Rural urbanisation and home gardening in southern Mexico: agrobiodiversity loss and alternative pathways

ABSTRACT

In southern Mexico, the home garden has played a key role in the livelihood security of rural families and in the conservation of agrobiodiversity. This article examines how, as a dominant trend, rural urbanisation has reduced the dependence on home gardening as a livelihood strategy, with a consequent decline in agrobiodiversity. Drawing upon a mixed methods approach involving household surveys and life stories, the article contributes to a better understanding of long-term dynamics in the relationship between agrobiodiversity management and people’s livelihoods, and provides novel evidence on alternative pathways that allow households to maintain greater home garden agrobiodiversity.

KEYWORDS:

• Urbanisation
• home gardening
• agrobiodiversity
• Mayan peoples
• livelihood diversification

1. Introduction

Mexico is one of the most biodiverse countries in the world (UNEP-WCMC 2016), and rural households have played a key role in the conservation and reproduction of this biodiversity. Rural households contribute in particular to the conservation and reproduction of agrobiodiversity, a subset of biodiversity comprising the crops and livestock managed by farmers, along with associated biota (Kontoleon, Pascual, and Smale 2009). The cosmovision, traditional knowledge, and practices of these households, many of which are inhabited by indigenous peoples, have allowed them to sustain high levels of agrobiodiversity in their food production systems (Moreno-Calles, Toledo, and Casas 2014; Mariaca Méndez 2012; Bellon et al. 2009).

However, as has been observed in the rest of Latin America, and across the globe, agrobiodiversity is threatened by several pressures, including demographic changes, such as urbanisation, migration, and the aging of the rural population; infrastructure development; the expansion of industrial agriculture; and climate change (Bellon et al. 2009; Moreno-Calles, Toledo, and Casas 2014; UNEP-WCMC 2016).

This article draws upon agrarian change studies to examine how transformations in rural spaces are shaping the conservation and reproduction of agrobiodiversity. In particular, this research focuses on the home garden, a small-scale agroforestry system that combines trees, shrubs, and herbs in association with domestic animals, and is usually located around the dwelling (Nair and Kumar 2006).

The neoliberal reforms implemented in the 1980s accelerated the transformations faced in the rural space. Urbanisation and policies favouring market liberalisation and industrial agriculture blurred the division between rural and urban areas and transformed the relationship of rural dwellers with land and labour (Klepeis and Vance 2003; Lerner, Eakin, and Sweeney 2013). Land dispossession and a cost-price squeeze, caused by the reduction in agricultural subsidies and increasing competition from highly subsidised products, pushed rural inhabitants to diversify their livelihoods and, in many cases, to migrate to the cities (Fry 2011).

Rural households have variously adopted a combination of market/non-market, capitalist/non-capitalist, and multi-sited urban/rural strategies to sustain their living (Fairbairn et al. 2014; Du Toit and Neves 2014). However, the de-agrarianisation process does not completely undermine the role of agriculture in rural livelihoods (Radel, Schmook, and Chowdhury 2010). Nonetheless, rural households are becoming increasingly ‘semi-proletarianized, semi-globalized and semi-urban’ (Hecht 2014, 878).

Livelihood diversification has also led to an increase in inequality in rural areas. For wealthier households, with assets to invest in off-farm activities and access to urban networks, diversification represents an accumulation strategy (Kay 2008). However, for poorer households, it is merely a survival strategy. Livelihood diversification drives the exploitation of those without means, as they end up taking precarious jobs in urban and peri-urban areas (Kay 2008; Kay 2015). Despite this increasing inequality, livelihood diversification also represents a resistance strategy for both poorer and wealthier households, as it allows them to sustain their self-provisioning agrarian activities (Isakson 2009).

Similar processes of agrarian change have been observed in other contexts across the Global South. In Indonesia, for example, forest conservation and agricultural intensification have led to the dispossession and impoverishment of rural villagers (Batubara et al. 2022). Meanwhile, in South Africa, as a result of the colonial and apartheid land dispossession and exclusion, poor households have relied on diversified livelihood strategies and networks of kinship across the rural-urban spectrum (du Toit 2018).

Intensifying and deepening the connections across the rural-urban spectrum, agrarian change not only transforms how people make their living in rural and urban areas, it also transforms how people live their lives. Of the various transformations involved in agrarian change, this article focuses on those related to the transformation of mainly rural-based livelihoods into more urbanised ones – that is, rural urbanisation (Zhijun 2004; Roberts 2016).

In this research, rural urbanisation is understood as a demographic, economic, social, and cultural transition which is manifested in population growth, improvements in infrastructure, increasing access to markets and formal education, intensified connection with urban areas, off-farm diversification, and changes in the family structure, lifestyles, and values (cf. Baños Ramírez 2001; Zhijun 2004; Cloke 2006).

This article examines how rural urbanisation is transforming the role of the home garden as a livelihood strategy and how this process affects agrobiodiversity. The geographical coverage of the research is Yucatán, a state located in the southeast of Mexico. In Yucatán, the home garden, together with the milpa, a swidden agriculture system based on the production of maize, beans, and squash, have traditionally played a key role in the livelihood security of rural families, producing an abundance of resources despite the shallow and stony soils of the region (García de Miguel 2000; Jiménez-Osornio, Ruenes Morales, and Gómez 2003).

This article contributes to filling two key gaps in the literature on agrarian change and the home garden: long-term dynamics and alternative pathways to avoiding agrobiodiversity loss. Long-term dynamics is an area that has received little attention. There are few studies analysing the transformations of the home gardens over a long period of time, and those that do address long-term dynamics are mostly based on document analysis of ethnographic and ethnobotanical records. This research draws upon life stories and panel survey data to capture the long-term dynamics of home garden management and livelihood diversification.

The second knowledge gap this research addresses is the study of the underlying factors that explain why some households manage to keep highly diverse home gardens in rural urbanising contexts. Rural urbanisation has undermined the agrobiodiversity and multifunctionality of the home garden, as a dominant trend; however, less is known about those households that have managed to maintain highly diverse home gardens (Mellisse et al. 2018; Peyre et al. 2006). This research examines both the dominant trend towards agrobiodiversity loss and the alternative pathways followed by households maintaining high levels of agrobiodiversity.

