2 Opening remarks and scheduled presentations

The workshop was called to order and officially opened by Dr. N. Mark Collins, Chief Executive Officer of WCMC. Dr. Collins thanked all participants for making themselves available to come to the workshop, and their perseverance in cases of complicated travel arrangements. Dr. Collins stressed the importance of having a framework of internationally accepted Global EZ's for pressing environmental issues, including global climate change, biodiversity conservation and their accompanying policy measures. This is not simple as ecological zoning is a very complex matter, including the consideration of climate, vegetation and land characteristics and their interactions. As any system that attempts to categorise nature or natural processes, it is a somewhat artificial tool for the convenience of people. This does not detract from its importance, but the Global Ecological Zoning system defined for the far-reaching work of FAO must be scientifically based, and accepted by both scientists and policy-makers in order to be effective.

Dr. Collins exhorted all participants to bring all their accumulated knowledge to the table and participate in the discussions of the workshop. The importance of having such a meeting steered and kept on course by an experienced chairperson was stressed. Dr. Collins invited Dr. Doug Williamson to chair the meeting, and Dr. Williamson kindly agreed. In conclusion, Dr. Collins warmly welcomed the participants to Cambridge and encouraged them to enjoy a visit to the city and its attractions.

2.1 Scheduled presentations

The workshop continued with the scheduled presentations. The session was opened by Dr. Williamson, who asked the participants to absorb the points in the presentations and to make notes as appropriate for subsequent discussion. Dr. Williamson also reminded the participants of the objectives of the workshop, as stipulated in the Introduction. The presentations are given in full in Annexes to this report (Appendix 3 and 4), but are summarised below. Some points raised after the presentations, for further discussion in the subsequent Working Groups, are presented in Section 2.2.

Presentations were as follows:

1. Henk Simons: The FRA 2000 Global Ecological Zone Map: Building on the FRA 1990 experience

2. Susan Iremonger: Review of existing Global Ecological Zoning systems

3. Karn Deo Singh: The Global/ Regional Ecological Zone Classification Concept and Definition

4. Zhiliang Zhu and Brad Smith: The Case Study of North America using the Proposed System

5. Marie-France Bellan: The Case Study on the use of the Existing Tropical Ecofloristic Zones maps in the Proposed System

and are summarised in sequence below.

1. Henk Simons, The FRA 2000 Global Ecological Zone Map: Building on the FRA 1990 experience

Dr. Simons gave an overview of the past work of FAO on the FRA 1990, and indicated how it hopes to build on this for the year 2000 report. The mission of the FRA programme of FAO is to provide the world community reliable information to describe and understand the situation of the world’s forests and related resources and how they change over time.

In attempting to fulfil this mission for the year 2000, FAO foresees providing four categories of information:

• Forest and land cover estimates: status 2000 and change
• Ecological aspects of the forest
• Economic potential of the forest
• Global maps

Describing the process and implementation methods of the survey, Dr. Simons described three main components of the assessment:

• Assessment based on existing information (usually national forest inventory and land cover information, which is standardised to common definitions).
• Remote sensing survey (based on interpretation of multi-date Landsat images, and providing information on rate and nature of deforestation, forest fragmentation and forest degradation).
• Special studies (including the development of global maps of forest and EZ's, reporting on the status of protected areas, biodiversity indicators, plantation forests and other subjects).

The goal and objectives of the global ecological zoning exercise were presented:

1. To develop a global classification concept and system for ecological zoning

2. To improve/update the EZ maps for the tropical regions

3. To develop new EZ maps for the non-tropical regions

4. To produce a global EZ database and map, and worldwide forest statistics by EZ.

The main purposes of the map and database were described as (a) the map as a product in itself, (b) the use of the map and database for reporting various forest parameters by EZ, and (c) support for the stratification in the remote sensing survey and modelling biomass.

The approach to the development of the map was explained, including the collaboration between many scientists and institutions. These would have input to the formation of the product, which will be implemented by FAO, WCMC, EDC and LET.

Dr. Simons concluded with some amusing remarks which highlighted the contrast of opinion on the Köppen climatic system. This is one of the major climatic classification systems and the one which has been used in the pilot EZ's mapping projects, to be described later in the presentations of Zhiliang Zhu and Brad Smith, and of Marie-France Bellan.

2. Susan Iremonger, Review of existing Global Ecological Zoning systems

Dr. Iremonger gave a review of currently available climate, EZ or potential vegetation maps that cover the whole globe. These global maps are generally based on macroclimate, with subdivisions based on the physiognomy of principal vegetation cover. The maps described are listed below.

