PART 1: GLOBAL OVERVIEW

INTRODUCTION

Mangroves are commonly found along sheltered coastlines in the tropics and sub-tropics where they fulfil important socio-economic and environmental functions. These include the provision of a large variety of wood and non-wood forest products; coastal protection against the effects of wind, waves and water currents; conservation of biological diversity – including a number of endangered mammals, reptiles, amphibians and birds; protection of coral reefs, sea grass beds and shipping lanes against siltation; and provision of habitat, spawning grounds and nutrients for a variety of fish and shellfish, including many commercial species.

High population pressure in coastal areas has, however, led to the conversion of many mangrove areas to other uses, including infrastructure, aquaculture, rice and salt production. Numerous case studies describe mangrove losses over time, but information on the status and trends of mangrove area extent at the global level is scarce.

The first attempt at estimating the total mangrove area in the world was undertaken as part of the FAO/UNEP Tropical Forest Resources Assessment in 1980, where the world total was estimated as 15.6 million hectares. More recent estimates range from 12 to 20 million ha (refer to Table 1 below). For many of these studies, countries with small areas of mangroves were excluded due to lack of information and because their combined area of mangroves would not significantly affect the world total.

Table 1. Previous estimates of global extent of mangroves.
	Reference	Reference year1	Number of countries included	Estimated total area (ha)
	FAO, UNEP, 1981	1980	51	15 642 673
	Saenger et al., 1983	1983	65	16 221 000
	FAO, 1994	1980-1985	56	16 500 000
	Groombridge, 1992	1992	87	19 847 861
	ITTO/ISME2, 1993	1993	54	12 429 115
	Fisher and Spalding, 1993	1993	91	19 881 800
	Spalding et al., 1997	1997	112	18 100 077
	Aizpuru et al., 2000	2000	1123	17 075 600

A recent FAO initiative aims at facilitating access by policy makers, planners and mangrove managers worldwide to comprehensive information on the current and past extent of mangroves in all countries and areas in which they exist. This document presents the preliminary results, which include an updated list of the most recent, reliable mangrove area estimate by country along with revised estimates for 1980 and 1990 and an extrapolated estimate for 2000 for each country.

All national level estimates for each country are presented along with a short description of the mangrove vegetation and graphs illustrating the trend in mangrove area changes over time.

MATERIALS AND METHODS

This study builds on the earlier FAO/UNEP assessment of the global mangrove area and on the recent FAO Global Forest Resources Assessment 2000 (FRA 2000), where all countries were asked to provide information on current forest area according to forest types, using their own classification system. Most of the countries that have mangroves provided specific information on their extent.

An extensive literature search and requests for information sent to mangrove networks and individual specialists worldwide yielded additional information. In addition to the past assessments listed in Table 1, the main bibliographical sources consulted include proceedings of conferences and other technical reports available in FAO’s libraries, the Aquatic Sciences and Fisheries Abstracts (ASFA) database and the CABI abstracts as well as information available on the internet.

Attention focused, in the first instance, on references with recent and reliable national level information on the extent of mangroves in a given country. A search was also made for past estimates in order to facilitate an analysis of area changes over time in each country. Sub-national level data (region/province/extended forest) were also included where available. In some cases, these sub-national level figures were used to create a composite national estimate. They were also very valuable for the trend analyses.

More than 2800 national and sub-national datasets were collected, covering 121 countries and areas where mangroves are known to exist, with the earliest estimates dating back to 1918. Crosschecking of data was done wherever possible and the information in the database has been analysed with the assistance of mangrove specialists throughout the world. (Refer to Annex 3 for a list of people contacted.)

One of the results of this analysis is an updated list of the most recent, reliable estimate for each country, mostly based on inventories or analysis of remote sensing imagery.

An initial screening of results included weeding out duplicates, discarding rough “guesstimates” and selecting just one estimate for those years where more than one was available for the country trend analyses. This was followed by regression analyses (best fit of linear, polynomial, logarithmic and power curves) of existing reliable data over time for each country, which provided estimates for 1990 and 1980, and an extrapolated estimate for 2000.

