The Boof Cattle Industry~ in th~e Eoi.aima Savan;unas:
A Potential Supplyr for Braril's fNoi-rt
THOMAS PISK; KELSEY
A DISSERTATION PRESENTED TO THE GRAD.lU;.TE COUNiM2L
OF THE UNIVTCERSITY`L OF FLO)RID1 ;iN PART~IALI
FULFILLMENTPI OF THE REQUIR~ENEUTSI' ~FO
THE DEGREE OF DOCTOR OF PHILOSOPHY
UNLIVERSITY OF FLORZID
UNIVERSITy OF FLORIOA
3 1262 08666 410 8
The experience of conducting field research and wri~jting~
a doctoral dissertation has no equal. The opportunity and
privilege of living on the frontier of northern Brazil has
been one that I shall never forget. To the warm, friendly
people of Brazil I express my heartfelt thanks.
I owe a tremendous debt of gratitude to many individuals
in many places. The leadership, guidance and encouragement
offer-ed by Professor Ralymond E. Crist, Chairman of m" Phi.D.
Committee, is impossible to record with proper thanks. Wlith-
out his efforts and enthusiasm this study. would not hav.e seen
the light of day. Profound appreciation is therefore grate-
To the members of my Committee, Dr. David L. Niddrie,
Dr. W. W. MczPherson, and Dr. Hugh L. Popenoe, I express my
appreciation for their help and encouragement, each in his
own constructive way. I thank Dr. E. E. Hegen who originally
suggested Roraima as a possible study area. The teaching of
the Portuguese language and the genuine concern shown by Pro-
fessor Alfred Hower and Professor Richard Preto-Rodas are
From my experience in Brazil, there is an endless list
of people to whom I would like to express my gratitude for
their friendship, help and interest in the study. I thank
His Excellency, Colonel Helio Campos, Governor of the Federal
Territory of Roraima, for permission to work in the territory,
near to the sensitive boundaries with Guyana and Venezuela.
Ilmo- Sr. Sebastiao da Silva, Mlinister of Agriculture, and
Captain Carlos Augusto de Goes e Silva, Director, Department
of Production, Lands-, and Colonization, were most helpful in
providing me with information from their files. Sr. Julio
Augusto M. Martins all but adopted my family and me. His aid
and personal friendship are of an unmeasurable magnitude. My
residence in Boa Vista was made most pleasant by its citizens,
especially Amazonas Brasil, Luis Av/elino, and Esdras Av/elino
Leit o. Of course, the fazendeiros of Roraima made it all
A special note of thanks goes to Dr. Saul Benchimol, in
Mlanaus, for his invaluable efforts to initiate my work in
My family and I express our appreciation to Dr. and Mrs.
Jerry Williams who sheltered and guided us through what would
have been a very trying experience of adjustment to life in
The study w~as financed by the Center of Tropical
Agriculture, Institute of Food and Agricultural Sciences,
Uni'.ersity of Florida, but any errors or misinterpretations
in thle dissertation are the responsibility of the author.
Finally, t~o miy w~ife, Dorothea, and my~ children, Niandy,
David, and ialrci, I express my~ sincere appreciation and
gratitue.1e for their perseverance and understanding during
many long day~s of my' absence from home and family.
This study was undertaken primarily to determine the
present and future possibilities of improving the quality and
increasing the quantity of beef cattle in Roraima for the
supply of urban consumers in Mlanaus. Roraima is truly one
of the "last frontiers" and precious little information has
been researched and written about this remote part of the
world. Therefore, a secondary purpose of the study was to
make basic information more available to those whose in-
terests lie ini the study of frontier areas.
Those who reside on the sav~annas of the Alto Rio Branco
and in the cities of Boa Vista and Mlanaus certainly are more
aware of their economic situations than this study ma imply.
As a foreigner, and especially as a North American, the
author found It extremely difficult to instill confidence in
the people, to convince them that he had no personal ambition
to purchase land and exploit the existing cattle interests
in Roraima. Just prior to the author's field investigation,
two North Americans had been ordered to leave the Territory
following a series of charges pertaining to business abuses.
As a result, people were very hesitant to volunteer
information on almost a7ny topic. At the other extreme, there
w~ere those individuals whi~o wanited to jnimpress upon their in-
quisitive visitor thiat their ranches were among the most
modern and efficient ji n Dazjl. Information and data given
by~ these people bore abisolut~elyr no resemblance to that which
the author observed.
All translations from Portuguese to English have been
made by the author and he maintains sole responsibility for
them. Since the terms whic-ih appear in the Glossar will be
used so often in this dissertation, they will not be
italicized after their initial use.
M~ost of mankind now? lives in a single w4orld system, and
because most of thle world's people, including Brazil's, have
found the Mletric System more practicable than alternative
measurement systems, the author has purposely been inconsis-
tent in stating units of measurement. As a gesture of respect
for the Brazilians, most of the units of measurement are
stated in the Metric System w~ith the United States Syrstem
stated in parentheses where immediate comparison is deemed
May this effort stimulate others to devote th-eir skills,
time, and effort to the problems and possibilities of pro-
ducing beef in a savanna environment and to this little known
part of the world.
I. INJTRODUCTI ON. ........ ...
II. METH-ODOLOGY .......
:II.. THE PHYSICAL SETTING. .........
IV. ACCESSIBILITY .............
River Transportation ........
Roads. . . . . . . . .
Air Transportation .........
V. THE PRESENCE OF BEEF CATTLE IN RORAlIMA,
Historical Background. .......
Cattle Population and Distribution
of Ranches . . . . . .
Cattle Breeds. ...........
Ranching Efficiency and Productivity
VI. FACTORS THAT LIM1IT PRODUCTION .. .
Inadequate Feeding .........
TABLE OF CON~TEN~TS
A C KNOW]L EDGMPENTTS. .
LIST OF TABLES ...
LIST OF FIGUR~ES. ..
. . .
. . .
. . .
. . .
. . .
TABLE OF CONTENTS (Continued)
Seasonal forage shortages. ....
Range and pasture problems ....
Diseases. . . . . . . .
Infectious and contagious disease.
Parasitic infestation. ...
Mineral deficiency problems. ...
Breed Improvement ..........
Other Limiting Factors. .......
Mla na gemen t . . . . . .
Land tenure. ... .. .. .
VI I. FROM ORANGE TO MAFRI;ET ..........
Slaughtering and Processing .....
Grading . . . . . . . .
Problems of Supplying Manaus. ....
Problems of SupplYing Boa Vista ..
VIII. W~HITHER THE BEEF INDUSTRY OF RORA~INA?.
GLOSSARY. . . . . . . . . . .
APPENDICES. . . . . . . . . . .
APPENDIX A SAMPLE QUESTIONNAIRE. ......
APPEr~DIX B PROPERTY AND HERD OF GADO DA
AMAFZONIA, S.A. (1967). .........
BIBLIOGRAPHY. . . . . . . . . .
BIOGRAPHICAL SKETCH ...............
LIST OF TABLES
1. Air Temperature Regimne--Boa Vista, Roraima.
2. Rainfall Regime--Boa Vista, Roraima ....
3. Radio-Telegraph Network; for Ror~aima
Territory,r 1954 . . . . . . .
4. Total Cattle Population for the Rio
Branco Savannas ... .... ..... ..
5. Ranch Facilities, Roraima Territory/, 1967 .
6. Rural Properties--Classes of Area
Territorio Federal de Roraima, 1967 ....
7. Rural Properties--Land Statute Tylpes
Territdrio Federal de Roraima, 1967 ....
8. Average Monthly Slaughtering Rates for
Mlanaus and Boa V/ista, June 1.963 M~ayy 1966.
9. Brazil: Per Capita Meat Consumption,
1948 1967 . . . . . . . .
10. Per Capita M~eat Consumption for Selected
Countries, 1967 .. ... . ... .
11. Bellm: Slaughter of Cattle and Buffalo
According to Region of Origin, 1962-1967. .
LIST OF FIGURES
1. Physical Features, Alto Rio Branco. .. .. 24
2. Climatic Regime, Boa Vista, Roraima
(19J58-1963) . . .. . 31
3. Climatic Regime, Manaus, Amazonas
(1958-1963) ... .. . ... . . . 32
4. Boa Vista in Isolation. .. . ... . 45
5. Access to the Alto Rio Branco .. . ... 47
6. The Rio Branco During Dry Season.. ... 48
7. The R~io Mlucajai . . 53
8. Crossing the Rio Mucajai. .......... 54
9. The Selva of Amazonas and Roraima .. .. 57
10. The First Fazendas, Alto Rio Branco,
ca. 1800 . . . 89
11. Alto Rio Branco Cattle Population: 1967. . 98
12. Example of the Area Mlap Show~ing the Locations
of Ranch Headquarters and the Cattle
Populations . .. .. .. .. . . 100
13. Example of the Area Flap Showing Locations of
Chutes, Refrigerators, Vehicles and Landing
Strips. .. .. .. .. .. .. . . . 101
14. PA Douro Cattle .. .. .. .. .. . . 106
15. A Nellore Bull. .. ... .. .. . . 108
LIST OF FIGURES (Contin ued)
16. The savanna Landscape After Repeated
Burning. ... .. .. .. ... .. 139
17. Uncontrolled Burning in the Heart of the
Dry' Season . . . . . . . .140
18. A Boieira. .. ... .. .. .. .. 200
19. A Balanca. .. .. .. .. ... .. 203
20. Alto Rio Branco Cattle Loading Points. . 204
21. The EMatadouro Industrial de Boa Vista. . 211
22. Average Ilonthly Slaughter of Cattle,
June 1963-May 1966, Boa Vista and
M~anaus . . . . . . . . .216
23. General Pattern of Cattle Movement
in Amaz~nia. . .. ... . .. .. 240
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of th~e
Requirements for the Degree of Doctor of Philosophy
THE BEEF CATTLE INDUSTRIAL IN~ THE RORAIIMA SAVANHAirS:
A POTENTIAL SUPPLY FOR BRAZIL'S NORTH
Thomas Fisk Kelsey
Chairman: Professor Rayrmond E. Crist
Major Department: Geography
Brazil is a nation currently undergoing rapid change.
While population growth continues, more and more people are
moving from rural areas to urban centers. Almost unique among
the developing nations is the fact that Brazil's economy is
moving even faster than the rapid population changes and the
urban consumers are gaining real purchasing power on a grand
There are, however, certain regions of Brazil which have
some very real supply problems in the face of promising eco-
nomic grow~th. Mlanaus, the booming urban center of Western
Amaz~nia, experiences annual critical shortages of beef, the
most desirable source of animal protein in the diets of its
people. As a greater share of Manaus' population achieves a
suitable income, the demand for beef, a more expensive item
than the traditional high-caloric food staples, increases to
true crisis proportions.
Ly'ing some 400 miles to the north, the savannas of
Roraima hav.e produced beef on an extensive grazing basis for
nearly twso centuries, yet little material progress in pro-
ductiv~ity' has been shown since the inception of ranching
there. The ranches of Roraima are completely isolated from
major markets, save for two months out of th~e y~ear, when
cattle can be sent by river barge to Mlanaus. An al~l-weather
road will soon bring this savanna area out of isolation and
offer an opportunity for the ranchers to supply Manaus during
its critical periods of shortage.
It is the purpose of this dissertation to investigate
the supply~ pattern of beef to Manaus and to determine the
probable impacts of the road on the beef cattle industry of
the sav~annas of Roraima. An attempt will be made to demon-
strate the present degree of inaccessibility to and from
Roraima, to trace the historical revolution of ranching in
that area, determine the nature and extent of production and
marketing problems, and assess the Importance of the new road
as a possible key to alleviating the beef shortage in Mlanaus.
The ranchers in Roraima welcome the road as an avenue of
improved access to material inputs required to improve the
quality of their cattle and the performance -of their produc-
tivity, but fear that the road will lead to the creation of a
cattle complex closer to Mlanaus, thus pre-empting their
present position with respect to that market. They vliew the
negat~ive turn of events which have occurred on Mlarajd Islan~d
as a result of the completion of the Bolem-Br;asilia road as
an omen of their future and look to distant foreign markets
as their salvation. Given the magnitude of th~e problem of
overcoming foot-and-mouth disease and the growing demand for
beef in M~anaus, and viewing the situation in light of the
economic principles of opportunityy cost and the law: of com-
parative disadvantage, it appears that ranching in the
savan~nas of Roraima w~ill benefit greatlyl by thec new land
surface connection to klanaus. The ranchers w~ill havle greater
access to material supplies at less cost and will be able~ to
reach the Manaus market w~ith their cattle on a 12-month basis.
I NT PRODUCT ION
Typical of some of the other larger: nations of Latin
America, Brazil is characterized by a relative abundance of
land, and w~hen view~edd in the long-run, still enjoys a situa-
tion of underpopulation rather than ov.erpopulation. In pro-
duction, area and population, Brazil is the largest country~
in Latin Amlerica.. The nation has over 93 million people,
occupying a mainly tropical area larger than the 48 contiguous
United States. Differing from most of the underdeveloped
world, Brazil does not y~et face the Malthusian dilemma of
serious population pressure on the food supply.
