Attempt at an assessment of the world's tropical moist forests

Adrian Sommer

ADRIAN SOMMER is a forester who is particularly concerned with research and collection of data about the extent of the world tropical forest. He compiled this study while assigned to the FAO by the Swiss government as an associate expert in forestry.

Under the influence of a euphoric belief in its unlimited growth, the area of tropical moist forests was seen until recently as an almost infinite resource, covering vast expanses of our planet and just waiting to be exploited or put to some other use. But rising concern for environment considerably changed this outlook. The international community became aware of the limited extension of these forest resources and of their gradual regression due to increasing human activities. For the first time, it became important to attempt an appraisal of the situation at the global level.

At this point, the difficulties began. The people involved in the task of gathering information were suddenly handicapped by an overabundance of data - a voluminous mass of confusing reports, scattered all over the world, yielding very few facts. There were good reasons for such a frustrating state of affairs. To begin with, most of the countries with tropical moist forests had been under colonial rule, and during these periods forestry exploitation concentrated particularly on few selected sites near coasts and rivers where economic prospects were most promising. These activities were still quite limited and it was possible to work, for many years, in relatively small areas. There was no need for extended surveys on larger areas, and therefore there was no intellectual motivation for developing better techniques.

But gradually the situation changed, the world demand for timber increased, prices went up, the areas with easy access were more and more overexploited, new exploitation techniques with far larger investments and production possibilities were developed and introduced, and the need of land for settlement and agriculture rose steadily. The integrated planning for future development of huge forest areas or of the entire forestry sector of whole countries, required careful appraisals of the actual situation. Various survey and inventory methods were developed and applied, generating abundant, heterogeneous and dispersed information. These traditional surveys were, and still are very costly and of limited use, for selected areas. It is only recently that, thanks to the new remote sensing techniques we have at last a tool which can give us an objective and accurate appraisal of the world's forest resources. But its successful adoption depends on the removal of a number of political obstacles, and other shortcomings such as the lack of qualified personnel, and adequate funds.

Therefore, a global appraisal of tropical moist forests undertaken at this time can only base its research on the material available - a mass of incomplete data and a number of assumptions. It will yield rather rough results. This applies in particular to this study, which does not produce any clear answers. The outcome of our research consists in estimates of a varying degree of reliability - their exactitude should not be overvalued.

It is hoped, though, that this attempt at getting a rounded picture may be retouched continuously through an increasing amount of precise data deriving from several remote sensing projects under way.

At this point, a summary of the major inadequacies characterizing the documentation examined seems in order:

· The required information is not available in the countries concerned.

· Not all existing information at country levels is available at FAO Headquarters.

· Existing and available information is often obsolete, not relevant or too uncertain.

· Available information is dispersed in various documents handled by different working units.

· The definitions and terms used in country appraisals vary considerably and it is very difficult to synthesize them.

· Data from forest inventories refer often only to a selected area and are not representative of whole countries.

· Repeated surveys for an assessment of current changes are still very rare.

· The same figures are repeatedly used in various reports and statements without knowledge of their origin and accuracy.

The starting point

In May 1974 the FAO Committee on Forest Development in the Tropics decided, at its 3rd session, to take a close look at the world's tropical forests. During the discussions it agreed on the following points:

TO DECIDE exactly which forest types below 1300 m altitude should be included and which should be excluded and to fix exactly where the borderlines between types can be drawn is a most arguable matter at world level and possibly a problem which cannot be solved.

THEREFORE, the interpretation of the term "tropical moist forest" should be left to individual countries, giving however the following guidelines:

1. Evergreen or partly evergreen forests (some trees may be deciduous) of the lower altitudes (up to 1300 m), never completely leafless, while in mature examples several more or less distinct strata exist.

2. Forest-savanna mosaic where patches of moist forest (not confined to streamsides) are surrounded by savanna of tall grass.

3. Coastal-savanna mosaic resembling the preceding type, less tall, lower rainfall.

These guidelines are derived from the Vegetation Map of Africa (AETFAT/Unesco, 1959) and refer mainly to the situation in the African region. For other continents, the conditions differ somewhat.

Despite the lack of consensus on the definition of the term "tropical moist forests," an attempt has been made in this study to present a more or less coherent qualitative and quantitative frame of reference by answering the following questions:

- What is the definition of the different forest types included within the area of the tropical moist forests?

- What is the total land area that could, and what is that which still bears tropical moist forests (climatic, climax areas)?

- Which are the present areas under tropical moist forests (broken down into different types if possible)?

- What is the total growing stock and annual growth potential of the existing tropical moist forests?

- What is the nature of conversions taking place at present and areas involved (present rate of regression)?

- What kinds of land use (and area of each) are being adopted within the tropical moist forest area?

Based on available information at FAO (vegetation and land use maps, atlases, different kinds of reports and general literature), a country-by-country evaluation of these questions has been attempted and the results regrouped for presentation by sub-continents and continents.

Criteria for selection

HUMID TROPICS ZONE. The opinion is still prevalent that the term "humid tropics" is mostly based on a climatic concept. The factors to consider are: the temperature, the precipitation, mainly with its distribution, and the length of the dry period. The main point to argue is how much seasonality as one of the principal factors determining the vegetation cover - can be allowed and an area still be classified as humid tropical. Available meteorological data are in general not yet adequate to determine valid boundaries of this humid tropical zone. In the near future, it will be possible to fill these gaps through the use of monitoring by remote sensing satellites, and thus define accurately the tropical humid area. Fosberg, Garnier and Küchler (1961) have attempted such a zoning based on already existing studies of various scientists and have proposed a map of the humid tropics that takes into consideration climatic and vegetational criteria. A more recent attempt was made by Troll and Pfaffen (1966) with their map, "Seasonal climates of the earth," showing the influence of various climates on their respective vegetation cover. We have based our selection of countries on these two studies.

The countries have been regrouped by continents and subcontinents. Where nothing is indicated to the contrary the same geographical units will be used throughout the entire study.

A. AFRICA

East Africa: Kenya, Madagascar, Mauritius, Reunion, Tanzania, Uganda

Central Africa: Cameroon, Central African Republic, People's Republic of Congo, Equatorial Guinea, Gabon, Zaire

West Africa: Angola and Cabinda, Benin, Ghana, Guinea, Guinea-Bissau, Ivory Coast, Liberia, Nigeria, Senegal, Sierra Leone, Togo

B. LATIN AMERICA

South America: Argentina (tropical part only), Bolivia, Brazil, Ecuador, French Guyana, Guyana, Paraguay (tropical part), Peru, Surinam, Venezuela

Central America and Caribbean: Belize, Costa Rica, Cuba, Dominican Republic, El Salvador, Guatemala, Haiti. Honduras, Jamaica, Mexico (tropical part only), Nicaragua, Panama, Trinidad and Tobago

C. ASIA

Pacific: Australia (tropical part only), British Solomon Islands, Fiji, Hawaii, New Caledonia, Papua New Guinea

Southeast Asia: Brunei, Burma, Cambodia, Indonesia, Lao, Malaysia, Philippines, Thailand, North Viet-Nam, Socialist Republic of Viet-Nam

South Asia: Bangladesh, India, Sri Lanka

In general, only those countries entirely located within the geographical tropical zone (exceptions are Australia, Argentina and Paraguay) have been included. Some other countries in the subtropical zone (part of China, Bhutan, and Nepal), partially covered by a kind of moist tropical forest, have not been taken into account. On the other land, India and Burma have been included as a whole, with their subtropical part.

The problem of vegetation classification systems

All the vegetation maps are based on a vegetation classification system elaborated according to varying criteria. Up to now, numerous researchers have developed different systems and terminologies, often based on their regional knowledge and adapted to regional conditions. The criteria mostly used were the physiognomy and structure (dense and open forest), the ecology (swamp forest, edaphic formations), the rhythm of the vegetation (evergreen, deciduous) and evolutionary aspects (primary and secondary forests). To our knowledge, there have been only two serious attempts to develop a generally recognized concept of vegetation classification: one made in 1956 for Africa which resulted in the so-called "Yangambi Classification" and the other, more recent, made by Unesco in 1973. But at the global level, scientists involved in vegetation research have not yet agreed on one uniform system for all vegetation classification work. This study does not intend to develop and propose some new approach in this field. We have to be content with selecting one classification proposal among those already existing, which can serve as a coherent basis for our work.