Where insufficient information was available, i.e. only one estimate within the last 30 years (representing around 0.1 percent of the total mangrove area), the area was assumed to have remained constant unless qualitative information indicated otherwise.

Where recent information was unavailable (about 3.5 percent of the total mangrove area), the extrapolation to year 2000 was based on the annual forest change rate for all forest types in the country for the period 1990 to 2000 applied to the latest reliable estimate. These change rates were reported in the Global Forest Resources Assessment 2000 (FAO, 2001).

All estimates and a review of all references consulted were entered in the Forestry Information System (FORIS), developed by FAO’s Forest Resources Assessment Programme. FORIS includes a relational database that enables the users to connect geographic units with references, data context and subject classification.

For details on reference sources and the estimates and regression used for individual countries refer to Part 2 and to FAO, in press. All primary datasets and reference reviews are accessible at the following website: http://www.fao.org/forestry/mangroves.

RESULTS

A total of 121 countries and areas were identified as containing one or more species of true mangroves (sensu Tomlinson, 1986). Table 2 below presents an overview of the main results of the analysis of the status and trends in mangrove area extent in the form of regional and world totals. The year given for the most recent, reliable estimate is the area-weighted average.

The most recent, reliable estimate for each country/area is shown in Table 3. A table with country level estimates for 1980, 1990 and 2000, based on trend analyses, can be found in Annex 1, while detailed information for each country/area is presented in Part 2.

Table 2. Status and trends in mangrove area extent by region.
	Region	Most recent reliable estimate		1980	1990	Annual change 1980-1990		2000	Annual change 1990-2000
		000 ha	Ref year	000 ha	000 ha	000 ha	%	000 ha	000 ha	%
	Africa	3 390	1993	3 659	3 470	-19	-0.5	3 351	-12	-0.3
	Asia	6 662	1991	7 857	6 689	-117	-1.6	5 833	-86	-1.4
	Oceania	1 578	1995	1 850	1 704	-15	-0.8	1 527	-18	-1.1
	N & C America	2 103	1994	2 641	2 296	-34	-1.4	1 968	-33	-1.5
	South America	2 030	1992	3 802	2 202	-160	-5.3	1 974	-23	-1.1
	World Total	15 763	1992	19 809	16 361	-345	-1.9	14 653	-171	-1.1

As can be seen from Table 2, the most extensive area of mangroves is found in Asia, followed by Africa and South America. Four countries (Indonesia, Brazil, Nigeria and Australia) account for about 41 percent of all mangroves and 60 percent of the total mangrove area is found in just ten countries. Refer to Figures 1, 2 and Table 3, which are all based on the most recent, reliable estimate identified for each country.

The results of the trend analyses indicate that the current mangrove area worldwide has now fallen below 15 million hectares, down from 19.8 million ha in 1980. The world has thus lost 5 million ha of mangroves over the last twenty years, or 25 percent of the extent found in 1980. As is also evident from Table 2, mangrove deforestation continues, albeit at a slightly lower rate in the 1990s (1.1 percent per annum) than in the 1980s (1.9 percent per annum).

The figures presented in Annex 1 suggest that, in absolute terms, the highest rates of deforestation in the 1980s were found in Brazil, Indonesia, Pakistan, Australia and Mexico. The same countries (except for Pakistan which has been replaced by Papua New Guinea) top the list of countries with the largest losses of mangroves in the 1990s, although the rates have diminished substantially in most of these.

Figure 1. Mangrove extent per country (hectares).

In relative terms, countries with high deforestation rates include Singapore, Benin, Dominica, Brazil and Côte d’Ivoire in the 1980s and Côte d’Ivoire, Honduras, China, Congo and Barbados in the 1990s. Conversely, a number of countries have registered a positive change over time, including Bangladesh, where the largest mangrove area (The Sundarbans Reserved Forest) is well protected and substantive and successful efforts have been made in coastal afforestation, particularly on land formed by accretion. Kuwait also registers a positive trend, as mangroves have been planted as an introduced species. For other countries, an apparent positive or negative trend may be due to changes in methodology or definition – or due to the discovery of new mangroves areas not previously included (e.g. United Arab Emirates).