This is not to say~ that Brazil does not havre a population
problem of enormous proportions. Perhaps the greatest tempta-
tion to lapse into a state of complacency~, surely a dangerous
attitude in this world of many problems, is in maintaining
the general belief that, given so much empty space, the nation
need not be concerned w~ith a high rate of population grow~th.
In this belief, Brazilians too easily~ forget that much of
this space is not yet physically productive (given present
available technical knowledge) or, ev.en if it is, it is
geographically too remote for years to come. M~ore important,
however, manyl Brazilians hal.e completely ignored the fact
that, in its economic effect, population growth places pres-
sure on the quality of the human resource as well as its
d is tribute ion Thus, even if one accepts the Iviewl that Bra-
zil's long-run optimum population might be 200 or even 300
million, the alacrityl with which they reach this size, i.e.,
the annual rate of growth, is a matter of urgent importance.
If the annual rate of natural increase continues at the
present level of 2.7 per cent, Brazil certainly wlill find it
almost impossible to ameliorate the population problem. At
such a rate of growth, it would appear that Brazil cannot
possibly allocate the massive resources required to expand
and improve its infrastructure--schools, public utilities,
public agricultural services, highways and communications,
and the like--and industrialize at the rate needed if it is
to absorb into economically worthwhile non-agricultural em-
ploy~ment a rapidly growing surplus of underemployed rural
people or unemployed urban dwellers. The agricultural sector
will suffer greatly as well. W4ith rural labor being kept
cheap and ignorant by high rates of population growth, the
modernization of agricultural techniques will seldom be
economically sound and will increase social problems b; dis-
placing labor by mechanization. Even at a much lower rate of
net population gr-owthh (for examplee 1.5 per cent), all of these
problems would still have to be confronted, but at least
gradual progress toward their solution would be feasible.l
Ai few statements are necessary to relate the previous
statements to the empirical situation. By 1960, after more
than 400 years of development, 90 per cent of Br~azil's 70.2
million people and 93 per cent of the cropland were located
in less than half the country's area and within 300 miles of
the coast. In terms of cultivated land, Brazil at about one
acre per capital did not r-ank~ far above some of the more densely
populated countries such as India and Pakistan.2 Despite con-
siderable progress since 1945, Brazil remains a dual economy
in two respects. On a geographical basis the nation is viewed
as the developed South and the under-developed Northeast and
Amazon regions. On a temporal basis, while a substantial
number of the people hav.e shared at least moderately in the
fruits of industrial and agricultural progress, the majority
remain poor. A large number of the latter earn meager incomes
in agriculture, many are underemployed in either agriculture
or in urban areas, and the remainder are unemployed. Ailong-
side this condition, the 1960 national census revealed an
illiteracy rate of 39 per cent and an average level of school-
ing of 2.4 years, emphasizing the existence of a skilled labor
shortage. Overall, annual per capital income is somewhat above
$300 U.S., but the regional distribution is very unev.en. In
the N~ortheast per capital income stands at $170,3 while in thie
state of Guan~abara (the Rio de Janeiro urban area) i~t is
almost three times thle national a.erage. This poses a serious
social and political problem and limits the effective sizce of
the consumer market.
Nonetheless, in the past two decades Brazil has made
tremendous strides. While population increased at an a\.erage
rate of sligh~tly ove~r 3 per cent in the 1950's and somewhat
under 3 per cent in the 1960's, real GN~P grew at an annual
average of 5.8 per cent in the 1949-1962 period. During the
1963-1966 period, GNP averaged only 3.25 per cent owing to
the inflation crisis during the 1962-1964 per:iod.4 NJineteen
sev.enty wa:s the third consecutive year in which Drazil
achieved close to a 9 per cent rise in real GNP. With the
population growth rate down to 2.7 per cent per annum, the
three year average GNP growth rate of 8.8 per cent translates
into an impressive annual rise of more than 6 per cent in real
per capital output.
The rapid rise of industry has greatlyr changed the
structure of the economy. Between 1949 and 1966 the combined
output of mining, manufacturing, construction and power
facilities rose about 256 per cent while agricultural pro-
duction rose by only 102 per cent.6 This has given rise to
an immense rural-urban migration and interregional migration.
About two-thirds of the population w~as rural in 1949, but the
1970 census showed that this had declined to 44 per cent.
Brazil's livestock sector as a whole lags behind other
sectors of the economy. During the last decade the prices
of livestock products hare increased fas~ter_ than the general
price level. The same has not been true for food crops. On
the demand side this reflects the strong desire of urban con-
sumers to substitute the more expensiv/e livestock products for
the cheaper cereals as their incomes rise. It also reflects
on the supply side the relatively inelastic response of
livestock production to rising prices, a co~nsequnce of the
technological backwardness and inefficiency of the livestockk
sector. Mlost assuredly, wherever urban market opportunities
exist and are not undermined by counterproductive price ceil-
ings, the producers of livestock products, even beef cattle,
are not only growing in number, but are also beginning to
follow many of the improved practices in use in more advanced
In general, however, livestock practices can still be
described in primitive or semi-colonial, chiefly characterized
by land monopoly, extensive land and livestock exploitation,
low technical standards and unsatisfactory labor relations.
Even those producers w~ho do strive for increased productive
efficiency find it difficult to achieve because they face
complex managerial problems (mnore so than crop producers) and
proceed on a purely empirical basis without reliable and
adequate technical direction. WJith an acute lack of techni-
cal and market information, improved practices are not
financially feasible, particularly since the prices of
modern agricultural inputs are rising more rapidly than
Within the category of livestock, beef is the major
source of animal protein in Brazil, accounting for nearly
three-fourths of all red meat and poultry consumed each year.
The overwhelming majority of beef is produced in the states
of Rio Grande do Sul, S~o Paulo, Mlato Grosso, Goia~s, and
Minas Gerais. The major markets lie in Brazil's South, in
the built up urban areas of Sho Paulo, Rio de Janeiro, Belo
Horizonte and P~rto Alegre. Any producing areas in the
thinly settled North are for all practical purposes excluded
from this market area by sheer distance if nothing else.
One area characterized by primitive ranching conditions
is the remotely situated Territdrio Federal de Roralma,
Brazil's northernmost political unit of state or territory.
size. Roraima contains natural grasslands to the extent of
some 16,000 square miles, which is approximately the total
combined area of Delaware and Mlaryland. It is possible that
this grassland could become the scene of modern ranching,
supplying the rapidly growing urban center of Mlanaus, Amazonas.
~IHuman population pressures pose no difficulties for this ex-
tensive activity as the population density for the entire
territory is less than 0.04 persons per square mile. Scrawny
range cattle have been grazed on these unimproved savannas
for nearly two centuries with the beef being sent to two
markets: Mlanaus (1970 population, 305,000); and Boa Vista,
the territorial capital (population estimates range between
14,000 and 30,000 with the former more likely to be repre-
sentative). Brazil's improving infrastructure, byt bringing
Roraima out of isolation, is now making it feasible to supply
beef and cattle products to areas of greater distance which
were formerly out of the market area owing to physical
barriers or prohibitive transport costs.
The objective of this study, therefore, is to examine in
detail, and in a geographical context, the present situation
of cattle-raising in Roraima and the prospects for further
development of this particular economic activity in that area.
Two major themes which the author believes are funda-
mental themes in geography will be pursued in this study.
One major theme is that of spatial arrangement and movement.
It is here that geographical elements may be characterized
as points, lines, or areas.' Furthermore, common geographical
problems may be thought of as a group of tensions created by
these geographical elements. Dimensional tension is created
between point- and area-occupying activities. For example,
on a national scale urban industrial centers may be thought
of as points, whereas agricultural activities are area-
Occupying. A dimensional tension is created by the fact
that man requires association with, and the products of, both
farmi and factory.
Another tension is that which exists between present
activities and past facilities and institutions, that is,
historical tension. In many studies past legacy may be the
single most important fact. Existing institutions and
facilities will always be lacking in suitability for the
present because man is always creating new activities which,
for greatest efficiency, necessitate new arrangements. This
tension is a fundamental geographical problem which arises
time after time, and, in fact, is a factor influencing the
beef cattle situation in Roraima Territory.
Scale, which can classify problems as local, regional,
national or world, is of primary importance. In the case of
Roraima, the spatial distribution on the surface of the
savannas cannot be ignored. This unit area dimension is a
basic concept. Mlany terms used by geographers, such as
density, rent, land value, crop yield, are compounds of the
ter~m "'unit area" with counts of other elements.
The second major theme is the relationship between man
and the env~ironment.8 In earlier times much attention of the
geographer wais focused on how the environment controls man's
behaior.Counter theories developed to this, but no example
of man's control of his environment was demonstrated by
acceptable method and, consequently, the concept is no longer
considered useful as a guide to geographic understanding.
Today the general concept accepted by geographers is
based on the notion that the physical character of the earth
means different things for different people. The significance
to mran of the physical environment is a function of the ob-
jectives, attitudes, and technical abilities of man himself.
The resource base provided by the earth takes a new meaning
with each chan e in any of the element of the human culture.
Thius, the geographer is interested in the process of inter-
action between man and the environment.
The choice of the specific tylpe of agricultural acti\.ity
at any location depends upon three prime factors. First are
site characteristics, i.e., characteristics of a particular
place such as soil, slope, micro-climate, and size and shape
of farm and field. Second are the situation characteristics,
which are defined as the elements of relati\.e position, such
as location with respect to market. The third and most
complex factor is that of technology and the socio-political
organization of the people.
Included in the socio-political organization of the
people is the polarization to subsistence societies and ex-
change (or commercial) societies. On then one hand, th~e sub-
sistence society can generally be characterized as having no
contact between groups, or at best only short range contacts.
Self-support is the rule and there is little exchange of
goods and ideas. Inherited culture prevails. Production is
labor intensive although productivity per worker is low.
Change comes about slowly~14 owing to an attitude of long-run
fatalism and the lack of an abstract image.
On the other hand, an exchange society can be described
as having long range contacts with a great exchange of
material goods and human thoughts. There is a high level of
adapted techniques. As opposed to the subsistence society's
labor intensiveness, the exchange society is capital intensive.
Labor productivity is high and rapid change is a characteris-
Within this framework of thinking, then, the beef
cattle industry of the Roraima savannas will be discussed in
the following chapters. But, of course, to make a valid con-
tribution to the learned world, and geography in particular,
suitable methods wouldd be selected to investigate the problem
in an orderly and scientific manner.
1These comments are based on those by William H.
Nicholls, "The Brazilian Food Supply: Problems and Pros-
Pects," Economic Development and Cultural Change, XIX
(April 1971), p. 379.
2U. S. Department of Agriculture, Economic Research
Serviice, Brazil's Position in World Agricultural Trade,
ERS-Foreign 190 (washington, D. C.: Goviernment Printing
Office, 1967), p. 1.
3"News Review: Brazil, Northeast, Income," Bank of
London and South America Reviewi, v, N~o. 57 (September 1971),
4Another indicator of progress revealed itself in the
1970 census as the rate of illiteracy had fallen from 39.5
per cent in 1960 to 33.1 per cent in 1970.
5Brazil: International Economic Surv/ey (New York:
Chemical Bank, September 1969), p. 4.
6Brazil: Economic Review and Outlook,Itentoa
Notes No. 173 (N~ew York: Chemical Bank, July 1971), p. 1.
7The ideas expressed here are drawn from lectures by,
conversations w~sith, and readings from Professor John D.
Nystuen, "Identification of Some Fundamental Spatial Concepts,"
Papers of the M~ichigan Academy of Science, Arts, and Letters,
XLVIII (1963), 373-84.
tThe order of discussion does not imply that one theme
is more important than the other. On the contrary, they are
of equal importance and inseparable in reality.
Outstanding researches and able formulation of bypo-
theses, like outstanding in-entions and skillful manufactur-
ing, fail to achieve notable results unless effectively
presented to a large public. It is often and truly said that
research is not complete until it is published. How!ever, as
Boring has pointed out, a badly~ written report may find no
public because it is too forbidding to be read. "What is
in the writer's mind becomes public opinion only when many
other persons read ,,and understand him." Thus, while effec-
tive presentation can never take the place of able invrestiga-
tion, it is the indispensable means of assuring full success
to any investigation. The following paragraphs will discuss
the selection of a method of inquiry and presentation.
Various levels of methodological sophistication warrants
the selection of the most appropriate method to investigate
the stated problem.2 As it is so aptly stated by Gould and
In these days of complex interregional,
nonlinear, and dynamic models one hesi-
tates to utilize simple means to approach
the problem under investigation.
