We should point out here that there are very few real hunter-gatherer societies in the area under study in this report. There are 6 known such cases, and they are either populations that have turned away from agriculture following contact with Europeans or conflicts with other indigenous populations (Akulio, Yuqui, Ava-Canoeiros, Guaja), or they are communities which have a symbiotic relationship with agricultural societies (Maku and, formerly, Mura).
Nearly all Amazonian populations are involved in hunting, fishing, gathering and agriculture. The only differences are in the relative importance given to each activity.
This relative importance is obviously deeply conditioned by seasonal rythms and influential factors such as the low water levels during the dry season, favourable to fishing, or the strong insolation of the dry season, indispensable for drying and burning new plots for cultivation ; or again, the fructification of plants during the rainy season which allows for concentrations of game and specific hunting strategies. Another common characteristic is the sexual division of labour, agriculture being generally carried out by women, and hunting and fishing by men, while both sexes are involved in gathering activities.
In the 20th century, indigenous populations are mostly concentrated in small communities which rarely exceed 150 people, the most basic unit being approximately 50 inhabitants, though even smaller groups do exist.
In past centuries, much larger communities could be found, but these were either sedentary populations living in the rich alluvial plains, or artificial concentrations of people set up by missionaries and the colonization process. In recent times however, there have been a few exceptions to the general trend : the populations in the Brazilian cerrado combined, until recently, small seasonal groups scattered during the dry season, and, concentrations in big villages which could house up to 2 000 people during the rainy season (Shavante, Apinayé, Northern Kayapo) (ZARUR, 1979).
In the course of the past 40 years, spatial organization has been altered by a policy of sedentarization implemented by governmental agencies and missionaries, thus encouraging a new emergence of large settlements (300 people or more).
The village community is the basic political, sociological and economic, and ecological relevant unit for the whole of Greater Amazonia. Whatever type of social organization it is based on, it is usually made up of people closely related, gathered around one leader whose power is generally weak. Strong levels of cooperation sanctioned by ritualized exchanges are generally opposed to highly individualistic behaviour which periodically leads to communities splitting up and recombining elsewhere.
Habitat varies between a type of communal housing and a group of houses based on family units. However, there seems to have been a large increase in the construction of these more complex settlements as contact was made with exogenous and non-indigenous populations, though some societies did also have groups of houses traditionally.
In stable environmental conditions, a total area of 250 km2 seems to be needed to sustain a community of 150 people. Such a big surface is not to be explained by the requirements of agriculture : it is necessary for hunting, fishing and food gathering activities, of course, but also to provide the infinite variety of raw materials used in crafts, to build houses, boats, etc. These raw materials have the added characteristic of being scattered over a vast area due to the high diversity of the flora in Greater Amazonia ; two isolated individual plants of the same species may be as much as several miles apart.
Population density is always low, usually around 0,2 inhab./km2 ; areas with more than 2 inhab./km2 are still today located either on extremely rich agricultural land (Kampa, Amueshe), or along the main rivers where the presence of rich soils is combined with the availability of protein resources (fish) (Shipibo, Tikuna).
The mobility of indigenous communities has been the source of many debates among specialists. In most cases, "semi-nomads" hardly applies and it would be more accurate to speak in terms of modest, or even cyclical, mobility. Several specialists (Carneiro, 1961) have most definitely shown that here again, mobility is not motivated by the fact that agricultural land may have been exhausted, but by a combination of two factors : the limited sources of protein and concern over not depleting such supplies as are available. The most mobile populations are of course those for whom hunting is more important, but over the last two centuries, the main factor inducing mobility has definitely been the continual attempt to get away from non-indigenous populations.
Traditionally, disputes over land between different groups only occurred when one group encroached on another group's territory to obtain some of the resources exploited or even monopolized by the latter. It is only since the experience of the massive invasion of their land by non-indigenous populations that native American Indians have been compelled to reconsider their position. The idea of boundaries for any single territory emerged and it is now followed by virulent demands that these boundaries be protected on a national, and these days international, level.
It is important to stress in this context that a policy aiming at setting up indigenous areas for native Amazonian populations is only compatible with indigenous concepts of spatial organization if it overlaps with the territory they actually exploit. But, as we shall see further on in this report, in areas along the 'pioneer frontier', such conditions are far from being met and are even impossible to implement now.
Most indigenous societies of Greater Amazonia have some form of agriculture.
Though it may seem foolish to try to draw up an overall picture of uniformity (some societies have been engaged in agriculture for longer than others ; the staple crop may vary, etc), indigenous agriculture, in optimum conditions, does have a certain amount of characteristic features (BECKERMAN, 1987 ; Human Ecology, 1983 ; GéLY, 1984).