The Unesco international classification system and mapping of vegetation (1973)

For the purpose of this report we have chosen the Unesco classification. It is the most recent system, proposed by a group of international experts, and is based on climax vegetation types where this is practical. It is basically physiognomic-structural in character, with supplementary ecological information integrated into its various categories and applicable to natural and semi-natural vegetation.

FOREST TYPES TO BE INCLUDED IN THE MOIST TROPICAL FOREST

The following terms and definitions are used in the Unesco system:

OMBROPHILOUS: mainly broad-leaved evergreen trees, forest canopy remains green all year though a few individual trees may be leafless for a few weeks

EVERGREEN SEASONAL: mainly broad-leaved evergreen trees, foliage reduction during dry season is noticeable

SEMI-DECIDUOUS: most of the upper canopy drought-deciduous, many of the understorey trees and shrubs evergreen

DECIDUOUS: majority of trees shed their foliage simultaneously in connection with the unfavourable season and foliage is shed regularly every year

We have included in this study the following forest types described in the Unesco system:

- tropical ombrophilous lowland forest
- tropical semi-deciduous lowland forest
- tropical ombrophilous sub-montane forest
- tropical semi-deciduous montane or cloud forest
- tropical ombrophilous montane forest
- tropical lowland and sub-montane evergreen needle-leaved forest
- tropical ombrophilous alluvial forest
- tropical ombrophilous swamp forest (without mangroves)
- tropical montane and subalpine evergreen needle-leaved forest
- tropical evergreen bog forest
- tropical drought-deciduous broad-leaved low-land and submontane forest
- tropical evergreen seasonal lowland forest
- tropical evergreen seasonal submontane and montane forest
- tropical drought-deciduous montane (and cloud) forest

Based on the delimitation of the humid tropics, the above listed forest types cover a much wider range than originally defined by the guidelines described at the beginning of this article. This makes possible the inclusion of important teak: and sat forests of Asia in this global review. Because the different altitudinal types are not clearly distinguished in the available data, it has been necessary to include the whole range of types from lowland to submontane and montane forests. For some countries in Central America, East Africa and Southeast Asia, mixed forests of broad-leaved and needle-leaved species are reported to exist. They are included in estimates of the climax and of the present area of tropical moist forests. Areas of needle-leaved forests as well as some secondary type of human induced fire-climax formation after destruction of the broad-leaved forests are included in the estimates of the climax areas.

A FOREST SEEN FROM THE SKY at last a tool that makes possible an accurate appraisal of the world's forest resources

The southwestern Atlantic coast of Senegal, including Gambia on either side of the wide, curving Gambia River (top), as it appears in imagery from a multispectral scanner mounted in a NASA Landsat satellite. This photo graph shows a transition zone between savanna land (dark grey to black), tropical forest (medium grey) and mangrove forest (light grey). It was made on February 21, 1973 in band seven, near infra red, at an altitude of 950 km and covers an area of about three-and-a-half million hectares. The density scale and numbers in the margin assist interpretation.

A FINITE RESOURCE the time has come for a global inventory

Areas which would be covered by tropical moist forests as climax vegetation under the present climatic conditions

The term climax in its various aspects has been widely discussed in past years and opinions vary. The assumption that the vegetation through different stages of evolution tends always towards a certain final stage only determined by climatic conditions (climatic climax) is too simple. The natural vegetation cover depends on the site conditions which are not only influenced by the climate. Edaphic and topographic conditions may be the decisive factors. Under extreme site conditions, the influence of the regional climate is restrained and the expected zonal vegetation can be replaced by an azonal one. For this reason, a delimitation of the climax area of the tropical moist forests based alone on climatic criteria is not realistic and would indicate too large an area. When the natural vegetation cover has been and is still being disturbed by human activities, it may be very difficult to evaluate the distribution of the former zonal natural vegetation, because under certain marginal site conditions and under the permanent influence of man, the former climatic climax vegetation may have been changed towards a new pseudoclimax vegetation which may be stable itself as long as the main stand factors are not altered. The detailed climax distribution of the tropical moist forests as formation type under present climatic conditions is therefore still imperfectly known.

It seems unlikely that the total area of the tropical moist forest has been constant even after the appearance of man. Under the influence of climatic fluctuations, the areas of the tropical moist forest have probably always undergone considerable modifications. In this context, the question of the general regression of the tropical moist forests and their replacement by savanna has provoked numerous discussions and speculations, especially for Africa. Even today, we cannot prove a massive regression of the dense forest in favour of the savanna in spite of strong arguments for it:

- a sharp borderline between the dense forest and savanna, which hardly seems to be natural;

- patches of dense forest among savanna vegetation under identical climatic conditions;

- mosaic vegetation of dense forest and savanna as transitional zone between the moist forest and the drier savannas of the Guinean type;

- presence of tall trees (single or in groups) in the savanna near dense forests.

Based on these arguments, a regression of the tropical moist forest in certain regions where it has been less stable for edaphic or climatic reasons seems probable or is evident (as in Guinea-Bissau). On the other hand, there are examples where, left under natural conditions, the tropical moist forest is progressing into savanna areas (South Cameroon, some parts of the ivory Coast). For our purpose, changes of relevance are those which have occurred in recent times and which can still be traced without any major doubts.

Results of the evaluation of the climax area of the tropical moist forests

AFRICAN REGION

Vegetation map used: Carte de la végétation au sud du tropique du Cancer, AETAT/Unesco 1959, 1:10000000. An updated edition of this map will be soon available enabling US to revise our work.

Map characteristics: It is a synthesis of various research work carried out at regional levels. It shows the actual distribution of the vegetation and not the presumed climax types.

Vegetation types included: Moist forests at low and medium altitudes, evergreen or partly evergreen; forest-savanna mosaic, patches of moist forests surrounded by savanna, probably a fire climax of the moist deciduous forests; coastal forest-savanna mosaic, resembling the preceding type.

Problems: a vegetation map at 1:10000000 cannot show details of the vegetation. It gives an overall picture of the main types. It could not be decided if the category of the "undifferentiated woodlands of relatively moist types" should be included or not. There is some evidence that most of the areas have been once covered by semi-evergreen forests as a transitional type vulnerable to fire between the evergreen and dry deciduous forests. In this study, these areas are excluded.

ESTIMATE OF AFRICAN SURFACE

Subcontinent	Total land area	Moist forest area	Forest-savanna mosaic area	Total moist forest area	Total land area
	million ha				percent
East Africa	236	13	12	25	10.6
Central Africa	408	197	72	269	65.9
West Africa	356	50	18	68	19.1
Total Africa	1000	260	102	362	36.2

LATIN AMERICAN REGION

SOUTH AMERICA

Vegetation map used: Vegetationskarte von Südamerika by Kurt Hueck, 1972, 1:8000000, used copy enlarged to approx. 1:5000000.

Map characteristics: It is based on regional and local research work and reports. It gives a very detailed picture of the actual vegetation cover broken down into 82 different types.

Vegetation types included: Evergreen tropical rainforests of the Amazon and Orinoco Basins. Evergreen montane rainforest of the lower parts of the Andes. - Evergreen and deciduous montane forest of the higher parts of the Andes. - Evergreen tropical rainforest of the Pacific, Caribbean and Atlantic Coast Region. - Deciduous mesophytic tropical forest, for example of Colombia and Paraguay.

Problems: For some savanna, palm savanna and palm forest types, it is not possible to determine the presumed climax vegetation. These types are not included in this study. In general, the map used is rather too detailed for our purpose her too detailed for our purpose.

Total land area	Lowland rain forest	Montane rain forest	Deciduous-mesophytic forest	Total moist forest area	Total land area
	million ha				percent
1401	538	62	150	750	53.5

CENTRAL AMERICA AND THE CARIBBEAN

Vegetation map used: Vegetation map prepared for the Soil Map of the World (FAO).

Map characteristics: It is mainly based on climatic, structural criteria and the rhythm of the vegetation. Where it is possible, it shows the presumed climax vegetation cover.

Vegetation types included: Evergreen rain forest; semi-evergreen forest; moist deciduous forest.