However, frequently these beautiful
pr-ogranmming models havep been con-
structed in a developed en-rironment
whlere pr.ofuse and dependable data,
both for the formulation of the model
and it-s empirical test, are close at
Inl underdeveloped countries reliable data are almost non-
exsistent and the categories of available recorded material
leave muich to be desired.
The inappropriateness of using a dynamic quantitive
model to aid in the studyl of the cattle situation in Roraimra
can be seen in the followings example. Halter and Dean have
used simulation to find some improved management policies
to deal w~ith uincertain environments in which decisions were
made about buying feeders for the range, trans herring feeders
to the feedlot, and selling fat stock. To construct a model
to simulate monthly~ range conditions over a period of several
years, the authors needed a data source which would reveal
actual range conditions at the beginning of each month.
Fortunately, the United States Crop Reporting Service repre-
sents th~e desired source of information. In Brazil no such
organization and service exist. Furthermore, the data needed
to formulate decisions on direct buying for the feedlot in-
cluded expected slaughter price, final weight, days on feed,
feed cost per head per day, initial weight and an uncertainty
factor margin. Ranching and becef production in Roraima
Territoryl have yet to r~each~ the level of advancement w~hereby~
such statistics as just mnentioned are recorded or can be
computed. wJiselyl, the authors of the article on simulation
In farm mana~gemntcll research at the
firm leel...some degree of caution
should be raised in recommending~ simu-
lation as a general melans of analysis.
It would be templLing to uise simulation
for almost: any~ size of problem;; but, it
should be made clear that unless simu-
lation proides answers to questions
which cannot be an~swered by simpler
techniques, it is dou.btful that it should
be used at the firm~ le.el.5
What Halter and Dean actually sho::ed was that ranch manage-
ment decisions did not, in fact, require the use of simula-
tion as an expectation model to improl.e the efficiency of
the decision to buy and sell stock.
Also, it can be safely said that the ranch managers in
the United States most likely have some degree of education
and are fully literate. In the example previously cited the
managers made almost all critical decisions of buying and
selling. To the contrary in Roraima, Ivery often the manager
is illiterate or can barely read and write and does not make
important decisions.. Instead, the absentee owner, very much
out of touch with the production aspects but more informed
about the market situation, makes the decisions. What is
missing in Roraima is the almost instantaneous flow of in-
formation .ia mass media which indeed exists in the U. S.
Gould and Sparks cry their tale of woe upon discovering
that in underdeveloped areas the quantitatil.e models hav.e
far outrun the a.ailable data. Because of this, simple models
may suffice to carry thie investigator past areas of unreliable
or incomplete data and smooth ov.er the pitfall of component
error. The authors concluded that the safest route of ap-
proach to the problem was to utilize a simple linear model.
One critical assumption they made was that the high prices
from large city markets reflect supply and demand in the near-
by' rural areas. In the case of Roraima this assumption would
not be valid, as all prices are set byl SUlaAB (Superintendincia
Nacional de Abastecimento), an autonomous government agency,
which, in fact, wreaks ha.oc on the free-w.heeling system of
supply and demand. In addition, the model would ha.e to
take into account the effect of political boundaries and
varying national economies. Thus, even a primitive." linear
model would still not operate if applied to the Roraima
Quoting the gentlemen mentioned abov.e, "In all charity,
perhaps we should remember that the Mlodel T frequently got
through and pro.ided useful service when badly potholed roads
tore sleeker models apart." Similarly, one does not send a
regiment to capture a thief wihen a few. policemen can accom-
plish the task. This author, for the purposes of this
dissertation, feels that a qualitativ.e approach is more
apropos than a quantitative approach. Owing to a lackr of
historicall; continuous records and unreliable existing data.8
the investigator turns to the process of deductive-inductive
analysis based on field evidence, observation, an~d interviews.
In the field, the investigator sought and acquired
published material, written in Portuguese an~d unavailable in
the United States, from official government offices and
agencies in Mianaus, Amnazonas and Boa Vista, Roraima. In
M~anaus, the lib-raryl of INJPA (Instituto NIacional de Pesquisas
da Amazo~nia), which is the NJational Institute for Amazonian
Research, pr~ov.ed to be a suitable beginniing point for further
library research. Other pertinent information was collected
from the Public Relations Sector, Secretary of Production,
State of Amazonas (Estado do Amazonas, Secretaria de Producao,
Setor de Relagoes Publicas). The official government sources
utilized in Boa Vista w~ere the Mlinistry of Agriculture of the
Territoryl of Roraima (M~iniste/rio da Agricultura), the office
of the Division of Production, Lands, and Colonization,
D.P.T.C. (Divis~io de Produg o, Terras, e Colonizag o), the
statistical section of Fundagio IBGE (Instituto Brasileiro
de Estatistica, Inspetoria Regional de Estatistica, Boa
Vista, Roraima), and the office of the secretary of the
mayor's administration of Boa Vista (Prefeitura Municipal de
Boa Vista, Secretaria de Administraio).
Personal contacts in the two capitals led to both the
acquisition of undistributed published material and fruitful
personal interviews. Of course, before embarking on the
actual field work in Roraima, the author paid a visit to His
Honor Francisco Zangerolame, Mayor of Boa Vista, and His
Excellency Governor Helio Campos, Governor of the Federal
Territory of Roraima, outlining his investigatory intentions.
One of the writer's original aims in gathering desired
unpublished material from ranch owners and managers was to
use a questionnaire which had been both simplified and
modified from its previous use in Costa Rica. This question-
naire is shown in Appendix A of this study. It was immediately
discovered that a gross difference in the level of advance-
ment of beef production existed between Roraima and Costa
Rica, the latter being far more advanced, and that even the
modified questionnaire was too sophisticated. Two principal
reasons for this were, first, a majority of the ranch owners
were not available for interviewing and second, most of the
ranch managers had only a minimum ability to read and write.
Furthermore, as the study will reveal, labor is so inexpensive
that machinery and fencing are still rarities in Roraima,
thus making several of the questions null and void. To com-
pensate for the loss of this useful tool the author relied
on conducting lengthy conversations with ranch owners,
managers, government officials, local businessmen, and per-
sonal friends. In each conversation a prepared set of ques-
tions was wrov.en into the exchange and hence, a ?.erbal ques-
tionnaire was, in fact, administered.
1Edwin G. Boring, "Another Note on Scientific writing,"
Science, LXXXIV (1936) 457-469.
The word complexityl may be substituted for the wrord
sophistication for those violentlyl opposed to t~he so-called
3Peter R. Gould and Jack P. Sparks,"TeGoapil
Context of Human D~iets in Southern Guatemala," Geographical
Review, LIX (Januaryl 1969), 66. Hereafter cited as Gould
4For the explanation of data needed to construct a
simulation model, see, for example, A. N. Halter and G. W.
Dean, "Use of Simulation in Evaluating Management Policies
under Uncertainty: A~pplication to a Large Scale Ranch,"
Journal of Farm Ec~onomics, XLVII (1965), 557-573.
5Ibid., p. 573.
6Gould and Sparks, p. 66.
eThe following is an example illustrating the fallacies
of published data pertaining to the Roraima study. The 1968
volume of Anuario Estatistico published byl the Instituto
Brasileiro de Geografia e Estatistica (IBGE) indicates that
Roraima Territory exported 9,000 head of cattle during July
and August of 1967. This figure is official and comes from
the office of the ma yor of Boa V/ista. In reality all of the
cattle in 1967 were sent down the Rio Branco to Mlanaus. The
suppliers (marchantes) who come from Mlanaus must pay a fee
(imposto de barreira) to the mayor's office for each head of
cattle he has purchased and intends to ship to Mlanaus. In
1967 the fee was NJCrS23.OO per head. Officiallyl the suppliers
paid fees for 9,000 head of cattle which was entered in the
official records. Unofficiallyl, 12,000 cattle were purchased
and rather than pay the full fee, the suppliers bribed a
minor official with the sum of NCrS5.00 for each head of
cattle he did not record, which amounted to 3,000 head!
Th~us, the export figure is understated by 25 per cent.
THE PHYSICAL SETTING
Roraima, a federal territory of Brazil, with an area
of 141,009 square miles, approximately the same size as the
state of S~o Paulo or the combined areas of the Federal
Republic of Germany and the German Democratic Republic, lies
between 50 16' 19" N and 10 27' OO" S Latitude and straddles
the northern and southern hemisphere tropics.l Longitudinally,
the territory~ extends from a western extreme of 640 39' 30" W~,
eastward to an extreme of 580 58' 30" W. In terms of angu-
lar distance extremes, the linear distance from north to
south is 461 miles and from east to west it measures 392
miles. The entire region lies within the time zone four
hours less than Greenwich.
Roraima Territory is bounded internationally by Venezuela
to the north and west for a distance of some 594 miles and
shares a common border with the Republic of Guyana to the
east for 430 miles. To the southwest, south, and southeast,
Roraima is bounded by the Brazilian states of Amazonas and
That portion of the territory containing the savannas
upon which this study will focus is known as Rio Branco or
the Alto Rio Branco, and forms an area some 16,000 square
miles in extent.2 (Figure 1) The Rio Branco savannas form
a part of the Rio Branco-Rupununi region; an area covering
some 21,000 square miles. The eastern portion, covering ap-
proximately 5,000 square miles, lies within Guyana and is
commonly known as the Rupununi.
The Alto Rio Branco-Rupununi sav~anna landscape lies at
elevations ranging from approximately 300 to 1,000 feet, ris-
ing somewhat towards the margins of the watersheds. These
savannas form a gently rolling plain of rather monotonous
appearance within which isolated peaks and low ridges
occasionally rise to view above the galleria forests which
follow the streams as they meander across the landscape.3
The savanna region as a whole is enclosed by~ a series of
mountain ranges and elevated surfaces which rise 1,000 to
4,000 feet with a few isolated peaks of even higher elevation,
and are aligned in such a way as to produce a basin-like re-
lief, with the southern boundary of the basin lower and less
continuous. (Figure 1) The physiographic unit is broken
by two large drainage outlets. To the northeast, the
Rupununi River, draining the Rupununi portion of the savannas,
flows through the Essequibo system to the Atlantic Coast.
To the south, the broad alluvial plain of the Rio Branco
.~;) EICIDI II 1..1
C7 nrsolora 11)51;la1
rn..l 01 IJrmo.
~o r, ,,,,
bL t;. I
ALTO RIO BRANCO
drains waters from the remainder of the region and passes it
on through the Rio Negro, the Rio Amazonas and ultimately
into the Atlantic Ocean. The savanna landscape occupies the
northern part of the floor of this basin-like formation.
In terms of geologic structure, the Rio Branco-Rupununi
savannas belong to the Guiana Shield, an ancient pre-Cambrian
mass that forms the nucleus of the northeastern part of South
America. Between the granitic rocks and gneisaic rocks of
the basin rim lie the alluvial lowrlands of recent fluvi~al-
lacustrine deposits which occupyl approximately half the basin
floor. Scattered areas of more elevated terrain, between
350 and 500 feet, follow along the edge of the mountain rim,
thus forming a discontinuous pattern. These areas are com-
posed of undulating gravel ridges largely of lateritic origin,
laterite plateau remnants, and long gentle sheetwash slopes.
The lowland is considered to represent part of a planation
surface provisionally dated Late Tertiary and upon this
bevelled surface the contemporary landscape has been fashioned
by Quaternary deposition and erosion.4
In modern times the main tributaries of the Amazon and
Essequibo river systems are incised into the savanna surface.
For example, at St. Ignatius Mlission, Guyana (Lat. 30 20'N,
Long. 590 50'W), the Takutu river is 10 to 15 meters (33 to
50 feet) below the general level of the surrounding savanna
surface and similar conditions prevail elsewhere. Y'e t, in
spite of this apparent, though slight, landscape rejuvena-
tion, the regional drainage system is unable to carry the
high olume of surfaces runoff during the height of the rainy
season. As a result the river levels rise at the onset of
the rains and in man: places overflow their banks. Conse-
quently, ground water drainage in the savanna is restricted
and flooding and waterlogging occur over extensive low-lying
regions. In more elevated areas, conditions are less extreme,
but even here ground water levels frequently rise close to
the surface. As the rainy season wanes, much of the surface
flood water rapi~dly disperses (by runoff and evaporation)
and is accompanied by regional lowering of ril.er levels. An
increased ground water gradient is thus established and
ground water drainage accelerates. This process is greatly
facilitated by the existence of porous sand and gravel sub-
strata, both in the low~-lying alluvial plain and in more
elevated areas to the south and east. In a few places
ground water drainage is impeded by heavy textured horizons,
and many shallow ponds and lakes may appear for several
months but during this occurrence there is no general profile
impedance and ground water levels fall fairly rapidly with
the lowering of the rivers. It is evident that the
hydrological imbalance across the watershed region contributes
to the seasonal contrast of the savanna landscape.
