GOLD IN MADAGASCAR
Dr Simon McDonald has purchased Camillo's extensive files and map collection, including a comprehensive data set for Madagascar. E-MAIL LINK Mobile Phone: +61 423 921 780
Very little is known about the potential gold deposits in Madagascar, but the country's estimated gold production is far from negligible. It runs to approximately 3-4 t/y and "could easily be doubled" (Mining Annual Review - 1995, p 149).
This is a large amount considering that all of it presently comes from "artisan" mining - or very
small syndicates - and mining is largely alluvial. Estimates of Madagascar's total gold
production since it began late last century is a highly respectable 70 tonnes.
Madagascar's primary gold deposits are thought to be of mesothermal "lode" quartz-hosted type,
and they are hosted in Proterozoic or, more frequently, Archean terrains. Lode gold deposits of
this type account for nearly 20% of world deposits, and are largely responsible for the vigorous
gold development of Australia, Canada, Brazil and Ghana.
There is no evidence in Madagascar of sediment-hosted gold deposits, pebble-conglomerate
deposits, gold-rich porphyry copper deposits or epithermal gold deposits. Although their presence
cannot be excluded given the current level of knowledge, their occurrence seems unlikely.
Madagascar's geological records are far more difficult to access than is usual with African
countries. This is particularly galling as, until 30 years ago, Madagascar was one of the best
geologically studied countries in Africa, with several thousand first-class records.
Unfortunately, virtually all these records are in French and it is very difficult to gain access
to them (Ph.D theses, obscure records buried in the archives of poorly organised institutions,
etc.). There is no evidence to suggest that this situation will change in the near future.
Like many other countries in Africa, Madagascar flirted for almost two decades with
socialist-inspired, centrally planned economic policies and this ended only very recently with
the demise of Soviet communism.
The country's economy suffered grievously from these policies. A more sober political line was
re-established in 1993 with a return to a form of liberal democracy.
French influence in all Francophone Africa, and Madagascar in particular, continued long after
the country's formal decolonisation in the early '60s. Indeed, it has only been in recent years
that France made clear that her interest in French-speaking Africa was decreasing. This process
started even before the devaluation and floating of currencies in ex-French African countries
(including the Malagasy Franc) in May 1994.
This long-overdue decision greatly boosted indigenous production, particularly agriculture, and
discouraged imports, particularly from France, which were directed essentially to an African
Until 1995, African politicians gave financial support to various French political parties. The
courtesy was reciprocated by very flexible bilateral co-operation agreements, which involved
The three factors mentioned above - stagnant economies, renunciation of socialism, and a lessened
French influence - have created new opportunities in Francophone Africa, particularly in mining.
This point has not been lost on the international community in west Africa, and IMR suggests that
this perception could be extended to Madagascar.
A significant gold producer now and in the past, Madagascar is about the size of Spain (or 41% of
South Africa), and contains prime Archean-Proterozoic terrains beneath its surface. The country
remains one of the best gold exploration targets in Africa.
However, French cultural heritage is well rooted in Madagascar, and perhaps rightly so. Most
Frenchmen had a special affinity with and liking of Madagascar, a fully independent and
functional country long before colonisation.
The grotesque "technicians" from North Korea that arrived in the '70s were received with mild
incredulity by a sophisticated Malagasy society. Although very poor, Madagascar never seriously
knew the cultural soul-searching of, say, Zaire or Ethiopia. The country remains to this day a
unique Malagasy (not African) state, with an overlay of French language, French background and
French culture. Any bullying will be resented. Overseas investors should take warning.
The present report aims to consolidate all available "open file" information on gold in
Madagascar. The information is presented in a modern and, wherever necessary, interpretative way,
and it is directed to major investors. For easy consultation, part of the material has been
presented in computer-generated databases that can easily be down-loaded in the reader's
computer system and edited or increased at will.
Several factors have radically affected Madagascar in the past few years. Among these have been:
- - The collapse of Soviet-type economies in the period 1989-1992.
- - South African renunciation of Apartheid politics, and its opening up to democracy and trade,
particularly with southern African countries.
- - Madagascar's adoption, on August 19, 1992, of a new constitution with a bicameral legislature,
a National Assembly elected by universal suffrage, and a president as constitutional Head of
State elected for a term of five years. The first elected president, Professor A. Zafy, was
inaugurated in 1993.
- - The renunciation by France in May 1994 of its policy to keep the Malagasy currency artificially
pegged to the French franc.
All these circumstances, coupled by a sustained gold price, has created a situation in which gold
exploration in a newly democratic country of great geological potential - such as Madagascar -
suddenly makes sense.
The present Database is intended to fill a void of relevant information about Madagascar. Its
objective is to present in a modern way what is known about gold in Madagascar from open-file
sources. Regrettably, some of these sources are no long available in Madagascar itself.
As the great majority of the potential mining investors are English speaking this is the language
which has been chosen for the Database.
The report is based on modern compilation methods. We have sought to use French reporting customs
in a way that Anglo-Saxon readers will understand. We liberally used computer techniques such a
tabulations, cross-indexing etc to provide the reader with immediately accessible information;
actually we feel that this has been the most valuable side of our work: to present in a
state-of-the-art framework information that can only be laboriously gathered in dusty files in
Europe or Antananarivo.
3. SCOPE OF THE DATABASE
No-one has seriously explored for gold in Madagascar for nearly 50 years. Although gold has been
mined in the country for more than a century, comparatively little is known about its objective
potential. For this reason, Madagascar's gold resources fall - by western standards - into the
"hypothetical" or "speculative" categories, not "proved" or "possible" reserves.
Gold exploration in Madagascar, particularly in the initial stage, would be quite risky, and the
crucial early phase of collating all the existing data would be very time-consuming, particularly
for non-French investors. But these are precisely the investors presently sought after by the
Malagasy authorities, understandably keen to enlarge their base of trading partners.
Madagascar's authorities are well aware of their island's gold potential, which is presently the
country's largest mineral export - albeit still clandestine and poorly regulated. The authorities
are keen to attract major gold explorers, within a legal and fair framework.
4. METHODS USED
The information on Madagascar's gold potential is rather dispersed. The main sources of
information for the present Database have been:
- 1) A computerised search on Madagascar's geology and metallogeny, utilising any CD-ROM geological
database available in 1995;
- 2) Previous reports by IMR staff who had worked in Madagascar;
- 3) A detailed field visit to Madagascar in 1995;
- 4) Verbal communication in Madagascar or France with personnel personally involved in gold
exploration in Madagascar;
- 5) Rapid examination of some satellite images of the country;
- 6) Study of various geological maps of Madagascar. The six sheets at 1:500,000 (1971) and the
three sheets at 1:1,000,000 (1964) have been particularly useful as most major gold occurrences
- 7) Articles and papers from professional journals and publications, not only on Madagascar but on
its geological neighbours, including Tanzania, Mozambique, India, Sri Lanka and Antarctica.
As the senior author of the present report spent two years in Madagascar in the 1970s, and has
made numerous visits since that time, the material above has been supplemented with personal
information and opinions.
In presenting the information care has been taken to distinguish factual information from IMR's
personal opinions and considerations. A clear distinction between "matter of fact" and "matter of
opinion" has been maintained throughout the Database.
The name of 216 recorded Malagasy gold prospects is given in three different databases ordered by
name, gold district and genetic type (Tables 5,6 and 7). The geographic position of at least 130
gold prospects and gold districts is given in illustrative sketches (sketches 7/2 through 7/12).
Madagascar's old French geographical names have now been changed to Malagasy ones (Fort Dauphin =
Tƒlanaro, etc.). This can be slightly confusing, particularly when old documents or maps are
consulted. The new "Malagasy" spelling of localities has been maintained throughout the Database.
In addition, until very recently the geographic co-ordinates in Madagascar were given using the
"Laborde" co-ordinates system (see sketch 5.2.1). "Laborde" co-ordinates can be converted through
a conveniently programmed Geographical Information System in any other modern co-ordinate system.
As technical French (or French/Malagasy) terms used in gold mining are not generally given in
normal English-French dictionaries, a brief glossary of these terms has been included in the
Database (Table 8a-8b). Its purpose is to allow people with even a very rudimentary knowledge of
French, or those using non-specialist interpreters, to be able to communicate on the subjects of
5. GEOGRAPHICAL AND SOCIAL FRAMEWORK
5.1 Physical and social geography - General
The Democratic Republic of Madagascar (Madagascar throughout this report) comprises the Island of
Madagascar, the world's fourth largest island, and several much smaller off-shore islands.
Madagascar lies about 400 km from the African mainland across the Mozambique channel. In reality
Madagascar is not part of Africa proper but a micro-continent of its own, stranded between Africa
and Southern Asia. This is apparent not only by its population but also by its unique fauna and
Madagascar extends 1,600 km from north to south and it is about 500 km wide. It covers an area of
587,000 km2, or slightly more than France and Belgium combined. Geologically, Madagascar is
composed essentially of crystalline rock, which forms the central highlands that rise abruptly
from the wide plains of the west coast. Once again, Madagascar's mineral provinces present
several similarities with S.E. Asian ones, particularly India and Sri Lanka.
It has been said that large islands (U.K., Japan, Cuba, etc) have a tendency to have "different"
behaviour. If that is true, Madagascar is no exception.
Madagascar was the only African country - with the dubious exception of Ethiopia - that was
independent long before colonial times. By the end of last century Madagascar had developed a
truly unified government, supported by a regular standing army, a crude but effective judicial
system, education, public works, administration and the like. The stone palace of the last
Madagascan queen (Ranavalona III) still proudly overlooks Antananarivo.
In a way, Madagascar was more Asian than a typical African country; a distinction that survives
to this day. During the "scramble for Africa", France, excluded from East Africa after the famous
"Fashoda" incident in Sudan, thought Madagascar an attractive prize. Its strategic position in
the Indian Ocean, its perceived richness in mineral resources (and particularly coal - then a
vital commodity for any major fleet) played an important part in this decision.
Madagascar was invaded by France in 1895 and become a French colony one year after. The country
was entirely subdued only 10 years later.
Colonisation followed a typical pattern: appropriation of land by French settlers and companies,
the exploitation of the peasantry through forced labour (although this practice predates French
arrival), the imposition of an import-export economy - essentially based on coffee and vanilla -
with the construction of roads and railroads to serve it. Gold exploitation at the beginning of
the French colonisation had a profound effect on what was up to then a near cashless society,
based on subsistence farming and barter.
Autocratic and alien as any colonial regime invariably is, there were positive aspects of French
rule on the island: hospitals were built, a very effective administration was put in place and a
Malagasy elite was trained to French intellectual standards. It is also fair to say that the
French colonial administration was better than those of Portugal, Germany or Italy on the other
side of the Mozambique channels, and more enlightened than the Belgians in the Congo.
Many Frenchmen took a sincere liking to Madagascar and some never returned to France (among them
H. Besairie, the father of Malagasy geology). Mixed marriages were common. France always referred
to its overseas African colonies as "Territoires d'Afrique et du Madagascar", stressing the
individuality of the latter. A nice and appreciated touch.
Malagasy units fighted with distinction in World War I.
It was during World War II that Madagascan history took a bizarre, and little known, turn. After
France was overrun by Germany in 1940, Hitler toyed with the idea of shipping the several million
undesirable European Jews to Madagascar (Gilbert, 1989 p.323) - a blander sort of "final
solution". A strongly anti-Semitic Vichy government showed an immediate and keen interest in the
project, but neither the French nor the Germans had adequate ships or enough control of the sea
to implement this odd scheme.
Throughout World War II, Madagascar - the "pearl" of the French colonial empire - was staunchly
pro-Axis and pro-Vichy.
After the fall of Singapore, a slightly paranoid British high command saw Madagascar as being the
next stepping-stone of Japan's expanding empire. On May 9, 1942, a British contingent landed in
the port of Diego Suarez (now Antiranana) under heavy French fire. On May 31, Japanese midget
submarines on a suicide mission sank a British merchant ship and damaged a battleship in Diego
Garcia (Gilbert, 1989, p.330).
It took weeks for the British forces - with many casualties - to fight every inch of the way to
reach Antananarivo and put Madagascar under the "Free France" of De Gaulle's rule.
At the end of WW2 a strong pro-independence movement took root in Madagascar. Open hostilities
erupted in 1947. But most of Madagascar's topography is unsuitable for guerrilla warfare and,
contrary to Vietnam or Algeria, Madagascar had no sympathetic bordering countries. The rising was
quashed with much bloodshed - more than 80,000 Malagasy casualties.
In 1956 the French "loi cadre" instituted universal suffrage and transferred a significant share
of executive power to the Malagasy, essentially to the c"tier-dominated Parti Social D‚mocrate
But in was not until the 1960s that De Gaulle's France bowed to the inevitable and restored
Madagascar to full independence, although on terms very favourable to France.
Under the Malagasy Republic's first president, the "c"tier" Philibert Tsiranana, the French were
allowed to retain their hold over trade and financial institutions and keep their military bases
on the island. Tsiranana also maintained a dialogue with South Africa and refused contacts with
The increasingly unpopular Tsiranana was overthrown by a military coup in 1972 and this initiated
a fundamental change in Malagasy policy. Aid agreements with France were renegotiated, the French
military bases were closed down and so was an important NASA tracking station.
The economy saw a return to the collective work system that had been practised in the rural areas
before colonisation. The new government also severed diplomatic relations with South Africa,
Taiwan and Israel and formed new links with China, the USSR and especially with North Korea.
Several military men succeeded each other in rapid succession and eventually the government
stabilised under Lt-Commdr. Didier Ratsiraka, a "c"tier", in 1975. Banks were nationalised, and a
network of public corporations was set up to deal with the agricultural, marketing and even
mining sectors. The petroleum refinery and most mineral resources were nationalised. By and large
the new leftish tendencies of the Government resulted in costly failures exacerbated by the
1981-82 international debt crisis.
By 1990 the Ratsiraka's regime, like most left-leaning regimes worldwide, was showing serious
signs of exhaustion and a lack of capacity to solve Madagascar's growing problems. An opposition
party, "Force Vives", took shape and a long political transition period started. Among other
measures the new Mining Code of August 1990 improved the terms for foreign investment.
A new constitution - endorsed by national referendum in August
1992 - was adopted. Legislative elections in June 1993 produced a majority in the Assembly for
the opposition party "Force Vives", which promptly elected a prime minister from its own ranks. A
President, Professor Albert Zafy, also from "Force Vives", was elected in a ballot between eight
candidates in February 1993. With two thirds of the vote he easily beat Ratsiraka, who gained
Although the problems affecting the Malagasy economy remain very serious, Madagascar's peaceful
transition from a military dictatorship to a parliamentary democracy bodes well for its future.
5.1.2 Population and Culture
History and geography account for the diversity of Madagascar's population which was estimated at
13,100,000 in mid-1995. The population is increasing at a annual rate of 3.1%, fast outstripping
Madagascar's capacity to feed and employ them.
The majority of Madagascar's population descends from Malay and Indonesian migrants. They began
arriving in the 6th century when the powerful Hindu-Sumatran empire of Srivijaya controlled much
of the maritime trade in the Indian Ocean.
These people brought with them the food crops of South-East Asia, so that even today Madagascar's
agriculture resembles Asia more than Africa. These earlier colonists brought with them a strong
dislike and suspicion of India which persist to this day.
There are 18 principal ethnic groups on the island. Here below is their distribution (1974
- Merina (Hova).............1,993,000
The dominant ethnic groups, the Merina and the Betsileo, who inhabit the most densely populated
central provinces of Antananarivo and Fianarantsoa, are of Asian origin. In the coastal areas
live the tribes collectively called "c"tiers" of which the most numerous are the Betsimisaraka on
the east coast. These tribes are of African and Arab origin. There is a lingering mistrust
between "c"tiers" and Merina-Betsileo people, the old ruling class in pre-French Madagascar.
Population density ranges from more than 30 inhabitants per km2 on the fertile central highlands
(where all Madagascar's gold occurrences are) down to two per km2 on the largely sterile west
French nationals, many married to Malagasy, numbered less than 15,000 in 1986 and their number is
decreasing. A minor, but persistent, point of friction between Madagascar and France are some
minor islands in the Mozambique channel that are claimed by both France and, reasonably enough,
A Chinese community, numbering about 10,000, is dispersed throughout the east-cost region, where
they are principally employed as grocers, small-scale bankers and traders. Also inhabiting the
coastal areas are 10,000 or so Indian nationals.
Indians are rather unpopular with the Malagasys because of their clannishness and their wealth,
acquired through control of the textile and jewellery trade and urban real estate. Both Indians
and Chinese are major players in the vigorous gold smuggling trade out of Madagascar (often
through Mauritius), which the Government is all but incapable of controlling.
More than 82% of the Malagasy still live in rural areas, but the towns are inevitably attracting
an ever-larger percentage of the fast-growing, under-employed rural population, thus seriously
aggravating socio-economic problems. Antananarivo the capital, far from being the gracious,
pleasant little city of the '70s, is now a sprawling city of more than 1,000,000 people, mostly
living in very poor conditions. The other six provincial capitals also continue to expand.
The universal languages of Madagascar are Malagasy and French (apart from the vernacular dialect
of any particular tribe), which are currently spoken by anybody who has attended primary school.
English is understood a little by the most educated levels of the population in the major cities,
but not at all in the countryside.
5.1.3 Currency and Banking
The national currency is the Malagasy Franc (MG Fr). In December 1995 the exchange rate was as
- 1 $US = 4,499.80 MG Fr
- 1 œStg = 6,967.80 MG Fr
- 1 DM = 3,038.66 MG Fr
Since 1994 the Malagasy Franc has not been pegged to the French Franc.
Liberalisation of the banking sector has attracted foreign private banks to Madagascar. Banque
Nationale de Paris is the main shareholder, owning 55% of Banque Malgache de l'OcŠan Indien
(BMOI), which was officially opened in January 1990. Belgian and German banks own 20% of BMOI,
and 25% belongs to about 300 Malagasy shareholders.
In 1991 a public investment program (1991-1993), financed by the World Bank, IMF and other
international donors, was initiated. The program, which cost an estimated $US 1,000M, again
placed emphasis on the liberalisation of trade and domestic sector, the encouragement of foreign
and domestic investment, a reduction in the role of parastatal bodies, and increased producer
Following a prolonged general strike in the second half of 1991, however, the program was
deferred, pending the fulfillment of certain conditions, which included the transfer of a number
of parastatal organisations to the private sector and a reduction of public expenditure.
The general strike of 1991 resulted in severe deterioration of the financial sector, with a major
fall in tax revenue and a rise in the budgetary deficit. That is now around 5% of GDP.
As matter of priority the new democratically elected president, Prof A. Zafy, established an
economic and financial co-operation committee in 1993 to conduct negotiations with international
donors, and resumed an economic program in conjunction with the IMF. The adoption of a number of
development projects, which included plans to rehabilitate infrastructure (with the assistance of
the World Bank) was also envisaged.
Madagascar's climate is temperate and rather pleasant; temperatures in Antananarivo are generally
between 8§ and 27§ C, with cooler, dryer weather between May and October. The coastal region is
tropical with an average daily temperature of 32§ C. The rainy season extends from November to
April in the highlands - average annual rainfall is 1,000 - 1,500mm - but is more prolonged on
the coast, where annual rainfall can reach 3,500mm.
Geological exploration can be carried out continuously throughout the year.
5.2. AVAILABLE DOCUMENTATION
Madagascar has fascinated a number of authors and researchers (and still does). This has created
a wealth of easily accessible backgound information; some of the most interesting works are given
in the bibliography of this Database.
Far less accessible is Madagascar's geological information. Mr H Besairie's classic, the
two-volume "Gites MinŠraux de Madagascar", is unfortunately out of print. The extensive gold
bibliography on Madagascar by French authors (appendix 1 to this Database) is very difficult to
access and some items have apparently been lost.
However the modern gold explorer in Madagascar still has (as of 1996) some reliable basic
information to guide exploration. Apart from the various works mentioned in the bibliograhy
(Chapter 9 of this Database) three sets of maps are strongly recommended for any exploration work
MADAGASCAR GEOLOGICAL MAPS AT 1 : 1,000,000
By H. Besairie
Projection: Conforme Laborde.
The point X = 800km and Y = 400km
has for geographical co-ord Lat 21§ & Long. 49§E, Paris.
See Sketch 5/2 in this Database for sheet coverage.
- A = North Sheet (Feuille du Nord)
- = Central Sheet (Feuille du Centre)
- C = South Sheet (Feuille du Sud)
N.B. Only 2,500 printed in 1965 at Antananarivo.
These maps may be purchased from: Service Geologique
BP 280 Antananarivo 101
MADAGASCAR GEOLOGICAL MAPS AT 1 : 500,000
Author: H Besairie
Projection: Conforme Laborde.
The point X = 800km and Y = 400km
has for geographical co-ord Lat 21§ & Long. 49§E, Paris.
See Sketch 5/2 in this Database for sheet coverage.
- 1 = Diego Suarez (now Antsiranana)
- 2 = Antalaha
- 3 = Majunga (now Mahajanga)
- 4 = Tamatave (now Toamasina)
- 5 = Tananarive (now Antananarivo)
- 6 = Morondava
- 7 = Fianarantsoa
- 8 = Ampanihy
N.B. Only 1,000 copies printed in 1972, Antananarivo.
These maps may be purchased from: Service Geologique
BP 280 Antananarivo 101
MADAGASCAR TOPOGRAPHICAL MAP (AERONAUTICAL) AT 1 : 1,000,000
Projection: Conformal Conical Projection,
Standard Parallels 1§20' and 6§40'
Convergence Factor . 06979
See Sketch 5/2/1 in this Database for sheet coverage.
- N6 = North of 15th Parallel
- P6 = Between 23rd and 15th Parallels
- O6 = South of 23rd Parallel
N.B. Each map is subdivided into four individual maps at 1:500,000 scale.
Civilian users may purchase these maps from the following address:
DMA Combat Support Center
Washington DC 20315-0020
These three products are of excellent and still unsurpassed quality. The first two can be
purchased in Antananarivo (although the supply will inevitably run out); the third can be ordered
in the USA.