The rest of the article is organised as follows. Section 2 discusses the different manifestations of rural urbanisation and how these have transformed the home garden. Section 3 describes the research sites and the methods used in data collection and data analysis. Section 4 draws upon data from a panel household survey as well as life stories to examine how transitions in home garden agrobiodiversity relate to shifts in occupation profiles and other household characteristics. Section 5 discusses the key findings on how rural urbanisation is transforming the role of the home garden as a livelihood activity and the effects this process is having on its agrobiodiversity. Lastly, Section 6 presents conclusions and policy implications.

2. Rural urbanisation and home garden dynamics

In Yucatán, Mexico, home gardens have played a key role in the livelihood security of rural families since pre-Hispanic times. Because of their great agrobiodiversity and intensive management, home gardens perform a variety of functions, including: ecological (nutrient cycling, enhanced pollination, biodiversity conservation); material provisioning (food, fuel, timber, fodder, medicinal); economic (income, savings repository); and social (transmission and conservation of traditional knowledge) (Mariaca Méndez 2012; Galhena, Freed, and Maredia 2013).

Nonetheless, home gardens are dynamic systems that have co-evolved with people’s lives. Changes in home gardens have accelerated since the 1980s, as part of the transformations occurring in the rural space as a result of the neoliberal reforms. These agroforestry systems have faced increasing land pressures, loss of species, abandonment of traditional practices, and loss of the related knowledge (Pietersen et al. 2018; Moreno-Calles, Toledo, and Casas 2014).

As Figure 1 summarises, this article focuses on how urbanisation has transformed home garden agrobiodiversity in increasingly urbanised rural spaces. No matter the context, a common trend observed in home gardening settings around the world is population growth with the consequent division of land (Soemarwoto and Conway 1992; Kumar and Ramachandran Nair 2004; Wiersum 2006; Chávez García 2012). This division of land has reduced the size of the home gardens and diminished the availability of land for open-field cultivation systems (Chávez García 2012; Wiersum 2006). Since home gardens tend to play a complementary role within the farming system, less land available for the main farming activities can lead either to an intensification of home garden cultivation or to an abandonment of the farming system altogether, as has been observed in Indonesia (Wiersum 2006) and Mexico (Chávez García 2012).

Figure 1. Rural urbanisation as a key driver of home garden transformations.

Source: Author’s elaboration based on the literature review presented in this section.

Improvements in infrastructure and other public services are another manifestation of urbanisation. An effect of these is the facilitation of access to goods, services, and labour markets. An increasing commercialisation of home garden products has been broadly documented in Ethiopia (Abdoellah et al. 2006), India (Kumar and Ramachandran Nair 2004), Indonesia (Soemarwoto and Conway 1992; Abdoellah et al. 2006; Wiersum 2006), and Spain (Kumar and Ramachandran Nair 2004).

A greater commercial role has generally meant specialisation and increases in home garden productivity in the short-term; however, some scholars have warned that these transformations threaten the sustainability of home gardens, increasing their dependence on external inputs and undermining their multifunctionality (Soemarwoto 1987; Abdoellah et al. 2006; Peyre et al. 2006). This occurs, for example, when home gardens are converted into private spaces, thus diminishing their social role (Soemarwoto 1987), and when the animal component of the system is neglected (Abdoellah et al. 2006; Mellisse et al. 2018). Although the greater emphasis on cash crops generally leads to reductions in home garden diversity, exceptions to this rule have been found in India (Peyre et al. 2006) and Ethiopia (Mellisse et al. 2018). However, these studies do not explain the reasons behind these alternative pathways.

Home gardens are increasingly interacting with off-farm livelihood activities, resulting in both complementarities and trade-offs. Although home gardens usually require small investments, households need a certain amount of financial and physical resources to establish and manage them successfully (Pritchard et al. 2019). Off-farm livelihood activities and participation in government programmes tend to ease resource constraints (Isakson 2009; Castillo Loeza et al. 2020). However, engagement in these activities may also diminish the time people have available for home garden management (Castañeda Navarrete 2019).

The changes in people’s livelihoods have been accompanied by cultural and social transformations in the organisation of the rural family in particular, and across rural society in general (Baños Ramírez 2001; Ellis 2006; Rigg 2006). An intensified connection with urban centres, together with broader access to formal education, television, mobile phones, and the internet, has modified people’s aspirations and resulted in acculturation, especially among young people (Baños Ramírez 2001; Vogl, Vogl-Lukasser, and Caballero 2002). Rural households are becoming smaller and based on the nuclear family, rather than the traditional extended family; consumption is being separated from production; and rural populations are increasingly shifting their preferences to industrialised food and off-farm livelihoods (Baños Ramírez 2002).

The consequences of these cultural and social transformations for home gardening range from the availability of more financial resources to invest in the home garden (Peñuelas and Guadalupe 2007), the introduction of new species and new techniques (Peñuelas and Guadalupe 2007; Kumar and Ramachandran Nair 2004), and the increasingly aesthetic function of the home garden (Wiersum 2006; Hernández Sánchez 2010), to the loss of traditional knowledge (Hoffmann 2003), a decrease in the use of medicinal plants (Kumar and Ramachandran Nair 2004; Cano-Ramírez et al. 2014), less interest in home gardening (Peñuelas and Guadalupe 2007), and the ageing of gardeners (Soini 2005).

Development interventions have also transformed dwelling spaces, creating new concrete structures, and introducing tap water, but at the same time they have displaced home garden species, contributed to the erosion of traditional knowledge, and increased the dependence on external inputs, threatening the sustainability of the home garden system (Soemarwoto 1987; Mariaca Méndez 2012; López Barreto 2017).

Despite the dominant de-agrarianisation trend, resistance is also to be found among rural dwellers, with some households, despite all the pressures imposed by rural urbanisation, managing to sustain high levels of agrobiodiversity in their home gardens while continuing to cultivate traditional food production systems, such as the milpa. The literature on home gardens has focused mainly on the pragmatic motives with which people, usually women, maintain their engagement in home gardening: from the benefits of the home garden as a source of food and the resulting savings from not having to buy these products from the market, to ornamental benefits and, to a lesser extent, personal preferences and social status (Cano-Ramírez et al. 2014; Campos and del Rosario 2012; Guerra Mukul 2005).

