CATFISH AND MULLETS: THE FOOD PREFERENCES AND TABOOS OF CAIÇARAS (SOUTHERN ATLANTIC FOREST COAST, BRAZIL)

Natalia Hanazaki and Alpina Begossi

Natalia Hanazaki. Ecologist. Doctoral degree in Ecology, Universidade de Campinas (UNICAMP), Brazil. Professor, Universidade Federal de Santa Catarina (UFSC), Brazil. Endereço: ECZ/CCB/UFSC, Campus Universitário Trindade, Florianópolis SC, Brazil, 88010-970. e-mail: natalia@ccb.ufsc.br

Alpina Begossi. Ecologist. Ph.D. in Ecology, University of California, Davis, USA. Researcher, Museu de História Natural, UNICAMP, Brazil. Endereço: NEPAM/UNICAMP, C.P. 6166, Campinas, SP, 13081-970, Brazil. e-mail: alpina@unicamp.br

Summary

An important feature of human diets is the establishment of food categories according to its preference or avoidance. The literature on fish food preferences points out a general pattern of preference for fish with scales and an avoidance or prohibition of scaleless fish. The food preferences, avoidances, and taboos on animal protein items among three caiçara communities from the Southeastern Brazilian coast were analyzed. Two aspects of the caiçaras choice of food were discussed, regarding their preferences, avoidances, and taboos on the consumption of catfish (Ariidae) and on different species of mullets (Mugilidae). This categorization can be explained both ecologically and culturally: through the environmental availability of the species, their position in the food web, or their importance in the economy and in the social relations within the community. A connection between resource conservation and food taboos about certain fish species are unlikely when compared to the possible relations between hunting taboos and conservation of the terrestrial fauna. For the studied communities, there was no nutritional deprivation resulting from food taboos on fish species.

ARIIDAE Y MUGILIDAE: LAS PREFERENCIAS Y LOS TABÚES ALIMENTARIOS DE LOS CAIÇARAS (COSTA ATLÁNTICA MERIDIONAL, BRASIL)

Resumen

Una característica importante de dietas humanas es el categorizar de alimentos según su preferencia o evitación. La literatura en preferencias alimenticias de los pescadores contempla un patrón general de preferencia por pescados con escamas y una aversión o prohibición de pescados sin escamas. Se analizaron las preferencias, evitaciones y tabúes sobre el alimento de origen animal en tres comunidades caiçaras de la costa brasileña del sudeste. Se discuten dos aspectos de la escogencia de alimento de los caiçaras, con respecto a sus preferencias, las aversiones, y los tabúes en el consumo de Ariidae y de varias especies de Mugilidae. Esta clasificación se puede explicar por factores ecológicos y culturales, como la disponibilidad ambiental de la especie, su posición en la cadena alimenticia, o su importancia en la economía y en las relaciones sociales dentro de la comunidad. Una conexión entre la conservación del recurso y los tabúes sobre ciertas especies de peces parecen más tenues que las posibles relaciones entre los tabúes de caza y la conservación de la fauna terrestre. Para las comunidades estudiadas, los tabúes sobre peces no resultaran en deficiencias nutricionales en la población humana.

BAGRES E MUGILÍDEOS: PREFERÊNCIAS E TABUS ALIMENTARES DE CAIÇARAS (LITORAL SUDESTE DA MATA ATLÂNTICA, BRASIL)

Resumo

Uma importante característica da dieta humana é a categorização de preferências e aversões alimentares. O objetivo deste artigo é analisar as preferências e tabus alimentares com relação a itens de proteína animal entre três comunidades caiçaras do litoral sudeste brasileiro. A literatura sobre preferências alimentares quanto ao consumo de pescado aponta para um padrão geral de preferência por peixes com escamas e aversões ou restrições de consumo por peixes sem escamas. Neste estudo, visamos discutir duas particularidades das escolhas alimentares entre caiçaras, especificamente quanto às preferências e aversões sobre o consumo de bagres (Ariidae) e sobre diferentes espécies do gênero Mugil (Mugilidae). Estas preferências e aversões alimentares podem ser explicadas por fatores ecológicos e culturais, como pela disponibilidade do recurso, pela posição da espécie na cadeia alimentar, ou através da importância destas espécies na economia e nas relações sociais dentro da comunidade. Relações entre conservação e tabus alimentares sobre certas espécies de peixe, ou grupos de espécies, parecem ser mais tênues do que as possíveis relações entre tabus de caça e conservação da fauna terrestre. Além disso, os tabus sobre espécies de peixes não resultam em deficiência nutricional nas comunidades estudadas.

KEY WORDS / Artisanal Fishery / Caiçara / Food Habits / Mugil gaimardianus / Taboos /

Received: 03/24/2005. Modified: 12/07/2005. Accepted: 12/16/2005.

Introduction

An important characteristic of human diet and food habits is the possibility of categorizing the food as edible or not, and as preferred or avoided (Messer, 1984; Roosevelt, 1987). Some food practices have a biological basis, such as the preference for sweet and the avoidance of bitter food (Armelagos, 1987). Besides this biological heritage, food preferences, avoidances and taboos may be the result of sociocultural and economic relations within a social framework (Grivetti, 1978; Simoons, 1978; Walker, 1995; MacBeth and Lawry, 1997). Such categorization of food items can be analyzed with complementary "emic" (or the explanation given by local people) and "etic" (or the explanations given by an outsider) perspectives (Zent, 1996; Begossi, 1998). According to Harris (1987a) the emic perspective includes idealist and mentalist approaches and the etic perspective includes the materialistic, biopsychological and behavioral approaches. In this sense, food preferences, avoidances and taboos may be the result of both symbolic and materialistic factors, being explained through both etic and emic factors.

