XII (L): L0l - Ll8 Kiel, Juni 1992

AMAZONIANA XII (l): l0l - ll8 Kiel, Juni 1992

Food taboos and folk medicine among fishermen from the Tocantins River (Brazil)

by

Alpina Begossi- and Francisco Manoel de Souza Braga

Dr. Alpina Begossi, Departamento de Ecologia, IB, Universidade Federal do Rio de Janeiro, C.P. 68020, 21941Rio de Janeiro, RI, Brazil. Prof. Dr. Francisco Manoel de Souza Braga, Departamento de Zoologia, Universidade Estadual Paulisø, C.P. 179, 13.500 Rio Claro, SP, Brazil. (Accepted for publicatian: August, 1990).

Abstract

Fish utilization for food and folk medicine, and fish preference of families from the Tocantins river were studied. Questionnaires werc used in the 234 interviews performed in cities, towns and scatteres houses located along 100 km of river stretch. Curimatí (Prochilodtu nigricans) is the most consumed fish and pacu-manteiga (Mylossorna duriventre) ¡he most preferred species. The fish species avoided are correlated with the species used in folk medicine (r, = 0,54, p < 0.02). Food t¡boos, orfish species not consumed during illness, are also cited. The usefulness of fish species for folk medicine and the piscivorous habits of most fish quoted as not consumed partially explain the food choices of fishermen. These explanations conform to materialist theories in cultural ecology.

Keywords: DIeÇ food teboog ñshlng communitleq Brezll, Rlo Toc¡ntins.

Introduction

Food preferences and avoidances have been a subject of many studies in Anthro- pology and Human Ecology. MESSER (1984) reviewed some of these studies and factors involved in the fmd choice of human populations. The avoidances of some food by people are considered to be based on ideological criteria (SAHLINS 1976) or on materialist reasons (HARRIS L977,1985). The ideologi- cal view considers that cultural patterns are arbitrary and derive from the human mind (ROSS 1987). Thus, we need to understand the internal logic of culturally specific symbol systems (BASSO 1978). The materialist point of view considers tho costs and

' Present address: NEPAÌví/IJMCAMP, C. P. I170, BR-13081 Campias, SP, Brazil.

ISSN 0065675511992t1011 @ MPI fü¡ Limnologie, AG Tropenökologie, Plön; INPA, Manaus

101 benefits involved in any decision making process, including food choices. For example, HARRIS (1985), based mainly on ecological and economical principles, explained why cor¡/s are avoided as food in India, why dogs are usually not eaten by Westerners and why insects are eaten and considered very østy by a variety of other human populations. A R,A As summarized by HARRIS (1987), food habits are explained by the idealist approach " /,/ as consequence of culture and by the materialist criæria by environmental, demographic, technological and political-economic factors. Food avoidances, or food taboos, have also been interpreted as a way of minimizing fhe overexploitation of certain animals G.OSS I MPERÂÎâ I Z r978). STATE This study focuses on fish preferences and avoidances of fisherman families on the 4æ V¡Rar.¡¡]ÃO Tocantins river, in Brazil. Our aim is to analyze fish utilization and understand why STATE some available fish are eaten where¿¡s others are rejected as food. In this study we use the terms "fish avoided" for species of fish which are usually not consumed and "fish a L^JË^OO taboos" for species of fish not consumed during illness.

TOCANTIN9 Methodology STATE

The study sites include cities, small towns and riverine settlements loc¡ted in the banks of the Tocantins river, between the cities of Imperatriz and Estreito. The extension of the area is abor¡t 100 km and it is located in the States of Mara¡rhão and Tocantins. Fieldwork included mainly interviews with fishermen and their wives, and was performed in altemate months from October 1987 to June 1988. The cities and towns studied at the Tocantins river are shown in figure l Imperatri", Porto Franco, Tocantinópolis and Estreito are considered. as "cities" and ltaguatins, Descarreto and lajeato.as "towns". Fig. 1: Riverine fishermen were sampled in both sides of the river. Map including the cities and towns studied.

1. Interviews lnterviews were based on questiørnaires. A preliminary questionnaire was prepared before going into the 2. Fish identification field and tested in the first visit. The, secqrd questionnaire focused on, among others, fish food øboos a¡rd fish Prof. J. C. CARAVELLO (Universidade Federal de São Carlos) collected and identified fish specimens, medicinal uses. Inten¡iews were performed panianlarly with fishermen. Wives were asked about rheir literacy, and checked many of them with fishermen. Prof. M. PETRERE Jr., from the Universidade Esødual Paulista, number of children and age. They also participated in questians about diet, food taboos and medicinal animals. contributed useful information. A list of common and scientific names of fish are in BEGOSSI & Wqnen were found to react more cørfortably than men in answering about food taboos as well as in quoting GARAVELLO (r990). medicinal animals. This probably occurs because women take care of children and prepare folk medicine at home. In the case of a fisherman's absence, the questionnaire was applied to his wife. The family was the unit 3. Fish feeding habits of study. Fish feeding habits werc observed through the analysis of stomach contents and through the literature. The "Colônias de pescadores" (fishermen's associatiurs) of Imperatriz and Estreito helped in estimating Unfortunately, there are cønparatively few studies on feeding habits of Br¿zilian fresh water fish. This kind the number of fishermen from each city and in locaring their residences. The affilation files were also of information for specific species was taken from BRITSKI (1972), NOMURA (1975), NOMURA et al. consulted, Three kinds of sampling methods were carried out, deper,ding upon the size of the settlements: (1981), NOMURA & NEMOTO (1983), NOMURA & TAVEIRA (1979), NOMURA et al. (1983), CATELLA a) cities - fisherman hor.¡ses were located after informal interviews with fishermen at the main harbor and & TORRES (1984), MOURÃO & TORRES (1984) and GOULDING & FERREIRA (1984). Trre dier of a after consulting the affiliation files of the "Colônias". Fishermen were found to live clustered in specific group of species of specific regiors were found in intemal reports, such as INPA/ÏLETRONORTE (1980/83), localities, usually close to the river. Interviews were undertaken in fishermen's homes orat the main harbor. IAIPU-BINACIONAL (1987) and ELETRONORTE/THEMAG (1987/89). Descripions of trophic interactiørs b) towns - these are small settlements, with fifteen thousands inhabiønts or less. Children and other among fishes and other organisms in the Amazon area were found in LOWE-McCONNEL (1975) and memþn of fisherman families helped in the location of residences. GOULDING (1980). General food habits of fresh water Brazilian fish and folk habits and beliefs are in c) riverine settlements - These include houses scattered along the banks of the Tocantins river. \ile used SANTOS (1962). motor boats of 25 and 45 HP. Houses werc sited f¡qn the boat and famfies from the first three consecutive houses interviewed; the next was skipped. By. this method, we tried to sample 75 lo of the fisherman families living along the river banks. Farmers werc not included, uriless they mentioned any kind of subsistence fishing.

