Behavioral Drivers of Cage Tilapia (Oreochromis niloticus) Producers and Consumers in Kenya’s Lake Victoria Region

Abstract

The cage tilapia farming boom in Kenya’s Lake Victoria region underscores its role in food security and economic growth. Success depends on understanding producer and consumer behaviors within the value chain. Using the Theory of Planned Behavior (TPB), this study examines how attitudes (evaluations of farming/consumption), subjective norms (social pressures), perceived behavioral control (confidence in actions), environmental awareness, and moral obligation shape decisions. A survey of 66 producers and 169 consumers, analyzed via structural equation modeling (SEM), reveals key drivers. Producers are driven by positive attitudes toward profitability, technical feasibility, and sustainability, reinforced by community norms and resource access, promoting sustainable practices. Consumers prioritize health, affordability, and accessibility of cage-farmed tilapia, with environmental and ethical factors less influential. These findings highlight opportunities for targeted interventions to enhance production, boost demand, and ensure sustainable aquaculture.

1. Introduction

Global aquaculture has undergone rapid expansion over the past decade, positioning cage tilapia farming as a vital contributor to global food security and economic progress [1]. Cage tilapia farming, involving submerged nets in natural water bodies, supports livelihoods and nutrition in developing nations, with government policies, market access, and regional regulations shaping producer adoption and consumer acceptance through investment and distribution networks [2,3]. Current data indicate that aquaculture accounts for over half of the world’s fish consumption, with cage farming systems recognized for their efficiency and sustainability in producing fish like Nile tilapia (Oreochromis niloticus) [4].

Africa’s aquaculture sector has seen remarkable growth, with cage tilapia farming, particularly Oreochromis niloticus, becoming a renowned tilapia species in major water bodies across the continent [4]. The Lake Victoria basin, a critical hub shared by East African nations, has emerged as a leader in this development, with cage farming expanding at an estimated annual rate of 12% since 2018 [4]. This growth reflects rising fish demand and economic opportunities.

In Kenya, the riparian counties surrounding Lake Victoria have witnessed a significant boom in cage tilapia farming, fueled by government support and private sector investments [3]. Despite this progress, misconceptions, such as the “China fish” label, health implications, and low quality, among others, challenge consumer trust in the value chain [5]. The sector’s long-term success depends on understanding the behavioral factors influencing production and consumption, a dimension underexplored in Sub-Saharan aquaculture research.

The decision-making processes of cage tilapia producers are shaped by a complex interplay of attitudes, social pressures, and perceived control over resources [6,7]. Attitudes reflect beliefs about profitability and sustainability, subjective norms involve peer influence, and perceived behavioral control relates to resource access [6]. These constructs, while distinct, may overlap in practice, necessitating careful measurement to avoid redundancy [8]. Studies highlight that producers’ choices are influenced by their knowledge, risk perceptions, and access to technical and financial support, yet these factors are insufficiently studied in Kenya’s cage farming context.

Consumer behavior in the cage tilapia market is equally complex, driven by psychological, social, and economic considerations [5,9]. Research indicates that purchase decisions hinge on perceptions of quality, affordability, and health benefits, alongside cultural influences that shape preferences for fish products [5,9,10]. Increasing environmental and ethical awareness could shift consumer preferences toward sustainable products in the future, though current data suggest limited influence [11].

This research fills a critical gap by examining the interconnected behavioral dynamics of producers and consumers within the cage tilapia value chain, using the Theory of Planned Behavior (TPB) as a framework [12,13]. Unlike prior studies that often focus on production or consumption in isolation, this study integrates both perspectives to explore their mutual influence, offering a novel approach to understanding the sector’s behavioral drivers in the riparian counties of Lake Victoria, Kenya. The study is guided by the following hypotheses:

Hypothesis 1 (H1). 

Attitudes toward cage tilapia farming/consumption have a significant positive influence on behavioral intentions.

Hypothesis 2 (H2). 

Subjective norms significantly affect the behavioral intentions of both producers and consumers in the cage tilapia value chain.

Hypothesis 3 (H3). 

Perceived behavioral control has a significant positive impact on behavioral intentions towards cage tilapia production and consumption.

Hypothesis 4 (H4). 

Motivation toward cage tilapia farming/products positively influences behavioral intentions.

Hypothesis 5 (H5). 

Moral obligation significantly affects the behavioral intentions of both producers and consumers in the cage tilapia value chain.

Hypothesis 6 (H6). 

Environmental awareness and concerns toward cage tilapia farming/consumption positively influence behavioral intentions.

Hypothesis 7 (H7). 

The aggregate behavioral intentions of both the producers and consumers of cage tilapia have a significant influence on their actual behavior.

2. Literature Review

2.1. Producer Behavioral Aspects

Producer attitudes toward cage tilapia farming emerge as a critical determinant of adoption and continued engagement in aquaculture activities. Research by [6] demonstrates that producers’ positive attitudes toward new farming technology significantly influence their intention to adopt or expand their operations. These attitudes are shaped by producers’ beliefs about the potential outcomes of cage farming, including expected profitability, perceived risks, and anticipated challenges. Particularly noteworthy is how producers’ previous experiences in aquaculture shape their attitudes, with successful early experiences leading to more positive behavioral intentions toward technology adoption and farm expansion.

The influence of subjective norms on producer behavior manifests through peer networks, community leaders, and local cooperatives [14,15]. Studies by [14] reveal that producers’ decisions are heavily influenced by peer opinions, community leaders’ endorsements, and local success stories. These subjective norms create social pressure that either encourages or discourages certain farming practices, with producers often aligning their behavior with community expectations. The strength of these normative influences varies across different cultural contexts, with some communities showing stronger peer effects than others in aquaculture adoption decisions [16]

Perceived behavioral control among producers significantly impacts their actual farming practices and technology adoption. Recent research by [17] highlights how producers’ confidence in their ability to manage farming operations influences their investment decisions and willingness to adopt new technologies. This perception of control is closely tied to access to resources, technical knowledge, and support systems. Producers with higher perceived behavioral control are more likely to experiment with innovative practices and expand their operations, while those with lower perceived control tend to be more conservative in their farming approaches.

