Mobility and evolving frontier settlements.
The case of Central Roraima State, Brazil
Alexandre M. A. Diniz
Assistant Professor
Department of Geography
UFMG - Brazil
Aldiniz@bhnet.com.br
Abstract
The purpose of this work is to assess the interplay between the evolution of frontiers and the associated
geographical mobility/circulation in the Brazilian Amazon at the settlement level. To address this objective, I
propose and test the mobility and evolving agriculture frontier settlements framework building on the
literature on Amazonian mobility and frontier evolution stage models. The proposed framework breaks up the
evolutionary process of individual agriculture settlements in four different phases – pioneer, transitional,
consolidated, and urbanized. Data to test the validity of the proposed model is gathered in Central Roraima
State, Brazil via survey instrument. Results demonstrate that the path of agriculture frontier settlement does
have an impact upon migration selectivity; migrant source regions; migration histories; predominant
migration type; migration pull factors; and circulation spatial and temporal pattern. In the end, a schematic
view of the entire process is presented, along a description of predominant mobility patterns and behaviors at
each level of evolution.
Key words: frontier evolution; human mobility; Amazon region; population geography; Roraima State.
1
Introduction
Frontier expansion has historically been an integral part of the development strategies of all countries sharing
the Amazon basin. It’s conquest has been portrayed as the panacea for many social and economic problems,
as it serves to redistribute population from densely inhabited areas, provides land and income to the land-less,
promotes regional development, and serves countries’ geopolitical interests (Mougeot, 1982a; Martine, 1984;
Sewastynowicz, 1986; Sternberg, 1987; Findley, 1988; Henriques, 1988; Foresta, 1992; Shrestha et al, 1993).
Based on these perspectives, a series of incentive plans for the occupation of the Amazon frontier has been set
forth by various governments over recent decades epitomized by incentive packages to boost domestic and
foreign capital investment and directed colonization programs.
The exploration of the Brazilian Amazon, however, overwhelms the experiences of other countries sharing
the region given the magnitude of centrally planned investments in roads, railroad, industries, mineral
resources exploration, agriculture, ranching, and colonization projects. Consequently, the regional economy
and population increased many-fold, as thousands of migrants flooded the Amazon in response to State
investments and incentives. Nonetheless, the Brazilian government policies have been highly inconsistent,
favoring peasants at times and national and multinational corporations at others. This erratic approach has led
to widespread conflicts and chronic displacement of settlers, re-migration, and circulation. Moreover, the
colonization projects implemented in the region have displayed low migrant retention rates, which in turn has
fueled the mobility system. In fact, some scholars argue that the Amazonian frontier no longer constitutes an
alternative to the social pressures taking place in rural Brazil, as most of the local resources have already been
claimed by the big players (Martins, 1980; Sawyer, 1981; 1984). Nevertheless, the Amazon region continues
to be perceived as the peasants’ save haven by many and migrants keep arriving. The years to come promise
an augmentation of conflicts over the local resources and an intensification of mobility, as the laissez-faire
approach adopted by recent administrations have done little to diminish the internal social problems.
The myriad of studies on the Brazilian Amazon display a bias for environmentally related investigations. At
this juncture, we know more about the ecological impact of settlements than we know about the evolution of
settlements and mobility. The literature on Amazonian mobility, on the other hand, is essentially descriptive
and focuses primarily on the characteristics of movers and migration patterns; whereas, the few theoretical
works pose sweeping generalizations about the displacement power of development and capitalist relations
without much empirical evidence. Similarly, the literature on frontier evolution is impregnated with constructs
that postulate that frontier development follows a set of hierarchical stages, moving from pre-capitalist to
capitalist oriented forms of production. These scholars suggest that as a given frontier settlement evolves
through time, in-migration wanes and earlier settlers are displaced by incoming firms and large scale farmers
to more backward parts of the frontier or into nearby cities (Martins, 1975; Foweraker, 1981; Browder and
Godfrey, 1990). These models, however, are essentially focused on the economic and urban evolution of
frontier settlements, failing to explore the impacts that the changing environment has on mobility.
The primary goal of this study is to fill some of these gaps by assessing the interplay between the evolution of
agricultural frontier settlements and the associated geographical mobility/circulation in agriculture settlements
of the Brazilian Amazon. Within this broader objective lies a few specific ones, namely to explore the
transformations promoted by the evolutionary process upon migration selectivity; migrant source regions;
migration histories; predominant migration type; migration pull factors; and circulation spatial and temporal
patterns.
While economic development has undoubtedly taken place in the Brazilian Amazon, social development lags
behind. Small-scale farmers and land-less peasants are constantly being reshuffled amidst different
evolutionary waves. Without intervention, the endless mobility is likely to be perpetuated. This study provides
a better understanding of the settlement system and the variegated forms of geographical mobility/circulation
embedded in the formation and demise of places. This understanding will contribute to policy debate and
discussion regarding the promotion of more stable settlements in the region and enhancement of quality of life
among settlers in the Brazilian Amazon.
2
Frontier evolution
Frontier areas have caught the collective imaginary of many societies with icons of pristine and abundant
lands. Nevertheless, the term “frontier” has different meanings depending on who is behind the glasses. For
some, it represents an escape valve for societies under social and economic pressure; a get away place where
well-to-do and stressed urban dwellers can turn to; and a means to promote individualism, which in turn
would lead to political freedom and democracy (Turner, 1920). For the countries sharing the Amazon basin,
frontier colonization has been an integral part of their economic development strategies, as it has been
portrayed as the solution for many social and economic problems. But before getting any further, it is
necessary to discuss the concept of "frontier".
Hennessy, (1978:3) suggests that the term “frontier” is associated with demographic expansion movements
over non-occupied or insufficiently occupied land. Mayhew (1997:184) furthers the concept by postulating
that “the frontier represents the part of the country which lies on the limit of the settled area”. Therefore, a
frontier settlement marks the furthest advance of civilization within a state. Neiva (1949), however, presents
a more sophisticated definition, by making a distinction between demographic and economic frontiers.
According to this view, the process of occupation of pristine areas is marked by the arrival of the
“demographic front” before the “economic one” makes itself present. The “demographic front” is composed
of small-scale producers, such as farmers and artisans, who tend to be the pioneer agents at frontier areas.
Capitalist enterprises and large-scale producers, on the other hand, represent the “economic front”.
Extending the work of Neiva, Martins (1996) envisions the frontier as an area marked by three well-defined
domains: demographic frontier; expansion front; and economic frontier. The demographic frontier represents
the limit of the settled area, beyond which lie indigenous populations and pristine areas. The economic
frontier, on the other hand, is marked by the extensive and territorial reproduction of capital, epitomized by
the conversion of land in merchandise, and Indian and peasants in waged laborers. Between the demographic
frontier and the economic frontier lies the expansion front, a transition zone marked by populations yet to be
incorporated by the economic frontier. The expansion front represents the movement of civilized population
and economic activities, somehow organized by the market. This front encompasses not only agricultural
entrepreneurs, businessmen, cities, political and juridical institutions, but also the poor non-Indigenous
populations, like garimpeiros1, cattlemen, rubber tapers, and peasants who work on the fringe of the market.
Nevertheless, the demographic frontier is also related to the expansion of capital, which cannot be regarded as
capitalist given the absence of organized labor and land markets.
This expansion front is characterized by the expansion of commerce and exchange networks, in which money
is regularly absent. These exchanges are arbitrated by those who maintain control over resources and the
manpower behind the exploitation of such resources, namely Indians and peasants. Market operates based on
monopolistic village entrepreneurs, who control not only commerce, but also working relations through
peonage, and enslavement by debt. Within the Brazilian context, it is difficult to identify the exact location of
the boundary between civilization and pristine lands; nonetheless, the Amazon region is widely recognized as
the frontier, incorporating within its limits areas dominated by expansion fronts, and demographic and
economic frontiers.
The exploration of frontier areas worldwide have been an important aspect of the evolution of many societies,
inspiring the construction of different theoretical proposals aimed at the comprehension of the process of
frontier occupation and evolution. Previous theoretical construes can be broadly classified with respect to the
scale in which they operate. Regional development frameworks (Neiva,1949; Bylund 1960; Hudson 1968;
De Vance,1970; Muller, 1977; Meyer 1980; Lisansky, 1990) are focused on the spatial diffusion and urban
hierarchical functions of a set of places across a given area. Such perspectives tend to envision the evolution
of entire frontier zones passing through a series of stages, moving from a few isolated settlements, to areas
marked by well-defined hierarchy of places. Such perspectives provide interesting ideas on the rise of cities
and urban systems, but given the scale in which they were conceptualized, these models are not directly
applicable to the present study. The focus here is on the interplay between the evolution of agricultural
frontier settlements and mobility.
1
Informal miners.
3
Place specific frameworks, on the other hand, are focused on the evolution of particular frontier settlements
over time, being conceptualized primarily based on the reality of the Amazon region. Theorists working on
this front postulate that frontier settlements evolve through a set of hierarchical stages, moving from a precapitalistic type economy, characterized by the lack of labor and land markets, to more capitalist oriented
forms of production. These models also tend to emphasize the clashes between different interest groups over
frontier resources, which leads to waning in-migration and the displacement of earlier settlers by incoming
entrepreneurs. Therefore, mobility is an explicit component; however, it is conceptualized as a mere response
to the structural forces of frontier evolution. Moreover, these models consistently overlook the variegated
forms of mobility and processes occurring at different stages of the evolutionary spectrum.
Foweraker´s (1981) construe emphasizes the process in which frontier areas become progressively connected
to the national economy. The transformation of pristine frontier areas into “productive societies“ is
understood as going through a three-phase transition: non-capitalist, pre-capitalist and capitalist society
stages. The idea of transition implies changes in production relations, and in the markets for goods, land, and
labor (Foweraker, 1981: 27). At the non-capitalist stage, the frontier economy based on extraction being
remarkably isolated. The sphere of exchange is limited to outside markets for one or two commodities locally
produced. This stage is marked by the lack of markets for land or labor; and social relations of production are
mainly servile. There is an emerging commodity sector, which will eventually favor the occupation of the
region by peasants.