r02 103 Results Tab. 1: lnterviews with fisherman families performed in the area of study; 'localities in the surroundings of Imperatriz: fishermen in this 1. Study site description and interviews category were interviewed in the market "O Povo Merece", during Imperatriz is the largest city snrdied (Table 1). Afær the construction of the fish landings; " source: IBGE @razilian Institute of Geography and Belém-Brasilia and São Luis-Imperatriz highways, migrants from many Brazilian states Suristics)(1987); "' rough estimarion of rhe number of fishermen came to Imperatriz (AROSO & LIMA 1984). Migration was suong during 1960-70, and (urban areas) from the Colônias de pescadores (fishermen's associa- contributed to the growth and commercial development of Imperatriz (KATZMAN trøs) Z-29 (Itnpe¡uriz) md Z-35 @streito). 1975). Fishermen of Imperatriz are locaæd in the vicinities of the market "O Povo Merece" (where hsh is sold to buyers). Most streets near fisherman houses are unpaved and houses are made of wood or are huts with mud walls. l,ocality Numbcr o? Tot¡l Number of Descarreto is found at the periphery of ltaguatins. It consists mainly of an uncovered Intervlews Populrtlon" Flshermen (thousends) main street, where huts made of stra'rvs or mud are found as well as a small church. Dry c') leaf palms are collecæd from the babaçú palm (Orbignya martiana) which is a very common palm tree in this region. It is used for roofs and walls. Imperatriz 52 235 Lajeado is near Lajeado a small town found the Belém-Brasilia highway and the Descarreto t7 15 T river. Porto Franco and Tocantinópolis are located at the Tocantins river banks, in Itaguatins 19 opposite sides. A ferry boat connects these cities. As in the other settlements, fisherman Lajeado l4 2 houses are found near the river banks, living in uncovered streets and in houses made Porto Franco 6 27 10 of straw or mud. Tocantinópolis 23 3t 40 The number of interviews are found in Table l. About 60 Vo of interviewed people Estreio 26 18 1 of both sexes are illiterate (including functional illiteraæs). Vy'omen have in average 4 Riverine 57 to 5 child¡en. Commercial fishermen are found mainly at Imperatriz, Porto Franco, Other' 20 Tocantinópolis and Esreito. In towns, such as Itaguatins, Descarreúo and lajeado, we Total found both commercial and part-time fishermen-peasants (or subsistence fishermen). 234 Riverine fishermen are usually fishermen-peasants. These categories tend ¡o use different fishing technologies. Commercial hshermen use motor boats and a variety of nets. Sub- sistence fishermen use paddled canoes, small nets and hook and line techniques Figure 4 shows the fish prefened as food by hshermen families. Pacu manteiga (BEGOSSI & PETRERE Jr. 1988). In addition, ,#omen and children often engage in (Mylossornn duriventre) and piabanha (Brycon sp.) were ciæd in all localities. Again, subsistence fishing with hook and lines along the river banks. Nevertheless, some women mandí-cabeça-de-ferro was specially mentioned by riverine fishermen (36 Eo). Curimatrá are engaged in commercial fishing. Important female activities are agriculture and and branquinha, the most importânt fîsh sold by fishermen, are not the most preferred babaçú collection (Figure 2). fish (quoted by 3 Vo and 16 7o of families, respecrively). Some fish are considered to have a bad taste, appe¿uance, or are simply said to be 2. Fish consumption and avoidances disgusting. These fish are usually not consumed. The results concerning the question Figures 3 o 7 show the fish ciæd by 5 7o or more of the interviewees. The scientific "which kind of fish you do not eat ?" are in figure 5. Cuiú-cuiú (species of Doradidae) names are found in Table 2. Mandí, piau and sardinha include a set of spe.cies, most are probably avoided due to their appeârance and hard texture; as the other Doradidae, important of which are mandí-cabeça-de-ferro (Pimelodus blochii) followed by is body is covered with bone plates. CandinÍ (Cetopsis sp.) eat worïns and dead animals. mandí-moela (Pimelodina fløvipinnis), piau-cabeça-gorda (Leporinus friderici) and As observed by one of the authors (FMSB) during fieldwork in the Amazon region, piau-vara (Schizodonvittatum). Sardinha is represented by species from different families fishermen from the INPA (National Institute of Amazon Research) consider the candiní and as such is not included in our discussion. very nasty, because they also eat dead human bodies. sANTos (1962), on the other Figure 3 shows fish ciæd by inærviewees as "fish consumed". Curimatii (Prochilodus hand, pointed out úat it is appreciated especially by native people. There are two kinds nigricans) was uniformly quoted as the most consumed in all localities. Mandicabeça- of candirú of common occurrence in Amazon: one type has few inches of length and is de-ferro was cited by 50 Vo of fishermen from Lajeado and by 57 7o of riverine a gill parasite and people are afraid of it because it can go inside the uretra and otier fishermen. This small fish is caught by hook and line and is a cheap fish in the market body cavities (family Trichomycteridae); the other, which is fhe one refened to in our (M. PETRERE Jr. pers. comm.). Branquinha (Psectrogaster arnazonica), an important study, is bigger and necrophagous (cetopsidae). other fish avoided are used in folk commercial fish in the Middle Tocantins, is consumed on a large scale in Imperatriz (60 medicine, such as pirara and raia (ray), among others (Table 3). Ray is probably under- 7o),where there are many commercial fishermen. represented in the interviews because some fishermen do not consider it ø be a fish.