Meanwhile, studies on the continutity of milpa cultivation – in which maize, the main Mexican staple, is produced – have also explored the roles of identity, social practice, personal fulfilment, and political resistance as powerful motives for people to continue to cultivate the milpa (de Frece and Poole 2008; Barrera-Bassols et al. 2009). For example, in Chiapas, Mexico, agrobiodiversity conservation in the milpa has been enhanced in the context of broader political and social movements in support of indigenous people’s autonomy (Hernández, Perales, and Jaffee 2022).

Furthermore, in urban and peri-urban centres, agriculture is found to constitute a risk-coping strategy against the uncertainty of job markets, and to compensate for low non-farm incomes (Satterthwaite, McGranahan, and Tacoli 2010; Lerner, Eakin, and Sweeney 2013). For excluded population groups, agriculture in urban and peri-urban contexts also signifies an act of resistance. In South Africa, millions of rural inhabitants were dispossessed of their land and pushed to cities that were unable to absorb them into wage employment (Jacobs 2018; Siebert 2020). There, urban citizens have organised themselves and claimed land for agriculture (Jacobs 2018).

3. Methods

3.1. Research sites

The research was conducted in three communities located in Yucatán, a state in the southeast of Mexico. Yucatán is in one of the most deprived regions of Mexico, with 48.9% of the population living in poverty (Consejo Nacional de Evaluación de la Política Social 2012). Yucatán has a population of 2.3 million inhabitants, 23.7% of whom speak an indigenous language, the third largest proportion among the Mexican states (Instituto Nacional de Estadística y Geografía 2020). The Mayas are the dominant indigenous group in Yucatán.

The three communities were selected to represent different modes of engagement in agriculture and distinct urbanisation transitions: Hocabá (peri-urban, sisal region), Sahcabá (semi-rural, sisal region), and Yaxcabá (semi-rural, milpa region). The criteria followed for this classification considers urbanisation as a gradient rather than a dichotomy (Tacoli 2003; Chomitz, Buys, and Thomas 2005; Lerner and Eakin 2011). To define these levels of urbanisation, a combination of three dimensions was used: community size (Anzaldo and Barrón 2009), remoteness (Chomitz, Buys, and Thomas 2005), and cultural practices (Cloke 2006). These dimensions were then operationalised using as indicators the following: absolute numbers of population and population density, distance to the metropolitan areas and access to infrastructure, and proportion of indigenous inhabitants (Unikel Spector 1968; Chomitz, Buys, and Thomas 2005; Lerner, Eakin, and Sweeney 2013).

The three communities are located within two municipalities: Hocabá and Yaxcabá. Hocabá is a municipality located in the northeast of the Yucatán Peninsula, in what was formerly known as the sisal region. It was in this region where sisal (Agave sisalana Perriné) export production took place for more than a century. Hocabá comprises a territory of 94.83 km² and a population of 6,514 inhabitants; however, 99.5% of the total population lives in two of its six localities: Hocabá (the municipality seat) and Sahcabá (Instituto Nacional de Estadística y Geografía 2020).

Yaxcabá is a municipality located in the eastern part of Yucatán, in a region known for milpera or maize growing. This region has been more isolated than the sisal region and its people’s livelihoods are still closely tied to traditional agriculture. Yaxcabá comprises a territory of 1,474 km² and a population of 16,350 inhabitants. It is divided into 45 localities, only eight of which have a population of more than 500 inhabitants. Yaxcabá, the municipality seat, is one of these localities (Instituto Nacional de Estadística y Geografía 2020).

3.2. Data collection and analysis

This research followed a mixed methods approach (Creswell and Plano Clark 2011), which involved a survey of 100 households conducted between December 2016 and April 2017, and life story interviews with a subsample of 16 of them conducted between October 2017 and January 2018. The household survey was a follow-up study of households and home gardens analysed in previous studies. Advice from local ethnobotanists was followed to identify relevant and rigorous previous studies. Data on Hocabá and Sahcabá (sisal region) came from Ortiz Pech (1999), having 1997 as base year; while the data on Yaxcabá (semi-rural, milpa region) came from Guerra Mukul (2005), having 2004 as base year.

Panel data sets were constructed using databases from the previous studies as first rounds, and the data collected from the household survey as a second round. Although the sample size of the panel data set was relatively small, this data set constitutes a subset of a larger cross-sectional database involving 262 households in the three research sites covered in this article. The larger data set allowed for confirming representativeness of the results at the community level.

The quantitative analysis of the panel data sets helped to identify different trajectories in the transitions of home garden agrobiodiversity and household occupations. Cluster analysis was applied to classify households according to the level of agrobiodiversity of their home gardens (low or high). Home garden agrobiodiversity was defined as planned agrobiodiversity, including only managed plant species and domesticated vertebrate animals. Occupation transitions were defined based on whether the main occupation of the household head was on-farm, such as farmer or milpero, or agriculture worker; or off-farm, such as construction worker, artisan, or housekeeper.

Communication with the researchers involved in the collection of agrobiodiversity and other household data from previous studies ensured the data were comparable, as similar data collection methods were followed (agrobiodiversity and household surveys). Nonetheless, agrobiodiversity levels were defined within each data set. This involved applying a separate cluster analysis to each of the two data sets obtained from previous studies as well as the author’s data set. This approach helped to compare agrobiodiversity categories between different data sets even though there may be discrepancies in how data were collected by each research team. However, this also means that, for example, a home garden that was classified as highly diverse in both data sets may show decreases (or increases) in its species richness or abundance. No matter their agrobiodiversity classification (high or low), as Table 1 illustrates, home gardens on average showed a decline in their species diversity.

Table 1. Changes in the number of plant species by agrobiodiversity transition.

Agrobiodiversity transition	Mean	Std. Dev.	Min	Max
Low – low	−7.2	9.6	−34	9
Low – high	−9.4	11.5	−37	11
High – high	−5.7	7.4	−21	9
High – low	−11.0	7.4	−28	1

Source: Author’s survey data (December 2016-April 2017); Guerra Mukul (2005); Ortiz Pech (1999).