The expansion of the commodity sector generates greater migratory flows into the region and more intensive
extractive activity, characterizing the second stage of frontier expansion. Land begins to be bought and sold,
but prices represent only what is on the land, rather than the land itself. These changes bring about regular
production of commodities and an emerging market for land; nevertheless labor markets are still absent.
Social relations of production are established by the growing commodity sector, or are mixed forms of servile
and capitalist relations. On the other hand, the final stage of the frontier expansion is identified by an intense
migratory flow into the region and established access to the national economy. The economy is no longer
based on extraction, but rather on agriculture, becoming increasingly capitalized. Land prices rise and land
ownership becomes concentrated. Capitalist relations of production are dominant and a labor market is finally
achieved. As the frontier moves into its final stage of evolution, economic activity generally becomes more
differentiated, whereas, social division of labor more complex (Foweraker, 1981: 27-39).
Henkel (1982) and Findley (1988) propose similar frontier settlement models focused on the behavior of
individuals. They postulate that the evolution of frontier settlements undergoes three incremental stages:
pioneering, commercialization, and abandonment/consolidation. During the pioneering stage, settlers are
concerned primarily with occupying the land and bringing it into production. Colonists rely on their large
families and on each other to clear the land. In order to fulfill their cash requirements, colonists engage in
wage employment at nearby farms. The commercialization stage is marked by the construction and
improvement of transportation linkages, which provide access to markets and stimulate production. The
abandonment/consolidation stage is characterized by a bifurcation of land tenure. In the process, earlier
arrivals sell their land, motivated by debts, lack of capital or environmentally related reasons, while the betteroff consolidators buy the improved land adding to their existing holdings. As large land holdings, usually
devoted to livestock or commercial crops, require little year round labor inputs, the population retention
potential of the colonization zones is compromised. As colonists who remained in the area still depend on
waged labor for their survival, they tend to leave in search of another plot of land. Thus, land concentration
leads to extensive population turnover and re-migration, not only among those who sold their plots, but also
among those who are willing stay.
Based upon the Brazilian Amazonian experience Browder and Godfrey (1990) developed the "Amazonian
landscape change and urban transition" model. This model postulates that Amazonian settlement typically
occurs in a progressive sequence, promoted by a gradual incorporation of a dependent peripheral region into
the larger national economy. The evolution of the Amazonian frontier is divided in five stages: native
subsistence economy, resource-extractive frontier, pioneer agricultural frontier, relict frontier and urban
primacy and rural depopulation phases. Despite its more unifying nature, Browder and Godfrey's proposal
was engendered primarily based on their experience in Southern Pará, which has its share of peculiarities. Its
4
relative proximity with the Northeast region, the abundant mineral and vegetal resources, the various large
scale federal projects, and the presence of various agents competing for power and resources made Southern
Pará the most complex and dynamic sub-region of the Amazon. Thus, at more remote areas of the Amazon, it
is possible that settlements do not undergo all five stages as postulated by Browder and Godfrey.
The models of frontier evolution or expansion described above, emphasize different aspects of the
transformations taking place at the settlement level. Neiva (1959) and Martins(1996) stress the conflicts
taking place along the process, namely the clashes between demographic and economic fronts. Henkel (1982)
and Findley (1988) emphasize the strategies adopted by colonists at the various stages of the evolutionary
process. Foweraker (1981) discusses primarily the structural aspects of the process, underlining foremost the
social and economic relations taking place at the frontier. Browder and Godfrey (1990), on the other hand
tackle the issue from a more integrative way, combining many elements that appear isolated in previous
models, while highlighting the process of genesis and growth of urban centers.
It is important to stress that at least a brief reference to mobility is present in all models of frontier evolution
discussed here. Yet, despite recognizing the presence and importance of human mobility in the region, these
models treat the phenomenon as a mere byproduct of the evolutionary process, neglecting to account for
different mobility patterns and processes taking place at the various stages of the process. A study that
combines such models, while stressing the selectivity, motivations, and strategies adopted by settlers at
varying levels of frontier evolution is yet to be advanced. This study attempts to fill this lacuna by working
from a composite of the main ideas rooted in the reviewed models, and advancing a series of hypotheses
relative to mobility in the ever-changing geography of frontier colonization settlements.
Amazonian mobility
The literature on frontier mobility in Brazil is based on descriptive studies focused on inter-regional and intraregional migration patterns. Emphasis is placed on the urbanization of the region (Almeida and David, 1981;
Martine, 1981; Sawyer, 1981,1982, 1984, 1986, and 1987; Wood and Schmink, 1983; Wood and Wilson,
1984; Santos, 1984; Martine, 1984; Oliveira, 1986; Da Silva, 1986; Bentes, 1986; Lavinas, 1987; Ferreira,
1987; Jardim, 1987; Vasconcelos, 1988; Martine, 1989; Becker, 1985; 1990; Godfrey, 1990; Volbeda, 1996).
These studies tend to be grounded on historical/structural perspectives, and like the place specific models of
frontier development, utilize as explanations for the massive intra-regional mobility the displacement power
of the advance of the capitalist mode of production, development and the inconsistent government policy
towards the region. Nonetheless, under these perspectives, mobility is generally treated in an unsystematic
fashion, and explanations are given in a non-empirical mode.
Another set of studies focusing on migration in the Brazilian Amazon explores the characteristics of migrants.
These studies tend to be based on survey instruments conducted in various areas of the Amazon (Henriques,
1985; 1986 and 1988; Sawyer and Carvalho, 1986; Oliveira, 1986; Da Silva, 1986; Bentes, 1986; Lisansky,
1990; Crócia 1994 and 1995; MacMillan, 1995; Diniz, 1997). Overall, migrants to rural Amazônia tend to be
young (15 to 35) and predominantly males given the hardship of the local environment. Migrants also display
fairly low educational attainment levels, for the most part less than four years of formal education. Rural
migrants engage chiefly in peasant agriculture and extractive activities, which tend to be conducted outside
the formal wage market. Circulation and occupational mobility are common survival strategies, as they
engage in seasonal employment in farms, ranches, and even in urban Amazônia in order to meet their cash
demands. Rural migrants display a tendency to come from rural areas of the Amazon region or from other
Brazilian regions and display long migration histories (large number of moves). On the other hand, a sizable
portion of migrants arriving in urban Amazônia display different characteristics. On average these better off
individuals are better educated and display a more gender-balanced nature. The majority of migrants arriving
in urban Amazônia is drawn from rural areas of the Amazon region, but also come from other urban places
inside and outside the region.
At this juncture, we know that mobility is ubiquitously intense in the Amazon. The literature on Amazonian
migration indicates that in-migrants are for the most part negatively selected, displaying intense circulation
behavior within the region. The transformations promoted by the frontier evolutionary process, commanded
5
foremost by the advance of the economic front and the capitalist mode of production are understood as the
chief processes behind the regional mobility. Nonetheless, such accounts are impregnated by biased eyes and
lack empirical evidence. Moreover, empirical studies tend to be descriptive in nature, lacking a more
analytical inclination. Therefore, there is room for a study that empirically explores how the differing levels
of frontier evolution engender different forms of migration and circulation at the settlement level.
Anticipated findings
Building on the literature on frontier evolution, and Amazonian mobility, an integrative framework is
advanced to study the various mobility patterns and processes taking place along the evolution of agriculture
frontier settlements. The proposed framework is based on the assumption that as a given frontier settlement
grows and changes through time, profound changes occur in terms of migration selectivity; migrant source
regions; migration histories; predominant migration type; migration pull factors; and circulation spatial and
temporal patterns.
The interplay between the structural forces underlying the evolution of frontiers and geographical mobility are
examined at four different types of settlement: pioneer, transitional, consolidated and urbanized. The
characterization of these stages incorporate a series of empirical and theoretical works discussed earlier in this
paper, most specially the ones pertaining to the place specific models of frontier evolution. Thus, the
definition of each level of evolution represents a condensation of all pertinent ideas set forth by the revised
literature on frontier evolution. The portrayal of each phase is followed by a series of hypotheses related to the
expected predominant mobility patterns and processes occurring at each evolutionary step.
Pioneer frontiers are marked by lack of markets for land and labor, rapid population growth via migration,
wide availability of land, poor transportation networks, predominantly rural populations, low levels of
deforestation, and economies based on extraction and peasant agriculture. Such places match Martin’s (1975)
demographic frontier, and Foweraker’s (1981) and Browder and Godfrey’s (1990) early development stages.
Here, settlers are concerned primarily with occupying the land and bringing it into production (Henkel, 1982;
Findley, 1988). Due to lack of capital and labor markets, settlers rely on each other during this labor-intensive
phase. Many settlers also seek off-farm employment at more developed portions of the frontier to meet their
cash needs.
Expected mobility
It is expected that due to the remoteness and the uncertainty associated with these frontiers, migration chains
will dominate the system, constituting the primary migration type, linking pioneer frontiers with very specific
rural origins in the Northeast and the Amazon region. Given the incipient nature of these frontiers, the lack of
market for land and labor, and the physical stamina involved in settling and surviving in the harsh equatorial
forest, colonists are expected to be poorly educated, young and predominantly male (Henriques, 1986;
Findley, 1988). As transportation networks are poor at this stage, circulation is expected to be spatially and
temporally restrained, as individuals should travel for brief periods of time within short distances, primarily
due to shifting agriculture and seasonal employment in farms and ranches in neighboring more developed
portions of the region.
Transitional frontiers are characterized by economies based on increasing commercial agriculture, and some
extractive activities. Land begins to be bought and sold, and an incipient labor market arise (Foweraker,
1981). Transportation and communication links are improved, easing the movement of crops to local and
regional markets and migration into the region. With the escalation of migration, land becomes scarce, and the
number of land-less peasants working as sharecroppers and land invaders increase (Martins, 1975; Browder
and Godfrey, 1990). The new roads also make land more valuable, attracting capitalized entrepreneurs,
marking the arrival or intensification of Martin’s (1975) “economic” frontier. Land values skyrocket and
become increasingly concentrated, as the rate of deforestation progresses. Earlier settlers begin out-migrating
in response to debt, legal battles over land ownership, lack of capital, inability to market their surpluses given
the presence of mass producers, and changes in the productive systems introduced by livestock and
6
commercial crop (Henkel, 1982; Findley, 1988). This phase is analogous to Hudson’s (1969) competition
phase, described as the struggle of individuals to maintain their domains and/or increase their land holdings.
