing countries identified threshing and winnowing as the most vulnerable stages of PHL. This issue is due to the lack of mechanization in post-harvest operations. According to these authors, the mechanization of operations with skilled labor can reduce soybean post-harvest losses by 75% to 100%. In Benin, producers still rely on traditional equipment and materials for these operations, and these activities remain largely manual.

It is important to note that winnowing and sorting operations are critical points for losses at all levels of the soybean production chain. This highlights the issue of insufficiently qualified personnel dedicated to these tasks, which also contributes to higher loss rates, as noted by Paulsen et al. [25]. The work of Arends-Kuenning et al. [26] corroborates this observation, showing that a lack of skill in using harvesters was the main cause of post-harvest losses in Brazil.

However, contrary to the findings of Caixeta-Filho and Péra [27], the present study did not identify transportation as a critical point for post-harvest losses. This may be attributed to policy efforts aimed at improving the condition of roads connecting various regions in the country or the proximity of markets.

4.2. Strategies for Reducing Critical Points for PHL

This study highlights several significant challenges in minimizing post-harvest losses in the soybean production chain. First, although timely harvesting is recognized as an effective strategy, the use of inappropriate tools during harvesting is a major cause of losses. This indicates the need to create and promote appropriate harvesting tools that minimize product damage. Additionally, processors wish to improve the quality of equipment and mechanize various steps in the processing process. However, these improvements require significant investment, which is beyond the capacity of most actors in the chain. As demonstrated by Paulsen et al. [25], most non-mechanized threshing/cleaning systems in developing countries like Benin lack adequate means for preserving harvested grains. These results underscore the need to develop financing and support mechanisms that make these investments more accessible.

It is evident that the current approaches implemented by traders are not sufficiently effective in reducing post-harvest losses. Several factors may contribute to this situation. On the one hand, the harvesting methods themselves may not be effective enough to reduce post-harvest losses. This suggests that a combination of methods is needed to be effective. For example, research by Coradi et al. [27] evaluated different sustainable strategies for managing soybean grain mass based on moisture content, optimizing drying and storage operations to improve grain flow and quality in large-scale storage units. They found that combined drying and dry aeration storage systems were the best option for achieving optimal soybean quality. Their experiments demonstrated that low-temperature drying and storage produced the best quality results for the grains. On the other hand, traders may lack the necessary knowledge or training, which could lead to inadequate implementation. Furthermore, external factors such as environmental or economic conditions may make it difficult to effectively apply these techniques. Therefore, it is crucial to continue research to identify the precise causes of this inadequacy and develop strategies for improvement.

5. Conclusions

Post-harvest losses are a major cause of economic and financial losses in the soybean sector, which is the fourth largest export crop in Benin. The aim of the study was to identify the most vulnerable stages and assess strategies for reducing post-harvest losses (PHLs) in the Benin soybean value chain in Benin. The findings revealed that the stages most critical stages along the value chains are in proportions ranging from 0.03% to 7.98%. Specifically, production recorded the highest losses at 7.98%, while marketing recorded the lowest at 0.03%. Threshing, winnowing, sorting and harvesting operations are the most sensitive to PHL. These losses are mainly due to inadequate mechanization and the lack of qualified personnel for post-harvest activities. The main coping strategies developed by stakeholders remain traditional.

Policymakers can set up an integrated post-harvest conservation and storage system that spans from the farmer’s field to the final consumer. This system could include appropriate and accessible harvesting and storage equipment or infrastructure, paired with a well-trained workforce skilled to conduct post-harvest operations efficiently. It would be desirable for future research to look at the other links not covered here such as transportation, packaging, and quality losses, and to include medium and large soybean-processing companies in Benin.