Introduction

Plant genetic resources for food and agriculture are the basis of human nutrition, industrial inputs and pharmacologically active ingredients. Cereals, fruits, roots and vegetables contribute significantly to agrobiodiversity (Bellon et al., 2009). The ex situ conservation of underutilized landraces and wild relatives has become a relevant source of unexplored genes, enriching germplasm banks and their use in breeding. Mexico is a megadiverse country, and rural communities manage a wide range of domesticated and semi-domesticated biological variants, cultivated and wild relatives which significantly increase agrobiodiversity (Casas & Vallejo, 2019). Field collections of species with recalcitrant seeds, such as chayotes (genus Sechium P. Br.), are among the strategies for ex situ conservation of agrobiodiversity. Chayote is an increasingly important vegetable in international markets, and morphotypes of the Mexican clade have gradually been collected (Barrera-Guzmán, Cadena-Iñiguez, Legaria-Solano, & Sahagún-Castellanos, 2021).

In Mexico, Sechium compositum (Donn. Sm.) C. Jeffrey, S. chinantlense Lira & F. Chiang, S. hintonii (Paul G. Wilson) C. Jeffrey and S. edule (Jacq.) Sw have been recorded as endemic species. Among these, only S. edule is edible since the other three species have bitter-tasting fruits. S. edule, particularly the smooth green var. virens levis variant, is distributed on all continents (Cadena-Iñiguez, 2005). Its success as a vegetable has transcended local markets, becoming an export product (Cadena-Iñiguez & Arévalo-Galarza, 2011). Different studies highlight S. edule as a species with notable intraspecific diversity, contributing to the feeding and rural local economy (Aguirre-Medina, Cadena-Iñiguez, Olguín-Hernández, Aguirre-Cadena, & Andrade-Luna, 2021). However, most published research only mentions S. edule without specifying the biological variant. For example, (Dire, Lima, Gomes, & Bernardo-Filho, 2003), (Setzer & Setzer, 2003), (Ordoñez, Gomez, Vattuone, & Lsla, 2006), (Loizzo et al., 2016), (Vieira, Pinho, Ferreira, & Delerue-Matos, 2019), among others, do not specify the variants studied, making it challenging to reproduce their findings. (Lira, Castrejón, Zamudio, & Rojas-Zenteno, 1999) performed a grouping analysis of edible morphotypes of S. edule and another for a wild type, without distinguishing or specifying the variation. This makes conservation actions difficult since morphotypes must be identified by some taxonomic method to be included in genebanks. In this case, the popular common name is not appropriate.

Morphological, anatomical, biochemical and genetic studies, carried out by various authors (Avendaño-Arrazate et al., 2012; Barrera-Guzmán et al., 2021; Barrera-Guzmán, Legaria-Solano, Cadena-Iñiguez, & Sahagún-Castellanos, 2021; Cadena-Iñiguez & Arévalo-Galarza, 2011; Cadena-Iñiguez, 2005; Donato, Cequea, & H, 1994; Iñiguez et al., 2011; Iñiguez-Luna et al., 2021; Machida-Hirano et al., 2015), have shown that there are stability, heritability and uniformity traits that allow the distinction of S. edule variants, which facilitate their conservation, management and research. These traits confer different and desirable characteristics for its use as food (Aguiñiga-Sánchez et al., 2015; Aguiñiga-Sánchez et al., 2017; Cadena-Iñiguez, Soto-Hernández, Arévalo-Galarza, Avendaño-Arrazate, & Aguirre-Medina, 2013; Salazar-Aguilar et al., 2017). Genebanks are crucial to avoiding diversity loss and maintaining this agrobiodiversity. In 2005, the National Germplasm Bank for Sechium edule (BANGESe) was founded to preserve endemic species and biological variants. To achieve this, fruits were collected from 12 Mexican states, as well as from Guatemala and Costa Rica. Currently, the BANGESe safeguards more than 300 accessions in a field collection since Sechium seeds are recalcitrant and do not respond to traditional conservation methods (Ramírez-Rodas et al., 2021).

The accessions are conserved in a taxonomic arrangement following the principles by (Stace, 1986) and adapted for S. edule. It integrates morphotypes as varietal groups that possess stable, heritable characteristics and fertile offspring. These varietal groups are albus minor, a. levis, a. dulcis and a. spinosum for plants with yellow, smooth and prickly fruits. For fruits with dark green epidermis with and without thorns are nigrum minor, n. conus, n. levis, n. xalapensis, n. spinosum and n. maxima. The morphotype virens levis, the most widely distributed in the world, reports light green fruit without spines, while the only representative with bitter fruits of S. edule is the amarus sylvestris morphotype. The above taxonomic nomenclature has facilitated the handling, distribution and distinction of accessions, allowing research, bioprospecting, breeding and preservation activities.