Prohibited food items, or food taboos, may be followed by all members of a community, or may be applied to a specific age group, gender, or time period (Rea, 1981; Colding, 1995). Taboos in this latter category are defined as segmentary taboos, and can play a role in the environmental conservation of the resources that are forbidden (Colding, 1995). For example, certain foods are forbidden during pregnancy or breast-feeding, even in urban communities (Eaton-Evans and Dugdale, 1986; Harris, 1987a; Coronios-Vargas et al., 1992).

Among fishermen populations it is common to observe taboos on certain species in specific situations, such as during illness, pregnancy or lactation (Smith, 1981, 1996; Madi and Begossi, 1997; Begossi, 1998). Several factors contribute to the emergence, adoption, and persistence of food taboos, particularly the fat content of the avoided species, how perishable it is, or the potential to accumulate toxins due to its position in the food web (Begossi, 1998; Begossi et al, 2004), as well as the social context of production of the fish resource (Pálsson, 1991). Many food taboos are emically related to bad luck or bad health, although the development and persistence of certain food habits is a combination of biopsychological, infrastructural, politic-ideological and economic factors (Harris, 1987a).

Tabooed fish are emically defined as reimoso, "heavy" or "hot" foods (Begossi, 1998). According to Maués and Motta-Maués (1977) the edible food items can be reimoso when that food "faz mal" (is deleterious). Only inhabitants in good health can consume such foods. On the other hand, foods called "mansa" (tame) are considered harmless to any health condition (Maués and Motta-Maués, 1977). The classification of foods as "hot" or "cold" is related to food preferences and avoidances in specific situations, and is associated with European concepts such as the Hippocratic humoral theory (Queiroz, 1984; Fleming-Moran, 1992; Strathern and Stewart, 1999).

This paper aims to analyze the food preferences, avoidances and segmentary taboos regarding the sources of animal protein among three caiçara communities on the southern coast of the Atlantic Forest, Brazil. Caiçaras are native inhabitants of the southeastern Brazilian coast, descendants of Amerindians and European colonizers (Begossi, 1992; Hanazaki et al., 1996). Based on the literature on food preferences regarding fish (Smith, 1981, 1996; Begossi and Braga, 1992; Begossi, 1998), which points out a general pattern of preference for fish with scales and avoidance or restriction for scaleless fish, we also aim to discuss two aspects of food choices among caiçaras. The first concerns their preference for consumption of catfish (Ariidae) in one of the studied communities, considering that this fish is scaleless. The second refers to both the preferences and segmentary taboos for different species of mullets (Mugilidae) found in all the communities studied.

Methods

Study site

The caiçara communities in this study are located on the estuarine system of Cananéia-Iguape, on the southern coast of São Paulo State, Brazil (Figure 1). This system is a mangrove-dominated estuary with shelf sands forming a barrier island (Comprida Island) that encloses narrow water bodies (Schaeffer-Novelli et al., 1990). The three communities included in this study were selected according to the following criteria: a) the majority of its full-time residents are caiçara; b) the communities are located on the margins of the estuary and, in some degree, depend on fisheries; and c) the communities have at least 10 resident families.

The selected communities were Icapara, Pedrinhas, and São Paulo Bagre (Figure 1). Icapara is located in the Iguape municipality. It has about 350-400 local families and approximately 200 houses owned by tourists (part-time residents) and newcomers. Pedrinhas is the main caiçara community of the Ilha Comprida municipality, has 60 local families and about 100 houses owned by tourists and non-native inhabitants. São Paulo Bagre has 17 local families and one house owned by a non-native. Fishing activities in Icapara are focused on Broadband anchovy fishery for the local industry (see fish identification and vernacular names in Table I). In Pedrinhas, fishing activities are mainly for subsistence and for local commerce. The species caught include mullet, snook, weakfish, croaker and white shrimp. In São Paulo Bagre, fishing focuses on shrimp, sold to tourists and sport fishermen as living baits (Hanazaki and Begossi, 2000a; Hanazaki, 2001) for snook and weakfish catch. The Portuguese name of this latter community literally means "Saint Paul Catfish".

Interviews

Data were collected through semi-structured interviews carried out between May and August 1998. Interviewees were adult men and women with at least two years of residence in the region. They were asked to list all the fish in four previously defined categories (Begossi, 1992; Begossi and Braga, 1992; Hanazaki et al., 1996; Madi and Begossi, 1997): a) frequently consumed (Which are the fish that you eat most frequently?); b) preferred (Which are the fish you most like to eat?); c) avoided (Which are the fish that you like least?); and d) segmentary taboos (Which are the fish that should not be eaten when a person is pregnant, lactating, or ill?). Throughout this work, segmentary taboos were considered in this sense. The last category was also associated with the terms "reimoso", "carregado" (heavy), "forte" (strong) and "quente" (hot).