The two sets of maps at 1:1,000,000 (geological and topographical) can be overlayed on each other
only locally due to a different projection system (the French-generated maps use the now obsolete
Laborde co-ordinates system.
If a Geographical Information system is used to generate one's own maps, IMR strongly recommends
using the Conforme Conical Projection used by the US Air Force.
These three documents alone (see sketch 5/2 and 5/2/1) are adequate for reconnaissance work. More
detailed geological maps at 1:200,000 are also available at Antananarivo but their coverage of
Madagascar is incomplete and several maps are now out of print.
Madagascar's "Institut GŠographique et Hydrographique National (Foiben-Taosarintanin'i
Madagasikara = FTM) publishes a series of maps of Madagascar at the 1:500,000 scale and an
excellent, informative, and up to date road map at the 1:2,000,000 scale.
Air photos are available but they are old, subject to restrictions and to tiring beaurocratic
haggles. On the other hand, excellent satellite images (particularly the French SPOT coverage)
are readily available outside Madagascar and can be conveniently blown up, up to the 1:50,000
scale. IMR strongly recommends their use: the morphology of Madagascar's highlands, where most of
its gold districts are, is ideally suited to satellite-image interpretation. Modern Global
Positioning System instrumentation can now assure a level of accuracy not previously available in
any published map of Madagascar.
5.3 INSTITUTIONAL AND ECONOMIC FRAMEWORK
5.3.1. The Government
Sketch 5/3 outlines the present political structure of Madagascar. The uncommonly large number of
ministries (21) reflects its French political heritage. Historically, the number of these
ministries is periodically pruned in times of political reform, then start to regrow immediately
afterwards to accommodate the need of political patronage, regardless of the political party in
The unusual combination of combining the offices of Prime Minister with the Ministry of Defence
is an insurance policy against military coups. So is the fact that the national police reports to
the Ministry of Defence (Prime Minister, really) rather than the Ministry of Interior.
This unusual constitutional arrangement has now been tested for only two and half years, but it
seems to work. The main area of friction remains between the respective powers of the
popularly-elected President and Head of State, and the National Assembly-elected Prime Minister.
The town of Antananarivo is chronically plagued with rumours of coups. This goes back for at
least 30 years. The inexperienced visitor is well advised to dismiss them and under no
circumstance demonstrate the least interest in the subject.
The Government is now seriously committed to regional decentralisation, with new regional
authorities known as Collectivit‚s Territoriales D‚centralise‚s (CTDs) elected by their own
special elections. It is impossible to say at this stage if the CTDs will have some control on
local mining operations. It is possible, particularly in environmental matters.
Mining and exploration activities remain controlled by the Ministry of Energy and Mining through
its two main Departments of Mines and Geology.
Unfortunately, a dualism of function remains in the mining sector with the presence of OMNIS,
which operates in the mining field, particularly in oil exploration and chromite production.
OMNIS was created directly by president Ratisraka in 1975 in the heyday of military government.
OMNIS originally meant "Office for Military National Strategic Industry", now renamed to "Office
for National Mining and Strategic Industry" and reportd directly to the President.
The status of OMNIS is somewhat anomalous in a civilian government, and Kronsten in 1994
mentioned "reforms urged by the World Bank, which exercises considerable influence on the current
coalition government, include the privatisation of the parastatal OMNIS". To this date (the last
quarter of1995) this privatisation has yet to take place.
Other ministries likely to affect any important mining operations are the Ministry of Finance and
Budget (export permits, taxation regime etc) and the Ministry of Civil Service (labour contracts
5.3.2. The Economy
Madagascar was classified by the World Bank in 1992 as one of the world's poorest countries. In
that year, according to estimates by the World Bank, Madagascar's gross national product (GNP) -
measured at average 1990-92 prices - was $US 2,809M, equivalent to about $230 per head.
In 1985-92, it was estimated that GNP per head declined, in real terms, by an annual average of
1.7%, while the population increased by an annual average of 3.1%. Madagascar's gross domestic
product (GDP) increased, in real terms, by an average of 1.1% per year in 1980-92. The growing
gap between population and economic growth is obvious.
The rural economy accounts for 80% of Madagascar's export revenues and supplies most of the raw
materials for industry. In 1992 the agricultural sector (including forestry and fishing)
accounted for 33% of GDP and engaged an estimated 76% of the country's labour force.
The production of both food and export crops, except cotton, either stagnated or declined after
President Ratsiraka took power in 1975, mainly as a result of the imposition of a doctrinaire
form of socialism. The presence of North Korean "advisers" obviously did not help. The imposition
of co-operative state farms on a reluctant peasantry also contributed to the agricultural
sector's poor performance.
Between 1980 and 1992 agricultural GDP increased by an annual average of 2.4%, still considerably
below the population growth rate.
The forestry sector has been badly neglected, and 81% of domestic fuel needs are supplied by wood
and charcoal. In 1991, according to a FAO estimates, about 26% of the land area was covered by
forests. A $66M aid arrangement was agreed to at a donors' meeting in Paris in February to
support the first stage of a long-term environmental action plan.
In August 1989 a "debt-for-nature" exchange, the first in Africa, was arranged: the USA gave the
Worldwide Fund for Nature $US 1M to buy $ 2.1M of government commercial debts at a 55% discount
from a consortium of international banks, and to use the money for environmental projects. This
imaginative scheme proved successful and further debt-conversion agreements, involving
international organisations, were subsequently negotiated.
Paddy rice is the main Malagasy crop. It is grown by 70% of the population whose basic food is
rice (the average annual consumption is about 135 kg per head, the highest in the world).
Paddy rice output in Madagascar averages about two tonnes per hectare. Madagascar, once a rice
exporter, has become a net importer since 1982, although improved economic policies can correct
this imbalance in the future. Other important staple crops are maize, cassava, bananas and sweet
Madagascar's main cash crops are coffee, vanilla and cloves. The most important is coffee (97% of
it robusta, although arabica production is encouraged) which in the 1980s accounted for about 24%
of total export earning and engaged 25% of the working population. In some Madagascan goldfields
coffee vies with artisanal gold mining in land use, with changes from one activity to the other
according to gold and coffee prices.
Madagascar's coffee production for the 1980s was generally about 82,000 t/y - almost the double
the export quota allocated by the International Coffee Organisation (ICO). Madagascar thus had
considerable quantities for disposal in the non-quota market.
Madagascar is the world's largest exporter of vanilla, which accounted for 19% of its total
export revenues in 1992, with about 1,000 t/y production. Madagascar operates a highly effective
price control system on vanilla with Reunion and Comoros. Production of cloves in the early 1980s
was around 10,000-12,000 t/y and accounted for about a quarter of export earnings.
Other significant Malagasy crops are cotton, sisal, sugar, groundnuts, pineapples, coconuts and
Industry accounted for 14% of Madagascar's GDP in 1992; the sector employs only about 3% of the
working population - not an uncommon figure for central Africa.
The island's major industrial centres, other than mines, are located in the High Plateau or near
Toamasina port. Food processing accounts for 49% of all industrial value added. Textiles was
formerly the second largest sector but it has been now superseded by brewing, paper and soap.
There are cement plants at Mahajanga and Toamasina, supplied with good quality local limestone.
However, in the late 1980s average annual production of cement was only about 40,000 t
necessitating imports of about 250,000 t/y.; by 1990 production had declined to a paltry 20,000
A fertiliser plant at Toamasina, which began operation in 1985, produces 90,000 t/y of urea and
ammonia-based fertilisers. Other industries include the manufacturing of wood products. Most
industrial plants in Madagascar have been operating recently at less than one-third of their
In 1986 the government introduced a new investment code, which provided incentives for domestic
and foreign private investment in activities outside the public sector, particularly in
manufacturing for the export market.
The country's fifth investment code, enacted in 1990, introduced incentives to attract foreign
private investors. This code was strongly opposed by many politicians and local business people.
Rules regarding foreign exchange and the number of expatriates which can be employed have been
relaxed and tax incentives were introduced, including a five-year holiday from corporation tax.
A number of export processing zones (EPZ) have been established and have attracted foreign
investors from South East Asia, France and Mauritius. Labour is cheaper and corporation laws are
lower in Madagascar than in France or Mauritius but the risks are higher for companies producing
for export. One common complaint by Asian investors has been the still modest education of
Malagasy labourers, as in the rest of Africa.
In the early 1990s the IDA approved a credit of $48M to finance the development of private
enterprise through training of personnel, the restructuring Madagascar's Chamber of Commerce, and
the establishment of an investment promotion agency. These initiatives, although unquestionably
praiseworthy, had practically no effect on the mining sector.
5.3.3. The Mining Sector
The high French expectation of Madagascar's minerals at the time of its colonisation (1895) never
The then vigourous gold production has ceased for decades to be an organised (taxable) activity
and does not now contribute to government revenue. The promising coal deposits of Sakoa (a major
preoccupation for the French steam fleet of the time) never really took off, the enthusiasm for
Madagascar's uranium was short lived and doomed.
Bulk commodities such as iron ore, bauxite and oil shales were condemned from the start by lack
of local industry or export infrastructure. Oil exploration has been particularly disappointing
with company after company drilling a few desultory holes in Madagascar's Karroo formations and
then withdrawing. In the '90s only chromite, graphite and mica survived as organised mining
activities and none was particularly wealthy. Madagascar is essentially an agricultural economy.
IMR's own impression is that Madagascar suffers "exploration fatigue" out of proportion with its
objective, and still largely untested, mineral potential. Foreign explorers eschew Madagascar.
Original geological work is minimal and no new ideas seem to be generated by an anaemic mining
sector. A country that only 30 years ago was one of the most advanced in central and east Africa
now languishes behind mosts in ideas, investments and energy; at least in the mineral sector.
The Malagasy Mining Code
In Appendix 2 to this Database is given the (barely readable) official version of Madagascar's
latest Mining Code (Law No. 88-06) of August 26, 1988. This bulky document consists of 121
articles. It has been drafted for the stated objectives of:
1) Creating a favourable climate, from the juridic, fiscal, technical and social perspectives, to
attract new investment in the [mining] sector and therefore to complete similar measures taken in
the whole national economy;
2) Decentralising the [mining] administration for better serving the general interests;
3) Increasing the professionalism of the [mining] sector, putting in places a set of clear and
appropriate rules in a way to offer a framework for the rational and optimum development of
national mineral resources.
This document unquestionably improves the terms for foreign investors but it still remains an
untested document, as Madagascar has failed to attract large new mining investors. Since the
nationalisation of the chromite industry in 1976 no major international mining firm has been
operating in Madagascar.
In addition, some important new mineral projects (oil, coal and petroleum exploration) have been
negotiated directly with the parastatal OMNIS organisation and not with the Ministry of Energy
and Mining (see sketch 5/3); this has tended to create some confusion.
In IMR's opinion the 1990 mining code is perfectly adequate and flexible enough to open serious
negotiations on exploration for large gold projects in Madagascar.
Chromite and graphite are the only minerals currently exploited in substantial volumes in
Madagascar. Their contribution to the GDP is less than 1%.
Madagascar is the world's tenth largest chromite producer with output at around 110,000 t in
1994, down from about 200,000 t/y in 1976 when OMNIS took over from Pechiney. This chromite
production, negligible by South African standards, is however larger than the production of
Australia, Japan, US and China combined.
With chromite prices (April 1995) about $ 70-80 per tonne Madagascar's total chromite production
is only worth about $9,000,000. This is equivalent to about 708 kg of gold or one fifth of
Madagascar's estimated "artisanal" gold production of 3-4 t Au/year.
Arguably then, even now gold is by far the most important mineral product in Madagascar, even if
the benefits from it do not reach the population at large. Which is, of course, deplorable.
Graphite production, the second largest "official" mineral product in Madagascar, totalled 10,600
t in 1992, down from 18,500 t in 1990.
The graphite industry in Madagascar is a very old and tired one. Without new investments,
Madagascar's graphite industry will be seriously affected by the competition of its more
aggressive neighbours. These neighbours - in a geological sense at least - are starting to
exploit the Malagasy type of graphite in their own territory: the Ancuabe mine (5,000 t/y) in
Mozambique, the Graphtan Merelani mine (15,000 t/y) in Tanzania, the Maranamadurai mine
(15,000-20,000 t/y) in Southern India. All these mines were commissioned in the period 1994-5 and
they are producing the same product (large flake graphite, >98% C) as Madagascar and therefore
significantly eroding Madagascar's world market share.
Along with Canada, Madagascar is one of the world's few producers of phlogopite, with exports of
1,800 tonnes in 1990, mostly by semi-artisanal operations around Tolanaro-Amboasary in the
extreme south-east of the island. A small percentage of the production is in the form of sheet
Madagascar also produces marble, a variety of construction materials, soil conditioners, and
semi-precious stones (amethysts, tourmaline, beryls, and garnets).
Bauxite deposits at Manantenina in the south east are estimated at 100 Mt and have been assessed
for decades by French, Soviet and more recently Italian and EU experts. The bauxite quality is
acceptable (barely: 38% Al2O3) but the total lack of infrastructure and Madagascar's fragile
environment work against its exploitation.
The coal deposits of Sakoa have been known for more than a century and the stated reserves are
>100 Mt. A number of companies including the Australian BHP have carried out evaluation in the
past but, once again, lack of infrastructure would make exploitation uneconomic.
The largest project under consideration is the extraction of ilmenite from an area of rainforest
and sand dunes at Toalagnaro on Madagascar's south-east coast.
Under a letter of intent signed in November 1994 by Quebec Iron and Titanium (QIT, a subsidiary
of the RTZ group) and the state mining group OMNIS, extraction could take place for 30 years and
it is estimated will earn up to $550 million.
The project will involve the construction of a new port and other infrastructure. The project has
attracted considerable flak from international environmental groups, concerned - among other
things - about the biodiversity of Madagascar's rainforest.
In the case of Malagasy mineral sands the alarm seems premature.
Madagascar's ilmenites present several extraction problems (Bartle, 1988). The deposit is low
grade and the ilmenite is of sub-standard quality. Its unacceptably high Cr2O3 and manganese
content and highly radioactive (Th/P) contaminants will require costly treatment processes before
a commercial product can be obtained.
These deposits have been evaluated for decades by a number of companies and individuals
(including the writer), with mixed reports. It would seem that Malagasy ilmenite is a rather
atypical ore and apparently only suitable for the production of a chlorinatable slag - not QIT's
preferred route for its products. It is therefore doubtful that Madagscar's ilmenite will be
exploited any time soon (IMR's opinion). The fact that Malagasy authorities should insist on a
vigorous and conclusive evaluation of its mineral sands (as Mozambique is doing) is entirely a
5.3.4. The gold sector
In Madagascar's socialistic climate of 1975, the country's gold industry was nationalised (Anon
1994). This was largely a symbolic move as official gold production in the 1970s had dwindled to
In reality the government did not have the resources to implement any policing of Malagasy
goldfields and as soon the gold prices picked up at the end of 1970s "artisanal" gold production
The distinction between illegal and "artisanal" gold production is a thorny one for any
developing country, and particularly for Africa. The authorities are aware that "artisanal" gold
mining is not either a sustainable or socially constructive activity but they are impotent to
Lack of information and an inadequate mining environment has deterred most foreign companies
from seriously committing themselves to gold exploration in Madagascar.
The Government's half-hearted attempts to entice foreign capital towards organised (taxable) gold
production in Madagascar came to little.
The public become seriously worried in 1993 when the government gave a little-known Swiss group
exclusive nationwide gold and prospecting rights for 25 years, in return for the placing of up to
$US2 billion in discounted promissory notes from the government (MJ Annual Review, 1995). The
scheme - of very dubious legal validity in any case - was shelved. If anything, this episode
proved the surprising degree of naivety of the Malagasy authorities on mining matters.
In late 1994 the government decided to form a gold trading agency through which the ministry of
Energy and Mining would buy the output of private prospectors (see appendix 3). This agency will
deposit sale proceeds with Malagasy central bank and it is designed to boost official revenue at
the expense of smuggling, particularly to India and mainland Africa.
Similar schemes have been implemented in other parts of Africa (ie. Tanzania, Angola, Zaire,
etc). The writer - who has participated in several UN or government-sponsored seminars on similar
matters - is unconvinced of the wisdom of such a move. The native gold producer will sell gold to
a government agency only if the price offered is superior to the market one (only China manages
to buy most of the gold from its miners at a significant discount from the international price,
thus creating a commercially viable agency).
In other words, to persuade individual miners to part with their gold a government has to
subsidise gold mining. This could mean subsidising an illegal, undesirable and unsustainable
practice with no return for the Government.
The problem is not new. It is occurring in dozens of African countries, in Latin America, in
Papua New Guinea. Each country will react to this problem (because a problem it is) according to
the possibilities and its own political and cultural framework. This is not the place to
elaborate on this.
However, it would not seem that individual miners would represent a direct threat to a major gold
explorer in Madagascar. Such an explorer is traditionally attracted by large - and therefore
comparatively low-grade - gold deposits with a substantive vertical expression, often requiring
sophisticated recovery plants. Superficial or alluvial gold deposits are not favourite targets
with such investors. Of 16 major new gold projects under development worlwide in 1995 (Thomas,
1995), none was alluvial.
It is IMR's firm opinion that large-scale gold deposits have simply not been sufficiently
explored in Madagascar and, at this stage of knowledge, they constitute a very good potential target.
Madagascar's mountainous topography has traditionally hindered communications. In 1991 there were
34,750 km of classified roads, of which 8,540 were main roads and 18,380 km secondary roads. Only
5,350 km were paved. The rest of the road network is occasionally unusable in the wet season
(November-April) unless using 4WD vehicles.
In 1989 the European Development Fund (EDF) granted $10M for road rehabilitation in the north and
west of the country. In mid-1994 a program to rehabilitate a further 500 km of main roads, at an
estimated cost of $8 M, was under consideration by the EDF.
The excellent "Carte RoutiŠre" of Madagascar at the scale 1:2,000,000, issued in 1990 by the IGM
(Insitut G‚ographique National) and available in any major bookshop, is all one needs for a
reliable and complete guide to road transport in Madagascar.
It should be stressed that all the major Madagascan goldfields (with the possible exception of
Andrarona) are easily accessible with standard 2x4 cars rented from commercial agencies in
5.4.2. Air transport
Domestic air services have always been important in Madagascar, on account of its size, difficult
terrain and often poor quality of other forms of transport. There are 211 government-approved
airfields (see "Carte RoutiŠre de Madagascar" for their location), two-thirds of which are
privately owned. The only international airport is at Antananarivo.
The national airline, Air Madagascar, is two-thirds owned by the government and one third by Air
France. Under the government's new liberalisation measures, Air Madagascar lost its monopoly on
domestic services and in 1992 its heavily subsidised external services were substantially
Any goldfield in Madagascar is a convenient distance (30-80 km) from an official airport, and
several air charter companies in Antananarivo offer reliable services. Air Madagascar links most
of the major townships with a fleet of Boeing 737 and Twin Otters; the timetable is given in
appendix 4. The service is totally reliable.
Charter plane and (slightly more difficult) helicopter hire are available in Antananarivo at
prices comparable with international standards.
Adequate telephone and facsimile facilities are available in all major towns. Communications
between Antananarivo and the rest of the world do not present any problem. Communications by
telephone within different localities in Madagascar can at times be difficult.
The use of two-way radios is restricted and requires a special permit.
Practically every village in Madagascar is equipped with a small post office. The largest ones
have telephone connections.
Railways and maritime transport play only a very subordinate role in gold mining and therefore
will be dealt very briefly. Madagascar has 1,095 km of railways. Three lines in the north of the
country primarily serve the capital, while the fourth, in the south, links Fianarantsoa to
In 1986 the World Bank agreed to lend $12m to finance the rehabilitation of the northern railway
and its extensions for a further 40 km.
Toamasina and Mahajanga, the principal seaport, suffer from a lack of storage space and
equipment. Toamasina port handles about 70% of Madagascar's foreign trade and was in the process
of being enlarged and modernised, until 80% of the port was destroyed by a cyclone in March 1986.
Germany, France, the UK and the World Bank are financing a ongoing project to rehabilitate 10
Malagasy ports. Toamasina is independently managed but the other ports are operated by the
Malagasy Ports Authority. The parastatal SocietŠ Malgache de Tranports Maritimes has four
ocean-going ships, while coastal shipping is mostly handled by private firms. The country has a
total of 18 ports handling a little less than 1 M tonnes of freight a year.
6.1. General information
Practically all geological work in Madagascar took place during the French colonial time
(1895-1960) - the golden age of Madagascar's geology. Many first-class geoscientists did their
doctorate theses on various aspects of Madagascar's geology or spent years in the then vigorous
Madagascar Geological Survey (Service Geologique). This service was headed from 1930 to
independence by the still legendary Henry Besairie.
Systematic geological mapping at various scales started in 1926. The geological map at 1:200,000
was completed in 1951 and in 1960 the geological maps at 1:100,000, covering almost two-thirds of
the country. Some of these maps are unfortunately out of print.
Between the two World Wars intensive exploration took place, identifying deposits of coal,
bitumen, graphite, mica, several precious stones, cement-quality limestone, glass-quality silica
sand and many other commercial minerals.
Immediately after WW2, exploration for oil and uranium was actively undertaken, respectively by
the S.P.M. (Societ‚ des P‚troles de Madagascar) and the C.E.A. (Commissariat … l'Energie
Atomique). For a short time Madagascar was a uranium producer, exporting several hundred tonnes
of urathorianites from the T"lan"ro (then Port Dauphin) region.
Since the end of colonial rule geological mapping and mineral exploration decreased dramatically
as any computeried literature search can confirm: less has been published on Madagascar's geology
in the 35 years after independence than in the last five years before independence. As a
consequence, and particularly in the non-French speaking world, the geological understanding of
Madagascar has remained grossly inadequate. This is despite the growing interest in Madagascar
because of its pivotal position in Gondwana (Katz and Premoli, 1978; de Wit, 1988; Kroner, 1991).