104 105 Tab. 2: Feeding habirs of rhe fish quoæd during interviews.

N¡me Famlly Maln diet local sclentlfic

An¡anã O s I eo glossuttt b íc i¡r hos uttt Osteoglossidae Arthropods, fish 1i ' Barbado Pinirampus pirrumpu Pimelodidae Fish Bico{e-paro Sorubím littu Pimelodidae Shrimp, ñsh Branquinha P s ec lr o gasl cr anazonicø Curimatidae Detritus, periph¡on T Cachorra H ydr olycns s c otnbc roidc s Charecid¡e Fish t) Candinú Cetopsis cf , caecutiens C.etopsidae Insecl.s, worms, mud, necrophagous Capadinho r- P arauche nip e nu g øl eatus Auchenipæridae f Algae, fruits, Bryozoa, Tocantinsia piresi Auchenþæridae I anhropods, fish Caranha Èu Colossoma brachyponun Serrasalmidae Fruits, seeds Corvinha g ' P la iosc ion sq uanos is s imus Sci¡enidae Insects, Crustacea, fish ñ ,$ -d P. surinattænsis Sciaenidae atæ Cuiú-cuiú Megalodoras irwini Doradidae Fruits Platydoras coslalus Doradidae Fruits Fis.2: Plerdoras granulosus Doradidae Fmits, Crustaceå, mdluscs Pseuddoras niger Doradidae A woman breaking balcørçú (Orbignya nartiana) after collection. Notice the wall of the hot¡ses as well as the Fruits, molluscs, shrimp Curimatá Prochilodw nigricans Curimatidae basket made of leaves of the palm tree. Detritus, periphyton Filhote B r ac hy p Ia ty s t o na fr I anu nt o s um Pimelodidae Fish Jaraquí Senaprochildus brana Curimatidae Detritus, periph¡on Jaú Paulicea Some fish are taboed in certain circumstances. These hsh are considered to be lutk¿ni Pimelodidae Fish Lampreia R anp hic ht hys narrnoratus "carregado" or "reimoso". "Canegado" is a name used by fishing communities from Rhamphichtidae Worms and insect larvae R. roslralus Rhamphichtidae found in the mud different parts of Brazil and these food a¡e thought to worsen illness or to cause wounds S ternar c horamp hus mull e r i Apteronotidae health problems (BEGOSSI, unpubl.). MORAN (1974) noticed food restric- and other Sternopygru macrrtua Stemopygidae tions during illness, pregnancy, lactation and mensruation in Amazon communities. S. obtusirostris Stemopygidae "reimoso" Families from the Middle Tocantins also explained that fish considered to be Mandí Pinulod¿lla cristata Pimelodidae Insects should be avoided especially in case of illness, wounds, menstruation and afær child Pimelodina flavipinnis Pirnelodidae Benthos birth. However, we found a gndient from women who never eat certain fish species, to Pittæloúu blochií Pimelodidae Detritus, fruits, arthro- women and men who avoid them periodically (menstruation/child birth or disease) to pods, fish men who never avoid these species. The diseases most commonly mentioned for "man Piabanha Brycon sp. Characidae Insects, frogs, fish avoidances" were sexually transmitted diseases and colds. Wounds were also cited. Pacr¡ Mylossotna dutiventre Serrasalmidae Seeds, fruits Amazon food øboos for fish species were mainly reported by SMITH (1981)' Food Piau Laenolyø petite Anostomidae Fruits, seeds, arthropods restrictions (including the restriction of cerûain fish species) during illness or related to Leporinus afínis Anostomidae and larvae L. pregnancy were also reported in other countries, such as in Malaysia (McKAY 1980; friderici Anostomidae L. trilasciatus Anostomid¡e WILSON 1980) and India (FERRO-LUZZI1980a, b, c). The fish tabooed ("carregado" Schizdon vittattnt Anostomidae Seeds, fruits, roots or "reimoso") at the Tocantins river are shown in Figure 6. Scaleless fish, called "peixes Pir¡rara P hra c to ce p løl ns he mil iopter us Pimelodidae Fish, turtles de couro" (su¡ubim - PseudoplatystomoÍasciatrm, ba¡bado - Pinirampus pirinampu, ja(t Poraquê E I e ct r op hor us e I e c tr ic us Electrophoridae Shrimp, fish - Paulicea lutkeni, and pirarara - Phraclocephalus lrcmiliopterus, afiroîg others) are Raia Potanolrygon moloro Potamotrygcridae Benthos considered as "reimoso". Some fish "with scales", such as curimatá and jaraqui usually P. hystrix Potamotrygonidae Benthos as having a (Semaprochilodus brama), are also taboo€d. These fish are considered Disceus thayeri Potamotryguridae Benthos "strong meat".