Rotation of plots is prevalent. This entails clearing every year or every two years, a patch of primary forest or high secondary forest (over 15 years old). The patch is burnt and setting fire to the forest provides a temporary form of fertilization for soils which are mostly poor and acid. Some land is regularly covered by alluvial deposits, and a few are good basic soils. The average size of a plot is 0,5 ha. Tree trunks are kept as firewood to last throughout the agricultural cycle. Planting starts with the first rains, setts with seeds, cuttings and other live plant parts are provided by plots that are already under cultivation. Plots have one staple crop, manioc in a majority of cases, which covers 80 to 90 % of the plot's surface area ; a small number of ethnic groups (Shipibo, Yanomami) prefer plantain, while a rare few grow maize (Araweté, Amawaka). A whole variety of secondary crops occupy the remaining area. They are either concentrated individually on tiny plots, intercropped, or planted in concentric rings around the staple crop (Human Ecology, ibid.). Varieties for each cultivated plant are often highly difersified though manioc, both the sweet and the bitter varieties, prevails (the extreme case being that of the Tukano who cultivate over one hundred different clones ; KERR & CLEMENT, 1980). The crops need little attention. Harvesting starts with maize but is spread out over 2 years for manioc and longer for plantain.
In any single year each family works on a young plot, a mature plot and an old plot. At the end of the agricultural cycle, the plot is abandonned and reverts to forest, this being the ultimate and crucial goal put forward by the native American Indians themselves as a way of stimulating higher concentrations of game species (LINARES, 1976 ; GRENAND, P., 1992).
Manioc is a very robust plant which can be harvested throughout the year, and which therefore does not require storage ; it provides high yields in carbohydrates (DUFOUR, 1988). With an agriculture based on manioc as the staple food crop, Amazonian societies are left with a lot of extra time that may thus be devoted to other activities. Proteins must be absorbed in sufficient quantities to counteract the manioc's residual toxins, and SPONSEL (1989) was able to show that it is consequently vital that men devote this extra time to fishing and hunting. On the other hand, detoxication of manioc and the culinary art involved in the preparation of its by-products, mobilizes a huge part of the women's labour time, thus making for a balanced sexual division of labour.
Manioc being one of the few high-yield plants (an average of 15 t/ha), it is clear that, when abiding by the requirements of forest regeneration in an Amazonian environment which has not been degraded (GRENAND, F., 1993), the combination of hunting, fishing, gathering and agriculture set up by these populations is an optimum ecological adaptation.
Most societies in Greater Amazonia combine both activities. However, some engage more exclusively in hunting (Kraho, Nambikwara,...) while others devote more time to fishing (Xinguanos).
The main weapons were (and still are to some extent) the bow and arrow with a great variety of arrowheads adapted to hunting different animal species. With such weapons one can catch terrestrial game (rodents, deer species, tapir and peccary), arboreal game (birds and monkeys), or water species (fish, turtles and caymans). However, in western Amazonia, hunting with blowpipes associated with the use of curares used to be widespread. Ichthyotoxic plants were and still are in general use for fishing in the whole area covered by this report. It is important to note that the techniques for setting traps are not nearly as developed in tropical America as they are in parts of Africa and Asia (COOPER, 1986).
For a long time, many scientists saw a correlation between the importance of hunting and/or fishing and the supposed inability of the environment to support sustained agricultural activities (Meggers, 1971, 1991). And the importance of hunting was attributed to such behaviour as semi-nomadism. The ever-increasing case-studies published over the last 25 years show that such correlations are the result of hasty conclusions, and that just about every combination does exist between the size of a community, group mobility and its exploitation of the environment. The factors underlying such combinations vary enormously and obviously involve the environment, together with sociological and political factors. However, the profusion of environmental factors does necessarily encourage the development of one particular activity rather than any other. The limited watercourses in the Brazilian cerrado imply that populations here are more likely to turn towards hunting ; similarly, the abundance of fish in white-water rivers, such as the Ucayali, explains why the Shipibo should have turned to fishing. But the reverse is equally true : the Tukano live along the black-water rivers and rely on them as their main source of protein though these rivers actually have few fish (ARHEM, 1976) ; the Xinguanos have as much game as fish on their territory, but have specialized in fishing. Such a choice is always linked to a strong symbolic representation, socially relevant and expressed in terms of food taboos.
Nowhere does hunting, fishing or the combination of both activities provide for more than 50 % of the diet of indigenous Amazonian populations (MILTON, 1991). But the ideological prevalence of these activities is an indication of how essential they are (CARNEIRO, 1970 ; BALéE, 1985). Not being able to engage in them accelerates, usually irrevocably, the collapse of the social structure.
In terms of the protection of indigenous populations in Greater Amazonia, it would be extremely dangerous to make a distinction between populations consuming vast amounts of wild products and others that would be more cautious in their consumption of such products. Whichever activity predominates, these societies have developed a way of life ensuring the reproduction of the resources they use, by regularly moving their villages for instance (BAKSH, 1985 : SPONSEL, 1992).
