Factors affecting ethnobotanical knowledge in a mestizo community of the Sierra de Huautla Biosphere Reserve, Mexico

This study was carried out at El Salto, a rural community located in the southern area of the state of Morelos, Mexico, within the Biosphere Reserve Sierra de Huautla. Its territory belongs to the mountain range known as Sierra Madre del Sur and it is part of the northern extreme of the Sierra de Huitzuco in the state of Guerrero, known as Cerro Frío. It is located at an altitude of 1,785 m between the parallels 18º20′30′ N and 99º17′21′ W, and it encompasses approximately 500 ha [42] (Figure 1).

The settlement is located in a transitional area between tropical deciduous forest and an oak forest [43]. Community surrounding areas include modified environments such as home gardens and farming lands. The dominant climate is semi-warm, semi-humid, with rains during the summer. The total annual rainfall is 924.3 mm and the average annual temperature is 28°C [44].

Inhabitants of El Salto are mestizo. The community was founded by farmers from the southern state of Morelos and by some migrant communities and adjoining villages of the state of Guerrero during the Mexican Revolution (1910–1920). The community is made up of 108 residents that belong to 25 households and each household houses around five people. All families share kinship networks with others, so there is a core of coexistence and knowledge originated from this cause. There are 55 women and 53 men, but children and young people (between 1 and 25 years of age) outnumber adults (60%). Information of 41 inhabitants (21 women and 20 men), ranging between 30 and 70 years of age (average ± SD: 51.4 ± 16.3 years), indicates that about 20% of the informants were born elsewhere, but they had lived at the locality for more than 40 ± 8.6 years. More than three fourths of the informants had finished primary education but 21.9% were illiterate. The festivities in the locality have a religious (mostly Catholic) and civic (school activities) character. The social organization is ruled by community assembly, whose highest authority is the Major’s Assistant.

The main economic activity, practiced by 85.3% of the informants, was subsistence agriculture. Main crops include corn (Zea mays L.), bean (Phaseolus vulgaris L.) and squash (Cucurbita argyrosperma subsp. argyrosperma) in an integrated system. Subsistence agriculture was complemented by a range of activities related to the appropriation of natural resources, such as gathering and hunting (4.8%), production of alcoholic beverages such as mezcal (made from Agave angustifolia Haw.) and “wine” (made from Vitis tiliifolia Humb. & Bonpl. ex Roem. & Schult.) (2.4%) and of products derived from cattle, such as milk and cheese (9.7%).

As much as 70.3% of the sample received economic support from the government, this support represented about 20% of their total income and therefore is an important complement to their main economic activities. The average monthly earnings per capita estimated for 2007 were $USD 133.98 ± 57.57. Nevertheless, 44% of the people were exclusively living on remittances coming from the US. Rates of youth migration to the US, mainly after the young people have completed their high-school, are high.

Between September 2005 and December 2006, we conducted open-ended and semi-structured interviews [45, 46] with all the household heads (n = 41: 21 women and 20 men). To avoid overestimating the knowledge about plants that might have been acquired elsewhere, we selected people older than 30 years old e.g., [47] who had been living in the study site for a minimum of 30 years [48, 49]. Interviews were carried out individually to prevent distortions due to the presence of a third person [48, 49]. The interviews focused on two aspects a) determining their knowledge of the plant resources (wild, cultivated and weedy species) in their community and b) characterizing the socio-economic conditions of the person being interviewed.

First, to quantify the degree of knowledge of plant resources and to identify the use value of each plant species, we conducted free listings by category of use [46]. The survey only included the theoretical dimension of ethnobotanical knowledge sensu[50]. Women’s and men’s knowledge of the plant richness of their community was defined as the number of species mentioned at the time of the interview. Second, we also asked about the following socio-economic information: 1) age, 2) gender, 3) education level (with or without primary studies); 4) origin (local or migrant), 5) main productive activities (farming or stockbreeding), 6) monthly expenditures, as a proxy for monetary earnings, 7) amount received in remittances, and 8) other economic activities (related to temporary earnings perceived as support from governmental institutional projects).