Additionally, systematic observations on the use of natural resources in the local diet were carried out in Pedrinhas and São Paulo Bagre between September 1998 and August 1999 (Hanazaki, 2001). The identification of zoological material was based on collected specimens, photographs, and pictures shown to local inhabitants. Collected fishes were identified through identification keys by Figueiredo (1977), Figueiredo and Menezes (1978, 1980), Menezes and Figueiredo (1980, 1985), Menezes (1983) and Serralheiro et al. (1994). The fish identifications were reviewed by J. L. Figueiredo, from the Zoology Museum, São Paulo University. Collected specimens were deposited at the Natural History Museum of Campinas University and at the Zoology Museum of São Paulo University. The identification of terrestrial mammals was based on Emmons (1990) and Fonseca et al. (1996), and reviewed by E. Z. F. Setz, from the Zoology Department, Campinas University.

Cluster analysis through UPGMA, or average-linkage (Sneath and Sokal, 1973), was done in order to verify the existence of clusters of fishes according to the four categories of consumption. In average-linkage clustering, the between-group (dis)similarity is defined as the average (dis)similarity between all possible pairs of members (one of each group). This method is most widely used in ecology and systematics (taxonomy). The algorithm maximizes the "cophenetic correlation", the correlation between the original (dis) similarities and the (dis) similarities between samples, as can be derived from a dendrogram. For any sample pair, it is the lowest dissimilarity (or highest similarity) required to join them in the dendrogram (Sneath and Sokal, 1973). The number of citations of 21 species of fish mentioned in at least three interviews (Table I) was transformed through log+1. Similarity distances were calculated through Euclidean distance.

Results

A total of 108 local inhabitants were interviewed, 51 from Icapara, 32 from Pedrinhas, and 25 from São Paulo Bagre. All the interviewees mentioned consumed fish, 86% mentioned preferred fish, 83% mentioned segmentary tabooed fish, and 69% mentioned avoided ones.

The cluster analysis divided the more frequently consumed fish and the less frequently consumed ones (Figure 2, groups 1 and 2, respectively). Among the frequently consumed fish (group 2), mullet (TAI) was also the preferred species, appearing as a separate subgroup. Less consumed fish (Figure 2, group 1) include two sub-groups: the fish widely mentioned as segmentary taboos of group 3 (ray ARR, shark CAC, and redeye mullet PPM), and the other less consumed fish seldom considered segmentary taboos (group 4).

Among the most cited fish in each community, mullet was mentioned as the most frequently consumed fish in two communities; but in São Paulo Bagre catfish was the most frequently consumed fish, followed by weakfish and white mullet (Figure 3a). Despite the preference for consuming catfish in São Paulo Bagre (Figure 3b), there is a strong avoidance for this fish in the other two communities (Figure 3c). Hanazaki (2001) followed the diet of 32 families in Pedrinhas and in São Paulo Bagre during one whole year and observed a catfish consumption frequency of 25% at São Paulo Bagre, while at Pedrinhas it was only 7%.

The genus Mugil is an outstanding group regarding food preferences and taboos. Three species of this genus have different categorizations: mullet (M. platanus) was reported as a frequently consumed (Figure 3a) and a preferred fish in all communities (Figure 3b), white mullet (M. curema) was frequently consumed (Figure 3a), while redeye mullet (M. gaimardianus) was mentioned as a segmentary taboo in both communities (Figure 3d), and was not mentioned in any other consumption category (Figure 3a, b, c).

When asked about items whose consumption is forbidden in cases of pregnancy, lactation, or illness, some interviewees also cited canned food and other animals, such as terrestrial mammals. Other foods considered reimoso are pork (10% of the interviewees) and game meat (7%). Game meat includes anteater, armadillo, paca, opossum, and coati. Prohibitions on the fish consumption (Figure 2, group 3, and Figure 3d) were confirmed and followed by all pregnant or lactating women during the fieldwork.

Discussion

Food preferences, avoidances and taboos: animal meat and fish

Despite the food preferences regarding species or groups of fish, there are species whose ratings in preference depend on the way they are cooked, because fish may be boiled or fried. Boiling is preferred for fish without scales, the big fish, and for fish with few bones. Fish with scales are preferred fried rather than boiled.

Regarding animal meat in general, beef is preferred to fish. One of the main activities of the local population is fishing, and fish is consumed more frequently than other meats, such as beef and poultry. According to the interviewees, the frequent consumption of certain foods can justify its preference, because they are used to it. Nevertheless, it is also the reason for the preference of other meats instead. The social role of the rare food items (MacBeth and Lawry, 1997), which gives them a higher status than a food frequently consumed (Harris, 1987a), also explains the preference for beef over fish.

Regarding the food taboos on fish, in Icapara, reimoso food was emically associated with the followers of Seventh-Day Adventism, which imposes religious beliefs on food restrictions towards the consumption of pork and on detritivorous and carnivorous fish, such as catfish. Under an etic explanation, the taboo on some fish species can be related to the presence of toxic substances when the fish deteriorates (Begossi, 1998), especially in the case of the Scombridae family (tunas and bonitos) and Elasmobranchs (sharks and rays). This explanation can be operating with or without juxtaposition to the religious beliefs. Sharks and rays were two of the three species more commonly referred as a segmentary taboo (Figure 2, group 3).