Table 3. Most recent, reliable mangrove area estimates.
AFRICA				ASIA				OCEANIA
Country/ Area	ha	Year		Country/ Area	ha	Year		Country/ Area	ha	Year
Angola	60 700	1992		Bahrain	100	1992		American Samoa	52	1976
Benin	1 700	1989		Bangladesh	622 482	1996		Australia	955 277	1997
Cameroon	227 500	2000		Brunei Darussalam	17 100	1992		Fiji	42 464	1991
Comoros	2 600	1976		Cambodia	72 835	1997		Guam	70	1993
Congo	12 000	1995		China	36 882	1994		Kiribati	258	1995
Côte d'Ivoire	15 000	1995		India	487 100	1997		Marshall Islands	n.a.	n.a.
Dem. Rep. of Congo	22 600	1995		Indonesia	3 493 110	1988		Micronesia	8 564	1983
Djibouti	1 000	1985		Islam. Rep. Of Iran	20 700	1994		Nauru	1	1993
Egypt	482	1998		Japan	400	1980		New Caledonia	20 250	1987
Equatorial Guinea	25 700	1995		Kuwait	2	2000		New Zealand	22 200	1996
Eritrea	6 400	1997		Malaysia	587 269	1995		Niue	3 000	1981
Gabon	115 000	2000		Maldives	n.a.	n.a.		Northern Mariana Islands	7	1984
Gambia	59 600	1993		Myanmar	452 492	1996		Palau	4 708	1985
Ghana	10 000	1995		Oman	2 000	1992		Papua New Guinea	464 000	1993
Guinea	296 300	1995		Pakistan	207 000	1990		Samoa	752	1993
Guinea-Bissau	248 400	1990		Philippines	127 610	1990		Solomon Islands	52 500	1995
Kenya	52 980	1995		Qatar	500	1992		Tokelau	n.a.	n.a.
Liberia	19 000	1995		Saudi Arabia	20 400	1985		Tonga	1 305	1990
Madagascar	325 560	1987		Singapore	500	1990		Tuvalu	40	1993
Mauritania	104	1993		Sri Lanka	8 688	1992		Vanuatu	2 519	1993
Mauritius	7	1991		Thailand	244 085	2000		Wallis and Futuna Islands	n.a.	n.a.
Mayotte	668	1989		Timor Leste	3 035	2000
Mozambique	392 749	1997		United Arab Emirates	4 000	1999
Nigeria	997 700	1995		Viet Nam	252 500	1983
Sao Tome and Principe	n.a.	n.a.		Yemen	927	1993
Senegal	182 400	1985
Seychelles	2 000	1995
Sierra Leone	156 500	1986
Somalia	10 000	1975
South Africa	673	1991
Sudan	500	1995
United Rep. of Tanzania	143 284	1987
Togo	1 000	1999
Total Africa	3 390 107	1993		Total Asia	6 661 717	1991		Total Oceania	1 577 967	1995

The relatively large mangrove deforestation rates in Asia, the Caribbean and Latin America in the 1980s reflect large scale conversion of mangroves for aquaculture and tourism infrastructure. Most countries have now banned the conversion of mangroves for aquaculture purposes and require environmental impact assessments prior to large-scale conversion of mangroves areas for other uses.

The study did not provide information on the rate of mangrove degradation or fragmentation.

DISCUSSION

As with all analyses, the reliability of the results depends on the quality of the inputs. In this particular case, lack of comprehensive information and changes in definitions and methodologies between different assessments make it difficult to compare results over time and establish reliable trends. For some countries, the most recent estimate is a composite of sub-national figures due to lack of recent national figures and for others, the borders have changed or additional mangrove areas have been discovered recently, making trend analyses very challenging.

The discussion below focuses on overall data availability and quality and on specific issues related to the determination of the most recent, reliable estimate and the trend analyses.

Overall data availability

FAO’s database on mangrove area estimates currently includes 2814 datasets sorted by the 121 countries and areas identified by this study as having mangroves. The total number of different references is 497. Of these, 43 are references providing information on mangrove areas at the global or regional level while the remaining 454 are country-specific references. An illustration of the information availability per country (number of country-specific references and number of national and sub-national level datasets) is given in the Figures below.