To determine the use-value of the local flora, we calculated the use-value (UV) index using the algorithm proposed by Phillips and Gentry [21, 30], modified by Rossato et al.[51] and Lucena et al.[52]. The calculation was obtained by counting all the uses mentioned by every person for a specific plant and dividing the result by the total number of informants. The use-value corresponds to the average use associated to each species in a specific community:

where U _is is the number of uses mentioned by an informant i, for each species s and n is the number of informants interviewed for each species.

We classified the local environment into five habitats: tropical deciduous forest, oak forest, riparian vegetation, home gardens and farming lands. The first three habitats correspond to vegetation types [43] and the other two to modified environments [53]. We visited each habitat repeatedly, with a different informant each time, to photograph and inventorying useful plants mentioned during the interviews and free listings. Subsequently, in a group meeting using an image projector, a photograph of each plant was showed to all interviewees to reach a consensus and verify that this was the correct etno-specie and avoid confusion by popular synonyms [48, 49]. Then, after the identification of plant species was completed, we estimated the richness as the number of plant species per use category per habitat. The habitat of one species was determined based on Rzedowski [43]. The identification of botanical material was carried out at the HUMO Herbarium of the Universidad Autónoma del Estado de Morelos and with experts from the MEXU Herbarium of the Universidad Nacional Autónoma de Mexico. Voucher specimens were deposited at the HUMO and the folio number for each species is shown in (Table 1).

To determine the association of socio-economic factors and ethnobotanical knowledge, as measured in the described interviews, we used a non-metric multidimensional scaling analysis (NMDS). This technique is appropriate for non-normal discontinuous data, such as the data used in this study. In our analysis, answers from the interviews were used as external variables to interpret ordination. Both the continuous variables (i.e., remittances, age, monthly expenditures), and the nominal variables, (i.e., main productive activities, education level, gender, origin, and other economic activities -earnings from a governmental source and origin of the person) were adjusted on the ordination. Two matrices were used to carry out the multivariate analysis. The main matrix included the useful species registered during open-ended and semi-structured interviews. The variables appear in rows and the interviewed people in columns. The secondary matrix included the socio-economic factors in rows and the interviewed people in columns.

The result from the analysis allowed for the representation of vectors as arrows that point in the direction towards which the variable being assessed changes the most; this is called direction of the gradient. The arrow’s length is proportional to the correlation between the ordination and the assessed variable, and it is called force of the gradient. The analysis estimates a value for r², which represents the goodness of fit of the vector. Significance or p-value is based on the random permutations of the data. In the case of the factors, the analysis also estimates a value for r² as goodness of fit, and a p-value that allows to test the significance of the factor on the ordination. The graphic representation included the main groups’ centroids. Only the vectors and factors that turned out to be significant (p < 0.05) were graphically represented. The “Vegan” module [54] from the R program [55] was used for the statistical analyses. The difference in terms of the knowledge of the number of species between the genders was assessed for the six main use categories (Table 1).

Paired t-tests by household and a Wilcoxon signed rank test were used to determine if there were differences between men and women in terms of the number of species they knew by use category and in terms of the total number of species known. The paired difference distribution of paired t-test was analyzed by means of the Shapiro–Wilk’s W normality tests.

A total of 185 species, belonging to 149 genera and 69 families (Table 1), were recorded. We distributed those species in 12 use categories. We found a total of 310 different uses for the 185 species; thus, the richness of use is greater than the richness of species since some of the species had more than one use (Table 2). The richest families were Asteraceae and Fabaceae and the richest genera were Bursera and Tagetes.

According to their life form, the greatest proportion of useful plants registered at the locality included herbaceous (47%), arboreal (38%), shrubby (11%) and climbing (4%) species. The most used plant structures comprised the stem (28%) and the complete plant (21%), but there is no consistent pattern across the use categories (Table 1). The distribution of the richness of species use per habitat was consistent across habitat only in the case of the medicinal and edible species, the two most common uses. The ranking arrangement of the other use categories changed according to the habitat; this could be related to the availability of the species in each habitat (Table 2).

Tropical deciduous forest was the habitat with the highest proportion of useful species known to the sample, as it is shown by the greatest richness value of the useful plants and the greater diversity of uses for this habitat (Tables 2 and 3).

