Factors Influencing Consumers’ Purchases of Snook (Centropomus viridis) and Red Snapper (Lutjanus peru) from Artisanal Aquaculture Cooperatives in Mexico

Abstract

Aquaculture in developing countries, including Mexico, primarily consists of artisanal activities characterized by low-scale production. No studies in these regions has analyzed consumer preferences regarding aquaculture products, specifically for snook and red snapper. Consequently, to explore consumer preferences, the primary objective of this study is to estimate a measure of welfare in the form of willingness to pay (WTP) for fish cultivated under small-scale production conditions in floating cages. To examine the variables impacting WTP, we employed the random utility model (RUM) theoretical framework and a grouping of econometric models belonging to the discrete choice framework. The results strongly suggest that product attributes, such as presentation, purchasing location, consumption-related characteristics, and certain socioeconomic variables, significantly influence the decision to select farmed products. Developing productive aquaculture projects in Mexico and other developing countries can enhance community development by providing producers with accurate information for decision-making and by expanding the fish supply in response to the growing consumer demand.

1. Introduction

According to FAO, global aquaculture production reached 50.92% of the total aquatic animal production in 2022, amounting to 94.4 million tons in live weight. This marked the first time in history that aquaculture production surpassed capture fisheries, which accounted for 49.08% (91 million tons in live weight). In the same year, global aquaculture accounted for 62.61% of production from inland waters and 37.39% from marine environments. The average annual growth rate of aquaculture in marine waters was 30.85% from the 1990s to 2022, increasing from 9.2 million to 35.3 million tons in live weight. By 2022, marine and inland aquaculture contributed approximately 35.3 million tons in live weight, with a 52.9% share in mollusk farming, 24.3% in finfish, 21.5% in crustaceans, and 1.3% in other aquatic animals [1].

In Mexico, the aquaculture sector accounted for 14% of the 2 million tons in live weight produced by the fishing sector in 2022, translating to approximately 0.292 million tons. In 2022, marine aquaculture represented 29.05% of total production, with key resources including crustaceans (shrimp accounting for 17.16%), mollusks (oysters, 9.75%), fish (tuna, 2.14%), and other products (0.84%, including ornamental fish and frogs) [2].

Globally, approximately 59.5 million people engage with fisheries and aquaculture, with 20.53 million working directly in aquaculture. Most individuals involved in fisheries and fish farming reside in developing countries, where small-scale artisanal fisheries actively participate in aquaculture activities [1]. According to the World Bank, artisanal fishing generates 16 times more jobs than large-scale fishing [3].

According to FAO, artisanal aquaculture “is characterized by low-income levels, limited investment, restricted access to technology and credit, and low productivity in small areas, often in remote locations. These enterprises frequently lack defined property rights and experience poor access to market information and pricing, resulting in limited bargaining power” [4] (p. 5). Martínez indicates that the configuration of marine fish aquaculture activities in Mexico is influenced by local oceanographic conditions and by the organization of producers, who may belong to the social sector composed of fishing cooperatives, medium-scale farmers, and technologically advanced companies, characteristics particularly distinct in the country’s northwest region [5].

The National Fisheries and Aquaculture Master Plan aims to produce 2.74 million tons (landed weight) by 2030. By 2030, the mariculture agenda includes plans to distribute various species, including snook and snapper. The target for snook production is 0.021 million tons in live weight, with seven production units established across Region IV (Campeche, Yucatán, and Quintana Roo, with four units), Region I (Baja California, Baja California Sur, Sonora, Sinaloa, and Nayarit, with one unit), and Region III (Tamaulipas and Veracruz, with three units). For the Pacific snapper, the strategy includes establishing 24 production units to produce 0.072 million tons, with eight units in Region I and 16 units in Region II (Jalisco, Colima, Michoacán, Guerrero, Oaxaca, and Chiapas) [6].

Aquaculture serves as a primary alternative and complementary economic activity to the commercial fishing sector, helping ensure food security amid population growth and the expanding diversity of preferences for aquatic products, as demand consistently rises [7]. Recently, researchers have intensified their efforts to measure the impacts of aquaculture consumption preferences. This surge in interest stems from the growth in aquaculture production and the shifting international consumption patterns that derive from the introduction of various products on the market [8,9]. Over the last twenty years, fifty studies have explored preferences for farmed fish, internationally. These initial studies originated primarily in Europe and the United States, with most contributions arising from these two geographic areas. However, our Supplementary Data (Table S1) also encompass research from regions such as Asia, Africa, and Latin America. Notably, 48% of all these studies (Table S1) utilized internet-based methods to interview participants and collect information.

A commonly used measure in the study of fish consumption preferences is willingness to pay (WTP), which quantifies the economic benefit that consumers derive from the difference between what they pay for a good and the maximum price they are willing to pay, known as consumer surplus. This economic benefit is expressed as the income that consumers are willing to forgo to acquire a product or its specific attributes. The WTP reflects market demand and represents the maximum amount that an individual is willing to pay to obtain a particular fish species or its attributes. Therefore, this measure is valuable for producers, as it provides insights into setting optimal prices for the products they intend to sell. It also allows producers to evaluate which goods will likely yield profitable investments to maximize their profits.

When analyzing consumer choices, researchers must explain the factors that cause variability in preferences which can be considered attributes affecting individuals’ utility. These factors can be categorized into observable and unobservable components, as some attributes relate directly to the good in question. In contrast, others are associated with consumer preferences expressed through socioeconomic predictors. The configuration of these factors enables the design of a choice model that accounts for the heterogeneity of consumer preferences, assuming that individuals seek to maximize their utility [8,9]. For these reasons, this study aims to estimate WTP, as it allows us to understand how each product attribute and socioeconomic variable, among other factors, influences the consumer decision-making process.

The literature has analyzed, thus far, fish consumption preferences to determine acceptance through willingness to pay (WTP) for farm-raised versus caught fish. Researchers have also explored additional topics, including product labels, certifications, and environmentally friendly practices. They have considered factors such as fish species color, organic labels, aquaculture management types, product quality, origin, the use of insect meal in fish feed, and the acceptance of genetically modified fish, among other attributes (see Table S1 for further details). WTP amounts fluctuate based on locality, country, fish welfare, and the specific questions posed to the target population.