Before discussing the climatic characteristics of the
Alto Rio Branco, the author wishes to express a word of
caution. The Federal Territory of Roraima contains only one
official climatic data collecting station, located at the
airport in Boa Vista, the territorial capital. Observations
are made by the N~eteorological Station of the Air Route Ser-
vice, Q. G. 1st Aereal Zone, Ministry of Aeronautics (Estac o
Metrolgic doServico Retas Aereas, Q. G. 1 Zona Aerea, do
Ministerio da Aeroniutica). The data collected at Boa Vrista
apply only to the region of the Alto Rio Branco, that is,
the region of savannas upon which the cattle-raising takes
place, and cannot be generalized to include the region to the
north and to the south. In the extreme north lies the
mountainous region with elevations reaching a maximum of
2,875 meters (i.e., Mt. Roraima, elevation 9,219 feet), while
the station at Boa Vista is situated approximately 120 meters
(400 feet) above sea level. Also, the data cannot be
generalized to include the lower reaches of the Rio Branco
because there climatic conditions are different. In the
lower Rio Branco the vegetation is that of a dense rain
forest while the Alto Rio Branco is dominated by savannas.
However, the data obtained from the weather station at Boa
Vista are indeed pertinent to this study.
The outstanding characteristic of the Alto Rio Branco
climate is the marked seasonality of the rainfall. Tables 1
and 2 and Figure 2 reveal that at Boa Vista 78 per cent of
the mean annual total of 1,740 milimeters (69.5 inches) fell
during the four months of May~, June, July' and August, 1958-
1963. By comparison, the rainy' season in the central Amazon
Basin to the south extends ov.er a much longer period. At
Manaus (Figure 3), an annual average total of 2,128 milimeters
(85 inches), over 100 milimetem per month, fell during nine
months of the y~ear ov~er the same period of time.
Mean monthly temperatures at Boa Vista are fairly typical
of the Amazon-Cuiana region, with an average maximum of
29.loC (84.4oF) in June, during the height of the rainy sea-
son, and an average maximum of 33.3oC (91.90F) for October,
in the heart of the dry season. The mean monthly lows for
the same two months wrere 23.3oC (73.90F) and 23.6oC (74.5 F),
respectively. Thle mean diurnal range, 6.5 CO (11.7 Fo) to
9.7 Co (17.5 Fo), which is one of the most attractive features
of the savanna climate, is high for a lowland location so
close to the equator. High wind speeds in the months follow-
ing the rainy season help to dry out the inundated savannas
but at the same time greatly/ reduce the effectiveness of the
Vegetation in the Alto Rio Branco can be placed into one
10 -I I co
BOA VISTA, RORAIMA
of four main categories.5 Most of the mountainous areas, at
least on their lower slopes, contain a dense corner of rain
forest. Occasionally, there are almost pure stands of pau-roxo
(purple-heart, Peltoqy~ne densiflora), a vJaluable timber tree
representing a potential resource for Roraima. The rain
forest merges into a fringing semi-deciduous forest, where
certain species lose their leaves during the dry season. A
similar type of forest occupies most of the "bush islands"
in the savanna.6 Second-growth associations tend to pre-
dominate over what are considered to be climax forms because
of the extensive practice of range grazing and repeated burn-
The savanna association is primarily of the open scrub
or woodland type. The density of the scrub varies from open
woodland formation to almost pure grassland. The dense scrub,
known as campo cerrado in the extensive savannas of Central
Brazil, is a rarity in the Alto Rio Branco. The dominant
scrub form is the fire-resistant pau-lixa, the sandpaper tree
(Curatella americana), and Trachlypogron plumosus is the
dominant grass. On iron-stone ridges and white sand the grass
grows alone, while on brown sand uplands it is joined by
other grasses such as Andropoqon anqustatus, Aristida seti-
folia, Axonopus chrqysites and Mesosetum loliiforme. The
grasses may reach over three feet in height and are found in
bunches six inches to one foot apart. On lower grounds where
there is likely to be flooding during the rainy~ season,
Trachy~pogon is supplanted by a mixed cover which includes
sedges and some low shrub forms.' In the baixas (extensive
depressions) the ground is occupied almost entirely by sedges.
Flood water may stand in these areas to a depth of several
feet for up to five months.
Although the above description and classification of
the regional vegetation is based on the excellent work. done
by Hills and others at McGill,8 other systems haver been applied
on the basis of regional terms for vegetation. The work of the
Brazilian geographer, Antinio Teixeira Guerra, exemplifies
the use of such terms.' He states that the vegetation cover
of that region is characterized by savannas that are more or
less "wooded."' The savannas of the Rio Branco represent a
series of different appearances which have regional names such
as campos agrestes, campos de baixada and campos cobertos.
The campos aqrestes is a type of vegetation having an
herbaceous cover, without shrubs and rarely with low shrubs.
The grasses are isolated or are dispersed in tufts, leaving
the soil, for the most part, uncovered. This term is also
applied to an area having a dense and continuous cover of
grass or herbaceous plants, which is not the case found in
Roraima. The campos agrestes, also called lavrados, correspond
to the true plains, being almost de.oid of trees or even
On the campos agrestes the existing grass is not pala-
table to the grazing cattle who look, instead, for the camp~os
do baixada, which corresponds to humid pasturles. Phlysiognomi-
cally the onlly feature which differentiates these areas fromn
the lal.1ado is that in the campos do baixada (or- simply
baixilr) the grass forms a continuous co.er, is tenderer, and
contains somec buriti plam, by virtue of th~e greater hulmidity.
Camouos cobertos is the name given bl the local inhabhi-
tants to the sav~annas up~on which trees and shrubs begin to
a~pper. This grades into, w~hat is known as the cerradoes,
w.hich is defined as w:aterless scrub wastelands.
Although the classification and description of the
v.egetation gi.en Guerra falls far short of than gi.:en- bl Hills
and others of the McGill Univ.ersitl Sarvanna Research~ Project,
it is included here to showI that regional vegetational termsn
do exist in the Alto Rio Branco and that the local people do
have an awareness of the varietl of their physical environ-
ment. The author found that among the ranch hands and rural
dwellers the term most frequ~entlly used in a blanket fashion
when asked to identify grasses, trees, soils', etc., was
The soils of the Alto Rio Branco can be characterized
as being highly acidic and veary low in all nutrients and
bases. Chemical weathering of tropical soils occurs at
approximately three times the rate of weathering in middle
latitudes. This leads to very deep senile soils. The deep
weathering occurrence is particularly~ the case where the
soil surface is easily penetrable, such as on the lateritic
ridges where drainage is rapid. The excessive eluviation of
clay and organic matter reduces the water retaining capacityl
of the surface soil. The resulting dcop desiccation in the
dry season further increases: the loss of soil constituents
and thus intensifies th~e detrimental cycle. If~ the soil sur-
face becomes compacted by~ h-eavy rainfall, high insolation, and
the removal of vegetation by~ fire, then sheet :ashl, gully~ing
and severe erosion generally occur. Hiumus is rapidly decom-
posed in the tropics owing to the increased activity of the
soil bacteria in higher temperatures.
Referring to the studies of the MCGill Univ.ersity Savranna
Research Project, there are perhaps three main intergrading
groups of soils in the Roraima savannas:
a. Soils of the Flats: These are the lowland soils.
They are pale colored silts and clays and
occasionally havre a thin organic cover. In some
places they become yellow in lower horizons, and
occasionally contain red iron or black manganese
nodules. These soils are flooded regularly,. then
desiccated, and the result of the fluctuating
water table appears to be the promotion of lateri-
b. Gra:.elly? Soils: These occur on upland sites and
are composed of secondary laterite or "peridigon."
These contain pebbles ranging from three inches
in diameter to large pisoliths, set in a matrix
of red earth which becomes paler in the lower
c. Gra; Soils: These occur also on upland sites,
but are composed of a mixture of gray sand,
and silt or silty clay. N~assivle or primary
laterite occasionally outcrop in both these
From this brief introduction to the landscape of the
Alto Rio Branco in Roraima it is evident that a far from ideal
natural environment for human land settlement anid occupation
exists. The soils are lacking in nutr-ients andr~ L!e soil
structures leave much to be desired. I-lost of the natural
vegetationn is not palatable to grazing animals and the distri-
bution of rainfall throughout the year is ex;tremely~ unfav.orable
both to the formation of a habitable environment and to man's
activities. In order to overcome such physical disadvantages
any form of economic activity would require technical
knowledge, skills, material supplies, and large capital in-
puts. In the chapters which follow the author hias I.iewed in
the field the consequences of this relationship between man
and the physical milieu.
The Federal Territoryl of Roraima was officially
created by Law 5,812 of September 13, 1943. Prior to this
time the area was part of the state of Amazonas. Although
the territory was legally authorized In 1943, it was not
until June 20, 1944 that the territory actually came into
existence. At that time it w~as known as the Federal Territory
of Rio Branco, taking its name from the river. In mid-1963
the official name of the territory was changed from Territorio
Federal do Rio Branco to Territorio Federal de Roraima.
SotAmicnSvnsComparative Studies, Llanos
and Gu:yana, MlcGill Univ.ersityl Savanna Research Project,
Savarnna Research Series, No. 5 (Montreal:: Department of
Geography, MlcGill Univ.ersity, December 1966) p. 10). (H~ere-
after cited as Llanos and Guyana); Antanio Teixcira Guerra,
Estudo Geogrifico do Territdrio do Rio Branco, Biblioteca
Geogra'fica Brasileira, Serie A ''Livros," No. 13 (Rio de
Janeiro: Instituto Brasileiro de Geografia e Estatistica,
Conselho Nacional de Geogra~fia, 1957), p. 2.
3These isolated peaks are inselburg-type landforms
which primarily take the form of small, usually rounded and
4Llanos and Guvana, p. 10.
5The classification of vegetation is extremely compli-
cated and confusing. For the sake of clarity and understand-
ing, the author will state what he feels are the most concise,
but accurate, statements pertaining to vegetation. For those
extremely interested in more detailed information about the
vegetation of this region, consult the McGill University
Savanna Research Project, Department of Geography, McGill
University, P.O., Canada.
6The "bush island" phenomenon is very common in the Rio
Branco-Rupununi savannas. As the term implies, the formation
of forest or dense scrub appears on the savranna landscape as
an island, often circular In shape. Most likely the bush
islands are remnants of earlier forest formations which have
been greatly reduced by burning and grazing activities. In
some bush islands forest regeneration begins, but, because
of a regional lowering of the water table, it is unlik~ely cou
progress very far. The bush islands probably~ will fall prey;
to savranna fires in the future.
7jMost likely, among the sedges would be found Rynchosgors
spp., Dichromena spp., Seleria spp., and Stenophy.llus spp.
The sedges are less strongly tufted and grow more closely: to-
gether than in the grass-dominant areas. Some shrub anid
woodyl herb forms, small and neither bushy nor dense, would
include Psidium spp., Palicourea spp., Casearea spp., Pandia:~_
spp., and Erlythroxylum spp.
8Llanos and Guvana.
Guerra, pp. 1413
10Guerra, writing in Portuguese, uses the term cer:.ade in
the phrase mais ou menos cerrados. The term cerrada has
several meanings, but w~hen used in reference to vegetation it
means a sav~anna w~ith scattered thickets of deciduous sc-rub
11Jacques Ourique (O Va'le do Rio Branco), describing such
a landscape, expressed himself in the following manner:
"The sufficiently elevated lands, the higher and more ru~ggd
the more they advanced in the direction of our frontiers with
British Guiana and Venezuela, die out along the banks of the
Rio Branco almost always in steep river banks, in the ma-
jority. of the cases elevated several meters. In that region,
so similar to the Argentine pampas and more beautiful because
of the distant chains of mountains that interrupt the hori-
zon, the entirety of nature amasses itself so that it may
become, one day, the principal and most abundant and most
diverse granaryr of this great state.
'There, beautiful green plains which extend, beyond
sight, in gentle undulations, in serene and sad continuity,
predominate, interrupted only by one or another isolated
stand of trees or by the rare groups of elegant miritizeiro
Fiber mauritia palm] The landscape is distinguished almost
always byl clear sand-bottomed lakes, devoid of vegetation,
where the thirsty~ steer quenches his thirst in the hours of
sul t rine ss ." (Ia na u s, 1906) p. 8.
12The appearance of the term agret in R~oraima is
most interesting. The overwhelming majority of people pre-
sently in Roraima, although no precise data are aailable,
ha-e either recently immigrated from or are first generation
immigrants from the sertac of the Northeast states of Ceara,
pioui, Pernambuco, Paraiba, Alagoas and rMaranh o. It is in
this region of the NJortheast that the true agreste exists.
E'.en though some sav.anna areas in Ror-aima .aguely resemble
the agreste of the NJortheast, the species of vegetation,
soils, and climatic conditions of the former differ from the
latter. What has happened then is that the uneducated people
coming from the ru~ra~l Northeast. have' spread7C the term aqreste,
but the true meaning of the term, at least in the case of
Roraima, has been abandoned.