As in many germplasm collections, BANGESe accessions come from direct collection from the areas of origin, distribution and domestication, as well as from donations by rural inhabitants who have locally safeguarded agrobiodiversity for a particular interest. This raises several important questions related to the management, design and regeneration strategies of these resources. For example: Which of the Sechium varietal groups have the broadest geographical distribution? Which biological variants (morphotypes) are the most important in terms of consumer preferences and what characteristics drive this? What is the origin of the accessions, and what implications does this have for their conservation? Which stakeholders are most engaged with conservation efforts? What are the primary uses of the fruits? How does the distribution of varietal groups vary with altitude? Finally, what are the morphological traits of the fruits that facilitate their visual distinction?

Genebanks are long-term conservation centres of genetic resources, especially to preserve agrobiodiversity. However, without thorough characterization and documentation of the accessions, making informed decisions about conservation, research, genetic improvement and potential use is challenging (Weise, Lohwasser, & Oppermann, 2020). Since Sechium is a cross-pollinated plant, asexual methods of multiplication are used to preserve the genetic identity of the accessions. The most successful are grafting, rooting of cuttings, and in vitro multiplication (Figure 1).

Materials and methods

Results

Descriptive analysis

BANGESe contains 231 accessions representing the intraspecific variability of S. edule in 12 varietal complexes: albus dulcis, albus levis, albus minor, albus spinosum, amarus sylvestris, nigrum conus, nigrum levis, nigrum maxima, nigrum minor, nigrum spinosum, nigrum xalapensis and virens levis. Also, there are four accessions of S. chinantlense, 12 of S. compositum and two of S. hintonii (the latter was not analyzed). The virens levis, n. spinosum and n. xalapensis varietal complexes are the most represented in the collection (Figure 2a). The major diversity of S. edule in Mexico is found in the state of Veracruz, with 110 accessions (47.6% of the total) (Figure 2b). Of the 231 accessions, 62.8% were collected from backyards and 23.3% from commercial orchards, while S. compositum, S. chinantlense and S. edule var. amarus sylvestris came from ruderal and forest areas (Figure 2c). Some accessions of virens levis, n. spinosum and n. xalapensis were from Costa Rica (CR) and Guatemala (Guat) (Figure 3).

It is well known that women play a key role in vegetable cultivation and plant breeding in domestic orchards. Our data showed that women managed 58.9% of the S. edule accessions, all of which are edible. Wild accessions are outside the interest of men and women; however, their conservation in genebanks is important due to their phytochemical potential for pharmacological use and as sources of genes for genetic improvement in the face of new challenges. The varietal complexes virens levis, nigrum xalapensis, n. spinosum, n. levis and albus dulcis were identified as man-enterprise, due to their large-scale commercialization. The 70.8% of the varietal complexes are cultivated for commercial purposes, while 29.2% are used for self-consumption, mainly focusing on fruit. This is especially relevant for the lesser-known variants, highlighting their potential to open new markets. The virens and nigrum groups are mainly used for commercialization, while albus is used for self-consumption. Bitter-flavoured varieties are generally not used by rural stakeholders (Figure 4).

Sechium sp. can be found at altitudes from 1,200–3,376m, due to its plasticity, which allows it to adapt to conditions different from its original habitat in mesophyll forests (Cadena-Iñiguez et al., 2008). The highest average altitude for nigrum spinosum was 1,993m. Outliers' presence affects some central tendency measures such as the mean, and some populations of the varietal complexes appear far from their optimal ranges for each specific variety.

Multiple correspondence analysis (MCA)

For this analysis, the phenotypic variables of fruits were included (colour, presence/absence of thorns, size, flavour and shape). The first two principal components explained 27.24% of the total obtained variation. Table 1 shows the eigenvalues and eigenvectors, respectively. For principal components 1 and 2 (PC1 and PC2), the variables contributing the most were fruit flavour, populations condition (forest, ruderal-wild and orchard), fruit colour (yellow and light green), small fruit size and flattened shape.

Table 1: Eigenvalues and variance are explained for the first ten components of the MCA. PC, principal component.