This information gap is actually widening. While more and more literature is generated on
Madagascar's geological neighbours (East Africa, south-west India, Sri Lanka, Antarctica), very
little new information is coming from Madagascar.
Another reason for the relative obscurity of Madagascar's geological and mineral endowment is
that although a bibliography on this subject lists several thousand titles (Besairie, 1971), this
literature is not generally available.
Mostly it consists of internal reports by Madagascar's Service de Mines and Service Geologique,
and they are difficult to obtain even in Antananarivo. Indeed, the principal sources of
information for this Database are Besairie's various maps and publications (some of them now
unfortunately out of print), supplemented by modern publications on Madagascar's regional geology
6.2. Madagascar's geomorphology
The geomorphology of Madagascar reflects its underlying geology and either one can be glanced at
in sketches 5/1 and 6/1.
The mountainous backbone of the island consists of Precambrian rocks. To the east lies a narrow
coastal plain defined by Cenozoic normal faults which strike NNE-SSW, and impart a striking
linearity of coastline.
The western shore is less regular. The Morondava basin lies south of Tanjona Vilanandro (former
Cap Saint Andr‚). To the north of this cape is the Majunga basin, and still farther north is the
Antsiranana (formerly Diego Suarez) basin. These three sedimentary basins are largely in faulted
contact with the Precambrian. The cuesta-forming beds dip gently seawards.
This area of Madagascar is sometimes called the "Karroo corridor" and, similarly to other Karroo
regions of Africa, it hosts interesting hydrocarbons, coal and uranium deposits (Premoli, 1995).
There are unexplained lacunae in Madagascar's geological records. After the Pan-African event,
principally consisting of thermo-metamorphism around 550 MA, there is essentially no record of
Phanerozoic rocks until the late Paleozoic (Karroo formations). This interval must include the
plant-bearing rocks in which Devonian or Carboniferous flora has been identified but whose
outcrops have not yet been found.
Late Paleozoic and younger rocks in the marginal basins show a succession of marine and
non-marine events. But since sedimentary basins play no role (alluvial deposits excluded) in
Madagascar's gold metallogeny they will not be discussed here in any detail.
All the known primary gold mineralisations in Madagascar predate the breaking up of Gondwana,
making regional tectonic models (India, Mozambique, Tanzania) particularly relevant. Equally so
is a good understanding of Madagascar's Basement complex.
6.3. Madagascar's Basement complex
Precambrian rocks form about two-thirds of Madagascar. They are exposed continuously from
Tolanaro in the south to Antsiranana in the north (see sketches 5/1 and 6/1). These Precambrian
terrains are often referred to as the Basement or the crystalline Basement of Madagascar.
The metamorphic grade of this Basement is generally high, and for almost the whole island varies
between the amphibolite and the granulite facies (Premoli, 1979). Greenschist facies are
comparatively rare. The tectonic history and geology of Madagascar presents several similarities
with both eastern Africa and peninsular India.
Madagascar's Basement underwent three major tectonic periods which can be summarised as follows
(from the youngest to the oldest):
- After 300 MA Era of platform cover and great
- Between 2,600 and 300 MA Era of intercratonic orogenesis
and great magmatic reactivations
- Before 2,600 MA Era of cratonisation
Madagascar's Basement rocks have been divided into three major systems. From the youngest to the
oldest these are:
- The Vohibory System
- The Graphite System
- The Androyen System
Each of these systems has been subdivided into several groups. These groups usually have outcrops
in different locations and consist of different lithologies. As a result the assignment of a
group to a particular system and the age relationship between groups in one system is still to be
determined with confidence in most cases. This problem is exacerbated by the general lack of
modern geochronological data in Madagascar from the last 35 years.
Madagascar's Basement rocks have been intruded by several igneous episodes. Some of them may have
contributed to the remobilisation and concentration of gold (model a4, see chapter 6.5). The
principal intrusive phases are indicated below (from younger to older):
- 90 MA: Cretaceous intrusive complex of Antampombato,
consisting of gabbros, syenites, pyroxenites,
pridotites and possibly the Itremo carbonatites
- 500 MA: pegmatites )
) of the Pan-African event
- 550 MA: granites )
- 850 MA: granites (poorly known) associated with
- 1125 MA: granites and syenites of Ambatofinandrahana
intrusive complex and associated nephelinitic
and gabbroic complexes. It compares with the
Kibarian orogeny in other parts of Africa
- 1400 MA:Andrianamena basic complex, including also
ultrabasic rocks (gabbros, pyroxenites,
peridotites) which have been metamorphosed to
orthopyroxenites, orthoamphibolites, talc-
schists and serpentinites
- 1890 MA: Intrusive episode (granite, granodiorite)
affecting the whole Madagascar basement.
A systematic and detailed study beween granitic intrusions and gold mineralisations (which has
proved highly successful in Australia) has never been attempted in Madagascar. Indeed only in the
last eight years has improved analytical and computer techniques made such study meaningful. Any
serious gold explorer in Madagascar should give early attention to gathering data in this field.
6.3.1. The Androyen System
This system is the oldest on the island and consists of highly metamorphosed, often granulitic,
rocks. This system is sometimes referred to in French literature as "Migmatitic-leptynitic Lower
Complex". The Androyen system outcrops are principally in the southern part of the island where
they occupy a roughly triangular area which extends northwards about 450 km (see Plate 1).
The Androyen System has been broken into three groups (Besairie, 1964), from the youngest to the
- A3 : Ampandrandava Group - Gneisses,
Pyroxenites, Charnokites, Marbles
- A2 : Tranomaro Group - Pyroxenites, Gneissess (Ca
- Mg), Leptinites, Marbles
- A1 : Fort Dauphin (now T"lan"ro) Group -
Leptynites with cordierite
Apparently, none of Madagascar's gold districts occur within the Androyen system and therefore
the interest of this system for gold exploration is rather limited.
6.3.2. The Graphite System
Rocks assigned to this system form the greatest part of Madagascar's Basement - and they are also
the most consistently mineralised in gold. The system is characterised by the presence of
graphite in greater or lesser abundance. The intensity of metamorphism varies according to
location; the different regional groups are still defined by their metamorphic grade rather than
their stratigraphic relationship.
Besairie (1964) distinguishes four groups within the Graphite System, some of which are closely
associated with gold districts. From the youngest to the oldest:
G1 B4 : Andriba Group - migmatites and gneisses
] B3 : Ambatolampy Group - micashists and gneisses
( with graphite
G2 ( B3 : Manampotsy Group - gneisses and migmatites
( with graphite
( B1 : Vohibory Group - Leptinites, gneisses
( amphibolic, amphibolites, marbles.
A more simple distinction can be made between migmatites without graphites (G1) and those that
are graphite-rich (G2). This easier differentation is particular relevant to gold exploration as
all the presently kown mineralisations occur in the graphite-rich zones of the Graphite System,
especially near Antanananrivo (see Plate 1, in this Database).
6.3.3. The Vohibory System
This system was first defined on Vohibory Mountain in the South-western part of Madagascar, hence
the name. Characteristic of the Vohibory System is the abundance of amphibolites in the form of
ortho-amphibolites, associated with basic to ultrabasic intrusions also metamorphosed to various
degrees. It would seem, however, that what the French geologists call an amphibolite their
Anglo-Saxon colleagues would call gneiss.
Besairie (1964) distinguished 10 groups within the graphite system, some of which are closely
associated with specific Madagascar gold districts. From the youngest to the oldest:
- C10 : Daraina Group - Amphibolites with
epidotes, gneiss, granodiorites
- C9 : Sambirano Group - Gneisses, quartzites,
- C8 : Antonigil Group - Epidote-rich migmatites
- C7 : Ambodiriana Group - Feldsphatic micashists
- C6 : Maentavanana Group - Greenschists,
magnetiferous quartzites, amphibolites,
- C5 : Beforona Group - Amphibolites, migmatites
- C4 : Manajary Group - Micaschists, gneiss,
- C3 : Amborompotsy Group - Amphibolic gneiss,
- C2 : Malakialina Group - micashists, marbles,
- C1 : Vohibory Group - Leptynites, amphibolic
gneiss, amphibolites, marbles.
A tentative correlation between this geological description, the Gondwana one and BRGM's
classification (edited and improved) is given in Table 2.
IMR would also like to bring to the attention of potential gold explorers two intriguing groups
in the Vohibory System:
i) C6, or the Maentavana group, with its most unusual magnetite-gold quartzites (nine alluvial
occurrences listed in Table 6). Their primary source could well be of economic interest in its
own right for the modern explorer.
ii) C4, or the Manajary group, what the French geologists have called greenschists. It may very
well be metamorphosed greenstone belts in the Vohilava-Ampasary gold district.
6.3.4. Concluding remarks
Of the three Archean "Systems" recognised by Besairie the Vohibory and the Graphite ones are
auriferous, the Androyen is not. Within the Graphite System the gold districts occur exclusively
within its graphite-rich parts and not in the much more extensive graphite-poor areas (see Plate
1). If a plausible rationale can be advanced for the lack of gold potential for the graphite-poor
groups within the Graphite System, then the area prospective for gold in Madagascar could be
Besairie considers the Betsiriry gold district as Archean (Vohybory System), but BRGM considers
it as Proterozoic (see sketch 7/2). However, this argument has little relevance with the
practical aspects of gold exploration in Madagascar at this stage.
In Madagascar there are obviously a lot of periods of tectonic-thermal overprints. Besairie sees
three major periods of gold mineralisation (from the youngest to the oldest):
140 MA - equivalent to Upper Cretaceous
1,125 MA - equivalent to Precambrian B (de Wit, 1988)
2,140 MA - equivalent to Precambrian C (de Wit, 1988)
IMR is sceptical about the period of mineralisation connected to the Upper Cretaceous period.
This time corresponds to tensile tectonics with associated alkaline intrusions, and particularly
carbonatites, throughout all Gondwana. These alkaline intrusions have important, and occasionally
economic, mineralisations (Cu, Fe, RE, Nb, F,) but so far no gold has been found in anything
near economic grades.
The impression (IMR's opinion) is that the 140 MA tectonic episode simply remobilised and
reconcentrated pre-existing gold mineralisations in the Basement, but did not introduce any
significant gold from mantle sources.
An attempt by IMR to conciliate the various different Precambrian stratigraphies of Madagascar
(Besairie, 1964, 1969; Jourde, 1971; Hottin, 1972; de Wit 1988) has been totally unsuccessful.
Considering that practically no new mapping or geochronological work has been carried out for
more than three decades and that Besairie' geological maps (particularly the ones at the scale of
1:500,000, see sketch 5/2) are the most useful on the field, Basairie's broad interpretation of
Madagascar's geology has been retained throughout this Database.
In the table below a comparison is given between the two main interpretations of the age of
Madagascar's known gold districts:
GOLD DISTRICT BRGM (1985) BESAIRIE (1966)
- BETSIAKA Basement/sedimentary Basement/sedimentary
- TSARANANA Archean, Fe-Ca-Mg series Vohybory System
- MAENTAVANA Archean, Fe-Ca-Mg series Vohybory System
- BETSIRIRI Proterozoic Vohybory System
- ITASY Archean, sialic series Graphite System
- AMBOSITRA Archean, sialic series Graphite System
- VOHILAVA Archean, sialic series Vohybory System
- BEFORONA Archean, Fe-Ca-Mg series Vohybory System
- ANDRARONA Basement/sedimentary ? Vohybory System
- VAVANTENINA Archean, sialic
- series Marble Series
This table has been expanded and elaborated on in Table 3, given in appendix to this Database.
How to usefully translate typical French concepts such as the sialic series (Series
"silico-alumineuses") and the Fe-Ca-Mg Series (Series "ferro-calco-magnesiennes") into the more
familiar western terms - calc-alkalic sequences, komatite-tholeiite, felsic volcanics, etc - will
no doubt keep many geologists busy in future.
6.4. Metallogeny of Madagascar's gold.
From what is known, all the primary gold deposits of Madagascar belong to the "mesothermal
lode-gold" deposits as defined by Hodgson (1993).
The reason why Madagascar's gold deposits are seldom or ever mentioned in the numerous review
papers on the world's gold deposits is that very little is known about them. But even a cursory
examination of now very obsolete literature shows that these deposits definitely do not conform
to epithermal, volcanogenic massive sulphides, quartz-pebble conglomerate or porphyry-style gold
Here below are the criteria that match conventional mesothermal lode-gold deposits with
Madagascar's own gold deposits:
- 1. The deposits are mostly quartz-vein related;
- 2. They occur in metamorphic terrains of all ages, not
excluding high-grade metamorphic terrains;
- 3. Minerals commonly associated with gold are pyrite (less
commonly pyrrhotite), base metals sulphides, arsenopyrite,
tourmaline and molybdenite;
- 4. Mineralisation may occur within most rock types but
typically in quartz veins, veinlet systems, or as
disseminations in heavily tectonised zones;
- 5. Most mineralisations are hosted by, and always related to,
steeply dipping reverse faults and shear zones;
- 6. At a regional scale, the mineralisations occur in
association with major transcrustal fault zones and small
felsic alkalic and trondhjemitic intrusions.
All these conditions, suggested by various workers in the field (Kirkham, 1993), have been
verified for Madagascar's gold mineralisations. But information is seriously lacking in the
A. Fluid inclusions (presumably CO2-bearing and with relatively
low salinity). There is no data whatsoever on Madagascar's
B. Stable isotope data - presumably in the range of +10 ë O
per mil (Kerrich, 1989). There is no data whatsoever on Madagascar's gold deposits;
C. Geotectonic environment - virtually all the papers on Madagascar's gold metallogeny predate
the acceptance of tectonic plates as a major ore-forming process. Particularly the association of
mesothermal gold deposits with major trascrustal faults (Hoffman 1991). Here a considerable
amount of field data exists but it needs a total reinterpretation, particularly utilising modern
It is obvious that any serious gold exploration in Madagascar should be aimed at filling the
above lacunae and putting their regional metallogeny into a modern prospective.
However, some general considerations can already be put forward.
Archean lode gold deposits similar to the ones found in Madagascar account for approximately 18%
of the world's gold production. Exploration for this type of deposit has increased dramatically
during the last 10 years, particularly in the Archean shields of Canada and Australia and, to a
lesser extent, in those of Southern and western Africa, Tanzania, Brazil and Venezuela. This
current phase of exploration has produced numerous discoveries, including several major deposits.
There is no reason, at the present level of knowledge, why this could not also be the case for
Until 1983, the great majority of the Archean lode gold discoveries in Canada and in Australia
resulted from surface prospecting. Virtually all of the remainder were discovered by the drilling
of mine extensions or geophysical anomalies.
Exploration success in the 1980s followed similar strategies, with considerable emphasis placed
on old mines. The Hemlo deposit (80 Mt with an average grade of 7.7 g/t Au), among the
outstanding lode gold discovery of the 1980s, had a history of exploration and minor development
dating from 1869 (Hodgson, 1993).
The West Australian gold rush at the end of the last century was led by prospectors and
culminated in the discovery of the Kalgoorlie deposits in 1893. The current gold rush (1978 to
the present) has seen the re-evaluation of many old gold occurrences, in most cases leading to
new discoveries. The economic viability of the new discoveries is strongly influenced by
sustained gold prices and new bulk-mining and processing technology, particularly where ore
occurs in the oxidised or lateritic profile as is unquestionably the case in Madagascar (see
A similar focus on old prospects and new technologies has formed the basis for recent gold mine
developments in Zimbabwe, Brazil and Ghana: all major and well-established gold producers.
This emphasis on re-evaluating old prospects when exploring for lode gold deposits suggests that
conceptual theories or genetic models have so far had little impact (Hodgson, 1993). This is
going to be particularly true for Madagascar with the imperfect and obsolete knowledge of its
Perhaps the most important empirical feature that has recently emerged in the exploration for
lode gold deposits is the association with subvertical structures, particularly shear zones.
The style of structure varies from brittle to ductile and shows a good correlation with the
metamorphic grade (the "front de migmatites" often quoted by Beasairie), the degree of
penetrative deformation, the style of veins, and the silicate and sulphide minerals.
Colvine et al., (1988) has schematically represented these features in a diagram that seems very
well suited to guide field exploration for lode gold deposits in Madagascar. It is given in
sketch 6/4. Other sensible exploration clues on how to proceed with lode-gold deposits of the
Madagascan type are given by Hodgson (1993) in both his paper and extensive bibliography.
6.5. Classification of Madagascar's gold deposits
The latest published classifications of Madagascar's gold occurrences and deposits has been given
by BRGM (1985). For simplicity a very similar classification is maintained here, but with some
adjustments to make it more readily understandable to Anglo-Saxon readers, and to accommodate
IMR's own opinions. Sketch 6/4 summarises schematically these gold deposits.
6.5.1 Primary deposits:
Madagascar's primary gold deposits, with the exception of the baryte veins of the Betsiaka at the
contact Basement and Permo-Triassic sediments in the extreme north-east of the island, all occur in
the Precambrian metamorphic terrains (gneiss or migmatites) as veins of gold-bearing quartz, or
finely disseminated in the various facies of the crystalline schists.
a) Archean primary deposits
These are the most interesting deposits. They occur mostly as quartz veins interstratified into
high-grade metamorphic terrains. There are a total of 105 recorded primary deposits in Madagascar
b) Proterozoic primary deposits
a1 - Associated with amphibolitic series and basic metamorphic rock (some of them could
actually be greenstones). Type examples: Alaotra, Ampasary, Andrianema, Maevantanana. There are
at least 13 recorded deposits of this type, particularly in the Maevantanana, Vohilava-Ampasary
and Beforona gold districts (Table 7, this Database);
a2 - Associated with magnetite-rich quartzites. Type examples: Maevatanana, Antananarivo
region, possibly Andrianema and Beforona-Alatroa. This is a comparatively rare type of deposit
(Table 7, this Database), but of obvious interest for large, low-grade resources utilising modern
(wet magnet separation) techniques. Sometimes these deposits form their secondary equivalent:
auriferous magnetite alluvions (Ankerika, Table 5);
a3 - Associated with the sialic series of the type Ambatolampy-Andriba (quartzites,
gneiss, migmatites, Al-rich micaschists - often graphitic). Type examples: Ambatolampy, Andriba,
Antananarivo-west, the series of Sahantana and Vavantenina, and possibly the series of the
Vohilava-Ampasary gold district. There are at least 34 recorded deposits of this type,
particularly in the Itasy, Ambositra-Antananarivo and Vavatenina gold districts (Table 7);
a4 - To the three types described above one can overlay another type, linked with later
granitic intrusions which locally effect all the above styles of mineralisations. These deposits
can take the form of peri-batholitic veins, stockworks and skarn-type mineralisations. It is
thought (BRGM, 1985) that these intrusion-linked gold remobilisations can constitute the most
favourable environment for gold deposits. There are at least 27 recorded deposits of this type,
particularly in the Vohilava-Ampasary and Andrarona gold districts (Table 7).
These deposits are associated with micaschists or quartzite facies of the
"schist-quartz-limestone" series (S‚rie shisto-quatzo-calcaire) and remobilised either by
regional or contact metamorphism. They generally consist of auriferous sulphides. More than 27
such deposits have been recorded (Table 7). Two type examples are the best known:
c: Mesozoic gold deposits
b1 - Betsiriry (east of Miavindrazo) where gold occurrences are particularly frequent in
the transition zone between the migmatitic gneiss and the epimetamorphic
"schist-quartz-limestone" series (the "front of migmatites" of the French authors).
b2 - The region of Itea where gold occurs in Si-rich skarn around the granitic intrusion
of Itea (see sketch 7/8).
c - Deposits associated with the Permo-Triassic tensional tectonics. These are "real" gold
veins, existing in the quartz-baryte filling of the fractures, with free-milling gold and
associated sulphides. This type of deposit has so far only been found in the Betsiaka gold
district to the extreme north-west of the island. Eleven deposits have been described (Table 7).
6.5.2. Secondary deposits
These deposits form near the meteoric alteration of the primary deposits and the reconcentration
of gold by surface waters, This alteration often tends to lateritise the primary gold-bearing
formations which can reach economic grades.
Aa - Eluvial deposits. Here the secondary material has been transported along slopes,
generally only short distances. Gravity concentration can form locally exploitable deposits even
if the primary gold sources are very low grade. More than 50 such deposits have been recorded.
These deposits often occurs in the same gold districts as deposits Ac and Ab.
Ab - Old alluvial deposits where the gold-bearing alluvions are more or less consolidated,
are in fossil terraces along present valleys. Type example: Ampasary River (sketch 7/12). Several
tens of such deposits have been recorded (Table 7);
Ac - Present day alluvial deposits. The auriferous sand and gravels are the actual river
beds. These have been traditionally the easiest deposits to exploit. More than 80% of
Madagascar's total gold production has originated from these deposits (Besairie, personal
communication 1976). Type example: Ambositra-Antananarivo and Vohilava-Ampasary gold districts.
More than 80 such deposits have been recorded (Table 7). Table 2 list some of the largest
Malagasy rivers with extensive alluvial gold.
L - "In situ" gold-bearing laterites. This type of deposit has been added to the above
classifications by IMR given the increasing importance of these potentially large, low-grade gold
deposits in lateritised Precambrian terrains (i.e. Boddington in W.A. = 428,000 t/y Au). This is
still a speculative type of deposit in Madagascar but of considerable potential. Type example:
Ranomandry, Tainangidina in the Maevatanana gold district.
7. GOLD IN MADAGASCAR: THE FACTS
7.1 History of gold mining
7.1.1 To 1895
Contrary to other east African countries, particularly Ethiopia and Eritrea, gold was not
exploited in Madagascar until comparatively recent times.
The first recorded gold discovery in Madagascar was in 1845 by J.Laborde, more than 50 years
before the French annexation of the island. Laborde, a French adventurer and adviser to the
Malagasy royals, is a legendary figure in Madagascar's history. The discovery took place during
the annual royal hunt in Maevatanana.
