Introduction

The knowledge and use of native species is not homogeneous. Rather, people tend to use certain species according to their preferences (Ghorbani et al. 2012; Cruz et al. 2014). Research on native plants highlights the importance of ecological factors such as accessibility, availability and abundance as influential in the selection of species (Phillips and Gentry 1993; Ladio and Lozada 2003; Ladio et al. 2007). Some researchers assume that common species that are frequently found near residences are more likely to be chosen (Albuquerque and Lucena 2005; Ladio et al. 2007).

In addition to environmental factors, versatility of species, i.e., the number of uses it can be put to, influences selection (Etkin 1997; Ogle et al. 2003; Termote et al. 2011. For example, studies conducted in Spain and Vietnam found that the use of a species for food is often continued because it is also used for medicinal purposes and/or as a source of income (Ogle et al. 2003; Rivera et al. 2007). Moreover, studies evaluating the overlap between food and medicinal uses have found that the use of food species for the prevention and treatment of illnesses is the result of biocultural adaptations (Etkin and Ross 1982).

Despite evidence of the importance of local availability and versatility of use in selecting a species, it remains unclear how the overlap between certain uses, such as food and medicines, can influence selection. Accordingly, our study focuses on preferred food species for three communities in northeastern Brazil to test the following hypotheses: 1) the preference for native food species is related to the number of uses for which species are cited; 2) the categories of food and medicinal use equally influence the selection of a native food species; and 3) the frequency and abundance of native food species in the collection area relate to the level of preference.

Study Area

This study was conducted in three rural communities near the Araripe National Forest (FLONA), in the southern region of Ceará, northeastern Brazil. The FLONA covers an area of 38,493.00 ha, distributed among the following vegetation types: cerrado (stricto sensu) and cerradão (Ribeiro and Walter 2008), carrasco (dry forest) and seasonal semi-deciduous forest (rainforest) (IBAMA 2004). Founded in 1946, the FLONA was the first conservation unit of sustainable use in Brazil. According to the Koppen classification system, the climate is humid tropical, with an average annual rainfall of 1090.9 mm, an average temperature of 24–26 °C and rainy seasons between December and April (Cavalcanti and Lopes 1994; INMET 2013).

The FLONA includes the municipalities of Barbalha, Crato, Jardim, Missão Velha and Santana do Cariri (IBAMA 2004). Near the municipalities, around the Araripe National Forest, are a number of communities that, although they practice agricultural activities and migrate at certain times of year to different regions of the country in search of work opportunities, have a strong link with native flora, which remain important in their diet, in the treatment of certain diseases and as sources of income (Sousa-Júnior et al. 2013; Lozano et al. 2014).

For this study, we selected three extractive communities located in three different municipalities. Baixa do Maracujá, in the municipality of Crato, lies approximately 5 km from the Araripe National Forest, and has a population of 357 among 120 households, generally engaged in agroforestry. Horizonte, in the municipality of Jardim, has a population of approximately 1120 people among 210 households, and is located approximately 10 km from the FLONA. Horizonte is a purely extractive community. Macaúba, in the municipality of Barbalha, has a population of 610 among 180 households, generally engaged in the extraction of babaçu palms (Attalea speciosa). The community is located 2 km from the FLONA (Fig. 1). The three communities are organized around social institutions, including churches, schools, health centers and Residents Associations.

Fig. 1
figure 1

Map of the study sites Baixa do Maracujá, Horizonte and Macaúba adjacent to the Araripe National Forest, Northeast Brazil

Full size image

Methods

A total of 317 informants for this study were identified from a previous study of native food plants in the three communities (80 in Baixa do Maracujá, 122 in Horizonte, and 115 in Macaúba) (Campos et al. 2015). Individuals were then selected based on their knowledge of native food plants, which we assessed from the quartile distribution of the number of food plants cited by each respondent, adopting the third quartile as a reference. Thus, respondents who cited eight plants or more were selected to participate in this research, for a total of 78 informants among the three communities (20 in Baixa do Maracujá, 31 in Horizonte, and 27 Macaúba).

