A Simplified Integrative Approach to Assessing Productive Sustainability and Livelihoods in the “Amazonian Chakra” in Ecuador

Abstract

This study integrates the Sustainable Livelihoods Framework (SLF) and the Sustainability Assessment of Food and Agriculture Systems (SAFA) to evaluate the sustainability and livelihood dynamics of the Amazonian Chakra system, recently designated as a Globally Important Agricultural Heritage System (GIAHS) by the FAO. Using data from 330 producers across three associations (Kallari, Wiñak, and Tsatsayaku) in the Ecuadorian Amazon, the study employed discriminant analysis to assess governance, environmental integrity, economic resilience, social well-being, and livelihood capitals. Results revealed significant disparities across associations in key sustainability dimensions. Kallari and Wiñak demonstrated stronger governance, environmental integrity and economic resilience, linked to mature organizational structures and effective governance mechanisms. In contrast, Tsatsayaku excelled in demographic diversity and larger landholdings but lagged in governance and environmental practices. Extreme poverty affected 82% of households, with Tsatsayaku having the lowest rate (69%) compared to Wiñak (89%) and Kallari (87%). Chakra income contributed significantly to livelihoods, accounting for 44% of total income in Kallari, 37% in Wiñak, but only 16% in Tsatsayaku, whose producers relied more on off-farm activities and livestock. The integration of SLF and SAFA methodologies offered a nuanced understanding of sustainability, highlighting the importance of governance, financial strategies, and environmental conservation in promoting resilience. Policies should prioritize participatory governance, market transparency, and credit access to address disparities and strengthen sustainability. These findings underscore the critical role of the Amazonian Chakra as a sustainable agroforestry system, providing economic and cultural benefits, while emphasizing the need for tailored interventions to enhance the sustainability of Amazonian producer associations.

1. Introduction

The sustainability of small-scale agriculture plays a crucial role in long-term food production, and its success depends not only on the availability of natural resources but also on how these resources are managed by the farmers and communities involved [1,2]. To enhance our understanding of this complex dynamic, the integration of the Sustainable Livelihoods Framework (SLF) is essential [3,4,5]. This framework emphasizes the role of multiple livelihood capitals—human, social, natural, physical, and financial—and how those capitals can be mobilized within a given institutional context to influence agroforestry practices and outcomes. By incorporating this holistic perspective, we can better grasp the socio-economic, organizational, and external factors that shape decision-making processes and impact sustainability in small-scale farming systems [6,7].

The concern for sustainability in agriculture emerged in response to environmental issues that began gaining attention in the 1950s and 1960s [8,9]. However, concepts and practices related to sustainability can be traced back to ancient texts from China, India, Greece, and Rome [10]. Today, sustainability is a central concept, particularly when evaluating agroforestry systems, which combine trees and crops in the same space. These systems offer numerous benefits, including biodiversity conservation, climate change mitigation, and soil quality improvement [11,12,13,14], as well as contributing to the diversification of rural incomes [15,16].

A comprehensive evaluation framework is essential for managing sustainability across agricultural, forestry, and livestock systems, whether collectively or individually. This framework should be cyclical, adaptable, and participatory, fostering interdisciplinary collaboration and feedback to identify intersections between environmental, social, and economic processes crucial to sustainability [17,18]. Sustainability assessments play a key role, using models to project the impacts of planned changes [19]. However, a one-size-fits-all approach is impractical, as evaluations must be tailored to specific contexts [20]. Nonetheless, these assessments provide a vital foundation for discussions, reflections, and adaptive decision-making, supporting more informed and resilient sustainability strategies [21].

The SAFA tool, developed by FAO, is used to assess sustainability in food and agricultural systems through four dimensions: good governance, environmental integrity, economic resilience, and social well-being. These dimensions encompass 21 themes, 58 sub-themes, and a total of 116 indicators, scored on a scale of 1 to 5 [22]. Additionally, SAFA aligns with the Bellagio sustainability assessment principles (STAMP), established in 1996 [23], and has been applied across various countries and contexts [24]. However, based on findings from previous research [24,25], we have observed that not all SAFA indicators are well-suited to the context of Amazonian producers. While SAFA is a robust tool, it is necessary to carefully adapt and select variables that better reflect the socio-economic and ecological specificities of EAR. In addition, incorporating SLF [3,5,26] into this process allows for a deeper understanding of how different capitals interact with sustainability goals. This integration creates a more nuanced and context-specific approach to sustainability, which is crucial for smallholder systems such as the Amazonian Chakra.

In this context, the present study aims to propose a simplified integrative approach to assessing productive sustainability and livelihoods in the “Amazonian Chakra” in Ecuador, for which the following partial objectives were developed: (a) select and apply the key indicators from the SAFA methodology that are applicable to the Amazonian Chakra system to evaluate its productive sustainability, as well as the main variables representing the capitals of sustainable livelihoods (human, social, natural, physical, and financial) and income; (b) assess sustainability through the application of these indicators to producers from three associations in the Ecuadorian Amazon, who predominantly utilize the Globally Important Agricultural Heritage System (GIAHS) known as the Amazonian Chakra; and (c) identify those indicators with the greatest discriminating power among the three producer associations in the Amazon. To address these objectives, the study is guided by the following research questions. How do key indicators from the SLF and SAFA frameworks reveal sustainability differences among producer associations? What are the critical livelihood capitals and sustainability dimensions influencing agricultural practices within the Amazonian Chakra system? How can the integration of SLF and SAFA methodologies inform localized policy and sustainability strategies?

The remainder of this article is organized as follows. The theoretical framework of the Amazonian Chakra is reviewed in the next section, drawing on the SLF, the SAFA, and the specific context of the Amazonian Chakra. The materials and methods section details the research design, including data sources, sampling methodology, and statistical analyses. The results and discussion section presents descriptive findings and evaluates the fit of the discriminant model, offering insights into the differences among producer associations. Finally, the conclusions section highlights the key findings and their implications for sustainability and livelihood management.