A subsample of 16 of these households (5 in Hocabá, 6 in Sahcabá, and 5 in Yaxcabá), representing different combinations of home garden agrobiodiversity level (low or high) and occupational transitions (on-farm or off-farm), was purposively selected to be further explored through life story interviews. The purpose of using life stories was to obtain a more comprehensive understanding of the motivations behind the different livelihood trajectories followed by the research participants. The interviews were largely unstructured, having as starting point the question: How has your home garden changed over time? In each case the main home garden manager or the household head was interviewed at least twice. Research participants’ names have been substituted by names of ancient Mayan noble families to protect their identities, but otherwise the descriptions are accurate representations of these households.

Where possible, sketches of the home gardens and their transformations were developed with the guidance of the research participants. Eric Alonso Méndez Salazar, Fernando Ismael Álvarez Frausto, and Zazil Yamile Gamboa Tun, recently graduated visual artists, provided support in making final copies of the home garden drawings, which were later shared with research participants. Some of these illustrations are included in this article.

4. Results

The overall trend observed in the home gardens was a reduction in their agrobiodiversity, as Figure 2 shows. Hocabá, the peri-urban community, experienced the largest reductions in the number of plant species; while Yaxcabá, the semi-rural community located in the milpa region, experienced the largest reductions in the number of animals.

Figure 2. Changes in home garden agrobiodiversity in Yucatan, Mexico.

Source: Author’s survey data (December 2016-April 2017); Ortiz Pech (1999); Guerra Mukul (2005).

As explained above in section 3.2, home gardens were classified according to their agrobiodiversity level (plants and animals) to study their changes over time and the underlying causes of these changes. Agrobiodiversity transitions differed between the communities studied (Table 2). The peri-urban community showed the largest proportion of households (50.9%) that moved from high to low agrobiodiversity. In contrast, in Sahcabá (semi-rural, sisal region) only 14.3% of the households showed this transition, and 26.9% did so in Yaxcabá (semi-rural, milpa region). Likewise, the peri-urban community showed the smallest proportion of households (3.8%) that moved from low to high agrobiodiversity, while about a third of the households of the other two communities were located in this category.

Table 2. Transition matrix of home garden agrobiodiversity, Yucatan, Mexico.

Agrobiodiversity transition	Hocabá (peri-urban, sisal region)	Sahcabá (semi-rural, sisal region)	Yaxcabá (semi-rural, milpa region)
Low – low	26.4%	38.1%	19.2%
Low – high	3.8%	33.3%	38.5%
High – high	18.9%	14.3%	15.4%
High – low	50.9%	14.3%	26.9%
Total	100.0%	100.0%	100.0%

Notes: Hocabá and Sahcabá, 1997-2017; Yaxcabá, 2004-2017. Chi-squared test, p-value =  0.001. Number of observations: Hocabá, 53; Sahcabá, 21 and Yaxcabá, 26.

Source: Author’s survey data (December 2016-April 2017); Guerra Mukul (2005); Ortiz Pech (1999).

It was found that the three communities studied experienced de-agrarianisation as the dominant livelihood transition, with many household heads shifting from on-farm to off-farm activities. In 1997 about 40% of the household heads were engaged in on-farm activities as their main livelihood in Hocabá (peri-urban, sisal region) and Sahcabá (semi-rural, sisal region), and close to 90% in Yaxcabá in 2004 (semi-rural, milpa region). However, by 2016/2017 less than 20% of the household heads were engaged in on-farm activities in Hocabá and Sahcabá, and about 70% in Yaxcabá.

In Hocabá (peri-urban, sisal region) and Sahcabá (semi-rural, sisal region), the main occupation performed by men was construction worker (15-28%). By contrast, in Yaxcabá (semi-rural, milpa region) over 40% of the men still worked mainly in agriculture. The main occupation among women in the three communities was ama de casa, or homemaker (52-80%), which usually involves several crucial reproductive and productive activities, including home gardening.

Handicraft making from sisal fibre was the second most common occupation in Sahcabá (semi-rural, sisal region), among both men and women. In Hocabá (peri-urban, sisal region) and Yaxcabá (semi-rural, milpa region), the second main occupation among women was housekeeper, mainly for houses located in Mérida, the capital city. Among men, labourer was the second main occupation in Hocabá (peri-urban, sisal region), while it was construction worker in Yaxcabá (semi-rural, milpa region).

Insights from the household survey and in-depth interviews revealed different complementarities and trade-offs between on-farm and off-farm occupations and home gardening management, and showed how these interact with demographic changes, such as ageing. The accounts of two research participants illustrate some of these:

When my husband finished his work in the milpa and there was much employment in Cancún, he left for three weeks or a month, while the next work in the milpa started. Then we bought pigs, raised them, and then sold them to invest in the house. We even raised cattle, we bought a bull calf. We raised it here in the backyard. We fed him with squash [from the milpa], breadnut tree leaves, and grass. I think we raised about seven big [bull calves]. From that we invested in the house. (Female research participant from Yaxcabá, Yucatán, 60 years old, 06/12/2018)

Before, she [my mother] did it [home gardening], but now because of her age she falls ill frequently, so she is leaving it. And we [women living in the house] do not have time; no one can take care of it. Almost all go to work [to Mérida, the capital city]. (Female research participant from Sahcabá, 44 years old, 14/11/2017)

Table 3 presents a matrix of agrobiodiversity and occupational transitions. Although differences in the proportion of households across all the combinations of diversity and occupation transitions were not all statistically significant, some associations between diversity and occupation transitions emerged from the analysis. Those households that transitioned towards off-farm occupations were more likely to present a reduction in their home garden agrobiodiversity, whereas those households where the head continued to work in agriculture showed a higher probability of maintaining a highly diverse home garden. An unexpected finding was that only 20% of the households that returned to on-farm occupations – that is, ‘off-farm–on-farm’ transition – experienced a transition from low to high agrobiodiversity.

Table 3. Transition matrix of home garden diversity and household head’s main occupation in the research sites, Yucatan, Mexico.