International studies have identified several factors influencing WTP, which can be classified into five main components: (1) product attributes and preferences, such as visibility, quality, processing type, market presentation, product information, deformities, shape ratings, brand perception, package size, flavor, and organic status, (2) considerations for sustainable aquaculture management, including product quality labels and certifications related to sanitary, health, and nutrition standards, cultivation practices, wastewater usage, adherence to specific fish farming standards, antibiotic application, survival and fattening rates, and genetic modification of the product, (3) comparisons among different consumer groups that encompass predictors such as household income levels, age, occupation, household size, education level, and marital status, (4) aspects related to traceability, such as product origin and consumption barriers, and (5) variables associated with WTP, including the maximum price consumers are willing to pay, the type of store where they purchase fish, and promotional factors such as discounts and product usage information.

Much of the research examining WTP for aquaculture fish has prioritized the discrete choice experiment (DCE) methodology, a well-established method in the stated preference (SP) segment which has gained popularity in recent decades for assessing environmental impacts and fish preferences [8,9,10]. Nevertheless, the contingent valuation method (CVM) continues to play a significant role in research with respect to consumption preferences for aquaculture fish, especially in relation to products that do not yet have a market. This method, along with the random utility maximization model (RUM) for private goods, offers ease of application, consistency with economic choices, and minimizes bias in survey design and implementation [11,12,13,14,15]. Other methods utilized to a lesser extent include conjoint analysis, experimental auction methods (EAM), label choice experiments (LCE), and hedonic scores (HS).

Econometric specifications frequently referenced in the WTP literature for farmed fish include the mixed logit (ML), latent class model (LCM), multinomial logit model (MNL), and conditional logit model (CLM), with the logit and probit models also being applied due to their statistical advantages, particularly in relation to the generation of consistent and efficient estimates [11,12,13,14,15]. Researchers continue to incorporate a variety of econometric treatments in their analyses, including ordinary least squares (OLS), weighted least squares (WLS), random effects (RE), heteroscedastic logit (HL), count data models, ordered choice models, and bivariate models.

To date, no study has examined the acceptance of aquaculture snook and red snapper consumption in either international or Mexican contexts. For Mexican artisanal cooperatives, aquaculture represents a viable production system and an opportunity to enhance income by marketing high-value marine products. In particular, the snook and red snapper can be cultured effectively in states such as Oaxaca and Colima. Thus, Mexican cooperatives require insights into consumer preferences and the attributes of the products they offer or intend to produce to understand market trends and identify potential opportunities. Therefore, this study aims to estimate the willingness to pay (WTP) for farmed snook and red snapper produced by artisanal Mexican cooperatives.

2. Materials and Methods

2.1. Consumer Acceptance of Artisanal Aquaculture Fish Using RUM as a Theoretical Framework

Economists formally analyze WTP responses in relation to aquaculture fish by implementing discrete/continuous choice (DCO) models. They express these models through the theory of random utility maximization (RUM), which describes mathematically how individuals’ preferences over two exclusive options guide their decision-making process to choose the alternative that will maximize their expected utility [16,17,18]. Given the individual’s utility function $U\left(·\right)$, the consumer decision-making process can be expressed in the following way:

$$U\left(A,X\right)-U\left(NA,X\right)\ge 0$$

(1)

where $A$ indicates that the individual accepts a scenario of paying the price for consuming aquaculture fish, $NA$ indicates maintaining the status quo and rejecting the proposed scenario, and $X$ is a set of explanatory variables. If the left term of function (1), which is called the indirect utility function $V\left(·\right)$ and is made up of the deterministic component and a stochastic or random component, is added, the error term $\epsilon$ emerges. Therefore, the function can be expressed as follows:

$$U\left(A, X\right)=V\left(A,X\right)+\epsilon$$

(2)

If the probability of obtaining a positive response in relation to the scenario of purchasing aquaculture fish by the individual is expressed, then we obtain:

$$Pr\left(Yes\right)=\left[V^1\left(A,X\right)\right]+{\epsilon }_1>V^0\left(NA,X\right)+{\epsilon }_0$$

(3)

where ${\epsilon }_1$ is the random component that includes non-deterministic preferences that are known to, but not observable by, the researcher, meaning that it represents the deviations that the rest of the individuals have with respect to the average individual. Following the above, the probability is expressed as:

$$Pr\left(Yes\right)={\left(1+e^{\Delta V}\right)}^{-1}$$

(4)

where $\Delta V=V^1-V^0$. In this sense, if the individual’s response manifests itself as a rejection of the scenario, that is, a willingness to remain in the status quo, the probability can be described as follows:

$$Pr\left(No\right)={\left(1+e^{\Delta V}\right)}^{-1}$$

(5)

Under this theoretical approach, the individual increases his welfare due to a change in his utility level when he accepts the scenario that consists of the acquisition of artisanal aquaculture snook and red snapper, a situation that causes a decrease in its income when this is destined to the purchase of the fish, a measure known as compensatory variation. The expected value of the willingness to pay (WTP) is based on the cumulative distribution function (CDF), as follows:

$$E\left(WTP\right)=\underset{0}{\overset{\infty }{\int }}\left[1-F\left(C\right)dC\right]$$

(6)

where $C$ is the proposed price. If $U_{i1}$ is the individual’s utility $i$ derived from choosing the proposed scenario and $U_{i0}$ is the individual’s utility $i$ derived from rejecting the scenario, then:

$$\begin{gathered} U_{i0}={\alpha }_0+X_{i0}\beta +{\epsilon }_{i0} \\ {U}_{i1}={\alpha }_0+X_{i1}\beta +{\epsilon }_{i1 } \end{gathered}$$

(7)

Such equations indicate that the individual chooses the proposed scenario (option 1) when the utility of that decision exceeds the alternative of rejecting the scenario (option 0) and vice versa. Consequently, the dependent variable is qualitative binary when Equation (8) is equivalent to the indirect utility function, with the terms on the right being the deterministic and stochastic components:

$$Y_i=X_i\beta +{\epsilon }_i \mathit{where}i:\mathrm{1,2}\dots N$$

(8)

In this study, the dependent variable took on a value of 1 for option 1 and a value of 0 for option 0. Therefore, this can be expressed as:

$$Y_i=\left\{\begin{array}{c}1 si U_{i1}>U_{i0}\\ 0 si U_{i1}<U_{i0}\end{array}\right.$$

(9)

Then, the probabilistic model can be expressed as:

$$P_i=Pr=\left(Y_i=1\right)=Pr\left(I^{\ast }>0\right)=Pr\left(X_i\beta +{\epsilon }_i>0\right)=F\left(X_i\beta \right)$$

(10)

F(·) corresponds to the cumulative distribution function related to a stochastic component, which can follow a logistic or standard normal distribution. The coefficients of the discrete choice models were estimated through the maximum likelihood function (

Table 1. Description of the variables of the discrete choice models.