2. Theoretical Framework

2.1. Sustainable Livelihoods Framework

This study employs the Sustainable Livelihoods Framework (SLF), promoted by the Department for International Development (DFID), a United Kingdom government department, in the late 1990s [3], which highlights the role of various forms of capital in rural development and well-being [3,5]. The SLF provides a comprehensive approach to understanding rural dynamics, addressing basic needs, human rights, and a more qualitative view of poverty [27,28,29] and insecurity [5]. It suggests that livelihoods are shaped by five key assets: human, social, natural, physical, and financial capital. These assets interact with vulnerability factors, such as crises, trends, and seasonality, as well as with institutional and cultural structures that influence livelihoods [30,31].

Additionally, the framework considers the livelihood strategies that individuals or households adopt, which are influenced by the combination of assets available to them and the broader institutional context [3]. These strategies aim to achieve livelihood outcomes, including improved well-being, reduced vulnerability, and sustainable use of natural resources [26,32,33]. This dynamic interaction highlights how individuals constantly adapt to changing conditions by reconfiguring their asset base and strategies.

Similarly, Masud et al. [34] argue that livelihoods extend beyond employment to include social relationships, property rights, and access to essential public services such as water, education, and healthcare. This aligns with Ellis [32], who emphasizes that livelihoods result from the combination of available assets, mediated by institutions and social relations, that enable individuals or households to sustain themselves. In resource dependent communities, access to these capitals plays a crucial role in determining resilience and the capacity to adapt to environmental and socio-economic changes [35].

The SLF also recognizes the feedback loops between different forms of capital; for instance, an increase in natural capital (such as improved access to land or water) can enhance financial capital, while a decline in one type of capital (such as social or natural capital) can erode the sustainability of livelihood outcomes [3]. The SLF is dynamic [36] and adaptable [37], allowing it to be combined with other approaches, such as those used to measure productive sustainability. By integrating elements like good governance and the pillars of sustainability (economic, social, and environmental). This adaptability helps design more effective and locally adapted policies and sustainable strategies, addressing the specific needs of rural communities and promoting long-term resilience.

2.2. Sustainability Assessment of Food and Agriculture Systems (SAFA)

Sustainability Assessment of Food and Agriculture Systems (SAFA), developed by the Food and Agriculture Organization (FAO) in 2013, provides a comprehensive framework to evaluate the sustainability of food and agricultural systems through three key tools. These include detailed guidelines outlining sustainability principles, a set of 116 indicators covering 58 sub-themes, 21 themes, and 4 sustainability dimensions [22], and software that analyzes and visualizes results [38]. The assessment uses a polygonal representation to classify systems across five sustainability levels, from “unacceptable sustainability” (red) to “optimal sustainability” (dark green). This integrated approach enables a rigorous evaluation of strengths, weaknesses, and opportunities for improving sustainability in food production systems, supporting the development of more resilient and sustainable alternatives.

Sustainability assessments benefit from a combination of methods and models to effectively capture the impacts of past or proposed changes within systems [39]. No single framework can fully address the diverse objectives of sustainability evaluation [40]. In this context, the SAFA methodology complements other approaches by providing a structured evaluation of sustainability that can be shaped into specific study areas. When combined with key qualitative and quantitative indicators, such as those from the SLF approach, SAFA enhances contextualization and offers a more holistic understanding of sustainability. These integrations allow researchers to perform more robust multivariate analyses and to explore and compare cases more effectively, identifying key patterns and insights across different contexts. This comprehensive approach strengthens the ability to generate evidence-based policy recommendations aimed at promoting sustainable development.

3. Materials and Methods

3.1. Study Area

The study area is located in the Napo province (Figure 1), within the Upper Amazon region, an area of great biodiversity significance that forms part of the Sumaco Biosphere Reserve (SBR), recognized by UNESCO in 2000 under the Man and the Biosphere (MAB) Programme. This region has been ancestrally inhabited by Kichwa Amazonian indigenous communities, with human presence dating back approximately 400 years. The territory is considered a biodiversity and endemism hotspot [41,42].

3.2. Context of the Amazonian Chakra System

In February 2023, the Amazonian Chakra was recognized by FAO as a Globally Important Agricultural Heritage System (GIAHS), defined as “a sustainable land-use model where productive spaces within farms are managed by families using organic and biodiverse approaches, valuing ancestral knowledge. The Amazonian Chakra, with its biological and cultural diversity, offers multiple services to populations, ranging from food security, ecosystem services, and the preservation of cultural values, to social cohesion and the maintenance of a megadiverse landscape” [43]. In the Napo province, several producer associations currently manage the Amazonian Chakra system (Figure 2a,b) within various agricultural markets. The typical landscape of the Amazonian Chakra system (Figure 2c) integrates diverse agroforestry components, including tall native trees, cacao plantations (visible in the foreground), and understory crops. These arrangements reflect the traditional, sustainable land-use practices of the Kichwa communities, where biodiversity conservation is harmonized with agroforestry production. The combination of native vegetation and cultivated crops represents the ecological and cultural balance achieved in these systems, which are central to our study on sustainability and livelihoods.

This study focuses on three key associations. (a) The Kallari Association, considered large in our study, with 980 producers, 95% of whom are Kichwa Amazonian. The association specializes in the processing and commercialization of products such as cacao (Theobroma cacao L.), vanilla (Vanilla spp.), guayusa (Ilex guayusa Loes.), and chocolate bars. These products are primarily exported to international markets, including the Czech Republic, Germany, France, the UK, and the USA. (b) The Wiñak Association, categorized as medium-sized in our study, with 355 Kichwa Amazonian producers. Wiñak focuses on the commercialization of cacao (Theobroma cacao L.), guayusa (Ilex guayusa Loes.), banana (Musa paradisiaca L.), cassava (Manihot esculenta Crantz), chocolate bars, and ground guayusa. Cacao, guayusa, and banana are the most relevant products, and their markets extend locally and internationally, including Italy, Japan, Mexico, and Spain. (c) The Tsatsayaku Association, classified as small, with 58 mestizo and Kichwa producers. Tsatsayaku primarily markets cacao paste (Theobroma cacao L.), chocolate, and chocolate nibs, with the first two products being the most prominent. Their commercialization is focused on the national market [24].

