Farmers’ Typologies for the Construction of a Technological and Socioeconomic Baseline of Industrial Cassava Organizations to Guide Research, Production, and Policy Design in Colombian Caribbean Region

Abstract

The cassava production system in Colombia is relevant for food security and agro-industrial development, especially in the Caribbean region. To optimize this system, it is necessary to understand the technical and socioeconomic particularities of the crop, which facilitates cost improvements and increased yields. Two main market segments are recognized in this country, cassava for human consumption and for industrial purposes (starch and animal feed). There is still no comprehensive study on the characterization of industrial cassava production systems; therefore, this work aims to identify typologies of farmers according to their technological and socioeconomic characteristics, to contribute to research and production planning. The information related to sociodemographic and technological conditions was collected through a structured survey applied to 53 organizations selected by random sampling that compiles consensual information of 1470 producers. The data were analyzed with descriptive statistics and a multivariate analysis of variables, such as cultivated area, yield, and proximity to markets, using the Gower method. The results indicated that in the Caribbean industrial cassava production, the average age of farmers is 47 years, with mostly men (72%), and a predominant educational level of secondary school (35%). Two groups of farmers were identified from multivariate analysis, from which group 2 (G2) represented four producer organizations (119 producers) and showed the highest yields and a larger cultivated area, suggesting that a larger scale of production is associated with better yields, which is associated with major access to technological tools and possibly technical knowledge. However, the most cassava producers (49 producer organizations and 1361 producers) had a lower cultivated area and yield, confirming that the technological level and management optimization is related to competitiveness and profitability.

1. Introduction

Cassava (Manihot esculenta Crantz) is a crop widely cultivated in tropical regions of the world. According to FAOSTAT 2024 [1], Nigeria was the world’s leading producer in 2022, with a production of 60,835,540 t. The Democratic Republic of the Congo ranked second with 48,774,623 t, followed by Thailand with 34,068,005 t, Ghana with 25,592,014 t, and Cambodia with 17,698,783 t. In terms of yield, Ghana led with 24.4 t/ha, followed by Cambodia with 23.3 t/ha and Thailand with 21.46 t/ha. In contrast, the Democratic Republic of the Congo and Nigeria achieved yields of 8.15 t/ha and 6.06 t/ha, respectively.

Cassava is a plant cultivated in tropical and subtropical regions, recognized for its excellent ability to adapt to unfavorable agroecological conditions and infertile soils [2]. Globally, the average yield of cassava is 12 t/ha of fresh roots; however, yield exceeding 60 t/ha can be achieved using optimal planting material and proper crop management [3,4]. In Colombia, total cassava cultivation extends across 32 of the 34 departments or provinces. According to the authors of [5], the cultivated area was estimated at 244,688 ha, with a total production of 2,440,638 tons and an average yield of 11.6 t/ha. The region with the largest production area is the Caribbean region, which covers approximately 133,809 ha, representing 54.68% of the total planted area. Among the main producing departments in this region are Bolívar, with 44,472 ha, followed by Córdoba (29,324 ha), Sucre (28,080 ha), Magdalena (21,372 ha), Cesar (12,920 ha), Atlántico (5795 ha), and La Guajira (4766 ha) [6].

Similar to several regions around the world, cassava production in Colombia is divided into two large segments. The first is recognized as sweet cassava, or cassava for human consumption [7], which corresponds to 92% of the cultivated area, while industrial cassava is used mainly to obtain starch, flour, and derivatives, and represents 8% of the total cultivated area (19,633 ha) [5,8]. Cassava production for industrial purposes is concentrated mainly in the Colombian Caribbean region (82%), which has climatic and soil conditions that are appropriate for this crop [8]. In 2023, in this region, the departments with the largest areas of industrial cassava were Sucre (10,982 ha), Córdoba (4310 ha), and Bolívar (779 ha) [5]. Regarding the average yield, the higher yield is found in Cordoba (18.83 t/ha), followed by Sucre (15.8 t/ha) and Bolivar (8.75 t/ha) [5].

