Patterns and Drivers of Sustainable Agricultural Practice Adoption Among Mennonite Farmers in a Closed Endorheic Basin of Northern Mexico

Abstract

Agricultural sustainability in closed endorheic basins is constrained by periodic water scarcity, hydrological isolation, and the cumulative effects of intensive land use. The Cuauhtémoc region of northern Mexico hosts long-established Mennonite farming communities whose mechanized and irrigation-dependent agriculture constitutes an important component of the regional economy and resource-use system. This study examined the prevalence, diversity, and drivers of sustainable agricultural practices among 95 Mennonite farmers in the Bustillos Lagoon region. Thirteen practices spanning soil and residue management, water conservation, biological soil-fertility management, and agrochemical stewardship were assessed, while a conservation-practice breadth indicator quantified the number of ten resource-management practices used by individual farmers. Among 92 farmers with complete practice data, conservation-practice breadth ranged from 2 to 8 practices (mean = 3.76 ± 1.06). Residue incorporation (98.9%), residue cover (89.4%), and rainwater collection (71.0%) were widespread, whereas zero tillage (1.1%), compost/leachate use (5.3%), mycorrhiza application (5.3%), and reduced tillage (8.5%) were uncommon. Robust Poisson regression showed that technical assistance was associated with a 37.8% higher expected conservation-practice count (IRR = 1.378, 95% CI 1.168-1.626, p < 0.001), while environmental concern was associated with an 18.7% higher count (IRR = 1.187, 95% CI 1.017-1.385, p = 0.030). Water-supply problems, farm area, producer-organization membership, and education showed no independent associations. The findings demonstrate that sustainable management is better characterized as a heterogeneous portfolio of practices than as a binary adoption decision and highlight institutional support and environmental orientation as important correlates of management diversification in water-constrained agricultural landscapes.

Introduction

Agriculture faces the dual challenge of maintaining food production while reducing pressures on soil, water, biodiversity, and other natural resources. This challenge is particularly acute in arid and semi-arid regions, where agricultural production frequently depends on limited water resources and climatic variability can intensify competition between productive and ecological water demands (FAO, 2024; IPCC, 2019, 2022; UNCCD, 2022; Muralikrishnan et al., 2021). Sustainable agriculture is understood here as farm management that sustains productive livelihoods while conserving soil and water resources and limiting avoidable environmental pressures over time. Within this broader concept, sustainable agricultural practices (SAPs) are observable management actions such as conservation tillage, crop rotation, residue management, water conservation, biological soil amendments, and responsible agrochemical management. Sustainable agricultural practices encompass a range of farm-management strategies intended to conserve natural resources while supporting agricultural production. Their adoption is rarely uniform across farms because decisions about individual practices are shaped by the interaction of farm resources, environmental conditions, knowledge, institutional support, farmer attitudes, and the characteristics and requirements of the practices themselves (Knowler and Bradshaw, 2007; Foguesatto et al., 2020; Kassie et al., 2013; Prokopy et al., 2019).

The heterogeneity of farmer responses has important implications for how sustainable agricultural adoption is conceptualized and measured. Binary classifications are useful when examining a single innovation but become restrictive when farmers simultaneously choose among multiple complementary or substitutable practices. Reviews of agricultural conservation adoption show that determinants vary substantially across practices and geographical settings and that socioeconomic, institutional, environmental, and behavioral factors operate together rather than independently (Foguesatto et al., 2020; Prokopy et al., 2019). A portfolio perspective can therefore provide a more informative representation of sustainability transitions than determining whether a farmer has adopted at least one environmentally oriented practice.

Access to knowledge and institutional support is especially important because many sustainable practices require technical skills, experimentation, or adaptation to local production conditions. Extension and technical-assistance programs can reduce informational barriers, strengthen farmers’ confidence in unfamiliar technologies, and facilitate continued adoption (Anderson and Elahi et al., 2021; Feder, 2004; Kassie et al., 2013; Mugari et al., 2025; Piñeiro et al., 2020). Farmer attitudes also influence whether available knowledge is translated into action. Behavioral research emphasizes that environmental concern, perceived benefits, norms, and other cognitive factors should be considered alongside structural constraints when explaining sustainable farming decisions (Kollmuss and Agyeman, 2002; Bwalya et al., 2023; Dessart et al., 2019).

These relationships become particularly important in water-limited agricultural systems. Dryland conditions create chronic constraints through limited and variable precipitation, whereas closed endorheic basins add a distinct hydrological characteristic: water and transported materials are retained internally rather than discharged through a permanent external outlet. Water scarcity can create incentives for conservation, but exposure to resource stress does not necessarily produce sustainable adaptation. Farmers’ responses depend on whether they possess the knowledge, infrastructure, financial resources, institutional connections, and feasible alternatives required to change established production practices (Muchara et al., 2014; Nath and Parameswar, 2025; Natarajan et al., 2022). Sustainability in this setting must therefore be understood both as a dryland adaptation challenge and as management within a basin where cumulative pressures can remain within the hydrological system.

Northern Mexico provides an important setting in which to examine these interactions. In this region, the state of Chihuahua has developed commercially important agriculture-based economies in arid and semi-arid areas with substantial dependence on groundwater. The valley of Cuauhtémoc, located on the central-western part of the state within the Bustillos Lagoon basin, is an illustrative example. This valley has developed into an intensive, groundwater-dependent agricultural landscape strongly associated with Mennonite settlement. Mennonite communities became established in Chihuahua during the early twentieth century and subsequently developed colony-based agricultural systems characterized by mechanization, irrigation, and commercial production (Dormady, 2014; Le Polain de Waroux et al., 2021). These communities are also distinguished by highly organized cooperative structures through which farmers share financial resources, agricultural inputs, machinery, infrastructure, and marketing networks, enabling coordinated decision-making and efficient agricultural production. As a result, farming practices and resource management decisions are often shaped not only at the individual farm level but also through collective institutions within colonies and cooperatives. Their contribution to the agricultural economy makes their management decisions relevant not only to farm productivity but also to regional land and water sustainability.

The environmental significance of these farming decisions is particularly apparent in the Bustillos Lagoon basin. The lagoon is an endorheic water body within the closed basins system of northern Chihuahua. Because these basins lack a permanent external surface outlet, water, sediments, nutrients, and contaminants transported from the surrounding landscape accumulate within the system (Ochoa-Rivero et al., 2017; Peña-Guerrero et al., 2020). Agricultural expansion and intensification, groundwater abstraction, land-use change, livestock production, and recurrent drought have increased concern about the preservation and long-term condition of the basin’s soil and water resources (Hruska, 2020; Rojas-Villalobos, 2019; Ochoa-Rivero et al., 2023; Valencia-Gaspar et al., 2024). In such a setting, agricultural sustainability concerns both farm productivity and the cumulative resource pressures generated across the basin.