106 107 Table 2: Continued Tab. 3: Folk medicinal uses of ñsh on the Tocantins river, based on interviews with fisherman families (quotations îrom 5 lo or more interviewees). Nemes in bold print refer to the most quoted fish part Name Family Main diet and disease. * called "rendição"; ** "dor de lado"; *** Brazilian rum made of sugar cane. "aU"O local sclentlfic **** wounds specially caused by Leishm¡niosis (called "Leisho") disease and by sexually transmitted diseases.

Sardinha many spP. Engraulidae Seeds, fruits and Characidae msects Flsh P¡rt used Dlse¡se Receipt Clupeidae

Surubim P s e ud o p la ty s I o rna fas c iat wt Pimelodid¡e Fish Surubim chicote S or ub imic htys pla nice ps Pimelodidae Fish Raia f¡t ¡sthme eaten or passed Traíra Hoplias n¿labørícus Erythrinidae Shrimp, fish cough, cold or massaged in the Voador Argonectes scapularís Hemiodqrtidae Detritus, benthos and earaches area affected

H e miodt¿s unittt¿c ulatw Hemiodqrtidae periph¡on skin bums H, argenteus Hemiodq¡tidae hemia* sPur preumonia** toasted and drunk wi¡h water or "cachaça"***

3. Fish used ¡n folk medicine Jaú f¡t skln burns eaten, Passed Rays are the most important fish used in folk medicine ffigure 7). The parts of rhe skin rhewnatism or massaged fish used rio prepare the medicines and the illnesses related ûo them are shot¡/n in Table asthma 3. The effectiveness of these medicines will not be discussed here. SANTOS (1962) cold eanches mentioned medicinal effects of the fermented fat of pirarara and the belief that a skin toothaches disease called "puru-puru" is caused by consumption of this fish. The meat of poraquê chilblains is seldom consumed and the electric discharges of the fish are considered effective in meat wounds**r. eaten the treatnent of rheumatic pains (SANTOS lf)62). Poraquê f¡t rheum¡tlsm passed, massaged Fish fat is commonly used in folk medicine. A few interviewees quoted ttre fat of ("Peixe elétrico") cold or eaten other fish such as cachora, dourada, piabanha and piau-cabeça-gorda (for earaches), asthma cuiú-cuiú (rheumatism), curimatá (eye pains), pacu manteiga (sexually transmitted aches diseases), surubim (cold) and of any scaleless fish ("peixes de couro") to Eeat skin bones snake-bites rea is drunk burns. Three interviewees told us that the fat of any fish could be used to cure skin Surubim-chicote me¡t Iælshm¡nlosis the cooked fish burns. A flour made from the scales of pirosca (Arapainw gigøs) serve against asthma tuberculosis is eaten and pneumonia and of curimatá to cure wounds. One interviewee believed that spitting Pirarara fat skin bums passed or in the mouth of cará (species of Cichlidae) and throwing the fish back into the river rheumatism massaged would cure tuberculosis. cough, cold Some fish, such as branquinha (Psectrogaster anuzonica) and ubarana (Anodus asthma wounds elongatus) are not used for any specific disease, but a¡e recommended to be eaten by ill hemia persons. Other animals a¡e also considered as having curative properties. The fat and Corvina otolith peinful a tea is made with skin of caiman are used against epilepsy and snake-bites; and the fat of "sucurí" (South urln¡tlon a flour from the American Anaconda) is used for rheumatism, common colds and skin burns. hemorrhages toasted stone Some fish seem more imporønt in certain localities than in others. For example, ray snake-bites was mentioned by 64 7o of fishermen from Estreito, 57 7o from Tocantinópolis, 50 7o Traíra f¡t e¡raches put in the ear from Imperatriz and only by 7 Vo from l"ajeado. Jaú was mentioned by 56 7o from diarrhe¡ the broth from the Estreito, 48 Vo from Tocantinópolis, 36 7o from Lajeado and only 3 Vo from Imperatriz. cooked fish is drunk Poraquê was mentioned by 80 7o of fishermen from Porto Franco and surubim chicote meat any disease eaten by 40 Vo of fishermen from Imperatriz (this fish was not mentioned at Porto Franco, Tocantinópolis, Estreito, and at the banks of the river from Ma¡anhão Søæ). It might

108 109 ."c .tt'*

aono' CON9UMEO ;.ôj' AVOIDED 7 cJ 6

..5 60 .ù' 50 (, 2(D =L ()E o 40 z z 4 l,rl l¡l G= =fr otd l¡J 3 l¡J ãq I I U' U) c.* L l! 2 L 20 ol! o

ðe o a o

o Fig.3: F I S H Fish consumed mentioned by at least 5 7o of the families interviewed (n = 233). Fig.5: F I s H Fish avoided as food and mendoned by at leasr 5 lo of the families interviewed (n 2zz). r(0 =

.¡uoo . òv ,¡s t0' ,o6'"ooo rSJ' ,""'

,od '"att" .o1' o90 _s 7 Qo' P R É.FE R R E D 50 .ss øù 60 TABOOED (9 o 40 4 z F 50 l- o õ z z l! 4 lr¡ (r l!= ft I 3 t¡, þp" I I L 2 2 lt olL 2 l! o ðC o òC

o o F ISH Fig. 6: FIS H Fig. 4: Fish classified as "tabooed": fish avoided during illness, mensrruarion or after child birth. Total families interviewed Fish preferred mentioned by at least 5 7o of the families interviewed (n = 185). = 196. "Peixes de cor¡ro" means ariy scaleless fish.