Segmentary taboos and nutrition

When segmentary food taboos are followed, the meat recommended for consumption is poultry. This is a widespread recommendation to pregnant or lactating women in Brazilian rural communities (Cândido, 1977; Fleming-Moran, 1992). According to Harris (1987a) there is no evidence that removing food taboos might result in improved nutrition for lactating women. Analyzing the nutritional impact of food taboos among communities from Ituri Forest, Zaire, Aunger (1992) found that these taboos had a minimum impact on their diet. In the case of lactating women in Queensland, Australia, Eaton-Evans and Dugdale (1986) found that the women who avoided some food items had significantly higher weight loss than those who followed a diet without restrictions, but there was no significant effect on the weight increase of their babies. Thus, there was no evidence that food aversions help babies or harm their mothers.

The connection between food taboos and nutritional quality should not have a significant effect on the studied caiçara populations, as there is no evidence of protein malnutrition in caiçara communities (Hanazaki and Begossi, 2000b; Hanazaki, 2001). Currently, the fact that there is access to other sources of animal protein, including purchased items such as poultry and beef, does not represent a problem that could result in nutrient deprivation.

Catfish: preferred or avoided?

Catfish (saltwater catfish from the Ariidae family and freshwater catfish from the Pimelodidae family) are often avoided by the fishing populations from both the Brazilian coast and inland (Begossi and Braga, 1992; Smith, 1996; Murrieta, 1998). These species are classified as "smooth" (scaleless) fish and considered reimoso or carregado.

Food taboos regarding scaleless fish are reported since biblical times (Douglas, 1969; Pálsson, 1991). Indeed, in Icapara the Seventh-Day Adventists consider catfish to be a forbidden food. Keeping this taboo may be interpreted as a symbolic link to religious purity (Douglas, 1969; 1999; Heald et al., 2004). However, to understand food restrictions linked to religious beliefs it is necessary to analyze them in the framework in which these religious codes arise (Grivetti, 1978; Harris, 1987b). Davis (1989) suggested another explanation for the avoidance of scaleless fish, arguing that the toxicity of pufferfish (Tetraodontidae) is so great that it could have influenced the widespread avoidance of scaleless fish in general. Pufferfish have no typical scales, but can have projections like spines in some parts of their body (Figueiredo and Menezes, 2000). Some genera such as Lagocephalus and Spheroides have few or no spines, and are considered scaleless by the fishermen. In the case of taboos on pufferfish, the etic explanation lies in their toxicity (Begossi, 1992). Another aspect of etic interpretations may the position of catfish in the food web: catfish are usually detritivorous or carnivorous, being prone to the accumulation of toxins (Begossi, 1992; Begossi and Braga, 1992). According to Mishima and Tanji (1982) the diet of saltwater catfish in this estuarine region (Ariidae: Cathorops spixii, Genidens genidens, Sciadeichthys luniscutis, Bagre bagre, B. marinus and Netuma barba) consists mainly of fishes (Stellifer sp., Macrodon sp., Cynoscion sp., Micropogon sp., C. spixii, among others), Decapoda (Litopenaeus sp., Alpheus sp., Portunus sp., Pinnixa sp. and Uca sp.), Annelida (Polichaeta), Mollusca (Littorinaceas, Tellina sp., Macoma sp., among others), zooplankton (Copepoda), macroalgae and macrophytes. All species of catfish have a broad feeding niche, with the largest individuals preferring the higher trophic levels and the smaller ones preferring the lower trophic levels (Mishima and Tanji, 1982).

According to Scarpin (1992), the name of the community of São Paulo Bagre could be related to the local abundance of bagre, or catfish. Names often reflect past experiences or circumstances; however, when looking for diachronical evidences of a higher abundance of catfish near this community, we do not find relevant data to support this explanation. According to the interviewed fishermen from the whole estuary, catfish was always an abundant resource. Moreover, data obtained in experimental fisheries on the estuarine region do not indicate a higher abundance of catfish near São Paulo Bagre compared to Pedrinhas (Begossi, 2000). Analyzing the spatial distribution of saltwater catfish in this estuary (Ariidae: C. spixii, G. genidens, S. luniscutis, B. bagre, B. marinus and N. barba), Mishima and Tanji (1983) found that the preferred areas for the juveniles of all catfish species are near the mouth of streams with sewer dumps and near the city of Cananéia, due to a high concentration of organic matter in those spots. The adult catfish had different preferences with reference to the salinity and depth, but are distributed all over the whole estuarine region. Mishima and Tanji (1983) found that in 14 collecting locations on the estuary, the average number of individuals per trawl was 85.6 catfish, ranging from 8.6 to 275.7 individuals. Near São Paulo Bagre the average was 99.9, and near Pedrinhas the average was 48.1 individuals per trawl. However, other collecting locations had higher average catches (169.6, 220.8 and 275.7 individuals per trawl, suggesting that although catfish are more abundant near São Paulo Bagre than near Pedrinhas, their abundance is only a little higher than the average for the estuary.