Agrobiodiversity transition	Occupational transition (Main occupation of the household head)
	On-farm–on-farm	On-farm–off-farm	Off-farm–off-farm	Off-farm–on-farm
Low – low	15.4%	36.7%	25.6%	40.0%
Low – high	38.5%	10.0%	12.8%	20.0%
High – high	23.1%	16.7%	15.4%	0.0%
High – low	23.1%	36.7%	46.2%	40.0%
Total	100.0%	100.0%	100.0%	100.0%

Notes: Hocabá and Sahcabá, 1997-2017; Yaxcabá, 2004-2017. Chi-squared test, p-value =  0.135. Number of observations: Hocabá, 53; Sahcabá, 21 and Yaxcabá, 26.

Source: Author’s survey data (December 2016-April 2017); Guerra Mukul (2005); Ortiz Pech (1999).

The relevance of on-farm activities in explaining home garden agrobiodiversity is illustrated by the case of the Xiu family (Box 1). Located in the semi-rural community of the milpa region, the Xiu family is a ‘high–high’ agrobiodiversity and ‘on-farm—on-farm’ occupational transition household. The engagement in milpa cultivation and an attachment to the Mayan culture have contributed to maintaining Mrs and Mr Xiu’s interest in home gardening. However, their case also shows how, even in this traditional maize growing region, livelihoods are changing, and their children are less likely to engage in home gardening.

Box 1. The Xiu family.

Table

Agrobiodiversity transition: high–high Occupational transition: on-farm—on-farm	Location: Yaxcabá (semi-rural, milpa region) Home garden size: 600 m^2
This family is formed by four adults. Mr Tutul Xiu, the household head, is over 70 years old, as is his spouse. He speaks both Mayan and Spanish, while his spouse only speaks Mayan. They have seven children. The two children who live with them are in their twenties and speak both Mayan and Spanish. The other five children are already married and have formed their own households.
Mr Tutul Xiu is illiterate. He lost his parents when he was a child and his uncle took care of him, but with some limitations. For example, his uncle did not send him to school, despite sending his own children to school. Mr Xiu has always depended on the milpa as his main livelihood, while his spouse works as a full-time homemaker. His sons learned how to cultivate the milpa, but they do not do it anymore. Recently things have changed at home, as their adult children were hired as workers by a middle-size pig farm. ‘Since they [children] get their fortnightly pay, they tell me that they can buy maize, but I like it [to cultivate the milpa].’
They have chickens, but do not keep pigs. They had to sell them, as required by the pig farm where their children work. ‘They may lose their job, as [keeping pigs at home] is forbidden.’
Mr Tutul Xiu comments:
They [children] are happy because they get money fortnightly. Because in Mérida [the capital city] it is the same, you have to pay rent, electricity and water bills; it is hard [to afford all the bills]. Here it is not. Here is better, because I have my milpa, my maize, they do not have to buy anything. If they earn money, it is for buying their own stuff, their clothes.

Table 4 presents selected household characteristics by occupational transition. Occupational transitions showed associations with specific household characteristics. The households that shifted into or remained in off-farm occupations reported higher education levels (‘pull-factor’) and a higher dependency ratio (‘push-factor’). These characteristics are likely indicative of a generational transition in these households. In contrast, the ‘on-farm–on-farm’ households tended to belong to elderly people, such as Mr Tutul Xiu. These households reported the highest mean age of household members, the lowest level of education, the smallest household size, and the lowest youth dependency ratio.

Table 4. Household characteristics by occupational transition, Yucatan, Mexico (2016-2017).

Household characteristics	Statistic	On-farm – on-farm	On-farm –off-farm	Off farm –off-farm	Off-farm – on-farm	Anova / Kruskal-Wallis H / Chi-squared (p value)
Average education of the adult members of the household (mean, years)	Mean	4.5	5.9	6.5	4.5	0.095
	Median	4.8	6.3	6.4	4.5	0.092
Average age of household members (mean, years)	Mean	56.9	45.5	48.9	52.1	0.008
	Median	55.4	45.5	47.2	58.3	0.009
Household size (mean, number of members)	Mean	3.9	6.9	5.6	6.6	0.073
	Median	3.0	6.0	5.0	3.0	0.046
Youth dependency ratio (Ratio of the children under 15 years old divided by the number of adults in the household)	Mean	0.1	0.5	0.5	0.2	0.027
	Median	0.0	0.4	0.0	0.0	0.131
Household head only speaks Mayan	Proportion	9.5	3.6	0.00	0.00	0.447
Household head speaks Spanish and Mayan	Proportion	85.7	92.9	89.5	100.0	0.447
Household head only speaks Spanish	Proportion	4.8	3.6	10.5	0.00	0.447
Percentage of household members working in urban jobs	Mean	0.2	0.3	0.3	0.5	0.043
	Median	0.00	0.3	0.3	0.5	0.060
Wealth index (0-1, 5 assets)	Mean	0.3	0.3	0.4	0.3	0.961
	Median	0.2	0.4	0.4	0.4	0.917
Household income, adult scale equivalent, MXN (USD)	Mean	1,426.0	1,545.2	1,546.1	1,879.2	0.864
		(75.9)	(82.3)	(82.3)	(100.1)
	Median	1,232.3	1,493.1	1,431.8	1,501.7	0.761
		(65.1)	(79.5)	(76.2)	(80.0)

Notes: Hocabá and Sahcabá, 1997-2017; Yaxcabá, 2004-2017. Number of observations: Hocabá, 53; Sahcabá, 21 and Yaxcabá, 26.

Source: Author’s survey data (December 2016-April 2017); Guerra Mukul (2005); Ortiz Pech (1999).

The ‘off-farm–on-farm’ occupational trajectory is of particular interest since it represents an alternative to the dominant pathway moving towards off-farm occupations. These households reported lower education levels and lower youth dependency ratios than the ‘on-farm–off-farm’ and ‘off-farm–off-farm’ households. Moreover, they reported the largest proportion of household members working in urban jobs and the highest household income. This means that even if the household head returned to work in agriculture, their children were likely working in urban jobs and contributing to the household income.