The species selected for this study were those identified with the Anthropac program (Borgatti 1996) as having greater cultural salience in a previous study (Campos et al. 2015). From the salience analysis results, 18 food native species were selected among the three communities (Table 1).

Table 1 Native food species used by the three study communities
Full size table

We prepared cards with images showing the size, flowers and fruits of all 18 species selected (Fig. 2), which were used primarily to confirm that the researcher and the respondent were referring to the identical taxonomic unit (Medeiros et al. 2014). During interviews, the specialist informants were asked to rank the species according to their preferences. They were then questioned about the motives for their choices and the chief uses for a species (Albuquerque et al. 2014). The preference for a species is linked to its utility; therefore, all categories and specific uses were recorded (Table 2).

Fig. 2
figure 2

Images of the 18 native food species. a Psidium laruotteanum; b Psidium guineense; c Psidium sobralianum (Photo: Alejandro Lozano); d Psidium myrsinites; e Eugenia sp; f Myrciaria sp.; g Mouriri pusa; h Hancornia speciosa; i Caryocar coriaceum; j Anacardium microcarpum; k Attalea speciosa l Anonna coriacea; m Hymenaea courbaril; n Syagrus cearensis; o Acrocomia aculleata; p Hymenea stigonocarpa (Photo: Marcelo Kuhlmann); q Passiflora silvestris (Photo: Maria Clara Cavalcanti) r Passiflora cincinnata

Full size image
Table 2 Criteria for classification of a native food species
Full size table

To refine the data collected and to determine the most important criteria for the classification of the 18 preferred species, a participatory workshop was conducted using the technique of “matrix of criteria and options” (Sieber et al. 2014). Participants assigned values on a scale from zero to 10 to the different categories of use identified for each species according to its importance, with the most important receiving the maximum value of 10. The workshop was conducted independently for each of the three communities, and all local specialists previously interviewed were consulted (14 from Baixa do Maracujá, 27 from Horizonte, and 12 from Macaúba).

Data on Vegetation Structure

The areas selected for the survey of vegetation structure were those indicated by informants as important locations for gathering. To verify whether the frequency and abundance of food species correlated to their preference status, 2 ha of vegetation, classified as “core-areas of collection,” were sampled. Within this area, 1 ha was cerradão, 0.5 ha was cerrado stricto sensu, and 0.5 ha was humid forest. In these core-areas, four plots of 0.5 ha were demarcated, and all plants with diameter at ground level (DGL) ≥ 3 cm were measured. The frequency and relative density for all species sampled in the study were calculated (Araújo and Ferraz 2014). All species in the plots were collected for later botanical determination and incorporation in the collection of the Herbarium Vasconcelos Sobrinho (PEUFR) of the Federal Rural University of Pernambuco (Universidade Federal Rural de Pernambuco -UFRPE).

Data Analysis

To determine the most preferred food species, a quantitative analysis of salience was conducted using Anthropac 4.0 software (Borgatti 1996). In this study, the most salient species were considered the most preferred. The salience results were used to perform the analysis of the main factors that influence the selection of these species. Using the vegetation structure measures, the data collected were processed and the basal area, frequency and relative density calculated for all species sampled in the study (Table 3).

Table 3 Statistical analyses used to test the hypotheses regarding the criteria linked to the preference for native food species
Full size table

Results

In relation to our first hypothesis that preference for a food species is related to the overall number of uses, and we found highly significant correlations between the preference for a food species and the number of individual uses listed for that species for each community: Baixa do Maracujá (rs = 0.6, p = 0.04), Horizonte (rs = 0.68, p = 0.01) and Macaúba (rs = 0.63, p = 0.01) (Table 4). Our results also supported our second hypothesis that the use categories Food and Medicinal have equal influence on the selection of a preferred species (Table 5).