110 111 ,00 1. Fish diet $0' The feeding habits of the fish mentioned by inærviewees are in Table 2. Comparing this table with Figures 3 to 6 we observe that camivorous fish, most of which with ""-:1î.ì"" ?" nJ--,rs piscivorous habis, are the fish avoided as food (75 Vo) or tâbooed (57 Vo).In contrast, 45 MEDICINAL 64 7o of fish consumed and 62 7o of fish prefened can be considered as herbivorous or i.o\on* \ as detrivorous. Secondary consumers are often avoided as food while primary consumers tend to be prefened as food. A similar pattern was observed in a previous resea¡ch in rt-^ 35 a maritime fishing community, locaæd in southeast Brazil Island) (BEGOSSI, o ^ô @úzios .¡rd unpubl.). The probability of acquiring toxins increases as one move.s up in the trophic E= o 30 levels (toxins may accumulaæ at top levels). Cases of toxic fish exist, such as the well .;ù l¡¡z known (LEWIS 2 ciguatera 1984). Thus, the avoidance of piscivorous hsh, especially in É, ^(ø case illness (fish tâbooed) t¡¡ ,oo of may be an adaptive behavior. I ¿U ed . a-t L -ô' lr 2. Medicinal fish, or "the drugstore hypothesis" L "tt- o BEGOSSI (1989a) argued that fish considered to be important in ttre treatrnent of

ðe o diseases of isolated populations may be tabooed as tio be available for folk medicine. Many people feel confolable when living in cities, near medical centres and drugstores. Nanre is the "drugstore" of isolated human populations. Plants used for medical care are o collecæd and also cultivated; hsh species used for this purpose should be preserved. Tocantins riverine fishermen seem to conform to this behavior. It should be pointed FISH out that in Brazil, living in or near cities does not guarantee medical care. Pharmaceutical Fig. 7: medicines are expensive and most poor people rely on home made medicines. For Medicinal fish. These fish (or their fat) are used to curr asthma, skin bums and rheumatic pains, among others. example, fishermen from Imperatriz, the largest city in our study, are also using fish as Interviews = 189. medicines. If the "drugstore hypothesis" holds, we should expect that the most. important fish species used for folk medical care would also be the most avoided as food. A Spearman be asked if these differences are related to environmental circumstances. Or, in other correlation was calculated for all the fish mentioned in interviews (Table 4). Results words, can these fïsh uses serve as "biological indicators" for local diseases? Iæishmani- indicate that the fish species most avoided are also used for medical purposes (r = 0.54, osis and malaria are common in this part of Brazil. The fact tttat fishermen from p < 0.02, df = 20). Imperatriz mentioned proportionality more the use of surubim chicote to cure this disease could represent the importânce of Leishmaniosis in this area. Without data on Tab. 4: Spearman's correlation among fish utilizatim ar rhe Tocantins river local tropical disease this hypothesis cannot be tested. These data are not easily available based qt interviews with fishermen. * 0.05 < p < 0.10; ** p < 0.02, because they are not recorded since many sick people do not go to hospitals, particularly df = ?-0. in isolated settlements. Medicinal animals are usually avoided as food, and this will be discussed in the next section. Medicinals Avoided

Discussion Avoided 0.54** Tabooed 0.û7 0.43* Two kind of explanations seem related to the food avoidances found at the Tocantins river. The first is the probable knowledge of fish food habits by fishermen. The second concerns the "usefulness" of fish or maintenance of available fish to be used for medical 3. Other aspects: parasits, lipids and toxins purposes (BEGOSSI 1989a). Position in the food web and importance in folk medicine are the most importånt facton explaining fish avoidances for consumption on the Tocantins. Other minor factors also play a role in fisherman diet" For example, internal organic feâtures of specific fish or fish "diseases" influence this behavior.