Variables	Description	Average	Standard Deviation	Min.	Max.
pricesnook	Continuous variable that indicates the monetary amount the individual would be willing to pay for farmed snook.	13.0936	2.2743	11.4533	20.0433
priceredsnapper	Continuous variable that indicates the monetary amount the individual would be willing to pay for farmed red snapper.	13.1316	2.2160	11.4533	20.0433
income2	Dichotomous variable that takes the value of 1 if the monthly family income is between $229.12 USD and $458.13 USD and takes the value of 0 otherwise.	0.1490	0.3567	0	1
income3	Dichotomous variable that takes the value of 1 if the monthly family income is between $458.19 USD and $687.20 USD and takes the value of 0 otherwise.	0.1192	0.3246	0	1
income4	Dichotomous variable that takes the value of 1 if the monthly family income is between $687.25 USD and $916.27 USD and takes the value of 0 otherwise.	0.1225	0.3284	0	1
income5	Dichotomous variable that takes the value of 1 if the monthly family income is between $916.32 USD and $1145.33 USD and takes the value of 0 otherwise.	0.1358	0.3431	0	1
income6	Dichotomous variable that takes the value of 1 if the monthly family income is greater than $1145.33 USD and takes the value of 0 otherwise.	0.3113	0.4638	0	1
specie1	Dichotomous variable that takes the value of 1 if the individual would like to have greater availability of local species and takes the value of 0 otherwise.	0.5728	0.4955	0	1
recipes	Dichotomous variable that takes the value of 1 if the individual would like to know recipes and cooking tips and takes the value of 0 otherwise.	0.3808	0.4864	0	1
promo	Dichotomous variable that takes the value of 1 if the individual would like to have discounts or promotions on fish products and takes the value of 0 otherwise.	0.3675	0.4829	0	1
ambient	Dichotomous variable that takes the value of 1 if the individual wants information about the sustainability and environmental impact of different species and takes the value of 0 otherwise.	0.3742	0.4847	0	1
source	Dichotomous variable that takes the value of 1 if the individual wants to know the origin and nutritional value and takes the value of 0 otherwise.	0.5066	0.5008	0	1
variety	Dichotomous variable that takes the value of 1 if the individual wants to have a greater variety of fish products in establishments and takes the value of 0 otherwise.	0.3311	0.4714	0	1
age1	Dichotomous variable that takes the value of 1 if the individual’s age is between 18 and 29 years and takes the value of 0 otherwise.	0.2086	0.4070	0	1
age2	Dichotomous variable that takes the value of 1 if the individual’s age is between 30 and 39 years and takes the value of 0 otherwise.	0.2152	0.4117	0	1
age3	Dichotomous variable that takes the value of 1 if the individual’s age is between 40 and 49 years and takes the value of 0 otherwise.	0.2152	0.4117	0	1
age5	Dichotomous variable that takes the value of 1 if the individual’s age is between 60 and 69 years and takes the value of 0 otherwise.	0.0960	0.2951	0	1
age6	Dichotomous variable that takes the value of 1 if the individual’s age is greater than 69 years and takes the value of 0 otherwise.	0.0331	0.1792	0	1
without viscera	Dichotomous variable that takes the value of 1 if the individual prefers to buy whole refrigerated fresh fish without viscera and takes the value of 0 otherwise.	0.0530	0.2244	0	1
viscera	Dichotomous variable that takes the value of 1 if the individual prefers to buy whole refrigerated fresh fish with viscera and takes the value of 0 otherwise.	0.3278	0.4702	0	1
chunks	Dichotomous variable that takes the value of 1 if the individual prefers to buy refrigerated fresh fish in chunks and takes the value of 0 otherwise.	0.0960	0.2951	0	1
skinless	Dichotomous variable that takes the value of 1 if the individual prefers to buy refrigerated fresh skinless fillet fish and takes the value of 0 otherwise.	0.5099	0.5007	0	1
skin	Dichotomous variable that takes the value of 1 if the individual prefers to buy refrigerated fresh fillet fish with skin on and takes the value of 0 otherwise.	0.1623	0.3693	0	1
medallions	Dichotomous variable that takes the value of 1 if the individual prefers to buy refrigerated fresh fish medallions and takes the value of 0 otherwise.	0.1589	0.3662	0	1
grounded	Dichotomous variable that takes the value of 1 if the individual prefers to buy refrigerated fresh grounded or chopped fish and takes the value of 0 otherwise.	0.0828	0.2760	0	1
smoked	Dichotomous variable that takes the value of 1 if the individual prefers to buy smoked fish and takes the value of 0 otherwise.	0.3344	0.4726	0	1
frozen	Dichotomous variable that takes the value of 1 if the individual prefers to buy frozen fish and takes the value of 0 otherwise.	0.4470	0.4980	0	1
canned	Dichotomous variable that takes the value of 1 if the individual prefers to buy canned fish and takes the value of 0 otherwise.	0.3046	0.4610	0	1
beach	Dichotomous variable that takes the value of 1 if the individual prefers to buy fish directly from the fisherman on the beach and takes the value of 0 otherwise.	0.6689	0.4714	0	1
self-employed	Dichotomous variable that takes the value of 1 if the individual is self-employed (independent) and takes the value of 0 otherwise.	0.1391	0.3466	0	1
unemployed	Dichotomous variable that takes the value of 1 if the individual belongs to the economically inactive population (unemployed) and takes the value of 0 otherwise.	0.2616	0.4402	0	1
freshness	Dichotomous variable that takes the value of 1 if the individual, when purchasing the fish, considers several indicators that determine “freshness” as the most important property and takes the value of 0 otherwise.	0.8046	0.3971	0	1

).