Parallel to land consolidation, embryonic urban areas within the settlement and nearby areas start to grow.
Expected mobility
Individuals moving into such places should rely on migration chains, but not to the same degree as pioneer
settlers, given the improvements in transportation and communication networks. These improvements should
also contribute to the arrival of a significant numbers of re-migrants or displaced peasants from various places
within the Amazon region, broadening the geographic scope of settlers’ origin places (Mougeot, 1982; 1983;
1986; Oliveira, 1986). Thus, repeat migration is expected to be the predominant migration type. Settlers
during this stage should still be primarily male, uneducated, and young due to the harshness of the area;
however, given the incipient land market, migrants might bring some financial resources. Given the presence
of improved transportation linkages, and the increasing need for cash introduced by the encroaching capitalist
mode of production, settlers are expected to circulate among a vast array of places, engaging in temporary
employment in various places in order to boost incomes and warrant subsistence (Becker, 1990).
Consolidated frontiers are marked by economies based on agriculture, costly land prices, land concentration,
presence of large-scale agricultural and ranching enterprises, predominantly capitalist relations of production,
and waged labor (Foweraker, 1981). Artificial pastures replace a large proportion of the properties, and the
natural vegetation is confined to the back of the rural properties. Land now serves large speculative holdings
and capital-intensive agribusiness. Consolidated frontiers are also characterized by sparse populations, as a
large portion of earlier settlers have already left for other areas. The remaining ones were converted into wage
laborers at farms and ranches, or struggle to maintain their plots. Such places also rely on year-round
transportation facilities and regular transport lines (Browder and Godfrey; 1990).
Expected mobility
Unlike landowners at other stages of evolution, consolidated ones should live outside the agriculture
settlement. Instead, farm and ranch employees should comprise the dwellers of such areas. These settlers
should be composed of former peasants, and late arriving migrants. However, owing to the nature of life and
work at consolidated frontier areas, these settlers should still display low levels of education, training, and
young age. Given the age of settlement, higher infrastructure levels, and better connections with other areas,
settlers should have been drawn from a variety of places, ranging from nearby settlements to various parts of
the country. As such individuals hold formal or informal jobs at large agricultural establishments, circulation
is expected to be negligible, as they have year-around work and stable incomes.
Urbanized frontiers are understood as the outcome of the embryonic urban nuclei created at the heart of
colonization and settlement projects2. The growth of urbanized frontiers is a natural outcome of the process of
frontier evolution at surrounding areas, thus becoming inseparable entities from the agricultural projects from
which they originated. As these areas progress in the evolutionary spectrum, rural-to-urban migration flows
intensify, augmenting the urban nuclei3. Such urbanized frontiers may in fact grow enough to become
município4 seats, materializing the claims of Becker (1990) and Godfrey (1992), who envision such places as
concentration and redistribution centers for displaced peasants5. On the other hand, these areas also attract a
pool of better-off migrants pulled by the installation of various governmental offices, military bases, and
service and commerce businesses (Sawyer and Carvalho, 1986; Da Silva, 1986; Diniz, 1997).
2 Article number 64 of the Land Statute states that parcels of land in colonization areas may be of two types: urban and rural. Rural parcels are destined to
agricultural work, where peasant families may also dwell. Urban plots are located at the center of the community, where administrative, cooperative,
commercial, artesian, public service and industrial activities congregate. Here peasant-families may also reside if they please. In this dichotomy lies the
foundation of many urban nuclei in the Amazon. Such embryonic urban centers become the major reference point for settlers within colonization and
settlement projects.
3 Truth is that in most cases, such urban places receive the same names of the originating colonization projects. Most municípios seats in Roraima State, for
instance, appeared in this fashion (Cantá, Mucajaí, São Luiz da Baliza and São João Anauá).
4
Brazilian sub-state political units.
5
For a detailed description of the evolutionary process of frontier towns see Volbeda (1996).
7
Expected mobility
Urbanized frontiers have good communication links with their hinterlands and with the rest of the country and
are expected to draw migrants from a much broader pool of rural and urban origins (Bentes, 1986). It is also
expected that such places should attract a variety of migrants, ranging from displaced peasants to better-off
urban migrants, who reach the settlement in search of formal tertiary jobs. Thus, the selectivity process is
more complex, as in-migrants can display low levels of human capital, and long migration histories on one
hand; and higher educational levels, training and short migration histories on another.
Circulation is expected to be intense at this evolutionary level, being facilitated by well-developed
transportation networks, tight urban job markets. Nonetheless, circulation patterns are expected to be
bifurcated, as negatively selected individuals, due their low skill levels and the demands imposed by the urban
life, are expected to circulate on a regular basis in search of urban and rural odd-jobs. On the other hand, jobseeking circulation is expected to be negligible among better-off migrants, as they hold formal and decently
paying jobs. Instead, circulation among these individuals should be conducted primarily leisure motivated.
Methods
The advanced set of hypotheses was designed to explore the reality of frontier communities in the Brazilian
Amazon, which have evolved from spontaneous or official agricultural settlements, in which small-scale
producers predominated at pioneering stages. Nonetheless, it would be an insurmountable task to test the
validity of the set of hypotheses in the entire Brazilian Amazon. Instead, evidence to test these anticipated
results was gathered at the colonization areas of Central Roraima State.
Roraima constitutes a prime study area for exploring the interplay between frontier evolution and mobility
(Figure 1). The State is one of the nine units of the Brazilian Amazon, located in the northernmost portion of
the country. The state encompasses 225,116,1 Km², having most of its land lying in the Northern Hemisphere.
Despite its remoteness, the state was the fastest growing area of Brazil during the 1980's and early 1990’s,
constituting the latest booming area in the Brazilian Amazon. Despite this tremendous population growth,
little is known of Roraima and most studies dealing with the Brazilian Amazon scarcely contain any reference
to the state (IBGE, 1992; MacMillan, 1995). Moreover, Roraima mirrors many current and past features and
problems of other Amazonian areas as it has experienced fast development, massive road building,
colonization programs, competition for land, destruction of natural vegetation, conflicts between indigenous
groups and settlers, and a rampant urbanization process (Furley and Mougeot, 1994).
As secondary data are unavailable at the settlement level, information to test the set of proposed hypotheses
had to be gathered via survey instruments. The selection of respondents proved to be a Herculean task, as the
colonization projects of Central Roraima had to be first classified in terms of their stage of evolution, before
an appropriate sampling design could be developed. With the composite four-stage frontier evolutionary path
in mind, information necessary for the classification of colonies was gauged via conversations with local
dwellers, personnel of land-granting agencies6, and thorough field observation. The discriminating factors
used in the classification of agricultural settlements were age of the colony/project, land concentration status,
evolution of land and labor markets, population density, deforestation rates, and implanted infrastructure. It is
important to stress that this qualitative approach was utilized given the lack of reliable data on the local
demographics, as well as on land and labor markets. Once the classification phase was complete, colonies
representative of each phase of the evolutionary path were sampled via multi-stage cluster sampling (Marconi
and Lakatos, 1996; and Hy, Feig and Regoli, 1983). In the process, three colonies (clusters) were randomly
selected among each evolutionary phase group.
6
INCRA and ITERAIMA.
8
Figure 1
Brazil, Political Division
Roraima
Amapá
Amazonas
Pará
Maranhão
Piauí
Acre
Tocantins
Rondônia
Mato Grosso
Ceará
R. Gde Norte
Paraíba
Pernambuco
Alagoas
Sergipe
Bahia
DF
&
Minas Gerais
Mato G do Sul
São Paulo
Espírito Santo
Rio de Janeiro
Paraná
Santa Catarina
400
800
Kilometers
Rio Grande do Sul
Once clusters were selected, survey respondents (household heads) were sampled, based on a systematic
approach of household units within each pre-selected settlement (Blalock, 1979). Colonization areas can be
quite large, blazed with kilometers of secondary roads, along which, lots are found. Thus, fieldwork attempted
to minimize, as much as possible, physical dislocation by working from carefully designed interview routes.
Along these routes, household heads were interviewed in accordance with the pre-established sampling
interval (every fifth plot along the way, for instance). Table 1 presents a list of all selected survey sites, along
with a description of their location and stage of evolution. Fieldwork took place during the dry season,
between November/97 and March/98, when roads are more reliable and most peasants are busy preparing
their plots for plantation. The survey was based on a semi-structured instrument, which sought to explore
various aspects of peasants’ social-economic conditions, as well as their past and present mobility behaviors.
Table 1
Selected agricultural settlements
Project name
Alto Alegre
Vila Iracema
Cantá
Alto Alegre
Confiança I
Vila Iracema
São Francisco
Confiança II
Roxinho
Confiança III
Maranhão
Sumaúma
Município
Alto Alegre
Vila Iracema
Cantá
Alto Alegre
Cantá
Iracema
Bonfim
Cantá
Iracema
Cantá
Iracema
Mucajaí
Stage of evolution # of interviews
Urbanized
30
Urbanized
30
Urbanized
30
Consolidated
30
Consolidated
30
Consolidated
30
Transitional
30
Transitional
30
Transitional
30
Pioneer
30
Pioneer
30
Pioneer
30
Total
360
*Estimates of INCRA and ITERAIMA officials
9
Settlers’ socioeconomic profile
Corroborating the results of earlier studies (Henriques, 1985; 1986 and 1988; Sawyer ad Carvalho, 1986;
Lisansky, 1990; Crócia 1994; MacMillan, 1995 and Diniz, 1997), survey results demonstrate that settlers of
Central Roraima are for the most part negatively selected. Nevertheless, the process of frontier evolution is
selective of certain socioeconomic traits, revealing a significant degree of heterogeneity among surveyed
subjects. It was detected that the vast majority of Central Roraima households are led by males (91.1%)
(Table 2); nevertheless, female-headed households tend to be more prevalent at urbanized frontier areas. This
phenomenon can be partially explained by a survival strategy adopted by a sizable number of Central
Roraima settlers, in which women and younger children settle at urbanized frontiers, where amenities,
services, and odd jobs are generally available; while men and older children take care of rural plots. As part of
the scheme, women living at urbanized frontiers supplement family incomes by performing service-type jobs
such as sowing, washing, ironing, and cleaning for better-off families. Nonetheless, during the key phases of
the agriculture cycle, women and younger children leave the urbanized realm to help men and older children
performing certain labor-intensive tasks such as harvesting and planting at rural areas.