The conditions in Colombia related to cassava cultivated area, yields, and associativity of industrial cassava producers are important elements that define differences among Caribbean departments. Therefore, studies to characterize production systems are essential to establish a technological and socioeconomic baseline [9], to determine the gaps that exist in the production and obtained quality, which weaken the cassava value chain. These measures seek to recognize the comprehensive scenario of agricultural systems from the perspective of livelihoods, mainly based on aspects related to human, natural (agriculture), physical, financial, and social capital [10]. This information allows us to identify the zero period or the current state of the cassava production system, especially the technical and economic characteristics, which facilitates the definition of more efficient production models and the development of technologies that optimize technical coefficients and return indicators. This, in turn, makes it possible to evaluate the effects of incorporating technologies during interventions [11]. Consequently, interventions by actors dedicated to rural development are more effective and generate a greater positive impact by accurately identifying implementation strategies under previously studied and integrated conditions [10]. Generally, technological interventions are related to high-yield improved varieties, use of fertilizers, weed management, and soil management, such as crop rotation and types of mechanization, among others [12].

The diversity of characteristics among farmers makes it difficult to analyze them individually, which in many cases prevents the identification of common strategies for scaling up, which are essential to achieve a greater impact in technological intervention [13]. Multivariate analyses allow the classification of producers or farms, and the detection of limitations that guide interventions, both in research and extension [13,14,15,16]. The classification of producers or farms, based on socioeconomic, biophysical, and geospatial factors, and the implementation of interventions adapted to smallholder farming systems, have proven to be effective and aligned with local requirements. This classification facilitates the formulation of more precise and relevant recommendations, optimizing the results of interventions [17,18].

In the case of cassava, advances in the characterization of this crop system made it possible to identify different types of producers or production systems. In Rwanda, the analysis based on 60 variables, which included demographic, agronomic, economic, and specific variables, on 390 cassava producers, resulted in 7 production typologies. Farmers more oriented to commercial-scale cassava cultivation had better access to formal sources of seeds; however, most farmers in all typologies accessed new varieties and quality cassava seeds through informal channels [19].

In general, there are different deterministic, probabilistic, participatory, and expert-based methods that have been widely applied to characterize agricultural systems [20], and although in cassava, so far, only the previously described report was found, which used multivariate methods that consecutively reduced the dimensionality of the data by applying principal component analysis (PCA) and the determination of subgroups through cluster analysis [19], other strategies have been identified that have been used to identify the main gaps and avenues for technical intervention in cassava cultivation worldwide. The determination of factors associated with productivity and their decomposition into technical efficiency change and technological change has also made it possible to identify key variables in the cassava production system to be considered in technical intervention [21,22,23].

Despite the importance of the cassava crop in Colombia, reports related to the characterization of this productive system are limited. The results of the analysis of 306 cassava producers from the department of Cauca in 2014 and 395 producers from the departments of Bolívar, Sucre, Córdoba, and Magdalena in 2017 indicated that the majority of the producers surveyed identify themselves as small farmers, with agriculture being their main occupation, with little or no access to extension services, and low yields were observed since low rates of adoption of technologies (improved varieties) were identified. Additionally, these households presented high vulnerability and low levels of education [24]. Similar conclusions were presented in the analysis of the cassava value chain in Colombia, recognizing the inefficiency of extension systems and access to technologies as the main weaknesses, which are deeply connected to socioeconomic and environmental factors, among others [25]. The classification of industrial cassava producers is recognized as a fundamental basis for generating intervention recommendations to improve production systems. For this reason, the objective of the present study was to identify typologies of farmers belonging to producer organizations, according to their technological and socioeconomic characteristics, to further contribute to research and production planning.

2. Materials and Methods

2.1. Study Design

The research was carried out during the year 2022 through characterization of 53 organizations of cassava producers with vocation for industrial cassava cultivation and that represent 1470 producers, which were randomly selected. Through field work, a structured survey was applied with the purpose of defining groups of producers that share similar characteristics from the socioeconomic and technological point of view to generate an approximation to the concept of recommendation domain. This study is in accordance with Rodríguez Osuna [26], who proposed the use of the deterministic sampling technique when dealing with small samples with a small universe or population where the general characteristics are known.