Details on each map were given in the presentation, indicating the suitability of the map for use in Drafting the FAO EZ map for FRA 2000. None of the maps was considered exactly appropriate to FAO’s needs. This was generally due to a number of factors, including:

• the resolution of the map was too coarse
• it was better developed for some regions of the globe than for others
• the definitions of the classes were not very clear
• it would not be a scheme that would be globally applicable
• more accurate data were now available than were used in making the map (e.g., vegetation formation extent)

The maps reviewed were:

• Map of Köppen and Trewartha's climate classification.
• Map of Holdridge’s Life Zone System.
• Map of Walter’s Zonobiomes.
• Bailey's Global Ecoregions map.
• Milanova and Kushlin's map of Present Day Landscapes.
• Olson and Watts' map of Major Ecosystem Complexes.
• Udvardy's map of the Biogeographical Provinces of the World.
• The WWF-US Terrestrial Ecoregions of the World.
• For the tropics: FAO/LET maps of Ecofloristic Zones.
• For Europe: Bohn and Katenina’s Map of Natural Vegetation.
• For Australia: Carnahan’s Natural Vegetation map.
• South America: Hueck’s Vegetation Map of South America.
• South America: UNESCO’s Vegetation Map of South America.
• Africa: White’s Map of the Vegetation of Africa.
• For East Asia: McKinnon's map of the Original Habitat Types

3. Karn Deo Singh, The Global/Regional Ecological Zone Classification Concept and Definition.

Dr. Singh presented some of the background and practical constraints influencing FAO’s approach to the proposed Global EZ's map and database. Information from two preliminary studies indicated that there was no possibility that FAO could make a completely new map and database, which would be up and running in time for use in the FRA 2000; nor was a completely new system necessarily desirable, considering the amount of scientific work on the subject that has already been done. FAO concentrated on finding an existing scheme of classification for Global EZ's and testing its adaptation to the requirements of the FRA

As macroclimate was found to be the basis of most large scale ecological zoning efforts, an acceptable climatic classification system was sought, and the Köppen system (as adapted by Trewartha) was selected for further testing in pilot studies. Reasons for its selection were that it seemed to demonstrate a good correspondence with major natural climax vegetation types and soils, as well as showing a good correspondence in terms of its classes to the needs of the FRA.

FAO, in conjunction with EDC and WCMC, drafted a classification system for mapping based on the Köppen climatic classification system. The proposed system displayed two hierarchical levels, where five major temperature domains were defined, and these were divided into 14 sub-types, using precipitation criteria. A third level was suggested, but not for mapping at the global scale, which included altitudinal factors. The second level, however, was proposed as the scheme for use at the global level for the EZ's map.

A process for implementing the drafting of the map was suggested, which included the involvement of regional and global experts, who should proceed according to the following steps (having approved the system):

• Identification of Köppen climatic types occurring in a region; which will approximate the level 2 EZ class of the FAO scheme.
• Establishment of correspondence between regional/national potential vegetation types and the global EZ's.
• Final definition and delineation of the global EZ's, using the maps and source data consulted in steps 1 and 2.
• Edge-matching between adjacent maps
• Validation

In conclusion, Dr. Singh indicated that the resulting output from a regional production should include the following: Arc/Info coverage of the EZ map with attributes of each level of classification, graphics (e.g., GIF images) of the two (or three) levels, and a table containing the levels of the classification and corresponding labels or codes of the input regional maps.

4. Zhiliang Zhu and Brad Smith, The Case Study of North America using the Proposed

System

The system proposed by FAO for the Global EZ map, outlined by K.D. Singh in his presentation at the workshop, was the subject of a pilot study, described here. The geographical area was Canada and the United States. The source data for Canada used for adaptation to the proposed global system were from an Ecoregions map for Canada produced by Environment Canada, and incorporated into the Ecological Regions of North America map by the Commission for Environmental Co-operation (1997). For the USA, Bailey’s Ecoregions Map of North America (1998) was used.

The methodology of using each of the source map units as a base for mapping level 2 of the proposed FAO system was presented in tabular and map format. Results indicated that in a process such as this, the cross-border Zone harmonisation operations should not be lengthy, and may consist of very minor adjustments. The re-classification of the two systems into the FAO system was found to be satisfactory in this study.

In addition to drafting the new EZ's map for Canada and the USA as a test case for the FRA 2000, the team also used the new map to report on forests as FAO would do for the FRA. This increased the importance of the case study, in that it showed that in addition to being a feasible system to map, it also provides a meaningful framework for reporting national forest resources information.

The presentation also described in some detail the technical procedures for adapting the existing maps for the FAO classification system, and what datasets should be used forest cover and administrative boundaries in other regions.

5. Marie-France Bellan The Case Study on the use of the Existing Tropical Ecofloristic Zones maps in the Proposed System

Ms. Bellan commenced by giving some background to the involvement of LET in the mapping for use in FAO’s FRA programme. LET was commissioned by FAO to produce Ecofloristic Zones maps for the tropical regions of the world, for use in the FRA 1990. Maps were produced for Africa, South America and tropical Asia, for use at a scale of 1:5 million.

For the FRA 2000, FAO needs a fully global map of EZ's. For the tropical regions it would be an advantage if this would be compatible with the tropical maps produced for the 1990 FRA, to promote continuity and comparability between the two Assessments. For this reason, FAO requested LET to carry out a study on one of the tropical areas it had mapped, to see if this was compatible with the system that FAO proposes using for the Global EZ map.