We defined three groups according to the species use-values (Figure 2). The first group (A) comprises three multi-purpose species (4 to 6 uses), wit use-values >1.5. Species in this group correspond to trees appreciated for their timber: tepehuaje (Lysiloma acapulcense (Kunth) Benth; 2,59), yellow oak (Quercus magnoliifolia Née; 2,54) and palo dulce (Eysenhardtia polystachya (Ortega) Sarg.; 1,95). The second group (B) comprises 39 multi-purpose species with use-values between 0.34 and 1.37. Species in this group range from one (Rosa chinensis Jacq.; 1,37) to six uses (Quercus castanea Née; 1,15), and many are used for medicinal purposes (11 species), timber (10 species), food (10 species), fuel (9 species), mystical-religious ceremonies (6 species), ornamental purposes (5 species), and commerce (2 species). The last group (C) comprises 143 species with use-values between 0 and 0.29, which show a distinct low diversity of uses per species. In fact this group includes only one taxon used for five purposes (Ipomoea murucoides Roem. & Schult.; 0,02). Species in this group are mainly used for medicinal needs (76 species), timber (55 species), food (35 species), religious activities (15 species), ornamental and fuel (12 species), and commercial activities (2 species). Four species on the group are toxic.

Results from the NMDS analysis and subsequent adjustment of socio-economic variables suggested that the variables age, gender, farming and stockbreeding were associated in a statistically significant way to the knowledge of plant resources (Figures 3 and 4, Table 4). In contrasts, education level, origin, and the three others variables related to economic status (i.e., monetary earnings, remittances, governmental monetary support) were not associated to the ethnobotanical knowledge of a person (Table 4).

The statistical analysis showed a clear spatial separation between two groups due to the differences in terms of men’s and women’s knowledge (Figure 3). Men from El Salto mentioned an average of 53.0 ± 10 (mean ± standard deviation) plant species, whereas women referred to an average of 38.9 ± 6 plant species. The difference was statistically significant (Student’s test, T₁₅ = 4.8, p < 0.001). However, when analyzing the indication of plant species according to the general use category, we found that women significantly mentioned more ornamental and mystical-religious plants species than men (Table 5). In contrast, men mentioned significantly more plant species used for building houses or fences, crafts, farming wooden tools, firewood, domestic wooden tools, fodder, poisons, and the commercialization of wild plants than women (Table 5). No differences were found between men’s and women’s responses in relation to edible or medicinal plants (Table 5).

Species used at El Salto represent only 2.6% of the useful plants previously reported by Caballero and Cortés [9] for peasant –indigenous and mestizo– communities in Mexico. However, if we compare our data with the number of useful species reported by Bye et al.[56] for a larger territory comprising 12 ejidos near the region of Chamela, Jalisco, Mexico, the number of useful species at El Salto is 12.4% higher than the number in those ejidos (185 vs. 162 species, respectively). This difference is more remarkable given that the ecological conditions are similar at both sites, suggesting that the size of the site is not the only factor that determines the number of useful species known by its inhabitants. Rather, historical, cultural and socioeconomic traits are essential to understand the local knowledge of useful plant diversity in a particular study area [57].

The plants used at El Salto include some botanical families (Asteraceae, Fabaceae, Burseraceae, Lamiaceae, Verbenaceae, Euphorbiaceae, Anacardiaceae and Solanaceae) which play an important role to satisfy local needs, as well as some considered of uppermost importance at the national, state, and regional levels [9, 22, 57]. While these families contributed many useful species, the highest percentage of species listed came from a wide range of families, as follows: 7.3% of the families contributed three species each, 17.4% of the families contributed two species each, and 56.5% of the families contributed one species each. The high incidence of new uses of plants reported in the community could be accounted by the occurrence of many rare species (n = 136 species), which in turn could be explained by the presence of two major types of vegetation in the area: tropical deciduous and oak forests. In terms of the life forms of the resources, there is a high variation within each use category. Furthermore, our results indicate the preponderance of herbaceous species over other life forms, especially in modified environments (i.e., home gardens and farming lands); this fact is in agreement with the pattern reported for Mexico [9].

When analyzing the use value of plants, we find a predominance of medicinal uses. Moreover, this category is the one that displays more uses per species. This result appears to be constant among mestizo communities in the country, as it has been suggested by Bye [58] on a review of case studies among mestizo and indigenous groups in Mexico. Furthermore, this author has pointed out to differences in terms of the use of alimentary plants among the social groups mentioned: edible plants are of immediate concern for the indigenous societies while they only play a secondary role for mestizos.