2.2. Study Area

The study was carried out in different municipalities of the departments of Córdoba, Sucre, and Bolívar, including the representative municipalities with cultivated areas of industrial cassava in the Caribbean region. According to the Köppen classification [27], in this subregion the climate is predominantly tropical savanna with dry winter (Aw), with an average annual temperature of 27 °C with alternating periods of drought (summer) and rain (winter) throughout the year [28]. The average annual rainfall is 1500 mm, and the relative humidity is between 75% and 85%, with an increase in the second half of the year [28]. In total, 53 surveys were applied, distributed as follows: Córdoba with 8 surveys in 3 municipalities, Sucre with 35 surveys in 17 municipalities, and Bolívar with 10 surveys in 6 municipalities (Figure 1).

2.3. Sociodemographic Characterization

Several sociodemographic variables were used to characterize the farmers that belong to organizations, and the registered data corresponded to consensus information that represented the majority of each organization. The variables included age, gender, educational level, experience in cassava production, and land property.

2.4. Methodological Framework for Farmers’ (Organization) Typologies Determination

The data grouping technique was applied by clusters, using variables such as cultivated area (ha), years of experience as a farmer, fresh cassava root yields (ton/ha), distance to the crop sales site (km), and price per ton of cassava. The objective was to form groups of farmers who shared similar conditions, highlighting the differences between the groups instead of within them. The purpose was to group individuals (observations) based on common characteristics, highlighting the differences between groups. It was essential to establish a similarity or distance index that allows reaching precise and reliable conclusions, thus reflecting reality more effectively [29].

To determine the contribution of each factor to cassava yields in the Caribbean region, a linear regression model was formulated. In this model, yields, expressed in tons per hectare, were considered as the dependent variable, and distance to the sales site, the age of the farmers, and their experience as independent variables. The mathematical model was formulated as follows:

Y = β₀ + β₁X₁ + β₂X₂ + β₃X₃ + ε1

where:

• Y = yield, in tons of cassava per hectare,
• X₁ = distance to the sales site, in kilometers,
• X₂ = age of farmers, in years,
• X₃ = experience, in years as a cassava farmer,
• β_i = parameters to estimate regression coefficients for each i variable,
• ε1 = disturbance term considered.

One of the most effective methods is the Gower similarity coefficient. When applied, the similarity between groups can be determined by measuring shared characteristics that are quantifiable [28]. Multivariate statistical analysis techniques, particularly the Gower similarity coefficient, constitute powerful tools to assess the similarity between individuals by allowing the simultaneous manipulation of quantitative and qualitative variables. This approach not only facilitates the grouping of individuals into homogeneous groups, but also highlights the differences between groups, allowing for more accurate and effective classification [30,31,32].

According to De la Fuente [29], the analysis begins with a single group that includes all cases, and through successive divisions, smaller and smaller groups are formed. This involves grouping farmers based on characteristics such as the size of the productive unit, the area cultivated in cassava, and the yields in kilograms per hectare per year. To carry out this process, Ward’s hierarchical clustering method was used through the SPSS v 8.0 program (Equation (1)):

$$S_{ij}={\displaystyle \frac{\sum \begin{array}{c}P1\\ h=1\end{array}\left(1-\frac{X_{ih}-X_{ih}}{Rh}+a+\alpha \right)}{P_1+\left(P_2-d\right)+P_3}}$$

(1)

where:

• S_ij is the Gower similarity coefficient,
• P₁ is the number of continuous quantitative variables,
• P₂ is the number of binary variables,
• P₃ is the number of qualitative (non-binary) variables,
• a is the number of matches (1, 1) in the binary variables,
• d is the number of matches (0, 0) in the binary variables,
• α is the number of matches in the qualitative (non-binary) variables,
• Rh is the range of the h-th quantitative variable.

The application of Ward’s hierarchical clustering multivariate analysis technique allowed the identification of two clearly defined groups of farmers in the cassava production system in the Caribbean region of Colombia. These groups were differentiated by the total area of the farm, the area dedicated to cassava cultivation, the yields in tons per hectare of fresh root per year, and the distance to the sales site in meters. In addition, technical variables related to farming activities and socioeconomic variables, such as infrastructure, available services, and the state of the roads, were considered.