The area chosen for comparison was South America. The new proposed FAO system of classification was compared to that used by the LET in the 1990 maps. The most significant differences in the classification systems concerned (a) a threshold of 20deg. C in the LET system as compared to 18 deg. C in that of FAO and (b) a difference in categories based on rainfall duration. A draft map of the FAO system and the 1990 LET system were compared and it was noted that there were no major differences. The presentation was supported by illustrations of the relevant parts of maps and data presented in tabular form.

In the LET system, the subhumid zone Aw has been subdivided into Awa (with a short Dry season) and Awb (with a long Dry season). On the LET map, this is the boundary between the evergreen seasonal forest and the cerrado region, which is covered by savannas or woodlands (semi-deciduous). This boundary is easy to draw according to vegetation, and is considered by LET to be of great enough significance in the tropics to be included in the FAO system. This will require an amendment to the FAO system.

In conclusion, Ms Bellan stated that:

• The differences between the 2 maps are mainly due to the fact that the Global EZ's map is only based on climate, without any adjustment using the vegetation types boundaries.
• The vegetation types boundaries allow to compensate for the lack of climate data (e.g., Amazonian Basin), and take into account some other factors than climate, like soils.
• If the new FAO Global EZ system and map were improved by using the vegetation boundaries, it would match rather well with the LET 1990 map.

2.2 Points raised in Plenary Discussions after the scheduled presentations

Concerns and requests for clarification ensued following the scheduled presentations, and these were addressed by the relevant delegate. Where points were leading to more substantive discussion, the Chairperson requested that the discussion should be renewed after the Working Group Sessions. The following represents the main points raised and some of the responses, in three separate categories, (a) Concepts and Classification issues, (b) Technical issues and (c) Implementation and political issues.

a. Concepts and Classification issues

• An observation was made that as the proposed method uses climatic characters predominantly it may be difficult to use information from different geographically accepted systems. A recommendation was that the map should be based on Potential vegetation characteristics and then overlaid with environmental factors. The importance of the quality of the primary information source was stressed - Potential vegetation. As this encompasses both physiognomy and climate, the zones would then be truly Ecological and not just Vegetational or Climatic.
• A clarification was that national or regional potential vegetation maps will be used as the primary source for the Drafting of the map, and it was confirmed that these would then be harmonised according to the proposed Köppen level 2 system. FAO’s goal is to identify broad homogeneous vegetation formations, and use climatic criteria as tools. The Köppen level 2 criteria were used in the production of the maps for the Canada and USA case study. As in this study, the knowledge from each region or country should be used in the formulation of the maps.
• A question that should be further discussed in the Working Groups and the subsequent Plenary was how Intrazonal and Azonal elements should be incorporated, and at what level? A possibility was level 3 orographic and level 4 edaphic. Land form, bogs and river valleys in particular were a concern, as in some regions of the globe these dominate the ecology. There was a general acceptance that these should probably not be shown at a level 2 of the hierarchy.
• It was noted that the higher levels of the system must be simple because there will otherwise be difficulties grouping regional units into these. However, the point was made that one division for Boreal Forests is not enough.
• It was noted that the human element should not be ignored. In some countries it was suggested that they reject the concept of potential vegetation because people have evolved along with the vegetation, and will not be leaving.
• It was noted that definition, for example of “forest” used by FAO and other UN organisations such as UNEP are not equal, and should be harmonised, particularly in the interest of clarity and comparability, but also for the convenience of National Reporting procedures.
• A clarification was made that the map is not primarily for use at national levels, but to Draw global pictures and to see each country and region in its global context. It therefore will not include characteristics such as density, height, biodiversity, or demographics.
• A statement was made that the terms “subtropical” and “rainforest” should be changed. Suggestion was “Warm Temperate” instead of “Subtropical”.

b. Technical issues

• A brief discussion on scale and resolution rejected that the map should be usable at a scale at which the smallest country is discernible, and accepted that it should be usable at a scale not greater than the resolution of the lowest input data. This will probably be 1:10million.
• It was noted that the most detailed level of the WWF-US system should be examined for use as subunits for Drafting the FAO map.

c. Implementation and political issues

• It was noted that the EZ system should be acceptable and usable by agricultural departments as well as forestry departments. The system will not be possible to use as reporting units for the national forest departments unless the zones are related to administrative boundaries. Most of the conflicts between agriculture and forestry are land use issues. In the USA and Canada all departments use the same EZ map.
• It was noted that the US Forest Service uses the Bailey system, but this may change (the following day it was discovered that the Omernik system was being considered seriously by the USFS). The EU already has biogeographically defined “reporting units” that all countries have to use. This system has been extended to all Council of Europe countries.
• A clarification was made that FAO is not using an already existing classification and map because some have certain more specific uses, and others are mostly more perfect or developed for a certain region (Bailey in N. America, Holdridge in L. America) or are generated without inputs from regions. The proposed FAO approach uses a common framework, and regional data and expertise.