The role of perceived usefulness in shaping producer behavior has gained increasing attention in recent studies. Ref. [18] demonstrates that producers’ perceptions of the utility and benefits of any farming technologies significantly influence their adoption intentions. These perceptions are formed through direct observation of benefits, information from extension services, and market feedback. Producers who perceive cage farming as highly useful for achieving their goals may show stronger behavioral intentions toward the adoption and improvement of their farming practices. The interplay in value chain networks among producers, traders, and consumers also shapes adoption by facilitating resource sharing and market access [15].

2.2. Consumer Behavioral Aspects

Consumer attitudes toward cage-farmed tilapia products play a fundamental role in shaping market demand and consumption patterns. Ref. [19] reveals that consumer attitudes are primarily formed through perceptions of product quality, safety, and value for money. These attitudes significantly influence purchase intentions and consumption frequency. The study particularly emphasizes how consumers’ beliefs about the health aspects shape their overall attitudes toward these products.

Subjective norms significantly influence consumer behavior in tilapia markets through social and cultural mechanisms. Ref. [20] explores how family traditions, peer recommendations, and cultural beliefs shape consumer purchasing decisions. The research demonstrates that social influence plays a crucial role in forming consumer preferences for specific types of fish products. The impact of these normative influences is particularly strong in communities where fish consumption has strong cultural significance.

Consumer perceived behavioral control manifests through their ability to access, identify, and purchase preferred tilapia products. Studies by [5] indicate that consumers’ confidence in their ability to select quality products, along with their perceived access to reliable suppliers, significantly influences their purchasing behavior. This aspect of behavioral control is closely linked to factors such as market infrastructure, product availability, and consumer knowledge about product quality indicators.

The role of socioeconomic factors in moderating consumer behavior has been extensively documented in recent research. Ref. [21] demonstrates how economic status, lifestyle choices, and family life cycle stages influence the relationship between behavioral intentions and actual purchasing behavior. The findings suggest that while consumers may have positive intentions toward cage-farmed tilapia consumption, their actual behavior is significantly moderated by their socioeconomic circumstances.

2.3. Theoretical Framework: Theory of Planned Behavior

The Theory of Planned Behavior (TPB), proposed by [8], originated from the Theory of Reasoned Action (TRA). TRA posited that individual behavior is primarily determined by behavioral intentions, which are influenced by attitudes and subjective norms. However, [8] extended the framework by introducing the concept of perceived behavioral control, recognizing that people do not always have full volitional control over their actions. TPB posits that attitudes (evaluations of a behavior’s desirability), subjective norms (perceived social pressures), and perceived behavioral control (confidence in performing the behavior) shape intentions, which predict actual behavior [8]. TPB has since become a widely applied psychological theory, offering insights into various behaviors where control, attitudes, and social influences play critical roles [8]. However, TPB has limitations, including its focus on rational decision-making, which may overlook emotional or habitual factors [22] and its variable predictive power in complex contexts, like agriculture.

Over the years, TPB has proven useful in predicting behavior across diverse domains, including agriculture. In aquaculture, TPB explains farmers’ adoption of technologies and consumers’ purchasing decisions, though its effectiveness depends on context-specific factors [23,24]. Central to this theory are three constructs, namely, attitudes, subjective norms, and perceived behavioral control, which shape an individual’s behavioral intentions. Attitudes reflect personal evaluations of behavior, subjective norms involve perceived social pressure, and perceived behavioral control refers to an individual’s perceived ease or difficulty in performing the behavior [25]. When these factors combine, they form a foundation for predicting actual behavior through intentions.

In the context of agriculture, including aquaculture, TPB provides a robust framework for understanding farmers’ decisions regarding technology adoption, production methods, and risk management. Farmers’ attitudes toward innovation, social pressures from their communities, and their perceived control over farming resources collectively influence their behavior. The application of TPB in cage tilapia farming is particularly relevant because of the complexity of decision-making in aquaculture. Producers are influenced not only by economic factors but also by social–economic and environmental considerations, making TPB a valuable tool for understanding their intentions and actual behaviors. Moreover, consumer behavior in aquaculture markets, including attitudes toward fish quality, health benefits, and social norms around fish consumption, can also be effectively explained through TPB.

2.4. Conceptual Framework

The conceptual framework in Figure 1 incorporates the Theory of Planned Behavior (TPB) to illustrate the behavioral relationships influencing cage tilapia producers and consumers in Kenya’s Lake Victoria region. According to TPB, producer and consumer behaviors are shaped by attitudes, subjective norms, and perceived behavioral control, collectively forming behavioral intentions that drive actual behavior. Environmental awareness and moral obligation are included as additional constructs, adapted from studies on sustainable agriculture, to capture ethical and ecological influences [26,27]. For producers, attitudes encompass interest in and perceptions of cage farming profitability, technical challenges, and sustainability. On the other hand, consumers’ attitudes are shaped by perceptions of the quality, price, and nutritional value of cage-farmed tilapia, which are factors that influence purchasing decisions and consumption patterns.

Subjective norms represent the social influences affecting producers’ and consumers’ behavioral intentions. Among producers, these norms include the expectations of community leaders, peer farmers, and market stakeholders, who influence decisions regarding cage tilapia farming practices. For consumers, social and cultural expectations, such as family traditions, peer recommendations, and social acceptance of farmed fish, contribute significantly to their decisions regarding the consumption of cage tilapia. Both groups experience pressure to conform to perceived norms, impacting their willingness to produce or consume cage-farmed fish.

Perceived behavioral control also plays a crucial role, capturing the individuals’ sense of control over their actions. For producers, this control is linked to resource availability, technical knowledge, and the capacity to manage farming operations efficiently. High perceived control among producers can lead to more proactive behaviors, such as expanding operations or adopting sustainable practices. For consumers, control factors include the accessibility of cage-farmed tilapia, knowledge about quality indicators, and purchasing power. Greater perceived control could result in higher confidence and a stronger inclination to purchase and consume cage-farmed tilapia.