Despite these pressures, relatively little farmer-level evidence examines the composition of sustainable agricultural management within the Cuauhtémoc-Bustillos closed-basin system. Existing scholarship provides valuable insight into Mennonite settlement, agricultural expansion and intensification, land-use change, and regional environmental pressure, but considerably less is known about the combinations of conservation-oriented practices used by individual farmers and the factors associated with maintaining broader practice portfolios. A key unresolved question is therefore how farmer-level management portfolios relate to institutional access, environmental orientation, and resource vulnerability within this hydrologically closed agricultural setting.

This gap also reflects a wider limitation in sustainable-agriculture research. Adoption studies have frequently examined socioeconomic characteristics and individual technologies, while the combined influence of environmental conditions, institutional support, and farmer perceptions has received less consistent attention (Knowler and Bradshaw, 2007; Foguesatto et al., 2020; Prokopy et al., 2019). At the same time, farmers commonly use different combinations of agricultural practices, reflecting variation in resource availability, production conditions, institutional access, and individual priorities. Understanding sustainable agricultural management therefore requires attention not only to individual practices but also to the diversity of management strategies employed within farming systems.

The present study addresses this question using primary survey data from Mennonite farmers in the Bustillos Lagoon region. It characterizes both the prevalence of individual practices and the breadth of sustainable management portfolios at the farmer level. Practices are considered according to their functional roles in soil and residue management, water management, biological soil-fertility management, and agrochemical stewardship. This portfolio-based approach reflects the multidimensional nature of sustainable agriculture and allows differences in management diversification to be examined without assuming that all practices provide equivalent environmental benefits.

The study is guided by the Sustainable Livelihoods Framework (SLF), which provides an integrated basis for examining how farmers’ resource endowments, institutional connections, human capacities, and vulnerability context shape livelihood strategies. Particular attention is given to technical assistance, environmental concerns, education, producer-organization membership, farm resources, and experienced irrigation-water constraints.

Specifically, the study seeks to: (1) characterize the prevalence, breadth, and functional composition of sustainable agricultural practices among Mennonite farmers in the Bustillos Lagoon region; (2) examine whether natural-resource constraints and farm resource endowments, particularly irrigation-water stress and irrigated farm area, are associated with the breadth of conservation-oriented agricultural practices; and (3) assess the relationships between human and institutional factors, particularly technical assistance, producer-organization membership, education, and environmental concern, and conservation-practice breadth.

The study contributes in three respects. First, this study provides both data and knowledge on agricultural practices and the complex web of socio-cultural, technological, and natural resource constraints that shape those practices within a commercially important yet comparatively understudied Mennonite agricultural system in northern Mexico. Second, it moves beyond binary adoption by distinguishing the breadth and composition of farmers’ management portfolios. Third, by examining institutional capacity, environmental orientation, farm resources, and water vulnerability within a common livelihood’s framework, it provides evidence relevant to designing context-specific agricultural sustainability interventions in Bustillos Lagoon and comparable water-constrained endorheic basins.

Theoretical Framework: Sustainable Livelihoods Framework

The Sustainable Livelihoods Framework provides the conceptual foundation for this study. Sustainable livelihood approaches conceptualize rural decision-making as the outcome of interactions among resources available to households, the vulnerability context in which livelihoods are pursued, and the institutional structures and processes that influence access to resources and opportunities (Scoones, 1998; DFID, 1999). Rather than explaining agricultural decisions through a single economic, demographic, or environmental factor, the framework recognizes that farmers combine different resources to pursue livelihood strategies under changing environmental and institutional conditions.

A central feature of the SLF is the livelihood asset structure, commonly represented through natural, human, physical, financial, and social capital. These capitals are not assumed to operate independently; rather, their configuration shapes the options available to households, while institutions can facilitate or constrain the conversion of assets into livelihood strategies (DFID, 1999; Scoones, 1998; Li et al., 2020; Natarajan et al., 2022). In agricultural settings, this means that access to natural resources or productive assets may be insufficient when farmers lack complementary knowledge, finance, infrastructure, or institutional support (Muchara et al., 2014).

Natural capital refers to the natural-resource base upon which livelihoods depend, including land, soil, and water. In the Bustillos system, farmer-reported soil conditions and irrigation-water availability characterize this dimension. Because soil properties in the survey were based on farmer assessment rather than standardized laboratory measurements and some categories had limited variation, they were used primarily to describe the natural-resource setting rather than as principal inferential predictors.

Human capital encompasses knowledge, skills, experience, and capacities that enable farmers to make and implement management decisions. The questionnaire captures this dimension through education, farming experience, and environmental concern. Prior research links education and cognitive capacity to agricultural innovation, while also showing that context-specific technical knowledge may be more consequential than formal schooling alone (Asfaw and Admassie, 2004; Elahi et al., 2021; Li et al., 2020). Environmental concern is interpreted narrowly as the respondent’s stated importance of caring for the environment rather than as a multidimensional psychometric measure of environmental awareness.

Physical capital concerns productive infrastructure and material resources. Irrigated farm area, irrigation systems, wells, and agricultural machinery provide indicators of the productive resources available to farmers. Financial capital encompasses economic resources such as agricultural income, credit, and production support. Previous applications of the SLF emphasize that physical and financial assets can condition whether farmers are technically and economically able to implement resource-conserving practices (Muchara et al., 2014; Zulfiqar and Thapa, 2017). These variables help characterize the livelihood system, although variables with insufficient variation were not forced into the multivariable model solely to represent every SLF capital.

Social capital refers to networks, relationships, and collective organization through which farmers obtain information and support. Producer-organization membership provides one indicator of this dimension. Technical assistance is treated as an institutional linkage because it represents farmers’ access to external agricultural knowledge and advisory support.

Beyond assets, the SLF emphasizes the vulnerability context: stresses, shocks, and trends over which livelihood actors may have limited direct control but to which they must respond. For farming households around Bustillos Lagoon, recurrent water constraints and dependence on groundwater are especially relevant. Dryland livelihood research shows that environmental stress can alter management priorities without necessarily producing a uniform shift toward conservation, particularly where adaptive capacity is uneven (Nath and Parameswar, 2025; Natarajan et al., 2022). Reported irrigation-water shortages are therefore conceptualized as an experienced manifestation of resource vulnerability rather than automatically as a determinant of sustainable behavior.