Component	Eigenvalue	Variance (%)	Cumulative variance (%)
PC 1	0.49	15.50	15.50
PC 2	0.42	13.48	28.98
PC 3	0.30	9.48	38.46
PC 4	0.24	7.79	46.24
PC 5	0.20	6.49	52.74
PC 6	0.19	6.16	58.90
PC 7	0.18	5.82	64.72
PC 8	0.16	5.20	69.92
PC 9	0.15	4.82	74.73
PC 10	0.15	4.70	79.44

The distribution of the accessions depended on the environmental conditions. Accessions of S. compositum, S. chinantlense and S. edule var. amarus sylvestris were distributed in wild and ruderal-wild forms. The virens levis complex was primarily found in orchard conditions and is the most commercially important variety of S. edule. The yellow-fruited varietal complexes were generally found in backyards. Ownership patterns indicate that the virens levis varietal group was more often associated with male farmers, while female tended to maintain the albus and nigrum groups.

Regarding fruit colour, most accessions exhibited various shades of green, except for the albus group, which had yellow fruits (Figure 5a). Figure 5b shows that all accessions of amarus sylvestris (AW), nigrum spinosum (NS) and albus spinosum (AS) had thorns on their fruits. This is relevant, considering that the wild ancestor S. edule (wild type) is dark green, bitter and densely spiny, suggesting that consumer preference is for smooth fruits and neutral or sweet flavour.

The fruit size of the chayote is highly variable, ranging from very large (15–22cm), medium (6–10cm), and small (3–4cm) (Figure 6a). Regarding flavour, the ellipses show a very noticeable separation, revealing that the fruits of S. chinantlense, S. compositum and amarus sylvestris are bitter, a trait attributed to their tetracyclic triterpenes content, mainly cucurbitacins (Aguiñiga-Sánchez et al., 2015; Aguiñiga-Sánchez et al., 2017; Salazar-Aguilar et al., 2017). Fruits of the albus group have a slightly sweet flavour (7.6–8.1°Brix), while nigrum (4.9–6.4°Brix) and virens groups have a neutral flavour (5.1°Brix) (Figure 6b) (Cadena-Iñiguez & Arévalo-Galarza, 2011; Cadena-Iñiguez et al., 2007).

Cucurbitacins are secondary metabolites of triterpene origin that give a bitter taste to plants (root, guides, leaves, and fruits) and have been recorded in all Sechium species in the BANGESe, although with notable differences in concentration and type of compound (aglycones and glycosides). For instance, the albus group contains one-hundredth of the triterpenes levels found in wild species, and ten times less than nigrum (Iñiguez-Luna et al., 2021; Uriostegui, 2014).

Fruit shape is highly variable and it is difficult to establish a clear grouping pattern (Figure 7); however, the commercially preferred shape is pyriform, which is the shape of virens levis and is the basis of the international standard Codex-Stan-83-993 (pyriform fruits, light green colour, smooth without thorns and grooves, 12–15cm long, 8–10cm equatorial width, 270–310g, neutral flavour, and free of pathogens). The pyriform shape may be a derivation of continued selection by cultivars, as the S. edule fruits of the wild type (ancestor) are small and ovate (Figure 8).

Morphological boundaries

Establishing morphological limits in intraspecific complexes is relatively difficult, since it requires to determine whether distinctive traits are stable and heritable. Variation in plants may be continuous (clinal) or discontinuous, consisting of distinct morphotypes or races (Styles, 1986). In Sechium, variation is clinal since it shows a gradient that allows any character to take values between two extremes (Cadena-Iñiguez, 2005). Environmental factors, including numerous physical, chemical and biological sub-variables, induce this variation by exerting extra pressure on the differentiation process, enabling organisms to express a sufficiently satisfactory plasticity response for survival. For example, in S. edule and S. compositum, fruit shape and size vary according to altitude, with accessions from lower elevations (60m) producing round and small fruits, while those from higher elevations (200m) produce medium pyriform fruits in evergreen forests. Figure 9 shows the morphological limits based on fruit shape, which helps to distinguish the varietal complexes of S. edule, S. chinantlense and S. compositum, based on the colour of the epidermis, spines, shape and size.

Discussion

The preference in the consumption of chayote varietal groups is based on the physical and chemical characteristics of its fruits. In Mexico, the most present in the markets are virens levis, nigrum xalapensis and nigrum spinosum, the latter recently introduced in the United States. For these markets, fruits are harvested at horticultural maturity (Aung, Ball, & Kushad, 1990) and have a watery consistency, neutral flavour and little fibre content. In regional markets, fruit with higher starch content in the endocarp, such as albus, or physiologically mature green fruits, are preferred. This promotes unconscious conservation actions in backyards by women, who determine the type of chayote to be used in the diet. Their decision to either conserve a single morphotype or diversify indirectly leads to reproductive isolation and inbreeding selection, which helps to fix characters and make them stable and heritable or may induce spontaneous crosses (Cadena-Iñiguez, Avendaño-Arrazate, Cisneros-Solano, & Campos-Rojas, 2010).