The case of the Itzáes family, presented in Box 2, illustrates a ‘high–low’ agrobiodiversity, ‘off-farm–on-farm’ occupational transition household. This household is located in a peri-urban community and exemplifies the relationship between family life cycle and the engagement in urban jobs. It also illustrates the division of the land as a consequence of population growth and how this relates to the high–low agrobiodiversity transition in their home garden.

Box 2. The Itzáes family.

Table

Agrobiodiversity transition: high–low Occupational transition: off-farm–on-farm	Location: Hocabá (peri-urban, sisal region) Home garden size: 600m^2
This family is formed by 3 adults. Mrs Itzáes, 45 years old; her husband, 46 years old; and their daughter, 23 years old. All of them speak Mayan and Spanish.
Mrs Itzáes explained that she moved to this plot to live with her husband and parents-in-law when she got married. They used to have more plants and animals in their home garden. ‘We used to plant coriander, radish and beetroot. Now I only have my orange trees. We also used to have more pigs, but the plot got small.’ The plot was evenly divided between the three children (Mr Itzáes, his brother and sister) when they got married. Nonetheless, Mrs Itzáes still plants herbs and vegetables in another plot she has.
Mr Itzáes started working in the milpa with his brother and father at a young age. When he was about 20 years old he began to work in Mérida as a loader in a factory. Then he found a job in a hotel. It was around that time when he met Mrs Itzáes and got married. Now Mr Itzáes works in the milpa, cultivating a non-mechanised hectare for his nuclear family’s consumption. He grows maize, wax beans, black beans, cowpeas, squash, cucumber, and watermelon.
Mrs Itzáes’ daughter completed an undergraduate degree and recently found a job. Mrs Itzáes explained: ‘Since she [daughter] completed her studies, her dad does not think that much on it [working in the city]. We helped her as far as we could.’

Figure 3 depicts the evolution of the Itzáes’ home garden over the last 35 years. The left panel shows how the home garden looked when Mr Itzáes was a child. They used to live in a traditional Mayan house and had more animals and plants. The right panel shows the plot where Mr Itzáes now lives with his wife and daughter. Their plot is a third of the size of Mr Itzáes’ parents’ plot. They live in a house made of concrete, and at the back of the plot they built an English toilet. They continue to raise pigs and keep them in a pigsty made of concrete blocks, instead of tethered to trees in the garden.

Figure 3. Evolution of the home garden of the Itzáes family.

Source: Drawings by Eric Alonso Méndez Salazar.

While positive associations were found between on-farm livelihood activities and greater levels of home garden agrobiodiversity, it was also noted that some households that depended mainly on off-farm activities were managing to maintain highly diverse home gardens. These households were examined in further detail. Table 5 presents selected characteristics of the ‘off-farm–off-farm’ households by diversity transition. Those households that maintained highly diverse home gardens – ‘low–high’ and ‘high–high’ agrobiodiversity transitions – seemed to possess better endowments: that is, they had a higher level of education, more household members, more assets, and higher income. However, differences were not statistically significant. Nonetheless, when analysed in pairs, significant differences were found in the mean income between ‘high–high’ and ‘high–low’ diversity transition households (0.05 significance level); and in the mean number of household members between ‘low–low’ and ‘low–high’ agrobiodiversity transition households (0.05 significance level).

Table 5. Household characteristics of ‘off-farm–off-farm’ households by home garden agrobiodiversity transition, Yucatan, Mexico (2016-2017).

Household characteristics	Statistic	Low–low	Low–high	High–high	High–low	Anova / Kruskal-Wallis H / Chi-squared (p value)
Average education of the adult members of the household (mean, years)	Mean	6.1	7.2	7.5	6.3	0.787
	Median	6.0	7.5	6.8	6.5	0.653
Average age of household members (mean, years)	Mean	53.7	39.8	45.8	50.1	0.224
	Median	55.0	38.5	41.5	49.3	0.223
Age of the household head (mean, years)	Mean	61.6	63.2	61.8	59.9	0.975
	Median	65.0	63.0	63.0	60.5	0.967
Household size (mean, number of members)	Mean	4.4	9.0	6.3	5.0	0.105
	Median	4.0	8.00	4.0	4.5	0.153
Youth dependency ratio (Ratio of the children under 15 years old divided by the number of adults in the household)	Mean	0.4	0.6	0.3	0.6	0.802
	Median	0.0	0.7	0.0	0.0	0.554
Household head speaks Spanish and Mayan	Proportion	66.7	100.0	100.0	94.4	0.081
Household head only speaks Spanish	Proportion	33.3	0.0	0.0	5.6	0.081
Wealth index (0-1, 5 assets)	Mean	0.3	0.3	0.5	0.4	0.578
	Median	0.2	0.4	0.4	0.4	0.457
Household income, adult scale equivalent, MXN (USD)	Mean	1486.9	2152.3	2153.6	1204.9	0.244
		(79.2)	(114.6)	(87.8)	(49.1)
	Median	1052.90	1750.47	1563.52	1388.06	0.142
		(56.1)	(93.2)	(83.3)	(73.9)

Notes: Hocabá and Sahcabá, 1997-2017; Yaxcabá, 2004-2017. Number of observations: Hocabá, 53; Sahcabá, 21 and Yaxcabá, 26.

Source: Author’s survey data (December 2016-April 2017); Guerra Mukul (2005); Ortiz Pech (1999).

These results likely indicate that better-off and larger households are more likely to diversify their livelihoods without undermining their home garden agrobiodiversity. Although off-farm occupations within the communities tended to be associated with lower incomes (MXN 2,038 = ∼USD 108.50 on average) than occupations outside the communities (MXN 4,396 = ∼USD 234.20 on average), they also seemed to allow households to better diversify their livelihoods, as the case of the Cocom family illustrates (Box 3).

Box 3. The Cocom family.