Table 2
Household head gender by stage of evolution
Male
Female
Total
Urbanized Consolidated Transitional Pioneer
71 (78.9)
87 (96.7)
83 (92.2)
87 (96.7)
19 (21.1)
3 (3.3)
7 (7.8)
3 (3.3)
90 (100)
90 (100)
90 (100)
90 (100)
Chi2 = 23.598 (0.0001) Tau B = 0.099 Eta = 0.256
Total
328 (91.1)
32 (8.9)
360(100)
Survey data also reveal that the majority of frontier migrants are formally or informally married (65.8%);
while, single individuals represent 26,7% of the researched universe (Table 3). When marital status at the time
of the survey is considered, there is no statistical difference among the stages of evolution, as the proportion
of married, single, divorced and widowed individuals are evenly distributed among the stages. Still, Table 3
shows a trend toward the concentration of married individuals at the pioneer stage. Such trend can be
explained by the fact that it is virtually impossible to overcome the harshness of such places, and the
strenuous work involved in traditional agriculture without the help of an established family, preferably with
numerous children.
Table 3
Marital status during survey by stage of frontier evolution
Marital Status
Single
Married*
Other
Total
Urbanized Consolidated Transitional Pioneer
24(26.7)
24(26.7)
27(30)
21(23.3)
59(65.5)
60(66.7)
51(56.7)
67(74.4)
7(7.8)
6(6.7)
12(13.3)
2(2.2)
90(100)
90(100)
90(100)
90(100)
Chi2 = 10.442 (0.107) Tau B = .019 Eta = 0.030
Total
96(26.7)
237(65.8)
27(7.5)
360(100)
Besides congregating a larger number of females, urbanized frontiers also display the largest households of
Central Roraima (Table 4). Notice how the average size of households at urbanized frontiers (4.96), is
significantly higher than the ones found in the remaining stages of frontier evolution. Conversely, pioneer and
consolidated areas are marked by the smaller households. Infrastructure levels, amenities, and the nature of
economic opportunities present at the more evolved frontier settings foster a different type of human
occupation, favoring the concentration of women and numerous children. Anova test confirms the difference;
while, Scheffe´s test (not shown) demonstrates that the statistical discrepancy lies exactly between the
urbanized frontiers and all the others stages (Table 4).
10
Table 4
Household size by stage of frontier evolution
Stage of frontier evolution Mean St. Dev.
Urbanized
4.96
2.34
Consolidated
3.60
2.61
Transitional
3.84
2.69
Pioneer
3.79
2.39
Total
4.05
2.56
F=5.391 (0.001)
CV
0.47
0.72
0.70
0.63
0.63
N
90
90
90
90
90
Frontier migrants arrived in the agricultural settlements of Central Roraima with an average age of 33.81
(Table 5). Survey results, however, frustrate the expectations of the proposed model, as the frontier evolution
process is not selective of migrants’ age as indicated by the statistical results presented in Table 5. Notice that
there is a slight, but not statistically significant, gradation among the stages of frontier evolution, with pioneer
settlers being the oldest group (average 35.52), followed by transitional (35.17), consolidated (33.33) and
urbanized (31.20) ones. Thus, despite the physical stamina demanded by the lifestyle at the least evolved
areas of the frontier - which should theoretically lower the mean age - one may be witnessing in Central
Roraima the materialization of a much broader migration trend. Namely, the fact that the migratory venture
tends to be dominated by young-adults.
Table 5
Age of household head during arrival at current destination
by stage of frontier evolution
Stage of evolution Mean St. Dev.
Urbanized
31.20 15.01
Consolidated
33.33 13.39
Transitional
35.17 12.38
Pioneer
35.52 11.25
Total
33.81 13.14
F= 2.067 (0.104)
CV
0.48
0.40
0.35
0.31
0.38
N
90
90
90
90
90
On the other hand, survey results corroborate the expectations embedded in the proposed model, attesting the
existence of an explicit relationship between the evolution of frontier settlements and settlers’ education. Over
75% of all subjects have less than 5 years of formal education; while, 36.1% are illiterate (Table 6). However,
better-educated individuals congregate at the more advanced stages; while, the least advanced stages are
dominated by uneducated settlers. Notice that at the urbanized and consolidated frontiers, an expressive
portion of in-migrants has more than 5 years of education (50%); whereas, at pioneer and transitional areas,
well-over 70% of migrants display less than 5 years of formal education. However, these data also
demonstrate the duality of urbanized frontiers, which attract and hold a sizable number of poorly educated
individuals and a host of better-off settlers, as anticipated.
Table 6
Years of schooling of the household head by stage of evolution
Schooling
Illiterate
1–4
5–8
≥9
Total
Urbanized Consolidated Transitional Pioneer
16 (17.8)
44(48.9)
27(30)
43(47.8)
29 (32.2)
27(30)
45(50)
40(44.4)
21(23.3)
17(18.9)
11(12.2)
6(6.7)
24(26.7)
2(2.2)
7(7.8)
1(1.1)
90(100)
90(100)
90(100)
90(100)
Chi2 = 72.772 (0.0001) Tau B = 0.80 Eta = 0.397
Total
130(36.1)
141(39.2)
55(15.3)
34(9.4)
360(100)
Survey results also show that most migrants living in rural Roraima own the plots of land they currently
occupy (61.5%) (Table 7). Nevertheless, the proportion of individuals dwelling in their own plots diminishes
with increasing levels of evolution. Note that 88.9% of migrants at pioneer frontiers own the land where they
live, contrasting with 54.4% of migrants in transitional frontiers and 41.1% in consolidated frontiers.
11
Therefore, these numbers confirm the expectations embedded in the models of frontier evolution (Henkel,
1982; Findley, 1988; Foweraker, 1981; and Browder and Godfrey, 1990) demonstrating an increasing
presence of land markets throughout the evolutionary process. One could dispute the claim that the absence of
landowners dwelling at the more evolved frontier areas is an indicator of the presence of land markets. After
all, original landowners could prosper enough to hire workers on a permanent basis, while dwelling at other
areas. Nevertheless, the reality of Central Roraima severely hampers the economic prosperity of peasants due
to a group of ecological, infrastructure and socioeconomic reasons. Within such context, absent landowners in
active rural properties, without fail, means that the property has exchanged owners. In fact, data presented by
Table 7 indicate that the path of evolution turns settlements once dominated by peasants and small-scale
farmers into areas marked by absent owners and large properties, which are occupied by waged workers.
Table 7
Land ownership status by stage of frontier evolution
Consolidated
Land Owner 37(41.1)
Worker
53(58.9)
Total
90(100)
Chi2 = 46.1990 (0.00001)
Transitional Pioneer
Total
49(54.4)
80(88.9) 166(61.5)
41(45.6)
10(11.1) 104(38.5)
90(100)
90(100)
270(100)
Tau B = 0.14220 Eta = 0.41365
Another evidence of the increasing presence of land markets has to do with the land acquisition process7.
Consolidated frontier settlers are more likely to have purchased the land they currently occupy (67.6%), than
migrants living at any other evolutionary phase, testifying the presence of a more developed land market in
such areas (Table 8). Transitional settlements display an intermediary position, with an equivalent proportion
of peasants that have purchased and invaded/claimed their plots. Similarly, a larger portion of peasants at
pioneer frontiers have acquired the plots they currently occupy through invasions/claims (41.3%); while, the
number of those who have formally purchased the plots (49.4%) is also expressive. These numbers challenge
the expectations embedded in the models of frontier evolution, indicating the presence of an incipient land
market at pioneer frontiers. In fact, it was observed at pioneer frontiers the buying and selling of land claims
much before the implementation of infrastructure by the land-granting institutions. However, given the
uncertainties associated with the legal aspect of such properties and the lack of access roads, land prices are
generally low8.
Table 8
Land acquisition process by stage of frontier evolution
Consolidated Transitional
Purchased
25(67.6)
25(51)
Donated by the land granting institutions
9(24.3)
5(10.2)
Invasion/Claim
3(8.1)
19(38.8)
Total
37(100)
49(100)
Chi2 = 15.535 (0.004) Tau B = 0.055 Eta = 0.249
Pioneer
32(40)
15(18.8)
33(41.3)
80(100)
Total
82(49.4)
29(17.5)
55(33.1)
166(100)
7
The questions regarding the status of land and production systems were made only to landowners, which total 166. Thus, in this portion of the analysis
urban dwellers and land-less rural workers are not contemplated.
8
A 50 ha land claim could be purchased in an invaded area during the fieldwork for US$ 150.00.
12
Migration9
Overall, the bulk of surveyed settlers originated in the North (Amazon) region (84.7%), most notably from
within the state of Roraima, which alone accounted for 72,2% of all interviewed in-migrants (Tables 9 and
10). This finding is symptomatic of difficulties associated with the establishment of stable livelihoods in
Roraima, besides denouncing the rampant intra-state mobility. In the process, migrants are drawn from
various states of the union, and after settling at any given portion of Roraima, are forced to re-migrate. On the
other hand, Northeast Brazil is the second most important migrant source region, being responsible for almost
12% of all settlers. Within the Northeast, Maranhão State plays a prominent role given its historical links with
Roraima State. Most Maranhenses10 originated in the northwestern portions of the state, an area where
frontier-like movements were prevalent during the 1950’s and 1960’s. During this period, Western Maranhão
attracted thousands of land-less peasants of Northeast Brazil, who reached the region in response to the vast
and pristine still-available lands covered with transitional forests. Thus, in a process analogous to the one
occurring in present day Roraima and other parts of the Amazon, land became increasingly concentrated in
western Maranhão. Incoming farmers and ranchers claimed and fenced large tracts of land, in spite the fact
that thousands of peasants had been cultivating and living in those areas for decades. Consequently, numerous
peasants were turned into sharecroppers, being forced to devote ½ of whatever they produced to the new
landowners. Others engaged in the babaçu nut collection and processing; while, some ventured in the
garimpos11 of the nearby Pará State, or sought land in Roraima and other areas of the Amazon.