Perhaps understandably, gold mining was immediately discouraged by the Malagasy government. The
two Malagasy contemporary legal codes (of the years 1868 & 1881) under queen Ranavalona II
punished any gold digger with 20 years' imprisonment. It was only in 1883 under Queen Ranavalona
III that gold exploitation was permitted, but by the government only.
The government's gold exploitations were initially largely manned by the compulsory corv‚e system
used for public works (noblemen were excepted). Thousands of people were employed. The first
kilogram of gold was produced in 1884 in the Ambositra district, possibly from Itea mine. In 1885
Madagascar's gold production had reached 75 kg (2,411 oz) in the Maevatan…na region alone.
The first non-government concession was given to the French entrepreneur Suberbie in 1886, once
again in the Maevatan…na district.
In 1988 a Mr Rigaud operated, on behalf of the Malagasy government, a series of plants to the
south of the capital Antananarivo: Behenjy, Ambatolampy and the rich alluvions of Androka,
Andraikibo, Hatsara and many other rivers, which are still artisanally exploited today.
The same year Mr Savaron started the exploitation of the Ikopa (12 kg of Au in 34 days). In 1894
there were more than six legal European gold concessionaires in Madagascar, including a Mr
Harrison Smith in the Antsianaka goldfield and Colonel Shervington in Ariovonimano. But the
French annexetion of Madagascar in 1895 discouraged Anglo-Saxon concessionaires.
By 1900 gold production in Madagascar was taking place in at least eight major gold districts
with a robust production of 1,114 kg (35,820 oz) of gold, mostly from the Mananjary district (see
With the exception of the Boer republics (Transvaal and the Orange Free State) this steady and
rapidly increasing gold production was unprecedented in Africa at the time. The perceived mineral
wealth of Madagascar played an unquestionable role (Grandidier G., 1987; Lacroix, 1900; Gallieni,
1901;) in the French government's decision to grant Madagascar "protectorate" status in 1895. One
year after that the protectorate became a colony.
7.1.2 From 1895 to 1960
The French government's high expectations of Madagascar's mineral potential (not only gold but
coal, gems and other materials) seemed justified at first.
The gold production easily doubled to 2460 kg (79,100 oz) in 1904. In 1906 the discovery of the
large and easily winnable deposits of Andavakoera in the extreme north of the island pushed
Madagascar's gold production to around 3 t/y, then on to an all-time high of 3.697 kg (118,874 oz
) in 1909.
By that time Madagascar was by far the largest gold producer in the enormous French colonial
French interest in Malagasy gold is reflected in at least 44 technical publications, some by the
most influential French geologists, on the subject in the period 1900-1910 alone (see Appendix
Most of Madagascar's gold production at the time was coming from rich but "unsustainable"
alluvial and eluvial deposits.
The French colonial authorities were obviously proud of introducing modern and efficient mining
methods in Madagascar, but disappointments quickly followed.
In 1904 a steam dredge was launched on the Tsiribihina river. This dredge never worked properly
and was abandoned near Miandrivazo. Other dredges were tried, without success, on the Mananjary
and Ikopa rivers. Only the plant of Andavakoera with five batteries of five hammers each and
several amalgamation tables functioned for a few years, but with very high costs.
Essentially a Madagascan alluvial field had a life span of 15-20 years before the production
dwindled to few kilograms/year produced by single operators or very small syndicates. Generally
cash crops then took place on the old goldfields.
Like most of colonial Africa production dwindled during and after WW1. In Madagascar production
decreased to less than 1 tonne in 1917 and did not pass that mark again until very recently - and
then in a totally uncontrolled way (see table 1).
In the period between the two World Wars a number of medium or small mines in primary ore were
developed in Madagascar, much as the Portuguese did in Mozambique or the Italians in Eritrea. But
this sector in Madagascar never attained the vitality of, for instance, Northern Rhodesia (now
This could be due to fundamental geological reasons but other causes are also apparent. The
miners seldom tried to extend exploitation below the oxidised, superficial zone. They were also
poorly equipped with plants able to deal with sulphide ore with acceptable recovery. Their
capacity to deal with water problems at depth was also limited. Actually, French underground
mines of the past seldom went deeper than the water table, apparently regardless of the presence
of further ore at depth.
Other causes, besides a constantly low gold price in the period, included the increasingly
profitable graphite mines, the coffee plantations and a variety of new opportunities created by a
very active French colonial administrations.
At independence in 1960 Madagascar's gold production had trickled down to only few kilograms per
7.1.3 From 1960 to the present
In 1966 Mr H Besairie (1966,) ex-chief geologist of the Madagascar Geological Survey, reviewed
and discussed the gold potential of Madagascar. This is the last organic study on the subject.
Besairie estimated that between 1897 and 1959 the total production on the island had been around
51 t (or about 1,800,000 oz).
Region of Tsinjoarivo (Ambositra-Antanananrivo goldfield)
At the time gold prices were at an all-time low, and the Nixon administration's decision to free
up gold prices was still eight years away.
Madagascar's official gold production (probably very close to the real one) languished to around
15-25 kg/year. It is difficult to say when the liberalisation of gold prices in 1971 seriously
stimulated gold production in Madagascar anew, but this probably started in the early '80s like
the rest of the western world.
The U.S. Bureau of Mines (Antonides 1988) estimated that in the year 1987-1988 Malagasy gold
production increased twenty fold. Probably by the year 1990 the Malagasy gold production,
virtually all unofficial, had reached the present estimated levels of 3-4 t/y.
For its part, since 1972 the Madagascar Geological Survey began "Operation Or", carrying out
various surveys in the regions of Masokoamena (East Maevantanana), Tsinjoarivo (East Ambatolampy)
and Dabolava (East Miandrizazo).
Region of Masokoamena:
Prospecting of alluvial and eluvial occurrences of the lower valley of the Kamoro River.
Installation and start-up of a plant exploiting the old alluvions of the Kamoro River. The
exiguity of the reserves, the poor recovery and, presumably other inadequacies in planning, led
to an early closure of the plant.
Prospecting of recent and old alluvions in the Onive River and its tributaries (sketch 7/8),
about 50 km south of Antananarivo. Study done on the Onive River into a meander upstream of
Tsinjorivo (sketch 7/8) in order to exploit the river bed in dry condition. This technique was
used with some success at the beginning of the exploitation of the Itasy goldfield. However, the
results obtained in the 1970s pointed to poor grades and reserves.
Region of Dabolava
Some surface sampling of the primary deposits on the axis Sarobaratra-Andravoravo. This is an
interesting theme and should be re-evaluated if the data is still available.
Prospecting on the old alluvions of the Dabolava River (Betsiriri gold district) with negative
Primary occurrence of Soavinarino
We note that all these activities were largely centred on old, well known goldfields and were
targeting essentially alluvial deposits: hardly a priority target for foreign investors. Very
little effort was devoted to original geological thinking and modern exploration aimed at the
identification of large primary deposits.
Another period of official activity took place in the period 1980-1983 when a Soviet mission,
assisted by personnel of the Geology and Mining Services (Service de la GŠologie, Service de
Mines) carried out some prospecting work. The two targets selected were the alluvial deposits on
the Maha River (West of Mananjary - see sketch 7/10) and on the primary deposit of Soavinarivo
(East of Antisarabe on Ambositra-Antananarivo goldfield).
By far the most interesting of the two targets, this exploration took place around the old
working of Soavinarino. Prospecting was based essentially on geochemistry, panning concentrates
and some geophysics.
Alluvions of the Maha River
These exploration activities were supplemented by a drilling program of six cored drill-holes
down from the old working where spot values of 3-5 g/t Au had been found. However (by admission
of the Malagasy authorities themselves) this drilling was very poorly carried out: four holes
out of the six were vertical (on vertical formations!!!) and probably missed the target.
The core recovery was poor, particularly in the zones presumably mineralised. This seems
consistent with Soviet exploration practices at the time. Also the idea of using surface
geochemistry in an area contaminated by decades of working is highly questionable (IMR's
This exploration program would therefore seem inconclusive.
The exploration took place essentially by geochemical means and panning concentrates over an area
of 10 km2, followed by trenches and pits on the anomalous areas. Only a very modest potential of
330 kg Au has been suggested.
A more serious attempt to re-evaluate Madagascar's gold potential was started by a program of
French bilateral aid. The agreement signed in 1984 by a mixed French-Malagasy commission had as
its objectives the "exploitation of the gold deposits of Madagascar".
The field activities were carried out by the BRGM which has maintained for decades - and still
does in 1996 - an office adjacent to the Geological Survey in Tananarive.
The objectives of this project were rather ambitious and intended to restudy, in a modern
fashion, the gold potential of all Madagascar on a regional scale. By geological and
metallogenetic means various priorities were selected for the 1984 field season:
Antsolobato (West of Antananrivo)
Andravoravo (South West of Ambatolampy)
Betsiaka (Andavakoera) Sketch 7/2, 7/4
Dabolava (East Miandrivazo)
Itea (South West Ambositra) Sketch 7/8/1, 7/9
Further reconnaissances in a rubber dinghy took place on the rivers Ikopa (sketch 7/7), Onive
(sketch 7/8) and Mahajilo to check the suitability of these rivers to a small type of suction
dredge the BRGM was trying to market world wide (sketch 7/1/1).
There is no public record of the results of these works but we note that the BRGM activities
sharply decreased after 1990, and there is not yet any organised or industrial gold exploitation
in Madagascar in spite of the 3-4 t/y of gold produced by "artisanal" methods.
7.1.4 The "artisanal" sector
As in other countries of East Africa - notably Tanzania, Ethiopia and Eritrea - Madagascar has a
thriving unofficial mining sector, locally called "orpillage" or "artisanal" mining. The
workforce employed in this activity varies according to the seasons but it has been realistically
estimated (Drean, 1994) at 150,000 to 200,000 individuals.
Typical Malagasy gold mining practices
This rate of production of around 20 gr Au per man/year is consistent with other African
countries with similar artisanal gold production. It is also consistent with logic. Madagascar's
GNP per individual is only $US230 per year, equal to about 18 gr of gold.
Also, Madagascar's countryside is almost cashless, surviving largely by subsistence farming and
barter. Most of the national income is generated and spent in the cities. The lure for a Malagasy
farmer to earn a reasonable living panning what would be would be a negligible quantity of gold
by western standards is simply irresistible.
Throughout history in the western world an ounce of gold was comparable to the wages of a skilled
labourer over two weeks (Govett, 1982). In most of rural Madagascar an ounce of gold is
equivalent to several months' wages for a skilled worker - provided he can find work at all.
However, IMR disagree with some authors (such as Mutematsaka, 1995) about the sustainability of
artisanal gold mining in Africa. The lesson learned from South America is that small-scale gold
mining is inherently unsustainable. Brazilian gold production, where artisanal mining was the
major source, has decreased from 101 t in 1989 to 75 t in 1994 and is still falling.
The idea of disciplining or controlling in any way this workforce is an illusion. It is
impossible to put a policeman behind any gold miner. Beside the artisanal miners ("orpilleurs"),
by virtue of their number, mobility and cash have traditionally enjoyed a political power way
superior to their cashless peers in the countryside. Madagascar is no exception.
When, in the '90s, gold was discovered at Rabeka, near Ambilobe -the birthplace of Madagascar's
president Prof. A. Zafy - the locality was immediately invaded by 15,000 unruly miners (Drean
The artisanal sector is important to Madagascar.
Several new areas are being worked (sketch 7/1) in regions well outside the historical
goldfields, and this points to the "artisanal" gold sectors energetic, if primitive, prospecting
activities. Some of these new gold prospects are in highly prospective geological settings.
Not all gold activities in Madagascar are uncontrolled. The Mines Service lists about 500
"Permissionaires" (Legal permit holders) and about 20 registered small gold mining companies
(Personal communication by P.J. Rakotoarimanana Chief of the Service of Mining Inspection,
The Malagasy government is trying very hard to control gold mining in Madagascar and several
provisions of the new Mining Code of August 1990 (Law No 90-017) are aimed at Madagascar's small
Some "artisanal" gold practices are unique to Madagascar and no ready English transalation is
available. Here below are some common examples:
"Aloramo" (Malagasy word): the practice of skin diving in auriferous river to recover paydirt.
This practice is particularly used in the Vohilava-Ampasary gold district.
"Lakantany" (Malagasy word): the practice of digging long derivation channels to collect gold
from the hills (when auriferous) during the rainy season. This practice is particularly well
suited to the denuded hilly morphology of Madagascar's highlands. Several "lankantany" are
clearly visible to this day from the highway between Antonanarivo and Antisarabe. At times the
"lakantany" are supplemented with sluices (boites chinoises).
"Lavaka" (Malagasy word): Natural erosion gullies typical of Madagascar's highlands. When cutting
auriferous structures they are significantly enriched in gold.
Pillonage (French word): Crude crushing of ore by hand-held stamps, often truck half-axles. This
method is rather effective in Madagascan conditions where labour is cheap and the primary gold
mineralisation consists of free milling gold in quartz veins.
7.1.5. The Gold Trading Agency (Comptoir de l'Or)
In late 1994 the government decided to form a national Gold Trading Agency (Comptoir de l'Or)
through which the Ministry of Energy and Mining would buy the output of private prospectors.
The Agency will deposit sale proceeds with the Central Bank. This is designed to boost official
revenue at the expense of smuggling.
The full text of the project of Law on the Gold Trading Agency is given in Appendix 3 of this
Database. The Decrete-Law is based on 48 articles and has been signed by the President of the
Republic and six Ministers, including, significantly, the Minister of Police and the Minister of
the Interior and Decentralisation.
It would be tedious and irrelevant to summarise the content of the 14 pages of this document.
Suffice to say that the objectives of the Gold Trading Agency are (Article 1):
a) to buy the gold from licensed prospectors and authorised buyers on the field. Unlicensed
prospectors or gold buyers become, by consequence, illegal;
b) to organise the sale, exportation or importation of all gold in Madagascar in case of
necessity justified by the authorities.
On the field these measures will essentially be policed by newly-created Local Security
It is too early to see how these policies will work. IMR has been involved in similar schemes in
Colombia, Zaire and Tanzania. Essentially the very small miner is persuaded to sell his gold
production to a government agency only when the Government is prepared to pay a substantial
premium on the prices paid by buyer-smugglers perpetually prowling about any artisanal goldfield
(China is a notable exception).
This is tantamount to subsidising an unsustainable and environmentally unsound form of mining.
The Gold Trading Agency will be largely immaterial to any future large-scale miner as,
obviously, the sale of any substantial gold production by a legitimate foreign investor could be
negotiated directly with Madagascar's central bank.
However, the Malagasy government would be well advised to use the inspectors of the Gold Trading
Agency to restrict or prohibit two gold mining practices:
1) The use of hydraulic mining (prohibited in U.S. since 1884), a particularly dangerous practice
in Madagascar's predominantly agricultural society for the irreversible damage it cause to
2) The use of mercury for gold extraction. This mercury is almost never retorted and inevitably
it accumulates in Madagascar's delicate biosphere with deleterious and long-term effects. in
1989, gold-producing Ghana passed a "Mercury Law" (PNDCL217, see also Otto, 1995; p 445) in order
to restrict or prohibit indiscriminate use of mercury.
7.2 MADAGASCAR's GOLD DISTRICTS
To avoid arbitrary new groupings of Madagascar's gold districts throughout the Database the
classification proposed by the BRGM in 1985 has been maintained. This classification comprises 10
distinct goldfields which are described in sketch 7/2.
The relationship of these gold districts with geology and tectonics is given respectively in
Plates 1 and 2 of this Database. Most of the descriptions of the individual deposits have,
however, been taken from various works by Besairie (1964, 1966, 1969) and supplemented by a
number of other publications.
7.2.1. Betsiaka District
Location and history
The district of Betsiaka (or Batsiaka-Andavakoera) is located in the extreme north of Madagascar
about 80 km south of Antsiranana, (formerly Diego Suarez). See sketch 7/13 for location.
Access is easy: by plane to Antsiranana (daily flight by Air Madagascar, Boeing 737, see Appendix
4 for timetable) and then 131 km of surfaced road to Ambilobe. This town (comfortable hotel,
telecommunications, airstrip) is central to the Andavakoera gold district.
The first gold occurrence in this district was discovered in 1906 by Mr Mortages near the village
of Betsiaka (31 km east of Ambilobe by surfaced road). In 1909 production had already reached
1.223 t/y and was the most important for the island (see table 1a). Production decreased rapidly
after the beginning of WWI and slipped below 100 kgs in 1917. The total recorded production of
the Betsiaka gold district has been 1/4 million oz - undoubtedly much larger if we consider the
"unofficial" production still taking place today.
The mineralisation consists of gold disseminated in veins of quartz and baryte (BRGM type c) near
the contact between the Basement and the Paleo-Mesozoic sedimentary sequence of the Diego basin
(see sketch 6/1).
The Basement consists of more or less migmatitic gneiss and micaschists in contact with a series
Triassic pelitic schists
The Isalo continental sandstones.
This sedimentary series is cut by intrusions of microsyenites, essexites, labradorites and
lamprophyres. The quartz-baryte veins occur into the Basement and the Permian formations but not
the Triassic schists. Galena and sphalerite are common accessory minerals in the veins, probably
representing a remobilisation of earlier mineralisations which Besairie (p 269a) estimated at 360
The vein's general strike is 50§ E. with several cross veins. The main vein system is generally
parallel to the tectonic of the area, (sketch 7/4) and indeed of northern Madagascar. Hot springs
(70§ C) are common in the area and their geochemical chracterisation would be unquestionably
important (IMR's opinion).
The gold veins have a thickness of some decimetres but occasionally can reach one or two metres
and strike for several kilometres. The gold-bearing quartz-baryte veins occur over a length of
about 130 km - from Anaborano to Ampisikinana - but they are heavily mineralised only in the
region of Andavakoera over about a 20 km strike. Probably because the presence of galena,
Andavakoera gold is always silver-bearing (electrum) with a finesse in the range of only 700-800.
Contrary to the majority of other Malagasy goldfields, Andavakoera's gold production was
essentially from primary sources. The mineralised quartz was extracted, crudely crushed and the
gold recovered by panning.
Besairie (1966) indicates at least eight "important" underground mines in the Betsiaka gold
district and they were (see sketch 7/4):
- Bereziky (*)
- Sarova (*)
- Tambo (*)
- Raphias (*)
- Ranomafana (*)
Only of five (*) have we some records: see sketches 7/5 and 7/6. From these it would seem that
Betsiaka gold was mainly hosted in basement gneiss and extremely closed to contact with Permian
sandstones. The underground works generally extended for only one of two levels 20-30 m apart,
probably to the water table.
Andavakoera's underground development was an impressive 17 km. From 1919 no rich gold zone was
intersected and 1,440 m of underground development turned out to be totally unproductive. It was
stopped in 1921 and the plant was closed in 1922 for lack of sufficient ore to be treated.
In 1934 an attempt to exploit (sluices) the Komakormalandy River alluvions was attempted but
In 1950 the Malagasy Mining Department (French at the time) carried out a new study of the gold
potential of the region but its results (probably thought to be negative) were not divulged.
Immediately afterwards even the artisanal production ceased, not to be seriously restarted until
the late 1970s.
No obvious attempt had been made to try to mine all the mineralised quartz, trying to exploit a
lower ore grade with larger tonnage. This underlines an uncanny resemblance with what happened in
the Western Australian goldfields at the beginning of the century, long before the gold boom of
the '80s showed that practically all the old "exhausted" mines could be worked economically at
much lower grades than previously thought. It is suggested that the Betsiaka gold district should
be re-evaluated with a similar approach (IMR's opinion).
Table 6 lists 11 recorded deposits for the Betsiaka gold district, all of them hard-rock
The Ranomafana gold vein was exploited by a large trench and two main underground works to about
40 m in depth (sketch 7/5). The mine produced about 1/4 t of gold. The nearby vein of Ambilon
produced about 400 kg of gold to a depth of 60 m.
In 1909 underground work began and a plant with a proper crusher was installed, but with
Apparently the mineralisation was very erratic and generally more consistent along strike rather
than in depth. The exploitable zones are of rather limited dimensions but occasionally some can
reach up to 200 m strike, with depth exceeding 60 m.
In the mineralised zones the gold grades are very irregular - from 0 to 300 g/t with an average
of 20 g/t over a 1,000 t [recorded] production. In some smaller zones the grades have been
extremely high: up to 150 kg for 10 t of quartz treated, or 500 oz/t.
The Ranomafana mine is the only one, for which records are available, that showed the depth of
the mine seemed to been connected with the exhaustion of the high grade ore (and the contact
gneiss/sandstones) rather than the meeting of the water table.
The Bereziky mine produced 800 kg of gold from a single pocket (pocket 1915, see sketch 7/6);
this underlies the extremely erratic grades of thr Betsiaka gold district. Between these very
rich pockets the grade varied between 0 and 3 g/t gold, at the time considered totally
The gold of Bereziky mine is entirely hosted in the Basement micaschists over a strike of
approximately 500 m.
7.2.2 TSARATANANA DISTRICT
Location and history
The gold district of Tsaratan…na is one of the oldest known in Madagascar although its production
statistics are often combined with the adjacent gold province on Maevatanana. The two provinces
can be considered one.
Access can be by car to Antananarivo (a 6 hr drive) or more conveniently by plane to Mahajanga
(direct flight from Antananarivo by Boeing 737, 7 days a week. See Appendix 4) and then by 153 km
of surface road to Ambondronamany (modest accommodation, telecommunications, no licensed
The Tsaratan…na gold district can be coveniently examined from here. A 4WD vehicle is
recommended, particularly in the rainy season (November-April). Conversely a charter plane can be
landed at Tsaratan…na township itself but no reliable vehicle hire or accommodation will be
Gold in the Tsaratan…na district occurs essentially in quartz veinlets. Paragenesis is complex:
tourmaline, barytine, pyrite, and silver, lead, bismuth, molybdenum, copper sulphides. These
gold-bearing quartz veinlets occur within a variety of lithologies: migmatites, leptinites and
above all gneiesses of the Archean Vohibory System, a Ca-Fe-Mg rich complex.