3. Results and Discussion

3.1. Socioeconomic Characteristics of Industrial Cassava Producers

The characterized organizations were integrated into several cassava producers (Figure 2). Sucre was the department with the higher number of organizations, with a total of 35 and a sum of 668 members, with the municipalities of San Marcos and San Antonio de Palmito, with a total of 4 organizations each, being those that presented the highest sum of active members, followed by Corozal, San Pedro, and San Juan de Betulia. It was followed by the department of Bolívar with 10 organizations and a total of 294 members, with San Juan de Nepomuceno being the municipality with the greatest representation, with 4 associations and a total of 150 active members. The department of Córdoba, with 8 organizations and 210 affiliates, was the department with minor organizations, of which 80% corresponded to the municipality of Ciénaga de Oro with 6 organizations. The present distribution is consistent with the cassava cultivated area [6], with the department of Sucre having the largest cassava area, followed by the municipality of Ciénaga de Oro in Córdoba. These results showed that several cassava farmers are organized in associations, and these models improve productive development at the agro-industrial level [33]. Associations manage to minimize the weaknesses that they would have individually and enhance their strengths when they are grouped with others, and this generates support in design, development, and business management [34]. These results coincide with those of Mendoza et al. [35], who identified three organizations for the department of Sucre with 152 active members dedicated to the cultivation of industrial cassava, which have influence in these municipalities.

As a component of the study, the sociodemographic characteristics of the surveyed sample were defined with variables such as age, education, and gender, which are presented in

Table 1. Socioeconomic characteristics of industrial cassava producers in the Caribbean region of Colombia.

Socioeconomic Status	Frequency	Percentage
Age (Average in years)	47	100
Gender
Male	39	72
Female	15	28
Educational level
None	3	6
Primary	14	26
Secondary	19	35
Technic	10	19
Professional	6	11
Postgraduate	2	4
Experience in cassava production	13	100
Land property
Family	1	2
Own	3	4
Leased	50	93

.

Among the sociodemographic variables, the age of the farmers was, on average, 47 years old, which shows that most of them were of mature age; however, it can be inferred that there is a guarantee of stability in the production system since the minimum age is 25 years. The male gender predominated, with 72% of the surveyed sample, however, 22% indicated the importance of female roles in the production systems. These results are similar to those reported by Ariza-Carcía in 2022 [36], who in a financial literacy work with cassava producers in the department of Sucre found that respondents between 18 and 30 years old are 10%, those between 31 and 40 years old are 25%, those between 41 and 50 years old are 15%, and those over 50 years old are 50%. Thus, technical efficiency in yam production among farmers in southern Adamawa State in Nigeria found that older producers were more efficient than younger ones, due to the experience acquired with age [37]. In contrast, Adanguidi and Amegnaglo [38] indicated that experience in agricultural production and age reduce the inefficiency of producers, due to young producers being more receptive to innovation in their system and because the generational change is fundamental in any organizational process, where the experiences and knowledge of older producers must be transmitted to the new generations in charge of continuing with the processes, creating synergies between the traditions and the innovation and development of technologies.

Most farmers received secondary education, with the highest percentage (35%) of the sample surveyed, and significant participation was seen at levels such as technician, with the presence of farmers with a professional degree, and two of the interviewees stated they had postgraduate degrees. These results contrast with those reported by the authors of [36], who found a maximum level of education of the respondents for cassava producers in the department of Sucre of 27.5%, 20%, 32.5%, and 20% for primary, secondary, tertiary, and other forms of education (such as learning a special skill), respectively. Rivera and other authors [24] found that cassava producers in Colombia have several differences and similarities between regions and agricultural systems, referencing the head of household being predominantly male, between 51 and 58 years old, and with an average of around 3 and a half years of formal education in the Caribbean region and 5 years in Cauca, with few households headed by women (6% for the Caribbean region and 12% for Cauca) among cassava producers. On average, farmers registered 13 years of experience in cassava cultivation, which indicates that it is feasible to accumulate the farmers’ experience to project the crop. Previous studies showed that a higher level of education encourages farmers to adopt and use innovations more effectively [39]; meanwhile, another study revealed that the aging of human capital negatively impacts total agricultural productivity by slowing technological advancement and decreasing the efficiency in the allocation of agricultural resources [40]. However, the experienced producer has more skills to decode explicit knowledge and adopt and implement innovations in his production system, hence the importance of strengthening knowledge and inter-institutional support that strengthen the productive chains of industrial cassava in the Caribbean region.