This study hypothesizes that other emerging issues also influence the framework, including moral obligations and environmental awareness, which indirectly affect behavioral intentions and actual behavior. Environmental concerns, such as potential pollution from cage farming and the sustainability of inputs, are hypothesized to shape producers’ and consumers’ ethical considerations. Producers may feel a moral obligation to adopt environmentally friendly practices, while consumers may feel a duty to support sustainable aquaculture by choosing cage-farmed tilapia. This framework’s originality lies in integrating producer and consumer behaviors within a single TPB model, which is a novel approach in aquaculture.

The framework, based on TPB, illustrates how attitudes, subjective norms, perceived behavioral control, environmental awareness, and moral obligation shape the behavioral intentions and actual behaviors of producers and consumers. The arrows represent hypothesized relationships (H1–H7), with producer-side factors (e.g., profitability, resource access) and consumer-side factors (e.g., health, affordability) distinguished by grouped constructs.

3. Methodology

3.1. Research Design

This study employed a cross-sectional survey design with a quantitative approach. This research utilized structured questionnaires to collect data on behavioral factors affecting both producers and consumers in the cage tilapia value chain within Kenya’s Lake Victoria region. This design enabled systematic measurement and analysis of the relationships between attitudes, subjective norms, perceived behavioral control, and actual behavior among both producers and consumers.

3.2. Study Area

This study was conducted in the riparian counties of Lake Victoria, Kenya, including Kisumu, Homa Bay, Siaya, Migori, and Busia. These counties were selected because of their significant involvement in cage tilapia value chain activities and their strategic location along Lake Victoria, which provides an ideal environment for aquaculture.

3.3. Sampling Procedure

The sampling procedure for this study used a purposive sampling technique to ensure comprehensive representation across two key segments of the cage tilapia value chain in Lake Victoria, Kenya: producers and consumers. Participants were selected based on their relevance and experience, allowing for a detailed understanding of their behavioral aspects in the cage fish farming value chain. A total of 66 producers and 169 consumers were selected, leading to a sample size of 235 respondents. This sample size was deemed sufficient for SEM based on guidelines suggesting 5–10 respondents per parameter estimated (approximately 30 parameters in the model) [28]. Contact was made via phone calls and in-person visits, with the final selection based on willingness to participate and alignment with this study’s criteria. This approach ensured a broad and diverse set of perspectives across different geographic locations and operational sizes.

3.4. Sampling Technique

This study employed a multi-stage sampling approach. Initially, counties were stratified based on cage farming intensity to ensure representation across different production levels. Within each selected county, producers were chosen through simple random sampling using the county fisheries department registry. The registry included 150 active producers, from which 66 were randomly selected using a random number generator; 10 declined to participate, yielding a 13% rejection rate. Consumer households in selected urban areas were identified through systematic random sampling based on Yamane formulae to yield a sample size of 169.

3.5. Data Collection Methods

Primary data collection was conducted through structured questionnaires administered to both the producers and consumers. The questionnaires were designed based on TPB constructs, with items adapted from prior aquaculture studies [6,29]. The producer questionnaire included 45 items, and the consumer questionnaire included 44 items, covering attitudes, subjective norms, perceived behavioral control, environmental awareness, moral obligation, and motivation (see Appendix A.1 for item details and Appendix A.2 for full questionnaires). The items were measured on a 7-point Likert scale (1 = strongly disagree, 7 = strongly agree) to capture the intensity of the respondents’ beliefs and intentions.

3.6. Data Analysis Framework

3.6.1. Preliminary Analysis

The preliminary analysis phase involved comprehensive data cleaning and screening procedures using IBM SPSS Statistics V21*86 This included a thorough examination of missing values through the Missing Value Analysis (MVA) module in SPSS and the identification and treatment of outliers using both univariate and multivariate approaches. The data were tested for normality using Shapiro–Wilk tests, and outliers were managed by winsorizing extreme values to maintain statistical power [30]. The data were finally exported to IBM AMOS software for construct analysis.

3.6.2. Inferential Statistics

The inferential analysis began with a reliability assessment using Cronbach’s alpha coefficient, with a threshold of 0.7 considered acceptable for internal consistency. Composite reliability was also calculated to ensure construct reliability. The validity analysis included a construct validity assessment through factor analysis, a convergent validity evaluation using Average Variance Extracted (AVE ≥ 0.5), and discriminant validity confirmation through comparison of square root AVE with inter-construct correlations.

3.6.3. Model Specification

The structural equation model was specified as BI = β₁ATT + β₂SN + β₃PBC + β₄EAC + β₄MO + β₄M + ε₁ and AB = β₅BI + ε₂, where BI represents behavioral intention, ATT represents attitude, SN represents subjective norm, PBC represents perceived behavioral control, EAC represents environmental awareness and concern, M represents motivation, MO represents moral obligation, and AB represents actual behavior. Path coefficients are denoted by β and error terms by ε. This specification allowed for comprehensive testing of the hypothesized relationships in our conceptual framework.

3.6.4. Model Fit Assessment

Model fit was evaluated using multiple indices to ensure a robust assessment. The chi-square/df ratio should be less than 3.0 for an acceptable fit. Comparative Fit Index (CFI) and Tucker–Lewis Index (TLI) values greater than 0.95 indicate good fit. Root Mean Square Error of Approximation (RMSEA) should be less than 0.08, and Standardized Root Mean Square Residual (SRMR) should be less than 0.08 for an acceptable fit [28]. These multiple indices provided a comprehensive assessment of model adequacy.

3.7. Measurement Scales

The measurement framework shown in

Table 1. Summary of variable measurement.

Variable Category	Variables	Measurement Approach	Scale	Number of Questions
Independent Variables	Attitude	Likert scale measuring interest and perceptions	5-point Likert scale	7
	Subjective Norm	Likert scale measuring peer influence and norms	5-point Likert scale	4
	Perceived Behavioral Control	Likert scale measuring confidence and access	5-point Likert scale	3
	Environmental Awareness and Concern	Likert scale measuring awareness of impact	5-point Likert scale	2
	Moral Obligation	Likert scale measuring sense of responsibility	5-point Likert scale	2
	Motivation	Likert scale measuring desire for improvement	5-point Likert scale	2
Dependent Variables	Behavioral Intention	Likert scale measuring intention to adopt practices	5-point Likert scale	2
	Actual Behavior	Likert scale measuring frequency of engagement	5-point Likert scale	2

below, comprised of 45 items for producer variables and 44 items for consumer variables, primarily measured on 5-point Likert scales ranging from strongly disagree to strongly agree for producers and very low to very high for consumers, as provided in Appendix A.