The Tsaratan…na goldfield seems to occur entirely within the Berofona Group (migmatites and
anphibolites), with ages between 1750 and 1890 MA (Pb method). These formations have later been
intruded by granite. Besairie (1964) indicates an age of 350 MA for these intrusions but this
seems very debatable.
Other minerals occurring in the region are obviously pegmatitic ones such as beryl, columbite and
quartz. Other mineralisations in the region are chromite, graphite and lead (economically
exploited underground at Besaky, 16 55" S -47 25" E). A time correlation between the lead and
gold mineralisation of the region is probable (IMR's opinion).
According to the geological map N.4 of Madagascar (1:500,000 Service Geologique) more than 20
gold prospects occur within a 100 km radius from Tsaratan…na village. Some, surprisingly, are
recorded over the sandstones of Isalo II sedimentary formations: these are, of course, alluvial
The gold-bearing veinlets of the primary occurrences strike N 45§ E and dip 35-60§ W. Their
thickness varies between 15 and 90 cm, with grades between 2 and 11 g/t, but occasionally much
higher: up to 10 oz/t at Ambatomisangana.
Table 6 lists 19 recorded deposits for the Tsaratanana gold district, most of them hard-rock
This is a small but very rich deposit located 4 km south of Tsaratan…na and 1 km east of
Ambatomitsangana, on the right shore of the Mahajamba R. The mineralisation is in a very rich
(several ounces/t of Au) of quartz and galena with gold. The vein was only 20 m long.
This deposit occurs 35 km south of Tsaratan…na. It has been prospected by the two small French
companies "Syndacat Lyonnais" and "SociŠtŠ des Mines d'Ambatobe". The mineralisation exists in a
vein of quartz and baryte with pyrite, chalcopyrite, sphaelarite
and occasionally bismuth and molybdenite. The presence of barytine is exceptional in a formation
of Basement schists. The vein is concordant with the foliation of the intruded gneiss.
The main vein, interbedded with the gneiss and migmatites, strikes on 400 m with a width between
15 and 90 cm. The vein's dip is between 35§ and 60§; no direction is given (Bes D, 271a) but it
probably dips to the west.
There is a strong correlation between gold and sulphides. Grades vary between 2 and 11 g/t but it
are much higher in the oxidation zone. This is, on average, 20 m deep. It is this zone that has
been selectively exploited by French miners.
The mine was still active in 1939, particularly in its southern sector, with four levels
developed on 90 m of galleries, adits and cross-cuts. These works met with a gold-bearing,
quartz-sulphide lens developed over more than 300 m strike and plunging in several locations to 100 m
below the outcrops. The vein's thickness ranged between 20 and 60 cm.
Other works at 600 m to the north of the main mine, and particularly an adit of 112 m, have shown
another vein with a thickness of about 60 cm. This and other mineralisations at 550 and 1950 m to
the north graded 14.3 g/t Au with a total of 6,600 t of quartz crushed; but the recovery was
In 1939 Mr Delaitre, a competent French engineer, advised works in order to establish the average
grade. His calculations indicated that a production based on 15,000 t/y would be economic with
gold grades of 7 g/t or above.
As was common for the period the working on Ambohipihaonana started on an abnormally rich
oxidised zone. The mine had an independent hydroelectric power plant which operated crushers and
mills ("bocards"). This plant was destroyed by a fire and the deposit was abandoned.
This deposit occur 40 km south of Tsaratan…na in a quartz-gold vein 0.25 to 0.40 m thick which
has been followed to a depth of 40 m and along a strike of 250 m. Finely disseminated tourmaline
and zircon are accessory minerals. The gold fineness is 858-900. Nearby a second quartz vein shows
molybdenite and numerous inclusions of rutile and zircon.
The rich vein of Ambararatakely is located 30 km south-west of Tsaratan…na on a rolling slope
below Mt Bejofo. The quartz-gold vein is rich in tourmaline and it has been exploited in the
past, like other deposits of the region, by the Syndacat Lyonnais, mainly by monitoring away the
gold-rich topsoil. The area is strongly lateritised, and some modern test work of these laterites
would seem warranted (IMR's opinion).
The deposits occurs 30 km south of Tsaratan…na in several quartz veins. The main orebody is a
subvertical lens with thickness of 0.5 - 0.75 m. and striking at least 30 m. Among the accessory
minerals there are actinolite-tremolite, chalcopyrite, azurite, pyrite, and native bismuth; the
gold fineness ranges between 750-800, possibly indicating the presence of lead-silver sulphides
This prospect is in three quartz veins of which the most important, well interstratified in the
schists, is about one metre thick. Chalcopyrite and baryte are accessory minerals. Apparently,
only one vein has been exploited (Besairie, 1966, p272a).
This prospect occurs 30 km south-west of Tsaratan…na in close stockwork of small quartz veins
only few centimetres thick. In some of them gold is directly visible to the naked eye. This
mineralisation seems of the "dissemination" type.
7.2.3 MAEVATANANA GOLD DISTRICT
Location and history
The Maevatan…na gold district is immediately to the west of Tsaratanana. The town of Maevatan…na
(comfortable hotel, telecommunications, no airport) is central to the goldfield and can be
reached by 330 km of good, surfaced road from Antananarivo (sketch 7/14).
Maevatan…na was the goldfield where the first concession to foreigners was granted by the
Malagasy government: to Mr Suberbie in 1886. The centre of these exploitations was at
Suberbieville, which is now the town of Maevatan…na.
The first systematic exploration of Suberbie's vast concession started in 1888 and was carried
out by competent personnel including several mining engineers.
A dozen small mines were operative by 1891. The first hard rock mine started on a large
concordant vein, with a thickness of 2.5 m with visible gold at Nandrojia, 1 km south of
Maevatan…na. In 1891 a plant was installed with mill and amalgamation tables. Power was provided
by two turbines of 80 CV on the Ikopa River (sketch 7/7).
In 1909 the Nandrojia vein was exploited by pits and cross-cuts but the operation was uneconomic
and was abandoned. The alluvions and the eluvions were then providing the largest part of the
production which reached 658 kgs (21,200 oz) in 1909. Gold production then decreased rapidly,
particularly after the start of WWI.
Maevatan…na's gold production was largely due to the efforts of Mr Suberbie and then by the
"Compagnie Occidentale", later to become the "Compagnie Franco-Malagache". There is no record of
the gold grades worked by these companies.
In 1914 attention was focused on the Ikopa black sands and the gold-bearing laterites. A small
hydroelectric plant was built on the Antafofo waterfalls of the Ikopa (sketch 7/7) and a relative
powerline. At least one work at Betaimby was equipped with monitors.
This large plant had only a limited life, and ended in total failure (Besairie, p 272b). Slightly
later two dredges were brought on the Ikopa, one of them at Antafofo waterfalls (see sketch 7/7).
Not surprisingly for the time, these dredges could not work properly as the Ikopa has very
irregular bedrock. It would be interesting to see how modern small suction dredges, let's say a
Keene with between a 2" and 6" suction head, would preform today in the same region.
In 1921 rumours started about the extreme richness of the laterites of Tainangidina and of
Ranomandry. A sober Mines Department Service downplayed expectations and carried out a survey,
which although insufficient by modern standards, did not show serious resources (at the time; but
in view of the successful lateritic gold deposit of Boddington in Western Australia a
re-evaluation of Maevantanana laterites is well warranted - IMR's opinion).
From 1929 the "Compagnie Franco-Malgache" devoted itself exclusively to the artisanal mining
(orpaillage). Before WWII seasonal dams were built to dry various sections of the Ikopa over a
length of 30 km which assured a good production. At least seven such barrages are recorded
(sketch 7/7). These workings started anew in 1949 at Bekaprija but were again abandoned in 1951.
Maevatan…na goldfield is underlain by the systems of Graphite and Vohibory. The gold originates
essentially from magnetiferous quartzites. The most productive district has been defined by the
polygone between the Ikopa R. the Menavava R. and the Betsiboka R. (see sketch 7/7), or an area
of about 7,500 km2, where at least 44 gold occurrences have been recorded (Besairie, 1966). In
this area gold was exploited mainly in alluvions but also by crushing the exposed quartz veins at
their contact with laterite.
Alluvial exploitations had been the most economic, but any attempt to mechanize them failed and
only "artisanal" mining managed to survive.
In the case of the "black sands" the gold occurs as inclusions into the magnetite, showing the
importance of magnetite-rich sandstones as the primary source of the alluvial deposits. As a
consequence, the richest alluvial basins are the ones where the exposures on magnetite-rich
sandstones are most abundant: the Ikopa Basin with its tributaries Nandrojia, Belambo, Mandraty,
and Firingalava, and the basin of Kamoro with Tsiandrarafa, Aneriga and Bejofo.
The alluvial terraces, especially in the Ikopa basin, were never considered important as they had
grades inadequate in relation with the thickness of the overburden. These terraces were seldom
exploited, and then only by pits. A modern re-evaluation taking in account present economic ore
grades and modern, low-cost stripping techniques coupled with efficient magnetic separation,
seems warranted for the Maevatan…na gold district. (IMR's opinion)
We note that Maevatanana is one of the districts where amalgamation was most extensivily used;
most probably without any retorting and recovering of the mercury. A study of the present
environmental conditions of the district would seem highly desirable.
Table 6 lists no less than 43 recorded deposits for the Maevatanana gold district, most of them
Below are listed some of the more interesting gold occurrences of the district.
This prospect occurs 50 km south of Maevatan…na and 2 km east of the Ikopa River. It is a very
old prospect, exploited since 1888, and which consists of numerous veinlets of gold-bearing
quartz distributed in small pockets along the same plan ("en chapelet") in the lateritised
Starting in 1914 these laterites were worked by monitors and sluices. Gold was recovered by
amalgamation. A hydroelectric plant was installed at Antafofo on the Ikopa River utilising a fall
of 17 m and generating 200 CV and a 7 km powerline to join it to Betaimby. The failure of this
venture has been total: 20,000 m3 of laterite produced only 1,146 gr Au, or a grade of 0.06
gr/m3. Available records do not mention which type of plant was used nor what gold recovery has
Only 1 km south of Maevatan…na, this prospect was the first hard-rock exploitation by Mr
Suberbie. A conformable quartz vein 2.5 metres thick was worked at the surface over 40 m and then
exploited underground in pits and cross-cuts. The quartz was milled. These works became
uneconomic in 1910 and were stopped the same year.
In the Ranomandry sector, 30 km south of Maevatan…na, several gold prospects occur between the
villages of Bejofo and Betanimanga. It has been suggested that these prospects consist in a
single, continuous vein; this is debatable.
There was a reconnaissance in 1928 over the old superficial works of Betanimanga. The 73 samples
collected gave an average grade of 18.6 g/t Au. An amalgamation test over two tonnes of quartz
has given a grade of 29.4 g/t Au. In 1928 the underground workings consisted of: an incline of
34 m, a cross-cut of 80 m and two galleries of 50 m. These workings have not been followed.
Much has been said about this deposit, situated 25 km south-east of Maevatan…na (sketch 7/7),
both for its vein and laterite potential.
The first study was carried out in 1915; others in 1923 and 1927. The deposit occurs in gently
rolling hills and consists of eight gold-bearing quartz veins concordant with the gneiss of the
Vohibory system. The eight veins cover a width of 200 m. A ninth vein is about 100 m away from
the main system. The average thickness of the veins is 30 cm and their length is between 50 and
150 m with the exception of one that strikes 650 m.
The average ore grade worked was around 18 g/t Au. All these veins, dipping 60§, are apparently
exploitable to a depth 100-200 m and one of them to 300 m, but this figure seems implausible
(IMR's opinion). A cross-cut 320 m long intersects the vein swarm.
On this basis a tonnage of 200,000 tonnes of quartz would be equivalent to almost 4 tonnes of
gold (Besairie, 1966 p273b).
However, these calculations result from rather risky extrapolations without systematic sampling
and no proof that the ore grades extend at depth. In other words, the prospect has been
considered a proper single vein deposit rather than a system of veins with large variation in
grades and continuity.
Resource estimates of the Tainangidina hill's laterites appear to be limited, contrary to
previous hopes. Also, the grades are rather low with the exception of the vicinity of the
gold-bearing quartz veins. All organised work on Tainangidina stopped in 1928.
The Ikopa river bed is unquestionably gold-rich but its exploitation is difficult. The river is
deep and the current strong. The river bed is irregular with boulders, channels and scours. An
attempt with two dredges in 1917 ended up in failure. Particularly galling was the fact that the
rainy season invariably destroyed the river dams and that work was possible only in the dry
Between 1930 and 1944 certain sectors of the Ikopa, Mandarti and Belambo rivers were dried by
means of dams and derivation channels. This technique permitted the exploitation of the dried
sectors for 2-3 months a year. The same barrages were built anew after one or two years and they
provided about 50 kg/year of gold. Sketch 7/7 shows some of the largest dams built on the Ikopa.
The last effort on the Ikopa River was made at Bekapirijy where a small rocky island favoured the
construction of a dam. This was followed by digging, on the western shore, a long derivation
channel and the exposure of the river-bed for a significant distance. This project, possibly the
largest ever established on the Ikopa River, worked intermittently between 1949 and 1951 with a
meagre recorded production of 7.1 kg of gold. No further barrages have been built on the Ikopa
although some other attempts have been made with suction dredges in recent times.
7.2.4 BETSIRIRY GOLD DISTRICT
Location and History
The Betsiriry gold district is located to the extreme west of Madagascar's crystalline Basement
(sketch 7/14) between 19§ and 20§ latitude south and 45§ to 30' longitude.
The town of Miandrivazo (airport, telecommunications, small hotel) can be reached by 250 km of
good, surfaced road from Antsirabe or directly by plane from Antananarivo (sketch 7/14).
This gold district produced 2.5 t (80,000 oz) of gold between 1901 and 1921. The greater
production took place between 1903-1914 with a peak production of 205 kg of gold in 1910.
As usual, attempts to dredge the major rivers of the district ended in failure. In particular, an
English company built a plant which collapsed before reaching the Mahajilo Falls where it was
supposed to operate.
The BRGM indicates this goldfield as being Proterozoic in age and so as not to complicate the
issue this has been maintained in sketch 7/2 of the present report.
In reality this assumption does not seem adequately supported. Besairie (1964, Feuille du Centre)
put these gold mineralisations firmly in the Amborompotsy Group of the Vohibory system. Hottin
(1972) proposes an age between 1,700 and 2,000 MA for these lithologies. On the basis of
available information the Betsitiry gold district does not look significantly different to the
great majority of the Madagascar's other gold mineralisations which are considered of Archean
The only distinctive character of the Betsiriry gold district is that the known gold
mineralisations appear to be located in the "migmatitic front" or in the passage zone between the
gneiss and the epimetamorphic series "schist-quartz-limestone".
In a radius of 50 km from Miandrivazo at least 14 gold occurrences are reported (Besairie, 1969,
Feuille 5). These prospects occur in gneiss and migmatites and they appear unrelated to later
The mineralisation is in free milling quartz often associated with magnetite and iron and copper
sulphides. In some deposits (Ankaboka) part of the gold is syncrystallized with the magnetite. In
others (Antasaily) gold is associated with mispickel.
The Betsiriry gold district may be considerably larger than indicated in past reports. We note
several gold prospects: Ankofoka, Antsiroroka, Antsevakely, Miandrarivo, almost 100 km north-east
of Miandrivazo district in what is essentially the same geological environment.
There is evidence that recent prospecting by "artisanal" miners has further extended the
boundaries of the Betsiriry gold district (sketch 7/1).
Table 6 lists only eight recorded deposits for the area is still little known. Significantly, all
of them are primary.
Individual gold deposits
Among the most important deposits of the Betsiriry goldfield is the Dabolova Group, 30 km
southeast of Miandrizazo, which has been in continual exploitation from 1885 to at least 1966
The quartz vein of Takodara I striking 2 km has been exploited over all its length. The recorded
grade varied between 10 and 50 gr/t Au. At Ambohipisaka, there are there are four main veins,
very irregular and in separate, lens-shaped gold-quartz occurrences ("en chapelet").
An exploitation of Dabolova was attempted between 1937 and 1939 with crushers, mills and
amalgamation tables. In 1938, 200 t of quartz yielded 2.8 kg of gold indicating a minimum grade
of about 14 g/t Au. We note that the plant was not equipped for the treatment of auriferous
pyrites. This may account for the low gold recovery (IMR's opinion).
Other prospects of Betsiriry goldfield still artisanally worked in 1966 were Analabe,
Ankarongana, Kiranomena, Antasaily and Andisiavo.
7.2.5. ITASY GOLD DISTRICT
Location and History
The gold district of Itasy-Tsiroanomandidy is located 50-150 km west of the capital,
Antananarivo, and produced 1,260 kg of gold between 1901 and 1921. Numerous vein deposits were
worked and rapidly abandoned. Its peak production was in 1904 with 174 kg produced, largely from
the Vohinambo deposit.
The Itasy gold district is a rather diffuse one, without a distinct geographical expression. The
village of Analavory (telecommunications, small hotel, emergency landing strip) can be reached
from Antananarivo by 124 km of good surfaced road (sketch 7/13). From this village most of the
gold prospects of the Itasy can be visited. The use of a 4WD vehicle is advisable.
There is relatively little geological information on the Itasy gold district and practically no
deposits marked even on the 1:500,000 geological maps. The district can be seen as the western
extension of the Ambositra-Antananarivo gold district.
The gold occurrences occur in the migmatites, gneiss (often garnet and sillimanite bearing),
quartzites and micaschists. All the gold prospects occur in the graphite-rich zones of the
Archean Graphite System. Granitic intrusions are frequent in the region but are unfortunately
undated. Some gold deposits seem controlled by the strong N-S and E-W tectonics of the area (see
Plate 2 in the present Database).
The quaternary basalts that form the Itasy massif itself are unrelated to the gold
Throughout the district there are several mentions of auriferous
quartzites with magnetite: i.e. Ambohimanoa, Ambohitsivalana, Behana. These are potentially large
if low-grade deposits very amenable to modern extraction techniques, particularly if the gold
occurs (as has been indicated) as inclusions in the magnetite. IMR is of the opinion that this is
well worth investigating.
Besairie (1966) lists 13 gold deposits occurring in a large arc between Ambatolampy and the Itasy
(see also Table 6); all primary. These deposits are described below:
The prospect is located at only 9 km to the south of Tananarive. It is a very old prospect in
several small veins of gold-bearing quartz in migmatites, gneiss and granites. The prospect has
been worked by a large trench producing between 1-2 kg gold per year.
This deposit is located 10 km south-west of Arivonimamo. The deposit was discovered by Mr Talbot
in 1904 who extracted 70 kg of gold in a few months. Once again the deposit consisted in series
of gold-bearing quartz veinlets of modest thickness interstratified into the gneiss. The
mineralisation is enriched at the contact with the overlying laterites. The occurrence has been
exploited on a trench of 80 m strike and 25 m depth and then by underground works. These works,
poorly executed, caved in and the deposit was abandoned.
20 km north-west of Tananarivo this deposit consists of a large quartzite lens in
sillimanite-rich lithologies. No quartz-gold vein as such has been recognised. Basairie considers
this mineralisation of a "disseminated" type; it certainly looks anomalous compared to the
prevalent style of gold deposits in Madagascar.
A sampling program, presumably carried out in the 1960s on the old working and consisting of 60 m
of channel sampling, produced five samples out of eight with grades between 1,5-2 g/t (by
amalgamation). The crushing of the other samples to 100 mesh did not indicate gold. Along the
hillsides quartzites have long been worked in the past by long hand-dug channels ("lakantany" in
Malagasy) which drained auriferous eluvions during the rainy season. On the hillside there is an
alluvial terrace which, in the past, has been exploited in its richest parts.
It is obvious that Ambohimanoa has been worked only with very crude artisanal methods; given the
mentioned "disseminated" nature of gold deposits in sedimentary formations a modern
quantification of the Ambohinamonoa prospect is well warranted (IMR's opinion).
This prospect occurs 20 km south of Arivonimamo village. Its old working date is around 1912. A
gully-dissected hill is the likely source of the gold in the Antosalobato creek. A bed of
mica-gneiss about 15 m thick hosts abundant quartz-gold veinlets concordant with the gneiss
foliation and dipping 60 degrees N. The largest vein, about 1.2 m thick, showed visible gold over
10-15 cm. Pyrite is also present. Underground works have failed to find extensions of this
This prospect occurs 50 km to the west of the Tsironomandidy village. Its placers were recognised
at the start of the century. Two hills, Ankafotra and, above all, Betanimena, show a substratum
of auriferous gneiss. Where these have been lateritised they formed local concentrations of very
fine gold, with occasional outcrops of auriferous quartz veinlets. In the alluvions of the
Ankafotra and Bekopaka rivers gold can occur as pepites (suggesting a goldfield where gold
detectors could be usefully employed by individual miners). The forested areas of these alluvions
have not been worked.
This prospect, located 40 km south-south-east of Tananarive, consists of altered and lateritised
gneiss with veinlets of auriferous quartz. The occurrence has been prospected over a 2 km strike
with trenches of various depth. These works are very old and probably date to the end of the last
century. Underground works were carried out in 1910-12; there is no record of the results.
A prospect 35 km to the east of Tsiroanomandidy village. The placers were recognised at the
beginning of this century.
Some early exploitation produced about 50 kg Au before 1895. Two quartz veins 40-80 cm thick
strike over 200 m and have been literally riddled with shafts of 12-15 m depth.
The quartz was artisanally crushed ("pillonn‚") and washed. The auriferous veins seem to
disappear immediately outside the worked zone. Underground works have not been successful. Nearby
there is quartzite with weakly auriferous magnetite which has formed low-grade alluvial deposits.
Gold recovery has been attempted with shaking tables. No mention of modern exploitation by wet
magnetic separator (IMR's observation).
A prospect 65 km east of Tsiroanomandidy. There is an old alluvial exploitation in a zone with
magnetiferous quartzite and amphibolites.