Despite the predominance of Roraima and Maranhão state in-migrants, the evolutionary path of frontier
settlements dramatically influences the way migration linkages are forged between settlements and various
source areas. As hypothesized, pioneer frontiers do draw migrants from a limited number of locations in the
Amazon region (98.9%), especially from within Roraima State. Nonetheless, data presented in Table 9
demonstrate that transitional frontiers also draw migrants primarily from the North (86.6) and Northeast
(12.2) regions, with the states of Roraima and Maranhão representing the chief source states. Conversely,
consolidated and urbanized frontiers attract migrants from a much broader range of origin areas, also
confirming the expectations of the proposed model. Notice how the relative participation of other Brazilian
regions is much larger among the more advanced stages of frontier, where better communications networks
and complex job markets prevail (Figure 2).
Another evidence of the massive intra-regional migration is the discrepancy between migrants’ region of birth
and last region of residence (Table 10). While a sizable portion of migrants were born in Northeast Brazil
(67.2); only 11.9% of surveyed settlers reached Central Roraima coming directly from their birthplaces.
Conversely, 24.7% of frontier colonists were born in the Amazon region; while, expressive 84.7% moved into
Roraima from within the Amazon. Embedded in these numbers is the regional repeat-migration process, in
which individuals born in Northeast Brazil, settle somewhere in the Amazon region, before arriving at
surveyed sites. This phenomenon, again, is emblematic of the great difficulties endured by negatively selected
settlers in establishing stable residences in the Amazon region.
9 Any change of residence with a length of stay of more than one month in any given destination is regarded as a migratory event, regardless of the distance
crossed.
10
Natives of Maranhão State.
11 11
Informal mining camps.
13
Table 9
Household heads’ last state of residence by stage of evolution *
State/Regions
Acre
Amapá
Amazonas
Pará
Rondônia
Roraima
Tocantins
North
Alagoas
Bahia
Ceara
Maranhão
Paraíba
Piauí
Pernambuco
R.G. Norte
Sergipe
Northeast
Distrito Federal
Goiás
Mato Grosso
Mato Grosso do Sul
Mid-West
Espirito Santo
Minas Gerais
Rio de Janeiro
São Paulo
Southeast
Paraná
Santa Catarina
R.G. Sul
South
Foreign
Total
Urbanized
1(1.1)
3(3.3)
11(12.2)
2(2.2)
45(50)
62(68.8)
3(3.3)
20(22.2)
1(1.1)
24(26.6)
1(1.1)
1(1.1)
1 (1.1)
1(1.1)
1 (1.1)
1 (1.1)
2 (2.2)
90(100)
Consolidated Transitional
2(2.2)
8(8.9)
3(3.3)
1(1.1)
66(73.3)
73(81.1)
1(1.1)
76(84.4)
78(86.6)
7(7.8)
11(12.2)
7(7.8)
11(12.2)
2 (2.2)
2 (2.2)
2(2.2)
2(2.2)
3(3.3)
1(1.1)
90(100)
90(100)
Pioneer
1(1.1)
11(12.2)
76(84.4)
1(1.1)
89(98.9)
1(1.1)
1(1.1)
90(100)
Total
1(0.3)
6(1.7)
33(9.2)
3(0.8)
260(72.2)
2(0.6)
305 (84.7)
3(0.8)
39(10.8)
1(0.3)
43(11.9)
2(0.6)
1(0.3)
3(0.8)
1(0.3)
1(0.3)
3(0.8)
1(0.3)
4(1.1)
4(1.1)
360(100)
* Values in parenthesis represent percentages
Table 10
Migrant household head source regions
Regions
North
Northeast
Mid-West
Southwest
South
Foreign countries
Total
Region of Birth
89 (24.7)
242(67.2)
5(1.4)
5(1.4)
14(3.9)
5(1.4)
360(100)
Region of last residence
305 (84.7)
43(11.9)
3(0.8)
1(0.3)
4(1.1)
4(1.1)
360 (100)
14
Figure 2
Settlers’ Last State of Residence
Consolidated
Urbanized
l
l
l
l
l
l
l l
l
l
l
l
l
l
l
l
l
l
l
Number of settlers
l
10
l
5
1
l
Transitional
l
l
l
Pioneer
l
l
l
l
l
l
15
The average number of moves performed by migrants before reaching present destinations is also indicative
of long migration histories. Overall, Central Roraima settlers have performed an average of 4 moves before
reaching surveyed areas; still, there is a statistically significant difference across the stages of frontier
evolution with respect to the average number of moves performed. Unlike anticipated, pioneer frontier
migrants are the most peripatetic individuals displaying an average of 4.68 moves (Table 11). Land
speculation strategies, coupled with the hardships endured by the rural poor do motivate a sizable numbers of
moves, partially explain this performance. Transitional frontier settlers come second with an average of 4.36
moves, followed by consolidated frontier individuals, with 3.56 moves; and urbanized frontier ones, with 3.40
moves. It is also important to notice how the average number of moves among urbanized frontiers’ settlers
present the highest degree of variation, which is produced by the variety of migrant groups arriving in such
areas. Among them the peripatetic displaced peasants, and the positively selected urban migrants, who may be
engaging in their first migratory experience (Diniz 1997).
Table 11
Number of settlers’ migratory events by stage of frontier evolution
Stage
Urbanized
Consolidated
Transitional
Pioneer
Total
Mean St. Dev.
3.40
2.73
3.56
2.24
4.36
2.59
4.68
2.69
4.00
2.62
F= 5.187 (0.002)
C.V.
0,80
0,62
0,59
0,57
0,65
N
90
90
90
90
90
Another way in which migration is influenced by the process of frontier evolution can be grasped by the
gradual unimportance of migration chains. Overall, 85.6% of pioneer frontier settlers had acquaintances in the
area before making the move (Table 12). The proportion of in-migrants with acquaintances, however, falls
significantly at the more evolved stages. Notice that migration chains influence 71.1% of transitional frontier
migrants; 48,9% of consolidated settlers, and 45,6% of urbanized frontier dwellers. These results confirm the
expectations embedded in the model of mobility in evolving agricultural frontier settlements, as better
endowed settlers, typical of the more evolved portions of the frontier, are less reliant upon friends and
relatives. In fact, these individuals can count on their higher educational attainment and financial resources to
access jobs and/or land markets.
Table 12
Settlers’ knowledge of friends and relatives
prior to move by stage of evolution
Acquaintances prior to move Urbanized Consolidated Transitional
With
41(45.6)
44(48.9)
64(71.1)
Without
49(54.4)
46(51.1)
26(28.9)
Total
90(100)
90(100)
90(100)
Chi2 = 41.511 (0.0000) Tau B = 0.093 Eta = 0.340
Pioneer
77(85.6)
13(14.4)
90(100)
Total
226(62.8)
134(37.2)
360(100)
When inquired about how settlers became familiar with present destinations, 91.1% pointed informal
communication channels, such as the word of friends and relatives passed on via letters, and phone calls, as
the primary source of information (Table 13). The proportion of migrants relying on informal communication
channels remains stable throughout the stages of frontier evolution. Nevertheless, the role of formal
communication channels is more pronounced among urbanized frontier settlements, where 18.8% of inmigrants heard about those locations via official advertisement put out by government agencies, television, or
companies operating in the area, as anticipated.
16
Table 13
Migration information system by stage of evolution
Information system
Informal
Formal
Total
Urbanized Consolidated Transitional Pioneer
74(82.2)
82(91.1)
86(95.6)
86(95.6)
16(17.8)
8(8.9)
4(4.4)
4(4.4)
90(100)
90(100)
90(100)
90(100)
Chi2 = 13.171 (0.004) Tau B = 0.057 Eta = 0.191
Total
328(91.1)
32(8.9)
360(100)
The changing geographies of frontier settlements also influence settlers’ primary reasons for moving into
Central Roraima. While the majority of settlers across pioneer frontiers (87.8%) pointed the desire of
acquiring land as the primary reason for migrating into the area; transitional frontiers in-migrants are divided.
Among these settlers, 57.8% were seeking land, and 41.1% employment (Table 14). At consolidated frontiers,
the proportion of migrants pursuing land diminishes a bit more (53.3%), while the extent of settlers in search
of jobs goes up to 43.3%. At urbanized frontiers one witness a more complex situation, in which individuals
can be grouped in three different categories: those pursuing land (27.8%), those seeking jobs (34.4%), and
those seeking the reunification with acquaintances, namely family members (37.8%).
Table 14
Household heads’ primary reason for in-migration by stage of evolution
Reason
Urbanized Consolidated Transitional Pioneer
Land
25(27.8)
48(53.3)
52(57.8)
79(87.8)
Job
31(34.4)
39(43.3)
37(41.1)
6(6.7)
Reunite acquaintances 34(37.8)
3(3.3)
1(1.1)
5(5.6)
Total
90(100)
90(100)
90(100)
90(100)
Chi2 = 121.20 (0.0000) Tau B = 0.178 Eta = 0.485
Total
204(56.7)
113(31.4)
43(11.9)
360(100)
Circulation12
Fieldwork results demonstrate that a large portion of settlers (78%) presently engage in some sort of regular
circulation (not shown). Among the major reasons for the intense intra-regional circulation is the performance
of odd jobs in order to boost household incomes or exchange services13 with fellow peasants during the more
labor intense phases of the agriculture production cycle. Nonetheless, the likelihood of performing odd jobs is
conditioned by the stage of evolution. Notice that over 81% of pioneer migrants regularly engage in odd jobs;
whereas, the proportion of transitional, consolidated and urbanized frontier migrants engaged in parallel jobs
is smaller (Table 15). This takes place because a larger portion of settlers of urbanized and consolidated
frontiers hold formal jobs, being committed to fix working schedules and regular pay. Peasants and other
autonomously employed individuals, prevalent across pioneer and transitional frontiers, and also present at
expressive numbers at urbanized settlements, enjoy greater time flexibility and erratic income, being therefore
freer to pursue odd jobs and needier of extra cash.
It was observed in the field that the months prior to the rainy season (December, January and February) are
marked by intense job-seeking circulation movements. This is a period of high labor demand in the frontier,
given the preparations for new agriculture fields. The process of clearing new tracts of forest and bringing
them into production is arduous, long, and divided in a series of phases. During the hiatus of the forest
clearing process, migrants take advantage of the high seasonal labor demand, to employ themselves
temporarily at nearby areas; while, some venture as far as Boa Vista14 and other urban places15. Circulation is
12 Circulation is regarded as any short-term move that does not involve declared permanent change of residence. This includes daily commuting, work trips, leisure and social trips, and seasonal, circulation among urban centers, farms, ranches
and garimpos.