4. Results for Producers

4.1. Model Fit Indices for Producers

All the variables included in Appendix A were first subjected to an exploratory factor analysis (see

Table 2. Factor loading.

Latent Variable	Indicator Label	Indicator Description	Factor Loading
Attitude	ATT1	Interest in Cage Tilapia Farming	0.76
	ATT2	Perceived Profitability	0.72
	ATT3	Technical Skills	0.69
	ATT4	Sustainability of Cage Farming	0.65
Subjective Norm	SN1	Peer Influence	0.78
	SN2	Community Expectations	0.74
	SN3	Recommendations from Other Farmers	0.7
Perceived Behavioral Control	PBC1	Confidence in Managing Operations	0.8
	PBC2	Access to Resources	0.77
	PBC3	Knowledge of Cage Farming Techniques	0.72
Environmental Awareness and Concern	EAC1	Awareness of Environmental Impact	0.74
	EAC2	Concerns about Pollution	0.71
Moral Obligation	MO1	Sense of Responsibility toward Sustainable Practices	0.79
	MO2	Guilt over Environmental Neglect	0.76
Motivation	MOT1	Desire to Improve Farming Practices	0.82
	MOT2	Economic Incentives	0.8
Behavioral Intention	BI1	Intention to Adopt New Practices	0.85
	BI2	Willingness to Expand Operations	0.83
Actual Behavior	AB1	Actual Adoption of Cage Farming Practices	0.88
	AB2	Frequency of Engagement in Farming Activities	0.86

). This was performed to identify the underlying latent constructs and refine the measurement scales for the producers’ behavioral factors. The factor analysis was conducted using maximum likelihood extraction with promax rotation. Variables with factor loadings less than 0.5 were considered weak indicators and were removed from the final measurement model. This trimming process helped ensure that each latent variable was well represented by its corresponding measurement items. The iterative factor analysis resulted in the final set of indicators presented in Section 4.5. All retained items had factor loadings above the 0.65 threshold, suggesting a strong alignment between the measured variables and their respective latent constructs. This high degree of factor loading confirms the validity and reliability of the measurement scales used in this study.

The factor analysis approach is consistent with the recommended best practices for scale development and validation in behavioral research [25,29]. The model fit indices in

Table 3. Model fit indices.

Fit Index	Value	Threshold	Interpretation
Chi-Square (χ2)	130.25	p < 0.01	Indicates overall model fit
Degrees of Freedom (df)	65	-	-
Chi-Square/df	2	<3	Acceptable fit
Root Mean Square Error of Approximation (RMSEA)	0.05	<0.08	Good fit
Comparative Fit Index (CFI)	0.94	>0.90	Good fit
Tucker–Lewis Index (TLI)	0.93	>0.90	Good fit
Standardized Root Mean Square Residual (SRMR)	0.07	<0.08	Acceptable fit

use standard thresholds (e.g., chi-square/df < 3, RMSEA < 0.08), ensuring robust SEM analysis [15].

The results show that the structural equation model for the producers has a good overall fit. The chi-square/df ratio is 1.63, which is less than the recommended threshold of 3.0. Other fit indices, such as CFI (0.94), TLI (0.93), RMSEA (0.05), and SRMR (0.07), also indicate the model has an acceptable to good fit. These results suggest the theoretical framework adequately represents the relationships among the behavioral factors for the cage tilapia producers.

4.2. Convergent Validity for Producers

The assessment of convergent validity through Average Variance Extracted (AVE) shows that all latent variables for the producers have AVE values greater than the 0.5 threshold as evident in

Table 4. Convergent validity.

Latent Variable	Indicator Count	AVE	Threshold	Interpretation
Attitude	4	0.6	≥0.5	Convergent validity confirmed
Subjective Norm	3	0.55	≥0.5	Convergent validity confirmed
Perceived Behavioral Control	3	0.62	≥0.5	Convergent validity confirmed
Environmental Awareness and Concern	2	0.57	≥0.5	Convergent validity confirmed
Moral Obligation	2	0.65	≥0.5	Convergent validity confirmed
Motivation	2	0.7	≥0.5	Convergent validity confirmed

. This confirms that the latent variables explain more than half of the variance in their respective indicators, demonstrating strong convergent validity.

4.3. Discriminant Validity for Producers

The analysis of discriminant validity indicates that the square root of the AVE for each latent variable is greater than its correlations with the other latent variables. This provides evidence that the latent variables are distinct and measure unique aspects of the producers’ behavioral factors. The factor loading results show that all indicators have loadings above 0.65, suggesting the latent variables are well represented by their respective measurement items as shown in

Table 5. Discriminant validity.

Latent Variable	Attitude	Subjective Norm	Perceived Behavioral Control	Environmental Awareness and Concern	Moral Obligation	Motivation
Attitude	0.78	0.35	0.42	0.3	0.38	0.45
Subjective Norm		0.74	0.4	0.33	0.37	0.41
Perceived Behavioral Control			0.79	0.36	0.39	0.44
Environmental Awareness and Concern				0.76	0.48	0.46
Moral Obligation					0.81	0.5
Motivation						0.84

. The high factor loadings confirm the validity of the measurement scales used in this study.

4.4. Producer Path Analysis

The path analysis for the producers indicates that all the hypothesized relationships significantly affect behavioral intentions and actual behavior, showing strong alignment with the Theory of Planned Behavior framework as presented in

Table 6. Path analysis results for cage tilapia producers.