Pálsson (1991) argued that one way to understand similarities and differences among fishing systems is to consider the social context of production, regarding the motivation of the producers and the destination of the products. When considering the domestic mode of production, it is motivated by the subsistence needs and the household values, which can alter the logic of the production market of capital accumulation (Pálsson, 1991). This alternate reason can also be considered for São Paulo Bagre, where a group of fish of low commercial value and often avoided as food is locally caught. This happens due to the local tradition of catfish fisheries as shown by the name of the community and by local food preferences. In the regional market, catfish are less valued than weakfish, mullets, and snooks, but a typical dish from this region is the bagre seco com banana (dried catfish with bananas). Tourists from nearby cities search for dried catfish in São Paulo Bagre. As discussed by Pálsson (1991), the definition of valuable fish is determined by many factors, including the domestic and local values, the market conditions and the external demand, all of which enter the local fishermen’s conceptual universe.

We are then left with two non-exclusive hypotheses. The first hypothesis suggests that a higher availability of catfish near São Paulo Bagre influences local food preferences, and the second hypothesis suggests that the local valorization of catfish is one way of enhancing the local market economy. In São Paulo Bagre, fishermen may be exploiting the environmental heterogeneity of the estuary, searching for the fishing spots with a greater abundance of catfish. The tradition of catfish fisheries should be intimately linked to the fishing skills and the local knowledge of this resource.

Mullets: consumed, preferred and tabooed

Fishes of Mugil genus are abundant in the Cananéia-Iguape estuarine system (Schaeffer-Novelli et al., 1990). For example, mullet is one of the leading species caught by fixed fish traps, mainly between April and August (Radasewsky, 1976). Mullet (M. plantanus) is also a valued species in other caiçara communities along the Brazilian coast (Hanazaki et al., 1996). Its symbolic importance may be related to its marked seasonality and to the communal labor historically involved in mullet fishing (Schmidt, 1947; Mussolini, 1980). Mullet fishing was once the main communal activity during the winter (between May and June) in caiçara communities from the southeastern Brazilian coast. It can still be found among fishing communities in Southern Brazil, such as the community from Pântano do Sul, at Florianópolis Island (Medeiros, 2001).

Two other species of the Mugil genus occur in the estuarine system of Cananéia-Iguape. They belong to the locally called group of paratís and are very similar to each other. The paratí comum, also known as paratí-puá or paratí-guaçú, is identified as M. curema (white mullet), and the paratí-pema is identified as M. gaimardianus (redeye mullet). When the fish catches are directed to the market, both species are sold together under the vernacular name of paratí. According to the fishermen, the main characteristics that differentiate redeye mullet from white mullet are the absence of horizontal ridges on the body, and the absence of a yellowish spot on the gill flap of redeye mullet. Fishermen also considered redeye mullet a rare species as compared to white mullet. According to Menezes (1983), redeye mullet can be rarer than other species of the same genus, such as white mullet and mullet. Some fishermen say that they prefer the open sea and are therefore rarely found in the estuary.

Despite their similar morphology, redeye mullet is often mentioned as a tabooed species (Table I). Fishing communities from Paranaguá Bay, located in the State of Paraná, on the southern border of the State of São Paulo also consider redeye mullet as a segmentary taboo, but eat white mullet (Fernandes-Pinto, 2001).

Detailed knowledge of Mugil has been observed in other Brazilian regions, such as in the estuary of Mundaú-Manguaba, in the State of Alagoas, northeastern Brazil (Marques, 1991). This knowledge is due to the important role that this genus plays for the local human population in terms of being a source of income and a way of making direct subsistence.

Fishermen from Mundaú-Manguaba identify the Mugilidae (M. liza, M. gaimardianus, M. trichodon, M. curema and M. incilis) as the family of peixes de umbigo, or "navel" fish (Marques, 1991). According to this author, the names given to the Mugilidae at Mundaú-Manguaba, such as curimã, patriaçu, tainha-negão, tainha-do-oito-amarelo, saúna and zereda arise from a complex system of at least 54 folk names. These species are considered "carregadas" or "não carregadas", which include an analogous meaning to "reimoso" or "not reimoso". In grading terms of this classification, the species M. gaimardianus is the most carregada at Mundaú-Manguaba. One of the characteristics that differentiate this species from M. incilis is its strong smell (Marques, 1991). However, not all the caiçaras avoid eating redeye mullet because it smells bad. Indeed, some fishermen and some of their wives state that there is no difference between the smell of redeye mullet and of white mullet, even during the cooking process.

According to Menezes and Figueiredo (1985) the Mugilidae feed mainly on vegetal matter from the muddy or sandy bottom of the estuary. This information was confirmed by the fishermen interviewed in this study, who refer to paratís as "fish that graze in the mud". However, based on the local knowledge of Mundaú-Manguaba fisherman, Marques (1991) suggested the existence of small variations in the habitat and foraging habits of Mugilidae. These small variations can result in alterations on the trophic position of these species, thus influencing the avoidance to M. gaimardianus. Mullets are migratory fish, entering the estuary as juveniles (Schaeffer-Novelli et al., 1990). Differences in the habitat in other regions outside the estuary can also affect the trophic position of these species.

Ecological effects of food preferences

Colding (1995) discussed the role of taboos on natural resources regarding environmental conservation and concluded that the taboos related to food items were those with greater ecological effects. The relation between food taboos and resource conservation was suggested by McDonald (1977) and Ross (1978) for Amerindian people from the Amazon, who used game as a source of protein. Food taboos could reduce the hunting intensity among these populations (McDonald, 1977). Criticism of such interpretations refer to the lack of empirical data (Alvard, 1998) and to the small nutritional impact of food taboos on the diet (Aunger, 1992). The avoidance or the taboo of a food item in order to conserve it can be related to the medicinal use of the species (Begossi, 1992). This was observed among caiçaras from Búzios Island (Brazil), where the lizard (Tupinambis sp.) is a taboo (Begossi, 1992), and among riverine people from Tocantins River (Brazil), where rays (Potamotrygonidae) and fish with high fat content were avoided (Begossi and Braga, 1992).