13 Job exchange groups are formed based on friendship and family ties, being facilitated by the arrival of different waves of related migrants in certain colonization areas. In the process, a given peasant (innovator) arrives at pioneering stages,
managing to claim adjacent lots of land for him/herself and for his/her friends and family members left behind. Once land has been secured, the second waive of peasants (followers) soon arrive at the frontier, occupying the reserved neighboring
plots. Mass in-migration is another strategy commonly adopted by settlers arriving at less evolved areas. Accordingly, a number of related families move concurrently into a given frontier area, claiming alongside plots, forming aid-groups. These
informal associations call for communal type work, in which efforts are concentrated at different lots in an alternate fashion. During the preparation of land for agriculture, for instance, a group of peasants, rotate the lots of all group members,
promoting specific tasks.
14 Roraima State Capital City.
17
also expressive during the rainy season (March to August), although not to the same degree as between
December and February. After planting is over, peasants are found in a very needy state, as most of the
previous harvest has been consumed and the scant financial resources were invested in seeds and tools for the
new crops. This needy period is matched by low labor demands in the lots, as crops have different maturing
times and harvesting is erratic passing over constant care. This state of affairs also stimulates odd-job-seeking
circulation.
Table 15
Odd jobs undertaken by settlers by stage of evolution
Urbanized Consolidated Transitional Pioneer
Perform odd jobs
37(41.1)
37(41.1)
44(48.9)
73(81.1)
Do not perform odd jobs 53(58.9)
53(58.9)
46(51.1)
17(18.9)
Total
90(100)
90(100)
90(100)
90(100)
Chi2 = 39.380 (0.000) Tau B = .084 Eta = 0.331
Total
191(53.1)
169(46.9)
360(100)
Corroborating the hypotheses embedded in the proposed model, survey results demonstrate that the place
where odd jobs are performed is influenced by the process of frontier evolution. Pioneer and transitional
settlers tend to travel less to perform such jobs. Notice that 83.6% of pioneer and 70,5% of transitional settlers
develop these activities within the confines of the agricultural project they live (Table 16). It was observed,
however, that the majority of odd jobs taking place in these areas is related to job-exchange informal groups.
On the other hand, a large proportion of individuals performing odd jobs at the more evolved portions of the
frontier travel further, as they count on reliable roads and regular transportation lines to move around. It is
also important to point out that the more evolved portions of the frontier are concentration points for displaced
peasants and recently arrived migrants, who after establishing residences locally, perform all sorts of informal
jobs at farms, ranches, and government projects, such as road construction (Becker, 1990).
Table 16
Place where odd-jobs are undertaken by stage of evolution
Local
Non-local
Total
Urbanized Consolidated Transitional Pioneer
14(37.8)
17(45.9)
31(70.5)
61(83.6)
23(62.2)
20(54.1)
13(29.5)
12(16.4)
37(100)
37(100)
44(100)
73(100)
Chi2 = 29.276 (0.000) Tau B = .120 Eta = 0.392
Total
123(64.4)
68(35.6)
191(100)
Furthermore, circulation in Central Roraima relies on well-established logistic support, as over 50% of
settlers, across all stages of evolution hold two or more active urban and/or rural properties within region (not
shown). Nonetheless, the periodicity in which circulation between residences takes place is also conditioned
by the evolutionary process, as anticipated by the proposed model. Table 17 demonstrates that settlers located
at the more evolved settlements tend to circulate between residences more often than those dwelling at less
evolved locales. Notice how only 15.2% of pioneer migrants circulate on a weekly-basis, contrasted with a
much larger proportion of individuals engaged in such movements at the more evolved frontier settings.
Conversely, 50% of pioneer frontier settlers circulate on a 15 to 30 day basis; whereas, 34.8% of them
circulate in intervals that exceed 30 days. Infra-structural levels present at the various stages of evolution
explain the regularity of inter-residence circulation. Places that are better endowed with access roads and
regular transportation lines, typical of more evolved frontiers, ease the seasonal movements and minimize the
temporal hiatus between them. On the other hand, low infrastructure levels present at the less evolved
frontiers, make migrants’ seasonal movements more difficult and less frequent.
Seasonal inter-residential movements typical of urbanized frontiers, are those performed by the women and
children of divided households. During the harvest of any given important crop such as rice, manioc, or beans,
folks living at urbanized places move temporarily to the rural lots in order to help their family members.
15 Those seeking seasonal jobs in larger urban places usually employ themselves in construction or informal commerce activities.
18
Harvesting, thus reunite the divided households at the rural areas, becoming a task undertaken by the entire
family. In fact, absence rates at local schools are so high that educational establishments have special
intermissions during harvest time, allowing children of schooling age to aid their peasant parents. Once the
task is completed, however, urbanized folks resume their urban lives, leaving their rural family members
behind. Also typical is the inter-residential circulation performed by peasants and small-scale producers, who
after adopting an urban residence still maintain strong linkages with the rural world, where subsistence is
earned. Originating in urbanized frontier settlements are the weekly leisurely-motivated sojourns undertaken
by the better-off urban dwellers at country places at rural areas during weekends and holidays. Thus, it
becomes clear that circulation pattern originating in urbanized areas is markedly different, depending on the
socioeconomic profile of settlers.
On the other hand, the circulatory moves originating in less evolved stages can be broadly classified
according to their destinations. Rural-to-urban circulation is usually performed by men and children of
divided households, who periodically leave agricultural plots in order to visit folks at urban and urbanized
frontier areas. Also motivating the rural-to-urban moves are the facilities located at the urban and urbanized
frontiers, which are regularly sought by peasants. In the urban realm they sell part of their production,
purchase goods, consult physicians and dentists, go to church, and so on. Conversely, rural-to-rural circulation
is associated with job exchanges within agricultural projects, and periodic visits made by peasants in order to
check on or claim plots of land.
Table 17
Periodicity of the inter-residential circulation by stage of frontier evolution
Periodicity
Weekly
15 to 30 Days
Over 30 days
Total
Urbanized Consolidated Transitional Pioneer
28(54.9)
50(54.1)
25(55.6)
7(15.2)
13(25.5)
13(35.1)
14(31.1)
23(50)
10(19.6)
4(10.8)
6(13.3)
16(34.8)
51(100)
37(100)
45(100)
46(100)
Chi2 = 24.037 (0.001) Tau B = 0.070 Eta = 0.341
Total
80(44.7)
63(35.2)
36(20.1)
179(100)
Mobility and evolving agriculture frontier settlements framework
This study explored the interplay between the evolution of agriculture frontier settlements and human
mobility. Rooted on the literature on frontier evolution stage models, and Amazonian human mobility, this
work proposed and tested various hypotheses related to mobility across four different types of settlement:
pioneer, transitional, consolidated and urbanized. The reality of Central Roraima corroborated most
expectations of the proposed model, confirming that human mobility is indeed influenced by the process of
frontier evolution. The overall relationship between human mobility and frontier evolution is portrayed by
Figure 3.
Evidence showed that at the least evolved portions of the frontier, settlers tend to be predominantly
uneducated, married, and male. Also intriguing is the fact that households are also much smaller at these
pioneer areas. Conversely, at more evolved settings there is an increasing presence of females, single and
educated individuals. Another curious finding is the fact that frontier evolution is not selective of settlers’ age,
as the majority of surveyed individuals arrived at present locales of residence with similar age. Therefore,
throughout the evolutionary path, the selectivity process remains relatively stable, as negatively selected
individuals dominate the system. At urbanized frontiers, however, one witnesses a more erratic trend, as
migrant socioeconomic profiles are more diversified.
Migratory patterns, however, are highly influenced by the evolutionary process. Recall that the more evolved
settings draw migrants from a broad range of places; whereas, the least evolved ones display strong links with
specific places. Thus, an incremental trend exists between the variety of in-migrants’ source areas and the
evolutionary process.
19
Figure 3
Mobility and evolving agriculture frontier settlements
Conversely, the role played by migration chains diminishes with increasing levels of evolution; while, the
importance of informal communication channels also dwindles. Another way in which migration has been
impacted by the changing geographies of agriculture frontier settlements has to do with the primary reason
disclosed by settlers for moving into various frontier types. At the least evolved settings land-seeking moves
dominate; while, family reunification and jobs are the primary motive behind the moves into urbanized areas.
Results demonstrate that circulation patterns are also sensitive to the frontier evolutionary process. First, odd
job seeking circulation is increasingly less prevalent at more evolved settings, and those who engage in such
practices at least evolved settings tend to travel less than those individuals dwelling at more evolved areas.
Fieldwork also unveiled that circulation is based on a well-orchestrated logistic support, as a sizable number
of settlers, regardless of stage of evolution, rely on dual or multi-residences. Nonetheless, the periodicity in
20
which settlers circulate between properties is influenced by the frontier evolutionary process, as the settlers at
more evolved areas tend to circulate more frequently than those living at less evolved areas.
The overall findings can also be organized in terms of a generalized model of mobility and evolving
agriculture frontier settlements at four different stages. The following paragraphs describe the predominant
mobility patterns and behaviors at each stage of evolution.
Pioneer frontiers
Given the incipient nature of pioneer settlements, the lack of market for land and labor, and the physical
stamina involved in settling and surviving in the harsh equatorial forest, households tend to be small, headed
by males, marked by low levels of education and long migration histories. Due to the remoteness and the
uncertainties associated with these areas, migration chains dominate the system, constituting the primary
migration type, linking pioneer frontiers with very specific rural origins. In the process, informal
communication channels strength the links between pioneer frontiers and very specific source areas. At this
stage, land is the primary migration pull factor. As transportation networks are poor at this stage, circulation is
spatially and temporally restrained, as individuals travel within short distances to perform odd-jobs and
exchange agricultural services with fellow peasants.