Path	Estimate	p-Value	Significance
Attitude→Behavioral Intention	0.65	<0.001	Significant
Subjective Norm→Behavioral Intention	0.59	<0.001	Significant
Perceived Behavioral Control→Behavioral Intention	0.6	<0.001	Significant
Environmental Awareness→Behavioral Intention	0.52	<0.01	Significant
Moral Obligation→Behavioral Intention	0.53	<0.01	Significant
Motivation→Behavioral Intention	0.58	<0.01	Significant
Behavioral Intention→Actual Behavior	0.79	<0.001	Significant

. Attitudes independently drive intentions through beliefs in profitability and sustainability, subjective norms via peer influence, and perceived behavioral control through resource confidence, with minimal overlap due to distinct measurement items [8].

The structural equation modeling results confirm that attitudes, subjective norms, and perceived behavioral control are pivotal in shaping producers’ intentions and practices in cage tilapia farming. Producers with favorable views on profitability, technical feasibility, and sustainability are more likely to adopt and expand cage farming operations. These attitudes are reinforced by social influences, such as peer endorsements and community expectations, which are particularly potent in the collectivist culture of the Lake Victoria region. Confidence in managing operations, bolstered by access to resources and technical knowledge, further drives producers to implement sustainable practices as demonstrated in Figure 2.

4.5. Covariance Between Latent Variables

The covariance estimates show strong relationships among the various behavioral factors. For instance, attitudes have relatively high covariances with subjective norms, perceived behavioral control, environmental awareness and concern, moral obligation, and motivation. These findings in

Table 7. Variable covariances.

Latent Variable Pair	Covariance Estimate
Attitude↔Subjective Norm	0.35
Attitude↔Perceived Behavioral Control	0.42
Attitude↔Environmental Awareness and Concern	0.3
Attitude↔Moral Obligation	0.38
Attitude↔Motivation	0.45
Attitude↔Behavioral Intention	0.65
Subjective Norm↔Perceived Behavioral Control	0.4
Subjective Norm↔Environmental Awareness and Concern	0.33
Subjective Norm↔Moral Obligation	0.37
Subjective Norm↔Motivation	0.41
Subjective Norm↔Behavioral Intention	0.59
Perceived Behavioral Control↔Environmental Awareness and Concern	0.36
Perceived Behavioral Control↔Moral Obligation	0.39
Perceived Behavioral Control↔Motivation	0.44
Perceived Behavioral Control↔Behavioral Intention	0.6
Environmental Awareness and Concern↔Moral Obligation	0.48
Environmental Awareness and Concern↔Motivation	0.46
Environmental Awareness and Concern↔Behavioral Intention	0.52
Moral Obligation↔Motivation	0.5
Moral Obligation↔Behavioral Intention	0.53
Motivation↔Behavioral Intention	0.58
Behavioral Intention↔Actual Behavior	0.79

suggest the interconnected nature of these determinants in shaping producers’ behavior.

4.6. Discussion

The assessment of convergent and discriminant validity provides strong evidence of the robustness and distinctiveness of the measurement model for the cage tilapia producers. The finding that all latent variables have AVE values exceeding the 0.5 threshold indicates that the constructs explain a substantial portion of the variance in their respective indicators, demonstrating sound convergent validity [25]. Furthermore, the analysis of discriminant validity shows that the square root of the AVE for each latent variable is greater than its correlations with other constructs. This suggests the behavioral factors are distinct and measure unique aspects of producers’ decision-making processes, rather than overlapping or redundant concepts (see

Table 8. Hypothesis testing supported by Table 6.

Hypothesis	Hypothesis Description	Results
H1	A positive attitude significantly influences behavioral intentions.	Supported
H2	Subjective norms positively affect behavioral intentions.	Supported
H3	Perceived behavioral control positively impacts behavioral intentions.	Supported
H4	Motivation positively influences behavioral intentions.	Supported
H5	Moral obligation positively affects behavioral intentions.	Supported
H6	Environmental awareness positively influences behavioral intentions.	Supported
H7	Behavioral intentions significantly influence actual behavior.	Supported

) [28].

The establishment of both convergent and discriminant validity enhances the confidence in this study’s findings and the interpretations derived from the structural relationships. It ensures the latent variables accurately capture the multidimensional nature of the producers’ attitudes, social influences, perceived control, environmental concerns, moral obligations, and motivation, without any conceptual ambiguity or confounding effects.

The findings from the structural equation modeling analysis for the producers align with and extend the existing research on the behavioral determinants of cage tilapia farming. The strong model fit indices corroborate the applicability of the Theory of Planned Behavior (TPB) framework in the aquaculture context, as evidenced by studies such as [31,32]. The path analysis results reveal the structural relationships among the key behavioral determinants of cage tilapia producers’ intentions and actual behavior. The findings demonstrate the central role of attitudes, subjective norms, and perceived behavioral control in shaping producers’ behavioral intentions, which, in turn, positively influence their actual farming practices.

This alignment with the core tenets of the Theory of Planned Behavior underscores the importance of addressing producers’ personal beliefs, social influences, and perceived control capabilities when aiming to promote the adoption and long-term sustainability of cage tilapia farming. Producers with more favorable attitudes toward cage farming, stronger perceptions of social pressure to engage in the activity, and higher confidence in their ability to manage the operations are more likely to exhibit intentions to expand their operations and implement best practices. These results are consistent with the existing literature on TPB applications in agriculture, which emphasizes the interplay of personal, social, and perceived control factors in shaping farmers’ intentions [23,24]. The indirect influences of environmental awareness and concern, as well as moral obligation, on behavioral intentions and actual behavior, provide further insights into the multidimensional nature of producers’ decision-making. These findings suggest that incorporating environmental sustainability and ethical considerations into extension programs and policy initiatives can positively shape producers’ behavior beyond the traditional focus on economic and technical factors.

These behavioral dynamics are shaped not only by internal factors but also indirectly by external influences, such as government policies that provide subsidies or technical support, enhancing producers’ ability to adopt best practices. Regional regulations on environmental standards also encourage a sense of responsibility, aligning farming practices with sustainability goals. These external factors implicitly interact with TPB internal constructs to create a supportive ecosystem for cage farming. Attitudes reflect producers’ beliefs about the economic and ecological viability of cage farming, driving their motivation to engage actively. Subjective norms channel community pressures, encouraging alignment with shared goals, while perceived behavioral control empowers producers to overcome operational challenges. Together, these factors enable producers to contribute to a resilient value chain, with interventions targeting these drivers likely to amplify sustainable production.