This prospect lies 52 km to the north-west of Tsiroanomandidy village and was exploited even
before 1895. Two placers of 1,800 x 800 m and 1,500 x 100 m produced 387 kg of gold between 1903
and 1909. Production then fell rapidly. The prospect occurs around the Betanimena hill which
consists of amphibolites, gneiss, micaschist and, once again, magnetiferous quartzite.
The amphibolites, largely formed by the cummingtonite-grunerite series, enclose a layer of
kyanite (disthene) with visible gold. The magnetiferous quartzites are auriferous and at times
show visible gold. Both eluvions and alluvions have been long exploited. Some essays by sluice
between 1909 and 1910 have been rather unsatisfactory with yields of about 1 kg Au a month. The
exploitation mainly took place by the Malagasy method of "lakantay".
7.2.6. AMBOSITRA-ANTANANARIVO GOLD DISTRICT
Location and history
The Ambositra-Antananarivo gold district is particularly developed in the province of Antsirabe
and officially produced 1,489 kg (48,000 oz.) between 1901 and 1921. The largest part of this
production came from the region south of the Onive River (sketch 7/8). Some deposits occurred
around Ambatolampy and in the Faratsilho sector.
Gold production was particularly important from 1890, when the exploitation was carried out by
administrative corve‚s. Around 1900 these corve‚s employed more than 2,000 workers in the
Sarobaratra River alone. The inevitable exhaustion of the alluvial resources took place
immediately afterwards and was a powerful incentive to the first exploitation of primary gold,
particularly where their contacts with the overlying laterites formed zones of strong enrichment.
This gave the district a new lease of life with a high production between 1904 and 1911. At that
time numerous exploitation of graphite were developed and attracted labour away from the less
lucrative gold workings. Production has been totally artisanal ever since.
It would be futile to list all the deposits that occur in the Ambositra-Antananarivo goldfield.
In any case records are very scarce, and they are mainly known by traces on ancient workings
(Besairie, 1966, p 276).
The heart of this goldfield can easily be reached by 100-180 km of excellent surfaced road from
Tananarive (sketch 7/15); most individual deposits can be reached only by 4WD vehicles,
particularly during the rainy season. The town of Antsirabe offers very good and pleasant
Once again the gold mineralisation of Ambositra-Antananarivo is hosted in migmatites, gneiss and
micaschists of the Ambatolampy group, Graphite System (see Plate 1), containing quartz-gold
veinlets with secondary alluvions and eluvions.
The largest part of these deposits is within the zones of micaschists or micaschist-gneiss but
they appear little affected by the migmatisation. On a regional scale gold seems to be
distributed on the migmatisation front.
Leucogranitic intrusions are common in the area but their relationship to the gold
mineralisation, if any, is unknown.
The regional tectonic is essentially N-S and E-W and some deposits appear fault controlled.
Sketch 7/8 shows some of the most notable workings of the south Onive district over an area of
approximately 700 km2. Besairie (1966) mentions at least five underground workings (Analasarotra,
Andranofito I and II, Ankanona, Miakatso), but there have certainly been many more.
The primary deposits of Andranofito, Tongarivo and Soavinarivo consist of concordant quartz
veins, locally rich and rapidly lost at depth where they reportedly had uneconomic grades (but at
the time, grades below 10 g/t in hard rock were considered uneconomic). Below are given some
details of the last deposits to be worked in the region.
Table 6 lists 39 recorded deposits for the Ambositra-Antananarivo gold district. About half of
them are primary.
Individual deposits (Antisarabe region)
Gold was exploited on the Onive River in the dry season, in particular downstream of the
Tsinjoarivo. Some projects, utilising sand-pumps, had been planned but were never carried out. An
attempt to dredge downstream of Antanifotsy failed.
The mount Sarobatra (1,867 m a.s.l.) is the highest point of a topographic range 25 km long
which terminates at the Onive River, north west of Tsinjoarivo village (Sketch 7/8). The gold
workings, very old, were developed on the Sarobatra River and its tributaries: the creeks Analava
and Anteritery. The alluvions were very well developed with thicknesses of between 2 and 4
The grade was initially rather high, reaching 4 - 5 gr/m3. These alluvions extended about 200 m.
The entire length of the wide valley of Sarobatra has been worked. It was the best alluvial
deposit in the region.
The latest alluvial exploitations were developed on Miakatso and Besorohitra creeks to the east
of Antofimbato; Ambodiala, Laptsihazo, Antanetibe and Saovinandriana to the east of the range;
the creeks Analalava and Anterivody to the south-west and Antasahondra to the south. The last
works took place at Ambohimanga (1,500 m north-east of Sarobatra) and Antanetibe (3 km to the
south-east). In 1969 there still was production from Antsofimbato, described below.
This prospect lies on the north-east extension of Sarobatra; the old workings of Antsofimbato in
the 1950s attracted attention because the continuity of its production: more than 1 kg gold a
year. Here is the synopsis by Besairie (1966, p 276b) of some investigation carried out in 1961:
"The formation worked consists of a bed of heavily lateritised micaschists, about 20m thick,
resting with a dip of 45§ over the quartzite which is the backbone of the mountain range. These
micaschists enclose several concordant quartz veinlets striking about 100 m and mineralised over
a thickness of 50 cm.
These veinlets are not continuous but rather in separate little lens-like pockets ('en chapelet'=
rosary-like) over a width of about 10 m in the micaschists. Gold is sometime visible to the naked
eye but more often is only visible after crushing and washing the quartz.
The micaschists contain traces of gold, very fine and very dispersed, with no [apparent] economic
grades. The rich zone has been exploited by trenches and some minor underground working. A 1960s
study by BRGM concluded that the grades were insufficient for mechanised exploitation of the
This abandoned gold prospect is located 30 km south-east of Ambatolampy in a series of graphitic
micashists. The prospect was examined in the 1960s by BRGM personnel (Besairie, 1966, p 277a).
The principal bed, exploited over 1 km, consists of a level of micaschists very rich in
sillimanite (local name = vatodidy) with an average thickness of 50 cm, intercalated between
quartzites on the footwall and a graphite-bearing layer on the hanging wall.
The deposit has been entirely exploited in its eluvial and alluvial parts and also in the
sillimanite-rich part more readily accessible by trenches. Channel samples in the
sillimanite-rich part of this prospect have indicated gold grades from 1 to 5 gr/t. We ignore
the analytical techniques for these determinations: probably nothing more that hand-crushing the
quartz followed by washing it with a gold-pan (IMR's opinion).
This gold prospect occurs 24 km west and slightly south of Tsinjoarivo. Its exploitation started
in 1889 by Mr Savaron - then one of the first legal French exploitations in Madagascar.
All the creeks running from the hills were more or less auriferous. To its south-west the deposit
was worked at the beginning of the century by Anasaha. Thick, bluish, quartz veins concordant
with the mica schists' foliation were the gold source: the gold generally being present in
association with pyrite. Unfortunately these veins were very discontinuous. The Belanitra,
Antkitsika and the Andranomiady rivers have been worked for alluvial gold.
This gold prospect occurs 9 km south of the Ambatomiady village and it is important both for its
alluvions and its underground workings - at least two of them (see sketch 7/8). Once again the
gold occurs in numerous quartz veins in the micaschists. The most important workings are around
the topographic peak of Andranofito with over 379 m of underground working. But the main vein,
with visible gold, was worked in a very rudimentary way, with crowbars.
A prospect 31 km ESE of Antisarabe. This deposit was the first to be worked underground in the
region. The prospect consists of a vein of gold-bearing pyritic quartz which has been exploited
by superficial workings over a length of 300 m.
A 25 m shaft was sunk to exploit the prospect at depth but was abandoned because of water
difficulties. The workings were abandoned in 1911. There are no records about ore grades. As the
ore was very rich in pyrite, a small plant for the crushing and washing of the pyrites was
installed. After the first recovery of the gold by panning, the residual pyrite was chemically
treated by the Chemical Laboratory of [Antananarivo's] Mines Service.
A gold prospect 12 km east of Antisarabe. Three quartz veins were worked both om the surface and
underground. Once again, all activities ceased in 1911.
A gold prospect about 20 km north-east of Faratsiho. There are important underground works in the
sectors of Faravoitra and above all at Antanifotsy. The gold is found as inclusions in the gneiss
and these have been exploited particularly along alteration [lateratization] zones. But the main
production came from alluvial deposits.
A gold prospect 45 km WSW of Antisirabe, on the upper course of the Isakelo River: a tributary of
the Mania River. The ancient workings extended over several km in the valley. A quartz vein with
visible gold has been totally mined out.
Individual deposits (Ambositra region)
A gold prospect 18 km south of Ambositra in the upper valley of the Ivato River, the prospect
showing important old workings. Two large trenches, one 400 m in length (probably visible in
Satellite images - note of IMR) were made to exploit an auriferous level, now partially covered
by slides, in the migmatites: probably the gold mineralisation was hosted in quartz veinlets as
this is the regional pattern. This occurrence has given place to important eluvial deposits and
to a mineralised [fluvial] terrace now totally exploited.
Itea is a rather different deposit from most of the region. It occurs 50 km to the south-east of
Ambositra (sketch 7/8/1). Its position is well to the west of the main Ambositra-Antananarivo
goldfield. Also, while the majority of the gold occurrences of this goldfield are in the
micaschists and the graphitic gneiss of the Ambatolampy Group of the "Graphite System", Itea gold
prospect is apparently hosted in the granite of Vavavato-Antofinandrahana which Besairie (1964)
estimates to be 700-1100 MA in age.
The Itea area is heavily mineralised with copper, lead and bastnaesite. Pegmatites are also
frequent in the area (quartz, beryl) and a few have been mined (i.e. Tambolehibe). The strike of
the pegmatites and some major faults in the region is N 40§E. It would seem that this fault has
at least partially controlled the Itea prospect.
The mineralisation apparently consists of gold impregnations into the micaschist and the
quartzite but also some alluvions. Itea's primary mineralisation was about 7 m thick and has been
worked underground, probably not below the water table, at about 22 m from the surface.
BRGM investigated Itea in the 1960s but the works were not apparently followed up and very little
is presently known of this prospect beside the poorly-drawn section given in sketch 7/9.
Given the anomalous nature of Itea's mineralisation a detailed and modern investigation of this
deposit seems well warranted.
7.2.7. VOHILAVA-AMPASARY GOLD FIELD (MANANJARY)
Location and History
This goldfield is often referred to in literature (ie. Besairie, 1966) as the Mananjary
goldfield. It has been one of the major producers in Madagascar with a recorded production of
8,580 t (276,800 oz) of gold between 1901 and 1921: an impressive performance.
There are no records of this goldfield producing before 1901 but at that date it was producing in
excess of 500 kg/y (16,100 oz) or about half of all Malagasy production at the time. Misguided
French investors and administrators (Gallieni, 1990) were at the time pinpointing this goldfield
as a potential rival of Witwatersrand in Transvaal. An entirely different geology of the two
goldfields rapidly disproved these wild expectations.
The gold exploitation of this goldfield was particularly intense until World War I with a peak of
744 kg/y of gold in 1903. It decreased sharply immediately after, particularly when coffee
plantations over the old gold alluvions proved more economic that artisanal mining.
The goldfield occurs around the pleasant coastal city of Mananjary, 350 km south of the main port
of Toamasina (ex Tamatave) on the Madagascar east coast - see sketch 7/15. Mananjary
(telecommunications, airport, hotel) can be conveniently reached from Antananarivo by commercial
flights with Air Madagascar's "twin otter" planes.
The Vohilava-Ampasary gold district is underlined by the gneiss and migmatites of the Mananjary
Group of the Vohibory system (Besairie 1964) and dated at 2,150 MA (Pb/Pb method). This group is
equivalent to the Ambatolampy Group in central Madagascar. Intrusive in the Mananjary migmatites
is the biotitic-monzonitic granite of Befody (see sketch 7/10).
In this region of Madagascar the Vohibory system is rather different from the rest of the island.
It can be split up into three sections: the Maha group in the east (sericite and chlorite-rich
schistes affected by a tectonic zones with tight folds, faults and mylonite zones), the Vohilava
group in the centre (mica-schists and geisses), and to the the west the Ifinadiana group
(closely-folded migmatites). Throughout the region there are numerous, small intrusions of
The impression that at least some of the Vohilava-Ampasary gold occurrences can be hosted in what
an Anglo-Saxon geologist would call greenstone belts cannot be be dismissed. Hottin (1972) seems
to suggest this possibility.
Geochronological determinations have given an age of 2,140 MA for gold-sulphide mineralisations
(Andrambo) and two possibilities close to 900 MA for a second mineralisation cycle: 890 MA for
the zircons of the Namolika granite to the extreme south of Befody massif (Pb/Pb), and 850 MA for
the alluvial monazite of Mananjary (Pb/Pb) (Besairie, 1966).
It would therefore seem that an age of around 850 MA can be safely suggested for the Befody
granite: this is in good accord with the general period of granitisation and charnockitisation of
Madagascar and peninsular India (See Plate 1 and sketch 6/3).
Sketch 7/10 shows the distribution of the known gold deposits of the Vohilava-Ampasary district.
We note to the west of Befody granite the auriferous zone Ampasary-Sakaleona, and to the east the
zone "de la Petite Falaise", and still further east the region Saka-Maha-Sahandrambo -
The primary gold mineralisations are geologically controlled: they are all below the magmatitic
front. In Ampasary-Sakaleona region the gold mineralisation is localised in quartz veins at the
contact between gneiss and the lenses of ultrabasics included in the gneiss. On the "Petite
Falaise", the gold occurrences are localised at the contact between granite and crystalline
schists. The auriferous quartz veins occur in both lithologies but generally at less than 100 m
from the contact. Significantly, all the creeks running across this contact are more or less
In the case of the Maha and Sahandrambo rivers the gold deposits are over large fault zones and
it is plausible that there is a correlation between these fault zones and gold mineralisations,
both primary and secondary. Some primary gold deposits occur as far as 30 km away from the known
surface expression of the Befody granite (sketch 7/10).
One of the characteristics of the region is the abundance of quartz and sulphide veins. Their
thickness varies from 0.2 to 1 m, and it is generally largest in the faulted zones. The
mineralisation consists of gold, pyrite and mispickel [arsenopyrite] commons.
Chalcopyrite and galena are rare and in variable proportions. Tourmaline and muscovite are
common. Some pegmatites are gold-bearing although this can probably be dismissed as an
unimportant remobilisation episode during the Pan-African event (480-550 MA. A major proportion
of the gold occurs as inclusion in sulphides.
From about 100 samples collected in a study carried out in 1952, all proved to be auriferous but
with grades at or below 0.5 g/t Au (Besairie, 1966); but the gold seems to be ubiquitous.
The regional morphology is important in the control of the placers. In the sharply cut valley the
rivers tend to have a torrential regimen. The alluvions are therefore very irregular and the gold
distribution is controlled mostly by the bedrock's shape and nature.
The evolution of the relief has given place to several ancient terraces, the oldest only 20 m
below the thalwegs. The surface of the alluvions is, in the majority of cases, smaller than 10
hectares and it reaches 50 hectares in very few cases. Larger deposits do exist along the Saka
and Maha rivers but their exploitation proved to be difficult.
In the region there is a secondary hydrographic network formed by small water courses of modest
importance but with frequent alluvial terraces. The placers of these tributary rivers have been
practically exhausted. The placers of the main river system have been exploited wherever
accessible by a shovel. The old river terraces, more or less at water table height during the dry
season, are covered by fine alluvions 5-8 m thick. These terraces have rarely been exploited.
Furthermore, they are now overgrown by coffee plantations.
In most of the auriferous alluvions of the Vohilava-Ampasary goldfield the gold-bearing pebble
horizon is covered by 3-5 m of fine alluvions and lateritic clays. But, in spite of this,
overburden exploitation is easy if water is readily available. Most of the richest terraces have
already been exploited.
The grades in the alluvions range from 1 to 2 g/m3 but exceptionally can reach 3 - 4 g/m3. The
thickness of the auriferous conglomerate varies between 0.5 and 2 m. The average grade for the
terraces still unexploited (1966) is around 1 g/m3 or less.
The gold recovered is generally coarse. Pepites are rare and of small dimensions. The gold
fineness never goes below 945 and occasionally reaches 995.
In the following chapter we will describe several gold deposits, mostly, but not all, from
Besairie (1966). Only a few of these deposits have been studied in any detail. It is these
mineralisations that - regardless of their grades - fed the secondary deposits, either alluvial
or eluvial, and these have produced most of the gold in the region.
Table 6 lists a record number of 58 known deposits for the Vohilava-Ampasary gold district.
Significantly, 27 of them are primary.
The small area of the Ampasary River (see sketches 7/10 and 7/11), 70 km north-west of
Mananjary, is the second biggest gold producing area in Madagascar, immediately after the
Andavakoera. The Ampasary course is sinuous and interrupted by numerous rapids. Important
auriferous terraces are developed around it in major meanders (sketch 7/12). On the east of the
river, on the hilly slopes, several prospections for hard-rock gold have been carried out with
success. But obviously this has been an under-explored target.
The richest zone starts to the south with the placers of Hanning and Betampona which have
produced respectively 119 and 132 kg of gold. More to the north Malazamasina and Alexander have
produced respectively 55 and 195 kg of gold: some terraces and flats in this area have not been
exploited due to heavy forestation.
Further north there are the placers of Antanbao which have produced 367 kg of gold: this section
of the river consists of a succession of rapids and deep pans where the exploitation was carried
out by diving ("alorano", in local parlance). A series of alluvial terraces occur on both sides
of the river: the richest have already been exploited.
Given the chaotic way of the old workings is difficult to estimate how much gold could still be
won from the Ampasary River but several meanders are certainly still very rich. The Ampasary
River ends with the sectors of d'Ambalavia and Ambodiara which produced 125 kg of gold. The
auriferous zone extends to the north with the basins of the Sakaelona.
Several primary gold deposits are known on the eastern shore of the Ampasary. They have been
exploited only by very primitive means, without the use of compressors or explosives and they
have been quickly abandoned at the first difficulty. Among these primary deposits we can quote
Ambalakaza-north, Tobilava, Antanasy and Antanambao-east and Ambalavia-Ambatomainty.
All these works were in gold and sulphide-bearing quartz veins in amphibolite and tremolite
schists. It would be interesting but probably near impossible to locate these old workings after
more than 80 years. The region and the style of mineralisation is rather favourable to regional
exploration by modern geochemistry: initially by stream sampling followed by grid soil sampling
on anomalous zones (IMR's opinion).
The Sakaleona River and its tributaries (Sahanana, Andranomanjaka, Tsaravinany, Nandravaona,
Tembinah and Sahakoriana (Besairie, 1966 p 279b) have yielded more than 500 kg of gold. See
sketch 7/10 for a general view of the area. Obviously, the geological setting is very similar to
the Ampasary goldfield and all gold mineralisations originate from the contact between the Befody
granite and the enclosing gneiss and micaschists.
La Petite Falaise
This series of deposits occurs along the eastern contact of Befody granite. Some individual
deposits are described below.
At Bebasy, 12 km south of Soavina village, a vein in the sericite-schists has been exploited at
the surface over 520 m, and underground for 350 m. The quartz was crushed by a battery with 10
stamps powered by a steam engine. The average ore grade was a respectable 13.5 g/t Au with a gold
finesse of 970.
Six km to the south the vein of Sahafandroana, another quartz vein, yielded gold grades superior
to Babasy but it was worked very little and prospected only by a trench of about 100 m in length.
In the same sector to the west of the Ambodilafa village there are other primary exploitations:
Ambavalero and Ambohitsara. The same sector also hosts several auriferous alluvions, particularly
along the tributaries of Sahavato: Sahafandroana, Sakandrekata, Sakafotsy Sahafanana and
Perhaps the most important primary gold deposit of the region is Andrambo-Maroantovo, 40 km to
the SSW of Soavina village. The deposit consists, as usual, of a lens-like quartz veins
concordant with the Mananjary gneiss. These gneiss are often dioritic and have lenses of
The deposit is well within the surface expression of the Befody granite (sketch 7/10). Galena
from this vein has been dated at 2,140 MA. The main vein strikes N 140§ and dips 70§ to 80§ with
an average thickness of only 20 cm; this vein has been followed on strike for about 4 km by
artisanal miners who have exploited only the richest sectors, with reported grades of 30-80 g/t
Au and even higher (Besairie, 1966, p 280b).
The southern sector of the vein is the better studied and the richest. BRGM reports an average
ore grade of 50 g/t. This sector has been worked over a strike of 1,400 m with trenches, adits
and cross-cuts. These workings were directed only to the zone above the water table.
In 1932 the conclusions from this exploration activities were: very erratic grades, average vein
thickness of about 15 cm, average grade of 20-30 g/t Au inversely proportional to the thickness,
gold finesse about 800 with a strong proportion of silver and some bismuth.
Above the water table the vein is enclosed in the laterite and altered gneiss; the presence of a
previous working made exploitation more difficult. 20,000 tonnes of mineralized quartz have been
estimated above the water table. There is no consistent information on the vein below the water
table, but the grades were thought to be decreasing (no enrichment by secondary alteration) and
mining conditions obviously more difficult.
All this did not bode well for a mechanised exploitation (at least in 1932) and no serious
investment took place. On the sector north of the Andrambo-Maroantovo prospect the average vein
thickness is larger (20 to 30 cm) but the grade falls to around 15 g/t Au.
Another recorded primary deposit is Andravoravo-Tsaramiadana which occurs 12 km to the NNW of
Vohilava village (see sketch 7/10). Important works were carried out by Mr Hanning around 1910
and were suspended in 1925. It consisted of adits and cross-cuts in laterite. The quartz vein,
besides visible gold, contained abundant sulphides with pyrite, chalcopyrite, galena, and
mispickel. The vein of Ambodivato slightly more to the north has been exploited in a large
This prospect occurs to the extreme south-west of the Befody granite, 8 km west of Vohilava
village. The prospect is very close to the granite-micaschists contact which has been explored by
adits (Besairie, 1966 p 281a). Nearby, the prospect Lavakianja is noteworthy for its abundance of
To the north of the Vohilava-Ampasary goldfield, the Itrozona River is a tributary of the
Vatovandana which flows into the Indian Ocean to the north of Masomameloka. At Beando, 3 km to
the south-east of Beando village, a 40 cm-thick quartz vein has been exploited by a trench 80 m
long with a few shafts.