Once the binary response models were estimated, their coefficients were used to calculate the expected value of the willingness to pay, as follows [19]:

$$E\left({WTP}_j|\alpha ,\beta ,z_j\right)={\displaystyle \frac{\alpha z_j}{\beta }}$$

(11)

where $\alpha$ is the constant term, $z$ is a vector of socioeconomic variables and attributes of the fish (

Table 2. Results of the discrete choice models.

Variables	Snook				Red Snapper
	Probit Model		Logit Model		Probit Model		Logit Model
	Coefficient	Z-Statistic	Coefficient	Z-Statistic	Coefficient	Z-Statistic	Coefficient	Z-Statistic
Constant	−1.4204 *	−1.84	−2.7322 *	−1.94	−0.8318	−1.05	−1.3718	−0.95
pricesnook	−0.0070 **	−2.49	−0.0113 **	−2.24
priceredsnapper					−0.0063 **	−2.27	−0.0125 **	−2.34
income2	0.4847	1.21	0.8288	1.23	0.6832	1.27	1.5081	1.46
income3	0.6038	1.30	1.1130	1.33	0.5576	1.21	1.2032	1.39
income4	0.9644 **	2.09	1.8048 **	2.04	1.4848 ***	3.20	2.6378 ***	3.14
income5	1.3389 ***	2.81	2.2406 **	2.48	2.7024 ***	4.86	4.8595 ***	4.37
income6	0.7743 *	1.86	1.3732 *	1.86	1.1733 ***	2.67	2.3368 ***	2.75
specie1	0.4094 *	1.92	0.6835 *	1.80	0.1919	1.00	0.3566	1.08
recipes1	−0.2484	−1.08	−0.4440	−1.01	0.6797 ***	2.69	1.1751 **	2.33
promo1	0.4862 **	2.12	0.9296 **	2.09	−0.5139 ***	−2.17	−0.8970 *	−1.93
ambient	0.6572 ***	2.91	1.1827 **	2.55	0.4176 ***	1.86	0.7413 *	1.70
source	−0.2304	−1.13	−0.3442	−0.92	0.1914	0.86	0.2474	0.57
variety	−0.1385	−0.60	−0.1134	−0.25	−0.5477 ***	−2.37	−0.9102 **	−2.10
age1	0.3249	0.92	0.5416	0.86	2.1765 ***	5.33	3.9890 **	4.50
age2	0.6666 **	2.13	1.0785 *	1.91	0.2524	0.86	0.4182	0.81
age3	0.1481	0.46	0.2500	0.41	1.6046 ***	4.20	2.9528 ***	3.57
age5	0.0249	0.07	0.0206	0.04	−0.1420	−0.39	−0.2617	−0.42
age6	1.3311 *	1.92	2.4753 *	1.80	−0.3269	−0.61	−0.5301	−0.56
without viscera	1.2012 ***	2.74	1.8596 **	2.43	−0.5115	−1.38	−1.0311	−1.62
viscera	0.0465	0.21	0.0600	0.14	0.3442 *	1.66	0.4851	1.27
chunks	1.3966 ***	3.28	2.2433 ***	2.88	0.6066 *	1.75	0.8490	1.31
skinless	0.5616 **	2.30	0.8407 *	1.76	1.1692 ***	4.98	2.0724 ***	4.28
skin	2.2912 ***	5.51	3.8164 ***	5.21	1.2000 ***	4.29	1.9677 ***	4.05
medallions	0.3276	1.12	0.5899	1.01	−0.0765	−0.24	−0.1163	−0.18
grounded	−0.5837 *	−1.74	−0.9486 *	−1.65	−1.1493 ***	−3.37	−1.9292 ***	−3.07
smoked	0.0770	0.35	0.2590	0.62	−0.9017 ***	−3.96	−1.5032 ***	−3.69
frozen	1.0751 ***	4.84	1.8552 ***	4.36	0.1766	0.86	0.3340	0.92
canned	−0.2903	−1.27	−0.5090	−1.16	−0.0609	−0.26	0.0173	0.04
beach	0.0443	0.19	0.1880	0.40	−0.6474 ***	−2.92	−1.1147 ***	−2.64
self-employed	0.8402 ***	2.86	1.3974 ***	2.69	−0.8305 ***	−2.78	−1.3767 **	−2.34
unemployed	0.2189	0.75	0.3859	0.73	−0.1099	−0.36	−0.2092	−0.37
freshness	0.9162 ***	2.98	1.5637 ***	2.70	0.9692 ***	3.37	1.7064 ***	3.11
Log likelihood	−119,017		−119,074		−111,095		−110,826
Log likelihood	−197,493		−197,493		−191,051		−191,051
restricted
Statistical LR	156,952		156,838		159,912		160,450
AIC	302.0341		302.1479		286.1903		285.6529
BIC	420.7677		420.8816		404.9240		404.3866
Count R2	0.795		0.808		0.828		0.844
Pseudo R2	0.3974		0.3971		0.4185		0.4199

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%.

), and $\beta$ is the coefficient of the proposed price.

2.2. Survey Description

Traditional data collection strategies, such as face-to-face interviews, telephone surveys, and mail questionnaires, often incur high costs and are time-consuming. However, technological advances have introduced innovative alternatives. Researchers have shown significant growth with respect to the collection of field data for scientific research through internet-based methods, such as online platforms and email, which typically involve lower costs.

One notable advantage of these data collection methods is their ability to quickly gather large quantities of information with high response rates in less time. Additionally, these methods provide greater access to a diverse range of groups and individuals which would be challenging to achieve through traditional means [20,21].