Problems: The vegetation map actually used seemed to be more appropriate than the ecologically oriented detailed life zone maps of Holdridge and Tosi in spite of their importance.

ESTIMATE OF CENTRAL AMERICAN AND CARIBBEAN

Total land area	Evergreen rain forest	Semi-evergreen forest	Moist deciduous forest	Total moist forest area	Total land area
	million ha				percent
166	27	26	-	53	31.9

ASIA AND PACIFIC REGION

Vegetation maps used: Australia - Welt Forst Atlas, Vegetationsformen von Australien, 1954, 1:13000000. Papua New Guinea - Vegetation map of Malaysia, van Steenis, 1958, 1:15000000. This map was produced before the setting up of Malaysia as a political unit and it covers the biogeographic area which used to be known as Malaya. Biogeographers have now decided to call this "Malesia" in order to avoid misunderstanding.

Map characteristics: They give the actual vegetation cover broken down into different structural types.

Vegetation types included: Welt Forst Atlas - tropical rain forest. Vegetation map of Malaysia - tropical rain forest (Dipterocarpus, Agathis, Borneo Ironwood). Also casuarina and pine forest; fresh water swamp and peat forest, secondary forest; monsoon or seasonal forest, teak forest.

Problems: For Hawaii, New Caledonia, Fiji and the British Solomon Islands, no relevant information could be found. An arbitrary estimate has been made, based on local climatic criteria.

ESTIMATE OF PACIFIC SURFACE

Total land area	Evergreen rain forest	Total moist forest area	Total land area
	million ha		percent
374	48	48	12.8

SOUTHEAST ASIA

Maps used, their characteristics and vegetation types included: Burma, Thailand - Welt Forst Atlas, Vegetationsformen, 1:8000000. The actual vegetation cover and land use is shown. The following vegetation types are included: evergreen rain forest, semi-deciduous and moist deciduous forest. For Burma, a part of these forests is located outside the tropics. Indonesia, Malaysia, Philippines Vegetation map of Malaysia, van Steenis, 1958, 1:5000000. (For details see under Papua New Guinea). Cambodia - Carte internationale du tapis végétal, P. Legris, 1970, 1:5000000. Based on climatic, physiognomic and structural criteria, the actual vegetation cover is shown. On a supplementary map, the evolutionary tendencies of the vegetation under elimination of human influences have been developed. Evergreen and semi-deciduous forests are singled out. Lao, both Viet-Nams - Vegetation map of the world, Ministry of Geology, Moscow, 1:25000000. Mainly based on climatic criteria and presents a very broad zonification which does not give much satisfaction. Only evergreen forests are singled out.

Problems: The lack of a coherent vegetation map based on the same criteria for the whole region constituted a big handicap. The various maps used are very different in their characteristics.

ESTIMATE OF SOUTHEAST ASIA'S SURFACE

Total land area	Evergreen rain forest	Semi-evergreen forest	Moist deciduous forest	Total moist forest area	Total land area
	million ha				percent
448	237	15	50	302	67.4

SOUTH ASIA

Maps used, characteristics and vegetation types included: Bangladesh - South Asia Vegetation Map prepared for the Soil Map of the World 1:10000000. Based on the habitat (climatic or edaphic), the physiognomy and structure of the vegetation. The tropical wet evergreen forests are singled out. India - the Vegetation Map of India Champion/Seth, 1968, approx. 1:19000000. Shows the different forest types based mainly on climatic criteria. Tropical wet evergreen, semi-evergreen and the moist deciduous forests are singled out. Sri Lanka - Carte internationale du tapis végétal, P. Legris, 1964, 1:1000000 and 1970, 1:5000000. (For details see under Cambodia).

Problems: Same as Southeast Asia.

ESTIMATE OF SOUTH ASIA'S SURFACE

Total land area	Wet evergreen forest	Semi-evergreen forest	Moist deciduous forest	Total moist forest area	Total land area
	million ha				percent
348	12	15	58	85	24.4

Present areas of tropical moist forests

Based on the available information at FAO Headquarters, an attempt has been made to evaluate the present areas of moist tropical forests for the countries included which conform to the definitions of forest types already discussed. Despite the fact that more and more areas are covered by forest inventories, information of statistical value and the required accuracy for whole countries is still lacking. Mainly carried out for economic development programmes, these inventories deal with selected areas which quite frequently are not representative of the whole forested area. The terminology and classifications used are very heterogeneous and difficult to standardize. A breakdown in distinct vegetational forest types is often not given. The information for different countries thus varies considerably in respect of quantity and quality. The figures include all those areas which are reported to be closed forests. A breakdown between already harvested forests and untouched forests is not possible. The figures for the total forest land area are quoted from the FAO Production Yearbook, Vol. 27, 1973. Their accuracy and actuality vary considerably. Where two or more variable figures were available the lowest one was taken into account. The main sources of information were: land use and vegetation maps; country, progress and project reports; country statistics, supplemented occasionally by oral statements. The estimates are regrouped by subcontinents and finally summarized at a global level.

EAST AFRICA

Total land area	Total forest land	Moist forest area (actual)	Total land area	Total forest land area
	million ha		percent
236	55	7	3	12.7

A breakdown into ombrophilous and evergreen-seasonal forests and into different attitudinal types in general is not given. All the lowland forests border directly on the savanna vegetation as a supposedly degraded stage of the semi-deciduous/moist deciduous forests which seem to exist only in very small relics throughout Africa.

For Kenya, Tanzania, and Uganda, the areas included are covered by tropical rain forests from low altitudes to montane ranges, from evergreen to semi-deciduous and from broad-leaved to needle-leaved (Podocarpus forests). Most of the remaining forests are declared forest reserves.

The accuracy of the figures for Madagascar is rather doubtful. The lowest figure (3.6 million ha) mentioned in recent reports is used here and includes very little or no degraded rain forest of the evergreen to the semi-deciduous types of the low altitude and submontane range.

CENTRAL AFRICA

Total land area	Total forest land	Moist forest area (actual)	Total land area	Total forest land area
	million ha		percent
408	176	149	36.5	84.6

The estimates of the closed forest area in Zaire vary considerably and are difficult to judge. The lowest figure (90 million ha) found in recent reports has been included. A recent review was carried out by ITALCONSULT on the basis of satellite photographs over 9.14 million ha (39% of the total land area) of the western part of Zaire.

ZAIRE, WESTERN PART

Dense equatorial forest	Gallery forest	Forest on hydromorphic soils	Total area of forest
	million ha		percent
36.9	166	5.9	59.4

This total area corresponds to 65% of the study area or 25.3% of the total land area. Taking these figures into consideration, the estimate of 90 million ha for the whole country may be quite realistic.

WEST AFRICA

Total land area	Total forest land	Moist forest area (actual)	Total land area	Total forest land area
	million ha		percent
356	103	19	5.3	18.4

The figures for Ghana (2 million ha) are quite uncertain because the area of unreserved moist forests is not known. It is reported that these unreserved forests are decreasing rapidly and will disappear in a few years.

The figure included for Ivory Coast in this study (5.4 million ha) is the most recent estimate by SODEFOR and takes only forest stands of 100 ha or more into account. Smaller patches are not included but are estimated to cover about 1 million ha. Most of these scattered forests will disappear in the near future and have not been taken into account.

LATIN AMERICA

Total land area	Total forest land	Moist forest area (actual)	Total land area	Total forest land area
	million ha		percent
1401	864	472	33.7	54.6

Relevant information for the countries of - this subcontinent is scarce and most figures are unreliable. Often, two different country reports for the same area present differing figures without stating the sources on which the estimates were based. Other figures are obsolete, ignoring the recent changes, reportedly enormous in certain regions. A breakdown into different types is seldom made, with one exception: different altitudinal types are sometimes distinguished.

For Argentina, only the Tucuman-Bolivian forest stand has been included. The reported figures vary from 2.6 to 3.6 million ha. The higher figure integrated into this study seems to include some type of moist transitional forest to drier types.

The estimates for the Amazon Basin in Brazil still vary from 200 to 300 million ha (a figure of 250 million ha is included in this study). The RADAM project is preparing a phyto-ecological map of the Amazon Basin which may clarify the situation in the near future and produce more reliable figures.

For Peru, two different figures are reported: 65 and 79 million ha. We chose the lower figure, based on the vegetation map of K. Hueck, which seems more probable.