The framework also identifies transforming structures and processes, including government agencies, extension systems, producer organizations, policies, and markets that influence how assets can be accessed and deployed. In this study, technical assistance and organizational membership represent observable points of interaction between farmers and this wider institutional environment.

Within this framework, sustainable agricultural practices are conceptualized as livelihood-management strategies rather than as a single behavioral state. Farmers may combine crop rotation, residue management, water conservation, biological inputs, and agrochemical stewardship in different ways. Practice breadth therefore represents the diversity of conservation-oriented management responses incorporated into the farming system; it is not interpreted as a direct ecological-performance score because individual practices differ in purpose, implementation quality, and environmental effect.

The SLF therefore serves as the organizing framework linking livelihood assets, institutional conditions, resource vulnerability, and agricultural management strategies (Fig. 1). Natural capital is represented mainly by the water-resource context and farmer-reported soil information; human capital by education, farming experience, and environmental concern; physical capital by irrigated farm area and irrigation infrastructure; financial capital by income, credit, and production-support information collected in the survey; and social/institutional capital by producer-organization membership and technical assistance. The three study objectives focus, respectively, on the livelihood strategy itself (the prevalence and breadth of SAP portfolios), the resource and vulnerability context (water-supply problems and farm scale), and human and institutional conditions (education, environmental concern, organization membership, and technical assistance). Financial-capital variables contribute to characterization of the livelihood system, while the multivariable model emphasizes variables with sufficient measurement quality, completeness, and variation for stable inference.

Fig. 1.

Application of the Sustainable Livelihoods Framework to the analysis of sustainable agricultural practice portfolios among Mennonite farmers in the Bustillos Lagoon region

Methodology

Research Design

The study employed a quantitative cross-sectional survey design, collecting farmer-level information at a single period to examine patterns and statistical associations among livelihood characteristics, environmental conditions, institutional support, and agricultural management practices (Creswell and Creswell, 2018). The design was appropriate because the study sought to characterize current management portfolios and identify factors associated with their breadth rather than establish causal effects. Cross-sectional survey data also enabled multiple dimensions of the SLF to be operationalized within a common analytical framework.

Study Area

The empirical study focused on the Mennonite agricultural zone from which participating farmers were sampled, situated within the wider 3288 km² Bustillos Lagoon watershed. The wider basin provides the hydrological context because groundwater use, runoff, and land-management decisions in the agricultural zone occur within the same endorheic system.

The Bustillos Lagoon basin is located in Cuauhtémoc, Chihuahua, northern Mexico (Fig. 2). The lagoon is situated at approximately 28°33′32″ N and 106°45′57″ W at an elevation of about 1975.5 m above sea level. The wider watershed covers approximately 3288 km² and forms part of Mexico’s Hydrological Region No. 34, the Closed Basins of the North. The basin lies about 90 km west of Chihuahua City (Ochoa-Rivero et al., 2017).

Fig. 2.

A Regional location of the Bustillos Lagoon Basin in Chihuahua, Mexico; B Bustillos Lagoon and the Mennonite agricultural valley

Bustillos Lagoon receives water from several perennial and intermittent streams and has no permanent natural surface outlet. Its endorheic character favors retention of materials transported from the contributing landscape. Previous research has documented pollutant occurrence in the lagoon and substantial land-use/land-cover change across the basin, highlighting the sensitivity of the system to cumulative agricultural and hydrological pressures (Ochoa-Rivero et al., 2017; Valencia-Gaspar et al., 2024).

The regional economy is strongly linked to agriculture, livestock production, and dairy farming. Cropping systems include maize, wheat, barley, oats, forage crops, apples, and vegetables, and agricultural production depends substantially on irrigation and groundwater resources. The regional climate is arid to semi-arid and characterized by marked interannual rainfall variability and recurrent drought, making irrigation central to production and exposing farmers to episodic water-supply constraints (Hruska, 2020; Muralikrishnan et al., 2021; Valencia-Gaspar et al., 2024). These climatic and hydrological conditions make the area appropriate for examining sustainable management within a water-constrained farming landscape.

Mennonite settlement has been an important component of agricultural development in the Cuauhtémoc region since the early twentieth century. Mennonite communities established colony-based farming settlements in Chihuahua and developed highly organized commercial agricultural systems characterized by mechanization, irrigation, and an organized credit financial support system known as FORTSCHRITT (Dormady, 2014; Le Polain de Waroux et al., 2021). The FORTSCHRITT sampling frame listed 839 active Mennonite farmers in the target agricultural zone. Surveyed farms varied substantially in operational scale, as described in the Results.

Study Population and Sampling

The target population comprised active Mennonite agricultural producers, hereafter referred to as farmers, managing agricultural plots in the target zone surrounding Bustillos Lagoon. The FORTSCHRITT registry provided the sampling frame and listed 839 eligible farmers at the time of the study.

The required sample size was estimated for a finite population under simple random sampling using the standard finite-population formula for a proportion (Téllez-Delgado et al., 2012):

$$n=\frac{N{Z}^{2}p\left(1-p\right)}{e^2\left(N-1\right)+Z^{2}p\left(1-p\right)}$$ 1

where n is the required sample size, N is the population size, Z is the standard-normal critical value for the selected confidence level, p is the assumed population proportion, and e is the allowable sampling error. With N = 839, Z = 1.96, p = 0.50, and e = 0.10, the calculation yields n ≈ 86.3. The target was increased to 95 farmers to improve precision and allow for incomplete responses.

Eligibility was based on active farming status and was independent of current use of sustainable agricultural practices. Because practice prevalence and practice breadth were study outcomes, SAP use was measured after participant selection. This design allowed the sample to represent the full range of management portfolios among eligible farmers and supported unbiased estimation of practice prevalence and farmer-level associations.

Individual farmers were selected from the registry using simple random sampling. Each eligible producer was assigned a unique identifier, and farmers were selected without replacement until the target sample size was reached, thereby ensuring that each producer in the sampling frame had an equal probability of selection. Following data compilation, the analytical dataset was screened for duplicate records. One exact duplicate identified in the source worksheet was removed, leaving 95 unique farmer records for analysis.

Data Collection Instrument and Variables

Data were collected through a structured questionnaire administered in person to the selected farmers. In-person administration enabled clarification of questions and immediate checking of incomplete or inconsistent responses, thereby supporting data completeness and consistency (Oppenheim, 2000; Bryman, 2016; Creswell and Creswell, 2018). The source questionnaire was administered and recorded in Spanish; variable labels and responses were subsequently translated into English for analysis while preserving the original questions, response categories, and coding structure. The English version of the questionnaire used for analysis is provided as Supplementary Material S1.