Given the limited budget for the research, we used surveys to organize a dissemination campaign through a “segmented campaign” on Facebook and Instagram. This tool increases the visibility and reach of a specific publication on the social networks used. Our team paid Facebook to show the post to a broader and more specific audience than that which would typically see the content organically. We organized the campaign so as to have a specific publication that used the Mexican states of Colima and Oaxaca as pilot sites, sharing the hyperlink of an online “SurveyMonkey” form through the Facebook social network. This platform offers a variety of tools and features that allow researchers to customize surveys, analyze data, create reports, and share results effectively. In short, “SurveyMonkey” is a versatile tool for obtaining valuable information through online surveys and questionnaires.

The campaign performed well with respect to total captured responses, with 302 responses collected across both states. Colima attained a higher proportion of responses, i.e., 59%, compared to the 41% attained in Oaxaca. It is essential to highlight that Facebook estimated the audience reach when designing the campaign, successfully reaching the entirety of the targeted audience within each state. In Colima, the projected reach was between 0.535 million and 0.629 million people, while, in Oaxaca, the potential reach ranged from 2.3 million to 2.7 million people. These data align with the Mexican 2020 Population and Housing Census from the Instituto Nacional de Estadística y Geografía (INEGI), which reported a population over 18 years of age of 0.52 million in Colima and 2.75 million in Oaxaca. Based on this information, researchers estimated the sample size using the technique of stratified random sampling with proportional allocation [22] to determine its representativeness. The result was approximately 302 surveys, with a margin of error of 5.64%.

One of the main limitations that researchers face in survey collection is the limited willingness of people to respond to electronic surveys. Today, information saturation and increased distrust among online users have created an environment where participation in electronic surveys is becoming more selective. Many individuals avoid sharing their opinions or personal data online, making it significantly more difficult for researchers to obtain valid responses. Additionally, choosing to conduct electronic surveys comes with its restrictions. By opting for this modality, researchers limit the target audience to only individuals with internet access [21] and a Facebook account. This approach excludes a significant portion of the population, including individuals without access to technology, individuals who do not use social networks, or individuals who prefer to keep their information private. Consequently, this reduces the pool of potential respondents and complicates the achievement of the proposed objectives.

Another factor that influenced our results was people’s time availability. Nowadays, people have increasingly busy schedules and multiple responsibilities, so their willingness to participate in surveys is limited. The lack of incentives or rewards may also have decreased individuals’ interest in completing electronic surveys. Cobanoglu, C., Cobanoglu, N., and Wright pointed out that specific incentives lead to faster survey response speeds [21,23].

The geographic diversity of study locations, such as Colima and Oaxaca, may have presented additional challenges. Cultural, economic, and social factors can influence people’s willingness to participate in surveys, and researchers must address these factors appropriately to achieve meaningful and accurate representation. The sample size also decreased due to the segmentation of the target audience into people with internet access and Facebook accounts. Therefore, the surveys conducted in this project represent a significant proportion of decision-making and trend analysis.

We followed the steps and recommendations required for adequate design and application of online surveys. We ensured that the survey instrument was as simple as possible to facilitate interviewees’ understanding. We appropriately selected individuals for the survey, preventing participants from generating multiple responses, using convenient data management, conducting pilot surveys, and addressing aspects of responsibilities and ethical standards [24,25,26,27].

It is relevant to point out that snook and red snapper are products of high commercial value which people with a specific purchasing power can afford. Therefore, it was assumed that the participating population would be able to access the internet. Moreover, different levels of education were believed to influence their use of platforms such as Facebook. Chinn and Fairlie, along with Močnik and Širec, revealed that the generation of infrastructure in information and communication technologies, skills with respect to managing online platforms, schooling, and income influence internet usage [28,29].

The survey we developed in this study comprised three sections: (a) socioeconomic variables, including age, family income, occupation, and gender, (b) product attributes, such as preferences for different types of presentations, type of establishment where the fish is purchased, the most critical factors considered when purchasing fish, and some aspects of preferences regarding fish consumption, and (c) a scenario that asks the central question corresponding to the binary dependent variable: “would you be willing to pay X amount of money (USD) for local artisanal aquaculture snook and red snapper?”. On the other hand, Mexican cooperatives propose fish production on a low scale, fattened in marine floating cages, with products distributed and marketed in establishments within Colima and Oaxaca. The local market prices for wild-caught snook and red snapper fillets per kilogram are $200 ($11.45 USD), $250 ($14.32 USD), $300 ($17.18 USD), and $350 ($20.04 USD). We proposed these same prices to the interviewees to measure the WTP for both species. We offered these values because farmed products need to be price-competitive on the market, as the literature indicates that wild-caught fish generally enjoys higher acceptance rates than farm-produced fish [13,30,31,32,33,34,35,36,37].

3. Results

We estimated the probit and logit discrete choice models based on the RUM from Equations (7)–(10) shown in the Materials and Methods section. We used these two econometric specifications for statistical inference purposes. On the one hand, the goodness-of-fit measure, as indicated by the pseudo-R², was slightly higher for the snook probit model (0.3974) and for the red snapper logit model (0.4199). This statistic indicates that the percentage of variation in WTP for the indicated species—39.74% for snook and 41.99% for red snapper—is explained by the model (Table 2). These values fall within the range suggested by studies analyzing WTP for aquaculture fish, typically ranging between 0.34 and 0.50 [11,12,13,14,15].

The overall significance tests of the regression, such as the likelihood ratio (LR) statistic, show that the probit model for snook (156.952) and the logit model for red snapper (160.450) had slightly higher values. The LRs were statistically significant at the 1% level. The values for the Akaike and Bayesian information criteria were lower in the probit model for snook (302.0341 and 420.7677, respectively) and in the logit model for red snapper (285.6529 and 404.3866, respectively), following the rule of selecting the econometric specification with the lowest quantities. These statistics suggest that the econometric analysis should focus on the versions that show superiority based on the pseudo-R² and the LR test and lower figures with respect to the information criteria. The results indicate that the percentage of correct predictions was 79.5% for the snook probit model and 84.4% for the red snapper logit model.