13Here the author is making an assumption, backed by
field observation and the findings or the !Instituto de
Pesquisas e Experimentaqi~o Agropecuarlo do NJorte, Eclem!!,
para'. Hills states in his article "'The Interior of British~
Guiana a~nd the Myl~th of El Dorado" (Canadian GErc~raphe, II,
19~61, 34) that the R~upununi sav.annais are mierely~ an appe.ndaige
of larger Roralmna or A~lto Rio Branco savannas. Those work-,
ing with~ the M~cGill Univ~ersity. Sav'anna Research Project hav.e
dr~awn froml their ow.n findings and the findings of e:-:tensivee
investigations by~ the Regional Research Centre of the
B3ritishi Caribbean (an organization based ini Trinidad which
car-riedi out ex~tensiv.e soil sampling in Rulpununi in the late
1950's) a number of generalizations, and ha\.e arrived at
three main groups of intergrading soils. The author is
assuming that these generalizations hold for the Alto Rio
Branco as w.ell. The possibility of this being an erroneous
assumption is acknowledged.
14Llanos and Guvana, pp. 11-12.
One of the key factors of modern exchange societies is
the rapid and efficient distribution of goods and information.
This is largely determined by the quality of highw~ay and
waterw;ay' networks, dependable mail and telegraph systems and~
by' the existence of an7 informed populace instru~c:ced bl' mass
media such~ as radio, television, and newspapers. For those
in Roraima, access to markets, goods, and information is most
difficult. The converse is also true for those mecrely. wish-
ing to travel to Roraima, or more significantly, to establish
business relations, or ship bulk items to or from the area.
The Territory of Roraima is within a region where, until
a relatively short time ago, the only mode of transportation
outside its boundaries was that of river navigation. As if
this were not enough of a limitation, travel was, and still
is, further impeded by the fact that during the "summer"
(i~.,the dry season which occurs from September to M~ay)
the waters of the Rio Branco fall so low that they eliminate,
for all practical purposes, commercial transportation.
The possibility of constructing a road or a railroad
connecting Mlanaus and Boa Vista has long been under considera-
tion. In a public address, the governor of the State of
Amazonas in 1905, referring to the construction of a railroad
connecting Manaus and Boa Vista, remarked inter alia:
In addition to the rail line I mentioned
to you previously, the contract was signed,
by virtue of Law 493 of October 23, 190O5,
w~ith the engineers Hlumberto Saboia de
Alburquerque and Hermano de vasconcelos
Bittencourt, for the construction of a
railroad which, starting from Campos Sales, ends
at the Jauaperi River. Upon reaching there,
by decree number 775-A of April 30 of the
same year, a decree which w:ill opportunely
be submitted to the consideration of this
Congress, the line would be extended to Boa
This particular endeavor never saw the light of day.
Aiugu'ste Plane (1903~), w~hen discussing the fun~ctions of
the Rio Branco in its connections w~ith Guylana (then Bri-tish~
Guiana) together with the difficulties caused by the dry
season and the numerous river rapids, stated: "If thie piro-
ject of the construction of a railroad from Manaus to the
Rio Branco were executed, this river would experience an
extremely rapid colonization."3 Once again, no railroad
and no rapid colonization along the Rio Branco came into
More recently, interest in reducing the isolation of
the Alto Rio Branco has intensified. In 1938, however, a
civil engineer, Raimundo Pereira da Silva, issued a
cautionary statement concerning the current desire to improve
travel and communications between the city of Mlanaus and Boa
Vista and the region of the Alto Rio Branco. After thorough
studies and the organization of a project (complete w:ith a
budget) for perennial stream navigation on the entire course
of the Rio Branro, from its delta into the Rio LHegro to th-e
confluence of the rivers M~au and Tacutu abov~e Bosa Viista, he
felt able to state that the clearing and channeling of the
Rio Branro's course was phy'sicallyl quite feasible and might
ha.e far--reaching economic conisequences. It would, hiowev.er,
be too great an expense, in his opinion, for the reallities
of this backwaerd region w.hich, although it did have' num~ero~us
and vast potential rich resources, did not ylet have a popula-
tion large enough to represent economic production capable
of justifying the huge capital investment necessary to under-
take such a project.4
Ev~en though the pessimistic v.iewe expressed byl Pereira
da Silv.a and others has continued to the present dayl, the
Brazilians have begun to carv.e a road through the rainforest
in an effort to bring Roraima out of its isolation. This
undertaking will be discussed later in this chapter.
The time necessary for human travel to Roraima has been
drastically reduced by the advent of commercial and military
air transportation. Boa Vista is now~ directly accessible
via the Brazilian airline Cruzoiro do Sul from Manaus three
times weekly, and Georgetown (Guy'ana) tw.ice weekly'. While
this has facilitated human access and the carriage of mail
and h ig h-a lue,/10w-b-ulk g oods shipments of building materials,
most consumer goods, and cattle remain prohibitivelyl ex-
pensivc owingj to re_7atil. e isolation or inaccessibilityl which
remains as Roraima~'s major impediment to joining the modern
Travel and com~munication difficulties present a serious
hindrance to ccenomic anid administrative activities. Those
living in the sa\annas of the territory become completely
ilhadas, that: is, isolated, during the rainyl season because
all roads and manyl of th~e air strips become inulndated.
Similarlyl, during th~e dryl season, Boa Vista faces grave
shor-tages ofi supplies ow.:ingj to thle extr~emely~ low level of
the Rio Branlco which becomes unnavigab-le.6
Even in terms of nic miles from Boa Vista to populated
areas representing markets and centers of industrial pro-
duction the region's extremely isolated situation is easily
discerned. (Figure 4)
To emphasize the absolute lack of access by land and
discontinuous modes of ingress and egress bj, other routes,
each category of transportation and communication may now be
Even though the airplane has brought the territory
closer to a major center, i.e., Manaus, in terms of travel
time, the river still r-emains as Roraima's life line with
thie exterior. As a transportation route, however, the river
leaves much to be desired. During the rainy season, M~ay
through- August, navigation is possible without hindrance from
M~anlaus as far north as Caracarai, a distance of some 420
miles. From Caracarai to Bsca da Estrada, a stretch includ-
in~g ma~ny rapids, navigation requires a very experienced and
skilled rive.r pilot whlo has a detailed knowledge of these
r-apids. Froml the latter point to Boa Vrista na.igation is
on~ce again free from hazards. (Figure 5) During the eight
m'1onLh dry season several sand beaches and islands appear as
th~e level of the river recedes. Accompanying this allu-:~iation
there is constant movement and relocation of navigable chan-
nols, making passage extremely difficult (Figure 6)
On the passage between Manaus and Boa Vista the average
trip upstream requires eight to ten days and the trip dow:n-
stream four or five days during the rainy season when the
river is at high levels. During the dry season anywhere from
15 to 20 days are required to ascend from Manaus to Boa Vista
while the return run takes from eight to ten days.
VENEZUELA ) -50N
~s~J-~) ~ Lethem
\ ~~~BOAI VIST~ B-r
.../ Baoc da Estrodl)
BR-174, BR-401 Federal
0 50 100 150 200 Km
O 25 50 75 100 M.
The Rio Branco During Dry Season
A view from Boa Vista looking southw~estward. This
portion of the river at this time is navigable only
for small craft. Further south, rapids impede
nearly all traffic at this time.
On the lower Rio Branco' launches and motor boats of
about fivle tons, dlead weight, and of shallow draught haul
barges which can carry cargoes of 50 to 60 tons. On the upper
reaches of th~e river the transports are generally smaller and
carry lesser loads.
An immediate consequence of the disrupted nature of the
navligation and the small load capacity of the river crafts
is the high price of commodities in Roraima caused by the
highi cost of freight. Since almost nothing is therefore pro-
duced in th-e territory and practically everything must b~e
imnport~ed from the ex3teri~or, the cost of living in Ror-aima
inevcitably h-as to be very high. It is not; at all unusual~7 to
pay' at least double wh:~at one would normally pay' in Rloo de
Janeiro or Shio Paulo for the majority of consumer items.
To date, the Alto Rio Branco does not have any roads
which allow access to the area from the exterior. This,
coupled with the seasonal problems of navigation, has kept
Roraima in isolation. Yet after hundreds of years of dis-
cussion and expectation,1 an all-wseather road is now, in
fact, being carved through the rainforest of Amazonia.
It was felt by many that the first link of a highways'
from Mlanaus northward to the Brazil-Venezuela border should
be a stretch of road connecting Caracarai and Boa Vista.
The village of Caracarai is located at the head of that
stretch of the Rio Branco which is freely navigable. From
Caracarai to B~ca da Estrada, a fluvial distance of some 30
miles, the smooth flow. of the river is interrupted by a
series of rapids and low falls, formed by resistant granites
and gneisses w:hichi outcrop in this area. From Baca da Estrada
northwa:ird the- river is once aIgain navigable for vessels hav-
ing a shallo:-' dr-aught. For at least eight months of the yecar
Caracarnii fu~ct~ions as Eoa Vrista's port. Prior to the exis-
tence of the IBca V.iSta~-CarTacar-ai road, supp~llies which arrived
in Caracarai frLomI fln .;us '.-'ere taken overland to Bica da
Estrada whe!lre they~ wer~e once again placed aboard waterciraft
for tr-ans-ship~ment to Boa Viista.
By virtu~e of the signing of Presidential Decree Number
18,50,9 of lNovembe~ljr 23, 1928, the Benedictine missionaries who
were in th~e territory initiated the road on behalf of the
Industrial Company of Rio Branco, which belonged to the
Monastery~ of S~io BEnto. This work progressed only a short
distance and was ultimately abandoned. During 1944 and 1945,
wrhen the road project wras renewed under the governorship of
Enio Garces, it was pushed further southward but still did
not reach Caracarai. After another halt in construction the
road finally reached Caracarai during the summer of 1948-1949.
The stretch approaching Caracarai apparently resembled a
trail. This road underwent tw~o more cycles of abandonment
and reopening in 1950 and 1953.
The principal reason for the slow and spasmodic con-
str~uction of the Boa Vista-Caracarai road was primarily poli-
tical. W~hen the National Highw~ay' Plan (Plano Rodov/iario
unc~~ional) was formulated in the 1940's, at least one road in
each of the territories in the A4mazon Basin, which would be
considered vital to national interests, wlas planned by' the
federal go.errnment. Unfortunately,, in the study of the dis-
tribution of the Na~tionsal H~ighw.ay Capital Fund, the Na~tional
Cong~esss ex;clu~ded the territories from the pr-iv.ileges of
draw.;ing funds from this source. Fur thermor e, thee D~epar tment
of National Highw1.ays (Departamento N\acional de Estradas de
Rodagens) w~as unable to undertake construction of roads in
the territories w~ith its allotted sum from the Ndational High-
w~ay Ca ital Fund because those tentative roads w~ere not in-
cluded in the program entitled the Plan of Primary Urgency
(Plano de Primeira Urgencia) which was also the brain-child
of the National Congress.13
The construction of the road itself has been no small
feat: considering the meager building resources at hand and
the terrain over which the road has been established. The
first stretch of the road, from Boa Vista to the Rio Mlucajai,
a distance of 35 miiles, passes ov.er savranna landscape, sav.e
the final two miles to the river, which is through dense selvra.
In order to make this road usable throughout the y~ear, how~-
ev.er (i.e., during the rainy season as well as the dry sea-
son), it was necessary to construct eight one-lane bridges
(obras de arte) ovrer streams and then to face the problems
of travelrsing the Rio Iducajai, a river of considerable size
and depth~. (Figure 7) It w~as decided that, given the sparse
population and traffic flow it would serve, a bridge would be
too costly to construct. The solution was to install a barge,
pow'.ereld by1 a motorized auxiliary craft, capable of transport-
ing fully-10aded trucks, one or two at: a time, dlepending on
thie si~ze of the truck;. figuree 8) From the Rio Mucajai to
Caracarai th~e road has been sliced through 55 miles of jungle,
crossing no less than 15 streams and shallow lakes, breaking
out onto savanna once again only four miles from Caracarai.
Until recently the entire 94 miles of this road was used
onlyl during parts of th~e dry~ season. Today, howsev.er, the
situation has changed. Presidential Decree N~umber 61,599 of
October 24, 1967, published in the Official Journal of the
Union (Dliario Oficial da Uniio) on October 26, 1967, has
secured the political means to execute the Preferential Plan
of Federal Road works (Plano Preferencial de Obras- Rodoviarios
Federal), giving absolute priority~ to the program of the
The Rio M~ucajai
This is one of many sizeable rivers feeding into
the Rio Branco which must be crossed by ferryl or
Crossing the Rio Muca jai
As shown, the barge is capable of carrying fully
loaded trucks. The tug is privately owned but it
is employed under a government contract. On the
far bank the structures for a bridge are beginning
to appear .
opening and implementing of fundamental roads needed for
national security/, economic development and the infra-
structure essential to the process of settlement, agricul-
tur-al production, and the betterment of health. Included
within thiis plan are the highw.ays BR-174, N~anaus-Boa Vista-
Vrenezuelan border (in the direction of Santa Helena), with
a branch to Lethem in the Repuiblic of Guyana and BR-319,
P3r~to velho-IMa-naus, which will integrate by highw~ay the State
of Ama~lZonas and the Terr-itolry of Roiraima to the rest of the
nation. (See Figure 5).