Table 4 Number of food species, and their other uses for the study communities
Full size table
Table 5 Average scores of the different categories of use determined with the “decision-matrix” method to rank the food species preferred by informants
Full size table

For the third hypothesis, we found that the preference for a food species is strongly correlated with its relative density (RD) and relative frequency (RF) in the communities for which the primary collection site is the FLONA: Horizonte (rs = 0.6, p = 0.03) and (rs = 0.62, p = 0.03) and Macaúba (rs = 0.8, p = 0.0008) and (rs = 0.78, p = 0.02), respectively. For Baixa do Maracujá, a community with agroforestry plantations as the main collection site, no correlation was found between preference and relative frequency (rs = 0.4 p = 0.15) or relative density (rs = 0.4 p = 0.1).

Discussion

Our results confirmed our hypothesis that preference for a food species relates to the number of uses for it. It is important to note that medicinal uses are the most common after food. We argue that two processes may be important to explain this finding. Firstly, the more uses a species has the more likely it is to be remembered and the knowledge passed down over generations. Secondly, the more easily a species is located and the more abundant it is, the more likely it is to be collected. Ladio and Lozada (2003) observed that a cost-benefit ratio would account for this. Thus, the collection of species with multiple uses can be seen as an adaptive strategy. We refer to these processes as the optimization of uses and the optimization of resource collection.

Studies conducted specifically with medicinal plants have found that in arid and semi-arid environments people generally access a set of highly versatile species, i.e., species that can be used to treat a range of ailments (Albuquerque and Oliveira 2007; Cartaxo et al. 2010; Ribeiro et al. 2014). Further, Albuquerque et al. (2009) argue that this multiplicity of uses may be the result of an optimization process in response to the significant climatic variations in these environments that affect the temporal and spatial distribution of many species. Some researchers have argued that individuals tend to remember information with potential adaptive value (Lavie 2005; Nairne et al. 2009; Broesch et al. 2014), and there is also evidence that humans tend to classify natural resources according to their utility (Hunn 1982).

The hypothesis that food and medicinal categories equally influence the selection of species as preferred was also confirmed by our results. Several studies have addressed the overlap between these two use categories (Etkin and Ross 1982; Ogle et al. 2003; Rivera et al. 2007; Cruz et al. 2014). Although these studies do not provide information to explain the factors that may be linked to this overlap, we argue that it occurs mainly because both food and medicines are closely linked to human survival. The same species can be treated as food or medicine simply depending on the amount and/or parts consumed and the method of preparation. Certain species, when added to the diet, are important not only in providing nutrients but also in the maintenance of health. For example, Jennings et al. (2014) observed that the seeds of a species of Fabaceae, widely used in food as a spice, are also used in natura to treat stomach problems and control diabetes.

While Etkin (2006) notes the difficulty of ascertaining the initial use of a species, some studies, using both historical and current data, describe how some native plants were first used as medicines in certain communities (Jennings et al. 2014; Toletin 2014). Our results found that Hancornia speciosa, popularly known as mangaba, is currently more widely used as a medicine to prevent the onset of diseases related to circulatory and gastrointestinal systems than as food. However, in the past mangaba was an important source of fruits used for in natura consumption and for vitamin-rich juices. The current predominantly medicinal use of H. speciosa is because of the latex extracted from the stem. This reported change might be evidence of local adaptation to the seasonal availability of mangaba (Albuquerque 2006; see also Molares and Ladio 2009).

The hypothesis that the most preferred species are also the most frequently occurring and abundant was confirmed by the two study communities most engaged in native resource extraction in the FLONA - Horizonte and Macaúba (see also Phillips and Gentry 1993; Ladio and Lozada 2004; Albuquerque and Lucena 2005). Other researchers have also noted that communities will favor preferred species through active landscape management (Voeks 1996). However this hypothesis was not supported for the agroforestry community of Baixa do Maracujá, since residents rarely collect native plants in the forest but rather cultivate favored species in their home gardens (see Kumar and Nair 2004).

Conclusions

This study provides evidence that the process of selecting a native food plant may be influenced by other uses of the species and by its abundance and ease of harvesting. Our results indicate that human populations tend to optimize both their ecological knowledge and their time and effort in harvesting and that these considerations lead to preferences for particular species. In addition, different livelihood strategies can influence the choice of a preferred species. The use of native plants may change over time, but medicinal use consistently has a strong association with food use, as reported elsewhere in the literature.