A total of 1010 datasets provide national level estimates (including those with breakdown to sub-national units), while 1804 datasets only provide sub-national figures, covering regions, provinces or important mangrove areas within a given country. Some of the entries are secondary sources repeating past estimates. The total number of different national level estimates is 756 - equalling an average of six different estimates over time per country.

Despite concerted efforts, no quantitative information was identified on the extent of the mangroves found in 5 countries/areas, i.e. Maldives, Marshall Islands, Sao Tome and Principe, Tokelau and Wallis and Futuna Islands. For ten countries, only a single estimate was identified. Conversely, more than ten different national level estimates over time were identified for 28 countries. For seven of these, more than 100 datasets (national and sub-national level) were identified.

Figure 3. World data availability map.

Figure 4. Data availability. Africa.

Figure 5. Data availability. Asia.

Figure 6. Data availability. Oceania.

Figure 7. Data availability. North and Central America.

Figure 8. Data availability. South America.

Data quality

The general quality of the data is illustrated in Table 4 below, which shows the methodology used for generating national level mangrove area estimates. The percentages given for the different methodologies refer to the number of national estimates, not the percentage of the mangrove area covered by this methodology.

Table 4. Data quality.
			All national level estimates									Most recent, reliable estimate
			Ground survey		Remote sensing/ Mapping		Expert estimate		Not known			GS		RS		EE		NK
	Region	No.	No.	%	No.	%	No.	%	No.	%		No.	%	No.	%	No.	%	No.	%
Africa		191	3	1	68	36	22	11	98	51		1	3	14	44	5	16	12	37
Asia		248	8	3	65	26	32	13	143	58		2	8	11	46	4	17	7	29
Oceania		60	2	3	19	32	6	10	33	55		2	11	9	50	0	0	7	39
N. & C. America		191	16	8	58	30	36	19	81	42		12	32	9	26	5	15	8	23
South America		66	0	0	31	47	6	9	29	44		0	0	3	37	2	25	3	37
World		756	29	4	241	32	102	13	384	51		17	15	46	40	16	14	37	32

Around 36 percent of all national level estimates were based on ground surveys or remote sensing and mapping. It is noteworthy that none of the national level estimates for countries in South America were based on ground surveys – probably due to the costs involved in conducting large scale inventories.

The high level of estimates without a known methodology is worrying. Where this is due to reliance on secondary sources attempts will be made to obtain copies of the original sources of information. However, the study also highlighted that there is a need to pay further attention to the description of the methodology used when reporting new figures.

As it is evident from the right hand side of the table, the selection of the most recent, reliable national level estimate gave priority to those estimates derived from ground surveys and/or remote sensing, where possible. More than half of the estimates selected are thus based on the use of these methodologies.

Definitions

In the present study, mangroves were broadly defined as areas containing one or more true mangroves according to Tomlinson, 1986. No attempt was made to re-classify areas defined as mangroves in the original sources; however waterways and other vegetation types existing within the larger mangrove area and areas of freshwater swamp substracted from the total area, when reported. All definitions used by the original sources have been inserted into the database. Given that mangroves are a relatively distinct and, in most cases, easily demarcated forest type, the problem of different definitions used over time is deemed to be less significant than the problems caused by the use of different methodologies and sampling intensities over time and possible interpretation errors.

Most recent, reliable estimate

Recent estimates, dating from 1995 or later, were identified for 45 countries (or 44 percent of all countries and areas) with estimates from an additional 48 countries dating from 1990 to 1994. Together, these two groups of countries account for 69 percent of the mangrove are in the world.

Conversely, for 23 countries and areas the most recent, reliable figure was dating from before 1990, and for five countries no estimate was identified at all. The majority of countries and areas where the most, recent reliable estimate dates from 1985 or earlier, are arid zone countries or Small Island Developing States (SIDS) i.e. countries with relatively small areas of mangroves.

The combined area of four countries (Indonesia, Brazil, Nigeria and Australia) has a strong influence on the global total, as these countries together account for an estimated 6.5 million ha or 41 percent of the total mangrove area in the world.