From the point of .iew of Presidential Decreei f)umber
61,599, highw~ay DR-174, w.hich will exrtend a distance of 582
miles (9~70 k~ilometers), has as its objectiv.e the more inten-
si.e exploitation of an extensive area of the extreme northern
section of the Br-azilian frontier. It is thus hoped that th-e
pezop~le~ of this region wJill bring about a permanent and con-
tinuous interchange with other areas of B~razil. As a part of
the international plan, the highwJay system described here
will form a section of the alter-nativ.e link of the Pan Ameri-
can Highw.ay System, connecting the Brazilian highway network
to bordering nations, initially with \.enezuela and ultimately
wJith the Republic of Guyana, the latter being called BR-401
(Boa Vista-Republic of Guyana), with a length of 84 miles
The actual implementation of these road projects is
well under ~ay, but construction moves ahead haltingly,. That
portion of the Mlanaus-Boa Vista-Santa Helena highw:ay (BR,-174)
which is to pass from Boa Vista northward to venezuela is
being constructed by a unit of the Brazilian A~rmy, a unit
created specifically ; for this particular undertaking. From~
Boa Vista to Manaus the project is headed by the Department
of National Highw~ays (DNER). The riv.alr; existing betw.een
these two organizations leaves the Braziliain Armyl unit wijth
decided advantages. The physical region thrLough wh~ich, theI
Army. must build a road is, on the one hand, mostly. open~
savanna. Much of the surface material is laterlite and
nearby there are igneous inselbeirgs offering good .;urfaicing
road metals. DNER, on the other hand, must build its portion
of th~e road through some 400 miles of dense tropical rain-
forest. (Figure 9) Less is known about this region and
materials suitable for surfacing the road must be brought in
from distant areas.
Advanced planning for the road has been minimal. Owing
to a paucity; of accurate accounts of terrain, surface
materials, stream drainage, and human settlement, financial
budgeting for this project has been most difficult. The
first phase of constructing highways through the rainforest
has been first to study photos from aereal traverses. Next,
The Selva of Amazonas and Roraima
It is through this terrain that BR-174 must pass be-
tween Manaus and Caracarai.
a small rigit~-of-w~ay is cut through the forest to permit
survey;ors to determine the topography~. The right-of-lay y is
thecn widened and thie surface is pushed into a road wiith the
aid OE huge bulldozers and earthmovers. After the surfacing
of the~ road w~ith~ crushed rock~, the final stages of construc-
tion atre the building of bridges or the installation of
ferries. Funds are requested from the federal government
onl; ?.hen it b~comies evident how much~ money will be required
to conpiece ea~ch, regmenlt of the road.
Since the beginning of the actual construction of this
road, rBrazll has been governed by its military~ establishment.
Thus i~t fol.low~s that: in the competition for road-buildingy
funds the federal government teldss to favor thie military~ unit
responsible for thle segment of road stretching from Boa Vista
to B~razil's border w~ith v~enezuela. Th~is has, in fact, proved
to bec so an~d this section should be completed by the end of
1970. Thie only major obstacle is that of bridging the Rio
Uraricoera, a project not yet begun. Partly because of the
difficult in procuring funds and partly because of difficult
terrain and hostile Indians, the section of road being con-
structed by; DN~ER is at least three to four years from comple-
The secondary roads servicing the cattle area of Roraimna
are passable for only part of the year, that being the dry
season. These "roads" range from roads comparable to that
stretch which passes between Boa Vista and Caracarai to what
in the United States would be labelled "'trails" or "'wheel
ruts." Even in the dryr season travel on these secondary
roads becomes difficult because of passage through small
streams and low, swampy depressions.
It stands undisputed that it was the airplane which
brought Boa Vista closer to the city of Manaus, the former's
nearest Brazilian urban neighbor. From a jaunt of several
days of trav~elling by river, the airplane has reduced the
travel time between Boa Vista and M~anaus to 100 minutes.
From as early as 1945, Boa Vista has been served by
regularly scheduled flights out of M~anaus byl the company
Servigos Aereos Cruzeiro do Sul, Ltda. In December 1953
Cruzeiro do Sul inaugurated an international service to
Georgetown, Guyana from Manaus, with a stop at Boa Vista.
Today the original DC-3's hav~e been replaced by a Japanese-
produced turbo-prop passenger liner having a far greater
passenger and cargo capacity than its predecessor. Boa Vista
now receives flights three times w~eeklyr from Mlanaus, two of
which go on to Georgetow.n and return the same day.
In addition, the Brazilian Air Force constantly flies
into Boa Vista, bringing troops and supplies to the army
battalion stationed there, as well as vehicles and materials
destined for both military and civilian use. It is in truth
the Air Force which is responsible for the operation of the
airport at Boa Vista.
The airport itself is small but. modern. The runway, ic
paved, kept in excellent condition, and, since it measures
6,500 feet in length (2,000 meters) by 131 feet in wlidth
(40 meters), is capable of receiving most jet aircraft,
even though it is not on the jet run at the present timie.
Moreover, this particular airport serves as a base for
several single-engined aircraft used by ranchers, diamond deal-
ers, and missionaries and the old amphibious twin-engined
Catalinas, which are used byl the Brazilian Air Force to dis-
tribute troops and supplies to border areas, to deliver mail,
to carry essentials to Indian tribes, and in general to
serve the public.
As part of a chain reaction, Mlanaus has also greatly
increased its contacts with the highly developed regions,
such as Rio de Janeiro, Brasilia, and Sio Paulo to the south.
The result has been that such things as newspapers and maga-
zines now reach Boa Vista from as far away as Sao Paulo and
even P~rto Alegre within two days.
With respect to human travel time, spread of information,
and shipment of high-value/lowJ-bulk items, the advent of the
airplane has obviously been a boon to Roraima. Yet in terms
of development and the free flow of goods and services to
and from the region, air travel has not proved to be the
definite and final answer.
To the great majority of people living in Roraima Terri-
tory the equivalent of $25.30 U.S. necessary for a round-trip
to Manaus by air is beyond their means. Thus, M~anaus remains
for them several day~s tral.el by uncomfortable river navilga-
tion. The cost of transporting essential goods such as rolls
of barbed wire, heavyl machinery, consumer goods in bulk; quan-
tities, animal feed supplements, and others by air continues;
to be prohibitive. Finally, the practice of bringing in much
needed fuels such as bottled gas and gasoline by air Is: for--
bidden. It stands, then, that while air transportation has
made great strides in bringing Roraima Territory~ out of iso-
lation, other lines of access must be improved or implemented
before this remote area of Brazil's N~orth can effectively join
Common icalt ions
The Federal Territory of Roraima does possess a system
of communications but its ov.5stence is somewhat tenuous.
Table 3 sh~ow;s the r-adio-tologyraphy~ netw~orkl and its state of
dev~elopmenlt in the y1ear1 19541.
Sin7ce 19J54 much~ ofT thc com~rmuni.cation system has ceased
to operate or has failca~ to Iincrease in eff~iciency~ of services.
Any' citizen of Boa V!ista will. readCily profess that the mail
services leaves much to ber desired.. Foir eight or nine months
of any given ycar- thie a~nnoulnCi' men1 bioard~ i~n the Manlaus Post
Office will1 ha;e, after the: Iicti;ng of surf ace mnail to B~oa
Vista, the w~ord parade e, pu) easetelvlo h
Rio Branco is too low for con.tinUuou nav.~g~igaton
Mluch of thle radio equipmentt~ ai~ainaiined by tle government
has become obsolete or Is nc.L ill work~ing order. Since
February 1968, there has beecn no: tele~ra~ph service out of
Boa Vista. When the telegraphp is functioning as intended
there should be contact: between Boa V~ista and Mlanaus. One
of the more dependable means of contact is that of an amateur
radio operator (ham) w~ho rela~ys messages not only to M~anaus,
but all over the world.
Taken at any level of the time-space continuum Roraima
and Boa Vista are inadequately prepared for the second half
Regions Operating Installed Projected
Radio-Telegraph Network~ for Roraima Territor;,
L ow;e r
Rio de Jan.
Baca do Rio
SIerra da Lua
C. Bras de
Ant~nio Teixeira Guerra, Estudo Geografio do Terri-
tdrio do Rio Branco, Biblioteca Geogra'fica Brasileira,
Publicagio No. 13, da Serie A "L~i\ros," Instituto
Br-asileiro de Geografia e Estatistica, Conselho Na-
cional de Geografia, Rio de Janeiro. 1957. p. 240.
The ultimate source w~as the Five-Year Plan presented
to the Superintendency~ of the Economic i'alorization
Plan of Amazinia by~ Dr. Valerio C. de Ilagalhies in 1954.
of thle tweintieth century. N~one of the present channels of
communication suffices in its present form to connect the
torritoryJ with th~e rest of Brazil or its foreign neighbors.
Inev.itably, in dealing w:ith the shortcomings of the ecoiinoic
infEr as t ruc tur e, increased productivity, and inno.ativre tech-
niquecs the author must rev~ert again and again to this there.
Nothngin fact, is more central to his purpose than h~igh-
lightinng each~ facet of accessibility, or rather the lack~ of
it, in such a promising "El Dorado."
The governor of the State of Amazonas during the year
1905 was the Honorabrle Dr. Anii Cntnin ei
Dr. Arntonio Constantino H3eri, "Message," read before
the Congress of A4mazonas at the opening of the third regular
session of the Fifth Legislature. Manaus, Amazonas, 1906,
3AugushE P lane L maztonie (Paris, 1903 ), p. 192.
4Raimundo Pereira da Silva, "Communicacies de Manaus
com o Alto Rio Branco," Revista do Clube de Engenhario,
October 1938, p. 2400.
5These excruciating circumstances have plagued the in-
habitants of FRoraima continually from the beginning of
settlement in the 1700's. See, for example, a petition to
the government given by the Chamber of Commerce of the State
of Amazon~as in 1.940, expressing itself as follows:
Another problem to solve in the area of beef
cattle activities in the state is the transport
of cattle for slaughter from the distant
ranches of the Alto Rio Branco to the city of
Mlanaus where the majority of the consumption
Despite the expensi.e and unsuccessful
attempts to construct a road, without appro-
priate planning, which claimed a heavy
sacrifice of federal lands, the difficulty
of transport, in the present situation, is
one of thie'major obstacles to the compensating
utilization of the large herds which exist
there. Diverse formulae and suggestions hav.e
been proposed byI technicians and studious
types, some favoring the construction of a
railroad and others favoring the planning
of a highw~ay. All, how~ever, have not been
accompanied by series of meticulous studies
which require further understanding. With-
out a doubt, the solution to the problem
of shipping cattle from the Alto Rio Branco
to Manaus lies in one of the-se tw~o w~ays or
a combination of both, wlithh the help,
furthermore, of river navigation. It seems
the opportune time for us to ask the govern-
ment of Your Excellency to study this
important question, from its several angles,
in order to decide upon the best solution.
In "Problemas da Amazonia." Petition from the Chamber of
Commerce of Amazonas (la naus Ama zonas) 1940, p. 63.
6While the author- was residing in Boa V'ista, in the
month of March (1969), he witnessed a resultant shortage of
sugar, coffee, pow..dered milk, k~er-osene, gasoline, bottled gas,
and beer. SevJeral of the author's trips to ranches in the
interior had to be temporarily postponed for lack of gasoline.
Indeed, it is most difficult to carryl on economically~ and
admin-istrativelyr with the occurrence of major interruptions
caused by material shortages.
7Obviously, and compar-ed w~ith Boa Vista, ITanaus is
indeed a major center (population 244,000 in 1967), buc: it
also suffers from isolation. Located some 1,000 miles up-
stream from the Atlantic Ocean and Beleom, its nearest urban
neighbor in terms of accessibility, Manaus is the commercial
center and capital for the State of Amazonas (898,000), with
a population density~ for the entire state of 0.02 inhabitant
per square mile. For a detailed discussion of the subject,
see: "The Functional Relationship of M'anaus to the Amazon
Basin." Unpublished dissertation (Univ.ersity of Florida,
1969) by Jerry R. W~illiams.