3.3. Sampling and Data Collection

The data were collected as part of the FAO-Ecuador project “Climate Intelligent Agriculture in Cacao Produced in Agroforestry Systems”, conducted between October and November 2020. Knowing the number of farmers in each of the three associations (small—Tsatsayaku; medium—Wiñak; large—Kallari), probabilistic sampling was employed to determine the sample size. The finite population formula was applied, considering a 95% confidence level and a 5% margin of error. Within each association, random sampling was used to select the rural households to be surveyed, based on a list provided by the president of each association. Initially, a total of 343 households were surveyed: 168 from Kallari, 130 from Wiñak, and 45 from Tsatsayaku [24]. After data cleaning, 13 incomplete records were removed, resulting in a final sample of 330 households distributed as follows: Kallari (156), Wiñak (129), and Tsatsayaku (45 households).

3.4. Determination of per Capita Income and Poverty Index

Per capita income and the extreme poverty level were calculated following the methodology proposed by [29], who determined the proportion of the population in extreme poverty using the Foster–Greer–Thorbecke (FGT) poverty index [44]. The formula used is Po = Np/N, where Po represents the proportion of the sample classified as extremely poor, Np is the number of households in extreme poverty, and N is the total number of households in each producer association. A household is classified as income-poor if it earns, on average, less than USD 2.96 per day, based on Ecuador’s national poverty line reported for 2022. A household with a total income below USD 1.67 per day is classified as extremely poor [45].

3.5. Research Design

To achieve the proposed objectives, a two-stage qualitative and quantitative mixed methodology was used (Figure 3). In the first stage, the relevant index in the Amazonian Chakra system was selected. This stage began with 116 indicators of the SAFA methodology, 15 livelihood indicators, and 5 income indicators of the Chakra (Supplementary Material Tables S1 and S2). The indicators were evaluated and grouped into dimensions based on the bibliography and by Delphi method with a panel of experts (n = 15). Subsequently, indicators were grouped into dimensions to deepen the knowledge of their relationships and the incidence on obtained variability.

Key livelihood variables and sustainability indicators for the Amazonian Chakra system were identified through a workshop involving 15 expert researchers and stakeholders engaged in sustainable development projects. During the workshop, both questionnaires were adapted to the Amazonian Chakra context, and the relevance of each question to the local setting was assessed. As a result, 100 indicators were selected, consisting of 80 from the SAFA questionnaire and 20 from the livelihoods and income survey. The experts evaluated each question using a Likert scale from 1 to 5 [46]. In the first round of evaluation, questions that received the highest score (five) from at least nine experts were retained, while those receiving the lowest score (one) from nine experts were discarded.

To determine the degree of agreement among the experts, the Ishikawa index was applied, which measures the level of consensus, as described in previous studies [47,48]. This index compares the responses given by each expert to each question in the SAFA questionnaire, thereby assessing the level of agreement. The calculation of the proportion of experts who agreed on each response resulted in a concordance value for each question. Subsequently, questions with a concordance level above 60% and an average score higher than 3.5 were selected. This concordance threshold was chosen to ensure that the selected questions reflected a high level of agreement among experts, while the average score served as an additional indicator of the quality of the selected questions.

In a second stage, fieldwork was carried out with two questionnaires and in situ field visits. A total of 330 households of the Amazonian Chakra were visited and two survey models were applied. The first survey characterizes the livelihoods and income of the studied population, using the methodology proposed by the Poverty and Environment Network (PEN) of the Center for International Forestry Research (CIFOR) [49,50]. This methodology employs a prototype survey designed to systematically collect data, allowing the assessment of livelihoods and income across diverse socioeconomic and biophysical contexts [50]. Fifteen variables were selected to represent the five capitals: human, social, natural, physical, and financial. Additionally, five variables derived from income calculations were used to gather detailed household income information, particularly in contexts where significant interactions exist between forests, the Amazonian Chakra, and human activity.

The second survey employed 80 sustainability indicators from the SAFA tool [22], specifically selected for the Amazonian Chakra context by a panel of experts in collaboration with sustainability specialists with local knowledge. Together, these two surveys, comprising a total of 100 indicators/variables, were applied to a random sample of 330 households belonging to three rural associations dedicated to cacao production in the Amazonian Chakra system. This combined approach enables a comprehensive evaluation of the livelihoods and sustainability of these productive systems, identifying opportunities for improvement and promoting sustainable management practices.

Finally, the database was analyzed using descriptive statistics, the Pearson total correlation matrix, and significant differences between associations for each variable. Ultimately, 56 variables were selected for discriminant analysis.

3.6. Statistical Analysis

SAFA livelihood and income variables were standardized according to Gonzalez Martinez et al. [51]. The Kolmogorov–Smirnov test and Bartlett test were performed to verify the normality and equality of the data variance (homoscedasticity). The KMO sampling adequacy test showed a value of 0.7 or greater, while the Bartlett test showed a satisfactory probability value (p < 0.001), thus indicating that the sampling was adequate [51,52].

The quantitative variables were compared using the analysis of variance (ANOVA), establishing the three associations (Kallari, Wiñak and Tsatsayaku) as a fixed effect, with two degrees of freedom. Qualitative variables (original and adjusted) were compared by a Kruskal–Wallis test, with three associations included as a fixed effect.

Later, a discriminant analysis was performed with all the significant transformed variables, including the classification matrix, as well as the graphic representation of the Mahalanobis distances through clusters and the spatial distribution of farms through a canonical scatterplot, establishing the association type as the grouping variable. A direct method of selection of variables was used at p-value < 0.05. Second, the selection of the most discriminant variables was made, applying the F of Snedecor, Wilks’ Lambda, and 1-Tolerance. High values of F for each variable indicate that the means of each group are widely separated and that these groups are homogeneous. Small Lambda values indicate that the variable discriminates well amongst groups. Finally, variables with a high percentage of tolerance (1-Toler) that reduced the redundant information were searched.