It is interesting to note that wherever non-indigenous populations have become self-subsistent, for various reasons beyond the scope of this study, they have usually adopted agricultural, halieutic and cynegetic techniques identical to those of neighbouring indigenous populations (PARKER, 1985 ; GRENAND, P. & GRENAND, F., 1990). But whenever indigenous societies were under increasing pressure from outside, led to drastically reducing their territory, and consequently lost their political and social structure, they in turn became the ones destroying the environment, just like non-indigenous populations have been seen to do (REDFORD & ROBINSON, 1987 ; ROBINSOn & REDFORD, 1991 ; MILTON, 1992). In this particular context, one should also bear in mind the technological alterations that have occurred over the last 50 years, mainly through the introduction of fire-arms, various nets, head lamps and speed boats. When indigenous societies have coherent internal strength, foreign technology is usually only used occasionally, to provide for their own consumption, and it is of no consequence for the general yearly cycle of activities. Disintegration begins when these same groups enter the vicious circle of commercial hunting and fishing.
The word 'gathering' tends to encompass a whole set of activities that cannot be included in any of the three main categories mentioned above, to wit, agriculture, hunting and fishing. Here, the scope of the term will be restricted to activities strictly connected with the annual cycle of these societies's self-subsistent activities.
Detailed studies have been carried out over the last few years on cross-sections of forest and the specific societies using them. They show that the people know most of the different plant species growing in the area, except the rare and unimportant ones. Besides, 1 hectare of forest contains an estimated 60 % of the species that are commonly used (BOOM, 1987 ; MILLIKEN et al., 1992 ; GRENAND, P., 1992).
To use an expression coined by POSEY (1982), this makes native American Indians the only true 'forest conservationists', in as much as no other population can claim to know more than they do on their Amazonian environment.
Moreover, recent research has shown that a number of natural habitats have actually been manipulated by humans thus aiming at increasing its rich diversity. BALéE (1988) has put forward evidence of this for the Guaja and the Urubu, and ANDERSON and POSEY (1985) for the Kayapo. New examples appear constantly, to such an extent that one may wonder whether the Amazon basin is not just simply one huge anthropized biome. It is difficult to apprehend the extent and impact of these manipulations today, both because of the lower numbers of indigenous individuals and because of the present disruptions the environment has to put up with (ROOSEVELT, 1989 ; BALéE & CAMPBELL, 1990). Another important indicator is the presence of systems of agroforestery, especially in the western half of the Amazon basin. The technique aims at reproducing on a small surface the natural stratified structure of the forest, but selecting only useful species (DENEVAN & PADOCH, 1987 ; GUILLAUMET et al., 1990).
The knowledge implied and the practices themselves show that human intervention can be a positive element in the management of the Amazonian forest . However, there is a serious lack of quantitative data for the weight which different species represent in these people's diets.
Gathering actually covers a wide variety of activities : collecting plant foodstuffs ; collecting animal foodstuffs (honeys, eggs, insects, shellfish, etc), gathering medicinal, hallucinogenic or poisonous plants and all the plants used for a technnological purpose : the natural environment is what provides native American Indians with weapons, the structure and roof of their houses, rafts and boats, the materials for making baskets, dyes, strings and ropes, handles for tools, clay for pots and an infinite variety of other products.
Finally, one should draw a clear distinction between this form of gathering and extractivist economies that have developed since the 18th century around the Amazonian forest's natural products (see III - C). Though originally based on knowledge learnt from indigenous populations, such economies responded, and still do, to an external demand : the characteristic of this demand is that it concentrates on a limited number of products at given times and in specific areas, thus leading to an excessive exploitation of these products. The truth of this pattern is illustrated today in the attempts aiming at integrating 'extrativism' into a form of sustainable development in Greater Amazonia : the aim is precisely to reintroduce diversification as a criteria for the selection of products, so as to reduce the destructive nature of such activities.
Though we still know little and certainly underestimate the scope of gathering and collecting activities, they must definitely be taken into account in defining the minimum area which is to be allotted to each indigenous society.
Native American populations traditionally have a varied, balanced and self-subsistent diet (BéNéFICE et al., 1989 ; DUFOUR, 1988). This would tend to show that such societies with a wide span of diversified activities not only do not harm the equilibrium of the Amazonian forest, but even contribute significantly to the environment.
It is important to remember that foraging activities or agriculture can only fit into this pattern if based on traditional methods. This is why it is virtually impossible to even try to perpetuate these societies without at the same time having an overall general policy aiming at protecting the Amazonian forest.
As soon as indigenous societies are tempted into commercializing their activities, the usual process sets in : the structure of the annual cycle of activities is undermined, the sexual division of labour is disrupted, there is no moderation on the exploitation of natural products, therefore leading to their rarefaction, poor diet, debts and increased poverty in the community, and destruction of the natural habitat. At this stage, because of their negative impact on the ecosystem, these populations are no different than their neighbouring mestizos or migrant populations.