3. Methodological issues

3.1 Main characteristics of the nfa approach

For a number of motives such - statistical, easy planning and implementation in the field, reliability of results, cost implications, etc - the systematic sampling was chosen as the most suitable technique for nationwide data collection on forest and tree resources. For equal sampling intensities, a systematic sampling usually yields higher precision of estimates compared to a non-stratified random sampling. More detailed information on the approach is available on “FAO's approach to support national forest assessments for Country Capacity Building” ( Saket & al. 2002)

The level of intensity of sampling for national forest and tree assessment is optimised to produce national estimates of main parameters at acceptable sampling error (<10%) for the key variables and maintain the exercise cost-effective. Sampling error is higher for rare population elements. The sampling intensity chosen to produce reliable information at the national level, ranges from 50 to 500 tracts, depending on country size and homogeneity. Stratification based on stable criteria such as ecological zone can be also applied.

Most of the information is collected on the ground in field plots (4 plots per tract) through measurements, observations and interviews to local community and key informants. To ensure long term monitoring of the resources the field plots are permanent and materialised on the ground.

The projects are to be implemented by national institutions and professional, and are founded on collaborative partnerships between the concerned government, donors and FAO. To guarantee a successful implementation of projects, commitments of the country’s institutions are essential. In addition to the generated information, country capacity building in forest inventory and assessment is a major outcome of the project.

To broaden the knowledge on forest and tree resources of a country the assessment covers not only variables related to the states of the resources, but also forest management and benefits (products and services) for all users. The approach for national forest assessments is designed on the basis of a standard core of global biophysical and management/use variables that are adjustable to national needs, a classification system based on common harmonized global classes and adaptable to national requirements, and common terminology and definitions.

Nfas produce considerable amount of data at different levels and for different purposes. A functional information system is fundamental to enter, store, process and manage the collected inventory data. Support to national forest assessments includes the development of a national information system on forest and tree resources. The information system will be based on a well-structured database, which includes various levels of internal relations. Many analyses can be carried out on the stored data. The system will permit storage of data from sequential inventories in order to detect and estimate changes and identify trends. The information system should therefore be allow receiving data from historic database, up-dating the information and have functions to analyse changes and trends. The system will enable reporting on national and global issues.

3.2 Reporting

A report layout was introduced to the participants of the workshop and discussed. There was a consensus that the report should cover to the extent possible the global variables for the Global Forest Resources Assessment (GFRA) as specified by the Advisory Group of the GFRA programme. The variables relate to criteria of sustainable forest management and will be used as a framework for country reporting during the next GFRA. These variables are:

- Land use;

- Land use change;

- Ownership;

- Designation;

-Forest characteristics;

- Forest characteristics change;

- Growing stock;

- Biomass;

- Carbon stock;

- Fires and Pests;

- Species occurrence;

- Forest tree species abundance;

- Wood supply balance;

- NWFP supply;

- Socio-cultural sites;

- Value of primary production of wood;

- Value of primary production of nwfp;

- Employment through primary activities in forests;

- Illegal harvesting.

The proposed report layout is given in the following frame.

It was accepted that the national reports should be produced in the official language of the country and be translated to English and French.

3.3 Presentation by participants

3.3.1 Findings of the consultancy on Field-testing and evaluating the Interview Component of FAO’s New Approach to nfa’s

Krister P. Andersson, Scientist from the Centre for the Study of Institutions, Population, and Environmental Change (CIPEC), Indiana University, Bloomington, IN, USA, was assigned by FAO the study of field-testing and evaluating the interview component of FAO’s approach to national forest assessments being implemented in Guatemala. A 10-day visit allowed producing a report (Andersson K.P., 2003). The review covered the data content and format as proposed by the FRA approach. In this respect, the findings of the study are that although the nfa has some issues to resolve, the information that the nfa makes available is of potentially great value both to national policy makers and to FAO, and it is much richer and more complete than the current forest assessments of most FAO member countries, whether developing or industrialized countries. Ultimately, the usefulness of this data is related to how the FRA end users perceive its overall quality. Thus, the issue of data quality is critical for the future success of the nfa, especially as the attributes of forest use tend to vary according to the locality’s bio-physical characteristic’s social, cultural and historical attributes of the community of users, as well as the rules and norms observed by each user. Three ways should be considered to determine the data quality, namely the reliability, validity, and uncertainty of estimates. For this, the author proposed a number of recommendations as to i) define a minimum standard for interview methods; and to ii) test the quality of information including test of data representativity, reliability and validity.

3.3.2 Introduction to knowledge reference for national forest assessments

Ylva Melin, from the Swedish University of Agricultural Sciences, Department of Forest Resource Management and Geomatics, delivered a presentation on knowledge reference for national forest assessments, which is an FAO-IUFRO initiative whose objective is to improve access, world-wide and in several languages, to state-of-the-art knowledge on a wide range of subjects relevant to national forest assessments.

The reference knowledge will be developed by:

- Lead authors for IUFRO network. Each chapter will have one responsible and co-authors;

- Representatives from countries for case studies;

- Institutions such FAO, IUFRO and the Swedish University of Agricultural Sciences, Department of Forest Resource Management and Geomatics.

Publications of this project will be made through the Internet and will reside within the FAO website. It will also be part of the Forestry Information System (FORIS). The contents will be divided into chapters structured in three parts. The tentative structure of each is the following:

§ Part I: Introduction

- Rationale;

- Cases.