Transitional frontiers
Household-heads at transitional frontiers are predominantly male, uneducated, and young, with long
migration histories. Household size is not very different from pioneer areas, but here one begins to notice the
increasing presence of farm and ranch workers to the detriment of landowners. Settlers are still drawn from a
limited number of places, and informal information systems prevail. Individuals moving into such places also
rely on migration chains, but not to the same degree as pioneer frontiers, given the improvements in
transportation and communication networks. Land remains the primary stimulus for in-migration, but farm
and ranch-jobs have also become important pull factors. Owing to improved transportation linkages, and
increasing need for cash fomented by the ever present capitalist mode of production, settlers begin circulating
among a vast array of places, within shorter intervals, engaging in temporary employment in farms, ranches,
garimpos and even urban areas.
Consolidated frontiers
Unlike other frontier areas, most consolidated settlers do not own the land they presently occupy. These
individuals tend to be late comers and displaced peasants, who employed themselves at the encroaching farms
and ranches, while looking for land at adjacent areas. These settlers tend to in-migrate individually and form
households with fellow workers. Therefore, household units are significantly smaller from the other stages of
evolution. Settlers at this stage are still mostly, male, young, and uneducated, with long migration histories.
Owing to their better infrastructure and more reliable transport lines, consolidated frontiers draw migrants
from a broad set of places. In-migrants still rely on unofficial information systems, but the importance of
migration chains diminishes substantially. The perspective of acquiring land remains important, but now jobseeking moves are also conspicuous. Job-seeking circulation occurs at a much lesser rate, due to the stable
employment and fixed schedules at farms and ranches. Nonetheless, circulation, in general, tends to occur in a
more frequent fashion (weekly), being spatially broader given the good road infrastructure and transport lines.
Urbanized frontier
Households at this stage are still primarily headed by males, although female-headed households are also
significant. Moreover, due to the presence of higher infrastructure levels, households tend to be large,
21
generally with numerous children. Given the prominence of urbanized frontiers, they do attract two different
major types of settlers: displaced peasants and better-off urban migrants. Thus, there is a bifurcation among
settles in terms of educational levels, age, occupation16, and migration histories. Urbanized frontiers have
good communication links with their hinterlands and with the rest of the country, drawing migrants from a
much broader pool of rural and urban origins. The role of migration chains is even smaller at this stage; while,
formal communication channels influence an increasing number of moves. Here a broader array of motives
inspires in-migration, among which land, job, and family reunification. Circulation is intense and spatially
broad being facilitated by well-developed transportation networks. While displaced peasants seek jobs at
surrounding areas with certain regularity; while, better-off in-migrants circulate on a regular basis especially
for leisure activities.
Discussion
Evidence for the proposed model was acquired in Central Roraima State, Brazil. Nonetheless, the model
should be applicable to various frontier settings given the scope of the claimed mobility and the premises in
which it was built upon. In fact, future studies should explore the validity of these claims in other areas of the
Amazon region in order to expand and calibrate the model presented here. A word of caution, however, has to
be voiced. The model of mobility and evolving agricultural frontier settlements targets the specifics of frontier
settlements, which have evolved from spontaneous or official agricultural colonies, or settlements, where
small-scale producers predominated at pioneering stages. Frontier areas where the demographic and economic
occupations have been promoted by other types of activities, such large scale resource extraction, namely
garimpos and logging, mega-construction projects, and large scale cattle ranching activities, follow a more
erratic and inconsistent evolutionary path, extrapolating the basic principles of the proposed model.
Moreover, the set of mobility claims advanced here is pertinent to the settlement level only. The complexities
embedded in the proposed model and the micro-scale dynamic entailed by the model prohibits its utilization
in meso or regional scales.
Despite being discredited by many scholars, I believe stage-thinking models remain important analytical
tools. They organize the temporal evolution of very complex phenomena in different segments, each being
composed of strikingly different characteristics, offering the means to center attention in the most prominent
aspects of reality. However, the proposed framework is not intended as a deterministic stage-model à la
Zelinsky’s (1971) “hypothesis of mobility transition”. Instead, it should be emphasized that settlements do not
necessarily follow a linear evolutionary path, going through all the previously identified stages in a
progressive fashion. Reality challenges all mechanistic views, as the pace and direction of settlements’
evolution are highly idiosyncratic. Therefore, the model should be utilized as a first step in the process of
understanding the chaotic Amazonian mobility, being imperative a foot in the mud approach in order to
bridge the theoretical construe with settlements’ reality.
Lastly, it is important to recall that the proposed model could be applied to both, cross-section and
longitudinal perspectives. The model was applied in this study in a cross-section fashion due to economic and
time restraints, as well as lack of historical secondary data. Still, it is important to alert those willing to apply
it on a longitudinal basis, that the timing and the direction in which settlements will pass through the
identified phases are erratic. Therefore, discretion and sensibility should be exercised when applying it in a
historical perspective.
16
Not shown. See forthcoming doctoral dissertation – Department of Geography – Arizona State University.
22
Bibliography
Almeida, O. 1992. The Colonization of the Amazon. Austin, University of Texas Press.
Almeida, A. and David, M. 1981. Tipos de Frontiera e Modelos de Colonização na Amazônia: Revisão de
Literatura e Especificação de Uma Pesquisa de Campo. IPEA, Instituto de Planejamento Econômico Social.
Texto para Discussão #38. Brasília, DF.
Aragón, L. 1980. Mobilidade dos Migrantes no Norte de Goiás. Seminário promovido pelo Setor de Pesquisa
(SEPEQ) do NAEA (Núcleo de Altos Estudos Amazônicos) em 31 de agosto de 1978. Série Seminários e
Debates, # 4. Belém, NAEA
1983. Mobilidade Geográfica e Ocupacional no Norte de Goiás: Um Exemplo de Migração por
Sobrevivência. In O Despovoamento do Território Amazônico: Contribuições Para Sua Interpretação.
Mougeot, L. and Aragón, L. (Editors.) p. 91-122 . Belém, UFPA/NAEA
Aragón, L. and L. Mougeot. 1986. Introdução. In Migrações Internas na Amazônia: Contribuições Teóricas e
Metodológicas, L. Aragón and L. Mougeot (eds.). pp. 17-53. Núcleo de Altos Estudos Amazônicos. Belém,
Falangola.
Becker, B. 1985. Fronteira e Urbanização Repensadas. Revista Brasileira de Geografia. Rio de Janeiro 47
(3/4): 357-371.
1990. Amazônia. São Paulo, Ática.
Bentes, R. 1986. Zona Franca, Desenvolvimento Regional e o Processo Migratório para Manaus. In
Migrações Internas na Amazônia: Contribuiçõs Teóricas e Metodológicas. By L. Aragón and L. Mougeot
(eds.). p. 220-254. Núcleo de Altos Estudos Amazônicos. Belém: Falangola.
Browder, J. and Godfrey, B. 1990. Frontier Urbanization in the Brazilian Amazon: A theoretical framework
for urban transition. Conference of Latin American Geographers, 16: 56-66.
Bylund, E. 1960. Theoretical Considerations Regarding the Distribution of Settlement in Inner North Sweden.
Geografiska Annaler B, 42, pp. 225-231.
Crócia, Nilson Barros. 1994. The Frontier Cycle: A Study of the Agricultural Frontier Settlement in the
Southeast of Roraima, Brazil. N.J. Cox, Series. Department of Geography, University of Durham UK.
1995. Roraima, Paisagens e Tempo na Amazônia Setentrional. Recife, Universidade Federal de
Pernambuco, Editora Universitária.
Da Silva, 1986. Retenção e Seleção migratórias em Rio Branco, Acre. In Migrações Internas na Amazônia:
Contribuições Teóricas e Metodológicas, L. Aragón and L. Mougeot (eds.). pp. 182-216. Núcleo de Altos
Estudos Amazônicos. Belém: Falangola.
Da Vance, J. 1970. The Merchant’s World : The Geography of Wholesaling. Englewood Cliffs: Prentice Hall.
Dawsey, C. 1992. Migration in Brazil: Research During the 1980s. Conference of Latin Americanist
Geographers, Vol. 17/18, p. 109-116.
Diniz, A. 1997. Growth and Urbanization in Roraima State, Brazil. Conference of Latin Americanist
Geographers. Forthcoming.
Ferreira, I. 1987. Expansão da Fronteira Agrícola e Urbanização. In A Urbanização da Fronteira. Lavinas,
Lena (Editor). Séries Monográficas #5, Volume 2. p. 1-26. Rio de Janeiro: IPPUR/UFRJ
23
Findley, S. 1988. Colonist Constraints, Strategies, and Mobility: Recent Trends in Latin American Frontier
Zones. In Oberai, A. (ed). Land Settlement Policies and Population Redistribution in Developing
Countries:271-316. New York, Praeger
Foresta, R. 1992. Amazônia and the Politics of Geopolitics. The Geographical Review .Vol 82 (2), p. 128142.
Foweraker, J.1981. The Struggle for Land. Cambridge: Cambridge University Press.
Furley, P. and Mougeot, L. 1994. Perspectives in the Forest Frontier, Settlement and Change in Brazilian
Roraima. Ed. by Peter Furley. New York, Routledge.
Godfrey, B. 1990. Boom Towns of the Amazon. The Geographical Review. Vol. 80 (2) p. 103-117.
1992. Migration to the Gold-Mining Frontier in Brazilian Amazonia. The Geographical Review
.Vol. 82 (4) p. 458-469.
Henkel, R. 1982. The Move to the Oriente: Colonization and Environmental Impact. In Modern Day Bolivia:
Legacy of the Revolution and Prospects for the Future. Edited by Jerry R. Ladman. Tempe: Center for Latin
American Studies, Arizona State University. pp.277-300.
Hennessy, Alistair. 1978. The frontier in Latin American history. London, Edward Arnold.
Henshall, J. and Momsen R. 1976. A Geography of Brazilian Development. London: G. Bell & Sons, LTD.
Henriques, M. 1984. A Politica de Colonização Dirigida no Brasil: um estudo de caso, Rondônia. Revista
Brasileira de Geografia 46(3-4) 393-414.
1985. A Dinâmica Demográfica de Uma Área de Fronteira: Rondônia. Revista Brasileira de
Geografia, Rio de Janeiro, 47 (3/4): 317-356.
1986. Os Colonos de Rondônia: Conquistas e Frustrações. Revista Brasileira de Geografia,
Rio de Janeiro, 48 (1): 3-42.