4. Results and Discussion

4.1. Livelihoods in the Amazonian Chakra

The analysis of human capital (L-HC) reveals important demographic differences among the associations. Household size averages around 5.2 members with no significant differences across groups, suggesting similar household structures. These findings align with the 5.6 members per household reported by [53] for timber producers in the lower region of Napo province. However, they are lower than the 6.6 members per household reported by [29] in a study on livelihood strategies among Kichwa and mestizo populations in the SBR. Wiñak producers have a significantly higher proportion of female household heads (68.2%) compared to Kallari (57.7%) and Tsatsayaku (51.1%) (p < 0.05), likely due to the Chakra system being traditionally managed by women to ensure food security, even though today it also includes marketable products like cacao, coffee, guayusa, etc. In terms of ethnicity, Wiñak (97.7%) and Kallari (94.2%) have a much higher proportion of Kichwa population than Tsatsayaku (73.3%) (p < 0.001), underscoring stronger indigenous cultural ties in the first two associations, which reinforces the role of the Amazonian Chakra for these populations. Additionally, Wiñak has a significantly younger average household head age (43.8 years) compared to Kallari (51.8 years) and Tsatsayaku (50.5 years) (p < 0.001), which may affect innovation and decision-making. Education levels are similar across all groups, with no significant differences (

Table 1. Mean of the main livelihoods and economic characteristics of livelihoods of producers’ associations, RBS, Ecuadorian Amazon.

Livelihood Variables	Mean n = 330	Kallari n = 156	Wiñak n = 129	Tsatsayaku n = 45	Sig.
Human capital (L-HC)
Gender/head of household (female%)	59	57.7 a	68.2 b	51.1 a	*
Ethnicity (% Kichwa)	88.4	94.2 b	97.7 b	73.3 a	***
Household head age (years)	48.7	51.8 a	43.8 b	50.5 a	***
Social capital (L-SC)
Training received (BMP) (%)	52.3	43.6 a	51.2 a	62.2 b	*
Natural capital (L-NC)
Chakra (ha)	2.2	2.1 a	1.9 a	2.7 b	***
Forest (ha)	5.7	4.1 a	1.2 a	11.8 b	***
Other crops (ha)	0.4	0.0 a	0.3 a	1 b	**
Total farm area (ha)	8.4	6.2 a	3.5 a	15.6 b	***
Financial capital (L-FC)
Access to credit (%)	8	9.6 a	4.7 a	24.4 b	**
Access to government bonus (%)	56	60.9 b	62.0 b	44.4 a	**
Physical capital (L-PhC)
Engine technology (yes = 1)	1.54	64.1 c	24.8 a	48.9 b	***
Cell phone (yes = 1)	69	59.6 a	62.0 a	77.8 b	*

p-value: * p < 0.05; ** p < 0.01; *** p < 0.001. Source: Authors computation from survey data. ANOVA for continuous variables and Chi square for categorical variables; different letters indicate significant differences between social groups at 5%.

).

The analysis of social capital (L-SC) shows significant differences in access to training on Best Management Practices (BMP) among the associations. Tsatsayaku has the highest percentage of producers who have received BMP training (62.2%), significantly higher than Wiñak (51.2%) and Kallari (43.6%) (p < 0.05). In addition to training, social capital also encompasses the associative activities of producers, which play a critical role in fostering sustainable development [30,31]. The higher level of training and involvement reflects a stronger capacity for collective action and resource management, key components for enhancing sustainability through improved agroforestry practices.

The results of natural capital (L-NC) reveal significant differences in land-use and management among the associations. Tsatsayaku has the largest Chakra area (2.7 ha), significantly more than Kallari (2.1 ha) and Wiñak (1.9 ha) (p < 0.001). Similarly, forest area is substantially greater in Tsatsayaku (11.8 ha), compared to Kallari (4.1 ha) and Wiñak (1.2 ha) (p < 0.001), indicating that Tsatsayaku producers manage larger forested lands. In terms of other crops, Tsatsayaku also leads with 1 ha, significantly more than Kallari and Wiñak (both below 0.3 ha) (p < 0.01). Overall, total farm area is largest in Tsatsayaku (15.6 ha), while Wiñak and Kallari manage smaller farms (3.5 ha and 6.2 ha, respectively) (p < 0.001). In comparison, Kichwa farmers dedicated to agriculture in Pastaza have significantly larger landholdings, averaging 64 hectares per household [54]. There were no significant differences in distance to the city or road access to the community among the associations. These findings highlight the larger landholdings and more extensive use of natural resources in Tsatsayaku, which could have implications for their capacity for sustainable land management.

Physical capital (L-PhC) reveals significant differences in access to essential agroforestry tools and communication devices among the associations. Access to engine technologies (use of strimmers, among others) shows the highest disparity, with Wiñak having the largest percentage of producers owning this tool (64.1%), followed by Kallari (48.9%) and Tsatsayaku (24.8%) (p < 0.001). This suggests that Wiñak and Kallari have better access to agricultural equipment necessary for sustainable land management. Cell phone ownership also differs significantly, with Tsatsayaku having the highest percentage of producers owning cell phones (77.8%), compared to Kallari (62.0%) and Wiñak (59.6%) (p < 0.05). This reflects better communication capabilities in Tsatsayaku. No significant differences were found in terms of household goods or vehicle ownership, with access to cars or motorcycles being relatively low across all associations.

4.2. Economic Welfare, Income, and Poverty Index in Households of the Amazonian Chakra

Two indicators were used to assess household economic welfare: per capita income (PCI) and the poverty index (FGT) [44,55]. Our results show that the average annual PCI was highest among cacao-producing households in the Tsatsayaku association (USD 714.15), largely due to larger landholdings and engagement in additional activities such as livestock farming. The PCI for Wiñak and Kallari producers was relatively similar, at USD 324.07 and USD 361.46, respectively. These figures are comparable to the USD 327 PCI reported by Torres et al. [29] for Amazonian producers who had adopted agriculture-oriented livelihood strategies.

Overall, extreme poverty was reported in 82% of households across the three producer associations (

Table 2. Income of producers belonging to the Amazonian Chakra of three local associative organizations, RBS, Ecuadorian Amazon.