This study contributes to TPB by validating its applicability in Kenyan aquaculture, confirming that attitudes, subjective norms, and perceived behavioral control strongly predict intentions [8]. Practically, these findings guide interventions: enhancing attitudes through profitability-focused training, strengthening norms via community cooperatives, and improving control through resource access (e.g., subsidies) can boost sustainable practices. For policymakers and industry stakeholders, these structural relationships offer valuable guidance for developing comprehensive strategies to support the cage tilapia sector. Interventions should also consider external factors, like government policies and market access, which shape adoption but were not modeled here, though future research could expand the TPB framework to include them [3].

The analysis of covariance between the latent variables reveals the interconnected nature of the behavioral determinants influencing cage tilapia producers’ decision-making. The positive effect of perceived behavioral control on producers’ intentions and actual behavior aligns with studies demonstrating the crucial role of farmers’ confidence in their ability to manage aquaculture operations [33]. The strong positive relationships among the constructs, such as the high covariances between attitudes and other factors, underscore the complex and multidimensional nature of producers’ behavioral dynamics.

These findings emphasize that the various behavioral drivers do not operate in isolation but rather interact and reinforce each other in shaping producers’ intentions and actual practices. For instance, producers with more favorable attitudes toward cage farming are likely to perceive stronger social pressure from their peers and community, feel a greater sense of control over their operations, and exhibit a stronger moral obligation to engage in sustainable farming practices.

The implications of these covariance patterns are twofold. First, they highlight the need for a holistic, systems-based approach to understanding and influencing producer behavior in the cage tilapia value chain. Interventions targeting a single factor, such as improving technical skills, may have limited effectiveness if they do not consider the interplay with other behavioral determinants. Second, the strong covariances suggest opportunities for synergistic and multiplier effects when designing integrated support programs. By simultaneously addressing producers’ attitudes, social networks, perceived control, environmental concerns, and ethical considerations, extension services and policymakers can potentially generate a more significant and lasting impact on the adoption of sustainable cage farming practices.

5. Consumers’ Results

5.1. Model Fit Indices for Consumers

The model fit indices indicate an acceptable to a good fit, suggesting that the conceptual model adequately represents the relationships influencing consumer behavior in the cage tilapia value chain as shown in the

Table 9. Model fit indices for consumers.

Fit Index	Value	Threshold	Interpretation
Chi-Square (χ2)	112.8	p < 0.01	Indicates overall fit
Degrees of Freedom (df)	168	-	-
Chi-Square/df	0.67	<3	Acceptable fit
Root Mean Square Error of Approximation (RMSEA)	0.04	<0.08	Good fit
Comparative Fit Index (CFI)	0.95	>0.90	Good fit
Tucker-Lewis Index (TLI)	0.94	>0.90	Good fit
Standardized Root Mean Square Residual (SRMR)	0.06	<0.08	Acceptable fit

5.2. Factor Loadings for Consumers

All factor loadings exceed the 0.65 threshold, confirming that the indicators are robust representations of the latent constructs within the model. High loadings reflect strong alignment, validated through exploratory factor analysis [25] as captured in

Table 10. Consumers’ factor loadings.

Latent Variable	Indicator Label	Indicator Description	Factor Loading
Attitude	ATT1	Interest in Consumption	0.78
	ATT2	Health Perception	0.73
	ATT3	Nutritional Value	0.76
Subjective Norm	SN1	Economic Status Influence	0.8
	SN2	Market Perception (“China Fish”)	0.77
Perceived Behavioral Control	PBC1	Ease of Consumption	0.74
	PBC2	Information Barriers	0.69
Environmental Awareness and Concern	EAC1	Health Risks Associated with Sales Practices	0.7
	EAC2	Waste Concerns	0.75
Moral Obligation	MO1	Support for Local Farmers	0.78
	MO2	Ambassadorial Role	0.74
Motivation	M1	Cage Tilapia as a Protein Source	0.82
	M2	Affordability	0.8
Behavioral Intention	BI1	Intent to Increase Intake	0.84
	BI2	Gathering More Information	0.81
Actual Behavior	AB1	Largest Cage Tilapia Consumption	0.87
	AB2	Regular Household Consumption	0.86

5.3. Convergent Validity for Consumers

Table 11. Consumer convergent validity.

Latent Variable	Indicator Count	AVE	Threshold	Interpretation
Attitude	3	0.64	≥0.5	Convergent validity confirmed
Subjective Norm	2	0.62	≥0.5	Convergent validity confirmed
Perceived Behavioral Control	2	0.6	≥0.5	Convergent validity confirmed
Environmental Awareness and Concern	2	0.65	≥0.5	Convergent validity confirmed
Moral Obligation	2	0.62	≥0.5	Convergent validity confirmed
Motivation	2	0.68	≥0.5	Convergent validity confirmed
Behavioral Intention	2	0.7	≥0.5	Convergent validity confirmed
Actual Behavior	2	0.75	≥0.5	Convergent validity confirmed

shows that the AVE values for each latent variable exceed the 0.5 threshold, indicating strong convergent validity for each construct in the consumer model.

5.4. Discriminant Validity for Consumers

The discriminant validity analysis confirms that each construct measures unique aspects of consumer behavior, with the square root of AVE for each variable being higher than its correlations with the other constructs as indicated in

Table 12. Consumer discriminant validity.

Latent Variable	Attitude	Subjective Norm	Perceived Behavioral Control	Environmental Awareness	Moral Obligation	Motivation
Attitude	0.8	0.4	0.45	0.36	0.39	0.42
Subjective Norm		0.79	0.35	0.38	0.36	0.4
Perceived Behavioral Control			0.77	0.4	0.38	0.41
Environmental Awareness and Concern				0.81	0.45	0.46
Moral Obligation					0.78	0.49
Motivation						0.82

.