Despite the very high number of sub-national and national level estimates available, estimating the area of mangroves for Indonesia proved particularly difficulties due to the large area and the lack of recent information covering the whole country. Sub-regional estimates for eight main groupings of islands (Bali, Irian Jaya, Java, Kalimantan, Maluku, Nusa Tenggara, Sulawesi and Sumatra) were thus used to generate a composite national level estimate as well as for the trend analysis. However, no recent estimate was identified for Irian Jaya which contains one of the largest mangrove areas. As a result, the area weighted average year of the composite estimate from Indonesia is 1988 (Refer to Part 2, Box 2 for details). For Brazil, the most recent, reliable estimate identified dates from 1991; for Nigeria, the most recent estimate is from 1995 and for Australia it is 1997.

Table 5 lists the countries and areas according to the date of the most recent, reliable estimate.

Table 5. Countries sorted by year of most recent, reliable estimates.
No quantitative information	y<1980	1980=y<1985	1985=y<1990	1990=y<1995	1995=y=2002
Maldives	American Samoa	French Guiana	Aruba	Angola	Australia
Marshall Isl.	Comoros	Japan	Benin	Anguilla	Bangladesh
Sao Tome and Principe	Somalia	Micronesia	Djibouti	Antigua and Barbuda	Belize
Tokelau		Netherlands Antilles	Indonesia	Bahamas	British Virgin Islands
Wallis and Futuna Isl.		Niue	Madagascar	Bahrain	Cambodia
		Northern Mariana Isl.	Mayotte	Barbados	Cameroon
		Viet Nam	New Caledonia	Bermuda	Colombia
			Palau	Brazil	Congo
			Saudi Arabia	Brunei Darussalam	Côte d'Ivoire
			Senegal	Cayman Islands	Dem. Rep. of Congo
			Sierra Leone	China	Dominican Republic
			United Rep. of Tanzania	Costa Rica	East Timor
			Venezuela	Cuba	Ecuador
				Dominica	Egypt
				El Salvador	Equatorial Guinea
				Fiji	Eritrea
				Gambia	Gabon
				Grenada	Ghana
				Guam	Guadeloupe
				Guinea-Bissau	Guatemala
				Guyana	Guinea
				Haiti	Honduras
				Islam. Rep. Of Iran	India
				Mauritania	Jamaica
				Mauritius	Kenya
				Mexico	Kiribati
				Montserrat	Kuwait
				Nauru	Liberia
				Nicaragua	Malaysia
				Oman	Martinique
				Pakistan	Mozambique
				Papua New Guinea	Myanmar
				Peru	New Zealand
				Philippines	Nigeria
				Qatar	Panama
				St Kitts and Nevis	Puerto Rico
				St Vincent and Grenadines	Saint Lucia
				Samoa	Seychelles
				Singapore	Solomon Isl.
				South Africa	Sudan
				Sri Lanka	Suriname
				Tonga	Thailand
				Trinidad and Tobago	Togo
				Turks and Caicos Islands	United Arab Emirates
				Tuvalu	United States
				US Virgin Islands.
				Vanuatu
				Yemen
Total number of countries per class
5	3	7	13	48	45
Percentage of total mangrove area per class
n.a.	0.1 %	2%	29%	25%	44%

Trends over time

The literature search covered a period of more than 80 years with the earliest assessments analysed dating back to 1918 (Philippines), 1921 (Madagascar) and 1928 (Singapore) and the most recent ones dating from 2002 (Saint Lucia).

Part 2 provides the list of all the national level estimates by country - including the full references as well as information related to the methodology and whether the estimate was used for the trend analysis. The results of the individual trend analyses for each country and area can also be found in Part 2. The complete lists of both national and sub national level estimates and references entered in the database are included in FAO (in press).

The trend analyses were hampered by three main problems: lack of comprehensive data over time, differences in definition over time and differences in assessment methodologies over time. The use of different methodologies over time is the largest problem in the current study. The apparent, fairly large mangrove deforestation rate in Australia, for example, may, at least partly, be due to the use of different methodologies over time. Although preference was given to estimates based on field level inventories and remote sensing when selecting estimates for the trend analysis, it was not always possible to exclude estimates where the methodology was unknown.