SThe late deputy Antinio Mlartins, in a speech delivered
before the Chamber of Deputies on October 7, 1947, expressed
the seasonal difficulties of navigation on the Rio Branco and
the resultant human hardships as follows: "It should be
pointed out that the lower Rio Branco, in its course of 400
kilometers, starting from its mouth, is only navigable with-
out great hindrance from Mlay to August up to the village of
Caracarai, on the right bank. In the remaining eight months
of the year when the river's volume of water is considerably
lower, navigation is permitted only to small engines, hauling
barges with reduced tonnage and having a draught of not more
than two and one-half feet, devoid of the most. fundamental
elements of comfort and hygienee"
9The lower Rio Branco is that portion of the rivecr's
course from its juncture with the Rio Negro northward to
Caracarai. The upper Rio Branco is that part of the river
lying north of Caracarai, passing Boa Vista to the point
where the Rio Uraricoera and the Rio Tacutu merge to become
the Rio Branco.
10Reference is made in Clovis Nova da Costa (O Vale do
Rio Branco to the length of time that better access to the
Alto Rio Branco has been considered. He states:
Whoevrer writes the history of the Rio
Branco will ha.e to reserve a special.
chapter to the Boa Vista-caracarai road,
whose attempted construction dates back
to the se'.'enteenth centurl when the
problems of transport stimulated the
imagination of the pioneers, looking for
the solution to the difficulties of access
and the fl.ow of production. The funda-
mental stage of the daring project of con-
necting Mlanaus and Boa VJista by road, only
after two hundred years of expectation, is
that which opened the expected w~ay of com-
munication, within a day~'s journey~, whose
principal facts we layr aside here (sic),
as the contribution to the history of the
major undertaking of the sort now: accom-
plished in the basin of the Rio Branco."
11The Brazilian planners have had a complete change of
rationalization with respect to highways construction.
Witness the thinking of Raimundo Pereira da Silva (see page 43)
as being typical of previous thinking. Prior to building
a road through an area there had to exist already a popula-
tion large enough to generate economic production capable
of justifying the allocation of capital necessary. to carry
out such an endeavor. The result of a turnabout in this
policy is v/iv~idly' brought to light in Dav~id Crease's article
on the new road connecting Belem andBasla (Jong
the Two Brazils," Geographiical M~agazine, XXL~VII I, No. 59J
(July' 1964), 184-197). The town of Imperatriz, some 500
miles up the Rio Tocantins in the State of Maranhdo, northern
B~razil, had a population of 6,000, one truck, and no road to
the exterior in 1960. By 1964 the Belemo-Brasilia highway con-
nected Imperatriz to the exterior both to the north and to
the southi. The 6,000 people there have now become mnore than
20,000. The rice crop has ju~mped in a few~ years from 8,000
sacks to 500,0001 The monthly count of vehicles ini transit
through Imnperatriz nowl exceeds 1,000.
1For example, the engineer, M1. Pacheco de Caralho, upon
studying the transportation problems of the territoryr, ex-
pressed his thoug~htsas follows:
Thle solution for the normal transport of
merchandise...is ev~idently the construction~
of a road connecting Caracarai to Boa Vista
which would allow transport by truck from
Carcarari to Boa Vrista in four years. Seeing
that the land is flat or gently rolling, it
allows the construction of a road hav.ing the
The construction of this road, whiich is of
vital federal importance since it is part of
BR-17 [no BR1 brings to mind the
colonization of the forest which runs alongside
the major part of its extension.
This was a note on the connection by highwlay between Boa Vista
and Caracarai as a stretch of BR-17 (unpublished) draw.n from
G~uerra, p. 234~.
In another statement by Dr. Pacheco de Carvalho, who was
also ex-director of the Dil.ision of Construction and Conservra-
tion of th!e Departmrent of N~ational Highways, he stated: "Att
this time the most necessary task is the construction of a
highway connection from Boa Vista to Caracarai, which is the
river port situated at the end of the freely navigable stretch
of the Rio Branco bearing vessels of up to three feet of
draught the ~entire dry season." "Plano Regional de Recuperac~io
e Colonizacao da Amazania," Boletim Geogrifico, X, No. 109
Captain Clo'vis Nova da Costa, former governor of Roraima,
referring to the importance of that road, has said:
The Boa Vista-Caracarai road constitutes
the axis of a system that will guarantee
the circulation w~ith the economy of the
bordering countries. From there, i.e.,
Boa Vista, there will be the construction
of another road, parting fr-om the capital,
on the left bank of the Rio Branco, going
to Concega~o do IMau, on the border with
British Guiana. Later, with a branch~ in
the direction of Santa Helena, r'enezuela,
the economies of the basins of the Amazon,
Demer-ara and Orinoco will become joined.
O Vrale do Rio Branco, Suas Realidades e Persp~ectivas (Ri
de Janeiro, 1949), p. 131.
13Manuel Pacheco de Car'.alho, "Plano Regional de
Recuperacao e Colonizacao da Amzni"Botm egaco
X;, N~o. 109 (1952), 416.
14Amazonas Brasil, and Ray A. C". Lins, "Projeto
Agropecuario, Fazondas Brasil, S.A:., Territorio Federal de
Roraim~a," Nlanaus, 1968, pp 12-13. (li meog ra phed )
THE PRC~SENJCE OF~ BEEF CATTLE 10~ ROR~ll-Us
Given th~e cultural background of the people andl the
nature of the physical milieu, it is not at all surprising
to find that exten~sivee livestock rais~ing has a~lway~:s been the
predominate economic activity on t.he sal.'ann~ias of Rtorailuia.
Latin Amrericanists have long recogrniZ~d th~at in7 tlhO Luso-
Hispanic colonizationi and de*.relopment of t!,0 j:n.Cricas cattle
raising has played ani important role.
It is a well documented fact th~at the entire system of
large-scale livestock raising was imported to the N~ew world
from the Iberian Peninsula.1 Cattle have been raised almost
everywlhere in Europe and America, but raising cattle and
cattle ranching are not the same thing! Bi shka provides a
definition of the latter by stating that ranching "implies
the ranging of cattle in considerable numbers over extensive
grazing grounds for the primary purpose of large-scale pro-
duction of beef and hides.2
Historically, as an extensive productive activity,
ranching required large tracts of grazing land and more
specialized techniques than those required for the care of a
few dairy cows or small herds which served as an adjunct to
agriculture. With the possible exception of the Hungarian
Plain and western portions of the British Isles,3 mnedieval
Iberia appears to hav.e been the only part, as it w~as un-
doubtedly the most important part, of medieval Europe to ad-
vance to this third level of cattle raising.4 Although the
precise circumstances are still obscure, the available
charters (cartas) and statute-law~s (fueiros) provide evidence
that a genuine ranch cattle industry evolved in the Iberian
Peninsula in the late elev/enth and twnelfth centuries, under
Alfonso VI and Alfonso VII of Leon-Castile. The birthplace
of this activity was that portion of the subhumid or arid
interior tableland of the Me~seta Central lying between the
middle course of the Duero (or Douro) River and the mountains
of Gata, Gredos and Guadarrama.
From this area of origin, cattle ranching, on an ever
expanding scale, spread southward along with the reconquista
colonization. By the late twelfth century~ it had moved into
the broad pasturelands of New Castile, Extremadura, both in
Spain, and Alentejo, in Portugal. Alentejo, by all available
accounts, is the cradle of the Portuguese ranching system
which was later extended into Algarve, the Archpelago of
Iladeirn and the B~razilian sertao. On the southern half of
thec mcsetai, primarily to the west of a line passing through
central new Castile, Castilian and Portuguese military orders,
noble,~- s and tow.nsmec n grazed thousands of cattle, although in
b~oth~ numbers and economic importance these w~ere less signifi-
cant thian the great sheep flocks of the blesta and other
owners. But this situation wias reversed after 1250, with~
Ferdlinanld II~i's reconquest of Andalusia, when royal regarti-
mienton~c assigned to cattlemen r~ather than sheep raisers the
bulk: of the extensive lands in the valley~ of the Guadalquivir
Rivr. s a result, theC Anldalusian plain (including part of
the PorLtuglueses Alentejo district) became in the latter Mliddle
Agejs the- on-e region of the Penin-sula, and perhaps of al~l
Europe, where pastoral life, and indeed agricultural life
in generall, was dominated by a thriving, highly~ organized
cattlel-ranchiing economy. The fact that, many of the early
colonists of the Canaries and the islands of Nladeira, and
later. the N~ew world, came from these southern Iberian cattle
kingdoms, which were at their heights in the fifteenth and
early sixteenth centuries, provides one significant clue
to the~ promotion of cattle over sheep ranching in the
Th~e types of cattle found in early Castilian and
Portuguese ranching were most likely the results of various
degrees of crossing between lighter-colored European types of
the all-purpose cow~ and theT w.ildl or semi-wild, black, dark
r-ed and dark birown descendacnts of a cattle strain unique to
Iberia, Bos taurus ibericus, the ancestor of the modern
fighting bull. These tw:o str:ainss mingled on the meseta and,
as the reconquista frontier:'~ wasush~ed southward, these tw;o
razas produced a v~ery-1 hardly Ilybri~d s~ock~, displaying an
;Imazing variety~ of color anld color comb~-inations from cream-s,
yellowis and gray~ish~~~-b:rons to deep b~rownis, reds and blacks,
This wias a stock charact~erize- d by:, markredly feral instincts
and of ten complete w~ildness. Su]ch cattle were :aluable
chief~ly for their tough hicdes andi str~iingy b-eef. These
cattle then, unlsuited for aniry or draft purposes, comnpelled
those w;ho pursued this producti.e aciiltivit in c'astile and
Portugal to abandon their "'cozy. little rolwpastures" for the
open range, to take to thr. hors~e for her-ding, to perfect
systematic methods of long-distance grazing (or eien traJns-
humance), periodical round-ups, branding, overland drives,
etc.--in short, to invent cattle ranching!
Across the sea in the ~ew~ WJorld, the Spanish found, i
what is now: northern M~exico and thie United States' Southw;est,
an area similar in many, respects to thie Ilese-ta C'entral, i.e.,
a semi-arid grassland poorly' suited to the growing of crops.
Being accustomed to the dry plains of Castile, the Spaniards
wer-e quick to realize the geographical limitations and
adapted thelmselves to the conditions of climate and topography
which~ theyc; enco~unteredd in the Americas. Liv.estock raising
presented a means for utilizing the available natural re-
sources to provide a liv.elihood and food supply~ for those
settlers w:illing to enidure th~e hardships of pioneer life.
In North Amiier'ica the Spanish Crown issued land grants
which co~rrsp-ondedd to the requ~iremenits of the cattle industry.
Far bJe~ttr tha~n the A~ngjlo-Amnerican~s, the Spaniards recognized
the need for largo tracts of land for grazing and pasture.6
w~alter~ Prescottl Hobc~b pointed out in~ his classic study~ of the
Great Plaj~in that the An3l~o:
thought in terms of the possibilities of
utilization and production. To him a
hundred acros was a sufficienc,.....No
pL~roirtion had been made in lawI for the
ranchmanc~ wh~o was using or trying to use
the semi-arid land.7
On the other hand, thle Spanish authorities, with ex-
per-ience in Elemi-a~rid environments, took into account the
climate and proposed land use in determining the size of the
land grants. The grants reflected either agricultural or
pastoral use; land suitable for irrigation and farming was
divided into small units, whlile dry land and pasture were
allowed in larger portions.8
The rise of ranching in Spanish Texas, as investigated
by' S. L. yrel'rs, wacs quite sim~ila7r to the course of the intro-
duction of ranching into Inoraila.' In northern M~exico the
Spanish system of landholding led to the establishment of
the haiciendia, a huge landed estate inl the possession of one
family that approximlated the feudal holding of Europe-. The
hacienda, not Infrequentlly containing more thian 300,0003
acres, became a self-suffic-ient oconomiic un~it which~ included
farming, mining, and a niumbr~h of commner-cial enterprises in
addition to stoch:raising.; Som~;e of th~e largest estat~es were
communitiesi in themsllves,'is withi field-s, filck~s anld herds,
wooded areas, flour: mills, forgos~c and wo~rk~shlops. Each had-
its church and so-.eeral hun~dre:1 o- e.len thousands of in-
habitants, w~ith v.illages foi the Indian~s and~r w.oreri~S, as
well as the casa de back~ndai o,.cup-ied byl thle haicendalo~ and
In Texas, ho-wev.er, thc Spran~ish '.e're forced to adapt their
methods of stockraisi~ng and lan~d utilization to frontier con-
ditions. Common pasturet anrd stockraising grants pro\.ided
grazing for large her'ds of li-.estock on the grasslands and
dry' plains, while the rivecr I.alley~s an~d well-w~atered areas
were reserved for farming and aIgriculture.10'
In addition to the individual land grants, collectivJe
holdings were available. For many centuries, the towns of
Castile and Portugal had been landholding bodies possessing
more or less extensive territories belonging to the towrns
and admiinistered byl their officials. Such lands included
proprios, rented out for cu~ltivation; e~jiidos, reserved for
a v.ariety of purposes such as dump grounds and slaughtering
pens and not used for buildings or cultivation; and debosas
or baldios, common pasturage on vacant lands in the open
range.1 Stockmen pastured animals on baldio lands under
license from the .icer-oy or governor until such~ time as thle
lands w~ere granted to individuals. In Tex>as ag~r~eements wecre
occasionallyr worked out between mi.-sioniaries and townspeople
as to the extent: of grazing p~rivileges on the dehosa betweeni
the Guadalupe and San Antonio rivers.12
During the period of early Spanish~ settlement, desp~it~e
the pleas of the Church, the missions in Tex:as were not
opened until Spanish borders w~ere menaced byl the French.