Table

Agrobiodiversity transition: high–high Occupational transition: on-farm–off-farm	Location: Hocabá (peri-urban, sisal region) Home garden size: 1,300m^2
This family is formed by three adults and two children. Mrs Cocom, the household head, is over 60 years old and speaks Mayan and Spanish. Her daughter is 40 years old and her son is 37 years old; both are single. Her daughter speaks Mayan and Spanish and her son only speaks Spanish. Her daughter has two sons who live in the house, the oldest being 16 years old and the youngest 4 years old. None of them speak Mayan.
Mrs Cocom’s father worked cultivating sisal. When Mrs Cocom was 14 years old, she started to commute to work in Mérida. She worked as a housekeeper in the same house for 7 years. Then she went back to Hocabá and got married with Mr. Cocom. They moved to this plot 41 years ago. The plot was a present from her mother-in-law.
Mrs Cocom is now a widow; her deceased husband worked in agriculture. After her husband passed away, she became the household head. Regarding her home garden, she remarked: ‘I like having plants, flowers, animals. My grandson also likes it, he started to raise animals again.’
Mrs Cocom comments:
Now is when we have more plants, because before, when I got married, I did not have [plants] because I had children. Now I have more plants because my daughter likes them. We have medicinal plants, such as basil, epazote, mint, peppermint. We have the chickens locked because they eat the plants.
Mrs Cocom and her daughter work selling food at their house. ‘We stay here to make a living.’ About how life has changed she explained:
Now everything is expensive. Before, it was cheap, because the maize was cultivated by the milperos. Everything came from the milpa: wax beans, black beans, squash, watermelon, tannia. There was harvest, there was food. People had tomatoes, radish, coriander in their home gardens. Now there is so much laziness and since you can buy it, people do not grow it anymore. There are people who still grow plants, because they like to eat fresh and better, but there are fewer (…) Now people use too many chemicals for growing the crops faster and that harms us. That is how they feed animals. Before, there were not chemicals. You raised the pig only with things from the milpa, with the squash, the breadnut tree and the maize.
They cut down several trees to build the bicycle repair shop where her grandson works. Her grandson learned the trade from Mr Cocom.

The case of the Cocom family represents a ‘high–high’ diversity and ‘on-farm–off-farm’ occupational transition household. It illustrates how off-farm livelihoods within the communities allow households to maintain their home garden agrobiodiversity. It also exemplifies how the ‘living space’ represented by the home garden, where the family conducts its everyday activities, evolves along with the changes in people’s livelihoods. Other factors that emerged from this case as explanations of the transformations in the home gardens are family life cycle, household size, household preferences, and cultural attachment. Cultural attachment here refers to adherence to Mayan culture, and the traditional knowledge and practices involved in it, including speaking the Mayan language.

Figure 4 depicts the home garden of the Cocom family. The Cocoms’ home garden features rooms made of concrete, the use of blocks and wire mesh in the construction of the poultry pen and the pigsty, and a bicycle repair shop.

Figure 4. Home garden of the Cocom family, 2017.

Source: Drawing by Eric Alonso Méndez Salazar.

Household preferences and cultural attachment contribute to explaining the different trajectories in home garden agrobiodiversity. In the households that reported ‘low–high’ and ‘high–high’ agrobiodiversity transitions, all the household heads spoke both Spanish and Mayan (Table 5). Their interest in learning and practicing the Mayan language may be interpreted as a sign of attachment to the ‘traditional’ culture, and thus a greater valuing of home garden agrobiodiversity than those household heads who only spoke Spanish. This cultural attachment, captured through the spoken language, also emerged as a relevant explanatory factor from the life stories.

5. Discussion

Rural urbanisation has shaped home gardening in various ways. Demographic pressure has resulted in land division and, consequently, less space for cultivating plants and raising animals. Moreover, population growth has meant less land for extensive agriculture (milpa), a driver of off-farm diversification. Lower participation in milpa cultivation has also resulted in lower animal diversity in the home garden, since the milpa used to be the main source of animal feeding. Today, its role in livelihood security has been undermined by multiple factors, among them: population growth; changes in weather patterns; use of agrochemicals and the consequent reduction in agrobiodiversity; lower soil fertility and unaffordable fertilisers (Baños Ramírez 2001; Sampson 2015).

Urbanisation and the related improvements in public infrastructure have facilitated work-related movements to urban areas. With intensified rural-urban interactions, young adults in the research sites now have less time for and less interest in home gardening. Moreover, with the expansion of public education facilities and government cash transfer programmes that are conditional on school attendance, children also have less time to contribute to home garden management. Those young families still interested in home gardening tend to prefer ornamental plants to vegetables and animals. Hernández Sánchez (2010) called this trend ‘gardenisation’ of the home gardens. Elderly people have become the main managers of home gardens; however, increased caring responsibilities due to the ageing of family members also undermine home garden agrobiodiversity. The general result of this trend is less plant and animal diversity than before, and the loss of traditional knowledge.

In other contexts, better communications infrastructure has meant broader participation of home gardeners in local markets, as discussed above in section 2. However, this was not observed in the research sites of the present study. Instead, access to markets has reduced the incentive for many households to produce their own food. Two likely explanations for this different pattern are the smaller size of the Yucatecan home gardens in comparison with those in other contexts – which hinders the generation of enough surplus to afford transportation costs – and the preference of some households for other livelihoods in order to earn an income. Correa Navarro (1997) documented the presence of commercial home gardens in the sisal region two decades ago. However, broader access to off-farm livelihoods appears to have reduced the relevance of home gardens as a means to earn an income for many.

Off-farm diversification has also driven changes in the structures of home gardens. Since home gardens are part of the dwelling, they are affected by changes in people’s livelihoods. Now bike repair shops, grocery shops, and areas for the preparation and sale of food have become part of the solar (the plot where the home garden is located), as Figure 4 shows. These structures reduce the vulnerability of the household in the event of tropical storms and hurricanes, and help to solve other dwelling problems, such as open defecation and overcrowding. However, they also reduce the space available for plants and animals. This replacement of traditional structures with ‘modern’ constructions reflects both a cultural change in peoples’ preferences and more availability of income. Urban jobs are the main source of income supporting the building of these structures; however, the government and NGOs have also provided support for the construction of toilets and rooms. Similar findings have been observed in other contexts, such as Myanmar, where engagement in off-farm livelihood activities has been found to involve both complementarities and trade-offs with home garden management (Pritchard et al. 2019).