tt2 ll3 The family Pimelodidae, one of tlre biggest families of Amazon hsh (SANTOS er conûast" herbivorous and detrivorous fish are preferred as food. The probability of al. 1984), is a very important group of fish avoided as food and,/or tabooed. Among 12 acquiring toxins in the food web could explain this pauern of food preference. Fat fish fish cited as avoided in interviews (only common names are considered: if mandí in- are well accepted by people and the calories obtained from them probably support it. cludes three species of Pimelodidae, we counted as one), 4 are Pimelodidae, and among Tab. 5: Fat ccntent of fìsh species (7o). Source: JUNK (1976) in SMIIH (1981) (= S) and JI,NK (1985) 13 ciæd as tabooed, 7 are Pimelodidae. tile found only Pimelodidae among fish I the (= Ð. O Species mentioned in inæwiews; I genera mentioned and O family menticned; consumed (out of l1) and prefened (out of 8). Pimelodidae are usually parasited by ce- * Pseudodoras niger. stoids. ITAIPU-BINACIONAL (198Ð, lisæd four pathogenic species of cestoids prot€o- cephalids of Pimelodidae: (Januella glandicepdus, Megathylacus brooki, Travassiela avitelina nd Peltidocotyle rugosa).In the same Itaipu reservoir, 605 Pimelodidae oi the Species l,oc¡l n¡me Fat (%) species Paulicea lwlæni, Pimelodus maculatus, Pinirampus pirinampu, Pseudoplaystoma corruscans, Sorubim lima, Hemisorubim platyrhynchus, Iheringichthys labrosus, Pseudopimelodus zwtgaro and Pimelodel/ø sp. were analyzrd. Among fhese,450 were I Brac hy platystona fl av icans 0.6 (Ð parasiæd by cestoids, found in their digestive tracL These parasiæs are responsible for O B.ftlameuosum Filhoæ 0.6 (Ð, 0.6 (s) I B. vaillanti 4.8 (I) localized sores and make the fish more susceptible to seconda¡y infectious diseases. I Brycon sp. Piabanha 22.8 (S) Another aspect is the lipid content of fish. Members of fisherman families during I Brycon cf. melanoplerus 15.2 (r) interviews commented that "reimoso" fish contain "gra*a" (fat). SMITH (1981) poinæd I Colossona bid¿ns 0.5 (Ð out that fat fish a¡e avoided in Amazon. Data on lipid content of Amazon fish species I C oloss ottu rutcro ponurn 1.5 (Ð are not readily available. However, JttNK (1976, cited in SMITH 1981) and JUNK Q H e mioà¿s wti¡nac ulat us r6.6 (Ð (1985) provided a list the lipid content some species. The of for fish most eaten and J lzporinus fasciatus Piau ls.8 (s) prefened by fisherman families (curimatá, mandí, piau, paci-manteiga and piabanha) I Mylossona sp. Pacr¡ 24.8 (s) have high lipid contents (Table 5). The most avoided and tabooed fish (cuiú-cuiú, O Osteoglossum bic irr hosum Aruanã 2.6 (s), 2.6 (Ð pirarara, surubim and curimatií) have low lipid contents, except for curimat¿í. Thus, O Orydoras niger * Cuiú-cuiú 0.2 (s), 0.4 (Ð contrarily to SMITH's (1981) hypothesis, fat fish are not r.he mosr tåbooed. Acnrally, if O P hractocephalw lumil iopterus Pirarara 0.2 (Ð there is any tendency, it, is that. fat ñsh are the most eaten and preferred fish. O Pbulodinaflavipímis Mandí 13.6 (Ð According úo evolutionary ecological models, such as "optimal foraging theory" O Pimelodt¿s blochii Mandí 10.2 (Ð @YKE 1984; STEPF{ENS & KREBS 1986), conventional foragers should maximize O P la g iosc ion s q uamos iss inus Cowina 5.8 (s) their net energy intake, which a¡e usually measu¡ed in calories. BEGOSSI (1989b) O Prochildus nigricans Curimatá 17.4 (S) O P s e udo platyst ona c iatun Surubim 2.0 (s), 0.3 (J) observed the imporønce of variables other than calories, such as fish bones and prices, fas I Semaprochilodus taenwus 8.2 (Ð in the diet (especially fish) of families from Búzios Island (SE Brazil). If fishermen from I S. uuignc e.4 (s) the Tocantins river are "optimal foragers", all other things being equal, they should be selecting the prey with the highest caloric contenq which appears to be the case. As lipids are more important than proteins for determining caloric contents flVHO 1973), eating fat fish would be an oprimal strategy. However, data on lipid content for many Medicinal fish are avoided as food. The theories of HARRIS (1985) seem úo fit in this Amazon fish species are lacking and the few data available do not allow us to reach food avoidance case. Animals that are too importânt for other tasks should not be eaten, reliable conclusions. and the only way ¡o avoid temptation is to classify them as "bad to eat". The presence Finally, certain fish can also be toxic. There âre no avarlable data on the natural of parasites in the Pimelodidae as well as their piscivorous habit reinforces an existing toxicity of Amazon fishes. However, in a list of toxic plants and animals found in taboo. Food avoidances may also not be related to animal conservation. Animals may ROSENBERG (1987), the ray Potamotrygon sp. and six species of carhsh are not be avoided or tabood especially to be preserved: some ¿ìre not consumed because mentioned. The toxicity of some fish can be related to fish food habits, since many ñsh there are better uses for them. This snrdy suggests that to understand food choice and acquire tlreir toxicity through the food web. However, more studies are needed to have taboos \ve must look for other uses of food items or for the context in which the a complete understanding of the complex relationships of feeding pattems of hsh and populaúon lives. of local fishermen,

4. Conclusions Summary Food preferences and avoidances among fisherman families from the Tocantins river can be partially explained by two major factors: fish food habits and fish "usefulness". Food choice pa¡tems arc a polemic subject in Human Ecology. We studied fish uses and choices Fish that are secondary consumers a¡e avoided æ food, especially piscivorous fish. In (consumption, medicine, preference and avoidance) among families of fishermen from the Middle Tocantins