Variables	Mean n = 330	Kallari n = 156	Wiñak n = 129	Tsatsayaku n = 45	Sig.
Income
Total income (USD)	1652.73	1540.68 a	1287.45 a	3096.94 b	***
Chakra income (USD)	575.69	673.34 b	487.38 a	490.36 a	*
Other income (USD)	565.54	305.12 a	299.03 a	2232.35 b	***
Per capita income (USD/annual)	394.94	361.46 a	324.07 a	714.15 b	**
Headcount index (% extremely poor)	82	86.5	89.1	68.9	n.s.

p-value: * p < 0.05; ** p < 0.01; *** p < 0.001. n.s. = no significant among groups. Source: Authors computation from survey data. ANOVA for continuous variables and Chi square for categorical variables; different letters indicate significant differences between social groups at 5%.

). However, Tsatsayaku producers had the lowest extreme poverty rate (69%), while Wiñak (89%) and Kallari (87%) exhibited higher rates, exceeding those reported by [29] for the Sumaco Biosphere Reserve (SBR).

Significant differences in income generated from the Amazonian Chakra were observed across the three associations (Table 2). Tsatsayaku reported an average income of USD 490.36, Wiñak USD 487.37, and Kallari led with an average income of USD 673.34. A deeper analysis of total income shows that income from the Chakra represents only 16% of the average total income in Tsatsayaku (USD 3096.94). However, for Wiñak producers, Chakra-related income accounts for 37% of their total income of USD 1284.45. For Kallari producers, this income constitutes 44% of their average household income of USD 1652.73 (Table 2), These findings are consistent with a report by GIZ [56], which highlighted that cacao income from the Chakra system in the Kallari association accounted for an average of 42% of total monetary income. This underscores the pivotal role of the Chakra system in sustaining the livelihoods of indigenous households. Given that cacao is the primary crop cultivated within the Chakra, these results align with findings from other tropical regions, which emphasize the economic importance of cacao-based agroforestry systems in providing income diversification and enhancing rural household resilience [57,58].

4.3. Sustainability Dimensions Assessment

Among the four sustainability dimensions analyzed using 80 SAFA indicators, we found that 44 out of the 80 evaluated indicators (55%) showed significant differences across the three producer groups. These results highlight the diverse sustainability practices and outcomes among the associations, emphasizing areas where specific groups excel or need improvement (

Table 3. SAFA indicators evaluated in the three producer’s associations evaluated in the Sumaco Biosphere Reserve, EAR.

		SAFA –Themes—Indicators	All n = 330	Kallari n = 156	Wiñak n = 129	Tsatsayaku n = 45	Sig.
Godd governance (14)	G1.1.1	Mission explicitness	3.08	3.21 b	3.27 b	2.09 a	***
	G1.1.2	Mission-driven	3.07	3.20 b	3.25 b	2.13 a	***
	G1.2.1	Due diligence	3.03	3.10 b	3.25 b	2.18 a	***
	G2.1.1	Holistic audits	3.11	3.14 b	3.40 b	2.18 a	***
	G2.2.1	Responsibility	3.02	3.03 b	3.33 b	2.11 a	***
	G2.3.1	Transparency	3.05	3.08 b	3.32 b	2.13 a	***
	G3.1.1	Stakeholder identification	3.23	3.28 b	3.46 b	2.38 a	***
	G3.1.2	Stakeholder engagement	3.25	3.24 b	3.48 b	2.60 a	***
	G3.1.4	Effective participation	3.05	3.13 b	3.13 b	2.51 a	**
	G3.2.1	Grievance procedures	3.11	3.27 b	3.09 b	2.64 a	**
	G4.2.1	Remedy, restoration, and prevention	3.34	3.32 a,b	3.52 b	2.87 a	*
	G4.3.1	Responsibility	3.19	3.03 a,b	3.48 b	2.91 a	***
	G5.1.1	Sustainability management plan	3.38	3.50 b	3.53 b	2.51 a	***
	G5.2.1	Full cost accounting	3.18	3.33 b	3.22 b	2.56 a	***
Ennvironmental integrity (15)	E1.1.1	GHG reduction target	3.48	3.68 b	3.44 b	2.91 a	***
	E1.1.2	GHG mitigation practices	3.58	3.70 b	3.57 a,b	3.20 a	*
	E2.1.2	Water conservation practices	3.26	3.10 a	3.27 a,b	3.78 b	*
	E2.2.2	Water pollution prevention practices	4.14	4.31 b	4.04 a,b	3.87 a	*
	E3.2.1	Land conservation and rehabilitation plan	3.68	3.86 b	3.66 b	3.13 a	***
	E3.2.2	Land conservation and rehabilitation practices	3.86	3.92 b	4.00 b	3.29 a	***
	E4.1.4	Ecosystem connectivity	3.96	4.06 b	3.98 b	3.53 a	**
	E4.1.5	Land-use and land-cover change	3.78	3.96 b	3.76 b	3.27	***
	E4.2.2	Species conservation practices	3.83	4.01 b	3.79 b	3.31 a	***
	E4.2.3	Diversity and abundance of key species	3.94	4.09 b	3.96 b	3.33 a	**
	E4.2.4	Diversity of production	4.00	4.14 b	4.01 b	3.49 a	***
	E4.3.1	Wild genetic diversity enhancing practices	4.06	4.10 b	4.19 b	3.51 a	***
	E4.3.2	Agro-biodiversity in situ conservation	4.07	4.10 b	4.14 b	3.73 a	*
	E5.2.1	Renewable energy use target	2.43	2.62 b	2.40 b	1.87 a	***
	E5.2.2	Energy saving practices	2.39	2.63 b	2.32 b	1.73 a	***
Economic resilience (11)	C1.2.1	Community investment	3.47	3.62 b	3.39 a,b	3.20 a	*
	C1.3.1	Long-term profitability	3.23	3.40 b	3.16 a,b	2.82 a	**
	C1.3.2	Business plan	3.15	3.42 c	3.05 b	2.51 a	***
	C1.4.1	Net income	3.19	3.40 b	3.06 a,b	2.82 a	***
	C1.4.2	Cost of production	3.15	3.36 b	3.03 a,b	2.80 a	***
	C1.4.3	Price determination	3.26	3.44 b	3.12 a,b	3.09 a	**
	C2.1.2	Product diversification	3.59	3.67 b	3.67 b	3.09 a	**
	C2.3.1	Stability of market	3.09	3.22 b	3.10 b	2.64 a	**
	C2.4.1	Net cash flow	3.09	3.25 b	3.08 b	2.60 a	***
	C2.4.2	Safety nets	2.96	3.17 b	2.93 b	2.31 a	***
	C2.5.1	Risk management	3.31	3.47 b	3.22 a,b	2.96 a	*
Social well-being (4)	S2.1.1	Fair pricing and transparent contracts	3.30	3.45 b	3.29 b	2.82 a	***
	S5.1.1	Safety and health training	3.47	3.74 b	3.36 a,b	2.84 a	***
	S5.2.1	Public health	4.19	4.12 a	4.24 b	4.29 b	***
	S6.1.1	Indigenous knowledge	4.26	4.44 b	4.36 b	4.21 a	***

p-value: * p < 0.05; ** p < 0.01; *** p < 0.001; different letters indicate significant differences between social groups at 5%.