A small 5-stamp mill was installed and operated by a hydraulic wheel. In 1912 this plant produced
36 kg of gold. In the same sector the placers of Androrangapetraka and of Ambodiara have been
worked. Other artisanal works took place at the Ambalafary and Sahaberiana rivers.
This locality, 18 km north-east of Vohilava, hosts some primary mineralisations in gold-bearing
quartz veins, lateritised and therefore enriched. An open cast mine was opened to exploit
sericite-schists abundantly intruded by quartz-pegmatitic veinlets in a mylonitic zone. There are
records (Besairie, 1966 p 281b) about the presence of quartz-sulphide veinlets up to 40 cm thick
with grades of 8 g/t Au and 9 g/t Ag. At Fotobato, 15 km to the NNE of Vohipeno village, there
occur large quartz lenses with traces of sulphides: these lenses are very irregularly distributed
and they are apparently grading 10 g/t Au.
An abundant sulphide mineralisation is located in the veins near the faulted zone of Andranotonga
12 km to the north of Vohilava.
Some quartz veins with biotite, tourmaline and fine pyrite and mispickel, weakly mineralised in
gold and silver, occur near the falls of the Sahamatrano River on the track
Andranotonga-Tanambao. On the Ambalavary and Ambatoharanana rivers wulfenite (PbMoO4) has been
found. The Besonjo creek is noteworthy for abundant quartz with sulphides; but no significant
gold values are reported.
On the upper course of the Sahandriambo River fluvial terraces and flats are almost nonexistant
and mining took place from the river itself. The actual river bed is certainly very rich as all
its southern tributaries are auriferous from the Ambodimanga R. onward.
The placers on the Ambatofotsy R. have produced 76 kg of gold between 1906 and 1928 after being
drained by means of a diversion canal 2,200 m long. Around 1910 two meanders upstream of the
Antapandrano R. were also drained. These and several other works aimed at tapping the region's
unquestionable alluvial gold potential have not apparently been overly successful. Besairie (1966
p 281b) mentions difficulties in bringing water to the various placers but this explanation
sounds unconvincing (IMR's opinion).
Several primary prospects are known in the Maha R. valley. Anajaridaina prospect, 2 km to the
west of Morafeno village, is in a mylonitic zone rich in quartz-pegmatitic veins which are rather
discontinuous. Another prospect, Manakana, 15 km west of Anajarindaina, is in numerous veins of
chlorite-rich schists: the paragenesis consisted of pyrite, chalcopyrite, rutile and some gold.
The two prospects of Ambia-north and Ambia-south, 7 km to the north-west of Morovato, were
exploited in the 1930s. The mineralisation was hosted in sulphide-rich quartz veins occasionally
associated with calcite or magnetite.
The alluvial deposits of the Maha are in terraces and flats. Most of the terraces have been more
or less exploited in the past. The flats are now largely covered by coffee plantations. In the
terraces a loose conglomerate of variable thickness (0.2 to 2 m) is overlaid by sterile lateritic
clays 3 to 7 m thick. The gold-bearing alluvions are very clayey and require a thorough washing
("d‚bourbage"). The ore grades range between 0.3 and 2.5 g/m3.
Several primary prospects are known in the valley of the Saka River. The already mentioned
Namolika prospect is one of these and it has been worked underground. Similarly at Tsaramiandana
a gold-bearing quartz vein has been exploited by shafts. The largest known alluvions are:
Andakatsiefo, Saka-Manama and Ambodiara. These and others are only a few km away from Vohilava
village (sketch 7/10). At Fadransana on the Bealana River, 34 kg of gold was produced between
1911 and 1928. Not all these alluvions have been exhausted. Some were still being worked by
"artisanal" means at the time of compilation of the present report (1995).
Prospection on the river-beds of Saka and Maha rivers
As this report is more oriented toward the hard-rock gold potential rather than the alluvial one,
the operations on the Saka and Maha rivers will receive only a rapid coverage mainly from
Besairie (1966, 1969) texts.
A exploration program was carried out on these rivers between 1939 and 1940 with a banka drill.
This work has resulted in the observations below, which are essentially applicable to most rivers
on Madagascar's east coast as the hydrographic regime is essentially the same.
The rivers of the Saka and Maha type have a bedrock of very irregular surface with numerous
rapids and waterfalls which explains the great irregularity in the thickness of the alluvions.
The alluvions are only coarsely classified and they are reworked at every flood, the gold bring
redispersed within the alluvions. Local concentrations vary from season to season. These
concentrations occur either at the surface of the alluvions or in the black sands [magnetite?]
within the alluvions. These concentrations are generally at the borders of the alluvions, and
slightly downstream of eroded banks or mined terraces.
Six exploration augers-drills on the Maha in 1939 indicated an average grade not above the 0.1
g/t Au. On the Saka River 48 Banka holes have been sunk. The only grades worth mentioning are:
1 g/m3 (two holes)
0.7 g/m3 (one hole)
0.5 g/m3 (two holes)
0.3 g/m3 (six holes)
These holes clearly showed that if the present alluvions appear rather rich at the surface, they
are disappointingly poor at depth. The alluvions' thickness ranges between 2 and 6 metres. Gold
mining by diving was limited to the upper part of the alluvions.
This exploration campaign indicated that the alluvions could not be exploited by conventional
dredges. In view of this the Madagascar Mines Service built special equipment (patent NeviŠres)
which could suck the alluvial sands at a given depth without being obliged to absorb its sterile
The dredge also was equipped with a water jet to facilitate the disgregation of the alluvions and
an suction duct. The production of this contraption was 6 m3/h in medium-sized gravels and up to
15 m3/h in sands and grits. During experimental tests at 4 m depth the dredge recovered, after a
water injection of one minute and an aspiration of one minute, 65 % of the lead and 55 % of the
gold that was sunk in the alluvions. It was estimated that this type of small dredge was likely
to give very satisfactory results in stable alluvial deposits even in coarse gravels.
The dredge was then tested directly on the Saka with the aim of exploiting some the sectors
indicated most favourable by the Banka drills. The results were extremely poor. They showed that
[predictably] there was no continuity between two holes with high gold grades. It was therefore
impossible to estimate an average grade from the drill-hole results and the average grades
obtained by this primitive suction dredge were always inferior to the grades indicated by Banka
In reality the description of the gold alluvions of the Maha and Saka rivers are no different to
the ones in tropical rain forests (the eastern part of Madagascar) with a hydrography draining
from gold-bearing laterites.
Some parts of Papua New Guinea, British Guyana and of Zaire's Kivu province present the same
conditions. The problems of efficient gold recovery from comparatively small and erratic placers
are not new. Literally thousands of small floating dredges, some of them very ingenious, have
been built in Australia, South-East Asia and South America for the exploitation of these placers,
invariably with mixed results.
In the 1980s the French BRGM (through its subsidiary Soditra - Sodiloc), built a state-of-the-art
small dredge which, although built and tested in Madagascar, had the world market in mind. These
dredges, called "Dragor 6"" and "Dragor 4" (sketch 7/1/1), are not significantly different from
US models developed two decades before.
The general geology and morphology of Madagascar does not seem favourable to the formation of
large alluvial gold deposits like those of Brazil, Venezuela or some parts of Papua New Guinea.
Exploration for these types of deposits will probably be outside the aims of large foreign
investors to whom this report is aimed.
7.2.8. BEFORONA GOLD DOSTRICT
Location and History
The Beforona gold district is bisected by the main road and railroad linking Antananarivo to
Toamasina (formerly Tamatave), the main Madagascan port on the east coast. This gold district can
easily be reached by 124 km of good, surfaced road from Antananarivo (sketch 7/15).
Beforona village, central to the gold district, has no facilities, but 30 km to the east the
small town of Moromanga (comfortable hotel, telecommunications) is a convenient base. Most of the
deposits of the Beforona gold district can be reached only by 4WD vehicles or on foot.
Most of the deposits of this goldfield were already productive in 1900 when this district was, in
terms of production, the second largest in Madagascar.
The gold deposits of the Beforona district are hosted in migmatites and amphibolites of the
Beforona group (C5) of the Vohibory system which have been dated at 2,700 MA (Besairie, 1964).
The gold mineralisation has probably been introduced or significantly remobilised in the major
tectonic event of 2,104 MA, but later remobilisations cannot be excluded.
Regionally, the gold-bearing migmatites and gneiss of the Beforona gold district seem to be the
extension of similar formations of the Mananjari system of the Vohilava-Ampasary gold district,
500 km to the south. The main difference seems to be that while the Vohilava-Ampasary gold is
hosted in a Fe-Ca-Mg metamorphic series, the Beforona gold is hosted in a more sialic
environment: possibly the metamorphic grade is also higher (migmatites rather than schists) than
the Vohilava-Ampasary series.
Also similar to the Vohilava-Amposary is prevalent north-south regional tectonics. The Beforona
district's known mineralisation consists of free-milling gold and gold in Cu-Fe sulphides
disseminated in quartz veins.
Table 6 lists only three recorded deposits, all primary, for the comparatively small and
understudied Beforona gold district. Significantly, 27 of them are primary.
The important Marovato deposit is only 15 km from Rogez Station on the Tananarive-Toamasina
railroad. It is one of the last primary gold deposits to have been industrially worked in
Madagascar: the deposit was abandoned in 1952. It was during Wold War II that Marovato attracted
new interest and investment from the Madagascar Service of Mines Exploitations (Mr A Lenoble).
The Marovato deposit consists of three gold-bearing quartz veins conformable within the
VEIN I: striking over 4 km and prospected by trenches, shafts and galleries at several levels
over a length of
405 m and a depth of 35 m. Its thickness ranges from 35 to 150 cm.
VEIN II: striking about 1,800 m: only small workings have taken place in this vein.
VEIN III: rather far apart and less rich than the previous two veins.
The ore is in quartz with auriferous pyrites and chalcopyrites and a few zones of free-milling
gold. In 1945, at Marovato, there was 520 m of underground development and 70 m of shafts. Power
was supplied by a hydraulic plant of 30 CV. Mr Lenoble estimated known reserves of 5.5 tonnes of
gold (177,400 oz) at unreported cut-off grades.
Preliminary tests over 250 tons of ore yielded 3,025 kg of gold or 12 g/t Au. At the time it was
estimated an industrial recovery of 10 g/t plus about 7 g/t Au in the sulphides. A further, well
quartered, sample treated by fire assay gave an average ore grade of 13-17 g/t Au. The plant was
destroyed by fire in 1947 although some minor prospecting activities took place until 1952 when
the mine was definitely abandoned.
The Grigri prospect, 45 km east of Moramanga, raised high hopes in 1905 but ended up as total
failure after some underground workings. The main deposit is in a large quartz vein 80 m long, 10
m thick at most, and at least 13 m deep. Underground surveys (tracages) were carried out at
several levels. This mass of white quartz showed an average grade of about 6-7 g/t Au with
occasional pyrite-rich zones with higher grades. Gold occurs either as an inclusion in pyrite or,
more rarely, as free-milling.
Other smaller, but not richer, gold-bearing quartz veins occur nearby the main Grigri prospect.
7.2.9. ANDRARONA GOLD DISTRICT
Location and History
The Andrarona gold district occurs in the extreme north-east of the island in the Masoala
peninsula, north of the Antongil Bay. This is a rather minor gold field, often quoted in the
literature as Antalaha, from the omonymous seaside town.
The Andrarona gold district can be reached either by a twice-weekly plane to Iharana (Vohemar) or
Sambava townships and then by by road, mostly unsurfaced to the village of Antalaha (with only
primitive accommodation). Conversely, it is possible to reach Antisarabato-Antalaha by charter
plane. The best base for the Andrarona gold district is the town of Sambava (hotel,
telecommunications and other services) 90 km to the north of the goldfield (see sketch 7/13).
This small gold district has officially produced only 270 kg Au between 1905 and 1920 - the
period of major production for Madagascar.
The geology of the Andrarona gold district is unique in Madagascar. Andrarona is the only
district where gold is clearly related to granitic intrusions. Although the gold mineralisation
occurs in Precambrian schists and micaschists of the Beforona group of the Graphite System, the
mineralising episode appears related to the emplacement of the Antonigil granites tentatively
dated at 770 MA (Besairie, 1964), but possibly older.
However, there are reservations about these Antonigil granites being the primary source of gold,
as in some works consulted these granites appear much older, and possibly cratonic (Hottin,
1972). On the other hand the BRGM (1985) points to a possible - and highly unlikely - Devonian
age for the Andrarona schists. The lack of modern geochronology of the region makes these
discussions rather futile. We also note that the consistent presence in the same goldfield of two
different gold finenesses (975 and 850) could indicate a multiple origin of the gold.
There is no obvious tectonic control in the Andrarona gold deposits but the region is strongly
tectonised, being in one of the regions of maximum stress linked with the break-up between India
The recent activities of artisanal miners in the previously unknown region of Marantsetra (see
sketches 5/1, 7/1 and 7/13), about 120 km to the south west, may indicate that the Androrona gold
district could be much larger than previously thought.
Table 6 lists 11 recorded deposits, all primary, for the Andrarona gold district, five of them
belonging to the comparatively rare a4 genetic type (see chapter 6.5).
This primary deposit, and relative alluvions, occurs 18 km south of Antalaha and, to date, has
been the largest known occurrence in the district. The deposit apparently lies at the contact
zone between a granitic massif (Masoala granite) and the schists and quartzites of the Beforona
group. The schists are cut by a swarm of quartz veinlets both conformable or uncomformable. The
quartz is milky or blue-coloured, often with tourmaline or rutile needles. Gold occurs in pyrite,
although alteration has occasionally liberated free-milling gold. Grades are reported at between
6 and 7 g/t with occasional peaks of 30 g/t.
Besairie (1966, p 270b) mentions this primary deposit occurring to the west of the Antsahivo
River, where a quartz vein was followed over 200 m by superficial working. Recorded grades were
around 30 g/t Au. This large vein, almost concordant with the schists, was dipping 75§ and is cut
by several veinlets forming a stockwork pattern with an average grade of 6-7 g/t Au over an
average thickness of about 1 m. Measured reserves have been calculated at 86 kg of gold with
possible reserves of 172 kg. We ignore the methods of these calculations (IMR's note).
At Andrarona, in the upper valley of Marambo, gold-bearing quartz (milky, blue or haline) veins
with rutile and tourmaline inclusions have been mentioned but without further details (Besairie,
1966 p 270a).
In the Andrarona goldfield several, presumably small, alluvial deposits have been recorded. Among
these: Andongozabe, Ambodihintsy, Ambalambato, Andranomena and Antsahabe.
7.2.10. VAVATENINA GOLD DISTRICT
Location and history
The Vavatenina gold district is located 80 km north of the provincial capital, and major port, of
Toamasina (formerly Tamatave). It can be reached by daily flights (Air Madagascar - Boeing 737)
to Toamasina and by 157 km of surfaced road to Vavatenina: but this village though central to the
goldfield offers no facilities. The best base to this goldfield is the pleasant seaside town of
Fenoarivo (hotel, telecommunications, landing strip) 40 km of surfaced road to the east of the
Vavatenina gold district is the smallest in Madagascar with a known extension of only a few tens
of km2. In the past (Besairie, 1966) the Vavantenina goldfield was considered jointly with the
much larger Berofona one and statistics are therefore rather confused.
In spite of its reduced area the Vavatenina district has been a significant gold producer in the
past: 40 kg of gold was produced in only six months in 1902. But production decreased sharply
from 1904 onward. Production took off again from 1941, with 8.8 kg produced in 1942 and 11 kg in
1944. Present production is entirely due to artisanal miners.
The geology of Vavatenina is charaterised by the predominance of of biotite-micaschists,
occasionally with kyanite. These schists are frequently cut by pegmatitic swarms and
quartz-feldspar veinlets which correspond to the migmatitic front in the Ambodiriana group (C7)
of in the Vohibory system. These micaschists have been dated at 2750 MA and the regional
tectonics show strong faulting striking 10§ to 15§ to the east (Besairie, 1964, Feuille du
Centre). There are no obvious granitic intrusions nearby.
The average gold finesse is 950, possibly pointing to a sulphide-poor environment (IMR's note).
Practically all rivers and streams in the region are auriferous, although they all are rather
small: never more than creeks, which limit their potential. A possible correlation of this
district with the Beforona gold district to the south in very similar geological and tectonic
conditions is tempting, as both are hosted in the Vohibory system (see Plate 1).
The only two (but primary) deposits listed in Table 6 point to the relative lack of knowledge of
the apparently small Vavatenina gold district.
The three individual gold deposits recorded are about 10-13 km west of the Vavatenina village
(Besairie 1969, feuille Tamatave). There no available descriptions of these deposits. A thermal
spring 8 km to the north of these prospects is well worth a geochemical characterisation (IMR's
8. GOLD IN MADAGASCAR: THE OPINIONS
8.1. POLITICAL RISK
As a consequence of poor information, Madagascar has been rather neglected by the wave of major
investors now scouring Africa for mining opportunities, particularly in gold; with South African,
North American and even Ghanan mining companies leading the pack.
Short-term political risk
Madagascar is still seen as an exotic country with untested democratic and market convictions.
This is not accurate. Since 1992 Madagascar's commitment to a liberal democracy has been
unquestionably sincere, although the legacy of nearly 20 years of centralised military rule is
The World Bank, among others, is a willing creditor to Madagascar. Madagascar is member of the
Multilateral Investment Guarantee Agency (MIGA). Taking these aspects into consideration, the
"risk" factor of Madagascar is not higher than for some other regions of Africa.
Less so the "geological risk". Here the western investors feel rather uncomfortable. A casual
business trip to Antananarivo does little to dispel this feeling. Gold exploration in Madagascar
requires both a serious commitment and plenty of homework. These aspects will be covered in the
But Madagascar's new Mining Code of August 1990 (law No. 90-017) and other investment laws are
adequate to cover potential mining investors.
It must be understood that since the demise of communism as a viable economic doctrine the
concept of "political risk" in mining ventures has changed radically.
Gone are the days where "repatriation of profits" was a major consideration. With even Castro's
Cuba now guaranteeing unrestricted repatriation of profits for foreign investors, the problem
doesn't seriously present itself any more. A country that does not accept that is simply not
serious about attracting mining investment in an increasingly competitive African scenario (Otto,
1995). A similar consideration is applicable to the other standard requirements for foreign
- mining investors:
Security of tenure,
- Management control
- Equity control
- Fixed tax terms
- Fixed environmental requirements
These requirements should in any case be firmly reviewed, point by point, with the appropriate
Malagasy authorities, and guarantees obtained before any major exploration commitment is made.
Large mining companies should avoid demanding a "one stop shop" for formal approval of their
exploration ventures. The often-mentioned "one shop stop" concept is simply unrealistic. In
Madagascar, as in most parts of the developing world, this approach can only encourage middlemen
and dubious "agencies", arouse public resentment, and open legal ways to later renege on the deal
concluded. There have been recent cases to this effect in India and some African countries.
The political risk in any country, especially an untested and unusual one such as Madagascar, can
be effectively reduced by gaining an intimate knowledge of the country's political climate. Every
day spent in aquiring a better understanding of Madagascar's objectives and aspirations will
lessen the political risk of any major venture. Conversely, to take any African country - and
Madagascar in particular - for granted could only invite unexpected difficulties.
Central to the concept of risk is, clearly, the absence or the presence of favourable geology.
However encouraging the local regime or favourable its commercial climate, no exploration
investment could be sanctioned unless the country had high prospects for the discovery of a gold
deposit of the size and profitability rate of return decided upon by the investor.
The self-explanatory sketch 8/1 gives a modern assessment for gold exploration in selected
African countries, taking into account:
- Gold Potential
- Political Ranking
- Commercial Ranking
In this assessment Madagascar ranks on the same level as Burkina Faso, Ethiopia and Tanzania. It
also ranks way above Zaire but, understandably, below Ghana and Zimbabwe.
A last point. Madagascar, in virtue of its unique and very fragile ecosystem, has long been the
darling of various conservation groups. Such groups have been vociferously hostile to, for
instance, a large scale ilmenite prject in the south of the island long before its economic
viability was proved (see chapter 5.3.3 of this Database). This type of risk can be minimised by
opening dialogue, on relevant issues, with the appropriate authorities (national, local or
international) at a very early stage of any gold project.
Long-term political risk
Long-term political risk is always more difficult to evaluate.
It should be remembered that the recent opening and liberalisation of most African countries is
only five years old and may very well be just an extended honeymoon.
There are massive, unaddressed problems in Africa: in particular its rapidly expanding
population, with vastly insufficient new investments to cope with it. In this scenario, the
menace of future coups d'Etat, civil unrest or ethnic violence cannot be easily dismissed.
In Madagascar's case, however, some advantages over other African countries are apparent:
- Madagascar's racial or ethnic tensions are not nearly as strong and divisive as in other parts
of Africa. The latent hostility between highland and coastal tribes (an hostility expolited in
the past by the French) has nothing of the intensity of the Hutu-Tutsi in Central Africa, the
Xhosa-Zulu in Natal or the clannish rivalries of the horn of Africa and Kenya.
- Even at the height of Malagasy military coups and counter-coups of the years 1972-1975, the
casualties were kept to a minimum. Malagasy people dislike violence.
- The fact that Madagascar was a cohesive society long before European colonisation adds an
element of stability rare in Africa.
- Madagascar's lack of borders and neighbours make any armed destabilisation more difficult than
the rest of Africa.
- Armed banditism on the scale of Mozambique or Uganda is not a present Malagasy feature and an
unlikely future one.
- After 20 years of disastrous central planning and declining living standards, the average
Malagasy is genuinely looking forward to stability and economic progress.
- Gold mining is a rather self-contained activity almost independent from heavy infrastructure,
which makes it far less vulnerable to external turmoil.