The questionnaire collected information on respondent demographics and farming experience; cultivated and irrigated farm characteristics; perceived soil and water conditions; irrigation infrastructure, machinery, and other productive resources; access to credit and agricultural support; producer-organization membership; technical assistance and training; environmental perceptions; and agricultural management practices. These variables were subsequently aligned with the Sustainable Livelihoods Framework (SLF), with education and farming experience representing human capital; soil and water conditions representing natural capital; irrigated area, irrigation systems, wells, and machinery representing physical capital; credit, agricultural income, and production support representing financial capital; and producer-organization membership and technical assistance representing social and institutional linkages. Variables were retained for inferential analysis only where their measurement quality, completeness, and variation were adequate.

The 13 agricultural management practices analyzed represented the complete set of practices included in the relevant section of the source questionnaire; no practices were selected or excluded on the basis of observed adoption patterns. Farmers reported the frequency with which they used reduced tillage, zero tillage, crop rotation, crop-residue cover, residue incorporation, rainwater collection, manure application, compost/leachate use, mycorrhiza application, biofertilizer use, permitted insecticide use, rational pesticide/herbicide use, and appropriate agrochemical-container management. These items define the practice domain captured by the original survey instrument and form the empirical basis for the practice-breadth measures used in the analysis.

For analytical interpretation, the 13 practices were organized according to their principal management function into four domains: soil and residue management (reduced tillage, zero tillage, crop rotation, residue cover, and residue incorporation); water management (rainwater collection); biological soil-fertility management (manure, compost/leachate, mycorrhiza, and biofertilizers); and agrochemical stewardship (permitted insecticides, rational pesticide/herbicide use, and appropriate agrochemical-container management). This functional classification was informed by the broader literature on conservation and sustainable agricultural practices (Knowler and Bradshaw, 2007; Foguesatto et al., 2020; Kassie et al., 2013) but was used primarily as an analytical grouping of the practices contained in the questionnaire. It does not imply that the practices are equivalent in cost, technical requirements, environmental effectiveness, or sustainability benefits.

Each practice retained the questionnaire’s original ordinal response categories of Always, Sometimes, and Never. For descriptive summaries, these responses were coded 2, 1, and 0, respectively, to preserve the reported frequency ordering. For the practice-breadth outcomes described in Section 3.5, the same responses were transformed into a binary indicator of current use, with Always or Sometimes = 1 and Never = 0. The ordinal coding therefore describes frequency of use, whereas the binary coding identifies whether a practice formed part of a farmer’s management portfolio.

Farmer-reported soil texture and soil fertility were also collected through the questionnaire. Because the source instrument did not document laboratory verification or the use of a standardized field texture-assessment protocol, these measures were treated as farmer-perceived soil characteristics rather than objective biophysical measurements. They were therefore used to characterize the natural-resource context but were not included as principal predictors in the multivariable model.

Operationalization of Sustainable Agricultural Practices

Farmer-level sustainable management was operationalized through the breadth of the agricultural practice portfolio recorded in the questionnaire. This measure captures variation in the number and combination of practices used by individual farmers and provides a direct indicator of management diversification across the surveyed population.

The original Always/Sometimes/Never responses were retained separately for descriptive reporting. For the primary breadth analysis only, each practice was then recoded to a binary use indicator: Always or Sometimes = 1 (practice used) and Never = 0 (practice not used). Thus, the binary recoding applies to each individual practice before summation; it does not classify the farmer as an overall adopter or non-adopter.

$$U_{ij}=\left\{\begin{array}{c}\hfill 1,\mathrm{ifresponse}\in \left\{\mathrm{Always},\mathrm{Sometimes}\right\}\hfill \\ \hfill 0,\mathrm{ifresponse}=\mathrm{Never}\hfill \end{array}\right)$$ 2

The conservation-practice breadth score for farmer i was then defined as:

$${CPB}_i={\sum }_{j=1}^{10}{U}_{ij},{CPB}_i\in \left\{0,1,\dots ,10\right\}$$ 3

where CPBᵢ is the number of the ten resource-management practices used by farmer i. The three agrochemical-stewardship items were excluded from the primary breadth score because they represent input-stewardship behavior and were nearly universal in the sample.

$${FPB}_i={\sum }_{j=1}^{13}{U}_{ij},{FPB}_i\in \left\{0,1,\dots ,13\right\}$$ 4

FPBᵢ denotes the full 13-practice breadth score used in the sensitivity analysis.

For descriptive analysis, the three original response categories were reported directly as frequencies rather than combined into an ordinal score. The practices were organized into four functional domains: soil and residue management (reduced tillage, zero tillage, crop rotation, residue cover, residue incorporation); water management (rainwater collection); biological soil-fertility management (manure, compost/leachate, mycorrhiza, biofertilizers); and agrochemical stewardship (permitted insecticides, rational pesticide/herbicide use, and agrochemical-container management). The domains describe management functions and do not imply equivalent ecological effects.

The primary outcome was conservation-practice breadth (CPB), the count of the first ten resource-management practices reported as used Always or Sometimes. Its theoretical range was 0-10. The three agrochemical-stewardship items were excluded from CPB because they represent input-stewardship behavior and showed near-universal use, providing little discriminatory information for the primary count. A full practice-breadth score (FPB; 0-13) including all 13 practices was estimated as a sensitivity outcome. Both measures represent management breadth, not ecological performance: a larger count denotes a more diverse practice portfolio but does not assume equal environmental benefit or equal implementation intensity across practices.

Explanatory Variables

Selection of explanatory variables was guided by the SLF and by the measurement quality and variation of variables captured in the questionnaire. Human and institutional dimensions were represented by educational attainment, technical assistance, producer-organization membership, and environmental concern. Environmental concern was derived from the questionnaire item, which asked respondents whether caring for the environment was important to them. It was therefore interpreted as a single-item indicator of pro-environmental orientation rather than as a multidimensional environmental-awareness scale. The vulnerability context was represented by reported irrigation-water supply problems during the preceding 20 years. Physical farm resources were represented by irrigated farm area, measured in hectares and natural-log transformed for multivariable analysis because of its positive skew. Education was consolidated into secondary-or-higher versus none/primary education because relatively few respondents had attained education beyond the secondary level. Farmer-reported soil texture and perceived fertility were retained for descriptive purposes but were not included as principal regression predictors because they were not independently verified and some categories exhibited limited variation.