CENTRAL AMERICA AND THE CARIBBEAN

Total land area	Total forest land	Moist forest area (actual)	Total land area	Total forest land area
	million ha		percent
166	100	34	20.5	34.0

The figures given for the countries of this subcontinent are not satisfactory. They are often contradictory and obsolete. It seems that the forest areas are decreasing rapidly.

The forest types included range from ombrophilous lowland to submontane and montane types, from evergreen-seasonal to semi-deciduous (moist deciduous forests are only reported for Belize), from broad-leaved to needle-leaved and mixed types. Where clearly indicated, needle-leaved forests have been excluded.

PACIFIC

Total land area	Total forest land	Moist forest area (actual)	Total land area	Total forest land area
	million ha		percent
374	78	36	9.6	46.2

The forests of Papua New Guinea cover alone 94.9% of the actual total moist forest area. The rest is scattered in stands of several 100000 ha or less over Australia, Fiji, New Caledonia, and the British Solomon Islands. The information seems to be quite accurate. Practically all of these forests are some type of tropical evergreen forest.

SOUTHEAST ASIA

Total land area	Total forest land	Evergreen forest	Moist deciduous forest	Total moist forest area	Total land area	Total forest land
	million ha				percent
448	268	146	41	187	41.7	69.8

In general, the figures given in various reports are homogeneous and seem to be up-to-date in spite of the rapid changes occurring in some countries of this subcontinent.

SOUTH ASIA

Total land area	Total forest land	Evergreen forest	Moist deciduous forest	Total moist forest area	Total land area	Total forest land
	million ha				percent
348	71	5	26	31	8.9	43.7

The information for India and Sri Lanka seems to be accurate whereas data for Burma are scarce and outdated. The moist deciduous forests of India are estimated to cover 74.4% of the total moist forest area of the subcontinent.

ESTIMATE FOR THE ENTIRE PRESENT AREA OF TROPICAL MOIST FORESTS

Subcontinent	Total land area	Total forest land	Moist forest area (actual)	Total land area	Total forest land	Total moist forest area (actual)
	million ha			percent
East Africa	236	55	7	3	12.7	0.7
Central Africa	408	176	149	36.5	84.6	15.9
West Africa	356	103	19	5.3	18.4	2.0
Total Africa	1000	334	175	17.5	52.4	18.7
Latin America	1401	864	472	33.7	54.6	50.5
Central American/ Caribbean Region	166	100	34	20.5	34.0	3.6
Total Latin America	1567	964	506	32.3	52.5	54.1
Pacific Region	374	78	36	9.6	46.2	3.8
Southeast Asia	448	268	187	41.7	69.8	20.0
South Asia	348	71	31	8.9	43.7	3.3
Total Asia	1170	417	254	21.7	60.9	27.1
TOTAL HUMID TROPICS	3737	1715	935	25.0	54.5	100.0

These figures must be considered a rough estimate based on present information which is still incomplete, in particular for Brazil, Peru and Zaire where 405 million ha or 43.4% of the total world area of moist tropical forest seem to be situated.

The 935 million ha of present tropical moist forest area are just one third of the world's total area of closed forests estimated by Persson (1974) at 2800 million ha.

Estimates of the growing stock and the annual growth potential of the present tropical moist forests

An estimate of the total growing stock and of the exploitable growing stock of the forests included in this study may be an interesting indicative figure giving the overall wood potential of these zones. But we have to take into account that only a part of this area is actually or will in the future be opened and developed for timber production (exploitable forests). The remaining part (unexploitable forests) are areas where the forest should be maintained for ecological reasons, where accessibility is not possible for technical reasons or where the wood potential is not attractive enough for investments for development. It seems that this concept of exploitable and unexploitable forests is becoming recognized more and more. However, with the information available, it is not yet possible to give a breakdown into these two categories.

Many inventories in tropical forests have already been carried out with varying concepts and methods, in different regions and forest types. They have produced a large quantity of different results which can hardly be compared without any special research. They usually enumerate only that part of the forest which is of interest for immediate harvesting of timber. Their results, therefore, are not or are only partially representative of the whole forested area, because they concern mainly preselected areas for economic and development reasons.

Total growing stock

We tried to give an estimate of the total growing stock for Africa, Latin America and Asia separately, based on the estimated areas of still existing moist tropical forests and the results of some selected inventories. Unless otherwise specified, the following volume concept has been adopted: gross stem volume of sound trees, over bark, min. DBH 10 cm or above buttresses and all species. Various authors agree that as a rule of thumb the proportion of bark for all species and diameters can be regarded to be of the order of magnitude of 10% in tropical moist forests.

We selected, for each continent, several inventory reports presenting results fitting this volume concept, inventories representative of high, medium and low volume forest types. In general, there are quite a few inventories taking into account all tree species with particular attention to trees with a min. DBH of 10 cm and more. This concept is relatively new and has replaced the previous one where only a selected number of tree species and diameter classes were counted. Reports from the following countries have been examined: Gabon, Ivory Coast, Nigeria, Ecuador, Guatemala, Peru, Venezuela, India, Malaysia, Philippines, and Sri Lanka. The summarized results follow.

GABON

INVENTORY A

Area: 72000 ha.

Forest type: tropical lowland evergreen forests; 80% of this area is classified as dense forest, the rest as exploited, degraded, secondary growth or swamp forests.

Gross volume: 358 m³/ha all species with a DBH of 15 cm and more; 5 m³/ha or 1.5% of the volume mentioned above can be added for all trees with a DBH of 10-15 cm; so the total volume amounts to 363 m³/ha.

INVENTORY B

Area: 100000 ha.

Forest type: see left; stands of secondary bush and regenerating forests represent more than 50% of the total area.

Gross volume: 261 m³/ha + 4 m³/ha = 265 m³/ha.

IVORY COAST (southwest region)

Area: 1.9 million ha

Forest type: tropical lowland evergreen, semi-evergreen and transitional forests; 20% of the total area affected by exploitation.

Gross volume/ha: 396 m³, min. DBH 15 cm, 6 m³/ha or 1.5% of the above mentioned volume may be added for the trees with a DBH of 10-15 cm; the total gross volume/he may be 402 m³.

NIGERIA (southeast region)

Area: 400000 ha.

Forest type: moist lowland evergreen forest with a transitional type in the drier parts of the area; 73% of the area is classified as closed high forest, the rest as exploited, low forests, bush and swamps.

Gross volume/ha: 186 m³ with a min. DBH of 15 cm; 3 m³/ha or 1.5% of the reported volume may be added for trees with a min. DBH 10-15 cm; the total gross volume/ha may be 189 m³.

Based on these selected examples for Africa it seems justifiable to adopt an average gross volume per ha of 175-225 m³ for this continent.

ECUADOR (northwest region)

Area: 823000 ha.

Forest type: tropical lowland and submontane evergreen forest.

Gross volume/ha: 140 m³.

GUATEMALA (Peten region)

Area: 2.2 million ha.

Forest type: tropical moist forests; 77% of the area is classified as commercial or potentially commercial forests of which 12% with tree heights of less than 15 m; 23% of the area is classified as non-commercial forests with tree heights of less than 15 m.

Gross volume/ha: 147 m³.

PERU (region of Pucallpa, Bosque de Humboldt)

Area: 200000 ha.

Forest type: tropical lowland and submontane moist forests; the 200000 ha represent the better part of the total area of 645000 ha.

Gross volume/ha: 162 m³.

VENEZUELA (Guyana region)

Area: 1.64 million ha.

Forest type: tropical lowland evergreen and semi-evergreen forests of low altitude.

Gross volume/ha: 91 m³.

Based on these examples an average gross volume/ha of 100-150 m³ may be assumed for Latin America.

INDIA (southern zone)

The volumes given in the inventory reports are under bark including all trees with a DBH of 5 cm and more, and they have been converted to volumes over bark including trees with a DBH of 10 cm and more in the above table. They are high and are confined to untouched stands.

Area	Forest types	Gross volume/ha
98000 ha	evergreen	421 m3
104000 ha	moist deciduous	190 m3
72000 ha	semi-evergreen	309 m3

MALAYSIA (West Malaysia)

Area: 7.4 million ha.