The predictors in probit and logit regression models do not provide direct interpretations; however, we can elucidate their meaning through the marginal effects. These effects allow us to examine the impact on the predicted probability of the dependent variable when an independent variable changes by one unit [38]. In the probit model for the snook, coefficients such as income2, income3, recipes, source, variety, age1, age3, age5, viscera, medallions, smoked, canned, beach, and unemployed were not statistically significant at conventional levels (1%, 5%, and 10%). Similarly, in the logit regression for the red snapper, marginal effects that lacked statistical significance included specie1, source, age2, age5, age6, viscera, without viscera, chunks, medallions, frozen, canned, and unemployed (

Table 3. Marginal effects of discrete choice models.

Variables	Snook				Red Snapper
	Probit Model		Logit Model		Probit Model		Logit Model
	Coefficient	Z-Statistic	Coefficient	Z-Statistic	Coefficient	Z-Statistic	Coefficient	Z-Statistic
pricesnook	−0.0024 **	−2.41	−0.0022 **	−2.11
priceredsnapper					−0.0020 **	−2.32	−0.0022 **	−2.51
income2	0.1453	1.35	0.1390	1.41	0.1784	1.61	0.1967 **	2.17
income3	0.1726	1.56	0.1730 *	1.70	0.1493	1.48	0.1634 **	1.99
income4	0.2452 ***	2.97	0.2421 ***	3.24	0.2863 ***	5.60	0.2648 ***	5.32
income5	0.3041 ***	4.39	0.2798 ***	3.70	0.3768 ***	8.94	0.3634 ***	7.71
income6	0.2345 **	2.05	0.2338 **	2.10	0.3087 ***	3.31	0.3270 ***	3.67
specie1	0.1397 *	1.91	0.1360 *	1.79	0.0614	1.00	0.0643	1.07
recipes	−0.0850	−1.06	−0.0887	−0.99	0.2017 ***	2.86	0.1932 **	2.48
promo	0.1567 **	2.28	0.1702 **	2.36	−0.1691 **	−2.13	−0.1688 *	−1.88
ambient	0.2084 ***	3.14	0.2133 ***	2.89	0.1272 **	1.97	0.1251 *	1.81
source	−0.0775	−1.13	−0.0671	−0.91	0.0607	0.86	0.0440	0.56
variety	−0.0473	−0.58	−0.0223	−0.24	−0.1826 **	−2.29	−0.1741 **	−2.01
age1	0.1030	0.98	0.0977	0.93	0.4034 ***	8.60	0.3962 ***	7.62
age2	0.1963 **	2.45	0.1792 **	2.16	0.0760	0.91	0.0696	0.87
age3	0.0487	0.48	0.0471	0.42	0.3435 ***	6.81	0.3335 ***	6.08
age5	0.0083	0.07	0.0040	0.04	−0.0467	−0.38	−0.0491	−0.40
age6	0.2651 ***	4.58	0.2503 ***	4.59	−0.1133	−0.57	−0.1062	−0.51
without viscera	0.2599 ***	5.04	0.2271 ***	3.89	−0.1829	−1.28	−0.2224	−1.47
viscera	0.0156	0.21	0.0117	0.14	0.1045 *	1.71	0.0823	1.27
chunks	0.2952 ***	6.09	0.2646 ***	4.84	0.1578 **	2.19	0.1236	1.59
skinless	0.1883 **	2.30	0.1639 *	1.69	0.3621 ***	5.49	0.3657 ***	5.00
skin	0.4113 ***	9.80	0.3878 ***	7.65	0.2684 ***	6.09	0.2381 ***	5.19
medallions	0.1026	1.21	0.1039	1.13	−0.0247	−0.23	−0.0211	−0.18
grounded	−0.2173 *	−1.65	−0.2129	−1.51	−0.4269 ***	−3.43	−0.4343 ***	−3.10
smoked	0.0258	0.35	0.0495	0.65	−0.3054 ***	−3.90	−0.2961 ***	−3.52
frozen	0.3403 ***	5.18	0.3388 ***	4.78	0.0556	0.87	0.0588	0.93
canned	−0.1007	−1.25	−0.1037	−1.12	−0.0195	−0.26	0.0031	0.04
beach	0.0150	0.19	0.0372	0.40	−0.1883 ***	−3.22	−0.1783 ***	−2.94
Self-employed	0.225 ***	3.57	0.2075 ***	3.31	−0.3011 ***	−2.64	−0.2971 **	−2.15
unemployed	0.0714	0.77	0.0720	0.76	−0.0355	−0.36	−0.0382	−0.36
freshness	0.3386 ***	2.95	0.3498 ***	2.65	0.3477 ***	3.27	0.3655 ***	2.99

Notes: * significant at 10%; ** significant at 5%; *** significant at 1%.

).

The price variable has a negative sign because it functions as a demand equation; this means that there is a lower willingness to pay as the offered payment increases. The probability associated with the willingness to pay (WTP) decreased by 0.24 for the snook and by 0.22 for the red snapper for each one-dollar increase in the proposed price. Regarding income, we omitted the segment of individuals earning less than or equal to $229.10 to highlight differences between income levels and avoid a dichotomous variable trap. The results show that individuals were more likely to pay for aquaculture snook and red snapper as the income level increased. The highest coefficients appeared for individuals earning between $919.32 USD and $1145.33 USD monthly (income5).

Individuals who expressed a desire for greater availability of local species experienced an increase of 13.97% in the probability associated with their WTP for snook. Those interested in recipes and fish preparation tips experienced an increase of 19.32% in the probability associated with their WTP for red snapper. Interviewees who wanted promotions and discounts for purchasing fish experienced an increase of 15.67% in their probability of consuming snook, although this effect was the opposite for red snapper. The group interested in sustainability and environmental impact experienced an increase of 20.84% and 12.51% in the probability of buying snook and red snapper, respectively.

Individuals attracted to various products in establishments experienced a decrease of 17.41% in their probability of preferring aquaculture red snapper. For the age variable, people aged 50–59 years served as a reference point due to being the largest segment. We found that those aged 30–39 years and over 69 years were more likely to prefer snook, while individuals aged 18–29 years and 40–49 years preferred red snapper, both produced through artisanal local aquaculture.