5.5. Path Analysis for Consumers

The path analysis reflects significant relationships between some key behavioral drivers and intentions, while moral obligation and environmental awareness do not show significant effects on behavioral intentions, indicating that H5 and H6 are not supported. Surprisingly, moral obligation (β = 0.09, p = 0.15) and environmental awareness (β = 0.10, p = 0.12) lack significant influence, likely because consumers prioritize immediate benefits like health and affordability over ethical or environmental concerns, possibly reflecting limited awareness or economic constraints. This contrasts with producers, where moral obligation (β = 0.53, p < 0.01) is significant, as producers bear direct responsibility for environmental impacts, unlike consumers’ distal role. These non-significant findings warrant cautious interpretation, suggesting that ethical considerations may gain traction with increased education but currently play a secondary role. Non-significant paths likely stem from consumers’ focus on immediate benefits, consistent with prior seafood studies [5,11] as indicated in Figure 3 and

Table 13. Path analysis results for consumers.

Path	Estimate	p-Value	Significance
Attitude→Behavioral Intention	0.62	<0.001	Significant
Subjective Norm→Behavioral Intention	0.5	<0.001	Significant
Perceived Behavioral Control→Behavioral Intention	0.58	<0.001	Significant
Environmental Awareness→Behavioral Intention	0.1	0.12	Not Significant
Moral Obligation→Behavioral Intention	0.09	0.15	Not Significant
Motivation→ Behavioral Intention	0.44	<0.001	Significant
Behavioral Intention→Actual Behavior	0.69	<0.001	Significant

below.

5.6. Covariance Between Latent Variables for Consumers

The covariance estimates in

Table 14. Covariance estimates.

Latent Variable Pair	Covariance Estimate
Attitude↔Subjective Norm	0.4
Attitude↔Perceived Behavioral Control	0.45
Attitude↔Environmental Awareness	0.36
Attitude↔Moral Obligation	0.39
Attitude↔Motivation	0.42
Subjective Norm↔Perceived Behavioral Control	0.35
Subjective Norm↔Environmental Awareness	0.38
Subjective Norm↔Moral Obligation	0.36
Subjective Norm↔Motivation	0.4
Perceived Behavioral Control↔Environmental Awareness	0.4
Perceived Behavioral Control↔Moral Obligation	0.38
Perceived Behavioral Control↔Motivation	0.41
Environmental Awareness↔Moral Obligation	0.45
Environmental Awareness↔Motivation	0.46
Moral Obligation↔Motivation	0.49
Behavioral Intention↔Actual Behavior	0.66

show positive relationships among various constructs, indicating that, despite the lack of direct influence on intentions, environmental awareness and moral obligation interact with other variables, affecting overall consumer behavior in indirect ways.

5.7. Discussion

Consumers’ intentions to purchase cage-farmed tilapia are primarily driven by positive attitudes toward its health benefits, nutritional value, and affordability, as evidenced by the structural equation modeling results. Social influences, including family traditions and peer recommendations, further shape these intentions, reflecting the cultural significance of fish consumption in the Lake Victoria region. Ease of access to tilapia and confidence in identifying quality products also facilitate purchasing decisions, highlighting the importance of perceived behavioral control. However, environmental awareness and moral obligation do not significantly influence consumer intentions, possibly because consumers prioritize immediate benefits like health and cost over long-term ecological concerns. Limited knowledge about the environmental impact of cage farming may further diminish these factors’ influence, suggesting a need for targeted education to bridge this gap.

Misconceptions, such as the “China fish” label, undermine consumer trust in cage-farmed tilapia’s quality and origin. Addressing these through transparent communication, certifications, and public campaigns can enhance market acceptance. Additionally, supply chain efficiency—encompassing transportation, storage, and market access—plays a critical role in ensuring product quality and availability, directly affecting consumer choices. These insights inform strategies for policymakers, extension services, and industry stakeholders, who can foster consumer confidence and demand by improving supply chains and promoting the benefits of locally produced tilapia.

These findings suggest that while consumer intentions are largely driven by personal attitudes, social influences, and a sense of control, moral obligation and environmental awareness do not play as significant roles in shaping intentions as they do for producers. For stakeholders, like NGOs and local governments, this indicates that marketing strategies should emphasize health, affordability, and accessibility, while educational campaigns could gradually increase environmental awareness to shift preferences [5]. Supply chain factors, such as transport and market access, also influence perceptions, particularly for high-volume consumers who may prioritize reliability [15]. Consumers dealing with higher fish volumes may exhibit stronger health-driven preferences, though economic models show uniform value perceptions across the sample [21].

Theoretically, this study extends TPB by highlighting its limitations in predicting consumer behavior when ethical constructs like moral obligation are non-significant, suggesting a need for context-specific adaptations. Practically, stakeholders can leverage these findings to design targeted interventions: NGOs can run radio campaigns to debunk “China fish” myths, while governments can improve market access to enhance perceived control. These strategies align with the significant paths (e.g., attitude, β = 0.62, p < 0.001) identified in the path analysis.

A key counterintuitive finding is the non-significant role of moral obligation and environmental awareness for consumers, unlike producers. Producers’ direct involvement in farming practices ties their moral obligation to tangible environmental outcomes (e.g., pollution control), whereas consumers’ distal role—purchasing rather than producing—reduces their sense of responsibility. This discrepancy suggests that consumer education on sustainability could bridge this gap, but immediate priorities like affordability currently dominate as shown in

Table 15. Hypothesis testing.

Hypothesis	Hypothesis Description	Results
H1	A positive attitude significantly influences behavioral intentions.	Supported
H2	Subjective norms positively affect behavioral intentions.	Supported
H3	Perceived behavioral control positively impacts behavioral intentions.	Supported
H4	Motivation positively influences behavioral intentions.	Supported
H5	Moral obligation positively affects behavioral intentions.	Not Supported
H6	Environmental awareness positively influences behavioral intentions.	Not Supported
H7	Behavioral intentions significantly influence actual behavior.	Supported

.

The investigation of consumer behavior in the cage tilapia market represents a critical extension of previous scholarly work examining food choice dynamics, building upon foundational research by [29] in seafood consumption patterns. Our findings resonate strongly with their seminal study, which similarly identified attitude, subjective norms, and perceived behavioral control as primary determinants of consumer intentions. However, our research provides a more nuanced exploration of these constructs within the specific context of cage-raised tilapia, offering unprecedented insights into the complex decision-making processes that drive food consumption.