t14 ll5 (Brazil). The study sites included cities, towns and riverine se$lemer¡ts along 100 km of the river. BEGOSSI, A. & M. PETRERE Ir. (1988): Utilização de rccr¡rsos aquíticos e tecnologia entre pescadores do Questionnaires were used to guide the interviews with the families (fishermen and/or wives). Curimatá Médio Tocantins (GO e MA). - In: DIEGUES, A.C. & R.R. SALES (eds.): II Encqrtro de Ciências (Prochilodtts nigrícans) is lhe most cqrsumed fish and pacu-merireiga (Mytossonø durivenrre) rhe mos¡. Sociais e o mar no Brasil, Coletânea de trabalhos apresent¡dos. São Paulo, IOUSP/F. FORD/UICN: preferred fish. Other fish, such as species of Doradidae (called "cuiú-cuiú"), are avoided as food. Surubim 158-169. (Pseudoplatystoma fasciatwm) is avoided as food in specific health ci¡cumstances. The fat of ray is ccnmonly BRffiKI, H.A. (1972): Peixes de água doce do Estado de São Paulo. Sistemática. - In: Poluição e Piscicr¡ltura. used to treat asthma. The main explanation for the avoidances of certain fish may be their "usefulness" (used - C.I.B.P.U. São Paulo, Faculdade de Saúde Priblica da USP, Instituto de Pesca: 79-108. for medicine, Spearman, r = 0.54, p < 0.û2) and their piscivorous habits. More research is needed to look for CATELLA, A.C. & G.E. TORRES (1984): Observações sobre o espectro e estratégia alimentar do a general pattem of fish avoidances. peixe-cachorro, Acestrorhynchus lacuslris (REINHARDT, 1974) (Characidae, Acestrorhynchini), do resewatório de Tres Marias - rio São Francisco, MG. - Anais do Serninário Regicral de Ecologia 4: 103-125. Resumo ELETRONORTE|IHEMAG ENGENHARLA (1987/89): Bacia do Médio Tocantins. Aproveitamento hidrclétrico do sítio Serra Quebrada - Sano Antonio: estudod de viabilidade (ictiofauna). - Relatórios Escolhas alimentares são um ponto polêmico ern Ecologia Humana. Estudamos a uriliz¿ção e escolha técnicos ne 1,2,3,4, 5, 6 and final (internal repons). (consumo, medicina caseira, preferåtcia e rcjeição) de peixes entre familias de pescadores do Médio Tocantins FERRO-LUZZ, G.E. (1980a). Food avoidances at prberty and menstruation in Tamilnad. - In: Food, ecology @rasil). As áreas de estudo incluíram cidades, vilas e cqnunidades ribeirinhas localizadas ao longo de l(X) and culture. New York, Gordqr & Bre¡ch, Science Publ. Inc.: 93-100. km do rio. Utilizamos questiørários para guiar as entrevistas. Curimatá (Præhilodus nigricaw) é o peixe mais FERRO-LU77I, G.E. (1980b): Food avoidances of pr€gnånt women in Tamilnad. - In: Food, ecology and consumido e pacu-manteiga (Mylossona durivenlre) o mais preferido. Outros peixes, como algumas est'cies culture. New York, Gordon & Brcach, Science Publ. Inc.: 101-108. de Doradidae (chamadas de cuiú-cuiú) são evitados como alimento. Peixes como o surubim (Pseudoplatystorna FERRO-LU77[, G.E. (1980c): Food avoidances during the puerperium and lactation in Tamilnad. - In: Food, fasciatum) são tabus alimenlares relacionados à saúde individual. A gordura da raia (Poråmotrygonidae) é ecology and culture. New York, Gordon & Bre¿ch, Sciatce Publ. Inc.: 109-118. (1980): bastante usada no tratamento då asma. As principais razões que @ern explicar porque certos peixes não são GOULDING, M. The fishes and the forcst: exploratians in Amazon natural history. - Berkeley, Univ. consumidos são a sua "utilidade" (medicina caseira, Spearman, r = 0.54, p < 0.02) e seus hábitos piscivoros. of Califomia Press. Entretanto, são necessários mais estudos para encontrar um padrão geral de tabus alimentares com relação à GOULDING, M. & E.J.G. FERREIRA (1984): Shrimp-eating fishes and a case of prey-switching in Amazon peixes. rive¡. - Revisø Brasil. Zool. 2:85-97. HARRIS, M. (1977): Cannibals & kings. - New York, Vintage Books. HARRIS, M. (1985): Good to eat: riddles of fæd and culture. - New York, Simon & Schuster. Acknowledgments HARRIS, M. (1987): Food ways: historic¿l overview and theoretical prolegomenon. - In: HARRIS, M. & E.B. ROSS (eds.): Food and Evolutior¡. - Philadelphia, Temple Univ. Press: 57-92. 'We IBGE (1987): Anuário Estatístico do Brasil, 1986. - Rio de Janeiro, Centro de Documentação e Disseminação are very graleful to ELETRONORTE-THEMAG ENGENHARIA for financial support of rhe de informações. fieldwork; to CNPq (Conselho Nacional de Desenvolvimento Cienrífico e Tecnológico) for a scholanhip to INPAÆLETRONORTE (1980/83): Estudo da ecologia e controle ambiental na região do reservatório da UHE AB and for the grant ne 402.228/89-8|ZOßVIPQ; to FÄPERI for rhe grant @-291170368,89) during writing de Tucurui - ictiofauna. - Relatórios técnicos no 1, 2,3 and final (intemal reports). this paper. To Íílio C. GARAVELLO (Universidade Federal de São Carlos) for rhe fish identificariur; Bniulio ITAIPU BINACIONAL (1987): Ictiofauna e Biologia pesqueira. - FUEM-SUREHMÀ/ITAIPU BINACIONAL, JIMENEZ (Oak Ridge Naticral Laboratory), É¿ca C¡,RAU¡,SCHI (Univenidade Federal do Rio de Ianeiro), 2 vols. Miguel PETRERE Jr. (Univenidade Estadual Paulista) and Scou W. HOEFLE @onrificia Univenidade JUNK, W.J. (1976): Biologia de água doce e pesc¡ no interior. - Relatório Anual do INPA, Manaus, Instituto Católica do Rio de Janeiro) for many helpful ccrnments and suggestians qr this manuscripr. Finally, to the Nacional de Pesquisas da Amaánia: 105 pp, "Colônias" Z-29 and Z-35, md to the conmunities on the Tocantins river for their kind coop€raticn. IIJNK, WJ. (1985): Temporary fat storage, an adaptatian of some fish species to the waterlevel fluctuations and related envirqrmental changes of the Amazon river. - Amazoniana 9(3): 315-351. KATzuAN, M.T. (1975): Regiaral developnent policy in Brazil: the role of growth poles and development References highways in Goias. - J. Ecm. Developrn. Cult. Change 24:15-lÚl. LEWIS, N.D. (l9M): Ciguatera-parameters of a tropical health problern - Human Ecology 12:253-274. AROSO, O.C.L. & O'C. UMA (1984): Áreas culturais linguísticas maranhenses. - Ciência e Cultura 36: LOWE-McCONNEL, R.H. (1975): Fish communities in trçical freshwaten: their distribution, ecology and 995-998. evolution. - [.ondon, Longman. BASSO, E.B. (1978): Cqnmenrs. - Currenr Anthropology 19: 16-17, McKAY, D.A. (1980): Food, illness and folk medicine: insights from Ulu Trengganu, West Malaysia. - In: BEGOSSI A' (1989a): Tabus aliment¿rcs na Ilha dos Búzios, uma comunidade de pescadores. - In: DIEGUES, Food, ecology and cr¡lture. - New York, Gordon & Breach, Science Publ. Inc.: 6l-66. A.C. (ed.): III Encontro de Ciências Sociais e o mar no Brasil, Coletânea de trabalhos apresentådos. São MESSER, E. (1984): Anthropological perspectives on diet. - Ann. Rev. Anthropol. 13:.2O5-249. Pado, IOUSP/F. FORD/IJICN: 253 -262, MOR.A,N, E. (1974): The adapive system of the Amazonian Caboclo. - I¡¡: WAGLEY, C. (ed.): Man in the BEGOSSI, A (1989b): Food diversity and droice, and æchnology in a Brazjlian fishing communtity @úzios Amazon. - Gainesville, Univ. Florida Press. Island, Sõo Paulo søte). - PhD Dissertarion, univ. califomia, Davis, USA (I.JMI # s919534). MOURÃO, G. de M. & G.E. TORRES (1984): Especrro alimenrâr e atividade predatória da cowina, BEGOSSL A. & J.C. GARAVELLO (1990): Nores on the ethnoicthyology of fishermen from the Tocantins Pøchyttnts squamipinnis Sciaenidae) no reservatório de Tres Marias, rio São Francisco, MG. - river. - Acta Ämazônica 20:341-351. @isces, Anais do Seminário Regional de Ecologia 4:295-3W.