Forest type: mixed tropical moist forests with Dipterocarpus as the main species group; the total area is composed of the following forest types: 11% superior hill forests, 16% good hill forests, 20% moderate hill forests, 17% logged hill forests, 23% disturbed hill forests, 3% under shifting cultivation, 6% poor hill forests, 4% upper and montane hill forests.

Gross volume/ha: 237 m³/min. DBH 15 cm, without bark; 3 m³ may be added for the trees with a DBH of 10-15 cm and 20 m³ for the bark; the total gross volume may be 260 m³/ha.

PHILIPPINES (Mindanao)

Area: 6.3 million ha.

Forest type: Dipterocarpus forests; 43% of the area is old growth forests, 27% residual or young growth forests and 30% is reproductive bush.

Gross volume/ha: 187 m³.

SRI LANKA (wet zone)

Area: 15000 ha.

Forest type: tropical lowland and montane moist evergreen forest; 33% of the area is classified as high and medium yield, 50% as low yield and 17% as non productive.

Gross volume/ha: 205 m³.

In view of these examples an average gross volume of 185-225 m³/ha may be assumed for evergreen and 100-150 m³ for moist deciduous forests in Asia.

Based on the estimated areas of still existing moist tropical forests and to the extent that the inventoried areas described constitute a representative sample within each continent, the total growing stock may be estimated as follows:

Continent	Area (million ha)	Growing stock (million m3)	Percent of global growing stock
Africa	175	30600 - 39400	24
Latin America	506	50600 - 75900	44
Asia
- Evergreen and semi evergreen	187	34600 - 42100	26
- Moist deciduous	67	6700 - 10100	6
Total tropical moist forests	935	122500 - 167400	100

Total exploitable growing stock

By the total exploitable growing stock we understand the part of the total growing stock contributed by all the trees with a diameter of more than 50-60 cm of which only a fraction is actually harvested, for these reasons:

- only a part of the total area still under tropical moist forests is or will be exploited,
- at present only a selected number of tree species is harvested and only selected logs are utilized.

Rollet (1974) gave indicative average figures of stand basal area per ha in moist lowland evergreen tropical forests.

All species with a diameter of more than 55-60 cm:

Africa	6.3 - 13.9 m2/ha
Latin America	5.7 - 10.3 m3/ha
Asia	8.2 - 20.0 m2/ha

The total exploitable growing stock (estimated to be at least ten times the stand basal area):

Africa	63 - 139 m3/ha av.: 100 m3/ha
Latin America	57 - 103 m3/ha av.: 80 m3/ha
Asia	80 - 200 m3/ha av.: 140 m3/ha

These figures cover quite a wide range of stand density and give, compared to the total growing stock per ha evaluated in the preceding section, reasonable estimates for the total exploitable growing stock for Africa and Asia. As for Latin America, the figure seems rather high and it is thought advisable to reduce it to 60 m³/ha. Concerning the moist deciduous tropical forests of Asia a figure of 80 m³/ha seems appropriate.

Continent	Area (million/ha)	Exploitable growing stock
		million m3
Africa	175	17500
Latin America	506	30400
Asia
- evergreen and semi evergreen	187	26200
- moist deciduous	67	5400
Total tropical moist forests	935	79500

The total exploitable growing stock thus corresponds to 65% of the lower estimate of the total growing stock (122500 million m³) and to 48% of the upper estimate (167400 million m³). The estimated roundwood production for 1973 (FAO Yearbook of Forest Products 1973) for the countries with tropical moist forests was approximately 151 million m³ or only 0.2% of the estimated total exploitable growing stock.

The estimated volume figures are determined from the inventory results of relatively small selected areas which have been extrapolated to the rest of the total area of tropical moist forests not covered by adequate surveys (Congo Basin, Amazon Basin, Indonesia and Papua New Guinea). Volume figures, however, in FAO regional and world resource summaries and other publications, are usually official government returns and are often based on other concepts of volume (higher diameter limits or without bark). They may differ therefore in some degree from the global figures given here.

Estimate of the annual growth potential

It is assumed that we can expect an annual increment only from disturbed or managed forests, whereas in undisturbed forests having reached their climax equilibrium it may be nil. Accordingly, it would be more appropriate to speak of the annual growth potential of the tropical moist forests, of which we know very little so far.

Based on general considerations in different forests, the annual growth potential can be estimated to be 1-2 m³/year/ha for Africa, 1-3 m³/year/ha for Latin America and 2-4 m³/year/ha for Asia.

Continent	Million ha	Total annual growth potential Million m3/year	Percent of growth potential as roundwood exploited
Africa	175	175 - 350	9 - 17.9
Latin America	506	506 - 1518	1.7 - 5.0
Asia	254	508 - 1016	9 - 18.1
Total world	935	1189 - 2884	5.2 - 12.1

Two things should be noted here. The first, that the relationship between the annual growth potential and the reported annual exploitations for Asia and Africa is quite similar whereas the figures for Latin America are lower, indicating a larger underexploited potential of forests for this continent. Also, that the real annual exploitation including all the illegal activities not reflected in the statistics must comprise far larger quantities, in particular for Latin America.

Past and present conversions in the zone of tropical moist forests and area involved

More and more of these ecosystems are disturbed by human activities in their natural evolution towards a stable equilibrium: repeated harvesting followed by cultivators invading illegally the areas through the logging roads; complete exploitation and conversion to forest monoculture plantations; clearing of forests for colonization and agriculture of a temporary or permanent character. The impact of these activities leads quite often to the complete destruction of large forest areas. Although we lack precise information, we can take it for granted that tropical moist forests are undergoing an accelerating process of depletion and regression.

Quantitative information at a global level is still very scarce and would need repeated monitoring of the forest areas of whole regions and continents. From time to lime, the results of case studies for single countries with a big regression rate (as in Ivory Coast) are used for very pessimistic extrapolations about the fate of tropical forests at a global level. In spite of the importance of the question, only a few countries concerned seem to have the situation under control and are able to present reliable figures. But the recently approved pilot study for a global tropical forest cover monitoring project, which aims to survey, on a permanent basis, the development of the tropical forest zones by modern remote sensing techniques is a step in the right direction. However, more and more countries are becoming aware of this critical situation and, for this reason, a noticeable area of tropical moist forests has already been declared forest reserves by governmental acts. In many cases, the services in charge are still too weak to carry out efficiently their new task. A global study of these reserves, the different types and their management, would be necessary. They are at present the only guarantee that some of these tropical moist forests may be saved for the future.

Comparison of the climax and actual areas of tropical moist forests

Such a comparison may be useful to give us some clue about the time involved in the regression of these forests. But it cannot clarify the relationship between these changes and the time-scale. It is therefore not possible to calculate, using these figures, a rate of regression and to make any extrapolations for future developments.

Subcontinent	Moist forest climax area	Moist forest actual area	Regression	Regression in percentage of climax area
	million hectares
East Africa	25	7	18	72.0
Central Africa	269	149	120	44.6
West Africa	68	19	49	72.0
Total Africa	362	175	187	51.6
South America	750	472	278	37.1
Central America	53	34	19	35.8
Total Latin America	803	506	297	37.0
Pacific Region	48	36	12	25.0
Southeast Asia	302	187	115	38.1
South Asia	85	31	54	63.5
Total Asia	435	254	181	41.6
TOTAL WORLD	1600	935	665	41.6

Based on these figures, a regression of 2/5 of the world's total area of tropical moist forests may be assumed. For the different subcontinents, these areas vary between 72.1% for East Africa and 25.0 for the Pacific Region. Furthermore, it is noticeable that Africa has already lost over 50% of its tropical moist forests, mainly the areas with a more or less pronounced drought period enabling man to attack and clear the forests by fire for his own purposes.

Direct conversion of once-harvested natural forests to monoculture plantations

Until recent times, this was hardly being applied on a large scale for various reasons:

- still enough abandoned agricultural land available for reafforestation with fast growing species
- only once-exploited forests contain still sufficient timber for a profitable second cycle
- existence of large areas of easily accessible and not yet exploited forests.