Previous studies by [34] in the aquaculture sector highlighted the significance of health perceptions in seafood consumption, a finding our research comprehensively validates. Our analysis reveals that consumers’ attitudes toward tilapia are predominantly shaped by three critical dimensions: nutritional value, health perception, and consumption interest. With factor loadings of 0.76, 0.73, and 0.78, respectively, these findings extend beyond the existing literature by demonstrating the complexity in the nature of consumer attitudes toward farmed fish. Notably, our research shows a more profound understanding of how these perceptions interact, suggesting that nutritional knowledge and health awareness are interconnected cognitive processes rather than discrete factors. Consumers dealing with higher fish volumes may exhibit stronger health-driven preferences, though economic models show uniform value perceptions across the sample.

The social dimension of food choice, explored extensively by [35] in their theoretical framework of planned behavior, finds robust support in our study. The subjective norm construct, with a path estimate of 0.50, reveals the profound impact of social and economic contexts on consumption intentions. Economic status influence and market perceptions emerge as powerful mediators, a finding that aligns with and expands upon Bourdieu’s concept of cultural capital. Unlike previous studies that treated social influences as peripheral, our research demonstrates how economic and social narratives are deeply embedded in consumer decision-making processes.

Particularly intriguing are the counterintuitive findings regarding environmental awareness and moral obligation. While [11] suggested direct correlations between environmental consciousness and consumption choices, our study reveals a more complex relationship. The non-significant direct paths of environmental awareness (0.10) and moral obligation (0.09) challenge simplistic linear models of consumer behavior. Instead, our analysis suggests these factors operate through subtle and indirect mechanisms, indicating a more delicate interaction between ethical considerations and food choices.

The predictive power of behavioral intention, confirmed by a robust path estimate of 0.69 between intention and actual behavior, builds upon and validates the methodological approach of [36] in consumer behavior research. This finding is particularly significant in the context of food consumption studies, where the gap between stated intentions and actual behavior has long been a critical research challenge. Our study provides empirical evidence of the strong correlation between consumer intentions and actual purchasing and consumption behaviors.

The practical implications of this study extend far beyond academic discourse. For stakeholders in the cage tilapia value chain, this research offers strategic insights that challenge existing marketing approaches. The findings suggest a shift from generic health messaging to more targeted communication strategies that address specific consumer attitudes, reduce perceived barriers to consumption, and leverage social and economic narratives. This approach represents a significant departure from the one-size-fits-all marketing strategies prevalent in the aquaculture sector.

The limitations of this study include the cross-sectional design, which limits causal inference and reliance on self-reported data, which may introduce bias [30]. Future studies could use longitudinal designs or objective consumption metrics to enhance robustness.

6. Summary of the Findings

This study reveals distinct yet complementary behavioral drivers for producers and consumers within the cage fish value chain. For producers, attitudes toward the economic and practical aspects of cage tilapia farming play a central role. Specifically, their “interest in cage tilapia production” and perceptions of its “capital intensity” and required “technical skills” shape their intentions to adopt and expand production. Additionally, subjective norms, such as “peer influence” and “community expectations,” strongly impact production choices by fostering an environment of community support and shared goals. Producers are further influenced by perceived behavioral control, including “confidence in managing operations” and “access to resources,” which enables them to navigate challenges effectively. Together, these factors encourage producers to adopt sustainable practices, while moral obligations and environmental awareness reinforce their commitment to responsible aquaculture.

Consumers, on the other hand, are primarily motivated by attitudes concerning the health, nutritional value, and affordability of cage-farmed tilapia. Indicators, such as “interest in consumption,” “health benefits,” and “affordability,” are key drivers in shaping consumer demand. Social influences also play a role, as perceptions like the “China fish” label affect views on product quality and authenticity. Perceived control factors, such as the “ease of access” and familiarity with “quality indicators,” further simplify purchasing decisions, making it easier for consumers to choose cage-farmed tilapia. These drivers indicate that consumers prioritize tangible benefits, such as health and quality, over environmental or moral considerations.

The relationship between production and consumption behaviors in the value chain reveals a feedback loop where each side reinforces the other. Sustainable practices adopted by producers, such as using “certified inputs” and environmentally “friendly inputs,” foster consumer trust and preference for cage-farmed tilapia, driving demand. In turn, consumer feedback on aspects like product quality and ethical standards informs producers, who can adapt their practices to better align with market demands. This reciprocal relationship strengthens the cage fish value chain, creating a responsive and resilient system that can adapt to evolving consumer preferences and environmental challenges, ultimately supporting the long-term stability of the sector.

7. Conclusions and Recommendations

This study highlights the complementary roles of producers and consumers in shaping the sustainability of the cage tilapia sector in Kenya’s Lake Victoria region. For producers, attitudes about profitability, technical demands, and sustainability, alongside social norms and high perceived behavioral control, encourage the adoption of responsible farming practices. Conversely, consumer behavior is primarily influenced by health, nutritional value, and affordability, with less emphasis on environmental and ethical considerations. This alignment of attitudes and motivations between both groups creates a feedback loop in which producer practices support consumer preferences and consumer demand, in turn, incentivizing producers to maintain quality and sustainable production standards. Together, these dynamics underscore the potential of a behaviorally informed approach to enhance the sustainability of cage tilapia farming.

Finally,

Table 16. Summary of recommendations for cage tilapia value chain stakeholders.

Stakeholder	Recommendation	Actionable Example
Producers	Enhance technical and sustainable practices	Participate in training programs like those offered by Kenya Marine and Fisheries Research Institute (KMFRI) on eco-friendly inputs
Consumers	Increase awareness of health and quality benefits	Launch campaigns via local radio to debunk “China fish” misconceptions and highlight nutritional value
Policymakers	Provide sustainability incentives	Offer subsidies for certified inputs, as piloted in Homa Bay County
NGOs	Facilitate feedback mechanisms	Establish community forums to share consumer preferences with producers
Industry	Strengthen supply chain reliability	Invest in cold chain transport to ensure consistent market access

shows the summary of recommendations and feasible action that can bring meaningful change in cage tilapia value chain.