No strong evidence of food preferences, avoidances and taboos resulting in resource conservation, was found in this study, not even for medicinal purposes. Catfish, avoided in Icapara and Pedrinhas, are quite abundant in the estuary (Mishima and Tanji, 1983). Even though there may appear to be a scarcity of redeye mullet when compared to other Mugil species, there is no reliable data on its abundance.

The relation between food taboos about fish species and resource conservation is improbable, when compared to the possible relations between game taboos and conservation of the terrestrial fauna. This kind of relationships should operate in a complex scenario with many causational factors acting together to result in a given conservative behavior. When discussing the food rules and taboos in Papua New Guinea, Whitehead (2000) argues against the one-level of analysis approach, suggesting that there are dynamically inter-relating mutually causative forces acting over the food rules (McPherson, 2003).

Conclusions

Along with other fishing populations, the caiçaras from the southern coast of the State of São Paulo have detailed food preferences, avoidances and taboos for animal protein, especially fish. These preferences and avoidances have both cultural and environmental explanations, which involve factors such as the species availability, their position in the food web, or their importance in the local context of a given community. The reasons that guide food preferences and avoidances seem to be the result of the interrelation of many of these factors.

The aversion to sharks and rays, a general trend found in fishing communities of Brazil, was also found among the studied communities. However, some particularities were observed, such as the preferential consumption of catfish in São Paulo Bagre. This preference is the opposite of that observed for the majority of fishing populations, most of whom avoided scaleless fish or showed a segmentary food taboo regarding catfish. The preference and consumption of catfish in São Paulo Bagre may be explained by a slightly higher availability in this locality as well by the local context of the catfish production. In São Paulo Bagre the tradition of catfish fisheries and the need to increase the value of a fish generally conceived as a ‘low status fish’ affects food preferences.

Another peculiarity was the consumption of Mugilidae. The abundance and position in the trophic chain suggests no specific reason for redeye mullet being widely considered a segmentary taboo. In this case, more data from other communities should be gathered to find an etic explanation for the emergence, adoption and persistence of this taboo.

A connection between resource conservation and food taboos about certain fish species is unlikely when compared to the possible relations between hunting taboos and conservation of the terrestrial fauna, because the tabooed and avoided fish are abundant on the estuarine region.

Acknowledgements

The authors thank N. Peroni for the help in data collection; J. L. Figueiredo and E. Z. F. Setz for reviewing specimen identifications; and E. F. Madi, C. S. Seixas, F. Castro, J. G. Marques, E. Moran and N. Miller for suggestions on a previous version of this manuscript. They are especially grateful to the caiçaras from the studied communities for their kind patience. FAPESP granted a doctoral scholarship (97/3668-8) to N. Hanazaki's and gave financial support; and CNPq granted a research productivity scholarship to A. Begossi.

References

1. Alvard MS (1998) Evolutionary ecology and resource conservation. Evol. Anthropol. 7: 62-74.        [ Links ]

2. Armelagos G (1987) Biocultural aspects of food choice. In Harris M, Ross E. (Eds.) Food and Evolution. Temple University Press. Philadelphia, PA, USA. pp. 579-594.        [ Links ]

3. Aunger R (1992) The nutritional consequences of rejecting food in the Ituri Forest of Zaire. Human Ecol. 20: 263-291.        [ Links ]

4. Begossi, A (1992) Food taboos at Búzios Island (Brazil): Their significance and relation to folk medicine. J Ethnobiol. 12: 117-139.        [ Links ]

5. Begossi A (1998) Food taboos - a scientific reason? In Pendergast HDV, Etkin N, Harris DR, Houghton PJ (Eds.) Plants for Food and Medicine. Royal Botanic Gardens. Kew, UK. pp. 41-46        [ Links ]

6. Begossi A (2000) Floresta e Mar: Usos e Conflitos no Vale do Ribeira e Litoral Sul de São Paulo. Second Report FAPESP 97/14514-1. Brazil.        [ Links ]

7. Begossi A, Braga FMS (1992) Food taboos and folk medicine among fishermen from the Tocantins River. Amazoniana 12: 101-118.        [ Links ]

8. Begossi A, Hanazaki N, Ramos RM (2004) Food chain and the reasons for fish food taboos among Amazonian and Atlantic Forest fishers (Brazil). Ecol. Applic. 14: 1334-1343.        [ Links ]

9. Cândido A (1977) Os Parceiros do Rio Bonito. 4a ed. Livraria Duas Cidades. São Paulo, Brazil. 284 pp.        [ Links ]

10. Colding J (1995) Taboos and the Conservation of Natural Resources, Species and Ecosystems. Thesis. Stockholm University. Sweden. 82 pp.        [ Links ]

11. Coronios-Vargas M, Toma RB, Tuveson RV, Schultz IM (1992) Cultural influences on food cravings and aversions during pregnancy. Ecol. Food Nutr. 27: 43-49.        [ Links ]