Because irrigated farm area was positively skewed, it entered the multivariable model on the natural-log scale:

$${A_i}^{*}=ln\left(A_i\right)$$ 5

where Aᵢ is irrigated farm area in hectares for farmer i.

Statistical Analysis

Data were screened for duplicate records, missing values, inconsistent coding, and implausible entries and were analyzed using IBM SPSS Statistics version 27. Descriptive statistics characterized the farming population and practice prevalence. Frequencies and percentages were used for categorical variables, while means, standard deviations, medians, and ranges were reported for continuous variables as appropriate. For objective-specific bivariate comparisons, Mann-Whitney U tests were used for two-group comparisons of practice breadth, the Kruskal-Wallis test was used for education categories, and Spearman’s rank correlation was used for irrigated farm area. These distribution-free procedures avoided imposing a normality assumption on the bounded count outcome.

Because the principal dependent variable is a non-negative count of conservation practices, a generalized linear model with a Poisson distribution and log link was used to estimate adjusted associations with conservation-practice breadth, a standard framework for count outcomes (Cameron and Trivedi, 1990). The model was specified as:

$$Y_i~Poisson\left({\mu }_i\right)$$ 6

with the conditional mean linked to the predictors through:

$$\ln \left({\mu }_i\right)={\beta }_0+{\beta }_1{TA}_i+{\beta }_2{EC}_i+{\beta }_3{W}_i+{\beta }_4{PO}_i+{\beta }_5{EDU}_i+{\beta }_{6}ln\left(A_i\right)$$ 7

where Yᵢ is conservation-practice breadth; TA is technical assistance; EC is environmental concern; WP is reported irrigation-water supply problems; PO is producer-organization membership; EDU is secondary-or-higher education; and A is irrigated farm area.

Exponentiated coefficients were reported as incidence-rate ratios (IRRs), defined as:

$${IRR}_k=exp\left({\beta }_k\right)$$ 8

An IRR greater than 1 indicates a higher expected practice count, whereas an IRR below 1 indicates a lower expected count, holding the other variables constant.

The Poisson mean-variance assumption was evaluated using the Pearson dispersion statistic (Cameron and Trivedi, 1990):

$$\widehat{\phi }=\frac{1}{n-p}{\sum }_{i=1}^n\frac{{\left(Y_i-{\mu }_{\widehat{i}}\right)}^2}{{\mu }_{\widehat{i}}}$$ 9

where n is the analytical sample size and p is the number of estimated model parameters. Values near 1 indicate approximate equidispersion, values greater than 1 indicate overdispersion, and values below 1 indicate underdispersion. The observed statistic was below 1, so robust sandwich standard errors were used for inference. Statistical significance was assessed using two-sided p < 0.05, results are presented with 95% confidence intervals complete-case analysis was used for the multivariable model, and the model was repeated using the 13-practice breadth score as a sensitivity analysis.

Results

Characteristics of the Farming Population and Livelihood Context

The demographic and livelihood characteristics of the surveyed Mennonite farmers are presented in Table 1. The sample was dominated by middle-aged farmers with substantial farming experience, while formal educational attainment was generally concentrated at the elementary level. Reported gender information also indicates that the surveyed agricultural producers were predominantly male.

Table 1.

Demographic, farm-resource, institutional, and environmental characteristics of the 95 surveyed Mennonite farmers in the Bustillos Lagoon agricultural zone

Characteristic	Category/summary	n	% or descriptive statistic
Age (years)	20–39	25	26.3%
	40–59	53	55.8%
	60–79	17	17.9%
	Mean ± SD	95	48.73 ± 12.41
	Range	95	20–76
Gender	Male	83	87.4%
	Female	0	0.0%
	No response	12	12.6%
Education level	No formal education	11	11.6%
	Primary/elementary	72	75.8%
	Secondary	10	10.5%
	High school	1	1.1%
	University	0	0.0%
	No response	1	1.1%
Farming experience (years)	Mean ± SD	95	25.4 ± 13.3
Irrigated farm area (ha)	Mean ± SD	92	123.5 ± 102.6
	Median	92	93.5
	Range	92	6.1–600
Technical assistance	Yes	11/92	12.0%
Producer-organization membership	Yes	57/91	62.6%
Environmental concern	Care for the environment considered important	66/86	76.7%
Irrigation-water supply problems	Reported problems during previous 20 years	30/89	33.7%

Denominators vary because of item non-response. n is number of respondents

The farming population was heterogeneous in operational scale, with considerable variation in irrigated farm area. Institutional connections also differed across respondents. Membership in producer organizations was relatively common, whereas access to agricultural technical assistance was limited. At the same time, environmental concern was widely expressed, although a notable proportion of farmers reported experiencing irrigation-water supply problems. These characteristics provide the demographic, productive, institutional, and environmental context for examining differences in the composition and breadth of sustainable agricultural practices.

Objective 1: Prevalence, Functional Composition, and Breadth of Sustainable Agricultural Practices

As shown in Table 2, adoption varied markedly across the 13 practices. Residue incorporation and residue cover were among the most widely used soil-management practices, whereas reduced and zero tillage were uncommon. Rainwater collection was also relatively widespread, while biological soil-fertility practices, particularly compost/leachate and mycorrhiza use, showed limited uptake. Agrochemical-stewardship practices were consistently reported across the sample.

Table 2.

Frequency of use of 13 sustainable agricultural management practices among surveyed Mennonite farmers

Practice	Always	Sometimes	Never	Valid n	Used ≥ sometimes (%)
Reduced tillage	3	5	86	94	8.5
Zero tillage	0	1	93	94	1.1
Crop rotation	31	19	44	94	53.2
Residue cover	24	60	10	94	89.4
Residue incorporation	76	17	1	94	98.9
Rainwater collection	62	4	27	93	71.0
Manure application	9	20	65	94	30.9
Compost/leachate	1	4	89	94	5.3
Mycorrhiza	4	1	89	94	5.3
Biofertilizers	9	6	78	93	16.1
Permitted insecticides	12	82	0	94	100.0
Rational pesticide/herbicide use	5	87	2	94	97.9
Agrochemical-container management	5	88	1	94	98.9

Valid n is the total number of respondents who answered that question

The functional-domain analysis further showed that sustainable management was uneven across domains (Table 3). Soil and residue management was moderately diversified, and water management was represented by substantial uptake of rainwater collection. In contrast, biological soil-fertility management was the least diversified domain, with most farmers using none or only one of the practices assessed. Agrochemical stewardship showed the highest and most uniform uptake.