Then, and only then, w~ere the missionaries allowed to begin
their wcork and garrisons w~ere constructed and ranches w~ere
established to protect and support them.
within this context it is important to note the develop-
ment of ranching as an appurtenance to the garrisons,1 those
tiny outposts of the Spanish Empire charged w~ith the responsi-
bility of protecting the missions and w~ardin'g off the rush of
intruders--French, Indian, and English--who desired to prey
upon Spanish possessions. As one of the most isolated of
Spain's frontier domains, Texas was far removed from main
communication and supply routes, and scarcity of food,
clothing and equipment plagued presidial commanders and pro-
vincial governors throughout the colonial period. Ranching
offered a partial answer to these problems not only by pro-
viding meat, but also oxen for plowing, mules for hauling
provisions and supplies, and horses for mounting and main-
taining the garrison troops.14
Contrast the introduction of ranching into the far-
flung Texas region of the Spanish Empire with the rise of
this same activity in wJhat is now Brazil's Northeast.15 The
history of the growth of cattle ranching in the Northeast,
beginning with the arrival of the first heads of cattle in
Salvador from the island of Madeira, is intricatrely tied to
the settlement of that region's interior. It was the breed-
ing and grazing of cattle, more than any other activity,
that promoted the advance and settlement of the Portuguese
into the Brazilian sertio. The entradas and the bandeiras
only exploited the region for Indians which they would sell
into slavePry along the coast, or they scoured the area look-
ing for mineral wealth and wer'e merely temporary dwellers.
The objective of each adventurer and miner was to make
his fortune as quickly as possible and return to the city so
that any settlement by these types of individuals was, at
best, of a temporary n~tur~e. On the other hand, the men wlho
were founding fazendas for cattle raising moved into the
interior continuously and in a "one-way"" fashion.
The lands w~ere obtained from th~e Portuguese Crow~n by
means of grants which wrere known as cartas do sesmai. 16
The actual amount granted wac~s called a sesmaria. A4t the
samie time thiat: the nobles wer.e dealing w.ith the authorities
in the sitting rooms of the Viceroy in Salv;ador, or the
Governor of Pernambtuco in Olinda, seeking final agreement on
the lands whose occupation theyr had been promised, the vagueiros
were receiving the word from those same nobles to push on with
land settlement as far as possible...''into all of th~e interior
of Bahia, Pernambuco, Alagoas, Sergipe, Rio Grande do iNorte,
Ceara, Piaui, large parts of Malranhao and Mlinas Gerais."l
With the passage of time the nobles received more cartas
de sesmaria and took possession of their newly acquired lands,
therebyl creating extensive latifundios. One example of the
benefactor of sesmrarias is the Grupo Garcia D'Avila, today
the wealthy family of the casa de Torre, in Salvador. This
family w:as so successful in accumulating lands that their
fazendas extended from coastal Bahia on into Piaui, Pernambuco
and Ceara. Along the banks of the Rio Sio Francisco, for
example, this family owned some 320 miles of land. They
owned territories larger than most European countries.l
Land distribution through crtas de sesmaria w~as dis-
criminatory; and unecqual. The r-ight of ownership was granted,
with grim consistency, to those who knew~ nothing about the
land. The process of granting the land, howe.er, w..as done
in a somewhat orderlyl~ mannier. The Portuguese Cr-own declared
that a fazenda would consist of three leagues' distance
(approximat~ely 12 miles) along a gi.en wa~tercourse. Th.er-e
would be one leaguec of landc left between fazendas so that
there w~ouldr' be no quel~sti~on about the property lines of a
given piece of land.l
Population con~tinued' to inc;-ease along the coast and
the resultant demand for me-at :;1 too large for th~e existing
ranches of th-e interiocr to fil~l. It was the ev.ergroowing
demand for meat tha- garve rise to thie rapid growth of the
fazendas in the inter-ior. Th-is numierrical increase w.as greatly
facilitated byr the extremely1 simple manner in which a fazenda
w.as es tabl ished. l that was:~ necessary~ was to build a
crude straw shelter for the .agueiro and hiis helpers, or
his family'; construct a .coryl crudre cor-ral where theyl would
place the cattle for the pur-pose of former os cascos, a
period of quarantine to tame the unruly, cattle for easier
handling on the range; after this the cattle wJould be turned
out into the w~ild caatinga, leaving them to feed themselves
and reproduce with almost no human care. In this way a
fazonda was established and a stretch of land about 12 miles
along a watercourse was "settled."
In a very narrow sense of the word, fazendeiro is
sylnonymous w.ith~ the owner of a fazenda. But in the early
times the landowners livecd on th~e coast andl only received
new.s and pLrofits occasionally from their fazendas. Thus,
for all practical purposes, it can be stated that the first
fazenidoiros that established themselves in th-e Northeast
we.~re the wiild ulaqu~eiros, pioneers who, with the cattle that
they received from the quarteacao, leased lands and began
ranch~ing for themselves.
Trhe rent from the sitios, leased lands generally a
squa~re league each (43.57 square kilometers or 16.8 square
miles, or 10,762 acres), was about 10 mil reis per year.
Th~is was during a period when one steer was worth- about tw~o
mril reiss so that payment corresponded to the value of five
steers per year.
In retrospect, the vaqueiro brought the cattle into a
given area, took on all the risks and performed all the
neces~saryl tasks of operating a fazenda, and finally saw co
the settlement of the leased area for which he had worked
10 to 15 years to obtain, often leasing for a high price
the very land that he himself had "conquered."
On a fazenda, initiated in this manner, the workers' lives
differed from those who lived on the large landed estates.
The former vlaquciro, upon establishing his ranch, was not
able to pay anythlin-g to the workers, who most often turned
out to be his children, who carried out the tasks of the
fabricas (~el~pers to the vaqueiro).
Trhe I;stem of qu~alrteac47~ o gve rise to a new class of
people in the sortho. Considering that there had been a
chiangc: onilyj in the name, to fazendeiro, from :aqueiro, to
en t r l: ep er uer fromn laborer,. but w~ith~out the possibil~ity of
pay'ingc~ :.or-kers, and the specific chlores were still the same,
thie p-oferr-ed namle f~or these fazendeir-os was fazendeiros de
yagueiro aiit~nomro.- It: should also be added that the
fazendeciirro de~- vaglueiro aut~jnomeo did niot care for his own
cattle any better than hie had cared for those belonging to
his f~ormeor Pa~traO.
whieni fcld crops began to appear on the fazendas it
became necerssary to build fences. The faizendeiro nowr
wocrriedl abouti his crops so that the cattle wer-e left: to
roam mnore unattended th~an ever, but now fenced in on one
part icul~ar fazenda. The use of fences brought an end to the
practice of apjart~apcio (rounding up and sorting out cattle
that had grlazed in common but w~ere owned by different
fazendeiros) because the cattle no longer bred and grazed on
As wars true throughout the A~mericas in the history of
ranching, the appearance of fences brought on a miore sophis-
ticated form of cattle ranching and gav.e new~ importance to
the fazendeiro de vaqruairo aut~nomo, w~ho tried to purchase
the lands he had been leasing. If the fazendeiiro lived in
a nearby~ city, he superv~ised a system of for-eman!;hip in ad3-
ministering the affairs of the fazenda, and he himself
tended to business in the city, such as acquiring breeding
stock, seeds and equipment, and selling thact which his
fazenda had produced.
W~hen the change came to the fazendas ow.ing to the ad.ent
of fences, the v~aquciro capstazi the foremnan of the city-
dwell1ing faztender ro, began receiving his salary in cash.
This occurred because the fazenda began a phase of more
sophisticated lil.estock; ranching, utilizing a system wh:ereby
it was possible to improl.e the quality of the herd by cross-
ing the cows with newly purchased breeding bulls. When the
quality of the cattle impro'.ed, the fazendeiro was no longer
interested in sharing his stock with his foreman by means of
the quarteaqio. In this way the -aqueiro experienced an
increase in importance of role on the fazenda but in economic
terms he was demoted because financially his real salary
turned out to be worth less, thus greatly reducing the possi-
bility of his el.er owning his own fazenda.
The origins of cattle ranching on the remote savannas
of Ror-aima have been obscured by~ time. At best, historical
accounts and eividen~ce of this are ex~tremcly sketchy and tend
to conflict with~ one another. HoIweverI while some of the
specific information is lackling, the general purpose and
pattern of establishiiing r:.nching as the primairy economic
activity~ in this regjio~n can be~ clearly' di~scerned. The early
situation in Roi~raimal beairs a sci-ron resemblance to wJhat had
occurred a century~ earlier in ,rpanish- Texas. The areas of
the upperu Rio Ne]glro, the~ Solimoes, anrd what is now\ R~oraima
werce regJions whi~erec the: colonial interests of Spain, Port-ugal
and the Dutch hadi cla~shed duringr the sevenentcnth and
There are fleeting rcefeences to European penetration
into the e:tireme nrorth of Br,~aziil in the eighteenth century,1
but the intentions o~f the5 earlly Intruders were not those of
establishirrg permanent settlement. For- example, Lawrence
Belfort, a wcealthy; Irishi entr-eproneur who had settled in
Mlaranh o, tra\elled to the interior of Amazinia seeking
slaves as labor for his various agricultural enterprises.
On one particular occasion, in 1740, he reached the valley
of the Rio Branco.2 Nicholas Hortsman, a German represent-
ing the interests of thle Dutch on the northeast coast of
South America, contacted the Indians of the R~io Negro in the
year 1741 by passing diown the Rio Branco system,24 advancing
from the Essequibo Region, which, at that time, w~as Dutch
territory anid later became a British possession.
Thle Treaty of IMadrid~ in 1750 w~as arranged to settle a
dispute b~etweren the Spanishi and Portuguesc Crowns which
centered on thle Plata region, but the regions of thle Or-inoco
and Ne~gro drivers far to the north w~ere not insensitive to
this highly tentarlivel agreement. As a result of thiis treaty,
Fr-anciso :av~ier de Me~ndonga furtado, Go'.ernor of the~ Cap-
t~aincy of Grand Paira, w~as appointed by the Crown~i as the prin-
cinal comm~iissioner- for- boundaries of the Ndorth in 1752.5
Hle was~ charged w.ith thie du~ty of meeting a Spanish expeditionary
Leam~ to delimit the boundaries in Amaz~nia between the Spanish
and portuguese possessions.26 The Treaty of M~adrid was de-
claredi nuill and v~oid in 1761 and tw~o yeare prior to t-hat
flendonsa Furtedoo had returned to Portugal, ne'.er having ful-
filled his boundary-delimiting mission.
The Treaty of Santo Ildefenso, in 1777, produced another
commission to fix the boundaries between the Spanish and
Portuguese possessions in the New World. The Portuguese
Crow~n appointed Joao Pereira Caldas, Gov~ernor of Grand Para,
to head its commission. Caldas took up residence in Barcelos,
the seat of the Captaincy of Sao Jose do Rio Negro, and
worked vligorously for a number of years there attempting to
promote'economic activity and permanent settlement which
would aid the Portuguese in their claims for extended bounda-
ries. Upon suffering a stroke, Caldas was replaced by his
former assistant in Grand Para, Mlanoel da Gama Lobo d'Almnda,
Governor of the Captaincy of Sayo Jose do Rio Negro.
Lobo d'Almada wias a most ambitious man. He explored
the upper Rio Negro and discovered major tributaries connect-
ing that river with the rivers Ja ura and Solim~jes. He
brought sparsely scattered Indians into settlements, he
developed the agricultural industry~ of growing mandioca for
the production of farinha, he acti;ated indigo grow~ing~ w.ith
an eye to shipping it to the European market, and he en-
couraged settlement in the savannas of the Rio Branco.2
As early as 1775, Spanish forces invaded the savanna
area in Roraima and began to establish a garrison to fortify~
the Rio Uraricoera.29 The Spanish had entered this area
seeking the fabled lake of El Dorado. The Portuguese sent
a force out fromn Barcelos led by Filipe Strum, a German in
the service of Portugal, which drove the Spaniards out that
same year. Besides forcing the Spaniards back, Strum seized
this opportunity to select a site and construct a fort to
defend the Rio Branco savannas from further intrusions by
the Spanish or the Dutch. He selected a site on the left
bank of the Tacutu river where it joins the Uraricoera to