12. Davis W (1989) Passage of Darkness: the Ethnobiology of the Haitian Zombie. The University of North Carolina Press. Chapel Hill, NC, USA. 356 pp.        [ Links ]

13. Douglas M (1969) Purity and Danger: an Analysis of Concepts of Pollution and Taboo. 2nd ed. Routledge & Kegan. London, UK. 188 pp.        [ Links ]

14. Douglas M (1999) Leviticus as Literature. Oxford University Press. Oxford, UK. 304 pp.        [ Links ]

15. Eaton-Evans J, Dugdale AE (1986) Food avoidance by breast feeding mothers in south east Queensland. Ecol. Food Nutr. 19: 123-129.        [ Links ]

16. Emmons LH (1990) Neotropical Rainforest Mammals. University of Chicago Press. Chicago, IL, USA. 295 pp.        [ Links ]

17. Fernandes-Pinto E (2001) Etnoictiologia dos Pescadores da Barra do Superagüi, Guaraqueçaba/PR: aspectos etnotaxonômicos, etnoecológicos e utilitários. Thesis. Universidadde Federal de São Carlos, Brazil. 138 pp.        [ Links ]

18. Figueiredo JL (1977) Manual de Peixes Marinhos do Sudeste do Brasil. I-Introdução. Cações, raias e quimeras. Museu de Zoologia/USP. São Paulo, Brazil. 104 pp.        [ Links ]

19. Figueiredo JL, Menezes NA (1978) Manual de Peixes Marinhos do sudeste do Brasil. II-Teleostei (1). Museu de Zoologia/USP. São Paulo, Brazil. 110 pp.        [ Links ]

20. Figueiredo JL, Menezes NA (1980) Manual de Peixes Marinhos do sudeste do Brasil. III-Teleostei (2). Museu de Zoologia/USP, São Paulo, Brazil. 90 pp.        [ Links ]

21. Figueiredo JL, Menezes NA (2000) Manual de Peixes Marinhos do sudeste do Brasil. VI-Teleostei (5). Museu de Zoologia/USP, São Paulo, Brazil. 116 pp.        [ Links ]

22. Fleming-Moran M (1992) The folk view of natural causation and disease in Brazil and its relation to traditional curing practices. Bol. Museu Paraense Emílio Goeldi, série Antropologia 8: 65-156.        [ Links ]

23. Fonseca GAB, Herrmann G, Leite YLR, Mittermeier RA, Rylands AB, Patton JL (1996) Lista Anotada dos Mamíferos do Brasil. Occasional papers in Conservation Biology, 4. Conservation International/Fundação Biodiversitas. Washington, DC, USA. 38 pp.        [ Links ]

24. Grivetti L (1978) Culture, diet, and nutrition: selected themes and topics. Bioscience 28: 171-177.        [ Links ]

25. Hanazaki N (2001) Ecologia de Caiçaras: uso de recursos e dieta. Thesis. UNICAMP. Campinas, Brazil. 193 pp.        [ Links ]

26. Hanazaki N, Begossi A (2000a) Caiçaras, mangroves and estuaries: an ethnoecological approach. Electronic papers of the Conference "Sustainable Use of Estuaries and mangroves: challenges and prospects". Recife, Brazil. CD-ROM.

27. Hanazaki N, Begossi A (2000b) Fishing and niche dimension for food consumption of caiçaras from Ponta do Almada (Brazil). Human Ecol. Rev. 7: 52-62.

28. Hanazaki N, Leitão-Filho HF, Begossi A (1996) The use of resources of the Brazilian Atlantic Forest: the case of Ponta do Almada (Ubatuba, Brasil). Interciencia 21: 268-276.        [ Links ]

29. Harris M (1987a) Foodways: historical overview and theoretical prolegomenon. In Harris M, Ross RB (Eds.) Food and Evolution. Temple University Press. Philadelphia, PA, USA. pp. 57-90.        [ Links ]

30. Harris M (1987b) The Sacred Cow and the Abominable Pig: riddles of food and culture. Simon and Schuster. New York, NY, USA. 288 pp.

31. Heald S, Grabbe LL, Handelman D, Segal A, Hendel RS (2004) Book review forum on Mary Douglas´s Leviticus as Literature. J Ritual Studies 18: 152-191.        [ Links ]

32. MacBeth H, Lawry S (1997) Food preferences and taste. In MacBeth H (Ed.) Food Preferences and Taste: Continuity and Change. Berghahn. Providence, RI, USA. pp. 1-13.        [ Links ]

33. Madi E, Begossi A (1997) Pollution and food taboos: a practical reason? J Human Ecol. 8: 405-408.        [ Links ]

34. Marques JGW (1991) Aspectos Ecológicos na Etnoictiologia dos Pescadores do Complexo Estuarino-lagunar Mundaú-Manguaba, Alagoas. Thesis. UNICAMP. Campinas, Brazil. 310 pp.        [ Links ]

35. Maués RH, Motta-Maués MA (1977) O modelo da "reima": representações alimentares em uma comunidade amazônica. Anuário Antropológico 77: 120-147.