Interviewees preferring fresh-refrigerated whole fish without viscera, in chunks, in the form of a fillet without skin and fillet with skin experienced an increase of 25.99%, 29.52%, 18.83%, and 41.13%, respectively, in their WTP probabilities with respect to the snook. With regard to the red snapper, probabilities increased only in relation to the last two presentations by 36.57% and 23.81%, respectively. Conversely, individuals likely to purchase fresh-ice-ground-chopped and smoked products experienced a decrease in their WTP probability of 43.43% and 29.61%, respectively, for the red snapper; for the snook, WTP only decreased by 21.73% in relation to the fresh-ice-ground-chopped products. People demanding frozen fish were more likely to purchase the snook, experiencing an increase in their probability of 34.03%.

Consumers preferring to buy fish directly from fishermen on the beach had a lower WTP for red snapper, experiencing a decrease in their WTP probability of 17.83%. Self-employment positively affected WTP for the snook but not for the red snapper. Individuals preferring fresh fish experienced an increase of 33.86% in their WTP probability in relation to the snook and of 36.55% in relation to the red snapper.

Table 4. Estimated willingness to pay for aquaculture fish.

Estimation Source	Extent	Method	WTP (Price of Fillet per Kg)	Confidence Intervals at 95%
				Lower Limit	Upper Limit
Snook
Probit	Mean	Krinsky and Robb	$310.99 MXN ($17.81 USD)	$264.56 MXN ($15.15 USD)	$578.92 MXN ($33.15 USD)
Logit	Mean	Krinsky and Robb	$318.70 MXN ($18.25 USD)	$260.73 MXN ($14.93 USD)	$709.79 MXN ($40.65 USD)
Probit	Mean	López-Feldman routine	$310.99 MXN ($17.81 USD)	$235.98 MXN ($13.51 USD)	$386.00 MXN ($22.10 USD)
Logit	Mean	López-Feldman routine	$318.70 MXN ($18.25 USD)	$226.26 MXN ($12.96 USD)	$411.14 MXN ($23.54 USD)
Red snapper
Probit	Mean	Krinsky and Robb	$336.94 MXN ($19.30 USD)	$281.74 MXN ($16.13 USD)	$715.75 MXN ($40.99 USD)
Logit	Mean	Krinsky and Robb	$325.56 MXN ($18.64 USD)	$279.34 MXN ($16.00 USD)	$616.49 MXN ($35.30 USD)
Probit	Mean	López-Feldman routine	$336.94 MXN ($19.30 USD)	$245.81 MXN ($14.08 USD)	$428.06 MXN ($24.51 USD)
Logit	Mean	López-Feldman routine	$325.56 MXN ($18.64 USD)	$251.31 MXN ($14.39 USD)	$399.81 MXN ($22.90 USD)

presents the average willingness to pay (WTP) and the 95% confidence intervals (CIs) for artisanal aquaculture snooks and snappers, expressed as both Mexican pesos (MXN) and US dollars. Our calculation of the mean welfare measure indicates that the average consumer was willing to pay $17.81 USD for snook and $18.64 USD for red snapper fillets per kilogram. We constructed the lower and upper limits of the WTP intervals using the Krinsky–Robb method [39] and López-Feldman’s STATA routine [40].

4. Discussion

This research empirically analyzed consumption preferences for snook and red snapper using a welfare measure, specifically the willingness to pay (WTP) for fish grown by Mexican cooperatives through low-scale artisanal production. We demonstrate that socioeconomic variables and product attributes are related to the WTP for both species. The estimated models for each species indicate that consumers displayed a high level of acceptance with respect to their willingness to pay (WTP). To emphasize the significance of these findings, we organized the discussion into sections.

4.1. Factors Influencing Consumers’ Willingness to Pay

According to the economic theory of the demand curve, acceptance rates among individuals typically decline as prices increase, with this decrease being slightly more pronounced in relation to the snook than in relation to the red snapper (Table 3). Variables related to the understanding of local fish availability, the acquisition of discounts and promotions for snook products, and the access to recipes and preparation tips for the red snapper positively influenced the choice of aquaculture goods. However, for the red snapper, opportunities for discounts and promotions and a wider variety of fish from various origins showed negative associations with WTP.

These results align with findings by Rickertsen and colleagues, who noted that consumers who occasionally purchase fish and those who eat neither species as part of their daily diet may increase their acceptance and consumption habits when recipes are readily available [41]. Additionally, a preference emerged for native and local aquaculture products [9,30,31,36,42,43,44,45,46,47,48], confirming a choice for Mexican cooperative artisanal farms.

An important factor highlighted by international evidence is the impact of product discount rates on WTP. A common practice among producers entering the market is implementing marketing strategies involving discounts, as there tends to be a greater preference for wild-caught fish [34,49]. In this study, market prices for fish from commercial fishing were provided to consumers, suggesting that implementing a discount strategy could substantially affect acceptance levels.

4.2. Preference Differences Among Consumer Groups

Income, a primary determinant of demand, positively impacted the consumption of both marine products. Empirical evidence shows a positive relationship between income and the purchase of farmed fish species [9,11,12,14,15,45,50]. In Mexico, three studies focused on fish consumption, specifically tuna, have revealed that this species is considered a normal good among high- and middle-income strata. At the same time, it is perceived as a luxury good in the low-income segment. This finding aligns with our research, as higher-income individuals demanded commercially valuable fish, classifying them as luxury goods [51,52,53].

The highest marginal effects for the red snapper appeared among the youngest age group, confirming patterns in studies that show that older segments are less willing to accept aquaculture products [34,54,55,56,57,58]. Conversely, the results in relation to the snook demonstrate an opposite trend, consistent with findings on aquaculture fish [11,15,45,59,60,61]. This phenomenon indicates that the literature does not clearly define the relationship between age and consumption of farmed fish. However, it is crucial to emphasize that the future market demand for the red snapper will rely on the preferences of the younger age segment, as they will eventually replace the older population. With regard to the snook, the greater interest from the older age segment necessitates that producers and policymakers implement marketing strategies to promote consumption among younger individuals, thereby preventing a contraction of future demand.