We also found that people recognize a greater richness of species and diversity of uses in wild habitats than in modified environments. This finding contrasts with some studies that show that home gardens and farming lands harbor a greater biological diversity than the one registered in the wild environment [53, 59]. Altieri et al.[53] show that tropical agroecosystems can contain more than 100 species; whereas Pulido et al.[59] point out that a family orchard with an average extension of 0.5-2.5 hectares located in the tropical deciduous forest holds a diversity comprising 92 species (61% of them are native to the area). The differences could be methodological, as our study refers to species recognized, whereas the other studies are based on plant inventories. The differences could also be explained by the existence of a direct relation between the relative diversity of species in modified environments and the availability of irrigation water, as was described by Villa and Caballero [60]. Taking into account this relation, we hypothesize that the extended dry season that characterizes the region inhabited by the community of El Salto severely limits the amount of irrigation water, which in turn restricts the diversity of useful plant species that can be maintained in managed environments.

The results from this study reveal the importance that wild ecosystems have for mestizo communities in terms of the development of basic rural subsistence activities in dry tropical areas. However, our research also shows that both, wild habitats and artificial environments, are valuable to understand a group’s ethnobotanical knowledge.

The highest use-values among species from groups A and B were registered in relation to timber, typically employed for house-building, firewood, and the manufacturing of farming tools, crafts, and household possessions. At present, these activities are not so frequent among the studied population, given the time and energy needed to manufacture products. However, even though these uses are less frequent, their knowledge persists, a situation that is similar to what Byg and Balslev [15] show for the use of palm species among the Shuar in Ecuador. For example, people maintain a body of knowledge related to timber species differentiating between spongy wood (hollow and brittle), solid wood (useful to manufacture farming tools), or wood that will become “good hot coal”, i.e., firewood that lasts longer ignited [57].

Even though groups B (i.e., UV from 1.37 to 0.34) and C (i.e., UV from 0.29 to 0.1) displayed the greatest floral diversity (182 species), there was a variation in terms of the multi-purposefulness and the number of uses was not stable. There was a tendency towards the use of timber yielding species in group B and of medicinal species in group C, suggesting that group B is probably the most important socially as it comprises the greatest number of species with the highest number of applications.

It is worth noting that none of the medicinal species displayed a high rate of use-value. In other words, the use of the most important plants for the community (i.e., the ones with highest use-value) is not related to the importance ranking by category of dominant use at the site. Such situation could be explained due to the multi-purposefulness that characterizes most species, since their inclusion in different use categories increases their potentiality, in other words, their use-value is enhanced while their exclusiveness for a specific use category decreases. Such speculation calls for a more complex analysis, since it is known that medicinal plants are culturally preeminent among mestizo communities [56].

Our findings suggest that the knowledge of plant resources is associated mainly to socio-economic activities, age and gender, which is consistent with other ethnobotanical investigations [6, 12, 15, 25]. In terms of the socio-economic aspects, despite the fact that occupation, i.e., farming and stockbreeding, was significantly associated to knowledge, it only accounted for 20% of the variation in the ethnobotanical knowledge. This suggests that there are other factors, not included in the analysis, which could influence ethnobotanical knowledge. Such other factors might entail cultural aspects such as ideological structures, ceremonies, significance and classification systems, production techniques and practices [5, 8, 14, 27, 37], or ecological ones, which have historically been poorly explored –e.g., density of useful species, floral heterogeneity at the site, dominant biological forms, altitudinal variations, types of vegetation, selective floral and fruit morphology and phenology– [16, 22, 26, 28, 52, 61, 62].

Farming and stockbreeding constitute common activities that seem to provide a particular contribution to ethnobotanical knowledge. Some studies show that conducting primary activities contributes to use and management of natural resources [25, 27]. The relation between animal rearing and ethnobotanical knowledge can be explained through several examples in this study. Thus, as livestock rearing constitutes the settlers’ main activity, there is a need for them to know the plant resources that are helpful in the treatment of cattle’s gastrointestinal diseases. This knowledge is based on the observation of the animals’ alimentary habits with respect to wild and fodder plants, as well as on the detection of the toxic species that are eliminated from the environment to avoid that cattle consume them. Ethnobotanical knowledge also allows farmers to use alternative fodder in times of economic shortage or draught and also contributes to livestock health through the prevention of common diseases. Farmers are also familiar with the species that are used to craft farming tools, for house-building activities (with specific traits such as resistance, flexibility, duration and pliability, as living fences and as tutors). Some species are also tolerated due to the benefits they offer such as shade, medicine and food [57].