Field leasing was the most predominant modality among farmers who grow cassava, with a participation of 93% of those surveyed. It is striking that 7% were landowners, and they generally acquired the land through succession or inheritance. The lack of land ownership is a limitation for access to credit, especially when one is not part of a system of advance contracts or letters of intent to purchase [25]. These results contrast with those found by Rivera and other authors in 2021 [24], who reported that 69% of productive land was owned by households, with Sucre being the department with the highest proportion of owned land (85%), followed by Bolívar (70%), and rented land being the second most common option, with 21%. These results showed that this could be the main constraint related to the low technological level in this agricultural system in this region.

Work carried out by Coronel and Ortuño in 2005 [41], in which they classified productive systems in Argentina, found that, from the socioeconomic perspective of the producer, it can be pointed out that 44% of the producers had been leading the same system for between 40 and 50 years. Only 47% resided on the rural farms, while the rest commuted daily from a nearby urban center. Two-thirds of the farmers had no higher education or had only completed primary school. Further, 47% of the farmers covered their economic needs with the income generated exclusively from rural activity, while the remaining 53% obtained additional income from formal transfers, such as retirement or pensions, or from other economic activities, such as the agricultural industry and commerce, categorizing these into two types of producers, which coincides with the results obtained in this study.

The average size of productive units dedicated to cassava cultivation was 50 hectares, with a range that varied between 2 (40 farmers) and 540 hectares (45 farmers; Figure 3). This finding demonstrated the wide differences among crop systems, since 1.46 hectares was the average reported in Nigeria [42]. However, here, it is important to remember that industrial purposes have different ways to sell the production, in contrast to cassava for human consumption, which are mainly connected to unique or few purchasers that are simultaneously processors. Furthermore, the information obtained on the cultivated area exceeded that reported by Itam in 2018 [43], which indicated a range of 0.5 to 0.9 hectares. This suggests that, while Itam referred to small-scale producers, our research focused on medium-sized farmers, which highlights a crucial difference in the productive structure of cassava in this region that must be recognized to technologically strengthen this crop system. Producers with a larger surface area in production tend to adopt higher levels of innovation and knowledge because of higher market requirements. On the other hand, land owners can count on a greater probability of investment, because there is a guarantee that the investment will return to the same productive unit in the medium or long term. The opposite happens with tenants, where they limit themselves to supplying the crop with what is necessary during the rental period, without considering the sustainability of the production system.

3.2. Cassava Farmers’ Typology Determination in Producer Organizations

The results showed that the independent variables were consistent in explaining the returns. The estimation of parameters was consistent, with significance at the 10% level. This implies that, in none of the 53 observations was the value of the variables expected to be zero, suggesting that the variables included in the model were relevant for its explanation. It is important to note that the estimated parameters comply with the desirable properties of linear estimators, such as linearity, unbiasedness, minimum variance, and exogeneity. These characteristics ensure that there is no serial correlation in the errors, which reinforces the validity and reliability of the estimated model. The estimation of the parameters to explain the quantities of cassava produced, expressed in kilograms per hectare, allowed us to appreciate that the distance to the sales site positively influenced the quantities produced, or there was a direct relationship between the distance to the sales site and the decisions to produce cassava (

Table 2. Ordinary least squares regression of the variables that explain cassava yields for farmer typification.

Variable	Coefficient	Standard Error	T Statistic	Probability
C	0.60	0.970	0.63	0.54
Distance to market	0.19	0.098	2.0	0.05
Age	0.27	0.020	1.32	0.10
Experience	−0.05	0.051	−1.03	0.05

). Even if production was between the fence and the sales site, the probability of increasing the quantities of cassava produced was 19.5%. In the case of age, the relationship was also direct: a 1% increase in age would generate a 27% increase in cassava production. The experience variable, measured in years, presented an inverse relationship, from which it can be understood that with a 1% increase in years of experience, the probability of increasing the quantities of cassava was −5%. It can also be seen in the contributions of the explanatory variables, such that cassava production was explained in 30% by experience in cultivation, age in 18%, and distance to the sales site in 5%, but in all cases the independent variables explained cassava production.