Self-employment of individuals positively impacted snook consumption, whereas, on the red snapper, it had a negative effect. Few studies in the aquaculture fish consumption literature include individuals’ occupations. Hynes and colleagues demonstrated that employed individuals prefer paying a premium price for salmon raised in sustainable management conditions [62]. Alternatively, other studies have shown no significant differences with respect to this variable [34,49]. In this study, the self-employment condition was contrasted with that of employed individuals. In relation to the snook, self-employed individuals were found to prefer aquaculture products, while the opposite was true for the red snapper. This phenomenon can be attributed to the larger proportion of self-employed individuals in the younger age group, while employed individuals spanned the intermediate to older age groups. These results corroborate the analysis of age groups between the two species.

4.3. Considerations for Sustainability

Studies that offer consumers alternatives between labeled and certified fish and those that do not indicate that fish raised under certification seals and labels is preferred and tends to lead to increased willingness to pay (WTP). This finding is significant, as providing individuals with information about farm practices, whether positive or negative, can influence their decisions regarding acceptance of consumption [9,31,36,42,45,46,47,49,59,62,63,64,65,66,67,68,69,70,71,72].

In this context, individuals expressing a need to know about the sustainable farming practices behind a given product in relation to both types of fish indicate that the provision of this information—whether through certifications on seals, labels, or ecolabels addressing the environmental impact—can create an expectation that, in the medium and long term, consumer awareness will increase. Incorporating this measure into marketing policies will enable the development of a market that enhances producers’ profits and contributes to raising fish without generating negative environmental impacts.

4.4. Impact of Product Attributes on Preferences

Consumers who preferred fresh-refrigerated fish without viscera and in chunks (snook) and fillets with skin and without skin (both species) exhibited a higher willingness to pay for aquaculture products. Conversely, ground or chopped fish (both species) and smoked red snapper were associated with a lower willingness to pay. The highest positive marginal effects were associated with fillets with the skin for the snook and fillets without the skin for the red snapper, reflecting preferences similar to Asche’s findings regarding salmon fillets presented in fresh and frozen forms [73]. This result aligns with studies analyzing consumer preferences for salmon, where characteristics that provide an attractive color to fillets play a significant role in purchasing decisions [74,75,76].

Freshness significantly influences the consumption of snooks and red snappers, a finding that is consistent with studies indicating that consumers prefer fresh fish [9,13,14,30,33,34,48,69,77,78]; with regard to consumers residing in coastal communities along the Pacific Ocean in Mexico, purchasing fish in this presentation is expected. However, transporting the product to locations further from the beach complicates efforts to maintain fish freshness, as alternative preservation methods become necessary. This preference presents a substantial challenge for small-scale producers, who often lack the technology and adequate financial resources to invest in solutions that meet consumer demands. These producers typically sell their products directly to consumers, without intermediaries.

In contrast, interviewees willing to purchase red snappers on the beach were less inclined to choose aquaculture options. This finding correlates with results suggesting that hypermarket shoppers and supermarkets prefer artisanal aquaculture fish [34,49].

4.5. WTP Confidence Intervals

Confidence interval (CI) estimations indicate that the true population values fell within specific ranges. With regard to the snook, the estimates ranged from $15.15 USD to $33.15 USD when using the Krinsky–Robb method and from $13.51 USD to $22.10 USD when using López-Feldman’s routine. For the red snapper, the confidence intervals ranged from $16.00 USD to $35.30 USD when using the Krinsky–Robb method and from $14.39 USD to $22.90 USD when using López-Feldman’s procedure. The willingness to pay (WTP) estimations were significantly greater than zero at the 95% confidence level, leading to the rejection of the null hypothesis of zero WTP for both the snook and the red snapper. Additionally, López-Feldman’s approach provided narrower estimations compared to the broader ranges produced by the Krinsky–Robb method.

5. Conclusions

This study analyzes consumer preferences using the willingness to pay (WTP) for locally sourced artisanal aquaculture snooks and red snappers as a welfare measure. Specific product presentations, the type of store where consumers acquire the goods, and particular consumption preferences shape the WTP. The findings suggest that most of the population is inclined to accept these two species produced through aquaculture. The average WTP values are $17.81 USD for the snook and $18.64 USD for the red snapper, with confidence intervals (CIs) being wider with the Krinsky–Robb method and narrower with López-Feldman’s routine.

Since consumer values for both species are influenced by their attributes, cooperatives can use these prices to reference the prevailing market values when preparing cost–benefit analyses. Therefore, this article provides relevant information on key market trends to assist small-scale producers in making informed decisions regarding snook and red snapper production in floating cages.

This constitutes pioneering work because it targets small-scale, low-income producers with limited access to market information, credit, and technology. The identification of pilot artisanal aquaculture projects for high-commercial-value species in Oaxaca and Colima can be replicated in developing countries and in other regions of Mexico where conditions support these economic activities, ultimately contributing to the improvement of living standards in these areas. Producers can benefit from these initiatives by diversifying and supplementing their income sources, especially as low remuneration and increasing uncertainty in wild fish capture threaten traditional livelihoods.

Consumers will benefit from establishing local artisanal farms for snooks and red snappers, ensuring a constant supply of these products at prices comparable to wild-caught fish. This will enhance the circular flow of income through aquaculture activities. Future research should investigate the consumption preferences for aquaculture fish and seafood in developing countries, where job opportunities in the broader aquaculture industry may be limited and where many individuals depend on artisanal aquaculture and fishing for their livelihoods.

Mexican cooperatives and policymakers must recognize the increasing demand for marine fish consumption. They must, however, consider that commercial fishing, particularly the capture of wild marine fish, has reached its maximum sustainable yield. This situation renders aquaculture a viable alternative to address the growing demand driven by population increases.

This study focuses on marine fish of high commercial value, targeting consumers with medium-to-high incomes who possess the purchasing power for both snooks and red snappers. Furthermore, special attention should be directed toward younger age groups, as they tend to be more open to the consumption of aquaculture fish. Their preferences will likely shape future market trends in light of generational shifts.

For small-scale producers, this presents an opportunity to diversify their operations, thereby enhancing their economic benefits and maximizing profits while implementing improved practices for environmental sustainability in fish farming. In this context, strategies may include differentiating product offerings through ecolabels and origin certifications. Such initiatives could strengthen consumer trust concerning the social and environmental impacts of the products, potentially increasing their market value. This study identifies key areas for investment in the processing of both species (i.e., presentations) to meet diverse consumer preferences.