____ _______ 1988. The Colonization Experience in Brazil. In Land Settlement Policies and Population
Redistribution in Developing Countries: 317-354 By Oberai (ed.). New York, Praeger.
Hudson, J. 1969, A Location Theory for Rural Settlement. Annals of the Association of American
Geographers 59, pp. 365-381.
IBGE 1992. Censo Demográfico de 1991. Rio de Janeiro: Fundação Institudo Brasileiro de Geografia e
Estatística.
Jardim, A. 1987. Aspéctos do Processo de Urbanização Recente na Região Centro-Oeste. In A Urbanização
da Fronteira. Lavinas, Lena (Editor). Séries Monográficas #5, Volume 2. p. 103-128 Rio de Janeiro:
IPPUR/UFRJ.
Lavinas, L. 1987. A Agro-Urbanização da Fronteira. In A Urbanização da Fronteira. By Lavinas, Lena (ed).
Series Monográficas #5, Volume 1. p. 91-108. Rio de Janeiro: IPPUR/UFRJ
Lewis, W. 1954. Economic Development with Unlimited Supplies of Labor. Journal of the Manchester
School of Economics and Social Studies. 20, pp. 139-92.
Lisansky, J. 1990. Migrants to Amazonia. Spontaneous colonization in the Brazilian frontier. Boulder, San
Francisco and London, Westview Press .
24
MacMillan, G. 1995. At the End of the Rainbow? Gold, Land and People in the Brazilian Amazon. London,
Earthscan Publications Ltd.
Martine, G. 1981. A Migração Repetida e a Busca de Sobrevivência: Alguns Padrões Brasileiros. Projeto de
Planejamento de Recursos Humanos. Relatório Técnico #52. Brasília.
1984. Frontier Expansion, Agricultural Modernization and Population Trends in Brazil.
Brasilia: IPEA/IPLAN/ CNRH.
____________ 1989. Internal Migration in Brazil. IPEA, Instituto de Planejamento Econômico Social. Texto
pare Discussão #13. Brasília, DF.
Martins, José de Souza. 1975. Frente Pioneira: Contribuição para uma caracterizacao sociológica, in
Capitalismo e Tradicionalismo pp. 43-5. By Velho, Otávio (ed.). São Paulo, Pioneira.
___________1984. The State and the Agrarian Question in Brazil. In Frontier Expansion in Amazonia, By
Marianne Schmink and Charles Wood (eds.). Gainesville, University of Florida Press, 463 - 490.
______________1996. O Tempo da Fronteira, retorno à controvérsia sobre o tempo histórico da frente de
expansão e da frente pioneira. Tempo social; Revista de Sociologia da USP. São Paulo, 8(1): 25-70, maio de
1996.
Meyer, D. 1980. A Dynamic Model of the Integration of Frontier Urban Places into the United States System
of Cities. Economic Geography 56, pp. 120-140.
Morrill, R. 1963. The Development of Spatial Distributions of Towns in Sweden: An Historical-Predictive
Approach. Annals of the Association of American Geographers 53, pp. 1-14.
Mougeot, L. 1982a. Ascensão Sócio-Econômica e Retenção Migratória na Fronteira. Série Seminários e
Debates (8). Belém, Universidade Federal do Pará - Núcleo de Altos Estudos Amazônicos.
1982b. Alternate Migration Targets and Brazilian Amazonia’s Closing Frontier. Paper
prepared for the Commission on Regional Systems and Policies. Latin American Regional Conference,
International Geographical Union. Belo Horizonte - Rio de Janeiro, Brazil. August, 9th -21st, 1982.
1983. Retenção Migratória das Cidades Pequenas, nas Frentes Amazônicas de Expansão:
Um Modelo Interpretativo. In O Despovoamento do Território Amazônico: Contribuições Para Sua
Interpretação. Mougeot, L. and Aragón, L. (Editors.) p. 123-146 . Belém: UFPA/NAEA
1986. A Recente Ocupação Humana da Região Amazônica: Causas, Durabilidade e
Utilidade Social. In Migrações Internas na Amazônia: Contribuições Teóricas e Metodológicas. Aragón, L
and Mougeot, L. (Editors) p. 17-53. Belém:UFPA, NAEA, CNPq.
Mougeot, L. and Aragón, L. (Editors.) 1983. Introdução. In O Despovoamento do Território Amazônico:
Contribuições Para Sua Interpretação. Introdução: 9-26. Belém: UFPA/NAEA
Muller, C. 1980. Frontier-based Agricultural Expansion: The Case of Rondonia. In Land, People and Planning
in Contemporary Amazonia. F. Barbira-Scazzocchio, ed. P. 141-53. Cambridge: University of Cambridge,
Center of Latin American Studies.
Muller, E. 1977. Regional Urbanization and the Selective Growth of Towns in North American Regions.
Journal of Historical Geography 3, pp. 21-39.
25
Neiva, Arthur Hehl. 1949. A imigração na política brasileira de povoamento. Revista brasileira de municípios,
ano II nº6, abril-junho, p.266.
Oliveira, C. 1986. O Campo Migratório de Soure, Pará. In Migrações Internas na Amazônia: Contribuições
Teóricas e Metodológicas, L. Aragón and L. Mougeot (eds.). pp. 148-181. Núcleo de Altos Estudos
Amazônicos. Belém: Falangola.
Santos, R. The Problem of Land in the Brazilian Amazon. In Frontier Expansion in Amazonia. Ed. Marianne
Schmink and Charles Wood. pp.439-462. Gainesville, University Press of Florida, University of Florida.
Sawyer, D. 1981. Ocupação e Desocupacao da Fronteira Agrícola no Brasil: Ensáio de Interpretação
Estrutural e Espacial. Trabalho apresentado no seminário sobre expansão da fronteira agropecuária e meioambiente na America Latina. Brasília, 10-13 de Novembro de 1981.
___________1982. Industrialization of Brazilian Agriculture and Debilitation of the Amazon Frontier. Paper
presented at the Seminar on Land Development in the Tropics. August, 8th-13th 1982. Belo Horizonte, Brazil.
______
_ 1984. Introduction. Frontier Expansion and Retraction in Brazil. In Frontier Expansion in
Amazonia, 180 - 203.By Marianne Schmink and Charles Wood (eds.). Gainesville, University of Florida
Press.
1986. A Fronteira Inacabada: Industrialização da Agricultura Brasileira e Debilitacão da
Fronteira Amazônica. In Migracões Internas na Amazônia: Contribuições Teóricas e Metodológicas, L.
Aragón and L. Mougeot (eds.). Núcleo de Altos Estudos Amazônicos pp. 54-91. Belém: Falangola.
___________1987. Urbanização da Fronteira Agrícola no Brasil. In A Urbanização da Fronteira. Lavinas,
Lena (Editor). Séries Monográficas #5, Volume 1. p. 43-60 Rio de Janeiro: IPPUR/UFRJ.
1989. Urbanization of the Brazilian Frontier. Presented in a Seminar on Urbanization in Large
Developing Countries, IUSSP and Gujarat Institute of Area Planning, Ahmedabad, India, 28 th of September
to October 1st.
Sawyer, D. and Carvalho, J. 1986. Os Migrantes em Rio Branco, Acre: Uma Análise a Partir de Dados
Primários. In Migrações Internas na Amazônia: Contribuições Teóricas e Metodológicas. Aragón, L and
Mougeot, L. (Editors) p. 112-147. Belém: UFPA, NAEA, CNPq.
Sawyer, D.; Torres, H.; Pereira, A. and R. Abers 1990. Fronteiras na Amazônia: Significado e Perspectivas.
Centro de Desenvolvimento e Planejamento Regional - CEDEPLAR - Universidade Federal de Minas Gerais .
Relatório final da primeira fase da pesquisa “A Dinâmica Demográfica da Região Amazônica Numa
Perspectiva Nacional” for ELETRONORTE
Sewastynowicz, 1986. Two-step Migration and Upward Mobility on the Frontier?: The Safety Valve Effect
in Pejibaye, Costa Rica. Economic Development and Cultural Change 34 (4) 731-753.
Shoemaker, R. 1981. The Peasants of El Dorado: Conflicts and contradiction in a Peruvian Frontier
Settlement Ithaca. NY: Cornell University Press.
Shrestha, N., Velu, R. and D. Conway. 1993. Frontier Migration and Upward Mobility: The Case of Nepal.
Economic Development and Cultural Change. Vol. 41 (4) pp. 787-816.
Sternberg, H. 1987. “Manifest Destiny” and the Brazilian Amazon: A Backdrop to Contemporary Security
and Development Issues. Conference of Latin Americanist Geographers. Vol. 13. p. 25-35.
Taaffe, E., R. Morrill, and P. Gould. 1963. Transport Expansion in Underdeveloped Countries: A
Comparative Analysis. Geographical Review 53: 503-529.
26
Volbeda, Sjoukje. 1996. Pioneer towns in the jungle, urbanization at an agricultural colonization frontier in
the Brazilian Amazon. Revista Geográfica # 104 - Jul/Dec. Pp.115-140.
Wood, C. 1982. Equilibrium and Historical-Structure Perspectives on Migration. International Migration
Review 16, pp. 298-319.
Woods, C. and Wilson, J. 1984. The Magnitude of Migration to the Brazilian Frontier. In Frontier Expansion
in Amazonia, 142-152. By Marianne Schmink and Charles Wood (eds.). Gainesville, University of Florida
Press.
Wood, C. and Schmink, M. 1983. Culpando a Vítima: Pequena Produção Agrícola em um Projeto de
Colonização na Amazônia. In O Despovoamento do Território Amazônico: Contribuições Para Sua
Interpretação. Mougeot, L. and Aragón, L. (Editors.) p. 70-90. Belém: UFPA/NAEA
Wood, C. and Wilson, J. 1984. The Magnitude of Migration to the Brazilian Frontier. In Frontier Expansion
in Amazonia. Ed. Marianne Schmink and Charles Wood. P. 143-152. Gainesville, University Press of Florida,
University of Florida.
Zeballos-Hurtado, H. 1975. From the Uplands to the Lowlands: An Economic Analysis of Bolivian RuralRural Migration. Ph.D. Diss, University of Wisconsin.
Zelinsky, W. 1971. The Hypothesis of the Mobility Transition. The Geographical Review Vol. 61 (2) pp. 131.
27