Even when the same method was used over time, different scales and sampling intensities may have been used.

Comparison with previous studies

Table 6 below presents an overview of the results of this study compared with previous global estimates.

Table 6. Comparison with previous studies.
	Extent (ha)
	Region	Ibid, 1980	Ibid, 1990	Ibid, 2000	Spalding et al. 1997	ISME/ ITTO 1993	Fisher & Spalding, 1993	Ibid 1992	Groombridge 1992	Saenger et al. 1983	FAO, UNEP 1981
	Africa	3 659 322	3 469 844	3 350 813	3 738 304	2 945 367	5 566 354	3 390 107	5 549 780	3 258 800	3 607 000
	Asia	7 856 500	6 689 280	5 832 737	7 580 932	5 091 913	7 719 081	6 661 717	7 682 335	5 179 600	5 581 651
	Oceania	1 850 067	1 703 949	1 526 934	1 878 894	38 543	1 514 457	1 577 967	1 492 157	1 698 002	553 000
	N. & C. America	2 641 289	2 296 402	1 968 407	2 479 425	1 931 650	3 314 940	2 102 886	3 309 940	2 012 400	2 151 700
	South America	3 801 600	2 202 000	1 974 300	2 457 900	2 459 546	1 813 823	2 030 330	1 813 823	4 072 200	3 749 322
	World Total	19 808 778	16 361 475	14 653 191	18 128 005	12 467 019	19 928 655	15 763 007	19 848 035	16 221 002	15 642 673

The present figures are obtained by adding all the country figures for each region

In summary, and keeping the caveats cited in the previous sections in mind, the results indicate that previous estimates of the total mangrove area made in the early 1980s were too low (primarily due to the limited number of countries included), whereas those made in the 1990s were too high. It should be noted that, while the new estimates provided by this study for 1980 and 1990 and the most recent, reliable estimate (area weighted average year: 1992) are presumed to be the best available and fairly reliable in that they are primarily based on information from inventories and aerial surveys, the extrapolation to year 2000 is, by nature, less reliable - particularly where few or no recent assessment results were available. This estimate is thus only indicative and is likely to change as and when results from on-going and future assessments become available.

CONCLUSIONS

Mangroves fulfil many extremely important productive, protective and social functions. Yet, increased population pressures in coastal areas and lack of awareness have lead to large scale conversion of mangroves to other uses. Numerous case studies and anecdotal evidence exist describing mangrove losses over time. However, access to comprehensive information on the status and trends of mangrove areas at the global level has been limited.

The initiative by the FAO described in this paper aims at facilitating access to comprehensive information on the current and past extent of mangroves in all countries and areas in which they exist.

More than 2800 national and sub-national data sets have, so far, been collected, covering 121 countries and areas where mangroves are known to exist, with the earliest estimates dating back to 1918. An updated list of the most recent, reliable estimate for each country has been compiled. Regression analyses based on earlier data provided revised estimates for 1990 and 1980 and an extrapolated estimate for 2000 for each country.

The results of the analysis of the most recent, reliable estimate identified for each country and area show that the area of mangroves vary from a few hectares to more than a million hectare in individual countries and that four countries (Indonesia, Brazil, Nigeria and Australia) account for about 41 percent of all mangroves and 60 percent of the total mangrove area is found in just ten countries.

The study illustrated the problems associated with generating reliable trends due to differences in methodology used over time as well as the lack of recent, reliable and comprehensive information. The estimate for 2000 is thus indicative only and is likely to change when results from on-going and future assessments become available.

Nevertheless, the trend analyses indicate that the current mangrove area worldwide has now fallen below 15 million hectares, down from 19.8 million ha in 1980. The world has thus lost 5 million ha of mangroves over the last twenty years, or 25 percent of the extent found in 1980.

The study also indicates that mangrove deforestation continues, albeit at a slightly lower rate in the 1990s (1.1 percent per annum) than in the 1980s (1.9 percent per annum), reflecting the fact that most countries have now banned the conversion of mangroves for aquaculture purposes and require environmental impact assessments prior to large-scale conversion of mangroves for other uses.