).

4.3.1. Good Governance

In the governance dimension, Wiñak (3.36) and Kallari (3.26) consistently outperformed Tsatsayaku (2.59), reflecting stronger governance structures in the former two associations. Tsatsayaku exhibited notably lower scores in key indicators such as mission explicitness (2.09) and responsibility (2.11), while Wiñak and Kallari demonstrated significantly higher values for these indicators (ranging between 3.20 and 3.33). This disparity may stem from Tsatsayaku being a younger association compared to the more established Wiñak and Kallari [59].

Similarly, indicators such as holistic audits and transparency highlighted these differences, with Tsatsayaku scoring approximately 2.10, compared to Kallari (~3.10) and Wiñak (~3.40). Stakeholder engagement and sustainability management plans followed similar trends, with Wiñak and Kallari scoring significantly higher (~3.48 and ~3.50, respectively) compared to Tsatsayaku (~2.60 and 2.51). A plausible explanation for this could be the organizational maturity and resources of more developed associations like Kallari, which have been able to secure funding and technical assistance from international cooperation, particularly for the specialization in cacao production within the Amazonian Chakra system [60,61].

4.3.2. Environmental Integrity

In the environmental dimension, Kallari (3.70) and Wiñak (3.64) consistently outperformed Tsatsayaku (3.27) across most indicators. Kallari led in key areas such as GHG reduction targets (3.68) and species conservation practices (4.01). A plausible explanation for this is that both Kallari and Wiñak are predominantly composed of indigenous populations, who traditionally manage Chakra systems with high tree diversity [62,63] and significant carbon storage potential [13,59]. Producers across all three associations demonstrated low perceptions of environmental impacts, which aligns with the inherent sustainability of the Amazonian Chakra system. This finding supports previous research by Caicedo [64], highlighting the ecological benefits of the Chakra system in minimizing environmental degradation while promoting biodiversity and carbon sequestration [65]. These results underscore the critical role of indigenous agroforestry practices in fostering environmental integrity and resilience.

For land conservation and rehabilitation practices, Kallari (3.92) and Wiñak (4.00) also scored higher than Tsatsayaku (3.29). Similarly, significant differences were observed in agro-biodiversity conservation and wild genetic diversity practices. Wiñak and Kallari achieved scores of around 4.10, whereas Tsatsayaku lagged at approximately 3.50. This disparity may also be attributed to the indigenous practices in Kallari and Wiñak, where other studies have reported that these populations manage highly agro-diverse Chakra systems, incorporating over 48 medicinal and spiritual plant species [66] and 110 species of trees, shrubs, and palms [67] in the Alto Napo region.

4.3.3. Economic Resilience

Although significant differences in income generated from the Amazonian Chakra were observed among the three associations (Table 2), Tsatsayaku reported a notably higher average total household income compared to Wiñak and Kallari. This disparity can be attributed to Tsatsayaku’s demographic composition, where 30% of the population comprises migrant settler farmers. These settlers often have greater access to off-farm employment opportunities and manage larger landholdings, which frequently include livestock systems that significantly enhance their overall income [68]. Moreover, this allows them to accumulate natural capital as a “savings account”, providing a buffer against income fluctuations [20,69,70]. However, producers’ perceptions of economic resilience—a critical dimension of sustainability [22]—presented a contrasting scenario, with Kallari (3.60) outperforming both Wiñak (3.46) and Tsatsayaku (3.18).

Key indicators such as net income (3.40), cost of production (3.36), and long-term profitability (3.40) were significantly higher for Kallari compared to Tsatsayaku, which scored lower in these areas (2.82–3.02). These differences may stem from the predominant indigenous culture in Kallari and Wiñak, where Chakra income constitutes a more substantial proportion of total income [59,68], and Kallari’s more advanced organizational strategies.

Wiñak showed intermediate performance across most indicators, closely aligning with Kallari in areas such as product diversification (3.67) and market stability (3.22). However, significant disparities emerged in indicators like business plan (3.42 vs. 2.51) and safety nets (3.17 vs. 2.31), where Kallari demonstrated stronger economic strategies. This may be attributed to Kallari’s implementation of Fair Trade strategies, tailored to its products and the specific needs of its producers [71]. These findings emphasize the role of economic strategies and demographic factors in shaping resilience and sustainability outcomes among producer associations.

4.3.4. Social Welfare

The social well-being dimension showed relatively similar overall scores among the associations, with Kallari (3.86), Wiñak (3.85), and Tsatsayaku (3.76) achieving comparable results. These findings highlight the importance of strengthening associative enterprises as a mechanism for promoting social sustainability. It is advisable to share these results with producers, local stakeholders, and decision-makers through initiatives such as field schools, which could help enhance medium- and long-term sustainability scores across other associations. Furthermore, these efforts should emphasize the promotion of productive associativity as a cornerstone for sustainable development in the Ecuadorian Amazon [37,59].

Despite overall similarities, Kallari excelled in key areas such as safety and health training (3.74), where Tsatsayaku lagged significantly behind (2.84). Public health scores also revealed disparities, with Wiñak (4.24) and Kallari (4.12) outperforming Tsatsayaku (3.87). While indicators such as gender equality, non-discrimination, and food sovereignty showed minimal variation among the associations, Tsatsayaku demonstrated slightly lower scores in areas related to support for vulnerable people and access to medical care.