Table 3.

Functional composition of sustainable agricultural practice portfolios, showing the number of practices used within soil/residue, water, biological soil-fertility, and agrochemical-stewardship domains

Functional domain	Valid n	Mean ± SD	Possible range	Dominant pattern
Soil/residue management	94	2.51 ± 0.63	0–5	52.1% used 2; 42.6% used 3
Water management	93	0.71 ± 0.46	0–1	71.0% used rainwater collection
Biological soil-fertility management	93	0.56 ± 0.77	0–4	57.0% used none; 33.3% used 1
Agrochemical stewardship	94	2.97 ± 0.23	0–3	97.9% used all 3

SD is standard deviation

The distribution of conservation-practice breadth was concentrated in the middle of the observed range (Table 4). Most farmers used a moderate rather than very narrow or very broad portfolio of resource-management practices, while only a small number reported more extensive diversification. The full 13-practice sensitivity measure showed a similarly compact distribution.

Table 4.

Distribution of the primary conservation-practice breadth score (0–10 theoretical range) among farmers with complete information for the ten resource-management practices (n = 92)

Conservation-practice breadth	n	%
2	9	9.8
3	29	31.5
4	34	37.0
5	18	19.6
7	1	1.1
8	1	1.1

Objective 2: Resource Conditions, Farm Endowments, and Conservation-Practice Breadth

Resource conditions and farm endowments were examined to determine whether differences in exposure to irrigation-water constraints and operational farm scale were associated with the breadth of conservation practices used by farmers. These variables represent two distinct aspects of the farming context: reported irrigation-water problems capture an experienced resource constraint, whereas irrigated farm area reflects variation in the scale of productive resources available to individual farmers.

At the bivariate level, there was little evidence that either factor differentiated conservation-practice breadth (Table 5). Farmers who reported experiencing irrigation-water supply problems had practice portfolios broadly comparable to those who did not report such problems. Similarly, the association between irrigated farm area and conservation-practice breadth was weak, indicating that larger irrigated holdings were not systematically characterized by either broader or narrower conservation-practice portfolios. Thus, neither exposure to water-supply difficulties nor operational scale alone appeared to distinguish the extent to which farmers combined the conservation practices assessed in this study.

Table 5.

Bivariate and adjusted associations of irrigation-water supply problems and irrigated farm area with conservation-practice breadth, including the 13-practice sensitivity estimates

Resource factor	Unadjusted evidence	Adjusted IRR (95% CI)	p-value	13-practice IRR	p-value
Irrigation-water problems	Mean 3.79 vs 3.70; MW p = 0.749	1.019 (0.910-1.142)	0.745	1.007	0.827
Irrigated farm area	Spearman ρ = 0.058; p = 0.586	0.990 (0.892-1.100)	0.855	1.002	0.928

MW Mann–Whitney U test, IRR incidence-rate ratio, CI confidence interval

Adjusted estimates were obtained from the robust Poisson model. Irrigated farm area was natural-log transformed for the multivariable analysis.

The multivariable analysis reinforced these patterns after accounting simultaneously for technical assistance, environmental concern, producer-organization membership, education, and the other resource-related factor (Table 5). Reported irrigation-water problems did not independently predict conservation-practice breadth, suggesting that farmers experiencing this form of resource pressure did not consistently maintain a different number of conservation practices once other measured characteristics were considered. Irrigated farm area likewise showed no independent association with the expected practice count. The absence of an association across the substantial variation in farm area observed in the sample indicates that the breadth of conservation management was not simply a function of operational scale.

The resource-related findings were stable when the outcome was broadened from the primary 10-practice conservation measure to the full 13-practice measure (Table 5). Neither irrigation-water problems nor irrigated farm area became statistically associated with practice breadth in the sensitivity analysis. The consistency across outcome specifications indicates that these resource-related variables had limited explanatory power for differences in sustainable-practice breadth within the surveyed population.

Objective 3: Human and Institutional Conditions and Conservation-Practice Breadth

Human and institutional conditions showed different relationships with the breadth of conservation practices used by farmers. At the bivariate level, farmers who received technical assistance tended to maintain broader conservation-practice portfolios than those without such support. A similar, although less pronounced, pattern was observed for environmental concern, with farmers who considered environmental care important tending to report somewhat broader practice portfolios. Neither difference was statistically detectable before adjustment, indicating that these simple group comparisons alone provided limited evidence of an association. Producer-organization membership showed little differentiation in conservation-practice breadth, while practice portfolios also varied only modestly across educational categories.

In the adjusted model, technical assistance showed the strongest positive association with conservation-practice breadth (Table 6). This association was estimated after accounting for environmental concern, water-supply problems, producer-organization membership, education, and irrigated farm area. Because relatively few farmers with complete outcome data reported receiving technical assistance, the estimate is presented with its confidence interval and interpreted cautiously.

Table 6.

Adjusted robust Poisson regression of the ten-practice conservation-breadth count among complete cases (N = 71)

Predictor	Adjusted IRR	95% CI	p-value
Technical assistance: Yes vs No	1.378	1.168-1.626	<0.001
Environmental concern: Yes vs No	1.187	1.017-1.385	0.030
Irrigation-water problem: Yes vs No	1.019	0.910-1.142	0.745
Producer-organization membership: Yes vs No	1.049	0.925-1.188	0.456
Secondary-or-higher education	1.153	0.958-1.387	0.132
Log irrigated farm area (ha)	0.990	0.892-1.100	0.855

IRR incidence-rate ratio, CI confidence interval.

Environmental concern also showed a positive independent association with conservation-practice breadth in the adjusted analysis (Table 6). The variable refers specifically to the single questionnaire item on the importance of environmental care.

Producer-organization membership and secondary-or-higher education did not show statistically detectable independent associations with conservation-practice breadth after adjustment (Table 6).

The 13-practice sensitivity analysis produced the same substantive pattern: the positive associations for technical assistance and environmental concern persisted, whereas the conclusions for producer-organization membership and education were unchanged.

Integrated Multivariable Model and Robustness Analysis

The integrated model jointly examined the resource, human, and institutional factors identified within the study framework, allowing their associations with conservation-practice breadth to be assessed while accounting for the other measured characteristics. The complete-case robust Poisson model included 71 farmers and is presented in Table 6. After adjustment, technical assistance emerged as the strongest positive correlate of conservation-practice breadth, with environmental concern also showing a positive independent association. This pattern suggests that access to agricultural knowledge and farmers’ environmental orientation differentiated the breadth of conservation practices more clearly than the resource and structural characteristics included in the model.