The informants’ age also seem to be associated with ethnobotanical knowledge. Older people knew more useful plant species than younger people, probably because ethnobotanical knowledge tends to accumulate through the life cycle, as has been found elsewhere [12, 15, 27, 28]. Garro [63] indicates that aging is naturally associate with the process of knowledge acquisition as the pass of time help individuals accumulate knowledge and experience. Furthermore, in some studies, age seem to be the only variable associated with knowledge [64], although some other authors have found no association between age and knowledge [15]. Although most studies highlight knowledge differences between young and old people, our results suggest that ethnobotanical knowledge continues to accumulate after 30 years of age.

We also found differences between men’s and women’s knowledge in relation to the plants they use at the interspecific and the intracategorical levels, as has been pointed out by other authors e.g., [6, 19, 28]. Women’s knowledge, in terms of the proportion of useful species they know, is closely associated to the treatment of diseases, the use of plants that embellish their household and of those related to rituals. The knowledge displayed by men is more diverse, since it includes a greater number of species used because of their wood quality to produce crafts and farming tools, to build houses, as fuel and household possessions, as well as species used as cattle fodder. Men are also more knowledgeable than women about plants that can cause bodily harm (swelling and irritation) during working days. Working with Raramuri indigenous people, Camou-Guerrero et al.[6] found that women had higher knowledge of medicinal and edible plants than men. As other authors [12], we did not find such differences in the community of El Salto. In sum, our results suggest that gendered division of labor within the family has resulted in constant interaction with the resources corresponding to specific activities. This phenomenon has determined how different species have acquired cultural importance for a specific gender in different cultural contexts [2, 65].

Other factor, such as the level of formal education, origin, and economic variables (income, remittances, and subsidies) are not associated to ethnobotanical knowledge. Previous research with young people has shown that the level of formal education bears a negative association with ethnobotanical knowledge [66], probably because time invested in schooling deflects from time invested in ethnobotanical knowledge also generating a lack of interest on the environment [26]. In contrast, Godoy [67] notes that formal education can lead to practices of use of more sustainable resources and the “environmental awareness”. In most of the studies it has been found that education is associated with the loss of language and ethnobotanical knowledge in indigenous communities and of mixed origin (mestizo-indigenous) [27]. If the loss of indigenous language is the main factor that drives the loss of ethnobotanical knowledge, this could help explain why in the studied Spanish-speaking mestizo community we do not find the expected negative association between schooling and ethnobotanical knowledge.

We did not find significant differences in ethnobotanical knowledge between people from the area and outsiders. The finding dovetails with what was reported by Byg and Balslev [15], who in their study in Ecuador find no differences between the ethnobotanical knowledge of indigenous peoples and colonists. In our case study, this could be due to the fact that people who have migrated to El Salto are coming from areas with the same type of vegetation and productive activities, and to the fact that most people migrated to the area during childhood.

Income, remittances and subsidies are all economic indicators of family well-being, which have also often been related to the loss of ethnobotanical knowledge, as income allows people to access market goods that substitute plant-made products [15, 34]. However, we did not find such a relation. We argue that this could be due to the lack of large differences in the sample. In most cases families depend on remittances and subsidies, which are then invested in primary activities. Reyes-García et al. [68] show that conducting forest and farm activities is associated with greater ethnobotanical skills and with greater theoretical ethnobotanical knowledge, even if those are market oriented, thus implying that some forms of economic development can take place without eroding local ecological knowledge.

Although the variables presented have been analyzed independently, they do not act in independent way, or always have a linear relation with ethnobotanical knowledge. The acquisition of ethnobotanical knowledge is a complex process and we can not assure that the variables analyzed are the only direct drivers of the transmission and acquisition of this knowledge. We suggest that the generation of ethnobotanical knowledge should be understood as a dynamic social process, driven by the current interaction with the ecosystem given the importance of multiple socioeconomic and cultural factors [69].