Women play a key role in the conservation of plant genetic resources mainly through the management of family orchards, where these resources are used for food and sometimes for local sale. Local marketing or exchange of backyard surpluses allows genetic diversity to flow at the regional level (Watson & Eyzaguirre, 2002). When phytogenetic resources are used for human consumption, strong selection pressures arise. For chayote, the preference tends to converge on green fruits with neutral or sweet flavour and low fibre, although preferences for fruits with starchy consistency (albus group) have also been identified (Cadena-Iñiguez et al., 2010).

The first wild populations of S. edule evolved in mesophyll forests, producing dark green, spiny and bitter fruits (Lira, 1996). This type of vegetation is conducive to intraspecific variation in S. edule (Iñiguez et al., 2011). Mesophilic forests, distributed across the highlands of the Sierra Madre Oriental, Sierra Norte de Chiapas and Sierra Madre del Sur, range from 600–3,100m in altitude, with annual precipitation of 1,000-3,000mm, and temperatures ranging between 12–23°C (Ruiz-Jiménez, Téllez-Valdés, Vega, & I, 2012).

Along with natural and artificial selections, migratory routes also cause variations in fruit colour and flavour. When wild populations of S. edule spread to low sub-evergreen forests, greater irradiance led to a change in chlorophyll content (Iñiguez et al., 2011), giving rise to some varietal complexes such as virens levis and those of the nigrum group. In the high plateaus and high valleys, it is common to find thorny green varietal complexes. In medium evergreen and low deciduous forest vegetation types, leaves and fruits began to change in colour and shape, later giving rise to varietal complexes of the albus group, where the fruits have low chlorophyll and cucurbitacin content but high carotenoid and ascorbic acid content (Iñiguez et al., 2011), presumably for photoprotection, since these fruits have high stomata density (Ramírez-Rodas et al., 2021). Also, the derivation of yellow fruits is strongly influenced by selection pressures. (Valenta et al., 2018) indicate that fruit colour plays an important role in seed dispersal and can be considered an evolutionary signal in plants. A contrast between leaf and fruit colour is important for zoochory success. In this regard, (Iñiguez et al., 2011) mention that yellow chayotes are considered the most evolved in terms of their environmental adaptation.

MCA of the fruit size shows that this variable tends to be dynamic. Wild populations of S. edule, S. chinantlense and S. compositum have relatively small fruits (Figure 9) and bitter taste. The latter is related to bi-1 and Bt-1 genes, which confer bitter taste to leaves and fruits (Valenta et al., 2018). Subsequently, with recurrent selection, this characteristic is reduced. In Sechium spp., the bitter taste is influenced by the concentration of cucurbitacins, higher in wild populations, and has applications in pharmacology particularly in cancer treatment (Cadena-Iñiguez et al., 2013; Rosado-Pérez, Aguiñiga-Sánchez, Santiago-Osorio, & Mendoza-Núñez, 2019).

Numerous examples illustrate the role of rural women in conserving plant diversity in home gardens for family food, which in turn supports broader biodiversity. In Indonesia, (Elfrida, Mubarak, & Suwardi, 2020) recorded the conservation of 39 fruit plant species of 23 genera and 17 botanical families by women. Other studies such as (Tefera & Kim, 2019) demonstrated the diversity of medicinal plants conserved by rural families in Ethiopia (52 families and 96 genera). Rural women are critical to conserving local knowledge, encompassing beliefs, medicine, food and economic livelihood. Conservation behaviour often arises from culture rather than formal education. (Ondiba & Matsui, 2021) mention that, from a sample of rural women surveyed, 98% expressed motivation to obtain economic income through acts of conservation, and 88% expressed a commitment to environmental conservation.

(George & Christopher, 2020) recorded the high species diversity in home orchards in Kerala, India; others, such as (Zhang et al., 2020) documented the plants and their local function, highlighting that altitude could be the most important variable that determines the composition of home gardens in China. (Galluzzi, Eyzaguirre, & Negri, 2010) reported the importance of home orchards in safeguarding agrobiodiversity, which reflects a complex structure.