Irrigation-water problems, producer-organization membership, education, and irrigated farm area did not show statistically detectable independent associations after simultaneous adjustment. Their confidence intervals included the null value, indicating that the available data did not distinguish their adjusted associations from zero within this sample. The estimates nevertheless provide information on direction and uncertainty and are interpreted alongside the broader livelihood context.

The model showed underdispersion relative to the conventional Poisson mean-variance assumption. Robust sandwich standard errors were therefore used to provide inference less dependent on the assumed variance structure. Given the complete-case sample size, interpretation was based on the direction, magnitude, confidence intervals, and consistency of the estimated associations rather than statistical significance alone.

The sensitivity analysis confirmed the stability of the primary model across alternative outcome specifications. When the outcome was expanded to include all 13 practices, technical assistance and environmental concern remained positively associated with practice breadth, while the conclusions for water-supply problems, producer-organization membership, education, and irrigated farm area were unchanged.

Discussion

Objective 1: Sustainable Agricultural Practice Portfolios among Mennonite Farmers

The findings demonstrate that sustainable agricultural management among Mennonite farmers is better understood as a portfolio of practices rather than as a single binary adopter/non-adopter classification. Farmers combined practices selectively, with stronger uptake of residue management, rainwater collection, and agrochemical stewardship but comparatively limited use of conservation tillage and biological soil-fertility practices. This uneven composition indicates that farmers occupy different positions along a continuum of management diversification. Such heterogeneity is consistent with evidence that agricultural practices are adopted selectively according to their compatibility with existing production systems, expected benefits, costs, technical requirements, and local constraints (Knowler and Bradshaw, 2007; Foguesatto et al., 2020; Prokopy et al., 2019).

The concentration of practices in particular management domains is especially important. Practices compatible with established production routines appear more embedded in the farming system, whereas biological inputs and conservation tillage remain comparatively marginal. Although the survey did not directly establish why individual practices were accepted or rejected, the pattern suggests that expanding sustainable management requires attention to which practices remain weakly integrated and why, rather than simply encouraging farmers to “adopt sustainable agriculture.” Importantly, practice breadth represents management diversification and should not be interpreted as a direct measure of ecological performance because the individual practices differ in their functions and environmental effects.

Objective 2: Resource Conditions and Farm Endowments

The absence of a detectable association between irrigation-water problems and conservation-practice breadth suggests that exposure to environmental stress does not automatically translate into broader conservation behavior. This does not diminish the importance of water scarcity in the Bustillos region. Instead, it supports the SLF proposition that vulnerability establishes the context within which livelihood decisions occur, while actual responses depend on the assets, knowledge, institutions, and feasible alternatives available to farmers (Scoones, 1998; DFID, 1999; Nath and Parameswar, 2025; Natarajan et al., 2022). Farmers experiencing similar resource pressures may therefore pursue different adaptation strategies, some of which may fall outside the practices measured in this study.

This distinction is particularly relevant to groundwater-dependent agriculture in the Bustillos Lagoon basin. Irrigation difficulties could stimulate responses involving pumping arrangements, infrastructure, crop choice, or cultivated area rather than necessarily increasing the number of conservation practices used. Similarly, the lack of an independent relationship between irrigated farm area and practice breadth suggests that productive scale alone does not determine management diversification. Larger operations may have greater investment capacity but also face higher transition costs, whereas smaller farms may have fewer resources but greater operational flexibility. The findings therefore caution against assuming that either resource scarcity or productive assets are sufficient drivers of sustainable management.

These results have wider environmental relevance because the farms operate within an endorheic basin where land- and water-management pressures can accumulate within the hydrological system. Previous studies have documented hydrological constraints, pollutant occurrence, and land-use change within the Bustillos Lagoon basin (Ochoa-Rivero et al., 2017; Rojas-Villalobos, 2019; Ochoa-Rivero et al., 2023; Valencia-Gaspar et al., 2024). Nevertheless, because the present study did not directly measure groundwater abstraction, water quality, erosion, nutrient loading, or pesticide concentrations, the observed management portfolios should be interpreted as potentially relevant agricultural behaviors rather than direct evidence of improved lagoon condition.

Objective 3: Human and Institutional Conditions Shaping Conservation-Practice Breadth

Technical assistance emerged as the most important institutional correlate of broader conservation-practice portfolios. This finding aligns with evidence that agricultural extension and advisory services can reduce information constraints, strengthen technical knowledge, support experimentation, and improve farmers’ capacity to implement sustainable land-management practices (Anderson and Elahi et al., 2021; Mugari et al., 2025; Feder, 2004; Kassie et al., 2013; Mugari et al., 2025; Piñeiro et al., 2020). Its importance is particularly noteworthy given the limited access to technical assistance within the surveyed population, suggesting that strengthening locally appropriate advisory services may offer a practical pathway for broadening management portfolios.

From an SLF perspective, technical assistance represents part of the transforming institutional environment through which available livelihood assets can be converted into management strategies (DFID, 1999; Scoones, 1998; Natarajan et al., 2022). Possession of land, machinery, irrigation infrastructure, or farming experience does not necessarily provide the specialized knowledge required to evaluate and implement less familiar conservation practices. Institutional support can therefore enhance the usability of existing human and physical capital, rather than functioning simply as an additional resource (Li et al., 2020; Muchara et al., 2014).

Environmental concern was also associated with broader practice portfolios, reinforcing the argument that sustainable agricultural behavior cannot be explained entirely by structural farm characteristics. Attitudes, perceived responsibility, environmental orientation, and other cognitive factors can influence management decisions alongside material opportunities and constraints (Kollmuss and Agyeman, 2002; Bwalya et al., 2023; Dessart et al., 2019). Because the questionnaire measured the stated importance of environmental care rather than a multidimensional awareness construct, this relationship should be interpreted specifically in terms of environmental concern, not comprehensive environmental knowledge.

Taken together, the findings for technical assistance and environmental concern suggest a potentially complementary relationship between motivation and implementation capacity. Environmental concern may encourage willingness to consider conservation-oriented management, while technical support can provide the knowledge required to translate that orientation into feasible practices. Although the cross-sectional design cannot establish causal sequencing, this pattern suggests that interventions combining environmental engagement with practical technical support may be more useful than relying on either component alone.