Overall, Kallari and Wiñak exhibited stronger outcomes in health and safety related indicators, reflecting their advanced organizational structures and access to resources. Meanwhile, Tsatsayaku displayed minor gaps in social well-being, suggesting the need for targeted interventions to strengthen these aspects and reduce disparities among associations. These results underscore the critical role of producer associations in enhancing social well-being and fostering sustainable livelihoods.

4.4. Multivariate Discriminant Analysis by Association (SAFA-SLF)

The multivariate discriminant analysis of 58 indicators from the SAFA and SLF revealed that several variables significantly differentiated the associations (

Table 4. Discriminant functions of livelihood and sustainability indicators of three producer associations involving the Amazonian Chakra, SBR, Ecuadorian Amazon.

No.	Parameters 1	Wilks’—Lambda	Partial—Lambda	F-Remove	p-Level 2	Toler	1-Toler
1	L-PhC.1	0.39	0.88	20.30	***	0.90	0.10
2	L-NC.1	0.35	0.98	2.50	*	0.04	0.96
3	C1.3.2	0.35	0.97	4.23	**	0.52	0.48
4	E2.1.2	0.36	0.94	8.98	***	0.72	0.28
5	L-HC.1	0.36	0.96	5.92	***	0.90	0.10
6	E2.2.2	0.36	0.95	7.25	***	0.54	0.46
7	L-HC.2	0.35	0.97	4.12	***	0.79	0.21
8	C2.4.2	0.35	0.97	4.28	**	0.32	0.68
9	C1.4.3	0.35	0.97	4.91	**	0.54	0.46
10	S5.2.1	0.35	0.97	5.06	**	0.62	0.38
11	S5.1.1	0.35	0.98	3.16	*	0.72	0.28
12	E4.3.1	0.35	0.97	4.91	**	0.19	0.81
13	L-FC.1	0.35	0.98	3.42	**	0.74	0.26
14	E4.3.2	0.35	0.98	3.38	*	0.22	0.78
15	L-FC.2	0.35	0.97	4.06	**	0.79	0.21
16	L-FC.3	0.35	0.98	3.28	*	0.83	0.17
17	G4.3.1	0.36	0.96	6.95	***	0.40	0.60
18	G5.2.1	0.35	0.98	3.01	*	0.73	0.27
19	G2.1.1	0.35	0.98	3.68	**	0.34	0.66

¹ L-PhC.1 = engine technology; L-NC.1 = farm size; C1.3.2 = business plan; E2.1.2 = water conservation practices; L-HC.1 = age household head; E2.2.2 = water pollution prevention practices; L-HC.2 = ethnicity; C2.4.2 = safety nets; C1.4.3 = price determination; S5.2.1 = public health; S5.1.1 = safety and health training; E4.3.1 = wild genetic diversity enhancing practices; L-FC.1 = total income; E4.3.2 = agro-biodiversity in situ conservation; L-FC.2 = Chakra income; L-FC.3 = access state bonus; G4.3.1 = civic responsibility; G5.2.1 = full-cost accounting; G2.1.1 = holistic audits. ² p-value: * p < 0.05; ** p < 0.01; *** p < 0.001.

). Wilks’ Lambda was used as a key measure, with values closer to 0 indicating stronger discriminating power.

Notably, the technicalization of tasks (use of strimmers, among others) (L-PhC.1) (Wilks’ Lambda = 0.39, p < 0.001) exhibited the highest discriminatory ability, suggesting that the availability and use of technology marked significant differences among associations, marking it as a critical factor in financial management and productivity. Of the 58 proposed indicators, the discriminant model accepted 33 and excluded 25 variables. Subsequently, according to Wilks’ Lambda, p-value, Snedecor’s F, and tolerance level, a final model was proposed with 19 indicators: 7 of (SLF) and 12 of the SAFA methodologies. Other key indicators, such as total income (L-FC.1) (Wilks’ Lambda = 0.98, p < 0.01) and Chakra income (L-FC.2) (Wilks’ Lambda = 0.97, p < 0.01), as well as business plan (C1.3.2) (Wilks’ Lambda = 0.97, p < 0.01) and price determination (C1.4.3) (Wilks’ Lambda = 0.97, p < 0.01), showed significant variation, pointing to disparities in income generation, business plan, and market dynamics across associations. Moreover, L-FC.1 includes both on-farm and off-farm income, such as employment and non-agricultural work, revealing disparities in access to alternative income sources between the associations. This highlights how some groups rely more heavily on diverse income streams beyond agriculture [15,72], which can impact their financial stability and resilience [1,25,35].

Furthermore, human capital and environmental management practices were also significant discriminators. For example, age of the household head (L-HC.1) (Wilks’ Lambda = 0.96, p < 0.001) and water pollution prevention practices (E2.2.2) (Wilks’ Lambda = 0.96, p < 0.001) highlighted the importance of demographic factors and environmental stewardship in shaping sustainability outcomes. Additionally, civic responsibility (G4.3.1) (Wilks’ Lambda = 0.95, p < 0.001) and safety nets (C2.4.2) (Wilks’ Lambda = 0.97, p < 0.01) were strong governance and economic resilience indicators, respectively, showing significant differences between the associations. These findings underscored the great value of considering livelihoods in a sustainability assessment and the need for targeted efforts to improve governance, environmental practices, and economic safety nets, particularly in associations that performed weaker in these areas.

Some indicators, such as farm size (L-NC.1) also emerged as a significant factor (Wilks’ Lambda = 0.98, p < 0.05), indicating that the size of landholdings is a key determinant of livelihood outcomes [68]. Larger farms enable greater diversification of production, which can promote economic stability and sustainability, while land scarcity may constrain the expansion of commercial Chakra systems [73]. In the area of governance, indicators such as civic responsibility (G4.3.1) (Wilks’ Lambda = 0.96, p < 0.001), full-cost accounting (G5.2.1) (Wilks’ Lambda = 0.98, p < 0.05), and holistic audits (G2.1.1) highlighted the importance of institutional strength and accountability. Associations with stronger governance frameworks were better able to manage resources sustainably and improve livelihood outcomes. This suggests that while certain governance and natural capital factors may be consistent, this needs to be further strengthened alongside other areas such as income diversification, financial access and market stability, need strategic intervention. By strengthening some elements of natural resource governance, positive impacts on other aspects of sustainability can be achieved [74].