Hierarchical analysis resulted in two groups, which were defined based on the characteristics of farm size, area cultivated in cassava, yields, and technological level (

Table 3. Socioeconomic and technological characteristics of cassava production in the Caribbean region of Colombia according to farmers’ typology.

Farmer Characteristics	Statistical Measurement	Group I	Group II
Gender	%	Male (64)	Male (63)
		Female (36)	Female (37)
Age (years)	Ẋ	46	48
Educational level	Frequency	None (5.45)	None (0.21)
		Primary (26)	Primary (0.98)
		Secondary (34)	Secondary (1.34)
		Technician (18)	Technician (0.70)
		Professional (11)	Professional (0.42)
		Postgraduate (4)	Posgraduate (0.14)
Experience as cassava producer (years)	Ẋ	11	36
Social characteristics
Distance to market	Meters	163	135.4
Technological level
Cassava crop area in the farm	Ẋ	33	250
Yield, t/ha	X	12	15
Economical return
Net income COP/ha		896,600	2,186,600
Profitability, %	Ẋ	21	51
Benefit/cost ratio		1.21	1.51

COP: Colombian Pesos; %: percentage; Ẋ: media.

). Group one (G1) grouped 49 producer organizations, with productive units with a total farm area average of 39 hectares, an area cultivated in cassava of, on average, 33 hectares, and with yields of 12 tons per hectare per year, who managed the crop with a medium technological level, corresponding to 92.4% of the sample surveyed.

Group two (G2) was formed by 3 organizations from Cienaga de Oro municipality (Cordoba) and 1 organization from Corozal municipality (Sucre), with productive units with a total farm area of 250 hectares, in which cassava crop was the total of this farm area, and with yields of 15 tons of fresh cassava roots per hectare per year, who managed the crop with a high technological level, corresponding to 7.5% of the surveyed sample.

Differences between G1 and G2 were more evident in technological variables than in socioeconomic variables, since both gender and schooling remained in similar proportions between the groups. Other previous studies identified several groups with similar characteristics on each farm. Among the similarities found were the proportion of men, 78% in [44], while for this study it was 64%, and 63% for groups 1 and 2, respectively. The age of the farmers was similar in both cases, at 57 years in [44], which indicates advanced age with low participation of young people in both cases, male and female, while in this study it was 47 years. In addition, the study in [44] reported that 16% of farmers had completed primary education, 60% had completed secondary education, 14% had a higher technical level, 1% had higher education, and 9% had no education at all. Rivera and other authors [24] found a large proportion of respondents with incomplete primary education. This proportion ranged between 28% and 48% for the departments of Cauca and Magdalena, respectively. The proportion of family members in the study sample who had completed post-secondary education (technical or university) was relatively low in all regions. The department of Bolívar again showed the lowest indicators, with only 3% of the working-age population surveyed and more with high-level education (technical only). In contrast, in this research, 6% of farmers had no formal education, 35% had completed secondary school, 26% had completed primary school, 19% had completed technical studies, 11% were professionals, and 4% had postgraduate degrees. These data indicate a higher level of education among cassava farmers in the Caribbean region compared to those reported by Anzules.

Significant differences between the two identified groups were related to technological characteristics, among which fresh root yields were a significant distinguishing factor. In Group 1 (G1), which included most farmers, yields reached an average of 12 tons per hectare, consistent with the historical average for the region according to Martínez [45]. In contrast, Group 2 (G2) presented yields of 15 tons per hectare, suggesting that the technological tools they have involved provide better conditions for cassava production. This development in recent years requires a higher level of capitalization, which explains the lower number of producers compared to G1. Compared with production of cassava for human consumption, these results are slightly higher than those reported by Rivera in 2021 [24], who found that the average yield for the entire sample was 7.5 t/ha and explained that the variation by production areas or type of producer could be due to the differences between the yield of improved and local varieties used respectively in each crop system. Moreover, we analyzed the area dedicated to cassava cultivation, which in the case of G1 was 33 hectares on average, while in G2 it rose to 250 hectares. This difference in the extension of cassava crops suggests that farmers in G2 adopted an entrepreneurial mentality, thus distancing themselves from the family farming and peasant economy approaches that predominated in G1. These results contrast with those reported by Rivera in 2021 [24], who observed few differences in the size of the land dedicated to cassava production, with almost 1.87 ha above the average for the Caribbean region and suggesting that most cassava producers are small producers for consumption and sale in local markets. In addition, they reported the presence of few large cassava producers, identifying two large producers with plots of 50 and 60 ha in Córdoba.