8.2. GEOLOGICAL POTENTIAL
This is the paramount consideration. If a gold deposit of the size and profitability sought by
the investing company does not exist in Madagascar there is no reason for the company to be there
at all, regardless of any other favourable circumstance. A simple point, but overlooked
Reflecting on gold's potential in Madagascar, H. Besairie (1966) presciently lamented that "...
on ne s'est pas souiciŠ de relever les teneurs moyennes..." [... The average grades of the
deposits have not been recorded... ] (Bes. 271b). This is only one of the many difficulties
facing an Anglo-Saxon geologist trying to form an informed and objective opinion on Madagascar's
Other constraints are:
- Lack of recent (post-1969) literature on Madagascar's goldfields;
- The fact that practically all the previous literature on Malagasy gold, although abundant, is
very difficult to gain access to, is over-descriptive and invariably in French, and technical
details such as cut-off grades, metallurgical plants used and recovery rates achieved etc. are
- The fact that Madagascar is essentially an agricultural country, rather than a mining, or even
an industrial country, make a meaningful dialogue with the authorities far more difficult than
for some other African countries;
- The neglect and the lack of funds that have affected both the Geological Survey and the Mining
Service of Madagascar in the last two decades.
Alluvial Gold Potential
Madagascar has produced more than 70 tonnes of gold from 10 documented goldfields, which in the
last 15 years have considerably expanded in size.
For various reasons, gold production in the past was mostly from alluvial operations, and this
long before suction dredges became the efficient tools they are today. This augurs well for the
development of a lively middle-sized local industry. Contrary to most east African countries,
Madagascar has more than 30 rivers where small gold dredges can operate (see also Table 2 and
A standard Keene 2«" suction dredge (or equivalent model) can easily go through m3 of gravel per
hour (Thoronthon, 1979; p 50) and operate in as little as 50 cm of water. In Madagascar there are
several hundred kilometres of river beds which have never been properly dredged.
Among the Malagasy gold districts that seem to offer outstanding potential for small river
dredges are the Maentavana, the Ambositra-Antananarivo and the Vohilava-Ampasary. These three
gold districts possess three important conditions for sustaining medium-scale dredging:
- 1. The area drained is large and it is heavily mineralised;
2. The fluvial regime is very dynamic;
3. The gold of the district is coarse in nature.
Regardless of its potential, alluvial gold mining in Madagascar is not recommended to the
reader of this Database.
Foreign investors are traditionally attracted by large primary gold deposits where the recent
lessons of, say, Australia and North America can easily be applied. A sharp separation between
small and medium-sized alluvial operations (better suited to Malagasy entrepreneurs) and large
primary ones (better suited to well-capitalised firms) will be beneficial to all parties
Primary gold potential
The primary gold deposits of Madagascar mostly occur in gold-bearing quartz-vein systems
frequently associated with sulphides. With the exception of the Betsiaka gold district in
Madagascar's extreme north, all other districts are in Precambrian terrains, which form
two-thirds of the island. Geologically, these terrains are related to other gold-producing areas
in southern Africa and probably India.
Madagascar's gold deposits are essentially of the "mesothermal lode gold" type, which account for
more than 18% of world production. In this respect they are, genetically at least, similar to
the deposits in the Archean shields of Canada and Australia, and to a lesser extent, those of
western Africa, Brazil, Venezuela and Tanzania. Exploration for these types of deposits has
increased dramatically in the last 10 years, with considerable success.
The traditional approach to the exploration for "lode gold deposits" is a thorough reassessment
of ancient workings. The Canadian Helmo deposit, an outstanding gold discovery of the 1980s, had
a history of minor development dating from 1869.
Most of Madagascar's old underground gold mines, more than 50 of them, some with several hundred
metres of underground development, were considered uneconomic by the early 1920s.
This was precisely the situation of Western Australia before the gold boom of the late 1970s. In
both cases grades below 10/g/t Au were seldom mined and the prospecting techniques were most
primitive. In Madagascar they often consisted of following the quartz veins at the surface and
exploiting only the rich secondary enrichments between quartz veins and overlying laterites.
Madagascar offers plenty of scope for modern exploration. Its well exposed terrain is ideally
suited to remote sensing, particularly because of the strong relationship between Madagascar's
goldfields and tectonics (Madagascar is arguably the most fractured fragment of Gondwana).
The enormous advancement in modern soil and rock geochemistry have all but passed Madagascar by.
So has the application of metallogenic models developed in the last 20 years in regions
geologically similar to Madagascar - in some cases with outstanding success. In spite of some
minor French and Soviet involvement, Madagascar's gold resources have not really been assessed in
In exploration for primary gold deposits the concept of size is very important. South African
gold mining companies, few in number, very large and bent on the necessity of replacing their
increasingly uncompetitive mines, will probably be uninterested in a project with less than 1M oz
of contained gold. A possible but a totally speculative type of deposit in underexplored
On the other hand the very successful and well diversified Australian gold sector operates more
than 131 mines with an average production of around 60,000 oz/y. (Pollock, 1995). This a far more
realistic target for Madagascar's presently known potential. Table 2 of this Database lists seven
gold districts thought to have high potential for this type of resource.
The exploration for medium-sized gold deposits (say 50.000 to 300.000 oz Au) seems to make
particularly good sense in Madagascar: they are statistically more frequent, easier to finance
and develop, gentler on the environment, more attuned to the decentralisation will of the
Malagasy authorities and likely to produce less economic and regional distortion of Madagascar's
In any case, exploration for primary gold in Madagascar will demand commitment. To reinterpret
the incredible wealth of accurate, if obsolete, geological information will require time and
skill. The country has a proud history of self-sufficiency and gold is already its major mineral
Predictably, the Malagasy authorities (already bitten by some recent, hare-brained schemes
regarding their gold resources) will require their potential investors to be serious. Given
Madagascar's unquestionable geological potential this is only fair.
9. CONCLUSIONS AND RECOMMENDATIONS
Australian experience is relevant to Madagascar in many ways.
Many goldfields which were very productive at the beginning of the century came to an abrupt halt in the second decade of this century. Classical example is Charter Tower gold field, the fifth largest in Australia where not an ounce of gold has been pulled out of the ground since 1917 and only recently being reinvestigated. In previous years 6.6 million oz was extracted from high-grade quartz veins which bear some similarities to most of Madagascar's goldfields. At today's price this means well in excess of US$ 2 billions. Then the gold price fell 50% between 1910 and 1920, mining operations were small and inefficient, and labor became scarce and expensive because of other opportunities.
Some authors (Borg 1995; p 74) have given a good account on how modern exploration techniques can be applicable to gold exploration and evaluation situations in a variety of different anomalous environment. Several relatively inexpensive techniques such as soil, rock and gossan sampling, detailed exam of old working and known mineralisation have given good results in almost all cases and particularly for the associations gold-pyrite which is very common in Madagascar.
In considering gold ventures in Africa the objectives of the explorer are all important: the five major South-African gold producers, bent in substituting what are now very deep and costly mines would not seriously consider a gold resource below 1 million oz gold (or about 30 kgs in situ resources). So probably will be the view of some of the major Australian mining houses: BHP,CRA, WMC, etc. On the contrary for some small speculative company from Perth or Toronto the main consideration would be the "profitability" of the venture and much smaller resources can be considered provided the grades are there: this indeed has been the philosophy in Western Australian goldfields for the last 15 years.
Lastly some medium-sized companies consider their strategic medium term objective to become a plus 250.000 ounce gold producer through mining development in Africa (Cowley, 1995; p 146). At this stage of our knowledge Madagascar offers scope to all these objectives.
The metallic associations of many Madagascar's gold districts
would make the modern use of soil and rock geochemistry highly effective, much as it has happened in geologically contiguous Tanzania.
Countries under consideration in Table XX - elaborated jointly by the private industry and the UN - show various elements of risks for mining investment. Without examining each item individually for Madagascar let's consider the most important.
Geological criteria - geological risk.
This brings in what South African geologists operating in Africa (an ever-growing breed) call geological risk. That is the geological uncertitude about the environment in which one is operating. With only minimal government funded geological exploration having taken place since independence of most African countries this geological risk is large and widening. By and large the geological and exploration models developed in Australia, North and South America in the last, say 30 years, have not yet been applied to most of Africa, Madagascar included. Mineral exploration has been poorly funded since the 1960's in Madagascar and often directed to energy resources: uranium, oil and gas, coal. At Government level Madagscar has amost missed out the gold exploration "boom" of the last 17 years.
This tends to put a gold exploration company in a "catch 22" situation: for knowing Madagascar's gold potential some exploration is unquestionably needed, but then one doesn't know if that exploration (with the inevitable arrangements with the Malagasy authorities that this entails) is warranted. It is up to any company, in view of its objectives and exploration philosophy to make this type of decision. In a way the present report has been made precisely with the view of lessening geological risk by combining together what is publically known about gold in Madagascar.
Two considerations are worth keeping in mind:
1) that for the forseeable future the geological risk on gold exploration in Madagascar is not going to be significantly decreased and
- 2) That medium sized gold deposits (say 50,000 - 500,000 oz Au) are statistically more frequent than very large ones (1,000,000 - plus oz Au) thus lessening the geological risk.
Agrawal, P. K., Pandey, O. P. and Negi, 1992. Madagascar: A continental fragment of the
paleo-super Dharwar craton of India. Geology Vol 20 p 543-546
Ainsworth, J., 1995. Mining development in Africa. in Conference "Africa, the New Mining
Frontier, J.B. Were & Son, Sydney, 11p
Andriamirado, R., 1976. Quelques remarques sur l'ancienne position de Madagascar par rapport …
l'Afrique. Acad‚mie des Sciences Comptes Rendues, ser D, v. 282, p 2041-2043
Anhaeusser, C. R., 1976. The nature and distribution of Archean gold mineralization in southern
Africa. Mineral Science and Engineering, 8 p. 46-84
Antonides, L. E. et al., 1988. The Mineral Industry of East Africa. In Mineral Yearbook 1988, Vol
3 ; Area Report International, U.S. Department of Interior, Bureau of Mines, Washington, p 75
Audley-Charles, M.G., Ballantine, P.D. and Hall, R., 1988. Mesozoic-Cenozoic rift-drift sequence
of Asian fragments from Gondwanaland. Tectonophysics 155, pp 317-330
Bartle, W.W., 1988. The Madagascar mineral sand project - its impact on the market. in
Proceedings, 8th Industrial Minerals International Congress , Boston, USA, G.M. Clarke Ed. pp
Besairie, H., 1964. Madagascar. Carte Geologique 1/1,000,000. In three sheets [in colour].
Besairie, H., 1969. Madagascar. Carte Geologique 1/500,000. In six sheets [in colour].
Besairie, H., 1966. Gites Min‚raux de Madagascar. Annales Geologiques de Madagascar, Fascicule
XXXIV, Tananarive (Two volumes: text anf Figures) p 437 + 93
Besairie, H., 1971. Geologie de Madagascar, 1. Ann. Geol. Madagascar, 35, p.463
Besairie, H., 1973. Precis de Geologie Malgache. Ann. Geol. Madagascar, 36 p. 143.
Bigioggero, B, Cadoppi, P., Costa, M., Omenetto, P., and Sacchi, R., 1989. Late Proterozoic
granites of Zambesia (Mozambique) IGCP n§ 225 Newsletter / Bulletin, 2 p 5-7
B.R.G.M., 1985. Or [In Madagascar]. Partial Rapport filed at the Library of the Service
Geologique, Antanananrivo p 329-350
Boast, J and Nairn A. E. M., 1982. An outline of the Geology of Madagascar. In Stelhi, F. G.
(Ed) " The Ocean Basisns and Margins", Vol 6, Plenum Press, p. 649-696
Borg, G., 1995. Multidisciplinary exploration approach for Archean gold and effects of
liberalization of mining and trade regulations in Tanzania. "in African Mining '95". The
Institution of Mining and Metallurgy, London. pp 63-78
Brandon, A. D. and Meen, J. K., 1995. Nd isotopic evidence for the position of southern Indian
terranes within East Gondwana. Precambrian Research 70., p 269-280
Coffin, M.F., and Rabinowitz, P.D., 1987. Reconstruction of Madagascar and Africa: Evidence from
the Davie fracture zone and western Somali Basin. Journal of Geophysical Research, v.92, p
Colvine, A.C. et al., 1988. Archean Lode Gold Deposits in Ontario. PartI: A Depositional Model.
Part II A Genetic Model: Mines and Minerals Division , Ontario Geological Survey, Miscellaneous
Paper 139, 135 p.
Cowley, P.N., 1995. Investment decisions by an African Gold miner. "in African Mining '95". The
Institution of Mining and Metallurgy, London. pp 499-414
De Wit, M., Jeffrery, M., Bergh, H., and Nicolayesen, L., 1988. Geological Map of Gondwana, scale
1:10,000,000 with explanation and references, American Association of Petroleum Geologists,
Drean. P., 1994. Madagascar dans la tempete du mŠtal jaune. Interview to DMD Tirelire, Tananarive
Embleton, B.J.J., and McElhinny, M.W., 1975. The paleoposition of Madagascar : paleomagnetic
evidence from the Isalo Group. Earth and Planetary Sciences Letters, v. 27, p. 329-341
Farah, H.H., 1995. Evaluation of private sector mining projects in Africa by the African
Development Bank. "in African Mining '95". The Institution of Mining and Metallurgy, London. pp
Gallieni, J. S., 1900, L'or … Madagascar. Bull. Soc. GŠogr. Comm. Paris
Garson, M. S., and Goossens, P. J. 1980. Comparative Gondwana metallogeny and structures in
south-western India, Sri Lanka and Madagascar. 26th Intern. Geological Congress - Resumes (26
Gilbert, M., 1989. Second World War. Fontana/Collins. p 846.
Govett, M.H. and Harrowell, M.R., 1982. Gold. Australian Mineral Economics P/L, Special
publication, Sydney, Australia,
Granadier, G., 1897, L'or … Madagascar. Bull. Com. Mad. t.3 Tananarive
Hester, B.W., 1990. Gold opportunities in Tanzania: DMBW Inc. (Denver), p 91
Hottin, G., 1964. Le gisement d'or d'Andravoravo. Rapp. ann. Serv. Geol. Mad. Tanarive
Hottin, G. 1976. Pr‚sentation te essai d'interpretation du Pr‚cambrien de Madagascar. With Map at
1:2,500,000; Bull. BRGM. (2ieme Ser.), sect.IV : p 117-153
Hodgson, C.J., 1993, Mesothermal lode-gold deposits. In: Kirkham, R. V. et al. "Mineral Deposits
Modelling" Geological Association of Canada, p 635-678.
Hoffman, P.F., 1991, On accretion of granite-greenstone terranes, in Robert, F et al. eds,
Greenstone Gold and Crustal Evolution: Geological Association of Canada, NUNA Conf. Volune,
Hottin, G., et al., 1966. Sur des gisement d'or du groupe d'Ambatolampi (centre de Madagascar).
Madagascar Sem.Geol. C. p 105-106
Humpries, M., 1995. Mineral Development in Southern Africa: the Road to recovery? in "African
Mining '95". The Institution of Mining and Metallurgy, London. pp 263-284
Katz, M. B. and Premoli, C., 1979 - India and Madagascar in Gondwanaland: a fit based on
Precambrian tectonic lineaments; Nature Vol. 1791, pp 312-315.
Kerrich, R., 1989, Geodynamic setting and hydraulic regimes of shear zone-hosted mesothermal gold
deposits, in Bursnall, J.T., ed Mineralization and Shear Zones: Geological Association of Canada,
Short Course Notes, v.6, p. 89-128
Kirkham, R.V., 1993. Mineral Deposit Modeling. Geological association of Canada Special Paper 40,
Kroner, A., 1991. African linkage of Pre-cambrian Sri Lanka. Geol.Rundsch., 80: p 429-440
Kronsten, G., 1994. Madagascar. In Mining Journal Annual Review, 1994, London p 142.
Kuehn, S., Ogala, J., and Sango, P., 1990. Regional setting and nature of gold mineralization in
Tanzania and SouthWest Kenya; Precambrian Research, vol. 46, p 71-82.
Johnson, C., 1990. Ranking countries for Mineral Exploration. Natural Resources Forum Vol 14 No.3
Jourde G., 1971 Madagascar Schema Stratigraque. Scale 1:2,500,000. Antananarivo
Lacroix, A., 1900. Sur le gneiss aurifŠres de Madagascar. Bull. Soc. Fr. Min t 23, Paris.
Lawver, L. A. and Scotese, C. R., 1987, A revised reconstruction of Gondwanaland, in McKenzie, G.
D., (ed.) Gondwana Six: Structure, tectonics and geophysics. American Geophysical Union Monograph
40, p. 17-40
Leitch, C.H.B., Godwin, C.I., Brown, T.H. and Taylor, B.E., 1991, Geochemistry of mineralizing
fluids in the Bralorne-Pioneer mesothermal gold deposits, British Columbia, Canada: Economic
Geology v. 86, p. 318-353
Matumatsaka, W., and Hangi, A., 1995. Environmentally sustainable artisanal gold mining in the
Lake Victiria regions, Tanzania. in "African Mining '95". The Institution of Mining and
Metallurgy, London. pp 443-456
Mc.Elhinney, M. W., Embleton, B.J., Daly, L., and Pozzi, J.P. 1976, Paleomagnetic evidence for
the location of Madagascar in Gondwanaland. Geology, v. 4, p. 455-457
Metcalfe, I., 1991. Southeast Asian Terranes: Gondwana origins and evolution, 8th International
Gondwana Symposium, Hobart, Abstracts, p.58-59
Metcalfe, I., 1994. Gondwanaland origin, dispersion, and accretion of East and Southeast Asian
continental terranes, Journal South American Earth Sci. Vol.7 pp. 333-347
Mukherjee, A.D. and Bhattacharyya, N.H., 1990. Basin evolution and metallogeny during
Mid-Proterozoic in peninsular India. IGCP n§255 Newsletter / Bulletin, 3 p 51-60
Nicollet, C., 1990. Crustal evolution of the granulites of Madagascar. in Vielzeuf, D., and
Vidal, P (eds.) "granulites and crustal Evolution: Dordrecht, Nederlands, Kluver, p 291-310
Noizet, G., and Rakotonanahary, R., 1966. Les gisement d'or de Marovato et de Gri-Gri (centre est
de Madagascar). Madagascar Sem Geol. C p 139-141
Otto, J M., 1995, International competition for mineral investment: implication for Africa. in
"African Mining '95". The Institution of Mining and Metallurgy, London. pp 443-456
Pollock, J., 1995. Bloomberg Gold Survey. The AusIMM Bulletin, Nov 1955 Melbourne, p 26-131
Powell, C. McA., Roots, S.R., and Veevers, J., 1988. Pre-breakup continental extension in East
Gondwanaland and the Early opening of the Eastern Indian Ocean. Tectonophysics, 155: p. 261-283
Premoli, C., 1976. Tectonic Map of Madagascar, Scale 1:2,000,000. Office National Militaire des
Industries Strategiques (OMNIS), Antananarivo, Madagascar, 1 map.
Premoli, C., 1995 - Mineral Potential of Central Africa: an Australian overview. in African
Mining '95. The Institution of Mining and Metallurgy, London. pp 499-414
Pretorious, D.A. and Maske, S., 1976. An issue devoted to mineral deposits in Southern Africa:
Econ. Geol. vol 71 N.1, p408
Razafimnantsoa, 1966. Inventaire et possibilit‚s des mineralisations auriferes alluviales.
Madagacar Sem Geol. , C. p 143-144.
Ralaimihoatra, E., 1966. Histore de Madagascar. Antonanarivo, SocietŠ Malgache d'Edition.
Roy, B. C., 1963. Metallogenic-Minerogenetic Map of India, 1:2,000,000. First Edition, Geological
Survey of India
Saager, R., and Meyer, M., 1984, Gold distribution in Archean granitoids and supracrustal rocks
from southern Africa: a comparison. in Foster, R. P., (ed.) "Gold'82: the geology, geochemistry
and genesis of gold deposits" A.A. Balkema / Rotterdam. p 53-70.
Scotese, C. R., Gahagan, L. M. and Roger L. Larson., 1988, Plate tectonic reconstructions of the
Cretaceous and Cenozoic ocean basins. Elsevier Sciences Publisher p XXXXXXXXs
Shiraishi, K., Ellis, D. J., Hiroi, Y., Fanning, C. M., Motoyoshi, Y., and Nakai, Y., 1994.
Cambrian orogenic belts in East Antartica and Sri Lanka: implications for Gondwana assembly. J.
Geol., 102, p. 47-65
Thomas, A., 1995. Major gold prjects. Mining Magazine, June i955, Londn p 344-346.
Thoronton, M., 1979. Dredging for gold. Keene Industries. Northridge, California. p 243 p.
van Straaten, H.P., 1994. Gold mineralizations in Tanzania - a review. in Foster, R. P., (ed.)
"Gold'82: the geology, geochemistry and genesis of gold deposits" A.A. Balkema / Rotterdam. p
VŠrin, P., 1990. Madagascar. Paris, Editions Karthala
Viljoen, J. M., 1994. Archean gold mineralization and komatiites in southern Africa. in Foster,
R. P., (ed.) "Gold'82: the geology, geochemistry and genesis of gold deposits" A.A. Balkema /
Rotterdam. p 595-628.
Wopfner, H., 1994. The Malagasy Rift, a chasm in the Tethyan margin of Gondwana. Journal of South
East Asia n Earth Sciences 9. (4). p 451-461
Wopfner, H., 1991. Permo-Triassic sedimentary basins in Australia and East Africa and their
relationship to Gondwanic stress patterns. Proceeding 7th Gondwana Congress, Sao Paulo, Brazil,
1988; p. 133-146
World Bank (the), 1992. Strategy for African Mining, Technical Paper n.181 Washington, 1-81 p.
Yoshida, M., Funaki, M., and Vitanage, P.W., 1992. Proterozoic to Mesozoic East Gondwana: the
Juxtaposition of India, Sri Lanka, and Antartica. Tectonics, 11: p 381-391