Focusing on indicators with significant discriminatory power, such as wild genetic diversity enhancing practices (E4.3.1) (Wilks’ Lambda = 0.97, p < 0.01), water conservation practices (E2.1.2) (Wilks’ Lambda = 0.94, p < 0.001), and public health (S5.2.1) (Wilks’ Lambda = 0.97, p < 0.01), can guide more personalized sustainability management practices [75]. These results demonstrate that associations focusing on water conservation and biodiversity enhancement are better equipped to maintain ecological balance, which is critical in the Amazonian Chakra system. Furthermore, agro-biodiversity in situ conservation (E4.3.2) (Wilks’ Lambda = 0.98, p < 0.01) was also significant, highlighting the importance of environmentally sustainable practices [63,76]. This approach will allow for better allocation of resources, improving sustainability outcomes in a context-specific manner across the different associations.

In Figure 4a, the canonical discriminant function plot shows clear separation among the producers of the three associations. Kallari is distinctly separated along Root 1, suggesting that indicators related to financial management and environmental practices are the most significant in differentiating this group from Wiñak and Tsatsayaku, which cluster more closely. Tsatsayaku shows greater internal variability along Root 2, indicating heterogeneity within the group, while Wiñak and Kallari present more cohesive clusters.

The cluster analysis in Figure 4b confirms these findings, with Wiñak and Kallari forming a closer cluster, reflecting similar sustainability practices, while Tsatsayaku is more distant, highlighting its distinct characteristics. These results suggest that Kallari and Wiñak share stronger governance and environmental resilience, while Tsatsayaku may benefit from targeted interventions in these areas to enhance its sustainability performance.

The sustainability assessment was carried out using 58 indicators: 44 SAFA, 14 livelihoods and income indicators. Finally, the proposed classification model used 19 indicators: 12 SAFA items and 7 livelihood and income items. The livelihood variables showed high discriminating power between associations. Two variables of great importance were the level of technological development and the size. The size is limited by the surface area of the protected space, which makes increasing it very complex. On the other hand, the technological level is associated with other variables included in the model, such as the age of the owner, ethnicity, access to credit, income, etc. According to Bastanchury et al. and De-Pablos-Heredero et al. [77,78] small farms can be viable through sustainable intensification by locating their production in the area of increasing returns.

Improving the level of technical development requires addressing different aspects. Indigenous people are more vulnerable and showed worse indicators. It is necessary to have the technology and enhance dynamic capacities to develop efficient use [78,79].

At the level of public policies, it would be interesting to facilitate access to credit for indigenous populations [80], promoting the succession of Chakras, which is a case of promoting family businesses [81] and women’s access to ownership of exploration [82]. These policies could promote technological improvement.

Given the importance of governance indicators, policies should also aim to strengthen the participatory and organizational capacity of productive associations [61], especially around internal financial accountability and resource management such as access to soft loans or access to special markets [37,83,84]. In addition, the promotion of sustainable environmental practices, such as water conservation, biodiversity protection [7,85], and carbon sequestration [13,86], will be essential to ensure long-term sustainability. In addition, all these related social, environmental, and good governance aspects can also be strengthened to attract new sustainable income through community and scientific [87] tourism around the Amazon Chakra system, which today is a GIAHS site [43].

Amongst the main limitations of the study, the following ones can be highlighted. This multivariate study is exploratory and should be complemented with other studies that quantify the causal relationship between sustainability and livelihood dimensions with economic results of the Amazonian Chakra. The sample is sufficient and broad but should be increased with traditional systems in other countries, which would allow the results to be extended to other contexts.

5. Conclusions

This study highlights the utility of the Sustainable Livelihoods Framework (SLF) and the Sustainability Assessment of Food and Agriculture Systems (SAFA) methodology as complementary tools for evaluating sustainability within the Amazonian Chakra system, a Globally Important Agricultural Heritage System (GIAHS) recognized by the FAO. By integrating these frameworks, the study provided a comprehensive assessment of livelihood capitals and sustainability dimensions, such as good governance, environmental integrity, economic resilience, and social well-being. The discriminant analysis was pivotal in identifying clear distinctions among the three producer associations, underscoring the unique strengths and challenges faced by each group and emphasizing the critical role of the Amazonian Chakra system in fostering sustainable agroforestry practices.

The discriminant analysis revealed significant differences in key indicators such as Chakra income, price determination, wild genetic diversity, and civic responsibility. Kallari and Wiñak consistently outperformed Tsatsayaku in governance, environmental integrity, and economic resilience, reflecting their more mature organizational structures and stronger governance mechanisms. Tsatsayaku, while exhibiting strengths in demographic diversity and larger landholdings, showed lower performance in critical sustainability indicators. Extreme poverty was prevalent across all associations, affecting 82% of households, with Tsatsayaku having the lowest poverty rate (69%). These findings emphasize the Chakra system’s economic and cultural significance for indigenous livelihoods, particularly for associations like Kallari and Wiñak, where Chakra-related income constitutes 44% and 37% of total income, respectively. In contrast, Tsatsayaku relied less on the Chakra system, with only 16% of total income derived from it, reflecting its engagement in other livelihood activities such as livestock farming and off-farm employment.

This study also highlights the disparities in resource access and governance among the associations, pointing to the need for tailored interventions. Policies should focus on financial support, credit access, and market transparency to address economic vulnerabilities. Participatory governance training and incentives for biodiversity conservation and sustainable land management practices are particularly crucial for associations with less-developed organizational structures like Tsatsayaku.

The integration of SLF and SAFA methodologies, complemented by discriminant analysis, offers a robust framework for assessing sustainability in agroforestry systems. This approach provides actionable insights for policymakers, emphasizing the importance of addressing disparities in income generation, resource access, and governance while leveraging the unique strengths of each association. By focusing on governance, financial strategies, and environmental practices, stakeholders can enhance the sustainability and resilience of Amazonian producer associations, ensuring the long-term viability of this culturally and ecologically significant system.