In addition to providing access to all primary data and thus facilitating the analysis of these by others, the database also gives an indication of where information is lacking and efforts should be made to obtain more recent and reliable area estimates.

Better information on both the extent and the condition of mangroves is needed as an aid to policy and decision making aimed at the conservation, management and sustainable use of the world’s remaining mangrove ecosystems.

Readers are thus strongly encouraged to provide feedback and additional information to help update and improve this database for the benefit of all those who may have an interest in mangroves.

BIBLIOGRAPHY

Aizpuru, M., Achard, F., and Blasco, F. 2000. Global assessment of cover change of the mangrove forests using satellite imagery at medium to high resolution. In: EEC Research project n 15017-1999-05 FIED ISP FR – Joint Research Centre, Ispra.

Clough, B.F. 1993. The Economic and environmental values of mangrove forests and their present state of conservation in the South-East Asia/Pacific Region. Mangrove Ecosystems Technical Reports vol.3 ITTO/ISME/JIAM Project PD71/89. Rev. 1(F) Okinawa, Japan, ISME. 202 pp.

Diop, E.S. 1993. Conservation and sustainable utilization of mangrove forests in Latin America and Africa regions, Part II – Africa. Mangrove Ecosystems Technical Reports vol.3 ITTO/ISME Project PD114/90. Okinawa, Japan, ISME. 262 pp.

FAO. in press. FAO’s database on mangrove area estimates. By Wilkie, M.L., Fortuna, S. and Souksavat, O. Forest Resources Assessment Working Paper No. 62. Forest Resources Division. FAO, Rome. (Unpublished)

FAO. 2001. Global Forest Resources Assessment 2000: Main report. FAO Forestry Paper 140. Rome. www.fao.org/forestry/fo/fra/main/index.jsp

FAO. 1994. Mangrove forest management guidelines. FAO Forestry Paper 117. Rome, 319 pp.

FAO, UNEP. 1981a. Los Recursos Forestales de la America Tropical Proyecto de Evaluación de los Recursos Forestales Tropicales (en el marco de SINUVIMA) FAO, UNEP, Rome. 349 pp.

FAO, UNEP. 1981b. Tropical Forest Resources Assessment Project, Forest Resources of Tropical Africa. Part II: Country Briefs FAO, UNEP, Rome. 586 pp.

FAO, UNEP. 1981c. Tropical Forest Resourcess Assesment Project, Forest Resources of Tropical Asia. FAO, UNEP, Rome. 475 pp.

Fisher, P. and Spalding, M.D. 1993. Protected areas with mangrove habitat. Draft Report World Conservation Centre, Cambridge, UK. 60pp.

Groombridge, B. 1992. Global biodiversity: status of the earth's living resources. WCMC / The National History Museum / IUCN / UNEP / WWF / WRI. Chapman & Hall, London. 594 pp

Lacerda, L.D. 1993. Conservation and sustainable utilization of mangrove forests in Latin America and Africa regions. Vol. 2 Part I- Latin America. Mangrove Ecosystems technical reports ITTO/ISME Project PD114/90 (F). Okinawa, Japan. 272 pp.

Saenger, P., Hegerl E.J. and J.D.S., Davie. 1983. Global status of mangrove ecosystems. Commission on Ecology Papers No.3. IUCN. Gland, Switzerland. 88 pp.

Spalding, M.D., Blasco, F. and Field, C.D., eds. 1997. World Mangrove Atlas. The International Society for Mangrove Ecosystems, Okinawa, Japan. 178 pp.

Tomlinson, P.B. 1986. The botany of mangroves. Cambridge University Press, Cambridge, UK. 419 pp.

¹ Except for FAO, UNEP 1981 and Aizpuru et al., 2000, the reference year is the year of the publications in which the estimate is cited, not the weighted average of all the national area estimates.

² Combined figure from 3 publications by Clough, 1993; Diop, 1993 and Lacerda 1993.

³ New estimates were provided for 21 countries, for the remaining countries the study relied on Spalding et al., 1997.