According to the passport data, the variability of Sechium accessions in BANGESe is mainly attributed to the actions of rural women in home gardens or backyards, despite conservation not being their conscious objective. However, and for the reasons documented in the passport data, aspects such as self-consumption, local trade, medicinal and ancestry uses, have fostered a wide agrobiodiversity, mainly determined by these plants’ value as food.

An important aspect that interacts with conservation, agrobiodiversity, bioprospective research, new applications, industry, rural players, home orchards and genebanks, are international treaties. The Nagoya Protocol (CBD, 2011) addresses several of these aspects and seeks, as far as possible, a fair and equitable distribution of benefits to the stakeholders that have safeguarded agrobiodiversity or have enriched collections in genebanks. In this regard, the analysis of passport data is a useful tool for identifying potential beneficiaries for bioprospective research and products over the medium or long term. (Varshney, Spiegel, Zyner, Tannahill, & Balasubramanian, 2020) proposed a 5G approach to conserved agrobiodiversity: Genome of each crop species, Genomically and agronomically characterized germplasm, Gene function identification, Genomic breeding methodologies, and finally Gene editing.

This approach is a desirable strategy for germplasm collections; however, it does not consider the participation of rural stakeholders, whose ancestral knowledge and practices have helped maintain high agrobiodiversity. Therefore, analyzing passport information is key to designing strategies for replacing and regenerating genotypes in ex situ collections, as well as for designing shared conservation models with rural stakeholders. Such models could include morpho-anatomical, phytochemical, genomic and agronomic characterization lines, as well as bioprospective studies to explore new applications and benefits. These efforts could also support descriptor guides for legal registration of new plant varieties or facilitate compliance with international treaties and material transfer agreements.

Wild populations with small, dark green fruits, thorns and bitter taste were the promoters of the morphological diversity that is currently known of S. edule. However, there is also evidence of introgression and hybridization with the species S. compositum and S. chinantlense (Barrera-Guzmán et al., 2021). Conserving this valuable resource begins with evaluating and characterizing genetic variability through morphological and molecular markers, as well as establishing evolutionary relationships among wild, cultivated populations and related species (Cadena-Iñiguez et al., 2007). Ecology studies in conjunction with ecological niche models also provide information on species adaptability, highlighting potential areas for conservation and zones threatened by climate change.

The study of biodiversity is also an incentive for participatory genetic improvement programmes of chayote, fostering collaboration and knowledge exchange between researchers and farmers. This encourages the continuous planting of native varieties or populations in traditional plots, and economic support and public policies are essential to maximize these efforts. Additionally, awareness campaigns are needed to inform the population about the importance of chayote as a plant genetic resource (Aguiñiga-Sánchez et al., 2017).

It is important to highlight that cytological information is missing or scarce in many accessions both in chromosome number and ploidy level as well as nuclear amount. The chromosomal level variegation such as translocation or inversion could explain species incompatibility (Olvera-Vazquez, Cadena-Iñiguez, Gilani, & Watanabe, 2019).

Conclusion

The morphological diversity of S. edule lies mainly in fruit characters, where traits such as bitterness, the presence of thorns and a dark green colour are representative of wild populations. Geographical and cultural richness also influence crop diversification, and the mountainous areas of Veracruz offer the ideal habitat for growth and development of new varieties. Women play a vital role in the conservation of native chayote populations through backyard cultivation. S. chinantlense and S. compositum are species related to S. edule, and from which important advances are being made in the production of cucurbitacins for the pharmaceutical sector.

Data availability statement

To facilitate access to the data from our chayote germplasm collection, information on how interested researchers can get full access is provided below. The data is hosted in our online database, which is accessible upon request. Researchers interested in accessing the full dataset can do so by contacting the corresponding author and providing a brief description of the intended use of the data. In addition, access requests will be reviewed to ensure proper use in accordance with established ethical guidelines for the conservation and research of plant genetic resources. Our goal is to support collaborative research efforts and encourage studies that expand the knowledge and utilization of this important collection of germplasm.

Authors contribution

Jorge Cadena-Iñiguez y Luis A. Barrera-Guzmán wrote the article and performed the statistical analysis of the database; Víctor M. Cisneros-Solano, the curator of the genebank, provided the information on the morphological aspects of the collection; Carlos H. Avendaño-Arrazate, Ma. de Lourdes C. Arévalo-Galarza, Kazuo N. Watanabe and Jorge D. Cadena-Zamudio, who have been developing the database over the years, also contributed to the discussion of this research.

Conflict of interest statement

The authors have no conflicts of interest to report.