When comparing the results of cassava production in the Caribbean region with those of Polo-Murcia and other authors in 2021 [46], the main difference was that in this case, three groups of producers were identified, each with their respective recommendation domains. What makes this study similar is the purpose, which consisted of evaluating the characteristics of each group to advance in the definition of the recommendation domains and evaluate the adoption of technologies once they were implemented in the production system.

In the analysis of economic returns, significant differences were identified between the two groups, attributable to variations in yields. With an average cost of Colombian Pesos (COP) 4,263,400/ha, the net income of G2 significantly exceeded that of G1. Specifically, average yields for farmers characterized as G2 reached 15 ton/ha, generating a net income of COP 2,186,600/ha, which translates into an advantage of COP 1,290,000/ha compared to G1. These findings reflect the efficiency and effectiveness of resource management in G2.

These results showed clear differences between the two groups. Farmers in G2 had an entrepreneurial mindset and a greater capacity for capitalization. They used recommended technologies that optimized their resources and increased their profits. In contrast, farmers in G1 used conventional technologies and combined their resources inefficiently, which negatively impacted their yields and return indicators, which is consistent with what was reported by Rivera and others in 2021 [24], who found low use of fertilizers (41%) and pest control (28%) in cassava production systems in the Caribbean region, and the level of agricultural assets used for cultivation was low, suggesting limited mechanization of agricultural production by producers. The difference among producers enriches the development of the territories [47,48], since techniques, processes, procedures, and forms of production or organizational systems are related to types of farmers. These results demonstrate that differences among farmers should be considered in strategies to promote rural development. The specific interactions along productive chain actors, entities, and institutions should be focused on producers and clients’ relations, thus building territorial systems of innovation, encouraging agreements between public and private actors that finally strengthen the cassava cultivation and transformation. The results obtained in this study coincide with those obtained by Frugoni [49] in Uruguay with the inclusion of rural development in agricultural public policies, indicating that after categorizing producers, it can be defined that for them to be inclusive, rural development policies should be integral, to create development policies and to promote actions of integration and organization of the social fabric and rural productivity. This must be framed by each type of producer according to their conditions. Martinez [50] stated that, throughout the 20th century, public policies toward cooperativism were promoted in Colombia, achieving the definition of norms, institutions, plans, and projects for its promotion and supervision. However, these public policies are rarely applied to the agricultural sector, and the knowledge and typology of the producers are not considered for productive development.

4. Conclusions

Comparing the levels of education among cassava farmers in the Caribbean region indicated a significant advance in academic training, with a decrease in the percentage of farmers without formal education and an increase in those with professional degrees. This progress has facilitated the adoption of agricultural technologies and contributed to improved yields in cassava production.

The results of the analysis indicated the differences in two groups of industrial cassava producers: G1 focused on family farming practices, obtaining yields that aligned with historical trends in the region, while G2 adopted a more industrialized and entrepreneurial approach, which translated into higher yields. However, the capitalization barriers and the size of cassava cultivated areas were linked to the production approaches and investment levels. G1, which operated on a smaller scale, faced constraints that affected its capacity for innovation and, therefore, its yields. In contrast, G2, by adopting an entrepreneurial model and farming on a larger scale, had access to better technologies and markets, allowing them to maximize their yield in their cassava fields.

Differences in economic returns were found in the groups. G2, with its entrepreneurial approach and use of appropriate technologies, achieved significantly higher yields. To improve their results, G1 farmers must adopt new technologies and optimize the management of their resources, which is essential to increase their competitiveness and profitability. However, government policy development related to land availability and rural development (to strengthen technologies and knowledge) could improve the current status of most cassava producers.