116 It7 NOMURA, H, (1975): Alimenøção de trcs est'cies de peixes do gênero Asryan¿-r BAIRD & GIRARD, 1854 (Osteichthyes, Characidae) do rio Mogi Guaçú. - Rev. Brasil. Biol. 35: 595-614. NOMURA, H. & L. NEMOTO (1983): Alguns caracteres meristicoc e da biologia do cascudo, Plecostomt¿s paulinus IHERING, 1905 @isces, Loricarüdae) do rio Mogi Guaçú, SP. - Rev. "O Solo" 75: 64-79. NOMIJRA, H. & A.C.D. TAVEIRA (1979): Biologia da sagvrn,Curbtatus elegans STEINDACHNER, 1974 do rio Mogi Guaçú, São Paulo (Osteichthyes, Curimatidae). - Rev. Br¿sil. Biol. 39:331-339. NOMURA, H., MENEZES, J.F.S. & M.V.F.A. SOUZA (1983): Evidências da ação predadora do tucunaÉ, Cichla ocellaris, sobrc a tilþia do Congo,Tiløpia retúalli @isces, Cichlidae). - Anais do Seminário Regional de Ecologia 3:223-231, NOMURA, H., NEMOTO, L. & I.M.M. MUELLER (1981): Alimentação de seis esffcies de peixes do gênero Plecostomt¿sV ALBAUM, 1782 @isces, Loricariidae) do rio Mogi-Guagi. - Anais do Seminário Regional de Ecologia 2:389-405. PYKE, G.H. (1984): Optimal foraging theory: a critical review. - Ann. Rev. Ecol. Sysr. l5:523-575. ROSENBERG, P. (ed.) (1987): Comman names index, poisanous ¡nimals, plants and bacteria. - Toxicon 25: 799-890. ROSS, E.B. (1978): Food taboos, diet, and hunting stntegy: the adaptation to animals in Amazqr cr¡ltural ecology. - Current Anthropology 19: l-36. ROSS, E.B. (1987): An overview of trends in dietary variation frun hunter-gatherer to modem capitalist societies. - In: HARRIS, M. & E.B. ROSS (eds.): Food and Evolution. - Philadelphia, Temple Univ. Press: 7 -57. SAHLINS, M. (1976): Cr¡lture and practical reason. - Chicago, Univ. Chicago Press. SANTOS, E. (1962): Peixes de água doce: vida e costumes dos peixes do Brasil, Znd ed. - Rio de Janeiro, Briguiet & Cia Editores. SANTOS, G.M., JÉGU, M. & B. MERONA (1984): Catálogo de peixes comerciais do Baixo Rio Tocanrins. - Manaus, ELETRONORTE/CNPq/INPA. SMITH, N.G.H. (1981): Man, fishes and the Amazon. - New York, Columbia Univ. Press: 87-91. STIPHENS, D.W. & J.R. KREBS (1986): Foraging rheory. - Princeton, Princeton Univ. Press. Wll-SON, C.S. (1980): Food taboos at childbirth: the Malay example. - In: Food, ecology and cuhure. - New York, Gordan & Breach, Science Publ. lnc.:67-74. WHO (1973): Energy and protein requirements. - World Health Organization Technical Report 522: l-121.

118