Nonetheless, alternative pathways to this dominant trend, of less dependence on home gardening and agrobiodiversity loss, were also identified. Better-off and larger households are managing to diversify their livelihoods without undermining their home garden agrobiodiversity. Better-off households are able to invest more in their home gardens, particularly in the livestock component, using them as a savings repository. Larger households likely benefit not only from a larger workforce, but also from stronger mutual support, characteristic of extended families. This means that some members are able to work in the city, while others stay at home to take care of the children and manage the home garden and the milpa.

Cultural attachment to the Mayan culture and its traditional practices is another factor influencing alternative pathways of households maintaining greater home garden agrobiodiversity. Households headed by people who speak Mayan showed more diverse home gardens. Off-farm diversification within the communities also explains higher levels of home garden agrobiodiversity. In comparison with urban jobs, off-farm activities within the communities allow households enough time to manage the home garden, while earning more than they would from on-farm activities. This livelihood diversification strategy allows them to buffer market, health, and weather shocks.

6. Conclusions

This research has found that, as a dominant trend, rural urbanisation has reduced the dependence on home gardening as a livelihood strategy in Yucatán, Mexico, with a consequent decline in agrobiodiversity. Urbanisation has meant less land and less interest in agriculture, but also more off-farm livelihood opportunities and better access to infrastructure, public services, and urban markets. In the past, rural households depended on the milpa as their main livelihood, and the home garden was part of this broader agricultural system. However, various factors have disincentivised engagement in the milpa system for many people. This, in turn, has affected home garden agrobiodiversity, although it was observed that the abandonment of the milpa has not necessarily led to the abandonment of the home garden. Men may work in urban jobs while their wives or mothers continue home gardening.

Nevertheless, the research has also identified households following an alternative pathway that allows them to maintain greater home garden agrobiodiversity. Off-farm diversification within the communities and attachment to the Mayan culture are two key factors influencing these alternative pathways. These findings suggest that more opportunities to earn an income within their communities or in intermediate towns, rather than urban jobs, would contribute to improving people’s wellbeing, saving the time and money they would otherwise spend in commuting to the urban areas.

Rural urbanisation is opening opportunities to participate in off-farm livelihoods; however, the related transformations also represent a risk which could threaten future livelihood options. In the research sites, traditional knowledge, labour, and land are the main resources households have at their disposal. As people become older, their chances to work in urban jobs are reduced. In a precarious economic environment, it is unlikely that many will have gainful, formal employment or receive a pension on retirement. Thus, if they lose their interest in and knowledge of agriculture, they also risk losing the long-term wellbeing of their households and communities.

While a discussion of discrimination against indigenous people in the region is beyond the scope of this paper, findings indicate that promoting an institutional environment that allows people to freely express and embrace their ethnicity without the risk of being discriminated against would contribute to preserving the biocultural richness of their home gardens.

Emerging evidence indicates that the COVID-19 pandemic may have reversed, to some extent, the trend of de-agrarianisation and agrobiodiversity loss. Many people returned to rural areas to work in agriculture in response to increasing unemployment, reduced wages, and lockdown measures (Observatorio Regional de Gobernanza y Coordinación Social ante el Covid-19 2020). However, the extent and permanence of this trend is less clear. Little research has been conducted in rural areas during the COVID-19 pandemic, as many communities did not allow access to non-locals, in order to prevent virus outbreaks. Nonetheless, emerging studies have revealed a revival of local markets, but also the complementarity between wage jobs and agrarian activities, since less income has also translated into fewer resources for maize cultivation (Amaro, María, and Flores Romero 2022).

Although the representativeness of this case study is limited to the Yucatán Peninsula region, some of the trends illustrated in the Yucatecan case have also been observed in other parts of Mexico and other contexts across the Global South, as discussed in the first sections of this article. The Yucatecan case may be useful to illustrate how the land issue has implications beyond providing a means of subsistence to traditional societies and social justice reparations. Ensuring access to cultivable land also contributes to the conservation and reproduction of agrobiodiversity. However, as the Yucatecan case shows, off-farm income has become as important as land in ensuring the participation of rural households in on-farm activities.

A key challenge of this study was to rely on data sets collected by different research teams. The author conducted detailed analyses to avoid driving conclusions from non-comparable data. This included defining agrobiodiversity levels within each data set, as explained in section 3.2. The data sets borrowed from previous research made it possible to track back a total of 100 households and to examine changes in home garden agrobiodiversity, occupational profiles, and family life cycles. However, due to data limitations, the analysis could not address how these transformations relate to changes in the wealth, food security, and other wellbeing dimensions of the households studied. Future research using longitudinal databases covering such variables would contribute to a better understanding of the wellbeing consequences of changes in livelihood activities and agrobiodiversity.

Acknowledgements

The author expresses her gratitude to all the research participants involved, as well as to †Dr Heriberto Cuanalo de la Cerda for his generous support facilitating fieldwork activities in Yaxcabá, Yucatán; and to Dr Rafael Ortiz Pech, MSc Rogelio Guerra Mukul, Dr Javier Becerril García; Dr Juan Jiménez Osornio; and MSc Jessica Quintana Loeza who kindly shared their research databases and helped me to understand better the local context. The author is also grateful for the comments provided by Dr Sung Kyu Kim to an earlier version of this article and for the valuable guidance provided by Dr Rachel Sabates-Wheeler and Dr John Thompson throughout this research.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by the Mexican Council of Science and Technology (CONACYT), the Ministry of Research, Innovation and Higher Education of Yucatán (SIIES) and the Mexican Ministry of Education (SEP), through their postgraduate research scholarship programmes.

Notes on contributors

Jennifer Castañeda-Navarrete

Jennifer Castañeda-Navarrete is a Development Economist with over ten years’ experience in policy analysis from both qualitative and quantitative perspectives. She works as Senior Policy Analyst at Policy Links, the knowledge exchange unit of the Centre for Science, Technology and Innovation Policy (CSTI), University of Cambridge. Jennifer provides expertise in development and feminist economics for the work conducted by Policy Links. Before joining the Policy Links Unit, Jennifer worked in academia and public administration in Mexico.