§ Part II: Inventory

- Sampling design;

- Observation and Measurement;

- Permanent plots;

- Remote sensing;

- Implementation and Organization.

§ Part III: Output

- Information management and Data registration;

- Biometry;

- Scenarios and Analysis;

- Policy influence.

The internet publication will include a chapter on specifications comprising:

- Overview;

- Lead author;

- Text addressing the subject of the paper;

- Annotated bibliography;

- Annex: Contacts;

- Annex: Proposed teaching materials;

- Annex: Acronyms, terms, definitions and thesaurus.

Seven persons have already shown interest in participating in the knowledge reference group. A meeting is planned for Mars 2003 with the lead authors to discuss developments in this project. Other follow up meetings will be organized.

3.3.3 Biodiversity indicators in national forest inventories

Valerie Kapos, from the WCMC, delivered a presentation entitled “Biodiversity indicators in national forest inventories”, stating that these indicators are required when:

- appropriate for use at the local scale, but can be readily aggregated at larger scales;

- can be generated from data collected by standard forest inventories;

- build on the many international initiatives that have developed biodiversity indicators such as ITTO, Pan-European (or ‘Helsinki’) Process, Montreal Process, Tarapoto, Lepaterique, Near East, Dry Zone Asia and Dry Zone Africa

The problems with existing forest biodiversity indicators include poor definition and difficulties of measuring a number of them, relevance of some at the national scale but not at the forest management unit and some, as indicated by a field evaluation of C&I undertaken by CIFOR, of little relevance to forest management. There is a number of indicators framework such as:

- Pressure-State-Response’(P-S-R) framework, developed by the OECD;

- Driving Force - State – Response’ (D-S-R) applied by the CSD (CSD 2001);

- Driver-Pressure-State-Impact-Response’ (D-P-S-I-R) developed by EEA;

- Plus others – Noss (1990), Stork et al. (1997);

Key biodiversity indicators include:

- Forest area by type, and successional stage relative to land area;

- Protected forest area by type, successional stage and protection category relative to total forest area;

- Degree of fragmentation of forest types -area by spatial integrity class;

- Rate of conversion of forest cover (by type) to other uses;

- Area and percentage of forests affected by anthropogenic and natural disturbance;

- Complexity and heterogeneity of forest structure, e.g. area by canopy height class, by canopy openness, dbh frequency distribution;

- Numbers of forest-dependent species;

- Conservation status of forest-dependent species.

Conclusions:

- Future global forest assessments should incorporate a greater emphasis on biodiversity;

- This could be partly achieved through use of indicators, appropriate for use at the local scale, but enabling information to be aggregated at larger scales;

- Information collected at the local scale may be aggregated by summarizing data in categorical form, and presenting them in relation to forest area;

- Such an approach would assist countries in contributing to C&I processes, as well as meeting their reporting obligations to international conventions;

- There is a widespread need for capacity building in biodiversity assessment and analysis.

3.3.4 Data entry and storage

Dan Altrell from the FRA presented the system developed for the entry, storage and analyse of the information collected in a nfa. A nfa produces considerable amount of data at different levels and for different purposes. A functional information system is fundamental to enter, store, process and manage the collected inventory data. Support to national forest inventories includes development of a national information system for forest and tree resources. The base for the information system will be a well-structured database, which will include various levels of internal relations. An undefined number of analyses can be carried out on the stored data. The system will permit storage of data from sequential surveys in order to detect and estimate changes and establish trends. The information system should therefore be a base for historic database, up-dated information and for changes and trend analysis. The system will enable reporting on global and national issues.

3.4 Conclusions and recommendations

a) In relation with the forest and tree resources uses and management variables, information that a nfa makes available is of potentially great value both to national policy makers to FAO, and it is much richer and more complete than the current forest assessments of most FAO member countries, whether developing countries or not. It is thus recommended that:

- A minimum standard for interview and processing methods is defined.

- Interviewers are well trained in interview methods.

- Quality of information is tested and proven reliable and representative.

b) In connection with biodiversity indicators, the NFA approach has the potential to provide valuable information. It is however recommended that:

- Future assessments incorporate a greater emphasis on biodiversity. This could be partly achieved through use of indicators, appropriate for use at the local scale, but enabling information to be aggregated at larger scales.

- Information collected at the local scale may be aggregated by summarizing data in categorical form, and presenting them in relation to forest area.

- Countries are enabled in contributing to C&I processes, as well as meeting their reporting obligations to international conventions.

- Capacity building in biodiversity assessment and analysis is assured by nfa projects.

c) The participants recognise the strong need for knowledge reference in national forest assessments. The interest of scientists from the IUFRO network to contribute as lead authors or co-authors is encouraging. The participants in the workshop recommended that:

- case studies should include the current nfa projects;

- access is provided to resource materials on interviews.

d) In connection with follow up activities, the participants:

- Considered that a working meeting on data analysis is necessary after completion of the ongoing projects. Such meeting should include decision makers from the countries.

- Recommended that an evaluation meeting is organised in the countries, at e.g. mid-term, to assess progress, adjust the approach where necessary and provide countries with feed back.

- Recommended that FAO considers, for new nfa project in new countries, specific national information requirements and involvement of representatives in ongoing nfa efforts in other countries through e.g. study tours.

- Recommended the development of a manual for data processing and synthesis.

e) Meeting outputs:

- The proceedings of the meeting as a working paper to be posted on nfa webpage