In contrast, formal education and producer-organization membership did not independently distinguish conservation-practice breadth. General schooling may not substitute for context-specific agricultural knowledge, particularly among farmers with substantial accumulated production experience (Asfaw and Admassie, 2004). Likewise, organizational membership alone reveals little about the purpose or quality of institutional engagement. Producer organizations oriented toward marketing, input procurement, credit, or production coordination may not necessarily facilitate conservation knowledge. Future research should therefore examine the functions and intensity of institutional relationships, rather than membership alone.

Cross-Cutting SLF Interpretation and Implications for the Bustillos Lagoon

Viewed collectively, the findings support the Sustainable Livelihoods Framework by showing that different livelihood dimensions contributed unevenly to sustainable management. The stronger relationships observed for technical assistance and environmental concern, alongside the absence of independent relationships for farm area, water problems, education, and organizational membership, suggest that management diversification depends on the configuration, accessibility, and usability of livelihood resources. In this study, livelihood strategy is therefore represented empirically as a diversified agricultural-management portfolio shaped by both institutional capacity and environmental context.

This interpretation is especially relevant to the Bustillos Lagoon because agricultural decisions occur within a closed basin where cumulative land- and water-management pressures have broader environmental implications. The existing uptake of residue management, rainwater collection, and agrochemical stewardship provides an established foundation on which conservation initiatives could build, while the limited diversification of biological soil-fertility management and conservation tillage identifies areas requiring further investigation. Rather than treating Mennonite agriculture as environmentally homogeneous, the findings demonstrate meaningful variation in management portfolios, environmental concern, and access to institutional support.

The results therefore point toward a targeted rather than generic sustainability strategy. Expanding technical assistance should be accompanied by practical assessment of the feasibility, costs, compatibility, and perceived risks of practices that remain uncommon. At the basin level, agricultural sustainability programs should also connect farm-management interventions with broader groundwater and lagoon-management objectives. Such integration is important because improving the number of practices used at farm level does not necessarily guarantee ecological improvement unless those practices effectively address the environmental pressures operating within the basin.

These interpretations remain subject to the study’s limitations. The cross-sectional design does not establish causality, self-reported practices were not independently verified, and practice breadth measures the number of practices rather than their implementation quality, spatial extent, duration, or ecological effectiveness. The relatively small sample and limited number of farmers receiving technical assistance also warrant caution. Future research should combine farmer surveys with soil measurements, groundwater and irrigation records, water-quality monitoring, and spatial land-management data, while longitudinal and qualitative studies could clarify why particular practices remain uncommon and how farmers evaluate technical advice, environmental responsibility, and resource risk.

Conclusion

This study characterizes sustainable agricultural management among Mennonite farmers in the Bustillos Lagoon region as a diversified and uneven portfolio of practices. Farmers widely incorporate some established practices, particularly residue management and rainwater collection, while conservation tillage and several biological soil-fertility measures remain uncommon. Broader conservation-practice portfolios were associated with technical assistance and environmental concern, whereas exposure to irrigation-water problems, farm scale, organizational membership, and formal education did not independently distinguish management breadth. These findings suggest that environmental pressure alone may not be sufficient to stimulate diversified sustainable management unless farmers also have access to relevant knowledge and institutional support and perceive environmental stewardship as important. From an SLF perspective, the results illustrate how livelihood strategies emerge from interactions between human and institutional capacities and environmental vulnerability rather than from resource scarcity alone. Although the study does not directly quantify effects on water quality, groundwater abstraction, or lagoon ecology, it identifies farm-management pathways and institutional conditions relevant to reducing agricultural pressures within the hydrologically sensitive Bustillos system. Its principal contribution is farmer-level evidence from an understudied Mennonite agricultural community and a portfolio-based account of how sustainable management varies within an intensive dryland farming system.

Environmental and Policy Implications

The strong association between technical assistance and broader conservation-practice portfolios identifies agricultural advisory services as a practical intervention point. The National Institute for Forestry, Agriculture and Livestock Research (INIFAP) and the Secretaría de Agricultura y Desarrollo Rural (SADER), working with Chihuahua state agricultural authorities, should expand locally appropriate technical support directed particularly toward practices that remain uncommon, including conservation tillage and biological soil-fertility management. Because relatively few respondents reported receiving technical assistance, expanding access may be more consequential than simply maintaining existing services. Delivery should account for the organizational and cultural characteristics of Mennonite farming communities and emphasize practical demonstration and locally tested options.

The association between environmental concern and broader practice portfolios suggests that technical interventions should be accompanied by environmental communication linking farm management with soil conservation, groundwater security, and basin-scale resource pressures. The Secretaría de Medio Ambiente y Recursos Naturales (SEMARNAT), the Comisión Nacional del Agua (CONAGUA), local environmental authorities, research institutions, academical institutions, agricultural organizations, and community representatives can collaborate on communication and training that make these connections locally meaningful. The absence of an independent association between reported water shortages and practice breadth further indicates that environmental stress should not be expected to generate conservation behavior automatically; farmers require viable technical and institutional pathways for responding to scarcity.

At basin scale, CONAGUA and environmental authorities should integrate farmer-level management initiatives with groundwater monitoring and hydrological planning. Conservation practices should complement, rather than substitute for, regulation and monitoring of groundwater abstraction and cumulative agricultural pressures. Universities, INIFAP, and environmental research institutions should link future farmer surveys with direct measurements of irrigation demand, soil condition, nutrient and pesticide movement, and lagoon water quality. Such monitoring would allow policy incentives to be directed toward combinations of practices that demonstrate measurable environmental benefits in the Bustillos system.

Research Involving Human Participants and/or Animals

This study involved human participants and did not involve animals. The study protocol was reviewed for ethical compliance in line with the regulations of the National Institute for Agriculture, Forestry and Livestock Research (INIFAP), Mexico. Participation was voluntary, no personally identifying information was retained in the analytical dataset, and participants could decline individual questions or withdraw without negative consequences.

Author Contributions

All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by EYO, JMOR, MBB, JAN, OCPG, SAA, and HAGJ. The first draft of the manuscript was written by EYO and supervised by HAGJ. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This research was supported by the National Institute for Agriculture, Forestry and Livestock Research (INIFAP), Mexico, and Pronatura Noreste, A.C. [Grant #1037304331, Aplicación de prácticas agrícolas sostenibles en la producción de maíz en comunidades menonitas del norte de México, 2021].

Data Availability

Data used in this study are available upon reasonable request from the corresponding author.

Compliance with Ethical Standards

Conflict of Interest

The authors declare no competing interests.

Informed Consent

Informed consent was obtained from all individual participants before questionnaire administration. Participants were informed about the purpose and procedures of the research and were assured of confidentiality and voluntary participation.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1007/s00267-026-02637-4.