This favourable situation may change to a certain extent in the near future. Accessible forests are being increasingly overexploited and contain only species of secondary value, and their natural regeneration is slow and uncertain. What will be the appropriate land use for such areas? Under certain conditions, a complete clearing with consecutive replanting by fast growing species could become an attractive solution for countries with limited forest resources. For instance, such a large scale project of direct conversion is under discussion for 140000 ha of natural forests in Liberia. In the Jari River area Pare/Brazil, the National Bulk Carrier Corporation, New York, has been operating for the last eight years a large scale industrial forest plantation scheme presently covering 100000 ha.

No noteworthy utilization of the virgin stands was reported and the areas were cleared by fire. In Gabon, a similar project covering 25000 ha is reported to be under way. In other countries, very often the combined agrosilvicultural plantations after clear-felling of the natural forests result in large monoculture plantations.

Definitive clearing of natural forests of permanent colonization and agriculture

Such activities should be planned carefully and can only be initiated where the climatic and edaphic conditions are favourable for human settlements. In Africa, there is at present no evidence of any larger colonization or resettlement projects within the forest area. But through the opening of remote areas for logging, large areas are afterwards taken over by illegal settlers. As an example of this development an indicative figure can be given from the Ivory Coast where during the period 1956-66 ha of dense forest per 5 m³ of roundwood exploited for export was cleared completely by consecutive settlers. The same relationship may be applicable for all the countries with a rather high demographic density and growth rate. Unfortunately, it is not possible to assess these changes on the basis of the information available at the global level.

Under growing demographic pressure, governments in Latin America are forced to release more and more forest areas for permanent settlement and agriculture. Quite often, such action takes the form of planned, long-term government-sponsored projects. The following table gives the projected figures of such activities for some countries in the period 1962-85.

PROJECTED NEW LAND DEVELOPMENT AND INVESTMENTS IN SEVEN TROPICAL LATIN AMERICAN COUNTRIES, 1962-85

Country	Development of new land (million ha)					Total investment ($ million)3
	Crop land from pasture1	Crop land from forest	Total	Pasture from forest	Total forest clearing
Colombia	1.0	1.0	222.0	6.6	7.6	440.0
Ecuador	0.3	0.3	0.6	0.9	1.2	92.5
Peru	0.4	0.4	0.8	0.7	1.1	70.0
Bolivia	0.2	0.1	20.3	4.2	4.3	120.0
Paraguay	0.2	0.2	0.4	3.7	3.9	147.5
Brazil	4.7	4.7	9.4	12.5	17.1	2120.0
Venezuela	0.2	0.2	20,4	3.8	4.0	270.0
Total	7.0	6.9	13.9	32.3	439.2	3260.0

¹ Assumption that 50 percent of new crop land will be derived from existing pasture.

² Total area in non-irrigated crops in Colombia is projected to increase by 1.7 million ha the additional 0.3 million compensate for projected irrigation of currently non-irrigated pasture and crop lands increases of 0.3 million ha are made for both Bolivia and Venezuela in similar compensation.

³ investment in development of new lands not associated with irrigation, drainage or flood control, at 1962 prices.

⁴ Annual rate of forest clearing would be 1.75 million ha.

Source: IWP Vol. II, FAO, Rome, 1968.

In 1969, an area of 10 million ha of natural forests or 2% of the still existing area in Latin America was reported to have been cleared for agricultural settlements and by uncontrolled fire (10th Meeting of the Latin American Forestry Commission, 1969). A definition of the term "natural forests" is not given.

Brazil warrants a separate mention, with its activities in the Amazon Basin. Hueck (1966) reports an area of 3 million ha of forest lost to agriculture in the Belém-Bragança region during recent years. A comparison of the road networks in the Amazon region for 1964 and 1974 gives some indication of the changes planned or now taking place. Between 10000 and 20000 km of roads have been constructed, are under construction or are planned during this period. Assuming that along all these roads a strip of 4 km (1 walking hour) on both sides would be cleared in the first stage, an area of 0.8-1.6 million ha or 0.3 to 0.6% of the total estimated area would be taken over per year. Similar developments are occurring in Southeast Asia, where in the Kalimantan region of Indonesia large areas of natural forests are replaced by rice fields, and 18 million ha of forest land will be transferred to agriculture over the next decade. Malaysia is expected to hand over to agriculture 2 million ha of forest during the next 20 years.

HOW TO SAVE TROPICAL MOIST FORESTS FOR THE FUTURE? global studies are the only guarantee

These are only some examples highlighting the relative importance of the areas involved in planned transformation processes.

Shifting cultivation and conversions through forest exploitation

Excepting some very vague estimates, such as 100 million ha of tropical high forest south of the Sahara being lost to shifting cultivation (Soils Bulletin No. 24, FAO, 1973), the best source of quantitative information is still the FAO study presented in 1969 to the 2nd Session of the Committee on Forestry Development in the Tropics. Based on a questionnaire sent to 80 countries in the tropics, answers from 20 could be fully or partially included. It became evident once more that the situation was very complex. The main result of interest here is the estimate that approximately 2.1% of the total forest area are cleared per year. A breakdown into forest types affected is not given. Based on these estimates an approximate area of 24 million ha of the total forest area of the countries included in our article is to be cleared (Production Yearbook, FAO 1973). It may be assumed that between 20-40% of these clearings affect intact natural forests or approximately 5-10 million ha equal to 0.5 to 0.6% of the still existing area of tropical moist forests.

TOTAL FOREST AREA AFFECTED BY EXPLOITATION DURING THE PERIOD 1964-73

Subcontinent	Roundwood production 1964-73 (1000 m3)	Percent of total roundwood production	M3 roundwood extracted by ha (estimate)	Area affected by exploitation (million ha)
	million hectares
East Africa	62132	5.5	10	6
Central Africa	140340	12.4	10	14
West Africa	99770	8.8	10	10
Total Africa	302242	26.8		30
South America	190869	16.9	10 - 30	6 - 19
Central American/ Caribbean Region	20076	2.3	10 - 30	1 - 3
Total Latin America	216945	19.2		7 - 22
Pacific Region	6632	0.6	30	0.2
Southeast Asia	504361	44.7	30 - 80	6 - 17
South Asia	97560	8.6	30	3
Total Asia	608553	54.0		9 - 20
Total world	1127740	100		46 - 72

INCREASE IN AREAS AFFECTED THROUGH EXPLOITATION
(based on roundwood production, 1964 and 1973)

Subcontinent	Roundwood production 1964 (1000 m3)	Areas affected by exploitation (million ha)	Roundwood production 1973 (1000 m3)	Areas affected by exploitation (million ha)	Increase in percentage of areas affected 1964
East Africa	3314	0.3	4114	0.4	24.1
Central Africa	12336	1.2	15511	1.6	25.7
West Africa	8542	0.9	11755	1.2	37.6
Total Africa	24192	2.4	31380	3.2	33.3
South America	15324	0.5 - 1.5	22331	0.7 - 2.2	45.7
Central America/ Caribbean Region	2111	0.1 - 0.2	3120	0.1 - 0.3	47.8
Total Latin America	17435	0.6 - 1.7	25451	0.8 - 2.5	46.0
Pacific Region	444	(0.01)	1227	(0.03)	176.4
Southeast Asia	33006	0.4 - 1.1	80638	1.0 - 2.7	144.3
South Asia	8425	0.3	9950	0.3	18.1
Total Asia	41875	0.7 - 1.4	91815	1.3 - 3.0	119.2
TOTAL WORLD	83502	3.7 - 5.5	148646	5.3 - 8.7	78

The direct impact of exploitation varies widely and depends on the intensity of the fellings. Slightly modified forests by only one felling to nearly completely destroyed stands through consecutive felling may be found. An attempt has been made to assess the total forest area affected by exploitation during a ten-year period, based on the total non-coniferous roundwood (without fuelwood) production for 1964-73 and the average quantity extracted.

Thus an average area of 4.6 to 7.2 million ha may have been affected per year based on the officially reported production. All the illegal fellings, in particular in Latin America and Asia, are not reflected in these figures. The extracted average quantities are well known for Africa whereas the figures for Latin America and Asia are quite uncertain. Assuming an area of 100 ha cleared per 1000 ha logged (figure from an FAO Project in the Philippines), a total cleared area of 4.8-7.3 million ha for the 10-year period would result alone from the logging activities. The consecutive clearings by the infiltration of permanent settlers or shifting cultivators cannot be evaluated at global level. Only a few examples (Ivory Coast) have been examined.