36. Mcdonald DR (1977) Food taboos: a primitive environmental protection agency (South America). Anthropos 72: 734-748.        [ Links ]

37. McPherson NM (2003) Book Review Forum on Harriet Whitehead’s Food Rules: Hunting, Sharing, and Tabooing Game in Papua New Guinea. J. Ritual Studies 17: 98-104.        [ Links ]

38. Medeiros R (2001) Estratégias de Pesca e usos dos Recursos em uma comunidade de pescadores artesanais da praia do Pântano do Sul (Florianópolis, Santa Catarina). Thesis. UNICAMP. Campinas, Brazil. 108 pp.        [ Links ]

39. Menezes NA (1983) Guia prático para conhecimento e identificação das tainhas e paratis (Pisces, Mugilidae) do litoral brasileiro. Rev. Bras. Zool. 2: 1-12.        [ Links ]

40. Menezes NA, Figueiredo JL (1980) Manual de Peixes Marinhos do Sudeste do Brasil. IV-Teleostei (3). Museu de Zoologia/USP. São Paulo, Brazil 96 pp.        [ Links ]

41. Menezes NA, Figueiredo JL (1985) Manual de Peixes Marinhos do Sudeste do Brasil. V-Teleostei (4). Museu de Zoologia/USP. São Paulo, Brazil. 105 pp.        [ Links ]

42. Messer E (1984). Anthropological Perspectives on Diet. Annu. Rev. Anthropol. 13: 205-249.        [ Links ]

43. Mishima M, Tanji S (1982) Nicho alimentar de bagres marinhos (Teleostei, Ariidae) no complexo estuarino-lagunar de Cananéia (25o S, 48o W). Bol. Inst. de Pesca 9: 131-140.        [ Links ]

44. Mishima M, Tanji S (1983) Fatores ambientais relacionados à distribuição e abundância de bagres marinhos (Osteichthyes, Ariidae) no complexo estuarino-lagunar de Cananéia (25o S, 48o W). Bol. Inst. de Pesca 10: 17-27.        [ Links ]

45. Murrieta RSS (1998) O dilema do papa-chibé: consumo alimentar, nutrição e práticas de intervenção na Ilha de Ituqui, baixo Amazonas, Pará. Rev. Antropol. 41: 97-150.        [ Links ]

46. Mussolini G (1980) Ensaios de Antropologia Indígena e Caiçara. Paz e Terra. Rio de Janeiro, Brazil. 289 pp.        [ Links ]

47. Pálsson G (1991) Coastal Economies, Cultural Accounts: Human Ecology and Icelandic Discourse. Manchester University Press. Manchester, UK. 224 pp.        [ Links ]

48. Queiroz MS (1984) Hot and cold classification in traditional Iguape medicine. Ethnology 23: 63-72.        [ Links ]

49. Rea AM (1981) Resource Utilization and Food Taboos of Sonoran Desert Peoples. J Ethnobiol. 1: 69-83.        [ Links ]

50. Roosevelt A (1987) The evolution of human subsistence. In Harris M, Ross EB (Eds.) Food and Evolution. Temple University Press. Philadelphia, PA, USA. pp. 565-578.        [ Links ]

51. Ross EB (1978) Food taboos, diet, and hunting strategy: the adaptation to animals in Amazon cultural ecology. Curr. Anthropol. 19: 1-36.        [ Links ]

52. Scarpin E (1992) O Imaginário Religioso num Bairro Rural de Cananéia. Thesis. FFLCH/USP. São Paulo, Brazil. 219 pp.        [ Links ]

53. Schaeffer-Novelli Y, Mesquita HSL, Cintrón-Molera G (1990) The Cananéia Lagoon Estuarine System, São Paulo, Brazil. Estuaries 13: 193-203.        [ Links ]

54. Schmidt CB (1947) Alguns aspectos da pesca no litoral paulista. Rev. Museu Paulista, Nova Série, 1: 181-212.        [ Links ]

55. Serralheiro PCS, Godinho HM, Paiva P (1994) Identificação de tainhas (Mugil sp.) da região estuarino-lagunar de Cananéia, SP, Brasil. Bol. Instituto de Pesca 21: 95-102.        [ Links ]

56. Simoons FJ (1978) Traditional use and avoidance of foods of animal origin: a culture historical view. Bioscience 28: 178-184.        [ Links ]

57. Smith NJH (1996) The Enchanted Amazon Rain Forest. Stories from a vanishing world. University of Florida Press. Gainesville, FL, USA. 194 pp.        [ Links ]

58. Smith NJH (1981) Man Fishes and the Amazon. Columbia University Press. New York, NY, USA. 180 pp.        [ Links ]

59. Sneath PHA, Sokal RR (1973) Numerical Taxonomy. Freeman. San Francisco, CA, USA. 573 pp.        [ Links ]

60. Strathern A, Stewart PJ (1999) Curing and Healing: Medical Anthropology in Global Perspective. Carolina Academic Press. Durham, NC, USA. 224 pp.        [ Links ]

61. Walker ARP (1995) Dietary advice: from folklore to present beliefs. Nutr. Rev. 53: 8-10.        [ Links ]

62. Whitehead H (2000) Food Rules: Hunting, Sharing, and Tabooing Game in Papua New Guinea. University of Michingan Press. Ann Arbor, MI, EEUU. 344 pp.        [ Links ]

63. Zent S (1996) Behavioral orientations toward ethnobotanical quantification. In Alexiades MN (Ed.) Selected Guidelines for Ethnobotanical Research: a field manual. New York Botanical Garden. New York, EEUU. pp. 199-239.        [ Links ]