It may therefore be concluded that in 1973 only 0.6-1.0% of the total actual area of tropical moist forest has been affected by the officially reported exploitations for the officially reported roundwood production (without fuelwood). The unknown areas under illegal felling activities have to be added to these figures.

Conversions through war effects

It has been estimated that during the Vietnamese war a total area of approximately 2 million ha of forests has been destroyed through chemical and conventional bombing.

Countries with reported decreasing rates in the tropical moist forest area

For the countries listed below, figures for the decreasing rates have been found in various reports. Some are official, some are based on individual estimates. In general, their quality varies considerably and is difficult to judge. They are presented here, not as officially recognized figures, but they can give some indication for the estimate of the decreasing rate at a global level. It is hoped that this list can be completed with supplementary figures.

Country	Area of forest reported to be lost per year
	million ha
Bangladesh	0.01
Colombia	0.25
Costa Rica	0.06
Ghana	0.05
Ivory Coast	0.4
Lao	0.3
Madagascar	0.3
Papua New Guinea	0.02
Philippines	0.26
Thailand	0.3
Venezuela	0.05 (?)
North Viet-Nam	0.01
Malaysia	0.15
Total	2.16

In most cases, the character of the clearings is not specified. These countries still contain 163 million ha of moist tropical forests. The various repeated losses reach at least 2 million ha every year or 1.2% of the actual forest area. Extrapolated to all countries and to the total actual area of tropical moist forests, we obtain an estimate of approximately 11 million ha as a minimal annual decreasing rate, a figure which will certainly increase progressively in the future, if no firm steps are taken to stop this development.

Reliable information on the regression rates of the tropical moist forest are still very scarce. Based on the information available, it is only possible to give some extrapolated estimates at the global level which can be summarized as follows:

- between 1964-73 an average area of 5.0-7.5 million ha per year at least has been affected by exploitations officially reported in the statistics;

- in 1964 an area of at least 4.0-5.5 million ha has been affected by exploitations; by 1973, this affected area had increased to 5.5-9.0 million ha;

- reported regression rates per year of the dense tropical moist forests for some 13 selected countries add up to 2.16 million ha or 1.2% of the still existing forests of these countries; extrapolated to the whole zone a regression rate of approximately 11 million ha per year may be assumed.

Present land use of the areas formerly covered by moist tropical forests

After definitive clearing of natural forests, the following land uses may occur: permanent settlement and agriculture for subsistence and cash crops, temporary settlements and agriculture under various forms of shifting cultivation. For most of the countries under review only a part of their total area is located within the tropical moist forest climax zone. Based on the information available it is hardly possible to study the actual land use of these selected areas alone. For this reason, we present, based on an arbitrary decision, a summarized review of the actual land use for the countries with more than 50% of their total land area within the tropical moist forest zone.

For Latin America, a review of the estimated land use within the humid tropics is taken from an FAO report. These figures are supplemented with the estimates of this study on the actual and climax area of tropical moist forests.

Finally, some selected countries with more detailed information available are presented as case studies.

Areas under shifting cultivation

Various estimates and guesses have been made under different assumptions for the areas under shifting cultivation at a global level. The most recent gives an area of 3600 million ha of the world's area, or 25% of the cultivated soils, feeding 250 million people or 7% of total mankind (Soils Bulletin No. 24, FAO, 1974). It did not give a breakdown into different vegetational or geographical zones, nor did it show how this figure was obtained. These reports hardly ever specified what type of vegetation is affected, nor was it clear whether the figures meant the total area altogether or only a section involved in some shifting cultivation activities; it therefore seems impossible to extract relevant figures from such inadequate data. The FAO study on shifting cultivation already mentioned estimates that an average area of 17% of the total forest area is under shifting cultivation, of which 21% in primary and 42% in secondary high forests. Based on these figures, an estimate for the countries included here gives the following results:

ESTIMATE OF AREAS UNDER SHIFTING CULTIVATION
(million ha)

Total forest area	17% of total forest area	Primary forest under shifting cultivation (21% of 304)	Secondary forest under shifting cultivation (42% of 304)	Total high forest under shifting cultivation
11787	304	64	128	191

¹ Production Yearbook, 1973: Forest and Woods.

These 191 million ha would represent 20.5% of the total actual area of tropical moist forest, which seems to be rather too high a figure.

APPROXIMATE AREAS OF DIFFERENT LAND USE OF THE ESTIMATED TROPICAL MOIST FOREST CLIMAX AREA IN LATIN AMERICA
(in million ha)

Country	Crops1	Pasture2	Bush fallow3	Moist forest (actual)4	Other land use5	Moist forest climax6
	a6	b6	c6	d6	e6	f6
Central America (without Mexico)	0.38	0.75	-	22.9	22.67	46.7
Bolivia	0.2	4.71	0.7	20.0	9.79	35.4
Brazil	17.22	82.2	-	250.0	114.78	464.2
Colombia	1.55	8.45	-	53.0	8.9	71.9
Ecuador	0.87	0.79	-	9.1	6.54	17.3
Paraguay	0.24	0.76	-	8.0	3.2	12.2
Peru	0.31	0.43	-	65.3	2.56	68.6
Venezuela	0.5	13.23	0.36	28.1	0.11	42.3
Total	21.27	111.32	1.06	456.4	168.55	758.6
	2.8%	14.7%	0.1%	60.2%	22.2%	100%

¹ No data is available on the area cropped twice, or crops under-sown. It is assumed that figures refer to cultivated area rather than area in crops.
² No distinction is made between native and improved pasture, nor is there indication in the statistics of whether any unexploited savannas are included.
³ The definition of bush-fallow is so vague that it is not normally reported in national censuses.
⁴ Figures from this study.
⁵ Figures from this study.
⁶ Figures calculated as follows: f - (a + b + c + d) = e.

Sources: Figures in ¹, ² and ³ are from the report on Public Policy for New Land Development in the Humid Tropics of Latin America, by M. Nelson, 1970.

The World's Forest - main vegetation zones

These figures indicate that only 2.8% of the given area is permanently cultivated. The type of cultivation is not distinguished. The area under the heading "other use" is, with more than 22%, significant. It can be assumed it includes large areas of fallow land and bush, as part of some shifting cultivation systems.

Actual land use of countries with 50% or more of their total land area within the tropical moist forest zone

The following countries have been included:

Africa: Cameroon, Congo, Equatorial Guinea, Gabon, Ivory Coast, Liberia, Sierra Leone, Zaire.

Latin America: Brazil, Colombia, Costa Rica, Cuba, Ecuador, El Salvador, French Guyana, Haiti, Jamaica, Nicaragua, Panama, Peru, Surinam, Trinidad and Tobago.

Asia and Pacific Region: British Solomon Islands, Brunei, Burma, Fiji, Indonesia, Cambodia, Malaysia, Papua New Guinea, Philippines, Thailand, Sri Lanka.

The following breakdown, based on the Production Yearbook, FAO 1973, was found, in million ha, for the listed countries:

Continent	Total land	Arable permanent cropping	Percent of total land	Meadows, pastures	Percent of total land	Forest, woodlands	Percent of total land	Other land1	Percent of total land
Africa	396.3 1210.5	28.3 54.1	7.1 4.5	63.1 162.8	15.9 13.5	184.4 726.9	46.6 59.9	120.5 266.7	30.4 22.1
Latin America	1210.5	54.1	4.5	162.8	13.5	726.9	59.9	266.7	22.1
Asia	446.6	69.4	15.5	13.9	3.1	280.7	62.9	82.6	18.5
Total	2053.4	151.8	7.4	239.8	11.7	1192.0	58.0	469.8	22.9

¹ Other land includes unused but potentially productive land, built-on areas, wasteland, parks, ornamental gardens, roads, lanes, barren land, water areas, and any other land use not specifically listed under arable land and woods and forests.

1276 million ha or 62% of the total land area of the countries included are located within the tropical moist forest climax area. This figure equals 80.1% of our estimate for the total tropical moist forest climax area of 1592 million ha It can be noted that only 7.4% of the total land area is reported to be arable or under permanent cropping. The distinction between the other land use categories is very often not clear. Shifting cultivation is probably included in all of them, depending on the classifications and definitions of the individual countries.

Actual land use in some selected countries of Latin America and Asia

COLOMBIA

Total land area: 113.8 million ha; while 71.9 million ha or 63.2% are estimated to be in the tropical moist forest climax zone, the actual area of tropical moist forest is 53 million ha.

LAND USE 1960

Type	Area in million ha	Percent of total land area
Arable, land, fruit trees, vines, orchards	3.7	3.2
Meadows, permanent grassland	9.7	8.5
Woods and forests	57.3	50.5
Rough grazing land	30.3	26.6
Non agricultural land	12.8	11.2
Total	113.8	100.0

ECUADOR

Total land area: 28.4 million ha; while 17.3 million ha or 60.9% are estimated to be in the tropical moist forest climax zone, the actual area of tropical moist forest is 9 million ha.

LAND USE 1964

Type	Area in million ha	Percent of total land area
Arable land	1.8	6.3
Fruit trees, vine, bush, orchards	0.3	1.0
Rough grazing land	2.2	7.7
Woods and forests	14.8	52.0
Non-agricultural land	9.4	33.0
Total	28.5	100.0

PAPUA NEW GUINEA

Total land area: 46.18 million ha; while 37.8 million ha or 81.8% are estimated to be in the tropical moist forest climax zone, the actual area of tropical moist forest is 33.7 million ha.

LAND USE 1964

Type	Area in million ha	Percent of total land area
Arable land	0.24	0.5
Coffee	0.013	0.03
Rubber	0.014	0.03
Fruit trees, vine, orchards	0.205	0.44
Woods and forests	32.319	70.0
Rough grazing and non-agricultural land	13.389	29.0
Total	46.18	100.0

SRI LANKA

Total land area: 6.5 million ha; while 4.7 million ha or 72.3% are estimated to be in the tropical moist forest climax zone, the actual area of tropical moist forest is 0.2 million ha.

LAND USE 1956

Type	Area in million ha	Percent of total land area
Arable land	0.55	8.5
Plantations	0.77	11.9
Gardens	0.57	8.8
Shifting cultivation	0.98	15.1
Permanent meadows and grass lands	0.42	6.4
Woods and forests	2.87	44.2
Non-agricultural land	0.33	5.1
Total	6.5	100.0

AREAS UNDER MAJOR CROPS

Crop	Area in million ha
Tea	0.231
Rubber	0.230
Coconut	0.466
Paddy	0.459
Total	1.386 or 21.3% of the total land area

THAILAND

Total land area: 51.4 million ha; while 50.4 million ha or 98% are estimated to be in the tropical moist forest climax zone, the actual area of tropical moist forest is 17.8 million ha.

LAND USE 1950 AND 1963

Type	Area in million ha 1950	Percent of total land area	Area in million ha 1963	Percent of total land area
Arable land	0.8	1.6	1.6	3.2
Rice	6.0	11.6	6.6	12.8
Fruit trees and orchards	0.9	1.7	1.7	3.3
Woods and forests, rough grazing land	30.9	60.2	27.1	52.8
Non-agricultural land	12.8	24.9	14.4	27.9
Total	51.4	100.0	51.4	100.0

These examples point out that only between 0.5% (Papua New Guinea) and 16% (Thailand, rice included) of the total area is classified as arable land. Woods and forests still cover between 44% (Sri Lanka) and 70% (Papua New Guinea). The portion of non-agricultural land varies between 5.1% (Sri Lanka) and 33% (Ecuador). Therefore, it may be concluded that large areas once covered by tropical moist forests, then cleared and put under some land use, are left again and represent a potential for more appropriate use.

At present, an accurate appraisal of the climax and actual areas of the moist tropical forest at the global level, based on the information available at FAO headquarters is not yet possible.

Before closing, we would like to stress once again that precision is not the main characteristic of this report, because of the poverty of reliable data. We may, however, try to synthesize the situation as best as possible, in this way:

· The total climatic climax area of the tropical moist forests has been estimated to be approximately 1600 million ha.

· The present area of tropical moist forests (evergreen, semi-deciduous, moist deciduous) is still approximately 935 million ha or 58.5% of the climax area or 54.5% of the statistically reported total forest land area.

· The total growing stock (exploitable and unexploitable forests) may vary between 122500 and 167400 million m³.

· The total exploitable growing stock (exploitable and unexploitable forests, all species) has been estimated to be 79500 million m³.

· The total annual growth potential may be assumed at 1200 to 2900 million m³.

· The roundwood production (without fuelwood) statistically reported for 1973 was approximately 150 million m³ or 5.2 to 12.5% of the annual growth potential.

· The estimated area affected by exploitation in the year 1973 may vary from 5.5 to 9.0 million ha.

· The reported regression rate of 13 countries representing 163 million ha or 18% of the total moist forest area is more than 2 million ha per year or 1.2% of their tropical moist forest area; extrapolated to the whole area an actual regression rate of approximately 11 million ha may be assumed.

We may add that the actual conversions taking place in the zone of moist tropical forests destroy sizable areas of forest at the local or regional level, which deeply upsets their ecological conditions. But, compared with the still existing total surface of tropical moist forests, the parts affected by increasing human activities seem rather small. This means that it is still possible, taking into account all current efforts, to keep the expanses of more or less untouched, or well managed, tropical moist forests. It can be achieved through the formulation and implementation of a clear development policy. As for the actual land use of already cleared forest areas, only small parts seem to be under different kinds of permanent agriculture, while large areas are probably part of some shifting cultivation system and may be more productive under some modified and more intensive exploitation.

Finally, the importance of relevant statistical data for the sound formulation of proposals for planning and development at various levels is evident.

Not long ago FAO initiated a series of regional appraisals of forest resources. This work is still being carried out and it should be expanded, so as to help the countries concerned in building up their store of documentation. The recently approved pilot study of the tropical forest cover monitoring projects, expected to include Benin, Cameroon, Nigeria and Togo is a very important step in this direction and should be supported by all interested countries and later expanded into a world study.

Meanwhile, it would be useful to collect, review and analyse this mass of information on forest resources in the humid tropics available at FAO but not easily found in one place.

Storing of information should be centralized and improved, to render it accessible to researchers, and it should be constantly updated with recent data supplied by the countries themselves or through bilateral and multilateral projects. This systematic approach could, eventually, lead to an invaluable computerized data-bank on the world's I vest resources. A complementary step would be the preparation of a systematic and uniform assessment of the actual state of forestry, to be updated at regular intervals. The purpose of all these efforts should be to build a pool of reliable documentation, indispensable to the formulation and planning of a sound policy on research and development.

But one important aspect should be kept in mind: a data bank is only as good as the information fed into it, and the job of supplying it is up to the countries concerned.

TROPICAL MOIST FOREST CLIMAX AREA OF THE WORLD
(summarized in millions of hectares)

Subcontinent	Total land area	Evergreen rain forest	Semi-deciduous forest	Moist deciduous forest or forest savanna mosaic1	Total moist forest area	Percent of total land areas	Percent of world total moist forest area
East Africa	236	13		12	25	10.6	1.6
Central Africa	408	197		72	269	65.9	16.8
West Africa	356	50		18	68	19.1	4.2
Total Africa	1000	2602		102	362	36.2	22.6
South America	1401	600		150	750	53.5	46.9
Central America/ Caribbean Region	166	27	26		53	31.9	3.3
Total Latin America	1567	627	26	150	803	51.2	50.2
Pacific Region	374	48			48	12.8	3.0
Southeast Asia	448	237	15	50	302	67.4	18.9
South Asia	348	12	15	58	85	24.4	5.3
Total Asia	1170	297	30	108	435	37.2	27.2
TOTAL WORLD	3737	1184	56	360	1600	42.8	100.0

¹ Refers only to Africa.
² Evergreen and semi-deciduous types.
(Source: author)

SUMMARY TABLE OF TOTAL LAND AREA, MOIST FOREST CLIMAX AREA, ACTUAL FOREST LAND AREA AND ACTUAL MOIST FOREST AREA (in million ha)

(Source: author)

SUMMARY TABLE OF GROWING STOCK, GROWTH POTENTIAL, ROUNDWOOD PRODUCTION 1964 AND 1973 AND AREAS AFFECTED

(Source: author)

STUDENT AT NATIONAL FORESTRY SCHOOL, HONDURAS, CHECKS WIND VELOCITY forestry development needs the development of foresters